L HISTORY OF CE VOLUME XI MALS OF CHINA A PART I FOR THEEEOPLE FOR EDVCATION FOR SCIENCE LIBRARY OF THE AMERICAN MUSEUM OF NATURAL HISTORY RAT Y i R) Ave an ) heh poles 4 i Lak CENTRAL ASIATIC EXPEDITIONS ROY CHAPMAN ANDREWS, Leader THE MAMMALS OF CHINA AND MONGOLIA BY GLOVER M. ALLEN, PrD. CURATOR OF MAMMALS MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS With 22 Distribution Maps and I other I Illustration in the Text and with 9 Plates NATURAL HISTORY OF CENTRAL ASIA VOL. XI, PART 1 WALTER GRANGER, DSc., Editor THE AMERICAN MUSEUM OF NATURAL HISTORY F. TRUBEE DAVISON, President NEW YORK 1938 THE MAMMALS OF CHINA AND MONGOLIA ~~ Copyright, 1938 by The American Museum of Natural History Published September 2, 1938 First Edition All rights reserved. This book, or parts thereof, must not be reproduced in any form without permission. 25RAL Pry 9s S1 - Made in the United States of America. PUBLICATION NOTE BECAUSE of the size of this report, treating as it does more than 500 forms of mammals, it has been found neces- sary to issue the volume in two parts, of which this is the first. Part 2, begin- ning with the typical rodents, will be brought out as soon as possible. The Bibliography of the entire work appears at the end of Part 1; the Index will be at the end of Part 2. [aa TRU Tab eae PAOUELE TARUT ATS PREFACE THE collections of Chinese and Mongolian mammals brought together through the field work of the Asiatic Expeditions of The American Museum of Natural History exceed any previously made, not only in the number of specimens secured, but also in the excellence of their preservation and the diversity of the localities which they represent. They form, therefore, a fairly adequate basis for a review of the mammalian fauna of these countries. In short preliminary papers, brief reports on this material have been published, group by group, during the past few years, with descriptions of species or races believed to be new. The present volume is intended to summarize our knowledge of the mammalian fauna of China and Mongolia and to serve as a handbook for the use of those interested. To this end, in addition to brief summary chapters, keys are given for the various orders, families, genera, and species, to facilitate the identification of specimens, while under each species or race in the systematic portion are given: the accepted Latin name, important synonymy, description of external and cranial characters, measurements, notes on nomenclature where necessary, the known facts of occurrence and habits within the area treated, and a list of localities from which specimens have been examined by the writer. While political boundaries usually form un- satisfactory units for the study of zodgeography, the general limits of China and Mongolia as set down in maps of the past decade have been taken, chiefly as a matter of convenience and partly because no significantly large collections have been available from without these limits. For these reasons, Korea and parts of Manchuria are excluded from consideration as well as the island of Formosa, the fauna of which, however, is essentially derived from that of the neighboring mainland. It should be added that all specimens referred to as obtained by the American Museum Asiatic Expeditions include those collected by Dr. Andrews and his associates in the course of their explorations in Fukien and Yunnan in 1916-17, and again in northern Mongolia in 1919. These two expeditions were officially designated as the First and Second Asiatic Expeditions. All specimens collected in China and in Mongolia during the work of 1921 to 1930 are credited to the Central Asiatic Expeditions, known in the early years of the organization as the Third Asiatic Expedition. Vv vi PREFACE In the preparation of this volume, a critical study has been made, to determine, so far as possible, the closer relationships of many eastern forms. Until recent years, it has frequently been the custom, at least among European naturalists, to give specific rank to mammals described as new, notwithstanding that the differences in comparison with other forms may be merely quantitative and the true relationships therefore much closer than between those forms differing qualitatively. An exaggerated conception of the number of distinct types or “‘species’’ is thus produced where relationships are really but sub- specific and better expressed by a trinomial. To demonstrate this relationship satisfactorily, however, is often not easy and must necessitate in most cases a “revision” of several related forms. New names have frequently been be- stowed on the basis of inadequate material; specimens of a single species prepared by different collectors at first sight often present slightly differing appearances; or differences of sex, age, season, or condition may mislead the investigator into regarding individuals of the same species as representing different races. Again, single specimens sometimes show individual peculiari- ties not duplicated in large series. A comparison with authentic material, types or topotypes, is often the only means of rectifying such mistakes, which are inevitable in the earlier stages of knowledge. Such steps are reflected in the synonymy of a species. It is too much to expect that all questions of relationship and nomenclature in the forms treated here are satisfactorily solved. Future workers will be able to correct many errors and extend the number of species or races known to occur within the limits of China and Mongolia. Nevertheless, in a general way it may fairly be said that the mam- mals of this area are now tolerably well known from a systematic point of view; many, however, remain rare in collections, while as to the more intimate points of life history we are still largely ignorant. Throughout the preparation of this volume, I have had the hearty and unfailing coéperation of Dr. H. E. Anthony, Curator of Mammals, and of Dr. Roy C. Andrews, leader of the Asiatic Expeditions of The American Museum of Natural History, as well as that of other members of the staff. The collec- tions of Chinese mammals in the Museum of Comparative Zodlogy, the Academy of Natural Sciences of Philadelphia, the University of Michigan, and at request, in the United States National Museum, have also been consulted. Through the aid of a grant from the Milton and Clark Fund of Harvard University, I was enabled in 1933 to spend several weeks at the British Museum, in the study of its unrivaled collection of historic specimens, types, and other material, without which it would have been impossible to correct many previous errors and misconceptions of my own and of other workers less fortunate. To all these institutions and their officers in charge, grateful thanks are due. PREFACE Vii In listing specimens examined, it is understood that where no abbreviation follows, the numbers if given, are those of the American Museum of Natural History. Abbreviations are: A. N.S. P., for Academy of Natural Sciences of Philadelphia; B. M., for British Museum (Natural History), London; M. C. Z., for Museum of Comparative Zodlogy, Cambridge, Massachusetts; U.S. N. M., for United States National Museum, at Washington, D. C.; Univ. Mich., for Museum of Zoédlogy, University of Michigan, Ann Arbor, Michigan. The Bibliography at the end of Part 1 of this Volume includes all literature on Chinese and Mongolian mammals that I have come upon, except incidental references, through 1937. Three new names are here proposed: Aéretes, a new genus for the flying squirrel Pteromys melanopterus; Eospalax, a new subgenus for the mole-rats typified by Myospalax fontanierti; and a new subspecies of Flying Squirrel. The photographic illustrations are selected chiefly from the large series made in the course of their work by Dr. Roy Chapman Andrews and Dr. Walter Granger, to whom I would express my sincere thanks for the use of them as well as for much other help. The cut illustrating the nose-leaves of the horseshoe bats, Hipposideros armiger and H. pratti, has been generously lent by the Field Museum of Chicago, through the kind offices of Dr. Wilfred H. Osgood, in whose paper of 1932 it first appeared. I am further indebted to Mr. Clifford H. Pope and Dr. Walter Granger of the American Museum of Natural History for many valuable notes, and particularly for their efforts to standardize the spelling of the place-names mentioned in the text. These so far as possible conform to the most modern usage as found in the List of Post Offices, thirteenth issue, Shanghai, 1932, and the Postal Map of China (Shanghai, 1920), as well as in other authentic sources. For many species, outline maps of distribution have been prepared as text-figures, but it should be well understood that these indicate the range of each in only the most general way, for there are still vast areas in China and Mongolia in which little or no collecting has been done, so that often large sections of the supposed range are included on inference only, while on small- scale maps details of local distribution, although often sharply marked, cannot well be plotted. Nevertheless, such maps are valuable as enabling a better visualization of the known or supposed areas of distribution, while at the same time indicating something of the extent of our ignorance of them. No doubt many of these maps will be found subject to considerable later correction or elaboration as the mammalian fauna of the region becomes known in more detail. Since the present report was written, many and great changes have taken place in the political organization of China and Mongolia, involving the re- viii PREFACE definition of certain of the former and more familiar provinces, so that in some cases their boundaries no longer coincide with those in use only a few years ago. Thus the western border of Szechwan is now drawn considerably to the eastward of its former course, so that specimens recorded from the well-known locality Tatsienlu are now to be regarded as from the new province, Hsikang, which stretches westward toward Tibet beyond the area covered in this book. The old province of Chihli, in North China, is now Hopei, and there are in addition the new provinces of Chinghai, Ningsia, Suiyuan and Chahar. All these changes are confusing, but have been taken account of in the distribu- tional maps in the text. I cannot too warmly express my indebtedness to Dr. Walter Granger, and his assistants, Miss Clara M. Beale and Miss Ruth Tyler, for the great care with which they have edited, checked and coérdinated the manuscript, thereby correcting many minor errors that might otherwise have escaped me. GLOVER M. ALLEN. Museum or ComPaARATIVE ZoOLocy, HARVARD UNIVERSITY, CAMBRIDGE, MASSACHUSETTS. SEPTEMBER I, 1937. CONTENTS OF PART 1 PREFACE. : F ; ‘ : : 2 : : 5 : ; 4 v List oF TExT FIGURES : : : : , : : é : : bol List oF PLATES . : é : ; : : ; : ‘ : 2 By 2.2.5i/ SECTION I—GENERAL INTRODUCTION ; . : S ; 3 : I CHAPTER I—COLLECTORS OF CHINESE AND MONGOLIAN MAMMALS : : : é 3 II.—FAUNAL AREAS OF CHINA AND MONGOLIA : é ; ¢ é 5 9 III—FAUNAL RELATIONS OF ASIA WITH NORTH AMERICA F : 5 20 SECTION II—SYSTEMATIC ACCOUNT OF THE MAMMALS OF CHINA AND MONGOLIA . F ; ‘ ‘ , 2 : ; : : Seay) IV.—ORDER INSECTIVORA. INSECTIVORES ; : ‘ ; 3 : : 29 Key to the Families of Chinese and Mongolian Insectivora ~ . ‘ 30) Family Tupaiide. Tree Shrews : é : : d : 2 830 Genus Tupaia Raffles : ‘ : ‘ £930 Key to Chinese Races of iin nen : 3 : net OE 1. Tupaia belangert chinensis J. Anderson . : ; Eee ah 2. Tupaia belangeri yunalis Thomas . ; : ; a Tiga 3. Tupaia belangeri modesta J. A. Allen . F : casa. Family Erinaceidez. Hedgehogs and their Allies : : 530 Key to the Genera of Chinese and Mongolian Hieneetdes : a ete Genus Hylomys S. Muller : ‘ : : 2 2) 037 4. Hylomys suillus peguensis Blyth : : : ; ee 37, Genus Neotetracus Trouessart . : , ‘ ‘ 3 yee) 5. Neotetracus sinensis ieopesaer : . : 3 =, 140 Genus Hemiechinus Fitzinger . : : : 2 42 6. Hemiechinus dauuricus dauuricus (Sundevall) : : wea 7. Hemuechinus dauuricus alaschanicus Satunin . : Biss Genus Erinaceus Linnzus j : : : ne 47, 8. Erinaceus europeus dealbatus Sraitine , 4 : Sy PA7 9. Erinaceus europeus miodon Thomas : : y a fh '52 Family Talpide. Moles : , ; : : : aT) 54 Key to the Genera of Chinese iaieidee : i : ; 5 yt Genus Uropsilus Milne-Edwards : d 4 5 ; 6 55 10. Uropsilus soricipes Milne-Edwards : : ; ae 7, Genus Rhynchonax Thomas A : : : 5 : Ree (9) ix CHAPTER CONTENTS 11. Rhynchonax andersoni andersonit Thomas 12. Rhynchonax andersoni atronates G. M. Allen 13. Rhynchonax andersoni nivatus G. M. Allen Genus Nasillus Thomas : : : ; 14. Nasillus gracilis Thomas 15. Nasillus investigator Thomas Genus Scaptonyx Milne-Edwards ? 16. Scaptonyx fusicaudatus Milne-Edwands . y 17. Scaptonyx fusicaudatus affinis Thomas Genus Talpa Linnzus , 18. Talpa longirostris Mileo-Edwanis. Genus Parascaptor Gill. 19. Parascaptor leucurus (Blyth) Genus Scaptochirus Milne-Edwards 20. Scaptochirus moschatus moschatus Milne-Edwards Scaptochirus moschatus gilliesi Thomas . Gene M ogera Pomel 22. Mogera latouchei Thomas 23. Mogera hainana Thomas Genus Scapanulus Thomas 24. Scapanulus oweni Thomas Family Soricide. Shrews : Key to the Genera of Chinese and ‘Mongolian Soricid: Genus Sorex Linnzeus ; Key to the Chinese and Mongolian Goeaten of Sores ; 25. Sorex araneus borealis Kastschenko 26. Sorex excelsus G. M. Allen 27. Sorex sinalis Thomas . ‘ 28. Sorex buxtoni buxtoni J. A. Allen . 29. Sorex buxtont cansulus Thomas 30. Sorex minutus thibetanus Kastschenko . 31. Sorex cylindricauda cylindricauda Milne-Bdwards 32. Sorex cylindricauda wardi Thomas : 33. Sorex cylindricauda gomphus G. M. Allen Genus Soriculus Blyth ‘ : : Key to Chinese Species of ad 34. Soriculus macrurus Blanford 35. Soriculus caudatus sacratus Thomas 36. Soriculus caudatus umbrinus G. M. Allen Genus Chodsigoa Kastschenko ; Key to Chinese Species of Chodsigoa ‘ 37. Chodsigoa hypsibia hypsibia (De ee and Styan) 38. Chodsigoa hypsibia larvarum Thomas 39. Chodsigoa hypsibia lamula Thomas 40. Chodsigoa parva G. M. Allen ; 41. Chodsigoa salenskii (Kastschenko) 42. Chodsigoa smithii smithit Thomas CONTENTS CHAPTER 43. Chodsigoa smith parca G. M. Allen Genus Blarinella Thomas : Key to Chinese Races of Blanelia é 44. Blarinella quadraticauda quadraticauda " (Milne- Bdwacis) 45. Blarinella quadraticauda griselda Thomas 46. Blarinella quadraticauda wardi Thomas . Genus Suncus Hemprich and Ehrenberg 47. Suncus murinus (Linnzus) Genus Crocidura Wagler ; Key to Chinese and Mongolian Bpenies oe Greciduea 48. Crocidura attenuata Milne-Edwards 49. Crocidura dracula dracula Thomas 50. Crocidura dracula grisescens A. B. Howell 51. Crocidura tlensis tlensis Miller 52. Crocidura ilensis lar G. M. Allen 53- Crocidura ilensis shantungensis Miller 54. Crocidura ilensis pheopus G. M. Allen 55. Crocidura vorax G. M. Allen 56. Crocidura rapax G. M. Allen Genus Anourosorex Milne-Edwards : : 57. Anourosorex squamipes Milne-Edwards . Genus Chimarrogale Anderson Key to Chinese Species and Gaheneeies oF Chameareeate 58. Chimarrogale himalayica himalayica (Gray) 59. Chimarrogale himalayica leander Thomas 60. Chimarrogale styani De Winton and Styan Genus Nectogale Milne-Edwards : j 61. Nectogale elegans Milne-Edwards V.—ORDER CHIROPTERA. BATS Key to the Families of Chinese and Mecr poten Giisgnten Family Pteropide. Fruit Bats or Flying Foxes Key to the Genera of Chinese Pteropidez [Pteropus chinensis Gray] [Pteropus formosus P. L. Sclater] Genus Cynopterus F. Cuvier. 62. Cynopterus sphinx sphinx (Vahl) 63. Cynopterus brachyotis angulatus Miller Genus Rousettus Gray 64. Rousettus leschenaulti ips Family Emballonuride. Sheath-tailed Bats Genus Taphozous Geoftroy 65. Taphozous melanopogon Cee Se Family Megadermide. Big-eared Bats Genus Lyroderma Peters 66. Lyroderma lyra sinensis ikem a Wroughton) Family Rhinolophide. Leaf-nosed Bats . xii CHAPTER CONTENTS Genus Rhinolophus Lacépéde = Key to Chinese Species of Rhinolophus . Rhinolophus rouxi sinicus Andersen . Rhinolophus affinis himalayanus Andersen Rhinolophus affinis macrurus Andersen . . Rhinolophus affinis hainanus J. A. Allen . Rhinolophus ferrum-equinum nippon Temminck . Rhinolophus ferrum-equinum tragatus Hodgson Rhinolophus lepidus shortridgei Andersen Rhinolophus cornutus pumilus Andersen . Rhinolophus blythi szechwanus Andersen Rhinolophus blythi calidus G. M. Allen . . Rhinolophus blythi parcus G. M. Allen . Rhinolophus pearsonii pearsonit Horsfield Rhinolophus pearsonii chinensis Andersen . Rhinolophus episcopus episcopus G. M. Allen . Rhinolophus episcopus caldwelli G. M. Allen . Rhinolophus lanosus lanosus Andersen Rhinolophus lanosus spurcus G. M. Allen Rhinolophus rex G. M. Allen : Family Pipacadenda! Horseshoe Bats . Key to the Genera of Chinese Hipposideridz Genus Hipposideros Gray : Key to the Chinese Species of Hipposideros 85. 86. 87. 88. - 89. Hipposideros armiger armiger (Hodgson) Hipposideros armiger swinhoii (Peters) . Hipposideros prattt Thomas . Hipposideros poutensis J. A. Allen Hipposideros gentilis sinensis Andersen . Genus Trienops Dobson go. Trienops wheeleri Osgood Genus Celops Blyth gl. 92. Celops inflata Miller Calops sinicus G. M. Allen . Family Vespertilionide. Vespertilionine Bats . Key to the Genera of Chinese and Mongolian Metereilicniiie ; Genus Myotis Kaup Key to the Chinese and Marsnasiess Boece: of M ne 93- : A 94. 95- 96. 97. 98. 99. 100. IOI. Myotis myotis ancilla Thomas Myotis altarium Thomas Myotis chinensis chinensis (eines) Myotis chinensis luctuosus G. M. Allen . Myotis formosus rufo-niger (Tomes) Myotis pequinius Thomas Myotis daubentonit (Kuhl) Myotis fimbriatus (Peters) Myotis mystacinus mystacinus (Kuhl) CHAPTER CONTENTS 102. Myotis mystacinus przewalskii Bobrinski 103. Myotis laniger (Peters) 104. Myotis frater G. M. Allen 105. Myotis muricola moupinensis (Milne-Edwards) 106. Myotis davidii (Peters) Genus Rickettia Bianchi 107. Rickettia pilosa (Peters) Genus Pipistrellus Kaup Key to Chinese Species of Pipisiealius 108. Pipistrellus pulveratus (Peters) 109. Pipistrellus abramus (Temminck) 110. Pipistrellus tralatitius tramatus Thomas Genus [a Thomas 111. Jato Thomas Genus Nyctalus Bowdich . : Key to the Chinese Forms of NV’ Worl 112. Nyctalus aviator Thomas ; 113. Nyctalus noctula plancyi (Gerbe) . 114. Nyctalus velutinus G. M. Allen Genus Eptesicus Rafinesque Key to Chinese and Mongolian Species of pee 115. Eptesicus nilssonii gobiensis Bobrinski . 116. Eptesicus serotinus pallens Miller 117. Eptesicus alashamicus Bobrinski . 118. Eptesicus andersoni Pe. Genus Vespertilio Linnzus 119. Vespertilio murinus murinus Ppneceis 120. Vespertilio murinus superans Thomas . Genus Tylonycteris Peters . 121. Tylonycteris pachypus pails (Blyth) . 122. Tylonycteris robustula Thomas. : = Genus Scotomanes Dobson. 123. Scotomanes ornatus sinensis Thomas Genus Scotophilus Leach Key to Chinese Species of Stoinsisilis 124. Scotophilus kuhlii Leach 125. Scotophilus heathii insularis J. A. ‘Allen 126. Scotophilus temminckwi consobrinus J. A. Allen Genus Barbastella Gray 127. Barbastella Wakeclincenses (Gisdeeen) Genus Plecotus Geoffroy Key to Chinese and Manette Busine of Pleeoius 128. Plecotus auritus auritus (Linnzus) 129. Plecotus auritus kozlovi Bobrinski 130. Plecotus auritus ariel Thomas Genus Miniopterus Bonaparte Key to Chinese Species of Miniopterus xiii PAGE 217 218 220 221 223 224 224 226 227 227 229 231 231 232 233 233 234 235 236 237 238 238 240 242 243 244 245 246 247 248 249 249 250 250 252 252 253 254 256 256 258 258 258 260 262 263 263 xiv CONTENTS CHAPTER 131. Muiniopterus schreibersit chinensis Thomas 132. Miniopterus schreibersit parvipes G. M. Allen 133. Miniopterus pusillus Dobson Genus Kerivoula Gray , ‘ Key to the Chinese Species of Ketone 134. Kerivoula picta bellissima Thomas 135. Kerivoula harduickit eigen Miller Genus Murina Gray i Key to Chinese Species of Mu urina . 136. Murina aurata Milne-Edwards 137. Murina cyclotis Dobson 138. Murina leucogaster Milne Bawards 139. Murina huttont rubella Thomas Family Molosside. Mastiff Bats Key to Genera of Chinese Molossidz Genus Cherephon Dobson 140. Cherephon plicatus (Bachssissidetacitiein) Genus Nyctinomus E. Geoffroy ; 141. Nyctinomus tentotis insignis Blyth 142. Nyctinomus teniotis cecata (Thomas) VI.—ORDER PRIMATES. LEMURS AND MONKEYS Key to the Genera of Chinese Primates Family Loriside. Slow Lemurs and Galagos Genus Nycticebus E. Geoffroy . 143. Nycticebus coucang cinereus Milue-Rawarils : Family Cercopithecidze. Baboons and Guenons Key to Genera of Chinese es a eae Genus Macaca Lacépéde Key to the Chinese Species of Macaca 144. Macaca assamensis (McClelland) . 145. Macaca mulatta (Zimmermann) Genus Lyssodes Gistel 146. Lyssodes speciosus ee (Milne-Edwards) 147. Lyssodes speciosus melli (Matschie) : Family Colobide. Langur Monkeys . Key to Genera of Chinese Colobidz Genus Pithecus Geoffroy and Cuvier . Key to Chinese Species of Pithecus : 148. Pithecus obscurus barbei (Blyth) . 149. Pithecus frangoist (Pousargues) 150. Pithecus nemeus (Linnzus) Genus Rhinopithecus Milne-Edwards . Key to the Chinese Species of Rhinopithecus . 151. Khinopithecus roxellane (Milne-Edwards) 152. Rhinopithecus bieti Milne-Edwards 153. Rhinopithecus brelichi Thomas CONTENTS CHAPTER Family Hylobatidz. Gibbons Genus Hylobates Iliger ; Key to the Chinese Species bf H ibatas : 154. Hylobates hoolock (Harlan) . 155. Hylobates concolor concolor (Harlan) VII—ORDER CARNIVORA. CARNIVORES Key to the Families of Chinese and Meuaotan eS Family Procyonide. Raccoons and their Kin . Genus Ailurus F. Cuvier . 156. Ailurus fulgens styani Theses Family Ailuropodide. The Giant Panda . Genus Ailuropoda Milne-Edwards 157. Azluropoda melanoleucus (David) Family Urside. Bears Key to Genera of Chinese fed icneokae Unsisie Genus Ursus Linnzeus : Key to the Chinese and Menaoltan Species of U rsus 158. Ursus pruinosus Blyth é 159. Ursus arctos lasiotus Gray . Genus Euarctos Gray 160. Euarctos thibetanus Sibeands (G. ee 161. Euarctos thibetanus melli (Matschie) Genus Helarctos Horsfield 162. Helarctos malayanus wardi Gydeeese Family Canide. Wolves, Dogs, and Foxes Key to the Genera of Chinese and Mongolian Gade Genus Canis Linnzeus : ; : 163. Canis lupus chanco rey Genus Nyctereutes Temminck : 164. Nyctereutes procyonoides roc jonoiaee (Gray) 165. Nyctereutes procyonoides orestes Thomas Genus Vulpes Oken : 166. Vulpes vulpes hoole ewichoe ; 167. Vulpes vulpes tschiliensis Matschie 168. Vulpes vulpes ?karagan (Erxleben) Genus Cynalopex Hamilton Smith : 169. Cynalopex corsac (Linnzus) Genus Cuon Hodgson 170. Cuon javanicus lepine Heude Family Mustelide. Martens, Weasels, Badgers, Otters Key to the Genera of Chinese and Mongolian Mustelidz Genus Charronia Gray 171. Charronia flavigula alkali (Boddaert) Genus Martes Pinel . Key to the Chinese and NMioneeiian Bpectes of M ares 172. Marites zibellina sajanensis Ognev XV PAGE 395 395 306 306 309 312 312 313 314 314 317 318 319 324 325 325 326 326 328 330 333 336 339 339 341 341 341 342 345 346 349 350 350 353 355 356 356 358 358 361 362 363 363 367 367 368 Xvi CONTENTS CHAPTER 173. Martes foina foina (Erxleben) Genus Mustela Linnzus Key to Chinese and Mivhigetinn Gpeniad a M astela 174. Mustela sibirica fontanierti (Milne-Edwards) 175. Mustela sibirica davidiana (Milne-Edwards) . 176. Mustela sibirica moupinensis (Milne-Edwards) 177. Mustela altaica altaica Pallas : : 178. Mustela altaica kathiah Hodgson . 179. Mustela rixosa pygmea (J. A. Allen) 180. Mustela russelliana Thomas 181. Mustela erminea mongolica Ognev 182. Mustela eversmanni tiarata Hollister Genus Vormela W. Blasius F 183. Vormela peregusna negans Miller Genus Helictis Gray : Key to the Chinese Species = val satis 184. Helictis moschata moschata Gray . 185. Helictis moschata ferreo-grisea Hilzheimer 186. Helictis taxilla sorella G. M. Allen Genus Meles Brisson : 187. Meles meles ecaiion Milne-Edwards Genus Arctonyx F. Cuvier : 188. Arctonyx collaris collaris F. Gavier : 189. Arctonyx collaris leucolemus (Milne-Bawards) Genus Lutra Brinnich Key to the Chinese Species of iien 190. Lutra lutra chinensis Gray . 191. Lutra lutra nair F. Cuvier . 192. Lutra tarayensis Hodgson Genus Micraonyx J. A. Allen 193. Micraonyx cinerea (iitivery Family Viverride. Civets and Mungooses Key to the Genera of Chinese Viverridze Genus Viverra Linnzeus 194. Viverra zibetha serial Svinhice Genus Viverricula Hodgson 195. Vuiverricula malaccensis Walnrats (Gann). 196. Viverricula malaccensis pallida (Gray) . Genus Paradoxurus F. Cuvier ‘ Key to Chinese Species of Paradoxurus 197. Paradoxurus hermaphroditus laotum Gyldenstoipe ? 198. Paradoxurus minor exitus Schwarz Genus Paguma Gray 199. Paguma larvata arcs (Eiamilton, Smith) 200. Paguma larvata intrudens Wroughton 201. Paguma larvata hainana Thomas Genus Herpestes Illiger CONTENTS CHAPTER Key to the Chinese Species of Herpestes . 202. Herpestes rubrifrons (J. A. sang 203. Herpestes urva (Hodgson) Family Felide. Cats Key to the Chinese and Mongotian F lids Genus Felis Linnzeus Subgenus Felis Linnzus 204. Felis chaus affinis Gray Subgenus Poliailurus Lonnberg 205. Felis bieti Milne-Edwards . °. Subgenus Trichelurus Satunin 206. Felis manul manul Pallas Subgenus Prionailurus Severtzov 207. Felis bengalensis bengalensis Kerr 208. Felis bengalensis chinensis Gray . Subgenus Profelis Severtzov . : 209. Felis temminckit tristis Waines ec weer Subgenus Neofelis Gray 210. Felis nebulosa Griffith Subgenus Panihera Oken : 211. Felis pardus fusca F. A. A. Meyer 212. Felis pardus fontanierit Milne-Edwards 213. Felis tigris amoyensis Hilzheimer 214. Felis tigris amurensis Dode Genus Lynx Kerr 215. Lynx lynx rliena (Blyth) VIII—ORDER PINNIPEDIA. SEALS, SEA-LIONS Family Otariide. Fur-seals and Sea-lions Genus Callorhinus Gray 216. Callorhinus curilensis Tastee and (once Family Phocide. Hair-seals Genus Phoca Linnzus 5 217. Phoca ?richardit (fea TX.—ORDER CETACEA. WHALES, DOLPHINS, PORPOISES Key to Families of Chinese Cetacea Suborder Odontoceti. Toothed Whales Family Iniide. River Dolphins Genus Lipotes Miller 218. Lipotes vexillifer Miller Family Delphinide. Dolphins and Porpoises Key to Genera of Chinese Delphinidze Genus Sotalia Gray . : 219. Sotalia chinensis (Osbeck) Genus Delphinus Linnzus Genus Tursiops Gervais Genus Neomeris Gray XVili CONTENTS 220. Neomeris phocenoides (G. Cuvier) Genus Grampus Gray 221. Grampus griseus (Cuvier) Genus Globicephala Lesson 222. Globicephala scammonit (Cope) Genus Orcinus Fitzinger . 223. Orcinus orca (Linnzus) Family Physeteride. Sperm Whales Genus Physeter Linnzeus ‘ 224. Physeter catodon Linnzus Suborder Mystacoceti. Whalebone Whales . Key to Whalebone Whales to be Expected in Chinese Seas X.—ORDER NOMARTHRA. SCALY ANTEATERS OR PANGOLINS Family Manide. Pangolins . Key to the Genera and Species of Ciiness Mamertiere Genus Manis Linnzus 225. Manis pentadactyla alae Sandievall 226. Manis pentadactyla pusilla J. A. Allen Genus Phatages Sundevall : 227. Phatages crassicaudata (Cay XI—ORDER RODENTIA. GNAWING MAMMALS ; Suborder Duplicidentata. Rabbits, Hares, and Migare shares Key to Families of Chinese and Mongolian Duplicidentata Family Ochotonide. Mouse-hares . Genus Ochotona Link ‘ Key to the Chinese and Mienecean Guess of Oona 228. Ochotona pallasti pallasii (Gray) 229. Ochotona pallasit pricei Thomas . 230. Ochotona hyperborea mantchurica Thomas 231. Ochotona alpina alpina (Pallas) 232. Ochotona alpina argentata A. B. Howell 233. Ochotona gloveri Thomas 234. Ochotona erythrotis (Buechner) . 235. Ochotona thibetana thibetana (Milne-Edwards) 236. Ochotona thibetana cansus Lyon 4 237. Ochotona thibetana huangensis (Matschie) 238. Ochotona thibetana sorella Thomas 239. Ochotona thibetana stevensi Osgood 240. Ochotona forresti Thomas 241. Ochotona roylei chinensis Thomas 242. Ochotona dauurica dauurica (Pallas) 243. Ochotona dauurica altaina Thomas 244. Ochotona dauurica bedfordi Thomas 245. Ochotona dauurica annectens Miller 246. Ochotona dauurica melanostoma (Buechner) Family Leporide. Hares and Rabbits CONTENTS CHAPTER Key to the Genera of Chinese and Mongolian Leporide . Genus Caprolagus Blyth 247. Caprolagus sinensis sinensis (Gray) 248. Caprolagus sinensis flaviventris G. M. Allen . 249. ?%Oryctolagus cuniculus (Linnzus) Genus Lepus Linnzeus Key to the Chinese and Mougoliati ismecies of Beas 250. Lepus europeus tolai Pallas 251. Lepus europeus swinhoet Thomas 252. Lepus europeus filchneri Matschie 253. Lepus europeus aurigineus Hollister 254. Lepus centrasiaticus Satunin 255. Lepus hainanus Swinhoe 256. Lepus otostolus otostolus Hodgson 257. Lepus oiostolus comus G. M. Allen 258. Lepus oiostolus grahami A. B. Howell SECTION III XII.—_BIBLIOGRAPHY, COMPLETE 579 x sdouaidovel iy Tah Cd: : VE Mo 7. paths sca ae enveal ae VN lh Wo 7 are Me ht + Vie ata ; ; ¢: ata +r One ahs WY La « dhiy le RU pid pe sates ayn oe owe ang P 8), ofertas A wwhonvi sal we (dy —_— Gedo. Al: mErain! Was at eben nal pitts be a, wie’ ay ht ery a } appa nae shoo) sipionne) er’ ™ y thers aap * TEXT FIGURES IGURE 1.—Distribution Map . Tupaia 1. T. belangert chinensis 2. T. belangert yunalis 3. T. belangeri modesta 2.—Distribution Map . Erinaceus 1. E. europeus dealbatus 2. E. europeus miodon Hemiechinus 3. H. dauuricus dauuricus 4. H. dauuricus alaschanicus 3.—Distribution Map . Uropsilus 1. U. soricipes Rhynchonax 2. R. andersoni andersoni 3. R. andersont nivatus 4. R. andersoni atronates Nasillus 5. N. gracilis 6. N. investigator 4.—Distribution Map . : Scaptochirus 1. S. moschatus moschatus 2. S. moschatus gilliesi Scapanulus 3. S. owenr 5.—Distribution Map . Mogera 1. M. latouchei 2. M. hainana 6.—Distribution Map . Sorex I. S. cylindricauda cylindricauda 2. S. cylindricauda wardi 3. S. cylindricauda gomphus PAGE 31 56 74 77 xxii TEXT FIGURES FIGURE PAGE 7.—Distribution Map . : : : : : : : : : : ~ 10g Chodsigoa 1. C. hypsibia hypsibia 2. C. hypsibia larvarum 3. C. hypsibia lamula 8.—Distribution Map . ; : : : : 3 : 3 : : - arg Blarinella 1. B. quadraticauda quadraticauda 2. B. quadraticauda griselda 3. B. quadraticauda wardi 9.—Distribution Map . : F é ; , : , : ; ‘ » 226 Crocidura 1. C. dracula dracula 2. C. dracula grisescens 10.—Distribution Map . : : , ; : : 3 ‘ , ; . \130 Crocidura 1. C. tlensis lar 2. C. ilensis shantungensis 3. C. ilensis pheopus 11.—Distribution Map . : : ‘ d , : é ; ; : 2 186 Anourosorex 1. A. squamipes 12.—Nose-leaves of Hipposideros armiger, male (left), and Hipposideros pratti, male (right) and female (center), natural size. After Dr. Wilfred H. Osgood (courtesy of Field Museum of Natural History) 194 13.—Distribution Map . ; ‘ : : : . : P : ‘ . 295 Pithecus 1. P. obscurus barber 2. P. frangoisi 3. P. nemeus 14.