^v. ^ y „/ ^7# /'•'' / -if '' V CORNELL UNIVERSITY LIBRARY FROM The Institution Cornell universiiy i-iwiaty QP 91.R35 The differentiation and specif icily of c 3 1924 003 153 263 i'®/ ^^ 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/cu31924003153263 THE DIFFERENTIATION AND SPECIFICITY OF CORRESPONDING PROTEINS AND OTHER VITAL SUBSTANCES IN RELATION TO BIOLOGICAL CLASSIFICATION AND ORGANIC EVOLUTION: THE CRYSTALLOGRAPHY OF HEMOGLOBINS. BY EDWAED TYBON EEICHEET, M.D., Professor of Physiology in the University of Pennsylvania, AND AMOS PEASLEE BEOWF, Ph.D., Professor of Mineralogy and Geology in the Univerdty of Pennsylvania. WASHINGTON, D. C. PUBtlSPBai BT THE CABNEOIE INSTITUTION OP WASHINGTON 1909 OAENEGIE INSTITUTION ^WASHINGTON PtTBLIOATION NO. 116 PRESS OF J. B. UPPINCOTT COMPANY FHILASELPHIA PREFACE. This research was begun by me in October, 1902, and after considerable preliminary laboratory investigation I found that in the solution of my problems the crystallographic method promised at the present time to be the most likely to yield sa^factory results. Not being an authority in the science of crystaUoiKaphy, I associated with me in 1904 one of my colleagues, Professor Amos Peaslee Brown, upon whom has fallen especiaUy that por- tion of the work which demanded the services of an expert crystallographer. The trend of modern biological science seems to be irresistibly toward the explanation of all vital phenomena on a physico-chemical basis, and this movement has already brought about the development of a physico- chemical physiology, tk>B^ysico-chemical pathology, and a physico-chemical therapeutics. The striking parallelisms that have been shown to exist in the properties and reactions of colloidal and crystalloidal matter in vitro and in the Uving organism lead to the assumption that protoplasm may be looked upon as consisting essentially of an extremely complex solution of interacting and interdependent colloids and crystalloids, and therefore that the phenomena of life are manifestations of colloidal and crystalloidal inter- actions in a peculiarly organized solution. We imagine this solution to con- sist mainly of proteins with various organic and inorganic substances. The constant presence of protein, fat, carbohydrate, and inorganic salts, together with the existence of protein-fat, protein-carbohydrate, and protein-inorganic salt combinations, justifies the belief that not only such substances, but also such combinations, are absolutely essential to the existence of Ufe. The very important fact that the physical, nutritive, or toxic properties of given substances may be greatly altered by a very slight change in the arrangement of the atoms or groups of molecules may be assumed to be conclusive evidence that a trifling modification in the chemical constitu- tion of a vital substance may give rise to even a profound alteration in its physiological properties. This, coupled with the fact that differences in cen- tesimal composition have proved very inadequate to explain the differences in the phenomena of living matter, implies that a much greater degree of importance is to be attached to pecuUarities of chemical constitution than is universally recognized. The possibilities of an inconceivable number of constitutional differences in any given protein are instanced in the fact that the serum albumin mole- cule may, as has been estimated, have as many as 1,000 million stereoisomers. If we assume that serum globulin, myoalbumin, and other of the highest protein^ may have a similar number, and that the simpler proteins and the fat& and carbohydrates, and perhaps other complex organic substances, may each have only a fraction of this number, it can readily be conceived how, primarily by differences in chemical constitution 6f vital substances, and secondarily by differences in chemical composition, th6re might be brought Jy PREFACE. about all of those differences which serve to charactenze genera, species, and individuals. Furthermore, since the factors which give rise to consti- tutional changes in one vital substance would probably operate at the same time to cause related changes in certain others, the alterations in one may logically be assumed to serve as a common index of all. In accordance with the foregoing statement it can readily he under- stood how environment, for instance, might so affect the individual's metar bolic processes as to give rise to modifications of the constitutions of certain corresponding proteins and other vital molecules which, even though they be of too subtle a character for the chemist to detect by his present methods, may nevertheless be sufficient to cause not only physiological and morpho- logical differentiations in the individual, but also become manifested physio- logically and morphologically in the offspring. Furthermore, if the corresponding proteins and other complex organic structural units of the different forms of protoplasm are not identical in chemical constitution it would seem to follow, as a corollary, that the homologous organic metabohtes should have specific dependent differ- ences. If this be so it is obvious that such differences should constitute a preeminently important means of determining the structural and physio- logical peculiarities of protoplasm. It was such germinal thoughts that led to the present research, which I began upon the hypothesis that if it should be found that corresponding vital substances are not identical, the alterations in one would doubtless be associated with related changes in others, and that if definite relationships could be shown to exist between these differences and peculiarities of the Uving organism, a fundamental principle of the utmost importance would be estabhshed in the explanation of heredity ,Tnutations, the influences of food and environment, the differentiation of sex, and other great problems of biology, nonnal and pathological. To what extent this hypothesis is well- founded may be judged from this partial report of the results of our investigations: It has been conclu- sively shown not only that corresponding hemoglobins are not identical, but also that their peculiarities are of positive generic specificity, and even much more sensitive in their differentiations than the " zooprecipitin test. " More- over, it has been found that one can with some certainty predict by these peculiarities, without previous knowledge of the species from which the hemo- globins were derived, whether or not interbreeding is probable or possible, and also certain characteristics of habit, etc., as will be seen by the context. The .question of interbreeding has, for instance, seemed perfectly clear in the case of CanidoB and Murid(B, and no difficulty was experienced in forecasting similarities and dissimil-arities of habit in Scmndce, Muridoe, ^eltdoB, etc,,. not because hemoglobin is per §e the determining, factor, but because,^according to this hypothesis, it serves a^. ah index (gross though It be, with our present very limited knowledge) of those physico-chemical properties which serve du-ectly or indirectly to differentiate genera, species, and indmduals. In otherwords, vital peculiarities may bl resolved tea pnysico-chemical basis. _ , . .. . " , Edwabd Tyson Reichert. CONTENTS. PAGE Chaptbb I. The Distribution op Hemoglobin and Allied Substances in the Aniual Kingdom . . . 1-27 Statement of the Distribution 2 Histohematins and Myohematins 3 Echinochrome, Hemerythrin, and Chlorocruorin, etc 4 Respiratory Metal-free Colorless Proteins 6 The Distribution of Hemocyanin 7 The Distribution of Hemoglobin in the Invertebrates 15 Causes of the Peculiarities in the Distribution of Respiratory Pigments 18 The Source of Hemoglobin Probably in Chlorophyl 19 Chemical Nature of Typical Respiratory Substances, etc 20 The Non-identity of Hemocyanins 27 The Identity or Non-identity of Corresponding Respiratory Substances 27 Chapter II. Specificity of the Blood of Vertebrates m Relation to Zoological Distinction . . 29-66 The Quantity of Blood in Relation to Body-weight in Reference to Genera 29 The Specific Gravity of the Blood in Relation to Genera 35 The Alkalinity of the Blood in Relation to Genera 36 The Proportions of Sodium and Potassiiuu in the Blood, Serum, and Corpuscles in Relation to Genera 39 The Phosphoric Acid of the Ash of the Corpuscles in Relation to Genera 41 The Proportions of Cholesterin in the Serum and Corpuscles in Relation to Genera 42 The Proteins of the Serum in Relation to Genera 43 The Proteins of Muscle Plasma and Seeds in Relation- to Genera 45 The ZoSprecipitins and Phytoprecipitins and Immune Sera in Relation to Genera 47 The Specificity of the Blood in Zoological Differentiation as Shown by the Phenomena of' Coagulation 47 The Leucocytes of the Blood in Relation to Genera 48 The Proportion of Corpuscles to Senmi in Relation to Genera 50 The Blood Platelets in Relation to Genera 51 The Form of the Erythrocytes in Relation to Genera 51 The Number of Erythrocytes in Relation to Genera 52 Table 20. The Number of Erythrocytes per Cubic Millimeter in Relation to Genera 53 The Size of the Erythrocytes in Relation to Genera 54 Table 21. The Sizes of the Erythrocytes in Different Genera according to the Measurements of Gulliver, Wormley, Treadwell, Formad, Welcker, and Malassez 55 Gulliver's Micrometry of the Red Blood Corpuscles (Plate A) 58-59 Certain Properties of the Erythrocyte in Relation to Genera 59 The Percentages of Hemoglobin in the Dry -Erythrocytes in Relation to Genera 60 The Percentage of Hemoglobin in the Moist Erythrocytes- in Relation to Genera 61 The Percentage of Hemoglobin in the Whole Blood in Relation to Genera 61 General Consideration of the ZoSlogical Specificities of the Blood 63 Chapter III. Hemoglobin; its General Chemical and Physical Characters, and its Specifici- ties 67-82 Constituents and Relations^to the Other Constituents of the Erythrocytes 67 The Elementary Composition of Hemoglobin ..i ; 70 The Molecular Formula and Weight of Hemoglobin 74 The Solubility of Hemoglobin ;....;,. 75 The Quantity of Water of Crjrstallization 76 The Extinction Coefficients and Quotients 77 The Differences in the Decomposability of the Hemoglobin of Different Species 80 Is the Oxyhemoglobin of the Blood of any Individual a Single Substance? 82 Chapter IV. The Preparation and Study of Hemoglobin Crystals Previous to the Investi- gations OF Preyer. ; .........._ ,..,... ••• • •* 83-92 v yi CONTENTS. Chaptbb v. The Investigations op Peeteb on the Cbtstallogbapht of Hemoglobin 93-106 Processes Used by Preyer for Obtaining Ciystals in Large Quantities 94 Processes Given by Preyer for Obtaining Crystals in Snaall Quantities 98 The Forms and Systems of Crystallization of Hemoglobins 101 Table 31. Preyer's Table Showing the Source of Hemoglobin Crystals, Crystalline Form, Crystalline System, etc 103-106 Chapter VI. The Pbbpabation and Cbystau-ogbapht op Hemoglobins Since Pbeteb's Inves- tioationb 107-130 Ceapteb VII. Cbtstallogbapht op Hemoglobin in Relation to Species, Accobdino to Pbe- viOTTS Investigatobs, with Explanations op Vabious Contbadictobt State- ments, ETC 131-140 Ceapteb VIII. Methods pob Pebpabing, Examining, and Measubing Cbtstals op the Hemo- globins Emplotbd in this Resbabce 141-147 Methods for Preparing Crystals of Hemoglobin 141 The Value of the Crystallographic Method of Investigation 144 The Petrographical Microscope and its Use 145 Chaptbb IX. Cbtstallogbapht op the Hemoglobins op Pisces, Batrachia, and Refthia. . 149-160 Pisces 149 Oxyhemoglobin of Raia bevis (Barndoor Skate) 149 Oxyhemoglobin of Acipenser sturio (Sturgeon) 150 Oxyhemoglobin of Alosa sapidissima (Shad) 152 Metoxyhemoglobin of Alosa sapidissima (Shad) 153 Reduced Hemoglobin of Alosa sapidissima (Shad) 154 Methemoglobin of Alosa sapidissima (Shad) 155 Metoxyhemoglobin of Cypriniis carpio (Carp) 166 Redu(^ Hemoglobin of Cyprinus carpio (Carp) 166 Batrachia 157 Oxyhemoglobin of Necturus maculatus (Necturus) 157 Reduced Hemoglobin of Necturus maculatus (Necturus) 158 Reptilia J58 o-Oxyhemoglobin of Python molurus (Indian Python) 159 ^-Oxyhemoglobin of Python molurus (Indian F^ihon) 160 Chaptbb X. Cbtstallogbapht op the Hemoglobins op Avbs 161-171 Aves jg» Methemoglobin (7) of Struthio camelus (African Ostrich) 162 Oxyhemoglobin of Casuarius galeatus (Cassowary) 162 Oxyhemoglobin of Anser anser (Goose) .'.'.'.'.'.'.'...... 163 Oxyhemoglobin of Olor buccinator (Trumpeter Swan) '.'.'...'....'....'..'..'..'.... 164 Oxyhemoglobin of Olor columbianus (Whistling Swan) .' 154 Oxyhemoglobin of Gallus domestica (Chicken) 155 Oxyhemoglobin of Colinus virginianus (Quail) !.!.'.!!!'...... 166 Oxyhemoglobin of Numida meleagris (Guinearfowl) .....'. I67 o-Oxyhemoglobin of Columba livia var. (Carrier I^wm) 168 Metoxyhemoglobin of Columba livia var. (Carrier Pigeon) 168 Reduced Hemoglobin of Columba Uvia var. (Carrier Pigeon) iao ^-Oxyhemoglobin of Columba Bvia var. (Carrier ^eon) ifio Oxyhemoglobin °"^*f • • \ snort diameter. A-P'^-- {te'SSer: Cryptobranchus Jap{l-g/^-^,- ; S«^--der {fe'iTmlL^.: gj f long diameter. . ishort diameter. Pisces: Trout ■['?fi'^>°'^*f- (.short diameter. Perch / '°°S diameter . \ short diameter. ViVa /long diameter. (.short diameter £gj I long diameter. I short diameter Lamprey* gpjg f long diameter . I short diameter Skate / '?.°1'^>™^*?'' • (.short diameter Torpedo i long diameter. "^ I short diameter Searhoise* 30.0 19.6 69.8 4L4 70.9 40.9 29.3 19.S 51.2 31.7 37.8 23.8 41.0 29.8 16.7 10.3 12.1 9.0 12.7 7.1 14.6 8.9 9.0 15.0 12.0 9.0 25.0 14.0 27.0 20.0. 15.0 * Circular. The figures recorded by different observers (Gram, Fortschr. d. Medicin, 1884, II, 33; Georgopulus, Zeit. f. klin. Medicin, 1906, lxiii, 322; White and Treadwell, Reference Handbook of the Medical Sciences, 1901, ii, 84; Gulli- ver, Proc. Zoolog. Society, London, 1875, 474; Wormley, Microchemistry of Poisons, 2d ed., Phila., 1888; Welcker, Zeit. f. rat. Medicin, Ser. 3, 1863, XX, 257; Malassez, Compt. rend. Acad. d. Sciences, 1872, lxxv, 1528; Formad, Comparative Studies of Mammalian Blood, Phila., 1888, and Journal of Comparative Medicine and Surgery, July, 1888, etc.) in their studies of given species differ in many instances. As a rule, Wormley's figures are somewhat higher than GuUiver's, while Formad's and Tread- well's are lower. These differences are not of importance, since on the whole they are remarkably close and entirely in accord in their indications of generic peculiarities. Human corpuscles have been more thoroughly studied than those of any other species. The extreme limits of measurements probably lie within 4 to 10 [I, but the ordinary range may be placed at about 6 to 9.5 ^. The mean is from 7.9 to 8 [i. The remarkably large proportion that measure close to the average is shown by the figures of Gram, Georgopulus, White, and others: Gram found that 82 per cent were of about the average meas- urements, 13 per cent larger, and 5 per cent smaller; Georgopulus records 73 per cent between 7 and 7.5 (i, 10 per cent between 8 and 8.5 /m, and 17 per cent between 6 and 6.5 {i; and White, 79.5 per cent between 7.5 and 8.5 (I, 12 per cent between 8.5 and 9.25 (i, and 8.5 per cent between 6.25 and 7.5 [i. IN RELATION TO ZOOLOGICAL DISTINCTION. 57 The diameters of the erythrocytes of different species have been made the subject of study especially by Gulliver, Wormley, Schmidt, Welcker, Malassez, Formad, and Treadwell. Gulliver's investigations extended over a period of 35 years and included studies of about 650 species. He frankly states that his tables can not pretend to absolute exactness, and are only offered for what they may be worth, and that in the estimation of their value allowance should be made for errors, whether instrumental or personal, more or less inevitable, notwithstanding the greatest care, in observations so extensive, and that the relative value of the measurements, though probably not unexceptionable, may be entitled to more confidence as fair approximation to the truth. He further states that in spite of little mis- takes or of variations in the dimensions of the corpuscles of this or that species, the comparative results will appear sufficiently uniform. Gulliver's measurements are so closely in accord with those of later observers that they are to be accepted as being sufficiently accurate to serve for purposes of comparison. His investigations were made from the point of view of the biologist, and he claims that the differences in the measurements of the erytlu-ocytes of different species constitute an important means of zoo- logical distinction. Thus, he states: If we compare the red corpuscles of species of one order or family, e. g., Tragulus and other ruminants, the corpuscles of the former animals will constantly prove the smallest; so, too, in Paradoxurus and Canis, in Hippopotamus and Elephas, in Mus and Hydrochoerus, in Dasypus villosus and Orycteropus capensis, in Rhea americana and Casuarius, in Zootica vivipara and Anguis fragilis, in Bufo viridis and Bufo vulgaris, in Osmerus eperlanus and Salmo salar. And in like manner the facts are equally clear in comparison of the different orders, so that the corpuscles are smaller in the Rumir nantia than in the Rodentia, in the Marsupialia than in the Edentata, in the Graminivora than in Rapaces, in Anura than in Urodela, in Sturiones than in Plagiostomi. Notwithstanding the foregoing positive statement, there seems to be a general, if not universal, belief that the size of the corpuscles is without much zoological importance, which is indicated by the very infrequent, casual, and scanty references to this subject. There is no doubt that the figures show unequivocally that the mean diameters of the erythrocytes of different genera, related or unrelated, may be practically the same, as, for instance, those of the monkey, hpped bear, hyena, and rhinoceros; and again, those of man, opossum, dingo, dog, wolf, whale, armadillo, beaver, capybara, guinea-pig, muskrat, etc. Nevertheless, it is clear that even among the members of a given order, or tribe, or genus, etc., the differences may be sufficient to be positively distinctive, and at times to have some other and more special zoological significance. Thus, in the primates there is seemingly an increase in the size of the corpuscles as the individual is higher in the scale of life, as, for instance, man, 7.9 (i; chim- panzee, 7Aix; monkey, 7.1 [i; lemur, 6.4 /tf. This relationship may be in rela- tion to differences in the sizes of the species (page 58) . Another interesting relationship is noted in the horse, mule, and ass, the mule being a hybrid and the corpuscles having an intermediate measurement. Then again, comparing representatives of classes of different orders, such as ruminants, felines, canines, etc., not only may each class be readily distinguished 58 SPECIFICITY OF THE BLOOD OF VERTEBRATES from the others by the mean sizes of the corpuscles, but even individuals belonging to each class. (Plate A.) The nearness of the diameters of the erythrocytes of certain of the domesticated animals to those of man is a matter of considerable impor- tance, chiefly because of its medico-legal bearing, and it is yet an open question if the corpuscles of the dog, and especially of the guinea-pig, can with positiveness be distinguished from those of man. The mean measurements found by White (expressed in [i) are: man 8.01, guinearpig 7.47, dog 6.87, pig 6.07, ox 5.44, sheep 4.75, and goat 3.69. Not only are these measurements sufficiently different to be significant, but the variations in the ranges in the sizes in the different species are peculiar. Particularly striking is the wide range in the pig and the narrow range in the goat, the limits of the former being 3.75 to 8.50 and of the latter only 3 to 4.5. 80.5 per cent of the corpuscles of man ranged between 7.5 and 8.5, 90 per cent of the dog between 5 and 7.5, 80 per cent of the pig between 5 and 7.25, 95 per cent of the ox between 4.75 and 6.25, 89 per cent of the sheep between 4 and 5.25, and 96 per cent of the goat between 3.45 and 4.25. There are also certain relationships between the mean size of the cor- puscles and the size of the species. GulUver states that, if "we confine the observations to small natural groups of the class, such a relation will plainly appear in a rule that the largest corpuscles occur in the largest species and the smallest corpuscles in the small species of a single order or family. This relation is well shown in ruminants, rodents, and edentates, and even in ferae, which offer some exceptions; the largest corpuscles are found in the big seals and the smallest in the little viverras and paradox- ures. In fine, though this rule is appUcable only to single orders or lower sections of apyrenaemata, it extends to the whole class of birds, but neither to the reptiles, batrachians, nor fishes, except in partial instances, which seem to be rather indeterminate or accidental than regular." Attempts to trace a relationship between the number and size of the corpuscles and the speed of the animal's movements have proven negative. The very small and numerous corpuscles of the chevrotain have been associated with the fleetness of the animal; while, on the other hand, the enormously large and comparatively few corpuscles in the amphiuma have been associated with sluggishness. Such assumptions have been founded upon insufficient or erroneous data. There is, as a rule, an inverse relationship between the number of corpuscles per cubic milUmeter and the mean diameter, but even in closely related genera this relationship may not exist. CERTAIN PROPERTIES OF THE ERYTHROCYTE IN RELATION TO GENERA. There are certain pecufiarities shown by the erythrocytes of different species which are of zoological significance. The well-known property of the erythrocytes of mammafian bloods to form rouleaux after the blood is shed has not been observed in the case of bloods having nucleated cells, except in the lamprey. This difference may be purely mechanical, and due to the nuclei preventing the approximation of the sides of the erythrocytes. There are certainly differences in the specific gravities and coloration of the erythrocytes of different species. Plate A I. MAN. II. QUADRUMANA. III. CHEIROPTERA. I ooo I ooo |ooo IV. FER,«. looOoOOoooo 1 p q r » t u w X y 2 V CETACEA VI. PACHYDERMATA. lOOOlOOOOoO Vll. RUMINANTIA. looo oooooOOOOOoDMQpn «bc de f g h iklran v y- * o VIII. RODENTIA. IX. EDENTATA. X. MARSUP. XI. MONOTI OOoo|OOoloo|ol > ■D < 3) m i^ > > XII. AVES. 3 4 5 6 7 8 9 10 U 12 13 14 XIII. REPTILIA ET BATRACHIA. CO , ^\ at LJJ > GymnopoduB. Crocodil, Laoert. Anguis. Coluber. Python. Bufo. XIV, Lissotriton . 13 -< ■J3 m ^^ > > Perca. G.Gu^^a^ ad nat, del, Tinea. Esox. Salmo. Gymnotusi Squalus. Aaim< ^th of ao (noh i i ^J i I i I i i I x 900 ' lOQoa Gulliver's micrometry of red blood corpuscles, all drawn to a uniform scale. Figures XIII, XIV, XV, XVI, XVIL and XVIII represent red blood corposdes of Reptilia and Batrachia, while under figure XIX those of the fishes are given. In ail these figures the names of the animals are inserted upon the plate, and do not require any comment at this place. It is evident that the blood corpuscles of the Amphiuma are so large that they can be perceived by the naked eye A. — Vebtebrata Aptben.«:mata. I. Homo ^nan); 1. Corpuscles lying flat 2. The same on edge • • • • 3. Membranous base of same after removal by « g water of coloring matter: it shows diminu- tion in diameter on account of acquired spherical shape II. Quadrumana (_7nonkeya): 4. Simla troglodytes (chimpanzee) 7.4 5. Ateles ater (black-faced spider monkey) 7.1 6. Lemur anguanensis 6.4 III. Cheiroptera (bats); 7. Cynonycteris coUaris (fruit bat) 6.6 8. Vespertilio noctula (large bat) 6.8 9. Vespertilio pipistrellus (common bat) 6.9 rV. Fera (.beasts of prey): (p) 10. Sorex tetragonurus (shrew) 6.6 (9) 11. Ursus labiatus (lipped bear) 6.8 (r) 12. Bassaris astuta (civet cat) 6.3 is) 13. Cercoleptes oaudivolvulus (kinkajou) 6.6 (() 14. Trichechus rosmarus (walrus) 9.2 (li) 15. Canis dingo (dog, Australian) 7.6 (w) 16. Mustela zorilla (weasel), Fehs leo (hon), Fehs leopardus (leopard) 5.9 (i) 17. Felis tigris (tiger) 6.0 (v) 18. Paradoxurus pallasii (Pallas paradoxure) 4.6 (z) 19. Paradoxurus bondar (Bondar paradoxure) .... 4.5 V. Ceiacea iwhalea): 20. Balsena (boops-whale) 8.2 21. Delphinus globioeps (caaing-whale) 7.9 22. Delphinus phocsna (porpoise) 6.6 VI. Pachydermata: 23. Elephas indicus (elephant) 9.2 24. Rhinoceros indicus (rhinoceros) 6.8 25. Tapirus indicus (tapir) 6.4 26. Equus caballus (horse) ., 6.6 27. Dicotyles torquatus (peccary) 6.7 28. Hyrax capensis (Cape hyrax) 7.7 VII. RuminarUia (ruminants'): (a) 29. Tragulus javanicus (Javan chevrotain, musk- deer) 2.1 (b) 30. Tragulus meminna (Indian chevrotain) 2.1 (c) 31. Tragulus stanleyanus (Stanleyan chevrotain) . . 2.3 (d) 32. Cervus nemorivagus (deer) ._. 3.6 (e) 33. Capra caucasica (Caucasian ibex) 3.6 (/) 34. Capra hircus (domestic goat) 4.0 Q) 35. Bos urus (represented by Gbillingham cattle). . 5.9 (h) 36. Camelopardalis (giraffe) 6.6 (i) 37. Auchenia (vicugna) | fc^^efer U (ft) 38. Auchenia paca (alpaca) { fc'diame^er 4.1 (I) 39. Auchenia glama (llama) | long diameter 7.6 ■^ &- V .» T short diameter 4.1 (m) 40. Camelus dromedarius (single- J long diameter 7.8 hump camel) I short diameter 3.7 (n) 41. Camelus bactrianua (double- j long diameter 8.1 hump camel) \ short diameter 4.3 (o) 42. Cervus mexicanus (deer, Mexican) 4.9 VIII. RoderUia (.Rodents): 43. HydrochcErus capybara (capybara) 8.0 44. Castor fiber (beaver) 7.6 46. Sciurus cinereus (squirrel) 6.4 46. Mus messoriuis (harvest mouse) 5.9 IX. Edentata: ■ 47. Myrmecophaga jubata (ant-eater) 9.2 48. Bradypus didactylus (sloth) 8.9 49. Dasypus villosuB (armadillo) 7.7 X, Marawpwlia: 50. Phascolomys (wombat) 7.3 51. Hj^siprymnus setosus (kangaroo rat) 6.4 XI, Monotremata: 52. Echidna histrix (echidna) 6.6 B. — ^Vehtebbata Ptken-emata. Long Short XII. Aves (jbiTde): diam. diam. 1. Struthio caxnelus (ostrich) 15.4 8.5 2. Same mside round and deprived of color by water. 3. Vanga destructor (East India shrike) . . . 12.6 6.5 4. Lanius excubitor (great gray shrike) .... 12,8 4.8 6. Bubo virginianus (homed owl) 13.8 6.4 6. Bymium nyctea (snowy owl) 16.3 6.3 7. Columba rufina (rufous pigeon) 11.0 7.6 8. Columba migratoria (wild pigeon) 13.3 6.6 9. Dolichonyx oryzivorus (rice bird) 10.6 6.1 10. Buceros rhinoceros (rhinoceros hombiU) 15.0 7.9 11. Fsittacus augustUB (august amazon) .... 12.2 7.0 12. Fhasianus superbuB (barrel-tailed pheas- ant) ^ 11.9 7.1 13. Felecanus onocrotalus (white pelican)... 14.3 7.6 14. Trochilus sp. (humming-bird) 9.4 6.6 IN RELATION TO ZOOLOGICAL DISTINCTION. 59 Certain obscure yet obvious differences have been described in the general microscopic appearances of the erythrocytes of different species, differences which have been expressed by Johnson and by Wormley by the term "stamp of individuality." The behavior of the erythrocytes of different species toward certain reagents shows zoological peculiarities. Thus, Haldane, Makgill, and Mav- rogordato (Journal of Physiology, 1897, xxi, 160) record that chlorates enter the erythrocytes of man and the dog to form methemoglobin, but not those of the mouse, guinea-pig, and rabbit. Other generic pecuUarities have been pointed out by Up de Graf (The Microscope, 1883, quoted by White, loc. cit.), who found that the corpuscles of the dog rupture less readily than those of man when subjected to water. Differences in the osmotic properties of erythrocytes of different species have been recorded by different observers, and the differences in the osmotic pressures of erythrocytes of different species are well known. The erythrocyte, in common with other protoplasmic structures, energetically decomposes H2O2. Bergengruen (Inaug. Dissert., Dorpat, 1888; Maly's Jahresb. ii. d. Fort. d. Thierchemie, 1888, 271) has shown that this property is due to the stroma, and, moreover, that the stromata of the corpuscles of the bullock, horse, and dog showed marked differ- ences in their energy. Studies of cytolysins and agglutinins show marked generic differences in the erythrocytes : Dog's serum hemolyzes the erythrocytes of the chicken, rabbit, sheep, ox, and guinea-pig. Horse serum is not hemolytic to chicken corpuscles, but chicken serum is hemolytic to horse corpuscles. Rabbit serum is not hemolytic to the corpuscles of the guinea-pig, bullock, or dog, but it is slightly hemolytic to bullock corpuscles. Bullock serum is faintly hemolytic to sheep corpuscles, and sheep serum is slightly hemolytic to dog corpuscles. Eel serum is a general strong hemolytic. THE PERCENTAGES OP HEMOGLOBIN IN THE DRY ERYTHROCYTES IN RELATION TO GENERA. The dry corpuscles consist almost wholly of protein, most of which in mammals and birds is hemoglobin. The proportion of hemoglobin varies in the corpuscles of different species, the non-nucleated cells containing a much larger percentage than the nucleated cells. The very limited data at hand indicate that in mammals the percentage will range from about 85 to 95 per cent, according to Hoppe-Seyler and Jiidell (Med. chem. TJnter- Table 22. — The percentages of hemoglobin in the dried erythrocytes of different genera, according to Hoppe-Seyler. Kind. PercentBge of hemoglobin. n, (I 86.79 94.30 86.50 92.25 62.65 46.70 M^njll . Dog.; ......:........::: Heaerebofir Goose Snake ^Coluber natriz) 60 SPECIFICITY OP THE BLOOD OP VERTEBRATES such., 1868, 391; Physiologische Chemie, 1881, 401) (table 22), varying doubtless in different genera, species, individuals of the same species, etc. In birds, taking the goose as the representative, the percentage is approxi- mately only two-thirds that in mammals, while in cold-blooded animals, as indicated by the snake, the proportion is only about half of that in mammals. THE PERCENTAGE OF HEMOGLOBIN IN THE MOIST ERYTHROCYTES IN RELATION TO GENERA. The percentages of hemoglobin in the moist corpuscles probably vary, according to the records of Abderhalden and others (Schmidt, Character, d. epidem. Cholera, Leipzig, 1850; Gorup-Besanez, Physiologische Chemie, 1878, 345; Bunge, Zeit. f. Biologie, 1876, xii, 191; Biemachi, Centralb. f. inner. Med., 1894, xv, 718; Kohler, Centralb. f. inner. Med., 1897, xviii, 724; Abderhalden, Zeit. f. physiolog. Chemie, 1898, xxv, 115, and xxvi, 65), within limits so narrow, varying as much in individuals as in species, that it seems futile to attempt any generic differentiation (table 23). In all the mammals examined the quantity of hemoglobin in reUable records approximates 32 per cent. Table 23. — The percentage of hemoglobin in the moist erythrocytes in relation to genera. Kind. Primate: Man j jj Man . . . . Camivora: Cat Cat Dog Ungnlata: Bullock. Bullock. Sheep. . . Goat. . . . Horae. . . Kg Pig Rodentia: Rabbit. . Percentage of hemogloDin. 15.96 31.11 29.68 29.8 32.99 32.81 31.67 28.00 31.26 32.40 31.68 26.1 32.68 33.19 Authority. Schmidt. Do. Biernachi. Kohler. Abderhalden. Do. Do. Bunge. Abderhalden. Do. Do. Bunge. Abderhalden. Do. THE PERCENTAGE OF HEMOGLOBIN IN THE WHOLE BLOOD IN RELATION TO GENERA. The percentage of hemoglobin in the whole blood has been made the subject of study by numerous investigators by various methods, but, owing to imperfect methods and other reasons, the figures for a given species are so variable as not to permit of close comparisons of different species. Some have made their estimates by the "extinction coeflacient," others by comparison by the colorimetric method, others by the percentage of iron, etc. In determinations by the "extinction coefficient," a weak solution of blood in a layer of given thickness is studied in relation to the extinction of some definite part of the hemoglobin spectrum, usually the second absorption band. These coefficients are proportional to the con- IN RELATION TO ZOOLOGICAL DISTINCTION. 61 centration of the hemoglobin solution. The coefficients for human blood based upon the values of Leichtenstem and others (Vierordt's Daten u. Tabellen, 1906, 220 et seq.) are for men 1.2359, for women 0.9559, and for children from 2 to 10 years of age 1.066. During the early weeks of life the coefficient is much higher than in the adult, but later it falls for a time to a point lower than in men, but higher than in women. Korniloff (Zeit. f . Biologie, 1876, xii, 515), in comparative studies by this method of the bloods of 110 vertebrates, including 44 species of both warm-blooded and cold- blooded animals, reports the following average figures: Mammals 0.9366, birds 0.7814, reptiles 0.4328, amphibia 0.3889, and fish 0.3564. Males, he found, have a higher hemoglobin capacity than females, and the old a higher capacity than the young. Tablk 24. — The percentages of hemoglobin in relation to genera. Kind. Peroentaee of hemoglobin. Authority. Kind. Fercentaee of hemoglobin. Authority. Primates: Man 12.09 to 16.07 11.67 to 13.69 Preyer. Ungulata-con'd: Horse Horse 12.06 to 13.19 10.4 to 10.88 Pelouze. Quinquaud. Woman Man 11.8 to 12.77 10.4 to 11.36 Quinquaud. Do. Horse Horse 13 11.34 Muller. Arronnet. Woman 13.00 12 to 14.6 12.05 to 12.78 MuUer. Henocque. Pelouze. Means deter- Pig 14.03 to 14.80 13.32 14.22 14.09 to 14.17 Preyer. Muller. Abderhalden. Pelouze. Man Pil Man Pie Pig Man 13.65 mined from Vierordt's Daten u.Ta- Rodentia: 8.85 8.6 to 9.22 Do Woman 12.67 Rat Quinquaud. bellen. Rat 8.68 to 9.12 Preyer. Monkey 6 to 14 Henocque. Guinea-pig . . . 14 Henocque. Carnivora: Rabbit 12.35 Abderhalden. Cat 14.32 13.12 to 13.46 10.51 14 to 14.5 Abderhalden. Preyer. Muller. Henocque. Rabbit, male . Rabbit, female Rabbit Aves: 10.51 8.77 8.41 Otto. Do. Subbotin. Doe Doe Doe Dog 13.34 to 14.66 14.16 AbderMden. Otto. Turkey Chicken 7.93 to 8.00 8.5 Pelouze. Do. Dog, male — Dog, female . . 13.76 Do. Chicken 9 to 9.92 Preyer, Doe 13.80 16.56 to 17.4 16.83 to 18.08 Subbotin. Fudakowski. Arronet. Goose Duck Duck Duck 8.26 to 8.76 8.14 to 8.19 9.16 to 9.42 8.2 Pelouze. Do. Preyer. Quinquaud. Doe Doe Ungulata: Bullock 13.33 to 13.96 Preyer. Pigeon 9 to 11.6 Henocque. Bullock 10.31 Abderhalden. Pigeon 7.31 to 12.66 Subbotm. Bullock 10.4 to 11.35 Quinquaud. Pigeon 7.09 to 8.03 Quinquaud. Bullock 10.21 MuUer. Sparrow 7.09 to 7.5 Do. Bullock 10.30 Abderhalden. Amphibia: Bullock 12.10 Subbotin. Sea-tortoise . . 6.94 to 8.98 Bardachzi. Bullock 11.43 to 13.02 Pelouze. Frog 10.12 Pelouze. Calf 10.11 to 10.68 Preyer. Quinquaud. Subbotin. Preyer. Muller. Frog Lizard 2.35 to 3.3 Quinquaud. Henocque. Calf 6.62 to 9.46 2 to 13 Calf 8 91 11.11 to li;53 Pisces: (according to season and condition) Sheep Sheep 10.93 Tench 2.36 to 3.78 Quinquaud. Sheep 9.29 to 10.28 Abderhalden. Skate 3.5 to 3.8 Harris. Sheep 6.62 to 8.98 Quinquaud. Invertebrata: Goat 11.26 Abderhalden. Annelids 3.4 VeUchi. Horse 16.69 Do. The estimates recorded by Preyer, Abderhalden, Henocque, and others by various methods are, broadly speaking, in accord with those of Korni- loff. (Preyer, Die Blutkrystalle, 1872, 116, and Annalen d. Chemie u. Phar., 1866, cxl, 187; Muller, Archiv f. Thierheilk., 1886, xii, 96; Henocque, 62 SPECIFICITY OF THE BLOOD OP VERTEBRATES Compt. rend. sec. biologie, 1886, cm, 493; Korniloff, Zeit. f. Biologie, 1876, XII, 515; Leichtenstern, Vierordt's Daten u. Tabellen, 1906, 220; Abderhalden, Zeit. f. physiolog. Chemie, 1898, xxvi, 65; Otto, Archiv f. ges. Physiologie, 1885, xxxv, 36; Subbotin, Zeit. f. Biologie, 1871, vn, 185; Pelouze, Compt. rend. soc. biologie, 1865, lx, 880; Otto et al, Vier- ordt's Daten u. Tabellen, 1906, 220 et seq. ; Jolyet, Gazette m^dicale, 1874, 383; Bardachzi, Zeit. f. physiolog. Chemie, 1906, xlix, 465; Quinquaud, Compt. rend. soc. biologie, 1873, lxxvii, 1489, and lxxvii, 487; Velichi, Inaug. Dissert., Berhn, 1900; Centralbl. f. Physiologie, 1901, xiv, 679; Fudakowski, Centralbl. f. med. Wissensch., 1866, iv, 705; Arronet, Inaug. Dissert., Dorpat, 1887; Harris, Journal of Physiology, 1903, xxx, 319.) In mammals the percentage ranges usually between 10 and 15; in birds, between 7 and 9; and in cold-blooded animals, from 2 to 10. Glancing over the figures for mammals (table 24), it appears that the percentages fall in the following order: pig, dog, cat, man, horse, bullock, goat, and sheep. The mean of rodents is notably lower than that of pri- mates, carnivora, and ungulates, the percentages being in the following order: guinea-pig, rabbit, and rat. In birds the percentage is, on the whole, distinctly lower than in rodents. Among cold-blooded animals it is prob- ably highest in the tortoise. Omitting the figures of Pelouze for the frog, which obviously are incorrect, the percentage in the frog is only about one-sixth to one-fourth of that in mammals. GENERAL CONSIDERATION OF THE ZOOLOGICAL SPECIFICITIES OF THE BLOOD. Zoological pecuharities of the blood or pseudo-blood are shown through- out the animal kingdom, vertebrate and invertebrate. The facts that have been brought together in this chapter show clearly not only marked zoo- logical differentiations, but also what important information is to be ex- pected from additional and detailed data along the same or related Hnes of inquiry. These distinctions are shown in part: (1) By the whole blood in differences in the proportions of blood to body-weight, in specific gravity, in alkalinity, in the proportions of cor- puscles to plasma, in the degree of coagulability and in the character of the fibrin, in the degree of decomposability, etc. (2) In differences in the plasma as regards especially the percentages and kinds of proteins and in the "protein quotient," in the percentage of cholesterin, in the pecuharities of the "precipitins," agglutinins, and hemolysins, etc. (3) In the leucocytes in respect to kind and relative numbers, in peculi- arities of the granular matter, in specific physiological pecuharities, etc. (4) In the erythrocytes in their size, structure, form, and number per cubic millimeter, in their behavior towards certain reagents, in their percent- ages of hemoglobin, sodium, potassium, phosphorus, and cholesterin, etc. In subsequent pages will be found additional evidence of specificities in the differences in the stroma, in the general chemical and physical prop- erties of the hemoglobins, and especially in the remarkable generic peculiar- ities of the hemoglobins as shown by their crystallographic properties. IN RELATION TO ZOOLOGICAL DISTINCTION. 63 These zoological differences are paralleled in the invertebrates, as will be manifest even after a most superficial inquiry, as, for instance: In the Protozoa and Porifera there is an entire absence of any fluid which we are justified in regarding as being even an analogue of the blood. In the Hy- drozoa, Actinozoa, and Echinodermata the perivisceral or chyUferous fluid is the simplest expression of a rudimentary blood. In the former this fluid scarcely differs from that in which the organism Uves, and it contains but few if any corpuscles, which are of a rudimentary character. In the Actin- ozoa and Echinodermata there is a distinct approach to a typical blood, there being both rudimentary and typical cells. The chylaqueous fluid of the Annelida contains typical corpuscles, and it is among the animals of the annuloid series, in Trichoscolices and Annelida, that we find the first appearance of hemoglobin, of colored corpuscles, and of a true blood — i.e., a circulatory fluid which combines the functions of both circulation and res- piration, and therefore that is comparable with the blood of the vertebrate. The first appearance of hemoglobin is, as far as known, in holothuridean {Thyonella gemmata), ophiuridean {Ophiactis virens), and turbellarian {Polia sanguirubra) echinoderms, although histohematins have been found in certain of the Porifera and Actiniidce. The first appearance of an ana- logue of the erythrocyte has been noted in the Gephyrea {Sipuncvhis nudus, S. balanorophus, S. echinxyrhynchics, Phxiscolosonm elongatum), in which are corpuscles having a distinct cell wall which incloses a colored fluid in which a nucleus is suspended. The coloring matter of the corpuscles is alUed to hemoglobin, but the corpuscles have no histological relationship to the vertebrate erythrocyte. The fluid of the pseudo-hemal system of certain Annelida contains a red coloring matter in the form of a histohematin which is closely allied to hemoglobin chemically and physiologically; that of others is green and colored by chlorocruorin, which also is closely aUied to hemoglobin; and that of others is colored by hemoglobin, etc. In invertebrata the blood or pseudo-blood may be colored or color- less, which differentiation is not related to the position of the organism in the scale of fife; but in aU of the non-generate vertebrates the blood is colored. In the invertebrates the coloring matter may be in solution in the circulatory fluid or in the corpuscles, but, as a rule, it is in the fluid, whereas in the vertebrates it is without exception in the erythrocytes. While the corpuscles of invertebrate blood are allied histologically to the leucocytes of vertebrate blood and not to the erythrocytes, when colored they may be (but usually are not) respiratory Hke the erythrocytes. In the vast majority of invertebrates the coloring matter of the blood is hemocyanin, which is, as far as known, invariably in solution in the plasma, the venous blood being colorless, or nearly so, and the arterial blood of various tints of blue or violet. Various other pigments have been found, as shown in the preceding chapter. In no instance has it been recorded, as far as we are aware, that both hemocyanin and hemoglobin coexist in the same blood or even in the same individual, although in certain organisms with a hemocyanin blood histohematins and myohematins have been found in certain of the body structures. Bloods that owe their coloration to other 64 SPECIFICITY OF THE BLOOD OF VERTEBRATES substances than hemocyanin may exhibit a variety of colors, which colors may be properties of the plasma or corpuscles or both. Even among animals of a given subkingdom the coloring matter may dififer very much. Thus, among Annelida, as stated, certain bloods are red owing to hemoglobin; in others the coloration tends to a reddish-brown owing to echinochrome; while in others it is green owing to chlorocruorin, etc. Among the Arthro- poda we may find hemoglobin, hemocyanin, pinnaglobin, or other pigment present. In Inseda we never find hemocyanin, and rarely hemoglobin; in Crustacea there may be hemocyanin or hemoglobin; in Decapoda and Stomatopoda the blood coloring matter is hemocyanin; while in Cladocera, Phyllopoda, Copepoda, and Ostracoda there is hemoglobin. Among Mol- lusca, in some the blood is colorless; in others is found pinnaglobuUn, which is closely related to hemoglobin, containing manganese in place of iron; in some there is hemoglobin; but in most of them the coloring matter is hemocyanin. In Gasteropoda and Cephalopoda hemocyanin is present. In the Chcetopoda hemoglobin has been found in quite a number of species, and chlorocruorin in a few. In Gephyrea, Nemertina, and Hirudinea hemo- globin has been noted. In Insecta the blood may be colorless or of various colors, but is usually colorless. The coloration is a property of the blood plasma, and except the blood of the larva of the dipterous insect Chironomus and Musca domestica there appears to be an absolute absence of hemoglobin from the bloods of these animals, although histohematins and myohematins have been found in various of the body structures of a number of them. In all the invertebrates which have hemocyanin this substance is, as far as known, in solution in the blood plasma; but the fact that copper has been found in leucocytes of the oyster and in other structures of inverte- brates, and that there are blue and violet corpuscles, leads to the belief that in certain invertebrates hemocyanin may be a component of both plasma and corpuscles, and even of other structures. In all invertebrates in which hemoglobin has been found it has been noted in solution in the blood plasma, excepting Glycera, Capitella, Phoronis, and Solen, in which it is a constituent of special blood corpuscles. Among the invertebrates the blood corpuscles may be colorless or colored, and when the latter they may be green, red, yellow, blue, violet, piuple, madder, mahogany, brown, hlac, etc.; and in certain organisms (Spatangus) a variety of corpuscles of different colors may be present in the same blood, such as yellow, green, brown, indigo-blue, and purple. In addition to these exceedingly interesting pecuUarities, there have been noted in the bloods of different invertebrate organisms differences in the specific gravity, in coagulability, and in the percentages of proteins, copper, and saUnes; and in the kinds, sizes, composition, and relative number of the corpuscles, etc. Gaskell (The Origin of Vertebrates, London, 1908) gives evidence and arguments upon geological, anatomical, and embryological grounds that lead to the belief that vertebrates may have had their origin from palseos- tracans. He states (p. 65) that the evidence of geology "points directly and strongly to the origin of vertebrates from the palseostraca-arthropod IN RELATION TO ZOOLOGICAL DISTINCTION, 65 forms, which were not crustacean and not arachnid, but gave origin both to the modern-day crustaceans and arachnids. The history of rocks further shows that these ancient fishes, when they first appeared, resembled in a remarkable manner members of the palaeostracan group (trilobites, higher scorpion and king-crab forms), so that again paleontologists have found great difficulty in determining whether a fossil is a fish or an arthropod. Fortunately, there is still alive on this earth one member of this remark- able group — the Ldmulus, or king-crab. * * * There are no trilobites still alive, but in Branchipus and Apus we possess the nearest approach to the trilobite organization among Hving crustaceans." In this connection it is of interest to note that hemocyanin has been found in the blood of Scorpiones (Scorpio), Xiphosura {Lirmdus), Decapoda (Homarus, Astacus, Cancer, Carcinus, Nephrops, Eriphia), and Stomato- poda (Squilla and Maia); and that hemoglobin has been found in the bloods of Diptera (Chironomus, Musca domestica), Ostracoda (Cypris), Copepoda (Lernanthropus) , Cladocera (Daphnia), and Phyllopoda (Apus and Branchipus). The great importance of hemoglobin in vertebrate life, as is indicated, for instance, in the fact of its universal presence in every living non-degen- erate vertebrate, suggests that if, as Gaskell contends, vertebrates had their origin from palseostraca, it was more likely from one of the group in which hemoglobin and not hemocyanin is the respiratory pigment of the blood. These facts, brought together as they are in so fragmentary and unsatis- factory a way, are nevertheless suflBcient to be convincing that the results of detailed inquiry, which has been denied us through lack of time, will prove of the utmost importance in zoological dififerentiation. CHAPTER III. HEMOGLOBIN; ITS GENERAL CHEMICAL AND PHYSICAL CHARACTERS, AND ITS SPECIFICITIES. CONSTITUENTS AND RELATIONS TO THE OTHER CONSTITUENTS OF THE ERYTHROCYTES. Hemoglobin, Cg4.57H7.22Ni6.38So-68Feo.33602o-40 (Jacquet, Zeit. f. physi- olog. Chemie, 1890, xiv, 289), is regarded as being composed of a colorless, strongly basic albuminous radical, termed globin, C54.97H7.2Ni6.89So-42020-52 (Schulz, Zeit. f. physiolog. Chemie, 1898, xxiv, 449), and a non-albuminous, colored radical termed hematin, C34H34N4Fe05 (Ktister, Zeit. f . physiolog. Chemie, 1904, xl, 391). The latter constitutes 4 to 4.5 per cent of the mole- cule (Schulz, loc. cit, and Lawrow, Zeit. f. physiolog. Chemie, 1898, xxvi, 343), and is, so far as known, probably absolutely identical in the hemo- globins of all animals; but the former is in all likeUhood not identical, as is indicated by certain differences in chemical composition and constitu- tion of the hemoglobins of different bloods; by the difference between hemoglobin and myohematin; by the fact that globin may be replaced by egg-white (Ham and Balean, Journal of Physiology, 1905, xxxii, 312), or by albumin from the blood of another species (Bertin-Sans et Moitessier, Compt. rend. soc. biologie, 1893, cxiv, 923; Bull. soc. chim., 1893, 5 Mai, 5 Sept.) ; and also by the fact, as stated by Schulz {loc. cit.), that the globins of the dog and horse are not identical with that of the goose. Globin is a histone-hke body; it has been isolated by Schulz and others; and its primary dissociation products have been studied by Fischer and Abderhalden (Zeit. f. physiolog. Chemie, 1902, xxxv, 268) and by Abderhalden (Zeit. f. physiolog. Chemie, 1903, xxxvii, 484). Only 72 per cent of these products have been accounted for — ^leucin 29, histidin 11, arginin 5.4, asparaginic acid 4.4, lysin 4.3, alanin 4.2, phenylalanin 4.28, prohn 2.3, glutaminic acid 1.7, tyrosin 1.3, oxyprolin 1, serin 0.6, cystin 0.3, ammonia 0.93 per cent. The sulphur is contained chiefly in the cystin, and the iron solely in the hematin. The union between globin and hematin is very feeble, the addition of weak acid being suflScient, as has been shown by Ham and Balean (loc. cit.), to cause immediate dissociation. The nature of this linking is in doubt. According to Hoppe-Seyler (Archiv f. path. Anat. u. Physiolog., 1864, XXIX, 233; Centralbl. f. med. Wissensch., 1864, 261, and 1865, 491), it is ester-Uke, while Hufner (Archiv f. Anat. u. Physiolog., 1899, 491) and Ham and Balean (loc. cit.) regard it as being through the agency of oxygen. According to the hypothesis of Ham and Balean, the formula for oxy- hemoglobin is: C,„H.,N,0. .0 »N,0/ \0— G 67 68 GENERAL CHEMICAL AND PHYSICAL CHARACTERS in which G represents the globin radical (page 26). According to Zinofifsky (Zeit. f. physiolog. Chemie, 1886, x, 16), the molecule may be regarded as consisting of two molecules of globin and one molecule of hematin. Whether or not globin and hematin are thus combined, or the hematin is linked with one or several molecules of globin; whether the globin is a simple or com- pound body; whether the hematin may be combined with polymeric or isomeric forms of globin; whether the hematin is with certainty a uniform substance, etc., are still open questions. If, as Miescher states, the albumin molecule with its 40 atoms of carbon may have as many as a biUion stereo- isomers, what may be the possibilities of hemoglobin or globin molecules with their hundreds of carbon atoms? Whatever may be the chemical relations between globin and hematin, they are so peculiarly associated that undecomposed hemoglobin gives neither albuminous nor iron color reactions. It is of incidental interest to note that, except the iron in hemoglobin, nearly all of the iron of the tissue cells is contained in the nucleoproteins, and that while these substances, unlike hemoglobin, yield the protein color reactions, they, like hemoglobin, do not yield iron reactions, show- ing that in both the iron is in a non-ionic or "masked" state. We are also in doubt as to the state or states in which hemoglobin exists in the erythrocytes, especially as to whether it is in a liquid, semi- liquid, or solid form, and as to the nature of the compound or compounds it probably forms with other constituents of the erythrocytes. The red corpuscles consist of a stroma and hemoglobin with other substances. The former is elastic, non-contractile, seemingly homogeneous, colorless, transparent, and albuminous. According to some, the stroma is in the form of minute sacs which contain hemoglobin and other substances in solution. According to others, it is in the form of a protoplasmic mass, throughout which the hemoglobin and other substances are distributed. That the hemoglobin is not in either crystalline or amorphous form has been shown by microscopic examination with high powers; and that it is not in solu- tion in a free state seems obvious from the fact that in the case at least of the very insoluble forms of hemoglobin, as in the guinea-pig, squirrel, rat, necturus, etc., not only are the water and the inorganic salts of the corpuscles wholly inadequate to dissolve or keep in solution the hemo- globin, but even the entire blood plasma is altogether insufficient to hold the hemoglobin in solution when freed from the corpuscles. The assumption of Preyer that the hemoglobin is held in solution in the corpuscles by virtue of potassium salts because of the presence of a relatively high percentage of these salts in comparison with the percentage in the plasma, and because of the higher solubiUty of the hemoglobin in water when these salts are present, is not worthy of consideration, inas- much as in certain bloods, for instance in those of the dog and cat, the per- centage of potassium in the corpuscles is practically the same as in the plasma, and yet in the dog crystallization takes place rapidly in the plasma upon the laking of the blood. Rywosch (Centralbl. f . Physiologic, 1905, xix, 388) believes that the hemoglobin is present in the corpuscles in a free state. He found, after destruction of the erythrocytes by grinding in sand, OP HEMOGLOBIN, AND ITS SPECIFICITIES. 69 that by mixing the pulp with an isotonic salt solution the hemoglobin was dissolved. This he holds would not occur " if the hemoglobin was in com- bination with the stroma." However, his method may have been the means of breaking up a hemoglobin-stroma union. Moreover, Stewart (see below) has found, in his experiments on the influences of various agents on the osmotic properties of the erythrocytes, that the hemoglobin can not exist in the corpuscles in ordinary aqueous solution. Hoppe-Seyler (Physiologische Chemie, 1877, 381; Zeit. f. physiolog. Chemie, 1889, xiii, 477) attempted to show, by various facts and arguments, that such differences exist between the behavior of the coloring matter of the blood as it exists in corpuscles and hemoglobin in solution that they can not be identical. Most of his deductions have, however, been found to be untenable. He distinguishes between the coloring matter of the blood, oxyhemoglobin, and reduced hemoglobin, regarding both oxyhemoglobin and reduced hemoglobin as cleavage products. He looks upon the " coloring mat- ter" of the blood as consisting of combinations of oxyhemoglobin and hemo- globin with lecithin, forming firm chemical unions. The coloring matter of arterial blood he distinguishes as arterin and that of venous blood as pkbin, the only difference between these two substances being a feebler combina- tion of oxygen in the former. While Hoppe-Seyler's hypothesis seems to have received a tacit acceptance, it has been opposed by Gamgee (Scha- fer's Text-book of Physiology, 1898, 1, 190) and questioned by others as being untenable; but it has been defended by Kobert (Das Wirbeltierblut, etc., Stuttgart, 1901, 5). Bohr (Zentralbl. f. Physiologie, 1904, xvii, 682, 688) believes that the coloring matter of the blood, which he terms hemochrome, is not identical with hemoglobin (which he prepared without the addition of alcohol), because the latter has a lower oxygen capacity. Recent evidence that hemoglobin exists in the corpuscles in some peculiar form of combination has been recorded by a number of investi- gators. Thus, Stewart (Journal of Physiology, 1899, xxiv, 211; Amer. Journal of Physiology, 1902, viii, 103) found, in a very interesting study of the effects of laking agents, "that the relations of the hemoglobin and the electrolytes of the corpuscle to some of the other constituents of the cor- puscle or to the envelop are such that under certain conditions hemoglobin may be Uberated while the electrolytes are retained; while under other conditions electrolytes may pass through an envelop which refuses passage to the hemoglobin, although in general it is easier for the hemoglobin, in spite of the great size of the molecule, to escape from the corpuscles than it is for the electrolytes. " He also found that, while hemoglobin may pass from the corpuscle, hemoglobin dissolved in the serum would not pass into the corpuscle. In explanation of these phenomena Stewart proposes four hypotheses as to the condition of the hemoglobin and the electrolytes in the corpuscles: (1) A portion of the electrolytes and of the hemoglobin is in solution as such; and the rest is in solution as compounds with other substances, such compounds being imable to pass through the envelop. 