w o V c y o n of W/# y ^ CORNELL LAB of ORNITHOLOGY LIBRARY At Sapsucker Woods Illustration of Snowy Owl by Louis Agassiz Fuertes M^Sj|MTY,UBRARY ■■i ,„,„ 3 1924 090 29 "056 DATE DUE GAYLOBD PRINTED IN U.SA 2274 SMITHSONIAN INSTITUTION UNITED STATES NATIONAL MUSEUM Bulletin 57 THE FAMILIES AND GENERA OF BATS BY GERRIT S. MILLER, Jr. Assistant Curator, Division of Mammals U. S. National Museum m^M WASHINGTON GOVERNMENT PRINTING OFFICE 1907 LABORATORY OF ORNITHOLOGY LIBRARY CORNELL UNIVERSITY LIBRARY ;£«. He* York 1485 r LIST OF ERRATA AND CORRECTIONS. On page xiii, line 6 from bottom, for trintatis read Prinitatis. On page xv, lines 8 and 18 from top, for Genoa read Geneva. On page xv, line 24 from top, for Ilarpiocephlus read Harpio- cephalus. On page xvn, line 7 from bottom, for Molussus i*ead Mvlossus. On page 47, line 14 from top, for Acerdon read Acerodon. On page 93, line 12 from top, for Taphonycteries read Taplionycteris. On page 151, line 12 from bottom, for effecting read affecting. On page 158, line 12 from top, for macconelli read macconneUi. On page 200, line 7 from bottom, for Myotus read Myotis. On page 206, between lines 1 and 2, insert: 1867. Alobus Peters, Monatsber. k. preuss. Akad. Wissensch., Berlin, p. 707 {temminckii=ruppellii), not Alobus Le Conte, 1856. On page 208, line 12 from bottom, for Vespertilio read Eptesicus. On page 208, line 17 from bottom, for Hungarlee read Hungarian On page 228, line 3 from bottom, for fucas vea.dfu.seus. On page 268, line 27 from bottom, strike out fucas, Miniop- terus 228. On page 268, between lines 9 and 10 from bottom, under fuscus, insert Miniopterus 228. On page 272, line 4 from bottom, for fucas read fuscus. 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/cu31924090297056 UBeraiory ©» Om»M»(W iB« Sapsuckar Woo* R°« d Cornell U»l*ef»ity Ittwea, New York 1485 • SMITHSONIAN INSTITUTION UNITED STATES NATIONAL MUSEUM Bulletin 57 THE FAMILIES AND GENERA OF BATS BY GERRIT S. MILLER, Jr. Assistant Curator, Division of Mammals U. S. National Museum WASHINGTON GOVERNMENT PRINTING OFFICE 1907 QU 737 Published June 29, 1907. ADVERTISEMENT. The scientific publications of the National Museum consist of two series — the Bulletin and the Proceedings. The Bulletin, publication of which was begun in 1875, is a series of more or less extensive works intended to illustrate the collections of the United States National Museum and, with the exception noted below, is issued separately. These bulletins are monographic in scope and are devoted principally to the discussion of large zoological and botanical groups, faunas and floras, contributions to anthropology, reports of expeditions, etc. They are usually of octavo size, although a quarto form, known as the Special Bulletin, has been adopted in a few instances in which a larger page was deemed indispensable. This work forms No. 57 of the Bulletin series. Since 1902 the volumes of the series known as " Contributions from the National Herbarium" and containing papers relating to the botanical collections of the Museum, have been published as bulletins. The Proceedings, the first volume of which was issued in 1878, are intended as a medium of publication of brief original papers based on the collections of the National Museum, and setting forth newly acquired facts in biology, anthropology, and geology derived there- from, or containing descriptions of new forms and revisions of lim- ited groups. A volume is issued annually, or oftener, for distribution to libraries and scientific establishments, and in view of the impor- tance of the more prompt dissemination of new facts a limited edition of each paper is printed in pamphlet form in advance. Charles D. Walcott, Secretary of the Smithsonian Institution. Washington, U. S. A., June 15, 1907. PREFACE. This classification of the families and genera of bats, primarily based on skeletal and dental characters, is chiefly the result of my studies of the collections in the United States National Museum. During two visits to Europe I have, however, been permitted with the utmost liberality to examine the material in the museums of Lon- don, Paris, Leiden, and Berlin, with the result that members of prac- tically all the known genera of Chiroptera have passed through my hands. To the authorities of these institutions I take pleasure in acknowledging my indebtedness. I am also under special obligations to Mr. Oldfield Thomas for the unique facilities that I have enjoyed for working in the British Museum, and to Mr. Knud Andersen for his unfailing kindness in verifying special characters in the specimens in London. Finally, to my assistant, Dr. Marcus W. Lyon, jr., acknowledgment is due for his care in superintending the drawing of some of the illustrations and assistance in reading proof during my absence from Washington. The pen drawings in the text are mostly by Miss Mary Mason Mitchell and the late Dr. J. C. McConnell; those in Plates XI to XIV are also by Doctor McConnell. The wash drawings of the teeth, forming Plates I to X, were made under my supervision by the late Mr. F. van Iterson. They should be regarded as semidiagram- matic. Owing to the fragmentary condition of the remains by which they are known, it has not seemed expedient to attempt to introduce the extinct genera in a system based on characters which the fossils have for the most part lost. It has also seemed inexpedient to deal with subgenera at a time when the species of bats are still very imperfectly known. All questions of nomenclature have been decided in accordance with the Code of Nomenclature of the American Ornithologists' Union, pending the final adoption of an international code. Gerrit S. Miller, Jr. <• Cormura, Stenoderma. and Amurphochihts are the only genera among the 173 here recognized that I have not seen. v CONTENTS. Page. Introduction 1 History 2 Anatomy 12 Wing 13 Humerus, shoulder, and. elbow 13 Phalanges 17 Sternum and shoulder girdle 17 Teeth 20 Milk dentition 20 Dental formula 23 Normal cusps 29 Interrelation of teeth - 33 Modifications of cusps 36 Classification 43 Order Chiroptera 43 Suborder Megachiroptera 44 Family Pteropidse 45 Subfamily Pteropinse 45 Genus Cynopterus F. Cuvier 47 Genus Niadius Miller 49 Genus Thoopterus Matschie 50 Genus Ptenochirus Peters 51 Genus Megaerops Peters 51 Genus Balionycteris Matschie 52 Genus Sphserias Miller 53 Genus Rousettus Gray 54 Genus Pterocyon Peters 55 Genus Pteropus Brisson 56 Genus Acerodon Jourdan 59 Genus Desmalopex, new genus 90 Genus Pteralopex Thomas 60 Genus Boneia Jentink 61 Genus Styloctenium Matschie 62 Genus Dobsonia Palmer 63 Genus Scotonycteris Matschie 64 Genus Epomophorus Bennett 65 Genus Hypsigiiathus H. Allen 67 Subfamily Kiodotina? 68 Genus Eonycteris Dobson 69 Genus Callinycteris Jentink 69 Genus Kiodotus Blyth 70 Genus Odontonycteris Jentink 71 Genus Syconycteris Matschie 72 Genus Trygenycteris Lydekker 73 Genus Melonycteris Dobson 73 Genus Nesonycteris Thomas 74 Genus Notopteris Gray 74 VII VIII CONTENTS. Classification — Continued. Page. Order Chiroptera — Continued. Suborder Megachiroptera — Continued. Family Pteropidse — Continued. Subfamily Nyctymeninse 75 Genus Nyctymene Bechstein 75 Subfamily Harpy ionycterinse 77 Genus Harpyionycteris Thomas 77 Suborder Microchiroptera 78 Family Rhinopomidae 80 Genus Rhinopoma Geoffroy 81 Family Emballonuridse 82 Subfamily Emballonurina? 8.1 Genus Emballonura Temniinck 86 Genus Coleura Peters 87 Genus Rhyncbiscus Miller 88 Genus Saccopteryx Illiger 89 Genus Cormura Peters 90 Genus Peropteryx Peters 90 Genus Peronymus Peters 90 Genus Centronycteris Gray 91 Genus Myropteryx Miller 91 Genus Balantiopteryx Peters 92 Genus Taphozous Geoffroy 93 Subfamily Diclidurinse 94 Genus Diclidurus Wied 95 Family Noctilionidse 95 Genus Noctilio Linnaeus 97 Genus Dirias Miller 99 Family Nycteridse 99 Genus Nycteris Geoffroy 101 Family Megadermidse 101 Genus Megaderma Geoffroy 103 Genus Lyroderma Peters 104 Genus Macroderma Miller 105 Genus Lavia Gray 105 Genus Cardioderma Peters 10*6 Family Rhinolophidoe 106 Genus Rhinolophus Lacepede 108 Family Hipposideridre 109 Genus Hipposideros Gray 110 Genus Asellia Gray 112 Genus Anthops Thomas 113 Genus Coelops Blyth 113 Genus Cloeotis Thomas 114 Genus Rhinonycteris Gray 114 Genus Tritenops Dobson 115 Family Phyllostomidse 116 Subfamily Chilonycterinaa 1 ;. 118 Genus Chilonycteris Gray 119 Genus Pteronotus Gray 120 Genus Mormoops Leach 121 CONTENTS. IX Classification — Continued. Page. Order Chiroptera — Continued. Suborder Microchiroptera — Continued. Family Phyllostomidse — Continued. Subfamily Phyllostominae 121 Genus Micronycteris Gray 123 Genus Xenoctenes, new genus 124 Genus Glyphonycteris Thomas 125 Genus Otopterus Lydekker 126 Genus Lonchorhina Tomes 127 Genus Dolichophyllum Lydekker 127 Genus Tonatia Gray 128 Genus Minion Gray 129 Genus Anthorhina Lydekker 129 Genus Phyllostomus LacepSde 130 Genus Phylloderma Peters 131 Genus Trachops Gray 132 Genus Chrotopterus Peters 133 Genus Vampyrus Leach 134 Subfamily Glossophaginse 136 Genus Glossophaga Geoffroy 137 Genus Lonchophylla Thomas 139 Genus Monophyllus Leach 139 Genus Anoura Gray 139 Genus Lonchoglossa Peters 140 Genus Choeronycteris Tschudi 141 Genus Hylonycteris Thomas 142 Genus Leptonycteris Lydekker 142 Genus Lichonycteris Thomas 143 Subfamily Hemiderminas 144 Genus Hemiderma Gervais 145 Genus Rhinophylla Peters 146 Subfamily Sturnirinse 147 Genus Sturnira Gray 148 Subfamily Stenoderminae 149 Genus Brachyphylla Gray 152 Genus Uroderma Peters 154 Genus Vampyrops Peters 155 Genus Vampyrodes Thomas 156 Genus Vampressa Thomas 156 Genus Vampyriscus Thomas 156 Genus Chiroderma Peters 157 Genus Mesophylla Thomas 158 Genus Ectophylla H. Allen 159 Genus Artibeus Leach 160 Genus Enchisthenes Andersen 162 Genus Ardops Miller 162 Genus Phyllops Peters. 104 Genus Ariteus Gray 165 Genus Stenoderma Geoffroy 165 Genus Pygoderma Peters 166 Genus Centurio Gray 168 Genus Sphseronycteris Peters 170 Genus Ametrida Gray 171 X CONTENTS. Classification — Continued. Page. Order Chiroptera — Continued. Suborder Microchiroptera — Continued. Family Phyllostomidae — Continued. Subfamily Phyllonycterinae 171 Genus Phyllonycteris Gundlach 172 Genus Reithronycteris Miller 174 Genus Erophylla Miller 175 Family Desmodontidse 176 Genus Desmodus Wied 177 Genus Disemus Miller 178 Genus Diphylla Spix 179 Family Natalidse 180 Genus Natalus Gray 183 Genus Phodotes Miller 184 Genus Chilonatalus Miller 185 Genus Nyctiellus Gervais 185 Family Furipteridse 186 Genus Furipterus Bonaparte 188 Genus Amorphochilus Peters 190 Family Thyropteridre 190 Genus Thyroptera Spix 192 Family Myzopodidse 193 Genus Myzopoda Milne Edwards and Grandidier — 194 Family Vespertilionidas 195 Subfamily Vespertilioninse 197 Genus Myotis Kaup 200 Genus Pizonyx Miller 202 Genus Lasionycteris Peters .__ 203 Genus Pipistrellus Kaup 204 Genus Glischropus Dobson 205 Genus Scotozous Dobson 206 Genus la Thomas 206 Genus Pterygistes Kaup 207 Genus Bptesicus Rafinesque 207 Genus Vespertilio Linnseus 209 Genus Rhinopterus Miller 210 Genus Hesperoptenus Peters 211 Genus Tylonycteris Peters 212 Genu's Mimetillus Thomas 213 Genus Philetor Thomas 213 Genus Histiotus Gervais 214 Genus Lsephotis Thomas 215 Genus Otonycteris Peters 215 Genus Nycticeius Rafinesque 216 Genus Scotcecus Thomas 217 Genus Scoteinus Dobson 217 Genus Scotomanes Dobson 217 Genus RhogeSssa H. Allen 218 Genus Bfflodon Miller 218 Genus Pachyotns Gray 219 Genus Chalinolobus Peters 219 Genus Glauconycteris Dobson 221 CONTENTS. XI Classification — Continued. Page. Order Chiroptera — Continued. Suborder Microchiroptera — Continued. Family Vespertilionida; — Continued. Subfamily Vesp'ertilioninie — Continued. Genus Lasiurus Gray 221 Genus Dasypterus Peters 222 Genus Barbastella Gray 223 Genus Plecotus Geoffroy 224 Genus Corynorhinus H. Allen 225 Genus Euderma H. Allen 225 Subfamily Miniopterinas 227 Genus Miniopterus Bonaparte 227 Subfamily Murininse 229 Genus Murina Gray , 22!) Genus Harpiocephalus Gray 230 Subfamily Kerivoulinse 232 Genus Kerivoula Gray 232 Genus Phoniseus Miller 233 Subfamily Nyctophilia? 234 Genus Antrozous H. Allen 235 Genus Nyctophilus Leach 236 Subfamily Tomopeatinse 237 Genus Tomopeas Miller 238 Family Mystacopidse 239 Genus Mystacops Lydekker 240 Family Molossidse 241 Genus Ctuerephon Dobson 244 Genus Eomops Thomas 245 Genus Molossops Peters 247 Genus Cheiromeles Horsfield 249 Genus Nyctinomus Geoffroy 251 Genus Mormopterus Peters 253 Genus Platymops Thomas 254 Genus Eumops Miller 257 Genus Promops Gervais 259 Genus Molossus Geoffroy 260 LIST OF ILLUSTRATIONS TEXT FIGURES. Pig. 1. Milk dentition of Cynopterus minutus. Cat No. 141271, U.S.N.M. Nias Island. X about 3 21 2. Milk dentition of Eptesicus fuscus. Cat. No. 84550, U.S.N.M. X about 4 22 8. Milk dentition of Promops fosteri. Cat. No. 105681, U.S.N.M. Villa Rica, Paraguay. X about 3 22 4. Typical molar teeth of an insectivorous bat. A, crown view of maxillary molar. B, crown view of mandibular molar. C, side view of maxillary molar 30 ecd. — entoconid. -mod. — metaconid. pcd. — paraconid. he. — hypocone. ms. — mesostyle. pre. — protocone. hod. — hypoconid. mts. — metastyle. prcd. — protoconid. me. — metacone. pc. — paracone. ps. — parastyle. 5. Cynopterus sphinx. Cat. No. 102430, U.S.N.M. Madras, India. Adult female, x 1J 48 6. Cynopterus montanoi. Cat. No. 102432, U.S.N.M. Singapore, Malay Peninsula. Adult female. X 1J 49 7. A, Rousettus amplexieaudatus. Cat. No. 37930, U.S.N.M. Caves near Maulmain, Burma. Adult female. X 2. B, Pterocyon stramineus. Cat. No. 102461, U.S.N.M. Robertsport, Liberia. Adult female. X 2 55 8. Pteropus lepidus. Cat. No. 101670, U.S.N.M. Saddle Island, South China Sea. Type. Adult female, x 1 58 9. Epomophorus franqueti. Cat. No. 38189, U.S.N.M. Liberia. X H- 66 10. Kiodotus lagochilus. Cat. No. 123440, U.S.N.M. Philippine Islands. Adult female. X 1J 71 11. Rhinopoma microphyllum. Cat. No. 37389, U.S.N.M. Egypt. Im- mature female. X 14 81 12. Emballonura penmsularis. Cat. No. 83556, U.S.N.M. Trong, Lower Siam. Adult male. X 2i 87 13. Rhynchiscus naso. Cat. No. 51565, U.S.N.M. Escondido River, Nicaragua, 50 miles from Blueflelds. Adult male. X 2 88 14. Saccopteryx MUneata. Cat. No. 6102, Amer. Mus. Nat. Hist. Caura, Trinidad. X 2 89 15. Peropteryx trintatis. Cat. No. 7496, Amer. Mus. Nat. Hist. Port of Spain, Trinidad. Adult female. Type. X 2 90 16. Balantiopteryx plicata. Cat. No. 51142, U.S.N.M. Morelos, Tehuan- tepec, Mexico. Adult female. X 2 92 17. Taphozous saccolaimus. Cat. No. 141092, U.S.N.M. Tarussan Bay, Sumatra. Adult female. X 2 94 XIII XIV LIST OF ILLUSTRATIONS. Page. Fig. 18. Diclidurus virgo. Cat. No. 120577, U.S.N.M. Champerico, Gua- temala. X 1J 95 19. Nycteris javanica. Cat. No. 112608, U.S.N.M. Seinbrong River, Johore. X 11 100 20. Megaderma spasma. Cat. No. 112733, U.S.N.M. Tanjong Sika Kap, Johore. X H 104 21. A, Asellia tridens. Cat. No. 38021, U.S.N.M. Egypt. Adult female. X 2. B, Hipposideros larvatu§. Cat. No. 83570, U.S.N.M. Lower Siam. Adult female. X 2 112 22. Triwnops persicus. Cat. No. 123439, U.S.N.M. Aden, Arabia. X 2. 115 23. Pygoderma bilabiatum. Cat. No. 105685, U.S.N.M. Sapucay, Para- guay. X If 167 24. Phyllonycteris poeyi. Cat. No. 103585, U.S.N.M. Guanajay, Cuba. Adult female. X 11 ' 173 25. Reithronycteris aphylla. Cat. No. 9, Museum, Institute of Jamaica. Type. Adult male. X If 174 26. Natalus mexioanus. Cat. No. 102509, U.S.N.M. Morelos, Mexico. Adult female. X 2J 184 27. Chilonataliis micropus, X 2§ 185 28. Nyctiellus lepidus. Cat. No. 103898, U.S.N.M. Isle of Pines, Cuba. Adult female. X 2i 186 29. Furipterus horrens, X 2§ 189 30. Thyroptera disoifera, X 2§ 192 31. Pizonyx vivesi. Cat. No. 123701, U.S.N.M. Guaymas, Mexico. X 1* 202 32. Pipistrellus pipistrellvs, X 2 205 33. Hesperoptenus tickelli. Cat. No. 123437, U.S.N.M. Ceylon, x 2__ 211 34. Tykmycteris padhypus. Cat. No. 83526, U.S.N.M. Trong, Lower Siam. Adult male. X li 212 35. Philetor rohui. Cat. No. 123438, U.S.N.M. Albert, New Guinea. X 2f 214 36. Ghalinolobus tuberculatus. Cat. No. 38031, U.S.N.M. Dunnedin, New Zealand. Adult male. X 2§ 220 37. Euderma macttlatum. Cat. No. 122545, U.S.N.M. Mesilla Park, New Mexico. Adult male. X 11 226 38. Kerivoula hardwickii. Cat. No. 141591, U.S.N.M. Western Java. Adult male. X 2| 233 39. Molossops planirostris. Cat. No. 37741, U.S.N.M. Paraguay. x2_ 248 40. Molossops temmmcJcii. No. 4530. Miller Coll. X 2 249 41. Nyctinom us brasiliensis. Cat. No. 37873, U.S.N.M. Parana, Brazil. X 2 252 42. Nyctinomus europs. Cat. No. 101502, U.S.N.M. Brazil. Para- type. Female. X 21 252 43. Nyctinomus macrotis. Jamaica, No. 3914. Miller Coll. X about 2J_ 253 44. Mormopterus minutus. Cat. No. 4915, Amer. Mus. Nat. Hist. Trinidad, Cuba. Adult male. X 2J 254 45. Eumops californicus. Cat. No. 88451, U.S.N.M. Tucson, Arizona. X 2 256 46. Eumops abrasus. Cat. No. 37456, U.S.N.M. Surinam. Adult male. X 2 257 47. Ewmops nanus. Cat. No. 0.7.11.99, British Museum. Bogava, Chi- riqui, Panama. Type Adult male. X 2 258 LIST OF i ILLUSTRATIONS. XV Page. Fig. 48. Promops fostcri. Cat. No. 105077, U.S.N.M. Villa Rica, Para- guay. Adult female. X 2\ 259 49. Molossus ritfu-t. Cat. No. 114885, U.S.N.M. Supacay. Paraguay. X2 -• 260 PLATES Plate I. Fig. 1. Upper tootlirow, X -V- (98). Rhinolophus ferrum-ctjitiiuim. Cat. No. 124390, U.S.N.M. Near Genoa, Switzerland. 2. Upper toothrow, X 8, Pachyotus kuhlii. Cat. No. 113402, U.S.N.M. Iligan, Mindanao, Philippine Islands. 3. Upper toothrow, X \ 2 - (4f), Vampyrus spectrum. Cat. No. 78127, U.S.N.M. Biol. Survey, Dept. Agric. Coll. Coatzacoalcos, Vera Cruz, Mexico. 4. Upper toothrow, X ty (Si), Harpiocephalus liarpia, Tomes Collection. Java. Plate II. Fig.,1. Lower toothrow, X -%"• (93), Rhinolophus ferrum-equiniim. Cat. No. 124390, U.S.N.M. Near Genoa, Switzerland. 2. Lower toothrow, X S, Pachyotus kuhlii. Cat. No. 113402, U.S.N.M. Iligan, Mindanao, Philippine Islands. 3. Lower toothrow, X \ 2 - (4g), Vampyrus spectrum. Cat. No. 78127, U.S.N.M. Biol. Survey, Dept. Agric. Coll. Coatzacoalcos, Vera Cruz, Mexico. i 4. Lower toothrow, X-j* (8$), Harpiocephlus harpia, Tomes Collection. Java. Plate III. Fig. 1. Upper toothrow, X 8, Micronycteris megalot'm. Cat. No. 102913, U.S.N.M. La Guaira, Venezuela. 2. Upper toothrow, X 8, Glossophaga longirostris. Cat. No. 102817, U.S.N.M. Macuto, Venezuela. 3. Upper toothrow, X 8, Sturnira Uliuni. Cat. No. 105591, U.S.N.M. Villa Rica, Paraguay. 4. Upper toothrow, X 8, Phyllonycteris poeyi. Cat. No. 113725, U.S.N.M. Baracoa, Cuba. Plate IV. Fig. 1. Lower toothrow, X 8, Micronycteris megalotis. Cat. No. 102913, U.S.N.M. La Guaira, Venezuela. 2. Lower toothrow, X 8, Glossophaga longirostris. Cat. No. 102817, U.S.N.M. ' Macuto, Venezuela. 3. Lower toothrow, X 8, Sturnira lilium. Cat. No. 105591, U.S.N.M. Villa Rica, Paraguay. 4. Lower toothrow, X 8, PhyUonycteris poeyi. Cat. No. 113725, U.S.N.M. Baracoa, Cuba. Plate V. Fig. 1. Upper toothrow, X >¥ (7£), Artibeus lituratus. Cat. No. 121450, U.S.N.M. Sapucay, Paraguay. 2. Lower toothrow, X V (7i), Artibeus lituratus. Cat. No. 121450, U.s'.N.M. Sapucay, Paraguay. 25733—07 m ii XVI LIST OF ILLUSTRATIONS. Plate VI. Fig. 1. Upper toothrow, x ^ (6§), Xeroderma convex urn. Cat. No. 111722, U.S.X.M. Colon, Panama. 2. Upper toothrow, X %« (6g), JCenturio xenea: Cat. No. 37786, U.S.X.M. Mirador, Mexico. 3. Upper toothrow, X - 3 / (7J), BrachypyUa cavernarum. Cat. No. 106085, U.S.N.M. St. Vincent, West Indies. Plate VII. Fig. 1. Upper toothrow, X $■ (1£), Pteropus vampymx. Cat. No. 101594, U.S.N.M. Linga Island, East Indies. 2. Upper toothrow, X 4, Nyctymene major. Cat. No. 124639, U.S.N.M. Duke of York Island. 3. Upper toothrow, X 4, Niadius princeps. Cat. No. 141234, U.S.N.M. Mojeia River, Nias Island, East Indies. 4. Upper toothrow, X 4, Melonycteris melanops. Cat. No. 124628, U.S.N.M. Duke of York Island. Plate VIII. Fig. 1. Lower toothrow, X §■ (If), Pteropus rampyrus. Cat. No. 101594. U.S.N.M. Linga Island, East Indies. 2. Lower toothrow, X 4, Nyctymene major. Cat. No. 124639, U.S.N.M. Duke of York Island. 3. Lower tcothrow, X 4, Niadius princeps. Cat. No. 141234, U.S.N.M. Mojeia River, Nias Island, East Indies. . 4. Lower toothrow, x 4, Melonycteris melanops. Cat. No. 124628, U.S.X.M. Duke of York Island. Plate IX. Fig. 1. Upper toothrow, X 8, Hemiderma subrufum. Cat. No. 123773, U.S.N.M. State of Vera Cruz, Mexico. 2. Upper toothrow, X 8, Eropliylla planifrons. Oat. No. 102055, U.S.N.M. Nassau, Bahama Islands, West Indies. 3. Upper toothrow, X V 1 (.7$), Desmodus rotundus. Cat. No. 114999, U.S.N.M. Sapucay, Paraguay.' Plate X. Fig. 1. Lower toothrow, X 8, Hemiderma subrufum. Cat. No. 123773, U.S.N.M. State of Vera Cruz, Mexico. 2. Lower toothrow, X 8, Erophylla planifrons. Cat. No. 102055, U.S.X.M. Nassau, Bahama Islands, West Indies. 3. Lower toothrow, X *f- (7i),Desmodus rotundus. Cat. No. 114999, U.S.N.M. Sapucay, Paraguay. Plate XI. Rhinopoma microphyllum, Cat. No. 18538, U.S.N.M., Egypt. All figures X li. Fig. 1. Sternum. 2. Lateral view of pelvis. 3. Dorsal view of pelvis. 4. Ventral view of pelvis. LIST OF ILLUSTRATIONS. XVII Plate XII. Diclidurus virgo, Cat. No'. 120577, U.S.N.M.. Guatemala. All figures, except 2, X li. Pig. 1. Leg bones. 2. Part of leg bones enlarged and cross section of same. 3. Sternum. 4. Lateral view of pelvis. 5. Ventral view of pelvis. C. Dorsal view of pelvis. Plate XIII. Noctilio leporinus, figs. 1, 2, 4, 5,' 6, Cat. No. 114006, U.S.N.M., male; fig. 3, Cat. No. 86708, U.S.N.M., female, Trinidad, West Indies. All figures X 1J. Pig. 1. Leg bones. 2. Sternum. 3. Ventral view of pelvis, female. 4. Lateral view of pelvis, male. 5. Dorsal view of pelvis, male. 6. Ventral view of pelvis, male. Plate XIV. Molussus pretiosus. Cat. No. 102778, U.S.N.M. La Guaira, Vene- zuela. All figures X 1J. Pig. 1. Sternum. 2. Leg bones. 3. Lateral view of pelvis. 4. Ventral view of pelvis. 5. Dorsal view of pelvis. THE FAMILIES AND GENERA OF BATS. By Geerit S. Miller, Jr., Assistant Curator, Division of Mammals, U. 8. National Museum. INTRODUCTION. In 1758, Linnaeus knew seven bats, all of which he placed in Ves- pertilio, the fourth and last genus of the order Primates. The work of the next fifty years, as recorded by Tiedemann in 1808, though it had resulted in the recognition of the order Chiroptera, had not increased the species beyond 14, while the number of genera, even with the addition of the flying lemur, was still only seven. The first subdivision of the' order into families appears to be due to Goldfuss, who, in 1820, arranged the genera in four groups, one of which still comprised the flying lemur. To each group he definitely applied the name " Familie." The real foundation for the current classifi- cation was, however, not laid by Goldfuss, but by Gray, who pub- lished the first of his many papers on bats in 1821. Gray excluded the flying lemur, recognized the two main subdivisions of the order, and applied to the names of families the system of nomenclature now in use. According to this scheme there were two suborders, the Fruc- tivorae and Insectivoras, the former containing the families Pteropidse and Cephalotidse, the latter the Noctilionidse and Vespertilionidse. Though Gray's system was not followed very closely during the suc- ceeding fifty years, it was finally given definite form by Gill in 1872 and Dobson in 1875, and since then has been almost universally adopted. The new classification now presented is, in fact, little more than an amplification of that founded by Gray. The increase in our knowledge of .the Chiroptera since 1821 has been, however, very great. * As we have seen, Linnaeus recognized only one genus of bats; Gray placed the number comprised in his four families at about 14. In 1865 Peters divided the group into 10 families and subfamilies, con- taining, in all, 59 genera. When Dobson published his Catalogue of the Chiroptera in the British Museum, in 1878, he described 401 25733— No. 57—07 m 1 1 2 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. species, 80 genera, and 14 families and subfamilies. As might have been anticipated, this work, the only complete special monograph of the order, was such a stimulus to the study of bats that since it appeared these animals have received more attention than ever before. In 1904 Trouessart recorded no less than 851 species, 122 genera, and 18 families and subfamilies. It appears, however, that even these numbers are much too small. Detailed study of the skeleton, par- ticularly of the wing and shoulder girdle and of the structure of the tooth cusps, leads me to the conclusion that among the known species at least 173 genera and 36 families and subfamilies should be recog- nized. With regard to the species, recent work " shows that an enor- mous increase is to be expected as the characters on which distinc- tions are based come to be better understood. It seems highly prob- able that the total number of recognized bats will eventually exceed 2,000 named forms. HISTORY. The following summaries are intended to give somewhat in detail the more important facts in the development of the classification of bats. The list, though incomplete, gives a sufficiently clear idea of the course that this work has followed. 