Sat dt 
Reith 


oastie, 
he 3, 
tries] 

fits ( 


; 
Baath 


2 tbe 
ea 
Meet 
He 
$i 


otad 
ESTO 


sitet ip 
etic fee 


Sah satiate Bias 
ds 


Mi 
Sikieest 


Peete: 
ei ys gles 


LIBRARY OF THE 


mk 
oe ‘SJ 
Pa Mtoe 


~ CENTRAL ASIATIC EXPEDITIONS OF THE AMERICAN MUSEUM 


OF NATURAL HISTORY, UNDER THE LEADERSHIP 
OF ROY CHAPMAN ANDREWS 


PRELIMINARY CONTRIBUTIONS 


IN 


GEOLOGY, PALEONTOLOGY 


AND ZOOLOGY 
1918-1925 


By R. C. ANDREWS 


J. T. NICHOLS C. C. Moox 

OuTRAM Bancs W. D. MatrHew 

H. W. Fow.er F. K. Morris 

W. GRANGER G. K. Nose 

C. P. BERKEY T. D. A. CockERELL 

W. K. Grecory K. P. Scumipt 

H. F. Osporn V. VAN STRAELEN 
G. G. SIMPSON 


G. M. ALLEN 


Reprints from the American Museum Bulletin and Novitates, also one from the 
Proceedings of the Biological Society of Washington, of the years 1918-1925 


VOLUME I 
Nos. | to 63 


List of additional articles, 1916-1925, 


not included in this volume 


4) 


‘ _ e ( 
rt ee ms 


4 co A 


- » » ae 
’ ‘ 
4 . , 


i 


TO THE BENEFACTORS OF THE 
CENTRAL ASIATIC EXPEDITIONS 


The Central Asiatic Expeditions of the American Museum of 
Natural History, under the leadership of Roy Chapman Andrews, 
actually began in the year 1916 with the first collecting tour of the 
Leader accompanied by Mrs. Andrews, into the western Chinese 
Province of Yun-nan. 

The exploration continued under. Doctor Andrews’ leadership in 
the First, Second and Third Asiatic Expeditions, and all the scientific 
results are now embraced under the general title of CenrrAL ASIATIC 
EXPEDITIONS. . 


Beginning in 1916, Doctor Andrews has contributed to Asia Maga- 
zine and other periodicals, as listed in the bibliography included in this 
volume, a series of articles on the organization, personnel, narrative 
and discoveries of the Expeditions, which will be comprised in 
Volume I, ‘ Narrative of the Expeditions,’ of the series of twelve volumes 
in which the scientific results of the Expedition will be republished in 
permanent form. Volume II of the series, ‘Geological Reconnaissance in 
Mongolia,’ covers the work of Messrs. Berkey and Morris to the end 
of the year 1925, and will appear during the present year. 


Meanwhile, sixty-three preliminary scientific papers, as listed in 
the bibliography of this volume, have been published, chiefly in the 
American Museum Bulletin and Nuvitates, reprints of which are collected 
and bound together in the present volume. This volume is issued 
especially for the convenience of specialists and investigators actually 
engaged in present researches in the same fields of Geology, Paleontology 
and Zoology. While designed for personal use, the volumes are addressed 
and donated to the institutions to which the recipients are attached, 
to become part of the permanent libraries of these institutions. 


A special branch of the AmerRIcAN Musrum LiBrary is to be de- 
voted to Central Asiatic exploration, especially in the regions covered 
by our expeditions, and libraries and individuals are invited to contrib- 
ute in return for the present volume, which is sent out in the spirit of 
international exchange and coéperation. 


Henry FarrFreELD OSBORN, 
President. : 


AMERICAN Musrum or NaTuRAL History, 
April 8, 1926. 


ee, 
out 
on & din f 
nv, v1 ate 


Kade nd Hw eau at ee htt i Lure 


pel 


20. 


21. 


ARTICLES INCLUDED IN THIS VOLUME 


Dedication. By Henry Fairfield Osborn. 
New Chinese Fishes. 1918. By John Treadwell Nichols. 


Description of a New Species of Serow from Yun-nan Province, China. 
1921. By Roy Chapman Andrews. 


The Birds of the American Museum of Natural History’s Asiatic 
Zoological Expedition of 1916-1917. 1921. By Outram Bangs. 


Description of a New Loach from North-eastern China. 1922. By Henry 
W. Fowler. 


Discovery of Cretaceous and Older Tertiary Strata in Mongolia. 1922. 
By Walter Granger and Charles P. Berkey. 


Protoceratops andrewsi, a Pre-Ceratopsian Dinosaur from Mongolia. 
1923. By Walter Granger and William K. Gregory. 


Later Sediments of the Desert Basins of Central Mongolia. 1923. By 
Charles P. Berkey and Walter Granger. 


Baluchitherium grangeri, a Giant Hornless Rhinoceros from Mongolia. 
1923. By Henry Fairfield Osborn. 


Description of a New Cyprinoid Fish from China. 1923. By Henry W. 
Fowler. 


New Chinese Bats. 1923. By Glover M. Allen. 


Titanotheres and Lophiodonts in Mongolia. 1923. By Henry Fairfield 
Osborn. 


Cadurcotherium from Mongolia. 1923. By Henry Fairfield Osborn. 


Two Lower Cretaceous Dinosaurs of Mongolia. 1923. By Henry 
Fairfield Osborn. 


Skull Characters of Alligator sinense Fauvel. 1923. By Charles C. Mook. 


New Fossil Mammals from the Pliocene of Sze-Chuan, China. 1923. 
By W. D. Matthew and Walter Granger. 


The Fauna of the Houldjin Gravels. 1923. By W. D. Matthew and Walter 
Granger. 


The Fauna of the Ardyn Obo Formation. 1923. By W. D. Matthew and 
Walter Granger. 


New Chinese Insectivores. 1923. By Glover M. Allen. 


New Bathyergide from the Oligocene of Mongolia. 1923. By W. D. 
Matthew and Walter Granger. 


Nine New Rodents from the Oligocene of Mongolia. 1923. By W. D. 
Matthew and Walter Granger. 


New Carnivora from the Tertiary of Mongolia. 1924. By W. D. Matthew 
and Walter Granger. 


¥ Mi att 1K 9 


‘ aha ay, y 


Pema nn irri 40x 


Aen’ 


22. 


23. 


24. 


25. 


26. 


27. 


28. 


29. 


30. 


31. 


32. 


33. 


34. 


35. 


36. 


37. 


38. 


39. 


40. 


41. 


Bibliography of Central Asiatic Expeditions 7 


New Inséctivores and Ruminants from the Tertiary of Mongolia, with 
Remarks on the Correlation. 1924. By W. D. Matthew and Walter 
Granger. 


A New Crocodilian from Mongolia. 1924. By Charles C. Mook. 


The Great Bathylith of Central Mongolia. 1924. By Charles P. Berkey 
and Frederick K. Morris. 


Psittacosaurus and Protiguanodon: Two Lower Cretaceous Iguanodonts 
from Mongolia. 1924. By Henry Fairfield Osborn. 


Sauropoda and Theropoda of the Lower Cretaceous of Mongolia. 1924. 
By Henry Fairfield Osborn. 


A New Spadefoot Toad from the Oligocene of Mongolia with a Summary 
of the Evolution of the Pelobatide. 1924. By G. K. Noble. 


Microtines Collected by the Asiatic Expeditions. 1924. By Glover M. 
Allen. 


Basin Structures in Mongolia. 1924. By Charles P. Berkey and Frederick 
K. Morris. 

Structural Elements of the Oldrock Floor of the Gobi Region. 1924. 
By Charles P. Berkey and Frederick K. Morris. 


The Peneplanes of Mongolia. 1924. By Charles P. Berkey and Frederick 
K. Morris. 

Three New Theropoda, Protoceratops Zone, Central Mongolia. 1924. 
By Henry Fairfield Osborn. 

Eudinoceras, Upper Eocene Amblypod of Mongolia. 1924. 
By Henry Fairfield Osborn. 

Andrewsarchus, Giant Mesonychid of Mongolia. 1924. 
By Henry Fairfield Osborn. 

Cadurcotherium ardynense, Oligocene, Mongolia. 1924. 
By Henry Fairfield Osborn. 

Serridentinus and Baluchitherium, Loh Formation, Mongolia. 1924. 
By Henry Fairfield Osborn. 

Fossils in the Ondai Sair Formation, Mongolia. 1924. By T. D. A. 
Cockerell. 

Some Fishes Collected by the Third Asiatic Expedition in China. 1924. 
By Henry W. Fowler. 

The Affinities of the Fish Lycoptera middendorffi. ,1925. By T. D. A. 
Cockerell. 

On Protoceratops, a Primitive Ceratopsian Dinosaur from the Lower 
Cretaceous of Mongolia. 1925. By William K. Gregory and Charles C. 
Mook. 


New Reptiles and a New Salamander from China. 1925. By Karl 
Patterson Schmidt. 


accor aia 


ore se a: i) 


42. 
43. 


44. 
45. 


46. 
47. 
48. 
49. 


50. 
51. 
52. 


53. 
54. 
55. 
56. 


57. 


58. 


59. 


60. 


61. 


62. 


63. 


Bibliography of Central Asiatic Expeditions 9 


Jerboas from Mongolia. 1925. By Glover M. Allen. 


Squirrels Collected by the American Museum Asiatic Expeditions. 1925. 
By Glover M. Allen. 


A New Homalopterin Loach from Fukien. 1925. By J. T. Nichols. 


An Analysis of Chinese Loaches of the Genus Misgurnus. 1925. By J. 
T. Nichols. 


The Two Chinese Loaches of the Genus Cobitis. 1925. By J. T. Nichols. 
Nemacheilus and Related Loaches in China. 1925. By J. T. Nichols. 
Homaloptera caldwelli, a New Chinese Loach. 1925. By J. T. Nichols. 


The Microstructure of the Dinosaurian Egg-Shells from the Cretaceous 
Beds of Mongolia. 1925. By Victor Van Straelen. 


New Chinese Amphibians and Reptiles. 1925. By Karl Patterson Schmidt. 
Some Chinese Fresh-Water Fishes. 1925. By J. T. Nichols. 


Hamsters Collected by the American Museum Asiatic Expeditions. 
1925. By Glover M. Allen. 


Some Chinese Fresh-Water Fishes. 1925. By J. T.. Nichols. 
Some Chinese Fresh-Water Fishes. 1925. By J. T. Nichols. 
Some Chinese Fresh-Water Fishes. 1925. By J. T. Nichols. 


Fauna and Correlation of the Gashato Formation of Mongolia. 1925. 
By W. D. Matthew and Walter Granger. 


New Creodonts and Rodents from the Ardyn Obo Formation of Mon- 
golia. 1925. By W. D. Matthew and Walter Granger. 


New Ungulates from the Ardyn Obo Formation of Mongolia. 1925. 
By W. D. Matthew and Walter Granger. 


New Mammals. from the Shara Murun Eocene of Mongolia. 1925. 
By W. D. Matthew and Walter Granger. 


New Mammals from the Irdin Manha Eocene of Mongolia. 1925. By 
W. D. Matthew and Walter Granger. 


The Smaller Perissodactyls of the Irdin Manha Formation, Eocene of 
Mongolia. 1925. By W. D. Matthew and Walter Granger. 