70 GENERAL CHEMICAL AND PHYSICAL CHARACTERS (2) A portion of the electrolytes and of the hemoglobin is in solution as such, and the rest exists in a soUd or semisoUd form united to some constituent of the stroma. (3) A portion of the electrolytes, but none of the hemoglobin, is in solu- tion as such; the whole of the hemoglobin and the rest of the electrolytes being in solution in the form of such compounds as are mentioned in (1). (4) A portion of the electrolytes, but none of the hemoglobin, is in solution as such; the rest of the electrolytes and all of the hemoglobin are united in the stroma. The last hypothesis, he thinks, best takes account of the facts of laking. Oxygen, it seems, serves as a connecting link not only between globin and hematin, but also between the stroma and hemoglobin. The removal of oxygen from the blood causes hemolysis. This phenomenon might, at first thought, be regarded as a mechanical effect due to the rapid dis- charge of from the erythrocytes when the blood is subjected to the vac- uum pump, but this is negatived by the fact that hemolysis occurs just the same when a continuous stream of CO2 is passed through the blood and the O thus driven off gradually. Even the Unkage between globin and hematin may be broken by CO2. That the hypothetical union between hemoglobin and the stroma must be a feeble one is evident in the readiness with which it is broken, by the removal of O from the blood, by minute quantities of foreign serum, snake venom, and certain bacterial products, by repeated freezing and thawing, etc. While it thus seems probable that the hemoglobin of the corpuscles is essentially or solely in some form or forms of union with the stroma, it is also probable, from the investigations of Hiifner (Archiv f . Anat. u. Physiologie, 1894, 135, 176), that the combination does not, in opposition to Hoppe-Seyler's statements, effect a marked alteration in the chemical nature of hemoglobin in so far as pertains to its relations to oxygen and to light, for he found that its behavior to oxygen and its spectropho- tometric properties are the same as when the hemoglobin is free, provided the solution be of the same degree of concentration. On the other hand, it is positive that at least the degree of solubility in relation to the plasma and the crystallizability are lessened to a marked degree, so much so that the crystallization may occur in the plasma of partially laked blood and not in the corpuscles, even though in the latter the concentration of the hemoglobin may be greatly higher. The corpuscles of the dog contain about 33 per cent of hemoglobin, while the highest percentage that could exist in the laked blood is about half of this; but while crystaUization does not occur in the corpuscles, it does occur rapidly in the laked blood. (See Chapters V and XV.) THE ELEMENTARY COMPOSITION OF HEMOGLOBIN. The determinations of the centesimal composition of hemoglobin of different species of animals differ sufficiently to indicate that all hemoglobins are not alike; but these differences are not on the whole greater than those noted in the analyses of specimens of blood from individuals of the same OF HEMOGLOBIN, AND ITS SPECIFICITIES. 71 species, and are therefore of little significance in indicating positive non- identity (table 25). In fact, the analyses, as a whole, are so discrepant that it must be admitted that hemoglobin is not a uniform substance. Table 25. — The centesimal composition of hemoglobin, according to various observers. Centesimal composition. Authority. C. H. N. S. Fe. o. Camivora: Cat 54.60 63.91 54.57 53.86 54.00 54.15 53.64 54.66 54.87 54.76 54.40 51.16 64.56 64.75 54.40 54.17 64.71 54.09 54.12 52.47 54.26 54.77 53.91 63.86 7.25 6.62 7.22 7.32 7.25 7.18 7.11 7.25 6.97 7.03 7.20 6.76 7.15 6.98 7.25 7.38 7.38 7.39 7.36 7.19 7.10 6.99 7.02 7.10 16.52 16.98 16.38 16.17 16.25 16.33 16.19 17.70 17.31 17.28 17.61 17.94 17.33 17.35 17.51 16.23 17.43 16.09 16.78 16.45 16.21 17.07 0.62 0.542 0.568 0.39 0.63 0.67 0.66 0.4 0.65 0.67 0.65 0.39 0.43 0.42 0.45 0.66 0.479 0.69 0.58 0.859 0.54 0.38 0.41 0.37 0.35 0.333 0.336 0.43 0.42 0.43 0.43 0.447 0.47 0.45 - 0.47 0.335 0.38 0.393 0.426 0.399 0.40 0.48. 0.335 0.43 0.41 20.66 22.62 20.93 21.84 21.45 21.24 20.03 19.543 19.73 19.81 19.67 23.42 20.12 19.85 21.634 19.602 21.44 20.68 22.50 20.69 Abderhalden, Phyeiologischen Chemie, 1906, 696. Jacquet, Zeit. f. physiol. Che- mie, 1888, XII, 285. Jacquet, Zeit. f. physiol. Che- mie, 1890, XIV, 289. Hoppe-Seyler, Med. chem. Un- tersuch., 1868, Heft 3, 366. Hufner, Jour. f. prakt. Chemie, 1880, XXII, 362. Schmidt, Preyer, Die Blut- krystalle, 1872, 65. Schmidt & Bottcher, Ueber Blutkrystalle; Inaug. Dis- sert., Dorpat, 1862. Hufner, Beitrage z. Physiol., C. Ludwig, Leipzig, 1887, 74. Kossel, Zeit. f. physiol. Che- mie, 1878-9, II, 149. Otto, Archiv. f. ges. Physiolo- gic, 1883, XXXI, 240. Httfner& Bucheler, Zeit.f. phys- iol. Chemie, 1884, viii, 358. Zinoffsky, Zeit. f . physiol. Che- mie, 1880, X, 16. Schulz, Zeit. f. physiol. Che- mie, 1898, XXIV, 449. Abderhalden, Zeit. f. physiol. Chemie, 1903, xxxvii, 494. Jutt, Inaug. Dissert., Dorpat, 1894; Maly's Jahresbr. lanes bounded by 6 sides. II- ustration in Funke'e Atlas, x, 4. Reichert, MdUer's Archiv, 1849, taf. iLfig. 6. Kunde, Zeitsohr.f. rat. Med., taf. ix, fig 2 (1852). Illustration in Funke's Atlas, x, 5. Lehmann's statement, that the crystals do not belong to the hexagonal system, is incorrect. Kunde gives an illustration in Zeitschr. f. rat. Med., taf. ix, fig, 3 (1852); also Kiihne, Lehrb., p. 200. Illustration, Zeitschr. f , wiss. Zool,, xn, taf. XXX, fig. 5. From the heart blood of the mouse I ob- tained only smallprismatic crys- tals, Kunde (Zeitschr. f, rat. Med,, N. F,, a, 1852, 285) ob- tained with water, without any additional mixture, needles and "prismatic plates," Hoppe^Seyler (Handbuch, 1865, p, 202) obtained crystals by simply diluting the blood with water. Compare Kunde in Zeitschr. f, rat. Med., 1852, N. F., ir, 276, Bisegger and Brach found the crystals "prismatic." (Ver- bandl. d. naturforsch, GSes. zu Basel, 1, 1857, 174.) Illustration in Zeitschr, f, wiss. Zool., xn, taf, 30, 2, and in Rol- lett, Vers. u. Beob. am Blute, Wien, 1862. Compare Ihe same, p. 25. Kunde (Zeitschr, f, rat, Med., 1882, p. 284) obtained the crystals simply by addition of water, as also did Teichmann (sameplace, 1853, 376), Budge, Spec, Physiol. 8 Aufl., p. 250, Illustration in Funke's Atlas, zx, 6. Kunde also saw the crystals. Valentin, in Moleschott's Unters, z. Naturl., 1863, ix, 131, W, Kahne, Med, Centralbl,, 1863, No, 53, p. 833. Illustration in the Zeitschr. f. rat. Med., N. F„ 1 Bd., 1851, taf. 1, figs. 4, 5, 6, Funke obtained the crystals from diluted Venous splenic blood, Kunde (same, 2 Bd., 1852, p. 285) from jugular ven- ous blood. Funke found the angle 60° 9' and 119° 32', I have seen prismatic crystals in the wether's blood after removal of the gases. But it is very diffi- cult to bring the blood to crys- tallization in any other way. See A. Schmidt, in Virchow's Arch., XXIX, p. 1, 1864. Funke also saw the crystals. Kunde obtained them by means of ether (Zeitschr. f. rat. Med., 1852, p. 284), Teichmann (ibid., 1853, 376) by allowing the blood to evaporate after dilution with 4 to 5 times its volume of water. ON THE CETSTALLOGKAPHr OP HEMOGLOBIN. 105 Table 31. — Preyer's table showing the source of hemoglobin crystals, etc. — Continued. Name of kind. Crystal form. Crystal system. Appearance. Solubility in water. Crystallizability. Remarks. Fig iSiuierofadomea- tica) Prisms (Preyer) . Intraglobtilar . Crystallites with extraordinary difficulty Owl (.Strix noctua) . 4-8ided plates (Preyer) Raven (.Cormts) . Sphenoids. Crow aCimmt corone) Lark iAlaudacrietata) Rhombic plates and comb^bap- ed and fan- Most probably rhomoio (Preyer) Very probably rhombic Same as above. Estiaglobular. Do. Do. Very diffieultly soluble in cold, not readily sol- uble in warm water (Bojan- owski) Crystallizes easily (Preyer) Crystallizes with difficulty Crystallizes easily shaped grouped __prism8 (Preyer) grou , OPre: Needle-shape crystals ending very pointedly Do. Sparrow. Pigeon... Like the lark crys- tals Very difficultly soluble in cold, very readily soluble in warm water (Bojan- owski) Sphenoids. Domestic goose. Large 4 or 6-sided rhombic plates Rhombic (?). Lacerta Turtle ^Tetludogrceca) Python (.Python achneideri) Python (Python bivU- tatut) Frog (Rana aeuUtUa) Prisms Needles and plates Prisms and plates Intraglobular . Extraglobular. Prisms. Intraglobular and extraglobular Intraglobular Crjrstallizes with much difficulty White-fish (Leueiacu* dobitia) Do. Intraglobular and extraglobular Crystallizes very easily Compare Funke, Joum. f. prakt. Ch., LTI, 19S, and Zeitschr. f . rat. Med., 1852, 201, and Klebs, Med. Centralbl., 1863, No. 54, p. 852. I also saw the crystals in every blood corpuscle. Funke refers to nets of crystal rods in the compounds. Meckel (Archlv f. d. Holl. Beitr. z. Nat.- u. Heilkunde) saw the intraglobu- lar crystals also. Teichmann obtained them by the evapora- tion of diluted blood (Zeitschr. f. rat. Med., 1853. 376). I obtained owl-blood ciTstals by letting a drop of blood, 2 days old, stand between the object- glass and cover-glass at room temperature. Illustration in Boianowski, Zeit- schr. f. wiss. Zool., XII, taf. 30, fig. 12. 1 obtained very large crystals from frozen heart blood. The crystal form is not clearly recognizable from the illustra- tion in Bojanowski, Zeitschr. f. wiss. Zool., XII, taf. 30, fig. 9. The crystals were produced by Bo- janowski, Zeitschr. f. wiss. Zool., XII, 334. Bojanowski found the pigeon- blood crystals similar to the raven-blood crystals, loc. dt.j p. 335. Hoppe-Seyler finds that dove-blood crystals are more easily produced pure than dog- blood crystals. Kunde, Zeit- . schr. f. rat. Med., N. F., ii, 285, and Teichmann (in the same place, 1853, 376). Hoppe-Seyler finds that goose- blood crystals, according to his method, can be more easily pro- duced pure than dog-blood crys- tals. According to Kelliker. Eunde, Zeitschr. f . rat. Med., N. F., n, p. 285. Berlin found, 1856, the blood of python crystallizable. He also saw crystcds in the stomach of the Amblyomma ezomatum, a blood-sucking parasite which the snake had Drought with it from Senegal to Euroi>e (Neder- landsch. Lancet, 3 eerie, 5 Jaar- gang, 1866-56, p. 739). Zeitschr. f. wiss. Zool., 1849, 1, 266 (Eolliker). Illustration in Virchow's Archiv, XXX, taf. 16, fig. 4, and Bullet, de I'Acad. de St. P^tersbourg, vin, 661-572. Teichmann (Zeit- schr. f. rat. Med., 1855, p. 379) obtained the crystals by mixing the defibrinated blood with very much water and allowing it to evaporate at a low temperature, but since he obtained them color- leas, it is doubtful if they con- sisted of hemoglobin. I saw the crystals in extravasated blood in the lymph-sac. Illustration in Funke's Atlas, taf. X, fig. 6. Funke saw the direct change of the blood corpuscles to crystals, and on the addition of water blood corpuscles were ageun formed. 106 INVESTIGATIONS OF PBETER. Table 31. — Preyer's table showing the source of hemoglobin crystals, etc- -Goncluded. Name of kind. Crystal fonn. Crystal system. Appearance. Solubility in water. Chystallixability. Remarks. Carp (Cj/prinus caX' Red-eyed roach (Cy- prinua erythroph- thabnua) Barbel (£ar&u« /!u- maiUit) White bream lAbra- mit bliceai Tench (.Tinea chry- silis) River bream (.Cypri- nu8 brama) River perch (Perca ftuviaiaU) Herring (Clupea harengus) Sole (.Plateita vul- pffcTjLox ludut) . . . Horn-fish (Belont Toatrata) Earthworm (.Immbri- CU8 terrestriB) Horseleech (7) (Ne- phelU) Soaly-ahaped crystals (Funke) Prisms Crystalliies veiy easily on the ad- dition of water Crystalliies very easily Funke, in Zeitschr. f. rat. Med., 1851, 191. Kunde, ibid., 1852, Remak (Mttller's Arch., 1852, 121) found the crystals 2 hours after death in the blood-vessels. Funke (Zeitschr. f. rat. Med., 18S2, 200) saw the change of the crystal containing blood cor- puscles into the ordinary ones on the addition of water. Funke saw the change of the blood corpuscles into crystals and the rechanging of the same on the addition of water, and could recrystallize them 3 to 4 times on the object-glass. Remak (Mailer's Archiv, 1852, 121) saw the crystals always 24 hours after the death of the ani/- mal in thick bundles in the vessels and in the heart. His statement that they are readily soluble in ether and alcohol rests on delusion (see Zeitschr. f. rat. Med., 1852, 213). The crystals can be recrystallized on the object-glass. Bojanowski, illustration in Zeit- schr. f. wiss. Zool., XII, t. 50, 4. KfiUiker saw the crystals, 1849. Eolliker saw the crystals first (Todd's Cyclop, of Anatomy and Physiology, 1849, pt. 36, Lond., p. 792, "Spleen"). Remak found them 2 hrs. after death in the blood-vessels (MuUer's Archiv, 1852, 121). Illust. in KBUiker's Handbuch der Gewebelehre, . 1863, 4 Aufl., p. 627). Illustration in Zeitschr. f. wisa. Zool., XII, taf. 30, 11 (Bojanow- ski). Ankersmit, Diss., p. S3. Illustration in Zeitschr. f. wiss. Zool., xn, 334, fig. 10 (Bojan- owski). Illustration in Zeitschr. f. wiss. Zool., XII, taf. 30, 10 (Bojan- owski). The crystals are formed if a drop of earthworm blood is allowed to evaporate slowly. Zeitschr. f. wiss. Zool., i, 1849, p. 116. Illustration in the same. Leydig, Lehrb. d. Histologic, 1857, 446, taf. 8, fig. 34 B. Eztraglobular (Remak) and intraglobular (Fiinkn) Even more read- ily soluble than the tench-blood crystals (Re- Spindle- and nee- dle-shaped crys- tals Intraglobular and extraglob- ular (Fimkfi) Extraglobular. . . . Do Crystalliies very readily Do Small, thin plates tapering at both ends Crystals dissolve very easily in water (Remak) Extraglobular and intraglob- ular Extraglobular. . , . Intraglobular .... As in the red-eyed roach (Remak) Very readily solu- ble Crystallizes very easily Crystallizes with extraordinary difficulty Plates and rods (Bojanowski) Very probably rhombic 4-sided prisms . . . Do Probably rhom- bic Do Do Very readily solu- ble Crystallizes eas- Very delicate nee- dle-shaped crys- tals (Preyer) Small tabular plates, little rods and col- umns (Leydig) Do In the stomach of Clepsine. CHAPTER VI. THE PREPARATION AND CRYSTALLOGRAPHY OF HEMOGLOBINS SINCE PREYER'S INVESTIGATIONS. Since the appearance of Preyer's monograph very Uttle progress has been made either in the methods of preparing the blood crystals or in the study of their crystalUne characters, although much has been added to our knowledge of hemoglobin in certain other directions. Such simple methods as have been described for preparing crystals in small quantities, together with Hoppe-Seyler's method for preparing them in large quantities, have proved so satisfactory for general laboratory purposes that there has been little encouragement to seek new processes; while Preyer's long list of crystals from different species, and the assignment of all of them to the rhombic system, except those of the squirrel, seem to have discouraged research along the lines of crystallography. In fact, with rare exceptions when crystals from a new species have been isolated, the observer has been content without further inquiry to record them as being rhombic. We are therefore now, so far as the crystallography of hemoglobin is concerned, virtually where we were when Preyer's monograph was published (1871). The method of furthering crystallization by the putrefactive process, already pursued by a number of observers, was adopted by Gscheidlen (Archiv f. ges. Physiologie, 1878, xvi, 421), who placed defibrinated blood in a glass vessel with little air, and kept it in an incubator until the absorp- tion spectrum showed the absence of oxyhemoglobin. When a drop of this blood was placed on an object-glass, allowed to evaporate sUghtly, and then covered with a cover-glass, crystals appeared under the eye of the operator. From dog's blood which had stood for several days in the incubator he obtained crystals from 3 to 4 mm. long. The rapid crystallization of the blood thus prepared he found to be due to putridity, since blood kept in sterilized vessels under the same conditions showed far less power of crys- talUzation. In guinea-pig's blood kept in the incubator with the admission of air he found not only large tetrahedra, but also rhombic plates and prisms. By this method Gscheidlen prepared crystals from the blood of the dog, guinea-pig, sheep, bullock, rabbit, and goose. He also noted that blood kept in hermetically sealed tubes for several years crystaUized in a short time upon exposure to the air. The readiness with which putrid blood crystallizes had already been noted by Schmidt, Bottcher, Blebs, and others. A method by Ktihne and Gamgee (Gamgee's Physiological Chemistry, 1880, 87) is as follows: 500 c.c. of defibrinated dog's blood are treated with 31 c.c. of ether and the mixture shaken for some minutes. It is then set aside in a cool place. After a period varjdng from 24 hours to 3 days the liquid becomes converted into a thick magma of crystals. The crystals 107 108 PEEPARATION AND CRYSTALLOGKAPHY OF HEMOGLOBINS may be separated by placing the mixture in tubes and using the centrif- ugal apparatus. The cakes of crystals thus obtained are mixed with water holding one-fourth its volume of alcohol, and again centrifugalized. By repeating this process, the crystals are said to be obtained free from serum- albumin. If requisite, the crystals are recrystallized by dissolving them in as small a quantity of water as possible at 25° to 30°, cooling the solution to 0°, and ad(fing a fourth of its volume of alcohol. It is better to place the fluid in a freezing mixture at a temperature of —10° to —20° for 24 hours. Crystals of reduced hemoglobin were prepared by Hiifner (Zeit. f. physiol. Chemie, 1880, iv) from human blood, diluted or not, by placing the blood in tubes from which air is excluded. After standing for a month or two at summer temperature the blood became of a beautiful purple color, and in many spots on the inner wall of the tubes there could be seen whole layers of purple-red crystals, which upon spectroscopic examination were found to give the characteristic bands of reduced hemoglobin. Wedl (Archiv f. path. Anat. u. Physiologie, 1880, lxxx, 172) obtained reduced hemoglobin crystals expeditiously by subjecting a solution of fresh or dried blood in a confined atmosphere in the presence of a solution of pyrogalUc acid. The acid absorbs the oxygen and thus reduces the hemoglobin. In this way crystals of reduced hemoglobin were prepared within 24 hours from the blood of man, the rabbit, hare, deer, pig, and sheep. Crystals of reduced hemoglobin were prepared in large quantities by Nencki and Sieber (Berichte d. d. chem. Ges., 1886, xix, 128, 410), who, however, make the erroneous statement that no one had up to that time prepared crystals of reduced hemoglobin. Kiihne (Archiv f. path. Anat. u. Physiol., 1865, xxxiv, 423), and shortly after Rollett (Sitzungsb. d. k. Akad. d. Wissensch., Wien, 1866, lii, 246), obtained crystals of reduced hemoglobin by reduction of concentrated solutions of oxyhemoglobin. Ktihne used a very concentrated solution of oxyhemoglobin in very dilute ammonia, which he subjected to a stream of pure dry hydrogen in a glass chamber. As evaporation proceeded crystals formed. Rollett (loc. cit.) prepared reduced hemoglobin by the aid of iron filings. Gscheidlen in 1878 Q^c. cit.) and Hiifner (Jx)c. cit.) and Wedl (J,oc. cit.) in 1880 also prepared crystals of reduced hemoglobin. Nencki and Sieber proceed in this way: Pure oxyhemoglobin crys- tals from the blood of the horse are dissolved in lukewarm water; the solution is then mixed with several cubic centimeters of decaying blood in a flask that is provided with an india-rubber stopper having two perfora- tions for tubes leading to and from the flask. The mixture is then freed from air by the passage of a stream of hydrogen, after which the two tubes are sealed by heat, and then the flask is set aside at a temperature of 20° to 25° fot- 8 to 14 days. After a time every trace of oxygen has dis- appeared, the fluid is of a beautiful violet-red color, and contains only re- duced hemoglobin. The solution is now cooled to 0°, an india-rubber tube is for some distance slipped over the outlet tube of the flask, and the other end of the tube is dipped in cold absolute alcohol. The flask is gently heated by immersing in lukewarm water, the end of the glass tube within SINCE preyer's investigations. 109 the rubber tube is broken off, and by alternate cooling and heating of the flask sufiicient alcohol is introduced so that the solution contains about 25 per cent of alcohol. The free end of the rubber tube is now closed by a screw clip and glass stopper and the solution is subjected to a temperature of 5° to 10°. After 12 to 24 hours the reduced hemoglobin has crystallized into glittering plates and prisms. When examined under the microscope at 0° in the mother-liquor, the crystals for the most part appear as 6-sided plates, of which some were from 2 to 3 mm. in diameter. In the micro- spectroscope every crystal showed only the one band of reduced hemo- globin. The prismatic crystals are doubly refracting. The color of the larger plates is a beautiful violet red; the smaller thin plates appeared greenish in transmitted light. The crystals were very sensitive to oxygen and warmth. At room temperature they quickly melt, and as quickly they lose their violet color and show by the microspectroscope the bands of oxyhemoglobin. In absolute alcohol they remain unchanged, at least in so far as their form is concerned. If the hemoglobin solution is mixed too soon with alcohol, before the bacteria have taken up the last traces of oxygen, both reduced-hemoglobin and oxyhemoglobin crystals are formed. Besides the differences they describe in the color and spectroscopic behavior Nencki and Sieber also make note of differences in the forms of oxyhemoglobin and reduced hemoglobin. From horse's blood they obtained oxyhemoglobin in long 4-sided columns, and the reduced hemoglobin in thin 6-sided plates which are more soluble in water than the oxyhemoglobin. In horse's blood which has decomposed in well-closed or sealed vessels, the reduced hemoglobin separates as a thick crystal pulp on the addition of alcohol after standing several hours at a temperature under 0°. Gamgee (Schafer's Text-book of Physiology, 1898, 1, 232) gives a method for preparing reduced hemoglobin which he states he employed 20 years previously, and which seems to him to possess some advantages: A magma of pure oxyhemoglobin crystals and a small quantity of the mother-Uquor are placed in a glass tube so as nearly to fill it, and the tube sealed and heated for some days in an incubator at about 35° and then set aside in a cool place. After some weeks of exposure at winter temperature crystals of reduced hemoglobin will be found. Crystals of reduced hemo- globin have also been prepared by Ewald, Frey, UhUk, Copemann, Dono- gdny, and others, as will be shown by subsequent references. The changes in solubility of crystals of hemoglobin that are caused by alcohol were studied by Struve (Ber. d. d. chem. Ges., 1881, xiv, 930; Jour. f. prakt. Chemie, 1884, xxix, 304), who found that fresh crystals placed in strong alcohol immediately became darker, without change of form, and insoluble in water. Upon treating crystals with dilute alcohol, they became faintly yellowish or completely decolorized. These and other phenomena led Struve to resurrect the long since abandoned view that the blood crystals are composed of a colorless albuminous substance which is stained or colored mechanically. After leeches have sucked blood and crystallization has begun, speci- mens of hemoglobin crystals may be obtained from time to time, as shown 110 PREPARATION AND CRYSTAI.LOGRAPHY OF HEMOGLOBINS by Stirling and Brito (Jour. Anat. and Physiology, 1882, xvi, 446), by causing the leeches to disgorge. To do this they appUed pressure, an 8 per cent salt solution, weak to strong acetic acid, 2 to 1000 solution of sulphuric acid, or galvanic or faradic shocks. Within 20 days, hemoglobin crystals appeared, which was much earlier, they state, than was noted by Budge and Bojanowski; but no hemin crystals were found, as were thought by Bojanowski to be present at times. Even after a year and a haU they found dusky-red purplish crystals of reduced hemoglobin of human blood in the form of 4-sided prisms, some of them nearly equal-sided, while others were oblong. From the stomach of the leech they obtained crystals from the blood of the common gold-fish, and also obtained crystals from sealed microscopic preparations of the diluted fish blood. From the blood of the frog they secured both colored and colorless crystals of exactly the same form. The former they describe as being very variable in size, highly refractive, acicular, and pointed at one extremity like the point of a pen. Stirling and Brito note that colorless crystals of frog's blood had also been discovered by Teichmann {loc. ciL), who mixed the defibrinated blood with water and evaporated at low temperature. Besides obtaining these crystals from the blood of the stomach of the leech they also prepared them by mixing 5 or 6 drops of the freshly drawn blood from the heart with one or two drops of distilled water, and then seaUng up the preparations with gold size. They state, however, that exposure to the air favors the formation of the crystals, which first form around and in the neighborhood of coagula. In the case of one of the leeches, on exposing some of the blood on the fourth day, they obtained blood which, when sealed up and allowed to stand, developed beautiful colored crystals of exactly the same shape as those which are colorless. The sole difference, they state, was in the color, and they therefore were inchned to regard the latter as being closely related to hemoglobin, if not identical with it. They did not find any crystals from the blood of the newt that had been ejected by the leech. Studies were also made of the influences of certain reagents on the crys- tallization of rat's blood. Stirling and Brito found that common salt and urine prevented the diffusion of hemoglobin from the corpuscles, and there- fore prevented crystallization; but a weak solution of pure urea behaved exactly like water, liberating the hemoglobin and thus permitting of crys- tallization. From this they conclude that the presence of common salt in the urine is sufficient to neutralize the effect of the urea. The crystals found in the solution were exactly the same as those formed after the addition of water. Crystals appeared in a few minutes when chloroform was freely mixed with a drop of rat's blood on a slide and covered and examined in the usual way, but the ordinary flattened prisms with beveled ends were shortened so as to be hexagonal. They also made the interesting observation that the passage of a galvanic current causes a deposition of crystals equally well at both negative and positive poles, but that the induced current was without effect. The use of chinolin to increase crystallizability was reported by Otto (Zeit. f . physiolog. Chemie, 1882, vii, 57) . He employed an alcoholic solution SINCE preyer's investigations. Ill or an aqueous solution of the hydrochlorate of chinolin, and by its aid pre- pared crystals of pig's blood. He notes that Hiifner previously found that the blood of the pig mixed immediately with one-third of its volume of a 1 per cent alcoholic solution of chinohn crystallizes beautifully when sub- jected to cold, the mixture after several days containing a mass of needles and plates which liquefied within an extremely short time when exposed at room temperature under the microscope. Otto used chinolin solutions and blood in varying proportions, as follows: (a) 100 c.c. of blood, 40 c.c. of 1 per cent chinolin hydrochlorate solution, and 30 c.c. of alcohol; (&) 100 c.c. of blood, 30 c.c. of chinoUn solution, and 30 c.c. of alcohol; (c) 100 c.c. of blood, 25 c.c. of chinolin solution, and 25 c.c. of alcohol. The mass of crys- tals which had collected during 8 days was washed on a filter-paper with alcohol (diluted 4 times) and then dissolved in a small quantity of water. Adding to this solution one-eighth its volume of alcohol, the mixture was again placed in the cold, whereupon crystals sometimes separated within a few days. As a rule, the second crystallization failed to occur, and instead a mass separated out in from 8 to 14 days, which was found to be met- hemoglobin. The unsatisfactory results of this method led Otto to adopt what is practically the Hoppe-Seyler method : The blood was diluted with salt solution and stood in a cyhndrical vessel for two days. The corpuscles were collected and dissolved in the smallest possible quantity of water at 50°, 300 c.c. of water being sufficient for the solution of the corpuscles from 1 liter of blood. Owing to the unusual solubility of the crystals of pig's blood, which liquefy at room temperature, it is very important, as Otto states, to avoid an excess of water in dissolving the corpuscles. The solution is filtered, cooled, mixed with cold absolute alcohol in the usual proportion of 4:1, and then subjected to cold. As a rule, after only one day a thick mass of fine, bright-red needles was found at the bottom of the cylinder. For the purpose of recrystallization, the crystals were collected upon a folded filter- paper, washed, and crystallized 3 times with dilute alcohol in the ice-chest. He prepared dog's crystals in the same way. The crystals were finally spread upon plates and dried under a bell-jar over sulphuric acid in the cold. The crystals of pig's blood thus prepared were then powdered, heated to 115°, and subjected to a stream of hydrogen, when they gave off 5.9 per cent of water. Those of the dog similarly treated lost only 4 per cent of water. Both kinds of crystals were subjected to elementary analyses (page 71). Otto also analyzed the methemoglobin of the pig. Studies were also made of the extinction coefficients (page 77) and of the oxygen capacities. The crystals were determined by the spectroscope to be oxyhemoglobin. In a later research (Archiv f. ges. Physiologie, 1883, xxxi, 240) Otto pre- pared crystals of horse's blood, which he also subjected to elementary analysis and spectroscopic examination. In his former investigation he determined that the extinction coefficients of the oxyhemoglobin of the pig and dog are the same (1.33), and in this inquiry the extinction coefficient of horse oxyhemoglobin was found to be 1.352. His elementary analyses are given on page 71. He also noted the observation of Hoppe-Seyler 112 PREPABATION AND CRYSTALLOGRAPHY OF HEMOGLOBINS (Zeit. f. phys. Chemie, 1878, 11, 149) that the crystals of horse hemoglobin appear to be of two kinds (needles and prisms) which differ in solubiUty — a difference which Hoppe-Seyler thought likely to be due to differences in the amount of water of crystaUization. Otto states that there continually appeared, besides little needles which are in relatively greater abundance in recrystaUization, long, thick prisms which prevail in the first crystallization. He endeavored, but failed, to separate the needles from the prisms by wash- ing with dilute alcohol, as Hoppe-Seyler states could be done. He also tried to determine differences in the water of crystallization, but he failed to obtain concordant results. Crystals from horse's blood were prepared by Hiifner and Biicheler (Zeit. f. phys. Chemie, 1884, viii, 355) by the ordinary alcoholic method and recrystallized three times in a refrigerator. Generally needles were obtained from 2 to 3 mm. long and 0.5 mm. wide. Once they found hex- agonal tablets of reduced hemoglobin, which changed quickly upon coming in contact with the air. Dried at 0° over sulphuric acid and anhydrous phosphoric acid, the crystals retained 3.94 per cent of water, which came off when the crystals were subjected to a stream of hydrogen at 115°. They made elementary analyses (page 71), calculated the molecular weight and formula (page 75), and determined the oxygen capacity. A new method for preparing hemoglobin crystals was reported by von Stein (Centralblatt f. med. Wissensch., 1884, xxii, 404; Archiv f. path. Anat. u. Physiologie, 1884, xcii, 483), which is applicable to small quan- tities of blood that are readily crystallizable. A drop of defibrinated blood or blood squeezed from a clot was placed on an object-glass and exposed to the air until it began to dry up at the margins. Canada balsam was then applied, first around the drop of blood, in order to prevent any possible escape, and then the remaining space above it was filled. It is to be observed, von Stein states, that the center of the drop of blood is pushed off to the periphery. In this way a clear space is made for crystallization, otherwise the crystals are so small that their outlines can not be made out. Too thick a layer of blood is to be avoided, because the balsam does not penetrate to the deeply Ijdng portions. Von Stein proceeded in another way, by not allowing the blood to evaporate, and by treating it immediately with the reagent and covering the mixture with a cover-glass. Canada balsam is best when it appears yellow and not entirely clear. In Uquid balsam the crystals form more quickly, and sometimes have larger dimensions, but they soon become brown (in one or two days), then dull and black, and in a short time are fissured to small pieces. Preparations can be made which retain their form and color for years if the balsam has been exposed to the air for a long time, or is evaporated to such a consistence that it can be drawn out into transparent but not milky threads when lifted with a glass rod. Whichever method is used, it is important that the preparation be left uncovered in the air until the crystallization has been completed, and until the odor of the balsam has completely disappeared, which lasts ordinarily a few days. Then with a knife iromersed in ether, turpentine, or oil of cloves (little should be used of either), the upper portion of the balsam is removed, SINCE preter's investigations. 113 and the whole covered with a cover-glass and sealed with asphalt or balsam. Crystals from human, horse, guinea-pig, and rat blood were obtained by the above methods. Von Stein's methods were extended by Smreker and Zoth (Sitzungsber. d. Wiener Acad., 1886, xcni, Abth. iii; Maly's Jahr. li. d. Fort. d. Thier- chemie, 1886, xvi, 102), who used Canada balsam, turpentine, Peru and other balsams; solutions of colophony, damar, and mastic dissolved in xylol; fixed oils; xylol solutions of roan; fatty acids, etc. The doubt as to whether or not hemoglobin is a chemical individual, together with the fact of the discrepancies in the centesimal analyses of hemoglobin, led Zinoffsky (Zeit. f. physiol. Chemie, 1886, x, 16) to prepare crystals of hemoglobin in several ways and to make careful determinations of the iron and sulphur contents. In preliminary experiments he found that the washing of the corpuscles by common salt solution, according to the directions of Hoppe-Seyler, is not only superfluous, but also undesir- able, because the washing introduces the danger of decomposition, owing to the fact that from 3 to 5 days are required in the process, and iDecause it is not of importance in removing the small quantity of protein in solution. In experiments in relation to the separation of the hemoglobin from the stromata he found that, when the corpuscle pulp is heated to 35° with 3 volumes of distilled water, the hemoglobin dissolves and crystalhzes and that the stromata remain undissolved and cUng so tenaciously to the hemoglobin crystals that they can not be removed by filtration. They must, therefore, be dissolved before the crystallization of the hemoglobin, either (1) by the addition of very Uttle ammonia to the fluid heated to 35°, which must then be carefully neutralized with dilute hydrochloric acid (according to the direction of Schmidt), or (2) by the addition of ether (30 e.c. of ether being sufficient for 9 liters of blood). To crystallize the hemoglobin the solution was cooled to 0°, mixed by titration with one-fourth its volume of absolute alcohol, and left standing for 72 hours. The crystals were washed by decantation with a mixture of 1 part of alcohol to 4 parts of water cooled to 0°. To obtain pure crystals, the crystals were dissolved in 3 volumes of distilled water at 35°, the solution was filtered, and the filtrate was mixed with dilute alcohol as before. Tests showed that two recrystallizations of the first product sufficed to obtain pure crystals, Zinoffsky also makes note of the fact that the drying of the crystals in vacuo at 0°, according to the directions of Hoppe-Seyler, is an exceptionally lengthy process, and that the crystals can be dried in about 8 hours at 18° to 20° without being placed in a vacuum. After these prehminary investi- gations he prepared crystals from horse blood by three methods, viz : First method: 20 liters of horse's blood were defibrinated; the blood- corpuscle pulp, which after 3 hours' standing in the cold had been deposited, was separated from the serum and mixed with 8 volumes of a 2 per cent solution of common salt. After 3 days the corpuscles were collected and placed in 3 volumes of distilled water at 35°, to which were then added 16 c.c. of one-tenth normal ammonia solution. After 5 minutes the ammonia was neutralized by titration with a very dilute hydrochloric acid. The 8 114 PEEPAEATION AND CRYSTALLOGRAPHY OF HEMOGLOBINS mixture was quickly cooled to 0°, to which was then added 1 volume of absolute alcohol at 0° to every 4 volumes of the solution. After 3 days' standing in an ice-and-salt mixture, the crystals were collected and washed twice with alcohol and water (1 : 4) at 0°, and then, in the way stated, recrys- tallized twice and finally dried by the aid of the air-pump. This preparation yielded 200 grams. Second method: The blood-corpuscle pulp obtained by decantation from 10 hters of blood, and without washing with salt solution, was dis- solved in 3 volumes of water at 35°, and then treated as in the first method. The yield was 520 grams. The very much larger quantity thus obtained led Zinoffsky to believe that washing with salt solution reduces the yield. Third method: The blood-corpuscle pulp from 9 liters of blood was dissolved immediately in 3 volumes of distilled water at 35°, then cooled; 30 c.c. of ether were added instead of ammonia, as in the first method, and then the solution was treated as in the first method. The product by the third process was the purest, the ash containing but a trace of chlorine, no alkalies, and only imponderable quantities of phosphorus, Ume, and magnesia. The product by the first process con- tained 0.0235 per cent of phosphoric acid. The second preparation was the least pure. It contained 0.0401 per cent of phosphoric acid, 0.0097 per cent of CaO, and 0.0131 per cent of MgO. Table 32. — Table from Zinoffsky, showing the 'percentages of sulphur and iron, the number of atoms of sulphur to each atom of iron, and the amount of sulphur and iron in the ash of hemoglobin. Material and author. Dog's blood: Schmidt Hoppe-Seyler Do Do Horse's blood: Biicheler Do Do Kossel Otto Per cent. 0.66 .375 .448 .359 .6532 .6443 .65 .67 Fe. Per cerd. 0.43 .45 .42 .42 .46370 .47238 .46720 .47 .45 Atoms of S to each atom of Fe. 2.686 1.60 2.427 2.42 2.60 Amounts of S and Fe in the ash of hemoglobin. 1.6637 1.9972 1.4033 1.8915 1.8481 1.8132 2.2876 Fe. 1.6637 1.7062 2.5122 4.7855 0.9439 4.0866 Zinoffsky in the earlier part of his article shows (table 32) the marked discrepancies in the results of the analyses by different observers of speci- mens of bloods from different individuals of the same species. They are also of particular interest in connection with the figures obtained by Zinoff- sky in this research. The two sulphur determinations of the first preparation were 0.3902 and 0.3916 per cent; of the second preparation, 0.3583 and 0.3658 per cent; and of the third preparation 0.3899 and 0.3881 per cent. In the determi- nations of iron he found in the first preparation 0.325 to 0.327 per cent and in the third preparation 0.334 to 0.338 per cent. These results show, he SINCE preyer's investigations. 115 states, that there are 2 atoms of sulphur to 1 atom of iron. The mean of his elementary analyses is C6i-15H6-76N17.94S0-3899F60.335O23.4251 and the molecular formula ^712111130^21482^60245 Comparing Zinoffsky's percentages with those of other analysts (see page 71), it seems as though there must be errors in his carbon and hydro- gen estimations. Moreover, his iron and sulphur determinations differ materially from those of others, yet his analyses were conducted in such a way as to warrant confidence in these figures. The low C content is cer- tainly suggestive of imperfect combustion, or, according to Hiifner, of contamination with stromata. Zinoffsky's work has been reviewed and supplemented by Hiifner (see later). The optical properties of oxyhemoglobin, reduced hemoglobin, met- hemoglobin, hemin, and CO-hemoglobin were studied by Ewald (Zeit. f. Biologie, 1886, xxii, 459). He laked the blood by repeated freezing and thawing, and then spread layers of varying thickness upon microscopic slides. The margins of the preparations soon dry, and then a cover-glass is placed directly on the blood or supported by a wedge of glass. If the preparations are examined immediately, only oxyhemoglobin crystals will be found; but after several days violet-purple spots appear which consist of reduced hemoglobin, but which soon pass into solution. He also obtained crystals of reduced hemoglobin by letting the blood stand in tubes for several days; in the deeper layers the oxygen disappears and crystals of reduced hemoglobin form. He found the crystals of oxyhemoglobin and reduced hemoglobin to be doubly refracting and pleochroic, and that the pleochroism is much more marked in reduced hemoglobin. In a research to determine whether or not the 6-sided crystals of certain rodents really belong to the hexagonal system, and to find an explanation of the difference in crystalline form that hemoglobin presents in different animals, HaUiburton (Jour, of Physiology, 1886, vii, Proc. Physiol. Soc. No. 1; Jour. Microscop. Science, 1887-88, xxviii, 181) carried out a series of observations chiefly with the bloods of the rat, guinea-pig, and squirrel. The rat was taken as a type of animals whose crystals are rhombic; the guinea-pig, of those whose crystals are tetragonal; and the squirrel, of those whose crystals are hexagonal. HaUiburton notes that Lehmann states, without giving any reason, that although the crystals of the squirrel are hexagonal in fonn they do not belong to the hexagonal system, and that von Lang, Kunde, and Preyer state that they do. In examinations of squirrel's crystals by polarized Ught he found evidence, as he believes, of their being true hexagons instead of their being, according to Lehmann, rhombic plates with an "hexagonal habit." It had already been found by von Lang that the tetrahedra of the guinea-pig belong to the rhombic system. In experiments instituted to show whether differences in crystaUine form are due to some agency extrinsic to the hemoglobin or to some property 116 PREPARATION AND CRYSTALLOGRAPHY OP HEMOGLOBINS inherent in the hemoglobin, he in some instances mixed serum, or the serum and the stromata, or the blood of one species of animal with the blood of another, or solutions of hemoglobins of different species with one another. The presence of the foreign serum, or serum and stromata, was without influence on crystalline form, and, while mixed bloods, or mixed hemoglobin solutions, did not affect crystalUne form, they sometimes caused modifica- tions in crystalline habit. Thus, in case of the bloods or solutions of hemo- globins of the rat and guinea-pig the crystals of the rat were rhombic with hexagonal habit, no needles or tetrahedra being present. (See table 33.) Table 33. — Forms of hemoglobin crystals in case of mixed bloods {from Halliburton). Blood of— • Mixed with that of — Form of hemoglobin crystals from the mixture. Rat Sauirrel Both rhombic prisms and hexagons present. No rhombic prisms of the shape usu- ally seen in rat's blood present. No tetrahedra. Crystals are all rhom- bic prisms \nth hexagonal habit. Hexagonal plates and tetrahedra both present. Many tetrahedra imper- fect. The tetrahedra were all re- duced to about half the size of those prepared from the unmixed blood of the same guinea-pigs. Fine rhombic needles and hexagonal plates both present in abundance. The greater number of the crystals formed are very small tetrahedra, about a quarter the size of those prepared from the blood of the same guinea-pigs. The optical properties are, however, the same. Rhombic prisms very slender, like those of dog's blood, also seen. Rat Guinearpig Guinea-pig Sauirrel "Doe Doe Guinearpig In another set of experiments HalUburton tried to break down the hexagonal constitution of the hemoglobin of squirrel's blood, first, by draw- ing off the water of crystallization and then adding water; second, by converting the hemoglobin into methemoglobin, and then by reducing agents to form once more hemoglobin, and to obtain crystals from this. Both attempts were unsuccessful. In opposition to the statement of Preyer that recrystalUzation does not alter the form of the crystals, HalUburton found that by recrystaUiza- tion of squirrel's hemoglobin, after 3 or 4 recrystallizations no 6-sided crystals were obtained, but a mixture of rhombic needles and tetrahedra, and that in some cases the latter were absent. In conclusion, the author states that the difference between the various forms of hemoglobin can not be a very deep or essential one, and that it seems to narrow itseK down to this, either we have a case of polymorphism or the crystaUine forms are due to the combination with varying proportions of water of crystalUzation. SINCE preyer's investigations. 117 In the second contribution referred to, Halliburton adds the following to our crystallographic data: Opossum {Didelphis cancrivora) . — Very large dark crystals can readily be obtained. They belong to the rhombic system. Kangaroo {Macropus giganteus). — Crystals are more soluble, and so less readily obtained. They are rhombic prisms, slenderer than in the opossum. Sugar squirrel {Belideus breviceps — a marsupial). — Crystals similar to those of opossum. Seal {Phoca vitvlina) . — Rhombic prisms, many of them very short and simulating hexa- gons. Easily obtained. Bear {JJrsus syriacus). — Bunches of rhombic needles, easily obtained. They are slen- derer than those obtained from dog's blood, as a rule, some being almost silken in appearance. White-bellied beaver rat (Hydromys leucogaster). — Rhombic prisms. White-whiskered swine {Siis leucomytax). — Rhombic prisms. Water vole (Arvicola aquatica). — Crystals are obtained easily by adding water to the blood. They are of the usual rhombic shape. The analyses of hemoglobin of horse's blood by Zinoffsky {loc. cit.) differed so much from those of previous observers that Hiifner (Beitrage z. Physiologie, Fest. f. Carl Ludwig, 1887, 74) was led to review and supple- ment Zinoffsky's work. Hiifner prepared hemoglobin crystals by a process that is a modification of Zinoffsky's to the extent essentially of separating the stromata of the corpuscles by mechanical instead of chemical means, that is by centrifugalization, so that the crystals could be freed from the stromata dissolved or undissolved and more expeditiously prepared. Crys- tals were obtained from the bloods of the pig and ox by centrifugaUzing the corpuscles, extracting the hemoglobin from them by distiUed water at 30° to 40°, cooling to 0°, centrifugalizing and treating by the usual method. After the crystals have formed they are centrifugaUzed in the cold to pre- vent their solution, and the hemoglobin is then three times crystalUzed by the usual method, and finally dried in an atmosphere at 0°. The ash of 10 grams of this product contained only an imponderable amount of phosphoric acid. The mean figures of his elementary analyses are as follows: Pig's oxyhemoglobin, C54.7iH7.38Ni7.43So.479Feo.3390i9-602 Ox's hemoglobin, C54.66H7.25N17.76S0.447Fe0.40O19.543 In comparing these figures with those of Otto {loc. cit), Hiifner states that the complete removal of the stromata in his preparations causes a higher percentage of C and N, Otto having found C54.17 and N16.23. Zinoffsky's C content (51.15) was very much lower than Hiifner's. Hiifner's analyses show the same ratio of S and Fe in both pig and dog hemoglobins, i.e., 2 of sulphur to 1 of iron, the same as Zinoffsky found with horse hemoglobin. The elementary analysis of the hemoglobin of the dog which was reported the following year by Jacquet (Zeit. f. physiol. Chemie, 1888, XII, 285) was of crystals prepared as follows: The corpuscles were centri- fugaUzed, then mixed with 2 volumes of water warmed to 35°, then cooled and shaken with ether and treated according to the Hoppe-Seyler process. The crystals were twice recrystallized, and then analyzed according to the methods pursued by Zinoffsky, but he endeavored to eliminate certain possible fallacies in the iron determinations. His analyses gave a mean C53.9iH6.62Ni5.98Feo.333022'62 118 PREPARATION AND CRYSTALLOGRAPHY OF HEMOGLOBINS These figures do not agree with those of Zinoffsky for the hemoglobin of the horse, or with those of Hiifner for the hemoglobins of the pig and ox. The relation of Fe to S in dog's hemoglobin, Jacquet found to be 1 : 2.85; Zinoffsky found 1 : 2 for the horse; and Hufner found 1 : 2 for the pig and ox. Jacquet beheves his sulphur value to be too small, and that there is 1 atom of Fe to 3 of S. Later (Zeit. f. physiolog. Chemie, 1889, XIV, 289) he analyzed crystals of the hemoglobin of the dog, which he prepared by a modification of his previous process. The earlier method was used, except that the solution of the corpuscles after the addition of the ether was centrifugaUzed in a machine running at from 1,600 to 2,000 revolutions, whereby the stromata could be partly done away with. The hemoglobin 3 times crystalUzed contained only a trace of phosphoric acid, the quantity not being estimated. The mean of his analyses was C54-67H7.22Ni6-38So-568^®0-336C>20-93 The amount of water of crystallization was 11.39 per cent. The ratio of Fe to S was 1 : 2.96, which was higher than in his preceding investigation. From the values obtained he gives the formula C75gHi203Ni95FeO2i8, and the molecular weight as 16669. He also analyzed the oxyhemoglobin of the chicken. In the prepara- tion of the crystals he made a special effort to get rid of the large amount of phosphoric acid (0.77 per cent) shown in the preparations of goose crystals by Hoppe-Seyler. He found that he could not treat the blood in precisely the same way as dog's blood, because when the corpuscles are agitated with ether a gelatinous mass is formed which could not be filtered. The corpuscles were therefore treated with an equal volume of water and one-third volume of ether. The mixture when heated to 35° formed into dark-red, gelatinous lumps which were separated from the fluid by centrifugahzation. The clear fluid thus obtained was readily filtered, and by the customary treatment very soluble needles of hemoglobin were formed. By recrystallization both rhombic plates and prisms were obtained. The three times crystallized hemoglobin was found upon analysis to have the following composition : C52-47H7.i9Ni6.45So.8586Feo-3353022-5Po-1973 The water of crystallization was 9.333 per cent, and the ratio of Fe to S 1 : 4.485. If the molecule be doubled the ratio is 2 : 9. In comparing the analyses of the oxyhemoglobin of the dog, chicken, and horse, he states that although these hemoglobins are different they have a similar iron capacity, which warrants the conclusion that the iron-containing group in the various hemoglobins is the same. Jacquet made ineffectual attempts to prepare crystals from fresh salmon blood, but succeeded when the blood was left to rot, there appearing clusters of crystals and beautiful single rhombic prisms. Jolin (Archiv f. Anat. u. Physiologie, 1889, 265) records that the hemo- globins of the dog and guinea-pig differ from that of the goose in their absorptive rapidity in relation to 0, as well as in the volume of absorbed. The increased crystallizabiUty of putrid blood has been noted by a number of observers and referred to in previous pages, and Bond (London SINCE pebyer's investigations. 119 Lancet, 1887, ii, 509, 557) has added to our knowledge in this particular by showing a relationship between crystaUizabihty and septic conditions in the body. He found that if a drop of blood were taken from the cleansed finger of a patient who is suffering severely from absorption of the prod- ucts of putrefaction, and that if such drop be placed between a sUde and a cover-glass and allowed to remain at room temperature (60° F.), in the course of 20 to 30 hours crystals of reduced hemoglobin of prismatic and needle form will be found, while within some corpuscles Uttle bars and needles may plainly be seen, apparently distinct from the enveloping stroma. He also found that adding putrid blood faciUtates crystalUzation, and that in cases of pernicious anemia crystalUzability seemed to be increased. The increased crystaUizabihty of human hemoglobin in pernicious anemia that was pointed out by Bond (]Loc. cit.) was later noted by Cope- mann (Journal of Physiology, 1890, xi, 401), who found that when a drop of blood from the finger of a patient thus affected was allowed to fall on a glass slide, the edge of the drop allowed to dry, and a cover-glass placed on the blood, crystals of hemoglobin gradually formed in from 10 to 48 hours. The only exception to this was in the case of patients who had been treated with arsenic for some days, although crystals were obtained upon the discontinuance of the arsenic. To imitate the influence of septicemia, as was also shown by Bond, Copemann treated the blood with decomposing serum. This method he found to be successful in the case of the bloods of the bullock, sheep, pig, dog, and cat, but unsuccessful for the blood of man, the monkey, rabbit, and squirrel. Except in the case of man and monkey the crystals were of oxyhemoglobin, and this notwithstanding that the decomposing serum invariably brought reduction of the oxyhemoglobin as it diffused from the corpuscles into the plasma. He states that this occurred to the greatest extent just inside of the edge of the cover-glass, but not extending to the edges where the layer is kept oxidized; and that it is in this intermediate zone of fuUy reduced hemoglobin that crystals are to be found in the great- est quantity, both in case of human and monkey blood and of that of the rabbit and squirrel; but in the latter the crystals are of oxyhemoglobin, while in the former they are of reduced hemoglobin. He also made the interesting observation that in specimens of squirrel's blood (species not stated) the crystals were in every instance in the form of fine needles and rhombic prisms, the needles sometimes being collected into bundles, while the usual hexagons were absolutely absent. Copemann also prepared crystals from the blood of the horse, bullock, sheep, pig, dog, cat, squirrel, rabbit, guinea-pig, rat, mouse, and chicken by the following simple process: The blood is shaken with ether (16 : 1) and then kept under an atmosphere of ether for some time, which may be accomplished by performing the agitation of the blood with ether in a stoppered bottle and gradually allowing the air to escape as the ether is volatiUzed. By this means the contained air is gradually replaced by ether vapor, while at the same time the small portion of blood which is forced out around the stopper of the bottle on drying fixes it in its place and so prevents 120 PREPARATION AND CRYSTALLOGRAPHY OF HEMOGLOBINS the ingress of air again. It seems also that it is better to leave the bottle in a room at ordinary temperature than to put it in a cool place, as advised by Gamgee. After a variable time, in case of most animals at least two days, a drop of blood is placed upon a slide, and when the margin of the drop is slightly dry a cover-glass is gently lowered on the surface of the drop. The formation of crystals will often be seen within an hour or so. Human blood subjected to the same process does not usually yield crystals, but when crystals do appear they invariably present the appearance of reduced hemoglobin. Copemann also obtained crystals from human blood by the use of bile' — ^preferably, as he states, cat's bile. Two methods for preparing hemoglobin crystals given by Mayet (Compt. rend. soc. biol., 1890, cix, 156), and stated by him to be improve- ments on the method of Hoppe-Seyler, are as follows: First method: The corpuscles are washed with sodium sulphate solu- tion (1.5 per cent solution of the anhydrous salt) instead of sodium chloride solution. To wash the corpuscles, a glass vessel having the capacity of 5 liters is used, the upper part of the vessel being of cyUndrical form and tapering conically, the lower part being in the form of a narrow cylinder which holds about 80 c.c. The latter part has at the bottom an opening which can be closed by a glass stopcock; a second opening, capable of being closed, is located where the upper conical and the lower cylindrical parts join. The treatment of the corpuscles with ether (one-fifth volume) is also performed in a special vessel consisting of a cylinder 3.5 mm. in diameter and 35 cm. long and extended by a conical part in the form of a narrow tube provided with a glass stopcock. To the blood solution is added one-fifth volume of absolute alcohol. This mixture is cooled at least 3 times for 12 hours at —14°. The crystals are separated and dissolved in water at 35°, the solution mixed with alcohol as before, and the hemoglobin at least 3 times crystallized by cooMng to — 14°. In this way crystals 1.5 mm. long were obtained from the bloods of the dog, horse, and ass. Second method: The corpuscles are washed as above, the corpuscle pulp is shaken with water (1 volume) and pure benzine (one^fifth volume), and kept 24 hours at 5° to 8°. Then the solution is gradually mixed with one-fifth volume of absolute alcohol and treated in the usual way. The yield by the second process is the greater. A study of the influences of various reagents upon the crystallization of oxyhemoglobin and reduced hemoglobin was made by Donog^y (Math- ematikai 6s termeszettudomdnyi ertesito, 1893, 11, 262; Maly's Jahr. u. d. Fort. d. Thierchemie, 1893, xxiii, 126), who prepared crystals from the bloods of the dog, cat, pig, mouse, ox, rabbit, duck, guinea-pig, horse, and man. 