CORNELL UNIVERSITY LIBRARY Cornell University Library QM 601.M66 Human embryology, 3 1924 003 131 251 Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31 9240031 31 251 HUMAN EMBRYOLOGY BY THE SAME AUTHOR A BIBLIOGRAPHY OF VERTEBRATE EMBRYOLOGY. Containing over three thousand titles classified by subjects and indexed by authors. 4to. Boston, i8g2. Published and for sale by the Boston Society of Natural History HUMAN EMBRYOLOGY BY CHARLES SEDGWICK MINOT Professor of Histology and Human Embryology Harvard Medical School, Boston FOUR HUNDRED AND SIXTY-THREE ILLUSTRATIONS NEW VORK WILLIAM WOOD AND COMPANY 1892 L. O. Dap. Ordar Div. Copyright By CHARLES SEDGWICK MINOT PREFACE. The following attempt to present a comprehensive summary of Embryology, as it bears upon the problems of human development, is the result of ten years' labor. I have endeavored to become famil- iar with the principal facts by my own observation, and with the results of the principal numerous investigations, working over the material into satisfactory form. The reader will find, nevertheless, imperfections of which I am conscious, and perhaps errors, for which I must be responsible. There is probably not a page which might not be enriched with facts already recorded by investigators ; certainly not a page which would not be improved by further revis- ion. Notwithstanding these defects, I have the hope that the book will be a useful contribution toward that final and exhaustive colla- tion of embryological facts which the future alone can give us. I have sought to form an unbiased judgment upon each ques- tion, to accept facts of observation without regard to their supposed theoretical bearings ; and to pay due attention to both Schools of Em- bryology, the Phylogenetic and the Anatomical, in the belief that both are justified. Whenever I have inserted a new observation or opinion, it is indicated as such by the use of the first person. In making my compilation, I have drawn constantly from the embryo- logical manuals of Kolliker, Oskar Hertwig, Balfour and Duval; from the researches of W. His, and from the writings, especially the " Entwickelungsgeschichte der Unke," of Alexander Groette. In regard to the technical terms, I have made certain innovations. It seems to me important to make the number of terms as small as is compatible with clearness, and to avoid duplication. Accordingly I have discarded the words " epiblast, mesoblast, and hypoblast." Further it has seemed to me that, as a thorough knowledge of Ger- man is indispensable to the student of embryology, it is justifiable, Vlll PREFACE. where no English equivalent is to be found, to adopt such imaltered German terms as have been fully established in embryological liter- ature. Where there has occurred an accepted term in English, French, or German, I have used it in preference to a Greek or Latin derivative. Whatever merit this work may possess should be attributed to the training in scientific research which I received in Germany and France. I cannot too gratefully acknowledge the unlimited kindness shown me while a student in Leipzig under Professor Carl Lud- wig and Professor Rudolph Leuckart ; in Paris under Professor Leon Ranvier; and in Wiirzburg under Professor Carl Semper. I would also here express my gratitude to Professor Wilhelm His, to whom I am particularly indebted for his great generosity in permitting me to study his unique embryological collection in Leipzig ; also to the large number of physicians, both in Europe and America, who have sup- plied me with material to carry on my investigations in human em- bryology. Charles Sedgwick Minot. Harvard Medical School, Boston, Mass., July 26, 1893. TABLE OF CONTENTS. CHAPTER PAGE Introddctiobt, ... . . . 1 I. The Uterus, . . . .1 II. General Outline of Human Development, .... 38 PART I. The Genital Products. III. History of the Genoblasts and the Theory of Sex, . 39 PART II. The Gebm-Layebs. IV. Segmentation ; Formation of the Diadei'm, .... 93 V. Concrescence : the Primitive Streak, .... 115 VI. The Mesoderm and the Coelom, 144 VII. General Remarks on the Germ-Layers, 159 PART III. The Embryo. VIII. The Medullary Groove, Notochord and Neurenteric Canals, 173 IX. Divisions of the Ccelom ; Origin of the Mesenchyma, . . 193 X. Origin of the Blood, Blood- Vessels and Heart, . , . 311 XI. Origin of the Urogenital System, 330 XII. The Archenteron and the Gill Clefts, . ... 354 XIII. The Germinal Area, the Embryo and its Appendages, . 371 PART IV. The Foetal Appendages. XIV. The Chorion, ... .... 317 XV. The Amnion and Proamnion, 333 XVI. The Yolk Sack, Allantois and Umbilical Cord, ... 346 XVII. The Placenta, . .... 364 TABLE OP CONTENTS. PART V. The Fcetus. CHAPTER PAGE' XVIII. Growth and External Development of the Embryo and Fcetus, . 381 XIX. The Mesenchymal Tissues, . . .... 397 XX. The Skeleton and Limbs, 422 XXI. The Muscular System, . . . ... 470 XXII. The Splanchnocoele and Diaphragm, ...... 480' XXIII. The Urogenital System, 490 XXIV. Transformations of the Heart and Blood- Vessels, ... 521 XXV. The Epidermal System, 548 XXVI. The Mouth Cavity and Face, 567 XXVII. The JSIervous System, . . ! .... 593 XXVIII. The Sense Organs, 706 XXIX. The Entodermal Canal, 743- LIST OF ILLUSTEATIO^S. »^IG. PAGE 1. Connective tissue of mucosa, uterus of pig, 3 3. Vertical section of the mucosa corpus uteri of the first day of men- struation, .... 5 3. Mucous membrane of a virgin uterus during the first day of men- struation, . . 7 4. Semi-diagrammatic outhne of an antero-posterior section of the gravid uterus and ovum of five -weelvs, 8 5. Uterus about forty days advanced in pregnancy, .... 9 6. Uterus one month pregnant ; outlines of the glands from a vertical section, ... 14 7. Uterus one mouth pregnant ; portion of the compact layer of the decidua seen in vertical section, 1.5 8. Uterus one month pregnant ; section of gland frorn cavernous layer, with the epithelium partly adherent to the walls, ... 16 9. Uterus one month pregnant ; section of gland from cavernous layer with the epithelium loosened from the walls, ... 16 10. Section of the decidua serotina, near the margin of the placenta ; normal uterus about seven months pregnant, .... 17 11. Decidual cells from the section represented in Fig. 10; stained with alum hsemotoxylin, and eosin, 18 13. Section of human decidua reflexa at two months, .... 30 13. Uterus twelve hours after artificial delivery at six months' preg- nancy, 32 14. Section of the placental area of the uterus three weeks post partum, 33 15. Vertical section through the wall of a uterus about seven months pregnant, with the foetal membranes in situ, 16. Human embryo, 4.3 mm. long, 17. Embryo, 3.15 mm. long, .... 18. Diagram of an embryo of fifteen to sixteen days, 19. Generalized diagram of an amniote vertebrate embryo, . 30. Generalized diagram of an amniote vertebrate embryo before the separation of the amnion, 21. Structure of a rat's spermatozoon, ... . . 33. Human spermatozoa, 33. Peripheral layer of the seminiferous tubule of a rat, . 34. Column of spermatocytes from the rat, .... 35. Developing spermatoblasts of the rat, . 36. Developing spermatozoa of a marsupial 37. Human spermatoblasts, to illustrate the rupture of the membrane, 46 38. Sertoli's column, with a basal nucleolated nucleus and a cluster of developing- spermatoblasts, . . . . . 47 39 33 33 34 34 85 41 '43 43 44 45 46 XU LIST OP ILLUSTRATIONS. Me. PAGE 39. Part of a cross-section of a seminiferous tubule of a rat, ... 47 30. Egg of Tendra zostericola, ' 49 31. Primary follicles from the ovary of a woman thirty-one years old, 51 33. Ovary of cat, ... 52 33. Egg-cell of Tengenaria domestioa, . . ... 54 84. Full-grown human ovum before maturation, ... 56 35. Part of the ovum of a mole, 57 36. Ovum of a sea urchin, Toxopneustes lividus, 58 37. Ovarion egg of haemops, 63 38. Egg of a leech (nephelis), three-quarters of an hour after being laid, 63 39. Ovum of nephelis (a leech), three hours after laying 64 40. Ovum of a rabbit ; taken from the middle of the oviduct about eighteen hours after coitus, 71 41. Anterior pole of the ovum of the petromyzon, with a spermatozoon, 72 43. Egg of nephelis, three hours after laying, ... .74 43. Ovum of sagitta with two pronuclei, ... 75 44. Two ova of the land-snail, arion, . ..... 76 45. Ovum of a i-abbit seventeen hours after coitus with the pronuclei about to conjugate, . . . ... .76 46. Ovum of Limax campestris during the first cleavage, ... 96 47. Blastula stage of Echinocardium cordatum twenty hours after impregnation, 97 48. Segmentation of the egg of the common frog, .... 98 49. Section of the segmented ovum of axolotl, ... . 99 50. Four stages of the segmentation of the hen's ovum, . . .100 51. Ovum of a flounder in transverse vertical section, .... 103 52. Ovum of a rabbit of twenty-four hours, . . ... 103 53. Rabbit's ovum of about seventy hours, 104 54. Ovum of a bat, Vgspertilis murina, with four segmentation spheres, 105 55. Ovum of Virginian opossum, with four segments, .... 105 56. Young blastodermic vesicle of a mole, 105 57. Sections through the inner mass of the blastodermic vesicle of the mole, at three successive stages, 106 58. Ovum of a rabbit, ninety-four hours after coitus, .... 107 59. Diagram of a segmented mammalian ovum, 108 60. Ovum of Amphioxus lanceolatus during segmentation stage, with eighty-eight cells, 110 01. Section of a gastrula of Toxopneustes lividus, 113 63. Diagrams of the principal modifications of the gastrula, . . . 114 63. Longitudinal section of an early stage of the gecko, . . . .115 64. Diagram illustrating the growth of the blastoderm and concrescence of its rim to form the primitive axis, . .... 116 65. Diagram of concrescence in a teleostean egg 118 66. Diagram of an elasmobraneh blastoderm to illustrate the formation of the marginal groove, . . 119 67. Diagram of a vertebrate blastoderm a little niore advanced than Fig. 96, 130 68. Ovum of. axolotl, . . 121 69. Ovum of petromyzon in longitudinal section, . . . 121 70. Longitudinal section of the ovum of a sturgeon after the formation of the entodermic cavity, . . . ' 122 71. Formation of the blastoporio canal in Lacerta muralis, . . . 123 LIST OF ILLUSTRATIONS. XIU FIG. PAGE 73. Hen's ovum ; incubated six, hours, . 134 73. Diagrammatic cross-section of a vertebrate ovum, in which con- crescence, is supposed to have been arrested, .... 136- 74. Dog-fish embryo, nearly in Balfour's stage C, . ... 136 75. Germinal area of a guinea-pig at thirteen days and twenty hours, 137 76. Diagram showing the relations of a vertebrate ovum with an em- bryo, in cross-section and a large yolk, . . . 138 77. Sections of axolotl eggs, 130 78. Area pellucida of a hen's egg, with completed primitive furrow, . 131 79. Longitudinal section of the region of the primitive streak of a hen's ovum incubated six hours, 133 80. Transverse sections of a germinatlve area, with half -formed primi- tive streak, of a hen's egg, . . 133 81. Transverse section of the anterior region of a fully developed primi- tive streak of a hen's ovum, 135 83. Blastodermic vesicle of a rabbit of seven days, ... . 136 83. Transverse section of the embryonic shield of the blastodermic vesicle of a sheep, 137 84. Central portion of a sheep's blastodermic vesicle of twelve to thir- teen days, . . 138 85. Embryonic shield, of a rabbit's ovum of five days, . . 139' 86. Section of the primitive streak of the mole, . . 140 87. Blastodermic vesicle of Mus sylvaticus, . . . 143 88. Axolotl embryo ; transverse section of an early stage, . 146 89. Diagrams of the embryonic area of the chick, . . . 150 90. Diagram of the embryonic area of a chick, 150 91. The mesodermal cavities of the germinal area of a chick of the third day, ■ 15^ 93. Section of a chicken embryo of about thirty-six hours, . 153 93. Transverse section of an amphioxus embryo, 156 94. Amphioxus embryo, . . IS''' 95. Opossum embryo of seventy-three hours ; transverse section at the level of the heart, I'J'S^ 96. Blastoderm of rabbit's ovum, 174 97. Chicken embryo with seven primitive segments, ... 175 98. Part of a transverse section of a young mole embryo, . . .176 99. Surface view of a young mole embryo, 176 100. Transverse section of a mole embryo, . . . . . 176 101. Early stage of Amblystoma punotatum, . . 177 103. Part of a transverse section of an axolotl embryo, . . .178 103. Transverse section of a rabbit embryo of eight days and two hours, 179 104. Part of a transverse section of an embryo of Lumbricus trapezoides, 180 105. Transverse section of a mole embryo, 183 106. Longitudinal section of the head end of a mole embryo, . . .188 107. Rabbit embryo of 6 mm. ; median longitudinal section of the head, 185 108. Longitudinal sections of the notochord of bombinator, . . .186 109. Degenerairing notochord tissue, from the central portion of the in- tervertebral disc of a cow's embryo, 187 110. Longitudinal section of a frog's ovum, shortly after closure of the medullary groove, .... 188 111. Transverse section of an embryo paroquet (melopsittacus) to show the anterior or true neurenteric canal, . . . 189 XIV LIST OF ILLUSTRATIONS. FIG PAGE 113. Chicken embryo with one segment, 193 113. Area vasculosa and embryo with eight segments of a hen's egg, . 193 114. Rabbit embryo with eight segments 194 115. Transverse section of a pristiurus embryo with fourteen segments, through the centre of the fourth segment 194 116. Transverse section through a recently formed primitive segment of a chick with eighteen and twenty segments, .... 196 117. Section of a chick with about twenty segments, . . . .198 118. Head of an embryo of Torpedo ocellata, in Balfour's stage J, . . 300 119. Longitudinal vertical section through five primitive segments of a rabbit embryo of nine days and seventeen hours, . . . 304 120. Longitudinal horizontal section through a segment of a rabbit embryo of ten and one-half days, 205 131- Transverse section through the upper part of a myotome of a chick of about seventy hours, 306 123. Pristiurus embryo with forty-five to forty -six segments, . . . 208 133. Diagram of a cross-section of a young amphioxus, .... 209 134. Surface view of a small part of the vascular network of an embryo chick of two days, 312 125. Vascular anlages of the area vasculosa of a chick of forty hours, . 318 136. Section of the area vasculosa of a chick, 314 137. Corpuscles from rabbits, from acanthias, from a chick, from a hu- man embryo, 319 138. Salamandra maculosa ; larva, very young ; transverse section to show the formation of the coelom in the heart region, . . 334 139. Salamandra maculosa, larva with branchial arches, .... 325 139A. Embryo chick ; section through the anlage of heart, . . . 336 130. Chick embryo, 337 131. Diagrammatic cross-section of a vertebrate to show the fundamen- tal relations of the urogenital system, 230 133. Rana temporaria. Tadpole of 13 mm. Cross-section through the pronephros, 231 133. Nephridium (or Wolffian tubule) of an acanthias embryo of 38.3 mm., seen from the caudal side ; reconstructed from the sections, 336 134. Section through a Wolffian tubule of a chick with primitive segments, . . . . 238 135. Wolffian tubule of a sheep embryo of 9 mm., 238 136. Coste's embryo of thirty-five days, 240 137. Transverse section of the Wolffian body or primitive kidney of a rabbit of thirteen days, 241 138. Longitudinal vertical section of the Wolffian body of a rabbit em- bryo of thirteen days, 242 139. Section through the testis of a human embryo of sixty-three to sixty- eight days 244 140. Transverse section through an advanced embryo of a shark, sycm- nus lichia ; from the abdominal region (dots represent nuclei), . 347 141. Section of the urogenital fold of a chick embryo of the fourth day, 248 142. Diagrammatic section of the yellow of a hen's egg at an early stage to show the relations of the arohenteron to the yolk-sac, . . 355 143. Diagrams to illustrate the separation of the embryo from the yolk, 256 144. Cross-section of a rabbit embryo of eight days and two hours, . 257 LIST OF ILLUSTRATIONS. XV FIG. _ PAGE 145. Longitudinal section of the posterior end of a sheep embryo of six- teen days, 358 146. Longitudinal median section of young chick embryo, . . 261 147. Transverse isection of the head of a chick embryo with seven segments, 363 148. Two views of a wax model of the cavity of' the pharynx of a rabbit embryo of eleven days, ... . ... 264 149. Aeanthias embryo of 17 mm. Horizontal section of the anterior half, 366 150. Chicken embryo of sixty-eight hours, 366 151. Aeanthias of 17 mm., .... . 267 153. Cross-section of a branchial arch of an advanced shark embryo, . 367 153. Longitudinal section of an embryo of Petromyzon planeri, four days old, reared at Naples, 268 154. Diagrams to indicate the fundamental relations of the archenteron, 270 155. Chicken embryo and germ area after twenty-seven hours' incubation, 273 156. Embryonic area of a rabbit of eleven days, with the placental area partly torn off 373 157. Diagram of the circulation in a chick at the end of the third day, as seen from the under or ventral side, ...... 374 158. Area vasculosa and embryo of a rabbit, . . ... 375 159. Transverse section of the rump of a dog-fish embryo 14 mm. long, . 380 160. Section through the rump of a rabbit embryo of eight days and three hours, 283 161. Transverse section of the rump of an embryo chick of the third day, 283 163. Diagrams to illustrate His' theory of the origin of the human amnion, 385 163. Keichert's ovum. Two views engraved from the original plate, . 388 164. Cross-section of Spee's embryo . . 393 165. Section passing through the blastopore of Spee's embryo, . 293 166. Diagram of His' embryo E : age fourteen (?) days; length about 3.3 mm., 393 167. Thomson's second ovum, .... ■ 294 168. Human embryo of thirteen to fourteen days, . . .395 169. Embryo of the beginning of third week, . . ... 396 170. Human embryoof 3.15mm.; anatomyreconstructedfromthesections 397 171. His' embryo L, 3.4 mm. long, . 399 173. Ovum supposed to be from fifteen to eighteen days old, . . 301 178. Embryo supposed to be from fifteen to eighteen days old, . 303 174. Fragment of the chorion of fig. 4, highly magnified, . . 303 175. His' embryo M, • • ^04 176. Digestive canal of His' embryo, . ■ • 305 177. Anterior wall of the pharynx of His' embryo BB, 3.8 mm. long, . 305 178. W. His' embryo M, . . . 306 179. Reconstruction of His' embryo BB, 3.3 mm. long, . . 306 180. Reconstruction of His' embryo, 307 181. Isolated terminal branch of a villus from the chorion of an embryo of twelve weeks, ^^^ 183. Villous stem from a placenta of the fifth month, . . . .330 183. Terminal villi of a placenta at full term, 330 184. Section of the chorion at three weeks, 331 185. Aborting villus froui a chorion of the second month, . . .331 186. Placental chorion of an embryo of seven months, . . . .334 XVI LIST OF ILLUSTRATIONS. FIG. PAGET 187. Section of the chorionic membrane of an ovum supposed to belong to the third week, 328 188. Section of the chorionic membrane of an embryo of three weeks, . 339 189. Section of the amnion and placental chorion of the fifth month, . 329 190. Adenoid tissue of a villus from a placenta of four months, . . 380 191. Section of the amnion covering the placenta of a two months em- bryo, 334 193. Two sections of the placental amnion, . 334 193. A natural group of nuclei from the mesoderm of the amnion of a foetus of the fifth month, 335 194. Mesodermic nuclei of the amnion of an embryo of about four months, 835 195. Surface view of the amniotic epithelium of an embryo of 144 days, 336 196. Diagram of the development of the foetal adnexa in the rabbit, . 343 197. Longitudinal median section of a jjetromyzon larva, . . . 346 198. "Wall of the yolk-sac in the area opaca of a chick of the second day, 347 199. Section of the yolk-sac of a human embryo, . . . . . 350 200. Diagram of the embryo and yolk-sac of a rabbit, .... 351 301. Vertical section of the wall of the yolk-sac of a rabbit embryo of thirteen days, 351 303. Diagram of an opossum embryo and its appendages, .... 351 303. Section of the allantois from the umbilical cord of an embryo of three months, 355 204. Diagrammatic, section of the bauchstiel of a human embryo, modi- fled from W. His, 356 305. Sections of human umbilical cords, 357 206. Connective tissue of the umbilical cord of an embryo of 31 mm., . 358 307. Connective tissue of the umbilical cord of a human embryo of about three months, 359 308. Epithelial covering of the umbilical cord of an embryo of three months, . . 360 309. Cross-section of an umbilical cord at term, ... . 363 310. Placenta at full term, doubly injected by Dr. H. P. Quincy to show the distribution of the vessels upon the surface, . . . 365 311. Placenta at lull term, 367 313. Mesenchymal tissue of a villus, from a placenta of four months, . 368 213. Section through a normal placenta of about seven months, in situ, 370 314. Portion of an injected villus from a placenta of about five months, 371 215. Placenta of about five months ; portion of a small villus, . . 371 316. His' embryo a, age probably twenty-three days, . . . 385 317. Fol's embryo of 5.6 mm., probably twenty-five days old, 386 318. His' Embryo A, 7.5 mm. long, . ... 387 319. Embryo of 9.8 mm., . 388 330. Embryo of about 14 mm., . .389 331. Dorsal view of an embryo of about 14 mm., . . . 389 333. Embryo of about thirty-five days, . . . 390 323. His' embryo XXXIV, ... . . 891 234. Embryo of 22 mm., . 392 235. Embryo of 38 mm., ... . 393 336. Embryo of 32 mm., ... . . .393 227. Embryo of 34 mm., ... .393 238. Embryo of 55 mm,, . . .... . 393 339. Embryo of 78 mm., . . . 394 LIST OF ILLUSTRATIONS. XVU PIG. PAGE 230. Front view of the head and face of the embryo, 394 231. Embryo of about 120 mm., . . . 394 232. Embryo of 118 mm., ... .395 233. Embryo of 155 mm., 396 234. Mesenchyma of a chick embryo of the third day from close to the otocyst, . ... 899 335. Omentum of a human embryo of five montlis, . . 400 236. Parietal bone of a human embryo of fourteen weeks, . . . 408 237. Transverse section of the mandible of a human embryo of the tenth week, 408 238. From a section of an ossifying vertebra of a human embryo of four months, .... 411 239. Section of a vertebra of the same embryo at right angles to the plane of fig. 238, and corresponding in level to the lower part of the bracket L, tig. 238, . .... . . 412 240. Artery from the allautois of a chick, surrounded by a network of lymphatics, ... 414 241. Section of the spleen of a human embryo of six months, . . . 416 242. Fat island from the skin of a human embryo of five months, . . 418 243. Reconstruction of the last occijiital, and first two cervical vertebra of a cow embryo of 8.8 mm., 425 244. Cross-section of the anlage of second cervical vertebra of a cow embryo of 8.8 mm., . 425 245. Longitudinal median section of the upper portion of the vertebral column of a cow embryo of 22.5 mm., 4S7 246. Frontal projection of the cephalic part of a vertebral column of a cow embryo, 4-31 247. Embryo pig of about 16 mm., ... 435 248. Embryo pig, one and one-third inch long, . . ... 436 249. Section of the anterior jxsrtion of the snout of an embryo pig, . 437 350. Embryo pig, six inches long, . . 439 251. Chondrocranium of an insectivorous mammal (Tatusia), . . . 442 353. Pectoral fin of a young embryo of sycUium in longitudinal and horizontal section, . 449 253. Scapula of a human embryo of five and one-half inches, dorsal view, 453 254. Vertical section of the ankle of a human embryo of nearly six months, 459 355. Isolated muscle fibres of a frog embryo, .... .471 256. A, transverse section ; B, longitudinal section of muscle fibres in the neck of a human embryo of sixty-three to sixty-eight days, . 473 357. Chick embryo, transverse section of the upper part of a myotome, 476 258. Transverse section of a branchial arch of a selachian embryo, . . 478 259. His' embryo R, 5 mm. Reconstruction to show the septum trans- versum, 360. Head of a rabbit embryo, with segments seen from the under side, 481 261. Rabbit embryo, eight and a half days, with eleven or twelve somites, cross-section, • 483 263. Model of part of the pleural and abdominal cavities of a rat embryo at a stage corresponding to a rabbit at fifteen days, . . .484 363. Section of the supra-renal body of a rabbit embryo of twenty-six days, . '^^'^ 364. Supra-renal capsule of a four months' human embryo, . . .488 265. Diagram of the indifferent stage of the urogenital system of amniota, 490 480 XVlll LIST OF ILLUSTRATIONS. FIG. PAGE 266. Diagram to illustrate the homologies of the sexual apparatus, . 491 267. Section of the testis of a human embryo of sixty-three to sixty-eight days, 493 268. Section of the ovary of a human embryo of 7 cm., 495 269. To illustrate the decensus testiculorum, 498 270. Cross-section of the ovary and WolflQan body of a human embryo of the third month, . . ... . . 499 371. Cross-section of the rectum, genital cord, and allantois of a male human embryo of about two months 502 272. Section of broad ligament of a female human embryo of four months, 504 273. Cross-section through the hind end of the left Wolffian body of a crocodile embryo of 12 mm., . . . . ,. . . 508 274. Section of a kidney, human embryo of about five months, . . 509 275. Semidiagrammatic figures of developing renal tubules of a mammal, 510 376. Section parallel to the medullary rays of the kidney of a human fcetus of about five months, . 511 277. Cross-section of the medullary tubules of the kidney of a human embryo of about five months, 513 278. Longitudinal median section of the cloaca of a sheep embryo of 18 mm., 517 379. Longitudinal section of the penis of a human embryo of about five months, . . 518 280. External genitalia, female embryo of 105 mm., 518 281. Section of the clitoris and labia majora of a human embryo of about four and one-half months, 519 282. External genitalia of the female human foetus at about four months, 519 283. Head of chick of thirty-eight hours, seen from the under side, . 522 384. Reconstruction of the heart and veins of a human embryo of 3.15 mm., 533 285. Endothelial heart of a human embryo of 3.15 mm.; seen from the left side, 533 286. Reconstructed side view of the endothelial heart of a human em- bryo of 4.3 mm., 528 287. Model of the muscular heart of a rabbit embryo of nine to nine and one-half days, seen from the left side, 533 388. Endothelial heart of a human embryo of 5 mm., .... 524 289. Inner surface of the heart of a human embryo of 10 mm., . . 524 390. Section of the heart and pericardial cavity of a rabbit embryo of ten and one-half days, 536 391. Section in the frontal plane through the heart of a rabbit embryo of thirteen days, 529 292. Oblique section of the heart of a human embryo of 8.5 mm., . . 530 393. Sections at different levels through the cardiac aorta of a human embryo of 11.5 mm., 531 294. A diagram of pharynx of an amniote vertebrate, .... 535 29.5. Anterior wall of the pharynx of a human embryo of 3.3 mm. length, 536 296. Aortic system of His' embryo BL, 4.35 mm., 536 237. Aortic system of His' embryo Si, 13.5 mm.; seen from the front, . 537 ^Si8. Aortic system of W. His' embryo Rg, 11.5 mm., 539 399. Reconstruction of the arteries of the head and neck of a rabbit em- bryo at the end of the eleventh day, 540 300. His' embryo Lr (4.3 mm.). Reconstruction to show the course of the blood-vessels, 543 LIST OF ILLUSTRATIONS. xix FIO. PAGE 301. Oross-seetjon through the hinder part of His' embryo R (5 mm.), . 543 303. Three diagrams to illustrate the transformation of the venous system, 543 303. Reoonstruotion of a human embryo (His' Bl.) of 4.25 mm., . 545 304. Reoonstruotion of the venous trunks and liver of His' embryo R, 5 mm., . . .546 305. Reoonstruotion of the venous system of His' embryo Rg, 11.5 mm., 547 306. Section of the skin of a human embryo of sixty-three to sixty-eight days, 548 307. Epidermis from the ocoiput of the liuman embryo of two and one- half months, . 550 308. Section of the skin of the under side of the right second toe of four months' embryo, 551 309. Epitriohium of a human embryo of the fifth month, .... 552 310. Vertical section of the skin of a human embryo of the fifth month, 554 311. Longitudinal section of the nail of the great toe of a human embryo of five months, . ... ... . 556 813. Development of hairs in a liuman embryo of about seven months, 558 318. Isolated epidermis of a human embryo of five to six months, . 559 314. Section of the sole of the foot of a foetus of the fifth month, to show the sweat glands, .... ... 563 315. Development of the mammary gland in the rabbit, .... 564 316. Acanthias embryo of 17 mm., under side, .... . 569 317. Blastoderm of a dog-fish, acanthias, with commencing t3onorescenoe, 570 318. Longitudinal median section of a recently hatched larva of petro- myzon, 573 319. Longitudinal section of an acanthias embryo of 18.3 mm., . . 573 320. Median section of the head of a rabbit embryo of thirteen and one- half days, ... . . 573 321. His' embryo A, 7.5 mm., 576 333. Facial region of a human embryo of 8 mm., front view, . . 576 338. Reconstruction of the face of His' embryo Sch, .... 577 334. View of the roof of the mouth of a human embryo, . . . .578 325. Frontal section of the oral and nasal chambers of a young cow embryo, 579 336. Frontal section of the nasal and oral cavities of a human embryo of three months, 581 337. Dental papilla of a dermal tooth of an acanthias embryo of 10 cm., 583 328. Section of the lower jaw of an acanthias embryo of 10 cm., . . 583 339. Section of part of the lower jaw of a human embryo of 40 mm., . 583 330. Explanation in text, ... 584 331. Vertical section of a molar tooth-germ of a human embryo of 160 mm. , 585 332. Part of the enamel organ of a new-born child, incisor germ, . 586 333. Odontoblasts from cow embryos. A, of 30 cm.; B, of 34 cm., . . 588 334. Section of the submaxillary gland of gi human embryo of sixty- three to sixty-eight days, . . 591 335. Reoonstruotion of the pharynx of a human embryo, . . . 592 386. Chick embryo of twenty-nine hours, . . .... 594 337. Cross-section through the fore-brain and optic vesicles of a lepidos- • teus embryo of eight days, 595 338. Brain of embryo No. 23, p. 397 596 339. Reconstruction of the brain of His' embryo Ko 597 840. Reconstructed median view of the fore-brain of His' embryo Ko, . 597 XX LIST OF ILLUSTRATIONS. FIG. Fi.aK 341. Brain of a human embryo of five weeks, 599 343. Hind-brain of a human embryo, 600 343. Dorsal view of the hind-brain of a human embryo of one month, . 600 344. Sections through the cervical part of the medulla of a human embryo with thirteen segments, 602 345. Longitudinal horizontal section of the wall of the hind-brain of a young embryo of a lizard (Anolis Sagrsei), 605 346. Diagrammatic section of the embryonic spinal cord, .... 607 347. Section of the medulla and otocysts, 608 348. Sections through the regions 3 and 5 of the hind-brain of His' embryo, 608 349. Sections through the region 3 of the hind-brain of His' embryo A, 609 350. Four sections of the brain of a human embryo of about five "weeks, 609 351. Brain of His' embryo Br. 3, 610 353. Neuroglia of the dorsal zone of the spinal cord of a human embryo of about three and one-half weeks, 613 353. Cross-section of the spinal cord of a human embryo of 14 mm., to show the neuroglia cells, 614 354. Part of a transverse section of the spinal cord of a human embryo of 33 cm., 615 355. From a section of the medulla oblongata of His' embryo Br'., . 617 356. Group of motor neuroblasts and nerve fibres from a transverse sec- tion of the spinal cord of a cat embryo of 6 mm., . . . 618 857. Bipolar cells from a spinal ganglion of an embryo, .... 619 358. Transverse section of the dorsal cord and ganglion of a chick of nine days, ... 630 359. Isolated nerve fibres from the sciatic nerve of a sheep embryo of 150 mm., 631 360. Part of the nerves of a human embryo of 13.8 mm., .... 623 361. Cells and nuclei from the cervical region of the spinal cord of a human embryo of one hundred and sixty days, .... 634 363. Spinal ganglion cells from a longitudinal horizontal section of » human embryo of the tenth week, 626 363. Peripheral nervous system of a human embryo of about 10 mm., . 638 864. Transverse section of a mouse embryo of about seventeen to eighteen days through the lumbar region, 629 365. Transverse section of the sympathetic cord from the lower dorsal region of a rat embryo of about thirteen days, .... 631 366. Sympathetic ganglia of one side of a human embryo of the fifth month, . 632 367. Transverse section through the posterior part of the mid-brain of a human embryo of five weelis, 639 368. Section of the brain of a five weeks' embryo, . ... 640 869. Section of the brain of a five weeks' embryo, 643 370. Section of the brain of a human embryo of five weeks, . . . 645 371. Otocysis and nerves of a human embryo of four and one-half weeks, 647 873 Acoustic ganglia of a human embryo of two months, . . . 647 373. Torpedo embryo of 13 mm., 653 374. Section of the medulla oblongata of a five weeks' human embryo, t;55 375. Lower end of tlie spinal cord of a human embryo of three months, 658 376. Section of the spinal cord of a human embryo of sixty-three to sixty-eight days, 660 LIST OF ILLUSTRATIONS. XXI JPIO- PAGE 377. Transverse section of the spinal cord from the upper dorsal region of a human embryo of six weelss, 661 378. Lower cervical cord of a human embrj'o of about five months, . 663 379. I'ransverse section of the medulla oblongata of His' embryo Ru, . 666 380. Transverse section of the medulla oblongata of His' embryo Mr, . 667 381. Section through the medulla oblongata of His' embryo CR, . 669 383. Median section of the brain of a chick embryo of about four days, 673 383. Longitudinal median section of the cerebellum of a chick of about twelve days 673 384. Section through the cerebellum and medulla oblongata of a human embryo of one hundred and sixty days, 674 385. Section of the cerebellum of a human embryo of one hundred and sixty days, ... . . ... 386. Median section of the head of a sheep embryo of .36 mm., S87. Brain, human foetus, five months, 388. Part of the brain of His' embryo CR, 13.6 mm., . 389. Section of the thalamencephalon of an embryo of five weeks, 390. Section of the fore-brain of a sheep embryo of 27 mm., 391. Brain of a human embryo of about three months, 393. Brain of a human embryo of the fourth month, . 393. Median view of a frog's brain, 394. Section through the fore-brain of a foetal guinea-pig, 395. Reconstruction of the brain of an embryo of about seven and half weeks, .... 396. Brain of a chick embryo, fourth day, 397. Human embryo of about four months ; brain in situ, 398. Section through the lateral wall of the cerebral hemisphere of a human embryo of four months, 694 399. View of the hemisphere of a human embryo from the early part of the third month, 695 400. Outlines of the fissure of Sylvius of human embryos at successive lunar months, . . 696 401. Median view of the fore-brain of a human embryo from the begin- ning of the third month, 403. Brain of human embryo of the fifth month after rejuoval of the right hemisphere, . . . ■ ... 698 403. Right hemisphere, natural size of a foetus of nearly seven months, 699 404. Under side of the brain of a human eiubryo of the fifth month, 700 405. Section of the fore-brain of a human embryo of nearly five weeks, 704 406. Horizontal section of the ciliary ganglion of a young torpedo em bry o, 707 407. Reconstruction to show the cephalic ganglia of a petromyzon larva '^07 4 mm. long, .... ... . . i\ji 408. Rabbit embryo of ten and one-half days ; section of head, . 711 409. Rabbit embryo of thirteen days ; section of the eye, . . . 713 410. Reconstruction from His' embryo Sch, 13.8 mm., ... 713 411. Section through the iris region of the eye of a chick of thirteen days, 714 4] 3 Rabbit embryo of ten and one-half days ; section of the lens anlage, 71.", 41;! Vertical section of the eye of a chick embryo of the third day, . 715 414. Section of the distal portion of the optic nerve of a rabbit embryo of thirteen days, 41,5 Sm-face view of the membrana limitans externa with the develop- inff rods and cones of a chick of fifteen to sixteen days, . . 720 675 678 678 679 681 681 683 683 684 685 686 688 691 096 18 X^li LIST OF ILLUSTRATIONS. PIG. PAGE 416. Injected vascular membrane of the retina of the eye of a pig embryo, 16 cm. long, ^31 417. Section through the iris region of the eye of a chick of thirteen days, . . . 735 418. United eyelids of a human embryo of about four months, seen in vertical section, . 736 419. Sections of human embryos showing the otocyst ; A, embryo of 3.4 mm.; B, embryo of 4 mm., 738 430. Horizontal section of the otocyst of a chick of the third day, . . 72& 431. Left otocyst of a human embryo of about four weeks ; A, from the inner, B, from the outer side, 739 433. Transverse section of the head of a rabbit embryo of ten and one- half days, . . 730 433. Left otocyst of a human embryo of about five weeks, seen from out- side and below, 730 434. Transverse section of the semicircular canal of an embryo rabbit of twenty-four days, 731 435. Left otocyst of a human embryo of about two months, . . . 733 436. Transverse section of scala media cochleae of a rabbit embryo of 55 mm , . . 783 437. Section through Corti's organ of the lower coil of the cochlea of a rabbit embryo of 75 mm., . . 734 438. Section through the internal ear of a sheep embryo, 28 mm., . . 736 429. Isolated right membranous labyrinth of human embryo of six months, seen from in front and outside, 737 430. Section through the region of the ear of a human embryo of three months 739 431. Development of the human external ear ; A, embryo of one month ; B, six weeks ; C, eight weeks ; D, ten weeks; E, fourteen weeks, 741 433. Keconstruction of the pharyngeal region of a human embryo of 11.5 mm., 744 433. From a section of a tonsil of a human embryo of five months, . . 745 434. Section through the third gill-cleft of a human embryo from the beginning of the third week, 746 435. Reconstruction of the pharyngeal region of a human embryo of 9.1 mm., 748 436. Reconstructions to show the development of the thyroid gland in the pig; A, embryo of 15 mm.; B, of 16 mm.; C, of 20 mm.; D, of 33.5 mm., 749 487. A, section of the thyroid gland of a human embryo of about four months ; B, a single acinus, more highly magnified, . . . 751 438. Reconstruction of His' embryo B ; the head is drawn as if erected, 753 439. Transverse section of the oesophagus of a human embryo ot four months, . 753 440. Highly magnified view of a small portion of the epithelium of fig. 439, 758 441. Reconstruction of FoPs embryo, 753 443. Epithelium of the greater curvature of the stomach of an embryo cat of 85 mm., 7,54 448. Peptic glands from the greater curvature of stomach of a human embryo from the end of the eighth lunar month, . . . 754 444. Digestive tracts of four human embryos. A, embryo of 4.3 mm. ■ B embryoof 7mm.; C, embryo of 13.8 mm.; D, embryo of 13.5 mm., 756 LIST OF ILLUSTRATIONS. XXlll I'IG. PAGE 445. Two front views of the entodermal canal. A, embryo Sch.. 1 of His'; B, His' embryo Sch. 2, 757 446. Part of the intestine of a human embryo of about six months, . 758 447. Section of the small intestine of a human embryo of sixty-three to sixty -eight days, 759 448. Section of the small intestine of a human embryo of three months, 759 449. Portion of a section of the liver of an aoanthias embryo of 29 mm., 761 450. Section through the liver of a rabbit embryo of thirteen days, 763 451. Section of a rabbit embryo of thirteen days through the region of the fore limbs and liver, .... . . . 764 452. Section of the pancreas of a human embryo of four months, . . 767 453. Two diagrams to illustrate morphological relations of the vertebrate mesentery ; A, earlier ; B, later condition, . . . 768 454. Diagram to illustrate the relations of the mesentery, . . 768 455. Diagram of the human mesentery in its primitive relations, . 769 456. Diagi'ams to illustrate the history of the human mesentery. A, earlier ; B, later condition, 770 457. Two diagrams to illustrate the history of the mesentery; A, earlier; B, later stages, .... . . . . 1^1' 1 458. Outline of the entodermal canal of His' embryo Lr., . . .773 459. Three views of the lungs of a human embryo of 10.5 mm., . 77.j 460. Lungs of a human embryo of five months, 776 461. Cross-section of the bronchial tube of a human embryo of sixty- three to sixty-eight days, . ... . 777 462. Section through the lung of a human embryo of the fourth month, 777 463. Epithelium and gland of the trachea of a four months' embryo, . 778 HUMAlSr EMBRYOLOGY. INTRODUCTION. CHAPTER I. THE UTERUS. The uterus enters in the mammalia into such intimate relations with the embryo, that a thorough knowledge of its structure is njec- cessary to the embryologist. The treatment of the uterus iri the text- books of human anatomy is usually too brief for the requirements of embryology. These considerations make it desirable to give a some- what detailed account of the human uterus. The uterus is the most variable organ within normal limits of the body, both as to size and structure. The virgin uterus is about three inches long and two inches wide at the upper part, where it is broadest ; it weighs about 40 grammes. At the end of pregnancy it is about ten inches long and nine wide, and weighs about 1,000 grammes. The walls of the virgin or resting uterus are tense and mainly muscular ; those of the pregnant organ are more spongy in texture and extremely, vascular, yet at the same time the muscular layers are greatly increased, though relatively less than the vascular layer. After a pregnancy the uterus never returns to its primitive condition, and its weight does not fall below two or three ounces ; from the gradual effects of advanced age, however, and independent of pregnancy, the uterus shrinks, becomes paler in color, and harder in texture. Finally at each successive recurrence of menstruation a complete removal of the superficial part of the mucous membrane takes place by a process, which we can describe but not explain. The removal is said to commence close to the cervix or at the os internum, and to progress toward the fundus during the remaining days of the flow of blood. As the shape and topographical relations are sufificiently described in the standard Anatomies, we confine our- selves principally to the histology. The descriptions are arranged in the following order : 1. Muscularis. 2. Mucosa corpus uteri. 3. Mucosa cervicis. 4. Blood-vessels. 5. Lymphatics. 1 . Muscular Coat. —The volume of the muscularis varies greatly with the condition of the uterus, for during pregnancy the muscles 1 ^ INTRODUCTION. undergo a progressive hypertrophy, which is so great that not only is there an enormous expansion corresponding to the dilatation of the uterus, but also a great thickening of the coat. The increase in volume is due — 1, to the growth of the single fibres (in length from 44-68 [!■ to 330-560 ij) ; 3, it is said also by the development of new muscle cells from small granular cells. After parturition the fibres in part return to their original size, in part undergo fatty degenera- tion (KoUiker, "Gewebelehre," 1867, p. 566). The disposition of the fibres is most readily elucidated in uteri near the end of gestation. Having made no original observations on this subject, I transcribe the following passage from Quain's " Anatomy " : " The external layer of th^ muscular coat forms a thin superficial sheet immediately beneath the peritoneum, and incomplete strata situated more deeply. A large share of these fibres, beginning as longitudinal bands at the cervix, arch transversely and obliquely over the fundus and adjoining part of the body of the organ, and pass on each side into the broad ligament. Of these some at either side run toward the commencement of the round ligaments, along which they are in part prolonged to the groin ; others pass off to the Fallopian tubes, and strong transverse bands from the anterior and posterior surfaces are extended into the ovarian ligaments. Other fibres run back from the cervix uteri beneath the recto-uterine folds of the peritoneum. The inner layer of the muscular coat, which is also thin, is composed of fibres which are found chiefiy on the back of the uterus, and stretch over the fundus and toward the sides, running somewhat irregularly between the ramifications of the blood- vessels. " On the inner boundary the mucosa is quite sharply set off from the muscularis ; an erroneous contrary statement is frequent in English and American works. It is commonly asserted that the muscular coat of the uterus is largely made up of the hypertrophied muscularis mucosce. The evidence for this view is not to be found either in the anatomy or in the developmental history of the uterus, but, so far as I can ascer- tain, solely in the preconception that every mucosa must have a special muscularis to itself, as is the case in the intestine, for exam- ple. Comparative anatomy, however, is conclusive on this point; for it is not rare to find a mucosa without the special muscle layer. The true morphological relations are probably the reverse of those which have been assumed by the view here criticised ; the primitive form is probably a mucosa composed of epithelium and sub-epithelial connective tissue resting on a muscular layer, as in the uterus ; the secondary form, that in which other muscular fibres have been dif- ferentiated to form a special layer, the muscularis mucosae. The muscle fibres have been shown by Elischer, 76.1, to differ somewhat from the forms known in other organs. They are elongated cells, often spindle-shaped, but frequently broad and stumpy ; in the pregnant uterus they are enlarged and flattened ; in length they increase from 40-60 /^- (virgin uterus) to 300-600 fi (uterus at term) ; in transverse section they are seen to be more or less distinctly polyhedral ; their ends and sometimes their sides bear branching processes ; they have one, sometimes two, or even more nuclei, which are usually oval, sometimes round, and usually nucleo- THE UTERUS. lated; the nucleolus is eccentric. The nucleus is surrounded by granular matter, which stretches out toward each end of the ceU ; often the granules are separated by a clear space from the nucleus. This space has been observed by various authors. Eimer has found it in several sorts of cells and gives it the name of hyaloid. It is a peculiarity of the uterus that its muscle cells vary greatly among themselves in appearance. S. Mucosa Corporis Uteri. (A). Virginalis.— At birth the mucosa of the body of the utei'us is about 0.3 mm. thick, soft, pale gray or reddish-gray; it consists of a covering ciliated cylinder epithelium and a connective-tissue layer ; it is without glands, the glands not appearing usually until the third or fourth year, and de- veloping very slowly up to the age of pubertj^. Wyder, 78.1, has shown that the time of the appearance of the glands is extremely variable. In the virgin resting uterus after puberty the mucosa is about 1 mm. in thickness. It is sharply marked off from the muscularis. The glands are tubular, often bifurcated in their lower third, round or oval in transverse sec- tion; they run more or less perpendicularly to the surface of the membrane, upon which they open; yet, strictly speaking, this is true of the glands in their upper half only, and even in that part their course is not straight but wavy . In their lower half they deviate much more, being more irregular and tortuous, the fundus curved sometimes even so much as to run parallel ' to the muscular layer (G. J . Engelmann, 75.1). These differences between the upper and lower parts of the glands are accentu- ated during menstruation and gravidity. The glands are invaginations of the uterine epithelium, are accordingly lined by ciliated cylinder cells, and have a nucleated basement membrane (Fig. 1, d), formed by a layer of anastomosing connective-tissue cells (Leopold, 74. 1) . Overlach, 85.1, however, expressly denies the existence of any such membrane in the human uterus examined by him. The glands reach to, and may even slightly penetrate, the muscularis. Between the glands is found a somewhat embryonic connective tissue consisting of elongated cells with oval nuclei and branching — ™»ift- FiG. ].— Connective tissue of mucosa, uterus of pig; a a, capillaries; b b, sheath of the same; c, uterine gland; d» gland-sheath. After Leopold. 4 INTRODUCTION. processes, which anastomose with one another * (Fig. 1) ; the spaces of the cellular network communicate, according to Leopold, l.s.c, with the Ijrmphatic vessels of the muscularis and external serosa, and may therefore be regarded as lymph roots or lymph spaces. The branching spindle-ceUs resemble somewhat those found in the um- bilical and other embryonic structures, and known under the name of mucous tissue. They tend to crowd together around the blood- vessels and glands. There do not appear to be any fibres in this layer, although some observers have so stated. Between the spindle-cells are small, round cells, probably wander- ing cells (leucocytes) , which vary greatly in number. The blood-vessels enter as veins and arteries from the muscularis, and take a winding course toward the surface ; the capillaries form a network around the glands and under the siirface of the mucosa. (B). Decidua Menstrualis. — The function of menstruation in- volves great changes in the mucosa of the body of the uterus. We distinguish three periods : 1, tumefaction of the mucosa, with accom- panying structural changes, taking 5 days, or, according to Heusen, 10 days ; 2, menstruation proper, about 4 days ; 3, restoration of the resting mucosa, about 7 days. The times given are approximative only. The whole cycle of changes covers about 16 days; as the monthly period is about four weeks, the period of rest as thus calcu- lated is onty about 13 days. 1. Tumefaction. — A few days before the menstrual flow the mucosa gradually thickens; the surface becomes irregular; the openings of the glands lie in depressions. The connective-tissue cells are increased in number, and it is said by some authors in size, but the increase in size I doubt; the number of round cells increases; the glands expand and become more irregular in their course; a short time before hemorrhage begins, the blood-vessels, especially the capillaries and veins, become greatly distended. We must assume that the connective-tissue cells proliferate, but we have no satisfactory observations upon their division. It was formerly asserted that the menstrual decidua contains decidual cells, but in all the specimens I have studied there are none present. 2. Menstruation. — When the changes just described are com- pleted, the decidua menstrualis is fully formed, and its partial dis- integration begins. The process commences with an infiltration of blood into the subepithelial tissues: this infiltration has hitherto been commonly explained as due to the rupture of the capillaries ; but as no ruptures at this period have been observed, Overlach, 85.1, very justly regards this explanation as inadmissible and thinks the infiltration occurs per diapedesin. It lasts for a day or two, and is apparently the immediate cause of a very rapid molecular disintegration of the superficial layers of the mucosa, which in con- sequence are lost ; the superficial blood-vessels are now exposed, and by rupturing cause the well-known hemorrhagia of menstruation ; by the disappearance of its upper portion the mucosa is left without any lining epithelium, and very much (and abruptly) reduced in thickness. Its surface is formed by connective tissue and exposed * Compare also Schmidt, Amer. Journal Obstet. , Jan., 1884. THE UTERUS. 5 blood-vessels. The third stage is the restoration of lost parts. Signs of fatty degeneration are found during the above-mentioned disinte- gration. Kundratand Engelmann, 73.1, supposed this degenera- tion to precede and cause the hemorrhage ; but this view has not been confirmed by subsequent investigation, it having been found that the degeneration begins later than the bleeding. Overlach, 85.1 suggests that the hemorrhage is caused by ttie gorging of the veins and capillaries, which in its turn is caused by the contraction of the muscles of the uterus compressing the thin-walled veins. Against this view I would urge that it is not shown that marked contraction of the muscles precedes the bloody discharge, and that if it does occur it cannot be assumed that it would cause sufficient com- pression of the veins to produce capillary ruptures. It is desirable to add a few words as to Williams' view, 75.1, 75.2. This author has maintained that the whole, or nearly the >- "Mus. Msa. Fig. 2, — Vertical section of the mucosa corpus uteri or' tlie first day of menstruation; after Leopold. Msc^ muscularis ; Muc, mucosa ; the blood-vessels (shaded dark) are much distended ; the glands much contorted ; there is a subepithelial blood infiltration, in consequence of which the epithelium is partly lost. whole, of the mucosa disappears from the body of the uterus during menstruation. This opinion is often cited as authoritative, espe- cially by English and American writers, but it is now definitely known to be erroneous (Leopold, 77.1, Underbill, 75.1, et al.). It was based upon — 1, failure to consider the effects of disease upon the uteri observed (c/. Wyder, 78.1, 24); 2, erroneous observa- tions ; 3, erroneous interpretations, involving a total disregard of the elementary laws of histogenesis. Minot, 98, -11 3-41 G, describes and figures a normal virgin uterus near the close of menstruation. " The mucous membrane is from 1.1-1.3 mm. thick; its surface is irregularly tumefied; the gland openings lie for the most part in the depressions. In the cavity of the uterus there was a small blood-clot. The mucosa is sharply limited against the muscularis, Fig. 3. In transverse sections one sees that the upper fourth of the mucosa is very much broken down and disintegrated, Fig. 3, d; the cells stain less than those 6 INTRODUCTION. of the deep portions of the membrane ; as represented in the figure the tissue is divided into numerous more or less separate small masses; some of the blood-vessels appear torn through, but it is difficult to make sure observation." Overlach, 85.1, considers it probable that the infiltration of blood takes place by diapedesin, not by rupture of the capillaries. The superficial epithelium, ep, is loosened everywhere ; in places fragments of it have fallen off, and in some parts it is gone altogether ; it stains readily with cochineal and its nuclei color well, the epithelium differing in this respect from the underlying connective tissue, which does not stain well; the blood-vessels in the disintegrated layer are for the most part small. The deeper layer of the mucosa is dense with crowded, weU-stained cells, which lie in groups separated by clearer lines ; in the figufe this grouping shows less plainly than in the preparation ; the lighter channels are perhaps lymph vessels — a suggestion which occurs to me because in so-called "moulds" one sometimes finds similar channels crowded with leucocytes. The cells appear to be the proliferated interglandular tissue ; there are very few leucocytes, so far as I can distinguish ; the cells have small, oval or elongated, darkly stained nuclei, with a very small granular protoplasmatic body each ; there is certainly no noticeable enlargement of the cells, but only a remark- able multiplication. 3. Restoration of the Mucosa. — At the close of menstruation the mucosa is 3-3 mm. thick ; the regeneration of the lost layers be- gins promptly and is completed in a variable time, probably five to ten days. The hypersemia rapidly disappears; the extravasated blood corpuscles are partly resorbed, partly cast off ; the spindle-cell network grows upward, while from the cylinder epithelium of the glands young ceUs grow up and produce a new epithelial covering ; new subepithelial capillaries appear. The details of these changes are imperfectly known ; they effect the return of the mucosa to its resting-stage. (C) . The decidua graviditatis is the decidua menstrualis pre- served ill situ, and considerably metamorphosed in consequence of pregnancy. The preservation is initiated hy the presence of a fertil- ized ovum in the upper end of the Fallopian tube, as is shown for various mammals by observation, and for man by conclusive infer- ence ; and the preservation is dependent for its continuance upon the further development of the ovum in utero. In the very youngest gravidity yet studied (twelve days) very great alterations have oc- curred, and we are reduced to hypotheses to explain how these alter- nations are effected. The ovum at this stage is already attached to the wall of the uterus, and is completely enclosed by a special cover- ing known as the decidual reflexa. The arrangement of the parts can also be followed in older ova, and is illustrated by the accom- panying woodcut, Fig. 4, which represents a median section of a uterus about five weeks pregnant. The whole uterus is considerably enlarged ; the mucosa lining the uterus is very greatly thickened ; to one part of it the ovum is attached ; the mucosa also rises all around the ovum, completely covering it in, so as to make a closed bag. The ovum itself is a sack, known as the chorionic- vesicle, which is covered on all parts by shaggy villi, and encloses the small THE UTERUS. |t ^iS*?;'M®8^pSif3ft.%^^^#4 INTRODUCTION. embryo in its interior ; it is very important to note that only the tips of the chorionic villi come in contact with the mucosa. The mucosa, we thus learn, is divided into three parts : 1, the decidua serotina, the area of the uterine wall, s s, to which the ovum is attached ; 2, the decidua vera, comprising all the remaining portions of the mucosa forming part of the walls of the body of the uterus j 3, the decidua reflexa, the arching dome of maternal tissue, r r, which rises from the walls of the uterus and completely encapsules the ovum. If the walls of the uterus are cut through and simply reflected, leaving the reflex intact, the appearances will be found essentially as in Fig. 5. The mucosa is enormously hypertrophied, and contains a great many dilated, irregular blood sin- uses. From one part hangs down a large bag, the de- cidua reflexa, D. ref., nearly fiUing the cavity of the uterus. The reflexa presents the same general appearance as the surface of the uterus ; if the reflexa be opened we come upon the villous chorion of the ovum, and find as previously stated that only the tips of the villi are united with the surface of the reflexa or serotina. To form the placenta the serotina and the parts of the villi and chorion con- nected with it (chorion f ron- dosum of later stages) un- dergo synchronous hyper- trophy and metamorphoses and become closely united, compare Chapter XVII. In gross appearance the decidua is reddish-gray, spongy or pulpy, soft and very moist ; after the fourth month it acquires, especially in the superficial layers, a duUer brownish color, which subsequently becomes more marked ; this coloration is due to the decidual cells. The vera and serotina are divided each into an upper or superficial more compact layer, and a deeper cavernous or spongy layer, Fig. 6; the two layers are usually of about equal thickness, but the cavernous layer sometimes encroaches upon the compact layer. After the fifth month, they are found very distinctly differentiated. The lumina of the deep layer are the cavities of the enlarged and irregu- FiG. 4. — Semi-diagrammatic outline of an antero- posterior section of the gravid uterus and ovum of five weeks; a, anterior surface; p, posterior surface; g, inner margin of metamorphosed mucosa ; s to s, area of the decidua serotina ; — all the parts of the mucosa adherent to the uterine walls and not included in the area of the serotina constitute the decidua vera; eft., chorion, within which is the embryo enclosed in the amnion, and attached to the walls of the chorion ; ap- pended to the embryo is the long-stalked yolk-sack; the chorion is covered in by the arching extension of the mucosa, which is the decidua reflexa, r r. After Allen Thompson. THE UTERUS. lar uterine glands. During the first two or three months the scat- tered openings of the uterine glands can still be distinguished over (C ^ M o p S'Sp. CD -■ p. -• tp crS P C3 atj the surface alike of the vera and serotina and over both surfaces of the reflexa. The surfaces of the vera and reflexa, though somewhat irregular, remain more or less smooth ; the inner surface of the reflexa 10 INTEODUCTION. is more irregular, and the protuberant parts are united with the tips of the foetal chorionic villi. The surface of the decidua serotina becomes very irregular dur- ing the progress of pregnancy. Rohr, 89. 1, has distinguished three kinds of projections in a uterus of the eighth month, viz. : 1, Hillocks, 1-4 mm. high, and with broad bases, their summits pointed, irregular or even branching ; 2, columns, beginning with a slightly expanded base, narrow stalk and often enlarged ends; the columns are long and stretch up toward the chorion, which they act- ually reach at least in the peripheral parts of the placenta ; in the central region they rise more or less vertically, but obliquely in the peripheral region ; in sections of the placenta they are often cut across, and give rise then to the appearance of islands of decidual tissue in the midst of the villi ; 3, septa, with wide bases rising irregularly to the height of 0.5 to 1.5 cm. ; it is by these septa that the placenta is divided into the so-called cotyledons, compare Chapter XVII. The origin of the decidua eeplexa is uncertain, there being no actual observations upon its genesis. The only view which has hitherto commanded attention is the following : When the ovum at- taches itself to the wall of the uterus, the mucosa (decidua) is sup- posed to form an annular upgrowth around it ; the upgrowth con- tinues making first a high waU, then arching over, and finally clos- ing at the top, dome-like. I do not know with whom this hypothesis originated. In certain rodents also there is a decidua reflexa. Selenka has shown that in them the ovum becomes completely buried in the uter- ine mucosa, and that the part of the mucosa covering in the ovum is converted into the reflexa as the ovum expands. In the hedgehog a reflexa is formed, according to Hubrecht, in a similar manner. Disappearance of the Decidua Reflexa. — A very important change in the disposition of the parts takes place usually during the fifth month, viz. : the reflexa, which, by its own expansion, corre- sponding to the growth of the ovum it encloses, is pressed close against the vera, disappears. Its disappearance has long been known, but until recently was unexplained ; it seems safe now to say that it degenerates and is resorbed, compare p. 19. In consequence of the disappearance of the reflexa the outermost layer (chorion Iceve) of the ovum comes into direct contact with the decidua vera. Be- fore the fifth month, if we cut through the uterine wall in the region of the vera, we come upon the decidua reflexa ; after the fifth month a. similar cut brings us upon the chorion of the foetus. The glands are already dilated in the menstrual mucosa; in pregnancy the dilatation is continued, but is still chiefly confined to the deeper parts of the glands. In the same proportion as the uterus expands the deep portions of the glands become stretched in their transverse diameter and appear during the latter half of pregnancy in sections of the decidua. Fig. 10, as narrow fissures; by the fifth month the glands can no longer be traced in the upper compact layer, their ducts being obliterated. The partitions left between the glands are quite thin. Fig. 10; they carry the blood-vessels and contain spindle cells, and, it is said, also multinucleate giant-cells after the fourth month. Compare the description below of the serotina of the THE UTERUS. H eighth month. The spindle cells, as stated by Langhaus, resemble smooth muscle cells in appearance, but when isolated are seen rather to be broad, round, and flat ; they ought probably to be regarded rather as true decidual cells than as merely enlarged connective-tissue cells. The epithelium of the glands verj' early breaks down, as described by Minot, compare below, p. 16. The epithelial cells at first lie scattered singly in the gland cavity, although patches of them still adhere to the walls; the cells disintegrate. I have observed this degeneration in every one of a large number of specimens which I have examined of all ages up to seven months. The degree of break- ing down may be said in a general way to advance with the duration of pregnancy, but even at term patches of intact epithelium and groups of single cells are always recognizable. The openings of the glands have been shown by Mogilowa, 91.1, to be closed by the growth of decidua; this fact is important, for it shows that the glands cannot discharge any secretion, and shows further that we must discard the suggestion made by Minot, 98, 426, that some of the persistent openings on the surface of the placental decidua are glandular and not vascular. The blood-vessels of the mucosa are all enlarged, those in the deeper parts to a lesser degree than the superficial capillaries and veins,' which are enormously Elated, forming huge, sinus-like cavi- ties in the upper stratum of the decidua. During the latter part of pregnancy the vessels are less conspicuous. The remarkable arrange- ment of the blood-vessels in the 'decidua serotina is fully described in Chapter XVII. ; it will suffice, therefore, to state now merely that the arteries and veins both open upon the surface of the decidua, so that the maternal blood circulates in the spaces between the villi of the placental chorion. The following changes in the blood-vessels must be noted, beside those already mentioned in describing the gross appearances. The vessels of the vera and reflexa reach their maximum development at the end of the second month, when they begin to atrophy, prepar- atory to finally disappearing. Apparently in the serotina, also, the blood-vessels are reduced in volume and number toward the end of pregnancy, but this alteration needs very much to be further inves- tigated. Growth of the Decidua. — With the growth of the foetus and the consequent dilatation of the uterus, the deciduse, of course, must increase rapidly in superficial extension. In fact there goes on a steady growth of the tissues, which however is not sufficient to efeect the expansion of the membrane throughout the whole period of pregnancy in both superficies and thickness. The growth begins by a thickening of the mucosa within the area of the uterine wall to which the ovum is attached, so that during the third and perhaps fourth week this area (serotina) is the thickest portion of the de- cidua (Kollmann, 79. 1) ; but the vera and reflexa also thicken, the former much the most, and soon outdo the serotina. By the end of the fifth week the reflexa measures nearly 2 mm. and the vera fully 1 cm. The absolute thickness of the serotina does not change much after this period, remaining 3 mm. or a little less up to the 12 INTRODUCTION. end of pregnancy. On the other hand, by the eighth month the reflexa has entirely disappeared, and the vera is reduced to about 3 mm. It must be added — 1, that the reflexa is thinner over the poles opposite the serotina than elsewhere, and 2, that the vera thins out toward the cervix and toward the opening of each Fallopian tube. The decidual cells are the most striking of the histological ele- ments of the decidua. They are very large, somewhat flattened, rounded, oval, or branching cells, which assume a characteristic brownish color after the fourth month ; they usually have a single, often nucleolated nucleus, but sometimes two, three, or more up to thirty or forty. Fig. 1 1 . They are exceedingly numerous and continue increasing in number up to nearly if not quite the termination of ges- tation. In size they vary from 0.03-0.1 mm. Kundrat and Engel- mann, 73. 1, and others maintain that the cells undergo fatty degeneration before delivery, and attribute the loosening of the pla- centa to the very fact of the fatty metamorphosis. This view is at best questionable, and it is even doubtful whether the fatty change is a constant phenomenon. Of the decidual cells, we notice particularly the very large ones (giant cells of Leopold) , with numerous nuclei and often with branching processes ; the number of nuclei varies, from ten to thirty and more. These giant cells are said by Leopold, 77. 1, to appear quite abruptly and abundantly during the fifth, month. They lie at first principally in the neighborhood of the blood-vessels of the deep parts of the decidua ; they do not occur in the reflexa, and are far less numerous in the vera than in the sero- tina. The multinucleate decidual' cells are perhaps only interme- diate stages in the multiplication of the uninuclear cells, each nucleus of the large cells finally separating from the parent with its share of the parent protoplasm to make a new decidual cell; if this is the case it accounts for the final disappearance of the giant cells. As regards the function of the multinucleate cells we know nothing; in the rabbit, however, the multinucleate decidual cells have a gly- cogenic function (see Chapter XVII.), but they differ very much in microscopic appearance from the human multinucleate cells, and perhaps differ equally in function. The decidual cells are most abundantly crowded together in the upper or compact layer, and contribute much to give that layer its main characteristics. By the eighth month they are found to have wandered into the cellular layer of the placental chorion, as is more fully described in the chapter on the chorion, apparently finding an entrance at the edge of the placenta. Scattered among the decidual . cells may be found a number of smaller cells which are more conspicuous during the earlier months, and are usually regarded as wandering cells (leucocytes). Lang- haus, 77. 1, 110, regards the leucocytes as the parents of the decid- ual cells — a view I cannot accept. The origin of the decidual cells was long uncertain. Three views contended for acceptance: 1st, they are modified leucocytes (Hennig, Langhaus just cited above, Sinety, 76.1,); 2d, they arise from the connective-tissue cells of the mucosa (Hegar und Maier, Leopold) ; 3d, they are produced by the epithelium. In favor of the first view there has never been, to my knowledge, any THE UJERUS. 13 evidence of importance. The second view has been definitely estab- lished by Minot, 98, 4:29. The epithelial origin was first advocated by Frommel {AerztUches Intelhgenzblatt, Miinchen, 1883, No. 31) for the mouse ; byOverlach, 85.1, for man. Overlach traced the decidual cells to their origin in the epithelium, but his observations are restricted to a, single uterus with pseudo-menstruation from acute phosphorus-poisoning. In the epithelium of the cervix of the uterus in question the following developmental stages of the decidual cells were found : 1, cells with a mother nucleus and one or several, up to fifteen smaller daughter nuclei; 2, cells with a little clearer though granular protoplasm col- lected around the daughter nucleus (or nuclei) ; 3, cells in which the protoplasm about the daughter nuclei has increased and is separated by a clear vacuole-like space from the protoplasm of the parent ; we have then a mother cell, much distended, with a vacuole partly filled by a daughter cell, or by several such huddled together; 4, young decidual ceUs, lying just under the epithelium and closely similar to the endogenous brood in the cells. The observations of A. Walker, 87. 1, on a case of abdominal pregnancy maybe taken as confirming Overlach. Walker found that the peritoneal epithelium at certain points in contact with the chorion had proliferated, forming several layers of cells, presenting an obvious similarity to true decidual cells. Isolated cells of a similar character were observed in the underlyint; connective tissue of the peritoneum. It thus appears that the ovum may cause in other epithelia than the uterine a cell growth analogous to that described by Overlach. Walker, it must be added, maintains that in his specimen the pseudo-decidual cells also arise in part by metamorphosis of the connective-tissue cells. I am inclined to interpret Overlach and Walker's observations as evidence of hyper- plastic degeneration, and not of the production of decidual cells. The manner in which the true decidual cells arise is described in the next section. For a description of the fully developed cells see p. 18. Uterus One Month Pregnant. — The specimen to be described came from a woman who committed suicide by violence. The speci- men was received in very fresh condition, but the reflexa was badly torn ; the embryo had been removed, and I was therefore unable to verify the age, or investigate the attachment of the villi of the chorion to the uterus. There was a beautiful corpus luteum in one ovary, quite similar to that figured by Dalton in his report on the corpxis luteum in the Transactions of the American Gynsecological Society for 1877, Fig. 9. My specimen enables me to confirm in most respects Turner's accurate description of two uteri of about the same age, 79.1, 546-648. The inner surface shows the hillocks {Inseln) described by Eeichert in the uterus of two weeks studied by him, which have been figured by Coste in slightly older specimens, and foimd by Turner also, 79.1, 540. The four illustrations given herewith are all from sections through what I suppose to be the placental region. There is an upper compact layer, Fig. 6, D, and a lower cavernous layer D' ; the caverns being gland cavities, which appear as rounded 14 INTRODUCTION. areolae partly lined with epithelium, partly filled with broken-down epithelial cells. The drawing, reproduced in Fig. 6, was obtained by drawing the outlines very carefully, stippling the areas occupied by the connective ?'o^;"r""^'!?^?!^v^v'v'^ tissue, represent- ;^;^;0::-:-^M-^;h;^;v);i;iiai-^ ing the blood-ves- -■■■^^-':''W''''^i^ ^!i:0^-^0^O^M sels by double out- '-T^<'-M-^::^^^ l™es, and omit- '-ii:-.-- ■':■'■/ i©/iu&0> ' f-5:S'i^;^:. :;'■* ting the glandular -^'^ /v ■-;-:---■ '''''ifv/' /v''^S-'^'/'''/"''^-v!S^^^ epithelium alto- i;;;!fe:;::p33:';.;i>-:-:';-;; ;■ : > -'''^./v' ;•, ::.^^ gether. It wiU be W)]^'^ x'': ":■^!>•C-<'■'"''':■.■^ ^'■^■^'/^'■''^^'^^■■^^^U'i^''^ noticed that about ^,. r'Ti/'' " ^!'-'<;^l!l---/ /^•■■' -;,;'i^^ J^::--^i^ three - fourths of 0:0A^;r1:'--^'' /■'}<: ■ "- ..--'^'iSK;: .,:^:^j""~"'''' the diameter of H;i;^i'~%v^' .■■'''■ 'v-r :■:'-.- I'L^-Wy'^"^',, ./"",- the mucosa is 00- N-^-v4s "'"-'-. '^!'"t''%^-"''2''-'5 ^r':''::^."V^r- cupied by the cav- '"O-^--- ::.C ' "^ f^fc;'S;.i.r„ t::\ ''^;;'^' ■■> :: ernous layer. If . g#^~'"~^^"'^3S:^'"'^l^ft The upper or r'^-'k ^ '''-v^~^^:^>,fMSS^-^:^'-"--* y:;li;^5C- • compact layer is >•v„^■*^:; :-"] ^'^i^=£i-3%S^$K•^/^ ■fc^t^-'-^,--; shown in Fig. 7. -r"%t->>-./^ ^5^^=i-I-5:'3'*^^t " The surface is "^jv;,.-/ C._:^%,_ y^-f^ ''""i^s4!:^^~.. ■'■-■:'x^'.!^>;i without any trace _,_ ""'X ,'■■'""--<}■■ ; v' "''"";;?^-^V---'''''%f9'^5^''*^^^r"'^ °^ epithelium, and ^^■'^■■:>r:^''^i^:^:r-^^t:.,^i'$MiiiSi^'' '% fT^r^?. is covered only by 7\ ,-<^-i:^!^:-^:.^>-^''^'^WXsf J^'-r-- i a thin fibrous and )...?'' ^:L-:y ViZ7' >|a|;5s^ 'fir^i;;^^';;^ ' granular coagu- ' /^K ..y ,f-\0;^^,.^:,^,^'%<-r-5^^':'--:. him., coagl; the ^'!^~S^P^!^{ ....•■'^i~--'^^'^X^^''^'^-&>--''''''"^ tissue itself con- ''■■■C.,.._?'"'';iJ-?7^ /' :v"- ■•■ib?"'^@ ( """J^f^'S'" sists almost exclu- ;:-^':5;--^i-r^'""^\-<-' ■ r'M-f0''^^-~'S:'^:vyp--'^^'f^'f::K sivdy of young t;t>^t^'?^-^'^if>^^-"/'3SS^ .-.--P-"^^^ decidual ceUs, d, |:;..l;:S?;:?:--j^"'', J ''^|i^€>-^ 'S;.'' ; >;v/ d', with a clear, f''':7i::i-:'' ' ":' '. ■;■ ' C' rs';:/^^^^^^ '; v_...< ':;::;--' homogeneous j^:vii^iS/:;:.;.;;,:V:.:^,h;i-- -r.- r ■,. ^ ■'•;--• matrix; here and \:^:^/0;:::;P::-/MW^r:r:f:if '-K '"■•;,■. ^-'X :',:;; -C,'"" there are leuco- i£iM^;^:;ii;c-;iS:;i:Ji;;S^^ cytes, but they are muse — -""• — ---" - — -— ~~'' - — nowhere numer- I'lG. b,— Uterus one month pregnant; outlines of the glands from a nuS" the decidual vertical section; to show the division of the mucosa into an upper Tl^' t, compact layer D' and a lower cavernous layer i5"; gi', gr, glands; art, cellS are ail Q Ulte spiralartery;m,«c,muscularis. ^^.^^^^ with their bodies deeply stained by the eosin ; the nuclei are round, oval, or slightly irregular in shape, coarsely granular, and sharp in outline; the cells themselves, though irregular and variable in shape, are all more or less rounded with processes running off in various directions ; scattered between the cells are many sections of their processes; occasionally it can be seen that two cells are connected ; in fact, we have in this tissue evidently a modified embryonic or so-called anastomosing connective tissue. Now, as we know through the ob- servations of Leopold, 77.1, which I have verified, the connective tissue of the uterine mucosa consists of anastomosing cells, and as stated in the previous section the cells are found proliferating in the menstruating uterus ; we have, therefore, only to imagine the cells n'X THE UTERUS. 15 enlarged with eertain accompanying modifications to obtain the tissue figui'ed in Fig. 7. There is no special formation of cells around the blood-vessels, where, according to Ercolani, the decidual tissue arises by new formation. In Turner's specimens the upper part of the compact layer was imperfectly preserved, but according to his description there appears to have been a coagulum similar to that which I have found, but thicker. In the deep part of the layer the cells are less enlarged, and when the cavernous layer is reached there occurs a rapid transition in the character of the cells, which become smaller and more fusiform, and their nuclei more elongate, smaller, and deeper stained by alum-cochineal. The gland openings upon the surface of the uterus lead into tubes, Fig. 6, gl', which run slightly obliquely through the compact layer, taking a more or less nearly straight course and joining the contorted gland tubes, Fig. 6, gl\ of the cavernous layer. The gland ducts are completely devoid of lining epithe- lium, which has disappeared ex- cept for a very loose cell, occa- sionally found lying free in the ducts; the cells have not fallen out from the sec- tions, but were lost before the tissue was im- bedded.* The ducts then are wide tubes run- ning nearly straight through the upper part of the decidua and bounded direct- ly by the decid- ual tissue; they c o m m u n icate below with the contorted cavi- ties. The cavern- ous layers con- tain numerous , ^ • j. spaces, the areola of Turner, 79.1, 547, who was uncertain as to their character, though he asc^tained that many of them belong to the glandular system. In my specimen it is perfectly clear that aU the larger areola belong to the glands, which must be extremely dis torted and distended to give the shapes shown m Fig. 6 Fig r— Uterus OBe month pregnant; poition oi the compact layer of the decidua seen in vertical section; coagl, coagulum upon the sur- face; d, d', decidual cells. X 445 diams. The thin flnr^^r^^^^^^f^S^^nlg^J^^J^-f^-r'-P-^^ 10 INTRODUCTION. dissepiments between the areolae are composed of connective tissue, the long dark nuclei of which, Fig. 8, are strikingly different from those of the cells of the compact layer, Fig. 7. The areolae present two extreme modifications and all intermediate phases between these two. The smaller areolae are lined by a well-preserved cylinder epithelium, or by one in which the cells are separat- ed by small fissures; in other areolae the cells are a little larger, Fig. 8, each for the most part cleft from its fel- lows, and some of them loos- ened from the wall and lying free in the cavity. The other extreme is represented in Fig. 9; the size of the areolae is much increased — compare Figs. 8 and 9 — both drawn on the same scale ; the epithelium is entirely loosened from the wall, and the cells lie separate- ly in the cavity which they fill ; the cells are greatly enlarged, their bodies having three or four times t^ie diameter of the cells in the small areolae; they have not the cylinder shape, but are irregular in outline; their protoplasm is finely granular and stains rather lightly; the nuclei are large, rounded, granular, and with sharp outlines; they are less darkly stained than the nuclei of the epithelium of Fig. S. The obvious interpretation of the appearances described is that Fig. 8.— Uterus one month pregnant; section of with the epithelium gland from cavernous layer., partly adherent to the walls. X 445 diamst r-' Fia. 9.— Uterus one month pregnant; section oiftland from cavernous layer with the epithe- lium loosened from the walls. X 445 diams. ihe glandular epithelium is breaking down, being lost altogether from theducts, but is still present in the deep portions of the glands ; in breaking down the cells separate from one another, and then from THE UTERUS. 17 the wall, and falling into the gland cavity there enlarge, the cavity- enlarging also. Similar appearances are also found in " moulds" of the second month ; very likely they have been often observed and mistaken for pathological changes. The blood-vessels of course lie in the dissepiments betvi^een the glands. I observed nothing to 'correspond with the " colossal capil- laries dilated into small sinuses," mentioned by Turner, 79.1, 548. Were not these supposed capillaries gland cavities, from which the epithelium had fallen out? Occasionally the sections pass through a spiral artery. Fig. 6, art, which is cut again and again as it twists around in its characteristic separate column of connective tissue. Decid.ua Serotina at Seven Months. — In a normal uterus about eight months pregnant I find the following relations : The serotina Fis. 10.— Section of the deeidua serotina, near the margin of the placenta; normal uterus about seven months pregnant, nic, Muscularis; D\ Z)", deeidua serotina; i>', cavernous or spongy layer, the spaces in which are the glands; /?", compact layer; Ft, scattered chorionic villi ; the intervillous spaces were filled with blood, which is not represented in the figure. is about 1.5 mm. thick, and contains an enormous number of decidual cells, Fig. 10 ; the cavernous, D', and compact layers, D", are very clearly separated ; the mucosa is sharply marked off from the mus- cularis, although scattered decidual cells have penetrated between the muscular fibres. The muscularis is about 10 mm. thick and is characterized by the presence of quite large and numerous venous thrombi, especially in the part toward the deeidua. The deeidua itself contains feiv blood-vessels. Upon the surface of the deeidua can be distinguished a special layer of denser decidual tissue, which in many places is interrupted by the ends of the chorionic villi which 2 18 INTRODUCTION. have penetrated it, as is well shown in the accompanying Figure 10. The gland cavities of the spongy layer, D', are long and slit-like ; they are filled for the most part with fine granular matter, which stains light blue with haematoxylin ; they also contain a little blood, and sometimes a few decidual cells. I have also seen in them a few oval bodies several times larger than any of the decidual cells, and presenting a vacuolated appearance. What these bodies are I have not ascertained; in a number of uteri over two months pregnant I have found them invariably present. In many places the glandu- lar epithelium is perfectly distinct ; its cells vary greatly in appear- ance, neighbors being often quite dissimilar ; nearly all are cuboidal, but some are flattened out ; of the former a number are small with darkly stained nuclei, but the majority of the cells are enlarged, with greatly enlarged hyaline, very refringent nuclei. There are also in many of the gland spaces isolated enlarged cells, which have detached themselves from the wall, and in some cases the detached cells nearly fill the gland cavity, very much as in Fig. 9. The decidual cells of the cavernous layer, Fig. 10, D, are smaller and more crowded than most of those of the compact layer. The largest cells are scattered through the compact layer, but are most numerous toward the surface. They extend around the margin of „ the placenta and have OnQ^f^T^ penetrated the chorion, in ^^ ^ the cellular layer of which they are very numerous (compare on this point the chapter on the Chorion) ; the immigration has im- parted to the chorionic layer in question some- what the appearance of a decidual membrane. Mis- led by this peculiarity, KoUiker and others have held this layer to be mater- nal in origin, and accord- ingly have described it as a "decidua subchorialis." The error was, so far as I am aware, first definitely corrected by Langhaus, , 77.1. The decidual cells exhibit great variety in their features. Fig. 11. They are nearly all oval discs, so that their outlines vary according as they are seen lying in the tissue turned one way or another ; they vary greatly in size ; the larger they are, the more nuclei they contain ; the nuclei are usually more or less elongated ; the contents of the cell granular. Some of the cells pre- sent another type, c ; these are more nearly round, are clear and transparent; the nucleus is round, stains lightly, and contains relatively few and small chromatin granules; such cells are most numerous about the placental margin. Fig. 11. — Decidual cells from the section represented in Fig. 10; stained with alum hasmatoxylin, ancf eosin; a, 6, d and/, various forms of cells, from the serotina; c. giant cell, from the margin of the placenta; e, clear cells from the chorion. At a, seven blood globules have been drawn in to scale to afford a ready measure of size. THE UTERUS. 19 Fate of the Decidua Eeflexa. — The decidua reflexa is a dis- tinct membrane up to the end, it is said, of the fifth month of gesta- tion, and after that period it can no longer be found. Exactly at what time it disappears is not established by observation, though the fact of the disappearance has long been known, nor have we had hitherto any definite knowledge as to how it disappears, although its gradual attenuation and increasing transparency during the first four or five months have been familiar to us since the publication of Coste's magnificent atlas. The view most generallj' accepted has been that it fused with the decidua vera, and that accordingly the layer of decidua nearest the chorion during the latter half of preg- nancy represents the decidua reflexa. I have had opportunity to study four well-preserved normal preg- nant uteri of two, three, five to six, and seven months' gestation respectively. These show that at two months the decidua reflexa is undergoing hyaline degeneration, that at three months the degener- ation is considerably more advanced, and that by the sixth and , seventh month the reflexa can no longer be found. These obser- vations justify the theory that the reflexa degenerates and is com- pletely resorbed. I will review briefly the actual observations : First, the reflexa at two months. It starts from the edge of the placental area with considerable thickness, which is rapidly lost, most of the reflexa being a thin membrane and the thinnest point being opposite the placenta. The examination of sections shows that the entire reflexa is undergoing degeneration, which is found to be the more advanced the more remote the part examined is from the placenta. The chorion laeve lies very near the reflexa, being separated only by chorionic villi, which are very much altered by degeneration, their ectoderm having become a hyaline tissue, which stains darkly, and their mesoderm showing clearly the partial loss of its cellular organization. In the region half-way between the base and the apex of the reflexa dome the tissue of the decidual membrane shows only vague traces of its original structure; only here and there can a distinct cell with its nucleus be made out, for most of the cells have broken down and fused into irregular masses without recogniz- able organization. Eamifying through the fused detritus there are two layers of so-called "flbrin," or, in other words, of a hyahne sub- stance, which like the " canalized flbrin" of the chorion stains very deeply with the ordinary histological dyes, carmine and logwood. The fibrin is much more developed upon the inner or chorionic than upon the outer side of the reflexa. It forms on the inner side a dense network, which on the one hand fuses with the degenerated ectoderm of the chorionic viUi wherever the villi are in contact with the decidua; and on the other handramifles more than half-way through the decidua, the ramifications being easily followed, owing to the hyaline character and deep staining of the "fibrin." Upon the out- side the fibrin forms a thinner layer, and shows its network structure in many sections much less clearly. All of these points are lUus- trated by the accompanying figure. .-,■,-, In the uterus three months pregnant I find essentially the same conditions, except that the degeneration is farther advanced, smce 20 INTRODUCTION. the traces of cellular structure in the reflexa are still more vague and the fibrin is more developed. The membrane is much thinner than at two months ; the thickness is about two-thirds of what it was. In the fresh specimen the membrane appeared much more transpar- ent than before. In all the parts examined I found leucocytes pres- ent, and in the region of the reflexa near the placenta they are very numerous and conspicuous; it is natural to conclude that they are concerned in the resorption of the degenerated tissue. In a section not far from the base of the reflexa the three layers are distinct as at two months, there being a thicker inner and a thinner outer fibrin layer, while between them is a stratum in which remains of cells are seen ; occasionally is an appearance which suggests a sur- viving decidual cell, and nearer the placenta the phantoms of cells Fig. 13.— Section of human decidua reflexa at two months. become distinctly cells, and true decidual cells can be made out. The inner fibrin layer is much denser and its meshes smaller than in the two months specimen, the trabeculee of fibrin having become thicker during the month elapsed. Those who conceive that there is a fusion between the reflexa and vera, are forced to seek for traces of the former membrane next the chorion. They may assume either that the epithelioid layer (cho- rionic ectoderm) is the remnant of the decidua, which forces them to leave the fate of the chorionic epithelium unexplained, or that the upper stratum of the decidua is the reflexa which is fused with and acquired the same structure as the underlying vera. If my obser- vations on the degeneration of the reflexa are correct, and corre- spond, as there is sufficient ground to believe they do, to normal con- ditions, then both assumptions as to the persistence of the reflexa involve the further and very improbable assumption that the degen- erated tissue is removed and replaced by fully organized cellular decidual tissue. It is obviously more in accordance with our knowl- edge of degenerative changes to assume that the hyaline metamor- phosis is necrotic and is succeeded by the disintegration and removal THE UTEKUS. 21 of the tissue. This accounts in a satisfactory manner for the absence of the decidua reflexa during the sixth and seventh month. The relations of the membranes at this period have been well described and figured by an admirable observer, Dr. G. Leopold, whose views and one of whose drawings have been incorporated by Prof. O. Hert- wig, in his " Entwickelungsgeschichte" (third edition, pp. 216-317, fig. 147) . Leopold holds that the epithelioid layer is the reflexa ; but what has just been said suffices, I think, to show that this view is untenable. That the membrana decidua reflexa should degenerate and disap- pear no longer seems strange, since recent investigations have shown that in many placental mammals there occurs an extensive pseudo- pathological destruction of the mucosa uteri during gestation. These changes, which are best known in the rabbit (c/. Minot, Biol. Cen- tralbl., X., 114) vary considerably in character and are exceedingly remarkable both for their extent and for their numerous modifica- tions, so that we need feel no surprise at the entire destruction of the decidua reflexa in man, nor at the form of the destruction being unlike the forms hitherto found in other mammals. As to the purpose or advantage of the sacrifices of maternal tissue we are in the dark. The same is true of the causation of the degen- eration, although we must regard it as the result of a reflex nervous activity. It is becoming more and more evident that the nerves have a profound influence upon organization, and it is no strained hypothesis which places the structure of the mucosa uteri under the immediate control of the nervous system. The changes in the decidua at parturition require special description. During labor a split occurs in the decidua serotina and vera ; all the parts within the split — that is, toward the chorion — are expelled, their expulsion being part of the act of delivery ; the term decidua or caduca refers to the fact that the membranes are cast off ; they are discharged after the foetus, and, together with the vera and foetal envelopes, constitute the so-called after-birth. There are thus removed the superficial portions of the vera and serotina. The split, according to Friedlander, 70.1, 76.1, usually occurs in the upper or compact layer just above the cavernous layer, leaving the surface of the uterus smooth and glistening, but the surface of the placental area is thrown into irregular hills and valleys. Some- times the split occurs at or just below the upper limit of the cavernous layer, in which case the surface of the uterus after parturition is jagged and irregular. In rarer cases the split occurs higher up in the compact layer, leaving consequently by far the greater part of the decidua in situ quo ante. In all normal cases, however, more' of the mucosa is lost than in menstruation, and a considerable portion is always left in utero; this latter portion contains the remnants of the uterine glands, and is the organ of regeneration for the entire mucosa; it has, of course, no epithelium upon its surface, which in- stead is formed by connective tissue and ruptured blood-vessel (and lymphatics?). The layer of vera left on the uterus is usually about 1 mm. thick; that of the serotina may be considerably less. The post-partum regeneration of the mucosa begins very soon, but varies greatly in the rate with which it progresses, being very 32 INTRODUCTION. rapid in vigorous, healthy women and slow in weakly women. The region of the vera is restored more rapidly than the placental area. The first step is the thickening of the mucosa to about 2 mm., owing to the contraction of the uterus, which of course reduces the superficial extent without altering the volume of the mucosa. In consequence of this change also the gland spaces become rounder and Fig. 13.— Uteinis twelve hours after artificial delivery at six months pregnancy; coagl^ blood clot; X>, decidua; m, muscularis. X 22diams. the course of the glands straighter. I wiU here interpolate a descrip- tion of a human uterus, twelve hours after abortion, see Minot, 98, 428. The uterus was apparently normal; it was already very much contracted; the mucosa measured about 1 mm. in thickness; the surface was ragged and more or less covered with clotted blood, pre- senting very much the appearance so superbly figured by Coste ("Devel. corps organises," pi. x. Espece humaine). Vertical sections. Fig. 13, show that the surfaces of the mucosa are very un- even ; on the free surface there is a thin layer of clotted blood, coagl; the upper or compact layer of the decidua has entirety disappeared, "leaving only the deep portion, D, permeated by numerous large gland spaces, between which are partitions containing the brownish and hyaline decidual cells, and a great many blood corpuscles, which lie in the tissues as weU as in the blood-vessels. The presence of blood corpuscles in the tissues is probably a constant feature of the decidua post partum." The second step is the restoration of the surface by the resorption of the blood and detritus, parallel with which advances the restora- tion of the glandular epithelium. These changes occupy apparently THE UTERUS. 23 from seven to fourteen days. The cuboidal gland cells at this time appear swollen, with indistinct intercellular boundaries ; the nuclei are almost all enlarged until they nearly fiU the cells; rapid ceU divi- sion is going on. At this time also venous thrombi are very con- spicuous, especially in the placental area, where they are found fresh and in various stages of progressing obliteration, Fig. 14. The thrombi persist for a long period (Leopold, 77. 1, xii., 185). The third step is the completion of the restoration of the glands up to their external openings, and the regrowth of the normal connec- tive tissue of the mucosa. The resulting stage was found by Leopold, 77. 1, xii., 199, to have been reached in a normal uterus three weeks after parturition. Of this specimen he gives the following descrip- tion, which refers to the placental region. " As shown by the illus- tration (Fig. 14) the young mucosa is composed mainly of fine short spindle cells, which form the interglandular tissue. They exhibit extraordinary proliferation, and are showing themselves in numerous processes (Zapfen) into the musculature, but still leaving the limits of muscularis and mucosa distinct as in every non-pregnant and preg- nant uterus. Secondly, between the young cells we find many blood- vessels, especially capillaries, in the neighborhood of which are col- lected blood corpuscles, hsematin crystals, and pigment. Many ap- pearances indicate the new formation of capillaries from simple cords t ^^ j^ ^m'- m T ?( / ,•' V Pis. 14.— Section of the placental aiea of the uterus three weeks post pai tu i Uuc mucosa; Msc. muscularis. After Leopold. of cells, which extend to the very surface. Thirdly, and most im- portant, we find the young glands, which are short vertical folhcles, imparting to the surface a more definite sieve-like appearance. Their cuboidal epithelium is spreading out from their mouths to re- cover the surface; but at this time the new epithelium is not yet completed. The mucosa is still a wounded tissue; for its complete restoration there is still lacking . . . the vascular network." The fourth step is a double one: the restoration of 1, the superficial 24 INTRODUCTION. epithelium, which is accomplished hy the spreading of the growing epithelium from the mouths of the glands, and of 3, the subepithelial network of capillaries. The completion of this, the last step in the restoration, has been observed in a normal uterus six weeks after parturition. A very different regenerative process is stated by Duval, 90.3, to occur in rodents ; he believes that in these animals the epithelium is reproduced nearly simultaneously over the rupture surface by a direct transformation of the connective-tissue cells of the placental decidua. 4. Mucosa Cervicis Uteri. — The mucosa of the cervix has been only very imperfectly investigated. It resembles somewhat that of the body of the organ ; but is distinguished first, by the possession of two kinds of glands, one agreeing with the utricular or uterine glands proper, the other of the "mulberry" type, there being numerous alveolar branches of the gla,nd cavity ; second, by the character of its lining epithelium, composed of enormous cylinder cells of many shapes, in length averaging fuUy 55 /x (c/. Overlach, 85.1, 214, 219ff.). The stratified epithelium of the vagina does not, it appears, nor- m^ally extend inside the os. The utricular glands are lined by an epithelium like that of the corpus, while the epithelial cells of the " mulberry" glands resemble those lining the cervix ; the latter glands are in fact strictly cervical, and apparently secrete only mucous matter ; they are very likely important contributors to the plug of mucus which closes the cervix during pregnancy. The cervix, except for this plug, remains open during gestation ; it also preserves its covering epithelium, and although it becomes tumefied during gravidity, and may, as claimed by Overlach, par- ticipate in the formation of decidual cells, it never, as far as j'et as- certained, forms a true deciduous membrane. A thorough investigation of the histology of the cervix in aU phases of the uterine functions would be extremely valuable. 4. Blood -Vessels. — The uterus is supplied from four arteries: two, the ovarian, running along the broad ligaments and giving each a considerable branch to the fundus; two, the uterine, derived from the internal iliacs, rimning to the cervix, and thence mounting by a very tortuous course toward the fundus to there join the ovarian arteries. The arteries give off very numerous branches, which take a characteristic spiral course through the muscularis, and form fre- quent anastomoses with one another. The arterial vessels of the uterus are further remarkable for the great development of their muscular walls, aU the more striking because the muscular coat of the capillaries and veins is slightly developed. The capillaries are wider in calibre than usual, and form specially distinct networks under the epithelium and around the glands of the mucosa. The veins are very wide, almost sinus-like, even in the resting uterus. During and just before the menstrual flow, and stiU more during the first half of pregnancy, the vessels are all dilated, and it is thought by some actually increased in number ; this latter opinion may be fairly doubted. The increase in the amount of blood is very obvious; indeed Eouget, 58.1, speaks of the tissue of the uterus as erectile, but this adjective is not applicable in the anatomical THE UTERUS. 25 sense, as Kolliker has very properly pointed out. The vascular enlargement afiEects principally the capillaries and veins (Turner). It is most marked during the second and third month of pregnancy ; in the fourth or fifth month the vessels begin to atrophy, and by the eighth month, as previously stated, the vessels are far less nu- merous; these changes require further investigation. A number of large venous sinuses remain, however, especially in the inner portion of the muscularis, and are highly characteristic of the latter half of the period of gestation. Large thrombi normally appear in these sinuses, becoming first noticeable during the eighth month and persisting several weeks post partum. Apparently they continue to arise during the eighth and ninth months and even after delivery (Leopold). The thrombi, which were first discovered by . Friedlander, 70.1, 76.1, and have been studied also by Leopold, are supposed by the latter au- thor to be directly caused by an immigration of giant cells into the veins. Leopold further supposes, 77.1, xi., 492-500, the presence of the thrombi to cause venous congestion of the uterus. Now, if it is true, as Brown-Sequard has maintained (" Experim. Researches Applied to Physiol, and Path.," 1853, 117, and Brown- Sequard's Journ. Physiol., i., 1858, 105), that carbonic acid excites toward the end of gestation uterine contractions very readily, then it is possible that the venous congestion above mentioned may be one of the proximate causes of parturition. Additional facts in regard to the blood-vessels during pregnancy are given, pp. 23, 27. 5. Lymphatics. — Our knowledge of this subject rests princi- pally upon the admirable memoir of Leopold, 74.1. The system begins in the intercellular spaces of the connective-tissue layer of the mucosa ; in this and in the muscular layer are lymph capillaries, which corhmunicate with the subserous (subperitoneal) network of lymphatics. Special Physiology of the Uterus.— Our anatomical study has shown us that the most remarkable changes of the uterus during its menstrual and gestative functions are: 1, the gradual thickening of the mucosa; 2, the removal of the superficial portions of the mu- cosa, in the one case during the menstrual flow and in the other dur- ing labor; 3, the appearance of an enormous number of the very characteristic and peculiar decidual cells during the thickening of the mucosa. The menstrual and gravidital changes follow the same cycle, and differ from one another essentially only in two points : 1, the time occupied, and 2, the extent of the changes. In fact the alterations, though of the same character, are greater in extent and occupy a longer period during gestation than during menstruation. These considerations force us to the conclusion that the gravid uterus is passing through the menstrual cycle prolonged and intensified. The function of gestation is a direct modification of the function of men- struation, and the two are physiologically homologous. The deduction is so evident that I have been surprised not to have yet encountered it clearly enunciated in any of the authors I have consulted. That the decidual cells perform some very important function seems to me likewise evident from their great prominence, but until 26 INTRODUCTION. their history has been elucidated even as to details, we can hardly hope to ascertain what that function is. We may surmise that they are either organs of regeneration, or of nutrition for the embryo, or of both functions. The cause of the formation of the decidua either in menstruation or in gestation is unknown. The presence of the impregnated ovum in the upper end of the Fallopian tube seems to be the cause of the arrest of the menstrual changes and the preservation of the decidua upon the uterine wall. How it produces this effect is unknown, but it is fair to assume that it takes place through the central nervous system. Experiment might demonstrate the nervous pathways fol- lowed by the irritation and the reflex, and perhaps discover a trophic centre in the cord for the uterus. That the impregnated ovum, when it exerts this influence, lies in the upper end of the oviduct quite remote from the uterus seems certain from analogy with mammals. Presumably the ovum undergoes rapid degeneration during its passage through the oviduct, and can be saved only by fertilization at the start. Lowenthal, 85.1, who shares the too frequent misap- prehensions of gynaecologists in regard to the site of impregnation, and thinks in his philosophy that it is impossible for a remote ovum to exert such a marked influence on the uterxis, has advanced the hj^pothesis that the ovum is fertilized in the uterus and affects it by direct contact. His critic, Wyhoff {Centralbl. f. Gyncek., 1885, No. 26, 401), thinks im.pregnation may occur either at the ovary, in the Fallopian tube, or in the uterus ! Such references to opinions on this subject, advanced without proper knowledge, might be readily multiplied. But if the decidua graviditatis is produced by the influence of the impregnated ovum on the menstrual membrane, we have still to ask, What causes the formation of the decidua menstrualis? To this no answer is possible. Pfliiger has advanced the theory, 65.1, that the ripening Graafian follicle exerts through the central nervous sys- tem a reflex action upon the uterus ; but, inasmuch as the attempt to establish a fixed relation in time between the ripening of the follicle and menstruation failed (Leopold, 83.1), it is impossible " to accept Pfliiger's theory at present. That menstruation is connected with ovulation appears probable, but that ovulation has a constant casual relation to the monthly period is by no means demonstrated. The belief in the connection is favored by the fact that the operative extir- pation of both ovaries usually, but not invariably, causes menstrua- tion to cease. Putnam-Jacobi has advanced a theory in regard to the cause of menstruation (see Amer. Journ. Obst., Apr., 1885), which is based upon singular false homologies between the ovary and uterus, and some physiological assumptions which are, I think, not admissible. Other theories, likewise not tenable in my judgment, have been advanced, but it seems undesirable to dwell upon specula- tive views. The cause of the formation of the reflexa is connected with the ovum, since wherever the ovum is attached the reflexa is formed around it ; how the ovum after its attachment exerts its influence, is unknown. Since the position of the ovum determines that of the reflexa it becomes the more interesting to put the question, What THE UTERUS. 27 determines the site of attachment of the ovum? which, unfortunately, is at present an unanswerable inquiry. The cause of delivery is not ascertained, but has been much de- bated. Various suggestions have been made to explain why the decidua cleaves in two, and why the uterus contracts to expel the foetus. Our inquiry as to the cause of birth may be resolved into two component questions : 1, What is the stimulus which causes the uterus to expel the foetus ; 3, What causes the stimulus to act at a certain period after conception, i.e., what determines the duration of pregnancy? The second question I hope to discuss elsewhere. As regards the first question. What stimulus causes delivery? it is weU known that various operative procedures can excite apparentlj- by reflex action contractions of the pregnant uterus which will result in the expulsion of the ovum. It is by taking advantage of this possibility that abortions (premature deliveries) are procured. Such stimulations as are referred to may be caused in the folloAving ways: 1, by rupturing the amnion and allowing the amniotic fluid to escape from the uterus; y, by the introduction of foreign bodies between the walls of the ovum and those of the uterus; 3, by me- chanical irritation of various parts, especially the cervix uteri, the external genitalia, or the breasts. With these facts in mind the hj'pothesis is unavoidable that the normal contractions of the uterus at full term are due to reflex stimulation. Varioiis authors have accepted this opinion and endeavored to ascertain the starting-jjoint of the stimulation. Mauriceau sought it in the uterus having reached the limit of its expansibility; Naegele in the irritation caused by the embryo, acting like a foreign body in the uterus ; Engelmann, at least partly in the degeneration of the decidual cells ; Harse and others, in the accumulation of carbonic acid in the blood of the uterus. None of these views are very well founded ; the two last deserve, however, a little more consideration. The fatty de- generation is not adequate, because in several instances it has been found wanting both before and immediately after birth (Sinety, 76.1, Meola, 84.1). The carbonic-acid theory is presented in its most plausible form by Leopold, and has been already stated (p. 43) . To what is there said may be added that it is not shown, 1, that venous thrombi cause the venous congestion of the uterus assumed by Leopold, and 2, that such congestion would charge the uterus with sufficient carbonic acid to excite contractions in it. Compare also Spiegelberg's "Lehrbuch," 1880, p. 120. We evidently have to do Avith a progressive maturation of the uterus — a series of changes we cannot explain, but which is, as al- ready pointed out, closely similar to the series of changes during menstruation. Hence it is probable that there is a common cause for the ending of the series (the casting off of the superficial part of the mucosa in both cases) ; in the delivery there is superadded the contraction of the uterus, and for this a\ e must see a cause also. Therefore it seems to me that it is undesirable to search for one cause only for the whole process of birth. The physiology of delivery does not fall within our scope ; for fur- ther information the reader is referred to Hensen's " Physiologie der Zeugung. " CHAPTEE II. GENERAL OUTLINE OF HUMAN DEVELOPMENT. This chapter is designed especially for the convenience of stu- dents of medicine and biology. Advanced students will find in it little of value to them, since all the subjects it considers are more fully treated in other portions of the volume. I. Retrogressive History of the Fcetus and its Envelopes. Uterus Eight Months Pregnant. — If we examine a pregnant uterus at any time during the sixth to ninth month of gestation, we find essentially the same relations of the parts — the most marked difference being in the size of the uterus, which increases with the duration of gestation, to correspond to the growth of the fcetus. A description of a uterus of the eighth month after conception will suffice, therefore, for our present purpose. Such a uterus is a large, rounded bag, with muscular waUs, and measures seven or eight inches in diameter. It renders the abdomen very protuberant. Examined externally it is remarkable especially for the numerous large sinus-like blood-vessels; its surface is smooth; the texture of the walls is firm to the touch, but the walls yield to- pressure, so that the position of the child can be felt. As the placenta is generally upon the dorsal side, it is usual to open the uterus by a crucial incision upon the ventral side. The walls are about one-half of an inch thick, sometimes more, sometimes less, and as soon as they are cut open we enter at once into the cavity of the uterus containing the foetus and nearly a pint of serous liquid — the liquid is the amniotic fluid. The foetus normally lies on one side, has the head bent forward, the arms crossed over the chest, the thighs drawn against the abdomen, and the legs crossed ; it resembles closely the child at birth, but is smaller ; its head is relatively to the size of the body larger ; the abdomen is more protuberant, and the limbs proportionately smaller. The inner surface of the uterus is smooth and glistening ; if it is touched with the finger it is found to be covered by a thin but rather tough membrane, called the amnion, which is only loosely attached. Examination of the uterine wall, where it has been cut through, shows that its thickness is formed principally by the muscular layer, which is made up by numerous laminae of fibres, between which are the large and crowded blood sinuses, similar to those distinguishable on the external surface of the uterus. About one-fifth or less of the wall inside the mus- cularis has a different texture and can be partly peeled off as two distinct membranes, the innermost of which is the amnion already mentioned, and the outer is the chorion united with the decidua. The amnion and chorion are appendages of the embryo; the de- OUTLINE OF HUMAN DEVELOPMENT. 29 cidua is the modified mucous membrane of the uterus. The inner portion of a microscopical section through the uterine wall is shown in Fig. 15. The amnion, a7n, consists of two layers, a cubical- celled epithelium facing the emhryo, and a connective-tissue stratum facing the uterus. The chorion, Cho, is likewise two-layered, but ■ ""0 " its epithelium, c, is next the uterus, its connective tissue next the amnion; the amnion and chorion are loosely held together by shreds and bands crossing from one membrane to the other. The decidua occupies the whole space between the chorion, Cho, and muscularis, muse; it contains blood-vessels, v, and remnants, gl, of gland cavi- ties. Let us return to the embryo. From its abdomen there sprmgs 30 INTRODUCTION. a long, whitish cord, known as the umbilical cord ; it is usually about one-third to one-half an inch in diameter and -10 cm. long, but its dimensions are extremely variable ; it always shows a spiral twist, and contains three large blood-vessels, two arteries, and one vein, all of which can be distinguished through the translucent tissue. The distal end of the cord is attached to the wall of the uterus, usu- ally near the middle of the dorsal side of the organ. It is easily seen that the blood-vessels of the umbilical cord radiate out from its end over the surface of the iiterus underneath the amnion, branching as they go; they spread, however, only over a circumscribed area, the placental, where the wall of the uterus is very much thickened. A vertical section through the placental area shows that the amnion and chorion are widely separated from the decidua and muscularis by a spongy mass soaked with maternal blood. This mass consists of numerous trees of tissue, which spring with comparatively thick stems from the chorion and branch again and again. In these stems and branches are to be found the final ramifications of the vessels of the umbilical cord ; the trees are known as chorionic or placental villi. Some of their end-twigs are very closely attached to the sur- face of the decidua. In the centre of the placental area the villi form a mass about three-fourths of an inch thick, but toward the edge of the area the mass gradually thins out until at the very edge the chorion and decidua come into immediate contact. The mass of villi, together with the overlying portions of the chorionic and am- niotic membranes and the underlying portion of the decidua, consti- tutes what is known as the placenta. The decidua of the placental area is called the decidua serotina ; the chorion of the placenta is known as the chorion frondosum. When birth takes place the whole placenta is expelled, after the delivery of the child ; the placenta of the obstetrician is therefore partly of foetal, partly of maternal, origin. Uterus Three Months Pregnant. — -The uterus measures about 3A inches in transverse diameter, and shows well-marked inlaid sinuses on its external surface. If it is opened, as before, bj' a cru- cial incision on the anterior side, its walls will be found about half an inch or more in thickness ; it contains a grayish-red bag (decidua reflexa) , which nearly fills the cavity of the uterus and incloses the embryo, so that upon opening the womb we do not encounter the foetus directly. The inner bag has a smooth surface, but shows a few small pores ; it is without blood-vessels and is attached to the dorsal wall of the uterus. The inner surface of the uterus shows a rich net- work of blood-vessels, many of which are large, irregular sinuses. The walls are seen to consist of an outer muscular layer, and an inner decidual layer, which takes up nearly half the thickness of the wall, and is known as the decidua vera. As compared with the eighth-month uterus the proportion of the layers shows us that dur- ing gestation the muscular layer increases and the decidual layer diminishes in thickness. The inner bag when opened shows the large cavity in which the embryo lies, floating in amniotic fluid. The bag is formed by three very distinct membranes, of which the outermost, decidua reflexa, is the thickest and opaque ; the two inner ones are thin and transparent; the innermost is the delicate amnion; the middle membrane is the chorion and is quite distinct from both OUTLINE OF HUMAN DEVELOPMENT. 31 the amnion and reflexa ; with the latter it is connected by a number of small branching villi scattered at some distance from one another over the surface ; the villi adhere firmly to the reflexa by their tips. The embryo resembles a child in its general appearance ; the length of the head and rump together is about eight centim.etres, and the head is approximately of equal bulk to the rump. The timbilical cord is 5-7 mm. in diameter and usually about 13 centimetres long. From its distal end the blood-vessels spread out over the placental area, and around the edge of the area rises the decidua reflexa, which does not extend on to the placenta. Floating in the amniotic fluid is a pear-shaped vesicle, the yolk-sack, which is about 8 mm. in diameter ; it has a fine network of blood-vessels upon its surface, and is connected at its pointed end with a long slender pedicle, the yolk- stalk, which runs to the placental end of the umbilical cord, there enters the cord itself, and runs through its entire length to its attach- ment to one of the coils of the intestine of the embryo. Over the whole of the placental area the chorion gives off large villous trunks, each of which has numerous branches, with ramifications of the foetal vessels ; the villi fill a space about one centimetre wide between the membrane of the chorion frondosum and the surface of the uter- ine decidua serotina, to which the tips of some of the villi are at- tached. With care the villi may be separated from the decidua, which is seen, when it is thus uncovered, to be cavernous ; the cav- erns are rounded in form and may be followed on the one hand until they connect with the blood sinvises of the uterus, and on the other until they open into the intervillous spaces, which therefore receive a direct supply of blood from the mother. The principal difference to be noted in the relations of parts be- tween the uterus before and that after the fifth month is in the pres- ence or absence of the decidua reflexa as a distinct membrane. Dur- ing the fourth month the reflexa stretches as the membranes expand and becomes thinner and thinner until by the end of the fourth month it is as delicate and transparent as the chorion and lies close against the wall of the uterus (decidua vera) . It is probable that the decidua reflexa degenerates and is resorbed, compare p. 19. Uterus Five Weeks Pregnant. — The relations are best shown by a median antero-posterior section. Fig. 4. The arrangement of the uterine parts is essentially the same a^ at three months. The mucosa uteri is changed into the decidua graviditatis. On the dor- sal side from s to s is the decidua serotina of the placental area, where the villi of the chorion are fastened by their tips to the uterus. From the edge of the placental area on all sides rises the decidua reflexa, r r, which is much thinner than the other parts of the de- cidua, and which forms a closed dome over the embryo ; hence when we pass through the cervix uteri, c, we enter, not the cavity contain- ing the ovum, but the fissure-like space between reflexa, r r, and the vera, g g, which includes the whole of the modified mucosa of the body of the uterus, except that part to which the ovum is at- tached and which produces the reflexa and serotina. The vera is that portion of the decidua which is not in direct contact with the ovum. No stage of gestation earlier than the completed formation of the reflexa has been observed. 32 INTRODUCTION. The embryo difEers greatly from the three months' foetus. Be- ginning with the envelopes we notice that the chorion is beset with well- developed villi over its entire surface, but the villi over the pla- cental area are larger than those over the parts adjacent to the de- cidua reflexa. The amnion does not lie close to the chorion, but close around the embryo, leaving a wide space between the two mem- branes, which space, as we have seen, is subsequently obliterated by the expansion of the amnion. The embryo itself is very small and not human in appearance, and its organs are only partially differen- tiated. The umbilical cord is very short ; the amnion springs from it at a short distance from the embryo. The yolk-stalk leaves the cord just beyond the amnion, is comparatively short, and ends in the pear-shaped yolk-sack, which is about the same size as at three months. Beyond the point where the amnion and yolk-stalk part from it, the umbilical cord continues a short distance with its blood- vessels, which ramify over the entire chorion and penetrate all the villi thereof. To produce the relations found at three months the blood-vessels and villi of the chorion must abort except over the pla- cental area ; the umbilical cord must elongate greatly ; the amnion must expand until it touches the chorion, and the foetus must grow ' and change. We must now trace back the his- I tory of the embryo still farther, in ^, ^ order to understand the relation of the embryo to the embryonic mem- branes and appendages. Ovum of Three Weeks.* — Ifor- mal human ova of this age very rarely indeed reach the embryolo- gists, but a few have been described. The chorion forms a closed vesicle beset on all sides with crowded, clumsily-branching villi; the vesicle measures about 3 cm. in diameter; the villi are about 3 mm. long, and as yet show no regional inequality in their development. If the vesicle is opened the embryo is found within rolled up, the back being convex ; it measures in its natural attitude about 44 mm. The head is bent toward the right; the caudal extremity to- ward the left ; the head and tail are almost in contact, so that it is diffi- cult to observe the insertion of the umbilical cord. With care this may be done, and it will then be seen that the amnion arises from the embryo, and is, in fact, a prolongation of the body-wall; the amnion itself is extremely thin and lies close about the embryo. The umbilical * For figures see Chapter XIII. 7 Fia. 16. — Human embryo, 4.2 mm. long (His' Lr). After W. His. Explanation in text. OUTLINE OF HUMAN DEVELOPMENT. 33 cord* unites with the abdomen; in front of it, i.e., headward, is a small opening through which the stalk of the yolk-sack enters the body to unite with the intestine ; it is from the edges of this opening that the amnion arises, and as the amnion passes around the umbili- cal cord, it may be said that the cord and the yolk-stalk both enter the body through the opening, but whereas the cord is in contact with the amnion the yolk-stalk is not. The opening may be called the umbilical foramen. The yolk-sack is pear-shaped, measures about 3 mm. in diameter, and is attached by its pointed end to a loop of the embryonic intestine. The yolk-stalk is developed by the subsequent prolongation of the pointed end of the sack. In an embryo a little younger the relations can be more clearly recognized. Fig. 16. The embryo is nearly straight, although both head and tail are bent over ventrally. The umbilical foramen, from the edge of which the amnion arises, is very wide and long ; at its /' s™ - ,,\^ tailward edge runs out the umbilical / V ■ ''.yf-r-j/J , \ cord (Bauchstiel) , to which the am- ,i j^''i-»>.lji nion is attached, and which joins k?;iA'<5:. / ■?v^"'t^'?,'';^'v\XV- the chorion a short distance from jA^i;;- 'SH | / Z**^;' ',;;" S ?^!i, the embryo^ The neck of the yolk- y";;' '^v^'t;; fcsSfEJ'''-;^ sack, FTcs, is also much wider; if \ ?/-/'';;---'siJ'-^'""'&il^A^^ the sack is cut open we find its neck to \.-'S;'ut]'"-\''^1\ ^:;i-,'"i/^jft. 'M be a large opening into the cavity of /^'*■;isSfef:B?i|?^'^ the intestine; in fact, the yolk-sack is A'}$?^%J:S:' ''jl*--^ ,:-;;■-: v' 'J^jl an appendage of the intestinal canal, -^xK °'f:i'-ii'f '^^''^'''W^i !::fiS$P^ which at this stage is very simple, !:^-f f'lfV^;' ;|§ "s^iSgSis* being hardly more than a straight \v*!^:;^;' i(?V;| tube running lengthwise; the open- ing between the sack and intestine may be called the vitelline foramen. " <><• ' 'rj" ' - "" The younger the embryo the vm. i7.-Embryo, a.is mm long. After longer are — relatively to the size ot the embryo — the foramen umbilicale and the foramen vitellinum, as is well shown in Fig. IT. The line of attachment of the amnion ex- tends almost the entire length of the embryo, beginning just in front of the heart, and ending upon the umbilical cord {Bauch- stiel or allantois-stalk), close to the chorion. The yolk-sack has also a long attachment, beginning just behind the heart, and extend- ing nearly to the allantois stalk, which now appears to the eye very much what it is morphologically, a prolongation of the posterior ex- tremity of the body of the embryo. Going back still farther, we find the relations to be as represented by the accompanying diagram. Fig. 18. The embryo, Emb, rests upon the yolk-sack, and is scarcely longer than the umbilical fora- men ; the end of the embryo is prolonged posteriorly as the relatively large allantois-stalk, Al, by which the embryo is attached to the chorion. The amnion springs from the sides of the embryo and of its allantois stalk, and forms a closed sack over the embryo. This stage is almost the youngest in the series of known human embryos, and has been only imperfectly d escribed. * At this stage more properly to lie called the BauclistieU see Chapter XVI. 3 34 INTRODUCTION. The following generalized diagram, Fig. 19, of a young ammote vertebrate embryo is intended to render clear the essential relations of the embryo and its appendages. The figure represents a trans- verse section of the embryo, together with all the membranes. The embryo consists of an axial mass, from which runs out on each side a lamina or plate of tissue, Som, to form the body-wall ; this plate extends beyond the embryo to form the amnion. Am; as the plate from one side joins that from the other, the amnion makes a closed sack over the back of the embryo. From the axial mass there run out two other plates, 8pl, to form the walls of the intestinal canal. In; these plates are likewise prolonged beyond the body to form the large yolk-sack. Yolk, upon the top of which the em- Fig. J8. — Diagram of an embryo of fifteen to sixteen days. Fio. 19.— Generalized diagram of an amniote vertebrate embryo, bryo rests. The space between the walls of the intestine and the body- walls is of course the bodj^-cavity, Coe. Where the body-wall, Som, passes over into the amnion, Am, there is an opening by which the body-cavity communicates directly with the space between the amnion and yolk-sack on one side and the chorion on the other; this opening is the umbilical foramen. Similarly there is a passage by which the cavity of the intestine. In, communicates with that of the yolk-sack. Yolk; this passage is the vitelline foramen. For our conceptions of the probable history of the human ovum up to the fourteenth day, we must rely mainly on analogy, drawn from our knowledge of the development of other mammals and of birds and reptiles. From these sources we learn that the amnion and chorion are originally portions of the same membrane, which is an extension of the body- wall of the embryo. In reality the differen- tiation of the amnion is quite a complex process, as is shown by the detailed history given in Chapter XV. The essential steps can be made clear, however, by a brief description. Fig. 20 is a diagram of a stage in the development of amniota a little earlier than that shown in Fig. 19. Both the vitelline and umbilical foramens are much wider than in the preceding figure. The body-wall of the embryo, Som, passes over as before into the amnion, Am, but the amnion of one side does not join that of the other, but instead bends OUTLINE OF HUMAN DEVELOPMENT. 35 amnion over and is continuous with the chorion, Gho. Thus the ^.^....j,, and chorion conjointly form a fold on each side of the embrvo ■ if the two folds enlarge and arch over the embryo until they meet and unite by their edges the condition illustrated by the preceding dia- gram, Fig. 19, will be established. Eeturning to the earlier condition. Fig. 20, we may say that the ovum consists of two closed vesicles S"'^^, united together by the axial mass of the embrj'o. The membrane, which forms the outer vesicle, is subdivided into three principal re- gions, to wit : the body- wall of the embryo, the amnion, the chorion, each having its separate history. The membrane which forms the inner vesicle is subdivided into two principal regions, to wit : the wall of the intestine and the wall of the yolk-sack, each having its separate history. It will be re- membered that the posterior end of the embryo is prolonged as the allantois-stalk, by means of which it remains permanently and directly united with the chorion. _ It is unnecessary, for our present purpose, to follow back .the earlier history step by step. Suffice it to say that in younger stages the two vesicles are represented only by one, and earlier yet there is merely a cluster of cells. The stages of development preceding this are not to be found in the uterus, but in the Fallopian tubes. They exhibit to us merely an agglomeration of a few cells, the so-called segmented ovum. The earlier the stage the fewer the cells, until we reach the condition when there are but few cells, then two, and finally one only. This cell is the impregnated ovum, the beginning of all development, but is itself formed of two separate parts, very different in their origin and constitution, namely, the egg-cell or ovum and the sper- matozoon, whose union is the act of impregnation — the beginning of a new existence. Fig. 20.— Generalized diagram of an amnioto vertebrate embryo before tne separation of the amnion, Am^ and chorion, Cho. II. Progressive History of the Fcetus and its Envelopes. The ovum enters the upper end of the Fallopian tube, and is there impregnated.* Very slowly it moves down the Fallopian tube, undergoing meanwhile the process of so-called segmentation, by which it is separated into a gradually increasing number of cells, that arrange themselves so as to begin the formation of the embryo and its appendages. Probably about the eighth day the ovum reaches the uterus, where it becomes adherent to the mucosa upon * It is possible that impregnation may occur while the ovum is passing from the ovary to the fimbriate opening of the Faflopian tube. 36 ' INTRODUCTION. the dorsal side of the uterus usually, and by an unknown process of agglutination. The decidua reflexa grows up around it by a pro- cess not yet observed. The amnion is differentiated from the cho- rion. The portion of the mucosa uteri in contact with the ovum is transformed into the decidua serotina ; the remaining portion of the mucosa becomes the decidua vera. The allantois-stalk unites the embryo with the chorion, and carries the blood-vessels of the foetus to ramify upon the chorion. The embryo is enclosed by the amnion ; the amnion is enclosed by the villous chorion ; the chorion is enclosed by the decidua reflexa and serotina. The vesicle formed by the close adherence of the chorion to the reflexa is suspended from the wall of the uterus. The mass of tissue resulting from the union of the chorion with the serotina forms the placenta. The umbilical cord (allantois-stalk) is always attached to the placental area, and later the ramifications of the umbilical vessels are restricted to that area. During the fifth month the decidua reflexa coalesces with the decidua vera, and the space between them is of course obliterated. Finally, we find that the amnion enlarges, lays itself against the chorion, and, uniting loosely with it, becomes the innermost constit- uent of the vesicle enclosing the embryo. PART I. THE GENITAL PRODUCTS. CHAPTEE III. THE HISTORY OF THE GENOBLASTS AND THE THEOEY OF SEX. The term genoblast is used to designate the sexual elements. I ^pply it exclusively to sexual elements proper, and not to the acces- sory parts with which those elements are associated. The spermato- zoon is a genoblast; a spermatophore is not. The egg-cell after maturation is a genoblast, but not before. I. Spermatozoa. 1. Summary. — The spermatozoa of mammals are filaments con- sisting of a short, thick end called the head, and a very long and delicate thread called the tail. The head varies greatly in shape, according to the species ; in man it is broad and thin. Fig. 23, and is widest at a little distance from the tail. The head contains chro- matin, and may be colored by the usual nuclear dyes. The tail consists of three parts: 1, the middle-piece, which is next the head, and the thickest of the three parts ; it contains an axial thread, and probably always has a very fine spiral thread running round it ; 2, the main-piece; and, 3, the end-piece, which is not more than a line, even as seen with very high magnifying powers. The human spermatozoon is 0.055 mm. long — the head being 0.005 mm., the tail 0.050, and the middle-piece 0.009. The development of the mammalian spermatozoa begins with a so- called parent or mother-cell, which lies near the outer wall of the seminiferous tubule. The mother-cell produces a number of daugh- ter-cells, which also multiply by division ; the daughter-cells break down, forming a column of matter (protoplasm) , in which lie their nuclei, and at the base of which lies the nucleus of the mother-cell; the nucleus of the mother-ceU and the column of matter both ulti- mately disappear, but exactly how is not determined ; the nuclei of the daughter-cells produce each a spermatozoon. The head and tail of the future spermatozoon become visible within the nuclear mem- brane ; the head is formed chiefly by the chromatin of the nucleus ; the nuclear membrane finally ruptures, and it as well as the contents of the nucleus which have not taken part in the formation of the spermatozoon are lost. Among the lost parts is a special round body of small size, which appears in the nucleus while the spermatozoon is developing ; this body may be stained by chloride of gold, but not b}' haematoxylin ; its significance is unknown. The long column holding the spermatozoa together has usually been regarded as a cell, and is the supporting cell auct., or Sertoli's column. 2. Sperm.atozoa are the essential fertilizing elements secreted by the male gland. They are minute bodies, capable of active loco- 40 THE GENITAL PRODUCTS. motion, and having a characteristic form in each species. In a few instances (certain snails, etc.) there are two distinct forms of sper- matozoon for a single species, but usually there is only one form, and that little variable. In a small number of animals the sper- matozoa, as in the nematods, are distinctly cell-like ; but in the great majority of animals, and, so far as I know, in all vertebrates, they are long and thread-like ; hence their common German name, Samen- fdden, first proposed, I think, by Kolliker. The mammalian spermatozoa are long, slender bodies, varying con- siderably in configuration, but all presenting at least the following features in common : One end is thickened and is called the head ; it has a strong affinity for nuclear staining fluids ; this afiBnity musti be attributed to the chromatin, which the head contains, as is shown by the history of its development ; the remainder of the spermatozoon is long and slender, and constitutes the tail; the tail consists of — 1, a middle part [Mittelstiick) , a little thicker than the rest, and sit- uated next to the head ; the middle part is traversed by a very fine axial thread, and ends abruptly; and, 2, a hind-piece, which, accord- ing to some writers, may be subdivided naturally into two segments, the main-piece (Hauptstuck) and end-piece. The spermatozoa of the various species differ in size in the pro- portions of the parts, and often very strikingly in the shape and structure of the head ; those of the opossum are especially remark- able for being double ; two apparently complete spermatozoa being united to a common plate by their heads (Selenka : " Studien liber Entwickelungsgeschichte," Heft IV., p. lOG). Twin spermatozoa have also been observed in the rat by Neumann, 75.1, 313, Taf. XVII., Fig. 16, b. Compare also Max von Brunn, 84.1, and Brock, 87.5. The largest known mammalian spermatozoon is perhaps that of the marsupial, Phascogale; the spermatozoon of this animal is 0.263 mm. long — the head, however, being only 0.013 mm. (Fiirst, 87.1, 354). The spermatozoon of the rat is 0.144 mm. long, the head 0.009, the tail 0.135, and the middle-piece 0.045 mm. La Vallette, 71.1, gives a synopsis concerning the forms of verte- brate spermatozoa nearly as follows : Fish : The spermatozoa of Am- phioxus are threads with round heads. In Petromyzon the head is rod-like or egg-shaped. The teleosts generally have pin-like sper- matozoa ; but in the salmonidse (Owsjannikow) the head is pointed and shaped like a heart-tip. The spermatozoa of selachians are much larger, with the head-end spindle-shaped and often spirally twisted. Amphibia : The head is long, generally pointed, the mid- dle-piece short, and the tail is often provided with an undulatory membrane (Retzius, 81.1). Reptiles and birds: The head is usu- ally long, often twisted. Mammals : The head is more or less elon- gated ; in ungulates the head is flattened and usually more or less egg-shaped in outline, the pointed end toward the tail. Among rodents there is considerable variety of form. In the dog the head is pear-shaped ; in the hedgehog the head is truncated inferiorly, and the tail is inserted laterally. No comprehensive summary of the observed forms of spermatozoa has been made since the publication of Wagner and Leuckart's article in "Todd's Cyclopaedia." SPERMATOZOA. 41 I i 5f B The most minutely studied mammalian spermatozoon is that of the rat? thanks especially to the patience of 0. S. Jensen, whose posthumous paper, 87.1, furnishes the basis of the ensuing de- scription. The rat's spermatozoon measures 144 m; its head, Fig. 21, C, is a broad hook, pointed at one end and obliquely truncated at the other ; from one corner of the truncated end starts the very long slender tail, which is divisible into the thicker middle-piece {Mittelstilck, or Jensen's Verbindungsstilck) and the thinner main-piece (Hauptstiick), Fig. 21, A, which terminates in a short and still finer thread called the end-piece {Endstiick) . The appearance of the spermatozoon varies according to its degree of development, it not attaining full maturity until it has left the seminiferous tubule. The changes referred to affect principally the head and the middle- piece. The head is covered, while the spermato- zoon remains in the seminiferous tubules, by a membranous cap, Fig. 21, A, which subsequently disappears. The middle-piece has a spiral thread running round its outside. Fig. 21, B. The spiral thread appears soon after the rupture of the nu- clear membrane, by which the developing sper- matozoon is set free (c/. infra). The thread is at first indistinct and makes only a few turns ; it rapidly becomes more distinct and the number of turns increases, until they become so numerovis that in a spermatozoon taken from the vas defer- ens only a series of thick-set cross-lines can be distinguished ; these lines have been seen by sev- eral observers and variously interpreted ; the spiral may run to the right or to the left. The thread becomes loosened off by the action of glycerin (1 part) and water (4 parts) , and is destroyed in one to two hours by 0.6 per cent salt solution, leaving then the axis uncovered. The thread can be stained by chloride of gold, though the axis can- not. The axis, when the spermatozoa are treated with acetic acid, often breaks up into threads (c/. Ballowitz, 86. 1) ; it shows a lighter line in its centre. These observations lead Jensen to the conclusion that the axis is formed by a wall of fibrillse surrounding a central core or cavity. The axis does not reach quite to the head, but ends with a little knob, leaving a small, perfectly transparent space between the knob and the head, Fig. 21, C. In some spermatozoa— e. gf., of horse and ox— though not in those of the rat, there -is a minute opening in the head called the microporus, and situated just opposite the knob of the axis. When the spermatozoa are stained with nuclear dyes, most of the head is colored, but the tip of the hook, which contains no chromatin, and is probably formed out of a scrap of the nuclear Fie. Structure of a rat's spermato- zoon. B, young sper- matozoon, end of the middle-piece and be- ginning of the main- piece to show the spiral thread — greatly magnified ; A, head, with part of the axial thread ; C, immature spermatozoon, a n t e- rior half only. After O. S. Jensen. 42 THE GENITAL PRODUCTS. membrane, remains uncolored : on the concave side of the tip a fine line can be distinguished, due, apparently, to a rod of substance. Sometimes a minute fragment of the nuclear membrane is left ad- herent to the lower end of the middle-piece; for the explanation of this possibility, compare the section below on de- velopment. The human spermatozoa are described by Retzius, 81.1, 85, as follows : The head, seen from the fiat side, appears oval. Fig. 22, A, with the front end generally tapering a little, but never pointed; the anterior half or two-thirds has a brighter and more transparent part. Seen on edge, Fig. 33, B, the head has a pointed form, with a posterior thicker, round, dark part. By adjustment of the focus it can be ascertained that the sides near the point are de- pressed somewhat like those of red blood corpuscles. Retzius could nowise succeed in demonstrating a special tip (Spiess) corresponding to that in the sala- mander, but Edw. M. Nelson {Journ. Quekett Club, 1889, III., 310) has observed a slender thread pro- longed from the head, and also a hook at the end of the thread ; these observations have been confirmed by Bardeleioen, 91.1. The latter also describes ad- ditional details of the structure of the head. The following piece (Schweigger-Seidel's Mittelstiick) is directly united with the head by a transverse joint ; there is no neck in Elmer's sense; the middle-piece is cylindrical and relatively small:^about as long, or a little longer, than the head; its surface is often granular or rough, and there cling to it a few shreds of protoplasm, as has been described by several in- vestigators ; the spiral thread was long overlooked, but has been recognized and figured by K. Bardele- ben, 91.1. The undulatory membrane, supposed by Gibbes, 79.1, and W. Krause, 81.4, to be present, was perhaps an abnormally loosened spiral thread. The main-piece of the tail is about half as thick as the "Mittelstiick," gradually tapers, and ends abruptly at the beginning of the still finer and very short end-piece. 3. Spermatogenesis. — The seminiferous tubules are cylindrical, i. e. , in cross-sections they appear round ; a large part of the tubule is filled with spermatozoa in various stages of development. The outer boundary is marked by a distinct line corresponding to the tunica propria, a layer of endothelial cells, with fiat oval nuclei (Neumann, 75.1, 306) . Next to the tunica comes a layer which, as far as known, presents pretty much the same appearances, whatever may be the stage of development of the spermatozoa within. This layer contains two kinds of cells : First, the large Sertoli's columns, as -they may be called, after their discoverer.* These cells are identical with Merkel's Stiitzzellen, La Vallette's spermatogonien, Swaen and Mas- quelin's cellules folliculaires. Second, small granular cells, vary- FiG. 23.— Humau spermatozoa. A, complete sperma- tozoon ; B, head seen from the side ; O, extremity of the tail. All highly magnified. After Retzius. * First described by Sertoli in ii. Morgagni (c/. Henle's Jahresberichte for 1864, p. 120). pare Sertoli, Arch, Sci. mediohe, ii., 107 (1877). Com- SPERMATOZOA. ing in appearance according to the exact stage of their develop- ment. Examined in surface views, Fig. 23 (compare also Figs. 5, 6, and 41 of Fiirst's paper, 87.1), the large cells are seen to be mostly hexagonal in outline, to touch one another, and to pass below, /. e. outside, the small cells; they have large, clear, oval nuclei with sharp outlines, and usually a single well-marked nu- cleolus. The nuclei lie quite near the tunica propria, but in man lie farther inward, and are in this case not so near the tunica as are the small cells. Around the nucleus there lie a few highly refractile granules which may be stained by arsenic acid, and are probably fat. The small cells lie in depressions or cups of the large cells. Fig. 23, B, and when the small cells are knocked out — as sometimes happens in teasing — -the partitions between the cups appear more distinctly and create a network figure, which formerly misled Von Ebner and others into describing a real network as constituting the layer. The large cells also have long columnar prolongations, as can be best seen in transverse sections of the tubules. Fig. 29 ; the prolongations are united with bundles of developing spermatoblasts. The small cells are very different ; they lie over the outlines of the large cells and between their centripetal prolongations. Fig. 29; they are gran- ular, have comparatively little protoplasm, and their nuclei are nearly spherical in shape. The nuclei vary considerably in appear- ance, as these cells multiply by indirect division ; usually they con- tain a chromatin network or a coiled chromatin cord ; sometimes the network is concentrated at one side of the nucleus, leaving the other side comparatively clear. At certain periods the nuclei are found in various stages of karyokinesis. The cells resulting from the division of the small cells form the packing between the inward columns of the large cells, hence in cross-sections we get alternating columns, Fig. 29. The descendants of the small cells pro- duce the spermato- blasts, and the sper- matoblasts are con- verted into the sper- matozoa. The small cells are then the parents of the sper- matozoa and may be called the parent - cells ; a great variety of names have been employed to designate them, such as mother-cells, spore-cells, ger- minative cells, Samenstammzellen, etc. The nomenclature of the small cells is very confused ; those of them in process of indirect division are often smaller than the others and have been designated as the "growing cells" by H. H. Brown, 85.1, and this term has been used by other writers since. The small cells in the resting stage are called " Stammzellen " by most German writers, as an equiv- alent for which I have ado-pted pare) d-cell. _ Formation of the Spermatoblasts. — The parent-cells divide Fig. 23.— Peripheral layer o( t'he seminiferous tubule of a rat. Two views from a teased preparation. After Neumann. 44 THE GENITAL PRODUCTS. and produce probably tbree cells, although the number has never been accurately ascertained. One cell remains as a parent-cell, and the other two are the mother-cells (Mutterzellen) and are well char- acterized by their appearance. According to Biondi, 85.1, the nucleus of the parent-cell remains and becomes like the nucleus of the large cells (Sertoli's or supporting cells). The mother-cells divide and their descendants also divide until there is produced a column of cells, Fig. 34, which stretches in a radial line from the mother- cell toward the centre of the tubule, and is packed in between the columnar centripetal prolongations of Sertoli's cells (c/. Figs. 34 and 39) . Probably, then, although investigators are not agreed in regard to this point, the parent-cells divide in such a way that the cells re- sulting from the division are unlike, one of them preserving the character of the parent-cell, and the others differing from it in hav- ing a relatively larger nucleus and a finer chro- matin network; the appearance of the nuclei varies, of course, according as they are in the resting or divisional (kinetic) phase.* The cell most like the original one, and which we may call still the parent-cell, lies at the outer edge of the tubule, while the others or mother-ceUs lie toward the centre, Fig. 24. The parent-cell, as already stated, produces at least a second and perhaps more mother-cells, so that the column grows centripetaily. The column also grows by multiplication of the mother-cells, but the cells thus formed lie in the innermost part of the col- umn; they are smaller, Fig. 34, than the first generation of (mother) cells ; they have relatively large nuclei, with the chromatin gathered into two or three spots — nucleoli. We thus have a column of cells in which we can distinguish three zones: 1, the outer zone of the parent-cell ; 3, the middle zone of the mother-cells ; 3, the inner zone of the daughter-cells. These zones remain more or less marked for a considerable period ; for, as the cells of the inner zone change into spermato- ■ blasts, those of the middle zone change into second daughter-cells, and as the inner spermatoblasts change into spermatozoa the cells of the second zone change into spermatoblasts; the innermost zone long continues one stage ahead. The trizonal arrangement is very conspicuous in cross-sections. The division of the mother and daughter-cells presents many pe- culiarities, and does not conform exactly to Flemming's well-known scheme of phases for indirect division. Attention was first directed to these peculiarities byCarnoy, in an important memoir, 85.1, and W. Flemming, 87. 1, has since confirmed these discoveries, in large part, by observations on the salamander, and gives a plate of dia- grams which is instructive as a facile means of comparison. La Fig. 24.— Column of spermatocytes from the rat ; a, parent cell ; 6, mother cells. After Binodi. X 600 diams. *For figures of the karyokinetic division of the daughter-cells, see Ftirst, 87. 1, Figs. 10-13. SPERMATOZOA. 45 Vallette, Niessing, 88. 1, p. 44, and others find that when the mother- cells multiply there is often a stage to be found where several nuclei (two to twelve) lie within one large cell. The multinucleate giant- cells are best found by teasing the fresh specimen. As to their place in the spermatogenetic history we possess no definite knowledge. The spermatoblasts arise from the nuclei of the daughter-cells (spermatocytes), and not as H. H. Brown, 85.1, and many others have, I think, erroneous^ be- % 111/ \ ( \ \ / i h lieved, each out of a whole cell. ^ - \ Biondi, 85.1, seems to me right ^j^ I in his statement that the bodies of the cells break down, or at any rate lose their boundaries, thus creating a granular proto- plasmatic column in which the nuclei lie. Compare also Nies- sing, 88.1. The protoplasm of the parent-cell participates in these changes, hence its nucleus comes to lie at the base of the column. This nucleus has meanwhile altered its charac- ter, and become large, clear, and nucleolated. Now, these columns are the same as the large Sertoli's or supporting cells above described. By no means all writers agree with this account of the origin of Sertoli's cells, but all other ex- planations that I have found appear to me vague and con- fused, and the history of the changes here advocated is clear, and accounts for the well-estab- lished grouping of the sperma- toblasts in the substance of Sertoli's column ; this essential phase is explained satisfactorily by no other theory. The nuclei congregate at the inner end of the column, and there change their character and become recognizable spermato- blasts. Figs. 25 and 29. Development of the Spermatoblasts into Spermatozoa.— The nuclei change iato spermatozoa as follows: The chromatin is at first unequally distributed throughout the nucleus; it then in great part accumulates at the end of the nucleus toward the outer wall of the tubule ; at this stage the chromatin is densest near the equator of the nucleus, where the edge of the chromatin is sharply marked, and toward the outer pole of the nucleus the chromatin is i 9 Fig. as. —Developing spermaioblasts of the rat: a, 6, c, d, «■, /, g, h, successive stages. X about 750 diameters. After H. H. Brown. 46 THE GENITAL, PRODUCTS. |) B less condensed (Niessing, 88.1, p. 40, Taf. I., Figs. 6, 7, and 8). It is from the equatorial plate that the future tail grows out at the start. Particles of the chromatin are said to remain in. other re- gions of the nucleus, and finally to gather together to form the small accessory corpuscle mentioned below. According to Platner, 89.2, 131, 132, the portion of the nucleus which forms the head of the spermatozoon in pulmonate snails is homologous with his Nebenkern. The main mass of the chro- matin is concerned in the formation of the head of the spermatozoon ; it is at first quite round, Fig. 25, a and fo, but soon begins to alter its shape, gradually assuming the form of the spermatozoon head, Fig. 25, c, d, e, /. The tail appears very early as a delicate fila- ment, growing out from the chromatin and lying entirely within the nucleus, Fig. 25, a, but shortly after is found to project beyond the nuclear membrane, h, and lengthens rap- idly, e, /, g. The nuclear membrane is very distinct; it elongates into an oval bag, b, c, one end of which lies close against the chro- matin, while the other surrounds part of the tail and is wide ; the lengthening continues, 6, /, Q, with accompanying changes of form, best indicated by the figures; the part of the tail within the nuclear membrane becomes the middle-piece. Fig. 26, but the spiral thread is not developed until later. The accessory body may be readily seen in the rat ; unlike the chromatin of the head it can be stained by chloride of gold : hence, if it is formed of chromatin at all, the chro- matin must have undergone alteration. Finally, the nuclear mem- brane ruptures, Fig. 27, a portion of the membrane remains upon the head, and the caudal bag sometimes endures longer, Fig. 25, g, but at last also disappears, except that in certain cases a trace of it remains visible as a fine cross-line at the end of the middle- -DevelopiiiK sperma- ial: Fig. 20. - -, - tozoa of a marsupial: Meta chirus Qiiica. A, B, C, differ ent stages. After Furst. piece. Fiirst and others think that the axis ' f V of the tail is formed from the chromatin, and that the sheath of the axis arises from the achromatine substance of the nucleus (caryoplasma) . After the rupture of the nuclear membrane the young spermatozoa still develop a little farther. The spermato- fig. 27. -Humaa spermatoblasts, 10 zoa are ultimately liberated, and, falling Sterwfeto-JX"'"^ ''**'' "''''"'''■^°''' into the lumen of the tubule, pass off. From their mode of development, it is evident that the sperma- tozoa necessarily lie in bundles, each bundle being held together by a Sertoli's column, Fig. 28; at first they lie at the inner end of the SPERMATOZOA. 47 X^ FlG.28. — Sertoli's col- umn, with a basal nu- cleolated nucleus and a cluster of developing spermatoblasts. After H. H. Brown. tion have been reason to suspect that the sperma- tozoa are oviparous; they are also stated to propagate b}- spontaneous fission, the sep aration taking place between the disc of the body and the caudal appendage, each of which develop the part required to form a perfect Whole." Meanwhile the inves- tigations of Spallanzani, Wagner, Czermak, and many others gradually increased the knowledge of the forms of the sper- matozoa. Dujardin was the first to consider the column, at a considerable distance from the basal nucleus, but as the nuclei (spermatoblasts) length- en, the heads push their way toward the base of the column, Fig. 39. Now as the development of the daughter-cells (spermatocytes) is continually progressing between Sertoli's columns, we obtain in sections the long-known, remarkable appear- ances shown in Fig. 20, of bundles of spermatozoa alternating with columns of proliferating cells. 4. Historical. — The seminal animalcules were, it is stated, first discovered by Ludwig Hamm, then a student at Leyden, in August, 1677. Leeuwenhoek claimed the merit of having made the discovery in November of the same year, and in 1678 Hartsoeker published an account of them, professing to have seen them as early as 1674. They were long considered to be probably parasites, and it was not until Prevost and Du- mas' researches that it was definitely ascertained that the " animalcules " were the essential fertiliz- ing element. Thus Richard Owen, in his article on "Entozoa" (1836), in Todd's "Cyclopaedia," includes the spermatozoa under that head, although he writes: "It is still undetermined whether they are to be regarded as analogous to the moving filaments of the pollen of plants or as inde- pendent organisms " (Vol. II., p. 412). But just ' after he adds: "Al though no distinct organs of genera - detected, there is Fig. 29. —Part of a cross-section of a seminiferous tubule of a rat. X about 750 diameters. After H. H. Brown. 48 THE GENITAL PRODUCTS. spermatozoa as generated from the inner layer of the seminiferous tubules, and therefore not as parasites. The discovery of the sperma- toblasts or immature spermatozoa by Von Siebold (Miiller's Archiv, 1836 and 1843), soon confirmed by KoUiker and Eeichert, marks an important step. Now follows a series of publications by which one detail after another was added to our knowledge. During the past twenty years there has been rapid progress, which may be said to have begun with Schweigger-Seidel's important memoir, 65.1, and to have made us acquainted with the minute structure of the sper- matozoa, and their development. Another line of investigation was opened by O. Herwig (1875), in following up the history of the sper- matozoon within the ovum after impregnation. For further histor- ical data, see Waldeyer's address, 87.3. II. Ova. Definition.— The term ovum is employed in various senses. It is applied — 1, to the cell distinguished as the ovarian cell, or immature ovum, out of which the female product or mature ovum is developed ; 2, to the mature ovum, or true female spore; 3, to the mature ovum plus the fecundating spermatozoon united with it — that is, to the impregnated ovum; 4, to various stages of development of the embryo. In this article we consider only the ovum in the strict sense — namely, as the female sexual product. Summary. — The ovum arises as a cell, which matures by a series of changes, of which the last and most striking is the expulsion of the so-called polar globules ; there are many important changes which occur earlier. The genesis of the mature ovum may be conveniently divided into three arbitrary stages : (1) Differentiation of the ovic cell ; (3) growth of the cell and accumulation of nutritive material in it ; (3) maturation proper. 1st. The Origin of the Primitive Ova (Ureier or Ovic Cells).* — It seems to me, in the light of the recent investigations of the origin of the ova in vertebrates, safe to assert that they arise from cells of the mesothelium (peritoneal epithelium) covering the genital ridge of the embryo, the ridge giving rise to the adult ovarj. On account of its function the epithelium of the genital ridge has been called the germinal epithelium {KeimepitheV) . In mammals, which alone will be here considered, according to the best authorities the ureier are developed as follows : Certain cells of the germinal epithelium become larger than the others ; these cells are soon carried into the interior of the ovary by being included in cord-like ingrowths of the epithelium. These cords are the Pfliiger'' schen Schlduche of German writers. The primitive ova exist in multiple in the cords, but each of them early becomes surrounded by a separate envelope of epithelial cells. A little later each ovum separates from its neighbors and appears as a round cell, with a clear nucleus and dis- tinct nucleolus, /, closely invested by a layer of cells smaller than itself. The young egg-cell, together with its epithelial envelope, constitutes the so-called primordial follicle (Fig. 31). » For a full discussion of this subject see Chapter XXIU, OVA. 49 2d. General G-rowth of the Ovum and Development of the Yolk. — The modifications which occur in growing egg-cells are as follows : 1st. Change of size ; the cell enlarges, it being a rule — no exception to which is, I believe, known — that the mature egg-cell is much larger than any of the other cells in the body of the parent. 2d. Change of shape ; the cell usually becomes nearly or quite spherical ; the shape of the egg does not necessarily remain spherical, but may be altered by external pressure, as in the uterus of Arion ("Hdbk.," Vol. IV., p. 7, Fig. 1815), or as when several are laid in one capsule (Lumbricus, Nephelis, Planaria, etc.), or when com- pressed by an unyielding shell. An instance of the last-mentioned kind has been described by Repiachoff {Z. f.wiss. Zool., XXX., Suppl.), who figures the egg of a European bryozoon found on eel- grass as fusiform, Fig. 30. 3d. The nucleus becomes larger, spher- ical, and assumes an eccentric position within the cell; the chro- matin usually gathers into one nucleolus, as in mammalia; the nu- cleolus is large, distinct, highly refringent, easily stained, and placed eccentrically within the nucleus. The achromatic substance or pro- toplasm of the nucleus develops into a coarse network, which radi- ates irregularly from the nucleolus as a centre. 4th. The cellular network becomes very distinct ; its interspaces become filled with ovoid, round or crystalline solid inclosures, which are usually, if not always, mainly of an albuminoid character. The inclosures form the part which is called the deutoplasm by Edouard van Beneden and others, fig. 30.— Egg ot Tendra The deutoplasm is the same as the yolk- substance chofl'^''^Magnined'^ Repia- of older writers, and is a store of nutritive mate- rial from which the protoplasm draws subsequently to support its growth. The term yolk has no very exact meaning, for it is used to designate sometimes the deutoplasm alone, sometimes the whole ovum proper, as when the segmentation of the yolk is spoken of. 5th. In all vertebrates an ovarian envelope, the zona radiata, is formed. 6th. It is probable that a vitelline or true cell-membrane is always formed inside the zona by the egg-cell before it reaches maturity. Primordial Ovum. * — In the ovary at birth, and thereafter up to the period of the climacteric, are small egg-cells, some of which develop from time to time into mature ova. At all ages these small egg-cells together with their follicles present a constant appearance. It is currently stated in text-books that there are some seventy thousand egg-cells in the human ovary at birth ; but upon what au- thority this assertion rests I do not know. In any case the number is very large, and it is probable that a good many of them never de-velop, but degenerate. These youngest egg-cells are known as the primordial ova ; they lie in a layer immediately below the albuginea of the ovary and never in the medullary region. Fig. 32. Thej^ are slightly irregular globules, usually 50-60 p- in diameter; 48 by 54 m, 54 by 58 a, 64 by 68 /-/., exemplify actual measurements. The proto- plasm is finelj' and evenly granular, and consists of a uniformly * In the ensuing account ot the ovarian ovum up to its maturation, I have been guided chiefly by Nagel's article, 88. 1 4 50 THE GENITAL PRODUCTS. clear matrix (hyaloplasma) and a fine reticulum, which may be brought out by eosine staining. In birds His, 68.1, has found " protagon " granules in the primordial ovum, and Ed. van Beneden affirms, 70. 1, that yolk-grains are present in the primordial ovum of various mammals; but in man this is not the case. The protoplasm is naked — that is, not enclosed in a cell membrane. The nucleus is round, lies in the centre of the cell, measures from 39 to 32 ij. in di- ameter, is bounded by a very distinct membrane, and contains a round excentrically-placed nucleolus, about 9 //. in diameter. Be- tween the nucleolus and the membrane there is a loose network of fibres, attached to both; the substance of the network is different from that of the nucleolus, as is shown by its different staining. The network was first observed by Flemming, 75.1, in Unio and Anodonta, and has since been often observed in many species ; it was first described in human ova by Trinchese (Mem. Acad. Sci. Bo- logna, Ser. III. , T. VII.) . Some of the primordial ova of veiy young children have no nucleolus, and in bats all of them are at first with- out it according to E. van Beneden. The position of the nucleolus is variable ; it may lie close to the membrane of the nucleus or nearly in the centre. A peculiarity worthy of mention is that once in a great while a primordial ovum has two or even three nuclei. This occurs so very rarely that it cannot be considered as any evidence of multiplication of the ova, but only as an extremely abnormal vari- ation (see Nagel, 88.1, 372-375). Each primordial ovum is sur- rounded by a very thin epithelial envelope. Fig. 31,/, with scattered fusiform nuclei easily distinguished in stained specimens from the similarly shaped nuclei of the neighboring connective tissue. The shape of the follicular nuclei has misled Schron, 63. 1, Foulis, 76.1, Klebs, 63.1, and others into maintaining that the follicle is derived from the stroma-cells, instead of, as is really the case, from the germinal epithelium ; KoUiker traces the origin of the follicular cells to the "' Markstrdnge" ; others, as, notably, Harz, 83.1, and Sabatier, 84.2, derive the follicular cells from the ovum. Both views must, it seems to me, be discarded (compare for details. Chapter XXIII.). The follicle forms a closed wall around the entire ovum, and not one with an opening, as certain authors have main- tained. The primary follicles of mammals were first described by Barry, 38. 1, under the name of ovisacs; his observations were soon confirmed by Bischoff , 42.3. They are now familiarly known to all histologists. Growth of the Ovum and Primary Follicle. — The follicles remain for a long time without change, but from time to time cer- tain ones of them develop. In a mature ovary we can find always several stages. The cause of the development of the follicles is un- known. The primary follicles are always near the surface; as they grow in size they move deeper into the stroma. The first step is the multiplication of the cells of the folHcle, Fig. 31, A, which converts the follicle into a layer of cubical cells with the nuclei at an even height. During this change in the follicle the primordial ovum does not alter in size. The second step is the elongation of the cells into a cubical form, with an accompanying enlargement of the ovum. The growth of OVA. 51 Fig. 31. — Primary follicles from the ovary of a woman, thirty-one years old: Th. connective-tissue layer; /, epithelical follicle; z, beginning, zona pellucida ; nu^ nucleus or ger- minative vesicle. After W. Nagel. the ovum affects the protoplasm, the nucleus, and the nucleolus, all of which increase their dimensions. The follicular wall steadily increases in thickness; at first it remains single-layered, but the nuclei take their places at various levels; a little later it be- comes several-layered, and then the formation of the first en- velope {zona pellucida) around the ovum begins. Nagel, 88.1, 380-382, calls attention to the large clear cells with large nuclei^ which show a distinct reticulum and one or several chromatin granules ; they are found in somewhat larger follicles, but only up to the time when the yolk granules begin to form in the ovum . Nagel inter- prets these cells as having a nutritive function, and calls them Ndhyzellen; he offers very little evidence in favor of his view. The cells in question measure 16- 31 //, and are much smaller than the prim- ordial ova, which they somewhat resemble in appearance. These cells have been seen by various authors, e.g.. Call and Exner (Sitzber. Wien. Akad. Wiss., etc., 15 April, 1865). It is more prob- able that these cells have to do with the formation of the liquor of the Graafian follicle, which begins while they are present. The cells of the granulosa multiply by indirect division, as has been shown by Harz and also Flemming {Arch, mikrosk. Anat., XXIV. 376-384). I have found the numerous karyokinetic figures in the follicles of the rabbit's ovary, though hardly quite as abundant as Flemming's description led me to expect. The mitoses have not been found in the first stages of follicular growth. During the growth of the follicle there is formed, as was first described by Schron, 63.1, 419, a network of blood-vessels close around the folli- cle; the layer of blood-vessels constitutes the so-called tunica vascu- losa or theca foUiculi; the first vessel is a simple loop, which embraces the young follicle; other loops approach and unite with their fellows to form a network. Development of the Graafian Follicle. — After the epithelium of the primary follicle has become many-layered, there appear in it rounded vacuolated spaces, which increase in size and finally become confluent, so that there is a space or fissure in the epithelium. Fig. 32, U. This fissure divides the epithelium into two layers, an inner one immediately surrounding the ovum, and an outer one next the stroma of the ovary. Since the fissure does not extend completely around the follicle there is one place where the two layers are united. Fig. 32 ; the place of union, though variable in position, is always on the side of the follicle away from the surface. The fissure is filled with a serous fluid known as the liquor follicuU. In man and most mammals there is a single continuous fissure; but in the rabbit, and perhaps other rodents, there are often cords of cells stretching across from the outer to the inner lamina of the epithelium ; the cords vary in number from two to ten ; they were first described by Barry, have been beautifully figured by Coste, 47. 1, Lapin, L. I., Fig. 2, and are 0-2 THE GENITAL PRODUCTS. known as the retinacula. The development of the fissure changes the primary into a Graafian follicle. The Graafian follicle is bounded by a layer of epithelium known as the membrana granulosa, from its appearance when examined in the fresh state ; it is surrounded by a vascular layer, characterized not only by its blood-vessels, but also by the condensation of the con- nective tissue composing it. The follicle lies a little below the layer of primordial ova. To a part of its walls on the side away from the surface of the ovary is attached a mass of cells more or less globular in shape; this mass is known as the discus or cumulus proUgerus; it encloses the ovum ; the cavity between the discus and granulosa is the cavity of the follicle, and contains the liquor foUiculi. The further history consists principally in growth and secondary modifi- cations. The follicular wall and the discus increase in thickness; there is added a very thin basement membrane, Waldeyer's mem- ,X\^^A^ (m^^'' 3 5 icesi Graafian follicles ; 11, corpus luteum; Ve, blood-vessels; Hi, hilus. After SchrBn. fia. 33.— Ovary of oat: 1-11, successive stages of the ova; 1, primordial ovum; 8, 9, 10, " " "" ^ ■ i; Fe, blood-vessels; Hi, nil brana propria, close around the outside of the granulosa and sep- arating it from the tunica vasculosa ; the membrana propria is said to be an endothelium derived from the connective-tissue cells of the ovary. The vascular membrane or theca foUiculi becomes differen- tiated into an outer fibrous layer (Henle's tunica fibrosa) carrying the larger blood-vessels, and an inner less fibrous layer carrying the smaller blood-vessels (tunica propria). The distinction between the membrana propria and tunica propria should not be overlooked. The smallest blood-vessels running around the follicle from below, and minutely subdivided on its upper surface, converge toward a point near the surface of the ovary ; this point is called the stigma, contains no blood-vessels, and marks the spot where finally the folli- cle is to rupture to allow the ovum to escape. The stigma, owing to OVA. 53 the absence of blood-vessels, is yellowish-white. In mammals and birds it is elongated and rounded in outline, but in lizards is angu- lar (Coste, 47. 1, 160) . The cells of the granulosa acquire, at least in the cow, highly characteristic forms (Lachi, 84. 1) ; there are, 1st, very narrow elongated cells, which stretch through the entire thick- ness of the layer, and present, when isolated, curious irregular forms; they have oval nuclei, about which there is usually a small amount of protoplasm ; the nuclei of these cells lie in the half of the granulosa next the cavity of the follicle. 3d, cells with rounded nuclei, larger cell-bodies, and a few fine processes of irregular shapes ; these cells lie between the processes of the others in the outer half of the membrane. 3d, cells that are probably immigrated leu- cocytes. The cells of the discus have not yet been minutely stud- ied; those next the ovum, are cylindroid, and radiate around the zona, constituting thus the so-called corona radiata of authors — compare Fig. 34. The cells of the outer layer of the discus are more rounded in form ; it is, of course, probable that the two forms of discus-cells resemble the cells of the granulosa in actual shape. Just before the primary follicle changes into the Graafian follicle the ovum, at least in man, has attained its full diameter, but still contains no yolk (deutoplasm) . At this time there appears a clear, delicate membrane close around the ovum, separating it from the cells of the follicular wall. In the Grraafian follicle this membrane steadily grows until it attains a diameter of 20-24 ij-\ it is called the zona radiata or pellucida; its structure is described in the subsequent section on the full-grown egg-cell. The first yolk-grains appear in the human species when the zona pellucida has attained a thickness of 1 ij- or more, and are situated always in the centre of the egg-cell (Nagel, 88.1, 385, 386) . In other mammals they are said to appear earlier. The j-olk-granules must be considered as the direct products of the vital activity of the egg- cell itself, and in my judgment thfere is no sufficient basis for any other view. Various hypotheses as to the origin of the yolk-grains have been advanced. Thus Waldeyer, 70.1, has maintained that the grains are produced by the cells of the follicle, and are trans- ferred from them across the zona into the ovum. It is not impossi- ble that very small young granules may arise in the follicular cells and be transmitted along the fine processes by which the cells are connected through the zona radiata with the ovum, and that these granules subsequently grow within the egg-cell, as Caldwell, 87.1, asserts is the case in monotremes and marsupials. Caldwell's state- ments are so aphoristic that the question must remain unsettled until more fully investigated. Lindgren, 77. 1 , asserts that the cells of the granulosa immigrate through the zona to form the yolk-gran- ules ; his observations were made on ova which had already been somewhat macerated, and which had the processes of the follicular cells swollen in consequence. That the yolk-grains are produced by the gradual enlargement of small ones has been shown by Sarasin's researches on reptiles, 83. 1. He found in Lacerta a central area of small granules which gradually enlarge ; this area {Herd der Dot- terbildung) persists even after the embryo has appeared, and the egg increases in volume and weight after the segmentation has begun. 54 THE GENITAL PRODUCTS. The characteristics of the human yolk-grain have not. been accu- rately investigated, nor have those of any of the higher mammalia been studied carefully. In the human ova the grains are 1 //- or less in diameter; highly refringent and of various kinds. In a sheep's ova Bonnet, 84.1, found small granules, fat-globules in considerable abundance, p. 178, and larger granules vsrhich stain with eosine, I. c, p. 183. Their accumulation continues centrifugally, forcing the nucleus of the ovum to an eccentric position ; v\rhen the maximum of the vitelline deposit is reached there is only a very thin layer of protoplasm around the outside of the egg-cell. Fig. 34, and a court of protoplasm around the nucleus. This disposition is particulai'ly well shown in the ova of the monotremes and marsupials. See Caldwell, 87.1, PI. XXIX. , Fig. 6. The cortical layer is readily dis- tinguished in fresh ova, but in hardened specimens is quite or whoUy indistinguishable. In the ovum of the placental mammalia the yolk never attains a great development ; but in most vertebrates the gran- ules gradually enlarge, and in some cases they are quite big. When they are thus developed it is easy to see that they are of various sorts. Thus in the hen's ovum there are two principal kinds of yolk-grains, the yellow and the white. The yellow grains are spheres of from 35 /* to 100 a in diameter, filled with numerous minute, highly refractile granules ; these spheres are very delicate, and easily destroyed by crushing. When boiled or otherwise hardened in situ, they assume a polyhedral form from mutual pressure. The white grains are vesicles, for the most part smaller (4 /i to 75 ij) than the spheres of the yellow yolk with a highly refractive body, often as small as 1 /* in the interior of each. There are also larger spheres, each of which contains a number of spherules similar to the smaller vesicles. The yolk-plates, or plagiostomes, which consist prin- cipally of lecithin and nuclein, are not present in the younger ova, but are present in great num- bers in the full-grown ones ; they are oval, barrel- shaped, or rectangular bodies, with rounded cor- ners and edges; the surface, especially in the larger plates, shows a fine transverse striation, corresponding to the laminate structure of the grain. As no thorough comparative investigation of the yolk-granules has been made, it is not worth while to enter into further details. Besides the yolk-grains there may also be present one or several large masses of nutritive material, such as the " oil-globules " of many teleosts, or the so-called yolk-nucleus. The " oil-globules " are produced by the liquefaction of the yolk, and are not oily. The yolk-nucleus has been described by Balbiani in the Arachnida. The eggs of some spiders contain, besides the nucleus, a second body (Fig. 33, k) , of about the same size as the nucleus, solid, resistant, and exhibiting indications of a series of concentric laminae ; this is the so-called yolk-nucleus, and is probably only a specialized form of deutoplasm, and might be compared, for instance, to the four large oil-globules described by Spengel in the eggs of Bonellia viridis. Fig. 33. — Egg-cell of Tegenaria domestica. tj-, nucleus ; fc, laminate yolk-nucleus. After Bal- biani. I OVA. 55 A yolk-nucleus has since been recorded in the ova of various verte- brates; thus Schtitz ("Ueber den Dotterkern," Diss. Inaug., Bonn, 1882) found in the ovarian ova of the pike, in September and Octo- ber, a round or oval body, not sharply delimited, clear, and more homogeneous than the protoplasm, and which increased in size with the growth of the egg. A yolk-nucleus, consisting of an accumula- tion of larger and smaller granules, has also been observed in the frog and newt (O. Schultze, 87. 1), but is apparently wanting in Bufo and Bombinator (Gotte) . The amount of yolk varies in different animals very greatly, and determines, apparently, the size of the ovum. It has been observed that the process of segmentation varies according to the amount of yolk, and this has led to the arbitrary division of ova into meroblas- tic and holoblastic (see Segmentation of the Ovum). The yolk usu- ally, perhaps always, leaves a peripheral layer of protoplasm free. In all vertebrate and in some invertebrate ova this layer of proto- plasm is thickened, often considerably, around one pole of the ovum, which is then distinguished as the animal pole, the opposite pole being called the vegetative. These are old terms, which have come down to us from the time when the ectoderm, which is produced during segmentation, principally from the substance of the animal pole, was called the animal layer and the entoderm the vegetative layer. It is at the animal pole that the extrusion of the polar glob- ules under normal conditions invariably takes place. The Graafian follicle grows very much more than the ovum, until it becomes a large cyst. Fig. 32, the position of which is marked by an external protuberance on the surface of the ovary. To the deep wall of this cyst is attached the discus proligerus with the ovum, which is now nearly full grown. The stigma is at the protuberant point of the follicle, which is covered by very little ovarial tissue, so that there is a very thin wall only separating the cavity of the folli- cle from that of the abdomen. The degeneration of the Graafian follicles with the contained ovum occurs normally in the ovary ; but as the process has no direct inter- est for the embryologist, it will suffice to refer to Frommann's very admirable summary (Eulenburg's "Real. Encyclop. Heilkunde," V., 602-604). Full-grown Ovum before Maturation.— The fuU-grown hu- man ovum is distinguished among mammalian ova for the clear development and ready visibility of all its parts — a peculiarity due chiefly to the small amount of the yolk and fewness of the fat-gran- ules it contains. Fig. 34 represents an ovum from a nearly mature Graafian follicle of a woman of thirty years ; the specimen was ob- tained by ovariotomy and examined and drawn in the fresh state, being kept in the liquor follicle. This specimen gave the following measures- The diameter of the whole ovum, including the zona radiata, 165-170 //; thickness of the zona, 20-24 p-; perivitelline fis- sure, 1.3 /J'.; the clear outer zone of the yolk, 4-6 //; the protoplasmic zone, 10-21 /j. ; the deutoplasm zone, 82-87 /j- ; the nucleus, 25-27 m. The corona radiata, cor. r., exhibits the features already described. The zona pellucida, Z, shows a distinct radial striation ; this is prob- ably due to the presence of minute pore canals running through the 66 THE GENITAL PRODUCTS. FiQ. 34. — Full-grown human ovum before matura tlon: cor. r, part of the corona radiata ; .2, zona pelhi- cida; Pi, protoplasm ; F, yolk; JVm, nucleus. After W. Nagel. zona, and which, at least in early stages, give passage to processes of the cells of the corona radiata, which unite with the ovum. These processes have not yet been observed in man in an altogether satis- factory manner, and indeed Nagel, 88.1, 403, expressly denies their existence, as well as that of the pore canals. The processes are, how- ever, readily seen in the low- er vertebrates, in the mono- tremes and marsupials, Cald- well, 87.1, and have been observed in the placental mammalia, Fig. 35. Hence it seems probable that they are present in man at least while the ovum is growing, though they may be obliter- ated at the stage we are now considering. Several observ- ers record " dumb-bell cells" * with the thin portion of the cell passing through the zona, and one knob lyingonthe out- side, the other on the inside, of the zona, compare H. Vir- chow, 85.1. But apparently such observations have been made solely on ova that had been somewhat macerated, and therefore the " dumb-bell cells" result probably from post-mortem changes, and cannot be interpreted as by Lindgren, 77. 1, to prove the actual normal passage of cells of the discus proligerus through the zona. The zona has no micropyle or special open channel for the entrance of the spermatozoon. For additional details, see the following section on the envelopes of the ovum. The ovum proper is separated by a narrow fissure, p v., the peri- vitelline space, from the zona, within which it lies free and loose, so that when a fresh specimen is examined the same side of the ovum — that containing the nucleus, which is the lightest part — is always found uppermost. The ovum has no vitelline membrane, according to Nagel, 88.1, 405 ; but in several mammals such a membrane has been described, appearing as a thin, delicate line about the time the ovum matures, Fig. 35, v.Tin. The body of the ovum may be divided into an inner kernel containing the yolk-granules and an outer protoplasmatic zone, of which the very outermost thin layer is clear, and therefore more, or less differentiated from the broader, deeper layer, which is granular and constitutes most of the zone. Frommann, 89.1, has shown that many of the granules in the ova of the sea-urchin are part of the protoplasmic reticulum; in the living egg they are incessantly changing in shape and in their connections, even disappearing and reappearing; the disappearance Frommann terms liquefaction, the reappearance a new formation. It seems to me possible that the * Nagelzellen, Spundzellen, Zwillingszellen, or Hant^lzellen of German writers. OVA. 57 changes seen are probably in part effects of contraction in the reticu- lum. The nucleus is nearly spherical, always eccentric in position, and has a nucleolus which in the fresh specimen shows amoeboid movements even at ordinary summer temperatures for several hours after removal from the ovary, Nagel, 88. 1, 407. In hardened speci- mens the nucleus shows its reticulum, as already described. In certain ova there has been observed a special band of proto- plasm leading from the surface of the ovum to the egg nucleus. This is found in the ovum of Petromyzon, having been first described and figured by Calberla, 78. 1, who, however, erroneously designated the nucleus as the female pronucleus, and interpreted it as the path- way performed for the passage of the spermatozoon — an error which Boehm has corrected by showing that the true pronucletis is formed later. As shown in the section on impregnation, p. 69, the path- way of the spermatozoon can be traced in certain amphibian ova. Peculiar names have been applied to the nucleus and nucleolus of the ovum, and are still in general use. The nucleus was first dis- covered in 1830 by Purkinge (" Sym- bolse ad ovarium historiam,"1830)in birds, and by Coste (1837) in mam- mals, and became known as the vesi- cula gerniinativa, ( Pa r k inje'sches Bldschen, or ger- minal vesicle). The nucleolus was first described in 1835 by R. Wag- ner, 35.1, and be- came known as the germinative or Wagnerian spot [Wagner'' scher Fleck). It was not, however, until 1839 that Theodore Schwann for the first time interpreted the ovum as a cell ; but before then the terms germinal vesicle and germinal spot had established themselves, and since then they have remained in general use. The Envelopes of the Ovum. — The eggs of different classes, and even species of animals, are, as is well known, extremely unlike m appearance. The dissimilarity refers chiefly to size, to the char- acter ot the yolk, and the nature and number of membranes or other envelopes, by which the ovum or egg-cell proper is surrounded. Thus in the hen's egg the yolk alone represents the part correspond- ing to the egg-celi, while the white of the egg and the egg-shell are only secondary envelopes, the former serving to nourish, the latter to protect, the so-called yolk, which is the essential part, the true egg. The various envelopes which eggs ever have may be classed under tour categories: First, a very thin and delicate one, the proper membrane ot the cell itself, and which ought always to be Fig. 36.— Part of the uvum o£ a mole: Ep. cells o£ corona radia- ta ; Z, zona pellucida ; v.m,., vitelline membrane ; After W. Heape , pore canals. 58 THE GENITAL PRODUCTS. distinguished as the vitelline membrane ; second, the ovarian enve- lopes, which are secreted around the egg-cell by the tissues of the ovary; third, the envelopes secreted by the oviduct, which may form a coating of nutritive material, or a protective shell, or both, as in the hen's egg, of which the nutritive white is secreted by the upper part, the calcareous shell by the middle part of the oviduct ; fourth, coverings secreted by accessory glands, such as the slime in which the eggs of snails are embedded, or the tough capsules in which leeches lay their eggs. By adhering to this classification it is possible to find one's way through the labyrinth of special descrip- tions. It is impossible to review here the manifold variations in the ovarian coverings of animals, and we shall attempt only to de- scribe those of the higher forms. All vertebrate ova probably have two envelopes : first, a very thin inner one, the vitelline membrane proper ; second, a thicker ovarian membrane, known as the zona radiata or pelluci da. The vitelline membrane is described by Heape, in the mole, as a very thin but distinct membrane (Fig. 35, v. m.), immediately against the yolk, separated by a narrow space from the zona , it is to be regarded as a product of the ovum itself. It appears a short time before the ovum matures, and is most distinct at the time of the formation of the polar globules; its fate during segmentation has not been ascer- tained. The so-called vitelline membrane (Dotterhaut) of amphibia is really the homologue of the zona (Frommann). Considerable doubt in regard to the presence of this membrane in vertebrates, and especially in mammals, has been expressed by various writers, but its existence seems to me to have been sufficiently demonstrated. It was first described by Reichert in 1841, and again by H. Meyer in 1842. In recent years it has been redescribed by Ed. van Beneden, by Heape, and others. Balfour, in his " Embryology," pronounces in favor of its occurrence. The zona radiata {pellucida of C. E. v. Baer), Fig. 35, Z, is a membrane, usually of considerable thickness, which can be distin- guished around the ovarian ovum quite early, being at first very thin, but gradually increasing in thickness until it attains in man a diameter of about 20 /i in the mature ovum. In the pig the diameter becomes 7 to 9 /J.; in the sheep, 7 to 12 //; in the cow, 7 to 8 /J. (Schulin) ; in the mole, 8 to 11 i^, ac- cording to Heape. In the mature ovum it is a tough, clear, glistening membrane, very Fig. se.-ovum of a sea ur- rcsistant to acids, and soluble in alkalies only chin, Toxopneustes lividus. .,i j./v^ ix tx • • .3 t_ After o. Hertwig. With dimculty. It IS picrced by numerous radiating pores, which produce the appear- ance to which the term zo7ia radiata refers, These pores were first observed by Johannes Miiller and Remak in fish eggs,* and they have since been observed in the ova of many other verte- brates, including several sjjecies of mammalia. It is probable thai they always exist, despite the doubts expressed by Schulin, Lind * The homologies ot the two envelopesaroundflshovaare6omewhacuncertain,seeE L Mark 9u. J. OVA. • 59 gren, Von Sehlen, Nagel, and others. While the ovum is still in the ovary it is surrounded by the cells of the discus proligerus ; these cells send processes through the pores of the zona (Fig. 35) . It is now commonly supposed that these processes are channels of nutri- tion for the ovum. The zona is somewhat granular in its outer por- tion, next the cells of the corona. Balfour has suggested, not very plausibly, I think, that the granular portion does not belong to the zona, but represents the remains of a hypothetical primary vitelline membrane, within which the zona proper arose subsequently. An- other very hypothetical homology is suggested by Caldwell, 87.1, who finds two membranes around the ovarian ovum of marsupials ; the inner membrane resembles the zona pellucida, and is termed by Caldwell erroneously the vitelline membrane ; the outer membrane is the proalbumen, which, during the passage of the ovum through the oviduct, swells up and becomes the albuminous envelope of the ■egg. Caldwell homologizes the inner clear layer of the zona of the placental m&mmals with the zona of marsupials, and the outer gran- ular layer with the proalbumen. In Petromyzon (Boehm, 88. 1), as in some teleosts (J. Brock, 78.1), there are two ovarian envelopes which are quite probably homologous with two envelopes found in marsupialia, but that the zona of the placentalia represents two envelopes united in one is, at least, very uncertain. It seems to me that the zona radiata is to be regarded as a modified intercellu- lar substance, and that the processes going through its pores are to be homologized with the ordinary intercellular protoplasmic bridges of epithelial cells. Heape thus describes the pores of the zona in the mole: "The radially striated appearance of the zona has long been shown to be due to a vast number of fine canals passing radially through it. The canals, I find, open on the inner side of the zona by a slightly dilated mouth, while on the outer side of the zona they communicate with the exterior by a considerably wider opening. Fig. 35. Into the external openings of these canals I have been able to trace prolonga- tions of those cells of the discus which are immediately in contact therewith. Fig. 35, and there appears to me no room to doubt that the contents of these follicular cells are thus rendered available for the nutriment and growth of the ovum." The term micropyle is used to designate a passage through the envelopes of the ovum, which serves to admit the spermatozoon. The micropyle is present in many invertebrate cA^a, notably in those of insects, and may have a quite complicated structure. In the vertebrates it is very rarely found, having been thus far positively demonstrated only in certain teleost eggs. Calberla, 78.1, affirmed that a micropyle was present in Petromyzon ; but Boehrn, after a later and more thorough investigation, 88. 1, expressly denies its ex- istence, Kupffer and Benecke, 78. S, having previously shown that «the spermatozoa penetrated the lamprey ovum at several points. Sundry authors from time to time have asserted that a micropyle was present in the mamamlian ovum, but the evidence against it seems to me conclusive. The corona radiata is the name given to the envelope of cells of the discus proligerus, which adheres for a short time to the zona 60 THE GENITAL PRODUCTS. radiata when the ovu.m is discharged from the Graafian follicle. The corona may be represented only by a few patches of cells, or may be a complete envelope; in either case the cells are entirely lost soon after the ovum begins its descent through the Fallopian tube. The egg of Lepidosteus has two envelopes ; the outer one is homologized by Beard with the corona radiata, but E. L. Mark, 90. 1 , denies this homology. The disappearance of the zona has been specially studied by Tour- neux et Hermann (C. R. Soc. Biol., Paris, 1887, p. 49), who found that it could be distinguished in rabbits' ova of ninety-five hours, but not in those of one hundred and sixteen hours. According to Hensen the zona in guinea-pigs is ruptured, and the ovum escapes during the descent through the oviduct. Polarity of tlie Ovum. — The mature egg-cell has a distinct axis, the two poles of which are unlike in character, while around the axis there is a complete radial symmetry so far as known. In my opinion the essential difference between the two pole§ is that the nucleus is nearer one than the other, and consequently the proto- plasha of the egg-cell is more concentrated at one pole than at the other; for, as is well known, the nucleus usually has an accumula- tion of protoplasm around it. The eccentric position of the nucleus is, I think, probably universal. Curiously it is frequently stated that the nucleus lies in simple ova in the centre, * and the notion is prevalent that the accumulation of yolk is the cause of the eccentric position in certain ova. This notion is not quite correct ; on the con- trary, we must assume that the position of the nucleus causes the eccentricity of the yolk material. There is unquestionably a strong tendency for nucleus and protoplasm to keep company : thus we see when cells are connected with one another by protoplasmatic bridges, a main cell-body around each nucleus. Again, within single cells, the protoplasm often forms a court around the nucleus and a looser net- work throughout the rest of the cell ; in ova with mcomplete segmen- tation each nucleus is imbedded in its special accumulation of proto- plasm; it appears to me, accordingly, that the disposition in the egg- cell is only a special instance of a more general principle. The eccentric position of the ovic nucleus is due to as yet un- known causes ; but being given it determines the accumulation of yolk-grains at the opposite pole ; it will be remembered that in the developing egg-cell the nucleus becomes eccentric before the yolk- grains appear. The amount of j'olk undoubtedly affects the degree of the nuclear eccentricity. The nucleus reigns over a compara- tively email territory, within which there is no, or but very little, yolk-matter developed ; in all vertebrate ova the perinuclear proto- plasm touches the vitelline membrane and ixiarks externally the site of the nuclear or so-called " animal" pole. In the rest of the egg-cell the yolk-grains may be freely developed, and as they increase in number and size there is a corresponding distention of the region of the cell which they occupy. This distention may go so far that, as in the birds'\ovum, the perinuclear territory is minute compared with the great bulk of the deutoplasmic territory, and consequently * For example O. Hertwig, 1888. 1, p. 8, says 'das Kelmblaschen lagert gewohnlich in der Mitte des Eies," yet his own figures correctly represent it as eccentric. OVA. 61 •the nucleus lies far away from the centre of the ovum.