A Mesozoic Mammal Skull from Mongolia. 1925. By George Gaylord 
Simpson. 


Upper Eocene and Lower Oligocene Titanotheres of Mongolia. 1925. 
By Henry Fairfield Osborn. 


Fei 4h Vine Peel 
ae LP ee 1% eT Ned ee 
ce oy sie bind 
forks ae r 4 
e 


A List or Articues (1916-1925) ReLatinc To THE ASIATIC 
EXPEDITIONS OF THE AMERICAN Museum or NaturRAL History 
Nort INCLUDED IN THis VOLUME 


1916 


The Asiatic Zodlogical Expedition of the American Museum of 
Natural History. By Roy Chapman Andrews. The American Museum 
Journal, Vol. XVI, No. 2, pp. 105-106. 

Reproductions in Duotone of Asiatic Photographs, taken by Roy 
Chapman Andrews. The American Museum Journal, Vol. XVI, No. 
2, pp. 107-114. . 

Note—The American Museum’s Asiatic Zodlogical Expedition 
secures the services of Mr. Edmund Heller. The American Museum 
Journal, Vol. XVI, No. 3, p. 208. 

Note—Advices from Mr. Andrews dated Shanghai, China, May 18, 
1916. The American Museum Journal, Vol. XVI, No. 5, p. 334. 

Note—Mr. Roy Chapman Andrews reports on the progress of the 
Museum’s Asiatic Zodélogical Expedition. The American Museum 
Journal, Vol. XVI, No. 6, p. 411. 

Note—Mr. Roy Chapman Andrews writes from Li-kiang-fu, 
Yunnan Province, China, October 7, 1916, on activities of the Museum’s 
Asiatic Zodlogical Expedition. The American Museum Journal, Vol. 
XVI, No. 8, p. 547. 


: 1917 


Report from the Asiatic Zodlogical Expedition. By J. A. Allen. 
The American Museum Journal, Vol. XVII, No. 2, pp. 144-145. 

Note—Mr. Andrews reports to the American Museum that the 
Asiatic Zodlogical Expedition will return about the end of September, 
1917. The American Museum Journal, Vol. XVII, No. 6, p. 422. 

Chinese Pagoda, near Tali-fu, Yunnan, Photographed by Yvette 
Borup Andrews on the American Museum’s Asiatic Zoédlogical Expedi- 
tion. The American Museum Journal, Vol: XVII, No. 8 (Cover). 

Frontispiece, A Typical Goral Cliff in the Province of Yunnan, 
China. The American Museum Journal, Vol. XVII, No. 8, p. 508. 

Little-known Mammals from China. By Roy Chapman Andrews. 
The American Museum Journal, Vol. XVII, No. 8, pp. 509-524. 

Picturesque Yunnan. Reproduction in Duotone from Photographs 
taken by Yvette Borup Andrews. The American Museum Journal, 
Volk XVII, No. 8, p. 525. 


wary 855 
’ a 


nA i ia yr, p 
ve A; 
he i Gaal 


* 
AL 


Bibliography of Central Asiatic Expeditions 13 


1918 

Whaling Off Asian Shores. By Roy Chapman Andrews. Asia 
Magazine, Vol. XVIII, pp. 227-232. 

Tracking the Chinese Chamois. By Roy Chapman Andrews. Asia 
Magazine, Vol. XVIII, pp. 473-479. 

Note—Reverend Harry R. Caldwell and Professor C. R. Kellogg 
elected to life membership in appreciation of assistance rendered to the 
Asiatic Zodlogical Expedition in Yunnan. The American Museum 
Journal, Vol. XVIII, No. 1, p. 69. 

China’s Ancient Monuments. By Roy Chapman Andrews. The 
American Museum Journal, Vol. XVIII, No. 4, pp. 251-262. 

Letter from Rev. A. Kok, of Li-kiang-fu, Yunnan, China, to Presi- 
dent Osborn tells of cordial relations that exist between members of 
Asiatic Expedition and Chinese people where Expedition worked. The 
American Museum Journal, Vol. XVIII, No. 4, p. 314. 

The Blue Tiger. By Roy Chapman Andrews. Illustrations by 
Yvette Borup Andrews. The American Museum Journal, Vol. XVIII, 
No. 5, pp. 372-380. 

Note—Roy Chapman Andrews writes September 18, 1918, to 
President Osborn describing journey to Urga. The American Museum 
Journal, Vol. XVIII, No. 7, p. 621. 

Note—Second Asiatic Zoélogica! Expedition to continue for another 
year. The American Museum Journal, Vol. XVIII, No. 8, p. 727. 

Camps and Trailsin China. By Roy Chapman Andrews and Yvette — 
Borup Andrews. D. Appleton & Co., New York. 

Traveling toward Tibet. By Roy Chapman Andrews. Harper’s 
Magazine, Vol. 136, pp. 617-32. 

Frontier of the Forbidden Land. By Roy Chapman Andrews. 
Harper’s Magazine, Vol. 136, pp. 894-905. 

Camps in China’s Tropics. By Roy Chapman Andrews. Harper’s 
Magazine, Vol. 137, pp. 124-37. 


1919 
Across Mongolia by Motor-Car. By Roy Chapman Andrews. 
Harper’s Magazine, Vol. 139, pp. 1-16. ; 
Photographs by Roy Chapman Andrews, depicting scenes in Peking 
on the signing of the Armistice at Close of World War. Natural History, 
Vol. XIX, No. 2, pp. 229-232. 
Note—Roy Chapman Andrews discusses with Mr. Kungpah T. 


one ei) oy 
eg Ane 


Bibliography of Central Asiatic Expeditions 15 


King ways and means of coéperation. Natural History, Vol. XIX, No. 
2, p. 229. 

Note—Roy Chapman Andrews writes to the Museum regarding the 
speed of antelopes. Natural History, Vol. XIX, No. 3, p. 355. 

Shooting Whales in the Far East. By Roy Chapman Andrews. 
Asia Magazine, Vol. XIX, pp. 592-596. 

Exploring Unknown Corners of the ‘‘Hermit Kingdom.” By Roy 
Chapman Andrews. National Geographic Magazine, Vol. XXXVI, 
pp. 25-48, 30 ills., 1 page map. 


1920 


Note—By invitation of the American Asiatic Association, Mr. 
Andrews delivers two lectures at Carnegie Hall upon the American 
Museum’s Asiatic Expeditions. Natural History, Vol. XX, No. 2, p. 222. 

Note—Mr. Andrews returns from China after an absence of two 
years where he had been in charge of the Second Asiatic Zodlogical 
Expedition. Natural History, Vol. XX, No. 1, p. 110. 

Note—Pending publication of extensive article in Natural History 
describing work of the American Museum’s Second Asiatic Expedition 
to North China and Mongolia, Mr. Andrews gives notes regarding the 
material obtained. Natural History, Vol. XX, No. 3, pp. 340-341. 

Frontispiece, Where the Mongolian Bighorn Sheep arefound. From 
a photograph by Yvette Borup Andrews. Natural History, Vol. XX, 
No. 4, p. 348. 

New Expedition to Central Asia. By Roy Chapman Andrews. 
Natural History, Vol. XX, No. 4, pp. 349-355. 

In Mongolia and North China. By Roy Chapman Andrews. 
Natural History, Vol. XX, No. 4, pp. 356-373. 

Big Game Hunting at the Eastern Tombs. By Roy Chapman 
Andrews. Asia Magazine, Vol. XX, No. 8, pp. 795-799. 

A New Search for the Oldest Man. By Roy Chapman Andrews. 
Asia Magazine, Vol. XX, No. 10, pp. 945-949. 

Urga, the Sacred City of the Living Buddha. By Roy Chapman 
Andrews. Harper’s Magazine, Vol. 141, pp. 145-156. 

Lure of the Mongolian Plains. By Roy Chapman Andrews. 
Harper’s Magazine, Vol. 141, pp. 480-441. 


1921 


The Motor Truck in Central Asia. By Roy Chapman Andrews. 
Natural History, Vol. X XI, No. 1, pp. 69-70. 


Bibliography of Central Asiatic Expeditions 17 


Note—Regarding scientific staff of the Expedition during first and 
second years. Natural History, Vol. XXI, No. 1, p. 100. 

Note—Assistance rendered to Mr. Andrews by Rev. Harry R. 
Caldwell on First and Second Asiatic Expeditions. Mr. Caldwell sends 
to American Museum collection of mammals, including exceptionally 
fine tiger skin. Natural History, Vol. XXI, No. 2, p. 211. 

Digging for the Roots of our Family Tree. By Roy Chapman 
Andrews. Asia Magazine, Vol. XXI, No. 5, pp. 489-444. 

Walter Granger and the Third Asiatic Expedition. Natural His- 
tory, Vol. XXI, No. 3, pp. 320-321. 

Note—Prof. C. P. Berkey prepares to join Third Asiatic Expedition 
in the spring of 1922. Natural History, Vol. X XI, No. 3, p. 326. 

_ Note—Mr. Clifford Pope sails from San Francisco, Calif., May 31, 
1921, with Mr. Walter Granger, to join the Third Asiatic Expedition. 
Natural History, Vol. XXI, No. 3, p. 326. 

Note—The Third Asiatic Expedition begins important work in 
China with every promise of success. Cordial interest manifested by : 
Dr. Yen and other members of the Cabinet in Peking. Dr. V. K. 
Ting and Dr. J. G. Andersson have given invaluable aid. Natural 
History, Vol. X XI, No. 5, pp. 545-546. 

Note—Report of the Third Asiatic Expedition. Letters from Roy 
Chapman Andrews and Walter Granger to President Osborn. Natural 
History, Vol. XXI, No. 6, pp. 649, 650. 

Across Mongolian Plains. By Roy Chapman Andrews. Photo- 
graphs by Yvette Borup Andrews. D. Appleton & Co., New York. 

Hunting the Great Ram of Mongolia. By Roy Chapman Andrews. 
Harper’s Magazine, Vol. 142, pp. 328-338. 


1922 


Note—Mr. Roy Chapman Andrews prepares a full account of 
activities in China to be featured in early number of Natural History 
Mr. Andrews’ letter of December 20, 1921, quoted. Natural History 
Vol. XXII, No. 1, p. 85. 

Note—Mr. Roy Chapman Andrews sends to President Osborn a 
series of water color sketches of Chinese reptiles and amphibians made 
from living specimens secured by the Expedition. Natural History, 
Vol. XXII, No. 2, pp. 181-182. 

Note—Mr. Walter Granger, paleontologist of Expedition, has been 
engaged for some months in explorations on the Upper Yangtze-kiang. 
Extracts from letter. Natural History, Vol. XXII, No. 2, pp. 184, 185. 


¥ 


ys 


baw i bois 


aT a 


+. 


Bibliography of Central Asiatic Expeditions 19 


Politics and Paleontology. .By Roy Chapman Andrews. Asia 
Magazine, Vol. XXII, No. 5, pp. 360-3864; 400-401. 

Scientific Work in Unsettled China. By Roy Chapman Andrews. 
Natural History, Vol. XXII, No. 3, pp. 213-223. 

- Note—Inaugural Meeting of the Geological Society of China held 
in Peking, March 23, 1922. Natural History, Vol. XXII, No. 3, p. 286. 

Note—April 16, Message from Roy Chapman Andrews, announces 
departure of Expedition for Mongolia. Discovery of important Creta- 
ceous and Tertiary beds with fragmentary fossils of mammals and 
dinosaurs. Natural History, Vol. XXII, No. 3, pp. 276-277. 