'Donog&D.j tested the usefulness of a number of the methods used for preparing oxyhemoglobin and reduced hemoglobin crystals, and he also made some examinations of the crystaUine forms. Several of the methods were modified. To obtain oxyhemoglobin crystals from dog's blood, the " Canada balsam method " (loc. cit.) is recommended. A method of his own, which he believes equally as good, is as follows : A drop of blood is treated with aUttle ethyl bromide, methylene chloride, or ethylidene chloride. SINCE preyer's investigations. 121 From cat's blood, Donegdny states, oxyhemoglobin crystals can be obtained by any of the usual methods except the methods of Gscheidlen, RoUett, and Wedl, by which only reduced hemoglobin can be produced. From horse's blood good results were recorded with Canada balsam, damar varnish, chloroform, amyl alcohol, pental, xylol, eolophonium dissolved in amyl alcohol, p3a'ogaUic acid, or by freezing. The crystals are doubly refracting, and they consist for the most part of oxyhemoglobin. With the methods of Gscheidlen and Wedl, crystals of reduced hemoglobin were obtained. If the RoUett method is used, combined with distilled water, a mass of reduced-hemoglobin and oxyhemoglobin crystals is formed. The blood of pigs, which is looked upon as crystaUizing with difficulty, he found crystallized readily by the use of ethereal oils. The formation of crystals went on slowly, and the crystals were large and well developed. The crystals were doubly refracting and consisted of oxyhemoglobin. From the blood of white mice crystals could not be produced by the aid of Canada balsam, distilled water, chloroform, ether, alcohol, or xylol. Ox blood did not crystalUze by treatment with Canada balsam, damar varnish, ether, amyl alcohol, xylol, chloroform, pental, ethereal oils, or pyrogaUic acid. By freezing and by Gscheidlen's method, combined with Canada balsam or damar varnish, only small needles could be obtained. From their fight color, Donogdny befieves that they were probably oxyhemo- globin. They were doubly refracting. The coloring matter of the blood of rabbits also crystalfized with difl&culty. The addition of ether, Canada balsam, chloroform, pental, ethereal oils, and acetone gave negative results. With the method of Gscheidlen and with damar varnish, only smaU needles could be obtained. With RoUett's method rather large needles were formed. The best result was obtained with pyrogalfic acid. The crystals formed, he states, consisted of reduced hemoglobin. The blood of the duck treated with damar varnish, xylol, ether, amyl alcohol, Canada balsam, chloroform, eolophonium solution, distiUed water, and by quick coofing, scarcely yielded crystals, and even in the most favorable instance only stunted ones. Gscheidlen's method, he writes, can be used with much better results, although here, too, crystalfization goes on slowly. The crystals were purple- red, almost blue, needles or prisms, and consisted of reduced hemoglobin. Later these crystals, under the influence of atmospheric air, changed to flesh-colored rhombic, even 6-comered tablets, which were doubly refract- ing, and consisted, perhaps, of oxyhemoglobin. From guinea-pig blood Donogdny produced crystals by means of Canada balsam. They formed quickly, and also became quite large if the Canada balsam used was not very thin and the preparation stood in a cool place. If form and size are not important good results can be obtained, he states, by means of ethyfidene chloride. With damar varnish crystalfization goes on somewhat slowly and at the sacrifice of sharp edges. Pjrrogalfic acid and valerian oil did not cause crystalfization. Ether, chloroform, xylol, amyl alcohol, acetone, Canada balsam dissolved in xylol, freezing, a mixture of water and alcohol, and repeated treatment with Canada bal- sam gave only poor results. With ethyl bromide, after the course of an 122 PREPARATION AND CRYSTALLOGRAPHY OF HEMOGLOBINS hour, the whole mass became crystalUne, yet the crystals, because of their smallness, were unsuited to investigation. The most beautiful crystals could be obtained with ethylidene chloride according to the following method: A drop of blood is thoroughly mixed with an equal amount of ethyUdene chloride, the cover-glass is placed on it, and the preparation set aside in a cool place. After the course of 10 to 12 hours it is entirely filled with crystals. With amyl nitrite or pyridine forms similar to these could not be obtained. All the crystals were oxyhemoglobin. Regarding the form of the oxyhemoglobin crystals of the guinea-pig, Donogdny adheres, on the basis of the geometric and optical characteristics, to the view of von Lang that they are sphenoids belonging to the rhombic system. In a later article (Zeit. f. Krystallographie, 1894, xxiii, 499) Donog^ny pubUshes the results of his crystallographic studies of hemoglobin crystals, which will be referred to in later chapters of this memoir. Crystals were easily obtained from human blood by pyrogaUic acid and by the aid of putrefaction. Donogdny first reduced the hemoglobin with a 10 per cent solution of sulphide of ammonium, which, however, is not necessary when using old decaying blood. After an interval of 5 to 6 hours crystals separate in the form of rather thick, flesh-colored or purple- red needles. After 12 to 24 hours the individual crystals are pretty well formed. Contrary to Wedl's assertion, Donogdny observed that the crystals can not be kept, since they burst in the course of 2 to 3 months in spite of being properly sealed. He states that the crystals produced in decajdng blood are of reduced hemoglobin and that they may be changed into oxy- hemoglobin without change of form. He succeeded in producing only reduced hemoglobin directly from the human blood, and he beUeves it doubtful whether by the influence of atmospheric air these crystals can be changed to oxyhemoglobin. _ Oxyhemoglobin was prepared by means of Canada balsam, xylol, damar varnish, chloroform, alcohol, amyl and methyl alcohols, acetone, valerian oil, methylene chloride, and ethylene chloride. Pyrogallic acid and freezing gave only reduced hemoglobin. The crystals, he states, belong to the rhombic system. Wedl had produced reduced hemoglobin crystals by means of pyrogallic acid from dried blood 3 days old, and Donogdny modified this method for the production of hemoglobin crystals from dry blood powder (1 year old). The powder was dissolved in a 5 to 10 per cent solution of sulphide of ammonium, pyrogallic acid was added, and crystals appeared after 10 to 12 hours. After the course of 24 to 48 hours crystallization had ceased. The crystals obtained from horse, cat, and rabbit blood in this way were very beautiful, and large crystals (1 cm. long) were not rare. The crystals were chiefly thin needles, broad prisms, and rhombic plates. In human blood, besides these forms, there appeared right-angled truncated prisms and forms similar to hexahedrons. Experiments with bloods of other animals gave less favorable results. The crystals were doubly refracting and consisted of reduced hemoglobin. The sulphate of ammonium process devised by Hofmeister (Zeit. f. physiol. Chemie, 1890, xxiv, 165) for preparing crystals of egg albumin, and subsequently used by Giirber (Wiirzburger physiol. medizin. Ges., SINCE preyer's investigations. 123 1894, 113) and others for crystallizing serum albumin, has been used by Dittrich (Archiv f. exper. Path. u. Pharm., 1892, xxix, 250) and others for preparing crystals of hemoglobin. Owing to the rapid conversion of hemo- globin into methemoglobin by this process, Dittrich used it also to prepare the latter. The blood of the horse was subjected to the Hoppe-Seyler process for preparing the blood-corpuscle pulp. The corpuscles were then dissolved in ether, the solution filtered, and then mixed with two volumes of a cold saturated solution of ammonium sulphate, filtered again, and then placed in flat vessels in the cold. Generally within 24 hours crystalUzation begins, but occasionally only after 2 to 3 days. The crystals could be recog- nized microscopically in transmitted light as glittering elongated prisms or broad plates. The crystals of the first crystallization were not pure; moreover, the mother-liquor contained, besides crystals, an amorphous precipitate which often could be separated only by repeated recrystalliza- tion. Generally a separation of "globulites" and spherocrystals preceded the formation of crystals. The most of the crystal mass of oxyhemoglobin changed on standing in the air, and through the processes of recrystalliza- tion, gradually and completely into methemoglobin. No further change, for example the formation of hematin, took place. The crystal pulp, recrystaUized several times from ammonium sulphate solution, was finally pressed between absorbent paper, and when dry was saved in this condi- tion. This method of production of methemoglobin renders superfluous the use of ferricyanide of potassium or any other agent, the ammonium sulphate in large quantities being sufficient to change the hemoglobin to methemoglobin. Finally, the crystal pulp with the contained ammonium sulphate is permanent, and its solubihty is not lost. If, however, the preparation is completely dried over sulphuric acid in vacuo, the largest part of the methemoglobin is changed to an insoluble modification. Schulz (Zeit. f. physiol. Chemie, 1899, xxiv, 454) used essentially the same process for preparing oxyhemoglobin for his studies of globin. Horse's blood was rendered incoagulable by ammonium oxalate, the corpuscles were collected by decantation and then diluted with 2 volumes of water. If the solution obtained in this way is mixed with a like volume of cold saturated ammonium sulphate solution, there is formed an ^abundant precipitate which consists essentially of fibrinogen and serum" globuUn. The precipitate after a time becomes so compact that it can be separated by filtration, but, since the hemoglobin begins to crystallize immediately, filtration is rendered difiicult because the pores of the filter become quickly clogged. In the completely clear filtrate crystaUization soon begins, but the quantity thus obtained is small because of the separation on the filter. If the hemoglobin solution and the ammonium sulphate solution are warmed to 40° before mixing, the separation of crystals takes place less quickly, so that the filtrate obtained is almost completely free from blood-coloring matter. If, on the other hand, both solutions are cooled in an ice-chest before the mixing, and the solution after the mixing is allowed to stand until the albuminous precipitate has completely settled, the crystallization of the hemoglobin is almost completely prevented before filtration. If the 124 PREPARATION AND CRYSTALLOGRAPHY OF HEMOGLOBINS solution is filtered in the ice-chest, a clear, dark-red filtrate > is obtained, which contains most of the coloring matter, and when brought to room temperature soon jdelds a rich crystal formation, which increases if a little concentrated ammonium sulphate solution is added. After several days the separation is complete — so complete that the filtrate appears almost colorless. It is purely crystalline, without amorphous admixtures. The crystals are without exception little rhombic plates, some of very consider- able size, while ordinarily crystals of horse hemoglobin, produced according to Hoppe-Seyler, separate in the form of long 4-sided prisms. The precipi- tate is filtered on a Biichner filter by the aid of a Sprengel pump and finally freed from the mother-hquor by pressing between filter-paper, then dis- solved in water, and again separated by the addition of an equal volume of a saturated solution of ammonium sulphate. In this way the hemoglobin can be recrystalhzed with ease several times. Ammonium sulphate efflo- resces on the surface of the firm cake that had been obtained by pressing, and can easily be removed. The cake when dried in the air can be crushed to a fine powder, which readily dissolves in water. The solution shows a pure oxyhemoglobin spectrum. Schulz states that in this way the hemoglobin may be separated from other proteins. Fibrinogen and serum globulin separate completely in a half- saturated ammonium sulphate solution, while the albumin separates only by a higher concentration of the ammonium sulphate than was used here. While as mentioned the oxyhemoglobin, according to the method used by Dittrich (J^c. cit), changes to methemoglobin, even during the recrystalliza- tion, a pure oxyhemoglobin can also be obtained by this method. The preparation thus obtained is, however, limited in stability; in one case it contained after about one year considerable methemoglobin. The limit of the quantity of ammonium sulphate required for the precipitation of the hemoglobin in the amorphous condition, incidentally noticed, is distinctly higher than that for crystallization. An amorphous precipitate occurred only when in 10 c.c. of the solution there were 6.5 c.c. of concentrated am- monium sulphate solution. In the tests which contained 5, 5.5, and 6 c.c, respectively, of the saturated ammonium sulphate solution in 10 c.c, no amorphous separation occurred, but after longer standing crystallization gradually took place. This method of preparation, according to Schulz, is good because of its convenience for experiments not depending on preparations free from salt. The ammonium sulphate method was also used by Spiro (Zeit. f. physiol. Chemie, 1899, xxviii, 182). The corpuscle pulp was obtained from oxalated horse's blood by decantation, diluted with 2 volumes of water, cooled in an ice-chest, after which the solution was agitated with ether in the proportion of 1,000 c.c. of blood-corpuscle pulp to 50 to 70 c.c. of ether. During continual stirring a saturated solution of ammonium sulphate in the proportion of 700 cc- to 1 liter of blood corpuscles was gradually added, the ammonium sulphate solution having the same temperature as that of the blood corpuscles. After 5 to 10 minutes the voluminous precipitate which has formed begins to rise; but if this does not occur more ether must be added, care being exercised to avoid a great excess, since hemoglobin SINCE preyer's investigations. 125 may be precipitated by it. Within several hours a Ught-red deposit has formed on the surface, while the fluid below appears clear and a dark granite-red. The mixture is filtered, and the filtrate is kept in an ice-chest. After 2 days only an insignificant quantity of crystals has formed. These crystals are suspended on the top of the mixture and are filtered off. The filtrate, which contains almost aU of the hemoglobin, is poured into large porcelain vessels and set aside at room temperature. The hemoglobin sepa^ rates at first as red and later as brownish crystals. After 3 days almost all of the hemoglobin has crystallized, so that the filtrate appears to be colored only shghtly brownish. Microscopically investigated, the crystals contain only sUght impurities which can eventually be eliminated by recrystalliza- tion. The hemoglobin is best drained on Biichner filters until the forma- tion of firm cakes. To recrystalUze, the crystals are dissolved in the least possible amount of water and mixed with ammonium sulphate (100 c.c. of the hemoglobin solution to 80 c.c. of saturated solution of ammonium sul- phate). The yield from 5 Hters of horse's blood was 1,500 grams. Fluoride of sodium was added by Arthus and Huber (Compt. rend. soc. biolog., 1893, XLV, 970) to the list of inorganic salts that favor the crystal- Uzation of hemoglobin. They found that when to normal or defibrinated blood there was added an equal volume of a 2 per cent solution of fluoride of sodium, and the solution allowed to stand at room temperature, crystals of oxyhemoglobin could be obtained within a few days. They also state that crystalUzation is accelerated by the addition of 0.1 to 0.5 per cent of hydrochloric acid and by increasing the temperature to 40". Crystals were prepared from the bloods of the dog, horse, cat, and guinea-pig. Guelfi (Rif . med., 1897, No. 10; Maly's Jahr. u. d. Fort. d. TMerchemie, 1897, xxvii, 149) also reports success with fluoride of sodium. He obtained crystals from the bloods of the dog and guinearpig by the addition of an equal volume of a 2 per cent solution of this salt and maintaining the mixture at a tem- perature of 40°. This method, he states, failed in the case of both arterial and venous human blood. The statement by Bohr {loc. cit.) of his beUef that oxyhemoglobin is not a homogeneous substance, and that it consists of a mixture of oxy- hemoglobins which differ in elementary composition, molecular weight, and combining capacity with O, has been shown by Hufner (Archiv f. Anat. u. Physiol., 1894, 130) to be untenable. Hiifner's researches proved that Bohr's methods for producing the several forms of oxyhemoglobin gave rise to mixtures of oxyhemoglobin with variable amounts of decom- position products. Hufner made new studies of the photometric constants of oxyhemoglobin, reduced hemoglobin, and carbon-monoxide hemoglobin, and determined the absorption coefl&cients for and CO. He concluded, from the constancy of the extinction coeflBcients, the O and CO capacities, and the percentage of iron, that in healthy fresh bullock's blood there is only one hemoglobin present, and that the blood-coloring matters of the higher animals have all, when freed from water, the same molecular weight and with it the same capacity for carbonic oxide and oxygen. Hufner also noted that when horse^s blood is crystalUzed in closed cyhnders there appear in great abundance dark-red 6-sided plates, together with the well-known 126 PREPARATION AND CRYSTALLOGRAPHY OF HEMOGLOBINS prisms. He states that if one observes under the microscope a drop of the fluid in which the crystals are suspended, before it is covered with the cover- glass, it will be seen that these hexagonal plates quickly melt, and that where they dissolve bundles of fine, bright-red, prismatic crystals suddenly shoot out. The dark-red hexagonal plates are crystals of reduced hemo- globin, as Nencki showed several years ago, while the bright-red prisms are of oxyhemoglobin. He found that horse's blood is particularly inclined to give crystals of reduced hemoglobin, and that in preparing crystals of horse's hemoglobin by the regular method, without particular exclusion of air, both forms appear at the same time. Hemoglobin crystals from the bloods of the horse, ox, pig, and dog were prepared by Frey (Inaug. Dissert., Wiirzburg, 1894; Jahr. ii. d. Fort. d. Thierchemie, 1895, xxv, 108) by means of the dialyzing method of Glirber. The corpuscles were separated from the defibrinated blood by centrifu- gaUzation, mixed with 2 volumes of water, and placed in a dialyzer which was suspended in 30 to 70 per cent alcohol. Beautiful crystals were obtained after 3 to 24 hours. If a drop of blood be placed on a slide under a cover- glass, crystals form (primary crop) which dissolve as the blood becomes fully laked, when occasionally a second crop forms. By reduction the blood became yellowish, and after 3 or 4 hours it was violet-red and venous, and at the same time granules appear which finally separate towards the margin as distinct crystals. These crystals, Frey states, are of reduced hemoglobin and in addition to these are clusters of colorless crystals. Horse's blood crystallized most readily, and then that of the dog, and finally that of the pig with difficulty. Kobert (Das Wirbelthierblut in mikrokristallographischer Hinsicht, 1901, 25) used the Glirber method to prepare crystals from the bloods of the dog and cat. Arthus (Compt. rend. soc. biolog., 1895, xlvii, 686) employed a similar method to obtain crystals of the horse and dog. The blood was prevented from coagulating by the addition of oxalate, and after the corpuscles had been separated by decantation they were mixed with 2 volumes of water and placed in a Kiihne membrane dialyzer suspended in 90 per cent alcohol. Large masses of crystals were formed. Arthus in a later research (Zeit. f. Biolog., 1897, xxxiv, 444) modified his previous method: The corpuscles from oxalated blood were dissolved in 2 volumes of water and filtered, and the filtrate was placed in a parchment-paper tube which was suspended in 17 to 33 per cent alcohol. At room temperature oxyhemoglobin crystals 7 to 8 mm. long with sharp edges were formed. When stronger alcohol was used the crystals were impure owing to an amorphous precipitate. Studies of the crystallographic characters of crystals from blood of the silkworm were made by Panebianco (Zeit. f . Krystallographie, 1897, xxviii, 1 98) . These crystals were colorless and it is doubtful if they were hemoglobin. Crystals from horse, dog, and cat were prepared by Abderhalden (Zeit. f. physiol. Chemie, 1898, xxiv, 545). Success in the production of crystals, he states, depends upon using the least possible amount of water necessary to dissolve the blood corpuscles which have been freed as much as possible from serum. To horse's corpuscles he added 3 volumes of water; to dog's SINCE preyer's investigations. 127 corpuscles, 2 volumes; and to cat's corpuscles, 1 volume. On the addition of alcohol in the cold, etc., according to the Hoppe-Seyler method, and twice recrystalUzation, he obtained, he states, absolutely pure crystals. His elementary analysis of cat oxyhemoglobin will be found on page 71. The very simple and satisfactory "Canada balsam" method of von Stein (loc. cit.) for preparing small quantities of crystals from readily crystalUzable bloods was again used by him in a later research (Archiv f. path. Anat. u. Phys., 1900, clxii, 477) in a determination of the effects of certain reagents on crystalUzation of guinea-pig's blood. The addition of sodium chloride up to 2 per cent aided crystallization, while quantities beyond this hindered ; calcium chloride, sulphureted hydrogen, and nitrous oxide hindered crystallization. Von Stein also noted variations in crystaUine form from the typical tetrahedra to forms ranging from truncated tetra- hedra to 6-sided plates. A painstaking study of the crystallography of the crystals of pigeon's blood was made by Schwantke (Zeit. f. physiolog. Chemie, 1900, xxix, 486). His results will be referred to fully in subsequent chapters. A new method of getting rid of the stromata, which whether in sus- pension or in solution hinder the crystalUzation of hemoglobin, was devised by Schuurmanns-Stekhoven (Zeit. f. physiolog. Chemie, 1901, xxxiii, 296). The blood corpuscles are washed with 1 per cent salt solution by centri- fugalization, and then shaken violently for 2 hours with asbestos wool. The blood-coloring matter passes into solution, while the stromata for the most part cUng to the asbestos and are removed by filtration. By this method the hemoglobin is not brought in contact with ether. The hemoglobin solution is placed in a parchment-paper dialyzer, which is suspended in 45 per cent alcohol, and put in an ice-chest. As soon as crystals begin to form on the wall of the dialyzer (after 24 to 48 hours) the contents of the dialyzer are placed in a cylindrical vessel and then set in an ice-chest until crystalUzation has been completed. The hemoglobin is not brought in contact with any more alcohol than is necessary for the crystallization. The crystal pulp is as far as possible freed by pressure from the mother-liquor, after which the crystals are dissolved in the smallest possible amount of water at 37°. This solution is again dis- solved and placed in the dialyzer in 45 per cent alcohol. CrystalUzation begins much more quickly than the first time. After crystalUzation has been completed the crystals are separated from the mother-Uquor and dried, first on porous plates and then in a porcelain bowl over chloride of calcium at room temperature. In the monographs by Schulz (Krystallization von Eiweisstoffen und ihre Bedeutung ftir die Eiweisschemie, Jena, 1901) and Robert {loc. cit.) much of the Uterature on the processes for preparing crystals of hemoglobin is referred to. The latter gives an account of blood crystals which he pre- pared from various species, and he attempts the support of the hypothesis of Hoppe-Seyler regarding the existence of the blood-coloring matter in the form of "arterin" and "phlebin." Stewart (American Journal of Physiology, 1903, viii, 102), in his studies of the actions of laking agents on the blood, found that intraglobular 128 PREPARATION AND CRYSTALLOGRAPHY OF HEMOGLOBINS crystallization of necturus blood is very readily obtained by the action of various hemolytic agents. A quick method for preparing crystals of oxyhemoglobin was reported by one of us (Reichert, American Journal of Physiology, 1903, rs, 97), who also made studies of the effects on crystallization by mixing the bloods of different species, etc. It was found that if to the blood of the dog there be added, before or after laking, from 1 to 5 per cent of ammonium oxalate, crystallization invariably begins immediately, and that any quantity of crystals can be obtained within a few hours at ordinary room temperature. If a drop of this blood be placed under the microscope, crystals will be seen to form at once near the margin of the drop, and to be deposited so rapidly that a sohd mass is formed in a few minutes. The blood of the horse does not yield quite so readily to this treatment. If a drop of blood so prepared be examined under the microscope, it will be found that crystallization wiU not begin usually at room temperature until after from 15 to 20 minutes or more, and that it will proceed slowly. Better results can be obtained if the blood be oxalated and centrifugalized, or set aside for the corpuscles to subside. The supernatant Uquid is then poured off, and the remaining corpuscles are laked with ether. Defibrinated blood of the rat, laked with water on a slide, and covered with a cover-glass after the margin of the drop has become dried, usually crystallizes very readily, as is well known. Quicker results can be obtained if the blood be oxalated before or after laking, and even more rapid crystal- Uzation occurs if the blood be laked with ether instead of water. Crystals form so rapidly in the oxalate-ether blood that a magma is formed in the test-tube within a few minutes. The oxalate-ether process applied to the blood of the guinear-pig gives most satisfactory results. Crystallization does not proceed quite so rapidly as in rat's blood, yet within a minute or two innumerable tetrahedra appear, and practically complete crystallization can be obtained within a couple of hours. The blood of the necturus crystallizes readily when so treated. The crystals resemble in form those of the triple phosphates. The rapidity with which crystallization begins and proceeds was found to be influenced decidedly both by the method of laking and the percentage of oxalate. Ethyl ether is a much better laking agent than water, and acetic ether is stronger than ethyl ether. The presence of any quantity of oxa- late up to saturation increases crystallizabiUty, but he found from 1 to 5 per cent to be the best; the larger the quantity the more is crystallization hastened. When more than 5 per cent is used, the oxalate also tends to crystallize upon the slide. If the blood be prevented from drying, as in the test-tube, the oxalate remains in solution. Asphyxial blood yields crystals more readily than normal blood. If to the blood of one species, the blood, plasma, or serum of another species be added, the laking of the blood may be retarded, accelerated, or unaffected, according to the character of the mixture. The period required for laking may be prolonged 6 minutes or more. The crystallization of the oxyhemoglobin may be hindered or prevented in such mixtures. Thus, by varying the proportions of a mixture of the bloods of the dog and guinea- SINCE peeyeb's investigations. 129 pig crystals from one or both may appear, but the process is invariably retarded, sometimes to a marked degree. If crystals of both kinds of oxy- hemoglolain are deposited, those of one usually begin forming some time before those of the other, and the crystaUization of one may be seemingly complete before crystals of the other are seen. The interesting observation was also made that the typical forms of the crystals of certain kinds of oxyhemoglobin may be modified or com- pletely changed when the bloods of two species are mixed. Thus, if to the blood of the rat there be added a definite percentage of the blood of the guinea-pig, crystals of the rat's oxyhemoglobin may appear in unaltered form, but most, if not all, of those from the guinea-pig's blood are changed; in fact, if any perfect tetrahedra are found, they wiU have been formed at the very end of the crystallization. If the proportions of the mixture be properly modified, not a single crystal of what can be identified as rat's oxyhemoglobin will appear, and all the crystals will be modified tetrahedra, spindles, and transitional forms between these. The spindles resemble Char- cot's crystals in form, but not in color; they vary in size, some being very large, and some may have small spindles attached to them; they can be obtained having sharp edges if crystalUzation has not been too rapid. This complete change in the form of the crystals of oxyhemoglobin when the bloods of two species are mixed, and the spindle-shaped form of the crystals, are, he believes, unique facts in the crystallography of this most important substance. (See Halliburton, page 115.) Moser (Vierteljahresschr. f. gerichtl. Medizin, 1901, xxii, 44) asserts that differences in crystalline form afford a positive means of recognition of the origin of the blood, and that positive distinction can be made between human and animal blood. He examined blood stains of the fresh blood of man and 10 species of mammals and fish, and gives drawings of their appearance under the microscope. From the differences in shapes he infers differences in crystaUization, which he reasons indicate differences in chem- ical constitution. No descriptions of the crystallographic or optical char- acters are given, and, as differences in the shapes of the crystals do not necessarily imply differences in crystal system, his conclusions are based upon insufficient data. Bonnel (These de Paris, 1903; Jahr. ii. d. Fort. d. Thierchemie, 1903, xxiii, 182) showed, however, that the method of Moser is not worthy of recommendation because by this method he obtained from human blood crystals of different shapes. Friboes (Archiv f. ges. Physiol- ogic, 1903, xcviii, 434) also found that human blood treated in the manner described by Moser crystalUzes in various forms. He notes that the bloods of certain animals show crystalline shapes which, with the exception of the bat and the goat, are distinguishable from human blood. He describes the different forms of the crystals he observed, but gives no definite crystallo- graphic data by which they may be recognized. (See Chapter VII.) In experiments with the blood of the horse, Uhlik (Archiv f . ges. Phys- iologie, 1904, civ, 64) found that as putrefaction progresses the usual rhombic crystals of oxyhemoglobin are replaced by hexagonal, holohedral crystals of reduced hemoglobin. Table 34 indicates the influences of the condition of the blood and temperature upon crystallization. 130 PREPARATION AND CRYSTALLOGRAPHY OP HEMOGLOBINS. Table 34. — Effects of the condition of the blood and temperature upon crystallization according to Uhlik. Condition of the blood. Temperature and crystal syetema. 0° 5° 10° 15° 20° Fresh Rhombic, abundant Rhombic, scarce or none Hexagonal, scarce or none Hexagonal, abundant Do Rhombic. . . Rhombic, scarce Rhombic. Hexagonal and rhom- bic. Partly reduced; not any decomposi- tion Reduced; beginning to decompose . . Decomnosed: insnissatins Hexagonal, abundant Do. Hexagonal and rhom- bic Hexagonal, abundant Hexagonal Uhlik also notes that Pregl found that a thrice-crystallized hemoglobin appeared as hexagonal crystals. Crystals of reduced hemoglobin have been prepared and described by a number of investigators, as stated in previous pages. The last hemoglobin to be obtained in crystalline form, excluding our own preparations, was prepared by Bardachzi (Zeit. f. physiolog. Chemie, 1906, xLix, 465) from the blood of the sea-tortoise {Thalassochelys corti- cata). The blood was centrifugalized, the corpuscles mixed with water, and then set aside for several hours at 50°. The solution was then filtered, one-fifth volume of alcohol added to the filtrate, and the mixture placed in an ice-chest. Crystallization occurred quickly and abundantly in the form of plates. The crystals were soluble with difficulty in cold water. For the purpose of analysis the crystals were dissolved in water at 40°, and after cooling one-seventh volume of alcohol was added, and crystalliza- tion obtained as before. The crystals were then centrifugalized off and dried in vacuum. The mean values of the elementary analyses were C54-77H6-99N17.07S0-38F60.41 The absence of phosphorus is striking, since previous observers failed to obtain hemoglobin free from phosphorus from bloods that contain nucleated erythrocytes. The optical investigation by means of the Hiifner spectro- photometer showed decided agreement with the blood-coloring matter of such other animals as have been closely investigated up to this time. The average quotient was e' : e= 1.561, while Hiifner found the quotient to be 1.578. The calculation of the extinction coefficients and quotients of hemo- globin and methemoglobin agreed, he states, with those of other oxyhemo- globins and methemoglobins, so that the coloring matter of the blood of the tortoise, Bardachzi holds, is identical with that of mammals. Abderhalden and Medigreceanu (Zeit. f. physiol. Chemie, 1909, lxix, 165) report their preparation of crystals of goose hemoglobin free from phosphorus. CHAPTER VII. CRYSTALLOGRAPHY OF HEMOGLOBIN IN RELATION TO SPECIES, ACCORDING TO PREVIOUS INVESTIGATORS, WITH EXPLANATIONS OF VARIOUS CONTRADIC- TORY STATEMENTS, ETC. As early as 1852 Kunde (Zeit. f. rat. Medicin, 1852, N. F., ii, 271) and Funke {ibid., 288) in coincident articles stated that the hemoglobin crystals of different species are different. Kunde prepared crystals from the bloods of a number of species, including the bat, dog, ox, horse, guinea- pig, squirrel, rat, mouse, rabbit, pigeon, and tortoise, and published some figures illustrating the shapes of the crystals. From these differences in the shape and from the differences in solubiUty he concluded that the blood crystals obtained from different species are not identical, but distinct and characteristic of the species. Funke was led to the same conclusion from the examination of the crystals from the blood of the horse, ox, pig, dog, cat, and several species of fish. While making no attempt to give an exact crystallographic description, Funke records a number of angles ob- served in two of the species examined. These contributions were almost immediately followed by an article by Teichmann {ibid., 1853, iii, 375), who states that from the same blood, and even in the same preparation, crystals of various forms may be obtained, from which and for other reasons he concludes that the differences are not in relationship to species, but accidental and due to exterior conditions. Teichmann's statement seems to have arrested further interest in this subject until 10 years later, when it was taken up by RoUett (Sitzungsb. Math.-nat. Klasse d. k. k. Akad., Wien, 1862, xlvi, Abth. ii, 85), and shortly after by Bojanowski (Zeit. f. wiss. Zoologie, 1863, xii, 312). RoUett pre- pared crystals from the bloods of man, the guinea-pig, dog, rabbit, squirrel, and cat, all of which preparations, with the exception of the last, he sub- mitted to von Lang, a crystallographer, for crystallographic investigation. Von Lang's examinations were made with the microscope, and in some cases the optical characters were examined and a few angles recorded. Von Lang determined the crystal system in each case, and from his data RoUett concluded that while the crystals from different species are different they may aU be included in two crystal systems, the orthorhombic and the hexagonal. The descriptions of von Lang are very brief, and no attempt at giving all of the crystallographic constants is made, but these are the first definite determinations on record of the systems of crystallization of hemoglobin. Bojanowski reviewed the Uterature of hemoglobin crystals and pre- pared crystals from the blood of rabbit, mouse, dog, cat, hedgehog, river 131 132 CRYSTALLOGRAPHY OF HEMOGLOBIN IN RELATION bream, pike, horn-fish, herring, lark, raven, and pigeon, and of man. He records that hemoglobin of various animals crystallizes in various forms and systems, and that he always obtained rhombic plates from the blood of man and many species of lower animals, regular 6-sided plates from the blood of the mouse and squirrel, tetrahedra from the blood of the guinea-pig, and prismatic crystals from the blood of the rabbit. Crystals from various kinds of blood which appear to possess a similar form still showed unmistakable differences in the sizes of the angles. From his investigations he reached the conclusion that the bloods of individual species have something specific and characteristic about them, so that it is occasionally even possible to determine the species of animal from whose blood the crystals were derived. Where, as in the case of human blood, as described by Funke, there appear to be two or more kinds of crystals in the same blood, Bojanowski considers that one of them is the characteristic form and the others undeveloped crystals. Thus, in human blood what he describes as the "right-angled plate" is, he believes, the characteristic form, while the "prisms and rhombic plates" are regarded as undeveloped forms of the right-angled plate. The descriptions given are very brief and incomplete: thus, the crystals from the dog are described as "rod-hke crystals forming closely woven nets," and from the cat as "very regular three-sided rods," etc. The description of the crystals of dog's blood would apply equally well to any species whose hemoglobin crystals are rather insoluble, if the hemoglobin crystallized in prisms, for such hemoglobins form felted masses of capillary or long pris- matic crystals. The prisms of reduced hemoglobin of the cat are not 3- sided, but nearly rectangular in section. After a latent period in the study of the crystallography of hemoglobins for the 5 succeeding years the first contribution by Preyer appeared (Archiv f. ges. Physiologic, 1868, i, 395), which was shortly followed by his now classic and authoritative memoir (Die Blutkrystalle, Jena, 1871). When the former contribution was pubUshed blood crystals from 47 species of vertebrates had been recorded, and of these in only 10 cases had the crys- tal system been recorded. In his memoir these 47 species are enumerated and the data concerning them are given. Preyer evidently regarded the crystals obtained from different species as differing from one another, but he concluded with RoUett that they may all be included in the two crystal systems, the orthorhombic and the hexagonal. He states that "besides the crystal system there are other distinctions, as, for instance, the sphe- noidal crystal of the guinea-pig, the 4-sided prisms of the dog, the 4-sided prisms and rhombic plates of man. These peculiar morphological shapes are obtained only from each animal, even after repeated recrystallizations ; a definite form is peculiar to each animal and can not be changed to another form. The same holds good with solutions of hemoglobin. Yet little im- portance is to be attached to statements on the crystallographic differences of the hemoglobin of different animals, because neither is the same method of crystallization always used, nor is the blood always capable of being compared, nor has the measure of the crystallizability of any optional sub- stance been found. It is the same with decomposabiUty as with crystaUiz- TO SPECIES, ACCOBDING TO PBEVIOUS INVESTIGATORS. 133 ability — ^both vary according to the species of animal; but the investigations undertaken in this direction suffer from so many and such large errors that they prove nothing beyond what has long been known, that is, the different species and individuals. " Preyer's statement that the form of the crystals can not be altered by repeated recrystallization, and that there is a constant and peculiar form in relation to each kind of animal, has been shown to be wrong by the records of Halliburton (page 115), Copemann (page 119), von Stein (page 127), Bonnel (page 129), Friboes (page 129), Moser (page 129), and Pregl (page 130). The work of Preyer was so painstaking and exhaustive that his con- clusions seem to have been accepted without question, and his dictum that all hemoglobins crystallize in the orthorhombic system with an exception which crystallizes in the hexagonal system seems to have absolutely dis- couraged investigation in the crystallography of hemoglobin, and such studies as have since been made have been chiefly with the view of dis- tinguishing human blood from that of domesticated animals, for medico-legal purposes. Of the papers treating of the crystallography of hemoglobin in rela- tion to species from this standpoint, those of Guelfi (Giomal di Med. Legale, 1898; Maly's Jahr. ii. d. Fort. d. Thierchemie, 1898, 145) and Moser (Viertel- jahr. ger. Med., 1901, xxii, 44) may here be noticed. Guelfi obtained " tetra- hedral crystals" from guinea-pig's blood and "prismatic crystals" from dog's blood, using both fresh and dried blood in each case. Comparing these with crystals obtained from partly dried human blood, which crystals he describes as "needle-shaped," he states that they can be distinguished from each other so that "it can be definitely stated that neither the tetra- hedra from the guinea-pig blood nor the prisms from the dog blood were from human blood. " Moser describes crystals obtained from the blood of about a dozen species of vertebrates including mammals and fish. His article is illustrated with drawings made from the appearances of the crystals under the micro- scope, but these are not accompanied by any exact crystallographic descrip- tions. The differences in the shapes of the crystals led him to the conclusion that differences in the forms of the crystals afford a positive means of recog- nition of the origin of the blood, and that in this way positive distinction can be made between human blood and the blood of other animals. The descriptions of the crystals are very brief and relate to their general mor- phology; this is true also of the drawings. No correlation of the different shapes of crystals found in the same species is attempted, and what are evidently different views of the same crystal are shown as different forms. It is obvious that he is distinguishing the different crystals merely and hazardously by their morphology. Moser's article has been the subject of adverse criticism, as will be pointed out. Various observers have studied the shapes of the crystals obtained from the bloods of different species, and in a few instances the crystal system has been determined by crystallographic study, and from these data they have arrived at the conclusion that the bloods of different species 134 CRYSTALLOGBAPHY OF HEMOGLOBIN IN RELATION may be distinguished by an examination of the hemoglobin crystals. On the other hand, this conclusion has been contradicted by many observers. Teichmann, for instance, as already stated, asserts that from the same blood, and even from the same preparation, he has obtained various crystal forms, and that still other forms may be produced by varying the method of prepa- ration, from which he naturally concludes that the form of the crystal is something entirely accidental and dependent upon exterior conditions and not an essential character of hemoglobin. Others have made the observa- tion that in the same blood several forms of crystals may be found. It has also been pointed out that crystals from the blood of a given species, as recorded by different investigators, are of different forms. Thus, Leh- mann described the crystals from the guinea-pig as isometric tetrahedra, he also describes them as isometric octahedra; Moleschott states that they are 6-sided plates. Von Lang writes that they are only seemingly isometric, and that, while the angle of the triangular face is so near 60° that they can not be distinguished from isometric tetrahedra, the optical characters make them orthorhombic. Donogdny measured the three angles of the triangular face of these crystals and records them as 64°11', 60°50', and 55°45', which three angles it will be noted do not add up to 180°. Of course the explana- tion of the record of tetrahedra in the one instance, of octahedra in another, and of 6-sided plates in a third is very simple. All of these observers were examining crystals of the same substance, and all were, as von Lang and Donogany state, orthorhombic sphenoidal: in the case of the simple "tetra- hedra" the right or left sphenoid only was observed; in the case of the "octahedra" the crystal was the combination of the right and left-handed sphenoids in approximate equihbrium; and in the last instance, of the 6- sided plates, the form seen was this combination observed normal to a sphenoid face upon which the crystal is flattened, causing the outUne to be hexagonal. The outline of an octahedron looked at as it Ues on one of its faces is hexagonal, but if it become flattened parallel to the face upon which it Ues it appears at a casual glance to be a hexagonal plate. Many such cases as that of the guinearpig crystals have been noted, where the blood of the same species by varying the treatment, or even according to different observers, furnished crystals of diverse form; and many observers have been led to the conclusion that was reached by Teich- mann, that the forms of hemoglobin crystals are variable in the same species, are perhaps even identical in different species, and that the differ- ences are not to be rehed upon for distinguishing the source of the blood in any given case. When the article by Moser appeared it apparently revived interest in the subject of the differentiation of the crystals of different species, but his results were soon attacked by Bonnel (Th^e de Paris, 1903; Maly's Jahr. ii. d. Fort. d. Thierchemie, 1903, xxiii, 182) and by Friboes (ArcWv f . ges. Physiologie, 1903, xcviii, 434). Bonnel argued that because human blood treated in the way described by Moser crystallizes in different shapes the method is of no value. He points out that the method is not to be recom- mended for the purpose of distinguishing human and animal blood (although TO SPECIES, ACCORDING TO PREVIOUS INVESTIGATORS. 135 differences between these are to be detected), because it is only applicable to fresh blood and can not be applied to blood stains, at least if they are more than two weeks old. Friboes also attacks Moser's conclusion that the crystals serve to distinguish between human and animal blood. He states that normal human blood treated in the way described by Moser crystallizes in various forms, and that the crystals from dried human blood are different from any of these. Human blood obtained from the splenic vein and the umbiUcal vessels is again different from these, so that a uniform crystal shape for human blood does not exist. Thus, from fresh human blood are obtained 4-sided doubly refracting prisms, also sharp-angled rods spUt into brush-Uke forms at the end, and very characteristic rectangular plates arranged in step-like aggregates. From the blood of a young child he obtained long rectangular plates which he regards as stiU different. From the splenic vein he found crystals showing composite aggregates of the step- Uke arrangement of the rectangular plates. From the blood of the umbilical vessels he prepared rosette aggregates of ray-hke crystals, and in this same blood he also noticed sheaf-Uke bundles of crystals and also isolated irregular crystals. The blood of other animals showed still other forms, which, however, are usually distinguishable from the crystals obtained from normal human blood, with the exception of those from the blood of the bat and goat. The distinction from human blood depends, he states, upon having a sufficient supply of blood and in obtaining it before it becomes dry. The article by Friboes is illustrated by excellent photomicrographic reproductions of some of the blood crystals examined, but his descriptions of the crystals are very brief and in many cases incorrect. Thus, in the description of the crystals from the cat he enumerates three kinds of crys- tals and illustrates them by two photomicrographs. These three types are (1) long, 3-sided prismatic rods, single or in bundles; (2) 4-sided prisms, rhombic; (3) fine needles. He points out in the photomicrographs what he designates the "3-sided rods," which are evidently only an edge view of what he reports as "4-sided prisms. " The fine needles are simply the same crystals in capillary form. All of these belong to the long prismatic type of crystal of cat reduced hemoglobin, and he appears not to have observed the short prismatic type nor the parallel growth aggregates that are usual in the preparation from cat's blood. His "fine needles" are generally the first crystals to appear, and his other two types, which he regards as distinct (one trigonal, the other rhombic), are but two views of the same crystal. The foregoing is simply an example of how an expert microscopist who is not a crystaUographer may be misled by different appearances that he is unable to reconcile. The objections recorded in opposition to the conclusion of Kunde and others that the blood crystals from different species are not identical and that they are characteristic of the species may be summarized briefly as fol- lows : The form of the crystal of any species may be entirely accidental and dependent upon exterior conditions, and hence can not be characteristic of the species. In the same species different forms of crystals may be seen 136 CRYSTALLOGRAPHY OF HEMOGLOBIN IN RELATION even in the same preparation, and by varying the method of preparation many forms of crystals may he obtained from a given species. Different forms of crystals have been obtained from the blood of different vessels of the same species or the same individual. Different observers have pro- duced quite different crystals from the blood of a given species, some of these closely resembhng or seemingly identical with those obtained from the blood of other species. There can not, therefore, be any one form of crystal that is characteristic of a given species. Preyer himself, while recognizing that crystallographic differences exist between the hemoglobins of different species, states that httle importance is to be attached to statements on the crystallographic dissimilarities of the hemoglobin of different species, because neither is the same method of crystalhzation used nor is the blood always capable of being compared. He might have added, that in very few cases have the crystallographic descriptions been at all adequate or even accurate, but this he probably failed to recognize. His statement that all hemoglobins crystallized in the orthorhombic system excepting that of the squirrel was doubtless taken by many as an argument in favor of the assump- tion of the identity of the blood crystals obtained from different species. We thus see that equally expert observers, working with the same data, have arrived at very diverse conclusions. Before attempting to reconcile these conflicting conclusions it will be of advantage to examine certain other observations that have been made on hemoglobin crystals. A number of the earUer investigators, including Lehmann, Teichmann, Weir Mitchell, and Bojanowski, and several of the later ones, such as Struve, and Stirling and Brito, have noted that the crystals obtained from the blood may be nearly or quite colorless, or may become so on standing; or, according to several of them, the deep-red crystals may be decolorized by washing them with alcohol, or with alcohol and water, or with other reagents. Thus, Bojanowski states that the blood crystals exposed to the air retain their form, but become paler and paler and finally completely colorless. The addition of sugar or gum produces the same result. Teich- mann had made similar observations on the loss of color of the deep-red crystals. Bojanowski's statement is a fairly accurate description of the paramorphous change of crystals of oxyhemoglobin to metoxyhemoglobin, many examples of which wUl be found in the records of this research. The color of the crystals of metoxyhemoglobin is very pale as compared with that of oxyhemoglobin, and when the crystals are thin they appear almost colorless. The very strong pleochroism of metoxyhemoglobin makes the crystals appear quite colorless in some positions. Weir Mitchell made similar observations on oxyhemoglobin crystals exposed to the air. The "colorless" crystals retain the form of the original oxyhemoglobin crystals, but after the change they are a different substance, and are in fact pseudo- morphs of the original oxyhemoglobin crystals, and if dissolved and recrys- tallized the form would probably be altered only sUghtly, not suflSciently to be noticed by casual observation. From the blood of the raven that had stood exposed to the air for 8 days, Bojanowski obtained " crystals which were partly bright yellow and TO SPECIES, ACCORDING TO PREVIOUS INVESTIGATORS. 