* The yolk- grains centre about the pole opposite the nucleus, which might therefore be called the vitelline or deutoplasmic pole, though it is still generally known by the inappropriate name of vegetative pole, which has come dow-n to us from long ago. F. M. Balfour, in his " Comparative Embryology," divided ova into three classes, as follows: 1st, alecithal, without any deutoplasm; 2d, telolecithal, with the deutoplasm collected opposite the animal pole; 3d, centrolecithal, with the deutoplasm in the centre sur- rounded by a cortex of protoplasm. It is probable that all ova are telolecithal in the sense that they have a nuclear pole, and that the yolk-matter is developed away from the nuclear pole. The alecithal ova are those in which the nuclear eccentricity is at a minimum; the centrolecithal ova, which occur only among invertebrates, are likely to prove to be really telolecithal. All known vertebrate ova are telolecithal. The polarity of the ovum dominates the process of the ripening of the egg-cell, and has a very important influence on the process of segmentation after impregnation. The extent of this domination has been thus summarized by E. L. Mark, 81.1, 515: "The migra- tion of the germinative vesicle toward a definite point of the sur- face; the radial position assumed by the maturation spindles; the waves of constriction which precede the formation of the polar globules, and the inequalities in the sizes of the latter ; the union of the pronuclei at a point nearer the primary than the secondary pole, and the consequently (?) eccentric position of the first seg- mentation spindle; the appearance of the first segmentation furrow earlier at the primary than at the opposite pole; the for- mation of pseudopodia-like elevations, often most conspicuous at the primary pole ; the accumulation of finely granular protoplasm at the secondary pole after the elimination of the polar globules ; and the appearance of ' polar rings ' and ' ring rays (Clepsine) at both ends of the primitive axis, are all indications of a polar differentia- tion of the egg." The polarity of the ovum also evinces itself in the difference of the specific gravity of the two poles ; usually, as in mammals, birds, amphibians, many fish and invertebrates, the deutoplasmic pole is heavier, and the ovum always presents the animal pole uppermost as soon as it is left free to turn ; in the ripe mammalian egg the yolk has room to turn within the zona : hence when the fresh ovum is examined under the microscope, the animal pole is toward the ob- server and the eccentric position of the nucleus cannot be observed. In various pelagic teleost ova the animal pole is the heavier, and the embryo develops accordingly on the under side of the egg. Maturation of the Ovum. — The term maturation is restricted by usage to the series of phenomena accompanying the expulsion of the polar globules which occurs after the egg-cell has attained its full size, and just before or just after the separation of the ovum from the ovary. A polar globule is a small, nucleated mass, extruded from a fully-grown egg-cell. When an ovum is about to mature its nucleus moves nearer that * J A. Ryder has published a semi-popular discussion of nuclear displacement. 8^. ]. 62 THE GENITAL PRODUCTS. point of the surface which may be regarded as the centre of the ani- mal pole, and there also occurs a contraction of the vitellus. The centrifugal movement of the nucleus was iirst observed by Von Baer, 27. 1, 29, in the hen's egg, and has since been seen by very numerous observers and in very numerous species ; it must, there- fore, be considered as an unvariable phenomenon. Concerning the force which moves the nucleus we have no definite conception ; for discussion of the question, see Whitman, 87.3. The contraction of the yolk is probably also a constant phenomenon ; it is apparently effected by the expulsion of fluid from the protoplasm, so that a clear space separates the zona and yolk. The observations have not been collated yet on this point, and it is impossible to state whether there is a constant rule as to the extent and epoch of the contraction. After reaching the surface the nucleus as such disappears. This fact was known to Purkinje, 30. 1, 15, the discoverer of the nucleus, and has been shown to occur in all eggs which have been accurately examined. K. E. von Baer maintained both in 1827 and subse- quently, 37. 1, 4 and 9, 37. 1, 28, 157, 297, the opinion that the dis- appearance of the germinal vesicle was connected with the maturation of the ovum — a conclusion which is now established beyond ques- tion. Reichert in 1846, 46.1, 199, 205, maintained that the disap- pearance was the first result of impregnation, and in this error he has had several followers (A. Miiller, Haeckel, Biitschli, and others). In birds the nucleus assumes a very large size, and migrates to the surface of the ovum, when it disappears as shown by Oellacher. M. HoU, 90. 1, records that in a newly hatched chick the ova meas- ured about 14 //. X 9 p-, while an ovum nearly ready to leave the ovary measured 40 X 35 mm ; in the former the nucleus was about 9 ij., in the latter 315 X 117 /■/- in diameter. jSTo polar globules have yet been observed in birds, though we must assume that they are formed. The disappearance of the germinal vesicle is only apparent, not actual, being in reality a metamorphosis. It is probable that the first step is the discharge of nuclear fluid {Kernsaft) into the surrounding protoplasm. This is indicated by two appearances — 1st, the shrink- ing of the nucleus, the outline of which becomes shrivelled ; 2d, a clear space which arises around the nucleus. The shrivelling of the nucleus has been observed in several mammals (Van Beneden, Rein, Bellonci, Tafani) in various vertebrates — as, for instance, in teleosts by Oellacher, 72.1, 3, in Amphibia by O. Schultze, 87.1, and in many invertebrates, e.g., Serpula by Schenk (Sitzber. Wien. Akad. LXX. Abth. 3, 291-294, 1875), in Hydra by Klemenberg, 72. 1, 42, in Asterocanthion by Ed. van Beneden, 76.1. The clear perinuclear space has been noticed especially in Anura by Gotte, 75.1, 20-32, and 0. Schultze, 87.1, 217. The second step is the dissolution of the mem- brane of the nucleus, so that the nuclear contents are brought into direct contact with, and partly mix with, the cell-plasma. Very likely this mixture of nuclear and cell substance is, as O. Schultze suggests, 87. 1, 215, one of the essential factors of maturation. The dissolution of the nuclear membrane has been found to occur in so many species that we may safely predicate it of all. We now find the contents of the nucleus lying together in the centre of the proto- plasm of the animal pole. The contents themselves are altered in OVA. 63 Fig. 37, —Ovarian egg of Hsemops ■ sp , nuclear spmdle ; character, the most noticeable change being the breaking up of the chromatin into separate granules ; in mammals the formation of the granules by the cleavage of the nucleolus occurs after the nucleus has begun its migration (van Beneden, Bellonci, Tafani) ; in Am- phibia the nucleus becomes multinucleate during the early growth of the ovum. The achromatic substance or reticulum of the nucleus appears as threads often very difficult to recognize. The threads and granules proceed to group themselves into a spindle-shaped body, the so- called nuclear spindle (Kernspindel) which lies more or less nearly in the radius of the ovum and has one of its ends close to the sur- face of the yolk, Fig. 3S, sp. The achromatic threads run from pole to pole of the spindle ; the chromatin granules lie in the centre of the spindle in one plane and produce the ap- pearance of a transverse band or disc (Strass- huTger' s Kernplatte) ; each chromatin granule fe?heOTary''"Aftyj?o''Hert^ is associated with one of the spindle-threads, wig. Each pointed end of the spindle lies just within a rounded clear space, from which, and not from the end of the spin- dle, radiates threads in the yolk, whence results a figure like a con- ventional sun. The whole spindle with the two suns has been named the amphiaster. As amphiasters occur in connection with ordinary indirect cell-division the distinctive term archianiphiaster has been proposed for those concerned in the production of the polar globules. Sometimes as in Limax, Mark, 81.1, the astral rays are not straight, but curved as in a turbine. In amphibian ova only a portion of the granules enter into the formation of the chromatin, while the majority of them are mingled with the yolk (0. Schultze) ; it is possible that this modification is connected with the large amount of yolk and will be found in other vertebrate ova. In the ova of mammals (all?) the chromatin enters into the " Kernplatte." The shape of the spindle varies, as does also the distribution of the granules of the nuclear plate, thus ; In the guinea- pig, the ends are pointed and the threads are straight, so the outline of the spindle is like a diamond ; in the bat the spindle is barrel-shaped and the threads are curved. In certain, possibly in all, cases the spindle, when first formed, lies ob- liquely, and subsequently becomes erect to the surface, as Whitman observed in the leech (Clepsine, 78.2) ; for further reference, see 0. Schultze, 87.1, 319-221. The reason for the obliquity and the following erection is unknown. The next changes may be followed with the help of Fig. 39. The spindle, driven by an undiscovered power, continues the centrifugal movement until it is partly extruded from the egg, as shown in the Fig. 38.— Egg of a leech (Nephelis), three-quarters of an hour after being laid ■ formation of the first polar glob- ule, p g. After O Hertwig. 64 THE GEXITAL PRODUCTS. figure ; the projecting end is enclosed in a distinct mass of protoplasm, p.g., which is constricted around its base. The fragments of chro- matin have each divided into two, and one-half of each fragment has moved toward one end, the other half toward the other end of the spindle. The half- fragments of each set move together, hence there seem to be two plates within the spindle. The translation of the groups of chromatin grains continues until they reach the ends of the spindle ; the achromatic threads then break through in the middle. % Thus the original nucleus, or at least part of it, has been divided. There are now two masses of nuclear substance — one in the ovum, the other in a little appendage to the ovum ; this appendage is the first polar globule ; its nuclear substance does not develop into a complete nucleus. The remnants of the egg-cell nucleus within the ovum undergo further changes. Usually when the amphiastral (indirect or kinetic) division of a nucleus is over, the separated nuclear masses resume the structure of a normal resting nucleus ; but in the ovum, as Plat- ner, 89.1, has especially noted, the nuclear remnants change directly into a second spindle, which lies as did the first within the protoplasm of the animal pole, and likewise gives rise to an amphiaster (second archiamphiaster, zweites RicMungsspin- del) . The second spindle even more clearly than the first has been observed to occupy an oblique position, as in mammals (Bellon- ci, 85.1), or even parallel with the surface, as in amphibians (O. Schultze, 87.1) and certain Crustacea, Weismann and Ischika- wa, 88.4. This spindle produces a second polar globule in similar manner to the first ; the globule is somewhat smaller than the Fig 39 ^Ovum of Nepheiis (a first, and is at least sometimes connected p'|?p^oia?gfobuSf/?flmifJp°?: both with the first globule and with the nucleus ; m, male pronucleus, ovum. Sometimes the first globule divides Aiter o. Hertwig. -^^^ ^^^^ y.^ gg^ ^^ ^^^ ^^^^ ^^^ ^^^^.^ connected together. The connection of the globules with the yolk persists for some time, and in the case of leeches is not dissolved until segmentation begins. The polar globules ultimately disappear — how is not exactly known. That they take no part in the further history of the ovum may be considered established ; for they break off and may often be seen in mammals knocking about within the zona, while the ovum is devel- oping after impregnation, and they then present a hyaline appear- ance, as if slowly degenerating. The number of polar globules, as Weismann and Ischikawa, 87.2, 88.4, first explicitly demonstrated, is two. According to these authors, 88.4, 590, two polar globules have been shown to occur in 8 species of coelenterates, 5 of plathelminths, 6 nemathelminths, 1 zephyrean, 10 annelids, 5 echinoderms, 22 mollusks, 6 tunicates, 1 bryozoon, 15 crustaceans, 6 insects, 11 vertebrates. It can hardly be doubted that two polar globules are necessary for the complete maturation of the ovum, and that until they are formed impregna- tion cannot take place. On the other hand, Blochmann discovered OVA. 65 that in a parthenogenetic ovum there is only one polar globule formed, and Weismann and Ischikawa, 88.4, have shown that this is true of many and presumably of all parthenogenetic ova — that is, of ova which develop without fertilization. For the theoretical consideration of the polar globules, see below. The polar globules appear to have been seen as long ago as 1837 by Dumortier in gasteropods, and in 1840 by the elder Van Beneden, and in 1842 in the rabbit by Bischoff . Fr. Miiller observed them more carefully in 18 tS, and detected their constant relation to the planes of segmentation, and gave them the current German name of Bichtungskorperchen. Robin, in 1862, termed them globules po- laires, which, translated, has become the accepted English designa- tion. Biitschli, 76.1, in 187G, first led the way toward a correct con- ception of the origin of the globules, and about the same time came the independent researches of 0. Hertwig, whose able memoirs, 75.1, 77.1, 77.2, 78.1, have formed the basis of all subsequent work. These were soon followed by the investigations of Fol and many others. From these studies we possess a tolerable general concep- tion of the origin of the polar globules, but the comparative study of the details and variations remains for the future. After the formation of the second polar globule there is a small group of chromatin elements and achromatic threads, which, since they have been halved twice, represent approximately one-fourth, not of the whole egg nucleus, but of so much thereof as entered into the formation of the first polar spindle. The nuclear remnant lies close to the animal pole and in the clear protoplasm ; it is the so-called female pj-onucleus, the history of which varies according to the species of animal. Three tendencies are known to affect the pro- nucleus — namely, to move toward a central position in the ovum ; to unite with the male pronucleus as soon as that is formed out of the spermatozoon, which enters the ovum to fertilize it, and to assume the character of a membranate nucleus. As the time of the forma- tion of the male pronucleus is variable, the other tendencies being more constant, the exact history of the female pronucleus may be said to depend principally upon the appearance of the male pro- nucleus. The earlier that event, the less does the female pronucleus move centripetally, and the less does it assume a nuclear form. In mammals as in echinoderms, the female pronucleus acquires a mem- brane, and lies, when the spermatozoon enters, near the centre. It is very much smaller than the egg nucleus (compare Figs. 36 and 39), and is remarkable for its homogeneous appearance and the ab- sence of nucleoli. In other animals, e.g. Petromyzon, it is merely a cluster of granules. For further details as to the pronuclei, see the following section on impregnation. The time when the polar globules are formed varies, and accord- ing to the animal may be before or after the egg-cell leaves the ovary. In placental mammals the maturation always begins, so far as known, in the ovary, and may be completed there, or it may go on in the Fallopian tube, as Tafani, 89. 1, 114, states is the case in white mice. Our knowledge of the maturation of the mammalian ovum is very imperfect, and rests almost exclusively upon observa- tions on bats and rodents (rabbits, mice, rats, and guinea-pigs), and 66 THE GENITAL PRODUCTS. even on these the observations are very incomplete. See Ed. van Beneden, 80.1, Van Beneden et Julin, Rein, 83.1, Bellonci, 85,1, Tafani, 89.1. III. Ovulation. The process of ovulation, or the discharge of the ovum from the ovary, has to be considered from both the morphological and physio- logical standpoint. The discharge results from structural changes in the Graafian follicle, and these changes continue after the departure of the ovum, transforming the Graafian follicle into a so-caUed cor- pus luteum. Concerning the physiology of ovulation we know al- most nothing beyond the coincidence in some species of mammals of the time of the bursting of the follicle with certain periodic changes in the uterus. Ovulational Metamorphosis of the Graafian Follicle. — The mature foUicle measures some 9 by 13 millimetres, being elon- gated in the same direction as the ovary, but its dimensions are variable. The granulosa is very thin, and its cells show signs of a fatty degeneration. It is probable, I think, that this degeneration progresses to a considerable extent, and involves the loosening of the granulosa cells ; for loose cells, granules, and fragments are found in the liquor foUiculi. The cavity of the follicle is very, large and filled with the fluid, which seems to be under pressure, since it spurts out with considerable force when the follicle is pricked. It is to the pres- sure of the liquid that Coste attributes the rupture of the foUicle. Wal- deyer in Strieker's " Gewebelehre," p. 571, describes a growth of the wall of the follicle, which causes it to form a series of folds which protrude into the follicle ; this ingrowth produces the force that expels the ovum. Unfortunately, Waldeyer does not state on what animal his observations were made; they certainly do not apply to the human species, for there is in man no considerable growth of the follicular wall until after the rupture. The stigma becomes, mean- while, very thin, and finally breaks through. Coste's observations, 47. 1, 172, on rabbits eight or ten hours after the coitus, showed that the rupture is not abrupt but gradual, the membranes of the follicle giving way first, and the peritoneum a little later. When the stigma breaks, the liquor foUiculi, together with the ovum surrounded by the discus proligerus, escapes, and the ovulation, sensu strictu, is com- pleted. The fate of the cells of the tunica granulosa is uncertain, though Benckiser, 84.1, has shown that in the pig they disappear at the time of or soon after the rupture. I consider it probable that they are lost in man at the time the ovum escapes ; it may be that they degenerate ; it must be mentioned that some writers maintain that the granulosa persists and takes part in the further metamor- phoses of the follicle. At the time of the rupture there occurs a hemorrhage of blood into the emptied follicle, and this blood forms a clot which fills up the entire follicle, and is known as the corpus hemorrhagicum. The hemorrhage may vary in amount or even be wanting altogether, as Benckiser, 84. 1, found in 8 cases out of 100 in the pig. Leopold expressly states, 83.1, that when the follicle rup- tures at the menstrual period it is always filled and distended by OVULATION. 67 blood filling it, but when the rupture occurs in the intermenstrual period the hemorrhage is small or altogether wanting ; the presence of blood is therefore not indispensable to the formation of the corpus. When the follicle contains no blood it is filled with a whitish coagu- lum of unknown origin (Coste, 47. 1, 1. 345). The coagulum, whether of blood or not, is rapidly penetrated by tissue which grows into it from the wall of the follicle, accompanied by numerous blood-vessels; the cells of this tissue have two principal forms, His, 65.2,186-187: first, spindle-shaped connective-tissue cells, which lie principally around the blood-vessels ; second, large cells, which contain granules of a pigment, called lutein from its color ; these cells are the lutein- cells, and are the characteristic elements of the metamorphosed clot, to the margin of which they impart a bright yellow color, whence the name corpus luteum. The ingrowing tissue is derived from the inner layer of the theca foUiculi. That the blood-vessels and spindle- cells have this origin has long been the generally accepted opinion, and though the origin of the lutein-cells is under dispute it is prob- able that they arise exclusively from the connective-tissue cells of the theca interna, which begin to enlarge even before the follicle finally bursts, and to charge themselves with lutein granules. Cer- tain writers attribute the origin of these cells to the granulosa either wholly (Exner and Call) or in part (Waldej^er). Peculiar is Beu- lin's view in his Konigsberg dissertation, 1877, that they are derived from the membrana propria folliculi. Benckiser's observations, 84.1, prove conclusively that in the pig the lutein-cells arise exclusively from the theca interna. This view I accept for man also, not only on account of the accuracy of the observations made in support of it (see His, 65.2, and Frommann, 86.3), but also because specimens of my own show that there is no granulosa in the human corpus hemorrhagicum, while the young lutein-cells can be easily recog- nized in the fibrous tunica propria. In consequence of their site of development the lutein-cells and vessels form a band around the coagulum, and owing to its own growth this yellow band soon be- comes folded. The central portion of the corpus luteum long re- mains distinguishable as a separate nucleus. The exact history of the corpus luteum varies according as ovula- tion is followed by pregnancy or not. In the latter case the corpus is entirely resorbed in a few weeks ; in the former it persists until after the birth of the child. We distinguish accordingly the corpus luteum of menstruation from the corpus luteum of pregnancy, or corpus luteum verum of authors. The corpus luteum of menstruation begins with a blood-clot. " The more recent the date of the menstrual flow, the fresher is the clot in the cavity of a ruptured Graafian follicle, and the less change has taken place in its surrounding wall, A few days later the wall be- gins to be enlarged and thickened, and this enlargement within a confined space causes it to become folded upon itself in short zigzag reduplications, mainly at the deeper part of the follicle. As the pro- cess goes on the entire wall participates in the hypertrophy. Its convolutions are extended and multiplied, often in a very compli- cated manner. They project into the cavity of the follicle, encroach upon the central clot, and become pressed against each other, form- 68 THE GENITAL PRODUCTS. ing by their coalescence a thickened, glandular-looking envelope. Previously to the rupture of a Graafian follicle its wall is a uniformly smooth, vascular membrane, not more than one-fourth of a milli- metre in thickness. After the rupture, its thickness increases to one-half a millimetre ; but as the foldings above described grow in number and in depth and crowd against each other laterally, the apparent thickness of the envelope thus formed becomes much greater, and may reach three or even four millimetres, especially at the deepest part of the follicle. In this way there is produced, dur- ing the intermenstrvial period, a corpus luteuin, occupying the sub- stance of the ovary immediately beneath the superficial cicatrix which marks the site of the ruptured follicle. At this time the cen- tral clot is red and gelatinous, while the convoluted wall is of a light rosy hue, mixed with more or less of a yellowish tint. Subsequently the whole structure diminishes in size, and the convoluted wall as- sumes a more decided yellow" (Dalton, 78.1, p. 18). Leopold, 83. 1, distinguishes between the typical and atypical cor- pora, the former being those which start at the menstrual epoch and have a blood-clot, a result probably of the ovarian hyperaemia, the latter beginning intermenstrually and having little or no blood. He says, I. c, p. 399: "The typical corpus luteum appears on the first day as a freshly ruptured follicle, which has filled itself with blood ; on the third day as an enormous blood-cavity ; about the eighth day a thin cortex and a clearer nucleus are marked in the clot. From the twelfth day on, the cortex thickens and becomes folded ; by the sixteenth day it becomes pale-red or yellowish. Toward the twen- tieth day the nucleus shrinks markedlj', the cortical band becomes more and more yellow, and shoots in toward the centre in rays and narrow folds, so as to leave by the twenty-fourth to thirty-fifth day only a small, pale nucleus enclosed in a much-convoluted bright- yellow shell." The corpus luteum of pregnancy begins in a similar manner to that of menstruation, but its growth continues. At the end of the first month its wall is convoluted, much thickened, and of a brilliant yel- low color; the central clot is nearly or quite decolorized and consti- tutes a white or whitish firm 'central mass, which in nearly one case out of three has a central cavity with well-defined, smooth walls. Sometimes a few fine blood-vessels penetrate through the lutein layer. The external convoluted wall continues to grow by encroach- ing upon the clot or white nucleus {corpus albicans), and at the same time the brilliancy of the yellow color diminishes. At term the white nucleus takes up about one-third of the diameter of the corpus and is still distinctly connected with the stigma, so that the lutein wall is interrupted at one point; the corpus as a whole is somewhat smaller than at from two to six mouths. After delivery, resorption goes on rapidly (Dalton, 78. 1). The brilliant yellow is especially characteristic of man ; in sheep the pigment is pale brown, in the cow dark orange, in the mouse brick-red, in the rabbit and pig fiesh-colored. Lutein is a crystal- line body, soluble in alcohol, ether, chloroform, and benzol, but of its chemical nature we have no exact knowledge. Physiology of Ovulation.— Concerning this subject and also IMPREGNATION. 09 concerning the functions of the corpora lutea, we possess scarcely any knowledge. We have to consider only the relation of ovulation to menstruation and coitus. Coste, 37.1, 454, 455, first showed that the discharge of the ova coincided with the period of heat in various animals. This was soon confirmed by Raciborski, 44.1, and since then by numerous ob- servers. Pouchet ("Theorie positive," etc.) attempted to prove that this is also true of the human species, the menstrual period being taken, correctly, as the equivalent of the rut. In this attempt Pou- chet has had many followers, especially among gynaecologists. Coste, however, demonstrated long ago, 47.1, 232, that the bursting of the Graafian follicles may occur before or after menstruation, though it is most apt to occur during the menses. This conclusion of Coste's has been fully confirmed by Leopold, 83.1, who made a very careful examination of twenty -five pairs of ovaries from women whose menstrual history was accurately known. It was Coste again, 47. 1, 183-185, who proved experimentally that coitus hastens in the rabbit the rupture of the Graafian follicles. Unfortunately he gives only two experiments, and since then tliej' have not been repeated, so far as I am aware, either upon rabbits or other animals. But there are statements by many authors, Bary, Reichert, Hensen, 88.1, 58, Van Beneden, 80.1, etc., to the effect that in the rabbit after coitus during heat the follicles are found to have burst during the tenth hour. IV. Impregnation. Impregnation is the union of the male and female elements to form a single new cell, capable of initiating by its own division a rapid succession of generations of descendent cells. The new cell is called the impregnated or fertilized ovum. The production of cells from it is called its segmentation. For the theory of the relation of the elements to one another and to cells, see the following section. In all multicellular animals, impregnation is effected by three suc- cessive steps: 1, the bringing together of the male and female ele- ments ; 2, the entrance of the spermatozoa into the ovum and forma- tion of the male pronucleus ; 3, fusion of the pronuclei to form the segmentation nucleus. We proceed to consider these steps in their order. 1. The Bringing Together of the Sexual Elements.— This is effected in a great variety of ways, which, however, fall into two groups according as the impregnation is effected, a, outside the body of the mother; or, b, inside. The simplest manner is the dis- charge of the male and female elements at the same time into the water, leaving their actual contact to chance, the method of the osse- ous fishes for the most part and of many invertebrates. An advance is the copulation of the Anura (frogs, etc.) ; the male embraces the female, and, as the latter discharges the ova, ejects the sperm upon them. In the higher vertebrates the seminal fluid is transferred from the male to the female passages during coitus. The physiol- ogy of this complicated function does not fall within the scope of this work. ' 70 THE GENITAL PRODUCTS. For a long time it was not known how the semen fertilized the ova ; the problem was fruitful of fruitless speculation. The first step toward gaining actual knowledge was the discovery of the possibil- ity of artificial fecundation by Jacobi in 1764. Spallanzani was the first to take advantage of this and to show that fecundation implied a material contact of the semen with the ova, and thus to set aside De Graaf's notion of the " aura seminalis." But not until fifty years later did the memorable experiments of Prevost and Dumas {Annales des Sciences Naturelles, 1824) es- tablish the fact that the spermatozoa are the essential factors of fer- tilization. Again, a little over fifty years later, Hertwig and Fol showed that one spermatozoon suffices to impregnate an ovum. We have then to consider how the spermatozoon, after the semen has been transferred to the female, attains the ovum. They are found in mammals after copulation in the vagina and even in the uterus, but it is not clearly ascertained how they get beyond the vagina. It is probable that they travel through the female passages partly by the movements thereof, partly by their own locomotion, and enter the Fallopian tubes, though why or how is really unknown, and pass upward to meet the ovum. They are found in considerable numbers in the Fallopian tubes. The ovum meanwhile travels down the ovi- duct, it probably being impelled by peristaltic movements of the duct. The meeting-point or site of impregnation in placental mammals is about one-third, perhaps one-half, way down from the fimbria to the uterus. It is remarkably constant for each species. Nothing positive is known as to the site of impregnation in man ; but there is no reason to suppose, as is unfortunately often done, that the site is variable or different from that in other mammalia. 2. The Entrance of the Spermatozoon into the Ovum and Formation of the Male Pronucleus. — With our present knowledge, the assumption appears unavoidable that the ovum exerts a specific attraction upon spermatozoa of the same animal species. We observe, in fact, when artificial fecundation is employed, that the spermatozoa swarm around the ova as if held by an irresistible impulse. This phenomenon occurs with every class of animals, even in mammals, whose freshly removed ova were examined on a warm stage under the microscope (Rein, 83.1). Stassano, 83.1, has main- tained that the eggs of echinoderms do exert such an attraction, and also a similar but less strong attraction upon the spermatozoa of allied species. But since the brothers Hertwig, 85. 1, have found by their experiments with sea-urchins that hybrid impregnation takes place more readily after the ova have been kept awhile, Stassano's view involves the further assumption that the specific nature of the attrac- tion fades away during a few hours. Very suggestive in this con- nection is Pfeffer's ("Untersuch. Bot. Inst." Tubingen, Bd. I., Hft. 3, 1884) discovery that certain chemical substances may attract moving spores, etc., to definite spots. It is conceivable that the ovum may draw the spermatozoa toward itself by chemical influ- ence, acting as an attracting stimulus. There may be mechanical devices to facilitate the entrance of the spermatozoon ; this is, perhaps, generally true of all ova with micro- IMPREGNATION. 71 pyles serving for the passage of the spermatozoa. A careful study of such devices in the cockroach has been made by J. Dewitz, 85. 1, who found that the motions of the spermatozoa of this insect are peculiar and adapted to increase the probability of their passing through one of the micropyles of the ovum. In ova without micro- pyles, among which those of mammals are included, the sperma- tozoa may, so far as we know, penetrate any part of the envelopes. In the rabbit (Rein, 83.1), about ten hours after coitus, the ovum is found nearly half-way through the oviduct and surrounded by many spermatozoa — perhaps a hundred, more or less. These are all, or nearly all, in active motion, for the most part pressing their heads against the zona radiata. Sev- eral of them make their way through into the interior of the ovum. According to Hen- I sen, 76.1, only those spermatozoa which en- 1 ter the zona along radial lines can make their way through; those which take oblique courses remain caught in the zona. Fig. 40, and may still be seen there during segmenta- - tion. As the ovum at this time is already , p» f° -Ovum of a Rabbit; „!, . 1,1 . ,, ,,'' taken from the middle of the luily matured, there is a space between the oviduct about eighteen hours contracted yolks and the zona. In this space, SudeSs"faire?d/ffSf the as well as in the zona itself, several sperma- ESLl^,~?rm^at?ioTiifb"otii tozoa may be observed at scattered points, in and within the zona. After The female pronucleus is present, having ^"^''^^ been re-formed since the expulsion of the second polar globule from the ovum while in the ovarJ^ One spermatozoon gets into the yolk proper, and its entrance apparently prevents the penetration of other spermatozoa — how is undetermined. The tail of the sperma- tozoon soon disappears, while the head enlarges, probably by the imbibition of fluid from the surrounding yolk, and thus becomes a nucleus-like body — the male pronucleus. The passage of the spermatozoa through the zona was first discov- ered by Martin Barry in lf-i43, and although his statement was re- ceived with considerable hesitation by his contemporaries, it has since had competent confirmation repeatedly. Warneck (Bull. Soc. Nati., Moscou, XXIII., 90) is said to have been the first (1850) to see the two pronuclei, but their significance was not perceived. The nature of the male pronucleus was first recognized by Oskar Hert- wig, who traced its genesis in the ova of echinoderms from the sper- matozoon. The fact that the male pronucleus is the metamorphosed spermatozoon has since been confirmed by Selenka ("Zool. Stud.," I.); Ed. van Beneden, 83.1; Nussbaum, 84.2; Eberth, 84.1; Platner, 86.1, and others. Although a number of the spermatozoa make their way into the perivitelline space, probably always one alone normally enters the yolk to there form a pronucleus. The best observers are agreed upon ithis point, and in all species the observations upon which have cov- ered the whole series of steps in the impregnation, there has been found in normal cases always a single male pronucleus. Schneider's statements to the contrary have been definitely corrected. Bambeke, 72 THE GENITAL PRODUCTS. 76.2, C. Kupffer, 82.1, and KupfEer und Benecke (" Befruchtung Neunauge," 1878), have observed that several spermatozoa actually enter the yolk in batrachians and Petromyzon. Hertwig, however, found only one male pronucleus in the frog, and there has as yet been no evidence adduced that several spermatozoa are concerned in the final phases of impregnation. Fol observed that star-fish eggs are nor- mally impregnated by one spermatozoon ; but if they are exposed to the action of carbonic acid they may, while so poisoned, be impreg- nated by several spermatozoa, and the subsequent development in this case is abnoi-mal : apparently each pronucleus becomes a separate centre of development. The manner in which additional spermatozoa are excluded after the first has entered is still under discussion. In cases where there is a single micropyle, which is used for entry, it is possible that a portion of the first spermatozoon may remain to close the passage, or that in going through it sets in action some mechanism by which the opening is automatically shut. Where there are several or many micropyles, as in some insects, or where the envelopes may be pierced at any point, as in mammals, there must be some other device. Fol has maintained that this is found in the star-fish in the rapid formation of a membrane around the yolk immediately after the entrance of the first spermatozoon ; but Hertwig affirms that this membrane pre-exists. Selenka {Biolog. Centralbl., v., 8) describes the fertilization of the ovum of a nemertean worm ; several sperma- tozoa enter within the vitelline membrane ; the yolk contracts slowly. After a time the two polar glob- al y' ules are expelled, and before ■\^\jM;^~ M r^f- they separate from the yolk one w~..i.^jM^^I-,-^^^ 'l^.f^ / spermatozoon passes into the #~'' ^UniJL^^ .'^^■'z ^ yolk between them ; the globules ,"~!^ then break off and are knocked \ about by the spermatozoa in the ^ ^^ ^ peri vitelline space. In this case \ ' jj ^" there seems to be a portal opened ^^^L ^t J'^®* long enough for one sper- ^^ -~a^ ^,„ >. ^ matozoon to enter. As the phe- ''^Hsk^:— '-^ " nomenon to be explained is com- FiG. 41.— Anterior pole of the ovum of the Petro i-nnn to all n-ua i+a r-anaQ+inn ia myzon, with a spermatozoon, sp., entering the l"On XO aii OVd, Its CaUSaXlOn IS micropyle, mi; p. v. perivitelline space ; z, zona presumably fundamentally iden- pellucida; a, pathway to female pronucleus, ill-i- ii n 3 ±t • 2/i:.,yoik. After caiberia tical m ail cases. Boyond this surmise our present knowledge does not permit us to go. The hypothesis may be suggested that the attractive power of the ovum is annulled or weakened by the formation of the male pronucleus. This hypothesis was first sug- gested by Minot (Buck's " Hdbk.," IV., 6), and has since been elabo- rated by Whitman, 87.3, 239-243. It is probable that the tail of the spermatozoon, when that appen- dage exists, disappears within the yolk. In a land-snail. Avion, Platner, 86. 1, has traced this process very clearly. Only a portion of the tail enters the yolk, but the part within acquires the property of staining readily, and so may easily be observed. He reports that the head and tail separate ; only the head conjugates with the female IMPREGNATION. 73 pronucleus, while the tail still remains distinct even after segmen- tation has been initiated, Fig. 44. The disappearance of the tail has been recorded by most observers. As Hertwig says {loc. cit., p. 23), aU these careful observations yield the assured conclusion that the head of the spermatozoon, and the head only, becomes the male pronucleus. While the spermatozoon is passing through the ovic envelopes, act- ive ahanges occur in the yolk. Of these the most constant, as well as the most obvious, is the formation of a slight protuberance on the surface of the yolk, rising up toward the spermatozoon. This pro- tuberance may remain, as in echinoderms, until the spermatozoon meets it and by penetrating it enters the ovum, or 'it may retract before the spermatozoon passes through the envelopes, and even with- draw, as in Petromyzon, Fig. 41, so far from the advancing sperma- tozoon as to form into a depression on its own surface, Fig. 41. The protuberance lasts only a few moments. In Bufo, according to Kupffer, several spermatozoa enter the yolk and a protuberance rises toward each one, as if the yolk were actively striving to reach the male element. The protuberance always consists of fine granu- lar protoplasm, which contains no deutoplasm, and is closely con- nected with the nucleus. The size of the protuberance is variable. In Petromyzon there is a large hummock of protoplasm, which con- tains the nucleus and in which both pronuclei form and unite; dur- ing these processes the protoplasm of the hummock is separated from that of the rest of the ovum by a special membrane, which disap- pears immediately after the pronuclear copulation. While the two pronuclei are meeting the hummock flattens out and the protoplasm forming it travels centripetally together with the pronuclei (Boehm, 88. 1, 650, 051). Whether the hummock in Petromyzon is homologous with the much smaller protuberance in other ova I am unable to say. The relative size of the two pronuclei varies considerably in differ- ent species, and is probably a secondary and unimportant relation. Each pronucleus when it first appears is small and gradually en- larges, apparently by the imbibition of fluids from the surrounding yolk. Now the time when the spermatozoon enters the yolk may be either after or at some stage during maturation of the ovum. If it enters early, as in Limax (Mark), the male pronucleus en- larges equally with the female. Fig. 42; but if late, as in the allied Arion (Platner), then it appears, Fig. 44, considerably smaller than the already swollen female pronucleus. O. Hertwig, in his third paper on maturation, p. 171, first gave this explanation and pointed out that in the star-fish (Asterias) , if the impregnation is prompt, the male pronacleus becomes as large as the female, but if impreg- nation is delayed for four hours the male pronucleus remains much the smaller of the two. Again in Hirudinea, Fig. 42, many Mol- lusca, Nematoidea, etc., impregnation usually takes place before the formation of the polar globules is completed, and the male pronu- cleus is accordingly as large as the female. In Echinus, on the other hand, where the polar globules are formed in the ovary, the male pronucleus is always small. Concerning the path of the male pronucleus we possess little infor- mation. 0. Hertwig and Bambeke have found that in certain am- 74 THE GENITAL PRODUCTS. phibian ova the spermatozoon leaves a trail {Pigment- Strasse) apparently by carrying along with itself some of the pigment gran- ules from the surface of the ovum. Eoux, 87.1, has studied this path in the frog's ovum and finds that it consists of one limb, the line of penetration through the cortical layer of the ovum, w^hich is nearly perpen- dicular to the ovic surface, and a second limb usually forming an angle with the first and leading directly to the female pronu- cleus. The force which draws the pronuclei to- gether is unknown. We can only say that, as Whitman has thoughtfully expounded, 87.3, there is a relation between the nu- FiG. 42 -ek« of Nepheiis, three cleus and the protoplasm of the ovum, such hours after layins: m, male /, that the nuclcus tcnds to take a Central posi- g1Sfufes^'°A£ter"o.' &rtwig!°^'^ tion. When the polar globules are formed the nucleus becomes repellant and drives itself centrifugally ; but the protoplasmic attraction remains and draws in the spermatozoon. Subsequently both pronuclei are at- tracted toward the centre and toward each other, and the curved routes the pronuclei often take are the resultants from these two attractions. 3. Fusion of the Pronuclei. — Each pronucleus is usually found surrounded by a space a little clearer than the rest of the yolk. Usu- ally the yolk around this clear space presents a radiating appearance, which is known as the aster. Fig. 43; but this appearance is not con- stant, nor is it known how it is caused. Frommann, 89. 1, 395, states that in the egg of Toxopneustes the astral rays are formed by very irregular rows of angular granules, which may lie separately, or be strung together by tine threads, or like a row of pearls, and are irregularly connected by cross-threads. The great regularity usu- ally pictured is purely diagrammatic. As the granules described by Frommann are part of the reticulum of the ovum, we may say that the astral figure results from the arrangement of the protoplasm. Mark, 81.1, was unable to see it in Limax, and Rein, 83.1, could not detect it in the rabbit. In Arion, as also in Petromyzon, according to Boehm, 88.1, apparently only the male pronucleus has an aster, Fig. 44. At one time it was assumed that the pronuclei acted as centres of attraction upon the j'olk, and that the asters were due to their direct influence ; but since, as in Arion, Fig. 44, the pronucleus may move away while the aster remains behind, it follows that the relations are more complex than this assumption indicates, since the aster exhibits a certain independence of the pronucleus. This is confirmed by Flemming's observations {loc. cit., p. 19), that when the asters first appear in echinoderms, the centre of radiation is not the pronucleus itself, but a clear space just alongside. Frommann, 88.1, 396, modifies this statement by recording that the position of the centre of the male aster varies in Toxopneustes and may be at one side or the other of the male pronucleus or coincident with it. Boehm, 88.1, 650,Taf. XXV., Fig. 30 d, notes the same peculiarity in the eggs of Petromyzon. These statements recall the fact that the IMPREGNATION. asters in indirect cell-division sometimes radiate from a clear spot at the tip of the spindle. Some writers have considered the aster an expression of magnetic force within the ovum — a fanciful notion without any evidence to support it. In the rabbit, Rein, 83.1, both pronuclei lie at first eccentrically, but they move toward each other and toward the centre, meeting, however, before the central position is attained. As they near one another, both pronuclei perform active amoeboid movements ; after they meet they still continue their amoeboid movements and move together to the centre of the ovum ; one of the pronuclei assumes a crescent shape and embraces the other, Fig. 45. At this time the yolk displays a radiate arrangement ; from analogy with other ani- mals it must be assumed that the two pronuclei fuse into a single nucleus, which is therefore an hermaphrodite structure, and which, after a certain period of repose, itself divides and so begins the cleav- age of the yolk. The place where the pronuclei meet varies. Apparently the female pronucleus of itself moves to the centre or near the centre of the ovum ; also the male pronucleus approaches the female as speedily as possi- ble. If now impregnation occurs early, the two pronuclei meet peripherally ; if late, they meet near the centre. In the former case they move together, as in the rabbit (Rein), to a central position. The observations so far made indicate that after they meet the pro- nuclei both perform active amoeboid movements, which continue for several minutes. Selenka maintains that the female pronucleus sends out processes which embrace the male pronucleus, but this has not been confirmed. Finally, the two pronuclei unite, but the process of union is very obscure, never having been satisfactorily observed. Apparently the membranes of the pronuclei, where the two are in contact, are dissolved away and the contents mix. The best account known to me of the fusion of the pronuclei is that given by Boehm in his memoir, 88.1, on Petromyzon. The outline of the female pronucleus is still diffuse a quarter of an hour after fertilization. The head of the spermatozoon (male pro- nucleus) breaks up into four, more rarely five, granules. The female pronucleus moves centripetally, and acquires a dis- tinct membrane. The pronuclei meet, the male granules having meanwhile multiplied by division. About this time the female pronucleus also breaks up into two pronuclei Kf^-r ,11 1 J. Arminfl each D] granules. We then have a clear spot which is the centre of an astral radia- tion, next this a bunch of male granules (Boehm's Spermatomeri- ten), and next that a bunch of female granules (Boehm's Ovomeri- ten), the whole making an elongated body lying at right angles to the radius of the ovum. Three hours after fertilization the two bunches are fused together and are no longer distinguishable. Each " Merit " consists of a body containing one or two chromatin specks Sagitta with .. _ ,._ . . After O. Hertwig. Around each pronucleus is shown the aster. 76 THE GENITAL PRODUCTS. {Microsomen) . In the Crustacea, according to Weismann and Ischi- kawa, 88.4, the two pronuclei, when they meet, resemble ordinary membranate nuclei ; where they come in contact with one another the membranes dissolve away and the contents of the pronuclei mingle. In Ascaris the process is more complicated. We may say, there- FiG. 44.— Two ova of a land-snail, Arion. After Platner. The ova are irregular in shape, as at this stage they are still in utero, and mutually compressed. A shows the segmentation nucleus, m, just formed; the two large "Karyosomen " in it are derived from the male pro- nucleus : the male aster still remains, as. B snows the commencing change of the segmentation nucleus into the first spindle. In both ova the tail, sp. of the spermatozoon is distinguishable. fore, that the fusion of the pronuclei is the essential phenomenon, and the method of the fusion is secondary in importance. Another point deserving mention is the rotation of the copulated nuclei. See Frommann's article on " Befruchtung " in " Eulenberg Cyclopsedia," p. 568. Now since the head of the spermatozoon is developed chiefly out of the chromatin of the nucleus of a spermatoblast, it follows that impregnation is essentially the addition of chromatin to the nucleus (female pronucleus) of the mature ovum. After the union of the pronuclei follows a period of repose, during which the yolk enlarges until it again fills or nearly fills the space within the zona radiata ; a little room is left, which is chiefly occupied by the polar globules. The significance of the contrac- tion of the mature and the expansion of the impregnated yolk is unknown. In certain cases the parts of the segmentation nucleus which are derived from the male pronucleus remain distinguishable. This is notably, according to Platner, the case with Arion. The segmentation nucleus contains a number of nucleolus-like bodies (Karyosomen of Platner, Fig. 44, A, n), with a distinct round outline, and a few granules of chromatin. These bodies are of two kinds. Fig. 44 ; the smaller and more numerous are produced by the female pronucleus, while the two larger ones arise from the division of the head of the spermatozoon. Fig. 45.— Ovum of a rabbit sev- enteen hours after coitus with, the pronuclei about to conjugate* pg, polar globules; s, zona pellucida. After Rein. THEOKY OP SEX. 77 In the later stage, when the nucleus is changing into the first seg- mentation spindle, Fig. 44, B, the two large male •' Karyosomen" are still distinct, and have each their chromatin gathered in little particles around the peripherJ^ Edouard van Beneden, 83.1, goes even further, stating that in Ascaris the chromatin from the two pronuclei can be distinguished in the nuclei of segmentation, and that when it divides, both the male and female chromatin loops divide also, so that the resulting nuclei are truly hermaphroditic. V. Theory of Sex. Sex is a term employed in two meanings, which are often confused but which it is indispensable to distinguish accurately. Originally sex was applied to the organism as a whole in recognition of the differentiation of the reproductive functions. Secondarily, sex, to- gether with the adjectives male and female, has been applied to the essential reproductive elements, spermatozoon and ovum, which it is the function of the respective sexual organisms (or organs) to pro- duce. According to a strict biological definition, sexuality is the characteristic of the male and female reproductive elements, and sex of the individuals, in which those elements arise. A man has sex, a spermatozoon sexuality. Sexuality is primitive and essential, and sex is dependent upon it. We have to consider, 1st, the nature of sexuality ; 2d, the origin of sexuality ; 3d, the nature of sex. Nature of Sexuality. — -The essential facts of sexual reproduc- tion are: That two bodies, partaking more or less plainly of the character of cells, fuse together into a single body, which seems in every known respect to be homologous with a uninucleate cell, and which undergoes a series of divisions (segmentation of the ovum) resulting in the production of an increasing number of new cells. In all the higher animals (and plants) the two bodies are obviously different. In all metazoa one body, the ovum, contains a store of nutritive material and has a special envelope of its own ; the other, the spermatozoon, is small and provided with means of active loco- motion ; the details of their fusion, which is known as the fertiliza- tion or impregnation of the ovum, have been described. The only hypothesis, as to the nature and mutual relations of the ovum and spermatozoon which rests, such is still my opinion, on much basis of fact is Minot's "Theory of the Genoblasts," 17.47. This hypothesis is based upon three categories of facts : 1st, Sexual reproduction is effected by the union of a male and female element, which produces a cell; this cell is, therefore, hermaphroditic, or, perhaps, one should say, asexual or neuter, since it is neither male nor female. 2d. When the cell, which gives rise to the female ele- ment matures into an ovum, it undergoes a remarkable process of unequal division, known as the extrusion of the polar globules. In other words, the cell divides into three bodies — a, two polar globules ; ft, a single female element. In some cases the polar globules sub- divide further. 3d. When a cell divides into the male elements there remains one cell which does not form a spermatozoon. In mammals it is probable that the parent-cell divides into three cells, one of which, h, remains to form the base of a Sertoli's column, and 78 THE GENITAL PRODUCTS. two of which, a, subdivide further to produce the spermatoblasts and ultimately the spermatozoa. Unfortunately our knowledge of the development of the spermatozoa is extremely unsatisfactory, no two authors agreeing, so that extreme caution is necessary. There are, however, reasons for thinking that the statements just made in re- gard to mammalian spermatogenesis correctly specify the essential steps, and it is probable that the essential steps are the same through- out the animal kingdom. Assuming then that the view of sperma- togenesis here adopted is correct, our further deductions from the premises are almost self-evident. In the cells proper both sexes are potentially present ; to produce sexual elements the cell divides into its sexual parts, the genoblasts ; in the case of the egg-cell the male polar-globules are cast off, leaving the female ovum (oospore of Bal- four) ; in the case of the sperm-cell the male spermatoblasts, which by the hypothesis of Minot are homologous with the polar globules, multiply considerably, and their descendants give rise by further spe- cialization (in mammals of their nuclei) to the male elements, while the parent-cell, or homologue of the oospore, atrophies. In both cases the sexual cell separates into a single female element or thely- hlast, and probably two male elements or arsenoblasts, which are capable of multiplication by division ; but in one case it is the thely- blast, in the other the arsenoblast, which is preserved, differentiated further, and utilized. To make a complete cell there must be a union of the male and female, and this is accomplished by " impreg- nation of the ovum." Minot's hypothesis is strictly morphological and offers us no in- sight at present into the physiological aspects of sexuality. It has been adopted by Balfour, and Ed. van Beneden, neither of whom cite Minot. Since the theory of the genoblasts was first advanced in 1877, it has been confirmed by important discoveries, especially by the series of investigations which have proven that polar globules, as stated in the section on the maturation of the ovum, occur in all classes of the animal kingdom, and, secondly, through the investi- gations on the relation of the polar globules to parthenogenesis. The general, may we not say the universal, occurrence of the formation of the polar globules as a necessary step in rendering the ovum capa- ble of impregnation, is, of course, a very important confirmation of the theory, since the theory assumes that the production of the polar globules is the essential step in converting the egg-cell into an oospore or thelyblast. Minot, in his original article, briefly indicated the application of his theory to parthenogenesis, and the question was more and ably discussed by Balfour in his "Comparative Embrj'ology," I., 63-64. In his article of 1883, 47, 367, Minot is more explicit. He says : " If one assumes that the ovum becomes female by the removal of the polar globules, then it must remain asexual so long as no globules are formed. If one further assumes that no polar globules are formed in ova, which develop parthenogenetically, then these ova would remain simple cells, and their reproductive process would depend on ordi- nary cell-division. If the globules are developed then impregna- tion is an unavoidable preliminary of further development." In other words, parthenogenesis is only an extreme case of asexual repro- THEORY OF SEX. 79 duction and in nowise the development of a female element (oospore or thelyblast) without impregnation. The correctness of this view has since become extremely probable through the observations of Blochmann, 87. 1, 88. 1, of Weismann, and of Weismann and Ischi- kawa, 88.2, 4, who find that in parthenogenetic ova there is only one polar globule formed, while in eggs requiring fertilization there are two. Now by Minot's theory the cells must be hermaphroditic in order to develop, and the egg-cell becomes a thelyblast by the ejection of two polar globules; if, therefore, one polar globule is removed and the other not, the egg-cell retains part of its male con- stituent. The significance of the two polar globules has already been discussed, p. 65 ; Weismann's interpretation is considered in the following section on Heredity. To my theory of the genoblasts, I feel justified in making an essential addition — namely, that sexuality is a relation of substances or forces, and not dependent on any special substance. The chief evidence in favor of this assumption is the fact that in all male ele- ments the proportion of protoplasm to the nucleus is small, while in female elements (thelyblasts) it is small; and, moreover, to produce spermatozoa there is an excessive growth of the nuclei, while to pro- duce ova there is an excessive growth of the protoplasm. It is remarkable, as Minot has demonstrated (Address, Proc. A. A. A. S., 1890), that a relative increase of protoplasm is the anatomical char- acteristic of senescence. The ovum resembles an old cell, the sper- matozoon a young cell, and these resemblances cannot be considered fortuitous. There is no material basis of sexuality in the sense that there is any visible male or female substance known to the biologist, nor is it probable that a male or female substance exists. The func- tions of life, according to our present conceptions, are not each con- nected with particular chemical compounds or with particular visible constituents of the cells, but rather depend upon the complex inter- relations of numerous different substances, which enter into the com- position of the cell. There are certain functions which are connected more intimately with one part than with another — as, for instance, contractility with the protoplasm, heredity with the nucleus; but even in these cases we cannot say that the functions in question could go on without the interplay of the other portions of the cells. The genoblasts contain nuclear substance, protoplasm, and enchy- lema, and we can ascertain the sex of a genoblast only l3y observing its history, not by any direct test. It is probable that male or female sexuality is au intracellular relation of parts, some modification of the interplay of forces within the cells, and for the present this view must hold against the opposite view that there is a male matter and a female matter. Several interpretations of the polar globules have been advocated, which are incompatible with Minot's theory. The first of these is that of Whitman,* who, in his first article on the development of Clepsine, 78.2, p. 48, maintains that a series of cell generations is produced by a series of divisions, and the separation of the polar * Compare also O. Hertwig. 90. 1. 80 THE GENITAL PRODUCTS. globules is merely the last of these divisions. Inasmuch as this view overlooks the fact that polar globules are part of the process of maturation and that no ovum can be impregnated until they are formed, and the further fact that the products of division (globules and oospore) are extremely unlike, while in ordinary divisions the two daughter-cells have close resemblance to one another — inasmuch as these fundamental facts are overlooked, it seems to me that Whit- man's explanation cannot be adopted. Allied to Whitman's view is that of Biitschli, 84.1, who, starting from the idea of a sexual colonj', such as is found in certain unicellu- lar animals (Flagellata) , considers that the tendency to form such colonies is preserved in the metazoa, and shows itself in the bundle of spermatoblasts and in a more rudimentary form in the egg-cell; forming a colony with its two polar globules. The essential objec- tion to this view is that it overlooks the fact that the divisions of cells to produce the sexual products are divisions into unlike bodies, while in the sexual colonies of the Flagellata the divisions are, so far as known at present, into like cells. O. Hertwig's criticisms, 90.1, against Minot are based on the study of the differentiation of the sexual elements of Ascaris. He overlooks the fact that the theory of Minot depends on the origin of the sexual elements, not on their differentiation; yet nothing is known as to the origin of the genoblasts in Ascaris. Besides the theory of sex already discussed, there are three others which must be noticed. The first of these has been advanced by Sabatier, and defended by him in a series of articles, several of which have been reprinted, making in their reprinted form the fifth volume of Sabatier's '' Travaux." * Sabatier considers that the cells are neuter or hermaphroditic, agreeing in this respect with Minot, and that the casting off of the male portion converts the cell into a female element, and vice versa, but he goes farther than Minot in attempting to specify which parts of the cell are male and which are female. He directs attention first to the fact that in certain inver- tebrates there is a central mass (Bloomfield's hlastophore) , to which there are attached spermatoblasts or spermatozoa. He endeavors to prove that this is the primitive method of spermatogenesis, and con- cludes that the male element is peripheral, and the product of a cen- trifugal action. He directs attention, second, to the various products that are thrown off from the cell, which ultimately forms the ovum. Summarizing his conclusions in regard to the egg (Z. c. V., 202-303) he says: "If we recapitulate now the various groups of globules which are eliminated from the ovule, commencing at the asexual cell-stage of its life up to the moment when it attains the complete dignity and signification of an egg, we see that there may be : " 1st. Globules precoces on du debut, which become usually the elements of the follicle and give, so to speak, the first impulse to the march of the cell toward sexuality. " 2d. Globules tardifs, which are at times formed well before the epoch of maturity, but are eliminated at a late period, and sometimes very near the maturity. They are all formed, as are the globules * "Travaux du Laboratoire de Zoologie de la Faculty des Sciences de Montpellier et de la Sta- tion Zoologique de Cette." Ire S6r., 5me vol. THEORY OP SEX. 81 precoces, by simple differentiation in the midst of the protoplasm and without karyokinetic phenomena. " 3d. Globules, which are contemporary with the period of com- plete maturity, and of which the elimination accentuates in the egg a very pronounced attraction for a male element coming from another cell, or even from another organism. These are the globules de maturation parfaife. Most of these globules result from phenom- ena of cellular division, and form the polar globules properly so-called. " From this quotation it will be clear that Sabatier classes together the follicular cells surrounding the ovum, the non-cellular masses excreted from the egg-cell during its development, and the polar globules. All of these are — so he maintains — thrown off from the central ovum, hence he concludes that the female element is central and the product of a centripetal action. In brief the male element represents a centrifugal force, the female element a centripetal force. A. Prenant has adopted a theory which is apparently a modifica- tion of Sabatier's, but until his memoir is published (Journ. de VAnat. et Physiol., 1892) discussion of his theory must be deferred. I am unable to accept Sabatier's theory for many reasons, of which the following may be mentioned : 1st. It cannot be shown that the differentiation of the spermatozoa does occur typically at the periph- ery ; on the contrary, in the great majority of cases, it is distinctly polar, since it takes place at the inner end of an epithelial cell. 2d. It is impossible to maintain a homology between cells and masses which are not nucleated at any period of their history, and Sabatier's views as to the maturation of the ovum oblige us to draw such an homology. 3d. Sabatier, to establish the centrifugal removals, which produce the ovum, relies largely upon the history of the glob- ules tardifs, which, therefore, must by his hypothesis be male. He bases his defence largely on observations on the " testa-cells " of Ascidians, which he considers to belong under the head of globules tardifs; but these observations have been called in question by Fol (Recueil Suises Zool., ISTo. 1),* so that there is doubt as to one of Sabatier's chief foundations. Now some of these globules — sup- posed to be male— contain no nuclear substance, yet all the sexual elements, which we know positively to be such, do contain nuclear substance. Balbiani's theory, 79. 1, is the exact inverse of the two previously mentioned ; for him every sexual element is the product of the copu- lation of two elements : 1st, the epithelial cells of the follicle, which are male; 2d, the Urei, which is always female. Balbiani has not observed any such copulation, nor has he any valid indirect evi- dence of it to bring forward; on the contrary, he disregards in several respects what others consider elementary principles of his- tolog3'. Nussbaum's theory appears to me valuable and suggestive. It was first advanced, so far as I know, in 1880, though similar con- ceptions are to be found in earlier writers. Nussbaum, 80. 1, starts with the conjugation of two similar unicellular individuals, as occurs in certain protozoa ; the two individuals fuse, and after fusion *For Sabatier's answer see same Recueil, No. 3, 83 THE GENITAL PKODtCTS. divide into successive generations of cells. He next points out that in the higher animals all the sexual differences are secondary not only in the so-called "secondary sexual characteristics," but also in the sexual organs themselves. He then goes on to emphasize the presence of the sexual cells {Ureter of German authors, Hamann's Urkeimzellen) ixi the embryo, and maintains that as these both give rise to the sexual products the ovum and spermatozoon are strictly homologous cells. He writes, p. 106 : " There come together during impregnation accordingly not two heterogeneous elements which complement one another and together form a whole, but rather there come together two homologous cells, of which one to facilitate conjugation is transformed into a more movable body ; the other is laden with nutritive material, and is furnished with protective de- vices." And again, y>. 113: "The differentiation of sex is not the transmission of two originally united functions to the differing descendants of a common original Anlagej it is rather the variation of homologous cells for the better achievement of their conjugation." The sexual elements, according to Nussbaum, are cells which are set apart for reproductive functions from the rest of the cells of the bod}', and there is no primary difference between male and female. He does not consider in any Avay the significance of the polar glob- ules or Sertoli's columns, and therefore does not argue directly against Minot's theory. His generalization that separate cells alike in character are set apart early in embrj'onic development to form both the male and female elements is a very important one, and has been adopted by embryologists. Weismann accepts it and applies it to his theory of heredity, and it has received a valuable confirmation in Hamann's paper, 87.1. But this generalization leaves the ques- tion of the final differentiation of the Ureier into sexual elements untouched, and is not necessarilj- in any way in conflict with the conception of that differentiation advocated by Minot. It seems to me, therefore, that although Minot's hypothesis cannot be proven at present, yet there is no other hypothesis of sex having nearly as strong evidence in its favor. Origin and Objects of Sexuality. — The origin of sexuality is involved in much obscurity. In the lowest unicellular organisms there is certainly no clear sexual differentiation, and some biologists assert that there is nothing comparable to sexual reproduction, but the observations are far too imperfect at present to justify any such assertion. The question involved is, whether sexualitj' is coexten- sive with life or not ; in the latter case it is the result of evolution from asexual organisms, and is a secondary and not a primary or essential characteristic of life. The problem is, therefore, a funda- mental one, but we cannot hope for its solution until our knowledge of the lowest organisms is greatly extended. The precursor of the sexual process is undoubtedly to be found in the conjugation of two similar cells, which fuse into a single organ- ism, as occurs in certain cryptogamous plants and among the pro- tozoa, notably the rhizopods. In the next stage the cells which fuse together are obviously different, as in the Flagellata. If now we pass to the colonies of the Flagellata we find that certain cells only act as conjugators, and thus we approach the disposition of the mul- THEORY OF SEX. 83 ticelkUar animals, Metazoa, which have bodies composed of cells, certain of which produce the sexual elements, and these elements conjugate. In conjugation and impregnation alike the process is the fusion of a nucleated protoplasm with another nucleated protoplasm of different origin. In plants also, as we ascend from the lower to the higher forms, we find the differences between the conjugating bodies to increase: thus in zygophytes the conjugating cells are alike, in phanerograms the pollen and ovicell are unlike. The ques- tion arises whether the conjugation of the like or of the unlike pro- toplasms (or, in other words, of similar cells, or of genoblasts) gives the clew to what is essential. Is the dissimilarity of the conjugating bodies essential? If Minot's theory of the genoblasts is correct it is probable that the dissimilarity is essential, in which case it is conceivable that when similar cells conjugate each cell contains both male and female, and the male of one saturates the female of the other, and vice versa. On the other hand, the whole tendency of evolution is from the simpler to the complex, and, a priori, it is more plausible to consider that complete sexuality is a differentiation of a simpler process rather than the mere separation of what was united in one cell. The last-mentioned conception is undoubtedly the one which would appeal to most biologists at the present time. Yet we see that the functions which exist in a cell do undergo sep- aration, so that they become excessively predominant in certain cells ; for instance, the nervous functions have been thus selected out for the superfluous endowment of certain cells, and it appears to me perfectly conceivable that male and female may be united in a unicellular organism just as completely as assimilative and nervous functions, and as these latter are differentiated, so, too, are the former. The above considerations, and others which might be given, were it worth while to lengthen the discussion of so obscure a subject, lead me to the hypothesis that sexuality is coextensive with life; that in protozoa * the male and female are united in each of the conjugating cells, and impregnation is double; and, finally, that in the metazoa the male and female of the cells separate to form genoblasts or true sexual elements, and impregnation is single. It need hardly be pointed out that this hypothesis is purely tentative, and may have to be rejected altogether when we have suflficient knowledge to decide as to its validity. The object of sexuality is, likewise, known only by hypothe- sis. Three views are to be considered: its purpose is, 1st, rejuvena- tion; 2d, to produce variability; 3d, to check variability. 1st. The theory that the purpose of sex is to produce a young organism is very old, and is based on every-day observation ; it involves, as its corollary, that organisms become old, and thereby incapable of maintaining their own existence. That sexual reproduction does produce a young organism is the universal law; it is also true that every young organism does possess certain morphological and physiological characteristics by which it may be distinguished from an old organism. When sexual reproduction occurs life pro- • * Very possibly this is not true for all protozoa for there may be protozoa with true geno- blasts. 84 THE GENITAL PRODUCTS. ceeds in cycles ; the sexual conjugation produces a single cell, which divides again and again, until at last the process cannot proceed further ; then a renewed conjugation follows and a new cycle of cell- generations ensues ; in the higher animals the cells remain together as they multiply ; in the protozoa the cells each lead a separate life, but in both the cell-cycle is dominant ; the body of a metazoon is comparable to the set of individual unicellular protozoa resulting from one sexual act. In one case the cells of a cycle remain together, in the other they separate. So far, then, as it is known to occur, the sexual process is a rejuvenating one ; but this does not prove that all living organisms require sexual rejuvenation from time to time, nor does it prove that there is no other means of rejuvenation. It may be that all cells as they divide asexually lose their growth-power, so that there comes a time when there must be a rejuvenation or restoration of the growth-power, but it is improbable that sexual reproduction is the only means to effect the necessary restoration of vitality. 2d. That the object of sex is to increase variability and so afford a wider scope for natural selection has been maintained by Weismann. At first sight the notion of the mingling of two hered- itary strains of different character producing variety in the offspring seems very plausible ; but the notion does not bear examination, for it renders the commencement of variability impossible, and fails to account for the divergence in the offspring of the same parents. 3d. The view that sexual reproduction checks variability has been advanced by Hatschek, 87.1, 380, who points out that the mingling of hereditary strains tends to restore the specific norm, since in the long run the variations counterbalance one another. Galton has shown that in human stature the tendency of heredity is to restore the normal height, and the same is presumably true of other char- acteristics. I am strongly inclined to accept Hatschek's theory, and to maintain with him that one result of sexual reproduction is to correct variations and so preserve the specific type. Nature of Sex. — Sex, as we encounter it in the human species, is the result of a long evolution affecting a large number of organs — perhaps all of the organs — so as to result in characteristic differences between the male and female ; but the essential difference is in the relation of the two sexes to the production of the genoblasts ; the male produces the spermatozoa, the female the ova, and in this lies the whole essence of the sexual differentiation; all other distinctive morphological and physiological traits of men and women are second- ary. Thus the structure and functions of the genital ducts, of the uterus, mammary glands, etc., though eminently characteristic of the sexes, in man are not from a biological point of view funda- mental. As we ascend the animal scale there is an increasing divergence between the sexes, owing to the increasing adaptation to the repro- ductive functions. It is generally believed that the primitive con- dition is hermaphroditic, and that the female is an individual in which the power of producing male elements is lost, and a male an individual in which the power of producing female elements is lost. In a certain sense this conception appears true, for in the embryo there is an indifferent stage in which the sexual glands are already HEREDITY. 85 differentiated, but in which the future sex is unrecognizable ; sub- sequently by unknown factors the sexual gland is converted into an ovary or a testis. In some cases, as in certain teleosts and in the snails, the sexual glands develop both ova and spermatozoa. These facts suggest that the primitive sexual gland is potentially herma- phroditic. It is to be remembered, however, that if hermaphroditism were the primitive form we should expect to find the lowest metazoa hermaphroditic ; but this is no^ the case either with all CcBlenterata or all sponges, although it is the case in some higher classes of the animal kingdom — as, for instance, the trematode worms and pul- monate gasteropods. These and other considerations have led me to the hypothesis that primitively each individual animal is sexually indifferent when young, and becomes either male or female when adult ; by a secondary modification in certain f oi-ms the individual becomes both male and female. This is contrary to the prevalent opinion that the hermaphroditic condition is the primitive one. VI. Heredity. In regard to the process of hereditary transmission there are two theories, each of which appears in several modifications- 1st, the theory of pangenesis; 2d, the theory of germinal continuity, The latter does, the former does net, appear to me to conform to our present knowledge. Pangenesis. — The theory of pangenesis was first formulated by Darwin, though it had been crudely foreshadowed by Buffon, Bon- net, and Herbert Spencer. The following quotation from Darwin's "Animals and Plants under Domestication" (Amer. edit., 1868, II., 448, 449) gives his statement of his theory: "I have now enumerated the chief facts which every one would desire to connect by some intelligible bond. This can be done, as it seems to me, if we make the following assumptions : if the first and chief one be not rejected, the others, from being supported by various physiological considerations, will not appear very improbable. It is almost uni- versally admitted that cells or the units of the body propagate them- selves by self-division or proliferation, retaining the same nature and ultimately becoming converted into the various tissues and sub- stances of the body. But besides this means of increase I assume that cells, before their conversion into completely passive or " form- material," throw off minute granules or atoms, which circulate freely throughout the system, and when supplied with proper nu- triment multiply by self-division, subsequently becoming developed into cells, like those from which they were derived. These granules, for the sake of distinctness, may be called cell-gemmules, or, as the cellular theory is not fully established, simply gemmules. They are supposed to be transmitted from the parents to the offspring, and are generally developed in the generation which immediately suc- ceeds, but are often transmitted in a dormant state during many generations and are then developed. Their development is supposed to depend on their union with other partially developed cells or gem- mules, which precede them in the regular course of growth. Why I use the term union will be seen when we discuss the direct action of 86 THE GENITAL PEODXJCTS. pollen on the tissues of the mother-plant. Gemmules are supposed to be thrown off by every cell or unit, not only during the adult state, but during all stages of development. Lastlj-, I assume that the gemmules in their dormant state have a mutual affinity for each other, leading to their aggregation either into buds or into the sexual elements. Hence, speaking strictly, it is not the reproductive ele- ments nor the buds which generate new organisms, but the cells themselves throughout the body. These assumptions constitute the provisional hypothesis which I have called Pangenesis." This hypothesis is the suggestion of a masterly mind, and, as a succinct and comprehensive expression of the facts of heredity, must always command admiration. But the real worth and real signifi- cance of the hypothesis have not been grasped by those who have tried to better it ; its value is not in explaining, for it does not do that, but in expressing heredity in hypothetical terms, which are at once suggestive and comprehensible. Haeckel, in an amusing pamphlet,* which no competent critic can assign the slightest value to, asserted that the gemmules are rhythmical vibrations, but he gives no reasons to justify his opinion. Elsberg has also written on the subject in the Proc. Amer. Assoc. Adv. Sci., XXV,, 178, and cites there earlier writings of his own.f Brooks' modification, 76. 1 ,of the theory of pangenesis well deserves consideration, although the subsequent progress of biology does not lead me to think it felicitous ; but we can now recognize it as a step toward N"ussbaum's valuable theory of germinal continuity, and also toward Weismann's conception that sexual reproduction has for its object the maintenance of variability. Brooks' theory is advocated in his book on " Heredity" (Baltimore, 1879) ; he states it succinctly as follows ■ X " This paper proposes a modification of Darwin's hy- pothesis of the same name (pangenesis) , removing most of its diffi- culties, but retaining all that is valuable. According to the hj''- pothesis in its modified form, characteristics which are constitutional and already hereditary are transmitted by the female organism by means of the ovum ; while new variations are transmitted by gem- mules, which are thrown off by the varying phsyiological units of the body, gathered up by the testicle and transmitted to the next generation by impregnation." If this theory was tenable, there should be — to mention a single objection — little variation in individ- uals produced by parthenogenesis , and they ought always to be fe- males, whereas they are sometimes males. There remains not a new theory of pangenesis, but the valuable suggestion that the maternal influence causes less variability than the paternal. I am, however, strongly disinclined to anticipate the confirmation of this suggestion, especially because the males are not more variable than the females, as we should expect. I have some extensive statistics which show *E. Haeckel "Pere^enesis der Plastidule." Berlin. 1876 For some criticisms which, consid- ering the character of this pamphlet, are very gentle see Eay Lankester in Nature. July 13th. 1876, xiv 235-288 + The perusal of Elsberg's article has not enabled me to recognize anything novel except the substitution of the term plastidule for gemmule. used by Darwin, and speculations as to the composition of plastidules as if he were groping after the conception of the unioella of NHgeli, with which he was apparently unacquainted tProo Amer Assoc Sc. Buffalo 1876 p 177. abstract of a paper read before the section of natural history HEREDITY. 87 that in mammals, at least, there are no essential differences between the sexes in variability. Even if Brooks' thesis should be established it would prove only that the inheritance from the mother is stronger than from the father, and there would lack reasons for his abstruse hypothesis. The theory of pangenesis is to be resigned, not so much on account of the direct arguments against it, as on account of the accumulation of evidence in favor of the theory of germinal continuity. Germinal Continuity. — There are various theories to be consid- ered under this head; but they all have in common the conception that there is a formative force in organisms — that the force depends upon a special material substratum, and that some of the supply of that substratum is given by the parent to the sexual elements it produces. The first important step toward the substitution of a new theory vice pangenesis was taken, so far as I am aware, by Moritz Nuss- baum, whose memoirs, 80.1, 84.2, on the differentiation of sex de- serve great attention. August Weismann* has adopted Nuss- baum's conception and defended it with insistent energy, adding also several modifications. Nussbaum pointed out that there is note- worthy evidence in the development of various animals tending to show that the germinal cells, from which the sexual products arise, are separated off very early from the other cells of the embryo and undergo very little alteration. Hence he concluded that some of the germ substance is directly abstracted from the developing ovum and preserved without essential alteration to become, by giving rise to the sexual elements, the germ substance of another generation. Weismann insists upon the corollary that the whole nature of the animal or plant depends upon its germinal substance {Keimplasma) , and that the reason why the offspring is like the parent is that in every genoblast some of the germinal matter is preserved unchanged. He calls this view the theory of the continuity of the germ-plasma. He follows ISTussbaum also in emphasizing the fact that this theory is inconsistent with the theory of pangenesis and with the theory that parental characteristics acquired through the influence of external causes are transmissible to the offspring. On these two points Weis- mann's second and third papers are especially important. Nussbaum and Weismann lay great stress upon the separation of the cells of the embryo into two kinds: 1, the germ-cells, which are converted into the sexual elements; 2, the somatic- cells, which constitute the body of the organism. The germ-cells descend directly from the ovum, according to Weismann, who has carried his speculations to a great extreme, and undergo little alteration, so that they have (in suspen- sion) the power to produce a whole organism, which the somatic-cells do not have. It is impossible to agree to this extraordinary, view. * Weismann's first papei- was read before the University of Freiburg as a Proreetoratsrede, and was published in pamphlet form at Jena in 1883, 83.1. A second paper was read before the German Naturforscherversammlung in 1885, and appeared in the Tageblatt of that Association, it was subsequently amplified and republished, 86.1. A third paper, 86.3, was likewise ad- dressed to the Naturforscherversammlung in 1886, and published at Jena the same year, A notice of this last is given by KoUmann, Biol. Cbl., v., 673 and 70.S. At thesame meeting of the Naturforscher, R. Virchow also delivered an address (see Virchow's Arch., ciii., 1, 305, 413, and abstract in Biol. Cbl., vi. 97, 129, 161,) in which he attacked Weismann. To Kollmann and Virchow Weismann has replied in Biolog. Centralbl. , vi. 38. 88 THE GENITAL PRODUCTS. Minot, 70, has expressly emphasized the fact that the formative force is certainly a diffused one, as is amply proven by the processes of regeneration, by the phenomenon of duplication of parts, and by asexual reproduction, since in all these cases the formation of a part or the whole of the organism proceeds without the participation of the sexual elements. KoUiker, also, 85.1, 44-46, clearly demonstrates that a sharp division between germ-cells and somatic-cells cannot be maintained. The same position has been adopted by Whitman, 87.3, and, of course, by many others. It is tobe further remem- bered that the cells for the different organs of the body are all set apart very early indeed, and in the case of vertebrates the germ- cells are among the very latest to become distinguishable ; thus the nerve-cells, muscle-cells, notochord-cells, etc., etc., all can be seen to precede the germ-cells in their differentiation. Weismann's assumption that the germ-ceUs are set apart specially early is simply false; all the organs have their cells set apart early, and that too while they are in the embryonic condition ; and it is not true that the germ-cells differ essentially as to their mode of origin or differen- tiation from the so-called somatic-cells. The early divergence of the cells according to the organs or parts they are destined for was pointed out explicitly by W. His many years ago, 74.1, 18. 19. Weismann's error consists in attributing a peculiar significance to a fact by connecting it only with the development of the sexual ele- ments, whereas it is a fact common to all parts of the body. All, therefore, of Weismann's further speculations as to the difference between germ plasma and " histogenes plasma" are without foun- dation. Nageli was probably the first to reach the definite conception of a material basis of heredity, to which basis he gave the name of idioplasma. This idioplasma is essentially identical, it seems to me, with Weismann's Keimplasma. Nageli's views are presented very fully in a large, abstruse, and little-studied volume, of which a useful abstract has been given by Kollmann {Biol. Cbl., IV., 488, 517). Nageli is led to the theory that there are in every living cell two substances, one of which, the idioplasma, alone car- ries on the function of hereditary transmission, while the other, the nutritive plasma (Ndhrplasma) is the seat of the remaining func- tions. In other words, Nageli put forward in a definite form the theory of germinal continuity, for he assumes the formative force to reside in a specific material substratum, which reproduces and per- petuates itself, occurs throughout the organism, and, therefore, in the genital products also. The argiunent in support of this theory is very able, and well deserves the cordial praise which KoUiker and others have bestowed upon it. Nageli did not specify what constituent of the cell corresponds to his idioplasma. 0. Hertwig, 85. 1, was the first to indicate the nu: cleus as the organ of hereditj-, and this view has been ably defended by KoUiker, 85. 1, Strassburger, and others. This notion rests upon the consideration of — 1st, various facts which suggest that the nucleus has special control over the organization of the cell ; 2d, the prob- ability that impregnation consists essentially in the fusion of the pronuclei; 3d, the development of the spermatozoon from the nu- HEREDITY. 89 cleus. That the nucleus presides over the cells is naturally suggested by the phenomena of cell-division, especially indirect division (karyokinesis, mitosis), for during the process the nucleus leads the way, and its visible alteration precedes that of the protoplasm ; the astral rays both during karyokinesis and those around the pronuclei during impregnation may be interpreted as results of nuclear control. The opposite conception that the protoplasm leads has not lacked de- fenders (see Auerbach, Biitschli, 76.1, Nussbaum, 86.1, 504, and Whitman, 88. 1) . I may point out that in interpreting the observa- tions bearing upon this discussion, we must not forget that the nu- cleus and protoplasm are interdependent, neither being able to main- ■fcain its existence permanently without the other. "The fact," says Minot, 85, 125, "that the visible alteration of the protoplasm in a certain rare case comes before that of the nucleus shows that the protoplasm probably has an active role in cell-division ; but since even then its arrangement depends on the position of the nucleus, the evidence of the superiority of nuclear control is, I think, not affected. " On the other hand, there are many observations, which may be interpreted as proofs, that the nuclei have a regulating power over the cells, especially as regards their division and organization. A few of these may be instanced: 1st. After a cell is formed, its nucleus enlarges first, and the cell-body follows it in growth. 2d. KoUiker, in his paper, 85. 1, on heredity (p. 29 &.), discusses the re- lation of nuclei to growth very fully and ably. The great extent of his learning has enabled hiiji to present the manifold aspects of the question more thoroughly than any other writer. His argumentation seems to me so satisfactory that it does not require the weight of his great authority to establish the conclusion that without nuclei there is no growth. Of this the most faith-compelling evidence is offered by the important experiments jf Nussbaum and Gruber,* who found "that when unicellular animals are artificially divided, the fragments containing nuclei continue to grow, while pieces without nuclei die •ofif. 3d. The large unicellular Thallophytes, such as Caulerpa and Codium, become multinuclear before they attain their adult size. Further illustrations are given by Kolliker (I. c, pp. 19, 20). 4th. Perhaps the most striking demonstration of the importance of the nucleus is afforded by the experimental alteration of the plane of division of the ovum. Pfiuger, 83.6, showed that the plane of the first division of the ovum is altered by tilting the ovum before the division begins, and keeping it in the same position during division ; normally the plane passes through the white pole, but when the ovum is fastened in an oblique position, the plane is not in the axis of the ovum, but in the line of gravity. Born,t 84.3, has continued these remarkable experiments, and discovered that the nucleus changes its position when the ovum is kept tilted, and that the site of the nucleus determines the plane of division of the ovum. The second and third points (the importance of the pronuclei and the nuclear origin of the * Science, vol. vi. , p. 4. See also Nussbaum's later paper in the Archiv f iir mikroskoj). Anat. , "xxvi. . p 485. Nussbaum also cites Fr. Schmitz's experiments on the artificial division of plants. Schmitz's paper I have not seen; it was published in 18711, in the Festschrift der natur- TOrschenden pesellschaft zu Halle. + I have not seen the original. There is an abstract in Hofmann und Schwalbe's Jahresbe- richt for 1884 p. 444. 90 THE GENITAL PRODUCTS. spermatozoon) have been sufficiently elucidated in previous divisions of this chapter. Now, it is obvious, since qualities may be inherited from the father, that the nucleus alone can furnish the means of transmission from parent to offspring ; and, since it can accomplish this on the paternal side, it is probable that it can do as much on the maternal side — an assumption against which no evidence has been brought forward ; hence the hypothesis that the nucleus is the organ of hereditary transmission. For criticism of this view see J. Frenzel, 86.5, p. 89, whose arguments have been controverted by Minot, 85, 137. We may go one step farther : Since the chromatin is the character- istic of the nucleus, and since spermatozoa in some cases consist almost exclusively of chromatin, it is probable, as maintained by Minot, 85, 127, that chromatin is the essential factor in the func- tion of heredity. It is my conviction that the hypothesis of pan- genesis, both in its original form and in all its subsequent modifica- tions, has been definitely set aside. In its place we have the theory that the nature of germ, i. e., of the impregnated ovum, is the same over and over again, not because there is in each case a similar collocation of gemmules or plastidules, but because the chromatin perpetuates itself so that the same kind of chromatin is found in the one generation as in the generations preceding it and following it. The child is like, the parents because its organization is reg- ulated by not merely similar, but by some of the same, chromatin as that of the parents. Perhaps instead of " chromatin " we ought to say, in order to avoid an unjustifiable explicitness, "nuclear substance." The validity of this hypothesis remains for the future to decide. There is one general objection to it — that of connecting a special function with a special substance, which is against the general con- ception of vital functions as the resultants of interlocking activities extending throughout each cell. Compare the remarks d propos of the theory of sex, ante, p. 79. The objection is, to my mind, a real and very serious one. PART II. - THE GEBM-LAYERS. CHAPTER IV. SEGMENTATION FORMATION OF THE DIADERM. There follows after impregnation a short pause, and then the ovum begins its process of repeated division, which is known as the "segmentation of the ovum, "the term having been introduced before it was knovsTi that each " segment" is a cell. The division or cleav- age {Furchung) of ova was described by Prevost and Dumas, 1824, and again by Rusconi, 36.1. By usage the term segmentation is restricted to the production of cells up to the period of development when the two primitive germ-layers are clearly differentiated and the first trace of organs is beginning to appear. Segmentation Nucleus.— The impregnated ovum has a single nucleus, which is known as the segmentation nucleus, and which is formed, as stated in Chapter III., by the union of the male and fe- male pronuclei.* It is the parent of all the nuclei subsequently found in the organism, and participates actively in the process of segmentation. It is very much smaller than the nucleus of the egg- cell before maturation ; it is usually membranate and has numerous fine granules of chromatin, microsoma, derived from the pronuclei ; in some cases the microsoma from the male pronucleus are distin- guishable from those of female pronucleus (see under Impregnation, ante, p. 76). In the rabbit the nucleus when first formed has in- distinct contours, irregular shape, and a homogeneous appearance (Ed. van Beneden, 75. 1, 699) ; it soon enlarges, becomes regular, and acquires a distinct centrally situated nucleolus (Bischoff, 42.1, 50, Coste, 47. 1, Lapin, PI. II., Fig. 4), presumably by the gathering to- gether of the microsoma. The position of the nucleus is always eccentric,! so far as known, and aproximately, if not exactly, the same as that of the egg-cell nu- cleus before maturation. Accordingly, the degree of eccentricity varies as the amount of yolk or deutoplasm, being least in alecithal and greatest in telolecithal ova. In brief, it may be said the nucleus tends to take the most central position possible with regard to the protoplasm of the ovum. The vitelline granules are not to be re- garded as protoplasm, hence their accumulation may produce a one- sided distension, without, however, in the least disturbing the uni- form radial distribution of the protoplasm. The nucleus is sur- rounded by protoplasm with few or no yolk-grains ; in telolecithal *fid. van Beneden in his first paper on Ascaris, 83.1, aiSrmed that there was no real union of the pronuclei in the impregnated ova of that species; but Carnoy, 86. 1, shows that Van Beneden 's observations were incomplete, and Zacharias has stated, 87.1, that they are so de- fective as to be fundamentally erroneous in regard to important phases, and he points out that in reality the eggs of Ascaris offer another proof of the actual union of the pronuclei. The im- pregnation in this nematod has since formed the subject of numerous articles ; see Van Beneden and Neyt 87.1, Carnoy 87.1, Boveri 88.1, O. Hertwig 90.1, etc. + It is often stated that the nucleus 'ies exactly in the centre, but 1 have been unable to find a single observation to justify the statement. 94 THE GERM-LAYERS, ova the perinuclear accumulation is the court of protoplasm at the animal pole. Period of Repose. — After the segmentation-nucleus is formed there occurs a pause, which lasts, according to observations on several invertebrates, from half to three-quarters of an hour. It is probable that a similar pause ensues in the mammalian ovum, but there are as yet no observations to shovs^ whether it occurs or not. During this period the yolk expands slightly, unless, indeed, the expansion observed is due to the influence of hardening agents, * and the mono- centric radiation, which is present when the nuclei copulate, grad- ually fades out, and is replaced by a dicentric radiation, which marks the end of the period of repose and the commencement of the first division of the ovum. Karyokinesis of the Ovum. — For convenience I interpolate a sketch of the process of cell-division as encountered in the ovum, based on O. Hertwig, 88. 1, 37, and C. Eabl, 84. 1. My sketch is by no means complete. It is probable that the resting nucleus has one pole at which the connection between the reticulum of the nucleus and the surrounding protoplasm is more intimate than elsewhere, as suggested by Rabl, 89.1. This pole is marked by a clearer spot outside the nucleus, close against it, and much smaller than it. This clear spot becomes the centre of a radiating arrangement of the protoplasm. It was, I believe, first observed by Flemming in the eggs of Echinoderms, has been seen in Ascaris megalocephala by Van Beneden and Neyt, 87.1, andbyBoveri, 88.1, in Siredon by KoUiker, 89.1, and in other cases. It is now designated as the sphere of attraction, f and is seen, at least in certain phases, to contain a separate central body (cen- trosoma of Boveri) . It is not improbable that the " sphere of at- traction" is identical with the Nebenkern of recent German writers. In a number of instances a small part of the nucleus is seen to separate off and to lie as a distinct body, Nebenkern, alongside the nucleus ; this body has a colorable portion, which is comparable to the "centrosoma." For an account of the scattered observations on the ISTebenkern, together with the relation of these bodies to Gaule's so-called cytozoa, see G. Platner, 86.3. For additional observations see Prenant, 88.1, andPlatner, 89.2. The sphere of attraction di- vides, as does also its central body, and its two parts move to op- posite sides of the nucleus. There thus appear two opposite accu- mulations of clear protoplasm, from each of which as a centre astral rays or radiating lines are formed in the cell-body. Meanwhile within the nucleus changes go on ; the threads of the intranuclear network radiate out from the pole, where the sphere of attraction lies before its division, and the chromatic substance forms a number of distinct grains. When the sphere of attraction divides and its halves go asunder the nuclear substance preserves its radiating re- lation to each sphere, and as the membrane of the nucleus disappears during these changes the final result is the transformation of the nu- * Van Beneden states that arsenic acid produces an artificial expansion of the ovum within the zona. t The history and significance of the spheres of attraction, as here presented, cannot by any means be regarded as final. The observations are few, and until recently the exact history of the spheres of attraction has received no attention from investigators SEGMENTATION: FORMATION OF THE DIADERM. 95 cletis into a spindle-shaped body, the points of which rest just within the clear centre of each astral system, so that the spindle stretches from one protoplasmic mass to the other. The spindle consists of fine threads extending from pole to pole and having almost no affin- ity for the dyes of the histologist — a peculiarity which causes them to be known as the achromatic threads. These threads are probably always compounded of a considerable number of exceedinglj- fine fibrillae (see Rabl, 89.1, 21,22) The colorable substance forms a number of separate grains, each of which is united with one of the achromatic threads, and all of which lie at the same level in the centre of the spindle; when the spindle is seen from the side, the chromatine grains appear to constitute a central band or disc (Strass- burger's Kernplatte) , but when the spindle is seen endwise the sep- arate grains are at once recognized. The shape of the grains is variable ; some authors without sufficient observational proof have advanced the opinion that the grains are always V-shaped.. The spindle, together with the polar accumulations of protoplasm and the two accompanying radiations, constitute a so-called amphuister. The domain of the radiation extends, the two protoplasmatic cen- tres move farther apart, the nuclear spindle elongates correspond- ingly, and the chromatin grains of the Kernplatte divide. Flem- ming maintains that the division is always lengthwise of the V-shaped grain, but this has been controverted by Carnoj'. How the division occurs in the mammalian ovum is unknown. By the division, however it is effected, the number of chromatin grains is doubled; they form two sets: one set moves toward one pole, the other toward the other pole; the grains of each set keep at the same level as they move until they reach the end of the spindle, where they appear as a polar disc (Carnoy's couronne polaire). Next the achromatic threads of the spindle break through and are apparently drawn in toward each polar crown. There are now two nuclear rnasses, each near, but not at, the centre of a radiation, and each consisting of chromatin and achromatic substance. Each mass develops into a complete membranate nucleus, but the steps of this process have yet to be followed in detail in the vertebrate ovum. The signs of division of the protoplasm usually become visible about the time the polar crowns are formed; but when the ovum contains much deutoplasm the division may be retarded. In the plane which passes through the equator of the nuclear spindle there appears a furrow on the surface of the ovum, which gradually spreads and deepens until it is a complete fissure around the cell ; it cuts in deeper until at last only a thin stalk connects the two halves of the cell, and thereupon the stalk breaks and the cell is divided. There next ensues a pause, during which the astral rays of the proto- plasm disappear in the. daughter-cells, and the daughter-nuclei assume each the form of an ordinary resting membranate nucleus. The external appearances of segmentation in the living ovum vary, of course, especially according to the amount and distribution of the yolk-material. The appearances in holoblastic ova with very little yolk are well exemplified by Limax campestris. Mark's de- scription, 81.1, is, nearly in his own words, as follows: In Limax, after impregnation, the region of the segmentation nucleus remains 96 THE GERM-LAYERS. more clear, but all that can be distinguished is a more or less circular, ill-defined area, which is less opaque than the surrounding portions of the vitellus. After a few moments this area grows less distinct. It finally appears elongated. Very soon this lengthening results in two light spots, which are inconspicuous at first, but which increase in size and distinctness, and presently become oval. If the outline of the egg be carefully watched, it is now seen to lengthen gradually in a direction corresponding to the line which joins the spots. As the latter enlarge the lengthening of the ovum increases, though not very conspicuously. Soon a slight flattening of the surface appears just under the polar globules ; the flattening changes to a depression. Fig. 46, which grows deeper and be- comes angular. A little later the furrow is seen to have extended , around on the sides of the yolk as a shallow depression, reaching something more than half- way toward the vegetable or inferior pole, and in four or five minutes after its appearance the depression extends completely around the yolk. This annular constriction now deepens on all sides, but most rapidly at the animal pole ; as it deepens it becomes narrower, almost a fis- sure. By the further deepening of the constric- tion on all sides there are formed two equal Campestri7durmg°thiS masses Connected by only a slender thread of cleavage. The envelopes protoplasm, situated nearer the vegetative than are not drawn in. After E. ii • i i j t_ * t, iT L. Mark, x 200 diams the animal pole, and which soon becomes more attenuated and finally parts. The first cleavage is now accomplished. Both segments undergo changes of form; they approach and flatten out against each other, and after a certain time themselves divide. Primitive Type of Segmentation. — In the lower animals there is not found that excessive amount of deutoplasm in the ovum which is so characteristic of the vertebrates, and in their ova we have what is undoubtedly the earlier and more primitive type of segmentation. In these cases the cleavage extends, as in the egg of Limax (see above) , through the whole of the dividing-cell. The two cells first produced are almost if not quite alike, and each of them produces two cells which are also very similar to one another ; then comes a division of the four cells into eight, four of which resemble one another and differ from the remaining cells which are also similar among themselves. Four of the cells are derived chiefly from the substance of the animal pole of the ovum and are very protoplasmatic ; and the other four cells are constituted out of the substance of the vegetable pole and accordingly contain most of the deutoplasm of the ovum The eight cells form an irregular spheroid, in the centre of which there is a space between the cells ; this space is known as the segmentation cavity. The four cells of the animal pole progress in their divisions more rapidly than the four of the vegetable pole ; but the latter, when the yolk matter is at a minimum, as, for instance, in echinoderms, do not lag much From their unequal rates of division the two sets of cells come to differ more and more in size, those of the animal pole being SEGMENTATION: FORMATION OF THE DIADERM. 97 much the smaller. The divisions of the cellfe take place so that the cells form a continuous layer of epithelium, one cell thick, stretching around the enlarged central segmentation cavity, Figs. 47 and 60; the epithelium consists of a larger area of the small cells of the animal pole and a small area of the large cells of the vegetable pole. This stage of segmentation is known as the blastida stage ; the small cells are destined to form the ectoderm of the embryo; the large cells the entodeniij the central space is the segmenta- tion cavity; the line along which the two parts V^^ /^J of the epithelium (ectoderm and entoderm) join Y^\ ,X^^/ is known as the ectental line. X//teiSffli«Sr\ /En Vertebrate Type of Segmentation. — In the vertebrates we find that segmentation ,j ,, „, ^ , , , li • i -ji 1- ; T 1 Fig. 47 -Blastula stage o£ also results m two epithelia, an ectoderm and Echinocardium cordatum, entoderm, joined at their edges, and surround- fo!'°rto1e?m"''En'oto: ing a segmentation cavity, but the resemblance 'j:t'^ ' /S; segmentation to the typical blastula is marked by changes in both ectoderm and entoderm: the vertebrate ectoderm when first fully differentiated consists of several layers of cells, and not merely of a single layer of cells, as in the primitive type of seg- mentation ; the entoderm contains a very large amount of nutritive material (deutoplasm) , and is represented either by a large mass of large cells (marsipobranchs, ganoids, amphibians) or a mass of pro- toplasm, not divided into cells or but partially divided into cells, and containing an enormous quantity of deutoplasm (sauropsidans and monotremes) . In the higher mammals there are further modifica- tions, described below. The more primitive form among vertebrates is, I think, presum- ably that in which the entoderm consists of separate cells; for this mode of segmentation is the one which most resembles that of inver- tebratesi and it occurs in the lowest vertebrates, and in ova which are not excessively charged with yolk. In the primitive form of vertebrate segmentation, which is preserved in the marsipobranchs, ganoids, and amphibia, there is a well-marked difference between the" cells of the two poles. The fol- lowing account refers especially to the frog's egg and is an adapta- tion of Balfour's summary (" Comp. Embryol.," I. , 78, 79) . The first formed furrow is vertical , it commences m the upper half of the ovum, which corresponds to the animal pole, and is characterized by the black pigment — the lower or vegetable pole being whitish. The first furrow extends rapidly through the upper, then more slowly through the lower half of the ovum, so that the divergence in the two polar rates of development is indicated already. As soon as the furrow has cleft the egg into halves, a second vertical furrow ap- pears at right angles to the first and behaves in the same way. Fig. 48. The next furrow is at right angles to both its predecessors, and there- fore parallel to the equator of the egg; but it is much nearer the animal than the vegetative pole. It extends rapidly around the egg and divides each of the four previous segments into two parts : one larger with a great deal of yolk and the other smaller with 7 98 THE GERM-LAYERS. very little yolk. The eight segments or cells have a small segmen- tation cavity in the centre between them. This cavity increases in size in subsequent stages, its roof being formed by the small cells further divided, and its floor by the large cells also multiplied by division, though to a less extent than the small cells. All the devel- opmental processes progress more rapidly at the animal pole. After the equatorial furrow there follow two vertical or meridional fur- rows, which begin at the animal pole and divide each of its four cells into two, making eight small cells. After a short period these furrows extend to the lower pole and divide each of the large cells into two, Fig. 48, i. The so-called meridional cleavages after the first and second are not true meridional cleavages, since they do not pass through the folds of the ovum, but through the poles of the cells (blastomeres) , which they divide (see Eauber, Morph. Jahrb., VIII. , 387) . A pause now ensues, after which the eight upper cells become divided by a furrow parallel to the equator, and somewhat later a similar furrow divides the eight lower segments. Each of the small cells is now again divided by a vertical furrow, which later divides also the corresponding large cell. The segmentation cavity Fig 48. —Segmentation of the egg of the common Frog. is, therefore, now bounded by 32 small and 33 large cells'. After this the upper cells (ectoderm) gain more and more in number beyond the lower cells (entoderm) . After the 64 segments are formed two equatorial furrows appear in the upper pole before a fresh furrow arises in the lower, making 128 ectodermal cells against only 33 entodermal. The regularity of the cleavage cannot be followed further, but the upper pole continues to undergo a more rapid seg- mentation than the lower. At the close of segmentation the egg forms a sphere containing an eccentric segmentation cavity, Fig. 49, s. c, composed of two unequal parts, an upper arch of several layers of cells, Bl, the primitive blastoderm of Minot or ectoderm, and a lower mass, Yolk, of large cells rich in protoplasm. At the edge of the mass of large cells, kw, there is a gradual passage in size to, the cells of the blastoderm, and it appears that the small cells receive additions at the expense of the large ones ; this zone corresponds to the so-called germinal wall of large vertebrate ova, and also to what we have defined as the ectental line. The secondary type of vertebrate segmentation differs from the primary principally in the retarded development of the entoderm, due apparently to the increase of the yolk-matter. The yolk-granules SEGMENTATION: FORMATION OF THE DIADERM. 99 are, as already mentioned, found to be situated not quite exclusively, though almost so, in those parts of the ovum out of which the ento- dermal cells are formed. Hence, when there is a great deal of yolk the anlage of the entoderm becomes bulky, and when it segments the entodermal cells it pro- duces are correspondingly big, as we have seen is the case in amphibian ova. On the other hand, when the amount of yolk is small, as in the primitive type of seg- mentation, e.g. echinoderms, the entodermal cells are small. In the reverse case, when the amount of yolk is exceedingly great, as in selachians, rep- tiles, and birds, the yolk may not divide into cells as fast as the nuclei multiply, so that it seems that the presence of the deutoplasm, though it does not affect the nuclear /qi"-nc, ■r^a■^^,■aAUr ^n^+oi^Kr ^"'- 49. —Section ot the Segmented ovum of Axolotl : divisions markedly, certainly bI, blastoderm: «, c, segmentation cavity: Yolk, impedes very much the di- if'ter'sellSnc'i^™ fcit; (Keimwall). germinal wall, vision of the protoplasm, and consequently in these ova we find, at certain stages of development, a multinucleate yolk. The impediment is not encountered by the protoplasm of the animal pole, hence we see the animal pole seg- menting while the yolk does not ; in this case the segmentation ap- pears confined to one portion of the ovum, and, accordingly, such ova are termed meroblastic in contradistinction to the holoblastic ova, in which the first cleavage furrow divides the whole ovum ; but the difference, it must be expressly remembered, is one of degree, not of kind. The best known example of a vertebrate meroblastic ovum is undoubtedly the hen's egg. The so-called yolk, or "yellow," is the ovum ; the white and the shell are both adventitious envelopes added by the oviduct as the ovum passes down after leaving the ovary. The segmentation begins while the ovum is passing through the lower part of the oviduct, and shortly before the formation of the shell commences. If an ovum from the upper part of the oviduct be examined it is found to be surrounded with more or less white (albumen) . Its animal pole is represented by a whitish disc from 3.5-3.5 mm. in diameter, and 0.30-0.35 mm. in thickness; this disc is known by many names : Formative yolk, germinal disc, cicatri- cula (Narbe, Hahnentritt, Keimscheibe, stratum s. discus pro- ligerus). The animal pole consists chiefly of protoplasm, and is peculiar only in its small size compared with the whole ovum ; it contains, when the ovum leaves the ovary, the egg-cell nucleus; the ovum then matures, impregnation occurs, and iinally segmentation begins. Viewing the ovum from above we see the first furrow appear as a groove running across the germinal disc, though not for 100 THE GERM-LAYERS. its whole width, and dividing it into halves; this furrow is developed in accompaniment with the division of the segmentation nucleus. The primary furrow is succeeded by a second furrow nearly at right angles to the first ; the surface of the germinal disc is cut up into four segments or quadrants, Fig. 50, A, which are not, however, sepa- rated from the underlying substance. The number of radiating furrows increases from four to seven or nine, when there arises a series of irregular cross-furrows, by which the central portion of each segment is cut oflf from the peripheral portion, giving rise to the appearance illustrated by Fig. 50, C ; there are now a number of small /"' -^v. •n& f- -n s / Fio. 50.— Four stages of the segmentation of tlie Hen's ovum. After Coste. Only the germinal disc seen from above and part of the surrounding yellow yolk are represented. central segments surrounded by larger wedge-shaped external seg- ments. Division of the segments now proceeds rapidly by means of furrows running in various directions. Not only are the small cen- tral segments divided into still smaller ones, D, but their num- ber is increased also by the addition of cells cleft off from the central ends of the large peripheral segments, which are themselves sub- divided by additional radiating furrows. Sections of the hard- ened germinal disc show that segmentation is not confined to SEGMENTATION: FORMATION OP THE DIADERM. 101 the surface, but extends through the protoplasmatic mass of the animal pole, there being deep-seated cleavages in planes parallel to the surface, of the ovum. According to Duval, 84. 1, when the first few small central cells are separated off, there is a small space between them and the underlying egg-substance (see Figs. 2, 3, 4, 5, and 6 of his PI. I.), and this space he calls the segmentation cavity ; but in this I think he is in error, for the cells formed below this space are incorporated in the ectoderm or primitive blastoderm ; the cells referred to are those marked in in Fig. 8 of Duval's PL I. The true segmentation-cavity, as we have seen, is bounded on one side by ectoderm, on the other side by entoderm. This fundamental characteristic Duval has entirely overlooked. From the processes described there results a disc of cells, which receives peripheral addi- tions ; the border from which these additions come is known as the segmenting zone. The whole mass of cells derived from the germi- nal disc represents the ectoderm, and the segmenting zone may be homologized with the cells around the edge of the primitive blasto- derm of the frog, Fif . 49, ktv. A section through the segmented germinal disc shows the following relations : The blastoderm is a disc of cells ; its upper layer is epithelioid ; its lower layers consist of rounded cells more or less irregularly disposed; at its edge it merges into the yolk, which continues to produce cells ; between the blastoderm and the yolk is a fissure, the segmentation cavity ; the yolk under the fissure contains a few nuclei, which have each a little protoplasm about them, but do not form parts of discrete cells. In reptiles the process of segmentation is very similar to that in birds. Our knowledge is based principally upon observations upon the eggs of the European lizards (Lacerta agilis and viridis) , which have been studied by KupfferandBenecke,78. 2, Balfour, 79.1, Sara- sin, 83.1, Weldon, 83.1, and Hofmann {Archives neerlandaises, XVI., 1881). Hofmann gives a resume in Brown's " Thierreich," VI.,Abth. III., pp. 1877-1881. The process is very irregular, for small cells are budded off singly and in scattered clusters from the larger segments. As Strahl, 87.1, 390, has pointed out, the blastoderm receives direct accretions from the underlying yolk, cells being sepa- rated off by horizontal cleavages. At the close of segmentation the germinal disc is converted into a membrane consisting, of several layers of cells and parted from the underlying yolk by a thin space, the segmentation cavity ; at its edge this membrane, the primitive blastoderm, is united with the yolk, it being immediately surrounded by a segmenting zone, from which it receives accretions. The layer of the yolk immediately under the segmentation cavity contains scat- tered nuclei, lying singly or in clusters ; each nucleus is surrounded by protoplasm; the nuclei are not all alike; some are verijlaxge, round with very distinct nuclear threads ; other are small and often bizarre in shape; probably the latter are budded off from the former. In Elasmobranchs the germinal disc is thicker, and consequently the mass of cells resulting from its segmentation cuts in quite deeply into the yolk (Balfour, "Comp. Embryol," I., Fig. 46; Riickert, 85.1, 38) . Kastschenko, 88. S, has shown that before the germmal disc is segmented into cells there are nuclei scattered through it, and he has rendered it probable, 88. 1, that these nuclei come from the seg- 10-2 THE GERM-LAYERS. mentation nucleus. It is possible that in other meroblastic verte- brates proliferation of the nuclei precedes the cleavage of the germinal disc into discrete cells. As segmentation progresses, the cells spread out into a layer which shows the same essential relations as have been described in birds and reptiles. There is the several-layered primitive blastoderm, with its edges connected with the yolk and itself overlying the segmentation cavity, the lower floor of which is formed by the multinucleate yolk, the representative of the cellular yolk-mass of the frog, Fig. 49, Yolk. The nuclei are confined to the layer immediately under the segmentation cavity, and this layer corresponds to the sub-germinal plate in teleost ova. Of the yolk- nuclei some are large, others are small as in reptiles ; they are the Parablastkerne of His, the Merocytenkerne of Riickert. In bony fishes also we find the same type, but modified somewhat. The process of segmentation has been very carefully studied by C. O. Whitman, 84. 1 , to whom I am indebted for the accompanying semi- diagrammatic figure of the segmented ovum of a flounder. The ovum is surrounded by a vitelline membran^, z, from which it has slightly withdrawn, notably at the upper pole, where lies the thick cap of cells constituting the blastoderm, Bl; in the stage represented the outer layer of cells is just beginning to assume an epithelioid character ; underneath the blastoderm is the well-marked segmenta- tion cavity, s. c.j everywhere at the edge of the blastoderm lies the segmenting zone, k iv, a ring of granular protoplasm with rapidly- dividing nuclei; the cells re- sulting from these divisions are added to the edge of the blastoderm, which thus en- larges peripherally. The pro- toplasm of the segmenting zone is prolonged inward, forming the floor of the seg- mentation cavity; this sheet of protoplasm, s.g., is known as the suh-germinal plate. The segmenting zone is, of course, the homologue of the similar zone in amniote ova, or the so-called germinal wall, but it is quite sharply defined against the yolk, and therein differs from the wall in the chick, because in the latter the germinal wall merges gradu- ally into the yolk. The process of segmentation differs from that in elasmobranchs and sauropsida in that the cleavage of the germinal disc is strikingly regular, and further in that the whole -vyidth and thickness of the germinal disc is involved in the segmen- tation from the very start. The segmentation in teleosts is further interesting as affording proof that all the nuclei, as shown by Whit- man's investigations, arise from the segmentation nucleus. Fig. 51. — Ovum of a Flouuaer in transverse verti- cal section; semi-diagrammatic figure by Dr. C. O. Whitman, z, vitelline membrane (or zona?); km., segmenting zone (Keimwall); Bl, blastoderm or primitive ectoderm ; s.c, segmentation cavity ; s.g., subgerminal plate; gl, oil globule of yolk. ■\ SEGMENTATION: FORMATION OF THE DIADERM. 103 To summarize : In vertebrate ova with a large yolk, which does not divide into cells until segmentation is considerably advanced, the substance of the animal pole segments completely, and produces several layers of cells (the uppermost becoming epithelioid) which are the ectoderm or primitive blastoderm ; the edge of the blastoderm touches the yolk, and is surrounded by a nucleated zone in which the production of cells is continuing; underneath the blastoderm is the fissure-like segmentation cavity ; the floor of this cavity is formed by the unsegmentated yolk (entoderm) which is furnished with scattered nuclei in the layer immediately underneath the yolk ; the yolk nuclei, at least in selachians and reptiles, are of two kinds, very large ones and smaller ones, which arise probably from the large nuclei ; the uninucleated layer may be termed the sub-germinal plate. Modified Segmentation of Placental Mammals. — The low- est mammals resemble the reptiles in many respects. Among other reptilian characteristics of the mono- tremes we find ova of large size and rich in deutoplasm. That these ova segment in similar manner to those of reptiles and during their passage through the oviduct was first ascertained by direct observation by Caldwell in 1884, 87.1. In marsupials and the placental mam- malia the amount of yolk-substance is ^ ' greatly reduced, and the ovum is of > ^ small size. It is, therefore, holoblastic, that is to say, the cleavage planes cut * — .™™- through the entire cell as in the prim- ^^^,,_^^^^ ^, , ^^^^ „j t^^^t^. itive type ot segmentation; but tne four hours. Atter coste. The erst arrangement of the cells at the close of S^!f™f|i'p^r&ra^?l?^lcefis*^e segmentation appears to be a direct in- Slolt/ife wiiwHreTonf ^euS: heritance from the reptilian ancestors cida. of the mammals. The segmentation of the mammalian ovum was first clearly recog- nized by Bischoff, though it had been previously seen and misinter- preted by Barry, 38.1, 39.1, 40.1; very beautiful figures of seg- mentation in the rabbit have been given by Coste, 47.1. More recently observations have been pubhshed by Hensen on the rab- bit, 76.1, Van Beneden on the rabbit, 76.1, 80.1, Kupffer on ro- dents, 8.23, Selenka on rodents, 83. 1, 83.1, 84.1, and opossums, 86. 1, Van Beneden and Julin on bats, 80. 1, Tafani on white mice, 89. 1 1 The ovum when discharged from the ovary is surrounded by the corona radiata (c/. ante, p. 59), which is lost when impregnation takes place. Segmentation begins when the ovum is one-half to two-thirds of the way through the oviduct. The ovum spends about seventy hours in the oviduct in the rabbit and about eight days m the dog. The first cleavage plane passes through the axis of the ovum, which is marked by the polar globules. When first formed the two segmentation spheres are oval and entirely separated from one another, but subsequently they fiatten against one another and be- come appressed— a remarkable phenomenon, of which we possess 104 THE GERM-LAYEES. no explanation whatever. The second cleavage plane is also meri- dional. The ovum next divides into eight and then into twelve segments, of which four are larger than the rest. The succeeding cleavages have never been followed accurately ; but from Heape's observations on the mole, 86. 1, 166, we know that the divisions progress with great irregularity, and it is probable that the commonly assumed regularity of mammalian segmentation does not exist in nature. After a time (in the rabbit about seventy hours) there is reached the stage termed Metagastrula by Van Beneden, 80.1, 153-160, in accordance with his view of the homologies of this stage. The metagastrula consists of a single layer of cuboidal hyaline cells lying close against the zona pellucida. Fig. 53, ew; the space within this layer contains an inner mass of cells, im, which are rounded or polygonal and densely granular. At one point the outer layer is interrupted and the space is filled by one of the granular segments of the inner mass, Fig. 53. . The nuclei of all the cells are some- what nodulated and have sev- eral highly ref ractile granules each. The granules in the bodies of the cells of the outer layer are somewhat concen- trated around the nucleus, leaving the cortices of the cells clear. Van Beneden, 76.1, 28, 29, has observed that sometimes (21 ova out of 29) the first two segmentation spheres are of unequal size in the rabbit, and similar variability occurs in the mole, Heape, 86.1, 165; Tafani, on the other hand, expressly denies its occurrence in white mice. It is, I think, very improbable that this differ- ence, which sometimes occurs and sometimes does not, has any fun- damental significance. Van Beneden, however, has maintained that the small cell gives rise in the rabbit to the inner mass of cells (see blow) , which he terms the entoderm, but which must, it seems to me, be homologized with the ectoderm, as explained below. That Van Beneden is in error as to the genetic relation of the small cell to the inner mass has been demonstrated by Heape, 86.1, 166. The second cleavage plane is probably also meridional, and is cer- tainly at right angles to the first, so that four similar cells are pro- duced as in the primitive type of segmentation,* Fig. 54. These four cells are also rounded at first and' probably become fitted against one another so as to produce the disposition observed by Tafani, 79. 1, 116, in mice ova at this stage. Tafani describes each cell as having the form of a three-sided pyramid with the apex at the cen- * The distinction here made between "primitive type of segmentation" and "primitive type of vertebrate segmentation " should be borne in mind by the reader Fig. 58. —Rabbit's ovum of about seventy hours. After E. van Beneden. z, zona pellucida ; Ec^ ento- derm; i.m., inner mass or granular cells. SEGMENTATION: FORMATION OF THE DIADERM. 105 tre of the ovum and a convex base forming part of the external sur- face of the yolk. That the two first cleavage planes are meridional is rendered probable by the arrangement in the four-cell stage observed by Selenka in the Virginian opossum, Fig. 55. During all these early stages the colls (segmentation spheres) are naked, i.e., without any mem- brane ; the nuclei, when not in karyokinetic stages, are large, clear, and vesicular; the yolk- granules are small, highly re- fractile, and more or less nearly spherical ; they show a marked tendency to lie in the cell half- way between the nucleus and the edge of the cell, or when the cells are large around the nu- cleus and at a little distance from it. It is at about this stage that ihe ovum passes from the Fal- lopian tube into the uterus, -where it dilates into what is known as the blastodermic vesicle. This dilatation is due principally to the multiplication and flattening out of the cells of the outer layer and, of course, in- volves the expansion and consequent thinning of the zona pellucida, compare Figs. 56 and 58. The inner mass meanwhile remains pas- sively attached to one point on the circumference of the vesicle, Fig. 56, i. m. By this process the thin fissure between the inner mass and the outer layer becomes a considerable space. Fig. 59, s. c, the cavity of the blastoderm or segmentation cavity (blastococle) . I.— Ovum of a Bat, Vespertilio murina, with four segmentation spheres. After Van Beu- eden and Julln. IJn. Pig. 55. — Ovum of Virginian Opossum, with four segments. After Selenka., Fig. 56. — Young blastodermic vesicle of a Mole. 2, Zona pellucida ; i.m, inner mass of cells; s.«., sub-zonal layer of cells. After W. Heape. The blastodermic vesicle continues to expand, and in the rabbit and mole there is a corresponding enlargement of the tubular uterus at the point where the vesicle is lodged. " It is clearly impossible for the delicate-walled ovum to expand in the form of a vesicle, and 106 THE GERM-LAYERS. distend the uterine walls by virtue of the growth of its cells ; it must be, therefore, concluded that it obtains some support. This support is rendered from within. The vesicle contains a transparent fluid, the nature of which I am only sufficiently conver- sant with to say that after treatment with alcohol a white precipitate is present in the vesicle. It is equally evident that this fluid can only have been obtained from the uterus, and that it is present within the vesicle at a very considerably greater pressure, than in the uterus itself. Such a condition is caused by means of the cells of the wall of the vesicle; they secrete the fluid within the vesicle, this function being performed against a pressure which is greater on their inner than on their outer side, exactly as the cells of the salivary glands are known to act. The uterine fluid is secreted by glands present in great numbers in the uterine tissue, and is poured through their open mouths into the cavity of the uterus. There is every probability it has. nutritive qualities, since it is thence taken up into the cavity of the embryonic vesicle, which eventually functions as a yolk-sac, in the walls of which embryonic blood-vessels ramify " (Heape) . The inner mass, Fig. 56, i. m., does not at first grow much and re- tains its rounded form, becoming, at least in the mole, nearly globu- lar. Fig. 57, A. The inner mass subsequently flattens out, becoming lens-shaped, thinner, and of larger area. Fig. 57, B. It continues spreading laterally and separates into three distinct layers. The ovum now consists of a very thin zona pellucida. Fig. 58, 2, close against which is a single layer of thin epithelial cells. En; at one pole this layer is interrupted by a lens-shaped mass, i. m., formed by three layers of cells. These three layers were first clearly described by E. van Beneden, 76.1, and have been since figured by him, 80. 1 ; Van Beneden identified these three layers with the three permanent germ- layers which do not arise until later. Rauber, however, showed that both the outer layers enter into the formation of the ectoderm, while the inner layer is concerned in the production of the permanent ento- derm; the outermost layer Rauber terms the Deckschicht. Lieber- kiihn, 79. 1, and others have since then confirmed Rauber's results. Fig. 57. —Sections through the inner mass of the blastodermic vesicle of the Mole at three successive stages. 0, zona pellucida: s.z.. subzonal layer: i m.. inner mass SEGMENTATION: FORMATION OF THE DIADEEM. 107 Homologies of the Mammalian Blastodermic Vesicle. — We have so little accurate information concerning the details of the formation of the blastodermic vesicle that any interpretation must be tentative. I still consider, however, the view which I brought forward in 1885, " Hdbk," I., 528, as the most satisfactory, and pre- . ferable to the similar explanation advanced independently and simul- taneously by Haddon, 85.1, and reproduced by him briefly in his "Practical Embryology," 47, 48. F. Keibel, 87.1, advocated similar interpretations two years later, but without quoting Minot or Haddon. I regard' the subzonal epithelium as the entoderm and the inner miass of cells as the primitive blastoderm or ectoderm ; by so doing the parts can be readily and exactly homologized with the parts in the frog's ovum, as will be evident at once if the diagram. Fig. 59, of the mammalian vesicle be compared with the section of a segmented amphibian ovum. Fig. 49. The primitive blastoderm Bl, or ectoderm, consists of several layers of cells rich in protoplasm ; below it is the large segmentation cavity, s. c, relatively much larger in the mammalian than in the amphibian ovum. At its edge the primitive blastoderm joins the entoderm Yolk., which in amphibia is a large mass, in mammals only a single layer of cells. Now, we know that the ancestors of the higher mammalia had ova with a large amount of deuto- plasm, which in the course of evo- lution has been lost, so that in the ova of the placen talia there is very little yolk-mate rial ; we know further that the readiness of cel- 1 u 1 a r divisions depends on the amount of yolk, hence, when the yolk is lost, we should expect to find the e n t o - derm, which, as we have seen, is derived from the vegetative substance of the ovum, to be represented by relatively small cells ■, if we imagine the number of entodermic cells m the frog's ovum, Fig. 49, Yolk, reduced, their connection with the prim- itive blastoderm and their character as a continuous layer bemg preserved, we obtain at once the characteristic arrangement of the mammalian blastodermic vesicle, Fig. 59. The homology here es- TiG .58 —Ovum of a Rabbit, ninety-four liours after coitus. -Vfter van Benedeti, £11. subzonal epithelium (entoderm) : Z, zona pelluci- da- i.m . inner mass of cells 108 THE GERM-LAYERS. tablished is further confirmed by the coarse network of protoplasm in the cells of the outer layer of the vesicle (Ed. van Beneden, 80. 1), suggesting at once the meshes which have been emptied of their deutoplasm. Adam Sedgwick, 86. 1, has shown that in the ova of Peripatus capensis the yolk-matter has been lost, though abundant in other species of the same genus, and the coarseness of the proto- plasmic network is preserved as evidence of the granules formerly present. This observation serves to confirm the view I have sug- gested as to the significance of the wide-meshed reticulum of the cells of the inammalian subzonal layer. Fig. 59, Yolk. The disposition of the animal pole in the ovum before segmenta- tion also conforms to the homologies here advocated. It will be remembered, afite, p. 55, that the protoplasm of the animal pole extends far into the ovum and is enveloped by a cup (deutoplasm zone) of the substance of the vegetable pole. Hence, when the animal pole forms cells, they lie as an inner mass, Fig. 56, i.m. If Minot's view be adopted, then the ectoderm lies within the entoderm at a certain stage of development, for the one cell which retains, as shown in Fig. 53, the connection of the ecto- derm with the exterior is sub- sequently overgrown by the outer layer of cells (Van Bene- den, Heape). There is, then, a complete inversion of the germ- layers in all (?) placental mam- malia. In most cases the inver- sion is temporary; the inner mass as described above flattens out, and probably flattens out inside the outer epithelial layer ; if this is the case then the ex- ternal layer of the lens -shaped mass. Fig. 57, B and C, is real- ly entoderm; this layer is Rauber's Deckschicht, which, as already stated, usually disappears, leaving the true inner mass or permanent ectoderm to form part of the surface of the blastodermic vesicle, so that with the exception of the reduction in the dimension of the entoderm the relations are the same as in other vertebrate ova. The inner layer of the flattened inner mass gives rise to the entoderm, and this at first sight appears to be conclusive evidence against the homology here drawn between the inner mass and the primitive ectoderm of other vertebrates. The same thing was formerly supposed to occur in the blastoderm of other vertebrates, but it is now known that the entoderm is added from another source to the under side of the primitive blastoderm or ectoderm, and though we possess no exact information whatever as to the origin of the entodermic cells under the primitive blastoderm of the mammalia, there is no reason to assume that they arise in a manner fundamen- tally different from that typical of other vertebrates. We may, Fig. 69.— Diagram of a segmented mammalian ovum; Z, zona pellucida; Bl. primitive blasto- derm; s.c, segmentation cavity; Yolk^ layers of cell representing the remnant of segmented yolk. SEGMENTATION: FORMATION OF THE DIADERM. 109 therefore, dismiss this objection. The origin of the entodermic cavity and its lining is described in the next chapter. Planes of Division During Segmentation. — The plane of the first division determines those of the subsequent divisions, and also perhaps the axes of the embryo ;* it is itself determined by the position of the long axis of the first amphiaster or nuclear spindle to which it is at right angles. It, therefore, is a matter of great interest to ascertain what factors determine the position of the first spindle, or, in other words, the axis of elongation of the segmentation nucleus. So far as at present known, there are two factors: 1, relation to the axis of the ovum; 2d, position of the path taken by male pronucleus to approach the female pronucleus. The axis of the ovum is fixed before impregnation ; it passes through the centre of the animal and that of the vegetable pole. Usually the nuclear spindle which leads to the formation of the polar globule has its long axis coincident with that of the ovum, hence the point of exit of the polar globule marks one end of the ovetic axis. The first amphi- aster or spindle is always at right angles to the ovic axis. This,- however, leaves the meridian plane undetermined. Roux, 87.1, from a series. of interesting experiments on frogs' ova, conchides that the plane is fixed by the path of the spermatozoon. So far as I know this idea was first suggested by Selenka in 1878, in his paper on "The Development of Toxopneusters Variegatus;" compare, also, Mark, 81.1, p. 500. In the frog's egg the path of the male pro- nucleus is marked by a line of pigment, as was first described by Van Bambecke, 70. 1, 65, and has been well figured by 0. Hertwig, 77.2, PI. v.. Fig. 48. The pigment renders it easy to ascertain the position of the male road even after the first cleavage of the ovum. This Roux has done in sectioned ova, and from experiments and observations reaches this result: The long axis of the first segmen- tation spindle lies in a plane, ivhich passes through the axis of the ovum and the path of the male pronucleus. If Roux's conclu- sion is confirmed, it will become of fundamental importance. Yet there must be other factors which can at least replace the male pro- nucleus in this special role, since the development of parthenogenetic oVa, in which there is no male pronucleus at all, is equally determinate. It is probable that the distribution of the protoplasm is the real cause determining the position of the nucleus; thus in oval eggs the spindle lies in the direction of the long axis ; it is quite probable that if the male pronucleus has the effect ascribed to it by Roux, it pro- duces it indirectly by altering the distribution of the protoplasm within the ovum ; that such alteration takes place is indicated by the occurrence of the male aster. That the first cleavage plane is determined by relations existing m the unimpregnated ovum, has been suggested by O. Schultze in consequence of his finding the germinal vesicle lying eccentrically in the eggs of the brown frog. Schultze suggests that the first plane passes through the ovic axis and the eccentric nucleus. Roux {Biol. * In certain oases, notably m birds as described above, the seRinentation is irregular; and It is therefore not known yet wliether the scheme of arrangement of the cleavage planes here given caf be applied to all ova or not. We may say, however that the scheme is tlie primitive one, from whFcli any modifications arose phylogenetically. The best discussion is by A. Agassiz and Whitman, 84.1, 34-41. 110 THE GERM-LAYERS. Cbl., VII., 420), maintains that this suggestion is set aside by his own observations cited above. For further discussion see Schultze's short note, 87.2, andEoux's rejoinder, 88.1. I think the question whether the first cleavage plane is determined by the ovum's struc- ture or not is still an open one. As already stated in the primitive segmentation, both invertebrate and vertebrate, the second cleavage plane is at right angles to the first and also meridional, while the third plane is at right angles to both the first and therefore equatorial. In meroblastic vertebrate •ova this regularity is entirely lost. Relation of the Segmentation Planes to the Embryonic Axis.- — It has been assumed by some writers that the first cleavage plane coincided with the future median plane of the embryo. This con- ception is rendered extremely improbable by the fact that the seg- ments of the ovum have been observed to migrate in various cases so as to destroy the symmetrical grouping. Miss Clapp's observa- tions, 91. 1, 499, on the toad-fish show that the median plane of the embryo may form almost any angle with the first cleavage plane. Diflferentiation of the Ectoderm and Entoderm. — As already pointed out, the essential feature of segmentation is the unlikeness of the cells produced; the manifold variations in the process of seg- mentation depend chiefly on the amount of yolk. Minot in 1877, 17, first established the generalization that in all animals the ovum undergoes a total segmenta- tion during which the cells of the ecto- derm divide faster and become smaller than the cells of the entoderm; com- pare Fig. 60. There are, however, a small, and I think diminishing, number of cases, where the process of segmenta- tion is imperfectly understood, and which cannot yet be shown to conform to this generalization. " All the known varia- by large entod'ermai the other*' by tions in the process of Segmentation de- pend merely upon : 1st, the degree of dif- ference in size between the two sets of cells ; 2d, the time when the difference appears; 3d, the mode of development, whether polar or by delamination,* either of which may or maj' not be accompanied by axial infolding. In Gasteropods, Planarians, " Calcispongiae, Gephyrea, Annelida, fish, birds, and Arthropods, the difference is great and appears early. In Echinoderms, most Coelenterates, some sponges, in Nematods, Amphibians, etc., it is less marked and ap- pears later." In most cases the entodermic cells are very decidedly larger and less_ numerous than those of the ectoderm. This distinction is obviously necessary on account of the mutual relations of the two primitive layers. The ectoderm has to grow around the entoderm, which it can do only by acq uiring a greater superficial extension ; "^ It does not occur among vertebrates. Fig. 60. — Ovum of Amphioxus lan- ■ceolatus during segmentation-stage, with 88 cells, x 280 diams. After B. Hatschek. One pole is occupied SEGMENTATION: FORMATION OP THE DIADERM. Ill this the ectoderm accomplishes by dividing very quickly at first into small cells. After the entoderm is fully enveloped it may then continue to grow until its superficies is much greater than that of the outer layer, within which, however, it still finds room by form- ing numerous folds ; thus is gradually reached the condition in the higher adult animals where the intestine sometimes has an enor- mous surface, but is nevertheless contained in body-walls covered by ectoderm presenting much less surface. It is, therefore, only during the early stages of segmentation that we find the entoderm expand- ing more slowly than the ectoderm. The terms holoblastic and ineroblastic are applied to ova accord- ing to their manner of segmentation. The first is employed for those ova in which there is either very little or only a moderate amount of yolk, so that the whole of the ovum splits up into distinct masses (cells) which enter into the composition of the embryo. The second designates ova with a very large amount of yolk, so that while the protoplasm, from which the ectoderm arises, divides rapidly into distinct cells, the entodermal portion merely develops nuclei at first, with the result that while one portion of the egg is " segmenting" another portion (the entodermal) remains unsegmented, so far as the external appearances are concerned. Eggs, then, with much yolk, undergo the so-called partial segmentation; hence the adjective meroblastic. Whatever the exact mode of segmentation there results always the same type of organization, to which Minot has applied the term diaderm; it is characterized by consisting of two plates of cells, differing in character, joined at their edge (ectental line) , and sur- rounding a central segmentation cavity ; the two plates or lamina are the two primitive germ-layers, the ectoderm and entoderm. The earliest form of the diaderm is that known as the hlastula, as Haeckel has felicitously named the first larval form of the lower animals. In the blastula we have a simple epithelial vesicle, the cavity of which is the large segmentation cavity. Fig. 47; the epithelial layer is one cell thick and divided into two regions ; ohe composed of smaller cells is the ectoderm, Ec, and the other of larger cells is the entoderm. This stage occurs with sundry modifications in a great many invertebrates. These modifications are due princi- pally to the increase in size of the entodermic cells, which, as already pointed out, results from the increase of the yolk-matter in the ovum. Thus in many mollusks the entodermic cells are very large and at first few in number. By a still further modification the cellular yolk is replaced by a mass rich in deutoplasm, but not divided into cells, while at the same time the segmentation cavity is reduced by the invasion of the yolk-mass. In vertebrates we have the additional modification that the cells are several layers deep in the ectoderm and primitively in the entoderm also; compare the section of the axolotl's ovum, Fig. 49; in certain forms, as we have seen, the entoderm is not divided into discrete cells, but remains one mass; this is the case in Elasmobranchs and the amniota, but in the highest amniota (Placentalia) the yolk is lost and the ento- derm is again represented by a single layer of cells. Fig. 59. It seems to me evident that the first step of development in the 113 THE GERM-LAYERS. segmenting ovum is the differentiation of the two germ-layers, ectoderm and entoderm, resulting in the diaderm stage. Diaderm is a term preferable to blastula, because the latter is applicable strictly- only to a special larval form, while the former is a general term which refers to the essential differentiation at this stage. It is important to remark that the two layers are distinct in the diaderm or blastula stage ; it is often erroneously affirmed that the blastula consists of a uniform layer of cells, part of which subsequently becomes the entoderm. The segmentation cavity comprises the whole space between the entoderm and ectoderm ; it is very early invaded by cells produced from the two primitive germ-layers. These cells are in vertebrates of many kinds and enter the segmentation cavity at various periods. It is customary to group the cells which enter early into this cavity under the common name of mesoderm, and to consider them as a third and distinct germ-layer. For convenience we may adopt this custom, for to a certain extent the mesoderm of authors is a separate germ-layer, but it by no means includes all the tissues which occupy the space between the two primitive germ-layers. As the space between the entoderm and ectoderm is always homologous with itself, it follows that the entire room between the epithelium (entoderm) of the digestive tract and its appendages on the one side and the epidermis on the other is homologous with the segmentation cavity. The mesoderm of authors comprises three tissues : 1, free wander- ing cells (tnesamoeboids) ; 2, embryonic connective tissue or cells connected together by processes (mesenchyma) ; 3, epithelium, which forms two or more separate sacs. The origin of the mesoderm and the relations of the three tissues it contains are discussed in the next chapter. The Gastrula Theory. — In invertebrates with holoblastic ova the blastula passes into a stage known as the gastrula. Gastrula is, properly speaking, a new name for a larval form called planula by older writers ; but the term is now generally employed to desig- nate an ideal embryonic stage, supposed to be common to all multi- cellular animals. The blastula changes into a gastrula by a process of invagination. The entodermal area of the blastula flattens out, the ectoderm mean- while expanding by multiplication of its cells ; after flattening, the entoderm turns inward, forming at first a shallow cup, then a pit which has an opening or mouth, the rim of which is the ectental line. The larva is now a double sac, and has an external wall or ectoderm and an internal wall or entoderm ; the entodermic cavity is entirely distinct from the segmentation cavity. The process of gastrulation is here described as it occurs among the lower inver- tebrates. Typical gastrulse are the free-swimming larvae of many marine invertebrates ; we may take as an example that of a sea-urchin. Fig. 61. The larva is round; at one pole it has an opening, in, the gastrula mouth leading into an internal cavity ; as this is a free- swimming larva it is provided with long cilia for organs of locomo- tion ; the cilia in many gastrulas are distributed over limited areas SEGMENTATION: FORMATION OP THE DIADERM. 113 Fio. 61.— Section of a gaatrula of Toxopneustes lividus ; after Selenka, ec, ectoderm ; en, ento- derm; mes^ mesoderm; m, mouth. or they may be wanting altogether. The larva consists of a double sac, a larger outer one of small epithelial cells, ec, the ectoderm, and a much smaller inner sac composed of larger entodermal epi- thelial cells, en; at the mouth, m, of the inner sac the two layers are continuous with one another; in the space between the two sacs, which corresponds to the segmentation cavity, are a few scattered cells, the first members of the mesoderm, mes. The entodermal sac of the gastrula is known as the arch- enteron; other terms are also in use, e. g., mid-gut, coelente- ron, urdarm, etc. The opening is known as the gastrula mouth (archistome, urmund, etc.) . The coelenterates preserve the gastrula organization through- out life, but in all higher classes the archenteron gives rise not only to the permanent digestive tract, but also to many appendages and derivatives thereof ; and, moreover, the gastrula mouth closes o\er, and in vertebrates the true mouth is an entirely new formation, which arises without any connection whatever, so far as known, with the gastrula mouth. By gastrula- tion the ectental line becomes the rim of the gastrula mouth. A line passing through the centre of the mouth and the opposite pole of the gastrula is the so-called axis. Now, if the mouth be elongated, there would at once be a new longitudinal axis marked out, and the gastrula would become bilaterally symmetrical. If, further, the mouth is pulled out into a slit, and in the process of evolution the lips come together and unite in their middle part, the animal would still have the two ends of the original mouth left open, and would so acquire two apertures to its archenteron — one anterior to serve as mouth, and one posterior to serve as anus. This hypothesis of the conversion of a gastrula into a bilaterally symmet- rical animal by the elongation of the mouth and concrescence of the lips or ectental line, was first suggested, so far as I am aware, by Rabl, 76.1. A very perfect exemplification of the process is afforded by the developing ova of Peripatus capensis, as shown by Balfour, 83.1, and Sedgwick, 85.2, PI. XXXII., Figs. 23-26. There are, how- ever, serious difficulties in applying the theory to bilateral inver- tebrates ; I am strongly inclined to think that further research will obviate these difficulties. In certain vertebrates and annelids the concrescence of the ectental line has been clearly demonstrated, but the process is rendered by secondary modifications much more complex than that described in the preceding paragraph — the detailed account of it forms the sub- ject of the next chapter. The gastrula theory is that all metazoa have a common inherited stage of development, which follows immediately after the diaderm ; 114 THE GEEM-LAYERS. this stage is characterized by there being an outer ectodermal sac with a perforation to the edge of which is attached the entoderm, which forms a closed inner sac, the archenteron. The embryology of coelenterates teaches us that the gastrula is a secondary type, and thus the interesting problem of the origin of the gastrula is to be solved by the invertebrate embryologist (see J. P. McMurrioh, 91.1,310.) The term gastrula was introduced by Haeckel, and is now univer- sally used by embryologists. The discovery of the importance of the gastrula is due to the brilliant researches of Kowalewski on various invertebrates, including Amphioxus, then supposed to be a verte- brate. Haeckel then seized upon the idea of the gastrula and wrote an essay, 74.2, (compare also 75. 1), upon it, which from its brilliant style 'attracted notice, and did much to direct attention to the impor- tant discovery of Kowalewski. Although Haeckel indulged his fantasy unduly and was misled into speculations which are now unheeded and almost forgotten, he did great good by starting the interest of zo5logists in the right direction. By a remarkable coinci- dence, Lankester published an essay, 73.1, of very similar purport to Haeckel's, at about the same time. The gastrula, like the diaderm, varies greatly, the chief modifica- tions depending on the amount of yolk present ; this is illustrated by the accompanying diagrams, Fig. 63; the mesoderm is intentionally omitted ; A corresponds to such a larva as Pig. 61; the difference in size between the two sets of cells is slight but evident. In B, the difference is more marked, and fairly represents a gastrula of Amphioxus. In C, the differ- ence is vei-y great and corresponds to that observed in certain gaste- ropod larvae. In D, the inner set is no longer separated into distinct cells, although there are a number of nuclei, each of which marks the centre of a future cell ; in such instances we must regard the whole inner portion as not yet transformed into a definite entodermic ceM-layer. This figure is par- ticularly instructive, because it shows that what we call the yolk is not something distinct from the germ, but really belongs to the inner layer of the embryo. E shows a similar egg, in which the outer set of cells has not yet grown around the yolk. F shows the same egg not in section, but seen from the outer surface in order to exhibit the cap of small cells (blastoderm) resting upon the yolk. Fig. 62, — Diagi'ams of the Principal Modiflca- tioDs of the Gastrula (see textj. A— E. repre- sents sections. CHAPTER V. CONCRESCENCE : ORIGIN OF THE PRIMITIVE STREAK AND ARCH- ENTERON. This chapter was published in a preliminary form in the American Naturalist, June-August, 1890. Since then the researches of Van Beneden, 88.3, on the rabbit, and of L. Will, 89.1, 90.1, 92.1, on reptiles have cleared up many obscure points. The chief gain, as Prenant has shown in his " Embryologie, " is the knowledge that probably in all, certainly in many vertebrates (excluding Amphioxus) , the entodermal canal arises by the fusion of two cavities ; one of these is the long-known notochordal or blastoporic canal, which communicates with the exterior by an opening (blastopore) at its posterior end; the other cavity is formed in the yolk immediately underneath the notochordal canal and is completely closed. Very early the partition between the two cavities disap- pearr and they fuse, making the definite cavity of the entodermal ca,nal. This primitive canal, from which the pharynx, lungs, and digestive organs are differentiated, is known as the archenteron. The relations with which we are now concerned are illustrated by Fig. 63. ^o a& S?. < I. The Law op Concrescence. Yolk Cavity. — Concerning the formation of the yolk-cavity we possess very imperfect knowl- edge. Undoubtedly a patient search might collate many facts from the literature of the early stages, but until such a collation shall be made and sup- plemented by further observations, no positive his- tory of the yolk cavity can be given. We can say that, when the notochordal canal begins to form, there is already a large cavity under the germ and entirely surrounded by entodermal material. In elasmobranchs and Sauropsida the floor of the cav- ity is the yolk itself, while the roof is formed by cellular material ; the cavity expands over a con- siderable area, but is flattened; it is completely separated from the segmentation cavity ; it is des- ignated often by the name of sub-germinal cavity, but unfortunately the same name is also applied to the morphologically different segmentation cavity, the yolk cavity has been recognized by 0. Schultze; it is not large. In Amphibia 116 THE GERM-LAYERS. In mammals the yolk-cavity, as soon as the entodermic layer is fully developed — see below — comprises the so-called cavity of the two- layered blastodermic vesicle ; owing to the reduction of the yolk, it is bounded wholly by a layer of cells, not partly by a mass of yolk, as in meroblastic ova, and is very large in proportion to the ovum. Concrescence. — The passage from the stage of segmentation to the first embryonic stage is effected in vertebrates by means of cer- tain migrations of embryonic material from lateral positions to median positions, and subsequent union in the middle line. This process of union is known as concrescence. It consists in the grow- ing together of the two halves of the ectental line to form the struct- ural axis of the future embryo. The process is somewhat complex, and needs therefore to be described in detail, the more so as it has still to be followed in mammals. The accompanying diagram may assist to render clear the process of concrescence, Fig. 64. It is intended to illustrate the spreading of the ectoderm (germi- nal disc, blastoderm, auct.) over the yolk and the simultaneous for- mation of the primitive axis. The whole ovum is represented as seen in projection; the propor- tions are such as have been suggested by the ova of flounders and frogs. Three successive stages of the expanding blastoderm are repre- sented ; the first position of the embryonic rim (ectental line) corre- sponds to the dotted line a" a"; the concres- cence reaches only to the point marked 1. The lateral margins, *S", which are to concresce later, still form part of the edge of the blasto- derm. At the next stage the ectoderm has grown very much and has moved its edge to a' a', while the margins, S, have coalesced so that the primitive axis extends to 2. The extension continues, bring- ing the ectental line to o a a ^, and carrying the primitive axis back to 3 ; behind the primitive streak a small area, Yk, of the yolk is still uncovered, and is homologous with, first, the anus of Rusconi in amphibian ova, and, second (in my belief), with the so-called primitive streak of the amniota. The portion of the ectental line bounding this area differs from that which is immediately concerned in the formation of the primitive streak, S; although it now lies be- hind the primitive streak, it was previously in front of it, when the Fio. 64.— Diagram illustrating the growth of the blastoderm and concre.scence of its rim to form the primitive axis, pr. s, iV, neural or medullary groove; n?% neural ridges; Bl, blastoderm ; S, part of the blastodermic rim, termed the jSicAeHn German ;'pr. s, primitive axis; bl\ blastopore; Ffc, uncovered yolk. Compare also the text. THE LAW OF CONCRESCENCE. 117 blastoderm covered only the minor portion of the ovum, see S" a"a". Ultimately the yolk is entirely covered by the blastoderm, thus fixing the length of the primitive streak. It is essential to notice that the blastodermic rim (ectental line) divides into two portions, one, s, which forms the primitive streak, and another, a"a", which over- grows the ovum and at last closes over the yolk behind the completed primitive axis. Historical Note. — The earliest observations on concrescence to form the embryonic axis are, so far as known to me, those of Rathke on leeches.* Nine years later Kowalewski (Mem. Acad. Sci., St. Petersburg, 7'"« Ser., XVI., 1871) recorded its occurrence among insects. Its recognition as a vertebrate mode of development we owe to the brilliant investigations of W. His; in his first paper, 76. 1, he describes very accurately and clearly the process of concres- cence in the salmon; in his second paper, 77. 1, he describes con- crescence in the sharks, and in his third and fourth papers, 77.2, 91.2, he discusses again the general bearing of his results. Semper, in his great work on the relationship of annelids and vertebrates, 76.3, 271, was the first to make a direct comparison of the pro- cesses of concrescence in annelids, insects, and vertebrates. Un- fortunately Balfour entirely failed to grasp the new conception, and by expressing himself very decidedly against it, "Comp. Embryol.," II., 306-308, led many embryologists to discredit the discovery. Whitman, 78.2, 91-94, has ably defended the com- parison made by Semper (see above); Rauber, 76.2, Kolhnan, 85.1, Ryder, 85.5, 9, and others have added to our knowledge of the phenomenon. Duval's researches on the chick, 84.1, demonstrate concrescence there also, though the author appears unacquainted with the results of his predecessors. Minot in the article "Foetus," in Buck's "Handbook," III., 172, 173, accepts concrescence as the typical mode of vertebrate development. Concrescence in Bony rishes.— At the close of segmentation the germinal disc forms a cap of cells on the yolk. The disc (primi- tive blastoderm) spreads over the yolk gradually ; when it begins to spread its edge is already thickened ; this thickened edge corresponds to the ectental line ; the thickening is known as the Randwulst; it is also called the blastodermic rim, which term Ryder and others have used. When the blastoderm has spread, so as to cover perhaps a sixth or less of the surface, one point of the rim ceases f to move ; consequently, as the expansion continues the edge of the disc bends in behind this point on each side, imtil two parts of the blastodermic rim meet as they come from opposite sides, and then grow together. This is illustrated by the accompanying diagram. Fig. 65; F is the outline of the yolk; bl is the outline of the blastoderm; a, the fixed point ; the expansion of the blastoderm has brought the parts 1 1 together and they have united; the parts 2 2 are about to meet and unite; then 3 3 will meet; i 4 and so on, until the two halves of the ectental line are brought together along their entire length ; their junction marks the axis of the future embryo, and pro- duces a longitudinal band of thicker tissue, which has long been * Rathke and Leuckart, "BeitrSge zur Entwickelungsgeschichte der Hirudineen ;" Leipzig, 1862. t Or perhaps merely moves more slowly. 118 THE GERM-LAYERS. Fig. 65. — Diagram of concrescence in a Teleostean egg: y, outline of yolk; 6Z, outline of blastoderm, 1 1, lateral parts already concresced; 2 2, lateral parts about to concresce; 3, 4, parts to con- cresce later. known to embryologists, and may be named the primitive axis. The fixed point of the blastodermic rim marks the head-end of the embryo; the parts of the ectental line which grow together next behind the fixed point develop into the head, those a little farther back into the neck, and those farthest back into the rump and tail. The parts of the circular rim most remote from the fixed point, a, of course concresce last. The destiny of each portion of the ectental line is fixed before concrescence occurs. In fact in certain cases the differentiation of the tissues advances to a consider- able degree in the " Randwulst " be- fore concrescence. This is strikingly the case in Elacate, in the ova of which the myotomes (or segmental divisions of the mesoderm) appear in the embryonic rim before its con- crescence (Ryder, 85.9) ; compare also Ryder's observations on Belone, 81.2. The development of the tele- ostean germ -layers is not yet fully worked out. For the best history of the entoderm and mesoderm, as well also for references to conflicting authorities, see M. Kowalew- ski, 86.1,2, who, however, pays no heed to the law of concrescence. That concrescence occurs in teleosts essentially as here described, seems to me evident from the figures given by AV. His, 76.1, C. Kupffer, 84.1, Coste, 47.1, and others. ITevertheless the concres- cence is denied by Henneguy, 88.1, H. V. Wilson, 91.1, 260, and others, but the arguments I have found against concrescence have not appeared to me valid. In the primitive axis is a mass of cells below the ectoderm ; this mass subsequently divides into mesoderm and entoderm. The entodermal cells form at first and for a considerable period a solid cord (c/. Balfour, "Comp. Embryol.," II., 75) in which, however, a lumen appears later ; this lumen I will tentatively homologize with the cavity of the notochordal canal of amniota. Concrescence in Elasmobranchs. — Our Imowledge rests mainly on the researches of His, 77.1, and his follower, KoUmann, 85.1. Fig. 66, A, is a generalized diagram of an elasmobranch ovum, representing the ectodermal disc, Bl, as seen from above rest- ing upon the yolk, which is not represented in the figure. The first change noticeable in the disc after the close of segmentation is a groove running completely around its margin between it and the yolk ; as the disc grows and expands the groove is no longer present along the front edge, a a, of the blastoderm, but only on the sides and behind. Abou' the same time there usually appears a distinct notch, n, which marks the fixed point of the margin and the pos- terior end of the disc. If now a section be made across the line, JT F, the relations will be found to be essentially as represented in the diagram. Fig. 66, B ; the disc rests on the yolk, F^, which con- THE LAW OF CONCRESCENCE. 119 tains numerous nuclei ; between the yolk and the ectoderm, Ec, is the segmentation cavity, sc; the groove is bounded above by a layer of cells. En, which are larger than those of the ectoderm, and have been produced by the yolk, Vi; sometimes there are cells lying in the segmentation cavity at this stage, the formation of the mesoderm having already begun. The essential point to note in this stage, is, as KoUmann has shown, the division of the margin of the "- ectodermal disc into two parts, one, a a, resting directly on the yolk, the other, S, directly continuous with a layer of en- todermal cells, B, En, forming a little groove under the mar- gin of the disc. The two por- tions of the ectental margin have entirely distinct func- tions, as already stated; the anterior, a a, is destined to grow over and cover the yolk by the extra-embryonic portion of the ectoderm ; the posterior, 8, is destined to form the primitive axis of the embryo. Fig. 67 is similar to Fig. 66, but represents a more ad- vanced stage. The ectodermal disc, Bl, is much enlarged, and its anterior grooveless mar- gin, a a a, is relatively much more increased than the poste- rior grooved margin, S; the centre of the notch. Fig. 66, n, has remained nearly if not quite stationary, Fig. 67, pr. s, while the margin, s s, of either side has been growing toward its fellow in the manner indicated by the arrows, and as they meet the two side-margins grow together in the median line, making a longi- tudinal structure. The manner and results of the concrescence of the margins from the two sides to form an axial structure become clearer in section. Fig. 67, B. The margin at the side, m, still shows the same relations as in Fig. 66, B; in the median line, however, the margins have met and intimately united, "so that what were originally two grooves have completely united to form a single canal, Ent, bounded above by entodermal cells, below by the entodermal yolk, Vi. This canal is the primitive entodermal cavity. Whether it represents, when first developed, merely the notochordal canal of the amniota or the fused notochordal canal and yolk cavity, we are unable to determine at present. A moment's consideration renders it evident that the canal must be open pos- teriorly; this opening is the blastopore, bl. There are some further details to be mentioned : where the ectental margins have united m Fig. 66.— Diagram of an Elasmobranch Blastoderm to illustrate the formation of the marginal groove. A, surface-view; Bl-, blastoderm; a a, anterior grooveless margin; S CSicheO, marginal CTOove; n, marginal notch ; X Y. line of section. B, section along the line X Y of A. Ec, ectoderm ; En, ento- derm; m, extental line; s.c. , segmentation cavity; Vi, yolk with nuclei. 130 THE GERM-LAYERS. the median line there appears a lateral outgrowth, mes, which is the beginning of the mesoderm ; in some cases this mesodermic tissue appears before the margins concresce ; when viewed from the sur- face the mesoderm can be seen through the ectoderm, as was ob- served long ago; it is this faint appearance which early writers call in anamniota the — primitive streak, it be- ing the foreshadowing of coming organiza- tion. Fig. 67, A, also shows in front of the primitive axis the first trace, N, of the central nervous system, which we shall describe later. The blastoderm is seen also to be divided al- ready into two parts, the lighter area pellu- cida, A. p., and the darker area opaca, 'A. o.; the latter also shows the first blood- islands. For further descriptions of these areas, see Chap. XIII. From their observa- tions. His, Kollmann, and others have in- ferred that at the an- terior ectental margin, a a a, there are pro- duced (from the yolk) cells, which grow in toward the embryo, and constitute part of the mesoderm and are especially concerned in forming the first blood, which is produced always in the extra- embryonic area. This mesoderm of peripheral origin His has named parablast — a term which, unfortunately, has been employed differ- ently by some subsequent writers. The ectoderm, entoderm, and axial mesoderm are grouped by His under the collective name of archiblast. This view of the double origin of the mesoderm, al- though it has been adopted in a modified form by the brothers Hert- wig, I am unable to accept. The question is discussed in Chapter VI. Concrescence in Marsipobranchs, Ganoids, and Amphib- ians. — As not only the constitution of the ovum, but also its early development, is very similar in the three classes named, we may con- sider them collectively in the present connection. The condition of the ovum at the close of segmentation has already been described, p. 99, and figured. Fig. 49. The ectental line is not sharply defined, Fig. 67.— DiaCTam of a Vertebrate Blastoderm a little more advanced than Fig. 66: A, surface-view. B, section along the a a a, anterior margin ; s s, pos- , area opaca; A.p.^ area pellucida; ; iV; neural or medullary groove ; pr. s. , prim- itive streak; feZTblastopore ; ^c. . ectoderm ; m, ectental margin; En^ entodermlo cells; Fi, yolk: mes, mesoderm ; s.c, segmen- tation cavity. line, X Y. Bl, blastoderm ; teriormargin (SicheO ; A,o.^ 71. r. , neural ridges ; THE LAW OF CONCRESCENCE. 121 nor does there appear any groove around the edge of the blastoderm as in meroblastic ova. The small-celled ectoderm spreads over the yolk ; while it is doing this a short notochordal canal appears at the hind edge of the blastoderm with a small opening to the exterior, known as the blastopore, Fig. 68, bl. The first indication of the canal in the frog is easily recognized, being the appearance of a curved area of pigmentation of semilunar outline amid the yolk-cells at the posterior pole ; the convexity of the area is directed toward the segmentation cavity ; the centre of the concavity corresponds to the dorsal lip of the blastopore (Robinson and Assheton, 91.1, 463). The canal runs forward toward the segmentation cavity, Fig. 68, ^.c; above and in front of the blastopore the cells have multiplied and accumulated to form the beginning of the primitive axis, Pr. In the lamprey there is at this stage no such axial accumulation of cells ; according to Shipley the ectoderm consists of a single layer of cells, and the notochordal canal is bounded on its dorsal side by a single layer of cells also, between which and the overlying blasto- derm there are no cells ; the gathering of cells corresponding to the primitive axis does not arise un- til later. The canal, according to Fia. 68.— Ovum of Axolotl. After Bellonci. liongitudinal section to show the commencing formation of the primitive axis, Pr. bl. blas- topore; Bl, blastoderm; s.c, segmentation cavity. Fig. 69. — Ovum of Petromyzon in longitudi- nal section. After Balfour, me, mesoderm of primitive axis ; bl, blastopore ; Ylc, yolk at the anus of Rusconi ; al, notochordal canal ; s.c, segmentation cavity. 0. Schultze, ultimately fuses with the yolk cavity to form the definite archenteron ; it is sometimes designated as the blastoporic invagina- tion. The canal in the same measure as the blastoderm spreads over the yolk-grooves at its hinder end away from the segmentation cavity. Fig. 69, s.c, just as in elasmobranchs. A stage is soon reached in which nearly the entire length of the archenteron is formed and nearly the whole yolk is covered. There is still a blastopore which leads into the cavity, and which has moved gradually backward from its original position. Behind the blastopore lies the uncovered yolk, Yk, which in the frog's ovum is very conspicuous, because its whitish color contrasts with the dark color of the heavily pigmented ectoderm around it ; this area of exposed yolk is the so-called anus of Eusconi. When the canal has completed its full length the 123 THE GERM-LAYBRS. following disposition of the parts is found, Fig. 70 : The archen- teron is bounded below by the large mass of yolk-cells, Vi, and above by the epithelium, Ent, of the entoderm ; its posterior end curves up to open at the blastopore, Bl, passing through a mass of cells, which constitute the end of the primitive streak; this portion of the archenteron is sometimes called the blastoporic canal. There is further a short prolongation, Al, of the cavity below the blastopore. This diverticulum has been homologized with the allan- tois, (see Chapter XII.). It is also very probably homologous with the more nearly spheri- cal diverticulum found in a similar position in teleosts, and now known as Kupffer's vesicle, from having been es- pecially studied by C. Kupf- fer, 66.1, 475, 68.1, who has interpreted it as the teleostean allantois. Com- pare D. Schwarz, 89.1, 197, Taf. XIII., Figs. 35, 37, etc. Around the blastopore is a mass of cells (primitive axis) continuous on the one side with the ectoderm, on the other with the epithelial en- toderm lining the archente- ron, and, thirdly, with a sheet of cells, Mes, between the ectoderm, Ec, and entoderm, Ent. The developmental phases just outlined seem to me to afford suf- ficient evidence of concrescence. Owing to the gradual transition between the ectoderm (blastoderm) and the entoderm (yolk-cells) there is no sharp ectental line, as in some other types. Moreover, there is no differentiation of the tissues at the blastodermic rim, but only after the cells are united in the axis ; hence we cannot distin- guish parts at the periphery of the blastoderm and follow their union in the primitive streak as we can in certain sharks and bony fishes. Nevertheless, we find all the essential features of concrescence ; the notochordal canal and the primitive axis begin at the edge of the blastoderm and grow at their posterior end away from the segmenta- tion cavity, and at the same rate the blastoderm overspreads the yolk. Concrescence in Sauropsida.— The early stages in Eeptilia have long been obscure. Clarke (Agassiz' "Contributions," II.), in his paper on the embryology of the turtle, mistook the commencement of the notochordal canal for the commencement of the amniotic fold. Weldon, 83. 1, Kupffer, 82. 1,84. 1, Strahl, 80. 1, 2, 3, 82. 1, 83.1, Hoffmann (Bronn's "Thierreich," VI., Abth. iii., 1892-1897), and others partly traced out the history of the canal. Will's observations, 90. 1, on the development of the gecko gave the key to the history of the canal in the reptiles. In tho gecko there is formed a notochordal canal, which is at first very short, but gradually lengthens out, apparently chiefly by growth at its hind or blastoporic Fig. 70. — Longitudinal section of the ovum of a Sturgeon after tfie formation of the entodermic cavitr : Ec^ ectoderm; Mes^ mesoderm; Ent^ entoderm: Bl^ blastopore \ Al, diverticulum of the archenteron ; Vi, yolls. After Salensky. THE LAW OF CONCRESCENCE. 133 end, Fig. 63, nch. c. The end of the canal, when the germinal area is examined in surface- views, is characterized by a transverse figure or sichel, which is well known in reptilian embryos of all orders, and which presumably represents the portions of the Randwulst which are to concresce and thereby lengthen the primitive axis and the notochordal canal inclosed by the cells of the axis. Underneath the notochordal canal is a layer of entodermal cells, Ent, which form the roof of the yolk cavity ; the figure does not show the inferior or lateral boundaries of the yolk cavity. In a little later stage, the tissue between the canal and the yolk cavity disappears and the two lumina fuse. In other reptiles the development is similar, though obscured by the peculiarity that the anterior part of the notochordal canal opens into the yolk cavity before the posterior part is formed. In such cases there is only a short section of the canal to be observed with complete boundaries at any one stage. In reptiles then concres- cence can only be inferred from the presence of the " sichel" and the B Fig 71 —Formation of the blastoporic canal in Lacerta muralis. After Weldon. B, C, longitudinal sections of two successive stages of the blastoderm, which in each case has been re- moved from the yolk ; the space under the entoderm, En, is the archenteric cavity. D, transverse section of the posterior part of the blastopore a little younger than C. Ec, ectoderm ; En, ento- derm; JV, primitive streak ; bl, blastopore: Ch, notochord; mes, mesoderm. growth backward of the primitive axis. Fig. 71 illustrates the formation of the canal in Lacerta, as described and figured by W. F. R. Weldon, 83.1. En is the entoderm forming the roof of the yolk cavity. B shows the notochordal canal, bl, just beginning to form. C is a stage considerably more advanced ; the anterior part of the canal has fused with the yolk cavity, and the dorsal wall of 124 THE GERM-LAYERS. the canal has produced the notochord, nch; only a short posterior end, bl, remains as a canal. D is a transverse action through the blastopore. _ The process of concrescence in birds was partly indicated by Koller s investigations, 79.1, 82.1, and has been carefully elucidated by Duval, 84. 1. The resemblance to concrescence as known in elasmo- branchs is very striking. Around the edge of the blastoderm ap- pears very early a small groove ; as the blastoderm expands the front portion loses the groove; one point, the centre of the grooved margin, ceases to move, or at least moves much more slowly than the remain- der of the blastodermic rim ; as the expansion continues the edges of the two halves of the groove coalesce gradually behind the fixed point, thus producing the entodermal canal in the same manner as in the sharks ; cells accumulate at the same time and make behind the blastopore the so-called primitive streak. There is some uncer- tainty in Duval's account, as, unfortunately, at the time he wrote the existence of a yolk cavity contributing to the formation of the archenteron had not been recognized. In birds (hen's ova) there is a further peculiarity, which_ is, I think, probably to be found in all amniota, namely : that portion of the edge of the ectoderm which does not share in concrescence and which corresponds to the edge of the anus of Rusconi closes over the yolk be- hind the primitive streak, so that the portion of the yolk which is left un- covered is remote from the embryonic Fia. 73. -Hen's ovum: incubated six area (or primitive stroak) . As a rare hours ; anterior edge of the ectoderm rest- v r 001 t ingon the yolk from a longitudinal sec- anomaly, See Wnitman, oo. 1, a line tion^^of the blastoderm m situ. After -^ ^-g.^^^ running in the sctoderm from the hind end of the primitive streak to the edge of the uncovered yolk ; this line is to be interpreted as evidence of the growing together of the ectoderm, behind the streak proper. The ectoderm, as it spreads over the yolk, receives no accre- tions from it, but accomplishes its expansion by proliferation of its own cells. Thus the uncovered yolk is bounded by the free edge of the ectoderm. Fig. 73. The area of uncovered yolk, which may be called the yolk blastopore,* is not homologous with the anus of Rusconi, from which it differs in position, being remote from instead of close (as is the anus of Rusconi) to the blastopore, for it is situated nearly opposite the embryonic area. In birds, according to Duval, 84. S, the yolk blastopore (DotternabeT) is never closed by ectoderm, but remains covered by the vitelline membrane only, until the mesoderm spreads over it. The growing edge of the ectoderm is somewhat thickened ; it finally is reflected around the edge of the yolk blastopore, forming, as it were, a funnel, at the bottom of which is the yolk (see Duval, I.e.). Concrescence in Mammals. — As shown below in the detailed history of the mammalian blastodermic vesicle, there is a fixed point (Hensen's knot) at which the formation of the primitive axis and notochordal canal begins, and from which they lengthen out back- * Duval applies to it the name of ombillc, ombilical. THE LAW OF CONCRESCENCE. 125 ward as they would do if formed by concrescence. The tfiain cavity of two-layered vesicle is the yolk cavity, and with it the notochordal canal subsequently fuses, cf. infra. The position and history of the ectental line being absolutely unknown in mammalia, it is of course impossible to form any definite notions as to the process of concres- cence in them. Concrescence : Summary. — The evidence that concrescence is the typical means ofi forming the primitive streak in vertebrates is : 1, detailed and conclusive observations upon elasmobranchs, teleosts, and birds ; 2, exact and extensive observations on marsipobranchs, ganoids, and amphibians, which concord with the theory of con- crescence; 3, a great probability of its occurrence in reptiles, owing to the similarity of their development with that of birds ; 4, a prob- ability of its occurrence in mammals, because of the resemblance in the growth and structure of the primitive axis to that in other verte- brates. The theoiy seems to me inevitable that the vertebrate primitive axis is formed by the growing together in the axial line of the future embryo of the two halves of the ectental line. The development of the primitive axis may be described in general terms as follows: At the close of segmentation the edge of the primitive blastoderm separates into two parts ; one part (the anterior) , as the blastoderm, expands, spreads over the yolk, gradually covering it with ectoderm ; the other part (the posterior) forms the primitive axis ; it has in its centre one fixed point ; consequently, when the blastoderm expands the two halves of the posterior part of the ectental line are brought together and gradually unite (concresce) along a line running from the fixed point backward (radially as re- gards the blastoderm). Consequently, the segmentation cavity, which is underneath the primitive blastoderm, lies in front of the developing axis. While this goes on cells grow out from the con- crescing part of the ectental line into the space between the ectoderm and entoderm (or yolk) ; underneath the line of junction a cavity is formed lined by entoderm ; this cavity is the notochordal canal ; it lengthens backward as concrescence progresses ; it has, whatever its length, a small entrance, the blastopore, at its hind end ; the blasto- pore is ultimately obliterated. The cells which grow out from the ectental line constitute the first anlage of the middle germinal layer or mesoderm, and shining through the ectoderm they produce the appearance of a whitish line, which has led to the name of primitive axis. The characteristics of the mesoderm are described in the next section. Along the line of junction there often appears a slight furrow in the ectoderm, which is known as the primitive groove. Significance of Concrescence. — It will at once be evident that if the process of concrescence went on without the actual meeting of the two portions of the ectental line the result would be to leave the archenteron open along its entire length; the borders of the opening would be the ectental line; and this line, as we have seen, corre- sponds to the lips of the gastrula mouth; consequently, we should have a gastrula with an elongated mouth. This condition is illus- trated by the accompanying diagram, Fig. 73. It agrees in aU re- spects with the gastrula type; its most noteworthy peculiarities are two : first, the enormous mass of yolk accumulated in the aboral 136 THE GERM-LAYERS. portion of the entoderm ; second, the elongation of the gastrula or archenteric cavity in a direction at right angles to the gastrula axis, xy. If now the lips of gastrula. Fig. 64, s, meet and unite we should obtain at once the vertebrate type. According to W. His' discovery, this is precisely what takes place — only the lips are brought together first at one end, where they at once unite, while behind they are widely separated ; but gradually they are brought together and unite throughout their entire length. Concrescence is, then, a modified method of uniting the lips of a greatly elongated gastrula mouth. Why this modification is estab- lished we cannot say with certainty, though we may surmise with confidence that it is consequent upon the great accumulation of yolk in vertebrate ova. The view here adopted enables us to speak positively as to the point where we are to look in vertebrates for the homologue of the invertebrate mouth. In annelids concrescence is very well marked, whenever the ova contains much yolk ; now in leeches and earth- worms the ectental line does not concresce along the entire axial line, / \ ^m->. % Fig. 74.— Dog-fish embryo, nearly in Balfour 's stage C. m, position of invertebrate moutii; B, rim of the germinative area. Fig, 73. — Diagrammatic cross-section of a vertebrate ovum, in which concrescence is supposed to have been arrested ; xy, median plane; Ach, archenteron; Ent, entoderm; Ect, ectoderm. but, on the contrary, as shown by Kleinenberg and Whitman, the foremost part of the germ bands (gastrula lips) do not unite, but leave a small opening ; when the permanent mouth is formed this opening is carried in and serves as the passage between the mouth cavity (Vorderdarm, stomodseum) and the archenteric cavity. The fore- most part of the line of concrescence lies, according to His' observa- tions, on fishes just where the optic outgrowths arise. Fig. 74, m; hence we have to search between the origins of the optic nerves for traces of the invertebrate mouths. Further reference to this question is made later in connection with the development of the nervous system. The Notochordal Canal.— The existence of this canal was, so far as I am aware, first satisfactorily recognized by Lieberkiihn, 82. 1, 84.1, who discovered that in mammals it produces the notochordj THE LAW OF CONCRESCENCE. 137 and by losing its lower walls fuses with the yolk cavity. The canal is a narrow tube which runs forward in the tissue of the primitive axis (KoUiker's head process) ; it ends blindly in front, but its posterior end communicates with the exterior by a funnel-shaped opening (the blastopore) through the ectoderm. Immediately behind the. blastopore lies the accumulation of cells, termed the primitive streak in amniota, the anus of Rusconi in amphibians. The canal is lined by epithelium, which is thickened on the dorsal side to form the anlage of the notochord. At the sides the epithelium merges into cells belonging to the mesoderm. The manner in which the canal is formed by concrescence is ex- plained in the preceding pages, and the manner in which it fuses with ^ the yolk cavity is described in the following section. For additional details and references see the history of the notochord in Chapter VIII. rusion of the Notochordal Canal and Yolk Cavity.— The fusion of these two cavities has been carefully studied in mammals and reptiles. The fusion in amphibi- ans is briefly mentioned by O. Schulze, 88.1. In the gecko (L. WiU, 90.1) and in mammals (Lieberkiihn, 82.1, 84. 1 , Van Beneden, 88.3, and others) , the canal becomes quite long, and then acquires a series of irregular openings, Fig. 75, nch, on its ventral side into the very large yolk cavity, which at this stage underlies the whole germinal area. The anlage of the notochord is already differentiated on the dorsal side of the canal. The ventral open- ings increase both in number and size until the entire canal has fused with the yolk cavity except at the hind end, where it persists for a while as the so- called blastoporic canal. The fusion occurs in guinea-pigs the fourteenth to fifteenth day, in rabbits the eighth day. In lizards (Strahl, Kupfiier) and tur- tles (Will) the fusion occurs in a simi- lar manner, but sooner, so that the anterior portion of the canal has fused with the yolk cavity before the posterior portion of the canal is completed. The union of the two cavities produces the definitive archenteron, which is a spacious cavity lined by entoderm, having the anlage of the notochord in its median dorsal line and opening to the exterior by the blastopore, which is situated at the caudal end of the primitive axis and the headward end of the primitive streak. Blastopore. — The blastopore is the small opening which leads into the notochordal canal, or, after the canal has fused with the yolk cavity, leads into the archenteron. It is situated at the hind end of the primitive axis (head-process), and marks the anterior boundary of the anus of Rusconi in amphibia, or of the primitive streak, properly so-called, of amniota. Fig. 71, B. Fig. 75. — Germinal area of a Guinea- pig at thirteen days and twenty hours, seen from the under-side. After Lieber- kiihn. ao^ area opaca; ap, areapellu- cida ; nch, anlage of the notochord as a canal, with several irregular openings on the entodermic side, x 24 diameters. 128 THE GERM-LAYERS. While the concrescence of the ectental line is going on the blasto- pore changes its position, being always at the end of the notochordal canal. When the canal fuses with the yolk cavity the end of the canal persists for a time as a passage at the end of the primitive axis, and this passage is sometimes designated as the blastoporic canal, see Figs. 70 and 71. The opening is finally obliterated. The blastopore is not homologous with the gastrula mouth, but is merely a small portion thereof ; in front of it the gastrula mouth is closed by concrescence ; while concrescence is going on there wiU be a part of the gastrula mouth open behind the blastopore ; when con- crescence is completed the blastopore is at the end of the elongated gastrula mouth, the lips of which are united throughout the remain- der of their length. The blastopore is not a fixed point, being merely the opening of the notochordal canal, and as by concrescence the canal is elongated, in precisely the same measure the blastopore travels backward. The Meroblastic Embryo. — Considerations of practical con- venience have led to the custom of distinguishing in the development of meroblastic ova the embryonic from the extra-embryonic por- tions. The distinction is in real- ity entirely arbitrary, for the whole of the ovum is included, morphologically speaking, within the body of the embryo. Custom has led to designating the two parts as the embryo and the yolk ; but the student should be careful not to allow himself to be misled by these terms. In the laboratory it is a general practice to remove the so-called " embryo " from the yolk, and in doing this the arch- enteric cavity loses its inferior wall, to wit : the entodermic yolk. Let the relations be represented by the accompanying diagram. Fig. 76, the embryo being drawn very much too large in proportion to the yolk, for the sake of clear- ness. Suppose the layers to be cut through on the lines x x; we could then remove the embryonic portion. This is what is actually done in practice. It is very im- portant to understand clearly that the yolk is part of the embryo, and that our sections usually represent only a torso. Fig. 76.— Diagram showing the relations of a vertebrate ovum with an embryo in cross-sec- tion and a large yolk. Ec, ectodei-m ; N, neural groove ; mes^ mesoderm ; s c ^ segmentation cav- ity ; Ent, archenterie cavity ; a a, ectodermal rim, where the ectoderm is growing over the yolk. II. The Primitive Axis and Streak. The term primitive axis is a new one, which it has seemed neces- sary to introduce to avoid confusion. It is nearly synonymous with the term head-process (KoUiker's Kopffortsatz) . It is applied in all vertebrates to the median band of cells which runs forward from the THE PRIMITIVE AXIS AND STREAK. 