'j^A. THE FAMILIES AND GENERA OF LIVING RODENTS BRITISH MUSEUM (NATURAL HISTORY) THE FAMILIES AND GENERA OF LIVING RODENTS ^(^^'^ BY a.OJ^'^ ^^^ |. R. ELLERMAN ' ^v' WITH A LIST OF N.\]\IED FORMS (1758 -1936) BY R. W. HAYMAN and G. W. C. HOLT VOLUME L RODENTS OTHER THAN MURIDAE LONDON: PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM Issued 8th yune, 1940] [Prke One Pound Fifteen Shillings Sold at The British Museum (Nait'r.\l History), Cromw-ell Road. S.W.y, and by B. QuARiTCH, Ltd.; Dulau & Co., Ltd.; and the Oxford University Press made and printed in great BRITAIN BY JARROLD AND SONS LTD. NORWICH THIS WORK IS DEDICATED TO THE MEMORY OF MY FRIEND WILLIAM COX PREFACE As the result of several years devoted to a careful study of the rich recent collections belonging to this museum, the author has prepared the following review of the families and genera of living Rodents; he has also supervised the preparation by Mr. R. W. Ilayman and Mr. G. W. C. Holt, members of the Museum staff, of a list of all the species and subspecies of rodents, described from the loth edition of Linnaeus's Systema Natura, I'j^d, down to the end of 1936. Such a review and such a list have long been two of the most needed desiderata of zoologists. The author has endeavoured in each case to find out not only what characters have been assigned to a given genus, but what characters it in fact possesses, and to test their value and constancy. Genera are recognized by their intrinsic characters ; a mere geographical differentiation of genera does not exist. When- ever the author has been able to study a genus for himself he has included it in his " Keys" ; when not seen personally he merely gives a note of the ascribed characters. The carrying through of this great examination has led to a con- siderable reduction in the number of genera here recognized ; thus, of 440 forms or groups given, at some time or another, full generic rank, only 343 (151 non-murine, 192 murine) are now regarded as valid genera. Indeed, had the Museum collections of North American Rodents been more extensive, and had it been possible to make a really detailed survey of the South American Muridae, a further much-needed reduction of genera would doubtless have taken place. With regard to major classification, the author excludes the Duplicidentata from the Order Rodentia. He reviews the more recent classifications of the Order (so restricted), to wit, those of Oldfield Thomas, 1896; TuUberg, 1899; Weber, 1904 and 1928; Miller & Gidley, 1918; and Winge, 1934. He then proposes a new classification of the families and genera, which, while necessarily sharing some features with one or other of the older systems, on the whole appears to be a great improvement upon any of the systems previously proposed. Of great interest is the chapter on distribution and the conclusion reached by the author that, as regards its peculiar Muridae, Australasia may be claimed as an evolutionary' centre. That view, in face of the characters of the group in question and the very high antiquity of the Order, appears to me to be perfectly sound. To conclude these general remarks the author is to be congratulated on having performed a colossal task, and we are indebted to him for providing an honest one-man view which cannot fail to be of service to all who wish to study this great and complex Order in future. With regard to the list of named forms, every endeavour has been made to make it complete and accurate, but it is too much to hope that nothing has escaped the compilers. viii PREFACE The beautiful drawings of skulls and teeth prepared by Mr. A. J. Engel Terzi greatlv enhance the value of this work; with the exception of Figs. 62, 63, 90. 96, 112, 115, iiS, 119, 146, 164, 167, 170, 173 and 184, bv the same artist, hut originally published in Miller's Catalogue of the Mammah of Western Europe, IQ12, all the figures have been specially drawn tor this volume. Owing to the heavy work undertaken by the artist and to the prevalent war conditions, considerable delay has occurred in the publication of Vol. I, which deals with the general matters discussed above and with the families and genera of non-murine rodents. Vol. II, dealing with the murine rodents, is already in type; in order to avoid further delay it is proposed to issue it immediately without waiting for the preparation of its full complement of figures. \'ol. Ill, to be published later, will be an "Atlas" containing all the figures of Vol. I repeated; the full complement of figures for Vol. II; to these it is hoped to add drawings of some of the more interesting and important external characters and of dissections ot jaw muscles. MARTIN A. C. HINTON, Keeper of the Department of Zoology. British Museum (Natural History) 25//; March, 1940 AUTHOR'S FOREWORD In 1896 Oldfield Thomas proposed a classification of Simplicidentate Rodents in which he recognized 156 genera. In 1904, Trouessart in his Catalogue of Mammals listed 205 genera of this section of mammals. At the present day, more than four hundred and forty forms have been given, at some time or another, full generic rank, so that in the last thirt\--five years there has been an increase of approximatelv two hundred and forty genera. The object of this work is primarily to inquire into the status of these named genera, and to give, in each family and subfamily, a key which will indicate as reasonably, clearly and briefly as possible, the differences between such genera as are supported on characters which appear constant through the various groups, and worthy of generic recognition. This has led on the one hand to a careful studv of the classification of the families and superfamilies or major groups of all authors who have attempted an arrangement of the Order based on adequate material and including all the principal leading genera of all families as recognized to-day (Thomas, 1896; Tullberg, 1899; Weber, 1904, 1928; Miller & Gidley, 1918; and Winge, 1 924), and on the other hand to the collection of a list of all named forms (species and subspecies) which have been named in the order since Linnaeus (i7:;8), up to and including the year 1936. It is not my intention to enter into a detailed description of the skeleton, soft parts, etc., of each genus; this has been done alreadv in a far more com- petent manner than I could attempt (Tullberg, 1899). Ninety-nine out of a hundred genera are based on cranial, dental, and external characters, or to put it more crudely, "skull and skin" characters. This work is based almost entirely on these characters. It must be noticed that in cranial, dental and external characters, ver\' many specimens of a species or genus may usually be at hand for examination, so that whether such a character is constant or not can generally be checked easily ; generic names based on skeleton, soft parts, or characters such as the baculum, which has been used for generic names (here not retained) in the Sciuridae, can as a rule only be examined for one species and often one specimen of a genus ; therefore it is not possible to give full notes on such a character throughout a whole genus or, if it is so, the notes refer to a restricted number of specimens only. Under these circumstances, it seems wiser not to pay too much attention to names which have been based solely on such characters. This work is based entirely on the collection of the British Museum, to the authorities of which I am much in debt for their kindness and consideration throughout the compilation of this work. No genus which is not represented in that collection is included in my keys; as it is difficult to include in a key any form which has not been examined, and in the case of certain Muridae, X AUTHOR'S FOREWORD impossible. However, there are only five non-Alurine genera at present un- represented in London, and about sixteen (out of nearly two hundred) Muridae; notes on these will be included, but generally speaking no comments. Of the genera which are included, I have endeavoured to give in each case the approxi- mate range, the number of forms at present recognized (to 193(1), a list of these forms, and a short description of the main cranial, dental and external characters. Concerning the dental characters, in this Order, great difference of opinion exists regarding the homologies of the various parts of each tooth. American authors use in the main a series of names for each cusp which are figured and explained by Goldman, North Anicr. Fauna, 43, p. 11, 191S; Miller & Gidley, iqiS, divide the greater part of the Order into "tritubercular" and "quadri- tubercular" series; Winge apparently uses quite a different theory which he takes throughout Mammalia; and Hinton uses still a different notation. With the exception of the Murinae, in which a series of eight or sometimes nine main cusps go through the entire subfamily, no attempt has been made in this work to use a definite formula; I have endeavoured to describe the dentition of each genus as I see it, and am content to leave the working out of cusp homolo- gies to those with more experience than I have. The view that the dental pattern of modern Rodentia is generally speaking derived from a much more complex pattern than is now present, expressed by Hinton, Monograph of Voles and Lemmings, 1926, Evolution of Molars, pp. 102-124, is here accepted. However, it is not the purpose of this paper to enter into an argument as to whether this view is correct, or the view frequently held that a complex dentition in a living Rodent is a secondarily acquired one. Take for instance two cases, Hapalomys, a very complex-toothed Rat (Murinae) as compared with Rattiis (normally a simple-toothed Rat); in this work, Hapalomys is considered the primitive type, Rattiis the specialized one; but turn these views round, and the two genera will still be at the opposite end of two extremes, which is broadly speaking what I set out to prove in each case. So far as the list of named forms of each genus is concerned, I list those which are named, making no attempt to guarantee the validity of any subspecies or species. How many of these names will ultimately be reduced to synonymy is not clear; but I believe that very many of them will prove invalid with more material available. I have attempted in each case in which a genus has not been revised, where possible, to divide the genera into "specific groups" as is now done by American authors. These groups indicate certain characters within the different species of each genus, but must be regarded as provisional. The attitude, however, is held that a list of names in some semblance of order, no matter how provisional, is better than a string of meaningless alphabetical names. At least I hope it gives a start to those who are interested in the characters w^hich run through the species of the various genera. I expect how- ever that a large number of South .\merican Mice (Cricetinae) will have for the time being to be abandoned, and listed alphabetically. It is perhaps not too much to hope that these lists will act as a deterrent to authors who rush to give names to new forms before consulting all the literature on the genus in question. Though great care has been taken I can give no guarantee that the list of AUTHOR'S FOREWORD xi named forms is complete, particularly in the case of some of the older synonyms, to which less attention has been paid than to names described more recently. The list was originally based on that of Trouessart (Cat. Mamm. Viv. et Foss, 1904); names which appear in synonymy in this work have in most cases been included here as they appear in Trouessart's list, and their position has not been verified. I have listed si.xty-four hundred forms which are supposed to be valid at the present day. Subspecies, except in cases of a genus which is definitely revised, are listed as far as possible geographically. Each named form is listed under its present accepted name, or the name which appears to be correct; in many cases not under the generic name under which it was described, for instance " Sminthus longer," Nathusius, 1840, is now listed as " Sicista subtilis loriger," etc., etc. The Order is absolutely dominated by one family, the Muridae, both in number of genera and named forms, as proved by the following figures, which must be taken as approximate. Twenty-two families of Rodents other than Muridae: 151 valid genera containing 2,773 named forms. Family Muridae : 192 valid genera containing 3,600 named forms. I have had therefore to divide the work into two volumes, the first of which contains all Rodents not belonging to the family Muridae, the second devoted entirely to that family. This work is based solely on Rodents which are living, or assumed to be living, though I have added short notes on the fossil history of each family, chiefly from a distributional point of view. My sincere thanks are due to the officials of the British Museum for their kind help and consideration throughout the time I have been preparing this work; especially I must mention Captain Guy Dollman, who originally made it possible for me to undertake this review; Mr. M. A. C. Hinton, who has under- taken the task of editing the work; Mr. Hayman and Mr. Holt, who have between them got together the references and type localities of more than sixty-four hundred names in the Order, and the former for much assistance in dealing with some of the species of the more unwieldy genera; Dr. Tchernavin, who has translated several papers from Russian, enabling me to give some details concerning the distribution of the various groups of Rodents occurring in the U.S.S.R.; Mr. A. J. Engel Terzi, who has made drawings of the more important genera; Mr. J. L. Chaworth-Musters, to whom I am indebted for nearly all my knowledge of Palaearctic Rodents; also I must thank Mr. E. R. Newman, who has given me much help throughout the compilation of the work, and Miss R. Blizard and Mr. F. C. Hitch, who have assisted during the later stages. Lastly I would thank my wife for numerous working drawings of specimens and continuous help in other ways. The indulgence of readers is sought for any typographical or other small errors in this work. The writing of the book itself was finished in June last, but the revision of the final prools had not been completed when war broke out. Since then it has been possible for me to give onlv the most cursorv and intermittent attention to such revision. b CONTENTS Preface Author's Foreword Limits of the Order Rodextia Variation in Rodents . Previous Classifications of the Order Oldfield Thomas, 1896 Tullberg, 1899 Weber, 1904, 1928 Miller & Gidley, 1918 Winge, 1924 .... Outline of Cl.\ssification here adopted Zygomasseteric Structure Distribution ..... Lists of Genera and Principal Species dealt Palaearctic .... Nearctic . . . . . Indo-Malayan .... African ..... Neotropical .... Order RODENTIA Key to Superfamilies Superfamily BATHYERGOIDAE Family Bathyergidae Genus Bathyergus, Illiger Genus Heliophubius, Peters Genus Georychus, Illiger Genus Cryptomys, Gray Genus Heterocephalus, Rijppell Superfamily IIYSTRICOID.AE . Family Echimvidae . Subfamily Echimyin.\e . Genus Echimys, Cuvier Genus Isothrix, Wagner Genus Diplomys, Thomas IN this V' olume PAGE vii i.\ I 3 5 5 8 II 13 18 22 42 47 57 57 60 64 66 69 77 77 79 79 83 84 86 86 94 96 lOI 106 108 "3 115 CONTENTS Genus Piorchinivs. Allen Genus Hoploinvs, Allen Genus Cercoinys, Cu\ier Genus Euryzygomatornys, Goeldi Genus Clyomvs, Thomas Genus Carterodon, Waterhouse Genus Mesomvs, Wagner Genus Loncliothrix, Thomas Subfamily Capromyinae Genus Capromvs, Desmarest Genus Geocapromys, Chapman Genus Procapromxs, Chapman Subfamily Plagiodontinae . Genus Plagiodontia, Cuvier Subfamily Dactylomyinae Genus Thrinacodus, Giinther Genus Dactvloinys, Geoffroy Genus Kaimahateomvs, Jentink Subfamily Myocastorinae Genus Mvocasior, Kerr Subfamily Thryonomyinae . Genus Tluyonomys, Fitzinger Subfamily Petromyinae Genus Petromus, Smith Subfamily Abrocominae Genus Ahrocoma, Waterhouse Subfamily Octodontinae Genus Octoiiiys, Thomas Genus Aconaemys, Ameghino Genus Octodon, Bennett Genus Octodoiitoinys, Palmer Genus Spalacopiis, Wagler Genus Ctenomys, Blainville Family Dinomyidae . Genus Dinomys, Peters Family Frethizontidae Subfamily Chaetomyinae Genus Cliaetomys, Gray PAGE "5 122 123 124 125 126 127 128 128 132 133 133 134 13s 136 137 139 140 144 144 149 149 152 156' 160 161 170 171 173 174 175 CONTENTS Subfamily Erethizontinae Genus Erethizon, Cuvier Genus Echinoprocia, Gray Genus Coendou, Lacepede Family Dasyproctidae Genus Dasyprocta, Illiger Genus Myoprocta, Thomas Family Hystricidae . Genus Trichys, Giinther Genus Atherurus, Cuvier Genus Thecurus, Lyon Genus Hystrix, Linnaeus Family Cuniculidae Genus Ctinicultis, Brisson Family Chinchillidae Genus Chinchilla, Bennett Genus Lagidium, Meyen Genus Lagostomus, Brooks Superfamily CAVIOIDAE Family Caviidae Subfamily Caviinae Genus Cavia, Pallas Genus Galea, Meyen Genus Caviella, Osgood Genus Kerodon, Cuvier Genus Dolichotis, Desmarest Subfamily Hydrochoerinae . Genus Hxdrochoerus, Brisson Superfamily APLODONTOIDAE Family Aplodontiidae Genus Aplodontia, Richardson Superfamily SCIUROIDAE Family Sciuridae Outline of previous classification of the family Classification here adopted . Genus Belomys, Thomas Genus Trogopterus, Heude . Genus Pteromyscus, Thomas XV PAGE •77 178 181 182 189 190 196 197 203 205 209 212 220 221 226 227 229 233 237 237 238 240 242 243 246 247 249 250 253 253 255 259 259 2fil 269 276 279 280 CONTENTS Genus Petaurista, Link Genus Aeromys, Robinson & Kloss Genus Pteromys, Cuvier Genus Glaucomys, Thomas . Genus Eoglaucomys, Howell Genus Hylopetes, Thomas . Genus Petinomxs, Thomas . Genus Petaurillus, Thomas . Genus lomvs, Thomas Genus Eupetaunis, Thomas Genus Mvosciuriis, Thomas Genus Nannosciurus, Trouessart Genus Sciurillus, Thomas Genus Microsciuriis, Allen . Genus Svntheosciuriis, Bangs Genus Sciurus, Linnaeus Genus Tamiasciunis, Trouessart Genus CaUosciurus. Gray- Genus Fuuamhulus, Lesson . Genus Dremomxs, Heude Genus Ratufa, Gray . Genus Menetes, Thomas Genus Lariscus, Thomas & Wroughton Genus Glyphotes, Thomas . Genus Wieithrosciurtis, Gray Genus Rhinosciurus, Gray Genus Hvosciuiiis, Tate & Archbold Genus Heliosciurus, Trouessart Genus Paraxerus, Forsyth Major Genus Funisciurus, Trouessart Genus Protoxenis, Forsyth Major Genus Mvrsilus, Thomas Genus Epixenis, Thomas Genus Xenis, llemprich &: Khrenberg Genus Atlantoxeriis, Forsyth Major Genus SpeiinuphHopsis, Blasius Genus Sciurotamias, Miller Genus Tamicn, Illiger PAGE 281 290 291 294 297 29S 300 302 3°3 304 312 313 317 319 321 321 345 348 376 3S0 383 390 391 393 393 395 39S 399 405 410 415 416 417 41S 422 423 425 426 CONTENTS Genus Citellus, Oken . Genus Marmota, Blumcnbach Genus Cynomys, Rafinesque Superfamily CASTOROIDAE Family Castoridae Genus Castor, Linnaeus Superfamily GEOMYOIDAE Family Heteromyidae Subfamily Heteromyinae Genus Heteromys, Desmarest Genus Liotnys, Merriam Subfamily Dipodomyinae Genus Perognathus, Wied Genus Microdipodops, Merriam Genus Dipodomys, Gray Family Geomyidae . Genus Thomomys, Wied Genus Geomys, Rafinesque . Genus Pappogeomys, Merriam Genus Cratogeomvs, Merriam Genus Platvgeomvs, Merriam Genus Ortliogeomys, Merriam Genus Heterogeomys, Merriam Genus Macrogeomvs, Merriam Genus Zygogeomys, Merriam Superfamily ANOMALUROID.\E Family Anomaluridae Subfamily AiNOMALURINAE Genus Anomalurus, Waterhouse Genus Anomalurops, Matschie Subfamily Idilrinae Genus Idiunis, Matschie Genus Zenkerella, Matschie Superfamily PEDETOIDAE Family Pedetidae Genus Pedetes, llliger Superfamily CTENODACTYLOIDAE Family Ctenodactylidae 437 454 461 464 464 465 468 470 471 472 476 479 480 492 494 505 507 524 527 528 530 531 532 533 534 535 535 537 537 541 542 543 546 547 547 549 553 553 CONTENTS Genus Pectinalor, Blvth Genus Ctetwdactyhis, Gray ■ Genus Massoutiera, Lataste Genus Felm-ia, Lataste Superfamily DIPODOIDAE Family Dipodidae . Subfamily Sicistinae Genus Sicista, Gray . Subfamily Zapodinaf . Genus Eozapus, Preble Genus Zapiis, Coues . Genus Xapaeozapiis, Preble . Subfamily Cardiocraminae . Genus Salpingotm, Vinogradov Genus Cardiocraiiiiis, Satunin Subfamily EfCHOREUTiNAi-: Genus Euclioietites, Sclater . Subfamily Difodinae . Genus Allactaga, Cuvier Genus Alactagidus, Xehring Genus Pygeretmiis, Gloger . Genus Parmiipiis, Vinogradov Genus Dipiis, Zimmermann . Genus Scirtopoda, Brandt Genus jfaciilus, Erxleben Genus Ereiiiodipiis, \'inogradoy Superfamily MUROIDAE , Family Muscardixidae Subfamily Graphiurinae Genus Graphiurus, Smuts Subfamily Musc.ardinixae Genus Eliomxs, \\'agner Genus Dvromys, Thomas Genus Gliruhis, Thomas Genus Glis, Brisson Genus Muscardinus, Kaup Genus Mvoiniiniis, Ognev Subfamily Platacaxthomyinae CONTENTS xix PAGE Genus Platacanthomys, Blyth ...... 627 Genus Typhlomys, Milne-Edwards . 629 Family Lophiomyidae . 632 Genus Lophiomys, Milne-Edwards . 632 Family Spalacidae .... • 636 Genus Spalax, Guldenstaedt ■ 638 Family Rhizomyidae • 644 Genus Rhizomys, Gray . 646 Genus Cannomys, Thomas . . . 651 LIST OF TEXT-FIGURES FIG PAGE I Bathyergiis suillus suillus, Schreber . . Skull X I 82 2 do. do. 83 3 do. Mandible x i 83 4 Cryptomys damarensis, Ogilby . . Skull X 2 88 S do. do. 89 6 Heleroccphalus glaber, Riippell . . Skull X3i 94 7 do. do. 95 8 Echimys armatus armatus, Geoffrey . . Skull X 4 110 9 do. do. no 10 do. Cheekteeth X 7 III 1 1 Proechimys cayennensis, Desmarest . . Skull xij 116 12 do. do. 116 13 do. Cheekteeth x 8 "7 H Geocapromys brownii, Fischer . . Skull x I 130 15 do. do. 130 16 do. Cheekteeth X 7 131 17 Kannabateomxs amblyonyx, Wagner . . Skull xij 138 18 do. do. 138 19 do. Cheekteeth X 5 139 20 Myocastor coypiis santaecruzae, Hollister . Skull X* 141 21 do. do. 142 22 do. Mandible X f 142 23 do. , Skull X i 143 24 do. Cheekteeth x 2 '43 25 Thryonomys gregorianus, Thomas . Skull X I 146 26 do. do. 146 27- do. do. 147 28 do. Cheekteeth X 4 147 29. Petrumus typicus. Smith .... . Skull X 2 150 30- ' do. do. ISO 31- do. Mandible X2 Cheekteeth X 7 151 32. Abrocoma bennetti, Waterhouse . Skull XI* 152 LIST OF TEXT-FIGURES 34- 35- 3b. 37- 38. 39- 40. 41- 4^- 43- 44- 45- 46. 47- 48. 49- 50. 51- 52- 53- 54- 55- 56, 57- 58. 59- 60. 61. 62. ^V 64. ^5- 66. 67. 68. 69. Ahrocorna hennetti, Waterhuuse do. OctoniYS iiiiniax, Thomas Spalacopus cyanus, Molina Ctenomys tuconax, Thomas do. do. Dinomvs hriviickii, Peters do. Chaeiomvs siibspinosiis, Kulil do. do. Ercthizon ipixaiitliuin myops, Merriam do. do. do. Coendou prehensilis holiviensis. Gray . do. do. do. Dasyprocta punctata isthmica, Alston do. do. Trichxs macrotis. Miller . do. At/wniiiis tiirni'ii, St. Legcr do. T/ieciiriis ciassispinis, Giinther . do. Hvstrix cristata, Linnaeus do. Cuniculus paca, Linnaeus do. Chinchjlla laniger, Molina do. Lagostomus maxinius, Desmarest do. PAGE Skull ■ lA 153 Cheekteeth X7 153 Cheekteeth x 8 157 Skull xzA 160 Skull xii 162 do. 162 Mandible x 1 J 163 Cheekteeth x 6 Skull xf 172 do. 172 Skull X I 17s do. 176 Cheekteeth X 3 176 Skull A I 179 do. 179 do. 180 Cheekteeth x 2 J- 180 Skull X I 183 do. 183 do. 184 Cheekteeth X3^ 184 Skull X I 191 do. 192 Cheekteeth X4 192 Skull ■, I 204 do. 204 Skull X I 207 do. 207 Skull X I 210 do. 210 Skull X I 215 Cheekteeth xi\ 216 Skull xr 222 Cheekteeth xzl 223 Skull X i\ 228 do. 228 Skull X \ 233 do. 234 LIST OF TEXT-FIGURES FIG. PAGE 70. Lagostomus maxitmis, Desmarest . Skull X 4 235 71- do. Cheekteeth X2j 235 72. Caviella atistralis joannia, Thomas . . Skull X 2 244 73- do. do. 24s 74- do. . Cheekteeth X 6 24s 75- Hydrochoerus hydrochaeris, Linnaeus . Skull xi 249 76. do. do. 250 77- do. do. 251 78. do. Cheekteeth x i \ 252 79- Aplodontia rufa, Rafinesque . . Skull X I 256 80. do. do. 257 81. do. Mandible x i Cheekteeth X4 257 82. Belomys pearsoni trichotis, Thomas . . Skull X 2 278 83- do. do. 278 84. Trogopterus xanthipes mordax, Thomas . Skull X 5 280 85. Petaurista philippensis, Elhot . . Skull X I 282 86. do. do. 283 87. do. Cheekteeth X 4 J, 283 88. Pteromys volans, Linnaeus . Skull X 2 292 89. do. do. 292 90. do. Cheekteeth X 5 293 91. Myosciurus pumilio, Le Conte . . Skull X 3^ 314 92 do. do. 314 93a. Myosciuriis pumilio, Le Conte . . Skull X 2 315 b. Sciurillus pusilhis, Desmarest . do. c. ,, rnurimis, MiJller & Schlegel do. d. Callosciurus tenuis mrdus, Miller do. 94 Sciurus vulgaris, Linnaeus . Skull xi| 328 95 do. do. 329 96 do. Cheekteeth x 5 329 97 Rheithrosciurus macrotis. Gray . Skull X I 394 98 do. do. 394 99 Rhinosciurus laticaudatus tupaioides, Gray . Skull X 2 397 100 do. do. 397 101 Funisciurus pyrrhopus leonis, Thomas . Skull X 2 410 102 do. do. 411 103 do. Cheekteeth X 8 411 XXIV Liai UF Ih.Vl-FlUUKt H FIG, PAGE 104. Xcnis niti/iis, C'rctzchmar . Skull XX J 419 105. do. do. 419 106. Sptrniopliilopsis leptodact vliis. Lichtenstein . Skull -I A 424 107. do. do. 424 loS. Tcvnias dorsalis, Baird .... . Skull ■,2i 427 log. do. do. 428 1 10. CitcHiis citfUus, I.iiinaeus . Skull . 2 43S III. do. do. 438 112. do. Cheekteeth ■5 439 113- Marmota marnuitii. Linnaeus . . Skull • I 456 114. do. do. 456 115. do. Cheekteeth . t 457 I If). Cxnomvs ludoTiciamis. Ord . Skull ■ I 4^" 117. do. Cheekteeth ■3 464rt 118. Castor fiber, Linnaeus .... . Cheekteeth ■ H 4*>4« 119. do. Skull X i 465(7 120. Iletcriimys aiioimilus, Thompson . Skull X 2* 473 121. do. do. 473 122. do. Cheekteeth 9 474 123- Pcrognathiis hispidus, Baird . Skull X3A 481 124. do. do. 48 1 125. do. Cheekteeth , 17 4S2 126. Dipodomvs merriaini iiieliinunis, Mcrriam . . Skull X 2* 495 127. do. do. 495 12S. do. Cheekteeth 13 496 129. T/ioiiiniiiys pcrpidlidiis, yi^vridm . Skull X 2* 508 130- do. do. 509 131- do. Cheekteeth ; ,6 509 132- Anomalurus fraseri jacksoni. de Winton . Skull xii 53S 133- do. do. 53S 134- do. Cheekteeth 5 539 135- Idiurus macriitis. Miller .... . Skull ■ 3i 544 136. do. do." 545 '37- do. Cheekteeth IS 545 138. Pi'detcs surddstcr larval/s, lloWistev . . Skull ■ I 55° ■39- do. do. 550 140. do. Cheekteeth 4 551 141. C'tciiodiictyhis rtiiiidi, Rothman . Skull xii 557 LIST OF TEXT-FIGURES . Ctenodactylus gundi, Rothman . Skull xij 557 do. Mandible x i J Cheekteeth x-j 558 . Sicista siiblilis loriger, Nathusius . Skull X 4 565 do. do. 565 do. Cheekteeth X 9 566 . Zaptis hudsonius, Zimmermann . Skull X3I 570 do. do. 570 do. Cheekteeth X 15 571 . Eiichoreutes naso, Sclater . Skull X 3 578 do. do. 578 do. Cheekteeth x 1 1 579 . Allactaga euphratica, Thomas . Skull X2i 58. do. do. 581 do. Cheekteeth X 8 582 . Jacultis jaculus, Linnaeus . Skull X2i 594 do. do. 594 do. Cheekteeth x 1 1 595 . Graphiurus hueti, Rochebrune . Skull X2i 605 do. do. 605 do. Cheekteeth x 10 606 . Eliomys quercinus, Linnaeus . . Skull X2i 614 do. do. 614 do. Cheekteeth x 10 615 . Dyrotnys nitedula, Pallas . Skull X 3^ 617 do. do. 617 do. Cheekteeth x 10 618 . Glis glis, Linnaeus . . . . . Skull X 2 621 do. do. 621 do. Cheekteeth x 10 622 . Miiscardinus avellanarius, Linnaeus . Skull X3 J 624 do. do. 624 do. Cheekteeth x 10 625 . Platacanthomys lasiurus, Blyth . Skull X2.\ 628 do. do. 628 do. Cheekteeth xio 629 . Typhlomys cinereiis, Milne-Edwards . Skull X4 630 do. do. 630 XXVI LIST OK TEXI-FIUU IRES FIG. PAGE 179. Typhlomys cinereus, Milne-Edwards . Cheekteeth X 14 63. 180. Lophiom\s imhausi, Milne-Edwards . . Skull:-: a ('33 181. do. do. 634 182. do. Cheekteeth x 5 634 183. Spalax monticola dolbrogcac, Miller . . Skull xii 639 184. do. Cheekteeth x 5 640 185. Rhizomvs pruinosus, Blyth . Skull xii 647 186. do. do. 647 187. do. do. 648 188. do. do. 64S 189. do. Cheekteeth x 5 649 (N.B. — In the figures of cheekteeth the left-hand figure (in a few cases lettered a) shows the upper tooth row of the right side seen from below; the right-hand figure (occasionally lettered b) shows the lower tooth row of the left side seen from above.) LIMITS OF THE ORDER RODENTIA In 1912 (Science, New York, n.s. XXXVI, p. 285), Gidley proposed a separate Order for the Rodentia Duplicidentata (Leporidae and Ochotonidae), and restricted the Order Rodentia to the great mass of animals usually known as Rodentia Simplicidentata. This division is currently accepted by American authors, usually not so elsewhere (excepting Flower, 1927, Vertebrate List, Zool. Soc. London, 1828-1927, Mammals, p. 239). Whether it is a classification that is likely to be universally followed seems open to question. Gregory' in his excellent work The Orders of Mammals, Bull. Amer. Mus. X.H. XXVII, 1910, inclined to the contrary opinion, though stating that no fossil forms have yet been discovered that will connect the Duplicidentata with the Simplicidentata. I can only say that in my opinion the Lagomorpha mav reasonably be regarded as an Order distinct, and that for the purposes of the present work they are regarded as such. The fundamental ditTerences in the appearance of those parts of the skull to which jaw-muscles are attached may surely at once be stated in the Lagomorpha to be a much more important character than the retention of the functionless second upper incisor which seems to be quoted always as the main difference between the two groups. To those of the contrarv' opinion I must plead guilt)' of saying: notes on the characters of three hundred and forty-five genera containing sixt\--four hundred forms proved enough work tor one; I do not look for extra forms to include in an Order as vast as this; and I have no intention of including in this Order forms which may very well not belong there. Before including the Lagomorpha in the Order, let us wait until an intermediate family is discovered fossil between the two groups. Because the fact that both Rodents and Lagomorphs are adapted for gnawing does not seem to prove conclusively that they must of necessity be so nearly related as to be included in the same Order. The Order Rodentia, therefore, as here understood, has been defined by Miller & Gidley as follows: "Terrestrial and fossorial, occasionally arboreal or semi-aquatic placental mammals with both brain and placentation generalized in type; feet unguicu- late; elbow joint always permitting free rotary motion of forearm; fibula never articulating with calcaneum; masseter muscle highly specialized, divided into three or more distinct portions, having slightly different functions; caecum without spiral fold; dental formula not knowTi to exceed i. j, c. y, p. t, m. 5=22 permanent teeth; incisors scalpriform growing from persistent pulp, the enamel of upper tooth not extending to posterior surface ; distance between mandibular and maxillar\- toothrows approximately equal, both pairs of rows capable of partial or complete opposition at the same time, the primarj- motion of the lower jaw in mastication longitudinal or oblique." For further notes on the essential characters of the Order see TuUberg, 1 — Livine Rodents — I I 2 LIMITS OF THE ORDER RODENTIA Uehcr das System der Nagetiere, 1899; Flower & Lydekker, .Mammals Lhing and Extinct, 1891, pp. 443-448; Gregory, Orders of Mammals, 1910, p. 323; Weber, Die Saugetiere, II, p. 238, 1928; Gidley, Science, n.s. XXXVI, p. 285 (the separation of the Order Lagomorpha from Rodentia, and differences between the two Orders); Parsons, 1894, Proc. Zool. Soc. London, p. 251, Myology of Sciuromorphine and Hystricomorphine Rodents, etc. VARIATION The great variation or diversity in structure within the Order is one of the features that makes it such an interesting study; it is perfectly safe to say that nowhere among all Orders of mammals is such div'ersity found in a single Order with the possible exception of the Marsupials. Great specialization has been attained independently again and again within this order for various modes of life. Among these may be mentioned extreme modification for aquatic life, most developed in Hvdromys and immediate allies (Muridae, Hvdromvinae); Ichtlivomvs and immediate allies (Muridae, Crice- tinae) ; to a lesser degree in Ondatra (Muridae, Microtinae), Myocastor (Echimy- idae), and Castor (Castoridae); modification attaining a high degree for arboreal life; two families contain "flying" genera, with a flying membrane which enables them to take flying leaps from tree to tree (Sciuridae, Pteromys group), and all genera of Anomaluridae except Zenkerella) ; in non-flying arboreal types, the Erethizontidae show great specialization of the feet, the hallux sometimes being replaced by a broad movable pad [Coendou, Echinoprocta, Chaetomys); and many Muridae (Murinae) of the Indo-Malayan region have a fully opposable hallux, as Hapalomys, Vandeleuria, Chiropodomys, Chiroinyscus, etc. -Many genera are fully modified for underground life, with Mole-like appearance, and either immensely developed incisors, or immensely developed claws for digging. In the Muridae, examples are Ellobiiis and Prometheomys (Micro- tinae), Xotiomys (Cricetinae), and Myospalax (Mvospalacinae). The Spalacidae (Spalax) go a step further than any in that even the eyes are suppressed. The Geomvidae, quite unrelated to the above, are all as highly specialized for underground life as any Murine; as are the African Bathyergidae, another isolated group. And in South America, certain Hystricoid genera, as Ctenomvs and Spalacopus, have taken to this form of life and become just as modified in external form. High specialization towards bipedal saltatorial life is developed in three or possibly four unrelated families, with Kangaroo-like form and elongated hindlimbs and tail; the Dipodidae (all genera except Sicista), the Pedetidae, the Heteromyidae (Dipodomys and Microdipodops), and probably one member of the Muridae, Xotomys (Murinae), from .\ustralia. Also perhaps some Gerbils. Cursorial specialization, with form like that of a primitive Ungulate, and reduction of the digits of the hindfoot to three has taken place twice in South -America, in the Dasyproctidae, and in the Caviidae (Dolkliotis). Great specialization of the covering of the body into spines is shown in the Hystricidae, being in this family at extreme development superior to that of any other Rodent and perhaps any other Mammal. Other families which have a more or less specialized spiny covering are the Erethizontidae (most effective!), certain genera of Echimyidae (not highly developed), and certain Muridae, as in some species of Rattus, and in Acomys; also to a lesser degree in the Cricetine 3 4 VARIATION genus Neacomvs. One genus of Muscardinidae, Platacantliomys, is spiny. And in two families of Rodentia extraordinarj' and abnormal development in the skull has taken place, the Cuniculidae, with their enormous bony cheekplates, and the Lophiomyidae, in which the temporal fossae are roofed in by bony plates. PREVIOUS CLASSIFICATIONS OF ORDER OLDFIELD THOMAS, 1896 (Proc. Zool. Soc. London, 1896, p. 1012) (Suborder SIMPLICIDENTATA) A. ANOMALURI Family Anomaluridae Anomahirus, Idiurus. B. SCIURO.MORPHA Family Sciuridae Subfamily Sciurinae (a) Rheithrosciurus, Xerus, Sciurus, Tamias, " Spermophilus" ( = Citellus), Cynomys, " Arctomys" ( = Marmota) (b) Eupetaurus, Petaurista, " Sciuropterus" { = Pteromys) Subfamily Nannosciurinae Nannosciuriis Family Castoridae Castor C. APLODONTIAE Family Aplodontiidae Aplodontia D. MYOMORPHA Family "Gliridae" Subfamily "Glirinae" Glis, Eliomys, Muscardinus, Graphiurus Subfamily Platacanthomyinae Phitacanthomys, Tvphlom \s Family IMlridae Subfamily Hydromyinae Hydromys, Xeromys, Chrotomys Subfamily Rhynchomyinae Rhytichumys Subfamily Phloeomyinae Phloeomxs Subfamily Gerbillinae Gerbilliis, Pachyuromys, Psammomys, Meriones, Rhombomys Subfamily Otomyinae Otomys, " Oreinomys" { = Oiomys) PREVIOUS CLASSIFICATIONS: THOMAS Subfamily Dendromyinae Deomvs, Dendromus, Stciitomxs, Mahiculhrix, l.cimiuomys Subt'aniilv Miirinae Miis, Xesukiti, Cricetomys, Malacoinys, Luplntromys, Saccostoiinis, Acoinys, Ariicantliis, Cjohiiida, lande/eiiiiii, C/iiropudoinvs. liat- omys, Carpomys, " Chiruromys" ( = Pogoiiomys), Hapalomvs, Pitheclieir, Cniteromys, " Craiiiot/nlx" { — Echiot/irix). Miistiic- onivs, I'romvs, Conihirus Subfamily Lophiomyinae Lup/lWI?!VS Subfamily " Sigmodontinae" "Hamster" { = Cricetiis), Mystromys, Bicuiivtcusoinvs, A'esomvs, " Ihilloiiivs" { = .\eso?iiys), Braclivuromxs, Hvpoi;eomvs, Eliurus, Gymnurumys, Onychomys, Peroinysctis, Rhipidomys, Tyloiiivs, Holuchihis, Sigmodou, Oryzomys, Reithrudontomys, Eligmodontia, Neoiomvs, Reithrotion, PhxUotis, Scapteroinys, Iclithyomys, Akodon, Oxvmvctenis, Bhiiiiwinys, Xotiomxs Subfamily Neotominae Neotomii, Xenomys, Hodomys Subfamily iXIicrotinae Plienacoinys, " Erotoinys" (--C/ethilunuinys), Mkiutiis, Syiuipt- oinys, Lemmus, Dicrostonyx, Ellohius Subfamily " Siphneinae" "Siphiieus" ( = Myospalax) Family Sfalacidah Subfamily Rhizomyinae Rliizoinys, Tacliyoryctes Subfamily Spalacinae Spalax Family Geomyidae Geomys, Thomomys Family IIeteromyidae Subfamily Dipodomyinae Dipodomys, " Perodipiis" { = Dlpodoiiiys), Microdipudops Subfamily Heteromyinae Perogniithus, Heteromxs Family Bathvergiuae Biitliyergus, Gcoiycliiis, " Myoscolops" (-^ Heliophohius), Hetero- cepludus) Family Dipodidae Subfamily "Sminthinae" " Smiiithus" ( = Sicista) Subfamily Zapodinae Zapiis Subfamily Dipodinae Dipiis, A/lactaga, " Platycercomys" (== Pygeietinus), Eiuhoreiites PREVIOUS CLASSIFICATIONS: THOMAS 7 E. HYSTRICOMORPHA Family Pedetidae Pedetes Family Octodontidae Subfamily Ctenodactylinae Ctrnodactylus, Massoiiliera, Pectinator, Petromvs Subfamily Octodontinac Ctenumys, Aconaemys, Spalacopus, Octudon, Ahrocuma Subfamily "Loncherinae" (a) Dactylomys, Thrinacodus, Kannabateomys, " Lonclwres" ( ^ Echimys) [b) " Thrichomys" { = Cercomys), Cercomys, Carterodon, Mesomys, " Echinomys" ( = Proechimys) Subfamily Capromvinae Myocastor, Capromys, Plagiodontia, Thryonomys Family Hystricidae Hystrix, Atherura, Tricky s Family Erethizontidae Subfamily Erethizontinac Erethizon, Coendou Subfamily Chaetomyinae Chaetomys Family Chinchillidae Chinchilla, Lagidium, Lagostomus Family Dasyproctidae Dasyprocta, " Coelogenys" { = Cuniculus) Family Dinomyidae Dinomvs Family Caviidae Cai'ia, Dolichotis, Hydruchoerus This classification is admittedly nothing more than a rearrangement and bringing up to date of an earlier classification of Alston. It may at once be discarded as unnatural, as being based mainly on the character of the fusion or separation of tibia and fibula. In the "Myomorpha" of Alston these bones are fused; in the " Sciuromorpha" and "llystricomorpha" they are separate. Alston states that "in the few cases in which the cranial differences fail us in separating the Sciurine Rodents from the Murine, and the latter from the I lystri- cine, the complete ankylosis of the lower part of the tibia and fibula in the second group comes to our aid." As Bathyergidae (which have Hystricine mandible formation) are placed in " Myomorpha " on account of the fibula structure, presumably this was considered the chief character in placing a Rodent systematically. But a Rodent did not become a Rodent because its fibula fused or remained separate. If Flower and Lydekker's book Mammals Living and Extinct is looked through with relation to the classification of other Orders, it may be seen S PREVIOUS CLASSIFICATIONS: TULLBERG that in one family of Insectivora, the Centetidae (now I believe known as Tenrecidae), in one branch of the family the tibia and fibula are described as fused (Oryzorictinae, p. 638), in the other branch they are distinct (Centetinae, p. 637). It therefore these two conditions may exist in the same family of Insectivora, the character is surelv one which can scarcely be used for super- familv arrangement in another Order. A Rodent becomes a Rodent because it gnaws, and its gnawing is done with its incisors (which do not vary throughout the genera of modern Rodentia sufficiently for any supertamily grouping to be arranged on this account), and with its jaw-muscles; and the jaw-muscles have modified those portions of the skull, to which they are attached, in various ways throughout the larger groups, as recognized by all other authors who have comparatively recently attempted a classification of the Order. I venture to suggest that if Raitiis, for example, had never taken to gnawing, whatever the condition of its hindleg bones it would not be classed in the present Order to-day; or if Orvsorictes had bv chance taken to this form of life and developed the characteristic Rodent skull, dentition, and jaw-muscle arrangement, whether or not its tibia and fibula were fused, it would automaticallv have come under the heading of "Rodent" as understood to-day. If the Order is to be dumped into three superfamilies or "waste-paper baskets" the " Sciuromorpha," "Myomorpha," and "Hvstrico- morpha" (into which as I see it the families will not naturally go), it should be on a verv different basis from that of Alston and Thomas, and more like that of Tullberg (below), with a separate group " Bathyergomorpha" for the Bathyer- gidae, and probably one also for the Aplodontiidae, the Sciuromorph branch to include Geomvidae, the Myomorph branch to include Ctenodactvlidae, Anomaluridae and Pedetidae, etc. TULLBERG, 1899 (Utber das .System der XaKcthierc. Nova .Acta Reg. Soc. Sci. Upsaliensis, vol. XVIII, No. I) (Suborder SIMPLICIDENT.'\TI) Tribus I. HYSTRICOGN.Vnil Subtribus i. BATHYERGOMORPHI Family B.\thverg!d.\e ( Georychiis, Biithyergns) Subtribus 2. HYSTRICOMORPHI Family Hystricidae {Hystiix, Athirura) Family C.wiidae (" Coelogenvs" { = CunkuIin), Dasypiocta, Ciivia, Dolichotis. Hydiochoena) Family Erethizontidaf. {Erethizun, Coendoii, Chaetumys) PREVIOUS CLASSIFICATIONS: TULLBERG 9 Family Ciiinchili.idae (Chinchilla, Lagidium, Lagostomus) Family " Ailacodidae" ("A ulacodus " ( = Thryonomys)) Family Echinomvidae Subfamily " .Myopotamini " (" .1 lyopotamus " ( = Myocastor)) Subfamily Echinomyini Echinomyes {" Echinomys" ( = Proechimys), " Nelomys" (based on a species of Cercomys), Kaimabateomys) Octodontes ("Habrocoma" { = Abrocoma), Octodon, Spalacopus, Ctenomys) F"amily Petromyidae (Petromys) Tribus2. SCIUROGNATHI Subtribus i. MYOMORPHI Sectio I. Ctenodactyloidei Family Ctenodactylidae {Ctenodactylus) Sectio 2. Atwmaluroidei Family Ano.maluridae (Anoimilurus) Family Pedetidae (Pedetes) Sectio 3. Myoidei Subsectio i. IVIyoxiformes Family "Myoxidae" (Graphiurus, " Myoxus" { = Glis), Eliomys, Muscar- dtnus) Subsectio 2. Dipodiformes Family Dipodidae ("Sminthus" { = Sicista), Zapus, Dipus (based on Jaculus), Allactaga) Subsectio 3. INIuriformes Family Spalacidae ("Siphneus" [^ Myospalax), Spalax, Rhisomvs, Tachyoryctes) Family Nesomyidae (Gymniiromys, Nesomys, Eliiirus, Brachyuromvs, Brachytarsomys) Family Cricetidae [Cricetus) lo PREVIOUS CLASSIFICATIONS: TULLBERG Family LoPHiOMYiDAn (iMphiovivs) Family Arvicolidae (Ellohius, Anicola, Neofiber, "Fiber" ( = Ondalra), " Ciiiiiculus" (^- Dicrostonyx), " Mvodes" ( = Lem- III lis)) Family Hesperomyidak (Hespeioiiivs (based on Peioiiivsciis), Neotoma, Siginodoii, Nectomvs, Oxvmvctenis) Family Muridae Subfamily Alurini (Miis, Nesokia, Cliiropodomys, " Hapalotis" (based on Notomvs), Hvdroiiivs, Dendioiiiiis, Steaiomvs, Saccostoiiius, Cricetomvs, Deoiiivs, Lopliiiromvs) Subfamily Phloeomyini (Phloeomxs) Subfamily Otomyini (Otomvs) Family Gerbillidae (Gerbillus, Psammomvs) Subtribus 2. SCILROMORPHI Sectio 1. Sciiiroidei Family ' ' H aplodontidae ' ' {" Haplodoii" ( = Aplodoiitia)) Family Sciuridae (Sciurus, " Sciuropteiiis" {^ Pteromvs), " Pteromvs" ( = Petaurista), " Arctoinvs" ( = Marmota), Cynomvs, " Spernioplii/us" { = Citelliis), Tainias) Sectio 2. Castoroidel Family Castoridae (Castor). Sectio 3. Geomyoidei Family Geomyidae Subfamily Dipodomyini {" Perodipiis" (- Dipodoinvs), Dipodonivs, Perogna- tlius, Heteromvs) Subfamily Geomyini (Geoiiivs, Tlwmoiiixs) This is in my mind perhaps the best classification of the Order that has been done. The only points which seem unnatural are the too close association of Aplodontiidae with Sciuridae, the lumping together of all the Old World Murine " borrowers " as a family Spalacidae (R/iizoiiiys, Myospahix, Tachy- oryctes, Spalax), and the lumping together of the Dasyproctidae and Cuniculidae with the Caviidae. PREVIOUS CLASSIFICATIONS: WEBER ii MAX WEBER, 1904, 2nd ed., 1928 (Die Saugetiere, 1928, II, p. 238) (Living forms only) (Suborder SIMPLICIDENTATA) Tribus 1. HAPLODONTOIDEA Family "Haplodontidae" {Aplodontia) Tribus 2. SCIUROIDEA Family Sciuridae (Sciurus, Neosciunis, Rheithrosciurus, Cullosciurus, Tamiasciurus, Ratufa, Heliosciurus, Funisciuriis, Nannosciurus, Funambiihis) Family Pteromyidae (Eupetaurus, Pteromys {—Petaurista), " Sciuropterus" (=Pter- omys), Glaucomys) Family Xeridae {Xerus, Geosciuriis) Family Tamiidae (Tamias, Eutaniias, Citellus) Family Marmotidae (Cynomys, Marmota) Tribus 3. CASTOROIDEA Family Castoridae {Castor) Tribus 4. GEO.MYOIDEA Family Heteromyidae {Dipodomys, Heteromys, Perognathus) Family Geomyidae {Geomys, Thomomys) Tribus 5. ANOMALUROIDEA Family Anomaluridae (Anomaluriis, Idiurus, Zenkerella) Family Pedetidae (Pedeies) Tribus 6. "AIYOXOIDEA" Family "Myoxidae" {Graphiurus, Muscardimis, Glis, " Dryoniys" {=Dyromys), Eliomvs) Family Platacanthomyidae (Platacanthotnys, Txphlomys) Tribus 7. DIPODOIDEA Family Sicistidae {Sicista) 12 PREVIOUS CLASSIFICATIONS: WEBER Family Dipodidae (Zapiis, Allactaga, " Scartiirus" {=AUactaga), Dipiis, Jacidus, Euchoreules, Pygerettmis) TribusS. MYOIDEA Family Spalacidae (Spalax, Rhizomys, Tachyoryctes, " Myotalpa" { = Myospalax)) Family Nesomyidae (Brachxuroinxs, Nesomys, Braclivtarsoinys, Eliuriis) Family Mlridae Subfamily Cricetinae (Cricetus, Mesocricetus, Crketulus, Mystromys, Hesperomys, Pero- mvsciis, Oryzotuys, Reithrodoiitomys, Sigmodon, Tylomys, Holo- cliilus, Xectomys, Eligmodontia, Ichthyomys) Subfamily Lophiomyinae (Lophiotnys) Subfamily Microtinae {Lemmus, Myopiis, Dicrostonyx (group Lemmi) ; EUobius (group EUobii); " Erotomys" (=Clethrionomys), "Fiber" (=Ondatra), Microtus, Pitymys, Arvicola (group Microti)) Subfamily Murinae (" Epimxs" (--Rattiis), Mm, Apodemus, Mkivmys, Nesokia, Pliloeomys, Pithecheir, Crketomys, Saccostoinus, Otomys, " Oreomvs" (=Otomvs), Dendromus, Deomvs, Mastacomys, Lepo- rillus, Uroiiivs, Mallomys, Conilurus, " Chiruromys" {^Pogo- nomxs)) Subfamily Cit-rbillinae (Gerbilbis, Pacliviiiomvs, Meriones, Rhombomys, Psammomys) Subfamily Hvdromyinae (Hydromys, Leptomys, Xeromys, Celaenomys. Chrntomys. Cru- noiiivs, Rliyncliomys, " Craurothrix" (=Echiothrix)) Tribus 9. BATHYERGOIDEA Family Bathyergidae {Buthyeigits, Georychus, " Myoscalops" { --IJe/iophohiiis), Hetero- ccpluilus) Tribus 10. HYSTRICOIDEA Family Hystricidae {Ilystrix, Atherwa, Trichys) Family Erethizontidae (Eretliizvn, Coendoii, Cliaetoinys) Family Cayiidae Subfamily Dinomyinae (Dinomys) Subfamily Dasyproctinac {"Coelogenys" (=Cuiiiciilus), Dasyprocta, Myoprocta) PREVIOUS CLASSIFICATIONS: MILLER & GIDLEY 13 Subfamily Caviinae (Cai-ia) Subfamily Hydrochoerinae {Ilydrochoenis, Doliclwtis) Family Chinchillidae (Chinchilla, Lagidium, " Vizcacia" {=Lagostomus)) Family Capromyidae (Myocastor, Capromys, Plagiodontia) Family Octodontidae {Echiniys, Octodon, Ctenomys) Family Ctenodactylidae (Ctenodactylus, Pectinator, Petromys) Family Thryonomyidae ( Thryonomys) So far as Superfamily grouping is concerned, this classification is followed in the present book, with some modifications, as for instance the separation of Pedetidae from Anomaluridae; also following Tullberg I cannot credit that the Ctenodactylidae should be referred to the Hystricoid branch [Petromiis is here transferred to the Echimyidae), and also on account of intermediate forms I am unable to find characters to keep the IMuscardinidae separate as a super- family from the Muroidae. Several of Weber's divisions into families appear unnecessary or unnatural, as the dividing up of the Sciuroidea into five "families"; the retention of the old family Spalacidae; the retention of the "family Nesomyidae"; the lumping together of Dinomvs, Cuniculus and Dasvprocta with the Caviidae; and the formation of a subfamily Hydrochoerinae including Dolichotis (obviouslv very nearly allied to Cavia) against the Caviinae with Cavia only. MILLER & GIDLEY, 1918 (Joum. Washington Acad. Sci., VIII, No. 13. p. 431, 1918) The Order, including Rodentia as here understood only (the Lagomorpha not included), is divided into five superfamilies based on zvgomasseteric structure. Superfamily SCIUROIDAE Three-cusped series Family Sciuridae Subfamily Sciurinae (the entire familv except the two follow- ing groups) Subfamily Nannosciurinae (Xannosciurus. Myosciurus, Sciurilliis) Subfamily Pteromyinae (Flying-Squirrels) Family Geo.myidae Subfamily Entoptychinae (fossil) [Entoptychiis) 14 PREVIOUS CLASSIFICATIONS; MILLER & GIDLEY Subfamily Geomyinae (North American Pocket-Gophers) Family Hetfro.myidae (North American Pocket-Mice and Kan- garoo-Rats, Oligocene (HcUscomys) to Recent) Four-cusped series Family Ahjidacmidae (fossil) {Adjidauino) Family Eutypomyidae (fossil) (Eutvpoviys) Family Chalicomvidae (fossil) (Chalicomvs (Steneofiber) and related genera; Tros^otio- therium, Palaeocastor, Eiicastor and related genera) F'amily Castoridae (Castor) Family Castoroididae (fossil) (Castoroides) Superfamily MUROIDAE Three-cusped series Family Muscardimdae (Eliomvs, Dvromvs, Glis, Muscaidiniis, also Leitliia (fossil)) Four-cusped series Family Ischyromvidae (fossil) {Ischvioinys) Family Cricetidae Subfamily Cricetinae (Cricctinae, Sigmodontinae, Neotominae and Nesomvinae of authors) Subfamily Gerbillinae (Gerbillinae of authors) Subfamily Microtinae (Microtinae of authors) Subfamily Lophiomvinae (Lopliiornvs) Family PLATACANTHOMYinAE (PldtacantJioinxs, Txphlomvs) Family Rhizomyidae Subfamily Tachyoryctinae ( Tachyoryctes) Subfamily Rhizomyinae {Rliizomvs and related genera) Subfamily Braminae (fossil) (Braiiiiis (-- EUohiui), a Microtine; see Hinton, Monogr. \'olfs & Lemmings, I, p. S7, 1926) Family Spalacidae Subfamily Myospalacinae (Myospahix) Subfamily Spalacinae (Spalax, Recent; Prospalax (fossil)) PREVIOUS CLASSIFICATIONS: MILLER & GIDLEY ,5 Family Muridae Subfamily Dendromyinae (Dendromvinae of authors) Subfamily Murinae (Murinae of authors) Subfamily Phloeomyinae (Phloeomys) Subfamily (jtomyinae (Otomvs) Subfamily Hydromvinae (Hydromvinae of authors! Superfamily DIPODOIDAE ' ' ' Three-cusped series Group A Family Pammyidae (fossil) {Paramys, Mysops, Prosciiirus and related genera) Group B o / Family Graphiuridae (Graphiuriis) Group C Family Allomyidae (fossil) (Allomys, Haplomys, Meniscomys, Mylagaiilodon) Family Aplodontiidae {Aplodontia, Recent; Liodontia, fossil) Family Cylindrodontidae (fossil) (Cylindrodoii) (Position of group doubtful) Four-cusped series Group A Family Pseudosciuridae (fossil) {Psendosciurus) Group B Family Mylagaulidae (fossil) (Mylagaulus, Ceratogaulus, Epigaulus) Group C Family Anomallridae [Auomalurus) Family Idiuridae Subfamily Idiurinae (Idiurus) Subfamily Zenkerellinae (Zenkerella) Group D Family Scilravidae (fossil) {Sciuravus) Family Zapodid.\e Subfamily Thcridomyinae (fossil) (Theridomyidae of authors) Subfamily Sicistinae {Sicista, Recent; }Eomys, fossil) i6 PREVIOUS CLASSIFICATIONS: MILLER & GIDLEY Subfamily Zapodinac (Eozdpiis, Zapiis, Sapaeozapus) Family Dipodidap, Subfamily Protiiptvchinae (fossil) (PriitoptycliHs) Subfamily Dipodinac (Dipodidae of authors, who recognize Zapodidae as a distinct family) Family Ctknodactylidae (Ctenodactylus and related genera) Family Pedktidak " (Pedctes) Superfamily BA'l'I lYERGOIDAE I'amily BATiiYnRciDAE (Bathyergidae of authors) Superfamily HYSTRICOIDAE Lateralis series. Group A. Family Hystricidae Subfamily Hystricinae (Hvs/n'x, A Clint /lion, Tlieciiriis) Subfamily Atherurinae (Atlicrurus, Trichys) Family Erethizontidae (New World Porcupines except Chaetomys; fossil genera, Asteromvs. Eosteiromvs, PmasteironiYs, Steiroinvs; ?PliiomYS, }iM('/llp/lioil!YS) Family Echimyidae Subfamily Echimyinae (Spiny Rats, provisionally including Chaetoniys; Hutias (Ciipromvs, Plagiodoiitiii, etc.); many extinct genera, among them AciireiiiYS, Boromvs, BrotoiiiYS, Colpostciiiiui, Eociirdia, Eoctodoii, GraphiiiiYS, Gyngiiophiis, Haplustroplia, Hetei- opsoiiiYS, Homopsomys, Isobolodon, Prospaiiiomys, Prottidel- plioiiiYS, Protacaremys, Sciiiinvs, Scleromys, Spaniomys, StichiniYS, Strophostephamis, Tribodon) Subfamily Octodontinae (CteiioiiiYs, Octodoii, OctodoiitomYs. Spahicopus. Amongst fossil genera are CeplinloiiiYS, Dicoe/ophoiiis, Eiicoe/oplionis, LitoduiituiiiYS, NeopIuinmiiYS, Palaeoctodoii, Plilorumys, Pitli- tiiioliiiiiYS, PlatoeoiiiYS, ScotuiiiYs) Family Petromyidae (PctrnmYs) Family Myocastoridae (MYociistor and related fossil genera) Family Tiiryonomyidae ( Thryonomys) PREVIOUS CLASSIFICATIONS: MILLER & GIDLEY 17 Family Dinomyidae (Includes the living Dwomji and extinct genera Amblyrhiza, Briaromys, Discolomys, Elasmodontomys, Gyriabrus, Me- gamys, .Xeoepiblema, Olenopsis, Potamarchus, Tetrastylus) Family Cunil(;lid.\e [Cuniculus) Family Hept.wodontidae (fossil) {Heptaxodon; JMurenia) Group B Family Dasyproctidae (Dasyproctidae of authors with Cuniculus removed, and Neoreomys (fossil) added) Family Chinchillidae (Chinchilla, Lagostomus, " Viscaccia" ( = Lagidium) ; extinct genera: Euphilus, Perimys, Pliolagostomus, Prolagostomus, Scotaeumys , Sphaeromys) Family Abrocomidae (Abrocoma) Medialis series Family Caviidae (Caviidac of authors, with Hydrochoerus and allies removed ; extinct genera : Anchimys, \eoprocavia, Orthoniyctera, Palaeocaiia, Phugatherium, Procardio- therium) Family Hydrochoeridae {Hydrochoerus and extinct allies, Plexochoeriis, Prohydro- choerus, Protohydrochoertis; perhaps Cardiomys, Cariodon and Cardial herium) Great attention has been paid to this classification, on which the present work was originally based. It attends much more strictly to detail characters than either those of Weber or Winge. But it seems to break down where the "Dipodoidae" (for instance, Graphiurus) and Muroidae are compared; it does not appear good classification to separate Graphiurus from the Muscar- dinidae so far that it is placed in another superfamily; moreover, it appears that the Murine genus Deomys has the "Dipodoid" zygomasseteric structure as defined by Miller & Gidley, and should be referred to that superfamilv if this classification was maintained. As already noted by Wood in his monograph of Heteromyidae, there is a very wide distinction between the zygomasseteric structure of Paraniys, and the Mylagaulidae and that of the Dipodidae, both referred to "Dipodoidae" of Miller iSc Gidley; from descriptions Paramvs and Mylagaulidae have similar zygomasseteric structure to that of Aplodontia, here considered the most primitive living Rodent as regards this arrangement; Wood states, "The character of zygomasseteric structure as given by Miller & Gidley . . . includes widely different types, which appear to have reached their present condition in widely different manners. In Paramys and the Mylagaulidae . . . the zygomatic plate is nearly horizontal because that is the primitive condition for Rodents, and the growing masseter has not as yet 2 — Living Rodents — I i8 PREVIOUS CLASSIFICATIONS: WINGE effected any great change. In the Dipodidae . . . the zygomatic plate is hori- zontal because the masseter has passed through the infraorbital fenestra and, on expanding, has forced the zygoma down until it becomes even lower than in the primitive forms." In Miller & Gidley, the families appear in some cases to be over-split. For the undesirahility of separating "Cricetidae" from Muridae, see Hinton, Monogr. \'oles & Lemmings, 1926, p. 121. WINGE, 1924 (Pattcdyr Slaegter, vol. II, Rodcntia, p. i) The Order is divided into nine families, one of which, the Leporidae, corresponds to the Suborder Duplicidentata or Order Lagomorpha. 1. Leporid.M' (Leporini: Palacolagus (fossil), Lepus ; "Lagomyini": " Lagomys" = Ocliotoiia) 2. "H.'