1758.' Linnaeus, Systema Naturae, I, 10th ed. Vespertilio, the only genus of bats recognized, appears as the fourth genus of Primates (p. 31.) The other genera are Homo, Simla, and Lemur. In the twelfth edition, 1766, Noctilio is separated from Vespertilio and placed among the Glires. This is apparently the only instance of one genus of bats being referred to a different order from the others. 1772. Briinnich, Zoologia Fundamenta. Vespertilio is here one of the genera of Ferae; the others are Hystrix, Lepus, Gavia, Castor, Mus, Sciurus, Etinaceus, Sorex, Phoca, Lutra, Hyaena, Felis, Canis, Cercopithecus, and Lemur. The orders recognized are : Bruta, Ferae, Pecora, Bellua, and Cetacea. 1777. Scopoli, Introductio ad Historiam Naturalem. The genus Vespertilio is placed between Bradypus and Lemur in the Ungui- culata. The position of this order in the system is as follows : Tribus XII. Mammalia. Gens I. Cetacea. Gens II. Quadrupedia. Divisio I. Aquatilia. Divisio II. Terrestria. Ordo I. Ungulata. Ordo II. Unguiculata. a See especially the papers (Proc. Zool. Soc. London, 1905, II, pp. 75-145, October 17, 1905, and Ann. Mag. Nat. Hist., 7th ser., XVI, pp. 648-662, December, 1905) by Mr. Knud Andersen on the Old World leaf -nosed groups. THE FAMILIES. AND GENERA OP BATS. 6 1777. Erxleben, Systema Regni Animalis. Two genera of bats are recognized, Pteropus and Vespertilio, both placed in the Linneean order Primates. 1779. Blumenbach, Handbuch der Naturgeschichte. The Mammalia are divided into 12 orders: I, Inermis; II, Pitheci; III, Bradypoda; IV, Sclerodermata ; V, Chiroptera; VI, Glires; VII, Ferse; VIII, Solidungua; IX, Bisulca; X, Belluse; XI, Palmatse, and XII, Cet- acea (pp. 57-59). Vespertilio, with 5 species, is the only genus of bats. 1800. Cuvier, Lecons d'Anatomie Comparee. The Cheiropteres are placed as a subdivision of the Carnivora. 1806. Dumeril, Zoologie Analytique. The Mammalia are divided into 14 families, the first three of which are Bimanes, Quadrumanes, and Chiropteres. Six genera of bats are recognized. 1808. Tiedemann, Zoologie. The bats form the " X Ordnung, Vogel- artige Saugthiere, Fliegende Saugthiere, Chiroptera (Chiropteres, Alipedes.)" There are 7 genera, one of which is Galeopithecus. 1816. Oken, Lehrbuch der Naturgeschichte, dritter Theil, Zoologie. The bats (17 genera) form the first division of " III Ordnung, Vogelsuke— Klauer," thus : III Ordnung, Vogelsuke — Klauer. A. Haltklauer. 1 S. Finkenklauer — Flere ; Fledermause. 2 S. Rabenklauer — Tatzer. ; Spitzm., Igel. 3 S. Spechtenklauer — Kraller ; Zahnerlose, Ameisenba- ren, Gurtelth., Schnabelth. B. Reissklauer. '4 S. Aukenklauer — Robben. 5 S. Reiherklauer — like ; Marder. 6 S. Hiihnerklauer — Koter ; Hunde, Katzen. 7 S. Trappenklauer — Baren. 1817. Cuvier, G., Le Regne Animal, I. In this work the " Cheirop- teres ". form the first family of " Carnassiers." They are all placed in the genus Vespertilio, but this is divided into two main groups, the Rousettes {Pteropus) and the true bats. Among the latter, 12 sub- divisions or subgenera are recognized. Essentially the same plan is followed in the second edition, published in 1829. 1820. Desmarest, Mammalogie. The classification is the same as that of Cuvier : Ord. Carnassiers. Premiere Famille, Cheiropteres. Premiere Tribu, Galeopitheques. Seconde Tribu, Chauve-souris. The minor groups of the former author are, however, recognized as 16 genera, while no less than 95 species are described. 4 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. 1820. Goldfuss, Handbuch der Zoologie. The arrangement of the bats is as follows : < 13 tB Ord., Chiroptera. 1 Familie, Noctiliones (Taphosous, Dysopes, Noctilio, Nycti- nomous, Myopterus, Plecotus, Vespertilio, Stenoderma.) 2 Familie, Harpyise (Harpyia, Pteropus) . 3 Familie, Phyllostomata (Megaderma, Rhinolophus, Rhino- poma, Nycteris, Phyllostoma) . 4 Familie, Galeopitheci. 1821. Gray, On the natural arrangement of Vertebrose Animals, London Medical Eepository, XV, pp. 296-310, April 1, 1821. In this little known work, Gray subdivided the vertebrates as follows : Sub-kingdom I. Vertebrosa. Class I. Bimanes. Order I. Primates. Class II. Quadrumanes. Order I. Platyonychse. Order II. Gampstonychse. Order III. Heteronychse. Class III. Cheiroptera. Order I.Fructivorse. Family I. Pteropidse (Pteropus, Rousettus) . Family II. Cephalotidae (G ephalotes) . Order II. Insectivorse. Family I. Noctilionidse (Molosses, Nyctinomes, Stenodermes, Noctilio, Vampyre, Phyllostoma). Family II. Vespertilionidse (Megadermes, Rhyno- lophus, Nycterus, Rhynopoma, Thaphosores, Ves- pertilio, Plecotus, Barbastella). Class IV. Quadrupedes. Subclass I. Unguiculata. Order I. Pterophorse. Order II. Plantigradse. Order III. Digitigradse. Order IV. Amphibise. Order V. Rosores. Order VI. Tardigradse. Order VII. Oligodontse. Order VIII. Edentulse. Order IX. Proboscidise. Order X. Tesserachense. Order XI. Trichenae. Order XII. Monochenas. Order XIII. Hydrophorse. Order XIV. Ruminantes. THE FAMILIES AND GENEKA OF BATS. 1821. Gray — Continued. Sub-kingdom I. Vertebrosa— -Continued. Class V. Pedimanes. Order I. Ferse. Order II. Brutae. Order III. Glires. Order IV. Rosores. Class VI. Cetaceas. Order I. Herbivorae. Order II. Carnivorse. 1823. Spix, Simiarum et Vespertilionum Brasiliensium Species Novse. Though dealing especially with the bats of Brazil the author proposed a general classification that influenced many subsequent writers. It was as follows : Fam. I. Anistiophori. Genus I. Noctilio. II. Molossus. III. Thyroptera. IV. Proboscidea. V. Vespertilio. Fam. II. Istiophori. Genus VI. Vampyrus. VII. Phyllostoma. VIII. Glossophagd. IX. Diphytta. 1827. Lesson, Manuel de Mammalogie. The flying lemur is once more associated with the bats, and the group thus formed is made a division of the carnivores : Ordre Carnassiers. Division Cheiropteres. Tribu Galeopitheques. Tribu Chauve-souris. The true bats are subdivided into two sections, the Istiophori con- taining the groups Phyllostomes and Rhinolophina, and the Anistio- phori, with the groups Vespertilionina, Noctilionina, and Pteropina. Twenty-three genera are recognized. 1829. Fischer, Synopsis Mammalium. The bats are placed as the second order, Chiroptera, without subdivision into secondary groups. 1831. Bonaparte, Saggio di una Distribuzione metodica degli Ani- mali Vertebrati. The bats form the second order, Chiroptera. They are all placed in the single family " Vespertiliones (Vespertilio- nidse) ," but this group is subdivided into five sections or subfamilies. Ord. 2. Chiroptera. Fam. 4. Vespertiliones (Vespertilionidse). Noctilionina (Cheiromeles Dysopes, Thyropte- rus Stenoderma, IMyopteris, Aello Tapho- zous ("Taphagous") Noctilio, Gcelano). 6 BULLETIN 57, UNITED STATES* NATIONAL MUSEUM. 1831. Bonaparte — Continued. Ord. 2. Chiroptera — Continued. Fam. 4. Vespertiliones (Vespertilionidse) — Continued. Vespertilionina (Proioscidea, Diclidurus, Ves- pertilio, Furia, Plecotus, Barbastellus) . Pteropina (Pteropus, Cynopterus, Macroglos- sus, Harpyia, Cephalotes). Phyllostomina (Phyllostoma, Vampyrus, Des- modus, Glossophaga, Rhinopoma, Mormops, Megaderma, Nyctophilus, Nyctinomus) . Khinolophina (Rhinolophus Phyllorrhina) . 1838. Bonaparte, Synopsis Vertebratorum Systematis. (Nuovi Annali delli Scienze Naturali, Bologna, Anno I, Tomo II, pp. 105- 113.) The subdivisions of the Chiroptera are now as follows : Pteropodidae. Pteropodina. Vespertilionidse. Noctilionina. Vespertilionina. Bhionolophina. Rhinopomina. Vampiridse. Vampirina. 1838. Gray, A revision of the genera of bats (Vespertilionidae), and the description of some new genera and species. (Mag. Zool. and Bot., II, pp. 486-505, December, 1838.) In again dealing with the classification of bats Gray abandoned his earlier plan and adopted the main divisions introduced by Spix. As thus modified the arrange- ment of the group was as follows. The number of genera had now been increased to 47 : Family Vespertilionidse. A. Istiophori. Tribe I. Phyllostomina. Tribe II. Rhinolophina. B. Anistiophori. Tribe III. Vespertilionina. Tribe IV. Noctilionina. Tribe V. Pteropina. 1840. Wagner, Schreber's Saugthiere. (Supplement, I.) The fly- ing lemurs are. again associated with the bats. Ord. Volitantia. I. Unterord. Dermoptera. II. Unterord. Chiroptera. Cornell University Library QL 737.C59M64 The families and genera of bate / 3 1924 022 537 322 01 a> I . of gativ CO -P 0) k g. fl o CO •• u cS •• to 0) 0) to •> n Xi 0) (U • o K h O «H 3 . <*-< O -P P o « ft « o o o C -H C -H o e o fc ■p O hO bD w •< •H .. fl h S (U h Oil o A) • • «H W) O 1 XJ CO a X) •H • o *H E) P C\J g s X K o H w p H d) X! O H •h m fe CM THE FAMILIES AND GENEEA OF BATS. 7 1840. Wagner — Continued. Ord. Volitantia — Continued.' II. Unterord. Chiroptera — Continued. I. Fam. Frugivora. II. Fam. Istiophora. I. Sippe. Desmodina. II. Sippe. Phyllostomata. III. Fam. Gymnorhina. I. Sippe. Brachyura. II. Sippe. Gymnura. III. Sippe. Vespertilionina. 1842. Lesson, Nouveau Tableau du Regne Animal, Mammiferes. The classification adopted by Lesson differs in many respects from those of previous writers. I. Classe. Mammiferes. I. Sous-Classe Mammiferes Normaux. I. Ordre Mastomonadelphie. I. Tribu Bimana. II. Tribu Quadrumana. III. Tribu Chiroptera. VI. Famille Pteropusidese. VII. Famille Noctilioninese. VIII. Famille Vespertilioneae. IX. Famille Phyllostomineae. X. Famille Rhinolophinese. 1854. Gervais, Histoire Naturelle des Mammiferes. The order Cheiropteres is divided into the four families Pteropodes, Phyllo- stomides, Rhinolophides, and Vespertilionides. 1855. Gervais, Documents Zoologiques pour servir a la Monographie des Cheiropteres Sud-Americains. Animaux Nouveaux ou Rares re- ceuillis dans l'Amerique du Sud sous la direction du Comte de Castel- nau, pp. 25-88. Only two families are dealt with, but these are sub- divided into groups that have been recognized by most subsequent writers. Cheiropteres Phyllostomides. Tribu Desmodina. Tribu Stenodermina. Tribu Glossophagina. Tribu Vampyrina. Cheiropteres Vespertilionides. Tribu Noctilionina. Tribu Molossina. Tribu Emballonurina. Tribu Nycticeina. Tribu Vespertilionina. 8 BULLETIN 51, UNITED STATES NATIONAL MUSEUM. 1865. Peters, Vorlage von Abbildungen zu einer Monographie der Chiropteren, und Ubersicht der von ihm befolgten systematischen Ordnung der hieher gehorigen Gattungen. (Monatsber. k. preuss. Akad. Wissensch., Berlin, 1865, pp. 256-258.) The following groups are recognized : I. Fam. Pteropi with 10 genera. II. Fam. Megadermata with 4 genera. III. Fam. Rhinolophi with 3 genera. IV. Fam. Phyllostomata. I. Subfam. Vampyri with 5 genera. II. Subfam. Glossophagse with 5 genera. i III. Subfam. Stenodermata with 5 genera. IV. Subfam. Desmodi with 2 genera. V. Subfam. Mormopes with 3 genera. V. Fam. Brachyura with 6 genera. VI. Fam. Molossi with 2 genera. VII. Fam. Vespertiliones with 14 genera. 1866. Gray, I. Synopsis of the genera of Vespertilionidse and Noc- tilionidse. (Ann. and Mag. Nat. Hist., 3d ser., XVII, pp. 89-93.) II. A revision of the genera of Pteropine Bats (Pteropidae) and descriptions of some apparently undescribed species. (Proc. Zool. Soc. London, 1866, pp. 62-67.) III. A revision of the genera of Rhinolophidse, or Horseshoe Bats. (Proc. Zool. Soc. London, 1866, pp. 81-83.) IV. Revision of the genera of Phyllostomidse, or Leaf- nosed Bats. (Proc. Zool. Soc. London, 1866, pp. 111-118.) In these four papers Gray again revised his classification of the bats. He recognized 5 families and 37 subfamilies or "tribes," but no larger groups. Family Pteropidse. Pteropina, 2 genera. Macroglossina, 4 genera. Cephalotina, 1 genus. Harpyiana, 1 genus. Cynopterina, 2 genera. Epomophorina, 3 genera. Family Rhinolophidse. Rhinolophina, 3 genera. Rhinonycterina, 1 genus. [Phyllorhinina] , 8 genera. Rhinopomina, 1 genus. Megadermina, 2 genera. Nycterina, 3 genera. Family Phyllostomidse. Lonchorhinina, 1 genus. Macrophyllina, 1 genus. THE FAMILIES AND GENERA OF BATS. W \ 1866. Gray— Continued. Family Phyllostqmidse — Continued. Vampyrina, 6 genera. Phyllostomina, 8 genera. Trachyopina, 1 genus. Brachyphyllina, 1 genus. Glossophagina, 3 genera. Stenodermina, 8 genera. Desmodina, 2 genera. Centurionina, 2 genera. Family Vespertilionida?. Scotophilina, 4 genera. Romiciana, 1 genus. Vespertilionina, 5 genera. Natalinia, 4 genera. ? Nycticellina, 1 genus. Plecotina, 2 genera. Nyctophilina, 1 genus. Nyctericina, 2 genera. Afterwards removed to the Rhino- lophidse. Furipterina, 1 genus. Emballonurina, 7 genera. Diclidurina, 1 genus. Family Noctilionidse. Noctilionina, 2 genera. Mormopsina, 1 genus. Phyllodiana, 3 genera. ? Spectrellina, 1 genus. Molossina, 4 genera. 1872. Gill, Arrangement of the Families of Mammals. (Smith- sonian Miscellaneous Collections, No. 230.) The bats are arranged as follows: Order Chiroptera. Suborder Animalivora. Family Desmodidse. . Family Phyllostomidse. Family Mormopidse. Family Rhinolophidse. Family Megadermidae. a. Vampyrinse. b. Glossophaginse. c. Stenoderminae. Family Vespertilionidae. a. Vespertilionina?. b. Nycticejinse. 10 BULLETIN 57, UNITED STATES) NATIONAL MUSEUM. 1872. Gill— Continued. Order Chiroptera — Continued. Suborder Animalivora — Continued. Family Molossidse. Family Noctilionidse. a. NoctilioninEe. b. EmballonurinEe. c. Furiina?. Suborder Frugivora. Family Pteropodidse. 1875. Dobson, Conspectus of the suborders, families and genera of Chiroptera arranged according to their natural affinities. (Ann. and Mag. Nat. Hist., 4th ser., XVI, pp. 345-357, November, 1875.) Dobson's classification has been the standard, little departed from except in minor details, during the past thirty years. Order Chiroptera. Suborder I. Megachiroptera. Family I. Pteropidse. Group I. Pteropi, 6 genera. Group II. Macroglossi, 3 genera. Suborder II. Microchiroptera. Family II. Rhinolophidae. Subfamily I. Rhinolophinse, 1 genus. Subfamily II. Phyllorhininse, 4 genera. Family III. Nycteridee. Subfamily I. Megaderminse, 1 genus. Subfamily II. Nycterinse, 1 genus. Family IV. Vespertilionidae. Group I. Plecoti, 7 genera. Group II. Vespertiliones, 8 genera. Group III. Miniopteri, 3 genera. Family V. Emballonuridae. Subfamily I. Emballonurinse. Group I. Emballonurae, 4 genera. Group II. Taphozoi, 3 genera. Group III. Rhinopoma, 1 genus. Group IV. Noctiliones, 1 genus. . Subfamily II. Molossinse. Group V. Molossi, 4 genera. Group VI. Mystacinse, 1 genus. Family VI. PhyllostomidEe. Subfamily I. Lobostominte. Group I. Mormopes, 3 genera. Subfamily II. Phyllostominsa. Group II. Vampyri, 10 genera. Group III. Glossophagce, 6 genera. Group IV. Stenodermata, 10 genera. Group V. Desmodontes, 2 genera. THE FAMILIES AND GENEEA OF BATS. 11 1886. Gill, Standard Natural History, V, pp. 159-177 Suborder Frugivora. Family Pteropodidse. Suborder Animalivora. Family Rhinolophidse. Subfamily Rhinolophinae. Subfamily Phyllorhininse. Family Megadermidse. Subfamily Megaderminse. Subfamily Nycterinse. Family Vespertilionidse. Group Plecoti. Group Vespertiliones. Group Miniopteri. Family Emballonuridse. Group Furiae. Group Emballonurse. Group Dicliduri. Group Rhinopomata. Family Molossidse. Family Noctilionidse. Family Phyllostomidae. Subfamily Phyllostomines. Subfamily Glossophagines. Subfamily Stenodermines. Family Mormopidse. Family Desmodontidse. 1892. Winge, Jordfundne og nulevende Flagermus (ChirOptera) fra Lagoa Santa, Minas Geraes, Brasilien. The classification adopted by Winge is the only recent one that differs notably from that of Dobson. It is as follows (p. 24) : Pteropodidae. Pteropodes. Macroglossi. Rhinolophidae. Megadermatini. Nycterides. Megadermata. Rhinolophini. Phyllorhinse. Rhinolophi. Phyllostomatidse. Phyllostomatini. Phyllostomata. Glossophagee. 12 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. 1892. Winge— Continued. Phyllostomatidae — Continued. Phyllostomatini — Continued. Stenodermata. Desmodontes. Mormopini. Emball onuridse. Rhinopomatini. Emballonurini. Emballonurse. Taphozoi. Vespertilionidse. Natalini. Vespertilionini. Molossini. 1904. Weber, Die Saugethiere. By a combination of Winge's results with those of earlier writers the following scheme is produced (pp. 400^01) : Megachiroptera. Pteropodidse. Pteropodinse. Macroglossinae. Microchiroptera. Rhinolophidae. Megadermatinse. Rhinolophinae. Phyllostomatidse. Lobostominse. Phyllostominse. Desmodontinse. Emballonuridae. Rhinopomatinae. Emballonurinse. Vespertilionidae. Natalinae. Vespertilioninse. Molossinae. ANATOMY. Though a general study of Chiropterine anatomy is still much needed, it is not my purpose in the present paper to undertake this work. Certain special parts, as the wing, shoulder girdle, sternum, and the cusps of the teeth, which have not hitherto been adequately described, but which appear to be of particular taxonomic impor- tance, will, however, be treated in some detail. THE FAMILIES AND GENEBA OF BATS. 13 Wing. The general structure of the wing in bats has been so frequently described that no detailed account is necessary here. Two special parts of the flying apparatus need, however, a few words. HUMERUS, SHOULDEB, AND ELBOW. As might be anticipated from its modification to serve as a wing, the anterior limb of the bats differs considerably from that of mam- mals that do not fly. The most obvious peculiarity is the great length- ening of the fingers to support the flying membrane, but the long bones and their two principal joints have also undergone considerable changes. The humerus, though in general without special modifica- tions in form or unusual development of ridges for muscular attach- ment, is peculiar in the large size of the trochiter (tuberculum majus) and trochin (tuberculum minus) , the former in some families being so greatly developed as to extend beyond the head and form a definite secondary articulation with the scapula, thus giving the shoulder joint a very unusual strength, while limiting its motion strictly to a single plane. The bicipital groove is usually deep and well defined, especially in the larger Pteropidse. Shaft slender, mostly subterete, though somewhat flattened distally, varying in form from distinctly sigmoid in the lower groups to nearly straight in the higher, its deltoid crest always present, low and broad in the Megachiroptera, high and flangelike in the Microchiroptera. Except for this crest the shaft is never marked by distinct ridges, though a faint trace of the supinator ridge is sometimes present. No supracondylar fora- men or supratrochlear perforation. Internal condyle usually large, though in the most specialized groups reduced to a mere base for the long spinous process. External condyle obliterated by the dis- placement outward of the articular surface. Capitellum large, with a well-developed external ridge and groove; trochlea reduced to a mere ridge scarcely more prominent than that on outer side of capi- tellum. The whole distal articular surface is in most bats so dis- placed outward as to be partly or in extreme cases almost wholly beyond axis of shaft. The elbow joint is formed almost exclusively by the humerus and radius, since the ulna is so reduced as to be nearly f unctionless. At its distal end the ulna is incomplete ; among the Megachiroptera it is continued as a threadlike cartilaginous strand to wrist, though in old age it may become fused with radius; in the Microchiroptera this fusion with radius is invariable, though the point at which it takes place is not always the same. Occasion- ally the proximal extremity also is anchylosed. Olecranon scarcely extending beyond radius and forming no definite part of joint, suc- ceeded by a sesamoid ossicle nearly equaling it in size. 14 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Among recent mammals the nearest approach to the conditions just described is found in the Colugidae. Here the trochiter and trochin, though less developed than in the Pteropidae, strongly- suggests the form characteristic of the bats as compared with' their much more rudimentary condition in Tupaia. Of the two processes the trochin is distinctly the larger, and it is obvious at a glance that the distal extremity of the humerus in Cynocephalus more nearly resembles that of Pteropus than it does that of Tupaia or than that of Pteropus does the most highly developed type in the Microchiroptera. This resemblance is further heightened by the size and form of the deltoid crest, which is rather closely alike in Cynocephalus and Pteropus, while in both Tupaia and the Microchiroptera it is noticeably dif- ferent. At the distal extremity of the humerus the resemblance be- tween Cynocephalus and the bats is less exclusive. In the Colugidae both supracondylar foramen and supratrochlear perforation are present and the supinator ridge is well developed, characters all of which are in common with the Insectivora. The actual surface of articulation, however, resembles that of the bats in the reduction of the trochlea and the large size of the capitellum, the outer edge of which shows the first suggestion of the groove and ridge which is so prominent a feature in the Chiroptera. As would be expected from the reduced condition of the trochlea, the ulna is much reduced from its normal form. It is throughout very slender, the distal half flattened against radius, into which it blends near wrist. Proximally it is slightly larger than in Pteropus, though very greatly reduced as compared with Tupaia, and the small, abruptly curved olecranon forms a definite part of the joint, its extremity fitting into the supra- trochlear perforation. It seems, therefore, that without touching on the question of the general relationships of the Dermoptera, it- may safely be said that the long bones and two principal joints of the anterior limb in this group are intermediate in structure between those of Insectivores and Bats, but distinctly nearer the latter, and that in this respect at least Cynocephalus represents a stage that was passed through by the near ancestors of Pteropus. In 1892, Winge " first called attention to the striking differences, already alluded to, in the humerus and shoulder joint of the Mega- chiroptera and of the more highly specialized Microchiroptera. A fact not mentioned by Winge is that, while the Microchiroptera, as a whole, show a much more specialized condition of the humerus than the Megachiroptera, different members of the group show different degrees of this modification, and these degrees are very largely charac- <> Jordfundne og Nulevende Flagermus (Chiroptera) fra Lagou Santa, Minas Geraes, Brasilien, p. 24. THE FAMILIES AND GENEKA OF BATS. 15 teristic of families. In a group like the bats, where divergence from ordinary mammals has been chiefly in the direction of changing the front limbs from legs to wings, it seems necessary to give special weight in classification to such characters as these. Pteropidw.