129 blastopore ; the central cells of the band are entodermal* and form the epithelial wall of the notochordal canal; the lateral cells of the band contribute to the production of the mesoderm. At the blast- opore the primitive axis merges into the primitive streak, sensu strictu, and on that account has been interpreted and described by many authors as the anterior prolongation of the primitive streak. After the ventral wall of the notochordal canal has disappeared and the canal has fused with the yolk cavity, the entire tissues of the primitive axis lie on the dorsal side of the archenteron. The term primitive streak may be conveniently and properly re- stricted hereafter to the accumulation of cells lying immediately behind the blastopore. In amphibia this accumulation is known as the anus of Eusconi ; it belongs to the entoderm (and later to the rnesoderm also) , and is very conspicuous owing to the absence of pigment in its cells. In amniota the corresponding accumulation comprises the cells in the region around the primitive groove, as described in detail below ; in amniota the accumulation has the yolk cavity (later archenteron) extending under it. Fig. 71, A B, pr, and it is therefore not directly continuous with the yolk proper, as in amphibians. The conceptions of the axis and streak above presented appear to me necessary consequences of our present knowledge, but until they are accepted by other embryologists, the reader must view them as largely my personal opinions, and must remember that morpholo- gists are not yet agreed as to the nature of the primitive streak. The Primitive Axis.— -As above defined, the primitive axis is the median band of cells resulting from concrescence and overly- ing the definitive archenteron. It is advisable to begin with the consideration of the arrangement as we find it in eggs of marsipobranchs, ganoids, and amphibians, since these eggs are probably more primitive in their mode of de- velopment than those of other vertebrates. The points of most im- portance in my judgment are illustrated in Fig. 77, A and B. In A we have a section through the middle portion of a young primi- tive axis of an axolotl, the axis still requiring considerable additions at its hinder end before attaining its full length ; the archenteric cavity, Ae, is a large space bounded above by an epithelium. En, and below by the large mass of yolk cells, Ykj the two-layered ectoderm, Uc, everywhere bounds the section ; above the archenteron and below the ectoderm lies the accumulation of cells constituting the primitive axis, Prj the lateral prolongations, Mes, of the axis represent the commencing mesodermic outgrowths; whether the mesoderm grows out from the primitive axis and subsequently ex- pands solely by its own proliferation, or whether it receives at its periphery accretions from the yolk cells is uncertain. I am inclined to think that the mesoderm does not receive additions from the yolk. In B we have a similar section, but of an older stage, and through the hind end of the nearly full-grown axis ; the general arrangement is the same as in A ; we note the following differences : the archen- teric cavity is a mere slit, Aej the primitive axis, Pr, is very thick *Prenant ("Embryologie") regards them as ectodermal, following 0. Hertwig's suggestion; the terminology in this case is largely a question o^ previous definition. 130 THE GERM-LAYERS. and composed of numerous small cells, and its lateral mesodermic expansion, Mes, extends farther around the ovum. In both sections we see that the cells of the primitive axis are not marked off from Fig. 77. —Sections of Axolotl eggs ; A, frontal section somewhat anterior to the blastopore, from an egg in which the archenteron was partly formed, hnt the anus of Rusconi not delimited. B, frontal section of an older ovum, with well-marked but large anus of Rusconi ; the section passes just in front of the blastopore. Ec^ ectoderm ; En^ entoderm ; ilfes, mesoderm ; Ae^ archenteric cavity; Yfc, yolk; pr, primitive axis. After Bellonci. those of the adjoining entoderm. In a longitudinal section, as is illustrated by that of a sturgeon. Fig. 70, the mesoderm of the primi- tive axis is seen to extend far forward from the blastopore, Bl. The disposition of the parts and the appearance of the cells vary in the three groups we are considering, but for our purpose it is unnecessary to describe these secondary differences. The points essential to note are that the primitive axis produces chiefly mesoderm, which is accumulated along the axial line, and is thickest around the blast' opore, where it joins the primitive streak, and which spreads lat- erally between the ectoderm and entoderm ; in the axial region the mesoderm is not separated from the entoderm. In elasmobranchs the differentiations of the axial tissues begins in the embryonic rim before concrescence takes place, so that while the type ajEfords peculiarly conclusive evidence of concrescence, it is less convenient for the study of the primitive axis, since the hind end of the primitive axis is, as it were, divided, being continued as the embryonic rim, right and left. The degree of differentiation in the rim varies extremely : in Pristiurus the mesoderm grows out ; in Scjdlium the mesoderm grows out and the differentiation of the notochord begins; in Torpedo (Riickert, 87.1, 101) the myotomes appear in the embryonic rim before concrescence, as in Elacate among teleosts. The relations are further complicated by the ad- vance in development of the axial structures while concrescence is going on, so that, as for instance in Pristiurus, Eabl, 89.2, 116-129, the axial notochord may be differentiated, while the mesoderm is stiU developing in the embryonic rim. The precocious changes in the embryonic rim demand especial attention when the origin of the mesoderm is discussed (c/. Chapter VI.). The ectoderm, as soon as it becomes one-layered (secondary blastoderm, see Chapter IV.), con- THE PEIMITIVE AXIS AND STREAK. 131 sists of high-cylinder cells. As development progresses the ectoderm thins out except on either side of the axial line. The mesoderm arises from the entoderm close to the ectental line and is there quite thick, but as it stretches away it thins out. Now if it be remembered that the ectental line becomes the axial line, when concrescence oc- curs, it is evident that this m.esoderniic thickening of the entoderm is in reality an axial thickening, and when concrescence takes place it fuses with the corresponding thickening of the opposite side and constitutes an actual axial thickening or true primitive streak ; but in elasmobranchs, as soon as the anterior axial structures have con- cresced, we find by precocious development that the notochord and medullary groove appear; now, as shown in Chapter VII., the ap- pearance of these structures causes the division of the axial mesoderm 4nto completely separated right and left portions. It is only by keep- ing the process of concrescence and the precocious development of the parts constantly in mind that we can understand the development in elasmobranchs or compare it rightly with that of other types. From what has been said it is clear that a section of the blastoder- mic rim from which the mesoderm was just growing out would cor- respond to half a section of, say, a bird's ovum, though the primitive axis, and upon comparison it will be found that all the essential re- lations are identical. The primitive axis and streak of birds have been much in- vestigated and discussed, and may be conveniently treated together. I follow in the main Duval, 78. 1, 84. 1, many of whose statements are confirmed by Zumstein, 87.1. Other important au- thorities to be consulted are Kol- liker in both his text-books ; His, 68.1, 77.2, 82.1, etc.; KoUer, 82. 1 ; Disse, 78. 1, 79. 1 ; Wal- deyer, 69.1, 83.1; M. Braun, 82.3; Gasser, 77.1, 79.1; Eauber, 76.2; C. Eabl, 89.2, et al. The following description ap- plies to the hen's egg. When the egg is laid the centre of the segmented blastodisc presents a circular area of lighter color; during the first few hours of in- cubation this area pellucida, as it is called, becomes more dis- tinct; as the area pellucida expands, the primitive streak appears in it eccentrically be- tween the eighth and twelfth hour. By the sixteenth hour the primitive streak has its . full length. The rate of development is extremely variable, au- tumn eggs developing more slowly than spring eggs; the eggs vary also individually, and are, moreover, much influenced by the tem- jp 30 Fio 78 —Area pellucida of a Hen s egg, with completed primitive furrow After Duval a o , areaopaca; ci, anterior crescent; a.jj., area pellu- cida; pt, primitive groove. X 80 diams. 133 THE GERM-LAYERS. perature of their incubation. For a fuller discussion of these varia- tions see His, 68.1, S6-63. Seen from the surface the area pel- lucida with completed streak presents the following features, Fig^ 78. The area pellucida, a. p., is considerably elongated and some- what pear-shaped, being widest at the anterior end of the primitive groove, pt. ; this groove is well marked as a narrow and shallow furrow, which begins some distance from the anterior edge of the area and ends just before reaching the posterior edge of the area ; the front end of the furrow usually bends slightly to the left, but not invariably, as KoUer and Rabl have maintained, for it sometimes bends to the right or is quite straight; aline of granules is sometimes noticeable above the primitive groove; they were seen byDursy, I.e., and are called by Duval, 78.1, 15, the filmnent epiaxial — compare Gasser, 79.1. The portion of the area pellucida immediately around , the primitive groove appears slightly darker than the rest. The anterior portion of the peUucida is further distinguished by the anterior crescent, ct, the "vordere Aussenfalte" of His, 68.1, and other German writers. The anterior crescent is a temporary ap- pearance, due, according to Duval, to a series of folds of the entoderm, which form a curving row of shallow pockets, that, shining through, mark out the crescent. The crescent disappears a little later, and there arises nearly, if not quite, in its, place a different fold, the amniotic. The similarity of position has led to the anterior cres- cent's being identified by some authors with the true amniotic fold. Longitudinal and transverse sections are very instructive. We Fig. 79. —Longitudinal section of the region of the primitive streak ot a Hen's ovum incubated six hours. After Duval. D, general view magnified about 40 diameters. A, B, C, details of D, with higher magnification. Efc, ectoderm; raeSy mesoderm; En^ entoderm; 6i, Duval's "blasto- pore;" fciy, germinal wall iKeiniwaXV) \ Ach^ archenteric cavity; sg.c^ segmentation cavity. begin with the examination of a longitudinal section of a somewhat younger stage. Later the ectoderm closes behind the primitive streak, as already stated and spreads backward over the yolk. The section shows that the yolk is not divided into cells, although nuclei are scattered throiigh it ; the nuclei are represented as black dots in THE PRIMITIVE AXIS AND STREAK. 133 A, B, and C. The cavity of the archenteron, Ach,* is enlarged by the formation of a deep pit in the yolk, while the posterior half of the cavity remains a narrow fissure between the cellular entoderm, JSn, and the yolk; the archenteron communicates, according to Duval, with the exterior by an opening, bl, which he calls the blas- topore ; as this supposed opening is apparently at the posterior ex- tremity of the primitive streak it cannot be the true blastopore. The entoderm is a loosely put together stratum of cells, which passes over anteriorly into a ridge of the yolk in which cells are being produced around the already accumulated nuclei ; this ridge, kw, is the ger- minal tvall. Posteriorly the cell layers are much thicker, A ; the ectoderm, He, is clearly differentiated from the underlying cells, which are all more or less alike; though they represent both the entoderm and mesoderm. From this connection and from the fact that the connection between the ectoderm and mesoderm which is so well known to exist after the primitive streak has attained its full length, Duval concludes that the mesoderm arises primitively from the entoderm. Transverse sections afford additional information. Ent C Ec Ent—. Fig. 80. — Transverse sections of a germinative area, with half -formed primitive streak, of a Hen's eg^. After Duval. A, through the anterior region of the area pellucida. B, through the primitive streak. C, part of A enlarged. Ec^ ectoderm ; mes^ mesoderm ; Ent^ entoderm ; bl, Diastopore; fcwi, germinal wall ^Keimwall); Ach, archenteric cavity; sg. c, segmentation cavity. The accompanjang Fig. 80 represents cross-sections of a germinal area, the primitive streak of which had attained about one-half its full length. The first section. Fig. 80, A, passes through the anterior region of the area pellucida, and therefore in front of the primitive groove; it shows the large cavern, Ach, of the archenteron (or yolk cavity?) hollowed out in the yolk; the entoderm, C, Ent, above the cavity is a thin layer of cells, connected laterally with a projecting shelf of yolk kiv (the bourrelet entodermo-vitelUn of Duval) , which is rich in nuclei ; it subsequently expands and acquires a more cellular character ; this shelf is the commencement, therefore, of the Keimwall of German writers. Immediately above the ento- derm, and intimately connected with it, are a few cells, which be- long to the mesoderm, C, mes; the ectoderm is quite thick, C, Ec, * As previously stated, Duval was unacquainted with the existence of yolk cavity ; it is prob- able that the cavity here termed archenteric is really the yolk cavity. 134 THE GESM-LAYEES. and consists of high columnar cells ; toward its periphery the ectoderm thins out, and its edge rests upon the yolk, with which it has no con- nection. In the region of the primitive streak. Fig. 80, B, there are important differences in the germ layers to note. The entodermic cavity, Ach, is very much smaller ; the mesoderm is much thicker and in the axial region fuses with both the outer layer of cells * and the entoderm, thus forming the Achsenstrang (axial cord) of Ger- man writers ; the mesoderm also spreads out over the yolk far beyond the archenteric cavity, and about one-third of the way from the axial line to the distal edge of the ectoderm : the ectoderm merges in the median line with the mesoderm, and presents externally a small notch, B, pr, corresponding to the primitive groove. Whether at the stage from which Fig. 80 is taken the formation of the primitive axis (head-process) has fairly begun is uncertain. In slightly older stages the "head-process" is present (KoUiker, "Grundriss," 2te Aufl., 36). During these changes the archenteron (yolk cavity?) expands rapidly, the entoderm becomes very thin in the area pellucida, and passes more and more abruptly, as development progresses, into the so-caUed germinal wall of the area opaca ; finally the ectoderm becomes thinner peripherally, so that the axial thicker part is gradually marked off more and more abruptly. Sections of a stage with the primitive groove at its maximum — a stage which is usually found toward the end of the first day of incubation — show these changes clearlj-. A cross-section through the area opaca in front of the area pellucida shows the thin ectoderm, the thick cellular entoderm overlying the archenteric cavity and charged with yolk granules ; the entodermic nuclei are very variable in form and ir- regular in distribution ; the cell boundaries are indistinct. There is no mesoderm. A cross-section near the front of the area pellucida likewise shows only ectoderm and entoderm ; the former is a high cylinder epithelium over the area pellucida and thins out toward the opaca on each side ; the latter is a thin layer over the area pellucida and passes quickly bvit not abruptly into the very thick yolk-bearing entoderm (or KeiniivalT) of the area opaca. Sections a short dis- tance in front of the primitive groove show that the head-process {Kopffortsatz) is a forward prolongation of the primitive streak, and consists of an axial accumulation of mesoderinic cells fused with the entoderm, and having broad extensions sideways to form the mesoderm between the outer and inner germ-layers; the lateral portions of the mesoderm have no connection with the other germ layers, and at its distal edge the mesoderm thins out and rests upon the entoderm of the opaca, but without being connected with it ; I cannot find any satisfactory evidence that it receives any additions from the opaca entoderm, as many authors have maintained. The ectoderm in the region of the " Kopffortsatz" resembles that iEurther forward, but it very soon shows a faint median furrow, the so-caUed dorsal groove {Riichenrinne) , which is the commencement of the medullary groove (see Chapter VII.) . In the anterior half of the primitive streak the relations are different from those in the " head- process." The outer layer shows the primitive groove. Fig. 81, prg, * This outer layer is usually termed ectoderm, but I hold that it is not ectoderm, but the homo- logue of the outer layer of yolk cells in the amphibian anus of Rusconi. THE MAMMALIAN BLASTODERMIC VESICLE. 135 and is fused with the axial cord (Achsenstrang) of the mesoderm ; laterally the outer layer passes into the true ectoderm, Ec. In the posterior region of the primitive groove the connection of the meso- derm with the inner germ-layer is dissolved. Behind the primitive groove the mesoderm extends, but lies free between the ectoderm and m Fig. 81. — Transverse section of the anterior region of a fully-developed primitive streak of a Hen's ovum: Ec^ ectoderm; mes, mesoderm; Ent^ entoderm; Pr. (/, primitive groove. The large black dots represent yolk grains. entoderm. To recapitulate : there is a long axial mesodermic thick- ening, which has the primitive groove over its posterior two-thirds ; the thickening in front of the groove is united with the entoderm, and constitutes the primitive axis ; tlie thickening under the front half of the groove is united with the entoderm ; in the median line its external surface is freely exposed, and laterally it merges into the ectoderm; the thickening under the hind half of the groos^e is not united with the entoderm. III. The Mammalian Blastodermic Vesicle. In all placental mammalia, owing presumably to the absence of the large amount of yolk present in the ova of other amniota, the early development is modified, and the germinal area instead of resting on a mass of yolk rests upon a vesicle. When the vesicle is fully de- veloped its main cavity is lined by entodermal cells, and must be, in my opinion, homologized with the yolk cavity of other vertebrates, for it fuses with the notochordal canal to develop the definitive archenteron. V/e may conveniently distinguish four stages of the vesicle, which are described below in order : 1, with one layer constituting the vesicle, except over the germinative area; 2, with two layers; 3, with primitive streak; 4, with "head-process," or primitive axis. 1, Vesicles with One Complete Layer. — After the close of segmentation we find that the inner mass becomes flattened out, and in the reigon it occupies we can distinguish three layers of cells, as previously described : first, counting from the outside, the thin layer of cells known as Rauber's " Deckschicht ;" second, a middle layer of cylindrical cells, which becomes the ectoderm; third, an inmost layer of thin flattened cells, which belong to the entoderm; the Deckschicht continues round the whole vesicle as a single layer; the other layers do not so continue, compare Figs. 57 and 58. The next step in development is the formation of a second layer, which spreads out in all directions from the region of the inner mass ; hence as far as the new layer reaches the blastodermic vesicle becomes two- 136 THE GERM-LAYBES. layered. Meanwhile the Deckschicht disappears, leaving two layers in the region of the inner mass ; it is to be remarked that the Deck- schicht is retained in certain rodents, undergoing special modification, as described in the section on inversion of the germ-layers. Rabbit's Vesicle at Six Days. — The following is a summary of Ed. van Beneden's description, 80.1, 185-300. The vesicle meas- ured 3.2 mm. in diameter; it was nearly spherical; the wall of one hemisphere consisted of one layer of cells; the other hemisphere had two layers of cells, and besides in its central portion a third layer intervening between the other two. The area with three layers Van Beneden designates as the tache embryonnairej it showed no trace of the primitive streak ; it was oval in outline and had one point, which the author identifies as Hensen' s knot, where the layers adhere together closely. Transverse sections show that the outer- most layer of cells is a low cjdinder epithelium, which, at the edge of the area, passes into a thin epithelium, quite abruptly ; it cor- responds to Rauber's Deckschicht, and has been said by him to flatten out and disappear, leaving the cells underneath as the per- manent outer layer of the embryonic vesicle. The cells of the inner- most layer are thin and wide ; they are called the hypoblast (ento- derm) by Van Beneden ; the cells themselves have round nuclei, around each of which is accumulated a court of granular proto- plasm ; the adjacent courts are connected by a coarse mesh work of protoplasmatic threads ; treatment with nitrate of silver brings out the cell boundaries and divides the reticulum into polygonal areas. The cells of the present outermost layer have distinct boundaries and contain granules and long bacilliform bodies, which Van Beneden saw also in the fresh specimens and found to be constant appearances. Similar bodies are found in the germinal vesicles of sheep, and are held by Bonnet, 84. 1, to be derived from the uterine milk; the rab- bit is not known to have uterine milk. The histological peculiari- ties of these two layers remain about the same from the fifth to the eighth day. The middle layer consists of rounded cells with nu- merous granules ; seen from the surface their diameter is greater than that of the cells outside them, but much less than that of the cells underlying them. While we know that the middle layers are ectodermal, it is uncertain whether the inner layer is ento- dermal or not. Blastocyst of the Rabbit of Seven Days. — The development is exceedingly variable, so that exact times cannot be given. The general appearance is illustrated by Fig. 83, from Kolliker. The vesicle figured was 4.4 mm. in length; the envelopes of the ovum are not shown, though they were still present ; at the upper pole is the Fig. 82. —Blastodermic vesicle of a Rabbit of seven days : og, area genninativa, or embryonic shield ; ye, line, above which the vesicle is two- layered. From Kolliker. THE MAMMALIAN BLASTODERMIC VESICLE. 137 small embryonic shield, corresponding in position to the region of the inner mass ; it is marked out by the greater thickness of the walls of the vesicle ; the developing second layer extends over more than half the vesicle, reaching to the line ge. S. Slastodermic Vesicle with Two Layers. — Of this stage we have several descriptions ; for the rabbit by KoUiker (" Grundriss," p. 91); Hensen, 76.1; C. Eabl, 89.2, 141; as well as the older accounts by Bischoff, 4S. 1, and Coste, 47. 1, and the brief mention by Heape in Foster and Balfour's " Embryology, " 2d edition, 316- 320; for the mole by Heape, 83.1; for the dog by Bischoflf, 45.1; for the cat by Schafer, 76.1; for the sheep by Bonnet, 84.1; and for several rodents, as indicated in the section on inversion of the germ- layers, p. 141. The two-layered stage is found in the rabbit about seven, in the sheep about thirteen, days after coitus. The dimensions for the sheep are about 4 mm. for the greatest diameter and 3.3 mm. for the lesser diameter. The two layers form each a closed sack ; the embryonic shield is well marked as a round spot, less translucent than the walls else- where. The outer layer has a distinctly epithelial character ; in the region of the shield its cells are columnar with spherical nuclei ; in the rabbit the cells are low and the nuclei lie nearly at one level; in the sheep the cells are taller and the nuclei are at various levels ; in the mole (for a good figure see Heape, 83.1, PL XXI., Fig. 49), and in various rodents there are several layers apparently, but perhaps in them also the epithelium is columnar, as it certainly is later. At the edge of the shield there is an abrupt change to a very thin layer, with widely expanded cells ; consequently, in the region of the shield the nuclei are close set, while outside the shield they are wider apart. The change at the edge of the shield is at first less abrupt, but at the present stage is very marked. A similar difference exists in the inner layer, although its cells are very much thinned out everywhere, yet the layer is slightly thicker in the region of the shield ; the nuclei of the inner layer are somewhat flattened, and they are larger and farther apart than the nuclei of the outer layer — a difference which is very obvious in surface views, both during this and the Fig. 83. —Transverse section of the embryonic shield of the •noY+ fr^^rrar\■r,a■ ■=dao'e^'i hlastodermie vesicle of a Sheep thirteen days premant. After next lOilOWing btageb. gonnet. n, outer layer of vesicle; 6, inner layer of vesicle. The inner layer has an epithelial character in the region of the shield, but farther away the cells move apart, and being connected by processes resemble embry- onic connective tissue (Bonnet, 84.1, 192; Hensen, 76.1, Figs. 15 and 11, B on Taf. VII. ; E. van Beneden, 80. 1). The relations are illustrated by the accompanying Fig. 83, representing the shield in the sheep at thirteen days and of a vesicle measuring 4 mm. by 2 mm. ; at the left of the figure the layers are accidentally folded. The next changes which occur are principally those of growth. 138 THE GERM-LAYERS. both of the vesicle as a whole and of the embryonic shield, which also begins to arch up ; the vesicle and shield both become oval ; usually the oval shield lies lengthwise, but in the deer, as shown by Bischoff, it lies transversely of the vesicle. The size of the shield is quite nearly uniform among the placental mammals in which it has been studied, but the size of the vesicle varies extremely ; especially note- worthy is the excessively rapid elongation in ungulates (pig, sheep, goat, and deer) ; in the sheep, for example, it trebles or sextuples its length in less than a single day after the shield appears. The next step is the appearance of a middle layer, at least in sheep (Bonnet, 84.1, 193-196, 89.1, 42), which shows in the fresh specimen as a slight turbidity. Fig. 84, mes, of the vesicular walls just outside the edge of the shield, while in the region of the shield there is no m.iddle layer whatever. Sections show that the new layer consists of loosely scattered cells connected by anastomosing processes ; it is everywhere absolutely distinct from the outer layer, but merges at many points with the inner layer ; from this con- nection Bonnet concludes that the middle layer is derived from the inner layer by what must be called a process of delamination. So far as known to me nothing analogous to this middle layer has yet been observed in other mammals. The next important step, again accord- ing to Bonnet, 81.1, 195, is the ap- pearance of Hensen's knot, which takes place while the peripheral middle layer is developing. The knot. Fig. 84, kn, is at first a small thickening on the under-side of the outer layer; it is situated on the middle line of the shield a little nearer one end than the other ; it is distinctly separated from the inner layer, but is connected with the cells of the middle layer, which have now developed themselves in the middle region of the shield also. Bonnet mantains that the knot gives off cells which contribute to the formation of the middle layer. The knot marks the front end of the future primitive streak, and is the beginning of the primitive axis. The appearances in a sheep's ovum at this stage are illustrated by Fig. 84 of a vesicle of twelve to thirteen days from a sheep ; the vesi- cle measured 55 mm. in length by about 1.5 in breadth, but the length of the vesicle is extremely variable at this stage; the specimen had been stained to bring out the small, close-set nuclei of the outer layer and the larger, more widely set nuclei of the inner layer. The upward arching embryonic shield, Sh, shows Hensen's knot, kn; around the edge of the shield, Sh, the middle layer makes an irregu- lar shadow, mes. '-"t'-'.i'K Fig. 84. —Central portion of a Sheep's blas- todermic vesicle or twelve to thirteen days. 8h^ shield; fcn, Hensen's knot; pr, trace of primitive streak ; mes, " Mesoblasthof . " Af- ter. Bonnet. X 34diams. THE MAMMALIAN BLASTODERMIC VESICLE. 139 A condition of the blastodermic vesicle similar to that described is figured by Coste for the rabbit, 47.1, by Bischofif for the rabbit, 42.1, Taf. IX., fig. 42, c, for the dog, 45.1, Taf. III., fig. 28, B; and the gradual extension of the second layer is recorded for the mole by Heape, 83. 1. Since it is known to occur in rodents, carnivora, and insectivora, it is probably true of all placental mammals that the one- layered vesicle becomes two-layered by the outgrowth of cells from the " inner mass" found at the close of segmentation ; this is the first step of development after segmentation. Rauber's Deckschicht has evidently great importance. It was first described by him in the rabbit, 75.2; and was also discovered by E. van Beneden, 76.1, who, however, made the error of consid- ering it as the permanent ectoderm, and the true ectoderm below it as the mesoderm ; this error has been amply corrected by Kolliker and is now admitted by Van Beneden (see Van Beneden and Julin, 84.1). Its disappearance in the rabbit has also been studied by Lieberkiihn, 79.1. Balfour ("Comp. Embryol.," II., 219) from in- vestigations on the rabbit by himself and Heape, concluded that the cells of the Desckchicht disappear by being incorporated in the true ectodermal layer becoming at the same time columnar; this view is verified by Lieberkuhn, 82.1, 400, 401. As already stated ihe rodent modification of .the Deckschicht is discussed below, p. 141. In the rabbit the Desckchicht disappears before the second layer of cells grows completely round the vesicle. 3. Blastodermic Vesicles "with. Primitive Streak. — The knot of Hensen marks the front end of the primitive streak, which lengthens backward; during the same period the vesicle as a whole enlarges ; in ruminants the enlargement is enormous and very rapid. * The primitive streak always lies in the long axis of the shield. The formation of the prim- itive axis begins with the union of Hensen's knot with the inner layer, so that at the knot all three layers are actually united — the condi- tion originally discovered by Hensen, 76.1, 268. The union of the knot with the inner layer spreads backward until it reaches the edge of the shield, thus generating the primitive streak. Next follows the elongation of the streak and shield, the latter becoming pointed at its hinder end. "We thus have a pear-shaped shield with the primitive streak running for- ward from its pointed end; the anterior end of ^^^ ^ _jimu, ^n.^^u,^,^ the primitive streak is somewhat enlarged and of'a^^bbit s'"ovum"'of"five the posterior end is considerably thickened ; the ^&*°andTh6'Ss^trib'ut.'on three layers are united along the primitive g^'^^^fiJl^^'^'j^'^ diamt'" streak. Fig. 85 represents the embryonic shield of a rabbit embryo; the shield measured 1.34 mm. in length and 0.85 mm. in width; the primitive streak is a broad band, corre- sponding to the axial thickening, and extends about two-thirds of * Bonnet states that in the sheep the blastodermic vesicle must elongate during this period at the rate of one centimetre an hour. FiQ 85 —Embryonic shield 140 THE GERM-LAYERS. the length of the shield; the mesoderm, m', in", occupies a circu- lar area around the hind end of the streak ; for a similar stage in the opossum see Selenka, 86.1, Taf. XVIII., Fig. 6; in the mole, Heape, 83.1, PL XXVIII., Fig. 12; in the sheep, Bonnet, 84.1, Taf. X., Fig. 39, 40. Cross-sections show the union of the three layers in the axis ; the greater width of the streak in front (to this wide anterior end of the streak the term Hensen's knot continues to he applied) ; and show also the increasing thickness of the streak posteriorly. The primitive groove, which is a shallow depression of the outer layer, appears first over Hensen's knot, and thence extends gradually backward along the median line of the primitive streak.. Fig. 86. — Section of the primitive streak of the Mole : p. (/r, primitive groove ; Ec^ ectoderm ;; mes, mesoderm ; En^ entoderm ; Pr, primitive streak. (In sections nearer the end of the groove^ Pr does not appear, and the inner layer is distinct, though not separated axially from the mid- dle layer. ) After Heape A transverse section through about the middle of the streak at this stage in the mole is represented in Fig. 86, and may be considered thoroughly typical. 4. Blastodermic Vesicles -with Primitive Streak and Head-Process. — In the stage we are now considering the axial thickening becomes subdivided into two parts, an anterior known as the head-process [Kopffortsatz) , and the true primitive streak. The two are distinguished by the fact that the axial thickening in the region of the process is separated from the outer layer but fused with the inner layer, while in the region of the streak it includes, as. in birds, the outer layer. Except at its anterior end, the axial thick- ening is not connected with the inner layer. Hence cross-sections may give us three different appearances according to the level at. which they are taken. The head-process was first distinguished, so far as I am aware, by KoUiker ("Entw.-Ges.," 1879,p. 271), also 83.1. Lieberkuhn, 82.1, first showed that in it appears a small longitudinal canal, the walls of which form the notochord. Heape, 83.1, discovered that the hinder end of this canal opens exteriorly in the mole, and Bonnet, 84.1, made the same observation on sheep. Strahl describes the "process" in the rabbit incidentally in his paper on the cloaca, 86.2; additional information is given by Bonnet, 89.1, concerning the sheep, and by C. Rabl, 89.2, concerning the rabbit. Especially valuable is Fr. Carius' dissertation, 88.1. In the guinea-pig, ac- cording to Carius, after the formation of the primitive streak the middle layer grows out in all directions and lies free between the- inner and outer layers. In front of the primitive streak the out- growth takes place in three divisions — one median, two lateral. The median outgrowth is the head-process proper, and it becomes later THE MAMMALIAN BLASTODERMIC VESICLE. 141 xinited with the inner layer, but at first lies entirely free (embryo of thirteen to fourteen days) . The first indications of the formation of a canal is an alteration of form in the cells, which elongate in directions at right angles to the axis of the head-process, so that their oval nuclei are radially placed; the change begins posteriorly and progresses forward ; while it is going on the anterior extremity of the head-process fuses with the inner layer. The radial cells move apart so that there arises a longitudinal canal ; subsequently the canal loses its inferior wall, so that it becomes continuous as a cavity with the cavity of the vesicle formed by the inner layer ; compare ante, p. 127. In the rabbit the head-process is also free at first, but very early imites with the inner layer, in which condition it was found by Carius, 18-19, at seven and a half days.* In the rabbit Hensen's knot presents at this stage a small depression (the front end of the primitive groove into which a small plug of tissue projects up from the underlying axial thickening (Carius' Fig. 7); Van Beneden homologizes this with the anus of Rusconi. The relations of the head-process in the sheep are very much as in the rabbit, Bonnet, 89.1, 65-67; the cells of the middle layer are at first free, as they grow forward to form the process, but subsequently are found united with the inner layer. The head-process (c/. Lieberkiihn) , 84. 1, probably always grows— as is certainly the case in the guinea-pig — at its hinder end and at the expense of the primitive streak ; it is, I think, in this manner that the often-noticed shortening and final disappearance of the streak is effected. This mode of growth concords with the concres- cence theory. Homologies of the Mammalian Blastocyst.— The homologies with corresponding stages of other vertebrates are uncertain. It seems clear that the main cavity of the two-layered vesicle corresponds to the yolk cavity and that the head-process is identical with the primitive axis. But the homologies during the stages of transition from the segmented ovum to the two-layered vesicle are uncertain, and must remain so until we understand the genesis, first of the yolk cavity, second of the primitive axis. Nor can the development be clear to us until the growth of the primitive axis by concrescence is elucidated. Inversion of the Germ-Layers in Rodents. — In many but hot in aU rodents the outer layer, Rauber's Deckschicht, of the embryonic shield undergoes a remarkable hypertrophy immediately after the close of segmentation proper ; the Deckschicht, together with the ectoderm underlying it, becomes a plug which pushes in the other layers, thereby profoundly altering the topography of the ovum. In the mole, Heape, 83.1, the hypertrophy is not very great and the plug disappears soon, so that there is no great change; in guinea- pigs, mice, and Arvicola, the plug becomes very large and remains for a long time. The plug is very long and the ovum elongates with it, changing into an almost cylindrical vesicle (Selenka's Keimcyl- *C. Eabl, 89.2, 143-145, states expressly that in the rabbit the axial thickening is not con- nected with the inner layer either under the head-process or under the primitive streak. He dif- fers from other investigators in this so much that I think his preparations were probably defec- tive; indeed, his own figures suggest at once that the inner layer has been artificially separated from the overlying one. 142 THE GERM-LAYERS. inder) . The plug becomes hollow, and the cells corresponding to the Deckschicht become separated from those which are to form the ectoderm of the embryo. Three modifications of the hollowing out of the plug and of the separation of its two parts are known. The changes referred to are very clearly illustrated by Selenka, 84.1, Taf. XVI., in a series of comparative diagrammatic figures. In the simplest case. Fig. 87, the plug acquires a single cavity, aj the cells around the upper end, TV, correspond to the Deckschicht and serve partly to attach the ovum to the uterine walls ; the cells, Ec, around the lower end of the cavity become the embryonic ectoderm ; all the cells around the cavity a are homologous with the outer layer of the embryonic shield of other mammals. The cavity c of the vesicle is very much reduced; the inner side of the shield, i.e. of the plug, is lined by an inner layer. En, which gives rise to the entoderm. The outer layer of the vesicle is very thin ; it unites very closely with the walls of the uterus, and later disappears. Hence, when the uterus is opened, only the hollow plug and its covering of entoderm can be removed ; as it makes a two- walled vesicle it was considered to represent by itself the two- layered stage of the blastodermic vesicle. Thus it came that Bischoff believed that in various ro- dents the ectoderm lies inside, the entoderm out- side. Bischoff 's observations, 52. 1, 70. 1, which were confirmed by Reichert, 62.1, are correct; but the inversion of the layers is apparent, not real. The actual homologies were not discovered - En until the improvements in microscopical technique enabled Selenka, 83.1, 84.1, and KupflEer, 82.3, to make sections of uteri with the ova in situ, and in their sections to follow the history of the outer vehicle oFM^^'syiTSi^ layer. Their results have been in the main con- ?t"/o£^4sTcitTk%nto- firmed by Fraser, 83.1, and extended to another derm; oi, outer layer; a, specics by Biehringer, 88.1, 91.1. toderm"^ A^fto'^^eSnka.' In Mus dccumanus the ectodermal cells early become a separate spherical mass, thus dividing the plug into two parts ; a cavity appears in each part ; these two cavities soon become confluent, and the inner layer of cells having meanwhile developed, the relations become essentially identical with those in Mus sylvaticus. Fig. 87. In Mus musculus the development is similar, but there is the additional peculiarity that the Deckschicht is regularly invaginated at first so as to form a small pit, into which uterine tissue grows. In Arvicola this invagination is more marked and lasts longer, but in both cases it is early obliterated. Arvicola represents the second modification mentioned above ; it has not only the invagination to distinguish it, but also the very early formation of the cavity of the plug as a fissure between the Deckschicht and the true ectodermal cells. The guinea-pig offers the third modification and, is characterized by the early complete separation of the plug into its two parts ; the Deckschicht remains at one end of the ovum and forms the Trager ; it acquires an independent cavity of its own ; the ectodermal portion THE MAMMALIAN BLASTODERMIC VESICLE. 143 of the plug forms a solid spherical mass which is transported to the opposite pole of the ovum ; it subsequently becomes hollowed out, pre- senting a space, which, as the later development shows, is the amniotic cavity. The inner layer passes from the edge of the Trager around the sphere of ectoderm ; if the two parts of the plug were connected the relations of the inner layer would be the same as in Mus sylvati- cus, Fig. 87. The subsequent development of the rodents with inverted layers is modified in various secondary features, which it will be unnecessary for us to study. In all typical respects the embryonic development agrees with that of other mammals even as to details. Duval, 90.2, has shown that in the rabbit the outer layer of the blastodermic vesicle degenerates and disappears, though at a niuch later stage than in the species just considered. Hence there is in the rabbit also potentially an inversion of the germ-layers. Graf Spee, 89.1, 170, suggests, and I think with considerable reason, that the earliest development of the human ovum takes place by inversion of the layers. If this hypothesis is correct, it explains many of the remarkable peculiarities of the youngest human ova known at the present time, CHAPTER VI. THE MESODERM AND THE CGELOM.* The morphology of the mesoderm is one of the most vexed ques- tions of the day. Scarcely an embryologist can be found who has not published opinions on this question considerably at variance with the opinions of others. It has been maintained that the mesoderm arises from the ectoderm ; that it arises from the entoderm, or from both ; from neither, but from two special segmentation spheres ; that it has a double origin, part coming from the blastoderm, part from the yolk, and even that there is no mesoderm. We now know positively that in all vertebrates there is a distinct and unmistakable mesoderm, which spreads out from the primitive streak in all directions and has distinctive histological character- istics. Two large and complex cavities appear in this mesoderm, one on each side of the median axial line. The mesodermic cells which bound these two cavities assume an epithelial arrangement, and are designated as the mesothelium; the cavities constitute the ccelom or primitive body cavity ; the mesothelium at various points throws off cells which compose the meseiichyma (embryonic connective tissue). We have, accordingly, three distinct phases to study, viz. : 1, the ori- gin of the mesoderm ; 2, formation of the ccelom and mesothelium ; 3, the origin of the mesenchyma. Finall}-, we must review the principal theories in regard to the morphological significance of the mesoderm. I. Origin of the Mesoderm. Mesoderm of Elasmobranchs. — In the cartilaginous fishes the mesoderm arises from the entoderm close to the ectental line. The observations of Balfour in his monograph, 78.3 (see also his works, I., 246-368), established the fact that the mesoderm appears after the two primary layers and is connected with the entoderm. This fact has since been abundantly confirmed, see Kollmann, 85.S Swaen, 87.1, Eiickert, 85.1, 87.1, Rabl, 89.2, D. Schwarz, 89. 1, et al. These later observations, particularly those of Riickert and Rabl, have settled the exact point, or rather area, of entoderm which is mesoblastogenic. Unfortunately Rabl overlooked the phenomena of concrescence, and consequently reached conclusions as to the development of the mesoderm which I feel no hesitation in pronouncing erroneous. The mesoderm is differentiated along the embryonic rim before concrescence takes place ; hence, when con- crescence is partly completed, there is an axial stretch of mesoderm. * This chapter has ah'eady been published in the American Naturalist, Oct. , 1890, but as here reprinted has been extensively altered. ORIGIN OF THE MESODERM. 145 and from the hind end of this the mesoderm stretches out toward each side along the embryonic rim in connection with the entoderm, as has been described in Chapter V. We can distinguish the axial mesoderm from the lateral mesoderm ; but later on, when concrescence has progressed farther, there is no lateral mesoderm, for it has be- come axial. Rabl, however, failed to study the later stages, and so came to consider that this temporary condition of the mesoderm signified a double origin ; accordingly, he distinguishes between the "gastral" (axial) and " peristomial" (lateral) mesoderm, and makes the unsuccessful attempt to show that the " gastral" and " peristomial" mesoderms are of essentially different origin in all vertebrates. Had Rabl understood concrescence he would certainly have not fallen into these errors. There is no positive evidence that there is an evag- ination of the entoderm as the Hertwigs maintain can be shown in the amphibians — see below. On the contrary, the cells grow forth from the entoderm so as to constitute a sheet between the primary germ-layers. Soon the connection with the entoderm is permanently severed. The fact that the mesoderm appears first in the embryonic rim renders it easy to make sure of its springing from the entoderm. Later, when concrescence moves the rim into the axial line, all three germ-layers are united in the primitive axis, and it becomes more difificult to decide which of the layers the mesoderm is specially con- nected with. To conclude : in elasmobranchs the mesoderm arises over a limited area of the entoderm near the ectental line ; it sepa- rates from the entoderm apparently by a process of delamination, but the exact means of separation have yet to be investigated ; it remains for a while connected with the entoderm along the embryonic axis ; after its separation from the entoderm the mesoderm expands by proliferation of its own cells and receives no accretions from the yolk, so far as at present known. Mesoderm of Teleosts. — So far as the published accounts go the middle layer of bony fishes arises as maintained by Balfour, '■ Comp. Embryol.," II., 74, from the entoderm. Such appears to be the significance of Ryder's observations, 84.1, 41, of A. Goette's, 73.1, E. Zielger's, Agassiz and Whitman's, 84.1, and of others. For a good description, together with citations of conflicting au- thorities, see M. Kowalewski, 86.1, 469-474. Apparently the blastodermic rim is turned under, and the turned-under portion yields the entoderm, and is intimately connected with the sheet of mesoder- mal cells, very much as in sharks ; the mesoderm is several layers thick and stretches in under the ectodermal blastoderm, gradually thinning out; the cells of the middle layer are at first closely com- pacted. Mesoderm of Am.pliibia. — Here it seems also clearly estab- lished that the mesoderm arises from the entoderm, principally along and alongside the median line, as a sheet of cells with no cavity (coelom), included between them; along the centre of the primitive axis and at the blastoporic margin the connection between the mesoderm and entoderm is both evident and intimate ; see Bellonci, 84.1, Tav. II., for figures showing this point in the axolotl, and 0. Schultze, 88.1, for similar figures of Rana fusca. These facts 10 146 THE GERM-LAYERS. have been recorded bj^ so many observers that there can be little doubt or none of their entire accuracy ; see the description and cuts, ante, p. 130. It may be considered as still uncertain whether the sheet of mesoderm receives accretions at its distal edge from the yolk cells (entodermic) upon which it rests. There usually is no sharp limit between the two, and therefore we must consider it probable that at first the mesoderm receives additions from the yolk ; later on it is found divided from the vitelline cells, and after it has split off it probably grows independently. The growth of the mesoderm at first from the yolk has been found in Petromyzon by A. E. Shipley, 88. 1,177, 178 (of " Studies"), although in later stages the mesoderm is severed from the yolk. In later stages the mesoderm is wanting in the median line, and thus constitutes two masses or two lateral sheets. This bilateral division is effected by the development of the medullary groove and notochord, as described in Chapter VIII. The mesodermic connec- tion with the entoderm is retained, but is double owing to the divis- ion. Along the median dorsal line of the archenteron runs the strip of entoderm which forms the notochord ; on each side of this strip runs the line of connection between entoderm and mesoderm. The study of this secondary condition has led many authors into the error of ascribing a double origin to the amphibian mesoderm, and inferentially to the vertebrate mesoderm in general. This brings us to the consideration of 0. Hertwig's views, which form one of the chief supports of the " Coelomtheorie" of the brothers Hertwig. For further discussion of this theory see p. 155. O. Hertwig, 82.1, 83.1, studied stages in which the notochord had appeared, and at this time, as O. Schultze, 88.1, has shown, the primitive relations of the lay- ers no longer exist, but Hertwig regarded the secondary arrange- ments in question as primary. He found no mesoderm in the axial line above the notochord; at the edge of the notochord, where it joins the undifferentiated epithe- lial entoderm of the archenteron, there is on each side a groove which in cross-sections appears as a notch. Fig. 88; the notch is of variable depth, is sometimes ab- sent, and is a temporary feature. In the neighborhood of the furrow alongside the notochord, the meso- FiG. 88.— Axoioti embryo; transverse section dermis Still intimately Connected of an early stage; £c, ectoderm; mes, meso- ,„-j.u 4.1, j. j rni i derm; Md, medullary groove; Cft, notochord; Wltn tne entoderm. Ihese rela- Serfluon™' ^^'''°^^' -^cft, archenteron. tjong are believed by Hertwig to indicate that the mesoderm arises as two masses, which is not the case, and that each mass is really a diverticulum of the archenteron, the furrow being the mouth of the diverticular cavity. Hertwig's figures, 83. 1, Taf. XIII., XIV., offer the plainest representations of the mesoderm in Triton as paired ORIGIN OF THE MESODERM. 147 diverticula ; but these figures * are evidently diagrammatic, and they must be termed inaccurate, I think, in the very respects vsrhich are essential to Hertwig's theory. This appears from the investigations of Goette, 75.1, Bellonci, 84.1, Bambeke, 68.1, 0. Schultze, 88.1, and others. Compare also K. Lampert, 83.1. The reader may compare, for instance, Hertwig's Fig. 10, I.e., Taf. XIII., with Bellonci's Fig. 11, I.e., Tav. III. 0. Schultze's detailed criticism. I.e., 314-349, of Hertwig's account seems to me entirely justified, and I accordingly accept it as a complete disproof. This criticism shows that Hertwig's conception is based upon insufiicient and erroneous observations ; insufficient because he did not investigate the early condition of the mesoderm, and failed to recognize the fugitive and unessential character of the parachordal grooves ; erroneous because the cavity in the mesoderm does not really communicate with that of the archenteron. There are other errors which Schultze points out and which are important. Robinson and Assheton, 91.1, 495, have also failed to verify Hertwig's statements. We find in Amphibia at a certain stage the axial (Rabl's gastrales) and lateral (Rabl's peristomales) mesoderm. The former is in the region of the completed concrescence, the latter round the edge of the anus of Rusconi. The former is connected with tlie entoderm alone, the latter with the ectoderm also, since the entoderm is connected with the ectoderm around the unconcresced blastoporic rim. The connection with the ectoderm renders it possible that the middle layer receives cell 3 from the outer layer, but there is no direct proof of this. When the concrescence is completed the mesoderm is said to sever, in the posterior axial region, its connection with the ento- derm, but to retain awhile its connection with the outer germ layer. The same phenomenon recurs in the amniota. It cannot, however, be taken to signify that the middle laj-er originates from the ecto- derm, since at an earlier stage it is clearly entodermal. Mesoderm of Sauropsida. — We may consider reptiles and birds together, since the early history of the middle layer is very similar in the two classes. In reptiles, so far as our present unsatisfactory knowledge en- ables us to judge, the mesoderm arises by delamination from the entoderm, but remains connected with it along the axial line ; in front {i.e., in the head-process) it is connected with the entoderm only, but posteriorly it is fused with the tissue of the primitive streak, and thereby is indirectly connected with the ectoderm. After its delamination the mesoderm expands independently of the other germ-layers except, perhaps, along the axis. That the relations are like those in birds has been shown clearly by Strahl, 83. 1, and also by Weldon, 83.1, whose Figure 1 is reproduced, ante. Fig. 71. The intimate connection of the mesoderm with the entoderm at the blastodermic rim before concrescence is sufficiently established by Kollmann, 84.3, 403-406, though his conception that this part of the mesoderm is a separate structure, which he terms akroblast, renders it difficult to foUow certain parts of his description. C. K. Hofmann may also be cited, though his account (Bronn's " Thier- *Some of them are reproduced in Hertwig's "Lehrbuch der Entwiekelungsgeschichte," 6tes Capitel. 148 THE GERM-LAYERS. reich, Reptilien," p. 1881) is of doubtful accuracy in several respects. L. Will, 89. 1, 1127, finds that in the gecko the mesoderm is united with the entoderm in the head-process, but omits to describe its exact connection with the primitive streak ; the stages showing the origin of the mesoderm he does not mention. The processes involved will undoubtedly be understood as soon as the concrescence of the axis has been worked out — a fundamental question, which as yet not a single investigator has heeded. In birds the exclusively entodermic origin of the mesoderm is, in my opinion, conclusively demonstrated by the researches of Duval, 84.1, 104-117; the entoderm gradually thickens by migrations of its cells over a considerable axial area; the upper stratum of this thickened area separates off as the mesoderm except that in the axial line it retains its connection with the entoderm ; when concrescence takes place the two layers form the primitive axis. In the region of the primitive streak there is a single large mass of cells. Fig. 71, Pr, which is continuous with all three germ-layers. Now if the homology maintained in the previous chapter be correct between the primitive streak and the anus of Rusconi, then the cells of the streak are also entodermal, and the middle germ-layer is connected in both axial regions directly only with the entoderm. After the mesoderm has separated from the entoderm except in the median line it may continue to receive accretions from the entoderm in the median line, but, as far as known, makes no peripheral additions except from its own growth. So far as heretofore observed the mesoderm receives no cells from the ectoderm. Mesoderm of Mammals. — In this class, according to the best recent investigations, the mesoderm appears to have a distinctly twofold origin. According to Bonnet, 84.1, 196, part of the meso- derm grows out from Hensen's knot at a time when the knot is a thickening of the outer layer and has not yet acquired any connection with the inner layer ; another portion is produced peripherally. Fig. 84, nies, by delamination from the inner layer; the two anlages grow toward one another and unite into one continuous mesoderm, in which all trace of the primitive double origin is obliterated. Kolliker has recorded the outgrowth of the mesoderm from Hensen's knot in the rabbit, and his statement has been confirmed by Fr. Carius, 88. 1, 17. In later stages we find the relations of the layers similar to those in Sauropsida, there being a head-process with the mesoderm connected axially with the inner layer, and a primitive streak, with which the mesoderm fuses ; the inner layer of the blas- todermic vesicle is connected with the front part of the streak. This stage is quite well known, cf. Heape, 83.1, on the mole, Bonnet on the sheep, 84.1, KoUiker on the rabbit (" Grundriss ") , Selenka on the opossum, 86.1, Lieberkuhn, 82.1, and others, especially the very careful descriptions of the rabbit's layers by C. Rabl, 89.2. At present it seems to me impossible to offer any satisfactory inter- pretation of the observed double origin of the mammalian mesoderm. The relations of the mesoderm to the primitive axis (head-process) and primitive streak are identical with those in birds and reptiles. The Vertebrate Type of Origin of the Mesoderm.— The preceding paragraphs show that in all classes of vertebrates the ORIGIN OF THE MESODERM. 149 origin of the mesoderm is essentially the same, except that in some mammals it begins in two regions of the entoderm almost simultane- ously. The relations in the mammals we do not understand. In the non-mammalian vertebrates the mesoderm first appears as a thicken- ing of the entoderm over a not inconsiderable area around the concres- cing blastodermic rim, and it becomes separated from the entoderm by the gradual parting of the upper cells to form the true mesoderm from the lower cells or permanent entoderm ; this delamination does not take place next the blastodermic rim (or — after concrescence — in the axial line) ; hence in the region of the primitive axis the three layers may be connected for a tim'e ; further, as the tissue of the primitive streak is at first connected with the ectoderm, the mesoderm is thereby indirectly continuous with the outer germ-layer during very early stages. It is important to note that the mesoderm arises over a considerable area during the same period ; that its formation may be more or less advanced before concrescence of the rim ; and that after concrescence it stretches across the axis of the embryo between the ectoderm and entoderm, thus becoming a continuous sheet or layer. This fact that the mesoderm is a single anlage needs to be specially emphasized. So far as known to me there is not a. single vertebrate which has been shown to lack this stage, but on the con- trary its occurrence is established for all classes and by so many observers that we may well assert that there are few facts in embry- ology better established. Later the mesoderm becomes divided in the axial line, * and consideration of this secondary condition has led to several theories of the mesoderm, which would hardly have been brought forward had their authors not neglected to take into account the earlier condition of the middle layer. Some of these theories are discussed below. After its delamination the mesoderm is a distinct layer and grows independently, receiving no accretions from the other layers except in the axial line, where it receives cells from the entoderm and in the region of the primitive streak. The edge of the expanding sheet of mesoderm is free, as has been pointed out in the previous chapter, resting upon the yolk but not fused with it. It is therefore, it seems to me, impossible to admit that there is a peripheral ingrowth of tissues arising from the yolk and entering the mesoderm to form the blood, etc. Compare below. Theories of the Mesoderin, p. 153. The primitive mesodermic cells are embryonic in character; that is, they have a large, usually nucleolated, nucleus, and very little protoplasm (Minot, 125) . They are connected together by fine threads, and may lie some distance apart, then presenting an obvious resemblance to the mesenchyma of later stages. The cells become more closely compacted as development progresses, and when the coelom appears they take on a distinctly epithelial arrangement to make the mesothelium. The cells frequently contain yolk grains— in the case of Amphibia numerous and large ones. In birds the yolk grains are few, but are easily observed, Fig. 81 ; in mammals they are almost entirely absent. *Mitsukuri, 91.1, has attempted to deny the views I have advanced, because in titles the mesoderm is divided, as shown by his own observations. He has overlooked the fact that his observations refer to the secondary stage only when the medullary groove and notochord are present, and that they have no bearing on the question of the earlier and primitive condition. 150 THE GERM-LAYERS. Expansion of the Mesoderm. — After the mesoderm is once formed as a distinct layer without connection with the primitive layers except in the axial line, it expands independently — that is, by the proliferation of its own cells. During its early expansion the -mesoderm assumes in all amniota a definite series of characteristic outlines. It is at first pear-shaped, Fig. 89, A, the anterior end be- ing pointed ; it extends a short distance only in front of the primitive streak, and is widest a little distance behind the area pellucida, aj). The same stage is found in mammals, see KoUiker, ("Grundriss," p. 93 and Fig. 71.) The condition in the chick at about the twentieth hour of incubation is indicated by Fig. 89, B, drawn on the same - C Fig. 89. —Diagrams of the embryonic area of the chick : Ao^ area opaca ; Ap^ area pellucida ; pr, prim- itive groove; mes, mesoderm. After Duval. Fig. 90. — Diagram of the embryonic area of a chick : Ao^ area opaca ; Ap^ area pellucida ; pr, primitive groove ; nies, mesoderm. After Duval. scale as A, and at the close of the first day by Fig. 90. In the last mentioned figure it will be noticed that the mesoderm is expanding unequally in front, having sent out two lateral wings which leave a median space between them without mesoderm. These wings con- tinue their growth and finally meet in front, so that in the anterior part of the area pellucida there is a small tract without any mesoderm, although there is mesoderm all around it ; this tract is the proamnion, of which a fuller history is given in Chapter XV. The expansion does not take place by any means with the exact regularity indicated by Figs. 89, 90, but, on the contrary, in birds, as shown by Zum- stein, 87.1, the outline of the middle layer is always irregular and more or less asymmetrical. Although there are not yet many ob- servations available as to the outline of the growing mesoderm, yet it is probable that the preceding description is essentially correct, not merely for birds but for all amniota. It is certainly so for the rabbit. Van Benedenet Julin, 84.1. II. Formation of the Ccelom and Mesothelium. Early in the course of development there appears in the mesoderm a complex series of cavities, which very soon become united so as to form two large cavities, one on each side, which together constitute the ccelom or embryonic body cavity. In the adult mammal the coelom is represented by the pericardial, pleural, and abdominal cavities ; the coelom also includes the cavities of the muscular seg- FORMATION OF THE CCELOM AND MESOTHELIUM. 151 ments (protovertebrse) and also certain tubular parts of the urogeni- tal system. But although its subsequent changes are complex the coelom consists at an early stage of a pair of tissures in the meso- derm. As the coelomatic cavities appear the cells bounding them take on a distmctly epithelial character. The mesodermic epithelium boundmg the coelom is termed the mesothelium, and it is probable —It we judge from our present imperfect knowledge— that the entire mesoderm is in all vertebrates first converted into mesothelium be- fore undergoing differentiation. _ Only one precise account of the mode of development of the coelom m mammals is known to me, namely, that of Bonnet, 84 1 303 for the sheep at about thirteen days. Around the embryo at' some distance from, the axis there appear a series of irregular fissures of rounded or elongated form, which may in part open on the mesoder- mic surface ; gradually the fissures enlarge and fuse, at the same time becoming more closely bounded by the mesodermic cells; thus there arises a continuous cavity in the mesoderm which is for a time crossed by cells and cell processes ; similar connections between the two leaves of the mesoderm while the coelom is forming and their subsequent rupture have been noticed in Amphibia by B. Solger, 85.1, 383 in Elasmobranchs by E. Ziegler, 88.1, 383, and I find similar phases with great distinctness in the chick; meanwhile the cells, which are loosely put together, form a compact layer of epithelium bounding the cavity, which we can now designate as the ccelom or primitive body cavity. By similar processes the ccelom grows toward the axial region, but never penetrates it, the primitive streak and head-process never developing a median ccelom. Albrecht Budge, 87. 1, has made a very exact study of the arrange- ment of the fissures in the mesoderm of the chick by means of in- jections of Prussian blue.* The fissures form a network of channels and by their fusion produce the coelomatic cavities. The channels appear first around the periphery of the area vasculosa, and thence their development progresses cen- trifugally,but most rapidly toward the head; the channels fuse first around the head to make the am- nio-cardial coelom {Paiietalhohle of His) ; now appears a circular sinus just inside the vena termi- nalis; the coelom grows back through the embryo and forms the body cavity of the rump ; alongside the rump, as shown in Fig. 91, appears a network of channels, which soon fuse to create the coelom under the lateral amniotic fold, and this unites with the coelom of the rump, forming the completed coelom continuous with that of the * The injections are made either into the amnio-cardial vesicles or into the circular fissure just inside the vena tcrminalis. Fig. 91.— The mesodermal cavities ot the germinal area of a chick of the third day, in- jected with Prussian blue. After A. Budge. 152 THE GERM-LAYERS. pericardium. The network of channels Budge regards as primary lymph spaces. Compare Chapter XIX. Whether in all vertebrates the coelom results from the fusion of numerous small spaces or not, is not yet determined by actual obser- vation. It is probable that it does so, and we may, therefore, say that the vertebrate ccelom is what Huxley terms a schizocoele, i.e., a cavity produced by splitting the mesoderm, compare p. 155. I con- sider it also probable that the ccslom always begins to appear at a little distance from the axis of the embryo and spreads both centrip- etally and centrifugally. Additional and important points in the earliest history of the coelom are treated in Chapter IX. We must add here that the coelomatic fissure divides the mesoderm on each side into an upper or outer leaf (Hautfaserblatt) and a lower or inner leaf (Darmfaser- blatt), Fig. 93. The upper leaf may be called the somatic meso- derm, Som, the lower leaf the splanchnic mesoderm, Spl, as pro- posed by Balfour. The upper leaf lies close against the ectoderm ; the two layers together form the somatopleure or body-wall. The lower leaf lies close against the entoderm ; these two layers together Ch. Fio. 93.— Section of a chicken embryo of about thirty-six hours : Ec, ectoderm; Som, som- atic mesoderm ; Sp!, splanchnic mesoderm; Ent, entoderm; W, WoltBan duct; m, mesenchymal cells; Md, medullary groove; v, vein; Cos, ccelom; MS, primitive segment; Oh, notochord. After W. Waldeyer. form the splanchnopleure or wall of the alimentary tract. Both the somatic leaf of mesoderm and the splanchnic consist at first solely of mesothelium, but very soon each contains mesenchyma also ; the latter arises from the mesothelium ; axially the two layers be- come continuous both with one another and with the axial meso- derm. The morphology of the coelom is so important that it is difficult to understand why so many investigators have slurred over the question of its embryonic development. Exact observations on its first appearance and on the first stages of its expansion in various types are urgently needed, and would certainly do more than any- thing else to throw light on the stiU obscure problem of the origin of the mesoderm. The histogenesis of the mesothelium varies somewhat in the different types. Primitively (marsiobranchs, amphibians) , the cells are rounded in form, contain considerable yolk, and are at first loosely aggregated, compare Fig. 88. When the coelom appears the cells become more closely appressed and so gradually assume more and more the characteristics of a cuboidal epithelium. In the amniota, on the other, hand, the mesodermal cells contain very little THE MESENCHYMA — THE MESODERM. 153 yolk, Fig. 81, in whicJi the yolk grains are shown as black dots; the cells are connected by their processes ; as the coelom develops the processes are shortened, and the cells become more closely packed, and thus gradually arrange themselves into a cuboidal mesothelium. III. Origin op the Mesbnchyma. The genesis of the mesenchyma is treated in Chapter IX. , as it cannot be understood without a knowledge of the development of the primitive segments. I will, therefore, merely state here the general methods of its production in order to render intelligible the following discussion of the theories of the mesoderm. By mesenchyma we understand the whole of the mesoderm of the embryo, except the mesothelial lining of the coelom. So far as at present demonstrated it arises solely from the mesothelium. Single •cells leave the mesothelium on the side away from the ccelom ; these cells remain connected with one another and with the mesothelial cells by protoplasmatic processes, but they do not lie close together as in an epithelium, so there is a considerable though variable amount of intercellular space. By the migration of the cells and their multiplication a large amount of mesodermic tissue is pro- duced, which fills up all the room between the mesothelium and the two primary germ-layers. At first no definite distinction between the mesothelium and the mesenchyma can be established, but ultimately they become and remain distinct tissues, with divergent histories. IV. Theories of the Mesoderm.* From the time of Von Baer's " Entwickelungsgeschichte," of which the first part appeared in 1828, until 1868, when W. His' great monograph on the chick, 68.1, was published, embryologists recog- nized the three layers, and regarded the mesoderm as a natural unit. His led the way to our present conception by a little-known article, 6 5 . 1 , on the membranes and cavities of the body, and his monograph, 68.1, above mentioned fully established the necessity of recogniz- ing two main groups of mesodermic tissues; accordingly he divided the mesoderm into two parts, the archiblast and parablast, corre- sponding respectively essentially to mesothelium and mesenchyma. Under archiblast His included not only the mesothelial tissues proper, but also the smooth or organic musculature; under parablast the mesenchymic tissue except the smooth muscle. The terms used cor- responded to his theory of the origin of the two parts of the meso- derm, for he believed that the archiblast arose in the axial region and was contained in the embryo from the start, while the parablast arose peripherally and grew in toward the embryo, a conception which was perhaps suggested by the appearance of the blood-vessels, first, outside the embryo proper. Seeking still farther for the source of the supposed peripheral parablast he believed he had found it in the germinal wall. The study of the relations of the wall in the <;hick induced him to think that the elements of the white yolk be- < Cf. ante, p. 149. 154 THE GERM-LAYERS. came parablast cells; moreover, the study of the hen's ovary led him. to the conclusion that the white yolk was developed from the granu- losa cells, and that these cells arise from leucocytes. He thus traced back the parablastic cells to maternal leucocytes. As subsequent chapters will show more fully, the granulosa cells are not leucocytes ; in Chapter III. it has already been shown that the granulosa cells do not enter the ovum ; the white yolk grains never become cells, for it has been proved that all nuclei of the segmenting ovum come from previous nuclei and lie in protoplasm, not in the yolk grains ; and, finally, it has been shown in this chapter that the mesoderm arises as a whole, not from double sources. Professor His' views as to the origin of the parablast must be given up, but this is no reason for overlooking, as certain writers have done, the fundamental significance of the distinction drawn between the two primary groups, of mesodermic tissues. Subsequent research has made only one im- portant change necessary — namely, the transfers of smooth muscula- ture from one group to the other. In view of this change, of the fact that parablast has been used with various other meanings, and of the unaptness of His' names — since we renounce the theory they correspond to — it will be well to use exclusively the newer terms, mesothelium and mesenchyma. The parablast theory has been defended by His, 76.2, and modified by him, 82.1. At present he holds to the distinction originally drawn, but is inclined to withdraw his hypothesis of the origin of the parablast. A number of writers have agreed with His as to the separate peripheral development of the mesenchyma (parablast). Among these maybe mentioned Rauber, 77.1, 83.4, and several authors who have dealt with the development of the blood, see Chap- ter X. The most important of the disciples of His is KoUmann, who, in a series of articles, 84.1, 3, 85.1, 2, has maintained the double origin of the mesoderm. Of these papers the most important is that on the "Randwulst," or germinal- wall, of the structure of which in the chick it gives an excellent description. KoUmann regards the germinal-wall not as part of the entoderm, but as a distinct organ composed of segmentation spheres, and destined to produce blood- vessels with blood, and probably also connective tissue ; this peripheral anlage {Eandkeim) he designates as acroblast, and the single ceUa. derived from it he names pore uten. Waldej-er, 83.1, has ac- cepted the parablast theory, but with a modification by which he seeks to reconcile conflicting observations. His article is written with characteristic clearness and exhaustive mastery of the literature, and will be found especially valuable by those who wish to pursue this subject farther. Waldeyer distinguishes between the primary and secondary segmentation ; the former producing the ectoderm, "entoderm, and archiblastic mesoderm, the latter occurring later and giving rise to the parablast. According to Waldeyer this remnant of the ovum (which in holoblastic ova consists of cells, in meroblastic ova of egg protoplasm) has its cell division (segmentation) retarded, and the cells thus tardily produced immigrate into and between the germ-layers already developed. The opposition to the parablast theory is the sum of numerous ob- servations, which, as pointed out in the previous part of this chapter, THEORIES OF THE MESODERM. 155 prove — it seems to me — that the mesoderm arises in all vertebrates (except mammals?) as a unit, and subsequently separates into meso- thelium and mesenchyma. The leading opponent of the separate origin of the parablast is KoUiker in both his text-books (" Entwick- elungsgeschichte," etc., and " Grundriss") , and in separate articles, see especially, 84.2, 4, andhiscriticism, 85.3, of Kollmann. lagree with Kolliker that it has been sufficiently demonstrated that the " akroblast" belongs to the entoderm, and after the delamination of the mesoderm is transformed into the epithelium of the yolk-sac ; for a conclusive demonstration that this is so in reptiles, see H. Strahl, 87.1. The ccelom theory of the brothers Hertwig includes a fundamental modification of the parablast theory. The main features of the ccelom theory are not original with the Hertwigs, but may be found in previous writers ; nevertheless they were the first to present the theory in a complete formula and with a backing of facts, both new and collated, from others so extensive as to compel attention. In justice to E. Ray Lankester it must be stated that he is really the author of the coelom theory, having in 1877 (77.1) published the hypothesis that the coslom is derived from the archenteron, and that the mesoderm of vertebrates represents solid entodermal diverticula. It is unfortunate that the Hertwigs have not made due acknowledg- ment of what they owed to Lankester and others. They made a series of investigations on the germ-layers of various representatives of the animal kingdom, and presented their general results in a com- prehensive article (0. and R. Hertwig, 81.1), and O. Hertwig has again expounded the theory in his text-book of embryology. The ccelom theory consists of two parts : 1, the coelom is formed by diver- ticula of the archenteron and its lining, the mesothelium, is part of ' the entoderm; 2, the mesenchyma consists of cells thrown off by the other germ-layers and is essentially distinct from the mesothe- lium. The value of this theory lay in the clearness of its formula- tion, thus facilitating discussion, and also in its bringing out the difference more clearly between the epithelial and the non-epithelial portions of the mesoderm. As we have seen, there is no evidence of a character to render even probable that the mesoderm of vertebrates represents archenteric diverticula, and the whole mesoderm appears as a single germ-layer, which is subsequently differentiated into mesenchyma and mesothelium. Hence, both essential parts of the coelom theory are inapplicable to vertebrates, at least in the present state of our knowledge. For further discussion of the difficulties of the Hertwigs' theory, see Rabl, 89.2, 198-202, also Alex. Goette, 90.1, 18, as well asp. 146. The Hertwigs recognized the signifi- cance of the parablast and added the important rectification, which Flemmings' observations, 78.2, had already rendered necessary, of separating the smooth muscles from the striated skeletal muscles — a separation the propriety of which was wrongly questioned by Bal- four, "Comp. Embryol. XL," 35'.). By this advance the two groups of mesodermal tissues became properly delimited. C. RabVs theory of the mesoderm is based, it seems to me, wholly upon his failure to understand the law of concrescence. That the mesoderm appears (perhaps in all vertebrates) while concrescence is 156 THE GERM-LAYERS. going on is well ascertained ; consequently, there is an axial meso- derm (Rabl's " gastrales mesoderm") where concrescence has taken place, and a lateral mesoderm (Rabl's " peristomales mesoderm") in the part of the blastodermic rim which has not concresced. Until Rabl proves that his " peristomales" mesoderm does not become axial mesoderm in later stages his theory can have no standing. Davidoflf, 90.1, 613, makes the best criticism of Rabl's theory which I have seen. Rabl's memoir brings out one point of very great importance for the elucidation of the early stages of vertebrates — namely, that the " peristomal" mesoderm, in other words, that of the blastodermic rim in selachians, and of the lips of the anus of Rusconi in amphibians, is represented in the amniota by the mesoderm of the primitive streak. If this interpretation, which is much strengthened by L. Will's researches on the gecko, 89.1, be verified, then the primitive streak is the homologue in amniota of the anus of Rusconi, and is the region where concrescence is incomplete ; the head-process is then the part where concrescence is finished ; this concords with the ob- served fact that the head-process grows at the expense of the primitive streak, as it would do if concrescence continued. Alexander Goette's theory, 90.1, 24-33, is that the walls of the archenteron in Amphioxus and the true vertebrates comprise a dorsal region which develops the notochord and mesoderm, and a ventral region which develops the digestive tract. Owing to the great amount of yolk in true vertebrates the dorsal region is spread so as to lie upon the yolk, hence it is separated from the yolk or entoderm by delamination instead of forming a true evagination as in Amphioxus. It occurs to me that Goette's theory raay be perhaps verified with the modification that the notochordal canal corresponds to his dorsal region, the yolk cavity to his ventral region of the archenteron. Hatschek^s germ-band theory o&ers, to my mind, the best-founded explanation of the vertebrate mesoderm, because it connects it with the mode of development of the middle layer in the annelids and other invertebrates. To understand the theory we must first consider the formation of the mesoderm in Amphioxus. The ovum of Amphioxus is discharged from the body and impregnated external- ly; it is about 0.105 mm. in diameter, and, as it contains only a small amount of yolk, undergoes a holoblastic segmen- tation, which results in a well-marked blastula stage. Fig. 60, followed by a gastrula stage. The gastrula elongates, the blastopore remaining open at the pos- terior extremity. Differentiations now take place by which the ectoderm forms the axial anlage of the nervous system, and the entoderm produces the notochord and the mesoderm — the three processes going on simultaneously. The accompanying Fig. 93 represents a cross-section of a larva'. Fig. 93. — Transverse section of an Amphioxus embryo: JWd, medullary plate; Ms, primitive segment; Ch. notoonord; Ent, entoderm; Ec, ecto- derm; Jn, archenteric cavity. After B. Hatsohek. THEORIES OF THE MESODERM. 157 The ectoderm, Ec, everywhere bounds the section ; on the dorsal side a portion of the ectoderm has been separated ofE to form the medul- lary plate, Md, above which is a small cavity. The cavity, In, of the archenteron is irregular, but symmetrical in outline ; the entoderm bounding it can be separated into four parts: 1, the lower portion, which forms the permanent entoderm, Ent; 2, the upper median por- tion, which becomes the notochord, Ch, compare Chapter VII. ; 3, 4th, two lateral portions constituting the diverticula, Ms; each diver- ticulum is a separate pouch, and as the development progresses there are formed a series of pairs of pouches, stretching on either side along the notochord ; later the pouches separate altogether from the archen- teron, each becoming a closed sack; the first pair of pouches, how- FiG. 94. — Amphioxus embryo: A, side view; B, ventral view. Ec, ectoderm; En^ ento- derm; a, neuropore; JV. nervous system; Mes, mesoderm; Mb, mesoblast; 1-5, segments. After B. Hatschek. ever, retain their connection for a considerable period with the archenteron, and have been described by older writers as glandular organs. The development of the pouches is, with the exception noted, most advanced anteriorlj^, and as we go tailward the pouches are less and less advanced in development, until, as shown in Fig. 94, they merge into the general entoderm as a band of cells, Mes, the last of which is the mesoblast, Mb, a large granular cell quite dis- tinct from the remaining cells of the band or pouches. The pouches are the primitive segments {Uisegineiite, mesoblastic somites of Bal- four). In Amphioxus, then, the mesoderm arises from the entoderm along two lines, and is divided into paired hollow segments before it is separated from the entoderm. Some writers, especially the brothers Hertwig, think this process of development to be primitive, and that the vertebrate type is derived from it. In true vertebrates the mesoderm arises on each side, but also in the axis, and becomes two masses when the medullary groove and notochord appear; in 158 THE GERM-LAYERS. Amphioxus the medullary plate and notochord appear very early, and the division of the mesoderm may be due to that fact. Amphioxus is undoubtedly a lower type, but whether it really preserves the older type of development in its purity is doubtful ; indeed it is prob- ably a tunicate rather than a vertebrate. Hatschek, in a series of brilliant investigations, has shown that in many bilaterally symmetrical invertebrates the mesoderm arises as two bands of cells, which subsequently divide into a series of closed sacks (segments) , and which during their own formation terminate each in a single large posterior cell (mesoblast) , which throws off cells to add to the mesodermal band (germ -band, Keimstreif) . This mesoblast, by its appearance and position, appears to be a derivative of the entoderm. As a matter of speculation we may assume that in Amphioxus we have the germ-bands, but characterized by an ex- ceedingly precocious segmentation. We can further assume that in vertebrates we have the germ-bands also, but that they are modified, 1, by the loss of the distinct terminal mesoblast ; 2, by precocious fusion in the axial line, and 3, by extremely retarded segmentation. A great deal may undoubtedly be said in favor of these three assump- tions, which together constitute that theory of the vertebrate meso- derm" which, of the many theories, that have been advanced, is most likely, in my opinion, to prove of permanent value. CHAPTER VII. GENERAL REMARKS ON THE GERM-LAYERS. In this chapter the general morphology and role of the germ- layers, the history of the theory of the germ-layers, and the laws of differentiation are briefly considered. I. Role op the Germ-Layers. It has long been known that the bodies of embryos consist of dis- tinct layers, which, in many cases, are separable from one another, so as to be recognized in gross as discrete membranes. It is now known that all such layers may be reduced to three primitive ones, named the ectoderm, mesoderm, and entoderm (by certain writers, epiblast, mesoblast, and hypoblast). The ectoderm is a layer of epithelium ; so also is the entoderm ; the mesoderm is more complex, and as we ascend the animal scale the mesoderm gradually acquires a greater predominance until in mammals nearly the whole bulk consists of mesoderm. But in spite of this change, the three layers are preserved throughout, and their essential relations are not altered, so that we are able to assert that unitj' of organization without which it would be impossible to accept the doctrine of evolution. The dem- onstration of the homologies of the germ-layers is the most important morphological generalization since the establishment of the cell- doctrine. As the history of all the organs is given in detail in other chapters, it is unnecessary to do more here than classify the tissues and organs of the human body according to the germ-layers from which they arise. Now, in classifying organs, it is best to rank them as belong- ing to that layer from which their functionally essential and char- acteristic part is derived. Thus, although the pancreas, ovary, and spinal cord all contain connective tissue, we do not caU them mesenchymal, but respectively entodermal, mesothelial, and ecto- dermal. The gland cells of the pancreas come from the entoderm ; the ova and the Graafian follicles come from the mesothelium ; the ganglion cells and nerve fibres (axis cylinders) from the ectodei^n. Adopting this principle we may classify the organs of the human body as follows : 160 THE GERM-LAYERS. ECTODERMAL. Skin (epidermis). Epidermal structures : — Hairs. Nails. Glands : — Sebaceous. Sudorific. Salivary. Mammary. Corneal epithelium. Lens of eye. Central nervous system :- Ganglia. Nerves. Eye,:— Optic vesicle. Optic nerve. Olfactory organ. Auditory organ. Mouth cavity : — Teeth. Hypophysis. Anus. Chorion : — Placenta. Amnion. MESODERMAL. 1. Mesothelium. Peritoneum. Pleurae. Pericardium. Urogenital. Wolffian bodj'. Kidney. Testes. Ovary. Oviduct. Uterus. Vagina, etc. Striated muscle. 3. Mesenchyma. Connective tissues. Blood. Blood-vessels. Lymphatics. Spleen. Smooth muscle. Fat- cells. Marrow. Skeleton. ENTODERMAL. Epithelium (of digestive tract) . Thyroid. Thymus. Tonsils. Trachea and lungs. (Esophagus. Stomach. Liver. Pancreas. Intestine. Yolk- sack. Caecum. Vermix. Colon. Rectum. AUantois : — (Bladder) . Notochord. The human body may be defined as two tubes of epithelium, one inside the other ; the outer tube (epidermal or ectodermal) is very irregular in its form ; the inner tube (entodermal) is much smaller in diameter, but much longer than the outer and has a number of branches (lung, pancreas, etc.), and is placed within the ectodermal tube. Between these two tubes is the very bulky mesoderm, which is divided by large cavities (abdominal and thoracic) into two main layers, one of which is closely associated with the epidermis and forms the body- wall, the somatopleure of embryologists ; the other joins with the entoderm to complete the walls of the splanchnic viscera, and constitutes the splanchnopleure of embryologists. The mesoderm is permeated by two sets of cavities: 1, the heart and blood-vessels ; 2, the lymphatic system. It is also differentiated into numerous tissues, muscle, tendon, bone, etc., and organs, urogenital system. The nervous system, although developed from the ecto- derm, is found separated from its site of origin, and completely en- cased in mesoderm. As we ascend the animal scale, we discover in all parts an increas- ing complexity ; especially in the nervous system is this marked, but it is even more strikingly shown by the evolution of the mesoderm in relative size and differentiation. This important correspondence between the organization of the mesoderm and the degree of evolu- tion of animals has not, to my knowledge, hitherto attracted express attention. II. Differentiation. The fundamental law of embryology is that the simple precedes the complex, the general and typical the special. Each germ-layer is at DIFFERENTIATION. 161 first a simple layer of cells of nearly uniform character. In order to develop out of the germ-layers the complex organs of the adult the layers have to be folded into various forms by unequal growth of their parts, and the cells composing them have to be specialized some in one way, some in another. This double process results in the differ- entiation of the organs. Differentiation may be defined as the proc- ess of change from homogeneous to heterogeneous structure, or as an increase of heterogeneity, since in living organisms there is no real homogeneity. From what has been just said it will be understood that under the present head we have to consider, 1, the laws of un- equal growth ; 2, the general laws of cellular differentiation, or, as it is called, histogenesis — the development of tissue. The delations of Surface to Mass. — However much the weight of an animal increases during its development, the ratio of -the free surface to the mass alters but slightly from the ratio estab- lished when the embrj'-o begins to take food from outside. It is only for convenience that I express this law in this precise form; in reality, about it our knowledge is scanty and our conceptions vague. According to a geometrical principle, when the bulk of a body bounded by a simple surface increases, the surface enlarges less than the mass — in the simplest case of a cube, the surface increases as the square, the mass as the cube, of the diameter. If in a cube of unit diameter one unit of surface bounds one unit of mass, then in a cube of three units diameter nine units of surface will bound twenty-seven units of mass; the proportion in the first cube is 1 : 1, in the second 1:3. To maintain the proper proportion in the embryo, simple en- largement is insufficient, therefore the surface increases by becoming more and more irregular. The irregularities are characteristic of each organ and part, and may be either large or microscopic. They may be conveniently grouped under two main heads — projections and invaginations. Projections are illustrated by the limbs, filaments of the gills in fishes, the villi of the intestine, folds of the stomach in ruminants, etc. In every case the projection is covered by an epithelium and has a core of mesodermic tissue. Invaginations exist in much more varied form and play the principal part in the differentiation of the animal body. They may be classified under four principal heads: 1, Dilatations: 2, Diver- ticula; 3, Glands; 4, Vesicles. X)i7atoh'oJis have considerable im- portance in embryology; the stomach, lungs, bladder, and uterus arise as gradual dilatations of canals or tubes of originally nearly uniform diameters. Diverticula in the sense of relatively large blind pouches also form important organs, such as the caecum and appendix vermiformis, or the gall bladder; these structures arise, each as a blind outgrowth of a canal, the walls of which at a certain point rapidly grow to form the pouch. Glands, which are, as first shown by Johannes Midler's classic researches, only small diverticu- la, which end blindly and appear in an immense variety of modifica- tions ; the manifold types of glands are discussed below in a separate paragraph ; they constitute the largest class of organs with which we have to deal. The glands are developed from epithelium and push their way into the mesoderm upon which the epithelium rests, while 11 163 THE GERM-LAYERS. in dilatations, and in diverticula, the epithelium and mesoderm expand together. Vesicles we call those epithelial sacs, which develop somewhat like glands by growing into the mesoderm, but the mouth of the invagination closes by the coalescence of the epithelium, thus shutting the cavity. The closed sac separates from the epithelium from which it arose, and connective tissue grows between the two ; the sac may then undergo various modifications. The membraneous labyrinth of the ear is developed from the ectoderm in this way, as is also the lens of the eye. We might perhaps also class the medul- lary canal under this head (c/. Chap. VIII.) if we choose to consider it as a v.esicle so much lengthened that it has become a tube. The Law of Unequal Growth.. — The changing shapes of the embryo and the development of those irregularities — projections and invaginations, which preserve the proper proportion between the sur- face and mass of the body, both depend upon the unequal growth of the germ-layers, especially in superficies. The expansion of a,germ- layer having the epithelial type of structure* may take place by three means: 1, the multiplication of the cells; 3, the flattening out of the cells ; 3, enlargement of the cells. In the early stages of de- velopment the influence of the first two factors predominates ; during the later stages, especially after birthy the latter. Of the three factors the first is the most important. The unequal multiplications of the cells in all embryonic epithelia is the fundamental factor of development, and we see it shaping out the embryo, its organs, and the parts of organs, before histological differentiation really begins. The distinct areas and centres of grovdih which are necessary to develop the human body out of the germ-layers are innumerable, and their distribution, limitations, and interactions make up a large part of the subject-matter of embryology. At every turn of our studies we encounter fresh illustrations. If in a limited area of a cellular membrane there occurs a growth or ex- pansion more rapid than in the neighboring parts, then that area is, as it were, bounded by a fixed ring, and can, therefore, find room for its own expansion only by rising above the level of the mem- brane ; thus when in the embryonic region of the blastodermic vesicle the growth becomes more rapid, the embryo begins to rise above the level of the vesicle ; when, at a certain point of the surface of the embryo, a steady and long-continued growth occurs, the limb ap- pears, gradually lengthens out, and enlarges from a small bud at first to a complete arm or leg. If the departure takes place the other way we have an invagination produced ; thus for every hair and every gland of the intestine there is a separate centre of growth. The reason for the unequal gi-owth is unknown. We have not even an hypothesis to offer as to why one group of cells multiplies or expands faster than another group of apparently similar cells close by in the same germ-layer. It is no real explanation to say that it is the result of heredity, for that leaves us as completely in the dark as ever as to the physiological factors at work in the developing in- dividual. The conception that the development of an animal depends funda- mentally upon the unequal expansion and consequent foldings and * By this limitation we exclude the mesenchyma, but not the mesothelimn. DIFFERENTIATION. 163 bendings of the germ-layers was first suggested by the researches of C. F. Wolff on the development of the intestine, and was more clearly recognized by Pander, who definitely asserted that the forma- tion of the embryo is affected by foldings of the germ-layers. In re- cent times His has studied the problem very intently, and in his memoir on the chick, 68.1, discussed it minutely. In this memoir is to be found most of what little we know concerning embryological mechanics. The Classification of Glands. — For a long time it has been customary to divide glands into tubular and acinous. W. Flem- m^ing, in an admirable article, 88. 1, has shown that this classification. as currently applied is untenable, and he proposes in its stead another, the basis of which is the branching of the glands ; he makes three primary divisions: Single glands (Einzeldrusen), which are unbranched; Single branching glands (verastelte Einzeldriisen) , with a single duct and the secretory portion branched ; Compound glands (zusammengesetzte Driisen), with both the ducts and the secretory portions branched. Under the first head he includes the follicles of the ovary, under the last the seminiferous tubules ; but the so-called sexual glands are not, properly speaking, glands at all, since their products arise as differentiations of the cells, not as secretions ; it can, I think, only perpetuate confusion to class them with the true glands. So, too, with regard to the principal organs of excretion — the lungs and the kidnej's ; the former can certainly not be regarded as a gland, since it produces no secretion, for the water and gases given off by the respiratory organs are not produced by the pulmonary epithelium. The kidneys have more claim to be classed with the glands, since their excretion is the direct product of the epithelium of the renal tubules ; the ureter represents the duct and the secretory portions (collecting tubules) branching, thus bring- ing them under the second of Flemming's headings. It seems to me more convenient to give the kidneys a place apart. Under the head of compound glands Flemming ranks the liver, but inasmuch as the gland cavities (gall-capillaries) of the liver form an anastomosing system of canals, it is better to put the liver in a class by itself, es- pecially as its development is unlike that of any other gland. For the sake of completeness we may add also the unicellular glands, such as are found in the lower vertebrates and in many invertebrates ; these constitute a group by themselves, distinct from the multicellular glands. The latter may be divided into four sub-groups : Simple, Branching, Compound, Anastomosing. _ A simple _ gland is one consisting of a single unbranched epithelial tube, ending blindly and opening upon the epithelial surface from which the gland has been developed; a simple gland maybe tubular, that is, a canal of alp- proximately even diameter; or alveolar, that is, with the blind end somewhat dilated ; or vesicular, that is, with the opening small, but the rest of the gland distended like a cyst. Even in the simple glands we usually find the portion of the epithelial tube near the orifice acting simply as a duct, while the deeper part alone performs the secretory office, or acts as the gland proper. The differentiation of the duct is to be regarded, generally speaking, as the earliest and most primitive specialization of a gland. A branching gland is a 164 THE GERM-LAYERS. simple gland with the addition of branching of the secretory portion proper ; under this head also we have tubular and alveolar glands. A compound gland is a branching gland with the addition of branch- ing of the duct. An anastomosing gland is a compound gland with the additional feature of the branches of the secretory portion united together so as to form a network. If we apply this classification to the glands of man, the result may be presented in a tabular form, as follows : GLANDS, A. Unicellular. (FouQd in ichthyopsida and invertebrata) . B. Multicellular ; 1. Simple glands. a. Tubular. 1. Liebei'kilhn's follicles. 3. Peptic glands. 3. Sweat glands. b. Alveolar. Small sebaceous glands. c. Vesicular. (Sub-epidermal glands, amphibia). 2. Branching glands. a. Tubular.* 1. Pyloric glands. 3. Brunner's glands. 3. Mucous glands. 4. Uterine gland. b. Alveolar. 1. Large sebaceous glands. 3. Meibomian glands. 3. Compound glands. a. Tubular 1. Salivary glands. 3 Pancreas. 3. Tear glands. 4. Cowper's glands. 5. Prostate glands. b. Alveolar, f Milk glands. 4. Anastomosing glands. Liver. This classification cannot be regarded as final, since it is based solely on the general shape of the epithelial invagination forming the glands. We may expect in its stead a better classification, based on other and more essential characteristics. The defects of the above arrangement are serious, as is strikingly illustrated by the unnatural separation of large and small sebaceous glands. The basis of classi- fication ought to be the phylogeny of the glands. Histological Differentiation. — The genesis of the tissues de- pends upon — 1, the multiplication of cells; 2d, the specialization of cells ; 3d, the development of intercellular substance. The first of the factors will be discussed in a later chapter. The second and third are to be considered here. The first tissue to appear is the epithelium of the ectoderm and * If the kidneys be con.sidered as glands they would come under this head, as branching tubular glands. i" If we consider the lung as a gland and the bronchi as ducts, the lung would come under this head as a compound alveolar gland. DIFFERENTIATION. 165 entoderm; the second form of tissue is the mesenchyma, for the mesothejial portion of the mesoderm is also epithelium. Histological differentiation, therefore, begins with epithelium and mesenchyma; these two primitive tissues we must consider separately. A. Epithelium. — In invertebrates the ectoderm and entoderm as soon as they become cellular consist each of a single row of polyhedral cells, which in the most primitive type are of equal height. The cells when viewed from the surface are always irregular in outline, usually five and six-sided, sometimes seven-sided or more, but prob- ably never four-sided, except occasionally isolated cells, which as- sume that outline. When the ceUs are not modified by the pres- ence of yolk, the round or nearly round nucleus lies in the centre of each cell. In every epithelial cell three axes may be distinguished, two parallel, with one perpendicular to the surface of the layer, of which the cell forms a part. In the primitive epithelium the three axis are approximately equal in length, hence the tissue is said to be composed of " cubical" (cuboidal) cells. There is very little substance between the cells, and it always remains relatively insignificant in epithelium in marked contrast to its development in the mesenchyma. In probably all vertebrates the ectoderm and entoderm during seg- mentation are both several-layered, but after the close of segmenta- tion they soon become each single-layered, as we have seen. The significance of this modification of the course of development is un- known. The further differentiation of the epithelial germ-layers depends on — 1, the formation of folds, already discussed, p. 161; 2, changes in the proportion of the cellular axes ; 3, structural changes in the cells; 4, arrangement of the cells in several strata. Concerning the latter factors a few words are necessary. The horizontal axis usually remain approximately equal in length, while the perpendicu- lar axis varies independently and to a much greater extent. That epithelial cells are primitively equiaxial may be accepted as an axiom. Yet in vertebrates there are marked departures from this type during very early stages. From the cuboidal type arise the principal modi- fications known as the " cylinder" epithelium and the " pavement" epithelium — names which are unfortunate. As regards the struct- ural differentiation, we must distinguish between the specializa- tion of single cells and that of groups of cells. The former is presumably the primitive form, since it predominates in coelen- terates; the later has been evolved, we must assume, by the grouping of specialized cells ; but in the development of a vertebrate we see al- ways a cluster of cells gradually differentiated from their fellows, and never the cells first specialized and then collected by migration or otherwise. Speaking generally we may say that the higher we ascend the animal scale the less specialization do we find of isolated cells, and the more of groups of cells. This noteworthy fact will, I think, be ultimately found to possess an important significance at present hidden from us. The development of additional strata, which is especially characteristic of the vertebrate ectoderm, is de- scribed in the chapter on the epidermis. B. Mesenchyma. — The first histological differentiation of the mesenchyma in vertebrates is the separation of a certain number of 160 THE GERM-LAYERS. cells from all attachment to their fellows ; these cells are capable of changing their site, and during further development they increase in number and variety. The first of these cells to appear are the blood-cells of the so-called blood islands. For all mesodermic cells not mechanically united to others, but capable of change of site, I have assumed that the primitive type was a cell capable of indepen- dent amoeboid movements, and have proposed for them (Minot, S3, 207) , the collective name of Mesamcehoids—as. a term at once appro- priate and corresponding to a natural class of tissues. The mesa- moeboids, then, I regard as a primitive form of the cells of the mesoderm, thus implying that When amoeboid cells are found in the higher metazoa we are dealing with those free mesodermic elements which have been least modified in the course of development. Ac- cording to this view the wander cells and white corpuscles in verte- brates represent one of the earliest tissues of the mesoderm. As already pointed out, the essential feature of the mesenchyma is that its cells lie somewhat apart and are connected together by protoplas- matic processes running from cell to cell ; the space between the cells is filled with a homogeneous, structureless, transparent substance, which is at first perhaps merely a serous fluid, and which is known as the basal substance (Grundsubstanz) or matrix. The mesenchy- mal matrix is the seat of numerous modifications, varying according to the special tissue formed out of the mesenchyma ; each modifica- tion of the matrix is associated with the corresponding specific change of the cells. III. History of the Theory of the Germ-Layers. The fundamental facts of the construction of the vertebrate body out of distinct layers of cells are collectively designated as the theory of the germ-layers. The theory is as important as the cell theory for the comprehension of the morphology of animals. The establish- ment of it is due principally to Carl Ernst von Baer, although it was first suggested half a century earlier by C. F. Wolff, and more clearly developed by Pander, from whom Von Baer drew his im- mediate inspiration. Since Von Baer's time numerous investigators have contributed to our knowledge of the germ-layers. If we leave out of consideration the introduction of the cell doctrine, which had a profound influence on embryology, as upon every department of biology, we may distinguish three principal steps in the. acquisition of our present notions concerning the germ-layers ; the first step was the recognition by Huxley that the coelenterates are built up of two layers, and the suggestion that these two layers are homologous with the germ-layers of the higher animals ; the second step was the formu- lation of the gastrula theory by Kowalewsky, and the third step was the discovery by His that the middle germ-layer comprises two dis- tinct groups of tissues. C. F. Wolff was the first investigator to recognize the embryonic germ-layers, which he did in the course of his study of the develop- ment of the digestive canal of the chick. His article was published in Latin in the "Commentaries of St. Petersburg Acad.," XII., XIII., 1768-1769, and shows that he suspected the far-reaching HISTORY OF THE THEORY OF THE GERM-LAYERS. 167 significance of the observations which taught him that the intestine is evolved out of a leaf-like sheet in the embryo. Wolff's article se- cured very little notice from his contemporaries, nor was it until it was translated into German by the elder Meckel, and published at Halle, in 1813, that its extraordinary merit became recognized. The translation seems to have awakened the interest of DoUinger, a pro- fessor at Wiirzburg in the early part of this century, who, though little known by his own works, has nevertheless become distinguished through his pupils, foremost among whom are Pander and Von Baer. The former in his dissertation (Wurzburg, 1817) gives a history of the metamorphosis of the hen's ovum during the first five days of incubation, and shortly after published his chief work (" Beitrage zur Bntwickelungsgeschichte des Hiihnchens im Eie," Wurzburg, 1817), the beautiful plates of which were prepared by his friend, D' Alton. Pander distinguished in the blastoderm at first a single layer, das Schleimblatt, external to which, after the twelfth hour, appears the serous layer, which is thinner and more transparent, and finally, at the end of the first day, a third layer, the GefdssschicM, between the mucous and the serous layers. Pander appears not to have con- tinued his embryological researches, but to have left that to his friend and fellow-student. Von Baer, who began his own studies in 1819, and continued them with some interruptions for ten years, extending them gradually to other vertebrates. In Von Baer's work we have the most profound, exhaustive, and original contri- bution to embryology, which has ever been made, and it is un- questionably one of the greatest achievements in the history of science. It ought to be read and pondered upon by every embryolo- gist. The work itself was entitled " Ueber Bntwickelungsgeschichte der Thiere, Beobachtung und Refiexion. " Never again have observa- tion and thought been so successfully combined in embryological work. The first part of Von Baer's " Bntwickelungsgeschichte" appeared in 1838, the second part in 1837. The second part was, however, incomplete and appeared with the announcement of the publishers, stating that they had begun to print the work in 1829, and after waiting five years for manuscript had carried the printing to the 315th page, and finally, after three years more waiting, pub- lished the incomplete second part. In 1888 the missing termination of Von Baer's work ^as published by Stieda. It seems that Von Baer had kept it back in the hope of filling up some gaps; not suc- ceeding in this he waited too long, and after the incomplete work had been issued, Von Baer seems to have lost his interest and to have laid aside his manuscript for the remainder of his long life. Von Baer worked out, almost as fully as was possible at this time, the genesis of all the principal organs from the germ-layers, instinct- ively getting at the truth as only a great genius could have done. Von Baer recognized the somatopleure, which he called animales Blatt, and splanchnopleure, which he called vegetatives Blatt, and further (as each of these Blatter consists of two layers) the animales Blatt had a Hautschicht (ectoderm) and a Fleischschicht (mesoderm), while the vegetatives Blatt had its Schleimschicht (entoderm) and GefdssschicM (mesoderm). With this generaliza- tion, and with the detail of development which he added, Von Baer 168 THE GERM-LAYERS. created modern embryology. It was not until after the cell doctrine was announced in 1838 by Schwann that any important progress was made; C. B. Reichert, 40.1, 43.1, added something to our knowledge, but the value of his work is greatly diminished by the imperfections of his observations, and still more by his errors of interpretation. Perhaps his greatest importance was in his influ- ence upon Remak, whose masterly investigations upon ' the differen- tiation of the uniform embryonic cells into the tissues of the adult at once converted embryology into a science closely allied to histology ; to Remak we owe also the recognition of the mesoderm as a unit, he having discovered that Von Baer's Fleischschicht and Oefdss- schicht are really parts of the same layer. There followed next a series of minor investigations by sundry authors, which, though not very numerous, nevertheless by their gradual accumulation afforded much knowledge. It is not until 1868, when His published his monograph on the chick, that anything fundamentally new was added to our notion of the germ -layers ; in that work His draws the distinction between the archiblast and parablast, see p. 153. From another side progress was being made by gathering materials by the comparative study of the germ-layers throughout the animal kingdom ; here Huxley led the way by discovering the two layers which compose the body of ccelenterates — a discovery which he announced in 1849, adding at the same time the fortunate suggestion that the two layers are homologous with the two primary germ- layers of vertebrates. Four years later (1853) Allman proposed for the two layers of ccelenterates the terms ectoderm and entoderm, which have since come into general use, not only for these layers, but for the corresponding germ-layers throughout the animal kingdom. Beginning about 1845 we have a series of researches on the embry- ology of invertebrates, especially of marine forms. The leader in these studies was Johannes Miiller, whose memoirs are classic and were published for the most part by the Berlin Academy, 1846-1854. He had numerous followers, among whom Alexander Agassiz and Metschnikoff may be mentioned. The naturalist, to whose work in this field we owe most as far as the development of the theory of the germ-layers is concerned, is Anton Kowalewsky, who, by a long series of well-known investigations accumulated a vast amount of evidence in favor of the homology of the gei^n-layers throughout the animal kingdom. Kowalewsky's investigations culminated in the theory that the planula, or, as it is now called, the gastrula, is the primitive embryonic type; he is the originator of the gastrula theory, an account of which has already been given, p. 112. Ernst Haeckel's two essays, 74.2, 75.1, contain, as already stated, exceed- ingly little that is reaUy original and valuable. Lankester's two essays, 73.1, 77.1, are more scientific, and are also noteworthy from having furnished us with a considerable number of terms, which have since become current in embryology. Lankester's essays are further remarkable for containing the first enunciation of the coelom theory. It will be remembered that Von Baer conceived the body cavity to be bounded by two distinct layers, the Fleischschicht and Gefdssschicht; Remak showed that the coelom is bounded by one layer only, the mesoderm; Huxley, 75.1, p. 54, attempted to make HISTOKY OF THE THEORY OF THE GERM-LAYERS. 169 clear the morphology of the body cavity by distinguishing three types thereof — 1, the enterocoele or body cavity, arising as a diverticulum of the alimentary canal, such as was then shown to be the case in the echinoderms and Sagitta ; 2, schizocosle, formed by simple splitting of the mesodei-m ; 3, epicoele, formed by invagina- tion of the outer wall of the body like the atrial chamber of Tunicata. Huxley suggests, p. 56, that the ccelom of vertebrates might be an epicoele. Lankester, 77.1, maintained the opposite view, that the vertebrate ccelom is an enterocoele; for the subsequent history of Lankester's theory, especially as modified by the Hertwigs, 81.1, see Chapter VL, p, 155. PART III. THE EMBRYO. CHAPTER VIII. THE MEDULLARY GROOVE, NOTOCHORD, CANALS. AND NEURBNTERIC In all vertebrates there occur two primary axial structures in very early embryonic stages : one is the medullary canal, derived from the ectoderm; the other is the notochord, derived from ^^ the entoderm : as soon as these two anlages have sK'S appeared the mesoderm disappears from the median if ^ line, and the previously continuous sheet of meso- p|" derm becomes divided into two wings. Connected j^b o with the early history of the medullary canal and gSg notochord are the temporary passages known as ^11 the neurenteric canals. For these reasons these 5-S.h three subjects are best treated together. |tij| I. The Medullary Groove. |= I. The Medullary Plate.— By this name we ^f designate the central axial portion of the ectoderm, 5.^ which early becomes distinguished by its greater 'g | thickness from the remaining portions of the layer o $ and which gives rise later to the nervous system. g:* The ectoderm of the mammalian embryonic shield "" | and of the sauropsidan embryonic area has at first, ^'^ it will be remembered, a considerable thickness, for ifg it consists of cuboidal or low cylinder epithelial fe;" cells. The stage which follows next after the for- -" 3 mation of the primitive axis is characterized by the i m gradual thinning out of the ectoderm over the peri- |,a pheral portions of the shield or area, while in the g § neighborhood of the axial line the full diameter of - 1 the outer germ-layer is not merely retained, but is I'S' actually increased. For a time there is a gradual '^ g passage between the thicker and thinner parts, but |- as development progresses the demarcation rapidly j? ^ becomes sharper, Fig. 95, Md. Soon after its for- ?. g mation, the interval varying according to the spe- ^w cies, the medullary plate increases its thickness 'a | everywhere except along the median line, thus be- o |^ coming double ; the thin median part often shows £?S: a slight groove which is known as the dorsal fur- ?■ s row (Muckenfiirche) . * This furrow does not extend -^ | clear to the cephalic end of the plate, because there 1 1 the lateral thicker bands are continuous with one ^"^ another, the front end of the plate being rounded and clearly limited. * Biickenfurche is also used as a synonym of medullary groove. 174 THE EMBRYO. The medullary plate appears only in the region of the head-process in amniota, and as the process grows backward at the expense of the primitive streak the medullary plate follows, hence it is unequally developed throughout its longitudinal extent, being always more advanced headward and less advanced tailward; hence it is that while it is developing its posterior extremity is always vague and fades out into the undifferentiated ectoderm. So great is this inequality in mammals that we find the front end of the plate trans- formed into the medullary groove before the hind end is differentiated. The stage of development in which there is a well-marked primi- tive streak and in front of it a medullary plate overlying the head- process occurs in the rabbit at the beginning of the eighth day . At its hind end the plate extends so as to partly cover the primitive streak, while in front its edges already rise slightly, so that it constitutes a minute shallow trough. For figures of a similar stage, age unknown, in the mole, see W: Heape, 83. 1, Figs. 13 and 14. In older writers we find figures representing the medullary plate (or groove) and the prim- itive streak as one structure, and the dorsal furrow in the middle of the plate as the continuation of the primitive groove. To illustrate this er- ror I present a copy, Fig. 96, of one of Bischoff's figures of the rabbit's ovum, in which no dis- tinction is made between the two grooves, although in reality the dorsal groove stops in front of the primitive groove, the anterior end of which is often bent to one side. In the Sauropsida the medullary plate is very similar to that of mammals. In both birds and lizards it can be seen that not the whole of the axial band of thicker ectoderm, but only the parts nearest the median line, share in the actual formation of the medul- lary groove. The differentiation is begun as in mammals by the thinning out of the ectoderm in the peripheral regions, until it becomes a thin pavement epithelium, while about the axis the cells become elongated vertically; pyramidal cells, with the apex external, alternating with those with the apex internal, thus producing a peculiar appearance on sections and causing the nuclei to form two layers ; the single cells are, of course, irregular in shape. In birds and reptiles, as in mammals, the medullary plate overlies the head- process and becomes well marked off in front, while it is still being differentiated posteriorly, compare Fig. 97. The Medullary Groove. — Almost or quite as soon as the medullary plate is formed, its edge becomes elevated in front and on each side ; hence it forms an open trough, known as the medul- FiG. 96. —Blastoderm of Babbit's Ovum; after Bischoff. The dorsal and primitive gi'ooves are rep- resented as a single continuous line. THE MEDULLARY GROOVE. 175 lary groove, Fig. 97, Md.gr. During this process the medullary ectoderm increases considerably in thickness, and at the same time the nuclei multiply and lie irregularly scattered at varying heights. The ectoderm alongside the medullary plate or groove thins out still farther. Inasmuch as the development is most rapid in the ii^c Fig. 97. — Chicken Embryo with Seven Primitive Segments (Minot Coll. , Embryo A J, sections 311, 212, 178, compare Figs. 81 and 147). A, Section through one of the segments; B, section posterior to the segments ; C, section just in front of the primitive streak. Md. gr^ medullary groove ; ncfe, notochord ; Ec, ectoderm ; mes, mesoderm ; JJn, entoderm. X about 230 diams. head end of the embryo, there comes a stage in which there is a well- marked medullary groove in front, a medullary plate behind that, and a primitive streak at the hind end of the embryo ; but when the streak has disappeared the medullary groove is found to extend the entire length of the embryo. There is then a stage in which, 176 THE EMBRYO. by means of a series of transverse sections, Fig.- 97, of the embryo, we may study the successive steps in the development of the medullary groove. This stage is found in the rabbit at nine days ; in the chick at thirty to forty hours of normal incubation. The medullary groove gradually deepens, its sides rising higher and higher and arching more and more toward one another until the edges meet and coalesce, thus changing the groove into a tube. The process is illustrated by the series of sections through a chicken embryo with seven segments shown in Fig. 97. In some mammals the medullary groove becomes well developed. Fig. 98, before the medullary plate is clearly marked off by the thinning out of the ectoderm alongside of it; the groove is also much larger. Fig. 98, in proportion to the size of the embryo, Fig. 99, than is the case in the large ova of birds and reptiles. The anterior end of the groove is wide open and ex- panded on each side; this lateral spreading is the anlage of the optic diverticulum, Fig. 99, op, and is transformed later into the optic vesicle, which is an essential component of the future eye. A sec- tion through the optic grooves of a mole embryo a trifle older than Heape's stage, F, Fig. 99, is shown in Fig. 100. The medullary plate is thickened anc shows a median lesser, and two lateral greater depressions ; the former, Md, is the medullary Fig. 98. — Part of a Transverse Section of a Yoxing Mole Embryo. After Heape. JWd, Medullary groove ; Ec^ ecto- derm: Mes, mesoderm: Ent, entoderm. — Surface View of a Young Mole Embryo (stage F, 196 mm.). After Heape. op, Optic diver- ticula; Jar, medullary ridge or edge of medul- lary groove ; Jlfd, medul- lary groove widely open. Fig. 100. — Tranverse Section of a Mole Embryo (Heape's stage F). Md, Medullary groove proper; op, optic gi-oove; Ec, ectoderm; Mes, mesoderm; En, entoderm ; nch, notochord. groove proper ; the latter, op, do not participate in the brain forma- tion, but in that of the eye ; at the edge of the optic anlage the plate passes abruptly into the much thinner entoderm. For some distance behind the optic anlage the edges of the medullary groove are almost in contact, Fig. 99, but farther back the grove is again wide open; this widely open part is known as the sinus rhomboidalis, which is not to be confused with the sinus rhomboidalis of the neck, for the term is also applied to the cavity of the embryonic fourth ventricle of the brain ; the sinus here described belongs to the future lumbar region. The swelling in the floor at the hind end of the sinus is caused by the mesoblast of the front end of the primitive streak. THE MEDULLARY GKOOVE. i I On either side of a rabbit or opossum embryo, in a stage a little more advanced than in Fig. 99, just behind the open anterior end of the canal, there extends a longitudinal ridge corresponding to the anlage of one of the two tubes which will eventually form the heart, see Chapter XI. The lateral heart anlage of the opossum is shown in section. Fig. 95, Ht. In a mole embryo, a little older than Fig. 99, the hinder portion of the medullary canal is much the same as before ; anteriorly, how- ever, development has progressed and the edges of the medullary folds have come together and partially fused at the anterior end of the embryo, owing to the more rapid growth of the sides than of the floor of the canal as pointed out above. At the extreme end, how- ever, a pore is left. At this stage, therefore, the neural canal is still open to the exterior, both anteriorly and posteriorly. The optic grooves are now closed, and have given rise to the optic vesicles ; these are shown as two bud-like vesicles projecting outward and back- ward and slightly downward from the front end of the neural tube ; behind them the swelling of the fore-brain is discernible, while still farther backward and at the edge of the body of the embryo the two tubes of the heart are indicated. The. folding off of the embryo from the yolk-sac has at this stage made some progress, and, indeed, the whole of the head of the embryo now projects freely above the blastodermic vesicle. In the next stage (H, embryo 3.2 mm.) of the mole the edges of the medullary plate have met and united, making the medullary groove in front into a canal, but the sinus rhomboidalis is still open, though beginning to close. The closure of the groove begins in the cervical region and spreads forward and more slowly backward ; where the closure takes place last in front is known as the neuroporus; the position of the neuroporus is presumably the same in all amniota if not in all vertebrates. Van Wijhe, 84. 1 , finds that in the duck the connection with the ectoderm is retained in front longest in the re- gion of the first cerebral vesicle, and not in that of the mid-brain, so that it has nothing to do, as some have suggested, with the development of the pineal gland (epiphysis) . This connec- tion represents the final anterior clos- ure; Van Wijhe speculates that it was an opening in the ancestors of vertebrates and terms it the anterior neuroporus. The medullary groove of Amphi- bia has been more fully studied than that of any other class . The most com- plete history is that given by Alex. Goette for Bombinator, 75.1, 158-176; see also Scott and Osborn, 79.1, S. F. Clarke, 80.1, Rusconi, Moquin-Tandon, 76.1, Ecker's " Icones," Taf. XXIII., and others. In all Amphibia the medullary plate is very wide, indeed. Fig. 101, being broadest in front. 12 Fig. 101. —Early Stage ot Amblystoma Punctatum. m.g, Median groove; m.p, medullary plate; Tnfc, cephalic portion ot medullary fold. After S. F. Clar Uarke. 178 THE EMBRYO. Fib. 102. —Part of a Transverse Section of an Axolotl Embryo. After Bellonci. Mes, Mesoderm ; coe, coelom; Md, medullary groove ; Ec, ectoderm; .Bnr, entodermal or archenteric cavity ; C/i, noto- chord ; Yollc, yolk cells. The deutoplasm is indi- cated in a few of the cells. Its margin is thrown up into a slight but broad ridge ; when the plate closes to form a canal the surfaces of the marginal ridges grow together and the surface of the plates within the ridges becomes the surface of the central canal. In all Amphibia the central dorsal groove, mg, is very distinct. As the ectoderm of the amphib- ian ovum very early becomes distinctly two-layered, it results that in the medullary plate the two layers can be recognized from the start ; the outer layer (Goette's Deckschicht, Bal- four's epidermic stratum) of course lines the medullary cav- ity and alone forms the epithe- lium of the cen tral canal. When the groove closes the lumen of the canal is nearly circular in section, but it soon changes into a narrow vertical slit similar to the lumen of the amniote canal. The round cavity is due to the way in which the medullary plates curl up, as shown in Fig. 102. As pointed out by Alex. Goette, 75.1, 160, the lateral portion of the meduUary plate arises by delamination, a peculiarity which has an important bearing, I think, on the discussion of the evolution of the medullary canal, see p. 179. Finally, in the Amphibia, the medullary plate extends to the middle of the blastopore, and, it is maintained by some writers, extends beyond it, so as to completely surround it. This point is recurred to in connection with the history of the anus. The Medullary Canal.— The medullary canal, as stated, arises by the closure of the groove. The canal closes in the cervical region first, hence it has at one time two free openings ; as the closure pro- gresses the anterior region is completed, while the sinus rhomboidalis is still open ; moreover, we see that the anterior end achieves consider- able differentiation before the posterior end of the canal is closed. Of the entire length of the primitive canal about one-half is the anlage of the brain, while the other half forms the spinal cord. In the development of the brain the transverse expansion of the canal is most conspicuous, while in the development of the spinal cord the elongation of the canal predominates. The dilatation of the brain part begins very early, and comprises at first a general dilatation of the whole anlage, and, second, special and greater dilatation of three regions ; the three dilatations are known as the three primary cere- bral vesicles (Hirnhlasen) , and are designated as /ore-&ram {Vor- derhirn, prosencephalon), mid-brain {Mittelhirn, mesencephalon), and hind-brain {Hinterhirn, metecephalon) , respectively. The first vesicle is much the widest, and appears in mammals and probably in all vertebrates very early ; in mammals it shows itself plainly in the medullary groove as already noted. When the groove closes the canal THE MEDULLARY GROOVE. 179 is of course attached to the ectoderm, Fig. 9-3, but this connection is soon severed, and the medullary, or, as it also called, neural canal, becomes an independent structure lying inside the external ectoderm of the embryo, and surrounded by mesodermic cells, which subse- quently grow in between the canal and the ectoderm so that the canal comes to lie farther and farther away from the surface, Fig. 103. The structure of the medullary canal in early stages has been as yet but imperfectly studied. The wall increases steadily in thick- ness, except in certain parts of the brain. Where it thickens its nu- clei multiply and form several irregular layers ; the cell bodies around the nuclei are small and connected by numerous processes, so as to produce a protoplasmic network ; the protoplasm and nuclei next the lumen early assume the character of epithelial cells, so that the cavity of the meduUary canal is lined by a distinct epithelial layer ; this layer corresponds to the outside layer of the ectoderm (epidermis) ; in some parts — as, for instance, the dorsal wall of the fourth ventricle — the single epithelial layers constitute the entire medullary wall. The Fie. 103.— Transverse Section of a Rabbit Embryo of Eight Days and Two Hours. Md, Medul- lary canal ; Seg, primitive segment ; Cho^ chorion ; Am, amnion : Sam, somatopleure ; Coe, coe- lom; Spl, splancnuopleurej JEhit-, entoderm; C/i, notochord; Ao, aorta. non-epithelial cells of the canal become, as described in Chapter XXVII., ganglion cells. The nuclei of the medullary canal wall are oval, their long axis being more or less nearly perpendicular to the surface of the canal ; each nucleus contains one or several nucleoli. The canal is primarily oval in section, but its lumen is a narrow fissure. Fig. 103, hence the walls are thickest at the sides, and thinner dorsally and ventrally ; this peculiarity dominates to a marked de- gree the subsequent development of the brain and spinal cord. Evolution of the Medullary Canal.* — ^Under this head we have to consider, first, what is the primitive vertebrate type of the central nervous system ; second, what genetic relation existed between the vertebrate and the invertebrate type. The opinion generally accepted by embryologists is that the typical vertebrate canal is formed by the closure of the medullary groove. * Originally published in the American Naturalist, Nov. , 1889. 180 THE EMBRYO. This view is advocated by Balfour, and has been so thoroughly accepted by Adam Sedgwick, that he has made it the basis of a spec- ulation, 83.1, on the original function of the canal; he supposes that it was open behind and excretory ; the cilia which are found in the central canal of the spinal cord originally served to produce the excre- tory current. Van Wijhe, 84. 1 , has advanced independently almost the same hypothesis. Both of these speculations overlook the serious difficulty of assuming that the canal is primitive, while in the lowest vertebrates it is clearly a secondary modification. In Petromyzon, Lepidosteus, and Teleosfs, the medullary plate, instead of becoming the fl-oor of an external groove, forms a solid keel-like projection toward the ventral surface. This keel subsequently becomes sepa- rated from the superficial layers of the ectoderm, and afterward a central canal is developed in it. In the ganoids, which approach the elasmobranchs in structure, there is, as shown by Salensky, 81.1, a medullary groove of peculiar form, which suggests a transition from the solid keel to the open groove ; again in Amphibia there is evidence that the delamination is still preserved to a slight extent in that group. These considerations lead me to the hypothesis that the ner- vous system of vertebrates was primitively a solid axial thickening of the ectoderm, and within the class of ganoids became modified into a groove perhaps simply by more precocious development of the central canal ; the groove type has been kept in elasmobranchs, amphibians, and amniota. Balfour, "Comp. Embryol.," II., 303, thus defends the opposite view : " It seems almost certain that the formation of thp central nervous system from a solid keel-like thickening of thi epidermis is a derived and secondary mode, and that the folding of the medullary plate into a canal is primitive. Apart from its greater frequency the latter mode of formation of the central nervous system is shown to be the primitive type by the fact that it offers a simple explanation of the presence of the central canal of the nervous sys- tem ; while the existence of such a canal cannot easily be explained on the assumption that the central nervous system was originally de- veloped as a keel-like thickening of the epiblast" (epiblast-ectoderm) . It is not possible at present to decide positively between rhe two views, but the view which I am inclined to adopt is further justified by the development of the central nervous system in Annelids, which is formed by the co- alescence of a pair of linear cords ; these cords arise each side of a ciliated longitudinal furrow, first as a single row of ectodermal cells, subse- as several rows; while still united to the ex- ternal ectoderm they extend toward one another inside the ciliated cells of the furrow, and unite in a single nervous band. The origin of the annelidan nerve cord is illustrated by Fie. 104, which renre- m.e Fig. 104.— Part ot a Transverse Section of an Embryo of Lumbricus Trapezoides. After Kleinenberg. En, Entoderm ; Ec^ ectoderm ; n n, anlagres of the nervous system; F, cells ot the ciliated hand separating the two parts of the nervous system; c, c\ parts of the Quentlv coelom; -ni.c, mesodermal cords. ^ , ._ ■/ THE NOTOCHORD. 181 sents a transverse section of the embryo of an earthworm, at a stage during which the cells, n n, that are to form the nerve cords are still part of the superficial ectoderm, Ec, though their future separa- tion is already indicated. In leeches and arthropods the develop- ment is very similar. In all these cases the bands split off from the ectoderm. It appears then that in the nearest* invertebrate allies of the vertebrates, the nervous system develops as a thickening along the inner surface of the ectoderm, and delaminates from that layer. It seems to me very natural to suppose, therefore, that the strikingly similar process in the lowest vertebrates is the primitive one, and that the canalization of the medullary plate was evolved within the vertebrate series. I have assumed that the ventral nerve cords of annelids are homolo- gous with the medullary canal, a view that is now generally accepted by embryologists. Balfour (Works, I., 393, and "Comp. Embryol." II., 311) has suggested a more complicated relation in his hypothesis that the lateral nerve trunks, which are known in many of the lower worms (e. g., nemerteans, have fused on the ventral side in annelids, but on the dorsal side of the body in the vermian ancestors of verte- brates. In favor of this ingenious surmise no evidence has since been found. Hubrecht denies the homology of the annelidan nerve chain and the vertebrate medulla; he considers, 87.1, 620-634, that the more primitive condition is represented by certain nemertean worms, which, beside two main lateral nerves, have a small longi- tudinal median nerve; the lateral nerves gave rise to the nervR chain of annelids by their fusion, the median nerve to the medulli of the ancestors of vertebrates. As no intermediate forms, either adult types or embryonic stages, are known to represent any phases of this double metamorphosis, I cannot admit that Hubrecht's bold specidation invalidates what seems to me the well-established ho- mology between annelids and vertebrates. The remarkable hypothesis of W. H. Gaskell, 90. 1, that the med- ullary canal is homologous with, and derived from, the entoder- mal canal of Crustacea, seems to me unwarrantable. II. The Notochoed. As the notochord is a purely embryonic structure, I present its complete history here. The notochord (chorda dorsalis, Wirhelsaite) is a rod of peculiar tissue, constituting the primitive axial skeleton of vertebrates. It begins immediately behind the pituitary body (hypophysis) and extends to the caudal extremity. It occurs as a permanent structure in the lower type, and as a temporary one in the embryos of amphibia and amniota, including man. Comparative embryology has shown that it is a greatly modified epithelial band which arises in the median dorsal line of the entoderm, being in position and mode of development analogous to the ectodermal medullary canal, or primi- tive tubular nervous system . Numerous embryological articles contain observations on the noto- chord. The following references may assist students. The best * With, of course, the possible exception of Amphioxus. 83 THE EMBRYO. eneral discussion is by Balfour, in his " Comparative Embryology;" le best observations on its origin in mammals is by Heape, 83. 1, jr descriptions of the chorda canal see Lieberkiihn, 82.1, 84.1; iarius, 88.1, and VanBeneden, 88.3; for its histology, W. Miiller, l.S; for its histogenesis, A. Goette, 75.1, 349-361; for its anterior natomical relations see Mihalkowics, 74.1, 75.1, Froriep, 82.1, Labl-Eiickhard, 80.1, and Romiti, 86. 1 ; for its atrophy in mammals 36 Leboucq, 80. 1; for its evolution see Ehlers, 85. 1. Origin from Notochordal Canal. — ^The notochord appears ery early in the course of development ; its differentiation from the ledian dorsal wall of the notochordal canal begins at the time when tie medullary groove is not fully marked out posteriorly, and is owhere closed. The notochordal anlage can be first detected just 1 front of the primitive streak as an axial band of cells, which at rst is not well marked off from the mesoderm ; this band forms the ledian dorsal wall of the blastoporic canal in all vertebrates in which bat canal has been identified. The differentiation of the notochordal eUs begins usually at the anterior end of the canal and. progresses ackward, as the blastopore moves backward during concrescence. 'he differentiation varies as to the time of its beginning ; it may egin in the unconcresced embryonic rim, as in Scyllium, or much iter, as in Lacerta. As the medullary groove (or keel) deepens, it pushes down toward be mid-gut until it comes into actual contact with the notochordal pithelial band, thus dividing the mesoderm into two lateral masses, 'ig. 97, one on each side; this also leads to the temporary transverse tretching of the notochord. Lieberkiihn, 82.1, 84. 1 , has directed attention to a special peculiar- ty in the early development of the notochord in mammals. The noto- hordal canal is formed throughout its length and then breaks through t various points to fuse with the yolk cavity, so that it may be escribed as a tube running along the median line, and having an [•regular series of openings on its ventral side. The canal is lined y epithelium, which is thickened on the dorsal side to form the nlage of the notochord. In transverse section the chorda appears ccording to the level of the section to constitute part of a furrow or , canal (compare also Heape, I. c, p. 441, Fig. 40, 41). Lieberkiihn alls thi^ canal mesoblastic, and Kolliker follows him in so doing, lut this opinion seems to me based upon misconception. Indeed, I Giacomini's researches, 88. 1, show that the canal terminates in he rabbit in a blastopore, and Van Beneden, 88.3, has emphasized he fact that the canal helps to form the definitive archenteron. Lfter the notochordal canal has fused with the yolk cavity, the noto- hordal anlage is, of course, incorporated in the entoderm of the nain archenteric cavity, and appears as the median dorsal portion if the entoderm. It early acquires a sharp demarcation and becomes onsiderably thicker. Fig. 105, than the adjoining entoderm, and orms a distinct though shallow groove. Separation from the Entoderm. — The notochordal band sep- .rates off and the entoderm proper closes across under it, so that he notochordal band lies between the entoderm and the floor of the nedullarv groove (or later canal) as shown in Fies. 106. 103. and 97 THE NOTOCHOED. 183 Fig. 105. — Transverse Section of a Mole Embryo, Stage H. After Heape. am, Amnion- Md^ medullary groove; My^ myotome; Coe^ eoelom or body cavity; JSn, entoderm; ncft, notochord; ao, aorta; vf.a, vitelline artery; Som^ somatic mesoderm; Spl^ splanclinic mesoderm. A. This separation does not take place at the anterior extremity of the chorda until somewhat later, so that for a considerable period its front end remains fused with the walls of the archenteron, Fig. 106. Selenka, 87. 1, observed that this front end of the notochord becomes dilated in the opossum M"^- and hollow; the hollow end subsequently forms an irregular sac opening into the anterior end of the intestinal cavity ; Selenka names the sac the Gaumenstasche j it opens behind the partition which closes the mouth and is entirely distinct from the hypophysal eva- gination. Further inves- tigations led to the dis- covery of traces of a similar canalization of the front end of the notochord in other verte- brates (Selenka, 88. 1). The peculiarity shows conclusively that the connection of the notochord with the hypophysis is secondary, and that, therefore, Hubrecht's hypothesis as to the evolution of the noto- chord is untenable. The separation from the entoderm is effected, at least in mammals, by the entoderm proper showing itself under the notochord toward the median line, and when the cells from one side meet those of the other they unite with them and form a continuous sheet of entoderm below the notochord cells. It is probable that the separation begins in all vertebrates, as it has been shown to do in several cases, be- fore the whole length of the noto- chord is formed, and progresses headward ; see, for example, Mcintosh and Prince's ac- count of the pro- cess in teleosts, 90.1, 743. So, also, in Triton al- pestris, Bambeke, 80.2, 90, found that the separation of the notochord from the entodei-m takes place earlier than in the Urodela, and progresses from in back forward. After the separation pigment granules appear in the central portion of the chorda, an important observation, since certain writers have held, I believe erroneously, that the presence of pigment proves that the notochord must be derived from the ectoderm, which is usually pigmented in amphibian ova. Fis. 106. —Longitudinal Section of the Head End of a Mole Embryo, Stage H. After Heape. Ec, Ectoderm ; En, entoderm ; pro. am. , pro- amnion; m.6., mid-brain; /. 6. , fore-brain; Ent, entodermic cavity ; ht, heart; mes, mesoderm; nch, notochord. 184 THE EMBRYO. After its separation the chorda is a narrow band of cells starting anteriorly from the wall of the alimentary tract and running back- ward to the blastopore. So long as the blastoporic canal is open, the chorda terminates in the entodermic epithelium lining the canal. For a certain period the chorda continues growing tailward by accretions of cells from the walls of the blastoporic passage, and after the canal is permanently obliterated the chorda may still con- tinue its lengthening by acquisitions at its caudal end of additional cells from the primitive streak ; such cells may, however, properly be regarded as coming from the entodermic lining of the blastopore. We can, then, distinguish two portions of the notochord ; the first arising from the epithelium of the notochordal canal, the second pre- sumably from the cellular wall of the obliterated blastopore. Braun and others have sought to attribute essential importance to these differences, but it seems to me improperly. It is more reasonable to say that the chorda arises in the amniota, as in the lower forms, di- rectly from the entoderm, but presents certain secondary modifications in its development. After it is once formed as a band of cells the notochord passes through various changes of form, but ultimately becomes a cylin- drical rod with tapering extremities. It attains considerable size in the embryos of most vertebrates, but in those of placental mammals it is always small, particularly so in the mole (Heape, 83. 1). It is probable that in mammals the notochord, when first separated from the entoderm, is a broad fiat band, as if compressed between the medul- lary canal and entoderm (c/. KQUiker, "Entwickelungsgesch.," Figs. 194 to 197, and also Heape, 86.2, PI. XIII. , Figs. 36 to 43) . The band then draws together, diminishing its transverse and increasing its vertical diameter, until it has acquired a rounded form ;* finally its outline becomes circular in cross-section. This series of changes begins near the anterior end of the chorda, and progresses both for- ward and backward. The nuclei of the notochord tend to gather at first in the central portion of the chorda, but in later stages (shark embryos with fifty and sixty myotomes) the nuclei are found situated peripherally, Rabl, 89.2, 349. The mesoderm early grows in between the entoderm and the notochord, which, however, for a considerable time remains close to the medullary tube. Later the mesoderm pen- etrates also between the notochord and medulla. The layer of meso- dermic cells immediately around the notochord, which are of the well-known mesenchymal tj'pe, forms a special sheath, which at first comprises only a single layer of cells, at least in batrachia (Goette, 75.1, 357, Fig. 187). This is the commencement of the so-called outer chorda sheath ; it subsequently becomes much thicker. In the lower types it is sometimes an important axial structure ; but in most cases it is replaced by cartilage, and in all the amniota the cartilage is replaced by the osseous vertebrae, the intervertebral ligaments, etc. The formation of the vertebral column involves the disappearance of the notochord as described below. Notochord of Teleosts. — The medullary keel or great neural axial thickening of teleosts extends to the entoderm ; the cells at the bottom of this keel next the entoderm give rise to the chorda. There * A splendid description of tlie selachian notochord at this staere is sriven by C. Rabl. 89.3, 213, 214. THE NOTOCHORD. 185 being, it is said, no open blastoporic canal in the bony fishes, we can only trace the cells back into the undifferentiated mass of cells with which ectoderm and entoderm also fuse, and which lies at the hind «nd of the embryo. According to the most generally received opin- ion, the cells of the notochord arise from the entoderm, and their fusion with the ectoderm of the medullary keel is temporary only. The teleostean chorda separates first from the mesoderm, second from the entoderm, and third from the ectoderm. The development in Lepidosteus is similar. The modifications we here encounter will probably be traced back to the general vertebrate type. For discus- sion of the subject and citations of earlier authorities, see Mcintosh and Prince, 90.1, 740-745. Shape and. Relations to Other Parts. —As soon as the head- bend (first cerebral flexure) appears. Fig. 107, the notochord becomes correspondingly bent, and its bend lies close to Eathke's pocket, Fig. 107, hy. From Selenka's Guamen- tasche there now runs upward and for- ward a short limb of the notochord, which subsequentlj' atrophies. This limb may remain regular or it may grow and become somewhat irregular before it atrophies; after it is gone the chorda has a netv or secondary ante- rior extremity, which Romiti, 86.1, finds in the chick embryo at the end of the fourth and during the fifth day of incubation to be united with an irregu- lar solid cord of cells which grows out from the epithelium of the hypophysis. The cord soon disappears. Its signifi- cance is quite unknown. Romiti sug- gests that it may produce a strain re- sulting in the pulling out of the hj-- pophysal evagination. This notion seems to me untenable, since the hypophysal invagination begins before there is any union with the notochord. The cranial portion of the notochord has not only the bend shown in Fig. 107, but also follows the other curves of the head ; it takes a sinuous course besides within the base of the cranium ; finally, in the region corresponding to the middle third of the spheno-occipital cartilage, it makes a great dip ventralward. The sheath of the notochord in the cranial region is converted into the spheno-occipital cartilage ; at the dip just mentioned, however, the notochord lies entirely below the cartilage close against the wall of the pharynx (Froriep, 82.1, Romiti, 86.1). Writers before Froriep had represented the chorda as having disappeared at the bottom of the dip. The anterior termination of the notochord has been carefully stud- ied by Prenant, 91.2, 203, who finds that it has (pig and rabbit) no connection with the hypophysis, but may have a secondary tem- porary connection with the entoderm just behind Seesel's pocket, and that the part of the notochord nearest the hypophysis very early ■degenerates, leaving the notochord to terminate above Seesel's pocket; Fig. 107.— Eabbit Embryo o£ 6 mm. ; Median Longitudinal Section of the Head. The connection between the mouth ilf, and pharynx eni, is just es- tablished; ncK notochord; Ab, hind- brain; m6, mid-brain; fb, fore-brain; pro.oni, proamnion; Ay, hypophysis cereiiii ; ht-, heart. After Mihalkovics. 186 THE EMBKYO. according to this view the so-called prae-chordal region primitively contains the notochord. The secondary anterior termination of the notochord is close to the infundibulum (and future pituitary body) , and it is customary for subsequent stages to divide the head and skull into a proe-pituitary and a post-pituitary region; the latter region alone contains the notochord, after very early stages. Histogenesis. — After the notochord has been formed as a rod of cells, its cells undergo a process of histological differentiation, unique in vertebrates. The cells at first become greatly compressed in the line of length of the chorda, and hence appear quite thin in longitu- dinal sections. Fig. 108, hardly greater in diameter than their own nuclei. The flattened cells are next converted into a highly charac- teristic reticulum by vacuolization. Thus, in the chick, by the third daj' some of the central cells become vacuolated, while the peripheral cells are still normal ; at first, as in the frog, there seems to be only one large vacuole in each cell, Fig. 108, B. Around the vacuole is Fig. 108.— Longitudinal Sections of the Notochord of Bombinator. After Goette. A, betor& the appearance of the vacuoles: B, after the appearance of the vacuoles; tich^ notochord; B}a^ entoderm. The cells, as is usual in amphibian embryos, are changed with yolk-granules. a peripheral layer of granular protoplasm, in which the nucleus lies imbedded, while the vacuoles themselves are filled with a perfectly clear and transparent material, which is supposed to be fluid in its natural condition. During the fourth day (chick) all the cells be- come vacuolated, with the exception of a single layer of flattened cells at the periphery. (In the anura, it is said, there is no distinct peripheral layer of protoplasmatic cells.) The vacuoles go on en- larging until by the sixth day they have so much increased at the expense of the protoplasm that only a very thin layer of the latter is left at the circumference of the cell, at one part of which, where there is generally more protoplasm than elsewhere, the remains of a nucleus may generally be detected. Thus the notochord becomes transformed into a spongy reticulum, the meshes of which correspond to the vacuoles of the cells and the septa to the remains of their cell walls (Foster and Balfour) . As Goette has pointed out, the process is accompanied by an expansion of the cells which is the main factor in the widening and lengthening of the notochord, which goes on pari passu with the growth of the surrounding tissue. The histogenetic process is stated to be essentially similar in mam- mals (W. Miiller, 71.2,337, 338). There is the central layer of vacuolated cells and the peripheral layer of protoplasmatic cells. THE NOTOCHORD. 187 Fig. lUU. — Degenerating Notochord Tissue,from the Central Portion of the Intervertebral Disk of After Leboucq. The latter are, however, ultimately converted into vacuolated cells. The cell walls are perforate, having fine pores that correspond prob- ably to intercellular bridges of protoplasm. The inner chorda sheath appears early, and is to be regarded as an anhistic basement membrane secreted by the notochordal cells. Disappearance. — The disappearance of the notochord in man commences with the second month of foetal life. The first step is an alteration of the characteristic histological structure, accompanied by shrinking of the tissues, so that a clear space appears around it. The inner chorda sheath is lost. The cell walls disappear, the tissue becomes granular, and breaks up into multinucleate, irregularly recticulate masses. Fig. 109, which are gradually resorbed (Leboucq, 80.1). In mammals the resorption progresses more rapidly in the cores of the verte- brae than in the intervertebral spaces, and again more rapidly at the ends than in the centre of each vertebra ; hence the chorda ^''co'^v'^Eiiibr^' persists a little longer in the cen- tre of the vertebra, and considerably longer in the intervertebral spaces ; in these last the final remnants of the chorda may be de- tected in man even after birth. The cavity between the vertebral cartilages is a new structure and is not the space left by the noto- chord, as has been sometimes asserted. It appears, however, that the resorption of the chorda may leave a small space, which becomes included in the intervertebral cavity. A peculiar feature is the fre- quent persistence of calcified cartilage immediately around the notochord in ossifying vertebrae. Morphology. — The notochord was for a long time supposed to be exclusively characteristic of vertebrates. It is now known to exist in Amphioxus, which is not a true vertebrate, and in the Tuni- cata. Morphologists have long believed that it must have some homologue among the organs of invertebrates. The development of the notochord in the lower vertebrates indicates very plainly what must have been the general character of such an homologous inver- tebrate organ. In certain fishes and amphibia the notochord has been asserted to arise as a furrow along the median dorsal line of the entoderm ; the furrow deepens and then closes over to form a rod separate from the entodermic canal proper. The notochordal rod retains for a time its anterior and posterior connections with the entoderm. It is usually regarded as morphologically a solid canal, a view very open to doubt. Ultimately the ends become detached, and so arises the solid isolated chorda. In the higher vertebrates the course of development is similar, although several of the primi- tive features in the formation of the chorda are obscured. Ehlers, 85. 1, has pointed out that in various invertebrates there is a similar canal, the " ISTebendarm " of German writers, which is derived from the entoderm and connected anteriorly and posteriorly with the en- todermal cavity. It is a very plausible suggestion, which homolo- 188 THE EMBRYO. gizes the vertebrate notochord with the invertebrate " Nebendarm." Hubrecht has sought to homologize the notochord with the proboscis of nemertean worms. There is not a single fact which seems to me to justify, even remotely, this attempt at guess-work phylogeny, nor can I find any resemblance of the notochord with the structure in Balanoglossus with which Bateson has sought to homologize it. III. Neurenteric Canals. The term neurenteric canal is used to designate an open commu- nication between the archenteric cavity and the medullary canal. Such communications are found only in early stages ; they always pass through the anterior end of the primitive streak and lead, therefore, into the posterior end of the medullary canal or groove ; they are present only during a short period. Much confusion has existed in regard to these canals, of which as many as three have been distinguished by M. Braun, 82.3, while several writers recog- nize two. The true Neurenteric Canal is probably the blastoporic canal proper, and is to be identified by the notochord terminating in its wall. As stated in Part II. of this chapter, the " chorda-canal" of mammals is the " blastoporic" canal, and therefore also in- cludes the neurenteric canal. As previously described, the blastopore is the opening of the notochordal canal at the ante- rior end of the primitive streak. The neural ridges or medullary folds extend around and behind or across the blastopore, which, therefore, opens into the poste- rior extremity of the medullary groove. If now the canal is open at the period of develop- ment when the medullary groove is deep or has already closed over, making the medul- lary canal, then there is a direct communication between the en- todermal canal on the one hand and the spinal canal on the other. We owe to Balfour the identification of this canal as the blastopore. It may with propriety be termed the true neurenteric canal, or the canal of Kowalewsky from its discoverer. Kowalewsky first found it in Amphioxus, and subsequently demonstrated its occurrence in vari- ous fish types. This canal is well known in Elasmobranchs and Sauropsida under Fig. 110. —Longitudinal Section of aFrog's Ovum shortly after the Closiire of the Medullary Groove. The anal canal, 6Z, is only partially cut, but was found open in neighboring sections, Md^ medul- lary canal; bl, anal portions of blastopore; Vit. yolk forming the floor of the entodermic cavity Ent; ne. , neurenteric canal. After Durham. NEURENTERIC CANALS. 189 the name of the blastoporic canal. It has recently been shown to be present in Petromyzon by A. Goette, 90. 1 . In teleosts it is rudi- mentary, the passage being only imperfectly indicated (see Mcintosh and Prince, 90. 1, 734-736). In the Amphibia its relations are more clearly understood than in any other type. According to H. E. Durham, 86.1, it can be well seen in longitudinal sections of early stages of the frog. Fig. 110, as a short canal, ne, opening widely into the entodermic cavity. This canal has also been described in Bom- binator by A. Goette, 75.1, and in Triton and Rana by F. Schanz, 87. 1. Schanz was the first to clearly discriminate between the anal or false blastoporic and the neurenteric or true blastoporic canal. In birds the neurenteric canal was first described by Gasser, 79.1, in goose embryos, and since then has been found by Braun in several other birds, though as an open passage it appears to be usually oblit- FiG. 1 11 . —Transverse Section of an Embryo Paroquet (Melopsittacus) to show the Anterior or true Neurenteric Canal. Ec, Ectoderm; My, myotome; Md, medullary canal; Ch, notochord pierced by the short neurenteric canal, ne; Ent, entoderm; mes, mesoderm. After Max Braun. erated, as in the chick. Fig. Ill represents a transverse section which passes through the Gasserian or neurenteric canal of the paroquet. Braun has maintained that in various birds there are two neuren- teric canals recognizable, which lie near together. Braun states that in the duck and MotaciUa the two canals a,re separated both_ in the times and position of their occurrence, and that in the Australian paroquet they are present simultaneously. D. Schwarz, 89.1, criti- cises Braun's observations and concludes that there is really no sec- ond canal. I am inclined to accept this conclusion. In mammals the open blastoporic canal has been seen by very few observers ; it has been carefully studied in the rabbit by C. Giaco- mini, 88. 1, who shows that the front portion persists for a time as the chorda-canal, while the hind portion, running through the prim- itive streak, corresponds to the neurenteric canal and is obliterated quite early. The Anal Canal is also sometimes called the neurenteric canal, and seems to have been especially the subject of misconception. Its full history is given later. Its morphological relations are prob- ably correctly indicated by the observations of Fr. Schanz, 87.1, which have been confirmed by T. H. Morgan, 90. S, by Robinson and 190 THE EMBRYO. Assheton, 91.1, and others. To explain these relations we may start from the stage in amphibian ova, in which the anus of Rusconi is almost closed over ; the true blastopore lies at its front edge. As the anus of Rusconi contracts its aperture appears more and more as a mere enlargement of true blastopore, and it is at this stage com- monly spoken of as the blastopore ; to preserve the distinction we may name this opening the secondary blastopore. Alice Johnson, 84. 1, had shown that the permanent anus is derived in Triton from this sscondary blastopore. H. E. Durham, 86.1, observed that there are two passages in the frog at a little later stage. Fig. 110. Schanz, I. c, found that the medullary ridges meet at their hind ends across the secondary blastopore and divide it into two openings, the anterior, the true blastoporic or neurenteric, and the posterior, the anal opening. In many Amphibia the anal canal is often tem- porarily closed by the tissue growing across it ; in later stages the ectoderm forms a slight invagination to develop the anus proper, the partition between it and the archenteron breaking through. The partition is called the anal plate. The relations of the anal canal in Saiiropsida are not yet well as- certained. It is represented by the anal plate, consisting only of ectoderm and entoderm. Although this anal plate (Afterhaut) has not been actually proved to be homologous with the tissue which tem- porarily closes the anal canal in amphibians, yet it is hardly possible to question the correctness of the homology, for it separates the anal invagination from the archenteron and subsequently ruptures. The anal membrane recurs in mammals, and if it represents the anal canal in one case it does also in the other. C Giacomini, 88.1, 287, 288, states that in rabbit embryos with several myotomes the anal plate is grown into a short cord of cells, in which there appears a temporary lumen — this lumen he calls the anal canal. After the canal has disappeared the anal membrane is again found to consist of two epithelial plates, the rupture of which forms the true anal perforation. Braun's Third Canal. — The third canal, which was first de- scribed by Braun, 82.3, is said to occur in older embryos. D. Schwarz, 89.1,211, denies its existence altogether. The "End- darm" of Gasser and KoUiker becomes the " Schwanzdarm" (post- anal gut, Balfour) of older embryos, which soon becomes divided, at least in birds, into a dilated terminal portion and a narrower neck communicating with the intestine proper. The posterior section then subdivides, and its narrow end-segment lengthens out and unites with the spinal cord. This passage we may designate as Braun's canal. It is not improbable that it is homologous with the amnio- allantoic canal of Gasser, 82.2, which Rauber, 83.2, has nicknamed Cochin-China canal, after a breed of hens in which it seems most constant. In the one case we may suppose the canal to open after, in the other before, the closure of the posterior end of the medullary groove. If the homology is correct it may be further said that the canal is identical with Kupffer's myelo-allantoidean canal; it can- not be brought into relation with the development of the allantois, as believed by Kupffer, 82.2, 83.1, as the allantois and end-darm are both formed before the canal appears. NEURENTERIC CANALS. 191 Significance of the Neurenteric Canal. — As to the mor- phology and physiology of the canal we know almost nothing. The suggestion of Sedgwick and Van Wijhe, that it is the excretory opening of the tubular nervous system, has already been noticed, p. 179 ; it will sufEce to recaU here that no valid evidence in favor of their hypotheses has been found yet. There is no adult form known in which the neurenteric canal persists ; were there such an animal we might hope to discover the function of the canal by observation. Morphologically the neurenteric canal, so far as I can judge from present evidence, is part of the persistent blastoporic canal, which is included in the medullary groove, and by the closure of the groove becomes shut off from the exterior. Why the secondary blastopore (prostoma) should be divided into the two openings, the neurenteric and anal, we do not know. It seems not impossible that a persistent neurenteric canal may occur as an excessively rare anomaly in the adult. CHAPTER IX. THE PRIMITIVE DIVISIONS OF THE CCELOM; ORIGIN OF THE MESENCHYMA. In all true vertebrates the ccBlom presents the peculiarity of con- sisting of an upper or dorsal segmented portion and a lower or ven- tral continuous unsegmented portion. The segmented ccelom consists of a series of discrete separate cavities, each of which communicates with the ventral coelom. Now, in annelids (and their arthropodous descendants) the ccelom consists only of separate paired cavities, so that the mesothelium is divided into distinct parts, each inclosing a space ; each division is known as a mesomere or mesoblastic somite. Hence we have the morphological question, how has the completely segmented ccelom of annelids become transformed, as we must assume it has, into the partially segmented coelom of vertebrates? The answer is probably given correctly by Hatschek's investigation of the changes of the mesothelium in Amphioxus, 88. 1. In Amphioxus the entire meso- derm becomes segmented ; the ventral cavities of the segments sub- sequently fuse, while the dorsal parts remain distinct. In the lower vertebrates the segmented coelom appears first and the unsegmented portion later ; whether the latter is temporarily segmented remains for future investigation to determine. In amniota the unsegmented portion of the coelom appears first, as described in Chapter VI. This must be regarded as a secondary modification, probably con- nected with the evolution of the amnion ; as explained in Chapter XV, the development of the amnion depends upon a precocious and exaggerated development of part of the coelom. With these general notions in mind, we can better appreciate the early history of the vertebrate coelom. We consider — 1, the primi- tive segments ; 2, the unsegmented coelom ; 3, division of the primi- tive segments; 4, the differentiation of the myotome; 5, origin of the mesenchyma ; 6, comparison with Amphioxus. The Primitive Segments.* — A segment consists of a pair of cavities symmetrically placed and bounded by mesothelium. The segments are permanent in many invertebrates, but they are greatly modified in all adult vertebrates, and so much modified in the amni- ota, that they can be said properly to exist only during embryonic stages, although they determine a large part of the adult structure. I have selected the term primitive segments as unlikely to lead to confusion, but numerous other names have been proposed ; the one most generally in use is protovertebra {Urivirbel}, which was intro- duced long ago under the erroneous notion that the segments were the direct precursors of the vertebrae, which they are not, properly speaking. The term protovertebrse is, however, more often used in * On the segments o£ the head, see p. 800. THE PRIMITIVE DIVISIONS OF THE COELOM. 193 a more restricted sense, viz. , for that part of the primitive segment which is called the myotome in the following pages. Other terms are mesohlastic somites, mesomeres, metameres, Ursegmente. The primitive segments appear very early ; the first pair can be recognized in the chick at twenty to twenty -two hours, in the rabbit at the beginning of the eighth day, or even earlier ; in both cases the medullary groove is still nowhere closed and the primitive streak is still present. In the anamniota the first segments appear at about the same early stage. In the amniote embryo just before the first segment appears the mesoderm forms a continuous sheet; surface views show that it forms two wings, being divided by the median down-growth of the medullary groove as stated, p. 149. The meso- derm on each side is considerably thicker alongside the axial line than farther away from it ; the distinction is well marked and ena- bles us to distinguish two zones, namely, the thicker segmental zone near the axis, and the thinner but much wider lateral or parietal zone; the segmental zone is the Stammzone of German writers, the Wirhelplatte of Remak, or vertebral plate of Balfour. The first noticeable indica- tion of the formation of the primitive segment is a loos- FiG. 118.— Chicken Embryo with one Segment. Ap, Area pelluoida; vAf, anterior crescent; vKf^ head fold; Ch, notochord ; Pz, parietal zone ; Stz, Stammzone. From Kolliker. (Com- pare text. ) Fig. 113. —Area Vasculosa and Embryo with Eight Seg- ments of a Hen's Egg. F7t, Fore-brain ; vAf^ anterior amniotic fold ; Pd, fovea cardiaca ; omr, anterior limit of open medullary groove; i2/, rhomboidal sinus of medullary groove ; Pr, primitive streak ; Bmj, margin of medullary groove; Ap^ area pellucida; Ao^ area opaca; XJw, first segment; Hft, hmd-brain ; Mh, mid- brain. From Kolliker. ening of the cells in the segmental zone along a narrow trans- verse line; in the chick this occurs about 0.14 mm. in front of the 13 194 THE EMBRYO. primitive streak, at a time when only a short stretch of the head- end of the medullary groove is formed. Very soon there appears a second transverse loosening of the cells and cleavage of the meso- dermic segmental zone takes place. According to A. Goette, 75.1, 303, the cleavage begins in teleosts and the chick and probably in other vertebrates with a small depression on the ectodermal side, and this depression gradually deepens to a cleft, which divides the segmental plate completely. The disposition of the fissures is such that they include on each side of the axis a cuboidal block of mesoderm, and this block with its fellow on the op- posite side constitutes the first prim- itive segment. Fig. 112. The site of Ent Stl) Coe Fig. 114. — Eabbit Embryo with Eight Seg- ments. Tjft., Fore-brain; ab, optic vesicles: a, heart; a/, amniotic fold; m/i, mid-brain; A/i, hind-brain; pz, parietai zone; stz, seg- mental zone; ap, area pellucida; rf, edge of open medullary groove; uw, primitive seg- ment; 1)0, venous end of heart; ft, heart; oft, pericardial cavity; vd, fovea cardiaca. From KOUilcer. Fig. 115. —Transverse Section of a Prlstiu- rus Embryo with Fourteen Segments, through the Centre of the Fourth Segment, md. Me- dullary groove; Ec, ectoderm; msth, meso- thelium; Coe, cavity of segment; ISnt, ento- derm ; nch, notochord. After C. Eabl. the first segment corresponds* to the posterior occipital region ; the second segment, at least in the chick, is formed immediately in front of the first; these two segments, according to Chiarugi, 90.1, 339, are the third and fourth occipital segments, and together with the first and second segments — subsequently found in front of them — abort in all amniota during early embryonic life, as was discovered by Froriep. According to Julia B. Piatt, 89.1, 179, the first divi- sion formed in the chick is that between the third and fourth occipital segments, and two segments are subsequently produced in front of this division, while seven are forming behind it. A chick with one segment is shown in Fig. 113. The medullary groove, Rf, is short and broad; the anterior end of the embryo, vKf, is already rising THE PRIMITIVE DIVISIONS OF THE CCELOM. 195 above the yolk; the parietal, Pz, and segmental zones, 8tz, are dis- tinct _ even in the region of the primitive streak, along which the primitive groove, Pr, is well marked ; the segment is well advanced and another has begun to form in front of it. A chicken embryo with eight segments is shown in Fig. 113, and a rabbit embryo also with eight segments, in Fig. 114; comparison shows many important differences between the two embryos. The examination of transverse sections shows that the primitive segments in all anamniote vertebrates are hollow and bound by mesothelium on all sides. The relations can be understood readily in elasmobranchs. Fig. 115 is from a Pristiurus embryo and shows the cavity of the segment very clearly; the embryo is much more separated from the yolk than is the case with amniote embryos at a corresponding stage, consequently the lateral or parietal zone of mesoderm lies nearly vertical, instead of resting horizontally, as it does upon the yolk of amniota ; in the parietal zone there is as yet no cavity (ccelom) ; the ventral or unsegmented coelom arises later. It is probable that the segmental cavities spread down into the pari- etal zone, and that their ventral (^. e. lower or so-called parietal) ends fuse together and form one large main body cavity. This probability, as C. Rabl has said, 89.2, rests upon the analogy with the ascertained process in Amphioxus, and upon the fact that the segmental cavities appear first, and expand outward or away from the axis. Whether, however, they do actually give rise to the main coelom by their partial lateral or ventral fusion or not, there are no observations at present to decide. Of the development of the segments in the primitive vertebrates (marsipobranchs, ganoids, and amphibians) , there is not much known, though there are many scattered observations recorded. There ap- pears, however, to be a distinct thickened zone of mesoblast on each side of the axis, and from this zone the segments are developed as pairs of cuboidal blocks of mesothelium ; the central cavity of the segment is very small ; its mesothelium is thick. The main coelom is at first a fissure farther away from the axis, and it has not yet been shown that there is from the start a communication between the segmental and the main coelom, although the mesoblast is con- tinuous. In Petromyzon, if I understand Goette aright, the meso- derm is at first solid; Goette states, 90.1, 48, that in the cervical region the main body cavity appears first, but in the rump the prim- itive segments acquire their cavities first ; that is, while the mesoderm of the parietal zone is still solid. This is important as foreshadow- ing the precocious development of the cervical coelom (cavity of the amnio-cardial vesicles) in amniota. In sections the primitive seg- ments in Petromyzon and Amphibia are triangular, filling out the space between the medullary canal and the adjacent ectoderm and entoderm on each side. In Bombinator, A. Goette, 75.1,203, and Petromyzon, A. Goette, 90.1, the first segment appears- near the middle of the embryo, and new segments are added in front to make the cervical region and a much larger number progressively back- ward to form the rump of the adult. In Sauropsida transverse sections of the primitive segments, when they are first formed, show no cavity, nor does any appear until con- 196 THE EMBRYO. siderably later. The development begins with the differentiation of the segmental zone (Remak's Urivirbelplatte, Balfour's vertebral plate), which is accomplished by the thickening of the mesoderm near the axis of the embryo. The process is intimately associated with the upward movement of the medullary plate to form the med- ullary groove, His, 68.1, 81, and Goette, 75.1; the space between the ectoderm and entoderm is enlarged by this movement, and is always nearly filled by the mesoderm ; consequently the segmental zone appears triangular in cross-sections, the base of the triangle being against the wall of the medullary groove, its two sides against the ectoderm and entoderm respectively, and its apex merging into the lateral mesoderm, which is very much thinner than the segmental plate. The changes just described show a very exact adjustment of the growth of the mesoderm to changes in the outer germ-layers. Such adjustments occur throughout all embryological developments, and are, I think, due to methods of growth rather than to simple mechanical conditions. His has attributed, 68.1, 81, 93, special in- fluence to the attachments of the mesoderm to the other layers, and the consequent strain upon the segmental plate as the medulla rises ; but the enlargement of the plate depends upon the multiplication of the cells, and we cannot assume that the strain causes cell prolifera- tion. At most, one might say that the strain determines the shape of the segmental plate. But is it not more natural to assume that the cells of the mesoderm simply spread out until they fill the avail- able space? The first sign of the mesomeres is the assumption by the cells, that are to form them, of a more distinctly epithelial arrangement, the cells radiating in ,, JgPSSil&9.-fte>. Som " '' My »jM^-.»u.^(^»(^t>_. all directions, but there remains in the centre a core or nucleus {Urwirbel- kern) of cells. Fig. 116, C, which have small bodies with anastomosing proto- plasmatic processes ; it is impossible, at least at present, to state whether this core of looser non- epithelial tissue is or two, but I am Fig. 116. —Transverse Section through a Recently Formed Primi- tive Segment of a Chicle with Eighteen and Twenty Segments. My^ Myotome : W. d. Wolffian duct ; Som, somatic mesoderm ; Coe, cce- lom (splanchnoccele) ; SpU splanchnic mesoderm ; iV, nephrotome or intermediate cell mass ; C, core of myotome. X 227 diams. an ingrowth from all sides or only from one strongly inclined to think that it is probably part of that side of the primitive segment which is next the medullary canal. The cross-sections further show that the mesomere is more complex than is apparent in surface views, in that it consists of a wider tri- angular part. Fig. 110, My, next the medullary groove, and a nar- rower lateral portion, N, next the parietal zone ; it is to the former that the term protovertebra ( Urwirbel) is restricted by most writers, while the latter is termed the intermediate cell mass, as proposed by Balfour ; both are parts of the primitive segment. The appear- THE PRIMITIVE DIVISIONS OF THE CCELOM. 197 ance of an epithelial arrangement of the cells is confined to the "protovertebra" sensu strictu. The square blocks seen in surface views correspond to the protovertebrse only and not to the whole segment. The most exact observations on the primitive segments of mam- mals known to me are those of Heape, 86.2, on the mole, of E. Bonnet, 89. 1, on the sheep, and of J. KoUmann, 91. 1, on the human embryo. The vertebral plate thickens as the medullary plate rises and becomes triangular in cross-section ; the mesodermal cells, which up to this point have been of the anastomosing type, become elon- gated and radiating, and gradually assume an epithelioid character, which becomes most distinct on the ectodermal side ; the ceUs grad- ually withdraw from the centre of the segment, leaving a cavity.* The cells of the segment multiply rapidly, most of the divisions tak- ing place in radial, but some in tangential, planes. The segments have the triangular form already noticed in other classes. The cells have branching prolongations, which extend out to the primary germ-layers, and are especially marked on the ectodermal side. In the sheep the cavities of the first four segments, and of them only (Bonnet, l.c.,^ 50), extend through the lateral portions of the segments and communicate with the main coelom ; these four segments Bonnet assigns to the occipital region. A similar series of communications have been recorded for the chick by S. Dexter, 90. 1 . The Ventral or Unsegmented Coelom. f— This portion of the coelom, which persists in the adult, gives rise to the pericardial, pleural, and abdominal cavities, which are morphologically parts of one continuous cavity, the ventral coelom. Many terms are in use to designate the ventral coelom ; by English embryologists it is usu- ally called the pleuro-peritoneal space or cavity ; or often simply body-cavity {Leibeshohle, cavite somatique) ; by German writers it is sometimes termed lateral coelom, sometimes the Parietalhohle, although the latter term is properly used only for the pleuro-peri- cardial division of it. Hatschek has proposed splanchnoccele, which is adopted in this work. The splanchnoccele appears in all cases in the parietal zone of the mesoblast as a narrow fissure, the method of origin of which has already been described, p. 151. The fissure rapidly widens and ex- tends toward the axis until it ahnost reaches the primitive segments and also spreads out laterally and into the so-called extra-embryonic region of the amniota, but there is for a considerable period a circu- lar area inclosing the region of the embryo like a ring, in which the mesoblast contains no coelom ; this mesodermic ring is known as the vascular area (area vasculosa, Gefdsshof) , and has for its special function the production of the first blood-vessels and blood-corpuscles ; see Chapter X. In later stages the coelom extends into the vascular area. The splanchnoccele is developed earlier, and acquires a greater dis- tention at first in the future cervical region. A. Goette, 90.1, 48-49, states that in Petromyzon it precedes in the region of the heart the appearance of the segmental cavity. In the Amphibia * See Bonnet's figure, I. c. Arch. Anat. und physiol. 1889, Taf . v. Fig. 5. + On tlie spianchnocoele of the liead, see aiso p. 199. 198 THE EMBRYO. the precocity of the cervical ccelom appears also, and it is perhaps true of other anamniota. The development of the main ccelom is still more hastened in all the amniota, being in them intimately associ- ated vsrith the development of the amnion. In the chick this is very well marked, because as probably in all sauropsida the splanchno- coele enlarges so rapidly in the cervical region that, even vs^hile the number of primitive segments is very small, we can recognize a vesicular space in the mesoderm on either side of the head of the embryo ; for these spaces, which are the Parietalhohlen of German embryologists, I propose the name of amnio-cardial vesicles. They are shown in Fig. 117, Ves. Their rapid expansion soon brings Md Fig. 117. — Section of a Chick with about Twenty Segments, B'f, Heart; Coc, coelom; Jlid, medullary tube ; Ph, pbarynx; Som, somatopleure ; Am., amnion; Ves, amnion-cardial vesicles; Cho, chorion ; Spl, splanchnopleure. X about 40 diams. them into contact, and then into fusion with one another under the neck of the embryo. The heart is lodged in this cavity, of which the lateral increase produces the so-called head-fold of the amnion. It is on account of this double destiny that the name amnio-cardial vesicles is proposed. The relations, which are traced out through the later stages in Chapter XV., may be more fully understood from Fig. 117, which is a cross-section of a chick through the heart region at an older stage than we are now considering. The posterior limit of the amnio-cardial coelom is marked by the course of the omphalo- mesaraic veins, which arise later and establish the communication between the area vasculosa and the venous end of the embryonic heart. The topographical relations are described in Chapter XIII., on the germinal area. In mammals the same peculiarity of the precocious dilatation of the amnio-cardial ccelom probably recurs, but has not yet been properly investigated. In the sheep (Bonnet, 89. 1), the amnion appears extraordinarily early, and, as it must be preceded by the formation of the coelom, we find in the sheep a huge ring of splanchnoccele around the embryo while it is still in the primitive streak stage. The splanchnoccele of the body proper — that is, of the region behind the neck and heart — appears after the primitive segments, even in the sheep, in which the extra-embryonic coelom is so very early developed. Moreover, in the body the main coelom expands more slowly than in the neck. The expansion takes place at first only in the part of the mesoderm next the primitive segments. Fig. 92. Already the thick- ening of the segmental plate {Urunrbelplatte), which accompanied the uprising of the medullary plate, has marked out partially the THE PRIMITIVE DIVISIONS OP THE CCELOM. 199 region of the embryo from that of the yolk, and now the distention of the splanchnoccele increases and finally completes the demarcation of the embryonic region from the extra-embryonic. The splanchno- coele extends in all amniota only part way through the mesoderm, until quite late in development, so that at a gradually increasing distance from the embryo there is a layer of mesoderm without any cavity, and the cells of which preserve the mesenchymal type. This undivided mesoderm develops the first blood and blood-vessels. After the first vessels of the area have appeared the splanchnocoele spreads out over them, so that the first vessels lie then below the ccelom — -i. e. , in the splanchnopleure. As the splanchnoccele develops, the mesodermal cells assume grad- ually a more and more distinctly epithelial character, so that the main coelom becomes bounded by mesothelium, as described in Chapter VI. , and the somatic leaf of the mesoderm is differentiated from the splanchnic ; toward the axis of the embryo the two leaves pass into one another, and also at the distal edge of the ccelom the two leaves pass without any distinct limit into the uncleft mesoderm of the area vasculosa. In conclusion, I wish to emphasize the fact that the splanchnoccele (pleuroperitoneal cavity) is almost, if not quite, from the start di- vided into a precociously enlarged cervical portion (amnio-cardial vesicles, Parietalhohle) and a rump portion (abdominal cavity) ; the boundary between the two portions is marked by the omphalo- mesaraic veins, which run from the area vasculosa into the embryo proper at nearly right angles to the embryonic axis. This primitive disposition is of fundamental morphological significance. Ccelom of the Head. — No thorough investigation of the history of the early stages of the mesoderm in the head has yet been made for any vertebrate. Until this is done we cannot hope to understand the morphology of the head, because the progress of research has demonstrated more and more clearly that the head is made up of series of greatly modified segments, but the number and metamor- phoses of the head segments can be determined only by knowing the entire history of the mesoderm. Balfour was the first, 78.3, to demonstrate the existence of the coelom in the head, and to partially work out its subdivisions. The subject was further advanced by A. Milnes Marshall, 81.2, Van Wijhe, 82.1, A. Dohrn, 90.2, and others. Marshall and Van Wijhe's results have been subjected by Gegenbaur, 88. 1, 3-8, to criticism, which seems to me by no means fortunate. Gegenbaur's conclusion, that the number of cephalic segments gives no trustworthy indication of the ancestral history I must entirely dissent from, since I believe that the number of meso- dermic segments in the head of the embryos of the lower vertebrates is the only trustworthy clew to the morphogeny of the head, which we can seek at present. A. Dohrn, 90.1, 3.35, made the discovery of a large number of segments in the head of vertebrate embryo, having observed seven- teen or eighteen in the head of Torpedo marmorata of 3 mm. Killian, 91.1, confirms and rectifies (l. c, p. 103) Dohrn 's observations, and describes seventeen to eighteen segments, Fig. 118, in the head of Elasmobranchs, as follows: Oral zone with two segments; mandibu- 200 THE EMBRYO. lar zone with three; spiracular zone with three, corresponding to the first gill cleft ; hyoid zone with four, in the region of the second gill cleft ; glossopharyngeal zone with two ; occipital zone with four. Killian observed these segments in Balfour's stages F and J, of Torpedo ocellata. In later stages Van Wijhe, 82.1, whose results have been verified, found only nine seg- ments. The number is presuma- bly reduced chiefly by abortion, but partly also by fusion. Van Wijhe's segments are as follows : The first or prse-oral is identical with Balfour's prse-mandibular cavity; and it is identified by Killian with his oral zone ; it is possible that the first segment of the oral zone is identical with the " new" head cavity described by Julia B. Piatt, 91.2; Van Wijhe's first segment is small and acquires its cavity late, being solid after the remaining eight myotomes have developed their cavities; it is connected by a short band of cells across the median line with its fellow of the opposite side; this band subse- quently (in Balfour's stage L) disappears ; the first segment pro- duces four muscles, the rectus superior, internus, and inferior, and the obliquus superior. The second or mandibular segment (Balfour's mandibular cavity) cor- responds with Killian's mandibular zone; its cavity disappears in Balfour's stage O; it produces the muscles of mastication; ac- cording to Killian, it is produced by the fusion of three segments. The third segment seems also to be the product of the fusion of three primitive segments of KiUian's spiracular zone; its cavity has a communication through the hyoid arch with the ventral coelom (pericardial cavity). Tlh.Q fourth segment corresponds in position over the second or hyoid gill cleft with the three segments of Kil- lian's hyoid zone (Dohrn's eleventh to thirteenth segments). The fifth segment corresponds to the two segments of Killian's glosso- pharyngeal zone. Killian's four occipital segments all persist inde- pendently of one another to constitute Van Wijhe's sixth to ninth segments, which I think are to be further identified with the four temporarily present hypoglossal or occipital segments which Froriep has discovered, 86.1, in amniote embryos. Van Wijhe regarded nine as the total maximum number of segments in the vertebrate head, and sought, 89.2, to identify nine corresponding segments in Amphioxus. Fig. 118. —Head of an Embryo of Torpedo Ocel- lata, in Balfour's Stage J. I-X, Anlages of the cepnalic ganglia and nerves; 1-18, cephalic primitive segments; 1-3, first three rvmip seg- ments; o.vl, oral plate; Sp, "spiracular''' cleft, or first gill cleft; Hy, hyoid cleft. The dotted circle below I and II indicates the optic vesicle. After Killian. THE PRIMITIVE DIVISIONS OF THE CGELOM. 201 That a series of coelomatic cavities exist in the head of the am- phibian embryo was, if I am not mistaken, first observed by Scott and Osborn, 79. 1. Houssay, 90. 1, has sought to identify the num- ber of cephalic myotomes in the axolotl. He accepts the idea of the exact correspondence between the branchial pouches and the myo- tomes in segmental order ; and as he maintains that there is a gill pouch, which corresponds to the auditory nerve and aborts during embryonic life, and further regards the nose, hypophysis, and mouth each as representing a separate segment, he finds that there must be at least eleven segments in the head of the axolotl, as follows : 1, nose; 2, hypophysis; 3, mouth; 4, "event;" 5, hyo-mandibular ; 6, hyoid; 7, ear; 8, first branchial; 9, second; 10, third; 11, fourth branchial; for each of these he assumes a separate myotome. He has actually observed, 90. 1, the nine somites corresponding to those described by Van Wijhe (see above), and further claims to have found evidence that the second and third of these are both really double, thus identifying eleven mesomeres, which, he says, 91.1, 58, appear in the following order : In position 1 23456789 10 11 In time 1 2 11 10 6345789 Or in groups 1' 2' 3" 2" 1" 1'" 2'" 3'" 4'" 5"' 6'" Van Bemmelen, 89. 1, 254, in a superb reconstruction of the head of a snake embryo, shows three myotomes belonging to the eyeball, but gives no information concerning them, and represents no other myotomes in the head until the hypoglossal region with its four myotomes is reached. A. Oppel, 90.1, describes the cephalic seg- ments in Anguis embryos; he has recorded the presence of Van Wijhe's first to third and sixth to ninth segments. The splanchnocoele of the head becomes the pericardial cavity of the adult ; its mesothelium, where it covers the heart, gives rise to the cardiac muscle, and it is supposed to extend between the gill pouches to produce the muscles of the branchial arches. Along the level of the branchial pouches the splanchnocoele becomes in part divided, as first shown by Balfour, 78.3, into a series of separate cavities by the outgrowth of the gill pouches and the union of the entoderm of each pouch with the ectoderm. Each of these cavities has an elongated form and communicates on the dorsal side with a myotome, and on the ventral side with the pericardial cavity (Van Wijhe, 82.1, Van Bemmelen, 90.1). We may distinguish, there- fore, the mandibular coelom, the hyoid coelom, and the branchial coelom (one cavity in each gill arch) . The connection of the cavities of the arches with both the myotomes and pericardial cavity is ap- parently lost, but as to the separation there are no definite observa- tions. The actual cavities in the arches are soon obliterated, but their mesothelial walls persist and produce the branchial muscles ; compare Chapter XXI. Division of the Primitive Segments. — The primitive seg- ments very early divide, each into two parts — the myotome (proto- vertebra of authors) next the medullary canal, and the smaller nephrotome {intermediate mass, KoUiker's Mittelplatte) ; next the lateral plates or mesothelium of the splanchnocoele. Fig. 116. The 302 THE EMBRYO. division is evidently indicated as soon as the. primitive segments are formed, the thicker proximal end being destined for the myotome, the thinner distal end for the nephrotome ; the latter originally unites the myotome with the lateral plates, hence its name of " intermediate cell mass" ; as its principal function is to develop the nephridia it may be more conveniently named the nephrotome, as proposed by Ruckert, 88.1. The nephrotome has to separate from the splanchnocoelic meso- thelium (lateral plates) on the one side and the myotome upon the other Unfortunately this double separation has been as yet very inadequately studied, except in the case of elasmobranchs, where the development of the nephridia has been carefully investigated; for details compare Chapter XI. In Bombinator a groove appears on the ectodermal side and gradually deepens until it separates the myotome from the rest of the mesoderm ; this groove does not pass through in the shortest direction, but extends obliquely upward, A. Goette, 75.1,213. The nephrotome loses its connection with the myotome relatively early, but retains, at least in some segments, the connection with the lateral plates for some time longer in most elasmobranchs and amphibians throughout life, but in amniota only during embryonic stages. The exact histological changes by which the nephrotome serves its double connections are still unknown. A. Goette, 90.1, 49, states that in Petromyzon the isolation of the neph- rotome takes place in the front end of the body when the mesoderm has a well-developed coelom, but in the rear part while the meso- derm has no coelom either in the vertebral or lateral plates. C. Rabl, 89.2, has directed especial attention to the fact that in elasmobranchs there is a special outgrowth of the wall of the primi- tive segments on the side nearest the chorda and from the point where the nephrotome joins the myotome, Fig. 122. This out- growth* is the beginning of the mesenchyma, and recurs, of course, segmentally, so that the term sclerotome may be applied to it, but all trace of segmental division is very soon lost, nor does the seg- mental origin of the axial mesenchyma, which is developed from. these outgrowths, determine the subsequent morphological differen- tiation, so far as yet known. Rabl likens this outgrowth to an evagination, and points out that the cavity of the nephrotome pres- ents a slight diverticulum at first, where the outgrowth takes place. He compares this evagination with the evagination at a corresponding point in Amphioxus, which has been described by Hatschek, 88.1, and is said to grow up between the myotome and the medulla; in Amphioxus, however, the cells retain an epithelial character, while in the vertebrate they are mesenchymal ; but as no strict line can be drawn between these two types of tissue, the histological difference cannot be held to invalidate the homology drawn by Rabl. The cavity of the primitive segment varies greatly in the various classes of vertebrates. In the primitive forms, Petromyzon, Am- phibians, etc., the myotomic portion is wedge-shaped, appearing triangular in cross-section, and considerably wider than the cavity of the nephrotome. In elasmobranchs, cf. C. Rabl, 89. S, Taf. X., Figs. 1-6, a similar difference exists at first, but very soon the two * Compare Eabl, I.e., "Morph. Jb. ,"xv., Taf. x. , Fig. 4, sk. THE PRIMITIVE DIVISIONS OF THE CCELOM. 