The Third Asiatic Expedition of the American Museum of Natural 
History. By Roy Chapman Andrews. Science, N.S., Vol. LV, No. 1431, 
pp. 584-587. 

The Quest of the Golden Fleece. I. The Wilds of Shensi. By Roy 
Chapman Andrews. Asia Magazine, Vol. XXII, No. 6, pp. 440-446. 

The Quest of the Golden Fleece. II. Takin on Their Rugged 
Peaks. By Roy Chapman Andrews. Asia Magazine, Vol. XXII, No. 
7, pp. 515-520; 568. 

Hunting Takin in the Mountains of Shensi. By Roy Chapman 
Andrews. Natural History, Vol. XXII, No. 4, pp. 292-300. 

Note—Progress of the Expedition. Letter to President Osborn 
from Roy Chapman Andrews from Urga, quoted. Natural History, 
Vol. XXII, No. 4, pp. 374-375. 

Discovery of Cretaceous and Older Tertiary Strata in Mongolia. 
By Henry Fairfield Osborn. (Abstract.) Science, N. 8., Vol. LVI, 
No. 1446, pp. 291-293. 

Note—Professor Osborn to visit the Far East. Natural History, 
Vol. XXII, No. 5, p. 471. 

Proving Asia the Mother of Continents. By Henry Fairfield Osborn. 
Asia Magazine, Vol. XXII, No. 9, pp. 721-724. 

Note—Discoveries in Mongolia—Expedition reports extraordinary 
success from summer’s explorations in Mongolia. Mr. Granger, palzeon- 
tologist of the Expedition, secured complete skeletons of small Cretaceous 
dinosaurs, a skull of the giant hornless rhinoceros Baluchitherium and 
numerous other important specimens. Natural History, Vol. XXII, 
No. 6, pp. 569-570. 

Gobi—A_ Desert ‘‘Wonder-House.”’ By Roy C. Andrews and 
W. D. Matthew. Asia Magazine, Vol. XXII, No. 12, pp. 959-962; 
1000-1002; 1005. 

Current Palzontological Research in China. By J. G. Andersson. 


Me 
ge iy teeth Poy) we r st oan 
up Shae au bit aye rie bifs ks wail “is | 
vl iriisty rerlen Ty Miah, Uri an yey: sharin: Lanett G fie 
is | ire dace he 
“all er: 


hice : 


Bibliography of Central Asiatic Expeditions PA | 


Bulletin of the American Museum of Natural History, Vol. XLVI, Art. 
te, pp. 727-737. 

China as a Field for Scientific Research. By Roy Chapman 
Andrews. Bulletin of the Geological Society of China, Vol. 1, No. 1-4, p. 8. 

_ The Wider Significance of Paleontological Research in China. By 

Walter Granger. Bulletin of the Geological Society of China, Vol. 1, No. 
1H, p. 8. : 

Some Points of Interest in the First Three Weeks’ Work of the 
Third Asiatic Expedition. By Davidson Black, Bulletin of the Geological 
Society of China, Vol. 1, No. 1+, p. 37. 


1923 


Note—What the Gobi Desert has Yielded. In a cable sent by Roy 
Chapman Andrews to Asza Magazine and published in the December 
issue, Mr. Andrews summarizes the remarkable results obtained from 
five months’ work in the Gobi Desert. Natural History, Vol. XXIII, 
No. 1, pp. 90-91. 

Note—Vertebrate Fossils—A Forerunner of the Horned Dinosaurs. 
A skull casually discovered by J. B. Shackelford, photographer of the 
Expedition, while on a trip to Mongolia, proves to be a type ancestral 
to the ceratopsians known only from Upper Cretaceous. Dr. Gregory 
and Mr. Granger prepare paper to be published in Novitates. Natural 
History, Vol. XXIII, No. 2, p. 192. 

Note—Motion pictures of the Expedition to Mongolia. Preliminary 
demonstration in the American Museum of the pictures secured during 
the trip. Natural History, Vol. XXIII, No. 2, p. 193. 

Fossil Bones of the Baluchitherium. (Wrontiapieceys Asia Magazine, 
Vol. XXIII, No. 4, p. 240. 

Setting Out for the Buried Treasure of Mongolia. By Roy Chap- 
man Andrews. Asia Magazine, Vol. XXIII, No. 4, pp. 241-245; 288. 

A Paradise for Dinosaurs. By Roy Chapman Andrews. Asia 
Magazine, Vol. XXIII, No. 5, pp. 336-340. 

The Extinct Giant Rhinoceros Baluchitherium of Western and 
Central Asia. By Henry Fairfield Osborn. Natural History, Vol. 
XXIII, No. 3, pp. 209-228. 

Untying Red Tape in Urga. By Roy Chapman Andrews. Asia 
Magazine, Vol. XXIII, No. 6, pp. 420-424; 459-460. 

Field Explorations of the American Museum during the Year 1922. 
By Henry Fairfield Osborn. Science, N.S., Vol. LVII, No. 1485, pp. 
681-683. 


D ene ae ‘ ee 7 e 
ae og eo AK een tes 3 
\ ; My iN ; 


Whee 


j“ebe chide 


| hy 


Bibliography of Central Asiatic Expeditions 23 


Ancient Fauna of Mongolia Discovered by the Third Asiatic Expedi- 
tion of the American Museum of Natural History. By Henry Fairfield 
Osborn. (Address to the New York Academy of Sciences at a special 
meeting, May 21, 1923). Science, N.S., Vol. LVII, No. 1487, pp. 729-732. 

Note—The Third Asiatic Expedition Resumes Work in Mongolia. 
Natural History, Vol. XXIII, No. 4, pp. 406-408. 

Tenting in Lama Land. By Roy Chapman Andrews. Asia 
Magazine, Vol. XXIII, No. 7, pp. 516-520. 

A Kentucky Derby in the Gobi Desert. By Roy Chapman Andrews. 
Asia Magazine, Vol. XXIII, No. 8, pp. 547-552; 598-600. 

Baluchitherium, Largest Known Land Mammal. (Frontispiece). 
Fourth Restoration painted under the direction of Henry Fairfield 
Osborn by Charles R. Knight. Asta Magazine, Vol. XXIII, No. 9, 
p. 624. 

Giant Beasts of Three Million Years Ago. By Henry Fairfield 
Osborn. Asia Magazine, Vol. XXIII, No. 9, pp. 625-630; 676. 

Baluchitheres browsing in the fertile savanna uplands of Mongolia, 
three million years ago. Third restoration drawn under direction of 
Henry Fairfield Osborn by Elizabeth M. Fulda. Asza Magazine, Vol. 
XXIII, No. 9, p. 629. 

Speech delivered by Dr. V. K. Ting at dinner given by Geological 
Society of China, September 27, in honor of Professor Henry Fairfield 
Osborn. The Peking Leader, September 29. 

The Broader Aspects of the Third Asiatic Expedition. By Henry 
Fairfield Osborn. The Peking Leader, September 29, p. 4. Reprinted, 
Bulletin of the Geological Society of China, Vol. 2, pp. 100-103. 

Note—The Paleontology of China. Dr. A. W. Grabau, Geological 
Survey of China, engaged in the description of all collections of inverte- 
brate fossils brought back to Peking by the Third Asiatic Expedition. 
Dr. Berkey, under date of June 9, pays tribute to Dr. Grabau’s work in 
the field. Natural History, Vol. X XIII, No. 5, pp. 521-522. 

Note—Mr. Roy Chapman Andrews’ letter written in Camp Erhlien 
(Iren Dabasu), May 15, 1923, quoted. Natural History, Vol. XXIII, 
No. 5, pp. 525-527. 

Note—Dinosaur Eggs Discovered—Cable report reaches the 
American Museum of remarkable find of no less than seventy skulls 
and ten skeletons of primitive horned Ceratopsian dinosaurs. Natural 
History, Vol. XXIII, No. 5, p. 536. 

A Fossil-Hunter’s Dream Come True. By Roy Chapman Andrews. 
Asia Magazine, Vol. XXIII, No. 10, pp. 746-751; 770. 


Bibliography of Central Asiatic Expeditions 25 


Why Mongolia may be the Home of Primitive Man. An Address 
(to the Wen Yu Hui). By Henry Fairfield Osborn. The Peking Leader, 
. October 10, pp. 2,4: Reprinted, Columbia Alumni News, Vol. XV, No. 
16, pp. 254-256 (February 1, 1924). 

The Explorations of the American Museum of Natural History in 
China and Mongolia. By Henry Fairfield Osborn. Proceedings of the 
American Philosophical Society, Vol. LXII, No. 3, pp. 90-94. 

Winter-Cooled Ardor for Fossils. By Roy Chapman Andrews. 
Asia Magazine, Vol. XXIII, No. 11, pp. 838-843; 867-869. 

Mongolia May Yield the Secret of Man’s Origin. Reported by 
James C. Young, The New York Times, November 11. 

Note—President Osborn’s Trip to Asia. President Osborn, accom- 
panied by Mrs. Osborn, sailed from Seattle on 8.8. Madison, August 18, 
1923, for Shanghai. While in Asia took prominent part in activities of 
Expedition. Made trip to the Gobi. Natural History, Vol. XXIII, No. 
6, p. 622. ; 

The Second Year Work of the Third Asiatic Expedition. By Roy 
C. Andrews. (Abstract.) Bulletin of the Geological Society of China, 
Vol. 2, No. 3-4, pp. 103-105. 

Paleontological Discoveries of the Third Asiatic Expedition. By 
Walter Granger. (Abstract.) Bulletin of the Geological Society of China, 
Vol. 2, No. 3+4, pp. 105-108. 

Physiography of Mongolia. By F. K. Morris. (Abstract.) Bulletin 
of the Geological Society of China, Vol. 2, No. 3-4, p. 109. 


1924 


The Significance of Recent Discoveries in Mongolia. By Henry 
Fairfield Osborn. The China Journal of Science and Arts, Vol. II, 
No. 1, pp. 39-43. 

Where the Dinosaur Hid Its Eggs. By Roy Chapman Andrews. 
Asia Magazine, Vol. XXIV, No. 1, pp. 9-15; 82-83. 

Discoveries during the Season of 1923 by the Third Asiatic Expedi- 
tion in Mongolia. By Henry Fairfield Osborn. Proceedings of the 
National Academy of Sciences, Vol. 10, pp. 23, 24. 

Notes—Welcome by the Geological Society of China. An Estimate 
of the Mongolian Discoveries. The Geological Survey of China. 
Welcomed Home bya Dinosaur. Replicas of Baluchitherium distributed. 
Natural History, Vol. XXIV, No. 1, pp. 112-114; 118, 119. 

Along the Trail with the Editor. Asia Magazine, Vol. XXIV, No. 
2, p. 89. 


Bibliography of Central Asiatic Expeditions 27 


The Lure of Mongolia. By Roy Chapman Andrews. Asia Maga- 
zine, Vol. XXIV, No. 2, pp. 137-144. 

Recent Explorations in China and Neighbouring Regions. By 
Arthur De C. Sowerby. The China Journal of Science and Arts, Vol. 
Mets. 1, No. 2, pp: 131-132. | 

Basin Structures in Mongolia. By Charles P. Berkey and Frederick 
K. Morris. (Abstract.) Bulletin of the Geological Society of America, 
Vol: 35, No. 1, pp. 59, 60. 