— The very characteristic structure of the humerus in this group shows no important variation in the numerous genera that I have examined. The head is broadly rounded, slightly oval in outline, well defined everywhere, except on the side next the trochiter, into which it merges. Trochiter ill-defined, low and broad, scarcely rising above base of head, to which it is joined by a level, smoothly rounded surface. Trochin slightly higher and narrower than troch- iter, and much better defined, a well-developed groove, about as wide 1 as its lateral diameter, separating it from head, and a similar though even more strongly defined groove isolating it dorsally from the low, broad, deltoid crest. Shaft somewhat triangular proximally, flat- tened distally, its sigmoid flexure distinct. Capitellum moved out- ward slightly beyond line of shaft and entirely obliterating epicon- dyle. Epitrochlea large, its lateral diameter more than half that of capitellum, its spinous process distinct, though short. Microchiroptera. — The least specialized type of humerus in the Microchiroptera is found in the Khinopomidae and Emballonuridse. In these families both trochin and trochiter have risen about to level of upper surface of head of humerus, a development which causes little change in the form of either tubercle as compared with its con- dition in the Pteropidse. The alteration in form is most noticeable in the trochiter, which is now well defined throughout, and separated from head by a distinct groove over which the smooth surface of the head is extended. This groove is narrower and better defined than that dividing the head from the trochin. Though so distinctly increased in size the trochiter is not large enough to reach the edge of the glenoid fossa of the scapula, the shoulder joint remaining therefore strictly single. Head globular in Bhinopoma, somewhat compressed in the Emballonuridse. As in all other Microchiroptera the deltoid crest is high and knifelike, very different from the low, rounded form in the Megachiroptera. The shaft of the bone is nearly terete throughout, except where distorted by the deltoid crest, and a little flattened distally. It has a slight single curvature. Capitellum essentially as in the Pteropidse. A slight modification of this type is found in the Noctilionidse. Here the head of the humerus is greatly compressed and strikingly ridgelike. The well-defined trochiter is very small and narrow, though rising fully as high as in the related groups, while the trochin is greatly increased in size, being nearly as large as the head. The groove .separating the trochin from the well-developed though rather small deltoid crest is ill-defined.- 16 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Another stage is represented by the Old World leaf-nosed groups, the Nycteridse, Megadermida?, Rhinolophidse, and Hipposideridse. In the first two of these the shoulder joint is still single, and the shaft of the humerus shows a slight though evident sigmoid flexure; the trochin and trochiter are as in the Emballonuridae, neither rising distinctly above head. In the Rhinolophidse and Hipposideridte the shaft of the humerus is nearly straight, with merely a slight upward curve in distal third, while the trochiter has increased in length sufficiently to exceed the head slightly and to form a definite secondary articulation with the scapula by a surface about half as large as glenoid fossa. In dze the trochiter is still distinctly ex- ceeded by the trochin, which, however, barely rises above head. At its distal extremity the humerus in all four families is peculiar in the lengthened epitrochlea, bearing near middle a well-developed peg- like process, and the very conspicuous displacement outward of the capitellum and consequent widening of the entire distal region. This displacement is greatest in 'the Megadermidse, where the articular surface is almost wholly exterior to the axis, and the width of the distal extremity is nearly one- fourth length of humerus. In the Nycteridse the displacement is rather less, though the proportion of width of extremity to length of bone remains about the same. In the Rhinolophidse and Hipposideridse it is still less, and the proportion of width to length is normal — that is, about as one to six. The displacement of the articular surface remains very evident, and the epitrochlea and its spinous process are of the same character as in the Nycteridse and Megadermidse, though less elongated. In the American leaf -nosed bats the shoulder joint is essentially like that of the Rhinolophidse and Hipposideridse, except that in the Chilonycterinse the trochiter is so shortened as to form no articulation with the scapula. In this subfamily the shaft of the humerus has a slight single curvature, while in the others the bend is often, though not invariably, double. Throughout the family the capitellum is slightly displaced outward and the epitrochlea is small or moderately developed, the spinous process, when present, at its distal extremity. A similar type of humerus occurs in the Desmodontidse, but both trochiter and trochin rise somewhat higher beyond head. The Natalidse, Furipteridse, Thyropteridse, and Myzopodidse agree with each other in the form of the humerus. This differs from the Phyllostomine type merely in the somewhat enlarged trochiter, which now about equals the trochin in size and slightly exceeds it in height ; distal extremity as in the Phyllostomidse ; flexure single or faintly double. From the last type as best shown in the Thyropteridse to the highly specialized form characteristic of the Vespertiliohidse, Mystacopidse, and Molossidse the transition is rather abrupt. In these families THE FAMILIES AND GENERA OF BATS. 17 the trochiter slightly exceeds the somewhat reduced trochin in actual sizei, while in height it extends conspicuously beyond it as well as beyond the head of the humerus. The double articulation with the scapula is now complete, and the surface on which the flangelike trochiter acts is nearly as large as the glenoid fossa. The shaft of the humerus is nearly straight, or with a moderate single curvature ; dis- tal extremity scarcely wider than shaft, the articular surface not displaced outward, the short epitrochlea scarcely more than the base to the high spinous process, which projects beyond distal extremity in much the same manner as the trochiter beyond head." PHALANGES. The first digit contains two phalanges, the distal- of which bears a claw, functional in all bats except the two genera of Furipteridae, In the second digit the metacarpal alone is present in the Embal- lonuridse, Nycteridse, Rhinolophidae, Hipposideridae, Natalidae, Fu- ripteridae, and Thyropteridae. One distinct bony phalanx is present in all other families of Microchiroptera except the Rhinopomidae, in which there are two. Three phalanges are present in this finger in all of the Megachiroptera, the terminal bone in most genera provided with a claw. Three is the usual number of phalanges in the third, fourth, and fifth digits. The distal phalanx is always cartilaginous in the fourth and fifth, though in carefully prepared material its joint with the second phalanx may be detected when both are present. In the third digit the terminal phalanx is ossified in the Phyllosr tomidse, Desmodontidae, Thyropteridse, Myzopodidse, and Mystaco- pidae. When not completely ossified its joint with the second phalanx is at least clearly indicated and partially of bone, a condition readily observed in the Vespertilionidae and Molossidse. No distinct third phalanx is present in the third, fourth, and fifth fingers of any Megachiroptera, or among the Rhinopomidae, Emballonuridse, Noc- tilionidse, Nycteridse, Megadermidae, Rhinolophidae, Hipposideridae, Natalidse, and Furipteridae, though it is probable that the terminal cartilage, when present, usually, if not always, represents this bone. Sternum and Shoulder Girdle. The general characteristics of the sternum and shoulder girdle (Plate. XI, fig. 1; Plate XII, fig. 3; Plate XIII, fig. 2; Plate XIV, fig. 1) are as follows: Sternum. — The presternum is large, strongly keeled, and with three anterior lobes, one of which is vertical and continuous with the keel (from which it is often scarcely to be distinguished), the two others a In the genus Eptesicus both processes are unusually short. 25733— No. 57—07 m 2 18 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. horizontal, their substance thickened and strengthened to receive the articulation of the clavicles, and their combined length eq.ual to or greater than longitudinal diameter of presternum. Mesosternum with a longitudinal median ridge, occasionally rising to a distinct keel, the segments of the bone always fused in adults, and their bound- aries usually obliterated. In cross section the depth of this part of the sternum is usually greater than the width, but to this rule there are conspicuous exceptions. Xiphisternum short, tapering or somewhat expanded posteriorly, its length usually greater than its width, its terminal cartilage well developed. The distinction be- tween the presternum and mesosternum is usually evident, but that between the mesosternum and xiphisternum seldom persists except in the Megachiroptera. Shoulder girdle. — The scapula is large, oval in form, the post- scapular fossa much larger than the anterior fossa, its surface divided into three secondary surfaces set at slight angles with each other. The spine is short and moderately high, with a large, strong acro- mion. Coracoid large, usually curved outward, but occasionally straight and directed inward; rarely bifid at tip. Clavicle curved, its length about equal to that of scapula or of longest ribs, its shaft somewhat compressed, in one genus (Diclidurus) (Plate XII, fig. 3) conspicuously expanded. The articulation of the clavicle with the enlarged horizontal anterior lobe of the presternum is by a broad, unusually definite surface, from which the clavicles project upward and outward over the thorax nearly at right angles with each other. Ribs. — The ribs decrease gradually in length from the longest to the second; between the second and the unusually shortened, thick- ened first, the reduction is much more noticeable, pr'oducing a break in the otherwise uniform series. Both portions of the first rib are shortened, but the modification, as compared with the others, is most noticeable in the sternal part, which is usually much expanded later- ally. The vertebra to which this rib is attached is usually free, both anteriorly and posteriorly, but not infrequently it becomes ' fused with the last cervical vertebra, even when, as in the Molossidae, no special modification of the shoulder girdle has taken place. These two vertebree, together with the first rib, are so intimately associated with the shoulder girdle in the changes which it undergoes that they may conveniently be treated as forming part of it. Modifications of the shoulder girdle and sternum. — The modifica- tions presented by the shoulder girdle and sternum of bats are, as might be anticipated, mostly connected with the mechanical problems of furnishing surfaces of attachment for the very large pectoral "For the homologies of the elements included in this region see Leche, Bihang Svensk. Akad Handl., V, No. 15 ; also, Flower, Introduction to the Osteology of the Mammalia, 1885, p. 253. THE FAMILIES AND GENERA OP BATS. 19 muscles and strengthening the arch by which the body is suspended to the wings during flight. Somewhat remarkably, however, the largest species, the great Pteropines, 01 in which the mechanical diffi- culties would be expected to be greatest, show little specialization except for the rather unusual development of the keel, particularly on mesosternum, a peculiarity which is most pronounced in the larger members of the group. The elements of the first rib and its attach- ments remain distinct, and except for the usual Chiropterine pecu- liarities show no characters worthy of note. Essentially the same conditions, though with much less development of the keel, are found in most of the Microchiroptera. So far as my observations go it is constant in the Bhinopomidse, Emballonuridse, Noctilionidae, Phyllostomidse, Desmodontidse, Myzopodidae, and in most of the Vespertilionidse. The first stage in the strengthening process is found in the Molossidae, the Mystacopidae and in one subfamily of Vesper tilionidse, the Tomopeatinse. It consists in the slightly greater thickening of the first rib, and the fusing of the last cervical and first dorsal vertebrae into a solid ring. There is no appreciable change in the form of any of the bones, and the keel of the mesoster- num remains very slightly indicated. A peculiar modification of this process occurs in the Thyropteridae where the fusion takes place between the first and second dorsals, the last cervical remaining free. In the Natalidse it is again the last cervical and first dorsal that fuse ; but in this family the supporting arch is further strengthened by a noticeable widening of the horizontal lobes of the presternum, so that the greatest anterior width exceeds the length of presternum and mesosternum together. The first rib is also thickened and its sternal segment is greatly reduced in length. Keel of mesosternum relatively as high as in the Pteropodidae, and a slight keel on xiphisternum. The most remarkable series of changes occurs in the four related families Nycteridae, Megadermidae, Rhinolophidae, and Hipposide- ridae. In the Nycteridse the conditions are not very different from those in the Natalidae. The keel is equally high and it extends simi- larly- on the short, wide xiphisternum. The presternum is, however, less broadened and strengthened, and the last cervical vertebra is not fused with the first dorsal. In the Megadermidae the mesosternum and xiphisternum are of the ordinary type, but the presternum is greatly widened, so that it is essentially a heavy transverse bar with a slight median backward projection. Laterally it is fused with the first rib, so that the only break in the continuity of the ring is at the point of articulation of the ribs with the vertebras. As in the Nycte- t»The weight of an alcoholic specimen of Pteropus from Tenasserim (Cat. No. 104451, U. S. N. M.) is 739.5 grams, while that of a Rhinolophus ferrumequmum from Genoa, Italy (Cat. No. 18472, U. S. N. M.), is only 16.8 grams. Yet in the smaller animal the strengthening process is carried to the extreme. 20 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. ridse the first dorsal vertebra is fused with the last cervical. In the Rhinolophidse the seventh cervical vertebra and first dorsal are so completely fused that their boundaries can not be detected ; this com- pound vertebra is fused with the first rib, which in turn is fused with presternum. The second rib fuses with first at about its point of attachment with presternum, to which it is joined throughout the rest of its course by a thin sheet of bone. At first sight, therefore, the presternum appears to be enlarged to a broadly crescentic plate applied to mesosternum at middle of its convex posterior border. The original elements of this mass, though, perfectly fused, can still be distinguished. In the Hipposideridse the strengthening process attains its highest degree of perfection. Here the general structure is as in the Rhinolophidse, but the fusion of the first and second ribs involves the entire bone to and including the corresponding dorsal vertebra?. A solid ring is thus formed, including three vertebrae, two ribs, and the presternum, the elements indicated by a slit- like vacuity between the upper halves of the ribs, a small median space between the two dorsal vertebra?, and two minute roundish apertures in the region corresponding to space between horizontal arm of presternum and sternal portion of second rib. The mechanical need for this remarkable strengthening in bats of this size is by no means apparent. Teeth. Though much has been published concerning the teeth of bats, par- ticularly by Winge," 1 the subject is still by no means exhausted. This is especially true of the modifications of the various cusps of the per- manent cheek teeth, which have hitherto received little attention. Before passing to this more important aspect of the dentition, how- ever, a few words on the milk teeth and the tooth formula may not be out of place. MILK DENTITION. The peculiar and very highly specialized milk dentition of bats has long been known in a general way, but the material has not yet been brought together for a complete study of its variations. The largest number of milk teeth that has been observed is 22, with the , - 2 3. 1. 1 2 . 2 - 2 1-1 2 - 2 „ x . A . . formula: 1 ^ g ^ 1 ^ («w g _ g , mc 1 _ v mp _ 2 =22). As this is present in Myotis, a genus with the maximum number of per 7 manent teeth, it probably represents the complete set. It also occurs in Eptesicus, in which the corresponding portion of the permanent den- _ g 3_ j_ 4. tition is jo 3 1 ' _ o _ 4 an< i in Lasiurus with the permanent teeth ajorfundne og nulevende Flagermus (Chiroptera) fra Lagoa Santa, Minas Geraes Brasilien, 1892, pp. 56-58. THE FAMILIES AND GENERA OF BATS. 21 1 no i _ o _ A - I Q Nyctincmvus brasiliensis, with corresponding part of permanent formula the same as in Lasiurus, it also appears to be present, though I have been unable to verify the occurrence of mi 3 . A slightly reduced formula 1 „ _' ' 9 =20, appears to be charac- teristic of the Desmodontida? and Phyllostomidse, though the pos- terior tooth both above and below is ^occasionally so small that its presence can not be positively determined. In Desmodus the doubtful tooth is mp z , while in Phyllostomus it is mp 2 . In two specimens of Artibeus (Nos. 38309 and 38310, U. S. Nat. Mus., Old Providence Island, Carribbean Sea), on the other hand, I can find no trace of mp j. For the Megachiroptera the complete milk dentition is prob- ably the same as that of the Phyllostomidee. At least, I have verified - 2 3. 1. 1 2 the formula ..— k — '-.'-. =^0 in Cynopterus and Pterocyon. In both of these the posterior tooth above and below corresponds with pm *, and pm 4 , respectively, pm 3 and pm s being conspicuously without a preceding functional deciduous tooth. In form the. milk teeth are for the most part quite unlike those of the permanent dentition. They are usually slender, minute spicules, with straight or re- curved tips, and their only function is prob- ably, as has been frequently suggested, to aid the young in clinging to the mother during flight. In size they are so insignificant that one may not infrequently persist in the edge of an alveolus until the permanent tooth is nearly r _, ,. .,. P i e ' Fie. 1.— Milk dentition of grown, lhese peculiarities ot size and form, cynopteeus minutus, nias together with the very early development of island, no. 141271. x the permanent dentition, preclude all possi- bility that the milk teeth can "be used in mastication, and indicate that, whatever may be the case with other mammals, in bats the de- cidous teeth can have had little if any influence on the modifications of the permanent set. The simplest form of milk tooth is a straight terete spicule, tapering to a sharp point, and with no evident differentiation between root and crown. This is found in the lower cheek teeth of Cynopterus. The next and more usual type differs in the definite recurving of the point. This curve may be slight and gradual (canines and cheek teeth of Cynopterus (fig. 1) and Pterocyon, "upper canine of Myotis yuman- ensis) or strong and abrupt, sometimes almost angular (canines and cheek teeth of Phyllostomidse, canines and incisors of Nyctinomus brasiliensis) . Occasionally this type is further modified by a slight but evident shortening and thickening of the crown, or the faint indi- 22 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Pig. 2. — Milk dentition of Eptesicus fuscus. No. 84550. x about 4. cation of incipient secondary cusps (mp 2 of Pteroycon, cheek teeth of Phyllonycteris) . The presence of a definite secondary cusp below and behind the main cusp is probably characteristic of the lower cheek teeth in the Vespertilionidse. It is very noticeable in Eptesi- cus fuscus, less so in Myotis albescens. Similar cusps occasionally, though less frequently, occur in the upper teeth and in both upper and lower canine {Eptesicus (fig. 2), Myotis). In Nyctinomus bra- siliensis the, canine has an additional secondary cusp situated on the anterior edge of the tooth somewhat nearer the tip. Contrary to what might be expected, the milk incisors present more varia- tions of form than any of the other deciduous teeth. The simplest incisors are perhaps those of Desmodus, in which the upper teeth are of the primitive recurved type, without trace of second- ary curvature or extra cusps, while the lower teeth are widened above, the outer slightly bifid. In Nyctinomus brasilien- sis the upper incisors are simple and terete, but with a distinct double curve inward and backward, like those of Promops fosteri (fig. 3). The lower incisors are, however, much like those of the permanent set, with well differentiated, flattened, bifid crowns, the inner lobe dis- tinctly the larger. In Eptesicus and Myotis both upper and lower incisors are trifid, those of the lower jaw scarcely recurved and very closely resembling the teeth of the permanent set, those of the upper jaw strongly recurved and with the cusps longer and more distinct than in the mandibular teeth. In the three genera of Phyllostomidae at hand, Phyllostomus, Artibeus, and Phyllonyc- teris, there is evident differentiation between the inner and outer incisor of the upper jaw. This is slightest in Phyllostomus, where the outer tooth is longer than the inner, its point bent forward and then back, the concavity directed outward and backward; inner tooth with crown flattened but tapering to a fine recurved point below which there is a slight concavity on outer side. Lower incisors not seen. In Arbiteus the upper teeth are of much the same form, but the inner is distinctly notched at apex, much as in its successor; lower incisor faintly trifid, the middle lobe highest. Upper incisors of somewhat the same type are found in Phyllonyc- teris, though the differentiation is less evident. Inner tooth with a distinct secondary cusp on outer side considerably below level of main cusp. The lower teeth are so minute that the details of their form can not be determined. The upper incisors of Cynopterus and the Fig. 3. — Milk denti- tion of PBOMOFS FOS- TERI, Villa Rica, Paraguay. No. 105681. x about 3. THE FAMILIES AND GENEKA OP BATS. 23 lower incisors or Pterocyon are of the simple, little recurved type; the lower incisors of the former are straight, with bluntly rounded, slightly thickened crowns. Upper incisors of Pterocyon not seen. DENTAL FORMULA. The nearest approach in bats to the typical complete dentition of higher mammals — that is, a set of 44 teeth — as expressed by the formula 1 2 3.1.128*567 .3-3 1-1 4-4 3-3 _ M t0rmula i23. 1.12345 67* 3^3' C T=V ^ m 4^4' m 3^3 ~ 44 ' is a set of 38, lacking one upper incisor and one upper and one lower - 2 3. 1. - 2 3 4 5 6 7 00 _, ., . , premolar: -ioq i _ q q 4, k a y = 38. lhe smallest number known is 20, represented by the formula ~ ~" ' ~~~ , „ = 20. Between -2 3. 1. -2-45 — these extremes occur no less than 50 formulas, representing totals of 24, 26, 28, 30, 32, 34, and 36. No bat is yet known with 22 teeth. The following table shows the distribution of the varipus total numbers of teeth among the 173 genera now recognized : 24 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. limits-fill 1 1 1 1 § 1-1 i , - III -If 1 III iillrlli asIiII in fin mill I rtiiiinins s . •S8 (go So § 111 i lit fllllltf litilhilf lis i Iiili S.s saesg S*£ ffs- 5:8.8 g o=s = o?§ a view's MS 2 s lilriiiiidllUl Htll c^d THE FAMILIES AND GENERA OF BATS. 25 The different formulas, with the genera by which they are repre- sented, are as follows : ■-•;:•>. I. - 2 :; 4 5 6 7 =38 Nyctie u USt Natalus, Chilonaialus, Phodoies, Thyrop- tera, Myzopoda, Myotis, Pizonyx, Kerivoula, Pho- nisous. 36. JPuripierus, Amorphochilus. 36. Lasionycteris, Plecotus, Corynorhinus, Miniopterus. =36. Odontonycteris. 34. Rousettus, Pterocyon, Pteropus, Acerodon, Desmalopex, Pteralopex, Eonycteris, Kiodotus, Syconycteris, Try- genycteris, Melonycteris. =34. Chilonycteris, Pteronolus, Mormoops, Microny clevis, Xenoctenes, Olyphonycteris, Otopterus, Lonchorhina, Dolichophyllum, Phylloderma, Trachops, Vampyrus, Glossophaga, Lonchophylla, Monophyllus. 23 1 —2—4567 =34, Pipistrellus, Glischropus, Scotozous, la, Pterygisles, Chalinolobus, Barbastella, Euderma, Murina, Har- piocephalus. =34. Emballonura. =32. Boneia. 