303 walls of the myotome come close together, Fig. 132, obliterating the cavity ; the nephrotomic portion, on the contrary, widens meanwhile. In Lepidosteiis the medullary and entodermal sides of the myotome are represented as several layers of cells thick by Balfour and Parker, 82.1, PL 23, Figs 38, 39, so that the myotome appears partly filled with cells belonging, however, to its inferior wall. We have in this case perhaps a transition to the amniote structure, in which the encroachment of cells is so great that no distinct cavity can be recognized in the myotome. Fig. 116; and since the nephro- tomic cavity appears very late, it results that in the amniota there is no distinct cavity whatsoever in the primitive segments, though there is a cavity later in both the myotome and nephrotome. The primitive aortae lie close below the myotomes on each side, Figs. 119, 132, 161, 105; a glance at any of these will show the reader that the mesoderm derived from the myotome from the very first comes into contact with and soon envelops the medullary tube, Md, the notochord, Ch, and the aorta, Ao, and also reaches over part of the entodermal wall of the archenteron. Shape of the Myotome. — As described above, the myotome, when first formed and even before it is separated from the nephro- tome, appears more or less nearly square in surface views and trian- gular in cross-section. Very soon it enlarges in Amphibia and amniota, so as to appear square in section also. Fig. 119. The cavity in Amphibia is very distinct and the epithelial character of the walls well marked ; but in all amniota, so far as known, the cavity at this stage is still obliterated by the core of cells (Remak's Urwirbelkern) . By the assumption of the cuboidal shape the myotome becomes more sharply marked off from the intermediate mass or nephrotome, and as the lateral or main coelom has been expanding during the same period, there is established a space above the nephrotome and be- tween the myotome and the lateral plates. It is in this space that the primitive longitudinal duct of the urogenital system. Fig. 116, W.d., is situated as soon as developed — a fact which led many writers to attribute the origin of the duct to a differentiation of the interme- diate cell mass. Differentiation of the Myotome. — We can distinguish three steps in the differentiation : 1, production of mesenchyma from the inner wall of the myotome. Fig. 119; "2, production of the true muscle plate. Fig. 130; 3, conversion of the outer wall into mesen- chyma to form the dermal layer, Fig. 121. The production of mesenchyma from the inner wall begins very early, and is marked by a loosening and moving apart of the meso- thelial ' cells until the entire inner wall, at least in amniota, is con- verted into tissue of the mesenchymal type, Fig. 119, mes. Owing to the moving apart of the cells the tissue occupies a large space and fills up the myotomic cavity. While the metamorphosis is going on the cells multiply rapidly. The course of this change of the inner wall has been carefully studied by W. Heape, 86.2, in the mole, by R. Bonnet, 89.1, 45-55, in the sheep, and by Erik Miiller, 88.1, in the chick. Miiller has further demonstrated that the muscular en- velope of the aorta comes from the mesenchyma produced by the inner myotomic wall. In elasmobranchs, according to C. Rabl, 204 THE EMBRYO. 89. S, the greater part of the inner wall of the myotome very early shows the differentiation of muscle fibres, the cells retaining the mesothelial type, Fig. 133, and the mesenchyma is produced only from that part of the inner wall, which is nearest the nephrotome {Mittelplatte of Remak) ; in elasmobranchs, therefore, the mesen- chyma appears more as an outgrowth from one point — a fact which THE PRIMITIVE DIVISIONS OF THE CCELOM. 205 leads Rabl to a significant comparison with Amphioxus, as stated above. In amniota the persistence of the outgrowth is indicated by the fact that the metamorphosis of the mesothelium of the inner wall begins near the nephrotome; it spreads, however, rapidly, so that nearly the entire wall undergoes the transformation. My own obser- vations are incomplete, but they indicate that in amniota the differ- entiation of myotomic muscles invariably follows later. Where the inner waU. joins the outer the cells retain the mesothelial arrange- ment for a very considerable period (see Figs. 119 and 131). The muscle plate proper arises from cells of the inner wall next the myotomic cavity, or we may say — since the cavity is obliterated Md Cu Fig. 130. —Longitudinal Horizontal Section through a Segment of a Rabbit Embryo of Ten and One-half Days. Md, Edge of the medullary wall ; Ec, ectoderm ; I. a, intersegmental artery ; mes, the so-called sclerotome or mesenehyma of the inner wall ; Cu, outer wall of the myotome (anlage of the cutis) ; M, muscle-plate. X 296 diams. —from the cells nearest the outer wall. The cells become elongated parallel with the longitudinal axis of the embryo, Fig. 130, M; the nuclei also elongate in the same direction, thus becoming oval, and as shown in the figure they are, at least in the chick, larger than the nuclei both of the neighboring mesenehyma, 7nes, and of the outer myotomic wall, Cu. The remainder of the inner wall, mes, is the sclerotome of recent German writers; it consists of mesen- chymal cells which are now entirely separated from the parts of the 206 THE EMBRYO. myotome which are still mesothelial. While the muscle-plate is forming the mesenchyma merges with it, but gradually it becomes sharply marked off from the muscle cells. The muscle-plate is con- tinuous at its edge with the outer wall, Cu, and retains the continu- ity for a very long period. The muscle-plate and outer mesothelium now form a single and highly characteristic structure, familiar to all embryologists ; the structure is a double plate, which takes an oblique position in the embryo ; as seen in cross-sections the double plate descends from near the dorsal border of the medullary tube downward and outward toward the somatopleure. The next change is the production of mesenchyma from the outer wall; the cells of the mesothelium move asunder until they come to lie quite far apart, Fig. 121, Cu, forming from the start a much ©e@ff*|fi^.^'-.«®- Fis. 121.— Transverse Section through the Upper Part of a Myotome of a CJhick of about Seventy Hours. Mes, mes', Mesenchyma of inner wall of myotome ; Ep, endothelial layer formed against the surface of the medullary tube; Tnsth^ mesothelial portion of the myotcme; Ec^ ecto- derm ; Ow, mesenchyma (cutis) from outer wall of myotome ; Mu, muscle-plate, the limits of which are much clearer in the preparation than in the engraving, x S96 diams. looser tissue than did the mesenchyma from the inner wall ; but at this stage, Fig. 121, the inner mesenchyma, mes, is spreading around the medullary canal, and as it spreads assumes also a looser texture. The mesothelium, msth, still persists around the four margins of the double plate, apparently as an organ to produce cells to be added on the one hand to the muscle plate proper, Mu, on the other to the cutis (dermal mesenchyma), Cu. In sections the mesothelium usu- ally makes a U-shaped figure, which is highly characteristic of all vertebrate embryos. In the primitive vertebrates as exemplified by Petromyzon (Goette, ORIGIN OF THE MESENCHYMA. 207 90.1, Taf. VI., Figs. 60-63), the flattened myotome consists of two closely appressed epithelial plates with a narrow fissure between them and passing over at their edges into one another ; the upper edge of the myotome is nearly on a level with the dorsal margin of the medulla ; the myotome inclines obliquely outward and downward and has its lower edge on the level of the archenteric cavity ; the outer layer of epithelium is the thinner, while the inner layer is con- siderably thickened ; as the myotome develops farther this difference between the two layers increases. The amphibian myotomes resemble very closely those of Petromj-- zon, but soon come to differ from them by the multiplication of cells of the inner layer (A. Goette, 75.1, 211, Figs. 138-140), which be- comes several cells thick and loses at the same time its distinctly epithelial character in the inner part of the layer, though it retains it in the outer part, there remaining, on the side nearest the entoderm, a single row of cells in epithelial form, so that we have here a con- dition established secondarily which in the amniota exists almost from the start — namely, a core of looser cells filling the myotomic cavity, but belonging to the entodermal side ; it is at this stage that in Bombinator the myotome separates from the remaining mesoderm. In later stages the amphibian myotome gives off from probably all parts of its wall cells to form part of the mesenchyma. while the cells which remain form the definite muscle-plate. Origin of the Mesencliyina. — The first author to trace the origin of the mesenchyma to the primitive mesothelium was Alex- ander Goette, who fully demonstrated the fact in his great work on the " Unke," 75.1. Goette designates the mesenchyma as Bildungs- gewebe, and seems to me to have been the first to fully recognize the morphological significance of the tissue. But his work has not hith- erto received its deserved attention. Scattered through numerous special papers are isolated observations which might be profitably collated, and which suffice to show that the mesenchyma arises from the mesothelium. In spite of this the brothers Hertwig advanced, 81.1, as stated previously, p. 155, the theory that the two mesodermal tissues are of different origin — a theory which we now know to be false, as, indeed, was proved by Goette six years before the Hertwigs' theory. That all parts of the mesoderm have a common origin was the view of the older embryologists, and, in fact, the differentiation of the middle layer was in the main correctly given by Eemak, 50. 1. The unity of the mesoderm has always been maintained by Kolliker in his text-books and articles, one of which, 84.4, contains a series of well-founded criticisms of other views and a sufficient defence of his own. Eecently the origin of the mesenchyma has been specially investigated by H. Ziegler, C. 88.1,Rabl, 89.2, and VanWijhe, 89. 1, in elasmobranchs, and by R. Bonnet, 89. 1, in the sheep. The mesenchyma rises from cells thrown off from the mesothelium. The entire mesothelium participates in this process, but not to an equal degree, nor at the same time throughout its whole extent. The first part to produce the mesenchymal cells in elasmobranchs is the splanchnic leaf at the point where the nephrotome unites with the myotome; at this point, as stated above, there are traces of 208 THE EMBRYO. an evagination. A little later, Fig. 133, the outer wall of the myotome throws off cells throughout its whole extent, and at the same time a much less active emigration is going on from the ne- phrotome, while it is not until much later that the walls of the splanchnoccEle contribute to the mesenchyna. Whether the meso- derm of the area vasculosa, in which there is at first no coelom, con- tributes directly to the mesenchyma is uncertain; it certainly produces (see Chapter X.) the blood-vessels, and whether the vessels ought to be considered as mesenchyma or as a distinct tissue is still under debate. An excellent diagram illustrating the mesothelial sources of the mes- enchyma is given by H. Ziegler, 88. 1, Taf. XIII., Fig. 1. For am- phibians we have Goette's detailed account; the mesenchyma arises from all parts of the mesothelium, the cells moving off from their epi- thelial union but remaining con- nected together by short thick pro- cesses, which are never mmierous, though variable in number ; the cells all contain a great deal of deuto- plasm; as development progresses the yolk grains disappear, the cells become entirely protoplasmatic, and the number of intercellular processes increases, the processes at the same time becoming finer and longer. There are regional distinctions in the density of the tissue, which are constant. The tissue increases by additions from the mesothelium dur- ing a certain period, and continu- ouslv by the proliferation of its own cells. Goette also, 75.1, 497-498, asserts that after the circulation begins leucocytes leave the blood- vessels and are transformed into Bildungsgeivebszellen; he does not seem to me to offer sufficient proof to justify this assertion. Goette attributes. I.e., 493, the moving ing apart of the cells, not, as seems to me most reasonable, to their own growth, but to the accumulation of intercellular fluid, which he assumes to be produced by transfusion from the archenteron. In mammals and birds the manner in which the myotome contributes to the mesenchyma is now pretty thoroughly understood, but the share taken by the nephrotome and lateral plates has still to be ascer- tained. In both classes the metamorphosis of the outer wall occurs much later than that of the inner wall, which very early becomes Fig. 122.— Pristiurus Embryo with Forty- five to Forty -six Segments; Cross-Section of the Anterior Part of the Body. Md, Medul- lary groove ; Gl-, anlage of the ganglion ; My, myotome; Coe^ coelom of the nephrotome; Ent., entoderm; Ao^ aorta; x, sub-notochor- dal rod or hypochorda; nch., notochord. After C. Eabl. ORIGIN OF THE MESENCHYMA. 309 considerably thickened by the multiplication of its cells. Heape, 86.2, describes the process in the mole nearly in the following words: The myotomes at Heape 's stage H commence first in the anterior region, and gradually assuming the same relations posteriorly, to divide into two portions, an outer arched epithelial portion and a thicker inner portion composed of anastomosing cells of distinctly mesenchymal type, which give rise to the axial mesenchyma, and participate in the formation of the definite muscle-plate. The myo- tomic cavity is very marked. In the next stage (J) the anterior myotomes exhibit still further changes ; the inner layer has grown very considerably, and the row of its cells next the cavity are more closely packed and so have assumed the epithelial form, while the re- mainder of the layer preserves the anastomosing character of the cells ; the inner layer of the myotome is therefore separated into its two parts ; the epithelial part becomes continuous with the outer layer, and the two epithelia together constitute the so-called double muscle- plate. Although arising from separate segments the axial mesen- chyma loses almost immediately every trace of segmental arrange- ment, and there is no real proof that its segmental origin has direct influence upon the segmental arrangement of the vertebral and other structures differentiated later from the mesenchyma. Ultimately, as in other vertebrates, the entire outer layer is converted into mesen- chyma, which forms the dermal layer, R. Bonnet, 89.1, 54. Comparison with Amphioxus. — Hatschek's observations, 88.1, on the differentiation of the mesoderm of Amphioxus show that there are many striking resemblances with the history of the vertebrate meso- derm as given above. The mesoderm con- sists at first of a series of paired mesothe- lial sacs ; the ventral portions of the sacs fuse into a continuous splanchnocoele ; in a larva several weeks old the inner wall of the dorsal segments is a thick epithelium, which produces the muscles on the inner or entodermal side of the cavity of the segment {myocoele of Hatschek) ; the me- sothelium becomes a thin pavement epi- thelium. After about three months of pelagic-life, the larva changes into Am- phioxus and takes to the sand. At this time the lower edge of the segment is found to have formed a diverticulum, which stretches upward beween the mus- cles on the one side and the medulla on the other. The segments have also ex- tended into the dorsal and ventral fins and have there formed cavities. These rela- tions are illustrated by the accompanying diagram. Fig. 133, after Hatschek. The points of special interest to us are four: 1, the formation of the splanchnocoele by the fusion of segmental cavities ; 2, the develop- ment of the muscles exclusively from the inner layer of the secon- 14 Fig. 13.3. —Diagram of a Cross Section of a Young Amphioxus. /, /, /, Parts of the coelom of the seg- ment; J/, splanchnoccele ; 1, outer layer of segment; 2, muscle layer; 3, 4, 5, 6, portions of sclerotomic diverticulum; 7, splanchnic meso- derm around the entoderm. After Hatschek. 210 THE EMBRYO. dary segments ; 3, the absence of differentiation in the outer layer of the segment ; 4, the outgrowth of mesothelium passing upward between the muscular layers and the axial structures, medulla, and notochord. It is probable that all these four peculiarities recur in the true vertebrates, though masked principally by the fact that the outer layer of the segment and the epiaxial diverticulum both lose, the former gradually, the latter almost from the start, all trace of epithelial structure, and become converted into mesenchyma. Of course the assumption that the vertebrate splanchnocoele arises in the same way as in Amphioxus, is at present entirely hypothetical. CHAPTER X. ORIGIN OF THE BLOOD, BLOOD-VESSELS, AND HEART. The circulatory system is developed from two anlages which are at first independent. The heart arises in the cervical region of the embryo ; the blood-vessels and first blood-cells in the extra-embryonic area vasculosa ; the blood-vessels subsequently grow into the embryo and unite with the heart. The heart begins to beat before the vessels are connected with it, so that as soon as the connection is established the circulation begins. The heart contains at first only a clear fluid; after the circulation has begun blood-cells come in through the ves- sels from the area vasculosa. The first blood-cells have a reddish color and a round nucleus. Somewhat later the colorless granular leucocytes appear, but where they arise is uncertain. In all verte- brates except mammals the red cells persist throughout life, but in mammals they are confined to the foetal period, during which they are gradually replaced by the non-nucleated red-blood globules (plas- tids) . ^Much confusion exists as to the nature and development of the blood, because the great majority of writers have ignored the important fact that the mammalian adult blood-globules are a new acquisition of that class and are not homologous with the red-blood corpuscles of other vertebrates. Mammals have three kinds of blood corpuscles : red cells, leucocytes, and the adult red globules ; all other vertebrates have two kinds only. An immense deal has been written on the development of the blood in the embr}'o, and there is perhaps no other question in em- brj^ology which has been so much studied and yet left with such a variety of opinions as to its right answer. In the following pages I have endeavored to collate what seem to me the best-established results ; but until some one subjects the literature of the subject to a critical revision, based on a thorough comparative investigation of the development of the blood and blood-vessels throughout the ver- tebrate series, we can hardly expect a satisfactory history o'f the embryonic blood. We have to distinguish between the primary and secondary vas- cular anlages. I. Blood-Vbssbls and Blood. Primary Vascular Anlages. — These are cords of cells which appear first in the area vasculosa and rapidly extend into the embryo; the cords form a network ; scattered clusters of cells in the cords very early assume the haemoglobin color and appear as reddish-yellow spots which have long been known, and are described by Pander, Von Baer, 28.2, Eemak, 50.1, Prevostet Lebert, 44.1, and others. We owe to His, 68. 1,* 95-103, the first exact account of the origin 212 THE EMBRYO. .1* *^ .^ S.t ^ ..?V^ of blood-vessels in the chick; since then the studies of Disse, 79.1, Gotte, 74. 1, Kolliker (" Entwickelungsges."), Balfour,73. 1, J. KoU- mann, 84.3, Uskow, 87.1, and others have added a little to the descriptions by His. It is now demonstrated that the blood arises in amniota from the mesoderm and not from the yolk, as was, I be- lieve, first suggested for teleosts by LerebouUet and recently by Ryder. The exact history of the first blood-vessels has yet to be studied in other amniota than the chick. In the chick the distal portion of' the mesoderm has no coelomatic cavity when the development of the blood begins; the mesoderm lies close against the entoderm or germinal wall (Keimwall). The juxtaposition of the two layers has led His and I f others to consider that the entoderm or yolk gave off the cells which form the mesoderm of the area vasculosa. This portion of the mesoderm was early distinguished by German writers under the name of Gefdss- schicht or vascular layer (feuillet angioplas- tique), and has been called the blood - germ (Blutkeim) ; by His it is identified as a stage of the parablast, see Chap- ter VI. The first indica- tion of the blood-vessels is a reticulate appearance of the layer, which can be recognized in surface views at the end of the first day and rapidly in- creases in extent and distinctness during the second day of incubation. As soon as there are several primitive segments the network shows traces of coloration in irregularly shaped reddish-yellow spots which are largest and most numerous around the caudal end of the embryo; these spots are the so-called blood-islands, Fig. 124. The network appearance is due to thickenings of the mesoderm, as is evident from sections. The two primary layers are separated so that the mesodermic thickenings lie between them. Between the thickenings are irregular lacunae, Fig. 124, b b, which are only partly filled with mesodermic cells ; these lacunae by their subsequent expansion and fusion develop during the latter half of the second day the ccelom of the area vasculosa, and always so that the thickenings (or blood- vessels) are on the entodermal side, Fig. 126. In other words, as soon as the two leaves of the mesoderm are differentiated in the area vasculosa the blood-vessels are found exclusively in the splanchnic leaf. In the sheep they appear also in the somatic leaf, R. Bonnet, 89. 1 , 56, or future amnion, but they soon disappear and never con- tain any blood-corpuscles. The network of blood-vessels of the L Fig. 134.— Surface View of a small Part of the Vascular Network of an Embryo CMck o£ two Days, a a, Blood-ves- sels ; 6 ft, mesoderm between the vessels ; c c, blood-islands. From KGlliker. BLOOD-VESSELS AND BLOOD. 213 vascular area form at first a thick network without distinction of stem or branch, and are all in one layer, none overlying the others (Kolliker, " Grundriss," p. 60), Fig. 125;theedgeof theareais ^ marked by a single large ->■ — vessel which is known as the vena, or better, sinus terminalis, Fig. 125, vt. I have spoken of vessels, /sf— , ( but up to this time the vascular anlages are solid. The vena termi- nalis persists for some time as the distal boun- dary of the area, while it , is spreading farther and i farther over the yolk, but i by the end of the fourth j day it is no longer dis- tinguishable as a distinct structure (Prevost et Le- bert, 44.3, 240). The vena terminalis ultimate- \ \j becomes connected \^' with the venous system ,,,.,= -,7 i » i . ^i, . ^ „ , , 1.1 1 J • 1 ' "^- ^^5. —Vascular Anlages of the Area Vasoulosa of a 01 the Cmck, but m rab- Chlck ot Forty Hours, ps, ps, Blood-islands ; rt, vena ter- bits with the arterial sys- '""^^"'- ^™'" ^°""^"'"- ^ '« ^'^'"^■ tem ; for this reason the term sinus is to be preferred to vena as ap- plied to this vessel. The blood- islands are spots where there is a cluster of cells which remain attached to the walls of the vessels in the area vasculosa (see Fig. 126, bl. is). The cells develop haemoglobin in their interior, hence the clusters have a reddish color, which renders the islands very conspicuous in surface views of fresh specimens. The blood- islands of the chick appear first in the area opaca, and almost imme- diately after in the pellucida also. They have at first a rounded or branching form, Fig. 1 24 ; in the inner part of the layer they are small and stand alone ; toward the periphery they are larger, closer set, and more united with one another ; their development is greater around the caudal end of the embryo. They are situated, chiefly, at the nodes of the vascular network. When the , solid vascular cords acquire a lumen, the islands, Fig. 120, bl. is, remain attached to one side of the vessel, like a thickening of its wall. The cells of the islands ultimately become free blood-corpuscles. The growth of the primary anlages takes place by the develop- ment of buds from the vessels already formed, as first shown accu- rately by Prevost et Lebert, 44.3, 239; these buds are rounded or pointed and elongated, forming as it were spurs; they often end by meeting one another and uniting ; they are usually hollow from the first, and after they meet one another or an adjacent vessel, the cavi- ties become continuous and thus the vascular network is extended. A. Goette, however, maintains, 75. 1, 497, that the network arrange- 214 THE EMBRYO. ment exists from the start in all vertebrates, and that the apparent budding is due to the progress of vascular differentiation into in- diflEerent mesenchymal cells. In mammals the solid primary anlages appear in the extra embryonic area vasculosa, and extend later into the embryo. So far as known to me there has been as yet no exact investiga- tion of their history. They pre- sent well-marked blood-islands, which are thickenings of the mesoderm, and make their first appearance in rabbit embryo of the eighth day just before the first primitive segments (Kolli- ker," Entwickelungsges.," 266). The growth of the network in the rabbit by the formation of solid buds which become hollow has been described by Wisso- sky, 77.1. Gro-wrth of the Vessels into the Embryo. — The fact that the vessels penetrate the embryo after they have appeared in the area vasculosa was first discovered by His, 68.1, 99, and is now a familiar pheno- menon. It is evident that this penetration may take place in two ways : it may be a progres- sive differentiation of cells al- ready present (c/. Goette, 75. 1, 539), or it may be an actual in- growth of vaso-formative tis- sue ; the balance of evidence is in favor of the latter alterna- tive, which accordingly, follow- ing His in this respect, the ma- jority of embryologists have adopted. In the chick the vas- cular differentiation extends from the area opaca to the area pellucida, and thence into the body proper of the embryo. But in the lizard (Strahl, Mar- burg, Sitzber., 1883, 60-71) the vessels appear first in the area pellucida and thence extend into the area opaca and the embryo. The entrance of the vessels into the embryo chick begins toward the end of the second day. It is effected, according to His, 68.1, 99, by buds, which are at first solid cords, and grow toward the embryo, uniting as they extend into a network; the hollowing BLOOD-VESSELS AND BLOOD. 215 out of the cords likewise progresses centripetally. The penetrating vessels foUow certain prescribed paths. A part of the vessels run along the posterior edge of the amnio-cardial vesicles, and enter into connection with the posterior end of the heart which has meanwhile been developed — owing to the early separation of the head end of the embryo from the yolk this is the only part of the heart the vessels can reach directly. While the vessels are approaching the heart their differentiation into various sizes is going on, the smallest ones to remain as capillaries, the larger ones to become arteries or veins ; this differentiation, which has yet to be foUowed step by step, leads to there being only two main vessels, the so-called onvphalo-mesaraic veins, which actually open into the hind end of the heart. Another set of vessels penetrates along the splanchnopleure of the rump on each side, until they attain the small space between the notochord, myo- tome, and entoderm, where they fuse (Turstig, 86.1), so as to form a longitudinal vessel, the anlage of the aorta descendens, which is primitively double. The aorta appears first at the head end of the rump and hence its development progresses backward ; it also grows forward over the heart, bends over ventrally just behind the mouth, and, passing around the blind end of the vorderdarm, approaches the median line and unites with the cephalic end of the tubular heart. An utterly different history of the origin of the aorta, namely, from the median dorsal wall of the archenteron, is asserted by C. K. Hoffmann, 92.1, for the dog-fish. The heart begins to beat before the vessels unite with it; the first blood-cells have already been formed ; hence as soon as the vmion is accomplished the blood- circulation starts up, the blood passing through the aorta to the rump, thence by numerous lateral branches to the area vasculosa, and returning by the two omphalo-mesaraic veins to the heart. The course and modifications of the primitive circulation are described and figured in Chapter XIV. , on the germinal area. Origin of the First Red Blood-Cells.* — I consider it proba- ble that the red blood-cells of all vertebrates arise, as has been main- tained byH. Ziegler, 89.1, by proliferation of the endothelial lining of the vessels. This conclusion is based — 1, upon the fact that in various vertebrates, notably in bony fishes, elasmobranchs, and all amniota, certain parts of the vascular system are at first solid cords of cells, and of these cords the central portion becomes blood-cells, the peripheral portion the vascular wall; it seems to me that the right interpretation is to regard the central cells as belonging with the outer cells, and therefore equivalent to the product of an endo- thelial proliferation ; '3, upon the origin of red cells from the walls of the venous capillaries of the bony marrow of birds (J. Denj^s, 87.1). In all these cases the blood-cell is a liberated specialized endothelial cell. A. Goette is the principal opponent of this view, and has maintained thatin Petromyzon, 90.1, 60, Amphibia, 75.1, 538, and birds, 74. 1, 180-186, the blood-cells of the embryo have an *It is singular that so close an observer as Balfour should have maintained, as he did. 73.1, that the blood-cells of sauropsida are metamorphosed nuclei, and this view is still adhered to in his "Elements," 3d ed. 1883., Balfour's error was due to the fact that the cells, when first set free, have a minimimi quantity of protoplasm around the nucleus, and this he did not observe : the nuclei have too at first a very distinct large nucleolus, which Balfour wrongly assumed to represent the nucleus of the future corpuscle. 216 THE EMBRYO. origin different from the endothelium, the former arising from the yolk, the latter from the mesoderm. Although Goette is one of the very best of embryological observers, I cannot agree with him on this point, for I feel satisfied that he is in error as regard the chick, while in regard to the lamprey and the land frogs it is possible that Goette's observations are incomplete — certainly his descriptions are less clear than those of the origin of the blood-cells within the vas- cular anlages. It must be added that Davidoff has maintained, 84. 1 , that in the salamander the blood-cells arise from the surface of the yolk ; but his statements need, I think, verification. The blood-cells of teleosts arise, at first at least, in certain large vessels within the embryo (Wenckebach, H. Ziegler, 87.1,89.1, compare also H. Ambert, 56.1), which are formed as solid cords, the central cells of which are metamorphosed into blood- corpuscles. At the time the circulation begins there are no blood-vessels over the yolk, but definite blood-channels, which are merely grooves on the yolk or passages between the yolk and the ectoderm ; these channels subsequently acquire mesenchymal walls when the mesoderm grows out over the yolk. Owing to this peculiarity of the early vitelline circulation blood-cells appear over the yolk before there are blood- vessels, and the observation of this fact seems to have led several observers to the error of attributing the origin of the blood-cells to the yolk or the superficial layer thereof (Kupffer's periblast) . For a sjmopsis of the various opinions see Mcintosh and Prince, 90^1, 782-783. In elasmohranchs (J. KoUmann, 85.1, 297) there are mesodermal blood-islands, which expand and unite, forming a net- work in the area opaca ; the vessels are at first solid, the central cells become blood-cells, the peripheral cells endothelial walls ; so far as observations go it is possible, however, that all the cells of the blood- islands become blood-cells, and that the endothelium is simply an overgrowth of mesenchyma, but in view of the development in other vertebrates this possibility has little probability. The development of the blood in reptiles and mammals needs thorough study, but we know that it is closely similar to that in birds. In the chick, as stated above, the cells of the blood-islands form the first blood-cells, and this statement probably applies also to all amniota. For the origin of blood-cells in the embryo see the following section. Secondary Vascular Anlages. — These are buds which arise from the vessels already present in the embryo, similar to the buds already described in the area vasculosa. There being no real division between the primary and secondary anlages the distinction is used merely for convenience of description. The secondary anlages, like the primary, give rise to the endothelium of the wall only ; when a vessel becomes an artery or a vein the media and adventitia are added by differentiation of the surrounding mesenchyma. The sec- ondary anlages can be found in mammals in various parts of the body during embryonic life, and even after birth, and in Amphibia may be studied during the larval period ; the tail of tadpoles being a favorite object for this purpose (Golubew, 69.1). The second- ary anlages were, so far as I know, first accurately described in batrachians by Prevost et Lebert, 44. 1 ; they were followed two BLOOD-VESSELS AND BLOOD. 217 years later by Kolliker, 46.2, see also Golubew, 69.1, Arnold, 71.1, and Ranvier's " Traite technique," 618, 633. In mammals they have been well described by Ranvier, 74. S, E. A. Schafer, 74. 1, Kolliker ("Entwickelungsges.," 171, "Grundriss," 63), and others. The secondary anlages appear as thorn-shaped points projecting more or less nearly at right angles from the walls of capillaries al- ready formed. A. Goette, 75.1, 544, has maintained that these are not real outgrowths, but differentiations of intercellular processes present ab initio in the mesenchyma. These points rapidly elongate into fine threads, which may join the wall of another capillary or the tip of another point ; Golubew states that when two points unite in the frog, they overlap and then unite by their sides ; while the point is growing the cavity of the parent capillary extends into the base of the point, and penetrates farther and farther, so that the thread-like point becomes gradually enlarged into a capillary blood- vessel. The capillaries formed in this way show a marked tendency to form loops. Very similar is the account quoted below by E. A. Schaeffer in Quain's " Anatomy," ninth edition, II., 198, 199) : " Within the body of the embryo vessels are formed in like manner from cells belonging to the connective tissue. One of the most favorable objects for the study of the development of the blood-vessels and their contained blood-corpuscles is afforded by the subcutaneous tissue of the new- born rat, especially those parts in which fat is being deposited. Here we may observe that many of the connective-tissue corpuscles are much vacuolated, and that the protoplasm of some of them pre- sents a decided reddish tinge. In others the red matter has become condensed in the form of globules within the cells, varying in size from minute specks to spheroids of the diameter of a blood-corpuscle or more. At some parts the tissue is completely studded with these cells, each containing a number of such spheroids, and forming, as it were, 'nests' of blood-corpuscles or minute 'blood-islands.' After a time the cells become elongated and pointed at their ends, sending out processes also to unite with neighboring cells. At the same time the vacuoles in their interior become enlarged, and coalesce to form a cavity with the cell in which the reddish globiiles, which are now becoming disc-shaped, are found. Finally, the cavity extends through the cell processes into those of neighboring cells and into those sent out from pre-existing capillaries, but a more or less exten- sive capillary network is often formed long before the connection with the rest of the vascular system is established. Young capilla- ries do not exhibit the well-known lines when treated with nitrate of silver for the differentiation of the hollowed cells and cell-processes into flattened cellular elements is usually a subsequent process. The mode of extension of the vascular system in growing parts of older animals, as well as in morbid new formations, is quite similar to that here described, except that blood-corpuscles are not developed within the cells which are forming the blood-vessels." The development of new capillaries in the manner just described also takes place from the vessels formed by vasoformative cells. The secondary vascular anlages of the foetal liver have been specially studied by P. Kuborn, 90.1; they correspond to the so- 218 THE EMBRYO. called foetal hepatic giant cells of early authors, and give rise to vascular walls, red cells, and later (embryos of three or four centi- metres) to the red plastids, compare p. 321. Vasoformative Cells. — In all secondary anlages of the vessels we have outgrowths of vessels already present ; there are also vessels developed from special vasoformative ceUs, which have no connection with previous vessels ; the origin of the vasoformative cells has still to be ascertained, but it may be safely asserted that they are derived from the mesenchyma. These cells were, I believe, discovered by L. Eanvier (74.2, and "Traite technique," 625), who studied them in the omentum of the rabbit before and after birth. He found small spots of milky appearance, which he designates as "taches laiteuses," and which contain ordinary connective-tissue corpuscles, and fibrillse, numerous leucocytes, and vasoformative cells. The last, in rabbits from two to eight weeks old, are finely granular, branching often anastomosing, elongated cells with elongated nuclei; earlier they are scattered, spindle-shaped cells. Soon a capillary from the neigh- borhood grows in and unites with the vasoformative network, and thereupon the excavation of the network begins, the lumen of the capillary gradually extending throughout the cluster of vasoforma- tive cells. Primitive Blood -Vessels. — The first vessels consist merely of a wall of protoplasm with scattered nuclei, and accordingly are all essentially alike in structure ; the first differentiation is one of size only, the vessels that are to become arteries and veins rapidly in- creasing their calibre, while the mesenchyma around them is still undifferentiated. The protoplasmatic wall in cross-sections of a vessel is thick enough to contain a nucleus. The next step in devel- opment is the thinning out of the layer, so that the nuclei become protuberant as in the adult endothelium ; at the same time the pro- toplasm becomes divided into distinct cell territories, and intercellu- lar lines are developed and may be impregnated with nitrate of silver, as in the adult. The vessels grow by the multiplication of the cells of their walls. W. Flemming, 90. 1, has shown that in the capillaries the nuclei undergo karyokinetic division, and that the division of the proto- plasm takes place later. The distribution and metamorphoses of the principal vessels are discussed in Chapter XIV. The. red blood-cells are the only elements contained in the blood during the earliest stages of the vertebrate embryo. When the circulation begins the number of corpuscles is small, but rapidly increases thereafter. The cells are at first round (in probably all vertebrates) ; in the chick they measure from 8.3 to 12.5 m. The nu- cleus is large, more or less nearly spherical, and surrounded by a layer of protoplasm (Minot, 122), which is so thin as to have been often overlooked. The cells at first are granular and slightly colored (Pre- vost et Lebert, 44.3, 241; KolMker, "Grrundriss," 63), and then be- come more colored and homogeneous, scarcely showing the nucleus during life, though it comes out very clearly as soon as the corpus- cles are removed from the vessels or acted upon by hardening re- agents. The nucleus in hardened corpuscles stains deeply. In BLOOD-VESSELS AND BLOOD. 210 amphibians the young blood-cells, like all the other cells of the em- bryo, contain numerous yolk granules ; as the granules disappear the >?■ CO ^ rm\' t?to BP j!3§ ® ® (®) W i*; i^ ta .S Ci (5^)^;?^ |x-j/s/ .® ' 35 i'S Bg. nuclei and bodies of the cells both acquire a more homogeneous and opalescent appearance, and at the same time become flattened, elongated, and colored (A. Goette, 75.1, 770). 220 THE EMBRYO. The primitive form of the vertebrate red-blood cell is probably spherical, or at least spheroidal, and the characteristic mature shape is not assumed until later, as I have learned from my own observa- tions on a considerable variety of embryos. This statement is further supported by A. Goette's observations on Petromyzon, 90. 1, 66, and Bombinator, 75.1,538. In the chick the mature elliptical form begins to predominate during the fourth day; the earlier round form is still encountered for several days, but it gradually becomes rarer (Prevost et Lebert, 44. 3, 342) . Minot, 123, has outlined the progressive differentiation of the red cells in sharks, salamanders, chicks, and rabbits. The following description refers primarily to the chick : By following the develop- ment we find that the protoplasm enlarges for several days, and that during the same time there is a progressive diminution in size of the nucleus, which, however, is completed before the layer of protoplasm reaches its ultimate size. The nucleus is at first granular, and its nucleolus, or nucleoli, stand out clearly ; as the nucleolus shrinks it becomes round and is colored darkly and almost uniformly by the usual nuclear stains. This species of blood-corpuscle occurs in all vertebrates and represents the genuine blood-cells. The blood-cells of mammals pass through the same metamorphoses as those of birds. For example, in rabbit embryos, the cells have reached the ichthy- opsidan stage on the eighth day ; two days later the nucleus is already smaller, and by the thirteenth day has shrunk to its fmal dimensions. According to the above description we can distinguish three principal stages: 1, young cells with very little protoplasm; 3, old cells with much protoplasm and granular nucleus; 3, modified cells with shrunken nucleus, which colors darkly and uniformly, Fig. 127. I do not know whether the first form occurs in any living adult verte- brate, although the assumption seems justified that they are the primitive form. On the other hand, the second stage is obviously characteristic of the Ichthyopsida in general, while the third form is typical for the Sauropsida. Therefore, the development of the blood- cells in amniota offers a new confirmation of Louis Agassiz' law (Haeckel's biogenetisches Grundgesetz) . Multiplication of the red cells by division was recorded by Remak, 50. 1, 164, and has since been frequently observed. Special attention was directed to its occurrence by Peremeschko in 1879, 79.1, 81.1, and byBizzozero {CM. med. Wiss., 1881, Moleschott's " Unters. zur Naturlehre," XIII.) in 1881, and has since been studied by Bizzozero et Torre, 84.1, Bizzozero, 84. l,Funcke, 80.1, Eberth and Aly, 85.1, A. Mosso, 88.2, and others. The division is indi- rect or karyokinetic, and takes place across the longitudinal axis of the corpuscle, with which the nuclear spindle is parallel. The process has been observed in bony fishes, amphibians, adult Sauropsida, and in amniote embryos. The division occurs only in young or partly differentiated corpuscles; the divisions, for example, are abundant in the blood of the chick of from three to five days ; the sixth day they are rarer, the tenth seldom, and after hatching are not found in the circulating blood at all (Funcke, I. c.) . It is, accordingly, safe to assume that the proliferation of the red cells is typical for all verte- brates. Their number is further increased by additions from various BLOOD-VESSELS AND BLOOD. 221 sources in the embryonic and (adult iion-mammalian) vertebrates ; but, so far as at present known, the mammals have only the red cells, which arise directly from the primary vascular anlages, therefore •the discussion of the maintenance of the supply of red cells falls out- side our scope. The problem has been much debated ; the investiga- tion which seems to me to have led to the best results is that of J. Denys, 87. 1. For the reader's convenience I cite also the following authorities, but the list is very incomplete : Bayerl, 84.1, W. H. Howell, 88.1; Lowit, 87.1, 91.1, E. Neumann, 74.1, Malassez, 82.1; Obrastzow, 81.1; G. Pouehet, 80.1; and Rindfleisch, 80.1. For additional references see Quain's "Anatomy," ninth edition, II., 40. Disappearance of the Red Cells.— The red cells form the permanent red-blood globules in all vertebrates except the mammals. In mammals they disappear during embryonic life or soon after birth. Although they persist for a long period, it will be convenient to state here what little is known of their history. How they disappear is not known, although several authors have main- tained that they are transformed into red plastids, but this opin- ion seems to me ill founded. W. H. Howell, 90.1, reports the interesting discovery that the nucleus of the mature red cells is ex- truded in mammals leaving the body of the cell ; in consequence he maintains the plausible conclusion that the extrusion is the means of developing the non-nucleated red corpuscles, but I am more inclined to regard it as a step in the degeneration and destruction of the red cells. In the human embryo at one month the red cells are the only blood-corpuscles; at two months they are the most numerous, al- though the plastids have begun to appear ; at three months they form only a small minority of the corpuscles. , Origin of Leucocytes. — The origin of the first colorless corpus- cles in the embryo is still uncertain. The blood is found to contain for some time only the red cells, the leucocytes appearing in the chick (Prevost et Lebert, 44.3, 243), about the eighth day of incuba- tion ; in the rabbit, it is said, about the ninth day, and in elasmo- branchs not until the embryo is well advanced in development, A. Mosso, 88. S. It is to be noted that after the blood-vessels and red- blood-cells the leucocytes are the first cells to be differentiated from the mesenchyma, the remaining mesenchymal tissues (Chapter XIX.) being differentiated gradually and to a large extent simultaneously. So far as I know, the subject has never been carefully investigated, nor is there even any exact description of the appearance and number of the first leucocytes. After the lymph-glands appear they probably assume the function of producing leucocytes ; but the process in embryonic glands has still to be studied, and accordingly for further information the reader is referred to the standard histologies. That the leucocytes multiply by direct or akinetic division has been recorded by several observers, L. Ranvier, J. Arnold, 84.1, and others. Origin of Mammalian Blood-Globules or Red Plastids. — There are many opinions as to the origin of the non-nucleated red blood-globules of mammals. The best-founded conclusion is,^it seems to me, that of E. A. Schafer, who traces them to local differ- 323 THE EMBRYO. entiations of the protoplasm of the vasifactive cells. This view makes the globules comparable to the plastids of botanists, such, for instance, as the chlorophyll granules. As the terms " globules" and " corpuscles" have been applied indiscriminately to all the formed elements of blood, and as it is desirable to have a simple term which shall also indicate the morphological separation from the other "blood-corpuscles," I shall apply the term "red plastids" to the non-nucleated mammalian adult red globules. The chief opinion rivalling Schafer's is that the red plastids are derived from nucleated corpuscles, which have lost their nuclei and shrunk, the plastids being always much smaller than the red cells. This view has been specially advocated by Kolliker, "Gewebelehre," 5te Aufl., 1867, p. 638, is found in several subsequent writers, and has been very re- cently brought forward by Casimoro Mondino, 88.1, but sufficient observation to justify it has not been furnished in my judgment. The strongest evidence in favor of the conversion of nucleated corpus- cles into plastids is that which is presented by Howell, 90.1, and men- tioned p. 331. Similar to this view is that which traces the plastids to modifications of leucocytes occurring after birth, F. Sanfelice, 89. 1; the white cells are supposed to shrink, lose their nuclei, and become charged with haemoglobin. Yet another opinion affirms that the marrow of bones produces from certain of its cells the red plastids, but the defenders of this opinion are by no means agreed among themselves as to how. For a good synopsis of the conflicting theories see Schafer in Quain's "Anatomy," tenth edition. Vol. I., Pt. II. The first red plastids certainly arise in the vasifactive cells in various parts of the embryo. Schafer in Quain's " Anatomy," ninth edition, II., 36-37, gives the following description of the process : " A part of the protoplasm of the cell acquires a reddish tinge, and after a time the colored substance becomes condensed in the form of glob- ules within the cells, varying in size from a minute speck to a sphe- roid of the diameter of a blood-corpuscle, or even larger; but grad- ually the size becomes more uniform. Some parts of the embryonic connective tissue, especially where a vascular tissue such as the fat is about to be developed, are completely studded with cells like these, occupied by a number of colored spheroids and forming nests of blood-corpuscles, or minute 'blood-islands.' After a time the cells become elongated and pointed at their ends, and processes grow out to join prolongations of neighboring blood-vessels or of similar cells. At the same time vacuoles form within them, and becoming enlarged coalesce to form a cavity filled with fiuid in which the reddish globules, which are now becoming disc-shaped, float. Finally, the cavity extends through the cell processes into those of neighboring cells, and a vascular network is produced, and this becomes eventu- ally united with pre-existing blood-vessels, so that the blood-corpus- cles which have been formed within the cells in the manner described get into the general circulation. This 'intracellular ' mode of devel- opment of red blood-corpuscles ceases in most animals before birth, although in those which, like the rat, are born very immature it m^y be continued for a few days after birth. Subsequently, although new vessels are found in the same v/ay, blood-corpuscles are not pro- BLOOD-VESSELS AND BLOOD. 233 duoed within them, and it becomes necessary to seek for some other source of origin of the red-blood discs, both during the remainder of the period of growth, and also during adult life, for it is certain that the blood-corpuscles are not exempted from the continual expenditure and fresh supply which affect all the other tissues of the body." Very early in embryonic life the liver, as first pointed out by Kol- liker, and more fully demonstrated by Neumann, 74. 1 , becomes the principal seat of blood formation. The secondary vascular anlages are very prominent in the foetal liver and in sheep embryos of four centimetres and more in length. P. Kuborn, 90.1, has traced the development of red plastids from the protoplasm only, as described by Schafer. A similar result is reached by R. Nicolaidos, 91.1, from studying the production of red plastids in the mesentery of young guinea-pigs, see also Wissosky, 77. 1. The process of plastid development is easily followed in the mesentery of the human foetus. It seems to me probable that research will ultimately establish the origin of red plastids in the adult also, as intracellular protoplasmatic bodies entirely distinct from the nuclei, and in no way to be homolo- gized with cells. Kultschitzki, however (see Hofmann-Sohwalbe's Jahresber., 1883, 58-59), asserts that in the lymph-glands of the rab- bit the red plastids arise within cells by metamorphosis of the nuclei ; to nuclei Balfour traced, he supposed, the red cells of birds, com- pare p. 315, foot-note. Origin of the Blood -Plates. — C. Mondino and L. Sala, 88.1, aflfirm that the blood-plates multiply by division, and being nucleated in the non-mammalian vertebrates, according to these authors, they divide karyokinetically ; while in mammals the plates have no nu- cleus, but the larger plates have chromatine granules, which, how- ever, divide as do the plates. They state that the plates are present in mammalian blood as soon as it begins to circulate. In the French resume of their work {Arch. Ital. Biol., XII., 304), they state that Fusari has confirmed their observations in an article in the Biforma medica, 13 Agosta, 1889. I question most decidedly the trustworthi- ness of these statements, for the author's figures suggest at once that they have mistaken distorted blood-globules for blood-plates. No other observations on foetal blood-plates are known to me. It should be added that L. Lilienfeld, 92. 1, has advanced the hypothe- sis that the plates are derived from leucocyte nuclei, while Howell, 90. 1, suggests that they are the extruded nuclei of red cells. Morphology of the Blood-Corpuscles. — The following con- ceptions have been advocated by Minot, 122. The preceding sections show that the vertebrate blood-corpuscles are of three kinds : 1, Red cells; 2, White cells; 3, Plastids. The red and white cells occur in all (?) vertebrates; the plastids are confined to the mammals. The red cells present three chief modifications ; whether the primiti^'e form occurs in any living adult vertebrate I do not know ; the second form is persistent in the Ichthyopsida, the third form in the Saurop- sida. According to this we must distinguish : A. One-celled blood, i. e., first stage by all vertebrates; the blood contains only red cells with little protoplasm. B. Two-celled blood, having red and white cells. The red cells have either a large, coarsely granular nucleus (Ichthyopsida) 224 THE BMBKYO. or a smaller, darkly staining nucleus (Sauropsida, mammalian embryos) . C. Plastid blood, without red cells, but with white cells and red plastids ; occurs only in adult mammals. Mammalian blood in its development passes through these stages, as well as through the two phases of stage B, all in their natural sequence ; the ontogenetic order follows the phylogenetic. It seems not improbable that an animal may yet be found with blood inter- mediate between B and C in the adult stage. II. Origin of the Heakt. The heart, as has been stated, is developed independently of the blood and blood-vessels ; it contains at first a clear fluid, and begins beating before the blood-vessels from the area vasculosa have joined it. The primitive form of the heart is a straight median tube on the ventral side of the cervical region ; the cephalic end of the tube is connected with the arterial system of the embryo, while the caudal end is primitively connected with the venous system of the yolk. These relations may be traced in all vertebrates, hence the heart arises as the active organ of communication between the yolk or primitive food supply and the embryo. Primitive Mode of Development of the Heart. — In re- gard to the development of the heart we have to distinguish the mode still preserved in the primitive vertebrates (marsipobranchs, ganoids, and amphibians) , elasmobranchs, and in some but not all teleosts (Mcintosh and Prince, 90.1, 775), from the mode in the amniota. In the first mode the heart arises in the median line ; in the second mode the heart arises from two lateral anlages, which subsequently unite in the median line. The difference is not a fun- damental one, but is correlated, as first pointed out by Balfour, with the earlier or later separation of the cephalic end of the embryo from the yolk ; when that separation is retarded the heart is differentiated before the neck of the embryo is folded off from the yolk, compare Chapter XIII. ; this delay occurs in varying degrees in all amniota. The following account of the origin of the heart in Amphibia is based on C. Rabl, Coe 6q Ei^ 86.1, who cites the Jo'°° o^ earlier authorities. The head of the em- bryo early becomes free and projects so far that the neck is free from the yolk also. The mesoderm extends forward on each side between ectoderm and ento- derm, and has a coelomatic cavity on each side. Fig. 128, Goe. The two wings of mesoderm do not, however, meet on the median ventral line, being separated by a ridge, En, of entoderm by which the inner germ-layer comes into immediate contact with the ectoderm, Ec. Fig. 1S8. — Salamandra Maculosa ; Larva, very Young ; Transverse section, to show the Formation of the Coelom in the Heart Region. Coe, Ccelom; En^ entoderm: .Ec, ectoderm; vies, mesothelium. After C. Babl. ORIGIN OF THE HEART. 325 Whether this ridge is preserved to form the endothelium of the heart or is resorbed into the general entoderm is not positively known. In a later stage, Fig. 129, the two mesodermic wings have met in mes Fig. 129. — Salamaudra Maculosa ; Larva with Branchial Arches. Coe^ Coelom ; vdsth^ mesothelium ; Ht^ endothelial heart ; i?c, ectoderm ; Dies, mesoderm ; Ent^ entoderm. After C. Eabl. the median line below the intestinal canal ; the coelom has expanded ; between the mesothelium of each side in the median line is a small mass of cells, Hi^ which soon shows a central lumen, which be- comes the cavity of the heart, while the cells around give rise to the future endothelium ; the endothelium is still in contact with the en- toderm. Below the heart the mesothelia are in actual contact, form- ing a double wall, which soon breaks through, so that the coelom on each side opens into the other, or, in other words, there is now a sin- gle pericardial cavity. The heart has become a two-layered tube ; the inner layer consists of endothelium, the origin of which is dis- cussed in a separate paragraph below; the outer layer consists of mesothelium, which gives rise to the muscular wall of the heart. Later the mesothelium closes over the dorsal side of the endothelium, thus finally separating it from the entoderm. Still later the tubular heart loses its suspension from the dorsal side of the pericardial cav- ity and is attached only at its anterior or cephalic and posterior or caudal extremities, and hence is free to bend and twist within the pericardial cavity in the manner necessary for the evolution of the heart's adult form. Axnniote Mode of Development of the Heart. — Observa- tions on the heart are to be found in many of the older writers on embryology, notably in Von Baer, Prevost et Lebert, Eemak, Bisch- off , and Coste, but until the introduction of section cutting the details of the process could not be observed. The foundations of our present knowledge'were laid by W. His, 68.1, 83-85, and the subject was further elucidated by Kolliker's invaluable observations on the chick and rabbit, recorded in his " Entwickelungsgeschichte ; " Gasser, 77.3, has published an admirable description with figures of the de- velopment in the chick; there are besides numerous references to the heart scattered in recent literature; see, for instance, Hensen, 76.1; Heape, 86. S; Selenka, 86.1, et al. In the amniota the cephalic coelom very early dilates to a much greater degree than the coslom elsewhere, thus developing on each side the so-called ParietalhoMe of German writers, for which I have 15 326 THE EMBKYO. proposed the name of amnio-cardial vesicle. In the chick the early and extreme dilatation of this cavity is weU known, and is intimately correlated both with the closure of the archenteron to form the cer- vical entoderm ic canal {Vorderdarm) , and also with the development of the heart and the origin of the amnion. In the chick the dilata- tion forces the splanchnopleure (splanchnic mesoblast and entoderm) downward on each side ; then bends the splanchnopleure in under the embryo until the two membranes meet in the median line and fuse ; their fusion shuts off the Vorderdarm from the yolk and leaves it as a flattened canal, Fig. 129A, Ph; for further details see Chapter XII. The layer of mesothelium bounding the coelom is everjrwhere distinct ; the mesenchyma is well developed all about the meduUary canal and notochord, Fig. 129A, but is almost entirely absent from the walls of amnio-cardial vesicles, until we reach the distal vascular area, consequently when the vesicles expand the mesothelium is brought close against that portion of the entoderm which is destined to form the Yorderdarm ; where the contact takes place there appear be- tween the entoderm and mesothelium a few very irregularlygrouped mesenchymal cells, Fig. 129 A, JEndo; these are theanlage of the endothelial lining of the heart, or EndotJielherz of German embryol- ogists. The mesothelium of each side meets its fellow in the median ventral line, forming a thin partition or ventral mesocardium. Fig. 129 A, which subsequently breaks through; from the ventral wall of Fig 129A.— Embryo Chick (Minot Coll. No. A J, Section 384) ; Section through the Anlage of the Heart. Md^ Medullary groove; Ec, ectoderm; -mes, mesenchyma; Am. ves, amniotic vesicle; PA, pharynx ; msth., mesothelium ; Endo^ cells to form the endothelium of the heart. the Vorderdarm, Ph, the mesothelium bulges out as a much-thick- ened layer, msi.^, which develops into the muscular wall of the heart, while between this wall and the entoderm of the Vorderdarm lie the mesenchymal cells. Development proceeds by the mesothelial fold becoming more protuberant on each side, and the mesenchymal cells assuming the endothelial character, coming to bound several irregular cavities on each side. Fig. 130, En.htj these cavities soon fuse into two main cavities running longitudinally ; as the two cavities enlarge they meet in the median line and remain separated at first by a wall of two layers of endothelium; this wall soon breaks through and ORIGIN OF THE HEART. 227 there results a single median tube of endothelium connected, by long processes of cells, across quite a wide space with the mesothelium. Excellent figures of all these changes are given by Gasser, 77.3. The heart is now a double tube, connected by the mesothelium with the tissues above and below ; but soon the connection on the ventral side is severed, and a little later that on the dorsal, but the attach- ments are retained as in amphibia at both ends of the tube. A sec- tion through the end of the heart is shown in Fig. 130; the ventral Fig. 130.— Chick Embryo CMinot Coll. No. AL, section 119). Md, Wall of medullary tube; nch, notochord; visth, mesothelium; P/i, pharynx; en.ht, endothelial heart: m./if, muscular heart. mesocardium is entirely lost ; the dorsal is preserved, as also at the opposite end of the heart, though not in its middle ; the thick meso- thelial wall or muscular heart is widely removed from the thin inner endothelial heart (Endothelherz) . From the preceding account it appears that, owing to the devel- opment of the heart beginning before the Vorderdarm closes, the heart is distinctly double in origin, though all trace of the duplex condition is quickly lost. In mammals the double stage lasts longer, the Vorderdarm being closed still later. Our knowledge of the origin of the heart in mammals rests chiefly on the observations of KoUiker upon rabbits; this paragraph is therefore based on the description given in KoUiker 's "Grundriss," p. 96, 120. Traces of the heart can be recognized in embryos with five protovertebratBB, and the two anlages are well advanced in em- bryos with eight to ten segments, and in surface views, Fig. 114, may be seen at either side of the head, bending anteriorly toward the median line, and each connected posteriorly with the developing omphalo-mesaraic vein of the same side ; one can also distinguish the parietal coelomatic cavity about the heart. A transverse section through the region of the heart presents a very uniform picture in all mammals thus far studied ; compare Fig. 95 of the opossum with Fig. 114 of the rabbit. The parietal ccelom or amnio-cardial vesicle 328 THE EMBRYO. is small as compared with that of the chick, Fig. 117, and lies quite distant from the median line; the splanchnic mesothelium f orms a large fold, which projects into and nearly fills up the ccelomatic cavity; this fold forms, as in the chick, one-half of the muscular heart ; in the interior of this fold lies the endothelial heart, which sends out processes by which it is connected with the surrounding mesothelium. By the bending down of the layers and the expansion of the coelom the Vorderdarm is shut off and the two lateral heart anlages are brought together in the median line below the Vorder- darm, and there fuse into a single structure ; the fusion takes place in such a manner that the two mesothelial folds unite by their edges to form a single thick tubular wall around the double endothelial heart ; it is not long, however, before th6 two endothelial tubes also fuse into one. As in the chick the two mesothelia, when the median heart arises, form a membrane (mesocardium) , by which the heart is attached to the tissues above and below ; both mesocardial mem- branes break through, putting the two ccelomatic cavities into com- munication and leaving the tubular heart suspended by its ends. In amniota the heart arises from a double anlage, which by the bending down of the splanchnopleure of the Vorderdarm becomes a single median anlage, as in amphibians; C. K. Hoffmann, 84.3, has asserted that in snakes the heart arises from one of the lateral anlages, but Junglow, 89.1, has rendered it probable that this is merely a blunder of observation. The median heart is at first a nearly straight tube attached by each end to the wall of the pericar- dial ccfilom, and connected in front with the aortae and behind with the omphalo-mesaraic veins ; the tube is double, consisting of a thin inner endothelial wall of mesenchymal origin separated by a consid- erable space from the outer thicker mesothelial layer, from which the muscular tissue of the heart arises. Origin of the Endothelium of the Heart. — This is still unsettled. As we have seen, the endothelium has upon its first ap- pearance nothing of an endothelial character, but resembles instead the cells of the mesenchyma at the time ; in amphibia they are large and rounded and charged with yolk granules; in amniota they are more like embryonic connecti'v e-tissue cells. These cells always appear between the entoderm of the cervical archenteron (Vorderdarm of Von Baer) and the mesoderm bounding the coelom, and when they first appear there are no other cells near them between the mesothe- lium and entoderm, compare Figs. 128 and 129. Whence do these cells come ? I consider it probable that they are the forward extension of the vascular anlages of the omphalo-mesaraic veins and that just as the endothelial aortse are formed by the ingrowth of loose strings of cells so are the two veins, and these uniting in the median line form the endothelial heart. This view is hypothetical. A variety of other conflicting views have been advanced, of which the follow- ing may be noted. Balfour, " Elements," 85, 89, thinks the cells come from the neighboring mesoblast, as Oellacher had previously consid- ered was probable in teleosts, 73. 1, 84. Goette has maintained that in Petromyzon, 90.1, teleosts and amphibians, 75.1, the cells come* directly from the entoderm, and C. K. Hoffmann, 92.1, maintains the origin of the heart to be entodermal in elasmobranchs. Rabl, ORIGIN OF THE HEART. 239 86.1, expresses himself very cautiously, but inclines to the view that the cells come from the entoderm, and in regard to the sharks he is uncertain, 89.2, 225. J. Riickert, 88.2, believes that the cells which become the endothelium are thrown off in elasmobranchs from both the entoderm and mesoderm at the points where the cells first appear. Finally, F. Schwink, 90.1, asserts that in amphibia the cells are derived neither from the neighboring entoderm nor mesoderm, but that they grow in from the mass of yolk-cells. Schwink's observations seem very careful, and may turn out to con- firm the hypothesis of the origin of the endothelial heart from the omphalo-mesaraic veins uniting. Origin of the Vascular System. — O. Biitschli, 83.3, has ad- vanced an hypothesis of the phjdogenetic origin of the heart and blood-vessels which has much plausibility. He suggests that the heart is a remnant of the primitive or segmentation cavity of the embryo, and is not derived from the secondary or permanent body cavity (schizocoele or enterccele) . He endeavors to reconcile this view with the accounts of the development of the heart in vertebrates, maintaining that it probably arises as a fissure in the mesoderm, remaining as a permanent part from the temporary primitive cavity. More support for the hypothesis is found in arthropods ; for it has been observed in several forms that the two edges of the mesoderm approach one another in the median dorsal line, leaving a space be- tween them which belongs to the primitive cavity. This space becomes the heart. Sometimes it is cut off before, sometimes after, the mesoderm is split into segments. These observations were upon the bee (Biitschli), Geopliilus (Metschinkoff) , and Branchipus (Claus) . An investigation to answer the problem propounded by Butschli would, it may be safely said, prove fruitful and interesting. For further speculations in this direction see Schimkevitsch, 85.1. As to the evolution of the vascular system the course of develop- ment in the embryo indicates, it seems to me, that the immediate ancestors of vertebrates had no capillary vessels, but only a few large afferent and efferent trunks with a few anastomoses, as is now found in many annelids. With the acquisition of the large yolk the devel- opmeiit of accessory blood-channels over the surface of the yolk pre- sumably followed to secure, more efficiently, nutrition for the embryo. These first channels were, if we may rely on the ontogenetic indi- cations, grooves on the surface of the j'olk bounded ou one side by mesenchymal cells, by the further differentiation of which the grooves become endotlaelial tubes ; in this manner we can account for the blood-vessels appearing first in the extra-embryonic area. Since the blood-cells are developed from the walls of the vessels, it is possible that the walls may have acquired haemoglobin, and the cells then have been set free by a further evolution, but it is perhaps equally possible that the isolation of the blood-cells from their matrix (the vascular wall) may have preceded the acquisition of haemoglobin. CHAPTER XI. ORIGIN OF THE UROGENITAL SYSTEM. The outlines of vertebrate morphology were given, in the main, correctly by the older anatomists, except as regards the urogenital system. In 1875 Carl Semper announced the discovery that the " excretory tubules of elasmobranchs have a funnel-shaped opening in the abdominal cavity — a fact discovered by Balfour, 78.3, at about the same time. Both authors recognized that this discovery was profoundly significant, but it is chiefly to Semper that we owe the reform of conceptions in this field. It is unnecessary to attempt a historical review ; the reader will find in Max Fiirbringer's admi- rable monograph, 7 8. 1, a thorough, critical, and trustworthy revision of all that had been done up to that time. For notices of the subse- quent literature see Riickert, 88.1, Van Wijhe, 89.1, and H. H. Field, 91.1. R. Semon's valuable memoir, 91.1, became accessible to me too late to enable me to remodel this chapter as his results render necessary. Fundamental Parts of tlie Urogenital System. — For a general explanatory description we may consider the fundamental parts to be four on each side of the vertebrate embryo, compare Fig. 131. The four parts are two lon- gitudinal ducts: the pronephric or Wolffian duct, W. D, and the MuUerian duct or o^aduct, M. D; and two ridges on the dorsal side of the body-cavity, Coe, into which they protrude ; each ridge is covered by mesothelium resting on mesenchyma. The smaller ridge, Gen, is called the genital, since it is transformed into the genital glands ; it lies nearest the median line; its cephalic end is probably identical with the so-called glomus of the prone- phros. The larger ridge, Ux, is called the Wolffian or nephridial ridge; it contains the transverse excretory tubules (segmental tu- bules, nephridia) which are de- veloped from the nephrotomes, the expansion of which probably causes the bulging of the mesothelium, which results in the forma- tion of the Wolffian ridge. The nephridia open into the pronephric •W.D M.D ■Msftj Fie. 131.— Diagrammatic Cross-Section of a Vertebrate to show the Fundamental Efilations ot the Urogenital System. Md, Medullary tube ; Nch^ notochord ; Ao, aorta ; Gen^ genital ridge; W. D, Wolffian duct; M. D, Mailer's duet; Ex, excretory or Wolffian ridge; Msth, mesothelium ; Coe, coelom ; Som, somatopleure : Ach, archenteron. ORIGIN OP THE UROGENITAL SYSTEM. 231 duct. The cephalic end of the nephridial or Wolffian ridge give rise to the pronephros, while the remainder of the ridge is for the chief part at least converted into the Wolffian body (primitive kidney, mesonephros, Urniere) . Head-kidney or Pronephros. — The head-kidney being the first part of the urogenital system to be differentiated in the verte- brate embryo, must be regarded as the phylogeneticaUy oldest part. It is found in the embryos of (probably) all vertebrates, but disap- pears before adult-life in selachians, some teleosts, and all amniota. The head-kidney is always situated in the segments immediately behind the heart, and is a paired organ with a longitudinal duct, which finally opens into the cloaca or hind end of the alimentary tract ; the duct has great morphological importance ; its development is described in the next section. The head-kidney consists of from one to five or more transverse tubules which are differentiated from the nephrotomes and have on the one hand an opening into the ven- tral coelom or abdominal cavity, and on the other into the longitu- dinal duct. Each tubule consists of epithelium and when well developed takes a convulated course. The number of these trans- verse tubules is said to be greatest in Myxine ; in Petromyzon there are four or five, in Torpedo six, Pristiurus four, Coecilia four, An- ura three, Urodela two; but in teleosts and cartilaginous ganoids one only. The head-kidney often protrudes somewhat into the body- cavity, and the part of the body-cavity into which it protrudes may become, as in teleosts and the lamprey, shut off from the remaining coelom. There is also developed a so-called glomus, which is a fold of the mesothelium arising near the base of the mesentery, and con- FiG. 132.— Eana Temporaria. Tadpole of 13 mm. Cross-Section through the pronephros. nch, Notochord; m, muscles; /, (unnel-shaped opening of tubule or second nephrotome; V, blood-vessels ; Ec, ectoderm ; t, tubule of pronephros ; gl, glomus ; Lu, lung. After M. Furbrln- ger. X 90 diams. taining numerous blood-vessels. The structure of the organ is well illustrated by Fig. 132. The development of the head-kidney varies considerably in the different classes of vertebrates, so that we are still uncertain as to what are the essential and typical features of its development. The 332 THE EMBRYO. confusion is probably due to the fact that it is only recently that we have gained the knowledge that between the myotome and the lateral plate comes, in every segment, the nephrotome, to which the origin of the transverse excretory tubule, both of the head-kidney and of the Wolffian body (mesonephros) has been traced in a number of cases. Since we have known that the essential question is, Do the pronephric tubules arise from the nephrotomes? sufficient investi- gations have not been undertaken. But it has been shoAvn in several cases that the nephrotomes do produce the tubules. The typical mode of development both for the pronephros and mesonephros is, I think, probably as follows: The nephrotomes typically contain a coelomatic cavity ; when they separate from the myotome the myo- tomic end of the nephrotomic cavity becomes closed, but the other end remains open and becomes the permanent nephrotome opening of the nephric tubules (Segmentalorgane of Semper) ; the nephrotome now lengthens out and unites secondarily with the pronephric or segmental duct. Until further research of a far more thorough character than anything we yet have shall decide the question, this hypothetical account is the best that can be presented. In the cylostomes, teleosts, and amphibia the pronephros is said to arise from the mesothelium of the ventral coelom ; but as this takes place so that the mesothelium is close to the myotome, it is more than possible that we have to do really with nephrotomic tissue. If Goette's account of the process in Petromyzon, 90.1, 54, 55, be correct, then it may be that in the lamprey the nephrotomic anlage separates from the myotomes, and while still connected with the lateral plates undergoes segmental division. In the lamprey (Goette, I. c), teleosts (Mcintosh and Prince, 90.1, 783-785), and amphib- ians (Fiirbringer, 78.1), the mesothelium, which produces the tubules, produces the longitudinal duct also, but in view of what is stated of other vertebrates this has been questioned. Our knowledge of t;he head-kidney in amphibia has been very much extended by the recent researches of Mollier, 90.1, Marshall and Bles, 90.1, R. Semon, 91.1, and H. H. Field, 91.1. My lack, of personal famil- iarity with the amphibian pronephros makes me unwilling to attempt a critical summary of their researches. , The pronephros of elasmobranchs begins to develop in (Pristiurus) embryos with twenty-seven segments ; the three foremost segments are subsequently included in the head, so the fourth is the first seg- ment of the rump (Van Wijhe, 89. 1, 473) . In the first four qoti End of a Sheep Embryo of Sixteen Days. After allantOlS m tUe CmCK: naS DCen R. Bonnet. 4mm, Amnion; a. m, anal membrane; studied in detail by E. GaSSer, jyr. s, primitive streak ; En^ entoderm ; Ach., ,„-- ,, t-. t>vi • archenteric cavity, or archenteron ; All, a\\a.u- 74.1, and by r. VOn Uobrymn, toic diverticulnm ; Mes, mesoderm. ^ j _ ^ _ j^ ^^^.^^ ^^^^^ ^^^ ^^^^ membrane is formed as a small pouch extending upward into the hind end of the primitive streak; the tip of this pouch lies just behind the bottom of the furrow, which marks off the caudal ex- tremity of the embryo ; the bottom of this furrow is the site of the anal plate; the pouch gradually enlarges and assumes the dipper shape, very much as in the sheep. Fig. 145, All; in the chick, how- ever, this stage is reached relatively later than in mammals, for in the chick we find the tail already far advanced, so that it not only projects freely but has begun to curl over downward so as to bring the allantois and anus on to the ventral side of the body as well as to cause the formation of the enddarm, which is a short extension of the archenteron into the caudal extremity*. The whole series of metamorphoses is admirably illustrated by Gasser, I. c, Taf. I. In mammals the formation of the tail is somewhat retarded, but in them also it results in curling over and so bringing both the allan- tois and the anal plate on to the ventral side, with the further result that the allantois now comes to lie headward of the anal plate, although before the curling over it lay behind it. It is important to note that the amnion arises between the anal plate and the allantois, and, as shown in Fig. 145, fuses with the wall of the allantois. The allantois is characterized by the rapid development of its mesoderm, which seems to be derived from the middle cells of the hind end of the primitive streak. The amount of the allantoic meso- derm is subject to much variation during the early stages of the organ, being much greater in mammals than in birds, so far as ob- served. The mesoderm is particularly conspicuous in rodents; in the rabbit it makes a distinct mound, compare Fig. 196; in the guinea-pig (E. Selenka, 84.1, Taf. XI.) it acquires an excessive size, becoming larger than all the rest of the embryo ; in Mus the precocious development is almost equally marked ; it is into this mass of mesoderm that the allantoic di"verticulum of the archenteron grows. In the opossum (Selenka, 87.1, 139) the amount of meso- derm is more nearly as in the rabbit. The mesoderm is also charac- terized in rodents, and perhaps in other mammals, by its precocious EARLY DEVELOPMENT OF THE ARCHBNTERON. 359 vascularization, which has been expressly emphasized for the rabbit by C. Rabl, 89.2, 153, Taf. IX., Fig. 14; the vessels give the tissue a spongy character. The protuberance caused by the allantoic meso- derm is termed Allantoishocker by recent German writers, the Allantoiswulst by KoUiker. The earliest stages of the human allantois are not known. There has been some discussion as to whether there is a free allantois, but no proof that such a stage occurs has been brought. The matter is discussed in the chapter on the youngest known human embryos, and in that on the umbilical cord, compare also Fr. Keibel, 91.4. Primitive Anus. — The terminal portion of the intestinal canal receives in early stages the urogenital ducts, a condition which is permanent in the Sauropsida; the portion of the archenteron com- mon to these ducts is known as the cloaca. The ectoderm in amniota forms very early a small anal invagination (proctodaeum) which grows in toward the cloaca until the ectoderm and entoderm come into contact; the membrane formed by the two epithelia breaks through and the cloaca thereby acquires an opening to the exterior ; this opening subsequently divides into two : 1, the urogenital open- ing ; 2, the permanent anus ; in distinction from the latter the clo- acal opening may be called ih.e primitive anus. In amniota the anal membrane arises in the anterior region of the primitive streak some distance behind the neurenteric or chorda canal. It has been studied in birds by Bornhaupt, 67.1, and more fully by Gasser, 80. 1. It has been noticed in Lacerta by H. Strahl, 86.2, 166, who states that it appears in that type at a much later stage than in birds or mammals. In mammals it was mentioned first. I think, by Kolliker, 83. 1, and has been since then studied by H. Strahl, 86.2, F. Keibel, 88.2, 410, R. Bonnet, 89.1, 90, Ket- terer, 90.2, Toumeux, 90.3, and especially by C. Giacomini, 88. 1, most of all these observations having been made on the rabbit. In rabbit embryos with five pairs of myotomes, the anal mem- brane can be distinctly recognized near the rear of the primi- tive streak, compare Strahl, I. c, Taf. IV., Fig. 6; it begins as a slight depression of the ectoderm ; behind it are situated the amniotic fold and the allantois ; the depression rapidly deepens, pushing away the mesodermic cells until the ectoderm comes into contact with the entoderm, which at this spot becomes, meanwhile, thickened into a cylinder epithelium ; when the contact takes place a slight entodermic depression appears. The two layers soon become indistinguishable, and by the proliferation of their cells produce a cord of cells; a simi- lar cord has been found in the sheep by R. Bonnet, 88. 1, and in the guinea-pig by F. Keibel, 88.2; the latter states that the cord is con- nected only with the ectoderm ; the cord is completely surrounded by typical primitive-streak tissue; according to Giacomini, 88.1, 287, the cord develops very soon a transient lumen, which he calls the "anal canal." While during the further development of the embryo the caudal extremity is rolled over ventralward, the cord changes in character, becoming a membrane, and at the same time it is brought on to the ventral side of the body and comes to lie be- hind, instead of in front, of the amnion as it did before the rolling Tip of the embryo. The change just referred to consists in rendering 2G0 THE EMBRYO. the two epithelia distinct again and converting each into a single cell-layer, making a double epithelial membrane from which the mesoderm is entirely excluded, and which has been appropriately named the anal membrane by Strahl. The membrane lies at the bottom of a shallow pit, which is commonly viewed as an ectodermal invagination, and has been called the Afterdarm by German, procto- daeum by some English writers. It is to be regarded as the rudi- mentary homologue of the well-developed invagination of annelids and other invertebrates, which forms in them a considerable portion of the digestive canal ; the anal invagination results in invertebrates in the formation of the so-called Hinterdarm (hind-gut of Foster and Balfour) , which must not be confused with the vertebrate Hin- terdarm, which is derived from the archenteron. The rupture of the anal membrane is said to occur in the rabbit about the twelfth (Kolliker, "Grundriss," 359) or thirteenth day (Strahl, 86.2, 165). I know of no exact description of the process in mam- mals. In the chick the epithelial cord arises and becomes perfo- rated, according to Gasser, without passing into the stage of anal membrane observed in mammals ; irregular cavities appear in the c'ord (Gasser, I. c, Taf. XIII., Figs. 6a, 7a); these cavities enlarge and fuse, the cells of the cord or plate meanwhile undergoing degen- erative changes ; the rupture is completed about the fifteenth day of incubation. The anal ectodermal invagination is somewhat more marked than in mammals and gives rise on its dorsal side to a con- siderable diverticulum, the bursa Fabricii, which is found in birds but not in mammals or reptiles. The anus of the love?' vertebrates arises, as has already been shown, in intimate relation with the blastopore. This fact was first discovered by Max Schultze, 56.1, in Petromyzon, ascertained in aiytes by Gasser, 82.3, in the newt by Alice Johnson, 84.1, in Rana by Durham, 86.1. The nature of this relation was first eluci- dated by Schanz, 87.1, and has since been worked out for various amphibia, as described, p. 189. Th3 Enddarm. — The prolongation of the archenteron into the tail of amniote embryos is generally known as the Enddarm, the German name most in use; it is also called Schvxmzdarm, tail-gut, and post-anal gut. It results from the differentiation and rolling over of the tail. The tail is produced by the growth of the tissue of the primitive streak between the anal membrane and the blastopore or neurenteric canal, compare Chapter XIII. ; the growth occurs in such a way that the tissue curls downward, and folds off the region of the archenteron underlying the primitive streak, and the disposi- tion becomes as shown in Fig. 102 of Kolliker's "' Grundriss," 3te Aufl., the enddarm extending into the tail behind the ventrallj"- situ- ated anal membrane. I consider the enddarm co be distinct from the neurenteric canal, with which Balfour ("Comp. Embryol.," II., 193, Fig. 124) brings it into relation. 0. Hertwig apparently agrees with Balfour, since he copies the latter's diagram ("Entwickelungsges.," 3te Aufl., Fig. 126). It seems to be confined to early embryonic life, but there are a few data as to its ultimate fate. Prenant, 91.2, 231, studying the rabbit found the post-anal gut to be a short wide pouch before the tail EARLY DEVELOPMENT OP THE ARCHENTERON. 2G1 develops ; as the tail develops, the gut extends into it and becomes long and narrow, and its posterior extremity merges with the fused anlages of the medullary tube and notochord. In still older em- bryos it degenerates. Origin of the Vorderdarm.— As is well known the first part of the embryo in vertebrates to project from the yolk is the head end. In the same measure as the head and neck become free the portion of the archenteron which pertains to them becomes closed below and shut off from the yolk. A longitudinal section of a chick in which the head has just become free is shown in Fig. 146. In consequence of the head end, H, having grown forward above the proamnion, pro. a, which overlies the extraembryonic archenteric cavity, it has become free on all sides, and at the same time the archenteron has been carried forward with the head, making the so-called Vorder- darm, Yd, of German authors. The term fore-gut has been proposed by Foster and Balfour as an equivalent English term, but has not come into general use, so I have prefered to use the German term. Vorderdarm is also used in invertebrate embryology, but in a dif- ferent sense, for it designates the oral invagination of the ectoderm, Fig. 146.— Longitudinal Median Section of Youns Chick Embryo. H, Head ; Vd, vorderclai-m ; mes, meso lerm ; /o, fovea cardiaca; 2Ji pericardial cavity; pro. a, proamnion ; ^Ic/i, archfa- teric cavity; Pr. s, primitive strealc. whereas the vertebrate vorderdarm is the cephalic portion of the archenteron. Even at the stage of Fig. 146, the vorderdarm has begun to be differentiated into an anterior division and a posterior, which at this time are distinguished chiefly by the coelom, p, being present only in the mesoderm below the posterior division. The anterior division forms the pharynx proper. The distinction between the two parts of the vorderdarm has long been recognized (see, for example, Goette's observations on Bombinator, 75.1, 221), but its morphological sig- nificance has been overlooked. The vorderdarm is a short canal under the anterior end of the medullary groove ; it ends blindly in front, but opens widely behind into the general archenteric cavity ; this opening is termed the /ouea cardiaca {rordere Darmpforte of KoUiker), having been so named by C. F. Wolff. The fovea is easily seen, when the chick embryo is removed from the 3'olk in the usual manner, and viewed from the under side ; its curving edge marks the end of the closed archenteron behind which the archenteric cavity of the embryo opens directly into the yolk-sac. In transverse sec- tions. Fig. 147, the vorderdarm appears widely expanded sideways, but compressed dorso-ventrally, and also bent, the concavity being upward ; it is, of course, completely lined by entoderm, the cells of which form a very thin layer on the dorsal side and a much thicker layer on the ventral side ; moreover, on the ventral side the entoderm is thickened toward the median line. These features are highly characteristic, but their significance is quite unknown. Are they ancestral in origin? 263 THE EMBRYO. In the explanations usually given, the development of the vorder- (larm is not attributed to the forward grov^th of the head, but to the down-folding of the splanchnopleures. Indeed if sections of succes- sive stages be compared the idea appears justified , for at first the cephalic archenteron opens widely into the yolk-sac, then as the head of the embryo begins to rise up and project forward from the yolk it seems as if the sides of the head were being tucked under ; but if it Tich. Fig. 147. — Transverse Section of the Head of a Chick Embryo with seven Segments, nch^ Noto- chord; 01, ganglion; md, medullary wall; mes, mesoderm; Ph, pharynx; Ec, ectoderm; En, entoderm ; Am. V. , amniotic vesicle. be remembered that the head is growing and that the opening be- tween the archenteron proper and the yolk enlarges very little, it will be clear that the growth of the head is the real cause of the formation of the vorderdarm. In mammals the process is the same as in birds, but the vorder- darm is less expanded laterally and less compressed dorso-ventrally than in the chick, hence the appearance in cross-section is somewhat different. In the opossum, however, there is a marked resemblance to the avian type in the shape of the vorderdarm, see Selenka, 86. 1, Taf. XXII., Figs. 9-10, and it is probable that more careful study will show that the mammalian vorderdarm passes through the flat- tened form before assuming its more familiar shape. The Oral Plate. — The fact that the anterior end of the vorder- darm lies against the ectoderm has long been known for advanced embryos. The two germ-layers, entoderm and ectoderm are soldered together with no mesoderm between them, thus forming a double epithelial plate (as shown in Figs. 106 and 170, o.pT), which separates the buccal from the archenteric cavity. The plate, which may be called the oral plate (membrana fauces, Eachenhaut, Mundrachen- haut) , by its subsequent rupture brings the mouth into communica- tion with the pharynx. Fr. Carius, 88.1, 22, has shown that the oral plate is present in the rabbit at a very early stage, the spot where the entoderm and ectoderm come into contact being distinguishable before the head is separated from the yolk. This spot lies just in front of the interior end of the medullary groove and of the chorda, the end of which EARLY DEVELOPMENT OP THE ABCHENTERON. 363 fuses with the entoderm of the membrane. As the head of the em- bryo grows forward and bends downward toward the yolk the oral plate is rolled over so as to lie on the ventral side of the embryo, and to constitute part of the ventral floor of the vorderdarm as shown in Fig. 106. Origin, of tlie Pharynx. — From what has been said in the pre- ceding section it appears that the vorderdarm very early divides into an anterior part without any splanchnocoele in the surrounding mesoderm and a posterior part, underneath which lies the pericardial division of the ccelom. The anterior division becomes the pharynx proper and is remarkable for its rapid enlargement during the earliest embryonic periods of amniota ; the large size of the pharynx is char- acteristic of the lower vertebrates, hence we have in the pharynx another illustration of the appearance in the embryo of a higher form of features characteristic of the adult lower forms. The posterior or epicardial division of the vorderdarm undergoes differentiation later than the pharynx, but ultimately gives rise to the oesophagus and stomach ; as the lungs arise near the junction of the two divi- sions, it is not quite certain, at present, whether they make part of the anterior or posterior division. The pharynx then is the anterior portion of the vorderdarm, and is further characterized by never having a continuous coelomatic cavity developed in the mesoderm surrounding it. The relations of the pharyngeal entoderm to the ectoderm are ex- tremely important to the morphologist, since they result in the formation, 1, of the oral plate and consequently of the mouth cavity; 2, of the gill-clefts, which in their turn determine to a large extent the complex morphology of the head. The Branchial Clefts, or gill-clefts, are permanent structures in the fishes and tailed amphibia, larval structures in anoura, and embryonic structures in amniota. They arise as a series of paired pouches from the sides of the pharynx. They are called Schlund-, Kiemen- or Visceral-spalten in German; fentes branchiales in French. The number of gill-clefts varies in the different classes of verte- brates. In mammals and birds there are four ; in reptiles, tailed am- phibians, and most fishes, five ; among the selachians, however, the number is variable, there being often six and in the Notidanidae eight, it is said. In the lamprey there are eight during larval life, but the first aborts when the larva (Ammoccetes) changes into the adult (Petromyzon) . In Amphioxus the pharynx has eighty to one hun- dred openings and even more. These facts have led to the general conclusion that within the vertebrate series the number of gill-clefts has been gradually reduced-^a hypothesis of great importance, from its bearing upon the solution of the morphology of the head. In all birds and mammals there are four pairs of gill pouches developed, all in essentially the same manner. The anterior pair appears first, the others in succession behind it. The entoderm of the pharynx forms a small outgrowth on each side, making a pouch, which expands until it reaches the ectoderm. Soon a second pair of outgrowths appear behind the first, and a third and a fourth. For a long time it was believed that the membrane formed by the ento- 264 THE EMBRYO. derm and ectoderm at the end of each pouch ruptured and converted each pouch into an actual cleft or opening by which a free passage was established through the side of the neck into the pharynx, as occurs in all Ichthyopsida. W. His pointed out, 81.1, 319, that this was open to question, and later showed that the membrane is not ruptured in birds and mammals — a conclusion which has since been confirmed by Born, 83.1, 275, KoUiker, "Grundriss," p. 77, and Piersol, and which is, I think, probably correct, for those who have called it in question (De Meuron, Kastschenko, and Liessner) seem to me to have offered insufficient evidence. Piersol, 88.1,162, studied the question with great care in the rabbit, and finds no sat- isfactory evidence of the closing membrane being ruptured in any of the branchial clefts at any time. The shape of the pharynx and its four pairs of branchial pouches has been carefully studied in the rabbit by G. A. Piersol, 88.1, by means of models of the cavity at various ages, constructed in wax by Bern's method. Two views of the model or cast of the pharyn- geal cavity at eleven days are given in Fig. 148. As the oral plate is already ruptured at this age, the buccal and pharyngeal cavities Fig. 148.— Two Views of a Wax Model of tlie Cavity of the Pharynx of a Babbit Embryo of Eleven Days. A, Showing the lateral and ventral surface ; B, showing dorsal and lateral surface. After Piersol. have fused, and the models shoAs^ also the oral evagination of the hypophysis, hy. The figures sufficiently indicate the complex configu- ration of the pouches with their wing-like expansions and ascending dorsal points, as well as the progressive diminution in size from the first to the fourth pouch. It must be borne in mind that while the gill-slits are developing the head is growing, and therefore lengthening, so that the pharyn- geal portion of the vorderdarm elongates. At the time the first gill- cleft is formed there is not room for the remaining clefts, but the growth of the pharynx provides the needed room soon. Thus in the chick there is at first only a very small distance between the region of the pericardium (and heart) and the anterior extremity of the embryo (see Fig. 146), but by the end of the third day there is a con- siderable interval between the anterior end of the heart and the actual head. This interval constitutes the embryonic neck, and cor- EARLY DBVKLOPMENT OF THE ARCHENTEKOX. 265 Tesponds to the pharyngeal region, and is characterized by two prin- cipal features: 1, the absence of a splanchnoccBle ; 3, the presence of the gill pouches. As soon as the pharyngeal evaginations reach the ectoderm they become attached to it, first on the dorsal side and then downward until the attachment is completed throughout the whole area of con- tact (A. Goette, 75.1, ^2 3). It seems now as if the ectoderm were actually held down where resting upon the entoderm, for we see as the next phase that the germ-layers grow freely in front and behind each gill pouch, thus producing columns, which are placed at the side of the pharynx and are separated from one another by the gill- clefts. As there are four gill-clefts it follows that there are five columns. These columns are known as the branchial arches, also as the gill or visceral arches {Kiemenbogen, Visceralbogen, arcs hranchiaux) . Each arch is marked out by projecting into the pha- rynx and upon the outside, and consequently soon after the gill pouches are developed the arches become easily distinguishable upon the exterior, and the depressions between them show the positions of the pouches. The depressions become part of the gill-clefts when the membrane (ectoderm and entoderm) breaks through; hence, when the clefts become, as in the lower vertebrates, open passages, their lining is partly of entodermic, partly of ectodermic origin, but as the epithelia fuse perfectly, the line of demarcation cannot be dis- tinguished in the open clefts. As to the time at which the gill-clefts appear, we need more exact information. C. Rabl, 89.2, 216, gives the following data for sela- chian embryos (Pristiurus) : Embryos with 18 myotomes show the first gill pouch. " 23-21 " the second pouch beginning. 26-27 " the second pouch well formed. '■ 31-32 " the third pouch well advanced. '■ 38-40 "' the fourth pouch beginning. " 45-16 " the fourth pouch completed, and the second breaking through. " 54-59 " the fifth pouch begun, and the first and third breaking through. " 66-68 " the filrst, second, and third pouches are clefts, the fourth is breaking through. " 74 " the sixth pouch is forming, the first four are open, the fifth opening. In the chick the gill-clefts begin to appear with third day, the fourth being present at the end of that day. In the rabbit the first pouch is seen the ninth day, and the fourth the tenth day. In man the pouches are developed during the beginning of the third week. The pharynx expands rapidly in all directions during the develop- ment of the branchial clefts, and there is a corresponding enlargement of the cervical region, whereby the form of the embryo is affected. The external features resulting from the development of the pharynx are described in Chapter XIII., to which the reader is referred. It may, however, help to make the fundamental relations of the pha- rynx clear, to insert here the figure of a longitudinal horizontal sec- 266 THE EMBRYO. tion of a dog-fish embryo. The pharynx is a very wide cavity, Pli, the sides of which are bounded by the five gill-arches; the gill-clefts behind each of the arches are already 6pen through ; the space in front of the first arch, J, is part of the opening of the mouth, which came into communication with the pharynx at a much earlier stage than that represented in the figure. The size of the pharynx forms a striking contrast with that of the intes- tinal canal. In; each branchial arch con- sists of a mass of connective tissue bounded by a layer of epithelium derived partly from the entoderm of the pharynx, partly from the ectoderm. The shapes and positions of the gill- slits are remarkably uniform in all verte- FiG. 149. — Acanthias Embryo of IT mni. Horizontal section of tlie anterior half. 3//>, Mid-brain; Of, otocyst; C, cochlea or lagena ; Ph, pharynx; i, H., III., IV,, V., gill- arches; Ht^ heart; Fe, vein; Ja, intestine. Li, liver. Fig. 150.— Chicken Embryo of Sixty-eight Hours. Ar, Vitelline artery; V. vitelline vein; iS, segment; Ao, aorta; or^, third branchial cleft; Of, otocyst; fib, hind brain ; JV/7), mid-brain; L. lens; H, hemisphere ; Hi, heart. brates. They are elongated dorso-ventrally and narrow in the di- rection of the longitudinal axis of the embryo. Fig. 150. The first is the largest and the remaining ones gradually diminish from in front backward. Viewed from the outside they are seen not to be strictly parallel, but to converge somewhat toward the ventral side, the angle between the first and second clefts being the largest. It EARLY DEVELOPMENT OP THE ARCHENTBEON. 367 Fig. 151.— Aoanthi- as of 17 mm. mb. Mid-brain; N, nasal pit; Afa;,maxilla; Jf, is also noteworthy that the lower edges of the clefts recede further and further from the median ventral line from the first to the last cleft, Fig. 151; the first clefts nearly meet on the ventral side, while the fourth and fifth clefts are far apart. The observation of this peculiarity has led to the supposition that the mouth may have been evolved by the meeting of two gill-clefts which have fused into one opening on the median line ; this hy- pothesis is discussed in the section on the evolution of the mouth. The Branchial Arches. — These are structures of great morphological importance, which undergo modifications of increasing complexity as we ascend the vertebrate series. They are also termed gill- arches and visceral arches {Kieinenbogen, Visceral- bogen) . In their earliest form they are merely the columns of tissue bordering the gill-clefts; in a horizontal section of the pharynx of an embryo they are cut transversely and are then seen to consist merely of a core of mesenchyma, surrounded by a layer of cylinder epithelium, derived in part from the ectoderm, in part from the entoderm, as explained above. In those cases where, as in the amniota, the gill-clefts do not become open, of course the ecto- mouth : Md, mdndi- derm from one arch passes across to the next, and the wf 'heartT"Yi''°yof^ entoderm likewise, but not the mesoderm, compare ^'*"^- Fig. 358. As previously stated the inner and outer layers together form a membrane ( Verschlussplatte) , which closes the gill-cleft. In more advanced stages additional parts are gradually differenti- ated in each gill-arch. Typically there are four principal structures developed, an aortic vessel, a downgrowth of the myotome overlying the dorsal end of the arch, two branches of nerves, and a rod of car- tilage — and they appear in the order named. The aortic vessels arise very early and establish a direct communication between the ventral and dorsal aortae, and are called the aortic arches. Their arrangement and metamorphoses are discussed in Chapter XXIV. Fig. 152 shows the aortic arch, ^, in a cross section of a gill-arch. The parts have a typical primitive arrangement from which all modifications are derived. The details are discussed in subsequent chapters. Viewed externally the gill-arches present the following peculiarities in amniote embryos at the stage when the gill-arches have their maximum typical development. The first arch divides the mouth from the first branchial cleft, and has its lower end enlarged and somewhat knob-like ; the second arch has a similar knob, but a little smaller ; at first the four knobs are quite distinct, but they soon fuse and f Fig. 153. — Cross Section of a Branchial Arch of an Advanced Shark Embryo. Pristiurus. /, Branchial filament; ^, aortic arch; msf/t, mesothelium. N, nerve; cart, cartilage* c, communicating vein; %\ vein. After A. Dohrn. 268 THE EMBRYO. become more or less indistinct ; the third and fourth arches, on the contrary, simply thin out and melt into the general ventral surface. The anterior (cranial) border of the mouth, after the buccal cavity has formed, is also thickened and its upper end joins the dorsal end of the first branchial arch and hence is sometimes called the maxil- lary process (Oberkieferfortsatz) of the first arch. Additional data and figures of the external appearances are given in Chapter XXVI. Seessel's Pocket. — This term is applied to a small diverticulum which appears in birds and mammals on the dorsal side of the phar- ynx. It was first described by Seessel, 78.1, and has been noted since by various observers, Piersol, 88.1, et al. Origin of the Liver. — The liver in the primitive type of devel- opment, as preserved in Petromyzon and amphibia, appears exceed- ingly early, Fig. 153 (compare also A. Goette's figures 75.1, Taf. II., Figs. 34-38). It is a diverticulum of the archenteron, Fig. 153, Li, near its anterior extremity, and projecting on the ventral side downward into the mass of yolk-cells. The short stretch of the archenteron in front of the hepatic evagination is the homologue of the vorderdarm, which shows, however, in this type of development no trace, as yet, of its sub- Md ,>. nch sequent division into pha- ryngeal and epi cardial re- gions. When, however, the heart appears the two regions of the vorderdarm become distinguishable, and the liver diverticulum is seen to lie immediately be- hind the posterior or venous extremity of the heart. It is probable from these facts that the liver is an older organ in the ancestral his- FlQ. 163.— Longitudinal Section of an Embryo of Pe- ^r\r^J r\f -jT-ATfeViTa+oc +'haTi tViA tromyzon Planeri, Four Days Old, Reared at Naples. ^^^J ^^ N eiteuraieb man Tne Jlfd, Medullary tube; .Ec, ectoderm; 6i, blastopore; mcA, pharynX Or even the heart. notochord; o.wZ, oral plate; ii, liver. After Kuplier. rni "^ ., ,. p ji t ihe situation of the liver causes it to lie close to the veins, which are subsequently developed to pass from the yolk to the heart ; these veins are especially devel- oped in amniota and are known as the omphalo-mesaraic veins. The further development, to be described later, brings the liver into peculiar intimate relations with the venous circulation. In elasmobranchs (Balfour, "Works," I., 455) the liver arises dur- ing stage I (^. e., three gill-pouches begun, but the first not open yet) as a ventral outgrowth at the hind end of the vorderdarm and immedi- ately in front of the union of the yolk-sac with the archenteron, or in other words just in front of the yolk-duct or umbilical canal, thus bringing the liver into proximity with the vitelline veins entering the heart. As the gill-pouches are present the pharynx is already differentiated, and, therefore, the liver arises relatively later than in Petromyzon and the amphibians. " Almost as soon as it is formed this outgrowth develops two lateral diverticula, opening into a median canal. The two diverticula are the rudimentary lobes of the EAELY DEVELOPMENT OF THE ARCHENTERON. 269 liver, and the median duct is the rudiment of tiie common bile duct {ductus choledochus) and gall bladder. By stage K the hepatic diverticula have begun to bud out a number of small hollow knobs." In teleosts the liver arises quite late, e. g. , in trout the twenty- fifth day — as a solid outgrowth from the archenteron close behind the heart — thus offering one of the many instances of a solid growth in the embryo replacing a hollow growth. (Mcintosh and Prince, 90.1, 774, give their own and cite some previous observations.) In amniota the anlage of the liver arises in the same position as in the anamnia, but has the peculiarity of showing its bifurcation almost, if not quite, from the start, at least in birds and mammals. The two forks embrace between them the omphalo-mesaraic or vitelline veins just before they empty into the sinus venosus. In the chick the anlage appears between the fifty-fifth and sixtieth hour (Foster and Balfour, "Elements," 178, 179), the right fork being in all cases of greater length but less diameter than the left. In the rabbit (Uskow, 83.2, 220) the anlage appears during the tenth day, and on the eleventh sends out branches; according to KoUiker ("Grundriss," 372) only the left branch appears on the tenth day, the right on the day following. In man the anlage is well marked in embryos of three millimetres (His, 81.1, Taf. XI.^ fig. 7-8, also " Anat. Menschl. Embry.," Heft III., lG-17). Hishasshown, 81.1, 322-323, that the liver anlage is a long strip on the ventral side of the vorderdarm, and that when the vorderdarm is separated off from the yolk-sac the most ventral part of the entoderm of the vor- derdarm already shows traces of the hepatic differentiation. In front of and above the heart the vorderdarm is completely shut off from the rest of the archenteron (cavity of the future yolk-sac) , but imme- diately behind the heart the entoderm, as it passes from the vorder- darm around the edge of the fovea cardiaca, and so out on to the extra-embryonic region, is caught, so to speak, and forms the anlage of the liver, so that the liver is initiated not so much by a local growth of the entoderm as by retention of the downward extension of the layer, which results from the manner by which the embryo is separated from the yolk. The point is important as an illustration of the comparatively simple mechanical factors of development. Relation of the Liver to the Septum Transversum. — The tissue through which the vitelline veins pass to enter the heart forms a transverse partition, which divides the pericardial coelom from the abdominal coelom. This partition is the rudiment of the diaphragm, and has been named the septum transversum by W. His. It lies just behind the heart, and forms the ventral edge of the fovea cardi- aca, or opening of the vorderdarm into the general archenteron ; it is overlaid in the median line by the hind end of the vorderdarm, and consequently the anlage of the liver is situated in the dorsal median portion of the septum. As the great veins also pass through the septum to reach the heart, the hepatic anlage comes into imme- diate contact with the veins ; in their further development the veins and entodermal liver are closely connected, with the resulb of com- plex modifications in both parts. Comparison of Mammalian and Amphibian Archenteron. — For the convenience of students I have inserted the accompanying 270 THE EMBRYO. diagrams, Fig. 154, A and B. They are extremely conventionalized and may be considered especially inaccurate in that they fail to show the way in which the head (and with it the vorderdarm) projects forward, and in that the heart and liver are omitted. Emb is the axis of the embryo represented in nature by the medullary tube and notochord ; hi is the blastopore or neurenteric canal, behind which the anal opening or anal p._L A plate should be added , ;4^.-.^ -V were the diagram to be W- ^-^^ .,.^1 pii^.A'i:^^ "^v I completed. All is the infra-blastoporic diverti- culum or allantois; Ent is the cavity of the arch- enteron — the letters being placed where the archen- teron of the embryo prop- er passes into that of the yolk-sac ; br indicates the four gill-slits. The yolk- sac, Vi, is represented as enveloped in mesoderm, indicated by a shaded layer and lined by ento- derm which is indicated by a broad black line ; it must be remembered that in amphibians, A, the cavity is really filled with yolk-cells, which are represented in mammals, B, only by a layer of epi- thelial cells. Ch is the chorion, consisting of a layer of ectoderm indi- cated by the outside black line, and a layer of meso- derm, indicated by shad- ing. Between the chorion and the yolk-sac lies a space which is the extra- embryonic coelom. In amphibia this part of the in man it is developed very early com- it never extends more than Fig. 154.— Diagrams to Indicate the Fundamental Rela tions of the Arcnenteron. A, in amphibia; B, in mammals. For explanation of the letters see the text. coelom develops gradually pletely around the yolk-sac; in rabbits half-way round, and other variations occur in other mammals ; to suggest these differences in mammals the lower half of the yolk-sac in B is drawn with a dotted line only ; vt. is the vena terminalis. These diagrams suffice to show that the closest homologies exist between the two types, however much the actual proportions may differ. The primitive homologies of the archenteron hold true of all vertebrates. CHAPTER XIII. THE GERMINAL AREA, THE EMBRYO AND ITS APPENDAGES. I. The Germinal Area. Definition. — The germinal area {area genninativa, area embry- onalis, Keimhof, aire germinative) is that portion of the meroblastic vertebrate ovum in the centre of which the embryo is differentiated. It therefore comprises both the embryo proper and the region imme- diately surrounding it. It exists in all amniota, but of course in the higher mammals, owing to the loss of yolk in the ovum, the primitive relations are less clear than in Sauropsida. The area is further char- acterized by various gradually developed peculiarities, three of which deserve special mention. To take them in the order of their appear- ance, the three peculiarities are, first, the extension of the archenteric cavity under nearly the whole of the area ; second, the extension of the coelom over nearly the whole of the area ; third, the development of blood-vessels and blood beginning peripherally in the splanchnic leaf of the mesoderm and extending gradually into the embryo. 1. Extension of the Archenteric Cavity. — -As shown in the pre- vious chapter, only a small part of the archenteron of amniota is taken up into the embryo, and the rest of the cavity remains as the cavity of the yolk-sac, and therefore the entoderm of the area belongs, for the most part, to the future yolk-sac. As pointed out in the sec- tion on the entodermal cells in the preceding chapter, it is only on the upper side of the expanded archenteron that the entoderm be- comes distinctly cellular ; on the lower side the yolk is multinucleate, but liot divided into discrete cells ; at the edge of the expanded cavity the upper cellular layer passes gradually into the yolk and the region of the transition is known as the germinal ivall, the structure of which is discussed in the chapter on the yolk-sac. As previously pointed out, the cells very early assume two forms, becoming thin and flattened in the central region of the area, and remaining as long cylinder cells in the peripheral zone ; this difference results in a greater transparency in the central zone, which has accordingly received the name of area pellucida, while the peripheral zone, owing to its rela- tively great opacity, has been named the area opaca. Another result of the extension of the archenteron is that all the layers above it can be easily removed from the rest of the ovum, keeping their natural connections otherwise intact ; they form when thus removed a thin membrane, which, following the terminology of the older embryologists, we commonly speak of as the blastoderm ; compare the section on the meroblastic embryo, p. 128. 2. The extension of the cceloni of course divides the mesoderm into an upper (somatic) and lower (splanchnic) layer. But the divi- sion does not take place in certain definite regions, which are, 1, the THE KMBRYO. primitive streak; 2, the axis of the embryo; 3, the proamniotic area, in which for a long period there is no mesoderm in amniota. It might also be added that as the mesoderm is excluded from the oral and anal membranes there is no coelom in them. Throughout the rest of the germinal area the coelom gradually extends, but for a long time it fails to reach, and in certain animals never reaches, the periphery of the constantly expanding mesoderm. The history of the embryonic coelom is given in special chapters, the history of the extra-embryonic coelom is indicated in the section of this chapter upon the origin of the amnion. 3. The appearance of the blood-vessels and blood has been con- sidered in Chaper X. ; it leads to the differentiation of the area vas- culosa {Gefdsshof, aire vascnlaire) , v\rhich is the region of the extra-embryonic circulation. As soon as the embryonic area con- tains a distinct vascular netvi'ork, there appears a peripheral vessel which marks the boundary of the area vasculosa, and is called the sinus terminalis. The vasculosa does not reach to the outer bound- ary of the germinal area, so that the region of the blood-vessels is inclosed in a ring which is known as the area vitellina. Topography. — The first differentiation in the germinal area, which can be clearly recognized by the naked eye, is the appearance of the area pellucida, which ^ is shortly followed by that of the primitive streak. Fig. 78, p. 131. Further prog- ress results in the gradual differentiation of the embryo, the steady expansion of the germinal area over the yolk, in the sharper demarcation of the area pellucida, which becomes pear-shaped, and in the appearance of the blood- vessels. Fig. 155 represents the embryonic area of a hen's ovum after about thirty hours' incubation. The em- bryo is well advanced in de- velopment, for although the primitive streak, pr, still re- mains in part and the medul- lary groove, Md, is still open behind, the brain is already marked out and the head has become partly free; along- side the medulla lie nine pairs of segments (proto-vertebrse, auct.) ; around the embryo one easily recognizes the pear-shaped area pellucida, ^-1 .p, and the darker area opaca, Ao, by which it is inclosed ; the area vasculosa stands out conspicuously and is bounded by the already distinguishable sinus terminalis, st; around and under- neath the head is the translucent proamniotic area, pro. am, from Fig. 155.— Chicken Embi,\o and Iti.ui Atea, after Twenty-seven Hours' Incubation, fov, Fovea cardia- ca; pro. a?rt, proamniotic area; a.c. v, amnio-cardial vesicle; st, smus terminalis; pr, primitive groove; . lo, area opaca ; Ap, area pellucida. After Duval. THE GERMINAL AREA. 373 which the mesoderm is altogether absent, and which therefore cannot contain any blood-vessels, nor are there at this stage any vessels in front of the proamnion. In the ovum of the mammalia there occurs a modification of the ectoderm, where that layer is attached to the walls of the maternal uterus. The region over which the attachment takes place gives rise in the higher mammals to the placenta. Hence the area of modified ectoderm may be called the placental area. It has been, as yet, very little studied. As it is not possible at present to speak in general terms of the embryonic area of mammals, I confine myself to a description of the area in the much-studied rabbit, following — a.,pl. ~v±. Fia. 156. — Embryonic Area of a Rabbit of; Eleven Days,witli the Placental Area Partly Torn OfP. After Van Beneden and Julin. pr. a. Proamnion; a.a^ amniotic area, approximately identical with the area pellucida ; a. v. , area vasculosa ; a, pl^ area placentalis ; v. r, sinus terminalis. Van Beneden and Julin, 84.1. The germinative area, Fig. 156, is nearly circular, and at the stage figured shows the following peculiarities. The nearly straight embryo lies in the centre and exhibits plainly the central nervous system and the proto- vertebra ; around the head of the embryo is a clear space, pr. a., the pro- amniotic area, over which no mesoderm is developed ; around the sides and hind end of the embryo is another light place which con- tains mesoderm, but is distinguished by the retarded vascularization ; this is the amniotic area, a. a., and is converted by a process of up- folding into the amnion, which covers the posterior portion of the rabbit embryo. The remainder of the germinal disc constitutes the area vasculosa, a. v., with the terminal sinus, blood-islands, etc. The area consists of two membranes, the upper, the somatopleure, 18 274 THE EMBRYO. the lower, the splanchnopleure ; a large portion of the former behind the embryo has been torn off, a. pi.; this defect is due to the fact that over this region villosities have appeared, and a close connection established between this region and the uterine wall ; it is by this means that the ovum is attached ; hence, when the embryo is re- moved from the uterus, this area of the splanchnopleure (chorion) remains adherent to the uterus. As development proceeds, the allantois grows up against this area, over which the differentiation of the placenta takes place ; hence the name, area placentalis. Area Vasculosa. — Soon after the capillary network of the area opaca and pellucida has penetrated the embryo, certain lines of the network begin to widen, and soon distinctly assume the size and functions of main trunks; some of these unite with the posterior Ara Om.A. Fig. 157. — Diagram of the Circulation in a Chick at the End of the Third Day, as seen from the Under or Ventral Side. The embryo, with the exception of the heart, H£. , is dotted ; Arc^ aortic arches; Z).C, ductus Cuvieri; JtAg. , jugular vein; card.^ cardinal vein. The remaining letters are explained in the text. The veins are black ; the arteries cross-lined. venous end of the heart, which has meanwhile been formed in the embryo, and others become connected with the anterior or aortic end ; even before this the heart has begun to beat, so that, as soon as all connections are made, the primitive circulation starts up. The arrangement of the vessels is not the same in birds and mammals, although commonly so stated. The disposition in birds is indicated by the diagram shown in Fig. 157, in which, it should be remem- bered, the embryo and the capillary network are drawn many times too large in proportion to the area vasculosa. The area is bounded by a broad circular vessel, the sinus terminalis, S.T., which consti- tutes a portion of the venous system in birds, for in front of the head of the embryo the sinus leaves a gap, and is reflected back along the sides of the body of the embryo to make two large veins, which, after uniting with other venous channels coming from various parts of THE GERMINAL AREA. 275 the area vasculosa on each side, enter the embryo as two large trunks, Om.V., known as the omphalo-mesaraic veins; these two veins unite in a median vessel, the sinus venosus, S.V., which runs straight forward and enters the posterior end of the heart. The sinus venosus also receives the veins from the body of the embryo, namely, the jugulars, Jug., and cardinals, card.; the former from in front unite each with the cardinal of the same side, making a short transverse trunk, known as the ductus Cuvieri, D. C. ; the two ducts empty into the sinus venosus. The entire venous current is thus brought to the heart in a united stream ; it passes out through the aorta ; the greater part ascends the aortic arches and passes back as shown in the figure, Ao., and divides at the posterior fork of the aorta, the bulk of the two currents passing out through omphalic arteries, Om.A., and thence to the capillaries of the area vasculosa and so on to the venous trunks again. As shown in the figure, which presents the under side of the area, the left omphalo-mesaraic vein preponderates, and in the latter stages this difference becomes more marked until finally the right stem is very inconsiderable in comparison with the great left vein. The time at which the dis- parity commences is extremely variable, as is also the degree of inequality between the two veins. The following description probably represents what was the prim- itive condition of vessels in the mammalian area vasculosa. It Fig. Area Vasculosa and Embryo of a Rabbit. After Van Beneden and Julin. applies to an early stage in the rabbit, which has been figured by Bischoff, 43.1, Tab. XIV., Fig. 60, whose figure is copied in Kolli- ker's "Grundriss," Fig. 90, p. 109. An essentially similar arrange- 276 THE EMBRYO. ment of the vessels exists also at a corresponding stage in the dog, BischofE, 45.1, Taf. VII., Fig. 37, C. The veins are much more symmetrical than in the chick, and have the same general plan ; the sinus terminalis belongs to the venous system, so that the connection with the arterial circulation, found later, is secondary ; the aorta of the embryo is double, and gives off on each side (segmentally ar- ranged?) transverse branches, one of which develops into the large trunk shown in Fig. 158; the network of small vessels forms two layers, of which the upper is connected with the arteries, the lower with the veins. The change from the earlier condition to the later has still to be followed. Selenka has figured the vascular area of an opossum, 86.1, Taf. XXIII., Fig. 3, in a condition which suggests at once a transition from between that just described and that described in the next paragraph; the figure shows the veins without direct connection with the sinus, while the aorta, though it gives off numerous small branches, has extended tailward of the embryo and joined the sinus. According to Van Beneden's recent researches on the rabbit the arrangement of the main vessels in the area vasculosa at a later stage is quite different. The sinus terminalis forms a complete ring. Fig. 158, and is connected with the arterial system by a single trunk, which corresponds to the left omphalic artery of the bird. For some time the connection between the embryonic arteries and the area vasculosa is entirely through capillaries, and the arterial trunk on the vascular area does not appear in the rabbit for several days. There are two veins, one arising from each side of the body and passing out on to the area vasculosa over the back of the embryo ; they are the two large upper vessels in the figure. Growth of the Vascular Area. — As the blood-vessels appear at first only in the splanchnic mesoderm, the vascular area belongs to the splanchnopleure, or, in other words, is part of the wall of the yolk-sac ; hence the circulation of the area is often spoken of as the vitelline circulation. The growth of the vascular area is therefore part of the history of the yolk-sac, and is considered now from con- venience merelj". The expansion of the vascular area is due to the growth and differentiation of the mesoderm, and in those mammals in which, as in the rabbit, the mesoderm extends only part way over the yolk, the vascular area cannot spread over the whole blastodermic vesicle ; but in those mammals in which, as in man, the mesoderm grows completely around the yolk, the vascular area may also extend completely around the yolk, with the consequence of the disappear- ance of the sinus terminalis. In the earliest known stages of man, the yolk-sac was found completely vascularized. The gradual spread of the area vasculosa over the yolk may be readily followed in the hen's egg. It is due, as just stated, to the growth and differentiation of the mesoderm. The size of the vascu- lar area is very variable, but the following table represents the approximate sizes, for several ages, as measured on blastoderms removed from the yolk, flattened and hardened ; the total circumfer- ence of the heii's yolk is about 90 mm. The area vasculosa of the chick measures — THE FORM OF THE EMBRYO. 277 At 2 days about 9 mm. in transverse diameter. " 2.5 " " 15 " " 3 " " 19 '• " 3.5 " " 23 " " 4.5 " " 30 " " 6 " " 40 " It is not until the seventeenth day of incubation that the yolk is completely overgrown by the vascular area, Duval " Atlas," Fig. 651. II. The Form of the Embryo. It has been pointed out already that among vertebrates there are two principal types of embryonic form : one, which is the ixiore prim- itive, characterized by the yolk-mass being included in the body of the embryo ; the other is secondary and characterized by the separa- tion of the embryo and the yolk. The primitive type of vertebrate embryo is found in the lampreys, ganoids, and amphibians ; the ventral side of the embryo is very much distended to allow room for the yolk, which consists, after the segmentation is completed, of a mass of cells, which lie for the most part below the archenteric cavity, as cross-sections show at once. As the development progresses, the embryo lengthens out, but the swelling caused by the yolk persists for a long period, the yolk material being only gradually resorbed by the embryo ; the swelling is readily recognized, even up to larval stages. The secondary type of vertebrate embryo is found in elasmobranchs and amniota. In elasmobranchs, when the embryo appears it occupies only a small part of the ovum, which is very large and contains much yolk. Soon after the appearance of the medul- lary groove, the head of the embryo begins to grow forward entirely free from and above the yolk ; and by the time the medullary groove is converted into the medullary canal the tail begins to grow back- ward in a similar manner independently of the yolk ; hence, only the central region of the embryo remains connected with the yolk. As the growth of the embryo continues, while the area of its body at- tached to the yolk increases very little in size, it follows that the connection becomes relatively smaller, until it becomes merely a narrow stalk as compared either with the embryo or the mass of yolk. The traditional and often-repeated description of the separa- tion of the embryo from the yolk attributes the separation to a fold- ing off of the embryo by the germ-layers being tucked in under the embryonic anlage, but it seems to me that the process is only appar- ent, and that it is by its own growth, as above described, that the embryo becomes partly separated from the yolk ; and I hold the same view as regards the amniota. The yolk is covered by the extra-embryonic extensions of the meso- derm and octoderm, the yolk proper being, of course, entoderm. If the mesoderm develops a coelomatic fissure around the yolk, we have the non-embryonic parts of the ovum converted into a double sac; an outer sac formed by the united ectoderm and mesoderm (somatopleure) , and an inner sac of mesoderm filled with the yolk- 278 THE EMBRYO. mass (vitelline entoderm) , the two representing the splanchnopleure. The enter sac in all vertebrates may be called the chorion, the name by which it is known in mammalia; the inner sac is the yolk-sac or umbilical vesicle. In ainniota, the separation of the embryo from the yolk takes place in the same general manner as just described for elasmobranchs, but there are additional complications due to the development of the am- nion and allantois taking place very early — see the following division of this chapter. Form of the Amniote Embryo. — It is not proposed to give here a comparative account of the forms of amniote embryos at suc- cessive stages, but merely to briefly indicate the characteristics of the stage in which all the principal anlages of the primary organs are present, but not specialized. The stage may be taken to be that of the hen's ovum at fifty to sixty hours of incubation, Fig. 150. The blastoderm reaches at this time over nearly half of the yolk, the extreme margin of the opaque area being near the equator, but the vascular area is much smaller, being only about 30mm. in diameter; still smaller is the pear-shaped area pellucida, in the centre of which lies the rapidly growing embryo. At this period the vascular area may be said to be in the stage of its most complete development ; for though it will afterward become larger, it will at the same time become less definite and relatively less important. The arterial system already has its main trunks. Fig. 167.A.V., and the main stems of the omphalo-mesaraic veins, om. V, are differentiated. As regards the embryo the most striking features are the advanced development of the head and the slight differentiation of the tail. The head has grown forward so as to be entirely free from the yolk, and is turned so that its left side rests upon the yolk, and as the tail end of the embryo still rests symmetrically upon the yolk, it follows that the intermediate portion of the body is twisted. This warping or tor- sion of the embryo, in order that the side of the flattened head may rest upon the yolk, occurs in Sauropsida and to a slight extent in pla- cental mammals, but not among any of the Ichthyopsida. We must, therefore, regard it as a special feature of the amniote embryo, which has been lost in the placental mammals, probably as a result of the loss of food yolk in the ovum. The head is remarkable for the advanced differentiation of its parts; the anlages of the eye, Fig. 150, L, and ear, Ot, are present; four iDranchial pouches are developed, 6rV the heart is large and already bent, Ht; the medullary tube is very much dilated and distinctly divided into its three primary vesicles, H, 31b, Hb. The head is also bent at the region of the mid-brain, Mb, so as to form almost a right angle with the axis of the hind-l3rain, Hb, and neck. This head-bend or cervical flexure is highly charac- teristic of all vertebrates ; it is beautifully shown in elasmobranch embryos, and can be easily recognized in all classes. It is a bend in the median plane of the embryo by which the end of the head is brought over toward the heart, Ht. Following along back- ward we encounter the first distinct segments just behind the oto- cyst, 0^, and can follow them some distance behind the vitelline arteries, until they merge into the undivided segmental zone, Ar; the limit of the body of the embryo is already indicated by the THE FORM OF THE EMBRYO. 279 parietal zone, but the zone will be encroached upon by the vas- cular area, and the whole zone of this stage is not destined to be included in the body of the embryo. In a sheep embryo,* although the fundamental characteristics are the same, there are many minor differences both from the chicken and the rabbit. The most striking peculiarities of the embryo are due to the foetal appendages, the development of which presents special modifications in ruminants, as more fully described in the next division of this chapter; the yolk-sac is long and narrow, and is connected by a broad twisted yolk-stalk with the embryo ; the allan- tois has already become a very large transversely expanded vesicle ; the amnion invests the embryo closely and gives off a long cord (Amnionstrang) , by which it is still attached to the chorion. The embryo, 5 mm. in length, is curving throughout its length; the head-bend is developed, and consequently the end of the head lies near the heart ; the torsion of the whole embryo is very marked, the dorsal side of the fore- brain facing us, of the neck being turned away from us, of the tail facing us again ; the embryo makes nearly one complete spiral turn. The head is small, laterally compressed, and less advanced than in the chick described above, for the anlage of the eye is only just begun ; that of the ear is not differentiated, and the first two visceral arches are present, while the third is only just beginning. The medullary groove is still open in the region of the forebrain, and widely open at its tail end, but closed throughout the rest of its length ; there are fourteen segments ; none of the vessels yet contain any red blood. Typical Embryo in Cross-Section. — For this purpose I select a dog-fish embryo. The following description is intended especially for the convenience of students. The body is bounded by a single layer of ectodermal cells, Ec, the anlage of the future epidermis ; the central nervous system, Mel, appears as a tube, with very much thickened cellular walls ; it lies on the dorsal side of the embryo, and although developed from the outer germ-layer, has no connection with the ectoderm; below the nervous system lies the very large notochord, nch, which contains a loose network in its centre, and a denser peripheral layer of cells; it is invested by a thin hyaline structureless sheath; the notochord as we ascend the vertebrate series diminishes in size; at corresponding stages in amphibians it is decidedly smaller in proportion to the medullary tube than in sharks — in birds its diameter is not more than a fifth — in mammals not more than a twelfth of the diameter of the medullary tube. Be- low the notochord comes the dorsal aorta, Ao, on either side of which, a little lower in position, may be seen a cardinal vein, c. V, while between the notochord and aorta is a small string of cells known as the subnotochordal rod or hypochorda, a structure which has not yet been observed in any of the amniota. The body-cavity proper, or splanchnoccele, Coe, is a wide space, bounded externally by the body walls, Som, and containing the intestinal canal. In, which has been developed from the splanchnopleures, and which is suspended from the dorsal wall by the membranous mesentery ; the cavity of the in- testine is lined by entoderm, En, and ta kes a spiral course which is *See Bonnet, 80.1, Fig. 13. 380 THE EMBRYO. characteristic of the elasmobranchs, but classes ; the abdominal cavity is lined by Fig. 159. —Transverse Section of the Rump of a Dog-Fish Embryo, 14 mm. long. Ec, Ectoderm; Md, medullary tube; My^ myotome; ncTi, notocbord; Mus, muscle; Ao^ aorta; c. "F, cardinal vein; s.s, segmental tubule; S. i>, segmental or Wolffian duct ; Coe, coelom ; Mst, mesentery (the reference line has been omitted) ; Som, somatopleure ; En^ entoderm ; si. F, sub-intestinal vein. is not encountered in other the epithelial mesoderm or mesothelium. The prim- itive longitudinal urogeni- tal duct appears in cross section just above the splanchnocoele, Coe, while near it on one side can be seen the opening of one of the transverse WolfSan or segmental tubules, st, which has been developed from the nephrotomic por- tion of the primitive seg- ment ; if the tubule is fol- lowed out its other end is found to open into the Wolffian duct ; in amniota the opening into the body- cavity is lost at a much earlier stage. The myo- tome, My, which also is developed from the prim- itive segment, is a double plate, its two walls being so closely appressed that the cavity between them is completely obliterated ; the inner wall is partly con- verted into muscular tis- sue. The mesenchyma, mes, has grown more than any other tissue, and con- stitutes in bulk the greater part of the embryo; it is destined before adult life is attained, to be differen- tiated into a large variety of tissues. III. Origin of the Fcetal Appendages. Under this head we have to consider the origins of the chorion, yolk-sac, allantois, proamnion and amnion, but as we have already considered the development of the yolk-sac, p. "2.55, the allantois, p. 357, and the proamnion, p. 150, we shall recur to them now inciden- tally only, and concern ourselves principally with chorion and amnion. Extension of the Extra-Embryonic Ccelom. — The distance to which the coelom can extend around the ovum depends upon the extension of the mesoderm, for of course the cavity cannot go farther than the layer within which it is developed. Now, as we have seen, the mesoderm expands gradually and a little more slowly than the ORIGIN OF THE FOSTAL APPENDAGES. 281 germinal area. This gradual expansion occurs in all vertebrates. In the primitive type (Petromyzon and amphibians) the mesoderm and the ccBlom both grow completely around the yolk ; and this was undoubtedly the primitive condition, but in the lower amniota the growth of the mesoderm has to be much greater in order to cover the enormous yolk mass ; hence in amniota the spread of the meso- derm is slow and long continued, and the embryo advances far in its development before the yolk is inclosed. In mammals the expan- sion of the mesoderm over the yolk-sac is also slow, and in rabbits (and probably in their allies) the mesoderm never extends over the "whole yolk-sac, but in man, on the contrary, the ccslom as well as the mesoderm are developed completely around the yolk-sac very early. No explanation of these differences among mammalia can be offered at present. In the lampreys and amphibians the appearance of the coelom around the yolk merely completes the separation of the body-wall or somatopleure of the embryo. In the amniota it also separates the somatopleure from the splanchnic mesoderm around the yolk, but owing to the division of the developing ovum into embryo proper and yolk-sac, only a small part of the somatopleure shares in the for- mation of the embryo, while the rest acts as a covering membrane of the yolk. This membrane in the mammalia is universally known as the chorion, and I shall apply this name to it hereafter for all vertebrates. Primitive Chorion. — The chorion has been defined by Minot (Buck's "Handb.," II., 143) to be the ivhole of that portion of the extra-embryonic somatopletire which is not concerned in the formation of the amnion. The term primitive chorion may be employed for the whole extra-embryonic somatopleure before the differentiation of the amnion from it, and the term chorion or true chorion be still used, as defined, for what remains of the membrane after the separation of the amnion. The somatopleure consists of two layers — the ectoderm and somatic mesoderm. The ectoderm consists of a single layer of epithelial cells. The mesoderm consists of a layer of mesothelium next the •coelom, and a thicker layer of mesenchyma between the mesothelium and ectoderm. The exact appearances of these layers are described with the aid of figures in the special chapters on the amnion and chorion. Origin of the Amnion. — The amnion is developed out of that part of the extra-embryonic somatopleure which immediately sur- rounds the embryo and the proamniotic area, or in other words, the amniotic region of the germinal area is part of the area pellucida, and perhaps includes the whole of the pellucida. The amnion owes its development to the expansion of the coelom. In the Sauropsida the process is about the same in all forms, but in mammals there are several modifications of the development known ; hence we consider first the sauropsidan, then the mammalian types. In the Sauropsida the formation of the amnion begins with the appearance of the amnio-cardial vesicles, p. 198, which form con- spicuous dilatations on either side of the neck, Fig. 117; the vesicles ■steadily enlarge and spread laterally and forward so as to inclose the 283 THE EMBRYO. proamniotio area, and finally fuse in front of it. The dilatation takes place in such a manner that the splanchnopleure is bent down slightly, while the somatopleure is bent upward to an extreme de- gree, forming a sort of dome. Transverse sections of a chick at this stage, at the level of the heart, show. Fig. 117, the amnio-cardial vesicle of each side fused with its fellow in the median line below the heart, Htj the somatopleure, Som, of the embryo makes a sharp turn outward and upward. Am, and then bends away again, Cho, from the embryo and finally joins the splanchnopleure of the yolk, Spl. As the upbending of the somatopleure goes on around the entire head of the embryo, it follows that the cephalic end of the embryo lies in a depression, the sides of which are formed by a part, Am, of the extra-embryonic somatopleure. While this is going on, the head of the embryo bends over, and the whole head gradually rolls over ventralward and thus is forced into the yolk, but since the proamni- FiG. 160. —Section through the Rump of a Rabbit Embryo of Eight Days and Three Hours. Md, Medullary tube; Seg, primitive segment; Cho, chorion; Am, amnion; Som, somatopleure of embryo; Coe, coelom: Spl, splanchnopleure; Ent, entoderm; Ch, notochord; Ao, aorta. otic area lies just here, it is invaginated along with the head, and consequently the head seems covered by a proamniotic membrane, which is known as the cephalic cap {Kopfkappe, capuchon cepha- lique) . ^ This cap is very noticeable in young chicks, for the head is hidden in it, while the rest of the embryo is uncovered. The actual relations are still further complicated by the singular fact that the edge of the cap is extended backward by the growth of the ectoderm alone, as shown by Duval; the backward growth of the ectoderm occurs also in turtle embryos, and to a much greater distance than in birds before the mesoderm follows it (K. Mitsukuri, 90.1). Sooner or later the mesoderm penetrates the 3ctodermal fold, and the ccelom appears in it as a forward extension of the cavities of the amnio-cardial vesicles. The cephalic end of the embryo now soon becomes completely cov- ered over by the extra-embryonic somatopleure ; this is due to the expansion of the coelom on all sides. The changes in the extra-embryonic somatopleure around the pos- ORIGIN OF THE PCETAL APPENDAGES. 283 terior half of the embryo, are similar in a general way to those around the anterior half, but the dilatation of the coelom is confined to the extra-embryonic region, hence the pictures obtailied from cross sections of the two parts of the embryo present certain essential differences. Fig. 161 is a section through the rump; here we see that the dilatation of the coelom causes the somatopleure to form a longitudinal fold along each side of the embryo; each fold, passing back- ward, joins its fel- low behind the em- bryo, so that they may be described conjointly as the tail-fold (Schiuanz- kappe, capuchoii, caudale) . The tail- fold is developed considerably later than the head-fold, but as one grows forward and the other grows back, they finally meet and constitute the complete amniotic fold around the en- tire embryo. The tail-fold gradually closes over the em- bryo ; the process may be understood from the accom- panying figures. Fig. 160 rei^resents a cross-section of a rabbit embryo. The somatopleure, Som, of the embryo bends over as the amnion. Am, so as to cover the embryo, above which it again bends outward as the chorion, Cho; we can already distinguish the embryonic, amniotic, and chorionic portions of the somatopleure from one another; where the amniotic portion joins the chorionic, the edge is prolonged by a thickening of the ectoderm, which re- minds us of the similar thickening at the edge of the cephalic cap ; the two edges have almost met over the back of the embryo; the 284 THE EMBKYO. asymmetry of the folds exists in all amniota and, as shown in the figure, is very marked in the rabbit, but is much less marked in the Sauropsida. In the next stage, Fig. 161, the folds have actually met; their edges grow together by their ectodermal thick- ening ; for some time the thickened ectoderm persists and offers in sections a characteristic feature ; after a time the mesoderm grows across, and the ectoderm of the amnion is entirely separated from that of the chorion ; still later the cavity of the chorion also pene- trates and completes the final separation of the amnion from the chorion, Fig. 19. The process of separation is essentially the same in the case of the cephalic amnion. The separation of the amnion from the chorion progresses most rapidly at the head end ; at the tail end it begins later and progresses forward ; hence the portion of the amnion over the middle of the rump is the last to be formed, as can at once be seen if the fresh ovum be examined. In surface views the gradual closure of the amniotic folds over the embryo can be beautifully followed; for example, in the hen's ovum incubated about sixty hours, we find the anterior half of the embryo entirely hidden by the cephalic cap, while the posterior third of the rump is also covered by the tail-fold, and at the sides of the rump the amniotic folds have partially arched over the embryo. These ar- rangements leave a small longitudinal oval opening through which we can look down upon the embryo. The opening gradually dimin- ishes as the edges of the folds advance, and is finally closed by the meeting of the edges from all sides. As the edges have their ecto- derm thickened, their final meeting is marked by a local thickening of the ectoderm, which persists for some time after the actual closure. In ruminants the connection between the amnion and chorion at the point of final closure is retained for a long time by means of tissue, which grows out into a long thread, the so-called amniotic cord {Amnionstrang) . A sonaewhat similar structure occurs in the opossum, Selenka, 87.1, Taf. XXV., Fig. 2. After the amniotic folds have closed, the embryo is surrounded by two membranes, both derived from the extra-embryonic somato- pleure. Of these the outer is the true chorion. Fig. 19, Clio, the inner the amnion ; from the manner of their formation the former has its ectodermal layer external, the latter its ectoderm internal or facing the embryo. The amnion. Fig. 19, Am, is the direct pro- longation of the somatopleure of the embryo ; the space between the amnion and the embryo is called the amniotic cavity; it is lined throughout by ectoderm. In mammals the development of the amnion was presumably at first like that in the Sauropsida, for not only do we find many traces of it still preserved, but also Selenka, 86.1, 130, has shown that in the opossum the sauropsidan stage is passed through, although some- what modified by the excessive development of the proamnion. The increased importance of the proamnion can be seen also in the rabbit (Van Beneden and Julin) , and is possibly characteristic of mammals as a class . In the two animals mentioned, more than half of the embryo is covered by the proamnion at the time the amnion closes, and hence the amnio-cardial vesicles cannot attain the size or importance they have in birds, and they are unable, in the opossum, to extend around ORIGIN OF THE FCETAI^ APPENDAGES. 285 the proamniotic area ; hence in front of that area there is no ccelom developed, the three germ-layers remaining in close contact and forming, as it were, a single membrane; in the rabbit the coelom does appear in front, as in birds. In ruminants the amnion appears very early, the folds being well advanced before the medullary groove appears. The formation of the amnion is induced by the precocious development of the extra- embryonic coelom, which, as Bonnet's researches on the sheep, 89. 1, have proven, extend very early around the embryo in a wide oval ring, which, by raising the somatopleure, forms an annular amniotic fold, before the embryo can be said to be differentiated ; these folds close over the anlage of the embryo, and by their union produce the two foetal membranes, amnion and chorion, in the same manner as in birds ; as already mentioned, there is formed at the point of closure a long cord of tissue {funiculus amnii) , hy which the two mem- branes remain united for a considerable period. In the rodents with so-called inversion of the germ-layers (e. g., guinea-pigs, rats, mice, etc.), the development of the amnion is ex- tremely modified from the original type. The cavity of the Trager, Fig. 87, a, becomes in part the cavity of the amnion. The ^, ^^, ^^^^__ J\^ manner in which this takes place and the way in which the process may be deduced from the primitive mode of development are both well il- lustrated by Selenka's dia- grams, 84.1, Taf. XVI. The human amnion in the earliest stages yet known has been found completely closed over the embryo, so that noth- ing is known as to its devel- opment by direct observation. The earliest known disposition was first described by W. His, whose account has been con- firmed by subsequent observ- ers. The embryo is from 2.5 to 3.0 mm. long; its relations to the rest of the ovum are in- dicated by the diagram, Fig. 162, B; it rests ui^on the large yolk-sac, V, and is connected by a short stalk, Al, with the chorion, Ch. The amnion arises under the head at the junction of the embryo and yolk-sac, and from the sides of the embryo and from the allantoic-stalk, and arching oyer the dorsal side of the embryo completely incloses it. To explain this disposition His has advanced the following hypothesis as to the Fig. lOS.— Diagrams to illustrate His' theory ot the Origin of the Human Amnion : A, First stage ; B, second stage, ^m, Amnion; ^Z, allantoic-stallc or Bauchstlel ; Ch, chorion, the villi of which are drawn smaller and more numerous than in nature ; F", yolk- sac. 286 THE EMBRYO. course of development. The embryo arises upon the surface of the blastodermic vesicle in the usual manner ; its somatopleure passes over into the primitive chorion, which is, at an extremely early age, completely separated from the yolk-sac; the chorion now forms a fold, as shown in Fig. 162, A, which arches backward over the head of the embryo; while the tail end of the embryo, retaining its direct connection with the chorion, becomes the allantoidean stalk, Al. The head-fold, of which the inner leaf is the amnion. Am, the outer leaf a part of the true chorion, grows backward over the embryo as indicated by the dotted line, Am', until it finally reaches the allantois-stalk, Al, and thus completely covers in the «mbryo. This hypothesis is probably correct, but it is possible that the amnion is preceded by a true proamnion, which becomes obliter- ated very early by the precocious development of the mesoderm and the coelom in the human ovum. If Graf Spee's plausible sugges- tion, 89. 1. 170, that there is a so-called inversion of the germ-layers in the human embryo be verified, then we shall probably find that the human amnion is developed according to the rodent type men- tioned above. The True Chorion is that portion of the extra-embryonic soma- topleure which remains around the ovum after the separation of the amnion ; it consists of an outer layer of ectoderm and an inner layer of mesoderm; the cavity within it is part of the coelom. By the closure of the amniotic folds the chorion becomes a membrane sur- rounding all the other parts of the ovum, and makes a complete bag, which is termed the chorionic vesicle. The chorion is the outermost of the foetal envelopes. It is sometimes termed the serous membrane or envelope {membrana serosa, serose Hulle) , especially in writings on sauropsidan embryology. Its relations may be rendered clear by the help of diagrams. Figs. 20 and 19. lY. Known Human Ova of the Second and Third Weeks. As no synopsis has ever been made of our knowledge of the early ■stages of man, I have attempted to collate all the descriptions of embryos not over three weeks. A summary of the descriptions is given, p. 308. Classification by Stages. — Any attempt to divide embryos into stages must necessarily establish artificial groups, for in nature there is no demarcation. Division into stages is for convenience, and ought, therefore, to be made by natural and obvious characteristics. After much deliberation I have chosen eight stages, which seem to me natural and convenient, and I have classified the thirty-eight embryos reviewed in the preceding pages, placing them according to my best judgment in their respective stages ; when the assignment is doubtful I have indicated it by an interrogation mark. First Stage: Appearance of the primitive streak. 1. Reichert's. 2. Breus'. 3. Wharton Jones'. 4. Ahlfeld's. 5. Beigel and Lowe's. KNOWN HUMAN OVA. 287 7. Kollmann's a. 8. " b. 9. Schwabe's. Second Stage: Appearance cf the medullary plate. 10. W. His'XLIV. (Bff). 11. Keibel's. 12. Spee's. Third Stage: Appearance of the medullary s^roove. 13. W. His' E. 14. Allen Thomson's No. I. 15. W. His' SR. 16. Allen Thomson's No. II. Fourth Stage: Formation of the heart and medullary canal. 17. Spee's second embryo. 18. Kollmann's embryo of a. 3 mm. 19. Von Baer's youngest ovum. Fifth Stage: First external gill-cleft. None. Sixth Stage: Two external gill-clefts. 20. Minot's No. 195. 21. " No. 143. 22. W. His' LXVIII. (Lg). 23. " " LXVI. (Sch. I.). 24. " " L. 24A. Janosik's. 25. Coste's. 26. Schroeder van der Kolk's. ? 27. Hennig's. ?? (9. Schwabe's.) ? 28. Remy's. Seventh Stage: Three external gill-clefts. 28A. Chiarugi's. 29. W. His', Rf. 30. " M. 31. " BE. 32. " Lr. 33. Allen Thompson's No. III. ? 35. Ecker's. ? 36. Hecker's. ? 5. Beigel's (abnormal). Eighth Stage: Four external gill-clefts. ? 34. Von Baer's. ? 37. Johannes Mtiller's. 38. R. Wagner's. Descriptions of the Kno-wnOva. 1. — Reichert's ovum, 73.1, was thought by him to be twelve or thirteen days old, and probably correctly so, as it was obtained at a post-mortem examination of a young German girl under circumstances which render the estimate of the age quite trustworthy. The ovum itself was very imperfectly examined by Reichert, whose very lengthy memoir deals largely with cognate subjects and contains much speculative matter. The actual 288 THE EMBRYO. Fig. 163. — Eeichert's Ovum. Two views engraved from the original plate description of the ovum is brief (pp. 25-28) ; but as far as he went Eeichert worked with exemplary accuracy, which gives value to his research. The ovum in question was a flattened sphere with a short diameter of 3.3 mm., and an equatorial diameter of 5.5 mm. ; smooth around both poles, and with a marginal or equatorial zone of villi separating the two smooth areas. The smaller and flatter of these two areas faced the uterine wall and bore on its inner sur- face (i-e., within the ovum) a small accu- mulation of rounded cells. The opposite area was more convex. The villi were short (0.2 mm.) thick cylinders with round- ed ends and no branches. The walls of the vesicle consisted only of epithelium, which also formed the simple hollow villi. The contents of the vesicle were: 1, The inner cell-mass lying, as before mentioned, at one pole ; 2, A network of threads, appar- ently the result of coagulation of the con- tained fluid, for no nuclei were found in it. Kollmann, 79. 1, 294, thinks that Eeichert's ovum must have had really two layers forming the vesicular walls — an inner one mesoderm (young connective tissue) and an outer one of true epithelium ; further, that the true epithelium had been lost, and that only the connective tissue remained, which Reichert mistook for epithelium. This siipposition is, I think, not probable. Reichert's ovum is presiimably younger than any other hitherto described, and maj' have been in the stage before the mesoderm had grown over the chorion. The villi are described as hollow by Reichert — a statement not compatible with the supposition that he mistook a solid core of mesoderm for the hollow shell of the ectoderm ; we know now that young villi usually contain no mesoderm at first. 2. Breus' ovum, 77. 1, must be considered further advanced than Reichert's, although the author fixes its age as presumably ten days. The total diameter of the ovum including the villi was only 5 mm. , and as the villi were about 1 mm. long, the diameter of the chorionic ve.sicle must have been about 3 mm. The villi, some branched, but mostly without branches, were thick set, but left one spot bald, agree- ing in this with Jones' ovum (see below). The chorion was smooth on its inner surface, and consisted of (1) an outer epithelial layer, and (2) an inner connective-tissue layer which sent out extensions partly filling the villi. The ovum contained a thready mass which Breus thinks was probably a product of coagulation, and an inner cell-mass about 1 mm. long and 0.5 mm. wide. The presence of villi and the existence of the mesodermic layer of the chorion, contrasted with the absence of any embryonic structure, led Breus to consider his ovum abnormal. But it is rather the contrary conclusion we must draw, since all our knowledge points to the deduction that, as compared with the embryo, the development of the chorion is very precocious in mammalia. I deem it, therefore, probable that Breus' KNOWN HUMAN OVA. 289 ovum was normal, and that the inner cell-mass he describes was in reality the embryo, compare Keibel's ovum. 3. Wharton Jones, 37.1, long ago described briefly a human ovum, the chorion of which measures in his figure (said to be nat- ural size) 6 by 4 mm. The following is all that can be gathered from Jones' description : The ovum was already covered by the de- cidua, and bore shaggy villi on the side toward the uterus, while the other side was bald. " The whole cavitj' of the chorion was filled with a fine gelatinous cellular tissue, imbedded in which, toward one extremity of the ovum, was a small round body ; it was evidently the vesicular blastoderma. On being taken and examined under the microscope, it presented the same friable, globular structure found in the vesicular blastoderma of the rabbit in the preceding observa- tion. There was no vitellary membrane to be seen. " To judge from the minute figure given, the villi were already branched; in Rei- chert's ovum they were still simple. 4. Ahlfeld's ovum, 78.1, represents perhaps the same age as Jones', but he does not give its diameter, which appears from inci- dental references to have been about 5 mm. The author's descrip- tion is not exhaustive by any means, but he mentions two points of great interest: first, the presence of a layer of connective tissue (mesoderm) underneath the chorionic epithelium, and extending into but only partially filling the villi of the chorion; second, the character of the villi, which are slightly branched and are constricted at the base, only their tips touched the surface of the decidua (reflexa and serotina) . He also states that the epithelium of the villi pre- cedes in its growth the connective tissue. This ovum was supposed to be fourteen to sixteen daj'S old (?). Owing to an accident, no observations of its internal contents were made. 5. 6. Beigel's ovum, 78.1, of which he maintains that it is the third smallest known, is, if we may judge from his plate, certainly abnormal to an extreme degree. I hold it to be a malformed ovum of the fifth or sixth week. The ovum described by Beigel and Lowe, 77.1, is of an even more questionable character. Moreover, their account is considered by Breus and Ahlfeld to be very inaccurate. It is noteworthy that Beigel and Lowe have also noticed the early pres- ence of the mesoderm under the'chorionic epithelium. Lowe, 79. 1, defends himself against Ahlfeld's attack, and insists with justice upon the presence of connective tissue on the inside of the chorion in ova of the second and third week. 7, 8. Kollmann's memoir, 79.1, is by far the most valuable which had appeared up to the time of its publication upon the structure of very young human ova. He describes two ova, a and b, both pre- served in the anatomical collection at Basle. Ovum a had been placed in glycerin and water, which preserved the form of the speci- men but ruined it histologically ; nothing was made out as to the contents of the chorionic vesicle. The vesicle itself measured 5.5 by 4.5 mm., and therefore was slightly flattened. This measure does not include the villi, which were from 1 to 1.2 mm. long, and re- peatedly branched. Ovum 6, 5. 5 mm. in diameter, was well preserved in alcohol ; the villi were somewhat branched ; the contents of the ovum were lost. On the other hand, the uterus belonging to this 19 290 THE EMBRYO. ovum was also preserved, and forms the basis of a very valuable description of the uterus in early pregnancy, to which I hope to recur on another occasion. KoUmann's two ova are both much more advanced than those of Reichert, Breus, and Jones, as is shown by their greater size and the branching of the villi. It is a matter of profound regret that only the chorion was left, but, fortunately, Kollmann has taken good advantage of his opportunity. His paper also gives an excellent critical analysis of nearly all the previous literature. He points out that the two primtive layers of the chorion are probably normally present at this stage. The chorion of his ova, he says, consists of " einer Lage jugendlichen, embryonalen Bin- degewebes, das zahlreiche Rund- und Spindelzellen enthalt, und das bedeckt wird von einer einfachen Lage platter Zellen" (p. 393) . He then passes the literature in review, and insists strongly upon the fact that the two layers have been distinguished in nearly all the very young human ova known except Reichert's. Kollmann, there- fore, as was mentioned above, questions, I think without sufficient foundation, the accuracy of Reichert's account. Concerning the connective-tissue layer Kollmann says but little. As regards the epithelium, he points out that the nuclei occupy a basal position so that the outer parts of the cells form a granular stratum, which some authors have considered a distinct membrane. The author supposes this granular stratum to become the cuticula described in later stages. Jassinsky, 67. 1, is the chief defender of the existence of a cuticula, which, however, he designates under the extraordinary name of tunica propria, extraordinary because the term is properly applied to the layer of connective tissue immediately upon which an epithelium rests. It is probable in the light of our present knowledge that Kolhnann saw the outer darker layer found in Spee's ovum, see below, and in many others a little older. This outer layer is nucleated, but the nuclei might be overlooked. Finally Kollmann adds (p. 397jf.) observations on the growth of the villi in ova of the fourth week. The outgrowth of branches is very rapid, and occurs with every de- gree of participation of the connective tissue. There are two extremes: 1. A bud consisting wholly of epithelium, which may stretch out into a process with a long thin peUicle and a thickened end, the whole remaining until it has become quite large without any connective tissue. 2. A thick bud with a well-developed core of connective tissue ; such a bud probably grows out as a nearly cylin- drical branch. Between these two extremes every intermediate state can be found. The various forms of growing branches may lie close together. Probably this complex mode of growth persists in older villi, which would explain the multiplicity of forms in the villous branches. 9. Schwabe, 79.1, has described an ovum which he considers thirteen to fifteen days old, but he is certainly mistaken, since both the data he gives as to the age and his account of the embryo shows that it is more advanced and belongs distinctly in the third week. In connection with KoUmann's observations we must notice those of Orth, 77.1, who has shown that at all ages, even at full term, the villi of the chorion in the placenta have epithelial buds, which are at first hollow and are afterward filled up with a vascular- KNOWN HUMAN OVA. 291 ized ingrowth of connective tissue. Apropos of this observation Orth discusses Boll's theory of growth, making the point that in this case the shaping of the parts depends primarily upon the growth of epithelium. Boll had maintained, as a general principle, that in the development of organs the shaping is dependent on the co-operation of the epithelial and connective tissues. 10. His' embryo, XLIV. (Bff), described in his " Anat. mensch- licher Embryonen," Heft II., pp. 32 and 87, belonged to a chorionic vesicle measuring 7 by 8 mm. ; the vesicle was somewhat flattened, and on one part had fewer villi than elsewhere; the villi were branched. Closely attached to the inner surface was a small body 1.4 mm. long in its greatest diameter; the body consisted apparently of a yolk-sac and closed amnion ; of the embryo no further descrip- tion has yet been published. 11. The ovum described by Keibel, 90.1, consisted of a some- what flattened chorionic vesicle more than half covered with little viUi and containing a somewhat distorted embryo. The vesicle measured 8.5 by 7.75 by 6.0 mm. The villi were arranged in a band or zone leaving the two flattened poles of the ovvrni smooth as in Reichert's ovum ; the smooth areas were of very unequal size, at the edge of the smaller one the embryo was attached by means of its allantois-stalk to the inner surface of the chorion. The embryo, about 1 mm. long, was found twisted at its hind end, which was continued as an allantois-stalk attached to the chorion ; the stalk was nearly or quite as large as the embryo proper ; the yolk was broadly attached along nearly the whole length of the embryo, and opposite the embryo the yolk-sac was attached to the chorion as if the coelora had not completely developed. Sections showed that there was no medullary groove yet formed, but the amnion was already closed over the embryo. Keibel places his embryo as intermediate between His' embryo, XLIV., and Spec's embryo. Keibel's ovum resem- bled externally those of Reichert and Wharton Jones, and as it con- tained an embryo, he suggests that it is probable that the ova of Reichert and Jones also contained an embryo without medullary groove, but with an allantoic-stalk nearly as large as the embryo. But it seems tome that since Keibel's ovum is nearly twice as large, it cannot be of the same stage ; the presence of the equatorial zone of villi is explainable as an instance of retarded development. The ex- cessive variability of embryos is well known. IS. Spec's embryo, 89.1, was contained in a chorionic vesicle measuring, including the villi, 8.5 by 10 by 6.5 mm. The tips of the villi were attached to the surface of the decidual capsule. The embryo was attached by a very short allantoic-stalk to the chorion, and was closely invested by the amnion ; the attachment of the yolk- sac occupied nearly the entire length of the embryo, for the head-end had scarcely begun to project; the embryo was 1.54 mm. long; its dorsal surface was occupied by the very broad medullary plate of thickened ectoderm; as seen from above the plate seemed somewhat constricted in the middle of the embryo, owing to the arching of the body at that region ; the centre of the plate showed a narrow longi- tudinal furrow. Fig. 164, /; at the caudal end this furrow widened out and disappeared; just behind it was the open and relatively 203 THE EMBRYO. ^-P Fig. 164.— Cross-Section of Spee's Embryo, tion in text, Explana- large neurenterio canal behind which was the short remnant of the primitive streak. The embryo was cut into transverse sections, of which there were about 180, counting the allantois-stalk (Bauchstiel) ; section 81, counted from the head, is represented in Fig. 164; the ectoderm, efc, is very much thickened to constitute the medullary plate, Md; the narrow central longitudinal furrow, /, mentioned above is very noticeable; outside of the embryo the ectoderm is reflected on to the amnion, ct, over the back of the em- bryo. The entoderm, en, is a thin layer of cells in the centre of which the noto- chordal band can be distin- guished ; in sections nearer the neurenteric canal the band is better marked, being there much thicker than the remaining entoderm. The mesoderm, me, is a distinct layer, although, as other sections show, it is fused in the median line of the primitive streak behind the neurenteric canal with both ectoderm and entoderm. The embryonic ccslom has only just begun to appear as a small fissure, p, but the extra embryonic ccslom is completed, so that out- side the embryo the mesoderm is completely divided into a somatic leaf, ct, which helps form the amnion and chorion, and a splanchnic leaf, df, which forms one layer of the wall, of the yolk-sac. The sections through the head-end show that the head had grown forward far enough to lead the separation of the very short vorderdarm; sections through the allantois-stalk showed that the allantoic diverticulum extended as a small canal through the great accumula- tion of mesoderm; throughout the rest of its extent the archenteron is nowhere differ- entiated from the yolk-sac. Fig. 165 is a section passing through the neurenteric canal, which leads through the centre of the medullary plate into the wide yolk-sac ; the part of the sac farthest from the embryo has its mesoderm thickened and vascularized, the vessels containing young blood-cells often in some stage of division. The cho- rion of Spee's embryo had a layer of meso- derm, with cells of a well-marked mesenchy- mal type, and an outer layer of ectoderm consisting of a thinner outer layer darkly stained, without distinct cell boundaries, but with Fig. 165. — Section Passing through the Blastopore of Spee's Embryo, am^ Amnion ; efc, ecto- derm; ct, amniotic mesoderm; g, meeting point of somatopleure and splanchnopleure; d/, meso- derm of yolk-sac ; o 5 6, blood vessels; eji, entoderm; n, blas- topore; d, cavity of yolk-sac. After Graf Spee. KNOWN HUMAN. OVA. 293 •iM ^^^^ small nuclei and an inner lighter layer of distinct cells with larger nuclei ; the ectoderm appeared somewhat as if ciliated. Unfortun- ately Spee gives no account of the villi beyond a few words to say that they resembled those of later stages. 13. We come now to the embryos with a well-developed medul- lary groove; the number of these is four. Their probable age is about fourteen days. The least advanced is His' embryo E (" Anat. mensch. Embryonen, " I. , Heft I. , p. 145) , of which only His' sketches are avail- able, the attempt to microtome the specimen not having been fortu- nate. The ovum was presumably normal; it measured 8.5 by 5.-5 mm., and was entirely covered by short branching villi. For the convenience of the reader I have ■constructed from the author's sketches and descriptions the ac- companying diagram. His states that the chorionic vesicle bore at one point a thick stalk, Al, which ran to the posterior end of the ■embryo ; the length of the embryo from the anterior extremity to the base of the stalk was 3.6 mm. The head-end of the embryo was somewhat thickened, and appar- ently showed the medullary groove still open. The small, round yolk- sac had a broad connection with the ventral surface of the embryo. The amnion sprang from the allan- tois and passed over the head of the embryo. The disposition of the caudal extremity was not made out. There were no limbs, gill- clefts, nor organs of any kind discernible— not even a protuberance between the head and yolk-sac, such as marks the position of the heart in older embrj-os. 14. Allen Thomson, 39.1, published an excellent article on young human ova in 18-30. He gives a very good critical review of what previous authors had written, and describes himself three embryos, which have become classical, for the figures and descriptions given of them by Thomson have been copied again and again. They are especially known by the reproductions in KoUiker's " Embryologies," and in Quain's "Anatomy." Two of these embryos (numbered I. and II. by Thomson) belong in the group we are now considering. I cannot, however, admit at present that either of them is certainly fully normal, though perhaps they are only slightly malformed. In number I., (see KoUiker's " Grundriss," 1884, Fig. 112, and "Ent- wickelungsgeschichte," 1879, Fig. 225) the yolk-sac was abnormally dilated and the characteristics of the embryo were not ascertained. His ("Anat. Emb." Heft II., pp. 35-36) has shown that the embryo proper was not observed, and that what Thomson called the embryo Was really only the amnion, springing from the allantois-stalk and passing over the embryo. Kolliker questions the accuracy of this in- terpretation, but upon what ground is not evident, for, so far as I can Diagram of His' Embryo E; Age fourteen (?) days; length about S.3 mm. The embryo is not repr^ented in quite its natural attitude ; the proportion of the parts is not ac- curate ; the villi of the chorion and the vessels on the yolk are purely diagrammatic as to their number and shape. .Em6, Embryo; Al, posed stalk of the allantois. . sup- 294 THE EMBRYO. see, it accords perfectly with our present knowledge. The embryo in question was presumably little advanced beyond His' embryo E, Fig. 165, but had an abnormally hypertrophied yolk-sac. As no sufficient description of the embryo exists, and as it is quite certain that the specimen was more or less abnormal, it cannot be longer regarded as a fair representative of a young ovum. 15. The third embryo of this group. His' SR (I., Heft I., 140-144) measured 2.2 mm. in length, and was probably fourteen days old. The chorion was 9 by 8 mm. in diameter. It shows considerable advance of development beyond the three embryos above considered. The neck of the yolk-sac is already somewhat contracted, or, in other words, .the connection between the embryo and the yolk-sac is no longer so broad and long as it was. The head is considerably en- larged ; between it and the anterior wall of the yolk-sac is a large thickening corresponding to the heart. From the under side of the caudal extremity runs off the stalk of the allantois, which is still short and thick; the amnion lies quite close to the embryo; the medullary ridges are still separated by an open, though deep, and relatively narrow groove; myotomes (protovertebrae, auct. ) are present, but their number was not ascertained. When the embryo is viewed in profile, the middle of the back shows a marked concavity which has been noticed in other older embryos, and is probably an artificial distortion. We shall have to return to this matter. Small openings were visible on the inner surface of the chorion. These I take to be the openings to the still hollow villi, such as have been seen in both younger and older ova. His attempted to obtain sec- tions of his specimen, but when cut the sections fell into fragments. 16. Much more valuable is the account of Thomson's second ovum, which he had better opportunities of studying. The original description has been supplemented by His, " Anat. Embry.," II., p- 34, who examined Thomson's original drawings, and called attention tO' an important error in the engraving in Thomson's plate. Kolliker, however, still reproduces the incorrect figure in the second edition of his "Grundriss," Fig. 114. An erroneous figure is also reproduced in Ecker's "Icones," Taf. XXV., Fig. 3. The chorionic vesicle measured O.CO by 0.45 of an inch, and was covered with branching villi . The contained embryo was very small ; according to Kolliker, only 2.5 mm. The embryo rested upon the round yolk-sac of 2.3 mm. The embryo consisted of two thick longitudinal ridges, Fig. 167, A, which curved round in front so as to become con- tinuous with one another, and were broken^ off posteriorly— an important fact noted by -Thomson's Second His (c/. sttp.) . Thcso ridges are presumably Bjemiir^ from'Sehind" ^ °^'^' the medullary folds. At the hind-end of the embryo was a tear, making a hole into the hollow yolk-sac. As His suggests, this is probably where the allantois was inserted and broken off. No amnion was observed, * It must be remembered that the term protovertebrae is an entire misnomer, and is inherited from the time when the primitive muscular segments (myotomes) were mistalren for the com- mencements of the vertebra. KNOWN HUMAN OVA. 395 and the nature of the connection of the embryo with the chorion was not ascertained. What we learn from this embryo is something more definite than is afforded by His' observations as to the size and disposition of the medullary ridges and the hoUowness of the yolk-sac. The apparent hypertrophy of the chorion enforces caution as to accepting the embryo as normal ; but it is not rare to find in abortions a small typical embryo with an enormously dilated chorion, so that it is not impossible that the embryo in the present case was quite normal. 17. Spee has briefly described a second ovum, but his ac- count is not now accessible to me. According to the notes given by Fr. Keibel, 90.1, 261, the chorionic vesicle measured 15x14x10 mm., ^he yolk-sac 3.5 mm. The embryo had seven myotomes, and its age in maximo was thirteen daj^s. 18. J. Kollmann, 89.1, 108-121, describes an embryo of about 2.3 mm. ; the yolk-sac was attached to the embryo, Fig. 168, for a dis- (M ^:.-J-.r ■:■■ --^ Yk^ ^.,.^ -""> Fio. 168.— Human Embryo of Thirteen to Fourteen Days, ^m, Amnion ; S. 7, seventh segment; Md, medullary groove, still open; Ht, heart; Yk.s, yolk-sac; Al, allantois-stalk. Alter J. Kollmann. tance of 1.5 mm., leaving the head to project 0.58 mm., the tail to project 0..3 mm. The head is already somewhat enlarged and slightly bent over ventralward; it forms at least a third of the whole embryo; there were thirteen* primitive segments which marked themselves externally ; the segmented region of the body is bent so that its dorsal outline is concave ; the medullary groove is open throughout the anterior two-thirds of its length, but the caudal third is closed ; the tail is slightly curled over, and is connected on its under side with a thick, short allantois-stalk, or Bauchstiel, by which the embryo is attached to the chorion ; there are no visceral or branchial arches, although the gill pouches may have begun form- ing in the pharynx ; no anlage of the eye or ear could be distin- guished; the oral invagination has formed, but the oral plate {Each- enhaut) is still intact; the heart is not straight but an already much bent tube, which receives at its hind end the two veins from the yolk-sac, which consists of vascularized mesoderm and an entodermal lining. The amnion was a thin, transparent membrane springing from the body of the embryo close around the yolk-sac, and envelop- ing the embryo very closely. The chorion formed a vesicle covered * The figure shows fifteen segments. 296 THE EMBRyO. externally by branching villi ; its diameter including the villi was 18. cm. Although the data were not very satisfactory, Kollmann esti- mated the age of this specimen to be thirteen to fourteen days. 19. The description of the ovum of thirteen to fourteen days by Von Baer, 88.1, was drawn up over sixty years ago. _ The ovum measured a little over three lines, and was covered with villi ; the embryo was about two-thirds of a line long ; Von Baer appears to have recognized the amnion and yolk-sac and to have seen the allan- toic-stalk (his Harnsack), though he did not observe its connection with the chorion ; as he states that the back was already formed, it is probable that the medullary groove was closed. It is with much hesitation that I place the embryo here in the series. There has been, so far as I am aware, no human embr^ with one gill-cleft described, unless, indeed, Coste's embryo described be- low was such. But sev- eral with two clefts marked externally have been described, most of them by His. Those of them which can be as- sumed to be normal present a remarkable bend in the back or dorsal flexure, by which their shape is so much altered from that of the slightly younger stage, and so unlike that of the next older stage, that the embryos with the dor- sal flexure might be considered abnormal had we not positive reasons to the contrary. In- deed it seems probable that embryos in this stage may have been, because assumed to be abnormal, discarded. His' embryo L, described below, and perhaps Coste's, p. 300, both probably belong in this stage and were artificially straightened out. Nothing similar to the dorsal flexure of the human embryo has been observed in any other vertebrate, though it may occur in apes and monkeys. 20, 21. Two specimens in my collection are in this stage. The younger of these is represented in Fig. 169, and is very near the embryo designated as Lg by His ; just behind the heart the whole body bends downward and then bends abruptly upward, so that the caudal end of the embryo runs nearly at right angles to the pharyn- geal region ; from the under side of the tail end runs off the thick allantoic-stalk by which the embryo was attached to the chorion. The other features observed are shown in the figures. Sections Pig. 169.— Embryo of the Beginning of Third Week (Minot Coll., No. 195). AIL AUantois; Am, amnion; 6r, branchial region; H, fore-brain; Hr, heart; Yk, yolk. KNOWN HUMAN OVA. 397 showed that the specimen was imperfectly preserved, and I cannot be sure that it was entirely normal in shape, though it differs but little from the certainly normal embryos of His. My second speci- men (Coll. No. 143) is a little older, I think, but as it is somewhat distorted, it is hardly worth figuring and describing separately. 22, 23. Far better preserved are the two embryos of His, which he has studied with such splendid thoroughness. He designates them as Lg (or LXVIII.) and Sch. 1, (orLXVI.), Fig. 17, p. 39. They resemble one another very closely, the most marked differences being that in Sch the heart is more exposed and the neck of the yolk- sac more constricted than in Lg. Lg measured 2.15 mm. ; Sch, 3.20 mm. The differences noted indicate that the latter is slightly more advanced. The following description applies especially to Lg. In «xtei'nal form the embryo is very similar to Minot's Fig. 169, but no trace of a third gill-cleft was visible externallj', and the amnion was attached along nearly the entire length of the allantois-stalk (His' Bauchstiel) . The anatomy can be understood from the accompany- ing Fig. 170. The head bend -being well marked, the central nervous system makes at the mid-brain, a bend at nearly a right angle, so that the fore- brain is brought very near the heart, which lies in the large pericardial sac, which protrudes conspicuously between the head of the embyro and the yolk- sac. Between the head and the peri- cardial sac is situated the oral invagi- nation or future mouth cavity, separated from the vorderdarm by an intact oral Plate {Rachenhaut) o.pl. As regards the archenteron we find the vorderdarm above the heart, Ht, with two gill pouches formed at its head-end and its lower end widened ; out of this wider part the lungs and the stomach are to be differentiated in later stages; the vorderdarm is compressed dorso-ven- trally but widely expanded transversely ; the middle portion of the archenteron opens widely into the j'olk-sac ; where the vorderdarm joins this middle divi- sion is found the outgrowth of the liver, Li, extending toward the heart ; in the posterior region of the embryo the arch- enteron has also become distinct from the yolk-sac and ends with a dilatation (His' bursa) in the tail of the embryo ; from the under side of the bursa runs out the allantoic diverticulum. All, which extends as a narrow tube of entoderm through the allantoic stalk to the level of the chorion where it ends blindly. The central nervous system forms in bulk a very large part of the embryo ; from the fore-brain the optic vesicles, Op, have grown out ; the mid-brain is only slightly dilated ; the hind- brain is as long as the mid- and fore-brain together, and is nearly as Fio. 170.— Human Embryo of 2.15 mm ; Anatomy Reconstructed from the Sections. Op, Optic vesicle ; o.y i, oral plate ; Ht, endothelial heart ; Li, liver ; Om, omphalo - mesaraic vein ; YJc, yolk-sac; All, allantoic diverticulum of archenteron ; Oi,otocyst; Ao, aorta; U.V., umbilical vein. After W. His. 398 THE EMBRYO. long as the vorderdarm, which it overlies ; near the centre of the hind- brain lies the open ectodermal invagination, Ot, destined to form the auditory vesicle or otocyst ; the remainder of the medullary canal corresponds to the future spinal cord and gradually tapers tailward; alongside it His was able to distinguish in Lg twenty-nine myotomes. The heart, Ht, is very largely and asymmetrically bent ; the heart at this stage and for some time later may be described as consisting of two tubes, a small inner one, Ht, formed of endothelial cells, and a larger outer one formed chieily of contractile elements, which are gradually differentiated into the striated muscles of the adult heart. The way in which the heart is bent can be best seen in front views ; the great veins enter the heart in the median line just above the liver ; the heart tube runs toward the head and the left side, making the auric- ular limbs ; then the tube bends to the ventral side and runs obliquely backward to the right side, making the ventricular limb, and finally takes a curving course as indicated in the figures to the median line, and ends close behind the mouth ; this third part is the aortic limb. The endothelial heart tube is continued beyond the pericardial cavity as the aorta, which soon divides into two branches oa each side, which pass up around the pharynx, one branch in front of each gill- cleft ; the front branch curves over, and, passing tailward, joins the second branch; the branches which pass around the pharynx are known as the aortic arches ; the united vessels run toward the tail on the dorsal side of the pharnyx ; they are called the dorsal aortae, and by uniting in the median line form the single dorsal aorta, which runs away back nearly to the tail of the embryo, where it forks, and its branches, passing one on each side of the intestinal canal, enter the allantois-stalk and run to the chorion, where they branch out. The veins of the embryo are tht jugular, which comes from the head and meets cardinal vein from the rump about at the level of the liver ; these two veins unite as a short stem, which runs transversely toward the venous end of the heart and is termed the ductus Cuvieri ; the ductus is joined, as in adult fishes, by the omphalo- mesaraic vein, Oin coming on the same side from the j-olk-sac, and the umbilical vein, ti.v, coming from the allantois; the four united veins meet their fellows from the opposite side and form with them the^median sinus reuniens, which communicates directly with the heart ; the course of the umbilical vein is curious, as it takes a short cut frora the allantois through the somatopleure along the base of the amnion to the heart ; how this course is possible can be under- stood by comparing figures 17 and 166. 24. We pass now to His' embryo L, and Coste's youngest em- bryo. It must be seriously doubted whether either of these embryos represent the normal shape. The former had two gill-slits and parts of it were torn away, so that we may surmise that it had had the dorsal flexure but was artificially straightened. Concerning Coste's embryO' see the next paragraph. His' embryo L is described in his " An at. menschl. Embryonen," Heft I., pp. 13.5-139. It measured 3.4 mm. in length, and was obtained from a chorionic vesicle of 8 to 9 mm. diameter. The specimen had been considerably injured, and no exact knowledge could be obtained in regard to the heart or the disposition of the allantois or the amnion. Precisely these three points are- KNOWN HUMAN OVA. 299 elucidated by Coete, while His has worked out the internal anatomy of his specimen ; in short, the two descriptions complement one an- other in a remarkable manner. Nearly all that His ascertained is represented in the accompanying illustrations, Fig. 171. A gives a side view showing the thickening of the head-end and the upward Fig. 171. — His^ Embryo Ij, 2.4 mm. long. A, Side view; B, ventral view; C, ventral view, with the walls of the body and intestine seen in frontal section ; D. dorsal view, showing the central nervous system, itf. Mouth ; Mx^ inferior maxilla or mandible ; 2, hyoid arch ; Fd, vor- derdarm ; F, splanchnopleura of the yolk-sac ; 2, 3, and 4, gill arches ; Coe, ccelom or primi- tive body-cavity ; Op, optic vesicle ; Au, auditory vesicle (otocyst) ; a. point where the medul- lary groove has not yet closed. curving of the tail, and the two gill-slits in the cervical region ; the mouth, M, is very large ; between it and the first gill-slit intervenes the thick ridge, Mx, of the first gill arch (branchial or visceral arch, auct.), which becomes the mandible; between the two slits is the second or hyoidean arch, in connection with which the hyoid bone afterward arises. A large body cavity is present, C, Coe; the walls of the body (somatopleures) pass over along an extended line into the amnion ; the connection between the embryo and the yolk-sac is al- ready much restricted compared with Coste's embryo. Fig. 172; at the side of the head a line and shadow mark the position of the optic vesicle. B is a ventral view ; it shows the large wide mouth, M, which, according to His, was apparently in communication with the intestinal canal, which is nothing but a straight tube with a great pharyngeal dilatation, and a wide open union with the yolk-sac ; the median light band shown at the back of the nwuth is the central nervous system shining through the covering tissue. C is intended to show the digestive tract, and is partly a horizontal section. Es- pecially to be noticed is the enormous size of the pharynx (the region of the branchial arches) , the straight, short intestine, and on each side of the latter the distinct body-cavity, Coe; there are indications of four visceral arches, Mx, 2, 3, and Jf; in front of the pharynx is 300 THE EMBRYO. shown the ventral surface of the fore-brain or first cerebral vesicle, with its lateral diverticula, the optic vesicles. D is a dorsal view of the brain and medullary canal which is still open at a. The brain and spinal cord are already differentiated by the dilatation of the former. The brain subdivides very early in all vertebrate embryos into three dilatations or primary vesicles; but in this embryo the two anterior dilatations are not yet clearly separated from one another, hence there is only one widening of the brain in front ; the front end is seen to bend downward and give off the conspicuous optic vesicles. Op, which, therefore, arise before there is any trace of the cerebral hemisphere — an important fact ; the posterior and larger dilatation is the primitive medulla oblongata; no trace of the cerebellum has appeared. The whole nervous system is a tube the walls of which are of nearly uniform thickness, except that the dorsal wall of the third vesicle (the cavity of which becomes the fourth ventricle of the adult) is very thin. This thin wall is persistent in the adult and never develops into nervous substance. On each side of the medulla lies a round cyst, the auditory sac, Au, the beginning of the adult membranous labyrinth. Three other points not shown in the figures remain to be noticed. 1 . In the tissue at the back of each body- cavity, Coe, was found a single longitudinal epithelial canal, the "Wolffian duct, the first part of the urogenital apparatus to be devel- oped. 2. Close below the nervous system lay a median' rod of cells with a small central cavity ; this rod is the notochord or chorda dorsalis, the primitive embryonic axis around which the vertebrae are formed later. 3. All the tissues are stiU embryonic — that is, the cells are not yet differentiated into tissues. Unfortunately, the number and disposition of the myotomes were not ascertained. 24A. Janosik, 87. 1 , describes an embryo with two gill pouches and three aortic arches, giving a few anatomical details. 25. Coste's embryo has been beautifully figured in his great work, 47.1. It is possible that it really belongs to an older stage with the dorsal bend, compare Fig. 169, and that it was stretched out by Coste ; the difficulty of assigning it its place is due to the entire uncertainty as to its actual dimensions. Coste's private collection is, I believe, now in the College of France, but upon search this particular specimen could not be found, so that His' inquiries to ascertain its actual length were resultless. KoUiker states that it was 4.4 mm. long, but his authority for the statement is not given; the measure was probably taken from Coste's figure, "grandeur naturelle." Since embryos of this length are far more developed than Coste's, it is probable that Coste's data as to the magnification of his figures are inaccurate. If we assume the embryo to have been really about 2.5 mm., it will then agree, except as to the great length of the rump, very closely with what we know otherwise of such young embryos. I give the accompanying figures, which are careful copies from the original plates published by Coste (4 "Esp^ce humaine," PI. II.), whose illustrations, made by his assistant, Gerbe, have never been surpassed for beauty and life-like accuracy. The embryo in question was inclosed in a villous chorion, Fig. 172, and was provided with a large vitelline sac, Vi, having a very broad connection with the embryo and covered with a network of vessels, in which was a KNOWN HUMAN OVA. 301 fluid not yet red. A thick allantois-stalk, Al, can be seen running^ from the under side of the embryo's tail to the chorion ; from the anterior side of tlae stalk springs the amnion, Am, completely inclos- ,^"'>n r/'l I "-A^ ii * Fig 172. —Ovum Supposed to be from Fifteen to Eighteen Days Old ; after Coste. The chorion has been opened and spread out to show the embryo and its annexa. Al, AUantois ; Am, amnion surrounding the embryo. ing the embryo. It is important to notice that in this, as in still older embryos, the disposition of the amnion is essentially the same as in the earliest stages {v. sup.) ; the line of attachment of the am- nion is down the sides of the allantois and around the embryo about 302 THE EMBRYO. on a line with the top of the yolk. As regards the embryo, it is drawn slightly canted on to its left side ; its back is concave ; the head-end is thickest and shows three gill-arches, hence there were probably two branchial clefts ; behind and be- low the gill-clefts can be seen the heart, already a bent tube, shining through ; behind the arches again, but on the dorsal side, the light-looking oesophagus is distinguishable ; in the figure a wedge-shaped shadow intervenes between the straight oesophagus and the bent heart; the heart causes a conspicuous bulging of the body between the head and the yolk-sac ; the caudal extremity i'j thick and rounded, and curves upward. Big. 173 is a ventral view of the same embryo after most of the yolk-sac has been cut off; its walls, Spl (splanchnopleure) , are seen to pass over without any break into those of the intestinal cavity. In the central line the chorda dorsalis, s, can be perceived through the trans- lucent dorsal wall of the intestinal cavity ; it is flanked on each side by the row of square mus- cular segments (myotomes). We see the large allantois, Al, behind, and in front the tubular heart, Ht, with a decided flexure to the right of the embryo ; the anterior end of the heart makes an opposite bend, separating off a limb, which becomes the bulbus aortce. The chorion con- sisted of two membranes, one of which passes tinuously over the inner surface of the chorion, while the other outer mem- brane alone forms the hol- low villi. Figs. 172 and 176; hence, in looking at the inside of the chorion, we see numerous round openings which do not penetrate the in- ner membrane. Fortunately we learn from KoUiker ("Entwickelungsgeschichte," 1879, p 309) who had an opportunity in 1861 to exam- ine the chorion, that the outer membrane was epithelial with cells of the same character as in the epithelium of older vascularized villi,* and that the inner layer consisted of developing connective tissue, and carried fine blood-vessels. It thus appears that Coste was the first to observe the role of the epi- iheliiun in the growth of the villi. 