137 partly colorless. " This is a description of the method of producing crystals of metoxyhemoglobin, and the colors described are such as would be found in metoxyhemoglobin crystals that were rather insoluble, as these are described as being. He made a similar observation upon the crystals from the cat. Weir Mitchell describes the production of crystals of oxyhemoglobin from the blood of the sturgeon, and states that their color may be com- pletely removed by alcohol and water without injury to the form, and that these decolorized crystals may be dissolved in water and recrystalUzed in the original form. Struve (Ber. d. d. chem. Ges., 1881, xiv, 930) decolorized blood crystals by treating them with dilute alcohol, but without causing any change of form. In a later communication (Jour. f. prakt. Chem., N. F., 1884, xxix, 304) he gives a more detailed description of his observations : Fresh blood crystals placed in an excess of alcohol change their color to a darker tint, without change of form, and become insoluble in water and alcohol. This, he states, is due to a loss of water of crystallization and going over into an amorphous condition. These altered crystals by treatment with ammoniacal alcohol, by glacial acetic acid, or by concentrated sulphuric acid are decol- orized without change of form. Struve did not dissolve and recrystallize them. The color extracted he regards as a hematin derivative, which he names hematin acid. His conclusion is that hemoglobin crystals are a col- orless albuminous substance, mechanically mixed with a coloring matter. On reading the descriptions of Struve it seems evident that the treat- ment with alcohol changes the crystals of hemoglobin by hardening them, an effect of alcohol upon albuminous substances generally; and if he started with oxyhemoglobin the darkened crystal treated with alcohol was already a different substance, a pseudomorph in fact. Such a pseudomorph nught retain its form even though the substance of which it was composed should be the original material decomposed. In inorganic substances we find for instance crystals of pyrite, FeS2, changed by pseudomorphism into limonite, Fe403(OH)6 without the shghtest change in outward form; fluorite, CaFg, in this way is changed to quartz, Si02. The colorless crystals obtained by treatment of the alcoholized crystals with the agents mentioned above are but skeletons of the original oxyhemoglobin crystals, and may have quite a different composition. As Struve states, they are amorphous and not really crystals at all. But Weir Mitchell's recrystalUzed colorless crystals are not of this kind, and are not to be explained in the Ught of our present knowledge. Colorless blood crystals are (with the exception of the recrystaUized colorless forms described by Weir Mitchell) to be accounted for by a change of oxyhemoglobin to metoxyhemoglobin, by pleochroism, or by pseudo- morphism in case of chemically treated crystals. Besides colorless and slightly colored crystals, other variations from the typically colored oxyhemoglobin crystals have been observed. Thus, we find records of "bluish," "purple," and "pink" crystals that are evidently reduced hemoglobin; and "yellowish" and "brownish" crystals that may be methemoglobin. The failure to distinguish between methemoglobin and 138 CRYSTALLOGRAPHY OP HEMOGLOBIN IN RELATION metoxyhemoglobin has given rise to much confusion. It is clear that dif- ferent observers of blood crystals have examined crystals of oxyhemoglo- bin, reduced hemoglobin, metoxyhemoglobin, and methemoglobin in many instances without making any distinction between them. Since these sub- stances in a given blood may form quite different crystals, a source of the variations in the recorded crystals of a given species is obvious. As has been shown, equally expert observers working with bloods of the same species have arrived at very different conclusions as to the specifi- city or non-specificity of hemoglobin crystals in relation to species, some claiming that the crystals are occasionally specific, others that they are always specific, and others that they are not specific because the same blood may yield crystals of very different forms and that the differences are probably accidental. Crystals of various colors and varying forms have been obtained from the same blood. It has been held in favor of specificity that recrystallization, even when frequently repeated, does not effect any change in form; but this has been contradicted by observers who point to final evidence to the contrary. How are these diverse conclusions to be reconciled? In the first place, it is evident that the substance under investigation was not always the same: sometimes it was oxyhemoglobin, or reduced hemoglobin, or metoxyhemoglobin, or methemoglobin, etc. Any one of these substances may appear in several forms of crystalUzation in the same blood, often as many as three of them in the blood of a given species ; and it is even probable that there are other forms of hemoglobins present which have not yet been isolated. But much more important even than these sources of variation in the crystals was the failure of the observer to in- terpret correctly his observations. The same crystal viewed in different aspects presents different appearances, and the same crystal combination may exist in different shapes due to the variation in crystal habit. The expert microscopist might learn to interpret the different aspects presented by a single crystal, but no one who is not a crystallographer would be likely to suspect that a long rod-like crystal and a thin tabular crystal might be the same combination of crystal forms. It was such failure to interpret the forms observed that has caused the confusion between the apparent octahedrons and the apparent 6-sided plates of the guinea-pig oxyhemo- globin crystals that have been mentioned. An octahedron lying on one of its faces and observed normal to this face has a hexagonal outline, and if it grows lying on this face it will develop into a 6-sided (or a 3-sided) plate, because it grows twice as fast parallel to the plane on which it Ues as it does normal to that plane, since it can not grow at all on the bottom plane. A tabular crystal seen on edge looks like a rod or prism, and has been so described by many observers. Actual errors in observation are very common. For instance, Bojan- owski, owing to the nearly square prisms of the cat hemoglobin when seen on edge, looks upon them as being 3-sided prisms; and Friboes falls into the same error, and even shows photomicrographs of the nearly square orthorhombic prisms of the same substance, and refers to them as "3-sided." TO SPECIES, ACCORDING TO PREVIOUS INVESTIGATORS. 139 Kunde and Lehmann observed "tetrahedra" in the "hemoglobin" of the black rat. These were doubtless the /^-oxyhemoglobin crystals, which are iso- metric, and appear as the three-sided plates that develop from the flattening of the octahedron. Such a crystal seen on edge would be described as a prism. When the different habits that the same crystal combination may assume are considered, the difficulty of interpreting the observations increases enormously. Thus, crystallization may begin with the formation of needle-hke or capillary crystals, and these may later become short prisms. Friboes describes these two forms of the same crystal as two kinds of crys- tals in the case of cat hemoglobin; and, as has been stated, by looking at the same crystal in two aspects at 45° to each other, he sees two kinds of prisms, thus making three kinds of crystals of the same identical crystal combination. In certain species of the cats the hemoglobin occurs in all of these variations of the prismatic type of crystal and also in the tabular form, yet the crystal forms shown may be the same in prism and plate. Under less pressure the crystals form as prisms; under greater pressure they form as plates. Crystals from the blood of the black rat have been described as tetra- hedra, prisms, elongated plates, and hexagonal plates. The tetrahedra have already been referred to, and they are evidently, as stated, /3-oxyhemoglobin. The prisms, elongated plates, and hexagonal plates are all the same combi- nation of crystal forms, the prism and macrodome, and are our a-oxyhemo- globin. When symmetrically developed the crystal is the squarish prism terminated by the dome. Flattening of the crystal on two opposite prism faces produces the "elongated plates" of Hoppe-Seyler, and shortening of this flattened prism produces the apparently hexagonal plate. Careful focusing would show at once that this plate is not bounded by vertical sides and that the angles are not hexagonal angles. All of these forms we have observed in the crystals from the blood of the common rat. The crystals are frequently interfered with by the slide and cover pro- ducing false planes, so that a tabular crystal on edge, thus confined, becomes a "prism." Many examples of crystals with such false planes have been figured, even as late as the work of Moser (1901). When it comes to the determination of the crystal system, we find that most of the observers make no attempt at it. Preyer states that in his table (page 103) five of the six crystal systems are recorded, the trichnic being the only one not included. The isometric, he writes, may be ruled out because all hemoglobin crystals are doubly refracting and because isometric crystals can not be doubly refracting. Crystallographers now recognize that the tetartohedral class of the isometric system is doubly refracting, and, as will be shown later, we have found singly refracting isometric crys- tals of hemoglobins. The tetragonal system he eliminates because the statements of Hoppe-Seyler in regard to the tetragonal character of the guinea-pig crystals were disproved by von Lang. Similarly he excludes the monoclinic because he states that Funke, who claims to have observed monoclinic crystals in the case of the cat and man, "supports his state- ment by nothing. " This leaves only the orthorhombic and the hexagonal. 140 CRYSTALLOGRAPHY OF HEMOGLOBIN IN RELATION TO SPECIES. Preyer also states that of the 47 species examined and recorded the system of crystallization is known in 10 instances, in only one of which are the crystals accredited to the hexagonal system. In fact, von Lang seems to have been the only professional crystallographer who examined blood crystals up to the time of Preyer, and his descriptions, as has been stated, are very brief. Since von Lang found only two crystal systems, so Preyer concludes there can be but two crystal systems to which the hemoglobin crystals belong. Nevertheless the fiye crystal systems mentioned by Preyer as having been recorded by various observers, of which he rejects three, are all represented by us in the hemoglobins included in this research. When we try to find how these investigators arrived at their conclusions as to the crystal system we are met by short, very incomplete descriptions, and we are led to the conclusion reached by Preyer in the case of Funke's monoclinic crystals, that "they support their statements by nothing." The work of von Lang was evidently accurate, although his crystallographic notes are brief; Donogdny confirmed von Lang's findings in the case of guinea-pig's crystals, but, as we have already pointed out, he records three angles of a triangle which sum up to 180° 46'. The only contribution that has appeared giving the crystallographic constants and an accurate descrip- tion of hemoglobin crystals is that of Schwantka (Zeit. f. physiol. Chemie, 1900, XXIX, 486) on the oxyhemoglobin of the pigeon, which will be found referred to at length under that species in a later chapter. The foregoing is in effect a brief statement of the status of the crystal- lography of hemoglobins at the inception of this research and up to the present time. CHAPTER VIII. METHODS FOR PREPARING, EXAMINING, AND MEASURING CRYST'ALS OF THE HEMOGLOBINS EMPLOYED IN THIS RESEARCH. METHODS FOR PREPARING CRYSTALS OF HEMOGLOBIN. The necessarily limited quantities of blood that have been furnished us led, as a consequence, to the study of only such methods as are especially applicable to very small supplies, such for instance as 1 to 5 c.c. of fluid or clotted blood, although several of our processes may be used to advan- tage in the preparation of very large quantities if a method be selected that is suited to the species and to the condition of the blood. In only a few instances were we unsuccessful in obtaining crystals, and when we failed it was owing to an inadvertent selection of a wrong method or to attendant conditions over which we had no control. Our difficulty was not so much in the way of securing crystals as it was in the preparation of specimens that were adapted to the peculiar requirements of our investiga- tion. We found, as we gained experience with the bloods of different species, that, while the blood of each species must be treated as an individ- ual, we could nevertheless depend with some confidence upon the guidance of certain generalizations in the selection of the best method to be pursued. Thus, we found that usually the hemoglobins of Rodentia and Canidce crystallize with great readiness, those of Marsupialia very readily, those of FelidoB readily, those of Ungulata not readily, those of Aves with difficulty, etc.; but there were so many unexpected exceptions that we were often misled, and, as a consequence, obtained inferior results, as a number of our photomicrographic reproductions show. Even in the case of species closely related, as, for instance, certain of the rats, we found striking exceptions : The blood of the common albino or white rat {Mus norvegicus var. cclbus)* and that of Mus decumanus Pall. {Mus norvegicus Erxleben — ^brown rat) crystallize with such readiness that we found it desirable to use a restrainer to obtain crystals of desirable size for study; on the other hand, the bloods of Mus rattus (black rat) and Mus cdexandrinus (alexandrine rat) crystallize much less readily, and hence should be treated in an entirely different way. We absolutely avoided the use of alcohol, because, notwithstanding the fact that it has proven one of the most widely used and most valuable agents in the preparation of hemoglobin crystals, it so deleteriously affects the hemoglobin molecule that even when present in dilute solution it lessens * Hatai (Biological Bulletin, Wistar Institute of Anatomy and Biology, Philadelphia, 1907, xii, 266) states, upon morphological grounds, that the albino rats of Chicago and Philadelphia are a variety of Mits norvegicus. 141 142 METHODS FOR PREPARING, EXAMINING, AND MEASURING solubility, alters the extinction coeflScient, gradually decolorizes the crystals, and doubtless affects the water of crystallization. Alkalies and mineral acids have Ukewise been avoided, because of their pernicious influences. Hemoglobin, whether in crystalline form or in solution, especially when in concentrated solution, undergoes rapid alteration; we therefore made our studies as soon as possible after we obtained satisfactory crystals, usually within a few hours. In none of our examinations have we used recrystal- hzed hemoglobin. Our specimens have been too small in quantity to permit of satisfactory recrystallization, and, moreover, the disadvantages of recrystalUzation, especially in so far as the methods of our investigation are concerned, quite outweigh the advantages. The injurious effects of recrystallization have been fully referred to in previous pages. At the inception of our research it seemed to us that the best results, on the whole, were to be obtained by the use of fluid blood, either defibri- nated or rendered incoagulable by oxalate, fluoride, or other anticoagulant, so that in the case of bloods which do not crystalUze readily the corpuscles could be collected from the serum or plasma by centrifugahzation, and thus eUminate certain substances in these fluids which retard crystallization and at the same time obtain a concentrated solution of hemoglobin. Since it seemed impracticable to obtain defibrinated blood, owing to the circum- stances under which our specimens were to be collected, and since one of us (Reichert, page 128) had already found that the presence of an anti- coagulant, such as neutral oxalate, was not only not injurious but actually beneficial, we made use of oxalate of ammonium in all of our preparations except in a very few instances, when for some special reason its absence was desirable or necessary. The addition of oxalate, in the proportion of 1 to 5 per cent of the dried powder, it was found, very much favors crystal- hzation; the larger the quantity up to the point of saturation the better the effect, saturation not being a disadvantage beyond the appearance of crystals of oxalate, which, however, are readily distinguishable from those of hemoglobin. In fact, in several instances these crystals appeared to be of advantage, because hemoglobin crystals formed on them, but not in other parts of the preparations. When we had defibrinated blood or clots to work with, oxalate was added at the proper time during our procedures of preparation. Since the presence of foreign bodies may, as is well known, not only augment or hinder crystallization, but also affect crystallization in other and even more important ways, we made appropriate tests to determine especially if the presence of the oxalate in any particular quantity affected either the type of the crystals or the optical properties of hemoglobin. The optical properties were not in any way appreciably affected. The habit of crystallization, as in the case of Necturus, seemed to be affected in the direction of causing the crystals to be shorter and thicker. The only im- portant influence of the oxalate, apart from the accelerating effect, we found in our experiments with the bloods of the horse and mule, in which we discovered that by modifications in the quantity of oxalate we could obtain a relative abundance of one or the other or of both kinds of oxy- CRYSTALS OF HEMOGLOBINS EMPLOYED IN THIS RESEARCH. 143 hemoglobins that are normally present in these bloods. In no instance did we find any evidence of any influence on the type of crystals that is pecuhar to the species. We did not make any investigations of the possible influ- ences of diseased conditions upon the form of crystallization, because, in the first place, of an insufficiency in our suppUes, and, secondly, because Dr. S. Weir Mitchell and others have found, as far as their studies have gone, that disease is without influence on the type of crystals. Assuming, however, that the presence of ammonium oxalate might have some undis- covered effect upon the morphological or optical properties of the crystals, we, with the few exceptions indicated, always introduced the oxalate, and as nearly as possible in the same proportion. Most of our specimens were in various stages of putrefaction. For- tunately, hemoglobin, in comparison with other proteins, is remarkably resistant to putrefaction, so that even when the plasma proteins are in an advanced state of decomposition the hemoglobin may have merely suffered a partial alteration to reduced hemoglobin and metoxyhempglobin. Upon exposure of the blood to the air, especially shaking with the air or with an atmosphere of pure oxygen, a rapid restoration of oxyhemoglobin is readily brought about, and unless the blood is excessively putrid the exposure of the drops of the prepared blood upon the slides antecedent to the covering with cover-glasses is sufficient to yield good preparations of oxyhemoglobin. Each of our processes is characterized by three major procedures, which are accompanied by such accessory procedures as conditions indicated: (1) the addition of oxalate; (2) laking with ethyl ether (Squibb's); (3) centrifugalization; (4) occasional accessory procedures, such as variations in temperature, the addition of asbestos wool, alumina, etc., to aid in the separation of the stromata or the nuclei of erythrocytes, keeping the prep- arations in a moist chamber, etc. Oxalate was added (a) to prevent coagu- lation, (&) to increase crystalhzability, or (c) to obtain one or another form of oxyhemoglobin present in the same blood. In the process of laking, the ether was usually added in three or four portions, the mixture being shaken vigorously after each addition, and sufficient ether being added to cause marked pressure within the test-tube when the opening of the tube is closed by the finger. When the blood is very putrid no more ether should be used than is absolutely necessary to cause complete laking, otherwise the hemo- globin is Ukely Jo be thrown down in the form of an insoluble precipitate. Caution must also be practised when working with bloods, oxalated or not, which contain nucleated erythrocytes. An excess of ether is Ukely to cause coagulation, and especially so the larger the quantity of oxalate present. One of us (Reichert, Journal of Experimental Medicine, April, 1905) has found that even oxalated defibrinated blood may be converted into a gelat- inous mass by the addition of ether. Centrifugahzation was practised, (a) to collect the corpuscles, and thus get rid of substances which hinder crystallization, and at the same time to secure a concentrated solution of hemoglobin; and (&) to clear the laked preparation of the stromata and other bodies in suspension. Occasionally our specimens were too small to centrifugahze. 144 METHODS FOR PREPARING, EXAMINING, AND MEASURING Our methods are briefly as follows: First method: The whole blood is laked and centrifugalized. If the blood had been defibrinated, oxalate was added before centrifugalization. Second method : The corpuscles are separated by centrifugalization and then laked, oxalate added to the solution, and the solution centrifugalized. Third method: The blood-clot is ground in sand, or the frozen clot corominuted to Uquef action; the fluid thus obtained is laked, oxalate is added, and then centrifugalized. Fourth method : In order to retard crystallization there may be added to the blood, before or after laking, such inert substances as plasma, serum, egg-white, water, glucose, gum, etc. The best results we have obtained by the use of plasma, serum, or a 50 per cent solution of egg-white. This latter is prepared by adding to the white of egg an equal volume of distilled water, shaking violently in a flask for a few moments, and then straining through linen. From 0.5 to 2 or more volumes may be added to the blood in accord- ance with the effect required. The mixture is then centrifugalized until a clear preparation is obtained. Occasionally the solution of egg-white was clarified by agitation with ether and then by centrifugalization before it was added to the already centrifugalized solutions of hemoglobin. Blood very readily crystallizable will often be changed within a few minutes into a magma of crystals, in which case excellent crystals can usually be obtained by using the mother-liquor which has been separated by centrifugaUzation. When the blood is badly decomposed it is better to complete the laking by repeated alternate freezing and thawing than by the addition of ether. The clear preparation obtained by these methods is placed upon slides, and after the margins of the drops have become sufficiently dried cover-glasses are put on, and in the course of an hour or two the covers sealed with Canada balsam. The first method is especially adapted to bloods that crystallize readily, such as those of Pisces and Marsupialia; the second, to those which do not crystallize so readily, as those of Felidce and Ungulata; the third, to those which crystallize with more or less difficulty, as those of Primates, Aves, and Reptilia; and the fourth, to those which tend to crystallize so rapidly as to yield crystals of too minute size, as those of Rodentia and Canidce. Various accessory incidental procedures will be referred to at the proper places. THE VALUE OF THE CRYSTALLOGRAPHIC METHOD OF INVESTIGATION. When a chemical compound solidifies from fusion, solution, or vapor under conditions which are favorable to the development of individuals, its particles tend to arrange themselves in regular order, so that a definite structure is produced. The external form of the individuals is also regular, being bounded by planes in definite relation to each other so that poly- hedral sohds are produced which are called crystals. The regular arrange- ment of the atoms among themselves, and of the molecules which they build up, is so characteristic of substances of definite composition that the crystalhne condition of dead matter is the normal condition. Differences of chemical constitution are accompanied by differences of physicail structure. CRYSTALS OF HEMOGLOBINS EMPLOYED IN THIS RESEARCH. 145 and the crystallographic test of differences of chemical constitution is recognized as the most delicate test of such differences. For instance, in the case of isomerides, the chemical differences between such substances consist in the differences in the arrangement of their constituent atoms, the position of a replaced hydrogen atom in a group (in which several such replacements are possible) altering the structure. Hence the following is true: "Isomeric substances possess different crystal structure." (Groth, Chemical Crystallography, trans, by Marshall, 1906, 63.) Such differences in crystal structure can generally be readily recognized, but to detect chem- ical differences between isomerides by any centesimal chemical analysis is obviously impossible. Chemical differences between such substances are detected by differences in solubility, in melting-point, in rotatory power, in reactions in which the substance is altered or decomposed; but when large numbers of isomerides are possible, as in the enormous molecules of the proteins, the detection of differences between them by purely chemical pro- cesses has thus far, except in rare instances, been found impossible. The crystallographic method is, of course, adapted to detecting differ- ences between substances that show differences in centesimal composition even better than between isomerides, for here the differences in structure may be more profound than in the case of the isomerides; and differences in centesimal composition must of necessity imply differences in structure. Hence the general law may be enunciated: Substances that show differences in crystallographic structure are different chemical substances. THE PETROGRAPHICAL MICROSCOPE AND ITS USB. The necessity of studying small crystals, especially sections of such crystals as are met with in rock sections, has resulted in the evolution of a form of microscope which is at once a goniometer, a polariscope, and an instrument for measuring optic axial angles — ^in short, for determining the physical crystallographic constants of small crystals. Of necessity, in some of its measurements it is not so exact as other instruments that may be employed for the same purpose, for its parts must be light, and its circles can not be read to the same degree of accuracy as those of the reflecting goniometers and spectrometers. The determination of the angle of a crys- tal by this instrument is, under favorable conditions, not accurate to less than 10' of arc. But when it is remembered that carefully made measure- ments on the reflecting goniometer often vary as much as this in different crystals of the same substance it is seen that data of value may be procured with the aid of such a microscope. The polariscope' portion of the petrographical microscope enables the observer to determine the position and relative value of the elasticity axes of crystals, to observe the position of the optic axes, and to determine their inclination to each other and to the elasticity axes. From these data the optical character of the crystal is determined. These optical reactions may be studied by this instrument with as much ease, and in general with as much accuracy, as with the larger and better graduated polariscope; and the data thus obtained are quite as accurate in most cases as those obtained 10 146 METHODS FOR PREPARING, EXAMINING, AND MEASURING by the use of the larger instruments. The use of the special eyepieces arranged with artificial twins of calcite or quartz enables the observer to determine the extinction angles of the crystals with as much accuracy as can be done with any form of polariscope. From such observations, made with the aid of this form of microscope, the following constants may be determined: (1) The plane angles of the crystals, in most cases the interfacial angles, giving the data from which the axial ratios are computed — ^in other words, the morphological constants of the single crystals. (2) The relation of the parts of the composite crystals or twins to each other, their angles, and the position of the twin plane, twin axis, compo- sition plane, and other constants of the twin crystals. (3) The pleochroism of the crystals, the character of the colors of the hght vibrating parallel to the elasticity axes in the crystal. This is effected by the use of the single polarizing prism below the stage. By analyzing this hght with the microspectroscope the differences of tint and color may be given quantitative values in wave lengths. (4) The position and relative values of the hght elasticity axes in the crystals, upon which depend the angles of extinction of the crystals, meas- ured from certain crystallographic axes or planes or edges. In uniaxial crystals (tetragonal and hexagonal systems) there are two such elasticity axes — the ordinary ray, designated as o, and the extraordinary ray, desig- nated as s. Either one of these may be the axis of greater or less elasticity, and according as the extraordinary ray is the axis of less elasticity or of greater elasticity the crystal is called optically positive or optically negative. In biaxial crystals (orthorhombic, monochnic, and triclinic systems) there are three elasticity axes at right angles to each other, and these are desig- nated as a, the axis of greatest elasticity; h, the axis of mean elasticity; and c, the axis of least elasticity. (5) The position and angle of inchnation of the optic axes or hnes of single refraction through the crystals. These always he in the plane of the elasticity axes a and c and the angles between the optic axes are bisected by the axes a and c. According as to whether c or a is the axis bisecting the acute angle, the acute bisectrix, Bxa, the crystal is called optically posi- tive or negative. Thus if Bxa=t, the optical character is positive. The apparent angle between the optic axes is determined by means of an eye- piece micrometer in an observation of the interference figure, looking along the acute bisectrix of the optic axes, and this angle is designated as 2E. The character of the double refraction may be determined by this angle. When good crystals were available for examination the physical data above enumerated could all be determined. Other data were recorded, such as the morphological habit, the character of the heterogeneous aggre- gates formed by the crystals, their relative dimensions, general color, etc. The character of the material under investigation was determined by the use of a Zeiss microspectroscope. As a general rule, the crystals were kept under examination until they ceased to change or until they were destroyed by bacterial decomposition. In this way the changes that the crystals went CRYSTALS OF HEMOGLOBINS EMPLOYED IN THIS RESEARCH. 147 through, the formation of successive crops of oxyhemoglobin and also reduced hemoglobin, methemoglobin, and metoxyhemoglobin were observed. Some crystals kept well for weeks, some altered inside of a day or two. In general the best results were obtained with crystals that formed at room temperature and did not dissolve on slight increase of temperature, but in many cases aU the observations had to be made near freezing tem- peratures. This was done by working in a cold room, at a temperature near 0° C. Even at such a temperature the heat of the body or breath often produced partial solution of some of the crystals, so that the measurements had to be made rapidly. Having usually a number of sUdes at hand, this could generally be done conveniently, as, when the crystals began to lose shape, the slide under investigation could be replaced by another from the supply kept at a temperature below freezing, and the crystals of the first slide would usually soon regain their form. When the blood crystals were not very soluble it was always found more advantageous to keep them at a temperature much above freezing rather than near the freezing-point; they could then be examined and photographed without fear of solution because of an increase of temperature. CHAPTER IX. CRYSTALLOGRAPHY OF THE HEMOGLOBINS OF PISCES, BATRACHIA, AND REPTILIA. While this research is mainly upon the hemoglobins of mammals, certain species of fish, batrachians, reptiles, and birds were investigated, in order that it might be seen whether the results obtained for mammals extended to these classes also. Only a very limited number of species were examined, including 4 fish, 1 batrachian, 1 reptile, and 10 birds; but the results are entirely in conformity with those obtained from the examination of mammaUan blood, as indicating the specific character of the properties of the hemoglobin in each species. In most cases the species were too far separated from each other to show similarities due to their zoological affinities, such as are shown in the hemoglobins of mam- malia, but in the case of the birds the relationship was sufficiently close to bring out family and generic characteristics. PISCES. Barndoor Skate, Raia Icevis. Plate 1. One species of the selachians or cartilaginous fishes, the barndoor skate, Raia Icevis, was examined. This skate was sent to the laboratory from the New York Aquarium, and the blood was extracted by the dis- section of the dead fish. The blood was oxalated and centrifugalized, the corpuscles were ether-laked, a httle water was added, and the mixture again centrifugalized. From the clear solution thus obtained the slide preparations were made. The crystals developed readily and were oxy- hemoglobin, as determined by spectroscopic examination. They occur in single crystals and in sheaf-hke or branching aggregates of these crystals. The individual crystals are rhombic plates. The crystallographic description is as follows: Oxyhemoglobin of Raia Icevis. Orthorhombic: Axial ratio a -.b :6 =0.6008 : 1 : 0.4024. Forms observed: Prism (110), base (001), and from twins the unit pyramid (111) and a macrodome (401). Angles: Prism angle 110AlT0=62°; prism to base 110 A 001 =90°; prism to pyramid 110 A 111=38°; macrodome to base 401 A 001 =69° (calculated 69° 33'). Habit thin tabular on the base (001), the crystals being a very short prism (text figures 1 and 2). Twinning is on the pyramid (111); this form was only observed in twins; or on the macrodome (401), a form also only observed in twins. Also in crystals grown together on the base into groups, which, seen on the edge aspect of the plate, 149 150 CRYSTALLOGRAPHY OF THE HEMOGLOBINS FlOB. 1, 2. Raia lavU Oxyhemoglobin, Fio. 3. Adpenter Bturio Osyhemoglobin, produce sheaf-shaped and fan-shaped groups, and on the flat show the irregularly overlapping plates (see plate 1, figs. 1, 2, 3) . Frequently the orientar tion of the crystals in a group is such that the entire group extinguishes simultaneously in polar- ized light. Extinction is symmetrical when the plate is viewed on the base (001), and parallel when viewed on edge (orthorhombic). The interference figure was not observed, but, from the pleochroism and the elasticities of the axes as observed, the orienta- tion of the elasticity axes with reference to the crystal axes is as follows (text figure 2) : o =6, 6 =a, c=6. From the fact that the figure could not be seen when looking along a and 6, and from the elasticities and pleochroism, the acute bisectrix of the optic axes Bxa=a. The optical character is hence negative. Three species of the Teleostomi or Bony Fishes were examined. In the order Actinopterygii the common sturgeon, Acipenser sturio, was examined. Stubgeon, Acipenser sturio. Plates 1 and 2. The blood was obtained from the living fish, oxalated, ether-laked, and centrifugaUzed, giving a clear solution, from which the preparations were made in the usual manner. The slides were prepared inside of 4 hours from the time of the bleeding of the fish. The crystals are deposited rapidly from the solution, and inside of a few minutes after the blood is placed upon the slide an abundant crop of small lath-shaped crystals has formed. After 1 hour the entire slide is filled with a felt of these crystals, and the solution is colorless. No particular change in the sUdes was noted after this condition was reached, showing that the crystals were quite insoluble in the plasma. In order to retard the rate of formation of the crystals a portion of the oxalated normal blood, which was found to be filled with crystals after standing 24 hours, was diluted with 5 volumes of water and centrifugalized. The clear supernatant hquid was then used to make sUde preparations, but while it readily crystallized, the crystals were not so perfect as those made from the whole blood. The crystals from the whole blood were therefore used for the examination. The color of the crystals, a brownish-red, indicated that they were oxyhemoglobin mixed with some methemoglobin, which latter was observed in the normal blood of other fish when obtained fresh from the living animal. The blood was not examined by the spectroscope. The crystallographic description of the crystals is as follows: Oxyhemoglobin of Acipenser sturio. Orthorhombic: Axial ratio not determined, only one fundamental angle being observed, and that a doubtful one. Forms observed: Macropinacoid (100), brachypinacoid (010), brachydome (Oil). OF PISCES, BATRACHIA, AND REPTILIA. 151 Habit of the crystals long lath-shaped (text figure 3), the length 20 to 30 times the breadth, which latter is about 4 times the thickness. The crystal appears to consist of the two pinacoids, terminated by a flat dome. This is probably the brachydome. Its interfacial angle was measured very roughly as 110°, giving the angle between face normals as 70°. The large plane in the prismatic zone is then taken as the brachypinacoid (010), and the smaller plane, visible when the crystal is on edge, is hence the macropina- coid. On the flat, the long lath-shaped crystal has square ends, measured as 90° with the sides; and on edge the flat brachydome terminates the crystal. Pleochroism is rather strong; for a, the direction of greatest elasticity, which is parallel to the length, the color is pale yellowish; b and c are nearly equal, and the color is reddish-brown. Extinction is straight in both the edge view and on the flat or side view. The polarization and pleochroie colors are not interfered with by the color of the plasma, as practically all of the hemoglobin crystallizes. Between crossed nicols, the colors seen on the flat ranged from blue-slate of the first order up to straw-yellow and orange of the first order in the different thicknesses of crystals, indicating a moderately strong double refraction. No interference figure was observed. Shad, Alosa sapidissima. Plates 2, 3, and 4. Shad blood was obtained by bleeding the Uving fish, and also by ob- taining blood from dead fish purchased in the market. In the former case it was either oxalated or allowed to clot; in the latter it was obtained in the form of clots from the larger vessels, etc. The clotted blood was treated by oxalating and ether-laking, and also by grinding the clot with sand and ether-laking, centrifugaUzing, and oxalating the clear blood. The fresh oxalated blood, either laked by ether and centrifugalized, or the corpuscles broken down by repeatedly freezing and thawing the blood, always showed a combination of methemoglobin and oxyhemoglobin, the material often described as "methemoglobin." Much difficulty was experienced in get- ting rid of the nuclei of the corpuscles by centrifugaUzing, but this was partly overcome by using thoroughly clotted blood and breaking up the clots by grinding in sand. On allowing the blood to stand in a corked tube in the refrigerator, the blood usually passed largely into reduced hemoglobin. The blood that was freely exposed to the air and had been kept for some days, or the deoxidized blood exposed to the air, and, finally, the clotted blood obtained from fish that had died in the air and had been dead a few days, as obtained in market, always contained some methemoglobin which would crystaUize as such and not as metoxyhemoglobin. It would seem, therefore, that the blood of the shad during the spawn- ing season, in which it is obtained in our rivers, contains a large proportion of methemoglobin in combination with oxyhemoglobin, or a substance inter- mediate between oxyhemoglobin, methemoglobin, and metoxyhemoglobin; and that further exposure to air changes this to pure methemoglobin in part, leaving a residue of pure oxyhemoglobin, which two substances may be crys- tallized simultaneously from the blood in distinct crystals of pure methemo- globin and pure oxyhemoglobin, which substances do not form in the freshly drawn, slightly oxidized blood. This occurrence of metoxyhemoglobin in the freshly drawn blood may be due to the state of inactivity that the digestive organs of the animal are in during the spawning season, as the fish do not feed during this time. The same condition in the blood is noted in the case of bears, for example, during the hibernating period. 152 CRYSTALLOGRAPHY OF THE HEMOGLOBINS From the blood of the shad, therefore, crystals of the following sub- stances were obtained: (1) Oxyhemoglobin, from blood that had been exposed to the air. (2) Metoxyhemoglobin from the fresh blood. (3) Methemoglobin from blood that had been exposed to the air, with (1) or from deoxidized blood, with (4), but not with (2). (4) Reduced hemo- globin * from stale blood that had not been exposed to the air. This was also formed by reduction of (2). (1) Oxyhemoglobin of Alosa sapidiasima. Monoclinic: Axial ratio a :b : 6 = 1.804: : 1 : 6; /3=68°. Forms observed: Prism (110) and base (001). Angles: On base (001) edges 110-00lAlT0-001=58°; edge of (110-lTOAOOl) =/?=68°. The angles of the plates varied, the acute angle being often not the supplement of the obtuse angle in the same plate, this difference being apparently due to some form of twinning. Thus, the acute angle often ranged up to 62°, and some were exactly 60°. The obtuse angle was always 120° or over, up to 124° and even 125° in a few cases. But in simple, untwinned crystals the angles of 58° and 122° seemed to be the average. Figs. 4, 5, 6, 7. Aloaa tapiditnma OzyheinoBlobin. Habit of the crystals (text figures 4 and 5) tabular on the base (001), the prism very short, making the crystal a rhomboidal plate with the plane of symmetry including the long diagonal of the plate. Twins with the base (001) as composition face, Manebach type (text figure 6), also in "homogeneous regular growth" like a twin on the pyramid, but with the two parts uniting on the base (001) and the prism edges (110) and (1 10) in juxtaposition (text figure 7). This twin, very common in all these tabular crystals of the monoclinic system, is called the "horse-type" of twin, and is fully described under Horse (p. 192). Regular growths with methemoglobin formed (heterogeneous regular growths) the methemoglobin crystals, which are hexagonal, forming in symmetrical position on the monoclinic oxyhemoglobin (text figure 14; see also plate 4, figs. 21, 22, and 23). The twinned crystals of the oxyhemoglobin with angles of 60° were usually found so overgrown; but this was not always the case, as may be seen by reference to the figures, see plate 4, fig. 24. Crystals are strongly pleochroic; a pale yellowish-red, b deep red, c deep red; the colors for light vibrating along b and c are about alike. Axial plane A. to axis b or in the plane of symmetry. Orientation of the elasticity axes is as follows: 6=6, aAo=6° in the obtuse angle, c A (o and the crystal is positive. From the characters of these methemoglobin crystals, it is very likely that the sub- stance is really a mimetic twin and only pseudohexagonal. It is not very permanent, but decomposes and produces a granular brownish precipitate, leaving the monoclinic crystals (usually now changed to reduced hemoglobin) unaltered. Fiaa. 13, 14, 15. Aloaa aapiditrima Oxyhemoglobin. 156 CRTSTALLOGEAPHT OF THE HEMOGLOBINS These methemoglobin crystals were found in blood of shad that had died in the air, and in freshly drawn blood that had been exposed to the air, and seem to be due to a separation of the metoxyhemoglobin into methemoglobin and oxyhemoglobin, which latter may be afterwards changed to reduced hemoglobin before the methemoglobin disappears. The formation of the pure methemoglobin, which crystallizes in these hex- agonal plates, is probably due to the further oxidation. Gabp, Cypiinus carpio. Plates 5 and 6. Blood of the carp was obtained from live fish caught at Gloucester, New Jersey. It was oxalated, ether-laked, and slides prepared within a few hours after it was collected. The blood had a brownish color, and was probably the metoxyhemoglobin mixture. After standing in a test-tube for 24 hours it was practically all converted into reduced hemoglobin. Preparations of this were also made and examined. Both the metoxyhemoglobin and the reduced hemoglobin crystallized readily, but without any separation of pure methemoglobin. They are isomorphous, having apparently almost the same axial ratio, and perhaps the axial ratios are actually identical. (1) Metoxyhemoglobin of Cyptinus carpio. Orthorhombic: Axial ratio, a :b : 6= 0.949 : 1 : 1.03. Forms observed: Prism (110), base (001), and, from twins on the macrodomes, also macrodomes (301), (201), (302). Angles: 110 A 110=93° (87° normals). 110 A 001=90°; from twins 302-502 = 68° 30'; 201 A 201=55° 15'; 301 A 301=37° 30'. Habit generally tabular, nearly square plates, formed by flattening on the base (001) in combination with the prism (110) (text figures 16 and 17); also long prismatic, formed by development of the prism in the ,C same combination (text figure 18). The prismatic crystals are the first to ap- pear; these are gradually absorbed as the plates develop. Twins are common in the prismatic habit, apparently on the macrodomes noted above; they are not so common in the tabular habit, but apparently the twin on (302) oc- curs. Parallel growths are common in = the tabular form; perhaps also homo- geneous regular growths occur. This parallel growth produces a piling up of the plates, and composite crystals and groups result. Pleochroism is marked; the colors \9 are shades of brownish-red. Orienta- tion of the elasticity axes was made out as follows: a=b, b=a, t=d. The interference figure was not observed. Extinction is symmetrical on the plates when examined on the base (001) and straight when examined on edge. (2) Reduced Hemoglobin of Cyprinus carpio. Orthorhombic: Axial ratio, a :b : er OxyWo- crystals showed a slight pleochroism. On the basal pinacoid giobin. the crystals are singly refracting when examined in parallel polarized light; in convergent light, the uniaxial interference figure is seen as a very dusky cross. Examined on edge, the vertical axis (the extraordinary ray) was seen to be the direction of less elasticity; hence s>(0 and the optical character is positive. When examined on edge, the crystals polarized as a whole and did not indicate any appearance of twinning, but from the fact that similar tetragonal characters are produced by what appears to be homogeneous regular growth in the case of the whistUng- swan blood — Olor columbianus — which is distinctly orthorhombic, it is quite possible that these blood crystals are only pseudo-tetragonal. Tkumpeteb Swan, Olor buccinator. Plates 9 and 10. This specimen of blood was received from the Philadelphia Zoological Gardens. The blood was oxalated, ether-laked, and centrifugaUzed, and preparations made as usual. The slides were kept at a temperature near the freezing-point, but no crystals developed until after 24 hours. Even then only scattered crystals appeared, excepting in two slides. The crystals 164 CRYSTALLOGEAPHY OF THE HEMOGLOBINS OF AVES. were very soluble and had to be kept at about 0° C. during the examination and photographing. Several other trials were subsequently made with the same blood, but they failed to yield crystals. The crystals obtained were found to be oxyhemoglobin by spectroscopic examination. Oxyhemoglobin of Olor buccinator. Tetragonal or pseudo-tetragonal: No axial ratio determined, as no pyramidal planes developed. Forms observed: Prism (110) and base (001). Angles: 110 A 1T0=90°; 110 A 001=90°. Habit thin to thick tabular, by development of (001) and (110) (text figure 36); in single crystals and in groups, often arborescent by growing together on the base or on a pyramid; also in clusters, but without definite appearance of twinning. Some of the large crystals were evidently composite, but did not show any appearance of twinning when examined on edge in polarized light. Pleochroism faint, apparently abnormal; the absorption for the direction of greater elasticity appears to be slightly greater than for the direction of less elasticity. Colors are deep oxyhemoglobin red; somewhat paler for cd. Uniaxial, singly refracting on the base in parallel polarized light, and showing a faint dusky cross in convergent light. Seen on edge the double refraction is very weak, but is observable with the aid of the quartz wedge, etc. ; when it is seen that w is the direction of less elasticity and hence oj > s and the optical character is negative. The fact that the double refraction is so weak would favor the suspicion that the crystals are composite and really only pseudo-tetragonal, as is the case with some crystals in the next species, Olor columbianus. C^ 36 39 Fio. 36. Olor buccinator Oxyhemoglobin. Fios. 37, 38, 39. Olor colvmbiama Oxyhemoglobin. Whistling Swan, Olor columbianus. Plate 10. The specimen was received from the Zoological Gardens at Washington and was clotted and in a very putrid condition. The clot was ground up with sand, etherized, and the liquid obtained oxidized by exposing it to the action of pure oxygen. It was then centrifugalized, and slides prepared as usual, the drops being allowed to become very concentrated before cover- ing. In only two slides out of some two dozen prepared did crystals appear. The crystals were rather dark, but were oxyhemoglobin. Oxyhemoglobin of Olor columbianus. Orthorhombic : Axial ratio a : b : i =0.9657 : 1 : 6; also pseudo-tetragonal by homogeneous regular growth. Forms observed: Prism (110), base (001). Angles: 110 A 1T0=88° (normals); 110 A 001=90°. CEYSTALLOGRAPHY OF THE HEMOGLOBINS OF AVES. 165 Habit thin tabular in simple crystals (text figure 37), becoming thicker by homo- geneous regular growth, in which the prism-base edge of one member of the group is in line with prism-base edge of the next following layer (text figure 38); the successive layers arranged in polysynthetic order (text figure 39) . Composite crystals are produced by this piling up of the successive individuals of the group, that finally develop angles of 90° for the plate, by averaging of the angles 88° and 92°; the result not being dis- tinguishable, from a tetragonal crystal when examined on the flat, and even giving a dusky uniaxial cross in convergent light. This kind of crystal shows the composite character by being less regularly developed than the simple crystal. On edge the sepa- rate individuals in the group are at once distinguished by pleochroism, using one nicol, and.with crossed nicols by the difference of double refraction. Some were seen consisting of two or three individuals, but many consisted of a much larger number in the poly- synthetic arrangement. The crystals were large, but did not grow into arborescent groups as in the case of those from the blood of Olor buccinator. Pleochroism is rather marked on the flat view, less so on the edge; the absorption is c > b > a. The colors are shades of the oxyhemoglobin red, a being yellowishjed. The orientation of the elasticity axes is a=a, 6=tS, c=&; the optic axes being in the plane of the basal pinacoid. No interference figure appears therefore on the flat view (001) ; but on edge, when looking nearly along a, one brush of the figure is seen. In the composite crystals this also is visible, the arrangement of their elasticities being as shown in text figure 39. From these figures it is seen that the axis b keeps its position in the regular growth, and the axes a and c alternate in the successive layers. In case of very thin layers in the composite crystals, so that the layers become too thin to show by the microscopic examination, this averaging of a and c would greatly reduce the amount of the double refraction, so that it might become almost zero, which is the condition in the species of swan examined, and leads to the suspicion that in the blood of Olor buc- cinator the crystals are only pseudo-tetragonal. From the position of the brush of the interference figure it is evident that the acute bisectrix Ba;a = ct, and the optical character is negative. The mimetic crystals produced by the homogeneous regular growth are singly refracting when examined on (001) and are not strongly doubly refracting when examined on edge view, especially when the individual layers are thin and not of the same thick- ness throughout; they are hence in some cases truly pseudo-tetragonal. The averaging of the angles of 88° and 92° to 90° makes them strictly tetragonal in form. Chicken, Gallus domestica. Plate 11. Blood was obtained from the living chicken, oxalated and centrifu- galized, and only the corpuscles used. The corpus- cles were treated by the usual method and crystal- lized at a temperature near the freezing-point. All examinations had to be made at the same low tem- perature, the room being kept at about the freezing- point or below. Very few slides showed crystals, and they were usually isolated or in small groups. Blood from two different birds was examined at different times, but the habit was about the same in both cases. The crystals were oxyhemoglobin. Oxyhemoglobin of Gallus domestica. Orthorhombic : Axial ratio a :b : 6 =0.949 : I : 6. Forms observed: Prism (110), base (001). Angles: 110 A 110=87° (normals); 110 A 001=90°. ^"»- **• ^Lm^iSbi^r*"^ °^''- 166 CEYSTALLOGRAPHY OP THE HEMOGLOBINS OF AVES. Habit tabular, the square tables (text figures 40 and 41) aggregated into groups by piling up of the plates or perhaps by twinning on an axis normal to the edge (110-001) ; also by what appears to be twinning on a dome; the crystals usually occur in isolated clusters. Sometimes skeleton crystals are seen that look tetragonal, but are orthorhombic according to their optical characters. Pleochroism is not very marked, hardly noticeable on the flat view, but stronger on the edge view, especially when looking along b. Orientation of the elasticity axes, a =6, h=a, c=6. The plane of the optic axes is the macropinacoid; Bxa=a, hence the crystal is optically negative. Absorption c> 6 > a. On looking along a in conver- gent light the interference figure is seen with the brushes rather widely separated. Quail, Colinus virginianus. Plate 12. The blood was obtained from the hving bird, and prepared in the usual manner. Corpuscles were used for extraction of the oxyhemoglobin which was tested by the spectroscope. The crystals form sparingly and melt readily at a little above 0° C. Oxyhemoglobin of Colinus virginianus. Orthorhombic: a:b -.6 =0.9657 : 1 : 6. Angles: 110 A 1T0=88° (normals); 110 A 001=90°. Forms observed: Prism (110), base (001). Habit thin to thick square tabular; the tables consisting of the above combination and varying in thickness from one-fourth to one-half of the width of the plate (text figures 42 and 43) ; the cry- stals grew singly and in groups, but did not grow in the radiating form of the chicken oxyhemoglobin. Perhaps they twin on the axis normal to the prism-base edge, as the plates pile up on the base and overlap somewhat irregularly. Examined on (001), the crystals show no perceptible pleochroism; on edge the pleochroism is weak, but notice- able. The angle of the prism is so near 90° and the plates so irregular, due to overlapping, etc., that it is difficult to determine the exact orientation of the optic axes; the extinction is straight on the edge view and symmetrical on the (001) view. One of the diagonals is readily made out by the quartz-wedge to be an axis of greater elastic- ity than the other, but on edge views it is seen that c=6. The plane of the optic axes is probably the macropina- coid, and when looking along a (in edge view) the inter- ference figure is seen, showing Bxa=a, and the optical Absorption is c=b (nearly) > a. Pleochroism: c and b Fioa. 42, 43. Colinus virginiantu Oxy- hemoglobin. character is hence negative. deep red, a paler red. Guinea-fowl, Numida meleagris. Plate 12. The blood was obtained by bleeding the living bird and was oxalated and prepared in the usual manner. Crystals formed readily and did not appear to be very soluble, as they remained in perfect condition at room temperature. The oxyhemoglobin crystaUizes readily at ordinary temper- ature in well-formed crystals, in contrast to the crystals obtained from the bloods of most of the birds examined. CRYSTALLOGRAPHY OF THE HEMOGLOBINS OF AVES. 167 Oxyhemoglobin of Numida meleagris. Orthorhombic : Axial ratio a : b : 6 =0.554 : 1 : 6 or a is still noticeable. Extinction is straight on all edge views and symmetrical on the base. The orientation of the elasticity axes is a =6, 6 =o, c =«5. The axis of least elasticity c appears to be the acute bisectrix; Bxa=c, hence the optical character is positive. These crystals were gradually converted by paramorphous change into pure methe- moglobin, giving the absorption band 630 fifi to 605 n/i and extending to 680 (ifi in the red; this change to pure methemoglobin appeared first in these crystals, although pure methemoglobin was later seen in the form of b-crystals also. (b) Metoxyhemoglobin of Columba livia. Orthorhombic: Axial ratio a :b : 6 =0.4615 : 1 : 6. Forms observed: Prism (110), base (001). Angles: 110 A 1T0=49° 33'; 110 A 001=90°. Habit tabular, in rather acute rhomboidal plates (text figures 51 and 52), usually occurring singly, elongating on the macrodiagonal by parallel growth or sometimes on the brachydiagonal; but not twinning in the way commonly seen in these tabular crystals, on an axis normal to the prism-base edge. After the crystals of this metoxyhemoglobin 6-type had passed into the pure methemoglobin, they formed a sort of regular growth with the tufts of needles of the reduced hemoglobin, the needles being arranged in tufts growii^ nearly normal to the prism-base edge of the crystals (see plate 13, fig. 77). These crystals showed a laminated structure parallel to the plane of symmetry, perhaps indicating a cleavage in that direction (see plate 13, fig. 76). The color of these crystals is reddish-brown, rather dark, and the spectrum showed absorption bands at 640 /ift to 615 ftfi, rather faint (the methemoglobin red band) ; and stronger bands at 580 nft to 565 fifi and 550 fifi to 530 fift, the oxyhemoglobin bands. When they finally became converted to methemoglobin, the oxyhemoglobin absorption bands disappeared entirely. CRYSTALLOGRAPHY OF THE HEMOGLOBINS OP AVES. 169 FiQS. 51, 62. Cohtmba livia Metoxybemoglobin. I^Q. 63. Colmiiba livia ^-Oxyhemoglobin. Pleochroism is very strong; a colorless or nearly so, b deep brownish-red, c very deep brownish-purple. The absorption for c and b is very strong. Double refraction strong; the extinction is straight on edge views and symmetrical on the base. Orienta- tion of the elasticity axes is a=b, b=a, c=6. On the base, traces of an interference figure are seen, but the brushes pass out of the field; looking along a the complete figure is seen, showing that the acute bisectrix Bxa= w, and the crystal is positive. On the base, in convergent light, the uniaxial cross is readily seen; in some cases the crystal is slightly biaxial, as in the nearly uniaxial micas, and the axis of least elasticity is normal to a prism face. The biaxial crystals are also distinctly positive. The separation of the brushes is only very slight, the angle 2E is very small. While these crystals are seen in all sizes, and do not appear to be composite, there can be little doubt that they are really mimetic hexagonal only and are twins of the a-oxyhemoglobin on the base in one of the two forms of twinning that have been described under the a-oxyhemoglobin. If the twin laminae were thin enough, the polarization test would not show the composite character and this would be especially true if, as is usually the case in these twins, the same layer did not run as a plane entirely across the basal surface. In looking through from side to side the different orientation of the layers would hence average, and neutralize each other. Of course, this averaging would happen on the flat view to a still greater degree, and the elasticity axes a and b in different orien- tation would completely extinguish each other, making a uniaxial effect. This may be done artificially with only three plates of mica, twinned as these a-oxyhemoglobin crystals OP THE MARSUPIALIA, EDENTATA, AND SIRENIA. 