32. CaUinycteris. =32. Nesonycleris. =32. Goleura, Rhynohiscus, Saccopteryx, Cormwra, Perop- 12 3. 1. -234567 - 2 3. 1. - - 3 4 5 6 7 12 3. 1. -234567 -2 3. 1. -2-4 56 7 12 3. 1. -234567 -2 3. 1. -23456 7 12- 1.-234567 -2 3. 1. -23456- 12-. 1. -234567 -2 3. 1. --34567 12-. 1. -234567 12 3. 1. -2-4567 -2 3. 1. -2-4567 12 3. 1. -2-45 6 7 -2-. 1. -2345 12-. 1. -234667 -2 3. 1. -23456 12-. 1. -23456 -2 3. 1. -23456 -2-. 1. -234567 - 2 -. 1. - 2 - 4 5 6 7 12 3. 1. -2-4567 -2 3. 1. 456 7 7 -=32. Rhinolophu, 12 3. 1.-2-45 67 -2-. 1.-2-4567 12- 1. -234567 -2 3. 1. --345 67 1--. 1. -234567 -2 3. 1. -234567 . 1. -234567 -2 3. 1. --34567 12-. 1. -2-4567 teryx, Peronymus, Centronycteris, Balantiopleryx, Diclidurus. =32. Nycteris. ■ 32. Tonatia, Chrotopterus. 32. Lonchoglossa, Anoura. =32. Phyllostomus, Hemiderma, Rhinophylla, Brachy- phylla, Artibeus (part), Enchisthenes, TJroderma, Vampyrops, Ardops, Phyllops, Stenoderma, Stur- nira, Phyllonycteris, Reithronycteris, Erophylla. 3 =32. Eptesicus, Vesp'ertilio, Rkinopterus, Hesperoptenus, 12 3. 1. -2-4567 -2-. 1. -2-4567 12 3. 1. -2-4567 - 2 -. 1. -23456 (2) -. 1. - 2 3 4 5 6 7 Tylonycteris, Mimetillus, PhUetor, Hwtiotus, Lsepho- tis, Glauconycteris. =32. Lasiurus, Nyctinomus (part). 30. Harpyionycieris. 26 BULLETIN 57, UNITED STATES NATIONAL MUSEUM -2 3. 1. -2345 12-. 1. -23456 -2 3. 1. -2345 6 -2-. 1. -23456- -2-1. -2-4567 12-. 1. -2-4567 . 1. -2-4567 12 3. 1. -2-456 7 -2-. 1. -2-4567 =30. Cynopterus, Niadius, Thoopterus, Sphssrias. =30. Balionycteris, Styloctenium. 30. Hipposideros, Anthops, Cwlops, Trisenops. . 30. Myropteryx. 1 2 -.' 1. - 2 - 4 5 6 7 =30. Taphozous. -2 3. 1. --3456 7 =30 Mimon AnthorMna. 1--. 1. -2-4567 -2 3. 1. T--34567 .1.-234567 -2 3. 1. --3 45 6- 1 2 -. 1. - 2 3 4 5 6 - -2 3. 1. --3456- 12-. 1. -2-4567 -2 3. 1. --3456 12-. 1.-2-456 7 =30. Chceronycteris, Hylonycteris. 30. Leptonycterls. 30. Artibeus (part). =30. Vampyrodes, Mesophylla, Ariteus. --l.il: '_3o_ Otonycteris, Nyct'ceius, Scoleinus, Scotcecus, Scoto- 12^ 1 -2-4567 i u u i manes, Rkoge'issa, Bieodon, Pachyotus, Dasypterus, Nyctophilias, Mormopterus (part). ~ 1 =30. Chserephon, Kyctinomus (part) Eumops (part), Pro- 1 Jt — . 1. — Zi — 4 y O I mops. _23 1 — 2345 —j- — '-—^ — — =28. Ptenochirwi, Megserops. — 2 — . 1. — 2 o 4 5 b — -2 3. 1. --345-- 1 2 -. 1. - 2 3 4 5 6 -' =28. Scotonycteris, Epomophorus, Hypsignathus. " "• ' - - 3 4 5 6 " = 28. Dobsonia. -2-. 1. -234567 -2 3. 1. -2345 - 2 -. 1. - 2 3 4 5 6 - -2-.1. 4567 12-. 1. -2-4567 =28. Nolopteris. =28. Rhinopoma. ;::'..!. 4 5 6 7 =28 NoctUio pirias. 1--. 1. -2-4567 .1.-2-4567 12- 1. -2-4567 -2-.1. 4567 12-. 1. -2-4567 -2 3. 1. --3 4 5 6- 1 2 -. 1. - 2 - 4 5 6 -" -2-1. 4567 1 2 -. 1. - 2 - 4 5 6 7 : -2-. 1. -2-4 567 1--. 1. -2-4 5 6 7' . 1. 4567 12-. 1. -2-4567 -2 3. 1. --3456 . 1. -23456- -2 3. 1. --3456 1--. 1. -2-456 =28. Megaderma, Lyroderma. =28. Asellia, Rhinonycteris, Cloeotis. =28. Artibeus (part), Vampyressa, Chiroderma, EctophyHa, Pygoderma, Centurio, Sphseronycteris, Ametrida. =28. Antrozous, Tomopeas, Molossops (part), Mormopterus (part), Platymops, Eumops (part). =28. Mystacops. =26. Macroderma, Lama, Cardioderma. =26. Lichonycteris. 26. Vampyriscus. THE FAMILIES AND GENERA OF BATS, 27 - 2 3. 1. 4 5 6 - OA „.. , „ 12-l.-2-456- = 26 - mphylla - 5 =26. Eomops, Molossops, (part), Cheiromeles, Molossus. 1--. 1. -2-4567 -2-. 1. -2345 .1.-2 345 6- -2-. 1. 45-- 12-. 1. - 2 - 4 5 - ■ =24. Nyctymene. 20. Desmodus, Disemus. It will be noticed at once that certain foramlas are characteristic of particular groups, or rather that a given formula never occurs in widely separated families. Thus, among the genera with 34 teeth the formula 7—= - — ' ' ~ „ . „ „ Z is confined to the Pteropidae, 12- 1. -234567 r ' -2 3. 1. --34567. ., D , ,, + -j ,-2 3. 1. -2-4567 12-1.-234567 t0 the ^^tormd*, and t a 8 . L _ a _ 4 6 6 7 to the Vespertilionidse. Similarly among the 32-toothed genera the formula ~ „ ~ — ' ~ ~ , „ , „ is peculiar to the Embalionuridae, 12 3. 1. -2-4567 y -2 3. 1. --34567. +u Tn.ii •* -^ j-23. 1. 4 5 6 7 12-1.-2-4^67 t0 the Ph y llostomid ^ and r23 n.-2-4567 to the Vesper tilionidse. The only formulas that occur in two or more families are 7 I j*' *' ~ !? « f I? j? 1 =38, found in the Natalidee, Thyr- 12 3. 1. — 234567 _ 2 — 1 4567 opteridse, Myzopodidse, and Vespertilionidse, ' ' — - — — — -— , - =30 found in 9 genera of Vespertilionidse and 1 of Molossidse (Mormop- _o_ i 4567 tencs), and ^— — '— - — - — ^ =28, found in two aberrant genera each of Vespertilionidse (Antrosous and Tomopeas) and Molossidse {Molossops, part, and Mormopterus, part). Even in these excep- tional cases, however, the families involved are not distantly allied. The only teeth that are never absent are the canines, the last pre- 1 t 45 molar and the first molar, ' ..* . ~ those never present the first upper incisor and the first upper and lower premolar, -231—234567 ., „ * ' — „ „ . K „ f, . The reductions by which the various formulas 13 3. 1. -234567 ' have been produced appear to be as follows : Incisors. — In the upper jaw the permanently missing incisor is without- much doubt the first. This is indicated, as shown by Winge, by the correspondence of the two upper teeth with the two outer of the lower jaw when the maximum set is present, and also, even more strongly, by the general tendency throughout the group for the pre- maxillaries to become reduced, particularly along the inner edge. This would inevitably result in eliminating that part of the bone in which the first incisor grows. Of the remaining upper incisors it is 28 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. invariably the outer that disappears first. This tooth may be seen in every stage of degeneration among the genera with 2-2 upper incisors, while in none is it distinctly more developed than its fellow, though in some of the long-tongued Pteropidse and in certain genera of Gloss- ophaginse it is slightly the larger. In the lower jaw the incisors become reduced from the outer side, a process mostly associated with narrowing of the anterior portion of the mandible. This process is especially well illustrated in the Molossidse. The third incisor exists in some species of Nyctinomus and in Mormopterus, though reduced in size and crowded beneath the cingulum of the canine, the prominent cusp of which has assumed its function. In Chcerephon, Eumops, Promops, and most species of Nyctinomus it has disappeared. In some species of Molossops the narrowing of the mandible has reduced and distorted the second incisor, the cingulum cusp of the canine in turn taking the functions of this tooth, while in other species of Molossops, in Eomops, Cheiro- meles, and Molossus the second incisor has disappeared. There is little doubt that a similar course" has been followed throughout the Microchiroptera. In the Megachiroptera no genera are known with 3-3 lower incisors, and it has been assumed by Winge that the first tooth is absent, a conclusion based chiefly on the correspondence in position of these teeth with, the second and third of the upper jaw. The probability of this view seems heightened by the almost univer- sally larger size of the outer tooth as compared with the inner, while in the Microchiroptera the reverse is normally the case. In the fruit- eating Phyllostomidse, however, there is a similar reduction in the size of the inner incisor as compared with the outer, probably due to the aotion of the tongue, and I prefer to assume that the frugivorous habits of the Pteropidse account for the relative size of these teeth also, and that the course of reduction in this group forms no exception to the rule, so far at least as regards the disappearance of the outer tooth first. The next incisor to disappear is, however, probably i t , acted, upon, as it must have been, by the tongue. In the genera Dobsonia, Nesonycteris, and Notopteris i , is therefore the remaining tooth. Premolars. — Both above and below it is probably the first pre- molar that is permanently absent, though of this there is no proof. In the upper jaw the next to disappear is either pm - or pm 6 . Three premolars are present in members of the families Pteropidse, Phyllostomidse, Natalidse, Thyropteridse, Myzopodidse, and Vesper- tilionidse. The anterior tooth {pm 2 ) is the more reduced in four of these, the Pteropidse, Phyllostomidse, Natalidae, and Thyrop- teridse, while the median (pm 3 ) is the smaller in the two others, ihe Myzopodidse and Vespertilionidse. While the evidence is there- fore not conclusive for any of the other families, it appears safe to THE FAMILIES AND GENERA OF BATS. 29 assume, from general 'considerations of relationship, that the small upper premolar is pm 3 in the Furipteridse and pm 2 in the Nyc- teridse, Megadermidse, Bhinolophidse, Hipposideridse, Mystacop- idse, and Molossidse. It seems probable that the degeneration of pm 2 is characteristic of the fruit-eating bats and their near relatives and that the reduction and disappearance of pm 3 occur in the insec- tivorous groups. As regards the small upper premolars, therefore, the families may be grouped as follows : Dominant tooth, pm * : Dominant tooth, pm * : Megadermidse. Pteropidse. Rhinolophidse. Phyllostomidse. Hipposideridse. Natalidse. Myzopodidse. Furipteridse. Vespertilionidse. Thyropteridaa. Mystacopidse. Molossidse. In the lower 'jaw pm 3 becomes reduced before pm 2 in all the groups of Microchiroptera, some members of which possess both teeth — the Khinolophidse, Phyllostomidse, Natalidse, Furipteridse, Thyropteridse, Myzopodidse, and Vespertilionidse. Hence it is prob- able that this is the normal process throughout the suborder. In the Megachiroptera, however, the opposite is taking place. In every known genus of Pteropidse both pm 2 and pm 3 are present, but pm 2 is invariably the smaller tooth. Molars. — While the posterior molar of both jaws is invariably reduced, it is never absent except in certain frugivorous and sanguiv- orous bats; that is, in various genera of Pteropidse and Phyllosto- midse, and in all of the Desmodontidae. The middle molar dis- appears in some of the Pteropidse and Desmodontidse, but is not yet known to be absent in any member of the Phyllostomidse. The vari- ous ways in which these teeth are suppressed are as follows : 5 6 - Harpyionycteris, Dobsonia, Artibeus (part), Vgmpyrodes, Mesophylla, 5 6 7' Ariteus. 5 6- Balionycteris, Styloctenium, Leptonycteris, Lichonycteris, Artibeus (part), 5 6-' Vampyriscus, Vampyressa, Chiroderma, Ectophylla, Pygoderma, Centu- rio, Sphseronycteris, Ametrida, Diphylla. 5 — Nyctymene, Oynopterus, Niadius, Thoopterus, Ptenochirus, Megserops, Sphse- 5 6-' Has, Scotonycteris, Epomophorus, Notopteris. g t- : Desmodus, Diazmw. NORMAL CUSPS. The typical Chiropterine dentition, which occurs essentially unmod- ified in at least some genera of every family except the Pteropidse and Desmodontidse, is as follows : Upper incisors subulate, the point slightly curved backward; cin- gulum well developed, particularly on outer side; a posterior or lat- eral secondary cusp usually present on one or both teeth. Lower 30 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. m ts incisors with crown well differentiated from root, its length greater than height or width, its cutting edge approximately horizontal, but with two evident notches. No striking differences in size or form between the teeth composing a series, except that the outer lower incisor is often wider than the others and with one or two blunt sup- plemental cusps posteriorly. Canines simple, the upper larger than the lower and usually the highest tooth in the entire series. Cingulum well developed, that of the lower tooth conspicuously oblique. Shaft subterete, that of the upper tooth flattened or slightly concave on inner side, that of lower tooth similarly flattened or concave posteriorly. The shaft of the upper tooth at base occupies almost entire crown, while that of lower appears to be situated a little in front of middle of crown so that the posterior portion P c f\ C forms a slight heel. pre Premolars except pm * essentially like the canines, though very much smaller and with height of crown not conspicu- ously greater than p IG , 4.— Typical molar teeth of an insectivorous bat. A.— length. Cmgulum Crown view of maxillary molak. B.— Crown view of y^-gll develoDed hori- TWATjnTWTTT.AR Mm.AP f!: — KtDF. VIEW OF MAXILLARY MOLAR, -y ' zontal or slightly ob- lique. Posterior up- per premolar much larger than either of the others, its cusp nearly as high as that of canine, from which it differs hi shape in the presence of a conspicuous flange-like postero-external extension, with well-developed cutting edge and supported by a second root. Inner side of this extension flat, continuous, with flattened postero-. internal surface of cusp; crown with a slight but evident postero- internal heel; cingulum well developed, often forming an anterior cusp at base of main cusp and occasionally another on heel. The first and second upper molars resemble each other, though the second is usually the larger of the two. The crown (fig. 4) is much wider than long or high, three-rooted, and set obliquely, so that the outer portion is higher than the inner. 6 Its outer border is a Length = diameter in axis of tooth row; width=diarneter perpendicular -to tooth row ; height=distance from lower edge of cingulum to extremity of high- est cusp. 6 Strictly speaking, it is lower, more ventral, but in the position in which the teeth are always examined it is higher. MANDIBULAR MOLAR. C— SIDE VIEW OF MAXILLARY MOLAR. ecd. — ENTOCONID. Wis. — METASTYLE. kC. — HYPOCONE. pC. — PARACONE. ftcd— HYPOCONID. pcd.— PARACONID. mc. — METACONE. pre. — PROTOCONE. mCd. — METACONID. prCd. — PROTOCONID. ms. — MESOSTYLE. pS. — PARASTYLE. THE FAMILIES AND GBNEKA OF BATS. 31 nearly straight, though usually marked by three projections and two indentations. The anterior border is unsymmetrically convex, the inner border variously truncate, rounded, or double convex, the posterior border with a concavity usually somewhat greater than con- vexity of anterior border. The inner posterior portion of the crown, bearing no large cusp, is usually flattened and often produced into a noticeable heel. The crown bears three main cusps, corresponding to those of the primitive tritubercular tooth, the inner anterior proto- cone (fig. 4, pre), the outer anterior paracone (fig. 4, pc), and the outer posterior metacone (fig. 4, roc). Not infrequently the "inner posterior hypocone (fig. 4, he.) is also present, but in bats this cusp never attains a size sufficient to obscure the tritubercular aspect of the crown. Very rarely (in some Phyllostomidse, particularly Brachy- phylla, Plate VI, fig. 3) an intermediate cusp, the protoconule, occurs between protocone and paracone, and another, the metaconule, at inner base of metacone. At the extreme outer edge of the tooth are three small cusps, the anterior parastyle (fig. 4, ps.), the median mesostyle (fig. 4, ms.), and the posterior metastyle (fig. 4, mts.) a . The styles are connected with the main cusps of the outer row by four conspicuous, trenchant, slightly concave ridges, or flu tings, the com- missures (fig. 4), the first extending from parastyle to paracone, the second from paracone to mesostyle, the third from mesostyle to metacone, the fourth from metacone to metastyle. The commissures are approximately equal in length, though increasing slightly from first to fourth. Together with the cusps, which they connect, these ridges form a conspicuous W-pattern, the variations in the form of which are of much systematic importance. Of the three main cusps the protocone is situated at a lower level than the others. In form it is more robust, though usually less elevated; frequently it occupies nearly the entire inner section of the tooth. A narrow commissure extends forward from anterior side of this cusp, past base of para- » In describing the molar teeth I have adopted the cusp nomenclature pro- posed by Osborn (See American Naturalist, XXII, p. 1072, December, 1888) as the most simple and convenient. The fact that it was based on a mistaken idea of the succession of cusps (See Gidley, Proc. Washington Acad. Sci., VIII, p. 106, July 10, 1906) is of little weight compared with the convenience of an exact name for each part of each tooth. The system of numbering the cusps (cusp No. l=ps, No. 2=ms, No. 3=m£s, No. 4=pc, No. 5=otc, No. 6=pc, No. 7=ftc, etc.) proposed by Winge (Vidensk. Meddel. Naturhist. Foren., Ki0ben- havn, 1882, pp. 15-19, pi. in) and recently adopted by Andersen and Wroughton (Ann. Mag. Nat. Hist, 7th ser., XIX, p. 129, February, 1907) appears to be equal!/ misleading as regards the succession of cusps, while it has the great disad vantage of furnishing no convenient names. With regard to the position of the primitive cusp: Winge placed It in the outer row (mesostyle), Osborn in the inner row (protocone), while the observations of Gidley, based on much more extensive material, shows almost conclusively that it is in the middle row (paracone). 32 cone to parastyle. A similar commissure occasionally extends pos- teriorly behind base of metacone to metastyle, but this is less con- stantly present. The paracone and metacone form the highest portion of the tooth; the metacone, together with third and fourth commissures, is usually larger than the anterior cusp and ridges. A faintly developed cingulum may usually be traced along anterior, inner, and posterior edges of crown. On outer side the cingulum is practically absent. The third molar is always smaller than either of the others, and some of its elements are reduced or absent. The process of reduction, which invariably proceeds from behind forward, varies, in teeth that can not be regarded as abnormal, from a mere shortening of the metastyle and fourth commissure, accompanied by diminution in height and diameter of the metacone, as seen in cer- tain Rhinolophidse (Plate I, fig. 1), to absence of everything except the parastyle, paracone, protocone, the first commissure, and a trace of the second, elements which are always present in this tooth. This condition is well illustrated by Pachyotus (Plate I, fig. 2). In the mandible the first and second molars are also alike in form, with the second usually a little the larger. Five cusps are present, the outer anterior protoconid (fig. 4, prcd.), the inner anterior para- conid (fig. 4, pcd.), the inner median metaconid (fig. 4, mcd.), the outer posterior hypoconid (fig. 4, hcd.), and the inner posterior entoconid (fig. 4, ecd.). In general appearance these teeth resemble the outer higher portion of the upper molars reversed, the three inner cusps corresponding in form to the styles, and the protoconid and hypoconid to the paracone and metacone. The hypoconid is, however, lower than the protoconid, though the width of the tooth through the base of this posterior cusp is normally greater than that through the protoconid. The points of the cusps of the two rows are nearer together than in the 'upper teeth, and the commissures are shorter and more strongly concave, the third extending down- ward from point of hypoconid to base of metaconid. The two seg- ments of the W are therefore disconnected, though this is not at first sight apparent when crowns are viewed from above. A well developed though not very prominent cingulum extends around outer base of crown from paraconid to entoconid. Third lower molar with posterior segment usually much smaller than the first, owing to the great reduction in size of both hypoconid and entoconid and the close approximation of these two cusps. In some Rhinolophidse and Emballonuridse, however, the tooth is practically identical with to ! and m 2 (Plate II, fig. 1). In the first and second upper molars of insectivorous bats the metacone may be said to be the dominant cusp. It is not only the largest and the first to appear above level of alveolus as the teeth grow, but in the modifications to which the form of the crown is subjected the metacone invariably THE FAMILIES AND GENERA OF BATS. 33 remains the most important cusp, drawing the protocone and para- cone toward it, so to speak, and gradually absorbing them. This process may continue until the molars approximate the large premolar in form, though no genus is yet known in which the protocone and paracone are entirely obliterated. In the third upper molar, how- ever, the paracone is the more permanent, as the successive short- ening of the tooth from behind soon eliminates the metacone. In all the mandibular molars the proto'conid is the chief cusp. Like the metacone of m x and m z it is the first to appear as the permanent teeth cut the gums, while it is the largest and most conspicuous cusp throughout all changes known to occur. INTERRELATION OF TEETH. Owing to the high cusps and deep hollows which cause the crowns to fit closely interlocked when the jaws are closed the interrelation between the teeth of the two jaws is very intimate, so much so that no modification can take place in a given region without its counter- part in the opposed structures. Lateral motion of the mandible, though supposed by at least one careful observer to be absent, un- doubtedly occurs to the extent of permitting the posterior surface of the lower canine to come in contact with the anterior surface of the corresponding upper tooth. In individuals with worn teeth the front face of the upper canine shows abrasion that could not otherwise be accounted for. This lateral motion is sufficient to give the opposed cusps and commissures of the molars the shearing action on whhh their effectiveness depends. In the typical condition with the maxi- mum number of teeth present the interrelation of the two sets are as follows : Incisors. — The inner upper incisor is opposed chiefly to i ,, the large median lobe of which fits into the notch formed by basal cusp of upper tooth ; extreme tip of i - nearly in contact with inner edge of median lobe of i ,. Outer upper incisor with long posterior sur- face in contact with anterior surface of lower canine, the two teeth playing against each other with a shearing motion as the jaws close. The extreme tip of the tooth almost touches notch at posterior outer base of i z when jaws are tightly shut. Inner lower incisor not op- posed to any tooth in upper jaw. Canines — The chief opposition of the canines is a shearing contact between the anterior inner edge of the upper tooth and the posterior outer edge of lower. The point and inner surface of the upper tooth also plays against the first lower premolar, or rather its inner pos- terior edge cuts any food material pressed upon it by the two small « Gosse, Ann. and Mag. Nat. Hist., XX, p. 426, December, 1847. The animal observed was a Noctilio, and the motion of the jaws is described as vertical only. 25733— No. 57—07 M 3 34 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. premolars together. In the genera with large, triangular- crowned trenchant premolars (most Phylldstomidae, Natalidse, etc.) the inner posterior edge of the upper canine has become knife-like. It is crossed by the anterior edge of the anterior lower premolar with a true shearing motion, while its cingulum is so formed as to fit the point and posterior edge of the smaller tooth. Premolars. — In bats with terete premolars the upper and lower teeth do not come actually in contact, those'of the mandible lying distinctly nearer to the sagittal plane than those of the upper jaw. The point of pm 2 comes slightly below and behind that of pm 3 , while that of pm i approaches pm 3 and the cingulum of pm 2 The large upper premolar {pm 4 ) shears with its interior inner sur- face close to but not in contact with the posterior outer surface of pm „ while its main cusp and posterior cutting edge play against the cusp and anterior outer surface of prd 1 °, the extreme point of which nearly touches surface of heel. When the premolars are trenchant the anterior edge of' each lower tooth shears against the posterior inner edge of the preceding upper tooth, while the posterior edge cuts upon the anterior edge of the corresponding tooth in the upper jaw. The relations of the first lower premolar to the upper canine have already been explained. Molars. — As the distance between the two row's of