Physiography of Mongolia. By Frederick K. Morris. (Abstract.) 
Bulletin of the Geological Society of America, Vol. 35, No. 1, pp. 87, 88. 

Fossils Mark Mongolian Waste as Cradle of the Human Race. 
By Henry Fairfield Osborn. The New York Herald, March 30, Section ~ 
i p. 4. 

Where Did Man Originate? By Henry Fairfield Osborn. Asia 
Magazine, Vol. XXIV, No. 6, pp. 427-431; 498, 499. | 

“Wind-Devils”’ and Warriors of the Gobi. By Roy Chapman 
Andrews. Asia Magazine, Vol. XXIV, No. 7, pp. 528-531. 

Andrews of Mongolia. By Louis D. Froelick. Asia Magazine, 
Vol. XXIV, No. 8, pp. 638-640. 

The Discovery of an Unknown Continent. By Henry Fairfield 
Osborn. Natural History, Vol. XXIV, No. 2, pp. 132-149. 

Living Animals of the Gobi Desert. By Roy Chapman Andrews. 
Natural History, Vol. XXIV, No. 2, pp. 150-159. * 

Geological Reconnaissance in Central Mongolia. By Charles P. 
Berkey. Natural History, Vol. XXIV, No. 2, pp. 160-173. 

Hainan. By Clifford H. Pope. Natural History, Vol. XXIV, No. 2, 
pp. 215-223. . 

Through the Yangtze Gorges to Wan Hsien. By Anna G. Granger. 
Natural History, Vol. XXIV, No. 2, pp. 224-235. 

Exploration for Fossil annie in Mongolia. (Abstract.) By 
W. D. Matthew. Bulletin of the Geological Society of cei Vol. 35, 
No. 1, pp. 187, 188. 

The Coming Five Years, 1924-1928, of the Third Asiatic Expedi- 
tion. By Roy Chapman Andrews. Natural History, Vol. XXIV, No. 2, 
pp. 256-257. 

Note—Addresses by Prof. Henry Fairfield Osborn in Peking, Sep- 
tember 21 to October 11, 1923. Natural History, Vol. XXIV, No. 2, 
pp. 260-265. 

Notes on the Mapping Program of the Third Asiatic Expedition 
in Mongolia. By Frederick K. Morris. The Geographical Review, Vol. 
XIV, No. 2, pp. 287-293. 


Bibliography of Central Asiatic Expeditions 29 


Explorations in the Desert Region of Central Asia. By Charles P. 
Berkey. Proceedings American Society of Civil Engineers, pp. 606-615. 

Geologic Explorations in the Gobi Desert. By Charles P. Berkey. 
Columbia Alumni News, issue of May 23. 

Eggs at $60,000 a Dozen. By Roy Chapman Andrews. The Satur- 
day Evening Post, May 24. 

American Men of the Dragon Bones. By Henry Fairfield Osborn. 
Natural History, Vol. XXIV, No. 3, pp. 350-365. 

Wintering over a Fire Basket in Szechuan Province. By Anna G. 
Granger. Natural History, Vol. XXIV, No. 3, pp. 366-380. 

Note on the Third Asiatic Expedition. Editorial Comments. The 
China Journal of Science and Arts, Vol. II, No. 5, p. 428. 

Asiatic Expeditions of the American Museum of Natural History. 
By Henry Fairfield Osborn. Nature, Vol. 114, No. 2866, pp. 504-507. 

Danger? (With illustrations). By Roy Chapman Andrews. Cos- 
mopolitan, pp. 42, 43; 180, 181, December. 

Flying Feet on Mongolian Hill and Plain. By Roy Chapman 
Andrews. Asia Magazine, Vol. XXIV, No. 12, pp. 975-979; 1018. 

Note—Roy Chapman Andrews returned in September to Peking 
from Urga, where he completed preliminary arrangements for the 1925 
phase of work. Other members of the Expedition will follow by motor in 
the spring. Asza Magazine, Vol. XXIV, No. 12, p. 937. - 


1925 


Note—Extract from letter of Roy Chapman Andrews: Things were 
lively in Urga when he was there in September. But he had no trouble 
_ and has arranged satisfactorily for work of the Expedition in the spring. 
Asia Magazine, Vol. XXV, No. 2, p. 97. 

Note—On March 7, Walter Granger and several members of the 
Asiatic Expedition sail from San Francisco, to meet Roy Chapman 
Andrews for work of the Expedition in Mongolia, starting from Peking, 
April 15, returning to Peking about September 20. Asza Magazine, 
Vol XXV, No: 3,.p. 187. 

Note—Archeological Research in Asia. By N.C. Nelson. Natural 
History, Vol. XXV, No. 3, p. 314. 

Note—The Structure of the Dinosaur Egg Shell. Natural History, 
Vol. XXV, No. 4, p. 422. 

Note—Letter from Mr. Clifford H. Pope, April 25, 1925. Natural 
History, Vol. XXV, No. 4, pp. 426-427. 

Note—Excerpts from letter of Roy Chapman Andrews, Asia Maga- 
zine, Vol. XXV, No. 10, p. 821. 


ef the ‘oo nf ie cs rn 


ai us pes be ate 


soit ee Si ite “aoe aad ror 
bs ; Me Ah bik I ae. cab Vin abut | 
bcs fi ii (% i I nena ean Nea oe sei ante 
oe my Se aaa Ba ma Wh Mitaateg 


pee: AOSD if Wet lits dnt ts - ay: Sr wry | Sia a 
aly Palio) Ae A ek Or ie = 

ULE! (ene ale AZ) ate lend gab 
Ms) 


Pe Ts nna ly ‘bel A 
. ey: 


. / indi 


‘ 
wwHiry? 
é 


we 
a 


i a! 
ey 


itt i 


Bibliography of Central Asiatic Expeditions 31 


Note on Third Asiatic Expedition. Natural History, Vol. XXV, No. 
6, pp. 622-623. 

Structural Nature and Origin of the Eastern Altai. By Charles P. 
Berkey and Frederick K. Morris. (Abstract.) Bulletin of the Geological 
Society of America, Vol. 36, No. 1, p. 133. 

Tectonic History of Central Asia. By Charles P. Berkey and Fred- 
erick K. Morris. (Abstract.) Bulletin of the Geological Society of 
America, Vol. 36, No. 1, p. 134. 

Central Asia in Cretaceous Time. By Charles P. Berkey and Fred- 
erick K. Morris. (Abstract.) Bulletin of the Geological Society of America, 
Vol. 36, No. 1, pp. 158, 159. 

Origin of Desert Depressions. By Charles P. Berkey and Frederick 
K. Morris. (Abstract.) Bulletin of the Geological Society of America, 
Vol. 36, No.-1, pp. 169-170. 

Protoceratops, a primitive Ceratopsian from the Lower Cretaceous of 
Mongolia. By W. K. Gregory and C.C. Mook. (Abstract.) Bulletin of 
the Geological Society of America, Vol. 36, No. 1, p. 228. 

Stratigraphy of China. By A. W. Grabau. Geological Survey 
of China. Peking. 

Palzogeographical Maps of Asia. By A. W. Grabau. Geological 
Survey of China. Peking. 


abt aT Sean 
‘¥ is 


Vol. 31, pp. 15-20 May 16, 1918 
PROCEEDINGS 


OF THE 


BIOLOGICAL SOCIETY OF WASHINGTON 


NEW CHINESE FISHES. 


BY JOHN TREADWELL NICHOLS. 


The American Museum of Natural History has recently ac- 
cumulated a small collection of fresh-water fishes from China. 
These have come from two widely separated sources, firstly from 
Yunnan-fu, Yunnan Province, collected by Mr. John Graham, 
secondly from Futsing, Fu-kien Province, collected by the 
Museum’s Asiatic Zoological Expedition of 1916-17, about 
July 25, 1916. 

Only one species is common to the two lots, the Goldfish, 
Carassius auratus (Linnaeus). 

Owing largely to small collections sent to the British Museum 
by Mr. Graham over a period of years, the fishes of Yunnan are 
comparatively well known. It will suffice merely to mention 
those species in the Yunnan lot already known from that local- 
ity: namely (Catfish) Silwrus mento Regan, Macrones medianalis 
Regan, Liobagrus nigricauda Regan; (Carps) Misgurnus anguil- 
licaudatus (Cantor), Nemachilus nigromaculatus Regan, Cyprinus 
carpio Linn., 2 barbelled form, Cyprinus micristius Regan, Barbus 
grahami Regan; (others) Ophiocephalus argus Cantor. Monopte- 
rus sp. (see beyond). 

Five species in the Fu-kien lot are well known fishes, namely 
(Carps) Acanthogobia maculatus (Bleeker), Zacco platypus (T. & - 
S.), known from Japan, Opsariichthys bidens Giinther; (others) 
Anguilla japonica T. & 8. Channa ocellata Peters. Mr. H. R. 
Caldwell contributes interesting data on Channa. This fish is 
abundant near sea level in the vicinity of Fu-chau, and also 
occurs, though in less numbers, at Yen-ping. It prefers stag- 
nant water and will travel across country to get from one pool 

5—Proo. BIoL. Soc. WasH., VOL. 31, 1918. (15) 


16 Proceedings of the Biological Society of Washington. 


to another or to invade therice fields. Natives call it nguok-la, 
‘“Moon-pike,’’ differentiating it from chau-la, “‘ Grass-pike ’’— 
Ophiocephalus. The reference is to the moon-like markings at 
the caudal base. It is a good food-fish. 

There remain to be considered at greater length two carps 
from Yunnan (a Hemiculter and Acanthorhodeus) apparently 
undescribed; two from Fu-kien, a Cobitis, and a fish allied to 
Leuciscus but with peculiar jaw,—and symbranch eels of the 
genus Monopterus from each locality. Monopterus from various 
parts of China is now referred to M. javanensis Lacépéde, but I 
find the material from Yunnan and Fu-kien separable, and see 
no reason why either form should be identical with’ that in 
Java, with the description of which neither agrees well. 


Cobitis dolichorhynchus, sp. nov. 


Resembles the striped loach, Cobitis taenia, which ranges from Europe 
to Japan, but is more elongate, especially the snout.. The type and only 
specimen No. 7026, American Museum of Natural History, was collected 
at Futsing, Fu-kien Province, China, by the Museum’s Asiatic Expedi- 
tion. It is 66 mm. long to base of caudal. Head 4.0 in this length, 
depth 5.8. Snout 2.0 in head, pectoral 1.5, ventral 1.7, longest dorsal 
ray 2.0, longest anal ray 1.9, caudal 1.5, depth of peduncle 2.0. Eye 3 
in snout. Interorbital 1.5 in eye. 

A strong, unequally forked spine recumbent in an elongate pit which 
extends forward from below the front of theeye. Body compressed, head 
and snout strongly compressed. Interorbital flat. Mouth small, well under 
the projecting snout, the gape reaching to below a point midway between 
tip of snout and nostril. A pair of barbels near the tip of snout, another 
at the end of the maxillaries, and a third in an intermediate position. 
Besides these six, a short barbel at the nostril. A broad, fleshy, weakly 
three-lobed membrane under the chin. Dorsal origin equidistant from 
chin and caudal base. Ventral slightly in advance of middle of dorsal. 
Pectoral reaching slightly more than half way to ventral, ventral slightly 
more than half way to anal, anal not reaching caudal by a distance equal 
to a third of head. Caudal rounded. Body covered with very small 
scales. Dorsal 9. Anal 7. 