\plodontidae" AUomyini (fossil) AUomys Ischyromvini (fossil) Paramys, Sciuravus, Ischyromys Mylagaulini (fossil) Mxlagaidus, Ceratogaulus "Haplodontini" " Hapludoii" = Aplodontia (sole living genus) 3. Anomaluridae Pseudosciurini (fossil) Pseiidosciurus, Sciuroides Trechomyini (fossil) Trecliovivs Anomalurini Aiwmalunis, " Aet/innis" = Zenkerelhi, Idiurus Theridomyini (fossil) Tlieridomys, Issiodoromys, Archaeomys Pedetini Pedetes 4. Dipodidae Eomyini (fossil) Eomys Dipodini " Sminthi" " Simiithus" = Sicista ; "Jactilus" based on Zapus Euchoreutae Etichoreiites Dipodes " Scirtetes" = Allactagu, Dipii^ PREVIOUS CLASSIFICATIONS: WINGE ,9 Spalacini Spalax 5. "Myoxidae" Graphiurini Graphiurus "Myoxini" "Myoxi" Eliomys, Leitbia (fossil), Hypnomys (fossil), " Myoxus" = Glis Muscurdiniis Platacanthomyes Platacantliomys 6. MURIDAE Rhizomyini Cricetodontes (fossil) Cncetodon, Eumys Rhizomyes Nesomys, Brac/iyfarsomys, Gvmnuromys, Eliurus, Brachvuromvs lachyoryctes, Rhizomys ' ' Cricetini Criceti Cricetus, Calomyscus, Lophiomys, " Siphneiis" = Myospdax Hesperomyes Hesperomys, Sigmodon, Neotoma, '' Habrothrix" (apparently based on_ Akodon), Oxymycterus, Ichthvomys, Scapteromys, ^alomys (species or genera referred to this group not clear) Hhtptdomys, Nectomys Arvicolae ''Hypudaeus" (=Cletlmonomys plus Dolomys plus Phenacomys), htber =Ondatra, Ellobiiis, Arvkola (including Microtus) Dtcrostonyx, " Myodes" = Lemmus Murini Mures Mus, " Spalacomys" =. Xesokia, PIdoeomvs, Crateromvs, Carpnmys tchiotmx, Rhynchomys, Leiwmys, Vandekuria, Chiropodomys Hapalomys; C/uiuromys, Notomvs, Mastacomvs, Zvzomvs, - Hapa- lotis" = Comhtrus ; Acomys, Cricetomys, Saccosto,nus,\Steatomys, Dendromiis, Deomys, Oenomys, Otomys Gerbilli Gerbillus, Rliombonns Hydromyes Xeromys, Crunomys, Hydromys, Chrotomys, Celaenomys 7. Hystricidae Bathyergini Bathyergiis, Heterocephalus, Georychus, Heliophobius 20 PREVIOUS CLASSIFICATIONS; WINGE Hystricini Hystrices Triciiys, Athenira, llvstiix " Sphinguri" Steiromys (fossil), Eietliisoii, " Sphtnguriis" = Coendou, Chaetomys Capromyini " Aulacodus" = Thryonomys, " Myopotamus" = Myocastor, Cap- romys, " Plagiodon" — Plagiodontia ; hobolodon (fossil) Ctenodactylini Petro)iivs, Pectinator, Ctenodactylus Dasvproctini Dinomyes Dinomys ; Elasmodontomys (fossil) Dasyproctac Dasypiocta ; Xeoicomys (fossil); "Coelogeiiys" =Ciinicii!iis Caviae Sclu'stomvs, Eocardia (both fossil), Cavia, Doliclio/is, Hvdio- clwerus "Eriomyini" = Chinchillini "Eriomves" " Erionivs" =Cliiiic/iil/a, LagidiiDii Lagostomi Scotaeumxs, Peiimvs (both fossil), Lagostomus. Octodontini Echinomyes Cercomys ( = what?, not Cercomys as here understood), Dacty- lomvs, Thrinacodus, " Lasitiiomys" — Isotlirix, " Loiicheres" = Echimxs, " Echinomxs" = Proecliimxs, " Nelomvs" =Cercomys, Mesowys (based evidently on Clyomys and Eioyzygomatomys); Carterodvn Octodontes Acareinxs, Sciainys (both fossil); Dicolpviiivs (fossil); " Hiihro- coina" = Abrocoiiiti, " Sclihodon" = Acontiemys, Spalncopus, Octo- don, Ctenomys 8. SCIURIDAE Castorini Eutxpomys, Steneofiher, Euliapsis (all fossil), Castor; Trogon- therium, Ccistoioides (fossil) Sciurini Sciuri Tamias, Otospennopliilus, Sciurus, Pteromys, Eiipctminis, Xeriis " Arctomyes" " Arctomys" — Maniiota, " Spennoplulus" =Citellus, Cynomys 9. " Saccomyidae" = Heteromyidae Gymnoptychini GymnoplychiiS { = Adjidniimo? fossil) PREVIOUS CLASSIFICATIONS: WINGE 21 "Saccomyini" Heliscomys (fossil), " Saccomys" = Heterotnys, Perognathus, Dipo- domys Geomyini Pleurolichus, Entoptychus (both fossil), Thomomys, Geomys. "I Form af Stamtrae" (phylogenetic tree) Myoxidae Mujidae Saccopyidae Dipodidae Hystricidae Sciiirjdae Anomaluridae Haplod Dntidae Leporidae The theory behind this work is brilliant, but the detailed characters are incorrect many times, and the nomenclature is deplorable. The classification of the "Hystricidae" seems in particular unnatural, with the inclusion of the Bathyergidae, the lumping together of Hystricidae and Erethizontidae, the whole family based more or less on the formation of the paroccipital process which in the British Museum material seems at any rate in Dactylomys and Capromys a variable character even within a genus. Further, if I understand the work rightly, the infraorbital foramen is supposed to transmit muscle (or to have transmitted muscle, which cannot be proved), in many groups in which it does not do so (Sciuridae, Castoridae, Geomyidae, Heteromyidae, Bathyer- gidae). Pedeta is regarded as with "fibula free," which is not so; M.i is re- garded as larger than M.2 in Cricetini and Murini, which is not constant {Anisomys agreeing with Winge's Rhizomyini in this character, as do some Neotropical Cricetines), etc., etc. OUTLINE OF CLASSIFICATION HERE ADOPTED The present classification is a combination as tar as possible of points which appear to he correct from the five classifications Hstcd above. It is based on characters which are constant throughout all the various genera referred to each familv, and has been the subject of great care and observation. The Order is here divided into twelve superfamilies. In case it may be thought that too man)' of these are retained, and that forms which share charac- ters have been too widely separated, I would quote a passage from Wood, who in his review of fossil and recent Heteromyidae, states, "The most important point to be emphasized is that ' parallelism, parallelism, more parallelism and still more parallelism' is the evolutionary motto of the Rodents in general." I have already stated that the Order is dominated to such a degree by one family, the Aluridae, that it is necessary to devote a separate volume to that family alone. It is necessary, therefore, for me to place this group last on the list (certainly not because it is considered the most highly specialized!); I, therefore, take first the group called Hystricognathi hv Tullberg, in which the mandible is more highly specialized, and last the Sciurognathi of Tullberg, which includes the Muridae, and in which the mandible is comparatively unmodified. With regard to the structure of the tibia and fibula, the following skeletons have been examined : "Fibula reduced, fully fused with the tibia in lower portion" Pedetidae : Pedetcs Bathyergidae: Batln'i'igiis, Georychus Geomyidae : Geomys Spalacidae : Spala.x Dipodidae: Alactagulus, Allactaga, Jaculus, Sicista Lophiomyidae : Lophiomys Aiuscardinidae: Glis, Eliomys, Muscaidiiiiis Muridae. (As this group is not dealt with in the present volume, less attention has been paid to them, as they are clearly separable from other families without this character. However, it has been checked in a very small number: Mus and Apademiis representing Murinae; Mesociicetus and PJiodopus representing Cricetinae; Gcrbilliis and Pacliyurumys representing Gerbillinae; and Tachyoryctes and Brachyiiromys representing Tachyoryctinae.) All examined are perfectly clear in this character, with the exception of Glis, in which the bones are widely separate, and only fused at extreme base. No Heteromyidae are available for examination, but Dipodoinys as figured by Howell agrees with the other Rodents. OUTLINE OF CLASSIFICATION HERE ADOPTED 23 " Fibula not reduced, not fully fused in lower portion w ith the tibia " Ctenodactylidae : Clenodactylus, Pectinator Anomaluridae : Anomahtrus Aplodontiidae : Aploduntia Castoridae: Castor Sciuridae : Petaiirista, Citellus, Spermophilopsis, Sciurus, Xenis, Tamias and Marmota Echimyidae : Dactylomys, Octodon, Aconaemys, Ctenomys, Capromys, Sivocastor, Tlir\onom\s, Isothrix, Petromus Erethizontidae : Erethizon, Coendou, Chaetomys Dasyproctidae : Dasyprocta Cuniculidae : Cunicithis Hystricidae: Hystrix, Tlieciirus, Atherurits Caviidae: Caiia, Dolichotis, Hydrochoerus Chinchillidae: Chinchilla, Lagidium, Lagostomus No skeletons of Dinomys nor of the Idiurinae are available. In this group. Castor appears in the adult to have fusion suggested in these bones, though not complete, and the Hystricoid Myocastor seems precisely similar, another interesting case of parallel development in these two un- related genera. The fibula is becoming reduced in Ctenomys. But much the most interesting result obtained is that in the skeletons seen of the three genera of Chinchillidae, Chinchilla, Lagidium and Lagostomus, the fibula, though still not fully fused, is excessively reduced, slender and threadlike; the reduction reaching an extreme degree in Lagostomus, which one might have expected to be not the case. It must be noted that, with regard to the formation of the forefoot, the ex- treme reduction, almost or completely to disappearing point of the pollex, is such a usual feature that no attention has been paid in the classification to this character. HYSTRICOGNATHI Lower jaw highly specialized, either by distortion outwards of the angular process by masseter lateralis, or by a conspicuous ridge extending along outside of mandible below level of the toothrow for attachment of masseter medialis. Angular portion of mandible never pulled inwards (inflected), and lower incisor root never forming conspicuous knob beside condylar process. 1. Bathyergomorph Series Infraorbital foramen not or scarcely transmitting muscle. First Superfamily. BATHYERGOID.\E Mandible with angular portion strongly distorted outwards by specialized limb of masseter lateralis. Infraorbital foramen not or scarcely transmitting 24 OUTLINE OF CLASSIFICATION HERE ADOPTED muscle. Zygomatic plate narrow, completely beneath the infraorbital foramen, not broadened for attachment of masseter lateralis. Skull much modified for fossorial life. Jugal moderately long, or in Hetero- cephalus shortened, and Murine. Kullae without special peculiarities. Cheekteeth i, t, or in Heliophohius, at full dentition, ;'; (in this genus the cheekteeth are normally not all in place together). Cheekteeth rooted, but extremely hypsodont, near simplification in pattern. A tendency present for the upper incisors to extend behind the toothrow, at extreme development into the pterygoids. Fibula fully fused with the tibia below. Digits of hindfoot five. External form much modified for underground life. Malleus and incus fused (Tullberg). Radiale and intermedium separate (Tullberg; fused in all other members of the Order examined by him except Ctenodactylidae). Family i. BATHYERGIDAE With the characters of the Superfamily. Group Bathyergi Bathyergus, Heliophobius, Georychiis, Cryptomys. Group Heterocephali Heterocephahis. (A key to all generic groups here recognized will be given below, when dealing with the families.) 2. Hystricomorph Series Infraorbital foramen much enlarged for muscle transmission. Second Superfamily. HYSTRICOIDAE Mandible with angular portion distorted outwards by specialized limb of masseter lateralis. Infraorbital foramen very large, always transmitting muscle. Zygomatic plate remaining narrow, and beneath the infraorbital foramen, not broadened for attachment of masseter lateralis. Skull usually broad, lacking interorbital constriction in the majority. Paroccipital process usually well developed, prominent. Jugal normally not approaching lachrymal ; frequently with downwardly or upwardly directed process present. Bullae variable; in certain groups much inflated. Cheekteeth |, the premolars not reduced in size in the majority; the pattern flatcrowned, reduced heptamerous, laminate, or sometimes approaching complete simplification. Fibula not becoming fused with the tibia high on the leg, and usually, but not alwavs, not specially reduced. Digits of hindfoot 5, 4, or 3. Malleus and incus fused (Tullberg). OUTLINE OF CLASSIFICATION HERE ADOPTED 25 Family 2. ECHIMYIDAE Cheekteeth rooted or rootless; when rootless or strongly hypsodont the pattern not a series of transverse plates. Digits of hindfoot 5, excepting Thryonomyinae. Bullae well inflated, sometimes extremely so. Feet never abnormally specialized for arboreal life. Spiny covering present or absent, but when present, never to an extremely specialized degree. External form never modified for cursorial life. Paroccipital process long, either curved forwards under the bullae, or lengthened and standing apart from bullae. Zygoma very generally complex, with downwardly or upwardly directed processes present. Subfamily FXHI.MYINAE Cheekteeth not hypsodont, and not simplified in pattern; usually the pattern is reduced heptamerous. Bullae not abnormally inflated, except Clyomys. Paroccipital process curved forwards under the bullae. Externally Rat-like or slightly modified for arboreal life, sometimes with spiny covering developed. Ecliimys, Isothrix, Diplomys ; Proecliimys, Hoplomys, Cercomys, Euryzygma- tomys, Clyomys, Carterodoti, Mesomys, Lotichothrix. Subfamily CAPROMYINAE Like the Echimyinae, but cheekteeth evergrowing, the re-entrant folds well filled with cement, so that simplification of pattern is suggested. Upper molars with more than one external re-entrant fold. Paroccipital process usuallv but not always standing apart from bullae. External form arboreal or terrestrial; fur not developing spines. Procapromys (not seen), Capromys, Geocapromys. Subfamily PLAGIODONTINAE Cheekteeth evergrowing, the upper series with only one external re-entrant fold, which is unusually deep and placed obliquely; the single inner re-entrant fold also unusually deep, and running parallel to the outer fold. Paroccipital process considerably lengthened. Zygoma simple. Plagiodontia. Subfamily DACTYLOMYINAE Cheekteeth rooted, abnormally broadened, nearly prismatic in appearance, and evidently with pattern not changing much during life. Paroccipital process either standing apart from the bullae, or curved forwards under them. Habits (said to be) arboreal; a tendency present towards elongation of the two central digits of forefoot and hindfoot. Thrinacodus, Dactylomys, Kannabateomys. Subfamily MYOCASTORINAE Cheekteeth extremely hypsodont, but not evergrowing, reduced heptamerous in pattern, the pattern changing little during life. Bullae reminiscent of the 26 OUTLINE OF CLASSIFICATION HERE ADOPTED type found in Castoridac, though less specialized than in that family. Skull prominently ridged for attachment of muscles. External form considerably specialized for aquatic life. Paroccipital process much lengthened, the lateral process of paroccipital process enlarged. Mvocastur. Subfamily ABROCOMINAE Cheekteeth evergrowing, the upper series more or less simple, the lower series prismatic, and complex. Auditory bullae much inflated, the paroccipital process curving forwards under them. Part of lachrymal canal open on side of rostrum. Al'iocoma. Subfamily OCTODONTINAE Cheekteeth evergrowing, both upper and lower series nearly or completely simplified. Bullae and paroccipital process as in Abrocominae. No part of lachrymal canal open on side of rostrum. External form generalized or modified for underground life. Octoinvs, Acoiiaemys, Octodoii, Octodontomys ; Spahicopus ; Ctenoiiiys. Subfamily PETROMYINAE Cheekteeth strongly hypsodont, nearly complete simplification in pattern, but not evergrowing. Auditory bullae much inflated. External form without special peculiarities. Petromus. Subfamily TMRYONOMYINAE Digits of hindfoot reduced to four. Cheekteeth moderately hypsodont, rooted, the re-entrant folds well marked and surrounded by heavy enamel. External form heavy, terrestrial-fossorial. Auditory bullae of moderate size; paroccipital process lengthened, standing apart from the bullae. Incisors much thickened, the upper ones heavily three-grooved. Skull massive, ex- cessively prominently ridged for muscle attachment. Shoulder-blade abnormal (TuUberg). Tlirvoiionivs. Family 3. DINOMYIDAE Cheekteeth evergrowing (?) or excessively hypsodont, the pattern a series of transverse plates. External form heavy, terrestrial. Digits ot hindfoot four. (Incisors thick, bullae moderate, paroccipital process not lengthened, no part of lachrymal canal open on side of rostrum, and angular portion of mandible powerfully distorted outwards, compare Chinchillidae.) Dinoinys. Family 4. ERETHIZONTIDAE Externally more specialized than in the Echimyidae; feet becoming abnor- mally modified for arboreal life, the hallux in progressive forms being replaced with a broad movable pad (and skeleton of foot correspondingly much modified). OUTLINE OF CLASSIFICATION HERE ADOPTED 27 Tail muscular, prehensile in progressive genera. Fur conspicuously spinous, the spines not long, and not modified into thick circular quills. Bullae relatively large; paroccipital process not lengthened. Zygoma usually simple. Check- teeth typically with wide re-entrant folds (parallel — .\nomaluridae), or in C/iae- tornys with structure much as in complex-toothed Echimyinae. Subfamily CIIAETOMYINAE Orbit almost surrounded by greatly thickened jugal and short postorbital process. Cheekteeth with narrow re-entrant folds. Spiny covering weakly developed. (Feet at highest specialization.) Chaetomys. Subfamily ERETHIZONTINAE Orbit large; skull without postorbital process; jugal not specially thickened. Cheekteeth with very wide re-entrant folds. Spiny covering at maximum specialization. (Feet moderately to extremely specialized.) Erethizon, Echinoprocta, Coendou. Family 5. DASYPROCTIDAE External form highly modified for cursorial life; digits of hindfoot reduced to three ; clavicles suppressed. Part of lachrymal canal open on side of rostrum. Bullae relatively large; paroccipital process not specially lengthened. Cheek- teeth strongly hypsodont, but not evergrowing, the re-entrant folds narrow, isolating on crown surface with wear as narrow islands. Myoprocta, Dasyprocta. Family 6. HYSTRICIDAE External form heavy, terrestrial; digits of hindfoot five. Fur always con- spicuously spinous, in progressive species attaining specialization in this respect not known elsewhere in the Order. Tail always with modified bristles or quills present. Spines of bodv usually partly modified into thick circular quills. Skull in progressive species characterized by great inflation of nasal bones. Auditory bullae relatively small. No part of lachrymal canal open on side of rostrum. Paroccipital process not specially lengthened. Zygoma simple. Cheekteeth moderately to extremely hypsodont, but not evergrowing, their pattern paralleling that present in Dasyproctidae. Clavicles present but in- complete. Lungs abnormal (Tullberg). Centrale not free (Tullberg; this character unique in the Order so far as known except Cuniculidae). Group Atheruri Trichys ; Atlieruriis. Group Hystrices Thecurus, Ilystrix. 28 OUTLINE OF CLASSIFICATION HERE ADOPTED Family 7. CUNICULIDAE Skull extremelv modified bv growth of conspicuous bony cheekplate, a structure not known elsewhere in the Order. Cheekteeth strongly hypsodont, but not evergrowing, their pattern hke that present in Dasyproctidae and Hystricidae, but rather more complex. External form heavy, terrestrial. Digits of hindfoot five. Clavicles present, but incomplete. Paroccipital process considerably lengthened. Centrale not free (Tullberg; on this character see remarks under family Hystricidae). Cuniciihis.^ Family 8. CHINCHILLIDAE Cheekteeth evergrowing, the pattern a series of transverse plates. Lower jaw with angular portion rather weaklv distorted outwards, the jaw to a certain degree transitionar}' towards that of the Cavioidae. Digits of hindfoot four or three (probablv three functional only in all genera). Jugal tending to approach the lachrymal, or to come in contact with it. Some part of lachrymal canal open on side of rostrum. Incisors relatively thin. A tendency present towards extreme inflation of mastoids and bullae. Paroccipital process relatively short (Chinchilla, Lagidium), or considerably lengthened (Lagosioiiiiis). Fibula extremely slender, much reduced (skeletons of the three genera have been examined). Group Chinchillae Chinchilla, Lagidiuin. Group Lagostomi Lagostomus. Third Superfamily. CAVIOIDAE Essential characters as in Hystricoidae except : angular portion of mandible not distorted outwards by specialized limb of masseter lateralis; masseter medialis the chief agent in modifying form of lower jaw, the outer side of which has a long and conspicuous ridge extending below and beside toothrows for attachment of this muscle. Cheekteeth evergrowing, relatively simple, but with sharp folds present, the effect more or less prismatic. Malleus and incus fused (Tullberg). Family 9. CAVIIDAE With the characters of the Superfamily. Paroccipital process moderately to extremelv enlarged; bullae normally prominent. Cheekteeth strongly uni- laterally hypsodont. Digits of hindfoot reduced to three; external form ambu- latory or cursorial. Clavicles suppressed. Tibia and fibula as in normal Hvstricoidae. Part of lachrymal canal open on side of rostrum, except in Dolicliotis. ^ = "Coelogeiiys.'' OUTLINE OF CLASSIFICATION HERE ADOPTED 29 Subfamily CAVIINAE Paroccipital process not excessively lengthened. M.3, upper series, not enlarged. Pattern of cheekteeth relatively simpler. Palate shortened from before backwards. Group Caviae Cavia, Galea, Caviella ; Kerodon. Group Doiichotides Dolichotis. Subfamily HYDROCHOERINAE Paroccipital process excessively lengthened. M.3 upper series extremely enlarged and elongated. Cheekteeth with more complex pattern. Palate not shortened from before backwards. (Size largest in the Order.) Hydiochoerus. SCIUROGNATHI Lower jaw not highly specialized, never with the angular portion distorted outwards, and never with long conspicuous ridge extending below level of toothrow for attachment of masseter medialis. Angular portion of mandible may be stronglv pulled inwards (inflected). In some genera, the root of the lower incisor forms conspicuous knob beside the condylar process. 1. Sciuromorph Series Infraorbital foramen not or scarcely transmitting muscle. Fourth Superfamily. APLODONTOIDAE Infraorbital foramen not transmitting muscle. Masseter lateralis not ex- tending attachment on outer side (forepart) of zygomatic plate, which remains narrow and unspecialized, completely below the infraorbital foramen. Mandible with angular portion inflected to an abnormal degree. Skull fossorial in aspect. Jugal lengthened. Bullae with neck directed horizontally outwards, and region of braincase greatly widened. Cheekteeth ;, evergrowing, near complete simplification of pattern. Fibula not reduced, nor fused with the tibia high on the leg. Malleus and incus free (Tullberg). Family 10. APLODONTIIDAE With the characters of the Superfamily. Aplodontia. (This family is doubtfully referred to the Sciuromorph series, and is regarded as one of the most isolated and primitive living Rodents.) 30 OUTLINE OF CLASSIFICATION HERE ADOPTED Fifth Superfamily. SCIUROIDAE Infraorbital foramen not or scarcely transmitting muscle. Masseter lateralis superficialis with anterior head distinct from zygoma, and masseter lateralis extending its line of attachment on to zygomatic plate, which is to a greater or lesser degree broadened and tilted upwards, the muscle typically rising upwards on zygomatic plate to superior border of rostrum. Mandible with angular process sometimes strongly inflected (Cynomvs); usually with a tendency for this formation to be present. Skull with, in progressive species, well developed postorbital processes present (these processes always traceable). Jugal long, usually approaching or reaching the lachrymal. Bullae without special modifications, usually prominent. Cheekteeth I or 4, cuspidate, very rarely approaching simplification, in which cases (Larisciis, Rheithrosciurus) the original pattern may usually be traced; the pattern normally a series of transverse ridges and corner cusps (in the upper series), the lower series most often basin-shaped, with cusps at each corner. Cheekteeth rooted, brachyodont or hypsodont. Fibula not fused with the tibia. Digits of hindfoot five. External form modified for arboreal or terrestrial life; a flying-membrane may be present, attached to sides; tail always completely haired. Malleus and incus free (Tullberg). Family 11. SCIURIDAE With the characters of the Superfamily. Group Pteromyes Belomvs, Trogoptenis, Pteromyscus, Petaiirista, Aeromys, Pteiomys, Hyhpetes, Petinomxs, Eoglaucomys, Glaucomys, Petaurillus, lomys, Eupetaurus . Group Sciuri Mxosciurus, Xaniwsciunis, Saurillus ; Microsciurm, Syntheosciurus, Sciunis, Tamiasciurus, Callosciwus, Funambuliis, Dremomys, Ratiifu, Menetes, l.arlsciis, Glxphotcs, Rheithrosciurus, Rliinosciiirus, Hyosciurus, Heliosciurus, Paraxcrus, Funisciurus, Protoxcrus, Myrsilus, Epixerus ; Atlantoxerus, Xerus, Spermo- philopsis ; Sciurotamias, Tawias ; Citelliis, Marmota, Cynomys. Sixth Superfamily. CASTOROIDAE Zygomasseteric structure essentially as in Sciuroidae. Mandible without special peculiarities. Skull with no well-marked postorbital process; jugal lengthened, approach- ing lachrymal, and extremely broadened anteriorly. Bullae with neck directed outwards and upwards. Cheekteeth \, extremely hypsodont, but not evergrowing, the pattern OUTLINE OF CLASSIFICATION HERE ADOPTED 31 flatcrowned, reduced heptamerous, the enamel folds narrow, the pattern chan^infj little during life. I'ihula not fused with the tibia, but tending to become reduced, so that in adult life fusion is suggested. Digits of hindfoot five. Externally much specialized for aquatic life. Caudal vertebrae broadened; tail naked, much broadened and flattened (unique in the Order in structure and appearance). Malleus and incus free (Tullberg). Family 12. CASTORIDAE With the characters of the Superfamily. Castor. Seventh Superfamily. GEOMYOIDAE Zygomasseteric structure closely resembling that of the Sciuroidae; infra- orbital foramen extremely reduced, and rather more modified; mandible with angular portion somewhat reduced. Large externally-opening cheekpouches present. Skull highly fossorial (Geomyidae), or becoming specialized by ex- treme inflation of auditory bullae and braincase, and narrowing of rostrum ("saltatorial type") (progressive Heteromyidae). Jugal always short; sometimes the zygomatic arch may be complete without it; in Heteromyidae, the whole zygoma is threadlike. Cheekteeth }, hypsodont, usually near complete simplification of pattern, and often evergrowing. Fibula (so far as known) fully fused with the tibia, high on the leg. Digits of hindfoot five, or in saltatorial genera may be reduced to four. Malleus and incus free (Tullberg). Family 13. HETEROMYIDAE Mastoids in progressive genera greatly inflated. Zygoma extremely narrowed. Infraorbital canal "with orifice protected from muscle pressure bv counter- sinking in a vacuity which extends transversely through rostrum" (Miller & Gidley). External form Murine or modified for saltatorial life. Cheekteeth rooted except in Dipodomys, and as a rule less simplified than in Geomyidae. Subfamily HETEROMYIN.\E Cheekteeth preserving pattern for a longer time; auditory bullae not specially inflated; form IXIurine. Heteromys, Liom\s. Subfamily DIPODOMYINAE Cheekteeth losing their pattern earlier; auditory bullae and mastoids con- siderably to abnormally inflated; external form Murine or saltatorial. (In Dipotlomys the cheekteeth are rootless and simple.) Group Perognathi Perognathus ; Microdipodops . 32 OUTLINE OF CLASSIFICATION HERE ADOPTED Group Dipodomyes Dipodomys. (The arrangement of this family is based on the classification of Wood, >935-) Family 14. GEOMYIDAE Mastoids not inflated. Zygoma robust, the jugal extremely shortened. Infraorbital foramen with its "orifice protected from muscle pressure by countersinking in an oblique sulcus" (Miller & Gidley). External form and cranial characters highly modified for underground life. Cheekteeth in Hying genera always eyergrowing and always, excepting the premolar, completely simple. Thoiiiomys, Geomvs, Piippogeomys, Cratogeom\s, Platygeomys, Orthogeomys, Heterogeomys, Macrogeomvs, Zygogeomys. 2. Myomorph Series Infraorbital foramen always clearly enlarged for muscle transmission. Eighth Superfamily. ANOMALUROIDAE Infraorbital foramen much enlarged for muscle transmission; masseter lateralis not extending its line of attachment on to forepart of zygomatic plate, which remains completely beneath the infraorbital foramen, and is narrow. Mandible without special peculiarities. (It may be noted that in this and the next three superfamilies, Pedetoidae, Ctenodactvloidae, and Dipodoidae, the zygomasseteric structure as regards the arrangement of the forepart of the skull (infraorbital foramen and zygomatic plate) is yery similar to that of Hystricoidae and Cayioidae). Skull not abnormal in the typical subfamily; in the Idiurinae, much modified by abnormal enlargement of infraorbital foramen, much constricted palate, thickened incisors, etc. Jugal long. Bullae large, but not abnormally inflated. Cheekteeth |, rooted, not hypsodont, the pattern typically reduced heptamerous, the re-entrant folds wide. Fibula, so far as known, not fused with the tibia (no skeletons of Zenkerella and Idiiirus available for examination). Digits of hindfocrt fiye. External form suited to arboreal life; usually a flying-membrane present, attached to sides. Underside of the tail with scaly outgrowths near the body. Malleus and incus free (Tullberg). Family 15. ANOMALURIDAE With the characters of the Superfamily. Subfamily ANOMALURINAE Infraorbital foramen not greatly enlarged. Palate not much narrowed. Incisors moderate; toothrows not reduced. "Anterior point of masseteric OUTLINE OF CLASSIFICATION HERE ADOPTED 33 insertion on mandible beneath hinder part of M.i " (Miller & Gidley). (Flying- membrane present.) A noiiui/iinis, Anomalurops. Subfamily lUlLKINAE Infraorbital foramen extremely enlarged, the zygomatic plate projected forwards to a point nearly immediately behind the incisors. Palate much narrowed. "Anterior point of masseteric insertion on mandible in front of P.4" (Miller & Gidley). Incisors much thickened from before backwards. Toothrows strongly reduced. (Flying-membrane present or absent.) Zenkerella; Idiurus. Ninth Superfamily. PEDETOIDAE Zygomasseteric structure not essentially different from that of the Anoma- luroidae (infraorbital foramen extremely enlarged, the zygomatic plate projected forwards, much as in Idiurinae). Skull with extremely broad frontals; jugal much thickened, lengthened, ascending to lachrymal; mastoids extremely inflated. Cheekteeth J, evergrowing, and near complete simplification in pattern. Fibula reduced and fully fused with tibia high on the leg. Digits of hind- foot four. Externally much modified for bipedal saltatorial life. Metatarsal bones normal, not tending to fuse (compare specialized Dipodidae). Malleus and incus free (Tullberg). Family 16. PEDETIDAE With the characters of the Superfamilv. Pedetes. Tenth Superfamily. CTENODACTYLOIDAE Zygomasseteric structure not essentially different from that of the Anoma- luroidae in so far as it affects the general shape of the skull. Infraorbital foramen large but not abnormally so. Mandible with the angular portion drawn backwards to a degree, but not distorted outwards; coronoid process absent; a weak short ridge may be developed, reminiscent of the "medialis ridge" of Cavioidae, though much less developed than in that superfamily. Skull considerably modified, flattened, the jugal long, the zygoma in two portions, a horizontal and a vertical (parallel — Dipodinae). Auditory bullae and mastoids much inflated. Cheekteeth at full dentition i, the premolars lost in the adult, or J (Pectinator). The teeth evergrowing, and near complete simplification of pattern. Fibula not fused with the tibia. Digits of hindfoot reduced to four. E.xternal form without special peculiarities; tail fully haired; some of the digits with stiff bristle-hairs present (parallel — Petromus, Octodontinae, Chinchillidae). 3 — Lning Kodents — 1 34 OUTLINE OF CLASSIFICATION HERE ADOPTED Malleus and incus fused (Tullberg). Radiale and intermedium can be separate (fused in all other Rodents examined by Tullberg except Bathyergidae). Family 17. CTENODACTYLIDAE With the characters of the Superfamily. Pectinator, Ctenodactyhis, Massoiitiera, Felovia. Eleventh Superfamily. DIPODOIDAE Zygomasseteric structure not essentially different from that of Anomaluroi- dae, the infraorbital foramen always large, sometimes extremely so, the zygomatic plate always completely beneath it. In primitive genera the infraorbital foramen is conspicuously wider below than above (compare the primitive Muroidae, Graphiurus, Deomys). Mandible with angular portion weak, sometimes perforated; occasionally strongly inflected, and usually with this formation suggested. Skull in progressive species characterized by much broadened frontals, greatlv inflated mastoids and auditory bullae, and specialized zygoma. In primitive species, the skull more or less Murine in general aspect (except the infraorbital foramen). Jugal long, usually extending to the lachrymal. Cheekteeth ^ or I, the extra premolar when present much reduced in size; the teeth rooted, cuspidate, with re-entrant folds which are fairly well open as a rule; occasionallv becoming flatcrowned, in which case the re-entrant folds are narrow (specialized Zapodinae). Fibula much reduced, fused high on the leg with the tibia. Digits of hind- foot in primitive species five; or may become reduced in progressive forms to three functional, or three only. External form much modified for bipedal saltatorial life, excepting the genus Sicisla. In specialized genera, the three central metatarsal bones fuse to form a cannonbone. Malleus and incus free (Tullberg). Family 18. DIPODIDAE With the characters of the Superfamily. Subfamily SICTSTINAE External form not modified for saltatorial life. Zygoma simple. Auditory bullae not inflated. Metatarsal bones normal. Cheekteeth quadritubercular, cuspidate, not tending to become flatcrowned. Sicistci. Subfamily ZAPODINAE External form modified for saltatorial life (as in all remaining members of the family). Cheekteeth semi-hypsodont, not quadritubercular, tending in pro- gressive species to become flatcrowned, in which case the re-entrant folds are OUTLINE OF CLASSIFICATION HERE ADOPTED 35 narrow, and usually isolate on crown surface as islands. Zygoma simple, un- specialized; auditory bullae not inflated; metatarsal bones normal. Eozapus, Zapus, Napaeozapus. Subfamily CARDIOCRANIINAE Auditory bullae and mastoids abnormally inflated. Zygoma in two portions, a horizontal and a vertical, these portions connected by a curvature. Metatarsal bones normal, not fused. Cheekteeth (apparently) as in Dipodinae. Digits of hindfoot mav be reduced to three. Cardiocranius (not seen), Salpingotus. Subfamily EUCHOREUTINAE The three central metatarsal bones fused to form cannonbone. Auditory bullae much inflated. Zygoma unspecialized, simple, like that of Zapodinae. Cheekteeth cuspidate, with unusually high cusps. Rostrum much lengthened. Ear much enlarged. M.3 (evidently) vestigial. (Hindfoot with three functional digits.) Euchoreutes. Subfamily DIPODINAE Three central metatarsal bones fused to form a cannonbone. Auditorv bullae and mastoids considerably to excessively inflated. Zygoma in two portions, a horizontal and a vertical, these portions not connected by a curvature. Cheekteeth with moderate cusps; rostrum weak; ear less enlarged; M.3 not vestigial. (Hindfoot with three functional digits.) Group Allactagae Allactaga, Alactagulus, Pygeretmus. Group Dipodes Paradipiis ; Dipiis, Scirtnpoda, Ereniodipus (the last not seen), Jaculus. Twelfth Superfamily. MUROIDAE Zygomasseteric structure primitively {Graphiurus, Deomys), nearly as in Sicistinae, except that the infraorbital foramen is less enlarged, and is not conspicuouslv wider below than above. In two hundred other genera examined belonging to the group, the zygomatic plate is broadened and tilted upwards to a greater or lesser degree; masseter lateralis extends its line of attachment on to zygomatic plate, and masseter lateralis superficialis has its anterior head distinct, so far as known, from zvgoma (see figures in Tullberg). Infraorbital foramen alwavs transmitting muscle, never extremely enlarged, usually less so than in the preceding families of Myomorph Rodents; and sometimes (Rhizomyidae) becoming much reduced. Mandible with angular portion not distorted outwards; in Muscardinidae this portion is relatively weak, may show signs of inflection, and may be perforated. Skull usually with constricted frontals; auditory bullae in the majority not 36 OUTLINE OF CLASSIFICATION HERE ADOPTED much enlarged, but may become so (Gerbillinae); or may be much reduced, as in Phloeomys, Lophiomys, certain species of Rattiis, and others. Jugal typically considerably shortened; but long in Muscardinidae, Tachxorvctes, etc. Cheekteeth \ (Muscardininae, Graphiurinae), f (the greater part of the superfamily), 'i (genus DcsmoiiiUiscus), or H (Rhvnchomys, some Hydromyine genera); rooted except in Mxoipalax, Rhomhomys, and the majority of the Microtinae. Fibula so tar as known reduced, and fused high on the leg. Digits of hind- foot fiye \yith one exception (Mahuotlnix). External form as a rule small, generalized; sometimes highly modified for underground lite (Spala.x, to a lesser degree Mxospcila.x, Ellohitis, Pronietlieomxs, Notiomxs, etc.); sometimes highly specialized for aquatic life, cranially as well as externally (Ichtlixomxs and allies, Hydromys, Crossuinys); sometimes specialized for arboreal life, with fully opposable hallux {Chiropodomys, Chirumyscus, Hapalomys and others); in one case, Notomys, apparently fully specialized for bipedal saltatorial life. Spiny coyering may be moderately deyeloped (Acomxs, some species of Rattiis, Plalacanthomys). Tail typically naked, scaly; uniformly haired in most Muscardinidae, CraUromxs, Lophiomys, one species of .Xcotonia, most Ger- billinae, and others. Malleus and incus free (Tullberg). Caecum suppressed in Muscardinidae, so far as known, except in Txph/omys; becoming much reduced in IcJitlixoinxs (Thomas). Family 19. MUSCARDINIDAE Skull without special modification; jugal usually relatiyclv long; bullae large and to a degree inflated except in Platacanthomyinae. Cheekteeth \ or jj, primitiyelv with basin-shaped crowns and comer cusps, becoming flatcrowned in progressiye genera, in which case they become a series of relatiyely narrow transyerse ridges (which are always traceable, eyen in primitiye torms), the general dental effect strongly reminiscent of that of the Sciuridae. Mandible with angular portion sometimes inflected, and sometimes with perforation. Externally slightly specialized for arboreal life; tail fully haired except in Mxoniimiis (not seen) and Txphlo)nxs. Cardiac portion of stomach with horny layer absent (Tullberg). (This character not known regarding Platacantho- myinae; but present, so far as knovyn, in all other members of the superfamily.) Caecum suppressed, so far as known, excepting in Txphlomys. Subfamily GRAPHIURINAE Zygomatic plate remaining completely beneath infraorbital toramen; masseter lateralis not extending its line of attachment on forepart of zygomatic plate. (Cheekteeth |). Graphiiirus. OUTLINE OF CLASSIFICATION HERE ADOPTED 37 Subfamily MUSCARDININAE Zygomatic plate tilted strongly upwards; masseter lateralis extending line of attachment on to forepart of zygomatic plate. Cheekteeth |. Bullae large, normally. Cheekteeth when flat-crowned with the depressions (between the ridges) not tending to become isolated with wear. Palate without a series of foramina (or a single large pair) situated between the toothrows. Myomimiis (not seen), Eliomys, Dyiomys, GUridus, Glis ; Muscardinus. Subfamily PLATACANTHOMYINAE Like the Muscardininae except: premolars suppressed, cheekteeth (flat- crowned) with the depressions (between the ridges) tending to become isolated on crown surface with wear; bullae small; palate with a large pair of foramina, or a series of foramina, between the toothrows. (Zygomatic plate much narrowed, parallel-Hydromyinae). (A caecum is present in Tvphlomys.) Platacanthomys, Typhlomys. Family 20. LOPHIOMYIDAE Like the Muridae (below, no. 23), but more specialized; skull with temporal fossae roofed over by bony plates rising from jugal, frontal, and parietal, a structure not known elsewhere in the Order. Cheekteeth i; pattern as in cuspidate Cricetinae. External form modified for arboreal life. (Hallux partly opposable; bullae much reduced.) Lophiomys. Family 21. SPALACIDAE Like the Muridae (below, no. 23), but more specialized; skull and external form extremely modified for underground life, the eyes suppressed. Infraorbital foramen relatively large, and zygomatic plate nearly horizontal and below it (secondarily acquired ? masseter lateralis superficialis with anterior head dis- tinct, as in Muridae, as figured by Tullberg). Cheekteeth with re-entrant folds which isolate on crown surface in adult. Spalax. Family 22. RHIZOMYIDAE Like the Muridae (below), but zygomasseteric structure unusual; infra- orbital foramen extremely reduced, owing to the fact that masseter lateralis rises to an abnormallv high degree on zygomatic plate (which is \er\ strongly tilted upwards), extending its line of attachment beside the infraorbital foramen on its inner side. Externally and cranially more or less modified for fossorial life; cheekteeth with re-entrant folds isolating as islands on crown surface in adult. (Infraorbital foramen not V-shaped below.) Rhizomys, Cannomys. (End of First Volume) 38 OUTLINE OF CLASSIFICATION HERE ADOPTED Family 23. MURIDAE Infraorbital foramen typically specialized into a wider upper portion for muscle transmission and a narrower lower one for nerve transmission, its lower border very generally V-shaped (not nearly straight as it is in Rhizomyidae). Jugal normally strongly reduced (except Tachyoryctes and some genera from Madagascar). Cheekteeth laminate, cuspidate, heptamerous or prismatic, but never reminiscent of those of Sciuridae, i.e. never agreeing in pattern with those of Muscardinidae. External form various, but when subfossorial, eyes retained, and zygomatic plate not specially narrowed nor turned downwards (compare Spalacidae). Temporal fossae never roofed in by bony plates (com- pare Lophiomvidae). Masseter muscle, so far as known, not rising beside infraorbital foramen on its inner side (compare Rhizomyidae). The order in which the subfamilies are listed here is provisional. Only valid genera which have been actually examined are included in the present list. Subfamily DEOMYINAE Zygomatic plate remaining completely beneath the infraorbital foramen. (Pattern of cheekteeth as in Dendromyinae). Deoinxs. Subfamily MURINAE Zygomatic plate (as in all remaining subfamilies) broadened and tilted upwards to a greater or lesser degree. Cheekteeth laminate or cuspidate; when laminate, the laminae tightly pressed together; when cuspidate, the cusps of the upper molars arranged in three longitudinal rows. Group Kliuri Eliunts. Group Anisomyes Anisoiiixs. Group Alures Hapalomvs, Ponoriomvs, Leiionivs, CJiiriipadoinys. V(iiiloduntia in the genus Cynomys (Sciuridae); and in the genus Cardiocranius (Dipodidae; not seen). The mandible of the Ctenodactylidae, so often placed in the Hystricoid series, is abnormal, but not in the least like the Hystricoid type. The coronoid is suppressed; the angular process drawn backwards to a degree; and a faint medialis-ridge, reminiscent of that of the Caviidae may be traced. I can call to mind no special peculiarities with regard to the mandible of the vast number of genera and species I have examined in the Muridae. Rarely the coronoid is suppressed. In the Aluscardinidae, and in certain Dipodidae, the angular portion mav be perforated. In the former family sometimes traces of the pulling inwards of the angular, so highly developed in Aplodontia, will be seen. The mandible mav be noted as weak in the Heteromyidae. In some genera with the "non-Hystricine" mandible, the lower incisor extends so far backwards that it forms a conspicuous process between the condylar and angular processes; examples are Spalax, Geomyidae, Rhizomyidae, Sesokia, etc. This never occurs in genera with Hystricine type of mandible. Pedetes, sometimes placed in the Hystricoid series (as by Thomas), has certainly not a Hystricine type of mandible, with its reduced relatively small angular process. In mandible structure, therefore, Rodents divide very broadly speaking into three classes : Angular process lifted up and distorted outwards by specialized limb of masseter lateralis superficialis; Bathyergidae, Ilystricidae, Erethizontidae, Echimyidae, Dinomy- idae, Cuniculidae, Dasyproctidac, Chinchillidae. Angular process never lifted up as above described. Lower jaw deeply modified by conspicuous ridge extending below level of toothrows on outer side, for attachment of masseter medialis: Caviidae. Lower jaw without extreme modifications, except in certain cases by root of lower incisor; or by strong inflection of angular process (e.xtreme only in Aplodontiidae): .\plodontiidae, Sciuridae, Geomyidae, Heteromyidae, Castoridae, Dipodidae, Ctenodactylidae, .Anomaluridae, Pedetidae, Muscar- dinidae, Spalacidae, Lophiomyidae, Rhizomyidae, .Muridae. The infraorbital foramen is enlarged to transmit the masseter muscle in a very large number of Rodents. Degree of enlargement, and shape and size of this foramen varies exceedinglv. Even in those forms which arc regarded here as not transmitting muscle, in two families, Sciuridae and particularly Bathyergidae, is certain variation 44 ZYGOMASSETERIC STRUCTURE found. In Protoxeriis and Tamias (Sciuridae), the infraorbital foramen is more enlarged than in the other Squirrels, and probably may transmit a small strand of the muscle. In these cases however it is not so far as I can judge anything like so enlarged and clear as in any Rodent which is regarded here as a form with muscle transmission of this foramen present. In the Bathyergidae, certain species of the genus Crvptoiiivs appear to be starting to transmit muscle through the infraorbital foramen ; it may rarelv, as in C. mcUandi, even be as much en- larged as in the much reduced type found in the Rhizomyidae. This is evidently a variable character, and in Cryptomys the foramen may even be more enlarged on one side of the skull than on the other, in individual cases. In the Aplo- dontiidae, most authors state that the canal does not transmit muscle; however in those examined it is on the large side for this section of the Order. The infraorbital foramen does not transmit muscle in Geomyidae, Hetero- mvidae (excessively reduced in these two families), Castoridae, .Sciuridae with the above noted exceptions, Aplodontiidae and Bathyergidae, with the above noted exceptions. In all other Rodents it is clearly enlarged to do so. There are then broadly speaking two types of infraorbital foramen structure to be discussed, with the one exception of the Rhizomvidae. In Hystricidae, Erethizontidae, Echimyidae, Dinomyidae, Dasyproctidae, Chinchillidae, Caviidae, Pedetidae, Anomaluridae, Ctenodactylidae and Dipodidae, it is round, completely above the zygomatic plate, and normally extremely enlarged. This enlargement reaches its greatest development probably in the Pedetidae, and in the Idiurine subfamily of Anomaluridae; and in certain sections of the Dipodidae. In the Cuniculidae, the infraorbital foramen is secondarily reduced by the growth of the enormous cheekfilates. In two genera of Rodents which are here referred to the Muroid superfamilv, Deoiins (Aluridae), and Gniphiiinis (Muscardinidae), the infra- orbital foramen, though not abnormallv enlarged, is completelv above the zygomatic plate. In the remainder of the Order, the infraorbital foramen, though sometimes varying in actual size of enlargement, is never abnormally enlarged; in vast sections of the family Muridae, it is specialized into a wider portion above for muscle transmission, and a narrower lower one for nerve transmission. In the Subfamily Microtinae, it has become, correlated probably with the increase in general strength of jaw-muscles in this group, much reduced. It is abnormally reduced in the Rhizomyidae {K/ihoiiivs and Cannom\s)\ in this group, the zvgomatic plate is strongly broadened and tilted upwards, and the foramen becomes reduced to a small aperture situated at the top of this plate; the masseter muscle rises up inside of it, a condition according to Tullberg not known elsewhere in the Order. It mav be noted that Winge puts forward the theorv that in all Rodentia living, Aplodotitia excepted, the infraorbital foramen has transmitted muscle, and has become secondarily closed in the Geomyidae, Sciuridae, Castoridae, Bathyergidae, Heteromyidae. Without wishing to enter into a discussion on matters such as these, it appears to me to be singularly unlikely that, having taken such a large step forward in evolution as the enlargement of this canal ZYGOMASSETERIC STRUCTURE 45 for muscle transmission (as it seems an unusual character among Mammalia to say the least), these families should go even further in evolution and, so to speak, develop covering over this canal so that it does not transmit again. There is not a wide difference in the arrangement of the zygomatic plate between Aplodontiu and a primitive Sciurine such as Belomys; it would seem so much more likely that the Sciurine arrangement of jaw-muscles was developed from a type not widely distinct from Aplodontia as regards arrangement of infraorbital foramen and zygomatic plate ; far more likely than that the infraorbital of Belomys is secondarily closed to muscle transmission. In the Spalacidae (Spalax alone), the infraorbital foramen is larger than usual for a Muroid Rodent and the zygomatic plate, though to a degree broadened, appears to be nearly flattened to a horizontal position. This how- ever may well have been brought about by the fossorial habits of this animal. Summarizing: the infraorbital foramen does not, or scarcely transmits muscle in Sciuridae, Geomyidae, lleteromyidae, Castoridae, Bathyergidae, Aplodontiidae. It is enlarged, and usuallv very much enlarged for muscle transmission in Hystricidae, Erethizontidae, Echimyidae, Dinomyidae, Cuniculidae (see note above), Dasyproctidae, Chinchillidae, Caviidae, Ctenodacty- lidae, Anomaluridae, Pedetidae, Dipodidae. It is enlarged, but very rarely much enlarged for muscle transmission in Muridae, Lophiomyidae, Spalacidae, Muscardinidae, and Rhizomyidae (see note above). The zygomatic plate is less variable in structure, broadlv speaking, than either the infraorbital foramen or the mandible. Among the Rodents it is found in two conditions only. It is narrow, usually very narrow, and strictly hori- zontal, remaining completely beneath the infraorbital foramen, in Aplo- dontiidae, Bathyergidae, Dipodidae, Anomaluridae, Ctenodactvlidae, Pedetidae, Hystricidae, Erethizontidae, Echimyidae, Dinomvidae, Dasyproctidae, Chin- chillidae, Caviidae, and in the genus Graphiurus (Muscardinidae), and Deomys (Muridae). In the Cuniculidae it is much distorted bv the growth of the bony cheek- plates. In other Rodentia, to a greater or lesser degree, it is broadened and tilted upwards. In these, according to Miller & Gidley, and supported by Tullberg's figures, masseter lateralis superficialis is distinct from the zygoma, "not attached to any part of zygoma except occasionally to a point at extreme base of zygo- matic plate." In the Sciuridae, Castoridae, Geomyidae and Heteromyidae, in which the infraorbital foramen does not transmit muscle, the zygomatic plate is very generally strongly broadened and tilted upwards, the only exceptions being found among the Sciuridae; such as Tamias, and most members of the Pteromvs group except Pteromys. In these families, the lateralis muscle rises to the superior border of rostrum and excludes masseter medialis from so doing. In the Muridae, so far as known, except Deomvs, the Muscardinidae, 46 ZYGOMASSETERIC STRUCTURE except Grapliiurus, the Lophionividae, the Spalacidae, and the Rhizomyidae, the zygomatic plate is broadened and tilted upwards to a certain degree, hut masseter mcdialis is transmitted through the infraorbital foramen so that it is not excluded from the superior border of the rostrum, and masseter lateralis as a rule does not extend so high on the forepart of the skull. The zygomatic plate in these families only approaches the Sciurine type of specialization as regards broadening in the Rhizomyidae. The degree of broadening, narrowing, and tilting upwards varies extremely through the Muridae, as might be expected in such a vast group. In Hydromyinae, though tilted up strongly, it is narrow. In such genera as Oxymycterus, and Lupjiuromvs, it is very little tilted upwards; but only in Deomys of the vast number examined does it appear to me to be absolutely indistinguishable from the Dipodoid type as defined by Miller & Gidley. Notwithstanding this, although Tullberg's figures show clearly that there is a wide distinction between Glis and Grapliiurus in the iVIuscardinidae, and between Deomys and Oxymycterus in the JMuridae, as regards zvgomasseteric structure of the forepart of the skull, I am not persuaded of the desirability of transferring Grapliiurus to a separate superfamily from Glis, as was done by Miller & Gidley, nor Deumys to a separate superfamily from the remainder of the Muridae, although it must be admitted that to identify the superfamily relationships of Deomys (here considered a Muroid), from Sicista (a primitive Dipodoid), is not possible on this character alone. It would seem however that the close resem- blance in all other main characters between Grapliiurus and sav Eliomvs, and between Deomys and sav Deiidromus indicate that the Murine type of zygomatic plate and arrangement of jaw-muscles has been derived from the Dipodoid type, or vice- versa. The zygomatic plate therefore in the Order is narrow, and completely beneath infraorbital foramen, showing no signs of becoming broadened and tilted upwards, in Hystricidae, Erethizontidae, Fxhimyidae, Dinomvidae, Dasvproctidae, Caviidae, Pedetidae, Ctenodactylidae, Anomaluridae, Dipodidae, Aplodontiidae, Hathvergidae, Chinchillidae, Muscardinidae, part, subfamily Graphiurinae, and Muridae, part, sub- family Deomyinae only. It is much modified by growth of cheekplate in Cuniculidae, but presumably possessed the above character originally as in the rest ot the Ilvstri- coidae. It is broadened and tilted upwards to a greater or lesser degree in Sciuridae, Castoridae, Geomyidae, Heteromyidae, Lophiomyidae, Rhizomyidae, Spalacidae (see note above, p. 45), Muscardinidae. part, except Graphiurinae, and Muridae, part, all except Deomvinae. The presumed relationship between Deomys and the Dendromyinae, and between Grapliiurus and the remainder of the Muscardinidae indicate that it is not wise to base superfamilv grouping on zygomasseteric structure alone, as was done by Miller & Gidlev. DISTRIBUTION The Rodentia is the only Order of non-Marsupial land mammals inhabiting Australia. The one family of Rodents, the Muridae, must have either got there early from South-east Asia, which is the view currently held, or evolved there, which is the view suggested in the present paper. It is curious that if the Muridae alone came from South-east Asia, some members at least of the families Tupaiidae, Soricidae, Erinaceidae, Galeopteridae, Viverridae, Musteli- dae, Tarsiidac, Cercopithecidae, Tragulidae, Cervidae, Sciuridae, Hystricidae and Manidae, to quote only some of the families widely or at least comfortably distributed throughout the Indo-Malayan islands to the north-west of New Guinea, did not do so. It is remarkable to say the least if not one genus of this vast assemblage entered the Australasian region, and yet such a large number of Muridae did so. For in New Guinea and Australia, and immediately adjacent islands such as Ceram, there are two hundred and forty-five named forms of Muridae, belonging to twenty--five genera and two subfamilies. It is to my mind as likely that a large section of the Muridae evolved in Australia and came into Asia via .some islands as Celebes and the Philippines which may have for a time been separated from Asia and joined Australia, but later separated from Australia and joined Asia, than that all these Australian types came from Asia unaccompanied bv anv other genus of non-murine mammal. (The presence of the genus Sus in New Guinea is usually held to be due to introduction.) 'I'his view suggests that the Muridae are among the most archaic of mammals, which appears on account of their universal distribution to be likely. The main Australasian genera of Aluridae are Rattus (many species totallv distinct from the "ship-rats" rattus and norvegicus, and including one group, concolor which ranges to some of the Pacific Islands), Uromys (doubtfully distinguishable from Rattus), and the isolated and distinct genera Zysornvs, Mesembriomvs, Notomvs, Conilunis, Leporillus, and .Mastacomvs (Australian or Tasmanian), and Mallotnys, Hvomys, Pogonnmys, Macruromys, Xesoromys and Anisoinys (New Guinea or Ceram). Leggadina (Australian), appears to represent a wild ally of the cosmopolitan genus .l/iw, which is I think not indigenous to the area under consideration. .-Ml these belong to the subfamily Murinae; the subfamily Hvdromvinae, which is probably derived from Murinae, and closely allied to it, has half a dozen representatives in the area, as Hydromys (Australia and New Guinea), and the more restricted Xeromys (Queensland), Crossomys, Parahydromys, Leptomys, etc. (New Guinea), most of which are little known and rare. The Indo-Malayan region presents few families, but great reduplication of species within the larger genera. Only Sciuridae, Hystricidae, Rhizomyidae, Muscardinidae (Malabar and South China), and Muridae have penetrated the area, and only Sciuridae, Hystricidae and Muridae to any great extent. Roughly twelve hundred forms are named from the area, about half of which arc Muridae. 