mandibular teeth is sensibly less than that between the opposed sets, 6 it follows that to obtain a uniformly shearing action of the molars the jaws must be thrown either to the right or left before the beginning of the stroke. When the mandible is at its extreme lateral position the outer and inner cusps of the lower teeth are closely opposed, respectively, to the styles and highest cusps of the upper teeth — that is, "in the outer rows; prcd 1 to ps 1, hcd 1 to ms 1, prcd 2 to mts 1 and ps 2, hcd 2 to ms 2, prcd 3 to mts 2 and ps 3, hcd 3 to ms 3, in the inner rows, pcd 1 to the apex of pm*, mcd 1 to pc 1, ecd 1 and pcd 2 to mc 1, mcd 2 to pc 2, ecd 2 and pcd 3 to mc 2, mcd 3 to pc 3, ecd 3 to mc 3. The triangles of the lower teeth with their concave blade-like edges (commissures) tilted slightly inward, and termi- nated externally by the recurved cusps face the spaces between the similarly concave but outward-tilted edges of the upper triangles, each lower triangle lying in front of the corresponding triangle of o For the sake of brevity the cusps of the three molars may be referred to by their abbreviation followed by the numbers 1, 2, and 3. Thus prd l=proto- conid of first lower molar ; mc 3=metacone of third upper molar. » In a specimen of Noctilio (No. 37435, Mona Island, West Indies), the greatest distance between outer edges of upper toothrows is 12.2 mm. ; between outer \edges of lower toothrows, 8.9 mm. In Myotis californicus (No. 25826, San Diego County, California) the same measurements are, respectively, 4.9 mm. and 2.8 mm. THE FAMILIES AND GENERA OF BATS. 35 the upper tooth. As the jaws close the cutting edges pass each other with a shearing motion, c 1 working against posterior edge of pm *, c 2 against c 1 , c 3 against c 2 , c 4 against c 3 , and so on to the end of the series. At the same time a less important though equally definite action takes place between the innermost cusps of both sets of teeth. This consists principally of the passing of the posterior inner side of mcd 1, 2, and 3 and anterior inner side of ecd 1, 2, and 3 across the inner surface of the three protocones, the. point of the large cusp in each instance fitting closely to the angle between the two small ones. The paraconid of m 2 and m , also comes within cutting dis- tance of the hypocone of to 1 and m 2 when it is large enough to be functional. As the stroke finishes each hypoconid scrapes through the hollow at middle of opposed crown, and finally comes in oppo- sition with inner side of corresponding protocone, across which it passes as the jaw is moved downward and toward the opposite side to take its position for the next stroke. At the same time each protoconid passes through the space at front of its corresponding upper tooth and comes similarly in opposition with posterior heel of the tooth in front, or with its hypocone when this cusp is present. While the full stroke just described is taking place on one side, a somewhat reversed half stroke occurs on the other, the inner surface of the long outer cusps of the lower teeth cutting against the outer side of the opposed inner cusps of the upper teeth, the action between them being the exact reverse to that which takes place near the end of the full stroke. As the recovery for the next stroke proceeds, the rounded convex outer surface of the protoconids and hypoconid s pass through the depressions between the paracones and metacones, while the similarly rounded inner surfaces of the upper teeth pass between the main cusps of the lower teeth, the result being a grind- ing and crushing action which must be very effective in the final trituration of small particles of food. The foregoing description of the molars in action is primarily based on the genus Noctilio, in which the molars are large enough to be readily examined and in which their effectiveness of structure is at its maximum. The cutting apparatus is essentially the same in all bats with normal teeth, though its apparent effectiveness is in some instances reduced without any considerable change in the form of the ■cusps. Thus,-in Myotis myotis the protoconids and hypoconids are so long and straight that they are less readily brought in contact with the styles, the stroke apparently losing thereby. In some Phyllosto- midse the lateral motion of the mandible is lessened by the large size of the canines, which in certain genera develop large cutting edges. When this occurs the outer portion of the upper molars appears to be invariably reduced, the corresponding change in the lower teeth showing itself in the lessened transverse diameter of the triangles. 36 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. This is well illustrated by the genera Anthorhina (canines small, lateral motion of mandible free, outer portion of upper teeth well developed), Phyllostomus (canines large, with strongly trenchant edges, lateral motion reduced, outer portion of upper teeth relatively less developed), and Vampyrus (canines very large, lateral motion nearly absent, outer portion of upper teeth noticeably degenerate), a series which also shows the exceedingly nice adjustment between the different parts of the cutting apparatus, since the gradually increas- ing size of the canines is closely followed by alterations in the cusps throughout the series of molar teeth. MODIFICATIONS OF CUSPS. While the majority of bats retain, together with their insectivo- rous habits, the typical primitive dentition almost unmodified, the members of certain groups show conspicuous changes in the form as well as in the function of the teeth. These alterations, which pro- ceed chiefly by suppression of the original cusps, follow two main lines, one leading to a flat-crowned type of tooth effective for crush- ing the pulp of fruits, the other to a narrow,, blade-like form most perfectly developed in those bats which subsist on blood. Among insectivorous bats a third line of divergence seems to be incipient. In this the molars tend to assume a single-cusped form closely resembling the fourth premolar. This is accomplished by the reduction of the styles and commissures and by the blending of the protocone and paracone with the metacone and of the paraconid and metaconid with the protoconid. In the most extreme instance known, however, it is not complete, as the three cusps are still recognizable. The first step in this direction is seen in various unrelated genera and consists in the reduction of the mesostyle together with the second and third commissures, accompanied by a distinct drawing together of the three main cusps, a noticeable diminution in the size of the protocone and paracone, and the partial fusion of the latter with the metacone. In the mandibular teeth the corresponding change is seen in the reduced area of the triangle formed by the protoconid, paraconid, and metaconid as compared with the basal area of the crown. This condition occurs in the Nycteridse, Megadermida;, the genus Vampyrus (Plates I, II, fig. 3) and in Pachyotus (Plates I, II, fig. 2). In the Megadermidse it shows various degrees of develop- ment, being well advanced in Megaderma, Macroderma, and Cardio- derma, merely suggested in Lavia. In Pachyotus (Plates I, II, fig. 2) it has reached the same stage as in Megaderma in the maxillary teeth, while the mandibular teeth are slightly further advanced, hav- ing the second triangle noticeably smaller than the first. The next stage occurs in the genus Harpiocephalus (Plates I, II, fig. 4). In this the mesostyle has disappeared, together with the second and third THE FAMILIES AND GENERA OP BATS. 37 commissures, and the parastyle and metastyle are closely approxi- mated, though distinct and unusually large. The three main cusps are much closer together than in any other genus of bats, and the protocone and paracone are so reduced as to appear like mere append- ages to the large metacone. The whole tooth thus rather closely resembles a normal fourth premolar, the first and fourth commissures representing the anterior and posterior cutting edges, respectively, and the metacone the main cusp. In the lower molars all of the cusps except the protoconid are so reduced that the teeth approximate the form of the lower premolars. While Harpiocephalus represents the extreme of this peculiar variation so far as at present known, it would not be surprising to find genera in which the metacone alone remains, thus forming the complete parallel to the similar changes which have resulted in the suppression of all but the paracone in such genera of Insectivora as Centetes, Ericulus, and others. The next and more important series of changes is that leading to the flat-crowned crushing tooth. This type of dentition, character- istic of fruit-eating bats, reaches its highest development in the Pteropidse, but is found in various intermediate conditions in mem- bers of the Phyllostomidse. As no transitory stages are known in the Pteropidse, the conditions in the Phyllostomidse present much the greater interest. Among the Chilonycterinse and Phyllostominse, the two most prim- itive groups of the Phyllostomidse, the teeth usually show no special modifications (Plates III, IV, fig. 1). The premolars, and occa- sionally the canines, are trenchant, and the molars retain their outer cusps and commissures well developed and functional, except where the lateral motion of the jaws is lessened by the excessive development of the canines, the extreme of which condition is seen in Vampyrus (Plates I, II, fig. 3). Such modification as occurs tends, therefore, rather toward the narrow, blade-like type. In the Glossophaginse the first stage in the flattening process is clearly seen (Plates III, IV, fig. 2) . The molars of the upper jaw have nearly lost the mesostyle, while the parastyle in m 1 and in m 2 is prac- tically indistinguishable from the base of the paracone. The meta- style, however, remains long, though low; fourth commissure well developed, but the other three reduced to a mere trace. In m 3 the parastyle remains long, while it is the metastyle that has disappeared. Paracone and metacone well developed, the metacone, as usual, the larger. Protocone low and broad, extending backward along inner margin of crown as a noticeable ridge. The three main cusps are wide apart, and the space between the protocone and those of the outer row is less encroached upon by the bases of the cusps. It thus suggests the first trace of the flattened crushing surface. The man- * See Gidley, Proc. Washington Acad. Sci., VIII, p. 94, July 10, 1906. 38 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. dibular molars retain all of their cusps, but these are low and rounded. This is especially true of the paraconid, which is much the lowest of the cusps on inner side. The protoconid and metaconid, as well at the hypoconid and entoconid, stand nearly opposite each other, the anterior pair united by a rather high ridge, the remnant of the second commissure. Another stage is represented by the Sturnirinse (Plates III, IV, fig. 3). Here the styles and commissures have completely disappeared, leaving a parallel-sided, squarish, or oblong crown, the median por- tion of which is occupied by a broad groove nearly continuous from one tooth to the next. In the upper teeth the three primitive cusps are present and in practically their normal position, though all are much hollowed out from the inner side. The protocone is large, ex- tending along entire lingual side of tooth. Paracone smaller, but better defined than metacone. Third molar with the elements so reduced that the homologies of the cusps are doubtful. In the lower teeth the reduction of the paraconid, begun in the. Glassophaginse, is carried so far that this cusp is absent in m 2 and m 3 , while in m ± it is decidedly the smallest of those on lingual side. The other cusps stand opposite each other in pairs, the metaconid facing the proto- conid and the entoconid facing the hypoconid. They are of much the same character as those of the upper teeth, being placed at ex- treme edge of crown and with the inner side hollowed out to form the longitudinal furrow. Traces of the four principal cusps are visible in the posterior tooth. In the typical genus of Phyllonycterinse (Plates III, IV, fig. 4) a stage is reached in which the lower molars closely resemble those of the Pteropidse, while the upper teeth remain more as in the Sturnirinse. From those of the latter group the maxillary teeth differ, principally in the lowering of the protocone, so that the longitudinal groove becomes rather a broad, nearly flat, crushing surface. The paracone and metacone are also flattened, but their distinctness remains evi- dent. In the slight hollow between them may often be seen the last trace of the mesostyle (not shown in the figure). In the mandible the crowns of the molars are flat, with a slightly raised rim, on which is faintly indicated a trace of the protoconid, metaconid, and hypoconid. So faint are these traces that they might readily escape notice. In m j the anterior portion of the crown is somewhat nar- rowed, bearing a rather prominent cusp, probably the protoconid, on the anterior inner side of which there is a slight concavity. This hol- low may represent the space between protoconid and paraconid. A glance at fig. 4, Plate IV and Plate VIII, will show the striking simi- larity between these teeth and those of the Pteropidse. Before pass- ing to the latter, however, it may be well to describe a very important aberrant type of crushing dentition found in the Phyllostomidse. THE FAMILIES AND GENERA OF BATS. 39 This aberrant type (Plates V and VI) is found throughout the large subfamily Stenoderminse. In the maxillary teeth the paracone and metacone, with occasionally an intermediate cusp, perhaps the rem- nant of the mesostyle, form a cutting edge at outer margin, usually with a distinct cin'gulum in the normal position and an equally well- developed ridge on inner side extending from point or base of paracone to posterior base of metacone; the crowns are increased in width to form a large crushing area, the surface of which is usually roughened by folds or wrinkles, from which may be developed a definite protoconule and metaconule (Plate VI, fig. 3) ; -and a very large hypocone is often as a conspicuous postero-internal heel, which, in the most extreme instances, occupies the entire lingual side of the tooth and assumes the appearance of a protocone, causing the latter to occupy the relative position of a protoconule, though always recog- nizable by its large size (Plate VI, fig. 2). In the mandibular teeth a similar widening and flattening of the crowns has taken place ; the surface of the enamel is of the same character; the paraconid is absent, and the metaconid and entoconid tend to assume a subulate form, rising abruptly from the flattened surface of the crown (Plate V,fig. 2). The Stenodermine dentition most nearly resembles the Sturnirine type, and it seems probable that the two had a common origin, though the Stenodermine type has now become much, the more aberrant. Among the different genera there is considerable variation in the details of the tooth structure, though never enough to obscure the peculiar appearance characteristic of the type. The nearest approach to the Sturnirinae is seen in Vampyrops and Chiroderma, both of which lack all definite trace of the hypocone. 11 In the former the surface of the crowns is nearly smooth and the inner cingulum of the outer cusps is well developed, while in the latter the crown surface is coarsely wrinkled and the inner cingulum is absent. Either genus is readily distinguishable from Sturnira by the increased breadth of the crowns and the consequent replacing of the longitudinal groove by a wide crushing area. In Uroderma (Plate VI, fig. 1), the second upper molar shows a small shelf -like postero-internal projection, while in the first molar this has increased in size and assumed a distinct cusp-like form. About the same stage is represented by the first molar of Pygoderma, while an evident trace of the small hypocone can usually be detected in the greatly reduced second molar. A very rudimentary metaconule is present in m 1 and m z of Uroderma, and a similar but larger cusp is represented in the a The faintest possible suggestion of this cusp, or rather of a cingulum in the position that the cusp occupies in other genera, is present in some specimens of Vampyrops lineatus. 40 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. same teeth of Stenoderma as figured by Peters. Artibeus (Plate V, fig. 1), shows the hypocone of m x large and in characteristic position for the group — that is, mostly lingual to the protocone. In m 2 , how- ever, the condition is about the same as in Uroderma. The genera Ariteus, Phyllops, Ardops, and Sphceronycteris have the hypocone of m 2 well developed and of the same form as that in m 1 , though not as large. This is accompanied by an increased widening of the crowns. The maximum of widening is seen in the first molar of Oenturio (Plate VI, fig. 2), where the distance from hypocone to protocone is nearly equal to that from protocone to paracone, and the points of these three cusps are exactly in line with each other. ■•Between the bases of the paracone and metacone in the first and- second molar of Uroderma (Plate VI, fig. 1) may be detected a slight depression, the middle of which is occupied by a faintly marked groove. This is probably the first trace of the peculiar structure that forms so conspicuous a feature in the second molar of each jaw in E ctophylla. As already pointed out, a minute metaconule is present in Uroderma and Stenoderma, apparently developed as a specialization of the rugosity of the crushing surface. The same cusp is present in the first and second molars of Brachyphylla (Plate VI, fig. 3) as a low though evident longitudinal or oblique ridge, best developed in m 2 . In the first molar there is a protoconule of somewhat the same size and form, though smaller and more strictly longitudinal. This cusp becomes of considerable importance in the second molar, where its height and its basal area are both fully half as great as in the protocone, which the new cusp resembles in form. So strong is this resemblance that were it not for its position on the crown and for the reversal of the relative sizes the protoconule might be mistaken for the protocone and the protocone for the hypocone of a somewhat aberrant tooth of the usual Stenodermine type. In the third molar a similar cusp is present, though not as high or as well defined. The inner margin of the teeth is without trace of cingulum or shelf-like rudiment of the hypocone. In the mandicular teeth the intermediate cusps are rather well developed in m ,, though absent or barely indicated in m t and in m 3 . When present they assume the form of low, subterete elevations at base of metaconid and entoconid. The very peculiar arrangement of cusps in this otherwise primitive genus is probably a development from a type in which the teeth resembled those of Ghiroderma. The Pteropine dentition (Plates VII, VIII) though at first sight strikingly different from that of the typical Microchiroptera, has undoubtedly been developed from a primitive type similar to that a Monatsber. k. preuss. Akad. Wissensch., Berlin, 1876, plate facing p. 434. THE FAMILIES AND GENERA OF BATS. 41 retained by the otherwise more highly modified insectivorous bats through a series of stages analogous to those that may now be seen in the Phyllostomidse. Owing to the early and complete assumption of frugivorous habits these intermediate stages have disappeared among the Megachiroptera , and no fossil forms have yet been found in which they may be traced. While no genus of Microchiroptera, however strictly frugivorous its members may be, has teeth exactly resembling those of the Pteropidae, the change that would be neces- sary to pass between such dentitions as those of Pteropus and Phyl- lonycteris is much less than those which can be observed step by step from the latter back to the normal primitive condition of the cusps. In its simplest and most characteristic form, as seen, for instance, in Pteropus (Plates VII, VIII, fig. 1) this dentition is immediately recognizable by the bluntly rounded incisors, large and conspicuously ridged canines, and by the striking uniformity of the cheek teeth. The first and last cheek tooth both above and below is usually small or deciduous, its terete crown flat, concave, or slightly cusped. The others have oblong or squarish crowns bearing a large outer and a small inner elevation, these elevations high and cusp-like in the more anterior teeth, becoming successively lower toward the posterior end of the series, where, particularly in the lower jaw, they may be scarcely more than the rims of the conspicuous longitudinal median groove. The distinction between the inner and outer cusp is not always evident, particularly in the first large tooth of the lower jaw, and the space between them is usually somewhat filled in by the ridge- like inner bases of the cusps. Anteriorly where the cusps are high the crushing surface is strongly oblique, while posteriorly, where the cusps are low, it is nearly horizontal. While ■ no intermediate stages are known directly connecting this type of dentition with normally cusped ancestral forms, it is safe to assume, from analogy with the frugivorous Phyllostomine bats, that in the upper molars the two cusps are the protocone and paracone and in the lower molars the protoconid and metaconid. Teeth essentially like those of Pteropus are found in many genera of Megachiroptera. In Styloctenium and the long-tongued genera simplification has taken place, in the first instance by broadening and flattening the cusps and ridges until a nearly cushion-shaped crown results, in the second (Plates VII, VIII, fig. 4) by reduction in the size of both crowns and elevations. The more usual type of divergence is seen in the tendency to produce additional cusps and ridges not homologous with the parts of the primitive insectivorous tooth. This is well illustrated by Nyctymene (Plates VII, VIII, fig. 2) and Niadius (Plates VII, VIII, fig. 3). In the former the first, second, and third of the large mandibular cheek teeth develop 42 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. an extra cusp behind the protoconid, while the upper canine has a new cusp on the outer side of its tip. In Niadius the extra canine cusp is at the posterior base, while the additions to the cheek teeth take the form of small terete cusps near middle of crushing surface. The latter region is occupied by a longitudinal ridge in Dolsonia. In Pteralopex and Harpyionycteris the tendency toward supernu- merary cusps is carried to the greatest extreme, the penultimate cheek tooth of the latter, both above and below, containing no less than three well-developed, subequal cusps on each side, the resulting form closely resembling that of the molars of the Jurassic Bolodon. The last series of changes leads to the cutting teeth of the sanguiv- orous bats. The chain of intermediate stages is here less complete than in that leading to the fruit-crushing type, as the teeth of the Desmodontidae are so excessively modified a? to stand quite alone. The first step in this direction may apparently be seen, however, in the Hemiderminse and in the genus Erophylla. The teeth of Hemiderma (Plates IX and X, fig. 1) are not con- spicuously different from those of Glossophaga. The mesostyles have, however, completely disappeared, and the parastyles and metastyles have practically lost their identity in the large bases of the paracones and metacones. The two outer main cusps are well developed, and their edges, together with what remains of the commissures, are strongly trenchant. Protocone low, short, and rounded, widely sepa- rated from the outer cusps and forming entire inner section of crown. In the lower molars the original cusps may all be traced, but those of the outer row stand not far from middle of crown, where they form a nearly straight cutting edge, while those of the inner row are much reduced. The paraconid is present, however, in all three teeth, though in the first it is less developed than in the others, while the entoconid, though low and indistinct, is present as a postero-internal rim in all but the last. The process of reduction begun in this genus is carried much further in Rhinophylla. The protocone has here prac- tically disappeared, being represented by a mere inner rim to the outer cusps, which now stand close to lingual side of crown. The paracone and metacone are low and very narrow, producing a faintly two-lobed cutting edge ; paracone lower than metacone. On the outer side the crown is relatively wide, and the basal remnants of the parastyle and metastyle are evident. The corresponding changes in the lower molars have resulted in a form of crown scarcely dis- tinguishable fiom that of the premolars, except by their slightly greater length and more distinct main cusp (protoconid). On close inspection the last trace of the hypoconid may be detected, appear- ing in profile as a mere backward prolongation of the median cutting edge, but showing a distinctly indicated thickened base when viewed from above. THE FAMILIES AND GENEKA OF BATS. 43 From such teeth as those of Rhinophylla to the excessively modi- fied molars of the Desmodontida; (Plates IX and X, fig. 3) the tran- sition could readily be made. The cutting edge in these minute teeth therefore probably represents the combined paracone and metacone in the maxillary teeth and the protoconid and hypoconid in the lower jaw.