—Distribution Map . : : é : ; : : ; : ; 3 5209 Rhinopithecus 1. R. roxellane 2. R. bieti 3. R. brelichi 15.—Distribution Map . j ; : : ; ¢ j : : : aegis Ailurus 1. A. fulgens styani 16.—Distribution Map . é ‘ ; ‘ ; : : : : ; sot Ailuropoda 1. A. melanoleucus TEXT FIGURES FIGURE 17.—Distribution Map . Mustela 1. M. sibirica fontanierit 2. M. sibirica davidiana 3. M. sibirica moupinensis 18.—Distribution Map . Viverra 1. V. zibetha ashtont 19.—Distribution Map . Paguma 1. P. larvata larvata 2. P. larvata intrudens 3. P. larvata hainana 20.—Distribution Map . Manis 1. M. pentadactyla dalmanni 2. M. pentadactyla pusilla 21.—Distribution Map . Ochotona 1. O. thibetana thibetana O. thibetana cansus O. thibetana huangensts 4. O. thibetana sorella O. thibetana stevenst 22.—Distribution Map . Ochotona 1. O. dauurica dauurica 2. O. dauurica annectens 3. O. dauurica bedfordi 4. O. dauurica altaina 23.—Distribution Map . Lepus 1. L. europeus tolat 2. L. europeus swinhoei 3. L. europeus filchneri 4. L. europeus aurigineus XXili PAGE 370 424 434 518 541 562 TT nigey aay Sh Oeigr apiindine Ra We ins Lowa’ \debndilinagl dae ak pil ts) te Ae ; 7 Alt ik weld 4) eis ede ali) of in ieee 1 yes? Sy chy cite he PLATES PLATE I.—Upper figure: The southern edge of the mixed forest zone, forty miles northeast of Urga, Outer Mongolia. Lower figure: Outpost larches at the edge of the forest zone, forty miles northeast of Urga, Outer Mongolia : 2 : ¥ : : : : II.—Sainnoin Khan, Outer Mongolia. Hills with patches of timber II].—Upper figure: Pass to the Mongolian Plateau at Kalgan. Lower figure: Mongolian Plateau above Kalgan, showing Chinese cultivation IV.—The Gobi. Looking out over the valley of oe Nor, from just below Baga Bogdo, Outer Mongolia : ; : : ? : V.—Yenchingkou near Wanhsien, eastern Szechwan. Site of bat caves; the lime- stone ridge in the background is the haunt of Edwards’s Giant Rat (Rattus edwardsi gigas) A : : : : : : VI.—Upper figure: Dr. Andrews’s camp and the Snow Mountain, western Yunnan, 12,000 feet. Lower figure: The Mekong valley, western Yunnan VII.—Upper figure: A tame Long-eared Hedgehog (Hemiechinus dauuricus alaschan- icus) at Tsagan Nor, in the Gobi, about to fold up. Lower figure: The hay-pile of a Pallas’s Mouse-hare asi ati Fed at Artsa Bogdo, in the Gobi. : ‘ : ; VIII.—Upper figure: Hoolock Gibbon (Hylobates hoolock), female, killed near the Burma border, Yunnan. Profile view. Lower figure: Front view of the same . IX.—Upper figure: A live Civet (Viverra ztbetha ashtont). Yunnan. Note the dorsal crest of erect hair. Lower figure: Head of a Civet (Viverra zibetha ashtoni). Mucheng, Yunnan XXV FACING PAGE 10 I2 16 18 46 306 420 THE MAMMALS OF CHINA AND MONGOLIA SECTION I GENERAL INTRODUCTION SECTION I GENERAL INTRODUCTION CHAPTER I—COLLECTORS OF CHINESE AND MONGOLIAN MAMMALS CHAPTER II—FAUNAL AREAS OF CHINA AND MONGOLIA - - - CHAPTER II—FAUNAL RELATIONS OF ASIA WITH NORTH AMERICA CHAPTER I COLLECTORS OF CHINESE AND MONGOLIAN MAMMALS THE first knowledge of the mammals of China and Mongolia by Europeans probably dates from the days of the Venetian traveler Marco Polo, who in the thirteenth century journeyed overland to the court of Kublai Khan and brought back parts of the Musk Deer (Moschus) which he seems to have met with on the western borders of those countries. Brief mention of other Chinese mam- mals is found in a few early works of travel, such as Du Halde’s ‘‘Description de la Chine,”’ 1735, which notices apparently for the first time the Mongolian Gazelle and the ‘‘Great Black Ape’’ or gibbon of Hainan; Peter Osbeck’s “Voyage to China and the East Indies’ (English edition, 1771) contains a “Faunula et Flora Sinensis,’ from which Thunberg (1823) copies a list of fifteen species of mammals, including “‘Simia nemea,” ‘‘Sus sinensis,” ‘‘Cervus alces,’”’ and ‘‘Cervus elaphus.”’ Pallas, too, who traveled in eastern Siberia in the latter half of the eighteenth century, described a number of species from near the borders of northern Mongolia that range into that country. The first important collections from China, however, did not reach Europe till much later. Among the earliest of these was perhaps that of John R. Reeves, who while stationed at Canton, sent back to the British Museum a number of mammals, of which J. E. Gray figured and named in his ‘Illustrations of Indian Zoology,” Canis procyonoides, Viverra pallida, Rhizomys sinensis, and Lepus sinensis. This work bears the date 1830-34, but some of the plates are lettered 1829. Gray soon after (1837) described Felis chinensis and Lutra chinensis, and Ogilby, in 1838, Cervus reevesit from the same source. The botanist R. Fortune (1853) who made two visits to China before the middle of the last century (1843, 1844), also brought back a few mammals to the British Museum, which subsequently proved new or valuable. Beginning about 1858, Robert Swinhoe, while on British consular service in China, took a very active interest in the zodlogy of that country and from time to time sent notes or specimens to the Zodlogical Society of London or to the British Museum. His work took him to Peiping, Chefoo, and Kalgan in North China, as well as to the island of Formosa, while later he was for some 3 4 THE MAMMALS OF CHINA AND MONGOLIA time in residence at Amoy and again at Ningpo in South China, and also made collections in Hainan. His published communications on the mammals of eastern China and Formosa extend over a period of some sixteen years to 1874, and include descriptions of various new species. ’ By far the most outstanding figure among collectors of Chinese mammals is the Jesuit missionary, Pere Armand David. He arrived in Peiping in July, 1862, and at once seems to have begun actively to make investigations and collections of the fauna. In 1864 he spent several months at Jehol, 200 kilometers north of Peiping, and in 1866-68 undertook a journey into what was then Chinese Mongolia, residing for some time at Saratsi, in northwestern Shansi. The collections he made, together with various specimens added by M. Fontanier, the French Consul at Peiping, were sent to the Paris Muscum, where they were studied by Alphonse Milne-Edwards, who described various new species among them. David’s most remarkable journey, however, was made about 1870 into the principality of Muping in Szechwan, then regarded as part of eastern Tibet. He was the first naturalist to reach these highlands, and in the course of nearly a year’s work made an extraord nary collection of mammals, large and small, including many wholly new and unknown types as well as other species previously undescribed. These included the new in- sectivore genera Anourosorex, Uropsilus, Nectogale, and Scaptonyx, and the Golden Monkey, Rhinopithecus. The extraordinary zeal of this indefatigable collector resulted in bringing together at Paris a representation of Chinese mammals excelling any previously made. In August, 1871, these collections were on special view to the public at the Muséum d’Histoire Naturelle at Paris and were made by Milne-Edwards the subject of a magnificent volume with an atlas of colored plates—the well known ‘‘Recherches” (1868-74). David himself published several brief reports on his collecting journeys (see David, A., 1867-71, 1872-73, 1872, 1872a). He had many narrow escapes,—several attempts were made to poison him,—but he survived all dangers in a miraculous way. In the course of his work he was accompanied by a faithful Chinese servant, “Yellow Tom,” whose memory Milne-Edwards essayed to perpetuate by naming in his honor a supposed new rat, “‘Mus huang-thome’’! Later David visited Chekiang and Kiangsi Provinces, but collected mainly in other branches than mammals. In 1873 he spent three and a half months in the mountains of Shensi, not far from Sianfu, making additional collections, and in the same year journeyed to Kuatun in the mountains of northwestern Fukien, where again he obtained various species of exceptional interest, including the first specimens of the curious genus Typhlomys, described by Milne-Edwards, and still unknown from other parts of China. Here, too, he collected the first examples of the microtine later named by Thomas as Microtus melanogaster colurnus, and notes on the label of one (now in the British Museum), ‘“‘rubigin- COLLECTORS OF CHINESE AND MONGOLIAN MAMMALS 5 osus n. sp.,’’ but for some reason the name never was published. The locality became a famous one for naturalists: both C. B. Rickett and J. D. La Touche had collectors working there for several years (see Thomas, O., 1898, p. 769), and later it was visited by Mr. Clifford H. Pope, who secured an excellent collection for the American Museum in the course of the Asiatic work. Mention should be made of the services of the French missionaries to science, for David was not alone in his interest in the mammals of China. Perhaps the foremost is P. M. Heude who in the course of a number of years from 1885-1901 published in his ‘‘Mémoires concernant l’Histoire Naturelle de l’Empire Chinois,” a series of papers on the ungulates of China as well as on the bears and other species. His intense interest in the details of tooth structure led him to bestow technical names freely on many minor and in- dividual variations in the skulls he brought together, resulting unfortunately in greatly burdening the nomenclature of many of these animals. His Sikawei Museum (now known as the Heude Museum) at Shanghai contains most of his type specimens, which have since been studied by Sowerby, while a general account of his life and work has been published by Courtois (1906). In Tatsienlu, Szechwan, several of the French missionaries have from time to time sent valuable specimens to the Paris Museum. Among others, the name of Monseigneur Biet is linked with his discovery of a striking species of Rhino- pithecus and that of R. P. Dejean with the Yunnan Sambar Deer. For many years following David’s time, western China was very little visited by naturalists. In 1907-08 the late Walter R. Zappey accompanied the veteran traveler E. H. Wilson into Hupeh and Szechwan, making extensive collections of birds and mammals for the Museum of Comparative Zodlogy, and eventually reached Tatsienlu and the extreme edge of the Tibetan plateau at Ramala Pass. Our first important knowledge of the mammals of extreme southwestern China is due to Dr. John Anderson (1879), who as naturalist and medical officer on two British expeditions from India to the borders of western Yunnan in 1867-68 and 1875, made collections in various branches of zodlogy, later embodying the results of his work in a sumptuous volume with atlas, giving descriptions and colored figures of many species. Most of these were pre- viously known in part from Indian representatives. Toward the close of the nineteenth century, Prince Henri d’Orléans under- took a journey from Tongking, across Yunnan to northern India, collecting a few mammals on the way, but not until comparatively recent years has system- atic collecting been done in this province. The explorations of F. Kingdon Ward in 1911, 1913-14, 1921-22, among the great snow peaks of southwestern Yunnan, and the subsequent work of George Forrest in 1922, resulted in many interesting discoveries, which with lists of the species obtained have been 6 THE MAMMALS OF CHINA AND MONGOLIA published by Oldfield Thomas (1922b, 1923). Meanwhile in 1916-17, Dr. Roy C. Andrews undertook the first of his Chinese expeditions, and accompanied by Mr. Edmund Heller, a skilled field naturalist, crossed Yunnan to Burma, returning with one of the largest and best prepared collections of large and small mammals ever made in this part of China. In the prosecution of this work in subsequent years, Dr. Andrews and his associates have done intensive collecting in various parts of China: eastern Szechwan, northern Hopei, northern Shansi, in the Taipai Shan of Shensi, and in Fukien. Mr. Clifford H. Pope has continued the work in the last-named area, visiting also the island of Hainan where large collections were made, To the late Oldfield Thomas, Keeper of Mammals in the British Museum, must be given high credit for the interest he aroused among his countrymen stationed in various parts of China or traveling through it, so that for many years correspondents from various parts of this vast area continued to send to the British Museum from time to time specimens of large and small mam- mals. In this way Thomas built up a considerable representation of Chinese species at London, until in 1904 His Grace the Duke of Bedford undertook to finance the work of a trained collector for more systematic exploration. The services of Malcolm P. Anderson were secured and he was sent first to the Japanese islands and Korea, then to Shantung, the Mongolian plateau, Hopei, Shansi, Shensi, Kansu, Szechwan, and Yunnan. The results were briefly re- ported upon by Thomas in a series of papers from 1906 to I912 and added greatly to a general knowledge of Chinese mammals. In some of this work Anderson was accompanied by Dr. J. A. C. Smith and Arthur de Carle Sowerby, who helped with the collection and preparation of specimens. The latter has himself done considerable collecting for both the British Museum and the U.S. National Museum, and through various popular books and more technical papers as well as in general articles in the China Journal (which he helped to found) has contributed greatly to furthering an interest in Chinese natural history, culminating in his recent successful attempt to establish a public natural history museum at Shanghai. Several Chinese zodlogists are now taking an active interest in the fauna of their native land, among whom mention should be made of Chausu Mc- Amicus Shih, who has made several collecting journeys into the less-known provinces of southern China,—Hunan, Kwangsi and Kwangtung. Important collections were made in the last-named region by R. Mell, an Austrian botanist and naturalist, in the years from 1916-21. These were in part reported on by Matschie in 1922. Another expedition into China led by Stétzner, in 1916, was accompanied as naturalist by Dr. Hugo Weigold whose collections eventu- ally reached the Dresden Museum and were reported on by Jacobi (1922). Weigold collected in northern Hopei, then proceeded up the Yangtze to the PLATE 1 The southern edge of the mixed forest zone, forty miles northeast of Urga, Outer Mongolia Outpost larches at the edge of the forest zone, forty miles northeast of Urga, Outer Mongolia COLLECTORS OF CHINESE AND MONGOLIAN MAMMALS ‘f Wa Shan region of Szechwan, previously visited by Wilson and Zappey. He finally reached Batang and northern Szechwan at Sungpan. China has thus been traversed by many collectors, chiefly from east to west, less often in the opposite direction. It is impracticable to mention here the names of all those who have added to the knowledge of Chinese mammals, but it may be pointed out that there are still large areas from which little is known, particularly in the southern provinces, although the western highlands and the northeastern parts of the country may now be thought of as fairly well worked. The excellent report of Dr. W. H. Osgood on the mammals obtained for the Field Museum by the Kelley-Roosevelts and Delacour Asiatic Expedi- tions, chiefly in Tongking, indicates that many species are to be expected along the extreme southern border of China, which hitherto have been taken but a short distance only across that line. Various Russian explorers have likewise done much to advance our knowl- edge of the mammals of Mongolia. Pallas in the latter half of the eighteenth century traversed Siberia, but his journey was mainly to the north of Mongolia. During the years 1855-59, Gustav Radde (1862) traveled extensively in Trans- baikalia, several times crossing the border into northeastern Mongolia, making collections and observations on the fauna. His volume on the Mammalia is of importance as giving much data on the habits and distribution of many species common to both countries, including several forms which he named as new. His specimens went to the Zodlogical Museum of the Academy of Sciences at St. Petersburg (now Leningrad). The four expeditions of Nikolai M. Przewalski into central Asia resulted in the discovery of many new mammals, which likewise were given to the Zo6- logical Museum of the Academy. After his return from the last of these journeys, a special exhibition of these was held in the new wing of the Museum and a catalogue of the collection (in Russian) was printed, which included 702 mammals, 5,010 birds and 1,199 reptiles and amphibians. His first expedition, from 1871-73, extended across northern Kansu and Mongolia to Kalgan; the second, 1876-77, was chiefly concerned with an exploration of the Tien Shan area; the third expedition, 1879-80, slightly more to the south, was to the Nan Shan, Kansu, and southern Alashan; while the fourth, in 1883-85, was to the country between the plateau of northeastern Tibet and the Tarim basin. The zoological results of these explorations were in part published by the Academy, but of the mammals only the rodents were in part completed by E. Buechner, with a later monograph by Zalensky on Przewalski’s Horse, discovered in the course of the third journey. The text, in Russian and German, is magnificently illustrated by a number of colored plates. Accounts of Przewalski’s explora- tions were published by him in Russian with later German and English editions - (see Przewalski, N. M., 1884; also the English edition translated by Morgan). 8 THE MAMMALS OF CHINA AND MONGOLIA In 1884-87, an expedition to Kansu and northwestern Szechwan was undertaken by G. Potanin and M. Berezovski for the Russian Geographic Society. The former visited the Ordos Desert and northern Shansi, touching also Kansu and Mongolia, while the latter confined his efforts mainly to the region about Ssigu and the adjacent parts of Kansu and Szechwan. Except for the specimens obtained by Przewalski, these collections were the first to be made in Kansu and included among other rarities, the Golden Monkey and the Giant Panda, apparently secured through native hunters. Buechner, who reported on the mammals, regretted that most of them were given to the museum at Tomsk rather than to the larger institution at Leningrad. Other Russian explorers, including P. K. Kozlov in 1909 and 1924, as well as the brothers Grum-Grzimailo, who made a journey to western China shortly before 1907, have collected a number of interesting species in extreme western Mon- golia, while in 1910-11 Douglas Carruthers, accompanied by J. H. Miller and M. P. Price, made a collecting trip for the British Museum, starting work at Minussinsk, Siberia, and working southeastward through the Syansk Moun- tains, crossed northwestern Mongolia, then proceeded westward through the Tannu Ola to the Great Altai and Barlik Mountains, northwestern Dzungaria, and along the Tien Shan to the Hami Mountains, and thence back to the Muzart valley, Kuldja, and Yarkand to India. Of the extended work carried on by Dr. Roy C. Andrews in eastern and central Mongolia, resulting in the collection of fine series of the desert species, and in northern Mongolia where the edge of the northern forest brings in an entirely different set of species, a full account is given in the narrative forming the first volume of the present series. Especial credit is due Dr. Walter Granger, who in the intervals between his many other duties, collected and prepared a large part of the splendid series of Mongolian and also Chinese small mammals in the course of several years’ work; also to Mr. Clifford H. Pope, whose intensive work in Hainan and Fukien resulted in large collections of small mammals from these important localities. Dr. Andrews himself worked in Fukien and in Yunnan with the aid of Mr. Edmund Heller in his earlier expedition and brought back a magnificent collection of the smaller species from these areas as well as a representative series of the larger mammals. Jaquit} Jo sayojzed YIM s][ty “eyjosuojy Jojng ‘ueyy ulouuTeS CHAPTER II FAUNAL AREAS OF CHINA AND MONGOLIA THE vast area included within the political boundaries of China and Mongolia is roughly some 2,000 miles square and constitutes about one-third of the continent of Asia, between approximately 22 and 52 degrees of north latitude and 90 and 120 degrees of east longitude. A wide range of climatic conditions is thus presented, correlated not only with latitudinal but also with topographical and other factors. Proceeding from north to south, the mam- malian fauna shows many striking contrasts between associations charac- teristic of one and another of these areas. On the other hand, some are wide- ranging species, such as tiger and black bear, and occur with but slight modifications under varying climatic conditions. The present distribution of the fauna is to be looked upon as the end result of many millions of years of changes, both climatic and geographic, some of them slow and by imper- ceptible degrees, others more rapid and spectacular. The composition of the fauna at any time is in part a result of the history of the area where it is found and in part a result of the food and other habitat preferences of the component species. If the local conditions undergo marked or gradual change, the species have three courses open: (1) they may gradually become adapted to the changes (evolution) ; (2) they may move off to other areas where the preferred conditions still continue (emigration); or (3) if appropriate response is impossible, they may remain and slowly die out as the changing conditions become more and more unfavorable (extinction). The history of the mammalian fauna of eastern Asia during the long past may be in part worked out through the study of its fossils and must be left largely for the paleontologist. The present fauna of the area shows the fol- lowing seven chief divisions: (1) the northern transcontinental forest which borders northern Mongolia, and to the east of the Khingan Mountains ex- tends around the eastern edge of the Gobi to Manchuria and northeastern China; (2) the intervening Gobi, including not only the grasslands that form a long east-west loop around the northern and eastern edges of Mongolia, but also the more arid and strictly desert area of the central Gobi; (3) the north- eastern part of China, including southern Hopei, Shantung, northern Shansi and Shensi, and Kansu, southward approximately to the borders of the Yangtze 9 10 THE MAMMALS OF CHINA AND MONGOLIA basin or roughly latitude 34°; (4) South China, including the eastern part of the country from the Yangtze basin to the southern border, characterized by many species that prefer a milder climate, many of them, as one proceeds farther south, of chiefly tropical and subtropical distribution; (5) the western highlands, which begin at about the longitude of western Hupeh or eastern Szechwan, and include the latter province from the borders of Kansu and Shensi south through Yunnan and probably parts of Kweichow; (6) the subtropical border of extreme southern China; and finally (7) the edge of the Tibetan plateau along its extreme eastern border, where it meets the western highlands of Szechwan. ! 1. The Northern Forest Fauna:—The mixed evergreen and hardwood forest of north temperate latitudes is practically continuous across northern Europe and Asia from the limit of tree growth southward. In the east the southward limits are somewhat restricted by the arid conditions of central Asia, so that it becomes more or less broken. The edge of this forest follows along the northern edge of Mongolia, giving way rather abruptly to the open grasslands and the more barren stretches of the Gobi; but to the eastward of the north-south Khingan Range it continues farther down across Manchuria to northern and central Hopei. In a general way it corresponds to the boreal forest of the Canadian Zone in North America. Prominent among the northern trees composing this mixed forest are larch, fir and pine as well as oaks and birches. It carries with it a rather characteristic series of mammals, many of them with closely similar represen- tatives in northern and central Europe, and in diminishing number, in northern North America. Among these, the insectivores are few, chiefly two species of small shrews (Sorex araneus borealis and S. buxtoni) and probably Neomys (as yet not known from Mongolia but recorded slightly north of its borders), as well as a race of the European hedgehog (Erinaceus); bats are represented by several species of the family Vespertilionide, belonging to the genera Myotis, Eptesicus, Plecotus, and Vespertilio, closely resembling corresponding European forms; the carnivores include the Brown Bear, Red Fox, the wide- ranging Wolf, the Ermine, and Pygmy Weasel (Mustela rixosa pygmea), Badger (Meles), Otter (Lutra lutra chinensis) and Lynx (Lynx lynx isabellina). These are associated with various northern rodents, including varieties of the European Tufted-eared Squirrel (Sciurus vulgaris) and the small northern flying squirrel (Pteromys volans buechneri) in the mixed forest, and in the more open parts the Eversmann’s Spermophile (Citellus eversmanni jacutensis). Beaver just reach the edge of Mongolia in the southern extension of the forest, but doubtless were formerly more widespread. In suitable situations the Mouse-hare (Ochotona hyperborea mantchurica) is found southward into this part of Mongolia, but the genus, although represented in western Europe PLATE III Pass to the Mongolian Plateau at Kalgan Mongolian Plateau above Kalgan, showing Chinese cultivation ¥ , a. Ls #3 a sy * { : i ° . , / . : aks i ‘ w- ; . a = = ny ‘ie pa ~*~ Me. - 2 1. ee 4 Dow 4 *.: om |) . Sindy 7 Pa: Rls a tah Wy he die FAUNAL AREAS OF CHINA AND MONGOLIA II in glacial times, has since died out in that continent. Other characteristic rodents are: the Red-backed Lemming (Myopus), the Field Vole (Microtus mongolicus representing the European M. arvalis), the red-backed mice (Cle- thrionomys), of two and perhaps three species corresponding to those of northern Europe, while of larger ungulates, the Pig, Elk or ‘‘Moose,”’ the Red Deer, and the Roe Deer are represented by forms differing but very slightly from their European relatives. Most of these species continue across to the Pacific coast following the northern forest and skirting the edge of northern Mon- golia. To the southward the limit for most of them is fixed by the intervening Gobi which extends eastward from the arid wastes of central Asia, and forms an effectual barrier. A few, however, such as one of the red-backed mice, flying squirrel, roe deer, appear again in northern China on the other side of the desert, having perhaps skirted its eastern end, following the forested parts east of the Khingan Range into Manchuria and northern Hopei. To the south and west of Peiping or thereabouts, the conditions are more arid and the forest broken, perhaps due in part to human destruction, making such a passage more difficult. Here, too, the northern-forest fauna appears to reach about its natural limit, and meets the species of more southern range that find their bounds in these latitudes. 2. The Gobi:—The fact that most of the boreal forms of mammals characteristic of the Mongolian forests do not reappear on the south side of the Gobi and are not found as relict colonies at high altitudes farther south, seems evidence that these species, if driven south by the cold of the glacial period, were unable to cross the intervening desert, which, therefore, must have existed for a very long period, forming an effective barrier to the forest- living types except those that were able to circle the eastern end. The Gobi forms a long tongue of desert extending eastward from the central Asiatic deserts, bringing in with it a series of mammals, particularly rodents, that have become adapted for desert life in a remarkable degree, implying a very long period of evolution in that type of environment. This arid area is more or less continuous from northern Africa, but most of the genera represented are of Asiatic affinities. The general east-west trend of the higher mountain chains enclosing the western end of the Gobi is perhaps a factor in permitting this extension of the desert by interposing a less effective barrier against the prevailing winds to extract their moisture. In the narrative volume of these reports, Dr. Roy C. Andrews and Dr. Walter Granger point out that the Gobi is not equally desert throughout, but that a broad belt of grasslands forms a wide, flat border around its northern, eastern and southeastern ends, while the central portions are much more arid or truly desert. Thus in crossing from southeast to northwest—Kalgan to Urga—one first passes through the southern grassland belt, then traverses the I2 THE MAMMALS OF CHINA AND MONGOLIA dry desert country of the central Gobi to Tuerin, where one emerges upon the northern grasslands, over which the caravan track proceeds until the forest edge is reached in northern Mongolia. These three divisions differ in their fauna. Characteristic of the southern grasslands are the Mongolian Gazelle (Prodorcas gutturosa) and the small spermophile (Citellus dauuricus mongolicus), which extends its range southward over parts of northern China. In the northern grasslands the spermophile is replaced by a larger species (C. pallidi- cauda). It is here, also, that the colonies of Bobac Marmot (Marmota bobak sibirica) are found (related to the form of central Europe), while preying upon them and the spermophiles is the Masked Polecat (Mustela eversmanni tiarata), corresponding in this relation to the Black-footed Ferret of North America, that frequents the colonies of “‘prairie dogs.’’ Brandt’s Field Vole (Microtus brandti) is abundant as well as M. poliakoffi, while a third species (M. tian- shanicus angustus) is rarer. Both the latter are members of the subgenus Stenocranius. The small hamster (Cricetiscus songarus campbelli) is a charac- teristic grassland species of both the southern and the northern portions. Extreme northwestern Mongolia is inhabited by several species that seem to be absent from the rest of the country. Among these are Przewalski’s Horse, which just reaches this area from farther west, and the vole (Microtus agrestis mongol) representing the group so widespread in North America, with an additional postero-internal loop on the middle upper molar. The large jerboa (Allactaga) is found over much of Mongolia, but the smaller genus (Alactagulus) has thus far been taken in the far northwestern part of the coun- try only. The Mouse-hare (Ochotona alpina) also reaches this end of Mon- golia from the west. Notable also is the extension of the Mountain Sheep (Ovis ammon darvini) and the Siberian Ibex (Capra sibirica) into the central Gobi, following the northwest-southeast trend of the mountain chains into the central part of the desert. It was perhaps by following these ranges that the sheep crossed to northern Shansi, and perhaps in former times extended into Shantung, but the ibex apparently did not penetrate so far, although both are found together in the western Gobi. The typical desert country occupies the central part of Mongolia and is largely barren and dry, as well described by Dr. Roy C. Andrews in the first volume of this series. Water is scarce, forage is sparse and cover hardly avail- able over great stretches. Yet a number of mammals are more or less confined to this desert. Of larger species the wild ass and the gazelle (Gazella sub- gutturosa hillieriana) are confined to this environment, the former sometimes gathering into considerable herds, but the latter consorting in small groups. Wolves are rare, and two species of fox, one a representative of the widespread Red Fox, the other resembling the American Kit Fox, are present in small numbers. One or two types of desert-living cats (Felis manul and probably E IV PLATE BY]OSUOTY JojnO ‘opsog vseg Mojaq qsnf wouy ‘ION uPSesy, Jo Kaljea 94} I9AO NO ZuryooT “Iqoy ayy, ere LVIPE EPO FAUNAL AREAS OF CHINA AND MONGOLIA 13 Felis bieti= pallida) must find at times abundant food in the shape of small rodents, such as the various species of jerboa and Przewalski’s Vole (Lagurus przewalskii). A burrowing adaptation of the microtine type is found in the genus Ellobius, of which forms closely allied to those of western and central Asia have penetrated well into this desert, where they have sought shelter through an underground existence. The small hamsters, Phodopus, have sim- ilarly developed subterranean habits, burrowing in the sand hills, and in corre- lation with these habits have nearly lost the tail and are extremely pallid in color. Most interesting of all are the various genera of leaping rodents that have developed the saltatorial habit in correlation with desert living, and have become long of foot and limb, with long tails and, usually, much enlarged audital bulle, sometimes associated with long external ears as in Allactaga and Euchoreutes, or with short ears as in Dipus, Salpingotus and Cardiocranius; all are characteristic of sandy deserts, developing stiff hairy pads under the toes for giving a secure footing on the shifting sand. Hares (Lepus europeus tolai) are common over parts of the Gobi as well as in North China, with the palest of the races in the desert. They often haunt old camp sites along the caravan track. Pallas’s Mouse-hare (Ochotona pallasii) lives among rocks in the desert, as does also the Cliff Mouse (Alticola) and a desert race of a northern bat (Eptesicus nilssonii gobiensis) while the Dauurian Mouse-hare makes burrows out in open grassy places. Hedgehogs of the long-eared group (Hemiechinus), very pale in color, take the place of the short-eared Erinaceus of the northern fauna, while shrews are almost lacking, although a single speci- men of a pale form of Crocidura (C. ilensis lar) was taken by the expedition at Tsagan Nor. 3. North China:—South of the Gobi and its extension the Ordos Desert, is a wide stretch of country comprising the northern parts of Kansu, Shensi, and Shansi, and the southern portion of Hopei, over which the conditions are semiarid, with occasional mountain ranges, partly wooded. Malcolm P. Anderson (in Thomas, 1909, p. 964) writes: ‘‘The provinces of Shan-si and Shen-si are quite different in character. The former may be briefly described as a mountainous country with occasional large upland plains. Some peaks in Shan-si rise above 10,000 ft., and are massive rocky mountains with only a comparatively thin coating of loess soil. Where the loess figures mostly is in the plains, of which that of Tai-Yuen-Fu, that of Ta-Tung-fu, and that of Hsiu-clou are the best examples. The streams of Shan-si flow only in the rainy season, with the exception of the larger rivers. Northern Shen-si, on the other hand, is a region of loess hills of almost uniform height; the skyline of Shen-si, seen from the mountains of its eastern neighbour, is a straight line declining very gradually as it passes from north to south.””. Of the mammals of this area, some are of northern transcontinental types, which may have 14 THE MAMMALS OF CHINA AND MONGOLIA reached this area either by extending their range around the eastern edge of the Gobi, or in case of species with a slightly more southern tendency, by having been able to spread widely east and west across the less barren parts of central Asia, in recent or earlier times. An interesting example is Clethrionomys rufocanus shanseius, the only one of the red-backed mice yet found in the area south of the Gobi, occurring in the spruce forests of northern Shansi. Another example is the northern flying squirrel (Pteromys volans buechneri), which probably occurs in northern Mon- golia, but hitherto is recorded only in Kansu and the forests of southern Shensi of the present area. The chipmunks (Eutamias) have a rather similar dis- tribution, for they are found in the open forest of northern Mongolia, around the eastern parts of the country, to Hopei, Shansi, Shensi, and Kansu. In a general way, these and a number of other species are limited in their southward tange by the east-west mountain ranges—the Min Shan in southern Kansu, and the Tsingling of southern Shensi and Shansi—and the northern borders of the Yangtze basin, with slightly differing limits in accordance with their particular requirements. Thus among the Insectivora, the mole genus Scaptochirus is characteristic of northeastern China; the hedgehog of this region is a close relative of the Eurasian species Erinaceus europeus, while the shrews of both red-toothed (Sorex) and white-toothed (Crocidura) types are rare, Sorex sinalis being known from Kansu, the small Crocidura ilensis shantungensis representing the latter genus in the northeast. The hamsters, including the larger Cricetulus triton group and the smaller C. barabensis races, do not extend south of these limits. Gerbils of the two species Meriones unguiculatus and M. psammophilus are both common in North China, and are partial to semi-desert. The mole-rats of two types are characteristic northern animals, the more northern Myospalax myospalax psilurus extending southward from extreme northeastern Mongolia into Hopei, the more southern M. fontanierit ranging south into Hupeh and northern Szechwan at high alti- tudes in slightly differing races. Other northern species, whose range is transcontinental in the north temperate zone and extends to northern China, include the Roebuck (Capreolus), the Red Deer (Cervus elaphus kansuensis), the Badger (Meles meles leptorynchus), the Otter (Lutra lutra chinensis), the Wolf (Canis lupus chanco), the Wild Boar (Sus scrofa subsp.) and various rodents such as voles of one or two species (e.g., Microtus ratticeps flaviventris), the Brown Rat (Rattus norvegicus socer), the Harvest Mouse (Micromys), and the small House Mouse (Mus bactrianus races). Other species of North China have a more restricted range, at least in modern times, and are peculiar to this part of China, though no doubt with once a wider range. Such are the River Deer (Hydropotes), confined in China to the Yangtze River bottoms but turning up once more in Korea; David’s Deer (Elaphurus) the antlers of which FAUNAL AREAS OF CHINA AND MONGOLIA 15 are associated with ancient human culture in Honan; and the Groove-toothed Flying Squirrel of the forests of Hopei. On the other hand, sundry species whose distribution is essentially southern here find their northern limit, as the Striped Tree Squirrel (Tamiops s. vestitus), the Large-toothed Flying Squirrel (Trogopterus), the common Yellow-bellied Rat (Rattus confucianus chihliensis and other races) and R. nitidus humiliatus. This meeting of the northern and the southern faunas is paralleled to some degree in North America. 4. South China:—In a general way the parallel of about 34° north marks the southern limit of the North China fauna, while south of that, from about the northern edge of the Yangtze drainage and the Tsingling Range, southward, this gives place to a more numerous assemblage of species whose distribution is largely southern, and corresponds in a general way to the austral fauna of North America. Many of the species are wide-ranging across the whole of southern China, while others are more strictly confined either to the lower country of the eastern parts or to the highlands of the western portion. One may thus distinguish the South China fauna and that of the western highlands. Of those species that are wide-ranging longitudi- nally and altitudinally, it is usual to find that the highlanders represent a more or less differentiated race from those of the lower coastal districts of southeastern China. There are thus subspecies of various bats in the lower country of South China differing subspecifically from their representatives in the west, as Rhinolophus blythi calidus and R. b. szechwanus, R. episcopus caldwelli and R.e. episcopus of east and west respectively, Myotis chinensis and M. c. luctuosus, and among rodents many others might be mentioned of the genera Dremomys, Callosciurus, Sciurotamias, Petaurista, Eothenomys, Rhizomys, Rattus, Mus, among the Carnivora the widespread weasels Mustela sibirica and M. alpina, and among ungulates the genera Capricornis and Nemorhedus, in all of which examples are found of subspecific differentiation into races characteristic of lowland and highland of eastern and western China. A number of other species or genera are not found in the western uplands but are confined to the southeastern lowlands. Such are the moles of the genus Mogera of Fukien and Hainan, the ferret-badgers (Helictis), the Crab- eating Mungoose (Herpestes urva), the Clouded Leopard (Felis nebulosa), the Scaly Anteater (Manis pentadactyla dalmanni), the distribution of which is doubtless largely governed by that of the termites on which it feeds, the curious Tufted-tailed Dormouse (Typhlomys), the Red-cheeked Squirrel (Dremomys rufigenis pyrrhomerus), Reeves’s Muntjac and the Tufted Muntjac (Mun- tuacus reevesit and M. crinifrons). Perhaps the common Rhesus Macaque (Macaca mulatta) might be included here, although it ranges across extreme southern China, and even reaches considerable altitudes in parts of Szechwan. 16 THE MAMMALS OF CHINA AND MONGOLIA The Sika Deer (Cervus nippon boschi) is another example of a species charac- teristic of southeastern China, although represented also in Korea. The genus Rattus is found chiefly in warmer countries of the Old World and is “represented in South China by a number of species which do not go much beyond the Yangtze basin in the east, although occurring also in the western highlands, such as R. losea, R. edwardsi, R. bowerst, and others. With all these more southern species the limits of northward distribution vary slightly in each case, but as a whole they represent a very distinctly austral element. 5. The Western Highlands:—By far the most interesting and remarkable of the faunal divisions of China is that of the western highlands, from approxi- mately southern Kansu and southern Shensi southward to include Szechwan and parts of northern Yunnan and probably Kweichow. This great area is well watered and mountainous, many of its ranges extending up to ten thousand feet or more, with many peaks even in western Yunnan running to 13,000 feet and perpetually snow-capped. The northern boundary is marked approxi- mately by the east-west ranges of the Min Shan (in Kansu) and Tsingling (southern Shensi), but on the west the mountain chains of the Szechwan and Yunnan borders trend north and south, thus opposing a barrier against the westerly winds and, by condensing moisture from these air currents, producing an abundant rainfall with its consequent forest growth. The forests are largely of fir and spruce with hardwoods at middle levels, but in the higher parts, 8,000 to 10,000 feet, the growth is lower, and is characterized by thickets of small bamboo and rhododendron. The north-south ranges of western China are believed to be older geologically than the east-west Himalayas with which they are contiguous, and seem to have afforded asylum for many peculiar, often primitive, types of mammals not known elsewhere to-day. On the other hand, their contact with the Himalayas allows a certain interchange or exten- sion of range for some of the faunal elements, as in the case of such genera as Soriculus, Nectogale, Chimarrogale, Ailurus, Budorcas, Nemorhedus and others, which are present in the Chinese highlands and in the eastern Himalayas. Peculiar to the former, are the following genera: among insectivores, Uropsilus and the related Rhynchonax and Nasillus, the more mole-like Scapanulus, and the shrews, Blarinella and Anourosorex (the latter extending into India); of the primates, the Golden Monkey (Rhinopithecus); of carnivores, the Giant Panda (Ailuropoda) ; of rodents, the Rock Squirrel (Rupestes), and the Jumping Mouse (Zapus), notable for its close relationship to the American Zapus, as well as the subgenus Neodon of Microtus. The voles of the genus Eothenomys, although extending in mountainous country to the coast ranges of Fukien and westward to the borders of Burma, are essentially characteristic of these highlands, as are also the mouse-hares, Ochotona thibetana and its races, a PLATE V sis TRS ee Yenchingkou near Wanhsien, eastern Szechwan. Site of bat caves; the limestone ridge in the background the haunt of Edwards’s Giant Rat (Rattus edwardsi gigas) is FAUNAL AREAS OF CHINA AND MONGOLIA 17 forest-dwelling species, and possibly O. glovert of the mountainous parts of western Szechwan. The presence of a mole of the genus Talpa (T. longirostris) here is interest- ing, as its affinities are more western, so that it possibly entered China by way of the Himalayan chain. In the opposite direction, a number of species seem to have spread into the Himalayan region, or to have been indigenous there. Thus the Takin and the Goral as well as the Serow reach their western limit in the Bhutan region; the shrews of the genera Soriculus and Nectogale as well as Chimarrogale occur in the Himalayas; the small Panda (Ailurus) is found there also, but not the Giant Panda; the widespread Rattus fulvescens extends into Nepal. It is interesting to find that the lofty mountain masses of Yunnan and parts of Szechwan that reach altitudes of over ten thousand feet with perpetual snow at their summits, show very little in the way of special alpine species. This is apparently because even during the Pleistocene period there was no way whereby boreal mammals, if pushed south by the advancing glaciers in the north, could cross the east-west deserts of the Gobi and central Asia, for the lack of north-south mountain chains afforded no such highway for migra- tion as we see, for example, in western North America, where boreal types of birds and mammals occur on the upper levels of the Sierra and Rocky Moun- tains, far south of their sea-level distribution. Perhaps in part because of this lack of competitors, in part because of favorable conditions of climate and food, these western highlands have remained to the present the home of many annectant or peculiar types that have elsewhere died out. Thus the genus Neotetracus may be thought of as an ancestral member of the Erinaceide; Uropsilus is a primitive shrew-like member of the Talpide; Parascaptor is perhaps a relative of the American mole Parascalops; Eothenomys, with its subgenera Anteliomys and Caryomys, shows many intermediate characters from which more advanced types of microtines may have developed, while the subgenus Neodon is practically a Pitymys without fossorial modifications. The absence of the latter genus from China is noteworthy. Zapus in these highlands is a relict, having died out elsewhere in the Old World, and the raccoon-like Ailurus is an annectant genus of the Procyonide that has survived here but disappeared elsewhere. The three genera of goat-antelopes (Budorcas, Capricornis, and Nemorhedus) all occur together here as more primitive members of the Bovide, represented elsewhere only by the Rocky Mountain Goat, and perhaps by the musk-oxen in America. Hemmed in by deserts on the north and west, and by low country on the east, the south has been the main source of intrusives, and these include for the most part species or genera whose main area of distribution is at lower and warmer latitudes (for example the Rhesus Monkey and Sambar Deer). 18 THE MAMMALS OF CHINA AND MONGOLIA 6. The Subtropical Fauna:—While the mammals of southern China are mostly of warmer-country types, often with wide longitudinal and latitudinal distribution, there are a number of species of distinctly more tropical preference, that just reach the southern edge of the country or penetrate a short distance beyond its borders. These may be regarded as constituting a subtropical element, which appears in extreme southern Yunnan, Kweichow, Kwangsi, and Kwangtung Provinces and in the island of Hainan. Of such are the tree- shrews (Tupaia), represented by local races of the single species T. belangeri, in Yunnan and Hainan, as well as the shrew-like Hylomys and Neotetracus in the former; other insectivores include the peculiar mole, Parascaptor (Yunnan), and several species of white-toothed shrews (Crocidura); a larger number of bats of tropical and subtropical distribution, including two genera, Cynopterus and Rousettus, of fruit bats; and various Microchiroptera, as Taphozous, an emballonurid; Lyroderma, a megadermid; species of Hipposideros and Trienops of the Hipposideride; Tylonycteris, Scotophilus, Scotomanes, and Kerivoula of the Vespertilionide; and Cherephon and Nyctinomus of the Molosside. At least three species of langur monkeys (Pithecus), and the gibbons (Hylobates hoolock and H. concolor), just enter the southern borderland, the last on the island of Hainan; possibly, too, Macaca assamensis and the Stump-tailed Macaque (Lyssodes) should be regarded as intrusives from the subtropics. Among the Carnivora there are a number of subtropical species such as the Lesser Ferret-badger (Helictis taxilla sorella), the Indian Otter (Lutra taray- ensis from Yunnan), two species of palm civets of the genus Paradoxurus, one of which (P. minor exitus) barely reaches the southeastern border, the other (P. hermaphroditus laotum) occurring in Hainan; mungooses (Herpestes) of two species, one in Hainan, the other (H. urva) more widely spread across the mainland of extreme southern China. Many rodents might be added to this list, including the giant squirrels (Ratufa, in Yunnan and Hainan); several flying squirrels, large and small, as Petaurista hainana, P. yunnanensis, P. punctatus marica, Pteromys (Petinomys) electilis (Hainan) and Belomys pearsoni; the mole-rat (Rhizomys pruinosus) just reaches southwestern Yunnan and is of more southern distribution than R. sinensis and its races, while other genera include such subtropical mammals as Leggada, Bandicota, Vandeleuria, Chiropodomys, and Hapalomys, members of the Muride, a family of essen- tially warm-country distribution. Among the ungulates, the larger muntjacs (Muntiacus vaginalis muntjak in Yunnan, and M. v. nigripes in Hainan) may be included as subtropical species, as well as the Panolia Deer (Rucervus platyceros hainanus). There is some evidence that in former times elephant and rhinoceros may also have reached the southern borders of China. No doubt future work will add considerably to this list, for the southern provinces, except Yunnan, have not been thoroughly collected. PLATE VI The Mekong valley, western Yunnan FAUNAL AREAS OF CHINA AND MONGOLIA 19 7. The Tibetan Plateau:—On its extreme western border, China includes the eastward edge of the great Tibetan plateau. The traveler emerges from the high passes of the Szechwan highlands, as near Tatsienlu, on the ancient caravan trail, wpon more open country, broken, yet scarcely mountainous, sparsely covered with grass and scattered vegetation, partly watered by swift narrow streams, trending northwest to southeast, some of which unite to form the Yangtze River flowing eastward to the Pacific Ocean, while others, turning southward, form a series of deep, narrow and nearly parallel valleys transecting western Yunnan. This plateau is semiarid and supports a characteristic fauna, although the species of mammals are relatively few, and include only such as can withstand the rigorous climatic conditions and find subsistence there. The following species are known from the borderland of China and range westward into Tibet: the Black Grizzly Bear (Ursus pruinosus) which subsists in part upon the burrowing mouse-hares that it digs up; of these, the red-eared species (Ochotona erythrotis) is found on the high mountains of western Szechwan; the Gray-tailed Hare (Lepus oiostolus) and races which occur along the western edge of Szechwan and Kansu; the Himalayan Marmot (Marmota himalayana robusta), hardly if at all distinguishable from the animal of the more western Himalayas; and several larger hoofed mammals, including the White-lipped Deer (Cervus albirostris), remarkable for its rough coat, the Tibetan Gazelle (Procapra picticaudata), which ranges northeastward into Kansu, where it is represented by a closely similar race, Przewalski’s Gazelle (P. picticaudata przewalskii), the Blue Sheep (Pseudots nayaur szechuanensis) with a like distribution, and probably too, the Wild Yak, which is known to occur close to the borders of Szechwan, and slightly farther north enters Chinese territory in northwestern Kansu. Possibly the Bactrian Camel was originally part of this fauna. Future investigation will doubtless show that the Snow Leopard (Uncia uncia), a species of these barren heights, also enters Chinese territory in the extreme west of Szechwan, for, although no specimens seem to have actually been taken in the province so far as recorded, it is well known to range across Tibet at higher levels, preying on young wild sheep and ibex, while the skins are frequently traded across into China. No doubt a few additional species will eventually enlarge this list. CHAPTER III FAUNAL RELATIONS OF ASIA WITH NORTH AMERICA A CERTAIN obvious similarity between the fauna and flora of northern Eurasia and those of northern North America has led zoégeographers to unite both in a single major division, the Holarctic Region. So far as the mammals go, however, this likeness is most marked amongst the boreal types of the arctic regions and of the transcontinental evergreen forest belt, southward of which the community of mammalian forms becomes less and less evident or is almost completely lost. This seems to indicate that the last land connection by way of northeastern Asia and Alaska, along which interchange must have taken place, was far to the northward and of relatively narrow extent, so that only species of northern distribution were enabled to cross from one continent to the other. The effect of a continuous land bridge of this sort would be to shut out any arctic current bringing cold water from the north through Bering Strait, while the warm Japanese Current washing its southern shores would perhaps affect but a narrow strip along its southern edge, bringing to it a warm, moist climate such as at present obtains along the shores of Alaska, British Columbia, and the State of Washington. The abundant moisture favors forest growth, so that mainly forest-living species might be expected to take advantage of the connection to make the crossing. On the other hand, ‘“‘the shutting off of the warmer southern currents from the polar region probably marked the commencement of a cooler northern climate’? (Osborn, H. F., “The Age of Mammals,’’ 1910, p. 244). There is evidence that such a connec- tion existed in early Pleistocene times and lasted till near the end of the glacial period, when Asia and North America were again completely separated. At an earlier epoch, in the Pliocene, there was a similar connection between the two continents, with evidence of a generally warmer climate farther to the north which would have enabled species of more southern latitudes to make the crossing, while a cooler climate in the north during subsequent times must have forced many of these southward to more congenial conditions. In the present distribution of eastern Asiatic mammals, there seems to be indication that at least two such periods of interchange took place, and as one would 20 FAUNAL RELATIONS OF ASIA WITH NORTH AMERICA 21 expect, the species of the earlier immigration are found farther to the south than those of the later one. The living species of mammals in northern Asia are in many cases repre- sented in northern North America by forms so closely similar that the differ- ences are hardly if at all more than of subspecific value. In other cases, how- ever, those of one continent are not now represented in the other, although there may be evidence of their former presence in the shape of fossils, as in the case of the musk-ox, once of circumpolar distribution, but now confined to Greenland and northern Canada. When the forest fauna of northern Mon- golia is compared with that of the ‘“‘Canadian Zone” of North America, its New World equivalent, many such correspondences appear, together with a smaller number of disharmonies. Among the Insectivora, the Saddle-backed Shrew, Sorex araneus borealis, is represented in northern America by the very similar S. arcticus group which is found from Alaska to Nova Scotia, but it seems uncertain whether S. minutus or S. buxtont has a similar New World relative. Among the bats, the Eurasian Myotis daubentonii is doubtless the counterpart of M. lucifugus, which reaches tree limit from Alaska to Labrador, but other species are for the great part more southern in distribution. Among these the short-nosed Myotis frater, described from Fukien, seems to be the Chinese representative of M. volans of western North America which ranges northward to extreme southern Alaska. Possibly at some former time the range was continuous across an ancient land bridge so that under very little warmer conditions the common ancestor was able to extend farther northward, although since then the increasing cold has forced the modern descendants somewhat to the south of those latitudes. The close correspondence between the small-footed and long-haired Myotis muricola moupinensis with its peculiarly dark-based fur and contrasting reddish tips, and the western American M. californicus group suggests a near relationship, and perhaps again indicates that the latter were derived from this Asiatic stock at a time when land connections and slightly warmer temperatures in the north made it possible for the species to cross from the Old to the New World. The present northward distribution of the two shows a wide gap from the Chinese highlands to extreme southern Alaska, for as in the previous case, the climatic conditions in eastern China are now more rigorous than in corresponding latitudes of the southern Alaskan coast. The interchange perhaps goes back to Pliocene times. Turning to the Carnivora, similar correspondences occur. An outstanding case is that of the family Procyonide, the raccoons and their kin. At the present time the group is well represented in tropical America, with a single species of Procyon extending northward in the United States to southern Canada in forested areas. Its only representative in Asia is the small Panda 22 THE MAMMALS OF CHINA AND MONGOLIA (Ailurus fulgens) at present confined to the highlands of southwestern China and the adjacent portions of the Himalayas. There must have been a time, probably far back in the later Tertiary, when the group was more widespread, and a land connection in the north enabled the ancestral forms to cross from one continent to the other. A similar case is that of Zapus to be mentioned later. Other eastern Carnivora show even closer kinship with American species. Thus the northern Brown Bear (Ursus arctos lasiotus) seems a close relative of the American Grizzly, with similar pale shoulder markings, and Euarctos thibetanus, in spite of its larger size, can hardly be other than the Asiatic representative of the American Black Bear (Euarctos americanus) with a comparable wide range from northern Mexico to Alaska. This latter fact and the marked size difference may indicate that the American animal came with an older emigration than Pleistocene, although the first presence of bears in America is supposed to be in relatively late Tertiary times. The Red Fox (Vulpes vulpes subsp.) and the Wolf of northern Asia are so similar to those of northern North America as hardly to merit specific distinction from their New World congeners, implying a separation of relatively short duration, although the canids are on the whole conservative types. The Old World Ermine (Mustela erminea and races) finds its counterpart in the closely similar M. cicognanii of northern North America, a species characteristic of the evergreen forest area, while the Pygmy Weasel (M. rixosa pygmea) is clearly but sub- specifically related to the M. rixosa and races of the same area. Probably, too, the Sable (Martes zibellina and races) stands in a similar relation to the American Pine Martens (M. americana and races), but the Beech Marten (M. foina) is not represented in the New World. The lynx of northern Mon- golia (Lynx lynx isabellina) is replaced in the evergreen forests of Alaska by the Canada Lynx, at present regarded as a distinct species but perhaps of closer relationship than this implies. Coming to the rodents, the tree squirrels show a striking contrast, for Sciurus vulgaris (and races) of the transcontinental forest of Eurasia has no close relative in the evergreen belt of America, but its place is occupied by the smaller red squirrel (Tamiasciurus). The flying squirrel of the northern Asiatic forests (Pteromys volans subsp.) is currently regarded as generically distinct from its counterpart (Glaucomys sabrinus) of the American fur countries, but the differences are probably best considered at most of subgeneric value. On the other hand, the common ground squirrel of northern Mongolia (Citellus eversmanni jacutensis) is so closely similar to the Alaskan members of the C. parryi group that the relationship can hardly be more than subspecific, and the