47 48 DISTRIBUTION In this family, of the typical subfamily, the genus Rat/us (largest genus in named forms in the Order), has its headquarters in the present region, with oyer three hundred and fifty named forms ranging over the whole area, and containing in the area about twenty specific groups, eight of which range through the greater part of the area except Peninsular India, and in some cases Celebes, two of \vhich are peculiar to Peninsular India, and several of which are confined either to Celebes or the Philippines. The genera Bandicota and Cliiropodomvs, and to a lesser degree Mus, range through most of the area except that Chiropo- domys is not known from Peninsular India nor Celebes, and Bandicota does not range further east than Java. Apart from these the Murine genera of the Malay Islands are rather different from those of the mainland. In the Philippines, highly specialized genera such as Phloeomvs, Crateromvs, Carpomvs and Cru- tiomys occur, and are not known outside the islands; they may be allied to certain Australasian types, as the New Guinea Mal/omvs, etc. The highly aberrant genus Echiothrix is confined to Celebes. A tew other rather unim- portant genera are named from Sumatra, Java, Borneo, closely allied either to Rattus or Mus, excepting Pithecheiy (Sumatra, Malacca). In the eastern portion of the mainland which constitutes this area (Burma, Siam, Indo-China region), a few genera as the isolated Hapalomys, and types such as Dacnomvs and Hadromvs are confined. Vaiide/eiina, wholly Indo- malayan, ranges into the area from Peninsular India. In South China, the Palaearctic genera Miiioiiivs and Apudcmtis occur. In this area, and Siam, the genus Alus appears to end its natural Eastern Range (except perhaps for its presence in the Philippines). Peninsular India appears to have types rather different from those of the eastern Indomalayan ; among these may be mentioned Goliiirda and Mil/ardia, which range more or less through the area, and in the north occurs a species of Acomys (African and Palaearctic chiefly), and Nesokia. A distinctly Australasian element is seen in the Philippines in the presence of the Hydromvinae (Chrotvmvs, Celaenomys). From the same island comes R/ivncIioinvs, which is here regarded as type of a subfamily the Rhynchomyinae. The Muridae of the Malay Islands, other than the Philippines, all belong to the typical subfamily. In the northern part of the mainland area, a few INIicro- tinae, as Eothenomvs (Southern China, Burma), Neoduii (Sikkim), and some others have their southernmost range limit in the Old World. In Peninsular India, the Gerbillinae are represented by Tatera which occurs throughout the area; the subfamily is not known from the remainder of the Indomalayan. The family Rhizomvidae {Rliisoinvs, Cainiomvs) is more or less confined to the area, ranging out of it only in parts of Szechuan, the group extends through South China and from Nepal south through Siam to Malacca and Sumatra. The Muscardinidae is represented by two rare genera which form a well- marked subfamily (Platacanthomyinae), and are confined to the Malabar coast of India (Platacantliomxs), and to South China (7\'phluiiivs). The Hystricidae very probably evolved in the present area since all the most primitive known types seem to be grouped in it. Two, Trichvs and I'ht'curus, are confined to the islands (Sumatra, Borneo, and in the case of Thectirtis, the Philippines) (Trichvs DISTRIBUTION 49 reaches Malacca). The more widelv ranging genera Atherurus and Hvstrix occur throughout much of the area; Hvstrix seems absent only from the Philip- pines and Celebes; Atherurus ranges from Sumatra at least, north to Assam and South China. The species of Hystrix in the area are with the exception of the form found in Peninsular India, which also ranges over much of Palae- arctic South-west Asia, of the more primitive type, at any rate as regards development of external covering of quills and spines. The Sciuridae present a great number of forms in the area, and a high degree of specialization. In Peninsular India, only two genera of non-flying squirrels occur, Funambulus (confined to the area), and Ratufa, which ranges over the whole region east to Borneo, but evidently not much in .South China, though known from the island of Hainan. In Nepal and Burma, many more genera occur; Callosciurus (not very clearly distinguishable from the Holarctic genus Sciurus), heading the list with about three hundred named forms. Mar- mota ranges into the area from the Palaearctic, in Nepal and Yunnan. Dremomys and Menetes may be mentioned as types typical of the eastern part of the main- land; the former ranges to Malacca and Borneo. When Malacca is reached, many new forms start their ranges, some of which are highly specialized. Lariscus and Rliiuosciwus, both ranging to Borneo, are among the more im- portant. The Pygmy Squirrels of the genus \' amiosciurus go through the whole of the larger Malay Islands, from Sumatra to the Philippines, except Celebes, where they are represented evidently by a species, murinus, which agrees more in characters with the allied genus Sciurillus. Other peculiar types are Rheithro- sciurus and Glyphutes, both of Borneo. None of the above-mentioned are known outside the Indomalayan region, except that Callosciurus has a few forms ranging into Palaearctic China. And it seems that the further south one goes the more highly specialized or aberrant become the distinct genera, though the more normal types as Callosciurus and Ratufa go through much of the area, the former even including Celebes and the Philippines. This is one of the few regions in the world where the named forms of Sciuridae actually exceeds the number of named forms of Murinae, for in addition to the above-mentioned, the area seems to be the headquarters of the Flying-squirrels; Petaurista, the giant Flying-squirrels, and smaller forms as Hylopetes and Petinomys, range more or less throughout the whole area except that Petaurista does not occur east of Borneo, Hylopetes does not enter Peninsular India, and neither Hylopetes nor Petinomvs appear to go verv far into South China. Belomys is an important genus confined to the north-eastern part of the area (Sikkim, Tongking, For- mosa, etc.). To the Malay Islands, some very distinct generic types are restricted, the most noteworthy being lomys. It will be seen that, as indicated above, only three families of Rodents have gained a real footing in this region, which is a very different state from that present in most of the other large areas of the World. Palaearctic Rodents. The Palaearctic as here understood contains all land in the Old World lying north of the Yangtsekiang River, the 30" line of latitude through northern India (i.e. including Kashmir), and broadly speaking 4 — Living Kodents — I so DISTRIBUTION the remainder of South-west Asia and the coastal regions of Africa which he north of this hne, or just south of it (as South Persia). Arabia should probably be regarded as forming part of the African rather than the Palaearctic region. From this area roughly eleven hundred forms are named. The Muridae is verv much the dominant family here in that seven hundred and fifty forms approximately belong to it. Six other families have a wide or moderate range in the area, and two, Ctenodactylidae (coastal regions of North Africa, Ctefio- dactvliis, Massoutiera) and Rhizomyidae (Szechuan, Rliizoniys), just touch it. The Microtinae is here the dominant suhfamilv of Muridae; the genus Microtus, which occurs almost throughout the whole region except most of North Africa, being the sole Rodent genus with more than a hundred forms named from the area alone. Other Microtine genera with a wide range are Clethrionomys, and Arvicola, the former like Microtus extending across into North America. Ellobim and Prometheomys, the two subfossorial \'oles, are restricted to the area. The Lemmings, Dicrostonyx (Arctic regions), Lemimis and Myopiis range across the northern portion. The two former also cross into North America. I'he most interesting of the rather numerous remaining genera in the area are Lagiinis, Alticola, Do/oiiiys, Pitvmys and Blanfordimys. Pitviins has a wide range in Continental Europe, but is not met with further east of the Caucasus until it turns up again in Eastern North America, though several forms as Neodon occurring to the East in the Palaearctic are closely allied to it. Lagurus is also known from America. Dolomvs and Blanfordimvs are rare and local (Montenegro region, and Afghanistan region respectively). The Murinae are well represented, but by only a very few genera, at any rate compared with the huge numbers of genera to be found in any of the tropical portions of the Old World. Indeed, only five have a real range in the area. Apodemus is probably naturally the most widely distributed, as well as one of the most primitive members of the group, and appears to extend its range even to Iceland. Mus and Rattiis are now cosmopolitan in the Palaearctic owing to artificial human distribution, but both probably have a naturally wide range in the area, especially the former. Micromys ranges intermittently from England to Japan ; and Xesokia is common in the more southern portions of the .Asiatic part of the area (Syria (into Egypt), across Persia and Russian Turkestan to Kashmir and Sinkiang). Of the rest, some four genera touch the coastal part of Africa, one of which, Acomvs, is known from Crete and Syria, and three genera range north from India into the Kashmir region, the most important of which is Goliinda. The Subfamily Gerbillinae has a wide range in the Palaearctic east of western Europe; Meriones is the main genus, having more named forms from the area than any other Palaearctic Rodent except Microtus and Apodemus; Rlwmbomxs, the most highlv specialized member of the subfamily, is from the Palaearctic only; but apart from these no member of the group ranges as far north as Siberia, being mostly confined to the Syrian-Persian region (as Tatera), or North Africa (Gerbillus. Psammomys, etc., both of which range into Syria). The peculiar " Fat-tailed (ierbils" of the genus Pnchyuromys seem more or less restricted to the Palaearctic portion of Africa. The subfamily Cricetinae has a fairly wide range in the area, though only DISTRIBUTION 5> five genera are met with, the group being primarily American ; Calomyscus, surely a very near relative of the American Perumyscus, is restricted to Persia, Russian Turkestan, and Baluchistan; the more typical Hamsters, which seem not to have very near allies in North America, have a wider range; Cricetus occurring from Central Europe as far west as Belgium, east to Central Siberia; Cricetiihis covering a very large part of China, as well as Greece, South Russia, Syria, S.W. Siberia, and Kashmir; the other genera occurring in the area being Mcsocricetiis and Phodopus. In addition to the four great subfamilies of Muridae being well represented as indicated above, there is a very interesting subfamily confined apparently to Palaearctic China and adjacent parts of Siberia only, the Myospalacinae, with one genus, Myospalax. The family Spalacidae, which is here restricted to the genus Spalax alone, is purely Palaearctic, ranging round the eastern end of the Mediterranean Sea from Hungary and the Balkan States to Egypt, and occurring in South Russia. The family Muscardinidae, represented by the typical subfamilv, is more or less western in general dis- tribution, though represented in Japan. The four better known genera, Eliomys, Dyromvs, Glis and Muscardinus all appear to meet in Central Europe, so far as their range is concerned. Dvi'omys goes east to Tianshan and Zungaria, but not west of Central Europe; Glis ranges to Spain and the Atlantic, also east to the Caucasus and Turkestan; E/iomvs does not range east of European Russia, but occurs again in Sinai and North Africa, as well as the Iberian Peninsula; Muscardinus is not known from Spain nor east of European Russia, but ranges naturally in England and in Scandinavia, which none of the others reach except by introduction. The family Dipodidae has its headquarters in the Palaearctic. Of the more primitive groups, the Sicistinae (Sicista) has the widest range, occurring from Scandinavia, the Balkans, and Hungary, more or less across the area evidently, in suitable localities. The Chinese Eozapus represents the American subfamily Zapodinae (the only subfamily occurring in that con- tinent). The Cardiocraniinae, containing two extremely rare types, Cardio- cranius and Salpingoius, appears to be restricted to the more inaccessible parts of Chinese Central Asia, except that a species of Salpingotus is known from Afghanistan. Of the more specialized groups, the Euchoreutinae {Euc/ioreutes) is restricted to the deserts of Inner China; the Dipodinae have, however, a wide range outside Western Europe. Alhictaga and Dipus both appear to range from South Russia across much of the Asiatic portion of the area, east more or less to the North Chinese Pacific coast; Jaculus ranges across North Africa from Morocco to Egypt and east as far as Persia; and generic types worthy ot note confined to the Palaearctic with more restricted ranges are Scirtopoda, Paradipus, and Pygeretmus. The Sciuridae have, as usual, a wide distribution in the area; only in contrast to the normal element (arboreal) in tropical areas, most of the Palaearctic genera are Ground-squirrels. Citellus and Marnwta have the widest ranges, both occurring in Europe as well as much of Asia, and both occurring again in North America. Tamias, principally American, ranges in North Russia, Siberia and China. Atlantoxerus, confined to Morocco and adjacent region, represents a somewhat different type of Ground-squirrel tound chiefly in 52 DISTRIBUTION Africa, and evidently not represented in either America or the Indo-Malayan; Spermuphilopsis from Russian Turkestan area is probably a distant ally. Another type of semi-terrestrial Squirrel is the Chinese Scitirotamitis, which seems nearest to Ttimias in relationships. Tree-squirrels are represented by Sciuriis, which occurs throughout Europe, across Russia and Northern (wooded) Siberia, and parts of Eastern China, as well as in the Caucasus, but is absent from North Africa, and much of the plains regions of S.W. Siberia. The Indomalayan Callosciurus sends a few forms north into China. Of the Flying-squirrels, Pteromys has the widest range, from Scandinavia across U.S.S.R. to Japan and N.E. China. Contrary to Thomas's classification of Flying-squirrels, the genus is here held to be an isolated specialized type with no very near allies, not a near ally either of the American Glaucomys or the Indomalayan Hv/opefes. Petaurista, from the Indomalayan, has a wide range in China, and includes Japan and the Kashmir region, but is not known west of Kashmir nor in any part of Siberia. Eupetaiirus is confined to Kashmir, and Trogoptcnis to parts of China (though this genus touches the Indomalayan in some parts of China south of the Yangtse). Eog/iiiicoinvs from Afghanistan completes the Palaearctic list of Sciuridae. Two other families occur in the area, the Castoridae (with one genus. Castor), now restricted to various localities in parts of Europe, such as Scandinavia, and some of the larger Central European rivers, parts of Russia, and the Mongolian Altai (the genus ranging across to North America), and the Hystricidae, yvith one genus, Hystiix, specialized species of which extend from India through Persia and Turkestan to Syria, and again in Italy, Sicily (where they might have been originally introduced), and North-western Africa. This section of the genus, however, finds its widest distribution in Africa south of the Sahara. The Nearctic regkix (Canada and the United States) contains roughly eleven hundred named forms of Rodents distributed among eight families. The Muridae are here in the minority as compared with all the others, only four hundred forms appro.ximateiy belonging to them, while about six hundred and ninety are named for the other families. This contrasts widely with the con- dition found in the Palaearctic. Only two subfamilies of Muridae reach America naturally at all (apart from the House-rats, and House-mice (Murinae), Raitiis and 71///.?, which were originally introduced accidentally by man). These two subfamilies are the Cricetinae and Microtinae. The Microtinae contain types mostly much like those of the Palaearctic, as for example Microtus, Clethrion- omys (ranging over much of the area), Dicrostonyx, Lemiiiiis (northern and Arctic), Laguriis and Pitynivs (with more restricted ranges), (ienera peculiar to the area are Ondatra (the largest member of the subfamily), Syiiaptoinxs (a Lemming), Phenacomvs (one of the most primitive known \'oles), and A'eofiber (confined to Florida). The Cricetinae of North America appear to have " come up from the south " rather than "in from the west," in that they are apparently more nearly related to South American types rather than to Palaearctic Cricetinae. Of the seven genera known north of Mexico, three only reach as far north as Canada, Peromys- cus (which appears to cover the entire continent), Ncotonui, and Onychomys. DISTRIBUTION S3 The genera Reithrodontomys, Sigmodon, and Orysomys have a fairly wide range in the warmer parts of the United States, and all the above without exception continue their ranges south into Mexico, all but Onychomys into Central America, and the three last-named range into South America. Outside the Muridae, three families occur which also range in the Palaearctic, and four are at the present day confined to the New World. The Sciuridac are represented, as in the Palaearctic, mostly by Ground- squirrels, of whicli Marmota, Cite/Ins, and Tamias (all also in the Palaearctic), each cover a great portion or most of the area. Cynomys, a rather isolated type of Ground-squirrel, is purely North American; Tree-squirrels are represented by Tamimciurus (American only), and the more widely ranging genus Sciurus, which, however, covers relatively little of the area. Flying-squirrels are represented only by Glaiicomys, which has, however, a very wide range. The Castoridae are represented by the sole genus. Castor, which is Holarctic in distribution. The Dipodidae are represented only by the primitive subfamily Zapodinae {Xapaeozapiis, Zapus), which, however, covers almost the whole area but seems not to occur south of it. We may now turn to four solely American families. The Aplodontiidae, with one genus, Aplodontia, represents an archaic type of Rodent confined at present to the western side of the Rocky Mountains, but known to have occurred formerly in East Asia. The Erethizon- tidae, a member of the Hystricoid branch, is represented by Erethison, note- worthy as being the only Hystricoid adapted for a life in cold climates. Other members of the family occur in the warmer portions of the Neotropical, from Mexico southwards. The Geomyoid branch of the Order contains two families both confined at present to America, extending south to Panama and Ecuador though chiefiy northern in distribution. Of these the Heteromyidae is widely distributed in the western portions of the United States, represented by the more primitive Perognathiis, which touches western Canada, and the highly specialized saltatorial Dipodomys and Microdipodops. The Geomvidae are repre- sented by Thomomys, with very many named forms from the western and central U.S.A., and ranging into western Canada, and Geomys, from the central and eastern United States, including Florida. So far as I have traced it this branch of the Order is not known fossil outside America. The NEOTROPICAL ."^REA has more named forms than any other of the great areas, if Mexico, the West Indies, and Central America are included in it. Notwithstanding this, although South America is currently referred to as being the " headquarters" of the Rodents, once Panama is passed there is a surprising general similarity of type through the various groups found in that continent, the members being cither Cricetine Muridae, or Sciuridae, or Caviidae, or members of the supcrfamilv llvstricoidae (all of which are more or less closely allied to each other). (A genus of Heteromyidae occurs in the extreme north.) There is a lack of that wide divergence of structure which makes the study of Palae- arctic, Nearctic, or African Rodents so interesting, and recalls the state of affairs present in the Indcmalayan. In fact, so far as the "superfamilies" recognized in this work are reckoned, fewer of them occur in South America than in any other of the great areas except the Indo-Malayan. 54 DISTRIBUTION Even in the Muridae there is incessant repetition of a single (Cricetine) type, very different from the interesting differences to be met with in the different subfamihes that occur in all the other great areas, even including the Indo- malayan. Of about fifteen hundred named forms, roughly eight hundred are Muridae, roughly seven hundred belong to other families. In Central America (with Mexico and the West Indies), already a great in- crease in Cricetinae and a great decrease in the more northern Microtinae is met with. In the Cricetinae north of Panama, as well as the North American genera Neotoma, Peiomvsciis, Oiivchoiiivs, Rcithrodoiitomys, Orvsomys and Sigmodon being well represented, about fourteen more genera start their ranges at once. Confined to the Central America area are Nvctomvs, Nelsonia, Ototy- lomys, Scotinomys, HoJuinvs and others; while Tvloinvs, Nectomvs, and one of the "Fishing-rats," Rluuiiivs, range north from northern South America. The Microtinae range south to Guatemala only; the chief genera in the area being Microtus and Pilvmys. The Sciuridae are very much the same, as regards genera, as in the Nearctic, except that Marmota and probably Tamimciuriis are absent. Few range south of Mexico; but the genus Sciurus has many named forms from this area, and extends into South America. G/aiicomys comes into Central America; while Microsciunis comes up into Nicaragua from South America; but apart from Sciurus and Microsciunis, no genus which occurs north of Panama crosses south of it. The Castoridae touch extreme North Mexico. The Heteromyidae are widely distributed through Central America, the primitive genera Hetcromvs and Lioiiiys occurring more or less throughout, while Dipodoinys and Perognathus are represented in Mexico. The Geomyidae are likewise common in the area, and one genus, Macrogeomxs, ranges as far south as Panama. This group (Heteromyidae -t- Geomyidae) is, however, not known in South America except for a few forms of Heteromvs from the extreme northern countries. Four families belonging to the Hystricoid branch occur in Central America ; of these the Erethizontidae is represented by Coendou (Mexico southwards), while the Dasyproctidae (Dasvproctti), and the Cuniculidae (Ciiiiiculus) start their range which is, as in the case of Coendou, from Mexico southwards over the greater part of tropical South America. The Echimyidae are represented by three subfamilies, two of which, Capromyinae (Geocapromys, Capromys), and Plaeiodontiinae (Plagiodontia), appear to be confined to the West Indies (where surprisingly few genera of Rodents occur), except that a member of the Capromyinae has been described from Venezuela ; the other, the more generalized Echimyinae, being represented by Proechimxs, Diplortivs, and Hoploinys, from Nicaragua southwards. The Caviidae are represented by Hydroclioerus which extends north to Panama. South of Panama, vast quantities of Cricetine Muridae swarm, belonging to a very large number of named genera, which are in many cases not or barely distinguishable from each other. These group themselves round the following main types: Oryzomvs, which appears to occur in all parts from Patagonia to Colombia, and has many close allies as Nectomvs, Rliipidoinvs, Tliomasomys, and perhaps leads to the specialized and isolated North Argentine Scapteromys; DISTRIBUTION 55 Akodon, with a very wide range in the continent, and with several allies the best known of which is Oxvmycterus; probably leading to the specialized subfossorial Notiomys of Patagonia; P/ivllotis, with its allies Hesperomys, Eligmodontia, which series may lead to such dentally highly modified types as Chinchillula and Irenomys; and Holochilus, with its allies Neotomys and Reithrodon which seem to correspond to the Nearctic Sigmodon type. By far the most interesting of the Neotropical Cricetines are the Fishing- rats, Ichthyomys and Anoiornys, which must be among the most specialized of all iMuridae, and in cranial characters parallel to a large degree the Australasian aquatic members of the Hydromyinae. The Squirrels of South America are, with the exception of Sciurillus from the Guianas, which appears to be a type which one might consider archaic and allied to the Indomalayan Naiviosciiirm, all essentially types which agree to such a degree with the northern Sciurus that there appears no necessity to separate them from that genus, except for a closely allied type Microsciurus . The family ranges south to Jujuy (North Argentine) only. The type of Squirrel found in South America suggests that the family has "come in" recently, comparatively speaking, from the north, and has not been isolated from more or less Eocene times or before in the Continent when it was (as generally admitted) an island, as I suggest most of the Cricetines and Hystricoid types have. The Caviidae, represented by two subfamilies Hydrochoerinae (Hvdroclioeriis only; tropical portions), and the Caviinae, containing the more specialized Dolichotis from the southern plains, and the more primitive Kerodon (Brazil), and Cavia and its immediate allies which between them cover the continent, are confined to the area. They are in this work not regarded as typical Hystricoidae, but referred to a separate superfamily on account of the formation of the lower jaw. The Hystricoid branch of the Rodents is represented in South America as follows : Northern tropical forest area: Family Echimyidae: two subfamilies, the Dactylomyinae, Thrinacodus and Dactylomvs; the Echimyinae, several genera among which the arboreal Echimys and Mesomys and the terrestrial Proechimys have the widest ranges. From South Brazil are known two rather distinct types referable to the latter subfamily in Clyomys and Carterodon. Family Erethizon- tidae : two very distinct subfamilies, the Chaetomyinae (Chaetomys only, distribution evidently local), and the Erethizontinae {Echinoprocta: Colombia; and Coendou, distribution general). Family Dasyproctidae : Dasyprocta and Myoprocta (distribution general). Family Cuniculidae: Cuniculus (distribution general?). The Family Dinomyidae (Dinnmys) is confined to Peru and Ecuador region. The Subfamily Octodontinae (Echimyidae) is represented by Octodon and Ctenomys as far north as Peru on the western side of the continent. In Peru also Ltigidiiini represents the Family Chinchillidae. In the plains and moun- tains of the southern part of the continent the following Hystricoid types occur: Family Chinchillidae {Ciiinchilla, Lagidium, Lagostomiis). Family Echimyidae : three subfamilies, Alyocastorinae (Myocastor only). Abrocominae (Abrocoma only), and the Octodontinae, of which Ctenonivs, Spalacopus, Acotuiemys, Octomys and Octoduii are the main genera. Besides these types, the genus Heteromys (Heteromyidae) occurs in Colombia, \'enezuela, and Ecuador. 56 DISTRIBUTION Africa is the only geographical area remaining to be discussed. This continent must surely be considered the present headquarters of the Order so tar as variation in character goes, in that it contains more superfamilies than any other area, four (out ot eight) of which are now confined to the continent. Roughly eleven hundred and fifty forms are named from the area; here once again the Muridae are very much the dominant feature in that about eight hundred of the above forms belong to the family. The African types, both of Squirrels and Rats and even Porcupines, have a rather difl^erent aspect from those of the Palaearctic or Indomalayan, and appear to be rather well separated from them in general. The Murinae possess a very large number of genera, most of which appear to have a very wide range on the continent, and very few of which are at present known from any other continent. The most distinct genera are Cricetoniys and Saccostomus. Other aberrant but more typically Murine types are Lophur- otnys, Acomys, Uraiiomvs, Mvlomys, Thamnomxs, Beamys, Dasxmys, Arvicanthis and its immediate allies (Aizictiiithis ranging north to Kgypt and occurring in Arabia, as does Acomvs), Oenoiiivs, Zelotomys, Colomys, etc. Besides these occur many indigenous wild species of A/w, and various groups of Rattiis, some of which have received generic names which appear quite unretainable. The Dendromyinae is a group confined to the continent, very closely allied to the Murinae, and containing Deiidromus and Steatumvs which have a wide range, and Piionomvs and Malacotlirix, which are more restricted, the latter being one of the most aberrant members of the whole family. Deomys, here regarded as type of a distinct subfamilv the Deomyinae, is confined to the Congo. The Otomyinae, with two valid genera Otvmys and Parotomys, are an interesting group confined to the area. The subfamily Gerbillinae is very widely distri- buted through the continent in suitable areas, containing more generalized types in Tatera, Gerbilliis, etc., and some more local specialized genera as Dismodillus, Dcsinodillisciis, Annnodilhis, etc. The Cricetinae is represented by one genus only, I\Jvstrot)iys, from the south; but the Microtinae are not known e.xcept in the Palaearctic coastal region. Even this does not exhaust the list of sub- families, as Tachyoryctes, type of the Tachyoryctinae, though not hitherto currently referred to the Muridae, is here regarded as a member of the family; this genus is known from the eastern and central portion of the continent. The Muscardinidae are represented by the subfamily Graphiurinae, the genus GrapliiiDus ranging over most of the continent. Eliomxs, a Palaearctic type belonging to the typical subfamily, ranges south to the Rio de Oro. Lophiomys, here regarded as type of a family the Lophio- myidae, is confined to the eastern part (Abyssinia, Somaliland, Kenya, Sudan). The Dipodidae is represented by Jaciilus in the Sahara and Somaliland. The family does not range south of this area. The Ctenodactylidae is another northern African group, with very much the same range collectively as the Dipodidae; the principal genera are Pectinator, Massoutiera and Ctenodactylus. The group is known fossil from South Europe, and from India. The Pedetidae, with one genus, Pedetes, is confined to the continent, ranging in the south and east. The Anomaluridae is another family DISTRIBUTION 57 peculiar to Africa, occurring mainly in the western forests, and containing two subfamilies, the Anomalurinae {Anomalurus, Anomalurops), and the Idiurinae {Zenkerella and Idiurus), these groups sometimes being given family rank. The African Sciuridae consist of a relatively small number of forms, one of which (Xeriis) is terrestrial, and is represented as indicated already in parts of the Palaearctic, one of which, Mvosciiirus, is a pygmy form perhaps not distantly related to the Indomalayan Sannosciiiriis, and the remainder of which are arboreal types ot which Protoxerus, Heliusciuriis, Funisciurus, and Paraxenis have the widest ranges. The Ilystricidae are widely distributed through the continent ; the genus Hystrix is the dominant form of this family and here reaches its highest degree of specialization ; while the genus Atherurus is found in the western and central forests, with species of a rather more advanced type than their Indomalayan cousins. Two other Hystricoid genera occur: T/irvonomys, which is best referred to the family Echimyidae (otherwise American) as type of a subfamily, which ranges through most of the continent, and which is known fossil from India; and Petromus, which seems best referred to the same family (as type of a subfamily), and which is known only from South-west Africa. Yet another exclusively African family, the Bathyergidae, some of the most isolated living Rodents, range collectively through most of the area: Cnptomys has the widest range; other more highly specialized but more restricted types are Bathyergus and Heterocephalus; and Heliophobius which appears unique in the whole Order in dental formula has a moderate range on the eastern side. In Madagascar Rodents are unknown save for half a dozen peculiar Muridae. It has been the custom of late years to refer these to a subfamily (or familv) the Nesomyinae or to place them in the Cricetinae. I have been able to find no characters which keep them apart as a distinct subfamily, nor do they all appear to be Cricetinae. In the present classification I have had provisionally to refer them to no less than five dilTerent subfamilies. The names of these genera are Eliurus, Brachyuromys, Xesomys, Gymnuromys, Hvpogeomvs and Brachytarsomys. Their status will be discussed in the volume set aside for RODENTS OF THE PALAEARCTIC (OTHER THAN MURIDAE) Genera, Prinxip.^l Species, and Approximate Ranges SCIURIDAE Genus Trogopterus xanthipes. China; Tibet to Chihli. Genus Petaurista alborufus. China; Tibet to Hupeh. stilcattis. China; Chihli. albiventer group. Kashmir; Japan, Manchuria; Szechuan. Genus Pteromys tolans. Scandinavia across Siberia to Japan, Kansu. Genus Eoglaucomys fimbriatus. .Afghanistan, Kashmir. 58 DISTRIBUTION Genus Eupetaurus cinereiis. Kashmir. Genus Sciiirus vulgaris group. All Europe; Siberia, to Manchuria, Chihli, Japan. anomalus group. Caucasus area. Cienus Callosciurus ntiic/el/andi group. Tibet; Chihli. erythraeus. Szechuan. (Indomalayan type.) Genus Dremomvs pernyi. China; Szechuan, Ilupeh. (Indomalayan type.) rufigems. China; Szechuan. (Indomalayan type.) Genus Funambuhis palmarum group. North Punjab. (Indomalayan type.) tjenus Atlantoxerus getulus. Morocco, (jenus Spermopliilopsis hptodactxlus. Afghanistan, Turkmenia. (jenus Sciurotamias davidianus. China; Kansu and Szechuan to Chihli. Genus Tamias sibiricus. North Russia, Siberia, China north of Yangtze, to Japan. Genus Citellus citelhis group. South-eastern Europe, Russia, Asia Minor; Shansi, Kansu, Mongolia, Transbaikalia (dauricus, etc.). suslicus. East Europe, South Russia. fulvus. East Russia, Turkestan, Persia. pygmaeus group. South Russia, Turkestan; Mongolia {pallidi- cauda). eversmanni . Russian Altai to East Siberia, North Mongolia. Genus Marmota mar/nota. Alps and Carpathians. bobak group. Poland, Russia, Altai, North Mongolia, Kansu, Transbaikalia, Tibet. caudata group. Kashmir, Afghanistan, Russian Turkestan, Chinese Turkestan. caligata group. North-east Siberia. CASTORIDAE Genus Castor fiber. Main rivers of Central Europe; Scandinavia; parts of European Russia; Mongolian Altai. CTENODACTYLIDAE Genus Ctenodactxlus gundi. North Algeria. Genus Massoiitiera mzabi group. North Algeria. DISTRIBUTION ^^ DIPODIDAE Oenus Stcista subtilis group. Scandinavia, Denmark, Hungary, Balkans, Russia Siberia to Lake Baikal. concolor group. Caucasus, Altai, Kashmir, Kansu, Manchuria aakhahn. ' Genus Eozapiis setchuanus. Szechuan, Kansu. Genus Cardiocrankis paradoxus. Nanshan, Sinkiang (China) Genus Salpingotus kozlmi. MongoHa; Gobi. crassicauda group. .Mongoha; Gobi, and in Afghanistan Genus Euclioreutes " naso. Chinese Turkestan, Inner Mongolia Genus Allactaga major group. Southern Russia, Russian Turkestan s,bmca group. Eastern Caspian region to Kansu, Mongolia, 1 ransbaikalia, Chihli. * elater group. Caucasus to Persia, Afghanistan, Kashmir, Chinese lurkestan; also Mesopotamia [tniphratka). wilhamsi. Caucasus, Asia Minor. tetradactyla. North Egjpt. Genus Alactagulus pumilio. Caucasus, Russian Turkestan, Inner Mongolia Genus Pygeretmus *' platyunis. Western Russian Turkestan. shitkovi. Eastern Russian Turkestan Genus Dipus '"S'tta. Caucasus across Russian Turkeston to -Mongolia and Chihli. " Genus Scirtopoda telum group. South Russia, Russian Turkestan, Mongolia Oenus Eremodipus ^ lichtensteini. Turkmenia. Genus Jaculus orientals. Palaearctic North Africa. jaculus. Across Palaearctic North Africa, Syria, Persia Uenus Faradtpus ctenodactylus. Turkmenia. HYSTRICIDAE Genus Atherurus Genus ;/• "'""'°"''"^- China; Szechuan. (Indo-Malayan type.) subcristatus. China; Szechuan. (Indo-Malayan type.) 60 DISTRIBUTION [Hystr:x) kuciita. Punjab, Afghanistan, Russian Turkestan, Transcaucasia, Asia Minor, Syria, Palestine. crislata. Sicily, South Italy, North-western Africa. iMUSCARDINIDAE Genus Mvomimus personatus. Transcaspia. Genus Eliumys quercimis group. Continental Europe south of Baltic; Syria, Northern Africa westwards from Tunis. Central and southern Russia. Genus Dyromys nitechda. Central Europe (Switzerland), eastwards across Russia, to N.W. Frontier (N. India), Tianshan, Zungaria. South in Europe to Greece. Asia Minor. Genus Glis glis. Continental Europe south of Baltic; Asia Minor, Caucasus, North Persia; Southern and Central Russia; South Turkmenia. Genus Glirulus japonicus. Japan. Genus Muscardimis uvelhmarius group. Europe, except Iberian Peninsula, including England, Sweden ; parts of Russia. SPALACIDAE Genus Spalax kirgisoriim group. Kirghiz Steppes(?), Syria, Palestine, North Egypt and Libya. monticolii group. Hungary, Roumania, Balkans, Asia Minor, Caucasus. nikrophthalmus group. Southern Russia, represented in Greece, Rumania, Poland. giganteus. Eastern Russia. RHIZOMYIDAE Genus Rliizomys vestitu$. Szechuan. RODENTS OF THE NEARCTIC (OTHER THAN MURIDAE) Genera, Principal Species, and Approximate Ranges APLODONTIIDAE Genus Aplodontia rufa. Western U.S.A., from California into southern British Columbia, Pacific side Rocky Mountains. DISTRIBUTION 6i SCIURIDAE Genus Glaucomys volans. Eastern U.S.A., from New York and Minnesota south- wards, including Florida. sabrintis. Labrador ; across most of Canada ; Alaska ; Pacific coastal States of U.S.A., cast to Idaho; Virginia. Genus Sciurus carolinensis. Eastern U.S.A. and southern East Canada, west to Minnesota, Oklahoma, including Florida. griseiis. California, Oregon. aherti. Colorado, Arizona, New Mexico. niger group. Eastern U.S.A., from Texas and South Wisconsin eastwards, including Florida; Arizona. Genus Tamiasciurus hudsonicus group. Most of Canada; Alaska; Western U.S.A., south to California, Arizona, New Mexico; Central U.S.A. (Minnesota, South Dakota, etc.); Eastern U.S.A., south to North Carolina at least. Genus Tamias alpinus. California. minimus group. Western U.S.A., east to Wisconsin, west to California and Pacific States; north to Yukon, Mackenzie, and to Lake Superior. amoenus group. Western U.S.A., coastal states, east to Montana, north into Alberta and British Columbia. quadrivittatus group. Montana, Idaho ; California east to Colorado ; Arizona, Texas, New Mexico. towiisendi group. Coastal states of western U.S.A., east to Utah, New Mexico. striattis group. Eastern U.S.A., west to Oklahoma and Minnesota, south to South Carolina, north to Canada (Ontario). Genus Ci tell us tozvnsendii group. Washington, Oregon, Idaho, Utah. 'aashingtoni group. Washington, Idaho. richardsoni group. California, Oregon, Nevada; Wyoming; Saskatchewan. parryii group. Oregon and Idaho north to Arctic Canada (xMac- kenzie and east to Hudson Bay), and Alaska. Iridecemlineatiis group. Western and West Central U.S.A., from Arizona, New Mexico and Texas north to Minnesota, the Dakotas, and Montana. spilosoma group. Arizona, New Mexico, Texas, north to (.')Nebraska. Jranklinii group. Saskatchewan south to Oklahoma and Illinois. variegatus group. Texas andColoradowesttoCaliforniaandOregon. 62 DISTRIBUTION (Citellus) harrisii group. Texas, Colorado, Arizona, California. tereticaiidiis group. California, Arizona, Colorado. lateralis group, .\rizona, Colorado, Wyoming and Montana west to Pacific coastal states, north into Canada (Alberta). Genus Marmota munax group. Across Canada from Labrador to Alaska, and Eastern U.S.A., south to North Alabama, west to Kansas and Minnesota. flaviventris group. Western U.S.A., from South Dakota, Colorado and New Mexico to Pacific states. Into British Columbia. caligaUi group. Western Canada and Alaska, south to \\'ashington and Montana, east to Alberta. Genus Cynomys ludovicianus group. The Dakotas and Montana south through West Central U.S.A. to Texas and Arizona. gunnisoni group. Slightly to the west of the range of /iidozicianin:; Wyoming south into Arizona and New Mexico. CASTORIDAE Genus Castor canadensis. "Most of North America from .Maska and Labrador to the Rio Grande" (Anthony). HLTEROMYIDAE Genus Liomvs irruratns. Southern Texas. Genus Perognatlius fasciatiis group. The Dakotas, Nebraska and Texas west to Wyoming, Colorado, Arizona. longitnembris group. California to L'tah and Arizona. parvus group. California and Pacific states north into British Columbia, east to Utah and Wyoming. formosiis group. Utah, California. hailevi group. Arizona, California. Iiispidiis group. Kansas, Oklahoma. penicillatus group. Texas, Arizona, California. intermedins group. New Mexico, Arizona, California. californiens group. California. spinatus group. California. Genus Microdipodops megacephaliis group. California, Nevada, Oregon. Genus Dipodomys lieermanni group. California, Oregon. spectabilis group. Arizona, New Mexico. phillipsii group. I'exas. merriami group. Texas and L'tah west to California. DISTRIBUTION g iDipodomys) ordii group. Oklahoma, Texas, Wyoming west to Oregon and Lalifornia. ° anilis group. California. compactiis group. Texas. tmcrops group. Oregon, California, Arizona desert! group. California. ^ .,„ GEOMYIDAE Lrenus 1 homomvs lowmendi group. California, Nevada, South Idaho 6o«<,. group. Oregon, Nevada, California, Colorado, Arizona Mew Mexico, I exas. alpinus group. California. perpalUdm group. New Mexico and Utah west to California Julvus group. Arizona, New Mexico, Texas. umbrinus group. Arizona. talpoides group. Colorado, Idaho and North Dakota, west to Washington, north to Canada (Saskatchewan) fossor group. California and Oregon east to Colorado and Wyoming. duuglasii group. Washington, Oregon. monticola group. California, Oregon ^"''' AlbeZ' ^'''""^'""' O^'^g""' ^daho, Montana, north mto bulbivortts group. Oregon. Genus Geomys tuza group Eastern U.S.A. (Alabama, Georgia, Florida) Tl Upper Mississippi ^•alIey, to Kansas, Missouri Illinois; v\est to Nebraska and the Dakotas breviceps group^ Central U.S.A.; Nebraska south to New Mexico Texas, and Louisiana. ' Genus Cratogeomys castanops. Colorado, New .Mexico, Texas. DIPODIDAE iicnus Zapus hudsonius group. Evidently most of Canada and U.S A east to Labrador and North Carolina, west to .\laska, British Columbia Genus Napaeo.aTJ ''''"' '" "' '''"'°™"' ^°"^' ^° -^^"' ^^^■^'-• '""""t^^'Tn S'""'*' ='"'' ^•^•^- ^^^"^ 0"'^"° ^"d Wisconsin to iNortn Carolina. „ ^ , ERETHIZONTIDAE Oenus Erethizon dorsatum group. "Most of forested North America north of 40' 64 DISTRIBUTION {Eretluzoti) and south ill the Rockies almost to Mexican bounciary" (Anthony). North to Labrador and Alaska. The more 1 examine North American faunal lists the more I am convinced that too manv specific groups arc admitted, at least in the Order Rodentia. RODENTS OF THE INDOMALAYAN REGION (OTHER THAN MURIDAE) Genfra, Principal Species, and Approxim.\tf. Ranges SCIURIDAE Genus Bclvmys pearsoiii. Sikkim, Assam, 'I'ongking, Formosa. Genus Trogoptenis xanthipes. Yunnan. (Palaearctic type.) Genus Fetaurista petauristu group. Fukien; Formosa; Malay Peninsula, Sumatra, Java, Borneo. alhonifus group. Yunnan, Formosa. puiiitdtiis. Malacca, Borneo. tilhkciiter group. Ceylon, Peninsular India, Nepal, Burma, Siam, Annam, Yunnan. Genus Pteromxscus pulverulenttis. Malacca, Sumatra, Borneo. Genus Aeromys tephromelas group. Malacca, Borneo. tliomasi. Borneo. Genus Hvlopetes alboiiiger group. Nepal, North Burma; Philippines (iiigripes). sagitia group. Burma, Laos, Malay Peninsula, Sumatra, Java, Borneo, Natunas. Genus Petinomys fuscocapillus group. Ceylon, South India. /;«^f«/' group. Sumatra; Philippines (cnw/ViH). geniharbis group. Java, Borneo, Malacca. Hainan {electilis). jf?ox!M group. Sumatra; Tenasserim. Genus Petaurillus hosei group. Malacca, Borneo. Genus lomys horsfieldi. Malacca, Sumatra, Borneo, Java. Genus Nannosciurus exilis group. Sumatra, Borneo, Philippines. whiteheadi. Borneo. melanotis group. Sumatra, Java, Borneo. Genus Sciiirillus muriniis. Philippines. DISTRIBUTION g Genus Callosciurus macleUandi group. Nepal, Assam, Burma, Yunnan, Fukien Haman, i'ormosa, Cochin-China, Siam, Annam erythraeus group. Hainan, Formosa, Yunnan, Kwantung, Assam Hurma, Siam, Annam, south to Pahang camceps gxonY>. Chekiang; Tongking, Tenassenm, Siam south to Malacca. pygerythrm group. Nepal, Bengal, Assam, Burma. quinquestnatus group. Yunnan, Burma. prevosti group. Malacca, Sumatra, Borneo, Celebes notatus group. Malacca, Sumatra, Java, Borneo hippurm group. Malacca, Sumatra, Borneo, Philippines teiiHis group. Malacca, Sumatra, Borneo. loKi group. Malacca, Sumatra, Borneo. leucomiis group. Celebes. rubriventer group. Celebes. Genus Dremomvs pernyi group. Burma, Yunnan to Fukien, Hainan. owstoni. Formosa. everetti. Borneo. lokriah group. Nepal, Assam, Burma. rufigenis group. Malacca, Tenasserim, Burma, Annam, Laos, Yunnan, Haman, Hupeh. Genus Funumbtdus palmarum group. Peninsular India, Ceylon. layardt. South India, Ceylon. siibliiieatus. South India, Cevlon. Genus Ratufa macroura. South India, Ceylon. indica. Peninsular India. gigantea. Assam, Nepal, Burma, Yunnan, Hainan btcolor and other species. Burma, Tenasserim, S,am, Annam, Malacca, Sumatra, Java, Borneo, Bali, Natunas Oenus menetes herdmorei. Burma, Tenasserim, Annam, Siam Oenus Lariscus insignis group. Malay Penii>sula (southern), Sumatra, Java, Borneo. hoset. Borneo. Genus Glyphotes simus. Borneo. Genus Rheitlirosciurus macrutis. Borneo. Genus Rhinosciunis r.n H l"tic"i'd<,tus. Malay Peninsula (southern), Sumatra, Borneo. Oenus Hyoscturus heinrichi. Celebes. 5 — Living Rodents — I 66 DISTRIBUTION Genus Mannutii hobak group (liimalayaiia); Nepal, Yunnan. (Palaearctic type.) Genus Sciurotamias Jorresti. Yunnan. HYSTRICIDAE Genus Triclivs lipura group. Malacca, Sumatra, Borneo. Genus Athenirus macrouriis. Hainan, Southern China, Tongking, Assam, Malacca, Tenasserim, Sumatra. Genus Theciirus piwiilus group. Philippines; Sumatra {siinuitrae). crassispims. Borneo. Genus Hystiix hraclivurus group. Malacca, Sumatra, Java, Borneo, Sumbawa, Floras. subcn'staius group. Sikkim, Assam, Burma, Tenasserim, Yunnan, Fukien, Anhwei, Hainan, (?)Annam. leuciiru group. Ceylon, Peninsulijr India, Nepal. MUSCARDINIDAE Genus Platacanthomvs lasiunis. South Peninsular India. Genus Typhlomys cinereus. I'ongking, Fukien. RHIZOxMYIDAE Genus Rliizomys vestitiis group. Burma; Fukien (davidi). sinensis group. Assam, Y'unnan, Kwantung; South Siam, Perak [paiinosus). siimatrensis group. Tenasserim to Sumatra. Genus Cannomys badiiis. Nepal, Burma, Siam. RODENTS OF AFRICA (OTHER THAN MURIDAE) (With Arabia, but not including Palaearctic North Coastal Area) GENKR.A, Principal Species, and Appro.ximate Ranges BATHYERGIDAE Genus Batliyergiis siiillus. South Africa; Cape Pnivuice. janetta. South Africa; Namaqualand. DISTRIBUTION 5_ Genus Georychus capensis. South Africa; Cape Colony. Genus Cryptomys mec/iou'i group. Angola, Rhodesia. huttetttottiis group. Cape, Rhodesia, Nyasaland, Tanganyika /fc//« section. North Nigeria, French Shari, North Congo Kala- hari, Rhodesia, Portuguese East Africa. zechi. West Africa (Togoland). ochraceocinereus. Sudan. bocagei. Angola. Genus Heliophobim argenteocinerem group. Portuguese East Africa, Rhodesia, Tan- ganyika. South Congo, Kenya. Genus Heterocephaliis glaber group. Somaliland, Abyssinia, North Kenya. SCIURIDAE Genus Alyosciurus pumilio. West Africa; Cameroons, Gaboon. Genus Heltosciurus gatnbianus. Gambia east to Abyssinia, south to Angola and Por- tuguese East Africa. ruwenzorii. Belgian Congo. Ruwenzori. poetuis. Fernando Po, Gold Coast, Gaboon. lucifer. Nyasaland. Genus Paraxerus cepapi group. Kenya {ochraceiis) south to Transvaal, Portuguese bast Africa, Kalahari, and Ovamboland palliaius group. Rhodesia, Portuguese East Africa, Zululand to Kenya, Somaliland. flavivittis group. Portuguese East .-yrica, to Tanganyika and Kenya. boehmi group. Congo, Sudan, Ruwenzori. Genus Funiscmrus lemniscatus group. Gaboon, Cameroons, Congo. congkus group. Angola and South Congo. pyrrhopus group. Sierra Leone east to Congo, .-Vngola, and Ruwenzori. Genus Protoxerus stangeri. Gold Coast and Nigeria east to Kenya, south through Congo to Angola. Genus Mynilus aubinii group. Liberia, .\shanti. Genus Epixerus wilsoni. Gaboon. ebii. Gold Coast. 6S DISTRIBUTION Cjenus Xrnis rutilus group. Somaliland, Eritrea, Abyssinia, Kenya. ervthropm group. Sudan, Kenya, Uganda, Sahara (Air), F^aive Chad, to Sierra Leone. capeiisis group. South and South-west Africa. ANOMALURIDAK Genus Anoinaliinis fmseri. Gold Coast east to Uganda (PKenya), Tanganyika, south through Congo to Angola. peli. Guinea Coast, Ashanti. piisillus. Congo, Gaboon. Genus Aiiomahiiops beecrofti. Sierra Leone eastwards to Congo. Genus Zenkerella !nsl£;ms. Spanish Guinea. Genus hiiiiiiis zenkeri. Canieroons, Congo. macrotis group. Canieroons, Congo. PEDETIDAE Genus Pedetes cafer group. Kenya and Angola to Cape Province. CTENODACTYLIDAE Genus Pectinatur spekei. Abyssinia, Somaliland, Eritrea. Genus Ctenodactylus guiidi group. Tripoli (northern Sahara, west to Morocco). Genus Alassoutiera mzabi group. Sahara, south to Asben. Genus Felovia vae. Senegal. DIPODIDAE Genus Jaculus jaciihis. Sahara, south to Asben, Sudan, Somaliland; and Arabia. (The other species, orientnlis, appears Palaearctic in distribution.) ECHIMYIDAE Genus Petiomus typicus. South-west Africa. Genus Thryotiomys swhideriamts group. liahr-el-ghazal and Uganda to Nigeria, Angola, South Africa. gregorianus group. Congo, Kenya, Nyasaland. DISTRIBUTION 69 HYSTRICIDAE Genus Atherurus africanus group. Gambia, Sierra Leone, Nigeria, Congo, to Kenya. Genus Hystrix leucura group. Arabia. cristata group. Senegal, Asben, Somaliland, Kenya, Uganda, Tanganyika. africaeaustralis group. South Africa, South-west Africa, Portu- guese East Africa, to Tanganyika. MUSCARDINIDAE Genus Graphiurus ocularis. Cape Province. platyops group. South-West Africa, Rhodesia. hueti group. Senegal, Liberia, Cameroons. crassicaudatus. Liberia, South Nigeria. surdus. French Congo. monardi. Angola. u'oosnami. Kalahari. murinus group. Asben, North Nigeria, Sudan, Somaliland, Kenya, Gold Coast south to Cape Province. Genus Eliomys quercinus group. (Prom Palaearctic) south to Rio de Oro. LOPHIOMYIDAE Genus Lophiomys imhausi. Somaliland, Abyssinia, Sudan (Kassala), and Kenya. RODENTS OF THE NEOTROPICAL (OTHER THAN MURIDAE) (According to Flower & Lydekker, Mexico should be included in this region.) Genera, Princip.\l Species, and Approximate Ranges SCIURIDAE Genus SciuHllus pusillus. (iuianas. Genus Syntheosciurus brochus. Panama Genus Microsciurus aljari group. Nicaragua south to Peru and Upper Amazon. Genus Sciuriu rariegatoides group. Mexico to Panama. deppei group. Mexico, Nicaragua. aberti group. Northern Mexico. 70 DISTRIBUTION (.Samus) niger group. Mexico. hoffmimi group. Nicaragua to Venezuela, Ecuador; Peru; North Argentine (Jujuy). aestuaiis group. Guianas, Venezuela, Eastern Brazil to Minas Geracs. stramincus group. Ecuador, Peru. piicherani group. Colombia, Peru, Bolivia. rhoadsi. Ecuador. jlaiiimijer . Venezuela. langsdorjfi group. Venezuela, Colombia, Peru, Ecuador, Bolivia, Brazil to Matte Grosso. Genus lamias quadrivittatus group. North Mexico. (Nearctic type.) Genus Citellus mexicanus. Mexico. spilosoma. Mexico. variegatus group. Mexico. annidatus group. Mexico. tereticaudiis. North Mexico. lateralis group. North Mexico. Genus Cynomys mexicanus. North Mexico. (Nearctic type.) Genus Glaucomys volans. Through Mexico to Honduras. CASTORIDAE Genus Castor canadensis. Extreme North Mexico. (Nearctic type.) HETEROMYIDAE Genus Heteromys anomalus group. Venezuela, Colombia, Ecuador. desmarestianus group. Mexico to Panama. gaumeri. Mexico (Yucatan). nelsoni. Mexico (Chiapas). Genus Liomys pictiis group. Mexico. crispus group. Mexico to Panama. irroratus group. Mexico. Genus Perogtiathus fasciatus group. North Mexico. (Nearctic type.) longimemhiis group. North Mexico. (Nearctic type.) baileyi group. North Mexico. (Nearctic type.) hispidiis group. Northern Mexico. penicillatus group. North-western Mexico. (Nearctic type.) intermedins group. Northern Mexico. 71 DISTRIBUTION Genus Dipodomys spectabilis group. Northern Mexico. phillipsii group. Mexico. merriami group. Northern Mexico. ordii group. Northern Mexico. deserti group. North Mexico (Sonora). (Nearctic type.) GEOMYIDAE Oenus 1 homomys hottae group. North Mexico, Sonora. (Nearctic type ) perpalhdus group. North Mexico, Sinaloa. (Nearctic tvpe ) timbrtnus group. Mexico. Genus Geomys r.n p ^/'■^^'■^''/'^ g'-o"P- N.E. Mexico (Tamaulipas). (Nearctic type.) yjenus. Fappo^eomvs }^'--J hulleri group. Mexico; Jalisco. Genus Cratogeornvs castanops group. Through Mexico. Genus Platygeomys gymnurus group. South Central Mexico. Genus Orthogeomvs gran'dis group. South Mexico to Honduras, Salvador uenus Heterogeomys hispidus group. Southern Mexico. Genus Zygogeomys trichopus. Mexico; Michoacan. Genus Macrogeomys heterodus group. Nicaragua, Costa Rica, Panama. . CAVIIDAE Genus Cavta aperea group. North Argentine (Tucuman, Corrientes), north to ,-. ^ , ^eru, Colombia, Venezuela, British Guiana. Genus Galea ./.mV^group. Southern Brazil, Bolivia, Argentine (to Upper Rio Genus Caviella australis. Argentine to Patagonia. shiptoni. Argentine (Catamarca). niata. Bolivia. Genus Kerodon rupestris. Eastern (?) Brazil. Genus Dolichotis patagona group. Argentine to Patagonia (Cordoba southwards ) saltmcola group. Argentine. 72 DISTRIBUTION Genus Hydrochoerus hydrochaeris group. WarmcT portions of South America (exact range not traced), north to Panama. Known to occur in Brazil, Paraguay, British Guiana, Venezuela. CHINCHILLIDAE Genus Chinchilla latiiger. Northern Chile. Genus Lagidium viscaccia group. Peru, Bolivia, Argentine, Chile (south to 50° S.). Genus Lagostomus maximus. Argentine. DINOMYIDAE Genus DiiKimvs hranickii. Peru, Colombia, Ecuador, Upper Amazonia. ECHIMYIDAE Genus Echimys dasythrix group. East Brazil; Bahia to Rio Grande do Sul. hlainzilki group. East Brazil; Bahia to Parana. thomasi. Island oflF Bahia, East Brazil. armatus group. Guianas, Brazil (North ?), Venezuela. chrvsurus group. Dutch Guiana, N.E. Brazil (Para). satiirnus. Ecuador. grandis group. Peru, Upper Amazonia. Genus Isothrix pictiis group. East Brazil (Bahia). histriatus group. Peru, Venezuela, Brazil south to Matto Grosso. Genus Diplomys caiiiceps group. Panama, Colombia. Genus Pioechimvs cayennensis group. Nicaragua southwards to Guianas, Peru, Bolivia, Minas Geraes. cauicollis. Colombia. iheringi. Island off Sao Paulo, Brazil. (Sao Sebastian Island.) setosus group. East Brazil; Bahia. Genus Hoplomvs gvmnuriis group. Nicaragua, Panama, Ecuador. Genus Cercomxs cunicularius. Paraguay, and East Brazil (Minas Geraes, Bahia, Pernambuco). Genus Eurvzxgomutomys spinosus. Paraguay, South-eastern Brazil. Genus Clvomvs lattceps. S.E. Brazil (Santa Catharina). DISTRIBUTION 73 Genus Carterodon sulcidens. South Brazil; (?Lagoa Santa). Oenus Mesomys Genus L.„.;^Kf '""^^ ^^^°"'^= '^"""^'"^ ^^^'^ *° Ecuador. Peru. emiliae. Central Brazil; Rio Tapajoz. Genus Procapromys geayi. Venezuela. Genus Capromys pilorides. Cuba. melanurm. Cuba. nana. Cuba. Genus Geocapromvs brounii. Jamaica. Genus P/.^iZ' " '""''■ '"'" '^'^"' ^"'^ "°"''"^^^)' -^ Bahamas. aediim. Dominican Republic. Genus Thrinacodiis albicauda group. Colombia, Venezuela Genus Dactylomys dactylinus. Ecuador, Bolivia, Brazil (Amazonia). peruanus. Peru. Genus Kannabatcumys amblyonyx. Paraguay, S.E. Brazil (Sao Paulo) Genus Myocastor ' corpus. Chile, Patagonia, Paraguay, Argentina. Genus Abrocoma bennetti. Chile. cinerea. Northern Argentina. Genus Octomys mimax. North Argentine (Catamarca, San Juan) Genus Aconaemys ' Genus Octodm'" ^'™"^' ^°"'^"" ^"''''^' Argentina (Andes). deous group. Chile, Peru. Genus Octodontomvs i^liroides. Bolivia. Genus Spalacoptts cyamis group. Chile. Genus Ctenomvs magellanicus section. Paraguay, North Argentina to Patagonia torquatus section. South Brazil, Bolivia, North and Central -Argentme. leiicodon. Bolivia. opimus section. South Peru, Bolivia to Chile and Patagonia bolivtensis section. Bolivia. 74 DISTRIBUTION ERETHIZONTIDAE Genus Cluietomys subspinosus. Brazil; tropical? (exact locality not traced). Genus Echinoprocta rufescens. Colombia. Genus C'oeiidoii prehensilis group. Colombia, Brazil (Matto Grosso, PPernambuco), Bolivia. bicolor group. Bolivia, Peru, Ecuador, Panama (rothschildi). mcxicaimm group. Mexico, Panama. paranaycnsis group. Paraguay, S.E. Brazil, Eastern Brazil ? vestitus. Colombia, Venezuela. DASYPROCTIDAE Genus Mvoprocta acoucliy. Cayenne, Amazonia. pratti. Peru, Ecuador, Colombia, east to Manaos region, Brazil. Genus Dasvprocta punctata. Mexico to Panama. rariegata. Peru, Ecuador, Colombia, Bolivia, Matto Grosso. aguti. Guianas, Brazil. Allied forms in Lesser Antilles. CUNICULIDAE Genus Citniculus paca group. Mexico, Panama, Ecuador, Colombia, Brazil, Cayenne. Probably south to Paraguav. taczauoivskii group. Ecuador, Venezuela. ?Peru. DISTRIBUTION OF RODENTS: SPECIAL WORKS OF REFERENCE Miller: Catalogue of Mammals of Western Europe, 1912. Vinogradov: Rodents of U.S.S.R., 1933, Tab. Anal. Faune de L'URSS, Inst. Zool. Ac. Sci. 10, p. I. Tate: Some Muridae of the Indn-Australian Region, Bull. .-^mer. Mus. Nat. Hist. LXXII, p. 501, IQ36. St. Leoer: Key to Families and Genera African Rodents, 1931, P.Z.S., p. 957. Hollister: Smiths. Inst. Bull. 99, 1919; East African Mammals in U.S. National Museum. Miller: List of North ."Xmencan Recent Mammals, 1923, Smiths. Inst. U.S. Nat. Mus. Bull. 128. Anthony: Field Book of North American Mammals, Putnam, 1928. Flower: Mammals of Egypt, P.Z.S., p. 369. 1932. Wrouchton: Indian Mammal Survey. Joum. Bombay N.H. .See. XXVI, No. 2, 191S {1919), p. 352. Aharoni ; Aluridac of Syria and Palestine, Zeitschr. filr Siiugethierk. Bd. 7, 1932. Robinson & Kloss : Nominal List of Oriental Sciuridae, 1918, Rec. Ind. Mus. XV, IV, p. 171. DISTRIBUTION 7S Robinson ; List of Sumatran Mammals, Journ. Fed. Malay States Mus., VIII, appendix, 1918. Taylor: Mammals of Philippine Islands, -Manila, 1934. Iredale & Troughton : Check List of Mammals recorded from Australia. Memoir VI, '934- Gyldenstolpe : Neotropical Cricetinae, Kungl. Svenska. Vetens. Hand. II, no. 3, 1932. Tate; Taxonomy of Neotropical Hystricoid Rodents, 1935; Bull. Amer. Mus. Nat. Hist., LXVIII, p. 295. Blanford: Fauna of British India, 188S, London. Jones : Mammals of South Australia, Adelaide, 1923. Shelford: a Naturalist in Borneo, igi6. Gee: Bull. Dep. Biol. Yenchiang Univ., 1929-30, 1,2; List of Mammals occurring in China. Shortridce: Mammals of South-West Africa, London, Heinemann, 1934. Allen, G.M.: Check List of African Mammals, Bull. Mus. Comp. Zool. LXXXIII, 1939. Rummler: Die Systematik und Verbreitung der Muriden Neuguineas, Send. Mitt. Zool. Mus. Berlin, 23, heft i, 1938. Osgood, W. H. : Mammals of the Kelley-Roosevelts and Delacour Asiatic Expeditions, Field. Mus. Publ. Zool. iS, 1932, pp. 193-339. Dammermann : On the Zoogeography of Java, Treubia, Vol. XI, livr. i, 1929, Appen- dix I, Mammals, pp. 33-39. ScHWARZ, E. : On the Evolution and Radiation of Mammalian Faunae. .\ct. Zool. Stockholm, 5, 1924, pp. 393^423. Order RODENTIA Key to Superfamilies here recognized Lower jaw much specialized, either by distortion outwards of the angular portion, by specialized limb of masseter lateralis superficialis, or by a conspicuous ridge extending along outside of jaw below level of toothrow, for attachment of masseter medialis. Infraorbital foramen not or scarcely transmitting muscle. Fibula reduced and fully fused with tibia. (Masseter lateralis chief agent in modifying form of mandible.) Superfamily Bathyergoidae Infraorbital foramen much enlarged for muscle transmission. Fibula rarely reduced, not fully fused with tibia. Masseter lateralis chief agent in modifying form of mandible. Superfamily Hystricoidae Masseter medialis chief agent in modifying form of mandible. Superfamily Cavioidae Lower jaw not much specialized, never with angular portion distorted outwards and never with deep ridge extending along outside of jaw for attachment of masseter medialis. Infraorbital foramen not or scarcelv transmitting muscle. Zygomatic plate completely beneath the infraorbital foramen. Superfamily Aplodontoidae Zygomatic plate more specialized, broadened and tilted upwards to a greater or lesser degree, never completely beneath infra- orbital foramen. Jugal bone long, usually extending to lachrymal; fibula so far as known not fully fused with tibia. No externally-opening cheekpouches present. Cheekteeth normally complex. Skull without well-marked postorbital processes ; jugal much broadened; cheekteeth extremely hypsodont, not cuspidate in pattern; external form much modified for aquatic life, tail broadened, flattened, naked, the vertebrae broadened. Superfamily C.astoroidae Skull with postorbital processes, well developed in the majority; jugal not specially broadened, cheek- teeth usually not hypsodont, cuspidate in pattern; 78 RODENTIA external form never modified for aquatic life, tail normal, always fully haired. Superfamily Sciuroidae' Jugal bone strongly reduced, never approaching lachrymal, the zygoma sometimes complete without it, the whole zygoma in some cases reduced, threadlike. Fibula, so far as known, fully fused with the tibia. Large externally- opening cheekpouches present. Cheekteeth with ten- dency to become simplified. Superfamily Geomyoid.'VE Infraorbital foramen always enlarged for muscle transmission. Zygomatic plate very generally tilted upwards and broadened to a greater or lesser degree (two exceptions out of approxi- mately two himdred genera). The infraorbital foramen never much enlarged. Fibula fused with the tibia. Superfamily Muroidae Zygomatic plate never tilted upwards, always narrow and com- pletely below the greatly enlarged infraorbital foramen. Premolars becoming suppressed, either absent, vestigial, or shed in the adult. Fibula fused with the tibia; cheekteeth rooted, complex in pattern; angular portion of mandible weak, not drawn backwards. Superfamily Dipodoidae Fibula free from tibia; cheekteeth evergrowing, simplified in pattern; angular portion of mandible strong, drawn backwards to a certain degree. Superfamily Ctenodactyloidae Premolars not suppressed, not shed in the adult, normally as large as the molars, and not reduced. Fibula fused with the tibia; cheekteeth evergrowing, simpli- fied in pattern; external form saltatorial; mastoids much inflated. Superfamily Pedetoidae Fibula free from tibia, so far as known; cheekteeth rooted, complex in pattern; external form arboreal; mas- toids not much inflated. Superfamily Anomaluroidae I would point out, before dealing with the families and genera, that sub- genera as here retained are equivalent in rank to sub-genera as understood by American authors; and are not groups which must at once be given full generic rank, as has been done so often by authors other than Americans, because they form "natural groups" or because of convenience. Some excellent remarks on the status of genera and sub-genera are given by Osgood in his revision of the American genus Peromysais, to which I would refer my readers. ' For the wide differences between Castor and the Sciuridae in external characters see Pocock, Proc. Zool. Soc. London, p. 1171, 1922. BATHYERGIDAE 79 The present author inclines to the view that systematic classification would be none the worse if sub-genera were abolished altogether. Superfamily BATHYERGOIDAE As here understood this contains one living family. Family BATHYERGIDAE 1S96. Thomas; Myomorpha, part; Family Bathyergidae. 