- Color in spirits pale, darker on top of head, a dark streak from eye to 
snout, dark specks on the cheek. Irregular dark cross blotches occupy- 
ing most of the back. Below these a narrow dark stripe extends back- 
ward from the nape, broken and mixed with the dorsal blotches behind 
the dorsal fin. A row of more or less oval dark blotches from the center 
of the side behind the gill cover to the lower caudal base, a faint dusky 
shade connecting them. Between the blotches and the stripe above, a 
ragged dark streak extends backward, terminating in widely spaced 


Nichols—New Chinese Fishes. 17 


specks under the dorsal. A narrow oblique vertical inky-black blotch at 
the upper base of the caudal. Dorsal and caudal with dark bars. 


Georgichthys, gen. nov. 


Type, Georgichthys scaphignathus, sp. nov. 

A cyprinid fish with superficial resemblance to certain species of Leucis- 
cus. Mouth small, slightly inferior. Rami parallel and fused, the lower 
jaw rather broad and rounded at its end, flat above, covered in front and 
above with a longitudinally fluted membrane. Lips rather thick, con- 
fined to the sides of the lower jaw. No barbels. Scales moderate. Lat- 
eral line complete, in the center of the peduncle. Dorsal without spinous 
ray, its origin slightly in advance of ventral. Dorsal and anal fins short. 
Teeth in a single row, five, hooked. This interesting minnow is named 
for George Borup Andrews. 


Georgichthys scaphignathus, sp. nov. 


The type and only specimen, No. 7038, American Museum of Natural 
History, collected at Futsing, Fu-kien Province, by the Museum’s Asiatic 
Expedition, is 70 mm. long to base of caudal. Head 4.0 in this measure, 
depth 3.5. Snout 3.3 in head, eye 4.0, interorbital 3.1, maxillary 4.0, not 
reaching eye, pectoral 1.3, vental 1.4, longest dorsal ray 1.3, longest anal 
ray 1.5, caudal 1.0. 

Body moderately compressed, peduncle 1.4 times as long as broad. 
Ventral not reaching anal, anal not reaching caudal, which is moderately 
forked. Scales normal on the rounded belly and elsewhere, with con- 
spicuous radiating and concentric marking, 39, 446 between lateral line 
and dorsal, 3 between lateral line and ventral. Dorsal 9. Anal 8. 

Color in spirits dark on the back, the sides with extensive irregular 
dark blotches, front of dorsal and center of caudal lobes blackish, under 
surface of body, and fins otherwise pale. (scaphignathus—spade-jaw. ) 


Hemiculter andrewsi, sp. nov. 


Scales smaller than in others of the genus (see monograph by War- 
pachowski. Bull. Ak. Sci. St. Petersb. 32, 1888, pp. 15 to 23, and Nikolski 
Annre. Mus. St. Petersb. 8, 1903, p. 359, for an additional species. ) Teeth 
in 3 rows, compressed, and slightly hooked, 4-4-2. 

The type, No. 7038, American Museum of Natural History, was col- 
lected at Yunnan-fu, Yunnan, China, by Mr. John Graham. It is 137 
mm. long to base of caudal. Head 3.8 in this length, depth 4.2. Eye 
4.2 in head, snout 4.0, interorbital 3.5, maxillary 3.4, depth of peduncle 
2.8, pectoral 1.5, ventral 2.0, longest dorsal ray 1.9, longest anal ray 2.9, 
caudal 1.4. 

Elongate, moderately compressed, peduncle twice as long as deep. 
Mouth moderate, rather oblique, lower jaw distinctly projecting, maxil- 
lary not quite to under front of eye, no barbels. Gill-rakers numerous 
and slender (the longest 1g eye), about 45 on the first arch, backed by a 


18 Proceedings of the Biological Society of Washington. 


second row of shorter ones. Dorsal without a spine, its second ray soft 
and segmented though not divided, its origin a little behind ventral base, 
equidistant from base of caudal and front of eye. Pectoral pointed, 
reaching % to ventral, ventral a little more than half way to anal, anal 
not reaching caudal by a distance equal to half head, caudal deeply forked. 
Apparently a low keel on belly behind ventrals and none in front of them 
(all three specimens are so cut as to make this difficult of determination). 
Lateral line complete, dropping to the lower part of the side with an 
abrupt flexure over end of pectoral, rising gradually on the peduncle to 
terminate in the center of same. Scales of moderate size on the body 
becoming smaller posteriorly, about 75, 11 between lateral line and dor- 
sal, 3 between lateral line and ventrals. Dorsal 9. Anal 15. 

Brownish along the back, silvery elsewhere. Named for Mr. and Mrs. 
Roy Chapman Andrews in recognition of their recent zoological explora- 
tions in Yunnan. Besides the type we have two smaller specimens of 
114 and 116 mm. with the same data. The type is a female with eggs. 


Acanthorhodeus grahami, sp. nov. 


Close to Acanthorhodeus atranalis Gunther, but more slender. Depth 
3.0 to 3.3 (instead of 2.5); 416 scales between lateral line and vent (in- 
stead of 54g). Dorsal soft rays 11 to 138, anal 10 to 11. Scales 37 to 39. 
Teeth in one row, 5, slender, hooked, denticulate. 

The type, No. 7029, American Museum of Natural History, was col- 
lected at Yunnan-fu, Yunnan, China, by Mr. John Graham. It is 54 
mm. long to base of caudal. Head 4.0 in this length, depth 3.1. Eye 
2.7 in head, snout 4.1, interorbital 3.4, maxillary 3.5, depth of peduncle 
2.5, dorsal spine 1.6, longest ray 1.4, anal spine 1.7, longest ray 1.6, pec- 
toral 1.3, ventral 1.5, caudal about 1.0. 

Body compressed. Mouth oblique, lower jaw slightly projecting, max- 
illary not quite reaching front of eye, upper jaw strongly protractile, 
angle of mandible under front of eye, no barbels. Upper profile straight 
to nape, thence convex to dorsal, thence concave to caudal; lower profile 
oblique to angle of mandible, thence horizontal to gill opening, thence 
gently convex to anal, thence slightly concave to caudal; peduncle taper- 
ing, 244 times as long as deep. Vent just behind base of ventrals with a 
conspicuous papilla. Pectoral narrow and pointed, just reaching ventral, 
ventral just not reaching anal, dorsal and anal slightly concave above, 
caudal forked. Dorsal origin behind ventrals, midway between snout 
and base of caudal; anal origin under middle of dorsal. Second dorsal 
and anal rays developed as strong sharp spines, but not serrate. Scales 
loosely attached, lateral line complete, below the center of the front of 
the peduncle. Scales 38, 44g between lateral line and vent, 644 between 
it and dorsal. Dorsal I1 12. Anal II 10. 

Color pale. Sides silvery, the silvery forming an ill-defined stripe pos- 
teriorly, with a faint dark central streak. Most of the anal occupied by 
a large black semi-oval terminal blotch, which does not extend on the 
spines or posterior rays. 


Nichols—New Chinese Fishes. 19 


Besides the type, a male, we have 4 males and 3 females of about the 
same size with the same data. Females lack the black anal blotch, have 
the ventral papilla more prominent, and a long vermiform oviduct. 


Monopterus cinereus Richardson. 


Ichthyology Voyage of the Sulphur, 1844, p. 117, plate 52. Woosung. 
Not Pneumabranchus cinereus McClelland. See Richardson, Ichthyology 
of China. Rept. Brit. Assoc. Adv. Sci. 1845 (1846), p. 315. 
Measurements of Yunnan specimens are as follows: 


Total length. Head in length. Tail from vent in length. Eye in snout. 


420 mm. 12.4 S5i/ 20 
338 eA 3.6 ire 
336 12.0 3.6 1.5 


The two smaller specimens (and a third with imperfect tail of about 
the same size) agree with Richardson’s excellent figure in Voyage of the 
Sulphur. The larger specimen is pale colored with occasional blackish 
blotches. They lack evident fins. 


Monopterus xanthognathus Richardson. 


Voyage of the Sulphur, 1844, p. 118, plate 52. Canton. 
Measurements of Fu-kien specimens are as follows: 


Total length. Head in length. Tail from vent in length. Eye in snout. 
284 mm. 13.5 4.2 wee 
Wee 13.8 Be PAA 


These specimens differ from cinereus just as does Reeves’ figure on 
which Richardson bases his description, in a shorter tail, smaller eye, and 
a high gibbous nape. 


AMERICAN MUSEUM NOVITATES 
No. 6 


DESCRIPTION OF A NEW SPECIES OF SEROW 
FROM YUN-NAN PROVINCE, CHINA 


By Roy CHAPMAN ANDREWS 


Issued March 24, 1921 


By OrprER oF THE TRUSTEES 
OF 


THE AMERICAN MUSEUM OF NATURAL HISTORY 
New York City 


AMERICAN MUSEUM NOVITATES 


Number 6 March 24, 1921 


59.9, 735C (51.3) 
DESCRIPTION OF A NEW SPECIES OF SEROW FROM YUN- 
NAN PROVINCE, CHINA 


By Roy CHAPMAN ANDREWS 


Although it is not the purpose to publish extensively upon the collec- 
tions obtained by the Asiatic Expeditions of the American Museum until 
the field work has been completed, it is desirable to describe such new 
species as may come to light from time to time in the preliminary ex- 
amination of certain groups. 

The splendid serow which I shot near eee China, not far 
from the Burma frontier, is the first animal to be described frira the 
Asiatic collections, and I take pleasure in proposing for it the name 
Capricornis osborni, in honor of the distinguished President of The 
American Museum of Natural History, Professor Henry Fairfield Osborn. 


Capricornis osborni, new species 


Type No. 43042, @ juv., Hui-yao (20 miles from Teng-yiieh), Yiin-nan Province, 
China, May 9, 1917; Roy Chapman Andrews. 

Forehead, cheeks, neck, breast; and body coal black. The white basal parts of 
the hairs show through to a certain extent but the general effect is jet black. A nar- 
row margin of white 6 mm. wide on upper lips from middle of snout to corner of mouth. 
Above this white band but below and behind the nostril on each side, is a triangular 
tawny patch. The lower lip is margined with white which occupies all except the 
central half of the chin and extends behind the corner of the mouth in a long, gradually 
narrowing streak; this almost reaches the white throat-patch which is about 40 mm. 
in width. 

The proximal part of each ear in front is strongly tinged with tawny but on the 
back this is less pronounced and the ear is largely black. The short mane is intense 
black like the body, stiff, erect and crest-like; the hairs are about 120 mm. in length. 
From the mane to the tail, the hair of the mid-dorsal line forms a well-defined ridge. 
The tail is black in the center with an admixture of tawny hairs; the tip is all black. 

Inside of fore legs to hoofs, tawny; externally, black to the knees; front of 
“cannon bones”’ black except at the knees where the black is indistinct and suffused 
with tawny. Just above the hoofs, the blackish area is thinly sprinkled with light 
buff and posteriorly, between the dew claws and the hoof, it is all light buff. Buttocks 
tinged with tawny. Thighs almost to hocks, black with slight admixture of tawny. 
Inner side of hind legs to hoofs tawny. From hocks to opposite dew claws, anteriorly, 
the legs are tawny but with a suggestion of blackish, due to the hairs which are black 
on the basal half and tawny on the tips. From opposite the dew claws to the hoofs the 
black is pronounced and thinly interspersed with buff-tipped hairs. The area between 
the dew claws and the hoofs, posteriorly, is all buff. 


a 


2 AMERICAN MUSEUM NOVITATES [No. 6 


There is no underfur present on any part of the body. 