chipmunks (Eutamias) are evidently near relatives on both sides. Among the microtines are several interesting contrasts. Thus, of the red- backed mice (Clethrionomys), forest-living species, the two species of northern FAUNAL RELATIONS OF ASIA WITH NORTH AMERICA 23 Mongolia are represented in Alaska by but one (C. dawsonz), for the C. rufo- canus type is not known from the New World. In the meadow mice (Microtus), the common form of northern Asia is of the M. arvalis type, represented in northern Mongolia by M. mongolicus, with but four prisms in the second upper molar. This type is found in North America represented by M. operarius of Alaska, with a similarly formed second upper molar. It has apparently not extended far into the continent, for elsewhere over most of Canada and the United States Microtus pennsyl- vanicus is the dominant species with an additional postero-internal loop on the second upper molar. This latter type is represented in northern Europe and Asia by M. agrestis, of which M. a. mongol is the Mongolian race. The abundance of this type in North America to the exclusion of the M. arvalis type may be evidence that it was the first to reach the New World or that it may have reached the Old from the New. An interesting parallel in association is found in the Bobac Marmot of the Mongolian grasslands, which is more or less colonial and has a special enemy in the Masked Polecat (Mustela eversmanni tiarata). In this relation it is similar to the American prairie dogs (Cynomys) which find in the related Black-footed Ferret (Mustela nigripes) a particular enemy. The two hosts are not very closely allied, but the predators are. Of the mouse-hares (Ochotona) only one of the three Asiatic subgenera is represented by the various forms of western North America, namely, the subgenus Pika to which the Mongolian O. hyperborea mantchurica belongs. The presence of this northern group on both sides of Bering Strait indicates again a close affinity between the living species of the two areas. Among the hoofed mammals, the genus Sus, wild swine, is wholly un- represented in North America, but it is mainly a group of the warmer parts of Asia and Europe, ranging northward to Mongolia and the Baikal region, but more abundant to the southward. The Musk Deer, which ranges north- eastward into eastern Siberia, might have been expected to appear in western North America, while the Roe Deer, now common in southern Siberia, might also have been looked for, but both are unrepresented in the New World. On the other hand, the Red Deer and the Elk are both represented on the American side by near relatives, the Wapiti and the Moose. The Wild Horse (Equus przewalskii), of extreme western Mongolia, as well as the Bactrian Camel, while no longer represented in the North American continent by related forms, were probably immigrants in early Pleis- tocene from that area, where both groups went through mo&t of their evolu- tionary history. Remains of both genera are known from deposits of the Ice Age in Alaska. These close relatives on both sides of Bering Strait indicate, then, a 24 THE MAMMALS OF CHINA AND MONGOLIA former continuity of range by way of a land bridge between the two continents, of so comparatively recent a date that in most cases, at least, no great differ- entiation has taken place between the representative forms of the two areas. A comparison of the mammals of northern China with those of corre- sponding latitudes of North America shows greater differences between related groups, indicative probably of an earlier land connection with a still warmer climate at higher latitudes, so that it was possible for their ancestors to range farther north and so avail themselves of the opportunity thus afforded to cross in either direction. This older connection was perhaps in the Pliocene, when, as shown by fossil remains, such trees as the bald cypress (Taxodium), the tulip tree (Liriodendron), honey locust (Gleditschia), sweet gum (Liquidambar) and others now found in southeastern United States were present also in western Europe and doubtless across the intervening area, for they also persist at present in southern China. In this connection, too, it is interesting to recall that the ginkgo tree, now confined to eastern Asia, has lately been found fossil in the late Tertiary of the State of Washington. Among the Insectivora of North China and North America, few similari- ties appear. No hedgehogs survive in North America, although there is evidence that they may have occurred in early Tertiary; the genus Crocidura is also quite lacking, although represented by a few species in eastern China and Korea. A further contrast is the apparent absence of the genus Sorex, a diversified group in North America, and represented in the Mongolian forest and Siberia. Of moles, the genus Scaptochirus of North China has no close relative in America, but Scapanulus, found in the highlands of Shensi, Kansu, and Szechwan, is believed to be not distantly related to Brewer’s Mole, Parascalops, of eastern North America, so that the wide gap now separating them is doubtless of long standing. A somewhat parallel case is that of the Eastern Chipmunk (Tamias striatus), of the eastern part of North America, now generically different from Eutamias, its close ally in eastern Asia and western North America. The mole (Scaptonyx), of the western Chinese highlands, is related rather closely to Neurotrichus of the American northwest coast. Among the rodents, the squirrel group of eastern and northern China shows little similarity to the North American members. The large flying squirrels are unrepresented in the latter continent; the genera Callosciurus, Rupestes, Dremomys, and Sciurotamias are also wholly confined to eastern Asia, the three last almost wholly to China. An interesting contrast is seen in the cricetines, for, whereas in North America the common white-footed mouse (Peromyscus) is widespread and in part a forest animal, with long tail, its closest allies in North China and Mongolia are all ground-living or burrowing forms with tails shortened or in some species reduced to a mere stump. Thus the genus Cricetulus is widespread over open country and invades the desert, FAUNAL RELATIONS OF ASIA WITH NORTH AMERICA 25 while Phodopus is a typically desert-living genus, with pallid coloring and the soles of the feet provided with pads of hair for progress on sand in which it burrows. The niche occupied by Peromyscus in the New World is taken by the wood mouse (A podemus), a murine, which may in part explain the absence of a forest-living cricetine. The latter genus is confined to the Old World like the rest of the Muride, including Mus, Rattus and related genera, most of them warm-climate types. Thus the genus Rattus is chiefly tropical and south temperate in distribution with only two or three species that reach north temperate latitudes under natural conditions. No relatives of the abundant pocket gophers (Thomomys) of western North America are found in Asia, but their niche is occupied in northern China and parts of northern Mongolia by the aryicoline genus Myospalax of similar subterranean habits, while gerbils and jerboas (Meriones, Rhombomys, Allactaga, Dipus, etc.) fill the ecological niches taken in the New World by leaping desert species of Dipodomys. Of special interest is the case of the jumping mouse (Zapus), known from the Szechwan highlands of China, but so closely related to the American Zapus that it is at most but subgenerically different. This again may be a remnant of a wider distribution that took place possibly in Pliocene times, for at present it is unknown from other intermediate areas. The tree porcupines of America, on the other hand, are unrepresented in the Old World, though two genera of ground porcupines occur in China. The distribution of the “‘goat-antelopes”’ at present centers in southeastern Asia, with three genera—Nemorhedus, Capricornis, and Budorcas—found in China. These are doubtless to be looked upon as survivals of primitive members of the Bovide, but although the first- named reaches the southern borders of Siberia in Amurland, none appears in North America, although the Rocky Mountain Goat (Oreamnos) is doubtless closely related. The musk-ox, probably also to be regarded as a member of this group, was nevertheless found in northern Asia in the Pleistocene, though now confined to arctic America and Greenland. In general it may then be said that, while the mammals of northern Mongolia of the forest zone correspond in many cases to their close relatives of northern distribution in the New World, those of more southern latitudes, in northern and western China, are either unrepresented or their counterparts are much more distantly related. The obvious conclusion is that the former group owes its similarity to a more recent continuity of land area, while the latter group has been separated for a far longer period, allowing time for evolu- tionary changes in some types and the extinction or increase in others, with a resulting lack of similarity at the present time. = i en MISA. WEAOK UTI Ale SD C@AUITAGRe SAAD ° 4d tere Getattos Hilde i .cnmy yilvirmes) vio? 1 aie a it Witter ni Views 10 yon | wl ties Te whe ae Dever a ete i ad Wit Yrosts ar ibis fl WW Yay i RAT ees ence. Gel Upeine rub vitott ye Po, Sureogali, iia Jae Agr (Parle antaiyy @0M wet h} } Aven eT! fara hho ie) adt or honllimo cf cue Petal ed alee aie 2 a iT) ae rote: Ceped anti aw, gorkatone erent arty Yo tem, ry . Figts | Perhyitt Wn writ: at TES earpies ovens igh wont (twin vhyi TU nh ere) “gli edie mutgaeteeretiy br Soiaepeie Ss) ae fron fugrts ! visas 4 anit pee Lrriiaie tata ates Stat GIsTH nt \ dnt Wiig) : aan nh eo. nite i Coen D) arindeyng it 7 - a attest Voonpiog. fey Aveo Orrnitrorm ar Rangiaia i enteter 4 L-. Madey of eth fart iifue tAllinie to s0yyneh) ebnde anileohgyie® as. A th54 sy ir (3 TOL LA, Sevier Soya) aorta 5 ‘ . Ww) wvew AO] re Mee aerial vei bho wel odin iol! aaa A ty) (tries r f OS ) reer wiki oth ta. sony Gh ye feeresdai we AMG f rt owlt at hothbr viaeahy oe ted anit To cheatitud cede ies iy) ney i sed, eerie Sted tenner Br a ‘ jreerty hi Vcr itl ie TTY USAT deat edie madriediateihy 4988 , means ry a CaM ol aa ae Nath: aye ryndind ansil ter conan erin r . Wi Giiovg trod i VA HUY gets tn HAIG bai ont bast eatto tga go Meegtitentcenh pee pa an help iT sar bh hai rtzop nenepiemod bi ce mye are muhro ayrtly difer ney: arlene i nein) fer Depts tlt at am Git horn Saat? echt) nf ftir Hoewnittl RSE Nite Age Lot RoC AAT to AAA tion Ws tavivian ome 7 We ee Kid a Beiter! ride iy sSivewl crorftisie Sly euisest b 7 See lint, si wy) f4e8) aitatal/ et why Haanelitn & vote a ky relocun 1 § ety rr yt ehh ian! (dn POT Lz 0-9 “aTt) WG 3 bo tator x faieni roiteblh ott A ied. ipod feoit ab teri ea gildiayon ayer qn heeled) had (Aiserh) sides Raaditiee melite: Wo eomerane adt lide ote Bina of world) Yan Te bevntig ' Reavsiinlot salonlod 1 eoeeo veniie ne Hitters AAs teayend artt Jo unite ‘ — - mm fuiiteal wornttoos with: UY aay bal wo, ofl? Gt iariitad ete Tro 6 7 2 Beyer roll SO Telnwepeite qi eri mete ho ree ° ‘ “—- | how a ? tortie! of? jurlt ob coteelyee eylivde aT” patain eligi atom Wo OY slid ivr ial heehee Wren sian iy at varratintin eft » thy to) orit ynivetls honey ahead 1h pal eA rel song. va Bit tdi il ashen woitaete’ ont? hte eigyd Sitter (30 #93 ratte 5 wiht Hest one 3& Vtialicers eat ial = Wy Pha a ay , sib i 4 j ra) ee ) BAS \ ; Pi gitne B= MOP Tee SECTION II SYSTEMATIC ACCOUNT OF THE MAMMALS OF CHINA AND MONGOLIA SECTION II SYSTEMATIC ACCOUNT OF THE MAMMALS OF CHINA AND MONGOLIA CHAPTER IV—ORDER INSECTIVORA. INSECTIVORES - CHAPTER V—ORDER CHIROPTERA. BATS - - - - CHAPTER VI—ORDER PRIMATES. LEMURS AND MONKEYS CHAPTER VII—ORDER CARNIVORA. CARNIVORES a eae CHAPTER VIII—ORDER PINNIPEDIA. SEALS, SEA-LIONS’ - CHAPTER IX—ORDER CETACEA. WHALES, DOLPHINS, PORPOISES CHAPTER X—ORDER NOMARTHRA. SCALY ANTEATERS OR PANGOLINS CHAPTER XI—ORDER RODENTIA. GNAWING MAMMALS Page 29 150 279 312 490 494 514 523 CHAPTER IV ORDER INSECTIVORA INSECTIVORES TuE Order Insectivora stands nearest among living mammalian groups to the stem from which the placental mammals have arisen. Its members are of small or even minute size; their reproductive system is of a generalized type, their teeth are nearly or quite of the full number (eleven in each jaw above and below) characteristic of placentals, and the molars are similar in structure to those of the carnivorous marsupials, which in sundry other respects, as in the formation of the brain and the presence of a distinct perforation (the entepicondylar foramen) at the inner side of the condyle of the humerus, they resemble. As a whole the Insectivora tend to have a tubular snout and a forceps-like action of the jaws, which results usually in an enlargement of the anterior incisors at the expense of the canines, and the arrangement of the incisors in lengthwise instead of transverse rows. The canines may be so reduced as to have disappeared altogether in some forms, although in a few, as the hedgehogs, they are well developed and may have double roots. Their food is chiefly of an animal nature, in which insects form a large part. The group falls readily into two divisions, by some regarded as almost the equiva- lent of separate orders, namely, a more primitive, the Menotyphla, in which the pubic symphysis is long, the postorbital process well indicated or even forming a complete ring about the eye, and the Lipotyphla, in which the pubic bones are barely or not at all in contact and the postorbital processes unde- veloped. The two groups differ in many other minute but significant details of structure, as admirably set forth by Gregory (1910). The Menotyphla are represented by two living families, the Tupaiide or tree shrews, confined to southeastern Asia and of arboreal habits, and the Macroscelidide or elephant shrews found in Africa mostly south of the Sahara. The former is represented in China by a single species. The Lipotyphla, which include the moles, shrews, and hedgehogs, on the other hand, are very well represented in eastern Asia, and include a few primitive forms, as Uropsilus, which show the steps in the origin of the mole-like from the shrew-like types. Three families of 29 30 THE MAMMALS OF CHINA AND MONGOLIA this group are found in China and Mongolia, representing respectively, the hedgehogs and their allies, the moles, and the shrews. KeEy TO THE FAMILIES OF CHINESE AND MONGOLIAN INSECTIVORA A. Form squirrel-like, tail well-haired and distichous; skull with orbit completely encircled: bya: POY sting isscvepne sce s che fayes eased 2) ate ee ee Tupaiide B. Form not squirrel-like, tail close-haired, orbit not encircled by a bony ring. a. Crowns of upper molars nearly square in outline, consisting of four sub- equal cusps, with a small central cusp... 06.60 e eee eee eee Erinaceide b. Crowns of upper molars without a fifth central cusp. a’. Zygomatic arch complete, form usually fossorial................... Talpide b’. Zygomatic arch incomplete through loss of the jugal; form more or less *SMOUISE=IIKE) so dan cvtrive sc imlore pengelccseec cant eeeessoVe.a olateaa Ifomeree rete lsreter eens Soricidz Family TUPAIIDZ TREE SHREWS Tree Shrews are squirrel-like in appearance and like squirrels are active by day. They are the most primitive of placental mammals, retaining arboreal habits with a generalized structure. Unlike squirrels, however, they have all five fingers of the hand and foot well developed and provided with a sharp claw; the muzzle is somewhat elongated, and the ears rather short. In the skull, the orbital ring is complete. Genus Tupaia Raffles Tupaia Raffles, Trans. Linn. Soc. London, vol. 13, p. 256, 1821. In his review of the tree shrews, Lyon (1913) restricts Tupaia to those species in which the tail is well haired throughout, and distichous, the snout not especially elongated, the naked area on top of the nose cut squarely across instead of being slightly prolonged backward in the midline, the lower lobe of the ear smaller than the upper portion, and scantily haired. The tooth formula consists of: ig c.t pm.§ m.g=38. The upper incisors are slightly en- larged, the lower narrow and proclivous, the canine above smaller than the incisors. The upper molars show very well the primitive pattern of cusps, with the W-shaped commissures. The genus is tropical and subtropical in distribution, confined to southeastern Asia, including the neighboring large islands. A single species reaches the southern borders of China and the island of Hainan, and is divisible into three rather poorly marked forms in this wide area. The type species of the genus is T. ferruginea Raffles (=T. glis ferru- ginea) of Sumatra. THE INSECTIVORES 31 Fic. 1. Distribution Map. Tupaia 1. TJ. belangeri chinensis 3. T. belangert modesta 2. T. belangeri yunalis Key To CHINESE Races or Tupaia belangeri A. Light shoulder stripe evident. aa Generalitintiabovetsnreenish) 255 sci. «em sieeve spielen eee oe T. belangert chinensis aa General mbit OrOwIiShyipeeeyt eth «) sieneyahsc jens (oieh wekeuey Vaid eusbad oceania: T. belangeri yunalis B. Shoulder stripe obsolete, general tint above grayer............... T. belangert modesta I. Tupaia belangeri chinensis J. Anderson Tupaia chinensis J. Anderson, Anat. and Zool. Researches Western Yunnan, p. 129, pl. 7, figs. 8, 9, 1879. Tupaia belangert chinensis Thomas, Ann. Mag. Nat. Hist., ser. 8, vol. 13, p. 243, 1914; ibid., vol. 14, p. 472, 1914. Type specimens:—Anderson mentions two individuals procured by him upon which he founded his description: No. 204a, from Ponsee, Burma, in 32 THE MAMMALS OF CHINA AND MONGOLIA alcohol, and No. 204b and c, the skin and skull of an adult from Muangla, Yunnan (Cat. Mammals Indian Mus., Calcutta, pt. 1, p. 155, 1881). Accord- ing to Dr. Lyon, these are probably still in the Indian Museum. Description:—General color above, from snout to tail, including the backs of hands and feet, an olive gray, very slightly darker in the middle region of the back where the black hairs are more numerous. The individual hairs are of two kinds: the shorter, with slaty bases and the terminal part pale greenish yellow or with a black ring dividing the greenish-yellow portion; and slightly longer, stouter hairs chiefly black, with or without a minute tip of greenish yellow. At the side of the neck a short, ill-defined pale streak extends back from the ear, where the terminal yellowish portion of the hairs is slightly paler and more extensive. The tail is like the back above, becoming slightly darker toward the tip; on its lower surface, the shaft is clothed with short appressed hairs colored like the rest of the tail in the apical third, but paler ochraceous buff basally. The entire under side from chin to vent, and the lower sides of the limbs, are pale ochraceous buff, the hairs with slaty bases except on the midline of the chest, the inguinal region, the interramal area, and arms. Ears similar to the back, but very thinly clad with short hair. Measurements :—Collectors’ measurements of two races are as follows: ° T. belangeri chinensis No. Total length Tail Hind foot Ear Sex 84929 358 175 44 16 ror 84935 338 178 42 16 a 84937 358 186 42 14 rol 84938 318 174 42 14 ci 84940 356 183 42 17 cS 84928 340 170 43 16 Q 84930 335 155 44 16 g 84931 320 170 40 Ly, ie) 84934 347 163 45 15 2 84936 344 172 42 16 Q T. belangeri modesta No. Head and body Tail Hind foot Ear Sex 59828 185 159 43 16 (et 59832 180 150 44 15 fou 59836 180 149 42 II rol 59838 185 150 45 10 rol 59840 189 150 46 13 rou 59829 185 159 43 16 g 59837 180 144 40 14 2 59839 185 165 45 17 2 59844 175 160 45 Io g 59850 195 155 44 a g THE INSECTIVORES 33 CRANIAL MEASUREMENTS OF TUPAIA Zygo- Width Upper Lower Greatest Basal Palatal matic Mastoid across tooth tooth No. length length length width width molars row row Locality T. belangeri chinensis 44280 46.7 41.3 24.2 23.5 16.7 15.9 23.7 22.9 Yunnan 44290 48.5 43-7 27.6 23-7 17.6 15.6 27.4 23-5 Yunnan 44293 47.2 41.6 24.4 24.5 17.5 15.5 24.6 23.0 Yunnan 84936 46.7 40.6 24.7 23.3 172 15.9 24.2 21.6 Yunnan 84929 48.6 42.2 25.6 24.6 ey) 14.6 24.2 22.0 Yunnan 84932 44.3 36.1 22.4 22.5 16.9 15.0 22.6 21.6 Yunnan 84934 47.4 41.7 25.1 24.7 18.2 16.0 25.1 23.5 Yunnan 84940 47-7 42.4 25-3 25.6 19.1 15.6 24.4 22.4 Yunnan T. belangeri yunalis 13685 Mcz 48.2 42.5 26.0 25.2 19.4 16.6 25-7 24.5 Yunnan 13686 MCZ 45-3 40.5 23.7 23.9 17.0 15.8 23.6 22.5 Yunnan 13687 MCz 45-5 40.7 23.6 23.4 18.5 15.5 24.1 22.2 Yunnan T. belangeri modesta 52829 48.8 42.7 25.8 23.8 18.0 16.6 25.7 24.5 Hainan 59843 49.0 43-9 27.0 25.2 18.6 16.0 26.5 24.9 Hainan 59848 48.6 42.2 26.0 24.5 17.9 16.8 25.7 237, Hainan Occurrence and Habits:—This tree shrew is characterized by its rather uniform olivaceous-gray color above, and as indicated by Lyon and later con- firmed by Thomas (1914, p. 243), is a slightly differentiated race of T. belangeri of southern Burma, differing in its almost complete lack of ferruginous coloring on the back. True T. b. chinensis occurs in the North Shan States and across the southwestern part of Yunnan. A large series collected by Dr. R. C. Andrews and Mr. Edmund Heller, as well as others collected for the British Museum and the U. S. National Museum, indicates that it has a wide alti- tudinal distribution, from the lower levels as at Namting River on the Burma border, up to 9,000 or 10,000 feet on the Likiang Range and the mountains to the east. Others are from the Mekong River and about Tali Lake, Yun- nanfu, and Yunnanyi. Specimens from the last three localities show inter- gradation with T. b. ywnalis in the very slightly warmer tone of the back. It does not extend much farther north than the Likiang Range. Little seems to be recorded concerning the habits of this species in China. It is in part tree-living, but apparently also spends much time on the ground. Anderson (1879) notes that the first one he saw on his Yunnan Expedition was in a grassy clearing close to patches of fruit, and that he at first mistook it for a squirrel. The stomach of one collected by Heller on the Mekong River con- tained seeds and leaves, indicating a partially vegetable diet, although insects and various other forms of animal life doubtless constitute its chief food. 34 THE MAMMALS OF CHINA AND MONGOLIA Specimens examined:—The following thirty-seven: Yunnan: Chiho, twenty miles south of Likiang, 2; Fengyang, 1; Hainkai, 1; Hsiohsien, 1; Likiang, 7; Mekong River, 3; Makaihsien, 2; Namting River, Burma border, 1; Peitai, 1; Mucheng, 1; Tali Lake and vicinity, 3; Wutinghsien, 1; Shihku, Yangtze River, 1; Yunnanfu, 1; one hundred and sixty miles west, 1; Yunnanyi, 5; Kaochiao, 6. 2. Tupaia belangeri yunalis Thomas Tupaia belangeri yunalis Thomas, Ann. Mag. Nat. Hist., ser. 8, vol. 13, p. 244, 1914. Tupatia ferruginea Shih, Bull. Dept. Biol., Sun Yatsen Univ., Canton, no. 4, p. 2, 1930. Type specimen:—A skin and skull, No. 12.7.25.45, British Museum, one of seven collected by Orii, July 10, 1910, at Mengtsz, Yunnan. Description:—Similar to T. b. chinensis, but the upper parts darker and of a much warmer tone, due to the more ochraceous subterminal rings of the particolored hairs, as well as to the more abundant black hairs which in some specimens darken the rump more than the back. The shoulder stripe is less conspicuous, shorter, and a nearly clear ochraceous, not so pale as in chinensis. Measurements :—See table under T. b. chinensis, page 33. Occurrence and Habits:—In southeastern Yunnan the tree shrew of the T. belangeri type becomes much less pale than in the western parts of the prov- ince, the greenish tone is warmer, the back darker, and the color a decided ochraceous. Thomas, who first recognized this difference, had specimens only from the extreme southern border, at Mengtsz, the type locality, whence also the Museum of Comparative Zodlogy has one specimen obtained by the same collector. Undoubtedly this form extends eastward along the extreme southern border of China for an undetermined distance, for Shih (1930, p. 2) records specimens secured by a Chinese expedition from Sun Yatsen University, Canton, in the Yao Shan area of Kwangsi, at Loshiang and Chinsiu. He observes that they are more ferruginous above than the Hainan tree shrew. There seems to be no record of the tree shrew in extreme southern Kwangtung, though its occurrence in Hainan implies its presence there; nor did Swinhoe learn of it near Amoy. Specimens examined:—Three, namely: 2 from Mengtsz, Yunnan, and I from Tongking (M.C.Z.). 3. Tupaia belangeri modesta J. A. Allen Tupaia modesta J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 22, p. 481, 1906. Tupatia belangeri modesta A. B. Howell, Proc. U. S. Nat. Mus., vol. 75, art. I, p. 5, 1929. Type specimen:—An adult male, skin and skull, No. 26654, American Museum of Natural History, from Leimuimon, island of Hainan, January 5, 1903. THE INSECTIVORES 35 Description:—Upper surface of body and tail an even and finely ticked mixture of the usual black hairs and particolored hairs with subterminal ochraceous ring; muzzle and shoulders slightly grayer than in T. b. yunalis. Under side of chin, throat, and arms buffy white to the roots of the hairs, this color continued to the central area of the belly and inguinal region, where, however, the hairs are slaty-based. Shoulder stripe obsolete, at most a barely indicated area where the pale tips of the hairs are unmixed with black; in some immature individuals it is better marked. The pelage of December specimens is shorter and less full than in winter skins of 7. b. chinensis from the Yunnan highlands. Measurements :—See tables on pages 32 and 33. Occurrence and Habits:—Notwithstanding the statement of its describer, that this ‘‘does not appear to be closely related to any of the previously de- scribed forms,” it is really very close indeed to yunalis, differing chiefly in a slightly grayer tinge to the muzzle and shoulders, an average character only. From the race chinensis of western Yunnan, it is at once separated by its much more brownish hue, lacking the peculiar greenish tint of the latter. The shoulder stripe may be hardly perceptible, but in some immature specimens it is clearly present. Swinhoe was the first naturalist to list the mammals of Hainan, but he evidently had no knowledge of the tree shrew of the island. It is, however, fairly common in suitable places. Dr. J. A. Allen (1906, p. 481), in recording it for the first time and publishing the original description, had seven speci- mens, five from Leimuimon in the mountains of central Hainan (the type locality), one from Utoshi, and one from Hoihow on the northern coast. He notes that there is a certain amount of individual variation in color, the youngest of the series having the throat more richly colored, tawny ochraceous. A. B. Howell (1929) mentions a specimen in the U. S. National Museum from Kachek, but otherwise there seem to be no other records of definite localities. Mr. Clifford H. Pope, who was successful in obtaining the large series noted below, from Nodoa and Namfong, 1922 and 1923, contributes the following interesting notes. ‘‘This little animal is of general distribution on the island, but nowhere occurs in great numbers. We first secured it in trapping—two of them caught in a thicket of high grass, small trees, and bushes. Later several more were caught, but never more than one or two at a time or in the same place in spite of dozens of traps set out. Those we caught came to bait of peanut butter. The fact that the traps were nearly always set in late after- noon and taken up early the following morning may in part account for our failure to take more, for they seem to be largely active by day. One was shot in the Mission compound, at Nodoa, near if not actually in one of the foreigners’ houses and I was told that on a previous occasion one had been found in the 36 THE MAMMALS OF CHINA AND MONGOLIA same house. Another appeared to live under the porch of one of the houses. I used to see it hopping about in the grass by the side of the house in the middle of the forenoon but if approached, it would run under the house and hide. They seem to be largely terrestrial, as well as diurnal, for my Chinese hunter shot a few, all in broad daylight. On occasion they may utter a shrill cry.”’ Specimens examined:—In all, twenty-eight, as follows: Hainan: Nodoa, 23; Namfong, 5. Family ERINACEID HEDGEHOGS AND THEIR ALLIES This family includes the hedgehogs and their relatives. They differ in many details of structure from the tree shrews, and constitute with the other living insectivores (except the latter and the elephant shrews) the suborder Lipotyphla. The pubic symphysis is very short, hardly more than a contact of the pubic bones. There is no complete bony ring surrounding the orbit as in tree shrews, but instead the postorbital processes are reduced to a mere point (in Neotetracus) or quite lacking. There is a marked constriction of the skull behind the orbits, and the zygomatic arch is slender. The snout is not especially elongate, and the palate retains the primitive trait of having its posterior border bounded by a raised transverse ridge. The first upper incisor is the longest, the canine is somewhat reduced in size, and usually two-rooted. The first premolar is large, its main outer cusp extending well below the level of the molars. The latter are of characteristic form, with the postero-internal cusp (hypocone) so enlarged as to give the crown of the tooth a nearly square outline. There are four blunt main cusps and a smaller fifth cusp in the center of the crown. The five digits are usually retained on each foot, the radius and ulna are free, but the tibia and fibula are fused together. Two subfamilies are recognized: the Gymnurine, including the more primi- tive members with nearly normal form, spineless coat and a longer or shorter tail; and the Erinaceine or Old World hedgehogs, with shortened skulls, com- pact, nearly tailless bodies, and a protective spiny coat. Key TO THE GENERA OF CHINESE AND MoNGOLIAN ERINACEIDZ A. Form rat- or mouse-like, fur normal, tail slender. a. Tail about half the length of head and body......................005- Neotetracus bx Dailgabouttasilonp as the hind footue. s0) eee Pole nee nee eter iae ele Hylomys B. Form stout, back with a spiny coat, tail a mere stump. a. Spines of the crown with a median parting, ears long, the postglenoid proc- esses of the skull as large as the mastoid processes and hollowed internally Hemiechinus b. Spines of the crown without a median parting, ears shorter............ Erinaceus THE INSECTIVORES $7 Genus Hylomys S. Miller Hylomys S. Muller, in Temminck, Verhand. Natuurl. Gesch. Nederl. Bezitt., vol. 1, Zoogd. Indisch. Archip., p- 50, 1839. In external form the body is slender, mouse-like, with a tapering snout and well-developed ears, but the tail is very short, about the length of the hind foot, slender and thinly haired. The hands and feet retain each five digits but the first and fifth are so short as barely to reach the bases of the three middle toes. The skull has the full placental tooth formula: i$c.ipm4m3=44. The anterior incisor is longest, the two others much smaller and shorter; the canine slightly larger and triangular in profile, with two roots; the three first pre- molars are all small, two-rooted, and hardly half as high as the fourth which is large with a prominent outer cusp, exceeding the molars in height. The latter are squarish in crown view, the two anterior of nearly equal size, the third much smaller with its outer posterior cusp much reduced. These small ground-living insectivores are known from the more tropical parts of the Malay Peninsula and the larger East Indian Islands. One of the forms barely reaches the southwestern border of China in Yunnan. The type species of the genus is Hylomys suillus Miller and Schlegel, of Java. 4. Hylomys suillus peguensis Blyth Hylomys peguensis Blyth, Journ. Asiatic Soc. Bengal, vol. 28, p. 294, 1859. Type specimens:—The species appears to have been based upon two in- dividuals from Pegu, Tenasserim, obtained by Blyth and probably still in the Indian Museum at Calcutta, where Anderson seems to have examined them and removed the skull of one previous to 1872. Description:—Upper surface of the head, body, and limbs a very uniform pale olive yellow and black mixed, giving a somewhat similar color to that of Tupaia b. chinensis. Individual hairs are either black throughout or slaty at the base with a short, pale-ochraceous terminal or subterminal band. The sides of the face about and below the eye are slightly clearer and brighter ochraceous, and the rump is a very little brighter than the middle of the back, but otherwise the coloring is very uniform. A certain sheen is imparted to the pelage by the burnished appearance of many of the longer hairs with their yellowish tips. The ears, feet, and tail are nearly naked, with a scattering of minute short hairs, those on the hind feet and tail brownish. Below, the color- ing of the upper surface merges rather suddenly with that of the under side, which is grayish white, with a very faint creamy tint on the lower throat and chest, the bases of the hairs slaty gray. The lower side of the metatarsus is clad with short stiff hair. The skull and skeleton have been very carefully described and figured by Anderson (1874) from a specimen he picked up on the Yunnan border near 38 THE MAMMALS OF CHINA AND MONGOLIA Ponsee, Burma. In form the skull differs from that of Tupaia in a slightly more elongate snout, less rounded, more flattened brain case, and the reduction of the bony ring about the eye to a mere projecting point on the frontal behind the eye. It is very light and delicate, and the zygomatic arch is thin and slender, especially at the posterior part. The palatal bones in the specimen examined contain several minute vacuities, and the palate itself ends in a transverse ridge with a short median projection. The palate is somewhat arched, the tympanic bullze incomplete. The teeth are of the full number found in placentals, 44. The anterior incisor is largest, the second about half its height, and the third about half the height of the second. The canine and fourth premolar are again larger, the former in the upper jaw double-rooted, the latter with its outer anterior cusp (paracone) much enlarged, while between them are three other much smaller premolars of subequal size and nearly half the height of the two larger teeth. The molars are three above and below, the first two upper ones squarish in outline, and consisting of four main cusps with a fifth smaller central cusp. The third molar has the metacone reduced, so that the outline is nearly triangular. As an anomaly, No. 44274 (now 20687 M.C.Z.) has but two upper incisors on the right side, the small third incisor having been lost. Measurements:—The six specimens were measured by the collector as follows: No. Head and body Tail Hind foot Ear Sex 44112 115 19 25 17.0 oy 44273 112 20 23 16.5 J 44275 115 23 24 17.0 rol 44113 IIO 22 23 17.0 °) 44272 105 24 23 16.0 2 44274 125 22 23 16.5 2 CRANIAL MEASUREMENTS OF HYLOMYS Zygo- Width Upper Lower Greatest Basal Palatal matic Mastoid across tooth tooth No. length length length width width molars TOW row Locality 44272 31.7 28.5 17.3 — 13.3 10.1 17.0 15.5 Yunnan 44274 32.9 30.0 1725) / 0 (065) p38 10.7 16.3 15.6 Yunnan 44275 32.4 29.8 17.0 16.5 12.7 10.4 16.5 15.6 Yunnan Occurrence and Habits:—Little is known of the habits of this species, but Anderson supposed that it was, in part at least, tree-living, although its elongate feet and shortened tail would rather indicate ground-living habits. The genus was first discovered in Java, but later on the mainland of the Malay Peninsula in a nearly identical form. Blyth in 1859 described as Hylomys peguensis the very similar animal from Pegu, Tenasserim, and Anderson (1874, 1879) during THE INSECTIVORES 39 his expedition to the borders of western Yunnan, picked up dead on the path at Ponsee, Upper Burma, a specimen that he referred to the same species and later used as the basis of his careful account of the skeleton with figures. The occurrence of Hylomys in Borneo was recorded by Thomas, who in 1888 named this island form H. suillus dorsalis. In recent years, the form on Sumatra has been distinguished as Hylomys parvus by Robinson and Kloss, and the latter author gave the name Hylomys siamensis to a specimen from Hinlap, eastern Siam, on the ground of paler, more buffy coloration and narrower nasals, while in 1925 Thomas named H. suzllus microtinus, a light-brown race from Thai-Nien, Tongking. In the lack of specimens for more minute comparison, I have re- ferred to H. suillus peguensis the series obtained in western Yunnan by the American Museum Asiatic Expeditions, following Anderson, who had seen the original specimens. It is not at all impossible, however, that they will even- tually be found to constitute a local race. Although Anderson (1879, p. 138) long ago recorded this animal from the very border of Yunnan, it remained for Dr. Andrews’s expedition actually to secure it in China. He and Edmund Heller discovered it at only one place, on the Namting River at the Burma border, at an elevation of 1,700 feet. In late February, 1917, six specimens were trapped, one of which contained two embryos. There are apparently no other records, and the species seems thus to be another of those subtropical mammals whose range just reaches the southern edge of China. Specimens examined:—Six, from Yunnan, Namting River, Burma border. Genus Neotetracus Trouessart Neotetracus Trouessart, Ann. Mag. Nat. Hist., ser. 8, vol. 4, p. 389, 1909. Externally the insectivores of this genus resemble Hylomys except that the tail is well developed, about half the length of head and body, thinly clad with minute hairs. The genus was so named by Trouessart in the belief that its tooth structure resembles that of the supposed fossil hedgehog (Tetracus) of the Eocene of France. It is, however, as Thomas (1911d, p. 162) has pointed out, ‘strictly a member of the Gymnurine,” and only distantly related to the hedgehogs. Itisanear relative of Hylomys, with which it agrees in the formula of both its milk and permanent dentitions, except in the absence of the minute second premolar, p2, showing, therefore, a more progressive state, so that the tooth formula is: i. $c. + pm. $m. $=40 in the adult, while in the milk den- tition there are only two upper incisors. On the other hand, its tail has under- gone less reduction, and the thumb is very slightly longer. The skull resembles that of Hylomys, but the rostrum is slightly shortened, correlated with the loss of one premolar, and there is a slightly more definite postorbital process. The palatal bones have better-developed slit-like vacu- 40 THE MAMMALS OF CHINA AND MONGOLIA ities, one in each, with a few irregular pores behind them. The upper first incisors are a trifle longer and more vertically placed. The type and only known species of the genus is the following. 5. Neotetracus sinensis Trouessart Neotetracus sinensis Trouessart, Ann. Mag. Nat. Hist., ser. 8, vol. 4, p. 389, 1909. Type specimens:—No one of the original series of seven skins and skulls from Tatsienlu, Szechwan, is designated as the type, nor are their numbers or location given. All are therefore to be regarded as cotypes, and are presum- ably still in the Muséum d’Histoire Naturelle in Paris. Description:—General external appearance much as in Hylomys suillus peguensis except that the tail is much longer, slender and mouse-like, and the mystacial bristles somewhat longer. The upper surfaces of head and body are of a general olive-brown tone, and the under parts are gray with a very light buffy wash. The individual hairs of the back and sides are of two sorts: some entirely black; others with blackish bases, then a band of light ochraceous and a fine black tip. On the sides the latter hairs are slightly more abundant and their ochraceous rings broader and polished, giving a lighter and somewhat glinting appearance. The sides of the head and neck have a richer tint through the more rusty color of the light band. The ears and upper side of the tail are dusky, dark brown, covered thinly with blackish-brown hairs visible with a lens only. The entire under side is pale gray, the hairs with slaty bases, the whole with a faint suffusion of buffy. The backs of the feet are thinly covered with pale hairs, among which on the metatarsal area are a few dark-brown hairs. The skull, while much resembling that of Hylomys, differs in many details, as noted above. Thus, in addition to the reduction of the premolars, from four above and below, to three in each jaw by the loss of one of the minute teeth, the canine is slightly smaller; the first upper incisors are actually larger and more pointed. The rostrum is relatively shorter through the reduction of the front end of the maxillaries and of the nasals; the palatal bones are more fenestrated, and the postorbital process is more developed and distinctly projects as a sharp spicular point. Measurements:—The following measurements were made in the field by the collector: No. Head and body Tail Hind foot Ear Sex 44238 105 64. 26.0 avyeild rot 44252 122 62 25-5 18 ro 44258 II5 60 24.0 17 fou 44269 125 70 26.0 18 rot 44266 110 64 24.5 17 9 THE INSECTIVORES 41 CRANIAL MEASUREMENTS OF NEOTETRACUS Zygo- Width Upper Lower Greatest Basal Palatal matic Mastoid across tooth tooth No. length length length width width molars TOW TOW Locality 44238 30.8 27.4 16.8 17.0 12.0 10.0 15.7 15.0 Yunnan 44239 29.9 27.4 16.1 16.5 12.4 9.5 14.7 14.0 Yunnan 44240 30.3 27.2 16.3 16.1 12.5 9.5 14.6 14.0 Yunnan 44249 29.3 26.0 15.6 15.6 12.4 9.4 14.4 14.0 Yunnan 44252 32.0 29.0 17.8 18.0 13.0 10.0 16.0 15.0 Yunnan : 44254 28.5 26.2 15.4 15.7 12.1 9.3 14.1 nae7) Yunnan 44255 29.8 27.4 16.3 17.0 12.5 9.8 15.0 14.4 Yunnan 44256 30.8 27.7 16.5 17.2 12.6 10.0 15.6 14.6 Yunnan 44258 31.0 28.0 16.2 17.0 12.0 9.7 15.3 14.6 Yunnan 44266 31.0 27.2 16.0 L73 12.3 10.0 15.0 14.8 Yunnan 44267 32.3 30.0 18.3 18.0 12.6 9.6 16.4. 15.4 Yunnan Occurrence and Habits:—The discovery of this remarkable insectivore is due to the enlightened interest of Monseigneur Biet of the Catholic Mission at Tatsienlu, Hsikang, who sent a series of seven from that locality to Professor E. L. Trouessart at Paris. The latter in a brief account of the animal (Troues- sart, 1909) announced his intention of giving later a more detailed description of its anatomy, but this has never appeared. The only other records for the species are those of Thomas (1911d, p. 162), who reports that the Duke of Bedford’s Expedition under the leadership of Malcolm P. Anderson secured a male forty-five miles southwest of Yachow (east of Tatsienlu) and thirteen at Omei Shan (just south of the latter locality) ; while in Yunnan, Mr. F. Kingdon Ward obtained a specimen at Yangpi, 7,000 feet. — The American Museum Asiatic Expeditions found this a common species in parts of western Yunnan, and obtained series at Mucheng, on the Salween drainage (7,000 ft.), Homushu Pass (8,000 ft.), and at Taipingpu on the Shweli River, 7,000 ft. Evidently it has a somewhat restricted range at inter- mediate altitudes in the northern half of Yunnan and the southern half of Szechwan, with probably a slight extension eastward. Anderson, who secured the series at Omei Shan, notes that they are found in damp forest away from water. The stomach contents of one consisted of “earthworms.”’ The stomach of another collected by Andrews and Heller contained, on the contrary,.‘‘vegetable matter,’’ indicating perhaps a mixed diet, for which the low-crowned molar teeth are well suited. Anderson found four embryos in one and five in another (August 10) female; four were found also by Andrews and Heller in a female taken April 10 at Taipingpu. The mammez are 2—2, or 8 in all (Thomas, rgrid). Specimens examined:—lIn all, forty-four, as follows: Yunnan: Mucheng, Salween drainage, 21; Homushu Pass, 8,000 feet, 17; Taipingpu, Shweli River, 7,000 feet, 6. 42 THE MAMMALS OF CHINA AND MONGOLIA Fic. 2. Distribution Map. Erinaceus P Hemiechinus 1. E. europeus dealbatus 3. H. dauuricus dauuricus 2. E. europeus miodon Poe se dauuricus alaschanicus Genus Hemiechinus Fitzinger Hemiechinus Fitzinger, Sitzungsb. Kaiserl. Akad. Wiss., Wien, math.-nat. Classe, vol. 54, pt: I, p. 565, 1866. Ericius Sundevall, Kong]. Vet.-Acad. Handlingar for 1841, Stockholm, pp. 223, 230-237, 1842. Lonnberg, Ann. Mag. Nat. Hist., ser. 9, vol. 9, p. 620, 1922 (preoccupied by Ericius Tilesius, 1813, for a genus of fishes). : THE INSECTIVORES 43 The hedgehogs of this genus differ from the typical members, of which Erinaceus europeus is an example, in having no median parting of the spines on the crown, while in the skull the postglenoid processes are as large as the mastoids and hollowed internally (Thomas, 1918, p. 193). The type species is H. platyotis of northeastern Africa, but in their main distribution they are characteristic of southeastern Europe and the desert region of central Asia. In this group the ears are more prominent than in the common European hedgehog, with which, however, it agrees in having five toes on both fore and hind feet, and differs from the African genus Atelerix with four toes only, on the hind foot. All the species are provided with a spiny armor, and if dis- turbed can roll themselves into a ball, presenting the spines in all directions. The large heavy skull, with shortened rostrum, no postorbital processes, large palatal slits, and well-developed tympanic rings is in contrast to that of the Gymnurinz. The dental formula is the same as in Erinaceus, namely: i¢ ct pm. m$=36. The first upper incisor is stout, terete, and quite twice as high as the second and third. The second upper incisor is smaller than the third and single-rooted, while the latter is two-rooted, as are also the canine and two anterior premolars. The outer anterior cusp of the last upper premolar is high and conspicuous. The two anterior molars are as typically in the family, squarish in outline, with four blunt cusps and a minute central one. The last upper molar, like that of Erinaceus, is reduced to a narrow crescent, but instead of having its long axis nearly in line with the outer border of m?, it is instead nearly transverse. Both lower incisors, the canine and the first premolar are single-rooted; the second premolar is large and high with two roots, and two prominent cusps, of which the anterior in profile is slightly the shorter. The last lower molar lies on the inner side of the tooth row and is reduced to a small oval in crown view, with indications of two very short cusps at its posterior border. Although several species of this group have been described from Mongolia and its borders, it seems unlikely that more than one species is really repre- sented, with perhaps one subspecies, the value of which is still not very clear. 6. Hemiechinus dauuricus dauuricus (Sundevall) Erinaceus dauuricus Sundevall, Kong. Vet.-Acad. Handlingar for 1841, Stockholm, p. 237, 1842. Dauuria. Erinaceus auritus Pallas, Zoographia Rosso-Asiat., vol. 1, p. 138, 1811; vol. I, p. 138, 1831 ed.; not Erinaceus auritus Linneus, 1758. Transbaikalia, Dauuria. Erinaceus dauricus Wagner, Arch. f. Naturgesch., vol. 9, pt. 2, p. 27, 1843. ?Hemiechinus przewalskit Satunin, Annuaire Mus. Zool. Acad. Imp. Sci. St. Pétersbourg, for 1906, vol. 11, p. 181, 1907. ‘‘Nord China?’ Hemiechinus (?) dauricus Satunin, ibid., p. 185. ?Ericius przewalskii Lonnberg, Ann. Mag. Nat. Hist., ser. 9, vol. 9, p. 626, 1922. 44 THE MAMMALS OF CHINA AND MONGOLIA Type specimen:—No type is specified. Sundevall’s name is based directly on the description by Pallas of specimens of a hedgehog from Dauuria, Trans- baikalia. Pallas, in his Zoographia, gives accounts of both the common European Hedgehog and the long-eared species, for the latter of which he uses Linnzus’s name, Erinaceus auritus, to include all the Asiatic forms of this group, with a general range (op. cit., p. 138): ‘In australioribus Tatariz magne atque Sibirie . . . usque ad Baicalem lacum frequens.’’ He makes special mention (op. cit., p. 139) of the fact that those from the latter area are larger and paler: ‘“‘In Daurico qui libr. rossicas 23g ponderabat, longitudo a summo naso ad anum 9”. 7’”’. aures 1’’. 4’. cauda 1”. 1’. . . . Dauuricis in genere vellus subtus fuscescente-cinereum, vel subgryseum ... Mongolis sunt in deliciis.”” When Sundevall (1842) published his synopsis of the hedgehogs, he stated that the Dauurian species was unknown to him by specimens, but on the basis of Pallas’s brief diagnosis, gave it the name Erinaceus dauuricus. It may | be questioned whether this hedgehog reaches the Transbaikalian district, for Radde (1862), who collected a series of hedgehogs there, in his account refers them to E. europeus and in his plate figures skulls of the Amur race of that species. Satunin, however, has examined his original set of five skins with fragmentary skulls and (19074, p. 186) gives a brief account of them, reviving Sundevall’s name. Description:—The spines begin on a line slightly behind the anterior base of the ears and extend to the area just above the tail. The spines are dark brown at the extreme base, then dull whitish for nearly one half their length, with next a band of brownish black and a contrasting white tip; the coloring of the hair of the lower parts is very pale, varying somewhat, but apparently, from Satunin’s account of the Dauurian specimens, the upper part of the head and the sides are pale brownish gray with faint rusty wash on back of snout and forehead. The chin, throat, and middle region of the belly are soiled white; the tail and the feet are chestnut brown, the latter mixed with gray. In some of the specimens the sides are brownish gray. Young individuals are said to lack the dark base of the spines. An immature specimen has the entire under side grayish brown with dark-brown feet. Ears covered with yellowish- white hairs on inner side and with pale brown on the outer. Measurements:—No measurements are available except those of Pallas, which, converted into millimeters, are: head and body, 244; tail, 28; ear, 34. Satunin gives the following skull measurements for one of Radde’s specimens from Dauuria: tip of rostrum to infraorbital foramen, 15.5 mm.; length of nasal suture, 14; combined width of nasals, 4; zygomatic width, 36; least interorbital width, 13.9; width outside first molars, 22.5; width of rostrum at level of first incisor, 7; distance between antorbital foramina, 14. THE INSECTIVORES 45 CRANIAL MEASUREMENTS OF HEMIECHINUS Upper Lower Zygo- Width tooth tooth Greatest Basal Palatal matic Mastoid across Tow, Tow, No. length length length width width molars alveoli alveoli Locality H. dauuricus dauuricus (after Lonnberg) 56 51-5 31-5 365 29.3 24 287 — Mongolia H. dauuricus alaschanicus 20683 MCZ 47 45 25 28.6 24 19 24 19 Mongolia Occurrence and Habits:—The range of the long-eared hedgehogs covers an enormous territory from southeastern Europe and Egypt to the eastern parts of Siberia. In all this area a number of species have been recognized, but it is not at all certain that they may not be geographic forms of one or two species. Thus the present animal may be merely a subspecies of Hemiechinus aurttus of Europe, but until this can be definitely shown it may stand as distinct. The long-eared hedgehog of Dauuria evidently has a restricted range in southern Transbaikalia, where it was taken by Radde in the region about the Tarei Nor on the northeastern edge of the Gobi. No doubt it occurs locally across Mongolia, and from his account, either in grass land or more usually in the tree- and bush-grown parts, probably avoiding the open desert. Recently Lonnberg (1922) has recorded under the name Ericius przewalsku three hedge- hogs from Mongolia, two taken by Professor J. G. Andersson at Bank Tsagan and Burtun Nor respectively, and a third from Tabool, received later from the same collector. These he regards as larger than the small pallid race, Hem- iechinus alaschanicus, and refers to Satunin’s species Hemiechinus przewalskit doubtfully. The latter was based on a specimen from an unknown locality, but labeled as from ‘‘?North China,’ collected by Przewalski in 1874, the year in which he procured the types of H. albulus alaschanicus. If it is not synony- mous with the latter, the name is probably best regarded as a synonym of H. dauuricus until some trenchant character can be pointed out. Loénnberg admits that it is ‘“‘hardly possible to tell”’ if it is really different. He gives the following measurements of two of his specimens: Hind foot Total length Tail (without claws) Ear 210 25 41 24 (dry) Tabool 266 31 43 33 The cranial measurements of Lonnberg’s Tabool specimen are reproduced in the table on this page. Specimens examined:—None. 46 THE MAMMALS OF CHINA AND MONGOLIA 7. Hemiechinus dauuricus alaschanicus Satunin Hemiechinus albulus alaschanicus Satunin, Annuaire Mus. Zool. Acad. Imp. Sci. St. Pétersbourg, for 1906, vol. 11, p. 181, 1907. Type specimens:—In his brief description, Satunin lists four specimens: No. 2020, male, from southern Gobi; 2018, 2019, 2021 from Alashan, all collected by Przewalski in 1874, and now in the Museum of the Academy of Sciences at Leningrad. Since no type is specified, all are cotypes. Description:—Similar to H. dauuricus but paler, the sides and under parts usually pure white instead of washed with grayish brown, and the forehead a pale rusty brown; backs of the feet dusky brown, mixed with gray. The spiny covering consists of spines of the usual type in the species, but their long white tips give a very pale effect. There is a slight amount of individual variation in color; youngish examples are whiter, less creamy. Measurements:—Hind foot, with claws, 35 mm., without claws, 30 mm. See table under preceding race for cranial measurements. Occurrence and Habits:—Although described as a form of H. “‘albulus,” there can be no doubt that this is merely a pale desert race of the Dauurian hedgehog (H. dauuricus), from which it chiefly differs in the more pallid colora- tion, with white sides and under parts, lacking the brownish wash. Hedgehogs of this species doubtless occur locally throughout the Gobi, but were secured at only two places by the Central Asiatic Expeditions, namely at Tsagan Nor, where five adults were taken, and at Artsa Bogdo, where on July 18, 1925, a young one only 90 mm. long was found. Howell (1929) mentions four speci- mens in the U. S. National Museum, from northwest of Ningsia, which agree in characters with this form. In his field notes concerning these hedgehogs, Dr. R. C. Andrews writes that in a river bottom at Tsagan Nor, one was captured alive in a trap, unin- jured, and after two days became very tame, allowing itself to be handled freely. It ate grasshoppers and beetles or any other insects voraciously, as well as raw meat, licking its mouth afterward. It drank water once or twice every day, taking a considerable quantity each time. It was possessed of great curiosity, and liked to poke about into corners, investigating carefully any new object. The skin of the back bearing the spines was very loose, except when the animal was frightened, when it stiffened, causing the quills to stand out at all angles. It walked very high on its legs, and entirely on the palms of its feet, holding the toes and nails high off the ground. In running it would go at an astonishing rate, and kept much closer to the ground. At Tsagan Nor the hedgehogs were attracted by some meat thrown into the grass near camp, and close to the water’s edge. Three or four were caught, as well as PLATE VII A tame Long-eared Hedgehog (Hemiechinus dauuricus alaschanicus) at Tsagan Nor, in the Gobi, about to fold up The hay-pile of a Pallas’s Mouse-hare (Ochotona pallasii pallasiz) at Artsa Bogdo, in the Gobi Yea THE INSECTIVORES 47 three young ones that came to traps baited with meat, so they must have been fairly abundant. Specimens examined:—The following seven: Mongolia: Tsagan Nor, 6; Artsa Bogdo, I. Genus Erinaceus Linnzus Erinaceus Linnzus, Syst. Nat., ed. 10, vol. 1, p. 52, 1758. The typical hedgehogs differ from the long-eared group chiefly in rather stouter build, with larger feet and stronger claws, in their shorter ears which do not exceed the adjacent spines, and in having the spines of the occipital region arranged in two clusters with a parting or naked strip between, not always apparent when the spines are spread. The skull has a blunter and less tapering snout; the nasals are extremely narrow, tapering to a fine point behind, while the tip of the ascending process of the premaxillary is broad and ends bluntly instead of in a slender point. The post-glenoid process is less hollowed out behind than in the related group. The teeth are more specialized, in that all the incisors, the canine and the first premolar are usually single-rooted, although occasionally the canine may have a double root. The last molar above is much reduced and usually stands with its long axis in line with the outer cusps of the second molar, instead of nearly transverse to the tooth row. The tooth formula is as in Hemiechinus. Probably two forms only are to be recognized from China, both here con- sidered subspecies of the European animal. 8. Erinaceus europeus dealbatus Swinhoe CHINESE HEDGEHOG Erinaceus dealbatus Swinhoe, Proc. Zool. Soc. London, 1870, pp. 450, 621. Erinaceus collaris Gray, Proc. Zool. Soc. London, 1861, p. 390 (part). Erinaceus europeus dealbatus Barrett-Hamilton, Ann. Mag. Nat. Hist., ser. 7, vol. 5, p. 367, 1900. Erinaceus kreyenbergi Matschie, Wiss. Ergebn. d. Exped. Filchner nach China u. Tibet 1903-05, vol. 10, pt. I, Pp- 135, 138, 1908. Shanghai Market. Erinaceus tschifuensis Matschie, ibid., p. 137. Chefoo, Shantung. Erinaceus hanensis Matschie, ibid., p. 138. Hankow, Hupeh. Erinaceus chinensis Satunin, Annuaire Mus. Zool. Acad. Imp. Sci. St. Pétersbourg, for 1906, vol. 11, p. 173, 1907. Khingan, Tyntza-intza. Erinaceus hught Thomas, Abstract Proc. Zool. Soc. London, December 15, 1908, p. 44; Proc. Zool. Soc. London, for 1908, p. 966, 1909. ; Type specimens:—Swinhoe specified no type specimen in his original description, but Barrett-Hamilton states (1900, p. 367) that it is No. 61.6.2.5, British Museum, from Peiping. The type of E. kreyenbergi was a specimen procured in the Shanghai Market, and is said to be in the Museum at Magde- 48 THE MAMMALS OF CHINA AND MONGOLIA burg, Germany. In recording the specimen, Hilzheimer (1906, p. 184) states that the head was lacking, but Matschie (1908) shows that this is not the case, and he examined the skull with the skin. The type specimen of E. tschifuensis is a skin and skull, No. 4625 in the Berlin Museum, from Chefoo, Shantung, and that of E. hanensis a skin, number not recorded, in the same institution from Hankow, Hupeh. The type of E. chinensis is in the Zoological Museum of the Academy of Sciences at Leningrad. E. hughi, from Paochi, Shensi, is a rather dark example of E. sassaatinagl lacking any conspicuous number of all- white spines. Description:—Similar to the European hedgehog, Erinaceus europeus, but paler in color and somewhat smaller; spines on the head in two groups, one on each side of the occiput, with a narrow bare space between; feet large with prominent claws on all the toes. The spines are of two sorts, some all white or white with a minute brownish point, but the greater part whitish basally with a broad band of light brown, not sharply defined, then an equal one of white, succeeded by a minute brown tip, giving a general brownish gray, pepper- and-salt effect. The face, limbs, sides and lower parts are clothed with coarse hair, rather uniform in color, varying from pinkish buff (as in a specimen from Ichang) to whitish, with a light brown wash on hands and feet. There is, however, much individual variation in color. Thus of two taken by the Central Asiatic Expeditions, one (from Yochow) has rather few dark spines, so that the paler tint predominates, while the other (from Wuhu, Anhwei) has a preponderance of dark spines. Matschie describes the lower parts of a specimen bought in Shanghai as ochraceous. The skull, while closely resembling that of the common European species in general form, is smaller on the average, and much slenderer. The pre- maxillaries, in contrast to the condition in Hemiechinus, are usually almost truncate vertically at their upper ends instead of attenuate. The nasals, as in E. europeus, vary individually from extremely narrow with a combined width of only 1 mm., to twice that width. The teeth, though smaller, are essentially like those of E. europeus, except that the third upper incisor seems slightly larger in proportion; the last upper molar is small and narrow as in the latter, but in some individuals stands more nearly transverse to the tooth row instead of at an angle of some 45° as it does in European specimens. Measurements:—This eastern hedgehog seems to average smaller than full-grown European animals. The following dimensions were taken from fresh specimens by the collector: No. Head and body Tail Hind foot Ear Locality 8.2.8.1 BM 217 42 41 25.5 Shantung 8.2.8.2 BM 215 45 40 26.0 Shantung THE INSECTIVORES 49 CRANIAL MEASUREMENTS OF ERINACEUS FROM CHINA Zygo- Width Upper Lower Greatest Basal Palatal matic Mastoid outside tooth tooth No. length length length width width molars row Tow Locality E. europeus dealbatus 25884 MCz 52.8 49.2 30.0 32.0 19.0 21.0 ey be 26.0 Shantung 25885 MCz 53.5 52.0 30.3 32.2 21.0 20.0 27.2 25.8 Shantung 4625 BERLIN 58.9 56.6 — 359 — 22.1 30.4 —— Shantung 8.2.8.1 BM —_— ss 31.0 —_ 2277, 28.5 26.8 Shantung LONNBERG 51.0 47.0 29.0 31.0 25.0 20.5 27.3 — Hopei 7132 MCZ 49.5 46.5 28.0 30.0 17.6 20.0 27.0 26.6 Hupeh 56508 U. MICH. 56.6 53-7 32.4 34.1 27.9 20.5 28.1 25.7 Kiangsu 56509 U. MICH. 53.5 49.7 30.9 (30) 25-7 19.3 27.6 25.4 Kiangsu E. europeus miodon 9.1.1.2 BM 54.6 50.1 28.5 32.1 26.0 21.1 26.5 23.9 Shensi 9.1.1.4 BM 53.2 48.6 28.5 32.0 25.6 20.5 26.0 24.5 Shensi 9.1.1.5 BM 53-7 49.5 29.3 33:2 25-7 21.6 27.5 25.5 Shensi 9.1.1.6 BM 50.9 47.1 27.0 30.8 26.0 20.3 26.0 24.6 Shensi 9.1.1.7 BM 49.3 44.7 27.0 28.7 24.3 20.5 26.3 24.6 Shensi 9.1.1.8 BM 51.9 47.0 27.4 32.0 25.0 20.6 25.7 23.4 Shensi 9.1.1.9 BM (type) 53.7 50.2 28.7 35-7 27.0 223 27.0 25.2 Shensi g.1.1.10 BM 55-2 50.2 29.1 32.3 25.9 21.5 28.3 25.6 Shensi Nomenclature:—After careful consideration of the various names applied to Chinese hedgehogs of this group, there seems to be little doubt that most of them are synonymous with E. dealbatus, based on variations that are really only individual. Thus Matschie (1908) describes as E. tschifuensis a specimen from Chefoo, Shantung, that differs from a Peiping specimen merely in having a somewhat larger skull, and the forehead and snout a darker brown, nearly “drab.’”’ It is, however, an aged example which would account for its large size, while the color, though darker than usual, is doubtless an individual variation. Thomas (1909, p. 966) in recording two other specimens of E. deal- batus from the same locality, does not regard them as different. The skulls of two other Shantung hedgehogs before me are slightly smaller. In the same paper, Matschie names as new species, E. kreyenbergi and E. hanensis. The former is based on a specimen purchased in the Shanghai Market and is characterized as having the snout, sides of head and the under side ochraceous, the chest brighter (‘‘ockerrot’’), forehead buff, feet dark ochraceous brown; the latter species is described from a skin from Hankow which seems to differ from the Peiping specimen in having the head, feet, and under side dark hair- brown mixed with gray, instead of being pale whitish. Nothing is said of the skulls of either, and both may for the present be best regarded as identical 50 THE MAMMALS OF CHINA AND MONGOLIA with E. dealbatus of which they represent color variations. I had previously (G. M. Allen, 1912, p. 242) referred to E. hanensis an immature male from Ichang, in which the lower surfaces are gray with a decided pinkish wash, but I now regard this as also a color variation of the same, for evidently there is much variety in the exact tint of the hairy coat. Satunin’s Erinaceus chinensis from Khingan, Manchuria (‘‘Tyntza-intza’’), although outside the area covered, may be briefly noticed here. The single specimen upon which it is based is evidently large, but in color does not appear to differ from what is sometimes found in the present form. Indeed, Satunin (1907a, p. 175) admits that he would not be surprised ‘‘wenn mit der Zeit, nach Untersuchung grésseren Materials, dieser Igel als identisch mit Er. deal- batus sich erweist.’’ It is possibly an intermediate toward the larger and darker eastern form E. amurensis. I would also include as a synonym, EF. hughit Thomas, the type of which I examined at the British Museum. It is an unusually dark individual, nearly lacking the all-white spines. Occurrence and Habits:—These hedgehogs occur locally over much of northern China, but seem to be somewhat sporadic in distribution, common in certain areas and rare or unknown in others. In general they are found in the northeastern part of the country as far south at least as the Yangtze basin, and as far west as the borders of the western highlands on the frontiers of Szechwan. Swinhoe (1870b) described the animal on the basis of specimens from Peiping, where he regarded it as common, adding (1870¢c, p. 621) that it is said to occur at Amoy and in Hainan, as well as “‘lately at Swatow,’’ Kwang- tung. Probably, however, it does not occur much to the south of the Yangtze basin, and its supposed presence in Hainan, as perhaps also in Swatow, has never been corroborated and is probably based on reports of porcupines. The evidence for its having been found at Amoy rests on the statement of Swinhoe (1864a) that hedgehogs, said to have been locally obtained, have been offered for sale in the market there. Probably, however, these captives were from much farther north, while the basis of the Swatow record is not further stated. The predilection of the Chinese for keeping pets is well known, so that little weight can be given to locality records based on animals purchased in the markets. Mell, who spent several years collecting in Kwangtung, makes no mention of these animals in the list of the mammals he found. The hedgehog is apparently commonest in the northern part of China, where it seems to thrive in spite of man and his works. In addition to Swinhoe’s record of its being common about Peiping (Hopei), Lonnberg (1922) mentions specimens from the same province (Miyuanhsien, Shunihsien, and Niulangshan), and Jacobi (1922, p. 2) records other specimens from Peiping, collected by the Stétzner THE INSECTIVORES 51 Expedition, while Dr. R. C. Andrews secured one at Eastern Tombs, to the northeast. Still farther east, Howell (1929, p. 6) records two from Tientsin, Hopei. This hedgehog seems to be still not uncommon in Shantung, despite the long period of intensive cultivation and denudation. In addition to the large individual from Chefoo, made by Matschie the type of his E. tschifuensis (in the Berlin Museum), Thomas (1908d, p. 6) records a male and a female from the same locality, adding the note of Malcolm P. Anderson, the collector, that these were purchased alive from peasants who had brought them in; he also mentions (1909, p. 966) one received from Swinhoe. Anderson says that, although he failed to secure any himself, in the course of some two months’ collecting, they were apparently ‘‘not uncommon” and strictly nocturnal. In the same province, Dr. A. Jacot has sent me the skull and a skeleton of this animal from Tsinan, where they were taken on the campus of Shantung Christian University. He writes: ‘‘We are in the center of a semiarid region that has been under intensive cultivation for 2000-4000 years, and the Chinese have long ago eliminated anything at all edible. There are no woodland tracts in this Province to my knowledge. The hills have been deforested and are grubbed monthly of shrubs for fuel. Thus only the most adaptable ani- mals have survived.” Farther south, Sowerby (1929c) who has had long experience in China, says that hedgehogs are found at least as far as Chekiang Province, coastwise; probably Matschie’s type specimen of EF. kreyenbergi: (in the Magdeburg Museum) purchased in the Shanghai Market, came from no great distance. Howell (1929) lists (under Erinaceus hanensis) two specimens from Shanghai in the U. S. National Museum, one from Ningpo, Chekiang, and others from Yochow, Hunan, whence also the American Museum has a specimen, and in addition one from Wuhu, Anhwei; these and the type of E. hanensis Matschie from Hankow, eastern Hupeh, and the specimen in the Museum of Compara- tive Zodlogy from the western part of the same province, at Ichang, very likely indicate roughly the southern range. The northwestward limits are perhaps in central Shansi. M. P. Anderson (in Thomas, 1909) states that it is unknown about Paotehchow, at the edge of the Ordos Desert, but he had reports of it at Ningwufu in the central parts of the province. In the more arid Shensi Province it doubtless grades into the following subspecies, E. e. miodon, if that proves eventually to be distinct. Sowerby (1914, p. 56), in brief notes on the Chinese hedgehogs, writes that in Hopei they are looked upon as sacred animals by the Chinese and so are not molested, but on the contrary, little shrines are often built for them. This may in part account for their comparative abundance in parts of this province, for elsewhere they are often eaten, and Sowerby recounts that in 52 THE MAMMALS OF CHINA AND MONGOLIA Manchuria the woodsmen prepare them for eating by first encasing them in a coating of mud, which, after the animal has been roasted whole in the embers of a wood fire, comes away with the spines, hair, and skin adhering to the clay, “leaving a very toothsome morsel of beautifully cooked meat.” He adds that foxes often kill hedgehogs by thrusting their snout under the spiny ball, and throwing the animal into the air. ‘This makes the hedgehog uncurl, and, before it can curl up again, the fox has nipped it in the unprotected vitals.’ Specimens examined:—In all, sixteen, as follows: Hopei: Eastern Tombs, 1; Peiping, 1 (type, B.M.); northeastern Hopei, 1 (B.M.). Hunan: Yochow, I. Anhwei: Wuhu, I. Hupeh: Ichang, 1 (M.C.Z.). Kiangsu: Nanking, 3 (Univ. Mich.); Shanghai, 1 (B.M., topotype of E. kreyenbergi). Shantung: Tsinan, 2 (M.C.Z.), 1 (B.M.); Chefoo, 2 (B.M.). ; Shensi: Paochi, 1 (B.M., type of E. hught). g. Erinaceus europeus miodon Thomas ~ Erinaceus miodon Thomas, Abstract Proc. Zool. Soc. London, Dec. 15, 1908, p. 44; Proc. Zool. Soc. London, for 1908, p. 965, 1909. Hemuechinus miodon Lénnberg, Ann. Mag. Nat. Hist., ser. 9, vol. 9, p. 626, 1922. Type specimen:—An adult male, skin and skull, No. 9.1.1.9, British Museum, from Yulinfu, Shensi, China, 4,000 feet altitude. Collected by Malcolm P. Anderson, in May, 1908. Description:—In size and general proportions resembling EF. europaeus dealbatus, but the spiny coat almost wholly lacks the all-white spines generally present in the latter. The spines are about 22-24 mm. long on the back, white for two-thirds their basal portion, then broadly ringed with blackish brown, the ring some 4 mm. wide, tipped with white for about an equal breadth, or sometimes the extreme tip minutely dark. The hairy coat of the head, sides, limbs, and tail is variable, as in E. e. dealbatus, in some a dull whitish or brownish white to distinctly brown (“broccoli brown’’). The belly is usually somewhat paler, varying to dull white. The type series is apparently in winter pelage still or in process of changing, so that the paler parts may represent the old fur of the latter. The skull differs from that of more typical E. e. dealbatus in having (in all but two of the nine specimens) the tip of the premaxillary process more slender and continued back to nearly or quite meet the anterior point of the frontal, completely shutting off the maxillary from contact with the nasals. Measurements:—The dimensions do not seem to be essentially different from those of E. e. dealbatus, although but few for the latter are available. The collector’s measurements for the type series are as follows: THE INSECTIVORES 53 : No. Head and body Tail Hind foot Ear Locality 9.1.1.2 BM 205 35 39 28.0 Shensi 9.1.1.3 BM 176 34 36 24.0 Shensi 9.1.1.4 BM 195 35 36 27.0 Shensi 9.1.1.5 BM 205 37 36 33.0 Shensi 9.1.1.6 BM 175 37 38 29.0 Shensi 9.1.1.7 BM 175 31 36 29.0 Shensi 9.1.1.8 BM 192 40 35 28.0 Shensi 9.1.1.9 BM (type) 215 46 40 34.5 Shensi g.1.1.10 BM 214 43 37 30.0 Shensi For cranial measurements of this series, see table under E. e. dealbatus. Nomenclature:—This hedgehog at first sight does not seem very different from some specimens of E. e. dealbatus, but the entire series is uniform in lacking the usual intermixture of all-white spines, common in that race, and the attenuated premaxillary extending rather far back is also different. Loénn- berg (1922) concluded that E. miodon is really one of the Hemiechinus group, but after seeing the original specimens in the British Museum, I agree with Thomas that its relationships are after all with Erinaceus e. dealbatus, and that it constitutes a local race of the desert country bordering the Ordos. Occurrence and Habits:—Hitherto this hedgehog is recorded from the type locality only, Yulinfu, in northwestern Shensi, close to the border of the Ordos Desert, and hence not far from the area inhabited by Hemiechinus dealbatus alaschanicus. Clark and Sowerby (1912, p. 22) write that ‘‘the country about Yu-lin Fu is wild and inexpressibly dreary. Very few trees are to be seen, and the bare brown cliffs and yellow sand are devoid of any vegetation, save an occasional tuft of some sage scrub. In places, especially where, as in the northeast, it rises to any prominence, gloomy chasms, with deadly quicksands lurking in their depths, gape in the sandstone and the half-formed shale. To north and west the prospect is heart-breaking. Sand-dunes and sand-dunes, and again sand-dunes—shifting with every storm and obliterating every landmark. Only here and there, as tiny islands in a sea of desolation, small clusters of mud huts, where some little oasis marks the site of a spring or well.’’ Malcolm P. Anderson, who collected the series of nine upon which the form was described, notes (Thomas, 1909, pp. 964, 966) that ‘‘the portion of Shen-si visited appears indeed like an extension of the plateau of which Ordos is part, only this extension has been cut into by a great many perennial streams, a process which is now taking place in southern Ordos.” Of the hedgehogs, he adds, “There appear to be large areas in North China where the Hedgehog is not found at all, and some places, of which the neighbourhood of Yu-lin-fu is one, where they are remarkably common. At the time we were at Yu-lin 54 THE MAMMALS OF CHINA AND MONGOLIA (April to May) the neighbouring desert was alive with several species of beetle upon which the Hedgehog fed. . . . Chinese name, “Tsi-wei’ (tsi- a thorn or spine).’’ Sowerby (Clark and Sowerby, 1912, p. 83), who visited the type locality shortly after, was unsuccessful in securing additional specimens, for on account of the lateness of the season (last of October), he found this and many other small mammals already gone into hibernation. Specimens examined:—Nine, the original series in the British Museum, from Yulinfu, northern Shensi. Family TALPIDA MOLES This family includes the mole-like insectivores, chiefly modified for a fossorial life through the specialization of the fore feet and anterior part of the body for pushing a way through loose soil. The head is tapering, the external ears small or absent, the neck short and muscular. The muscles of shoulder and arm are enlarged and powerful, their bones short and strong, with the clavicle and humerus especially well developed, the latter, however, retaining its entepicondylar foramen. All five fingers are present on the fore feet, each provided with a stout claw, and in the more strictly fossorial types, the entire hand is so rotated that the foot seems to be set on edge, with the palm turned out. The skull lacks prominent ridges for muscle attachment, but its com- ponent bones fuse at an early age. The zygomatic arch is still present though slender, and the teeth retain a primitive sectorial type, with sharp cusps on the upper molars tending to form a W-pattern. The tympanic bone fuses to the skull, forming a low, rounded bulla. Thomas (1912d), in a discussion of the family, regards it as divisible into five subfamilies, of which the Desmanine and the Condylurine are aquatic in habits, the former confined to western Siberia and southern Europe, the latter to eastern North America. Representatives of the three remaining sub- families occur in China, the most primitive, the Uropsiline, including species of the Chinese highlands that retain an almost shrew-like form, without special fossorial modifications; the two other groups, the Talpinz or more typical moles, and the Scalopinz with one Chinese genus and other American forms, are modified for underground life. Key TO THE GENERA OF CHINESE TALPIDAE A. Form shrew-like, with long snout, slender tail, feet not modified for digging; 2 anterior upper incisors much larger than the canine and anterior premolars immediately following. a. Nine teeth in the upper jaw, eight in the lower...................- Uropsilus b. Ten upper and nine lower teeth. a’. With two lower incisors and three lower premolars............ Rhynchonax b’: With only one lower incisor and four lower premolars.......... Nasillus THE INSECTIVORES 55 B. Form thickset for burrowing, snout shorter, tail about twice the length of hind foot or much less; fore feet broadened and the fore claws lengthened for digging. a. Anterior incisors small, followed by an enlarged canine; tail about as long as hind foot. a lever CecuImieaCh AW astctn tartater mie nit rstve riers tres ral Talpa b’. Teeth less than eleven in each jaw. 1. Upper premolars three, two small and one large; lower four. Parascaptor 2. Both upper and lower premolars three only.............. Scaptochirus 3. Teeth as in Talpa but lower canine lacking, making the fourth tooth close behind the upper canine................. Mogera b. Anterior incisors larger than the succeeding teeth, tail twice as long as the hind foot. a’. Upper teeth eleven, lower ten in number, fore feet less broadened. Scaptonyx b’. Upper teeth nine, lower nine in number, fore feet broader...... Scapanulus Genus Uropsilus Milne-Edwards Uropsilus Milne-Edwards, in David, Nouy. Arch. Mus. d’Hist. Nat. Paris, vol. 7, Bull., p. 92, 1871. Milne Edwards, Recherches pour servir a l’Hist. Nat. des Mammiféres, p. 272, 1868-74. The insectivores of this group are of special interest as representing the most primitive of the living mole-like species, in which the bodily form is still essentially shrew-like, with none of the characteristic adaptations of the burrowing types. The snout is very long, with a cartilaginous tubular pro- longation beyond the fore part of the skull; the external ears are well developed and reach the height of the surrounding fur; the tail is long, slender and covered with rings of small scales, while the fore feet are provided with slender toes, each with a compressed instead of a flattened claw. The backs of both hands and feet are scaly, and according to Milne-Edwards the bones of the ungual phalanges are entire instead of bifurcate as in the typical moles. The skull with its rounded outlines and complete zygomatic arch is essen- tially mole-like, however, and the teeth resemble those of most moles in the enlargement of the first incisors and reduction of the succeeding teeth, without the extreme lengthening of the first lower incisors, found in the soricids. In the type species, U. soricipes, described by Milne-Edwards from Muping, Szechwan, the tooth formula is given as of nine upper and eight lower teeth on each side, as follows: i.¢ c.t pm. m.3=34. These are interpreted by Thomas (1912e, p. 129) as representing the following teeth: 1.>2 c.t pm.}+49-++- m.+2+ There seems, however, to be a curious variation in the number of the small premolars and lower incisors that may be present, variations which are apparently of a definite sort, although accompanied by little if any external difference. The commoner of these seems to be the presence of four upper premolars associated with two lower incisors, while the less usual one is the presence of four upper and four lower premolars, but in other respects the formule are as given above. Thomas has regarded these 56 THE MAMMALS OF CHINA AND MONGOLIA as indicating two additional genera, which he names Rhynchonax and Nasillus respectively. It is still uncertain whether they express merely fluctuating conditions in the presence or absence of nearly functionless teeth in process of disappearance in a single generic type, or whether they are correlated with other differences that would distinguish related groups, which, though per- haps occurring together in the same general area, nevertheless do not intercross and so are to be regarded as distinct. A conservative course might be either Fic. 3. Distribution Map. Uropsilus Rhynchonax Nasillus 1. U. soricipes 2. R. andersoni andersont 5. WN. gractlis 3. R. andersoni nivatus 6. WN. investigator 4. . andersoni atronates THE INSECTIVORES 57 to consider all three described genera as one, with definite variations among the disappearing teeth of a dentition in course of reduction, or to recognize all three at present as representing truly distinct groups of a primitive stock, until sufficient evidence is obtained to warrant a definite conclusion. The second alternative is here accepted, and the three genera treated as valid until they can be shown to be otherwise. The fact that, so far as the available specimens indicate, these three groups seem to have fairly distinct areas of geographical distribution, may be interpreted as favoring their distinctness. 10. Uropsilus soricipes Milne-Edwards Uropsilus soricipes Milne-Edwards, in David, Nouv. Arch. Mus. d’Hist. Nat. Paris, vol. 7, Bull., p. 92, 1871. Milne-Edwards, Recherches pour servir 4 l’Hist. Nat. des Mammiféres, p. 272, pl. 40, fig. 1; pl. 40A, fig. 1, 1868-74. Type specimen:—No type specimen is mentioned in the original descrip- tion, which, however, was evidently based on an individual sent in alcohol to the Muséum d’Histoire Naturelle at Paris by Pére Armand David, who col- lected it in the principality of Muping, Szechwan, China. Description:—Upper surface of head and body dark brown, near ‘‘Prout’s brown”’ of Ridgway; below dark slaty. Tail and backs of feet dark brown, scaly, with minute blackish hairs growing from between the scales. The skull has a full, rounded brain case, slender up-curved zygomatic arches, and a tapering rostrum which is slightly depressed anterior to the orbits. In side view, as seen in Milne-Edwards’s Plate 40A, the first incisor of each side above is only slightly larger than the second, but set with its broad axis transverse to the long axis of the skull so that the two of opposite sides together form a sharp cutting edge. The two teeth following the second incisor are hardly one-third the size of that tooth, single-rooted, and interpreted as a canine and premolar 1, short and bluntly conical. The two other premolars are larger, practically in contact, and the fourth exceeds the one in front of it, interpreted as pm*. The two anterior molars have the usual cusps well developed, forming with their commissures a W-pattern, which in the third molar is reduced through the loss of the posterior commissure of the metacone and the suppression of the hypocone. In the lower jaw the single large incisor corresponds to the upper, and is succeeded by three unicuspids, increasing slightly in size from first to last (regarded as a canine and two premolars), while the third premolar is larger, and in close contact with that in front. Concerning the skeleton, Milne-Edwards states that the vertebree number seven cervicals, thirteen dorsals, seven lumbars, five sacrals, and fourteen caudals, total 46. The manubrium of the sternum is laterally compressed below, but does not show the definite keel characteristic of the typical moles, and the hand lacks the falciform bone. The clavicles are relatively weak, but the 58 THE MAMMALS OF CHINA AND MONGOLIA humerus, though slender and shrew-like, has nevertheless prominent muscle crests. The terminal phalanges of the hand are normal in form without the cleft appearance found in the moles, in which the heavy claws require this extra means of support. . Measurements:—The measurements of the type specimen and the two others secured by the Duke of Bedford’s Expedition are as follows: No. Head and body Tail Hind foot Ear Locality PARIS 63 64 15 — Szechwan I1.9.8.12 BM 72 65 15 10 Szechwan CRANIAL MEASUREMENTS OF UROPSILUS Zygo- Width Upper Lower Greatest Basal Palatal matic Mastoid outside cheek cheek No. length length length width width molars teeth teeth Locality 11.9.8.11 BM 105° 10.9% "=—— "80 ° 9.7 ‘9.0 Szechwan II.9.8.12 BM 21.2 17.5 10.5 II. 11.6 7.0 9.6 9.1 Szechwan PARIS (type) 21.0 —_>= ——_—-—— 11.0 —_—_-— —- Szechwan Occurrence. and Habits:—This remarkable species, a sort of annectant type between the moles and the more primitive insectivores from which they must have sprung, was first discovered by Pére David, who in the course of his exploration in western Szechwan, in 1870, sent back specimens to the Paris Museum, which were later described and figured by Milne-Edwards. No further record of the animal occurs for some thirty years, until Pousargues (1896a, p. 179) recorded a specimen from Yunnan, brought back by Prince Henri d’Orléans. However, the later discovery of the very similar genus Rhynchonax as a common species in that province raises the presumption that in the light of more recent knowledge, this specimen might prove to be a mem- ber of the latter genus. It was not until 1912 that Thomas (19I2e, p. 129) discovered the fact that typical Uropsilus, as shown by examination of examples from Pére David’s original series, has a slightly different tooth formula from the one more common in specimens he had hitherto referred to this genus, leading him to reéxamine the entire lot acquired shortly before by the Duke of Bedford’s exploration under Malcolm P. Anderson. He showed that Uropsilus soricipes, characterized by the possession of nine upper and eight lower teeth, is known only from the original series from Muping and from two males having the same formula, taken by Anderson only a short distance to the northeast of the original locality, at Weichow, sixty miles northwest of Chengtu, Szechwan, in the narrow valley of the Sungpan Ho. Apparently, then, this variation, in which the small upper premolar, between the two large ones, and the second lower incisor are lacking, is confined to the region of Muping and slightly to the east, while to the southeast and southwest two other variations occur, each of which Thomas has regarded as typifying distinct genera, Nasillus and THE INSECTIVORES 59 Rhynchonax, respectively. In a previous paper (G. M. Allen, 1912), I re- garded all these variations as merely individual and indicative of a tooth formula in active process of reduction through the variable presence or absence of one or two of the minute teeth in one or both jaws. The fact that these variations seem in some degree correlated with an area of distribution, how- ever, lends color to Thomas’s view that they represent distinct genera, but the external similarity of the animals points to their essential specific unity. Awaiting further knowledge, therefore, it seems best to let Thomas’s genera stand provisionally as a group of closely related forms, of which typical Urop- stlus soricipes is the most northeastern in distribution, confined to a narrow area in Szechwan. Specimens examined:—Two, from Weichow, Szechwan (B.M.). Genus Rhynchonax Thomas Rhynchonax Thomas, Proc. Zool. Soc. London, 1912, p. 129. Type, Rhynchonax andersoni Thomas. Uropsilus Thomas, Proc. Zool. Soc. London, 1911, p. 163. G. M. Allen, Mem. Mus. Comp. Zool., vol. 40, p. 239, 1912 (in part). This genus chiefly differs from other members of the subfamily in the re- tention, usually, of a minute upper premolar (pm‘) and a lower incisor (i;), in addition to the number present in Uropsilus, making in all ten upper and nine lower teeth. In external characters the two are alike. As mentioned under Uropsilus, it is a question whether the presence or absence of the extra minute tooth above and below is to be regarded as sufficient basis in this case for a generic separation. In 1912, I inclined to the view that it was not, especially since a series from Wa Shan, somewhat to the southward of Muping, showed an additional variation in sometimes lacking the minute upper premolar, although with nine lower teeth. Thomas, who later examined one of the specimens, wrote me that he regarded them all, nevertheless, as Rhynchonax. Thus it appears that the latter may sometimes lose the upper minute premolar (Thomas writes that its alveolus can be made out in the specimen sent him from our series), while still retaining the extra lower one; or, as in some of the Wa Shan series, the reduction may go so far that not only is there no alveolus to be made out, but the space where the tooth stood is nearly obliterated by the close approximation of the two large premolars. Thomas believes further that the color of the two genera is slightly but recog- nizably different. 11. Rhynchonax andersoni andersoni Thomas Rhynchonax andersoni Thomas, Abstract Proc. Zool. Soc. London, October 31, 1911, p. 49; Proc. Zool. Soc: London, 1912, p. 130. Uropsilus soricipes Thomas, Proc. Zool. Soc. London, 1911, p. 163. G. M. Allen, Mem. Mus. Comp. Zodl., vol. 40, p. 239, 1912 (not Milne-Edwards). ? Pousargues, Bull. Mus. d’Hist. Nat., Paris, vol. 2, p. 179, 1896 (? in part). 60 THE MAMMALS OF CHINA AND MONGOLIA Type specimen:—Adult male, skin and skull, No. 11.2.1.25, British Mu- seum, from Omei Shan, Omei Hsien, southern Szechwan, 9,500 feet. Description:—General color above a dark brown, near ‘‘clove-brown’’ to “‘bister’’ (Thomas); below dark slaty. Tail dark brown all around, with rings of minute scales, between which are very small bristles of about the length of a single ring, except toward the tip, where they are longer, tending to form a short pencil. As already noted, the tooth formula is normally greater by one upper pre- molar and one lower incisor than that of Uropsilus, and the color is darker. Measurements:—The following are measurements of the British Museum series, those of the exterior made in the field by the collector. No. Head and body Tail Hind foot Locality II.2.1.26 BM 67 65 14.5 Szechwan II.2.1.27 BM 70 67 15.0 Szechwan 1I.2.1.28 BM 71 65 15.0 Szechwan 11.2.1.29 BM 67 68 15.5 Szechwan II.2.1.31 BM 69 68 15.5 Szechwan I1.2.1.25 BM (type) 70 67 15.5 Szechwan CRANIAL MEASUREMENTS OF RHYNCHONAX Zygo- Width Upper Lower Greatest Basal Palatal matic Mastoid across - cheek cheek No. length length length width width molars teeth teeth Locality R. andersoni andersonti II.2.1.25 BM 21.0 17.3 10.1 II. 11.5 6.6 9.2 8.6 Szechwan II.2.1.26 BM 21.7 16.7 9.8 10.6 11.5 6.3 9.4 8.6 Szechwan I1.2.1.30 BM 22.1 17.3 10.3 10.4 11.5 6.8 9.7 9.1 Szechwan II.2.1.31 BM 21.6 17.3 9.9 10.4 11.6 6.8 9.6 9.4 Szechwan 7520 MCZ 21.7 17.1 10.5 10.5 11.2 6.6 9.6 9.0 Szechwan 7521 MCz 21.1 16.7 9.8 10.5 11.5 6.1 9.5 8.6 Szechwan 7523 MCZ 20.4 16.6 9.8 9.9 11.2 6.5 9.3 8.6 Szechwan 7524 MCZ 22.0 17.5 10.2 II. D7, 6.8 9.8 8.6 Szechwan 7527 MCZ 22.0 17.5 10.2 10.4 11.4 6.7 9.7 9.0 Szechwan R. andersoni atronates 44344 20.0 16.6 9.4 10.5 5.8 8.7 8.2 Yunnan 44340 20.0 16.6 9.2 — 10.9 6.3 8.5 7.8 Yunnan R. andersoni nivatus 44361 20.3 16.2 9.8 11.3 6.2 8.9 8.0 Yunnan 44364 19.9 16.0 9.3 9.7 10.5 6.0 9.2 8.0 Yunnan Occurrence and Habits:—So far as at present known, this seems to be a slightly more southern animal than Uropsilus in its general distribution. The type and eight others were taken by Malcolm P. Anderson at Omei Shan, Omei Hsien, in central Szechwan. In addition, a small series of seven was secured for the Museum of Comparative Zodlogy by Walter R. Zappey, in 1908, at THE INSECTIVORES 61 Wa Shan, slightly south of the same place, and a single one each from Liang- hokow and Tachiao in the same general area (not Taochow, Kansu, as Howell, 1929, p. 7, suggests). Probably the animal occurs in typical form over central Szechwan into northern Yunnan, but there seem to be no other authen- tic records. Specimens secured by the American Museum Asiatic Expeditions in southern Yunnan are sufficiently different to warrant separation as sub- species. Specimens examined:—In all, eighteen, as follows: Szechwan: Lianghokow, 1 (B.M.); Omei Shan, 9 (B.M.); Tachiao, 1 (U.S.N.M.); Wa Shan, 7 (M.C.Z.). 