1899. Tullberg: Hystricognathi ; Bathyergoniorpha, Family Bathyergidae. 1918. Miller & Gidley : Superfamily Bathy'Ergoidae ; Family Bathyergidae. 1924. Winge: Family Hystricidae, part, Bathyergini. 1928. Weber: Bathyercoidea ; Family Bathyergidae. Geographical Distributio.n. — Africa: from Sudan, Abyssinia and Somali- land, and from Gold Coast to the Cape. Number of Genera. — Five. Characters. — Zygomasseteric structure unique in the order; mandible with angular portion distorted outwards to "allow passage of a specialized and enlarged distal anterior limb of masseter lateralis super- ficialis" (Miller & Gidley); paralleling the Hystricoidae in this respect, but if anything even more developed than in the most specialized of these. Infra- orbital foramen small, not or scarcely transmitting muscle; if so, only a small strand in certain species, the degree of enlargement of infraorbital foramen evidently in some cases variable individually. Skull and external form much modified for fossorial life. Number of cheekteeth vary-ing in the different genera; cheekteeth strongly hypsodont, but not evergrowing; normally simplified to ring-pattern in adult (excepting the genus Georychus). A tendency present for the upper incisors to extend into the pterygoids. Fibula reduced, fully fused with the tibia. Remarks. — The peculiar jaw-muscle structure combined with the varia- bility of the number of cheekteeth, and the variability of the infraorbital foramen serve to isolate the Bathyergidae completely among living rodents. Elsewhere, there is a strong uniformity in the dental formula of any one group; in some cases, as Sciuridae (cheekteeth ; or J), Dipodidae (cheekteeth ' or :'), there is a difference in the formula, it is true; but in almost all cases the extra premolar retained is vestigial and going; in a vast group like the Muridae the formula of 'i cheekteeth is very general, only a very few Australian and Philippine genera having it reduced to H. But in this family, three completely different dental formulas, or possibly e\en four, are to be found in five genera. These rodents certainly cannot be lumped in "Myomorpha," as was done by Thomas and earlier authors, on account of the comparatively trivial character of the fusion of the tibia and fibula ; nor can they be transferred to the Hystri- coidea, "Hystricidae," as was done by Winge, presumably on account of the similarity of the lower jaw in the two groups, though in VVinge's Hystricidae 8o BATHYERGIDAE the Ctenodactylidae are included, which do not possess the Hystricoid type of mandible. Nor does the infraorbital foramen transmit muscle here, as appar- ently Winge is of the opinion that it does (or did), except to a very small degree occasionally, as discussed below; nor in the Mystricoidae are the tibia and fibula fused, though this is a character which Winge has used elsewhere as a division in other families (Anomaluridae against Dipodidae, etc., page 7). In zygomasseteric structure the Bathyergoidae differ from the Hystricoidae chiefly in that in the latter group the infraorbital foramen is always very much enlarged to transmit muscle, whereas in the present group it is usually not enlarged at all ; this fact , combined with the lack of broadening of the zygomatic plate present, appears to be a primitive condition. According to Tullberg's figures, the temporalis muscle in this family appears less reduced than is usual, taking up the whole of the hinder part of superior portion of skull, and extending forwards nearly to level of anterior zygomatic root (Georyclius capensis). Digits of forefoot and hindfoot five, none reduced. According to Tullberg, the radiale and intermedium of members of this family are separate, alone of rodents (examined by him) except Ctenodactylidae. Malleus and incus fused according to Tullberg, as in Hystricoidae, Cteno- dactylidae, but unlike the remainder of the order. Thomas, Ann. Mag. Nat Hist., 8, IV, p. 111, 1909, suggested that the cheekteeth present in the various genera are probably as follows: Heliopliohius b ^15: 1.2.3. m. 1.2.3. Bathyergus and Georyclius 4 ^ 4 " 5: 1.2. m. 1.2. Heterocephalus 3 3 '•■ 3'::: I. m. — I. " Foiiiarina " {=Hetervccphalus T p. 34- 34- phillipsi) Miller & Gidley, with reference to Heliophubius which exceeds their highest formula for a rodent (i), state: "In the Genus Heliophubius, with the greatest number of teeth, there are never more than j' functional at one time; the apparent addition of one tooth in the upper jaw and two in the lower jaw to the maximtun Rodent formula is probably due to a specialized condition of the milk-dentition." Discussion of Genera. — The genus Bathyergus appears to have evolved in a rather different way from the remainder of the family in that the digging is done not so much with the incisors as with the foreclaws. This has led to great enlargement of these claws, but not to any great lengthening of the upper incisor roots, so that the upper incisors do not show any inclination to extend to the back ot the palate, or the pterygoids. A EATHYERGIDAE 8i parallel to this, between Bathyergus (a "claw-digger") and Georychus (a "tooth- digger"), is seen in Spalax against Myospalax; the two fossorial Microtinae Ellobius against Prometheomys; etc. In all other Bathyergidae, the claws remain relatively small, but the upper incisors extend over the cheekteeth to the back of the palate or at extreme development into the pterygoids. lleUophobius is remarkable in that, as indicated above, it is the only rodent known with l cheekteeth, and appears to be erupting teeth more or less through life. Cryptomys and Georyc/ius are closely allied types, with a dental formula of i ; Georychus, confined like Bathyergus to a small range in South Africa, is the only member of the family without simplified cheekteeth in the adult; Cryptomys with a large number of named forms extends over most of the Continent. Heterocephalus, from Abyssinia and Somaliland, is a most extraordinary animal; alone among the rodents it has become practically naked, having lost the fur almost entirely. Various other characters such as the fact that D.3 in the manus is noticeably longer than D.4, the more strongly shortened and Murine jugal, and the reduction of the cheekteeth to i, or even sometimes |, leads me to believe that it should be separated from the rest as a "generic group." Key to the Generic Groups of Bathyergidae Fur reduced to a few scattered hairs. D.3 in manus markedly longer than D.4. Jugal short, supported anteriorly by the zygomatic process of the maxillary, its general form Murine. Cheekteeth becoming reduced numerically: ^ or - Heterocephalus Group (Heterocephali) 3 2 Fur normal. D.3 in manus never markedly longer than D.4, usually slightly or considerably shorter. Jugal long, forming the greater part of the zygoma. Cheekteeth not becoming reduced numerically: -, or in 6 + one genus, at full dentition, j. Bathyergus Group (Bathyergi) The Bathyergus Group Fur normal; eyes and ears, as usual in the family, greatly reduced; usually D.2 in manus longer than D.3, the digits reduced in size from D.2 to D.5 evenly; pollex not vestigial, clawed. Hindfoot with D.3 remaining main digit, except in Heliophobius. General cranial characters as follows: Skull with frontals moderately or rarely strongly constricted, nasals usually narrow; posterior root of zygoma noticeably broad, and zygomata widely spreading. .\ prominent ridge developed in all genera extending along centre of skull from posterior part of nasals to lambdoid crest. Occipital region usually prominent, outstanding and strongly ridged. Jugal long. Bullae small-moderate, not abnormal. Palate normally excessively constricted between toothrows ; extending behind level of toothrows, in which position it is broader, excepting Heliophobius. Incisive foramina 6 — Living KoUents — I 82 BATHVERGIDAE obsolete. Angle of mandible powerfully distorted outwards; usually not pro- duced far backwards, except in Bathyergus. Incisors thick, pro-odont. Key to the Genera of the Bathyergus Group Cheekteeth i. .4. , . Upper incisors not extending behind toothrows, and heavily grooved. Foreclaws much enlarged. Angular portion of mandible pro- duced considerably backwards. Bathyergus Upper incisors extending behind the toothrows, in extreme develop- ment into pterygoids, not grooved. Foreclaws not specially enlarged. Angular portion of mandible not produced far back- wards. Cheekteeth simplified to ring-pattern in adult. Posterior tooth cut early in life. Cryptomys Cheekteeth retaining one inner, one outer fold to old age; posterior tooth cut late in life. Georychus Cheekteeth at full dentition -. (Upper incisor extends mto pterj'goids; cheekteeth ring-shaped; fore- claws not enlarged; angular portion of mandible not produced far backwards.) Heliophobius 1^-- te. Fig. I. B.^THYERGUS suiLLL's suiLLUS, Schreber. B.M. No. 5.8.10.10. o ; -■ I. BATHYERGIDAE: BATHYERGUS 83 Fig. 2. Bathyergus suillus suillus, Schreber. B.M. No. 5. 8. 10.10, (J; X I. Fig. 3. Bathyergu-s suillus, Schreber. Mandible from below ; X i. Genus i. BATHYERGUS, Illiger 1811. Bathyergus, Illiger, Prodr. Syst. Mamm., p. 86. Type Species. — Mus maritimus, Gmelin. Range. — South Africa: Cape Province, and Namaqualand; coastlands. 84 BATHYERGUS— HELIOPHOBIUS Number of Forms. — Three. Ch.\racters. — Skull essentially as described above: frontals in the rs'pe species much constricted; occipital region extremely ridged and powerful in old age. .\ngular portion of mandible produced considerably backwards, the mandible being perhaps proportionately larger in relation to the upper part of the skull than in any other member of the order. Upper incisors one-grooved, lowers plain; roots of upper incisors not extending to nor approaching posterior part of toothrow. Cheekteeth hypsodont, wider than long, when cut, with an inner and outer fold, which quickly wear down so that the tooth is ring-shaped in adult. Infra- orbital foramen normal (small). Size largest tor the family in the type species; ear conch absent: tail as long as hindtoot, thick and flat, with long hairs growing outwards each side giving a feather-like effect. Claws immensely developed in forefoot, particularly of digits 2. I and 4; D.2 longer than D.3. Hindfoot with the centre digit longest, D.2 shghtlv longer than D.4, hallux slightly longer than D.5. Claws of hindfoot medium. Pollex with short claw, not vestigial. Two well-marked species are known, the "giant" suillus, and the moderate- sized ^(jHf/^rt, which appears to have a less heavily ridged skull. Forms examined: suillus. janetta. List of X.\.med Forms I. B.\THYERGUS SUILLUS SUILLUS, Schreher 1782. Saugt. IV, p. 715. pi. 204B. South Africa: Cape. S>"non\-m: maritimus. Gmelin, 17SS, Linn. Syst. Nat. i, p. 140. africatm. Lamarck. Voyages de Thunberg au Japon. etc.. 4, 34S, 179b. ;. B.-\THYERGUS SUILLUS INTERMEDIUS, Roberts 1926. .Ann. Transvaal Mus. XL p. 261. Klaver, Cape Province. 5. B.A.THYERGUS J.A.NETT.\, Thomas S; Schwann 1904. .-^bstr. Proc. Zool. Soc. London, no. 2, p. 6. Port Xolloth, Little Xamaqualand. Genus 2. HELIOPHOBIUS, Peters 1846. Heliophobius, Peters, Monats. Her. Akad. Berlin, p. 259. 1890. Myosc.\lops. Thomas, Proc. Zool. Soc. London, p. 448. Xew name to replace Heliophobius on the assumption that it was preoccupied by Heliophobus, Boisduval. Type Species. — Heliophobius argenteocinereus, Peters. Range. — Eastern and Central Tropical Africa: Kenya, Tanganyika, South Congo, Xorth Rhodesia, Xyasaland. Number of Forms. — Eight or nine are recognized. Char.\cters. — Cheekteeth at full dentition ". The teeth are very infre- quently all in place together; the anterior premolars being shed before the posterior molars are cut. In fifty skulls available for examina- tion only one No. 18. 6. 15. 6 has all six teeth in place together (one side of the jaw HELIOPHOBIUS 85 only). 1 have not seen one with the six lower teeth in place together. The normal number in place at once appears to be either i or ', but sometimes there may be ', etc. and frequently there will be 5 teeth on one side of the jaw and 4 on the other. The last tootli appears to be cut late in life. The teeth when cut are with one external and one internal fold, but soon simplify to a ring-pattern. Upper incisor roots extending into pterygoids. Palate excessively narrow, differing from that of Bathyergus, Cryptomys and Georychiis in that it does not e.xtend behind the toothrows. Infraorbital foramen very small. Other essential characters of skull as already described. Tail and ears obsolete. Claws not excessively lengthened. Hindfoot differ- ing from that of liatliycraus in that D.2 is the main digit rather than D.3, as in the forefoot, though the hallux remains slightly longer than D.5. Forms seen : albifrons, angoniciis, argenteocinereus, emini, kapiti, marungensis, robustus, spalax. List of Named Forms Mr. R. W. Hayman has been kind enough to look through this genus for me and reports that all the named forms "cannot in my view be more than races of argenteocinereus, except spalax, which has the narrow posterior nares reaching the level of last molars, and is distinguishable on this from all the others." I fully agree with this conclusion. 1. HELIOPHOBIUS SPALAX, Thomas 1910. Ann. Mag. Xat. Hist. 8, VI, p. 315. Taveta, near Kilimanjaro. 2. HELIOPHOBIUS .\RGENTEOCINEREUS .ARGENTEOCINEREUS, Peters 1852. Reise nach Mozambique, ZooL Saug. p. 140. Tette, Lower Zambesi. 3. HELIOPHOBIUS ARGENTEOCINEREUS ANGONICUS, Thomas 1917. Ann. Mag. Nat. Hist. 8, XX, p. 314. Bua River, Angoniland, East Rhodesia. 4. HELIOPHOBIUS ARGENTEOCINEREUS ROBUSTUS, Thomas 1906. Ann. Mag. Nat. Hist. 7, XVII, p. 179. Mpika, N.E. Rhodesia. 5. HELIOPHOBIUS ARGENTEOCINEREUS MARUNGENSIS, Noack 1887. Zool. Jahrb. Syst. II, p. 223, pi. ix, fig. 25. Marungu, South-east Congo. 6. HELIOPHOBIUS ARGENTEOCINEREUS EMINI, Noack 1894. Zool. Jahrb. Syst. VII, p. 559. Simba Mucnna, near Mpwapwa, Tanganyika. 7. HELIOPHOBIUS .\RGENTEOCINEREUS K.\PITI, Hillcr 1909. Smiths. Misc. Coll. LII, part 4, p. 469. Kapiti Plains, Kenya. 8. HELIOPHOBIUS ARGENTEOCINEREUS ALBIFRONS, Gray 1864. Proc. Zool. Soc. London, p. 123. "East .Africa." Synonym : ? pallidus. Gray, 1864, P.Z.S. London, 9. 124. "East Africa." 86 GEORYCHUS— CRYPTOMYS 9. HELIOPHOBIUS MOTTOULEI, Schouteden. (Not seen) 1913. Rev. Zool. Afr. 2, p. 203. Kilong\ve, near Lake Kisale, Belgian Congo. Genus 3. GEORYCHUS, Illiger 181 1. Georvchus, Illiger. Prodr. Syst. Mamni. p. 87. Type Species. — Miis capensis, Pallas. Range. — South Africa: Natal and Namaqualand to the Cape. Number of Forms. — Three. Characters. — Like Cryptomys, next to be described, but upper cheek- teeth with one narrow inner and outer fold each, the folds persisting; lower cheekteeth with one outer fold persistent and one inner fold which tends to become weak or obsolete. Posterior cheekteeth cut late in life. Upper incisor roots extending into pterygoids. Infraorbital foramen normal (small). Externally with no special peculiarities; claws not enlarged; the digits arranged about as in Bathyergus. Forms seen: caperisis, canescens. List of Named Forms 1. GEORYCHUS CAPENSIS CAPENSIS, Pallas 1779. Glires, pp. 76. 172, pi. VII. Cape Colony. Synonyms: buffoni. Cuvier. Ann. Sci. Nat. 1834, i, p. iq6. hucops, Lichtenstein, Forsters Desc. Arnrn. Iter. ad. Maris Aust. Teras Suscepto, p. 364, 1844. 2. GEORYCHCS CAPENSIS CANESCENS, Thomas & Schwann iqo6. Proc. Zool. See. London, p. 165. Knysna, South Cape Colony. 3. GEORYCHUS CAPENSIS Y.\TESI, Roberts IQ13. Ann. Transvaal Mus. IV, p. 92. Trans\aal. Genus 4. CRYPTOMYS, Gray 1864. Cryptoiuys, Gray, Proc. Zool. Soc. London, p. 124. 1S64. CoETOMYS, Gray. Proc. Zool. Soc. Loiidon, p. 125- (Based on coeciitwns and da?nayetisis). Type Species. — Georvchus holosericeus, Wagner. Range. — Africa, widely distributed: Togoland, Nigeria and Bahr-el-Ghazal to the Cape; evidently not occurring in Kenya, nor Abyssinia, nor vSomaliland. Number of Forms. — Approximately forty-nine have been named. Characters. — Skull, excepting in some species the infraorbital foramen, without special peculiarity; about as usual in the family; frontals not much constricted; mandible with angular portion not much pro- duced backwards. Upper incisors plain, their roots extending behind the CRYPTOMYS 87 toothrow. Cheekteeth }, a simple ring in adult, one inner, one outer fold when unworn. Claws normal; length of digits as Bathyergus, or with considerable tendency towards D.2 and D.3 in the manus being subequal, or even in some seen D.3 is very slightly the longer. D.2 and D.3 in the hindfoot also often subequal. Mammae usually 2 — i = 6 (Thomas). Tail shorter than hindfoot. Infraorbital foramen variable, sometimes even individually. In some cases, as in the giant species, tuellandi, it becomes as large relatively as in some Muroid rodents, as Rhizomys, and surely must transmit muscle. Skull no. 20.1 1.3.227 at the British Museum shows a specimen in this state. In C. coecutiens, as figured by Tullberg, a small strand of muscle passes through the foramen. The infraorbital foramen is normal (small), or but very slightly enlarged in damaremis, lugardi, beirae, zeclii (type not seen), molyneaiixi, micklemi,foxi, lechei, kummi, and tchytei (slightly enlarged); in the type of nimrodi, the infraorbital foramen of one side of the skull is small, on the other side slightly enlarged (which proves that no specific groups may be based on this character) ; it is very little enlarged in ansorgei (one of the giant mechoui group) ; moderate in blainei (same group); relatively large in bocagei, coecutiens (type not seen), darlingi, jorisseni; largest in amatus, hottentotus (type not seen), mellandi and mechowi (type not seen). The above notes, except when stated to the contrary, are based on type skulls; there may be some individual variation perhaps within some of the species. Four forms, mechowi, mellandi, ansorgei and blainei, separate rather sharply as a group from the others on account of their relatively very large size. Forms seen: amatus, ansorgei, beirae, blainei, bocagei, coecutiens, damarensis, darlingi, foxi, holosericeus, hottentotus, jorisseni, kummi, lechei, lugardi, mechowi, mellandi, micklemi, molyneauxi, nimrodi, talpoides, whytei, zechi. List of Named Forms Mr. R. W. Hayman has kindly looked through the large collection of the genus Cryptomys at the British Museum, with a view to getting it into some semblance of order, and reports as follows : "The British Museum material of this genus seems to me to be divisible into five groups, based primarily on presence or absence of white head-spot (this is more reliable than was expected), secondarily on colour and size. Cranial characters seem to be unreliable and cannot be correlated with the groupings given here. It is obvious that manv of the so-called species listed here will eventually be relegated to sub-specific rank. I. Without head-spot. (a) mechmci group. Large to very large, head and body 200 up to 260. \'ery pale brown in all forms. 1. mechou'i, Peters. North Angola. 2. mellandi, Thomas. North Rhodesia and Angola. 3. ansorgei, Thomas & Wroughton. Central Angola. 4. blainei, Hinton. Central .\ngola. 88 CRYPTOMYS (b) hotteritoius group. Small to medium-sized, head and body 100-150. Drab or lawn, exceptionally blackish (talpoides). 5. hottentotus, Lesson. Cape Colony and Natal. 6. /(. talpoides, Thomas & Schwann. Cape Colony. 7. h. occlusus, Allen & Loveridge. S.W. Tanganyika Terr. 8. Ii. zobytei, Thomas. N.W. Nyasa, N.E. Rhodesia. (Treated as a race of hottentotus by Allen & Loveridge, Bull. Mus. Comp. Zool. Harv., LXXV, No. 2, p. 125.) Fig. 4. Cryptomys d.\m.4Rensis, Ogilby. B.M. No. 25.12.4.190, $; y 2. 9. eoecutiens, Lichtenstein. Cape Colony. 10. holosericeus, Wagner. Cape Colony, Orange Free State, Transvaal. 11. jorisseni, Jameson. Transvaal. 12. nimrodi, de Winton. S. Rhodesia. 13. amatus, Wroughton. N. Rhodesia and Katanga. With head-spot. (a) lechei group. Size medium to large, head and body about 125 to 200. Colour ranging from blackish through seal-brown to slate and silvery-fawn. 14. lechei, Thomas. N.E. Congo, N. Uganda. (Only 3 Uganda skins seen, all lacked spot.) 15. kummi, Thomas. French Shari. 16. foxi, Thomas. North Nigeria. CRYPTOMYS 89 17. lugardi, dc Winton. S.W. Africa and Kalahari. 18. micklemi, Chubb. N.W. Rhodesia. (St. Leger, Proc. Zool. Soc. London, 1932, p. 964, considers micklemi lugardi.) 19. molyneauxi, Chubb. N.W. Rhodesia. 20. darlingi, Thomas. S. Rhodesia. 21. beirae, Thomas. Portuguese East Africa. Fig. 5. CRYPTONrys DA^L\RENSIS, Ogilby. B.M. No. 25.12.4.190, 9; X 2. (A) diimarensis group. Size medium; head and body about 150. Pale sandy brown. 22. damarensis, Ogilby. S.W. .\frica. 23. ochraccocinereus, Heuglin. Bahr-el-Ghazal, Sudan. 24. zechi, iVIatschie. Togoland, West Africa. Authentic examples of 23 {ochraceocinereus) not seen. This is periiaps a more artificial group than the preceding one, but Nos. 22, 23 and 24 do not seem to fit elsewhere. 3. With or without head-spot. bocagei group. Colour cinnamon to drab; size small to medium, up to 150. Head spot very variable. 25. bocagei, de Winton. Angola. 26. kubtingensis, IVIonard. Angola." On account of the variability of the head-spot I think it will be desirable to treat the hottentotus, lechi, damarensis and bocagei groups as sections of one specific group, particularly bearing in mind the amount of variability met with in the genus lleliupliobius in this character. I must add that it was I who origin- ally suggested to Mr. Hayman that the presence or absence of this spot might be used to divide Cryptomys into groups. go CRYPTOMYS The mechowi group is unquestionably very distinct from the remainder. mechoioi Group 1. CRYPTOMYS MECHOWI, Piters 1881. Sitz. Ber. Ges. Nat. Fr. Berlin, p. 133. Malanje, North Angola. 2. CRYPTOMYS MliLLANDI, Thomas igo6. Ann. Mag. Nat. Hist. 7, XVII, p. 178. Mpika, N.-E. Rliodesia. 3. CRYPTOMYS .YNSORGEl, Thomas & Wroughtoii 1905. Ann. Mag. Nat. Hist. 7, XVI, p. 175. Bihe, Central .\ngola. 4. CRYPTOMYS BLAINEI, Hiiit.)n 1921. Ann. Mag. Nat. Hist, g, VII, p. 372. Loando River, Central .\ngola. hottentotus Group (typical section) 5. CRYPTOMYS HOTTENTOTUS HOTTENTOTCS, Lesson 1826. Voy. Coq. Zool. i, p. 166, pi. ii, fig. 2. Paarl, Cape. 6. CRYPTOMYS HOTTENTOTUS T.-\LPOIDES, Thomas & Schwann igo6. Proc. Zool. Soc. London, p. 166. Knysna, Cape Colony. 7. CRYPTOMYS HOTTENTOTUS OCCLUSUS, Allen & Loveridge 1933. Bull. Mus. Comp. Zool. LXXV, no. 2, p. 125. Uzung%ve Mountains, S.-W. Tanganyika. 8. CRYPTOMYS HOTTENTOTUS WHYTEI, Thomas 1897. Proc. Zool. Soc. London, p. 432. Karonga, Lake Nyasa. g. CRYPTOMYS COECUTIENS, Lichtenstein 1827. Brants. Muiz. p. 37. Natal. Synonym: hiilnigi, Smitli, 1829, Zool. Journ., p. 439. A synonym of C. hottentotus, fide G. M. Allen. 10. CRYPTOMYS HOLOSERICEUS, Wagner 1842. Schreb. Saugt. Suppl. Ill, p. 373. Graaf Reinet, Cape Colony. 11. CRYPTOMYS NIMRODI, de Winton i8g6. Proc. Zool. Soc. London, p. 808. Bulawayo, Rhodesia. 12. CRYPTOMYS .\MATUS, Wroughton 1907. Manchester Mem. 51, no. 5, p. 28. .Alala Plateau, North Rhodesia. 13. CRYPTOMYS JORISSENI, Jameson igog. Ann. Mag. Nat. Hist. 8, IV, p. 466. Waynek, Waterhurg District, Transvaal. CRYPTOMYS 9" (lechei section) 14. CRYPTOMYS I.ECHEI, Thomas 1895. Ann. Mag. Nat. Hist. 6, XVI, p. 241. Bellima, Monbuttu, N.E. Congo. 15. CRYPTOMYS KUMMI, Thomas igii. Ann. Mag. Nat. Hist. 8, VH, p. 592. French Shari Protectorate, Ironside Plateau, about 8° N. 22' E. 16. CRYPTOMYS FOXI, Thomas 1911. Ann. Mag. Nat. Hist. 8, VII, p. 462. Panyani, North Nigeria. 17. CRYPTOMYS LUG.U^DI, de Winton 1898. Ann. Mag. Nat. Hist. 7, I, p. 253. Kalahari, between Palapye and Ngami. Synonym: micklemi, Chubb, 1909, Ann. Mag. Nat. Hist. 8, III, p. 35. Upper Zambesi. i8. CRYPTOMY'S MOLYNEAUXI, Chutb 1908. Ann. Mag. Nat. Hist. 8, II, p. 451. Loano Valley, N.-\V. FLhodesia. IQ. CRYPTOMYS DARLINGI, Thomas 1895. Ann. Mag. Nat. Hist. 6, XVI, p. 239. Salisbury, Rhodesia. 20. CRYPTOMYS BEIR.\E, Thomas & Wroughton 1907. Proc. Zool. Soc. London, p. 780. Beira, Portuguese East Africa. {damarensis section) 21. CRYPTOMYS DAMARENSIS, Ogilby 1838. Proc. Zool. Soc. London, p. 5. Damaraland. 22. CRYPTOMYS ZECm, Matschic 1900. Sitz. Ber. Ges. Nat. Fr. Berlin, no. 4, p. 146. Middle Volta, Togohind. 23. CRYPTOMYS OCHI^ACECJCINEREUS, Heuglin 1864. Nov. Act. Ak. Caes. Leop. Dresden, XXXI, p. 3. Bahr-el-Ghazal, Sudan. (bocagei section) 24. CRYPTOMYS BOCAGEI, de Winton 1897. Ann. Mag. Nat. Hist. 6, XX, p. 323. Hanha, Angola. 25. CRY'PTOMYS KLB.VNGENSIS, Monard 1933. Bull. Soc. Neuchatel. Sci. Nat. 57, p. 58. Cubangu River, Mossamedes, -Angola. There then remain to be discussed twenty-three "species" (?) of Roberts. Some comments on some of these have already been made by Oldfield Thomas, Ann. Mag. Nat. Hist. 8, XX, p. 444, 1917. 92 CRVPTOMYS It is useless attempting any remarks on these, as all are unrepresented; they are therefore listed alphabetically. 2b. CRYPTOMYS ABERUANS. Roberts 1913. Ann. Transv. Mus. IV, p. 98. Port St. Johns, Cape Province. -7. CRYPTOMYS ALDUS, Roberts 1913. .Ann. Transv. Mus. IV, p. 100. Wynberg, Cape Colony. 28. CRYPTOMYS ANOMALUS, Roberts 1913. Ann. Transv. Mus. IV, p. 96. Transvaal, Pretoria. 2v. CRYPTOMYS AREN.ARIUS, Roberts 1913. .Ann. Transv. Mus. IV, p. 96. Transvaal, Pretoria. 30. CRYPTOMYS BIG.\LKEI, Roberts 1924. Ann. Transv. Mus. X, p. 73. Glen, Orange Free State. 31. CRYPTO\n'S CRADOCKENSIS, Roberts 1924. Ann. Transv. Mus. X, p. 73. Cradock, Cape Province. 32. CRYPTOMYS JAMESONI, Roberts 1913. .Ann. Transv. Mus. IV, p. 95. Trans\'aal, Johannesburg. 33. CRYPTOMYS JUNODI, Roberts 1926. .Ann. Transv. Mus. XI, p. 260. Masiene, Portuguese East Africa. 34. CRYPTOMYS KOMATIENSIS, Roberts 191 7. .Ann. Transv. Mus. V, p. 272. Amhemburg, Transvaal. 35. CRYPTOMYS LANGI, Roberts 1929. .Ann. Transv. Mus. XIII, p. 119. Keerkloof, Natal. 36. CRYPTOMYS MAHALI, Roberts 1913. .Ann. Transv. Mus. IV, p. loS. Transvaal. 37. CRYPTOMYS MELANOTICUS. Roberts 1926. .Ann. Transv. Mus. XI, p. 260. Makoetsi River, N.E. Transvaal. 38. CRYPTOMYS MONT.ANUS, Roberts 1926. Ann. Transv. Mus. XI, p. 260. Klapperklop, Pretoria, Transvaal. 39. CRYPTOMYS NATALENSIS, Roberts 1913. Ann. Transv. Mus. IV, p. 94. Natal, Wakkcrstroom, Transvaal. CRYPTOMYS 93 40. CRYPTOMYS ORANGIAE, Roberts 1926. Ann. Transv. Mus. XI, p. 259. (Jlcn, Orange Free State. 41. CRYPTOMYS PAI.KI, Roberts 1917. Ann. Transv. Mus. VI, p. $■ Vaal River, Transvaal. 42. CRYPTOMYS PRKTORIAE, Rolierts 1913. .Ann. Transv. Mus. IV, p. 99. Transvaal, Pretoria. 43. CRYPTOMYS RUFULUS, Roberts 1917. Ann. Transv. Mus. V, p. 272. Tzaneen, Transvaal. 44. CRYPTOMYS STELLATUS, Roberts 1917. Ann. Transv. Mus. V, p. 272. Komatipoort, Transvaal. 45. CRYPTOMY'S TR.ANSVAALENS1S, Roberts 1924. .Ann. Transv. Mus. X, p. 73. Pretoria district. 46. CRYPTOMYS VANDAiMI, Roberts 1917. Ann. Transv. Mus. V, p. 273. Leydsdorp, Transvaal. 47. CRYPTOMYS VETENSIS, Roberts 1926. Ann. Transv. Mus. XI, p. 259. Vet River, Orange Free State. 48. CRYPTOMYS VRYBURGENSIS, Roberts 191 7. Ann. Transv. Mus. V, p. 274. Vryburg, British Bechuanaland. Addendum ; CRYPTOMYS NATALENSIS NEMO, G. .M. Allen. 1939. Bull. Mus. Comp. Zool. LXXXIII, p. 429. Manetsi River, near Malala, Zoutspansberg district, Transvaal. Synon>Tn: pallidus, Roberts. 1917, Ann. Trans. Mus. V, p. 278. Not of Gray. The Heterocephaliis Group Cheekteeth ij or |, simplified in adult. Size smaller than in other members of the family. Fur practically absent, the hairs occurring singly, scattered, throughout the body, most developed on the feet. Tail longer than hindfoot. Eyes and ears extremely small, no ear conch. Forefoot with five digits, the centre of which is the longest. D.5 and especially the pollex shorter than D.4 and D.2, which are subequal. Hindfoot like forefoot, but hallux about as long as D.5. Essential cranial characters as in Bathyergus group, but jugal reduced; more Murine in appearance; palate not continuing behind molars, and in appearance rather less constricted normally than in other genera; upper incisors extending behind the toothrows. 94 HETEROCEPHALUS Genus 5. HETEROCEPHALUS, Ruppell 1S42. Heterocephalis, Ruppell, Mus. Scnckenberg. Abh. 3, 1903. FoRNARiNA, Thomas, Proc. Zool. Soc. London, p. phillipsi, Thomas.) Type Species. — Heterocephalus glaber, Ruppell. R.^NGE. — Known from Abyssinia, Somaliland, Kenya. Number of Forms. — Four are here listed. Heft 2, p. 99. 336. {Heterocephalus Fig. 6. Heterocephalus glaber glaber, Ruppell. B.M. No. 32.2.19.9, ?; > 2h- CH.'UiACTERS. — As indicated above. Frontals little constricted; nasals appear rather broader than in other genera; palate shorter than in other genera except Ihliophobiiis. Cheekteeth simple in adult, the usual folds found elsewhere in the family present when unworn ; normally ij; in H. phillipsi, so far as known, reduced to f , evidently at a certain age or stage of wear; this species is represented in the British Museum only by three skulls; two of these have two upper teeth on each side, the third has three upper cheek- teeth on one side, the posterior one minute, two on the other side, the posterior one apparently having been shed. This species was made the type of a genus "Foriiaiina" by Thomas, but much more evidence on the condition of this form is required before any generic separation can be done; I should be HETEROCEPHALUS 9S quite content to assume that if enough specimens could be brought to hand phillipsi would turn out to be no more than a race of H. glaber^ In the few available for examination, including skulls which have been made types of two or three "species," there is much variation in the size of the cheek- teeth, M.3 being in some only slightly, and some very considerably smaller than M.2, and of the incisors, which reach their maximum size in the type of dunni. There is also variation in the form of the coronoid. Hollister, 1919, East African Mammals in the U.S. National Museum, synonymizcs several forms with the typical race; this classification is here followed. Provisionally I list all named forms as either synonyms or races of the type. Forms seen : ansorgei, glaber, dunni, phillipsi. Fig. 7. Heteroceph.\ll"s glaber glaber, Ruppell. B.M. No. 32.2.19.9, 9; X 3 J. List of Named Forms I. HETEROCEPHALUS GLABER GL.A.BER, RUppell 1842. Mus. Senckenberg, .Abb. 3, Heft 2, p. 99. Shoa, .Abyssinia. Synonj-m: glaber progrediens, LSnnbcrg, 191 1, Kungl. Sv. Vet. Akad. Handl.Bd.48,no.5,p. 102. North of GuasoNyiro, Kenya. ansorgei, Thomas, 1903, Proc. Zool. Soc. London, p. 336. Makindu district, Kenya. stygius, .Allien, 1912, Bull. Mus. Comp. Zool. LIV, p. 444. Neumann's Boma, Nth. Guaso Nyiro, Kenya. * Since the above was written, the Check List of .African Mammals of G. ^L Allen has been published ; in this H. phillipsi is considered a synonym of H. glaber. 96 HYSTRICOIDAE 2. HETEROCKPHALLIS GI.ABICR SCORTICCCII, de Heaux 1934. Atti. Soc. Ital, Sci. Nat. LXXIII, p. 2S3. Gardo, Italian Somaliland. (A synonym of o'. glaber, according to G. M. Allen) 3. HE'l'EROCEPHALUS GLABER DUNNI, Thomas ujoq. Ann. Mag. Nat. Hist. S, IV, p. log. Wardairi, Central Somaliland. (A synonym of ^. glaber, according to G. M. Allen) 4. HETEROCEPHALUS GLABER PHILLIPS!, Thomas iSSj. Proc. Zool. Soc. London, p. 612. Somaliland. {A synonym of ,i^. glaher, according to G. M. Allen) The references and type localities for all members of the family Bathyer- gidae are the work of Mr. R. W. I layman. The family have been described fossil from the Oligocene of Mongolia. This indicates a former wide distribution for the group, and contrasts with some of the other African families as Anomaluridae and Pedetidae which do not seem to be known outside the Continent. BATHYERGIDAE: SPECIAL WORKS OF REFERENCE TuLLBERG, Nova .Acta Reg. Soc. Sci. Upsaliensis. XVIII, ser. 3, no. i, 1899. HoLLiSTER, .Smiths. Inst. Bull. 99, p. 159, 1919; East African Mammals in the U.S. Nat. Mus. Note on status of some forms of Heterocephalus. St. Leger, Key to Families and Genera of African Rodentia, Proc. Zool. Soc. London, 1 93 1, p. 976. Thomas, Ann. Mag. Nat. Hist. ser. 8, vol. IV, p. 1 10, IQ09. Note on dental formula in the family. Superfamily HYSTRICOIDAE This group is equal in rank to the "lateralis-scries" of the superfamily Hystricoidae of Miller & Gidley; or to the "Hystricomorpha" of authors not including Caviidae (Cavia, Galea, Caviella, Kerodoii, Dolichatis, Ilydrochomis) nor Ctenodactylidae; it is here divided into seven families. 1896. Thomas: Hystricomorpha, part, included Pedetidae, Caviidae, Ctenodactylidae. 1899. TuUberg: HvsTRicoGNATHi : Hystricomorpha; part, included Caviidae. 1918. Miller & Gidley: Superfamily Hystricoid.'\e, part, lateralis series. (Included, as medialis series, the Caviidae.) 1924. Winge : Family Hystricidae, part, included IJathyergidae, Ctenodactylidae and Caviidae of this work. 1928. Weber: HYSTRICOIDAE, part; included Caviidae, Ctenodactylidae. Geographical Distribution. — The greater part of the American Con- tinent from Canada to Patagonia (evidently absent only from certain areas of southern U.S.A.); the greater part of the African Continent; the Indo-Malayan region, from the Himalayas to Ceylon and from Southern China to Borneo and the Philippine Islands; represented in the Palaearctic in Italy, coastal regions North-west Africa, and in South- western Asia (north into southern Siberia). HYSTRICOIDAE 97 Characters. — Zygomasseteric structure differing from that of all members of the order, except Bathyergidae, in that the lower jaw, paralleling the Bathyergidae, has the angular portion of the mandible distorted outwards, to a greater or lesser degree, "to allow passage of a specialized and enlarged distal anterior limb of masseter lateralis superficialis, its general direction parallel with zygoma" (Miller & Gidlev); combined with the fact that the infraorbital foramen is very much enlarged to allow passage of masseter medialis; the zygomatic plate is narrow, and remaining completely below it, the general arrangement of the forepart of the skull as regards muscle insertion (infraorbital foramen, zygomatic plate), essentially as in Anomaluridae, Cteno- dactylidae, Pedetidae and Dipodidae. Skull normally specialized, with broad frontals, which rarely show much signs of interorbital constriction, a tendency present towards complexity of zygoma (Echimyidae), and lengthening and specialization of paroccipital processes. Dental formula: i. — c. -> p. -> m. - ^ 20. I o "^ I 3 Cheekteeth usually flatcrowned, usually hypsodont, often evergrowing, not cuspidate in adult. A tendency present towards reduction of the digits of the hindfoot (in some forms, Dasyproctidae, Lagostomits, hindfoot with three digits only). The malleus and incus are fused according to Tullberg, though in some cases apparently not completely so. The tibia and fibula remain distinct, or are not fully fused. Remarks. — This group has been recognized as one of the major groups of the order by all authorities. But many forms which appear to me not to belong are currently included in it. The Hystricomorpha or Hystri- coid series are always described as with the angular process distorted outwards, as indicated above; if this is a sufficiently important character on which to base superfamily grouping, and it apparently is so (Tullberg divided the whole order into two great groups, Hvstricognathi and Sciurognathi, based on its presence or absence), it seems clear that forms which do not agree in mandible structure with typical Hystricoidae must be excluded from that superfamily, no matter what their ancestors may have been. This takes the Caviidae into another branch of the order, as thev cannot by the longest stretch of imagination be regarded as with typically Hystricoid mandible formation. The close associa- tion ot Caviidae with such forms as Dasyprocta and Cunicuhis by manv authors, Tullberg among them, has long struck me as extremely unnatural; Cuniculus is of course one of the most isolated and aberrantly specialized living rodents, and has not even the feet structure and external specialization of Caviidae and Dasyprocta; but the last two named, both clearly parallels in evolution, both highly modified for cursorial life, with digits of hindfoot reduced to three, are yet so clearly totallv different when lower jaw structure and dental structure are looked into. Such an association appears as unnatural to me as dumping Castor and Myocastur into a family together because both swim ! Of the forty-three genera belonging to the group, thirty-six are confined to the neotropical region. One is peculiar to North America ; of the remainder, two 7 — Living Rodents — I 98 HYSTRICOIDAE are African, two (Hystricidae) the Malay Islands, the remaining two (Hystricidae) cover a wide area in the southern palaearctic, the African and the Indo-malayan regions. This group, together with the Cavioidae, contains all the giants of the order (except Castor), and exhibits some extreme tvpes of external specialization. Taking all their characters into consideration, it appears to me that this group is without doubt, broadly speaking, the most highly specialized and progressive branch of the order, particularly such forms as Ilvstrix cristata, though in all cases such species grade down quicklv to relatively low primitive allies. The division of the group into families is not easy. As many as thirteen have been recognized by various authors. It appears to me to be both unneces- sary and inconvenient to divide them into as many as this; particularly taking into account that all seem closely allied to each other, and that elsewhere vast groups like the Muridae (with Cricetinae, Microtinae, Gerbillinae, Myospa- lacinae, etc., etc.), are usually retained as one family. Flower & Lydekker, 1891, Mammals Living and Extinct, recognized (of Hystricoidae as here under- stood) only five families, the Octodontidae (=Echimyidae of this paper), Hystricidae, Chinchillidae, Dasyproctidae and Dinomyidae. Thomas in 1896 very properly separated the American "Porcupines" from the Hystricidae as a distinct family; and Miller & Gidley, Pocock and others have recognized Cuniculiis as type of a distinct family. With these two modifications I retain the classification of Flower & Lydekker. The Echimyidae as here understood contain a large quantity of neotropical rodents and two African genera ( Thrvoiioinys, Petroinm) ; broadly speaking this group contains forms which have not become much modified externally (excep- tions: Mvocastor, Tlirvoiioinvs, Dactv/oiiivinae, part), but which seem to have their specialization in the skull characters (lengthened or specialized paroccipital process, some tendency to enlargement of bullae, very general tendency for complex zygoma, etc.); and in dental characters (such as the rootless simple teeth in Octodontinae). Normally the size is relatively small, though Mvocastor and Tlii'vonomxs provide exceptions to this. The relationships of the various groups will be discussed below. Apart from the evergrowing plain laminate cheekteeth, there is little to dis- tinguish the Dinomyidae from them; and certain species of Echimyidae mav attain laminate cheekteeth, though this not combined with extreme hypsodonty. But Dinomys is evidently rather an isolated specialized type, and is best retained as a family. The Chinchillidae are dentally like Dinomys, but cranially with their poorly-ridged mandible, the tendency either to lengthening of paroccipital process or extreme inflation of mastoids and bullae, they stand rather apart not onlv from Dinomvs but as a group from all other Hystricoids apparently; the functional digits of the hindfoot are reduced to three. The lirethizontidae are more primitive than the Echimyidae cranially and dentally (as a rule), but very much more highly specialized externally, the feet attaining arboreal specializa- tion not seen elsewhere in the order, and the spiny covering of the body being in a very different class from that of the few spiny members of the Echimyidae, in all but the very lowest. The Hystricidae are held not to be closely related HYSTRICOIDAE 99 to the Erethizontidae, but rather to Dasyproctidae ; once again their lower members are less specialized than in Echimyidae, as regards cranial char- acters, their higher ones very much more so, with tendency to extreme inflation of nasals without parallel in living rodents. The external covering presents extreme specialization in development of spiny covering, hut goes through an interesting series of grades of development, so that the lowest is much less specialized in spiny covering than the higher members. The Dasyproctidae are very similar to the Ilystricidae in cranial and dental characters; externally they are very different, and not less specialized in their way, being modified for cursorial life, with three digits only to the hindfoot, and in external form calling to mind a type that primitive ancestral ungulates must have at one time passed through in their evolutionary history. Finally the Cuniculidae in the develop- ment of their vast bony cheek-plate present cranial characters very widely different from any other rodent. Key to thk Families of Hystricoidae Entire zygomatic region abnormally modified by growth of bony cheek- plates. (Cheekteeth strongly hypsodont, the folds of the teeth isolating as islands on crown surface; form heavy; digits not numerically reduced on hindfoot, the claws thick, more or less hoof-like.) Family Cuniculidae Zygomatic region not abnormal, always without bony cheek-plates. Cheekteeth evergrowing, or extremely hypsodont, the pattern one of a series of transverse plates. Mandible with angular process strongly distorted outwards ; incisors powerful; form heavy, limbs not lengthened; hindfoot with four well-developed digits; no tendency present for exces- sive enlargement of bullae and mastoids nor for lengthening of paroccipital process. Family Dino.myidae Mandible with angular process rather weakly distorted outwards; incisors relatively thin or medium; form more slender, limbs to a degree lengthened; hindfoot with three well- developed digits; a marked tendency present either to excessive enlargement of bullae and mastoids, or to lengthening of paroccipital process. Family Chinchillidae Cheekteeth when evergrowing or extremely hypsodont never with pattern of a series of transverse plates. Hindfeet excessivelv specialized for arboreal life, or becoming so; the hallux being replaced by a broad movable pad. (Fur conspicuously spinous, the spines short; bullae rather large, prominent; cheekteeth either with very wide re-entrant folds, or nearly laminate in structure, rooted; no lengthening of paroccipital process; zygoma simple.) Familv ERETHIZONTIDAE loo HYSTRICOIDAE llindteet never excessively modified for arboreal lite, hallux never replaced by broad movable pad, and never suppressed in arboreal genera. External form modified for cursorial life, the digits of the hind- foot reduced to three, the claws thick, more or less hoof- like. (Cheekteeth semi-rooted, the re-entrant folds isola- ting as narrow islands on crown surface in adult; fur not developing spines.) Family D.\syproctid.\e External form never modified for cursorial life; digits of hind- foot never reduced to three; claws not hoof-like in struc- ture. Externally showing a progressive series of modification of fur into spiny covering, at extreme development highly specialized (the spines long); bullae relatively small; paroccipital process not specially lengthened nor modified; zygoma simple. (Cheekteeth semi-rooted or rooted, the re-entrant folds isolating as narrow islands on crown surface in adult; a tendency present towards extreme inflation of nasals in progressive species; form heavy, terrestrial.) Family Hystricidae Externally without extreme modifications (in one case to a degree specialized for aquatic life); spiny covering of body when present relatively weak (as compared w ith Hystricidae) ; bullae relatively large ; paroccipital process enlarged, either curving forward under the bullae or lengthened, tending to stand apart from them; zygoma very generally with upwardly directed process on posterior border, or downwanilv directed process on posterior border, or both. (Cheekteeth various; sometimes evergrowing, when evergrowing most often approaching or reaching complete simpli- fication of pattern ; w^hen with a pattern of islands isolating on crown surface in adult, usually brachv- odont.) Family I'Iciiimyidae These families it will be noticed are based chiefly on the external characters. So many fossils are known belonging to this group that care must be taken if defining the families on cranial and dental characters, as it may be that in many cases fossil forms will prove intermediate between certain groups, or break these characters down. Also the cranial characters of these Flystricoids are, generally speaking, so similar that if not known living, all except Ciiiiiculus could readily be referred to one family. The extreme external specializations reached by some members of this group are in my opinion just as important as any cranial or dental character. ECHIMYIDAE loi Family ECHIMYIDAE i8q6. Thomas: Hystricomorfha, part, Family Octodontidae, part (included Cteno- dactylidae) ; subfamilies Echimyinae (included Dactylomyinae), Capromyinae {Capromys, Plagiodontia, Myocaslor, Thryonomys), Octodontinae (included Abrocoma), Ctenodactylinae, part {Petromus). 1899. Tullberg: Hystricomorfha, part, Family Echinomyidae ; subfamily "Myopo- tamini" (= Myocastorinae) ; subfamily Echinomyini (groups, Echinomyes, Echimyinae and Dactylomyinae as here understood) ; and Octodontes (Octo- dontinae and Abrocominae as here understood). Family Aulacodidae ( = Thryo- nomyidae). Family Pctromyidae. igi8. Miller & Gidley. Superfamily Hystricoidae, part; Family Echimyidae, part (included Chaetomys); Subfamily Echimyinae ("Spiny-Rats, Hutias," etc.); subfamily Octodontinae. Family Petromyidae. Family Myocastoridae. Family Thryonomyidae. Family Abrocomidae. 1924. Winge: Family Hystricidae, part; Capromyini (Capromys, Plagiodontia, Myocastor, Thryonomys); Octodontini, groups Octodontes (Octodontinae and Abrocominae as here understood), Echinomyes (Dactylomyinae and Echimyinae as here understood); Ctenodactylini, part (Petromus). 1928. Weber: Hystricoidea, part. Family Capromyidae (Capromys, Plagiodontia, Myocastor); Family Octodontidae (Ecliimys, Octodon, Ctenomys); Family Cteno- dactylidae, part (Petromus) ; Family Thr\onomyidae. Geographic.\l Distribution. — Neotropical region, from Nicaragua south- wards to Patagonia; Cuba and the West Indies; Africa widel)- distributed south of the Sahara. NiMBER OF Genera. — As here understood the family contains twenty-eight genera, one of which, Procapromys, has not been examined and is not represented in London. Char.\cters. — Zygomasseteric structure typically Hystricoid in formation. Cheekteeth when evergrowing never a series of transverse plates (compare Chinchillidae, Dinomyidae); feet never abnormally modified for arboreal life (compare Erethizontidae); zygomatic region without bony cheek-plate (compare Cuniculidae); external form never modified for cursorial life, digits of hindfoot more than three (always five except four in Thr\onom\s) (compare Dasyproctidae); bullae prominent, and paroccipital process length- ened, and zvgoma usually more angular, also tail never with specialized quills or bristles in spiny genera, and spinv covering when present usually not highly developed (compare Hystricidae). As thus defined the group includes the great central mass of genera of Hystri- coid rodents which have not become abnormally specialized in any external particulars. The cheekteeth may be evergrowing (.\brocominae, Octodontinae, Plagiodontinae, Capromvinae), extremely hypsodont (Myocastorinae, Petro- myinae), or moderately so but rooted (the remainder). In the Octodontinae, the structure approaches complete simplification of pattern; in the Abrocominae the upper cheekteeth are simplified, but the lower series remains complex; certain simplification has taken place in Plagiodontinae, which appears unique as regards dental characters, and in Petromyinae. These subfamilies have one external fold only in the upper cheekteeth; all other subfamilies have 102 ECHIMYIDAE more than one, typically three. The skull is normally remarkable for the par- occipital processes, which may be extremely lengthened', as in Myocastorinae, moderately so and standing apart from the bullae (Caprominyae, Plagiodontinae, Thryonomvinae), or curved forwards to a greater or lesser degree under the bullae (Echimvinae, Petromvinae, Abrocominae, Octodontinae). This fact has led some authors to form two families, Capromyidae and Octodontidae, but in the Dactylomyinae, as proved by British Museum material, either condition may exist; Kannabateomxs and Tlirinacudus, and some specimens of Dactxlomxs agree with the Echimyinae, but some large skulls of Dactxlomxs are quite indistinguish- able in paroccipital structure from Capromyinae, in which group there is also some variation apparently. The zygoma is usually rather broad, and frequently of a complex type, with a downwardly directed process on posterior lower border, and sometimes an upwardly directed one present above also. The bullae may be much inflated, as in Abrocominae, Octodontinae, Petromvinae; they are relatively large as a general rule. The external form varies; in Myocastorinae, which contain relatively very large forms, it is modified for aquatic life, with enlarged hindfeet, most of the digits of which are webbed; this group has also bullae, which recall the type found in Castoridae, though less specialized than in that family. Elsewhere the genera are not aquatic. Some genera of Octodontinae are remarkable as being the only Hystricoids which have taken to a subfossorial life {Ctenomys, Spalacopiis, Acouaemxs); Spalacopus must be about the smallest living Hystri- coid genus. A tendency to develop spiny covering, most pronounced in Mesomxs, Lonc/iothrix and Hoplomxs, is present in some of the Echimvinae; the spiny covering is, broadly speaking, very rudimentary compared with Hystri- cidae ( Tn'clixs perhaps excepted), and even Erethizontidae. The Dactylomyinae present a curious feature in that except in Tluiniicodiis the third and fourth digits of fore- and hindfeet are much elongated; these animals are said to be arboreal in habit. In all the genera of the family four functional digits in manus and five functional digits in pes are present excepting Thryono- myinae, in which the manus has three main digits only, D.5 being excessively shortened, though bearing a thick claw, and the hallux is entirely suppressed; this group possesses an abnormally massive and heavily ridged skull, perhaps more so than in any other living rodent, and extremely thickened three-grooved upper incisors. The form is heavy, and of terrestrial or slightly fossorial type. I provisionally divide the group into nme subfamilies, most of which have at some time or other been regarded as distinct families. But if a vast group like the Muridae are kept together as one family, and the Hystricoid branch is divided up into about seventeen families, the classification of the Order does not appear very consistent. The simple-toothed Octodontinae are connected with the main branch by such forms as Abrocoma and Plai;iodonlia\ the African genera appear to possess no essential characters which will keep them apart as families distinct. Petromus has of late been associated with the Ctenodactylidae by some authors, but the typical Hystricoid mandible of Petromus differs very widely from that of Ctenodactylidae, which are fully discussed elsewhere and ECHIMYIDAE 103 which are here considered not related to the Hystricoid group. Thryonomys is undoubtedly a very distinct type, but the only character which seems valid to keep it apart as a family is the formation of the shoulder-blade, which, accord- ing to Tullberg, differs considerably from that of other Echimyidae; but un- fortunately this character cannot be examined throughout the genera at the British Museum, Petromus, for one (one of the most important genera), being not represented so far as this character is concerned; so that until the skeleton of all the genera here included can be examined it seems more reasonable not to separate any group on this structure alone. The digits of Thryonomys are reduced, but this is an acquired character. Myocastor, which is sometimes made the type of a family, is undoubtedly a highly specialized form, but aquatic specialization alone is not sufficient to base family groups on unless accompanied by some definite cranial or dental characters (compare, for instance, other families, many of which have aquatic offshoots beside normal generalized types, as Hydromyinae, Cricetinae, etc.); and the structure of the paroccipital process, though highly specialized in Myocastor, is too variable elsewhere in the group. Miller & Gidley divided the Hystricoid lateralis series (^Hystricoidae as here understood) into two groups based on the lachrymal canal, which is stated to be closed in one branch, but open in the other branch on the side of the ros- trum. Abrocoma as thus defined comes in a different group from the remainder of those here. But this character seems not too constant elsewhere, for instance both conditions are to be found in the Caviidae of Miller & Gidley, and I do not attach too great importance to this character. The family as here under- stood in fact is the family Octodontidae of Flower & Lydekker, and earlier authors, except that of course the Ctenodactylidae are removed. The subfamilies must be regarded as provisional; though easily recognizable in living genera, it may be that among the large number of neotropical Hvstri- coid fossil rodents some forms would be found which are intermediate between some of the groups in the characters here noted. Key to the Subf.^milies of Echimyidae Cheekteeth becoming strongly simplified, the outer side of the upper series with not more than one re-entrant fold. The lower cheekteeth prismatic and angular in appearance; the upper series eight-shaped. Part of lachrymal canal open on side of rostrum. (Cheekteeth evergrowing; bullae much inflated; zygoma more or less simple; form terrestrial; digits of hindfoot five; skull not heavily ridged for muscle attachment.) Subfamily Abrocomin.^e [Abrocoma) The lower cheekteeth not essentially different in pattern from the upper scries; no part ot lachrymal canal open on side of rostrum. Cheekteeth rooted, the pattern ultimately wearing out; inner side of upper and outer side of lower teeth elevated. (Bullae I04 ECHIMYIDAE much inflated, the paroccipital process joining them; digits of hindtoot five; zygoma more or less simple; skull not heavily ridged for muscle attachment; form terrestrial, tail fully haired.) Subfamily Petromyinae (Petronnis) Cheekteeth evergrowing, the pattern not or scarcely changing during lite. Inner side of upper and outer side of lower teeth not elevated. Folds of cheekteeth very deep and long, set obliquely; each upper tooth with a well-marked outwardly pointing projection on external side. (Paroccipital process length- ened, standing apart from bullae, which arc not extremely inflated; zygoma simple; form generalized, tail naked; skull not heavily ridged for muscle attachment.) Subfamily Plagiodontinae {Plagiodontia) Folds of cheekteeth not specially deep and long, set less obliquely, or not so. Upper cheekteeth either eight- shaped, or "kidney-shaped." (Bullae normally much inflated, the paroccipital process curved forwards under them to a greater or lesser degree, and joining them; zygoma complex, always with an upwardly pointing process on posterior border; skull normally not heavily ridged for muscle attachment (sagittal crest if present weak); external form terrestrial, or modified for sub- fossorial life; digits of hindfoot five.) Subfamily Octodontinae (Oitoinvs, Octoilontoiiivs, Ociodon, Acoiiaemvs ; SpdlcHopiis ; Ctenoinvs) Cheekteeth less simplified, the outer side of the upper molars with at least two, typically three, re-entrant folds. Externally considerably modified for acjuatic life (bodily size largest of family); paroccipital process much elongated; cheekteeth extremely hypsodont, but not evergrowing, with well-marked re-entrant folds which are long retained, the teeth decreasing in size markedly from M.3 forwards; skull considerably ridged for muscle attachment (palate con- stricted anteriorly; digits of hindtoot five; zygoma complex). Subfamily Myocastorinae (Mvocdstor) External form never modified for aquatic life; paroccipital process less or not elongated; cheekteeth less or not decreasing in size from M.3 forwards. ECHIMVIUAE 105 Skull massive, abnormally heavily ridged for muscle attach- ment; incisors thick, the upper ones heavily three- grooved; forefoot with three functional digits (D.5 clawed but extremely shortened); hindfoot with four digits only; cheekteeth rooted; external form heavy, terrestrial-tossorial (bullae not much inflated, paroccipital process lengthened to a degree, and standing apart from bullae; zygoma much broadened, jugal nearly reaching lachrymal). (Shoulder-blade differing from American genera (Tullberg).) [Subfamily Thryonomyinae ( Tliryonomys) Skull much less heavily ridged for muscle attachment ; incisors not three-grooved (plain except in Carterodon); forefoot ahvays with four and hindfoot always with five functional digits; jugal usually complex, with upwardly or down- wardly projecting process present on posterior border, and normally not approaching lachrymal. Cheekteeth extremelv broadened, the re-entrant folds deep and persistent, the teeth rooted; palate with a ten- dency towards anterior constriction; paroccipital processes tending to stand apart from bullae in larger species, or curved forwards under them in smaller forms; tendency present towards considerable elong- ation of middle digits of forefoot and hindfoot; (fur not spiny). Subfamily Dactylomyinae {Thrinacodus, Dactylomys, Kannabateomys) Cheekteeth not speciallv broadened; no tendency present towards any lengthening of the digits. Cheekteeth evergrowing, the folds filled up with cement, the teeth llatcrowned; fur not developing spines; paroccipital processes usually standing apart from the bullae (not always); palate slightly constricted anteriorly; form usually not rat-like (modified for arboreal or terrestrial life). Subfamily Capromyin'ae (Procapromys (not seen), Capromys, Geocapromys) Cheekteeth rooted, the folds normally isolating as narrow islands in adult, or rarely {Echimys and allies) more persistent, the dental pattern more complex, or in extreme development becoming a series of transverse plates. External form usually rat-like; modified for arboreal, terres- trial, or slightly for fossorial life; a tendency present for the fur to be spiny; paroccipital io6 ECHIMYIDAE: ECHIMYINAE processes curved forwards under the bullae, which are large but not abnormally so, excepting CIvoinys; palate not constricted anteriorly. Subfamily Echimyinae (Efhtinys, ho/hrix, Diploniys ; Proechimvs, Hdplomxs, Eiirvivgoiiiat<»iiYS, (Ivomys, Cartcr- odon, Cercomvs, Mesomvs, Lonchothrix) Subfamily ECHIMYINAE Geographical Distribution. — Neotropical region from Nicaragua south to Paraguay and South Brazil. Number of Genera. — Eleven. Char.