Skull badly broken. Measurements of skull: condylo-basal length, 257 mm.; 
least orbital width, 68; width of palate between first premolars, 37; length of horn 
on curve, 117; circumference of horn at base, 95. External measurements of type: 
head and body, 1350; tail, 180; hind foot, 390; ear, 175; height at shoulder, 950. 

Capricornis osborni is undoubtedly allied to our specimens from Li- 
chiang, Yiin-nan Province, which I have identified as C. milne-edwardszt. 
Its chief distinguishing characters are its coal-black body and head, its 
short black mane and the greater amount of black on the lower part of 
the legs. Our four specimens of C. milne-edwardsi all have brownish- 
black bodies and heads, long whitish manes, and little or no black upon 
the lower legs. In the very heavy mat-like gray mane, my two speci- 
mens of C. argyrochetes from Fukien Province, China, differ strikingly 
from osborni, although in the amount and disposition of the black on the 
lower legs the two somewhat resemble each other. C. swettenhami of the 
Malay Peninsula is distinguished from osborni by the black legs and the 
mane, which is a mixture of whitish, black and reddish hairs. 

In discussing C. milne-edwardsi Mr. R. I. Pocock! has remarked: 
“A closely allied form apparently resembling typical mzlne-edwardsz 
in color except that the fronts of the cannon bones appear to be black 
has been recorded by Mr. H. Shaw Dunn from Kyonklongyi and other 
localities in the North Shan States of Upper Burma where it lives mostly 
in evergreen forests at altitudes of from 4,500 to 6,000 feet (Field, Jan. 
9, 1909.).”’ 

I have not been able to discover Mr. Dunn’s communication in the 
‘Field’ but I have no doubt that the race I am now describing is the one 
to which he refers. 

The serow which Lieut. R. C. Beavan? described as inhabiting the 
vicinity of Moul-mein, Burma, and which Mr. Pocock referred provi- 
sionally to milne-edwardsi may be this new form. While the affinities of 
osborni are toward milne-edwardsi, it is interesting as showing an ap- 
proach toward swettenhami of the Malay Peninsula in the considerable 
amount of black on the legs and the short black mane. 

Near Genkang, Yiin-nan Province, we purchased from a native a flat 
serow skin which lacks the head and lower legs. This specimen was said 
to have come from the mountains of Keng-ma about 200 miles southeast 
of Teng-yiieh and not far from the Burma frontier. It is brownish black 


l The Serows, Gorals and Takins of British India and the Straits Settlements.’ By R. I. Pocock. 
Part II. Journal, Bombay Natural History Society, XXII, pp. 307-308. 


21866, Proc. Zool. Soc. London, p. 4. 


1921] A NEW SEROW FROM CHINA | 3 


in general color, has a short crest-like, brownish-black mane, similar in 
character to that of osborni, and what remains of the skin shows that both 
the fore and hind legs were whitish or light buff below the knees and 
hocks. 

This specimen may possibly represent the male of C. osborni, for 
the differences are somewhat similar to those between the male and 
female of our C. argyrochetes from Fukien. 

I shot C. osborni near the village of Hui-yao while hunting monkeys 
on the precipitous bank of the river. The cliff was almost perpendicular 
and was covered with a tangled jungle growth. Now and then the rock 
wall would become less precipitous and the thick cover give place to an 
open grassy slope. It was when I was about to cross such an opening 
that the serow dashed out of the bushes where it had evidently been 
feeding. I fired just before it disappeared over the rim of the gorge and it 
sank in its tracks, gave a convulsive twist, and plunged into the canyon. 
It was recovered with considerable difficulty. 

Although the natives knew that serows lived in this part of the gorge, 
few of them had ever seen one and it was an object of great curiosity in 
the village. 

; There is little change in the country between Hui-yao and the Burma 
frontier and no reason why C. osborni should not have an unrestricted 
range into Burma. 


4 


oe Cert yore A ees Lee ee 
“The American Museum of Nat mer History 
. | Frane B. berg paee 


ihe : York City 
“Mammatocy 


- 201. Riots. Collected on: sie Roosevelt -Biizitian ‘Expedition with’ 
- ‘Field Notes by Leo E. Miller. ~ By J. “A, Allen, 1916, - Bulletin, © 
Sp ee ho SR RV Art! 30, pie BognaL0, 2 Re 
~ 202. Shrews. Collécted bv the Congo Rxpedition of the Shek be Wedecuiti.?? 
ses ene oor N. Hollister, 1916, Bulletin, XKXYV, at 35, ‘PP. 663-680,- « 
pa Pete S, VII-XI, 2) < 
203: The Sei’ Whale (Balenoptera harealex Lesson). ‘By Roy ‘Cy “Andrews. 
‘and Hermann von. W. Schulte, 1916, Memoirs, N. 8. fT pare: 6," 


riche Oa Dae pp. 289-502, Pls. xxrx-Lvm, 48 text figures, ./ 
2042 A New Rabbit and a New Bat from Neotropical Regions. | By 5 soe 
ene bee Zanes 1917, ‘Bulletin, Darcie ee Art, 13, pp. 335-837, Ply. 


aos : 2085. The Skull « of Kogia breviceps Blainville, By H. Fen Schulte. i917, 
Be es. Bulletin, = SOC VS “Att AY,” pp: 361-404, Bis? XXXV-XEIE, 


; 2 text figures. 
206. The American Museum Cnn Daveditien Guliorbinn of Bats. By ators 
Lae Fy J. A. Allen, Herbert Lang, and James P. Chapin, 1917, Bulletin, ~ e 
rosea ‘Art. 18, pp. 405-963, Pls, AOE 26 text figures, a 50 


© 2207, The. Selceal: Chatsckane of Scutisorex Thomas. By J. A. Allen, ‘with 
; Field Notes by H: Lang. 1917, Bulletin, XXXVII, Art. 28, PP. 
: 769-784, Pls. EXxxIx—xcu, 8 text figures, 3 
208. A-Note on the Lumbar Vertebre of Scutisorex Thomas. By H. von 
gi oss W. Schulte, 1917, Bulletin, XXXVI, Art.-29, pp. 785-792. es 
209. -The External Characters, Skeletal Muscles, and ‘Peripheral Nerves ~ z 
of Kogia breviceps (Blainville). ‘By. H.;von W. Schulte and - : 
M. de Forest Smith, 1918, Bulletin, XXXVI, Art. 2, PP: 7-72, 
hes 21 text figures. . — 
210... Memoranda Upon the Anatomy: of the Respiratory Brant, Foregut,. 
: and Thoracic Viscera of a Foetal: Kogia breviceps. By John D. 
~ Kernan, Jr: and H. von W: Schulte, 1918, Bulletin, AXXVUL, . 
Rie Soe Art..8, pp. 231-267, 16 text figures, . ~~ 
241. The Skull of LZiphius cavirostris: By John D. Kernan, 1918, Bulletin, : 
XXXVITI, Art. 11, pp. 349-394, Pls. xx-XxxIl > f 
212.°.The Indigenous Land Mammals of Porto Rico, Living and- Extinct. 
~ By H.#. Anthony. 1918, Memoirs; N.S:, If, part pp- 831-435, - 
poe Pls. tv-uxxtv, 55 text figures. 
_-213.° ‘Severtzow's Classification of: the Felidw.. “By: J. ‘A, Allen, 1919, 
ss ~» Bulletin, XLI, Art. 6, pp. 335-340. 
214. Notes on the Synonymy and Nomenelature of the Smaller ‘Spotted 
i Cats of Tropical America. By J. A. Allen, 1919, Bulletin, XU, 
;  -Art.7, pp. 341-419, 31 text figures. “100. 
-215;. Mammals ‘Collected in Eastern Cuba in 1917, with Descriptions Of = ee ee 
ae Nae Two New Species. By H. E. Anthony, 1919, Bulletin, XU, @ ee 
i : Art..20, pp. 625-643, Pls. xxxv-xXXXVII. oUe: 
216. New Mammals from Jamaica. “ By H.E. Derioias, 1920, Bulletin, has 
Sees ee nee 12, pp: 469°475, Pl. XXXuI, a teat REE Phas at aL Oe 


59.82(5) 
Article XX—THE BIRDS OF THE AMERICAN MUSEUM OF 
NATURALHISTORY’S ASIATIC ZOOLOGICAL 
EXPEDITION OF 1916-1917 


By Outram Bancs 


The birds collected by Roy Chapman Andrews and Edmund Heller 
in Burma, Yunnan and Fokien during the course of The American Mu- 
seum of Natural History’s Asiatic Zoological Expedition of 1916-1917, 
were very kindly placed in my hands for identification by Dr. Frank M. 
Chapman, and I now take pleasure in submitting the following annotated 
list. . ; 

The collecting of mammals was the primary object of the expedi- 
tion and birds were to some extent a secondary consideration, which 
accounts for the short series, many of the species being represented by 
only a single specimen. The shortness of the series renders subspecific 
identification in a few instances a matter of some uncertainty. 

The Expedition traveled along the border of Burma and Yunnan, 
and in western Yunnan to the Snow Mountains, and made one trip east- 
ward to Yunnan Fu. An interesting detailed account of the wanderings 
and experience of the members of the party, with descriptions of all sta- 
tions at which specimens were collected, and illustrated by many photo- 
graphs and a sketch map of the route, has been published by Mr. 
Andrews. 

The collection of birds made in Fokien was supplemented by a 
series of skins received on the spot from Rev. Harry R. Caldwell. This 
collection is included here with the others. 

I have kept the birds from Fokien apart from those of Burma and 
Yunnan, giving two separate lists. This has caused very little repeti- 
tion of names and has made no faunal confusion, which would have been 
the case had all been listed together. 

Long after this paper was originally written Lord Rothschild pub- 
lished an article (Novitates Zoologicee, XX VIII, pp. 14-67, May 1921) 
on a collection of birds numbering 1442, made in 1918 and 1919 by George 
Forrest in west-central and north-western Yunnan. 

Many changes in current names occur in this article and, as might be 
expected, much that I had said in my original MS. is anticipated. All 
this has now been rewritten and so far as the material in the Andrews 
Collection will allow I follow Lord Rothschild’s opinions. 


“Camps and Trails in China.’ Appleton « Co. New York, 1918. 
575 


576 Bulletin American Museum of Natural History [Vol. XLIV 


BIRDS FROM THE BURMA BORDER AND YUNNAN 
PHASIANIDZE 
Francolinus pintadeanus phayrei (Blyth) 

Two males: Malipa, Burma, March 13, 1917; and Namting River, 
Burma border, March 4, 1917. These skins agree with one from 
Mengtsz, Yunnan, and one from Siam in the Museum of Comparative 
Zoology in being much smaller in all dimensions than specimens from 
southern China (Fokien, ete.). The wing ranges from 138 to 143 mm., 
with all other measurements proportionately small. I cannot detect 
any constant differences in color in the two races. Blyth’s type was 
from Arakan, and his description reads: ‘Closely resembling in plum- 
age the Pintado Partridge of the Mauritius, Francolinus perlatus, but 
of a less robust form and the male armed with well-developed spurs.” 
The measurements given lately by Robinson and Kloss (1919, Ibis, (11) 
I, July, p. 408) for birds from South Annam are also small. 