- It is useless, however, to attempt to suggest exact homologies for the three cusps indicated in the posterior lower cheek tooth of Desmodus, as it is probable that these are mere secondary growths. Another suggestion as to the probable origin of the Desmodontine molars is furnished by the teeth of ErophyMa (Plates IX, X, fig. 2). These differ from those of Phyllonycteris (Plates III, IV, fig. 4) in their distinct narrowing and in the development of a well-defined cut- ting edge along the outer portion of the crowns, both above and below. Order CHIROPTERA. 1779. Chiroptera Blumenbach, Handbuch der Naturgeschichte, p. 74. Geographic distribution. — Eastern and western hemispheres to the northern and southern limits of tree growth. In the Pacific Ocean the range of the order extends to the Galapagos Islands and Hawaii from America, and to New Zealand, Samoa, the Caroline and Ladrone Islands from Asia. Characters. — Mammals with the front limbs modified for true flight, the fingers greatly elongated (the third usually at least equal to head and body) and joined together by a membrane which extends to sides of body and legs ; shoulder girdle much more developed than pelvis, the sternum usually keeled ; knee directed backward owing to rotation of leg for support of wing membrane. Number of forms. — There are at present recognized about 900 forms of Chiroptera, a number probably representing considerably less than half of what will eventually be known. Principal subdivisions. — The families of bats fall naturally into two main groups representing, as Winge has shown, two distinct stages in the specialization of the anterior limbs for flight. They may-be distinguished as follows: KEY TO THE SUBORDERS OF CHIROPTERA. Second finger retaining an evident degree of independence, its ungual pha- lanx present; humerus with trochiter and trochin small, the former never articulating with the scapula ; mandible with angular process broad and low or practically absent ; margin- of ear forming a com- plete ring MEGACHIROPTERA, p. 44. Second finger scarcely if at all independent from third, its ungual phalanx absent ; humerus with trochiter and trochin large, the former usually articulating with the scapula ; mandible with angular process well developed, long and narrow ; margin of ear not forming a complete ring MICROCHIROPTERA, p. 78. 44 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Remarks. — The distinction between these two groups is so sharp and definite that it is a matter of great convenience to recognize the Megachiroptera and Microchiroptera as suborders, though it may be. questioned whether they are of equal structural importance with other suborders of Mammalia. Suborder MEGACHIROPTERA. 1821. Frnctivorw Gray, London Medical Repository, XV, p. 299, April 1,-, 1821. 1872. Frugivora Gill, Arrangement of the Families of Mammals, p. 18, No- vember, 1872. 1875. Megachiroptera Dobson, Ann. and Mag. Nat. Hist., 4th ser., XVI, p. 340, November, 1875. 1878. Megachiroptera Dobson, Catal. Chiropt. Brit. Mus., p. 2. 1899. Megachiroptera Matschie, Flederm. des Berliner Mus. fur Naturk., p.l. Geographic distribution. — Tropical and subtropical regions of the Old World, east to Australia, Samoa, and the Caroline Islands. Characters. — In addition to the characters given in the key (p. 43) the following may be mentioned: Tragus never present; skull with rostral portion variable in length, but never specialized in form ; pre- maxillary well developed, usually free, always without palatal branch ; postorbital processes well developed; teeth very highly modified for frugivorism, the cheek teeth of upper and lower jaws closely resem- bling each other in form; molars normally with two blunt cusps on anterior portion of crown, these representing the protocone and para- cone in upper teeth, the protoconid and metaconid in lower; mandi- bular incisors never more than 2-2. Number of forms. — According to Matschie about 150 recognizable forms of Megachiroptera are now known. Principal subdivisions. — The Megachiroptera are all members of one family, the Pteropidse. Remarks. — Though the structure of the teeth presents a high de- gree of specialization, the development of the wings and the form of the skull represent an evolutionary stage much nearer to normal mam- mals than that which has been reached by the Microchiroptera. The index finger retains its ungual phalanx and much of its primitive in- dependence from the third digit; the humerus has not yet developed a high, flange-like deltoid crest for muscular attachment, nor has it acquired a secondary articulation with the scapula. Finally the whole general appearance of the skull is more nearly that of an ordi- nary mammal and less distinctively that of a bat. On the other hand, the molar teeth have nearly lost all distinct traces of their primitive structure. That this fact is of relatively little importance is shown, however, by the existence in a family of Microchiroptera, the Phyl- lostomidse, of a complete series of stages connecting the normal form THE FAMILIES AND GENEEA OF BATS. 45 of the teeeth with one nearly resembling that of the Megachiroptera. But, even if this were not true, in a group of characteristically volant animals the chief taxonomic importance must be assigned to the development of the wings. Family PTEROPID^E Gray. 1821. Pteropidw Gray, London Medical Repository, XV, p. 299, April 1, 1821. 1827. Pteropina Lesson, Man. de Mammalogie, p. 99. 1831. Pteropina Bonapabte, Saggio di una distrib. metodica degli Anim. Vert, p. 15. 1838. Pteropodidw Bonaparte, Nuovi Annali delle Scienze Naturali, Bo- logna, Anno I, Tomo II, p. 112. 1838. Pteropina Gray, Mag. Zool. and Bot, II, p. 502, December, 1838. 1854. [Pteropidw] " Pteropides " Gervais, Hist. Nat. des Mamm., p. 184. 1865. Pteropi Peters, Monatsber. k. preuss. Akad. Wissensch.. Berlin, p. 256. 1872. Pteropodidw Gill, Arrangement of the Families of Mammals, p. 18. 1875. Pteropidw Dobson, Ann. and Mag. Nat. Hist., 4th ser., XVI, p. 346, November, 1875. 1878. Pteropodidw Dobson, Catal. Chiropt. Brit. Mus., p. 3. 1892. Pteropodidw Winge, Jordfundne og nulevende Plagermus (Chiroptera) fra Lagoa Santa, Minas Geraes, Brasilien, p. 24. 1899. Pteropodidw Matschie, Flederm. des Berliner Mus. fur Naturk., p. 1. Geographic distribution. — Same as that of the suborder Mega- chiroptera. Characters. — As in the suborder. Principal, subdivisions. — The family Pteropidae is divisible into four groups worthy of recognition as subfamilies. KEY to the subfamilies of pteropid^e. . Premaxillaries broadly united anteriorly. Molars with blunt, indistinct cusps Nyctymenin^b, p. 75. Molars with at least 5 sharp cusps each Harpyionycterin^e, p. 77. Premaxillaries separate or barely united anteriorly. Tongue not highly extensible ; molars and incisors not specially reduced in size ; braincase never strongly deflected ; upper surface of mandib- ular symphysis forming conspicuous angle with alveolar line, Pteropin^e, p. 45. Tongue highly extensible ; molars and incisors considerably reduced in size ; braincase strongly deflected ; upper surface of mandibular sym- physis parallel with alveolar line Kicdotin^e, p. 68. Subfamily PTEROPINJE. 1878. Pteropodidw Dobson, Catal. Chiropt. Brit. Mus., p. 3 (part ; Pteropi part). 1891. Pteropodinw Flower and Lydekker, Mammals living and extinct, p. 650. 1899. Pteropodidw Matschie, Flederm. des Berliner Mus. fur Naturk^p. 1 (part) . 46 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Geographic distribution. — The same as that of the Megachiroptera (p. 44). Characters: — Premaxillaries separate, though usually in contact anteriorly, retaining their distinctness throughout life; bony palate, narrowing gradually behind tooth rows, the width of interpterygoid fossa, including hamulars, distinctly less than distance between pos- terior molars; canines parallel when jaws are closed; cheek teeth well developed, without unusual development of cusps ; tongue not specially elongated. Principal subdivisions. — The Pteropinse, the largest group of the Pteropidae, includes at least 19 genera. KEY TO THE GENERA OF PTEBOPIN.*:. Occiput so little deflected that alveolar line projected backward falls out- side of skull or at most passes through condyle. Upper cheek teeth more than 3-3. Upper incisors with widened, trenchant crowns ; calcar absent. Sphwrias, p. 53. Upper incisors styliform ; calcar present. Lower incisors 2-2; outer upper incisor nearly as long as inner. Dentition normal ; canines with secondary cusp ; crown of cheek-teeth without median ridge. Crowns of molars longer than broad, without evident median cusp Gynopterus, p. 47. Crowns of molars subquadrate, pm t and m - x with distinct terete cusp slightly in front of middle of crushing surface Niadius, p. 49. Dentition unusually heavy; canines without secondary cusp ; crown of pm t and m 1 with distinct median ridge. Thoopterus, p. 50. Lower, incisors 1-1 ; outer upper incisor conspicuously shorter than inner. Upper cheek teeth 5-r5 , Balionyctcris, p. 52. Upper cheek teeth 4-4. Rostrum nearly straight above ; tail present. v_ Ptenochirus, p. 51. Rostrum strongly concave above; tail absent. Megwrops, p. 51. Upper cheek teeth 3-^3. Cheek teeth all subterete Scotonyeteris, p. 64. Cheek teeth not all subterete. Dorsal and ventral profiles of skull strongly converging ante- riorly ; lower molars with one cusp on outer side. Epomophorus, p. 65. Dorsal and ventral profiles of skull nearly parallel; lower molars with two cusps on outer side Hypsignathus, p. 67. THE FAMILIES AND GENERA OF BATS. 47 Occiput so much deflected that alveolar line projected backward passes through or above root of zygoma. Posterior portion of occiput not distinctly elongated and tubular. Audital bulla with annular meatus ; first lower molar scarcely longer than second Rousettus, p. 54. Audital bulla with tubular meatus ; first lower molar nearly as long as second and third together Pterocyon, p. 55. Posterior portion of occiput distinctly elongated and tubular. 2-2 Incisors =-=. Canines with well-developed secondary cusps-Pteralopex, p. 60. Canines without secondary cusps. First and second lower molars with broad, flat postero- internal heel in addition to the two longitudinal ridges Acerdon, p. 59. First and second lower molars without postero-internal heel. Upper molariform teeth (pro 4 m 1 and m 2 ) subquad- rate in outline, essentially alike in form ; crown diameter of inner mandibular incisor about I that of outer . Desm,alopex, p. 60. Upper molariform teeth (pm 4 m 1 and m 2 ) not sub- quadrate, differing notably from each other in form ; crown diameter of inner mandibular incisor about i that of outer___' Pteropus, p. 56. Incisors less than -^ . Lower incisors 2-2 ; upper incisors wide apart Boneia, p. 61. Lower incisors 1-1 ; upper incisors close together. Upper incisors 1-1 ; cheek teeth normal Doosonia, p. 63. Upper incisors 2-2 ; cheek teeth with rounded crowns and broad low ridges Styloctenium, p. 62. Genus CYNOPTERUS F. Cuvier. 1825. Cynopterus F. Cuvier, Des Dents des Mammiferes, p. 248 (Pteropus marginatus Geoflvoy =Vespertilio sphinx Vahl). 1828. Pachysoma Is. Geoffkoy, Diet. Class. Hist. Nat., XIV, p. 703, Septem- ber, 1828. (Pteropus melanocephalus, P. titthwcheilus, Pachysoma diardii, P. duvaucelii, P. brevicaudatum) . Not Pachysoma MacLeay 1821. 1878. Cynopterus Dobson, Catal. Chiropt. Brit. Mus., p. 80 (part; mar- ginatus, scherzeri, brachysvma). 1899. Cynopterus Matschie, Flederm. des Berliner Mus. fiir Naturk., p. 74 (subgenus). „_ Type-species. — Pteropus marginatus Geoff roy =Vespertilio sphinx Vahl. Geographic distribution. — India and the Malay region, east to the Philippine Islands. Number of forms. — As here restricted, the genus Cynopterus con- tains about a dozen known forms,, most of them closely related to C. sphinx. 48 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Characters. — Dental formula : -.2 3. 1. -2345-- -2-2 1-1 1 2 -. 1. -2 3 4 5 6 2-2' 1-1 pm m 1-1 3-3 3_3' "' 2-2 =30. Teeth without special modifications. Upper incisors in contact with each other, but separated from canine by wide diastema, nearly terete, their crowns slightly flattened antero-posteriorly, the inner tooth of each pair slightly longer than the outer. Lower incisors somewhat flattened laterally, the outer of each pair slightly the larger; cutting edge blunt, nearly flat, though divided by a shallow notch into two in- conspicuous lobes. Canines both above and below with a secondary cusp on inner side formed by a slight prolonga- tion of the cingulum and some- what variable in its develop- ment in different species. An- terior upper premolar perma- nent, but scarcely larger than incisor in cross section and so short as barely to penetrate the gum. Its crown is nearly flat, though with a faintly in- dicated antero-external cusp. The corresponding lower tooth is considerably larger, its disk- shaped crown somewhat con- cave except for the minute antero - external cusp. The three suceeding cheek teeth almost exactly resemble each other above and below ; pm z and pm 3 are higher and shorter than the others, and the inner ridge tends to form a blunt cusp opposite base of main cusp ; pm i ,m 1 , pm±, and w^ have the longitudinal furrow well de- ' veloped and bounded on the inner side by a distinct, though low, ridge, and on the outer side by a much higher ridge, which rises anteriorly into a blunt cusp. Their crowns are noticeably longer than broad. The sec- ond lower molar closely resembles the anterior lower premolar, except that it lacks the small antero-internal cusp. The skull (figs. 5, 6) very primitive in form, the ventral profile nearly straight, the dorsal pro- file showing no special peculiarities. Floor of brain case scarcely Fig. 5. — Cynopterus sphinx. Adult female. Mad- ras, India. No. 102430. XU. THE FAMILIES AND GENERA OP BATS. 49 deflected, though placed at a slight angle with palate. Occipital region without trace of tubular elongation. Alveolar line projected backward barely touches audital bulla and occipital condyle. Ros- trum short (least distance from orbit to nares less than lachrymal width), its dorsal profile slightly concave. Premaxillaries in con- tact anteriorly, but not fused, slender throughout, and tapering above, to a recurved point distinctly below upper rim of nares. Palate nor- mal. Audital bullae well developed, but without tube or lip at meatus. Mandible with broad low angular process. Exter- nal form short and heavy. Nostrils very prominent, almost tubular. Upper lip divided by a deep nar- row vertical groove. Sec- ond finger with well-de- veloped claw. Wings from sides of back. Tail distinct, its terminal half free from the interfemo-. ral membrane. Calcar weak, its length about equal to breadth of foot. Species examined, — Gy- nopterus sphinx (Vahl), G. angulatus Miller, C. montanoi Robin, C. tii- thcecheilus Temminck, C. sp. (Celebes), G. scherzeri (Zelebor), G. melanocephalus (Temminck), C. major Miller, C. pagensis Miller, C. brachyotis (Miiller). Fig. 6.— Cynopterus Montanoi. pore, Malay Peninsula. Adult Female. Singa- No. 102432. xl}. Genus NIADIUS Miller. 1900. Xiadius Miller, Proc. Biol. Soc. Washington, XIX, p. 83, June 4, 1906. Type-species. — Cynopterus princeps Miller. Geographic distribution. — Nias Island, off west coast of Sumatra. Number of forms. — The type is the only species known. Characters. — Like Cynopterus,- but with the larger cheek teeth broader and more squarish in outline; crown of £>m 4 .and m 1 with 25733— No. 57—07 m 4 50 BULLETIN 5*7, UNITED STATES NATIONAL MUSEUM. distinct terete cusp slightly in front of middle of crushing surface (Plate VIII, fig. 3). Species examined. — Niadius princeps (Miller). Remarks. — In the somewhat increased size and the squarish out- line of the cheek teeth this genus shows a slight approach to Thoop- terus; but the terete cusp in pm t and m 1 differs conspicuously from the ridge which occupies the same position in the related group. Genus THOOPTERUS Matsehie. 1878. Gynopterus Dobson, Catal. Cliiropt. Brit. Mus., p. 80 (part, latidens= nigrescens) . 1899. Thoopterus Matschie, Flederm. des Berliner Mus. fur Naturk., p. 77 (part). Subgenus of Gynopterus to contain nigrescens, melanocephalus, and Wanfordi. Type-species. — Gynopterus nigrescens Gray (G. latidens Dobson). Geographic distribution. — Borneo, Celebes, Morotai, and Luzon. Number of forms. — Only two species of Thoopterus are known, T. nigrescens (Gray) and T. spadiceus Thomas. Characters. — Like Gynopterus in external characters (tail and cal- car present). Skull as in Gynopterus, except that interorbital region is narrower and zygomata stand out more widely. Teeth as in Gy- nopterus., but incisors smaller (though of the same proportions) ; upper canine without trace of secondary cusp, except a slight angular elevation of the cingulum postero- internally ; lower canine with a rather abrupt lobule on inner side, developed from cingulum and not forming a distinct cusp; and cheek teeth (except the small pm 2 and pm 2 ) greatly enlarged, their crowns subquadrate in outline. The crown of pm 4 and m t is crossed by a distinct ridge, this better developed in Thoopterus nigrescens than in T. spadiceus, in which it is somewhat obscured by the unusually large inner lobe. In the type of T. spadiceus the smaller upper premolar is absent and no trace of its alveolus can be detected. Species examined. — Thoopterus nigrescens (Gray), T. spadiceus (Thomas). Remarks. — This genus is at once recognizable by its general re- semblance to Gynopterus (though with the tail somewhat reduced), and the enlarged, subquadrate cheek teeth. As here used the limits of the group differ from those originally assigned to it by Matschie in the exclusion of the species melanocephalus and blanfordi; the first of which is a Gynopterus; while the second I have made the type of the genus Sphcerias. THE FAMILIES AND GENERA OP BATS. 51 Genus PTENOCHIRUS Peters. 1861. Ptenochirus Peters, Monatsber. k. preuss. Akad. Weissensch. Berlin, p. 707 (jagori, subgenus oiPachysoma = Cynopterus) . 1878. Ptenochirus Dobson, Catal. Chiropt. Brit. Mus., p. 81 (subgenus of Cynopterus) . 1899. Ptenochirus Matschie, Flederm. des Berliner Mus. fur Naturk., p. 77 (part, subgenus). Type-species. — Pachysoma {Ptenochirus) jagori Peters. Geographic distribution. — Philippines, Borneo. Number of forms. — Two species of Ptenochirus are now known. Characters.— External and cranial characters like Cynopterus (tail and calcar present) . Dental formula : -23.1.-2345-- .2-2, 1-1 3-3 1-1 OQ -2-1-88*6 -6^ * rf C ri' Pm 3^3' m 2^T ^ Upper incisors smaller and more pointed than those of Cynopterus, distinctly though narrowly separated from each' other, the outer tooth of each pair scarcely more than half as long as inner; lower incisor smaller than the corresponding tooth (i 2 ) in Cynopterus, nearly terete (not widened antero-posteriorly) . Canines as in Cynopterus, but without secondary cusps. Cheek teeth differing from those of Cynopterus in a general widening and flattening of the crowns, the outline of which becomes subquadrate in pm i , pm 4 , and m „ and the less distinctness of the median furrow encroached upon by the thickened inner ridge. Species examined. — Ptenochirus lucasi Dobson, P. jagori Peters. Remarks. — This genus differs from Cynopterus in the absence of the inner pair of lower incisors, the reduction in size of the outer upper incisors, and in the broader, natter molariform teeth. The last character suggests Thoopterus, but is less pronounced and the teeth are not increased in size. Genus MEG^EROPS Peters. 1841. Megera Temminck, Monogr. de Mamm., II, p. 274 (ecaudata) . 1841. Megara Temminck, Monogr. de Mamm., II, p. 359 (ecaudata) ; not of Wagler, 1830, or of Robineau-Devoidy, 1830. 1865. Megcerops Peters, Monatsber. k. preuss. Akad. Wissensch. Berlin, p. 256. (Substitute for Megara.) 1878. Megmrops Dobson, Cat. Chir. Brit. Mus., p. 87 (subgenus of Cynop- terus). 1899. Megmrops Matschie, Flederm. des Berliner Mus. fur Naturk., p. 79 (subgenus of Ptenochirus). Type-species. — Megcera ecaudata Temminck. - Geographic distribution. — Sumatra, Borneo. Number of forms. — Only the type-species is known. 52 BULLETIN 51, UNITED STATES NATIONAL MUSEUM. Characters. — Like Ptenochirus, but with no visible tail and with rostral portion of skull strongly concave both dorsally and laterally. The peculiar concave dorsal profile is shown in Matschie's figure." In this skull the rostrum is also concave laterally, or perhaps the form might be better described as deeply compressed in and below the mid-nasal region. Species examined. — Megasrops ecaudatus (Temminck) from Bor- neo. Remarks. — Though M'egcerops has been generally regarded as a subgenus of Cynopterus, and Matschie has placed it under Ptenochi- rus, I think that it may more naturally be allowed to stand as a dis- tinct genus. By the tooth formula and the character of the cheek teeth it is rather closely allied to Ptenochirus, but the absence of any external tail and the peculiar form of the rostrum sufficiently differ- entiate it. The type of " Gynonycteris " grandidieri Peters, which I have ex- amined in Paris, somewhat closely resembles Megcerops in the form of the rostrum (see Matschie's plate already referred to), but differs from this genus in the presence of 2-2 lower incisors. As Matschie suggests, 6 the supposed m 1 and m 2 are the two roots of the single molar tooth, the crown having been entirely worn away. In the lower jaw the anomalous " m /' and " m" are likewise the remnants of the normal second molar. The animal had, therefore, the dental formula of Cynopterus. In its bad condition, however, the structure of the teeth being quite obscured by their excessive wear, I prefer not to attempt to assign the species to any genus. Genus BALIONYCTERIS Matschie. 1899. Balionycteris Matschie, Flederm. des Berliner Mus. fur Naturk., p. 80. (maculatus.) Type-species. — Cynopterus maculatus Thomas. Geographic distribution. — Borneo. Number of forms. — One, the type species. Characters. — Like Ptenochirus, but without external tail and with 5-5 upper cheek teeth. Dental formula : - 2 3. 1. - 2 3 4 5 6 - -2-2 1-1 3-3 2-2 OA -2-1.-23456- ' * 1=1' ° 1=1' Pm 3=3' m 2=2 =3 °- Teeth closely resembling those of Ptenochirus except for the pres- ence of the minute second upper molar. Incisors styliform; canines "Die Flederm. des Berliner Mus. fur Naturk., pi. [vin], fig. 5. It is even more- pronounced in a Bornean specimen in the British Museum (No. 95.10.4.1, Kina Balu, John Whitehead). & Die Flederm. des. Berliner Mus. fur Naturk., p. 71. THE FAMILIES AND GENERA OF BATS. 53 without supplemental cusps; crowns of larger cheek teeth subquad- rate in outline, the median groove, as in Ptenochirus, less distinct than in Cynopterus. General form of skull as in Cynopterus and Ptenochirus. Species examined. — Balionycteris maculatus (Thomas) . Remarks.