\cters. — Cheekteeth rooted, not specially broadened, flatcrowned with inner and outer re-entrant folds which become isolated on crown surface with wear; or with a heavier dentition, more complex, and tending to become a series of transverse plates (Echiinys, Isotlirix, Diplomys). Bullae prominent ; paroccipital processes curved forward under the bullae. Skull with broad frontals, little or not constricted. Externally more or less rat-like; a tendency present for development of spiny covering, which is in rare cases strongly developed. Compared with the Capromvinae, the cheekteeth are brachyodont and rooted, and of a less simple general appearance, the pattern changing during the animal's life; compared with the Dactylomyinae, there is no broadening of the cheek- teeth, which never show traces of the almost prismatic pattern peculiar to Dactylomyinae; and no digit elongation takes place; compared with the Myo- castorinae, cheekteeth more brachyodont, and pattern not long maintained; also externally not modified for aquatic life; skull much less heavily ridged, size smaller, form much less heavy, and paroccipital processes not lengthened; com- pared with Thryonomyinae, skull much less heavily ridged; forefoot with four and hindfoot with five clearly developed digits. In the remainder of the subfamilies the cheekteeth are more simplified. There seems in this group a tendency for the tail to be lost during life, paralleled by the Old World porcupine Trichys; in the case of Trichys it is sug- gested that the males may bite oft the tail of the female when courting; but whether this might apply to this group is not known. The subfamily contains two sections, in one of which the teeth are much heavier, more complex, the folds usually not isolating as islands; in extreme development the teeth become a series of transverse plates. The genera contained in this section have the feet modified tor arboreal life. EcHlMYS is the main genus, with five or more rather well-defined specific groups, and a wide range in South America ; a species, E. arimitiis, is said to occur in the Lesser Antilles (Martinique). The fur may be strongly spinous, or weakly so; IsoTHRix appears to be indistinguishable cranially or dentally from Echiniys, but has soft fur, and a bushy tail; Diplomys is a (chiefly) Central American ECHIMYINAE 107 genus in which hoth lower and upper molars are a series of transverse plates, though in a section (jf Kchini\s the upper molars arc already so, so that Diplomys is not widely separated from Echimvs. The simpler-toothed branch of the subfamily, in which the dentition is lighter, and the folds isolate as islands in the adult, contains eight genera; I'ROEClllMYS is much the most widely distributed, having a range coincident with with that of the subfamily, and very many named forms; this genus has spiny fur in adult, though not highly developed, and is terrestrial; Hoplomvs stands near Proechimys, but the spiny covering is very much more developed, and the cheekteeth are more complex; the genus ranges from Nicaragua to Ecuador. EuRVZYGOMATOMYS is like Proechimys, but is more modified for fossorial life (though not highly so); the tail is strongly reduced; the cheekteeth are more simplified than in normal Proechimys, and the skull is less ridged posteriorly, but the zygoma is greatly broadened. Clyomys is near Euryzygomatomys, but alone in the group has abnormally inflated bullae. Carterodon is not unlike the last two named in cranial characters, but has deeply grooved incisors (unique in the subfamily), and softer fur. These three genera are rare and have a restricted range in Brazil, the last two being very little known. Cercomy'S agrees with the last named in essential cranial and dental characters, but is not in any way modified for fossorial life; the tail is long and fully haired, the fur soft. Mesomys is an arboreal type ranging across the tropical portions of north- ern South America; the fur is densely spiny, the teeth are of the Proechimys type. Finally Lonchothrix, rare and little known, is much like Mesomys externally except for the heavily tufted tail, but dentally differs from all in the considerable width of the re-entrant folds of the molars. Key to the Genera of Echimyinae Cheekteeth lighter, with narrow folds, which typically become isolated as islands in adult; the teeth never tending to become a series of transverse plates. The hindfcet broadened, noticeably modified for arboreal life. (Fur densely spiny; D.5 hindfoot relatively long; tail long, usually longer than head and body.) Cheekteeth with strong wide folds; tail conspicuously tufted terminally. LoNCHOTHRIX Cheekteeth with weaker narrower folds; tail not or scarcely tufted terminally. Mesomy'S The hindfeet narrow, long, terrestrial in type. (D.5 hindfoot relatively short; tail in general shorter than head and body.) Jugal thickened anteriorlv; zygomatic region noticeably broadened, Tail strongly reduced. Upper incisors grooved ; fur soft ; foreclaws less enlarged. Carterodon Upper incisors plain; fur bristly; foreclaws to a greater or lesser degree enlarged. io8 ECHIMYINAE: ECHIMYS Bullae abnormally inflated. Clyomys Bullae moderate. Euryzygomatomys Jugal not or rarely thickened anteriorly, zygomatic region narrower. Tail not strongly reduced. (Upper incisors plain; bullae never abnormally inflated.) Fur soft; tail well haired; palatal foramina broadened; folds of cheekteeth tend less to isolate. Cercomys Fur spinv; tail poorly haired; palatal foramina not noticeably broadened; folds of cheekteeth tend to isolate to a greater degree. Modification of hair into spines at highest development; cheekteeth more complex, outer side of upper cheek- teeth with clear traces of four folds. Hoplomys Modification of hair into spines much less developed ; cheek- teeth relatively simpler, outer side of upper cheek- teeth with clear trace of three folds or less. Proixhimys Cheekteeth heavier, with more persistent folds, the general effect comple.x; the folds isolating late, or not isolating; or cheekteeth tending to become a series of transverse plates. (Feet adapted for arboreal life.) The lower molars a series of transverse plates, as well as the upper series. Diplomys The lower molars not becoming specialized into a series of transverse plates; (the upper cheekteeth may or may not be so). Fur bristly or spiny; tail not bushy. EcHIMYS Fur soft; tail bushy, Isothrix (k-nus I. I'X'HIMYS, Cuvier 1S09. EcHiMVS, Cuvier, Bull. Soc. Philom. XXIV. p. 3134. 181 1. LoNCHERES, Illigcr, Prodr. Syst. Mamm. et Avium, p. 90. (Myoxiis clirysurns, Zimmermanii). 1837. Nelomys, Cuvier, Ann. Sci. Nat. Paris, 2, VIII, p. 370. (Nelomys blaiiivitlei, Cuvier). 1841. Phvllomvs, Lund, ,Afh. K. Danske Vid. Selsk, 4, VIII, p. 243. (FhyUoiiiys brazilicnsis, Waterhouse.) 1840. EcHINOMYS, Wagner, .Abb. Bayer. Akad. 3, p. 203. (Emendation of Eciiimys.) Type Species. — Myoxus chrysurus, Zimmermann. Range. — Neotropical; Colombia, Ecuador, Peru, Southern Brazil, Eastern Brazil, Amazon region, the Guianas, Venezuela. E. armatus is recorded from Martinique, Lesser Antilles. Number of F'orms. — Twenty-two are named. Characters. — Skull with broad frontals, and as a rule well marked supra- orbital ridges; in larger species the parietals are well ridged. Infraorbital foramen with no canal for transmission of nerve. Bullae prominent, the paroccipitals curved forward under them. Palate varying in width in ECHIMYS 109 different species, but tends to be narrow. Palatal foramina normally short. Zygoma relatively broad, the jugal usually with process both above and below on posterior border. Mandible well twisted and ridged. Cheekteeth complex; apparently usually there are two outer and two inner folds in the upper series, and there is a strong tendency for these teeth to be divided into a series of transverse plates. They are completely so in some Southern Brazilian types as mediits, thumasi, lamarum, dasythrix, etc., for which Thomas revived the generic name Nelomys. Tate synonymizes this with Echimys; some of the northern species come so near this formation that I do not think the name can be used. The lower cheekteeth are characterized by one outer and two inner folds, except the premolar, a complex tooth with three or sometimes four inner folds; the folds of the lower teeth are usually rather well open. Externally, the feet are broad and modified for arboreal life, with D.5 relatively long, and claws prominent; the fur is always spiny, though the spines vary in their development in different groups, being in some very strong, in others relatively weak. The length of the tail is little shorter than head and body to longer than this measurement; it may be well haired or nearly naked. Not very much material of this interesting genus is available for examina- tion, but the forms seen seem to divide sharply and easily into five main groups. These groups should be regarded, however, as provisional, as it may be with more material that some of the characters would not be constant. The blaiinillei group as here understood is equivalent to part of the genus Nelomys of Thomas; containing species in which the upper molars are appar- ently completely specialized into transverse plates; the tail is longer than the head and body, and is more or less completely haired; blainvillei has the spines strongly developed in the one skin seen; medius has the spines weakly developed; thomasi is like medius but much larger (hindfoot 45 as against about 38 or less in allies). E. braziliensis is listed as "hairy-tailed" by Tate, and is provisionally included here, as it was based on the genus " PhvUomys" which is considered by Thomas synonymous with Xelomys. The dasythrix group contains the rest of the genus Nelomys of Thomas; it is closely allied to the last, and from the same area (Southern Brazil), but the tail appears to be not longer than the head and body, and is naked. The spines are strong. It may be that blainvillei might be considered an intermediate form between these two groups. The armatus groups contains forms in which the upper teeth are normally not separated into transverse plates; the tail is naked, or in punctatus rather intermediate between the "hairy-tailed" and "naked-tailed" types. The spines are strong, well developed. E. carrikeri, not seen, appears to stand near punctatus from the description ; longirostris is stated to be near armatus, as is obscura, according to Tate. I can see no specific difference between armatus and occasius, and treat the latter as a subspecies. £. y?aM(yi« is described as an insular form of punctatus. The clirysurus group contains two striking forms; the tail is longer than the head and body, coloured white from about half its length or more to the ECHIMYS end; a white headspot present; spiny covering strong to extreme (at maximum for the genus). The tail is well haired ; the dentition agrees with arniatus. Fig. 8. Echimys armatus armatvs, GeofFroy. B.M. No. 3.4.5.40, ?; ,■ li. Fig. 9. Echimys armatus armatus, GeofFroy. B.M. No. 3.4.5.40, 9; > I J. E. satiirmis, a giant species (hindfoot 51), was thought by Thomas to be near the above group; it is evidently known by one skin only. I am inclined for now to refer it to xht grandis group (below), on account of the much less development ECHIMYS 1 1 i of the spiny covering, which is relatively weak. There is no white headspot. The tail is only white at the extreme tip. With a larger series it may be that saturnus could be shown to belong to the chrysurus group, but with the limited material at present at hand there is no doubt that it is very different from that group. (The tail is long and fully haired.) The grundis group contains western forms with tail sub-equal to or shorter than head and body, not white terminally, and well haired. Spiny covering weak or weak-medium. E. grandis is a very large form (hindfoot 53); dentition usually as in armatus group. \^^^ a b Fig. 10. EcHiMVs armatus armatus, Geoffroy. Cheekteeth: a, upper; b, lower. B.M. No. 3.4.5.40, ?; X 7. The spines may vary through the animal's life, and be an age character throughout this family, within closely related species; but I do not think there is any question of such a difference between groups listed here as "strong- spined" or " weak-spined." In adult grandis, for instance, the spines are very weak; in adult chrysurus abnormally strong. Other species not represented at the British Museum I am unable to allocate to groups. Forms seen; armatus, blaimillei, braziliensis, chrysurus, dasythrix, "gutanae," grandis, lamarum, medius, occasius,paleacea,punctatus, rhipidurus, saturnus, thomasi. List of N'.\.med Forms (The references and type localities for all members of the family Echimyidae have been collected for me bv Mr. G. W. C. Holt.) ,12 ECHIMYS Not seen uiitl not allocated to t^roups 1. I-;CH1MVS SEMIVILLOSUS, Gtoffroy 1838. Revue Zool. i, p. lOi. New Grenada, Colonibia. 2. ECHIMYS MACRL'RA, Wagner 1842. Archiv. fur. Naturg. i, p. 360. Borba, Rio Madeira, Brazil. 3. ECHIMYS UNICOLOR, Wanner 1842. Archiv. filr. Naturg. i, p. 361. Brazil. 4. ECHIMYS NIGRISPINA, Wagner 1842. Archiv. fiir. Naturg. i, p. 361. Ypanema, Sao Paulo, Brazil. dasythrix Group 5. ECHIMYS DASYTHRIX. Ilensel 1872. Abh. Akad. Berlin, p. 49. Rio Grande do Sul, Brazil. 6. ECHIMYS LAMARUM, Thomas 1(316. Ann. Mag. Nat. Hi.st. 8, XVIII, p. 297. Lamarao, Bahia, Brazil. blainvillei Group 7. ECHIMYS BLAINVILLEI, Cuvrer 1837. Ann. Sci. Nat. Paris, 2, VIII, p. 371. Small Island near Bahia, Brazil. S. ECHIMYS MEDIUS, Thomas 1909. Ann. Mag. Nat. Hist. 8, IV, p. 239. Roca Nova, Parana, Brazil. 9. ECHIMYS THOMASI. Ihirint: 1S97. Revista Paulista, ii, p. 171. Island of Sao Sebastiao, near Bahia, Brazil. 10. IXHIMYS BRAZILIENSIS. Waterhnuse 1848. Nat. Hist. Mamni. ii, p. 330. Lagoa Santa, Mmas Geraes, Brazil. armatus Group 11. IX'HIMYS ARMATUS ARMATCS, Geoffroy 1838. Revue Zooliigique, i, p. loi. Cavenne, I'rench Guiana. Synonym: guiamie, Thomas, 1888, Ann. Mag. Nat. Hist. 6, II, p. 326. British Guiana. castaneiis, Allen & Chapman, 1893, Bull. Amer. Mus. N. H. V, p. 222. Princetown, Trinidad. 12. ECHIMYS ARMATL'S OCCASIUS, Thomas 1921. Ann. Mag. Nat. Hist. 9, VII, p. 450. Gualea, Mt. Pichincha. Ecuador. (This locality is queried by Tate, I935-) ECHIMYS— ISOTHRIX 113 .3. IXHIMYS LONGIROSTRIS, Anthony 1921. Amer. Mus. Nov. no. 19, p. 5. Kartabo, Uritish Guiana. 14. IXHIMY.S ()BSCUR.'\, Wanner 1840. Abh. Akad. Wiss. Munch, iii, p. 196. Brazil. 15. KCHIMYS PUNCTATUS, Thomas 1899. Ann. Mag. Nat. Hist. 7, III, p. 153. Caicara, Rio Orinoco, Venezuela. 16. ECHIMYS KL.WIDUS, Hollister 1914. Proc. Biol. Soc. Washington, XXVII, p. 143. El Valle, Margarita Island, Venezuela. 17. ECHIMYS C.-XHRIKERI, Allen 191 1. Bull. Amer, Mus, Nat. Hist. XXX, p. 251. San Esteban, near Venezuela. chrysurus Group 18. ECHIMYS CHRYSURUS, Zimmermann 1780. Geogr. Gesch. ii, p. 352. Surinam (Dutch Guiana). Synonym: cristatus, Desmarest, 1817, Nouv. Diet. d'Hist. Nat. 2nd ed., X, p. 55. 19. ECHIMYS PALEACEA, Illiger 181 1. Prodr, Syst, Mamm. et Avium, nom. nud. 1820, Lichtenstein, Abh. Ak. Wis. Berlin, p. 187. Province of Para, Brazil. grandis Group {satiirnus section) 20. ECHIMYS SATURNUS, Thomas 1928. Ann. Mag. Nat. Hist. 10, II, p. 409. Ecuador, Rio Napo, Prov. del Oriente. (typical section) 21. ECHIMYS GRANDIS, Wagner 1845. Archiv. fiir Naturg. i, p. 146. Managueri, Upper Amazon, Brazil. 22. ECHIMYS RHIPIUURUS, Thomas 1928. Ann. Mag. Nat. Hist. 10, II, p. 291. Pebas, Rio Maranon, Peru. Genus 2. ISOTHRIX, Wagner 1845. IsoTHRlx, Wagner, .^rchiv. fiir. Naturg. i, p. 145. 1852. Lasiiromys, Deville, Rev. Mag. Zool. 2, IV, p. 353. {Lasiuromys villosus, Deville.) Type Species. — hothrix histriatus, Wagner. Range. — Venezuela, Brazil and Peru. (South evidently to Matto Grosso.) Number of Forms. — Eight are named. 8 — LivinR Rodents — I 114 ISOTHRIX Characters. — Like luhiinvs cranially and dentally; (parietals ridged; upper cheekteeth nut tending to become separated into transverse plates). Fur soft, showing no tendency to develop bristles or spines. Tail long, bushy, almost Sciurine. Feet of arboreal type. Remarks. — In a group of this description where considerable specialization is sometimes present towards modification of fur into spines, I think the difl'erence in the coat between Echimys and hot/nix is sufficient for their generic separation. Forms seen : mol/iae, iiegreiisis, oriiioci, pagurus, pictiis, rillosiis. Two groups may be recognized among the material examined, pictiis, with its highly specialized black (or dark brown) and white cok)ur pattern, and the rest which are much more sober in coloration. Mr. Tate stated that pictiis is an Echimys; but it definitely belongs here according to our specimens; some months ago when he was in London we looked at the species together, and he was in agreement with me on this point. List of Named Forms pictiis Group 1. ISOTHRIX PICTUS, Pictet 1 841. Notice An. Nouv. Mus. Geneve, p. zg. Bahia, Brazil. bistriatiis Group 2. ISOTHRIX BISTRI.'^TUS BISTRl.^TUS, Wagner 1845. Arch. Naturg. i, p. 146. Rio Guapore, Brazil. 3. ISOTHRIX BISTRIATUS ORINOCI, Thomas 1899. Ann. Mag. Nat. Hist. 7, IV, p. 382. Maipures, Upper Orinoco, Venezuela. 4. ISOTHRIX BISTRIATUS NEGRENSIS, Thomas 1920. Ann. Mag. Nat. Hist. 9, VI, p. 277. Acajutuba, Lower Rio Negro, Brazil. 5. ISOTHRIX PACHYURA, Wagner 1845. Archiv. fiir Naturg. i, p. 146. Cuyaba, Matto Grosso, Brazil. Synonym: crassicaudiis, Wagner, Abli. .\kad. Miinch., p. 291, 1847. Brazil. 6. ISOTHRIX PAGURLIS, Wagner 1S45. Archiv. fiir Naturg, i, p. 146. Borba, Rio Madeira, Brazil. 7. ISOTHRIX VILLOSUS VILLOSUS, Deville 1852. Rev. Mag. Zool. 2, IV, p. 560. Mission de Sarayacu, Rio Urubamba, Peru. 8. ISOTHRIX VILLOSUS M( )LLIAE, Thomas 1924. Ann. Mag. Nat. Hist. 9, XIII, p. 534. Tushemo, Masisea, Rio Ucayali, Peru. Tiie "hirsiitiis" of Burmeister which has been confused with this genus is a Siginodon (Cricetinae). DIPLOMYS— PROECHIMYS iiS Genus 3. DIPLOMYS, Thomas 1916. DiPLOMYS, Thomas, Ann. Mag. Nat. Hist. 8, XVIII, p. 240. Type Species. — Loncheres caniceps, Giinther. Range. — Panama and Colombia. Number of Forms. — Four. Characters. — In general Hke Echimys, but the lower cheekteeth in a series of transverse plates as well as the upper teeth. There are four laminae in each of the upper teeth; in the lower series there are four in P.4, three in the molars. Externally, the fur is harsh but not spiny; tail moderately haired, but with the scales visible ; feet of arboreal type. Remarks. — The genus is not well represented in London; I assume the dental characters to be constant. Forms seen : caniceps, labilis. List of Named Forms 1. DIPLOIVrV'S CANICEPS, Gunther 1876. Proc. Zool. Soc. London, p. 745. Medellin, Colombia. 2. DIPLOMYS LABILIS, Bangs 1901. Amer. Naturalist, XXXV, p. 638. San Miguel Island, Panama. 3. DIPLO.MYS D.\RLINGI, Goldman 1912. Smiths. Misc. Coll. LX, no. 2, p. 12. Marraganti, Rio Tuyra, East Panama. 4. DIPLOMYS RUFODORSALIS, .'\llen 1899. Bull. Amer. Mus. Nat. Hist. XII, p. 197. Onaca, Santa Marta district, Colombia. Genus 4. PROECHIMYS, Allen 1899. PROECHIMYS, Allen, Bull. Amer. Mus. Nat. Hist., XII, p. 257. 1921. Trinomys, Thomas, .•\nn. Mag. Nat. Hist. 9, VIII, p. 140. Proechimys albi- spiirns, GeofFroy. Valid as a subgenus. Type Species. — Echimys trinitatis, Allen & Chapman. Range. — From Nicaragua, Costa Rica and Panama to Colombia, Ecuador, Peru, Bolivia, Southern Brazil, East Brazil, the Amazon region, the Guianas, Venezuela, Trinidad. Number of Forms. — Appro.ximately fifty are named. Characters. — Rostrum relatively narrow and pointed; incisors typically opisthodont; supraorbital ridges present and usually well developed; parietals usually ridged. Canal for transmission of nerve in infra- orbital foramen present, weak or absent, never strongly developed. Palatal foramina well open, but not excessive; palate relatively broader as a rule than in Echimys; toothrow rather short, far forward in skull, the pterygoid fossae long. Jugal normally thin, but ridged posteriorly and tending to have a weak I'ROECHIMYS process on posterior lower border; thickened only, so far as seen, in iheringi (considerably, but not as extremely as in Euiyzygomatomys), to a degree in the subgenus Trinomys, and to a degree in dimidiatus. Bullae largish; paroccipital Fig. II. Proechimys cayennensis, Desmarest. B.M. No. 3.4.5-44. 9; X I J. Fig. 12. Proechimys cayennensis, Desmarest. B.M. No. 3-4.5-44. S; ' i|- PROECHIMYS 117 processes curved forwards under them, as is usual in the subfamily. Mandible strongly ridged, the angular process clearly lifted outwards, its lower border broad; coronoid process low; angular portion slightly drawn backwards at lower border. Upper cheekteeth normally with three outer and one inner folds each,' these soon becoming isolated as islands. A few species, which will be discussed below, vary slightly in pattern. Lower cheekteeth normally reverse the pattern of the Fig. 13. Proechimys cayennensis, Desmarest. Cheekteeth: a, upper; b, lower. B.M. No. 3.4.5.44, $; x 8. upper series. P. 4 sometimes with extra island anteriorly. Spines always present in adult, but not highly developed, at any rate as compared with such genera as Mesomxs and Hoplomys. Tail rather shorter than head and body, naked or moderately haired. Hindfeet long and narrow, the outer digits shorter than the central three, hallux shorter than D.5; forefoot not abnormal, pollex rudi- mentar\\ Tri.nomys was erected as a subgenus for the species albispimis and setosus, on the ground mainly that the folds of the cheekteeth are in these species reduced from three (external) to two. (One skull seen in which the teeth are cutting has a normal P. 4, however). The cheekteeth may varv slightly elsewhere in the genus; P. lacillator has it appears M.2 and M.3 as Trinomrs; the type of P. dimidiatus appears to be going the same way, though rather worn. One skull seen of P. canicollis seems also to be transitionary towards the Trinomxs type. It should also be noted that P. iheringi seems not quite normal dentally; the teeth in this case appearing a little more complex than usual. ii8 PROECHIMYS Other characters of Trinomxs are cranial, "shorter muzzle, less developed supraorbital ridges, and orthodont or slightly pro-odont incisors." Also the spiny covering seems to be much more developed than in the other species. Forms seen : albispiiuis, ho/iiiainis, brevicauda, calidior, cavciincnsis, canicollis, centralis, cherriei, chiriquinus, chrysaeoliis, coloinbianiis, decumaniis, diinidiatus, goeldii, gorgomie, guiarae, gidaris, hendeei, hilda, iheringi, longicaiidatus, mincae, oris, pac/iitii, paruimensis, rattiinis, roberti, rosa, securiis, semispinostis, setostis, sertotntis, sitnonsi, trinitatis, iirichi, vacillator, warreni. Of the fifty named forms of this genus, twelve have not been seen. Most of the remainder are standing at the present day as "distinct species." A very large number, however, seem to conform to one essential type, so that neither in external nor cranial characters, so far as I can see, are they more than racially distinct. P. caxeiiiieiisis, Desmarest, 1817, is the oldest name for this type of Spiny- Rat. It is true that Tomes has pointed out characters by which his semispinosm may be distinguished from cayeiieimsis; but these seem to me to be relatively unimportant subspecific characters. All forms occurring north of Panama are at present regarded as races of semispinosus. The type and a large or moderate series of skins have been examined in clirysaeolus, deciiiininiis, rosa, warreni, cherriei, gularis, brevicaiida, simonsi, pdchita, hilda, bulivianus, securiis, oris and roberti, and a moderate or large series of skins of mincae, guiarae, trinitatis, uriclii, gorgonae, longicaiidatus, and goeldii. Not one of the above names seem to me to be retainable as full species. There are slight colour variations present ; there are quite noticeable size variations (but the smallest connected with the largest by intermediate forms); there are variations in the length of the tail, though in animals of this kind, which frequently lose the tail during life, this seems to be a character to which too much attention should not be paid. I propose provisionally to treat all the above-mentioned forms, and semi- spinosus and its races, as subspecies of cayenuensis. Should this prove in any case incorrect, it may be noted that numbers 14 to 20 are regarded now as semi- spinosus and its races, and numbers 21 to 40 as "distinct species." P. vacillator is kept separate on account of the dental characters noticed above, though it must be admitted that on external characters alone it would certainly be regarded as a race of cavennensis. P. canicollis differs in colour from the above; hendeei and rattiinis are darker than usual, and with rather weak spines; Thomas regarded these as forming a section of the genus; they are accordingly kept apart as species. P. iheringi is kept apart as a group on account of the characters of the zygoma. P. dimidiatus, of which one skull only has been seen, might belong to that group, or to the typical one; until more specimens come to hand the question must remain open. P. albispiniis differs from the only other species referred to the subgenus Trinomys in colour, so far as seen. The forms here referred as races to cavennensis are mostly supposed by Thomas to be "species" on trifling skull characters, such as the absence of the parietal ridges (age character .'), the length of palatal foramina, " narrow muzzle," etc. PROECHIMYS 119 List or Named Forms Subgenus Proecliimys, Allen Not seen, and not allocated to group. 1. I'ROKCHIMYS OCHRACEUS, Osgood 191 2. Field Mus. Pub. Zool. Ser. X, p. 56. El Panorama, Rio Aurare, Zulia, Venezuela. 2. PROECHIMYS MACROURUS, Jentink 1879. Notes Leyden Mus. i, note 23, p. 97. Surinam. 3. PROECHIMYS O'CONNELLI, Allen 1913. Bull. Amer. Mus. Nat. Hist. XXXII, p. 479. Villavicencio, Colombia. 4. PROECHIMYS POLIOPUS, Osgood 1914. Field Mus. Nat. Hist. Zool. ser. X, p. 141. San Juan de Colon, Tachira, Venezuela. 5. PROECHIMYS STEEREI, Goldman 191 1. Proc. Biol. Soc. Washington, XXIV, p. 238. Hyutanaham, Rio Purus, Brazil. 6. PROECHIMYS KERMITI, Allen 1915. Bull. Amer. Mus. Nat. Hist. XXXIV, p. 629. Lower Solimoes, Brazil. 7. PROECHIMYS BOIMENSIS, Allen 1916. Bull. Amer. Mus. Nat. Hist. XXXV, p. 523. Boim, Rio Tapajoz, Brazil. 8. PROECHIMYS ELEGANS, Lund 1841. Afh. K. Danske Vid. Selsk, 4, VIII, p. 245. Lagoa Santa, Minas Geraes, Brazil. 9. PROECHIMYS LEUCOMYSTAX, Ribeiro 1914. Comm. Linhas. Telegr. Annex. 5, p. 43. Utiarity, Rio Papagaio, Matto Grosso, Brazil. 10. PROECHIMYS MYOSUROS, Lichtenstein 1820. Abh. Akad. Wiss. Berlin (1818-1819), p. 192. Bahia, Brazil. 11. PROECHIMYS LEPTOSOMA, Brants 1827. Muizen, p. 150. Bahia and Sao Paulo, Brazil. Synonym: cinnamomeus, Lichtenstein, 1830, Darstellung, pi. 36, fig. 2. 12. PROECHIMYS FULIGINOSUS, Wagner 1842. Schreber Saug. Suppl. Ill, p. 343. Brazil. From the descriptions, ochraceus (no. 1), o'connelli (no. 3), steerei (no. 5), kermiti (no. 6), appear near caycnncnsis or perhaps races ; poliopus (no. 4) is probably distinct, hoimcnsis (no. 7), is clearly distinct from others, and macrourus (no. 2), is described as a form with an unusually long tail (head and body 221, tail 320). I20 PROECHIMYS ca\ennensis Group 13. PRtJECHlMYS CAYENNENSIS CAYIiNNENSIS, Dcsmarcsl 1817. Nouv. Diet. d'Hist. Nat. 2d Ed. X, p. 59. Guiana. 14. PROECHIMYS CAYENNENSIS SEMISPINOSUS, Tomes i860. Proc. Zool. Soc. London, p. 265. Gualaquiza, Eastern Ecuador. 15. PROECHIMYS C.JiYENNENSIS BURRUS, Bangs 1901. Amer. Naturalist, XXXV, p. 640. San Migue! Island, Panama, lb. PROECHIMYS CAYENNENSIS CENTRALIS, Thomas 1896. Ann. Mag. Nat. Hist. 6, XVIII, p. 312. San Emilio, Lake Nicaragua, Nicaragua. 17. PROECHIMYS C.\YENNI:NSIS PANAMENSIS, Thomas 1900. .\nn. Mag. Nat. Hist. 7, V, p. 220. City of Panama, Panama. Synonym: centralis chiriquimis, Thomas, 1900, Ann. Mag. Nat. Hist. 7, V, p. 220. Bogava, Chiriqui, Panama. 18. PROECHIMYS CAYENNENSIS RUBELLUS, HolHster 1914. Proc. Biol. Soc. Washington, XXVII, p. 57. Angostura Valley, Costa Rica, ig. PROECHIMYS CAYENNENSIS COLOMBIANUS, Thomas 1914. .Ann. Mag. Nat. Hist, 8, XIV, p. 60. Condoto, Choco, Western Colombia. 20. PROECHIMYS CAYENNENSIS CALIDIOR. Thomas 1911. Ann. Mag. Nat. Hist. 8, VIII, p. 254. San Javier, Lower Rio Cachavi, Ecuador. 21. PROECHIMYS CAYENNENSIS CHRYSAEOLUS, Thomas 1898. Ann. Mag. Nat. Hist. 7, I. P- 244- Muzo, north of Bogota, Colombia. 22. PROECHIMYS CAYENNENSIS MINCAE. Allen 1899. Bull. Amer. Mus. Nat. Hist. XII, p. 198. Minca, Santa Marta district, Colombia. 23. PROECHIMYS CAYENNENSIS GORGONAE, Bangs 1905. Bull. Mus. Comp. Zool. Harvard, 46, p. 89. Gorgona Island, Colombia. 24. PROECHIMYS CAYENNENSIS DECUM.ANUS, Thomas 1899. Ann. Mag. Nat. Hist. 7, IV, p. 282. Chongon, Prov. Guayas, Ecuador. 25. PROECHIMYS CAYENNENSIS ROSA, Thomas 1900. .Ann. Mag. Nat. Hist. 7, V, p. 219. Santa Rosa, South-west Ecuador. 26. PROECHIMYS CAYENNENSIS GULARIS, Thomas 191 1. Ann. Mag. Nat. Hist. 8, VIII, p. 253. Canelos, Rio Bobonaza, Ecuador. 27. PROECHIMYS CAYENNENSIS BREVICAUDA, Gunther 1876. Proc. Zool. Soc. London, p. 748. Chamicuros, Huallaga River. Peru. PROECHIMYS 28. PROECHIMYS CAYENNENSIS SIMONSI, Thomas iQoo. Ann. Map. Nat. Hist. 7, VI, p. 300. Perene River, Prov. Junin, Peru. 29. PROECHIMYS CAYENNKN.SIS P.-\CHITA, Thomas 1923. Ann. Mag. Nat. Hist. 9, XH, p. 694. iPucrto Leguia, Rio Pachita, Peru. 30. PROECHIMYS CAYENNENSIS HILDA, Thomas 1924. Ann. Mag. Nat. Hist. 9, XIII, p. 534. San Lorenzo, Rio Maranon, Peru. 31. PROECHIMYS CAYENNENSIS BOIJVl.-\NUS, Thomas 1901. Ann. Mag. Nat. Hist. 7, VIII, p. 537. Mapiri, Upper Rio Beni, Bolivia. 32. PROECHIMYS CAYENNENSIS SECURUS, Thomas 1902. Ann. Mag. Nat. Hist. 7, IX, p. 140. Charuplaya, Secure River, Bolivia. 33. PROECHIMYS CAYENNENSIS WARRENI, Thomas 1905. Ann. Mag. Nat. Hist. 7, XVI, p. 312. Comaccka, Demerara River, British Guiana. 34. PROECHIMYS CAYEiNNENSlS GLAIRAE, Thomas igoi. Proc. Biol, Soc. Washington, XIV, p. 27. La Guaira, Venezuela. 35. PROECHIMYS CAYENNENSIS URICHI, Allen 1899. Bull. Amer. Mus. Nat. Hist. XII, p. 199. Quebrada Seca, Prov. Sucre, Venezuela. 36. PROECHIMYS CAYENNENSIS CHERRIEI, Thomas 1899. Ann. Mag. Nat. Hist. 7, IV, p. 381. Munduapo, Upper River Orinoco, Venezuela. 37. PROECHIMYS CAYENNENSIS TRINITATIS, Allen & Chapman 1893. Bull, Amer. Mus. Nat. Hist. V, p. 223. Princestown, Trinidad. 38. PROECHIMYS CAYENNENSIS GOELDII, Thomas 1905. Ann. Mag. Nat. Hist. 7, XV, p. 587. Santarem, Rio Tapajoz, Brazil. 39. PROECHIMYS CAYENNENSIS ORIS, Thomas 1904. Ann. Mag. Nat. Hist. 7, XIV. p. 195. Igarape-Assu, near Para, Brazil. 40. PROECHIMYS CAYENNENSIS ROBERTI, Thomas igoi. Ann. Mag. Nat. Hist. 7, VIII, p. 531. Rio Jordao, Araguar\' district, Minas Geraes, Brazil. 41. PROECHIMYS CAYENNENSIS LONGICAUDATUS, Rengger 1830. Naturg. Saug. Paraguay, p. 236. North of Paraguay, Matto Grosso, Brazil. 42. PROECHIMYS VACILL.ATOR, Thomas 1903. Ann. Mag. Nat. Hist. 7, XI, p. 490. Kanuku Mountains, British Guiana. 43. PROECHIMYS HENDEEI, Thomas 1926. Ann. Mag. Nat. Hist, 9, XVIII, p. 162. Puco Tambo, 50 miles cast of Chachapoyas, Peru. 122 PROECHIMYS— HOPLOMYS 44. PROICCHIMYS RATTINUS, Thomas 1926. Ann. Mag. Nat. Hist, q, XVIII, p. 164. Tushcmo, Masisea, Rio Ucayali, Peru. 45. PROECHIMYS CANICOLLIS, Allen iSgg. Bull. Amer. Mus. Nat. Hist. XII, p. 200. Ronda, Santa Marta district, Colombia. 46. I'ROECHIMY.S DIMIDIATUS, Gunther 1S76. Proc. Zool. Soc. London, p. 747. South Brazil. iheringi Group 47. PROECHIMYS IHERINGI, Thomas 191 1. Ann. Mag. Nat. Hist. 8, VIII, p. 252. Island of Sao Sebastiao, Sao Paulo, Brazil. Subgenus 'J'rinomvs, Thomas 48. PROECHIMYS ALBISPINUS ALBISPINUS, Geoffrey 1838. Ann. Sci. Nat. X, p. 125. Ilha de Deos, near Bahia, Brazil. 49. PROECHIMYS ALBISPINUS SERTONIUS, Thomas 1921. Ann. Mag. Nat. Hist. 9, VIII, p. 142. Lamarao, Bahia, Brazil. 50. PROECHIMYS SETOSUS, Geoffrey 1817. Desmarest, Nouv. Diet. d'Hist. Nat. X, p. 59. Brazil. The anoinalm of Kuhl, 1820, Beitr. Zool., p. 17, is, according to Tate, based on Heteromys anomahis, Thompson. Genus 5. HOPLOMYS, Allen 1908. HoPLOMYS, Allen, Bull. Amer. Mus. Nat. Hist. XXIV, p. 649. Type Species. — Hoplomys tniei, Allen. Range. — Known from Nicaragua, Panama and Ecuador. Number of Forms. — Three. Characters. — Skull much like Proechimvs; bullae appear a little smaller. Cheekteeth with four (at least) outer folds in the upper series, longer than Pioec/iimvs, placed more obliquely, but isolating in the same way, the folds stretching further across the tooth, and sometimes tending to divide up more when isolated. Lower molars reversing the pattern of the upper series. Frontals and parietals strongly ridged. Feet as Proechimvs; spines much more developed, at maximum for the family, more or less concealing the fur, coarse and strong. Tail shorter than head and body, scaly, naked. Size about largest of subfamily (triiei: head and body 380 mm.). Forms seen: gvmnurus. HOPLOMYS— CERCOMYS 123 According to Goldman (Smiths. Misc. Coll., LXIX, no. 5, p. 124, 1920) all forms may be regarded as races of the oldest name, gynmurus. List of N.ivmed Forms 1. HOPLOMYS GVMNURUS GYMNURUS, Thomas 1897. Ann. Man. Nat. Hist. 6, XX, p. 550. Cachavi, North Ecuador. 2. HOPLOMYS GYMNURUS GOETHALSI, Goldman 191 2. Smiths. Misc. Coll. LVI, no. 36, p. 10. Rio Indio, near Gatun, Canal Zone, Panama. 3. HOPLOMYS GYMNURUS TRUEI, Allen 1908. Bull. Amer. Mus. Nat. Hist. XXIV, p. 650. Lavala, Matagalpa, Nicaragua. Genus 6. CERCOMYS, Cuvier 1829. Cercomys, Cuvier, Hist. Nat. Mammalia, iii, pi. 60. 1881. Thrichomys, Trouessart, Cat. Mamm. Bull. Soc. litudes Sci. Angers, p. 179. (Nelomys antricola, Lund). Type Species. — Cercomys cunicularius, Cuvier. Range. — East Brazil (Pernambuco) southwards (?); known also from Bahia, Lagoa Santa and Paraguay. Number of Forms. — Four. Characters. — Skull broad, rather less heavily ridged than Proechimys and Hoplomys, braincase appearing rather broader, and parietals not or scarcely ridged. Jugal not thickened anteriorly, zygoma narrow, and usually with weak process on lower posterior border. Infraorbital foramen with separate canal for nerve transmission. Bullae large. Upper cheekteeth with two outer folds and one inner one, the folds usually clear and straight, not tending to isolate so completely as in Hoplomys and Proeclmnys. Lower cheekteeth reversing the pattern of the upper series; P. 4 often with a small extra inner fold. Palatal foramina usually abnormally broadened. E.xternally with soft fur, showing no signs of developing bristles; tail slightly shorter than head and body, thickly haired. Feet essentially as Proechimys. Remarks. — This genus has in the past been compared with Dactylomys and Myocasior; but from the dental characters and the characters of the feet I am convinced that it has nothing to do with these genera, but seems to bear very nearly the same relationship to Proechimys that Isothrix does to Echimys. namely a hairy-tailed soft-furred representative. A paper has been published (Boker, 1929, Vcrh. Anat. Ges. Jena, XXXVIII, p. 19) on the bipedal leaping adaptations of a captivity specimen of this genus. Forms seen : hiurentius, fostcri, apereoides. 124 CERCOMYS— EURYZYGOMATOMYS I am convinced that all the three forms seen are not more than racially distinct from each other, though the first two were described as species. Thomas states (Proc. Biol. Soc. Washington, 1912, XX\', p. 115) that apcreuides is synonymous with the earlier described ctiiiicularius. According to Thomas there are four mammae {Iciurcntius). List of N.-\med Forms 1. CERCOMYS CUNICULARIUS CUNICULARIUS, Cuvier 1829. Hist. Nat. Mamm. ill, fig. 276. "Capitanerie des Mines," Brazil. Synonym; afiercoides, Lund, 1841, Afh. K. Danskc Vidensk Selsk 4, VIII, p. 98. Lagoa Santa, Minas Geraes, Brazil. aniricola, Lund, 1841, Afh. K. Danske Vidensk Selsk 4, VIII, p. 242. Brazil. 2. CERCOMYS CUNICULARIUS LAURENTIUS, Thomas 1904. Ann. Mag. Nat. Hist. 7, XII, p. 254. Sao Louren90, near Pernambuco, Brazil. 3. CERCOMYS CUNICULARIUS FOSTER!, Thomas 1903. Ann. Mag. Nat. Hist. 7, XI, p. 227. Sapucay, Paraguay. 4. CERCOMYS INERMIS, Pictct. (Not seen) 1841. Notice An. Nouv. Mus. Gene%'e, ii, p. 33. Bahia, Brazil. Genus 7. EURYZYGOMATOMYS, Goeldi igoi. EuRYZYGOMATo.MYs, CJneldi, Bol. Mus. Paraense, III, p. 179. Type Species. — Echirnvs spinosus, Rengger. (See Tate, 1935, Taxonomy of Neotropical Hvstricoid Rodents). Range. — "Probably throughout the pampas country of Paraguay, northern Corrientes, Parana, Santa Catharina and Rio Grande do Sol" (Tate). Number of Forms. — Three are named. Ch.xracters. — Skull broad, with poorly marked supraorbital ridges, rela- tively broad rostrum, prominent bullae (these not excessively inflated). Palate narrow and rather short; palatal foramina short and broad. Infraorbital foramen with a separate canal for nerve transmission. Jugal long, greatly thickened anteriorly, but with posterior projecting process not well marked; the zygoma more robust than in the genera dealt with above. Mandible heavily ridged and twisted; coronoid higher than in Proechhnys. Llpper cheek- teeth with two outer, one inner folds, becoming isolated as islands with wear; general effect nearer Cercomvs than Proecliinivs. Lower teeth with this pattern reversed. Fur bristly, spines about as well developed as in the less spiny members of Proechimxs, or perhaps less so. Feet narrow, essentially of Proechimys type; EURYZYGOMATOMYS— CLYOMYS 125 claws of forefoot slightly elongated. Tail strongly reduced, not much longer than hindfoot, well haired. I'orins seen : s{>inosiis, cutelliis. 1 do nut think that the above two forms are more than racially distinct from each other. List ov N.\med Forms 1. I;LRYZYGOM.'>iTOMYS SPINOSUS SPINOSUS, Desmarest 1817. Nouv. Diet. D'Hist. Nat. 2d Ed. X, p. 57. Atira, 8 leagues east of Asuncion, Paraguay. Synonym: brachyurus, Wagner, 1843, Schreber Saug. Suppl. iii, p. 346. Brazil. rufa, Lichtenstein, 1818, Abh. Akad. Berlin, p. iy2. Brazil. 2. KURYZYGOMATOMYS SPINOSUS CATELLUS, Thomas 1916. Ann. Mag. Nat. Hist. 8, XVIII, p. 301. Joinville, Santa Catharina, Brazil. 3. KURYZYGOMATOMYS GUIARA, Brandt (Not seen) 1S35. Mem. Acad. St. Petersb. 6, III, p. 432. Ypaneme, Sao Paulo, Brazil. The status of this form appears doubtful (Tate, Bull. Amer. Mus. Nat. Hist. LXVIII.p. 405). Genus 8. CLYOMYS, Thomas 1916. Clyomys, Thomas, Ann. Mag. Nat. Hist. 8, XVIII, p. 300. Type Species. — Echimys laticeps, Thomas. Range. — Described from Joinville, Santa Catharina, Brazil. Number of Forms. — One. Characters. — Essentiallv like Euryzvgomatomys in cranial characters except that the bullae are abnormally inflated, very much more so than in other members of this subfamily, a great part visible external to the paroccipitals when viewed from behind. Dental characters of the one skull seen too worn for notes. Externally like Euryzygomatomys, but foreclaws noticeably more developed. Forms seen : laticeps. List of N.\med Forms I. CLYOMYS LATICEPS, Lund, 1S41, nom. nud., Thomas 1909. Thomas, Ann. Mag. Nat. Hist. 8, IV, p. 240. Joinville, Santa Catharina, Brazil. (1841, Lund, nom. nud.. Afh. K. Danske Vid. Selsk. 4, VIII, p. 99) Genus 9. CARTERODON, Waterhouse 1848. Carterouon, Waterhouse, Nat. Hist. Mammalia, ii, p. 351. Type Species. — Echimys sulcidcns, Lund. 126 CARTERODON— MESOMYS Range. — Brazil (? Lagoa Santa). Number of Forms. — One. Ch.\racters. — Jugal thickened anteriorly, as in Eiiryzvgomatomys and C/voinys; supraorbital ridges developed, and slight inter- orbital constriction present in the one skull examined. Nasals broad. Bullae prominent, hut not extreme. No canal tor nerve transmission in the infraorbital foramen. Mandible heavily twisted. Zygoma with moderate process on posterior lower border. Upper incisors one-grooved. l"hc outer side of these teeth yellow, the inner side white, as remarked bv Waterhouse. Lower incisors plain. Upper cheekteeth with two outer, one inner folds, the enamel sur- rounding them thick; lower teeth reversing the pattern. Size relatively small; fur soft, at any rate compared with most of the genera of this group; tail shortened, well haired, evidently not so reduced as in Eury- zygomatomys; feet narrow and long, as in Proechimys; claws moderate. Forms seen : sukidens. List of N.^med Forms I. CARTERODON SULCIDENS, Lund 1841. Mh. K. Danske Vid. Selsk, 4, VIII, p. 99. (Originally described fossil from Lagoa .Santa, Brazil.) This genus was originally described from fossil remains, but subsequently found living. Genus ID. MESOMYS, Wagner 1845. Mesomys, Wagner, .-^rch. fur Naturg. i, p. 145. Type Species. — Mesomys ecaiidatus, Wagner. R.\NGE. — Amazonia; "From the Tocantins River to eastern Peru and Ecuador" (Tate). Number of Forms. — Approximately seven. Ch,\racters. — This genus was described by Thomas as having the skull, ears and feet of Echiwvs, but the teeth of Proechimys. Skull with short and narrow rostrum, and well marked supraorbital ridges. Frontals tending to be very broad, parietals not or scarcely ridged. No canal for trans- mission of nerve in infraorbital foramen. Bullae relatively large; jugal not specially broadened, with weak process on posterior border both below and sometimes above. Palatal foramina short; toothrow far forward in skull. L'pper cheekteeth of Proechimys tvpe, with narrow folds, usually traces of four external in the upper series, the lower teeth reversing the pattern. .Size small, usually or always under 200 mm. head and body, fur heavily spiny, comparable to that of llophjmys; hindfeet very broad, of arboreal type, D.5 long; claws prominent. Tail usually slightly longer than head and body, scaly, poorly haired except terminally. Forms seen : ferrugiiieus, liispidus. leiiiceps, spicatus, stimiihix. The few species admitted are all very closely allied to each other. MESOMYS— LONCHOTHRIX 127 List of Named Forms 1. MESOMYS HISPIDUS, Desmarest 1817. Nouv. Diet. D'Hist. Nat. 2d. Ed. X, p. 58. "South America." Synon>Tn: ecaudatiis, Wagner, 1845, Arch. fiir. Naturg. i, p. 145. Borba, Rio Madeira, Brazil. (For status see Thomas, Ann. Mag. Nat. Hist. 8, XVI 11, p. 298, 1916.) 2. MESOMYS FERRUGINEUS FERRCGINEUS, Giinther 1876. Proc. Zool. Soc. London, p. 750. Chamicuros, Rio Huallaga, Peru. 3. MESOMYS FERRUGINEUS SPICATUS, Thomas 1924. Ann. Mag. Nat. Hist. 9, XIII, p. 535. Tushemo, near Masisea, Rio Ucayali, Peru. 4. MESOMYS LENICEPS, Thomas 1926. Ann. Mag. Nat. Hist. 9, XVIII, p. 348. Yambrasbamba, Amazonas, Peru. 5. MESOMYS STIMULAX, Thomas 191 1. Ann. Mag. Nat. Hist. 8, VII, p. 607. Cameta, Lower Tocantins, Brazil. 6. MESOMYS DIDELPHOIDES, Desmarest. (Not seen) 1817. Nouv. Diet. d'Hist, Nat. 2nd Ed. X, p. 58. Probably from Brazil. 7. MESOMYS OBSCURUS, Wagner. (Not seen) 1840. Abh. Akad. Wiss. Munch, iii, p. 196. Brazil. Genus 11. LONCHOTHRIX, Thomas 1920. LoNCHOTimix, Thomas, Ann. Mag. Nat. Hist. 9, VI, p. 113. Type Species. — Lonchothrix emiliae, Thomas. Range. — Described from Villa Braga, Rio Tapajoz, Brazil. Number of Forms. — One. Characters. — Skull closely similar to that of Mesomys. Cheekteeth com- pared by Thomas to those of a small Erethizon ; upper molars with three outer, one inner folds, these noticeably wide and deep; lower cheek- teeth evidently with only two inner folds, one outer in all molars; P. 4 with traces of four folds (three main, one vestigial); the folds wide in the lower series. Externally the form is striking owing to the heavily tufted tail, which is considerably longer than the head and body, but hairy at the end only, the upper part scaly and covered with short spines. Fore and hindfeet broad, of arboreal type, as in Mesomys. Spines of body veni' highly developed com- paratively, even the belly being semi-spinous. Little appears to be known of this genus; the teeth do not appear very 128 CAPROMYINAE: CAPROMYS typical of this section; I include it here on account of the resemblance to Mesomys in cranial and external characters. Forms seen : emiliae. List of Named Forms 1. LONCHOTHRIX EMILIAE, Thomas 1920. Ann. Mag. Nat. Hist, 9, VI, p. 114. Villa liraga, Rio Tapajoz. Brazil. Subfamily CAPROMYINAE Geogr.aphic.^i. Distribution. — Cuba, Jamaica, Bahama Islands, Swan Island (Gulf of Honduras), and one form named from Venezuela. Number of Genera. — As here understood the group contains three genera, of which one is not represented in the British Museum. Characters. — Not essentially different from the Echimyinae, but cheek- teeth evergrowing, characterized by two outer, one inner folds in the upper series, the folds in adult completely filled with cement, the teeth flatcrowned and changing little or not at all during the animal's life. Paroccipital processes usually, not always, tending to stand apart from the bullae. Form usually robust, not Rat-like; tail haired, long and prehensile or strongly reduced; habits terrestrial or arboreal. Key to the Genera of Capromyinae, not including the genus Pruccipromvs (not seen) Tail considerably longer than hindfoot; claws more prominent; (habits arboreal). (Tail prehensile, constant ?). Capro.my'S Tail scarcely longer than hindfoot; claws less prominent; (habits terres- trial). Geqcafromys Genus I. CAPROMYS, Desmarest 1822. Capbomys, Desmarest, Bull. Soc. Philom. Paris, p. 185. Type Species.— Cupromvs foiiniieri, Desmarest=Isudoii pilorides, Say. Range. — Cuba, including the Isle of Pines. Number of Forms. — Six. Ch.\racters. — Skull long and rather flat, a postorbital-like ridge can be present; parietals may be well ridged; jugal with well marked and strong backwardly directed process. Bullae prominent. Paroccipital pro- cesses usually slightly lengthened, and standing apart; in the one skull seen of C. nana (adult female), the paroccipital processes join the bullae, about as in Echimyinae. CAPROMYS 129 Infraorbital foramen with no canal for nerve transmission. Palate slightly constricted anteriorly, but less so than in Myocastor or Dactylomys; palatal foramina medium. Mandible with angular process drawn backwards, and stronglv lifted outwards; condyle high; coronoid process low. Incisors narrow. Upper cheekteeth as already described; the lower series reverse the pattern of the upper series, the premolar has also a vestigial e.xtra inner fold. Externally rather large as a rule; fur harsh; feet broad, of arboreal type, or more or less; claws prominent, D.5 relatively long, hallux medium. A tendency in the few skins examined for D.4 to be a little longer than D.3. Forefoot with four digits well developed, pollex small. Tail long, haired, said to be prehensile in at least one species, and may be so throughout the genus. The species of Capromys were revised bv Chapman, 1901, Bull. Amer. Mus. N'at. Hist., vol. XI\', p. 313. Forms seen : prehensilis, pilorides, melamirm, nana. C. Jiana, of which one skull alone is available, seems considerably smaller than the remainder; it was originally described as fossil, but subsequently found living. C. melanurus, of which one specimen only has been seen, has a much more heavily haired tail than the remainder. C. prehensilis and C. pilorides are distinguishable from each other on length of tail, and the latter is stated to have a much heavier skull. The genus is not very well represented in London. List of Named Forms (The references and type localities of all Capromyinae are the work of Mr. G. W. C. Holt.) 1. C.-\PROMY.S PILORIDKS PILORIDES, Say 1822. Joum. Acad. Philadelphia, ii, p. 333. Cuba. Synonym; Joitrnieri, Desmarest, 1822, Mem. Soc. Hist. Nat. i, p. 43. Cuba. ? qiiemi, Fischer, Add. ad Synops. Mamm. 1830, p. 389. 2. CAPRO.MYS PILORIDKS RKLICTLS, .•\llen 191 1. Bull. Mus. Comp. Zool. Harvard Univ. 54, p. 207. Casas Mountains, Nueva Gerona, Isle of Pines, Cuba. 3. CWPRO.MYS PREHENSILIS PREHENSILIS, Poeppig 1824. Joum. Acad. Philadelphia, 4, p. 11. Wooded parts of Southern Cuba. Synonym: poeyi, Guerin, 1834, Mag. Zool. IV, PI. XV, 5 pp., and poeppingi, Lesson, 1842, Nouv. Tabl. Regn. .-Xnim. p. 124. 4. CAPROMYS PREHENSILIS GUNDL.\CHI, Chapman igoi. Bull. .Amer. Mus. Nat. Hist. XIV, p. 317. Nueva Gerona, Isle of Pines. 5. CAPROMYS MEI.ANL RLS, Peters 1864. Mon. Ber. .■ikad. Wiss. Berlin, p. 384. Manzanillo, Cuba. Synonym: pallidus. Peters, 1864, Mon. Ber. .Akad. Wiss. Berlin, p. 384. Cuba. 9 — Living Rodents — I p-IG. 14. GeOCAPROMYS BROWNII, Fischcr. B.M. No. 1334 C.; I. Fig. 15. GEOCiPROMYS brownii, Fischer. B.M. No. 1334 t.; / I. CAPROMYS— GEOCAPROMYS 131 6. CAPROMYS NANA, G. M. Allen 1917. Proc. New England Zool. Club, VI, p. 54. Sierra de Hato Xucvo, Province of Matanzas, Cuba. The " Capromys" elegans of Cabrera, 1901, is a member of the Murine genus Phloeomys. „ b Fig. 16. Geocapromys brownii, Fischer. Cheekteeth: B.M. No. 1334C.; X7. Genus 2. GEOCAPROMYS, Chapman 1901. Geocapromys, Chapman, Bull. Amer. Mus. Nat. Hist. XIV, p. 313. Type Species. — Capromys brozcni, Fischer. Range. — Jamaica, Swan Island, and the Bahamas. Number of Forms. — Three. Ch.'VR.'^cters. — Like Capromys, but tail strongly shortened, scarcely longer than hindfoot; feet with less prominent claws, and pollex more reduced (the feet, however, do not seem very different in formation from those of Capromys); habits terrestrial. "Dentition and cranium as Capromys, but ascending portion of maxillary arch of zygoma wider, superior margin of squamosal narrower, and without process, and occipital region lower" (Chap- man). 1 am inclined to doubt the constancy of these cranial characters between the two groups, particularly when C. nana, not known in Chapman's dav, is compared. The skull may have a sagittal ridge in adult; it may be that this is 132 GEOCAPROMYS— PROCAPROMYS present also in Caproinvs, but not in our small series; the paroccipitals in Geo- capromys appear relatively short. The dentition is as in Capromvs; in the cutting teeth of a newly born animal, the folds are quite open and well marked, hut even as early in life as that the dentition is relatively simple. In M.i lower, the vestigial inner extra front fold of Capromxs is more clearly marked. The group was proposed as a subgenus, but has since been given generic rank; the differences in habit and tail characters between the two groups seem, I think, to warrant their separation. The broader ascending portion of the maxillary seems very well marked in all our series of Gcocaproiiivs with one exception, which might be wrongly identified; other than this it seems a clear distinction between Geocaproinxs and Caproiins. The species were revised by Chapman, 1901, Bull. Amer. AIus. Nat. Hist., vol. XI\', pp. 313-323. hrowni seems rather larger, and with a rather heavier skull than thoracatus and ingrahami, which appear very doubtfully distinct from each other. Certain cranial characters are said to distinguish hrozcnii from thoracatus, and the form of the ear. Forms seen : hrou-nii, thoracatus, ingraliaiiii. List of Named Forms 1. GEOC-^PROMYS BROWXII, Fischer 1829. Syn. Mamm. Addenda, page 3S9 ( -= p. 589). Jamaica. Synonym: hrachyurus, Plill, 1S51, in Gosse, Nat. Sojourn in Jamaica, p. 471. Jamaica. 2. GEOCAPROMYS THORACYTU.S, True 1SS8. Proc. U.S. Nat. Mus. XI, p. 469. Little Swan Island, Gulf of Honduras. 3. GEOCAPROMYS INGRAHAMI, Allen 1S91. Bull. lAmer. Mus. Nat. Hist. Ill, p. 329. Plana Keys, between Acklin Island and Mariguana, Bahama Islands. Genus 3. PROCAPROMYS, Chapman lyoi. Proc.\pro.mvs. Chapman, Bull. .Amer. Mus. Nat. Hist. XIV, p. 322. Type Species. — Capromxs geaxi, Pousargues. Range. — Described from \'enezuela, central coastal region. Nl-.mber oe F0RM.S. — One. Rem.\rks. — Not represented in the British Museum; differing m dental details from Capromxs and Geocapromxs. " Size smaller than the smallest known species of Capromxs, tail half as long as head and body — enamel outline in the first three upper molars continuous, with two external and one internal folds; the fourth, last molar with three dis- tinct and disconnected transverse enamel ellipses, the posterior one about half the size of either of the anterior two; enamel outline in the four lower molars continuous, the first molar with three internal and one external folds, the first PLAGIODONTINAE: PLAGIODONTIA 133 and second interior folds being more extended than in the corresponding tooth of Capramys; the remainini; three lower molars each with two internal and one external folds, the enamel enclosed space on the posterior margin of the last molar being scarcely wider than the enamel itself" (Chapman). Chapman suggests that this represents the ancestral mainland type from which Capromys and Geocapromys descended. List of Named Forms I. PROCAPROMYS GEAYI, Pousargues 1899. Bull. Mus. Paris, p. 150. Mountainous coastal region on the slopes of the range which separates the town of Caracas from the port of La Guaira, Venezuela. Subfamily PLAGIODONTINAE Geographical Distribution. — Dominican Republic. Number of Genera. — One. Characters. — Not unlike the Capromyinae, but cheekteeth differing markedly from any Hystricoid Rodent examined; the upper molars with only one fold each side, these folds yery long and deep, penetrating far into tooth, running parallel to each other and set obliquely; folds well filled with cement, as in Capromyinae; each upper tooth with (in the one examined) a strong outwardly-pointing external projection on the outer side, adjacent to the external border of the outer fold. Cheekteeth eyergrowing. Lower cheek- teeth with two long, deep inner and one shallow outer folds. Paroccipital pro- cesses much lengthened. Jugal simpler than in Capromyinae, without processes on upper or lower border. Tail naked. Rem.\rks. — The status of this genus must remain provisional ; it does not appear to agree with Capromyinae sufficiently to be included in the same subfamily, in dental characters; but only one skull is available for examination. Genus I. PLAGIODONTL-\, Cuvier 1836. Plagiodontia, Cuvier, Ann. Sci. Nat. Paris, 2, VI, p. 347. Type Species. — Plagiodontia aedium, Cu\ier. Range. — As in the subfamily. Number of Forms. — Two. Characters. — Miller compared his P. hylueum with Geocapromys broumii; the most noteworthy differences quoted were (in Plagio- dontia) the less breadth between the lachrymals, the more anterior positions of the swellings caused by the frontal sinuses, the zygoma much more slender, the upper part not bearing an orbital process, the jugal slender, without posterior concavity and posteroinferior process, the excessively long paroccipital processes, the smaller incisive foramina, the greater width of the mandibular masseteric 134 PLAGIODONTIA— DACTYLOMYINAE ridge. Some interorbital constriction is apparent. There is no canal in the infraorbital foramen for nerve transmission. The skull appears depressed, or slanting downwards, posteriorly; the palate is slightly constricted anteriorly. The cheekteeth are as described above. The feet are heavy, the tail naked, of moderate length, the ears small. Claws well developed; D.5 hindfoot relatively long. Miller suggested that the animal is more nearly related to Ade/pliuinva, a Patagonian fossil, than to living Hutias, with which it is currently associated. Forms seen: liyhuuin. List of Named Forms 1. PLAGIODONTIA AKDIUM, Cuvicr 1S36. Ann. Sci. Nat. Paris. 2, VI, p. 347. Dominican Republic. 2. PLAGIODONTIA HYLAKL'iM. Miller 1927. Proc. U.S. Nat. Mus. LXXVII, no. 16, p. 4. Guarabo, lo miles east of Jovero, Samana Province. Dominican Republic. The type species does not appear to be known at present as a living animal. Originally described in 1836, little more was heard of the genus until Miller described livlcieum ninety years later. In 1916 Miller described some bones taken in the Dominican Republic; the impression at that time was that the animal was extinct. Subfamily DACTYLOMYINAE Gf.ographical Distribution. — South America: \'enezuela, Colombia, Amazonia, Ecuador, Peru, Bolivia, Para- guay, S.FL Brazil, etc. Number of Genera. — Three. Characters. — Cheekteeth brachyodont, excessively broad and heavy, the pattern essentially consisting of a deep re-entrant fold in the middle of each upper molar more or less completely dividing each tooth into two lobes, each of which is subdivided by a broad external fold. The pattern varies slightly within the genera, but the general somewhat prismatic effect is unmistakable. There is a strong tendency towards anterior constriction ot the palate, as in M\ociistur\ the paroccipital processes are usually as in Ec/iimys, i.e. curved forward under the bullae, but in some specimens of Dactylomys dactylinus, they stand apart from the bullae and cannot be distinguished from those of Geocapromys, which makes the former separation of this section of Rodents into forms with large paroccipitals (family "Capromyidae") and forms with paroccipitals curved under the bullae (family Echimyidae, hitherto including Dactylomys) unretainable. Skull number 22.5.4.4. in the British Museum appears just as Geocaproinvs, in paroccipital structure. The palatal foramina are small or nearly obsolete; the palate is very narrow and extends to a level with hinder part ot .M.3 or slightly behind it. DACTYLOMYINAE: THRINACODUS 135 The fur is soft, not developing spines. A feature of the group is the extreme elongation of certain digits in the manus and pes of all except Thrinacodus, a character very unusual or unique within the Order. I am told that these are climbing animals, and that they grasp the branches between their third and fourth digits. The tail is usually much longer than the head and body, and may be heavily haired, or naked and reptilian in appearance. If the families Capromyidae, Myocastoridae and Thryonomyidae are to be retained as distinct from the Echimyidae, I suggest the present group should also form a special family, Dactylomyidae. For the purposes of the present work, however, all these groups are kept within one family, as already noted. Key to the Gener.a of D.actylomyinae Digits three and four of both fore and hindfeet not specially elongated, and not broadened. THRiN.'iCODUS Digits three and four of both fore and hindfeet much elongated, and considerably broadened. Palate much constricted anteriorlv; main lobes of upper cheekteeth not united by enamel bridges. Dactylomys Palate scarcely constricted anteriorly; main lobes of upper cheekteeth united by narrow enamel bridges. Kannabateomys Genus i. THRINACODUS, Gunther 1879. Thrinacodus, Gunther, Proc. Zool. Soc. London, p. 144. Type Species. — Thrinacodus albicauda, Gunther. R.