Arboricola torqueola (Valenc.) 


One adult female: No-mu-shu Pass, Yunnan, 8000 feet altitude, 
April 7, 1917. The oviduct contained eggs nearly ready for laying. 


Arboricola brunneipectus brunneipectus Tickell 
One adult male: Namting River, Burma border, March 2, 1917. 


Coturnix coturnix japonica Temminck and Schlegel 
_ One adult female: Malipa, Burma, March 13, 1917. 


Bambusicola fytchei fytchei Anderson 


Two adult females: Mu-cheng, Salwin drainage, February 10, 
1917; Teng-Yueh, Yunnan, April 22, 1917. 

These skins like Rothschild’s are wholly referable to Bambusicola 
fytcher fytcher Anderson, type locality Pouse, western Yunnan, and differ 
from B. f. hopkinsont Godwin-Austen of Assam, etec., as pointed out by 
Rothschild. 

Tam loath, however, to throw the Mengtsz bird B. f. oleaginea Bangs 
and Phillips into the synonymy of true fytchei as Rothschild was inclined 
to do. The type, to be sure, is the only individual I have seen, but it 
differs more from fytchez than does fytchet from hopkinsoni. The spots 
on the upper parts are much blacker, these black spots extending even 
all over the hind neck; the ground color of the upper parts is darker olive 


. 


1921] Bangs, Birds of the Asiatic Expedition 577 


and the top of the head much darker; all the wing coverts, the scapulars 
and the back are much more uniform in color, hardly at all varied with 
paler, grayish cross markings and vermiculations; the chest is darker 
and very uniform—but little spotted. 

Until a series from Mengtsz, which might, of course, prove the type 
to be an exceptional specimen, is available, I prefer to recognize three 
forms. 


Tragopan temmincki (J. EH. Gray) 
One adult male: Ho-mu-shu Pass, Yunnan, 8000 feet altitude, 
April 8, 1917. 


Gennaeus nycthemerus ripponi Sharpe 

One fine adult, male: Ho-mu-shu Pass, Yunnan, 8000 feet altitude, 
April 7, 1917. 

Phasianus colchicus elegans Elliot 

Three adult males: Li-chiang Fu, Yunnan, 9000 feet altitude, 
October 25 and November 4, 1916. 

Rothschild regretted his inability to compare Szechwan and Yunnan 
adults of this species and so be sure that P. elegans and P. sladeni 
(nomen nudum) were identical. This I have done, comparing four adult 
males from Szechwan with four adult malesfrom Yunnan, and can find no 
constant differences either in size or in color. 


Calophasis humiz Hume 
One female: Teng-yueh Ting, Yunnan, April 22, 1917. 


Chrysolophus amherstie (Leadbeater) 
Four adult males: Wan-tien, 7000 feet; Pei-ti-ping, Mekong River 
drainage, 9000 feet; and Li-chiang Fu, 11,000 feet altitude, November 
and December 1916 and May 14, 1917. 


Gallus gallus ferrugineus (Gmelin) 

Five specimens, three adult males, two adult females; Chang-lung, 
Salwin River, Yunnan; Namtung River, Burma border; and Malipa, 
Burma, February and March 1917. The distinction Stuart Baker (1917, 
Journ. Nat. Hist. Soc. of Bombay, XXV, p. 18) makes between the 
Indian and the Chino-Malayan Jungle Fowls is shown to a certain 
extent in the series of birds I have examined, though I find it very difficult 
to distinguish some specimens. 


578 Bulletin American Museum of Natural History [Vol. XLIV 


Pavo muticus Linnzus 


One adult female: Chang-lung, Salwin River, Yunnan, 2000 feet 
altitude, March 21, 1917. The oviduct contained partly formed eggs. 


TURNICIDE 

Turnix pugnax rostrata Swinhoe 
One adult female: Chu-tung, Yung-ping Ho, Yunnan, 5000 feet 
altitude, January 17, 1917. 


CoLUMBIDE 
Columba hodgsonii Vigors 


Three males: Chang-lung, Yunnan, 2000 feet altitude, and Malipa, 
Burma, March 1917. 


Streptopelia orientalis orientalis (Latham) 


Three specimens, two males and a female: Ho-mu-shu Pass, Yun- 
nan, 8000 feet altitude, April 8, 1917, and Malipa, Burma border, 3000 
feet, March 10 and 14, 1917. 

Rothschild referred his one specimen to this form without comment. 
Our three skins show a decided approach to S.-o. agricola (Tickell); 
No. 143299 from Ho-mu-shu Pass in particular. This specimen, I 
think, might almost as well be referred to one form as to the other. 


Streptopelia chinensis tigrina (Temminck and Knip) 


One adult female: Namting River, Burma border, February 28, 
1917. This example is perfectly typical. 


RALLIDZ 
Amaurornis phoenicura chinensis (Boddaert) 


Three adults, two males and a female: Namting River, Burma 
border, March 4, 1917; Malipa, Burma, March 14, 1917; and Meng- 
peng, Salwin drainage, March 17, 1917. 


CHARADRIIDA 
Hoplopterus ventralis (Wagler) 
Two adult females: Meng-ting, Yunnan, February 16 and 17, 
1917. 
Scolopax rusticola rusticola Linnzus 
One male: Namting River, Burma border, March 1, 1917. 


1921] Bangs, Birds of the Asiatic Expedition 579 


ARDEIDZ 
Bubulcus ibis coromandus (Boddaert) 
One immature female: Lung-ling, Yunnan, March 28, 1917. 


FALCONIDZ 
Lophospiza trivirgatus rufitinctus (McClellan) 


One male, Namting River, Burma border, February 24, 1917. This 
specimen has a wing of 230 mm. 


Spilornis cheela ricketti Sclater 


One adult male: Malipa, Burma border, March 14, 1917. This 
is a large bird with a wing of 455 mm. and without much doubt belongs 
to this race lately described by Sclater. 


Cerchneis tinnunculus saturatus (Blyth) 
One ‘‘male”’ (female): Hung-chang, Yunnan, January 28, 1917. 
; BUBONIDZE 
- Glaucidium cuculoides cuculoides (Gould) 
One adult female: Namting River, Burma border, March 1, 1917. 


CORACIIDA 
Coracias affinis McClellan 


Three specimens, two males and a female: Hsiao, Meng-ting and 
Cheng-kang, Salwin drainage, February 3 and 6; and Shui-chai, Mekong 
River, Yunnan, January 19, 1917. All three are large birds like those 
from the eastern Himalayas, with wings ranging from 193 to 196 mm. 
Specimens in the Museum of Comparative Zoology from Cochin China 
are smaller, as also is one listed by Kloss from Siam. The smaller form 
of Cochin China and Siam, if really separable, should be known as C. 
affinis theresiz Parrot. It, however, was not recognized by Kloss (1918, 
Ibis, (10) VI, January, p. 91), nor by Robinson and Kloss (1919, Ibis, 
(11) I, July, p. 421). 


ALCEDINID 
Halcyon smyrnensis fusca (Boddaert) 
One adult female, Meng-ting, Burma border, February 18, 1917. 


580 Bulletin American Museum of Natural History [Vol. XLIV 


BUCEROTIDZE 
Anthracoceros malabaricus affinis (Blyth) 
Two specimens, an adult male and.an immature (sex not deter- 
mined): Namting River, Burma border, February 28, 1917. These birds 


belong to the large Himalayan form, the adult male having a wing of 
308 mm. 


UPUPIDE 
Upupa epops saturata Lonnberg 
One adult male: Yung-chang Fu, Yunnan, January 28,1917. This 


is a large bird, with a wing of 154 mm. In size, as well as in other re- 
spects, it is an extreme example of the northern saturata. 


MEROPIDZ 
Melittophagus leschenaulti swinehoei (Hume) 


Three adults, a male and two females: Chang-lung, Salwin River, 
Yunnan, March 18, 19, and 21, 1917. 


TROGONIDE 
Pyrotrogon erythrocephalus erythrocephalus (Gould) 
One adult male: Namting River, Burma ea March 30, 1917. 
Rothschild referred a male and a female’from Shweli-Salwin Divide 
to P. e. yamakanensis (Rickett) of Fokien. Our specimen certainly does 
not represent that form, of which I have seen one fully adult male. I 
have no hesitation in calling it true erythrocephalus. 


CUCULIDE 
Cuculus canorus bakeri Hartert 

Two adult males: Teng-yueh Ting and Wa-hui, Yunnan, April 22 
and May 16, 1917. On the label of one (killed April 22 at Teng-yueh 
Ting) is written: ‘‘ Note ku-ku-calling throughout the day.’’ These two 
specimens seem to me to be bakerz. They are quite as heavily barred 
below as in C. canorus canorus, and the color ot the upper parts is as 
dark as in C. optatus. They have small bills, smaller than in C. optatus. 


Centropus sinensis intermedius (Hume) 

Five specimens, both sexes: Namting River, Burma border; Chang- 
lung, Salwin River; Meng-ting, Yunnan; February 18, 22, and 28, and 
March 2 and 22, 1917. These specimens undoubtedly belong to the 
smaller form; the wing in the four females ranges from 205 to 216 mm. 
In the single male it is 200 mm. 


1921] Bangs, Birds of the Asiatic Expedition 581 


Rhopodytes tristis tristis (Lesson) 

One adult male: Chang-lung, Salwin River, Yunnan, March 20, 
1917. This specimen, with a wing of 163 mm., I refer to the larger 
northern form. 

CAPITONIDZ 
Cyanops asiatica Latham 


Two specimens, male and female: Chang-lung, Salwin River, 
Yunnan, March 20 and 21, 1917. 


Cyanops franklini franklini (Blyth) 
One adult male: Tai-ping-pu, Yunnan, April 12, 1917. 


Xantholaema hemacephala indica (Latham) 
One adult male: Namting River, Burma border, February 28, 1917. 


PICIDE 
Picus canus sordidior (Rippon) 


Three adults, two males and a female: Hui-yao, Yunnan, 5000 and 
5500 feet altitude, May 7, 1917; and Malipa, Burma border, 3200 feet 
altitude, March 14, 1917. 


Hypopicus hyperythrus subrufinus (Cabanis and Heine) 


Two adult males: Li-chiang Fu, Snow Mountains, Yunnan, 10,000 
feet altitude, November 16, 1916. 


Dryobates pernyi pernyi (Verreaux) 
One adult male: Li-chiang Fu, Snow Mountains, Yunnan, 10,000 
feet altitude, November 16, 1916. 


Chrysocolaptes guttacristatus sultaneus Hodgson 


Two adult males: Malipa, Burma border, February 22, 1917. 
These are large birds, with the wings 176 and 178 mm. respectively and 
with heavy bills, and must, therefore, I suppose, be referred to this form. 


Thriponax javensis feddeni (Blanford) 
One adult male: Malipa, Burma border, March 15, 1917. 


Jynx torquilla japonica Bonaparte 
One adult female: Yung-chang Fu, Yunnan, January 28, 1917. 


582 Bulletin American Museum of Natural History [Vol. XLIV 


EURYLAIMIDZ 
Serilophus lunatus lunatus Gould 


One male: Meng-ting, Burma border, February 19, 1917. Appar- 
ently referable to this form. 