177 twin, and has been observed in the oxyhemoglobin of many species that twin in this way, where it is easily seen that the crystal is composite. If both kinds of twins formed in the same crystal, the averaging of the elasticities might be perfect. But, as has been observed in other species that form hexagonal plates (compare rats, squirrels, etc.), the growth of the composite plate by this form of twinning produces an averaging of the angles, so that prism angles that are nearly 60° (58° 30' as in this a-oxyhemoglobin) become exactly 60° in the twin. It might be possible that this crystal was an averaging of right- and left-handed forms, resulting in the more symmetrical mimetic twin. From the forms of twinning assumed, the elasticity axis, c, remains always in the same position in all of the members of the composite crystal, and hence the vertical axis, e, becomes the axis of less elasticity, and the composite remains positive. If the above view of these crystals is correct the substance of the a-oxyhemoglobin and of the /?-oxyhemoglobin may be the same unless perhaps the /?-oxyhemoglobin is a union of right-handed and left-handed crystals of the a-oxyhemoglobin. Reduced Hemoglobin of Didelphis virginiana. Monoclinic: Axial ratio a : b : 6 =1.963 : 1 : i; ^=66°. Forms observed: Prism (110), base (001). Angles: Prism angle traces on the base of 110 A 1T0=54°; prism edge to base llO-lIO A 001=^=66°; base to plane of symmetry or to side prism edge 001 A 110- 1T0=90°- a Figs. 66, 67. Didelphis virginuma Reduced Hemoglobin. Figs. 68, 60. D. virginiana o-Carbon-mono^dde Hemoglobin. Habit, rhombic plates with oblique sides, composed of base and prism, the crystals generally very perfect and sharp (text figures 66 and 67). They usually occur singly, but also twin with the normal to the base as the plane of twinning and the twin axis normal to the prism-base edge, the composition face being the basal pinacoid. This type of twin, " horse-type," is seen in a-oxyhemoglobin (text figure 51) and is the common twin on the base in all of these monoclinic reduced hemoglobins and oxyhemoglobins, especially when the prism angle is near 60°- These twins are often complex and the polysynthetic arrangement is very common. The crystals are readily distinguished from the a-oxyhemoglobin by their color, and by the fact that they occur singly and not in parallel growth, as is so commonly the case in the a-oxyhemoglobin. In the photographs they appear as lighter-colored, more transparent crystals than the oxyhemoglobin crystals. Pleochroism is very strong; a very pale violet, nearly colorless; b deep reddish; c deep claret-color to purple. Ex- tinction is symmetrical or nearly so on the flat basal face; on edge looking along the axis b it is oblique; the extinction angle is a A a = 13°, in the obtuse angle. The orientation of the elasticity axes is as follows: The axial plane is the plane of symmetry; a a a = 13° in the obtuse angle, b = &, CA w, and the optical character is positive. Compared with the jS-oxyhemoglobin these crystals are seen to have identical characters, and there is probably no doubt that if the ^-oxyhemoglobin crystals are mimetic twins these are also. The forms of twinning noted for the a-oxyhemoglobin would produce such mimetic forms if the twinning was repeated or polysynthetic. Such twins, with the twin axis lying in the basal pinacoid and normal to the prism-base edge, and the base as the composition face, were apparently observed in the a-CO-hemoglobin along with the type of twin already described. The close resemblance of the oxyhemoglobins and the CO-hemoglobins in this species is what might be expected from the other resemblances between these compounds. It will be noted, however, that the reduced hemoglobin varies from either in the incli- nation of the base and in the prism angle. Table 38 shows these differences plainly. Table 38. — Differences of the oxyhemoglobins and CO-hemoglobins in Didelphis virginiana. Substance. Axial ratio. Angle ?. Prism angle. Gxtino- tion angle. Optical character. -System. a-oxvhemo&rlobin 0=1.7856 0=1.804 = 1.963 O 48 (437) 41 66 90 90 o t 58 30 58 30 54 60 60 o 17 13 13 Positive Do. Do. Do. Do. Monoclinic. Do. Do. Hexagonal. Do. jd-oxvh.einofirlobin Tasmanian Devil, Sarcophilus ursinus. Plate 18. The specimen was obtained from the National Zoological Park at Washington, District of Columbia, and consisted of about 2 c.c. of oxalated blood preserved in our usual collecting tube. The blood was centrifugal- ized and the corpuscles separated and laked with ether. Preparations were made in the usual manner. The blood crystallized very readily and photographs could be taken within 2 hours of making the preparations. The blood being in good condition, the crystals were oxyhemoglobin, as determined by the spectroscope. 180 CETSTALLOGKAPHT OP THE HEMOGLOBINS Oxyhemoglobin of Sarcophilus ursinus. Monoclinic: Axial ratio a:b: 6 =1,804 :1: 6; /?=69° (111°)- Forms observed: Prism (110), base (001), clinopinacoid (010); also in twins (111) a positive hemipyramid, of which the angle was not determined- Angles: Traces of prism on the base, edge 110-001 A 110-001=58° (122°); edge of 110-lTO A 001=/? =69° (111°). Habit prismatic, also tabular; the first crystals to appear are usually long thin prisms and also shorter prisms with the length varying from 20 times to 6 times the thickness (text figure 71); later nearly equidimensional blocks appear, short prisms with ratio of length to thickness varying from 2 : 1 to 1 : 2 (text figure 72) ; finally a few crystals become tabular on the base, with the prism reduced to perhaps one-tenth of the thickness or width of plate (text figures 73 and 74). Twinning was observed with the plane of twinning and composition plane the prism face (twin on the prism), both contact (text figure 75) and interpenetrant (text figure 76) twins; also long prisms twinned on a positive hemipyramid, of which the angle was not determined, called (Til), as the plane of twinning and the twiuning axis normal to this plane; these were inter- penetrant twins. a v^ 71 01 a / NIX72 I^Vfl ^^ U If / Figs. 71, 72, 73, 74, 75, 78. SarcophUut urtinut Oxyhemoglobm. Pleochroism strong; a pale yellowish, nearly colorless; B red, c deep red. Extinc- tion is symmetrical on the base in the plates and all sections in zone of 6; on side view, on (010) the extinction is oblique with an extinction angle of 10° from the prism edge, or 11° from the trace of the base (001). The axial plane is the plane of symmetry (010) and the orientation of the elasticity axes is c A <5 = 10° lying in the obtuse angle; a A a = 11° in the obtuse angle; b=b. The interference figure is seen on this basal section, with the brushes slightly unsymmetrical and rather widely separated. 2.E= above 80°, and the angle between c and the normal of (001) is 11°. The acute bisectrix, Bxa=c, and the optical character is hence positive. Spotted Dastxtrb, Dasyurus maculatus. Plate 18. The blood as received was thick and dark colored, somewhat putrid. It was prepared by oxalating and freezing and thawing, then centrifugaliz- ing. The best crystals were obtained by diluting the centrifugalized blood with an equal volume of water. The needle-like, prismatic crystals appeared soon after making the preparations. After 48 hours the crystals had developed into large plates. The crystals were all reduced hemoglobin, and even in ordinary light varied much in color, owing to the strong pleochroism. OP THE MARSUPIALIA, EDENTATA, AND SIRENIA. 181 Redticed Hemoglobin of Dasyurus maculaiiis. Monoclinic: Axial ratio a : b : b > a. On the base, the crystal extinguishes symmetrically, with the axis of greater elasticity bisecting the acute angle of the plate. On the clinopinacoid sections the extinction is oblique, making an angle of 9° with the edge 010-001 or with the clino-axis, a, in the obtuse angle. The plane of the optic axes is normal to the plane of symmetry and a—b. The orientation of the other elasticity axes is, B a a=9° in the obtuse angle, the extinc- tion angle; and c A (5=26°, in the obtuse angle. Traces of the brushes of an interfer- ence figure show on some of the plates that are tilted, with the orientation as above for the elasticity axes. On the clinopinacoid sections the interference figure shows when looking along a, the brushes being only slightly separated. The acute bisectrix is hence the axis of greatest elasticity, Bxa^f^, and the optical character is hence negative. Rat-kangaroo, JE-pyprymnus rufescens. Plates 21 and 22. The specimen was received from the National Zoological Park at Washington. The blood was putrid, but, being collected in oxalate, was liquid. The contents of the tube were centrifugahzed. The separated cor- puscles were then laked with ether and again centrifugahzed, and slides prepared in the usual manner. CrystaUization was rapid; the crystals formed readily at room temperature and appeared to be relatively insoluble. The color of the solution was almost discharged, showing that the crystal- lization was very complete. Negatives were made within 5 hours after the sUdes were prepared. The crystals were oxyhemoglobin as determined by the spectroscope. OF THE MARSUPIALIA, EDENTATA, AND SIRENIA. 185 Oxyhemoglobin of JEpyprymnus rufescens. Monoclinic: Axial ratio a : b : 6 =1.4825 : 1 : 1.338; )S=67°. Forms observed: Prism (110), clinopinacoid (010), base (001); and, in twins, positive hemiorthodome (lOl), orthopinacoid (100). Angles: Prism angle 110aT10 = 112° (68°); 001 A 010=90°; 100 A 001=/? = 67°; 100 A T01= 51°, from twin. Habit prismatic, elongated on the vertical axis, with the prism (110) and base (001) alone (text figure 89) or in combination with the clinopinacoid (010) (text figure 90), and frequently flattened on the clinopinacoid; the base produces an obUque termi- nation in all crystals. The prisms are sometimes long, but in the greater number of crystals are not more than 4 to 6 times as long as wide. The orthopinacoid is not de- veloped as a face and only appears in twins; hence no square crystals are seen. Twin- ning is of several types, interpenetrant twins on the prism being common (text figure 91), see plate 22, fig. 127; twins also form on the orthopinacoid (text figure 92) (gypsum type) and on the clinopinacoid similar to the carlsbad type (text figure 93). The twin on the positive hemiorthodome (lOl), from which the value of 6 was obtained, is an interpenetrant twin, often seen in crystals with the corhbination (110) (010) (001). The base of one member is nearly parallel with the orthopinacoid or prism edge of the other and the acute angles are opposed with the obtuse angles pointing outward (see text figure 94). In the twins of the carlsbad type the opposed prism faces on either side of the plane of twinning appear to be developed more than the other pair in each case. The twins on the prism are not only interpenetrant, they are frequently juxtaposed, and even in this case polysynthetic (see plate 22, fig. 128). C • P\ Oi M ifs ^ \^ 89 V 90 91 92 <;; \y 9J FiQB. 89, 90, 91, 92, 93, 94. Mpuvrymma rufacmi Oxyhemoglobin. Pleochroism is very strong; a pale yellowish-red, b rather deep red, c very deep red; the pleochroism is readily observed on account of the almost complete crystalliza- tion of the oxyhemoglobin, leaving a nearly colorless solution. Extinction on all of the usual aspects is oblique, the crystals generally presenting the clinopinacoid or prism face. In the twins on the orthopinacoid, prism, and orthodome the extinction is sym- metrical with the plane of twinning. The plane of the optic axes is the plane of symmetry, the orientation of the elasticity axes is a A o = ll° in the obtuse angle; 6=6, c A «! = 12° in the obtuse angle. Only brushes of the interference figure in unsymmetrical arrange- ment were seen on such optical sections as could be observed; the optical character appeared to be negative, or BXa = ci. Kangaroo, Macropus giganteus (?). Plate 22. The specimen was received from the National Zoological Park at Washington during the summer and was kept frozen in the original collect- ing tube until examined. The blood was rather putrid and contained many- small clots and amorphous matter. It had been drawn into a collecting tube supplied with oxalate, hence did not clot after the specimen was 186 CRYSTALLOGRAPHY OF THE HEMOGLOBINS placed in the tube. Ether added to lake the blood appeared to increase the precipitate of amorphous granular matter, and centrifugalizing did not entirely free the blood from this precipitate. The crystals formed readily at ordinary room temperature and did not appear to be very soluble. Pho- tographs could be made inside of 4 hours after the slides were prepared. Examination with the spectroscope showed the crystals to be oxyhemo- globin. Oxyhemoglobin of Macropus giganteus. Monoclinic: Axial ratio a : 6 =1 : 0.497; ^9=87° (93°). Forms observed: Orthopinacoid (100), clinopinacoid (010), base (001); also an orthodome (lOl). Some crystals apparently showed a square prism (110) in place of the two vertical pinacoids; this was not clearly made out. Angles: 100 A 001=j9=87° (93°); 001 A 010=90°; 100 A 101=66°. The posi- tive hemiorthodome (TOl) was also observed as a plane of twinning with angle between the 6 axes of the twui of 48° 30' as measured. This gives the angle 100 A TOl =65° 45'. Habit lath-shaped crystals, consisting of orthopinacoid (the principal plane) and clinopinacoid; the two forming a flattened prismatic crystal elongated along the vertical axis and terminated by the base (001) (text figure 95). They, being flattened on (100), generally present this aspect and hence appear to be terminated by a plane normal to the length, but the square end is produced by the aspect in which the crystal is usually presented, and on an edge view, looking along the 6-axis, the end is seen to be oblique. The orthodome was seen in a few crystals only. The lath-shaped crystals aggregate into sheaf- like bundles and stellate radiating groups, and also grow singly. Twins on the orthopina- coid (100), gjrpsum type, were occasionally seen (text figure 96) and one distinct twin on the hemiorthodome (TOl) was observed with the angle as given above (text figure 97). Pleochroism is very pronounced; a pale pink, 6 rose-pink, c deep rose-pink. Extinc- tion is straight or nearly so in both side and edge views. The end view was not seen. The orientation of the elasticity axes is e= 6 > a. Extinction is straight on (100) and on (010). The orientation of the elasticity axes is a=a; i=b; c=6. No interference figure was seen on b; but on a, or looking along a, traces of a figure were seen, the brushes passing out of the field in open position. The aspect looking along c could not obtained, but there seems to be no doubt that Bxa=c, or that the optical character is positive. The axis of <5=c is evidently the axis of least cohesion, as is shown by the cleavage and by the solution of the crystals in this direction. The axis of greatest elasticity a is the direction in which the crystals develop after the full length has been attained. SIRENIA. Manatee, Manatus americanus. Oxalated blood was received from the New York Zoological Gardens, and was prepared as usual. The crystals are very soluble and dissolve rapidly when brought into a warm room. Oxyhemoglobin of Manatus americanus. 103 p — ■ ' ^'•'^^ J. Orthorhombic : Axial ratio 0.949 :!:(}. "^-. ^ ^ J — ■ Forms observed: Unit prism (110), macropinacoid (100), base (001). Angles: Prism angle 110 A 1T0=95° (85° normals) ; prism to base 110 A 001 =90°. Habit tabular, elongated parallel to the macro-axis, the prism faces and the macropinacoid often in equilib- rium (text figures 103 and 104). Pleochroism was difl&cult to observe, owing to the high color of the plasma, but evidently the color is deeper parallel to the 6-axis (c). Extinction is symmetrical on the fiat views of the plates and straight on edge views. Orientation of the elasticity axes is a=* 112 Figs. 108, 109, 110, 111, 112, 113. Eguut caballua /3-Oxyhemoglobm. common gypsum twin on the orthopinacoid; and even becoming elongated along the common edge, but not to the same extent as in the case when the base is the composition face. Another hemitrope twin of this type is possible in which the twin axis is the com- mon prism-base edge and the composition face is the base or the normal to the base. This also appears to occur, but not so commonly. There is a third kind of twinning that was occasionally seen, of the Manebach type (text figure 112), when the base is the com- position face and twinning plane, or the normal to the base is the twin axis. From the way in which these twins develop it would seem better to assume a twin axis parallel to the edge 010-001, and then the twinning plane would be the plane normal to the base in the zone of (lOO)-(OOl). This plane actually appears to occur as a composition face in these twins. In the first type of twinning (horse-type) especially, when the base is the composition plane, they twin several times on the different prism edges so that three or more may occur in a group (text figure 113), in partial polysynthetic order. In some cases this produces a six-pointed star with three individuals, or more often with four. These monoclinic crystals are produced in great numbers in the blood to which oxalate has been added, and they increase in proportion to the amount of oxalate added, and also in inverse proportion to the number of a-oxyhemoglobin crystals. But they are produced in blood to which no oxalate has been added, although in comparatively small numbers. The oxalate does not alter the habit, form, or other characteristics of the crystals at all. They form evidently from concentrated or dense solutions of the hemoglobin, and the function of the oxalate is perhaps to increase the pressure of the solution or to make it more concentrated by taking some of the water. Slow evaporation of non-oxalated blood has the same effect of making the solution more dense. It is also possible that the oxalate helps to convert one isomer into another. CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE UNGULATES. 193 Pleochroism is strong; a pale yellowish-red, b rather bright red, c deep blood-red. Extinction is symmetrical on the base, but obHque on the clinopinacoid sections, as well as on all prism faces. The orientation of the elasticity axes is a A a = 13°, in the obtuse angle; 6=5; c A i=5°, in the obtuse angle; extinction on the clinopinacoid section is hence 13° from the trace of the base, and 5° from the prism edge. The plane of the optic axes is the clinopinacoid; and on the basal section one brush of the interference figure is seen, somewhat out of the center of the field, revolving as the crystal is revolved. The other brush is out of the field. The axial angle 2E is evidently more than 50°, per- haps about 60°. The acute bisectrix is the axis of least elasticity Bxa=i, and the crystal is optically positive. CARBON-MONOXIDE HEMOGLOBIN OF HORSE. The CO-hemoglobin was made by exposing the blood to illuminating gas (water-gas) for several hours, and in laked bloods crystals form during the time of exposure to the gas at the ordinary room temperature. The solution, from which the crystals have been separated by centrifugalizing, crystallizes very rapidly; too rapidly, in fact, when making slide prepara- tions in the usual manner. In order to make the crystallization less rapid the blood was diluted with a 50 per cent solution of egg-white and from this the best crystals were obtained. As in the case of the oxyhemoglobin, the addition of oxalate causes the development of the monoclinic crystal, while in absence of oxalate mainly the orthorhombic crystals are formed; but the addition of egg-white only retards the formation of the crystals, and makes them finer in development, without in any way altering their characters. The heat of the hand was found to be sufficient to dissolve the crystals that had formed in the flask, during exposure to CO. To produce the orthorhombic crystals, the laked blood, after exposure for some hours to illuminating gas, was simply warmed by the hand, or by immersing the tube in water at body temperature, and, after centrifugalizing a few minutes, the preparations made. Only the orthorhombic crystals developed at first. The best crystals were made by mixing one part of the CO-blood with one part of egg-white, shaking with excess of ether and centrifugalizing a few minutes. By addition of oxalate to either the undiluted CO-blood, or the CO- blood diluted with one part of a 50 per cent solution of egg-white, and warmed in each case to body temperature, the monoclinic form only developed. The crystals were of the same habit in preparations from the undiluted blood and from the diluted blood. All of these CO-hemoglobin preparations were made from fresh blood that had been ether-laked and centrifugahzed, so that it was clear before exposure to the water-gas. There are hence two forms of the CO-hemoglobin, as was the case with the oxyhemoglobin. These have been distinguished as a-CO-hemoglobin and /?-CO-hemoglobin. a-CO-hemoglobin ofEquus caballus. Orthorhombic: Axial ratio a :b :i = 0.7332 : 1 : 0.4106. Forms observed: Prism (110), macrodome (101). Angles: Prism angle 110 A ITO =72° 30' (normals) ; dome angle 101 A TOl =58° 30' (normals). Habit long or short prismatic (text figures 114 and 115), prisms from 2 to 10 times as long as they are thick, and terminated by the macrodome, with generally equal de- 13 194 CRYSTALLOGRAPHY OF HEMOGLOBINS OP THE UNGULATES. velopment of the dome faces; the prisms are often flattened on two opposite faces, but usually symmetrically developed. They frequently grow in radiating groups and in some cases appear to be twinned on a pyramid of the unit series (text figure 116) but the angle of this pyramid was not determined. Parallel growths of two crystals, side by side with a common dome edge, are frequently seen. Some of the prismatic crystals are very long, the ratio of length to thickness being 20 : 1 or more. 114 Figs. 114, 115, 116. Egima cabdUua o-Carbon-monoxide Eemoglobin. Pleochroism is not so strong as in the oxyhemoglobin, but a is pale rose-pink and i and c nearly equal and rose-red. The spectrum was that of CO-hemoglobin as deter- mined by the microspectroscope. The orientation of the elasticity axes is a =6, 6=6, c=o; the plane of the optic axes is the brachypinacoid; the optical character, judging from the pleochroism, is negative, and the acute bisectrix, Bxa=a. P-CO-hemogloMn of Equus caballus. Monoclinic: Axial ratio a : b : (5 = 1.664 : 1 : 6; /3=68°. Forms: Unit prism (110), positive hemiorthodome (lOl), base (001). Angles: Prism angle, traces of the prism on the base, edges 110-001 A 110-001 = 62°; angle of the hemiorthodome on the base not obtained; prism edge to base, edge 110-110 A 001=68°. Fios. 117, 118, 119. Eg\m» cdhdttut ^-Carbon-monoxide Hemoglobin, Habit thin tabular on the base, consisting of the base cut by the unit prism (110) (text figures 117 and 118) and in some cases by the hemiorthodome (101) also (text figure 119), often this form appearing on one end of the plate only. This hemiorthodome was also observed in some of the twins. Twinning is normal in these crystals; an un- twinned crystal is exceptional. The usual twin (horse-type) is on the normal to a common prism-base edge as twin axis, the normal lying in the base and the plane of twinning, hence a plane including the prism-base edge and normal to the base as already described under ^-oxyhemoglobin of the horse. The majority of these crystals consist of at least CRYSTALLOGRAPHY OP HEMOGLOBINS OF THE UNGULATES. 195 three individuals; often more, up to six. They frequently form complicated groups. The twinned crystals are generally elongated in the direction of the common edge, as was the case with the /?-oxyhemoglobin. Pleochroism rather marked; a pale pink, B deep rose-pink, c very deep rose-red. Extinction is symmetrical on the base and straight on edge, looking along the clino-axis on edge, but looking along the ortho-axis, the extinction angle is about 15° from a. The orientation of the elasticity axes is a A a = 15°, in the obtuse angle; b=b; c A 6=7°, in the obtuse angle. The plane of the optic axes is the chnopinacoid and on the base in convergent Hght a biaxial figure is seen, with rather widely separated brushes, showing that the acute bisectrix Bxa = c, and the optical character is positive. The twinning pro- duces apparent optical anomalies; the twins, consisting of three, show a nearly uniaxial figure, and, even with two, the twin sometimes shows two symmetrically placed brushes as though orthorhombic. In the more complicated groups the apparently uniaxial figure is normal, but in all of these the interference cross opens slightly upon revolution of the crystals. Mule, Equus adnus (male) X Equus cahallus (female). Plates 27-29. The blood was obtained fresh and was not oxalated when collected, but defibrinated by beating. Centrifugalized corpuscles were laked and centrifugalized again, preparations being made both with and without oxalate. As with horse, the crystals are dimorphous; and the non-oxa- lated blood crystallized principally in the orthorhombic system, while the oxalated blood crystallizes principally in the monoclinic system; but in each, oxalated and unoxalated, both kinds of crystals appeared in the slides. For example, the corpuscles were laked with a large excess of ether and centrifugalized for a few minutes; preparations from this treatment showed only the orthorhombic crystals at first, but inside of 20 hours the monoclinic plates had developed sparingly, a few to the slide, in large, well- formed crystals. In the same way, the orthorhombic prisms appeared only sparingly in the preparations with a large amount of oxalate. The crystals were all oxyhemoglobin, as determined by the microspectroscope. The two forms are distinguished as a-oxyhemoglobin and |S-oxyhemoglobin. a-Oxy hemoglobin of Mule. Orthorhombic: Axial ratio 0.7813 : 1 : 0.4198. Forms observed: Unit prism (110), macrodome (101). Angles: Prism angle 110 A 1T0=76° (normals); dome angle 101 A 101=56° 30' (normals) . Habit prismatic, either long or short, elongated along (110) (text figures 120, 121, and 122), the ratio of the length of the prism to its thickness being about 2 : 1 in the short prisms (text figure 120) and vary- ing up to 20 : 1 in long prisms (text figure 121); some are even almost hair-like. The prisms grow in irregular groups in the protein ring and along the cover edge, and are also commonly found scattered singly through the slides. The macrodome is usually un- equally developed, one face larger than the other, giving a rather monoclinic aspect to the crystals (text figure 123). The prism faces are also sometimes unequally developed, the prism being flattened on two opposite faces. Twinning was not definitely made out. 121 fv^ 122 123 Fios. 120, 121, 122, 123. Mule a-Oxyhemoglobm. 196 CRYSTALLOGRAPHY OP HEMOGLOBINS OF THE UNGULATES. Pleochroism is readily observed; a pale pinkish, 6 rather moderately deep red, c deep red, i and c being nearly equal. Double refraction is not very strong; extinction is straight in all aspects. The orientation of the elasticity axes is a = b > a. From the double refraction and the pleochroism, as well as from the absorption, which show c and b to be nearly equal, it would seem that the acute bisectrix Bxa=a, and the optical character is negative. Reduced Hemoglobin of Sus scrofa, Domesticated Variety. In the slides after standing for 24 hours there always developed numerous long prismatic crystals of reduced hemoglobin, which appear at first around the margin of the cover, and later throughout the body of the slides. They show straight extinction on most aspects, but have a decidedly monoclinic habit, being terminated obliquely in many cases. Some appeared to have square ends; others, a single plane like a basal pinacoid, but oblique. They appear to be monoclinic. They grow in tufts and in sheaf-like aggre- gates, sometimes even in feathery groups. They appear to twin on a dome or pyramid, and also on the prism. The terminal plane is usually very imperfect, due to a fibrous character which the crystals show, the ends of the fibers making a rough plane. Smaller crystals and short stout prisms show a very monoclinic aspect. Muis Deek or Chevhotain, Tragvlus meminna. Plate 34. The specimen was obtained from the post mortem of an animal that died in the Philadelphia Zoological Gardens. The blood was oxalated, ether-laked, and centrifugalized; the shdes were prepared in the usual manner. Crystals formed readily and did not show a tendency to dissolve on bringing them into a warm room. They were oxyhemoglobin. Later, the same sUdes developed crystals of reduced hemoglobin, along with those of the oxyhemoglobin; these latter being relatively enormous. Both kinds of crystals were very sharp and well defined. CRYSTALLOGRAPHY OF HEMOGLOBINS OP THE UNGULATES. 201 Oxyhemoglobin of Tragulus meminna. Monoclinic: Axial ratio a : b : 6 =-1.804 : I : 6; /?=63<'. Forms observed: Unit prism (110), base (001). Angles: Traces of prism on the base, edges 110-001 A 1T0-001=59°; true angle 110 A ITO = 64° 50' (calculated); prism edge to base, edge 110-lTO A 001 = 63° (normals) =^. Habit of the single crystals tabular on the base (text figures 142, 143), the plate bounded by the prism faces, generally symmetrical or nearly so; but most of the crystals are twinned with the prism-base edge (110-001) as twin edge and a normal to this edge in the plane of the base as the twin axis (text figures 144 and 145) . In these twins the composition face is the base and along one of the prism-base edges, where they unite, there is a reentrant angle, while on the opposite edge there is an ordinary dihedral angle. In these twins (horse-type), which are common in all hemoglobins with angles that approximate 60°, the compound crystal in this species is usually elongated along the common edge, and the two crystals overlap each other at the ends of this elongated crystal forming reentrant angles in the outlines of these ends. The twinning is frequently repeated in polysynthetic order; and it is often complicated by parallel growth in one or more of the members of the twin. It does not appear to tend to produce hexagonal forms by twinning on more than one pair of the prism-base edges, however, as is com- monly the case in this kind of twinning. Twinning on the base as twin plane is also found apparently, but it is rare. This twinning seems to tend to make the angle of the plate nearer 60°. In some cases the opposite prism-base edges do not appear to be par- allel, due perhaps to a vicinal prism face in one member of the twin; this non-parallelism would tend to average the angles to near 60°- FiGS. 142, 143, 144, 145. TraeiUxu meminna Oxyhemoglobin. Fiss. 146, 147. Tragulut meminna Reduced Hemoglobin. Pleochroism is strong; a is pale yellowish with a reddish tinge, 6 is a blood-red and c is still deeper red than B. Extinction seems to be symmetrical with the sides of the plate; in some cases it appeared a little oblique, but probably the plates were some- what tilted. In twins with symmetrical extinction the angle was about 10° from the prism-base edge. Looking along the symmetry axis, it was about 15° from the trace of the base. The orientation of the elasticity axes is a A a = 15°, in the obtuse angle, b=b, c A 6 = 12° (about), in the obtuse angle. The axial plane lies in the plane of symmetry; the acute bisectrix of the optic axes is c, hence the optical character is positive. Looking at the crystal normal to the base, in convergent light, the biaxial figure is seen, with one brush constantly in the field, the other passing out on rotating the crystal. The axial angle is hence large. The oxyhemoglobin crystals changed to metoxyhemoglobin by paramorphism, without apparent change of form, and they retained their optical orientation and all optical characters, except the pleochroism, which no longer showed any blood-reds, but dull brownish-reds. Otherwise they resembled the normal oxyhemoglobin. This change took place after the reduced-hemoglobin crystals began to appear. 202 CRYSTALLOGRAPHY OP HEMOGLOBINS OP THE UNGULATES. Reduced Hemoglobin of Tragulus meminna. Orthorhombic: Axial ratio a : b : 6 =0.5205 : 1 : 6. Forms observed: Unit prism (110), base (001). Angles: Prism angle 110 A 1T0=55° (normals); base to prism (001) A 110=90°. Habit tabular on the base, the combination of the prism and base making the rhombic plate (text figures 146 and 147). The orientation of the axes is changed from that in the oxyhemoglobin, the symmetry axis of the monoclinic crystal becoming the macro-axis of the reduced hemoglobin crystal, so that to compare them with the oxy- hemoglobin crystals, the position of these axes must be reversed. Making such reversal the axial ratio of the hemoglobin would be a : 6 =1.9209 : 1 as against 1.804 : 1 in the oxyhemoglobin; but this difference would be still greater if the true cross-section of the oxyhemoglobin prism were taken. The difference is not only seen in the prism angles, however; the reduced hemoglobin crystals do not show the twinning so characteristic of the oxyhemoglobin. Pleochroism is very strong; a is colorless to pale pinkish; B is deep rose-pink; c is deep ruby-red. The extinction is symmetrical on the base and straight on edge views. Looking along the b axis, in convergent light, the biaxial interference figure is seen; the brushes are rather widely separated. The orientation of the elasticity axes is a=b, b=o, c=i, analogous to the arrangement of the elasticity axes in the monoclinic oxy- hemoglobin. The plane of the optic axes is the macropinacoid, the acute bisectrix BXa= o :b-= 0.5205 :l,/3 -90° Positive, Bxa=c Negative, Bxa=o Elk OB Wapiti, Cervus canadensis. Plates 34 and 35. Two specimens were examined, one probably from the Philadelphia Zoological Gardens and the other from the National Zoological Park at Washington. The first specimen was putrid; the last was in better condi- tion. Both gave crystals of oxyhemoglobin. The putrid blood was pre- pared without the use of ether, which is likely to produce precipitation in such blood; it was simply repeatedly frozen and thawed, until the cor- puscles were broken down, and then centrifugaUzed. The crystals formed readily after the slides were covered, and were large enough to photograph inside of a few hours. Oxyhemoglobin of Cervus canadensis. Tetragonal: Axial ratio a : 6 =1 : 0.7133. Forms observed: Unit pyramid (111), diametral prism (100). Angles: 110 A 1T0=90° from outlines of the pyramid in plane; 101 A 101 = 109° (or 71° normals) from outlines of pyramid in elevation; the other elevation looking along the diagonal axis gave about 56° (normals) (55° 15' by calculation about). This angle as ordinarily presented appears to be somewhat higher, up to 60° or more. CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE UNGULATES. 203 Habit bipyramidal (text figure 148); also combinations of the diametral prism with the unit pyramid (text figure 149) ; these latter occur more rarely. The crystals were usually single and isolated; but Lq some cases formed irregular linear aggregates, straight or curved, some being evidently parallel growths. Twinning on the unit pyra- mid, as the plane of twinning, was observed, but was not common; the twins are pene- tration twins, but otherwise resemble the spinel twins somewhat (text figure 150). FiOB. 148, 149, 150. Cerma canadmtit Ozyhemoglobin. The color was oxyhemoglobin red and the pleochroism is not very marked; absorp- tion for CO greater than for e. The double refraction is not very strong, showing better in the prismatic crystals; extinction is parallel to the vertical axis 6, which is the axis of greater elasticity. On the square sections, looking along the polar axis, the crystals are singly refracting and do not polarize; and in convergent light they show a uniaxial figure in the shape of a dusky cross. The optical character is negative, cj > e. Red-backed Deeb (Probably the Red Brocket, Cariacus rufus). Plate 35. The specimen of blood was received from the Philadelphia Zoological Gardens, and was beginning to putrefy. It was treated by oxalating and freezing, then laked with ether and centrifugalized. The slides were pre- pared in the usual manner and kept at a temperature near the freezing-point. It crystallized readily, the crystals that were the first to form being long lath-shaped rods; these were followed by large rectangular plates, very thin, and evidently the same as the rods, but of a tabular habit. When the sUdes were brought into a warm room the plates were rapidly dissolved; the rods showed more resistance to solution, but it was found necessary to examine the slides in the cold, and all photographs were taken at a temperature near the freezing-point. The crystals gave the spectrum of reduced hemoglobin. Reduced Hemoglobin of Cariacus rufus. Monoclinic: Axial ratio could not be determined, as only pinacoids were seen. Angle p appears to be about 90°. Forms observed: Base (001), clinopinacoid (010), orthopinacoid (100). Angles: Clinopinacoid to orthopinacoid, the outUne of the plates, 010 A 100=90°; base to orthopinacoid, angle /9, about 90°, perhaps exactly 90°. The third angle was not observed, but must be 90° in this system. Habit lath-shaped crystals elongated on the clino-axis, and flattened on the base; these, by elongation on the ortho-axis also, became rectangular plates, which are fre- quently as long as the lath-shaped rods and perhaps 30 times as wide. The plates appear to be produced by the piling up of narrow plates all in parallel position and this produces striation on most of these plates parallel to the a axis. The rods grow in tufts, radiating or brush-like, generally united with each other on the base, or in the zone of (OOl)-(lOO). The composite plates are produced in the same way, by uniting on the base. No definite twinning was observed, but possibly the piled-up plates may be polysynthetic twins. 204 CRYSTALLOGRAi'HY OF HEMOGLOBINS OF THE UNGULATES. Pleochroism was strong, as is common in reduced hemoglobin; a pale pink, nearly- colorless; i purplish, deeper than a; c deep purplish-red. Absorption was in the order c > b > a. Extinction on the base was straight; on edge views the extinction angle was about 30° with the length of the rod-like section. The plane of the optic axes is the plane of symmetry; the orientation of the elasticity axes is a A a =30°, 6=6, c A 6=30°. On the base, in convergent polarized light, a single brush of a biaxial interference figure is seen, the optic axis emerging at a small angle with c or the normal to the plate ; the acute bisectrix is hence evidently c, and the optical character is positive. Venezuela Deeh, Mazama americana savannarum (?). Plate 36. The specimen of blood was received from the National Zoological Park at Washington during the summer and was kept frozen until examined. The quantity of blood was small, and it was quite thick and putrid, and full of extraneous matter. This latter was centrifugalized off as far as possible, but the specimen was not thoroughly cleansed, owing to an accident to the centrifugal machine. The sUdes were prepared in the usual manner, and crystals formed readily in the cold. They were not dissolved at a tempera- ture of 10° C. A spectroscopic examination of the plasma showed the presence of oxyhemoglobin, but only crystals of reduced hemoglobin were obtained, they being determined as such by the microspectroscope. Reduced Hemoglobin of Mazama ammcana savannarum. Monoclinic: Axial ratio not determinable as the pinacoids only are developed. The angle j3 seems to be 90°. Forms observed: Base (001), cUnopinacoid (010), orthopinacoid (100). Angles: Clinopinacoid to orthopinacoid, the outUne of the plates 010 A 100=90°; base to orthopinacoid 001 A 100=90°=/?. The third angle could not be obtained, but is necessarily 90°. Habit broad or narrow lath-shaped, flattened on the base and elongated parallel to the cUno-axis (text figure 151) ; the lath-shaped crystals by development along the sym- metry axis b become broad plates. When the plate-like habit is assumed, the tabular crystals are seen to be composite, by par- allel growth and uniting on the base, producing strong striation parallel to the clino-axis. The crystals grow in tufts, radiat- ing from a center, and the majority of the crystals are broad '^¥ 151 Mamma americana lath-shaped or tabular, with the length of the plate 2 to 3 times "^'mvannarwn Reduced the width. On edge vicw they do uot show the usual tendency Hemoglobin. ^^ radiate in a brush-like manner to any very marked degree. The color is reduced hemoglobin purple; the pleochroism, as usual in hemoglobin, is very strong; a is pale rose-pink; 6 is strong rose-pink; c is deep rose-red. On the flat the extinction is straight, parallel to the edges of the plate or lath-shaped crystals; on edge it is oblique, about 30° measured from the length of the rod. The plane of the optic axes is the plane of symmetry; and the orientation of the elasticity axis is o a a = 30°, the extinction angle; h=b; c A e and the optical character is negative. Reduced Hemoglobin of Cervus dama. Probably orthorhombic, perhaps monoclinic; the crystals were very imperfect. They showed sometimes a roughly four-sided cross-section, with an angle of perhaps 85°, but most of them seemed to be rather lath-shaped. They were generally not ter- minated; a few showed square-cut ends, but the majority were merely shred-like masses, with more or less straight sides and splintery looking ends. They were also in spheru- litic masses and are probably parallel or radiating groups of smaller prisms; the parallel groups forming the prism-like shreddy crystals and the radiating groups the spherulites and tufts of crystals. The parallel masses show straight extinction, the spherulites extinguish parallel to the fibers and show the usual extinction cross of spherulitic masses of crystals in polarized light. The straight groups of crystals show the length of the prisms to be the direction of greater elasticity and the direction normal to this to be the 206 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE UNGULATES. direction of less elasticity. Pleochroism is strong, parallel to the length of the fibers; a is colorless, normal to the length, b and c are purple. In the spherulitic masses the pleochroism shows strongly, dividing the spherulite into sectors, the opposite sectors being of the same color; a colorless, c and 6 purple, as in the prisms. MuNTjAK, Cervulus muntjak. Plates 36 and 37. The specimen of blood was received from the New York Zoological Gardens and was in a somewhat putrid condition. The blood, containing oxalate, was laked with ether and centrifugalized and from the clear solu- tion slides were prepared in the usual manner. The blood crystallized slowly and the crystallization was therefore carried on at temperatures near 0° C. The first crystals to form were oxyhemoglobin, short prisms with very oblique terminations; later, crystals of reduced hemoglobin formed in the shape of long square-ended rods or lath-shaped crystals, and in curving arborescent forms. The crystals of the oxyhemoglobin with- stood changes of temperature fairly well, but the crystals of reduced hemo- globin were rapidly dissolved when brought into the warm room. Oxyhemoglobin of Cervulus muntjak. Monoclinic: Axial ratio a : b : 6 =1.303 : 1 : 6; /3=52°. Forms observed: Unit prism (110), base (001). Angles: Traces of the prism angle on the base, or plane angle of the basal section, edges 110-001 A lTO-001 =75° (105° normals), base to prism edge=^=52°, actual prism angle 110 A 1T0=91° 30' (calcu- lated 91° 45'). Habit short prismatic, the unit prism terminated by the oblique basal pinacoid; sometimes in equilibrium (text figure 153), and then looking rather like a rhombohedron. The crystals grow isolated or Fig. 153. Cenmitu munt- crowdcd together in great numbers, and sometimes appear to twin on jofc Oxyhemoglobin. -j tj-i. -I ■ i. i- • xif j a pyramid of the unit series or perhaps on a hemiorthodome. Pleochroism is rather marked; a nearly colorless, somewhat yellowish; 6 strong red, c deep red. On all sections in the zone of (OOl)-(lOO) the extinction is symmetrical; looking along the symmetry axis b the extinction angle is 30° with the prism edge. The plane of the optic axes is the plane of symmetry, and the orientation of the elasticity axes is as follows: a A a = 8°, in the obtuse angle; b=b; c A (5=30°, in the obtuse angle. In convergent light, looking nearly along c, the biaxial interference figure was seen with the brushes rather widely separated. The optical character is hence positive, as the acute bisectrix is c. Reduced Hemoglobin of Cervulus muntjak. Monoclinic: Axial ratio can not be determined, as only the pinacoids are developed. The angle P seems to be 90°. Forms observed; Base (001), orthopinacoid (100), clino- pinaCOid (010). .,„,,„, Fm. 154. Ceroulu. muntfot Reduced Angles of orthopinacoids and clmopinacoids, 90°; base Hemoglobin, to orthopinacoid about 90°- Habit lath-shaped, flattened on (001) and elongated along a (text figure 154), also hair-like, in feathery and arborescent groups and tufts; when broad lath-shaped, the crystals compound on the base. The crystals melt or dissolve very rapidly on being brought into a warm room. They are very much larger than the oxyhemoglobin crystals. Pleochroism is rather marked; a very pale bluish-lilac; b purplish, pale but stronger than a; c deep reddish-purple. Extinction is straight on the flat and about 18° with the CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE UNGULATES. 207 length on edge. The axial plane lies in the plane of symmetry and the orientation of the elasticity axes is a A a = 18°, h=b,c A 6 = 18°. The interference figure was not observed, but from the character of the pleochroism it may be judged that the crystals are optically positive. Indian Antelope, Antilope cervicapra. Plate 37. The specimen was received from the Philadelphia Zoological Gardens. The blood was partly in clots and somewhat putrid. It was ground with sand and ether, and centrifugalized. After it had cleared, some oxalate was added and the solution again centrifugalized. From the clear solution, the slide preparations were made in the usual manner. Crystals formed readily in the dried protein ring and along the edge of the cover. The photographs were made on the following day. Examination with the microspectroscope showed that the crystals were oxyhemoglobin. They appeared quite insoluble, showing no tendency to dissolve on bringing the shdes into a warm room. ^ a/1 155 156 Flos. 165, 156. Antilope cer- vicapra Oxyhemoglobin. Oxyhemoglobin of Antilope cervicapra. £° /^ Monoclinic: Axial ratio a : b : 6 =1.887 : 1 : 6; /?=71° 45'. Forms observed: Unit prism (110), base (001). Angles: Prism angle on cross-section of prism 110 A 1T0 = 58° 20'; angle of prism traces on the base, edge 110-001 A ITO- 001 = 56° (about); angle of prism edge to base 110-110 A 001 = 71° 45' (normals) =i3. Habit prismatic, elongated on the vertical axis, the acute prism obliquely terminated by the base (text figure 155) . In many crystals the prism appears to be flattened on two of the opposite faces, becoming lath-shaped. The faces of the prism are vertically striated, and there is an appearance of cleavage, parallel to the prism faces. The crystals are large, and the length is very great in proportion to the width. When seen in the aspect looking along a, the termination of the prism frequently appears to be square. The crystals grow in groups and slightly divergent tufts. Twins on the prism face as the twin plane were noted; they were rare (text figure 156). They seemed to be usually juxtaposed twins, but a few interpenetrant twins were seen. The color is oxyhemoglobin red, but the pleochroism is strong and the crystals look light or dark according as the aspect presented is normal to a or parallel to a. The colors are: a pale pinkish, b red, c deep red. Extinction seems to be parallel to the prism edges in all positions of the prism normal to the 6 axis, and symmetrical on the cross-sections. The orientation of the elasticity axes is a a a = 18° 15', in the obtuse angle; b—b; c=6; the plane of the optic axes is hence the plane of symmetry and the angle between c and 6 is 0°. Looking along a, traces of the interference brushes are seen, but pass far out of the field; this would seem to indicate that the axis of least elasticity is the acute bisectrix, Bxa—C, and the optical character is positive. From the straight extinction, it looks as though this might be an orthorhombic crystal with one pair of macrodome faces developed; but such symmetry as the crystals show is plainly monoclinic. Redttnca Antelope, Nagor, Cervicapra redunca. Plate 38. The blood was from a specimen of the nagor that died in the New York Zoological Gardens, and was received from New York in good condi- tion. It was oxalated, laked with ether, and centrifugalized; the slide preparations were made in the usual manner. The crystals formed readily, 208 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE UNGULATES. rather faster than in horse blood for example, and the first crystals to form were rather long rods. These formed at a temperature of about 24° C. ; on placing the sUdes in the cold, a second crop of larger and stouter crystals formed, which appear to be more soluble than those of the first crop, dis- solving readily when brought into a temperature much above 12° C. When placed in the cold, the corroded edges and angles that were produced by the increase of temperature are soon repaired. The crystals of the first crop do not seem to be affected by such temperature change as would cause solution of the second crop. Examination with the spectroscope failed to show any difference in the spectra of these two t3^es of crystals; and they have identical axial ratios, although the habit is quite different. After some days the terminations of the crystals of the first crop became imperfect or even dissolved; but this occurred while they were kept in the cold. The crystals of the second crop seemed to keep very well, so long as the temperature was kept below 10° C. ; at a few degrees above this tempera^ ture they began to dissolve. About two weeks after the slides were pre- pared they showed mainly very large crystals of prismatic habit, somewhat similar to the first crop, but with the ends corroded. From cross-sections these were evidently unit prisms, of the character of those formed in the first crop of crystals. Oxyhemoglobin of Cervicapra redunca. Orthorhombic: Axial ratio a : b : 6 =0.839 : 1 : 0.5877. Forms observed: Unit prism (110), brachypinacoid (010), macropinacoid (100) (?), macrodome (101), brachydome (032), unit pyramid (111). Angles: Prism angle 110 A 1T0=80° (normals); macrodome 101aT01=69° (normals); brachydome 032 A 032=43° (normals); unit pyramid edges over the pole or unit brachydome Oil A Oil =54°. The value for 6, calculated from the macrodome (first crop), was 0.5870 and from the pole edge angle of the unit pyramid (second crop) was 0.5877, which are substantially identical. Habit of the first crop crystals (text figure 157) is long prismatic, consisting of the unit prism (110) in equilibrium with the brachypinacoid (010), and termi- nated by the unit macrodome (101) ; the second crop crystals are short prismatic, becoming tabular by de- velopment of the brachypinacoid as they become larger, and consisting of the unit prism (110) and the brachy- pinacoid (010) in the prismatic zone, terminated by the brachydome (032) (text figure 158); or in some cases by the unit pyramid (111) (text figure 159). In some crystals the macropinacoid seemed to be devel- oped, probably by pressure of the cover. The prismatic crystals, found in the slides that had been standing in the cold for two weeks, were of the first type; but the brachypinacoid was much reduced in size and the terminations were wanting. Small crystals developed a few days after the slides were prepared, which showed the same habit, and, even with the brachypinacoid entirely absent, these were terminated by the unit macrodome. The first crop of crystals grew from the edges of the cover and from the protein ring in irregular tufts, also in radiating stellate groups through the body of the slide; and the mass of the crystals into which the protein ring was converted were of this type. The second-crop crystals appeared near the edge of the cover, singly or in roughly fs^ A 157 d:\ \i^» ^=t^ 159 Figs. 157, 158, 159. Cervicapra redunca Oxy- hemoglobin. CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE UNGULATES. 