\N"GE. — Colombia and Venezuela. Number of Forms. — Three. Characters. — Essential cranial and dental characters as Dactylomys, ne.xt to be described. Palate as Dactylomys. Digit elongation at minimum for the subfamily; the foreclaws sharper than in related genera; polle.x as usual in the group scarcely traceable. Digits narrow; D.3 and D.4 longer than the outer digits in forefoot; D.2 longer than D.5. Hindfoot with digits not abnormal, essentially like those of forefoot except that a short hallu.x is present. Fur thick and soft; tail longer than head and body, moderately to poorlv haired. (The paroccipital processes agree with those of the Echimyinae.) Forms seen : eda.x, albicauda. The three forms are at present regarded as species ; I am not very- convinced as to their distinctness from each other, and they should perhaps be regarded as races. List of Named Forms I. THRIN.'VCODLS .\LBlCAUD.-\, Gunther 1879. Proc. Zool. Soc. London, p. 144. Near Medellin, Colombia. 136 THRINACODUS— DACTYLOMYS 2. THRINACODUS APOLINARI, Allen 1914. Bull. Amer. Mus. Nat. Hist., XXXIII, p. 387. Tomeque, Bogota district, Colombia. 3. THRINACODUS ED.A.X, Thom.ns 1916. Ann. Mag. Nat. Hist. 8, XVIII, p. 299. Sierra de Merida, Venezuela. Genus 2. DACTYLOMYS, Geoffroy 1S38. D.\CTYLOMYS, CJeotfroy, .^nn. Sci. Nat. Paris, 2, X, p. 126. 1916. L.\CHNOMYS, Thomas, Ann. Mag. Nat. Hist. 8, XVIII, p. 298. (Dnclylomys peruanus, Allen). Valid as a subgenus. Typi; Sphcies. — Dactvlomvs tvpiis, Geotfroy — iic/z/wivv dtictvliinis, I)es- marest. R.'iNGE. — Known from Ecuador, Peru, .Amazonia, and Bolivia. Number of Forms. — Five. Characters. — Skull typically with weak postorhital-like ridges, and a sagittal crest developed in adult; paroccipital processes, as noted above, in the type species tending to stand apart from the bullae (age character?), or in smaller forms about as in Ecliimys. Jugal not thickened anteriorly, but relatively broad, with small process on upper and lower border posteriorly. Bullae moderately large. Infraorbital foramen with no canal for nerve transmission. Palate constricted anteriorly so that the premolars almost touch each other. Upper cheekteeth as described above, the inner side of the lobes narrow and contracted ; lower molars with two inner folds, the hinder one completely divitling the tooth; lower premolar with a small extra lobe anteriorly, behind which it is not so completely divided into lobes as are the molars. Size rather large; fur soft; tail longer than head and body, tvpically almost completely naked e.xcept the portion joining the bodv, which is well haired. Forefoot with the two central digits greatly lengthened, broadened to a degree; D.2 also considerablv lengthened; D.5 short; poUex untraceable normally. Claws weak, nail-like. Ilindfoot not very different from forefoot except that the hallux is moderately developed, and the claws are more prominent. Lachnomys, proposed as a subgenus bv Thomas for D. peiiiaiiiis, is given generic rank by Tate, though it scarcely seems even a valid subgenus. The fur is much thicker, and the tail is fully haired throughout. The dental details given by Thomas to divide the two subgenera are not clear to me in our series. Forms seen : dactvliiius, caiiesceiii, peiiianus. The species bolivicnsis appears from description to be very closely allied to the type species. List of Named Forms Subgenus Dactvlomvs, Geoffroy 1. DACTYLOMYS D.\CTYIJNUS DACTYLINUS, Desmarest 1S17. Nouv. Diet. d'Hist. Nat. 2nd Ed. X, p. 57. No locality in original description. Synonym: typtis, CJeoffrny, 1S38, Ann. Sci. Nat. Paris, 2, X, p. 127. Brazil (?) DACTYLOMYS— KANNABATEOMYS 137 2. DACTYLOMYS DACTYI.INL'S CANESCENS, Thomas 1912. Ann. Mag. Nat. Hist. 8, XI, p. 87. Itacoatiara, Middle .Amazons, Brazil (below Manaos). 3. DACTVLO.MYS D.^CTYLINU-S MODESTUS, Lonnbcrg 1921. Archiv. fiir Zool. XIV, no. 4, p. 38. Banks of Rio Curaray, Ecuador (Prov. del Oriente). 4. DACTYLO.MYS B(:)LIVIEN.SI.S, Anthony 1920. Journ. Mamm. Baltimore, I, p. 82. Mission San Antonio, Cochabamba, Bolivia. Subgenus Lachnomys, Thomas 5. DACTYLOMYS PERUANUS, Allen 1900. Bull. Amer. Mus. Nat. Hist. XIII, p. 220. Juliaca, Peru. Genus 3. KANNABATEOMYS, Jentink i8gi. Kannabateomys, Jentink, Notes Leyden Mus. XIII, p. 109. Type Specie.?. — Dactylomys amblyonyx, Wagner. Range. — Paraguay and South-eastern Brazil. Number of Forms. — ^Tvvo. Char.^cters. — The palate very slightly constricted anteriorly; the check- teeth not completely divided into lobes, the lobes being connected bv a small bridge; the enamel folds more nearly perpendicular to the molar series than in Dactylomxs. The lower cheekteeth with an anterior V-shaped fold, and a posterior elongated one, as in Dactylomxs, but the lobes thus formed united by a small bridge. Lower premolar like that of Dactvlomys, but anterior lobe larger. Skull much like that of Dactvlomys; apparently a sagittal ridge is not formed ; the paroccipital processes curve under the bullae. Externally rather smaller than typical Dactylomys; the fur thick, soft. Forefeet much as in Dactylomys; hindfoot relatively broad, essential digit arrangement as in Dactylomxs. Tail very long, relatively well haired. Remarks. — Very closely allied to Dactylomys. The character of the palate is perhaps the most important in keeping the two genera separate. Forms seen : amblyonyx, pallidior. List of Named Forms 1. KANNABATEOMYS AMBLYOViTC AMBLYONY'X, Wagner 1845. Archiv. fiir N'aturg. i, p. 146. Ypanema, Province of Sao Paulo, Brazil. 2. KANNAB.VnCOMYS AMBLYONV.X I'ALI.IDIOR. Thomas 1903. Ann. Mag. Nat. Hist. 7, XI, p. 489. Sapucay, Paraguay. Fic. 17. Kannabatf.omys amblyonyx amblyonyx, Wagner. B.M. No. 1.6.6.67, V; ■ li. Fig. 18. Kannab.^teomys amblyonyx amblyonyx, Wagner. li.M. No. 1.6.(1.67, V; X li. MYOCASTORINAE 139 Fig. 19. Kannabateomys amblyon^-x amblyonyx, Wagner. Cheekteeth: B.M. No. 1.6.6.67, 9; X5- Subfamily MYOCASTORINAE Geographical Distribution. Number of Genera. — One. -Southern South America. Characters. — The external form robust and heavy, the size larger than in other members of the family; the genus is quite one of the giants of the Order. The external characters show strong specialization towards aquatic life; the hindfeet have four of the toes webbed; D.5 is free, and perhaps used for combing the fur. The hindfeet are much larger than the forefeet, which bear a rather rudimentary pollex, and four well-developed main digits; all the digits are armed with sharp large claws. The skull is more heavily ridged for attachment of muscles than in other Neotropical Echimyidae, and is the only member of the familv which tends in this character to approach the .\frican Thryonomys. The paroccipital processes are greatly elongated; the lateral process of each stands well apart from the main downwardly pointing bone. The cheekteeth decrease markedly in size from M.3 forwards; thev are semi-rooted, and broadened, with strong inner and outer re-entrant folds, which are long retained; the palate is strongly constricted anteriorly. I40 MYOCASTOR (;cnus I. MYOCASTOR, Kerr. 1792. Mvoc.'VSTOU, Kerr, Anini. KiiiRd., p. 225. 1S05. MvopOTAMUs, (ieoffroy, Ann. Mus. d'Hist. Nat. VI, p. S2. (Myopotoiinis boiiarieiisis, Geoff roy.) Type Species. — Mas Kjypus, .Molina. R.'\NGE. — Southern South America; Holiister in a review of races repre- sented in the American Museum quotes as localities: Chile, the Straits of Magellan, Buenos Ayres, Santa Fe and Paraguay, Parana River; Rio Negro and Rio Salados, Patagonia. Whether the genus ranges farther north than any of these has not been ascertained; quoted by Waterhouse from Peru. Number of Forms. — Three. Cn.\R.\CTERS. — The nasals are somew hat arched, the frontals broad and flat, the parietals deeplv depressed, and in adults a very strong sagittal ridge is present. There is a sharply pointed but short squamosal process and a small postorbital process to the frontals. The anterior zygomatic root is placed farther back than normal, over the middle of the toothrow. The occipital region is high and prominent. The bullae tend to spread sidewavs, with the neck pointing outwards and upwards, approaching the type found in Castor, though much less developed than in that genus. The hamular processes are thick, the palate very narrow anteriorly, broad posteriorly. Jugal thick, broader posteriorly, with an upwardly projecting process on posterior border. There is no special canal for nerve transmission in the infraorbital foramen. The mandible is immensely heavily ridged and distorted outwards, the angular process sharply drawn backwards. The coronoid process is obsolete. The cheekteeth are extremely hvpsodont ; the fundamental pattern of the upper series, judging by a young specimen, appears to be two external re-entrant folds, the front one placed far forwards, the second one about in the middle of the tooth, and two internal folds, the first almost meeting the second outer one, the second placed posteriorly, rapidly extending across the tooth and cutting off the posterior part altogether. The enamel surrounding the folds is wide, the general effect of the dental pattern rather complex, probably not changing much tiuring the animal's life. In the lower series, there are three inner and one outer re-entrant folds; P. 4 has one small extra inner fold. M.3 is in both jaws con- siderably the largest tooth, in the adult; M.2 is markedly larger than the anterior two teeth, which tend to wear down in old age. In these front teeth, the folds tend to isolate, but the effect is considerably different from such types as Eiirvzvgiiiiiiitdmys in which as the folds isolate the pattern tends to become simpler. The general effect of the teeth is reminiscent to a degree of that of Castor, perhaps owing to the similar life which these two unrelated animals lead. The incis(jrs are broad and powertvd. The essential external characters are described above; the fur is soft and MYOCASTOR 141 thick, and of some commercial value ("Nutria"). The tail is moderate in length, scalv and poorly haired. The largest of a small series of skins at the British Museum is 586 mm. head and body; whether this would represent about the extreme development for the genus I do not know. Forms seen : coy pus. Fig. 20. Myocastor corpus santaecruz.\e, HoUister. B.M. No. 16. 10.3. 85; X {. List of N.mvied Forms (References and type localities by Mr. G. \V. C. Holt.) I. MYOCASTOR COVPLS COYPLS, Molina 1782. Sagg. Stor. Natur. Chile, p. 287. Chile. Synonym: popelairi, Wesmael, 1841. Hull. .Ac. Roy. Brux. \III, 2 p. 61. chilensis. Lesson, 1842, Nouv. Tabl. Ri-gn. Anim. p. 126. Fig. 21. Myocastor coypus s.\ntaecrl'z.\e, Hollister. B.M. No. 16.10.3.85; X J. Fio. 22. Myocastor coy'pus santaecruzae, Hollister. Mandible from below: B.M. No. 16. 10.3. 85 ; • *. Fio. 23. Myocastor coypus santaecruzae, Hollister B.M. No. 16.10.3.8s; X i. Cheekteeth: B.M. No. 16.10.3.85; ,< 2. 144 THRYOXOMYINAE: THRYONOMYS 2. MYOCASTOR COVI'LS BONARIHNSIS, Geoffroy iSof) (1805). Ann. Mus. d'Hist. Nat. VI, p. 82. Paraguay. SjTionym: castorides, Barrow. 1815, Trans. Linn. .Soc. London, XI, p. 167. Brazil (?) 3. MYOCASTOR COYPUS SANTAECRUZAE, Hullister IQ14. Proc. Biol. Soc. Washington XXVII, p. 57. Rio Salado, near Los Palmares, Santa Cniz, Arcentina. Subfamily THRYONOMYINAE Gfogr.\phral Distribution. — Africa, widely distributed south of the Sahara. "Central and East Africa from Uahr-el-Ghazal and Uganda to Eastern Cape Province" (St. Leger); Nigeria; Angola; T. swindcritiiuis group; Kenya, Uganda, North Congo and Nyasaland, r. i^regoriamis group. Number of Genera. — One. Characters. — Skull massive, excessively prominently ridged; cheekteeth rooted, similar in pattern to some of the genera of Echimv- inae; incisors powerful, the upper ones heavily three-grooved; paroccipital processes elongated; occipital region of skull extremely powerfully developed. Arrangement of digits of fore and hindtoot perissodactyle; hallux entirely suppressed ; D.5 of manus vestigial ; claws thick and heavy, more or less fossorial. The shoulder-blade as described by Tullberg is apparently peculiar, and not like that of the other Echimyidae examined by him. Remarks. — As indicated above, \yithout comparing the shoulder-blade of this animal with all other genera included here, it is not wise to base a separate family on this alone. The digit reduction, unique in the present family, is too uncertain a character in other groups to base family characters on. Nevertheless I am not sure that this animal is rightly referred to the present family, or if it is an entirely distinct otfshoot; it seems to stand alone rather in the Ilvstricoid group, though having no very striking characters to separate it off from the remainder of the more normal genera. Genus i. THRYONOMYS, Fitzinger 1827. AuL.'iCODUS, Temminck, Mon. Mam. Tab. ML'th., p. xxvi. (Not of Eschscholtz.) {AulacodiiS S7iinderiaitus, Temminck.) 1867. Thryonomvs, Fitzmger, Sitz.-B. K. Akad. Wiss.Wien, Math. Nat. CI., 56, p. 141. 1922. Choeromys, Thomas, .Ann. Mag. Nat. Hist. 9, IX, p. 390. {Thryonottiys gre- goriamts, Thomas.) Type Species. — Auhicodus scniipdlnuitus, Hcuglin. Range. — As in the subfamily Thrvonomyinae. Number of Forms. — Ten. Characters. — Skull very prominently ridged; rostrum high, broad, rather reminiscent of Pedetes except that the nasals are less arched and the zygomatic plate is not specially projected forwards; jugal and zygoma THRYONOMYS 145 thick, not markedly angular, the jugal nearly in contact with the iachryjnal, the zygomatic region bearing some resemblance to that of Pedetes. Frontals broad, with sharp angular depression immediately in front of the suture formed bv the frontals and parietals each side, in adult. Infraorbital foramen very large, with well marked canal for nerve transmission. Parietals converging into an exces- sively high sagittal ridge; occipital region high and prominent; bullae moderate in size; paroccipital processes considerably lengthened (less so than Myocastor, probably more so than other Echimyidae). Bony palate e.xtending slightly behind M.3; the palate straight, broad; palatal foramina verv broad and large. Mandible with moderate coronoid process, angular portion low, drawn back- wards to a degree, the mandible very heavily ridged and distorted outwards. Cheekteeth semi-hypsodont, broad and hea\'y; the enamel surrounding the folds thick, the folds broad originally, tending to become narrower with wear, evidently not isolating on crown surface to any degree. Upper cheekteeth with two outer, one inner folds; lower teeth reversing the pattern, P. 4 with small extra inner fold. In old age, the pattern wears out. Incisors very broad and powerful, probably more so than in any other Rodent, the upper ones three-grooved, the main groove normally placed centrally, the second and third placed between this and the inner edge of the teeth. Externally, size rather large (perhaps approaching 600 mm. head and body); form heavy; fur harsh and bristly. Tail not long, comparatively well haired. Forefoot with three main digits, the centre one longest, a minute pollex, and D.5 so reduced that it must be almost functionless, though the claw is about as well developed as those of digits 2, 3 and 4. Hindfoot lacking hallux; the digits otherwise like those of forefoot, but longer; D.5 greatly reduced. I am told that a specimen kept at the London Zoological Gardens "shed its tail" when picked up, thus recalling a feature which is common in the Echimvinae. In the su'inderianus group, the skull is much arched anteriorly; the gregorianiis group was given the generic name " Choeromys" by Thomas on account of the "almost complete absence of the large frontal sinuses present in Thrxonomxs, and so developed as to produce a totally different shape of the opening that leads from the cerebral to the olfactory fossa of the skull. The opening is narrow below, broad above in Choeromys, broad below, narrow above in Thryonomys, where its upper corners have been compressed by the large frontal sinuses; owing to this absence of sinuses the frontal area is flat instead of convex." (The tail was also stated to be more reduced in "Choeromys," but in this character T. sclateri is intermediate, having the tail nearly as long as in the typical group. But in any case the tail is strongly reduced comparatively in the whole genus.) A cranial character such as this, though clearly marked, does not seem of generic importance when one takes into account the differences to be found in the skull of other Hvstricoid genera, for instance, Coendou, in which closely related forms (as laemitutn and mexuanum) may ha\e the skull, in the one case arched, in the other flat; or llystrLx, in which the nasals vary extremely, even in 10 — Living Kotlents — I Fig. 25. iThryonomys gregorianus gregorianus, Thomas. B.M. No. 34.6.2.68, ?: .-, I. Fig. 26. Thryonomys gregorianus gregorianus, Thomas. B.M. No. 34.6.2.68, V; ■ I- Fig. 27. Thrvonomys cregorianus gregorianus, Thomas. B.M. No. 34.6.2.68, ?; X i. Fig. 28. Thrvo.nomys grecori.\nus cregorianus, Thomas. Cheektteth: B.M. No. 34.6.2.68, 9; x 4. 148 THRYONOMYS the closely allied African Crested Porcupines {cristatu compared with ajricae- atistralis); these two groups of Thryonomys are so essentially similar in all other characters that I do not think Choeromys is worth retaining even as a subgenus. Forms seen: aiigolac, congicus, grcguiiaiiiis, harrisoiii, raptoriim, sclateri, su'iihhriauus, TCiriigaiiis. List ok Named Forms (References and type localities hy Mr. G. W. C. Holt.) szcindcriainis ( j roup I. THRYONOMYS SWINDERL-WUS SWINDKRLANUS. TLiiimmck 1827. Monogr. Mamm. i, p. 248. Sierra Leone. (For full range of specific group see "Subfamily Thr\'onomyinae," page 144.) z. THRYONOMYS SWINDERIANUS VARIKGATUS, Peters 1S52. Reise nach Mozambique, Zool. Saug, p. 13S. Africa. Was first mentioned in Manuscript, Peters, 1S45, •''nd recorded from Tette, Macanga, Sena and Boror. Synonym: calamophagus, de Beerst, iSgy, Pousargues. Bull. Mus. Paris, p. 160. Nyasa, Central Africa. seiiiipahiialiis. Heuglin, 1864, Nov. Act. Acad. Leop. Dres- den, XXXL P- 6- Central Africa. 3. THRYONOMYS SWINDERIANUS RAPTORUM, Thomas 1922. Ann. Mag. Nat. Hist. 9, IX, p. 392. Nigeria, Lagos. 4. THRYONOMYS SWINDERIANUS ANGOLAE, Thomas 1922. Ann. Mag. Nat. Hist, g, IX, p. 392. Angola, junction of Luandu and Cuje Rivers. grcgoriatius Group 5. THRYONOMYS RUTSHURICUS, Lunnbcrg 1918. Stockholm. Vet, Ak. Handl. 58, no. 2, p. 78. Central Africa, Rutshuru, east of Rutshuru River, half-way between Lake Albert Edward and Lake Kivu. h. THRYONOMYS C;REGORIANLS GREGORIANUS. Thomas 1894. Ann. Mag. Nat. Hist. 6, XIII, p. 202. Luiji Reru River, Kiroyo, Kenya, 7. THRYONOMYS (;REC;ORIANUS PUSILLUS, Heller 191 2. Smiths. Misc. Coll. LIX, no. 16, p. 17. Ndi, Taita Hills, Kenya. S. THRYONOMYS HARRISONI HARRISON!, Thomas & Wroughton 1907. .Ann. Mag. Nat. Hist. 7, XIX, p. 384. Lado (Anglo-Egyptian Sudan), Loka, 60 miles S.-\V. of Fort Berkeley. 9. THRYONOMYS HARRISONI CONGICUS, Thomas 1922. Ann. Mag. Nat. Hist. 9, IX, p. 390. Uele River, Belgian Congo. 10 THRYONOMYS SCLATERI, Thomas 1897. Proc. Zoo!. See. London, p. 432. Nvika Plateau, Nvasaland. PETROMYINAE: PETROMUS i49 Addenda (gregorianus group): THKYONOMYS I.OGONENSIS, Jeannin. 1936. Manim. Sauvagus du Camcroun, Encycl. Biol. 16, p. 178. IJorders of Logone, Chad district, French Equatorial Africa. J', nitshuricus, not seen, is described as a very short-tailed form, probably nearest the gregorianus group. T. harrisoni has a narrower skull than in allies. Subfamily PETROMYINAE Geographical Distribution. — South-west Africa. Number of Genera. — One. Characters. — Cheekteeth rooted, but showing considerable simplification of pattern; only one fold on each side in the upper series; the internal side of the upper series and the external side of the lower series marked by two elevations, the teeth strongly hypsodont. External form small, general- ized except for the bushy tail. Bullae much inflated; skull flattened; mandible tvpically Ilystricoid in formation. Genus i. PETROMUS, Smith 1831. Petromus, Smith, South Afr. Quart. Journ. i, no. 5, p. 10. Type Svzci^.— Petromus tvpicus. Smith. Range. — As in the subfamily. Number of Forms. — Four. Characters. — Skull broad and flat, without any constriction in the inter- orbital region ; infraorbital foramen with canal for nerve transmission; bullae considerably inflated, the paroccipital processes joining them; palatal foramina deep and long, well open, extending to toothrow; palate extending slightly behind the toothrows, relatively narrow. Angular portion of mandible slanting downwards posteriorly. Incisors opisthodont. Cheekteeth as described above, the elevations clear and well marked, the teeth set obliquely, the folds broad. Lower teeth with one fold on each side in adult, as in the upper series. Pattern ultimately obliterated with wear. External form more or less Rat-like except for the tail, which is bushy, and not verv' much shorter than the head and body. Feet narrow, with short claws; four main digits well developed on both fore and hindfeet, D.5 nearly as long as the others; pollex vestigial, hallux short. Some stiff bristle hairs present on hindtoes, as in Octodontinae. The zygoma is relatively broad, sometimes with weak process on the lower border. The mandible is clearlv distorted outwards in the angular process, like typical llystricoids, but unlike Ctenodactylidae, with which this genus has been associated; the coronoid is low, the angular portion drawn backwards. Forms seen : tvpicus, trupicidis, cunealis. cunealis was described as a species, but is probably best regarded as a race as there seems very little essential difference between it and tvpicus. Fig. 29. Petromus typicus typicus, Smith. B.M. No. 12.4.25. 12 Tionym: poeppigi, Wagler, 1832, Isis, XXV, p. 1219. Quintero, Rio Aconcagua, Chile. ater, Cuvier, 1834, Ann. Sci. Nat. i, p. 323. Coquimbo. noctivagus, Poeppig, 1835, Arch. Naturg. i, p. 252. Quin- tero, Rio Aconcagua, Chile. 2. SPALACOPUS TABANUS, Thomas 1925. Ann. Mag. Nat. Hist. 0, XV, p. 585. South Chile. Genus 6. CTENOMYS, Blainville 1826. Ctenomys, Blainville, Bull. Soc. Philom. p. 62. 1916. Haptomys, Thomas, Ann. Mag. Nat. Hist. 8, XVUI, p. 305; subgenus for C. leucodon, Waterhouse. Type Species. — Ctenomys brasiliensis, Blainville. Range. — South Brazil (Matto Grosso), Bolivia, Paraguay, Argentina (Buenos Ayres region, Jujuy, Salta, Tucuman, Catamarca, San Juan, Cordoba, Mendoza, etc.), Patagonia south to Tierra del Fuego; Chile. Number of Forms. — Approximately sixty-one are named. Characters. — Skull with broad rostrum, postorbital process usually present to frontals, their development variable; parietals well ridged, though evidently most often a sagittal crest is not formed; lambdoid crest prominent; bullae large, pear-shaped, spread sideways; paroccipital processes large, curved under them (the bullae show prominently on each side when skull is viewed from behind). Palate essentially as in other Octodontinae; palatal foramina usually short; jugal with extremely prominent upwardly projecting process in larger forms; this process always present, usually well developed. II — Living Kodents — i Fig. 37. Ctenomys tuconax, Thomas. B.j\l. No. 25.3,1.19, 'i; X iJ. Fig. 3S. Ctenomy,s tucona.\, Thomas. B.M. No. 25.3.1.19, 5; X iS. CTENOMYS 163 Infraorbital foramen with no canal for nerve transmission. Upper incisor root extending far backwards, and showing on inner border of infraorbital foramen, though not so extremely as in Spalacopiis. Mandible with angular processes widely spreading, sharply distorted outwards; coronoid process moderate. Incisors much thickened, usually not pro-odont, except in leucodon and lewisi. Cheekteeth like Octodon, but simpler, the small inner fold obsolete in the upper molars. iM. I vestigial. Fig. 39. Ctenomys tuconax, Thomas. Mandible from below, X ij; Cheekteeth, X 6: B.M. No. 25. 3.1. 19, §. Externally much modified for subfossorial life; eyes and ears reduced; forefoot with extremely large claws (pollex less reduced than is normal) ; hindfoot with moderate claws; hallux rather less reduced than in other Octodontinae, otherwise general arrangement of digits like allied genera ; tail strongly shortened though not vestigial, moderately or poorly haired. Forms seen : antonii, azarae, barbariis, hergi, boliviensis, budini, coludo, dorsalis, emiliamis, fochi, fodax, frater, fuegiiius, fiimosiis, fulviis, goodfellowi, liaigi,johamiis, juris, knighti, latru, lentulus, leucodon, leivisi, luteolus, niagelhmicus, mendocina, mordosus, nigriceps, occidtus, opimiis, perrensi, pontifex, porteousi, recessus, saltarius, sericeiis, steinbuchi, syhamis, talariim, torquatus, tuconax, tucutnanus, tulduco, utibilis, viperinus. 'I'his genus is undoubtedly in great need of revision. The forms seem ex- tremely closely allied to each other, generally speaking, though most of them are 1 64 CTENOMYS standing at present as distinct "species." There is great difference in size be- tween some of the forms, emilianus, tuconax, and iiigiiceps having a hindfoot measurement of 3S mm. and terms like recessiis and occuliiis only 26 mm. But intermediate forms exist between both e.xtremes, so that all hindfoot measure- ment figures exist within the genus between the figures 38 mm. and 26 mm. leucodon and lewisi, the latter described as semi-aquatic, have more pro-odont upper incisors than the others. Three large Bolivian types, boliviensis, good- fellowi and steinbachi, appear to have a skull which is broader than normal, particularly in the region of the muzzle. The genus has been reviewed by Rusconi, 1928 (Anal. Soc. Arg. Geogr. "Gaea," III, p. 235), who shows the subgenus " Haptomys" to be no longer retainable. I propose lor the purposes of the present work to divide the genus into sections. Xo attempt is made to reduce forms to subspecies, the genus being far too big for a revision to be attempted in the present work; undoubtedly very many "species" now standing will ultimately be regarded as races. There are many forms not represented in the British Museum, though in the case of those that have been seen, except in verv few cases, a large and representative series of skins have been examined. So far as Patagonia is concerned, on British Museum material, there are two well-marked groups, very small types like magellaiiiciis, and very large types like fodax present only. But elsewhere, there are the "small," "medium," and " large " sections living apparently more or less side by side, the measurements of which grade into each other. The sections here recognized are as follows, though it must be borne in mind that the plan followed here is no more than provisional, and an attempt to get some order out of considerable chaos. 1. magellaiiiciis section: small forms, smallest of genus; hindfoot usually under 30, rarelv exceeding this measurement, never more than 32; often 24, 25, 26. 2. torquatus section: moderate-sized forms, not becoming very large; hind- foot rarely under 30, never less than 28, usually measurement 31-35 ; never more than 37. 3. opimus section: like the last, but becoming large, approaching maximum for the genus; hindfoot usually over 36, not under 35 excepting one race of opimus {luteolus), which agrees with the larger members of section 2. At maximum, hindfoot up to 48 (fodax); in others as a rule not more than 39. 4. boliviensis section : agreeing in measurement with the last, but skull unusually broadened, particularly in the muzzle region. (Bolivia: boliviensis, goodfellowi, steinbachi.) 5. leucodon section : incisors strongly pro-odont ; hindfoot measurement about 30 (not many seen). 6. /f««/ section : incisors also pro-odont. hindfoot measurement 32-37; water-side dwelling type. Thomas suggested that this was not a near ally of leucodon. CTENOMYS 1 6s List of Named Forms magellanicus section 1. CTENOMYS HAIGI HAIGI, Thomas 1919. Ann. Mag. Nat. Hist. 9, III, p. 210. Maiten, Western Chubut, Argentina. 2. CTENOMYS HAIGI LENTLI.US, Thomas 1919. Ann. Mag. Nat. Hist. 9, III, p. 211. Pilcaneu, Upper Rio Negro, Argentina. 3. CTENOMYS SERICEUS, Allen 1903. Bull. Amer. Mus. XIX, p. 187. Cordilleras, upper Rio Chico de Santa Cruz, Patagonia. 4. CTENOMYS MAGELL.-VNICUS, Bennett 1835. Proc. Zool. Soc. London, p. 190. Port Gregory, Straits of Magellan. Synonym: neglectus, Nehring, 1900, Zool. Anz. XXIII, p. 535. Pata- gonia. 5. CTENOMYS TALARUM TALARUM, Thomas 1898. Ann. Mag. Nat. Hist. 7, I, p. 2S5. Los Talas, Ensenada, La Plata, Argentina. 6. CTENOMYS TALARUM ANTOMI, Thomas 1910. Ann. Mag. Nat. Hist. 8, V, p. 242. Los Yngleses ranch, Ajo, eastern Buenos Ayres, Argentina. 7. CTENOMYS TALARUM RECESSUS, Thomas 1912. Ann. Mag. Nat. Hist. 8, IX, p. 241. Bahia Blanca, .Argentina. 8. CTENOMYS MENDOCINA, Phihppi 1869. Arch, fiir Naturg. p. 38. Mendoza, Argentina. 9. CTENOMYS PONTIFEX, Thomas 1918. .-Xnn. Mag. Nat. Hist. 9, I, p. 39. East side of Andes, Province of Mendoza, Argentina (near Fort San Rafael). 10. CTENOIVrVS BERGI, Thomas 1902. -Ann. Mag. Nat. Hist. 7, IX, p. 241. Cruz de Eje, Cordova, .Argentina. 11. CTE.NOMYS FOCHI, Thomas 1919. .Ann. Mag. Nat. Hist. 9, III, p. 117. Chumbicha, Catamarca, Argentina. 12. CTENOMYS TUCUMANUS, Thomas 1900. .Ann. Mag. Nat. Hist. 7, VI, p. 301. Tucuman, Argentina. 13. CTENOMYS L.ATRO, Thomas 1918. Ann. Mag. Nat. Hist. 9, I, p. 38. Tapia, Tucuman, .Argentina. i66 CTENOMYS 14. CTENOMYS ()CCL:i/rL'S, Thomas 1920. Ann. Mat;. Nat. Hist, g, VI, p. 243. Monteagudo, 80 kilometres south-east of Tucunum City, Argentina. 15. CTENOMYS SALTARIUS, Thomas 1912. .Ann. Mag. Nat. Hist. 8, X, p. 639. Salta. Northern Argentina. ih. CTENOMYS JURIS, Thomas 1920. Ann. Mag. Nat. Hist. 9, V, p. 194. El Chaguaral, Jujuy, Argentina, 20 kilometres east of San Pedro de Jujuy, between San Pedro and Villa Carolina. 17. CTENOMYS DORSALIS, Thomas igoo. Ann. Mag. Nat. Hist. 7, VI, p. 385. Northern Chaco, Paraguay. torqiiatiis section 18. CTENOMYS PERRENSI. Thomas 1896. Ann. Mag. Nat. Hist. 6, XVIII, p. 311. Goya, Corrientes, Argentina. ig. CTENOMYS AZAR.'^E, Thomas 1903. .Ann. Mag. Nat. Hist. 7, XI, p. 228. 37 45' S., 65' W., 780 kiloinetres south-west of Buenos Ayres, Buenos .Ayres Province, Argentina. 20. CTENOMYS PORTEOUSI PORTEOUSI, Thomas 191 6. .Ann. Mag. Nat. Hist. 8, XVIII, p. 304. Bonifacio, South-west Buenos .Ayres, Argentina. 21. CTENOMYS PORTEOUSI AUSTRALIS, Ruscom 1934. Rev. Chili. Nat. Hist. 38, p. 108. Province Buenos Ayres, Argentina. 22. CTENOMYS TULDUCO, Thomas 1921. .Ann. Mag. Nat. Hist. 9, VIII, p. 218. Los Sombreros, Sierra Tontal, .San Juan, Argentina. 23. CTENOMYS FAMOSUS, Thomas 1920. Ann. Mag. Nat. Hist. 9, VI, p. 420. Potrerillo, Rioja, Argentina. 24. CTENOMYS COLUDO COLUDO, Thomas 1920. Ann. Mag. Nat. Hist, g, VI, p. 119. La Puntilla, Tinogasta, Catamarca, Argentina. 25. CTENOMYS COLUDO JOHANNIS, Thomas 1921. .Ann. Mag. Nat. Hist. g. VII, p. 523. Caiiada Honda, San Juan, Argentina. 26. CTENOMYS VIPERINUS, Thomas ig26. Ann. Mag. Nat. Hist, g, XVII, p. 605. Tablelands above Norco, Vipos, Dept. of Trancas, Tucuman, Argen- tina. 27. CTENOMYS SYLVANUS SYLVANUS, Thomas igig. .Ann. Mag. Nat. Hist. g. IV, p. 155. Tartagal, Pro\ince Salta, .Argentina. CTENOMYS 167 28. CTENOMYS SYLVANUS UTIBILIS, Thomas 1920. Ann. Mag. Nat. Hist. 9, V, p. 193. Yuto, Rio San Francisco, Argentina, 20 kilometres east of San Pedro de Jujuy. 29. CTENOMYS SYLVANUS MORDOSUS, Thomas 1926. .'Xnn. Mag. Nat. Hist. 9, XVH, p. 325. Tambo, 75 kilometres cast of Tarija, Bolivia. 30. CTENOMYS DUDINI BUDINI, Thomas 1913. .\nn. Mag. Nat. Hist. 8, XI, p. 141. Cerro de Lagunita, Jujuy, Argentina. 31. CTENOMYS BUDINI BAKBARUS, Thomas 1 92 1. .Ann. Mag. Nat. Hist. 9, VII, p. 185. Sunchal, Jujuy, ;\rgentina. 3=. CTENOMYS FRATER, Thomas 1902. Ann. Mag. Nat. Hist. 7, IX, p. 228. Potosi, Bolivia. 33 CTENOMYS TORQUATUS, Lichtenstein 1830. Darstell. Saugethiere, text of PI. XXXI. Southern Provinces of Brazil and banks of Uruguay River. leucodon section 34. CTENOMYS LEUCODON, Waterhouse 1848. Nat. Hist. Mammalia, II, p. 281. San Andres de Machaca, Bolivia (Dept. of La Paz). lewisi section 35. CTENOMYS LEWISI, Thomas 1926. Ann. Mag. Nat. Hist. 9, XVII, p. 323. Sama, 50 kilometres west of Tarija, Bolivia. opimiis section 36. CTENOMYS EMILIANUS, Thomas & St. Leger 1926. Ann. Mag. Nat. Hist. 9, XVIII, p. 637. Chos Malal, Neuquen, Argentina. 37- CTENOMYS FODAX, Thomas 1910. Ann. Mag. Nat. Hist. 8, V, p. 243. Valle de Lago Blanco, Chubut, Patagonia. 3S. CTENOMYS FUEGINUS, Philippi 1880. Arch, fiir Naturg. p. 276. Eastern Island of Tierra Del Fuego. 39. CTENOMYS FULVUS, Philippi i860. Reise. .Atacama Halle, p. 157. Desert of .Atacama, Chile. 40. CTENOMYS KNIGHTI, Thomas 1919- Ann. Mag. Nat. Hist. 9, III, p. 498. Otro Cerro, 45 kilometres west of Chumbicha, Catamarca, Argentina. 1 68 CTENOMYS 41. CTENOMYS TLCONAX, Thomas 1925. Ann. Mag. Nat. Hist, g, XV, p. 583- Concepcion, Tucuman, Argentina. 42. CTENOMYS OPIMUS OPIMUS, Wagner 1S4S. Archiv. fur Naturg. i, p. 75- Bolivia. 43. CTENOMYS OPIMUS NIGRICEPS, Thomas 1900. .Ann. Mag. Nat. Hist. 7, VI, p. 383. Tetiri, Puno Moquegua Road, South Peru. 44. CTENOMYS OPIMUS LUTEOLUS, Thomas 1900. Ann. Mag. Nat. Hist. 7. VI, p. 384- Cordilleras of Jujuy. .Argentina. bolivieiisis section 45. CTENOMYS BOLIVIENSIS, Waterhouse 1848. Nat. Hist. Mammalia, ii, p. 278. Plains of Santa Cruz de la Sierra, Bolivia. 46. CTENOMYS STEINB.ACHI. Thomas 1907. .Ann. Mag. Nat. Hist. 7, XX, p. 164. Campo of Province Sara, Bolivia. 47. CTENOMYS GOODFELLOWI, Thomas 1921. .Ann. Mag. Nat. Hist. 9, VII, p. 136. Esperanza, Concepcion, Eastern Bolivia. \ot seen, and not allocated to section 48. CTENOMYS BRASILIENSIS, BlainvillL- 1826. Bull. Soc. Philom. p. 62. Minas Geraes, Brazil. (Waterhouse treats torquatus, number 33, as a synonym of this species.) 49. CTENOMYS OSGOODI, Allen 1905 Report Princetown Univ. Exped. to Patagonia, p. 191- Rio Chico de Santa Cruz, Patagonia. „ „ ,^ x- . u- . Svnonvm: robustus, Allen, not of Philippi, 1903. Bull. Mus. Nat. H,st. XIX, p. 185. Patagonia. (According to measurements from description, this species will belong in the moRel- lanictis section.) 50. CTENOMYS COLBURNI, Allen IQ03. Bull. Amer. Mus. Nat. Hist. XIX, p. 18S. „ . • ' ^ Arroyo Aike, 50 miles south-east of Lake Buenos Ayres, Patagonia. (According to measurements from description this species probably belongs in magcUanicus section.) 51. CTENOMYS MINUTCS, Nchring 1 887. Sitz. Ber. Ges. Nat. Fr. Berlin, p. 47. , j c 1 n 1 "Campos," East of Mundo Novo, Rio Grande do bul, Brazil. 52. CTENOMYS RONDONI, Ribeiro 1914. Comm. Linhas, Tel. Annexo, 5, p. 39. Juruena, Matto Grosso. Brazil. CTENOMYS 169 53. CTENOMYS BICOLOR, Kibtiro 1914. Comm. Linhas. Tel. Annexo, 5, p. 41. Matto Grosso, Brazil. 54. CTENOMYS NATTERERI, Wagner 1848. Archiv. fiir. Naturg. i, p. 72. Caissora, Matto Grosso, Brazil. 55. CTENOMYS PUNDTI, NL-hrinc 1900. Zool. Anz. XXIII, p. 420. Alejo Ledensa, Cordova, Argentina. 56. CTENOMYS ATACAMENSIS, Philippi i860. Reise. Atacama Halle, p. 157. Desert of .Atacama, Chile. 57. CTENOMYS ROBUSTUS, Philippi 1896. An. Mus. Nac. Chile, no. 13, p. 11. Canchones, near Pica, Tarapaca, Chile. 58. CTENOMYS PALLIDUS, Philippi 1896. An. Mus. Nac. Chile, no. 13, p. 13. Breas, desert of Atacama, Chile. 59. CTENOM\'S PERNIX, Philippi 1896. An. Mus. Nac. Chile, 13, p. 15. Near Aguas Calientes, Chile. 60. CTENOMYS CHILENSIS, Philippi 1896. An. Mus. Nac. Chile, 13, p. 16. Linares, Chile. 61. CTENOMYS MAULINUS, Philippi 1872. Zeitschr. f. ges. Naturw. XL, p. 442. High Andes of Province of Maule, Chile. Tate lists also a "cinerea" Thomas, which is evidently a mistake for Abrocoma cinerea, Thomas. The family Echimyidae contains according to Miller & Gidley very many Neotropical fossil genera. The Octodontinae are quoted from the Oligocene; one of the genera, Cephalomys, had a deciduous P. 4 (Gregory, Orders of Mammals, 1910), a character not known in living Hystricoids; the Echimyinae (with which Miller & Gidlev include Capromyinae and Dactviomyinae) are quoted from the Miocene; some of the genera, as Isobolodoti (Porto Rico), Brotomys (Dominican Republic), and Boromys (Cuba), are thought to have existed recently. The Thryonomyinae have been described from the Miocene of India. ECHIMYIDAE: GENERAL IVORKS OF REFERENCE VVaterhouse, 1848, Natural Histor>- of Mammalia: Rodentia. General review of all forms then known. T.^TE, 1935, Ta.\onomv of Neotropical Hystricoid Rodents, BiJI. Amer. Mus. Nat. Hist. LXVIII, p. 295. PococK, Proc. Zool. Soc. London, 1922, p. 365; external characters of Hystricomorph Rodents (notes on Octodon, Capromys, Myocastor, Daclylomys, Ctenomys, Thryo- nomys). I70 DINOMYIDAE TuLLBERc, Nova Acta Reg. Soc. Sci. Upsaliensis, XVIII, ser. 3, no. i, iSgg. Ch.\pm.^,\', iqoi, Bull. Amer. Mus. Nat. Hist. vol. XIV, p. 313. Revision of Hutias (Caproniyinae). RuscoNi, Review of Ctenomys, 1928, An. Soc. Arg. Geogr. "Gaea," III, p. 235. Thom.\s, races of Thryoiiomys siLinderiiDuis, Ann. Mag. Nat. Hist., 9, IX, p. 392. 1922. Miller, Proc. U.S. Nat. Mus, LXXII, no. 16, p. 4, 1927 (Plaaiodoiitui hylaeum). Jentink, Notes Leyden Museum, XIII, 1891, p. 105. On Dactylomys dactyUnus and Kannahateomys amhlyouyx. And numerous papers by Oldficld Thomas (Echimyinae, Octodontinae). Family DINOMYIDAE 1896. Thomas: Hystricomorpha; Family Dinomyidae. 1899. TuUberg; Hystricomorpha; (?) Family Dinomyidae. 1918. Miller S: Gidley: Hystricoidae; Family Dinomyidae. 1924. Winge: Family Hystricidae; Dasyproctini, part, group Dinomyes. 1928. Weber: Hy'stricoidea; Family Caviidae, part, subfamily Dinomyinae. Geogr.^phical Distribution. South America; Peru, Colomhia, Ecuador and Western Amazonia. Number of Genera. — One. Characters. — Cheekteeth extremely hypsodont, or probably evergrowing, a series of transverse plates. External form heavy, terrestrial; forefeet and hindfeet with four digits, the feet broad, the claws long and power- ful. Limbs not lengthened. Palate constricted anteriorly. Zygomasseteric structure typically Hystricoid, as regards the formation of the lower jaw. Remarks. — Except by those authors who merge Cunicidus and Dasyprocta with the Caviidae and who have regarded this genus also as a member of that family, Diiiomys has usually been regarded as an isolated type among Hystricoidae. There is not the slightest reason to suppose that the animal is near the Caviidae, the lower jaw being typically Hystricoid in forma- tion, and therefore differing from that family; nor does the genus seem closely connected either with Ciinicnhis or Dnsyprocta, differing from botli in tooth formation as well as the feet and digits. The palate and cheekteeth are similar to those of the Chinchillidae, but from these Dinomys differs by its typically ridged and distorted angular portion of the mandible, the general external form, the absence of part of the lachrymal canal open on the side of the rostrum, as well as by no tendency to great inflation of bullae. Goeldi in a paper on some captivity specimens states that the animals are slow-moving, unlike Dusvpioctii and the Chinchillidae. He mentions the fact that like Dasyprocta but unlike Cunicuhis they will sit up on their haunches and use the front paws when feeding. The claws on dried skins of Dinomys appear to be fossorial in type, but Goeldi states that he has not seen the captivity specimens use the claws for digging. The breadth of the manubrium has been used as a character to distinguish this genus as a family or subfamily from Dasyproctidae or Caviidae (Winge and others); it should be noted that this character, according to Tullberg's notes, may vary within some of the other families. DINOMYIDAE: DINOMYS 171 Long and Narrow Broad Laiiostomus (Chinchillidac) Chinchilla (Chinchillidae) Dolichotis (Caviidac) Cavia (Ca\iidae) It is stated to be long and narrow in Dasyprocta and Citnictdus, broad in Dinomys. 'l"he clavicles in Dinomys are stated to be complete. Genus i. DINOMYS, Peters 1873. Dinomys, Peters, Mon. Ber. Ak. Wiss. Berlin, p. 551. TvPH Species. — Dinomys branickii, Peters. Range. — As in the family Dinomyidae. Number of Forms. — One only is now recognized; revised by Sanborn, 1931, Field. Mus. N.H., zool. ser. XVIII, p. 149. Characters. — Skull heavy and broad, with long broad frontals; the parietals are depressed for muscular attachment, but a sagittal crest is not formed in any of the few skulls examined. No separate canal in infra- orbital foramen for nerve transmission. Bullae medium sized; paroccipital processes not lengthened. Jugal long, broad, but evidently simple. Palate of a similar type to that found in Chinchillidae, but mesopterygoid fossa much broader; the palate is continued farther backwards, to slightly behind the tooth- rows. Palatal foramina small. Lachrymal large. Incisors broad and heavy; cheekteeth a series of transverse plates; four of these in each upper tooth ; four evidently in the lower teeth, but the anterior one vestigial. Externally large, heavy, bearing a superficial resemblance to Cunicubis; but tail longer than hindfoot (fully haired). Hindlimbs not lengthened; the feet broad, the claws long and heavy; no great discrepancy between the lengths of the (four) digits; forefoot with four digits, the claws large and powerful, though apparently narrower than in Cuniculus. The genus is not well represented at the British Museum. Forms seen: branickii, " occidenialis." List of Named Forms (The references and type localities are the work of Mr. G. W. C. Holt.) I. DINOMYS BRANICKII, Peters 1873. Mon. Ber. .\kad. Berlin, p. 552. Central Peru; Montana de Vitoc, Colonia .Amable Maria. Synonym: branickii occidentalis, Lonnberg, 1921, .■\rk. Zool. XIV, no. 4, p. 49. Ilambo, near Gualea, Ecuador. gigas, .Ajithony, 1921, .Amer. Mus. Nov. no. 19. p. 6. Colombia. pacarana, Ribeiro, 1919, Arch. Escola Sup. Agric. Med. Vet. 2, p. 13. .\mazon, Brazil. Fig. 40. DiNOMVS br^nickii, Peters. B.M. No. 34. 9. 10. 191, o; • 1- Fig. 41. DiNOMYS branickii, Peters. B.M. No. 34.9.10.191, cJ; •, i'. ERETHIZONTIDAE 173 The family Dinomyidae as defined by Miller & Gidley (Lachr^-mal canal closed in front of orbit; like the Echimyidae, but cheekteeth combining a multilaminar structure with excessive hypsodonty), is known fossil from the Miocene of South America and the Greater Antilles; many extinct genera are quoted by these authors. Family ERETHIZONTIDAE iSq6. Thomas: HvsTRiroMORPHA, part; Family Erethizontidae, with subfamilies Erethizontinae and Chaetomyinae. 1899. TullbtTg: HvsTRicoMOBPHA, part; Family Erethizontidae. 1Q18. Miller & Gidley: Superfamily HvsTRicom.^E; Family Erethizontidae. Family Echimyidae, subfamily Echimyinae, part (Chaetomys). 1 924. Winge: Family Hystricidae; Hystricini, part; "Sphinguri." 1928. Weber: HvsTRlcomEA, part; Family Erethizontidae. Geographical Distribution. — America; Canada, Western United States; Mexico, Central America, and the greater part of Tropical South America. Number of Genera. — Four. Characters. — Not essentially different from the Echimyidae, but externally more highly specialized ; feet becoming abnormally modified for arboreal life ; function of hallux being taken over in specialized forms by a broad movable pad, the sole becoming abnormally wide; body hair modified partly or completely into short sharp spines. Bullae prominent, but paroccipital processes not lengthened. Cheekteeth rooted, typically with the re-entrant folds extremely wide; external form thickset, heax-y. Remarks. — Presumably because these animals are also known as "Porcu- pines," or because their fur is spiny, most earlier authors placed them in the family Hystricidae. Thomas very properly formed a distinct family for them, and most subsequent authors have retained the distinction. TuUberg states that there is hardly a single common feature between the Old World and New World Porcupines except the spines, and even these are of a consider- ably different structure. The two families differ entirely in the structure of the feet, the structure of the cheekteeth, tlie formation of the bullae, the structure of the tail; even in the essential arrangement of spiny covering. They agree in zygomasseteric structure, which proclaims them both members of the Hystri- coidea, but this seems about all they have in common. In fact, it would seem from cranial and dental characters, at least, that the American representatives of the Hvstricidae (if that family has American representatives, and the resemblance is not due to convergence) are the Dasyproctidae; certainly not the present group. In Erethizontidae the paroccipital processes are less lengthened and evidently of a more generalized structure than in Echimyidae. The zygoma is simpler than in that family. The cheekteeth of the typical subfamily, which contains Erethizon, Echinoprocta, and Coendou, and which has been incorrectly split into 174 ERETHIZONTIDAE: CHAETOMYINAE two subfamilies bv some authors (see notes on Echinoprocta below), are remari<- able for the width of their reentrant folds, paralleling in this formation certain Squirrels as Fiiiiischinis, also the Anomaluridae, and to a degree reminiscent of some of the more complex-toothed Neotropical Cricetinae. Chcietomys, on the other hand, has teeth more like those of Echimys. This is an isolated type, the relationships of which are by no means clear, so that it might be quite correct to refer it to a distinct family Chaetomyidae. Agreeing with most specialized Erethizontinae in the structure of the feet, it differs to a very wide degree from them in cranial and dental characters. The orbit is almost completely surrounded by bone, a very rare feature in the Order; and in no member of the Order which I have seen is this specialization so nearly complete. Moreover, the teeth are not in the least like those of Erethizon and Coendou. It is a rare genus, the exact locality of which I have so far been unable to trace, and evidently little is known about it. The spiny covering of the body is very poorly developed compared with other members of the family. Miller & Gidley transferred it to the Echimyidae, but it seems not to belong there in cranial characters, and the feet are as highly specialized as in Coendou, and evidently in exactly the same manner. Thomas expressed the opinion that mainly on this accoimt it might be re- tained in this family, and formed a subfamily for its reception. This view is here adopted. Key to the Subfamilies of F-Irethizontidae Orbit almost surrounded by extremely thickened jugal and short post- orbital process of frontals. Cheekteeth with narrow re-entrant folds, the structure of the upper series not far removed from laminate. Subfamily Chaetomyinae CI>aeto7nys Orbit large; frontal w-ithout postorbital process; jugal not specially thickened. Cheekteeth with wide re-entrant folds. Subfamily Eretiiizontin.\e Erethizon, Eclivio[>ioctii, Coendou The mandible in this family is characterized by the length of the symphysis; the angular process is not so conspicuously distorted as in the Echimyidae, and a weak ridge below the condyle similar to that sometimes found in Chinchillidae, and presumably for the attachment of masseter medialis, foreshadowing that which is so much lengthened and such an important feature of the jaw in Caviidae, can be present. In Erethizon the lower border of the angular process is conspicuously broadened. Subfamily CHAETOMYINAE Geographical Distribution. — ? Brazil. Number of Genera. — One. Characters. — As indicated in the above key. CHAETOMYS 175 Genus i. CHAETOMYS, Gray 1843. CHAETOMYS, Gray, List Specimens Mamm. in Coll. Brit. Mus. p. 123. Type Species. — Hystrix subspinosa, Kuhl. Range. — Brazil; exact locality apparently not known. Number of Forms.— One. Characters.- — Frontals extremely broad, but some narrowing present in front of the well-marked postorbital process. Parietals strongly ridged, but the posterior part of the skull broad, and parietal ridges showing no signs of coming together. Palate relatively narrow; short. Palatal foramina very short, far in front of toothrows. Bullae relatively large, the meatus produced sharply sideways, forming sharp angle. Paroccipital processes short. Nasal chamber appears less open than is usual in Erethizontinae. Jugal with anterior part immensely broadened, nearly in contact with the postorbital process; the jugal nearly extending to the lachry- mal. No canal for nerve transmission in infraorbital foramen. Mandible with low coronoid; angular process relatively small, the lower border not specially widened, but this part of the jaw clearly distorted outwards. Fig. 42. CtMETOMVs subspinosus, Kuhl. B.M. No. 3.9.4-86, S; X 1. Fig. 43. Chaetomys subspinosus, Kuhl. B.M. No. 3.9-4-86. ?; ;' >■ Fig. 44. Ch.-\f,to\iy.s sub.spinosis, Kuhl. Cheekteeth: B. M- No. 3. 9.4.S(,, V; 3. CHAETOMYS— ERETHIZONTINAE 177 Upper cheekteeth are divided into three lobes, the middle one being simple, straight, separated from the front and the hind ones, which are each subdivided by a well-marked outer re-entrant fold. The lower cheekteeth are not unlike those of Kchimvs; there are one outer and two inner folds present. Incisors narrow. The spiny covering of the back is rudimentary, consisting of long wavy bristles only; the head is densely covered with sharp spines, which extend over the neck and forelimb. Feet as specialized as in any member of the family; four digits present on fore and hindfeet, these bearing long curved claws, the poUcx and hallux minute, replaced by a broad pad. Tail relatively long, scaly, and moderately haired, the underside, near the body, clothed with stiff bristles, as in Coendou. Forms seen : suhspinosiis. List of Named Forms (The references and type localities of all Erethizontidae are the work of Mr. G. W. C. Hoh.) I. CHAETOMYS SUBSPINOSUS, Kuhl. 1820. Beitr. Zool. Manim. p. 71. Brazil (?). Synonym: tortilis, Olfers, 1820, Neue Bibl. Reis. XV, p. 211. Brazil. moricandi, Pictet, 1843, Rev. Zool. p. 227. Brazil. Subfamily ERETHIZONTINAE Geographical Distribution. — As in the subfamily. Number of Genera. — Three. Characters. — Differing from the Chaetomyinae in the large orbit, the lack of postorbital process, the lack of extreme thickening of the jugal, the pattern of the cheekteeth, which are with wide inner and outer re- entrant folds (three outer, one inner in the upper series), and the greater develop- ment of spiny covering of the back. The feet may be highly specialized, or in Eretliison less so. The tail may be short {Erethizon, Echinoprocta), or long and prehensile [Coendou). Remarks. — Pocock in 1922 proposed to divide this group into two sub- families, Erethizontinae and Coendinae; he does not include in his key the genus Ecliiiiopructa which is precisely intermediate in the main character (the tail), between Pocock's two "subfamilies." Key to the Genera of Erethizontinae Hallux well developed, and no well-marked pad taking its place on the hindfoot; inner side of forefoot not or less expanded; tail short, non-prehensile. Erethizon llalhiv vestigial or absent, its function taken over by a broad movable pad; inner side of forefoot more expanded. 12 — Living Rodents — 1 178 ERETHIZONTINAE: ERETHIZON Tail short, non-prehensile, httle longer than hindfoot. Echinoprocta Tail long, prehensile (as far as known), much longer than hindfoot. CoENDOi; Genus i. ERETHIZON, Cuvier 1S22. ERETHIZON, Cuvier, Mem. Mus. Hist. Nat. IX, p. 425. Type Species. — Hvstrix duisatu, Linnaeus. Range. — North America; "Most of forested North America north of 40° and south in the Rocky Mountains almost to Mexican boundary" (Anthony). Forms named from Labrador, Nebraska, California, Arizona, British Columbia, Alaska. Number of Forms. — Seven. Characters. — Nasals wide, widely open anteriorly; frontals broad, strongly ridged, these ridges extending backwards to form a sharp sagittal crest. Palate narrow anteriorly, very broad behind, and short. Zygoma simple, jugal broader anteriorly. Bullae large, the external meatus produced slightly sideways. Palatal foramina medium in size. Incisors relatively thin. Mandible with low coronoid, relatively low condylar process, this thickened; the area beside the condyle with noticeable ridge presumably for attachment of masseter medialis; this short, not so pronounced as in some Chinchillidae. Angular portion distorted outwards fairly strongly, the lower border abnormally thickened. LIpper cheekteeth with one external and one internal main persistent folds, the other two outer folds (anterior and posterior) tending to isolate, and to take up most of the lobes formed by the central folds. L'sually a trace of a small posterior fold in the hack of each tooth. The lower teeth reverse the pattern of the upper series. The infraorbital foramen has no separate canal for transmission of nerve, in this respect agreeing with all other members of the subfamily. Entire body, limbs, head, tail, and sides of feet covered with thick hair which completely conceals the highly effective spiny covering below it. The spines are short, with barbed tip, and detach very easily; once sticking in an object they are sometimes quite difficult to take out (this feature common to all Erethizontinae). Tail short and bushy, covered with spines more or less throughout. Hind- foot lacking the inner pad characteristic of Coendoii, and with a well-developed hallux, which is, however, shorter than the remaining four digits; claws curved, powerful. Forefoot broad, with four functional digits. Mammae 4 (Anthony). Size relatively large; up to about 34 inches head and body. The genus is noteworthy as being the only Hystricoid adapted ior life in cold climates. Two closely allied species are admitted. Forms seen: dorsatum, epixtwthum, iiivops. Fig. 45. Erethizon epixanthum myops, Merriam. B.M. No. 4. II. 30.1; X I. Fig. 46. Erethizon epix.\nthvm mvops, Merriam. B.M. No. 4.1 1.30. 1 ■ ■; I. Fig. 47. Erethizon epixanthum myops, Merriam. B.M. No. 4. II. 30.1 ; I. Fig. 48. Erethizon epixanthum mvops, Merriam. (.'heckteeth: 4. 11.30. 1 ; zk. ERETHIZON— ECHINOPROCTA i8i List of Named Forms 1. ERETHIZON DORSATUM DORSATUM, Linnaeus 1758. Syst. Nat. i, p. 57. Eastern Canada. Synonjin: hudsonis, Brisson, 1756, Regn. Anim. Quadr. p. 128. America. Trouessart quotes as synonym : pilosus americanus, Catesby, 1 73 1, Nat. Hist. Carolina, i, xxx. 2. ERETHIZON DORSATUM PICINUM, Bangs 1900. Proc. New Engl. Zool. Club, II, p. 37. L'Anse au Loup, Strait of Belle Isle, Labrador. 3. ERETHIZON EPIXANTHUM EPIXANTHUM, Brandt 1835. Mem. Acad. St. Petersb. pi. i, p. 390. California. 4. ERETHIZON EPIX.\NTHUM BRUNERI, Swenk 1916. Univ. Studies Lincoln Nebr. vol. XVI, p. 3. 3 miles east of Mitchell, Scotsbluff County, Nebraska. 5. ERETHIZON EPIXANTHUM COUESI, Mearns 1897. Proc. U.S. Nat. Mus. XIX, p. 723. Fort Whipple, Yavapai County, Arizona. 6. ERETHIZON EPIX.\NTHUM NIGRESCENS, .-Mien 1903. Bull. Amer. Mus. Nat. Hist. XIX, p. 558. Shesley River, British Columbia, Canada. 7. ERETHIZON EPIXANTHUM MYOPS, Merriam 1900. Proc. Washington Acad. Sci. II, p. 27. Portage Bay, Alaska Peninsula, Alaska. Genus 2. ECHINOPROCTA. Gray 1865. ECHINOPROCTA, Gray, Proc. Zool. Soc. London, p. 321. Type Species. — Erethizon rufescens. Gray. Range. — Colombia. Number of Forms. — One. Characters. — Cranially and dentally not essentially different from small species of Coendou (next to be described); a sagittal crest evidently formed; f rentals not specially inflated; zygoma simple; a little con- striction noticeable in interorbital region. Size smaller than is normal in Coendou; the spines of the back long and bristly, gradually becoming thicker and stronger as they approach the rump, on which they are as strong as in Coendou. Head covered with sharp spines. Feet, including the bone formation of the specialized pad of the hindfoot (as figured by Trouessart), similar to Coendou; hallux suppressed. Tail short, little longer than hindfoot, hairy, non-prehensile. Forms seen : rufescens. List of Named Forms 1. ECHINOPROCTA RUFESCENS, Gray 1865. Proc. Zool. Soc. London, p. 322. Colombia. 1 82 COENDOU Genus 3. COENDOU, Lacepede 1 799. CoENDOU, Lacepede, Tabl. des Divisions des Mamm. p. 11. 1S25. Sphiggurus, Cuvier, Dents. Mamm., p. 256. (Sphiggure, 1S22, Mem. Mus. Nat. Hist. Paris, IX, p. 427.) (Sphiggurus spinosus, Cuvier.) V.'vlid .\s a SrB(;ENUS. 1825. SlNOETHEBis, Cuvier, Dents Mamm. p. 256. (Hystrix preheiisilis, Linnaeus.) 1835. Cercol.