12. Rhynchonax andersoni atronates G. M. Allen Rhynchonax andersoni atronates G. M. Allen, Amer. Mus. Novitates, no. 100, p. 2, December 28, 1923. Type specimen:—A female, skin and skull, No. 44343, American Museum of Natural History, from Mucheng, Salween drainage, southwestern Yunnan, China, altitude 7,000 feet. Collected February 13, 1917, by Dr. R. C. An- drews and Edmund Heller.. Description:—A dark form, the rump nearly unmixed slaty black; skull smaller, with the third upper premolar less reduced than in typical R. andersont. General color above nearly ‘‘Prout’s brown” (of Ridgway). The pelage consists of shining black hairs, mixed with others that are blackish slate basally, tipped with hazel. On the rump the latter hairs are few or absent, giving a strikingly blackish appearance to this region. The lower surfaces of body and limbs are uniform blackish slate. Backs of feet and entire tail scaly, with minute scattered blackish hairs; the tail usually not paler underneath. The skull is smaller than in typical R. andersoni of Szechwan, and the teeth are smaller throughout, except that the third upper premolar, which in the latter is minute or sometimes wanting altogether (with about half the crown area of the first and barely reaching the level of the cingulum of the two adjoining it), is in this Yunnan animal much larger, of about the same crown area as the first premolar, with cingulum and crown well developed, the tip of the tooth standing well above the general cingulum level. In the lower jaw, the second incisor, instead of being minute, is as large as the canine, and the small second premolar (p:), though slightly smaller than the canine, is never- theless much better developed than in typical R. andersoni, in which it is very minute or sometimes altogether absent. Measurements:—The external measurements are practically the same as in R. andersoni. The type measures: head and body, 67 mm.; tail, 57; hind foot, 14; ear, 10. For skull measurements, see table under R. andersoni andersoni. 62 THE MAMMALS OF CHINA AND MONGOLIA Occurrence and Habits:—Specimens of this genus were obtained at only two localities in southwestern Yunnan by the American Museum Asiatic Expeditions, at Mucheng on the Salween drainage (sixteen), and again at Peitaiping, where a single one was secured. The slight differences in color are of less importance probably than the less reduced condition of the teeth in this more southern animal. Specimens examined:—Seventeen, as follows: Yunnan: Mucheng, 6,000-7,000 feet, 16; Mekong drainage, Peitaiping, 9,000 feet, 1 (in alcohol). 13. Rhynchonax andersoni nivatus G. M. Allen Rhynchonax andersoni nivatus G. M. Allen, Amer. Mus. Novitates, no. 100, p. 2, December 28, 1923. Type specimen:—Male, skin and skull, No. 44352, American Museum of Natural History, from Ssu Shan (Snow Mountain), Likiang Range, western Yunnan, China, at 12,000 feet altitude. Collected October 22, 1916, by Dr. R. C. Andrews and Mr. Edmund Heller. Description:—Similar to typical R. andersoni. but much paler brownish, with a smaller skull and larger third upper premolar. In size and cranial characters, this subspecies resembles the preceding but is much paler brown above, nearly light cinnamon brown, almost of the same tint as our Sorex cinereus; it lacks the blackish rump of R. andersoni atronates, and the tail is indistinctly bicolor. The fore legs and under parts are ‘‘deep neutral gray” (Ridgway), much paler than in R. andersoni from Szechwan. In certain lights the tips of the hairs appear glistening. Tail fuscous above, paler below. The skull is smaller than in the typical form and closely resembles that of R. andersoni atronates. The third upper small premolar is as large as the first, - stands well in the tooth row with its cingulum level with those of the adjacent teeth, and has a distinct crown about as high as the width of the cingulum. The second lower incisor is larger than the lower canine, and hence much larger than in the typical form. Measurements:—The type measured in the flesh: head and body, 68 mm.; tail, 60; hind foot, 15. For cranial measurements, see table (p. 60) under R. a. andersoni. Occurrence and Habits:—The discovery of this animal on the isolated Likiang Range at altitudes of from 10,000 to 12,000 feet by Dr. Andrews’s expedition extends the known range of the genus considerably to the southwest. No doubt intergradation with the preceding race takes place at lower levels to the northward, but as with many other species this seems to have developed a local form on this range cut off in the loop of the Yangtze. While resembling THE INSECTIVORES 63 R. a. atronates in cranial and tooth characters, the color is so strikingly different that the Likiang specimens can be picked out at once from a mixed series. The large size of the third upper premolar and the second lower incisor in these southern races, as contrasted with their reduced size in the typical animal to the northward, lends color to the argument that we have in these and Uropst- lus a progressive series in reduction of these teeth, culminating at the north- ward limits of the range in their loss, in what is here called Uropsilus soricipes; so that it may be that we have to do really with but this single species, of which the forms of Rhynchonax may be regarded as subspecies. It is peculiar that all the specimens taken were males. Specimens examined:—Nine, from Yunnan: Likiang Range, Ssu Shan (Snow Mountain), 10,000—12,000 feet. Genus Nasillus Thomas Nasillus Thomas, Abstract Proc. Zool. Soc. London, October 31, 1911, p. 49; Proc. Zool. Soc. London, 1912, p- 129. While having the same number of teeth, ten above and nine below, as Rhynchonax, the formula differs in that but one instead of two lower incisors is present, with all four lower premolars instead of only three. Thomas fas interpreted the teeth a§ follows: 1.4-2-4- ct pm. ame OS or 38inall. In general appearance this genus externally is similar to Uropsilus but slightly smaller. It may again seem possible that the variation in tooth formula is not a matter of generic value, but Thomas, after careful study of the three types of formule, has reached the conclusion that in the present case a separate genus should be recognized, with Nasillus gracilis as its type. 14. Nasillus gracilis Thomas Nasillus gracilis Thomas, Abstract Proc. Zool. Soc. London, October 31, 1911, p. 49; Proc. Zool. Soc. London, 1912, p. 130. Type specimen:—A female, skin and skull, No. 11.9.1.13, British Museum, from Chinfu Shan, near Nanchwan, southeastern Szechwan, China. Description:—Externally resembling Uropsilus and Rhynchonax, the general color above is near “‘sepia,’’ much as in Uropsilus soricipes, not so dark as in Rhynchonax andersoni; below slaty. Hands and feet pale brown, the tail uniformly brown. The skull is shorter and decidedly narrower than in either of the two related genera, with a less-expanded brain case. The upper third premolar (absent in Uropsilus and minute in Rhynchonax andersoni) has the crown as large as the small anterior premolar; in the lower jaw the minute incisor stand- 64 THE MAMMALS OF CHINA AND MONGOLIA ing just back of the large anterior incisor in the latter species, is quite lacking, but there is a very minute second lower premolar, the smallest tooth in the jaw. Hence, although the same number of teeth is present as in Rhynchonax, their interpretation is different. Measurements:—Thomas gives the following measurements of the type and only known specimen, and points out that the hind foot is proportionately smaller than in the related genera Uropsilus and Rhynchonax: head and body, 66 mm.; tail, 55; hind foot, 13.5; ear, 9. For cranial measurements see the table under JN. investigator. Occurrence and Habits:—This species is known only from the single speci- men taken at Chinfu Shan, near Nanchwan, in southeastern Szechwan, which is apparently the most eastern point known for the occurrence of a member of this group. The altitude is about 4,000 feet. Specimens examined:—One, the type, from Nanchwan, Szechwan (B.M.). 15. Nasillus investigator Thomas Nasillus investigator Thomas, Ann. Mag. Nat. Hist., ser. 9, vol. 10, p. 393, 1922. Type specimen:—Skin and skull, No. 22.9.1.16, British Museum, from the Kiukiang-Salween divide, at 28° north, northwestern Yunnan, China. Collected July 24, 1921, by George Forrest. Description:—Externally similar to N. gracilis, ‘indeed, all the members of the three genera Uropsilus, Rhynchonax and Nasillus are hardly distinguish- able from each other’ (Thomas), externally, except that the members of Rhynchonax are noticeably more blackish. The skull differs from that of N. gracilis in being longer with a wider brain case. The skull of the type, though still retaining milk teeth, is said by its describer to be quite of full size, and in its tooth formula agrees with that of N. gracilis, as do also the five other specimens taken at the same time and place. Measurements:—The original series and another from Gomba-la were measured by the collector as follows: No. Head and body Tail Hind foot Ear Locality 22.9.1.13 BM 80 54 16 9 Yunnan 22.9.1.14 BM 83 59 14 10 Yunnan 22.9.1.15 BM 80 66 15 8 Yunnan 22.9.1.16 BM (type) 88 62 14 10 Yunnan 22.9.1.17 BM 74. 74 14 9 Yunnan 22.9.1.18 BM 78 75 15 9 Yunnan 22.9.1.19 BM 67 64 15 10 Yunnan 23.3.7.7 BM 70 72 13 10 Yunnan THE INSECTIVORES 65 CRANIAL MEASUREMENTS OF NASILLUS Great- Zygo- Mas- Width Upper Lower est Basal Palatal matic toid across tooth tooth No. length length length width width molars row tow Locality N. investigator 22.9.1.15 BM 2) 9.9 — 67 9.7 90 Yunnan 22.9.1.16 BM (type) 21.3 — 06) 102) ) 11-1 6.59 (O25 8.6 Yunnan 22.9.1.17 BM —_s> 9.7 m— — 66 94 4x88 Yunnan 22.9.1.18 BM 2i-5, 17.4. 10.1 QO 25.5) 6.8) O78) o.9)) Munnan 22.9.1.19 BM 21.8 — 9.7 10.8 D163) 6:5) 10:27) 4316.) Yunnan N. gracilis 11.9.8.13 BM (type) 20-31) 16:3 9.2 96 104 62 89 18.1 Szechwan Occurrence and Habits:—Seven specimens in all were obtained by Forrest at or near the type locality, all of which agree in tooth formula, thus addi- tionally confirming Thomas’s view of the distinctness of the genus. The locality is at an altitude of 11,000 feet, on the Kiukiang-Salween divide at about latitude 28° N. The same collector later secured one at Gomba-la, on the Mekong-Salween divide. Nothing further is known of it, but undoubtedly it will eventually prove to be at most a slightly larger subspecies of N. gracilis, of which it is obviously a close relative. The luck of-collecting appears again in this case where Forrest secured six specimens, while Andrews and Heller working in the same general area secured numbers of Rhynchonax instead. From the brief notes entered on the labels of the original series, it appears that all were trapped on open alpine meadows at altitudes of from 11,000- 14,000 feet, except one which was caught in Abzes forest at a similar height. It is thus apparently a high-alpine species. Specimens examined:—In all, eight, namely: Yunnan: Salween-Kiukiang divide, 7 (B.M.); Mekong-Salween divide, 1 (B.M.). Genus Scaptonyx Milne-Edwards Scaptonyx Milne-Edwards, Recherches pour servir 4 1’Hist. Nat. des Mammiféres, p. 278, pl. 38B, fig. 4; pl. 40B, fig. 2, 1868-74. This is a somewhat more mole-like member of the Uropsiline than Uropsilus, with a long cartilaginous snout, slightly broadened hands with large, nearly straight claws, of which the first and fifth are much shorter than the three middle ones. There is practically no external ear and the tail is shorter and more thickened than in that genus, thinly clad with stiff short hairs that do not conceal the rings of scales. It resembles the Japanese Urotrichus and the western American Neurotrichus, to both of which it is related. In the skeleton, Milne-Edwards has shown that the terminal 66 THE MAMMALS OF CHINA AND MONGOLIA phalanges of the hands are bifid as in the moles, and the humerus lacks the entepicondylar perforation for the brachial nerve. In the skull the zygomatic arches are complete though slender and thread- like, and the tooth formula has one lower incisor on each side less than the full placental number. The interpretation of the lower teeth was tentatively published by Milne-Edwards as three incisors, a canine, three premolars and three molars, but Thomas (1912b) has shown that it is really one of the incisors (the anteriormost) that is missing, so that the tooth formula is: i.$c.ipm.4m.% = 42. The first upper incisor of each side is the largest, and the two are set with their chisel-like edges transverse to the tooth row; the two succeeding upper incisors are smaller, the canine conical, and relatively small, about as high as the first incisor; it is followed by three small, conical, double-rooted premolars, and a fourth much larger consisting of a high tri- angular cusp and asmallinner lobe. In the lower jaw the two anterior incisors on each side are small, chisel-like and project slightly forward, the canine is very small, while the first premolar as in the true moles is enlarged and conical, with a minute posterior cusp at the base. The two small premolars behind it are similar but smaller, and double-rooted. ; So far as known the genus is confined to the Chinese highlands. This interesting genus, as Milne-Edwards says, looks like a mole with the feet of Urotrichus, or like a Urotrichus with the head of a mole. It gives the annectant stage between the more shrew-like moles, as represented by Urop- silus, and the more typical genera, adapted for a wholly subterranean life. The presence of this genus isolated in the highlands of western China is another instance of the persistence here of a primitive type, while its additional interest lies in the fact that it is represented on the Pacific coast of North America by a related but more progressive genus, Neurotrichus, and in Japan by Uro- trichus. Only one species is known, with one subspecies. 16. Scaptonyx fusicaudatus Milne-Edwards Scaptonyx fusicaudatus Milne-Edwards, Recherches pour servir a l’Hist. Nat. des Mammiferes, p. 278, pl. 38B, fig. 4; pl. 40B, fig. 2, 1868-74. Scaptonyx fusicauda David, Nouv. Arch. Mus. d’Hist. Nat. Paris, vol. 7, Bull., p. 92, 1871 (lapsus calami). Type specimen:—The type was a specimen sent by Pére Armand David who secured it ‘‘sur les confins du Kokonoor et du Sé-tschouan.”’ Unfortu- nately, the flask in which it was preserved in alcohol was broken in transit to France, so that the specimen arrived in very poor condition, and the skull was so injured that Milne-Edwards found it possible to figure the teeth only. The specimen is still preserved in the Muséum d’Histoire Naturelle at Paris. Description:—The general form is mole-like, with the fore feet only slightly broadened, but with stout flattened fossorial claws, the hind feet more THE INSECTIVORES 67 slender, with long, compressed claws, the backs of both hands and feet covered with scales between which are short scattered black hairs. Fur soft, short and mole-like, of a uniform dark slate color throughout. The tail is hardly twice the length of the hind foot, thickened, and slightly tapering from the enlargement near the base (fusiform), thinly and evenly clad with long project- ing blackish hairs. The skull is lightly built, with slender and complete zygomata, the sade tapering to a blunt point. The teeth have been briefly described under the generic characters. Measurements:—The type specimen measured: total length, 108 mm.; tail, 45. No skull measurements of the typical form are available. Occurrence and Habits:—The type is the only recorded specimen. Specimens examined:—None. 17. Scaptonyx fusicaudatus affinis Thomas Scaptonyx fusicaudatus affinis Thomas, Ann. Mag. Nat. Hist., ser. 8, vol. 9, p. 514, 1912. Type specimen:—A male, skin and skull, No. 12.3.18.1, British Museum, from twelve miles south of Atuntze, northwestern Yunnan, at 13,500 feet altitude. Collected June 22, 1911, by F. Kingdon Ward. Description:—Similar to the typical form, but the upper canine slightly smaller, and with other minute differences in the teeth, as follows: third upper premolar not larger than second incisor; first and second upper premolars sub- equal, and smaller than the third; fourth upper premolar slightly shorter horizontally but of about the same breadth as the third; lower tooth row shortened, the incisors less spatulate; canine (the third tooth in the jaw) shorter and more slender than the posterior incisor; first and fourth lower premolars nearly equal in size but “‘rather lighter” than in the typical form, but the second and third lower premolars conspicuously smaller, the latter not a quarter the bulk and only about half the height of the fourth; the second lower incisor again about one-half its bulk and three-quarters its height, both the second and third single-rooted (Thomas, 1912b, p. 514). Measurements:—The following dimensions are available from specimens in the British Museum, and others taken by Edmund Heller of the fresh speci- mens collected by the American Museum Asiatic Expeditions: No. Head and body Tail Hind foot Ear Sex 12.3.18.1 BM (type) 90 31 15.5 35 of 44517 88 26 15.0 — rol 44518 ~ - — — o 44519 — — we ee wee J 22.9.1.12 BM 80 29 12.0 — 68 THE MAMMALS OF CHINA AND MONGOLIA CRANIAL MEASUREMENTS OF SCAPTONYX Zygo- Breadth Breadth Upper Lower Greatest Basal Palatal matic ofbrain outside tooth tooth No. length length length breadth case molars TOW row Locality 12°3710.0 BM, | 24.3°) "202 II.0 8.6 10.7 6.8 10.5 9.1 Yunnan 44517 24.0 19.4 10.5 8.5 — 6.7 10.5 9.8 Yunnan 44518 —_ — — — — — — Yunnan 44519 — ———— — = — —— — Yunnan 22.9.1.I2 BM 22.4 18.6 10.0 — 9.9 6.0 10.1 9.3 Yunnan The skull is extremely slender, with thread-like zygomata, which are less in width than the width of the brain case. The canine and unicuspid premolars of the upper jaw are all double-rooted. Thomas, who made actual comparison of the skull of the type of S. fusi- caudatus with that of his specimen from Atuntze, has pointed out the minute differences mentioned above in the proportions of the teeth, and on that basis has regarded the latter animal as a distinct form. It must be said, however, that the characters are of the most trivial nature, and without comparison of more than these two individuals it is difficult to tell how valuable they may be as a basis for subspecific distinction, since there is no way of knowing how great may be the individual variation in the typical form. Of the three specimens secured by the American Museum Asiatic Expeditions, No. 44517 has only three lower premolars, lacking apparently the second, of which there is not even an alveolus. Occurrence and Habits:—The type specimen of this subspecies, taken some twelve miles south of Atuntze, in northwestern Yunnan in 1912, was the second known example of the genus, and enabled Thomas to correct the impression of Milne-Edwards that the snout was not especially elongated, for it is ap- parently nearly as attenuate as in Uropsilus. Ten years later, Thomas (1922b, p. 393) recorded a second specimen, also a male, secured by George Forrest, on the Mekong-Salween divide, 28° north, at an elevation of 7,000 to 8,000 feet, practically a topotype. The only other specimens known seem to be the three secured by Dr. Andrews and Mr. Heller in 1914, one not far from the original locality, at Tomulang in the Chungtien district, northwestern Yunnan, at 10,000 feet, and two others on the Snow Mountain of the Likiang Range, to the southward, at from 12,000 to 13,000 feet. Probably this is a species of more or less forested country, for Thomas records of his first specimen that it was taken on a mossy bank in a fir forest. Specimens examined:—Five, as follows: Yunnan: Tomulang, Chungtien district, 1; Likiang Range, Snow Mountain, 2; Atuntze, 1 (B.M., the type); Mekong valley, 1 (B.M.). THE INSECTIVORES 69 Genus Talpa Linnzeus Talpa Linnzus, Syst. Nat., ed. 10, vol. 1, p. 52, 1758. The typical genus of moles shows a body specialized for subterranean burrowing life, in the pointed head, heavy musculature of neck and arms, the shortened tail, fore feet enlarged and broadened with the palms turned outward, and the five claws stout, flattened and elongate for digging. The short, plush-like fur, lacking a definite grain, is suitable for living in close quarters, and the eye is much reduced in correlation partly with life under- ground. The teeth in Talpa are of the full number characteristic of placental mammals, with the formula: i. ct pm.4m.$=44. The three incisors above and below are subequal, small and with chisel-like edges set nearly transverse to the long axis of the skull; the canines are well differentiated by their size, the upper especially large in contrast to the reduced condition in the Uropsi- linze, and are two-rooted. The three small premolars that follow are of nearly equal size, the fourth considerably larger, while the three molars all show the four primary cusps, of which the two outer form a W with their commissures. The last molar is the smallest, with its inner cusps slightly reduced. The type species is Talpa europea Linneus. As a genus this is chiefly found in the temperate portions of Europe and western Asia, as far as the high- lands of India, while a single species reaches western China. 18. Talpa longirostris Milne-Edwards Talpa longirostris Milne-Edwards, Compt. Rend. Acad. Sci., Paris, vol. 70, p. 341, 1870; Recherches pour servir a l’Hist. Nat. des Mammiféres, p. 281, pl. 17A, fig. 2 (skull and teeth); pl. 38, fig. 2 (exterior), 1868-74. Type specimen:—No type is indicated in the original description, but the specimen figured and described by Milne-Edwards is presumably still in the Muséum d’Histoire Naturelle at Paris, having been sent by Pére Armand David from ‘‘Thibet oriental,’ that is, probably from the mountains of Muping, Szechwan, China. Description:—This mole is slightly smaller than the European species, with, for a mole, a fairly well-developed tail, terete, with long sparse hairs, the more terminal of which are some 12.5 mm. long. The general color in fresh pelage is slaty black, with a slightly brownish tinge. The skull has a narrow rostrum, and narrow almost cylindrical inter- orbital region, from which the wide, oval brain case rather abruptly expands. The upper incisors are subequal, the third slightly the narrowest; the second upper premolar is slightly smaller than the first, and recurved; the third and fourth increase regularly in size, but the fourth has a distinct basal cusp posteriorly. 70 THE MAMMALS OF CHINA AND MONGOLIA Measurements :—In addition to the measurements given by Milne-Edwards (1868-74, p. 283) for the type of this species, those of the two in the British Museum are added: No. Head and body Tail Hind foot Locality PARIS (type) 105 20 20.0 (c.u.) Szechwan 99.3.1.9 BM ——— —_ 18.5 (c.u.) Szechwan II.2.1.24 BM 105 20 14.5 (s.u.) Szechwan CRANIAL MEASUREMENTS OF TALPA LONGIROSTRIS Zygo- Width Upper Lower Greatest Basal Palatal matic Mastoid across tooth tooth No. length length length width width molars row row Locality 99.3.1.9 BM 129 —- — 7 13.1 12.4 Szechwan II.2.1.24 BM 32.1 26.8 12.9 9.0 14.5 6.8 12.6 12.0 Szechwan Milne-Edwards’s excellent figures of the skull and dentition leave no doubt that this is a true Talpa. In comparison with the European species, how- ever, the upper premolars seem slightly stouter, and the anterior two perhaps single-rooted instead of with two roots, while the large first lower premolar is shown in contact with the lower canine. Milne-Edwards states that there are four lower incisors, but in fact there are only three. The lower canine in side view resembles the incisor next it in its low crown and short cutting edge, but in crown view its base extends much farther inward; the first lower premolar has become caniniform, and of large size, but that it is not the canine is shown by the fact that it lies just behind the upper canine when the jaws are closed. Occurrence and Habits:—Although said by Milne-Edwards to be apparently “assez commune” in the mountains of Szechwan and eastern Tibet, this continues to be a rare species in collections. It was discovered by Pére David in the course of his journey to Szechwan in 1870, whence he sent back the first specimens to the Paris Museum. This area was at that time a borderland included as part of Tibet, but there seems to be no evidence that the species actually occurs in that country as at present understood. No specimens were again reported until 1899, when De Winton and Styan recorded a male secured by the latter at Yunglipa (Yangliupa) in northwestern Szechwan; they de- scribe its color as uniformly black. This specimen remained unique in the British Museum collection until 1911, when Thomas (1911d, p. 163) recorded a second male secured by Malcolm P. Anderson at Omei Shan in the same province. Slightly to the southward the later expedition of the Dresden Museum secured two additional specimens in alcohol, from the isolated Wa Shan Range, the first bought from natives at Wa Shan, the second caught by a dog at Hwanglungtse, in an alpine forest of red-barked birches, among mossy boulders (Jacobi, 1922, p. 2; Weigold, 1923, p. 71). THE INSECTIVORES 71 Anderson’s specimen from Omei Shan was caught on a “mossy bank in damp forest,”’ so that the slight available evidence seems to indicate that it is a forest-dweller. Specimens examined:—Two, namely: Szechwan: Omei Shan, 1 (B.M.); Yangliupa, 1 (B.M.). Genus Parascaptor Gill Parascaptor Gill, Bull. U. S. Geol. and Geogr. Surv. Terr., vol. 1 (ser. 2), p. 110, 1875. Type species, Talpa leucura Blyth. This genus is closely similar to Talpa, of which by many writers it has been regarded as merely a subgenus, but is well distinguished by the tooth formula, which differs from that of Talpa in lacking one of the small upper premolars, making a total of three premolars above instead of the four present in Talpa, as follows: 1.3 c.+ pm. m.3 =42. In external appearance it resembles the common mole, but the tail is very short and club-shaped. Its range includes Assam and Burma, eastward to the borders of Yunnan, and Siam. But the one species is known from eastern Asia, although Milne-Edwards at one time referred to it the mole described from North China by Thomas as Talpa leptura, in which the tooth formula was similar, but which, as later pointed out by Thomas (1910), proved on further examination to be really _ a Scaptochirus, with the characteristic broad heavy skull, but abnormally with an extra lower premolar. Milne-Edwards (1884) has also described as Para- scaptor davidianus a mole said to have been collected by Pére Armand David in Syria, but nothing further seems to be known about it. 19. Parascaptor leucurus (Blyth) Talpa leucura Blyth, Journ. Asiatic Soc. Bengal, vol. 19, p. 215, pl. 4, fig. 1, 1850. Parascaptor leucurus Milne-Edwards, Compt. Rend. Acad. Sci., Paris, vol. 99, p. 1142, 18845 Type specimen:—The original specimen was from Cherra Punji, Assam, British India, and is presumably in the Indian Museum at Calcutta. _ Description:—Similar in general appearance to the common European mole, but the tail very short, club-shaped, about one-twelfth of the total length. Color a uniform brown, varying to black, the basal part of ~ fur “leaden black”’; tail hairs very short, white or whitish. The skull is of the more delicate type, like that of Talpa, with tapering rostrum. The teeth resemble those of the latter, but lack one of the small unicuspid premolars, so that only two of these small teeth intervene between the well-developed upper canine and the large posteriormost premolar. In the lower jaw, the canine in side view is low, with a chisel-like edge, and closely resembles the incisors, while the first premolar as in Talpa is enlarged and 72 THE MAMMALS OF CHINA AND MONGOLIA caniniform, closing behind the upper canine; it is succeeded by two small premolars set close together, while the fourth is large, equaling in size the first. Measurements:—Blanford gives the following dimensions (here reduced to millimeters) for an alcoholic specimen from India, and I have added those of the Suki specimen: No. Head and body Tail Hind foot Locality 14.10.32.2 BM 110 15.5 15 Yunnan BLANFORD 105 10.0 — India CRANIAL MEASUREMENTS OF PARASCAPTOR Zygo- Mas- Width Upper Lower Greatest Basal Palatal matic toid across tooth tooth No. length length length width width molars Tow Tow Locality 14.10.23.2 BM 27:80" 24240 11-71 +(955) 13:6)N 73 11.5 11.5 Yunnan 22.10.21.1 BM 28:5 in24.3 IT:6. 