— The only known species of this well-characterized genus is further recognizable by its excessively small size (forearm • only 41 mm.) and by the conspicuous yellowish spotting of the wings. Genus SPH^ERIAS Miller. 1899. Thoopterus Matschie, Flederm. des. Berliner Mus. fiir Naturk., p. 77 (part). 1906. Sphwrias Miller, Proc. Biol. Soc. Washington, XIX, p. 83, June 4, 1906. Type-species. — Cynopterus blanfordi Thomas. Geographic distribution. — Burma. Number of forms. — The type-species. Characters. — Like Cynopterus, but without calcar and external tail; incisors with conspicuously developed, sharply pointed, cut- ting edges. Dental formula, as in Cynopterus: - 2 3. 1. - 2 3 4 5 - - . 2-2 1-1 irvm 3-3 1-1 _„ A 12"-. 1. - 2 3 4 5 6 - * 2=2' ' 1=1' P ™ 3=3' m 2^~ 6 °- Cheek teeth and canines as in Cynopterus, except that the canines show no trace of secondary cusps. Incisors better developed .than in any of the related genera, and of conspicuously different form. In Cynopterus the cutting edge of the lower incisors is flat, though divided by a shallow emargination into two faintly indicated equal lobes. In Sphmrias the outer lobe is practically absent, and the inner lobe is elongated and sharply pointed, so that the series of teeth, when viewed from in front, shows four conspicuous serrations instead of six low, ill-defined points, or four rounded or flattened surfaces. The upper incisors are even more peculiar, the crown being sharp edged and angular, well differentiated from the shaft, and having a large main cusp near the middle. Skull as in Cynopterus. Species examined. — Sphmrias blanfordi (Thomas). Remarks. — The type and only known species of this genus was placed by Matschie in Thoopterus, but is strikingly different from all of its allies, particularly in the .absence of the calcar, the very narrow interfemoral membrane, and the remarkably developed in- cisors. Though perhaps most nearly allied to Cynopterus, its rela- tionships to the genera of the group to which it belongs can not be regarded as very intimate. 54 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Genus ROUSETTUS Gray. 1821. Rousettus Gray, London Medical Repository, XV, p. 299, April 1, 1821. (wgyptiacus.) 1829. Cercopteropus Burnett, Quart. Journ. Sci. Lit. and Art, XXVII, p. 269, April-June, 1829. (wgyptiacus, amplexicaudatus.) 1843. Xantharpyia Gray, List Spec. Mamrn. Brit. Mus., p. 37. (amplexi- caudata. ) 1843. Eleutherura Gray, Voyage of the Sulphur. Mamm., Pt. 2, p. 29. (Pterop'us hottentottus Temminck=P. collaris Illiger.) 1852. Gynonycteris Peters, Reise nach Mossambique, Zool., I, Saugeth., p. 25. (collaris.) 1870. Senonycteris Gray, Catal. Monkeys, Lemurs, and Fruit-eating Bats, Brit. Mus., p. 115. (Pteropus seminudus Kelnavt= Rousettus am- plexicaudatus.) 1878. Gynonycteris Dobson, Catal. Chiropt. Brit. Mus., p. 70 (part). 1899. Myonycteris Matschie, Plederm. des Berliner Mus. fur Naturk., p. 63. ( torquatus. ) 1899. Xantharpyia Matschie, Plederm. des Berliner Mus. fur Naturk., p. 65. (brachycephala, brachyotis, wgyptiaca, collaris, amplexicaudata.) Type-species. — Pteropus wgyptiacus Geoffroy. Geographic distribution. — Africa; Southern Asia to the Hima- layas; Malayan region to the Philippines; New Guinea, Bismark Archipelago, Solomon Islands. Number of forms. — Eight species are currently recognized. Characters. — Dental formula : -2 3. 1. - 23 456- .2-2 1-1 3-3 2-2 12-. 1. -234567 % 2^2' ° 1=?*™ 3^3' m 3^3 =M - Teeth essentially as in Cynopterus, but cheek teeth with cusps, ridges, and grooves less strongly developed. Upper incisors low, widely spaced, the central pair not higher than the outer teeth, the crowns bluntly rounded but distinctly differentiated from shafts ; lower incisors with front surface marked by a groove which in some species is so deep as to produce distinct emargination of the cutting edge. Skull (fig. 7) differing from that of Cynopterus in the slightly greater deflection of the occipital region, the alveolar line projected backward, passing through middle of occiptal condyle, and in the longer rostrum (distance from orbit to nares greater than lachrymal breadth) . External characters essentially as in Cynopterus, but muz- zle less shortened. Second finger with well-developed claw; tail and calcar present. Species examined. — Rousettus wgyptiacus (Geoffroy), R.amplexi- caudatus (Geoffroy), R. philippinensis (Gray), R. collaris (Illiger), R. brachyotis (Dobson), and R. torquatus (Dobson). Remarks. — As here understood the genus Rousettus includes the forms placed by Matschie in the subgenera Rousettus and Myonycteris. THE FAMILIES AND GENERA OP BATS. Genus PTEROCYON Peters. 55 1861. Pterocyon Peters, Monatsber. k. preuss. Akad. Wissensch., Berlin, I, p. 423 (paleaceus=stramineus) . 1878. Cynonycteris Dobson, Catal. Chiropt. Brit. Mus., p. 70 (part). 1881. Leiponyx Jentink, Notes from the Leydeu Museum, III, p. 59. (Leiponyx buttikoferi.) 1899. Pterocyon Matschie, Flederm. des Berliner Mus. fur Naturk., p. 62. (Subgenus of Xantharpyia=Rousettus.) 1899. Leiponyx Matschie, Flederm. des Berliner Mus. ftir Naturk., p. 85. (genus.) Type-species. — Pteropus stramineus Geoffroy. Geographic distribution. — Arabia; Africa south of the Sahara; Madagascar. Number of forms. — Three species of Pterocyon are now known. Characters. — Dental formula 2 3. 1. 2345 6- .2 2 1 c 1. -234567 2-2' 1-1 pin 3-3 2-2 '"3^3' m 3^3 = 34. Teeth as in Rousettus ex- cept that the crowns of the lower incisors are not grooved on the anterior face, and the cutting edges are uniformly rounded." The skull (fig. 7) , while in general closely resembling that of Rousettus is distin- guished by the remarkable development of the audital bullae, the outer portion of which is distinctly differen- tiated from the inner as a prominent lip or short tube surrounding the meatus. ( Fig. 7. ) Nothing compar- able to this structure occurs in the related genera or in any of the bats that I have examined Rousettus. Fig. "I— A, Rousettus amplexicaudatus. adult female. Caves near Maulmain, Burma. No. 37930. x2. B, Pterycyonstramineus. Adult female. Robertsport, Liberia. No. 102461. x2. External chraacters as in a In the species of Rousettus now known the crown of the first lower molar is scarcely longer than that of the second, while in Pterocyon stramineus and P. dupreanus it is about as long as the second and third together. This char- acter has been made very prominent by Matschie, but I do not consider it of special taxonomic importance, however convenient it may be as a means of recognizing members of the two genera. 56 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Species examined. — Pterocyon buttikoferi ( Jentink) P dupreanus (Pollen) and P. Stramineus (Geoffroy). Remarks. — This genus is distinguishable from all other groups of Chiroptera by the structure of the audital bullae, and from its nearest ally, Rousettus, by the entire-edged lower incisors as well. The type specimen of Leiponyx buttikoferi, which I have examined in company with Doctor Jentink, appears to be unquestionably a Pterocyon in which the claw of the index finger is lacking. The individual is so old that the crowns of the molars are completely worn away, a fact that may have some bearing on the absence of the claws. While the form of the molars can not be determined, the incisors, canines, and anterior premolars agree perfectly with those of Pterocyon, as do also the general aspect of the skull, the peculiar structure of the audital bullae, and, with the one exception already noted, the details of external form. Genus PTEROPUS Brisson. 1762. Pteropus Beisson, Regn. Anim. in Classes IX distrib., 2 ed., p. 13 ( vampyrus ) . 1799. Spectrum Lacepede, Tabl. des Div. Sousdiv.' Orders et Genres, des Mammif., p. 15 {vampyrus). 1866. Eunycteris Gray, Proc. Zool. Soc. London, p. 64 (phaiops). 1870. Pselaphon Gbay, Catal. Monkeys, Lemurs and Fruit-eating Bats Brit. Mus., p. 110 (ursirnis=pselaphon). 1878. Pteropus Doeson, Catal. Chiropt. Brit. Mus., p. 15 (part, included Acerodon and Styloctenium) . 1899. Pteropus Matschie, Flederrn. des Berliner Mus. fiir Naturk., p: 5 (part, included Pteralopex and Acerodon). 1899. Sericonycteris Matschie, Flederm. des Berliner Mus. fiir Naturk;, p. 30 (subgenus of Pteropus containing the species rubricollis, temmincki, jpetersi,personatus, capistratus, molossinus, ivoodfordi). 1899. Eunycteris Matschie, Flederm. des Berliner Mus. fiir Xaturk., p. 11 (subgenus of Pteropus containing the species melanopogon, papua- nus, degener, neohioernicus) . 1899. Spectrum, Matschie, Flederm. des Berliner Mus. fiir Xaturk., p. 19 (sub- genus of Pteropus containing the species poliocephalus, leucopterus, pselaphon, vetulus, vampyrus, anetianus, rayueri, samoensis, dasy- mallus, loochooensis, fonnosus, rodricenxix, lotiihocensis, brunneus. scapulatus, Hvingstoni, macrotis, epularls, assamensis, hypomela- nus, mariannus. Type-species. — V espertilio vampyrus Linnseus. Geographic distribution. — From the Comoro Islands, Madagascar, and India, to Tasmania, Samoa, the Ladrones, and southern Japan (Kiusiu). a Specimens with one or both claws very imperfectly developed, or perhaps reduced by disease or injury, have been observed in Pteropus capistratus (No. 122124, U.S.N.M-.), P. hypomelanvs (No. 105440), P. modiglianii (No. 140992), P. poliocephalus (No. 13594), and Acerodon sp. (No. 105454). THE FAMILIES AND GENEEA OF BATS. 57 Number of forms. — About 70 forms of Pteropus are now recog- nized. Characters. — Dental formula : -23. 1. -2345 6- .2-2 1-1 3-3 2-2 „. 12-1.-234567 2-2' 1-1'^ 3-3' 3-3" Incisors (Plates VII, VIII, fig. 1) , terete with slightly differentiated crowns, the lower bluntly rounded, the upper with a fairly defined cutting edge. The upper incisors are in contact or nearly in contact with each other, but are separated from canine by wide diastema. They form a slightly convex row, and the outer tooth of each pair is distinctly shorter than the inner. Lower incisors separated from each other and from canines by minute spaces, the row slightly convex and the inner tooth of each pair smaller than the outer. Canines sim- ple, without secondary cusps, the cingulum rather prominently developed on posterior and inner sides, and the shaft, especially of the upper teeth, marked by conspicuous longitudinal furrows. Anterior premolar (pm 2 and pm 2 ) small, that in upper jaw much smaller than incisors and deciduous, its crown bluntly pointed, that in lower jaw slightly larger than outer incisor, per- manent, its crown rounded in outline, the surface concave, the outer edge usually somewhat higher than inner. Posterior molars (m * and m 3 ) essentially like anterior lower premolar in size and form, but m 2 usually with a more prominent inner edge and often with a distinct outer anterior cusp or tubercle. The remaining cheek teeth (pm 3 , pm 4 , m % and pm 3 , pm 4 , m 15 m 2 ) rather closely resem- ble each other and are all formed on the same general plan of a sub- quadrate or somewhat elongated crown with distinct median furrow in long axis of jaw and a well developed outer and inner ridge, each ridge tending to rise into a cusp a little in front of middle. The outer ridge and cusp is. always larger than the inner, and in the region between the cusps the longitudinal groove is usually con- stricted or partly obliterated. This type of structure is most per- fectly shown by m * and m t , the ridges and cusps tending to become obsolete in the more posterior teeth, while they successively increase at the expense of the median furrow in the two large premolars, so that pm * and pm 3 assume almost the aspect of short canines, each with a well developed, or at least distinctly indicated, secondary cusp on inner side, the longitudinal groove having been made almost vertical by the elongation of the cusps. Skull (fig. 8) moderately elongate, the length of rostrum considerably greater than lachrymal breadth. Premaxillaries well developed, in contact anteriorly but not fused, their breadth along nasal suture at least equal to that at alveolus of outer incisor. Occipital region strongly deflected and distinctly tubular, the floor of the braincase forming so great an angle 58 BULLETIN 51, UNITED STATES NATIONAL MUSEUM. with palate that alveolor line projected backward passes through tympanic region or base of zygoma. Audital bulla reduced to a mere ring. Externally characterized by absence of tail, narrow mter- femoral membrane, well developed calcar, and large claw on index finger. Species examined. — Pteropus admiralitatum, Thomas, P. aldaoren- sis True, P. anetianus Gray, P. Brunneus Dobson, P. cagayanus Mearns, P capistratus Peters, P chrysoproc- tus Temminck, P. con- spicillatus Gould, P- coronatus Thomas, P. dasymdllus Temminck, , P. edwardsi Geoffroy, P. faunulus Miller, P. fuscus Dobson, P- ge- minorum Miller, P. gi- ganteus ( Briinnich ) , P. gouldi Peters, P grandis Thomas, P. hypomela nus Tem- minck, P. keraundren Quoy and Gaimard, P. lanensis Mearns, P. lanigera H. Allen, P. lepidus Miller, P. leu- copterus Temminck, P. livingstoni Gray, P. lombocensis Dobson, P. loochooensis Gray, P. me I an o p o g on Schlegel, P. modigli- anii Thomas, P. molos- Fig. 8. — Ptekopus lepidus. Adult female. Saddle Island, Sinus South China Sea. No. 101670, type. XI. Temminck. P. natalis Thomas, P- nicooaricus Zelebor, P. personatus Temminck, P. poliocephalus Tem- minck, P pselaphon Say, P. rayneri Gray, P. rubricollis Geoffroy, P. samoensis Peale, P. scapulatus Peters, P seychellensis Milne Edwards, P. temminckii Peters, P. vampyrus Linnaeus, P. woodfordi Thomas. Remarks. — Members of the genus Pteropus may be recognized by the large number and simple structure of the teeth, combined with the absence of the tail, and the presence of a well developed claw on the index finger. The species are usually large, amono- them some of the largest bats. THE FAMILIES AND GENERA OF BATS. 59 In extent of distribution and in the number of its species Pteropus exceeds all other genera of Megachiroptera. It also has the most complete and least modified dentition. These two circumstances have caused the gerilis to be generally regarded as the most primitive of the group; but it seems probable that the deflection and tubularity of the brain case, the great reduction of the audital bullae, and the complete absence of the tail are of more importance than the large number of teeth, and that Pteropus is in reality less primitive than Cynopterus, Rousettus, and Pterocyon. Genus ACERODON Jourdan. 1837. Acerodon Joubdan, Ann. Sci. Nat., Paris, 2" ser., VIII, Zool., p. 369 (jubatus). 1878. Pteropus Dobson, Catal. Chiropt. Brit. Mus., p. 15 (part). 1899. Acerodon Matschie, Plederm. des Berliner Mus. fiir Naturk., p. 9 (sub- genus of Pteropus. Type-species. — Pteropus -jubatus Eschscholtz. Geographic distribution. — Eastern part of the Malayan region (Philippines, Djilolo, Batjan, Celebes, Flores, Timor, Sumba). Number of forms. — Seven or eight forms of Acerodon have been described. Characters. — Like Pteropus, but with smaller canines and larger, more complex molariform teeth. Dental formula as in Pteropus. Upper incisors as in Pteropus, but somewhat more elongate and less bluntly pointed. Lower incisors differing from those of Pteropus in the much greater contrast in size between the inner and outer tooth of each pair. Canines much shortened as compared with Pteropus, the mandibular canine little exceeding the height of pm 3 . Though reduced in length the canines retain their thickness, and the cingulum is even better developed than in the related genus. Small premolars as in Pteropus. Upper molariform teeth differing from those of Pteropus in relatively greater size, greater breadth of crowns, greater distinctness of cusps, and a peculiar trenchant character of the edges and ridges. First large lower premolar {pm 3 ) essentially as in Pteropus. The three succeeding teeth differ from those of all other Pteropidas in the presence of a broad, flat postero-internal heel, sharply differentiated from the two ridges and extending considerably beyond them, especially behind. Cranial and external characters as in Pteropus. Species examined. — Acerodon jubatus (Eschscholtz) and several undetermined forms. Remarks. — Although not generally recognized as distinct from Pteropus this genus is well characterized by the form of the lower cheek teeth. Additional characters are furnished by the reduced size of the canines and of the inner lower incisors. 60 BULLETIN 5*7, UNITED STATES NATIONAL MUSEUM. DESMALOPEX, new genus. Type-species. — Pteropus leucopterus Temminck (as understood by Dobf;on, Catal. Chiropt. Brit. Mus., p. 32. 1878). Geographical distribution. — Philippine Islands. Number of forms. — The type is the only species known. Characters. — Like Pteropus; but skull with broader rostrum and palate, orbits directed slightly more upward, and teeth showing sev- eral peculiarities that suggest Pteralopex. Upper incisors subequal, distinctly larger than in Pteropus, the cross section of the crown nearly one-third that of canine, the cingulum produced into a notice- able shelf posteriorly. Lower incisors very unequal, the crown area of the outer nearly one-half that of canine, that of the inner scarcely more than one-half as great. Canines not peculiar, the upper with- out secondary cusps. Small upper premolar well developed, not deciduous, its diameter nearly half that of upper incisor, its crown flat. Small lower premolar relatively larger than in Pteropus, but smaller than in Pteralopex, its outer edge raised but not distinctly cuspidate. The remaining premolars, both above and below, agree with those of Pteropus, except that pm 3 shows no trace of cusp on inner side. Molars, both above and below, subquadrate in outline, the length of the crown never conspicuously greater than the width (to 3 and to * not elongated as in Pteropus) . In the maxillary teeth the cusps are like those of Pteropus except that they are higher rel- atively to length of crown. Lower molars peculiar in that the ridges of to x and m 2 are each divided into two low but distinct rounded cusps. The quadritubercular form resulting from this is very noticeable in to a , less so in to 2 . Third lower molar as in Pteropus. Species examined. — Desmalopex leucopterus (Temminck). Remarks. — This genus is intermediate between Pteropus and Pte- ralopex, though nearer the former. In the broadened rostrum and slightly upturned • orbits the skull distinctly suggests Pteralopex, while the same tendency is shown by the form and relative size of the incisors, the well-developed small upper premolar, the squarish out- line of the molars, and the extra cusps of to l and m 2 . Externally there are no characters worthy of note. Genus PTERALOPEX Thomas. 1888. Pteralopex Thomas, Ann. and Mag. Nat. Hist., Oth ser., I, p. 155, February, 1888. (atrata.) 1899. Pteralopex Matschie, Flederm. des Berliner Mus. fur Naturk., p. 11. (Subgenus of Pteropus.) Type-species. — Pteralopex atrata Thomas. Geographic distribution. — Solomon Islands. Number of forms. — Only the type species has been thus far dis- covered. THE FAMILIES AND GENEKA OP BATS. 61 Characters. — In general like Pteropus, but skull with broader, nearly parallel-sided rostrum, more upwardly directed orbits, and high sagittal crest, somewhat obscuring tubularity of occipital re- gion, and both canines and cheek teeth conspicuously several-cusped. Except as noted, the skull does not differ particularly from that of Pteropus. The posterior portion of brain case is distinctly tubular and so deflected that the alveolar line, projected backward, passes just below root of zygoma. Premaxillaries rather shorter and broader than in Pteropus, and in contact for about 2 mm. anteriorly. Audital bullae as in Pteropus. Dental formula as in Pteropus, but pm 2 larger and apparently persistent. " Upper incisors with broad posterior ledges. Upper canines short vertically, enormously thick antero-posteriorly, each with one stout secondary cusp halfway up its posterior edge, and two smaller postero-internal basal cusps. Pre- molars and molars short and broad, their anterior and posterior basal ledges so developed and their main cusps so conical as to destroy all the appearance of longitudinal grooving characteristic of the genus Pteropus. Lower incisors extremely disproportionate in size, the outer not less than about twenty times the bulk of the inner. Canines very short vertically, with a simple posterior basal ledge. Cheek teeth markedly cuspidate, the general longitudinal grooving quite obliterated. Posterior premolar and first molar each with three high anterior cusps, and a low posterior basal ledge." a Species examined. — Pteralopex atrata Thomas. Remarks. — The dentition of Pteralopex represents one of the ex- treme phases of that tendency to develop secondary cusps, which appears to be everywhere latent in the Pteropidse. The stages through which it has passed are indicated in the genera Acerodon and Des- malopex. The much-reduced inner mandibular incisor, the broad- ened rostrum, and the complete, strongly upturned orbits are also highly specialized characters. Genus BONEIA Jentink. 1879. Boneia Jentink, Notes from the Leyden Museum, I, p. 117. (Hdens.) 1899. Boneia Matsohie, Flederm. des Berliner_Mus. fur Naturk., p. 69. Type-species. — Boneia bidens Jentink. Geographic distribution. — Celebes. Number of forms. — Two species are now known. Characters. — Like Pteropus, but with a distinct tail; only 1-1 up- per incisors ; cheek teeth with crowns much flattened. Dental formula : - 2 -. 1. - 2 3 4 5 6 - . 1-1 1-1 3-3 2-2 on TT 12-. 1.-984 5-67 * 2=2 °' 1=1' ^ 3=3' m 2=2 = 32< U PP er lncl " a Thomas, Proc. Zool. Soc. London, 1888, p. 473. 62 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. sor very small, the inner extremity of the premaxillary projecting beside it and appearing at first sight like a second tooth. Lower inci- sors as in Pteropus, but disproportion in size between outer and inner more evident. Canines not peculiar, except that the lower are more strongly curved backward and outward than in the related genera. Small upper premolar (pm 2 ) apparently permanent, con- siderably larger than in Pteropus. Except for the greater size of pm 2 , the two anterior cheek teeth in each jaw are essentially as in Pteropus. The three posterior above (pm 4 , m 1 , and m 2 ) and the four posterior below pm 4 , m 15 m „ and m 3 ) are remarkable for their very flat crowns, with very low, though sharp, peripheral ridges. The crown of m 1 and of m 1 is crossed by a distinct transverse ridge be- tween the two very rudimentary cusps. General outline of crowns as in Pteropus. Skull differing from that of Pteropus in the slightly greater deflection of the brain case (alveolar line projected backward passes slightly above root of zygoma) and in the low, weak posterior section of mandible, the coronoid process forming an unusually slight angle with the ramus. External characters as in Pteropus, except for the presence of a well-developed tail projecting distinctly beyond inter femoral membrane. Index finger with claw. Species examined. — Boneia bidens Jentink and B. menadensis Thomas. Genus STYLOCTENIUM Matschie. 1878. Pteropus Dobsok, Catal. Cbiropt Brit. Mus., p. 15 (part). 1899. St uloctenium Matschie, Flederm. des Berliner Mus. fiir Naturk., p. 33. (wallacei.) Type-species. — Pteropus wallacei Gray. Geographic distribution. — Celebes. Number of forms. — One, the type species." Characters. — Like Pteropus, but with only 1-1 lower incisors and 2-2 lower molars ; cheek teeth subterete in outline and with smoothly rounded crowns. Dental formula: -2 3. 1. -234 56- .2-2 1-1 3-3 2-2 - 2 -. 