\bes, Brandt. Mem. Acad. St. Petersburg, C>, iii, p. 301. New name for Coendou, Lacepede. Type Species. — Hystrix preheiisilis, Linnaeus. Range. — Mexico (through Central America.'), to Panama; Venezuela, Colombia, Ecuador, Peru, Bolivia, Brazil south to Parana and Rio Grande do Sul. One form named from Chile, and one "said to be from the West Indies"; there seems reason to doubt both these localities. Number of Forms. — Twenty-nine. CHAR.'iCTERS. — Skull broad, sometimes characterized by somewhat extreme inflation of frontals (this most developed in preJiensilis group, also to a certain degree in bicolor and mexicanum); the skull in these species sloping sharply downwards in front, and more gradually so behind; in smaller species as paragayensis, the portion of the skull over the posterior zygomatic root is the highest part ; between these extremes exist intermediate forms. Nasals well open anteriorly, usually short. Parietals ridged, and a sagittal crest may be formed (this evidently a variable character). Palate wide, especially pos- teriorly; hamulars thick, usually joining the bullae, which are prominent. Palatal foramina usually relatively short. Paroccipital processes not lengthened. Jugal rather long; zygoma simple. Mandible like Eretliizon except that the lower border is usually less extremely broadened; there is a tendency in this group for the degree of distortion outwards of the angular process to be weak. Cheekteeth essentially as Eretliizon. Externally the body is covered in short thick spines, which probably do not much exceed four inches in length at highest development. Tail prehensile, so far as known; its length variable in the different species; sometimes slightly longer than the head and body, but usually rather shorter. The lower part at the end is naked, curling upwards when grasping an object. The underside near the body is covered with stiff sharp bristles, which it has been suggested perform a similar function to the caudal scales of the Anomaluridae, to assist the animal's balance when resting on a branch. The upper part of the tail near the body is spiny. The feet are very highly specialized, the pad on the hindfoot at its highest development; the claws are long and curved; both fore and hindfeet with four functional digits only; there is no very marked discrepancy in their lengths. The pads of the hindtoot are supported by a bony structure, which is well described and figured bv Water- house, 1848, Nat. Hist. Mamm., p. 405, and pi. 18, tig. 4. Some forms have the spiny covering of the back mixed with or covered by long thick fur; the hair of the chest and belly is usually in these forms less bristly, or soft. For these the subgeneric name Sphiggurus is used by Tate; it is here retained. Fig. 49. CoENDOu prehensilis boliviensis. Gray. B.M. No. 50.6.5.2; X I. Fig. 50. CoENDOu prehensilis boliviensis. Gray. B.M. No. so.6.5.2; X I. Fig. 51. CoENDOU prehensilis boliviensis, Gray. B.M. No. 50.6.5.2; , I. Fic. 52. CoENDOU PREHENSILIS BOLIVIENSIS, Gray. Cheekteeth: B.M. No. 50.6.5.2; x 3*. COENDOU i8s Forms seen: bicolur, holiviensis, centralis, "couiy" [—paragayemis), insidiosus, laetiatufn, melanurus, mexicanum, pallidus, prehensilis, pruinosus, quichua, roberii, rothschildi, simonsi, tricolor, vestitiis, villosiis, yucataniae. The genus is in need of revision; it appears to be in a more chaotic state even than is usual among these Neotropical Rodents, many "species" being apparently based on only one skin, with exact locality unknown. It appears to me to divide, broadly speaking, into four or possibly five groups. Subgenus Coendou (the spines not mixed with hairy covering) : prehensilis group : large animals (largest of genus), with frontals normally at maximum inflation for the genus; general effect silvery as regards colour, the spines white terminally (evidently main spines black terminally in all others). Long-tailed types. With prehensilis, holiviensis, which is probably a synonym, or at most a race of prehensilis, centralis, described as near brandti; brandti, which seems very near prehensilis; and tricolor, the status of which is doubtful, the type skull (broken) appears to be less arched in the region of the frontals than in allies, and the colour of the one skin seen, on the identi- fication of which there is some doubt, rather different. Perhaps this species should be placed incertae sedis. C. sanctaemartae, not seen, is described as a member of the group. bicolor group : presenting typically the following features : spines black terminally, general effect of animal dark; head and shoulders covered by a profuse mantle of moderately long thick bristle-like spines (not sharp, nor effective as weapons of defence). Relatively large; frontals markedly inflated, but less extremely than in prehensilis group. C. simonsi is evidently not more than a subspecies of bicolor. C. quichua, a smaller form, with less developed mantle on head and shoulders; the mantle- spines white-tipped in the type skin. Skull not arched in frontal region. C. rothschildi, near quichua, differing in colour. Incertae sedis species: platycentrotus, near prehensilis according to Waterhouse, but placed in neighbourhood of bicolor group by Tate; and nycthemera, stated by Waterhouse to be synonymous with bicolor, but this identification questioned by Thomas; and listed by Tate as a member of subgenus Sphiggurus. Subgenus Sphiggurus (the spines mixed with and typically covered by long woolly hair). mexicanum group. Larger, very dark types, from Central America. Typi- cally the skull considerablv inflated in the frontal region (about as in bicolor). Includes laenatuin, in which the frontals are Hat, with no trace of inflation. Goldman (Mammals Panama, Smiths. Misc. Coll. 69, 5, p. 133, 1920) states that intergradation may take place here, and refers laenatum to mexicanum as a subspecies; this indicates that too much attention should not be paid to cranial characters in this group, as the skulls of laenatum and the mexicanum (with yucataniae) examined are verv distinct from each other. i86 COENDOU paragayensis group: normally smaller lighter types. (It is not easy to give exact measurements of these species, as comparatively few of the skins examined bear measurements.) So far as seen the frontals never inflated. paragayensis (or the skins bearing the name " coniy" which accordini,' to I'ate must be regarded as a synonym of paragayensis), if identified rightly, are remarkable for the fact that the spines of the head and shoulders are exceptionally strong, and not covered by any hair, which is present, however, on the lower part of the back. The type skin of roberti is similar, but even less hairy on head and shoulders. insidiosiis (or skins bearing this name) have the head and back normally hairy; the skulls of these three last-mentioned species are very similar. Tate suggests that spinosus of Cuvier is probably a synonym of paragay- ensis. C. villosiis is probably a synonym of insidiosiis according to Water- house; it appears to be treated as such at the British Museum. C. nigricans, not seen, is considered near villosus bv Waterhouse. C. mela- niirus is a type much like the above-mentioned, but with a jet-black tail; the skull is flat. C. pallidiis, based on a young animal "said to be from the West Indies," is a similar type of animal, but much lighter coloured (albinistic?); rather short-tailed. A rather distinct section is seen in vestitus and pruiiiosus, which differ from each other in colour; both have no inflation of the frontals; the spines are of two kinds, the normal mixed with longer "bristle-spines"; covered as usual in the subgenus with thick hair. In vestitus the tail appears shorter than in any other; it is still, however, considerably longer than Echinoprocta, and partly naked, as in Coendou. Finally eliilensis, ajfinis, and sericeus are not represented in London; the first-named was said to come from Chile, but Tate, p. 299, states: "It seems improbable that any Porcupine exists in the wild state in Chile." List of Named Forms Subgenus Coendou, Lacepede prehensilis Group 1. COENDOl' PREHENSILIS PREHENSILIS, Linnaeus 1758. Syst. Xat. 10th Ed. p. 57. Brazil. (Probably near Purnainbuco.) Synoin-m; ciiiiiu/ii, Desmarest. 1S22, Ency. Mc'th. (Mamni.). 2, p. 346. Hrazil. longicaudatus, Lacepede, 1709. Tahl. des Div. dcs Mamm. p. I. Cayenne. 2. COENDOU PREHENSILIS BOLIVIENSIS. Gray 1850. .Ann. Nat. Hist. V. p. 3S0, Bolivia. 3. COENDOU CENTRALIS, Thomas 1903. Proc. ZooL Soc. London, p. 240. Chapada, Matto Grosso, Brazil. COENDOU 187 4. COENDOU BRANDTII, Jentink 1879. Notes Leydcn Mus. i, p. 96. Matto Grosso, Brazil (?). 5. COENDOU SANCTAI•:^L■\RTAE, Allen 1904. Bull. Arner. Mus. Nat. Hist. XX, p. 441. Bonda, Santa Marta district, Colombia. 6. COENDOU TRICOLOR, Gray 1850. Ann. Nat. Hist. V, p. 381. Bolivia (?). bicolor Group 7. COENDOU BICOLOR BICOLOR, Tschudi 1845. Fauna Peruana, p. 186. Woods between Rivers Tullamayo and Chanchamayo, Peru. 8. COENDOU BICOLOR SIMONSI, Thomas 1902. Ann. Mag. Nat. Hist. 7, IX, p. 141. Charuplaya, Secure River, Yungas, Bolivia, y. COENDOU QUICHUA QUICHUA, Thomas 1899. Ann. Mag. Nat. Hist. 7, IV, p. 283. Puembo, Pichincha, Ecuador. 10. COENDOU QUICHUA RICH-^RDSONI, Allen 1913. Bull. Amer. Mus. Nat. Hist. XXXII, p. 478. Esmeraldas, Ecuador. 11. COENDOU ROTHSCHILDI, Thomas ig02. .Ann. Mag. Nat. Hist. 7, X, p. 169. Sevilla Island, off Chiriqui, Panama. incertae sedis 12. COENDOU PL.ATYCENTROTUS, Brandt 1835. Mem. Acad. St. Petersb. p. 399. "Ainerica australis." 13. COENDOU NYCTHEMERA, Kuhl 1820. Beitr. Zool. Mamm. p. 71. No locality. Brazil (?). Subgenus Sphiggurus, Cuvier mexicanum Group 14. COENDOU MEXICANUM MEXICANUM, Kerr 1792. Anim. Kingd. p. 214. Mountains of Mexico. Synonym: novaehispaniae, Brisson, 1756, Reg. -Anim. p. 127. Me.iico. (For status of Brisson's specific names see Tate, Bull. Amer. Mus. N.H. LXVIII, 1935, p. 297.) liebmanni, Reinhart. 1844, Arch. Naturg. p. 241. Me.vico. 15. COENDOU MEXICANUM YUC.\TANL\E, Thomas 1902. .Ann. Mag. Nat. Hist. 7, X, p. 249. Yucatan, Mexico (probably near Izamal). 16. COENDOU L.\EN".VrUM. Thomas 1903. Ann. Mag. Nat. Hist. 7, XI, p. 381. Boquete, Chiriqui, Panama. i88 COENDOU paragayensis Group (Typical Section) 17. COENDOU PARAGAYENSIS, Oken 1S16. Lchrbuch der ZooIokIc. p. S70. Paraguay. Synon>'in: couiy, Desmarest, 1S22, Mammalogic, ii, p. 345. Brazil. 18. COENDOU SI'INOSUS, Cuvier 1822. Mem. Mus. Hist. Nat., IX, p. 433. No locality (? Brazil) (? Based on paragayensis, Oken). 19. COENDOU ROBERTI, Thomas 1902. .-^nn. Mag. Nat. Hist. 7, IX, p. 63. Rofa Nova, Parana, Brazil. 20. COENDOU INSIDIOSUS, Kuhl 1S20. Beitr. Zool. Mamm. p. 71. No locality (? Brazil). 21. COENDOU VILLOSUS Cuviir 1822. M^m. Mus. Hist. Nat. IX, p. 434. Brazil. 22. COENDOU NIGRICANS, Brandt 1835. Mem. .Acad. St. Pctersb. p. 403. Brazil. 23. COENDOU .MELANURUS, Wagner 1S42. .Archiv. fur Naturg. i, p. 360. Barra, Rio Negro, Brazil. 24. COENDOU PALLIDUM, Waterhouse 1848. Nat. Hist. Mamm. ii, p. 434. "Said to be the West Indies." (restittis section) 25. COENDOU VESTITUS, Thomas 1899. .Ann. Mag. Nat. Hist. 7, IV, p. 284. Colombia. 26. COENDOU PRUINOSUS, Thomas 1905. Ann. Mag. Nat. Hist. 7, XVI, p. 310. Montafias de la Pedregosa, Merida, Venezuela. Not allocated to group ; not seen 27. COENDOU AFFINIS, Brandt 1835. Mem. Acad. St. Petersb. p. 412. Brazil. 28. COENDOU SERICEUS, Cope 1889. Amer. Naturalist, XXIII, p. 136. Sao Joao do Monte Negro, Rio Grande do Sul, Brazil. 29. COENDOU CHILENSIS, Molina 1S09. Geogr. Nat. and Civil Hist, of Chile, p. 242. Chile. Numbers 27, 28, 29 have not been seen; they are listed by Tate as members of the subgenus Sphiggunis. DASYPROCTIDAE 189 ERETHIZONTIDAE: SPECIAL WORKS OF REFERENCE VV'ATEBHOiSE, 1 848, Natural History Mammalia, Rodentia. Tatf, 1935, Taxonomy of Neotropical Hvstricoid Rodents, Bull. Amer. Mus. Nat. Hist. LXVIH, p. 295. ' I'ocoiK, Proc. Zool. Soc. London, 1922, p. 365. External Characters of some Hystrico- morph Rodents (Coendou, Erethizon). 'i"ROl'ESS.\RT, Echiiwprocla, Bull. Mus. Hist. Nat. 1920, no. 6, p. 44S. Al-LEN, North American Rodentia, p. 385, 1876. " Hystricidae " (Erethizon). TuLLBERG, Nova Acta Reg. Soc. Sci. Upsaliensis, XVHI, 3, i, 1899. The family is known fossil from the Oligocene from America. Family DASYPROCTIDAE 1896. Thomas: Hystricomorph.\ : Family Dasyproctidae, part, included Cuniculus (^" Coelogenys"). 1899. Tullberg: Hystricomorpha : Family Caviidae, part. 1918. Miller & Gidley: Hystricoid.\e : Family Dasyproctidae. 1924. Winge; Family Hystricidae, part, Dasyproctini, part, Dasyproctae, part (included Cufiiculus). 1928. Weber: Hystricoidea : Family Caviidae, part, subfamily Dasyproctinae, part (included Citniculus). Geogr.\phical Distribution. — Tropical America, from Mexico through Central America to Bolivia and Paraguay, Ecuador and Peru. Trinidad. Lesser Antilles. Number of Genera. — Two. Characters. — E.xternal form much modified for cursorial life; hindlimbs lengthened; hindfeet with three digits, forefoot with four functional digits, the claws hooflike. Cheekteeth semi-rooted, extremely hypsodont, closely paralleling the structure present in the Hystricidae. Clavicles undeveloped. Remarks. — This family has (together with Cuniculus) often been united with the Caviidae. This association appears most unnatural. The lower jaw is totally distinct in the two groups, Dasyprocta being typically Hystricoid in this formation; the cheekteeth show an entirely different pattern in the two groups. The similarities between such genera as Dasyprocta and Dolichotis in the arrangement of digits and parts of the skeleton for swift running appear to be parallel evolution brought about by a similar mode of life, comparable to the similarities between such types as Bathyergus and, say, Geomys, which resemble each other externally to a large degree and yet in which the zygomasseteric structure and the cheekteeth are totally different. One of the reasons which has been advocated for classing Dasyprocta with the Caviidae is the formation of the penis, which is said to be armed with a pair of horny spikes in these genera; it is therefore interesting to note that according to Pocock the penis of Dolichotis and ot Hydrochoerus, both members of Caviidae, lack these spikes, disagreeing in this character from Cazia and other members of the Caviidae, as well as from iQO DASVPROCTIDAE: DASYPROCTA Dasypructti and Cuiiiciiliis. But in any case the penis does not furnish a suffi- ciently rehable character on which to base family distinctions. Key to the Genera of Dasyproctidae Tail not obsolete, approaching half length of hindleg; toothrow reduced, "teeth smaller both relatively and absolutely than in any species oi Dasyprocta" (Thomas). ^IY0PR0CTA Tail obsolete; teeth relatively larger, and toothrow less reduced. Dasyprocta These two genera are not well marked, and might be regarded as subgenera of one genus. Mxoprocta contains much smaller forms than is normal in Dasyprocta. Ccnus I. DASYPROCTA, Illiger iSii. Dasyproct.a, Illiger, Prodr. Syst. Mamm. et Avium, p. 03. Type Species. — Mus aguti, Linnaeus. Range. — As in the familv Dasyproctidae; south to South Brazil. Number of Forms. — About forty-six are named. Char.\CTERS. — The skull is less ridged than in other Hystricoids of a similar size; the nasals shorter than the frontals, which are broad, flat, and with a well-marked though short postorbital process at suture of frontals and parietals. A weak short sagittal crest is developed in the adult. The bullae are moderately large ; the paroccipital processes prominent, though not so length- ened as in CiDiiciiliis. The palate is straight, and extends back to level of M.3; the hinder part being formed much as in the Hystricidae. Palatal foramina short, far in front of toothrow. Lachr\'mal much enlarged, forming most of upper zygomatic root, and part of the lachr\-mal canal is open on the side of the rostrum, immediately in front of the anterior part of toothrow. There is no canal for nerve transmission in the infraorbital foramen, which is of medium size. Zygoma generally simple. Mandible with angular portion powerfully distorted outwards, and its lower border slightly drawn backwards; coronoid process low; condylar process rather broad. Cheekteeth strongly hypsodont; like those of the Hvstricidae in essential pattern; one more or less persistent narrow inner fold in the upper series; the outer folds soon isolate as islands, and there is a tendency tor the islands to divide on the surface of the tooth, so that there may be seven or eight or more minute islands in a worn tooth. Lower cheekteeth like the upper series, but with the pattern reversed. Incisors relatively thin, compressed. Externally the form is slender, cursorial, the hindlimbs lengthened, the hindfoot very long and narrow, with three digits which bear sharp hoot-like claws; the central digit is the longest, D.4 is a little shorter than D.2; the sole is naked; in the skeleton of the foot the metatarsal bones for the outer digits are absent or vestigial. The forefoot is less elongated than the hindtoot; the digits are four, but the appearance of the foot is perissodactyie owing to D.5 being considerably reduced; the pollex is represented by a knob. The fur on DASYPROCTA 191 the hinder part of the body is very long and thick. The ears are of medium size. The tail is obsolete. The head and body measurement may approach 580 mm. Forms seen : aguti, azarae, boliviae, catrinae, cayennae, coibae, cristata, croconota, flavescens, fuliginosa, isthmica, lucifer, liinaris, maraxica, nigra, para- guayensis, pandora, prymnolopha, punctata, ruatanica, rubrata, variegata, yungarum. FlC. 53. D.\SYPROCTA PUNCTATA ISTHMICA, Alston. B.M. No. 98. 1 1.6. 10; X I. List of Named Forms (References and type localities for all forms of Dasyproctidae are the work of Mr. G. W. C. Holt.) Tate divides the genus into three sections: " Eastern or red-rumped Agoutis," "Central American Agoutis," and "Dark-grey Agoutis." The material ex- amined does not support these divisions; moreover, I have been quite unable to Fig. 54. Dasyprocta punctata isthmica, .Alston. B.M. No. 98.11. 6.10; y I. Fig. 55. D.^YPROCTA punctata isthmica, Alston. ChecktL-eth: B.M. No. 98.11.6.10; • 4. (In the lower jaw the anterior tooth is the much worn milk molar — seen in profile in fig. 5 4 ; in the upper jaw the anterior tooth is the newly cut premolar, and the larce inner root of the shed milk molar is seen to the right or inner side of this tooth.) I DASYPROCTA 193 get this large and unwieldy genus into any definite order, and therefore list geographically. The real dark-grey types like colombiana or fuliginos a grade quickly into dark unicolorous types, which in turn grade into reddish-rumped types, which seem to grade into the red types like aguti. Sometimes a blackish middorsal area may be present, as mprymnolopha. How many species should be recognized I am not prepared to say, but it seems clear that far too many are at present standing, and many could be reduced to the rank of subspecies. It is to be hoped that someone will attempt a revision of this genus, which is much needed. 1. DASYPROCTA NOBLEI, M\en 1914. Proc. New. Engl. Zool. Club. V. p. 6g. Goyave, Guadeloupe, Lesser Antilles. 2. DASYPROCTA ALBIDA, Gray 1842. Ann. Mag. Nat. Hist, i, X, p. 264. St. Vincent, Lesser Antilles. 3. DASYPROCTA ANTILLENSIS, Sdater 1874. Proc. Zool. See. London, p. 666. St. Lucia, Lesser Antilles. 4. DASYPROCTA RUBR.ATA RUBRATA, Thomas 1898. Ann. Mag. Nat. Hist. 7, H, p. 272. Savannah Grande, Trinidad. 5. DASYPROCTA RUBR.^TA FL.WESCENS, Thomas 1898. Ann. Mag. Nat. Hist. 7, H. p. 273. Caripe, Cumana, Venezuela. 6. DASYPROCTA LUCIFER LUCIFER, Thomas 1903. .-^nn. Mag. Nat. Hist. 7, XI, p. 491. Caicara, Rio Orinoco, Venezuela. 7. DASYPROCTA LUCIFER C.^YEN.NAE, Thomas 1903. .\nn. Mag. Nat. Hist. 7, XI, p. 492. Approvague, Cayenne. 8. DASYPROCTA CAY'ANUS, Lacep^de 1802. Tabl. des Div. des Mamm. p. 78. Cayenne. 9. DASYPROCTA PRYMNOLOPHA, Wagler 1831. Isis, XXIV, p. 619. Guiana. 10. DASYPROCTA NTGRICLUNIS, Osgood 1915. Field. Mus. Nat. Hist. Publ, Zool. ser. X^ p. 192. Sao Marcello, upper Rio Preto, Bahia, Brazil. 11. DASYPROCTA CROCONOTA, Wagler 1831. Isis, XXIV, p. 618. Amazon River, Brazil (mouth of Rio Madeira). 12. DASYPROCTA AGUTI AGUTI, Linnaeus 1766. Syst. Nat. 12th. ed. p. 80. Brazil. 13 — Living Rodents — I 194 DASYl'ROCTA (D. aguti aguti) Synonym: (?) leporina. Linnaeus, 1758, Syst. Nat. 10th. ed. p. 59. Unknown (probably unidentifiable, according to Tate.) bicolor, Boddaert, 1785, Elenchus Anun. p. 103. 13. DASYPROCTA AGUTI MARAXICA, Thomas 1923. Ann. Mag. Nat. Hist. 9, XII, p. 341. Marajo Island, Amazon River, Brazil. 14. DASYPROCTA AGUTI LUNARIS, Thomas 1917. Ann. Mag. Nat. Hist. 8, XX, p. 259. Moon Mountains, British Guiana. 15. DASYPROCTA AZARAE AZARAE, Lichtenstein 1823. Doubl. Zool. Mus. Berlin, p. 3. Sao Paulo, Brazil. 16. DASYPROCTA AZARAE C.ATRINAF.. Thomas 1917. Ann. Mag. Nat. Hist. 8, XX, p. 311. Santa Catharina, Southern Brazil. 17. DASYPROCTA AUREA, Cope 1889. Amer. Naturalist, p. 13S. Chapada, Matto Grosso, Brazil. 18. DASYPROCTA CAUDATA, Lund 1841. Afh. K. Danske. Vid. Selsk. 4, viii, p. 286. Rio das Velhas, Minas Geraes, Brazil. ig. DASYPROCTA PARAGUAYENSIS, Liais 1872. Climats, Geologic. Faune et Geographic Botanique du Bresil, p. 536. Paraguay. Synonym: felicia, Thomas, 1917, Ann. Mag. Nat. Hist. 8, XX, p. 310. Near Concepcion, Paraguay. 20. DASYPROCTA MEXICANA, Saussure i860. Rev. Mag. Zool. 2, XII, p. 53. "Hot zone of Mexico," probably in State of Vera Cruz. 21. DASYPROCTA PUNCTATA PUNCTATA, Gray 1842. Ann. Mag. Nat. Hist, i, X, p. 264. Realejo, west coast of Nicaragua. 22. DASYPROCTA PUNCTATA RICHMONDI, Goldman 191 7. Proc. Bio!. Soc. Washington XXX, p. 114. Escondido River, 50 miles above Bluefields, Nicaragua. 23. DASYPROCTA PUNCTATA CHIAPENSIS, GoIdni;in 1913. Smiths. Misc. Coll. LX, no. 22, p. 13. Huehuetan, Chiapas, Mexico. 24. DASYPROCTA PUNCTATA YUCATANICA, Goldman 1913. Smiths. Misc. Coll. LX, no. 22, p. 12. Apazote, Campeche, Mexico. 25. DASYPROCTA PUNCTATA UNDERWOODI, Goldman 1931. Journ. Washington Acad. Sci. XXI, p. 481. San Geronimo, district of Pirris, West Costa Rica. 2fa. DASYPROCTA PUNCTATA DARIENSIS, Goldman 1913. Smiths. Misc. Coll. LX, no. 22, p. 11. Near head of Rio Limon, Mt. Pirri, Eastern Panama. DASYPROCTA 195 27. DASYPROCTA PUNCTATA NL'CHALIS, Goldman 1917. Proc. Biol. Soc. Washington XXX, p. 113. Divala, Chiriqui, Panama. 28. DASYPROCTA PUNCTATA ISTHMICA, Alston 1876. Proc. Zool. Soc. London, p. 347. Colon, Panama. 29. DASYPROCTA RU.^TANICA, Thomas 1901. Ann. Mag. Nat. Hist. 7, VIII, p. 272. Ruatan Island, Bay of Honduras. 30. DASYPROCTA CALLIDA, Bangs 1901. Amer. Naturalist, XXXV, p. 635. San Miguel Island, Panama. 31. DASYPROCTA COIBAE, Thomas 1902. Nov. Zool. IX, p. 136. Coiba Island, Panama. 32. DASYPROCTA PANDORA. Thomas 1917. Ann. Mag. Nat. Hist. 8, XX, p. 313. Gorgona Island, off Colombia. 33. DASYPROCTA VARIEGATA VARIEG.\TA, Tschudi 1845. Fauna Peruana, p. 190. Chanchamayo region. Eastern Peru. 34. DASYPROCTA VARIEGATA ZAMORAE, Allen 1915. Bull. Amer. Mus. Nat. Hist., XXXIV, p. 627. Zamora, Eastern Eucador. 35- DASYPROCTA VARIEGATA CHOCOENSIS, .\llen 1915. Bull. Amer. Mus. Nat. Hist. XXXIV, p. 627. Los Cisneros, Choco district, Colombia. 36. DASYPROCTA VARIEGATA COLOMBIANA, Bangs 1898. Proc. Biol. Soc. Washington XII, p. 163. Santa Marta, Colombia. 37. DASYPROCTA VARIEG.A.TA VUNGARUM, Thomas 19:0. Ann. Mag. Nat. Hist. 8, VI, p. 505. Chimosi, Vungas, Bolivia. 38. DASYPROCT.-\ VARIEG.VPA nOI.IVI.\E, Thomas 1917. Ann. Mag. Nat. Hist. 8, XX, p. 312. Charuplaya, Bolivia. 39. DASYPROCTA VARIEG.'^TA URUCUMA, Allen 1915. Bull. Amer. Mus. Nat. Hist. XXXIV, p. 634. Urucum, near Curumba, Matto Grosso, Brazil. 40. DASYPROCTA FULIGINOSA FULIGINOSA, Wagler 1832. Isis, XXV, p. 1220. Near .Amazon River, Brazil. (Borba, Rio Madeira.) Synonym: nigra, Gray, .^nn. Mag. Nat. Hist, i, X, p. 264, 1842. nigricans, Wagner, 1842, .Archiv. fur Naturg. i, p. 362. Borba, R. .Madeira, Brazil. caroliniensis, Cuvier, Ger\ais, Mamm. i, 1854, p. 329. 196 DASYPROCTA— MVOPROCTA ^i. DASYPROCTA FULIGINOSA CANDELENSIS, Allen IQ15. Bull. Amer. Mus. Nat. Hist. XXXIV, p. 625. La Candcia, Huila, Colombia. 42. DASYPROCTA FULIGINOSA MESATIA, Cabrera 1917. Madrid Trab. Mus. Nac. Ci. Nat. 31, p. 53. Tarapote, Ecuador. 43. DASYPROCTA CRISTATA, Desmarest 1816. Nouv. Diet. d'Hist. Nat. 2d. Ed. i, p. 213. Surinam, Dutch Guiana. 44. DASYPROCTA KALINOWSKII, Thomas 1897. Ann. Mag. Nat. Hist. 6, XX, p. 219. Idma, valley of Santa ."Vna, Cuzco, Peru. Genus 2. MYOPROCTA, Thomas 1903. Myoprocta, Thomas, Ann. Mag. Nat. Hist. 7, XH, p. 464. Type Species. — Cavia acoucliw Erxleben. Range. — Guianas, Brazil, Colombia, Ecuador, Peru. Number of Forms. — Ten. Ch.\r.\cters. — Like Dasxprocta, but smaller (about 380 or less, head and body) ; the tail less reduced, approaching about half the length of the hindleg, slender, hairy. Essential cranial and dental characters as in Dasyprocta, but toothrow reduced, the teeth "smaller both relatively and absolutely than in any species of Dasyprocta" (Thomas); the sagittal crest appears to be in old individuals rather longer than in Dasyprocta. There may be a small backwardly directed process on anterior portion of jugal, just behind its point of junction with zygomatic process of maxillary. Forms seen: acoitcliv, cavmainiin, lepttira, liiiuiniis, milleri, pratti, puralis. There appears some doubt on the status of exilis. Apart from this the forms divide into two groups, the type species, reddish above, and the pratti group, duller greenish types. List of N.\med Forms (acoucliv Section) 1. MYOPROCTA ACOLXHY, Erxleben 1777- Syst. Regn. Anim. p. 354. Cayenne. 2. MYOPROCTA LICPTURA, Wapner 1844. Schreber Saug. .Suppl. IV, p. 49. Rio Negro, Brazil. (pratti Section) 3. MYOPROCTA PRATTI PRATTI, Pocock 1913. Ann. Mag. Nat. Hist. 8, XII, p. no. Pongo de Rentema, Rio Marafion, Peru. MYOPROCTA— HYSTRICIDAE 197 4. MYOPROCTA PRATTI ARCHIDONAE, Lonnberg 1925. Journ. Manim. Baltimore, VI, p. 274. Archidona, Province Oriente, Ecuador. 5. MYOPROCTA PRATTI PLKALIS, Thomas 1926. Ann. Mag. Nat. Hist, g, XVII, p. 639. .■\yapua, about 300 kilometres south-west of Manaos, Brazil. 6. MYOPROCTA PRATTI CAY.MAXLM, Thomas 1926. .^nn. Mag. Nat. Hist. 9, XVII, p. 638. Canabouca, Parana de Jacare, 120 kilometres south-west of Manaos, Brazil. 7. MYOPROCTA PR.-\TTI LIMANUS, Thomas 1920. .\nn. Mag. Nat. Hist. 9, VI, p. 279. Acajutuba, Rio Negro, Brazil. 8. MYOPROCTA MILLERI, .'^llen 1913. Bull. .\mer. Mus. Nat. Hist. XXXII, p. 477. La Murelia, Caqueta, Colombia. Not allocated to section ; not seen 9. MYOPROCTA EXILIS EXILIS, Wagler 1831. Isis, XXIV, p. 621. .\mazon, Brazil. 10. MYOPROCTA EXILl.S PARVA, Lonnberg 1921. Ark. Zool. XIV, no. 4, p. 41. Rio Curaray, Prov. Oriente, Ecuador. DASYPROCTIDAE: SPECIAL WORKS OF REFERENCE Waterhovse, 1848, Natural History' Mammalia, vol. II, Rodentia. Tate, 1935, Bull. .\mer. Mus. Nat. Hist. LXVIII, p. 295. Ta.\onomy of Neotropical Hystricoid Rodents. PococK, Proc. Zool. Soc. London, p. 365, 1922. External Characters of some Hystrico- niorph Rodents. The Dasyproctidae are known fossil from the Miocene, from the Neotropical region only. Family HYSTRICIDAE 1896. Thomas: Hvstricomorpha : Family Hystricidae. 1899. Tullberg: Hvstricomorpha: Family Hystricidae. 1918. Miller & Gidley: Superfamily Hystricoidae: Family Hystricidae, with sub- families Hystricinae and Atherurinae. 1924. Winge: Family Hvstricid.ae, part, Hystricini, part, Hystrices. 1928. Weber; Hvstricoidea: Family Hystricidae. Geographical Distribution. — Tropical parts of the Old World ; the greater part of the African Continent; Italy and Sicily; Transcaucasia; Southern Palaearctic Asia (Arabia, Syria, Persia, Sleso- potamia, .\fghanistan, Russian Turkestan); Peninsular India and Cevlon; igS HYSTRICIDAE South China (south of the Yangtsekiang), Hainan, Assam, Burma southwards to Alalacca, Sumatra, Java and islands to the east of it (Sumbawa, Flores); Borneo; represented in the Phihppine Islands. Number of Genera. — Four. Ch.\r.\cters. — E.xternal form heavy, terrestrial-fossorial; modification of hair into spiny covering always well developed, at extreme development reaching a grade of specialization not seen elsewhere in the Order; tail always bearing a group of modified quills or bristles; digits of hindfoot five. Cheekteeth moderatelv to e.xtremely hypsodont, semi-rooted in progressive genera, the re-entrant folds isolating early on crown surface as narrow enamel islands; bullae relatively small, and paroccipital processes not lengthened. Occipital region of skull strongly ridged, prominent; zygoma simple; a tendency present towards extreme inflation and lengthening of nasals; clavicles imperfect; habits entirely terrestrial. According to Tullberg, the carpus has no free centrale [Atherurus, Hystrix), unique in the Order among those examined by him except in Cuniculidae. According also to this author, the lungs are abnormal, being divided into a number of small lobes. Rem.\rks. — Lyon in 1907 proposed to divide the family into two subfamilies, the Hystricinae and the Atherurinae, on account of the differ- ences of the length and structure of the tail between the two groups, the fact that there are only three sacral vertebrae present in Atherurinae as against four in the Hystricinae (at the same time remarking that there is apparently some variation in the number of the vertebrae, especially lumbar, sacral, and caudal). But although the Atherurus group is much less specialized than the Hxstrix group both in spinv covering, reduction of tail, and in the more brachyodont cheekteeth, there appear to be too manv essential characters common to both groups for this division to be maintained; the pattern of the cheekteeth and the structure of the feet, for example, are essentiallv similar in both groups; and the cranial characters of Thecurus, the lowest member of the Hvsirix group, are very similar to those of Atherurus, including the rather important character of length of nasals. Cheekteeth. — The cheekteeth vary individually, but the essential pattern throughout the family is, in the upper series, one inner and three outer folds, the folds isolating as islands almost immediately on the flat crown surface ; like the Dasvproctidae, there is a strong tendency for the isolated folds to divide, particularly the posterior one, so that on the outer side of the tooth there are usuallv at least four isolated islands in the adult. The lower cheekteeth reverse the pattern of the upper series. The milk premolars are shed comparatively late in life. Gener.\l Extern.al Ch.ar.acters of principal species, as regards the development of spiny covering. For note on details of formation of hollow "rattling-quills" (caudal quills) of fivstn'x group, see p. 208. HYSTRICIDAE 199 Trichvs lipura is the most primitive species in development of spiny cover- ing, in all respects. No true quills are developed; the body is covered with relatively short flattened weakly developed spines; the tail is more or less naked, and bearing a brush of unmodified bristles at the end; the head is hairy. Atherurus macrourus presents the next stage of development; the body is covered with spines of a similar nature, but sharper, longer, and evidently more effective as weapons of defence; the head is similar to Trichvs; the tail long, less naked, covered with spinv short hairs, and the end bears a cluster of much more specialized bristles, these being alternatively expanded and contracted; but, as far as seen, no quills are as yet developed. Atherurus africanus and related African forms present a higher stage of development in that among the spines of the back there are present a few thick circular quills, of the type found in all higher Porcupines; these var\' in their development; thev mav be quite strong or, in some skins I have seen, verv weak, so that perhaps if a really large series of skins came to hand from Africa it might be that some "quill-less" ones would be among them. Otherwise the external characters are as in the Asiatic Atherurus. It is interesting that, judging bv a specimen of Atherurus at the London Zoological Gardens, verv much the same noise can be made by the rattling of the tail bristles as that produced bv the smaller species of Hystrix. Thecurus pumilis, from the Philippines, to which T. sumatrae evidently bears a close resemblance in general spine characters, appears to be the lowest true Porcupine; in this species, as in all the Hystrix group, the tail has become strongly reduced and its end bears a cluster of small and very poorly developed hollow "rattling-quills" which reach such a high stage of specialization in the higher Porcupines ; some of these tend to be closed at the tip. A certain number of short thick true quills are developed; the spines of the body e.xtend to the rump from the upper part of the back ; the head is hairy, much as in the Brush-tailed Porcupines. Hystrix {Acanthion) jaraniann represents the next stage; the caudal quills are very weakly de\eIoped, essentially as in Thecurus pumilis; the head remains haip*-; the quills of the back are perhaps slightly more developed than in Thecurus pumilis; judging by the few skins seen they appear, as in Thecurus, to be tightly wedged in the body, so that I should imagine they are very infrequently shed, but I have not seen this animal in cap- tivity; on dried skins they do not give to the touch as do the quills of most higher Porcupines. Thecurus crassispinis from Borneo stands rather alone in development of spines, and presents a mixture of generalization mixed with extremely high specialization. The quills of the back are enormously thick, rela- tively as thick as those even of the most highlv developed forms of Hystrix, or so it seems to me. But no long thin quills are developed to cover them, as they are in Hystrix cristata and leucura groups; the head !oo HYSTRICIDAE remains hairy, with no trace of a crest; and the caudal quills remain at their lowest development. Hystrix (Acniitliioii) hodgsoiu represents a stage of development typically not very much higher than in jiivaniciim; the quills, though not as well developed as in Thecurus crassispiiiis, are profuse and well developed, less tightly wedged in the body apparently than m javanicum; there is a certain growth of long hair-like quills on the back, not met with in those described heretofore; a vestigial crest may be present (or suggested) on the head ; but the caudal quills remain very poorly developed. Whether certain intergradation takes place between this and such forms as klossi from the same area I do not know; there seems to be a rather distinct difference as a rule between skins seen of Iiodgsoni, as compared with klossi, as regards crest, caudal quills, etc. ; comparable to the difference seen between jai-iiniciim and biachvunis. Hystrix {Acanthiou) bracliyurus (with lungicauda, miilleri) reaches a rather higher state of specialization; the caudal quills are as a rule larger than in hodgsoni, more open, and apparently less primitive; the quills of the body are powerful and profuse, though not attaining any great length. Hvxtrix (Acanthiou) klossi is very similar in external characters to hrachyurus, though on cranial characters belonging to a different group; there are a few long liair-like quills present, as in hodgsoni; the crest tends to become less abortive, and quite well marked; this tvpe, I believe, leads on to the Chinese Porcupine sidKiistcitiis, in which the crest is said to be quite well developed, but skins of which I have not seen unless a very small juvenile labelled '' viiniuiiioisis," from Annam, which has for its size surprisingly developed caudal quills and quite conspicuous crest, represents this species, as from these characters I suspect it may do. Hvstrix leuciira (with hirsutirostris) marks the highest development to be attained in a Porcupine; a long crest of hairs is present on the head; the quills are exceedingly profuse on the back; the short ones found in the preceding species being more or less covered by an outgrowth of long thin quills, each with several rings instead of only one as in the above species. The caudal quills are large, well open, and at their highest development ; there are many short white (ordinary) quills in the neighbourhood of the tail. Sometimes the long quills tend to be narrower than in cristata and other African Porcupines. The bodily size is usually larger than in other .Asiatic Porcupines. Hystrix cristata, H. galeata, H. africaeaiistralis are indistinguishable from one another externally; the quills may tend to be thicker, perhaps longer and more powerful than in leucura, otherwise the external covering is essentially similar, including the long crest and powerful tail-quills; the size, in galeata, becomes the largest in the genus. For the last thirteen vears these animals have been a special hobby ot the author, in the London Zoological Gardens, and a tew words on their captivity habits may not be amiss. The temperament of //. cristata compared with the smaller //. hrachyurus HYSTRICIDAE 201 type of animal is as different as that of a dog from a cat. The smaller Porcupines never display, so far as I have observed them, the slightest nervousness, and generally seem to tame down and come to hand almost on arrival, and to be the most friendly of animals, though occasionally exhibiting an unpleasant streak in their character which I have only once observed in cristata (an old specimen newly imported from aiiroad which might have been ill-treated). On the other hand all cristata Porcupines I have ever seen are most abnor- mally nervy animals, extremely hard to tame; it is nothing, for instance, for a specimen of this kind to take sixteen weeks before ever feeding from hand. But once they get over their first nerves, and brace their courage sufficiently to come to hand, they are most engaging animals, with an excellent temperament, and, I think, with a good memory for people; although it has been my experience that they never completely lose their distrust, so that the least thing outside routine, such, for instance, as a sudden movement, or a sneeze, will send them scurrying for shelter in a panic, and if it is an animal which does not know one well, it will be some time before the animal can be coaxed back. One cristata only I have known who allowed himself to be stroked. This was an individual who seemed to delight in being scratched and rubbed all over; on this animal, the belly, shoulders, and limbs are covered with bristles which are not harsh to the touch; the skin of the back when reached through the mass of quills is completely naked, flesh-coloured, soft and velvety. This nakedness is probably not a constant character; I have noticed a certain growth of hair under the quills of the back on other Crested Porcupines, which is also present in //. hodgsoni. H. cristata definitely sheds the quills much more freely and frequently than the hodgsum-brachviirus type of animal; and on occasion can do very much more damage with them; but any Porcupine will draw blood immediately (even accidentally), or if touched or handled when not in the mood, and speaking from personal experience it can be agonizingly painful. Moreover, they can attack with their spiny covering by running sideways or backwards into their enemy, usually leaving some of the quills in anything they run into. The quills in cristata may reach extreme length; I once had one in my possession measuring twenty-one inches. But it is not these, but the short thick quills which do the damage. The young have sharp little bristles at birth; and these will develop into sufficiently sharp spines within ten days to make handling impossible. In captivity there are one or at most two young in a litter, so far as my experience goes. H. hodgsoni, which has the caudal rattling quills poorly developed, does not use them except in moments of great excitement, and even then the noise produced is feeble; but in cristata, in which these hollow quills are at maximum development, they are normally used con- stantly, though I have known more than one specimen which appeared quite unable to make any sound with them at all. The sound produced by cristata is entirely different from that produced by hodgsoni, and in moments of anger or excitement it can be terrific; a noisy motor-bicycle is the only thing to which I can compare it; but I have never been able to ascertain whether all this noise is caused by the quills, or whether the animal roars at the same time. It must be stated, not for the first time, that the belief that these animals shoot their 202 HYSTRICIDAE quills is a mvth. But the origin of that story may be explained as follows: a Porcupine will normally be carrying a few loose quills on the body (sometimes these may even be picked out in very tame specimens, as in "Joe," a famous hodgsoni {or javcviicum;) which lived in the London Zoo for about twelve years or more). When the animal wakes up suddenly, he shakes himself, and on rare occasions one of these loose quills is shot out and hurtles across the cage, giving the effect that the animal has shot it. The captivity lite is good; best of all Rodents, according to Flower's valuable paper on the subject; in this paper a specimen is mentioned which attained the great age of twenty years; but normally I suppose twelve or fifteen would be the absolute limit. It may be added that their gnawing powers are prodigious and that they use extreme ingenuity on certain occasions; for instance, I once saw one trying to shake open a door; after some ineffective pawing attempts, he ran backwards and then took a run at the door, just as a human being might do, which I thought showed high reasoning power. Key to Generic Groups Tail relatively long, its end bearing a tuft of bristles; cheekteeth rooted, more brachyodont; spiny covering not highly developed (thick circular quills on back most often absent); anterior zygomatic root over P.4. .Atherurus Group. Atheruri Trichys ; Atheruriis Tail short, bearing a cluster of highly modified hollow quills; cheekteeth usually strongly hypsodont, semi-rooted; spiny covering highly developed, always with thick circular quills present on back; anterior zygomatic root over middle of toothrow. Hystrix Group. Hystrices Tlieciirus, Hystrix The Alhenirus Group Ch.\r.^cters. — .\s indicated in the above key. The size is relatively smaller usually than in the Hvstrix group; the spiny covering less specialized. Skull with no inflation of nasals, and usually no inflation of frontals. The thick quills on the back, characteristic of true Porcupines, are not developed in Trichys nor in the Asiatic section of Atherurus; in the African section of the latter genus they are usually present. This indicates a higher grade of specializa- tion for the African forms of this genus, which is paralleled by the Hystrix group, w^hich are at their lowest in the Indo-Malayan region, and at their highest in Africa. Key to the Gener.\ of the Atherurus Group Skull with well-marked postorbital process, and strong interorbital con- striction; lower incisors more compressed; tail long, scaly, its tip bearing a cluster of parallel-sided bristles; body clothed with flexible spines; a prominent horizontal groove on surface of jugal. Trichys TRICHYS 203 Skull with scarcely marked postorbital process (or this absent), and little interorbital constriction; lower incisors not compressed; tail long, less naked, its tip bearing a cluster of bristles which are alternately expanded and contracted ; body clothed with sharper bristles, some- times mixed with quills; no groove on surface of jugal. Atherurus Genus i. TRICHYS, Gunther 1876. Trichys, Gunther, Proc. Zool. Soc. London, p. 739. Type Species. — Trichys lipura, Gunther. R.\NGE. — Sumatra, Borneo, and southern Malay Peninsula (Malacca, Perak, etc.). Number of Forms. — Two. Characters. — Skull rather long and narrow; a prominent sagittal ridge present in adult; well-marked postorbital processes present, behind which skull is considerably constricted; nasals short, narrow, shorter than the frontals, extending back to anterior zygomatic root. Rostrum slender. Occipital region high, strong; paroccipital processes not lengthened; bullae relatively small. Palate straight, moderately wide; hamular process long; palatal foramina very short, situated far in front of toothrow. Infraorbital foramen relatively small for a Hystricoid Rodent; no canal for nerve transmission. Zygoma simple; jugal long, though not extending to lachrv'mal, a prominent groove running along it, on exterior border, throughout most of its length. Mandible with angular portion well distorted outwards; the hinder part of the jaw flattened, as in all Hystricidae (i.e. no backward pro- longation of angular process). Coronoid high, nearly as high as condyle in type skull. Cheekteeth as already described, rooted; toothrow short. Externally covered with fiat grooved spines scarcely more developed than in some Neotropical Echimyidae; head and underparts hairy. Tail moderately long, poorly haired, scaly, its end bearing a cluster of straight bristles. There is a tendency present for this animal to lose the tail during life. Forefoot rather broad, with four well-developed digits bearing thick claws, the two central digits slightly longer than the outer ones. PoUex rudimentar}-. Hindfeet similar to the forefeet, but longer, and the hallux more developed than the pollex ; otherwise digits like those of forefoot. Size smaller than is usual in the family. Forms seen : lipura, macrotis. The two species are verv' closely allied, and might be considered as not more than races; their chief difference lies in the size of the ear. List of Named Forms (References and type localities of all Hystricidae are the work of Mr. G. W. C. Holt.) Fig. 56. Trichys macrotis, Miller. B.M. No. 16. II. 16-2; >'• I- Fig. 57. Trichys m.acrotis, Miller. B.M. No. 16. II. 16. 2; :, I. I TRICHYS— ATHERURUS 205 1. TRICHYS LIPURA, Gunther 1876. Proc. Zool. Soc. London, p. 739. Borneo. Synonym: guentheri, Thomas, 1889, Proc. Zool. Soc. London, p. 235. Kina Balu, Borneo. 2. TRICHYS MACROTIS, Miller 1903. Proc. U.S. Nat. Mus. XXVI, p. 469. N.-W. Sumatra, Tapanuli Bay. Genus 2. ATHERURUS, Cuvier 1829. .ATHERURUS, Cuvier, Diet. Sci. Nat. LIX, p. 483. Type Species. — Hystrix macrourus, Linnaeus. (See Lyon, Proc. U.S. Nat. Mus. XXII, 1907, p. 584). Range. — Indo-Malayan : Southern China, Tongking, Hainan, Assam, Tenasserim, Malay Peninsula, Sumatra. (Other Malay Islands ?) Africa: Nigeria, Fernando Po, Sierra Leone, Senegambia, Congo, Uganda, Kenya. (In China occurring as far north as Szechuan.) Number of Forms. — Thirteen. Characters. — Skull with short nasals, extending hack about to anterior margin of infraorbital foramen; frontals long, becoming arched and slightly inflated in africanus ; usually no postorbital process, but this can be slightly marked, and usually ver}' little interorbital constriction. Sagittal crest formed in adult. Zygoma simple, but jugal broader anteriorly, and tending to be longer in African species than in Malayan ones; in manv cases it reaches the lachrymal in African group. Bullae relatively small. Palate essentially as Trichys. Occipital region of skull as in Trichys. Mandible with low coronoid, rather low condyle; angular process moderately to weakly dis- torted outwards; traces of a ridge beside the condyle, as found in Erethizontidae and Chinchillidae, may be present. Infraorbital foramen without canal for nerve transmission; relatively small for a Hystricoid Rodent, particularly in africanus, in which the frontal inflation is present. Palatal foramina much reduced. Cheekteeth as described above; rooted. Externally covered with spines, stronger and more pow^erful than in Trichys; head hairy. Tail moderately long, with short spiny hairs on the scales, and a thick tuft of bristles at the end, the bristles alternately expanded and con- tracted (up to about five times or more); in the African species these expansions tend to be nearer to each other and rather more numerous than in the Indo- Malayan type. Feet essentially as in Trichys. Head and body length up to 525 mm. or perhaps more. These may be described as much faster-moving animals than members of the genus Hystrix, judging by specimens observed in the London Zoological Gardens. In the African species the thick round quills, characteristic of all the mem- bers of the Hystrix group, are present; these vary in their development from scarcelv traceable to rather strong; it may be that with a large series, African 2o6 ATHERURUS types might come to hand which lack tiicm. They are always absent in the Indo-Malayan forms so far as I have examined. Forms seen : africaiius, assomensis, biirrozvsi, centralis, macrourus, sterensi, tionis, turiieri, zygomaticiis. The terms are here provisionally treated as two distinct groups, based on the presence (African species) or absence (Malayan types) of true quills. Of the Indo-AIalayan forms, from descriptions and from those I have examined I cannot credit that there is more than one species; trivial skull characters have for the most part been used to separate the various forms, and probably all are best regarded as races. The African types divide sharply into two on skull characters, though the form turneri seems to me to be intermediate between the two types to a certain degree; it must also be noted as a form in which apparently the quills are weak. In africaniis, the trontals are arched, the skull is broader, the infraorbital foramen, due no doubt to the inflation of the frontals, seems smaller, and the jugal appears consistently to extend up to the lachrvmal. In centralis, the frontals are not or less arched, the skull is narrower, and more like the Asiatic tvpes, and the jugal may or may not extend so far anteriorly. A. Imrrowsi, based on a skull without a skin, is best regarded as a race oi centralis. List of N.^med Forms macrourus Group 1. ATHERURUS MACROURUS MACROURUS, Linnaeus 1758. Syst. Nat. loth Ed. no. 4, p. 57. No exact locality'. Synonym: {?)Jasciciilata, Shaw, 1801, Gen. Zool. ii, i, p. 11, pi. 124. (This name used by Lyon, 1907, for the Malaccan Trichys.) orieiitalis, Brisson, 1756, Regn. Anini. p. 131. East Indies. 2. ATHERURUS MACROURUS PEMANGILIS, Robinson 1912. .Ann. Mag. Nat. Hist. 8, X, p. 590. Johore Archipelago (Malaya) ; Pulau Pemanggil, between Pulau Aor and Pulao Tioman, S. China Sea. 3. ATHERURUS I\L\CROURUS STEVENSI, Thomas 1925. Proc. Zool. Soc. London, p. 505. Ngai-Tio, 7'ongking. 4. ATHERURUS MACROURUS ASSAMENSIS. Thomas 1921. Joum. Bombay Nat. Hist. Soc. XXVH, p. 598. Cherrapunji, Assam. 5. .■\THERURUS M.^CROURUS H,4INANUS, Allen 1906. Bull. Amer. Mus. Nat. Hist. XXH, p. 470. Hainan. ft. ATHERURUS MACROURUS TERUTAUS, Lyon 1907. Proc. U.S. Nat. Mus. XXXII, p. 587. Pulou Ten-itau (Malay Peninsula). Fig. 58. Atherurus turneri, St. Leger. B.M. No. 34.6.2.77, (J; X I. L Fig. 59. .Atheriris tlrneri, St. Lcger. B.M. No. 34.6.2.77, (J; X I. 2o8 ATHERURUS 7. ATHERURUS MACROURUS TIONIS, Thomas iqoiS. Joum. Fed. Malay States Mus. Vol. II, no. 3, p. 105. Juara Bay, Tioman Island (Malay Peninsula). 8. ATHKRURUS MACROURUS ZYGOMATICUS, Miller 1903. Smiths. Misc. Coll. no. 1420, 45, p. 42. Pulau Aor, Johore, Malay Peninsula. africanus Group 9. ATHERURUS CENTRALIS CENTRALIS, Thomas iSqs. Ann. Mag. Nat. Hist. 6, XV, p. 89. Monbuttu, Congo, Central Africa. 10. ATHERURUS CENTRALIS BURROWSI, Thomas 1902. Ann. Mag. Nat. Hist. 7, IX, p. 271. Lower Aruwimi River, Congo. 11. .-ATHERURUS TURNERI, St. Leger 1932. .Ann. \Iag. Nat. Hist. 10, X, p. 231. West Kenya Colony; Kakamega Forest, near Kaimosi, North Kavi- rondo. 12. .-XTHERURUS AFRICANUS, Gray 1842. .Ann. Mag, Nat. Hist. X, p. 261. Sierra Leone. incertae sedi's 13. ATH1:RL'RUS ARMATUS. Gervais. (Not seen.) 1S54. Nat. Hist. Mamni. i, p. 334. Senegambia. (Description evidently based on e.xternal characters only.) The genus Atlierurus provides a good example of discontinuous distribution, no form being known living between Assam and Kenya. It is interesting to note that the genus lives side by side with llystrix throughout much of its range; and that the primitive member of the Atheninis group {Trichys) is restricted to the Malav Islands area, exactly as is the primitive member of the Hystrix group (Tlieciirus). The Hystrix Group The Hystrix group differs from the Atheninis group primarily in the reduc- tion of the tail, which is short, and as indicated above bears a cluster of hollow "rattling-quills," developed to a greater or lesser degree. These quills have a flower-like effect, being secured to the tail by a stalk, above which they open out into the hollow terminal part. When the animal is nervous, the tail is apparently shaken, and the quills, being lightly attached, are rattled together to produce the warning signal. I have written at some length above on this fact, and the habits of certain species of the genus concerning it. There are always on the back some thick circular quills, the extremities of which are very sharp. The cheekteeth in this section are extremely hypsodont, usually semi-rooted. THECURUS 209 Key to the Genera of the Ilystrix Group Nasals narrower, shorter, confined to rostrum (essentially Atherurus-likc in appearance), extending back to about anterior margin of infra- orbital foramen, shorter than the frontals. Thecurus Nasals broader, longer, not confined to rostrum, extending back about to the level of the lachrymal in primitive species, or in progressive species tending to reach level of posterior zygomatic root; frontals shorter than the nasals. Hystrix The nasals of all II\strix seen are considerably broadened, even in primitive forms \i]i.e javanictim. So far as I can trace, and from our skulls, the percentage of the nasal length against the occipitonasal length does not exceed 33 per cent in Thecurus, and averages 31 in our series, 32 in a series of sumatrae the measure- ments (taken against the "total length" of skull) published by Lyon. In Hystrix, its lowest adult appears to be a specimen from Flores with the per- centage 394; two specimens from Java, juvenile or subadult have the percentage 39-6 and 39'8; others measured are over 40 per cent. Some measurements included below. Genus 3. THECURUS, Lyon 1907. Thecurus, Lyon, Proc. U.S. Nat. Mus. XXXII, p. 382. Type Species. — Thecurus sumatrae, Lyon. Range.— Sumatra, Borneo, and Paragua, Philippine Islands. Number of Forms. — Three. Characters. — Like smaller species of Hystrix (next to be described), but skull transitionary towards the Atherurus type, the nasals narrow, extending back only to about the level of anterior margin of infraorbital foramen. Rostrum narrower; skull appearing flatter than in any Hystrix seen. Frontals longer than nasals; interorbital constriction may be suggested. Palate not narrowed in front of toothrows; palatal foramina as usual very small; infraorbital foramen, as in Hystrix, with no canal for nerve transmission. Bullae relatively small. Cheekteeth strongly hypsodont; pattern as already described (ultimately the pattern of the teeth appears to become obliterated). Remarks. — On account of the Athenirus-hke skull (regarding the area of the nasals) this group may, I think, stand as a genus, though it is not very well separated from smaller species of Hystrix like javanicum. There is, however, an undoubted and clear difference in the skulls of these two types. Lyon in his original genus description compared Thecurus to " Acanthion" and remarked that the caudal rattling-quills are much smaller; but, as indicated above, they are essentially in formation as those of H. javanicum (which Lyon evidently did not see), and H. hodgsoni from Nepal, which Lvon did not use in his notes for comparing the two tvpes. He also mentioned a number of skeletal characters in which the two groups 14 — Livin;; Rodents — I Fig. 6o, Thecurus crassispinis, Gunther. B.M, No. 92.9.6.171 X I. Fig. 61. Thecurus crassispinis, Gunther. B.M. No. 92.9.6.17; X I. THECURUS 211 differed; such as, for instance (compared with Acanthion), "instead of having a large laterally compressed neural spine on the axis, that vertebra bears a relatively short triprismatic spine, not compressed laterally any more than it is anteroposteriorly," etc. But before accepting such characters as these for generic purposes it is surely necessary to examine skeletons of all known Hystrix-group Porcupines. As indicated above in my notes on external characters of main species in the family, two very well-marked groups may be recognized: the pumilis group, in which the covering is at its lowest development among true Porcupmes, and the crassispinis group, containing a larger animal in which the quills of the back reach a surprising degree of thickness for a relatively generalized Porcupine of this type. Forms seen : crassispinis, pumilis, and a recently acquired specimen from Sumatra which I take to represent sumatrae. List of Named Forms pumilis Group 1. THECURUS PUMILLS, Gunthcr 1879. Ann. Mag. Nat. Hist. 5, IV, p. 106. Paragua, Philippine Islands. 2. THECURUS SUMATR.^E, Lyon 1907. Proc. U.S. Nat. Mus. XXXII, p. 583. Aru Bay, east coast of Sumatra. crassispinis Group 3. THECURUS CR.ASSISPINIS, Gunthcr 1876. Proc. Zool. Soc. London, p. 736. Mainland of Borneo, opposite Labuan. Nasal, Front.\l and Occipitonasal Measurements (in mm.) of British Museum Series of Skulls (other than broken ones) OCCIPITONASAL SPECIES NUMBER NASAL LENGTH FRONTAL LENGTH LENGTH T. pumilis 79-5-317 25 <^- 34 87 94.2. 1. 1 5 27 32 91 T. crassispinis 76.9.20.15 (the type) 33 38 c. no 92.9.6.17 37 40 114 92.10.1.5 34 39 III 95-5-7-7 35 37 lo? 84.5.19.7 35 36 106 Measurements for comparison, of hrachytirus Group of Hystrix in British Museum (small species with nasal percentage of occipitonasal length less than 50 per cent) H. jaiamcum, juvenile 50.1 2.2. 16 43-5 29 I09 H. brachyurus 5.9.27.1 51 36 I2Z OCCII'ITONASAL NASAL LHNGIH FRONTAL LENGTH LENGTH 63 34 130 62-5 34 128 43 31 109 45 31 109 ^I'S left 31 5 114 48 r ght 44 31 5 104 42 32 loO 212 THECURUS— HYSTRIX Sl'ICIFS NUMIIEU! H. hniihvurus 3.2.6.74 H. wiillmO), from Borneo 93. 3. 4. 8 Frost Collkction, 1938. Not yet registered. Flores, adult female jtn'aiiiciini Flores, adult nva\c jmaniciim East Java, adult male javaiiicum East Java, old female javaniciiiii East Java, sub-adult y«f(7/H'c»/« All other Hystrix, so far as traced, have the percentage of nasals against occipitonasal length more than 50 per cent e.xcept sometimes H. Icucura (^•j,fide Lonnberg). Genus 4. HYSTRIX, Linnaeus 1758. Hystrix, Linnaeus, Syst. Nat. loth. Ed. i, p. 56. 1822. AcANTHiON, Cuvier, Mem. Mus. Hist. Nat. IX, p. 425. (Acaiithion javaiiicum, Cuvier.) Valid as a subgenus. Type Species. — Hystrix cristata, Linnaeus. Range. — The greater part of the African Continent (Morocco, Asben (Sahara), Upper Egypt, Senegal, Gambia; Uganda, Kenya, Somaliland, Tanganyika, Portuguese East Africa, South-west Africa, South Africa); Italy and Sicily; Palestine, Syria, Asia Minor, Mesopotamia, South Arabia (specimens in B.M.), Afghanistan, probably Persia; Transcaucasia (Talysh); South-western Siberia (Turkmenia, Semirechia, Kopet-Dag moun- tains, Karakum), (Vinogradov); Baluchistan; Peninsular India (Punjab, Rajpu- tana. Central Provinces, Palanpur, Cutch, Kathiawar, Deccan, Mysore, Coorg, Nilgiris, Malabar); Ceylon; Nepal, Sikkim, Bhutan, Assam; Burma, Tenas- serim. South China (Szechuan, Y'unnan, Fukien, Anhwei); Hainan; Malay Peninsula, Sumatra, Java, Borneo, Sumbawa, Flores. Number of Forms. — .Approximately thirty-five. Ch.\r.