9:3 14.3) «73 11.6 II.I Yunnan Occurrence and Habits:—Blanford gives the distribution of this mole as Sylhet, the Khasi and Naga Hills, south of Assam, and probably locally throughout Burma, reaching an altitude of as high as 10,000 feet. It is also present in northern Siam, whence it is recorded by Gyldenstolpe (1919) as having been obtained by Eisenhofer southeast of Chiengmai. It is, there- fore, not surprising to find it extending into the extreme western border of Yunnan, whence Thomas (1914b, p. 473) has lately recorded the first known Chinese specimen, collected by F. Kingdon Ward at Suki, in the Salween valley, 7,000 feet altitude, in north latitude 27° 30’. There is also in the British Museum a second specimen from Yunnan, taken at Tengyueh, and received in 1922. Ward notes on the label of his specimen that it was captured where there were “numerous burrows in the river bed, amongst grass and shrubs.” No other specimens have been taken in China and nothing is recorded of its soil preference or food. Specimens examined:—In addition to specimens in the British Museum from the Shan States, Khasi Hills, and Assam in British India, two, from Yunnan,—Suki and Tengyueh respectively (B. M.). Genus Scaptochirus Milne-Edwards Scaptochirus Milne-Edwards, Ann. des Sci. Nat., Zool., ser. 5, vol. 7, p. 375, 1867. Talpa David, Nouv. Arch. Mus. d’Hist. Nat. Paris, vol. 3, Bull., p. 26, 1867 (not Linnzus, 1758). Chiroscaptor Heude, Mém. concern. l'Hist. Nat. de l'Emp. Chin., vol. 4, pt. 1, p. 36, pl. 9, figs. 1-1c, 1898. The moles of this genus differ externally from Talpa in the more reduced tail, which is very short and slender (about two-thirds the length of the hind foot), and thinly haired. THE INSECTIVORES 72 The skull is less delicate, with a shorter, broader rostrum, and the molar teeth are larger with higher crowns. In its dentition, Scaptochirus has one less premolar both above and below, on account of the loss of one of the small intermediate teeth in each jaw, so that the formula is: 1.3 c.t pm. m.3=40. The lower canine is in side view like an incisor, while the first lower premolar is enlarged and caniniform. The type species is Scaptochirus moschatus Milne-Edwards, and the genus seems to be confined to eastern China and the borders of southeastern Mongolia. Only a single species is known from this area, with a subspecies of which little is yet made out. 20. Scaptochirus moschatus moschatus Milne-Edwards Scaptochirus moschatus Milne-Edwards, Ann. des Sci. Nat., Zool., ser. 5, vol. 7, p. 375, 1867; Recherches pour servir a l’Hist. Nat. des Mammifeéres, p. 173, pl. 17, fig. 4; pl. 17A, figs. 1-1c, 1868-74. Talpa europea David, Nouv. Arch. Mus. d’Hist. Nat. Paris, vol. 3, Bull., p. 26, 1867 (not of Linnzus, 1758). Talpa leucura ? Swinhoe, Proc. Zool. Soc. London, 1861, p. 135 (not of Blyth). Scaptochirus davidianus Swinhoe, ibid., 1870, p. 620 (errorim). Talpa leptura Thomas, Ann. Mag. Nat. Hist., ser. 5, vol. 7, p. 470, 1881. Parascaptor leptura Milne-Edwards, Compt. Rend. Acad. Sci., Paris, vol. 99, p. 1142, 1884. Chiroscaptor sinensis Heude, Mém. concern. 1’Hist. Nat. de 1’Emp. Chin., vol. 4, pt. 1, p. 36, 1898. Scaptochirus moschiferus Heude, ibid., p. 40, pl. 9, figs. 2-2c (errorim). Scaptochirus leptura Thomas, Ann. Mag. Nat. Hist., ser. 8, vol. 5, p. 350, 1910. Scaptochirus lepturus Thomas, loc. cit. Type specimen:—The original specimen was collected by Pére Armand David ‘‘en Mongolie” and is presumably still in the collection of the Muséum d’Histoire Naturelle at Paris. In David’s time the term Mongolia was used to include what are now the western parts of Hopei and Shansi, so that Thomas has suggested that the type locality may therefore be considered Suanhwafu, about one hundred miles northwest of Peiping, whence much of David’s “Mongolian” material came. Description:—In color this mole is a nearly uniform clear grayish brown above and below, slightly paler about the mouth and on the lower side, glisten- ing silvery in certain lights. The fur is short, with slaty bases and minute brownish tips. The snout is thinly haired, with a lengthwise groove ventrally. The tail is short, about two-thirds the length of the hind foot, with short hairs, hardly hiding the scales, and forming a short terminal tuft. Backs of the broad fore feet and narrower hind feet thinly haired, or in old animals nearly naked. Measurements:—No fresh measurements of this mole are at hand. Milne- Edwards notes that the head and body of his alcoholic specimen, following the curve of the back, measured 140 mm., and the tail less than a centimeter. The hind foot of the type of S. leptura is 20 mm.; its tail, perhaps stretched in the mounted specimen, is 15 mm. 74 THE MAMMALS OF CHINA AND MONGOLIA Fic. 4. Distribution Map. < : Scaptochirus Scapanulus 1. S. moschatus moschatus 3. S. owent 2. S. moschatus gilliest CRANIAL MEASUREMENTS OF SCAPTOCHIRUS 4 E S < 3 s 5 8 gs a eel a ee ant Meant Pras Wee ee z Si oak ae OS aie OBS S. moschatus moschatus 28.1.6.2 BM 34.2 29.6 14.5 14.7 18.0 10.8 14.3 28.1.6.3 BM 34.3,.29.3. 14.7 —— 17.8 10.7 14.3 28.1.6.4 BM 34.0 28.9 14.2 14.3 18.1 II.I 14.5 8.8.11.15 BM 33-2 —— 14.7 —— 17.4 10.5 14.4 16.1.1.4 BM 30.7 27.2 13.3 —— 17.2 10.1 13.0 61.6.2.4 BM 34.0 —— 14.9 —— 18.1 10:5 14.6 (type of T. leptura) 25883 MCz 35:7 30.9 15.3 —— 17.9 II.1 15.7 S. moschatus gilliest 10.3.13.I BM 30:2) 2514) 124133 161g" "O17 e12:5 Lower tooth row Locality 13.6 Shantung 14.0 Shantung 13.5 Shantung 13.7 Hopei —— Hopei 13.3 Hopei 14.4 Shantung 12.5 Shansi THE INSECTIVORES 75 Nomenclature:—The first specimen of this mole to reach Europe was one sent by Consul Swinhoe to the British Museum in 1860, from Peiping. This, as Swinhoe mentions, was pronounced a new species by Gray, who gave it the manuscript name (on the museum stand) Talpa chinensis. This, how- ever, seems never to have been published, so that it remained for Milne- Edwards to study and describe it as a new genus and species in 1867, on the basis of a specimen sent by Pére David, probably, as Thomas suggests, from Suanhwafu, about ninety miles northwest of Peiping. Later, in 1870, Swinhoe in his account of Chinese mammals again mentions the species, but inad- vertently called it’ Scaptochirus davidianus, a specific name which, curiously, Milne-Edwards gave in 1884 to a mole from Syria. In 1881, Thomas having removed the skull from the original specimen sent by Swinhoe, found that the number of teeth agreed with Talpa instead of with Scaptochirus, and therefore described the specimen from Peiping as Talpa leptura. He later discovered, however, that the lack of the additional tooth characteristic of the genus Scaptochirus was evidently an abnormality, for the teeth agreed otherwise; nevertheless he assumed that the Peiping animal might be main- tained as distinct on the basis of a longer tail, 15 mm. in the specimen as mounted, but this seems highly improbable, and as the skulls are quite the same, I am regarding T. Jeptura as a synonym of S. moschatus. In 1898, Heude described as a new genus and species, Chiroscaptor sinensis, a mole from Hopei which he believed differed from S. moschatus (inadvertently written moschiferus!) in that the cusps of the lower molars projected much more forward. An examination of his figure and description, however, leaves no doubt that the differences he observed were due to the fact that he compared the fresh, unworn dentition of his type with the much worn teeth of an older individual. Occurrence and Habits:—Specimens of this mole seem to be uncommon in collections. In addition to the original specimen sent by David to Paris, and assumed to have come from northwest of Peiping, the British Museum has two from the neighborhood of the latter city, and another from Chihfeng, 200 miles to the north. Apparently its range extends farther to the north- east, for A. B. Howell (1929) records a specimen in the U. S. National Museum from Heisui, Manchuria. The only other locality yet reported is Weihsien in Shantung. Specimens examined:—Nine, as follows: Hopei: Peiping, 2 (B.M., including type of Talpa leptura); Chihfeng, 1 (B.M.). Shantung: Weihsien, 6 (A.M.N.H., M.C.Z., B.M.). 76 THE MAMMALS OF CHINA AND MONGOLIA 21. Scaptochirus moschatus gilliesi Thomas Scaptochirus gilliesi Thomas, Ann. Mag. Nat. Hist., ser. 8, vol. 5, p. 350, 1910. Scaptochirus gillesei Sowerby, in Clark and Sowerby, Through Shén-Kan, p. 172, 1912 (Japsus calami). Type specimen:—An adult, skin and skull, No. 10.3.13.1, British Museum, from Hotsin, southwestern Shansi, China. Collected November, 1909, by Robert Gillies. Description:—Similar to the typical race of Hopei and Shantung, but smaller, with smaller skull, although the tail and foot measurements are nearly identical. The general color of the type may be a slight shade darker, about ‘‘broccoli brown.” The skull is markedly smaller with less narrowed middle region, smaller and lighter teeth. Measurements:—No measurements of total length are available, but the tail is about 16 mm. long; hind foot with claws, 19.5; breadth of fore foot, 12.5. Cranial measurements of the type are given in the table under S. moschatus. Occurrence and Habits:—Moles referred to this slightly smaller race have been taken at several localities in Shansi and Shensi. In addition to the original specimen sent from Hotsin, southwestern Shansi, the American Museum Asiatic Expeditions secured one at Maitaichao, some forty miles east of Paotow; A. B: Howell (1929) records four in the U. S. National Museum from twenty miles west of Ningwufu and a single one from Taiyuanfu, whence also the Museum of Comparative Zodlogy has two specimens collected by F. R. Wulsin; in northwestern Shansi, Sowerby (Clark and Sowerby, 1912, p. 172) mentions its capture on the plains of Wuchai, a dry, sandy area. In Shensi, Sowerby secured it at Yulinfu on the borders of the sandy Ordos Desert, where it was rare and “‘unexpected.’’ He writes (Sowerby, 1914, p. 59) that this mole ‘‘seems to have adapted itself to an existence under more or less desert conditions, in which there certainly cannot be any abundance of worms.” Possibly in the absence of these, it finds subsistence in the shape of beetle larve, with the adults of which the same author found the desert swarming at certain times of year. Of the two specimens from Taiyuanfu, one taken August 10, 1921, is a young animal with the hair just beginning to grow out, and so short that it hardly projects from the skin. The type specimen in the British Museum is considerably smaller than those from about Peiping, yet is quite adult, with basal suture wholly obliterated. Specimens examined:—In all, four, as follows: Shansi: Maitaichao, 1; Taiyuanfu, 2 (M.C.Z.); Hotsin, 1 (B.M., the type). THE INSECTIVORES + HINA SEA POU ean HAI) Fic. 5. Distribution Map. Mogera 1. M. latouchet 2. M. hainana Genus Mogera Pomel Mogera Pomel, Arch. des Sci. Phys. et Nat., vol. 9, p. 247, 1848. The moles of this genus are characterized by the loss of the lower canine from the dentition, so that the total number of teeth present is one less on each side than in Talpa, in which the full number typical of placental mammals, namely 44, is found. The tooth formula is, therefore: 1.3 c.¢ pm.t m.3 =42. The upper incisors form a slightly convex transverse row; the upper canine is strong and broad, the upper premolars 1-3 are usually double-rooted. In the lower jaw there is a space where the missing canine should be, while the first premolar is large, double-rooted, and functions in place of the canine, although its true homology is evident from the fact that it closes behind the upper canine tooth. Externally the moles of this genus resemble Jalpa in having the usual form and a short stout tail, well clothed with hair. The range of the genus in general is complementary to that of Scaptochirus, for whereas the latter is the mole of northern and northwestern China, the former 78 THE MAMMALS OF CHINA AND MONGOLIA is found in the extreme eastern and southeastern part of the country. No doubt there is a difference in soil preference or food, inducing the former to seek the sandy dry areas of the loess formation, while the latter is more coastal, and perhaps is partial to a different type of country. In China this genus is at present known chiefly from the southern portion, extending westward to Szechwan, and eastward through Formosa and the Japanese islands into Korea and Manchuria, but apparently avoiding altogether the provinces of North China. The type species is the Japanese Talpa (=Mogera) wogura Temminck. 22. Mogera latouchei Thomas LA TOUCHE’S MOLE Mogera latouchei Thomas, Proc. Zool. Soc. London, 1907, p. 463. Talpa sp., Swinhoe, Proc. Zool. Soc. London, 1870, p. 620. Talpa wogura Thomas, Proc. Zool. Soc. London, 1898, p. 771 (not of Temminck). Mogera mogera Shih, Bull. Dept. Biol., Sun Yatsen Univ., Canton, no. 4, p. 3, 1930. Type specimen:—The type is a skin and skull, No. 98.8.17.1, British Museum, from Kuatun, northwestern Fukien, China. J. D. La Touche, col- lector, 1898. Description:—A small mole, of a nearly uniform slate color, faintly washed above with dark brown; below more smoky, the throat and upper chest paler, dark gray. Backs of the hands and feet and the tail thinly clad with dull whitish hairs, those on the tail longer. The tail is slightly longer than the hind foot. The skull is smaller as compared with that of the more northern forms, and is specially characterized by having the first upper premolar single-rooted and usually shorter than the second. . Measurements:—The following measurements of the exterior were made by the collector, Mr. Clifford H. Pope, of a series from Chunganhsien: No. Head and body Tail Hind foot Sex IIo 15 14.0 2) 84805 93 19 14.0 rol 84806 108 18 14.0 fos 84808 106 14 14.0 rot 84809 87 16 13.5 rou 84810 95 20 14.0 rot 84811 115 19 14.0 THE INSECTIVORES 79 CRANIAL MEASUREMENTS OF MOGERA Upper Lower Greatest Basal Palatal Greatest Across tooth tooth No. length length length width molars TOW Tow Locality M. latouchet 84805 29.0 25.0 12.0 14.0 vie) 12.0 11.5 Fukien 84809 27.7 24.0 11.5 13.0 7.0 11.5 11.0 Fukien 84811 28.5 24.0 11.7 13.5 6.8 11.6 — Fukien 84808 29.5 25.0 12.0 15.0 7.0 11.8 11.0 Fukien 96.8.17.1 BM 28.8 24.0 11.6 13.9 Tee 11.6 10.8 Fukien M. hainana 10.4.25.4 BM 30.0 — 12.0 14.3 7.6 Lie7, DL Hainan Occurrence and Habits:—The occurrence of a mole in southern China seems first to have been recorded by Robert Swinhoe (1870c, p. 620), who secured one from Foochow, Fukien. It was next obtained in the northwestern part of the same province by J. D. La Touche and C. B. Rickett, who secured six (four skins and two in alcohol) at Kuatun, where it was found to be ‘‘tolerably common’”’ in the hill country. One of these specimens subsequently became the type of this species. In addition, Cabrera (1922, p. 163) has recorded it also from Foochow, and the American Museum of Natural Nistory has speci- mens from Fuching and from Shaowu on the Min River in the same province, as well as a series from Chunganhsien, northwestern Fukien, practically topo- types, secured by Mr. Clifford H. Pope, who writes that it is common there in the high mountains. At Kuatun, he says, the Chinese know it as ‘“‘fan pa chang”’ (reversed palm) and ‘‘pu chien t’ien’’ (not see heaven, 7. e., blind). Westward from Fukien, it is found sparingly, having been recorded from the southwestern border of Hunan (Shih, 1930b, p. 2) and still farther south, from the Yao Shan district of Kwangsi (Shih, 1930, p. 3); while the most west- erly locality known is Tseogiakeo, Szechwan, south of Suifu, on the Yunnan border, whence A. B. Howell (1929, p. 7) records a single specimen in the col- lection of the U. S. National Museum. Concerning the last, Howell states that it has a broader interpterygoid and smaller bulle than a specimen from Kuatun, indicating that possibly the more western animal is slightly different, although without a series for comparison, the value of these differences cannot be determined. He adds further that in his specimen from Kuatun the first upper premolar is a trifle longer than the second, instead of shorter; it is slender and recurved. Specimens examined:—Sixteen, including skins with or without skulls, and a separate skull, namely: Fukien: Fuchinghsien, 1 skull; Chunganhsien, 8; Shaowu, Min River, 4; Kuatun, 1 (B.M., the type); northwestern Fukien, 2. 80 THE MAMMALS OF CHINA AND MONGOLIA 23. Mogera hainana Thomas THE HAINAN MOLE Mogera hainana Thomas, Ann. Mag. Nat. Hist., ser. 8, vol. 5, p. 535, 1910. Mogera insularis hainana Kloss, Journ. Nat. Hist. Soc. Siam, Bangkok, vol. 6, p. 213, 1923. Type specimen:—The type is an adult female, skin and skull, No. 10.4.25.4, British Museum, from Mount Wuchih, Hainan, China, taken by Alan Owston’s collector. Description:—Larger and browner than M. latouchei, the general color of the head and body above, nearly ‘mummy brown’’ of Ridgway, slightly more smoky below, and paler grayish about the head above and below. In the specimen at hand, the sides of the muzzle are white. Backs of the hands and feet with scattered short whitish hairs; tail very short, shorter than hind foot, and well covered with short hairs like the back in color. As usual in moles, a silvery sheen appears in certain lights. The size is intermediate between M. insularis of Formosa and M. latouchei, both of which agree in having the anterior premolar of the upper jaw smaller than the second and single-rooted, while M. hainana agrees with the more northern members of the genus in having that tooth larger than the second and double-rooted. Measurements:—The following field measurements were made by Mr. Clifford H. Pope: No. Head and body Tail Hind foot Locality 59913 120 9 14 Hainan 59914 II5 14 17 Hainan 59915 122 10 II Hainan 59916 125 9 8 Hainan 59917 134 12 18 Hainan For cranial measurements, see table under M. latouchei, page 79. Occurrence and Habits:—The occurrence of a mole on the island of Hainan was apparently unsuspected by Swinhoe, or by later collectors, until, in 1910, Thomas obtained it from Mount Wuchih through Alan Owston’s collectors, and named it Mogera hainana. By a curious chance, Kloss (1923), thirteen years later, overlooking Thomas’s description, renamed the animal Mogera insularis hainana, regarding it as a subspecies of the Formosan animal, M. insularis. His specimen also came from Five-finger Mountain (Wuchih). It is apparently local in its distribution and difficult to secure. Mr. Clifford H. Pope tells me that he set traps in many places, and frequently found the little ridges of earth pushed up by the animals, but succeeded in catching only two, although four others were purchased from natives. His trapper THE INSECTIVORES 81 averred that they burrow at a deeper level than the moles of North China (Scaptochirus) with which he was familiar. Specimens examined:—Seven, as follows: Hainan: Nodoa, 1; Namfong, 5; Mount Wuchih, 1 (B.M., the type.) Genus Scapanulus Thomas Scapanulus Thomas, Ann. Mag. Nat. Hist., ser. 8, vol. 10, p. 396, 1912. The special interest of this genus of moles is that it represents in eastern Asia the group to which the moles of North America belong (except of course, the star-nosed mole), namely, the subfamily Scalopodine (“‘Scalopine’’), in which the incisors become enlarged and the canine of the upper jaw reduced, so that the anterior teeth of the jaw bear the brunt of its forceps-like action. Exteriorly the form is mole-like, but the hands, though more expanded than in Scaptonyx, are not quite so wide proportionally as in Talpa, etc., while the claws, though long and flattened, are rather slender. The first digit of the hind foot, as first noticed by Thomas, is set outward at a slight angle to the remaining toes, and its claw is stouter and more sharply curved than the others, which are long, slender, and nearly straight. The tail is relatively long, about twice the length of the hind foot, stout and thickly haired. The snout is taper- ing, fairly long, and grooved on its lower side in the middle. In the skull, the pterygoids are better developed than in Scapanus, and the tympanic bone is incomplete. The interparietal is broad, and less tapering forward than in Urotrichus. The tooth formula includes nine teeth above and nine below on each side, a number present elsewhere only in Neurotrichus, as follows: 1.3 c.t pm.$ m.3=36. Thomas is confident that it is the third incisor in each jaw and probably the second premolar that are lacking. The lower incisors are proclivous and both abut against the large upper first incisor. Thomas states that the second upper incisor and the first upper premolar are subequal and smaller than the canine that stands spaced between them. In our specimen, however, the second incisor is nearly as large as the canine, and the first premolar is slightly the smallest of all. The third premolar is as high as the canine but more broadly triangular in side view. The upper canine and the two small premolars are all double-rooted. The fourth upper premolar is largest, with a distinct posterior shoulder, but our specimen does not show an internal cusp as did that of the type. The internal ledge-like protocone of the upper molars is faintly three-lobed. The type and only known species of the genus is Scapanulus oweni Thomas. 24. Scapanulus oweni Thomas Scapanulus oweni Thomas, Ann. Mag. Nat. Hist., ser. 8, vol. 10, p. 396, 1912. Type specimen:—A male, skin and skull, No. 12.8.5.2, British Museum, 82 THE MAMMALS OF CHINA AND MONGOLIA from southeast of Taochow, Kansu, China. Collected October 31, 1911, by G. Fenwick Owen. Description:—General color above and below drab gray, with a silvery appearance in certain lights, very similar to the color of Blarina. On close inspection it is seen that the individual hairs are slaty with a minute tip of brown. Area at base of snout slightly paler. Backs of hands and feet thinly haired, their central area colored like the back, the fingers whitish. The stout club-shaped tail is all of twice the length of the hind foot with claw, and so thickly clothed with rather long hairs as to hide the scales; its tip and lower side are slightly paler than the rest, which is colored like the back. The peculiarities of the skull and teeth have been noted in the generic description. Measurements:—Thomas published the following dimensions of the type, taken in the flesh: head and body, 108 mm. ; tail, 38; hind foot, 14. A specimen in the Museum of Comparative Zodlogy measured: total length, 136 mm.; tail, 38. SKULL MEASUREMENTS Zygo- Breadth Breadth Upper Lower Greatest Basal Palatal matic ofbrain across tooth tooth No. length length length breadth case molars row TOW Locality BM (type) Phe ee Ayes LG | 9.6 13.0 8.0 12.3 11.5 Kansu 23915 MCZ 2710) 20-5) el 2rI 10.0 13.2 8.0 11.7 11.0 Kansu Occurrence and Habits:—As already noted, this mole is an Asiatic repre- sentative of the subfamily to which the American moles (except Condylura) belong, in which the anterior incisors are enlarged and the canine reduced. It much resembles Scapanus or Parascalops externally, but is smaller with a longer, well-haired tail. The curious distortion of the first hind toe and its peculiar curved nail were mentioned by Thomas in his original description and are evident in our specimen too. The discovery of this remarkable mole in northwestern China is another of the notable finds in recent years, linking the fauna of eastern Asia with that of America. Apparently this is a forest-living mole, for G. Fenwick Owen, who collected the two specimens originally recorded, states that they were taken in mossy undergrowth in fir forest. In addition to these, from twenty-three and forty- six miles respectively, southeast of Taochow, in Kansu, only a few other speci- mens are known, as follows: one from Archuen, Min Shan, in the U. S. National Museum (A. B. Howell, 1929, p. 8), and a second from Choni, in the Museum of Comparative Zodlogy, both localities in southeastern Kansu; a third in the THE INSECTIVORES 83 American Museum of Natural History, collected by Dr. R. C. Andrews, at Taipai Shan, Tsingling Mountains, Shensi, the last extending the known range slightly to the eastward; and a fourth taken in northwestern Szechwan, at Hoangshuikwan, north of Sungpan, July 16, 1931, by the Brooke Dolan Expedition, and now in the collection of the Museum of Comparative Zodlogy. Specimens examined:—Four, as follows: Shensi: Taipai Shan, Tsingling Mountains, 1. Kansu: Choni, 1 (M.C.Z.); southeast of Taochow, 1 (B.M., the type). Szechwan: Hoangshuikwan, 1 (M.C.Z.). Family SORICIDA SHREWS While both the Talpidez and the Soricide are believed to have originated from some common stock, the former have retained the zygomatic arch of the skull and to a large extent an unspecialized dentition, although in some of its members the anterior incisors are slightly enlarged and the canines correspond- ingly reduced. The genus Uropsilus, as the most primitive of living talpids, indicates in general what this common stock may have been like. The Soricide, instead of developing modifications for digging, have retained the lightly built, slender body and limbs, with unenlarged feet and claws, a long tapering head and snout, and a low but evident externalear. -The skull is especially characterized by the loss of the zygomatic arches, and the great enlargement of the first incisor of both jaws, together with reduction of the canine, producing a forceps-like modification, foreshadowed in the den- tition of Uropsilus. The tympanic is a delicate ring of bone, unsolidified to the skull, in contrast to the bulla-like structure in the Talpide. A cloaca is frequently present, again a primitive character. Two subfamilies are recognized: one, the typical Soricine, in which the teeth are red-tipped; the other, the Crocidurine, in which the teeth are white. The former is mainly of northern distribution and occurs throughout most of the holarctic region; the latter is confined to the eastern hemisphere and is most abundantly represented in tropical and subtropical areas. In both subfamilies, as parallel modifications, aquatic members have been developed as well as semifossorial forms with reduced tails. In China and Mongolia, the Soricine are found throughout the more northern wooded area, southward at higher levels in the western highlands, while the Crocidurine are abundant in the southern parts of the area, even at high altitudes, but are few in number of species in northern China and Mongolia. 84 THE MAMMALS OF CHINA AND MONGOLIA Key TO THE GENERA OF CHINESE AND MONGOLIAN SORICIDZ A. Teeth with their cusps pigmented dark chestnut............... Subfamily Soricinz a. Tail about as long as head and body or less; upper unicuspids five; large lower incisor with three crenulations on its cutting edge. a’. Form slender, ears obvious, claws delicate, tail about half the length of head and body or longer.................. Sorex b’. Form stouter, ears not evident, claws slightly lengthened, tail about half the length of head and body............. Blarinella b. Tail long, about equaling or exceeding head and body; upper unicuspids less than five; large lower incisor with a single prom- inent denticle on its cutting edge. EMaal Ohojoleraibuab ely oy enoaad saa GoM O MOONS na os non Ac Soriculus . ib AiUppenitinicuspids:s) poe ae ret ere ata elses ycinacye hoon token talents Chodsigoa B. Teeth white, their cusps unpigmented......................-- Subfamily Crocidurine a. Tail at least half as long as head and body. a’. Ears prominent, projecting well above fur; tail with longer scattered bristles, especially on its basal half. a) Se Uppentinictispidsy4s cess cect | tee alae a eioree Suncus bY.) Upperiainicuspidsts i Mee) Hd, Pe SR Crocidura b’. Ears reduced, not projecting; aquatic, with fringes of short stiff hairs on sides of feet and toes. a’’, Feet not webbed, tail cylindrical................... Chimarrogale b”’. Feet webbed to base of third joint; tail with vertical andMlateralicrestss ek sites: notekt tere. eke tees RUN Nectogale b. Tail very short, not exceeding hind foot in length; ears much reduced; upper unicuspids 2 only..................0e cece Anourosorex In addition to these genera it can hardly be doubted that the genus Neomys, an aquatic group of the Soricine, will eventually be found in the northern part of Mongolia, for it occurs over much of continental Europe and the British Isles, eastward into northern Asia, and has recently been recorded from the shores of Lake Baikal, at no very great distance to the north of the Mongolian border (Ognev, Annuaire Mus. Zool. Acad. Sci. de Russie for 1917- 21, vol. 22, p. 346, 1921). Genus Sorex Linnzus Sorex Linnzus, Syst. Nat., ed. 10, vol. 1, p. 53, 1758. The members of this genus include some of the smallest living mammals. The limbs and feet are delicately formed, all four feet with five toes well developed, the first shorter, and all provided with sharp compressed claws. The snout is elongate and tapering, the vibrissz long, and the tail slender and thinly haired, about as long as the body. The skull is low and broad, of delicate structure, the brain case relatively large, the rostrum pointed and narrow. The first upper incisor is much en- THE INSECTIVORES 85 larged with two cusps one behind the other, the posterior cusp equaling in side view the two succeeding unicuspid conical incisors; the canine resembles these incisors in shape but is smaller, and is followed by two small premolars of which the second is much smaller than the first, and stands in the angle between the first and the much larger third premolar (p‘), which is somewhat molariform with a well-developed antero-internal cusp. The molars exhibit the primitive structure, with a well-marked W-pattern formed by the two outer cusps (paracone and metacone) and their commissures, and a large inner cusp (proto- cone) with a postero-internal ledge (hypocone) produced distinctly backward. In the lower jaw the first incisor is also enlarged, long and scalpriform, with its long axis in the direction of the tooth row and its upper cutting edge pro- vided with three serrations; it is followed by two smaller teeth, triangular in side view, which are interpreted as a canine anda premolar. The lower molars are three in number, so that the tooth formula is: i.¢ c.+ pm.{ m.3 = 32. This genus is of holarctic distribution in forested areas of the northern parts of Asia, Europe and North America. Several species occur in China and the wooded parts of Mongolia. In general they are of a nearly uniform brownish color above, paler below; one group, however, has a narrow black median stripe. Type species, Sorex araneus Linneus. Key TO THE CHINESE AND MONGOLIAN SPECIES OF Sorex A. Back a uniform shade of brown without black median stripe. a. Larger, hind foot with claws 13-14 mm. a’. Lower surfaces whitish-tipped. EES Alpail eilovrths ZK) yaobsabee en Gone Agate ec Be ot OO OOe Sorex araneus borealis bYaiDanltaboutisopmmycsecveicee ante Hecate dee S. excelsus b’. Lower surfaces distinctly brownish................ S. sinalis b. Smaller, hind foot with claws 12 mm. or less. . a’. Skull length about 18 mm. Aa OTLOWIEL fee pc mey earn sie hee cre isiegeeoeliis S. buxtont buxtont Bivegs GIVE ayer eitra trc cree ey ete Pra ccererrs cee eer S. buxtoni cansulus baskulllensthaboutersimimere so: 422 222+ ss ese ee S. minutus thibetanus B. Back with a blackish median stripe. a. Dark brown above. a’. Larger, skull length 17.5-18.0 mm................- S. cylindricauda cylindricauda Dawsmeallen skullllength 166mm 4-0 4.0 eee ele S. cylindricauda gomphus by ealer: oravish: brown aboves o.css.csc sc «