209 parallel grouping with the vertical axis nearly normal to the cover edge. The relation of the first and second crops seems to be, that from the first supersaturated solution the first crop crystals developed, until the solution was in equilibrium for that temperature; and when the shdes were placed in the cold a second crop developed gradually, probably under increased pressure as the balsam seal of the slides hardened, until equilibrium had been reached in the saturation of the solution for the lower temperature of about 10° or less. Of course, on bringing the slides into the warmth of a heated room, this equilibrium would be disturbed and resolution of the last-formed crystals would take place. As will be seen from reference to the data in regard to the position of the elasticity axes, the relative shortening of the crystal axes in the second-crop crystals is in the inverse order of their elasticities; a, the axis of greatest elasticity, shortening more than 6, the axis of mean elasticity; while c, the axis of least elasticity, lengthens with respect to the other two. The color of the crystals was the usual oxyhemoglobin red. Pleochroism is rather strong; a is pale yellowish-red; b is a pale red, somewhat rose-pink; c is deep blood-red. Absorption is in order, c > 6 > a. Extinction is straight or symmetrical in all aspects. The plane of the optic axes is the basal pinacoid and the orientation of the elasticity axes is a=b, 6=tS, c=a. On side views of the prism or on the vertical pinacoids traces of an interference figure are seen in convergent Ught; on (010) the two brushes of the biaxial figure are seen in the field, but they are widely separated. On (100) two brushes show also, but pass out of the field in the diagonal position. The acute bisectrix is hence evidently a, and the crystals are optically negative. This is indicated also by the pleo- chroism, for 6 and c are much nearer together than 6 and a. Dorcas Gazelle, Gazella dorcas. Plate 39. This specimen was received from the Philadelphia Zoological Gardens. The blood was clotted, but in good condition. The clot was ground in sand with ether and the mixture centrifugalized; from the solution the slide preparations were made in the usual manner. Crystals formed readily in the dried protein ring, but were gradually dissolved as equilibrium was established in the solution after covering; and then they reformed along the cover edge, as they were dissolved from the protein ring, until the solution in the slides was homogeneous. When the condition of equilib- rium was reached in the solution, the crystals showed no sign of being dissolved and were in good condition for days. The first crystals to form are small rectangular plates, but with them are long rods; both seem to be the same, however, and both are oxyhemoglobin, as determined by the spectroscope. Oxyhemoglobin of Gazella dorcas. Orthorhombic: Axial ratio a : b : 6 =0.3639 : 1 : 0.4452. Forms observed: Brachypinacoid (010), base (001), unit prism (110), brachydome (Oil), and, without measurement of angles, the macrodome (101) and the macropinacoid (100). Angles: Prism angle 110 A llO =40° (normals) ; brachydome angle Oil A OTl = 48° (normals); outline of plates, 100 A 001=90°. Habit tabular on (010), the plate bounded by the other two pinacoids (100) and (001) when the crystals begin to grow (text figure 160) or by the combination of the prism (110) and the brachydome (010) in the larger crystals (text figure 161). The macropinacoid disappears as the crystals increase in size, but the base sometimes appears to persist, although it is generally replaced by the brachydome. The crystals are usually elongated along the vertical axis, so that on the brachypinacoid aspect the length is 14 210 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE UNGULATES. about double the width; but some are much more elongated, and others are reduced to nearly square tabular plates. They frequently show parallel growth, uniting on the brachypinacoid or on the macropinacoid and also grow in irregular, somewhat radiating aggregates. Pleochroism is rather strong; a is pale yellowish-red; B is deep blood-red, c is deep red, somewhat deeper than B. Extinction is straight in all aspects. The orientation of the elasticity axes iaa=a,'b=6, t=b. The plane of the optic axes is the base, and a is the acute bisectrix, Bxa =«. Looking along this axis of a, the interference figure is seen in convergent light, with the angle between the optic axes, 2E =40°. The optical character is negative, la 1^ f^ J>\ -C Fios. 160, 161. GazeUa dorcat Oxyhemoglobin. Fis. 162. Cephdlophvt grimmi o-Oxyhemoglobin. Fio. 163. Cephalophui grimmi /3-Oxyhemoglobin. DuiCKEBBOK, Cephalophus grimmi. Plate 40. The specimen was received from the National Zoological Park at Washington. The blood was oxalated, laked with ether, and centrifu- galized, and preparations were made in the usual manner. The slides were kept at a temperature of about 0° C, and soon were filled with the small pyramidal crystals, which showed a tendency to dissolve, and were hence examined and photographed at temperatures near the freezing-point. Later, the slides developed the second kind of ciystals, the hexagonal plates. Both kinds of crystals were oxyhemoglobin. a-Oxyhemaglobin of Cephalophus grimmi. Tetragonal: Axial ratio a : 6 =1 : 0.8687. Forms observed: Unit pyramid (111), also traces of (100). Angles: Between the pyramid edges in the horizontal plane, normal to the axes = 110 A 1T0=90°; between the pyramid edges in the vertical axial plane (calculated) = 101 A 101=98°, observed angle of pyramid over the pole = 111 A Til =77°. Habit pyramidal (text figure 162), the crystals occurring singly or in irregular groups and in parallel growths. Twinning seems to occur on the pyramid as the plane of twinning. The color is the normal oxyhemoglobin red; pleochroism is scarcely noticeable. Double refraction is weak, but the extinction is symmetrical on the aspects normal to the vertical axis. Looking along this axis, the crystals are singly refracting in parallel polarized light and do not polarize; but in convergent light they show a faint dusky uniaxial cross. By the quartz wedge it is seen that the vertical axis is the direction of greater elasticity, hence w > e and the optical character is negative. CBYSTALLOGBAPHY OP HEMOGLOBINS OF THE UNGULATES. 211 ^-Oxyhemoglobin of Cephalophus grimmi. These crystals develop after the pyramids of the a-oxyhemoglobin. They are in the form of very thin hexagonal plates (text figure 163), occurring both singly and also growing in groups, often with the orientation of parallel growth. They are evidently hexagonal, the angle of the plate being 120° (60° normals) and the sides square with the terminal plane. No axial ratio could be determined, as the combination of forms is simply unit prism (lOTO) and base (0001). They are singly refracting when viewed on the base, but too thin to give an interference figure. On edge view they show very weak double refraction and extinguish parallel to the base. The optical character could not be determined owing to the very weak double refraction. Sheep, Ovis aries. Plates 40-42. The fresh blood was collected in oxalate from the abattoir, and centri- fugalized to throw down the corpuscles. The plasma was drained away, the corpuscles were laked with ether, oxalate added almost to saturation, and the solution centrifugaUzed for 2 hours. From the clear Hquid the slide preparations were made as usual. The preparation crystallized at room temperature, and the crystals showed no tendency to dissolve. Some crystals were obtained within 5 hours of making the preparations. The crystals at first formed were fine needles, but soon tabular crystals began to appear. Several other preparations were made from the same blood, and in all the crystals kept well. After about a week, crystals of reduced hemoglobin began to make their appearance, along with the crystals of oxyhemoglobin, which formed in the freshly prepared slides. These crystals of reduced hemoglobin, Uke the oxyhemoglobin, were not dissolved on slight increase of temperature. The slides were kept cool, at about 10° C, except when under examination. Both the oxyhemoglobin and the reduced hemoglobin were identified by the spectroscope. Oxyhemoglobin of Ovis aries. Monoolinic: Axial ratio a : b : 6 =1.140 : 1 : 0.970; /?=54°. Forms observed: Unit prism (110), positive hemiorthodome (lOl), base (001), clinopinacoid (010), orthopinacoid (100). Angles: Prism angle 110 A ITO, traces on the base, or angle of edges 110-001 A 110-001=82° 30' (actual angle); orthodome to orthopinacoid TOl A 100=72°; ortho- pinacoid to base 100 A 001 =54°=|l? (normals). Habit of the first crystals to form minute needles without definite outlines, tapering to a point at either end; with these soon appear tabular crystals consisting of the base with a very short prism, tabular on the base (text figure 164). After about a day, long prismatic crystals appear consisting of the three pinacoids, elongated parallel to the vertical axis and generally flattened on the orthopinacoid. These sometimes show the prism as a bevel on the edges (text figure 165), but more often are simply the three pina- coids. These crystals twin and form networks of rods, and frequently on the orthopina- coid faces twin growths develop, producing a cross-banded effect due to the strong pleochroism. Twins are hemitrope, on the orthopinacoid (100) and on the hemiortho- dome (lOT), the two occurring together and making fivelings of exactly pentagonal shape (text figure 166). These little fivelings grow on the sides of the long crystals, or singly, scattered through the slides; and they grouped themselves along the crystals of oxalate that formed in some of the slides, strung like beads along the needles of the oxalate. In this occurrence they present edge views to the observer. When seen in side view they are generally more or less perfect pentagons, divided by the contact planes into five 212 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE UNGULATES. sectors meeting at a point in the center; sometimes, by irregularity of growth, 7 individuals are seen in the group. The following more detailed description will make the arrangement plain. Suppose a fiveling, the members taken in cyclic order around from 1 to 5. Nos. 1 and 2 twin on the orthopinacoid (100) ; 3 is twinned to 2 on the hemiorthodome (lOT) and to 4 on the orthopinacoid; 5 is twinned to 4 on the hemiortho- dome which brings its orthopinacoid in parallel position with the hemiorthodome of 1, thus completing the cycle. In each member of the twin the base forms one of the sides of the pentagon. Fias. 164, 165, 166. Ovi> aria Ozyhemoglobm. Fios. 167, 168. Om» ansa Reduced Hemoglobin. These twins are often seen terminating a long prismatic crystal and, as already noted, they sprout out of the side planes of the long crystals, producing the banded appearance seen in the photographs. Pleochroism is very strong; a nearly colorless, 6 moderately strong red, c very deep blood-red. In the plates, the extinction is symmetrical on the base, and straight on the aspect looking along a; on the (010) aspect, the extinction angle is 30° with the prism edge. The long prismatic crystals show the same extinction as the plates. The orientation of the elasticity axes is a A a=6°, b=6, c A 6=30°. The axial plane is the plane of symmetry, a is probably the acute bisectrix, but the interference figure was not observed. The optical character is probably negative. Reduced Hemoglobin of Ovis aries. Orthorhombic: Axial ratio a:b: 6 = 0.7813 : 1 : <5. Forms observed: Unit prism (110), base (001). Angles: Prism angle 110 A 1T0 = 76° (normals); prism to base 110 A 001 = 99°. Habit tabular on the base, the crystals consisting of a very short prism and the basal pinacoid (text figures 167 and 168), the ratio of the length of the prism to the short diag- onal of the base being about 1 : 5. The crystals occurred singly and did not appear to form twins. They appeared rather sparingly in the slides about a week after the preparations were made and seemed to show no tendency to dissolve on slight increase of temperature. Pleochroism is strong; a nearly colorless, h deep red, c deep purplish-red; the colors of i and c are nearly equal. Extinction is straight in all positions on edge and symmet- rical on the base. The orientation of the elasticity axes ia a=b,b =6, c=a. The plane of the optic axes is the basal pinacoid and the acute bisectrix, Bxa=CL. The optical character is hence negative. BuRBELL OB Bhabal, Ovis nahura. Plates 42 and 43. The specimen was received from the National Zoological Park at Washington, and consisted of a small quantity of blood in the shape of clots, with a small amount of liquid. It contained much foreign matter in suspension. The blood was oxalated, treated with ether and centrifugalized CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE UNGULATES. 213 in the usual way, yielding a small amount of clear solution, from which slide preparations were made as usual. The blood began to crystallize in needles soon after the preparations were covered. These formed in the protein ring, but afterwards dissolved as the solution under the cover came to an equilibrimn, while new rod-like and tabular crystals appeared throughout the slides. These crystals seemed to be not very soluble, al- though rise of temperature caused a loss of the planes on the ends of the long lath-shaped crystals; and with the oxyhemoglobin crystals were mixed occasional crystals of reduced hemoglobin. The absorption spectra of both kinds of crystals were examined by the microspectroscope. The photo- graphs were made 3 days and 5 days after the preparation of the slides. Oxyhemoglobin of Ovis nahura. Monoelinic: Axial ratio a : h : 6 =1.232 : 1 : e and the optical character is slightly positive. On the basal aspect a very faint dusky cross in convergent light shows the uniaxial character of the crystals. ^-Oxyhemoglobin of Sciurus vulgaris, from the Putrid Blood. Orthorhombic: Axial ratio a : b : 6 =0.577 : 1 : 6. Forms observed: Unit prism (110), base (001). Angles: Prism angle 110 A 1T0=60°; prism to base 110 A 001=90°. 220 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. Habit very thin tabular on the base, with the very short prism determining the outline of the rhombic plate (text figures 181 and 182). The crystals rarely occur singly, however, but in twins of the usual type on an axis in the base and normal to a prism-base edge, the composition face being the basal pinacoid. These occur on any or all of the prism-base edges, making a group that is composed of many individuals and generally of a roughly hexagonal outline (text figure 183). Hexagonal plates of normal develop- ment occur with these twins, similar to the plates of a-oxyhemoglobin; and they are probably mimetic twins of the /?-oxyhemoglobin, due to the complicated groups above described becoming developed into hexagonal plates. Twins on pyramid faces, as seen in the hexagonal a-oxyhemoglobin, were not observed. The color was the usual oxyhemoglobin red, perhaps a little darker than for the corresponding thickness in the a-crystals. Pleochroism is very slight on edge and not noticeable on the basal aspect; a deep red, b=c somewhat deeper red. Double refrac- tion on the base is not noticeable, even with the quartz wedge (6 = c) ; on edge the extinc- tion is straight and the relative elasticities may be made out with the quartz wedge. On the base no interference figure of any kind could be detected in convergent light, but it is evident that the vertical axis is the axis of greatest elasticity, and that b=c; hence the acute bisectrix Bxa=o., and the optical character is negative. On comparing these two types it is evident that the characters of the j3-oxyhemo- globin are such that it would readily become hexagonal by mimetic twinning, the prism angle being exactly 60°, and the double refraction of the ^-modification is such that, but for the form of the crystal, it might be hexagonal. In the mimetic twins, produced by piling up of the rhombic plates to build a hexagonal composite plate, it might readily happen, with the very weak double refraction, that the crystal might become more dense in the direction in which the plates are piled, and hence the vertical axis, or normal to the plates, become the axis of greater density or less elasticity, when the pseudohexagonal crystal would become positive. It is hence entirely probable that the two modifications are really one and the same, the a-oxyhemoglobin being a mimetic twin of the /9-oxyhemo- globin and only pseudohexagonal. Fox-SQUiEBEL, Sdurus rufiventer negledus, Plate 46. The specimen was purchased from a collector at Orlando, Florida, and was bled in the laboratory, oxalated, ether-laked, centrifugalized, and the sUde preparations made as usual. Crystals formed rapidly in the slides, and showed no tendency to dissolve. The blood crystallizes more readily than that of the related gray squirrel. The crystals were shown to be typical oxyhemoglobin by the spectroscope. Oxyhemoglobin of Sciurus rufiventer negledus. Hexagonal: Axial ratio not determinable. Forms: Unit prism (lOTO), base (0001). Angles: Prism angle 60°, prism to base 90°. Habit, thin tabular on the base, very symmetrical hexagonal plates consisting of the short prism and the basal pinacoid (text figure 184). The crystals occur singly, or in parallel growths and piled groups on the base, the smaller crystals piled concentrically on a larger crystal. In single crystals the thickness of the plate is one-tenth to one- twentieth of the width, but this is very variable. Many single perfect plates are seen, but this varies in different slides. In some cases the plates elongated on two prism faces or along the diameter of the hexagon parallel to a crystal axis, becoming somewhat orthorhombic looking; most of them are almost perfect hexagons. Twins on a first- order pyramid occur, mostly contact twins (text figure 185) . The color of the plates is variable with the thickness, but pleochroism is very slight. On the base they are singly refracting, and polarize very faintly on edge; the double CRYSTALLOGRAPHY OF HEMOGLOBINS OP THE RODENTIA. 221 refraction is very weak. The elasticity for the ordinary ray, a>, is somewhat greater than for the extraordinary ray or e > w in refraction indices, and the optical character is hence weakly positive. Gray Squierel, Sciurus carolinensis. Plates 46 and 47. The living animal was obtained from a collector at Newport News, Virginia, and was bled in the laboratory. The oxalated blood was laked with ether and centrifugalized, and the slide preparations made in the usual manner. Crystals formed more slowly than with the other squirrels examined; they were larger and showed more tendency to produce com- posite crystals than in the other species. They showed no tendency to dissolve, however, and are evidently quite difficultly soluble in the plasma. Examination with the microspectroscope shows that these crystals are typical oxyhemoglobin. 186 Fios. 184, 185. Sciurut rufiverUer negleetus Oxyhemoglobin. Fios. 186, 187. Sciurut carolinennt Oxyhemoglobin. Oxyhemoglobin of Sciurus carolinensis. Orthorhombic; pseudohexagonal : Axial ratio a : b : 6 =0.577 : 1 : i. Forms observed: Unit prism (110), brachypinacoid (010), basal pinacoid (001); or, as pseudohexagonal, prism and base. Angles: Prism angle 110 A 1T0=60° (normals); prism to brachypinacoid 110 A 010=60° (normals), the two making a perfect hexagonal plate; prism to base 110 A 001=90°. Habit pseudohexagonal, tabular on the base, and with the prism and brachypina- coid faces in equilibrium, so that the plate is a perfect hexagon; sometimes, however, the plate is elongated on the brachy-axis, producing a distinctly orthorhombic habit (text figures 186, 187). The plates are large and perfect hexagons, but are not often simple; they produce groups by piling up on the base, more or less concentrically, and often with curving of the crystals, producing the form of the "eisen rose" of hematite, (see plate 47, fig. 277) . The parallel growths on the base may, however, start from several centers, and it is very common to see a small group of this kind near one side of a large plate, not central, but in perfect orientation with the large plate. Twinning seems to be on a brachydome. In the protein ring the crystals form spherulitic masses of the radiating plates, and when these are seen on edge, or interfered with by the cover-glass, they look like lath-shaped crystals. When the piled-up plates are seen on edge, in section, they present a sheaf-shaped appearance. The color varies much with the thickness, but in the thicker crystals it shows the normal oxyhemoglobin red. Pleochroism does not show on the flat aspect, the crystal 222 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. acting like a hexagonal crystal; but on edge the pleochroism is noticeable, the color being deeper normal to the plate than parallel to it. On the base the crystal is nearly singly refracting; but by use of the quartz wedge it is seen to be very weakly doubly refracting; on edge the double refraction is easily seen, and extinction is parallel with the base. In convergent Ught, a uniaxial cross shows in most aspects; but on revolution the brushes open slightly and the crystal is as strongly biaxial as some of the biotite micas. The orientation of the elasticity is a =6, B=o, c=6. The macropinacoid is the plane of the optic axes and the acute bisectrix Bxa = a. The optical character is hence negative. Flying-squibbel, Sciuropterus volans. Plate 47. Blood was obtained from the living animal, oxalated, ether-laked, and centrifugalized, and the slide preparations made as usual. Crystal- lization begins as soon as the blood is laked, and proceeds with great rapid- ity, so that the preparations are soon full of minute scales or tabular crystals. To retard the formation of the crystals somewhat, and permit them to grow to a larger size, some preparations were made by diluting the blood with about 3 times its volume of the blood plasma; but, even with this dilution, the crystals begin to form immediately upon laking the corpuscles. They are always small, much smaller than in other species of squirrels, but other- wise resemble those formed in the squirrel bloods in general. They were oxyhemoglobin, as determined by the microspectroscope. VI [0»| Oxyhemoglobin of Sciuropterus volans. Hexagonal. No axial ratio determinable. Forms observed: Unit prism (1010), base (0001). Angles : Prism angle 60° (normals) ; crys- tals are so thin that prism to base could K not be measured with any exactness, but it appears to be 90° Habit very thin tabular on the base (text figure 188), minute hexagonal scales or plates, with very little color, owing to their being so thin. They develop in enormous 191 numbers, the slides becoming completely filled with them. They generally occur singly or in irregular groups, but a twin on a pyramid of the second order seems to occur, interpenetrant and of the same general form as the tridymite twin (text figure 189) . The crystals are very faintly colored, when seen on the flat, but on edge have the red color of oxyhemoglobin, and show pleochroism; co deep red, e very pale red. On the flat, the crystals are singly refracting; on edge they polarize, but not strongly. The direction of the vertical axis or of the optic axis s is the direction of greater elasticity; hence w > e, and the crystals are negative. 190 FiGB. 188, 189. Sciuropterut volana Oxyhemoglobin. FiOB. 190, 191. Tamicu ttriatut Oxyhemoglobia. Gbottnd-squibeel OB Hackee, Tamias striatus. Plate 48. The specimen was purchased from a collector in eastern Pennsylvania. The animal was bled into oxalate, the blood laked with ether and centrif- ugalized, and slide preparations made in the usual manner. The blood crystallized readily at a temperature of 22° C. The crystals are quite insol- CETSTALLOGKAPHY OF HEMOGLOBINS OF THE RODENTIA. 223 uble and keep well at ordinary room temperature, showing no tendency to dissolve, even in the rays of the electric arc lamp, when making the photo- micrographs. They are the usual oxyhemoglobin red, and were determined as oxyhemoglobin by the spectroscope. Oxyhemoglobin of Tamias striatus. Orthorhombic (?): Axial ratio about a : h : 6 =0.9246 : 1 : 0.589. Forms observed: Unit prism (110), macrodome (101), brachydome (Oil). Angles: Macrodome angle 101 A T01=65° (normals); brachydome angle Oil A 011=61° (normals); prism angle (calculated) 110 A 1T0=85° 30'. The prism angle was not observed, but was calculated from the two dome angles which were measured, but not very satisfactorily; hence the uncertainty as to the exact axial ratio. Habit prismatic on the vertical axis; the first prisms that develop are very long and slender; later, stouter crystals form on which some measurements of the terminal planes can be made. The common termination is the macrodome, one face much more developed than the other, giving thei crystal a very monoclinic aspect (text figure 190). It may in fact be monoclinic, but the measurements of prism edge to macrodome seemed to be symmetrical in the crystals examined, and extinction is straight in all aspects. The prisms range in ratio of length to thickness from 15 : 1 to 100 : 1, and in most of them the terminal macrodome is unsymmetrically developed. In some a brachydome appears (text figure 191) and, some days after the slides were prepared, the two domes were seen in equilibrium, in a few cases. The crystals grow in radiating tufts from the protein ring and cover edge, and also scattered irregularly through the body of the slide; but they do not appear to form twins. Pleochroism is rather pronounced; a pale yellowish-red, 6 pale rose-pink, c deep red. The orientation of the elasticity axes is apparently a=a, 6=6, c = <5; but no inter- ference figure was made out. As stated above, the extinction is straight in all aspects of the crystals that could be examined. The optical character could not be determined, but, from the pleochroism, it should be positive. Praieie-dog, Cynomys ludovicianus. Plate 48. Specimens of prairie-dogs were purchased from collectors in Ohio and in Kansas City, and the animals were bled in the laboratory. Prep- arations were made from the corpuscles, but not from the whole blood, which probably prevented the characteristic plate-hke crystals, common in rodent blood, from developing. The corpuscles were oxalated, ether- laked, and centrifugalized and from the clear solution the sUde prepara- tions were made as usual. Only one type of crystals developed and these were not very favorable for observing the characters. They were oxy- hemoglobin. Oxyhemoglobin of Cynomys ludovicianus. Probably orthorhombic : No axial ratio determinable. Forms observed: Evidently a unit prism, but the terminations were not perfect. Angles: No angles of the crystals could be measured. Habit of the crystals obtained was long prismatic, practically hair-like, and taper- ing gradually to an acute point; but, in the larger crystals, a high power showed that they were four-sided prisms, with a lozenge-shaped cross-section; and they probably are orthorhombic, possibly tetragonal, but certainly not hexagonal. The polarization characters showed that they must be one of these three systems. The needles grow in tufts, radiating from a center, the adjacent tufts penetrating each other and forming networks of interlacing fibers. 224 CRYSTALLOGRAPHY OP HEMOGLOBINS OF THE RODENTIA. The needles in the dense tufts show the oxyhemoglobin color, but individual needles are very pale owing to their tenuity. Pleochroism is noticeable, the direction of the length of the needles showing more color than the normal to the length. The elasticity is greater normal to the length of the fiber and less parallel to the length. They are so thin that no characters can be made out in convergent light; but extinction is straight in all aspects; and this, with the four-sided cross-section, reduces the possible crystal systems to two, orthorhombic and tetragonal. The lozenge-shaped section indicates that the crystallization is orthorhombic. The blood was examined before we had devel- oped our methods of retarding crystalUzation in order to produce better crystals, and hence this blood should be further investigated. Gbound-hog ok WooDCHncK, Marmota monax. Plates 49 and 50. Specimens of this animal were purchased at different times from collectors in eastern and central Pennsylvania, and were bled in the lab- oratory. The blood was collected in oxalate. The first preparations were made by laking the oxalated corpuscles, and centrifugalizing, and from the clear solution preparing the slides as usual. As these preparations produced mainly long needles, that did not show the crystallographic characters definitely, and as the hexagonal plates that finally appeared were so im- perfect that better preparations seemed necessary, others were made, using the whole blood, the preparations being made as above described. In these preparations from the whole blood, the first crystallization in the dried protein ring is in the form of minute hexagonal plates; these soon become covered by the rapidly developing needles, and in part dissolve; so that the slides finally contain only masses of the needles. A preliminary trial of diluting with the blood plasma, and etherizing strongly before centrifugalizing, proving satisfactory in developing the plates, preparations were made by diluting the whole blood with an equal volume of the blood plasma and laking, and carrying out the preparation as above described. In this diluted blood, the plates developed readily and grew to large size, with only a slight development of the rods. The hexagonal plates kept well and passed by paramorphous change into reduced hemoglobin and also into metoxyhemoglobin. The crystals at first formed were, in all cases, oxyhemoglobin. Crystals form very readily in solutions of either the corpuscles, the whole blood, or the whole blood diluted with plasma; but much more rapidly, of course, in solutions of the corpuscles alone than in the less concentrated solutions. The development of the needles, or of the plates, can be controlled at will by the amount of dilution. The same principle appUes to other bloods that develop needles or hair-like crystals from the whole blood. Unfortunately, however, the amount of blood in the samples received was rarely enough to try the experiment, or the plasma was not in good condition owing to putrescence. In rodents in general, dilution of the blood by the plasma or serum will probably be found advantageous. Two kinds of tabular crystals were observed in the blood of the ground-hog; the one, hexagonal plates, that are probably only pseudohexagonal and mimetic twins of the second kind, which latter are in the form of rhombic plates, belonging to the monoclinic system. These two will be described as a-oxyhemoglobin and y-oxyhemoglobin, CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. 225 respectively. The rods are possibly a form of the y-oxyhemoglobin with prismatic development, but they appear to be orthorhombic, and will be called /^-oxyhemoglobin. The other forms observed, which are reduced hemoglobin and metoxyhemoglobin, were simply paramorphous alterations of the normal crystals. They appeared mainly in the ot-oxyhemoglobin form, and only in slides that had been kept for some days. orOxyhemoglobin of Marmota monax. Hexagonal or pseudohexagonal: Axial ratio not deter- minable, as no pyramidal forms were observed. ^'' ' ^ 192 Forms observed: Unit prism (lOTO), base (0001). P'°- "2. Mamota^naa; o-oxy Angles: Prism angle 60°; prism to jjase 90°. ^ ° "*' Habit thin tabular; in the whole blood preparations, the first crystals to appear are very minute hexagonal plates in the protein ring; these are later dissolved with development of the needles of /?-oxyhemoglobin. In diluted blood, the typical a-oxy- hemoglobin plates are developed; they are large, well-formed, and very regular hexagonal plates (text figure 192), occurring singly or in complicated groups in parallel growth orientation, either piled on the base (plate 49, figs. 290, 291, and 292) or in arborescent forms (plate 50, fig. 297) ; also in partial orientation, which looks complete on the base, but is seen to be partial in edge view, the plates radiating from the center of the main groups as though twinned in the zone of two opposite unit-pyramid faces (plate 59, figs. 295 and 296) . Interference with the slide and cover produces in these groups on edge broad lath-shaped individuals which look rather orthorhombic. Often a single large plate may have on its basal surface several small concentric groups, all in perfect parallel growth orientation with the main large crystal. Twins are on the unit pyramid, but owing to the tendency to produce radiating groups the angle of the pyramid could not be determined with any certainty. The color of the plates varies with the thickness, but they show rather strong pleo- chroism; w deep red, s pale reddish to colorless. On the base, the crystal is singly refract- ing; on edge, the double refraction is quite strong, and the extinction is straight parallel to the base. In convergent light, a dusky cross appears on the basal aspect, showing the uniaxial character. The vertical axis is the direction of greater elasticity, w > e, and the optical character is negative. ^-Oxyhemoglobin of Marmota monax. Orthorhombic or monoclinic: No axial ratio is determinable. Forms observed: Apparently two vertical pinacoids and a terminal dome or some- times one plane of such a dome. Angles: The crystals were not perfect enough to measure angles with exactness; the angle of the terminal dome seemed to be about 58° (normals), and the two pinacoids at right angles. Habit ordinarily hair-like, the ends tapering to a point, without any definite plane terminations; some larger crystals were lath-shaped and showed the dome or oblique termination described above (text figure 193). The crystals grew in tufts, radiating slightly; or in groups of such tufts, sometimes radiating from a center like the spokes of a wheel; along the protein ring they shoot out normal to the surface and form a con- tinuous mass of hairs on the inside of the ring; outside of it the crystals are larger and longer and the tufts more dense. The crystals are quite elastic, bending considerably before they break. They reach a large size, the individual tufts of hairs being easily seen with the unaided eye. When the crystals are lath-shaped, the flat surface of the lath is usually presented; and on this surface the pleochroism is quite marked. The length of the lath is apparently c, the width 6, and the thickness a. On this aspect above described the axes b and c 15 226 CRYSTALLOGBAPHY OF HEMOGLOBINS OF THE BODENTIA. show. The crystal is very thin so that the a direction is very short. The pleochroism on the flat is B pale yellowish-red, c pale red; or, when very thin, 6 colorless and c pale pink. On edge view the crystal shows probably from 5 to 10 times the thickness seen on the flat, and the colors are deeper; a pale pink, c deep red. On the flat aspect in con- vergent light a pair of dusky brushes of a biaxial interference figure shows; the conjugate axis is the long dimension of the lath c, but the brushes pass out of the field on rotation of the crystal. It seems probable from this that the acute bisectrix of the optic axes Bxa=t, and the optical character is positive. Calling the flat side of the lath the macro- pinacoid, the narrow edge the brachypinacoid and the dome a brachydome, the orienta- tion of the elasticity axes is a=o, 6=6, c=(i; and the axial plane is the brachypinacoid. Y-Oxyhemoglobin of Marmota monax. Monoclinic: Axial ratio a : h : 6 =1.804 : 1 : w, and the optical character was negative. The metoxyhemoglobin showed the mixed spectrum of oxyhemoglobin and methemoglobin that is often described as methemoglobin. It seemed to be the final paramorphous change, following the change to reduced hemo- globin. The metoxyhemoglobin is also quite strongly pleochroic; e is colorless or pale yellow, tt> is deep reddish-brown. The elasticity and optical character are as in the reduced hemoglobin. Beaver, Castor canadensis. Plate 51. The specimen was received from the Philadelphia Zoological Gardens in a putrid condition. The usual method of preparation was employed. Crystals formed readily after the sUdes were covered, and were at first long needle-like rods; but soon they became lath-shaped, and then plate- like crystals began to appear. They were not very stable, many crystals disintegrated and were dissolved within 24 hours after making the prep- aration. The crystals were oxyhemoglobin. Oxyhemoglobin of Castor canadensis. Monoclinic: Axial ratio a : b : 6 =1.732 : I : 6; /3=78° (about). Forms observed: Unit prism (110), base (001), orthopinacoid (100). Angles: Prism angle 110 A 1T0=60° (or very nearly); orthopinacoid to base 010 A 100=78° (about) =/3; prism edge to base, edge 110-TlO A 001=90°. Habit tabular on the base, the combination being prism and base with a greater or less development of the orthopinacoid, making generally hexagonal plates or truncated lozenge-shaped plates (text figures 197, 198). The first crystals to appear are needles; when these attain dimensions to show planes they are generally seen to be twinned, and are the twin on a twin axis in the base and normal to a prism-base edge, along which the crystal appears to be elongated (text figure 199). In some cases the twin in these prismatic crystals seems to be on a unit pyramid, interpenetrant and forming an oblique cross-shape in the cross-section (text figure 200). The plates appear when these twins 228 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. are visible; some are rhomboidal plates, apparently untwinned, but most of them seem to be twinned and in the twins the orthopinacoid planes seem to be more developed (text figure 201). By this twinning being several times repeated the crystals become nearly symmetrically hexagonal in outline, and perfect hexagonal plates appear sparingly (along with the obviously twinned crystals) that are apparently mimetic twins and really hexagonal in symmetry (text figure 202). The rhomboidal plates tend to grow into groups, by piling up of the plates (plate 51, fig. 303), and, as these are nearly all hexagonal in outline, due to development of (100), these groups closely resemble the similar forms seen in the hexagonal plates of other rodents, as the squirrels for example. 199 ^^ Fios. 197, 108, 199, 200, 201, 202. Catlor canadentit Oxyhemoglobin. The color of the crystals is a bright scarlet or blood-red. Pleochroism on the basal aspect is hardly noticeable, but probably most of the crystals examined were twinned. The colors were: a yellowish, b yellowish-red, rather a strong color; c deep blood-red. On the basal aspect the double refraction is very weak, and extinction is very hard to observe; it is, however, symmetrical. On the edge view, the double refraction is stronger, and looking along a the extinction is straight; along b it is 8° from the trace of the base or from the clino-axis, a. On the base in convergent light the interference figure is readily seen — a nearly uniaxial cross, which opens and closes as the crystal is revolved, showing the crystal to be biaxial. The angle of the optic axes, 2E is not above 7° or 8°- ~a The orientation of the elasticity axes is a A a=8°, the extinction angle; b=b, c A 6=4:° (about). The plane of the optic axes is the plane of symmetry, and the acute bisectrix is the axis of least elasticity, JSa;„ =c. The optical character is hence positive. As nearly all of the crystals examined were twinned, and as these mimetic twins tend to become uniaxial, it is possible that the above-described interference figure is due to twinning; but if so, the orientation of the optic axes is not altered nor is the optical character. MusKKAT, Fiber zibethicus. Plates 51 and 52. The living animal was procured from a collector and bled in the lab- oratory. The blood was oxalated, laked with ether and centrifugalized; from the clear solution slide preparations were made in the usual manner. The crystals formed readily soon after covering the slides; at first, the crystals were fine needles, but afterwards these became lath-shaped, or flat prismatic crystals appeared amongst the needles; and, at the same time, tabular crystals began to appear. They kept well, showing no sign of dis- solving. Crystallization continued after sealing the slides, until practically the entire slide was filled with crystals. The crystals are oxyhemoglobin. CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTLA. Oxyhemoglobin of Fiber ziiethicus. Axial ratio a:b: 6 =1.6318 : 1 : i; ^=68°. 229 Monoclinic: Forms observed: Unit prism (110), base (001). Angles: Prism angle 110 A llO =63°; prism edge to base, edge 110-lTO A 001 =68° =^. FiQB. 203, 204, 205, 206. Fiber tibethicu* Ozyhemoglobin. Habit thin tabular on the base, the crystal consisting of the base (001) bounded by the prism (110). The first crystals to form are long needles or lath-shaped crystals with oblique ends; among these soon appear blade-like crystals of the same habit, often twinned; and, at about the same time, lozenge-shaped tabular crystals appear all through the slides. These latter (text figures 203 and 204), which are undistorted crystals, are very symmetrical rhombic plates, sometimes untwinned, sometimes twinned repeatedly. They do not seem to form mimetic twins and develop into hexagonal plates, as is so commonly the case with rodents. The twins are of the usual horse-type (text figure 205), on a twin axis normal to a prism-base edge and lying in the base. But in the blade-like crystals these appear as contact twins with the common prism-like edge parallel to the length of the blade-like crystal; the blades being elongated in the direction of two oppo- site prism faces, and consisting, therefore, of the two basal faces and two prism faces and terminated by the prism faces. This elongation produces in the untwinned crystals a tricHnic appearance. In these same blade-like crystals another kind of twinning is very commonly seen, on a unit pyramid as the plane of twinning, the twin being inter- penetrant and showing an X-shaped cross-section (text figure 206). These were also observed in the plates. The twins of the plates are, as stated, of the usual horse-tjrpe, but the plates being very symmetrical the group formed by the twinning has often the outline of a truncated triangle, and sometimes is nearly triangular. By parallel growth the plates become greatly elongated in the direction of the clino-axis, forming parallel growth groups, and they also grow together on the base in groups, extending in the direc- tion of the same axis. Sometimes the rhombic plates form radiating groups by uniting on the base, the radial character showing when the plates are seen on edge. On the base, the color of the crystals is a deep scarlet, owing to the very slight pleochroism on this aspect. Pleochroism is weak on the base, but strong when the edge aspect is presented; a pale reddish-orange, B blood-red; c blood-red, somewhat deeper than h, but the two practically equal. Double refraction is so weak on the basal aspect that the quartz wedge scarcely shows the difiference between b and c. On the edge view, looking along c or b, however, the double refraction is quite strong. On the edge view looking along b the extinction is oblique, about 15° in the acute angle. The orientation of the optic axes is a A 6=37°, in the obtuse angle; b=b; c A a = 15°, in the acute angle. The plane of the optic axes is the plane of symmetry (010), and the acute bisec- trix is the axis of least elasticity, BXa=ci; the optical character is hence negaiive. 230 CRYSTALLOGEAPHY OF HEMOGLOBINS OF THE RODENTIA. White Rat, Albino op Mus norvegicus (Mus norvegicus var. albus Hatai) . Plate 53. A number of specimens were examined at different times, the living animals being bled in the laboratory. The general method of preparation was to bleed the animal into oxalate, lake the whole blood with ether, and centrifugaUze. From the clear solution slide preparations were made as usual. Modifications of the method above described, using corpuscles and adding variable amounts of plasma in excess of the normal, gave about the same results as the preparations of the whole blood. The blood crystaUizes very readily, so much so that the crystals are usually small unless methods of preparation are used that retard crystaUization. Being small, they show but Uttle color; the crystals examined were, in each case, determined to be oxyhemoglobin by the microspectroscope. A superficial examination shows that the crystals are of several habits, and they look as though they were of different systems. Careful study shows, however, that they are all of the same crystaUization, although one type, a hexagonal plate, seems to be sometimes a mimetic twin of the normal crystals. Oxyhemoglobin of Albino of Mus norvegicus. Orthorhombic : The axial ratio was calculated from the traces of the macrodome on the prism, assuming the same prism as in Mus norvegicus; iiia a \h : 6 =0.7829 : 1 : 0.7332. Forms observed: Unit prism (110), brachydome (101). Angles: The only angles that can ordinarily be observed are the plane angles between the edges produced by a prism face intersecting the two brachydome faces, that is edges 110-011 A 110-011 = 120°. The half of this angle can also often be meas- ured, and it is 60° actual angle. In twins of the stellate shape, the edges are inclined to each other at 60°. As this dome angle on the prism is the same as in Mus norvegicus, the true dome angle Oil A Oil is assumed to be the same also, 72° 30', which makes the prism angle 110 A 1T0=76° 7'. Habit thin tabular, elongated along the vertical axis and the crystal tabular on two opposite faces of the unit prism (110), the end being formed by the faces of the brachydome (text figure 207). The tabular crystals are thus roughly six-sided with two sides longer than the other four. Some symmetrically developed crystals of the combination of prism and brachydome were seen (text figure 208), but they were always very small. Generally when the prism faces were equally developed, which occasionally happened, the dome faces were unsymmetrical, two opposite faces being larger and the other two smaller, but usually two opposite prism faces were larger, making the tabular crystal, and the other two prism faces were smaller (text figure 209). The prism is very nearly square, although evidently not quite so; but no cross- sections of it could be obtained for measurement. The ratio given o : b =0.7829 : 1 was calculated by assuming the same prism that was determined for the Norway rat crystal, which gave the same plane angle of macrodome on prism face as in this albino variety. The usual crystals are hence like vertical sections of the prism, parallel to one pair of faces; and, as the section approaches the exterior of the symmetrical crystal, the outline of the section becomes nearly four-sided; whereas a median section is nearly regularly six-sided, with four short and two long sides. This flattening of the prism produces the tabular effect, and there is, therefore, a flat view and an edge view of each crystal possible. The above descriptions refer to the flat view, but the edge view is quite analogous. The crystals twin by growing together on a prism face either on the flat or on the edge aspect, with the prism edges of the individuals of the trilling (which it usually is), at almost exactly 60° with each other. When this is on the flat aspect the crystals seem to pile up on each other at the 60° angle (text figure CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. 231 210) and there may be more than three in the combination. This kind of grouping pro- duces roughly hexagonal plates, and even fairly regular hexagonal tabular crystals, in which the composite character can, however, usually be made out. Rarely, this com- posite character almost disappears when the members of the twin are many and very thin, and the crystal then becomes pseudohexagonal; this is the normal hexagonal crystal of the rodents. When, on the other hand, the crystals twin on edge, they seem to be more interpenetrant; although in these, too, there is often the appearance of piling up. The twins of crystals on edge are in the form of six-rayed star-shaped groups (text figure 211). In some cases the two aspects of the crystals are presented in the same group, as is natural, for there is no essential difference of structure on the two kinds of prism faces, the broad face and the narrow face. The twin on the flat seems to be on a brachypyramid nearly, or quite in the zone of the prism-dome edge, but the composition face is the unit prism; in the star-shaped twins, the twin plane and composition face are a pyramid of the unit series. FiOB. 207, 208, 209, 210, 211. Miu norvegiciu aUnu Oxyhemoglobin. The crystals therefore occur in five habits or forms: (1) Prismatic crystals, long or short, consisting of the symmetrically developed prism and generally unsymmetrically developed brachydome. Rather rare, but the most nearly symmetrical crystal. (2) Elongated six-sided plates formed from (1) by flattening on two opposite prism faces. The common single crystal. (3) Composite and rough hexagonal plates, twins of (2) on the brachypyramid, presenting the tabular aspect. The common crystal. (4) Six-pointed star-shaped twins, produced by twinning on the unit pyramid, presenting the narrow planes or edge aspect. Almost as common as (3) . (5) More or less regular hexagonal plates, mimetic twins of the type of (3). Rather rarely observed. The color of the crystals is rather pale, owing to their very small size, but pleo- chroism is quite noticeable; a and b nearly equal, almost colorless in these minute crys- tals; c is reddish-orange to deeper red. The orientation of the elasticity axes is only approximately made out for a and h, which are nearly equal, but c=d. The peculiar development of the crystals renders it nearly impossible to get views along axes a and b; there can be little doubt, however, that the acute bisectrix Bxa=c, which would make the optical character positive. 232 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. Norway oh Bkown Rat, Mus norvegicus. Plate 54. The specimen of blood was received from the Wistar Institute of Anat- omy, Philadelphia. The animal was bled into oxalate, and the blood used immediately. On laking, the blood at once began to crystallize, and within a few minutes a considerable amount of the crystals of oxyhemoglobin had formed in the tube. These were separated by centrifugaUzation, and from the clear mother-Uquor the shde preparations were made. After cover- ing the shdes crystals formed rapidly, and they were quite insoluble. The color of the plasma was almost entirely discharged, showing the ciystal- Uzation to be nearly complete. The crystals were small and thin, as in the case of the white rat. They kept well and showed no tendency to dissolve Figs. 212, 213, 214, 215, 216, 217. Mut norvegieiu a-Oxyhemoglobin. upon moderate increase of temperature. Even after a month the form of the crystals in the shdes had not changed materially. The crystals were oxyhemoglobin. Two forms of the oxyhemoglobin appeared: one pris- matic, and probably orthorhombic, hke the white-rat oxyhemoglobin; the other isotropic and apparently isometric, but showing hexagonal out- lines. The prismatic form was the first to appear; the isotropic form developing later may be an isomer of the first form or a mimetic twin. They are distinguished as a-oxyhemoglobin and |3-oxyhemoglobin. a-Oxyhemoglobin of Mus norvegicus. Orthorhombic: Axial ratio a : b : 6 =0.7829 : 1 : 0.7332. Angles: Brachydome angle Oil A OTl =72° 30'; the prism angle was not observed, but was calculated as 76° 7'; profile of dome edges over pole when the prism lies on its side, edges 110-011 A TTO-OTl = 120°; this is the plane angle of the dome on the prism. Habit of the first crystals to form prismatic and generally flattened on two opposite prism faces, making a six-sided tabular crystal elongated on the vertical axis, as is com- mon in the white rat (text figure 212). Some symmetrically developed prismatic crystals were observed that showed the dome termination in symmetrical development; these looked like normal orthorhombic crystals (text figure 213). But the distorted crystals, CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. 233 flattened on two opposite prism faces (text figure 214), have a decidedly monoclinic aspect. Nothing that was observed of the optical characters could determine that these crystals were not orthorhombic ; but no end views, looking along the length of the crystal, could be obtained, and hence they may be really monocUnic. Twins on the brachy and unit pyramids formed as has been described for the white rat; on the flat aspect, twinned on the brachypyramid, they make pseudo-hexagonal groups; and twinned on the unit pyramid, with the edge of the flattened prism presented, they make six-pointed star- shaped groups (text figures 210, 211). What appears to be a twin on the prism also occurs, in some cases producing the effect of a carpenter's miterbox, where the two crystals on edge appear on either side of one presenting the flat aspect (text figure 215). As the crystals continue to develop, short prismatic crystals, flattened on two prism faces, appear, and they produce hexagonal plates, owing to the angle of 120° of the dome profile, which is of course the same as the profile of dome to prism outline (text figure 216). These crystals show much less double refraction than the elongated crystals and are sometimes practically isotropic. When the prism is symmetrically developed and in equilibrium with the dome (text figure 217) the crystals resemble octahedra, and they appear to pass into isometric octahedra, the ^-oxyhemoglobin crystal. Pleochroism is marked in the elongated prismatic crystals, but wanting in the hexag- onal plates and in the equidimensional, octahedral-looking, prism-dome combinations. The colors are a=b (about), pale yellowish-red to pale red; c deeper red. No end views were seen, so that the pleochroism of a and h could not be differentiated. Double refrac- tion is strong in the long crystals, but very weak or entirely wanting in the equidimen- sional crystals and in the hexagonal plates. The symmetrical crystals in convergent polarized light showed traces of the brushes of an interference figure, looking along the macro-axis, the brushes passing out of the field upon rotation of the stage, showing that the observation was being made on the obtuse bisectrix of the optic axes. The orienta- tion of the elasticity axes is a =6; b=a; c=i. No observation of the interference figure, looking along the acute bisectrix was possible, but the acute bisectrix is the axis of least elasticity, Bxa^c, and the optical character is positive. ^-Oxyhemoglobin of Mua norvegicus. Isometric or pseudo-isometric. Forms observed: Octahedron (111). Angles: The angle over the pole of the octahedron 111 A TTl —71° (about). Figs. 218, 219, 220. liut mmiegiciu IS^xyhemoglobin. Habit octahedral; symmetrical isometric octahedra, or distorted octahedra formed by the crystal lying on one face and hence developing into forms with a nearly triangular to almost hexagonal profile (text figures 218, 219, 220). These hexagonal sections of the octahedron closely resemble the hexagonal-looking tabular crystals of the a-oxyhemo- globin, but may be distinguished from them by careful examination. The crystals show the oxyhemoglobin red of the white and Norway rats, rather a pale brownish or yellowish-red. There is no pleochroism nor double refraction; the crystals appear to be absolutely isotropic. 