\cters. — Nasals, even in the most primitive forms, longer and broader than in Thectirits, extending about to lachrymal level in brachyurus group, progressively lengthened in most other species; broader to a degree and much longer in suhcristatus group; relatively short in leucnra group, and not wider behind than in front; relatively short in afrkaeustralis group but enormously broadened, much wider behind than in front; considerably broader behind, and also much lengthened in cristata group, ultimately approaching the level of the posterior zygomatic root. In these larger African species the skull becomes much arched. A prominent sagittal crest normally present in adult. Occiput thick, strongly ridged, prominent. Bullae relatively small; paroccipital processes not much lengthened. Palate broad, extending about to end of toothrows behind, and straight; not depressed in front of toothrows; HYSTRIX 213 palatal foramina short, far in front of toothrow. Infraorbital foramen of moderate size for a Hystricoid Rodent, relatively small in some in which the nasals reach their maximum inflation ; no canal for nerve transmission. Zygoma broad but simple; jugal not approaching lachrymal as a rule; zygomatic plate projected forwards, appearing as an anterior prolongation of zygoma; lachrymal moderately large. Incisors broad. Mandible with hinder part flattened, the angular portion powerfully ridged and distorted outwards. Coronoid process low'. Externally becoming very large in progressive species ("38 inches" and 810 mm. the largest (measured) skins seen, galeata type). Forefeet broad, with four well-developed digits bearing thick claws; hindfeet longer but essentially similar except that the hallux is quite well developed, the two central digits slightly longer than the outer ones. I have already written at length on the external characters as regards arrange- ment of head-crest, body quills and spines, and caudal rattling-quills, of the various species, on pp. 199, 200. The cheekteeth are essentially as already described under the heading "Family Hystricidae" (p. 198). They are strongly hypsodont, the pattern is long preserved though ultimately obliterated, and the premolar is shed late in life. Remarks. — The genus is frequently divided into two, Hystrix and Acan- thion. Great as are the differences between the highly specialized H. cristata (type of Hystrix), and the relatively primitive H. javanicitm (type of Acanthion), it becomes clear that so many intermediate forms exist that this classification cannot be retained. This is made very clear in a paper by Lonn- berg, 1923 (on the Chinese Porcupine H. subcristatus, Swinhoe with remarks on other members of the genus, Arkiv. for Zoologi, Band 15, No. 18, pp. i-io), and in other papers by this author. In 1912 Aliller restricted the genus Hystrix to the European and African species only, on account of the "inflation of facial regions of skull at maximum for the family; nasal bones extending to glenoid level." But this appears to include in Hystrix the Chinese subcristatus, currently referred to Acanthion, and to exclude the African crested species africaeaustralis, which is naturally a Hystrix. It is quite clear that on nasal structure alone this genus will not divide into two. In my opinion, in an animal of this description, the external characters (development of quills, etc.) must be regarded as being just as important as any cranial character. There is a definite break in the species between cristata, africaeaustralis, leucura groups (Crested Porcupines) on the one hand, and subcristatus {}) (skin not seen, but description fits in with Acanthion as here suggested), klossi, brachyurus, javanicum types on the other. For the present I suggest that Acanthion may be used subgenerically for the latter group (with crest poorly developed, vestigial or absent; caudal-quills moderately to poorly developed; body quills thick but without the profuse mantle of manv-ringed longer quills present and covering them). If on the other hand subcristatus proves to be an intermediate between the two groups, the name Acanthion will have to be placed in svnonymv. 214 HYSTRIX Subgenus ACANTHION (With characters as just indicated; quills with one ring only.) There are two groups here recognized, a "short-nasal" group containing bidchxurus, of the Malay Peninsula and Islands (with loiigicauda and iniiUeri as races or synonvms), and jai-aiiiciim, and a "long-nasal" group. H. braclivurus group (nasals in percentage of occipitonasal length averaging 45-3 in our specimens; 49 in five measured by Lyon (nasals against "total length" of skull). H. javankum has an average percentage of 40-9 in the few adult skulls repre- sented in London ; sunibazcae, Schwarz, from Sumbawa has a slightly lower measurement, 36-2, in the figures published. This might or might not be a race oi jaraniciim; specimens received from Flores (which is beyond Sumbawa east- wards from Java), appear quite indistinguishable from typical javanicum. I divide this group into two sections, the typical, and the javanicum section (several skins of which have been seen), these sections differing from each other in the external characters indicated on pp. 199, 200. Although between these two sections there is quite a clear difference in the material examined, perhaps if enough material was collected, the two types would intergrade. //. subcristatiis group (with long nasals) contains siibcristatiis, klossi (with mills!, based on skulls the external characters of which are unknown, pro- visionally treated as a race), and hodgsoni. Nasals in percentage of occipitonasal length in subcristattis S^'(^57'7 (Lonnberg), (and mesopterygoid space unusually wide in our skulls); the percentage in hodgsoni averages 55-6 in four measured; the same in klossi (four measured), is 53-6. This group is divided into two sections, in precisely the same way as the braclivurus group; in hodgsoni the external characters are more primitive than in klossi \ but probably intergradation would take place in these external characters, if enough specimens came to hand. These characters have been noted on p. 200. The measurements of the skulls just quoted are: OCCIPITONASAL SPECIES NUMBER NASAL LENGTH FRONTAL LENGTH LENG1 H. klossi 14. 12.8.224 80 31 139 ,- -, (type) 14. 12. 8. 223 75 34 145 1) 7» I 5. 1 1.4.220 71 30 i3'3 ,, ,, (nnllsi\ type 21.7. 16.4 68 Zi 128 H. hodgsoni S3.8.16.I I 64 24 IIS ») )» 45.1.8.8 6S 20 116 )j ») 21.10.4.3s 66 23 115 i< ti 79.11. 21. 637 61 24 114 Two of the main difficulties of dealing with animals of this description are (i) the rarity, and (2) the frequently bad condition, in which they come to hand. But even if these notes are based on insufficient material, they do at least give a preliminary survey of all the main species of the whole genus, and not sections of it (for instance, Lonnberg's paper compares only suhcristatus with the larger crested types; Lyon's paper compares only the Indo-Malayan ones. I have seen no paper which compares klossi, hodgsoni, etc., with either of these, or either of these groups with each other). HYSTRIX 215 Subgenus HYSTRIX (With more highly specialized development of external characters; crest long, fully developed; caudal rattling-quills at maximum development; a profuse Fig. 62. Hystrix cristata, Linnaeus. (From Miller's Catalogue of the Mammah of Western Europe.) xj. growth of very long many-ringed quills covering the short thick ones of Acanthion. For detail notes see p. 213.) Lonnberg states that these Porcupines (with subcristattis) should divide into three groups, as follows: 2i6 HYSTRIX "Nasal cavity widened chiefly by means of prolongation backwards of nasals. Proc. nasales of premaxillaries truncate behind, only little widened, suhciistatiis." "Nasal cavity much enlarged by extremely broad proc. nasales of pre- maxillaries . Iciiaiiii, ga/eata." (He suggests this group might divide into two, one for kuciiici (Indian, Asiatic types), the other for the African galeata.) " Nasal cavity enlarged by expansion of nasals, proc. nasales of premaxillaries wedge-shaped behind, not or only moderately enlarged, africae- australis, cristata." (He suggests that this group might also divide into two, one for cristata and seiiegalica type, one for ajricaeaiistralis.) But there is a very profound difference betw'een leiicura and galeata. Skulls referred to galeata in London seem to vary individually in the shape of the nasals. I believe that all species of African Porcupines excepting africacaiistralis would prove to be referable to one species cristata if enough of them came to hand, and that the shape of the nasals would be found to vary individually so that galeata would become merged into cristata. H. leuctira with hirsutirostris seems to me to be a perfectly natural group, sharply differentiated on nasal structure from all African Porcupines. The nasals are not or scarcely broader posteriorly than anteriorly; the whole skull lacks that broadening characteristic of African Crested Porcupines. Further, the nasals in percentage of occipito- nasal length are short; (47-49 kiiciira, 48- 2-49-6 liirsiitirostris) (fide Lonnberg); a leiicura measured for comparison with these figures has the percentage Sf". This is markedly shorter even than in africacaiistralis. Russian authors give hirsutirostris full specific rank, but there seems no reason to believe that it is distinct from the Indian leucura. I am therefore treating all named forms of this group as subspecies of the earlier name leucura. The africaeaustralis group, from South and Southern Africa, appears on the material examined to be clearly separable from the cristata group as here understood. Compared with leucura, the nasals are much broadened, always as far as seen considerably broader posteriorly i^^Sy than anteriorly; compared with the cristata group, they are short (percentage of occipitonasal-nasal length 54- ^^■2, fide Lonnberg; this percentage slightly exceeded (55-9), in British Museum material). H. stegmanni, not seen, appears from Lonnberg's percentage figures and remarks to belong in this group. The cristata group contains all other African Porcu- pines, as here understood. The nasal lengthis maximum for the genus, and is combined with great broadening, but the shape is very variable both individually and between some of the species (cristata, per- centage 58-9-65, senegalica, 69; galeata, 6i-i-66-8; galeata ambigua. 60-7, Fig. 63. HvsTRix CRISTATA, Linnaeus. Cheekteeth; i^. HYSTRIX 217 fide Lonnberg). The nasals extend much farther back than in the africae- australis group. //. aeritla, from Ashen, is a small form allied to senegalica. In my opinion probably all North African Porcupines are no more than races of cristata, including uccidanea, not seen, as figured by Cabrera. Forms seen: aeriilti, africacaustralis, " hengalensis," brachyurus, cristata, cuneiceps, "curieri," galeata, hirsutirostris, hodgsoni, javanicum, klossi, leucura, millsi, miilleri, senegalica, schmidti, somaliensis, subcristatus. List of N.ameu Forms Subgenus Acanthion, Cuvier brachyurus Group (javanicum section) 1. HYSTRIX JAVANICUM, Cuvier 1822. Mem. Mus. Xat. Hist. IX, p. 431. Java. (Occurs also in Flores.) Synonym: torquata, van der Hoev, 1836, Tijdschr. iii, p. 110. Java. breiispinosa, Wagner, 1844, Schreber, Saug. Suppl. \\ , p. 20. Java. ecaudala, van der Hoev, 1836, Tijdschr. iii, p. no. Java. 2. HYSTRIX SUMBAWAE, Schwarz 191 1. Ann. Mag. Xat. Hist. 8, VH, p. 639. Dompu, Sumbawa, East Indian Archipelago. (typical section) 3. HYSTRIX BRACHYURUS BR.ACHYURUS, Linnaeus 1758. Syst. Nat. loth Ed. p. 57, no. 5. Malacca. Synonym : ^o/fi. Gray, 1866, Proc. Zool. Soc. London, pi. 31, p. 310. flemiiigi. Gray, 1847, Proc. Zool. Soc. London, p. 103. 4. HYSTRIX BR.ACHYURUS LONGICAUDA, Marsden 181 1. Hist. Nat. Sumatra, 3rd Ed. p. u8. Sumatra. 5. HYSTRIX BR.ACHYURUS MCLLERI, Marshall 1871. Proc. Zool. Soc. London, p. 235. (See Lvon, Proc. U.S. Nat. Mus. XXXII, 1907, p. 580.) Padang, Sumatra. (An animal of this type occurs in Borneo.) suhcristattts Group {hodgsoni section) 6. HYSTRIX HODG.SONI, Gray 1847. Proc. Zool. Soc. London, p. loi. Nepal, Himalayas. Synonym: bengalensis. Blyth, 1851, Joum. .As. Soc. Bengal, XX, p. 170. alophus, Hodgson, 1847, Joum. .As. Soc. Bengal, XVI, p. 771. Himalayas. 2i8 HYSTRIX (typical section) 7. HYSTRIX KLUSSI KLOSSI, Thomas 1916. Ann. Mag. Nat. Hist. 8, XVII, p. 139. Tenasserim Town. S. HYSTRIX KLOSSI MILLSI, Thomas 1922. Journ. Bombay Nat. Hist. Soc. XXVIII, no. 2, p. 431. Sangrachu, Assam. 9. HYSTRIX SUBCRISTATUS SUBCRISTATUS, Swinhoe 1870. Proc. Zool. Soc. London, p. 638. South China; Fokien Province. 10. HYSTRIX SUBCRISTATUS PAPAE, Allen IQ27. Amer. Mus. Nov. no. 290, p. 3. Hainan. Not allocated to Group 11. HYSTRIX YUNNANENSIS, Anderson 1878. Anat. and. Zool. Res. Yunnan, p. 332. West Yunnan. (According to Thomas based on a short-nosed species allied to javanicum.) Subgenus Hystrix, Linnaeus leiiciira Group 12. HYSTRIX LEUCURA LEUCURA, Sykes 1 83 1. Proc. Zool. Soc. London, p. 103. India; Sayul of Mahrattas. Synonym: zeylonensis, Blyth, 1851, Journ. As. Soc. Bengal, XX, p. 171. Ceylon. malaharica, Sclater. 1865, Proc. Zool. Soc. London, p. 353. Cochin, India. indica, Gray & Hardwicke, 1833-34, 111. Indian Zool. ii, pi. 14. 13. HYSTRIX LEUCURA CUNEICEPS, Wroughton 1912. Journ. Bombay Nat. Hist. Soc. XXI, p. 771. Nokania, Cutch, India. 14. HYSTRIX LEUCURA HIRSUTIROSTRIS, Brandt 1835. Mamm. Exot. Nov. p. 39. Afghanistan. 15. HYSTRIX LEUCURA BLANFORDI, Muller 191 1. Sitz. Ber. Ges. Nat. Freunde Berlin, p. 121. Jalk, Baluchistan. 16. HYSTRIX LEUCURA SATUNINI, Mullei 1911. Sitz. Ber. Ges. Nat. Freunde Berlin, p. 117. Transcaspia, Geok Tepe, east of Caspian Sea, 75^' E., 38" N. 17. HY.STRIX LEUCURA MERSINAE, Muller 191 1. Sitz. Ber. Ges. Nat. Freunde Berlin, p. 122. Mersina, south-east of Taurus, Asia Minor. HYSTRIX 219 i8. HYSTRIX LEUCURA AHARONII, Muller 191 1. Sitz. Ber. Ges. Xat. Freunde Berlin, p. 123. Palestine; Emmaus, west of Jerusalem. 19. HYSTRIX LEUCL'RA SCHMITZI, Muller 191 1. Sitz. Ber. Ges. Nat. Freunde Berlin, p. 126. Palestine; Ain Dcheier, N.-W. of Dead Sea, Jordan valley. 20. HYSTRIX LEUCURA NARYXENSIS, Muller 1919. Sitz. Ber. Ges. Nat. Freunde Berlin, p. 67. Naryn, Turkestan. 21. HYSTRIX LEUCURA MESOPOTAMICA, Muller 1920. Zool. Anzeiger, 51, p. 198. Jebel .-Xbdul Azir, N.-E. Syria; 40° 20' E., 36° 20' N. africaeaustralis Group 22. HYSTRIX AFRIC.\E.AUSTR.\LIS AFRICAEAUSTRALIS, Peters 1852. Reise nach Mossambique, Saugeth. p. 170. South-East Africa; Querimba, Tette; 11° to 17° south. Synonym: copensis. Grill, 1858, Zool. Anteckningar af. J. F. Victorin, p. 19. 23. HYSTRIX AFRICAEAUSTR.A.LIS PRITTWITZI, Muller 1910. Sitz. Ber. Ges. Nat. Freunde Berlin, p. 311. Tabora, Tanganyika Territon.-. 24. HYSTRIX AFRICAEAUSTRALIS ZULUENSIS, Roberts 1936. Ann. Trans. Mus. XVIII, p. 240. Zululand, White Umfolosi River. 25. HYSTRIX STEG.M.AXNI. Muller 1910. .\rch. f. Naturgesch. Jahrg. 76, Band i. p. 186. Kissenji, north-east of Lake Kivu, Tanganyika. cristata Group 26. HYSTRIX CRISTATA CRISTATA, Linnaeus 1758. Syst. Nat. i, loth Ed. p. 56. Near Rome, Italy. (See Miller, Cat. Mamm. W. Europe, 1912, p. 543.) Synonym: cuiieri. Gray, 1847, Proc. Zool. Soc. London, p. !02. Locality not known. (?) datibentoni, Cuvier, 1822, Mem. Mus. Nat. Hist. IX, p. 431. Locality unknown; perhaps best regarded as unidentifiable. alba, de Selys-Longchamps, 1839, Etudes de Micro- mamm. p. 152, nom, nud. europaea, Kerr, 1792, Anim. Kingd. p. 213. (Some specimens in B.M. labelled " moroccana" ; the reference to this name has not been traced.) 27. HYSTRIX CRIST.\TA OCCIDANEA, Cabrera 1924. Bol. Soc. Esp. Hist. Nat. XXIV, p. 220. Mogador, West Morocco. 28. HYSTRIX CRIST.ATA SEXEG.A^LICA, Cuvier 1822. Mem. Mus. Hist. Nat. IX, p. 430. Senegal, West .\frica. 220 HYSTRIX— CUNICULIDAE 29. HYSTRIX CRISTATA AKRULA, Thomas 1925. Ann. Mae. Nat. Hist, f), XVI, p. 196. Aouderas, Asben, Sahara. 30. HYSTRIX GALEATA GALEATA, Thomas 1S93. Ann. Mag. Nat. Hist. 6, XI, p. 230. Lamu, Kenya. 31. HYSTRIX GALEATA SOMALENSIS, Lbnnberg 1912. Kungl. Sv. Vet. Akad. Handl. Band. 48, no. 5, p. 109. Njoro, Guaso Nyiro, North Kenya. 32. HYSTRIX GALEATA AMBIGUA, Lbnnberg 1908. Sjost. Kilimanj. Meru. Exp. p. 29. Kibonoto, Kihmanjaro, East Africa. 33. HYSTRIX GALEATA LADEMANNI, Muller 1910. Sitz. Ber. Ges. Nat. Freunde Berlin, p. 314. Kondoa-Irangi, Tanganyika. 34. HYSTRIX GALI:ATA CONRADSI, Muller IQIO. Sitz. Ber. Ges. Nat. Freunde Berlin, p. 314. Ukerewe Island, Lake Victoria. 35. HYSTRIX GALEATA LONNBERGI, Muller 1910. Sitz. Ber. Ges. Nat Freunde Berlin, p. 315. Kilimanjaro, East Africa. The family Hvstricidae is known fossil from the Upper Miocene, from the Old World only. ' HYSTRICIDAE : SPECIAL WORKS OF REFERENCE Waterhoi'se, 184S, Natural Historj- Mammalia; Rodentia (Vol. II). PococK, 1922, Proc. Zool. Soc. London, 1922, p. 365. External Characters of some Hystricomorph Rodents. TuLLBERG, 1899, Nova Acta Reg. Soc. Sci. Upsaliensis, XVIII, 3, i. Lyon, Porcupines of the Malay Peninsula; Proc. U.S. Nat. Mus. XXXII, 1907, pp. 575-594- LoNNBERi;. On the Chinese Porcupine Hystrix siihcristatiis. Swinhoe, .Arkiv. for Zoologi, Bd. 15, no. 18, p. I, 1923. Miller, Catalogue of Mammals of Western Europe, 1912, p. 542. Hvstricidae. {Hystrix cristata). Family CUNICULIDAE 1S96. Thomas: Hvstricomorpha: Dasyproctidae, part. 1899. TuUberg: Hvstricomorph.a : Caviidae, part. 1918. Miller S: Gidley: Hv.sTRicoiD.y- : Family Cuniculidae. 1924. Winge: Family Hvstricidae; Dasyproctini, part, Dasyproctae, part. 1928. Weber: Hystricoidea: Caviidae, part; subfamily Dasyproctinae, part. Geographical Distribution. — Tropical America, from Mexico to South Brazil. Number of Genera. — One. Characters. — Skull highly abnormal, the greater part of the maxillary and jugal expanded to form large bony cheek-plates, the surface CUNICULIDAE: CUNICULUS 221 of which tends to become rugose. Cheekteeth hypsodont, semi-rooted, charac- terized by deep re-entrant folds which isolate as long islands on crown surface. External form heavy, terrestrial, the limbs not lengthened; feet with digits of sub-ungulate type, the claws extremely thick; hindfoot perissodactyle, with three main digits, but both D.5 and hallux present though strongly reduced; forefoot artiodactyle, with four main digits. Clavicles not suppressed, but (said to be) incomplete. According to Tull- berg, the carpus lacks a free centrale, alone of all Rodents examined by him excepting the Hystricidae. Rhm.^rks. — The diflFerences between this genus and the Dasyproctidae with which it has often been associated have been discussed at length by Pocock (Proc. Zool. Soc. London, 1922, p. 424), who, following Miller & Gidley, refers the animal to a distinct family. I am entirely in agreement with this classification. The unique skull structure in the family indicates evolutionary development along a very different line not only from that of Dasyproctidae but from all other Rodents; the feet do not agree with those of the Dasyproctidae, nor are the limbs lengthened; the cheekteeth do not agree exactly with those of the Dasyproctidae. (The relationships of Dasyproctidae as compared with Caviidae, with the latter of which Cuniculus has also been associated, have been fully discussed when dealing with the family Dasy- proctidae.) As indicated already, the cranial characters of this genus difTer entirely from those of any other Rodent; in this particular it must be looked upon as the most aberrantly specialized member of the whole Order. Genus i. CUNICULUS, Brisson 1762. Cuniculus, Brisson, Regn. Anim. ed. 2, p. 13. 1799. Agouti, Lacepede, Ordres et Genres Mamm. 9. Agouti paca (^Mus paca, Linnaeus). 1807. CoELOGENUs, Cuvier, Ann. Mus. Hist. Nat. Paris, X, p. 203. (Mus paca, Linnaeus.) 1924. Stictomys, Thomas, Ann. Mag. Nat. Hist. 9, XHI, p. 238. {Coelogenys tacza- ncnuskii, Stolzmann.) Type Species. — Mus paca, Linnaeus. R.WGE. — Forms are named from Mexico, Panama, Cayenne, Ecuador, Colombia, Brazil, Venezuela. Specimens in British Museum from Paraguay, Peru; according to Thomas ranging to South Brazil. Number of Forms. — Ten approximately. Characters. — Skull with zygomatic region abnormally modified by out- growth from maxillary and jugal, of bone forming a cheek- plate which extends downwards and conceals a large part of the mandible, the maxillary' part ot this plate being deeply hollowed internally. The infraorbital foramen is smaller than in other Hystricoidae, in adults becoming strongly reduced, being dwarfed by the cheek-plate. Infraorbital foramen with a separate Fig. 64. CrsicuLUS paca, Linnaeus. B.M. No. 13.10.24.61, a; ■ slightly more than 1. CUNICULUS 233 canal for nerve transmission. Nasals broad, relatively short; frontals broad, very long; parietals depressed, and a sagittal ridge formed in adult; well marked postorbital processes occur at the suture of the frontals and parietals. Parocci- pital processes thick and rather long; bullae relatively small. Palate broad, not constricted anteriorly, the anterior part extending beyond the toothrow as a i Fic. 65. Cltmiculus paca, Linnaeus. Cheekteeth: B. M. No. 13. 10. 24. 61, (J; x 2J. narrow shelf bordered by two high longitudinal ridges which extend nearly to the incisors, and on either side of which lie the enormous cavities caused by the cheekplates. Palatal foramina obsolete. On account of the cheek-plate, the skull of the typical species appears nearly as broad as long. There is a marked tendency for the cheek-plate in old specimens to become rugose, and for the frontals to assume a similar character, this being apparently especially marked 224 CUNICULUS in males. The hinder part ot the mandible is more or less flattened, as in Ilystricidae; coronoid process relatively low; degree of distortion outwards of angidar process moderate. Cheekteeth rather complex; upper series apparently with two inner and three outer re-entrant folds, except M,3, which is the largest tooth of the series, and in which the number of folds appears to be reversed. Sometimes there is a tendency for M.i to become reduced in size and elements, with wear. Most of the folds isolate as long persistent islands almost immedi- ately; the unworn tooth shows, as usual in this Order, an extremely complex pattern. Some of the isolated folds become suppressed with wear. Lower teeth with one outer, three inner folds each; the premolar may have an e.xtra inner fold. Incisors thin, compressed. Externally typicallv the fur is harsh, the sides of the body with longitudinal rows of spots. Hindtoot with D.5 much reduced, but with moderate claw (though less strong than those of D.2, 3, 4); hallux rudimentary; three main digits long, bearing very sharp somewhat hoof-like claws. F"orefoot with four main digits; D.2 is longer than D.5, but shorter than the central pair; pollex represented by a knob. Tail obsolete. Form heavy, and size large, one of the largest members of the Order. (The largest skin examined is 6S5 mm. head and body, but this measurement probably may be exceeded.) It is probable that the habits of these animals are not cursorial (com- pared with Dasyproctidae); Pacas are said to take to the water when alarmed. In the tdcsdiiuicskii group (Mountain Pacas), the fur is thicker, less harsh. This group was referred to a distinct genus by Thomas, on account of "the narrow compressed claws and much more profusely granulated soles; cranially by the proportionately longer nasals, much smaller orbits, more anteriorly situated postorbitals — the zygomata narrower, generally much less rugose, though as usual there is much variation in this respect — finally the incisors are orthodont." Mammae i — 1 =4 (sierrae), (Thomas). But there are far too many essential characters shared by the two groups for there to be any question of even subgeneric separation, in my opinion. These characters indicate that the plains Pacas and mountain Pacas belong to distinct species. It may be stated that in a skin of sierrtif, the claws seem even thicker than a specimen oi paca with which it was compared; there is certainly no generic difterence so far as claws are concerned; (compare, tor instance, Chimliilla with Lagoitomiis; Ciiviti with Kerodun); when two groups have gone so far together in specialization (cranial characters), it not only seems unneces- sary but bad classification to give them generic names on small cranial differences such as the above. An interesting account of the formation of the cheek-pouches of the genus is given by Mr. R. I. Pocock (under the name of Cotlogoiys), Proc. Zool. Soc. London, 1922, p. 37(1. Forms seen : paca, fiiiantii, taczaiimcskii, sierrac. I can see no specific difference between the last two forms, the latter being described as a distinct species. CUNICULUS aas List of Named Forms (References and type localities collected by Mr. G. W. C. Holt.) paca Group 1. CUNICULU.S PACA PACA, Linnaeus 1766. Syst. Nat. i, p. 81. Cayenne. .Synonym: fulvus, Cuvier, 1807, .\nn. Mus. X, p. 207. suhniger, Cuvier, 1807, Ann. Mus. X, p. 206. Tobago. 2. CUNICULUS PACA NELSONI, Goldman 1913. Smiths. Misc. Coll. LX, no. 22, p. g. Catemaco, Southern Vera Cruz, Mexico. 3. CUNICULUS PACA VIRGATUS, Bangs igo2. Bull. Mus. Comp. Zoo!. Har\-ard Univ. XXXIX, p. 47. Divala, Chiriqui, Panama. 4. CUNICULUS PACA ALBA, Kerr 1792. Anim. Kingd. p. 217. .St. Francis River, Brazil. 5. CUNICULUS PACA MEXIANAE, Hagmann 1908. Arch. Rassenbiol. 5, p. 25. Mexiana Island, Amazonian estuary, Brazil. 6. CUNICULUS PACA GUANTA, Lonnberg 1921. .\rk. f. Zool. Band XIV, no. 4, p. 45. Pacto, below Gualea, Ecuador. 7. CUNICULUS PACA SUBLAEVIS, Gervais 1854. Ger%'ais Mamm. i, p. 326. Colombia. taczanowskii Group 8. CUNICULUS TACZANOWSKII T.-VCZANOWSKII, Stolzmann 1885. Proc. Zool. Soc. London, p. 161. Ecuador; forests on either slope of the .Andes, between 6,000 and 10,000 ft. 9. CUNICULUS TACZANOWSKII ANDINA, Lonnberg 1913. Ark. f. Zool. Band VIII, no. 16, p. 28. Mount Pichincha, Ecuador. 10. CUNICULUS T.A.CZ.ANOWSKII SIERRAE, Thomas 1905. .Ann. Mag. Nat. Hist. ser. 7, XV, p. 589. I'edregosa Mountains, Sierra de Merida, Venezuela. Tate quotes " thomasi," nom. nud. (.') ex Thomas, Bull. Amer. Mus. Nat. Hist., LXVHI, 2, p. 314, 1935. CUNICULIDAE: SPECIAL WORKS OF REFERENCE POCOCK, 1922, Proc. Zool. Soc. London, 1922, p. 365. External Characters of some Hystricomorph Rodents; p. 376, account of the mechanics of the cheek-pouches. 15 — Living Rodents — I 226 CHINCHILLIDAE Watebhoise, 1848, Natural History Mammalia, Rodentia. Tate, 1935, Taxonomy Neotropical Hvstricoid Rodents, Bull. Amcr. Mus. Nat. Hist. LXVHI, p. 295. TvLLBERG, Nova Acta Reg. Soc. Sci. Upsaliensis, XVHI, 3, i, 1S99. Family CHINCHILLIDAE 1896. Thomas: Hvstricomorpha: Family Chinchillidae. 1899, TullberK: Hystbitomorpha: Family Chinchillidae. 1918. Miller &: Gidley: Hvstbicoidae: Family Chinchillidae. 1924. Winge: F'amily Hystricidae, part; "Eriomyini" (= Chinchillini). 1928. Weber: Hystricoidea: Family Chinchillidae. Geographical Distribution. — Western and Southern South America, from Peru, Bolivia and North Argentina southwards. Number of Genera. — Three. Characters. — Cheekteeth evergrowing, the pattern one of transverse plates. Mandible with no sharply defined ridge for attachment of masseter lateralis; the angular portion less strongly distorted outwards than is usual in Hystricoidae; a weak ridge below condylar process presumably for attachment for masseter medialis may be present, foreshadowing the structure present in Cavioidae, but much shorter and less developed than in that group. Jugal usually in contact with lachrymal; zygoma simple, but normally thickened anteriorly. A tendency present towards great inflation of inastoids and bullae; paroccipital processes relatively long. Palate much constricted anteriorly; palatal foramina usually very long, narrow. Incisors relatively narrow. External form slender, the hindfeet lengthened (semi-saltatorial or cursorial Rabbit-like types); digits of hindfoot three or four, D.5 when present extremely reduced, perhaps functionless. The lachrymal is large; part of the lachrymal canal is open on side of rostrum in front of orbit. A skeleton has been examined in each of the three genera, and each presents the feature that the fibula, though not fused with the tibia, is excessively reduced, a structure rather different from that of other Hystricoid Rodents examined for this character. The Chinchillidae fall into two well-marked groups, one containing Lago- stomiis only, the other ChiiicliiUa and Lagidiiiiii. The difierences between the skulls and the digits of these groups are rather extreme, and they have been regarded as subfamilies (Pocock, 1922). But these difl^erences seem rather adaptive; and I have seen it stated that Chinchilla has bred with "the much larger but related Vizcacha" (Jcnnison, 1929). They are therefore here treated as groups only. Key to the Groups of Chinchillidae Paroccipital processes long, standing apart trom bullae, which are not specially inflated; occipital region of skull strong, prominent; skull CHINCHILLIDAE: CHINCHILLA 227 more prominently ridged for muscle attachment; digits of hind- foot three, the claws heavy, prominent, excessively sharp; palatal foramina shorter; cheekteeth, excepting M.3, upper series, bila- minate. Lagostomus Group. Lagostomi Lagostomus Paroccipital processes closely applied to bullae, not or less elongated; bullae considerably to extremely inflated; occipital region of skull weak; skull not prominently ridged for muscle attachment; digits of hindfoot four; the claws blunt and weak; palatal foramina long and narrow; cheekteeth trilaminate. Chinchilla Group. Chinchillae Chinchilla, Lagidium The Chinchilla Group Characters. — As indicated in the above key. Key to the Genera of the Chinxhilla Group Bullae abnormally inflated, the mastoids showing prominently in superior aspect of skull. Jugal usually not in contact with lachrvmal. Laminae of cheekteeth straight. Chinchilla Bullae less abnormally inflated, the mastoids scarcely showring in superior aspect of skull. Jugal in contact with lachrj'mal. Laminae of cheekteeth curved. L.agidium Genus i. CHINCHILLA, Bennett 1829. Chinchilla, Bennett, Gard. and Menagerie Zool. Soc. i, p. i. 1830. Eriomys Lichtenstein, Darstell. Saug. VI, pi. 28. (Eriomys chinchilla, Lichten- stein). Type Species.— Mw laniger, Molina. Range. — Chile. ? Bolivia. Number of Forms. — One is recognized. Characters. — Mastoids and bullae abnormally inflated, the mastoids showing prominently each side and at back of skull. Con- siderable interorbital constriction evident. No canal for nerve transmission in infraorbital foramen. Palatal foramina very long; palate narrow, considerably so anteriorly. Jugal usuallv not extending to lachrvmal, broad. Paroccipital processes moderate, closely applied to and dwarfed by the bullae. Mandible with narrow angular process, which is sharply drawn backwards; the ridge outside the condylar process weak. Cheekteeth like Lagidium (next to be described), but the laminae straighter; three lobes per tooth, the hinder one in M.3 pointing backwards, as a heel. The anterior lobe of the lower teeth short, reduced. Externally, with very soft fur; the tail long, though not as long as head and body, heavily haired throughout. Hindfeet long and narrow; stiff bristle hairs Fig. 66. Chinchilla laniger, Molina. B.M. No. 1. 8. 24. 1, S; ■■ '■i- Fig. 67. Chinchilla laniger, Molina. B.M. No. 1.8.24.1, . /«a/. IS- p- subrosea. 16. p. saturata. 17. /). pnnensis. 18. /). arequipae. 17 and 18 doubtfully distinct. 19. 6. AOATZ'. 20. b. sarae. 21. woljfsohni." Note. — L. zvolffsolmi differs clearly in colour pattern from all the remainder. List of Named Forms 1. L..\GIDIUM VISCACCIA VISCACCIA, Molina 1782. Storr. Nat. Chile, p. 307. Chile. Synonym: luiescetis, Philippi, 1S96, .Ann. Mus. Chile, 13, p. 8. Tacapuca, Northern Chile. cuvieri, Bennett, 1833, Proc. Zool. Soc. London, p. 58. Peru. aureus, Geoffroy & D'Orbigny, 1830, Ann. Sci. Nat. XXL p. 291. Corrientes, Buenos .Ayres. criniger. Gay, 1847, Fauna Chile, i, p. 49. Chile. crassidens, Phihppi, 1896, Ann. Mus. Nac. Chile, 13, p. 10. chilemis, Oken, 1816, Lehrbuch Naturgesch. ii, p. 836, Chile. crimgerum, Philippi, 1896, Ann. Mus. Nac. Chile, 13, p. 10. Chile. viscaccica, Brandis, 1786, Versuch einer Naturgesch. von Chile, p. 272. 2. LAGIDIUM VISCACCIA LUTEA, Thomas 1907. Ann. Mag. Nat. Hist. 7, XIX, p. 443. Esperanza, Mount Sajania, Bolivia. 3. LAGIDIUM VISCACCIA CUSCUS, Thomas 1907. Ann. Mag. Nat. Hist. 7, XIX, p. 443. Paratani, Bolivia. 4. L.\GIDIUM VISC.A.CCIA PERLUTEA, Thomas 1907. .Ann. Mag. Nat. Hist. 7, XIX, p. 443. Pampa .Aulliaga, Bolivia. 5. LAGIDIUM VISCACCI.A VULCA.NI, Thomas 1919. Ann. Mag. Nat. Hist. 9, IV, p. 133. Cerro Casabindo, Jujuy, .Argentina. 6. L.AGIDIUM VISC.ACCI.A TUCUMANA, Thomas 1907. -Ann. Mag. Nat. Hist. 7, XIX, p. 444. Cumbre de Mala-Mala, Sierra de Tucuman, .Argentina. 7. LAGIDIUM VISCACCIA LOCKWOODI, Thomas 1919. Ann. Mag. Nat. Hist. 9, III, p. 499. Otro Cerro, Rioja, .Argentina. 232 LAGIDIUM S, LAGIDIUM VISCACCIA KAMATINAK, Thomas 1920. Ann. Mag. Nat. Hist. 9, VI, p. 421. La In\'ernada, Rioja, Argentina. Q. LAGIDIUM VISCACCIA TONTALIS, Thomas 1921. Ann. Mag. Nat. Hist. 9, VHI, p. 219. Los Sombreros, Sierra Tontal, west of San Juan, Argentina. 10. LAGIDIUM VISCACCIA VIATORUM, Thomas 1 92 1. Ann. Mag. Nat. Hist. 9, VI H, p. 220. Punta dc Vacas, Mendoza, Argentina. 11. LAGIDIUM VISCACCIA MORLNI, Thomas 1897. Ann. Mag. Nat. Hist. 6, XIX, p. 467. Hills near Chubut, Argentina. 12. LAGIDIUM PERUANUM PERUANUM, Meyen 1833. Nova Acta Ac. Nat. Cur. XVI, p. 57S. Southern Peru. 13. LAGIDIUM PERUANUM PALLIPES, Bennett 1835. Proc. Zool. Soc. London, p. 67. (Believed to be) Chilian Andes. 14. LAGIDIUM PERUANUM INCA, Thomas 1907. Ann. Mag. Nat. Hist. 7, XIX, p. 442. Incapirca, Zezioro, Junin, Peru. 15. LAGIDIUM PERUANUM SUBROSEA, Thomas 1907. Ann. Mag. Nat. Hist. 7, XIX, p. 442. Galera, west of Oroya, Dept. Lima, Peru, lb. LAGIDIUM PERUANT'M SATURATA, Thomas 1907. Ann. Mag. Nat. Hist. 7, XIX, p. 442. Limbane, Inambari, Dept. of Puno, Peru. 17. LAGIDIUM PERUANUM PUNENSIS, Thomas 1907. Ann. Mag. Nat. Hist. 7, XIX, p. 443. Puno, Lake Titicaca, Peru. 18. LAGIDIUM PERUANUM AREQUIPAE, Thomas 1907. Ann. Mag. Nat. Hist. 7, XIX, p. 442. Sumbay, near Arequipa, Peru. 19. LAGIDIUM BOXI BOXI, Thomas 1921. Ann. Mag. Nat. Hist. 9, VII, p. 179. Pilcaneu, Rio Negro, Argentina. 20. LAGIDIUM BOXI SARAE, Thomas & St. Leger 1926. .\nn. Mag. Nat. Hist. 9, XVIII, p. 639. Pino Hachado, Neuquen, Argentina. 21. LAGIDIUM WOLFFSOHNI, Thomas 1907. .Ann. Mag. Nat. Hist. 7, XIX, p. 440. Sierra de los Bagualcs y de las Vizcachas, 50 50' S., 72 20' \V., on boundary between Chile and Argentina. The Lagostomus Group Differing chiefly from the Chinchilla group in the bullae, whicli are not greatly inflated, the paroccipital processes, whicli are lengthened and stand apart from the bullae, the extremely sharp claws of the feet, the complete suppression of D.^ on the hindtoot, the skull more heavily ridged for muscle attachment. LAGOSTOMUS 233 Fig. 68. Lagostomus maximus m-wimus, Desmarest. B.M. No. 17.5.2.18; X t- Genus 3. LAGOSTOMUS, Brooks 1828. L.\GOSTOMUS, Brooks, Trans. Linn. See. XVI, p. 96. 1824. VizcACi.A, Schinz. Naturg. und Abbild. Sijugeth. p. 243. (This name is not to be used as it is a homonym of Viscaccia, Okcn, (Tate).) Type Species. — Lagostomus tricltodactylus, Brooks=£>ipui maximus, Des- marest. Range. — Argentina. One form, from Peru, is probably extinct. 234 LAGOSTOMUS Number of Forms. — Four. Characters. — Skull flat, with broad frontals, which bear quite well-marked postorbital processes. Nasals relatively long and narrow. Fig. 69. Lagostomus maximus m.\.ximus, Desmarest. B.M. No. 17. 5. 2. 18; ■, J. A well-marked sagittal crest present. Paroccipital processes lengthened (probably about as much as in T/irvoiiomvs); bullae not much inflated compared with the Chinchilla group, though appearing to a certain degree behind, each side of occipital region. A prominent canal present in infraorbital foramen for nerve Fig. 70. Lagostomls maximus maximus, Desmarest. B.M. No. 17.5.2. 18; X t- Fig. 71. Lacostoml's m.wimis maximus, Desmarest. Cheekteeth: B.M. No. 17.5.2.18; X 2}. 236 LAGOSTOMUS transmission. Palate strongly constricted anteriorly. Jugal slanting sharply upwards anteriorly, and in contact with lachrymal. As in other members of the family, the upper zygomatic root is placed far backwards, over the middle of hinder part of the toothrow. Mandible with rather strong ridge beside condyle for attachment of masseter medialis; this, however, much shorter than in Caviidae. Coronoid low; condylar process relatively low; angular process distorted outwards rather weakly. Cheekteeth set at an angle, the upper series with their outer edge pointing forw'ards; all upper and lower teeth with two laminae only except M.3 upper series, which is the largest tooth in the series and has three laminae. Incisors medium in width, their surfaces covered with faint longitudinal grooves. According to Pocock the penis differs considerably from that of other Hystricoid genera examined bv him, including ChinchiUa; see also note on breadth of manubrium on p. 171, which indicates yet another wide distinction from the Chinchillae. Externally relatively large (up to 470 mm. head and body in London collection); fur less soft than in Chinchilla group; tail not long, fully haired; forefoot with four digits armed with stout claws; hindfoot rather long, with three digits only, the claws in their development comparable to those of Cuiii- culits, excessively thick and sharp. Stiff bristle-hairs present on D.3, which is the longest digit; D.2 is shorter than D.4. Forms seen : crassus, inunoUis, inaximus. List of Named Forms 1. LAGOSTOMUS M.AXIMUS MAXIMUS, Desmarest 1S17. Nouv. Diet. d'Hist. Nat. xiii, p. 117. Argentina {?). (Localirj- unknown.) Synonym: tricliodactyhis, Brooks, Trans. Linn. Soc. XVI, p. gfi, 1828. .■\rgentina. diana, Griffith in Cuvier, 1827, Anim. Kingd. III. p. 170. viscaccia, GeofFroy & D'Orbigny, 1830, .\nn. Sei. Nat. xxi, p. 291. aiiiiger. Lesson, 1842, Nouv. Tabl. Regne. Anim. p. 105. poniparum, Schinz, 1825 (1S24) Naturg. und Abbild. Saugeth. p. 244. americana, Sehinz, 1825, Cuviers Thierreich, IV, p. 429. 2. LAGOSTOMl'S AL-VXIMUS IMMOLLIS, Thomas 1910. Ann. Mag. Nat. Hist. ser. 8, V, p. 245. Tapia, Tucuman, Argentina. 3. LAGOSTOMUS M.'iXIMUS PETILIDENS, Holl.ster 1914. Proc. Biol. Soc. Washington XXVII, p. 58. 8 miles south of Camien de Patagones, Southern Argentina. 4. LAGOSTOMUS CRASSUS, Thomas. (Extinct?) 1910. Ann. Mag. Nat. Hist. ser. 8, V, p. 246. Santa .^na, C'uzcn, Peru. (Known from cranial characters only.) CAVIOIDAE: CAVIIDAE 237 CIIINCHILLIDAE: SPECIAL WORKS OF REFERENCE Waterhouse, 1848, Natural Historj' Mammalia, Rodentia. Tate, 1935, Taxonomy of Neotropical Hystricoid Rodents, Bull. Amer. Mus. Nat. Hist. LXVIII, p. 295. PococK, Proc. Zool. Soc. London, 1922, p. 365. External Characters of some Hystri- comorph Rodents. TuLLBERC, Nova -Acta Reg. Soc. Sci. Upsaliensis, XVIII, 3, no. i, 1899. The family is known fossil from the IMiocene, from the Neotropical region only. Miller & Gidley quoted several extinct genera. Superfamily CAVIOIDAE 1896. Thomas: Hystrromorpha, part. 1899. TuUberg: Hystricogn.^thi, Hystricomorpha, part. 1918 Miller & Gidley: Superfamily Hystricoid.\e, part, medialis series. 1924. Winge: Family Hystricidae, part. 1928. Weber: Hystricoidea, part. This Superfamily is equal to the medialis series of the Hystricoidae of Miller & Gidley, and contains one family, the Caviidae, containing the genera and subgenera Cavia, Galea, Caviella, Monticavia, Nanocavia, Kerodon, Dolichotis, Paradolichotis and Hydrochoerus only. I have elsewhere, when dealing with the Hystricoidae (Hystricoidae, p. 97) remarked on the desirability of removing the Caviidae from the tj'pical Hystricoid series, on account of the different formation of the lower jaw. Apart from this structure and the formation of the cheekteeth they appear to agree in essential characteristics with the Hystricoidae; but although Chin- chillidae may show a certain resemblance in mandible formation to Caviidae, I am unable to regard the Caviidae as typical Hystricoidae now, whatever their ancestors may have been. FamUy CAVIIDAE 1896. Thomas: Hystricomorph.a : Family Caviidae. 1899. TuUberg: Hystricomorpha : Family Caviidae, part, included Dasyprocta and Cuniculus ( ' ' Coelogenys "). 1918. Miller iSc Gidley: Hvstricoid.\e (Medialis series). Family Caviidae; and Family Hydrochoeridae (Hydrochoerus and fossil allies). 1924. Winge: Family Hystricidae, Dasyproctini, part, group Caviae. 1928. Weber: Hystricoidea: Family Caviidae, part, subfamilies Caviinae {Cavia), and Hydrochoerinae (Hydrochoerus, Dolichotis). Geographical Distribution. — The greater part or the whole of South America; extending north to Panama. Number of Genera. — Six. Ch.\racters. — Zygomasseteric structure differing from that of the Hystri- coidae in the formation of the lower jaw, which has the angular process drawn backwards but not distorted outwards, and possesses a deep horizontal ridge, for the insertion of masseter medialis (according to 23S CAVIIDAE: CAVIINAE Miller & tiidley) present on side of mandible sliglitiy below alveolar level and extending from the level of the condylar process to about as far as the hinder part of M.I. Infraorbital foramen very large, as in Hystricoidae, and zygomatic plate narrow, remaining completely beneath it. Dental formula i. J, c. ;!, p. r, ni. if = 20. Cheekteeth evergrowing, unilat- erally hypsodont, normally comparatively simplified in structure, but with sharp folds and angular projections, the general effect more or less prismatic. 'Fibia and fibula not fully fused. Clavicles suppressed. External form ambulatory or cursorial; digits of hindfoot reduced to three. Tail obsolete. Lachrymal large; part of lachrymal canal open on side of the rostrum, except in the genus Dolicliotis. Two well-marked subfamilies may be recognized, which are sometimes considered as families; but which, notwithstanding the high specializations of the Hydrochoerinae, appear to agree in very many essential features. Key to the Subeamilies of the Cavud.^e M.3 not greatly enlarged; pattern of cheekteeth comparatively simple; palate short to extremely short (from before backwards); parocci- pital processes not abnormallv elongated. Subfamily Caviinae [Caria, Galea, Caviella, Kirodon ; Dolicliotis) M.3 extremely enlarged (upper series); pattern of cheekteeth compara- tively complex; palate not short (from before backwards); parocci- pital processes abnormallv lengthened. Subfamily Hvhrochoerinae (Hydroclioerus) Subfamily CAVIINAE Geographical Distribution. — As in the family, except not known from Panama. Number of Genera. — Five. Remarks. — The Caviinae fall into two well-marked groups, the Cavia group, smaller genera with short limbs, shorter ears, and moderate claws (or in Kerodon blunt nails), and the Dolicliotis group, containing a single genus, with larger size, long limbs, long cars, sharp hoof-like claws, the external form more modified for cursorial life. Doliclujtis seems to be too nearly allied to the Caviae for these groups to be regarded as subfamilies, as has been done (Pocock, Tate). The Cavia group was revised by Osgood in 1915 (Field. Mus. Nat. Hist. Publ. Zool. ser. X, no. 13, pp. 194, 195), who rightly restricted the genus Kerodon to the species nipestris, and proposed the subgenera Galea and Caviella for those species of Cavia with more simple cheekteeth. Thomas in 1916 treated Galea and Caviella as full genera, and erected Monticaria for the species niata (referred by Osgood to Caviella). Later Thomas erected Naiiocavia for a new species shiptoni, allied to Caviella and Monticavia. The Caviinae have recentlv been revised at some length by Kraglievitch, 193 1, Ann. Mus. Nac. Buenos Aires, XXXVL p. 77. He divides the subfamily CAVIINAE 239 Caviinae into the two groups here recognized. The genera Cavia, Galea, Caviella, Monticavia, and Kerodun are retained in the Cavia group; Nanocavia he reduced to a subgenus of Monticavia. On this account I retain Galea and Caviella as full genera, though I am bound to say that I feel convinced that it would be wiser to retain Osgood's original classification, in which these two groups are regarded as of subgeneric value only, as the characters which separate them from each other are of very doubtful value. But Monticavia is so closely allied to Caviella that I cannot treat it as more than a subgenus. The main differences between Monticavia and Caviella are that the heel in M.3 of Monticavia is less sharply defined, and that the incisors are in Monticavia more pro-odont. But Kraglievitch gives the measurement of the angle of the incisors of Caviella as between 88° and 1 10°, and that of Monti- cavia (Xanocavia) shiptoni as 1 11°, so that the difference in this respect appears to amount to one degree between the two "genera," which is hardly sufficient to base a generic name on ! Kerodon is a distinct genus, which cannot be confused with any of the other members of the Cavia group, whatever their status may be. Key to the Groups of the Caviin.\e The limbs shorter; ears short; claws not hooflike, less broadened, or may be blunt; nasals not narrowed and pointed anteriorly; interorbital region narrower; paroccipital processes less lengthened. CAVL^i Group (Caviae) {Cavia, Galea, Caviella; Kerodon) The limbs longer; ears long; claws powerful and hoof like; nasals markedly narrowed and pointed anteriorlv; interorbital region very broad; paroccipital processes more lengthened. Dolichotis Group (Dolichotides) {Dolichotis) The Cavia Group Characters as indicated in the above key. Size medium or small, not becoming large. In all the genera, the jugal is broad but rather short, and zygoma not angular; incisors relatively short, narrow; palate extremely constricted anteriorlv, the premolars almost touching; upper cheekteeth much higher on inner side than outer side, the lower cheekteeth much higher on outer side than inner side. Mandible with coronoid process obsolete; condylar process of medium height; angular process drawn far backwards, but not distorted outwards. Beside and below the condylar process and extending forwards about to level of hinder part of M.i is an extremely deep and prominent ridge for insertion of masseter medialis. Key to the Ge.nera of the Cavia Group Claws blunt. Sternum narrow and rounded (Osgood). Kerodon 240 CAVIA Claws sharp. Sternum broad and Hat (Osgood). Posterior lobe of upper cheekteeth with a clear and deep outer re- entrant fold; dental pattern less simplified. Cavi.^ Posterior lobe of upper cheekteeth with no re-entrant fold; dental pattern more simplified. Orbital branch of maxillae completely interrupted by the lachrymal ; incisors pigmented; skull not bowed. G.alea Orbital branch of maxillae not completely interrupted by the lachrymal; incisors not pigmented; skull bowed to a greater or lesser degree. Caviella The ditferences between Galea and Caviella are based on characters which are in other groups very variable; for instance, in Dolichotis the interruption of the orbital branch of the maxillae by the lachrymal may be present or absent. The incisors may or may not be pigmented within many genera elsewhere in the Order, for instance, Ctenoiiivs, Xenis (Geosciunis), and others. The orbit is more circular in Caviella than in Galea. Genus i. CAVIA, Pallas 1766. Cavia, Pallas, Misc. Zool. p. 30. Type Species. — Cavia cubava, Pallas = Miis porcellus, Linnaeus. Range. — South America; Brazil, the Guianas, Venezuela, Colombia, Peru, Bolivia south to Northern Argentina. Number of Forms. — Approximately seventeen. Characters. — Skull with some interorbital constriction apparent, and a sagittal crest developed in the adult. Infraorbital foramen broader below than above; a canal present for transmission of nerve. Bullae relatively large. Paroccipital processes noticeably elongated. Palate short, ex- tending about to front of M.3. Palatal foramina short, narrow. Jugal medium, not approaching the lachrymal. Incisors not pigmented. Upper cheekteeth divided into two lobes by inner re-entrant fold in the upper series, the hinder lobe larger than the anterior one, and with a deeply indenting fold in its outer border. M.3 with posterior projection. Lower cheekteeth with one deep outer fold dividing tooth into two lobes and with an inner fold in the posterior lobe. Mandible as already described. Externally the limbs not specially elongated, the hindfeet long, with three digits, the central digit the longest; the claws sharp. Forefeet with four digits, D.3 the longest, D.5 the shortest; D.4 rather longer than D.2. Ears relatively short. This genus is quite well differentiated from Galea and Caviella by the more complex cheekteeth. Forms seen : anolaiinae, apereti, azarae, festiiui, fulgiiia, niiianae, nana, pain- parnm, pallida, porcellus, rosida, stolida, tsclnidii, umhrata. For notes on the species of Cavia see I'homas, Ann. Mag. Nat. Hist. 8, XIX, p. 152, 1917. CAVIA 241 List of Named Forms (References and type localities of all named forms for the Caviidac are the work of I\Ir. G. W. C. Holt.) 1. CAVIA GUIANAK, Thomas 1901. Ann. Mac. Nat. Hist. 7, VIII, p. 152. Kanuku Mountains, British Guiana. 2. CAVIA VENKZUELAIC, Allen 191 1. Bull. Amer. Mus. Nat. Hist. XXX, p. 250. Altagracia, Immataca district, Venezuela. Considered by Thomas as doubtfully distinguishable from guianae. 3. CAVIA .\PI;REA APEREA, Erxleben 1777. Syst. Regn. .Anim. i, p. 348. Brazil. Synonym: leiicopyga, Brandt, 1835, M^m. Acad. St. Petersb. 6, iii, p. 436. Brazil. 4. CAVIA APEREA AZ-iVR-^E, Lichtenstein 1823. Doublet. Zool. Mus. Berlin, p. 3. Ypamena, Province Sao Paulo, Brazil. 5. CAVIA ROSIDA, Thomas 1917. .Ann. Mag. Nat. Hist. 8, XIX, p. 154. Roca Nova, East Parana, Brazil. 6. CAVIA PAMPARUM, Thomas 1901. .Ann. Mag. Nat. Hist. 7, VIII, p. 539. Goya, Corrientes, Argentina. 7. CAVIA TSCHUDII TSCHUDII, Fitzinger 1867. Sitz.-Ber. K. Akad. Wien (Math. Nat.), LVI, p. 154. Cit>' of Yea, 70 miles east of Pisco, Western Peru. 8. CAVIA TSCHUDII UMBRATA, Thomas 1917. .Ann. Mag. Nat. Hist. 8, XIX, p. 157. Incapirca, Zezioro, Central Peru. 9. CAVIA TSCHUDII AREQUIPAE, Osgood 1919. Journ. Mamm. Baltimore, p. 34. .Arequipa, Peru. Synonym: tschudii pallidior, Thomas, 1917, Ann. Mag. Nat. Hist, 8, XIX, p. 158. 'Sox. (niata) pallidior, Thomas. The name arequipae was proposed in case Monticavia was regarded as not distinguishable generically from Cavia. Perhaps pallidior should stand in the present work, as niata pallidior is regarded as a Caiietla. 10. CAVIA TSCHUDII STOI.IDA, Thomas 1926. .Ann. Mag. Nat. Hist. 9, XVIII, p. 166. Rio Utcubamba, 15 miles south of Chachapoyas, Peru. 11. CAVIA TSCHUDII FESTINA, Thomas 1927. Ann. Mag. Nat. Hist. 9, XX, p. 604. Huariaca, Junin, Peru. 12. CAVIA TSCHUDII SODALIS, Thomas 1926. Ann. Mag. Nat. Hist. 9, XVII, p. 607. Norco, 20 kilometres north-west of Vipos, Prov. Tucuman, .Argentina. 16 — Living Rodents — 1 242 CAVIA— GALEA 13. CAVIA TSCHUDII ATAHUALPAE, Osgood 1Q13. Field Mus. Nat. Hist. Publ. Zool. ser. X, no. 13, p. gS. Cajamarca, Peru. 14. CAVIA NANA. Thomas 191 7. Ann. Mag. Nat. Hist. !S, XIX, p. 158. Bolivian Highlands (Chulumani, Yungas). 15. CAVIA FULGIDA, Waglcr 1 83 1. Isis, XXIV, p. 512. .Amazonia. Synonym: rufescens, Lund, 1841, Afh. K. Danske. Vid. Selsk. 4, VIII, p. 284. Lagoa Santa, Brazil. nigricans, Wagner, 1841, Schreber, Saug. Suppl. IV, p. 64. Brazil. 16. CAVIA .ANOLAIMAE, Allen 1916. Bull. .\mer. Mus. Nat. Hist. XXXV, p. 85. Anolainia, west of Bogota, Colombia; on a branch of River Bogota. 17. CAVIA PORCELLUS, Linnaeus. (Domestic) 1758. Syst. Nat. loth ed., i, p. 59. Brazil. Synonym: aperoides, Lund, Blik. Dyr. pi. 25. Brazil. robiista, Lund, 1841, Blik. Dyr. pi. 25, fig. 16. brasilieiisis, Linnaeus, 1754, Mus. Adolphi. Friederici, p. 9. gracilis, Lund, 1841, Blik. Dyr. pi. 25. ciitleri, Bennett, 1835, Proc. Zool. Soc. London p. 191. Lima, Peru. cobaya, Marcgrave, 1648, Hist. Nat. Bras. p. 224. Peru. longipilis, P'itzinger, 1879, Sitz.-Ber. K. Akad. Wien (Math. Nat.), LXXX, Ab. i , p. 43 1 . Japan. Genus 2. GALEA, Meyen 1833. Galea, Meyen, Nova Acta Ak. Caes. Leop. XVI, 2, p. 597. Type Species. — Galea musteloides, Meyen. Range. — Bolivia, North Argentina, Chile and Brazil. Number of Forms. — Ten. Characters. — Like Cavia but cheekteeth simpler, each upper tooth cut into two lobes by one inner fold; M.^ with weak backwardly projecting heel. Lower teeth two-lohed; P. 4 with short anterior prolongation. Orbital branch of maxillary completely interrupted by lachrymal. Incisors pigmented. Forms seen : aiiceps, bolivieiisis, copies, demissa, flavidiis, littoialis, iiegremis, palustris, spixii. List of Named Forms I. GALEA MUSTELOIDES MUSTELOIDES, I\Teyen 1833. Nova Acta Ak. Caes. Leop. XVI, 2, p. 598. Pass of Tacara and Tajori, Andes, North-west Bolivia. Synonym: bolizicnsis, Waterhouse, 1848, Nat. Hist. Mamm. 11, p. 175. Bolivia, highlands between Cochabamba and La Paz. comes, Thomas, 1019, .•\nn. Mag. Nat. Hist. 9, IV, p. 134. Maimara, Jujuy, Argentina. GALEA— CAVIELLA 243 2. GALEA MUSTKLOIDIiS Li;UCOBLKPH.\RA, Burmeister 1861. Reise durch La Plata, H, p. 425. Mendoza to Tucuman, Argentina. 3. GALEA MUSTELOIDES LITTORALIS, Thomas igoi. Ann. Mag. Nat. Hist. 7, VU, p. 195. Bahia Blanca, Argentina. Synonym: miisteloides negrensis, Thomas, 19 19, Ann. Mag. Nat. Hist. 9, HL p. 211. Pilcaneu, Upper Rio Negro, .-Argentina. 4. GALEA MUSTELOIDES DEMISSA, Thomas 1921. Ann. Mag. Nat. Hist. 9, VIII, p. 623. San Antonio, Parapiti, Bohvia. 5. GALEA MUSTELOIDES AUCEPS, Thomas 1911. Ann. Mag. Nat. Hist. 8, VIII, p. 255. Guarina, Lake Titicaca, Bolivia. 6. GALEA MINIMUS, Molina 1782. Sagg. Stor. Nat. Chili, ist ed., p. 306. Chile. Considered a subspecies of miisteloides by Tate ; if this is so, the name antedates miisteloides, and all races must be regarded as races of minimus. 7. GALEA SPIXII, Wagler 1831. Isis, XXIV, p. 512. Brazil. Synonym: saxatilis, Lund, 1841, Afh. K. Danske Vid. Selsk. 4, VIII, p. 286. Lagoa Santa, Brazil. 8. GALEA WELLSI, Osgood 1915. Field Mus. Nat. Hist. Publ. Zool. ser. X, no. 13, p. 196. Sao Marcello, Bahia, Brazil. 0. GALEA PALUSTRIS, Thomas 191 1. Ann. Mag. Nat. Hist. 8, VII, p. 608. Cameta, lower Rio Tocantins, Brazil. 10. GALEA FLAVIDENS, Brandt 1835. M^m. Acad. St. Petersb. p. 439. Brazil. Synonym: obscurus, Lichtenstein, 1823, Doublet. Z. Mus. Berlin, p. 3. Brazil. bilobidens, Lund, 1841, Afh. K. Danske Vid. Selsk. 4, VIII, p. 286. Brazil. Genus 3. CAVIELL.A, Osgood 1915. C.^vlELL.^, Osgood, Field. Mus. Nat. Hist. Publ. Zool. ser. X, no. 13, p. 194. Regarded by Kraglievitch as indistinguishable from Microcavia, Genais and Ameghino, 1880, .Mamm. Foss. Anier. Sud. p. 50, a fossil genus. 1916. MoNTiCAViA, Thomas, Ann. Mag. Nat. Hist. 8, XVIII, p. 303. Cavia niata, Thomas. Valid as a subgenus. 1925. Nanocavia, Thomas, Ann. Mag. Nat. Hist. 9, XV, p. 41S. Manocavia shiptoni, Thomas. Valid as a subgenus. 244 CAVIELLA Type Speciks. — Cavia amtralis, Geoffroy & D'OrhiKiiy. Range. — Bolivia and Argentina, south to Patagonia. Number of Forms. — Eight. Characters. — Si\ull with rostrum slanting downwards anteriorly, more bowed than in allies, the highest part of the skull usually about over posterior zygomatic root. Palatal foramina larger than in preceding Fig. 72. Caviella australis joannia, Thomas. B.M. No. 71. 12. 29. 12, 9; X 2. genera, triangular, placed more closely to toothrows. Sagittal crest present in old age. Bullae relatively larger, and orbit more circular than in Curia and Galea. Incisors without pigment. Cheekteeth like Galea, but usually I\1.3 with deeper posterior fold. Monticai'ia, here regarded as a subgenus of Caviella, has more pro-odont incisors, the angle with the line of toothrow about 1 1 s . M.3 is less complicated, the heel a short projection, without internal notch. Skull more bowed anteriorly. Nafiocavia, as remarked above, is intermediate between typical Caviella and Monticavia in the angle of the incisors; the bullae are considerably smaller than in either, the portion appearing on occipital surface of skull practically uninflated. Fig. 73. Caviella australis joannia, Thomas. B.M. No. 71. 12. 29. 12, ?; X 2. Fig. 74. C.-\vi£LL.\ australis joannia, Thomas. Cheekteeth: B.M. No. 71. 12.29. 12, ?; x 6. 246 CAVIELLA— KERODON Forms seen: aiistnilis, juaiinia, maenas, niata, " iiigiuinti," pa/lidiur, salinia, shiptoni. List of Named Forms Subgenus Cai-ie/hi, Osgood 1. CAVIELLA AUSTRALIS AUSTRALIS, Gcoffroy S: D'Orbigny 1833. Mag. Zool. I. pi. 12. Rio Negro, Patagonia. Synonym: australis nigriana, Thomas, i(j2i, Ann. Mag. Nat. Hist. 9, VII, p. 446. Neuquen, Rio Negro, Argentina. 2. CAVIELLA AUSTR.ALIS KINGII, Bennett 1835. Proc. Zool. Soc. London, p. iqo. Port Desire, Patagonia. 3. CAVIELLA AUSTRALIS JOANNIA, Thomas 1921. Ann. Mag. Nat. Hist. 9, VII, p. 446. Canada Honda, San Juan, Argentina. 4. C.WIELLA AUSTRALIS MAENAS, Thomas 1898. Ann. Mag. Nat. Hist. 7, I, p. 284. Chilecito, Rioja, Argentina. 5. CAVIELL.A AUSTRALIS SALINIA, Thomas 1921. Ann. Mag. Nat. Hist. 9, VII, p. 447. South-east Catamarca, .Argentina. Subgenus Nanocavia, Thomas 6. CAVIELLA SHIPTONI, Thomas 1925. Ann. Mag. Nat. Hist. 9, XV, p. 419. Laguna Blanca, Catamarca, Argentina. Subgenus Monticavia, Thomas 7. CAVIELLA NIATA NIATA, Thomas 1898. Ann. Mag. Nat. Hist. 7, I, p. 282. Espcranza, 50 km. from Mt. Sahama, Bolivia. 8. CAVIELLA NIATA PALLIDIOR, Thomas 1902. Ann. Mag. Nat. Hist. 7, IX, p. 229. Pampa .Aullaga, Lake Poopo, Bolivia. Genus 4. KERODON, F. Cuvier 1825. I'iERODON, F. Cuvier, Dents, des Mamm. p. 151. Type .Species. — Cavia nipcstiis, Wied. Range. — Brazil. (British Museum specimens froin Bahia.) Number of Forms. — One. Characters. — Much like Curia cranially. Sagittal ridge feeble or absent. Infraorbital foramen with no canal for nerve transmission. KERODON— DOLICHOTIS 247 Palatal foramina excessively narrowed. Rostrum relatively narrow. Bullae moderate; paroccipital processes rather long. Upper cheekteeth two-lobed; 1VI.3 with a weak backwardly projecting heel. P.4 lower with a well-marked e.xtra anterior projection; heel of M.3 (lower) poorly defined. Differing, according to Osgood, in several details of the skeleton from Cavia and allies, the chief of which is that the sternum in this genus is narrow and rounded instead of broad and flat, the spinous processes of the lumbar vertebrae are thick, heavy, and depressed, and the large neural spine of the axis fully overlaps the first cervical. Externally differing markedly from all allies in the fact that the digits are armed only with blunt nails. Remarks. — Whatever the status of Galea and Caviella compared with Cavia, there is no doubt that on account even of the nails alone, this genus is distinct from that group. Forms seen : rupestris. List of Named Forms I. KERODON RUPESTRIS, Wied 1820. Isis, VI, p. 43. Rio Grande de Belmont, Rio Pardo, etc., Brazil. Synonym: moco, F. Cuvier, Dents, des Mamm. 1825, p. 151. Brazil. sciureus, Geoffroy, 1826, Diet. Class. IX, p. 120. Brazil. The Dolichotis Group Becoming larger than the Cavia group; to very large (head and body 690, or more?); hindlimbs lengthened, general form modified for cursorial life. Hindfoot very long, with three digits bearing hooflike claws; arrangement of digits perissodactyle. A rudimentary tail present. Forefoot artiodactyle, the four digits armed with sharp claws. Frontals much broadened, and nasals considerably specialized. Genus 5. DOLICHOTI