26, 37, 28. It will be as well to mention here, rather than later, Fig. 173. — Embryo Sup- posed to be from Fifteen to Eighteen Days Old; after Coste. Ventral view; the vitelline sac has been re- moved. Am, Amnion: Ht, heart; SpZ, splanchnopleure, ■extending beyond the em- bryo to form the yolk-sac; S, chorda dorsalis with a row of myotomes on each side. Al, stalk of the allan- tois. con- <; Fig. 174. —Fragment of the Chorion of Fig. 4, highly magnified. Ec, Epithelial layer, Mes, connective-tis- sue layer; Fi,chorionic villi, formed wholly of epithe- lium. * Hierbei Zeigte sich, dass die Zotten und die sie tragende Haut ganz und gar aus epithelarti- gen Zellen, vonderselhen Beschaflenheit wie des Epithels der spateren gefasshaltigen Choriou- -zotten bestehen. "— Kijlliker, I. c. KNOWN HUMAN OVA. 303 iliree descriptions of young embryos, which either belong in this ' stage or are a little older. Of these descriptions Remy's alone brings much of any positive information, but the size and age of his embryo can only be guessed at. The first of the embryos is Schroder van der Kolk's (51.1, p. 106 j^^., with figures on PI. II.). Kolk's figures are not very clear. He states that his specimen had two gill-clefts and measured 1.8 mm. in length; one can but ask, Was it not really larger? Kolk's figure suggests that the specimen was doubled up ; if this was the case, the embryo, when straightened out, would agree fairly well with His' embryo L, above described. Professor His, for reasons not clear to me, considers Kolk's specimen as somewhat older, but to this opinion I am unwilling to accede. The second embryo is that of Hennig, whose description, 73.1, leaves very much, and whose figures leave everything to be desired. From this paper we can gather very little, except confirmation of Coste's state- ments in regard to, (1) the disposition of the amnion and its connec- tion with the stalk of the allantois ; (2) the absence of a yolk-stalk. Schwabe's, 79. 1, embryo, to which reference has alreadj' been made, and which he assumes to be thirteen to fifteen days old, was probably sixteen to twenty days old, as shown both by his own data and by the description of the ovum. Very likely it was a little younger than Coste's embryo, v. sup. There were a well-developed yolk-sac and an amnion closely investing the embryo, which was connected with the chorion by a short allantoic stem. The chorionic villi were considerably branched and entirely filled with mesoderm ; their tips had little thickenings of the epithelium by which they were attached to the decidua; this was the only connection between the foetal and maternal tissues. This last fact is an interesting confirmation of the observations of Ahlfeld and Langhans. Remy's embryo, 80. 1, was also a young one, but its exact age is not stated, nor are the measures of its length given except in the title, where it is called " long d'un centimetre." From the stage of development, and from the state- ment in the text that the chorionic cavity measured 30x10 mm., it seems impossible that the embryo was so large ; we should rather expect an embryo of 3 mm. Remy's figure is too inexact for one to make out the form of the embryo. If he gives the length cor- rectly, the specimen must have been a month old. As to its struc- ure, Remy gives the following details : The medullary canal was still united with the ectoderm at its lower end, and extensively so over the fourth ventricle, which was entirely closed. The heart already had muscular striae. The epidermis had two layers of cells, the outer somewhat fiattened, the inner cuboidal. The cutis was not differentiated. The epithelium of the chorion he describes as maternal — a common error. He also distinguished the inner mem- brane of the chorion, the allantoic. He has also seen, apparentlj-, what is known as Langhans' cellular layer, but has taken it for a deep portion of the epithelium, which he accordingly calls many-layered. The stage with three gill-clefts is known through five embryos, four of which have been studied by His, and belong to the end of this stage, since in all, except one (Rf) , of which we have no detailed description, the fourth gill-pouch of the pharynx was partly formed, and in all there were five aortic arches. The fifth embryo is de- 304 THE EMBRYO. scribed by Chiarugi, and had three gill-clefts and three aortic arches j it therefore belongs to the beginning of this stage. 28 A. Chiarugi's embryo, 88. 1, had a very marked dorsal flexure {insenatura dorsale); its greatest length was 2.6 mm.; its chori- onic vesicle measured 15x12x8 mm. ; the villi were much longer (1.5 mm. ) than upon the other. The embryo had three gill-clefts showing externally, and unlike the two embryos of His, BB, Lr, only three internal gill-pouches and three aortic arches ; the otocyst was closed but stiU connected with the ectoderm ; the yolk-sac had a broad con- nection with embryo, and measured in vertical diameter 1.9 mm.; in transverse, 1.8 mm. ; in antero-posterior, 1.6 mm. These points show that the embryo was intermediate between His' L and M. In Chiarugi's specimen the WolflSan bodies had become protuberant; the cephalic and spinal ganglia were present, but the spinal motor roots were not developed ; the notochord measiired 30/j- in transverse, 24/Ji in dorso-ventral diameter, and its caudal termination was indis- tinct. Chiarugi gives a full and admirable description of all the parts, but as in the respects not specially mentioned above, the structure is very similar to that of other embryos with three gill- clefts, further details may be omitted. 29—32. The four embryos with three gill-clefts described by His have been designated by him as Rf ; M, Fig. 175 — BB, and Lr, Fig. 16 — they being named in the presumable order of development. M and Lr are pro- bably the most perfect ; Rf is I somewhatroUed up; BB has a "A distinct dorsal ^ ' I ^- flexure, but, as ' ' '"■ ?(!. ,j;^i? His himself re- " f s / " '"■ ' ' marks, this was ' /* ^ ij ^ ^ . \ probably due to -»*.!! .,«Mr ..^-aaegs-^ ^ ^ mecha n i c a 1 ' strain and is ar- ; tificial; hence we may assume that in all em- bryos of this stage the dorsal flexure has dis- appeared and the back has be- c o m e convex. The four embryos are described and figured in His' "Anatomie menschlicher Embryonen, " Heft I. -III. " Of M a systematic anatom- ical description is given (Heft I., 166-134), and additional details con- cerning BB and Lr are scattered through Heft III. The lengths are : M, 2.6 mm. ; BB, 3.2 mm. ; Lr, 4.2 mm. ; Rf being rolled up could not be measured satisfactorily. The chorionic vesicle of M measured 7.5x8.0 mm. ; of BB, 11x14 mm. From the data given by His, the age of BB may be estimated at probably twenty to twenty-one days. r" >• -His' embryo M. KNOWN HUMAN OVA. 305 Digestive Canal of His' Embryo Lr, 4.2 mm. lon^. (Compare Fig. 16, p. 38, and Fig. 444.) The head is bent down, the back very convex, and the caudal ex- tremity is rolled up and turned toward the right — in Lr, however, to the left — while the head is twisted slightly toward the left ; the long axis of the body, therefore, describes a large segment of a spiral revolution; the spiral form is more marked in embryos a little older ; it is, of course, produced by the more rapid growth of one side; in view of the differences between right and left in the adult, it is very interesting to find differences between symmetrical parts showing so very early in the heart of the embryo and the twisting of the body. The caudal end of the body has grown very much; the allantois- stalk has presumably lengthened ; the neck of the yolk-sac is much constricted ; the gill- clefts can be distinguished externally; the otocyst, Fig. 178, ot, has become somewhat pear-shaped. The neural canal is completely closed; the mid-brain and fore-brain have become perfectly distinct, and the latter has begun to form the hemispheres in front. The mouth is large, and at its upper corner the protuberance of the maxillary process is marked; the mandibular process is very prominent. Fig. 176, a geometrical recon- struction from the sections, shows the anatomy of the entodermic canal. The pharynx, bounded on each side by four branchial arches, is still very large and tapers down posteriorly ; the intestine is turned to the left and opens into the broad canal, Tics, of the yolk-sac ; just in front of the yolk-sac there is a small ventral diverticu- lum, Li. , the commencement of the liver ; behind the yolk-sac the cylindrical intestine runs over into the tail, where it expands into the biirsa of His, and gives off a cylin- drical canal, which has very thick connec- tive-tissue walls, and is the allantoic-stalk, Al, which carries the two allantoic veins and the two large allantoic arteries. Fig. 178. Fig. 177 gives a view of the anterior wall of the pharynx of BB; in front is the M^-ivKw.a;:,;. large opening of the mouth, M, the oral T, ^".-^ . . ^- -,,7 „' «.v plate between the mouth cavity and the Fig. 177.— Anterior Wall of the J^,. . . ,. tH ^^ Pharynx of His' Embryo BB, 3.2 vorderdarm having disappeared ; the wide Sdi ionteini^g™''fortfo'areh pharynx shows four gill-pouches, and at its shown by dotted lines; 5, fifth lower end gradually contracts and passes aortic arch; M, mouth ; Oe, . ^ -^ , mi j.- oesophagus or vorderdarm; Coe, mto the narrOW CBSOphagUS. Ihe aortlC coeiom. After w. His. vessels are indicated by dotted hnes; the cardiac aorta reaches the pharynx between the bases of the second and third gill-arches, and divides into two branches on each side ; the an- terior branch forks and runs through the first and second arches; the 30 30G THE EMBRYO. posterior branch forks, one fork going to the third, and the other after again forking supplies the fourth and fifth arches ; this arrange- ment of the aorta is typical. Between the bases of the first and second arches is a small protuberance which is the anlage of the tongue, and is named by His the tuherculum inpar. The body- cavity of the abdomen has on each side of its dorsal surface a longi- tudinal ridge, the commencement of the Wolffian body ; the ridge already contains traces of the canals of the Wolffian body. Of spe- cial interest is the arrangement of the circulatory apparatus. Fig. 178. In the figure the arteries are °P' shaded dark. The heart is an S- "K\ shaped tube, the venous end is con- Fio. 178.— W. His' Embryo M. op, Optic vesicle; A^ aorta; Orn^ omphalo-mesaraic vein ; Au^ arteria.' umbilicales; All, allantois; Car, cardinal veins; Vh, right umbilical vein ; Aq, dorsal aorta ; Jg, jug- ular vein ; of, otocyst. After W. His. Fig. 179.— Reconstruction of His' Em- bryo BB, 3. 2 mm. long, to show the Course of the Endothelial Heart, Ht, and aortic arches. Op., Optic vesicle; Ht, heart; Li, liver: 1^, aortic arches; V, allantoic vein; Ait, auricle. After W. His. vex toward the head, the arterial end convex toward the tail; when viewed from in front the venous portion is seen on the left, Fig. 179, the arterial portion on the right of the embryo. The heart is con- tinued forward by the large aorta, Ao, which gives off five branches on each side of the neck ; these branches unite again on the dorsal side and run backward to unite with the fellow-stem, and so form the single median dorsal aorta, Ao, which runs way back and terminates in two large branches. Fig. 178, Au, which curving round pass out through the allantois-stalk. The five branches in the neck are known as the aortic arches, and the column of tissue around- each branch constitutes a so-called branchial or visceral arch; between the five arches are four spaces, in each of which a gill-cleft is ulti- mately formed. The reconstruction of Lr in a side view. Fig. 180, affords further information concerning the disposition of the heart and large blood-vessels. The veins, as is there shown, are, 1, the jugular, J, and cardinal, car; which unite and form a single KNOWN HUMAN OVA. 307 transverse stem, the ductus Cuvieri, D.C.; the cardinal veins receive chiefly the blood from the Wolffian bodies and atrophy later with those bodies; 2, the large umbilical veins which pass up, Al.v. from the allantois and also open into the ducti Cuvieri, but nearer the heart than the jugulars and cardinals; 3, the omphalo-mesaraic veins, Om, which come up from the yolk-sac. More precise details of the course of the veins through the region of the liver will be found in Chapter XXIX. The conformation of the body- cavity (splanchnocosle) can be better con- sidered in connection with the his- tory of the septum transversum, Chapter XXII. 33—36. Of other embryos about the stage of those described in the preceding pages several are known . D c His has referred the following to this stage : 1. Allen Thomson's ovum III. (2), 39.1. 2. C. E. von Baer's described in his " Entwickelungsgeschichte," Bd. II., 361-363, Taf. VI., Figs. 15-19 ; also in Von Siebold's Jour- nal fiir Geburtshiilfe (1834), XIV., 409. 3. Schroeder van der Kolk's (5) , 51.1. 4. Alexander Ecker's (9) 73.1. 5. Prof. Hecker's {vide infra). 6. Beigel's {vide infra). 7. Bruch's (10). Of these Thomson's embryo, Fir. 180.— Reconstruction of His' Embryo, £r, (Fig. 16). Ot, Otoeyst; J, jugular; car, ,i™ f T ■ 1 -, 1 • "i" carotid; /, first aortic arch ; .4m, auricle; Feii, the ngure or which reduced m scale ventricle; U, llver; om, omphalo-mesaraic be found in His <■" Ar,al- yein;^^ allantoic divertlpulum; 4,-i, allan- toic artery; Al. v, allantoic vein; Am, origin of the amnion; D. C, ductus Cuvieri. After W. His. may be found an Mis ("Anat. menschl. Embryonen," Heft II., Fig. 18, p. 33, marked A. T.3), is the only one deserving much attention. Thomson's embryo re- sembles His's M (see below) quite closely, not only in general form but also in the possession of distinct gill-clefts and the great prom- inence of the heart. Its length is given by Thomson at one- eighth of an inch, about 3 mm. Von Baer's embryo, on the contrary, was only 2 mm. long; it was surrounded by an amnion of about 4.5 mm. diameter, which is abnormally large; Von Baer observed four open gill-slits ; the hind end of the body was partially atrophied, which accounts for the short length. Van der Kolk's embryo, as I have already stated, I refer not to this but to the pre- vious stage, perhaps mistakenly, but I think not. In Ecker's ovum the chorion measured 12 by 9 mm., and the embryo only 2 mm; the author's description is very meagre and his figures not distinct; Ecker expressly compares it with an ovum of Wagner's, figured in 308 THE EMBRYO. Wagner's "Icones Physiologicae," and again in Ecker's "Icones Physiologicae," Taf. XXV., Fig. V. ; but ttie comparison apparently refers only to the chorion, for Wagner's embryo was evidently older, being 4.5 mm. long and having external traces of limbs. Hecker's ovum (5) I know only through Prof. His' reference, which leaves the impression that Hecker's description is so unsatisfactory as to render it a matter of surmise exactly what stage of development the specimen had reached. In regard to Beigel's ovum I have already expressed, p. 289, my ojoinion that it is a much older and abnormal embryo ; I do not differ from Prof. His as to the slight value attaching to Beigel's description. Bruch's embryo (Abh.Senck.Ges.VI.,Taf. X. [40] ) appears to me from his description and plate to have been very- abnormal. Of these seven embryos Kolk's and Beigel's do not belong to this stage; Von Baer's and Bruch's were abnormal; Hecker's is questionable, Ecker's somewhat uncertain, and Thomson's the only satisfactory one. Of Thomson's only the general appearance is described, but that confirms what we learn from His' observations on this stage. 37, 38. Of embryos with four gill-clef ts we possess no satisfactory descriptions, unless, indeed, we regard His' embryo Lr, described above, as belonging to this stage, since the fourth pharyngeal gill- pouch is found in it. To this stage may perhaps be assigned the embryo described by Johannes Miiller (" Physiologie, " 4te Aufl.,Bd. II., 713, Taf., and in Miiller's Archiv, 1834, p. 8, and 1836, p. clxvii), and also Wagner's embryo (Wagner's "Icones Physiol.," Taf. VIII., Figs. 2 and 3, also in Ecker's "Icones Physiol.," Taf. XXV., Fig. 5) ; important critical remarks on these two embryos may be found in His' "Anat. menschlicher Embryonen," Heft 1, 162-163, and Heft II., 41-43) . Miiller's embryo was about 5.5 mm. alone, Wagner's 4.5 mm. They both had a marked dorsal flexure resembling that normally present in embryos with two gill-clefts ; but this flexure was probably produced artificially by a strain upon the yolk-sac pulling the back down ; the flexure is to be regarded as artificial, because in embryos which were certainly normal it was not found in the next younger or the next older stage. How easily the flexure may be produced is shown by His' observation of its occurring in his embryo W, while he was manipulating it. Neither of the two embryos under consideration are described or figured with sufficient accuracy of detail to justify a fuller description of them. As Von Baer states of his embryo, 34, that it had four clefts, it must be held to belong in this stage probably. Summary. Known Young: Human Ova. — The detailed de- scriptions of the preceding pages, 287 to 308, are summarized in the following paragraphs. First Stage:* Primitive Streak. — No human ovum has been observed to have a primitive streak, but there are several which are younger than the embryos with open medullary groove, and there- fore presumably are in this stage ; unfortunately there is a satisfac- tory description of the embryonic structures in no one of them. To this group have been assigned the embryos 1 to 9, but of these Beigel's (5) is certainly abnormal, and Schwabe's (9) is probably * For deflaitioDS of the stages, see p. 286, ante. KNOWN HUMAN OVA. 309 both abnormal and much older. From the preceding review of the remaining seven ova the following conclusions may be drawn : The human ovum by the twelfth or thirteenth day is a rounded, somewhat flattened sac of three to fovir millimetres in diameter, bearing an equatorial zone of short unbranched villi; the villi are probably formed by the ectoderm only ; the wall of the sac is ectoderm, whether underlaid by somatic mesoderm or not is uncertain; to the inner wall of the sac over one of the bare poles of tiie ovum is attached a mass of cells, constituting the anlage of the embryo ; as to the ar- rangement of these cells we possess no knowledge. In the next stage the villi have spread over the germinal area and have become slightly branched ; the villi next appear over the oppo- site pole of the ovum and rapidly increase their length and ramifi- cations. The germinal area faces the uterine wall (Jones' ovum, 3). By the time villi are present over the whole vesicle there is probably always a layer of connective tissue underlying the epithelium (Breus 2, Ahfeld 4, Lowe 5, etc.), but no embryonic structures have been recognized. The ova of twelve to fourteen days are already completely inclosed by the decidua (reflexa and serotina) ; only the tips of the villi adhere to, or are even in contact with, the decidua ; this is the only connection between the maternal and fcetal tissue, for neither does the uterine mucosa grow in between the villi, nor do the villi penetrate the cavities of the uterine glands. The epithelium of the chorion and villi is only imperfectly marked with boundaries for the single cells ; its nuclei all occupy a basal position, leaving a distinct outer layer, often mistaken for a separate structure. The epithelium forms buds which become branches of the villi. These buds may grow out to a considerable size without connective tissue (hollow villi), or the connective tissue may penetrate into them from the start (solid villi). The human ovum, then, is remarkable for its precocious development of the chorion, both as regards the villi and the connective tissue or mesodermic layer, and for its early complete encapsulation by the decidua. All these events (according to the scanty observations yet made) precede the appearance of the embryo. It is also noteworthy that the villi are first developed around the equator, next over the germinal area pole, and last over the area of the opposite pole. Second Stage: Medullary Plate. — To this stage I assign the embryos. His' XLIV. or Bff (10), Keibel's, 11, and Spec's, 12, and I think they belong in the order named. The chorionic vesicle is rounded and somewhat fiattened ; in its greatest diameter it meas- ures 8-10 mm. ; it is beset with short branching villi, which are present over the entire surface except in one case, where they formed an equatorial band as in Eeichert's ovum 1. The chorion had a distinct ectodermal and a distinct mesodermal layer ; the former, at least in Spee's embryo, had two strata of cells, as is characteristic of the chorion. To the inner surface of the chorion was attached a thick allantois-stalk (BauclistieT) , which, curving slightly, passed over without any demarcation into the embryo, which in Keibel's ovum measured about 1 mm, in Spee's about 1.5 mm. From the sides of the allantois-stalk and of the embryo sprang the thin am- nion, which- was completely closed. Along nearly the entire length 310 THE EMBRYO. of the ventral surface of the embryo was attached the yolk-sac, which was of rounded form and about equal in diameter to the length of the embryo; in Keibel's ovum the yolk-sac had blood-vessels con- taining nucleated blood-corpuscles, and was a hollow vesicle whose thin walls were composed of a fine lining of entoderm, and a thicker sheet of mesoderm. Spee was able to study his embryo in detail; it had a well-marked medullary plate with a median furrow. Fig. 164 ; at the posterior end of the plate was the primitive streak, and at the anterior end of the primitive streak was an opening (named by Spee the neurenteric canal) leading into the entodermal cavity ; the head had grown forward sufficiently to indicate the development of the vorderdarm ; the notochord was present, as a median band of entodermal cells, running forward from the neurenteric canal; the allantoic diverticulum extended as a narrow tube of entoderm through the allantois-stalk to the chorion ; the coelom had not ap- peared in the embryo proper ; the anlage of the heart was not present. This stage is, therefore, characterized by the size of the chorionic vesicle, 8-10 mm., the completed development of the extra-embryonic coelom, and the absence of the embryonic coelom and heart anlage ; by the presence of the medullary plate, neurenteric (or blastoporic) canal, notochordal band in the entoderm, the vascularized yolk-sac, the thick allantois-stalk with the tubular allantoic diverticulum. The general arrangement can be understood from the diagram, Fig. 166. Third Stage : Medullary Groove. — The development of both the embryo and ixs appendages has advanced. Particularly note- worthy are the large size of the medullary ridges and the precocious differentiation of the chorion and amnion. The youngest embryos of this group are in the neighborhood of 2.2 mm. in length (Thom- son gives the length of his embryo I. as 3.5 mm., but the criticisms made above render it plain that this measure probably refers to the length of the amnion plus the allantois-stalk) ; the embryo not seen by Thomson was presumably shorter. The embryo has a broad attachment to the yolk-sac, which in diameter nearly equals the length of the embryo and is already furnished with blood-vessels. The most conspicuous character of the embryo is the presence of two very thick dorsal ridges — medullary folds, running the whole length of the embryo and inclosing the medullary groove, central nervous system to be, between them; the cephalic extremity is somewhat thickened ; from the ventral side of the caudal extremity springs the short and thick allantois-stalk, the opposite end of which is inserted into the chorion. The amnion completely incloses the embryo, and is attached on the one hand to the allantois-stalk, on the other to the embryo nearly parallel to the junction of the embryo and the yolk- sac. The next change involves not merely the growth of the embryo, but also the thickening of its cephalic end, the development of the great heart protuberance between the yolk-sac and the head, the concave flexion of the back, and the deepening of the medullary groove, which, however, still remains open. The chorion forms a relatively large vesicle, its average diameter being about 8 mm., but the four specimens vary from 5.7 to 15 mm. The chorion bears villi over its whole surface; the villi are considerably branched. KNOWN HUMAN OVA. 311 Probably the villi are formed chiefly if not solely by epithelium, and probably, also, there is a layer of connective tissue, very likely al- ready vascular, which lines the chorion, but does not extend into the villi. There are many still unsolved problems as to the develop- ment of man. It will be observed that not a single one of the ova hitherto noticed has been adequately investigated, and that no speci- mens have yet been studied at all, showing the first appearance of the embryo, the origin of the amnion or of the allantois, or of the yolk-sac ; and finally, that of all the earliest stages our knowledge is extremely imperfect. It is, therefore, much to be hoped that all who obtain available specimens will carefully preserve them and intrust them to a competent investigator. From the above considerations it is also evident that the summary just given can be only tentative. Fourth Stage: The Heart. — In this stage the embryo is prob- ably 2.2 to 2.5 mm. long; the head projects in front of the yolk, and on the under side of the cervical region the heart has appeared ; the deep neural groove is partly closed to form the medullary canal, but is open along the cephalic region ; the dorsal outline is slightly con- cave ; the myotomes have appeared, the number varying ; Spee found seven, KoUmann thirteen ; the caudal end of the embryo also projects beyond the yolk, but less than does the head ; the auditory invag- ination is probably not yet formed ; there are no gill-clefts showing externally. Concerning the chorionic vesicle at this stage, satisfactory data are lacking. Fifth Stage: One Gill-cleft. — No human embryo with only one gill-cleft showing externally is known. Sixth Stage: Two Gill-clefts and Dorsal Flexure.— To this stage we must assign not only my two specimens referred to above, 20 and 21, and His' Lg, 23, and Sch 1, 23, but also His' L, 24, and probably Coste's, although in neither of the latter does the dorsal flexure appear. It is possible that Schroeder van der Kolk's ovum, 26, Hennig's, 27, Schwabe's, 9, and Remy's, 28, also belong in this stage, but for reasons given above in detail the position of these four is very doubtful, that of Schwabe's especially so. In His' embryo L, and in Coste's, the dorsal flexure was probably obliterated artificially, leaving only the four embryos, 20-23, upon which the following synopsis is based, with the addition of some anatomical facts derived from Nos. 24 and 25. The general shape of the embryo and its remarkable dorsal flexure can be best understood from Fig. 17. The head bend is very marked and the tail end of the embryo is also bent over ventralward; the yolk-sac extends from the heart backward to where the body of the embryo turns to make the dorsal flexure; the heart is large and very protuberant ; it is bent so that we can clearly distinguish the auricular, ventricular, and aortic limbs, and it consists of a smaller inner tube, the endothelial heart, or endocardium (which is con- tinuous at one end with the walls of the veins, at the other with the walls of the aorta), and of an outer larger tube, the muscular heart or myocardium ; between the two heart tubes is a considerable space ; there are two gill-clefts and, at least in the youngest specimens, only two aortic arches, one in front of each cleft ; between the head 313 THE EMBRYO. and the heart the oral invagination has been formed but is still sepa- rated by the oral plate (Rachenhaut) from the vorderdarm ; above the gill-clefts is the open ectodermal invagination of the otocyst, which in His' embryo L, 34, had become a closed vesicle. The cen- tral nervous system is very large compared with the whole embryo ; the brain comprises in length about one-half of the medullary canal ; the optic vesicles are large, and the optic stalks are well differenti- ated; the head bend takes place in the region of the mid-brain, which is imperfectly separated from the fore-brain ; the hind-brain is about equal to the fore and mid brains together in length; there were twenty-nine myotomes in His' embryo Lg, 22. The vorderdarm is flattened dorso-ventrally ; the liver is developing in the septum trans- versum ; the middle portion of the intestine opens into the yolk-sac, the posterior portion is closed and at its caudal termination is dilated to form the bursa of His, and curves over to pass as the narrow tubular allantoic diverticulum through the allantois-stalk to the level of the chorion. The veins show the typical arrangement, the jugu- lars joined by the cardinals form the ducti Cuvieri, and these after receiving the omphalo-mesaraic (or vitelline) and the umbilical (or allantoic) veins unite in the median line as the sinus reuniens ; the course of the allantoic veins is peculiar and may be described as a short cut through the somatopleure along the line where the body wall of the embryo is deflected back to form the amnion. Seventh Stage: Three Gill-clefts. — All the accurately known embryos, except one, 28A, belonging to this stage, belong to the end of it, and one of them, His' Lr, 33, is so far advanced that it might almost be classed in the next stage. Five good em- bryos, 29-33, are to be placed here, and foiir others, 34, 35, 36, and 5 have been associated with them, but the latter are all doubtful cases; the best of them being Von Baer's, 34, which probably should be put in the eighth stage. For reasons stated in the section on the dorsal flexure, p. 313, the flexure is probably normally absent in embryos at the close of the seventh stage. The described embryos vary from 2.6 to 4.3 mm. inlength; His' M, 30, was 2.6 mm. long, and its chorionic vesicle measured 11 by 14 mm. His' BB, 31, was 3.3 mm. long, and its chorionic vesicle measured 11 by 14 mm : the age of BB was probably twenty to twenty-one days. The back of the embryo is normally (or at least usually) convex ; the head is bent to one side (usually to the right) and the tail to the other, the whole embryo having a spiral twist; there are three gill-clefts showing externally ; the tail end has grown considerably and the allantois- stalk has lengthened ; the yolk-stalk (neck of the yolk-sac) is both relatively and absolutely smaller than in the previous stage, but the embryo is larger. The heart has grown very much ; in the older specimens the development of the auricular pouches has begun. The otocyst is a closed pear-shaped vesicle, its apex pointing toward the dorsal side. The mouth cavity has deepened, the oral plate is ruptured; above the mouth the maxillary process can be distin- guished. The pharynx is wide, compressed dorso-ventrally, and has in the known specimens four gill-pouches, and on its median ventral floor a small prominence, His' tuberculum impar, the anlage of the tongue ; the diverticulum of the liver is well marked in the youngest, KNOWN HUMAN OVA. 313 and enlarged and branching in the oldest specimens ; the Wolffian ridge is distinguishable and contains Wolffian tubules, but as to the number and form of these we possess no exact information. The medullary canal is closed throughout its length ; the mid and fore brains have become clearly separated since the sixth stage. As regards the circulatory system, besides the appearance of the auricles and the general advance of the heart, we have to note that the great veins passing through the septum transversum have begun their transformations into the hepatic system, and that the aorta has five aortic arches, the two first coming from one branch, the remaining three from another branch on each side ; no embryos are known with only four aortic arches. Eighth Stage: Four Gill-clefts. — The three embryos, 34, 37, 38, which were apparently in this stage, are so imperfectly known that there is practically nothing definite to say in regard to their anatomy. Wagner's specimen, 38, measured 4.") mm.; Miiller's, 37, 5.5 mm. Th.e Dorsal Flexure. — In a number of embryos with from two to four gill-clefts there has been observed a deep bend in the rump, which suggests at once the effect of a pull upon the yolk having pro- duced a sharp conca-vity in the back, compare Fig. 169. In embryos with two gill-clefts this bend, for which I propose the term dorsal flexure (Riickeiikrihumung), has been shown by His to be normal. In older embryos it seems to be abnormal, for in one with three clefts and the dorsal flexure, 31, the tissues in the region of the bend were lacerated, and in a still older specimen (W of His) the bend was artificially produced while the embryo was being manipulated. The facts indicate that the back is too long for the somatopleure at the side of the body, and that it finds room at the stage with two gill- clefts by becoming concave; later it springs into a new position of equilibrium by becoming convex ; it is possible that the change from the concave to the convex position is very abrupt, and it is probable that the time of its occurrence is very variable, so that we may find hereafter embryos in the seventh and eighth stagey, which are per- fectly normal though still having the dorsal flexure. PAKT IV. THE F(ETAL APPENDAGES. CHAPTER XIV. THE HUMAN CHORION. The human chorion has been the object of greater misconception than perhaps any other organ of the body. Even at the present time there prevail numerous false notions concerning it, and many of these errors are repeated in some of the best accredited text-books. The literature of the subject includes a majority of papers of little value, and often remarkable for the gross crudity of the observations they record and for the ignorance displayed by their authors of other and better observations. The literature also includes numerous pa- pers by investigators of exceptional accuracy and intelligence, such as Coste, Farre, Turner, Langhans, Waldeyer, etc., by which we are enabled to give a fairly complete history of the chorion. This chapter is based chiefly on the account given in my paper on " The Uterus and Embryo," 98. General Description. — The chorion has already been defined as the whole of that portion of the extra-embryonic somatopleure, which is not concerned in the formation of the amnion. The human chorion, as stated above, p. 281, is remarkable for its very early complete separation from the yolk-sac and for the precocious appear- ance of its viUi. As shown in the previous chapter, both of these developments have taken place in all the very young human ova hitherto obtained, even in Reichert's ovum, which is suppposed to be the youngest known, see p. 287. At about twelve days the cho- rionic vesicle is a closed sac, somewhat flattened and 3-4 mm. in greatest diameter ; around its equator there is a broad zone of simple villi. About a day later the vesicle is 1 mm. larger, and the villi, which are beginning to branch, have developed over one of the polar areas, and soon after develop over the second pole also, though less thickly than elsewhere. The fourteenth day the vesicle measures about 5 mm. In all these cases the vesicle is found completely encapsuled in the decidua, and the decidua reflexa is closed over it. The tips of the villi alone touch the decidual surface, to which they are lightly at- tached, for they can be pulled off from it without their breaking, Spee, 89. 1, 160. This arrangement leaves a space which is bounded on the one side by the maternal decidua, on the other by the foetal chorion, and which is crossed by the foetal villi; it is commonly designated as the intervillous space. As to the growth of the chorionic vesicle I have failed to find any extended observations, and can only express the hope that the omis- sion may be soon supplied. The growth at first is very rapid, so that during the second month there is always a considerable space around the embryo and amnion, but after the second month the space is 318 THE FCBTAL APPENDAGES. relatively diminished by the growth of the fcBtus. W. His gives the following table (" Anat. Menschl. Embryonen," Heft II., 31). Diameter of Chorion, < 1.5 cm. ; Embryo, 2-4 mm. 1.5-3.0 " " 4-10 " 2.5-4.0 " " 10-15 " 3.5-5.0 " " 15-20 " 4.0-6.0 " " 20-25 " The contents of the vesicle are, first, the embryo with its allantois- stalk and yolk-sac, and, second, the chorionic fluid; concerning the latter I know of no exact observations, but it probably resembles, if indeed it be not identical with, the amniotic fluid, compare p. 337. The history of the villi is given below in detail ; in this paragraph we need refer only to the changes in the villi, by which the mem- brane is differentiated into the chorion Iceve and the chorion fron- dosum. I consider it doubtful whether the number of viUi increases at all after a comparatively early stage, but over all that part of the chorion which overlies the decidua serotina (c/. Chapter I.) the villi continue to grow both in size and in the number of their branches for a long time — perhaps through the entire period of pregnancy ; this area of enlarged villi presents a shaggy appearance and hence is called the chorion frondosum ; it participates in the formation of the placenta ; the allantois-stalk (or later the umbilical cord) is always inserted into the chorion frondosum. Over all the remainder of the chorion, which lies against the decidua reflexa, the villi gradually atrophy during the second month, so that this region becomes smooth, and hence is termed the chorion laeve. The chorion consists histologically of an external layer of epithelial ectoderm and an inner thicker layer of mesoderm ; whether the mes- oderm is divisible into a mesenchyma and mesothelium, as the devel- opment of the chorion out of the somatopleure leads us to expect, is uncertain, but if there is an interior layer of epithelium on the meso- dermal surface it must be extremely thin, for I cannot detect it in my sections ; the bulk of the mesoderm is undoubtedly mesenchymal. The ectoderm in the earliest stages known consists of two clearly differentiated layers, a thinner outer one with small nuclei and with- out recognizable cell boundaries, and an inner one consisting of dis- tinct cells with large nuclei. The outer layer has been regarded by some authors as maternal tissue — an opinion discussed in the section on the histology, p. 322. The chorion is at first vascular throughout its entire extent, re- ceiving its blood from the embryo via the allantois-stalk through two arteries, and returning it by the same route through two veins, see Chapter XVII. The vessels early penetrate the villi, but as the villi disappear from the chorion laeve the blood-vessels also abort there and remain only over the chorion frondosum (compare Chapter XVII.), to maintain the circulation of the foetal placenta. Chorionic Villi. Development. — As has been stated both in the review of the youngest known human ova, and in the general description of the chorion, the villi arise in a broad zone around the equator of the somewhat flattened chorionic vesicle, and soon after appear over both polar areas; they are at first clumsy cylinders which may grow to a millimetre in length before they begin branch- THE HUMAN CHORION. 319 ing. They arise, as shown long ago by the observations of Costs, as outgrowths of the ectoderm only. Fig. 174; the hollowness of the villi and their clumsy shape are to be especially noted. The meso- derm grows into them subsequently. The openings into the villi can also be seen in Fig. 173, scattered over the surface of the chorion. Branching.— The branches of the villi grow out in a similar manner, the process being led, as it were, by the ectoderm. Orth in a special paper, 78. 1, has used these facts to argue against Boll's "Princip des Wachsthums." KoUmann's observations, 79.1, 397, on the growth of villi during the fourth week are particularly in- structive. The outgrowth of the branches is very rapid and occurs with every degree of participation of the connective tissue. The two extremes are: 1, a bud consisting wholly of epithelium, which may become a process with a long thin pedicle, and a thickened free end remaining entirely without mesoderm ; 2, a thick bud with a well- developed core of connective tissue, and having a nearly cylindrical form. Between these extremes everj^ intermediate state can be found. Other observers have noted this peculiar manner of growth, which I have found still going on in the placental chorion during the fourth month. Robin, 54. 1 , appears also to have crudely observed both the young hollow villi and the solid epithelial buds. The blood-vessels he traces to the division of the cavity of the villi into an artery and a vein ; from the nature of things he offers no observa- tions in support of this assertion. Only the tips of the villi touch the surface of the decidua either at first or subsequently, except, of course, over the chorion Iseve during the abortion of the villi. The tips of the villi are attached to the uterine surface ; they penetrate the decidua for a short distance, but even in the placental area at the close of gestation the penetration is slight and the villi make their way only into the surface stratum of the decidua serotina. There is no evidence of any sort that the villi penetrate the glands at any period. The relation of the villi to the decidua has now been so accurately ascertained that there can be, I think, no longer any question whatsoever on this point. The best discussion is by Langhans, 77.1, p. 331 J^. The shape of the villi varies according to the part of the chorion and the age of the embryo. They gradually abort over the chorion laeve, and gradually grow over the chorion frondosum. Let us begin with the placental villi : At first they are short, thick-set bodies of irregular shape, as shown in Fig. 174; at twelve weeks their form is extremely characteristic, Fig. 181 ; the main stem gives off nu- merous branches at more or less acute angles, and these again other branches, imtil at last the terminal twigs are reached ; the whole of the space between the chorion and decidua is occupied by these ram- ifications; the branches and twigs, as the illustration shows, are extremely irregular and variable, although in general they may be described as club-shaped, being more or less constricted at their bases. The branches may be bigger than the trunk which bears them, or of any less size ; some of the smallest are merely slender outgrowths of the epithelial covering of the villus, such as have already been alluded to. Gradually there is a change. During the fifth month we find the irregularity, though still very marked, decidedly less 320 THE PCETAL. APPENDAGES. exaggerated, Fig. 183; the branches tend to go off at more nearly right angles ; one finds very numerous free ends, as of course only a small proportion of the branches touch the decidual surface; the Fig. 181.— Isolated Terminal Branch o( aVillus Fig. 183.— Villous Stem from a Placenta of from the Chorion of an Embryo of Twelve the Fifth Month. X 9 diams. Weeks. branches, too, are less out of proportion to the stems, less constricted at their bases, or, in other words, less remote from the cylindrical form; the awkward cucumber- shapes of the twelfth week are no longer found except here and there. The change continues in the same direction ; that is, is toward greater regularity of configura- tion. It is hardly necessary to describe the intermediate phases that have been exam- ined, but it will suf- fice to describe the form at full term. Fig. 183, when the branches are long, slender, and less closely set, as well as less subdivided, than at earlier stages ; they have nodular projec- tions like branches arrested at the begin- ning of their develop- ment; there are nu- merous spots upon the surfaces of the villi; microscopical examination shows that these spots are proliferation islands, as we may call them, or little thick- enings of the ectoderm with crowded nuclei. It appears that not all ^^^^ XE) The Fig. 183.— Terminal Villi of a Placenta at Full Term, little spots represent the proliferation islands of the covering epithelium. THE HUMAN CHORION. > 321 the villi change to the slender form ; for some villi, having still the earlier, thicker form, are found even in the mature placenta, a fact already noticed by Jassinsky, 67.1, 346. These thick villi usually show also a distinct " cellular layer" in their ectoderm, a peculiarity to be considered below again. Seller, 32. 1, has given figures of the villi at various ages, but fails to show the characteristic forms. Langhans has observed X / efe^. ''.tCi :x '^ M I ^ ^J^TIOJ"-/!."**'^'' EP „f" * VI i^'J < ^^ \3 Vi Fig. 184. —Section of the Chorion at Three Weeks, a, layer of coagulum; &, mesoderm of chorion; Ep^ epithelium, also extending over the villi, Vi and Vi' ; the mesoderm, ft, con- tains a number of blood-vessels, nearly all in transverse sec- tion. X 65 diams. the alteration in the villi, 77.1, 199, and even justly remarks that many of the villi in so- called " moulds" are not pathological, as they have been frequently considered, but normal young villi. The differ- ences in the villi accord- ing to age are very con- spicuous in sections. The sections should, of course, be made so that the fragments of the villi will remain in situ; imbedding in cel- lodine is convenient for this purpose ; if this end be attained, one finds below the chorionic membrane numerous sec- tions of villi; if the specimen be a young chorion — -first to third month — the villi are large, with a good deal of room between them ; their outlines are very irregular and there are relatively few small branches. Fig. 184. The older the specimen, the larger the propor- tion of small branches. In an old chorion — seventh to ninth month — the number of small villi of nearly uniform size is very striking (see the figure of a section through a placenta in situ, given in Fig. 313). The abolition of the villi of the chorion laeve takes place by an arrest of development and a subsequent slow degeneration of the tissues, which lose all recognizable organization in the protoplasm, and to a large extent of the nuclei ; at the same time they alter their shape. Fig. 185, becoming more and more filamentous ; by the fourth month only a few tapering threads, with very few branches, re- main The villi disappear almost completely from the laeve, except near the edge of the placenta, where they are to be found, even in the after-birth, imbedded in the degenerated epithelium of the chorion and the upper layers of the decidua, as shown by Minot, 98, the epithelium and decidua being so fused at this point that it is impos- sible to determine any line of demarcation between them. 31 Fia. 185. — Aborting Villus from a Chorion of the Second Month. 332 THE FCBTAL APPENDAGES. Histology of the Chorion. — The chorion being a portion of the somatopleure consists, of course, of two primary layers, the meso- derm and ectoderm. During the second half of the first month, the earliest period concerning which we have any accurate knowledge, the mesoderm is already a vascular layer of considerable thickness (Figs. 184 and 188, 7nes), and the epithelium (ectoderm) has two layers of cells Fig. 188, a and b; of which the outer is the darker in speci- mens stained with osmic acid, carmine, cochineal, or haematoxylin, and has also smaller and more granular nuclei. The same distinc- tion exists in the two-layered stage of the ectoderm of the umbilical cord. Fig. 208, and of the foetal skin. Hitherto most authors have entirely overlooked the inner layer at early stages. It was first clearly recognized by Langhans, who directed attention to it in a special memoir, '82. 1, he having already described its later history, 77.1. In some earlier writers are allusions to the layer. Kast- schenko, in his paper on the chorionic epithelium, has also described it, although he has not followed its history very far. The interpre- tation to be offered seems to me clearly to be that the chorionic epi- thelium advances in its difi^erentiation to a stage equivalent to the two-layered stage of the epidermis and there stops;' whatever further change occurs is degenerative. The two prirnitive laj'ers of the chorionic epithelium have been more or less clearly observed at later stages by several anatomists, and have been variously interpreted. Ercolani and Turner regard them as absolutely distinct, assigning the deep layer to the chorion as its true and only epithelium, and the outer layer to the uterus, thus enabling themselves to conceive the villi as covered by maternal as well as a fcetal epithelium, so that maternal blood found between the villi is still within the maternal tissue. After accepting the outer layer as maternal, the question as to its origin still remained. Some authors affirmed it to be the uterine epithelium, others to be the lining of expanded uterine blood sinuses. So far as I am aware, no one has made observations to show by the developmental history of the layer that one or the other of the last-mentioned hypotheses :' )rrect. When we consider the precision and exactitude of Kast- I nko's observations, which actual specimens enable one to verify, e is in my judgment no reason left for differing from the con- clusion that both layers are parts of the foetal ectoderm. Governed by the difficulty of accounting for the presence of ma- ternal blood in the intravillous spaces, and therefore apparently outside the maternal tissues, several investigators have been led to seek for at least an endothelium outside the chorionic epithelium. Some authors, as, for instance, Winkler, haye asserted the existence of such an endothelium, but after a prolonged and careful search I fail to find anything of the kind, and in this result it seems to me the best observers are agreed. Waldeyer, 90.1, 33, has recently again advocated the existence of an endothelium, but from his description it appears to me that his supposed endothelium is only the outer layer of the ectoderm. Keibel, 89.2, reports a very different observation; in a young ovum (twenty-five days?) he found a very thin endo- thelial layer outside of the two layers of ectoderm, and enclosing the maternal blood. Now in the rabbit the placental villi grow down THE HUMAN CHORION. 333 into ths uterine mucosa ; the intervillous maternal tissue disappears, leaving only the maternal capillaries, w^hich become enormously hypertrophied and take up the entire intervillous room ; in conse- quence the capillary endothelium immediately covers the villi ; later this endothelium also aborts, leaving the blood of the uterus to circu- late in channels bounded by the chorionic epithelium. If we assume that the process of development is similar in man, but is completed very precociously, 'we can understand both Keibel's observation and the failure to detect any true endothelium in later stages. For a full review of the many conflicting opinions concerning the covering of the villi, see W. Waldeyer, 90.1, 33-47. Differentiation of the Ectoderm. — The epithelium of the chorion becomes differentiated in three different ways : 1, upon the chorion frondosum; 2, upon the chorion laeve; 3, upon the villi. For a correct knowledge of the remarkable changes which the epi- thelium undergoes, particularly in the placenta, we are indebted to the unusually exact investigations of Langhans, 77.1, and 82.1. This author left two points of importance unsettled; namely, the origin of his '' Zellschicht," and of the " canalisirtes Fibrin." Kastschenko has traced the cellular layer {Zellschicht) to the epithe- lium, as already stated; compare pp. 403-469 of his memoir, 85.1. My own observations show, I think, conclusively that the canalized fibrin arises through a degenerative metamorphosis of the epithe- lium, which begins in the outer layer and may invade the inner layer (Langhans' Zellschicht) . Let us consider separately the three series of modifications of the chorionic ectoderm. In the region of the chorion frondosum the inner layer of the ecto- derm (the cellular laj^er of Langhans) becomes irregularly thickened in patches, which present every possible degree of variation as to number and as to their breadth and thickness. Although at first the cellular layer is more or less continuous and composed of uni- form cells, this is not the case in later stages. We must assume that with the growth of the membrane the epithelium increases in area, but remains in many places single-layered, developing no Zell- schicht. The patches of cells have been well described by Lang- hans, 77.1, and Kastschenko, 85.1, 466, and are represented with lower power in Fig. 189, c, and with a higher power in Fig. 186, c. They vary much in appearance ; the cells are more distinct in the small patches, but are less individual in the large patches, owing to the spread of the process of degeneration into the layer. Fig. 186, c. The cell bodies are lightly stained, and the granular nuclei are not very sharply defined and vary in size and shape. The cellular layer is always sharply defined against the stroma, although there is no true basement membrane, but toward the outer layer of the ectoderm its boundary is sometimes distinct, sometimes lost in a gradual transition. The outer layer of the ectoderm of the frondosum is even more variable. As stated by Kastschenko, it is primitively a dense pro- toplasmic reticulum, with nuclei in a single layer and without any cell boundaries. In the chorion frondosum at four months and after I find spots where this structure still prevails either with or without an underlying cellular layer ; in other spots the layer is thickened and 324 THE FOSTAL APPENDAGES. contains an increased number of nuclei, which are sometimes crowded in a bunch ; elsewhere the layer is thinned out and has no nuclei ; in still other spots the thickening has gone on much further, and usually but not always, where the outer layer is much thickened the cellular layer under it is also thickened ; wherever it is thickened, and occasionally where it is thin, the outer layer of the ectoderm shows a marked tendency to degenerate into canalized fibrin. Fig. 189, Fbr, and Fig. 186, fb. It is not difficult to assure one's-self that the fibrin arises by direct metamorphosis of the ectoderm. I now think that its formation begins in the outer layer and thence ^^■' Fig. • 186. —Placental Chorion of an Embryo of Seven Months ; Vertical Section through the Ecto- derm and Portion of the Adjacent Stroma, mes, Mesodermic stroma; c, cell layer; fb fibrin layer ; ep, remnant of epithelium. X 445 diams. spreads into the cellular layer; for, in fact, when both layers are distinguishable, as in Fig. 186, the fibrin layer, fb, is always external, and the external layer of nucleated protoplasm has either totally dis- appeared or is represented by mere remnants as in Fig. 186, ep. The fibrin layer consists of a hyaline very refringent substance permeated by numerous channels. Fig. 186, fb; the substance has a violent affinity for carmine and hsematoxylin, and is always the most deeply colored part of a stained section ; the channels tend to run more or less parallel to the surface of the chorion, and are connected by THE HUMAN CHORION. 325 iiixmerous cross-channels; some of the channels contain cells or nuclei. This complex system of canals is by no means of uniform appearance in all parts of the placenta, both the spaces and dissepi- ments varying in size and shape. The fibrin often sends, as shown in Fig. 186, long outshoots into the cellular layers upon which it seems to encroach. The frequency of these images in my prepara- tions led me to the opinion* that the fibrin arises from the cellular layer only, and I concluded that the ectoderm was first transformed into the so-called cellular layer, which was then transformed into fibrin. It still appears to me that much of the degeneration goes by these stages ; but, on the other hand, it seems clear that the de- generation begins, as above stated, in the outer layer. Another appearance is presented by the ectoderm where it is thickened and wholly transformed into the cellular layer. In brief: the ectodei-m of the placental chorionic mesoderm undergoes patchwise manifold changes ; it exists in three chief forms : 1 , the nucleated protoplasm ; 2, the cellular layer; 3, canalized fibrin. A patch of the ecto- derm may consist of any one of these modifications, of any two, or of all three, but they have fixed relative positions, for when the nucle- ated protoplasm is present it always covers the free surface of the chorion ; when the cellular layer is present it always lies next the mesoderm ; and when all three forms are present over the same part, the fibrin is always the middle stratum. In general terms it may 1)6 said that the amount of canalized fibrin increases with the age of the placenta, but it is very variable in its degree of development. The peculiar layer into which the ectoderm is transformed, has long puzzled anatomists. E. H. Weber recognized the fibrin layer and described its appearance correctly ; it has probably been often seen, but generally regarded as either pathological or a blood coagulum. Robin, for instance, may be cited, 54. 1, 70-71, as one who saw with- out observing correctly and understandingly the tissue in question. An important gain was made when Winkler recognized the modi- fied ectoderm as a constant layer, and in 1872 directed especial attention to it under the name of " Schlussplatte," 72. 1. K5lliker ("Entwickelungeschichte," 2te Aufl. 337) added essentially to our knowledge of its structure, but it is to Langhans that we owe the first clear light. Meanwhile other writers, following the lead of Ercolani and Turner, 79.1, 551-553, have been influenced chiefly by the presence of the cellular layer, in the large size of the elements of which they found a resemblance to the decidual cells which has guided them to the conclusion that the cellular layer is derived from the wall of the uterus. This error has been definitely corrected by Kastschenko, as already stated. In further support of the conclu- sion that the chorionic cellular layer is not decidual, may be brought forward the fact that there is a certain immigration of decidual cells into the placenta at its margin ; but they remain entirely distinct from the cells of the cellular layer. This is readily seen in radial sections through the margin of a placenta from a normal afterbirth ; compare, below, the account of the ectoderm of the chorion laeve. The origin of the canalized fibrin from blood, which Langhans left in his first paper as an open possibility, and which even so recent a * Anatom. Anzeiger, ii. , 28. 326 THE FCETAL APPENDAGES. writer as Ruge, 86.1, 123 and 130, has advocated, cannot be main- tained. Of course there may be a deposit of blood fibrin (coagu- lum), but it would be pathological, and, therefore, to be distinguished from the normal fibrin of ectodermal origin. Moreover, the micro- scopic appearance of a blood clot or thrombus is so extremely char- acteristic that one can readily distinguish it from the placental canalized fibrin. The ectoderm of the villi of the placenta differs from that of the chorionic membrane in several respects : 1. The cellular layer after the first month becomes less and less conspicuous, and after the fourth month is present only in a few isolated patches, known as the Zell- Icnoten, and carefully described by Langhans and Kastschenko ; both of these authors were impressed by the resemblance of the cells to those of the decidua serotina; Langhans concludes that the Zell- knoten arise from the serotina, but Kastschenko, having traced their development from the chorionic epithelium, denies his predecessor's conclusion, but, still clinging to the idea of a genetic connection be- tween the Zellknoten and the decidua, reverses the reasoning and concludes that the decidual cells arise in part, at least, from the Knoten. Neither of these authors have found the intermediate forms between the tw;o types of cells, and when we examine their descrip- tions, critically we find that they have really no evidence except the likeness of the cells to offer in favor of their genetic relationship, and accordingly Langhans expresses himself with characteristic caution. To me the resemblance appears altogether superficial; hence my conclusion that the Zellknoten are remnants of the cellular layer. 2. For the most part the villi remain covered by the nucleated proto- plasm, which in many places is thickened. In the later stages these thickenings are small and numerous, constituting the so-called "Proliferations-inseln;" compare Fig. 183. Many of the little thickenings appear in sections of the villi, and here and there are converted into fibrin. I have interpreted them (Wood's "Refer- ence Handbook of the Medical Sciences," V., 695) as commencing buds, and consider that in earlier stages they grow into branches, but in later stages are in part, at least, arrested in their develop- ment. 3. The proliferation islands are converted into canalized fibrin, and at the same time grow and fuse, forming larger patches, particularly on the larger stems ; in this manner are produced the large areas and columns of fibrin found in the placenta at four months and after ; they have been well described by Langhans, and form a striking feature in sections of placentae. Some of the columns, as stated by Langhans, stretch along the villi from the chorionic membrane to the surface of the serotina as if to act as supports. Ercolani appears, if I understand his account, to have seen the fibrin columns without, however, ascertaining either their structure or their origin. 4. Over the tips of the villi, which are bent consid- erably where they are imbedded in the decidua serotina, the relations are not clear ; the epithelium is certainly not present in its original form over the imbedded ends of the villi, which are, however, sur- rounded by a hyaline tissue of the character of the canalized fibrin, except that the canals are often indistinct or even wanting; the hya- line tissue forms an almost continuous coat over the decidual surface; THE HUMAN CHORION. 337 in earlier stages the ectoderm of the terminal villi is often consider- ably expanded. The natural interpretation of these facts is that the ectoderm of the villi expands over the decidua serotina and degener- ates. In this manner we account for both the absence of any cellu- lar ectoderm over the ends of the villi and the presence of canalized fibrin upon the serotina! surface; but the hypothesis must await the final test by observation. The ectoderm of the chorion Iseve loses by the seventh month all traces of the protoplasmic layer, and is without any canalized fibrin, except near the placenta; cf. infra. It is transformed into a Zell- schicht. In a section of the laeve, in situ, at seven months. Fig. 15, the chorionic epithelium, c, rests directly upon the decidua, which has none of its own. The ectodermal cells lie two or three deep ; they are described by Kolliker and Langhans, the former designat- ing them as the chorionic epithelium, while the latter doubtfully traces their origin to the uterus. That Kolliker (" Entwickelungsge- schichte"), 2te Aufl., p. 322) is right, I am confident. It is easy to foUow the layer of cells in question at the edge of the placenta, and see that it is directly continuous with the cellular layer of the fron- dosum, which it resembles in character. On the other hand the ectodermal cells of the laeve are distinct in character from the decid- ual cells next to them, Fig. 15, having smaller and more darkly stained nuclei, and much more coarsely granular protoplasm; the ectodermal cells are much smaller than the decidual. The ectoderm is sharply marked off from the decidua, but its surface is often cor- rugated, and then the line of separation between the tissiies is irreg- ular, and in sections it may even appear that there is a true inter- penetration and mingling of the decidual and ectodermal cells ; but it is only apparent, and the demarcation is always preserved. Diflferentiation of the Mesoderm. — The further history of the chorionic mesoderm is so fully given by Langhans in his invaluable memoir, 77.1, and Kastschenko, 85.1, that my own observations have afforded little to be added. In the earliest stage I have been able to examine, an ovum of the third week, the matrix of the cho- rionic connective tissue in a preparation stained with cochineal or haematoxylin, and imbedded in paraffin for cutting, appears hya- line and glistening, owing to its refrangibility. Fig. 187; it has lacunte in which the cells lie ; the cell bodies are either shrunken or colorless, so that the lacunae, except for the staining of their contained nuclei, are clear and light. This appearance I find again in speci- mens a little older. The image is entirely distinct from that of the same layer later, for then the cells are stained darker than the matrix, which at the same time has lost its homogeneous character, and ac- quired a fibrillated look. Very different from my own sections are several which I owe to the kindness of Professor Langhans of Bern, and which that distinguished investigator informs me are from a three weeks' ovum, which had been preserved in osmic acid, Fig. 188. In Professor Langhans' preparations the cells are all stained much deeper than the matrix ; they have an elongated form, and run in various directions more or less parallel to the epi- thelium, ect; hence many of them are cut transversely or obliquely. Whether the differences noted are due to the methods of preparation 328 THE FCETAL APPENDAGES. must be decided by preserving the same chorion in part with osmic acid, in part with Miiller's fluid or picrosulphuric acid, the latter being the reagents I have used. In specimens of the tenth week the matrix of the chorionic mesoderm has quite altered in character, being no longer homogeneous, and at the same time it has increased in thickness. For the most part the matrix stains lightly, and where it is lighter it contains fibrils of extreme fineness and running curly courses; there are also streaks of lightly stained matrix, giving the impression of fibres resulting from portions of the primitive colorable matrix being left. In other parts of the layer the primitive matrix is still present, and we find a homogeneous well-colored basal sub- stance, the cell lacunae of which appear light by contrast, as in Fig. 187. One can distinguish also the commencement of the perivascular 1 '^^ ^c -^ 9¥i> \^ 1i ^# t o >" M df'°i>>^^''^ '^^fL^'-^ Fio. 187.— Section of the Chorionic Membrane of an Ovum Supposed to Belong to the Third Week, ect. Ectoderm; ines^ mesoderm; a, outer, 6, inner layer of ectoderm : stained with alum- cochineal. X 445 diams. coats, at least of the larger vessels, the matrix being quite dense around them and the cells elongated almost into fibres, and possessing a slightly increased affinity for coloring-matters. The larger blood- vessels and unmetamorphosed part of the layer occupy a middle por- tion between the two surfaces, but the smaller blood-vessels lie near the ectoderm (compare Fig. 187, v), thus presaging the formation of Langhans' vascular layer {OefdssschicM) . The development of the mesoderm of the chorion laeve stops at about this stage, or at the stage when the matrix has completely changed from its first stage ; in the region of the frondosum, however, development proceeds much further by the production of fibres throughout the whole of the layer; usually, but not invariably, the fibres become much more numerous near the ectoderm than in the inner parts of the mesoderm. THE HUMAN CHORION. 329 thus differentiating a well-marked sub-epithelial fibrillar layer, Fig. 189, fib, from the deeper and wider stroma, Str. The fibrillar layer is that commonly spoken of as the connective tissue layer of the niEg -^-^^'l ^^}^' 'J.^^^^^-'^' Fig. 188. — Section of the Chorionic Membrane of an Embryo of Three Weeks; stained with os- mic acid, mes^ Mesoderm ; ect. ectoderm ; a outer, b, inner layer of ectoderm. From a section prepared by Prof. Theodor Langhans. X 445 diams. chorion; for details of its structure, including the " Gefdssschicht," see Langhans and Kastschenko. The inner layer, Str, is called the Gallertschich t by many G-erman writers, and seems to be what KoUi- -ker ("Entwickelungsgeschichte," 2te Aufl., p. 322) designates as .-■■>..■» a \ '^^.'^" Fig. -Section of the Amnion and Placental Chorion of the Fifth Month. Ep, Amniotic epi- __ ft, amnion; Str^ stroma; Mb^ fibrillar layer; /&)-, fibrin layer; c, cellular layer; Fi, villi. (From a section cut in celloidin, and stained with Weigert's hasmatoxylin. The draw- thelium; -4m, amnion; Str, stroma; Mb^ fibrillar layer; /&)-, fibrin layer; c, cellular layer; Fi, villi. (From a section cut in celloidin, and stained with Weij " ' ' '■ "" ing is only approximately correct as to details.) X 71 diams. " Gallertgewebe zwischen Chorion und Amnion;" it usually con- -tains a considerable number of large granular wandering cells. 330 THE FOETAL APPENDAGES. Jungbluth, 69. 1, describes a network of capillaries which exist dur- ing the first half of pregnancy, apparently in the upper part of the stroma — i. e. , next the amnion — but I fail to find any. Where th& amnion conies into contact with the chorion the adjacent parts of the two membranes are more or less loosened, forming a network of strands by which the membranes are united ; most of the uniting strands appear to belong rather to the chorion than to the amnion. This loose tissue is perhaps that which Kolliker designates as a. Gallertgewebe distinct from the chorion. Although the chorion bounds the ccelom, I have observed no mes- othelium upon its mesodermic surface ; but I have not made search for it by any special methods. In the rabbit, it will be remembered, the mesothelium is very evident over the placenta, but the rabbit differs from man by the absence of union between the amnion and chorion. Nor have I been able to find any basement membrane, prop- erly So-called, under the chorionic ectoderm. As to the appearance which suggests it, I accept Kastschenko's explanation, 85. 1, 455. The mesoderm in the villi is differentiated otherwise than that of the membrane of the chorion. In the youngest stage I have exam- ined there is some of the primitive matrix present in the villi; and I presume that earlier the whole naesoderm has the same character. In my specimen (three weeks) the change is progressing. I have not succeeded in sat- isfying myself as to the process of change which takes place, but I think it prob- ably essentially as follows : The cells gradually develop large bodies and ac- quire a more decid- ed affinity for color- ing-matters ; meanwhile vacuoles appear in the matrix, presumably by its modification into a new substance ; the vacuoles increase in size and number, transforming the matrix into a network and ulti- mately causing its total disappearance, leaving the intercellular spaces filled entirely with the new substance, which has come from a metamorphosis of the original matrix; probably this new su)d- stance is more or less fluid, since wandering cells are scattered freely through it. Leaving this half-hypothetical history, let us pass on to direct observations. In the placental villi of embryos of four months and older, the mesoderm exists in two principal forms — adenoid tissue and fibre-cell tissue around the blood-vessels. The- adenoid tissue, Fig. 190, is that of which the supposed development. Fig. 190.— Adenoid Tissue of a Villus from a Placenta of Four Months. Ill, "Wandering cells ; v v, capillary blood-vesselt, : d, finer meshwork from near a capillary. X 352 diams. THE HUMAN CHORION. 331 has just been sketched ; it may be considered as the proper tissue of the villus. It consists of a network of protoplasmic threads, which start from nucleated masses (cells) . There are many large meshes which are partly occupied by the coarsely granular wandering cells, l.l, which are scattered about, and are usually present in large num- bers. About the capillaries the network is much more finely spun. Kastschenko, 85.1, 454, found the wandering cells most abundant near the epithelium, but I have . noticed no such peculiarity except that they do not often enter the dense perivascular tissue; and as the blood-vessels are centrally situated, the adenoid tissue and the wan- dering cells in it are of course more peripheral. It seems to me that the leucocytes are distributed more or less evenly throughout the adenoid tissue. I fail to recognize any intercellular substance. The abundance of nuclei deserves special mention. Around all the non- capillary vessels the mesoderm is very different, for it exhibits dis- tinct intercellular substance, with a tendency to fibrillar differentia- tion in quite a wide zone around the blood-vessels ; in this zone the cells become elongated and irregularly fusiform ; around the larger vessels the cells are grouped in lamina, making the structure similar to that already described in the walls of the vessels of the umbilical cord; after the perivascular coats acquire a certain thickness the cells of the inner layers are more elongated, more regularly fusiform, and more closely packed than those of the outer layer ; the transition from the denser to the looser tissue is gradual. We are perhaps entitled to recognize in the denser inner layer the media, in the outer looser layer the adventitia, although neither of the layers has by any means the full histological differentiation characteristic of the like-named layers of the blood-vessels of the adult. Blood-Vessels of tlie Chorion. — As already stated the entire chorion is vascular at an early stage, but the vessels abort very soon over the chorion Iseve, while over the frondosum they acquire a great development in connection with the formation of the placenta; it seems to me more convenient to deal with them in connection with that organ, and accordingly the reader is referred to Chapter XVII. Fluid Contents of the Cliorionic Vesicle. — In early stages, as we have seen, there is a large chorionic cavity, which in later stages is obliterated by the expansion of the amnion. The space between the chorion on the one hand and the amnion and the yolk- sac on the other is filled with a fluid, which is coagulated by the action of the hardening agents, making a network of threads. This obser- vation, which has been often verified, is all that we know concerning the nature of the chorionic fluid ; it is probably of a serous character and may very likely be found to contain free connective-tissue cells (wandering cells or leucocytes) . Evolution of tlie Chorion. — There can be little doubt, if any, that the chorion arose by the growth and expansion of the abdominal somatopleure, in result of the increase of the yolk-material in the earliest amniota. It can, therefore, not be regarded as originally a new organ. When the amniote type of development was evolved a portion of the original chorion was differentiated and separated as the amnion from the primitive membrane, leaving the rest as the true chorion (false amnion or membrana serosa), enclosing all the other 332 THE FCETAL APPENDAGES. parts of the embryo and making the chorionic vesicle. This vesicle, therefore, results from the development of the amnion, or perhaps the formation of the amnion is a result of the development of the vesicle. It is customary to refer to the amnion as playing the lead- ing role, but of this there is no certain proof, though the conception is natural and plausible. The possession of a true chorion is as characteristic as the possession of an amnion or allantois in the higher vertebrates, so that from a morphological standpoint the term Chorionida vi^ould be as appropriate and justifiable as the terms AUantoidea or the more generally used Amniota. '' In the mammalia the chorion, being the outermost member of the ovum, is brought into immediate contact with the uterine wall, and has consequently undergone many and complex modifications in con- nection with the evolution of the placenta. But while the chorion in the placental mammals is the organ of communication between the mother (uterus) and embryo, its vascular connection with the latter is maintained through the intervention of the allantois, which thus co-operates in an essential manner in developing the placenta, though, strictly speaking, it does not participate in forming the actual pla- centa, meaning by placenta the organ produced through the intimate union of foetal and maternal tissues. It is evident that, as Minot has maintained, the placenta is necessarily chorionic. Further re- marks on this subject will be found in Chapter XVII., "The Pla- centa." CHAPTER XV. THE AMNION AND PEOAMNION. Definition of the Amnidn. — The amnion is a thin, pellucid, non- vascular membrane, and is the innermost of the envelopes enclos- ing the embryo or foetus. Its origin and formation have been described alreadj-, p. 381. Morphologically it is a part of the body- wall (somatopleure) of the foetus, and therefore consists, as we have seen, of two layers, one epithelial continuous with the ectoderm (seu epidermis) of the embryo, the second of loose connective tissue con- tinuous with the somatic mesoderm (outer leaf of mesoderm after the appearance of the body-cavity) . The epithelial layer is turned toward the embryo, and the connective-tissue layer consequently lies upon the outside of the amnion away from the embryo, and toward the chorion and the uterine wall. Growth of the Amnion. — Concerning the growth of the am- nion I know of no exact measurements. During the first three weeks it stands off a little from the embryo, but during the fourth week the latter grows so rapidly that it takes up nearly the whole of the amniotic cavity ; during the second month the amnion enlarges rap- idly so as to leave considerable space for the amniotic fluid; the amnion continues, of course, to expand during all the following months, but after the fourth month it fits pretty closely around the embryo, but is kept distended by the amniotic fluid. The amnion does not grow around the allantois-stalk or umbilical cord of man as it is commonly stated to do, but, on the contrary, springs from the stalk in the same manner as from the body of the embryo, and is separated from the stalk in the course of development, as is described more fully below, in connection with the histology of the allantois-stalk. Histology of the Amnion. — For a certain period after it is first formed the amnion, in all embryos I have been able to examine, consists of two layers of cells, both very thin and with the nuclei considerable distances apart, but sometimes in little groups ; between the two layers is a distinct space. The ectodermal layer is the most regular and the best defined as to its inner boundary. The meso- dermal layer is more or less irregular and sends at intervals a process across the space between the two layers to be attached to the ectoderm. In a human amnion of a normal two months' embryo. Fig. 191, the mesoderm has become very much thicker, and is readily seen to be separated into two parts, the thin mesothelial layer, msth, cover- ing the surface of the amnion toward the chorion, and a mesenchymal layer, mes, which makes up the greater part of the membrane ; the mesenchyma is probably derived from the mesothelium by prolifer- ation and migration ; I have noticed many indications of the process, but have never studied it carefully. The ectoderm, Ec, is very 334 THE FCETAL APPENDAGES. much in the condition just described for the earlier stage, but in specimens of three months' amnia it has become thicker, and its cells are beginning to change into the cuboidal form of later stages. No blood-vessels or nerves are known to exist in the amnion of i Msth I'iG. 191. — Section of the Amnion Covering the Placenta of a Two Months' Embryo. Jfc, Ecto- derm : mes^ mesoderm (mesenchymal) ; msth^ mesothelium. x 250 diams. the human embryo, although in sheep embryos in very early stages the vessels have been noticed by Bonnet to extend a short distance into the amnion from the body-wall. Histological Differentiation. — The tissues of the amnion do not progress beyond an early embryonic stage ; the ectoderm remain- ing at the one-layered stage. the mesoderm preserving much of the primitive mat- rix. Emery ("Arch. Ital. Biol.,"' III., 37) has directed attention to the primitive homogeneous matrix of the vertebrate mesoderm, and es- pecially to the separate sub- epidermal layer of the em- bryo, which contains no cells at first. In the human am- nion there is a non-cellular layer under the epithelium, as is well shown in Fig. 193, A and B. Sometimes this layer is invaded to a certain extent by connective tissue- cells, B ; in other cases the portion of the matrix toward the chorion acquires a fibril- lar character. A, as if par- tially resorbed, but in no case have I seen the matrix entirely altered from its primitive character. The cells of the mesoderm lie in lacu- nae ; they are flattened in the plane parallel to the surface, and hence in vertical sections, Fig. 192, appear more or less fusiform. They present no special features, so far as I have observed, to distinguish them from other embryonic connective-tissue cells. Their bodies have little afiinitj' for coloring-matters, hence it is difficult to follow the processes by which the cells are united. Their nuclei are at first round or oval. After the third month they often show a great vari- ety of alterations in shape and size. Figs. 193, 194; some of the nuclei are then very large, with a distinct network, d; others are smaller and djffer but slightly from the normal ; some are very irregu- lar, b, and others again strangely elongated, a; many other forms Fig. 192.— Two Sections of the Placental Amnion; A, from an embryo of the eighth month; B, at term. ■€ci, Ectoderm; mes^ mesoderm; a, layer of meso- dermic cells. X 340 diams. THE AMNION AND PROAMNION. 335 ^^ 193. — A Natural Group of Nuclei from the Meso- Foetus of the Fifth mouth. Fig. derm of the Amnion of X 1225 diams. beside those represented in Fig. 193, are to be found. The changes indicated I consider of a degenerative character, and in fact many of the nuclei are breaking down, for one finds in some specimens every stage between a nucleus and scattered granules — nuclei, nuclei with indistinct mem- branes, nuclei without mem- branes, masses of granular matter, clusters of granules crowded together, and final- ly other clusters more or less scattered. This degenera- tive process may be com- pared with that described by Phisalix {Arch. Zool. Expt., Ser. 11., T. III., 382) as occurring in the blood-cells of the spleen of teleosts. Compare also the chromatine degeneration observed by Flemming to occur in ova of the verte- brate ovary (His and Braune's Archiv, 1885, 321-244:). In the human amnion the nuclear degeneration described is not always to be recognized so clearlj^, although the nuclei in all amnia older than three months, which I have observed, are more or less irregular and distorted. Finally it is to be added that not infrequently the cells form a distinct epithelioid layer upon the surface of the amnion next the chorion as represented in Fig. 192, B, a. The epithelium of the amnion varies in appearance, as seen in transverse sections. Usu- ally the cells are cuboidal or low cylinders. Fig. 192, A, each with a rounded top, in which is situated the more or less nearly spherical micleus; sometimes, however, the nuclei lie deeper do-wn. Less frequently the epi- thelium is thin. Fig. 192, B, and its nuclei, which are trans- versely elongated, lie further apart. It is probable that those differences are not structural, but conditional upon the greater or less degree to which the am- nion is stretched. I have ob- served no constant differences between the placental and the remaining amnion. The most interesting peculiarity of the epithelium is best seen in surface views; namely, the intercellular bridges. They display themselves with a clearness which I have never seen in other epi- thelia; see Fig. 195. Fig. 194. — Meeodermic Nuclei of the Amnion of an Embryo of about Four Months. X 713 diams. 336 THE FCETAL APPENDAGES. nu. The nuclei, nu, are relatively large, rounded, with distinct out- lines ; they have a more or less well marked intra-nuclear network, with thickened nodes, and a small number of deeply stained granules, which are probably chromatin. Each nucleus is surrounded by a cell-body, pi, and the adjacent cell-bodies are separated from one another by clear spaces. With high powers, as represented in the figure, one sees that io maan zon Larva. JSc, Ectoderm ; Md, medulla ; nch, notocjiord ; ^- "^ • . ■^° »Oun as Uie meSv/- o.pl, oral plate; Li, liver pouch, extending into the mass derm IS developed COmnlcte- 01 the yollc cells; 6(, blastopore. After 0. Kuptfer. i j j_i •, i. ly around the ovum it of course separates the yolk and the ectoderm, and as soon as the ccelom is developed in the abdominal region there is a layer of mesoderm enclosing the yolk ; now as the yolk is entodermal it follows that the yolk, together with the mesoderm layer around, are morphologically part of the splanchnopleure. This splanchnopleuric bag is the homo- logue of the yolk-sac. In the meroblastic anamniota* (elasmobranchs) there is a separation of the yolk-sac from the embryo, and it hangs down from the intestinal canal of the embryo by a small stalk; but it is covered by the somatopleure just as in the more primitive types, so that the true yolk-sac is inclosed in a second membrane. The same arrangement exists in the amniota ; there is an inner or true yolk-sac formed by the vitelline entoderm and splanchnic mesoderm, and an * For some further details see P. Mayer, S7.1, 346. THE YOLK-SAC. 347 outer somatopleuric sac, homologous with the external membrane of the elasmobranch, but-commonly known as the membrana serosa in Sauropsida, and as the pri^nitive chorion in mammals. The term yolk-sac, as applied to the el^smobranchs, includes both the inner or true yolk-sac and the outer somatopleuric covering, homologous with the chorion ; but as applied to amniota, it commonly refers only to the inner sac, to the exclusion of the chorion. Yolk -Sac of Sauropsida. — The manner in which the embry- onic archenteron is separated from the yolk-sac has already been described, p. 355, and we saw that the peripheral part of the area pellucida, the whole of the area opaca, of the so-called germinal wall and of the yolk-mass are included in the yolk-sac ; all the parts men- tioned constituting collectively the entodermal lining of the yolk- sac. The whole of the vitelline entoderm tends to assume a distinctly epithelial structure ; the change begins in the region of the embryo and thence spreads gradually in all directions ; in the region of the area pellucida the vitelline epithelium (Dottersackepithel) has thin wide cells ; in the region of the area opaca the cells are high cylinder cells, Fig. 198, c, of somewhat irregular shape, containing a loose Fio. 198.— Wall of the Yolk-Sac in the Area Opaca of a Chick of the Second Day. Mes, Meso- derna; F, V, blood-vessels containing a few young blood-cells; Ent, entoderm; c, four entoder- mal cells showing distinctly. (Compare with Fig. 301.) network of granular protoplasm; the lower ends of the cells are rounded and projecting and have a well-marked border of dense pro- toplasm ; the nuclei are variable in size, but for the most part large, often three or four times greater in diameter than the neighboring mesodermic nuclei ; they have usually one, sometimes two, conspic- uous nucleoli ; the nuclei always lie on the upper or basal ends of the cells, generally near one side — a point best made out in surface views; the cells further contain yolk grains, which appear to be undergoing resorption ; near the area pellucida the cells are smaller, the network of protoplasm closer, and .the yolk grains either absent altogether, or, if present, small in size and few in number ; the transi- tion to the thin entoderm of the area pellucida is quite abrupt, accord- ing to H. Virchow, 75. 1, but I have found in some cases a gradual change.' Toward the periphery of the area opaca the entodermal cells become larger and richer in yolk-grains and pass gradually into the germinal wall. The cylinder cells of the opaca entoderm stand at various inclinations, so that they are cut obliquely for the most part ; consequently only here and there can we recognize them clearly. 348 THE FCETAIi APPENDAGES. as in Fig." 198, c. The germinal wall is the connecting link between the epithelium on the dorsal side of the cavity of the yolk-sac and the yolk forming the floor on the ventral side of the cavity. The structure and metamorphoses of the germinal wall have been the subjects of much discussion, leading to very little result, for many authors have sought in the germinal wall the origin of mesodermal and even of ectodermal cells ; that all such views are erroneous was demonstrated by H. Virchow, 75.1; it would have saved a great deal of confusion if his admirable little paper had received the atten- tion it deserves. H. Virchow has since confirmed and amplified his results in two valuable memoirs, 91.1, 92.1.* The germinal wall is the transition from the cellular Opaca entoderm to the non-cellular yolk, hence it consists of protoplasm charged with yolk grains and having numerous nuclei, which toward the embryo become situated in discrete cells, which, as we pass to the opaca, gradually take on a more and more epithelial character ; the non-cellular yolk has nuclei also, but they are further apart than those of the germ wall ; these nuclei are the so-called parablastic nuclei (see p. 353). As develop- ment proceeds we see the area pellucida encroach upon the opaca, the area opaca upon the germinal wall, and the germinal wall upon the yolk proper ; the whole series of changes may be described as a centrifugal metamorphosis. The mesoderm of the yolk-sac is a thin layer which gradually spreads over the yolk, and so slowly that, according to M. Duval ("Atlas," Fig. 652), it does not completely enclose the yolk tmtil the seventeenth day in the chick. The early appearance of blood- vessels in it marks out the area vasculosa, which is a part of the 3'olk-sac ; the expansion of the vascular area has already been de- scribed, p. 276. A further peculiarity of the mesoderm is that it sends down partitions into the mass of yolk, carrying along the blood-vessels, and thus increasing the absorbent surface ; the parti- tions in the chick begin to appear during the sixth day, and continue multiplying and growing for at least ten days. As the yolk-sac contains the nutritive material for the embryo, it diminishes in size as the latter grows ; the shrinkage causes the sac to become, the sixth or seventh day in the chick, fiaccid and some- what irregular in shape, two peculiarities whigh become more and more marked as development progresses. By the eighteenth day the sac is very much smaller ; by the nineteenth the reduction is still more striking and the sac begins to be withdrawn within the body of the chick, and before hatching the embryo takes in. the yolk-sac completely through the umbilical opening; during its retraction the sac has a characteristic hour-glass shape, owing to the narrowness of the umbilicus. Concerning the structure of the yolk-mass during the resorption of the yolk material we know very little, and of the physiology of the assimilation of the yolk, almost nothing. Von Baer pointed out, 28.3, that the yolk becomes more fluid in the chick, and H. Eathke, 39.1, 113, that in the snake the separate yolk-granules disappear, and the yolk becomes a greenish-yellow homogeneous translucent * I regret that these memoirs came to my hands too late to enable me to incorporate Vlrchow'S results in the text. THE YOLK-SAC. 349 fluid. H. Strahl, 87. 1, gives an important account of the yolk-sac in the lizard, showing that the dissepiments of mesoderm are covered with large yolk-cells — the whole yolk apparently becoming cellular in later stages ; the cavity of the sac is very distinct ; the sac itself be- comes flattened ; and it is only on the inferior side that the dissepiments acquire a considerable development, and on this lower side the cellular structure is perhaps never fully attained. The regular form of the yolk-sac persists in the lizard, but in the snake, H. Eathke, 39.1, 183-184, it becomes flaccid and irregular. Yolk-Sac of Mammals. — In order to imderstand clearly the development of the mammalian yolk-sac, it is best to start with the two-layered blastodermic vesicle, with a small embryonic area in which there is mesoderm ; the inner layer of the vesicle is the homo- logue of the yolk-mass of Sauropsida, and is able to assume the cellu- lar structure owing to the loss of yolk, which is undoubtedly also the cause of the large size of the cavity of the vesicle — this cavity being, as we have seen, the vitelline cavity ; the inner vesicle then is the homologue of the entodermal part of the yolk-sac. The extra- embryonic mesoderm and coelom are extremely variable in extent in the mammalian ovum; in man, as we have seen, the mesoderm is very early developed completely around the yolk- vesicle, and so is the ccElom, so that in the earliest accurately known of human stages the yolk-sac and chorion are completely differentiated. In the sheep, and probably, in all ruminants, there is a similar early separation of the yolk-sac and chorion. In the rabbit the mesoderm never extends over more than about half of the blastodermic vesicle, but the coelom extends nearly to the periphery of the sheet of mesoderm ; hence we have a half-way separation of the yolk-sac and chorion. In the opossum the mesoderm extends about half-way over the blastodermic vesicle, but the coelom is developed only around the allantois, so that there is only a very partial separation of the yolk-sac and the chorion. In both rabbit and opossum the lower half of the yolk-vesicle is in direct contact with the ectoderm, preserving to this extent the con- dition of the stage of the two-layered blastodermic vesicles. That the partial extension of the mesoderm represents a modified condition is evident, since in all non-mammalian vertebrates both mesoderm and coelom extend completely around the yolk. _ Hence, the complete separation of the yolk-sac in man and the sheep is nearer the ancestral type than the relations of the extra-embryonic germ- layers to one another in the rabbit and the opossum. The question as to what was the primitive mammalian arrangement must be left open ; we cannot say whether the opossum or man most nearly rep- resents the ancestral type. Man. — The human yolk-sac is an appendage of the digestive canal formed by the extra-embryonic somatopleure. At the beginning of the third week the diameter of the yolk-sac is about equal to the length of the embryo. By the middle of the third week the sac has become distinctly pear-shaped and is attached by its pointed end to the intestinal canal of the embryo, Fig. 17. The sac continues growing up to the end of the fourth week, after which it enlarges very slightly, if at all; its diameter is from 7-11 mm. It is then a pear-shaped vesicle attached by a long stalk to the intestine, the 350 THE FCETAL APPENDAGES. stalk having been formed by the lengthening of the neck of the yolk- sac, Fig. 169. Sections show that the sac is hollow, with a lining of entoderinal cells, and a thicker layer of mesoderm, containing blood-vessels ; the network of vessels imparts a characteristic appear- ance to the external or mesodermic surface of the yolk-sac, compare Fig. 175. The accompanying Fig. 199 represents a section of the yolk-sac of an embryo of about 1 mm., after Keibel. The cavity of the yolk-sac extends at first through the stalk to the intestine, but it early becomes obliterated in the stalk. The ento- derm disappears altogether and quite early in the yolk-stalk ; thus in an embryo of 13.5 mm., His ("Anat. menschl. Embryonen," III., 20) found only remnants of it in the ^ ,nn o ..• * tv -tr 1, <= f stslk. lu tho voslcle itself the en- FiG. 199.— Section of the Tolk-Sao of a . , , . Human Embryo (No. 11, p. 290. Ent, toderm consisted m a Very young fet°e'?l.™ifte'?Fr. feTbet™' ^' ^'°°''" ovum of a single layer of cuboidal cells (Graf Spee, 90.1, 163), but is said to become fatty and to change into a pavement epithelium, which is also thrown up into vascular villi (Kolliker). In regard to the further contents of the yolk-sac, Von Baer states, 37. 1, 372, that in young ova (at six to seven weeks) the contents are sometimes as thick and yellow as the yolk of a bird's egg; in ova of this period the thinner the contents the more rounded and fuUy distended is the yolk-sac. A little later the contents are always fluid, but at the end of pregnancy, according to B. S. Schultze, 61.1, when the sac has shrunk to 4-7 mm. in diameter, it contains variable quantities of fatty substances and carbonates. It thus appears that during the first month, at least, the yolk-sac does contain more or less true yolk — an idea which is confirmed by Rauber's observations on the rab- bit's ovum. It seems indeed probable that the rudimentary yolk-sac of man still performs for a short period the function of a food reser- voir for the embrj'o, amnion, and the chorion. (B. S. Schultze, 61.1). Sheep. — The two-layered blastodermic vesicle, with an embryonic shield, has an elongated form ; the mesoderm spreads out gradually between the ectoderm and entoderm (yolk-vesicle) starting from the shield ; the coelom is developed in it as it spreads, so that by the thirteenth day (E. Bonnet, 89.1, Taf. VI., Fig. 3) about a third of the ovum is furnished with mesoderm, and in this third the splanch- nopleure of the yolk-sac is completely separated from the chorionic somatopleure, while elsewhere the yolk entoderm is stiU directly in contact with the ectoderm ; this stage, see Fig. 200, is as far as de- velopment progresses in the rabbit. In the sheep the development of the mesoderm and the coelom proceeds, until about the seventeenth day the yolk-sac is completely separated from the chorion ; the yolk- stalk remains short, but the yolk-sac proper becomes drawn out and twisted, following in its growth the characteristic elongation of the ruminant chorionic \esicle. THE YOLK-SAC. 351 Rabbit. — The development of the extra-embryonic mesoderm and coe- lom is entirelj- arrested at about the stage reached by the sheep on the thirteenth day, so that the yolk-sac and chorion are never differentiated over more than half the ovum, the inferior hemisphere of which re- mains in the stage of a two-layered blastodermic vesicle, and is said by Duval to degenerate and be resorbed. The accompanying diagram will suffice to render the disposition clear ; it will be seen at once that the cho- rion, Cho, exists only part- way round the ovum.. I introduce here Fig. 301 of a section through the wall of the yolk-sac of a rabbit embryo of thirteen days; the structure closely resembles that of the area ■Ent Fig. SOO. — Diagram of the Embryo and Tolk-Sao of a Eabbit. eoe^ Ooelom ; Clio, chorion; YTc, yollc-sac; mes, mesoderm, v.ty vena terminalis; Entt entoderm; Ec, ectoderm. Fig. SOI.— Vertical Section of the Wall of the Yolli-Sao of a Rabbit Embryo of Thirteen Days. FF, Blood-vessels; 6/, blood-cells; mes, mesoderm. opaca of the bird's yolk-sac, Fig. 198, except that the entodermal cylinder epithelium is composed of much smaller cells in the rabbit, owing to the absence of Am ^K r' yolk. Opossum. — Our knowl- edge rests mainly upon the observations of Selenka, 87.1, whose diagram I have copied. Fig. 202; the embryo, Emh, is almost en- tirely covered by the pro- amnion. Pro. am, the am- nion. Am, being very much reduced; the allantois. All, projects also into the yolk- sac cavity, Yk; owing to the development of the pro- amnion and allantois the cavity, Yk, of the yolk-sac acquires a very complicated form; the extra-embryonic ccslom, Coe, is hardly more Ppo.ani hs.t Ent Fig. 302.- its agi -- - .- .J, Bt, Sinus terminalis ; Oho, chorion ; Am, amnion ; Ec, ectoderm ; mes, mesoderm ; Enib, embryo ; Pro. am, pro-amnion ; Ent, entoderm ; Yh, cavity of yolk-sac; >Cii, allantois; Coe, ccelom. After E. Selenka. 353 THE FfETAL APPENDAGES. than a space around the allantois, and consequently the true cho- rion is reduced to an insignificant area, Choj the extra-embry- onic mesoderm, mes, extends over nearly half the ovum, from st to 5^, but contains — except around the allantois — no coelom; in this sheet of mesoderm the blood-vessels of the area vasculosa are devel- oped ; and as there is no coelom over the area, the vessels are almost as closely related to the ectoderm, Ec, as to the entoderm, Ent. Here, then, we have the mesoderm spreading out as in the rabbit, but the development of the coelom is arrested. Although the opos- sum stands low in the mammalian scale, its foetal membranes show many changes from the sauropsidan type and are probably modified in an aberrant manner, differently from mammals of other classes. For the peculiar relations of the yolk-sac to the allantois, see the description of the latter organ. The So-called Parablastic Nuclei of the Yolk. — In mero- blastic vertebrate ova, after the embryo is formed, there appear in the yolk near its surface underneath the extra-embryonic blastoderm peculiar large nuclei, which are commonly designated as the para- blastic nuclei. The following description applies to Pristiurus.* The extra-embryonic ectoderm is a thin, much-fiattened epithelium lying close to the yolk; below the ectoderm is the superficial layer of the yolk, a broad stratum of protoplasm with scattered small yolk granules ; a little deeper down a row of irregular vascular spaces, and again a little deeper a layer of very big nuclei, each with a dis- tinct intranuclear network and several deeply -stained nucleoli ; the nuclei vary in size, being from two to five times the diameter of ';he nuclei in the embryo. The upper part of the protoplasmatic stratum contains numerous small and a few larger yolk-grains, and contains near and under the embryo small nuclei; the middle part of the stratum contains the vacuoles, the big nuclei, and but few yolk grains ; the deepest part contains larger granules and merges grad- ally into the yolk proper. (See also His, 82.1, 75, and Riickert, 85. 1.) Riickert designates these nuclei as " Merocytenkerne," and the cells which they represent as " Merocyten." H. E. Ziegler, 87. 1, states that the parablastic nuclei of teleosts multiply up to a certain stage by indirect division, but later they assume a peculiar habitus and multiply by indirect division, and assume various shapes. These changes are perhaps connected with the death of the nuclei, their active functions being completed. The special function of the protoplasmic layer appears to be the assimilation of the nutritive yolk. Riickert also maintains, but without proper evidence, it seems to me, that merocytes become cells, some of which join the ectoderm, some the entoderm, and yet others the mesenchyma. In the Sau- ropsida we find similar nuclei and similar relations of the nucleated layer, but in them the protoplasmic layer becomes the epithelium of the yolk — see especially H. Strahl, 87. 1 — and I consider it probable that these parablast nuclei in all meroblastic ova belong to the vitel- line entoderm. J. Riickert, 92.3, clain^s that some of the "para- blast nuclei" are derived from spermatozoa, which enter the yolk but do not unite with the female pronucleus ; it is doubtful whether * From sections in the collection o£ Professor His, which he generously permitted me to study. THE ALLANTOIS. 353 or not any of these spermatozoa nuclei share in the production of embryonic tissue. In holoblastic mammalian ova the vitelline entoderm is cellular, and no nuclei are known similar to large " parablastic " nuclei of meroblastic ova. II. The Allantois. The origin of the allantois we have already described, p. 257. It arises as an entodermal evagination behind or below the blastopore and anus, and extending into the anterior end of the primitive streak. Allantois of Sauropsida.— The allantois becomes rapidly dis- tended by the enlargement of its entodermal cavity, and hence comes to project freely into the coelom as a vesicle, attached by a pedicle to the anal end of the intestinal canal. This vesicle is of course formed by the splanchnopleure, and therefore lined by entoderm, with an outer layer of mesoderm. In the chick the vesicle is about as large as the eye by the middle of the fourth day, and after that grows very rapidly, becoming bent so as to project on the right side of the em- bryo ; -by the end of the fourth day it is already about as large as the mid-brain at that stage (c/. Duval, " Atlas," Fig. 122). During this expansion its mesoblastic walls, which are at first very thick, become thinner, and at the same time the allantoic blood circulation becomes important. The blood is supplied directly from the dorsal aorta, which terminates in a fork, of which each branch is an allan- toic artery, and the blood is returned by two allantoic veins, which run along in the body walls. By the third day in the chick they are found, after having united into a single trunk, to open into the vitelline vein, close behind the liver. The allantois continues enlarg- ing, and pushes its way very rapidly into the extra-embryonic coelom, between the amnion and chorion (serosa or false amnion) . Curving up around the right side of the embryo the vesicle comes to lie on the dorsal side, above the amnion, and separated from the shell by nothing more than the thin chorion. In this position its rapid growth continues ; it forms a flattened bag, covering the right side of the embryo, and rapidly spreading out in all directions. It is filled with iiuid, so that in spite of its flattened form its opposite walls are dis- tinctly separated from one another. The expansion steadily contin- ues, so that by the ninth day the allantoic vesicle nearly surrounds the yolk; during the eleventh day the outer wall of the allantois begins to grow, together with the chorion ; hence in opening an egg during the later stages of incubation, there is much danger of tear- ing the allantois when the shell membrane is removed. The embryo may now be said to be surrounded by two new membranes — the outer and inner walls of the allantois. About the sixteenth day the allan- toic sac completely envelops the ovum, and by the seventeenth its edges fuse. The closure, according to Duval, 84.2, takes place in such a manner "that there is formed a sac of ectoderm, inclosing the remnant of white at the pointed end of the ovum ; this sac, as well as the yolk, is inclosed by the allantois. Histology. — ^Concerning the tissues of the allantois we possess 33 354 THE FCETAL APPENDAGES. very little information; the entodermal lining appears, at least in advanced stages, as a low cuboidal epithelium, while the mesoderm is thicker and consists of more or less widely separated mesenchymal cells, covered by a thin mesothelium ; the mesoderm contains blood- vessels ; and since contractile pulsatioms have been observed in the allantois of the chick toward the close of ^incubation, it is probable that some of the mesenchymal cells assume the form of smooth mus- cle fibres. Where the allantois fuses with the chorion (membrana serosa) the mesothelium of both layers disappears, and there is no demarcation or difference between the allantoic and chorionic mesen- chyma — ^compare Duval, 84.2, PI. IX., Fig. 8. Allantois in Mammals. — The allantois is very variously devel- oped in the mammalia, being a distinct vesicle in most forms, but never growing around the embryo and yolk, as in birds. In the ojjossuiii, Selenka, 87.1, 141, the allantois does not even come in contact with the chorion, but invaginates the wall of the yolk-sac, as shown by the diagram. Fig. 202 ; the wall of the yolk-sac forms a pocket in which the allantoic vesicle is lodged ; the walls of the two organs do not unite ; the pocket in the yolk-sac has curious relations to the main blood-vessels running from the embryo to the area vas- culosa, for the two omphalo-mesaraic veins run straight back from the embryo along the edges of the mouth of the pocket, while the single omphalo-mesaraic artery runs in a great arch in the median plane round the bottom of the pocket. These features are beautifully ihustrated by Selenka, 87.1, Taf. XXVII., Figs. 1-4. _ The aUan- toic wall is figured by him. Fig. 4, Taf. XXV., as consisting, in an embryo six days old, of an inner layer of entodermal cuboidal epi- thelium, a thin outer layer of mesothelium, and a thicker layer of vascular mesenchyma. In the rabbit (and probably all rodents) the allantois becomes a moderate-sized vesicle, Fig. 196, All, which grows out until it reaches the placenta chorion, with which it unites to co-operate in the development of the placenta. In insectivora the allantois seems to resemble that of the rodents, though it acquires greater size; exact investigations are much needed. In ruminants the allantois expands very earlj^, growing out transversely and con- tinuing to enlarge with extraordinary rapidity until it takes up most of the chorionic vesicle, thus becoming, relatively to the embryo, of enormous size. For further details see R. Bonnet, 89.1, 26-36, and Bischoff, 54.1. A few histological facts may be gleaned from the very verbose article on the allantois by A. Dastre, 76.1, 17-44. Man.- — The allantois in man and other primates is essentially different from that of any other known amniote. It never becomes a free vesicle, but always remains a narrow tubular diverticulum. In man the embryo, when the amnion is formed, becomes every- where separated from the chorion, except at the hind end, where the accumulation of mesodermal cells into which the allantoic diverticu- lurn extends, see Figs. ITO, ISO, and 222, constitutes a thick stalk. This stalk has been named the Bauchstiel by W. His ; it may be regarded as a direct prolongation of the body of the embryo ; it per- manently connects the embryo proper with the chorion. The amnion springs from the sides of the Bauchstiel in the same manner as from THE ALT.ANTOIS. 355 •TTIBS ~-Ent the body of the embryo. In man, therefore, there is no free allan- tois, and the history of the organ may be said to be reduced to that of the entodermal allantoic diverticulum. The entodermal allantois is a small, long, epithelial tube extending, as we have seen, to the chorion, p. 397, Fig. 170. The tube increases very little in diam- eter, after the second month; compare A and B, Fig. 444. It is very persistent, appearing usuallj- even in the cord at fuU term, at least in the proximal end, according to KoUiker (" Entwickelungs- gesch. , " 3te Aufl. , p. 34) . After the second month it is a small group of epithelioid cells, with distinct walls, irregularly granular contents, and round nuclei; around the cells. Fig. 303, etit, which may or may not show a remnant of the central cavity, there is a slight condensation, mes, of the connective tissue to form, as it were, an envelope. This structure has been regarded by Ahlfeld and others as the persistent yolk-sac. I think the correct interpretation was first suggested by KoUiker. It has been supposed by some writers that the human allantois grew out as a free vesicle. Haeckel even went so far as to prophesy that when a human embryo of the right stage should be obtained, this condition would be found. Shortly after this W. Krause published a description, 76.1, of an embryo, which he said was human and had a free allantois. mistaken, the former through hasty and unfounded speculation, the latter through an error as to the identity of his embryo. W. His has shown that it was certainly not human, probably not even mam- malian, but avian, 80.1, 72. Krause still maintained that it was human. The discussion as to this specimen was a long and animated one, but has now little interest except historically. See Krause, 80.1, 81.1, 2, KoUiker, "Entwickelungsgesch.," 1879, 3Q6, 1013, Ahlfeld {Cbl. fiir Gynak., 1880, No. 35), Krause, ibid., 81.1, and Ecker in His' Arch. f. Anat., 1880, 405. Allantoic Fluid. — The fluid contents of the allantois cannot be weU studied in man, owing to the minute size of the cavity of the organ ; but when the aUantois becomes a large sac, as in the cow and pig, the fluid is readily collected. There are many observations recorded concerning the chemical composition of the fluid, but the best work known to me is that of Doderlein, 90.1, on the fluid in cow embryos. His results may be summarized as foUows : Fig. 203. — Section of the Allantois from the Umbilical Cord of an Embryo of Three Months. ent. Entoderm; mes^ mesoderm. X 340 diams. Both Haeckel and Krause were Allantoic Fluid (Cow). XaCl. NaO. KO Ca. Mg. Average per cent 0,244 0.163 0.093 0.015 0.049 No. of obs 8 8 8 10 , 8 356 THE FCETAL APPENDAGES. Embryo. Fluid. 1. 2. 3. 4. 5. 6. 7. No. Wt. grms. c.c. Per cent of Wt. of emb. Total N. Proteid N. N— Prot. N. 1 83 73 837 3 87 90 108 3 876 800 73 0.136 0.049 0.086 4 860 400 111 5 480 860 177 6 600 850 141 7 1,880 3,400 178 0.124 0.093 0.032 8 1,700 1,800 76 9 1,800 1,400 77 0.164 0.141 0.023 10 5,183 8,000 38 0.371 0124 0.147 11 6,690 2,460 36 0.848 0141 0.107 IS 6,700 3,500 52 0.196 0.122 0.074 13 8,850 6,600 77 0.808 0.141 0.061 14 11,300 5,000 44 0.389 0187 0.142 16 14,900 6,600 •44 0.489 0.820 0.209 Am The allantoic fluid differs markedly from the amniotic — compare the tables above with those on p. 339 — and shows in its composition that it is an excretory product of the foetus, coming from the Wolffian bodies and the kidneys. In the chick, by the sixteenth day, deposits of water become abundant in the fluid (Foster and Balfour's " Ele- ments," second edition, 280). Notices of several of the earlier investigations on the allantoic fluid may be found in A. Dastre, 76.1, 45-61, together with some results of his own. III. The Umbilical Cord. Bauch.stiel. — As the Bauchstiel is the anlage of the human um- bilical cord, we must consider its structure and relations. As we have already seen, it is the prolon- gation of the tail of the embryo, Fig. lG(j, Al, running to the chorion and containing the tubular allantoic diverticulum, Fig. 170, Al ; it con- sists mainly of mesoderm, and from its side springs the amnion. Prof. His (" Anat. menschl. Embryonen," Heft III., 222-226) has made a spe- cial comparison which shows that the fundamental morphological rela- tions are the same in the human Bauchstiel as in the embryo proper, and that there are even traces of a rudimentary medullary groove. The resemblance is illustrated by the ac- companying Fig. 204. The am- nion, Am, arches over the dorsal side, which is covered over by thickened ectoderm, md, which His regards as the rudiment of the rnedullary groove; the archenteron is represented by the allantoic diverticulum, All, lined by the entodermal epithelium ; in the meso- Fio. 804.— Diagrammatic Section of the Bauchstiel of a Human Embryo, modified from W. His. Am^ Amnion ; md, medul- lary groove ; V, F, umbilical veins ; A, A, umbilical arteries; All, allantois; me, coelom. • THE UMBILICAL COKD. 357 derm run the two large allantoic veins, V, V, and the two smaller axteries, A, A; the space around the cord is of course part of the embryonic ccelom, coe; the amnion represents the somatopleure, the walls of the allantois the closed splanchnopleure. To convert the " Bauchstiel " into the umbilical cord, the somato- pleure bends down on each side, and finally closing on the ventral side below the allantois, shutting in a portion of the ccelom, and becomes separated from the amnion. The amnion separates from the embryo first, then from the embryonic end of the Bauchstiel, and last of all from the distal end of the Bauchstiel ; hence, when the closure of the somatopleure is completed the amnion arises no longer from the embryo, but only from the end of the cord, where it joins the chorion. The closure of the Bauchstiel forms a long tube run- ning from the embryo to the chorion ; the cavity of this tube is part of the ccelom ; the whole tube is known as the umbilical cord. When the ccelom of the cord is shut off, it is shut off in such a way that the long, narrow stalk of the yolk-sac, or the so-called vitelline duct is included, compare Fig. 223, v.s. This is possible owing to the rolling up of the embryo, which, as shown in Figs. 169, 172, 175, and 222, brings the Bauchstiel into, close proximity with the neck of the yolk-sac. The development of the cord shows that it is never covered by the amnion, which, on the contrary, is always separate from the cord proper. This point is important to note, because in most text- books the cord is erroneously described as covered by the amnion — compare Minot, 98, 380. Development of the Cord. — The origin of the cord from the Bauchstiel has been described in the preceding section. The Fig. 805. —Sections of Human Umbilicai Cords. A, Embryo of 21 mm ; B, embryo of sixty- four and sixty-nine days ; v, left umbilical vein; Ar, arteries; All, allantois; Coe, coelom; Ys, yolk-stalk or vitelline duct. structure and growth of the cord may be best studied in cross-sec- tions. Fig. 205. The coelom, Coe, is a large cavity and bontains the yolk- stalk, Ys, with its two vessels, and its entodermic cavity entirely obliterated; near the embryo the coelom may become much enlarged, and is often found during the second month, and even later, 358 THE pCETAL APPENDAGES. to contain a few coils of the intestine ; above the body -cavity is the duct of the allantois, All, lined by entodermal epithelium ; and in this region are situated the two arteries and single vein ; the section is bounded bj"- ectoderm.* The further development of the cord de- pends upon three factors: 1, the growth of the connective tissue and blood-vessels ; 2, the abortion of the coelom yolk-stalk and allantois duct, in the order named ; 3, differentiation of the connective tissue and of the ectoderm. The growth and differentiation of the mesoderm proceed rapidly, encroaching upon the coelom, which is obliterated (early in the fourth month). At first the connective tissue, Fig. 306, is composed merely ;Wf ^9^A -W- ^^H^^ ft. If. i r iK!4j/', Fig. 206.- -Connective Tissue of the Umbilical Cord of an Embryo of 81 mm. X 540 diameters. Stained with alum, cochineal, and eosin. of numerous cells embedded in a clear substance ; the cells form a complex network, of which the filaments and meshes are extremely variable in size; the nuclei are oval, granular, and do not have always accumulations of protoplasm about them, forming main cell- bodies, t I notice also a few cells, which I suppose to be leucocytes, but see no other structures. By the end of the third month the cells have assumed nearly their definite form; the protoplasm has increased in amount and forms a large cell-body around each nucleus. Fig. 207. The network has become simpler and coarser, the meshes bigger, and the filaments fewer and thicker ; in the matrix are numerous connec- * The ectoderm is often wanting, owing to its frequent destruction post mortem. t It is possible that the reticulum here described as cellular is, in part at least, composed of remnants of an early matrix, which shrinks up and is replaced by the clear matrix here de- scribed; my observations do not settle this question of the nature of the reticulum. THE UMBILICAL CORD. 3.V.) tive-tissue fibrils, not yet disposed in bundles except here and there ; as they curl in all directions many of them are cut transversely, and therefore appear as dots. In older cords there is an obvious increase in the number of fibrils, and they form wavy bundles. In the cord at term the matrix contains mucin, and may be stained by alum hsema- toxylin ; at what period the reaction is first developed I have not ascer- tained. I have observed nothing to indicate the presence of special lymph- channels in the cord at any period, but I have not investigated the point. Tait's lymph-channels are merely the intercellular spaces. V Tig. 20".— Connective Tissue of the Umbilical Cord of a Human Embryo of about thi-ee Months. X 511 diameters. Stained with alum, cochineal, and eosin. The ectoderm is at first a single layer of cells, a condition which is permanent over the amnion ; in an embryo of three months I find the two-layered stage. Fig. 208. The outer layer is granular, and in some parts each cell protrudes like a dome;* the inner layer consists of larger, clearer cells. By the fifth month the stratification of the epithelium becomes more evident and cornification begins. The ectoderm {Ec) , therefore, develops like the epidermi proper, although much more slowly, so that it never gets beyond the stage which the true epidermis reaches by perhaps the fourth month ; on the other hand it differs entirely from the amniotic epithelium. * From the investigations of Dr. J. T. Bowen on the development of the epidermis, which lie has been carrying on in the histological laboratory of the Harvard Medical School, it seems to me probable that this external layer is homologous with epitrichium. 360 THE FCETAL APPENDAGES. The blood-vessels steadily enlarge and acquire thick muscular walls. In the cord of an embryo of 21 mm., Fig. 205, the arterial iimscularis is well marked, the venous muscularis just beginning to show. At sixty-three days I find the coat thickened on all the vessels ; there is a gradual passage from the muscle cells to the sur- rounding connective tissue, so that one wins the impression that the Fig. 208.— Epithelial Covering o£ the Umbilical Cord of an Embryo o£ three Months. X 545 ' diameters. connective-tissue cells are being directly metamorphosed into mus- cular cells. By the tifth month the demarcation of the muscular coats is quite sharp, and it is probable that the further growth of the layer depends upon the growth of the elements it already contains and not upon the accretion of new ones ; that the muscle-cells do actually become bigger is easily ascertained by direct observation.* The obliteration of the coelom goes on rapidly during the second and third months, and by the beginning of the fourth is nearly or quite completed. The vitelline duct persists longer, but seems to disappear by the sixth month ; for a time it is distinguishable as a shrunken remnant in the midst of the connective-tissue cells of the cord. The allantoic duct occupies usually a position between the two arteries; it attains its maximum diameter about the fifth week, when it is a small epithelial tube, Fig. 203, of irregular width, as which it remains for some time without noticeable alteration ; during the third month it loses this character and becomes solid, by the enlargement of its epithelial cells ; the duct persists up to birth in this form, though losing, according to KoUiker, its complete conti- nuity; after it becomes solid there is a slight condensation of tissue around it. The Human XTmbilicaJ Cord at Birth. — The human cord is a long twisted rope of tissue, whitish in color, and attached by one end to the navel of the embryo, by the other to the surface of the placenta. Its dimensions are extremely variable at all periods ; at birth it is usually about fifty-five centimetres long and twelve milli- metres thick ; it is said that cords only fifteen centimetres long as one extreme, and over one hundred and sixty centimetres long as the * This offers another example of the rule that growth and cell mutiplication may be distinct processes. Compare Merk's remarks, "Denkschr. Wien. Akad.," liii., pp. 34^1, 1887. THE UMBILICAL COHD. 361 other extreme, have been observed. Its surface is smooth and glis- tening, except at the constricted foetal end, where the epidermis stretches about one centimetre on to the cord. The placental end expands to fuse with the chorionic m.embrane. The placental inser- tion is generally eccentric, that is, the cord joins the placenta at a point between the centre and margin of the letter organ ; usually the eccentricity is well marked, and not infrequently is so great that insertion becomes marginal ; in still rarer cases the cord joins the chorion outside the region of the placenta {insertio velmnentosa). Occasionally the cord forks before joining the chorion {insertio furcata) . The twisting of the cord is alwaj^s well marked externally at the time of birth by the spiral ridges within each of which a large blood- vessel runs. I have observed the number of spirals to vary from three to thirty-two ; the turns, beginning at the embryo, go usually from left to right, but sometimes from, right to left. The cause of the twisting, which begins about the middle of the second month, has been much and very unprofitably discussed. Of the many theo- ries on the subject which have been advanced, there is not one, so far as I know, having the slightest claim to acceptance. These vagaries have been collated by Hyrtl, 70.1, and also less, fully by Lawson Tait, 76. 1, who adds to them. All we can say is that the vessels grow faster in length than the cord as a whole, and therefore assume the spiral disposition ; the cause of this inequality is as com- pletely unknown to us as the causes of all the other inequalities of growth which occur in the embryo. One point must be specially mentioned, namely, that there is no reason to suppose that the cord as a whole actually twists any more than the spiral intestine of a shark is formed by twisting; many writers have falsely assum^ed the occurrence of this twisting motion, and have dissertated at no little length on the revolutions of the embryo in utero. There is no evidence that such rev(3lutions occur, nor have we any ground for assuming that the twisted appearance of the cord is due to an actual twisting like that of a rope; if a long rubber tube forms a coil within a short glass cylinder, it does not indicate that the cylinder has been twisted. The cord is covered by a layer of epithelium which is continuous at the distal end with the epithelium of the amnion. Its interior consists of a peculiar embryonic connective tissue known as Whar- ton's jelly, which is described below; in this jelly are found at birth three large blood-vessels, and usually a few degenerated remnants of the epithelium of the allantois. There are no capillaries except close to the navel, and, in spite of the opinion of some writers, it appears safe to say that there are no lymph-vessels,* and no nerves in the distal part of the cord. Schott, 36.1, claims to have followed branches of the hepatic plexus along the vein three or four centi- metres into the cord, and branches of the plexus of the colon and uterus an equal distance along the arteries. Valentin has found nerves even further, 8-11 ctm. from the navel. As Kolliker remarks * Wanderins cells occur in the intercellular spaces of Wharton's jelly, and it is possible that there are lymph channels in the matrix, though no vessels. Compare particularly Koster s paper. 302 THE FCETAL APPENDAGES. in his larger text-book, 79.2, p. 347, the absence of nerves in the distal portion of the cord and in the chorion is of no little physiologi- cal interest, since the blood-vessels are so contractile. In a cross sec- tion, Fig. 209, as usually obtained., the vessels are found contracted, the arteries, AA, with their cavities almost obliterated. The vessels have thick v^ralls composed of a muscular coat and rudimentary intima, but without any special external connective-tissue layer. The vessels differ from adult vessels of similar calibre in many re- spects ; there is no elastic tissue so far as I have observed in any part; the muscle-cells are short, fusiform, loosely arranged, and run in various directions ; next the intima the fibres are longitudi- nal in trend ; in the rest of the coat they are grouped in laminae, which have the fibres obliquely in one direction or an- other, thus giving rise to the appear- coats, noticed by Lawson Tait, 76.1 Fig. 209. — Cross-Section of an Umbi- lical Cord at Term, x about 12 diame- ters. Y, Remnant of the allantois ; F, omphalo-mesaraic vein; A, A, umbili- cal arteries. alternating spiral ance of (p. 434 and Plate XIII., Figs. 17 and 18)'. The muscular coat passes over without any sharp demarcation into the surrounding tissue, known as Wharton's jelly, which consists of scattered anas- tomosing cells, compare Fig. 207, and a muciparous matrix with very numerous connective-tissue fibres. The cells and fibres tend to arrange themselves in concentric lines around the blood-vessels and parallel to the surface of the cord, Fig. 209, so that we may speak of four systems; within each system the cells tend to an elongated form, but where the systems approach one another the cells become more triangular, as seen in section. Fig. 209, and show three or four main processes. These triangular cells form, of course, long columns which are more or less distinct from the tissue encom- passing the vessels; to these columns the name of chordae funiculm has been applied by Hyrtl ; they are said to have been noticed by Woortwyck over a century ago. The external covering of the cord is a stratified epithelium, of which the outer layer is distinctly cor- neous; sometimes there are spaces without cells, which have been regarded as true lymph stomata (Koster and also Tait) ; the mid- dle layer is composed of clear cells, and the basal layer of granular cuboidal cells ; in section the appearances are closely comparable to those of the embryonic epidermis from parts where there are no hairs, and at the time when the horny layer begins to appear. As there is no differentiated connective-tissue layer beneath the epi- thelium, the covering of the cord is best described as embryonic skin. According to current descriptions the cord is said to be cov- ered by the amnion, but this is obviously an error, as shown by His' observations upon the development, and my own upon the histology of the cord. There is usuallj' to be seen in sections of the cord at term, accord- ing toKolliker, 79.2, p. 344, especially in sections from the proximal end and middle region, a small group of epithelioid cells, wi+h dis- THE UMBILICAL CORD. 363 -fcinct walls, irregular granular contents, and rounded nuclei ; around the cells, Fig. 203, there is a slight condensation of the connective tissue to form, as it were, an envelope. This structure has been regarded by some writers as the persistent yolk-stalk, as, for exam- ple, by Ahlfeld {Arch, fiir Gynak., VIII., 363). Kolliker, 79.2, p. 344, considered it to be the remnant of the allantoic cavity — a sup- position which my own observations confirm. CHAPTER XVII. THE PLACENTA. For convenience the placenta may be considered as • an organ hy itself rather than as a derivative of the chorion and of the decidua, which it must be considered from a strictly morphological standpoint. I give as full an account of the human placenta as possible. I. The Human Placenta. Placenta at Full Term. — The human placenta {Mutterkuchen} is a disc of tissue to which the umbilical cord of the child is attached by its distal end. As a result of normal labor the amnion and cho- rion, by which the fcetus in utero is surrounded, are ruptured; the child is then expelled, but by means of the long umbilical cord re- mains attached to the uterus ; after an interval the placenta with which the cord retains its connection is loosened from the uterine wall and expelled, together with the foetal envelopes and portions of the decidual membrane (uterine mucosa) of the mother; the parts thus thrown off secondarily constitute the so-called afterbirth of obstetricians. The placenta at full term, as thus obtained by natural expulsion, is a moist mass, containing a great deal of blood, spongy in texture, about seven inches in diameter, but very variable in size, being roughly proportionate to the bulk of the child ; usually oval, some- times round, but not infrequently irregular in shape. One surface is smooth and covered by a pellucid membrane (the amnion), and reddish-gray in color ;. to this surface the umbilical cord is attached, and it shows the arteries and veins branghing out irregularly from the cord over the surface of the placenta. Fig. 210. The opposite surface is rough, lacerated, and is usually covered irregularly with more or less blood, which is often dark and clotted. When the blood is removed the surface is seen to be crossed by a system of grooves which divide the placental tissue into irregular areas, each perhaps an inch or so in diameter ; these areas are called cotyledons. The placenta is about twenty-five or thirty millimetres thick, but thins out rapidly at the edges, and its tissue passes over without a break into thin foetal membranes, which accordingly spring, as it were, from the margin of the placenta. When in situ the placenta is fastened to the walls of the uterus by its rough or cotyledonary surface ; its smooth, amniotic surface faces the cavity in which the foetus lies. A more detailed examination of the gross appearances of a placenta discharged at term leads to the following additional observations : The color is a reddish or purplish gray, varying in tint according to the condition of the blood, and mottled between the divaricating: THE HUMAN PLACENTA. 365 Hood- vessels by patches and networks of pale yellowish or flesh color. The light pattern is produced by the tissue of the villi shining through the membrane of the chorion. These appearances are less distinct when the placenta, as is usually the case, is covered by the thin amnion. The amnion, however, is very easily detached as far as the insertion of the umbilical cord, but from the latter it cannot be pulled oflf. The blood-vessels run out in all directions from the end of the cord ; each vessel produces a ridge upon the placental surface Fig. 210. — Placenta at full Term, doubly Injected by Dr. H. P. Quincy to show the Distribution of the Vessels upon the Sui'face. so that its course is readily followed. The arteries and veins are more easily distinguished after double injection, as is shown in Fig. 210. The two kinds of vessels do not run together ; the arteries lie nearer the surface, the veins deeper ; the arteries fork repeatedly, until they are represented only by small branches and fine vessels ; some of the small branches disappear quite suddenly by dipping down into the deeper-lying tissue in order to pass into the villi. The veins. Fig. 310, are considerably larger than the arteries ; they branch in a sim- 36(5 THE FCETAL APPENDAGES. ilar manner, but some of the trunks disappear from the surface more abruptly than is the case with the arteries. There is the greatest possible variability in the vessels of the placenta; I have never seen two placentae with vessels alike. So far as I have observed, the in- sertion of the cord is always obviously eccentric ; the degree of eccentricity is variable and is easily seen to be related to the distri- bution of the vessels.' The insertion of bhe cord may even be entirely outside the placenta, which yet, as B. S. Schultze has shown, may otherwise be normally developed. Such insertions are caUed velamentous. The usual type is shown in Fig. 210. The arteries come down together from the cord ; they usually, but not always, anastomose by a short transverse vessel, which lies about half an inch above the surface of the pla- centa; it could not be shown in Fig. 310. I have never noticed any arterial or venous anastomoses on the surface of the placenta. The arteries there spread out in a manner which may be described as roughly symmetrical. The veins par- tially foUows the course of the arteries. When the cord is inserted near the margin the symmetry of the placental vessels is greater ; when the insertion is nearer the centre the symmetry is less than in Fig. 310. The reverse or uterine surface of the placenta is rough and divided into numerous rounded oval or angular portions termed lobes or coty- ledons,* as stated above. These vary from half an inch to an inch and a half in diameter. The whole of this surface consists of a thin, soft, somewhat leathery investment of decidual membrane, which dips down in various parts to form the grooves that separate the cotyledons from each other. This layer is a portion of the decidua serotina, which, as long as the parts are in situ, constitutes the boundary between the placenta and the muscular substance of the uterus, but which at the time of labor becomes split asunder, so that while a portion is carried off along with the placenta and constitutes its external membrane, the rest remains attached to the inner surface of the uterus. If a placenta is cut through it is found to consist of a spongy mass containing a large quantity of blood and bounded by two membranes, each less than a millimetre thick ; the upper one is the chorion covered by the still thinner amnion, and greatly thick- ened where the vessels lie in it ; the lower one is the decidual tissue together with the ends of the villi imbedded in it (cf. especially p. 17 and Fig. 311) ; it represents only a portion of the decidua, the other portion has remained upon the titerine wall. The spongy mass is found upon examination to consist of an immense number of tufts of fine rods of tissue, which are irregularly cylindrical in shape. Further examination shows that they are twigs. Fig. 183, with rounded ends and springing from little branchlets, which in their turn arise from branches, and so on, until a large main stem is found, which starts from the chorion. This branching system is richly supplied with blood from the foetal vessels on the surface of the placenta. The villi are interwoven so that the twigs of one branch are interlaced with those of another, and apparently separate * The division of the placenta into cotyledons is not primary, but, on the contrary , is not de- veloped until the fourth or fifth month. THE HUMAN PLACENTA. 367 twigB may grow together and their vessels anastomose ; but on this point I am unable to speak positively. The villous twigs next the surface of the decidua penetrate that tissue a slight distance. The intervillous spaces are filled, or nearly so, with blood ; they form a complex system of channels. The intervillous blood, as we know from the researches of Farre, Turner, and Waldeyer, is ma- ternal. Farresays, in his article in Todd's "Cyclopaedia," V. Suppl., p. 716, in reference to the placental decidua: " Numerous valve-like apertures are observed upon all parts of the surface ; they are the Fig Sll.— Placenta at Full Term. A, Vertical section through the margin; D, decidua: m, aborted villi outside the placenta; Cho, chorion; Ni. sinus; Vi, placental villi; Mb, fibrin; B, portion of A, more magnitted to show the decidual tissue near b; w, blood-vessel ; d,d', decidual cells ; d, with one nucleus ; d', with several nuclei. orifices of the veins, which have been torn off from the uterus. _ A probe passed into any one of these, after taking an oblique direction, enters at once into the placental substance. Small arteries about half an inch in length are also everywhere observed imbedded in this layer. After making several sharp spiral turns they likewise suddenly open into the placenta;" and on p. 719 he adds: "These venous orifices occupy three situations. The first and most numer- ous are scattered over the inner side of the general layer of decidua which constitutes the upper boundary of the placenta ; the_ second form openings upon the sides of the decidual prolongations or 368 THE FCETAL APPENDAGES. dissepiments which separate the lobes (cotyledons) from each other; while the third lead directly into the interrupted channel in the margin, termed the circular sinus." The circular sinus (Fig. 311) is merely a space at the edge of the placenta which is left com- paratively free from the villi. It is not a continuous channel, but is interrupted here and there. Susbequent writers have gone but little beyond Farre's account, which has been entirely overlooked by most recent German investigators, who accordingly announce facts kno_wn to Farre as new discoveries. Under these circumstances it seems" no more than just to direct renewed attention to Farre's masterly article. To study the histology of the placenta sections are best made after imbedding the organ in celloidin. Fig. 211 represents parts of a section of a placenta at term from which the amnion was removed. Fig. 211, A, represents the placental margin magnified thirteen diameters ; B, a portion of the decidua near b in A, but more highly magnified. The chorion, Cho, and decidua, D, are in immediate contact at the left of the figure, that is, outside the placenta, though remnants of a few aborted villi, in, are still plainly recognizable; but they are found only close about the placenta. At the margin of the placenta and in its neighborhood the chorion and decidua are not clearly delimited, but, on the contrary, the decidual cells find here an opportunity to penetrate the chorionic membrane. The placental chorion exhibits its characteristic stratification a short distance within the margin. I have found, however, that the distinctness of that stratification varies considerably, not only in different placentae, but also in different parts of the same placenta. The decidua, D, outsidethe placenta is 1 ^ I 1 ;» Cj very thick, but at the edge it begins to thin out, and, as it passes over the under side of the placenta, rap- idly becomes so much reduced as to be even less in thickness than the chorion, Cho. The decidua is char- acterized by an im- mense number of large decidual cells, not scattered about as in Fig. 10, but densely packed. Fig. 211, B, the cells are irregularly oval in outline, clear, or somewhat granular, and have usually a single nucleus ; a few are larger, more granular and multinucleate. At the edge of the placenta the chorion and decidua separate; where they first part there are very few villi, Fig. 211, Si, but else- where the room between them is occupied by innumerable branches of Fig. S12.— Mesenchymal Tissue of a Villus, from a Placenta of four Months. ? ?, Leucocytes; v v^ capillary blood-vessels; d, finer mesh-work from near a capillary. THE HUMAN PLACENTA. 369 villi, Vi, Vi, with narrow spaces between for the blood ; the sections of the villi are of all sizes and shapes ; they all contain blood-vessels, but only the larger ones can be distinguished with the magnification of Fig. 211, A, where they have been made as distinct as possible by being drawn black. The spaces between the villi have been left white, the blood which partially filled them not being represented. Placenta in Situ. — The placenta in its natural position in the uterus follows the curvature of the uterine walls, hence its free or amniotic surface is slightly concave, its decidual surface is strongly convex ; it is thickest in its centre and thins out gradually toward its edge. There is no definite boundary between the portion of the decidua serotina which is to be torn off with the placenta, and the part which is to remain in the uterus after delivery. Vertical sections through the uterus with the placenta in place are very instructive. Fig. 313 represents such a section through a pla- centa of about seven months. The thin amnion. Am, clothes the inner surface of the chorionic membrane, Cho; this membrane is separated from the decidua, D, by a dense forest of villi; in the younger specimens the distance between the chorion is considerably less than the thickness of the uterine wall, D, Mc, but in the pres- ent specimen. Fig. 213, it is much greater; in younger stages the villi are much less numerous, and much smaller than in the older one ; these differences correspond to the growth of the placenta and to the changes in shape of the chorionic viUi already described, p. 319. The ends of some of the villi touch and are imbedded in the decidual tissue; these imbedded ends are without covering epithe- lium, but their connective tissue is immediately surrounded by hyaline substance, which is probably the degenerated epithelium. The de- cidua is plainly divided into an upper compact, and a lower cavernous layer, see p. 8. The section passes through a wide arterial ves- sel, Ve. Foetal Circulation of the Placenta. — The following para- graph refers to the placenta during the later months of pregnancy; it is copied almost without change from my article on the placenta in Buck's "Eef. Handb. Med. Sci.," V., 696-697. To follow the course of the foetal blood-vessels within the placenta, the best method is by corrosion injections. These may be made either with fusible metal, wax, or celloidin. The first is specially suited for the study of the large trunks ; the latter for that of the smaller vessels also. I have a very beautiful celloidin injection by Dr. S. J. Mixter, which, with others of wax and metals, has served as the basis of the following description : The veins leave the surface somewhat more abruptly than do the arteries, which gives off more small branches to the surface than do the veins. Fig. 210. Both kinds of vessels leave the surface by curving downward for a short distance into the trunk of a villus; the vessels then divide, and their branches again take a more horizontal course; the branches then curl over downward, and after a second short descent toward the decidua, again send out horizontal branches. The result of this arrangement is a terrace-like appearance in the course of the vessels ; they approach the uterine side of the placenta in this very character- istic manner. The number of terraces is variable ; usually there are 34 370 THE FCETAL APPENDAGES. Fie 213. -Section through a normal Placenta of about seven Months tn situ. Am, Amnion, Clw, chorion; Vi, villus trunk; vi, sections of villi in the substance of the Pla«enta,; a deciaua. iMc, musoularis; Xl', compact layer of deoidua; Ve, uterine blood-vessels (or fl??*?) opening inti the placentk The f Setal blood-vessels are drawn b ack ; the maternal blood sPfMS are left white; tte chorionic tissue is stippled except the canalized fibrin, which is shaded by lines ; the remnants of the gland cavities in D" are stippled dark. (Drawn from nature by J . H. Emerton. ) THE HUMAN PLACENTA. 371 two or three, but sometimes there is only one, or they may number four or even five. Arrived at the end of its terraces the main vessel takes a more nearly perpendicular course, and rapidly subdivides. Immediately after entering the villi, the arteries and veins give off but few capillaries, but after a short course in the main stalk of the villus the vessels give rise to many branchlets, and gradually the character of the circulation changes until in the smallest villous twigs there are capillaries only. Fig. 214. The vascular trunks have a marked tendency to dichotomous divi- sion which is maintained within the villi to a certain extent ; the arterioles and vein- lets in the mature placenta go from their trunks at wide angles for the most part, and subdivide in the same manner, so that they spread out through the whole sub- stance of the placenta. The vessels next FiQ. 214.— Portion of an iniected Villus from a Placenta of about five Months; magnified 210 diams. Fig. 215. — Placenta of about five Months ; Portion of a small Villus to show the Central Vessels and Su- perficial Capillaries. X 105 diams. the decidua take a more horizontal trend, like the top branches of a wind-swept tree. As the vessels run in the villi, of course the way in which the latter branch out determine the paths of the former ; hence, by following the distribution of the vessels we inform ourselves as to the ramifications of the villi. Thus the horizontal course of the vessels on the uterine side of the placenta corresponds to the well- known fact that the ends of the villi attached to the uterus become bent and adhere by their sides to the decidual surface. The capillaries of the villi are remarkable for their large size, and on this account have been described as arteries or veins by E. H. Weber, Groodsir, and other writers. Their calibre is often sufificient for from four to six blood-discs abreast. They are very variable in diameter, and also peculiar in exhibiting sudden constrictions and dilatations. Fig. 314. In the short bud-like branches there is often only a single capillary loop, but as the branch becomes larger the 372 THE PCETAL APPENDAGES. number of loops increases, and they form anastomoses. In branches large enough to serve as a stem, some one or two of the vessels may be enlarged, as may be seen in Fig. 314 ; in the branches large enough to admit of it, there are two (or sometimes only one) longitudinal central vessels, an artery and vein, and a superficial network of capillaries, Fig. 215. Goodsir and other early writers laid great stress on the formation of the capillary loops, but this feature is a common one in the development of the foetal vascular system, as is also the width of the capillaries. In my opinion these peculiarities are characteristic rather of the foetus than specifically of the placenta. In some of the older writers (Goodsir, Farre, et al.) it is asserted that the true capillary systems disappear toward the end of gesta- tion. I am unable to confirm this, but find instead that in the slen- der terminal villi of the placenta at term there is often only a single, sometimes long, capillary loop ; the capillary is very wide, and its width is probably the reason of its having been held formerly to be a vein or an artery. Maternal Circulation of the Placenta. — ^The course of the maternal blood in the placenta has been the subject of nearly con- stant debate for a -century past, and the problem has received its final answer only within the last few years. The discovery of the facts belongs to so many authors that it seems not worth while to attempt to cite the authorities for each detail, accordingly I give a summary of what is known, and in an historical note refer to the principal investigations. The arteries and veins both open upon the surface of the decidual serotina, at least during the later half of pregnancy ; concerning the circulation during the first half of pregnancy we possess no positive information, although the fundamental arrangements are presuma- bly the same. The blood, which is poured out from the arteries, circulates in the intervillous spaces, which act as maternal blood channels. Both arteries and veins change the character of their walls as they approach the surface of the decidua ; when they enter the decidua they are nearly or quite without muscular walls, and can, therefore, be recognized as arteries or veins, not by their histological structure, but only by their size and their continuity with undoubted arteries and veins in the muscularis ; during their passage through the de- cidua their walls gradually become reduced to the endothelial layer ; but the arteries have, what the veins do not have, a thin clear layer just outside the endothelium ; this layer colors readily with carmine, contains a few scattered nuclei, and is probably the result of degen- eration; it ceases before the artery actually reaches the surface. The endothelial nuclei of the veins project distinctly into the lumen of the vessel. Waldeyer, 90.1, 33, summarizes the differences be- tween the arteries and veins as follows: The arfeHes are smaller; they take a spiral course and run within special columns of fibrous connective tissue; they make numerous turns within the decidua, and lie in the broad ridges of the membrane ; they usually do not branch but terminate with a single opening, which generally lies in the upper or lateral part of a decidual ridge ; the opening is narrow and the villi do not project into it at all or taut slightly ; the terminal THE HUMAN PLACENTA. 373 piece of the artery is round in cross-section ; the artery in the decidua has a special layer outside the endothelium, to within a short dis- tance of the opening. The vems are, generally speaking, wider; they have no special sheaths, and do not run in spiral, but in direct courses, more or less parallel to the surface ; their openings lie, for the most part, between the ridges (septa) and never at the summits of the ridges; from the border vein (Grenzvene, Waldeyer) run out terminal branches which open on the surface and are usually numer- ous ; the chorionic villi project into the mouths of the veins and reach down even into the "Grenzvenen;" the mouths of the veins are irregularly shaped, and the veins themselves are irregular in cross-section, never circular. The position of the vascular openings is such that the arterial blood flows out from the septa, while the venous blood flows off through the surface between the septa ; hence, as pointed out by Bumm, 90. 1 , each cotyledon represents a more or less distinct circulatory region, the blood entering at the sides and leaving at the bottom. Historical Note. — ^The long prevalent erroneous view that there is a direct communication between the maternal and foetal circula- tions originated I believe with Haller ("Elementa Physiologiae," VIII.). It was revived again by Flourens, 36. 1, and though long since entirely disproved is still encountered from time to time. The first important evidence of the circulation of the blood in the inter- villous spaces was brought by F<. H. Weber, whose investigations were published in Hildebrandt's " Handbuch der Anatomie des Men- schen," 4te Auflage, IV., 490. Weber's doctrine was adopted by most subsequent investigators. The most important additions to his observations were made by Farre, 58.1, and Turner, 73.1, 76.1, 76.3, 77. 1, 77.2, 88. 1, imtil we come to the recent researches of Langhans, 77.1, 82.1, and his pupils, Mtabuch, 87.1, Eohr, 89.1, etc.; of Waldeyer, 87.1, 90.1, of Bumm, 90.1, Minot, 98, Bloch, 89.1, and others, which have finally settled the problem. That the intervillous spaces normally contain blood was seriously questioned by Braxton Hicks, 72.1, whose doubts were again brought prominently forward by C. Euge, 86.1. Ruge's position I was inclined at first to adopt (see Minot, Anat. Anzeiger, II., 19), but I have since become entirely convinced of the correctness of Weber's doctrine as established by Langhans, Waldeyer, etc. A thorough and very valuable critical review of the whole subject is given by Waldeyer, 90.1, upon whose citations this note is based, but I have referred only to a few of the numerous authorities quoted by Waldeyer. Nutrition of the Foetus. — The mechanism of the transfer of nourishment from the uterus to the child is not well understood. It is evident that the supply must come from the maternal blood and reach the foetus through the veins of the umbilical cord ; although the amniotic fluid may be a source of supply, as some have main- tained, yet at most its role can be only secondary and the main transfer of material must take place through the placenta. Our pres- ent knowledge of the structure of the organ renders it unnecessary to discuss the old theory recently revived by Currie, of a direct com- munication between the maternal and fcetal vessels, for we know 374 THE FCETAL APPENDAGES. positively that no such communication exists. This theory has been put forward again with the modification that the vascular walls will let small solid particles through. Thus Koubassoff, on the basis of some inconclusive experiments, sought to maintain that microbes, and ergo other solid particles, could pass from mother to embryo (see Comptes Rendus Acad. Paris, t. CI., 508-510). More care- ful tests by Marie Miropolsky failed to confirm this (Arch, de Physiol., n. etp., 1885, 101-108). A second theory, at pres- ent the best accredited, is that of diffusion, which finds its chief basis in the elaborate arrangements found in all placental types for bringing the foetal and maternal blood into immediate proximity. A third theory is that Eauber, 79, who attributes the chief role in the nutrition of the embryo to the immigration, by way of the pla- centa, of maternal leucocytes. A fourth theory attributes an active part to the utricular glands, which are supposed to pour out a nu- trient secretion into the intervillous spaces, where it is taken up by the chorionic villi. It is impossible at present to form a final judg- ment upon these theories. As we have seen, the intervillous spaces are probably maternal blood-channels at all periods, so that, from a very early stage on, the conditions for the transfer of material, either by a migration of leucocytes or by simple transfusion, are established. Rauber's leucocyte theory has not commended itself to me, and I incline to accept the transfusion theory. That the uterine milk exists in man has not been proven, and the occurrence of such a secretion is not compatible with the degeneration of the glandular epitheliiun observed by Minot, see p. 10. II. Theory op the Placenta.* Attachment of the Embryo. — That the rabbit embryo is at- tached to the surface of the uterus by a thickened region (area pla- centalis) of the ectoderm of the germinative area was first shown by Van Beneden and Julin, 84. 1 ; this discovery has since been con- firmed by Minot, 98, Masius, 89.1, Duval, 89.1, and others. That a similar method of attachment exists in other mammals has been shown by Strahl, 89. 1, 4, 90. 1 ; in the dog it has been recorded by G. Heinricius, 89.1. In aU these cases the thickened ectoderm is found to be closely adherent to the uterine surface, upon which it is apt to remain when the ovum is forcibly removed ; it fits exactly to the surface of the maternal epithelium ; there is no visible layer of cement, and we do not know how the adherence is made so close. It is probable that we have here the primitive form of attachment, and that therefore the evolution of the placenta began with the dif- ferentiation of the ectoderm of the area placentalis. There is another type of attachment found in the hedgehog and in rodent ova with inversion of the germ-layers, characterized by the ovum being so closely invested by the uterine mucosa that the whole surface of the ovum comes in contact with the maternal tis- ues (see E. Selenka, 84. 1, and Hubrecht's superb monograph on the placenta of the hedgehog, 89. 1). * Reference is made especially to the true chorionic placenta. THEORY OF THE PLACENTA. 375 Degeneration of Uterine Tissues. — Over the region of the placenta] attachment, which varies in different animals as to position, there occurs an extensive degeneration of the tissues of the uterine mucosa, affecting hoth the covering epithelium, the glands, and the connective tissue. The degeneration takes place most rapidly in the epithelium and glands, while the connective tissue undergoes an exten- sive hypertrophic metamorphosis, usually in the form of the develop- ment of decidual cells, before the degenerative change acquires the upper hand. The nature and extent of the degenerative changes have become known for various types by investigations published since 1888, several of which appeared during 1889 (Minot, 89, 98, Masius, 89.1, Heinricius, 89.1, Duval, 89.1, Hubrecht, 89.1, Strahl, 89.1, 4, etc.), and represent simultaneous and independent re- searches. In view of what we now know it must be considered probable that in all placental mammals, or at least in the orders of the unguiculate series, the uterine degeneration is an invariable factor in the development oJE the placenta. The form of degeneration is not fixed, but yaries greatly. This is illustrated by the history of the decidua in man and in the rabbit. Other modifications occur in the dog, the hedgehog, the mole, and doubtless in other animals. The result of the degeneration is : first, to bring the chorionic ectoderm of the embryo into direct contact with the connective tissue of the mucosa uteri in consequence of the degeneration and resorp- tion of the epithelium, including the glands; second, to allow the maternal vessels by simple expansion to come into contact with the foetal chorion. In the rodents the degeneration goes so far in the neighborhood of the chorion that aU (or nearly all) the maternal tis- sue disappears, leaving the maternal blood to bathe the surface of the chorion, or, to speak more exactly, of the chorionic villi. It is probable that similar changes take place in man, but in the earliest stages yet studied- they appear to have been already completed, so that in the region of the viUi the maternal tissues have completely disappeared, unless the endothelial layer described by Keibel be ma- ternal, V.S., p. 323. Heinricius has maintained that in the dog part of the glandular epithelium remains. Outgrowth of Chorionic Villi. — These are restricted at first to the small placental area, but as that area may itself grow and take up more and more of the chorion, we get various modifications of the villous area. The more primitive types preserve the discoidal plan, illustrated by the rabbit ; in other cases the placental or villous area expands until it forms a belt or zone around the ovum (carniv- ora) ; but the development in the dog shows that the discoidal form is the earlier, and changes into the zonary ; in man the placental area spreads over the whole chorion. The villi appear to arise as outgrowths of the ectoderm only ; after the outgrowths have attained a certain size the mesoderm of the chorion grows into them. The villi grow into the maternal tissues, and acquire great length and numerous branches, by which their form- becomes extremely complicated. Their form is highly char- acteristic of the various orders ; it is known exactly only in man, but is certainly very different in various animals. 376 THE FCETAL APPENDAGES. The villi occupy only a part of the mucosa, there being always a considerable layer of decidual membrane left between the end of the villi and the muscularis. The villi, as here described, consist of a core of mesoderm covered by fcetal ectoderm, and are essentially different from the ectodermal outgrowths assumed by Duval * to exist in the rabbit. TJnion of the AUantois with the Chorion. — We know two principal modifications of the union of the allantois with the cho- rion : 1. The allantois joins the chorion early, and serves as the stalk, connecting the embryo with chorion ; in this type the allantois brings the blood-vessels to the chorion and the vessels then ramify over the chorion itself, which has therefore its own circulation, though de- pendent upon the allantois ; this modification is characteristic of the unguiculate series of mammals. 2. The allantois grows out into a large vesicle, which has for some time no connection with the chorion but maintains a well-developed circulation of its own ; its expansion brings it ultimately into contact with the chorion, and its outer or mesodermic layer grows together with the inner or mesodermic layer of the chorion (Bonnet, 89. 1) which thus becomes indirectly vascu- larized ; this modification is characteristic of the ungulate series of mammals. How far other modifications, distinct from these, may exist, we cannot say at present. We have then two types: 1, the chorion has its own vessels (un- guiculates) ; 2, the chorion acquires vessels by growing together with the vascular walls of the allantoic vesicle (ungulates) . In both cases th'e chorion is the part of the foetus and the only part in direct contact with the uterine wall, and therefore in both cases it is the essential part of the foetal placenta. In unguiculates the chorion, after it receives its blood-vessels, has its own blood sup- ply and circulation, and therefore suflfices to develop the placenta. In ungulates the circulation is entirely allantoic, and the walls of the allantois are essential to maintain the fcetal circulation of the placenta ; the chorion, therefore, does not sufl&ce to develop the fce- tal placenta. While we recognize that the chorion is always the means of union between the mother and the offspring, we may con- veniently distinguish the unguiculate type as having a true cho- rionic placenta, and the ungulate type as having an allantoic pla- centa. Evolution of the Placenta.^ — As regards the evolution of the placenta, our conceptions are still very obscure. The opinion was long, and perhaps still is, generally prevalent that the placenta is primarily an organ of the allantois. This notion was one of those theories which sometimes become current without ever having been supported by adequate proof, and are repeated until tradition has rendered them venerable and age gives them a dignity their worth does not entitle them to. The principal support of this theory was de- rived from the fact that the allantois is connected with the placental circulation. Balfour in 1881 ("Works," I., 743) sought to modify this view by attributing importance to the relations of the yolk-sac, which he believed to be the means of maintaining the circulation. * Erroneously, as I believe. THEORY OP THE PLACENTA. 37? In his "Comparative Embryology," II., 349, Balfour reprints most of the article cited. Minot, 98, 433, laid stress upon the role of the chorion and upon the fact that the placenta is necessarily always a product of the chorion, and further upon the fact that the allantois in man is permanently (and in the rabbit primarily) merely a stalk of connection between the embryo and the chorion. The investigations mentioned in this chapter which have been recently published seem to me to greatly strengthen my view. It is by the chorion that the ovum is attached, except in certain rodents in which the development has obviously been modified. It is from the chorion that the foetal villi grow out. On the other hand, it is evident that the yolk-sac is primitively a product of the splanchnopleure and distinct from the somatopleuric chorion; the failure of the mesoderm and coelom to spread completely over the yolk (entoderm of the blastodermic vesicle) in certain mammals does not alter the fundamental relations. It is true that in certain marsupials the chorion is very imperfectly sep- arated from the yolk-sac, but it does not appear that this represents an ancestral stage of the mammalia ; on the contrary, it is probably a purely marsupial modification. I am therefore unable to recog- nize any reason for connecting the evolution of the placenta with the yolk-sac or vitelline circulation. The role of the aUantois is second- ary ; it serves as a medium of blood supply, either, as we have seen, as a carrier of vascular trunks to supply the circulation of the cho- rion (unguiculates) or bringing its own circulation into play by growing together with a non-vascular chorion. The question remains whether the unguiculate or the ungulate type of placenta is to be regarded as the more primitive. At first thought the resemblance of the foetal envelopes of ungulates to those of Sauropsida leads us to conclude that the allantoic placenta must be the more primitive ; the resemblance referred to consists in the early complete separation of the chorion (serosa) from the other parts and in the development of the allantois as a large free vesicle. But the ungulates are highly modified mammals not related closely to the lower placentalia, while the unguiculates do merge into a generalized mammalian type. When we consider further that the lower un- guiculates show the typical chorionic placenta in its full perfection, the conclusion is unavoidable that this is the nearer type to the ancestral. In fact, the placenta appears in animals with the chori- onic type of the organ before the allantois becomes free, and the great size of the allantoic blood-vessels is connected primitively, not with the allantois, but with the already important chorionic circula- tion; the placenta is here interpolated in the ontogeny before the specialization of the allantois, which functions as the vascular path- way between the chorion and embryo, both primitively and perma- nently. The enlargement of the allantois in ungulate mammals is a supervening change, effected perhaps by an atavistic recurrence to reptilian ontogeny. Ryder, 87.6, has advanced the theory that the zonary placenta is older than the discoidal, but Minot, 98, 434, has shown that this view is untenable. The degenerative changes in the uterus occur, so far at present known, only in connection with the chorionic placenta ; in the un- 378 THE POETAL APPENDAGES. gulates the uterine mucosa is modified in structure in connection with the development of the placenta, but the modifications are not known to be degenerative ; hence in the allantoic placenta the ma- ternal blood fiows in maternal blood-vessels, and it is always sepa- rated by maternal connective tissue and epithelium from the chorion. Theory of the Placenta. — -According to the views explained in the preceding pages, I hold the placenta to be an organ of the chorion ; that primitively the chorion had its own circulation, and formed the discoidal placenta by developing villi which grew down into the degenerating uterine mucosa ; by the degeneration of the maternal tissues the maternal blood is brought closer to the villi, and the degeneration may go so far that all the tissue of the uterus be- tween the villi disappears ; a layer of the mucosa is preserved between the ends of the villi and the muscularis uteri to form the so-called decidua ; the placenta receives its foetal blood by the means of large vessels running in the mesoderm of the allantois. From this dis- coidal chorionic placenta the zonary placenta of carnivora, the diffuse placenta of the lower primates, and the metadiscoidal placenta of man have been evolved. A second type of placenta, perhaps evolved from the first, is found in ungulates, and is characterized by a vascular allantoic vesicle uniting with a non- vascular chorion to form the fcetal placenta, and by the absence of degeneration in the maternal tissue. This type is the allantoic placenta, which offers many interesting modifications. PART V. THE FCETUS. CHAPTER XVIII. GROWTH AND EXTERNAL DEVELOPMENT OF THE HUMAN EMBRYO AND FCETUS. The two sections following on the growth of the foetus and the weight at birth are taken from my article on " Growth " in Buck's "Reference Handb. Med. Sci.," III., 394. A more accurate concep- tion of the growth of the embryo can, however, be gathered from the figures in the latter part of this chapter. Growth of the Pcetus.— The difficulty of determining the age of the human foetus and of obtaining specimens certainly fresh and normal has prevented our having any definite information on this subject. Preyer has compiled the following table of the length of the human embryo in centimetres : Lunar Month. Toldt. Hennig. Hecker. (300 obs.) (100 obs.) First, 1.5 (1.3) 0.75 Second, 3.5 .4.0 Third, 7.0 8.4 4 to 9 Fourth, 12.0 16.2 10 to 17 Fifth, 20.0 37.5 18 to 37 Sixth, 30.0 35.25 38 to 34 Seventh, 35.0 40.35 35 to 38 Eighth, 40.0 44.3 39 to 41 Ninth, 45.0 47.3 43 to 44 Tenth, 50.0 (49.0) 45 to 47 If the absolute length at they end of each month is divided by the increase during that month we obtain what Preyer calls the relative growth. Hennig's figures give the following relative growth for each month: First, 1,000; second, 0.812; third, 0.523; fourth, 0.419; fifth, 0.410; sixth, 0.219; seventh, 0.124; eighth, 0.093; ninth, 0.069; tenth, 0.037. All the above data are obviously inexact. Toldt's are evidently cooked up and not derived from observation, nor do the lengths mean the same thing, for of the early stages the head and trunk only were measured ; of the later stages the head, trunk, and legs. A falser and more misleading device for studying growth has never been put in practice. The foetus, too, being spirally coiled in early stages cannot have its length determined accurately. Far better would it be to always determine the weight. The growth of the foetus in weight has been most inadequately stud- ied, although the weight is the only available measure of the growth of the foetus as a whole. Hecker's data are perhaps the best. The weights are in grammes ; 383 THE FCETTJS. Month. Maximum. Minimum. Third, 30 5 Fourth, 120 10 Fifth, 500 75 Sixth, 1,380 375 Seventh, 3,350 780 Eighth, 3,438 1,093 Ninth, 3,906 1,500 Tenth, 1.563 Average. 11 57 384 634 1,218 1,569 1,971 The range of the maxima and the minima suggests that errors in the determination of the ages may have occurred — such errors of a month are not rare with obstetricians. Appended here are Hecker's data as to the weight of the placenta in grammes, and the length of the umbilical cord in centimetres : Month. No. of obs. Placenta. Cord. Third, 3 36 7 Fourth, 17 80 19 Fifth, 34 178 31 Sixth, 14 273 37 Seventh, 19 374 43 Eighth, 33 451 46 Ninth, 45 461 47 Tenth, 62 481 51 2. Weight of the New-Born Child. — It is subject to very consider- able variations. For middle Europe the average may be held to be about 3,340 grammes for boys, 3,190 for girls, the latter being some- what lighter. The variations are very great, ranging from 1,000 to 5, 000 grammes. For instance, the following table is given by Pfann- kuch, who unfortunately jumbles the two sexes together : Kilos, 1.50 to 2.00 3.00 to 3.35 3.25 to 2.50 2. 50 to 3.75 3.75 to 3.00 Obs. 33 36 52 90 110 Kilos. 3.00 to 3.25 3.35 to 3.50 3 50 to 3.75 3.75 to 4.00 4.00 to 4.50 Obs. 150 115 79 46 13 It will be noticed that the maximijm number of cases (150) falls between 3.00 and 3.25 kilos., and that the further the weight is removed on either side, above or below from this mean, the fewer are the cases. The tables by other authors show the same general results, with usually slight differences in the quantitative values. For the most part these tables cannot be combined with one another, for they nearly all fail to fulfil some obvious requirement of good statistics ; indeed, amateur statistics are generally provoking to the expert. It is, therefore, not desirable to attempt an analysis of the recorded data. As an example of statistics at once valuable and grossly defective, the following table is given after Siebold. The author gives the weights in pounds, but has neglected to say, as is necessary in Germany, what kind of pounds, hence the metric equiva- lents cannot be calculated. Moreover the number of cases weighing even pounds and half-pounds is far in excess of those weighing pounds and one-fourth or three-fourths, which shows inaccurate weighing, of course. To correct this the quarter-pound groups of original table are condensed into half-pound groups : GROWTH AND EXTERNAL DEVELOPMENT. 383 Weight in lbs. Boys. Girls. Weight in lbs. Boys. Girls. 40 to 4.5 4 10 7.5 to 8.0 286 200 4.5 to 5.0 19 34 8.0 to 8.5 101 44 5.0 to 5.5 44 53 8.5 to 9.0 79 42 5.5 to 6.0 172 195 9. to 9.5 15 14 6.0 to .6. 5 . 230 235 9.5 to 10.0 7 3 6.5 to 7.0 353 353 10.0 to 10.5 1 7.0 to 7.5 386 240 10. 5 to 11.0 1 The extremes recorded in medical literature are very far apart, and statements of excessively large size are by no means rare, but can be received with incredulity only, as, for instance, the case re- ported of a still-born child weighing 8,250 gms. {Berlin, klin. Wochenschr., 1878, No. 14) ! Vierordt gives as the accredited extremes 717 gms. (Riter), and 6,123 gms. (Wright.) The factors which determine the weight at birth are very obscure. It is, of course, safe to say vaguely that it depends on the nutrition of the foetus ; it is probable that individual differences in the rate of growth exist before as well as after birth, and it is probable that the length of gestation is the most influential single factor, to judge from my own experiments on the growth of mammals. It has been demonstrated that the variations in the weight of the child depend upon various maternal circumstances. First. It is correlated with the age of the mother, as is shown in the following table, giving the weight of the children in grammes according to three observers : Age of mother. Ingersley. Fassbender. Petersson. 15 to 19 years 3,241 3,271 3,451 20 to 24 " 3,299 3,240 3,485 35 to 29 " 3,343 3,333 3,591 30 to 34 " 3,375 3,367 4,062 35 to 39 " 3,438 ] 3,292 ■ 3,591 40 to 44 " 3,326 3,676 From such tables we learn that very young mothers have the smallest children, and those of about thirty-five years the heaviest. It is much to be regretted that the tables do not show the correlation by single years and also the number of observations. Second. The weight of the child increases with the weight (Gass- ner) and length (Frankenhauser) of the mother. Gassner states that the weight of the child is to that of the mother as 1 to 19.13, or 5.23 per cent of the maternal weight. Frankenhauser states that if the height of the mother is less than 4 feet 6 inches the child weighs 6 lb. 15 oz. ; if it is 4 feet 6 inches to 4 feet 11 inches, the child weighs 6 lbs. 25 oz. ; if it is more than 4 feet 11 inches, the child weighs 7 lb. 3 oz. Third. The weight of the child increases according to the number of previous pregnancies, as indicated by the following table : Number of pregnancies. One, Two, Three, Four, Five, Six, Hecker. Ingersley. CGrms.) CGrms.) 3,201 3,254 3,330 3,391 3,358 3,400 3,360 3,424 3,412 3,500 3,353 384 THE FOETUS. Here again we encounter faulty statistics, for it is not shown that we have any other effect than that of age, for the conclusion claimed cannot be established until it is proved that primiparae have smaller children than multiparas of the same age. Fourth. Negri has maintained {Annali di Obstetrica, 1885) that the compilation of three hundred and thirty -three observations show that the children of women whose first menstruation is early are larger than the children of those whose first menstruation is late. Fifth and Sixth. The influence of race and climate, which have not yet been subjected to any proper exact study. In conclusion I may add that it seems to me probable that all these influences produce their effect principally by prolonging or abbrevi- ating the period of gestation. In other words, the variations in the weight of children at birth are to be referred immediately to two principal causes: 1, Differences in the age at birth; 2, individual differences of the rate of growth in utero. Measuring the Length, of Embryos. — Owing to many changes in the curvatures of the longitudinal axis of the human em- bryo it is impracticable to employ any one system of measurement, to obtain comparable results for all ages. On this account I have adopted the system of giving in all cases the greatest length along a straight line, the embryo being measured in its natural attitude — excluding, however, the limbs from the measurement. His adopts for embryos of four to ten weeks what he calls the Nackenldnge ("Anat. menschl. Embryonen," Heft II., 5) or the distance in a straight line from the neck bend to the caudal bend, but as this cannot be measured accurately in later stages I have thought it best to give up this measure. Hence it results that the length of an embryo as given by His is often different from that given in this work. Embryos of Known Ages.— As already pointed out, we have to reckon from the last day of the menstrual period as the date of conception, but this date is never quite certain, hence there is always some doubt as to the age of every embryo. We owe to Pro- fessor His most of our information in regard to the form and size of the embryo at successive ages during the first two months, see his " Anatomie menschl. Embryonen," Heft II., 1882, especially pp. 25 and 72, also Heft I., 166, Heft III., 236-254, and Taf. X., which gives figures at a uniform scale of twenty-five embryos of the first two months. The development of the embryo during the first three weeks has already been described and illustrated. Up to the end of the ninth week the form and size of the embryo undergo a correlated develop- ment, so that generally an embryo, at a given stage of development in/orm, wiU agree with its fellows in size; but to this rule there are not infrequent exceptions, and sometimes an embryo is found much larger than others at the same stage (His, I.e., Heft III., 240). Moreover the variability of embryos is very great, for in specimens otherwise alike we find this or that organ retarded or advanced in development, as compared with the embryo as a whole. Neverthe- less it is possible with the information at present at command to determine the age of an embryo within two days plus or minus up GROWTH AND EXTERNAL DEVELOPMENT. 385 to the end of the ninth week. For the development during the third month we possess as yet no satisfactory information, but embryos three months old are quite frequently obtained, and my own collec- tion gives a good series of specimens up to the end of the fifth month. Twenty-three Days.— The first figure I give is that of His' em- bryo a. Fig. 316, described by him in his " Anat. menschl. Embry- onen," Heft I., 100-115. The specimen was from a chorionic vesicle measuring 3.5 by 3.0 cm. ; the greatest length of the embryo was 4 mm., measured from the end of the hind-brain, iv, to the four- teenth segment of the rump. It lay with its left side against the chorion, with which it was connected by a short allantoic stalk ; the yolk-sac measured 3.7 by 3.0 mm., and had a short pedicle. His states that the probable age of the specimen was twenty-three da>-s, and apparently bases the deter- , mination upon comparison with slightly younger and older spe- cimens of known age. The shape of the embryo is very dif- ferent from that of Fig. 180, p. 307, owing to the whole body having become rolled up, so that the dorsal outline describes more than a complete circle ; the body has a marked spiral twist, the head being bent to the right, the tail to the left ; the bending of the body is specially marked at the region of the mid-brain (head-bend) and at the posterior limit of the hind-brain (neck- bend, Nackenkriimmiing) . The primitive segments show externally ; the anlages of both pairs of limbs have appeared as outgrowths of the so-called "Wolflfian ridge, but the leg is less developed than the arm. In the region of the head the divisions of the cerebral vesicles can be recognized. The optic vesicles are indicated by small protuberances. The oval oto- cyst lies about at the level of the second gill-cleft. The cephalic border of the mouth has become ridge-like; the dorsal end of the ridge joins the dorsal end of first visceral arch, which is known as the mandibular arch; the ridge on the cephalic side is known as the maxillary process. The second, third, and fourth gill-arches are distinct and behind each the imperforate gill-cleft can be distin- guished, but the fifth arch is indistinct. The heart forms a marked protuberance, the bulbus aortse showing most on the left side. Fig. 316, and the ventricles on the right. An embryo of C. Rabl's, very similar to His' a, just described, is figured by O. Hertwig, "Entwickelungsges.," 3te Aufi., Fig. 137. His points out that the embryos numbered by him XXVI., (D 3), LVI., (W), and LVII., (R), are very near the one just described, though a little older. In the same group belongs the embryo of Coste, 47.1, PI. II., Fig. 5, of which the age was determined at twenty to twenty-five days ; also Thompson's fourth embryo, figured Fig, 816. —His' Embryo a. Age i three Days, X about 10 < twenty- 386 THE PCETUS. and described by Kolliker, "Entwickelungsges.," p. 311, Fig. 231; also that described by Hensen, 77.1, and finally Ecker's specimen, 80.1. Of all these Coste's most deserves attention on account of the superb manner in which it has been figured.^ Concerning His' embryo R, some data about the coelom have been given by His, 81.1, 311. Twenty-five Days. — Embryos of this age are extremely rare. Fol has given a full but not w^hoUy satisfactory description, 84.2, of an embryo presumably of this age, though no data were obtained in re- gard to it. The embryo. Fig. 317, as compared with that at twenty- three days, has grown rapidly ; its greatest length is 5.6 mm.; its form has changed by the body hav- ing partially unrolled, but the head-bend and neck-bend remain and are more prominent than be- fore, owing to the embryo as a whole being less curved. The region of the fore-brain is brought close to the heart, the head being still bent to the right; the limbs are a little larger and there is a well developed, dis- tinct tail . The other pr in- cipal change is that only three gill - arches show externally, md, 1, 2, the third and fourth being already invaginated in connection with the for- mation of the cervical sinus. It must be added that this embryo was not quite normal, as is shown especially by the condition of its veins. The representation of the external form of the head in the figure is probably not entirely correct. Twenty-six Days. — Mall, 91.3, gives a superb figure and com- plete anatomical description of an embryo, the probable age of which he fixes at twenty-six days. Twenty-seven to Twenty-eight Days. — Embryos of this age are characterized by the extreme development of the neck-bend. Fig. 218, the apex of which forms, as it were, the summit of the embryo; the greatest length from this apex is 7-8 mm. To the age of about twenty-eight days are to be assigned the embryo described by Johan- nes Miiller, 30.2, one figured by Coste, 47.1, PI. III., one de- scribed by Waldeyer, 63.1, and four embryos in His' collection. Fig. 31T.— Fol's Embryo of 5.6 mm., probably twenty- five Days old. op, Position of the optic vesicle ; /r, fossa rhomboidalis; s.s, segments; a, I, anterior limb ; pi, pos- terior limb ; nwi, mandibular, 2, hyoid ; 2, bronchial arch. GROWTH AND EXTERNAL DEVELOPMENT. 387 menschl. Embryonen," Heft II., 8), (A), and XL. (3tt.) ; concerning XL. numbered by him ("Anat, L (B), LXI. (Eck. I.), 11. , see His, I.e., pp. 24, 93. Of A and B His has published a detailed anatomical ac- count ( " Anat. menschl. Embryonen," Heft I., 14- 99). I choose for my illustra- tion Fig. 218, His' embryo A, because it shows the neck-bend most perfectly ; how entirely the prominence of the neck-bend alters the shape of the embryo will appear immediately if Fig. 218 be compared with Fig. 175. As changes since the twenty-fifth day we note especially the distinctness of the olfactory pit {JRiech- gruhe) and of the still open invagination to form the lens of the eye, the deepen- ing of the cervical sinus (sinus praecervicalis of His) , and the partial closure of the allantois-stalk {Bauch- stieT) around the proximal part of the now narrow pedicle of the yolk-sac ; the closure of the Bauchstiel forms the umbilical cord, but the cord itself is veryshort and in proportion to the embryo very thick. In all parts there has been an obvious development since the twenty-fifth day. Fig. 217, but further details may be omitted. Comparison of this embryo with others of the same stage show that there is a considerable variation as to the nature and degree of curvature of the back, in consequence of which the specimens differ somewhat in general form, though agreeing closely in structure. Twenty-nine to Thirty Days. — Embryos 8-10 mm. A number of specimens, which probably belong to the middle of the fifth week are Imown. For my illustration I give a drawing, Fig. 219, of an em-- bryo sent to me by Dr. H. J. Garigues of New York; the data suffice only to determine the age as the fifth week ; the specimen appeared- normal and v/ell-preserved, but upon microtoming it, it was found to be in poor condition histologically ; it has interest because it show^ with especial clearness the relations of the foetal appendages. The embryo proper has begun to straighten its body, and as the outline over the region of the medulla oblongata, compare Fig. 217, Fr, has become less curved, the head begin sto appear to form a right angle with the body; the plfactory pit, ol, has deepened; the lens of the eye, op, is well n^arked, as is also the lachrymal groove descending from the eye; the cervical sinus, c.s, has deepened but is still open; Fig. 218.- -His' Embryo A, 7. 5 mm. long, twenty-eight days. - - -- '• Probable age X 10 diams. 388 THE FOETUS. , the limbs have lengthened and in other specimens begin to show the differentiation of the hand and foot. About two-thirds of the allan- tois-stalk has closed to form the umbili- cal cord, Tim, from the end of which ex- tends the amnion, Am. The long j^olk- stalk, Vi.s, ends in the pear-shaped yolk- sac, Vij the allantois-stalk or Bauchstiel, Yj / Bs, which runs to the chorion, Cho. In this group belong the embryo of Rabl (0. Hert- wig, "Entwickelungsges.," 3te Aufl., Fig. 158), the em- bryo of 10 mm. of which Phisalix gives a detailed anatomical description, 88. 1, — also seven embryos, enumerated bj' His, " Anat. menschl. Embryonen," Heft II., 8, and described there p. 45-7, and His' embryo Pr., I. c, p. 10, 238, Taf. X., Fig. 8, Taf. XIIL,Fig. 47 — this last by far the most perfect drawing of this stage which we possess. Thirty,one to Thirty-two Days.— EmhT JOS of 10-13 mm. (see His, I.e., Heft 11., 47-51 and, for a list of ten specimens, p. 8-9, Taf. X., Figs. 13, 14, and 15, Taf. XIII. , Fig. 6) . The age of the embryo at this stage can as yet only be estimated, as in no case have we data sufficient for a reliable determination. For a typical illustration of this stage we may take His' Br. 1, I.e., Taf. XIII., Fig. 6, which measured 11 mm. The back has straightened out, though the lower end of the body is still rolled over; the head has risen somewhat and enlarged both absolutely and in proportion to the rest of the body. Be- tween the end of the region of the hind-brain and the level of the arm the outline has become slightly concave; this concavity His designates as the Nackengruhe. The cervical sinus is so deep that the second, third, and fourth gill-clefts have disappeared from the external surface ; the first gill-cleft remains and can al- ready be recognized as the anlage of the external auditory meatus; it is separated from the mouth by the prominent mandibular arch. On the cephalic side of the mouth the maxillary process has become more prominent, but the two processes do not yet meet in the median line. The primitive segments are still marked externally. The limbs show the tripartite division ; the fore limb is more ad- vanced than the hind limb ; the division of the digits of the hand is just indicated. The abdomen bulges out owing to the growth of the liver. There is a true tail, which is now near its maximum develop- Fis. 2f9.— Embryo of 9.8 mm. Minot Collection No. 145. Probable age thirty days. X 5 diams. GROWTH AND EXTERNAL DEVELOPMENT. 389 ment. The umbilical cord has lengthened and shows the commenc- ing spiral twisting ; the amnion springs from the end of the cord, leaving only a short stretch of the allantois-stalk between the cord proper and the chorion ; the amnion lies close to the embryo. In the fresh specimen something can be seen of the shape of the brain ; especiaUy noteworthy, among the points thus to be recognized, is the sharp bend {Briickenkrummung) at the deep-lying anterior end of the hind-brain or region of the sinus rhomboidalis. In embryos a little older than these the changes in form above mentioned have progressed further. The specimens measure 12-13 mm. The body is straighter ; the head is larger and has risen so as to be at about right angles to the body ; the concavity below the hind- brain in the outline of the neck (Nackenkrummimg) is more marked ; the limbs are longer, the fingers more distinctly marked; the tail is at its maximum development as a free appendage ; where the man- dibles meet in the median line the separation of lip and chin has begun ; the second gill-cleft is invaginated into the cervical sinus and can no longer be seen on the outside. Thirty-five Days. — Embryos of 14 mm. The correlation of age and size of this stage cannot be regarded as absolute, though we can Fie. 220. —Embryo of about 14 mm. Minot Collection No. 120. sumed age, thirty-five days. X 5 diams. _;■.■■!■;;;■ Ijl^l-yiS!?' W Fig. 221.— Dorsal View of an Embryo of about 14 mm. Minot Collection, No'. ISO. Assumed age, thirty -five days. X 5 diams. (Compare Fig. 320). % say (His, I.e., Heft III., '^39) that embryos of this length are about five weeks old. The body is now nearly straight; the limbs project beyond the outline of the body in profile views; the abdomen, owmg to the large size of the heart and liver, bulges far out; in side views the area of the head is about equal to that of the rest of the body; the outline of the head shows the head-bend and neck-bend most clearly marked; the neck-bend is characterized by the prominence at that 390 THE FCETUS. point ; the prominence is often less than in Fig. 220. The umbilioal cord frequently contains one or several coils of the intestine and makes one or two spiral turns. The stalk of the yolk-sac is long, and projects quite far from the end of the cord between the amnion and chorion. In a dorsal view we can see that the limbs are some- what flattened and in a plane nearly parallel with the longitudinal axis of the embryo, but the planes of the arms are inclined so as to meet above the head, and the planes of the legs are inclined to as to meet below the tail. Owing to the flattening of the limbs we can already distinguish the inner or palmar surfaces from the outer. GROWTH AND EXTERNAL DEVELOPMENT. 391 Noteworthy is the irregularly crenulated appearance of the walls of the medullary tube or spinal cord. Fig. 222 is copied from Coste, and is valuable on account of the very large number of anatomical facts which it records. Coste gives no data but states that the specimen was "about thirty-five days old." Thirty-eight Days.— EmhTjo of 15 mm., in a chorionic vesicle of 45 by 40 mm. The age of this specimen, Fig. 233, is known by estimate only. It has been su- perbly figured by His (" Anat. menschl. Embryonen," Taf. XIV., Fig. 5). This stage represents the transition from the embryo to the foetus, be- cause after the fortieth day the form is distinctly human. The head has risen • considerably, and the back has straightened still more, the rectangular neck- bend thus becoming empha- sized. The body has become still more protuberant on the ventral side, and in side views the arms no longer reach to the outline of the body. Forty Days. — Embryos of 19 mm. The hea(J has risen far toward its definite position, with the result of a very rapid apparent increase in the length of the embryo. The change of position of the head results in bringing the mid-brain finally directly above the hind-brain, the embryo being conceived as having the body vertical. Durink the elevation of the head the concavity {Nackengrube) at the bacg of the neck is gradually obliterated. In both head and rump the external modelling, which in earlier stages indicated more or less the position of the internal organs, has become blurred and in the next stage is found to have nearly or quite disappeared. The max- illary processes have met and united in the median line. The an- lages of the eyelids have developed. The concha of the ear is indi- cated. The arm reaches beyond the heart; the fingers appear as separate outgrowths. Fifty Days. — Embryo of 21 mm. I have a fair specimen which came into my possession with no history whatever, but it agrees very closely with Fig. 33, Taf. X., in His' "Anat. menschl. Embryonen," of His' embryo Ltz, of which he fixes the probable age as just over seven weeks. The head is nearer its final position than in Fig. 323, and relatively larger in proportion to the body. In the eye, cornea and conjunctiva are clearly separated ; the face has the foetal form, the nose, mouth, and chin being fully marked off. The arms are •clearly divided into upper and lower segments ; the five digits are Fia. S33.— His' Embryo XXXIV. (Dr.), 15 mm. long from the Neck-bend to the Coccygeal Bend. Age estimated at thirty-seven to thirty-eight days, X 5 diams. 393 THE FCETUS. well developed; the hands rest over the heart and nearly touch one another. In the specimen figured the outline of the abdomen is ab- normal. The leg shows the tripartite division; the toes are just be- ginning to be free, but the hind limb is much less advanced than the fore limb. The tail is still a freely projecting appendage. Fifty-three Days.— Embryo of 22 mm. The specimen, Fig. 224, is probably not quite normal, but except for the extreme and unusual curvature of the back it agrees closely with His' embryo Zw, which is figured by him. I.e., Fig. 24, Taf. X., as a normal embryo of pre- sumably about seven and one-half weeks. My specimen I received in 1884 with the following history. " Menstruation began January 26th. February and March slight show every few days. Abortion March 30th," which is insufficient to de- FiG 224.— Embryo of 22 mm. Minot Collec- tion, No. 54. Probable age, fifty-three days, X 3 diam» Fig. 225. —Embryo of 28 mm. No. 144 of Minot Collection. Assumed age, sixty days. X 3 diams. termine the age. As compared with the last stage there are com- paratively few changes of external form ; the most noteworthy are perhaps the increased development of the legs and feet and the com- mencing disappearance of the free tail. At this time the protrusion of the coils of the intestine into the ccelom of the umbilical cord is about at its maximum. Sixty Days. — Embryo of 28 mm. The specimen figured re- sembles closely in form, though larger than, His' embryo Wt (Fig. 25, Taf. X., I.e.), which he has determined as a normal embryo of about eight and one-half weeks. My specimen. Fig. 225, came to me with no data. The head is stiU larger in proportion to the body than in Fig. 223. The face shows the two lines, which, as seen in profile, mark the two ridges which run over the cheek, one alongside the" nose to the corner of the mouth, the other from the eye j these ridges are highly characteristic of the ninth week, and traces of them. GEOWTH AND EXTERNAL DEVELOPMENT. 