234 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. The passage of the orthorhombic crystal of a-oxyhemoglobin into the isometric or pseudo-isometric crystal of /?-oxyhemoglobin, that was described above, depends upon the pseudo-isometric character of the a-oxyhemoglobin (its angles being near the angle of the octahedron), if no change in substance occurs in the change from the dis- tinctly orthorhombic crystal to the pseudo-isometric crystal of the hexagonal or equi- dimensional type. More probably there is a change of substance and an isomer or polymer is formed (or may already exist in the solution) and then the hexagonal or the equidimen- sional form is a mixed crystal containing both a-oxyhemoglobin and /3-oxyhemoglobin. The influence of the j9-oxyhemoglobin on the a-oxyhemoglobin, or the concentration of the solution, determines the conversion of the a-oxyhemoglobin into the other isomer (or polymer) ^-oxyhemoglobin. Black Rat, Mua rattus. Plates 54 and 55. Specimens of the blood of the black rat were obtained from the Wistar Institute of Anatomy, of Philadelphia. The animal was bled into oxalate and the blood used inmiediately. The corpuscles, separated by centrif- ugalizing, were laked with ether, oxalated, and the solution again centrif- ugalized. From the clear solution thus obtained the slides were prepared. The blood crystallized very readily; in fact, it is probable that better prep- arations would have been obtained if the whole blood had been used. The crystals do not form so readily as those of the white rat or the Norway rat, but they are quite permanent, show no signs of dissolving on slight increase of temperature, and they keep for weeks in the slides. They are not nearly so insoluble as the crystals of the white or Norway rats, however, and upon an increase of temperature, up to a temperature of 25" C, they begin to dissolve, so that they can not be satisfactorily studied in warm weather. This character is in sharp distinction from the insolubility of the oxyhem- oglobin crystals of the white and Norway rats, which are permanent at temperatures up to 35° C. The crystallization is not so complete as in the case of the other rats mentioned, so that the fluid remains of a strong red color, showing much oxyhemoglobin still in solution. Oxyhemoglobin of Mus rattus. Orthorhombic: Axial ratio a : b : (5=0.7829 : 1 : 0.5864. Forms observed: Unit prism (110), macrodome (101). Angles: No cross-sections of the prism could be observed. The only angle that can be determined is the plane angle of the brachydome on the prism face, edges 110- 011 A 110-011 = 130° 26', average of nine measurements. Assuming the same prism for this rat that was determined for the Norway rat from the true dome angle and the plane angle of the dome on the prism, the axial ratio was calculated. This makes the macrodome of the black rat (101), the macrodome (405) on the axial ratio of the Norway rat, the average measured angle for the edges, 130° 26', agreeing exactly with the calculated value. Habit tabular on two faces of the prism, the crystal consisting of the prism (110) and the brachydome (Oil). The prism is flattened on two opposite faces, as is common in the rats (text figure 221), and the dome termination may be of four equally developed dome faces, or two large and two small dome faces, or even of two equally developed faces on one end and one large and one small face on the other end. In some crystals two dome faces appear at one end of the prism and only one at the other end, making a five-sided plate (text figure 222) , When two dome faces on the same side of the crystal are developed (one at each end), the plate becomes unsymmetrically four-sided (text CRYSTALLOGRAPHY OP HEMOGLOBINS OP THE RODENTIA. 235 figure 223). By shortening of the prism the tabular crystal becomes hexagonal in outline (text figure 224) ; the crystals do not elongate into the long tabular crystals so common in the Norway and albino rats. Twinning on the flat and on edge, as seen in the white rat, was observed, but it did not occur so frequently as in the case of that species. The twinning is upon a pyramid in each case, as is common in the rats. The twins on edge were of two individuals only, as a rule, and did not form the six-pointed star like Norway and white rat twins (text figure 225). On the flat, the twin consists of two individuals also (text figure 226), and does not result in the formation of a hexagonal plate, as in the case of the Norway and white rat crystals. The difference in the twins is no doubt due to the fact that in the twin on the brachypyramid the prism edge of one individual is parallel to the prism-dome trace of the other; but in the corresponding twins of the white rat and the Norway rat crystals the angle of the twin of the prism edges in the three members is 120°, while in the black rat (and Alexandrine rat) the angle of the prism edges in the two members of the twin is about 130° 25', being for the Alexandrine rat 130° 19' and for the black rat 130° 26'. Three crystals of the Norway or white rat could twin at the angle of 120° to make a regular hexagonal plate; but three crystals of the black or Alexandrine rat so twinned could not produce a regular hexagonal plate. Fias. 221, 222, 223, 224, 225, 226. Mu» railut Oxyhemoglobin. Pleochroism is fairly strong, but from the positions of the crystals presented o and 6 can not be directly observed. The colors of a and b are evidently close together, rang- ing from pale yellowish-red to deeper red, according to the thickness of the crystal. The color of c is always much deeper; even in the thinner plates it is a deep red. Double refraction is strong, extinction is straight in all aspects presented. The orientation of a and i could not be observed; it is probably the same as in the Norway rat, a=h,h=a; the axis of least elasticity c^ti. From the fact that the elasticities of a and 6 are nearly the same, it is probable that the axis of least elasticity c is the acute bisectrix, Bxa=f, this makes the optical character positive. No interference figure could be observed. Alexandrine Rat, Mus alexandrinus. Plate 55. Specimens of the blood of the Alexandrine rat were obtained from the Wistar Institute of Anatomy, of Philadelphia. The blood was collected in oxalate and was used immediately. The corpuscles were separated by centrifugalization, laked with ether, additional oxalate added, and the solution centrifugalized. The slide preparations were made as usual. Crystallization proceeded rapidly before and after covering the sUdes; 236 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. the crystals kept well at temperatures under 15° C, but did not form well at 25° C. or higher. At moderate temperatures they showed no tendency to dissolve. They were oxyhemoglobin. In solubility they resemble the crystals of the black rat, being perhaps somewhat more soluble, but they are very much more soluble than the crystals of the white rat or the Nor- way rat. CrystaUization was not complete, the mother-liquor remaining quite strongly colored. The crystals in the slides were in good condition a month after the preparations were made. Only one kind of crystal was observed, corresponding to the a-oxyhemoglobin of the Norway rat. Oxyhemoglobin of Mus alexandrinus. Orthorhombic: Axial ratio a:b: 6-0.7829 : 1 : 0.5880. Forms observed: Unit prism (110), brachydome (Oil). Angles: From the aspects presented by the crystals the prism angle could not be measured; it was assumed as the same as that determined for the Norway rat, 110 A 110=76° 7'. The only angle that could be measured was the plane angle of the brachy- dome on the prism face; this was, edges 110-011 A 110-011 = 130° 19', the average of a number of measurements. From this angle and the assumed prism angle of 76° 7' the axial ratio was calculated. The true brachydome angle Oil A OTl could not be observed. 227 228 229 Fios. 227, 228, 220. Mut dltxandTmut Oxyhemoglobin. Habit tabular on two opposite prism faces and somewhat elongated parallel to the vertical axis (text figure 227); the crystal consists of the flattened prism terminated by the flat brachydome. Distorted crystals, in which the two brachydome faces on one side of the prism are alone developed (text figure 228), or crystals with one end so devel- oped and the other showing the two dome faces (text figure 229), are very common; these distorted crystals are rather more common in the crystals of this species than in those of the black rat. There are thus formed four, five, and six-sided plates, the four and five-sided tabular crystals having a decidedly hemimorphic aspect. By shortening of the prism the tabular crystal becomes a hexagonal plate, but owing to the two angles of 130° 19' it is not a regular hexagon. The crystals are considerably larger than those 7 Figs. 230, 231. Mu» alexanirinui Oxyhemoglobin. of the black rat in preparations made under the same conditions, indicating that they are more soluble than those of the black rat. Twinning occurs on the flat aspect; two crystals united on the prism face with the orientation that of a twin in the zone of the brachypyramid, which would bevel the dome-prism edge (text figure 230); but such twins are very rare. The stellate twin on a pyramid of the unit series is more common. CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. 237 sometimes forming a six-rayed star, like text figure 211; more commonly forming X-shaped twins by the interpenetration of two prisms, like text figure 225; and also often showing interpenetration, but with one or two of the rays developed on one side of the prism only (text figure 231). The twin in the zone of the brachypyramid can not form mimetic hexagonal crystals because of the flat dome angle. The axial ratio of this species is related to that of the Mus norvegicus groups of rats by the same 5 : 4 ratio that was true of the black rat. Indeed, except for the difference in solubility and in habit of crystal there is no great difference between crystals of the black and the Alexandrine rats, and these rats look like varieties of the same species. The difference in angle of dome on prism, that makes the slight difference in axial ratio, is within the limit of error of the method of observation, and the axial ratios of the crystals of the two species are probably identical. Pleochroism is rather strong, but, owing to the positions of the crystals presented, it is not possible to distinguish between a and b, the colors of which are apparently nearly alike. The pleochroic colors are a and 6 pale to deeper yellowish-red; c rather deep red. Double refraction is strong and extinction is straight in all aspects that could be observed. The orientation of a and 6 could not be observed, but may be assumed to be the same as in the Norway rat, a =&, h=a; the third axis could be observed and gave c =c. No inter- ference figure could be observed, but from the fact that a=h (nearly) it is probable that the axis of least elasticity c is the acute bisectrix, Bxa=e; and the optical character is positive. Reviewing these four species of rats it will be seen that they may be arranged in two groups: (1), the Miis norvegicus group, comprising Mus norvegicus and its albino, Mus norvegicus alhus Hatai, in which the crystals are orthorhombic and pseudo-isometric, with an axial ratio of 0.7829 : 1 : 0.7332; and (2), the Mus rattus group, comprising Mus rcdtus and Mus alexan- drinus, in which the crystals are orthorhombic, but not pseudo-isometric, and the axial ratio is 0.7829 : 1 : 0.5864 or 0.7829 : 1 : 0.5880. As noted above, these ratios for the vertical axis stand in the ratio of 5 : 4; but that they are different axial ratios, and not simply different crystal habits, is shown by the twins in the brachypyramid zones following these ratios in each case, thus indicating that the difference is one of the form of structure of the crystal and not simply a difference of development. The oxyhemoglobin of the Mus norvegicus group is not the same substance as the oxyhemoglobin of the Mus rattus-alexandrinus group. Canadian Porcupine, Erethizon dorsatus. Plates 56 and 57. The specimen was received from the Philadelphia Zoological Gardens and was in the form of rather hard clots. The clotted blood was ground up in sand and ether, a httle oxalate and water added, and the mixture centrif- ugaUzed. From the clear liquid thus obtained the slide preparations were made as usual. Crystals began to appear within about half an hour after the slides were covered; they were rather small and thin and showed a tendency to dissolve in the solution. These are described as a-oxyhemo- globin. Inside of 20 hours, they had mostly been dissolved from the slides, and their place was taken by the crystals of /8-oxy hemoglobin; in only a few slides were there any of the a-oxyhemoglobin crystals remaining. The second crop of crystals, /8-oxyhemoglobin, appears to be less soluble and more stable than the a-oxyhemoglobin crystals. A second preparation was made the next day, a new lot of slides being mounted from the cleansed blood above described, to which an additional 238 CRYSTALLOGRAPHY OP HEMOGLOBINS OP THE RODENTIA. amount of oxalate had been added. This blood was highly charged with ether (from the method of preparation) and this prevented the develop- ment of bacteria, so that the blood kept well. In this series, the first crys- tals to appear were the a-oxyhemoglobin crystals, as before, but of a differ- ent type; and soon after they appeared the /3-crystals began to form, of the same habit as in the first experiment. All the crystals examined were determined to be oxyhemoglobin by the microspectroscope. a-Oxyhemoglobin of Erethizon dorsaius. Monoclinic: Axial ratio a : b : (i= 0.5543 : 1 : 6; /}=56°. Forms observed: Unit prism (110), orthopinacoid (100), clinopinacoid (010), basal pinacoid (001). Angles: Prism angle, traces of the prism on the base, edges 110-001 A 1T0-001 = 58° (normals); orthopinacoid to base 100 A 001=56°=^. 234 235 Fios. 232, 233, 234, 235. Erethizon dortaliu a-Oxyhemoglobin. Habit tabular on the base; in the untwinned crystals of the first preparation, type (a), the combination was the base cut by one-half of the prism and one face of the ortho- pinacoid, with the two faces of the clinopinacoid (text figures 232, 233), much in the habit of the clinohedrite type of the monoclinic system (domatic class) . In the twinned crystals of the second preparation, from blood containing more oxalate, type (b), while the corresponding angles are the same, the crystal is a twin on a normal to the prism-base edge, but so developed that it looks as though the composition plane was the normal to the base that included this common prism-base edge (text figure 234). The obtuse prism angle of 122° (or 58° normals) appears symmetrically four times on these twins, while the double of the acute angle 116° (or 58° X 2 =64° normals) appears twice in sym- metrical position. In some crystals the development is as usual in this horse-type of twin; the two parts united on the base, elongated on the common prism-base edge, and with the ends overlapping (text figure 235). These twins are evidently repeated in polysynthetic order, and the optical characters for the twin are abnormal, being the summation of the optical characters of its members, which vary in their orientation. Hence the twin has a plane of symmetry for its optical characters as seen on the flat, which is normal to the base, and includes the prism-base edge, the twin plane in short. The crystals are rather strong oxyhemoglobin red; the simple crystals, being thin, are paler than the twins. On the base, the simple crystals of type (a) are distinctly pleochroic, but not strongly so; on edge, the pleochroism is quite strong. The pleochroic colors in these thin crystals of type (a) are a pale yellowish-red, b rather pale scarlet, c (on edge) deep red. In the second type of crystals, the pleochroism on the base is very slight, owing to the averaging of the elasticities in the composite crystal; and the apparent a comes much nearer to the apparent B. On edge, these still show considerable pleochro- ism; but the contrast is not so strong as in the untwinned crystals. Extinction in type CRYSTALLOGRAPHY OF HEMOGLOBINS OP THE RODENTIA. 239 (a) is symmetrical on the base and straight looking along the a-axis; but is 19° to 20° (perhaps higher) on the edge view, looking along b. In the twins of the second type of crystal, the extinction is symmetrical with the outline and parallel to the twin axis. In the untwinned crystals the interference may be easily seen in traces when the crystal does not lie fiat; but is not readily observed when the crystal is flat, only traces of it showing. This is due to the orientation of the elasticity axes, which is a = 6; b A a =20°, in the acute angle; c A 6=54:°, in the obtuse angle. The plane of the optic axes is normal to the plane of symmetry and inclined to the orthopinacoid at an angle of 54°, in the obtuse angle; it is hence inclined 34° to the normal to the base. The acute bisec- trix is the axis of least elasticity, Bxa=(, and the optical character is positive. ^-Oxyhemoglobin of Erethizon dorsatus. Orthorhombic : Axial ratio a : b : (5=0.8170 : 1 : 6. Forms observed: Unit prism (110), base (001). Angles: Prism angle 110 A 1T0=78°30' (normals); prism to base 110 A 001=90°. FlQB. 236, 237, 238. Erethiion dormlus ^-Oxyhemoglobin. Habit tabular on the base; consisting of the short prism (110) cut by the basal pinacoid (001) (text figures 236 and 237) ; the ratio of length to breadth of the prism being generally about 1 : 3, but the tabular crystals thicken and become equidimensional in some cases. The crystals are generally more or less hopper-shaped on the base and evi- dently grow more actively on the prism faces and edges and less so on the base; the result is often the appearance of a skeleton crystal, looking down upon (001), with the crystal axes marked out in more solid substance. On edge views and sections, the solid central part is seen to run out from the center to the four coigns of the basal face; the interven- ing parts, between these four directions and the outer surfaces of the prism, having a porous aspect. These crystals are relatively large, much larger than the crystals of the a-oxyhemoglobin that have been described. The crystals showed parallel growth; and a twin, of the interpenetrant type, on a brachydome (text figure 238), was seen, the two parts crossing at an angle of nearly 90° The crystals are a rather bright scarlet color, and on the flat as well as on edge they are quite pleochroic. The pleochroism is a nearly colorless to yellowish-red, depend- ing on the thickness; b deep scarlet to blood-red; c deep blood-red. On the base, the extinction is symmetrical; and on edge views it is straight in all aspects, when the section is normal to the base. On the base, in convergent light, a biaxial figure is seen, symmet- rically placed, with the separation of the brushes wide; 2E is above 75°, but was not exactly measured. The orientation of the elasticity axes is: a = 6; b=a; c = > e and hence the optical character is negative. CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. 243 ^-Oxyhemoglobin of Hydrochaerus capyvara. Orthorhombic: Axial ratio not determinable. Forms: Unit prism (110)? terminations were wanting. Angles: No angles could be measured. Habit prismatic, the crystals appear to be a prism, much elongated; but perhaps the planes may be two pinacoids. When examined, a few hours after the crystals had begun to form, the terminations were imperfect and not measurable. Later they were lost by resolution. Pleochroism was weak; c deep red, a (or b) somewhat paler. Extinction is parallel to the length of the crystal in all aspects that were examined. On the side view, in some cases, a biaxial figure was seen, with the plane of the optic axes including the vertical axis, which is evidently c; this must, therefore, have been seen when looking along a and the optical character is negative, as in the case of the tetragonal crystals of a-oxyhemoglobin. The crystallization is evidently orthorhombic, but the crystallographic constants can not be determined beyond those already stated. Domestic Rabbit, Lepus cunicvliis. Plates 59-61. The living animal was purchased and bled into oxalate in the labora- tory. The corpuscles were separated from the plasma by centrifugalizing; and the preparations were made from the corpuscles by laking with ether and centrifugaUzing for 3 hours. The sUde preparations were made from the clear centrifugalized blood as usual. Crystallization begins in the protein ring soon after the sUdes are covered, and it proceeds rapidly at room temperature. As the solution under the cover comes to an equiUb- rium, these first crystals dissolve and disappear from the slides. Upon putting the slides in the cold at near 0° C. a second crop of crystals appears, some of which are like the first crop and some are of a different type. These are distinguished as a-oxyhemoglobin and /^-oxyhemoglobin. The a-crystals are less soluble and may be examined at near room temperature; but the jS-crystals are much more soluble, and dissolve rapidly when the tempera- ture is raised a few degrees above 0° C. They had to be examined and photographed in a room temperature near the freezing-point. Another preparation, made from the same blood, was evidently not evaporated quite to the same point as the first before applying the cover; for, while two types of crystals developed, both tended to dissolve upon increasing the temperature a few degrees above 0° C, and they therefore had to be exam- ined at about this temperature. So long as the preparations were kept at about the freezing-point the crystals continued in excellent condition. a-Oxy hemoglobin of Lepus cuniculus. Monoclinic: Axial ratio a : b : 6 =0.643 : 1 : 0.797; /9=85°- Forms observed: Unit prism (110), clinoprism (320), chnodome (Oil), clinopina- coid (010), orthopinacoid (100). Angles: Unit prism 110 A 1T0=65°30' (normals); clinoprism 320 A 320=88° (normals); clinodome Oil A 011=77°; prism edge to dome edge, in the plane of sym- metry =85° (normals) =/?, or 95° actual angle. Two habits of crystals develop: (a) the first crystals to appear are usually pris- matic, consisting of the two pinacoids and the clinodome, elongated vertically and flat- tened on (100) (text figure 247), and with or without the unit prism (110) (text figure 248) ; (b) the second type, which is much more symmetrically developed, consists of the clinoprism (320) in combination with the clinodome and the clinopinacoid (text figure 249). Type (a) crystals are elongated vertically and striated on the orthopinacoid; 244 CRYSTALLOGRAPHY OF HEMOGLOBINS OF THE RODENTIA. by the disappearance of this plane and the development of the cUnoprism, while, at the same time, the clinopinacoid becomes larger, they pass into the second type of crystals. Type (b) crystals are tabular on the clinopinacoid, elongated vertically, and generally smooth, not striated, as in the type (a) crystals. Both kinds of crystals of the a-oxyhemo- globin are much smaller than the crystals of the /?-oxyhemoglobin. The (b) type of crystals form parallel growths and also seem to twin on the orthopinacoid; the twinning was observed in polarized light. The color is the usual oxyhemoglobin red, but the crystals are quite pleochroic; a colorless or pale yellowish; b rose-pink to pale red; c deep red. Double refraction is moderately strong on most aspects; in all sections in the zone of 100-001 the extinction is straight or symmetrical; on the plane of symmetry, looking along b, the extinction is obUque; 15°, in the obtuse angle, from the prism edge. On this aspect in some crystals, a biaxial interference figure was seen with the above orientation, the plane of the optic axes being inclined to the vertical axis 6 at 15°. The orientation of the elas- ticity axes is a A 6 = 15°, in the obtuse angle; 6 A a = 10°, in the acute angle; c=6. The plane of the optic axes is hence normal to the plane of symmetry. The angle between the optic axes was not accurately measured; the separation was considerable, however. The acute bisectrix emerges normal to the plane of symmetry; it is c and hence the optical character is positive. /A /^ 247 \y V y^ a- d 248 249 w FlQS. 247, 248, 249. Lepua euniculua o-Oxyhemoglobin. Fios. 250, 251. Lepitt cuniculua /3-Oxyhemoglobin. ^-Oxyhemoglobin of Lepus cunicvlus. Orthorhombic: Axial ratio a : 6 : i =0.5317 : 1 : 6, but no cross-sections were found from which this angle of the prism could be obtained. Habit (a) prismatic on the vertical axis, the crystal consisting of the unit prism, clinopinacoid, and base (text figure 295); the crystals rather large and well-formed; also (b) tabular on the clinopinacoid, the crystal consisting of the three jf pinacoids only, with the base and orthopinacoid in equilibrium; thus making a rhomboidal plate (text figure 296). Both kinds of crys- tals, types (a) and (b) , were found very sparingly, in a preparation of defibrinated blood without oxalate; the plate-like crystals of type (b) d- were seen still more sparingly in a preparation in which oxalate was used. Often preparations of defi- brinated blood failed to develop these crystals. Pleochroism is very strong; a pale yellowish-red, b rose-pink, c deep blood-red. Double refraction is strong, and the extinction, meas- ured from the prism edge, is 15°. The orientation of the optic axes is as follows: a Ao =27° in the obtuse angle ; b = b ; c A * Ii?if J 295 A^ - 294 ^ VJ.JJ' J 297 293 Figs. 293, 294. Cani» familiaris a-Oxyhemoglobin. Fias. 295, 296, 297. Canit familiarit ^-Oxyhemoglobin. 268 CRYSTALLOGRAPHY OP THE HEMOGLOBINS twin plane (text figure 297). In such twins, the extinction angle is, of course, less than that recorded above, about 7° or 8°. Chow Doq, Canis familiaris var. Plate 73. The specimen of blood was received from the Philadelphia Zoological Gardens, and was in a clotted and rather putrid condition. The specimen was ground in sand and etherized and then centrifugaUzed for several hours; and from the clear solution thus obtained the slide preparations were made as usual. Crystals form rapidly and readily at room temperature, and show no sign of dissolving. Within 3 hours after the slide preparations were made, satisfactory photographs were procured. The crystals were oxyhemoglobin, and resemble those of the common domestic dogs very closely; appearing, however, to differ slightly in angles. Oxyhemoglobin of Canis familiaris var. Orthorhombic: Axial ratio a : b : 6 =0.6696 : 1 : 0.2878; a: 6 =1 : 0.4348. Forms observed: Unit prism (110), macrodome (101), base (001). Angles: Prism angle 110 A 1T0=67°; macrodome angle 101 A TOl =47°. Habit long prismatic on the vertical axis, the crystals consist- ing of the unit prism terminated by the macrodome and sometimes also by the base (text figures 298 and 299). The first crystals to form are long and hair-like or needle-like; but, as they grow, they develop greater thickness; so that the normal and fuUy developed crystal has a ratio of length to thickness of about 25 : 1 to 15 : 1, and sometimes somewhat less, down to 10 : 1. Doubly terminated crystals of measurable quahty are comparatively common and the crystals are well developed and sharp in outline. Cross-sections of the crystals are, however, difficult to find, owing to the great length of the crystals in proportion to their thickness. Parallel growth is normal on the prism faces and on the brachypinacoid, and the crystals flatten in this way in the macropinacoid direction. A very common feature is the development of double crystals in this way, two growing side by side in parallel orientation and united on the brachjrpinacoid. The hair-Uke crystals which have a ratio of length to thickness of 100 : 1, or even 200 : 1, do not show this tendency to parallel growth until they have considerably increased in thickness; the most perfect crystals are generally found in the ratio of about 20 : 1. The crystals grow singly through the slide, or in irregular groups, and often in radiating clusters; small thin rods, attached in such radiating groups to larger composite crystals, being particularly common. In the protein ring, and along the cover edge, they grow usually more or less normal to the surface from which they spring; but radiating tufts are common here also. Twinning was not definitely observed; but indications of twinning upon the brachydome and upon the pyramid were seen; and twinning on the prism appeared to be present in many of the composite groups. These last could hardly be made out with certainty without observations of cross-sections, and these were impossible to find. Pleochroism is marked and is as follows: a yellowish-red, i pale red (medium oxy- hemoglobin red), c deep red. Double refraction is strong, and extinction is parallel to the vertical axis in all side views; on cross-sections extinction is symmetrical, and parallel to the crystal axes. Looking along the brachy-axis in convergent fight, the interference figure is seen, with widely separated brushes, the plane of the axes being the brachypina- coid. The orientation of the elasticity axes is a =o; b=6; c=6. The acute bisectrix of the optic axes is the axis of greatest elasticity, Ba;<,=o, and the optical character is hence negative. 298 299 FiQS. 298, 299. Oxyhemo. globin of "Chow Dog." OF THE DOGS, WOLVES, AND FOXES. 269 Cross between Colue Dog and Coyote, Canis familiaris and Canis latrans. Plate 73. This specimen of the blood of a hybrid coyote was from an animal 8 years of age that was killed fighting with the other coyotes with which it was confined; and was received from the Zoological Garden at Lincoln Park, Chicago. The blood was not clotted; it was dark purphsh in color and quite putrid. It was ether-laked, mixed with an equal volume of 50 per cent solution of egg-white, and centrifugalized. From the clear solution thus obtained the sUde preparations were made in the usual man- ner. The blood crystallized very readily, and inside of 2 hours after the slides were covered satisfactory photomicrographs were obtained. The crystallization was very complete, and the crystals showed no signs of dissolving. The crystals were oxyhemoglobin. Oxyhemoglobin of cross between C. familiaris and C. latrans. Orthorhombic: Axial ratio a : b : 6 =0.6619 : 1 : 0.2912; a: 6 =1 : 0.4400. Forms observed: Unit prism (110), macrodome (101), (706), base (001). Angles: Prism angle about 67°, not measured exactly; macro- dome angle 101 A T01=47°30'; macrodome 706 A 706=54° .30'; prism to base 110 A 001=90°. Habit medium and long prismatic to capillary; the first-formed crystals are trichites without much dimension, aside from length; these soon increase in diameter, and become measurable crystals, with a ratio of length to thickness of 20 : 1 or less (down to 10 : 1) (text figure 300) ; some of the first-formed crystals, even, show such relative dimensions. The crystals of large size show, usually, parallel growth on the prismatic axis, producing the usual groups of two, side by side, and united on the brachypinacoid; but much more often they form irregular groups in parallel growth, looking like bundles of crystals; and these, especially, grow in such a manner that the dome faces are sup- pressed and a pseudo-basal pinacoid develops. These groups appear cut off square on the ends by this apparent base (text figure 301). Whether the real basal pinacoid actually occurs, or whether the apparent base is always this pseudo-plane, could not be determined. Both the capillary and the stouter crystals are found growing in irregularly radiating tufts, as is common in the dog crystals. Twins were only doubtfully observed; the apparent twins seemed to be on the pyramid. Pleochroism is rather pronounced; a nearly colorless, b rather strong red, c rather deep red; but the colors of S and c are not so very different. Double refraction is strong and extinction straight in all side views. In convergent light, looking along the brachy- axis, the biaxial interference figure is seen, with widely separated brushes, and the orien- tation such that the plane of the optic axes is the brachypinacoid. The orientation of the elasticity axes is a=a, b=b, c=(}. The acute bisectrix of the optic axes is the axis of greatest elasticity, Bxa=a, and the optical character is negative. 300 301 Figs. 300, 301. Oxy- hemoglobin of Croea of Collie and Coyote. Gray Wolf, Canis lupus mexicanus. Plate 74. This specimen was received from the Philadelphia Zoological Gardens. The blood was of a brownish color and was rather thin and watery. It was oxalated and ether-laked and then centrifugalized for several hours. From the clear solution thus obtained the slide preparations were made as 270 CRYSTALLOGRAPHY OF THE HEMOGLOBINS usual. The blood crystallized very readily, and the crystals showed no tendency to dissolve in the solution. They were brownish in color, and the spectroscope showed them to be metoxyhemoglobin. Later, a sort of second crop of crystals appeared, of a somewhat different habit; but evidently the same material, and with the same axial ratio. The morphological characters of these crystals, including the angles, compare very closely with those of other species of Canis, and evidently this metoxyhemoglobin crystallizes much as the oxyhemoglobin does in this species, as is found to be the case in other genera where both substances were observed in one species and could be directly compared. Metoxyhemoglobin of Canis lupus mexicanus. a-' 302 30} — a- 304 FiGB. 302, 303. Canit luput mexicanut Metoxyhemoglobin. Fio. 304. Canit latrana Oxyhemoglobin. Orthorhombic: Axial ratio a : b : 6 =0.6576 : 1 : 0.2863; a : 6 =1 : 0.4272. Forms observed: Prism (670), macrodomes (101), (403). Angles: Brachy-prism angle 670 A 670=75°; unit prism (computed) 110 A TlO = 66° 40'; macrodome 403 A 303=53°; macrodome 101 A 101 (computed) 46° 16'; measured roughly as 46°. Habit long prismatic on the vertical axis (text figures 302 and 303), the crystals as they increase in size becoming strongly striated, due to composition of many individuals in parallel growth. Capillary crystals not so common as is usual in the Canidce, but the thicker crystals rather long in proportion to the thickness, with a ratio of length to thickness ranging from 50 : 1 to 15 : 1. The second-crop crystals are proportionately longer, and many of them have a ratio exceeding 100 : 1. Among these rod-like crystals one or two obUque sections of the prism (670) were seen developed as plates. The crystals of the first crop showed a decided tendency to arrange themselves in groups, radiating in all directions from a center; or frequently the rods would develop brush-like ends, due to the same tendency to form divergent groups. In the second-crop crystals, this tendency to form brush-Uke ends was especially pronounced; and they also formed various tufted arborescent groupings, but without producing the circular radiating clusters found so commonly in those of the first crop. The long, sUghtly divergent tufts of the second-crop crystals, and the spheruUtic radiating groups of the first crop, are charac- teristic of this species; as are also the particular forms that are present, the prism (670) and the dome (403) . The unit macrodome was seen in crystals from a second preparation from the same blood, and probably also the unit prism, but this last was not measured. Pleochroism is not very marked, as the colors are not bright; it is quite noticeable, however, and in some positions rather strong; the colors are shades of brownish, as follows: a pale brownish, b deeper brownish, c deep brown. Double refraction is fairly strong and extinction is straight in all aspects normal to the prism, and symmetrical on cross-sections. The orientation of the elasticity axes is a=a, 6=6, c=6. The interfer- ence figure was not observed, but the relative elasticities appear to indicate that the axis of greatest elasticity is the acute bisectrix, Bxa=a, and hence the optical character is negative. Coyote ok Pbairie Wolf, Cania latrans. Plate 74. The specimen of blood was received from the National Zoological Park at Washington, District of Columbia. The blood was stale, but not putrid. It was laked with ether and centrifugalized for several hours; OF THE DOGS, WOLVES, AND FOXES. 271 and from the clear solution thus obtained slide preparations were made as usual. Crystallization proceeded rapidly after covering the slides, the blood crystallizing more readily than that of the dog. The crystals were small, but gradually increased in size and showed no tendency to dissolve for several days. Finally, however, they did dissolve in the plasma. They were oxyhemoglobin. Oxyhemoglobin of Canis latrans. Orthorhombic: Axial ratio a : b : 6 =1 : 0.4254, from (504). Forms observed: Prism, probably (110), macrodome (504). Angles: The only angle obtained satisfactorily was that of the macrodome (504) = 56°. The prism angle was not observed in measurable position. From (504) the angle 101 A TOl =46° 5' was calculated. Habit of the crystals at first comparatively short capillary, elongated on the prism and terminated by a dome which appears to be (504). Later, the crystals along the protein ring and the cover edge, as well as scattered crystals throughout the body of the slide, became thicker, without increase of length, until the ratio of length to thickness was about 5 : 1 (text figure 304); whereas in the capillary crystals at first developed this ratio was 200 : 1 or more. The larger crystals show very distinctly the striated character, parallel to the length, that is common in the genus Canis, and this is evidently due, as usual, to parallel growth on the vertical axis. The capillary crystals form dense felted masses throughout the slide, or grow in spherulitic tufts, radiating from a common center. The larger crystals are much shorter than is common in the dog tribe, but they were never so thick as is common with most of the dogs. It is evident that the oxyhemoglobin is more insoluble than is usual in this genus, and this fact would account for the differences noted. Pleochroism is marked and double refraction strong. The colors are: a pale yel- lowish-red, 6 deeper red, c deep red. Extinction is straight in all aspects seen; no cross- sections of the prism were observed. The interference figure was not observed. The orientation of the elasticity axes appears to be a=a, b=b, c=6 as usual. The axis of greatest elasticity is probably the acute bisectrix, Bxa=(i, which would make the optical character negative. jACKiL, Canis aureus. Plate 75. The specimen was received from the National Zoological Park at Washington, District of Columbia, and was not putrid, but had the color of stale blood. The blood was oxalated, frozen and thawed, ether-laked, and centrifugalized for several hours, and from the clear solution thus obtained the usual slide preparations were made. The blood crystalUzed very readily and the crystals appeared to be rather insoluble, showing no tendency to dissolve in the solution. Inside of a few hours the crystals had reached measurable dimensions, and after 24 hours many of them were quite large. They formed at first along the protein ring and the cover edge, but soon the entire body of the slide became filled with crystals, and deposition continued until the solution was nearly colorless. The crystals were oxyhemoglobin. Oxyhemoglobin of Canis aureus. Orthorhombic : Axial ratio a: 6 =1 : 0.4245. Forms observed: Prism, probably the unit prism (110), macrodome (101). Angles: The only angle that was satisfactorily determined was that of the macro- dome, 101 A 101 =46°. The prism angle was not made out. Q- 272 CRYSTALLOGRAPHY OP THE HEMOQLOBINS Habit at first (when the crystallization is proceeding rapidly) capillary, long tri- chites which are quite flexible and have a ratio of length to thickness of 1000 : 1 or more; as the crystals be^n to form more slowly the thickness increases and in crystals in the protein ring and along the cover edge, as well as in scattered crystals throughout the body of the slide, this ratio falls to 10 : 1 and even to 5 : 1. The hair-like crystals grow felted together irregularly and in spherulitic groupings radiating from a common center (see plate 75, fig. 445). In some cases these trichites are jesolved, but in most of the slides they persisted for days. The thicker normal crystals are strongly striated parallel with the length (parallel to the vertical axis, 6), as is usual in this genus; and they were evidently bundles of crystals aggregated together in parallel growth. But some of the crystals appearing throughout the body of the slide were apparently single crystals (text figure 305), although of measurable size. They were not vertically striated. All of the larger crystals showed only the long prism (110) and the macrodome (101); but frequently in the groups the dome became so reduced that the crystal was seen to be terminated by a pseudo-base due to the parallel grouping. When the parallel growth resulted in the members of I the group uniting on the brachypinacoid (and hence the group flattening par- j 305 allel to the macropinacoid) the domes were not so noticeable, and again the ^'^ crystal appeared to be terminated by the base. No definite twins were observed. CmUa^au Plcochroism was not very marked when looking normal to the macro- oxyhemogio- pinacoid, but when looking along the macro-axis it was very strong. The colors **""■ were: a pale yellowish-red; b rose-pink; c pale to deep red, according to the thickness. When the crystal was observed on the brachypinacoid aspect, looking along the macro-axis and with o and c in the field, the colors were: a pale yellowish-red, c deep (scarlet) red. In the aspect at 90° to this, with b and c in the field, the colors were: b rose-pink, c pale scarlet, and the colors of b and c were nearly equally strong. Extinction is straight in all aspects normal to the vertical axis and no cross-sections were seen. The double refraction is rather strong. No interference figure was observed. The orientation of the elasticity axes is a=a; b=b;c=6. The acute bisectrix of the optic axes is evidently the axis of greatest elasticity, Bxa=a, and the optical character is negative. Dingo ob Australian Wild Dog, Canis dingo. Plates 75 and 76. The specimen of blood was received from the National Zoological Park at Washington, District of Columbia, and was from a pup. The blood was clotted and putrid, and was evidently stale when placed in our collecting tube with oxalate. The specimen was ground in sand with ether to destroy the clot, a Uttle normal saline solution added, and the ground mixture centrifugalized for several hours. From the clear solution thus obtained the slide preparations were made as usual. Crystallization proceeds very rapidly after the slides are covered, the first crystals to form being rather short rods, but later longer crystals developed. The crystals formed at room temperature show no signs of dissolving, but the slides kept overnight at a temperature below freezing developed crystals that dissolved until equilibrium in the solution was reestablished. When kept at room temperar ture fairly large crystals formed in the protein ring and along the cover edge, and short trichites appeared through the body of the sUde. The crystals are evidently quite as insoluble as is commonly the case in this genus. They are oxyhemoglobin as determined by the microspectroscope. OP THE DOGS, WOLVES, AND FOXES. 273 ^ a." a- a.' Oxyhemoglobin of Cania dingo. Orthorhombic: Axial ratio a : b : 6=0.6009 : 1 : 0.2582; a : 6 =1 : 0.4296. Forms observed: Unit prism (110), macrodomes (101), (504), (302); basal pina- coid (001). Angles: Unit prism angle 110AlT0=61° 30'; macrodome 101aT01=46°; macro- dome 504 A 504 =56° 30'; macrodome 302 A 302 =66°, about (65° 36' calculated). Habit prismatic on the vertical axis, the crystal consisting of the unit prism (110) terminated by a zone of macrodomes (101), (504), (302) (text figure 306), of which either one may predominate, but the common ter- mination is the unit dome (101) (text figure 307). The crystals are normally much shorter than is common in this genus and vary in ratio of length to thickness from 20 : 1 to 4 : 1 or even less. The normal crystals usually range between 20 : 1 and 15 : 1. Along the protein ring and the cover edge the crystals are much larger and stouter, and are vertically striated, due to parallel growth. Groups of two crystals united on the brachy- pinacoid in parallel growth are particularly common among these larger crystals, but the majority of them are more complicated groups. The greater part of the crystals occurring in the body of the slide are simple and unstriated, and doubly terminated. They appear to twin upon a pyramid face as contact twins, but do not unite in spherulitic or radiating groups, as is common in the dogs. Fleochroism is marked when a and c lie in the section, but weak when h and c are in the section, indicating that b and c are near together in index, but a is rather far from either. The pleochroic colors are: a pale rose; b deep rose, c deep blood-red. Inordinary (unpolarized) light the color of the crystals is the usual oxyhemoglobin red. Looking along a the double refraction is quite weak, but in other aspects it is moderately strong. Extinction is straight in all aspects. On the fiat of the prism, looking along a the biaxial interference figure is seen, with not very widely separated brushes. The orientation of the elasticity axes ia a=a;i=b; c=d. The plane of the optic axes is the brachypinacoid. The acute bisectrix of the optic axes is the axis of greatest elasticity, Bxa=(i, and the optical character is hence negative. 306 307 308 109 Flos. 306, 307. CanU dingo Oxyhemoglobin. Fios. 308, 309. CanU azara Oxyhemoglobin. Azara's Wild Dog, Canis azarce. Plate 76. The specimen of blood was received during the summer from the New York Zoological Park, and was kept frozen in the refrigerating plant until examined. The blood was very putrid and was practically a mass of crys- tals when the collecting tube was taken from the cold storage. The blood was mixed with an equal volxmie of a 50 per cent solution of egg-white and subjected to action of oxygen until thoroughly saturated and the color had changed to bright red. It was then centrifugalized for 2 hours, and the sUde preparations made as usual. Crystals of oxyhemoglobin formed very readily and showed no signs of dissolving. About 20 hours after the slides were prepared the photomicrographs were made. Spectroscopic examina- tion showed the presence of reduced hemoglobin in the solution, but the crystals showed only the spectrum of oxyhemoglobin. 18 274 CRYSTALLOGRAPHY OP THE HEMOGLOBINS Oxyhemoglobin of Canis azarce. Orthorhombic: Axial ratio a : 6 =1 : 0.4328 from (706) or 1 : 0.4348 from (101). Forms observed: Unit prism (110), macrodomes (101), (706), (304). Angles: The prism angle was not obtained. The most satisfactory angle was the macrodome 706 A 706 =53° 35'; the other macrodomes were 101 A TOl =47° measured, 46° 50' computed from (706) ; also the dome 304 A 304 =36°, which agrees almost exactly with calculation from dome (706), as a : <5 for this dome from (706) is 1 : 0.3246 and from 36° measured is 1 : 0.3249. Habit at first long capillary, later stout prismatic crystals form (text figure 309). The normal crystals are long prismatic (text figure 308), elongated on the vertical axis and longitudinally striated; as is usual in the genus Canis, the striations are produced by parallel growth. Double crystals, two prisms growing side by side in parallel growth and united on the brachypinacoid, are common in the normal crystals. Owing to this tendency to parallel growth, the crystals become flattened in the direction of the macro- pinacoid, and the square-ended aspect is the common one; the crystal on edge is less frequently seen. Cross-sections were not observed, so that the angle of the unit prism was not obtained. The capillary crystals grow in somewhat radiating tufts, but the tendency of all crystals, whether capillary or thicker, is to aggregate into masses in nearly parallel growth, so that quite large groups are formed, the crystals growing as though united on the brachjrpinacoid. The circular radiating tufts and spherulitic groups, usually seen in this genus, were not observed in this species. The crystals that formed after the solution came to an equilibrium, which were hence formed more slowly, were shorter and stouter than the larger crystals formed during the first crystallization. The ratio of length to thickness in the normal crystals may be taken at about 20 : 1 on the average; but in these later crystals it was often as low as 5 : 1 or even less. No definite twins were observed. The color of the crystals in ordinary light is the usual oxyhemoglobin red. Pleo- chroism is not very strong; the colors are: a pale pink; b and c nearly equal and ranging from pale to deep red. The double refraction is fairly strong, and the extinction is straight in all positions. The orientation of the elasticity axes is a=a, B=&, c=i. The inter- ference figure was not observed, but the pleochroism and the double refraction indicate that the acute bisectrix is the axis of greatest elasticity, Bxa=Ci, and the optical character is hence probably negative. Swiss Fox,* Vtdpes vulpes (?). Plate 78. The specimen of blood was received from the National Zoological Park at Washington, District of Columbia. The blood was very thick and clotted and quite putrid. The clots were destroyed by grinding in sand and the ground mass was diluted with twice its volume of a 50 per cent aqueous solution of egg-white, 1:1; the mixture was then centrifugalized, and from the clear solution thus obtained the slides were prepared as usual. Decomposition continued, however, and considerable granular matter sepa- rated, due to breaking down of the materials in the solution. Crystalliza- tion proceeded rapidly and well-formed crystals of oxyhemoglobin formed along the protein ring and the cover edge, as well as throughout the body of the slide. After 24 hours many of these crystals had passed into reduced hemoglobin by paramorphous change, and many were partly dissolved along the protein ring. Along the cover edge, the crystals were in rather good condition when the photomicrographs were made, about 28 hours * This may be'the swift fox, Vrdpes velox, but was marked " Swiss fox," which presumably would be the European fox, Vidpes wdpe*. OF THE DOGS, WOLVES, AND FOXES. 275 after the slides were prepared. No difference in form or optical characters was observed between the crystals of oxyhemoglobin and the paramorphs of reduced hemoglobin. Oxyhemoglobin of Vvlpes wipes. Orthorhombic: Axial ratio a : 6 =1 : 0.4245. Forms observed: Unit prism (110), macrodome (101). Angles: The prism angle could not be observed as no cross-sections were seen. A fair measurement of the macrodome angle gave 101 A 101=46°. Habit long prismatic on the vertical axis, the crystals ranging from long hairs, 1,000 times as long as thick, to more nearly normal crystals with the ratio of length to thickness varying from 40 : 1 to 15 : 1 (text figure 310). The larger crystals are verti- cally striated along the length, due to parallel growth; but they appear to grow together on the prism planes as well as on the brachypinacoid, so that the composite crystals, produced by parallel growth, are not so flattened on the macropinacoid as they often are in this genus. The hair-like crystals grow throughout the body of the slide and form felted masses of hairs, but do not grow in the spherulitic and radiating tufts, as is usual in this genus. The larger crystals grow in parallel groups, rather than radiating, and the tufts of crystals thus formed are only slightly divergent. Twins were not observed. Pleochroism was noticeable, the double refraction was distinct and extinction straight in all aspects. The orientation of the elasticity axes appears to be as usual in the dogs, a =0, b =&, c =(J. The optical character is probably negative. a- 310 Q' /A ^sjm 311 Fio. 310. Vvlpa vulpa Oxyhemoglobin. Figb. 311, 312, 313, 314. VtUpa fvlvut Ozyhemoglobin. Red Fox, Vvlpes fulvus. Plate 77. Examination was made of two specimens, both of fresh blood, the first specimen being obtained from a zoological garden and the second by purchase of the hAring animal from a collector. The first specimen was oxalated and repeatedly frozen and thawed and then ether-laked and centrifugaUzed; and from the clear solution the slides were prepared. The crystals obtained in this series of preparations were rather more perfect than those obtained by either of the other methods of preparation used. The blood from the second specimen was treated in several ways: (a) The whole blood was oxalated and ether-laked as in the regular method of preparation, (6) the corpuscles were centrifugalized and a mixture of three- fourths plasma and one-fourth corpuscles was oxalated and ether-laked; 276 CRYSTALLOGRAPHY OF THE HEMOGLOBINS both were then centrifugalized, and from the clear solution the slides were prepared. The blood crystallized very rapidly in all cases, and the slides soon became filled with crystals. These were seen to vary in habit some- what according to the method of preparation, but not more than the crys- tals varied in the same sUde, according as to how long after covering they appeared. In other words, the habit of the crystals was conditioned by the strength of solution, rather than by its composition. The first crystals to form are capillary, as a rule; later these grow thicker, or stouter crystals appear; but in the first preparation some well-formed tabular crystals finally made their appearance, which at first sight seemed to be different from the prisms. After a study of their angles and optical characters they were finally determined to be identical with the prisms. All of the crystals observed were oxyhemoglobin. Oxyhemoglobin of Vulpes fulvus. Orthorhombic: Axial ratio a:h: (5=0.6494 : 1 : 0.2824; a : 6 = 1 : 0.4348. Forms observed: Unit prism (110), macrodome (101), brachypinacoid (010). Angles: Prism angle 110AlT0=66°; macrodome 101 A 101=47°. Habit prismatic, elongated on the vertical axis (text figure 311), and the prism striated in the same direction, due to parallel growth; or tabular on the brachypinacoid, the crystal in the prismatic habit consisting of the unit prism with the macrodome, and, in the tabular habit, the same faces with the brachypinacoid (text figures 312, 313); but in some cases one pair of opposite dome faces much developed while the alternate pair are much reduced in size or even wanting, giving the crystal a monoclinic aspect (text figure 314). The first crystals to form are generally capillary; and in some of the preparations, especially in those that were diluted with plasma, nearly all of the crystals retained the relative dimensions of the capillary crystals, until they attained a length of more than 3 mm. As the crystals continue to deposit from the solution, stouter crystals appear, and these resemble more closely the usual crystals seen in the blood of the species of this genus. The thin prisms are usually single crystals until they attain large size; they grow in slightly divergent tufts or more rarely form groups radiating in all directions from a center; in very many cases the groups are so slightly divergent that the individual crystals appear parallel, and the group looks like a parallel growth. Single crystals become covered by smaller ones that are actually arranged in parallel growth; but they do not seem to grow together on the brachypinacoid and hence flatten on the macro- pinacoid; in some cases the groups may contain individuals in twin position on the prism. It is this tendency to form composite groups, with the vertical axes parallel, that produces the characteristic striation in this direction. Pleochroism is moderately strong when observed along b with a and c in the field; but when looking along a it is rather weak. The colors are a pale reddish, somewhat yellowish-red; h and c nearly equal and deeper red. The crystals are not highly colored, as they are slender. Double refraction is not very strong except when a and c are in the field; the extinction is straight in all aspects. Traces of a biaxial interference figure were observed. The orientation of the elasticity axes is a=a, b=b, e = 6; and the axis of greatest elasticity appears to be the acute bisectrix, Bxa=a; the optical character is hence negative. Blue ob Arctic Fox, Vvlpes lagopus. Plate 78. Two specimens of the blood of this species were received from the National Zoological Park at Washington, District of Columbia, one during warm weather and the other during the winter. The former was kept OF THE DOGS, WOLVES, AND FOXES. 