1. - 2 3 i 5 6 - 1 1=1' c l=l^ m 3=3' m 2=2 =3 °- Incisors both above and below terete, very straight, the crowns abruptly pointed and scarcely differentiated from shafts. Canines as in Pteropus, but rather short and thick. Small upper premolar (pm 2 ) deciduous, similar in size and form to that of Pteropus. Other anterior cheek teeth not peculiar, though rather less pointed than in Pteropus. Posterior three teeth in each jaw (pm *,m 1 ,m 2 , and pm 4 , to 1; m 2 ) , with crowns nearly circular in outline and 'with evenly rounded surface, the cusps and ridges, though normal in posi- tion, so flattened and broadened as to be scarcely noticeable. Skull THE FAMILIES AND GENEBA OP BATS. 63 as in Pteropus. Posterior portion of brain case strongly tubular, and occipital region so deflected that alveolar line projected backward passes through root of zygoma. Audita! bullae as in Pteropus. Ex- ternal characters as in Pteropus. Species examined. — Stylo cienium wallacei (Gray). Remarks. — This genus is well differentiated from its allies by its peculiar dental formula and by the structure of the incisors and pos- terior cheek teeth. Genus DOBSONIA Palmer. 1810. Cephalotes Geoffboy, Ann. Mus. d'Hist. Nat. Paris, XV, p. 104 (part). 1828. Hypoderma Is. Geoffkoy, Diet. Class. d'Hist. Nat., XIV, p. 706 (not of Latreille, 1825). 1840. Hypodermis Blyth, Cuvier's Animal Kingdom, p. 69. s 1878. Cephalotes Dobson, Catal. Chiropt. Brit. Mus., p. 91. 1898. Dobsonia Palmer, Proc. Biol. Soc. Washington, XII, p. 114, April 30, 1898. 1899. Cephalotes Matschie, Plederm. des Berliner Mus. fur Naturk., p. 85. 1902. Dobsonia Thomas, Proc. Biol. Soc. Washington, XV, p. 198, October 10, 1902. Type-species. — Cephalotes peronii Geoffroy. Geographic distribution. — Celebes, Moluccas, New Guinea, Bis- mark Archipelago, Solomon Islands. Number of forms. — Three species are now recognized. — — Characters. — Differs from Pteropus in absence of claw of index finger, presence of well-developed tail, attachment , of membranes along middle of back, and in the presence of only 28 teeth. Dental formula : -2-.1. --3456- .1-1 1-1 2-2 2 - 2_ oa -2 -.1. -234567 b T^T ° l^V pm 3^3* m 3^3 *' Upper incisors short, but well developed, in contact or nearly so, the crowns about as wide as long and with distinct cutting edge. Lower incisors almost structureless spicules with barely indicated blunt crowns. Canines rather short, not peculiar in form, without second- ary cusps, those of the lower jaw nearly in contact. No small upper premolar (pm 2 ). The corresponding tooth in the lower jaw is rela- tively larger than in Pteropus, and is not separated from canine or from first large premolar by any appreciable space. The other pre- molars both above and below differ from those of Pteropus in greater development of the cusps and in more trenchant ridges. Small posterior molars (m 2 and m 3 ) as in Pteropus. Anterior molars (m x and m iy m 2 ) resembling the large premolars in their general charac- teristics as compared with the corresponding teeth in Pteropus, the crown of each with a distinct median longitudinal ridge at middle of furrow. This ridge is continuous nearly from end to end of m 2 , 64 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. in m 1 and m t it is confined to the region behind the cusps, and in some specimens is obsolete in the first lower molar. Skull much shorter and more heavily built than that of Pteropus, but of essen- tially the same type. Distance from orbit to nares about equal to lachrymal width. Posterior portion of brain case slightly tubular, and occipital region so deflected that alveolar line projected back- ward, passes just below root of zygoma. Premaxillaries very narrow, the width scarcely more than that of crown of incisor. The inner extremities barely come in contact. Audital bullae somewhat wider than in Pteropus, but less developed than in Cynopterus. External characters unlike those of Pteropus in the absence of the claw of index finger, the presence of a distinct tail, and the attachment of the wing membranes along middle of back. Species examined. — -Dobsonia palliata (Geoffroy), D. minor (Dob- son), and D. magna (Thomas). Remarks. — The genus Dobsonia is one of the most aberrant of those associated with Pteropus by the form of the skull. Its chief modifi- cation is in the direction of a general shortening of the facial part of the skull, with crowding of the tooth rows and a tendency to eliminate the less important anterior teeth. Genus SCOTONYCTERIS Matsehie. 1894. Scotonycteris Matschie, Sitz.-Ber. Gesellsch. uaturforsch. Freunde. Berlin, No. 8, p. 200. 1899. Scotonycteris Matschie, Flederm. des Berliner Mus. fur Naturk., p. 70. Type-species. — Scotonycteris senkeri Matschie. Geographic distribution. — West Africa (Kamerun and Fernando Po). Number of forms. — Two species are now known, the type and S. bedfordi Thomas. Characters. — Dental formula as in Epomophorus and Hypsigna- thus. Incisors and canines essentially as in Epomophorus, the former styliform and without noticeable contrast in size, the latter relatively larger than in the allied genera, though of the same form, the anterior surface smoothly rounded. Cheek teeth all subterete, essentially alike in form, the small lower premolar (pm 2 ) tapering to a point, the two lower molars flattish ; other cheek teeth (pm 3 , pm *, m l , pm pm 4 ) with obliquely sloping crowns and no distinct flat crushing sur face (their form closely resembling that of pm 3 in Epomophorus comptus). Skull not flattened as in Epomophorus, but resembling that of Cynopterus except that rostrum is compressed in region of diastema and the premaxillaries are spatulate instead of tapering above. Mandible remarkably slender and weak, its form almost exactly as in Epomophorus. Audital bullae as in Epomophorus. Externally much as in Epomophorus, except that the lips are less 35 THE FAMILIES AND GENERA OF BATS. 65 developed. There is a rudimentary papilla-like tail, the calcar is present, as is the claw of the index finger, the wings are attached low on the sides, as in Epomophorus, and the fur is slightly woolly. Species examined. — Scotonycteris bedfprdi Thomas. Remarks. — This genus has been placed near Cynopterus by Mats- chie, but it appears to be more closely related to Epomophorus. Though the skull more nearly resembles that of the former, this char- acter seems to indicate merely a less specialized stage than has been reached by Epomophorus. Similarly primitive are the slightly de- veloped lips. The number of teeth, the form of the mandible, and the broadly hairy back all resemble Epomophorus, while the struc- ture of the teeth could easily be derived from that of the related genus by a process nearly the opposite to that which has taken place in Hypsignathus. Genus EPOMOPHORUS Bennett. 1836. Epomophorus Bennett, Proc. Zool. Soc, London (1835), p. 149,. Feb- ruary 12, 1836. (macrocephalus.) 1866. Epomops Gray, Proc. Zool. Soc, London,-p. 65 (franqueti). 1878. Epomophorus Dobson, Catal. Chiropt. Brit. Mus., p. 4 (part). 1899. Epomophorus Matschie, Flederm. des Berliner Mus. filr Naturk., p. 34 (part). 1899. Epomops Matschie, Flederm. des Berliner Mus. fur Naturk., p. 56 (subgenus for E. oompttis, E. franqueti, and E. dobsonii). 1899. Micropteropas Matschie, Flederm-. des Berliner Mus. fur Naturk.. p. 57 (subgenus for E. pusillus). 1899. Nanonycteris Matschie, Flederm. des Berliner Mus. fur Naturk., p. 58 (subgenus for E. velcUcampii) . Type-species. — Pteropus epomophorus Bennett=P. macrocephalus Ogilby. Geographic distribution. — Africa, south of the Sahara. Number of forms. — About two dozen forms of Epomophorus are now known. Next to Pteropus this is the largest genus of Mega- chiroptera. Characters. — General appearance much as in Pteropus, but lips fuller; tail present, though usually not visible externally; occipital portion of skull (fig. 9) slightly deflected and not tubular, and upper cheek teeth only 3-3. Dental formula : -2 3. 1. -"-3 4 5-- . 2-2 1-1 2-2 1-1 12-. 1. -23456-*2-2' c l- l'^ m 3 - 3' m 2 - 2~ ■ ' Upper incisors small, stylif orm ; the crowns pointed, but not much differentiated from shafts; the outer tooth of each pair shorter than inner, but not strongly contrasted in size. The incisors are sepa- rated from each other by narrow spaces and from the canine by a wide diastema. Lower incisors equal in size, forming a nearly straight row between the canines, separated from each other and 25733— No. 57—07 m 5 66 BULLETIN 51, UNITED STATES NATIONAL MUSEUM. from canines by narrow spaces, their crowns bifid or notched, rather distinctly marked off from shafts, sometimes by an evident con- striction. Canines small and weak, not peculiar in form, without secondary cusps or distinct ridges, the anterior surface without trace of longitudinal furrows. Except that the mandibular canines are smaller, they are almost identical in appearance with those of the upper jaw. No small maxillary premolar or molar (pm 2 , m 2 ). The remaining maxil- lary teeth (pm s , pm 4 , and m 1 ) are essentially as in Pteropus, except that they are rela- tively smaller, and the cusps and ridges, though sharp and distinct, are not as large. Anterior premolar separated from ca- nine by a notice- able diastema, and from next premo- lar by spaces rela- tively wider than in Pteropus. Small lower pre- molar • (pm 2 ) of about the same relative size as in Pteropus (dis- tinctly larger than incisors) , but with an outer cusp and obliquely flattened surface. Other mandibular teeth differing from those of Pteropus about as in the case of the corresponding maxillary teeth, the small molar (m 2 ) closely resembling m % of Pteropus. Skull (fig. 9) broad and flattened, the depth of brain case usually little more than half its width. Deflection of occipital region slight, the alveolar line con- aVery noticeable in an immature E. franqneti from Mount Coffee, Liberia (Cat. No. 83798, U. S. N. M.). Fig. 9.— Epomofhobus franqtjeti. Liberia. No. xli. THE FAMILIES AND GENERA OF BATS. 67 tinued backward usually passing through audital bulla and occipital condyle. Audital bullae about as in Gynopterus, much better de- veloped than in Pteropus. Mandible slender and weak, its depth between pm a and pm± scarcely greater than length of pm s ; coronoid low and very gradually sloping. In its external characters the genus differs from Pteropus chiefly in the very large full lips and the pres- ence of a distinct, though very rudimentary tail consisting of two vertebrae and readily detected by touch. Males usually with con- spicuous glandular mass and tuft of modified hairs on shoulder. Species examined. — Epomophorus comptus H. Allen, E. crypturus Peters, E. dobsonii Bocage, E. franqueti Tomes, E. gambianus Jen- tink, E. labiatus (Temminck), E. macrocephalus (Ogilby), E. minor Dobson, E. neumanni Matschie, E. pusillus Peters, E. wahlbergi Sun- devall; also several undetermined forms. Genus HYPSIGNATHUS H. Allen. 1861. Hypsignatlius H. Allen, Proc. Acad. Nat. Sci. Philadelphia, p. 156. (monstrosus.) 1862. Sphyrocephalus Murray, Proc. Zool. Soc. London, 1862, p. 8. (labro- sus=monstrosus. ) 1862. Zygmnocephalus Murray, Proc. Zool. Soc. London, pi. i (misprint). 1878. Hypsignathus Dobson, Catal. Chiropt. Brit. Mus., p. 6 (subgenus). 1899. Hypsignathus Matschie, Flederm. des Berliner Mus. fur Naturk., p. 42 (subgenus). Type-species. — Hypsignathus monstrosus H. Allen. Geographic distribution. — Western and central Africa from Gam- bia to the Uelle region and French Kongo. Number of forms. — The type species only. Characters. — Like Epomophorus, but with lips much more devel- oped, the upper lip thrown into conspicuous folds anteriorly and about nostrils; beneath upper lip the mouth communicates with a paired sac extending from extremity of muzzle to halfway between eyes and ears. No shoulder glands. Tail reduced to a tubercle quite hidden beneath the skin and barely perceptible to the touch. Skull like that of Epomophorus, but with dorsal and ventral profiles nearly parallel, the depth of rostrum in males greater at diastema than in lachrymal region. Teeth as in Epomophorus, but incisors more widely spaced, lower incisors very obscurely or not bilobed, canines even more reduced in size, small lower premolar barely pier- cing gum, pm *, to 1 , pm 4 , m i , and m 2 , with inner ridge much more developed than in Epomophorus and median furrow correspondingly deepened, outer ridge of lower molars divided into two distinct blunt cusps. Species examined. — Hypsignathus monstrosus H. Allen. Remarks. — The peculiarities of Hypsignathus monstrosus', espe- cially those of the teeth, are too great to allow the animal to be placed in the genus Epomophorus, 68 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Subfamily KIODOTINiE. 1875. Maoroglossi Dobson, Ann. and Mag. Nat. Hist., 4th ser., XVI, p. 346, November, 1875 (group of I'teropidm) . 1878. Macroglossi Dobson, Catal. Chiropt Brit. Mus., p. 4 (group of Ptero- 1891. Garponyoteriinai Flower and Lydekkeb, Mammals living and extinct, p. 654. 1898. Kiodotinm Palmer, Proc. Biol. Soc. Washington, XII, p. Ill, April 30, 1898. 1899. Pteropodidce Matschie, Flederm. Berliner Mus. Naturk., p. 1 (part). Geographic distribution. — Malay region from Darjiling to the Philippines ; also New Guinea and northern Australia, and the neigh- boring islands east to the Fiji and Caroline groups. A single genus occurs in West Africa. Characters. — Premaxillaries at first separate, but uniting and losing their distinctness later in life; bony palate narrowing gradually behind toothrows ; mandibular symphysis elongated, its upper surface parallel with alveolar line; tongue highly extensible; teeth (except canines) much reduced in size. Principal subdivisions. — Nine genera of Kiodotinae are at present known. KEY TO THE GENERA OF KIODOTINiE. Occipital portion of skull slightly deflected, the alveolar line when projected backward passing through root of zygoma Eonycteris, p. 69. Occipital portion of skull much deflected, the alveolar line when projected backward passing considerably above root of zygoma. Lower cheek teeth 5-5 ; tail present. 2-2 Incisors j^j; upper cheek teeth 4-4; tail extending to heel. Notopterw, p. 74. Incisors g^; upper cheek teeth 5-5; tail not extending to knee. Cattinycteris, p. 69. Lower cheek teeth 6-6 ; tail absent. 2-2 Incisors j^f Nesonycteris, p. 74. Incisors g^s- Incisors well developed, the lower forming a continuous line between canines Syconycteris, p. 72. Incisors minute, the lower not forming a continuous line between canines. First three maxillary cheek teeth approximately alike in size and form. Upper cheek teeth 6-6 Odontonycteris, p. 71. Upper cheek teeth 5-5 Kiodotus, p. 70. First three maxilliary cheek teeth differing conspicuously among themselves in size and form. Length of mandibular symphysis nearly twice distance between outer surface of canines; lower incisors not bifid Melonycteris, p. 73. Length of mandibular symphysis scarcely greater than distance between outer surface of canines; lower incisors conspicuously bifid Trygcnycteris, p. 73. THE FAMILIES AND GENERA OE BATS. 69 Genus EONYCTERIS Dobson. 1873. Eonycteris Dobson, Journ. Asiat. Soc. Bengal, XLII, Pt. 2, p. 204. 1878. Eonycteris Dobson, Catal. Chiropt. Brit. Mm, p. 94. 1899. Eonycteris Matschie, Flederm. des Berliner Mus. fur Naturk., p. 89. Type-species. — Macroglossus spelceus Dobson. Geographic distribution. — Malay Peninsula, Sumatra, Java, and Borneo. Number of forms. — The type is the only species yet known. Characters. — Externally like Rousettus, except that muzzle is more elongate and the index finger has no claw. Tail and calcar well developed. Tongue with the usual characters of the subfamily, sharply pointed, and highly extensible. Skull not strikingly differ- ent from that of Rousettus, but more slender, and occipital region more deflected, the alveolar line when continued backward passing through or slightly above root of zygoma. Mandible not as heavy as in Rousettus and coronoid not rising as abruptly; otherwise differ- ing only in the character of the symphysis peculiar to the sub- family. Dental formula : -2 3. 1. -23456- - 2-2 1-1 3-3 2-2 12-. 1. -234567*2-2' C l-V pm 3-3' m 3 - 3~ d4 ' Upper incisors of equal size, simple, shorter than in Rousettus, but not distinctly different in form or position. Lower incisors ap- parently similar to upper (none seen with unworn crowns) , directed distinctly forward, separated from each other and from canines by subequal spaces about as wide as diameter of incisor at alveolus. Canines slender and not. very large, without secondary cusps and with barely indicated cingulum, the front surface of the upper canine marked by a deep longitudinal groove. Cheek teeth essentially as in Rousettus, but lateral ridges and median grooves less distinct. Species examined. — Eonycteris spelcea (Dobson). Remarks. — While this genus by the structure of the tongue and an- terior portion of the mandible is evidently a member of the subfamily Kiodotinse, it is the least characteristic of the group, lacking the noticeable deflection of the occipital region that occurs in all the others and showing in the form of the teeth and rostrum a stage of development intermediate between the more typical genera and Rousettus, which may be regarded as occupying a position near the point of departure of the long-tongued group. Genus CALLINYCTERIS Jentink. 1889. Gallinycieris Jentink, Notes from the Leyden Museum, XI, p. 209, November, 1889. 1899. Callinycteris Matschie, Flederm. des Berliner Mus. fur Naturk., p. 90. Type-species. — Callinycteris rosenbergii Jentink. Geographic distribution. — Celebes. 70 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. Number of forms. — The genus is represented by the type species only. Characters. — Like Eonycteris, but with wing attached to second toe instead of to first, and with lower molars 2-2 instead of 3-3. The teeth are throughout more robust than those of Eonycteris, and the crowns of the last upper premolar and first upper molar are distinctly broader in proportion to their length. Species examined. — Gallinycteris rosenbergii Jentink. Remarks. — Though closely related to Eonycteris this genus ap- pears to be well characterized by its heavier dentition and by the absence of the small m 3 . Doctor Jentink has kindly sent me a photograph of the skull of the type, which shows that the specimen is a young adult with the teeth unworn. Genus KIODOTUS Blyth. 1824. Maoroglossus Schinz, Naturgesch. und Abbild. Saugeth., p. 71 (not Macroglossum Scopoli 1777). 1827. Macroglossa Lesson, Man. de Mammalogie, p. 115. 1840. Eiodotus Blyth, Cuvier's Animal Kingdom, p. 69. 1848. Rhynchocyon Gistel, Naturgesch. Thierreiclis fiir hohere Schulen, p. ix (not of Peters, 1847). 1878. Macroglossus Dobson, Catal. Chiropt. Brit. Mus., p. 95. 1891. Carponycteris Lydekkee, in Flower and Lydekker, Mammals Living and Extinct, p. 654. 1899. Macroglossus Matschie, Flederm. des Berliner Mus. fiir Naturk., p. 95 (part). 1898. Kiodotus Palmer, Proc. Biol. Soc. Washington, XII, p. Ill, April 30, 1898. Type-species. — Pteropus minimus Geoffroy. Geographic distribution. — Malay region from Darjiling to the Phil- ippines, New Guinea, and the Solomon Islands. Number of forms. — Three or four species are currently recognized. Characters. — Dental formula : -23. 1.-23456- .2-2 1-1 3-3 2-2 n , 12-.1.-234567 % 2^2' ° \=V pm 3^3' m S^ 54 " Upper incisors subequal, minute, simple, distinctly projecting for- ward, the diameter of the crowns much less than that of posterior molars, the teeth of each pair separated by a space about equal to diameter of tooth at alveolus, the pairs separated from each other by a slightly wider space and from canines by diastemata double as wide. Lower incisors similar to upper in form and position, but outer tooth appreciably larger than inner and median space much wider than that between outer incisor and canine, the latter interval about equal to that between the incisors of each pair. Canines strong THE f amilieS and genera 6f bats. 71 and well developed, without secondary cusps or prominent cingula, the maxillary canines with deep longitudinal groove on anterior sur- face. Cheek teeth small, without distinct contrasts of size or form, the two anterior in each jaw {pm 2 , pm 3 , pm 2 , pm 8 ) with the crowns compressed and elevated into distinct though blunt cusp, the others {pm 4 , m 1 , m 2 , pm 4 , m u m 2 , m 8 ) with narrow nearly flat orowns bounded by indistinct ridges, of which that on the outer side is the less developed. Anterior premolar, both above and below, close to canine, but separated from succeeding premolar by space about double as great as width of its crown. In both jaws the anterior premolar is smaller than the next. Skull (fig. 10) with a weak rostrum and large brain case, the occipital region not tubular but so strongly deflected that the alveolar line projected backward passes between root of zygoma and middle of braincase. Postorbital processes short and little curved. Development of audital bul- lae about as in Gynopterus. Mandible very weak, and ramus distinctly con- cave beneath. Symphysis long, its greatest diameter equal to about twice the distance between outer sides of ca- nines. Externally characterized by presence of well-developed claw on in- dex finger, attachment of wing mem- brane to base of fourth toe, and rather low on sides of body. No external tail. Species examined. — Kiodotus minimus (Geoffroy) and K. lago- chilus (Matschie). Genus ODONTONYCTERIS Jentink. Pio. 10.— Kiodotus lagochilus. Adult female. Philippine Islands. No. 123440. x H. 1902. Odontonycteris Jentink, No. 3 (July, 1901), p. 140. Notes from the Leyden Museum, XXIII, July 15, 1902. Type-species. — Odontonycteris meyeri Jentink. Geographic distribution. — Known only ~from two small islands south of the Philippines, Great Sangir and Cagayan Sulu. Number of forms. — Only the type-species is known. Characters. — Like Kiodotus, but upper molars 3-3 and papillae of tongue less developed. The dental formula ,pm -2 3. 1. -234567 ,-2-2 1 12-. 1. -2 3 456 7 2-2' 1-1' m - — -=36 is unique among Pteropine bats. 3-3' "" 3-3 In actual form, however, the teeth 72 BULLETIN 57, UNITED STATES NATIONAL MUSEUM. do not differ very noticeably from those of Kiodotus. Incisors and canines quite as in that genus. Cheek teeth both above and below uniformly smaller than in Kiodotus (compared with a Philippine specimen of the lagocMlus type) , a character that is especially notice- able in the upper molars, the combined crown length of which (3 teeth) is scarcely more than in the related genus (2 teeth). Third upper molar well developed, nearly as large as m 3 , its crown sub- circular in outline. External characters as in Kiodotus. Tongue with the papiike smaller than in Kiodotus (compared with Philip- pine lagochilus only), those at side and tip (except extreme anterior margin) so widely spaced that the surface of the tongue is plainly visible between them (in Kiodotus they overlap so that the surface of the tongue is quite hidden) . Species examined. — Odontonycteris meyeri Jentink or a closely related species from Cagayan Sulu (north of Sandakan, Borneo). Remarks. — While the characters of the single specimen of 'Odon- tonycteris that I have examined (male, ad. Cat. No. 125316, U.S. KM., collected by Dr. Edgar A. Mearns, February 25, 1904) , might suggest an abnormal Kiodotus, they are too important to permit the name to be placed in synonymy. This individual has in the left mandible a small supplemental tooth behind the third molar. Genus SYCONYCTERIS Matschie. 1899. Syconycteris Matschie, Flederm. des Berliner Mus. fiir Naturk., p. 98. (Subgenus of Macroglossus=Kiodotus.) 1902. Syconycteris Jentink, Notes from the Leyden Museum, XXIII, No. 3 (July, 1901), p. 131, July 15, 1902. (Genus.) Type-species. — Macroglossus minimus var. australis Peters. Geographic distribution. — New Guinea, northen Australia, and the neighboring islands (Bismarck Archipelago, Aru, Louisiade Islands) . Number of forms. — Four species are recognized. Characters. — Like Kiodotus, but with much larger incisors, the height of those in upper jaw greater than width of premaxillaries, those in the lower jaw forming a continuous line between canines, and all, except the middle pair below (which are less than half as large as outer) , approximately equal in diameter to posterior molars (m 3 and m 3 ). Occiput not as much deflected as in Kiodotus, the alveolar line when projected backward passing distinctly below mid- dle of braincase. Species examined. — Syconycteris crassa (Thomas). Remarks. — As Doctor Jentink has pointed out, this group, origi- nally defined as a subgenus of Kiodotus to contain the species austra- lis, papuanus, finschi, and crassus, seems quite worthy of reco