393 not rarely persist in the adult physiognomy. The limbs have grown considerably, the hands being lifted toward the face; at the elbow there is a considerable bend; the toes are all free and the soles of the feet are turned toward one another. The tail has disappeared as a free appen- dage. The external genitalia are considerably developed; the cli- toris-penis projects some dis- tance. Sixty-four Days. — Embryo of 33 mm. The specimen, Fig. 336, came to roe with the following history : " January 4th, 1886, last flow began; March 13th, 1886, abortion ; " between these two dates are sixty-eight days; but as the flow took place conception probably occurred af- ter menstruation, therefore if we deduct four days, making the age sixty-four days, we shall probably not be far wrong. It will be no- ticed that the head has not yet mr da™'^x°3 assumed its final angle with the body. On the other hand the protuberance of the abdomen is much reduced, so that the body as a whole has begun to have a more slender form than in earlier stages. In this specimen the eyelids have not even begun to meet ; in another I have they have met, Fig. 337, ex- cept just in the centre where is still a loophole. This specimen was brought to me with the statement that it was just sixty days. I en- deavored, unsuccess- fully, to get Fig. 226 —Embryo of 32 mm Collection. Probable age sixty-four days, diams. Fig. 227. —Embryo of 34 mm. Collection. Front View of Face. x3diams° the exact Fig. 228. — Embryo of 55 mm. No. 97 Minot Collection. Assumed age, seventy - five days. Natural size. 394 THE FCETUS. data. The large size, 43 mm., and advanced development of the embryo led me to consider the age given as erroneous, and to be- lieve the true age to be perhaps six- ty-seven days. Seventy -five Days . — Embryo of 55 mm. I fig- ure next. Fig. 338, a foetus concerning which I possess no data. Comparison with embryos of two and three months leads me to place it a little under half-way be- tween them. The specimen has essentially the configuration of the young child; but Fig. 230. -Front View of the Head and Face of the Embryo, Fig. 229. FlQ. 229.— Embryo of 78 mm. No. 74 Minot Collection. Age three months. the head is very large, and the body slender ; the posi- tion of the limbs is typi- cal ; the upper arm is bent down, the forearm extends toward the chin ; the knee is bent so as to throw the foot toward the median line; the soles of the feet are placed obliquely facing one another; the anlages of the nails can be recog- nized on both the fingers and toes. Embrj'os of the eleventh and twelfth weeks are very rarely obtained. I have never had a normal one of this period with data to determine the age. Three Months. — Embryos of 78-80 mm. In my experience there Fig. 281.— Embryo of about 120 mm. No. 61 Minot Col- lection. Assumed age, one hundred and ten days. Nat- ural size. GROWTH AND EXTERNAL DEVELOPMENT. 395 is no other age at which abortion of normal embryos occurs so fre- quently as at three months, and I possess a number of specimens of this age, which agree very closely with another in size and form. The foetus drawn in Fig. 239 may be taken to represent very accu- rately the form and size of the human embryo at three months. The position of the limbs is typical for this age, but there are slight variations in that the hands, one or both, may project more or be higher or lower ; usually the right foot lies in front of the left, but not always. Fig. 230 gives the front view of face of the same embryo to show the closed eyelids, the broad triangular nose, the thick lips and pointed chin. Three and One - half Months . — Embryos of 108-110 mm. I have sev- eral specimens which rep- resent this age. I figure two of them, one to show the natural attitude, Fig. 231, in utero, the other to show the natural attitude assumed by the embryo when released from its membranes. The first specimen came to me with no history, but as it is cer- tainly a little larger than several other foetus of about one hundred and six days it is probably a little older. The head is bent forward, Fig. 231; the back is curved ; the arms and legs are both raised toward the face ; the right leg is nearlj' straight so that the toes are brought against the forehead, while the left leg is bent at the knee, bringing the left foot against the right thigh. In this attitude the embryo fiUs out as ^^ ^ -Embryo of iis perfectly as possible an oval space and fits. mm. /o^ Jf^Mmot^coUe<> therefore the cavity of the uterus, ihe second six days. Natural size. specimen. Fig. 232, shows the attitude assumed by the embryo when free, and proves that the position in utero. Fig. 331, is a constrained one. This foetus was received November 30th, 1883. The delivery took place on the morning of that day, and the 396 THE FCETUS. last menstruation had ceased one hun- dred and six days previously ; the re- markahly fresh con- dition of the foetus indicated that it had been dead only a very short time, so that we cannot be far wrong in putting its. exact age at one hun- dred and six days. Four Months. — Embryo 155 mm. The foetus, shown in Fig. 233, came to me in a very fresh con- dition, January 3d, 1887, with the state- ment : " Conception, said to have taken place September 1st, 1886 ; foetus came- away January 2d, about noon." The embryo is typical in size and development for four months, ex- cept that the crown is higher than usual, and the antero-poste- rior diameter of th& head somewhat be- low the average. The n'atural atti- tude in utero is sim- ilar to that of Fig. 231, the attitude shown is that as- sumed by the foetus, when released from, the membranes. Fia. 233.— Embryo of 155 mm. No. 180 Minot Collection ; Age, one hundred and twenty-three days. Natural size. CHAPTER XIX. THE MESENCHYMAL TISSUES. As the numerous tissues which result from the differentiation of the mesenchyma enter to a greater or less extent into the formation of the organs of which the main parts are derived from the ectoderm, entoderm, or mesothelium, it is desirable to begin the study of the organs with a general review of the mesenchyma. The development of the skeleton is treated in the next chapter, p. 422. ClassiflcatioxL of Mesenchymal Tissues. — The fundamental and essential characteristic of the mesenchyma is, that the cells are some distance apart, but connected together by their own protoplas- matic processes. The tissue is made up of anastomosing cells. The spaces left between the cells are filled with intercellular substance, which, owing to the size of the spaces, constitutes a large part of the tissue. . In this respect the mesenchyma offers a marked contrast to all epithelia, for the latter have the intercellular substance reduced to a minimum. The intercellular substance is an extremely important factor in the difEerentiation of the mesenchymal tissues; in fact so important that it affords a better basis for the classification of the ' tissues than the cells themselves. To these fundamental conceptions I attribute a great value. In the primitive stage we have cells with small protoplasmatic bodies, connected by few processes and imbedded in a homogeneous matrix (intercellular substance) . We can distinguish in subsequent changes three main sets of modifications: 1, those which are spe- cially characterized by changes in the basal substance ; 3, those char- acterized chiefly by changes in the cells ; 3, those characterized by the special arrangement of the tissues produced by the differentiations of the mesenchyma. In the first series I put the development of connective-tissue fibrils and fibres, of the intercellular network both elastic and non-elastic, of mucin, as in Wharton's jeUy, of cartilage (chondrification), of bone (ossification), and also the disappearance (or liquefaction?) of the basal substance, and finally its hypertrophy. In the second series I put the development of the blood-vessels, of the lymphatic vessels, muscle-cells, fat-cells, pigment-ceUs, and of the marrow of bones. In the third series I put the development of the connective-tissue cavities such as the synovial, bursal, and subarachnoid, and the formation of special layers such as the subepithelial basement niern- branes, the submucosa, the cutis, and so forth. What little there is to be said in regard to the special layers wiU be foun^ in connection with the history of the special organs of which they form parts. The following table gives the classification adopted. It must be 398 THE FCETUS. borne in mind that the classification is somewhat arbitrary, since in. all the tissues modifications occur in both the cells and the inter- cellular substance; moreover, several differentiations may occur simultaneously or successively in the same part; for instance, the fibrillsB and network are usually found together ; cartilage may or may not have fibrillae and elastic tissue. Mesenchymal Tissues. First Series. Second Series. Third Series. (Changes in matrix) . (Changes in Cells.) (Special arrangements) 1. Fibrils. 1. Blood vessels. 1. Cavities. 2. Network. 2. Lymphatics. a. synovial. a. yellow elastic. 3. Muscle-cells. b. bursal. b. white non-elastic. 4. Fat-cells. c. subarachnoid. 3. Mucin. 5. Pigment-cells. 2. Membranes. 4. Chondrification. 6. Marrow. a. basement. 5. Ossification. b. submucous. 6. Disappearance. ? by liquefaction. c. dermal. etc. 7. Hypertrophy. 8. Ligaments. 4. Tendons. Embryonic Mesenchyma. — Concerning the very early history of the mesenchyma we have little satisfactory knowledge beyond the fact that the cells of the mesoderm are at first closely crowded and as they move apart are seen to remain connected together by proto- plasmatic processes. As regards the shape of the cells I distinguish two stages, of which the earlier has not hitherto been definitely recognized. In the first stage, which I have observed to occur to elasmobranchs, birds, and mammals, the protoplasm forms a complex network in which the nuclei are scattered at irregular intervals ; around the nuclei there is often little or no condensation of protoplasm, so that there are, properly speaking, no perinuclear cell bodies. The tissue corre- sponds, therefore, very poorly to our conventional conceptions. This stage is well represented by the mesoderm of the umbilical cord in a human embryo of about seven weeks. Fig. 206, p. 358. The form of the cells — -or, if the expression be preferred, of the nodes of the reticulum— varies greatly, but in a definite manner in the various regions of the embryo ; the variations depend chiefly upon the den- sity of the tissue and its trend ; for instance, in amniote embryos with two to four gill-clefts there is always a distinct contrast between the dermal mesenchyma, which is of loose texture and with no de- cided trend, and the mesenchyma between fhe muscle-plate and the medullary tube, which is dense and has elongated cells. The differ- ences have never been comprehensively studied, and we can only say that they are the expression of unlike conditions of origin and growth of the various parts of the mesenchyma. In the second stage, which seems to be reached by all the cells of the mesenchyma sooner or later in aU vertebrates, the protoplasm has formed distinct cell- bodies around the nuclei, and there are no considerable accumula- tions of protoplasm except aroimd the nuclei. This stage is illus- rated by the human umbilical cord at three months. Fig. 207, p. 359, and is still more typically and characteristically shown by the mes- oderm of a chick of the third or fourth day, or in a rabbit embryo THE MESENCHYMAL TISSUES. 399 of ten or eleven days ; in the dog-fish this stage is not reached until considerably later in the development than in the amniote embryo. In the chick, Fig. 234, the cells have a large nucleus of rounded form, with a distinct intranuclear reti- culum of protoplasm and one or several granules of chromatin; the nucleus is surrounded by granular protoplasm, constitut- ing a small cell-body, which sends off tapering processes to unite with similar processes of other cells ; the processes are sometimes very short, but vary in length up to two or three times the diameter of the cell-bodies. The IpTiD-fVi rif ihp ■nrnfpe.iiP^ nl«r> vn-riAa ^i<*- 234.— Mesenchyma of a Chick Embryo of lengm OI tne processes aiSO varies ^-^^ xhird Day from close to the Otocyst. A, a in different regions, so that the nucleus in karyoldnesis; the chromatin loops n . *--'.' are seen in cross-section. cells m one region are more or less crowded than in others; the cells also vary in shape, being elongated in certain districts ; these differences are all significant as the results of previous development and as establishing conditions for the subsequent development. In young mammalian embryos the cell-bodies are less well marked than in the chick, and the processes form a network of fine threads between the cells, as can be seen in places in rabbit embryos, as late at least as the seventeenth day. The matrix is perfectly clear, homogeneous, colorless, and structure- less ; it is of slight consistency, and scarcely stains with any of the most used histological dyes. Intercellular DiflFerentiation. — The means by which differ- entiation of the mesenchymal matrix is effected are little understood. If we accept the view, which is, however, as yet by no means be- yond doubt, that the fibrils and network arise from the cells, we can account for a part, but only for a part, of the intercellular structures. If, on the other hand, we hold that all intercellular structures are of intercellular origin, then we can assume that there is some general principle in accordance with which they are all produced. Even in this case the cells must have some influence, since their presence and vitality are essential conditions. Experiments published by Harting are suggestive in this connec- tion. Connective-Tissue Fibrils. — The fine fibrils of the adult con- nective tissue appear quite early in the embryo in the intercellular substance. There are two theories of the origin of the fibrils : 1, they arise from cells; 2, they arise from the matrix. Their origin from cells was the view of the founder of the modern cell theory, Theodore Schwann, 39. 1, who thought that the cells grew in length and split into bundles of fibrils. Various modifications of this theory have since appeared; thus Obersteiner (Sitzungsber. Wien. Akad., LVI., 351) traces the fibrillse to outgrowth of spindle-shaped mesen- chymal cells. Max Schultze (Eeichert's Archiv, 1861, 13) thought that the cells fused together and their fused parts formed the fibrillffi as well as the intercellular substance, thus tracing the fibrillse 400 THE FCETUS. to a differentiation of the peripheral parts of the cells— a view which, somewhat modified, has been revived by B. Lwoff, 89.1, who maintains that the fibriUae arise from the surface of the cells, nearly the whole body of each cell being converted into fibrillse, which ex- tend along whole rows of cells and along their processes, enveloping the protoplasm. The origin of the fibrils by deposition in the matrix was first maintained by Henle ("Allgemeine Anatomie," Erste Aufl., 379). and was, in my judgment, demonstrated by Reliefs investigations, recorded in Strieker's "Grewebelehre," 1871, 62-67, upon the omentum, and by Ranvier's later observations ("Traite technique d'Histologie," 405-411). KoUiker, whose judgments upon histological problems are rarely mistaken, has accepted in his " Gewebelehre," 6te Aufl., 133, the intercellular origin of the fibrils. If we examine a tissue in which the fibrils are just beginning to appear, as, for instance, the human umbilical cord toward the end of the third month. Fig. 207, p. 359, or the omentum of a sheep embryo of 17 cm., we find the fibrils running singly and in every direction, both parallel with the cells and their processes and at aU angles with them. The omentum, as pointed out by Rollet, is a particu- larly favorable object, for we are sure of having the entire length of the fibres. The cells of the omentum gradually assume (sheep embryos 4-7 cm.) an elon- gated spindle form, remain- ing connected together only by very few processes, which arise chiefiy from the end of the ceUs ; the nuclei become oval, and when stained with haematoxylin have a dis- tinct membrane, and consist of a clear outer layer and a dark granular central part. Between the cells, and for the most part remote from them, appear the fibrils, which grow in length and number. In later stages. Fig. 335, the ceUs of the omentum are more attenu- ated, and their ends are united so as to form a net- work, though some of the cells appear to terminate without any con- nection with their fellows ; the nuclei are more finely granular and have lost the clear outer zone, characteristic of earlier stages. The iibrillse have grown in length and increased enormously in number; Fig. 23d —Omentum of a Human Embryo o£ five Months, c c. Connective cells forming a network; leu, leucocyte : fb. flbrillfe x 363 diams. THE MESENCHYMAL TISSUES. 401 they form bundles which take a wavy course ; these bundles frequently subdivide and unite, so that they form a network ; their course and arrangement are independent of the trend of the cells, and there is nothing to suggest any genetic connection between the cells and the bundles of fibrils. Scattered about there are also usually a few leucocytes, Fig. 235, leu, which are readily distinguishable from the true mesenchymal cells or so-called connective-tissue corpuscles, c c. The bundles of fibrils correspond to the connective-tissue "fibres" of the adult ; each fibre consists of a large number of fibrils. The em- bryonic fibrils differ from those of the adult in staining much more readily. The growth of the fibres depends upon multiplication of the fibrils for Harting (" Eecherches micrometriques sur le developpe- ment des Tissus," etc., 1845, p. 53) found that the fibrils measured 0.0010-0.0014 mm. in the foetus and from 0.0007-0.0017 mm. in the adult ; as, therefore, the fibrils do not thicken they must increase in number as the bundles or fibres enlarge. Ranvier, I.e., finds that the fibrillse have no connection with the cells in three tissues, which he has studied in regard to this point, namely, the embryonic dermis, the developing tendon, and the scle- rotic cartilage of rays. E. A. Schafer (Quain's " Anatonjy," ninth edition, II., 73) writes as follows: "The view which supposes that a direct conversion of the protoplasm of the connective-tissue cells takes place into fibres, both white and elastic, has of late years been widely adopted, but it seems to rest less upon observation than upon a desire to interpret the facts in accordance with the conceptions of Beale and M. Schultze, according to which every part of an organized body consists either of protoplasm (formative matter), or of material which has been protoplasm (formed material), and the idea of deposition or change occurring outside the cells in the intercellular substance is excluded. But it is not difiicult to show that a formation of fibres may occur in soft substances in the animal organism, independently of the direct agency of cells, although the materials for such forma- tion may be furnished by cells. Thus in those coelenterate animals in which a low form of connective tissue first makes its appearance, this is distinguished by a total absence of cellular elements, the ground-substance being first developed and fibres becoming formed in it. Again, the fibres of the shell-membrane of the bird's egg are certainly not formed by the direct conversion of the protoplasm of the cells which line the oviduct, although it is probably in matter secreted by those cells, and through their agency, that the deposit occurs in a fibrous form." Intercellular Network or Elastic Tissue.— The intercellu- lar substance of the adult contains in most parts of the mesenchyma a network which varies greatly in appearance. This network has hithero been described usually as being formed of elastic fibres ; now since the material which forms the network does not always resem- ble fibres, but often rather lamellae, and since, as shown by F. Mall, 88.3, 91.1, some parts of the network do not contain elastin, it seems very undesirable to continue the use of the term elastic fibres, which is entirely misleading. I shall therefore speak of the two forms of tissue as yellow elastic network and white non-elastic network respectively. Mall states that there is a non-elastic mate- 36 403 THE FCETUS. rial which alone forms the white network, but which in the yellow network forms a sheath around the elastic core. Concerning the development of the network we possess little accu- rate knowledge. Just as with regard to the intercellular fibrils, p. 399, there are two theories: according to one, the network arises by metamorphosis of the cells; according to the other, by differenti- ation of the matrix. The origin from ramifying cells was the old theory and seems at first thought plausible — see Bonders' remarks inZeit. wissensch. Zool. ,111., 358 — for if we assume the cell processes to be converted into elastin a network would result. The attempt, however, to demonstrate the actual metamorphosis has hitherto been unsuccessful . Kuskow ,87.1, found that in the ligamentum nuchas of the embryo, after digestion in cold pepsin solution, the elastic fibres could be seen uniting with the elongated mesenchymal nuclei, usually with the ends, sometimes with the sides of the nuclei. Heller, whose paper I know only from the abstract in Hofmann-Schwalbe's Jahres- bericht for 1887, 126-137, is said to have seen the connection with nuclei both in the ligamentum nuchas and in the very young arytenoid cartilage of the embryo; in the cartilage of the ear, on the other hand. Heller states that there is no connection of the elastic fibres with either the nuclei or the cells. These observations show that the elastic tissue may enter into special relation to the nuclei, but throw no light on the significance of the connection ; we do not yet know whether the fibres develop independently and afterward unite with the nuclei, or are united with them from the start. Kuskow's sug- gestion that the elastic network is formed by the nuclei is not likely to be verified, because nuclei never form outgrowths or unite with one another to make reticula, so far as heretofore known. If the connection With the nuclei is secondary, then the network may be intercellular in origin. Ranvier, "Traite technique," 401, 411, has shown that the elastic tissue first appears in the form of rows of granules or minute glob- ules, probably of elastin, which subsequently fuse together into a net- work the lines of which are marked out by the original deposition of the globules. To form an elastic membrane the globules, instead of being arranged in lines, are deposited in small patches, and by their fusion form a plate. In elastic cartilage the granules first make their appearance, it is true, in the immediate neighborhood of the carti- lage cells. This renders it not improbable that the deposition of the granules is infiuenced by the cells, but does not prove that they are foiTiied by a direct conversion of the cell-protoplasm. Indeed the subsequent extension of the fibres into those parts of the matrix that were previously clear of them and in which no such direct conversion of the protoplasm seems possible is a strong argument in favor of the deposition hypothesis. For an admirable discussion of the two views seeH. Rabl-Rlickhard, 63.1. As to the time when the elastic fibres appear we may saj' in gen- eral that it is quite late. They appear in the ligamentum nuchse of cow embrj'os of 15 cm. ; in the cartilage of the ear in embryos of 30 to 33 cm., and human embryos of five months, Rabl-Riick- hard, 63.1, 43; in the arytenoid cartilage in cow embryos of 55 cm. ; in adult fibro-elastic cartilage the elastic network is still de- THE MESENCHYMAL TISSUES. 403 veloping, and is not formed at all in the sheaths of nerves until adult life. The elastic network grows by thickening the fibres and plates, which are found much larger in diameter in the adult than in the foetus. In this respect it forms a striking contrast with the inter- cellular fibrillae, which grow princially by multiplication. Concerning the development of the white non-elastic network we know almost nothing. Mucous Tissue or "Wharton's Jelly. — In man this tissue oc- curs only in the umbilical cord. It is characterized by the develop- ment of mucin in the intercellular substance. The tissue has already been described, p. 358, and I have only to add that the mucin is present in a diffuse form, and has, so far as yet known, no special structural arrangement. Mucous tissue is said to occur in various parts of the body in fishes, but unless it contains intercellular mu- cin it cannot be regarded as true mucous tissue, in the sense here considered. Cartilage. — Cartilage begins to be differentiated earlier than any other of the mesenchymal tissues, except the blood-vessels, which are developed much earlier, and perhaps the smooth muscle-cells. It is probably older phylogenetically than any of the other tissues of the group except the two mentioned, for not only does it appear very early in the embryo, but is found in invertebrates. It is for conven- ience only that I consider cartilage after the fibrillse and elastic net- work, for both of these intercellular structures appear in certain forms of cartilage. In this section the history of cartilage is considered under the following heads: 1, condensation of the mesenchymal tis- sue to form the anlage of the cartilage ; 2, appearance of the matrix ; 3, young cartilage; 4, growth of cartilage; 5, mature cartilage; 6, fibrillar cartilage; 7, elastic network cartilage. 1. Condensation of the Tissue. — This takes place simply by the cells becoming very much more closely crowded together than in the surrounding mesoderm; at first merely the central portion of the anlage is thus marked out and there is a very gradual transition to the looser mesenchyma about ; for every piece of cartilage there is a separate anlage, which is distinct from the start ; one exception to this rule occurs in the case of the vertebrse, as has been stated by Gegenbaur, and has been shown with great precision for birds and mammals by A. Froriep, 83. 1, 86.1. Another exception is offered by cartilages of the limbs of amphibia, which Goette and H. Strasser, 79. 1, have shown to be coalesced, when they first appear. As development progresses, the anlage becomes more and more sharply defined until at last its limit can be determined within a cell or two. The cartilage cells are now so crowded that the nuclei, as has been observed in all classes of vertebrates, seem almost to actu- ally touch one another— see H. Strasser, 79.1, 245, and A. Froriep, 86.1, 73. When the anlage is completed its peripheral cells become elon- gated and form the anlage of the perichondrium ; while the central cells, by taking on the rounded form, begin their metamorphosis into cartilage cells; the perichondrium is a thin layer. C. Hasse, 79. 1, 2, thinks that the cells assume a spindle shape first, and afterward 404 THE FCETUS. take on the rounded form, at least in elasmobranchs. It is uncertain whether the two stages can be distinguished in the higher vertebrates. The first cartilaginous anlages appear in the chick during the fifth day, and in the rabbit, I think, about the thirteenth day. The ver- tebral are probably always the first cartilages to be indicated by completed anlages. The other cartilages become recognizable later ; the exact times need to be determined by closer study than has yet been attempted. 2. Appearance of the Matrix. — Prce-cartilage (prochondrium, Vorknorpel). — The intercellular substance, as the cells begin to move apart and lose their connections with one another, gradually assumes a greater density and finally becomes highly retractile and quite resistant mechanically and chemically, and at the same time acquires, at least in many cases, a great affinity for carmine and hsematoxylin. Hasse, 79.2, states that this color-reaction always appears in the young cartilage of elasmobranchs, and therefore he proposes to dis- tinguish the stage as a distinct one, since the matrix of the fully differentiated hyaline cartilage does not stain ; for the young car- tilage with colorable matrix he proposes the term Vorknorpel, which I have rendered prcB-cartilage. Hasse states that in the prae-carti- lage of elasmobranchs the matrix consists of numerous fibriUae held together by a cementing substance. This is now generally held to be the structure of the matrix in adult hyaline cartilage — see, for in- stance, Spronck, 87.1, and Kolster, 87.1, who both give references to the preceding literature. Hasse further states that in prse-carti- lage the matrix is of uniform structure throughout, and that there are no capsules around the cells. The cells of young cartilage are said to contain glycogen; Rouget claims to have found it in the sheep embryo at two months. Many authors have held that the matrix was formed as a series of capsules, one around each cell ; the capsules grow and fuse. In support of this view there are no satis- factory observations known to me. As it is adopted in Quain's "Anatomy" by Schaefer (ninth edition, II., 84), I presume it rests upon some good authority, which I have overlooked. When the condensed mesenchyma is beginning to change, dark ir- regular masses appear among the cells ; these are the " prechondral elements" of H. Strasser, 79.1. Alice Johnson, 83.1, 400, states that they may be seen in the hind limb of the chick about the sixth day, and she interprets them as degenerated cells which have lost their nuclei. 3. Young hyaline cartilage differs but little from that just de- scribed, except that the matrix has increased and the cells are slightly larger. It is to be considered as the primitive form of tissue, from which all the modifications of adult cartilage are derived. In the thyroid cartilage of a three-months human embryo I find the cells farther apart and a little larger than in younger stages ; the cells are still small and are here and there in groups of two ; they are not round but more or less compressed in shape, and some of them appear to contain fat granules. In the same cartilage at four months the general appearance is the same as before, but the matrix stains un- evenly ; around the cells it is light, but between the cells intervenes a darker-colored portion which forms a network through the tissue. THE MESENCHYMAL, TISSUES. 405 In the neighborhood of the prochondrium the matrix is altogether light and the cells are in part larger, rounder, and with distinct spher- ical nuclei. In the tracheal cartilage of embryos of about seven months the cells are decidedly larger than those of the thyroid just described ; the rounded nuclei are very distinct ; the protoplasm is granular and entirely fills the cell space (lacuna) of the matrix ; the cells exhibit only a very slight tendency to form cell groups as in mature cartilage, nor are there any signs I can recognize as such, of the degenerative changes which can be seen in the adult. 4. Growth of Cartilage. — The matrix presumably grows by in- tussusception, and not, as some authors have maintained, by the continual conversion of the superficial protoplasm of the cells into matrix. If such a conversion took place we should expect to see the cells diminish in size, whereas they increase. The cells increase in number by division, and by additions from the perichondrium ; of the two factors the latter is probably the more important. The division of cartilage cells has been especially studied by W. Schleicher, 79.1. The division is indirect. The nuclear membrane first of all disappears or is converted into filaments which soon be- come lost among the other filaments developed within the nucleus. The filaments are at first short and irregular, but soon take on a stellate arrangement, and the chromatin becomes grouped into an equatorial plate, which divides, one group of chromatin elements moving toward one pole, the other toward the opposite pole. The division of the protoplasm is not effected as usual in animal cells, but by means of a cell-plate, as in many vegetable tissues ; the cell-plate forms a partition in the middle of the elongated binucleate cell ; the plate grows and becomes the matrix between the two daughter cells. As the plate thickens slowly the cells remain near together for some time, and one or both them may again divide with the result that there is a group of three or four cells. This grouping is highly characteristic of adult cartilage, but exactly when it first appears I do not know. It does not appear in embryonic cartilage, so that we must assume either that in the embryo the cartilage cells do not divide, or else that they divide and move apart very rapidly. In either case the grouping of the cells remains a sign of age, and ought perhaps to be regarded as the expression of a diminished vitality. Concerning the exact history of the perichondrial cells as they change into cartilage cells special investigation is needed. At pres- ent we can say hardly more than that the change takes place. 5. Mature Hyaline Cartilage. — The hyaline cartilage of the adult exists in two principal modifications, both characterized by the great development of the matrix and by having the cells for the most part in groups of two, three, or four, but distinguished by having in the one case large cells with round nuclei and well-developed protoplas- matic bodies, and in the other cells which have shrunk somewhat and are often compressed, with nuclei which are often indistinct and irreg- ular, and protoplasm which frequently contains fat globules. I believe that we have to do with two stages in the life-history of cartilage, and that the first modification, in which the cells are large, is the earlier stage, and represents the maximum of development, while the second, in which the cells are shrunk and fatty, represents a later 406 THE PCETUS. stage, with more or less degeneration. Dekhuyzen, whose papers I know only through the abstract prepared by himself for Hofmann- Schwalbe, Jahresbericht f. 1889, 82-83, was the first to interpret the mature cartilage as a degenerating tissue. In deciding upon the order of the two stages I have been guided chiefly by my observa- tions upon the growing cartilages of the lung in rodents, for in them the large, round, protoplasmatic cells lie between the connective-tissue cells on the one hand, and the fatty, compressed cartilage cells on the other, and clearly present a transitional stage of the transformation of the perichondria! cell into the old cartilage cell, and by the fur- ther observation that in the child at birth the bronchial cartilage consists entirely of large, rounded cells with spherical nuclei. The changes which are here noted as degenerative begin very early; thus Dekhuyzen states that they are well advanced in the epiglottis of the dog at birth. Little has been done upon the development of the matrix, but numerous researches have been made upon the structure and chemical composition of the adult matrix. A little upon the chemical develop- ment after birth may be found in Moner (Schwalbe's Jahresber. f. 1889, 81-82). 6. Fihro-cartilage appears first in the form of hyaline cartilage, and the fibrillaB, which appear in the matrix and seem to be homolo- gous with the ordinary intercellular connective-tissue fibrillse, are developed earlier or later. 7. Elastic cartilage also appears as hyaline cartilage, in which an elastic network is subsequently developed. Degeneration of Ossifying Cartilage. ^ — Besides the changes of a degenerative character, above referred to, the skeletal cartilages undergo a complete resorption, whenever in the course of develop- ment they are to be replaced by bone, except that in a few parts the cartilage is changed directly into bone. There are two forms of the resorption of cartilage, the direct and the indirect. The direct re- sorption occurs in only a few cases, as for instance in the greater part of Meckel's cartilage, and is characterized by the gradual dis- appearance of the cartilage without any preceding striking change in it. The indirect resorption occurs whenever the development of bone begins in the interior of a cartilage, and is accompanied by very remarkable structural alterations in the cartilage. So far as I know no exact study of the direct resorption of cartilage has yet been made, while the indirect resorption has been investigated again and again. The indirect resorption begins in the centre of the cartilage; the first step in the process is an enlargement of the single cartilage cells, without much or any change in the amount of the matrix between them, but the matrix assumes a granular appearance and acquires a gritty feel to the knife owing to the formation of calcareous deposits. Meanwhile the cartilage above and below the centre of degeneration becomes enlarged and piled up in elongated groups or columns which radiate from the centre for a certain distance toward either end. The radiating columns of cells taper toward their ends away from the centre, the end cells being smaller. In the matrix between the col- umns calcification takes place, so that calcified partitions separate THE MESENCHYMAL TISSUES. 407 the columns from one another. Turning now to the cells we find that thej' are undergoing a hypertrophic degeneration, for their enlarge- ment precedes their breaking down. There has been no sufficient study of the changes in the cells, but so far as my own observations enable me to judge the changes are probably as follows, Fig. 238. The cell enlarges and its protoplasm becomes granular; the enlargement continues and the cell appears to encroach upon the matrix more and more until ultimately adjacent cell-cavities coalesce; while this cor- rosion of the matrix is progressing the protoplasm of the cell becomes vacuolated; its nucleus becomes irregular and indistinct, and sooner or later disintegrates ; the cell then contracts and forms a flattened body, which stains more or less, but exhibits no distinct structure, unless now and then some trace of the original nucleus ; after the cells have shrunk their cavities fuse together, and sooner or later the eel Is break down into mere granular detritus. The coalescence of the cell-cavities does not take place equally in all directions, but principally as shown in Fig. 238, along radiating lines ; hence there arise a series of radiating cavities separated by partitions formed by the calcified matrix. While these changes are going on in the inte- rior of the cartilage, columns of the surrounding connective tissue go into the cartilage at various points, but always toward the de- generating tissue; each column contains blood-vessels also. As to why or how these columns penetrate the firm cartilage with their own soft tissues, we know nothing. The coh;mns reach the centre of degeneration just as the cells of the cartilage break down and the ingrowing new connective tissue at once fills the spaces formed in the cartilage and proceeds in its new site to produce bone. The degen- erative process now extends toward both ends of the cartilage and is followed by the formation of bone. The whole series of changes is commonly termed the ossification of cartilage, but this is incorrect, for the cartilage is destroyed, not ossified. The conjunction of the two sets of processes. Fig. 238, creates very singular microscopical pictures, which for a long time puzzled investigators. For further details see the following section on ossification. Ossification. — Bone is a mesenchymatous tissue, in which the cells have a characteristic shape and the matrix or intercellular sub- stance is large in amount and calcified. It is derived always by a direct metamorphosis of embryonic connective tissue or of embry- onic cartilage, and of periosteum. The ossification of cartilage plays a small part — for instance, at the angle of the jaw it has been observed to occur by J. Brock, 76.1, who found the cartilage cells changing into bone cells there, though nowhere else in the mandible. Most bones are formed by the ossification of the connective-tissue cells or undifferentiated mesenchyma, and by layers of bone added by the ossification of the periosteum. Bony tissue after it is once formed does not grow except by additions to its surface. In the sim- plest form of ossification we have a layer or membrane of connective tissue, in which the tissue changes into bone; this is called intra- membranous, direct, or metaplastic ossification. The direct ossi- fication of cartilage should also be placed under this head. As a modification of the simple ossification we must regard the ossification to replace cartilage, which is termed the intra-cartilaginous, indirect, 408 THE FOETUS. Fig. 236.— Parietal Bone of a Human Embryo of fourteen Weeks. After KoUiker. X 18 diams. or neoplastic ossification. In both types the actual processes of ossification are essentially the same, and the bone is completed by the co-operation of the periost. Metaplastic Ossification. — This may be conveniently studied in the parietal bone of the human embryo. About the end of the third month, in the parietal ^ j^ ^ region of the raembran- ' ^ ous skuU there appear minute calcified spic- ules, which rapidly in- crease in number and grow both in diameter and length so that they soon fuse together and form an irregular net- work, Fig. 236. The meshes of the network are fiUed with mesen- chymal cells, which are continually forming bone upon the surface of the spicules. Some- what later the fibrous periosteum appears upon the surface of the young bone, and adds osseous tissue to it. The transformation of the mesenchyma into bone is illustrated by Fig. 337, which represents a transverse section of the foetal mandible. The man- dible is closely invested by the fibrous periosteum, per, which is in part artificially separated from the bone, Os, the irregular bars of which have already acquired con- siderable thickness ; the spaces in the interior of the bony mandible are fiUed with a loose mesenchyma, the cells of which have large rounded, finely granular nuclei, with but little proto- plasm forming cell-bodies; the cells are connected by a rich network of fine granu- lar threads with one an- other. Some of the cells lie directly against the bone, either just touching it, or half or wholly imbedded in it; those which are in the obi' I* Fig. 237.— Transverse Section of the Mandible of a Human Embryo of the tenth Week. Minot Collection, No. 138. Os, Bone; vies^ mesenchyma; oW, o6Z', osteo- blasts ; per, periosteum, x 227 diams. bone are true bone cells, and are easily recognized as modified embry- onic connective-tissue cells, which have gradually accumulated more THE MESENCHYMAL TISSUES. 409 and more protoplasm so that, since the cells begin to enlarge as soon as they touch the bone, they are found to have grown considerably by the time they are completely imbedded. The connective-tissue cells, which lie against the bone, are known as osteoblasts, a name proposed by Gegenbaur in 1864; though often close together they always are separated by distinct spaces from one another ; they are rounded, polyhedral, or triangular in form, and frequently are so crowded over the surface of the bone that they present a distinctly epithelioid arrangement; the nucleus usually lies toward one side or end of the cell. The osteoblasts become imbedded in the bony matrix and thereby converted into bone cells, not by migration, but by the growth of the calcified matrix, the formation of which goes on first on the side of the osteoblast toward the bone, and gradually advanc- ing overgrows the osteoblast and continues beyond it. The history of the intercellular threads of protoplasm during the transformation of the connective-tissue cell into an osteoblast, and then into a bone- cell, has, so far as I am aware, never been followed out. It seems to me probable that the threads are preserved and lead to the develop- ment of the canaliculi, just as the cell-bodies produce the so-called lacunae. Whether threads of protoplasm run through the canaliculi in the mature bone or not is still under dispute. Beside the osteo- blasts in the interior of mandible there are others, Fig. 237, obV, which are derived from the cells of the periost, per, and although the peri- osteal cells are of a very different character from those of the mesen- chyma, mes, in the interior of the mandible, yet all the osteoblasts are alike. E. A. Schafer has directed attention to what he calls the osteogenetic fibres* Upon close observation of the growing spicules of the parietal bone the calcified parts appear granular, and from them Schafer finds that there run out for a little way soft and pliant bundles of transparent fibres. They exhibit a faint fibrillation and have been compared to bundles of white connective-tissue fibrils, with which in some situations they appear to be continuous. But although similar in chemical composition, they are somewhat differ- ent from these in appearance, having a stiffer aspect and straighter course, besides being less distinctly fibrillated. The fibres become calcified by the deposition within them of earthy salts in the form of minute globules, which produce a darkish granular opacity, until the interstices between the globules also become calcified, and the minute globules, becoming thus fused together, the bone again looks comparatively clear. It is stated that the fibrils themselves are not calcified, but the calcification affects the portion of matrix which unites them into the osteogenic fibres, so that these may be described as being calcified. The bundles of osteogenic fibres which prolong the bony spicules generally spread out from the end of each spicule so as to come in contact with those from adjacent spicules. When this happens, the innermost or proximal fibres frequently grow to- gether, while the other fibres partially intercross as they grow further into the membrane. The ossific process extends into the osteogenic fibres pari passu with their growth, and thus new bony spicules become continually formed by calcification of the groups or bundles * This account of the osteogenetic fibres is taken with some slight changes from Quain's "Anatomy," ninth edition. 410 THE FCETUS. of osteogenic fibres. The earthy deposit not only involves the osteo- genic fibres, but also the ground-substance of the tissue in which they lie. It occasionally appears in an isolated patch here and there on some of the osteogenic fibres in advance of the main area of ossi- fication. The osteogenic fibres become comparatively indistinct as they and the substance between them calcifies; they appear, how- ever, to persist in the form of decussating fibres, such as are seen in the adult bone, although in the embryonic bone their disposition is less lamellated, the bony matter having a somewhat coarsely reticular structure. Neoplastic Ossification. — When bone replaces degenerated car- tilage, the method of bone formation is essentially the same as when of ossification takes place in connective tissue, except for one fea- ture, namely, that the bone is first deposited against the calcified remnants of the cartilaginous matrix as soon as the cartilage cells have disappeared. A section through an ossifying long bone or ver- tebra. Fig. 238, presents a highly characteristic picture, and if the section be made as in the figure, parallel to the columns of cartilage cells, all the phases can be seen in a single successful preparation. The section figured was stained with Beale's carmine and alum haematoxylin, by which method not only are the cells and nuclei brought out clearly, but also the calcified cartilage is made deep blue, while the bone is deep red. In the upper part of the figure, C, the cartilage cells are just forming groups or columns, which a little lower down, C, are very distinct; lower down again, C", the cavi- ties, in which the columns of cartilage cells lie, have fused together into large spaces ; in these spaces the cartilage cells, c, are scattered in various stages of disintegration ; the adjacent spaces are separated from one another by partitions formed of ossified cartilaginous ma- trix,_ Ma, which appears a deep blue in marked contrast to the un- calcified matrix of the upper part of the figure, where the matrix is almost uncolored. The remnants of calcified matrix extend far be- low the cartilage. At the level indicated by the bracket, L, the new mesenchyma, mes, is found penetrating the spaces between the blue partitions. Ma; the mesenchyma is accompanied by blood-vessels, which are easily recognized, V, by their endothelial walls. Some of the invading mesenchymal cells lay themselves against the surfaces of the calcified partitions, become osteoblasts and produce bone, which thickens by additions to its surface. Thus the calcified remains of the cartilage become coated with bone, which in the preparation de- scribed has a rich red stain. As in the lower part of the figure, the development is more advanced, we find there the layer of bone, B, much thicker than nearer the cartilage. Fig. 339 is a very accurate drawmg of part of a section of a vertebra of a four-months' embryo so made that the columns of cartilage cells are cut at right angles; the level of the section corresponds to the lower part of bracket L, Fig. 238. The cartilage cells have disappeared and have been replaced by the invading mesenchyma ; the network of partitions formed by the remnants of the calcified matrix, Ma, of the cartilage is at once recognized, as can also be recognized the transformation of the cells mto osteoblasts, ohl, and the deposit of bone, B, upon the partition; noteworthy are also the osteoclasts, Osc, to which THE MESENCHYMAL TISSUES. 411 fuller reference is made in the following paragraph on the growth of bone. In the long bones the periosteal ossification has great importance, ("^^^ £-Si rr — Fig 238 —From a Section o( an Ossifying Vertebra of a Human Bmbi yo of four Months. Minot Collection, No. 35. C, C. C". Cartilage; C, region where the cells are beginning to form rows; C, region of the cells in columns; C", region where the cells are breaking down, and where the cell spaces are separated by calcified cartilaginous matrix. Ma, c, degenerated carti- lage cell ; B, layer of bone ; V, blood-vessels ; mes, ingrowing mesenchyma ; L, level of ossifica- tion. X 173 diams. and as it proceeds very rapidly at first in the central part of the bone, most of the shaft is formed from the periost— compare Quain's " Anat.," ninth edition, II., Fig. 109. 413 THE PCBTUS. We have learned that the development of bone may take place from embryonic connective fibrillar tissue (periost), or from cartilage, but whatever its origin, it has always nearly, if not quite, the same structure. This is true both of the cells and the matrix. Historical Note.— I have purposely abstained from attempting a full history of ossification. For full and comprehensive accounts I refer to Quain's "Anatomy," Ranvier's " Traite technique d'Histo- logie," KoUiker's " Gewebelehre," Krause's " Anatomie," etc. For a goodreviewof the literature up to 1858, seeH. Miiller, 58.2, and for Oic"- PiG. 239.— Section of a Vertebra of the same Embryo at right Angles to the Plane of Fig. 238, and corresponding in level to the lower part of me bracket i. Fig. 238. Osc^ Osc\ Osc'\ osteoclasts; S, Done; obl^ osteoblasts: Ma^ calcified matrix of cartilage. Stained with hsema- toxylin and eosine. X 356 diams. authorities see Ranvier's ■Gewebelehre," and Mas - references to the more important later "Traite," RoUett's chapter in Strieker's quelin's "Memoir." Growth, of Bone. — It is a well-known fact that the bones do not grow in the ordinary sense ; the bone cells cannot multiply ; the ap- parent growth of bone is accomplished by the destruction of the bone already formed and the production of new bone. The destruction of the bone is effected by means of large multinucleate cells. Fig. 239, Osc, which are derived from the mesenchymal cells, but just how is not clear. The cells in question have been named myeloplaxes (or myeloplacques), by Robin and French histologists, and osteoclasts THE MESENCHYMAL TISSUES. 413 (bone-destroyers) by Kolliker. They are frequently found against the surface of the bone, on cartilage, Fig. 239, Osc', and in that case lie in a little concavity formed by the eating away of the bone. As the development of these cells is not known and as their functions have been but little studied in the embryo, the detailed examination of their structure and history may be omitted here. Full accounts of the growth of bone may be found in all the standard histologies. Disappearance of Intercellular Substance. — In the adult there are various spaces in the mesenchymal tissues, which are in the natural condition filled with fluid, such as the so-called lymph spaces and lymph channels ; these spaces have no cellular walls. In the lymph glands also there is much fluid between the cells and retic- ulum of the gland. We must, therefore, assume that the intercel- lular substance has in some way been replaced, but whether it has been liquefied, or resorbed and fluid supplied in its stead, or simply cavities developed in it, we do not know. We can, therefore, do nothing more than note the gap in our knowledge. Hypertrophy of Intercellular Substance. — By this I do not mean the increase which occurs in connection with the development of fibriUsB, elastic network, or cartilage, but the hypertrophy of the clear homogeneous matrix of the young mesenchyma or embryonic connective tissue. Such an hypertrophy occurs in the amnion, in the young cutis, and elsewhere, and it is probably the most impor- tant factor in the histogenesis of the vitreous and aqueous humors ; as to how this hypertrophy is effected nothing is known. For the history of the vitreous humor see Chapter XXVIII. Blood-Vessels are the earliest of the mesenchymal tissues to be differentiated. Their history has already been given in full. See Chapter X. Lymphatic System consists of lymph spaces, lymphatic ves- sels, and lymph glands. The lymph spaces are merely channels in the intercellular substance, concerning the development of which nothing has been ascertained, and not much is known concerning the development of the vessels or glands. Lymphatic Vessels. — Kolliker ("Gewebelehre," 5te Aufl., 599- 600) states that in tails of tadpoles the lymph vessels can be seen de- veloping, in similar manner to the blood-vessels, by the hollowing out of mesenchymal cells. Klein has come to the same conclusion from the study of the development of lymphatics in serous mem- branes. According to Klein a vacuole is formed within one of the cells of the connective tissue, and becomes gradually larger, so as ultimately to produce a cavity fiUed with fluid, while the protoplasm of the cell thins out to form the wall around the cavity. He also adds that from this wall portions bud inward into the cavity, and detaching themselves become lymph corpuscles — but this history cannot be accepted without better foundation. To form vessels the vesicular cells become connected together. The protoplasmatic walls become multinucleate and are dijBferentiated into the lining endothe- lium. A. Budge's incompleted investigation of the development of the lymphatics in the chick, 87.1, was published posthumously by Professor W. His, and is an admirable piece of thorough work. The main part of the published memoir is devoted to the history of the 414 THE FCETUS. formation of the coelom by the fusion of a network of channels in the mesoderm, see p. 151. Budge states that after the coelom is devel- oped some of the channels are still found in the somatopleure, and represent the primitive lymphatics ; the somatopleure at this stage has no blood-vessels and the splanchnopleure no lymph-vessels. The primitive lymph-vessels communicate directly with the ccelom. Later on the ductus thoracicus appears and establishes the commu- nication between the lymphatics and the blood-vessels. Unfortunately the published paper contains no details about the development of the ductus. In a short note {Centralbl. Med. Wiss., 1881, No. 34) Budge has reported that in the allantois of a chick of eighteen to twenty days there are abundant lymphatics which can be injected with a subcutaneous syringe. The vessels accompanying the arteries, forming networks around them, Fig. 240, extend along the arte- rise umbilicales to enter the body and run along the aorta (see Budge, 87.1, 60, and Taf. VI., Fig. 2) as paired ducti, which are connected with one another by smaller cross stems, and unite in the upper part of the thorax into a single duct, which, however, again forks and has a double opening into the veins. The right umbilical lymph stem appears to atrophy later. In connection with the allantoic Ij'mphatics Budge has found (see His and Braune's Arch. f. Anat., 1882, 350) in chick embryos of ten to twenty days, lymph hearts, which lie in the angle between the pelvis and coccyx. Lymph- Glands. — Concerning the development of the glands I know of three papers, Sertoli, 66.1, Chievitz, 81.1, and a disserta- tion by Orth (Bonn, 1870) , which last I have not seen. Kolliker quotes alsoBreschet (" Le Systeme lymphatique, " Paris,1836, 185) andEngel {Prag. Viertelj., II., Ill, 1850) as maintaining that the glands arise each as a plexus of lymph vessels — a view which the observations of Sertoli have set aside. To study the early stages, glands must be chosen, the exact position of which in relation to other parts can be determined, in order that the condition of the tissue before the differ- entiation of the gland can be ascertained. With this in view Sertoli selected the mesenterial glands in cow embryos and Chievitz the same in the pig and the inguinal gland in man. Sertoli found in four-inch embryos fissures in the connective tissue of the mesentery where the glands were to appear ; in four-inch embryos these spots were further marked out by the crowding of nuclei around them. In six-and-a-half -inch embryos the anlages were pear-shaped, the pointed end being toward the radix mesenterii ; the pointed end alone contains lymph spaces, while the blunt end in which the nuclei are crowded is the anlage of the future cortex of the gland. Somewhat later the fibrous envelopes of the glands are differentiated, and as soon as their formation begins, the growth of the glands by accession from the surrounding mesenchyma ceases. Chievitz studied the Fig. 340.— Artery from the Allantois of a Chick, sur- rounded by a Network of Lymphatics. After Albrecht Budge. THE MESENCHYMAL TISSUES. 415 human inguinal gland; its anlage is clearly recognizable at about three months, and at three and one-half months the cortical portion with crowded nuclei can be distinguished from the medullary, in which there are spaces ; the gland is separated from the surrounding tissue by a fissure which is crossed by a few threads ; the fissure does not extend across the part of the gland corresponding to the future hilus ; the cells of the glands have large granular nuclei, and are easily distinguished from the lymphoid cells, which are much smaller with spherical refringent nuclei ; at first there are very few lymphoid cells, but they increase in number. Concerning the devel- opment of the reticulum, which Ranvier (" Traite technique," 678) has shown to be distinct in the mature glands 'from the branching cells, we have no information. Spleen. — Although the development of the spleen must offer many points of great interest, it has received very little attention. In 0. Hertwig's text-book no mention of the spleen is made; KoUiker, in both his text-books, dismisses the organ with a single brief para- graph. A little fuller is the notice by W. Miiller, in Strieker's "Handbuch der Gewebelehre," 260. Of special investigations there are three short ones, Peremeschko, 67.1, 2, and F. Maurer, 90.1, and the longer article on the spleen in fishes by E. Laguesse, 90. 1. The spleen is developed out of a mesenchymal anlage, which be- comes recognizable in the human embryo toward the end of the sec- ond month. In all amniota it is situated in the mesogastrium near the pancreas, and close to large arterial vessels. Its first differen- tiation appears to be due to an accumulation of rather large lymph- oid cells with large granular nuclei, and to the moving apart of the mesenchymal cells, which are much smaller than the lymphoid. Concerning the origin of the lymphoid cells we have only the obser- vations of F. Maurer, 90, who found in tadpoles, measuring from 6-8 mm. from mouth to anus, of the frog, Rana femporaria, that the entoderm gives off cells which pass into the mesenchyma and give rise to the first lymphoid cells. Maurer also obtained evidence that the same process occurs in the tailed amphibians. During the third month, in man (KSUiker), the blood-vessels penetrate the organ, which soon becomes rich in blood. W. Miiller states that the fur- ther development proceeds rapidly, so that in the human foetus of eight centimetres in length the various constituents are already differ- entiated. The cells lying beneath the peritoneal epithelium become elongated, and form fusiform nucleated bodies, and similar ones at an early period invest the larger vessels. From both small processes are given off which grow toward one another and represent the com- mencement of the trabecular system. Along the arterial branches denser accumulations of small nucleated cells may already be dis- cerned, which are conspicuous in tinted preparations by their deep color, and these form by far the chief constituent of the pulp. This consists of cells with from one to three nuclei and a delicate inter- cellular substance, forming plexuses, the interstices of which are constantly filled with blood-corpuscles. According to Peremeschko, there are now developed larger protoplasmic corpuscles in the tissue of the pulp containing from two to six nuclei, that are capable of performing amoeboid movements, and which, toward the end of 416 THE FOETUS. embryonic life, atrophy. In the further course of development the several constituents increase in volume, and a part of the fusiform cells of the capsule and the vascular sheaths develop into smooth muscular tissue. The arterial sheaths, containing numerous cells, are clearly distinguishable from the pulp, and from the middle of embryonic life the Malpighian corpuscles are recognizable. Concern- ing the size of the foetal spleen I know only of the statement by KoUiker," Grundriss," 380, that in man by the eighth week the anlage measures 0.62 x 0.31 mm., and in the third month 1.7 x 1.13 mm. In the embryo at six months the spleen already has its triangular form in outline ; the fibrous sheath or capsule, 0, is differentiated ; Fig. 241.— Section of the Spleen of a Human Embryo of six Months. C, Capsule; Hi, hilus; V V, blood-vessels. (The Embryo is Minot Collection, No. 8.) the hilus. Hi, is wide ; the main blood-vessels are remarkable for their size, and are encased in the sheaths of muscle fibres as in the adult ; the differentiation of the Malpighian corpuscles is indicated by the scattered areas, in which the cells are more crowded, which therefore appear darker in the stained specimen. In a thin section (0.01 mm.) of a somewhat younger spleen the reticulum of the spleen, the abundant blood capillaries, and the immense number of pulp- cells I find all well shown ; the pulp cells have round, finely granular nuclei with a very small amount of protoplasm ; I see also a much less number of smaller oval nuclei, which seem to belong to the reticulum. Laguesse's monograph, 90. 1, on the spleen of fishes is a conscien- tious and valuable work. The spleen appears late, some time after the pancreas, in the mesenchymal wall of the duodenum close to and on the left side of the insertion of the mesentery, and in close relation with the subintestinal vein. The anlage is first recognizable by the condensation of the tissue and the accumulation of free cells in its meshes. The developing spleen gradually comes into closer relations with the stomach and separated from the duodenum, and THE MESENCHYMAL TISSUES. 417 is ultimately situated in the mesogastrium. The origin of the free cells was not ascertained, but the author is inclined to trace them to the mesenchyma rather than to accept F. Maurer's view. They are small, have rounded granular nuclei (Laguesse's noyau d'origine) and very little protoplasm ; according to Laguesse they give rise some to leucocytes, others to red cells ; but in regard to this I think there is need of further evidence, for in other cases we know that leucocytes and red blood-cells {heviaties) have different origins. The network is produced rn situ by the mesenchymal cells, the processes of which gradually become more resistant, refringent, and homoge- neous, while the nuclei gradually disappear more or less completely. This confirms the view so long defended by Kolliker, as to the nature of the reticulum of the spleen. The cavities of the spleen form a rich network, which very soon enters into direct communication with branches which develop from the subintestinal (portal) vein, but the similar connection with the arteries is not established until later ; after the arteries have penetrated there is a circulation through the spleen and many of its free cells are carried off, but in places aside from the currents there remain accumulations of multiplying free cells ; such accumulations are found especially around the large arteries. The veins in the spleen consist only of an endothelium, but in the adiilt are in part encased in a sort of basement membrane formed by condensation of the spleen reticulum around the larger vessels. Smooth-Muscle Fibres. — That these are simply modified mes- enchymal cells seem to me no longer open to doubt, as explained in Chapter VI. on the mesoderm. This implies that the hypothesis so long upheld by His, that the muscles are genetically distinct from the connective-tissue elements, must be definitely laid aside. His classed the rauscles as archiblastic elements His' pupil, Erik Miiller, has sought in a special article, 88.1, to justify His' view, but the his- tory he gives is, that the inner mesothelium of the primitive segment breaks up into mesenchyma, and that some of these mesenchymal cells form the peri-endothelial walls of the aorta — a fact I can verify from my own observations on birds and mammals — but others of the cells, coming from the inner wall of the segment, form connective tissue, so that in this instance we have a proof of the identical mesenchymal origin of the two tissues. So also in the umbilical cord, it can be seen after the third month that the vessels are surrounded by smooth muscle cells, which gradually pass into mesenchymal cells proper ; as the muscular walls thicken with age it seems evident that the transi- tion represents an actual transformation of the connective-tissue cells into muscle cells, but the details of the process have still to be worked out. The earliest definite proof, known to me, that no line can be drawn between smooth muscle and connective tissue is that afforded by Flemming's observations, 78.2, on the bladder of salamanders, in which both tissues with all intermediate forms occur. Concerning the histogenetic transformation of mesenchyma into smooth muscle we possess no detailed or accurate information. Fat-cells first appear in the human embryo, it is said, about the fourteenth week, and after their first appearance gradually increase in size and number up to the time of birth, when, however, the fat 27 418 THE PCETUS. cells are still much smaller than in the adult. The fat cells are derived from the embryonic connective-tissue cells or mesenchyma, as has been demonstrated by Flemming, 71.1, 71.2, whose view was questioned by L; Eanvier (" Traite"), and the Hoggans, 79. 1. Ean- vier's observations were incomplete, in that he did not ascertain the origin of the cell which forms the fat-cells, as Flemming has pointed out in his reply, 79. 1, to the criticisms upon his work. The inves- tigations of the Hoggans appear to me untrustworthy. The fat cells are always developed in groups or clusters,^ and each cluster is supplied with an abundant network of blood capillaries. Fig. 342. —Fat Island from the Skin of a Human Embryo of five Months. Minot Collection, No. 95. Fe.. Blood-vessel; Jtfes, mesenchyma or embryonic connective tissue; F^ fat cells. X 310 diams. The fat cells always occur in the neighborhood of blood-vessels, so that one is almost compelled to conclude that superabundant food sup- ply is an essential condition of their development. Some interesting studies on the circulation in fat tissue have been published by J. Schobl, 85.1. The clusters of fat cells may be called fat islands, a term less likely to mislead students than that of fat globule, which has been used. Fig. 243 represents a section of a fat island in the embryonic cutis, drawn very exactly from the preparation, which had been stained with alum cochineal and eosine ; the mesenchymal cells, Mes, are scattered around and completely isolate the fat islands THE MESENCHYMAL TISSUES. 419 from one another ; the fat cells, F, form a group by themselves ; each cell has a large globule of fat surrounded by a thin layer of proto- plasm, which is thickened on one side, where the nucleus is situated ; the smaller the cell the more distinctly does the layer of protoplasm stand out ; the nuclei are compressed, smaller than those of the sur- rounding mesenchyma, and more darkly stained ; the difference be- tween the staining of the fat-cell and the other nuclei is exaggerated in the drawing. By their subsequent growth and expansion the fat islands may fuse together, thus forming a more or less continuous fatty layer. As regards the history of the single cells our knowledge rests chiefly on the admirable researches of Flemming, I.e. The cells lose their connections with one another and assume a somewhat rounded form, and the amount of protoplasm increases ; the nucleus comes to lie on one side of the cell either before the fat granules are developed or just as they are beginning to appear ; according as the nucleus is peripheral or central the fat is at first on one side or around the periphery of the cell. In either case the fat soon collects in one main globule, with other small ones about it in the protoplasm, and thus the condition of the young fat cells, as in Fig. 242, is attained. Soon after the nucleiis has been forced to one side by the fat the membrane of the cell appears. It is probable that the fat is accumulated within the cells before it becomes microscopically visible as granules, for Stolnikow {Arch. Anat. u. Physiol., Suppl., 1887, p. 1), has observed that the fat in the liver cells of frogs after phosphorus poisoning may be present in considerable quantities without appear- ing in granules. Upon this stress has been laid by Gaule, 90. 1, as indicating that the fat is bound to some other compound, perhaps leci- thin. This lends support to the suggestion of Poljakoff, 88.1, that the dull {" matten") granules, which appear in the protoplasm before or along with the first minute fat granules and disappear a"s the fat increases, are made up of fat combined with some albuminoid. The degeneration or regression of fat cells has been studied by Flemming and Poljakoff, but as the change does not occur before birth it does not fall within our scope, beyond noticing the suggestion that Ehrlich's Mastzellen (plasma cells) are regressive stages of fat cells. Pigment Cells. — Concerning their development in the embryo I knowof no exact investigation. What Goette gives, 75. 1, 531-522, is largely speculative. Flemming, 90.1, has shown that the pig- ment cells multiply by indirect division in salamander larvae, and that the scission of the protoplasm may be delayed. K. W. Zim- mermann, 90. 1, has given some further details. The divisional process offers several interesting features. The pigment granules which give color to the epidermis are not of epidermal origin, but arise in mesenchymal cells, which wander in from the underlying cutis. The source of the pigment was dis- covered by Aeby, 85.1, whose observations have been extended and confirmed by Kolliker, 87.2, 3, "Gewebelehre," 6teAufl., 302, List, 89.1, and Piersol, 90.2. Kodis, 89.1, on the other hand, has maintained that the pigment cells are formed in the epidermis and wander thence into the cutis, but Kodis fails, I think, to prove his point. In amniota the first pigment appears in small granular cells 430 THE PCETUS. in the basal layer of the epidermis (lizard, 40 mm. ; chick of ten days; cat, 47 mm.). These cells resemble leucocytes so much that Kodis has designated them as " leucocytoide Zellenj" they lie be- tween the true epidermal cells; the protoplasm is small in amotmt when the pigment begins to appear, but as the pigment increases the cell enlarges and passes from an apparently round to a distinctly stellate form. In mamiiiaJs the bodies of the cells are composed at first of clear, homogeneous, faintly granular protoplasm, in the midst of which sharply defined oval nuclei are seen ; in short, they resem- ble the cells of the underlying cutis and are probably immigrant mesenchymal cells. The earliest pigment particles are sparingly and irregularly distributed, but soon evince a tendency to aggregate about the nucleus, around which a brown wreath is soon formed. Subsequently pigment cells a^^pear also in the cutis and exhibit a strong tendency to collect beneath the epidermis and to form there rich networks. These cells send processes into the epithelium, to be followed often by the greater part of the cell ; it is thus that the pictures of immigrating pigment cells arise. As to the source of the pigment granules, they seem to be formed within the pigment cells and not to be taken up, as some writers have siiggested, as preformed particles from outside. It is possible that the pigment is connected genetically with the hsemoglobin, but of this there is no definite proof. For a discussion of the source of pigment granules see Maass {Arch. f. mtkrosk. ^naf., XXXIV., 453) and Piersol, I.e. Marrow. — The marrow of bone is derived from the mesenchyma, which, as above described, p. 410, enters the space left by the degen- erating cartilage; some of these mesenchymal cells become osteo- blasts, while the remainder produce the marrow of the future bone. The marrow has a verj'' complex structure in the adult, and numerous investigations upon its adult structure have been published. In these publications are scattered a good many observations on the foetal marrow, but as they have never been properly collated, and as there is no comprehensive research upon the development of the foetal marrow, I reluctantly forego the attempt to describe the histogenesis of the tissue — a subject which would certainly well repay competent thorough study. I will only add that the suggestion made by Ean- vier (" Traite technique, " 439), that the cells of the degenerating carti- lage produce marrow cells, cannot in my opinion be upheld, for it appears to me unquestionable that the cells of the cartilage are dis- integrated. Mesenchymal Cavities.— Under this head I do not include the blood-vessels, nor lymph-vessels, nor the lymph channels of the inter- cellular substance and lymph spaces of the lymphatic glands, but only those spaces which have, so to speak, passive functions, are filled with serous fluid, and are entirely bounded by mesenchyma. For example : the channels around the membranous labyrinth of the ear (compare the second division on the ear in Chapter XXVII.) , the subarachnoid space, the synovial and bursal cavities. These are probably all formed by the cells breaking apart, and are further characterized by the tendency of the layer of mesenchymal cells im- mediately round the cavity to become crowded until they form a dis- THE MESENCHYMAL TISSUES. 421 tinct lining endothelium. The degree to which this tendency is evinced varies extremely, and we may have the cells either simply somewhat crowded, or converted into an endothelium in patches, or wholly endothelium. The transition from one form of tissue to the other can be seen in the adult synovial cavities, and is important as additional evidence of the slight real difference between mesenchyma and epithelium. I know no observations on the development of the arachnoid spaces. Synovial and Bursal Cavities. — The development of the synovial cavities has been studied by Hagen-Torn, 82.1. Between the car- tilages of the limbs there is left undifferentiated mesenchyma, which very early acquires blood-vessels and shows later an increased vascu- larity. The formation of the cavity begins in the centre between the cartilages, and is first indicated by the tissue becoming less dense there (rabbit embryos 19-'20 mm.) ; some of the central cells undergo a mucoid degeneration and disappear, others become spindle-shaped and change into cartilage cells, with the result that the ends of the skeletal cartilages are now separated from one another only by a slight space. At the sides of the cavity the mesenchyma forms the synovial membrane, Avhich is merely very vascular, fibrillar con- nective tissue; upon the synovial surface patches of endothelium are developed. Villi, if formed at all, appear in later stages and al- ways at the sides of the cavity by the synovial membrane proper. Membranes. — The development of the various membranes and special mesenchymal layers, such as the submucosa, dermis, etc. , is considered in connection with the various organs, to which they be- long. There is one general feature which may be mentioned here, namely, the so-called basement membranes. By this term is now generally understood the layers of endothelioid cells found imme- diately underneath various epithelia ; for instance, under the ento- derm (epithelium) of the intestine, ' around the Graafian follicles of the ovary, around the seminiferous tubules, and the urinary tubules. These membranes, often designated as tunicse proprise, are undoubt- edly the product of the mesenchyma, though nothing is known of their development. They have the general morphological interest of demonstrating the tendency of the mesenchyma to revert to the epi- thelioid type. Ligaments and Tendons. — Both structures are modifications of fibrillse and elastic connective tissue. The tendons consist almost wholly of fibrillse running all in the same direction. The ligaments vary more, and may consist either of fibrillar or elastic tissue or both. The development of the ligaments has scarcely been studied; that of tendons has been investigated by L. Eanvier, 74.1, also his " Traite technique," 407 ; the regeneration and growth of the tendon tissue in the adult has been studied by several authorities— see A. Beltzow, 83.1. We learn, however, little beyond the fact that where tendon is to be formed the cells arrange themselves in rows, parallel with the length of the future tendon ; the fibrillee are devel- oped between the rows and parallel to them, and gradually increase until they occupy the entire space between the cells. By what stages the cells pass from the condition of simple mesenchyma to the sin- gular shapes of the adult tendon cells is unknown. CHAPTEK XX. THE SKELETON. The literature of the skeleton is very extensive as regards both its development and comparative anatomy. The ease with which skele- tons can be prepared and the importance of the hard parts to the palaeontologist has long given the skeleton a prominence in morpho- logical research far in excess of its importance as compared with the other systems. Athough the skeleton is in the mechanical sense the framework of the body, it is not so in the inorphological sense, because so far is it from being the framework upon which the body is built up, that its development is entirely subsidiary to the devel- opment of other systems, and is dominated by the arrangement of other tissues and organs, which have been formed and arranged be- fore the skeleton even begins to appear. In this chapter there is no attempt to give an exhaustive treatise upon the development, but by following the summaries given by KoUiker (" Entwickelungsgeschichte, " 2te Aufl., 401-503), Hertwig ("Lehrbuch," 3te Aufl., 492-543), and W K. Parker ("Morphology of the Skull "), and consultation of the more important original au- thorities, I have endeavored to write a comprehensive accotmt of the subject. Stages of the Skeleton.— We must distinguish between the stages of the skeleton as a whole, and the stages in the histogenesis of the bones. It must also be constantly borne in mind that the verte- brates have two morphologically distinct skeletons, the primary cartilaginous skeleton, which in the higher forms becomes partly ossified, and the secondary skeleton composed of dermal bones. 1. Notochordal Stage. — Permanent in amphioxus. In this stage the only skeleton is the axial rod of the notochord, and it is found to be the first stage in all vertebrate embryos. 2. The Membranous Stage. — The second stage in all true verte- brates and the permanent one in marsipobranchs. The mesenchyma is condensed around the notochord and strengthens thus the axis. 3. The Primary Cartilaginous Stage. — The principal parts of the primary skeleton are represented by separate cartilages. 4. The Completed Cartilaginous Stage. — In which aU the parts of the primary skeleton are present in the form of cartilages. No definite line can be drawn between this stage and the preceding, nor between it and the following. 5. Stage of the dermal skeleton, characterized by the develop- ment of sundry bones in the dermis. Dermal bones begin to develop before the cartilages ossify, and are present in cartilaginous fishes, hence they must be considered as older, and therefore belonging to an earlier stage, than the bones replacing cartilages. THE SKELETON. 433 6. Stage luith osseous primary skeleton, characterized by the primary cartilages being replaced by bone. The replacement is very gradual and never becomes complete; it begins in some of the cartilages before others are developed ; it is, accordingly, impossible to establish any definite limit in time for this stage. The most logical treatment would be to deal with these six stages in their natural sequence, but it has appeared to me more convenient to give the complete history of the notochord by itself (see p. 181), to add a section upon the membranous stage, and then to present the entire history of the primary skeleton under two main heads, the axial skeleton, p. 434, and the appendicular, p. 448; leaving the dermal skeleton till the last, p. 461, although it is ontogenetically and phylogenetically older than the osseous primary skeleton. The chapter closes, p. 465, with some general remarks on the morphology of the skull. Membranous Stage.— As we have already seen, the mesothe- lium of the inner side of the primitive segments produces the mes- enchymal cells, which invest the notochord and medullary canal. Recent writers have tended to regard this periaxial mesenchyma as segmented, and Van Wijhe even proposes to bestow the special name of sklerotome upon each of the mesenchymal segments. It is true that owing to its segmented origin the tissue does show for a time traces of metameric division, but the division becomes unrecog- nizable long before there is any mesenchymal skeleton indicated. The primary segmentation plays no immediate part in the develop- ment of the separate vertebrse. These considerations render it un- justifiable to regard the periaxial mesenchyma as segmented. We ought not to speak of sklerotomes unless we are prepared to speak of dermotomes, because the anlage of the dermal mesenchyma is as much segmented as the anlage of the periaxial mesenchyma. The ques- tion under consideration arose from a mistake of the older embry- ologists, who believed that the primitive segments were the direct anlages of the vertebrae, and accordingly named them protovertebrse (Urwirbel) ; unfortunately this misleading term is still in use. Then came the discovery that the true vertebrae are developed apparently between the primitive segments or in alternation with them. Re- mak formulated the hypothesis of resegmentation of the skele- ton {Neugliederung des Axenskelefs), which is wrong in assuming that the segmentation of the skeleton is not parallel with the primary segments, but is right in assuming that there is a primary segmen- tation of the skeleton, corresponding to the original mesothelial segments. Remak's conception has perpetuated itself to this day, and is carefully repeated in current text- books ; were it correct in its entirety then the membranous stage we are now considering would not occur. The first step toward the development of the perichordal skeleton is the fusion of the loose mesenchyma, derived from the segmented mesothelium, into a continuous mass of cells, which grow around the notochord and separate it first from the entoderm and later from the medullary canal, and grow around the medullary canal and close over it slowly, and also grow around the primitive aortae, see Figs. 161 and 103. This mesenchyma is of a loose but not quite uniform 424 THE FCETUS. character, and the cells early begin to condense in the immediate neighborhod of the notochord and nervous system. Around the notoohord the cells gradually become very closely crowded and form what is known in the lower vertebrates as the chorda sheath, in the amniote embryo as the investing mass, but in the amniota the uni- forfai continuous sheath exists only around the anterior end of the notochord where the investing mass participates in the formation of the cranium, while throughout the remainder of the embryo, as has been shown by A. Froriep,. the condensed mesenchymal an- lage is divided from the start more or less distinctly into separate vertebral masses, which in stained sections stand out conspicuous^. Froriep has studied the development of the vertebrae in the chick, 83. 1, and mammals (cow embryos), 86. 1. I. Axial Skeleton. Vertebral Column. — -As to how far forward the vertebral col- umn extends in the head we have no means of deciding positively, but as the occipital region of the skull is developed by the fusion of vertebrae, and as these vertebrae appear less and less distinctly as we pass forward from the neck, and as the number of occipital ver- tebras is greater in birds than in mammals, we cannot avoid the supposition that the number of vertebrae fused in the head was once greater than now appears in the mammalian embryo. There is accordingly much uncertainty as to the number of cephalic vertebrae. But though the number of vertebrae is not exactly known, we can fix the position of the cephalic end of the vertebral column, as coin- cident with the cephalic end of the notochord, which is close to the hypophysis or pituitary body. The notochord becomes invested almost up to its cephalic extremity by the condensed mesenchymal sheath, which is found in the occipital region, as in the body, to be the blastema out of which are differentiated the anlages of the verte- brae; it appears, therefore, no mere imagination to regard this as homologous with the vertebral column throughout, but with the development of the vertebrae inhibited entirely in the anterior, par- tially in the posterior occipital region. In front of the pituitary body the notochord and consequently the investing mass do not extend. We must in fact divide the head into a prae-pituitary unvertebrated and a post-pituitary vertebrated region. The remaining vertebrae to the end of the tail develop all much alike. They assume, how- ever, modified forms in the various regions, but the origin in the embryo of the differences between the cervical, dorsal, and lumbar vertebrae has never been worked out. Special modifications of the first and second cervical vertebrae take place in mammals to form the atlas and epistropheus or axis, in the five sacral vertebrae to form the sacrum, and in the caudal vertebrae to form the coccyx. Typical Development of a Vertebra.— Our exact knowl- edge rests mainly upon the investigations of August Froriep, 83. 1, 86. 1, on chick and cow embryos. The investing mass or condensed perichordal mesenchyma forms a continuous sheath around the noto- chord. At a point corresponding to the centre of each mesodermic segment, or a little on the cephalic side of each segment, the investing AXIAL SKELETON. 425 mass becomes thicker in diameter and its tissue more condensed; the condensation is very noticeable in stained sections and is the first sign of the vertebral formation; the condensation spreads rapidly, extending sideways, upward, and backward with the result of forming a bow of dense mesenchyma, the primitive vertebral bow ( Wirbel- bogen) of Froriep. These bows are distinct from the bodies of the vertebrae, which arise later from separate anlages. The bows pass on the ventral side of the notochord, and thence arch on each side, Fig. 343, tailward and dorsalward, so as to end at the caudal edge of the muscle plate of the segment to which they belong, and ending, therefore, just in front of the intersegmental arterj% v, and of the spinal nerve, N, from the sensory ganglion of the next following seg- ment. We see here that the vertebrae are strictly segmental struc- tures and not intersegmental as' has been commonly assumed since nch Md Fig. 943. — Reconsti'uctiun of the Last Occip- ital, and First Two Cervical Vertebrae of a Cow Embryo of 8.8 mm., the notochord and axis being assumed to be straight. ?7f7i, Notochord ; s, sheath of notochord : 6, bow of -occipital vertebra: u, segmental artery; N, nerve ; My, myotome. After A. Froriep. X 33 diams. Fig. 244. —Cross-Section of the Anlage of Second Cervical Vertebra of a Cow Embryo of 8.8 mm. J/c/, Medullary canal ; G/, ganglion of the second cervical nerve ; Jl/it, muscle plate of the second cervical segment (Froriep's first cervical muscle plate) ; Jvc/i, notochord; Vert, anlage of the vertebra ; Ao, aorta. After Fro- riep. Remak. The course of the bow, as compared with the transverse plane of the body of the embryo, is oblique, so that while the centre of the bow next the notochord is near the centre of the segment, the tips of the bow lie at the caudal limit of the segment and ultimately separate the muscle plate of its own segment from that of the next following. The obliquity of the bow appears to me to be deter- mined primarily bj'' the arrangement of the spinal ganglia, the dor- sal ends of which fill out the width of the segment, while the lower pointed end is carried forward to the anterior border of the segment ; this disposition leaves the caudal side of the segment free for the mes- enchyma and the differentiation of the vertebral bow ; the obliquity is further assisted by the form of the muscle plate, as can be seen in Fig. 243. The portion of the bow underneath the chorda in the median line is termed the hypochordal brace (Spange) and in its ultiinate development differs considerably from the rest of the bow. The investing mass around the notochord on the caudal side of the 436 THE PCETUS. bow and above it becomes later the anlage of the body of the verte- bra. The vertebral bow may be regarded as the primitive stage; it is found in the chick from the middle of the fourth to the middle of the fifth day; in cow embryos of 7-11 mm. The vertebral bow is destined to form the processes of the verte- bra, and the manner in which its ends spread out against the muscle plate can be weU seen in a cross-section, Fig. 244. At the time the bow is differentiated the muscle plate has become protuberant toward the notochord, and when the dense mesenchyma forming the bow spreads out it is forced by the muscle plate to grow dorsalward, and ventralward, and thereby to become, as it were, branched ; the dorsal branch is the anlage of the neural arch; the ventral branch the anlage of the transverse or costal process, because it grows out still farther to form the anlage of the rib. There follows a transitional state which is characterized by the gradual development of the cartilaginous vertebra. This stage ex- tends in the chick from the middle of the fifth to the middle of the sixth day, and is found in cow embryos of 13-17 mm. The noto- chord exhibits signs of retrogressive change, and is contracted at the level of the vertebral bow. The part of the investing mass (peri- chordal mesenchyma) immediately over the centre of the bow or hypochordal brace becomes the anlage of the intervertebral liga- ment, its cells becoming first less crowded and then acquiring an elongated form; out of this anlage the adult ligament is slowlj- differentiated, chiefly by the development of connective-tissue fibril- lae. The investing mass behind the hypochordal brace develops into- the cartilaginous body of the vertebra, in the mammal before, in the bird after, cartilage begins to appear in the vertebral bow. In the- mammals there are two centres of chondrification, which may be recognized in the bird also, although they are in the latter connected with one another under the chorda. The process of chondrification continues until out of the investing mass the separate vertebral body is differentiated. Meanwhile the chondrification goes on in the ver- tebral bow, and in birds the whole bow is converted into cartilage- and unites with the body to form the completed vertebra. In mam- mals except in the occipital and anterior cervical vertebrse the cen- tral part does not form cartilage but remains as a dense mesenchymal band, which can be recognized as a more or less distinct structure for some time, but is ultimately lost in the substance of the inter- vertebral ligament. A median longitudinal section of a cow embryo a little more advanced, Fig. 245, shows the persistence of the hypo- chordal brace. The permanent stage is reached by the fusion of the cartilage of the bow with that of the body, which may be said to be completed in the chick by the middle of the seventh day, and in cow embryos- of 22 mm. In the chick the whole bow is differentiated into carti- lage, and its central part fuses with the vertebral body. In mammals this fusion does not take place except in the occiput, but the two ends of each bow become cartilaginous and fuse with the cor- responding vertebral body, except in the case of the first cervical vertebra, see p. 430. The central portion of the bow in all vertebrse below the first cervical disappears and is lost in the intervertebral AXIAL SKELETON. 427 ligament. In a longitudinal section, Fig. 245, it can be seen that the first bow is a well -developed cartilaginous piece, 8c, while the second, is only partially chondrified, while the third and fourth are almost lost in the intervertebral ligament. The first bow, as just stated, forms the atlas. During the development of the cartilage the vertebra continues growing and the arches extend farther from the body ; the neural arches |.\ \, ncii Vrt b gradually ' -^ close over the medulla ry canal, the closure tak- ing place much earlier in the chick than in the mammal In the human embryo the neural arches extend at eight weeks only a short distance up the side of the spinal cord , at three months they have come in contact on the dor- sal side in the dorsal region, but are still quite far apart m the lumbar and sacral regions (KoUiker, " Grundriss, " 191) and by the fourth month all the arches have closed The development of the spinous process needs to be fur- ther investigated. The ventral pro- cesses, Fig. 244, spread downward and are brought, owing to the primitive inclination of the vertebral bow, to the caudal boundary of the segment to which they belong, and as they lie at the caudal edge of the muscle plate of their respective segments, they con- tribute to separate that plate from the next following. These processes lose their continuity with the vertebra proper, but remain connected with it by ligaments ; they thus become the independent anlages of the ribs, where true ribs are developed. Another point deserving attention is the relation of the vertebrae to the vertebral artery which arises, as described in Chapter XXIV., as a series of longitudinal anastomoses between the intersegmental ar- teries; the vertebral artery begins to appear in cow embryos of 12 mm., and is a continuous stem in those of 15 mm. The vessel form- ing the anastomosis grows through the mass of the vertebral bow during the transitional stage, while the mesenchyma is not very dense at the point penetrated by the artery. This discovery, which we owe to Froriep, sets aside the statement, which has become traditional, that the developing vertebra grows around the artery, and shows instead that the artery grows through the developing anlage of the Fig. 345. —Longitudinal Median Sec- tion of the Upper Portion of the Verte- bral Column of a Cow Embryo of 22.5 mm. Nch, notochord ; Oc^ occipital car- tilage; Art.b, arteria basilaris; Eps, Eps^. C3, C, bodies of the first and sec- ond cervical vertebrae; C\ C", bodies of third and fourth vertebrae ; sc, anlage of Atlas. After Froriep. 428 THE FCETUS. vertebra. The artery, by its position, may be said to mark approx- imately the boundary between the neural and costal processes of the vertebra. The ossification of the vertebrae does not alter the morphology of the cartilaginous stage, and it is doubtful whether it is accom- panied by any noteworthy change in the form of the single skeletal pieces. The ossification begins with two centres, one in each neural arch, and is continued by a third centre in the body of the vertebra. The centres in the neural arches lie near the body proper ; that of the body appears in man about the seventh week. The centres of ossification of the body become recognizable first in the dorsal region, and from there their differentiation progresses successively from vertebra to vertebra, both headward and tailward. The centre is situated at first on the dorsal side of the chorda (Robin), but as the centre extends it incloses the notochord, which is gradually obliterated so that it can no longer be distinguished after the actual formation of bone has commenced. The progress of ossification is very slow; thus the preliminary degeneration covers the period, in cow embryos, in which their length increases from 2.2 to CO cm., and it is not until the latter length has been attained that the actual deposit of bone begins (Froriep, 86. 1, 130) . In man the centres do not attain the surface of the cartilage until the fourth or fifth month. Ultimately * the three deposits of bone fuse into a single osseous vertebra, but for a long period before this cartilage remains between the bony arches and the bony body, and on the dorsal side between the arches; these cartilaginous areas act as growing zones. The epiphyses are separate centres of ossification, which appear one on the cranial side, one on the caudal side of the body of each vertebra, but not until after birth. The development of the epiphyses and their fusion with the main body have been investigated by Schwegel, 58.1. To complete the adult bony vertebra there are five centres of ossification requisite. Summary. — Every vertebra is developed within the limits of a single segment, that is, out of the mesenchyma produced from the inner wall of a single segment. This point is especially important because it is commonly stated that each vertebra is derived from ad- jacent parts of two segments. Each' vertebra has two distinct parts, the vertebral bow {Wirhelbogen) and the vertebral body {Wirbel- korper) ; both parts in their first stage consist of condensed mesen- chymal tissue. The bow appears first and is an arched band of tissue passing under the notochord, thence running obliquely backward and terminating on the caudal side of the muscle plate of the seg- ment. The body appears later in each segment just behind the me- dian part of the bow. The bow and the body both chondrify and fuse with one another, except in the first cervical segment ; in birds the whole bow becomes cartilaginous, but in mammals the middle part of the bow atrophies, except in the first cervical segment. The lateral portions of the bow form both the neural and costal arches ; the distal parts of the latter separate from the vertebra proper to form the an- lages of the ribs. The morphology of the vertebral column is com- * During the first year after birth the arches unite dorsally, between the third and eighth year the arches unite with the body. AXIAL SKELETON. 439 pletely determined while it is in the cartilaginous stage ; ossification is merelj^ a supplementary process and produces no important change in the form or anatomical relations of the vertebrfe. Froriep's discovery that the vertebral bow and body are distinct pieces must be considered very important, and at once suggests com- parison with those palaeozoic reptiles in which centra and intercentra have been distinguished in the vertebral columns, but this compari- son has yet to be worked out. For a general paper on the intercen- trum see Cope, 86.4, also G. Baur, 86.1, for a discussion of the morphogeny of vertebrae from the j)alseontological point of view. Evolution of Vertebrse. — We have no positive knowledge, nor even valuable theories, as to the causes which first led to the evolution of vertebrae, though unscientific hj-potheses have been abundant. There is one important consideration which has been rather neglected, though almost self-evident, namely, that vertebrae have arisen within the vertebrate series, the perichordal mesenchyma in the lowest vertebrates not being divided into vertebrae, there being, in short, so-called vertebrates without vertebrae. As the higher fishes have vertebrae, it is evident that the vertebral column was evolved within the class of fishes. The embryological development of the vertebrae indicates that thej' are compound bodies, as above shown. We are thus led to distin- guish four stages in the differentiation of the axial skeleton ; 1. Notochordal stage. 2. Perichordal stage. 3. Froriep's stage (vertebral bow and centre not united). 4. Vertebral stage (vertebral bow and centre united). The first stage is permanent in Amphioxus ; the second is perma- nent in Petromyzon ; the third will perhaps be found permanent in Chimaera ; the fourth is permanent in Amphibia and Amniota. The skull may be looked upon as in part a modification of the second stage in the head region. Occipital Vertebrse. — The occipital bone of the adult is the final outcome of the fusion and ossification of an uncertain number of vertebrae. The investing mass of the cephalic portion of the notochord forms the anlage of the occipital skeleton. This anlage terminates a short distance behind the hypophysis. In birds and mammalia it may be divided into two regions, comprising each about half the length of the anlage ; the anterior or pituitary half does not offer, even in the earliest embryonic stages, so far as known, any trace of division into separate vertebral masses ; the posterior or cervical half does show clear division at an early stage into four vertebrae (in the chick into five vertebra) , but of these only the last appears as a perfectly distinct, well-differentiated vertebrae, but even this vertebra, when its chondrification begins, merges into the gen- eral occipital mass (A. Froriep, 83. 1, 86. 1). The vertebrae of the mammalian occiput correspond to four segments, of which the hypoglossus represents the nerves. Fig. 246 is a frontal projection of the cephalic end of the mesenchymal vertebral column of an em- bryo, 15.5 mm. long, from a cow. The nerves, N, mark the divis- ions between the vertebrae, as do also the intersegmental arteries, vj the anterior vertebrae are already fused, Oc, but the fourth is 430 THE FCETUS. Fig. 246. —Frontal Projection of the Cephalic Part of a Vertebral Column of a Cow Embryo, 15. 5 mm. long. perfectly differentiated and closely similar to the succeeding ver- tebrae. Three hypoglossal nerves traverse the occipital anlage. In embryos of 18.5 mm. the occipital vertebra is found to have fused with the occipital mass, though the ends of its vertebral bow project enough to still indicate the original division of which all trace is lost in slightly older embryos. In the occipital mass chondrification begins on each side of the notochord, nch, just as it does in the bodies of the indi- vidual vertebrae, and it begins before the fourth vertebra (Froriep's occipital ver- tebra) unites with those in front. The result of the chondrification is to produce two bars of cartilage which extend along- side the occipital notochord, but of course, as the histogenetic change spreads, the cartilage unites and finally extends through the entire anlage. The bars of cartilage are known as the parachordals, and are commonly, but erroneously, described as the primitive anlage of the occipital cra- nium, whereas in reality they indicate only the growth of the cen- tres of chondrification in the anlage. I can recognize no grounds at present for assigning any special morphological meaning to the parachordals. Atlas and-Epistroplieus. — The first and second cervical verte- brae undergo remarkable modifications, which are established during the transitional stage of the vertebrae — in other words, while the ver- tebral anlages are chondrifying. In mammals the first cervical vertebra develops two cartilages, one of which is formed out of the whole vertebral bow and gives rise to the atlas, and the other is formed out of the vertebral body. The later cartilage fuses with the second vertebra and with it forms the epistropheus or axis. Our precise knowledge of the development of these two vertebrae rests principally upon the admirable researches of A. Froriep, 83.1, 86.1, though previous investigators had established that the first vertebra forms the so-called odontoid process of the epistropheus, see Ch. Robin, 64.1, and C. Hasse, 73.1. In birds, but not in mam- mals, the central portion of the vertebral bow of the second cervical segment also contributes to the formation of the epistropheus; in mammals it disappears or is merged in the intervertebral ligament. Owing to this differences the atlanto-epistrophic articulation is not strictly homologous in the two classes, being formed in birds by the vertebral bow of the second segment ; in mammals by the expanded caudal part of the vertebral body of the first segment of the neck. The specialization of the two vertebrae begins when their chondrifi- cation, is well advanced (cow embryos, 17-18 mm.), for we see then that the whole of the first vertebral bow is changing into cartilage to form the atlas, and that it does not grow together with the body. Meanwhile in mammals the body of the first vertebra is changing form, its cephalic end becoming conical to make the anlage of the AXIAL SKELETON. 431 odontoid process, and the caudal part broadening out, and making a shoulder laterally and ventrally around the base of the odontoid process ; this shoulder forms the articulation with the atlas. The expansion of the first vertebral body forces the vertebral artery and the second cervical nerve out laterally ; the bend of the artery thus produced is permanent ; the expansion also brings the first body into contact with the bases of the transverse processes of the second vertebra; the intervertebral tissue (ligament) between them dis- appearing ; the two vertebrae unite by their two points of contact, and thereafter their fusion progresses toward the median line, until all the tissue of the intervertebral ligament is obliterated and the two cartilages have fused into one, the epistropheus. The atlas ossifies from three centres, two of which correspond to and appear about the same time as those of the neighboring vertebral bows (neural arches), while the third does not appear until after birth, and is situated in the middle of the ventral arch of the atlas (corresponding to the primitive hypochordal brace, Froriep's Spange). Often there is also a separate centre for the spinous process. The two primitive centres unite on the dorsal side during the third year, and with the ventral centre in the fifth to sixth year. The epistropheus, in accordance with its development, has four centres, one for the body of its first vertebra or the odontoid pro- cess, one for its own body, and two for its neural arches. The two first-named centres appear during the fourth or fifth month. Tho fusion of the centres may not be completed until the sixth or seventh year, and up to that age the tip of the odontoid process remains unossified. Sacral Vertebrse. — In man there are five vertebras, characterized by their peculiar form and by their articulation with the pelvis, and which begin at eighteen years to slowly unite into a single bone known in anatomy as the os sacrum. In other animals, however, the sacrum is not formed out of the same vertebrae, if we count from the last cervical vertebra, nor out of the same number of vertebrae. Various attempts have been made to explain these divergences — see especially Rosenberg, 76.1 — but no certain result has yet been reached. Of the history of these vertebrae we have no such exact knowledge as Froriep's researches have given us concerning the cervical vertebrae. The processes form neural arches and lateral processes (Seiten- fortsdtze) which were commonly homologized with the costal pro- cesses of other vertebrae, principally upon comparative-anatomical grounds. The chief embryological evidence in favor of this homology was the fact that the lateral processes have a separate centre of ossi- fication, making, together with the three usual centres, five primary centres for each sacral vertebra. In 1875 Rosenberg, 75.1, showed that the anlages of the sacral ribs can be seen in human embryos, and that the proximal ends of these change into cartilage and fuse with the true transverse processes of the vertebrae— very much as happens with the thirteenth rib in man. Coccygeal and Caudal Vertebrse. — Behind the sacrum there are nine segments to be found in the human embryo of 8-9 mm., as discovered by H. Fol, 85.1. From the sacrum tailward they are 432 THE FCETUS. found progressively more and more rudimentary, and only from three to five of the segments immediately following the sacrum devel- oped ossified vertebrae. These are the so-called coccygeal vertebrae, concerning the embryology of which we know nothing. It is prob- able that some of the segments behind the coccyx form at least mesenchymal, if not cartilaginous, vertebrae, and Fol's observations suggest that the last coccygeal vertebra is really the product of the fusion of several caudal vertebrae. Only the first coccygeal vertebra begins to ossify before birth. This, the thirtieth vertebra, has been shown by E. Rosenberg, 76.1, to be in the embryo really a sacral vertebra, but it separates in the course of development from the sacrum, and becomes the first of the coccygeal series. Kibs and Stermim. — The ribs and sternum are vertebral struc- tures, and therefore strictly segmental. This statement seems to me an unavoidable deduction from Froriep's observations on the devel- opment of the costal processes of the vertebrae, but it is directly opposed to the conception current among raorphologists, according to which the ribs are intersegmental. That the sternum is a mor- phological product of the ribs is, I believe, the accepted opinion of both comparative anatomists and embryologists. That it is so in man has been put beyond doubt by G. Ruge's investigations, 80. 1, see also C. K. Hofmann, 80. 1. 1. Ribs. — Comparative anatomy renders it probable that every vertebra had ribs primitively, and most of them have still in the human embryo the anlages of ribs. In man there are only twelve vertebrae (eighth to nineteenth) of which the costal anlages are repre- sented in the adult by true ribs; traces of a thirteenth pair of ribs belonging to the twentieth vertebra appear in the human embryo, and as a. rare anomaly the thirteenth pair occurs in the adult. In the cervical region there are found costal processes of the vertebrae, also in the lumbar and sacral region ; in the last-named region the processes acquire a certain independence, but soon lose it and fuse with the vertebrtB proper. These variations should be borne in mind while reading the foUovv^ing paragraph, which attempts to give the general history of a typical rib. The ends of the vertebral bows grow out as shown by Froriep, 86. 1, until they corns in contact with the muscle plates of their own seg- ments. By the bulging of the plate the end of the bow is forced to expand dorso-ventrally, and there is thus given the primary divi- sion into dorsal or neural, and ventral or costal process. The spinal ganglion forces the end of the bow, compare Fig. 243, p. 425, to grow toward the posterior limit of the segment, and this permits the costal process to grow out past the caudal edge of the muscle plate and to there become the anlage of the rib, which is not therefore an inter- segmental structure, as current tradition has it, but truly segmental ; the rib and the mj'-otome headward of it belong to the same som- ite, and the rib owes its apparently intersegmental position to its situation at the caudal limit of the segment, behind the muscular an- lage. Whether the costal anlage is produced as an actual outgrowth of the condensed mesenchyma of the vertebral blastema or by differ- entiation of the mesenchyma in loco, we do not know ; nor do we AXIAL SKELETON. 433 know what limits the rib in the transverse plane so that it is merely a rod and not a wide and high partition wall. In this stage the rib is directly continuous with the vertebra, but when by changing into cartilage it passes into the next stage, it separates f^m the ver- tebra by the development of a fibrous ligament, forming le primary articulation between the rib and spinal column. The division takes place obliquely, thus allowing the head of the rib to come in contact with the body of the vertebra, and to articulate, by its dorsal sur- face, with the ventral surface of the future transverse process. In the course of its further development the single primitive articulation becomes divided and the secondary, or adult condition, is established with one articulation with the transverse process, and a second with the body of the vertebra. In the case of the ribs, which become rudimentary, the development ceases at this stage, and only the proximal end of the rib chondrifies ; the small remnant of cartilage unites with the transverse process of the vertebra, re-establishing by a secondary union the primary connection. The true ribs, as those belonging to the dorsal vertebrae of mam- mals are called, extend a considerable distance through the somato- pleure toward the median ventral line, but, as discovered by H. Rathke, 38.2, 365, before they reach the middle ventral line the ribs produce the anlages of the sternum, and of the intercostal ligament, at first as condensed mesenchyma, which afterward becomes histolog- ically differentiated — see the next section on the sternum. The ribs extend to unequal distances, the first coming nearest the ventral line, the last terminating farthest from it. In the human embryo of from 2 to 3 cm. there is present a thirteenth true rib (Rosen- berg, 75. 1, 89-91) ; the proximal end chondrifies and fuses with the vertebra. This valuable observation shows that the so-called first lumbar vertebra of man is really the last dorsal vertebra, and in its embryonic stage is strictly comparable with the thirteenth dorsal vertebra of Troglodytes. As in Hylobates the twenty-first vertebra sometimes has ribs, the evidence within the primates suffices to prove that the lumbar region was evolved at the expense of the dorsal. The ribs are only partly ossified, hence the osseous rib of the adult represents only a portion of the whole primitive rib, the most distal part of which has been reserved to contribute to the sternum (or intercostal ligament), and another part of which remains in the cartilaginous stage to unite the costal bone with the sternum or intercostal ligament. Each primitive rib is therefore divided into three parts: 1, the proximal bony division, the rib of human anat- omists; 2, the middle cartilaginous division, the costal cartilage; 3, the distal sternal or ligamentous division. By the differentiation of fibrillar tissue out of the original costal anlage articulations are developed for the costal cartilages at their proximal ends with the bony ribs, and at their distal ends with the sternum. The exact history of these differentiations has still to be worked out. The ossification of the ribs begins during the second month, ac- cording to KoUiker, and there is but a single centre. Schwegel, 58.1, states that epiphyseal centres appear eight to fourteen years after birth in the head and tubercle, that is, for both vertebral artic- 38 434 THE FCETUS. ulations; the epiphyses do not unite with the main bone until later; often not until the twenty-fifth year. 2. Sternum. — The breast bone is developed from the ends of the ribs, but the early stages have still to be ascertained by following out the relations while the anlages are in the mesenchymal stages. Hitherto investigations have begun only with the cartilaginous stage. It seems probable that the costal anlages grow beyond the ventral limits of the muscle plates and then bend headward, and by uniting, form a longitudinal sternal anlage on each side at some lit- tle distance from the median line. The cartilaginous half-sternum apears in rabbits the seventeenth day ; they are still separate in chicks of the eighth day, in pig embryos of about 27 mm. in human em- bryos of 24 mm. In the chick the halves are uniting during the seventh day, and in pig embryos of about 50 mm., the halves are fully united. The sternal anlages (Ruge's Sternalleisten) arise from the ends of the first to seventh ribs, and accordingly are nearest together toward the head and diverge tailward. My own observations lead me to think it probable that the connection reaUy extends to all the ribs, but between the seventh and twelfth ribs it becomes fibrillar, and gives rise to the intercostal ligament, which, therefore, is morphologically the prolongation of the sternum. The sternal halves gradually coalesce, beginning at their upper ends. In many mammals the sternum shows plainly its metameric origin and consists of separate pieces metamerically arranged, and there is a separate centre of ossification for each piece. In man, on the con- trary, the originally continuous cartilage forms three pieces, the uppermost of which belongs only to the first sternal segment or first pair of ribs according to G. Euge, 80. 1, but, according to W. K. Parker, also is formed partly at the expense of the aborted last cer- vical rib ; the middle piece corresponding to the second to seventh segment; and the third piece, which remains chiefly or wholly car- tilaginous. The first piece is the manubrium, the second piece is the body of the sternum, and the third piece is the ensiform or xiphoid cartilage. G. Ruge, 80.1, found in human embryos two small suprasternal cartilages which fuse with one another and then with the manubrium ; the significance of these cartilages is uncertain. The sternum ossifies with one centre in the manubrium, and in man with an irregular number of centres in the body. Its ossifica- tion does not begin until the sixth month. The double origin of the sternum and its dependence upon the ribs was discovered by H. Eathke, 38.2, 363. This discovery was con- firmed and extended twenty years later by W. K. Parker, 58.S, and more recently by A. Goette, Hofmann, 80.1, and G. Ruge, 80.1; the last is an admirable investigation of the development of the sternum in man. Trabeculae Cranii. — H. Rathke discovered that at the same time -that the cartilaginous tissue develops in the occipital skeleton there appear two curved bars of cartilage in front of the notochord. These cartilages by their fusion and expansion form the whole of the prae- chordal chondrocranium, and were named by Eathke the trabeculae cranii. All subsequent writers have made Rathke's discovery the starting-point of their accounts of the development of the anterior AXIAL SKELETON. 435 part of the skull. But the morphological differentiation of the skel- eton, as we have already seen in the case of the vertebrae, etc. , is given by condensed mesenchyma, and the cartilage, when it first appears and for a considerable period afterward, does not by any means cor- respond to the real shape of the skeletal piece. Now nearly all the information we possess as to the early stages of the skull is concern- ing the progress of the so-called chondrocranium, and since this is really for a considerable period merely the history of the progress of chondrification in the already formed mesenchymal skeleton of the cranium, it results that concerning the early stages of the skull we have almost no available information, nor can we hope to understand the morphology of the skull until its developmental history through the mesenchymal stages shall have been followed, as has been that of the cervical vertebrse by Froriep. Concerning the history of the car- tilage of the skull we possess an immense fund of information, owing chiefly to the long series of splendid monographs by W. Kitchen Par- ker (1866-1886), the chief results of which have been summed up by himself and Mr. Bettany in a single comprehensive volume, 77. 1. From what has been said it is clear that the shape of the prse- chordal cartilaginous skull has very little morphological significance until the mesenchymal skull is completely chondrified ; until then the growth of the cartilage represents merely the advance of a histological modification within the skeletal piece. Unfortunately it is impossible at present to say when the cartilage does begin to represent the shape of the cranium. As the history of the early stages of the prse-chordal cartilage has very little morphological value it may be very briefly given. The trabeculsB cranii of the pig* may be taken as typical representatives of the mammalian trabeculse, and show essentially the same arrangement as are found in all other vertebrates, although the form and proportions vary from class to class. In pig embryos of about 16 mm., the trabeculae cranii appear as two curving rods of cartilage, united in front, but separated behind ; in general shape they resemble calipers; they lie anteriorly between the olfactory pits fio. 247. -Embryo pig ot about 16 ■,,,■,•'. T» I, j_iIj. mm. Cranial elements seen from be- and the brain, and form from the start low, if, olfactory pit; iv, trabecuia; a skeletal partition between these two f?i' p ar",^?£^7T tloih^sisf ^^; structures. As shown in Fig. 347, Tr, ^^ '^^^,^^'J^i^, TI, the trabeculse separate posteriorly some notochord; 7, facial nerve? 9, giosso- distance in front of the hypophysis,_B; g^S.^'=^'' ^°' ™^^- ^'"^^ ^- ^• then curving toward the median line, taper and end in points immediately behind the hypophysis. The anterior end of each trabecula is bent over outward and down- ward, forming the cornu trabeculce, and causing a projection on * I follow the account given in Parker and Bettany's "Morphology of theSlniU," chapter viii. 436 THE PCETUS. per Fig. 248. — Embryo Pig, One and one-third Inch long ; Median Longitudinal Section of the Head; the nasal septum and brain have been removed. After W. K. Parker. (For explanation of lettering see text. ) either side of the palate in the mouth cavity behind the olfactory- pits. These pits are situated entirely in front of the trabeculae at this stage, but between them there is an internasal septum of mes- enchyma, and into this septum there already extend two cartilagi- nous laminae which are the prolongations of the trabeculse. In the course of their further development the trabeculse fuse throughout their entire extent. In pigs one inch long the internasal cartilages have nearly or quite fused into a single median piece, and the trabe- culse proper are united also except around the hypophy- sis, which they closely em- brace. At this stage we see further that the trabecular cartilage is extending side- ways, outward and upward around the brain, outward and downward around the olfactory pits. In embryos an inch and a third long the posterior ends of the trabeculse have united with the anterior end of the occipital car- tilage, thus forming a continuous floor of cartilage, which under- lies the brain, and in front overlies the olfactory pits, and has also odevelped under the hypophysis, which thus becomes definitely sepa- rated from the mouth cavity and inclosed within the brain case. We find at this stage also that the cartilaginous periotic capsules have begun to fuse with the lateral portions of the occipital cartilage, thus making one continuous skeletal piece, which is known as the primitive chondrocranium, but it does not correspond to the real cra- nium at this stage, for beyond the limits of the cartilage the skeleton around the brain and olfactory pits is already formed as condensed mesenchyma. The general arrangement and the outgrowths from the trabecular mass are shown in Fig. 248. The hypophysis, Hy, lies in a deep fossa, which remains in the adult and is known as the sella turcica; on the caudal side of the hypophysis the fused ends of the trabeculse have risen as a transverse plate, the posterior clinoid ridge, p.cl, and in front of the hypophysis is the much smaller anterior clinoid ridge ; p.sp indicates the region of the future prse-sphenoid bone, the cartilage of which is continued directly for- ward in the nasal septum as the ethmoidal plate ; from the sides of cartilage there spring two lateral plates, which curve upward and outward around the brain ; the anterior and larger of these plates is the orbito-sphenoid, O.sp, which spreads out between the brain and the eyeball, and extends far back toward the periotic capsule, per ; during its development the orbito-sphenoid cartilage grows around the optic nerve, thus forming the optic foramen, Op, which is near the base of the plate ; the smaller of the plates is the ali-sphenoid, and springs from the region of the two clinoid ridges ; it is short and thick and has a downward process, which extends to the palato- AXIAL, SKELETON. 437 pterygoid bar and represents the external pterygoid cartilage ; this process being external does not show in the figure. Between the ali-sphenoid and the periotic capsule is a shallow fossa for the Gasserian ganglion, and from the ganglion the main stem of the fifth or trigeminal nerve passes out through a foramen. The sup- erior maxillary division of the trigeminal passes out between the orbito- and ali-sphenoids. The nasal cavities are large and com- plex ; they already occupy more than half the length of the head, and in part underlie the brain ; the partition which separates the nasal cavity from the overlying olfactory lobes is composed of undifferen- tiated mesenchyma, which is traversed by the olfactory nerve fibres, but at the present stage, or a little later, the partition chondrifies by an extension of the cartilage of the ethmoidal plate, with the result of producing the cribriform plate, cr. p. The shape of the nasal chambers is rendered complex by the turbinal prominences on the lateral wall of each chamber as described in Chapter XXVIII. Already in the previous stage the median ethmoidal plate had sent outgrowing laminae of cartilage one on each side over the top and down on the outside of each nasal cavity, and from the lateral car- tilage there appear ingrowths into each turbinal prominence. The relations of the cartilage to the nasal chambers can be more readily ^Vv*;tV- ^.;.„,,i;sj^i,a:-,- Fig 249. —Section of the Anterior Portion of the Snout of an Embryo Pig. iTiftm, Median ethmoidal plate; lat, lateral nasal cartilage; i.tb. inferior turbinal prominence into which the cartilage has begun to penetrate ; N, nasal cavity. understood in a cross-section, Fig 249, which calls for no further de- scription than is afforded above and in the explanation of the figure. As partly indicated by Fig. 248, there are five turbinal promi- nences, the ali-nasal, the inferior, it, the middle, nt, and the upper u. th—ihe last two mentioned being, however, hardly distinct from one another at this stage. It now remains only to add that at the 438 THE FCETUS. ventral side of the anterior edge of the ethmoidal plate the cornua trabecularum are still present; the cornua are the anlages of the ali-nasal cartilages. In man the history of the chondrocranium is very similar to that just given for the pig, as we know through the investigations of Spondli, 46,1, Vrolik, 73.1, Virchow, 57.1, and Van ISToorden, 87.1, and others. The general significance of the chondrocranium is discussed in the section on the morphology of the skulj, p. 465. Periotic Capsules. — This name has been employed by Huxley, and may be conveniently retained, to designate the independent car- tilages, which appear very early around the otocysts, and later be- come integral parts of the primitive chondrocranium by coalescing with the occipital cartilage. In pig embryos of about 16 mm. they appear as two rounded masses, Fig. 347, per, close alongside the anterior half of the occipital cartilage, against which they lie with a nearly straight margin, while the rest of their outline is rounded. The aqueductus vestibuli is left as an opening in the cartilage on the upper and inner edge ; the facial nerve, 7, enters the capsule a little behind this, its passage being the aqueductus Fallopii. As regards the inclosed otocyst we find that the semicircular canals and cochlea are just budding forth. At this stage there is sort of plug of non- cartilaginous mesenchyma stiU left in the external wall of the cap- sule. The neighboring cranial nerves show a characteristic relation to the caps»les. The trigeminus passes out between the capsule and caudal extremities of the trabeculse. In the angle between the cap- sule and the occipital cartilage there pass out three nerves, the glosso- pharyngeal, 9, the vagus, 10, and the hypoglossus, 11. In embryo pigs of one inch the capsules have begun to coalesce posteriorly with the occipital cartilage, and in those an inch and a third long they are found coalesced along nearly the whole line of contact between the capsules and the basilar plate. Concerning the origin of the periotic capsules we possess no accurate knowledge, and cannot even say whether they represent primarily distinct skeletal pieces or merely separate centres of chondrification in a larger mesenchymal skeletal piece. The latter appears to me the more probable alternative, and it may be further suggested that the capsules are differentiations of the lateral outgrowths of the in- vesting mass of the cephalic notochord. The questions raised can be answered only by a careful investigation of the mesenchymal cranium. TJltimate History of the Chondrocranium. — The primitive cartilaginous skull is formed by the fusion and expansion of the oc- cipital cartilage, the trabeculae cranii, and the periotic capsules. It occupies the floor of the cranial cavity and the roof of the olfactory cavities, and has certain lateral expansions. The arrangement of these can be understood from the accompanying Fig. 250, although the figure represents a stage in which ossification has begun. Be- tween the nasal cavities lies the mesethmoid septum from the dorsal side of which spring the ali-nasals, oln, covering the dorsal and lateral parts of the nasal cavities ; from the mesethmoid extend also the plates forming the ali-ethmoids and middle turbinal, mth ; AXIAL SKELETON. 439 peiy also the cribriform plate, cr, through which the olfactory nerve passes. The orbito-sphenoidal wings, obs, are large and rise from the prse-sphenoid ; the ali-sphenoidal wings are smaller, al j between the two sphenoid wings is the foramen lacerum; the periotic capsules are large and fill out nearly the whole space between the ali-sphenoids and the wings of the occipital. The occipital has expanded completely around the foramen magnum, /. m, through which the spinal cord enters the brain-case, so as to form on the dorsal side the supra-occipital, s.oc. In the fishes the chondrocranium passes through a stage corresponding closely to that just described, except that in them there is no bone formed ; but whereas in the mammal the chondrocranium does not pass beyond this stage, in the fishes it continues growing until the brain is completely inclosed and there is a perfect cartilaginous skull, at least in the lower forms, marsipobranchs, ganoids, and selachians. We must, then, distinguish two types of chondrocranium, according as it does or does not completely encase the brain. The latter is the type exclusively found in mammalia. The mammalian chondrocranium is repre- sented in the adult by a number of distinct bones, which represent also a still larger number of bones of lower types. As to how the originally continuous cartilage becomes divided into separate bones, our notions are somewhat vague. In the division the centres of ossification play a leading role, of course, but not in the sense that every centre invariably results in the formation of a separate bone. The second important factor is the development of the sutures, which form the boundaries of the bones. The sutures are of two kinds, those marked out by the edges of the chondrocranium itself, and those produced in the cartilage. Although a knowledge of the history of the sutures must be considered of the utmost importance for the elucidation of the morphoj. p-y of the skull, such knowledge appears never to have been sought. besides those parts of the cartilaginous skull which make bones there are certain others, few in number and small in size, which atrophy. We have then to present the history of the ossification and partial atrophy of the chondrocranium. Ossification. — The occipital region begins to ossify during the early part of the third month in human embryos; comparative anat- omy teaches that the occipital bone of man is homologous with five bones — the median ventral basi-occipital bordering the front or ven- tral side of the foramen magnum, the paired lateral ex-occipitals bor- dering the sides of the foramen and including the condyles by which the occiput articulates with the axis, and the paired supra-occipitals, which, however, are often united into a dorsal median bone; in bOC Fig. 260. — Embryo Pig, six Inches Long. Partly Ossified Chondrocranium seen from above. oZ?i, Ali-nasal ; ei/i, eth- moid ; m. *6, middle turbinal ; cr, cribriform plate; obs, orbito- sphenoid ; al, ali-sphenoid ; per, periotic capsules; 6. oc,basi-oc- cipital ; /ni, foramen magnum, s.oc-, supra-occipital. Natural size. After W. K. Parker. 440 ' THE FCETUS. agreement with this homology there are five centres in the occipital cranium, namely, the basi-occipital, the two ex-occipital or condylar, and two supra-occipital, which, however, very soon unite ; according to KoUiker there is also later a small deposit of dermal bone added to the supra-occipital. The ex-occipitals do not unite with the supra- occipitals until one or two years after birth, nor with the basi-occip- itals until the fifth or sixth year. In the sphenoid region ossifica- tion begins during the second half of the third month in the human embryo, and takes place from six principal centres corresponding to the six bones with which the human sphenoid bone is homologized by comparative anatomists. The six centres are : 1, the basi-sphe- noid in the neighborhood of the hypophj^sis, and .said by KoUiker to be due to the fusion of two minor centres ; 3, the pre-sphenoid, which appears in the median line near the optic foramina, and is likewise said to consist of two minor fused centres; the pre-sphenoid, at least in the pig, is the last of the six centres to appear ; 3, 4, the ali-sphe- noid centres, one in each wing. Fig. 250, al j they appear a little later than the basi-sphenoid centre; 5, 6, the orbito-sphenoid centres, which unite with the prae-sphenoid after the fifth month ; the prse- sphenoid and basi-sphenoid do not unite until several j'ears after birth, and even at thirteen years Virchow has found remnants of cartilage between the two bones. In th.Q per iotic region there are three main centres, which are taken to represent as many distinct bones, al- though they unite in mammals into a single bone, the os petrosum; in man the petrous bone is found to have fused with the dermal bone, known as the squamosum, and also with the ring of bone formed around the tympanum of the ear, and known as the annulus tym- panicus ; from the union of these five bones arises the temporal bone of human anatomy. The three centres which appear in the periotic capsules are termed the pro-otic, opisthotic, and epiotic, and are con- sidered to represent the separate bones bearing the same names in lower vertebral js; the pro-otic centre is by its position in close rela- tion with the anterior vertical semicircular canal, between which and the exit of the third division of the fifth nerve it lies ; in pig em- bryos of six inches it forms a patch of bone lying under the fore part of the cochlea above and in front of the fenestra ovalis, and extend- ing to the junction of the anterior and posterior semicircular canals ; the opisthotic centre is on the lower and posterior surface of the capsule, placed so that most of the bulbous portion of the cochlea lies dorsal to it; one of its processes lies between the fenestra ovalis and the fenestra rotunda, close in front of the head of the stylo-hyal car- tilage ; the epiotic centre develops somewhat more tardily ; it is in especial relation with the posterior vertical semicircular canal, and when it first appears (pig embryos of six inches) is a small piece just above the stylo-hyal process and foramen rotundum, and behind both the foramen ovale and the above-mentioned opisthotic process. According to A. J. Vrolik, 73.1, the ossification of the periotic cap- sules proceeds somewhat differently in man, there being four centres which coalesce by the sixth month of foetal life. In the ethmoidal region, including the cribriform plate, the lateral nasal and turbinal cartilages, ossification takes place very late, and the morphological significance or homologies of the various centres is little understood. AXIAL SKELETON. 441 In the pig at birth the median cartilage is unossified, the cribriform plate is about to begin ossification, being invaded by vascular mesen- chyma, the upper and middle turbinals are partially ossified, the in- ferior turbinals almost completely ossified. In man a similar con- dition is reached about the seventh month of fcetal life. The human ethmoid proper does not ossify until the first year after birth. Atrophy. — There are certain parts of the chondrocranium which do not ossify, but are lost in the adult. The exact process by which they are resorbed is not known. The following parts are said to disappear: the cornua trabeculae; 2, the cartilage under the nasals; 3, Spondli's so-called frontal plate, or that portion of the orbito-sphe- noid outside of which the frontal bone is developed ; 4, the parietal plate or a small portion of the ex-occipital outside of which the parietal bone is developed ; 6, a small portion of the ali -sphenoid (ala magna) outside of which the parietal bone is developed ; 6, the cartilaginous capsules of the sphenoidal, maxillary, and frontal sinuses ; 7, parts of the turbinal cartilages. Dursy, 69. 1, 203, has maintained that some of these cartilages do not really disappear by atrophy, but by becoming ossified and united with the dermal bones overlying them. Kolliker (" Entwickelungs- geschichte," 456), without absolutely denying the correctness of Dursy's view, states that he has been unable to confirm it by his own observations. The following description of the primordial skull of Tatusia (one of the Insectivora) in W. K. Parker's own words, 86. 1, 7-10, brings out many points of morphological importance : * "So great is the uni- formity of the early chondrocranium in the eutheria or placental mammals, that the drawing. Fig. 251, made from the skull of an outlying and low type, might serve as a diagram wherewith to illustrate the skull at this stage of the tj'pes of this order, and of all the orders above it. The figure of a chondrocranium like this, but a little less advanced, before the osseous centres have commenced in it — that of the mole — will be given in my next paper ; and such a skull is very near to that of a shark, or, still better, of a skate. The parts, or rather, regions, of which it is composed, correspond very exactly with what is seen in those generalized, but not loic, fishes ; and in this specimen with long centres appearing, the level is ob- tained which is permanent in the skull of the dipnoi, and of such a low ganoid as the paddle-fish (Polyodon).f As in cartilaginous fishes and amphibians, the chondrocranium may be compared to a basin or a boat, the upper part being unfinished, leaving a mem- branous fontanelle of greater or less extent; this is only partially filled in, at present, by the investing bones, the frontals and parie- tals (f., p.). The outline of this sectional view is very elegant, and quite similar to that of a vertical section of a bird's skull at a like stage, except that the nasal roof -cartilages run on along the whole extent of the median keeled bar — the intertrabecula ; in the bird they stop short, leaving a free cartilaginous rostrum, like that of a shark or skate, which, however, only lasts until it has served as a model * Compare also Parker, 86.2, "On the Skull of Insectivora." t See Bridge, "On the Skull of the Pulgodon Felium," Phil. Trans., 1872. Plates 55-57, pp. 683-863. 443 THE FCETUS. on which the huge premaxillaries of the bird are formed. In the sides of this hollow cartilaginous structure near the hind part the large oval auditory capsules {a.sc, chl) are seen to have great distinct- ness ; they are, however, confluent with the chondrocranium proper at various points — above, behind, and below, as the section will show. These are the only sense capsules displayed in a preparation of this kind, for the eyeballs are quite free from the solid cranial structure (and are, indeed, outside in such a view as this) and the left nasal labyrinth has been removed. Before describing this figure in detail there is one remark to be made, namely, that here we have clearly shown the true diagnostic mark of a mammalian skull. This mark is the rupture of the side walls, due to the pressure of the large lateral masses of the cerebrum. In front of the auditory capsules there is a large, elegantly semicircular opening, the crown Fig. 251. —Chondrocranium of an Insectivorous Mammal (Tatusia). After W. K. Parker. Explanation in text. of the arch looking upward and forward. Only the lower half of the wall has thus broken outward; this 'fault' forms the ali- sphenoid, while the orbito-sphenoid (o.s), the so-called 'lesser wing, ' is many times its size and is continuous, over the archways, with the cartilage that runs on backward, into the supra-occipital region (so). There is nothing similar to this in that sauropsidan skull which comes nearest to that of the mammal, the skull of the crocodile (see Trans. Zool. Soc, Vol. XI., Plate 65), while in birds the orbito-sphenoids are very small, even when Jhey are most developed, as in Struthio (see Phil. Trans., 1866, Plate 7), and in that class the ali-sphenoids almost finish the cranial cavity, being turned inward toward each other, on each side of the back part of the orbital septum. I lay special stress upon this rupture outward of the ali-sphenoid, and on the fact that the nasal roofs utilize the whole of the huge high-crested inbertrabecula, because these are the most distinctive marks of the mammalian skull, and they arise out of two things in which the mammal shows its great superiority to even the highest Sauropsida, namely, the huge volume of the cere- brum, and the tenfold complexity of the nasal labyrinth. A third clear diagnostic is seen in this very figure ; this is the peculiar de- velopment of the antero-inferior part of the oblique auditory capsule, AXIAL SKELETON. 443 due to the development of the coils of the cochlea. So that, at once correlated with the sudden expansion, so to speak, of the cerebrum, we have these new and most important improvements in the organs of sniell and of hearing. At first sight, seeing how large the median bar (intertrabecula) is, with its internasal crest (perpendicular eth- moid and septum nasi— pe, s.n), it might be supposed that the mammalian skull was of the high kind, like that seen in many teleostean fishes, in lizards, and in birds. It is not so, however, but belongs to the low kind, seen in selachians and amphibians ; and, like theirs, is hinged on the spine by a pair of occipital condyles. Hence the eyeballs are kept far apart, instead of coming very near each other as in most birds, where often nothing but a membranous fenestra is found between the right and left capsules and their spe- cial muscular apparatus. But the face as well as the skull of the mammal shows marks of excellence, such as are not seen in the Sauropsida, even in the higher kinds as crocodiles and birds. The great development of the nasal organs is correlated with a most remarkable growth of the bones of the upper jaw and the palate to form the 'hard palate.' This is found in rudiment even in the chelonia and in birds ; but especially in the crocodilia, where, how- ever, its excessive development — as in certain Edentata, e. g. Mijr- mecophaga — is not dependent upon or correlated with any great improvement in the organs of smell, but has to do with the peculiar manner in which these monsters take their prey. " Branchial Skeleton.— Every branchial arch contains a skeletal element, which in its primitive form in all vertebrate embryos * is a bar or rod of condensed mesenchyma, which very early changes into cartilage. The number of these bars of course depends upon the number of gill-arches, compare p. 263, and hence in the mammalia there are five branchial cartilages on each side, which begin dorsally near the cranium, and curving around the sides of the pharynx end near the median ventral line. Fig. 177. The position of the carti- lage can be seen in a section of a branchial arch. Pig. 153, to be alongside of the artery or aortic arch, and on the pharyngeal side of the coelom of the branchial arch. The constant recurrence of the simple stage just described in all vertebrates (except, perhaps, in marsipobranchs), renders it highly probable that forms existed at one time with such a branchial skeleton ; but no such forms are known to exist at the present day. It will be convenient to state the divisions which comparative anatomy teaches us may be considered typical for each branchial cartilage. The divisions are usually given as four: 1, pharyngo- branchial, or dorsal segments, which has usually a horizontal course; 2, the epi-branchial, and, 3, cerato-branchial, both at the sides of the pharynx ; 4, the hypo-branchial or ventral segment, which typically articulates with a median unpaired cartilage known as the basi- branchial, or copula. In the aquatic vertebrates the bars usually send out supporting cartilages into the branchial lamellae, but in mammals there is no trace of any similar outgrowths even during embryonic periods. * Except, perhaps, in the marsipobranchs, the branchial skeleton of which is possibly not homologous with that of the higher vertebrates. See, however, Anton Dohrn, 84. 1. 444 THE FCETUS. In mammals the earliest stage of the branchial skeleton has never been accurately described ; this is because investigators have hitherto been content to begin with the cartilaginous stage, instead of the mesenchymal sta.ge, and, consequently we are left with no definite information as to the bars of the fourth and fifth arches, and with insufficient information as to the origin of the bars of the first to third arches. In selachians, according to Anton Dohrn, 84.1, 110- 111, the differentiation of the cartilage of the branchial arches begins shortly after the branchial filaments have appeared as a condensa- tion of the mesenchyma. Fig. 152, G, situated on the pharyngeal side of the arch and tailward of the mesothelial anlage, In.m, of the inner muscles. For the further history see Dohrn, I.e., 114. In regard to the history of the branchial skeleton from the cartilaginous stage on, we have very full information, chiefly owing to the exten- sive investigations of W. K. Parker, also in part through KoUiker, DoUo, Salensky, 80. 1, Fraser, 82. 1, and others. Each pair of bars passes through a distinct series of modifications, therefore it will be convenient to present the history of each pair separately. We shall call the skeleton of the first arch the mandibular bars, that of the sec- ond the hyoid bars, of the third the thyro-hyal bars. Mandibular Bars. — The adaptations of both the mandibular and hyoid bars to functions entirely different from those which they primitively served, are most remarkable. In mammals the mandib- ular bar becomes primarily divided into two parts, a dorsal piece corresponding to the palatoquadrate of comparative anatomy, and a ventral piece known as Meckel's cartilage. The commencement of the corresponding division of the mandibular bar may be seen in a dog-fish embryo of about 23 mm., the upper end of the bar being enlarged and sending out a process which runs forward on the cranial side of the mouth and later joins the trabecula ; this process is the palato-pterygoid ; another process, the meta- pterygoid, runs upward ; the wider part uniting the two processes is homologous with the quadrate; in elasmobranchs the meta-pterygoid process becomes ligamentous. In mammals the early stages have not been worked out. Parker states that in embryo pigs of about 16 mm. the cartila- ginous palato-pterygoid bars. Fig. 247, are less definitely developed than the other skeletal elements present at this stage, but are more or less distinct from the rest of the mandibular bar ; the palato-ptery- goids are situated in the maxillary process, so that, starting from the dorsal end of the mandibular arches, they run obliquely downward and forward toward the anterior end of the trabeculae ; anteriorly, they converge toward the median line, but do not meet. In the mandibular arch itself is the rod-like Meckel's cartilage, Fig. 247, Md. Between the pterygoid plate and the cartilage of Meckel is a space in which Parker figures no skeletal element, but which is oc- cupied by the quadrate element, which in mammals is the anlage of incus. At the same stage (embryo pig, 16 mm.) the lower divisions of mandibular bar or the Meckel's cartilages are much stouter and are better differentiated from the mesenchyma than the palato-ptery- goids; they are situated in the mandibular processes, and do not meet in the median line. Each Meckel's cartilage is a rounded rod, but its dorsal extremity forms a hook, is somewhat enlarged, and is AXIAL SKELETON. 445 situated close to the upper border of" the first branchial cleft. In pigs a little older (25 mm.) the hook is longer and the end of the cartilage is thicker, making it easy to recognize in it the anlage of the malleus, the hook being the future manubrium or handle of the malleus. In pigs two and one-half inches long the malleus is sepa- rately ossified, but is not separated from the cartilage of the jaw. When the final separation takes place I do not know. MeckeVs cartilage proper may be defined as the ventral segment of the first branchial bar. In mammals the two cartilages alwa}-s unite in the median line, although in man the actual union is said not to have been observed. The lower portions of the cartilage ossify metaplastically but not to the median line, and this ossification be- gins in man during the third month. The bony part is incorporated in the permanent mandible, but the rest of the cartilage atrophies and entirely disappears except a small portion of the end next the malleus, which becomes changed into fibrillar tissue and remains, according to KoUiker, "Grundriss," 320, as the ligamentum laterale internum maxillae inferioris. Meckel's cartilage is the homologue of the cartilaginous mandible of the lower fishes, but is not homolo- gous with the bony mandible of the amniota, which is developed later and belongs to the class of the dermal bones. Summary. — The primitive cartilaginous rod of the first branchial arch gives rise first to a palato-quadrate dorsal segment and a ven- tral or Meckelian segment. The palato-quadrate segment subdivides into the palato-pterygoid plate and the quadrate or incus. In the earliest accurately known mammalian stage the palato-pterygoid and incus are already separate, but it may be safel}^ assumed that in a still earlier stage they constitute one piece. The Meckelian seg- ment subdivides into the malleus and the Meckelian cartilage proper ; the latter unites in the median ventral line with its fellow. One inevitably inclines to homologize the parts with a typical branchial arch as follows : The palato-pterygoid is the pharyngo-branchial ; the incus is the epi-branchial ; the malleus is the cerato-branchial; the Meckel's cartilage is the hypo- branchial ; the united ends of the cartilages are the copula. These homologies are, however, some- what hypothetical, principally because the homologies of the malleus are not clearly ascertained, and we cannot say what element of the lower vertebrates it represents. The course of the palato-pterygoid at such a marked angle to the Meckel's cartilage is probably due to the head-bend. Very likely the head-bend is causally connected also with the peculiar forms assumed by the incus and malleus. Hyoid bars, or Eeichert's cartilages, as they have been named by Kolliker, are the skeletal elements of the second or hyoid branch- ial arch, and they are typically divided, like the other bars in the lower vertebrates, into four parts, the dorsal one of which (pharyngo- branchial) fuses quite early with the cartilaginous periotic capsules, and becoming ossified appears in the human adult as the styloid process; the second part (epi-branchial) becomes partly ligamentous in all placental mammals, and perhaps wholly ligamentous in man ; the third part (cerato-branchial) and fourth part (hypo-branchial) both become cartilaginous and ossify early, so as to form a single 446 THE FCETUS. piece of bone, which perhaps includes also some bone derived from the second part also. This single piece of bone is known in the adult as the lesser horn of the hyoid. The adult hyoid bar then comprises the styloid process, the stylo-hyal ligament, and the lesser hyoid cornua. The main body of the hyoid probably belongs to the next branchial arch, but the hyoid bars unite with it very early. It was long maintained by Huxley, 69.1, and W. K. Parker (Parker and Bettany, " Skull") that the incus was derived from the hyoid bar, but since Salensky, 80. 1, showed that the incus is devel- oped from the mandibular bar, Parker, 86.1, 10, has retracted his fornier opinion. Reickert, 37.1, thought that the stapes was de- rived from the hyoid bar, but recent investigations show that this is not the case, although Rabl, 87. 1, has shown that the stapes is de- veloped within the territory of the second branchial arch. O. Hert- wig ("Lehrbuch," 3te Aufl., 509) suggested that the stapes was a double bone, one part of which is derived from the branchial skele- ton, but Staderini, 91.1, has proved that this suggestion cannot be adopted — see Chapter XXVIII. The following quotation from W. K. Parker, 86.1, 10, 11, gives some insight into the discussion about the incus, which may be said to have ended with the admissions made in the course of the quoted sentences. " But that great improvement just spoken of as appear- ing in the organ of hearing in the mammal has wrought a change in the hinder face that has tivo most important bearings. From the first promise of an ear-drum in the tailed Amphibia, to its highest fulfilment in the noblest of the oviparous tribes — the birds that nes- tle on high ('aves altrices ') — the only element from the visceral arches that is used for carrying the vibrations of the air inward to the organ of hearing is the uppermost part of the hyoid arch — ^the ' pharyngo-branchial ' element of the second postoral arch, to speak morphologically. From the salamandroids to the singing birds, all through the Amphibia and Sauropsida, the first postoral arch which forms both the upper and lower jaw is only segmented once, that is, into an epi-branchial and a cerato-branchial element or joint. The upper piece is specially termed the ' quadrate ' and the lower the 'articulo Meckelian;' the one forms the swinging piece, hinge, or pier, to the ' compound lower jaw,' and the other its axis or pith, the part which becomes covered with more or fewer ' investing bones. ' In these low ' Eutheria ' and also in both the ' Metatheria ' and the ' Prototheria ' (Marsupials and Monotremes), the modified visceral rod that runs through the drum cavity has two new elements added to the one (single or variously segmented) element derived from the hyoid arch. This is an apparently sudden change, for we have it in the lowest or teatless mammals ; their ancestry that should show us the earlier steps of the change are unfortunately all extinct. In this dilemma not only zoology, but palaeontology also, fails us utterly, but embryology comes in with every stage and every link. I have worked out the early conditions of these parts in several kinds of Marsupials, and in the young of Ornithorhynchus ; but even in the lower Euthreia, the Edentata, now to be described, and in the large and varied group of the Insectivora, I have been able to trace every step in the transformation of these parts. I am now satisfied that AXIAL, SKELETON. 447 the incus is the upper element of the first or mandibular arch ; both Professor Salensky's and Professor Praser's researches put this, I think, beyond doubt ; and my own attempts for a long time to make the hyoid theory of this part agree with facts, only kept the subject in hopeless confusion. The new elements of the ear-chain are then the arrested quadrate or incus, and the arrested and amputated articular region of the articulo-Meckelian rod or primary lower jaw. The bony part of the ' ramus ' is the well-known dentary with the coronoid and splenial bones in a sub-distinct state ; the cartilage for the neiv articulation oi the lower jaw is derived from a large super- ficial slab — a ' lower labial ' — the like of which is not found again until we get as low down as the Chimaeroids. Prom this is derived the hinder half of the ramus by transformation of its substance into bone ; and from this we get the cartilage, both of the condyle and the glenoid cavity, and also of the intervening ' meniscus. ' Of course the drum cavity is the ' first cleft, ' and the concha auris with its segmented meatus-tube — the tympanic bone, the tympanic bulla, and the cartilaginous lining of the Eustachian tube — all these are parts of a curiously specialized opercular growth belonging to the hinder edge of the first visceral fold and arch. This last assertion has not been made as a stride across the types from the mammal to the elasmobranch, but is the result of a very slow step-by-step process, made during many years 'along all the lines' of vertebrate mor- phology." Thyeo-HTOID Bars. — Whether these bars extend in the mesen- chymal stage through the entire length of the third branchial arches or not is not known, but their lower ends are chondrified and later ossified to form the principal part of the hyoid bone. There appears very early a median azygous element or copula, which in pigs of 16 mm. is already cartilaginous and united with not only the thyro- hyoid bars but also with the recurved ends of the hyoid bars. This copula is called the basi-hyal, and is the anlage of the main body of the hyoid bone; it is said to belong to the third branchial arch, al- though the hyoid bars unite with it. It is at first small in size, but as development progresses it enlarges considerably, while the ventral ends of the hyoid bars grow but little ; it results that the relative size of the parts is changed, and the rudiments of the hyoid bars, which start nearly equal in diameter to the basi-hyal, appear in the adult as the lesser horns. The thyro-hyoid cartilages, on the other hand, grow at about the same rate as the basi-hyal and become the greater horns of the adult hyoid bone. The hyoid bone of mammals is formed by the ventral portions of the hyoid bars (lesser cornua) , the ventral portions of the thyro-hyoid bars and the copula of the third pair of branchial arches. In ac- cordance with its development the hyoid bone has five centres of ossification, one for the body and one for each of its four horns. Ossification begins in man in the great cornua and body during the last month of foetal life, and in the small cornua during the first year after birth. The great cornua and body do not unite until mid- dle life, and the lesser cornua usually remain distinct, though some- times found united with the body at advanced ages. 448 THE FCETUS. II. The Limbs and Appendicular Skeleton. Origin of Vertebrate Limbs. — The morphological value of the limbs of vertebrates has long been the subject of discussion and speculation, and at the present time the solution of the problem is theoretical rather than positive. It is unnecessary to give a resume of the older hypotheses as to the archtj'pe of the limbs, though I may refer those interested to Owen's article " On the Nature of Limbs," and Goodsir's essay " On the Morphological Constitution of Limbs," Edinburgh, Neio Philos. Journ., 1857. Gegenbaur has ad- vanced an hypothesis of the origin of limbs in support of which his memoir, 76. 1, brought very scanty evidence. According to this hypothesis the limbs are modified branchial skeletons, the shoulder and pelvic girdles representing the branchial bar, and the skeletal pieces of the limbs proper representing branchial rays ; the central ray formed the axis of the limb, and the remaining rays gradually became articulated with the axial ray, and thus produced the type of limb found in Ceratodus, and which Gegenbaur regards as the primi- tive type from which all vertebrate limbs are derived. This theory, which was adopted by Huxley (on Ceratodus, Proc. Zool. Soc, London, 1876), has attracted great attention, although it has been definitely set aside by the observations of Balfour, 81.1, on the de- velopment of the limbs of Scyllium, which demonstrated that the limbs arise as parts of a longitudinal fold, which runs along the side of body, both fore and hind limb being part of the same fold. Were Gegenbaur's hypothesis correct, the limbs should arise as transverse or vertical folds. Under these circumstances it seems to me that Gegenbaur's theory has merely historical interest. The only theory having any standing at present is the one adopted by Balfour (" Comp. Embryology," II.) according to which the limbs are specialized portions of a lateral fin-fold, similar to the dorsal and ventral median fin-folds of fishes. The resemblance of the lateral fins or true limbs to the median fins in general structure is obvious in many fishes, and especially in teleosts, and renders direct compari- son very natural. Such comparison is suggested by several writers, but was first definitely worked out by J. K. Thacker, 77. 1, and at about the same time advocated by St. George Mivart, 79.1, both these authors basing their conclusions upon comparative anatomical studies. Their general result was that the structure of limbs could be explained by assuming that they are specialized portions of lateral fin-folds, having a structure similar to that of the median fin-folds. At about the same time appeared the chapter of Balfour's mono- *■ graph on the development of elasmobranch fishes, in which he ad- vocated a similar theory upon embryological grounds, and by his observations put the theory upon a firm basis. It is a remarkable coincidence that the same hypothesis was formulated independently and published at about the same time by three investigators. These views were attacked by Von Davidoff, 79. 1, then a pupil of Gegen- baur's, and to Davidoff's paper Gegenbaur added a note upholding his theory ; these criticisms were adequately answered by Balfour, 81.1 ("Reprinted Works," I., 714). From the manner of their development it is obvious that the limbs THE LIMBS AND APPENDICULAR SKELETON. 449 have a flattened form and a dorsal (or extensor) surface, and a ven- tral (or flexor) surface, and as soon as they project from the body, as they do at right angles, there is an anterior or cranial border and a posterior or caudal border. The development of the limbs in Scyl- lium, as described by Balfour, throws important light on the primitive position of these borders. Balfour ("Comp. Embryol.," II., 612) says : " The direction of the original ridge which connects the two fins of each side is nearly, though not quite, longitudinal, sloping somewhat obliquely downward. It thus comes about that the attach- ment of each pair of limbs is somewhat on a slant, and that the pel- vic pair nearly meet each other in the median ventral line a little way behind the anus. The elongated ridge, forming the rudiment of each fin, gradually projects more and more, and so becomes broader in proportion to its length, but at the same time its actual attachment to the side of the body becomes shortened from behind forward, so that what was originally the attached border becomes in part converted va.\x>^e posterior border. This process is much •more completely carried out in the case of the pectoral fins than in that of the pelvic, and the changes of form undergone by the pectoral fin in its development may be gathered from my figures. In Scyllium the development of both the pectoral and pelvic fins is very similar. In both fins the skeleton in its earliest stage consists of a bar spring- ing from the posterior side of the pectoral or pelvic girder, and running backward paral- lel to the long axis of the body. The outer side of this bar is continued into a plate which extends into the fin, and which becomes very early segmented into a series ^ , ^. , ,, „ ^ . „ , ~ •' f? , J. • 1 J. < Fig. 258.— Pectoral Fin of a Young Embryo of Sycl- 01 parallel raj^S at rignt an- Hum in Longitudinal and Horizontal Section. The n-loa +r. +Via Innfrif nrlinal Viar skeleton of the fin was Still in the condition of em- gieS to xne longltuainai oar. ^^^^^^^^ cartilage; 6. p., basl -pterygium ("eventual In other words, the primitive meta-pterygulm) ; /r, flnrays: p.g. pectoral girdle in T T , ^ 1 J.1 j_i j2 transverse section ;/, foramen in pectoral girdle- p. c. skeleton of both the hnS con- wall of peritoneal cavity. sists of a longitudinal bar running along the base of the fin and giving off at right angles a series of rays which pass into the fin. The longitudinal bar, which may be called the basi-pterygium, is, moreover, continuous in front with the pectoral or pelvic girdle as the case may be. My obser- vations show that the embryonic skeleton of the paired fin con- sists of a series of parallel rays similar to those of the unpaired fins. These rays support the soft part of the fin, which has the form of a longitudinal ridge, and are continuous at their base with a longitudi- nal bar, which may very probably be due to secondary development. As pointed out by Mivart, a longitudinal bar is also occasionally formed to support the cartilaginous rays of unpaired fins." Balfour's observations show that there was a primitive longitudi- nal skeletal piece at the base of the limb-fold, and that from this rays are developed which run out into the fold ; Mivart assumed 29 450 THE FCETUS. that the rays were primitive and the longitudinal piece the product of the fusion of the bases of the rays. As the limb grows out its base becomes free and constitutes the posterior border, and the basal skeletal piece appears as the axis of the limb, while the fin-rays run off from one side toward the primitive outer or ultimate cephalic border of the fin ; on the caudal side of the axis there are necessa- rilly no fin-rays. If we assume, as we must, that Scyllium illustrates the general type of fin development, then a condition in which, as in the fins of the adult Ceratodus, there are rays on both sides of the axis must be considered a secondary condition. The Ceratodus type is known as the archipterygium, and, as already stated, has been held by Gegenbaur to be the ancestral form of vertebrate limbs. But our knowledge of the development and morphology of fins ren- ders it impossible to accept this view, at least at present. The archipterygium may be defined as a skeletal limb axis with rays coming off on both sides; no such fins are known among the lower fishes, but only among the higher (Dipnoi) ; this fact offers another serious obstacle to regarding the archipterygium as the prim- itive ancestral form, but suggests that it may represent the ancestral form of the pentadactyle limbs of amphibia and mammals. I think much may be said in favor of this suggestion, and indeed it is largely on account of the possibility of deducing the pentadactyle limbs from it that the archipterygiiim has been regarded as an archtype by Gegenbaur and his followers. The cheiropterygiiim is the archtype or ancestral form of the pentadactyle limb. Its essential characteristic is the division into four segments : , ( Upper arm. „ j Forearm. „ S Carpus. , ( Hand. ■ ( Upper leg. ' j Lower leg. • } Tarsus. ] Foot. The upper segment contains one long bone (humerus or femur) ; the second segment contains two long bones (radius or tibia, and ulna or fibula) ; the third segment contains nine small bones (carpals or tar- sals) ; the fourth segment consists of separate digits, five in number, hence the term pentadactyle applied to this type of limb ; each digit has a proximal or basal bone (metacarpal or metatarsal) upon which follows a linear series of phalanges, separate 'bones variable in num- ber. It is convenient always to count the digits in the same way, commencing from the radial or tibial side ; thus the thumb is the first digit of the hand, the great toe the first digit of the foot. The arrangement of the carpal and tarsal bones is greatly modified not only in the amniota but also in many of the amphibia, both by the suppression of some of the nine bones and by fusions among them. The nine bones are the intermedium between the distal ends of the radius and ulna, the radiale and ulnare at the distal ends of the radius and ulna respectively ; the two centralia, on the distal side of the ntermedium ; between these four and the meta- carpals (or metatarsals) follow the five carpalia or tarsalia, . In most pentadactyle limbs the two centralia are fused into one bone, the centrale. In many cases some of the bones are suppressed. The following table shows the homologies in man : THE LIMBS AND APPENDICULAR SKELETON. 451 Ulnare (fibulare) . Cuneiforme. Calcaneum. Intermedium. Lunare. I * * i Eadiale (tibiale). Scaplioid. ^Astragalus. Centralia. ( ?) Naviculare. 1. 1. Carpale. 1. Tavsale. 2. (Carpalia). 2. 3. 3. (Tarsalia). " 3. 3. 4. 5. ■ Unciforme. [ Cuboides. The pisiforme is a sesamoid bone developed in the tendon of the flexor carpi ulnaris, and has nothing to do with the primitive carpus. It is generallj- taught that there is one series of bones which represents the true axis of the limb, and that the other bones repre- sent a series of rays coming off from it. This supposed axis begins with the humerus (femur), is continued through the ulna (fibula), and terminates with one of the digits, but which digit authorities are not agreed; thus Gegenbaur carries the axis through the ulnare fifth metacarpal and fifth digit, which makes the first ray pass off from the humerus and include the radius, radiale, first carpal, and first digit; the second ray arises from the ulna and includes the intermedium, one centrale, and the second digit ; the third ray springs from the ulnare and includes one centrale and the third digit ; the fourth ray springs from the fifth carpale and includes the fourth car- pale and the fourth digit; similarly, changing the names, in the hind limb, see Gegenbaur, " Grundriss d. vergl. Anatomie," 1878, 513, Fig. 273. Wiedersheim, on the contrary, carries the axis (see his " Grundriss der vergl. Anatomie," 2te Aufi., Fig. 110) through the ulna (fibula), intermedium, both centralia, second carpale (tar- sale) , and second digit. Such divergences of opinion raise doubts as to the existence of any true axis at all. A full discussion of the morphology of the limbs does not fall within the scope of this work, because our conceptions are not based upon embryological observations. I shall, therefore, merely refer to the recent papers of G. B. Howes, 87.1, J. A. Eyder, 87.1, D'Arcy Thompson, 86.1, Hatschek, 89.1, and E. E. Prince, 90.1. Relation to the Somites. — Each limb arises along the territory of several somites, and receives outgrowths from the muscle plates of several successive segments, and with these outgrowths, which pro- duce the muscles of the limbs, come the nerves of several segments, so that the fact that the limb arises along a considerable length of the body explains several important features in the development of limbs — features which remain inexplicable if we accept Gegenbaur's theory of the evolution of the limb from a branchial arch, because this theory confines the primitive limb to a single segment, whereas at its verj' earliest stage it is already related to several segments. As to the exact number of limb somites we are in doubt. Balfour's observations indicate that each limb was originally attached along a considei-able number of segments, but that on the caudal side the attachment becomes shortened. As it is not until this restriction of the base has taken place, that the muscle plates penetrate the limb, it follows that the iimscles of the limb are derived from a less num- ber of segments than corresponded to the primitive attachment. 452 THE FCETUS. This reduced number is probably five in the amniota, but certainty on this point is yet to be reached. Concerning the position of the limbs, as regards their distance from the head and the segments to which they belong, we have little exact knowledge. A. M. Paterson, 91.2, holds that the position in this sense is not uniform among the mammalia ; he bases this opin- ion upon the innervation which is variable. The variation is much less for the fore than the hind limb; the former is, as a rule, inner- vated from the lower cervical and upper thoracic segments; the twenty-fifth spinal nerve is the only one invariably present in the hind limb of mammals, while the nerve plexus may begin, according to the species, with any of the nerves from the twenty-first to the twenty-fifth, and, as it usually comprises five or six spinal nerves, it ends with the twenty -fifth to twenty -ninth nerve. It is thus prob- able that the hind limb readily shifts its position. As the sacrum is always developed in connection with the limb it follows that the number of prae-sacral vertebrae must vary, although there is no in- tercalation or. obliteration of vertebrae. Position of the Limbs. — The primitive position of the limbs is at right angles to the body in a plane nearly parallel with the longi- tudinal body-axis. The first change is the appearance of two bends which give the limb the position which is permanent in amphibia ; the bends are similar in the fore and hind limbs. The first bend (elbow or knee) is at the end of the upper limb (humerus or femur) , and is such that the lower limb is flexed downward (ventralward) and toward the median line ; the second bend is at the carpus (tar- sus) and is in the opposite direction or outward. Thus the ventral aspects of the forearms and lower legs come to look inwardly and their dorsal aspects outwardly; Avhile the ventral aspects of the hands and feet look downward and their dorsal aspects upward. This change is obviously correlated with the change from aquatic to terrestrial life and the consequent substitution of legs for fins. When the position of the limbs has been no further altered than this, the radius and tibia are found on the cranial side, the ulna and fibula on the caudal side of their respective limbs. The second step is the torsion of the limbs, which is similar in both pairs and occurs in all mammalia, the result of which is that the digits point headward, the first digit being in both hind and fore limbs toward the median line. This is the arrangement which is permanent in the reptilia and in the lower mammalia. The torsion, bj^ which the change is effected, does not take place in the arm or leg itself, but at the shoulder or hip. The third change is the torsion of the upper arm (not known to occur in the leg) by which the distal end of the humerus is twisted over through an angle in man of nearly or quite one hundred and eighty degrees ; by this torsion the head of the radius, which before the change was on the inner side of the arm, is brought across in front of the ulna to the outer side, with the result that if the hand is kept in its primi- tive position, palm down, the forearm is twisted in the reverse direc- tion to the upper arm ; this third change is accompanied by accessory modifications in the joints and muscles by which the radius becomes so movable that it can be employed to turn the hand with the palm either up (supination) or down (pronation) . THE LIMBS AND APPENDICULAR SKELETON. 453 Anatomists are not entirely agreed as to the alterations in the positions of the limbs. The above formularization of the changes is based partly upon that given by Huxley in his " Anatomy of Ver- tebrated Animals," 32-33, and on Hatschek, 89. l,partl}' on a f ew observations I have made on skeletons. It is to be expected that the limbs of the higher mammalia pass through the three stages of limb position which may be conveniently designated as amphibian, reptilian, and mammalian. Unfortu- nately there are no observations as yet to show whether this is the case or not. This gap in our knowledge offers a favorable opportu- nity for a research. Shoulder Girdle. — The anlage of the shoulder girdle is probably continuous in all vertebrates, as it has been shown to be in the fishes, with the anlage of the base of the limb, but in the amniota it early becomes a separate cartilage, lying in one plane and extending dorso- ventrally. In mammals there is a large dorsal segment of this car- tilage above the articulation with the humerus (glenoid fossa) and a much smaller segment below the articulation. The dorsal segment develops into the large shoulder blade, while the ventral segment forms merely the small coracoid process, although it is the homo- logue of the large and independent coracoid bone of sauropsida and amphibia. It is to be noted that Sabatier, 80.1, has homologized the "coracoid" process with the prae-coracoid, and holds that the upper third of the mammalian glenoid fossa, which ossifies from a separate centre, represents the true coracoid, but this view has not been accepted. Little is known concerning the development of the scapula in mammalia beyond what is given in W. K. Parker's monograph, 68.1, a work which has by no means received the attention it deserves. Owing to the reduc- tion of the coracoid in mam- malia the history of the scapula is practically that of the entire shoulder girdle. Parker, /. c, p. 223-224, records some obser- vations on the scapula of hu- man embryos. In an embryo 5^ inches long, the scapula, Fig. 253, already has much of its persistent form and is ossi- fied through about half its extent, b; the small size of the prse-scapular region, p. sc, and the great size of the acro- mion, Acr, are features in which the embryonic shoulder blade differs strikingly from that of the adult; in an earlier stage, embryo of 2^ inches the prae- scapula is proportionately still smaller, and the acromion thicker and more curved toward the clavicle. The meso-scapular ridge, msc, is the thick prolongation of the acromion. The coracoid process. S.Sc Fio. S53.— Suapula of a Human Embryo of five and one-half Inches, Dorsal View. Natural size. J?p, Goette's episternal element OParker's omo- sternum) ; c, cartilage at end of clavicle ; CT, termed by Parker ]5rae-coracoid; c' cartilage at scapular end of clavicle, termed by Parker meso- scapular segment: Acr, acromion; Ci\ coracoid process; gl.f, glenoid fossa; 6, bony portion; msc, meso-scapula ; S.Sc, supra-scapula ; p.sc, prffi-soapula. After W. K. Parker. 454 THE FCETUS. Cr, is small and slightly curved ; it is connected by a fibrous band with the end of the clavicle, but the cartilaginous end of the clavicle (Parker's so-called meso-scapular segment) is articulated by a synovial joint at this stage with the end of the acromion. The coracoid has its own centre of ossification, to which are added at the tirne of puberty two epiphysal centres (Rambaud and Renault) — its ossifica- tion thus indicating its morphological individuality. The acromion has two, sometimes three, centres, which appear between the four- teenth and sixteenth years and soon coalesce, but the ossified acro- mion does not unite with the scapula until eight to ten years later. There is a separate centre for the inferior angle (supra- scapular) and for the upper part of the glenoid cavity. Clavicle. — Opinions differ as to whether the clavicle is a dermal bone or an integral portion of the scapular arch. It is, as discovered by C. Bruch, 53.1, 3ri-.373, the first bone formed in the human embryo its ossification going on during the seventh week. Geg- enbaur, has shown that the bone commences by ossification of mesenchyma ; then cartilaginous masses appear at each end, which are, however, softer and have less basal substance than most em- bryonic cartilage ; these cartilages serve to maintain the growth in length of the clavicle. KoUiker states (" Entwickelungsgeschichte," 1879, p. 495) that he has verified on rabbit embryos Gegenbaur's observations, though he regards the tissue of the anlage as inter- mediate between mesenchyma and true cartilage. KoUiker adds that there is a separate centre of ossification, which may be com- pared to an epiphysis at the sternal end. This epiphysal piece was first described by ^^' . K. Parker, 68. 1, 223-'224, and was shown by him to become distinct while still cartilage ; Parker terms it the prae- coracoid, although this name is properly applied to an entirely differ- ent bone. These peculiarities in the development of the clavicle, together with Rathke's statement that the clavicular anlage is at first continuous with that of the coraco-scapular arch, and certain observations of his own, have led Alex. Goette, 77.1, to maintain that the clavicle is an element of the shoulder. Goette's observa- tions have been in part confirmed by C. K. Hofmann, 79. 1. Gegen- baur regards the mammalian clavicle as a compound bone homolo- gous with both the true dermal clavicle (Decknochen des Procora- coids) and the cartilaginous proooracoid of fishes, the two originally separate skeletal elements having united with one another ; by this double homology Gegenbaur explains the peculiar development of the bone; compare his " Grundriss d. vergl. Anatomic," 2te Aufl., 501. It is possible, however, that we attribute too great morphological meaning to the appearance of cartilage, and that partial chondrifi- cation of the clavicular anlage does not mean, as Gegenbaur thinks, a separate element of the skeleton, or, as Goette thinks, connection with the shoulder girdle, but is merely a modification of the histo- genetic development^ — compare the paragraph on the mandible, p. 444. We cannot hope to understand the homologies of the clavicle until its development shall have been completely traced, beginning with the earliest mesenchymal stage. Episternum. — "Whether there is any episternum in the human embryo is uncertain. Perhaps the suprasternal cartilages just men- THE LIMBS AND APPENDICULAR SKELETON. 455 tioned as having been described by Q. Ruge, 80.1, are its repre- sentatives. K. Bardeleben, 79.1, has sought to homologize the deep portion of the interclavicular ligament as the rudiment of the human episternum. A. Goette, vs^ho has w^orked out, 77.1, the development of the parts more fully than any other anatomist, finds that "paired interclavicular elements grow out backward from the ventral ends of the clavicles, and uniting together form a somewhat T-shaped interclavicle overlying the front end of the sternum. This condition is permanent in the Ornithodelphia ex- cept that the anterior part of the sternum undergoes atrophy. But in the higher forms the interclavicle becomes almost at once divided into three parts, of which the two lateral remain distinct, while the median element fuses with the subjacent part of the sternum and constitutes with it the presternum {manubrium sterni). If Goette's facts are to be trusted, and they have been to a large extent confirmed by Hofmann, his homologies appear to be satisfactorily estab- lished. "_ (Balfour.) Pelvic Girdle. — The pelvic girdle resembles the pectoral; it consists of a bar of cartilage which articulates with the femur; the articular cavity is known as the acetabulum and divides the girdle into a dorsal and ventral segment, as the glenoid fossa divides the scapular arch. The dorsal pelvic division is called the iliac section, the ventral division the pubic section. The iliac section has no connection with the vertebral column in fishes, but is articulated with the sacral vertebree in amphibia and amniota. The pubic section meets its fellow in the median ventral line; in amphibians it be- comes more expanded and plate-like, and there appears an inter- ruption of the cartilage by which the obturator foramen is formed ; this foramen divides the pubic section into a cephalic portion or pubis, and a caudal portion or ischium. In mammals the foramen is enlarged so that ischium and pubis are more distinct than in am- phibia. Balfour ("Comp. Embryology," II., 600) found that the mode of development of the pelvic girdle in Scyllium is very similar to that of the pectoral girdle. There is a bar on each side continuous on its posterior border with the basal element of the fin (Figs. 345 and 347). This bar meets and unites with its fellow. Concerning the early development of the girdle in amniota I know of only the observations of A. Bunge, whose dissertation I have not seen, and those of Alice Johnson, 83. 1, on the chick. The lat- ter shows that the girdle is continuous with the femur at first; the ischium and pubis grow out separately from the acetabular region, both growing ventralward, the former on the caudal, the latter on the cephalic, side of the crural nerve ; if the ischium and pubis were to unite distally, which, however, they do not do in the chick, they would inclose a space homologous with the obturator foramen. This observation renders it improbable that the ischium >and pubis are together homologous with the pubic section of the girdle in fishes, and indicates that one of them is a new element— added in the amphibia perhaps. The pubis sends out a process headward from just below the acetabulum ; this process is the pre-pubis ; it is well *"Entwiek. Beckengui-tels. " Inaug. Diss., Dorpat, 1880. 456 THE FOETUS. developed in the Ornithorhynchus, but is rudimentary in the higher mammalia. Skeleton of the Arm. — Our knowledge of the development of the skeleton of the fore limb in mammalia is very imperfect. It rests chiefly on the data furnished by Henke and Reyher, 74.1, supplemented by E. Rosenberg's valuable investigations of the cen- trale carpi in man, 75. 1, and a few observations recorded by KoUi- ker in his "Entwickelungsgeschichte," 3te Aufl., and by C. Emery, 90. 1. To these references ought to be added one to the paper on the development of ungulate limbs by Alexander Rosenberg, 73.1, which, however, has less direct interest for us. The skeleton of the arm in mammals (as in amphibia also, H. Strasser, 79. 1) in its earliest mesenchymal stage forms an uninter- rupted anlage (KoUiker, I.e., 491), with no indication of its future subdivision, and is, moreover, proijably continuous with the anlage of the pectoral girdle. As soon as chondrifications begin the indi- vidual skeletal pieces are indicated by corresponding separate centres of chondrification, which begin near the centre of each piece and spread toward its periphery. The separation of each digital series is given in the primitive mesenchymal anlage, Avhich also shows, according to C. Emery, 90. 1 , 296, traces of a sixth digit (prae-poUux) in front of the thumb ; the sixth digit persists as a rudiment and only for a short time. The condensed mesenchyma between two adjacent cartilages becomes fibrillar and produces the articulations. On the development of the joints, see p. 460. When, however, two car- tilages fuse into one, as occurs in man with several of the carpals, the fusion takes place very early and no articulation is formed. It may be noted here that the joints are not differentiated until six or eight weeks after chondrification begins. In the human embryo at six weeks nearly all the skeletal pieces are present ; the ends of the humerus are somewhat enlarged ; the ulna has a processus anconseus already ; the radius shows both head and neck; the metacarpals are beginning to chondrify. By the eighth week the phalanges are cartilaginous, having begun to chon- drify (Kolliker, "Entwickelungsgeschichte," 2te Aufl., 491) when the five digits became distinctly indicated by marginal notches in the hand, and in the humerus the calcification of the cartilage, pre- liminary to its degeneration and replacement by bone, has begun ; the articular surfaces of the cartilages are becoming more sharply defined (Henke u. Reyher, 74.1, 224-230). These authors discov- ered, I. c, p. 268, that the centrale exists as a separate structure in embryos of the second month. E. Rosenberg, 75.1, 172-191, has traced out the history of the centrale very carefully ; it is character- ized by having less intercellular substance than the other carpal car- tilages, and by never changing into bone, except as a rare anomaly ; normally it is gradually absorbed in older embryos and disappears, the space it occupied being taken up by the enlargement of the radiale (scaphoid) . Henke and Reyher have observed a tenth carpal also which was perhaps merely a transitory (Gegenbaur's " radial sesamoid") bone — at least this suggestion of E. Rosenberg's is a plausible explanation. Ossification. — " In the humerus a nucleus appears near the mid- THE LIMBS AND APPENDICULAR SKELETON. 457 die of the shaft in the eighth week. It gradually extends, until at Ijirth only the ends of the bone are cartilaginous. In the first year the nucleus of the head appears, and during the third year that for the great tuberosity. The lesser tuberosity is either ossified from a distinct nucleus, which appears in the fifth year, or by extension of ossification from the great tuberosity. These nuclei join together about the sixth year to form an epiphysis which is not united to the shaft till the twentieth year. In the cartilage of the lower end of the bone four separate nuclei are seen, the first appearing in the capitellum in the third year. The nucleus of the internal condyle ■ appears in the fifth year, that of the trochlear in the eleventh or twelfth year, and that of the external condyle in the thirteenth or fourteenth year. The nucleus of the internal condyle forms a distinct epiphysis, which unites with the shaft in the eighteenth year; the other three jiuclei coalesce to form an epiphysis, which is united to the shaft in the sixteenth or seventeenth year. " The radius is developed from a nucleus which appears in the mid- dle of the shaft :"n the eighth week, and from an epiphysal nucleus in each extremity which only appears some time after birth. The nucleus in the carpal extremity appears at the end of the second year, while that of the head is not seen till the fifth or sixth year. The superior epiphysis and shaft unite about the seventeentla or eigh- teenth year; the inferior epiphysis and shaft unite about the twen- tieth year. " The ulna is ossified similarly to the radius but begins a little later. The nucleus of the shaft appears about the eighth week, that of the carpal extremitj' in the fourth or fifth year. The upper extremity grows mainly from the shaft, but at the end of the olecranon a small •epiphysis is formed from a nucleus which appears in the tenth year. This epiphysis is united to the shaft about the seventeenth year ; the inferior epiphysis about the twentieth year. " From what is stated above it appears that in the bones of the arm .and forearm the epiphyses which meet at the elbow-joint begin to -ossify later, and unite with their shafts earlier, than those at the opposite ends of the bones; whereas in the bones of the thigh and leg the epiphyses at the knee-joint are the soonest to ossify (except in the fibula) and the latest to unite with their shafts. In the bones of the .arm and forearm the arterial foramina are directed toward the el- bow ; in those of the thigh and leg they are directed awaj- from the knee. Thus, in each bone the epiphysis of the extremity toward which the canal of the medullary artery is directed is the first to be united to the shaft. It is found also that while the elongation of the long bones is chiefly the result of addition to the shaft at the epiphysial synchondroses, the growth takes place more rapidly, and is continued longer, at the end where the epiphysis is last united ; and the oblique direction of the vascular canals is due to this in- equality of growth, which causes a shifting of the investing perios- teum, and so draws the proximal portion of the medullary artery to- ward the more rapidly growing end. " The carpus is entirely cartilaginous at birth. Each carpal bone is ossified from a single nucleus. The nucleus of the os magnum ap- pears in the first year ; that of the unciform in the first or second 458 THE FOETUS. year ; that of the pyramidal in the third year ; those of the trapezium and the lunar bone in the fifth year ; that of the scaphoi in the sixth or seventh year ; that of the trapezoid in the seventh or eighth year ; and that of the pisiform in the twelfth year. " The metacarpal hones and phalanges are usually formed each from a principal centre for the shaft and one epiphysis. The ossi- fication of the shaft begins about the eighth or ninth week. In the inner four metacarpal bones the epiphysis is at the distal extremity, while in the metacarpal bone of the thumb and in the phalanges it is placed at the proximal extremity. In many instances, however, there is also a distal epiphysis visible in the first metacarpal bone at the age of seven or eight years, and there are even traces of a prox- imal epiphysis in the second metacarpal. In the seal and some other animals there are always two epiphyses in these bones. The epiphy- ses begin to be ossified from the third to the fifth year, and are united to their respective shafts about the twentieth year. The terminal phalanges of the digits present the remarkable peculiarity that the ossification of their shafts commences at the distal extremity, instead of in the middle of their length, as is the case with the other phalanges and with the long bones generally (F. A. Dixey)." (G. D. Thane in Quain's " Anat.," tenth edition.) Skeleton of the Leg.^ — The primitive mesenchymal anlages of the skeleton of the leg, like that of the arm, is continuous throughout in amphibia, H. Strasser, 79.1, and birds, Alice Johnson, 85.1, and therefore probably in mammals also, and in birds it is continu- ous also with the pelvic girdle, which appears as an outgrowth of the skeletal anlage of the limb proper. As in the arm chondrifica- tion blocks out the separate skeletal pieces. The formation of car- tilage begins in the chick the sixth day and becomes well marked by the seventh day, when Strasser's "prochondral elements," p. 404, have already disappeared (Johnson, I. c). In the human embryo at six weeks aU the skeletal parts are mapped out in cartilage, except the terminal phalanges, which are still entirely mesenchymal. The plan of structure is essentially the aftme as in the arm at the same age, but the difEerentiation is less advanced; the femur has already neck and trochanter, is slightly curved, and its lower end is enlarged, with two condyles and the incisura intercondyloidea recognizable ; the tibia has broad condyles at its upper end and is suddenly restricted immediately below, and slowly increases in diameter toward the tarsus, to end with a sur- face so oblique as to be nearly parallel with the length of the limb; the astragalus (talus) consists of a lower main portion, the homo- logue of the tibials, and an upper process lying between the tibia and fibula, and homologous with the intermedium; the fibulare (calcaneum) is not so long as the astragalus, and is separated by articular mesenchyma from both the fibula and astragalus, alongside of which last it is situated, but this situation is found to alter grad- ually, beginning to alter in embryos but little over six weeks. In the digital rays the metatarsi and first phalanges only are differen- tiated (Henke and Reyher, 74. 1,230-234). In an embryo of nearly six months the ankle has, I have found,, essentially the adult form. As shown in a vertical section, Fig. 254,. THE LIMBS AND APPENDICULAR SKELETON. irj9 .ip^?m Ast. Cal Vt- Fb the lower ends of the tibia, Tb, and fibula, Fh, are still cartilaginous ; the astragalus, Astr, and calcaneum or os calcis, Cal, are wholly cartilaginous, although penetrated by vessels preparatory to their later ossification. Tlie astraga- lus, Astr, is in quite different relations from those found at six weeks; it underlies the tibia, and shows clearly the subdivision of its tibial articulation into the joint with the main shaft, Tb, and with the internal malleolus, m ; by its external surface it ar- ticulates with the fibula. Fib; by its lower surface with the os calcis. Gal. All of these articu- lations are well differentiated. At its lower internal angle the cartilage of the astragalus is in- terrupted to allow the irruption of the vascular mesenchyma. Ossification. — " The femur is developed from one principal ossific centre for the shaft which appears in the seventh week, and from four epiphyses, the centres for which appear in the following order: A single nucleus for the lower extremity appears shortly before birth, one for the head appears in the first year, one for the great trochanter in the fourth year, and one for the small trochanter in the thirteenth or fourteenth year. These epiphyses become united to the shaft in an order the reverse of that of their appearance. The small trochanter is united abovit the seventeenth year, the great tro- chanter about the eighteenth year, the head from the eighteenth to the nineteenth year, and the lower extremity soon after the twen- tieth year. The neck of the femur is formed by extension of ossifi- cation from the shaft. "The iihla and fibula each present, besides the principal centre for the shaft, a superior and an inferior epiphysis. In the tibia the cen- tre for the shaft appears in the seventh week ; that for the upper ex- tremity including both tuberosities and the tubercle, appears most frequently before, but sometimes after birth ; and that for the infe- rior extremity and internal malleous appears in the second year. The tubercle is occasionally formed from a separate centre. The lower epiphysis and shaft unite in the eighteenth or nineteenth year, the upper epiphysis and shaft in the twenty-first or twenty-second year. In the fibula the centre for the shaft appears rather later than in the tibia; that for the lower extremity appears in the second year, and that for the upper, unlike that of the tibia, not till the third or fourth year. The lower epiphysis and shaft unite about the twenty-first year, the upper epiphysis and shaft about the twenty- fourth year. Fig. 254. —Vertical Section of the Ankle of a Human Embryo of nearly six Months. Minot Collection, No. 109. Tb, Tibia. Fb, fibula; m, intei-nal malleolus of the tibia; Astr, astragalus; Cal, calcaneum (os calcis). X 3 diams. 