277 frozen in the refrigerating plant until examined; it was in good condition, and evidently was collected when the animal was but recently dead. The second specimen was somewhat putrid and thick; and had probably been taken some time after death. Both had been collected in oxalate, in our regular collecting tubes. The two specimens were treated in the same man- ner; laked with ether and centrifugaUzed, and from the clear solution thus obtained the shde preparations were made in the usual way. The blood crystaUizes very readily, and within a few hours after the preparations were made the slides were in condition for examination. The stale blood crys- tallized rather faster than the fresh specimen, due no doubt to its being in a more concentrated condition. The portion of clear solution remaining in the tube after the slides were prepared from this stale specimen became a mass of crystals within 2 hours after the preparations were made. The crystals keep well and show no tendency to dissolve in the solution. They were oxyhemoglobin in the case of the second sample of blood, which was the one from which the measurements were obtained on which the crystal- lographic constants were determined. The other sample appeared to have been converted into the acid form of metoxyhemoglobin. Its crystallo- graphic constants were the same as those of the second specimen, so far as they were recorded, but the habit of growth of the crystals was somewhat different. Oxyhemoglobin of Vulpes lago-pus. Orthorhombic: Axial ratio a : e and the optical character is negative. Hemoglobin of Scalops aguaticus. Hexagonal, only observed in paramorphs after the oxyhemoglobin. The forms and angles are hence identical in the two substances. The reduced hemoglobin para- morph is produced by bacterial action. The bacteria enter at the basal depressions and frequently penetrate the crystal from end to end, which then becomes like a short hex- agonal bead with a central perforation. They work through the substance of the crystal and completely honeycomb it, but usually leave a shell of unaltered oxyhemoglobin on the exterior, including the pyramidal planes, but not the base, which is completely eaten away. In some cases the crystals were thus completely converted to reduced hemo- globin and the channels made by the bacteria were even repaired and filled up by re- crystallized hemoglobin, making quite perfect crystals. Pleochroism was very strong, a=e, nearly colorless, pale lilac; €=&>, deep purplish- red. The double refraction is strong and extinction straight. The axis of greatest elasticity is the vertical axis, w > e; hence the optical character is negative, the same as in the oxyhemoglobin. CHIROPTERA. Fox-bat oe Flting-fox, Pteropus medius. Plate 94. The specimen was received from the Philadelphia Zoological Gardens, and was in a putrid condition. It was oxalated, a little ether added, and preparations made as usual. The blood crystaUized readily, and the crys- tals did not appear to dissolve at first, but after a few hours they began to break down and by the next day had disappeared from the slides. The photographs were taken inside of 4 hours after the preparations were made. The crystals were oxyhemoglobin. The examination of the crystals was incomplete, owing to their disappearing from the slides so rapidly; and hence the crystallographic constants were imperfectly determined. Oxyhemoglobin of Pteropus medius. Monoclinic (or perhaps triclinic): Axial ratio a :b : i =1 :b : 1.2808; ^=56° 30'. Forms observed: Unit prism (110), orthodome (TOl), clinopinacoid (010), base (001). Angles: The prism angle was not obtained. Hemiorthodome to prism edge or orthopinacoid TOl A 100=50°; hemiorthodome to base 101 A 001 =73° 30'; ortho- pinacoid to base (or prism-edge to base) 100 A 001 =56° 30'=^. 302 CRYSTALLOGRAPHY OF THE HEMOGLOBINS Habit generally tabular on the clinopinacoid (text figures 363 and 364) ; also short prismatic to tabular on the base. The planes seem to be irregularly developed, and may represent triclinic symmetry. The cover-edge crystals are elongated, apparently on the vertical axis, and generally flattened on the plane of symmetry; they grow crowded together, but as irregular aggregates, not apparently twinned. Pleochroism was only observed on the clinopinacoid aspect, and is a pale yellowish, nearly colorless; c deep red. Double refraction is strong; extinction is nearly or quite parallel to the prism edges. The plane of the optic axes appears to be the plane of sym- metry, or parallel to the plane taken as the clinopinacoid. The optical character can not be determined because of insufficient data. FioB. 363, 364. Pteropiu mediwi Oxyhemoglobm. Fios. 365, 366. Vetpertilio futeua Oxyhemoglobin. Brown Bat, Vespertilio fuscus. Plate 95. The specimen was bled in the laboratory. The few drops of blood ob- tained were caught in oxalate, and ether-laked. The quantity was not enough to centrifugalize. The slide preparations were made with the laked blood in the usual manner. Crystallization proceeded rapidly, and the crystals appeared to be rather insoluble, keeping well and showing no tendency to dissolve. Crystalhzation was quite complete, but Uttle color remaining in the solution. The crystals were shown to be oxyhemoglobin by the microspectroscope. Oxyhemoglobin of Vespertilio fuscus. Monoclinic: Axial ratio not determinable; /?=81°. Forms observed: The three pinacoids only, orthopinacoid (100), clinopinacoid (010), base (001). Angles: Base to orthopinacoid, 001 A 100=81°=|9; clinopinacoid, 010 A 100=90°. Habit tabular on the base, and elongated on the clino-axis, producing broad lath- shaped crystals (text figures 365 and 366), with a ratio of length to width on the base of about 8 : 1 to 5 : 1. The tabular crystals are thin, the thickness is one-twentieth of the length or less. They usually grow singly or sometimes in parallel growth on the base. In some cases the plates pile up on the base into parallel groups or bundles of crystals. No evidence of twinning was observed. Pleochroism moderate; a pale yellpwish-red, 6 rather deep red, c very deep red. Double refraction fairly strong, extinction straight on the basal aspect; on the clinopina- coid aspect the extinction is 16°, measured from the basal edge. The orientation of the elasticity axes is a A violet. ,,^^ 406 FiQB. 406, 407. Papio anubit y-Oxy- hemoglobin. jr-Oxyhemoglobin of Papio anubis. Orthorhombic: Axial ratio a :b : 6 =0.3268 : 1 : TnnEf|f>h!Ti or the barndoor Skate {Raia Icevis), showing sheaf-shaped crystal aggregates seen z, 6. i?anie, seen iii uiie nat aspect. 4. Same, showing twinning. 5, 6. Oxyhemoglobin of the Sturgeon {Acipenser sturio), showing brachypinacoid aspect. PLATE 2 7. Oxyhemoglobin of the Sturgeon {Acipenser sturio), showing general view of smaller crystals. 8. Same, showing brachypinacoid aspect in large cr_ystals. 9. Oxyhemoglobin of the Shad (Alosa sapidissima) , shovnng aggregates produced by twinning, horse-type. 10. Same, showing large tT\-in aggregate, horse-type. 11. Ox3'hemoglobin and Methemoglobin of the Shad, shomng twins of oxyhemoglobin. 12. Same, showing star-shaped twins of horse-type. PLATES 17 13. Metoxyhemoglobin of the Shad (Alosa sapidissinia), showing aggregate produced by parallel growth and twinning. 14. Same, showing single crystals and twins. 15. Same, showing aggregate group of twinned crystals in various aspects. 16. Same, showing parallel growth and twinning. IS. Reduced Hemoglobin of the Shad, showing various aspects of simple crystals. PLATE 4 ■11 ^- f.i .' ,■..•.,; ife ■,! ■',»■; %■< ■.■■»*• ,.-■ 22 19, 20. Reduced Hemoglobin of the Shad {Ahmo sapidissimo), showing simple crystals. 21. Oxyhemoglobin and Methemoglobin of the Shad, showing basal aspect of regular growth. 22. Oxyhemoglobin and Methemoglobin of the Shad, showing edge aspect of a crystal of oxyhemo- globin inclosed by methemoglobin in regular growth and twinned. 23. Oxyhemoglobin and Methemoglobin of the Shad, showing the two substances crystallized separately and also in regular growth. 24. Reduced Hemoglobin and Methemoglobin of the Shad, showing separate crystals, not in regular growth; also prismatic cleavage of reduced hemoglobin. PLATE 5 25. Metoxyhemoglobin of the Carp (Cyprinua carpio), shoTving sheaf-like aggregates. 26. Same, showing parallel growth of plates, in polarized light. 27. Same, showing irregular aggregates, produced by piling up of plates. 28. Same, showing parallel growth of tabular crystals. 29. Reduced Hemoglobin of the Carp, showing long prismatic crystals with acute macrodome (401), anc: also tabular crystals. 30. Same, showing prismatic crystals growing in sheaf-shaped tufts. PLATE 6 Reduced Hemoglobin of the Carp (Cyprinus cnrpio), showing smaller tabular crystals as appear when they first begin to develop. 32. Same, showing larger tabular crystals aggregated in parallel growth. 33-36. Oxj'hemoglobin of the Necturus {Nectunis maculatus), showing normally developed crystals, s twinned. \^0 PLATE 7 37. Oxyhemoglobin of the Necturus (Xecturiis tnaculatiis), showing long prismatic type of crj'stal. 38. Same, showing short prismatic development. 39. a-Oxyhemoglobin of the Python (Python molurus), showing habit (b), the talxilar crystal in flat view, and oblique section of crystal. 40. a-Oxyhemoglobin of the Python, habit (b), showing parallel growth on base. 41, 42. a-Oxyhemoglobin of the Python, habit (b), showing flat and edge views. PLATE 8 ,^' ;#WM-:^^ i \^ /■v / 43, 44. /5-Oxyhemoglobin of the Python (Python molurus), showing small pyramidal crystals as second crop on a-oxyhemoglobin crystals. 45, 46. Oxyhemoglobin of the Ostrich {Struthio camelus), showing small rhombic tabular crystals. 47, 48. Oxyhemoglobin of the Cassowary {Casuarius galecdus), showing prismatic crystals (dark), with crystals of ammonium oxalate (white). PLATE 9 50 49. Oxyhemoglobin of the Goose (Anser nnser), showing chisters of tabular crystals, twinned in zone of pyramid. 50. Same, showing aggregates like figure 49, but with more individuals, producing rosette-shaped group. 51, 52. Same, tabular crystals seen on the basal asjiect. 53. Oxyhemoglobin of the Trumpeter Swan (Olor huccinator), showing first-formed simple crystals. 54. Same, showing large irregular aggregate of later growth. PLATE 10 .-s^»s^:^!srr?r?3 i-Vf/ 59 60 55. Oxyhemoglobin of the Trumpeter Swan (Olor huccincdor) , showing hirge simple crystal. 56. Same, showing arborescent aggregates. 57, 58. O.xyhemoglobin of the Whistling Swan (Olor Columbian us), showing large twinned crystals in flat and edge aspects. 59, 60. Same, showing large composite crystals as seen on basal aspect. PLATE 11 / 66 61-63. Oxyhemoglobin of the Chicken (Gallus domestica) , showing tabular composite crystals. 6-t. Same, showing single twinned crystal. 65, 66. Same, showing composite crystals in groups showing basal and edge aspects. PLATE 12 2 67, (58. Oxyhemoglobin of the Quail (Colinus virginiamis), showing same composite tabular crystal in two different stages of development. G9. Same, shoOTng single and composite tabular crystals. 70, 71. Oxyhemoglobin of the Guinea-fowl {Numida meleagris), showing small simple crystals. 72. Same, larger crystals, showing twin on pyramid. PLATE 13 74 73. Oxyhemoglobin of the Pigeon (Cohimha livia), showing comparatively simple twinned aggregate, in the basal aspect. 7-t. 75. Same, showing more complex aggregates, some on edge or inclined. 76. Metoxyhemoglobin and Oxyhemoglobin of the Pigeon, showing disintegration of oxyhemoglobin crystals as metoxyhemoglobin crystals develop. 77, 78. Metoxyhemoglobin and Reduced Hemoglobin of the Pigeon, showing the neeille-like crystals of reduced hemoglobin growing in tufts from crystals of metoxyhemoglobin. PLATE 14 79. Oxyhemoglobin of the Crow {Corvus americanu.?), showing single, nearly eijuidiiiiensional crystal attached to cover by false plane near (100). 80-82. Same, showing groups of crj^stals presenting various orientations. 83. Reduced Hemoglobin of the Crow, showing group of elongated crystals. 84. Same, showing usual tabular crystal. PLATE 15 85. Oxyhemoglobin ami Rpduced Heninglobiii of the Crow {Cnrvus americariiia}, showing; regular growth of reduced hemoglobin on oxyhemoglobin, in basal aspect. S6. Oxyhemoglobin and Reduced Hemoglobin of the Crow, showing regular growth of reduced hemoglobin on oxyhemoglobin in edge and basal aspects. Reduced hemoglobin crystals seen on end.s of oxy- hemoglobin crystals in parallel position. 87. a-Oxyhemoglobin of the Opossum (Didelphis virginiana), showing small untninned crystals. 88. Same, showing large and small crystals. One large crystal shows unsynmietrical development of the hemiorthodome. 89. Same, showing sini[)le crystal (100) (001). 90. Same, tabular crystals in parallel growth, showing hemiorthodome. PLATE 16 ^ "^^^^^^^^^ "^l" •4- ^^^. ■ >C4 95 91. ci-Oxj'hemoglobin of the Opossum (Didelphis virginiana) , large crystal showing parallel growth and unsjiiimetrical development of heniiorthodome. 92. Same, showing large simple crystal and horse-type twin. 93. Same, showing twin on heniiorthodome and unsjTiimetrical development of this dome. 94. Same, showing unsymmetrical crystals and a twin on edge. 95. Reduced Hemoglobin of the Opossum, showing thick and thin tabular crystals and parallel growth. 96. Same, showing Reduced Hemoglobin with crystals of oxyhemoglobin. PLATE 17 97. |3-OxyhenioKlohin of the Opossum, showing tabular crystals in various orientations. 98. bame, showing large composite crystal. 99. Same, showirig basal aspect of tab{ilar crystals with a crystal of a-oxvhemoglobin in partial orien- ,n« ^ tation mth it, also edge yiew of a larger crystal of the 3-oxvhemosIohin 100. hame, showing single large tabular crystal. 101, 102. a-CO-Hemoglobin of the Opossum, showing thin tabular crystals in different aspects. ^^^^^^* PLATE 18 103. Oxyhemoglobin of the Tasmanian Devil (Sarcophiliis ursinus), showing tabular and short pris- matic types of crystals. 104, 105. Same, showhig short prismatic crystals with slender, long prismatic type. 106. Same, showing different types of prismatic crystals, some twinned. 107. Reduced Hemoglobin of the Spotted Dasyure (Dasyurus maculatus), showing prismatic and tab- ular types of crystals. 108. Same, showing prismatic type of crystals. PLATE 19 iUy. Oxyhemoglobin of the Australian Cat {Dasijtiru-s riverrinus), showing larger crystals, tabular on two opposite prism faces. V-shaped twin seen in one crystal. 110. Same, showing crystals in different aspects. Ill, 112. Same, showing very much flattened crystals and V-shaped twin on prism. 113. Same, showing symmetrical and distorted prisms. 114. Same, showing group of long prismatic crystals from protein ring. PLATE 20 ^'^*^i>0imw? i^li5#SS;^ 117 «r 119 "Ha ^^^ 120 115. a-Oxyhemoglobin of the Tasmanian Wolf (Tlujlnctinus cytJnccphalus), showinu; groups of plates in parallel growth. 116. Same, showing large tabular crystals, presenting basal and oblique aspects. 117. Same. shoAving group of crystals along protein ring, some showing edge aspect. 118. /3-Oxyhemoglobin of the Tasmanian Wolf, showing small dodecahedral crystals. 119. Same, showing dodecahedron in combination with the cube. 120. Same, showing large dodecahedra flattened by the cover and some in combination with the cube. PLATE 21 121. Oxyhemoglobin of the Vulpine Phalanger {Tricliosurus rulpeculu), showing tabular habit. 122. Same, showing short prismatic habit. 123. Same, showing long prismatic crystals and edge views of plates growing from cover edge. 124. Same, showing stout crystals in region of cover edge. 125. Oxyhemoglobin of the Rat-kangaroo {Jlpyprymnun nifescens) , showing smaller crystals, many in twinned position. 126. Same, showing medium-sized crvstals, some twinned on the orthodome. PLATE 22 127. Oxyhemoglobin of the Rat-kangaroo {.iipyprymnus rufeticens), large crystals growing from cover edge and some showing interpenetrant twin on prism in the flattened crystals. 128. Same, showing gypsimi type of twin. These large crystals are flattened between cover and slide. 129, 130. Oxyhemoglobin of the Kangaroo {Macropu^ giganieus), showing lath-shaped crystals growing singly and aggregated into sheaf-like bundles, lol. Same, showing stellate group of crystals along cover edge. 132. Same, showing larger lath-shaped crystals of parallel growth. PLATE 23 133. Oxyhemoglobin of the Rock-kanfiaroo {Petrogale sp.), showing short lath-shaped crystals. 134, 135. Same, showing long and short lath-shaped crystals and capillary crystals. 136. Same, large crystals of long, lath-shaped type, showing parallel growth. 137. Oxyhemoglobin of the Ant-eater (Mi/rmecophaga ?), showing capillary crystals growing from pro- tein ring and large lath-shaped crystals. 138. Same, showing large lath-shaped crystals gromng from cover edge. PLATE 24 i:?9. rt-Oxvliemoglobin of the Horse {Equus cahaUiif!). showing group of longer prismatic crystals (preparation without oxalate). 140. Same, showing long prismatic crystals growing from protein ring (preparation without oxalate). 141. Same, showing shorter prismatic crystals near protein ring (preparation with oxalate). 142. Same, showing twin on pyramid (preparation with oxalate). Some crystals of the 3-oxyhemo- globin are seen near edge of field. 143, 144. Same, showing stout prismatic crystals along protein ring and tufts of first-formed capillary crystals (preparation with oxalate). PLATE 25 . ' 146 MSJ ^ ^ d,. ^ ' 145. /3-Oxyhemoglobiii of the Horse (Equux caballux). .showing horse-type twiiis consists of four individuahs, that on the left above, of two individuals. 146. Same, showing small composite horse-type twins. 147. Same, showing large com|iosite horse-type twins, some seen in edge view. 14S. Same, showing irregular composite horse-type twins. 149. Same, showing large composite twinned group and elongated horse-type twin on rigid 150. Same, showing single crystals and large twinned group, with smaller twins on edge. Twin on the right above PLATE 26 151. c(-Oxyhemou;lol)in and ,5-Oxyhemo,2;lol)in of the Horse (Eqiius cahallu-f). showing the two in large crys- tals. The a-oxyhemoglol)in shows twin on pyramid, the (9-oxyhemoglobin shows usual horse- type twin. l.')2. /?-Oxyhemoglobin of the Horse, showing large twinned group. Large crystal above to right shows orthopinacoid on one end. ].■>:?. «-C'0-Hemoglobin of the Horse, showing long prismatic crystals growing from |jrotein ring. l.")4. Same, showing stout prismatic crystals growing from protein ring and also capillary crystals. \55. ;?-C'0-Hemoglobin of the Horse, showing composite twins of horse-type in various aspects. 156. Same, showing elongated horse-type twin. PLATE 27 >^' " -VJI 157. a-Oxyhenioglobiii of the Mule {Equus asinu.s J X Equm Cdhalhix ' ), showing short prismatic crystals. 158. Same, showing long prismatic crystals growing from i^rutein rmg. l.)9, 160. Same, showing large stout prismatic crystals growing along cover eilgc. 161. ;9-Oxyhemoglobin of the Mule, showing equidimensional development of crystals, mostly un- twinned. 162. Same, showing elongated horse-type twins. PLATE 28 |:> 163 164 3-Oxvliemoglobin of the ^Fule (Eqini.t asiniis S X Equus cabalhi.i I), showing a large single cryslal. Same show-ing large horse-type twin with composition face normal to base and including common 163 164. _ - prism-base edge. 165. Same, showing large horse-type twin in various aspects. 166. Same, showing smaller horse-type twins. Hi7. Same, show-ing large elongated horse-type twins. 168. Same, showing large horse-type twins. PLATE 29 169 yc ^ 169. /?-Oxj'hemoglobin of the Mule (Equus asinu.s c? X Equu.s cnhallua i ), showing composite twnn of horse-tyi)e. 170. Same, showing prismatic and tabular types of crystals. 171, 172. ,3-CO-Hemoglobin of the Mule, showing composite horse-type twins ami simple crystals, normal type. 173. Same, showing elongated horse-type twin. 174. a-C'0-Hemoglobin and ,3-CO-Hemoglobin of the Mule, showing ends of n-CO-hemoglobin crystals growing from protein ring, and elongated horse-type twin of ;?-C'0-hemoglohin. PLATE 30 175 Oxvhemo<^lobin of the Hippopotamus {Hippopotamus amphibim). showing smaller prismatic ' ' ' untwTnneil crystals, some tabular, some resembling rhombohedra and some longer prismatic. 176. Same, showing smaller prismatic type of crystals, some twinned. 177! Same! showing trapezoidal form of horse-type twin. 17S Same, showing large twinned aggregates. 179 ISO. Same, showing apparently hemimorphic crystals produced l>y elongated twins of trapezoidal horse -type. PLATE 31 181 r 183 '? J& ISl, 1.S2. Oxvhemugldbin of the Hippopotamus {Hippnpnfnmiix nmplilhhiK), showiiif; firo\ips of large twinned crystals. 183. Same, showing single large twinned group of horse-type twin. 1,S4. 8ame, showing twin aggregates and ladder-like form produced \>y twinning. 18o, ISti. Oxyhemoglobin of the Peccary (Dicotylex lahiatii.s). showing dodecahedron-like combination of short diametral prism with unit pyramid. PLATE 32 189'^ ^^^ % .««' a 1S7, 1S8. Oxvhenioglohin of the Collared Peccary (Dicoli/lex fajacii), showing small pjramidal crj'slals. 1S9, 190. Oxvliemoglobin of the Pig (Shs scrofa), showing single crystals consisting of unit prism and brachydome. growing in i)rotein ring. 191. Same, showing large crystals in various orientations. 192. Same, showing grouj) of crystals and aggregates from protein ring. PLATE 33 19:!. (Ixyliemoololiin ol' the Pig; (Sus srro/«). showing small crystals from |)rotein ring. 1!) 1-190. Reduced Hemoglobin of the Pig, showing long prismatic crystals, some terniiTiated by base, growing in sheaf-like grou|-)s and sometimes twinned. 197. Same, showing thick and thin prismatic crystals at basal termination. 19S. Same, showing feathery groups of crystals. PLATE 34 ^.^$' «^^ i»% ^ .?v;sp; .^.' #: r« c- <^ <* .ri 1!)9. Oxyhemoglobin of tlie Muis Deer (Tragtdus ineminna), showing single tubular crystals. 21)1), 201. Same, showing horse-type twin. 202. Reilnced Hemoglobin of the Muis Deer, showing part of very large crystal. 203, 204. Oxyhemoglobin of the Elk (Cervus canadensis), showing simple pyramidal form of crystals. n PLATE 35 f » « m^ ir 206 205, 206. Oxyhemoglobin of the Elk {L'ervus canadensis) , showing flat pyraniiil in three aspects, plan and side elevations. 207, 20S. Reduced Hemoglobin of the Red Brocket (Cariacxt-s rufus), showing composite tabular crystals of parallel growth; also narrow lath-shaped crystals. 209. Same, showing simple tabular crystals and lath-shaped crystals. 210. Same, showing rods growing in sheaf-like tufts. PLATE 36 Jlil, z 1 L '':mi'/j-p' mm <0 ^ 214 a^ •^'^ ?!. , .? i:^*^ f I?-. c "<, ^•'i^s;:^ ^ 213 ■'•^,4 211. Reduced Hemoglobin of the Venezuela Deer [Moihiiki nincricunu sniyiniiornni), showing diver- gent group of narrow lath-shaped crystals. 212. Same, showing group of broader tabular crj-stals. 21i). Same, .showing edge and flat views of broader crystals. 214. Oxyhemoglobin of the Fallow Deer (Cerrtis dama), showing simple, pyramidal crystals. 21.1. 216. Oxyhemoglobin of the Muntjak (Ccrridiis muntjak). showing prism terminated oliliquely by base. E 37 217. Reduced Hemoglobin of the Muntjak {C'ervidus muntjak), showing latli-shaped crystals. 218. Same, showing hair-hke crystals in feathery groups. 219-222. Oxyhemoglobin of the Indian Antelope {Antilope cervicapra), showing long prismatic crystals terminated bv base. PLATE 38 223. Oxyhemoglobin of the Reduncu or Xagor (Cervicapra reduncu), showiiifi long prismatic crystals oi' the first crop, growing from protein ring. 224. Same, showing short crystals of second crop developed in protein ring. 225. Same, showing cross-sections of the prismatic crystals. 226. Same, shorter prismatic crystals of second crop, showing macrodome termination. 227. Same, showing large tabular crystals of second crop, flattened on brachypinacoid. 22S. Same, showing large prismatic crystals of second crop, Hatteiied on brachypinacoid. PLATE 39 V fit .^mr^*^- -^^ 220. Oxylienioglobiii of the Dorcas Gazelle (Gazella dorcas). showing small first-fonueil crystals in different orientations. Prism and dome can be seen in many crystals. 230-232. Same, showing larger crystals growing from the protein ring, presenting brachypinacoid and edge Niews. 233, 234. Same, showing large tabular crystals in parallel growth. PLATE 40 235 r %: -■ «-:^ 235-2:57. n-0\yliemoglobin of the Duickerbok (Ccpliidopliiis qrimmi). showing pimple, pyruiniilal crystals. 238. /3-Oxyhemoglol)iii of the Duickerbok, showing hexagonal plates on flat and on edge, also several groups of u-oxyhemoglobin crystals in parallel growth. 239. Oxyhemoglobin ofthe Sheep (Oris aries). showing first-formed needle-like crystals and small tabular crystals. 240. Same, showing long prismatic crystals that develop after 24 hoiirs. PLATE 41 :-r3^ '*■ -.- S.^to-1^ 241. Oxyhemoglobin of the Sheep {Ovis aries), showing network produced by twinning and beginning of pentagon twins. 2-J2. Same, showing isolated pentagon twins in side and edse views. 243. Same, showing composite groups produced by twinning. 244. Same, showing cross-banded effect on pinacoid produced by twinning. 245. Same, showing isolated pentagon twins strung like beads on needles of oxalata 246. Same, showing pentagon twins seen in polarized light. PLATE 42 247, 248. Reduced Hemoglobin of the Sheep (OvU arie.s), showing tabular crystals in various orientations. 249-251. Oxyhemoglobin of the Bharal (Oris nahiira). showing long prismatic crystals, many flattened on two opposite prism faces, thus producing a triciinic aspect. 252. Same, showing flattened prisms shortened to almost rhombic plates. PLATE 43 /m^W i ^;\., 253. Oxyhemoglobin of the Bharal Ifjvu nn/iiira), showing network of prisms produced Ijv twinnin". 254. Same, showing barred effect produced on prisms by twinning. 255, 256. Same, showing isolated pentagon twins in various as|)ects. 257. 25S. Same, showing jientagon twins growing on antl capping prisms. PLATE 44 259. Oxyhemoglobin of the Bullock (S0.5 taunis). sho-ning small, hrst-formed crystals consisting of unit prism and brachyilome. 250. Same, showing larger prismatic crystals with unec(iially developed dome faces. 261. Same, sliort stout crystals, some showing hracliyprism in combination with unit prism. 262. Samp, showing group of crystals growing attached to an oxalate crystal. 263. Oxyhemoglobin of the Bison (Bos hison), showing irregular aggregate of thin lath-shaped crystals in protein ring. 264. Same, sliowing long crystals growing in tufts from protein ring. PLATE 45 265-269. n-(~)xylieinoglobin of the European Red Squirrel (Sciurus rulgnns), showing hexagonal tabular crystals, many in parallel growth. 270. 9-Oxvhemoglobin of the European "Red Squirrel, showing composite crystals. PLATE 46 271-274. Oxyliemo<;lobin of the Fox-squirrel (Sciurus rufirenter neglectus), showing single crystals and groups in parallel growth. 275. Oxyhemoglobin of the Gray Squirrel (Sciurus carolinensis), showing p.seudohexagonal single crystals and groups in piarallel growth. 276. Same, showing large single crystals with smaller groups attached in parallel growth and radiating, producing a sheaf-like appearance. PLATE 47 r^^'^- «*»> ■9- 3.' , 11- ' *■ " ''i.W- ... « • » IP - » 1 • • _ V^ 280 4 O'i, *--^ "*282 277. Oxvhenioirlobin of the Gray Squirrel {Sciurus carolintmis), showing "eisen rose" groups and sphenehtic masses of crystals in protein ring. 27S. Same, large crystals shomng parallel growth. 279-2S2. Oxyhemoglobin of the Flying Squirrel (Sciiiropterus volans), showing small, simple hexagonal crystals characteristic of the species. PLATE 48 283 •• / 284 2S3, 2S4. Oxvhoinoslobin of the Ground-squirrel (Tamia3 striatn-i), showing irregular groups of crystals growing in protein ring 285 286. Same showing long crystals growing in radiating groups from protein rins. 287. O-xvhemoglobin of the Prairie Dog (Cynnmi/.i ludnricianiK). showing divergent tufts of liair- like crystals, growiim from cover cilge. 288. Same, showing divergent groups of larger acicular crystals, seen in polarized light. PLATE 49 289. (S-Oxyhemoglobin of the Ground-hog (Marmota monax), showing radiating tufts of hair-like crystals. 290. n-Oxyliemoglobin and /3-oxyhemoglohin of same, showing t>ifts of ,3-oxyhemoglobin and small hexagonal plates of first-formed a-oxybemoglohin crystals. 291. Same, showing single large crystal with groujis of smaller crystals on it in parallel growth. 292 293. Same, showing large simple and composite crystals, the parallel growth preserving general hex- agonal outline. 294. Same, showing irregular aggre'gate of hexagonal plates, all in parallel growth. PLATE 50 295. a-Ox}-hemo<;lol)in of the Ground-hog {Mnrmota ninnaj:). showing two groups of hexagonal plates in partial orientation, looking complete as seen on base, and one group seen in side view, showing radiating character due to partial orientation. 29(3. Same, showing larger group in this partial orientation, as seen on base and on edge. In these two figures the lath-shaped section of plates is illustrated. 207. Same, a \'ery complex grou]i in edge -view, showing arborescent form of group. 29S. )-OxYhemogIobin, showing arborescent and feathery forms of groups of crystals. 299. Same, showing simpler group of crystals on base and on edge. 300. Same, showing tninned group of crystals in polarized light. In 299 and 300 the imperfect curved '•':./_ :' crystals are sho\\^l. PLATE 51 301. Oxyhemoglobin of the Beaver {Castnr canailensis), showing rhomboidal and hexagonal tabular crystals, on base and edge. 302. Same, showing four and six-sided tabular crystals in different aspects and small hexagonal plates of mimetic twin. 303. Same, large hexagonal plates, showing twinning and parallel growth. 304. Same, elongated crystals showing twinning. 305, 300. Oxyhemoglobin of the Muskrat (Fiber zibethicus). showing needle-like first-formed crystals with elongated horse-type twins of lath-shaped crystals. PLATE 52 307-309. Oxyhemoglobin of the iluskrat {Fiber zibeOttcus), showing lozenge-shapeil tabular crystals in different orientations and some horse-type twins. 310-312. Same, sliowing different kinds of crystal aggregates in parallel growth, arborescent grouping and sheaf-like forms being produced. PLATE 53 4|S -ca^;<7«S c;^^^ 318 313-315. Oxyhemoiflobin of the White Rat (Albuio of Mus norvegicus) , showing eiongate'l six-sided plate prodticei-1 by flattening of prism; rotighly hexagonal grotips due to twinning on brachy- pyramid and interpenetrant twins on unit pyramids. 316. Same, showing star-shaped twins on unit pyramid. 317. Same, shotting hexagonal composites with higher magnification. 31S. Same, showing star-shaped twin with higher magnification. PLATE 54 319. n-Oxyhemoglobin of the Norway Rat [Miis nonegicus), showing symmetrical ami flattened prisms and hexagonal plates produced by shortening of prism. Star-shaped twin on imit pyramid seen at two places in lower part of field. 320. Same, showing especially nearly hexagonal plates produced by shortening of flattened jirisni. 321. ,9-Oxyhenioglobin of the Norway Rat, showing .symmetrical and flattened octahedra in different aspects, with higher magnification. 322. n-Oxyhemoglobin of the Norway Rat, showing symmetrical and unsynimetrical ijrismatic crystals and pseudo-hexagonal jilates, with some hexagonal plates of ^-oxyhemoglobin due to flatten- ing of octahedron. 5!0'? -nA n,.,.i, — '""lobin of Black Rat {Mus riitlu.'!). showing flattened prism terminated by dome, with PLATE 55 325. Oxyhemoglobin of the Black Rat (Mus rattm), showing twin on the flat, consisting of two individ- uals and not producing a hexagonal plate as in White Rat twin. 326. Same, showing thicker crystals and oblique termination of dome faces. 327. Oxyhemoglobin of the Alexandrine Rat {ilus alcxandrhius), showing unsymmetrical flattened prisms and a twin on flat aspiect to lower left of field. 32S. Same, showing four-, five-, and six-sided tabular crystals, due to uns^^nmetrical development of dome faces. 329. Same, showing star-sha]ied twins. 330. Same, showing larger crystals. PLATE 56 331. n-Oxyliemoglobin of the Porcupine (Eretlrizon dursalus), .showing first-formed talnilar crystals. 332, 333. a-Oxylienioglobin and .5-t)xyhemoglobin of tlie Porcupine, sliowiiig bundles of elongated horse- type twins of a-oxyheniofflobin and thick tabular crystals of ,3-oxyhemoglobin. 334. ,9-Oxylienioglobin of the Porcupine, showing skeleton crystal apiiearance. 33.5, 33lj. 8anie, showing large crystals in various orientations. PLATE 57 -^ m ^v, >r|i> v^^^^^^ ^ 5¥ 1« ^ ^^ / ^ •J- V T f^^ ^^^ ri:57. 3-t)xylieniui;lohiii of tliu l^>icu|jiiie (Erethiznn dnrfotuf:), showing large crystals in Lasal aspect. o3S. Sanif. showing basal and edge views. .'«;), :U0. Oxyhenioglubin of the (hiinea-pig {('arid culli rl), showing small single crystals and twins of types a and f, :U1. Same, medimn-size crystals, some twimieil. showing tyjies n. Ii. and c. M2. Same, showing large ciystals with twins of types a and 6. 343 ^ 4 PLATE 58 ► ► N A A 344 4e^'l^ ■•■■0 fe -^ # ^ "^' Of) I 345 1/ ^^^^ ^' ,c« 0>^ o 343. Oxyheniofiloliin of the Guinea-pig (Caria ciitleri), showing single crystals and twins of types b and c. 344. Same, showing twin of type a in side ^iew to right center of field. 345-347. n-Oxyhenioglobin of the Capybara (Hydrnchtirux capyvaro), showing pyramidal crystals in various orientations. 345. .^-Oxyhemoglobin of the Capybara, showing long prismatic crystals. PLATE 59 ,mi 350 349. ci-Oxyheinoglobin of the Capybara (Hydroclucnis ccijiyvara). [ihotographed in polarized light with one nicol to show pileochroism. 350. Same, showing larger crystals. 351. Same, showing large crystals in ordinary light. 352. Same crystals seen in 351. in polarized light with one nicol showing pleochroism. 353. a-Ox3'hemoglobin of the Rabbit {Lepus cunicidus), showing prismatic type of crystaf. 354. Same, sho\\ing prismatic and tabular types of crystal. PLATE 60 357 358 359 355. n-Oxyhemoglobin of the EaLhit (I.tinis ciiiuculu.-i), showing prismatic (type a) crystals glowing from protein ring. 356. Same, showing prismatic (tyjie a) crystals in various orientations. 357-359. k>anie. showing tabular (type h) crystals. 360. u-C)xyhemogIobiu type a and ;?-Oxyhemoglobin of the Rabbit, showing jikochioism of fi-Uxyhenio- globin crystals in ordinary lisht. PLATE 61 ...(vi;'^"' 361. /3-Oxyhemoglobin of the Rabbit (Lepus cuniculus), showing oblique sections of crystals, 362. Same, showing oblique sections; crystals are pleochroic in ordinary light, 363, 364, Same, showing groups of large crystals in oblique section. 364 shows decided pleochroism, 365. a-Oxyhemoglobin of the Belgian Hare {Lepus europoeus), showing prismatic crystals of second crop, growing in radiating groups from protein ring. A few show single oblique termination. 366. Same, showing doubly terminated prismatic crystals from protein ring. PLATE 62 367, 368. ((-Oxyhemoglobin of the Belgian Hare (Lepiis eun)pausj, showing large prismatic crystals uf second crop. 369. Oxyhemoglobin of the Harbor Seal {Phuca vitulinu), showing first-formed, small tabular crystals. Their hemimorpliic character is easily observed. 370. Same, showing larger tabular crystals, some exhibiting parallel growth. Ilemimorphism is evi- dent in large central crystal. 371. Same, showing crystals elongated into prisms by development of orthopinacoid. 372. Same, showing simple symmetrical crystals consisting of base (001), unit prism (110), and unit pyramid (111). PLATE 63 .\ o73. Oxyhemoo;Iobin of California .Sea-lion (Otarin gillespii). showing small, first-formed crystals, mostly simple and untwinned. 374. Same, showing larger single crystal. Hemimorphic character well shown in this crystal. 375. Same, showing Sea-lion twin in edge view, consisting in this case of two indixiduals. •wb. Same, showing parallel growth and an edge view of Sea-lion twin in parallel growth. 37/. Same, .showing flat view of crystals in parallel growth. Cross-barring is due to twinning of a number of individuals in parallelpo.sition. 378. Same, showing parallel growth in twins seen on edge, producing comb-like appearance and cross- barring when seen on the fiat. PLATE 64 373. CO-Hemoglol:iin of the California Sea-lion (Otaria gillespv). showing small, first-foimeil crystals, twirnied. 350. Same, showing the crystals seen on edge and in section. 351. Same, large crystals showing parallel growth. ■i.S2. Same, showing Sea-lion twin in edge view, consisting of three nearly symmetrically develo|)ed viduals. ■■>^)5. Same, showing parallel growth. 384 Same, showing single large crystal in symmetrical position. Small crystal below to the lett ; profile view looking along ortho-axis. indi- hows PLATE 65 385 ./ 385. 3S6. 387. 388. 389, 390. Oxyhemoglobin of the Skunk [Mephitis mephitica putida), showing long prismatic cr\-stals j iiig in radiating groups. Same, showing long prismatic crystals and brush-hke tuft. Same, showing irregular aggregate of shorter prismatic crystals growing in protein ring. Same, showing large groups growing in parallel orientation. Oxyhemoglobin of the Ferret (Mmtela pulnriwi), showing tabular crystals and twins. PLATE 66 391, Oxyhemoglobin of the Ferret (Mustela putorim), showing small, first-formed somatic crystals in protein ring. 392, Same, showing larger tabular crystals in various orientations, growing in protein ring. 393-395, Oxyhemoglobin of the Otter {Lutra canadensis}, showing tabular crystals in different positions, many in twin position. Sphenoidal or heminior|ihic character may be seen in crystal on edge near middle of field in 39,3. 396, Same, showing thicker tabular crystals in basal and edge views; basal aspect shows parallel growth. -^■Ix PLATE 67 ""^m^^M mi '• ir'p:' 398 IJS* ^^ ^ <*. I*]/ t^' '4 ^3f ^ ^ ti. 47 399 \^> 1 v»S fi) ^ /> /*/ pO^- ^>5k ,#",£?^ vj«^^' \i\':^^ - ill. ^"./^ J'i^^ M^" ^^ 400 397. Oxyhemoglobin of the Badger {Taxidea amerwana), type n, showing small, first-formed tabular crystals, five-sided, and consisting of unit prism and base with one face of hemiorthodome. Badger twin shown in many of these crystals. 398. Same, type a, showing tabular crystals more symmetrically developed and four-sided. 399. Same, type a, showing more prismatic development of crystal consisting of prism, base, and hemior- thodome; also penetration twins. 400. Same, showing crystals of type a in short prismatic form and long prisms of type b growing from crystals of oxalate. ■iOl. Same, showing long prisms of type h growing in a tuft. 402. Same, showing radiating group of type h crystals growing from an oxalate crystal. PLATE 68 403, 404. Oxyhemoglobin of tlie Kinkajou (Cercaleples caiirientation, divideil by large crystals in obhqne section. 417, 41,S. Oxyhemoglobin of the Polar Bear (tVsiw maritimux), showing small, first-formed crystals tvdnned in bear-type twin, growing in protein ring. 419. Same, single crystals showing hemimorphisni and bear-type twins, near cover edge. 420. Same, larger tabular crystals along cover edge, showing parallel growth. PLATE 71 %' ^t Cc/ < V ;^: *£l.'-*f^#'^' 421, 422. Ox3'heinoglubin of the .Sloth Bear {Melwsus ursinii.s), showing small, hrst-furmed single crystals and bear-type twins. Combination is base cut by the unit prism at one end of ortho-a.xis and by unit pyramid at the other end, and with or without orthopinacoid. -123. Same, larger cry.stals, most of them showing orthojiinacoid. ■^24, 425. Same, larger second-crop crystals from along cover edge, a few showing bear-type twin. 426. Same, group of larger crystals from near cover edge, showing jjarallel growth and cross-sections of crystals. PLATE 72 427. Oxyhemoglobin of the Sloth Beur (Melursus ursiniis), showins group of crystals in parallel growth along ortho-axis. The white crystals are o.xalate. 428. Same, showing irregular groups in parallel growth orientation. 429. a-Oxyhemoglobiu of the Dog (Cania faituliaris), showing mass of capillary crystals jjroduced in thick slides. 4.">(). Same, showing striated jirismatic crystals growing from protein ring. 131, 4o2. .Same, showing divergent groups of long prismatic crystals. Plate 73 433. Oxyhemoglobin of the Chow Dog {Cams familiaris var.), showiiifi; ca])illary and long prismatic crystals doubly terminated. 434. Same, showing shorter prismatic crystals in divergent tufts and overgrown by sphenelitic groups of smaller prisms. 435. Same, .showing radiating cluster of larger prisms growing from |notein ring. 436. Same, showing large crystals growing from protein ring. 437. Oxyhemoglobin of the cross between Collie (Canis familiaris) and Coyote (Canis latrans), showing capillary and thicker prisms, several being the group of two and others more composite. 438. Same, showing groups of crystals, some probably in twin orientation. PLATE 74 439. Metoxyhemoglobin of the Wolf {Canis lii]>us mexkanus), showing lachating groups of crystals. 440. .Same, showing brush-like ends on some crystals. 441. O.xyhemoglobin of the Coyote (Canis latrans), showing mass of capillary crystals. 442. Same, showing short capillary crystals and single shorter composite crystal. 44.1 Same, showing radiating groups of crystals. 444, Same, showing mass of short needle-like crystals. PLATE 75 4' \ - . '^k 1 '^^i==:^^ i^" ^^%.;^/.,; ^^^K ^%W ^i^^..^. "Mii^-:^' 449 7 -- .45. oxyhemoglobin ^J'^l.'jae.al (Ca.. «^-'^)-^^^r^^" ^'^^ ^^'"""" "'^"^^ "^ ^"""■' crystals. Flexibility of crystals shown by cur^ a ures. 446. Same, showing capillary and shorter prismatic crystal. ^ 447. Same, showing shorter prismatic crystals along ote ' r | ,,1 ,enninated, com|.os.te cr>-.- aK 448. Same showing part of the protein ring that l^;'^"''^;^;";" prismatic doubly tcrnunated crystals, 449. Oxyhemoglobin of the Dingo (Cam.s ^''f ?"\' j^^;??' "."of X l ■ somein twin position ^s they occur through bod o U le ^ 450. Same, showing short composite prisms that de\elop 1 PLATE 76 ■lol. Oxyhemos'-obiii of the Dingo (Canis diiigu), sliowiiig stout and thin prisms. 452. Same, showing medium thick prisms along protein ring. 453. O.xyhemoglobin of Azara's Dog (Cam's azarm), showing divergent tufts of long capillary crystals. 454-456. Sariie, showing masses of crystals aggregated in approximately parallel growth along proteni ring or cover edge. PLATE 77 )c^' 9'^y''S'"oglobin of the Red Fox (Vulpes fulvua), showing radiating groups ot thin jirismatic crystals. •15S. Same, showing radiating groups of bundles of crystals, probably in twin orientation. ■io9. Same, showing mass of long prismatic crystals. j60. Same, showing bundles of slightly divergent crystals. T*?. "^S'lis. showing apt)ro.ximate |)arallelism of a large groii)) of the crystals. 462. Same, showing groups of crystals in approximately parallel growth. PLATE 78 40:!, 464. Oxyhemoslobin of the Swiss Fox (I'ulpcs nilpcs), showing long striated ]irisnis ami capillary crystals along protein ring. 465. Oxyhemoglobin of the Blue Fox {Viilpes lagnpus), showing cajjillary and stout |jnsniatic crystals. 466. Same, show-ing group of short stout prisms. 467. Same, showing long square-ended prisms. 468. Metoxvhcmoglobin of the Blue P'ox, showing radiating grouiJS of crystals in jirotcm ruig. PLATE 79 469. 470. 471, 472. 47.3, 474. Oxyhemoglobin of the Gray Fox {Urocyon cinereo-argenteu.s) , showing splienehtic groups of tliin prismatic crystals, an occurrence rarely seen. Same, showing large masses of crystals in parallel growth, a characteristic aggregate in this species. Same, showing radiating tufts that develop in retarded crystallizations. Same, showing shorter thicker crystals developed in undiluted preparations. These crystals formed in protein ring, and in 474 one of the small rare sphenelitic groujjs is seen. PLATE 80 475. Oxyliemoglobin of the Lion {l''elis leo), showing type a crystals occurring singly and exhibiting parallel growth. Short prismatic crystals in this |)late are reduced hemoglobin of hrst crojj. 4^6. Same, showing an irregular aggregate of type a crystals. 477-480. Same, showing crystals of type a and type b along cover edge. Characteristic form of type b crystals may be seen to lower left of field in figure 480. All figures show reduced liemoglol)in crystals. PLATE 81 481. Oxyhemoglobin of the Lion (Felis leo), showing group of type a, rhombic crystals, some exhibiting parallel growth. 482. Oxyhemoglobin and Reduced Hemoglobin of the Lion, showing rhombic plates of oxyhemogluljin and brush-hke aggregates of hemoglobin needles of second crop. 483. Same, showing large crystals of oxyhemoglobin showing parallel growth, embedded in tuft of crystals of reduced hemoglobin of second crop. 484. Oxyhemoglobin of Lion, showing large type a crj'stal with .smaller crystals on it in parallel growth. 485. Oxyhemoglobin and Reduced Hemoglobin of Lion, showing large type a oxyhemoglobin crystals on itaining small embedded crystals of reduced hemoglobin. Needle-like -.__ _ ■ .. ,. . :.;■;: .:.. ;d hemoglobin. of the short prismatic type; pleochroism very distinct. PLATE 82 487. Reduced Hemoglobin of the Lion (Felis leo), showing smaller, more normally developed first -formed crystals of third crop in various orientations. 4o8. bame, showing larger crystals of third crop. Traces of needles of secoml-crop hen.oglobin crystals may be seen in crystals at top of figure, penetrating the large crystals. 489. Same, showing one of these large, third-crop crystals penetrated by needles of second-crop crystals. 4J0. Same, various sections of these large, third-croj) crystals, showing traces of penetration by needles of second-crop reduced hemoglobin. Ann' J'^J'-'ced Hemoglobin of the Tiger (Felis tigris), showing small, first-crop crystals. r\f firot nmi-i coi^ia Hictr^rtin/I ]^\' o.vn\vtVi nr hv nnnti^ot with qIkIp nr pnvi^r Mif^' PLATE 83 493. Reduced Hemoglobin of the Tiger {Felis tigris), showing various aspects of medium-sized crystals. Angle of macrodome may be seen in crystal to left of center of field. 494, 495. Same, showing different orientations of medium-sized crystals. 496. Same, showing large crystals. 497, 498. Same, showing large crystals of second crop, along with small crystals of first crop. PLATE 84 • J# ~1^^^:-y^:*^^< ^> f ?\ 504 499. n-Oxyheaioglobin of the Jaguar (Felin oiica), two groups of tabular crystals showing parallel growth. <30U. Same, showing large group of crystals all in parallel growth orientation. Rod-like crystals are reduced hemoglobin. 501. Same, .showing three groups of tabular crystals, each in parallel growth orientation. ^02. Same, showing large grou]) in parallel growth orientation. o03. /J-Oxyheraoglobin of tlie Jaguar, showing dodecahedral crj'stals. ^04. Same, showing octahedral crystals. PLATE 85 505 «# f '^-9^ ii^ (L tl %# ^ ^ *^%^ % ^ 509 505. /3-Oxyhemoglobin of the Jaguar (Felis onca), showing octahedral crystals. 506. Reduced Hemoglobin of the Jaguar, showing long, square, four-sided prisms ending in a brush of fibers. "07, 508. o-Oxyhemoglobin of the Mountain Lion (Felis concolor), showing type a crystal, consisting of unit prism and brachydome. 509. Same, showing type 6 crystal. 510. Same, showing type 6 crystal, some passing into /?-Oxyhemoglobin. PLATE 86 516 511. Oxyhemoglobin of the Mountain Lion {Felts concohr), showing type c crystal. al-. /3-Oxyhemoglobin and a-Oxyhemoglobin of the Mountain Lion, showing type b crystal of a-Oxyhemo- globin twinned and passing into /3-Oxyhemoglobin octahedron. old. a-CO-Hemoglobin of the Mountain Lion, showing type c crystal. 014. CO-Hemoglobin of the Mountain Lion, showing tabular type c crystals of a-CO-Hemoglobin and crystals of /3-CC)-Hemoglobin that have grown on a-CO-Hemoglobin crystals and gradually absorbed them. , , , , o'o. Reduced Hemoglobin of the Leopard-cat (Felis hengalemis), showing small, first-formed crystals. olb. Same, showing symmetrical crystal consisting of unit prism and macrodome. PLATE 87 517, 518. Reduced Hemoglobin of the Leopard-cat (Felis hengalermji), showing single crystals in various orientations. In 518 a number of different sections of the prism are shown. 519. Same, showing two skeleton groups in parallel growth orientation. 520. Same, showing group of large crystals growing from a square cross-section of a prism as a nucleus. Cross-section of a prism seen near bottom of figure. 521. Same, showing group similar to 520, with single large prism and oblique section of another pn.sm that resembles an acute rhombohedron. 522. S.«imp a)ir.w;r,,r =;„„]„ i„-gg crystal with outgrowth of smaller crystals. 523. Reduced Hemoglobin of the Ocelot {Felis pardalis), showing small, first-formed crystals, growing singly or twinned. 524. Same, showing large crystals covered with radiating tufts of smaller crystals. 525, 526. Same, large crystals, many in cross-section, showing uiacropinacoid. 527. Reduced Hemoglobin of the Cat (Felis dom.estica), showing first detiaite crystals to form and parallel growth aggregates of reduced hemoglobin roils. These groups are later absorlied as symmetrical crystals develop. 528. Same, showing network of long prismatic crystals of second crop. PLATE 89 " '" ' '.-,■■. *• ~ - ■^^^^, 531 532 533 529. Reduced Hemoglobin of the Cat (FeUs dorne.Uca), showmg rough parallel growth aggregates like those of first-formed crystals. .'530. Same, showing network of large prismatic crystals. 531. Same, showing single crystal lying on face of pnsni. .,„*;„,, nf lirc^e crystal. 532. Same! showin| smaller, 'later-crop crystals, gro^yu.g rom f^^^^^^^^l^!^-' "^-'' 533. Same, showing smaller crystals growing in --adiating form from U^^^^^ ^^ ^^.^^ 534. Same, showing dome symmetrically and unsymmetrically de^elopea, aiu. .v Cat {Lynx rufus). PLATE 90 540 S^ 535. Reduced Hemoglobin of the Cat (Felis dnmestica) , showing short prismatic type of crystal. oo6. Same, showing cross-section of prism. 5.37. Same, showing group of short type of crystals growing on an oblique cross-section of prism; also prism ^'^'^ °"^ dome face developed making a monoclinic-Iooking crystal, fon' ^'^™^) showing radiating groujxs of smaller crystals growing on larger prisms. lin ?^™^' showing prismatic type of crystal. 540. Same, showing tabular type of crystal. PLATE 91 541, 542. Oxyhemoglobin of the Wild Cat (Lijtix rufus), showing tabular type of crystal, the base bounded by unit prism and two pinacoids. 543. Same, showing large prismatic crystals covered with a secondary growth. I nsymnietncal dome termination seen in large crystal below middle of plate. 544. Same, showing larger prismatic crystals covered by growth of smaller crystals. 545, 546. Reduced Henioglobin of the Wild Cat, showing long prismatic type of crystal and (ui j44) forin- ••■" •"*■;•':.-':. Manv of these are in tw'in positions. PLATE 92 547. Reduced Hemoglobin of the Wild Cat (Lynx rufus), showing long prismatic crystals, some with imsyrametrical ends, some with an overgrowth of radiatmg smaller crystals, llie two crys- tals" to upper left are in twinned position. , 548. Same, showing larger long prismatic crystals with attached overgrowth oi smaller crystals. 549. Same, showing short type of prismatic crystal. , , . ,, , , 550. Same, showing twin on brachvdome in upper left and immediately below it a paralle grow h showing group extending iu direction of macro-axis. Crystals on either side ot middle indi- vidual have unsvmmetrical development of brachvdome, but this is arranged symmetrically PLATE 93 i 558 553. Reduced Hemoglobin of the Lynx (Lynx canadensis var.), showing type a crystals, consisting of brachy- prisin and macrodome; in some the prism is very short. 554. Same, showing type a crystals, some showing macropinacoid in addition to brachyprism and macro- dome. Distorted crystal with unequally developed dome faces seen on left. 555. Same, showing type h crystal w'ith long type a prisms growing out from it. Several smaller type h crystals are seen in this figure. 556. Same, showing parallel growth in groups of type a crystals. 557. Same, showing large type h crystal with a parallel growth group of type a crystals. PLATE 94 ^^ 560 ^kza. /. ^ M '> C %-- ~\ ■tl oo9, 560. Oxyhemoglobin of the Mole (Scalops aquaticus), showing small barrel-shaped crystals, some in twinned position. 561, 562. Same, showing large crystals along with crystals of first crop. 563, 564. Oxyhemoglobin of the Fox-bat (Pteropiis medium), showing small tabular crystals. PLATE 95 ))Ho, 566. Oxyhemoglobin of the Brown Bat {Vet^pertiliofusciin), showing broad lath-shaped crystals, tabvi- lar on base. •567. Oxyhemoglobin of the Ring-tailed I.emur {Lemur catla), showing first crystals to form imperfect prisms consisting of bundles of fibers crossing each other at definite angles and also small imperfect tabular crystals. 568. Same, showing hexagonal plate produced by twinning. Other crystals are seen growing on base of main plate, not all in exact orientation. 569. Same, showing mimetic hexagonal tabular crystals. 570. Same, showing rosette-shaped groups of crystals. PLATE 96 >«. ♦ #v-- a :::m' 571. «-Oxyhemoglobin of the Yellow Baboon (Fapio babuin), showing small tabular crystals along protein ring. 572. Same, showing thicker tabular, nearly cubical crystals. 573, 574. /3-Oxvhemoglobin of the Yellow Baboon, showing tabular and prismatic types of crystals. 575, 576. Oxyhemoglobin of the Drill {Papio leucop/ne us), ^Rhovfins. long rod-like crystals growmg m diver- gent tufts and irregular aggregates. PLATE 97 577, 578. 579. 580. 581. 582. ; -Oxyhemoglobin of the Guinea Baboon (Papio sphinx), showing long diamond-shapetl tabular crystals, growing in divergent tufts and piling up in approximately parallel growth. ra-Oxyhemoglobin and /3-Oxyhemoglobin of the Long-armed Baboon (Papio langheldi). showing short lath-shaped crystals of a-Oxyhemoglobin and rhombic tabular crystals of /9-Ox^hemo- globin. a-Oxyhemoglobin of the Long-armed Baboon, showing thicker tabular crystals. Same, showing twin on brachydome. ;3-0xyhemoglobiu of the Long-armed Baboon, showing thick tabular and prismatic crystals. Different depth of shading is due to pleoehroism. -^ PLATE 98 -.~3 i # '^^Ji^ osd, p-Oxyhemoglobin of the Long-armed Baboon {Papio langheldi), showing tabular and prismatic types of crystals. Oo4. Same, showing tabular and prismatic types of crystals. Horse-type twin on edge is seen near lower part of field. ?sfi' ^^"^^' showing horse-type twin on right of field. ^S? ^^^' showing long and short prismatic crystals. obi. a-Oxyhemoglobin of the C'hacma (Papio po'rcarius), showing tabular crystals. Difference in tint is due to oleoehroism. 088. Same, showing partial parallel growth and fan-shaped appearance of group when seen on edge. PLATE 99 a-Oxyhemoglobin of the Chacnia (Papio porcarius), showing thick tabular crystals in various orienta- tions and illustrating pleochroism as crystals are viewed in different positions, ^arne, showing parallel growth and fan-shaped aggregates. ,a-0.xyhemoglobin of the Chacma, showing side view of horse-type twin. Same, showing parallel growth in direction of ortho-axis. ""Oxyhemoglobin and /3-Oxyhemoglobin of the t'hacma, showing horse-type twins of ,9-Oxyheraoglobin. 594. a-Oxyhemoglobin and /?-Oxyhemoglobin of the Chacma, showing parallel growth and horse-type twinning. 590, 591, 592. 593. PLATE 100 .r^ 595. /3-Oxyhemoglobin of the Anubis Baboon {Papio anubis), showing tabular crvstals with unequal devel- , opment of orthopinacoid faces. CQ7 o "^*^' **^°"■i"g tabular crystals in horse-type twins, on edge and on flat. 2qo- o ™''' showing single tabular crystals. TOQ t^"^^' ^^°^^™g prismatic type of crystal. fim'i'o ,?^' ^'^o^^'i'iS large crystals with tufted groups of thin crvstals of -/-Oxyhemoglobin. ouu. /^-Oxyhemoglobin and )-Oxyhemoglobin of the Anubis.