460 THE FCETUS. " The tarsal bones are ossified in cartilage, each from a single nu- cleus with the exception of the os calcis, which in addition to its proper osseous centre has an epiphysis upon its posterior extremity. The principal nucleus of the os calcis appears in the sixth month of foetal life ; its epiphysis begins to be ossified in the tenth year, and is united to the tuberosity in the fifteenth or sixteenth year. The nucleus of the astragalus appears in the seventh month ; that of the cuboid about the time of birth; that of the external cuneiform in the first year ; that of the internal cuneiform in the third year ; that of the middle cuneiform in the fourth year, and that of the navicular in the fourth or fifth year. "The metatarsal bones siadi phalanges agree respectively with the corresponding bones of the hand, in the mode of their ossification. Each bone is formed from a principal piece and one epiphysis ; and while in the four outer metatarsal bones the epiphysis is at the dis- tal extremity, in the metatarsal bone of the great toe and in the phalanges it is placed at the proximal extremity. In the first meta- tarsal bone there is also to be observed, as in the first metacarpal, a tendency to the formation of a second or distal epiphysis (A. Thom- son) . In the metatarsal bones the nuclei of the shafts appear in the eighth or ninth week. The epiphyses appear from the th^ird to the eighth year, and unite with the shafts from the eighteenth to the twentieth year. The nuclei of the shafts of the phalanges appear in the ninth or tenth week. The epiphyses appear from the fourth to the eighth year, and unite with the shafts from the nineteenth to the twenty-first year." (G. D. Thane in Quain's "Anatomy," ninth edition.) Joints of the Limbs. — Our knowledge of the development of the joints is based chiefly upon the researches of Henke and Reyher, 74.1, Bernays, 78.1, and Hepburn, 89.1; Hagen-Torn's article, 82.1, is chiefly on the histogenesis of the synovial membrane, see p. 431. Where a joint is to be formed the cells become elongated at right angles to the axis of the anlage {synarthrodial stage), the tissue becomes fibrillar and in its midst the cavity appears (diarthro- dial stage) ; chondrification soon extends to the cavity, the articu- lating surfaces thus becoming cartilaginous. The development of the joints is very gradual, but by the end of the third month there are true articulating surfaces, which gradually become better devel- oped; the development of the joints progresses distally, thus the elbow- joint is developed much earlier than the finger-joints; the articulations of the arm appear sooner than the corresponding ones in the leg, thus the knee-joint appears later than the elbow-joint. Bernays, 78. 1, states that the synarthrodial stage of the knee begins in a human embryo of 2 cm., and still persists in one of 3 cm.; in the latter, although there is still no articular cavity, yet the artic- ular ends of the femur and tibia are shaped nearly as at birth — an important observation because it shows that the articulating surfaces are shaped before any free motions can begin. In the three-centi- metre embryo the growth of the lateral tibial condyle has already forced the fibula out of its intimate connection with the femur, which is characteristic both for the earlier stage in man and for ancestral types. By comparative anatomy Bernays has sought to prove that DERMAL BONES. 461 synarthrodia! joints are characteristic of the fishes, imperfect diar- throdial joints of the amphibia, perfect ones of the amniota. Hep- burn, 89.1, adds but little to our knowledge, but his paper is valuable for an admirable synopsis of the stages of joint differenti- ation and of the classification of joints from the embryological stand- point. Hepburn's classification is essentially as follows : Syndes- mosis, synchondrosis, primitive articular cavity, amphiarthrosis, diarthrosis (simple, double with meniscus) ; the diarthroses show the following stages : 1, surfaces become cartilaginous; 2, capsular ligament formed ; 3, other ligaments formed; 4, synovial membrane developed. III. Dermal Bones. It has been long known that not all the bones are prseformed in cartilage, and that some of them, especially of the head, are devel- oped from soft tissue. The latter were known to the older anato- mists as membrane hones. In the years 1845-50 the origin of the membrane bones was actively debated, and at that time the term secondary hones was substituted for the earlier designation, and the terms Belegknochen and Deckknochen were introduced by KoUiker, whose investigations played the principal part in demonstrating that the membrane bones are developed by the direct ossification of young connective tissue, or — as we should now say — of mesenchyma. Those who wish to follow this discussion are referred to Kolliker, 50.2, where references are given to various authorities of the time, and also to KoUiker's "Bericht der Zootom. Anstalt in Wiirzburg," and his "Entwickelungsgeschichte," 2te Aufl., 463. The dermal bones of the head may lie close against the cartilage (or bone) of the pri- mordial skull, and in that case are often called splint bones or splenial hones. In the lower vertebrates the membrane bones acquire a greater development than in higher forms, and in certain ganoids and tele- osts are developed over nearly the entire body, whereas in the amni- ota they are confined to the head. O. Hertwig's brilliant researches, 74.1,2, 76.1, 79.1, have demonstrated that the dermal bones are homologous with the plates formed by the fusion of epidermal teeth or so-called placoid scales. The placoid scales are true teeth developed in the skin and supported by a base of bone ; by the fusion of adjacent bony bases we may have an osseous plate developed in the cutis. In tailed amphibia several of the membrane bones arise as dentiferous plates, but later in the development the teeth are resorbed leaving merely the bony plate, but in anoura the homologous bones are developed without teeth being formed at all. The inevitable conclusion from these facts is that the dermal skeleton has been evolved through three principal stages : 1, scattered independent dermal teeth (placoid scales) ; 2, teeth -bearing plates formed by the fusion of the expanded bases of adjacent teeth (exo-skeleton) ; 3, membrane bones developing without teeth appearing (dermal bones of tailless amphibia and amniota) . The plates or bones of the dermal skeleton are not the same throughout the vertebrate series; among the fishes there are numer- 462 THE FCETUS. ous modifications, the liomologies of v/hich have not yet been thor- oughly elucidated ; in the amphibia we encounter all the elements of the dermal skeleton of the amniote head, and comparative anatomists have succeeded in homologizing some of these elements with plates in fishes, but as much remains to be done, and as the conclusions have not hitherto been based upon much embryological evidence, I shall not attempt to enter into these difficult discussions. Typical Dermal Bones of Amniota. — In amniota the dermal bones are confined to the skull and face. There are, 1, four pairs of bones on the dorsal side, namely, the nasals overlying the olfactory chambers ; the frontals overlying the anterior part of the brain cav- ity; the parietals overlying the middle part of the brain cavity, and the interparietals overlying the anterior part of the occipital region; the frontals, parietals, and interparietals, together with the supra- occipital, constitute the roof of the skull; when the cartilaginous skull spreads upward it goes under the territory of the frontals, parietals, and interparietals, and when it ossifies it may contribute to a greater or less extent to the bones in question, so that they are not exclusively membranous in origin (Dursy). Between the parie- tals and supra-occipital is the interparietal; 2. The small lachrymals situated between the nasals, frontals, and the eye on each side (in certain reptiles there are additional periorbital bones), and the squamosal, occupying the space between the parietals, ali-sphenoids, and occipitals, and overlying that portion of the mandibular bar which forms the quadrate of reptilia (incus of mammalia) ; the squa- mosal is perhaps the homologue of the prae-opercular of fishes, as maintained by Huxley, or perhaps of the ganoid supratemporal as suggested by Balfour, "Comp. EmbryoL," II., 593. 3. The bones associated with the mandibular branchial bar ; these are, first, those associated with the palato-quadrate bars and appearing in the roof of the mouth, the vomer, palatines, and pterygoids; second, a series associated with Meckel's cartilage, and consisting primarily, accord- ing to comparative anatomists, of three dermal bones, the distal den- tale, the smaller articidare, and in the angle between these two the small angulare: but in mammals there is only a single bone devel- oped from the mesenchyma around Meckel's cartilage, which evi- dently represents the dentale, but whether or not it also represents the articulare and angulare ha? not been definitely settled. 4. The series associated with the maxillary processes, four on each side form- ing a row ; beginning at the ventral end of the process these four bones are the prce- max ilia, maxilla, jugal, and quadrato-jugal. 5. The median par a- sphenoid, which is developed in the roof of the mouth in many fishes (but not in elasmobranchs or marsipobranchs), in amphibia, and in sauropsida, in which last it is less important and becomes indistinguishably fused with the sphenoid in the adult ; in mammalia it has not been found, though probably morphologically present in the sphenoid — a probability which it would be worth test- ing by a special investigation. 6. The tympanal bone formed around the drum of the ear. The Dermal Bones in Man.— The numerous dermal bones, mentioned as characteristic for the amniota at large, have all been identified in the adult human skull, except the articulare, angulare, DERMAL BONES. 463 quadrato-jugular, and para-sphenoid. The four bones mentioned are, however, all probably represented by definite parts as follows : the interparietal bj^ the upper median portion of the supra-occipital ; the articulare and angulare by parts of the adult mandible ; the quad- rato-jugular by one of the ossificatory centres of the jugal, and the para-sphenoid by part of the sphenoid. The nasals, parietals, lach- rymals, vomer, and jugal remain independent bones, while the frontals and palatines are also independent except that each pair forms but a single bone. On the other hand the squamosals, pterj-- goids, dentals, are united with certain parts of the primordial skull. Finally the prse-maxillaries and maxillaries fuse into a single bone, of which the part bearing the four upper incisors corresponds to the prae-maxillaries. A tabular view of the homologies of the human skull is given on p. 465. The following data afford additional information concerning the development of the single dermal bones. 1. Nasals are each ossified from a single centre which appears about the eighth week. 2. Frontal is ossified from two centres, one for each frontal ap- pearing about the seventh week. At birth the frontals are still en- tirely distinct, but they become united during the first year after birth by the median " frontal" suture, which usually becomes obliter- ated by osseous union taking place from below upward during the second year, but not infrequently the suture persists throughout life. 3. Parietals are each ossified from a single centre which appears in the site of the parietal eminence about the seventh week. The eminence is very conspicuous in tlie young bone and gives a marked character to the form of the skull for a number of years in early life. 4. Interparietals are represented by the iipper pair of centres of the supra-occipital region ; these centres appear during the seventh week in the mesenchyma overlying the supra-occipital cartilage. The interparietals usually unite with the true occipitals, but oc- casionally they remain distinct and are then separated from the supra-occipital by a suture running transversely from one lateral angle of the occipital bone to the other. 6. Lachrymals are each ossified from a single centre, which ap- pears about the eighth week. 0. Squamosals are each ossified from a single centre, which ap- pears in its lower part about the seventh or eighth week; ossification spreads upward in the squamosum proper, and outward into the zygo- matic process. At birth the squamosal is still separated from the periotic capsules, but during the first year bony union is effected and the squamosal becomes a part of the temporal bone of the adult. 7. Vomer is ossified from a single nucleus appearing at the hinder part about the eighth week. From this nucleus two laminae are de- veloped, which pass up on either side of the median line and embrace the lower part of the cartilaginous internasal septum. These laminae gradually coalesce from behind forward till the age of puberty, thus foi-ming a mesial plate, with only a groove remaining on its superior and anterior margins. 8. Palatine is ossified from a single centre which appears in the 46i THE FCETUS. seventh or eighth week at the angle between its horizontal and ascending parts. 9. Pterygoids are each ossified from a single centre which appears during the fourth month ; during the fifth or sixth month the pterj'- goids unite with the ossified pterygoid processes (future external pterygoid plates) of the ali-sphenoids and thus become the internal pterygoid plates of the adult basi-sphenoid. 10. Prce-maxillaries have been studied by Th. KoUiker, 83.1; they ossify later than the maxillaries and appear just before the pal- ate fissure closes, and after the fissure has closed they are found imited with the maxillaries so that the period of their independent existence is very short; but in the ninth week traces of the primitive division are still present, and even these traces disappear by end of the tenth week. The prae-maxillaries carry the four upper incisors. A spe- cial interest attaches to these bones because their homologies in man were ascertained by Goette. 11. Maxillaries begin to ossify toward the end of the second month and offer the peculiarity of starting from several spots, which, however, speedily fuse and cannot be regarded as separate centres. This peculiarity was first recorded by Beclard, 20.1, and his obser- vation has been confirmed by Rambaud et Renault, and more re- cently by Callender, 70.1, 163. As stated above, the maxillaries and prsB-maxillaries are united before the tenth week. 12. Jugals, or malars, begin to ossify about the eighth week. According to Rambaud et Renault, ossification begins from three points, which are found united by the fourth month. 13. Mandible. The mandible of the adult is a compound bone, for it includes both the dermal bone and the ossified lower ends of Meckel's cartilage, most of which, however, is resorbed, and it is further peculiar in having cartilage developed at the ends of both the coronoid and condylar processes. The two mandibles are distinct at birth, but during the first year their lower or ventral ends unite, but in a pig embryo of two and a half inches Parker (" Morphology of the Skull," 390) describes the ends of Meckel's cartilages as united, and it is probable that the cartilaginous jaws of the human embryo are similarly united. The development of the human man- dible has been studied by Masquelin, 78. 1; in an embryo of 5 cm. the cartilage of Meckel is entirely surrounded by mesenchymal bone, and in embryos of 17 cm. there are only slight calcified remains of the cartilage, except in the lower ends near the symphysis, where, as shown by KoUiker, the cartilage participates in the ossification of the mandible ; the cartilage of the coronoid process was found in embryos of 7.5 and 9.5 cm., and in the later cartilage along the alveolar border ; the cartilage of the condyle is developed still earlier ; the three cartilages upon each mandible undergo direct ossification. Strelzoff, 73.1, was led by the observation of these cartilages to maintain that the entire jaw is preformed in cartilage, but that this view is erroneous was demonstrated .in an admirable paper by J. Brock, 76. 1. It is evident that the accessory cartilage of the man- dible is morphologically distinct from that of the primordial skeleton. 14. Tympanals develop during the third month each from a cen- tre which appears in the lower part of the external membranous MOEPHOLOGY OP THE SKULL. 465 wall of the tympanum and extends upward until a nearly complete bony ring is formed, inclosing the tympanic membrane ; before birth the ends of the open ring become united with the squamosal, and thus incorporated in the great temporal bone of the adult. The Fontanelles. — These are membranous intervals between the incomplete angles of the parietal and neighboring bones, in some of which movements of the soft wall of the cranium may be observed in connection with variations in the state of the circulation and respiration. They are at the time of birth six in number, two me- dian, anterior and posterior, and four lateral. The anterior fon- tanelle, situated between the antero-superior angles of the parietal bones and the superior angles of the ununited halves of the frontal bone, is quadrangular in form and remains open for some time after birth. The posterior fontanelle, situated between the postero- superior angles of the parietal bones and the superior angle of the occipital bone, is triangular in shape. It is filled up before birth, but the edges of the bones being united by membrane only are still freely movable upon each other. The lateral fontanelles, small and of irregular form, are situated at the inferior angles of the parietal hones. The fontanelles are gradually filled up by the extension of ossification into the membrane which occupies them, thus complet- ing the angles of the bones and forming the sutures. The closure, especially of the posterior and lateral, is often assisted by the devel- opment of Wormian bones in these situations. All traces of these unossified spaces disappear before the age of four years. IV. Morphology of the Skull. We are now in a position to consider several questions concerning the skull as a whole. What is presented on these questions I have divided under the following headings into sections: 1. Homologies of the bones of the human skull. 2. Relations of the primary and secondary skull. 3. Position of the facial apparatus. 4. Signifi- cance of the trabeculse cranii. 5. Theories of the skull. The de- tailed history of each element of the skull is given, as fully as practicable, in the preceding pages. Homologies of the Bones of the Human Skull.— These have been discussed in the preceding pages of this chapter, but it will be convenient to present the conclusions arrived at in a tabular form: HUMAN. TYPICAL AMNIOTE. A. Cranial.— Ethmoid and turbinals. Ethmoid and turbinals. Prse-sphenoid. Prae-sphenoid, orbito-sphenoids (alse) minores), and pterygoids. Basi-sphenoid. Basi-sphenoid, ali sphenoids (alas ma- jores) (? and para-sphenoid). Occipital. Basi-occipital, ex-occipitals, supraoc- cipitals, and interparietal. Temporal. Periotic capsule (pro-otic, opisthotic, epiotic) squamosal, annulus tym- panicus, and styloid process (upper end of hyoid bar) . 30 466 THE FCETUS. HUMAN. TYPICAL AMNIOTE. B. Facial,. —Nasals. Nasals. Lachrymals. Lachrymals. Jugal. Jugal. Superior maxillary. Prsemaxillse and maxillae. Vomer. Vomer. Palatine. Palatines. Mandible. Dentale (? Articulare and angulare) and lower part of Meckel's cartilage. Relations of the Primary and Secondary Skull. — Com- parative anatomy and embryology alike teach us that we must attribute to the skull a double origin, or rather that there are two skulls, one outside the other, which were primitively distinct from one another, but in the progress of evolution from the earliest fish type to the higher mammalia the union between the two skulls be- comes more and more intimate. The inner skull is what is known as the primordial skull, with which I include the branchial skeleton ; the outer skull comprises the series of dermal bones of the cranial and facial regions. The primary skull appears first as the continuation into the region of the head of the axial mesenchymal skeleton, which in the neck and rump is the anlage of the vertebrae. That the mesenchymal skull represents in part, at least, a series of vertebrae is certain, and we find it sending dorsal outgrowths to inclose the brain just as the true vertebrae cover in the spinal cord. The mesenchymal skull also extends in front of the hypophysis, where it produces the trabeculae cranii. What little can be surmised concerning the original homolo- gies of this part of the skull is given in the section on the trabeculae, p. 434. The mesenchymal skull grows so as to completely incase the brain and partially incase the olfactory chambers. While it is growing six centres of chondrification appear in it : namely, two tra- becular, two parachordal, and two periotic; each centre forms a cartilage, which is extraordinarilj^' uniform in shape and relations throughout the entire vertebrate series; the six cartilages remain distinct for a very short time only ; the two trabeculae unite first, the two parachordals next, third the united parachordals (or occipital) coalesce with the periotic capsules and later with the caudal ends of the trabeciilae, thus forming a large floor of cartilage under the brain. In the lower forms chondrification spreads until the entire primary skull becomes cartilaginous, and it is in this condition we find the skull in many of the fishes. In the amphibia and amniota there is a progressive reduction of the cartilaginous skull by which its development as a roof over the brain is more and more diminished. This reduction leaves an open- ing as it were on the dorsal side, and at once increases the impor- tance of the covering dermal bones — frontals, parietals, and inter- parietals. In Sauropsida the opening is larger than in amphibia, and in the mammalia there is further progressive increase in size, as shown by Parker's observations, the opening being larger in pigs than in insectivora and edentates. In mammals there is a further loss, which is not found in other classes, namely, an absence of chondrification in the region between the ali-sphenoids and periotic capsules, by which the importance of the squamosal — the dermal MORPHOLOGY OF THE SKULL. 467 bone of the region — is increased; see W. K. Parker, 86.1, 8, who speaks of the disappearance of the cartilage under the squamosal as "the true diagnostic mark" of the mammalian chondrocranium. Reduction of the cartilages of the branchial skeleton also progresses from the lower to the higher vertebrates. This shows itself in mammals not only in the total disappearance of the cartilages of the fourth and fifth arches, but also in the partial disappearance of the thyro-hyoid bars and the imperfect development of the hyoid bars. It shows itself further in the reduction of the mandibulars, for not only is the greater part of Meckel's cartilage resorbed as in all am- niota, but also the palato-quadrate is very much reduced. As the palato-quadrate is an important part of the skull in the amphibia, the palatines and pterygoids appear as true splint bones, whereas in mammalia they have greater independence. It is clear from the above that the evolution of the mammalian skull has depended to a large extent upon the reduction or partial degeneration of the inner skull, or primordial chondrocranium. The secondary or outer skull is not so old as the inner skull, and originated in the higher fishes as a series of dermal bony plates, which overlaid the primary skull, and probably formed a nearly complete case for the head, including the face. The definite arrange- ment of the plates, as perpetuated and modified in mammalia, appears in the amphibia, and was perhaps evolved during the transition from the fish to the amphibian type. The dermal plates (membrane bones) may either remain as splint-bones, as for instance is the case with the vomer, or they may coalesce with the underlying portions of the chondrocranium, as for instance occurs with the interparietals in primates, or they may remain where the cartilage disappears beneath them, as for instance the frontals. Already in the amphibians the co-ordination and fusion of the inner and outer skulls into one com- plex skull is established, and in the amniota the welding together is carried still further, and the elements of the outer skull, i.e. the dermal bones, acquire increased importance as the inner skull, i. e. chondrocranium, is reduced. In brief, the evolution of the mammal- ian skull has depended largely upon increased morphological promi- nence of the dermal bones. If we designate the formation of the chondrocranium as the first stage, and the formation of the dermal bones as the second stage in the evolution of the skull, we may designate the ossification of the primordial chondrocranium as the third stage. As to what caused that ossification, we have not even an hypothesis, and we are equally in the dark as to how the number of separate bones, or centres of ossification, was determined. It is noteworthy that the number of the primordial bones is extraordinarily constant. Finally, let me emphasize the fact that, given the full number of bones, there is a sustained tendency to reduce them_ by fusion. The number of skull bones is less in the amphibia than in the teleosts, in edentates than in amphibians, in man than in edentates. A thorough comparative study of the number of the skull bones is much to be desired. Position of the Facial Apparatus.— Owing to the head-bend of the embryo, the oral invagination, or mouth cavity, is brought 468 THE FCBTTJS. between the fore-brain and the heart, and upon the ventral surface, and this is the permanent position in the sharks. If we follow through the vertebrate series, or the development of an amniote, we find in either case a steady increase in the region of the olfactory and oral invaginations, in consequence of which it projects more and more, and further by a throwing of the whole head upward the face is brought forward and projects in front of the brain. In man this condition is again modified : first, because the upright position renders it unnecessary to bend the head as in quadrupeds, and, there- fore the head is left facing ventralward; second, because the enormous development of cerebral hemispheres has rendered an enlargement of the brain cavity necessary, and this enlargement has taken place by extending the cavity over the olfactory regions as well as by enlarging the whole cranium ; third, because the develop- ment of the facial apparatus is arrested at an embryonic stage, the production of a long snout being really an advance of development (Minot, 35), which does not take place in man. Significance of tlie Trabeculae Cranii. — ^Conceming the nature of the trabeculae we have no satisfactory conceptions. They are a temporary stage of the chondrification of the mesenchymal skeleton in front of the notochord, and it seems to me as improbable that they have any important morphological significance, as it is improbable that the rounded form of the bony centre of a half-ossi- fied vertebra has any important meaning. It cannot be too strongly insisted upon that the morphological condition is determined by the shape of the mesenchymal anlage, of which the trabeculae are merely a part. So far as I am aware, not a single investigator has described this anlage accurately and fully. I consider it not improbable that the axial perichordal mesen- chymal skeleton sends an outgrowth past the hypophysis to inclose the fore-brain, and that, assuming that the infundibulum marks the true anterior limit of the medullary canal, the trabecular anlage is not a prolongation of the floor of the cranium, but an upgrowth, which owing to the head-bend has come to lie in the line of the cranial axis. Theories of the Skull. — It was noticed a long time ago that the skull has resemblance to vertebrae; the skuU has the greatest thickness on the ventral side of the brain and arches over the central nervous system, and thus possesses two of the chief characteristics of the vertebrae. It was, therefore, natural to seek to compare the skull homologically with vertebrae. It is said that during the eighteenth century this comparison acquired greater prominence and was definitely formulated by Vicq d'Azyr.* These comparisons of Vicq d'Azyr and others proceeded upon a false basis, and it was not until 1872, when Gegenbaur, 72.1, opened an entirely new method of solving the morphology of the head, that correct views began to be formed. Another great stride was made by Froriep's observations on the development of the occiput, p. 429. I have placed what I have to say under the three headings of Vicq d'Azyr's the- ory, Gegenbaur's theory, and Froriep's law. 1. Vicq d'Azyr's Theory.— According to this theory the skull consists of several vertebrae. Whether d'Azyr really originated it, * I have not succeeded in finding anything in Vicq d'Azyr's "CEuvres"to justify this state- ment. MORPHOLOGY OF THE SKULL. 469 I cannot say. It was taken up by Oken, who is often quoted as the founder of it, and later also by Goette, who by some authors has been cited as the father of the theory. The history of the theory and of the modification it underwent is given by R. Virchow (" Goette als Naturforscher ") . One of the earliest suggestions of the vertebral theory is that of Burdin, independently made about the same time by Heilmeyer. These authors compared the skull to a single complex vertebra. Oken conceived that there were four cranial vertebrae, and this was the notion most in favor until 1858. Goette counted six vertebrae, of which three belonged to the facial apparatus. As to the number of these supposed vertebrae there is a very extensive literature, which possesses an interest purely historical. Let it suffice, therefore, to state aphoristically that three vertebras were advocated by Spix, Meckel, Burdach and Carus; four by Oken, Bojanus, and Owen ; six by McClise; seven by Geoffrey. The death-blow to this long-lived error was dealt by Huxley in his Croonian lecture delivered in 1858, 58. 1 — a great achievement, for it at once terminated the history of the old vertebral theory of the skull, and paved the way for Gegenbaur. Gegenbaur's Theory. — In 1872 Gegenbaur published his great work, 72.1, on the cephalic skeleton of Selachians, in which he took the ground that the skull does not represent a series of vertebrae, but that it arose out of the axial or perichordal skeleton before dis- tinct vertebrae were formed in the axial region ; he further main- tained that the head includes a number of segmentig, which he sought to ascertain by determining the segmental arrangement of the cranial nerves. This was a great step and in the right direction. F. M. Balfour, 78.3, was, I believe, the first to endeavor to trace out the actual number of segments (mesoblastic somites) in the head of embryos. A vast amount of labor has been expended by subse- quent writers in investigating the development of the cephalic myo- tomes and cranial nerves, but much remains to be done before the morphological constitution of the head shall be understood, but we are already in a position to say that Gegenbaur's thesis — that the primary or inner skull. is developed from the axial skeleton but not from vertebrae — is correct except as regards the hypoglossal region. For further observations on the segmentation of the head see Chap- ter XXVI. Feoriep's Law. — Froriep's investigations, p. 429, have demon- strated that the skull has extended itself, in the amniota at least, by the annexation of true vertebrae, corresponding to segments of which the hypoglossus represents the nerve. The head has grown at the expense of the neck. Present Theory of the Skull. — The primary skull was de- veloped out of the axial (perichordal) skeleton, in the region of the brain, where the dorsal and ventral nerve roots are not united into a single nerve for each segment ; the primary skull has grown at least in the amniota by the annexation of several cervical vertebrae; a secondary skull was developed outside the primary cartilaginous skull by the formation of dermal bones. In the higher forms the primary skull partly disappears ; what remains, together with the secondary or dermal skull, constitutes the actual skull of the adult. CHAPTER XXI. THE MESOTHELIAL MUSCLES. The muscle fibres fall into two main classes, the smooth or mesen- chymal fibres, which have been already considered, p. ill, and stri- ated or mesothelial muscles. The latter fall into three groups; 1, the skeletal muscles ; 2, the branchial muscles ; 3, the cardiac mus- cles. The first are developed from the epithelial muscle plates, the origin of which from the mesothelial primitive segments has been already described, p. 205 ; the second are developed from the meso- thelium of the branchial coelomatic cavities (head cavities of Balfour) see p. 478 ; the latter are developed from the mesothelial wall of the heart of the embryo and constitute the so-called " muscular heart " {Muskelherz), see p. 227. The Segmental or Skeletal Muscle Fibre. — Remak, 50.1, was the first, if I am not mistaken, to show that the primitive seg- ment, or as he termed it the " protovertebra, " forms both the anlage of the axial mesenchyma and of the muscles ; he also thought that the " protovertebra" formed the spinal ganglion, an error which was corrected by His, 68.1. To the myotome, or the two layers of the mesothelium remaining after the differentiation of the periaxial mesenchyma (Van Wijhe's sclerotome), Remak applied the term Ruckenplatte. After Remak (1850) followed a series of investiga- tions and discussions as to the histogenesis of the striated muscle fibre. The chief differences of opinion were as to whether, as origin- ally maintained by Remak, each fibre is developed from a single cell, or, as suggested by Theodore Schwann, out of the fusion of several cells. The latter view was advocated by Margo, 59.1, in 1859; Margo studied the muscle corpuscles, terming them sarcoplasts, and regarding them as so many separate cells which had united to form the muscle fibre. That Remak was right was maintained by KoUi- ker in 1857, 57. 1, on the basis of his own observations, and also by Max Schultze in a masterly essay, 61.1, which at the same time laid the foundation of the modern doctrine of cells, and anticipated Heitzmann's observations on the union of the cells by over twenty years. In the same year, 1861, appeared Deiters' paper, 61.1, and the year after, F. E. Schulze's, 62.1, who together with Max Schultze conclusively established Remak' s opinion as correct. Never- theless we find the Schwann-Margo hypothesis reappearing from time to time, although it has never had any sound observational basis to rest upon. A synopsis of various papers upon the development of striated muscle fibres is given by Calberla, 76.1, and more fully by G. Born in his dissertation, 73.1. That the striated mus- cles have an epithelial origin was first emphasized by the two Hert- wigs, 81.1, 61-66, who demonstrated at the same time that only the THE MESOTHELIAL MUSCLES. 471 ^^ inner layer of the myotome forms muscle, not both plates as had been wrongly stated by Balfour, "Comp. Embryology," II., to be the case in elasmobranchs. Since then it has been ascertained beyond question that the outer layer gives rise to the dermis (compare p. 206) J Kaestner's contrary conclusions, 90.1, being attributable, in my judgment, to his imperfect observation. The single muscle fibre arises from a single epithelial (mesothelial) cell of the muscle plate or inner wall of the myotome. In the am- phibia each cell elongates in a direction parallel with the axis of the body until, as shown by F. S E. Schultze, 62.1, it stretches the entire length of ^ the segment ; it seems to me that each cell extends the entire length (cephalo-caudal) of the segment in sharks and chick embryos also, but I have not studied the point sufficiently. Paterson, 87.1, asserts that in chicks the cells lengthen but remain shorter than the segment. In amphibia the cells are crowded with yolk granules, which, however, are gradually resorbed ; thus in the frog they at first hide the nuclei, but by the fourth day are sufficiently reduced to allow the nuclei to be seen easily in the fresh unstained specimen (Calberla, 76.1); in amniota, on the other hand, there are exceedingly few yolk grains left in the muscle plate. The first evidence of striation appears in the frog toward the end of the fifth day, on 07ie side of the cell. Fig. 225, as first recorded by F. E. Schultze, 63. 1, and since frequently confirmed {e.g., by Calberla and Ranvier, "Traite technique d'Histologie," 516). The side upon which the striation first appears has been observed in elas- mobranchs by C. Eabl, 89.2, 239, to be the side toward the notochord, or farthest from the cavity of the myotome. In Petromyzon, A. Goette, 90.1, 50, the fibrillBB are found to form a peri- pheral layer so very early that it is doubtful whether they first appear on one side of the cell only or not. The striation continues to develop until it passes completely around the cell, forming a peripheral layer (Deiters, 61.1, KoUiker, " Ge- webelehre," 6te Aufl., p. 401) as illustrated in Fig. 256. At about this time, perhaps sooner in some forms and later in others, the nucleus divides, and by repetitions of the divisional process the cell soon becomes multinucleate. C. Rabl, 89.2, 242, directs attention to the fact that the nuclei of the muscle-plate in sharks stain more lightly than the mesenchymal nuclei and contain an elongated chromatine granule ; in the chick I have observed the same nuclear peculiarity. Later the nuclei lose this main granule and have instead a number of smaller ones, Fig. 256, m?i. The muscle fibre acquires its membranous sheath, sarcolemriia, some time later. As to the exact time I have found no positive data; % Fig. 255. — Isolated Muscle Fibres of a Frog Embryo. A, Showing yolk grains (partly re- sorbed) and nucleus ; B, with, the muscular stria- tion ^ust appearing on one side. 472 THE FCETtrS. but authorities are agreed that the fibre remains naked for a consid- erable period. During the early stages of their differentiation, the muscle fibres retain the epithelial arrangement, that is, remain closely packed ; not long after the appearance of the fibriUas and stri- ation, the fibres begin to separate and connective tissue grows in between them. During their separation the fibres become massed in bundles instead of in epithelial layers. The central portion of the young muscle fibre is granular, and contains not only the nuclei and the remains of the yolk material, but also a considerable quantity of glycogen (Eanvier, "Traite tech- nique d'Histologie," 515) . As this substance is very readily dissolved out, it is probable that the clear empty appearance of the central Fig. 2o6 — a, Transverse; B, Longitudinal Section of Muscle Fibres in the Neck of a Human Embryo of sixty-three to sixty-eight Days. Minot Collection, No. 138. m.m m Muscle fibres; nwi, muscle fibres showing the central nuclei; mes, mesenchymal nuclei. A X about 750 portion of the fibres, which is so striking in sections of hardened em- bryos, see Fig. 256, A, is due to the loss of the glycogen. The mantle of striated muscle-substance gradually increases until the whole fibre is fibrillated and the muscle no longer appears hollow. The time at which the muscle fibres become " solid " varies from embryo to em- bryo and from muscle to muscle. In the human embryo at the end of the fifth or beginning of the sixth month most of the fibres of the upper extremity are solid, but it is not until the seventh month that those of the lower extremity become solid, W. Felix, 89.1 232. The nuclei have at first an axial position, but toward the end of the third month some of them lie in the mantle, compare also p. 474. The size of the fibres is smaller in the embryo than in the adult but Felix, 89.1, 233, points out that up to latter part of the third THE MESOTHELIAL MUSCLES. 473 month in man the fibres increase in size, many having at two and one-half months a diameter of from 13-1 9m, but later they are smaller owing probably to the division of the fibres, and it is not until birth that the same diameter is again reached by the single fibres. Fibrillce. — Before discussing the origin of the fibrillsB it is neces- sary to remove an unfortunate confusion which has prevailed in the use of the term. By fibrillae is sometimes meant the longitudinal threads of protoplasm, but more often is meant the material between adjacent longitudinal threads. Corresponding to the two usages of the word "fibrilla," there are two essentially different conceptions of the structure of the adult muscle fibre. According to the older view, which is currently repeated in text-books of anatomy and his- tology, the " primitive fibrillee " into which a muscle fibre may be mechanically divided under certain conditions are the contractile portions of the fibre ; the ends of the " primitive fibrillae" are the so- called Cohnheim's areas, and the " sarcous elements" are the divisions of the "primitive fibrillae." Henle ("AUgemeine Anatomic," 1841, p. 580) recognized that there was substance left between the fibrillee; later Leydig (Muller's Arch., 1856, 156) and Kolliker (Zeit. Wiss. Zool., VIII., 316) pointed oxit its general occurrence. Max Schultze, 61.1, 3, shows that this material was the derivative of the proto- plasm of the muscle cells. L. Gerlach * was the first, so far as I know, to demonstrate that this interfibrillar material formed a re- ticulum, but he regarded it as the prolongation of the nerve. The reticulate structure appears to have been recognized also by G. Thin {Quart. Journ. Microsc. Sci., 1876, XVI. 351). No special signifi- cance seems to have been attributed to all these observations until 1881, when Retzius, 81.2, proposed the new conception of the structure of muscle fibres, according to which the material between the so-called "primitive fibrillse" is the essential part of the fibre; this material is part of the protoplasmatic network of the cell which makes the muscle. According to Eetzius, the essential feature of the muscle fibre is the peculiar and characteristic arrangement of this network, by which the striation is caused. The fibrillae of embryol- ogists are threads of protoplasm and are not the same as the "primi- tive fibrillee " of histologists, but are characterized by staining read- ily. That the fibrillation was developed by a metamorphosis of the protoplasm of the young muscle cell has long been the conception of embryologists, see for example Max Schultze's article, 61.1, pub- lished in 1861. L. Bremer, 83. 1, was the first to place this concep- tion upon a firm basis of observation, by tracing out further than had been done before the transformation of the protoplasmatic net- work of the developing fibre. Retzius' results were extended and made the basis of a theory of the structure and contraction of the muscle fibre by B. Melland, 85. 1, and C. F. Marshall, 87. 1, 9a 1 both working in A. Mihies Marshall's laboratory at the Owens Col- lege; compare also Biitschli und Schewiakoff, 91.1. This series of investigations render it necessary to accept Retzius' view— although Kolliker in the sixth edition of his " Gewebelehre " throws the weight of his great authority ag ainst it. As it now stands Retzms view * Gerlach, "DasVerhaltniss der Nervenzu denMuskeln der Wirbelthiere," Leipzig, 1874. See also Arch. £.' Microsc. Anat., xiii., 1877, 397. 474 THE FCETUS. may be summarized thus: fibrillse and sarcous elements are post- mortem effects due to the cleavage of the matrix ; the muscle fibre really consists of a homogeneous matrix which is traversed by a very regular reticulum, made of longitudinal threads connected at regu- lar intervals by transverse threads, corresponding in position to Krause's membrane; at the nodes, where the cross and long threads unite, there are little thickenings. The thickenings correspond to the balls of Schafer's dumb-bells, the handles of which are the long threads, compare Schafer, Philosophical Transactions, 1873. Be- tween every two Krause's membranes are numerous fine cross threads, which cause the appearance of the dark bands and therefore of the transverse striation. The transformation of the reticulum of the multinucleate cell of the myotome into the network of the adult muscle fibre has been most carefully described by L. Bremer, 83.1, whose results may be summarized as follows : The nucleus of the muscle fibre, together with the protoplasm surrounding it, constitutes the so-caUed muscle corpuscle ; the corpuscle is much more prominent in young than in old muscle, for its protoplasm is gradually differentiated into muscu- lar substance ; a small number of corpuscles enter into the formation of each fibre ; the substance of the muscle forms a network, which was first partially recognized by Heitzmann (Wien. Sitzungsber. , XVII., Abth. 3, 1873) ; the meshes of this network appear polyg- onal in transverse — rectangular in longitudinal, sections; the net- work is a modification of the protoplasmatic network of the cor- puscles, and is so arranged that there are alternating rows, both transverse and longitudinal, of fine knots and large knots (corre- sponding to the fine and broad striae) ; the fine knots are connected by fine threads, and the large knots by coarse threads ; hence there is a fine and a coarse net. Multiplication of Muscle Fibres. — That the muscle fibres multiply during embryonic life can hardly be questioned at present. Two methods of accounting for the multiplication have been advo- cated, the first that it is effected by the intervention of sarcoplasts (Margo) , the second that it is by a direct longitudinal fusion of the fibre (Weismann, 61.1). I consider the latter view the correct one. 1. Margo's Theory.— BremeT's results, 83.1, on this question are as follows : The post-embryonic multiplication of fibres takes place by means of the structures described by Margo (59.1, 239) under the name of sarcoplasten; these are lines or chains of muscle cor- puscles, united by the protoplasm net, and derived by proliferation from the corpuscles of the original fibres ; the sarcoplast gradually separates from the parent fibre, undergoing muscular differentiation meanwhile, and also becoming connected with the nerve. The growth of the fibre is initiated by a multiplication of the corpuscles ; the sarcolemma is not present at first, but appears later, being prob- ably formed by the fused cell membranes of the corpuscles, to which appears to be added a coat of connective tissue, and also around the motor plate between the two sarcolemmic coats appears an extension of Henle's sheath of the nerve. Paneth has recently renewed, 85.1 Marge's observations, 59.1, giving a careful description of the sar- coplasts and maintaining that they are the agents of fibre multipli- THE MESOTHELIAL MUSCLES. 475 cation. Sigmund Mayer, 86.1, attacked Paneth, because he found muscle corpuscles abundant in the fibres of the tail in tadpoles during the process of resorption, and hence concluded thatt he corpuscles are muscle destroyers {sarcolytes) . This opinion has been accepted by Barfurth, 87.1, but the mere presence of the corpuscles, while the muscle fibres are becoming destroyed, is, as Paneth justly replied, 87.1, no evidence whatever that they have a sarcolytic function. There remains, however, another hypothesis which has been advanced by Felix, 89.1, 253, namely, that the so-called sarcoplasts represent muscle fibres partly degenerated. Felix's interpretation is the one which has most commended itself to me. 2. Weismann's Theory. — Felix's conclusions are, that from the middle of the third month until the end of foetal life there are, in the muscles, fibres with multiplied nuclei, which are arranged in rows. These fibres with multiple nuclei are of two kinds, those with a single and those with several rows of nuclei. In the first kind the nuclei are central, color deeply, lie transversely, and differ but little from one another ; fibres of this kind do not divide though they may grow; some of them degenerate and form Margo's sarco- plasts. The second kind of fibres have several rows of nuclei in the mantle or fibrillar layer ; in the middle part of the rows the nuclei are closely crowded and compressed into all possible forms; this crowding probably marks the centre of proliferation. The fibre divides into daughter fibres, one for each row of nuclei. The fibre becomes enveloped in a sheath, rich in nuclei and vessels, and this sheath persists while the fibre is dividing ; afterwards it disappears. The daughter fibres may also divide, but apparently usually into two only. The areas, in which the nuclei are crowded together, have long been known, though imperfectly described. They are usually termed Huskelknospen or Muskelspindel by German writers, and they mark the point where the union with the nerve is established. They were known to Weismann in 1861, 61.1, and were shortly after described by Kolliker, 62.1, in amphibia. Von Franque, 90.1, records some observations upon them and gives references to the scattered observations upon them made by a number of writers. The Muscle Plates. — The development of the muscle plates has already been described. There is unfortunately little to be added at present concerning their later history. When the outer leaf of the myotome is changed into the mesenchyma of the dermis, the cells nearest the muscle plate on all sides retain for a con- siderable period their epithelial arrangement, and appear to act as a growing layer, and presumably contribute both to the mesen- chyma on one side (compare Fig. 257) and the muscle plate on the other. Certainly the muscle plate continues to grow in all directions, but most rapidly dorsal ward over the rpedullary canal and ventralward into the somatopleure. At the same time the muscle plates not only lengthen out as the whole trunk lengthens, but each one grows forward under the one in front, and thus is produced the stage so characteristic of fishes, with the mus- cular segments oblique and the hind border of each overlapping the 476 THE FCETUS. segment next behind. That the imbrication is produced as stated seems to me clear from the study of shark embryos, in which the original position of the segment is indicated by the nerves, ganglia, and inter-segmental arteries ; the hind edge of the muscle plate coin- FiG. 257. —Chick Embryo, Transverse Section of the Upper Part of a Myotome, mes, mes', Mesenchyma; msth^ mesothelium; £c, ectoderm; Cw, cutis; Mu^ muscle-plate; ^p^ epithelioid layer, x 396 diams. cides with posterior limit of its segment thus determined, while the anterior edge is clearly within the territory of the next segment in front. In the region of the limbs the muscle plates send in elasmobranchs buds into the limbs to produce their muscles, as discovered by Bal; four, "Comp. Embryology," II., 673. According to Dohrn, 84.1, 163, this budding takes place after all the gill-clefts have become open, and the cartilage is just begining to appear in the branchial arches ; each myotome produces an anterior and posterior bud, which both point outward and downward ; the buds have at first a spherical form, but soon separate from the parent muscle plate, and elongate, and later divide each into two, a dorsal and ventral secondary bud, so that from each myotome there are produced four buds. The main muscle plate continues its growth into the abdominal somatopleure. The number of myotomes which contribute to the limbs is uncertain, but there are several. It is probable that in all amniota the myo- tomes also send buds to form the muscles of the limbs. Van Bem- melen, 89.1, 343, has shown that in snake embryos with the fifth gill-cleft just formed, the myotome of the second to tenth post-occip- ital segments send downgrowths into the limbs, and continue on in THE MESOTHELIAL MUSCLES. 477 the somatopleure ventralward. Of the eight segments the first three have their outgrowths oblique to enter the limbs. Paterson, 87.1, has expressly denied this origin for the chick, but as he was able to distinguish only a confused mass of mesoderm in the young limbs, his opinion cannot carry weight, but must be considered based upon imperfect observation. Abortion. — A certain number of muscle plates disappear during early embryonic life. Thus Froriep has shown, 86.1, that in the cow embryo there are four rudimentary nauscle plates in the region of the occiput or hypoglossus, which, however, all disappear very early. It is probable that there were once other muscle plates in the head which have now disappeared, compare p. 200. Further, it is probable that in man there are rudimentary muscle plates in the em- bryonic tail, which has been shown by Fol to contain at least nine rudimentary segments, some of which may advance into the muscle- plate stage. Myotomic Muscles.— There is no part of embryology so obscure at present as the development of the muscular system. Scarcely the most elementary questions have been answered. Not only has the development of the single muscles from the mesothelial plates scarcely been studied, but also the very significance and the arrangement of these plates in the head is wrapped in uncertainty, see p. 200. The following points in regard to the cephalic myotomes have been ascertained. Of Van Wijhe's nine myotomes, seen in elasmobranchs, the first comes to lie against the optic vesicle and gives rise to the rectus superior, rectus inferior, and obliquus inferior of the eye; the second produces the obliquus superior, and the third the rectus ex- ternus; a good figure of the three myotomes which form the eye muscles, as observed in an elasmobranch embryo, is given by A. Froriep, Anat. Anzeiger, N. 56, see also Miss Piatt's figures 91.2; the fourth, fifth, and sixth disappear; the eighth, ninth, and tenth produce muscles running from the skull to the sh6ulder girdle. Froriep, 86.1, has shown that there are four myotomes in the occipital hypoglossal region of mammals, which early become rudimentary, but Van Bemmelen has observed, 89.1, 241, that these four myotomes together with that of the first cervical (atlas) segment grow obliquely ventralward, so as to meet and unite into a single cord which de- scends behind the last (in reptiles the fifth) gill-cleft, accornpanied by the hypoglossal nerves, and then curving forward grows into the tongue and there produces the lingual musculature. This explana- tion of -the origin of the muscles of the tongue is probably correct, but it differs from that offered by Froriep, 85. 1, and still more from that of His ("Anat. menschl. Embryonen," III., 92). According to Van Wijhe, 82. 1, the coracohyoid muscle of sharks arises, like the mammalian lingual muscles, from the downgrowth of the posterior cephalic and anterior cervical myotomes. As regards the development of the muscles of the rump and limbs we possess, so far as I am aware, practically no information beyond the little which has been noticed in connection with the history of the muscle plates, p. 200. Muscles of the Branch.ial Arches. — That these muscles all arise from the mesothelium of the arches is now generally believed 478 THE FCETUS. although by no means rigidly demonstrated, except for elasmobranchs ; Van Wijhe, 82.1, states that the corocobranchialis and coraco- mandibularis muscles of sharks are developed from the pericardial mesothelium. Anton Dohrn, 84. 1, 109-114, finds in selachians that the mesothelial tube lengthens with the whole arch and by expand- ing in the transverse plane becomes a plate. Fig. 258, msth, which stretches across the arch between the nerve in front and the anlage of the cartilage behind ; the coelomatic cavity is obliterated except on the outer side of the arch ; the plate then divides close to the nerve. The further history is complicated and niGth /'^i^^?^X need not be presented here, as noth- \;f^^Misi&^^r,°-^'^ ing definite is known as to the ho- /^^?^^^^H?'^^i^tt-c'vll mologies of the branchial muscles of A - J^"'W ^'V.'-.;!j K- v8j'^J-^-'ll sharks with those of amniota. f '»'°=*^.V~Uv&-i?^t^-'-!"l?BB°'^^^ -^^^ (" Anat. menschlicher Embry- ..-.-i^Aiii^lj^^^o^Xv-fj.^^^^ onen," Heft III., 93) has endeavored '^iifcisiili^^'^^^ to indicate to which arches certain V c muscles belong, but has not worked FiG.258.— Transverse Section of a Branch- OUt the actual development. He as- ial Arch of a Selachian Embryo /.Branch- „• , ;i„ „„i„4.„™i„„„ „ „i i i ialfllament:mstt. mesothelium; .4, artery; Signs the palatoglossus, StylogloSSUS ^•,.fn^^i°''v^i»; ^„- <=°°°<«="°s ^«'°= '^"'■*' and levator palati mollis to the second cartilage, i\, nerve. , - \ ., i-i i arch (counting the mandibular as the first) ; the stylopharyngeus and perhaps both the palato-pharyn- geus and hypoglossus to the third arch. Of the constrictors of the pharynx the upper probably belongs to the third arch, but the mid- dle and lower to the fourth arch. C. Rabl, 87. 1, maintains that the myothelium of the hyoid arch forms the embryonic platysma, which spreads out in front of and behind the external ear (hyoid cleft) and breaks up into the individ- ual superficial muscles of the face and epicranium. The stapedius muscle also belongs to the hyoid, according to Rabl. Mandibular Muscles. — Their development in the chick has been studied bj' Kaczander, 85. 1. The muscles form at first a continuous mass, which grows Ijy multiplication of the fibres. The mass is divided into separate muscles by the ingrowth of fibrillar connective- tissue partitions, and by the development of the osseous mandible, which separates the muscles attached to the connective tissue from those having an insertion on the Meckel's cartilage. The change in the direction of the course of the fibres ■ results from the muscles adapting themselves to changes in the form of the jaw. The inser- tion into the mesenchymal anlage of the mandible remains unaltered when the anlage ossifies. There is no migration of the insertions. Dohrn, 84.1, 113, states that in sharks the developmental history of the mandibular muscles is quite different from that of the muscles of the succeeding arches. Muscles of tlie Heart.— The exact history of the genesis of the cardiac muscle fibre has still to be worked out. In the rabbit (KoUiker, "Grundriss," 2te Aufl., 383) the musculature of the heart appears the ninth day, and by the tenth or eleventh day is developed over the entire organ, including the bulbus aortae. The muscles soon arrange themselves into a spongy structure, each web of the sponge- work being covered by endothelium. Fig, 290, but during the third THE MESOTHELIAL MUSCLES. 479 and fourth month the musculature gradually becomes more compact, so that at the beginning of the fifth month the spongy structure is confined to the innermost layers of the muscular wall. The stria- tions appear, according to A. C. Bernays, 76.1. 487, upon one side of the branching, protoplasmatic muscle cells (embryo calf of 13-16 mm.) and later around the periphery of the cells somewhat as in the myotomic muscle cell. CHAPTER XXII. THE SPLANCHNOCCELE AND SEPTUM TRANSVERSUM. RENAL CAPSULES. THE SUPRA- The history of the splanchnocoele of the head has already been given as fully as our present knowledge permits, see p. 301, and ex- cept of that part which forms the pericardium. In this chapter the subdivision of the main ventral coelom or splanchnocoele into the pericardial, pleural, and abdominal cavities, and in connection there- with the development of the diaphragm, is described. It is to be remembered that the coelomatic cavities of the gill arches are possi- bly part of the splanchnocoele. Development of the Septum Transversum. — The term sep- tum transversum was first introduced by Wilhelm His in his description of his embryo M (" Anatomie menschlicher Embryonen," Heft I.). Our present knowledge of its formation and metamor- phoses rests chiefly upon the investigations of His, I.e., Heft III., also 81. 1, and of Ed. Ravn, 89.2. The septum transversum is the primary partition across the body, the heart being on the cranial side, the abdomen on the caudal side of the partition ; while above it the coelom forms a passage on each side of the median plane; these passages become the pleural cavi- ties. The septum is quite thick; it includes the anlage of liver ; by it all the veins make their entrance into the heart ; later on the anlages of the supra-renal capsules also ap- pear in it; it is itself the anlage of both the diaphragm and of the membrane separating the pericar- dial and pleural cavities of the adult. The character and general relations of the septum can be un- derstood from Fig. 259. The sep- tum divides the pericardial cav- ity, p. c, from the main abdominal cavity, Ab. c; the heart is sup- posed to be in great part removed. The septum appears to be much enlarged by the growth of the liver, which at the stage represented has become, as it were, an appendage, Li, of the septum proper, which may be conveniently defined as the layer of connective tissue next the pericardial sac. Above the sep- tum is the small passage, PI, into which the lung, Pul, is beginning to project, and which becomes the pleural cavity of the adult; at this DnLv. Fio. 259. —His' Embryo R, 5 mm. Eecon- struction to show tlie Septum Transversum. p. c, Pericardial cavity; Md. mandible: Ao^ aorta ; Au, auricular end of heart ; v.j, vena jugularis; D, C, ductus Cuvieri; Pul, lung; PI, pleural cavity; lA, liver; Ah.c, abdominal cavity ;_ In, intestine; C/m.f, umbilical vein; vi.v, vitelline vein. SPLANCHNOCCELE AND SEPTUM TRANSVERSUM. 481 stage it is termed the recessus parietalis dorsalis by His. It will be recalled, that the ccelom forms originally two splanchnoccelic cavities; in the region of the heart the partition disappears, leaving a single pericardial cavity; in the abdomen the partition (ventral mes- entery) below the intestine disappears, so that the two cavities are brought into communication, while on the dorsal side of the septum the partition remains ; but the two pleural splanchnocceles always are distinct and never communicate directly with one another. The arrangement of the veins is important. The jugular, v.j, coming from the head, and the cardinal, coming from the rump (Wolflfian body) , unite on the dorsal side of the cephalic end of the pleural cavity into a single stem, the ductus Cuvieri (future vena cava superior), which passes in the somatopleure around the outside of the pleural cavity to join the other veins in the dorsal part of the septum. The ductus Cuvieri, D. C, is just at the boundary of the pericardial and pleural cavities, and its growth is the essential factor in shutting the opening. The umbilical vein. Urn. v, joins the ductus just as it enters the septum. The vitelline or omphalo-mesaraic veins enter the septum nearer the median line; the four veins which are thus united form the large sinus reuniens (see Chapter XXIV.) from which the blood is poured into the heart. The origin of the septum in mammals * has been studied as yet only in the rabbit by Uskow, 83. 1, His, and Eavn. The following description is based on Ravn's 89.2, 124-139. The head, H, of the em- bryo early grows forward, Fig. 2G0, so as to intrude upon the region of the proamnion. Pro. A, and hence, as will be evident by an inspection of the figure, the head is bounded in front and at the sides by the proam- nion, and therefore the ccelom of the head cannot communicate with the extra - embryonic ccelom directly. Around the edge of the proamnion runs the omphalo-mesaraic vein, om.v, the continuation of which is the anlage of the heart, Ht; on the double origin of the rabbit's heart, see p. 227. The vein projects con- siderably above the level of the splanchnopleure in which it runs, and this projection gradually increases until the wall of the vein reaches to and unites with the overlying soma- topleure, and thus divides the ccelom into two parts, the recessus pari- etalis dorsalis, r.p.cL, and the recessus parietalis ventralis, r.p.v. This division is confined to the region of communication between the pericardial cavity, P, and the remaining ccelom, as indicated by the position of the reference letters r.p.d, and r.p.v. A cross-section Fig. 260.. Head ot & Rabbit Embryo, with Segments, seen from tlie under Side. Pro.A^ Outline of proamnion; H, head; om.v, omphalo-mesaraic vein; Ht^ anlage of heart ; a, margin of the opening of the Vorderdarm; s, primitive segments; Md, medullary canal ; r.p.v^ recessus parieta- lis ventralis; r.p.d, recessus parietalis dorsalis; P, pericardial cavity, x 25 diams. After Ed. Eavn. * On the development of the diaphTagm in the chick see Loclcwood. 88. 1 ; in lizards see Ravn, 1.3. 31 483 THE FCETUS. Pro.a rpd •" rp.v. Fig. 261.— Rabbit Embryo, Eight and a Half Days, with Eleven or Twelve Som- ites: Cross-Section. Pro.a^ Pro-amnion; V. car^ cardinal vein; r.p.d^ recessus pa- rietalis dorsalis (pleural cavity) ; Ht^ anlage of the heart; r. p.v, recessus pari- etalis ventralis; Pft, pharynx, diams. After Ravn. of a little older stage with the vorderdarm just closing is represented in Fig. 361, and will help to elucidate the disposition of the parts. The ventral recessus early becomes closed at its hinder extremity and is thereby converted into a third pocket of the pericardial coelom, which His has described under the name of the bursa parietalis. The bursas sub- sequently become merged with the pericardial cavity. The dorsal reces- sus, Fig. 261, r.p.d., is the anlage of the pleural cavity and persists for some time open at both ends. The par- tition dividing the two recessi from one another, and containing the om- phalo-mesaraic veins, is the anlage of X 40 the lateral portions of septum trans- versum (Cadiat's cloison mesoder- mique, Kolliker's mesocardium laterale, Uskow's Verwachsungs- briicke) . By the further growth of the embryo the head lengthens and with it the median heart formed by the union of the two heart anlages. The splanchnopleuric wall at /o, Fig. 146, bounds not only the open- ing of the vorderdarm into the yolk-sac, but also the posterior wall of the pericardial cavity, and is the anlage of the median portion of the septum transversum. As the liver is developed at the hind end of the vorderdarm it has to grow out into this wall, /o, and conse- quently contributes to the thickening of the septum transversum. The septum is further expanded by the development of the remain- ing veins, (jugulars, cardinals, and umbilicals), which are all ulti- matel}'' united with the omphalo-mesaraics to constitute the great sinus reuniens. In brief : the septum transversum includes the median part of the splanchnopleuric wall separating the pericardial cavity from the neck of the yolk-sac, and the two lateral parts resulting from the two up- growths of the splanchnopleure to carry the omphalo-mesaraic or vitelline veins to the heart. It is, therefore, entirely a product of the splanchnopleure. Separation of the Pleural and Pericardial Cavities. — The septum transversum separates the two cavities as soon as it is formed, and in the adult the primitive arrangement is easily traced in part, despite the great expansion of the pulmonary coelom. The septmn leaves, however, a direct communication open as shown in Fig. 259, where the ducts of Cuvier, D.C, descend from the dorsal to the ventral side. The figure further shows that the septum is oblique, so that the pericardial cavity in part underlies the pleural cavity. As development progresses, the three cavities all expand, and more and more of the pericardial cavity comes to lie on the ventral side of the pleural cavities, leaving a part of the septum transversum as a partition, which, of course, runs as far headward as the duc- tus Cuvieri, D. C. This partition early becomes thin, and is the membrana pleuro-pericardiaca which was partly described by F. SPLANCHNOCCELE AND SEPTUM TRANSVERSUM. 483 T. Schmidt, 70.1, and Uskow, 83. 1, and more fully by His, 81.1, 313, and Ravn, 89.2, 136. The anterior passage is closed by the growth of the ductus Cuvieri, which, like all the chief veins of the embryo, has an enormous size ; it causes, therefore, a projection which ultimately shuts the passage to the pericardium completely. Exactly at what time the shutting off occurs is not stated, but probably dur- ing the fifth week in the human embryo, and in the rabbit by the fifteenth day. The separation of the pleural from the abdominal cavity takes place much later. Expansion of the Pleural Cavities. — Concerning the gradual enlargement of the pleural cavities very little is known beyond the fact that they enlarge at the same rate as the lungs. In the rabbit at fifteen days they are together about half as large as the pericardial cavity. As stated above, the primitive pleural cavity is on the dorsal side of the septum, and the cephalic limit of the septum is marked by the ductus Cuvieri, or future vena cava superior. Part of the septum is used to develop the pleuro-pericardial membrane, while the re- mainder, which includes the hepatic attachment, develops into the diaphragm; beyond the caudal boundary of the septum the lungs never project. These considerations show that the pleural cavities lie entirely within the territory of the septum, and that their expan- sion takes place within the septum. This conception renders it necessary to regard the thorax of the adult as chiefly occupied by the distended septum transversum, and involves important changes in our morphological notions concerning the adult condition. Mesocardium, Mediastinum, and Mesentery. — These mem- branes are the remains of the tissue which originally divides the ccelom of one side from that of the other. The tissue disappears for the most part around the heart, so that the pericardial cavity is con- tinuous on both the dorsal and ventral sides of the heart. In the abdomen this continuity is established only on the ventral, not on the dorsal, side of the intestinal canal, and the tissue between the two lateral coeloms remains to form a very thin membrane, the mes- entery, by which the intestine is attached to the median dorsal wall of the abdomen. Between the two pleural cavities the tissue remains, forming a thick partition, the mediastinum. Concerning the gene- sis of these membranes little is known. Sac of the Omentum and Foramen of Winslow. *— In the chick soon after the lungs have grown out from the oesophagus, and just when the first forking has begun, the abdominal ccelom is found to form two blind diverticula lined by well-marked mesothelium and extending until they come into direct contact with the pulmonary entoderm. Of these -diverticula I have found no mention. Similar ones have been observed in the rabbit by Ravn, 89. S, 139; their formation is connected with the prolongation of the ridge of meso- derm on the side of the oesophagus. The ridge on the left side, and with it the diverticulum, disappears very early, but that on the right side persists and enlarges, the vena cava inferior being developed within it, on which account Ravn terms it the "vena cava Falte." This fold extends down into the abdomen ; the coelomatic diverticu- * Compare also chapter xxix. 484 THE PCETUS. V.om Som lum between it and the intestinal canal, Fig. 363, is the "re- cessus superior sacci omenti" of W. His (" Anat. menschl. Em- bryonen," Heft I., p. 65). "While this growth of the " vena cava Falte" is taking place the stomach has been developing its great bend to the left as indicated by the dotted lines in Fig. 362, carrying with it, of course, the mesogastrium and mesentery of the duodenum, and thus forming a sac, the entrance to which is partially closed by the vena cava Falte. The sac is the saccus omenti, the entrance to it is the foramen of Winslow, F. W. The saccus is bounded on the dorsal side and on the left by the meso- gastrium; on the ventral side by the stomach, the position of which at a level nearer the observer may be easily imagined from the figure ; and on the right by the vena cava Falte, v.c.i. In the rabbit, according to Ravn, 89.2, 146-147, the anterior end of the recessus. Fig. 363, becomes sep- arated, as a closed sac, about the seventeenth day, and forms a cavity between the oesophagus and the so- called lobus inferior medialis of the right lung, and persists in the adult. A similar cavity (Schleimscheide) is found also in rats and mice, and is presumably developed in the same way. Ravn thinks it probable that an homologous cavity is present in the human em- bryo, but aborts. Separation of the Pleural and Abdominal Cavities. — This takes place much later than the separation of the pleural and pericardial cavities, for it is not effected in the rabbit until the seven- teenth day, and KoUiker records that it had not taken place in a two- months' human embryo. This agrees with the fact that the separa- tion takes place only in the mammals, not in other vertebrates. Ravn, 89.2, 147, is the only investigator who has attempted to fol- low out the process accurately. A fold is formed. Fig. 363, which lies obliquely between the lungs and the Wolffian body on each side, and which in the rabbit at fifteen days is found to somewhat contract the opening between the pleural and abdominal cavities ; the fold extends almost if not completely around the opening, making as it were a circular shelf. Another factor, as pointed out by His, is the expansion of the liver. I have observed also, in studying Pro- fessor His' embryo Zw,* that the anlage of the supra-renal capsule had appeared in the septum transversum on the ventral side of, and close to, the peritoneal opening of the pleural cavity, so that the con- FiG. 263. —Model of Part o( the Pleural and Abdominal Cavities of a Rat Embryo at a Stage Corresponding to a Rabbit at fifteen •Days, oe, CEsophagus ; i, lung ; D, anlage of fold to form the diaphragm ; Om, omen- tum; Som^ somatopleure; V.om^ omphalo- mesaralc vein ; F, >F, foramen of Winslow ; V, c. i, vein in the plica venae cavse of Bavn. After Ravn. * My grateful acknowledgments are due to Professor His for the very generous manner in which he placed his material at my disposal, during a few weeks I had the pleasure to spend in Leipzig in 1887. *^ SUPRA-RENAL, CAPSULES. 485 elusion was inevitable that the final factor in completing the closure of the opening was the growth of the supra-renal capsule. Diaphragm. — The diaphragm {Ztverchfell) is developed from that portion of the septum transversum which intervenes between the pericardial and abdominal cavities, and from the fold which shuts off the connection of the pleural cavities with the abdominal. The veins pass through the diaphragm to the heart, and to the area around the veins the liver is permanentlj- attached ; it is out of the remainder of the diaphragm that the muscular part and the centrum tendineum are developed, but concerning their development no obser- vations whatever are known to me. Lining Membranes of the Splanchnoccele. — These mem- branes are the pericardial, pleural, and peritoneal. They each con- sist of a layer of specialized connective tissue and the mesothelium, which is found in the adult to have lost its primitive character of a cuboidal epithelium and to have become a thin layer or endothelium. Concerning the manner in which the transformation is effected, there are few reliable observations — compare Chapter XXIX. The Supra-renal Capsules. It is only with considerable hesitation that I have decided to treat the supra-renal capsules as organs developed in the septum transver- sum on the ventral side of the pleuro-peritoneal opening. I have made observations which lead me to think this view necessary from the facts of development, but I have not been able hitherto to continue the re- search to a satisfactory conclusion. As the kidney grows forward until it reaches the dorsal pillars of the diaphragm, the supra-renals would come in juxtaposition with the upper end of the kidneys, whether the capsules began their development on the dorsal side of the pleuro-peri- toneal opening or on the ventral side, for in the latter case the closure of the opening would bring the capsules near the kidneys. At present I am inclined to the belief that the mesenchymal portion of the supra- renals arises on the ventral side of the opening and the sympathetic portion on the dorsal side. That this view is right is confirmed by the observation that the capsule lies entirely on the ventral side of the kidney in the human embryo at two months and at three. Mesenchymal Anlage. — The mesothelium in the region of the vena cava inferior and septum transversum throws off cells to con- tribute to the mesenchyma. Janosik, 83.1, who observed this pro- cess at the point where the supra-renals develop in mammals, concluded that it was a special process and that the supra-renal capsules must, therefore, be considered as derivatives of the perito- neum . The recognition since then of the genetic relation of the whole mesenchyma to mesothelium renders it unnecessary to assume a special relation for a single mesenchymal organ. The same criticism appHes also to Weldon, 85.1, who, having observed the production of mesenchyma from the mesothelium of the nephrotomes, or seg- mental vesicles, in lizards and sharks, concludes that there is a special genetic relation between the supra-renals and the segmental organs. In reptiles, soon after the vena cava is formed, there appears on each side of that vein a small cluster of crowded mesenchymal 486 THE PCETUS. cells (Max Braun, 83. 1), which increases quite rapidly ; the cells of the cluster gradually arrange themselves in cords which become more and more twisted and united ; numerous blood-vessels are early developed between the cells, probably by ingrowth from the adjacent Wolffian bodies. The nearness of the first trace of the supra-renals to the vena cava has also been noted by Gottschau, 83, 1, by Mitsu- kuri, 82. 1, andWeldon, 85. 1. In the rabbit the first distinct trace of crowding of the cells and of their enlargement to form the anlage of the rupra-renals may be seen on the twelfth day ; on the four- teenth day the anlages are well marked (Mitsukuri, 82. 1) ; by the sixteenth day the sympathetic anlage is surrounded by the mesen- chymal. In the sheep (Gottschau, 83.1, 449) the anlage can be rec- ognized in embryos 9 mm. long ; it is in contact with the sympathetic ganglion tissue in embryos of 11 mm. and in those of 13 mm. has become quite sharply defined against the surrounding mesenchyma. In the pig the first trace is seen in 9 mm. embryos according to Gott- schau, 83.1, 452. Balfour, 81.3, homologies the mesenchymal anlage with the so- called inter-renal bodies of elasmobranchs. Sympathetic Anlage. — On the dorsal side and somewhat toward the median line of each mesenchymal anlage appear a cluster of small cells, which are stained brown by bichromate of potassium, as first observed by M. Braun, 82.1, 25, in reptile embryos. These cells are derived from the chain of sympathetic ganglia, and are characterized by being smaller and more granular and by having smaller nuclei than the cells of the mesenchymal anlage. I have noticed that in specimens colored with alum-cochineal they stand out conspicuously owing to their deeper staining. In rabbits of fourteen days, the sympathetic anlage has become very distinct and has increased in size, and in those of sixteen days it is found surround- ing the mesenchymal supra- renal and more or less separated from the ganglion proper. At this time traces of young ganglion cells and of nerve-fibres are said to be clearly recognized. F. M. Balfour, 81.3, and in his " Comparative Embryology, " II. , 664, homologizes the sym- pathetic anlage with the so-called "true supra-renals" of elasmo- branchs, bodies which develop from the sympathetic ganglia. Ac- cording to Balfour (monograph of Elasmobranchs, " Works," I., 472), who greatly extended Semper 's observations, 75.2, in shark embryos in Balfour's stage L the ganglia of the sympathetic chain are par- tially divided into two parts : one the future ganglion proper, the other the anlage of the supra-renal, which receives a direct artery from the aorta. By stage O these supra-renal anlages have acquired a distinct mesenchymal investment, which penetrates into their interior and divides it, especially in the case of the anterior anlages, into a number of distinct alveoli. By stage Q, the cells are differen- tiated into larger (ganglionic?) cells and smaller ones, which Balfour holds to form the true supra-renal tissue. The observations thus far made indicate that the sympathetic anlage is derived from a series of spinal ganglia, which give off a series of supra-renal parts ; these parts remain distinct in elasmo- branchs, but fuse into one mass on each side in amniota. Rudimen- tary ganglion cells arise, but soon abort. SUPRA-RENAL CAPSULES. 487 sym. —Section of the Supra-renal Body Eabbit Embryo of twenty-six Days. Tnes^ Mesencbymal portion; sym^ sympathetic portion; v, blood-vessels: s, mesenchymal Union and Ultimate Fate of the Two Anlages.— The mesenchymal and sympathetic portions very soon come into contact (sheep of 1 1 mm. , rabbits of the fourteenth day) . At first, in amniota at least, the sympathetic anlage grows most rapidly and partially surrounds the mesenchymal portion, but soon the relations are re- versed and gradually the mesenchymal portion completely invests the synipathetic part, but for some time there remains a hilus on the inner side toward the posterior end of the organ. Fig. 363 shows a transverse section of the left supra- renal taken about the middle of the body from a rabbit embryo of twenty-six days. The cortex is already made up of distinct cell- groups, which, however, are not yet differentiated into the adult cortical and medullary zones. Capillaries are well formed be- tween the adjacent cell-groups. The sympathetic portion, sym, is divided into irregular groups of cells, which stain readily and are situated exclusively in the cen- tral region ; between these groups are relatively large blood-vessels, v. The connective tissue has formed a sheath, s, around the organ. Mitsukuri states that the masses of nervous origin are now full of " distinct ganglion cells, supported in a connective-tissue framework ; scattered among the larger cells are smaller cells." This may be regarded as perhaps the Sauropsidan condition, since according to Hans Rabl, 91.1, the two supra-renal tissues persist in birds throughout life, interlaced with one an- other. Mitsukuri believed that the medulla of the adult capsules arises from the sympathetic anlage, but Gottschau, 83. 1, showed that this was not the case, though he failed to ascertain what became of the sympathetic masses. By a considerable series of observations on the supra-renal capsules of human embryos, I have ascertained that there are groups of cells which gradually disappear and take no part in the production of the adult organ. The cells are in clusters in the central portion of the organ and stain very readily, so that they stand out conspicuously in the sections. In appearance they resemble the cells assigned to a sympathetic origin in the rabbit, and I should feel no doubt that they are the same were it not that I fail to find them in embryos of the second month, so that if they are really of sympathetic origin then the union of the two anlages must take place at a considerably later stage in man than in other animals. These groups of cells are readily seen in the three-months' embryo, but in the four-months' embryo they are disappearing and many of the clus- ters are hollow, their cavities being filled with what is apparently a coagulum ; by the seventh month the clusters have, so far as I have hitherto observed, entirely disappeared. That both the cortex and sheath cavity. msth^ mesothelium lining the body- After Mitsulturi. 488 THE FCETITS. medulla of the adult organ are formed in man from the mesenchymal cells, as Gottschau, 83.1, showed was the casein several mammals, is I think, beyond question. The cords of cells are at first uniform throughout, but I fmd that toward the end of the second month the cells of the cords multiply and become smaller, while at the same time the cords assume a more radial position and regular arrange- ment around the periphery; there is thus developed a cortex, char- acterized by radiating, smaU-celled cords and a medulla, characterized by irregular, large-celled cords. In the cortex the cords are wide and contain numerous cells; toward the interior the cords break up into small ones, which pursue the same radial course and consist of cells which gradually increase in size toward the centre of the organ. The cords are marked off by wide capillaries, with distinct endothe- lial walls, between which and the supra-renal cords there appears to be no connective tissue, although in the medulla there is more or Fig. 264. —Supra-renal Capsule of a four-months' Human Embryo. Minot Collection, No. 35. Cross section of the medulla, x about 500 diams. less connective tissue developed early around the vessels, Fig. 264. It seems to me that the cortex grows at the expense of the medulla, the deep-lying large cells dividing into smaller ones. The medulla of a four-months' embryo is represented in Fig. 264. The cords of supra-renal cells are very irregular and often con- nected together, but are readily seen to be directly continuous with the cortical cords. The medullary cords are much more widely separated than those of the cortex from one another, the spaces be- tween them being filled with connective tissue and blood-vessels, none of which have any adventitial or muscular walls. The great variety of appearances presented by the cells of the cords is indi- SUPRA-RENAL CAPSULES. 489 ■cated in the figure; large and smal], regularly and irregularly shaped, uninucleate and multinucleate, light-stained and darkly- stained cells lie jumbled together without obvious law of arrangement. The significance of this strange picture is unknown. It should be noted that the nuclei of the cord-cells are all, or nearly all, decidedly larger than those of the adjacent connective tissue. As develop- ment proceeds the cells become gradually more uniform in appear- ance, and offer by the seventh m.onth little variety ; also the conti- nuity of the cords is lessened in the medulla and the blood-vessels become larger. It is evident that there is no fundamental difference between cortex and medulla — in the former the cords have a radial trend, in the latter they run irregularly ; the medulla is also char- acterized by having larger supra-renal cells and a richer blood supply. Form and Size of the Supra-renals. — The supra-renal capsules have at first a rounded form and lie on the ventral side of the cephalic end of the kidney. Probably about the third month they begin to spread on to the dorsal side of the kidney, the head end of which they invest like a cap. The capsules grow at first very rapidly, afterward more slowly, and as the kidneys grow more steadily the relative size of the capsules compared with the kidneys passes through striking changes. CHAPTER XXIII. THE UROGENITAL SYSTEM. The early history of the urogenital system has already been given, Chapter XL, p. 230. We have now to consider the differentiation of the male and female type from the indifferent condition. In or- der to render the complex changes clear, it has seemed to me advisa- ble to give first a general history of the metamorphoses, so as to bring out first the homologies in the two sexes, next to present the special histories of the single parts, and finally to append an account of the external genitalia. I. General History, The Indiflferent Stage. I. — The early history of the urogenital system has been given in Chapter XI. ; nevertheless it will be con- venient to present here a generalized diagram of the indifferent stage, for comparison with diagrams of the differentiated system male and female. The indifferent stage is characterized by all the organs being contained in two longitudinal urogenital ridges Fig. 365, one on each side of the body and projecting from the dorsal wall into the peritoneal cavity. At the caudal end of the abdomen the two ridges draw closer together and finally come into contact with the anal region of the intestinal canal. The ridge is constituted chiefly by the Wolffian body, w.b, and it therefore contains the Wolffian tubules and the Wolffian duct, W.D, which is situated on the side of the ridge farthest from the mesentery, mes. Close alongside the Wolffian duct lies the Miillerian duct, M. D. Both ducts open into the cloaca, CI, or ter- minal division of the intestine. Changes in Both Sexes.— The essential or fundamental difference between the two sexes is the change of the genital ridge into an ovary or testis according to the sex. The secondary differences are chiefly in the modifications of the ducts, and as regards these the most important changes are that in the male the Wolffian duct becomes the gen- ital duct (duct of the epididymis, vas deferens, and ductus ejacula- torius) , while in the female the Miillerian duct becomes the genital duct (Fallopian tube or oviduct, uterus, and vagina) . Before con- sidering the changes more in detail it will be convenient to divide them into two groups; 1, common to both sexes; 2, characteristic of one sex. Fig. 265. —Diagram of the Indifferent Stage of the Urogen- ital System of Am- niota. Explanation in text. GENERAL HISTORY. 491 k.Mcble B.FemaZe 1. Changes Common to Both Sexes.— There are three im- portant changes from the indifferent stage to be noted under this head : A, the union of the caudal ends of the urogenital ridges to form a single median genital cord; B, the anterior end of the Wolffian body persists and undergoes modification in connection with the genital glands, by which two separate organs are produced in each sex; C, in the course of development the genital organs become restricted to the lower (or caudal) end of the abdomen, and do not continue to stretch the whole length of the abdomen as at first. Another important series of changes is that by which the cephalic portion of the urogenital ridge acquires in the female a transverse position, in consequence of which the upper or cephalic end of the Miillerian duct, or in other words the future Fallopian tube, runs transversely. This change occurs in the male also, but is less noticeable and is, to a certain extent, masked by the migration of the testis from the abdomen through the inguinal ring into the scrotum. 2. Changes Characteristic of One Sex Only. — A. Male. — The general plan of the urogenital ridge in the male is indicated in the diagram Fig. 366. In the male, as stated above, the sexual gland becomes a testis by the development of seminiferous tubules, and the Wolffian duct becomes the genital duct. The connection between the Wolffian duct and the seminiferous tubules is established by means of the anterior tubules of the Wolffian body. There are special extensions of these tubules into the testis, which unite with the semi- niferous tubules and form a series of anastomoses with one another within the testis (compare Fig. 266), constituting the rete testis, while the tubules proper of the anterior part of the Wolffian body remain to serve as the channels of connection (vasa efferentia) be- tween the rete testis and the Wolf- fian duct, which is thus enabled to serve as the spermiduct. A portion of the anterior Wolffian tubules persist as a separate group, which is known as the organ of Giraldes, or paradidymis of Wal- deyer. The spermiduct becomes differentiated into three principal divisions: 1, the coiled portion ^ „„ „. , t,, ^ . *>, ti , ,, , ,. ,., /• ,1 Fig. 266. —Diagram to Illustrate the Homolo- nearest the testis COnStltutmg the gles of the Sexual Apparatus, fl-j/d. Hydatid; j,.„f „f tVio orvirlirlTrmi«- '1 tViplnno' ""■ ''/' '^^^ efferentia; Epd, duct of epididy- auct or tne epiaiaymis , ,,T;neiong ^^^.^ jr.AWoifflan duct; M.D,Miiiier's duct; VaS deferens running through the G.c, genital cord; m.m utems mascullnus; ., , » IT , °i .IX Ic, testis; iJefe, rete Halleri; Porad, paradi- UrOgenital told, to wnere tne two dymis; /, fimbria; parv, parovarium or epo- folds unite to constitute the genital ?P^?™ro.fiv"a'?y™'p^art5/.^Ti™ephoron.*'^' cord; 3, the ductus ejaculatorius, developed below the point where the seminal vesicles are formed and within the genital cord. The Miillerian ducts remain rudimentary and their middle portions usually abort, leaving the upper fimbriate -ParoopTi 492 THE FCETUS. ends to develop into the so-called hydatids of Morgagni, and the lower or caudal ends to unite within the genital cord to form the so-called uterus masculinus (prostatic vesicle), a rudimentary repre- sentative of the female uterus and vagina. B. Female.— The general plan of the urogenital ridge in the fe- male is indicated by the diagram, Fig. 2(36. In the female, as stated above, the sexual gland becomes an ovary, by the development of ovic follicles, and the Miillerian duct becomes the genital duct. The Wolffian duct remains rudimentary and in part disappears ; it persists in the genital cord as the duct of Gartner, but does not, so far as known, unite with its fellow ; it persists also at its upper or cephalic end as the duct of the parovarium (epoophoron, organ of Eosenmiiller) , which comprises the group of Wolffian tubules in the female homolo- gous with the vasa efferentia of the male. There also persists a group of Wolffian tubules, which has been named the paroophoron by Wal- deyer and is homologous with the male organ of Giraldes. The Miil- lerian ductus unite within the genital cord to a single median duct, which enlarges greatly and is differentiated into the uterus and vagina ; the upper or cephalic portions remain separate and form the Fallopian tubes or oviducts proper ; the Miillerian funnel becomes the fimbriate opening of the Fallopian tube. Homologies bet-ween the Sexes. — These may be readily fol- lowed by means of the accompanying diagrams. Fig. 366, A and B, and the table given below. The diagrams call for no further expla- nation than is given on the figures. Tabular View of the Homologies op the Human Urogenital Apparatus in the Two Sexes. Indifferent Stagse. Male. Female. Genital ridge. Wolffian tubules. Wolffian ducts. Testis. 1. Epididymis. 3. Paradidymis. 3. Vas deferens. Ovary. 1. Epoophoron. 2. Parovarium. 3. Duct of Gartner. Miillerian duct. (Vesicula seminalis. ) 4. Ductvjs ejaoulatorius. 5. Hydatid of Morgagni. 6. (Usually aborts. ) 7. Uterus masculinus. 4. (Usually aborts.) 5. Fimbriate opening. 6. Fallopian tube. 7. Uterus. 8. (Usually undeveloped. 9. Verum montanum. 8. Vagina. 9. Hymen. Urogenital sinus. 10. Urethra. 10. Urethra and vestibule. Genital eminence. 11. Cowper's glands. 13. Penis. 11. Bartholini's glands. 12. Clitoris and nymphae. External labia. 13. Scrotum. 13. Labia majora. II. Special Histories op the Urogenital Organs. Sexual Glands. — A. Male. — ^The testis becomes recognizable by its histological character in the human embryo according to W. Nagel, 89.3, 309, at five weeks; according to Benda, 89. 1, at six weeks. It can be distinguished from the ovaries by its external form in the three months' embryo. By the abortion of the Wolffian body and by the growth of the testis the latter becomes the principal organ of the urogenital fold in the male. The Wolffian part of the fold re- mains to form the mesorchium, the lower or caudal portion of the SPECIAL HISTORIES OF THE UROGENITAL ORGANS. 493 fold remaining as the gubernaculum. By the fourth month the testis has assumed its permanent form, but its growth continues. The role of the Wolffian tubules in the genesis of the testis is described below, p. 500. Histogenesis. — The subsequent account of the development of the testis follows Nagel, 89.3, closely, whose results I accept, both as regards his observations and his criticism of previous investigations, although they require modification owing to what has become known concerning the genetic relation of the mesenchyma to the mesothe- lium of the genital ridge, see p. 248. As described in Chapter XII. the genital mesothelium throws off cells, which at first assume entirely the character of loose mesenchj^ma, and later remain crowded to- gether with scarcely a trace of division from the parent epithelium; in this denser tissue appear large cells, the so-called "Ureier." Out of this anlage are developed epithelioid cords, the sexual cords, which in chide some of the ureier, and become more and more separated from one another by the development of loose mesen- chyma or embryonic connective tissue between them. Nagel finds that the male sexual gland, Fig. 267, may be recognized even in Sc Fig. 267.— Section of the Testis of a Human Embryo of sixty thiee to sixty eight Days. A'.c, Sexual cords; lir, ureier; coTin, mesenchyma. an embryo of 13 mm. by the small number of the ureier as com- pared with the ovary of corresponding age. In the testis at this stage (human embryo of 13 mm.), the sexual cords are not yet very distinct and are connected with the superficial epithelium. In an embryo of nine weeks. Fig. 207, the sexual gland * is covered by a regular cuboidal epithelium, distin ctly bounded against the under- *In accordance with Kolliker's description and figure C"Gi-undriss," Fig. 288), this gland would be an ovary, but Von Ackeren states that KoUiker has become doubtful m regard fo his Fig. 888, and I think it must be regarded as the section of a testis. 494 THE FCETUS. lying tissue, which is composed of mesenchyma with sexual cords, S.c, which are not connected with the mesothelium; the submeso- thelial layer is the anlage of the tunica albugineaj as no corre- sponding layer exists in the ovary, its presence in the male gland at this stage establishes one of the most characteristic features of the testis ; in the albuginea, connective-tissue fibrillse are just beginning to appear. The central portion of the testis is occupied by sharply defined sexual cords, which frequently anastomose with one another and contain here and there an " Urei," or sexual cell of Mihalko- vics ; the sexual cells are clearer and larger than the other cells of the cords, measuring 11/-! with nuclei of 8/^- diameter. In an embryo of 35 mm. the general structure is much the same, but the albuginea is thicker and more fibrillar, and the cords are more regular in their arrangement ; the cords are about 33//- thick and their cells show a somewhat epithelioid arrangement ; the few sexual cells they con- tain now measure 14-1 6/j'.. In an embryo of ten centimetres a new feature is found in the presence of the interstitial cells. These are large cells which lie between the sexual cords, and are probably developed by the enlargement of the connective-tissue cells between the cords : they are spindle-shaped or polyhedral, with several pro- cesses each; their protoplasm offers a peculiar mat appearance ; their nuclei are large, with one or two nucleoli and a distinct intranuclear network. The cords are the solid anlages of the seminiferous tubules. The question has been debated at great length whether they are differ- entiated from the stroma or the epithelium of the genital ridge — compare the synopsis of opinions given by Nagel — but as the epi- thelium (mesothelium) produces the mesenchyma or stroma, the question appears to me insignificant. The further history of the sexual cords (future seminiferous tubules) has been most fully stud- ied by C. Benda, 89.1, compare also Prenant, 89.1, 90.1. The cords remain solid throughout foetal life, the smaller cells having a radial position and epithelioid arrangement, but the nuclei are irreg- ularly placed, so that it is difficult to decide whether the cells are in a single row or not around the centre of the cord. The large ureier are irregularly distributed — less irregularly in man than in other animals — but they are always completely imbedded in the other cells and show a tendency to lie near the periphery of the cord in man, rodents, dogs, and cats, near the centre in ruminants (ox). As to their number, few ureier are found in the cords of man, while in rodents they are very numerous ; dogs and cats occupying an inter- mediate position as to number. The condition described is attained in man about the sixth week, in the rabbit the seventeenth day, and persists with little change not only throughout the foetal period, but until the time of puberty, when the cords change to seminiferous tubules. The conversion of the male sexual cords into the seminiferous tubules, being post-foetal, does not fall within the scope of this work. The reader is referred to the investigations of Prenant, 89. 1, Benda, 89. 1, and F. Hermann, 89.2. According to Benda the epithelioid cells give rise to the columns of Sertoli (Benda's Fusszellen) and the ureier to the spermatocytes (Benda's Samenstammzellen) . This SPECIAL HISTORIES OF THE UROGENITAL ORGANS. 495 is in accordance with Benda's hypothesis that the spermatocytes have no genetic relation with Sertoli's columns, an hypothesis which is not yet established firmly — compare Chapter III. B. Female. — The ovary becomes histologically recognizable about the same time as the testis, i.e. six weeks; it can be readily distinguished from the testis in the three months' human embryo by its external form. In consequence of the abortion of the Wolffian body and of its own growth the ovary is already the principal organ of the urogenital fold at three months. As the greater part of the fold has thinned out to constitute the broad ligament, the relations found in the adult are established at the age under consideration. Histogenesis of the Ovary. — ^ According to Nagel the ovary may be distinguished from the testis in human embryos of only 12-13 mm. by the greater abundance of the developed and developing ureier. In an embryo of 12mm. ]Sragel,89.3,305, describes the ovary as consisting of the prolifer- ated germinal epitheli- um ; the proliferated cells are of two kinds, the more numerous are smaller, and have more darkly stained nuclei; the less numerous are the young ureier with lighter stained granular nuclei. In an embryo of 20 mm. the ovary projects a little from the surface of the urogenital ridge, and is filled with the cells from the epithelium, the two kinds being present, and, as before, with numerous transitional stages be- tween them; the ureier measure 10-16/i with a nucleus of 8/t diame- ter — the smaller cells 8/j- with a nucleus of 5/j. ; in the centre, spindle- shaped connective-tissue cells are appearing. In an embryo of 30 mm. the ovary projects still more from the Wolffian body; the ureier are larger, 16/^, and the connective tissue or stroma is more devel- oped and has capillaries. Nagel has studied also embryos between 3 and 7 cm. in length, but we may pass at once to the latter. In embryos of 7 cm. the ovary is triangular in section, the apex of the triangle corresponding to the attachment to the Wolffian body or future broad ligament. The connective tissue now forms partitions, which divide the remaining cells into groups, Fig. 268, but the par- titions fade out toward the surface, which is covered by a single layer of cells, which has begun to assume the character of an epi- thelium entirely distinct from the underlying cells. In an embryo of 11 cm. the covering epithelium of the ovary has become more sharply bounded and the development of the stroma has extended Fig. 268. —Section of the Ovary of a Human Embryo of 7 cm. Mstli, Mesothelium ; Ue, ureier ; cc, proliferated small cells; Str, stroma or connective tissue. After W. Nagel. 496 THE FCETUS. quite to the surface, dividing the proliferated cells into rounded groups of small cells and ureier, which are at this stage very nu- merous, and indeed outnumber the small cells in the balls. These balls are a highly characteristic feature of the young mammalian ovary, but their arrangement and connections with one another have been as yet only very imperfectly studied ; nevertheless it seems safe to say that they are not separate masses, but, as seen under the microscope, sections of contorted and anastomosing cords. If this view is correct then there is an evident resemblance between the testis and ovary, there being in both cords derived from the ger- minal epithelium, containing ureier and separated from one another by vascular connective tissue. The ovary differs from the testis in having larger cords and a much larger absolute and proportionate number of ureier. That we have to do with sexual cords is evident in later stages, where the cords are verj'- distinct and are found still connected with the covering mesothelium (Waldeyer's KeiinepitheT) ; in their later stage, the ovarian sexual cords are known as Pfliiger's cords (Pfluger'schen Schlauche), being named after their discov- erer, and they differ considerably from their earlier stage in that they include a large number of small or follicular cells, which com- pletely surround the ureier and separate them from one another, by constituting an epithelioid laj'er or follicle around each urei. The transition from the stage of the balls, as we may call it, to the stage of Pfliiger's cords has not been clearly ascertained, because the origin of the small or follicular cells is still uncertain but I agreewith 0. Hertwig ("Lehrbuch d. Entwickelungsgesch.," 3te Aufl., 321) that they are cells of the original cords derived from the mesothelium of the ovary, although Eouget and so eminent an authority as Kol- liker ("Grundriss," 423) have maintained that the medullary cords grow around the ureier and produce the follicles ; KoUiker seems to me not to have offered sufficient evidence to render his view probable. Another view is that advocated by Foulis, 76.1, who believes that the ureier becomes entirely free and that the follicles are developed trom the stroma cells — a conception which cannot be maintained. If we assume, as we apparently must, that the follicular cells arise from the sexual cords, the question would still remain, whether they are derived from some of the original small cells or from the ureier; that the latter derivation is the actual one is to mj- mind probable, because there appears to be a stage in the development of the sexual cords of the mammalian ovary in which all the cells are converted into ureier ; but until further investigations shall have decided it, the question of the origin of the follicular cells must be considered an open one. Mihalkovics, 85. 1, 449, discusses carefully the origin of these cells, but owing to the distinction he draws between the sexual cords and the proliferation of the germinal epithelium to form the ureier, it is impossible to follow his own account : Mihalkovics also gives an admirable review, pp. 423-428, of the literature upon the development of the ovary. Gubernaculum, Processus Vaginalis, and Descent of the Testis. — The descent of the testis begins very early, the change in position being evident by the tenth week, but the passage into the scrotum does not begin until the seventh month. The testis SPECIAL HISTORIES OF THE UROGENITAL ORGANS. 497 makes three movements: 1, backward to near the inguinal ring; 2, forward a short distance, during the period of the formation of the rouscular gubernaculum ; 3, downward into the processus vaginalis. The processus does not extend completely into the scrotum during foetal life, hence the foetal scrotum has no cavity and never contains the testis, but on the contrary is filled by a very vascular connective tissue like the labia of the female. At birth the processus lies partially in the scrotum. The cause of the descent of the testis has been much discussed and many fanciful explanations have been propounded. There is no reason for supposing that these movements are in anjj- wise different from the numerous other movements of organs and changes of form occurring during the course of development. These changes are all due to inequalities of growth in the tissues, but the causes of these inequalities are not yet ascertained. A long-prevalent tendency has tainted the study of the generative organs with mysticism, and it must be attributed to this tendency that so many far-fetched expla- nations of the descensus testiculorum have been published. The changes in the gubernaculum are probably the immediate causes of a part of the changes in the position of the testis ; the growth of the gubernaculum accounts for the forward movement, and its atrophy for the passage along the wall of the processus vaginalis ; it must be added here that the testis also accompanies the downgrowth of the processus, and is not dragged down merely by the shortening of the gubernaculum. Some writers have supposed that the muscles of the gubernaculum effect the descent by their contraction, but this view lacks foundation. The descensus has been carefully studied in the human embryo by Bramann, 84.1. The details of the process are as follows: The urogenital fold is a long structure reaching to the posterior or caudal end of the abdomen. The greater part forms the Wolffian body, and when this atrophies the fold is much reduced ; toward the head end it contains the testis and the remnant of the Wolffian body (epididy- mis) , the portion of the fold dorsal of these acting as a suspensory membrane to which the name of mesorchium has been given, and which is comparable to the mesentery ; it is quite thick, but finally disappears. The part of the urogenital fold tailward of the testis contains the Wolffian duct (vas deferens) and runs to the point of the abdomen, where the inguinal ring is developed. A portion of this region of the fold is converted into the gubernaculum Hunteri, by an ingrowth of muscular fibres from the obliquus internus and obi. transversus. The mesorchium, together with the posterior part of the fold, including the gubernaculum, is the homologue in the male of the broad ligament of the female. To complete the statement of the homologies, it may be added that the gubernaculum becomes the cremaster, and is said to be the equivalent of the round ligament of the uterus in the other sex; the latter identification needs confirma- tion. 1 , J. The first change which occurs is the nearly complete disappear- ance of the long piece of the urogenital fold, which lies tailward of the testis. Accordingly we find the male gland at the end of the second month has moved into the immediate neighborhood of the 33 498 THE FCETUS. inguinal ring, with which it is connected by the short remnant of the fold, Fig. 269, A. The vas deferens has a nearly horizontal transverse course. The second change is the conversion of this hind remnant of the urogenital fold into the gubemaculum, a process which begins with the fourth and ends with the sixth month, it requiring about two months for the gubernaculum to attain its maxi- mum size. To form the structure in question, the fold behind the testis enlarges both longitudinally and transversely until it measures 8-9 mm. by 3-4 mm. ; the testis moves forward meanwhile a corre- sponding distance. At first the gubernaculum consists only of the Fig. 269 (To Illustrate the Descensus Testiculorum).— A, Foetus of Fourteen to Fifteen Weeks. X 2 diams. B, Foetus of the first half of the seventh month with the processus vagi- nalis opened. Te, Testis; epd, epididymis; v.d, vas deferens; p.v, processus vaginalis; r, rec- tum; r.w, bladder; v.sp, vaaa spermatica. After Bramann, peritoneum and the inclosed connective tissue, but soon muscular fibres appear in its caudal portion ; these fibres can be traced to a connection with the obliquus internus and obliquus transversus; they form a sheath or mantle underneath the peritoneum and around a central core of connective tissue ; at first they do not reach to the testis, but stop at that point where the gubernaculum is crossed by the vas deferens. They appear to extend farther forward later. The fibres are not parallel, but quite irregular in their courses. At the lower end of the muscle a bundle of connective fibres extends beyond the gubernaculum into the side of the processus (see below) . The gubernaculum is now completely differentiated. This stage of the organ is permanent in some rodents and other mammals, low in the series, and must be considered of great phylogenetic significance. Attention is directed to the fact that its muscular fibres are striped, and to its shape shown in the figure, because both points accentuate the resemblance to the rodent cremaster. While the gubernaculum is being formed there appears at its caudal end a little pouch made by an evagination of the peritoneum at the inguinal ring. This is ±h.e anlage of the processus vaginalis; it lies laterally and ventrally ■of the end of the gubernaculum ; it enlarges very slowly up to the ■end of the sixth month, but after that more rapidly. The third change is the true descent of the testis ; the evagination of the pro- cessus includes not only a considerable peritoneal surface, but also the gubernaculum, and later the testis ; in other words the urogen- ital fold extends down the processus and forms, indeed, the dorso- medial wall of the sac ; as the sac grows down, the fold (gubernacu- SPECIAL HISTORIES OF THE UROGENITAL ORGANS. 499 lum and testis) grows with it, Fig. 269, B. In a transverse section the lumen of the processus vaginalis appears somewhat crescent-like, the concave wall corresponding to the protuberance of the urogenital fold, the convex wall to the peritoneal covering. The mesorchium disappears during the descent into the processus. Early in the sev- enth month, the testis is drawn into the mouth of the sac. Fig. 269, B, and shortly after lies wholly in the interior thereof. But the testis descends to the bottom of the processus; this translation is accomplished during the seventh and eighth months, first by a shortening of the gubernaculum, second by a slipping down of the testis over the muscles ; the portion of the gubernaculum between the testis and the base of the processus is reduced to an inconspicuous band of connective tissue. The muscle now lies between the testis and the base of the penis and is developed in that position into the cremaster. At last the processus enters the scrotum, and an entirely new relation of parts is established ; the transportation of the testis into the scrotum represents a very advanced stage, since it takes place only in the higher mammals, and accordingly we find it to occur very late in the development of man. The Broad Liigament. — The broad ligament of human anat- omy is the persistent urogenital fold, reduced to a relatively thin suspensory membrane for the ovary and uterus by the abortion of the ^^x, Ov Wolffian tubules. The fundamental y ..■■'-£^S ''''■■"'••. i\'-'"^N relations here involved may be readily r? ' i^|''«J?) \*^fe. \;^:::«i. .A understood from Fig. 270, which rep- |-:' '*?:<^° °"''' :';'^f~""--v:^-L^^^^ resents a section through the urogen- '] "' J 1 • T-> of the Third Month. Pm\ Parovarium ; IS the anlage or the parovarium, I^ar, spo, epoophoron; m.d. Mailer's duct or the other, near Miiller's duct, 3Id, ^|^°f''"' '"'^<'' '^' °''^^- ^^^"^ ^^■ is the anlage of the epoophoron, Epo. The ovary, Ov, projects from what was originally the medial side of the Wolffian body, with which it is connected by a thin mesovarium. As in the adult the broad ligament contains the parovarium and epoophoron, it is evident that it is really the Wolffian body, con- verted into a suspensory membrane, most of the Wolffian tubules being aborted. The development of the broad ligament is accompanied by a change of position, first of the whole genital fold, second of that part of the fold which forms the ligament. It will be remembered that the two folds unite in part to form the genital cord, out of which the uterus and vagina are developed ; the remainder of each urogenital ridge is transformed in the female into the broad ligament and ovary. As the foetus grows, the urogenital ridge fails to grow proportion- ately, and after the second month becomes more and more restricted to the caudal or pelvic end of the abdomen. Its relative position is so rapidly shifted that by the end of the third month it already occu- 500 THE FCETUS. pies its permanent situation. While this modification is being established the Wolffian body in large part aborts, and the portion of the fold in front of the genital cords comes to occupy an oblique and finally a nearly transverse position, which is permanently retained, so that the broad ligament is always obliquely transverse. At three months the ovary is as long (3 mm.) as the Fallopian tube, and stretches in an obliquely transverse direction from the upper or cephalic end of the genital cord (future uterus) to the Miillerian funnel or fimbriate opening of the Fallopian tube. By the fourth month the transverse position is more marked, and since the ovary is originally on the medial side of the urogenital ridge, it remains on that side, and accordingly is situated on what is known in human anatomy as the dorsal side of the ligament. The development of the round ligament and of the ligament of the ovary have been but little studied by modern methods ; Mihalkovics, 85.1, 418, and G. Wieger, 85.1, have shown that they are parts of the same cord of tissues, and that by the assumption by the ovary of its transverse position this cord of tissue is subdivided into the two ligaments by becoming bent at the summit of the uterus. The primitive ligament is usually regarded as the homologue of the gubernaculum of the male. Epididymis and Epooplioron. — It is desirable to treat this organ, which is known under different names in the two sexes, as a single organ, not, as is often done, as a distinct organ in each sex. In both sexes there is a small number of permanently preserved and considerably modified Wolffian tubules from the anterior part of the urogenital ridge, which remain permanently connected with the ceph- alic or upper end of Wolffian duct. The organ thus formed becomes in both sexes very closely associated with, indeed we might better say incorporated in, the sexual gland. In the female the organ is rudimentary and has been variously named ; as it was first accu- rately described by Rosenmuller, 02.1, it has been widely known as the "organ of Rosenmiiller ;" Kobelt, who demonstrated, 47.1, that it was a remnant of the primitive kidney, introduced the term "parovarium." Waldeyer has proposed, 70.1, 142, "epoophoron" to be comparable with the epididymis, with which he recognized the parovarium to be homologous. In the male the organ has great functional importance, for its tubules serve to convey the sperma- tozoa from the seminiferous tubules to the Wolffian duct, and accordingly it is in the male that the full development of the organ is attained. A. Epididymis.— In the male human embryo of the third month there are found from ten to twenty tubules in the anterior part of the Wolffian body, which have become connected with sexual cords or future seminiferous tubules of the .testis, and have retained also their connection with the Wolffian duct. These tubules constitute the epididymis, and the portion of the Wolflfian duct which follows immediately below them, by becoming very much convoluted gives rise to the so-called head of the epididymis. At three months (K61- liker, "Grundriss," 3te Aufl., 436) traces of glomeruli can be still found in the primitive kidney, and the epithelial tubules anastomose with one another in the region between the Wolffian body proper SPECIAL HISTORIES OP THE UROGENITAL ORGANS. 501 and the testis proper. These anastomoses constitute the rete Halleri, while the Wolffian tubules become the vasa efferentia of the adult. According to Kolliker, I.e., the vasa become convoluted during the fourth and fifth month, and thereby develop the so-called C07ii vasculosi. The early development of the epididymis is known chiefly through Braun's observations, 77.4, 149, on reptile embryos. Solid out- growths appear early from the walls of the Malpighian corpuscles of the Wolffian body, and these penetrate toward the testis as cords, which subsequently acquire a lumen. The primitive connection is between the tubules of the testis and the mesonephric glomeruli — ■ a disposition which is permanent in some of the amphibia (see J. W. Spengel), but in all amniota the glomeruli disappear. C. K. Hof- mann (Bronn's " Thierreich," VI., III. Abth., p. 2062) asserts, in op- position to Braun, that the glomeruli persist in Lacerta agilis at least one year after hatching. In mammals, Mihalkovics, 85. 1 , 473, found the outgrowths from the glomeruli in cat, dog, and rabbit em- bryos of 5-6 cm., but the Malpighian corpuscles disappear early during embryonic life. B. Epoophoron (or organ of Rosenmiiller.) — Beyond tracing out the general history far enough to establish the homology with the epididymis (Waldeyer, 70. 1, 142), little has been done to eluci- date the development of the organ in the embryo. It has been already pointed out that the medullary cords of the ovary are pre- sumably parts of the epoophoron. The epoophoron is formed from perhaps ten to fifteen Wolffian tubules, and the outgrowths from the Malpighian corpuscles remain, in part at, least, solid cellular cords; the Malpighian corpuscles of the organ disappear very early in the human embryo (? third month). F. Tourneux, 88.3, has described the epoophoron in various mammals and in the human species at birth and in the adult, and has shown that its structure entirely con- firms its homology with the epididymis. Paradidymis and Paroophoron. — By these names is desig- nated, in males and females respectively, the organ constituted by the persistent tubules of the posterior part of the Wolffian body. The organ was first described in the male by Giraldes, 61.1, under the name of the "corps innomine," and was first described in the female by Waldeyer, 70. 1, 142. The persistent rudimentary meso- nephros of the human embryo has a yellowish color ; the tubules are wide, their cells pale with indistinct nuclei, and have not only no connection with the sexual gland, but have lost their original con- nection with the Wolffian duct. The position of the organ in the male and female human embryo of about three months has been figured by Waldeyer. The interesting post-fcetal changes have been made the subject of an excellent paper by Czerny, 89. 1. Genital Cord. — ^That the posterior (lower or caudal) ends of the two urogenital ridges unite into a single median mass, the genital cord, has been pointed out above, p. 491. The genital cord is a structure peculiarly characteristic of the placental mammalia, being found only in them and in certain marsupials. It does not occur in monotremes or Sauro'psida. The genital cord and its significance were first recognized by Thiersch, 53. 1. The fullest history of the 503 THE FCETUS. cord yet published is that given by Mihalkovics, 85.1, 324-347, upon which this section is based. The pelvic portions of the two urogenital ridges unite so as to form a transverse partition (rabbit embryo of about 14 mm., pig embryo of about 30 mm.). This partition is the genital cord {Cfeni- talstrang) of Thiersch. It stretches across between the rectum, which is on the dorsal side, and the allantois on the ventral side, compare Fig. 271 ; it is thick and the mesenchyma of which it is chiefly composed is a dense tissue. At the time the two ridges unite, their pelvic ends contain only the Wolffian ducts, hence the Miillerian ducts, as they develop, grow into the already formed gen- ital cord, and in the female (human embryo of 3 cm.) begin to unite almost immediately after they appear in the cord. The formation and position of the partition is well illustrated by Mihalkovics, 85.1, Figs. 114 and 115. After the genital cord is once formed, it is drawn more and more into the pelvis, and as the ccelom extends farther into the pelvis on the dorsal than on the ventral side of the Coe All -■ Fig. 271.— Cross Section of the Rectum, Genital Cord, and Allantois of a Male Human Embryo of about two Months. B, Rectum ; Coe, ccElom ; Qc, genital cord, with the two WolfBan ducts, and the median united Miillerian ducts between them; All, allantois. cord, we obtain in sections the picture reproducd in Fig. 271, which shows the typical relations of the genital cord in the indifferent stage ; the cord consists chiefly of a very dense mesenchyma, and is quite sharply bounded, except against the allantois, and it contains three longitudinal epithelial tubes, of which the median represents the united Miillerian ducts, the two lateral the Wolffian ducts, W.d. "Wolfi&an Duct. — The cephalic end of the duct remains, as we have seen, in connection with the anterior Wolffian tubules as the duct of the epididymis of the male, of the epoophoron of the female. In the male it also forms the adult spermiduct and the vesiculse seminales SPECIAL HISTORIES OF THE UROGENITAL ORGANS. 503 in the female the rudimentary duct of Gartner. It is important to note that in both sexes the Wolffian duct contributes to the formation of the utero- vaginal canal (fused Miillerian ducts) according to the observations of Mihalkovics, 85.1, and Tourneux, 87.2, upon the male, and of Van Ackeren, 89. 1, upon the female. A. Spbrmiduct of the Male.— Since the demonstration, by Johannes Miiller, that the Wolffian duct becomes the spermiduct, little has been done upon the history of the male canal. Thiersch in 1853 drew attention to the union of the caudal parts of the urogenital fold as the genital cord, see above. This cord exists temporarily in the embryo of man, and while it lasts the two spermiducts run through it, together with the two Miillerian ducts, which partially abort later. This stage is described and figured by Kolliker (" Ent- wickelungsgeschichte," 1879, p. 985 and Fig. 598). Later the geni- tal cord divides, and its dense tissue forms a thick wall around each epithelial Wolffian duct. The vesicula seminalis arises as a lateral evagination of the Wolffian duct. At five months the evagination is a simple sac about 1 mm. long, and is situated entirely within the genital cord. The evagination passes at first out horizontally, and then bends upward (Mihalkovics, 85.1, 379). B. Gartner's Canals of the Female, so named after their discoverer, are epithelial tubes which are sometimes found in the walls of the uterus, and even of the vagina, one on each side. Their significance is said to have been first recognized by Jacobson in 1830 and to have been clearly demonstrated by Kobelt in 1847. It is known that the Wolffian ducts always run through the genital cord. Fig. 371, and can be usually seen in cross sections of the uterine portion of the genital cord of the female human embryo of four to five months, and occasionally in older specimens, and even in the adult. On the disappearance of the Wolffian duct see Van Ackeren, 89.1, 34. In the foetus the Wolffian ducts open into the vagina during the fourth month ; their ends dilate and the dilated cavities fuse with the lumen of the vagina. Miillerian Duct. — The history of this duct is the reverse of that of the Wolffian duct, since it becomes rudimentary in the male, and the functional sexual duct in the female. A. Male. — Hydatid op Morgagni and Uterus Masculinus. — In the male the middle part of the Miillerian duct usually aborts, leaving the upper part with its open funnel close to the testis and the lower part within the genital cord. The upper part gives rise to the hydatid of Morgagni as maintained by Kobelt, 47.1, and later by Waldeyer, 70.1, 137. This explanation of the origin of the hydatid has by no means been put beyond question by strict observations — but we need no additional evidence to set aside the notion of Fleischl and Krause that the hydatid is the homologue of the ovary ( !) . The lower part of the Miillerian duct is contained within the genital cord, where it unites with its fellow to form a single median canal between the two Wolffian ducts — compare Kol- liker, "Entwickelungsgeschichte," 1879, Fig. 598, m. This canal corresponds to the cavity of the uterus and vagina in the female. It varies greatly in its degree of development in individuals, but 504 THE PCETTJS. usually persists in the adult as a small sac, known as the uterus masculinus or vesicula prostatica. According to E. H. Weber, the vesicula, which is rudimentary in man, is well developed in various mammals. For references to the literature of the subject, and an account of the organ in the rabbit, see V. von Mihalkovics, 85. 1, 364-378. The WolflSan duct contributes to the formation of the uterus masculinus, as it does to the formation of the vagina in the female, see p. 506. B. Female. — Fallopian Tube, Uterus, and Vagina. — These are developed from the Miillerian ducts, but it is to be remembered that, strictly speaking, the epithelial Miillerian ducts produce only the epithelial lining of the adult tuba, uterus, and vagina, and that the connective tissue, which forms the thickest part of the walls in the adult, is developed from the mesenchyma of the urogenital fold. Fallopian Tube. — The fullest account is that given by Mihalko- vics, 85.1, 304-306. The tube is developed from that part of the Miillerian duct which runs along the Wolffian body and is not included in the genital cord. The epithelium becomes much thinner except in the funnel, where it retains its cylindrical character. Later — in chicks about the eighth to tenth day — ^the mesenchyma begins to condense around the duct, thus initiating the development of the con- ^ nective - tissue ' """" '. ^ ' ^ coats of the . j^^ . „ •■■;'/ tube; shortly ^p, ' ll "„-■=- "'*-' ">"' y^ after the mes- ^tO «-'"«, *^ A' --— _ ^^'^ enchyma wall n 2i ""'W7 Oj ^— ^— begins to de- ^— ^^ *" veloptheMul- \^ lerian funnel becomes larger, and „ ^^^A r its surface thi own into folds — the a - -!"'■*» ^ anlages of the fimbriae. As the \ '^ " --^ Wolffian body atrophies and is^- changes mto the transverse broad Fig. ar«. -Section ot Broad Ligament of a ligament, the Fallopian tube ap- Femaie Human Embryo ot four Months, pears more and more at the cdsre Minot Collection, No. 35. To show the Fal- tf +i,„ -j. ij; ij j i lopian tube, F.T, and woifBan duct, w.d. 9 Urogenital lold, and changes its primitive longitudinal course to a transverse one — the primitive course being retained until the end of the third month. After the third month the tube elongates faster than the broad ligament and consequently assumes a sinuous course. By the sixth or seventh month, the definite transverse position is attained. By the fourth month. Fig. 272, the folds at the ovarial end of the tube, F. T, are well developed, but the thick dense mesenchy- mal coat is not yet divided into muscular and adventitial layers ; at this time the small Wolffian duct, W.d, still persists, though later it usually disappears. Uterus and Vagina.— As stated above, p. 503, the genital cord contains four ducts, compare Fig. 371: the two laterally placed Wolffian ducts, and the two Miillerian ducts, which lie nearer the median line and more dorsally. In man the genital cord is the an- lage of both the uterus and the vagina; within the cord the two Miillerian ducts unite in the median line, forming a single canal- SPECIAL HISTORIES OF THE UROGENITAL ORGANS. 505 the cephalad portion of this canal becomes dilated into the uter- ine cavity, and its epithelium becomes the lining of the uterus; the caudad portion develops into the vagina ; the mesodermic tissue of the cord is converted into the muscular and connective tissue layers of the adult passages; finally the Wolffian ducts atrophy, ususally completely, but they sometimes persist to a greater or less extent as rudiments, known as Gartner's canals, which lie on one or both sides in the walls of the uterus. Our knowledge of the development of the uterus and vagina is based upon numerous investigations. The fusion of the Miillerian ducts was known to Johannes Miiller, 30.1, in 1830, but he failed to ascertain that the fusion produced not only the uterus, but also the vagina ; the latter was regarded for some time as a derivative of the urogenital sinus. This error was corrected by Bischofif in his manual of "Embryology." Important advances were made by Thiersch, 52.1, and by Leuckart in his important article " Zeugung, " in Wagner's " Handworterbuch," 1853. We pass to the modern period of investigations based chiefly on the microscopic study of sections. L. Fiirst described in 1867 the fusion of the MiJllerian ducts very accurately. H. Dohrn's researches (" Schriften Ges. N"at.- wiss.," Marburg, 18U9, No. 3, also Bd. IX., 1871, p. 255) confirmed Fiirst's observations and did much to elucidate the history of the uterus. Finally may be mentioned the very admirable monographs of Tourneux and Legay, 84.1, of Mihalkovics, 85.1, 333 and 317, and W. Nagel, 91.2, which have been my chief guides in the pre- paration of the following account, and to which I refer the reader for a fuller index of the literature. Interesting additional details have been recorded by Van Ackeren, 89. 1. The genital cord extends by the fourth month from the insertion of the Hunterian or round ligaments to the urogenital sinus. The Miiller's ducts fuse in the median line between these two points, ex- cept at the upper end ; that is to say, the ducts diverge, after the complete fusion, a little below the round ligaments and these diver- gent portions become the horns of the uterus. The fusion commences at the end of the eighth week about two-thirds of the way down from the cephalad end of the cord to the urogenital sinus, and pro- gresses from that point both upward and downward, but the upper two-thirds are united before the lower extremities. The process is completed according to Fiirst by the end of the third month. In the pig, mouse, and rabbit, the fusion commences at the same relative point, but in the sheep it begins higher up. The single canal thus produced is known as the genital canal, or better as the utero- vaginal canal. A failure of the lower ends to fuse leaves two openings (double or biperforate hymen) . W. Nagel, 91.1, has pointed out that the genital cord becomes bent very early in the human embryo, so as to divide the cord into an upper or uterine limb, which is inclined ventralward over the bladder, and lower or vaginal limb, which runs longitudinally be- tween the bladder and rectum. At the end of the third month, the simple epithelium lining the cavity of the canal changes its charac- ter in its lower third, becoming there a stratified pavement epithe- lium, which passes over very gradually into the cylindrical epithelium 506 THE FCBTUS. of the upper portion. The change progresses upward, and as it ad- vances, the demarcation between the two kinds of epithelium becomes sharper. By the eighth month the passage is abrupt and occurs at the middle of the canal. The stratified epithelium lines the vaginal limb, which occupies half the genital cord at birth. After birth the uterine limb enlarges more rapidly than the vaginal. Vagina.— During the fourth month the vaginal limb expands lat- erally and becomes flattened dorso-ventrally. Its two epithelial surfaces meet and grow together, closing the lumen of the vagina and forming an epithelial lamina, the cells of which now commence a rapid proliferation which thickens the vagina and forces down its lower end, thus forming the hymen because the actual diameter of the vagina, where it is connected with the sinus, does not share in the general dilatation. The epithelial plate of the vagina has two features requiring special mention: 1. A short distance above the sinus it is T-shaped in transverse section ; the two side portions are probably remnants of the Wolffian ducts which unite with the vagina at this point. In this connection it is significant to observe that in the cow the persistent ducts of Gartner (Wolff) open into the vagina ; the question arises whether this connection is not general in the Placentalia. 2. At its upper end the lamina forms a cup-shaped outgrowth, which embraces the lower end of the uterus. Every- where between the two points thus specialized the lamina is crescen- tic in section, the concavity facing the back. The anlages of the rugae of the vagina appear during the end of the fourth month as budding ridges on the outside of the lamina. Finally the permanent lumen of the vagina begins to appear during the sixth month and is formed by the breaking down of the central cells of the lamina. This pro- cess penetrates the cup-shaped outgrowth just described, so that the lower end of the uterus protrudes into the vagina, in consequence, be it remarked, of the vagina growing up around the extremity of the uterus. The stratified epithelium often extends a short dis- tance inside the os uteri. Uterus. — The cavity of the uterine limb is always open, and its epithelium composed of a single layer of cells, which diminish in height from 60a (third month) to 25a (eighth month). A short time before birth the epithelium of the cervix develops into beaker cells. This transformation has been well described by Moricke, 82. 1. The cells increase in length and the nuclei move toward the base of the layer; the upper portion becomes clear and no longer stains with picrocarmine owing to the formation of mucus. These cells secrete the mucous plug which fills the cervix at birth. As far as ascertained there are no cilia in the foetal uterus. The develop- ment of the arbor vitse of the uterus commences at the end of the fourth month with the appearance of the main stems (rachis) , which extend from a little above the future os nearly to the fundus. Their disposition is asymmetrical, the two stems of the posterior wall lying to the left, of the anterior wall to the right ; hence the cavity of the uterus is somewhat S-shaped in section. The arbor vitae is merely a set of folds of the uterine mucosa. The mesoderm of the genital cord differentiates very slowly. The first noticeable change is the increased vascularity of the part next SPECIAL HISTORIES OF THE UROGENITAL ORGANS. 507 the epithelium; this vascular layer becomes the mucosa, and the tissue outside it the muscularis. The latter does not become distinct histologically until the close of the fifth month. The muscular fibres are very irregularly disposed ; however, the trend of the inner ones is circular, of the outer longitudinal. The glands of the uterus and vagina do not appear until after birth, except in the cervix uteri, the glands of which arise the middle of the fifth month (Van Ackeren, 89.1). Cadiat, 84.1, maintains that those of the corpus uterus arise during foetal life. This opinion I consider erroneous ; has not Cadiat mistaken folds of the arbor vitse for the anlages of glands? The following table indicates the growth of the uterus and vagina : Foetus from ver- tex to coccyx. Supposed Age. Canal, Length. Vagina, Length. Uterus. 7. cm 9.0 5. mm. 6.5 13.0 31. 39. 3.3 mm. 3.5 6.5 10. 16. 3. 8 mm. 3.0 13.5 16.0 30. .. 6.5 '11. 13. Child Eight days. Four months. Three years. 35. 50. 65. 10. 30. 40. 35. Child 30, Child 25 A few words must be added concerning the comparative morphol- ogy of the uterus. The round or Hunter's ligaments mark in all mammals the division between the Fallopian and the uterine por- tion of Miiller's ducts. In man the whole of the uterine division is included in the genital cord and participates in the formation of the single median uterus. Hymen. — The hymen is said to be the homologue of the verum montanum of the male urethra. It appears about the beginning of the fifth month as a transverse ridge situated on the ventral side of the vestibular end of the vagina, and projecting into the urogenital sinus (vestibulum) . At this time the vagina begins its dilatation, and as it widens it appears to force down the hymen, which is thereby rendered more protuberant. The hymen is a thin non-muscular fold covered on one surface by the epithelium of the sinus, and on the other by the epithelium of the vagina, the latter being much thicker than the former. The hymen grows rapidly after its first appear- ance. When, as may happen by an arrest of development, the lower ends of the Miillerian ducts do not fuse, the hymen presents two orifices leading into a single vagina (H. Dohrn, 75.1; Tour- neux and Legay, 84.1, 345). The development of the hymen has been studied by H. Dohrn, 75.1, 78.1, Tourneux et Legay, 84.1, Mihalkovics, 85.1, 349, and Van Ackeren, 89.1, 30. Development of the Kidney.* — The true or permanent ma- niote kidney has no homologue, so far as known, in the amniota, the so-called kidneys of the latter being Wolffian bodies (mesonephri). We are entirely unable at present to trace the probable evolution of the kidney, for the view advanced by Semper, 75.2, that it is a modi- fication of the hind end of the Wolffian body is negatived by the * For further details see Ove Hamburger, 90. 1. 508 THE FCBTUS. embryonic development of the kidney. Nor do we possess any light as to the factors by which the development of the kidneys is initiated in embryos. In short, we are compelled to confine ourselves to a bare narration of the actual development, as known at present. The Renal Anlage. — The renal anlage consists of three parts, the epithelial evagination of the Wolfiian duct, the condensed mesen- chyma, and Braun's cords, which appear in the order named. The epithelial evagination appears on the dorsal side of the Wolffian duct near the opening of the duct into the intestinal canal (cloaca) . The evagination appears in the chick at the end of the fourth day, in crocodiles of 13 mm., in Lacerta agilis about eight daj's after the eggs are laid, in the rabbit the eleventh day, in sheep embryos of 8 mm., in human embryos of 5 mm. The evagination rapidly changes in character : first, by elongating forward and by the en- largement of its cephalic end. Fig. 444, N; second, by acquiring (in the chick by the sixth day) a direct opening into the urogenital sinus or hind end of the intestinal canal. The enlarged blind end is the anlage of the epithelial portions of the kidnsy, that is to say, of the lining of the pelvis and of the renal tubules ;_ the remainder of the evagination becomes a long nafrowTube7 which may be at once designated as the ureter, although it corresponds, of course, only to the epithelial lining of the adult ureter. The way in which the evagination grows is well illustrated in Fig. 444, B,D,C. The blind end of the renal evagination grows forward on the dor- sal side of the Wolffian body and continues this growth while it is developing into the kidnej'', so that the more advanced the kidney in its differentiation, the more of the Wolffian body is covered dor sally by it. The mesenchyma around the blind end very soon becomes condensed, but the condensation, at least in crocodiles, occurs chiefly on the medial side of the renal tube. The relations just described are well illustrated in Fig. 273. The condensed mesenchyma can be followed some distance along Fis. ?r3. -Cross Section through the Hind '^^^ t^reter and there gradually be- End of the Left Wolffian Body of a Crocodile COmeS loOSer, and its COnCCntric Embryo of 13 mm. iV, Nerve; ^o, aorta; M, , %. /txt- j mesentery; in, intestinal canal; *.<, Wolffian arrangement disappears (WiederS- tubules; W.d, Wolffian duct; msth, mesothe- rttim 90 S AAR^ TVio Tn-irvii+iiTo lium; car, cardinal vein; fct, evagination to "^1™) »U.O, ttb) . ine primitive form the kidney ; mes, condensed mesenchyma. anlage of the kidnev, therefore. After Wiedersheim. '^ . , , , .i , % , „ comprises the dilated end of an epithelial tube and condensed mesenchyma. It is convenient to consider the history of the two separately. Mesenchyma. — The histogenesis of the mesenchymal portions of the kidney is almost unknown. It seems to me particularly desira- ble that the history of the blood-vessels should be ascertained. Grolgi, 89. 1, 341, observed that in the fcetal kidney the arteries subdivide and form an irregular network of capillaries in the peripheral portion car-' SPECIAL HISTORIES OF THE UROGENITAL ORGANS. 509 of the orgaa, and when the glomerulus begins to form it contains a single loop of this network, and later from this primary loop second- ary loops bud forth until the circulation of the glomerulus is com- pleted. It is important to note that the fibrous capsule is developed very early, before there are any glomeruli — for instance, it is present in the kidney of an embryo rabbit of fifteen days, and at sixteen days is figured as quite thick by Kolliker ("Entwickelungsgesch.," 1879, Fig. 581) . The capsule is definitely present in human embryos of 30-25 mm. length and is formed of spindle-shaped anastomosing cells (W. Nagel, 89.3, 367). My observations have led me to be- lieve that the capsule is the essential mechanical condition for the development of the glomeruli, compare below. Tubules and Malpighian Corpuscules. — The tubules arise as branches of the blind end of the renal evagination and the blind ends of the branches form the so-called Malpighian corpuscles. The branching begins very early, compare Fig. 444, D, C, and gradually a distinction becomes recognizable between the enlarged end of the ureter, destined to form the pelvis and the tubules proper — a distinction which be- comes more and more marked as develop- ment progresses. The branches are at first __ short but wide, and form wide angles with ^ another ; their walls are a rather high one cylinder epithelium. At an early period — in the rabbit by the fourteenth day — the branches reach the capsule, which has mean- while been differentiated from the surround- ing mesenchyma. The capsule seems to prevent the further elongation of the branch in its line of growth, and to force the end of the branch to curl over, thus by a simple mechanical condition causing the formation of the anlage of the Malpighian corpuscle. This role of the capsule has not been noticed hitherto, so far as I am aware. My atten- tion was called to it by observing that in older kidneys (human embryos of three, four, and five months) the formation of the Malpighian corpuscles always goes on close against the capsule. Fig. 274; one sees a straight collecting tubule, which runs to the capsule and there bends over into the anlage of the convoluted tubule and Malpi- ^fig. 274^ -section of a Kidney, , . " 1 ,1 ,1 1 Human Embryo of about five ghian corpuscle ; the younger the corpuscle Months. Minot collection, No. 34. the nearer is it to the capsule. To explain %^^'l'i^^^''tX^^^^g the difference in position, we must assume tubuie. that the corpuscles remain approximately where they arise and that the capsule enlarges, and thereby gives opportunity for new Malpighian corpuscles to be developed out- side of the older ones— examination of the carefully drawn Fig. 270 will make the distribution of the corpuscles clear. The collect- 510 THE FCETTJS. ing tubules appear to all arise as branches, at first from. the end of the ureter, after that from the collecting tubules already formed — the details of their development have still to be ascertained ; at first the branches devaricate at wide angles, but later they show the characteristic U-shaped fork of the adult, compare Fig. 276, col. The convoluted tubules and Malpighian corpuscles develop accord- ing to Golgi, 89.1, as follows: The end of the tubules bend over, Fig. 376, into an S-shape; in Golgi's diagrams each main tubule is represented as forming two convoluted tubules at once ; whether this is the case is not quite clear ,from his text, but it is probably true, I think, of the first-formed Malpighian corpuscles, but later each straight tubule forms, so far as I can observe, only one corpuscle and convoluted tubule. The different parts of the S-shaped tubule have each their fixed destiny. The end of the S (in the diagrams the lower part) receives the vascular loop, which gives rise to the blood- vessels of the future glomerulus, gl; the lower limb of the S, a, elongates enormously and forms the first division of the convoluted tubule including the loop of Henle, H; the upper limb, 6, of the S also elongates very much — though less than the lower limb — and is Ftg. 275.— Semidiagrammatic Figures of Developing Benal Tubules of a Mammal 1 2 3 4, 5,6, Successive stages; gl, blood-vessels of glomerulus; a, first, b, second portion of convo- luted tubule ; H, Heme's loop. After Golgi. the anlage of the second division of the convoluted tubule; where the two limbs join the tubule passes close to the Malpighian cor- puscle and seems to be intimately attached to it. This attachment is preserved, according to Golgi, in the adult kidney. During de- velopment it acts as a fixed point, which parts the convoluted tubule into two primary divisions, which, as is well known, are persistent. Henle's loop rapidly elongates in the direction parallel to the straight or collecting tubule and toward the medulla, its elongation perhaps explaining why it increases in diameter less rapidly than the remain- ing parts of the tubules. The development of the corpuscles has been described quite fully by Toldt, 74.1, and also by KoUiker in his " Entwickelungsgeschichte, 1879," 949, but it is to be noted that the S-shaped tubule is not merely the anlage of the Malpighian corpuscles as supposed by these authors, but also of the convoluted tubule! The blind end alone forms the corpuscle ; the wall of this end is pushed in by the very formation of the S, and the end assumes somewhat the shape of a rubber ball with one side pushed in (Toldt) in the concavity of which a network of capillaries appears, Fig. 375 ql In older kidneys of the human embryo, the concave side is always SPECIAL HISTORIES OP THE UROGENITAL ORGANS. 511 turned away from the straight collecting tubule with which the cor- puscle is connected, Fig. 276. The epithelium upon the convex side is much thinner than that of the concave side, and as development progresses this difference becomes more marked ; the space of the ; Cap* C-^ j3 r H- M< o gj ^ \ !t u V r. '5 col col Fie. 376. —Section Parallel to the Medullary Rays of the Kidney of a Human Foetus of about flTO Months. Minot Collection, No. 34. Explanation in text. tubule is the cavity of the corpuscle ; the thin epithelium is the lining of Bowman's capsule; the thicker epithelium covers the glomerulus. The further differentiation depends chiefly upon the assumption of the spherical form and upon the growth of the glomerulus and its vessels. The original area, by which the vessels enter the glomer- 513 THE FCETUS. ulus, remains about the same, or perhaps even diminishes in size, but the Malpighian corpuscle grows, and hence the neck by which the vessels enter becomes relatively much smaller. The corpuscles continue their growth for a long period, and are smaller in the foetus than in the adult, therefore they must continue to grow after birth. Some authors have maintained that there is an atrophy of some of the tubules of the foetal kidney, but I agree with Golgi, 89, 1, in believing that of this there is no valid evidence. I present figures of two typical sections of human foetal kidneys, Figs. 376, 377. The first, Fig. 376, represents a radial section of a ^J Fie. 277. —Cross Sections of the Medullary Tubules of the Kidney of a Human Embryo of about Five Months. Minot Collection, No. 34. Explanation in text. kidney at about five months. The capsule, Cap, is fibrous and thick. The separation of the cortical, C, and medullary zones, M, is given by the distribution of the Malpighian corpuscles, of which the youngest stages are found near, the oldest farthest from, the capsule; between the two zones are situated the main blood-vessels, vv, drawn dark in the figure; the medullary rays, S, are distinct, but consist each of only a few tubules : the convoluted tubules, cc, are very pale and not all of them are represented ; to render the figure clearer they are drawn without nuclei ; Henle's loops, H, are found at all levels, and show, as yet, no very distinctive histological features; SPECIAL HISTORIES OF THE UROGENITAL ORGANS. 513 the collecting tubules, coll, are large and show the typical branching with great perfection. Especially characteristic of the foetal kidney is the large proportion of connective tissue, and the consequent wide separation of the tubules. The second section is through the medulla at right angles to the direction of the tubules, Fig. 277. Here the wide separation of the tubules by connective tissue is more apparent than in the previous iigure. The collecting tubules are large and have a cylinder epithelium with evenly placed nuclei; the Henle's tubules, H, are much smaller, but vary greatly in size; as Golgi has pointed out, it is sometimes the ascending, sometimes the descending, limb which is small. Every collecting tubule is sur- rounded by a space, which at first I thought artificial, but as I find it in all specimens, including the freshest and best preserved, I con- clude that it exists during life, and regard it as probably a lymph space. Braun's Cords.^ — Max Braun, 77.4, 199-201, described cords of cells, which extend in very early stages of the embryos of lizards through the renal anlage. These cords differ but little from the rest of the mesenchyma, except in having their cells more closely crowded together, and that they can be traced to a direct connection with the mesothelium. This observation has since been confirmed — on chicks by A. Sedgwick, 80.1, on crocodile and turtle embryos by R. Wiedersheim, 90.3. The cords I must regard from Braun's own descriptions as merely the beginning of the condensed mesenchyma of the renal anlage. The three authors who have observed the cords regard them as the anlages of the convoluted tubules, though they bring no direct proof in support of this view, and since it has been positively demonstrated that the convoluted tubules arise from the collecting tubules, the view in question must be abandoned. Shape. — The kidney is early marked out definitely by the devel- opment of its capsule, and in its first form is already "kidney- shaped," and has a smooth surface. When the development of the Malpighian corpuscles begins, the surface of the kidney changes, and at ten weeks (Burdach) is already divided into lobes, separated from one another by shallow but sharply defined depressions. The number of lobes is usually about eighteen in the human embryo, but Burdach ("Physiologie," Bd. II., 1828) describes eight lobes at ten weeks. The lobate stage is found in all amniota and is permanent in Sauropsida and cetaceans, but in most mammalia is confined to the foetal period. In man the lobes disappear soon after birth and the surface of the kidney again becomes smooth. Each lobe corre- sponds to the base of a Malpighian pyramid. The appearance of the foetal kidney is also affected by its upper end being covered by the relatively large hood formed by the supra- renal capsule. Human Kidney. — The following dates as to the development in man are taken chiefly from Kolliker's " Entwickelungsgeschichte," 1879, p. 952. In an embryo six or seven weeks old the kidney meas- ured If mm., was flattened, bean-shaped, and overlaid the WolfiBan body. In the eighth week, it measured 2.5 mm. long, and lay far behind the large supra-renal capsule, with which it comes in contact during the third month. The lobules, as first fully described by 33 514 THE FOETUS. Toldt, 74.1, appear during the second month and remain marked upon the external renal surface throughout foetal life. The Malpi- ghian corpuscles begin to form toward the end of the second month, and continue forming until a few weeks after birth. The Henle's loops, as shown by Golgi, 89.1, begin their development immedi- ately after the corpuscles appear, and may be recognized in three- months' embryos, as I have observed, but are not well developed until the fourth month. Ureter. — Concerning the embryonic history of the ureter little is known. KupfEer, 65.1, 66.1, has shown that the stretch of the Wolffian duct between the original evagination and the urogenital sinus elongates somewhat, but as development proceeds this part becomes included more and more in the sinus, with the result that the two canals open separately. During these changes, the ureter becomes twisted so that its opening is situated in front of that of the Wolffian duct. As to the histogenesis of the ureter I know of no observations. Historical Note. — Remak, 50.1, was the first to describe cor- rectly the development of the kidney ; he observed the forward growth of the ureter from the cloaca, the enlargement of the end of the ureter, and the outgrowth from it to produce the collecting and con- voluted tubules. Kupffer, 65.1, 66.1, showed that the ureter was an evagination of the Wolffian duct near the cloaca, and this has since been confirmed by numerous observations on all classes of am- niota ; Kupffer added also the erroneous notion that the uriniferous tubules do not all arise as products of the ureter. Unfortunately Kupffer's error was upheld by Bornhaupt, 67.1, Colberg (CM. Med. Wiss.,186d). Goette,Thaysen, 73. 1, Braun, 77.4, Sedgwick, 80. 1, Balfour, Riedel, 74.1, and Emery, 83.1, and even Wiedersheim, 90.3. The authors since Braun have been largely infiuenced by theoretical considerations, especially by the wish to demonstrate that the true kidney is developed similarly to the Wolffian body (mesonephros) , in other words that its secretory tubules are different in origin from its ducts. Remak's original view found few uphold- ers, of whom Waldeyer, 70.1, 132, Toldt, 74.1, Kolliker ("Ent- wickelungsgeschichte," 1879), and Golgi, 89.1, and W. ISTagel, 89.3, 365, are certainly the most important. Golgi may be said to have put the matter beyond debate so far as mammals are concerned. My own observations enable me to affirm with confidence that the tubules arise as evaginations of the ureter, and that in man the con- voluted tubules and Malpighian corpuscles arise as branches of the collecting tubules. The facts are so clear that it is difficult now to understand how the opinion could have been entertained that the convoluted tubules arose from the blastema, and not as outgrowths of the collecting tubules. Allantois and Bladder.— That portion of the allantois which lies within the body of the embryo, and extends from the anus to the umbilicus, becomes the bladder. It has been mentioned already that the ureters very early separate entirely in mammals from the Wolffian ducts and come to open into the neck of the allantois. The dilatation of the embryonic portion of the allantois to a fusiform vesicle begins in man during the second month ; one end of the vesicle is connected SPECIAL HISTORIES OP THE UROGENITAL ORGANS. 515 with the anal end of the intestinal canal, while the other end tapers out and is prolonged as the so-called urachus, into which the cavity of the vesicle is prolonged, but at some time not yet definitely ascer- tained the cavity of the urachus disappears, though seldom com- pletely, for Luschka (Virchow's ^4 rc/i., XXIII.) found remnants of it even in the adult. The urachus is transformed into the ligamentum vesicae medium (Kolliker, "Entwickelungsgesch.," 1879, p. 953). The main vesicle becomes the bladder. The entoderm of the allantois becomes the epithelium, and the mesenchyma becomes the connective tissue and muscular walls of the bladder. The histogen- esis and changes in shape of the embryonic bladder have still to be investigated. Recently Retterer, 90. 1, and Keibel, 91.1, have revived Rathke's conception, 32.1, I., 57, that the bladder is an outgrowth of the cloaca, which becomes early divided into a dorsal or intestinal and ventral or allantoic division. The distinction seems to me arbitrary between this notion and the view adopted above, since the allantois is in any case a prolongation of the entodermal canal, and neither Retterer nor Keibel show that there is a true division of the cloaca. Urogenital Sinus. — I base this section upon Mihalkovics' mon- ograph, 85.1, 307-32J:. As shown in Fig. 444, the allantois is the direct continuation of the intestinal canal, and the urogenital ducts open into the allantoic portion or the part of the canal on the ventral side of the future anus. After the anus is formed, there is a terminal portion, the so-called cloaca, into which both the intestinal canal proper and the allantoic canal open. The greater part of the allan- tois dilates into the bladder, but between the bladder and the cloaca the allantoic canal remains narrower ; it is into this narrower portion that the Miillerian ducts open; the stretch between the bladder proper and the opening of Mliller's duct is the urethra in the strict sense, while the part below received from Johannes Miiller the name of simis urogenitalis. The female adult urethra corresponds to the urethra as here defined,* but the male urethra includes both the urethra and the sinus. This may be called the monotreme stage, and is characterized by there being merely a single or cloacal opening, through which the excrement, urine, and genital products are dis- charged ; the stage is the permanent one in non-mammalian verte- brates and in monotremata. An important advance is made in pla- cental mammals by the subdivision of the cloacal opening into the ventral urogenital opening and the dorsal anal opening, which takes place in the human embryo about the fourteenth week, and involves the complete separation of the urogenital sinus from the intestinal canal. In the male the closure of the raphe penis converts the sinus into the prolongation of the urethra proper, as we may term the neck of the allantois or bladder above the opening of the fused Miillerian ducts (uterus masculinus) . In the female the sinus persists as the vestibulum into which the urethra and vagina both open. The sepa- rated urogenital and anal openings lie at first in a shallow fossa or recess, the raised edges of which represent the anlages of the external genitalia ; see the following section. Prostatic Gland. — This gland is present during the fourth * The relations are wpU shown by KBUiker, "Grundriss," Fig. 295. 516 THE PCETTJS. (Kolliker) or fifth month (Mihalkovics) as a series of branchings evaginations of the epithelium of the upper end of the urogenital sinus, which expand into wide irregular cavities. The muscular tissue is developed much later from the mesenchyma of the walls of the sinus (Kolliker, "Entwickelungsges.," 1879, p. 1000, and Mihal- kovics, 85.1, 378). The evaginations make their first appearance, according to Tourneux, 89.1, 257, about the twelfth or thirteenth week, and persist in the female, though more or less in a rudimen- tary condition (Tourneux, Soc. Biol., Paris, Jan., 1888). Cowper's and Bartholini's Glands.*— These names are ap- phed to the same glands in the male and female respectively ; they arise as paired evaginations of the lower part of the urogenital sinus. According to Van Ackeren, 89.1, 44, the glands of Bartholini be- gin their development in man toward the end of the fourth month ^ during the fifth month the branches {acini) increase in number and are found separated from one another by considerable mesenchymal tissues ; by the sixth month, as already described by R. Geigel, 83. 1, they form considerable organs 1X1.8 mm., of rounded form, but the left gland is a little smaller than the right ; the acini now lie close togetlier. III. External Genitalia. The main facts in the development of the external genitalia, and the homologies between the two sexes, were worked out by Tiede- mann — see his "Anatomie der kopflosen Misgeburten, " Landshut, 1813, p. 84. A very good description of the foetal penis and clitoris is given by Joh. Fr. Meckel in his " Handbuch der menschlichen Anatomie," 1815-1820, so that Johannes Miiller in 1830, 30.1, could add but little. Some further details have been given by H. Rathke, 33.2, and by Kolliker. Ecker in his " Icones Physiologicse" has given a. series of figures, which have been extensively copied in text-books, and have been reproduced in the well-known and some- what inaccurate wax models of Ziegler. In 1888-89 appeared Tourneux's admirable monographs, 88.1, 89.1, upon which the following account is based. We have to consider the history, 1, of the genital tubercle (penis-clitoris) , 2, of the genital labia (scrotum- labia majora). The external genitals are homologous in the two sexes, but in the male they are more specialized than in the female ; the condition in the adult female corresponds to that of the foetal male. Genital Tubercle. — The anal plate becomes very much thick- ened until it constitutes — sheep embryos 18-25 mm., pig embryos 14-20 mm. — a thick plug of epithelium, on the dorsal side of which appears an external invagination, Fig. 378, the vestibule anaJe of Tourneux, which gradually penetrates until it leaves only a thin epithelial membrane to close the rectum, while the main plug {houchon cloacal) closes the urogenital sinus or pedicle of the allan- tois, 8. ug. The accumulation of mesenchyma on the ventral side of the epithelial plug, pi, of the cloaca is indicated by an external prominence, which may be already designated as the genital tuber- cle, t. g. As development progresses the genital tubercle rapidly ♦The paper by Swieoicki in Gerlach's "Beitriige," 1883, I have not seen. EXTERNAL GENITALIA. 517 lengthens and the epithelium upon its dorsal side is reduced from a great plug to a thin layer, and by the disappearance of the plug both rectum and urogenital sinus become open to the exterior ; the mass of tissue between the two openings is termed by Tourneux the repli or eperon perineal. The genital tubercle owes its origin to the thickening of the anal plate, it gives rise to the penis in the male, to the clitoris and nymphse in the female. The tubercle is quite prom- inent — measuring 1.5 mm. in length— in the human embryo by the ■tenth week and is then found to have its end knob-like, indicating the future glans, and its dorsal or anal side with a shallow groove, which directly prolongs the channel of the urogenital sinus, but only as far as the knob of the glans. About the tenth week there appear two eminences alongside the genital tubercle and the urogenital opening, which we may call the genital labia — compare Fig. 380, lab. The labia are solid hillocks of mesenchyma with a covering of epitheli- um, see Fig. 281. In the female they persist as the labia majora and in the male as the scrotum. Penis. — ^In the male the genital tuber- cle continues elongating as follows : Foetus. . . .40. 50. 60. 105. mm. Penis 3. 2.5 3. 3.5 mm. and its dorsal groove not only deepens while it remains very narrow, but also closes, beginning at the base ; the line of closure remains permanent- ly marked by the raphe of the penis ; the effect of the closure is to form an epithelial canal which prolongs the urogenital sinus (or urethra) into the penis; the epithelial canal separates from the epithelium covering the penis, except just below the glans, where the permanent opening is established. During the third month there appears first an epithelial ridge upon the glans, as in Fig. 280; this ridge lies in the prolongation of the groove; it soon disappears and the groove extends gradually on to the glans. It is at this stage (end of the third month) that the thick prepuce of the glans begins to develop, but the groove on the anal side prevents its forming a complete ring around the organ. The prepuce appears as a slight ridge which overgrows the glans, the epithelium of the inner surface uniting, as the fold extends, with the epithelium of the glans — the two epithelia fusing into one solid plate. Fig. 297, ep. Later the groove becomes closed to a canal, and the terminal opening of the canal is shut by the growth of the epithelium, Fig. 279, a, which plugs up the oriiice. This fact is important from its bearing on the question of the origin of the amniotic fluid, p. 340. The two epithelial layers of the prepuce are separated by mesoderm. The relations which now exist can be better explained by reference to Fig. 279, which shows the glans in longitudinal section ; observe the thickened epithelium, a, closing the orifice of the urethra ; also Fig. 278. — Longitudinal Median Seo tion of the Cloaca of a Sheep Embryo of 18 mm. Coe, Coelom; Al, allantois; JEc, ectoderm ; S. ug^ sinus urogenita- lis; t.g, genital tubercle; pi, epithe- lium : an, anal plate ; JUn, entoderm. X 22 diams. After Tourneux. 518 THE PCETUS. that the epithelial plate— and consequently the prepuce also— extends further on the ventral side, ep, than on the dorsal, and that though- the glans, G, is very vascu- lar, the corpus cavernosum is remarkable for the absence of blood-vessels. Finally there may be noticed in the epithe- lial plate, ep, several places where the cells are arranged more or less concentrically; the appearances have been thought by Schweigger-Sei- del, 66.1, to be connected with the cleavage of the epi- thelial plate to form the epi- thelium of the prepuce and of the glans. This cleavage does not take place until after birth, but just when is not known. The corpus cavernosum develops slowly; it is first marked out as a dense mes- enchyma, in which the blood capillaries develop more and more, beginning in the third month ; but it is not until the sixth month that the capil- laries begin to show any marked dilatation. The cor- pus spongiosum develops also chiefly by the expansion of capillaries, but considerably later than the cavernosum. Retterer (Soc. Biol., Paris, 1889, p. 399) states that in various mammals the corpus caver- nosum is very dense and fibrous before any blood-vessels appear in it. Clitoris and Nymphs. — The development of the genital tubercle in the female is in all respects similar to that of the male, but it does Fig. S79. —Longitudinal Section of the Penis of a Human Embryo of about five Months. Minot Coll. No. 84. a, Epithelial plug ; Pp, prepuce; ep, epithe- lial lamina between prepuce and glans; ff, glans; (7, urethra. /A ci -lab not pass beyond the stage in which there is an open urethral groove. The glans and prepuce are formed, as in the male, to constitute the clitoris, but the borders of the urethral groove do not unite, as they do in the male to form the stalk of the penis, but remain as elevated ridges known as the labia minora or nymphse in the adult. During the third month the dif- ' v -•' ference between the male and female tubercle ^ , ' an becomes more and more clearly marked, and < *^ ' the distance between the urogenital and anal ;„^i? aso — E-rtemai oemtai- T^ ,, ="1 ^ ., ,i.t'^^ Female Embryo of 105 mm. openings increases. By the end or the third gi- Glans; ny, nymphae; lab, month, Fig. 280, the glans measures about 1 !S-ogeSt\r%emn|°"'.kf'^r mm. , while the lips (anlages of the nymphse) Toumeux. x about 5 diams. of the urethral groove measure about 3 mm. in length; around the base of the glans can be seen the commencing fold of the preputium, EXTERNAL GENITALIA. 519 and upon the glans can be seen the median epithelial crest, which subsequently disappears, the urethral groove extending on to the glans during the fourth month. The groove persists, so that in the adult the prepuce does not ex- tend, as in the male, completely around the glans, but is cleft on the Fig. 381. — Section of the Clitoris and Labia Majora of a Human Embryo of about four and one- balf Months. (Minot Collection No. 49. ) anal side. Tourneux, 89.1, 254, observed in two cases epithelial ingrowths from the epithelium of the groove of the glans ; these in- growths he regards as the anlages of the " glande clitoridienne" of Werthheimer 0ourn. de VAnat. et Physiol., 1883), the homologue of the mucous glands of the sinus of Guerin in the male. The mes- ^ ^- — A \ \ Fig. 282.— External Genitalia of the Female Human Foetus at about four Months. Minot Col- lection No. 57. A, Ventral, B, perineal view. cJ, Clitoris; Za, labia majora ; An, alius. enohyma of the glans persists at a stage corresponding approxi- mately to that of the homologous tissue in the male at eight months, but the corpora cavernosa develop as in the male into true erectile tissue. The accompanying Fig. 281 represents a section of the clitoris and labia majores of a foetus of the fifth month, the urethra 530 THE FCETTTS. extends into the glans, which is covered by the prepuce ; the glans is ahnost buried between the large labia majora. Scrotum and Labia Majora. — There appear two prominences during the tenth week, one on each side of the genital tubercle. These prominences, which are merely hillocks, so to speak, of mesoderm covered by fcBtal epidermis, are the anlages.of the male scrotum and the female labia majora. Their relations are well shown in Fig. 380. In both sexes the genital labia attain considerable size ; in the female the foetal type, Fig. 383, is but slightly modified, but in the male the two labia meet and unite during the fourth month between the base of the penis and the anus to form the scrotum ; the raphe marks in the adult the line of junction ; as stated above, p. 498, the vaginal processes grow into the scrotum and later the testis descends into it, p. 499. CHAPTER XXIV. TRANSFORMATIONS OF THE HEART AND BLOOD-VESSELS. We have already considered, Chapter X., the origin and early iistory of the heart and blood-vessels, and have now to consider the metamorphoses of the foetal organs of circulation to the time of birth. We shall take up, 1, the heart; 2, the arteries; 3, the veins. I. Transformation of the Heart. We left the heart, p. 288, as a median longitudinal tube, with double walls, the inner endothelium, and the outer mesothelial or muscular ; the double tube was free except at its ends, which were attached to the walls of the pericardium; the anterior end communi- cated with the aortic vessels, the posterior (caudal) end with the veins united in the septum transversum. To develop the adult heart out of this simple tube, five principal sets of changes occur: 1, the bending of the tubular heart; 2, the outgrowth of the auricles; 3, changes in the thickness and histological constitution of the walls ; 4, the development of valves ; 5, the appearance of secondary parti- tions dividing the right heart from the left. The literature upon the heart is very extensive, but the history given by His, "Anat. menschl. Embryonen," Heft III., 129-184, of the human heart, and that given by Born of the rabbit's heart, 89. 1, are so thorough that I have relied chiefly on these two authors. Special mention ought also to be made of Carl Rose's dissertation, 88.1. Besides the special papers on the heart, there are numer- ous observations scattered in general works. The general develop- ment of the chick's heart is described in the text-books of KoUiker and of Foster and Balfour, and in J. Masius' excellent article, 89.2, based upon models constructed by Bern's method. Of earlier papers that of Lindes, 65. 1, is specially noteworthy, and, as pointed out by His and Born, far in advance of his time. Bending of tlie Heart and Formation of the Auricles. — After the disappearance of the mesocardium on both the dorsal and ventral side of the primitive heart, the heart is attached only by its aortic and venous ends. The early enlargement of the pericardial cavity has been already described ; its size is important as affording the heart room to elongate, bend, and enlarge. The straight median heart grows rapidly, and to find room bends to the right; in the chick the bending begins at the close of the first day and increases very rapidly during the second day, Fig. 283, and becoming at the same time more complicated by the assumption of an irregular S-shape. In mammals the same form is assumed, and is found in the rabbit at nine days, in human embryos of 2.15 mm.. Fig. 284. 522 THE FCETUS. Pro. A The venous end of the heart Fig. 384, V.l, lies somewhat to the left and extends for a short distance toward the head and then passes into the ventricular portion of the tube, which curves, as shown in the cut, V. I, on to the ventral side, where it crosses obliquely to the right side, and then bending dorsalward finally runs toward the head and becoming narrower passes into the bulbus aortae, A.b, or division of the heart tube, which passes in the median line into the trunk of the aorta. At this stage there is a very short venous or auricular division of the heart, a very long, thick, and much bent ventricular division, and a bulbous di- vision of intermediate dimensions. The dif- ferentiation of these divisions comes out more clearly from the study of the endothelial heart (or heart cavities) at this stage, Fig. 285. The general course of the heart may be best understood by combining this figure with the preceding, remembering that Fig. 284 shows the muscular heart, which at this stage is still separated by a considerable space from the endothelial heart, and is much larger than the inner tube. In Fig. 285 the divi- sions of the heart are clearly marked; the auricular division, V.h, is very short and receives the omphalo-mesaraic, v.o.m, and umbilical, v.u, veins and the ducti Cuvieri on each side; it is continued ■^■"v Fie 283 —Head of Cluck of Thirty eight Houis seen from the Under Side. Op, Optic vesi- cles; Pro, A, proamnion; a.c.v., amnio - cardial vesicle ; Ht^ heart. X 26 diams. After Du- val. FiQ. 284. —Reconstruction of the Heart and Veins of a Human Embryo of 2.15 mm. (His' embryo Lg.). M6, Mouth; ^. 6, bulbus aort^ ; V.m, left ventricle; F.c, vena cava superior; S.r, sinus reuniens; v.u, vena umbilicalis; V. I, venous limb of heart ; Ho, anlage of auri- cle; i>, septum transversum; Lb, liver; Lbg, hepatic duct, x 40 diams After W. His. V.h r.c. Fig. 286.— Endothelial Heart of a Human' Embi-yo of 2.13 mm., seen from the Left Side (His' Embryo Lg.). ^.6, Bulbus aortae; Fr, fretum Halleri; V, ventricle; V.h, auricle; V.o.m, vena ompnalo-mesaraica; V.u, vena umbilicalis; V.c, vena cava; C.a, auricular canal. X 40 diams. After W. His. headward by a portion, C.a, the auricular canal, wliich connects the auricle with the ventricle, V. The ventricle, V, is the widest and longest division of the heart ; it describes a somewhat complex curve TRANSFORMATION OF THE HEART. 53S' from left to right, and is then continued headward on the right side of the embryo by a very narrow division, the fretum Halleri, Fr, which leads into the somewhat wider and curving bulbus aortse. In a slightly older stage, Fig. 286, the lateral outgrowths of the auricular division have appeared, Vh, and are the anlages of the true auric- ular cavities ; the two limbs of the ventricle are now nearer together and where they join have a distinct apex, which, owing to the in- creased bending of the heart, lies a little below (in the figure) the level of the auricles, Vh. The irregularly S-like course of the heart is very evident in this figure ; one loop of the S is constituted by the auricular division, Vh, the auricular canal, G.a, and part of the ven- tricle F; the second loop of the S is constituted by the whole of the ventricle, by the fretum Halleri, Fr, and the -bulbus, A.h. As the auricular division comprises only about a sixth of the length of the heart tube and consequently only about a third of one loop of the S, we cannot say that the heart consists of a venous and an arterial C. a. Fig. 286. — Reconstructed Side View of the Endo- thelial Heart of a Human Embryo of 4.8 mm. (His' Embryo Lr.). C. a. Auricular canal ; A.b, bulbus aor- ta3 ; Fr, fretum Halleri ; F, ventricle ; P, wall of pericardium; u.u, vena_ umbilicalis; v.o.in, vena omphalo - mesaraica ; V.cd, cardinal vein; F.j, .iugular vein ; F7i, auricle. X 40 diams. After W. His. All '«f^^^ '"'ti-'-i^ Fig. 38". —Model of the Muscular Heart of a Rabbit Embryo of nine to nine and one-half Days, seen from the left side. C, Auricular canal; B.a, bulbus aortse; A^, first, A'', second aortic arch; D.A, dorsal aorta ; D. C, ductus Cuvieri ; v. iim, vena umbilicalis: v.om, vena omphalo- me.saraica; Aa, auricle; F, ventricle. X 60 diams. After G. Born. loop. This mode of description has unfortunately been often used, and has led to much unnecessary confusion. The heart of the rabbit agrees closely with that of man. Fig. 287 is a side view of the rabbit's heart at nine to nine and one-half days, with the first aortic arch, A\ fully developed, and the second. A', just forming. The model shows especially well the union of the three venous trunks of each side in the large median sinus reuniens, which opens into the auricle; later the sinus merges with the right auricle and so disappears as a separate division. It will be noticed upon comparison of Figs. 288 and 287 that the muscular heart shows the division of the cardiac tube far less distinctly than does the endothelial heart, nevertheless the auricles, Au, auricular canal, C, and ventricle, V, are perfectly distinguishable in the model of the muscular heart. Fig. 287. The most conspicuous of the changes which now follow are, first, the descent of the ventricle and second 524 THE PCETUS. the enlargement of the two diverticula of the auricles. Both changes are well illustrated by the heart of a rabbit at twelve and one-half days — ^see Born, I. c, Fig. 19. Comparison of this figure with the preceding renders evident that the ventricle has descended so as to lie below, i.e, farther tailward, nevertheless the arterial exit of the heart tube (or the transit to the aorta) lies, as before, above or head- ward of the auricles, so that the descent of the ventricle has depended upon or been accompanied by — we may express it either way — the lengthening of the bulbus aortae, B.a. Fig. 288 represents the endo- thelial heart of a human embryo at about the stage we are now con- sidering, and illustrates how the auricles, Ho, enlarge on each side and embrace the bulbus aortse between them, and also how between the two sides of the ventricle the heart tube is somewhat constricted, forming, as it were, a narrow passage. Into the space left by this Fig. 888.— Endothelial Heart of a Human Embryo of 5 mm. (His' Embryo E). S.v, Si- nus venosus ; Ho^ auricle; C.a^ auricular canal; F.Z, ventricle; p, passage between the two sides of the ventricle; C.s^ conus arterio- sus. X 40 diams. After W. His. Fig. 289. —Inner Surface of the Heart of a Human Embryo of 10 mm. (His' Embryo Pr). V.cs, Vena cava superior; S.sp, septum spu- rium of His ; Ss, septum superius ; L,Au, auri- cle of left side; Ave, auricular canal; L.V, ventricle; S.i, septum inferlus, V.E, Eusta- chian valve ; Ai, area Interposita. x 34 diams. After W. His. constriction grows tissue from the wall of the muscular heart, which tissue gives rise to the septum inferius, that plays the chief part in the ultimate division of the right from the left ventricle. We turn now to the consideration of the interior of the heart at a slightly more advanced stage, in which the muscular and endothelial hearts are closely conjoined, owing chiefly to the growth of the muscu- lar heart having obliterated the space between it and the endothelium. Fig. 289 exhibits a view of the inside of the heart of a human em- bryo of ten millimetres. The two auricles have expanded so as to meet above, leaving, however, a partition, which is known as the septum superius, S.s; between the two sides of the ventricle is another partly developed partition, produced as just described and known as the septum inferius ; in the auricular canal there is also a projecting cushion, which in conjunction with its fellow tends to divide the right from the left side of the auricular canal. We TRANSFORMATION OF THE HEART. 525 thus encounter at three points the commencements of the ultimate division of the heart into right and left sides. The opening of the venous sinus reunions is no longer in, the median line, but upon the right side of the heart, or in other words, into the right auricle. The opening itself is bordered by two thin folds or rudimentary- valves, of which the fateral one, V.E, is the anlage of both the valvula Eustachii and the valvula Thebesii, while the medial fold ultimately disappears ; as it exists in the embryo it has been named by His the valvula vestihuli sinistra. Above the venous orifice is a small septum, S.sp, which disappears early in foetal life and is therefore known as the septum spurium. The septum spurium may be regarded as the prolongation upward of the united right and left venous valves. The space between the spurious and the superior septa is named by Born the spatium interseptale ; it is indicated for a time by a bulge upon the exterior; it merges with the general cavity of the right auricle, when the spurious septum disappears. Below the valvula sinistra, and between it and the septum superius, is the spini vestihuli, Ai, of His; it is identical with the so-called area interposita of earlier stages. The area interposita is com- posed of connective tissue and contains no muscle fibres; it is wedge- shaped, and as seen in the interior of the heart. Fig. 289, presents a triangular outline. It belongs, strictly speaking, to the septum transversum, and corresponds to part of the area by which the venous end of the heart is permanently attached to the septum trans- versum or diaphragm. The septum superius or interauricular par- tition extends on to the area interposita, and there fades out. The Primitive Ventricle. — The ventricle is at first simply a bent tube ; it may therefore be described as consisting of two limbs, which pass into one another at the apex of the ventricle. Figs. 286 and 288. The connection between the two limbs is originally very narrow, but it early widens out so much that the two limbs may be said to fuse into one general cavity. This may be called the stage of the primitive ventricle, since it is characteristic of the ichthyopsida. While the two limbs are fusing, the junction with the ventricle of the aorta (fretum Halleri, as the narrow part of the aorta is called) moves toward the median line and takes up its per- manent position just in front of the auricular canal. The change in position of the beginning or ventricular end of the aorta allows the aorta, Fig. 288, C.s, to take a nearly straight course between the auricles. The apex of the primitive ventricle is rounded, and it is not until some time after the heart is completely divided that, the ventricle assumes the adult pointed shape. Changes in the Walls of the Heart. — We consider here, 1, the histogenesis of the heart; 2, thickness of the walls; 3, the special connective tissue or non-muscular areas. 1. Histogenesis. — For what little is known concerning the devel- opment of muscle fibres see p. 478. The heart consists originally of the endothelial tube and the outer muscular tube. The endothelium, Fig. 290, endo, retains its primitive character as a thin layer lining the cavity of the heart, but the exact appearance of the cells at suc- cessive stages has still to be observed ; so far as known the endo- thelium does not give off any cells to fill up the space between it and 526 THE FCETUS. the muscular heart. As soon as the inner surface of the ventricle becomes irregular, Fig. 290, Ven, we find the endothelium close against the muscular wall and following it exactly. The muscular heart or outer heart tube produces the pericardial covering (mesothelium) of the heart, the muscle fibres, and the con- nective tissue ; Fig. 390 illustrates the general course of these modi- fications. The muscular heart tube begins to thicken, and throws off a certain number of cells which assume a mesenchymal charac- ter, and stretch across the space between the outer and inner heart, mibl»[ Fig. ^00. — Section of the Heart and Pericardial Cavity of a Rabbit Embryo of ten and one- half Days. Ec, Ectoderm; P, pericardial cavity; jilsth, mesothelium; B.a, bulbus aortse; endo, endothelium; Fe?i, ventricle; Ao^ aorta. as is shown in the bulbus aortae, Ba, of Fig. 290. Gradually the number of these cells increases until the entire space is occupied and the muscular heart is a compact wall reaching to the endothelium. During its growth we see the muscular heart acquire a more and more clearly differentiated external layer of endothelioid cells; as indicated in the figure the layer is particularly distinct in the ten and a half days' rabbit over the ventricle. The remaining cells be- come for the most part muscle fibres, but others retain the mesen- chymal character and give rise to the connective tissue, and perhaps TRANSFORMATION OF THE HEART. 537 also to the blood-vessels of the heart — when the blood-vessels first appear in the cardiac walls I do not know. 2. Thickness of the Walls. — From what has been said it is evident that the thickness of the walls depends upon the growth of the muscular heart, which takes place so that each division of the bent heart has its characteristic thickness of wall. In the auricles the walls never become very thick, and are always of about the same diameter throughout, excepting where the veins enter and the heart is attached to the septum transversum. In the auricular canal the walls become considerably thicker than in the auricles, and much less thick than in the ventricles, where the walls are most developed, and form many irregular projections into the interior of the heart so that the tissue assumes a spongy appearance. Fig. 290, Ven, which early becomes one of the most marked characteristics of the ventri- cles. G. A. Gibson, 91.1, found that during foetal life the walls of both ventricles are approximately equal in thickness, therefore the thickness of the adult left ventricle is acquired after birth. In the bulbus aortae the walls become a little thicker than in the auricles. 3. Non-Muscular Areas of the Heart. — There are several spots where connective tissue is developed to the complete or partial exclusion of the muscle-fibres of the outer heart. These spots have great importance in the differentiation of the heart. They are : the area interposita ; the thickened edge of the septum superius ; the bolsters of the auricular canal ; and the ridges in the bulbus aortse. Sinus Venosus. — A venous sinus, more or less distinct from the auricles and formed by the union of the large veins entering the heart, is found temporarily in mammalian embryos, and represents the adult condition of reptiles. At first the sinus, for which His uses the name sinus reiniiens, is symmetrically placed in the septum transversum at the venous end of the heart. As soon as the heart has become bent and the descent of the ventricles has occurred, the sinus necessarily lies on the dorsal side of the auricular division of the heart and appears partly free from the septum transversum as a short piece. Fig. 287, between the septum and the auricle. The sinus is long in the transverse direction, narrow in the longi- tudinal and dorso-ventral direction of the embryo. But as the lat- eral outgrowths forming the auricles are developed, the lateral ends of the sinus are bent headward, so that it becomes somewhat horse- shoe-shaped — the convexity being toward the apex of the heart. At the same time the sinus grows much less rapidly than the auricles ; thus it becomes proportionately smaller in later stages — in a rabbit of fourteen days its length is equal to only half the width of the auricles. Into the ends of the sinus open the ducts of Cuvier, Fig. 287, D.C., and on each side but nearer the median line the omphalo- mesaraic and umbilical veins (rabbits of eleven days). The two mesaraic and umbilical openings are, however, soon replaced by a single vein, the vena cava inferior, which opens on the right side of the sinus. Fig. 287. The cava inferior is present in rabbits of twelve and one-half days, and its development is described later in this chapter. By the time the cava inferior is developed, the sinus is no longer found opening into the heart in the median line, but upon the right side, Fig. 387 ; this change Born attributes to the manner 528 THE FCETUS. in which the partial separation of the sinus from the septum trans- versum is effected; the furrow or groove, which produces the sepa- ration, cutting in deeper on the left than on the right side, thus forcing the veins from the left side over to the right. The actual opening of the sinus into the right auricle is elongated and oblique, as shown in Fig. 289, and is bordered by two valves, which unite at the upper end of the heart and are continued as the septum spurium. The history of the valves is given below, p. 533. The sinus as a whole bulges somewhat into the interior of the auricle. The stage to which we have now traced the sinus venosus is found in the human embryo of 10 mm. In the course of its further development the mammalian sinus merges into the right auricle and entirely disappears as a distinct division. The modification is accomplished very gradually, by the expansion of the right auricle backward and downward ;* it thus embraces the whole of the right horn of the sinus, converting the right horn into a part of the auricular cavity, and the dorsal or posterior wall of the horn into an integral part of the auricular wall, consequently the valves of the venous opening appear to spring from the posterior wall of the heart. The three permanent body veins open as before with a common oblique mouth ; compare Fig. 389. The upper end of this orifice corresponds to the opening of the vena cava superior dextra, the lower end to the opening of the vena cava inferior, the middle to the opening of vena cava superior sinistra. The sinus is found almost completely merged in the auricle in rabbit embryos of about twenty days, and its limits can be traced in considerably older stages, and according to His even in the adult human heart. The left horn of the sinus remains outside the auricle and becomes the coronary sinus of the adult. Division into Right and Left Hearts. — The developmental conditions which result in the complete division of the heart are established by the primitive bending of the heart and the outgrowth of the auricles; the former initiates the division of the ventricle, the latter the final separation of the two auricles. The division is supplemented by that of the auricular canal and aorta. Accordingly we may consider the division of the heart under four heads : division of, 1, the auricles; 2, the auricular canals; 3, the ventricle; 4, the aorta. 1. Division of the Auricles. — The histories of the process: given by His and Born, differ in several essential points. I follow the latter authority, I. c, pp. 308-312, as giving the presumably correct account. When the two auricles grow forth, they expand upward, but there remains between them a partition, Fig. 391, to which His applies the name of septum siiperins, Born the name septum primum. As the auricles continue to expand, the septum of course increases by the continued meeting of the auricles, and it also in- creases, without doubt, by its own growth. The septum early acquires a very characteristic appearance by the thickening of its lower edge, Fig. 393, just above the auricular canal; the thin part of the partition contains muscle fibres, but the thickened edge * I follow Born, 89. 1, 326, but not intelligently, iSr I have been unable to understand fully his account of the merging of the sinus in the auricles. TRANSFORMATION OF THE HEART. 529 Peri.- 7 t^^'^, — Left- Fig. 291. — Section in the Frontal Plane through the Heart of a Eahbit Embryo of thirteen Days, c.au, Canalis auricularis; s.sp, septum spurium; ss, septum superius; B.V,^ right Ten- tricle ; Peri, pericardial cavity. X 117 diams. consists of embryonic connective tissue; the septum is, of course, covered with endothelium. Seen from the side, the edge of the sep- tum presents a curved outline, being concave toward the ventricle. The only connection between the auricle is Y ..=- ^ . ._,, now under the edge of the septum. This communication has been homologized with the foramen ovale of the foetal heart in later stages. Born has shown that this homology, which was maintained by his predecessors, is incorrect, and that the septum grows down to the auricular canal, C, and by unit- ing with the partition developed in the canal closes permanently the primary commu- nication {ostium pri- mum of Bom). The true foramen ovale is developed as a perforation of the upper part of the septum superius. This perforation is termed by Born the ostium secundum, I. c, p. 311. It appears in rabbit embryos of 14 mm. (about fifteen days) ; it is small at first and situated close to the wall of the auricle ; as to how it is developed, Born gives no in- formation. It soon enlarges, and in rabbits of 7.3 mm. or nearly thirteen days is about the same size as the earlier communication {ostium primum), which; from this stage on, gradually contracts until in rabbits of 10-13 mm. it has closed. In rabbits of ten milli- metres (thirteen and one-fourth days) a new septum appears above trie foramen ovale ; it is crescentic in shape, and belongs to the right auricle, since it springs a little to the right of the insertion of the septum superius. This new partition {septum secundum) was first recognized by Born, and can be followed a little way alongside the septum superius: it is also distinguished by being thicker than the septum superius; its edge forms part {limbus Vieusenii) of the boundary of the foramen ovale. The foramen ovale remains open during foetal life, and in man is not completely closed until some time after birth. On the posterior wall of the auricle the septum superius runs on to the area interposita of His, see p. 535, and can for part of its extent be regarded as an upgrowth of that area. Born, in opposition to His, attributes little special importance to this relation. The closure of the primary communication between the auricles is better de- scribed in connection with the division of the auricular canal. 2. Division of the Auricular Canal. — The auricular canal 84 530 THE FCETUS. in human embryos of 8 mm. is found, as it were, invaginated into the ventricle, Fig. 292, c, c'. There appear also two prominences of connective tissue, one on the posterior, one on the anterior wall of the canal. These prominences are the Endothelkissen of F. F. Schmidt,70. 1, the Endocardkissen of Bom, I.e., 330. They increase in height until they meet and unite (rabbit embryos of about thirteen days) so as to divide the passage of the auricular canal into two channels, c and c'. The prom- inences are wide; hence, when they meet, the greater part of the canal is closed and the channels are relatively small. Each channel maintains the direct connection between the auricle and ventricle of its own side, and is triangular in sec- tion. The triangular section is a necessary consequence of the mode of formation ; each prom- inence forms a side, and the original wall of the canal makes the third side . While the prom- inences are joining one another, the edge of the septum superius also unites with them, so that, except for the open foramen ovale, both auricles and the auricular canal are divided be- fore the ventricles. His has proposed so designate the septum thus formed by the term septum intermedium, but a special new term seems to me superfluous. As shown in Fig. 392, the septum consists of a thinner part between the auricles and a thicker part in the auricular canal. His attrib- utes considerable importance to the area interposita, as contributing to unite the septum superius with the prominences of the auricul^ canal; compare p. 535. The auricular canal soon ceases to be a recognizably distinct part of the heart, and is represented only by the openings between the auricles and ventricles {ostia atrio-ventriculares) , and by the atrio- ventricular valves. 3. Division op the Ventricles. — The two limbs of the ventricle are, it wiU be remembered, at first entirely distinct, Fig. 386, and even after the ventricle has grown considerably and the connection b)etween the two limbs has widened so much that they form essen- tially one continuous cavity, the original division between the left limb and the right limb is marked by a groove on the external sur- face. This groove corresponds to a fold of the cardiac wall, and hence is represented in the interior of the heart by a projection which grows, as development proceeds, although the external groove is gradually obliterated. The growth of the projection establishes the partition between the ventricles, which is known as the septum Sao' Pig. 292.— Oblique Section of the Heart of a Human Embryo of 8.5 mm. (His' Embryo I.). CE, (Esopliagus; Lg^ lung; svs, sinus venosus sin- ister; L.Au, left auricle; B.Au^ right auricle; c, c', channels of auricular canal ; a;, connective tissue; Sp.i^ septum inferius; Sao^ anlages of aortic septum ; V,E, Eustachian valve; svd, sinus venosus dexter. X 24 diams. After W. His. TRANSFORMATION OF THE HEAR'i. 631 inferius, Fig. 292, Sp.i. This septum is thick, and consists chiefly of muscle fibres ; it has a partially trabecular structure, and certain of its trabeculae are ultimately transformed into chordae of the atrio- ventricular valves. In a side view the upper edge of the septum is seen to be curving, the septum as a whole being crescent-shaped ; it is situated somewhat to the right side of the median line. Fig. 291. After it is fully developed (rabbit embryos of 10 mm.) the septum reaches nearly to the auricular canal and if it were prolonged it would join the right-hand side of the partition in the auricular canal ; on the posterior side of the heart the septum does actually join the auricular canal, but on the anterior side it fades out toward the aorta. In brief, the broad communication between the two ventricles becomes an interventricular foramen bounded by the partition of the auricular canal and by the edge of the septum inferius; it repeats for the ventricles the role of the foramen ovale for the auricles, p. 529, but were it to close over, as does the foramen ovale, the left ventricle would have no exit, because, as already described (compare Fig. 288) , the aorta is the prolongation of the right limb of the ventri- cle. In order to furnish the necessary outlet the aorta is divided into two vessels, and one of these (aorta vera) becomes connected through the interventricular foramen exclusively with the left ven- tricle, thereby rendering the separation of the ventricles complete. Accordingly, to fully understand this separation we must follow the history of the division of the aorta. 4. Division of the Aorta. — The cardiac aorta comprises the fretum Halleri and bulbus aortse, which at an early stage differ in the width of their cavities. Fig. 286. This difference is soon lost, and the cavity (endothelial aorta) becomes flattened except in the truncus aortse or upper part of the bulbus, where the cylindrical form is retained. The plane of the flattened cavity changes ; it is sagittal where the aorta arises from the conus arteriosus of the right ventri- cle, and as we ascend along the aorta we find the anterior edge of the cavity moving toward the left until the plane of the flattened cavity becomes transverse. Meanwhile the muscular wall of the aortic heart has developed, partly into muscle, partly into connective tjssue, and this connective tissue develops into a ridge on each side of the flattened cavity. The ridges increase and unite, thus dividing the aorta into two channels, the anterior or left channel becoming Fig. 293.— Sections at Different Levels through the Cardiac Aorta of a Human Embryo of 11.5 mm. (His' EmbiTO Eg). The lowest section is on the left, the highest on the right, a, Aorta; P, pulmonary division. X 15 diams. After W. His. that of the pulmonary artery, the posterior or right channel becoming the permanent true aortic cavity. The union of the ridges begins just where the aorta divides to form the aortic arches, and the parti- tion at this point is sagittal, cf. Fig. 293. The formation of the partition progresses downward toward the ventricle, the plane of 533 THE PCETUS. the partition gradually changing to transverse. The two ridges are found to extend into the ventricle, and participate in the closure of the interventricular foramen, by developing an oblique partition which grows down to the edge of the septum inferius, and thus con- verts the interventricular foramen into the orifice of the true aorta. The blood in leaving the left ventricle must now pass through the foramen, then across a space which originally belonged to the right ventricle, but which has been shut off by the down-growth of the septum aorticum. The ventricular extension of the aortic partition is effected chiefly by the left or anterior ridge, the right or posterior ridge passing out more on to the lateral wall of the ventricle where it fades out; the left ridge (rabbit embryos of 14 mm.) runs onto the edge of the septum inferius. The division of the aorta and ventricle is completed in rabbit embryos of about sixteen days. At the upper end of the aorta the partition extends so that the fifth aortic arches are connected only with the pulmonary aorta, while the remaining arches are connected with the true aorta only. Soon after the internal partition is formed, the external division commences as two grooves on the outside of the aorta, beginning just between the fourth and fifth aortic arches. The two grooves extend to the ventricle and gradually deepen, until the aorta is com- pletely divided into two vessels (Born, I.e., 337), which have, as soon as they are separated, both their connections with the heart and their relative positions to one another essentially as in the adult. The heart is now completely divided. Valves of the Heart. — The entrances of the pulmonary veins have no valves ; the entrances of the body veins have two valves in the embryo, of which the left disappears and the right persists as the Eustachian valve and Thebesian valve ; the right atrioventricular passage has the tricuspid, the left the bicuspid or mitral valve. The entrances of the pulmonary or right, and true or left aorta are each guarded by three semilunar valves. As is well known, all these valves are set so as to favor the flow of blood toward the arteries and prevent its flow toward the veins. We shall consider, 1, the venous valves : 3, the atrioventricular valves : 3, the aortic valves. 1. The Venous Valves.— In a human embryo of 10 mm., the opening of the body veins or sinus venosus into the right auricle is guarded, as shown in Fig. 389, by two valves or thin flaps of the heart walls; at the upper side of the oblique opening the two valves unite and are continued as the septum spurium, S.spj the left valve lies near the septum superius and merges into the area interposita; the right valve is from the start much larger than the left, and de- velops into the valvula Eustachii and valvula Thebesii. The venous valves owe their origin to the sinus venosus being pushed into the right auricle and in consequence forming a fold which projects around the venous orifice into the cavity of the heart. The edge of this fold grows considerably and becomes the anlage of the venous valves. The left valve gradually disappears — probably completely or nearly so. But His thought it contributed to form part of the edge of the foramen ovale. Born's later observations, I.e., 331, suggest rather that it never unites with the septum superius (inter-auricular parti- tion) but simply aborts, and for a time (embryos of the fourth month) TRANSFORMATION OF THE HEART. 533 can be recognized as a slight ridge on the posterior wall of the auricle. The right valve, which is always larger than the left, persists in greater part. Early in its development it begins to grow unequally, so that there is a larger upper flap bounding the main venous open- ings, and a smaller lower flap bounding the mouth of the coronary vein; the two flaps are, of course, continuous with one another though separated by a notch ; the upper flap is the anlage of the Eustachian, the lower of the Thebesian valve. The Eustachian valve does not include the whole upper division of the primitive valve, for the uppermost part aborts, though it can still be traced in the human embryo of four months and even at seven months (Born, p. 332). The septum spurium is to be regarded as the prolongation of the united right and left venous valves. As it contains muscular fibres, its probable function is, as suggested by Born, to draw the two valves together and prevent the back flow of the blood, a func- tion of great importance in the embryonic heart before the atrioven- tricular valves are developed. In a human embryo of 34 mm. (beginning of the third month) , the septum is so much reduced that it would not be recognized without knowledge of the preceding stages, and at this time we find the tricuspid and mitral valves in action. 2. The Atrioventricular Valves. — Their development has been studied by Bernays, 76.1, whose results have been confirmed by Born, 89.1, 340. W. His' observations (" Anat. menschl. Em- bryonen," Heft III., 15 •2-160) also are important. The valves proper — in distinction to the muscles and tendons, which belong to the ventricle — are to be regarded as morphologically modifications of the walls of the auricular canal, the canal being to a certain extent invaginated into the ventricles (W. His, I.e., Fig. 105). Theinvag- inated portions of the canal become the anlages of the atrioventricular valves, on the left side the mitral, and on the right the tricus- pid. When the auricular canal divides into the two atrioventricular channels, each channel or ostium is triangular in section, and as this form is preserved on the right side of the heart, there are three valves developed, one as the prolongation of each of the three walls of the ostium, but on the left side, in consequence of as yet unde- termined conditions, there are developed only two, the mitral valves. In each case the lateral valves are developed from a fold of the heart wall, which, as indicated at x in Fig. 292, is formed partly by the wall of the auricular canal, partly by the wall of the ventricle, and partly by connective tissue in the interior of the fold. The medial valves— one on the left side, two on the right— may be described as prolongations of the septum intermedium. Fig. 292. The mus- cular trabeculae of the ventricle are, almost from the start, con- nected with the ventricular surfaces of the atrioventricular valves ; out of these trabeculse are developed the chords of the valves, known in the adult as the papillary muscles and chordae tendinese. The trabeculse are originally very irregular in their arrangement, but as development progresses those which are connected with the valves become longer and slenderer, and descend in main lines directly from 634 THE FCETUS. the valves to the ventricular walls, but preserve the network charac- ter. A little later (pig and calf embryos of 45-60 mm.) the valvular trabeculae become very slender though still muscular, in the neigh- borhood of the valves, but toward the apex of the heart the fine trabeculae unite into plump bundles, the papillary muscles (Bernays, 76. 1, 495. At this stage each papillary muscle breaks up into some six or eight muscular cords, which are inserted into the valves. In older embryos (in man during the fifth month) the muscular cords change into tendinous chords ; the muscular tissue in them disap- pears and is replaced by mesenchyma, which becomes fibrillar ; hence each papillary muscle is connected by several filamentous tendons with its valve. The slender tendons are the chordae tendineae of the human anatomy. 3. The Aortic or Semilunar Valves. — Before the bulbus aortas completely divides into the true and the pulmonary aortae, there appear four small protuberances at its ventricular orifice. Each protuberance is a mass of connective tissue covered by endo- thelium ; two of them are merely the ends of the ridges, described p. 531, by which the aorta is divided. When the division is completed, the ends of the two ridges are also divided, making four protuber- ances, or in all six — three for each aortic trunk. These protuber- ances are the anlages of the semilunar valves, and may be seen in a human embryo of seven weeks. They grow until they meet so as to close the aortic entrances, and assume the adult form by becoming concave. Their exact history has still to be worked out ; compare Tonge, 70. 1, 387, on the semilunar valves of the embryo chick. II. The Arterial System. We left the arterial system consisting of the cardiac aorta, the five aortic arches, and four carotids, the dorsal aorta, vitelline or om- phalo-mesaraic arteries, and allantoic arteries (see p. 274-376) and have now to trace the changes which result in the adult system of arteries — changes which are very numerous. Aortic Arches. — The general scheme of the metamorphosis of the great arteries of the five gill-arches is indicated by the diagrams. Fig. 294. A is the primitive condition: The wide pharynx, Ph, is shaded to suggest its rounded form ; the four gill-clefts of the left side, are also indicated, 1, 2, 3, 4. From the heart, Ht, runs out the aorta, which soon forks; each fork gives off five branches, I, II, III, IV, V, one in front of each cleft and a fifth behind the fourth cleft. On the dorsal side the five arches unite into a common trunk, which joins the corresponding trunk from the opposite side to form the median dorsal aorta, Ao. Now, as the clefts develop from in front backward, so the first branchial arch arises first, the second next, and so on, until the series is completed;' shortly after each arch is formed the aortic vessel appears in it. The disposition in the human embryo corresponds entirely to the diagram, for the relations are all the same, although, owing to the rolling up of the embryo, the primitive topography is disturbed : thus in Fig. 300, we at once recognize the four clefts and the five arches. The homologies of this complicated aortic system with that of the THE ARTEKIAL SYSTEM. 535 adult mammal are shown in the diagram Fig. 394, B. The shaded parts are preserved in the adult; the others disappear. The parts lost are the first and second arches ; the dorsal connection between the third and fourth left arches ; the upper part of the left fifth arch ; there disappear on the right side the upper part of the fourth and the whole of the fifth arch, and also the dorsal connection of the arches with the median dorsal aorta, Ao. There remain parts as follows: 1. The heart aorta, which by an internal septum is divided into two aortas (p. 531), one of which maintains a communication with the right ventricle and is continuous headward with the fifth arch of the left side ; from the middle of this arch springs a vessel which soon forks to make the two pulmonary arteries, P; during A Fig. 294. — ^A, Diagram of Pharynx of an Amniote Vertebrate. 1, 2, 3, 4, Gill pouches (clefts) of the pharynx, Ph; Oe, oesophagus; I, II, III, IV, V, aortic arches springing from the fork of the aorta or the heart, Ht; on the dorsal side the five arches again unite into a single trunk, which joins its opposite fellow to form the median dorsal aorta, Ao : Jkf, invagination of the ec- toderm to form the mouth; Ex.c^ external carotid springing from the ventral side of the first aortic arch; J?i.c, Internal carotid, springing from the dorsal side of the first aortic arch : om, omphalo-mesaraic veins emptying into the heart. The arrows indicate the direction of the blood-currents. B, diagram of gill-arches as preserved in mammalia; the shaded portions are those retained, the unshaded vessels are lost; da, ductus arteriosus; P, pulmonary artery. The other letters are the same as above. foetal life the upper part of this arch, da, persists as the well-known ductus arteriosus, so that there is a direct communication between the pulmonary and the body aorta. Soon after birth the lumen of the ductus is obliterated. 2. The left fourth arch, which is very much enlarged, to constitute the permanent aortic arch ; as shown in the diagram, the obliteration of parts is such that the left fourth arch is the only permanent channel of communication between the heart and the dorsal aorta, Ao; hence the aorta of the adult springs from the heart, and gives off to the right a branch, then makes itself a great arch on the left side up to the back, where it is continued down, i.e. tailward. 3. The third arches on both sides, appearing, as the figure clearly shows they must, as portions of the internal carotid, In.c; the ventral stem between the third and fourth arches is the common carotid of the adult on each side, while the continua- 536 THE FCETUS. tion of that stem headward becomes part of the external carotid. 4. The right fork of the aorta becomes the arteria innommata, a; part of the right fourth arch remains as the right subclavian artery, bj the corresponding left subclavian being given off from the corre- sponding left arch, that is to say, by the great arch of the aorta. The ventral stem between the right third and fourth arches becomes the common carotid of the right side. But, though the connections and metamorphoses of the aortic arches are sufficiently illustrated by Fig. 394 to elucidate the homol- ogies, yet the actual course of the arches is somewhat diflEerent, Fig. 295, the branching taking place as described p. 306. The cardiac aorta at first opens under the pharynx between the bases of the man- dibular and hyoid arches, but by the time the five aortic arches are developed it has moved tailward; finally when during the second Fig. 395. —Anterior Wall of the Pharynx of a Human Embryo of 8.2 mm. length. 1 to 4, Fio. 896. — Aortic System of His' Embryo Gill-pouches ; the ectodermal pouches are sep- Bl, 4. 25 mm. I-V, Aortic arches ; Ufc, mandi- arat^ by thin walls from the entodermal ; the ble; Sd, thyroid gland; K, main aorta: P, fill-arches show the aortic arches drawn in pulmonary artery ; igr, lung; Oe, oesophagus; otted lines' and arising from the heart aorta, X 36 diams. After W. His. Ao; M, mouth; Oe, oesophagus; Coe, body cavity. X 50 diams. After His. month the head is bent back or raised, compare Chapter XVIII., Figs. 323, 326, and the front of the neck elongates, the heart remains on the level with the thorax, and the position of the aorta is relatively lowered. The five aortic arches are found in human embryos of 3.6-3.3 mm., and all persist for a short time, but as soon as the neck bend begins to develop (embryos of 4 mm.) the disappearance of the first aortic arch occurs, Fig. 296, to be very soon followed by the dis- appearance of the third arch, but the dorsal part of these arches per- sists, as already explained, as the internal carotid, while the ventral part persists as the stem of the external carotid, which gives off in the region of the hyoid arch a branch, and in the region of the man- dibular arch a second branch. The branches are designated by His("Anat. menschl. Embryonen," Heft III., 187) as the arteria lingualis and arteria maxillaris communis respectively. The ar- rangement with three arches open, the first and second closed, is shown in Fig. 398. As both the third arches and the left fourth per- THE ARTERIAL SYSTEM. 537 Fig. 297. — Aortic System of His' Embryo Si, 12. 5 mm. , seen from the front. F, Vertebral artery; Ao^ aorta; 5d, thyroid id ; " gland Cc, carotis communis; sist, we have next to consider the modification of the right fourth arch {^orfo descendens dextra) , which in embryos of 3 mm. , and even less, is smaller in diameter than the corresponding left arch, a differ- ence which His is inclined to attribute to the oblique insertion of the cardiac aorta rendering the left arch the more direct continuation of the cardiac aorta. Curiously enough the difference is lost tempora- rily (embryos of 7-10 mm.), but becomes very marked again in those of 11-12 mm., Fig. 297, so that it now is hardly more than a branch of the aorta, supplying the carotid and vertebral arteries, v, of the right side. In an embryo of 13.8 mm. the right fifth arch has disappeared, and with it the piece connecting it with aorta descendens dextra. The disposition of the main stems persists at this stage, with little change except in their diameters until after birth. The cardiac aorta {aorta ascend-ens) divides into, 1, the smaller left arch {arteria OMonyina) which is continued as the arteria subclavia and gives off as a branch the stem leading to the first, second, and third arches of the right side ; this stem is the right carotis commu- nis; and into, 2, the larger left arch, arcus aortce, which is homologous with the anony- ma and like it gives off the carotis communis and subclavia of its side, and is then contin- ued on to the permanent dorsal aorta. The connection of the ano- nyma (right fourth arch) with the dorsal aorta is preserved for some time. The history of the fifth arches is given in the section .on the pul- monary arteries, p. 538. Development of the Aortic Wall. — The aortse, like all other blood-vessels, consist at first of a simple endothelium, to which are added the muscular and adventitial walls by differentiation of the surrounding mesenchyma, which begins to condense around the aortae by the end of the second month, and during the second month the separation of the mesenchymal coat into tunica media and tunica adventitia becomes apparent (see His, " Anat. menschl. Embryonen," Heft m., 198, also Morpurgo, 85.1). Erik Muller's paper, 88.1, describes, strictly speaking, not the origin of the muscular tissue of the aorta, but of the primitive mesenchyma from the inner wall of the primitive segments. Aortic Arches in Branchiate Vertebrates.— In aquatic vertebrates the aortic arches do not remain as large vessels, but they break up into smaller vessels and capillaries, which are distrib- uted through the branchial filaments, or respiratory outgrowths of the gill-arches. When this modification occurs the ventral end of each aortic arch acts as the afferent stem (branchial artery) and the dorsal end as the efferent stem (branchial vein) of the gill. It is evi- dent that the branchial veins are morphologically distinct from the true veins, and belong not to the venous, but to the arterial, system. r.», truncus pulmonalis; ,R pulmonary artery. X 24 diams. ^fter W. His. 538 THE FCETUS. Evolution of the Aortic Arches. — That there were in the early vertebrates more gill-arches than are preserved in the amniota, has been stated already. But the exact number is uncertain, and as there must have been one aortic trunk in each gill-arch the number of the aortic arches is uncertain. It seems, however, probable that there were at least nine as indicated by the structure of marsipobranchs (Julin, 87.3) and Chlamydoselachus (Howard Ayers, 89.1). Indeed, von Boas, 87. 1, has adduced weighty evidence to support his belief that even in amniota the number of aortic arches is six, a belief which Zimmerman, 89.1, has supported. To settle this question must be left to research based upon very extended comparative anatomical and embryological observations. It is improbable, as Ayers, 89. 1, has demonstrated, that the united dorsal ends of the aortic arches, which form two stems, represent the forward continuation of the median aorta, but that rather there was primitively a median dorsal aorta extending over the pharynx to the hypophysis, and that there were lateral anastomoses which have been preserved while the cephalic median aorta has disappeared. Ayer's hypothesis, which seems to me well justified, is incompatible with the current notion that the dorsal aorta represents two stems fused in the median line — a notion which has been specially advo- cated by Macalister {Jour. Anat. and Physiol., X.X.., 193, 1886). The special importance of the question at present is its bearing on the comparison of the arterial systems of vertebrates and annelids. The abortion of the aortic arches is attributed by general consent to the head-bend, and consequent cramping of the branchial region, but the factors which have caused the modification of the five partially preserved arches of mammals have still to be ascertained. Hoch- stetter, 90.1, 577, suggests that the development of a new trunk of supply — the internal mammary — for the anterior intercostal arteries may have been concerned in the abortion of the right aortic root and the changed position of the left aortic root, but leaves his thought unexplained. Internal Carotids. — As indicated in the diagram, Fig. 394, the internal carotids are developed out of the first, second, and third aortic arches ; the third arch loses its connection on the dorsal side with the fourth arch, but keeps its connection with the second and first ; there is thus a direct blood-channel from the cardiac aorta to the vessel, which runs from the dorsal end of the first arch to the head and brain. Fig. 294, In. c. The internal carotid of the adult comprises the third aortic arch, the dorsal part of the second arch, the dorsal part of the first arch, and the whole of the true internal carotid of the embryo. His ("Anat. mensch. Embryonen," Heft III., 192) states that in man the dorsal connection between the fourth and fifth arches is lost during the fifth week ; and points out that, as the heart and cardiac aorta descend, the position of the third arch becomes more and more oblique, compare Fig. 298. Pulmonary Aorta and Arteries.— In Sauropsida the fifth aortic arches are preserved on both sides, in reptiles completely, in birds partially, but in mammals the fifth arch entirely disappears on the right side and partially on the left, as established by the classic investigations of Heinrich Rathke, 57. 1. In all amniota the lungs THE ARTERIAL SYSTEM. 539 Fig. 298. —Aortic System of W. His' Embryo Hg, 11.5 mm. C7A-, Mandible; Zg^ tongue; I-V, aortic arclies; Av^ vertebral artery; P, pulmonary artery; igr, lung; oesophagus; T, truncus pulmonalis; Ao^ aorta, x Oe, 18 diams. After W. His. are supplied by arterial branches springing from the middle of the fifth aortic arch, in Sauropsida on both sides, in mammals on the left side only, Fig. 298, P. The right fifth arch disappears in man very early, but the left persists throughout foetal life. Concerning the development of the pul- monary artery proper, i.e. the branch from the arch to the lungs, Fig. 298, P, little is known. His ("Anat. mensch. Embry- onen," Heft II., 186) finds the reptilian condition — the right and left fifth arches, each producing a branch to the lungs— in an embryo of 4.2 mm. and more distinctly developed in embryos of 5-6 mm. but later both pulmonary arteries are found to spring by a common stem from the left fifth arch. How the change comes about I do not know, and I have found no explanation of it. The arteries have a special relation to the bronchi, as is explained in the section on the lungs in Chapter XXIX. Returning now to Fig. 298, it will be observed that the pulmonary artery, P, divides the fifth aortic arch into a lower part, T, connected with the heart, and an upper part, F, connected with the dorsal aorta. The lower part is the future trunk of the pulmonary aorta, and as the lungs develop the pulmonary artery increases in calibre until it equals the trunk, T, in diameter. The upper part, F, is known as the ductus arteri- osus ov ductus Botalli {BotallischerGang) audit remains through- out the foetal period as an open channel, so that blood from the right ventricle flows in part to the lungs, in part into the dorsal aorta. As stated above, the lumen of the ductus arteriosus disappears soon after birth. Dorsal Aorta and Its Branch.es. — There are many valuable observations on the foetal arteries scattered in the works of the older embryologists, in the descriptions of human embryos (Chapter XVIII.) and in articles dealing with the development of special organs, but these observations have never been collated, nor has any attempt been made, so far as I am aware, to study comprehensively the morphology of the dorsal aorta and its branches. This is the more singular as much labor as been expended upon the aortic arches and veins. An exception has been made in the case of the interseg- mental and vertebral arteries, see below. That the dorsal aorta is formed very early by the ingrowth of the omphalo-mesaraic arteries and that these arteries are the primitive branches of the aorta has been already explained. The next branches to be formed are the umbilical or allantoic, which very early acquire 540 THE FCETUS. a large size and appear as the main branches of the aorta, but the dorsal aorta is prolonged to the tail, and in tailed vertebrates persists as a permanent and considerable vessel {arteria caudalis) but in man it remains only as a small vessel, the sacra media. From the umbilical arteries, as soon as the anlages of the legs appear, arise branches, the iliac arteries, one on each side to supply the corre- sponding limbs. With the progress of development the iliacs become the main branches, and the allantoic vessels are very much reduced, becoming the relatively small hypogastric arteries of the adult. Of the omphalo-mesaraic or vitelline arteries the left aborts very early, while the right persists, and soon develops the arteria mesenterica superior as a small branch, which ultimately becomes the principal continuation of the main stem. Intersegmental Arteries. — The first branches of the aorta to appear in the embryo are a series of small vessels, which pass upward and outward on each side of the embryo. One of these vessels is to be found between every adjacent pair of myotomes, and hence they have been called the interprotovertebral arteries. In the region of the pharynx where the aorta is double, each aorta gives rise to the intersegmental arteries of its own side. Farther from the head the vessels arise in pairs from the dorsal aorta. In longitudinal horizontal (i.e. frontal) sections of the primitive seg- ments the intersegmental arteries show very well, compare Fig. 119, Is. The metamorphoses of the vessels under con- sideration have been worked out for the region of the head and neck by Froriep, 86.1 (pp, 89, 96, 103, 108, 139), and Fr. Hochstetter, 90.1, 90.3. There are six intersegmental arteries between the seven cervical segments ; of these the sixth gives rise to the arteria subclavia as a branch. There are also two segmental arteries headward of the cervical ones; these two lie respectively between the first cervical and the last occipital seg- ments, and between the last and the penultimate occipital segments. Of these eight arteries the first very early aborts, the second gives rise to a vessel which ted'etolVel^^^lSter/faltS: ™°.« forward in the head to the mid- bit Embryo at the end of the eleventh brain and there joins tho internal carotid, 5?;f-arteria°^"SrlfisT'^oT°£te: ^ig. 399. A scries of anastomes are ane?/r.'o°1r<;iavii°n*^rtely?in ^T <5eveloped between the intersegmen- IV y, aortic arches, IV showing the tal artenes of the neck and united and t™rT.'°i^ochsteS:'™"°'''^'='^- ^'- enlarged anastomosing vessels. Fig. 299, -4.V, appear as a prolongation through the neck of the vertebral artery. The intersegmental branches rapidly abort, except the sixth in the neck which persists as the stem. Fig. 299, s.cl, of the vertebral artery, and as soon as the fore limb buds out (rab- bits of eleven days) sends a branch to it, which becomes the subclavian THE VENOUS SYSTEM. 541 artery. The artery between the sixth and seventh cervical vertebrae is thus seen to acquire a special importance, as it becomes the stem of the sub-clavian and vertebral arteries of the adult. We also learn that the vertebral artery is the earlier developed, and that, therefore, the sub- clavian is morphologically a branch of the vertebral artery, instead of the vertebral being a branch of the subclavian, as usually described in human anatomy. The small original intersegmental arteries persist on the dorsal side of the vertebral artery in the neck, and supply in the adult the circulation of the vertebral column. The next following intersegmental arteries, i. e. those between the seventh cervical and first thoracic, and between the first four or five thoracic segments, undergo a similar change, a secondary longitudinal vessel being developed between them also (rabbits of thirteen days) , and as they disappear, this vessel becomes a branch — intercostalis superior of human anatomy — of the common stem of the vertebral and sub- clavian arteries. Hochstetter states, 90.1, 577, that the internal mammary arises as a branch of the subclavian at about the same time as the superior intercostal. The subclavian does not long retain its original position, but en- larges and migrates from the dorsal to the ventral side of the sympathetic ganglion chain (Hochstetter, 90.1, 578-580). The remaining intersegmental arteries of the thorax are said to give rise to the intercostal arteries. The vertebral arteries unite in the occipital region (human embryo of 10 mm. according to W. His, I.e., 193) to form the arteria basi- laris, Fig. 345, while further forward they remain distinct, resulting in the development of the circulus Willisii. Umbilical, Arteries. — These acquire a large size in the human embryo and owing to the reduction of the caudal artery {sacra media) appear as the terminal forks of the dorsal aorta. They curve around past the cloaca, run in the walls of the allantois or anlage of the bladder, to the umbilicus, and thence through the umbilical cord to the placenta. They develop each a branch, which runs to the hind limb as soon as it buds forth. Until birth the umbilical artery per- sists as the main stem, but after birth, having lost its main function, it ceases to develop and becomes the hypogastric artery of the adult. The branch to the leg (the common iliac) continues to enlarge and after birth becomes more and more the chief vessel, so that the root of the umbilical artery is converted into the beginning of the iliac artery and the hypogastric into a branch of the iliac. The precise history of these vessels has still to be worked out thoroughly. HI. The Venous System. The Primitive Veins. — By this heading I mean the jugular, cardinal, vitelline, and umbilical veins, or main venous stems of the first completed embryonic circulation. The initial arrangement of the four pairs of trunk veins can be studied in a human embryo of 4.2 mm.. Fig. 300. From the head, where it extends to the fore- brain and has several branches, comes the jugular vein, Jg, descend- ing nearly to the level of the septum transversum. From the tail comes the cardinal vein — the posterior cardinal of comparative 643 THE FCETUS. DC anatomy — to meet the jugular vein. Only part of the cardinal vein is drawn in the figure ; in reality it extends the vrhole length of the rump and ends in the tail. In a cross section the cardinal vein is seen to be situated originally in the splanchnopleure of the embryo, just at the level of the nephrotomes (or intermediate cell masses). This position being kept brings the vein, as soon as the Woffian tubules are developed, to lie just above the Wolffian body, and late- ral of the aorta, compare Figs. 301, 135, and 137. The jugular vein occupies the corresponding situation in the neck, but at the level of the segments, which in the chick shows an open connection with the splanchnoccele (S. Dexter, 91.1), crosses from the splanchno- pleure between the myotomes and splanchnoccele to the somatopleure and runs forward to the head. The jugular and cardinal veins unite forming a common trunk. Fig. 300, Z>.C— the ductus Cu- vieri — which passes in an oblique, transverse direction in the somato- pleure to the anterior edge of the septum transversum, and there bends toward the median ventral line to empty into the venous end of the heart by way of the sinus venosus. From the yolk-sac come up the two vitelline (omphalo-mesaraic) veins, one on each side, om, and from the allantois stalk pass up through the somatopleure the two al- lantoic veins, also one on each side. Fig. 300, A I. v. A cross section through the rump shows the differ- ence in situation of the cardinal vein. Fig. 301, C, in the splanchnopleure above the Wolffian body, and the umbilical vein, Uv, in the somatopleure. The umbilical vein empties into the ductus Cuvieri ; the vitelline vein into the sinus venosus. For good, fig- ures of the relation of the primitive veins to the rabbit's heart, see Born, 89.1, Taf. XX., Fig. 15. The veins, as they approach the heart, pass by the anlage of the liver, and as this organ develops it en- ters into intimate relations with the vessels, which undergo numerous modifications. It will be conve- nient to consider the changes in the hepatic veins collectively, and therefore we take up first those Fig. 300.— His' Embryo Lr (4.2 mm.). E«con- struction to show the Course of the Blood-Ves- sels. t7, Jugular vein ; Oi, otocyst ; D. C, ductus cuvieri: Am, edge of amnion; Al.v, allantoic vein ; car, internal carotid ; J, first aortic arch ; Au, auricle J Ven, ventricle; Li, liver; om, vitelline vein; Al, allantoic diverticulum; Art, allantoic artery. After W. His. Coe Fio. 301 . —Cross Sec- tion through the Hinder Part of His' Embryo E C5 mm.). My, Myotome ; C, car- dinal vein; TJv, um- bilical vein ; Coe, cce- lom. X 30 diams. After W. His. THE VENOUS SYSTEM. 543 changes in the primary veins which are not associated with the de- velopment of the liver. But to do this we must present the early history of the vena cava inferior. Vena Cava Inferior.— This is a large unpaired vessel, which is developed somewat later— in rabbits not until the twelfth day— than the four pairs of primary veins. Our present knowledge of its development rests chiefly upon F. Hochstetter's admirable investi- gations, 87.2, 88.1, 88.3. It arises as a small vessel from the ductus venosus of the liver and running through the hepatic sub- stance is continued on the right side ventrad of the aorta in the tissue between the two primitive kidneys, Fig. 302, A, ci, to a point a little beyond the aortic origin of the superior mesenteric artery. It gradually enlarges and forms two fine branches, which pass around the aorta and anastomose with the cardinal veins, the communication A :b c I^lV-'" Fig. 304— Three Diagrams to illustrate the Transformation of the Venous System. After O. Hertwig. (Explanations in the text.) being established about at the origin of the renal vein. Fig. 302, A, r. By the thirteenth day the anterior portion of the cardinal vein is nearly aborted. The lower part of the right cardinal appears now as the direct continuation of the enlarged vena cava, and in fact is the anlage of the lower part of the adult cava inferior, Fig. 303, C. By the fourteenth day the renal veins appear as branches of the cava, and the caudal ends of the two cardinals are united, thus con- verting the lower branches of both these veins into branches of the cava inferior. But in man this fusion of the cardinal veins does not take place, but instead there is developed a cross anastomosis by which the lower ramifications of the left cardinal become branches of the cava. Fig. 302. C in Fig. 302 represents diagrammatically the permanent condition. The true vena cava inferior extends only to the renal veins, r, which are persistent segmental branches of the cardinal veins; beyond this point the cava is really the persistent right cardinal vein ; a cross anastomosis, ilcs, becomes the left com- mon iliac, while the terminal branches of the cardinals are converted into the external and internal iliacs on each side, and empty their 544 THE FCETUS. blood into the right cardinal, or lower segment of the adult cava inferior. Metamorphoses of the Primitive Veins. — By a series of changes beginning very early indeed in the embryo the four pairs of symmetrically placed veins take on an asymmetrical arratigement. The chief factors of the change are, 1, the development of new cross trunks, which become main stems ; 2, the abortion of parts of the primitive veins ; 3, migration of the vessels. The changes which occur, in the venous sinus have been already indicated ; those which occur in the liver are described in a separate section below. Changes of the Ductus Cuvieri and their Connections. — We have already noticed the relations of the ductus to the horns of the sinus venosus, p. 527, and the role of the ductus in shutting off the pleural from the pericardial cavity, p. 483. The transformation of the ductus begins with a change in their position, their course be- coming steeper, in consequence of the descent of the heart, and at the same time thej' project across the opening of the pleural cavity into the pericardial cavity, and by finally closing across this opening the ductus are enabled to tmite with the medias- tinum, thus bringing the two veins nearer together. Of the veins supplying the ductus the jugulars continue to develop and with the growth of the head to acquire an increasing importance, while the cardinal veins have their circulation impeded owing to the com- petition of the vena cava inferior ; the preponderance of the jugular is further increased by the vein of the fore limb, the subclavian, Fig. 303, A. s, emptying into it. The two sides of the sinus venosus early become asymmetrical, and, owing to the migration of the sinus toward the right side of the heart, p. 537, the right ductus (the future vena cava superior dextra) has a shorter and more direct course to the heart than the left ductus, which has to bend around the left auricle toward the right. The left ductus runs along the coronary groove of the heart, and there receives the coronary vein, concerning the development of which we have no definite information. This may be called the Sauropsidan stage, since it is permanent in all reptiles and birds; but it is said to be retained in certain mammals. In man, however, a further stage is reached by the partial abortion of the left ductus {vena cava superior sinistra). The reduction begins with the development of a cross anastomosis. Fig. 303, B, as, between the two jugulars. The anastomosing vessel, which is the future vena anonyma sinistra, runs obliquely from the left to the right jugular, where the conditions for the return of blood to the heart are more favorable ; the cross vessel enlarges and in the same measure the right ductus enlarges also, with the further consequence that the right cava usurps more and more of the blood from the left jugular. This leads to the gradual closure of the left ductus Cuvieri . (cava sinistra) except of the end next the heart, which persists as the vein delivering the vena coronaria into the right auricle. We thus learn that the cardiac orifice of the coronary vein is really the mouth of the vena cava superior sinistra. The development of the valve (valvula Thebesii) of this orifice is described p. 533. The cardinal veins undergo a similar change to the jugulars, see THE VENOUS SYSTEM. 545 Fig. 302, C, in that a cross vein appears which takes the blood of the left cardinal into the right, so that the stream of both cardinals is poured into the right ductus Cuvieri (cava sup. dextra). In the account of the vena cava inferior it has been explained how the lower parts of the two cardinals are changed, and only the upper parts left. As the main function of the cardinals appears to be to maintain the circulation of the Wolffian bodies, the cardinals lose their importance as the bodies abort. They persist, however, in part to give rise to the azygos and hemiazygos veins of the adult, as sufficiently indicated by Fig. 303, C, az, hz\ hz. Veins of the Hand and Foot. — Fr. Hochstetter, 91.1, has shown that in all amniota there is a vein (Bandvene) which runs around the edge of the hand (or foot) but when the digits appear this "randvene" is divided and gradually disappears. The veins are de- veloped as a network of capillaries, connected with the randvene and the venous trunk of the limb. As the digits grow out, the rand- vene persists on each side of each digit, but is interrupted at the apex. The randvene thus gives rise to the digital veins and proba- bly also is continued on the ulnar side as the permanent vein of the arm, and correspondingly on the leg. Hepatic Veins. — The following account is an abstract of His' researches (" Anat. menschl. Embryonen," Heft III., 200-310). The liver grows out into the septum transversum and by its enlarge- ment comes very soon into con- tact with the vitelline and um- bilical veins on their way to the sinus venosus. The hepatic cy- linders grow into the veins, push- ing, however, the vascular endo- thelium before them, and dividing the veins into numerous channels, which constitute a network of fine branches. The four vessels are thus broken up into smaller vessels, but for a while they per- sist in part as larger stems lead- ing from the liver to the sinus venosus. The liver is now sup- plied with all the blood from the chorion (placenta) and the yolk- sac. This stage is found in a human embryo of 4. 35 mm., Fig. 303. The united umbilical veins of the allantoic-stalk, All, pass up to the liver in the somatopleure of each side of the body ; the left umbilical, v.tt".s,is already decid- edly larger than the right; both veins break up within or near the liver into small vessels. The two vitelline veins, Vi, run in the splanchnopleure or wall of the intestine and unite just before they attain the liver, then separate and pass 35 An. Fig. .303. —Reconstruction of a Human Embryo (His' Bl.) of 4. 2.5 mm. Front view. V.j, Jugular vein : card^ cardinal vein ; D. C, ductus Cuvieri ; Li, liver: S.r. sinus venosus; f.l, fore limb; v.u.s, upper part of left umbilical vein; v.u".s, lower part of same; All, allantoic stalk; In, intestine; vu.d, riprht umbilical vein; Vi, vitel- line vein. X 18 diams. After W. His. 546 THE FCETXTS. /V-Ar vu around the entodermal intestinal canal to unite again on its dorsal side, making a complete venous ring; they then again separate and pass back again aroxmd the intestine, forming a second complete ring before they break up into small hepatic vessels. On the right side the umbilical and vitelline trunks remain separate as they leave the liver, and open separately into the sinus venosus, but on the left side the two trunks unite, as shown in the figure, and empty by a com- mon stem into the venous sinus, S.r. In the next stage the lower part of the right umbilical has no longer any connection with the upper part of the same vessel , and, therefore, since it continues to act as a venous path, its stream is directed downward. The left umbilical vein, on the contrary, has in- creased in size, Fig. 304, V.us, and is prolonged within the liver by a large stem, which joins the left side of the upper venous ring formed by the vitelline veins, vi. The upper ring is connected by a newly developed large trunk, V.ar., the vena ascendens, or vena Aranti — as to the origin of which we possess as yet no satis- FiG. 304. — Reconstruction of the Venous factorv data. Remnants of the Trunks and Liver of His' Embryo B, 5 mm. ,. •' j: ±\. ^^■^• ^ j ■ F.p, Portal vein; «.«, right umbilical vein; pOrtlOnS OI the UmblllCal and, VI- F. ^4r, vena Arantii : F. US, vena umbilicalis sin- +QlKTizi -xToina TST-Tiir*!! in i\\cx nrci- istra; vi, vitelline veins; v.u.d, vena urn- Tenine veins, wnicn in tne pre bilicalis dextra. The vessels left white are VIOUS Stage tOok the blood from aborted. X 40 diams. After W. His. .-, i . *^ . , i the liver to the sinus venosus, still persists. It will be seen that the essential difference between this stage and the preceding is, that whereas previously all the blood passed the liver through small vessels, now only part of it flows through small vessels, the rest through large trunks directly to the heart. The third stage is established by developing the single portal vein out of the two vitelline veins. This is accomplished as indicated by the diagram. Fig. 304, which is to be compared with the previous figure. The left side of the upper ring formed by the vitelline veins, Fig. 303, vi, and the right side of the lower ring persist, leaving parts of each ring to form a single continuous vessel, the fenapor ice, which from its mode of origin necessarily makes one complete spiral turn around the intestine. Herewith the condition is reached which per- sists throughout foetal life. Fig. 305. The portal vein and left umbilical vein supply the liver with venous blood, and also form within the liver near its lower surface two large stems which unite and are continued forward by the single vena Arantii. These three great veins after the third month are found to lie near the median plane, and to follow straighter courses than in Fig. 305. The final stage is not reached until after birth, when the umbilical vein rapidly aborts. A little later the large channel formed within the liver by the vense portse and Arantii also disappears, except that the part between the union of the vena cava inferior with the Vena .ixs. THE VENOUS SYSTEM. 547 Arantii (ductus venosus) and the heart is retained and functions as the cardiac end of the adult cava inferior. In the fourth or adult stage, the liver is supplied by the portal vein, the representative of the vitelline or omphalo-mesaraic veins of the embryo, and all the por- tal blood passes through the liver in small vessels (capillaries) , though, of course, larger venous branches per- sist to distribute the blood to, and collect it from, the capillaries of the hepatic lobules. Pulmonary Veins. — It was first shown by Fr. Schmidt, 70.1, that the pulmonary veins are four vessels, which unite into a short common stem emptying into the left auricle. Their history has been further eluci- dated by His, 87.3, 103, and G. Born, 89. 1, 313, 334. The common stem appears first as a capillary ves- sel arising from the left auricle near the interauricular septum (twelve days' rabbits) ; the small vessel runs through the mesocardium posterius directly toward the anlage of the lungs; by enlarging and branching this vessel forms the system of the pulmonary veins, but for some time after its appearance it remains small. The development is not the same in the rabbit and in man; in the lat- ter the common stem enlarges and merges into the auricular cavity, at first as a recess, later without demar- cation; hence the four pulmonary veins open into the heart by two orifices, the two veins on each side uniting before they empty. Still later (two months' embryo) the four veins each open separately, more of the vein being annexed by the heart. In the rabbit the primitive condition is permanent, and the four pulmonary veins unite before joining the heart. The course of the four veins in the lungs has been described by His, 87.3, 103. They run from the central stem one to each lobe of the lung ; in other words, from the start there is an upper and a lower vein in each lung; the pulmonary veins are situated below the forking of the trachea, and this relative position the main stems retain throughout life, — compare Fig. 459. Fig, ■Reconstruction of the Venous System of His' Embryo Rg, 11.5 mm. CHAPTER XXV. THE EPIDERMAL SYSTEM. That portion of the ectoderm which remains upon the surface of the embryo is called the epidermis; it constitutes the outer skin; for convenience the inner skin (cutis or dermis) is treated in connection with the history of the true skin in this chapter. We have also to consider the development of the following epidermal appendages: nails, hairs, and glands. I. The Skin. Epidermis. — The ectoderm of all amniote vertebrates is at first a single layer of cells, which presents considerable variations in ap- pearance not only in different classes, but also at different stages of the same species, and even in different parts of the same embryo. Since in all invertebrates the ectoderm consists of a single epithelial layer, we may call the first stage of the vertebrate epidermis the invertebrate stage. The appearance of the ectoderm while in this stage has been indicated by the figures and descriptions scattered through Chapters V.-XV., and until a comprehensive study of the ectoderm of amniota in the one-layered condition shall have been made, it is impossible to give a minute description of it possessing much value or any interest. The epidermis of Amphioxus and the ectoderm of the amnion never pass beyond the one-layered stage, p. 335. In its second stage the epidermis becomes two-layered. The cells of the single layer become irregularly placed ; some have their nuclei nearer the outer, others nearer the inner, surface of the ectoderm. The difference rapidly in- creases, and though for a time the cells stretch through the whole thick- ness of the layer, yet they gradually draw away, some from the upper, others from the low( surface, until they ha definitely arranged thei selves in two distinct layers, Fig. 306. This stage is established in the human embryo by the end of the first month, and persists over part certainly of the embryo, at least until the close of the second month. Fig. 306. — Section of the Skin of a Human Embryo of sixty-three to sixty-eight Days. Minot Collection, No. 138. a, Outer layer of epidermis; 6, inner layer of epidermis; c, cutis. THE SKIN. 549 In stained sections the outer layer, Fig. 306, a, is composed of some- what flattened cells, with irrea;ularly shaped, slightly granular nu- clei, and are darker and thicker walls than the cells of the inner layer. These latter. Fig. 306, &, are larger and clearer, and have larger, more granular nuclei of round shapes. The appearance of the outer cells suggests a necrotic change. Bowen's careful researches, 88. 1, render it probable that the outer layer is the epitrichium, compare below. It is a remarkable fact that the primitive blastoderm in amphibia, teleosts, and ganoids never passes from the several-layered to the one-layered condition, but only to the two-layered condition. For description of this stage in Bombinator, see A. Goette, 75.1, and in teleosts see M'Intosh and Prince, 90.1, 739, in Lepidosteus, Balfour and Parker, 83.1. The development, therefore, in this group of forms, offers a marked difference from that found in mar- sipobranchs and amniota, but since in Petromyzon we encounter the one-layered stage, we must consider the succession of stages adopted in this chapter as the primitive one, and conclude that the precocious appearance of the two-layered stage in amphibians, etc. , is a second- ary modification, the cause of which is unknown. That the two layers of the epidermis are homologous throughout the vertebrate series, we have no reason to doubt (Balfour, "Comp. Embryol.," II., 300). Where the epidermis has an initial division into two layers, the inner is commonly termed the nervous layer, and it has the main share in forming all the organs derived from the epidermis ; the outer layer, according to homologies I hold to be probable, must be identified with amniota epitrichium, although unlike the true epitrichium it disappears as a distinct layer, its cells showing them- selves between those of the inner layer (Goette, 75.1, 158). The ectoderm of the chorion and umbilical cord never advances beyond this stage, unless we regard the formation of the chorionic cellular layer as such an advance. The third stage is very gradually reached by the increase in the number of layers until there are several. I consider it probable that this stage is established in two ways — one, the more primitive, in- volves the disappearance of a distinct outer layer, as in amphibia; the other depends upon the preservation of the outer layer, as the epitrichium. This view can be advanced, at present, only as an hypothesis. 1. The primitive method is maintained in amniota only over very limited special regions ; as such I venture to designate the cornea, the nasal pits, the mouth cavity and lips, and the anal ectoderm. Over these parts the distinct outer layer disappears as such, and we have developed a stratified epithelium, which never produces a true horny layer, but consists of a basal row of protoplasmatic cells, and several layers of cells above, which are clear in appearance and have thickened walls. The details of the process of differentiation have not yet been worked out. 2. The secondary method of forming the several-layered epider- mis is established over the skin proper. It can be well seen in the human embryo of the third month. In' an embryo of two and one- half months. Fig. 307, there are four to five layers of cells. The 550 THE FCETUS. basal layer, b, is composed of a single row of cuboidal cells, which are rich in protoplasm, though small in size, and which have round nuclei.^ This basal layer persists throughout life in all amniota, and is one of the most characteristic features of the amniote epidermis. F19. 307.— Epidermis from the Occiput of the Human Embryo of two and one-half Months. JSptr, Epitrichial layer; to, Malpighian layer; b, basal layer. After Bowen. Above the basal comes the middle layer, which varies from two to three cells in thickness ; its cells are irregular in shape and size, and are so large that the nuclei of many of them do not appear in the section. The outermost layer, Eptr, is the epitrichium, and con- sists of a single layer of large dark cells, which from their arching up may be termed dome cells. It is probable that the epitrichium is the outer layer of the second stage preserved and modified, and that all the middle cells come from the inner layer of the previous stage, but conclusive proof of this identification is still required. The his- tory of the epitrichium is treated in the next section. The fourth stage is characterized by the presence of a horny layer (stratum corneum). The stratum corneum presents marked varia- tions in structure, and it is probable that, as explained in the follow- ing paragraph, at least two morphologically distinct layers have been confused under a common name. Unfortunately almost noth- ing is known concerning the genesis of the horny layer. Bowen's observations, 89.1, render it probable that it arises from the epi- trichium, but if this view be adopted we encounter certain difScul- ties which our present knowledge cannot remove. If Bowen's hypothesis is correct, we must define the fourth stage as characterized by the cornification of the thickened epitrichium. Concerning the process of cornification we possess some information, which is re- ferred to more fully under the head of nails, p. 555. When the horny layer is produced the skin is considerably thickened and the number of layers of cells which it comprises is much increased. The line of division between the horny layer and the underlying mucous or Malpighian layer becomes quite sharp. It must be as- sumed that cells of the deep layer are added to the horny layer. The fifth stage is established by the development of the stratum lucidum. Bowen has made the important discovery that the stratum lucidum of the human embryo lies immediately underneath the epi- trichium, and is directly continuous with the nail, and the epitri- chium is continuous with the horny layer outside the stratum luci- dum. Bowen suggests, 89. 1, 449, that, where there is no epitrichial layer nor characteristic stratum lucidum (Zander's Typus B, 88.1), the stratum really extends over the Malpighian layer, being modified and constituting the horny layer of those parts. The essential char- acteristic of the stratum lucidum is that its cells are solidly cornified, their nuclei being obliterated. When the epitrichial cells cornify they acquire thickened walls, but remain hollow (Zander's Typus A, 86.1, 88.1). The histogenesis of the stratum lucidum is de- THE SKIN. 551 scribed in the section on the nails, p. 555, the process having scarcely- been studied except in connection with the investigation of the nails. The ridges {retes d'Henle) on the under or dermal side of the epidermis begin to appear on the hairless parts, according to Blaschko, 87.1, about the fourth month, but on the hairy parts, where they are always rudimentary, they do not appear until toward the end of foetal life. There are primary and secondary ridges. The former are the first developed, and from them the solid out- growths to form the sweat glands originate. Fig. 308 represents Fio. 308.— Section of the Skin of the under Side of the Eight Second Toe of four months' Em- bryo, Minot Collection, No. 123. Ep, Epidermis ; Bi, primary ridge of epidermis ; S, sweat gland ; Cu, cutis. a section across the primary ridges : the epidermis is some seven or eight cells thick, its outer surface irregular, but not thrown into folds or ridges ; the structure of the superficial layer is indistinct but the epitrichium seems to have disappeared ; the dermal surface is thrown up into regular rounded equidistant ridges, Ri, from which grow out here and there the solid anlages of sweat-glands, 8. These ridges do not arise all at the same time, but their forma- tion spreads from sundry centres, nor do the ridges run in straight lines altogether, but on the contrary in parallel curves. The ridges under the nails appear first (three and one-half months) under their distal and lateral borders, later under their central and proximal portions; additional ridges appear between those first formed (F. Curtis, 89.2, 179). In the next stage, which is assumed by the epidermis only upon the palms and soles, the outer surface forms a low ridge over each of the inner ridges. The external ridges with the openings of the sweat glands upon them are easily seen upon the adult hand. When the external ridges are developed there appear also secondary ridges on the dermal side, between the primary ridges. The secondary are much smaller than the primary ridges and under- lie the grooves separating the external ridges. 552 THE PCBTUS. The origin of epidermal pigment has been already discussed, p. 419. Epitrichium. — The external layer of the skin is known to be stratified in all amniota, but the homologies of the strata have never been satisfactorily determined. That the mucous or Malpighian layer is the same in all classes is evident, but that the horny layer comprises two distinct strata is, I think, extremely probable, as stated above. One stratum may be homologized with the stratum lucidum, the other with the epitrichium- Where the epitrichium is lost (nails and hairy skin) the stratum lucidum forms the superficial layer of the epidermis, but when the epitrichium is preserved, it forms the outer layer and the stratum lucidum underlies it. The history of the epitrichium is the key to the morphology of the am- niote epidermis. The epitrichium was discovered by Welcker, 64. 1, in the embryos of a sloth (Bradypus) , where it forms a continuous membrane over- lying the hairs. Welcker found the layer in several mammals, including man, and demonstrated that it belongs to the epidermis, becoming separated from the rest of the outer skin, when the hairs grow forth. In the sloth it forms, so to speak, an extra foetal en- velope, which we find mentioned by Eschricht and Ebsen (Miiller's Arch., 1837, 41) and and by Simon (Miiller's Arch., 1841, 370-373), but these authors did not ascertain its origin. Kerbert, 77. 1, dem- onstrated the epitrichium in reptiles ; Jeffries, 83.1, and G-ardiner, 84.1, in birds — the latter author adding also considerably to our knowledge of the layer in mammals. KoUiker failed to recognize the layer in man (see his " Entwickelungsges.," 1879, and " Gewebe- lehre," 6te Aufl., 204). Minot, 86, showed that the layer is present in the human embryo at certain stages and is absolutely distinct from the underlying horny layer. The history of the human epitri- chium has been quite fully worked out by J. T. Bow- en, 89.1. The epitrichium becomes well marked during the third month, as a single layer of cells of large size, and each arching up from the surface. Fig. 307, Eptr. Over the hairy parts of the skin the development does not seem to progress beyond this stage. The cells of the epitrichium enlarge and gradually flatten down, but before they are completely flattened there intervenes a condition in which the ex- panded cells are flattened ex- cept m their central part, which forms a dome-like projection on each cell ; into this dome the nucleus and protoplasm of the cell are found withdrawn and degenerating. Later the cells are very large, Fig. 309, Fig. 309.— Epitrichiiim of a Human Embryo of the Fifth Month. a,b. Cells of two layers of the underlying horny layer drawn on the same scale. THE SKTN. 55:3 three to six times the diameter of the underlying epidermal cells ; there are no transitional forms, as Kolliker has erroneously main- tained, between the epitrichial and the underlying cells. The out- lines of the ceUs are polygonal and very distinct ; in the middle of each cell is an irregular lump of degenerated protoplasm, in which the nucleus can sometimes be distinguished. The epitrichium over- lies the hairs ; those hairs which project from their follicles lie be- tween the epitrichium and the rest of the epidermis. Over the hairless parts of the' skin the epitrichium probably per- sists and becomes several-layered, except that it disappears in great part over the nails, see p. 555. Thus, in an embryo of three months, there appear on the palms several layers of cells, all of which have the vesicular character and dark look of the cells of the single-layered stage. It is unknown how this growth of the epitrichium is effected ; the primitive epitrichial cells have so much the appearance of degen- erating tissue, that it is improbable that they proliferate, hence we must assume that the growth is effected by the addition of cells from the deeper layers. It was indicated above that in other parts the many-layered epitrichium probably undergoes cornification accord- ing to Zander's Typus A, and forms the stratum corneum of authors, which is found overlying the stratum lucidum. This probability rests chiefly upon Bowen's observation that the epitrichium over the developing nail is continuous with the horny layer. If we accept this interpretation, we must say that the epitrichial cell cornifies so as to form a thick-walled vesicle, while the underlying cells cornify so as to form solid scales (Zander's Typus B, 88.1). That the epitrichium in birds and mammals may become horny was demon- strated by Gardiner's careful researches, 84.1. Dermis. — Although the dermis or cutis is of exclusively mesen- chymal origin, it is convenient to consider its development in con- nection with that of the external skin. In very early stages the mesenchyma extends to the ectoderm, but shows no trace of a special layer under the epidermis. This layer is, however, well marked in embryos of two months by the condensation of the dermal mesen- chyma, the cells becoming flattened in a plane parallel with the surface, and hence they appear somewhat elongated in vertical sec- tions of the skin, Fig. 306, c; the nuclei are granular, the protoplasm forms a rich network of great delicacy. Later the protoplasm is, I find, more condensed around the nuclei, and the cells have more indi- viduality; at the same time the protoplasmatic network becomes coarser and simpler in character. During the third month (K511iker, "Entwickelungsges.," 1879, p. 774) the primitive dermis becomes differentiated into two layers, the true dermal (corium, Lederhaid) and the subdermal (Unterhautzellgeivebe) , the tissue being more condensed in the former and more fibrillar in the latter. During the latter half of the fourth month fat cells arise in the subdermal layer and steadily increase thereafter in both number and size, and by the end of the fifth month the whitish fat islands are conspicuous to the naked eye. The skin now comprises. Fig. 310, the epidermis, Ep, the dermis or cutis, Cu, and the fat-layer F; below is loose con- nective tissue, c. The hairs grow to the bottom of the fatty layer. The origin of the columnse adiposae (J. C. Warren, 77.1), calls for 554 THE FCETUS. investigation. The papillae of the dermis can be first seen during the sixth month (KoUiker, I. c. ) on the hand and feet, forming a double i^sss Fig. 310.— Vertical Section of the Slcin of a Human Embryo of thie fifth Month. Up, Epider- mis ; Cm, cutis ; F, fat layer : c, loose connective tissue. row between every pair of primary ridges, F-ig. 308, Hi. The elas- tic fibres appear during the seventh month (K511iker," Entwickelungs- ges.," 2te Aufl., 776). II. Nails and Hairs. Ifails. — A nail is a modified area of the stratum lucidum, situ- ated upon the upper side of the terminal joint of a digit and laid bare by the loss of the overlying epitrichium. This definition is essentially different from that hitherto current, and is based on Bowen's discoveries, 89.1. The first indication of the nails may be seen in the human embryo at the beginning of the third month as a thickening of the epitrichium over the end of the digit. In most mammals this position is per- manent and there is developed a terminal claw, but in man, as dis- covered by Zander, 84.1, the terminal position is transitory, and the ungual area migrates on to the dorsal side of the digit. The change of position is attributed by Kolliker, 88.3, 25, to the growth and expa,nsion of the palmar side of the finger-tip. A secondary result of the migration of the nail is the transfer of the terminal branches of the two digital nerves of the palmar surface to the back of the finger (toe) tips. Zander, 84. 1. The nail area is marked out quite definitely by a limiting groove or depression which persists more or less distinctly throughout life. As soon as the nail area has reached its dorsal permanent position, there appears at its proximal edge an oblique ingrowth of the Mal- pighian layer of the epidermis, to form the so-called root of the nail. The epitrichium over the nail is much thickened — see Bowen, I.e., Fig. 3 — but is thickest near and beyond the distal edge of the nail. The primary ridges of the Malpighian layer now appear, but only over the palmar surface of the finger or toe tip, and as they do not appear until much later under the nail, they establish a marked difference between the epidermis surrounding and that covering the NAILS AND HAIRS. 555 nail area. The epitrichial layer over the area has received the special name of eponychium from Unna, 76.1. Until the fourth month there is little change except that the anlage of the root of the nail grows considerably, and at the same time becomes more and more inclined toward a horizontal position, a change which pro- gresses until by the eighth month the nail-root is horizontal, i.e., in the same plane with the nail-bed proper — compare Fig. 311. About the beginning of the fourth month there appear, KoUiker, 88.2, 4, granules in the uppermost cells of the Malpighian layer. The granules are rounded in form, variable in size and have a decided -affinity for coloring matters, especially for acid fuchsin, Zander, 86.1, 285. Very soon the cells form a stratum lucidum, which appears first in the distal part of the ungual area and is very thin, thence spreads proximalward, and, last of all, appears in the nail- root, being there also preceded by the granular cells. By the middle of the fourth month the stratum lucidum is present over the whole nail and also extends on to the palmar surface. Fig. 311, s.l. The gran- ules have been supposed to be identical with eleidin, but on this point there has been some discussion, which is summarized by K61- liker ("Gewebelehre," 6te Aufl., 216); Ranvier ("Traite technique d'Histologie," 886) was the first to observe that the granules differ somewhat from true eleidin. There can be little question, if any, that the granules are directly connected with the cornification of the cells to form the nail proper. The granules were described by Brook in 1883, in a paper (Schenk's "Mitth.," II., 159), which I have not seen, and their relation to keratosis was more fully studied by Zander, 86.1, whose results have been in the main confirmed by KoHiker, 88. S, and F. Curtis, 89.3. The walls of the granular cells gradually become thickened (marginal keratinization of Curtis), the cell be- comes flattened, its nucleus disappears, the walls unite, and there is thus produced a horny scale in the place of the cell. By the trans- formation of additional cells, the horny stratum lucidum is constantly thickened on its under side — compare Fig. 163, in Kolliker's " Gewe- belehre," 6te Aufl. During the fifth month the development of the stratum gradually extends beyond the nail area over the rest of the finger-tips, and more slowly into the nail-root. The epitrichium disappears over the nail at about five months, first in the centre, then toward the base, sides, and distal end, but a small band persists as the perionix across the root of the nail. Fig. 311, Ep', and a large mass, Ep", forms a conspicuous ridge after the fifth month, across the distal end of the nail, and is continued over the palmar surface of the digit, as a considerable horny layer covering the stratum lucidum, s.l. The nail, N, although the direct continuation of the stratum lucidum, has, of course, its surface ex- posed. The epitrichium varies greatly in appearance, for it may either preserve more or less the vesicular form of its cells, or its cells may be more or less cornified and fiattened. It is probably owing to the frequency of the latter modification that the nature of the layer has been overlooked. The cornification of the epitrichium is pre- ceded by the appearance of eleidin granules in its cells, Curtis, 89. S, 17. The final step in the development of the nail is the change by 556 THE FCETUS. which its distal edge becomes free, according to KoUiker, 88.2, 7, by desquamation of the stratum lucidum at the point where the nail passes distally into the stratum of the palmar surface. Fig. 311.— Longitudinal Section of the Nail of the Great Toe of a Hirman Embryo of five Months. Minot Coll. No. 95. Ep, Remnant of epitriehlum; Ep\ distal ridee of epitriehium; N, nail ; s. ^ stratum lucidum ; 6, bone. From a section by Dr. Bowen, stained with acid f uchsin. Morphology. — The discovery that the nails are modified portions of the stratum lucidum gives the question of their evolution an en- tirely new aspect. It renders it probable that the claws and hoofs are also derived from the stratum lucidum, and that the develop- ment and changes of this layer of the epidermis will have to be care- fully investigated in the lowest am n iota before we can hope to understand the origin of claws. It may be safely assumed that the nail is a modified claw. Zan- der, 84.1, having observed the primitive terminal position of the nail area [Nagelfeld) in the human embryo, and its subsequent mi- gration to the dorsal side of the digit, concluded that the human nail represented a terminal claw flattened out, and that the centre of the nail must correspond to the point of the claw. Boas, 84.1, from comparative anatomical studies on claws, hoofs, and nails, established a distinction between the volar side and the palmar side of claws and hoofs, and homologized the nail with the volar side of a claw, which may therefore be termed the nail-plate (^Nagelplatte) ; Boas further maintained that the palmar side (sole-plate, Sohlen- horn) of the claw becomes rudimentary in man, and believed that its representative is the small area of epidermis under the edge of the nail in the adult ; this area probably corresponds to that which in the embryo is covered by the epitrichial ridge, Fig. 311, Ep", at the distal edge of the nail. This interpretation has been adopted by Gegenbaur, 85. 1, in whose laboratory Boas' researches were carried out. In view of our present knowledge it seems to me that Boas' conception must be accepted, with the modification, however, that the stratum lucidum covered by epitriehium over the end of the digit, must be considered the homologue of the sole- plate (Sohlenhorn), and that not merely the epitrichial ridge at the edge represents the NAILS AND HAIRS. 557 sole-plate. To decide the question, we must acquire exact knowl- edge of the relation of the sole-plate to the stratum lucidum in clawed and hoofed mammals. Hairs. — A hair is a long downgrowth of the mucous layer of the epidermis into the cutis, Fig. 312, A; into the enlarged end of the downgrowth extends a papilla, p, of mesenchymal tissue ; the down- growth separates into two parts, the axial or hair proper, H, which grows upward and projects above the surface, and a peripheral part or follicle, /. At the base of the hair, the hair itself and the follicle unite. The hairs arise in man as solid processes of the epidermis, the ends of which very soon expand, Fig. 313, 5, 6, and acquire the dermal papilla, 7. In other cases, as has been observed by Alexander Goette 68.1, and also, it is said, by Reissner and Feiertag, the papilla is formed first, as a slight projection of the dermis into the Malpighian layer of the ectoderm; the overlying epidermis then forms a downgrowth, which carries the papilla with it; in other respects the hair develops as in man. O. Hertwig ("Entwicke- lungsgeschichte," 3te Aufl.,436) regards the type of development in which the papiUa appears first, as the more primitive; this view is plausible, and enables us to assume that the hairs were evolved by modifications of the epidermis, overlying special dermal papillae. Hertwig fortifies his hypothesis by comparison with the teeth, which in the lower vertebrates are developed from dermal papillae, while in the higher forms there is a deep ingrowth of the epidermis before the mesenchymal papilla of the dental germ appears. The hair anlages appear in the human embryo at about three months, and can be first seen over the forehead and eyebrows, but very soon (sixteenth to seventeenth week) are developed over the en- tire head, and a little later the rest of the body, so far as it is ever hairy— on the limbs the hairs appear about the twentieth week. By the end of the fifth month, aU the hairy areas are marked out. From the third to the seventh month at least — my observations do not go further — new hair anlages continue to arise, so that one finds various stages at once. It is thus possible to study in one preparation the gradual differentiation of the hair. In embryos of five to seven months, which have died and been retained in utero, the epidermis is usually loosened and may be isolated.* Such a piece of epidermis stained with alum hsematoxylin and viewed from the under side is represented in Fig. 313. I distinguish two kinds of nuclei, those which are more darkly stained and those which are lighter. Some of the light nuclei appear dark because of the epitrichial cells un- derlying them. The darkly stained nuclei all belong to cells which participate in the formation of hairs. At first the dark nuclei make a little cluster, as at 1 and 2 ; the clusters grow in size^one a little larger is seen just to the left of that numbered 2, one a good deal larger is shown at 3. Sections show that such clusters are on the under-side of the epidermis and form slight protuberances or rudi- mentary papillae; the papillae lengthen out and acquire rounded ends, 4 ; they grow rapidly down into the cutis, and by the contrac- •The process may be imitated by soaking the skin of a foetus for several days in a 0.75 per cent salt solution to which a little thymol has been added to render it aseptic. 558 THE FCETUS. NAILS AND HAIRS. 55!) tion of their upper part become club-shaped, 5 and 6. The next step is the formation of the dermal papillae of the hair, 7 ; a little notch arises at the thick end of the epidermal ingrowth, and the tissue fill- ing this notch is the so-called dermal papilla. The figure presents also a well-developed hair ; here the axial portion of the papilla has Fig. 813.— Isolated Epidermis of a Human Embryo of five to six Months. 1-7, Hair anlages in successive stages ; /i, nair ; F^ follicle from which the hair has been pulled out; Gl^ anlage of sebaceous gland; /, wall of follicle; ft', bulb of hair. formed the hair, h, while the cortical portion has formed the follicle, /; the end of the hair is thickened, h', as the so-called hair-bulb ; the sebaceous gland, Gl, has begun to grow out from the follicular walls. In the upper part of the follicle the hair lies quite free, hence in sev- eral places where the hairs have been forcibly torn off the upper part 660 THE FCETUS. of the follicle, F, still remains, while the lower part attached to the hair is gone. The differentiation of the hair in the axis of the downgrowth be- gins about three to five weeks after the anlage appears, when the anlages are from 0.25-0.40 mm. long, and before the dermal papilla is recognizable. Two changes mark the commencing differentiation of the hair and the follicle : 1, the axial cells elongate in the direction of the future hair : 3, the outermost layer of cells assumes the char- acter of a cuboidal epithelium. The next step is the formation of the papilla, Fig. 313, 7, which is followed by the separation, in an- lages of 0.6-0.7 mm., of the axial mass of elongated cells into a smaller darker central portion, the hair proper. Fig. 313, H, and a lighter portion, which constitutes the inner follicular sheath, s. It is at this stage that the sebaceous glands, Fig. 313, Gl, and Fig. 313, A, gl, buds from the follicular tissue. At the enlarged base of the hair the layers all merge into one another. The hair proper grows in length very much, in diameter very little, and by its elon- gation penetrates the epidermal layers, being accompanied by the inner follicular sheath. As all the hair anlages descend obliquely, the hair penetrates the epidermis obliquely and within the epidermis is bent down. By its continued elongation it finally reaches the surface of the skin, and its tip remains covered only by the epitri- chium (Minot, 83), and when that disappears the hair is free. The detailed histbry of the hair follicles calls for much further study. I have observed the following details : In a longitudinal section of a fully developed hair. Fig. 313, A, the upper part of the follicle, F, is seen to have a central cavity, which is partly filled by the frag- ments of the broken-down inner follicular sheath ; on the lower side of the hair, and at the end of the hollow division of the follicle, is the anlage of the sebaceous gland, gl; from this point down there is no space between the wall of the follicle and the hair ; immediately below the gland is an eminence, m. i, which is formed by a thicken- ing of the follicle, and serves for the insertion of the slender muscle, muse, the erector pili. How this muscle arises is unknown. The thickening of the follicle where the muscle is attached is not men- tioned in the text-books I have consulted. From repeated observa- tions I conclude that it is a typical feature of the human hair. It has been described and figured by Unna, 76.1. Below the muscular insertion the follicle is differentiated into three layers, which are better shown under a higher power, Fig. 313, D; there is an inner- most sharply limited hornj- layer, S, with no trace of cellular struct- ure, a middle layer of granular cells, c, and an outermost layer of clear epithelioid cells, Ep, having their nuclei in their bases toward the hair, h. The follicle is incased in a fibrous mesenchymal tunica propria, tu. Returning to Fig. 313, A, the two outer layers of the follicle are seen to merge into one another toward the base of the hair, and to thin out and disappear ; the inner sheath, s, on the contrary, thickens, becomes more and more distinctly cellular, and finally expands as the hair bulb around the papilla. The hair proper, H, is of nearly uniform diameter until it reaches the bulb, where it expands to embrace the papilla, pa, and fuses with the iimer follicu- lar sheath. A network of blood-vessels, v, in the tunica propria NAILS AND HAIRS. 561 is spuii around the bulb, but vessels have not, in the stage figured, penetrated the papilla itself, pa. Lanugo is the term applied to the first coat of hairs in the embryo. This coat is a conspicuous feature at seven months. It is to be re- garded as the embryonic reproduction in man of an ancestral simian characteristic (Darwin, "Descent of Man," Chap. I.). The hairs are fine, compared with those of the adult, and are therefore usually described as woolly hairs ; they are lost from most parts of the body, and replaced by larger and coarser hairs. Over the face the lanugo persists throughout life, but owing to its fineness and loss of color is not usually noticed. Loss and Renewal of Hairs. — The length of life of a single hair is not long, for, as is well known, the hairs are continually shed. In many mammals the shedding is an annual process, but in man it takes place constantly. As the number of hairs, except in cases of baldness, does not diminish sensibly, it follows that new hairs must be continually formed. The loss of hairs begins during foetal life. The hairs shed by the foetus fall into the amniotic fluid and are sometimes swallowed by the embryo and found in the meconium, see Chap. XXIX. Imme- diately after birth the shedding of the lanugo occurs, its place being taken in certain parts by coarser hairs. The shedding of the hair is initiated by changes in the hair bulb, or expanded end of the hair fitted over the papilla; the multiplication of cells in the bulb, by means of which the growth of the hair is maintained, ceases, and the bulb atrophies, separates from the papilla, and breaks up into a bun- dle of fibres ; the hairs in which the bulbs have become fibrous are the Kolhenhaare of J. Henle, the Beethaare of P. Unna, 76.1. That these hairs which have no papillge cannot grow has been demonstrated experimentally by L. Ranvier. For a time the hair is still retained in place by the sheath of the follicle pressing against it. . It is finally either pulled out by some outside force, or pressed out by the secondary hair (Ersatzhaare^ ; there is also an actual shortening of the follicle of the atrophying hair, a fact observed by von Ebner, 76.1, and confirmed by KoUiker (" Gewebelehre," 6te Aufl., 241. As to the development of the secondary or replacement hairs {Ersatzhaare) , authors are not agreed. That there is a long contin- ued production of new hair-germs during foetal life is well known, and that the process is continued after birth has been maintained by several writers, but such hairs cannot be regarded as secondary- but only as primary hairs. The true secondary hairs are those which arise from the follicles of previous hairs. According to some authors, the old papilla is preserved and the new hair is formed over it, but this opinion does not appear to me to rest upon satisfactory observations. Far better founded is the view of Kolliker (" Gewe- belehre," 5te Aufl., 1867, p. 137), that the new hairs are developed from buds, which spring from the base of the old follicles soon after the old hair bulb has atrophied ; the buds are small in diameter, and lengthen out the old follicle ; the cells show, at first, no differentia- tion, the bud resembling closely a young hair germ ; in it a new hair is developed in the same way as in the primary hair germs. The 3fi 562 THE FCETUS. figure showing the development of the secondary hairs given by Kolliker in his "Gewebelehre," 5te Aufl., have been reproduced by him in the sixth edition, Figs. 186 and 187, also in his "Entwicke- lungsgeschichte," 1879, Figs. 476, 477. Sebaceous Glands. — As the sebaceous glands are outgrowths of the hair follicles, they are appropriately treated here. They ap- pear as thickenings of the follicles of the hair germs, about the time the hair proper reaches the level of the epidermis. The thick- enings are solid, and as they enlarge become somewhat lobulated, Fig. 312, A, gl; they usually are situated on the under side of the hair. Fig. 312, A, C, but sometimes spring laterally. Even before the lobulation begins, the anlage is seen to be differentiated into an outer layer. Fig. 312, C, cor, in which the cells retain their original character, and are small and granular : and a central mass of larger modified cells, Sb. The latter increase in number until they find an exit into the cavity of the follicle, Fig. 313, A, gl. According to Kolliker ("Entwickelungsgesch.," 1879, p. 797) the central cells contain fat globules and are discharged into the follicle, thereby becoming the secretion of the gland ; the cortical layer persists as the germinating bed of new fatty central cells. In specimens hard- ened in alcohol, stained in alum cochineal, and cut in paraffine, the central cells of the sebaceous glands of the foetus present a highly characteristic appearance; they are rounded or oval, and much larger than the cortical cells. Fig. 312, C. Under a high power. Fig. 312, B, each cell is seen to be separated from its fellows, to have a distinct outline, a coarse intracellular network and a finely granular rounded nucleus, lying in a perinuclear space, which is darker than the rest of the cells. The further development of the gland consists principally in the addition of lobules, which arise as buds of the cortical layer, the fatty central cells developing later in each bud (alveolus) ; the neck connecting the lobules with the hair follicle becomes the duct of the gland. The growing gland spreads around its hair f oUicle, but the position of its duct permanently indi- cates its origin from the under side of the hair. As the development of the sebaceous glands begins at a definite stage of the hairs, and as the hair germs continue arising throughout foetal life, so we encounter, at any time after the fifth month, glands in various stages. The first glands, according to Kolliker, appear on the head at about four and one-half months, on the body at about five months. Vernix Caseosa. — As we have learned from their development the sebaceous glands begin their secretory activity by the end of the fifth month. Their fatty secretion is discharged on the surface, and, together with the shed portions of the epidermis, usually forms a more or less extensive coating of the embryo. Minot, 83, has sug- gested that the persistence of the epitrichium may be a factor in the formation of the coating, which is known as the vernix caseosa {smegma embryonum, Kcisefirniss, FrucMschmiere) . Simon (" Med. Chemie, " II. , 486) is said to have been the first to show that the vernix consists entirely of sebaceous cells, fat globules and epidermal cells, and therefore could not be a product, as some of the older writers imagined, of the amniotic fluid. Quantitatively the epidermal cells GLANDS OF THE SKIN. 563 are the chief components. The vernix becomes conspicuous during the sixth month and increases until birth. It is extremely variable in amount. KoUiker states that Buck {" De vernice caseosa," Halis, 1844) found it might increase to 3.5 drachms in weight. In other cases it is almost entirely absent. Elsasser (Schmidt's Jahrbiicher, Bd. VII., 1833) found that about half the children of both sexes are born without vernix caseosa, the other half with a varying amount. On the chemical composition of the vernix see Davy (London Med. Gazette, 1844) and Buck ("De vernice caseosa," Halis, 1844). The vernix contains nine to ten per cent fats and seventy-eight to eighty-four per cent water. III. Glands op the Skin. The development of the sebaceous glands of the hair has been described, p. 562 ; concerning the development of the other sebaceous glands, such as those of the external ear and of the prepuce, little is known ; the glands of the eyeballs and eyelids are treated in Chapter XXVIII. ; there remain to be considered here the sweat glands and the mammary glands. Sweat Glands. — They arise as solid ingrowths of the Malpighian layer of the epidermis, somewhat similar at first to young hair- germs. They may be distinguished from hair-germs by their descending perpendicularly instead of obliquely, and by appearing in the fresh state — not whitish, like hair-germs, but yeUowish. They appear on the hairless parts (soles and palms) early during the fifth month, but not until much later on the hairy parts. KoUi- ker, 88.2, 15, has observed that the sweat glands are developed earlier on the under than on the upper side of the digits, and earlier on the third digit than on the others. The ingrowths arise on the soles and palms from the primary ridges. Fig. 308, S. The lower end is somewhat thicker than the upper part of the ingrowth, which rapidly elon- gates, passes through the dermis proper, and when it reaches the fatty layer or sub- dermal tissue, the anlage of the gland begins to assume a contorted course, the ■end of the gland roll- ing over toward the epidermis, Fig. 314. The lumen of the gland can be readily distinguished at this time, but does not extend through the epidermis until later— on the foot, not until the seventh month (Kolliker) . This fact is important, because it sets aside the notion, formerly advanced, that the sweat glands produce the liquor amnii. At the time of birth, the glands are longer, more coiled, and their ducts take a spiral course, but the spiral turns are by no means so close together or so numerous as in the adult. Fig 314, —Section of the Sole o£ the Foot of a Foetus of the fifth konth, to show the Sweat Glands, which arise from the in- ner or Malpighian layer of the epidermis. 564 THE FCETUS. ^?g%, ooOOdc'^oo'^ B Mammary Glands. — The milk glands vary in position. It is- probable that there were typically two rows of glands, and that different portions of these rows are preserved in different mammals, e. g., the headward portions in primates, the tailward portions in ruminants. According to 0. Schultze, 92.1, the first trace of the mammse may be observed, in pig embryos of 15 mm. and rabbit embryos of thirteen to fourteen days, as a continuous ridge-like thickening (Milchlinie) running from the fore-limb to the inguinal fold. In the next stage (20 mm.) the ridge is specially thickened — in the pig^ at 5-7 points, at each of which a mamma is developed ; each local thickening becomes separate and assumes a rounded form. The local thickening of the epidermis is the anlage of a milk gland, and this anlage has been long known and marks the site of the future nipple. In man the thickening may be observed toward the end of the sec- ond month. It is at first very slight, though it causes a discernible ex- ternal protuberance. Later it pro- jects from the epidermis into the dermis. The thickening commences when the epidermis is two-layered and solely at the expense of the in- ner layer, the outer layer persisting for a time as a distinct epitrichium, Fig. 315, A, Eptr. The epithelial ingrowth, Fig. 315, B, Ep.in, en- larges, and the cells in its central portion gradually cornify and fall out, so that the ingrowth becomes hollow; but the excavation pro- gresses verj' slowly and sometimes is not completed until after birth. Soon after the hollowing has begun the ingrowth sends out buds, which resemble, in their appearance and early development, true sweat glands. The organ may be said to be now in the monotreme stage. C. Gegenbaur showed in 1886 that in Echidna the mamma is a de- pressed area of the skin, from which spring a number of lacteal glands resembling the sweat glands in appearance. The depressed area Gegenbaur terms the Driisenfeld (gland area) . It seems to me be- yond possible question that the thickening of the outer skin to form the depressed area by a subsequent loss of cells in no wise militates against the homology here maintained, and which was first advanced by Gegenbaur. Epi Eptr ^••'}':A-^:ti'T'^^^f^^r%i:i'<\:"'((v.'\ Pig. 815.— Development of the Mammary Gland in the Rabbit. A, Embryo of 17 mm B, embryo of 39 mm. C, embryo of 75-80 mm. Eptr, epitrichium ; Ei^, epidermis ; .Ep in, epidermal ingrowth; Cii, dermis; ql, milk glands proper ; Mm, anlage of muscle ; sir, stroma of gland. (A and B are much more highly magnified than C.) After Eein. GLANDS OF THE SKIN. 565 In the stage of Fig. 315, C, all the parts of the adult gland may be recognized. The tissue around the epithelial ingrowth, Ep. in, is destined to form the protuberant nipple, of which the dermal tis- sue is clearly differentiated during foetal life, although the nipple does not usually become protuberant, according to Eein, 82. 1, until after birth. The boundary of the dermal tissue of the nipple is marked by a distinct layer of smooth muscle fibres, insc. Outside or below the muscular layer is the fibrillar connective-tissue stroma, str, into which the glands grow, and within which they are differ- entiated. The next stage of development is reached by a series of changes, of which the most important are : 1, the obliteration of the depres- sion, which arose by the hollowing out of the epithelial ingrowth ; 2, the development of branches from, and cavities in, the milk glands proper ; 3, the development of the fat layer under the gland ; and, 4, the growth of the nipple. The branching of the glands begins with the seventh month, and even at the time of birth is very slightly advanced. The lumen of the glands appears first in their enlarged lower ends, not long before birth, and then extends toward the mouth of the glands. In each gland we can distinguish: 1, the terminal branched glandular portion, and, 3, the duct. The duct consists of a wide part, sinus lacteus of authors, next the secretory portion, and a narrow part, which extends into the nipple and opens there on the apex ; the orifi^ce of the duct is funnel-shaped, and hence is termed the pars infundibular is. The fat layer is a continuation of that of the skin, locally thickened; about five or eight years after birth fat develops also in the stroma of the mamma between the gland tubules. The course of development has been shown by Rein to be essen- tially the same in several classes of mammals as in man. There are, however, noteworthy secondary differences, particularly in rumi- nants ; in them the nipple is precociously developed and the epithelial ingrowth carried up on to its apex before the gland buds appear ; the central cells of the ingrowth disappear as in man, but the de- pression left by their loss is not obliterated, but is permanent. Moreover, there is only a single gland bud developed, which grows out to a considerable length to attain the base of the long nipple or teat, where it branches. Consequently in ruminants there is but a single duct through the nipple, instead of several as in man and most mammals. In the horse (Rein, 82.2, 685) the epithelial in- growth forms two buds, hence there are two ducts in the adult. Milk at Birth. — Although the mammary gland is immaturely developed at birth, yet, as is well known, there is frequently a secre- tion discharged from it for a few days after birth. Scanzoni, de Sinety, and Rein, 82. 1, 464, have shown that this secretion is true milk. It is known in German as Hexenmilch. Montgomery's glands have been shown by Rein, 82.1, 470, to be accessory rudimentary milk glands. Evolution of the Mammary Gland. — That the mammary gland arose through specialization of a group of epidermal glands, is a necessary deduction from the facts of comparative anatomy and embryology. Several authors have thought that the milk gland was ■evolved from the sebaceous glands, others from the sweat glands. 566 THE FCETUS. The latter opinion rests upon strong evidence, the former principally upon the analogy of there being considerable fat in both the seba- ceous and lacteal secretions. Haidenhain (Hermann's " Physiologie, " Bd. v., 380) has shown that in the milk glands there is no fatty metamorphosis of the central cells, as in sebaceous glands, but a secretion from the gland walls, as in the sweat glands, so that there is nothing in the structure or function of the adult gland to justify a comparison with the sebaceous type. As regards the embryolog- ical development, the primary epithelial ingrowth. Fig. 315, A,, Ep.in, is, I think, to be regarded merely as the result of a modified method of developing the depressed glandular area (Drilsenfeld) ; the glands, sensu strictu, arise as solid, long, slender ingrowths of the Malpighian layer, and resemble closely the true sweat-gland anlages and not the sebaceous glands. Another point of importance is the resemblance, which has been observed by Gegenbaur, 86.1, between the milk glands of the lowest mammalia and the sweat glands. The derivation of the milk glands from the sweat glands is indicated by the structure and mode of secretion of the adult mamma, by the development of the gland, and by the structure of the gland in the Echidna. Gegenbaur, 75.1, 86.1, has maintained that there are two types of milk glands, one type modified sweat glands, the other type modified sebaceous glands ; he has maintained, also, that there are two types of nipple. The embryology of the organ shows that both the nipple and the gland are of one type, certainly in most, probably in all, mammalia. Gegenbaur's conception that there are two mor- phologically distinct forms of nipple was based upon Huss' obser- vations, which are inaccurate in several important respects. Literature. — Our knowledge of the development of the mammae was derived chiefly from the observations of Langer, 52.1, and of Kolliker{"Gewebelehre,"1867), until Huss in 1873, 73.1, greatly widened our acquaintance with the early stages in man and rumi- nants. Huss' memoir contained important errors, especially as to the origin of the ruminant teat, and these errors led Gegenbaur, 73.1, 75. 1, to his notion of two types of teats — a notion which has passed into the text-books, although shown by Eein to be untenable. H. Klaatsch, 84.1, argues against Rein in favor of Gegenbaur, but does not, it seems to me, invalidate either Rein's observations or conclusions. Rein's investigations, 82.1, 82.2, easily take the first place. Creighton's paper, 77.1, added but little, how little may be judged from his conclusion that the glands are developed from the mesoderm. CHAPTER XXVI. THE MOUTH CAVITY AND FACE. The face may be said to be a 'characteristic of the higher verte- brates, and to acquire increased importance as we ascend the series. In the marsipobranchs, ganoids, and selachians, the face does not form a projecting apparatus, there being merely an area on the ven- tral side of the head, which is distinguished by including the mouth and the nasal pits. The primitive arrangement is somewhat masked in the marsipobranchs by the modification of the mouth into a large sucking apparatus, but in ganoids and elasmobranchs it is preserved throughout life with little alteration. That the vertebrate mouth belongs primitively on the under side of the head, and is at first a simple transversely expanded orifice, is clearly established by the embryology of every vertebrate class. Balfour ("Comparative Embryology," II., 317) seems to have been the first to definitely for- mulate this generalization. The evolution of the face, so far as we could judge at present, depended, first, upon the enlargement and fusion of the oral and nasal cavities, which involved a change of site for the hypophysis ; second, upon the partial separation of the nasal and oral cavities, leaving the posterior nares open ; th ird, upon the growth and specialization of the facial region, of which the elongation of the jaws is the most conspicuous indication; fourth, upon the development of a prominent external nose. At the same time there occur modifications of position in the face in relation to the brain and its case or cranium, which it will be well to mention briefly in order to render the following sections of this chapter clearer. The position of the face, or oral region, is originally determined by the head-bend, as is more fully explained in the following section, see also Fig. 31 i). If we imagine a median longitudinal section of the head to occupy a rectangular area divided into quarters, then we may say the lower posterior quarter corresponds to the mouth region, the other three quarters to the brain. As development progresses, the oral quarter enlarges out of proportion to the rest of the head so as to project forward in front of the fore-brain ; in this stage, which is represented by the adult amphibians, the bulk of the facial appa- ratus is very great proportionately to the cranium. In the reptiles, the oral region is elongated still further in front of the brain case, resulting in the long snout. In mammals a third stage is established by the great increase in size of the brain, especially of the cerebral hemispheres, in consequence of which the brain comes to extend over the snout, as it were ; in man, whose brain has the maximum enlargement, the facial apparatus is almost entirely covered by the brain. The modifications involved in the increase of the bram in 568 THE FOETUS. mammalia, so far as the skull is concerned, have been considered p. 467 ; they are well indicated by Wiedersheim in his " Grundriss der vergleichenden Anatomie," 2te Aufl., Fig. 84. In brief, the facial apparatus, 1, underlies the hind brain, as in elasmobranchs ; 2, pro- jects in front of the brain (amphibia, reptiles) ; 3, is covered by the cerebrum (mammals). Formation of the Oral Cavity.— When the medullary tube enlarges to form the brain— see Chapter XXVII.— the end of the head bends over to make room for that enlargement. The bending of the head carries the oral plate over on to the ventral side of the freely projecting head, compare p. ^62. In Fig. 106, the head-bend is just developing; Ent, indicates the anterior extremity of the entodermal canal, and the reference line crosses the oral plate, or membrane formed by the union of the entoderm and ectoderm ; the oral plate occupies the entire space between the fore-brain, fb, and heart, ht, and there is as yet, properly speaking, no oral cavity, but it arises by the next changes which occur. The changes which de- velop the mouth cavity are the growth of the brain and of the peri- cardial cavity, both of which expand ventrally, leaving a space — the mouth cavity^ — ^between them. Fig. 170. Laterally the cavity is bounded by a wall or sheet of tissue, which stretches from the peri- cardial somatopleure to the head and is the anlage of the cheek ; it may be called the cheek plate ( Wangenplatte) . The mouth cavity is now a shallow fossa between the head and the heart, and still with- out connection with the entodermal canal (human embryo of 2.15 mm. with two aortic arches) . The fossa cannot, strictly speaking, be regarded as an invagination, such as is the invertebrate vorder- darm, p. 261, but is rather the result of the growth of the parts sur- rounding the oral plate. The oral pit is lined by ectoderm. "While the oral fossa is developing, the formation of the gill pouches begins. About the time the third branchial arch is formed, the oral plate ruptures in the human embryo, and the oral fossa communi- cates widely with the pharynx, Fig. 320. Upon the lateral and ventral sides no boundary can be found later, but upon the dorsal or cranial side a projection persists, Fig. 319, in front of which appears an evagination of the oral fossa, to constitute the anlage of the hypophysis cerebri or pituitary body (see below, p. 571), and behind which appears a second evagination from the pharynx to con-^titute the so-called Seesel's pocket, p. 268. The oral cavity proper and the pharynx are now merged into a single wide cavity, Fig. 320, for which we have in English no special term — in German it is some- times called the Mundrachenraum (His, " Anat. menschl. Embry- onen," Heft III., 26). The ectodermal mouth cavity, or oral fossa, does not correspond to the mouth cavity of the adult, for the adult cavity must include part of the pharynx, since it includes the tongue, which is developed from the floor of the pharynx, and in fact His has shown ("Anat. m'enschl. Embryonen," Heft III., 31) that the arcus palato-glossi, which are taken as the boundary between mouth and pharynx in the adult, are derived from the second or hyoid arches of the pharynx. Hence the adult mouth cavity includes the ectodermal oral fossa plus the region of the first gill-arches of the pharynx THE MOUTH CAVITY AND FACE. 569 In human embryos of the third week the mouth is a five-sided orifice, and I observe that the same shape appears in other mammalian embryos, and also in both amphibian and elasmobranch embryos, Fig. 316, hence it is probably characteristic of all vertebrates in the stage with five unmodified aortic arches. The mouth is bounded (His, "Anat. menschl. Embryonen," Heft III., 30) anteriorly by the wall {Stir7iwulst) of the head covering the fore- brain between the nasal pits, Fig. 316, lY, laterally by the maxillary processes, Mx, and latero-posteri- orly by the mandibular processes, Mel; the latter are the first branchial arches, and unlike the fol- lowing arches, br, they meet in the median ventral line. Another important factor in the development of the oral region is the descent or migration of the heart. It will be remembered that the aortic end of the heart moves from the anterior or buccal end of the pharynx, tailward. The change in the heart's position leaves the greater part of the pharynx free to be differentiated in intimate asso- ciation with the oral region, and the change also separates the mouth and the heart, so that very early we find the caudal or lower boundary of the mouth to be no longer the pericardial somatopleure, but the mandibular processes or arches, the ventral ends of which are developed between the mouth and heart. In certain teleosts, some time after the first pair of gill-pouches develop, the mouth breaks through in the ventral region of these pockets as a bi-lateral involution of the ectoderm, fusing with the entoderm and opening on each side of a central partition ; neither involution crosses the median lino. The double oral invagination was discovered by A. Dohrn, 82. 1, and the discovery has been confirmed by J. B. Piatt, 91.1, 263. In other teleosts (Mcintosh and Prince, 90.1, 773) the mouth is single and median in origin, as in the remaining vertebrates. The significance of the aberrant double origin is unknown, though Dohrn interprets it as evidence of the evolution of the mouth by the fusion of two gill-clefts. The Evolution of the "Vertebrate Mouth is still one of the -most puzzling of the unsettled problems of morphology. The in- crease of extent of the mouth cavity in the higher as compared with the lower vertebrates is discussed in the next section on the hypo- physis. The present section treats only of the origin of the verte- brate mouth. The first question is, necessarily, whether the mouth of vertebrates is homologous with the mouth of invertebrates or is a new structure. The formation of the embryo by concrescence enables us, I think, to decide between these alternatives. In Peripatus, the leeches, and the annelids with well-marked concrescence, the union of the ectental lines is incomplete, the anterior and posterior ends not meeting, but leaving the two ends of the elongated gastrula mouth Fig. 316.— Acanthias Embryo of 17 mm., underside. m6. Mid- brain; Ny nasal pit; Mr, maxillary pro- cess; ilf, mouth; Mrf, mandibular process; br, branchial arches; ht, position of heart ; F, yolk stalk cut across. 570 THE FCETUS. open, to form the mouth and anus respectively ; the mouth is car- ried inward by the invagination of the vorderdarm, and the primi- tive mouth is thereafter merely the opening of the vorderdarm or oesophagus into the archenteric canal.* In those invertebrates in which the process of concrescence is plainly marked, the mouth is seen to be the anterior extremity of the gastrula mouth, and to be bounded by the ectental line; the site of the invertebrate mouth is where concrescence begins, and it is, therefore surrounded by the ectodermal neural plate, f forming the brain (Sc/ie ^YeZpZaWe), oesoph- ageal commissures, and ,......-■- — ^ ventral nerve chain (Bauch- ^^ '\ ganglienkette) . The corre- /^ \ spending point in the verte- / "\ brate embryo is easily found, /' \ being between the optic j \ evaginations at the place ( \ marked m, in Fig. 317, and \ ' which probably corresponds \ j to the future infundibulum \^ m / in position. So far as I am ■'«. ^^ /' aware, the relations at this. '- ,..^ ■'''^ ^' point during early stages in " -" -. R """""^ vertebrates have never been FIG. 317.-Blastoderm of a Dog-Fish, Acanthias, thoroughly studied with the with commencing Concrescence. J/, Point corre- intention ot ascertaining sponding to invertebrate mouth; ij, blastodermic whether any traces of a Com- munication with the archen- teron could be found. Until this is done, there can be, in my judg- ment, little hope of our knowing what has become of the invertebrate mouth. The above determination of the site where we have to search ■ for the original mouth may be accepted with considerable confidence. If it is correct, it sets aside two hypotheses which have attracted attention : first, the hypothesis that the vertebrate mouth is identical with that of the invertebrate, and, second, the hypothesis that the old mouth is represented by the hypophysis, % for neither of these structures are derived from any part of the gastrula mouth. That both these hypotheses are untenable is evidenced by the deductions involved in their adoption. The annelid brain lies in front of the mouth ; if, therefore, either the hypophysis or the mouth of vertebrates is identical with the annelid mouth, then the brain and spinal cord must correspond to the ventral nerve chain only, and the annelid brain must have entirely disappeared. The vertebrate brain and eyes thus become new structures — a conception which seems to me * The meaning of the double origin of the mouth described by C. Semper in budding annelids and by Kleinenberg in Lopadorhynchus has not been explained. That it has the significance attributed to it by Kleinenberg can hardly be admitted, for there is no evidence that it represents- a primitive mode of development. t It seems to me justifiable to speak of this as a continuous neural plate, although there is a certain independence of development between the " Scheitelplatte " and ventral chain and al- though the commissures develop later. t That the hypophysis represents the annelid oesophagus was first suggested by A. Dohm 72 3 but he has since withdrawn his opinion. Similar was Richard Owen's infelicitous homologv of the hypophysis, infundibulum, and pineal gland with the old oesophagus (Proc. Linn Soc London, xvi). Beard has revived Dohrn's theory, hut has not succeeded in rendering it more plausible, to my judgment. Compare A. Dohm, 83.1. THE MOUTH CAVITY AND FACE. 571 indefensible. Another deduction involved in the vievsrs under dis- cussion is that a line of concrescence runs from the hypophysis or mouth to the fore-brain, representing the closure of the gastrula along that distance — yet of such a line not a trace can be detected. As the infundibulum is an invagination of the ectoderm tovirard the archenteron developed at or near the point vi^here the invertebrate mouth lay, it is quite possible that it corresponds to the oral invagi- nation (vorderdarm) of annelids. This identification has been more or less in the minds of morphologists for twenty years past, but no one has yet brought decisive evidence to justify it; nevertheless, its plausibility must be admitted. Since the vertebrate mouth is regarded as a nev^r structure, the second question comes: Hov^r did it arise? As we have seen, the first trace of the mouth is the oral plate, p. 262, formed by the union of the ectoderm and entoderm over a small area without mesoderm in front of the brain ; by the development of the head-bend the plate is carried over on to the ventral side and the oral cavity is developed. There is nothing in this history which we can recognize as a clew to the origin of the mouth, but, on the other hand, there is nothing in it strictly incompatible with Anton Dohrn's hypothesis that the mouth of vertebrates represents two gill-slits united in the median ventral line. The chief facts in favor of Dohrn's suggestion seem to me to be : first, that the trigeminal nerve shows the same relation to the mouth as other cranial nerves (facial, glosso-pharyngeal, and vagus) to the gill-clefts: second, that the gill-clefts approach the median line anteriorly, the first pair being nearest, the last pair furth- est from the middle plane; third, that the oral plate is formed like the membrane across a gill-cleft {Verschlussplatte) , p. 264, of ecto- derm and entoderm united without mesoderm. Dohrn has recurred repeatedly to this hypothesis in his " Studien. " Two other theories have to be mentioned, namely, Semper's and Balfour's. The former, 76.3, 336, observed a small ectodermal pit on the dorsal side of the head of a leech, which suggested the possi- bility of such a pit deepening and becoming connected with the archenteron, and so creating a new mouth. Balfour ("Works," I., 392-394) has suggested that vertebrates and annelids arose from an ancestor with lateral nerve cords, and that in annelids the cords united to form a median ventral chain, in vertebrates to form a median dorsal chain, so that in the former there is, in the latter there is not, an oesophageal ring. The development of both types by concrescence proves that the neural sides are identical in annelids and vertebrates. Therefore Balfour's hypothesis falls — and with it of course, Gregenbaur's — that the brain is the same in both types, but that the vertebrate spinal cord is an outgrowth of the annelidan supra-oesophageal ganglion, the annelidan ventral chain being lost in vertebrates. Hypophysis. — The hypophysis cerebri, Rathke's pocket or pitui- tary body (Hirnanhang) , is a structure of very problematical sig- nificance, which has been much studied and speculated upon by embryologists. It arises in all vertebrates as an evagination of the ectoderm near the dorsal border of the oral plate, but is separated from the plate by a fold of the ectoderm. In Petromyzon, Fig. 318, 572 THE FCETUS. the fold, F, acquires great size, and is shown by Dohrn, 83.1, to develop into the roof of the mouth and the upper hp ; accordingly the hypophvsal invagination, hy, is outside the oral cavity proper, ■'^ ^ " and more intimately associated with the olfactory area and nasal pit, N. The hypophysis runs toward the end of the notochord, nch, and is nearly met by a small blind diverti- culum of the archenteron, which is presumably homo- logous with Seesel's pocket, p. 268, of amniota; in the lamp- rey the hypophysis early gives rise to glandular diverticula, and itself becomes the adult nasal duct of authors. In elasmobranchs, owing, prob- ably, to the increased head- bend and size of the fore-brain, the region between the nose and oral plate is turned in so as to be almost wholly included in the oral cavity, and accordingly the fold. Fig. 319, F, and hypophysis, hxj, nov/ appear as appendages of the oral cavity, for I homologize the transverse fold. Fig. 319, F, which borders the hypophysis in shark embryos, with the fold. Fig. 318, F, which forms the upper lip in the lamprey. In amphibians, according to A. Goette's observations, 75.1, 288, 317, upon Bom- binator, the hypophysis arises as a solid ingrowth from the nervous layer (c/. p. 549) of the ectoderm, in front of the mouth, and, as development proceeds, there follows the inclusion of the hypophysal Fig 318.— Longitudinal Median Section of a re- cently hatched Larva of Petromyzon. fb. Fore- brain; pin, _pineal gland; inb^ mid-brain ; JVc/t, notochord; Ent, entoderm; mes, mesoderm; M, mouth cavity ; F, fold between hypopliysis and mouth ; hy, hypophysis ; N, nasal pit. After C. Kupffer, \. ■''^'■t- Fig. 319. — Lougicudinal Section of an Acanthias Embryo of 13.2 mm. f.b. Fore-brain; m.b, mid-brain ; m. tZ, medulla oblongata ; hy^ hypophysis evagination : if', fold separating hypophysis and archenteron; P/t, pharynx; ht^ heart; Li, anlage of liver; Yk.s^ yolk-stalk. area in the general mouth cavity ; there is no distinct fold between the hypophysis and the oral plate. In amniota nearly the whole ecto- dermal area between the oral plate and the nasal pits is turned in and incorporated with the mouth cavity before the evagination to form the hypophysis appears ; hence, the organ develops as an out- THE MOUTH CAVITY AND FACE. 573 gro\yth of the oral chamber. The comparative embryology of the pituitary body teaches us that the mouth cavity increases, as we ascend the vertebrate series, by the annexation of neighboring terri- tory, and that the primitive upper lip of vertebrates disappears, with the further consequence that in cyclostomes the homologue of the maxillary process is to be sought, not in the lip, but between the hypophysis and nasal pits. In mammals the hypophysis is first indicated (KoUiker, "Ent- wickelungsgesch.," 1879, p. 302) by a slight groove a little in front of the oral plate, but it does not have the form of a distinct evagina- tion imtil after the oral plate (Rachenhauf) is ruptured. The ecto- derm of the mouth over the hypophysal area lies against, and is apparently intimately soldered to, the ectoderm of the brain, a point Fig. 330. — Median Section of the Head of a Rabbit Embryo of thirteen and one-half Days. /b. Fore-brain ; mb, mid-brain ; c6i, cerebellum ; /i6, hind-brain ; nc/i, notochord ; hy^ hypophy- sis ; F^ fold corresponding to the lip of Petromyzon ; Ec^ ectoderm ; P, somatopleuric wall of pericardium ; Md^ mandime ; Ao^ wall of the aorta. which has been generally overlooked, but which seems to me of great importance. It is commonly stated, e. g. , by Kraushaar, 85.1, 87, that, when the oral plate ruptures, a portion of it persists upon the dorsal side, and is the beginning of the fold which separates the hypophysis from the pharynx. I think that this is probably not the case, but that all trace of the oral plate disappears and that a new fold arises as a duplicature of the ectoderm filled with mesoderm, Fig. 330, F. This new fold I homologize with the lip of Petromyzon, Fig. 318, F. The hypophysis is now, Fig. 320, hy, a diverticulum of the oral cavity, with one wall attached to the brain, and the other formed by a fold dividing the hypophysis from the mouth. The 574 THE FCETUS. epithelium of the mouth is one-layered, and not thickened, as is that of the hypophysis ; the cells are multiplying rapidly in the stage fig- ured, there being numerous karyokinetic figures, which, so far as I have seen, are always near the free surface of the epithelium. The relations of the notochord to the hypophysal wall have been dis- cussed, p. 183 ; in the specimen figured above, there is a connection between the chorda and the lower posterior part of the hypophysis. The organ in the stage of open invagination was described by Rathke, hence the invagination is often called " Rathke' s pocket;" Rathke supposed, erroneously, that it was developed from the archenteron (pharynx) . The hypophysal diverticulum now elongates and its upper end expands to a considerable vesicle, the lower end remaining narrow as the pedicle. At the same time the floor of the brain forms an outgrowth behind the hypophysis, which is the anlage of the infun- dibulum— compare Chapter XXVII. The two diverticula have their walls united. It is probable that the cementing together over tBe hypophysal area of the buccal and cerebral ectoderm is the mechanical condition causing the formation of the two diverticula. The hypophysis now grows rapidly ; the pedicle elongates and its lumen is obliterated ; the mesenchyma meanwhile condenses to form the base of the skull (sphenoid) ; the pedicle aborts entirely (in the rabbit by the sixteenth day) but the position for its passage through the sphenoid is marked a little longer, but is ultimately obliterated by the growth of the sphenoidal cartilage. According to Miklucho- Maclay (70.1, 40, Anm.) the passage persists in sharks. Lanzert (see Henle's Jahresbericht, 1868, p. 88) found traces of the passage, named by him canalis crcmio-pharyngeus, in children at birth in ten cases out of one hundred. There is then left merely the upper end of the hypophysis as a closed epithelial vesicle lying in the fu- ture sella turcica close to the infundibulum. The vesicle becomes flattened in the longitudinal direction, and the flattened vesicle soon acquires, at least in the pig, a yoke shape in section by becoming -first convex toward the fore-brain, then concave in its centre, toward the infundibulum, as may be observed in a pig embryo of 18 mm. (Kolliker, "Emwickelungsges.," 2te Aufl., Fig. 329.) The vesicle completes its development by sending out hollow buds from its ante- rior wall (rabbits, 20-30 mm.); in birds, according to W. Miiller, 71.4, and Mihalkovios, 77.1, buds arise from both walls. The buds elongate and branch; numerous blood-vessels are developed between them; the buds separate from the parent vesicle (rabbits of 40 mm.), but continue to grow; their lumen disappears, and they produce a highly vascularized complex of hypophysal cords. Kolliker thinks ("Entwickelungsgeschichte," 1879, 531) that the main vesi- cle persists recognizably in man into adult life. The infundibulum also contributes to the production of the adult hypophysis of mammals, although in lower vertebrates it persists as an integral portion of the brain, and is differentiated into ganglionic tissue. As first shown by W. Miiller, 71.4, the pointed end of the infundibulum undergoes in amniota an enlargement, beginning in sheep embryos of 35 mm., in pig embryos of 32 mm. (Kolliker, "Entwickelungsges.," 1879, 531). The knob-like enlargement loses THE MOUTH CAVITY AND PACE. 575 its cavitj-, and although the differentiation of nervous tissue begins in it, its cells early acquire an indifferent character, and it is pene- trated by blood-vessels and connective tissue; the connection with the brain is permanently retained. The knob is designated in the adult as the posterior lobe of the hypophysis, although it can in no sense be regarded as part of the true hypophysis. Historical Note. — ^The following memoranda are taken from Mihalkovics, 77.1, and Kraushaar, 85.1. The older authors re- garded the hypophysis as part of the brain ; this conception was held by Von Baer "Entw.-Ges.," I., 104, 103, and II., 393, and found as late as 1863 a defender in F. Schmidt, 62.1, 51, although Rathke had discovered the hypophysal evagination in 1838, 38. 1, and Rathke 's discovery had been confirmed by KoUiker ("Entwickelungsges.," 1861, p. 242) . Rathke subsequently, 61.1, 100, withdrew his opinion that the evagination formed the hypophysis, but W. Mxiller, 71.1, demonstrated that it was unquestionably correct, but retained the erroneous opinion that the evagination was developed from the arch- enteron. That the evagination belongs to the oral cavity was finally proven for amphibia by A. Goette, 75.1, and for mammals by Mihalkovics, whose researches, 77. 1, 83-94, are the most important yet made on the organ. Mihalkovics' results on mammalia have been confirmed by Kolliker, 79.2, Kraushaar, 85.1 (His, "Anat. menschl. Embryonen ") , and others. The development of the hypo- physis in the lamprey has been especially studied by Dohrn, 83. 1, whose results have been confirmed by subsequent investigators (Scott, 83.2, Shipley, 88.1, Kupffer, 90.1). That the notochord had some connection with the hypophysis has been held by several authors. C. B. Reichert, 40.1, 179, regarded the pituitary body as the end of the notochord, but twice later, 1861 and 1878, changed his opinion. Dursy, 69.1, maintained that the notochord was united with the pocket of Rathke, and formed part of the hypophysis ; see also J. B. Piatt, 91.1. Nasal Pits. — In this section the development of the cavity of the nose is taken up — for the history of the olfactory organ proper, see Chapter XXVIII. The formation of the nasal pits begins with the differentiation of the olfactory plates, which are two areas of thick- ened epidermis situated just in front of the mouth and in actual contact with the wall of the fore-brain. The plates give rise to the olfactory epithelium of the adult. In Petromyzon instead of two plates there is a single median one, which extends to the anlage of the hypophysis, Fig. 318. This fact renders it probable that primi- tively there was a single median plate in vertebrates, which has become divided ; in the lamprey such division is established later. H. Ayers, 90. 1, 240, however, states that the nasal area or olfactory plate of the larval lamprey is divided by a median non-olfactive raphe into two lateral pockets, right and left, to which the right and left olfactory nerves are respectively distributed. It is possible that more •exact observation will show that in all vertebrates there is at first a single plate, which is early divided. Balfour, "Comp. Embryol.," II., 533, regards the condition in Petromyzon as secondary, but gives no evidence to support his opinion, which was, perhaps, really due to the tradition which says the vertebrate olfactory organ is paired. 576 THE FCETUS. The nasal pits proper are developed, as pointed out by A. Goette, 75. 1, not by the invagination of the olfactory plate, which is apposed to the brain ab initio, but by the upgrowth of the ectoderm and mesoderm around the plate. The upgrowth takes place on the medial, upper, and lateral side of each plato, and hence forms two pits with a partition, the future septum narii, between them. They are the nasal pits and communicate along- their whole lower side di- rectly with the mouth cav- ity, Fig. 323. The mode of development of the nasal pits or sacs renders it highly probable, it seems to me, that the essential mechani- cal condition is, as with the hypophysis, the union of the epidermal plate with the brain wall. The nasal pit is at first very shallow, Fig. 321, and the olfactory. plate is exposed laterally; and there can be seen at its lower part a small depression, the anlage of the organ of Jacobson. The growth of the nasal pits in man has been described by His ("Anat. menschl. Embryonen," Heft III., 45-55). There are two principal changes, 1, the growth of the tissues around the olfactory plate ; 2, the migration of the pits away from the brain. Fig. 322 gives a view of an early stage in which the pits are small and shallow and the tissue is forming a ridge around them, which, however, does not extend on to the oral side, so that the pits open freely into the mouth cavity. The nasal pits are widely separated by a projecting mass of tissue, which I propose to call the nasal process, and which is the Stirnfortsatz of German embryo- logists'. Between the nasal pit on each side and the mouth the anlage of the nasal process is thickened and rounded, making a protuberance — the processus globularis of His. The nasal process includes the partition between the two nasal chambers, the anlage of the future nose and of the future intermaxillary region of the upper lip. The maxillary process extends be- tween the mouth and eye, toward the nasal pit, and later by joining the processus globularis begins the separation of the nasal and buccal Fig. 3S1.— His' Embryo A, 7.5 mm. N Fig. 322,— Facial Region, of a Human Embryo of 8 mm. ; front view. After W. His. X 10 diams. THE MOUTH CAVITY AND PACE. 577 chambers and completes the permanent upper border of the mouth — compare Fig. 324, L, Mx. As development proceeds, the lateral ridge, see Fig. 321, grows forward and covers in the nasal pit from the side, and may therefore be regarded as the anlage of the wing of the adult nose. We now have the two external nares. Turning to the growth of the nasal chambers, we find that they enlarge as the whole face enlarges, and that they occupy an increas- ing_ space, Fig. 323, Nh, opening widely into the mouth cavity above the palate shelf. The figure shows that the palate develops from the walls of the mouth cavity, and the space above it is, therefore, oral, not na- sal; hence the nasal cavity of the adult includes more than the nasal pit of the embryo. It is from the nasal pits proper that the so-called labyrinth of the nose is formed. The development of the labyrinth begins with the appearance — in man during the third month — ^of three projecting folds on the lateral wall of each nasal chamber. Fig. 326, the folds are the upper, middle, and lower turbinal folds {Naseti- mioschelii) and consist at first each of a duplication of the ectoderm filled with indifferent mesenchyma, which, however, very early changes into cartilage ; the turbinal cartilage is a consequence, not a cause, as often stated, of the development of the turbinal fold. The formation of the labyrinth advances by the formation of outgrowths, which become the ethmoidal sinuses, by the appearance — in man during the sixth month — of the antrum Highmorii, or expansion of the nasal cavity, into the region of the superior maxilla, and finally by the evaginations to form the sphenoidal and frontal sinus, which, how- ever, do not arise in man until after birth. Finally we consider the separation of the olfactory plate from the brain. This does not take place until the olfactory ganglion develops from the epithelium (ectoderm) of the plate. The olfactory nerve fibres are developed very early, in the chick during the third day — compare Chapter XXVII. The fibres lengthen, the olfactory and neural epithelia separate, and ultimately the osseous cribriform plate is developed between them. For observations on the development of the posterior nares, see Fr. Hochstetter, 91.2. Jacobson's Organ. — The organ of Jacobson arises very early as a small distinct invagination, on the medial wall of the nasal pit, as first stated by Dursy, 69. 1. Our knowledge of its development is due chiefly to KoUiker, 77. S, 79. S, 766, and Fleischer, 78.1. At four months it is a cylindrical blind canal, running from its 37 Fig. 32.3.— Reconstruction of the Face of His' Embryo Sch. N.of, Olfactory nerve; Nh, nasal cavity; R.T., Eathke's pocket; Ch, notochord; T, tonsil; F.g, processus flobularis; GZjalate anlage; Uic, mandi- le. After W. His, 578 THE PCETUS. original oriiice backward in the septum narii. It is surrounded by a small cartilage (Jacobson's cartilage) near its orifice; this separate cartilage is derived from a growth of the main cartilage of the sep- tum. The canal is innervated by the olfactory nerves, and in certain mammals it is much more developed than in man. The external nose is developed toward the end of the second month by a growth of the nasal process (His, " Anat. menschl. Embry- onen, " III. , 35) . It is at first short and broad, having at three months very nearly the shape which is permanent in certain negro races. The external nares and wings of the nose are carried forward with the general nasal upgrowth. At three months the external nares are usually completely closed by the growth of their epithelium, which forms a plug of gelatinous consistency. The plug disappears after the fifth month (KoUiker, "Entwickelungsges.," 1879, 767). Maxillary Process. — ^Eef erence has already been made to that thickening of the upper edge of the mouth, which appears almost as a continuation of the mandibular arch, and which is known as the maxiUary process, or sometimes as the superior maxillary process (Oberkieferfortsatz). It is termed a process, because from its small size and position it appears at first like a bud from the mandibular arch. Later it stretches farther forward, and when the mouth has changed from its original pentagonal shape to a transverse slit, Fig. 322, the maxillary process no longer appears specially con- nected with the mandibular arch, but is united with the edge of the nasal process as above described, p. 57G. A thorough study of the primitive relations and growth of the maxillary process is much needed. It is possible that, as several authorities have maintained, it is morphologically the upper part of the mandibular arch, which, in consequence of the head-bend, makes an angle with the mandible proper. Although this hypothesis commends itself to the embryolo- gist, it needs a firmer basis than it yet has to stand upofa. Mandibular Arch..— The first branchial arch forms the lower boundary of the mouth, and by its long-continued growth develops into the projecting lower jaw. The x^" " - " history of the skeleton and muscles of the lower jaw are treated, p. 444 and 478, respectively. The chin is at first retreating and does not be- -^ >' <" . 1 . . " Mx come distinctly prominent until the fifth month. The growth of the jaws increases the separation of the mouth from the heart. Lips and Gums. — Very soon after the upper jaw has been formed ^ „„, „ by the union of the maxillary and ^ Tie. 324. -View of the Roof of the Mouth riQool r^r-r^nc^cicac i4-„ ^ ^ £ j of a Human Embryo, na, Nares; Op, eye; Hasai prOCeSSCS, itS Oral Surface de- «;„S2^?°?^°' ''?'^®™i'iE'"* '''°™,'^®i'*^?i velops two parallel ridees Fis 324 process ; Jlfx, portion of the upper lips devel- /. x.- i. xi, "^^ ^'"^gc, j. ig. o«*, oped from the maxillary process : D, dental 01 WniCn tne OUfcer and more bnlkv groove on the gums: Pai. nalatfi vfiriiatna 7- nr.. •_ jt ■. _"io i^uxivj, After W. His. / Op Pal, palate. X 8 diams. L, Mx, is the anlage of the upper lip, and the inner and smaller, D, the anlage of the gums (gingivce, dental ridge) . At about the same time, or a little later, similar ridges develop on the lower jaws THE MOUTH CAVITY AND FACE. 579 The histogenesis of the lips and gums has not been investigated. From the study of sections of the lower lip of a fcstus of six months, which I have prepared, I consider it probable that the peculiar epi- thelium of the lips arises, 1, by the disappearance of both the epitri- chium and stratum lucidum, and, 2, the distention of the remaining cells — a basal growing layer being retained. In a rabbit of thirteen days, the epitrichium runs over the region of the future lip. In a pig embryo of about 3.5 cm., the epitrichium is still present, and the cells below are enlarging and beginning to cornify. The glands of the lips, according to Kolliker, " Entwickelungsges.," 1876, 828, arise during the fourth month as solid ingrowths of the epithelium, and later send out each eight to ten branches, which, while still solid, form a pretty rosette. Formation of tlie Palate. — As soon as the external nares are separated from the mouth, there is a partition between the nasal pits Fig. 325.— Frontal Section of the Oral and Nasal Chambers of a young Cow Embryo, cat. Palate; lat, lateral cartilage; Eth, ethmoid plate; ii6, inferior turbinal; T, tongue; d, dental germ; a, oral angle; M, Meckel's cartilage; Md, mandible. and the mouth. This partition, in which the intermaxillary bone is differentiated later, is supplemented by another partition, the true palate. Fig. 324, Pal, which shuts off the upper part of the oral cavity from the lower, thus adding the upper part to the nasal chambers. The palate is a secondary structure, which divides the mouth into an 580 THE FCETUS. upper respiratory passage and a lower lingual or digestive passage. The palate arises as two shelf -like growths of the inner side ot each maxillary process, Fig. 324, Pal, and is completed by the union ot the two shelves in the median line. As seen in a side view the shelves are represented in Fig. 323, Gl, they arch so as to descend Fig 326 —Frontal Section of the Nasal and Oral Cavicies of a Human Embryo of three Months (Minot Coll. No. 411. a certain distance into the pharynx, but in the pharynx their growth is arrested, though they may be still recognized in the adult. In the region of the tongue, which includes more than the primitive oral cavity, the palate shelves continue growing. At first they project obliquely downward toward the floor of the mouth, Fig. 325, pat, and the tongue, T, rises high between them, and appears in sections which, like the one represented in Fig. 325, pass through the internal nares, to be about to join the internasal septum. As the lower jaw grows, the floor of the mouth is lowered and the tongue is thereby brought further away from the internasal septum. At the same time the palate shelves take a more horizontal position and pass toward one another above the tongue and below the nasal sep- tum, and meet in the middle line where they unite. From their original position, see Fig. 325, pat, the shelves necessarily meet in. front (toward the lips) first, and unite behind (toward the pharynx) later. In the human embryo the union begins at eight weeks, and at nine weeks is completed for the region of the future hard palate, and by eleven weeks is usually completed for the soft palate also. The palate shelves extend back across the second and third branchial arches ; by the migration of the first gill pouch, or, in other words. THE MOUTH CAVITY AND FACE. 581 of the Eustachian tube, the Eustachian opening comes to lie above the palate (uvula) while the second cleft remains lower down and lies below the palate, as the anlage of the tonsil. His, "Anat. menschl. Embryonen," Heft III., 82. The uvula appears during the latter half of the third month as a projection of the border of the soft palate. Soon after the two palatal shelves have united with one another the nasal septum unites with the palate also, Fig. 336, and thereby the permanent or adult relations of the cavi- ties are established. Lachrymal Duct. — The canal which leads from the corner of the eye to the nose {Thrdnennasengang of G. Born) is not found in fishes, but only in amphibia and amniota. The site of this duct is very early marked out by the lachrymal groove, Fig. 323, running down from the eye to the invagination, or to the nasal pit as soon as the latter appears. This groove is bordered above by what is known as the lateral nasal process or prominent surface between the nasal pit and the eye — compare Fig. 333 — and is bordered below by the maxillary process. This groove soon disappears and leaves, so far as known, no trace. It was supposed by Kolliker (" Entwickelungs- gesch.," 1879, 469) to be the anlage of the duct — an opinion which Bern's observations on amphibians, 76.1, and on Sauropsida, 78,1, 83.3, followed by those of Legal, 81.1, 83. 1, on mammals, showed to be erroneous. The duct arises along the line of the lachrymal groove as a thicken- ing of the under side of the epidermis, which appears about the time that the cartilage develops around the nasal cavities — in man, ac- cording to Ewetzky, 88. 1, at the end of the fifth or beginning of the sixth week. The thickening increases until it forms a ridge, which finally separates as a solid cord from the epidermis, except at each end ; the cord then acquires a lumen, thereby becoming an epithelial canal. In man the upper end of the solid cord broadens out at the inner canthus and then divides into two forks, each of which acquires a lumen, with the result of producing a bifurcation of the duct (Ewetzky) . In the pig, the bifurcation is developed, but one fork aborts, according to Legal, 83. 1. Teeth. — The development of the teeth in man and other mam- mals has been much studied, and has been repeatedly described by competent authorities in comprehensive summaries. I have, there- fore, deemed it unnecessary to go over many of the original articles carefully, and instead base the following synopsis chiefly upon Waldeyer, 73.1, Kolliker, 79.2, 815, Tomes ("Dental Anatomy"), Von Ebner, 90.1, and O. Hertwig. The list of authorities is given in my "Bibliography" under " Teeth," but it is far from com- plete ; for further lists see Waldeyer and Von Ebner. For a very admirable critical synopsis of the various notions that have been advanced concerning the histogenesis of the teeth, see Von Ebner, 90.1, 349-353. It must be remembered that most of the articles upon the human teeth are by more or less incompetent writers. Dermal Teeth of Sharks. — The teeth were primitively organs of the skin and widely developed over the surface of the body, and as stated before, p. 461, they have played an important role in the genesis of the skeleton. It is, therefore, to the fishes that we must Fig. SSU. —Dental Papilla of a Dermal Tooth of an Acanthias Embryo of 10 cm. En, Enamel organ; p, papilla; Ep, epidermis; Ou, dermis. After O. Hertwig. 583 THE FCETUS. turn to ascertain the primitive mode of tooth formation, choosing the sharks, since they have been the most thoroughly studied in this regard, thanks chiefly to O. Hertwig, 74.1,2. The teeth of sharks are generally known as placoid scales. The tooth begins as a mesenchymal papilla, Fig. 337, composed of crowded cells and projecting into the epider- mis. The layer of epidermal cells overlying the papilla changes in character, its cells gradually lengthening into very long cylinders, and be- comes the enamel organ. By further development the epi- dermis thickens, the papilla projects into it, and becoming narrow and longer, and taking an oblique position, gradually assumes the shape of the tooth. Ossification now begins over the surface of the papilla ; there arises a layer of epithelioid osteo- blasts, and between these and the enamel organ the development of bone, or, as it is called in teeth, of ivory, begins ; the osteoblasts per- sist, and the bony structure is developed onlj- between them and the epidermis, forming a stratum which grows in thickness. At the same time the enamel organ begins to deposit the calcified layer, known as enamel, over the papilla. Later the tooth acquires a sup- port by the direct ossification of the connective tissue at its base, and is then a completed "placoid scale." The teeth of the mouth depart from this primitive mode of devel- opment, for they do not arise on the surface, but deep down, Fig. 338, because the dentiferous epithe- EP\ Hum grows down into the dermis, forming an oblique shelf, which may be regarded as a spe- cial tooth - forming organ. On the un- der side of the shelf the teeth are devel- -^ F n oped in the same n D p D.s ■wn-o- n<5 nvpr +>io „FiG. 328.— Section of the Lower Jaw of an Acanthias Embryo of way db over me lO cm. r, Tooth; em, enamel cap; Wp, epidermis; ^dentinef^jt. Skm, although they «°amel cells; P, dental papllla; X>.sr dental shelf. After O. Hert- are much larger. The teeth are, however, in various stages of development, and only one is fully exposed ; when, as happens in time, it is lost, the next tooth behind replaces it, and since the production of new tooth germs goes on in adult life, the replacement of teeth in the shark's jaw continues indefinitely; hence sharks are termed polyphyodont. Mammals have two sets of teeth, and hence are called diphyodont. ' THE MOUTH CAVITY AND FACE. 583 Ep.i o.ep papilla. iter C. Eose. dental shelf; Pp, "VVe learn from the sharks that a tooth is a papilla which projects into the epidermis, and, ossifying in a peculiar way, changes into ivory around the soft core or pulp : to the papilla the epidermis adds a layer of enamel. The tooth proper unites with a small plate of dermal bone at its base. By a modification in the jaws, the epider- mis first grows into the dermis, and then the dermal tooth papilla is developed. In the higher vertebrates the teeth of the jaws on\j are developed, and they arise in the modified way we have noted in the selachian jaw. Amniote Tooth-Germs. — The first indication of the development of tooth-germs in mammals is the appearance of a thickening of the epithelium covering the jaw ; the thickening forms a curving ridge on the under side of the epithelium. According to C. Rose, 91.2, 461, the ridge appears in the human em- bryo during the sixth week. The ridge expands, Fig. 329, and subdi- vides into an outer portion, L.gr, the anlage of the groove between the lip and gum, and an inner por- tion, d.sh, the dental shelf, which grows obliquely inward ; on the un- der side (in the upper jaw on the corresponding upper side) of the shelf arise the dental papillse. Pp. The dental shelf (Zahnleiste) is homologous with the similar structure in the shark. Its history in the human embryo has been investigated by C Rose, 91.2. The papillae for the milk-teeth are formed on the under side of the shelf. Fig. 329, and it is thus possi- ble for the shelf to continue growing toward the lingual side, so that a second set of germs is developed for the permanent teeth. The end of the shelf toward the articulation of the jaws is prolonged without retaining the direct connection with the epithelium, and from this prolongation arise the enamel organs for the three permanent molars. "Wherever a tooth-germ arises, the dental shelf is locally enlarged, and the local enlargement constitutes an enamel organ which projects from the under side of the shelf. The portions of the shelf between the enamel organs gradually break up, forming first an irregular- network, and later separate fragments, which may persist throughout life and lead to various pathological structures ; while the permanent germs are forming the shelf is solid between them, although it has assumed the reticulate structure between the germs of the milk- teeth. In consequence of the reticular formation, the fully developed enamel organs have several bands or threads, by which they are connected with the dental shelf proper. Fig. 330 represents the under side of a model of the epithelium of the gum of the upper jaw of a human embryo of 40 mm. reconstructed by G. Rose from the sections. Fig. 329. L.gr is the ridge corre- sponding to the groove between the lip and gum; pal is the surface of the palate; d.sh is the dental shelf, the ten cups or depressions on which correspond to the papillee for the ten milk-teeth. Fig. 329. — Section of Part of the Lower Jaw of a Human Embryo of 40 mm. Ep.K Epi- thelium of lip; o.ep, oral epithelium; Lj;*", anlage of lip groove ; d. s/t, der Afte ' " " 584 THE FOETUS. a.sh After the shelf has developed somewhat, its line of connection with the epithelium of the gum becomes marked by a superficial groove, as may be seen in the human embryo of eight to ten weeks. Fag. 334 D. This groove was formerly supposed to be the first trace of ' the dental shelf, but Rose's observations correct the sup- position. The second step in mam- mals is the formation of out- growths (in man ten in each jaw) from the under side of the dental shelf; each out- growth is the ajilage of an enamel organ for a milk- FiG. 330. -Explanation in text. ^^^^^_ r^^^ derivation of the enamel organ from the epidermis was discovered by Kolliker. The outgrowth is covered toward the mesoderm by a layer of cy- lindrical epithelial cells, the continuation of the basal layer of the epidermis, while the core is filled with polygonal cells, which resemble those of the middle part of the Malpighian layer of the skin. The outgrowths, after penetrating a short distance, expand at their lower ends, but remain each connected by a narrow neck with the overlying epidermis. The expanded end is the enamel germ proper ; it very soon assumes a triangular outline as seen in sections, owing to the flattening of its under side, and at the same time it moves somewhat toward the lips. Meanwhile the shelf con- tinues growing on the lingual side of each ingrowth, to produce the enamel organs destined for the second or permanent teeth. At this stage we notice that the mesenchyma under the flattened end of the enamel organ has become more dense, to form the anlage of the dental papilla, and is beginning to develop fibrillae around both the enamel germ and the papillary anlage. The fibrillar envelope is the future dental follicle {Zahnsack) . The third step is the final differentiation of the enamel organ and the accompanying shaping of the papilla. The enamel organ, Fig. 331, continues growing and becomes concave on its under side, so that the mesoderm underneath acquires the shape of a papilla. It is now that the form of the tooth is determined by the form assumed by the papilla, which in its turn is probably determined by the growth of the enamel organ. Von Brunn, 87.1, has shown that the enamel organ extends over the papilla of various mammals not only as far as the enamel is formed, but also as a thin layer to the base of the papilla, or over the future root. Over the root, after the tooth is shaped, the enamel organ aborts. The apex of the root is never covered. C. Rose, 91.2, has shown that in man also the enamel organ extends at first over the root, but subsequently aborts. A fully developed tooth germ consists of, 1, the follicle, 3, the. enamel germ with its neck running to the dental shelf, the edge of which grows on, Fig. 331, B, to form the secondary teeth, and, 3, the papiUa. The follicle is merely an envelope of connective tissue, Fig. 331, in which we can distinguish, according to Kolliker, an outer THE MOUTH CAVITY AND FACE. 585 denser and inner looser layer; in the latter the cells are more distinct and the fibrillae are less numerous than in the former. A rich net- work of capillarj' vessels is developed in the follicle, Fig. 331, v, and appears in part as a series of villous-like growths into the «namel organ. The follicle develops first over the lower part of the .■iS^i- ii-B Fio. 331.— Vertical Section of a Molar Tooth Germ of a Human Embryo of 160 mm. Ep, Epi- thelium of the dental furrow ; B^ bud for secondary germ ; En^ central cells of the enamel organ ; c, enamel cells; p, mesenchymal papilla; v, follicular envelope with blood-vessels. papilla, then over the enamel organ, the neck of which aborts and the follicle closes over, completely separating the enamel organ from its parent epidermis. The enamel organ changes greatly in appearance. The layer of cylinder cells is well preserved only over the concave side, Fig. 331, c, where the epithelium is in contact with the dental papilla. In the neck the cells become appressed and irregular in form. Over the convex surface of the enamel organ the cells become lower and 586 THE PCETtrS. cuboidal, and ultimately atrophy and flatten out, but, so far as I know, no exact study has yet been made of the changes they pass through. The convex surface becomes very irregular by upgrowths of cells, crowded together; it is between these upgrowths that the vascular villi of the follicle are formed. The layer of cylinder cells over the papilla become much elongated and as their nuclei, after the enamel has begun to form, are nearly all placed at about the same level, they constitute one of the most beautifully regular epi- thelial layers known. These cells covering the papilla are known as the enamel cells {Schmelzzellen, ameloblasts, membrana adaman- tina of Easchkow) because they produce the enamel, as described below. The enamel cells average about 40/^. in length, and at birth about 6-7/j. in width; their outer ends, i. e. away from the papilla, are furnished with prickles or thread-bridges by which the cells are connected. Fig. 332, with one another and the neighboring cells of the enamel organ; the bodies of the cells are finely granular, and not infrequently have larger glistening granules at their lower or papillary ends; their nuclei are elliptical and 10-13/^- long; before the enamel appears they lie at various levels ; after it appears they are found, with rare exceptions. Fig. 332, 6, near the upper ends of the cells, all at one level. The lower or papillary ends have the processes of Tomes, so named from their discoverer ; these appear when the enamel begins to form ; they are short, thick, and tapering, one on each cell; they often seem fibrillated, and are always separated "from the cell proper by a small cuticular border; while in situ Tomes' pro- cesses are fitted into sockets on the surface of the enamel. The enamel ceUs have, probably, no membrane on their sides. After the for- mation of the enamel is completed the enamel cells degenerate and are lost, except, 1, that their border persists as a horny membrane, cuticula eboris, covering the enamel, and, 2, that a few groups of cells may remain for a long time as isolated epithelial bodies in the dental follicle (Malassez). The cells in the centre of the enamel organ undergo a very peculiar metamorphosis. They remain united together by a few thread-like processes, and, therefore, have a certain degree of resemblance to the embryonic connective tissue cells, but the intercellular spaces do not contain in the enamel organ any homogeneous matrix, but merely fluid. The steps by which this metamorphosis of the central cells is accomplished are still imperfectly known. A few layers of the central cells of the enamel organ retain more of their primitive character, Fig. 332, c. These cells constitute the intermediate layer of KoUiker ; they are polygonal, granular, and connected with one another by intercellular threads (prickles). Fig. 332.— Part of the Enamel Or^an ot a New-Boru Child; Incisor Germ, a, Tomes' pro- cesses; b, enamel cells; c, mid- dle layer of prickle cells ; d, central or pulp-cells. After V. von Ebner. X about 550 diams. THE MOUTH CAVITY AND FACE. 687 The DENTAL PAPILLA consists at first, as stated above, of crowded mesenchymal cells. Blood-vessels appear in it very soon after the enamel organ has become concave on the lower side. The papilla acquires very nearly its permanent shape before any further differ- entiation of its tissue begins. The shape of the papilla is probably determined entirely by the enamel organ, by which it is completely embraced, see above. During the fourth mouth the cells nearest the surface enlarge — principally by the growth of their protoplasm. They appear as a continuous layer next the enamel organ ; their function is to produce the dentine between themselves and the enamel organ, hence they are called odontoblasts {membrana ehoris, KoUiker) ; they are to be regarded as modified osteoblasts. The deposit of dentine begins in the milk-teeth toward the end of the fourth month. In a vertical section of a developing papilla, one can see several stages, because the development advances more rapidly toward the apex and more slowly toward the base of the papilla. The tissues underneath the odontoblast layer constitute the so-called pulp of the tooth. The connective-tissue cells of the embryonic pulp are small and have numerous very fine and branching processes which impart a fibril- lated appearance to the tissue, but so far as known there are no true intercellular fibrillse in the pulp. The cells are somewhat more crowded directly under the odontoblasts than in the interior of the papilla. Enamel. — The deposit of enamel begins on the milk-teeth toward the end of the fourth month. According to our present knowledge, the formation of enamel must be conceived about as follows : Each enamel cell forms an enamel prism by the metamorphosis of the lower end of the cell into a calcified column ; a cement, which is also calcified, holds the prisms together; the cement is presumably a derivative of the inter-cellular substance between the enamel cells. Enamel is, therefore, essentially different from bone and dentine, in neither of which do the cells calcify, yet the enamel cells resemble odontoblasts in many respects. The first step toward the production of an enamel prism is the change of the protoplasm at the lower or papillary end of the enamel ceU into a homogeneous mass, resembling a cuticular cell border ; by the union of the borders of adjacent cells, a continuous membrane or cuticula is generated. We must assume that this membrane grows upon its upper side by apposition from the enamel cells, and becomes modified on its lower or papillary side at nearly the same rate. The modification consists in the production of the fibrous tuft. Fig. 332, a, described above, at the end of each enamel cell. The lower end of this tuft (Tomes' process) cal- cifies and becomes the beginning of the enamel prism. The enamel prisms begin small in diameter with considerable cementing sub- stance between, but, as they lengthen, their diameter increases so much that there is little or no space for cementing substance between them. The enamel prisms lengthen by apposition on their ends adjoining the enamel cells, yet for a long time the cells maintain their size, perhaps nourishing themselves at the expense of the cen- tral cells of the enamel organ, which gradually atrophies as the enamel thickens. From their mode of growth, it follows that the prisms stretch through the whole thickness of the layer of enamel. 588 THE FCETUS. Since the enamel prisms widen out toward the surface of the tooth, it is probable that the enamel cells increase in diameter as the enamel is deposited. The cells cease multiplying by the time the enamel begins to form. The enamel prisms undergo further changes after birth. They become harder and thicker at the expense of the ce- menting substance between them. At birth it is still relatively easy to break up the enamel into . its prisms, and to a, certain extent to break the prisms so as to obtain indications of fibriilated structure. Dentine. — The odontoblasts, as stated above, are modified mesen- chymal cells, which form an epithelioid layer over the surface of the papilla. The odontoblasts are, at first, short cylinder cells, each with an oval nucleus toward the end of the cell farthest from the enamel organ. They keep their mesenchymal character in that they are connected by processes with one another and with the underlying cells of the papilla. The first change in the odontoblasts prepara- tory to the deposit of dentine is the appearance of the so-called membrana prceformativa, a clear homogeneous membrane consist- ing apparently of anisotropic intercellular substance. The mem- brana always lies next the odontoblasts and is best interpreted as the layer of uncalcified dentine, see C. Eose, 91.2, 470. There now arise the dental processes, which are prolonga- ^ , j I tions of the odontoblasts toward the enamel organ ' ' I as far as the membrana praeformativa. The pro- : cesses vary much in size, but are generally about " , ' one-sixth to one-fourth the diameter of the cells ; ^ f-, - ' each cell usually has one dentinal process only, ajL^**^ , but sometimes there are two, and even as many Wp ' as six have been seen by Boll. Between the den- Ij^l I. I I final processes a clear anisotropic substance is formed, which gradually increases in thickness, the processes lengthening correspondingly, until a considerable layer, which may be described as uncalcified dentine, intervenes between the odon- ' N toblasts and the enamel organ. Calcification sets in next the enamel and progresses toward the papilla ; at the same time the deposit of uncalcified dentine is continued by the odontoblasts. The calcification is incomplete; the uncalcified spots are known in the adult tooth as the interglobular spaces. The membrana praeformativa cannot, as suggested by Von Ebner, 90.1, 244, be resorbed by the enamel organ, since it is not in contact with it, but it is to be observed in well-developed teeth, and is perhaps present throughout life. It has given rise to many misconceptions. The matrix of the dentine was supposed by Waldeyer blasts from c?w°Em: *« ^^ produced by a metamorphosis of the proto- B ^°of A' cL ^ Atiki P^^^™ °^ *.^® odontoblasts, but this point is open Franz Boll. ' to discussion. The question is part of the more general one— What is the origin of intercellular substance? Compare p. 399. As the dentine increases in thickness the odontoblasts become longer and narrower. Fig. 333, B and the THE MOUTH CAVITY AND FACE. 589 dentinal processes finer, more numerous and branching, the branches anastomosing with one another. The processes persist and never calcify, the spaces they occupy being the dental canaliculi of the adult. The ends of the odontoblasts toward the dentine become, for the most part, as it were, squared off, while the lower ends become more or less pointed. Fig. 333. The odontoblasts lose much of their regularity of arrangement, as the dentine nears completion, but they are still found in the adult. In old age they become comparatively inconspicuous and assume a rounded or ovoid shape (Tomes' " Den- tal Anat.," 1876, p. 97). The cement is merely a layer of bone developed by ossification of the dental follicle over the root of the tooth. It differs from ordinary hone by the greater abundance of Sharpey's fibres in it. Its develop- ment begins on the milk-teeth during the fifth month, and takes place after the type of periosteal ossification. Age of Development. — The following table indicates approx- imately the ages at which the various stages of development are passed bj- the different teeth. To complete the table it must be added, 1, that the first permanent molar arises the fifteenth or six- teenth week like a milk-tooth as a bud from the epithelium of the dental groove; 2, that the second molar begins as a bud from the neck of the first about the third month after birth, and, 3, that, ac- cording to Magitot, the germ of the third molar, or wisdom-tooth, begins as an enamel bud from the neck of the second molar, about the third year (C. S. Tomes, "Dental Anat.," 1876, p. 128.) Age, Weets. Jlilk teeth. Permanent teeth (except molars). First molars. 7th 8th 9th 10th 15th Dental groove and ridge. Enamel organs bud. Enamel organ concaves. Follicular wall. J Enamel organ fully differen- 1 tiated I Follicle closes above germ. { Neck of enamel organ re- ( sorbed 1 Dentine appears on incisors J Enamel bud 1 appears. 16th irth Papilla. 18th i Dentine appears on first and Follicle. 20th 26th 28th S2d 36th 39th Dentine caps, 0.(VM).06 in. high. 0.05-0.07 0.08-0.09 0.09-0.11 O.n-O.l'J 0.12-0.14 Papilla formed Follicle closes. ) Enamel organ fully differen- 1 tiated ; follicle well formed. Follicle closes above germ. Cusps coalesce. Aftpr hirth Enamel and dentine appear. Double Dentition of Mammals. — The manner in which the teeth are renewed in the shark's jaw has been described, p. 582, Fig. 328 ; the new tooth-germs arise as outgrowths on the lingual side of the old. In mammals there is the same relation between the earlier milk-teeth and the later permanent teeth. It is, therefore, justifiable to assume that the diphyodont mammal preserves in a reduced degree the piscian power of renewing the teeth, and that the milk-teeth rep- resent the primary dentition. Such, however, is not the view of Flower, 67. 1, who considers that the present mammals are derived 590 THE PCETUS. from monophyodont ancestors, and have acquired the milk-teeth secondarily by interpolation. This conception has been more recently adopted and defended by Oldfield Thomas (Phil. Trans., 1887, 451). For criticisms of these authors see Lataste, 89. 1, who also advances a more complicated hypothesis. Flower's hypothesis was based on the belief that marsupials, which have only one set of teeth, possess the permanent set, but W. Kiikenthal, 91.1, has found that the teeth of Didelphys (opossum) correspond to the milk-teeth, and that the germs of the permanent teeth are present in the embryo and abort without forming any tooth except the third praemolar (so-called first molar) of the upper jaw, which belongs to the second dentition. As to the evolution of the complicated forms assumed by the teeth of mammaHa, see E. D. Cope, 74.1, and H. F. Osbom, 88.2, Salivary Glands. — The mouth cavity of amniotes is furnished with numerous glands, which in Sauropsida are found in part vari- ously gathered into groups, in part scattered singly. In mammals scattered single glands are found, but instead of groups of glands there are three pairs of large glands, each with a long single duct. The three pairs are the salivary glands and are known only in mam- mals. It has been suggested that each salivary gland corresponds to a group of oral glands in reptiles, but the attempts to determine the homologies involved in this assumption have failed, compare Reichel, 83.1, and Ercole Giaccomini, 90.1. On the other hand the development, I believe, indicates clearly that each salivary gland is a single oral gland greatly enlarged, for it arises from a single invagination and in an early stage has a marked resemblance to an ordinary branching gland of the mouth. Concerning the development of the small oral glands in man, a few observations are recorded by KoUiker (" Mikrosk. Anat.," II., 2, and "Entwickelungsges.," 1879, 828) who also gives a few data con- cerning the sali varies. The development of the latter glands is known to us chiefly through the researches of J. H. Chiewitz, 85. 1. The glands appear in the following order : submaxillary, sublingual, parotid. The submaxillary anlage can be seen in a pig embryo of 21 mm. and in a human enibrj^o of about six weeks; the parotid appears in man ^ by the end of the eighth week. As to the position of the anlages : the mouth at the time they appear has a character- istic shape in section. Fig. 325, being — if we imagine the tongue removed — -like an inverted j_, and there is at each side an angle, a; it is from the epithelium along this angle that the solid outgrowth to form the parotis takes place. The base of the tongue forms an angle on each side with the floor of the mouth. Fig. 325 ; it is from this angle that the solid outgrowths of the buccal epithelium take place to form the sublingual and submaxillary glands, the former near the front, the latter near the back of the tongtie. The anlages of the parotid and submaxillary are at first at about the same dis- tance back from the frenulum of the tongue, but as development proceeds the submaxillary orifice migrates forward, the parotid back- ward. The following measurements are from Chiewitz, 85.1 422. Age of embryo in weeks Submaxillary gland, distance from frenulum. . Parotid gland 6 8 10 12 u.5a 32 0.36 0.13 mm. 0.34 1.08 1.10 mm. THE MOUTH CAVITY AND PACE. 591 The outgrowth of the salivary anlage is at first a cylinder, which, however, soon begins to lengthen and branch; the ends of the branches enlarge, and ultimately develop into the alveoli. The gland is now further characterized by 'the condensation of the con- nective tissue about its branches into a globular mass, which is sharply defined, Fig. 334, a, against the neighboring looser con- Alv. Fig. 834. Days. -Section of the Submaxillary Gland of a Human Embryo of sixty-three to sixty -eight Minot Coll. No. 138. Alv^ Alveolus; o, connective-tissue sheath of gland; Z>, duct. nective tissue. The lumen of the gland appedrs first in the main duct, then in its branches, and, last of all, in the alveoli ; it develops, not by the abortion of the cells in the centre, but by the cells moving asunder so as to leave a central cavity, while they themselves assume an epithelial arrangement. The alveoli are still solid at the begin- ning of the fifth month, but in an embryo of twenty-two weeks were found by Chiewitz, I.e., 427, to be all hollow. At this time the epi- thelium consists of a single layer of cylinder cells; in the ducts the nuclei are so placed that they form, as in earlier stages also. Fig. 334, Z), two rows; the nuclei of the outer row are somewhat smaller and stain more readily than those of the inner row ; in the alveoli the cells are at first all alike, but after the alveoli become hollow some of the cells become enlarged to form muciparous beaker- cells. 592 THE FCETUS. while others remain smaller and protoplasmatic; these smaller cells become partly covered in by the neighboring beaker-cells, and thus develop into the semilunar cells of the adult. Between the anlages of the- sublingual and submaxillary glands, there appear later— twelfth week in man— some eleven to thirteen gland anlages, which in their mode of development resemble small salivary glands, Chiewitz, 85.1, 423. These are termed by Chie- witz alveolingual glands, and have been often confounded with the true sublingual gland. Tongue. — Although the tongue is developed from the floor of the pharynx, yet it becomes so entirely an appendage of the mouth that it may be appropriately treated here. Our knowledge of the devel- opment of the tongue is derived chiefly from Dursy, 69. 1, and His, ("Anat. menschl. Embryonen," III., 64-81). The first distinct trace of the tongue is a small tubercle which appears in the middle line on the floor of the pharynx between the ends of the first and second (i.e., mandibular and hyoid) arches. It was supposed by Dursy to be formed by the fusion of the lower ends of the mandibular arches, but His has shown that it is single and median, and accordingly has termed it tuberculum impar,¥ig. 177. Immediately behind the tu- bercle appears the evagination to form the thyroid gland, see Chap- ter XXIX. Very soon after the tubercle has appeared the lower ends of the second and third arches fuse — human embrj-os of 7 nam. — and their fused ends constitute the anlages of the back of the tongue. Tn.tk Fig. 385.— Reconstruction of the Pharynx of a Human Embryo (His' Pr, 10 mm.). Tg, Tongue; I, H^ ni, IV, branchial arches; Si, sinus cervicalis; Vg, vagus nerve; Epg, epi- ,^,,-71 n • m i glottis; m.«A, median thyroid aniage. After The tubercie now rapidly enlarges, W.His. x24diams. p.g_ ggg^ y^^ ^^^ becomes easily recognizable as the front part of the tongue. The site of the thyroid evagination remains as a fixed point, which is often marked by a small depression, the foramen caecum of Morgagni ; the duct of the thyroid sometimes persists and is then found starting from the fora- men coecum. The front and back of the tongue are marked off. Fig. 335, by two oblique lines, which start from the foramen, and together form a widely open V. This Y can be traced — -as pointed out by His, I.e., 79 — in the adult tongue; the part behind the V has its surface thrown into ridges, and over it there are glands, which appear dur- ing the third month ; the part in front has papillae developed under its epithelium, a.nd the papillae circumvallatee are situated a little (5-8 mm.) in front of the V, but in lines parallel with it ; the cireum- vaUate papiUse do not, therefore, represent the division line between the front and back of the tongue. The largest part of the tongue is developed from the tuberculum impar, the less part from the region of the second and third branchial arches— hence the tongue is a derivative of the pharynx and not of the oral cavity. CHAPTER XXVII. THE NERVOUS SYSTEM. The formation of the vertebrate cerebrospinal axis has already been treated at length, pp. 173-181. In its first stage it appears as the medullary tube with ectodermal walls. The second stage is the dif- ferentiation of the brain from the spinal cord by the enlargement of the anterior end of the tube. The sharp distinction which we have just drawn ^between the stages does not maintain itself in the am- niota. In fact the medullary groove widens at its cephalic end before it closes to form a tube, so that the brain is indicated in the embryo before the medullary tube is formed. Moreover the development of the brain progresses while the groove is closing, so that the brain is already quite advanced before the medullary tube is closed at its caudal end. These irregularities in the development of the central nervous system render it impossible to decide at present whether the simple medullary tube without a brain enlargement, or a (perhaps solid) central nervous system with a brain enlargement, represents the phylogenetically primitive condition. The difficulty of reaching a decision is still further increased by the fact that the tubular con- dition of the nervous system was probably acquired within the verte- brate series, see p. 180. Definition of the Brain. — ^The vertebrate brain is the anterior portion of the medullary tube, and is characterized by two primary features: 1, the enlargement of the tube; 3, its special associations with higher sense organs (olfactory, visual, and auditory). The brain is further characterized in all true vertebrates : 1, by having three principal enlargements separated from one another by two con- strictions (H. Ayers, 90.1, claims that the three enlargements can be traced in Amphioxus also) ; 3, by being bent at the region of the second enlargement (mid-brain) owing to the development of the head-bend of the embryo ; 3, by containing the principal centres for the co-ordination of sensations and movements.. All modifications of the brain can be traced back to this primitive type, and it seems probable that the evolution of the brain has been dominated by the advantages of more perfect co-ordinating apparatus, as the special senses on the one hand and the locomotive organization on the other acquired a higher development. Cerebral Vesicles. — The enlargement which produces the brain extends about half the length of the embryo, compare Figs. 114 and 155, and takes place unevenly, so that there are produced three suc- cessive lobes, which are known as the primary cerebral vesicles, Fig. 113 and 114; the second and third vesicles (mid-brain and hind- brain) are often imperfectly divided from one another. The three vesicles subsequently subdivide, so as to form — to follow the tradi- 38 694 THE PCETUS. tional description — five secondary vesicles. It has long been cus- tomary to describe the medulla as dilating to form the three and later five vesicles, but unfortunately the descriptions have been so much conventionalized in subservience to tradition that they are misleading in several important respects. The attempt is here made to give an untrammelled objective account. Optic Evaginations. — The first indication of brain formation seems to me to be the widening of the extreme anterior end of the medullary plate or groove, which can be recognized in all vertebrate embryos at a very early stage. In elasmobranchs it appears to me evident that the widening is due to the very process of concrescence itself, and is initiated while the ectental lines are approaching one another, and is fully marked before the longitudinal axis of the embryo is completed by concrescence. Fig. 317 represents a dog-fish embryo ; in is the point at which concrescence has begun ; it will be observed that the embryonic rim curves around this point and in consequence is spread out laterally; in later stages the lateral protrusion, which we see initiated in Fig. 317, at in, becomes stiU more marked and can be followed until it is evidently the optic diverticulum. In mammals we find the meduUary groove specially widened at its anterior end — noticeably so in the mole, Fig. 99, op. A cross section through the optic vesicle at this stage offers a very singular appearance. Fig. 100; the entoderm. En, has not closed over, although the notochord, nch, is already distinguishable under the medul- lary groove; the ectoderm, Ec, is greatly thickened on the dorsal side to form the very wide medullary plate, which has a median depression, Mp, corresponding to the medul- lary groove proper, and two lateral depres- sions corresponding each to an optic vesicle. If we imagine the medullary plate to bend upward and to close over itself, then the two edges of the optic depressions, op, which are outerrnost in Fig. 100, wiU meet in the me- dian line, and as soon as the groove, by clos- ing, becomes a tube, there will be at this point two lateral diverticula, having the same char- acteristically thickened ectodermal lining as the rest of the medullary tube. These diver- ticula are the so-called optic vesicles, which are ultimately transformed into the optic nerve, retina, and choroid of the eye. In the chick the optic vesicles become clear- ly indicated by the twenty-fourth hour, when there are from five to seven distinct pairs of primitive segments, and the head projects shghtly over the proamniotic area. Before the medullary groove has closed anywhere the optic diverticula are quite distinct. In a Fig. 336.— Chick Embryo of twenty-nine Hours, op. Optic Tesicle; pro, proamnion; V^, second cerebral vesicle or mid- brain; F3, hind -brain; v.om, omphalo - mesaraic vein ; seg, primitive segment; Afd, me- dullary tube; pr.g, primitive groove. After M. Duval. THE NERVOUS SYSTEM. 595 chick of twenty-nine hours, Fig. 336, the vesicles, op, are very large, their growth being an important factor in the precocious distention of the head. Widening of the Medullary Tube.— While the optic vesi- cles are developing the medullary tube expands in diameter through- out its cranial or anterior half, without there being at first much change in the structure of its walls or much evidence of subdivision, but very soon the expansion becomes unequal, so that the tube is slightly constricted immediately behind the optic vesicles. Fig. 336, op; then follows a slight dilatation, V\ the mid-brain (Mittelhirn) , which is separated by a second constriction from the long and large hind-brain, F' (Hinterhirn) , which is widest in front and gradually diminishes in diameter, and merges without distinct boundary into the posterior unexpanded portion of the medullary tube or future spinal cord. Transverse sections show that the widening, by which the brain is differentiated from the cord, is due chiefly to the en- largement of the medullary cavity, a'nd that the walls change but little in thickness until the three vesicles are differentiated, when the walls begin a series of characteristic modifications. The three primary vesicles {Gehirnbldschen, vesiculce cere- brates) were known to Malpighi and Haller according to Tiedemann, 61.1, 9. Bischoff, 45.1, 170, appears to have been the first to observe that they are formed before the medullary groove is entirely closed in the cephalic region. Owing to the fact that the optic ves- icles grow out so early and that the remainder of the brain as a whole widens out, we ought, ■ perhaps, to accept A. Goette's view, 75.1, 280, that a double division precedes the triple. In this case we should have to describe the mid-brain and hind-brain as arising by the subdivision of the second primary enlargement. 1. The Fore-Brain. — As we have seen above, the fore-brain orig- inally includes the optic vesicles, which primitivelj' show no trace of any de- marcation from the central portion of the fore-brain. This condition, how- ever, does not last long, for the central portion of the fore-brain soon begins to expand upward and forward, making a separate central enlargement, which may be designated as the permanent fore-brain. Meanwhile the distal ends of the optic diverticula also dilate rap- idly, while the part of each diverticulum nearest the fore -brain proper grows slowly. It is often erroneously stated that part of the optic vesicle is con- ../■^■i . TJ.-J. 1 i.u ^v. Fig. 337.— Cross Section through the Stricted: m reality it enlarges, though Fore-Brain and optic vesicles of a relatively slowly. From these modifi- k^P'^ttbrlin'^rlrectode'^'^'i^^Sl cations there are developed a wide me- laKeof lens; op, optic vesicle; jsi, en- j. J. ,. JJ.1J. 1 J.- toderm. After Balfour and Parker. dian rore-bram and two lateral optic vesicles connected by tubular stalks with the ventral side of the brain proper. Fig. 337.* In short, the primitive vesicle is divided into three * Compare also Figs. 170, 171, and 179. 596 THE FCETXJS. parts, one median and two lateral, and it is only the median part that enters into the formation of the brain. The history of the median division is, therefore, treated in this chapter, while that of the two lateral divisions is dealt with in Chapter XXVIII., on the organs of sense. It may, however, be stated now, in order to facilitate the comprehension of the figures, that the optic vesicles expand dorsr.l- ward. Fig. 337, op. It shoiild be noted that the walls of the fore- brain and optic vesicle are still nearly uniform in thickness, and, so far as yet observed, in structure. The changes described in this paragraph occur in the chick at about thirty-two to forty hours, in the rabbit the ninth day, in man about the eighteenth day. The next series of changes in the fore-brain lead to the differentia- tion of the cerebral hemispheres. By a long-continued tradition it has become customary to describe the process as the subdivision of the primary vesicle into two secondary vesicles, designated as the fore-brain proper {Vorderhirn, prosencephalon) and 'tween-brain (Zwischenhirn, tlialamencephalon) . Such a description, however, seems to me hardly justified either by embryology or comparative anatomy, and to be especially apt to mislead and confuse. In fact every embryologist must admit that , ' ^^ it is scarcely correct to say that the ^"^"^ ''' (-.|^ fore-brain divides into two vesicles, from the anterior of which the cere- bral hemispheres grow out. It is more in accord with the actual facts to describe the hemispheres as ap- pendages of the fore-brain, that is to say, of the so-called Zwischenhirn \ or thalamencephalon. Accordingly ]^ I present the history of the origin of '• i,^^ the cerebral hemispheres somewhat I ;jj differently from usual, though, of j: ' ■:?f. course, without changing the facts. Sp.c I ;j^ For convenience I defer mention of i, ;^ the head-bend (see p. 600), which Fig. m-Brain of Embryo No. 22, p. 297 develops while the hemispheres are <^His'Lg). H.Aniage of hemispheres; iif 6, arising. Soon after the optic vcsi- mid-bram ; 06, future cerebellum ;£», epen- i i ° i , ti -, ,-, '^ , dyma; ot, auditory vesicle; Sp.cf spinal cies have become Stalked the extreme ^°diknS'' cEmp;r?Fig. A'!^'^' ^'^^ ^ anterior end of the first vesicle en- larges and pushes itself, so to speak, forward and, owing to the head-bend, downward. The flexure is at first slight, but increases as development proceeds, compare p; 600. The enlarged end of the medullary tube is in no way divided off from the first cerebral vesicle until the end begins to dilate to- ward each side to produce the hemispheres. The manner in which the hemispheres grow out can be better understood from the Figs. 338, 339, and 340, than from any mere description. At first, as just indicated, they form an undivided common anterior enlargement, but the lateral expansion begins very early, and with it the anlages of the two hemispheres are given. If the position of the hemispheres is observed carefully, Fig. 338, H, it will be seen at once that it is the product of the dorsal side, and that the ventral half of the primi- THE NBRVOITS SYSTEM. 597 ^'W> /:? W- fb. w 01 Fig. 339.— Reconstruction of the Brain ot His' Embryo Ko CNackenldnge, 10.2 mm.). M6, Mid-brain; /6, fore-brain; iJ, hemisphere; 01, olfactory lobe; Op, optic nerve; hy, hypophy- sis ; cm, corpus mammilare. After W. His. tive fore-brain, as shown by W. His, 89.4, does not participate in the outgrowth. The consideration of this important fact demon- strates that the hemispheres cannot be strictly compared with one of the primary vesicles, each of which includes a ventral as well as a dorsal portion of the medullary tube. The origin of the hemispheres from the dorsal side __ ,„: — , has so great impor- tance morphologically that special emphasis must be laid upon the .. ,,„,, fact. The ventral ^'^i- .:iMmiSMMMriW :::}:m:B :■ H. boundary of the hemi- spheres must be placed near the optic stalks, so that the hemi- spheres include that _ _,_ , , , ... ..^, , portion of the brain ' c.m '"'''■'-^:„ 'x v:^— -Op wall which unites with ^y' the ectoderm to form the olfactory plate, al- ready described, p. 575. The cerebral hemispheres grow more rapidly than any other part of the brain, see Fig. 339, H, but their growth is principally in their distal parts, so that, like the optic vesicles, they become large pouches connected by relatively smaU hollow stalks with the fore-brain. The stalk is short. The passage through the stalk is called the foramen of Munroe, Fig. 340, f.in. As this foramen enlarges but little, while the brain increases enormously, it appears in the adult as a small opening in proportion to the size of the whole brain. Although the foramen enlarges absolute^, it is sometimes described errone- ously as becoming smaller during development. While the hemi- spheres are expanding the olfactory plate, Fig. 339, 01, acquires a more marked differ- entiation beneath them Lind shows traces of di- vision into a dorsal or anterior, and ventral or posterior, lobe. Even at the stage of Fig. 339, it can still be recognized that the ol- factory region corre- sponds to what was, before the brain was bent, part of the ex- treme anterior wall of the fore-brain. But the olfactory region is already paired, and is associated in its development with the hemispheres. This leaves a part of the wall of the fore-brain in the median line. Fig. 340 (between the reference lines /.m and E.o), which is known as the lamina terminalis and represents throughout Fig. 840.— Reconstructed Median View of the Fore-Bram of His' Embryo Ko iNackenlange, 10.2 mm.). H. Hemisphere; f.m, foramen of Munroe; B.o, recessus opticus; t.c, tuber cine- reum ; m, corpus mammilare ; mJ>, mid-brain. After W. His. 598 THE FCETUS. life the extreme anterior wall of the fore-brain. As seen in Fig. 340, it extends from the level of the foramen of Munroe to the level of the optic stalks. In the same figure it can also be seen that the hemispheres and olfactory lobe project further forward than the lamina. The hemispheres expand, not only upward and forward in regard to the longitudinal axis of the fore-brain, but also back- ward, as can be well seen in Figs. 339 and 341. The history of the hemispheres is given more fully and for later stages below, p. 690. The primary differentiations of the floor or ventral wall of the fore-brain are also clearly indicated in a human embryo of 10-13 mm. {Nackenlange) , Figs. 339 and 341. The lower part _ of the fore-brain has expanded, forming, as it were, a hangmg pouch, Fig. 339, from which pass off the optic stalks. Op. Follow- ing the median wall of the pouch around from the mid- brain to the level of the foramen of Monroe, Fig. 340, f.m, we find, _.^rs^, a pro- tuberance, m, which extends nearly half-way to the optic stalk, and indicates the future mammillary bodies ; second, a slight swelling, t.c, which marks the future tuber cinereum; third, the future apex of the infundibulum ; fourth, the area of the brain wall united with the hypophysis ; and fifth, the lamina terminalis, just beyond the recessus opticus, B.o. 2. The Mid-Brain. — The second cerebral vesicle undergoes less modification than the first and third. Its walls are at first of nearly uniform thickness, see Duval, "Atlas," Fig. 255. It is oval or round in transverse section. It is situated at the point where the head-bend takes place (compare p. 600), and by the head-bend its shape is profoundly altered, its dorsal surface becoming more arched and expanded, Fig. 338, Mb, while its ventral wall as seen in profile becomes concave ; further, the dorsal wall becomes relatively much thinner than the ventral wall. The cavity of the mid-brain remains very large, and during the early expansion of the brain the commu- nication between the fore-brain and mid-brain enlarges more than does the passage between the mid- and hind-brain. This is commonly expressed by saying that the constriction between the first and second cerebral vesicles is much less marked than between the second and third. In the lower vertebrates the fore-brain and hind-brain do not ad- vance either in growth or complication as in the amniota. In birds and reptiles the mid-brain develops to a greater extent than in mam- mals, and in the embryo early acquires great size, see Fig. 396, II. In mammals, on the other hand, the mid-brain grows more slowly. Roughly speaking, then, we may say that the importance of the mid- brain diminishes as we ascend the vertebrate series, and that it does not participate in the advance of organization which characterizes the first and third cerebral vesicles. 3. Hind-Brain. — The third cerebral vesicle is especially charac- terized by the great expansion of its very thin dorsal wall, by the thickening of the dorsal wall immediately behind the constriction separating the second from the third vesicle, and by the great and prominent bend formed by the ventral wall of the hind-brain. Fig. 341, Hb. The thin dorsal wall corresponds to the epithelial epen- dyma of the adult ; its morphological significance is explained in the THE NERVOUS SYSTEM. 599 section on the zones of His, p. 606. The dorsal thickening is the aniage of the cerebellum and corresponds to a commissure found in the lower vertebrates. The apex of the ventral flexure is the aniage of the pons Varolii of the adult. The thickened floor of the hind-brain, between the pons and the spinal cord, sp.c, gives rise to the medulla oblongata. We thus have the four chief struct- ures, which develop from the hind-brain, definitely mapped out by the earliest changes. The modifications which result in this four-fold difi'erentiation all take place simultaneously \r "'^ I and are interdependent. They ' ' ^" QTd +1-10 ToanH- nf +xlrr^ fo/nf/-^T-c . 1 ^"'- 341-— Brain o( a Human Embryo of five- aietne result OI two tactors. i, weeks CHIs' Embryo Sch). fi^^ Hemisphere; i?, the unequal develonment of dif- olfactory lobe; Op, optic nerve; ilift mid-brain; , , ^ . ,. .i- - ,, i/6, limd-bram ; Sp. c, spmal cord. After W. His. rerent regions of the medullary walls ;^2, the appearance of the Varolian bend {Bruckenkriimmung) . These factors are considered later. It is usually stated that the hind-brain subdivides into two vesi- cles, for which the names secondary hind-brain and after-brain (Nachhirn) have been employed; the Nachhirn is the part nearest the spinal cord. In fact, it is convenient to designate the anterior part of the hind-brain, out of which the cerebellum and pons Varolii arise, as the hind-brain proper (metencephalon) and the posterior part as the Nachhirn {epencephalon or myelenceplialon) or, better, as the medulla oblongata. On the other hand, it is incorrect to speak of the primitive hind-brain as forming two secondary vesicles. This error goes back to the time of Von Baer, II., 106, who observed such division in the chicken embryo. It has also been described and fig- ured by Mihalkovics, 77.1, 25, Taf. IV., Fig. 33, in a chick of fifty- eight hours. These authors, and most others who have written on the subject, assumed that their observations were upon a constant and typical condition. In reality there is great irregularity in the growth of the walls of the hind-brain, and sometimes in birds and per- haps in reptiles the third cerebral vesicle is temporarily more dilated at its anterior end than elsewhere. The dilatation soon disappears, and no proof has been brought yet, to my knowledge, to establish an identity between it and the region corresponding to the cerebellum and pons — it seems to take in more than the cerebellum, less than the pons. In chicken embryos the separate dilatation is usually wanting, and it has, so far as I know, never been observed in any mammalian or ichthyopsidan embryo. It is interesting to note that Balfour, "Comp. Embryol.," II., 424, though he doe| not expressly mention the error of the traditional description, yet skilfully avoids adopting it in his account of the hind-brain. The shape of the hind-brain requires more detailed description. As seen in Fig. 338, the hind-brain at the time of the development of the head-bend is more than equal to all the rest of the brain in length. It begins with the constriction or isthmus behind the mid- 600 THE PCETUS. brain and at first widens rapidly, tlien gradually tapers to the neck- bend, where it passes into the spinal cord. Viewed from the dorsal side Fig. 343, the anterior constriction or isthmus is still more no- ticeable, and we can also see the kite-shaped outhne of the thin roof. Comparison of the figure with the following. Fig. 343, represent- ing a slightly older stage, affords an idea of the widening of the me- dulla, while comparison of Figs. 338 and 341 will indicate its modi- fications as seen in profile. Itis important to observe that there is, as yet, no cerebellum, but only a thickening of the dorsal wall close to the isthmus. This thickening is the anlage of the cerebellum, and is to be homologized with the commissure found in the corre- sponding position in Ichthyopsida. Cerebral Flexures. — The axis of the neuron may be described as Brain of a Human Em- . -, , « . , V , ^l Dryo or one Montn Straight, for it is actually very csis' eu). m, Mid- the stage when '"■"°= ^"^^ ''^'■^'""- 343. —Dorsal of the Hind- nearly so, up to the optic vesicles begin to be con- tncie; sp-c, spinal stricted off— see Figs. 99 and 336. °'''^- ^^'^'^^■H'^- lum; rV, fourth ven- Fio. 342. —Hind-Brain of a Human Embryo (No.22, p.29r,His'Lg), seen from the dorsal side, mb. Mid-brain; C6, cerebellum ; IV, fourth ventricle ; Sp.c, spinal cord. After W. His. Compare Fig. 338. While the dilatation to form the second cerebral vesicle or mid-brain is taking place, the primary head-bend of the embryo is established, involving the brain. The bend of the brain takes place at the level of the mid-brain ; the fore-brain is bent over ventralward until it forms a right angle with the hind-brain. Fig. 338, the actual flexure being almost confined to the mid-brain, in which, as can be seen in the figure, the cerebral axis curves very much, while in the hind- brain it remains nearly straight, and in the fore-brain is slightly bowed only. This bend may be called the mid-brain or primary fiexure.* During the early stages of the hemispheral outgrowths the flexure increases until the axis of the fore-brain forms an acute angle with that of the hind-brain. Fig. 320. Mihalkovics, 77.1, 39, proposes to distinguish the right-angled stage as the Haken- kriimmung, and the later acute-angled stage as the Kofpbeuge. Such a distinction is entirely arbitrary, and the suggestion has not been adopted. The angle becomes ultimately so sharp that the floor of the fore-brain becomes nearly parallel with that of the hind-brain. The second bend to appear is at the junction of the hind-brain (medulla oblongata) and spinal cord, Fig. 338, and is termed the neck-bend {Nackenkriimmung) . Like the primary bend it affects the whole head ; the summit of its angle appears in the embryo when seen in profile, compare Figs. 220 and 223, during several early stages as a projection (His' Nackenliocker) , which is, however, soon obliterated. The neck-bend develops later than the head-bend, not appearing in mammals until the hemisphere anlages have begun * It is called by Reichert Oesichtskopfbeuge ; by Dursy, Kopfbeuge ; by KoUiker, vordere Kopfkrummung ; by His, Scheitelkfummung. THE NERVOUS SYSTEM. 601 to grow out separately. It is very slight in the Ichthyopsida ; in the reptiles and birds it is more developed, but it attains its maxi- mum only in the mammalia, and notably in man. In human embryos the neck-bend increases from the third to the end of the fifth week, when it reaches its maximum, the hind-brain then form- ing nearly a right angle with the spinal cord, Fig. 341. Later the bend becomes less again, owing to the gradual erection of the head as already described and illustrated in Chapter XVIII. for the human embryo. The third cerebral flexure is known as the Varolian bend (K5lli- ker's Briickenkriiinmung) and is essentially different from the two flexures just described, for it is not a bend of the whole medullary tube, as are they, but a bend of the ventral side of the hind-brain, Fig. 341, the dorsal side remaining as seen in profile, nearly straight. As already mentioned, the greater part of the dorsal wall of the hind- brain is a thin membrane, and this membrane takes no part in the formation of the Varolian bend, which depends on the growth of the thick walls of the floor of the hind-brain, and with this growth the bend increases, its formation being accompanied by the lateral ex- pansion of the hind-brain at its anterior or cerebellar end, Fig. 343. The cause of aU the cerebral flexures is, of course, the unequal growth of the various parts. Herein the growth of the brain is cer- tainly the principal factor in determining the result. The general conception of the influence of the unequal growth of the brain dates back to Von Baer, and was revived by Rathke. W. His was the first to attempt an analysis of the mechanical conditions, and to demonstrate that the shaping of the brain depends to a large degree upon these conditions, which are many of them relatively obvious and simple. His has given in his semi-popular work, "Unsere Korperform," 74.1, pp. 93-118, an admirable presentation of his re- sults, which have not yet received from embryologists the attention which their exceptional importance demands. Origin of the Sensory Ganglia. — To fully understand the history of the ganglia the reader should consult the section on the ganglionic sense-organs in the following chapter. The origin of the ganglia has been carefully traced in a human embryo * with thirteen segments, by M. von Lenhossek, 91.1, three of whose figures I reproduce. Fig. 344. As seen in A, the ectodermal cells, Gl, which immediately adjoin the medullary plate, differ in size and by their rounded form from the cells of the neighboring ectoderm and of the medulla. These cells constitute two bands, which unite in a single median band when the medullary groove closes. The median band has been termed the Zwischenstrang in the chick embryo by His, but is more usually termed the neural crest or ridge {Neuralleiste) , as proposed by Balfour. In B, the cells are about to unite in the median line. In C they have united, and though incorporated in the medulla and separated entirely from the external ectoderm are readily distinguished from the cells of the medullary plate proper. The cells are also growing out on each side, Gl, toward the myotome. The emigration continues until all the cells are transferred from the median line to the lateral masses, Gl, which are th e anlages of the * This embryo is the one designated as No. 18, and described p. 295. 6oa THE FCETUS. sensory ganglia. As the cells depart from the neural crest the medullary plate proper closes over in the median dorsal line. The ganglionic lateral masses exhibit a segmental arrangement very early, so that the cells ap- A .,