HIRUNDINIDZ 
Hirundo rustica gutturalis Scopoli 
One adult male: Meng-ting, Burma border, February 18, 1917. 


Hirundo daurica nipalensis Hodgson 


One adult, sex not determined: Meng-ting, Salwin drainage, Yun- 
nan, February 19, 1917. 


Ptyonoprogne rupestris (Scopoli) 


One female: Chen-kang, Salwin drainage, Yunnan, February 6, 
1917. 


MUSCICAPIDA 
Cyornis tickellie whitei Harington 


One adult male: Chang-lung, Salwin River, Yunnan, 2000 feet 
altitude, March 21, 1917. 


Niltava sundara denotata Bangs and Phillips 

Three adult males: Chang-lung, Salwin River and Tai-ping-pu, 
Yunnan, March 18 and 20, and April 9, 1917. These skins exactly match 
the type of the subspecies and differ from true N. sundara in having the 
back blacker, less purplish, and the under parts much paler and yellower 
and in longer wing. 

Rothschild hesitates to recognize denotata, but again going over all 
material available to me I am still inclined to do so. 


Muscicapula melanoleuca melanoleuca (Blyth) 
Four specimens: Tai-ping-pu, Yunnan, April 12 and 13, 1917. 


Rhipidura albicollis albicollis (Vieillot) 


Three adults, two males and a female: Namting River, Burma 
border, and Mu-cheng, Salwin drainage, Yunnan, February 25 and 
March 5, 1917. 


1921] Bangs, Birds of the Asiatic Expedition 583 


As well as I can determine these skins with the limited Indian 
material available to me, which does not include an example from the 
region assigned by Baker to his form, stanleyi, they do not belong to the 
northern and north-eastern race, characterized by Baker (1913, Records 
of the Indian Museum, VIII, part 3, September, p. 275) as Rhipidura 
albicollis kempi, new subspecies. The specimens recorded by Phillips 
and myself from Mengtsz are quite the same in color as those in the 
present collection. 

The name kempi given by Baker was preoccupied by Rhipidura 
rufifrons kempi Mathews (1912, Nov. Zool., X VIII, January, p. 320) and 
was changed later by Baker to stanley. It was, however, again used for 
still another bird, Rhipidura flabellifera kempi Mathews and Iredale of 
North Island, New Zealand (Ibis, 1913, p. 441), which may be called 
Rhipidura flabellifera placabilis, new name. 


Culicicapa ceylonensis ceylonensis (Swainson) 
Three adults, two male and a female; Namting River and Malipa, 


Burma border, and Tai-ping-pu, Yunnan, February 22, March 16, and 
April 12, 1917. 


Stoparola thalassina thalassina (Swainson)! 


One adult male: Mu-cheng, Salwin drainage, Yunnan, 5000 feet 
altitude, February 15, 1917. 


CAMPEPHAGIDE 
Pericrocotus speciosus speciosus (Latham) 


One adult male: Ta-shui-tang, Salwin drainage, Yunnan, 6000 feet 
altitude, February 2, 1917. 


Pericrocotus yvettz,” new species 


Two specimens, an adult male and an adult female: Malipa, Burma 
border, and Taiping-pu, Yunnan, March 10 and April 12, 1917. 

Typr.—Amer. Mus. of Nat. Hist. No. 143365; adult male; Malipa, Burma 
border, 3000 feet altitude, March 10, 1917; R. C. Andrews and E. Heller. 

Craracters.—Adult male similar to the adult male of P. xanthogaster xantho- 
gaster (Raffles) and with the four outer primaries without red but slightly larger, 
and red on secondaries continuous from base to near tip on outer webs of feathers 


1For change in the specific name from melanops Vig. to thalassina (Swainson), see Oberholzer, 
1919, Proc. Biol. Soc. Wash., XXXII, p. 240, December 31. 
2Named in honor of Mrs. Andrews. 


584 Bulletin American Museum of Natural History [Vol. XLIV 


(in P. xanthogaster the red on the secondaries is arranged in spots near the tips of 
the feathers which are separated from the red bases by black); red of under parts 
nearly scarlet-red (nearly scarlet in P. xanthogaster). Adult female similar to the 
female of P. xanthogaster. 


MEASUREMENTS 
A. M.N.H.No. Sex Wing Tail Tarsus Culmen 
143365 ; Oo ad. 94 mm. 105 mm. 14.5 mm. 11 mm. 
143367 @ ad 87 94 14.0 Tal 


I feel a little hesitation in describing this form on the strength of one 
adult male, still I have been unable to match this specimen at all nearly, 
nor can I find the description of a Minivet which could apply to this one. 
The female, which I refer to this species because it certainly is not the 
female of brevirostris or speciosus, I cannot distinguish from females of P. 
zantogasier from Sumatra in color or markings, except that in this one 
example there is no yellow on the secondaries except at the base. 

The red markings in the wing of the male are very striking, giving 
two broad red stripes down the closed wing, with black between them, 
one along the primaries, the other along the secondaries. 


Pericrocotus brevirostris ethologus Bangs and Phillips 


One adult male: Tai-ping-pu, Yunnan, April 12, 1917. This is an 
intensely colored individual not extremely typical of this form. When 
Phillips and I (1914, Bull. Mus. Comp. Zool., LVIIT, No. 6, pp. 282-283) 
divided this species into three subspecies we allowed a typographical 
error to creep into one of our new names, which we did not detect till 
long afterward, probably because the printed jl and f look so alike. The 
name we gave the more western subspecies should be Pericrocotus breviros- 
irzs favillaceus (favillaceus, ike glowing ashes or embers) not ‘flavil- 
laceus,’’ as it appeared, which has no meaning. I believe even at this 
late date the rules of nomenclature allow such a correction to be made. 


PYCNONOTIDZ 
githina tiphia tiphia Linnzeus 
One male in green plumage: Chang-lung, Yunnan, March 19, 1917. 


Chloropsis hardwickii (Jardin and Selby) 


Four adults, both sexes: Chang-lung and Mu-cheng, Salwin River, 
Yunnan, February 15 and March 18 and 21, 1917. 


Chloropsis icterocephala chlorocephala (Walden) 


| Two males: Namting River, Burma border, February 21 and 28, 
1917. 


1921] Bangs, Birds of the Asiatic Expedition 585 


Hypsipetes leucocephalus (Gmelin) 

Two adults, male and female: Namting River, Burma border, and 
Yoakuan, Yunnan, January 21 and February 21, 1917. 

The male of this pair has a pure white head, neck, and chest; the 
rest of the under parts are clear ashy gray with no black intermixed. 
Such a condition of plumage is apparently rare. Most birds with pure 
white heads have the under parts black or much mixed with black. One 
or two, however, in a long series from other parts of China match this one. 

The female I at first inclined to refer to some other species. Like 
the male, it has no black on the under parts, which are wholly gray, but 
there is no white at all on its head. I have seen some skins, however, 
from Hupeh and Mengtsz, Yunnan, that have hardly any white on 
the head and that almost match it. It has no black on the cheek and 
therefore cannot be referred to H. concolor Blyth. 

The extraordinary range of variation in color in this species seems to 
be individual rather than due to age. I have before me black-breasted 
birds with wholly white heads and others with only a few white feathers 
in the head, and gray-breasted birds with and without white heads, and 
all sorts and kinds of intermediates. 


Hemixus flavala flavala (Hodgson) 
One male: Chang-lung, Salwin River, Yunnan, March 19, 1917. 


Iole maclellandi similis Rothschild 
One male: Ta-shiu-tang, Salwin drainage, Yunnan, February 3, 1917. 


Alcurus striatus (Blyth) 
One adult female: Tai-ping-pu, Yunnan, April 9, 1917. 


Molpastes cafer burmanicus (Sharpe) 


Three adults, two males, one with sex not determined: Yung-chang 
Fu, Yunnan, January 24, 27, and 28, 1917. 


Pycnonotus xanthorrhous xanthorrhous J. Anderson 


Four adults, both sexes: Wan-tien, Li-chiang Fu, Chang-lung, and 
Hui-yao, Yunnan, November 11, 1916, March 19, and May 1, 16 and 
17,1917. These are, of course, true zanthorrhous of Anderson. A large 
series collected by Zappey in Hupeh represents quite a different form, 
distinguished by being slightly larger, paler brown above, and with the 
brown band across the chest much paler and less sharply contrasted. 
This is Pycnonotus xanthorrhous andersoni (Swinhoe); type locality, 
Ichang. 


586 Bulletin American Museum of Natural History [Vol. XLIV 


Otocompsa emeria emeria (Linneus) 


Six adults, both sexes: Malipa, Burma; Chang-lung and Meng- 
ting Yunnan, February 17, March 12 and 21, 1917. 


Otocompsa flaviventris flaviventris (Tickell) 
One adult male: Chang-lung, Yunnan, March 19, 1917. 


Spizixos canifrons Blyth 


Five aults, both sexes: Tai-ping-pu and Chen-kang, Yunnan, 
February 7 and April 8, 9, and 12, 1917. 


TIMELIIDE 
Ianthocincla lanceolata lanceolata J. Verreaux 
One adult male: Mu-cheng, Salwin drainage, Yunnan, February 13, 
1917. I cannot distinguish this skin in any way from specimens from 
the mountains of Hupeh. 


Ianthocincla lanceolata bonvaloti (Oustalet) 

One adult male: Li-chiang, Snow Mountains, Yunnan, 10,000 feet 
altitude, November 16, 1916. This specimen is indistinguishable from 
examples taken by Zappey in the high mountains, 9000 to 14,000 feet 
altitude, of extreme western Szechwan. 

These two specimens are both in unworn, fresh plumage and I have 
had ample material, collected by Zappey in Hupeh and Szechwan, with 
which to compare them. 

It is interesting to get both forms in Yunnan, and there as in south- 
western China to find the small paler form, lanceolata, at lower and the 
large darker form, bonvalotz, at higher altitudes. 

In the series now before me, in 7 adults of lanceolata the wing ranges 
from 91 to 98 mm.; in 4 adults of bonvaloti from 106-113. In fresh 
plumage bonvaloti is darker than lanceolata with the browns of the head 
and back deeper and richer, and the stripes on the under parts both 
- heavier and darker in color. 


Ianthocincla erythrocephala woodi (Stuart Baker) 
One adult male: Mu-cheng,. Salwin drainage, Yunnan, February 
14, 1917; ‘‘caught in a steel trap set in the forest.”” This skin agrees 
very well with Stuart Baker’s description and undoubtedly represents 
the very distinct form recently described by him. It, No. 143413 Amer. 
Mus. of Nat. Hist., affords the following measurements: Wing, 104; 
tail, 119; tarsus, 41.5; exposed culmen, 20 mm. 


1921] Bangs, Birds of the Asiatic Expedition 587 


Ianthocincla ellioti (Verreaux) 

One adult male: Li-chiang, Snow mountains, Yunnan, 10,000 feet 
altitude, November 9, 1916. After a very careful comparison of this’ 
skin with the long series collected by Zappey in Hupeh and Szechwan, I 
can detect not the slightest difference. 

Rothschild reached the same conclusions from a study of his 
material, but hesitated to throw I. elliott honoripeta Hartert, the sup- 
posedly darker Yunnan form into synonymy, thinking it might be a 
form that reached the province only on migration. I think there is 
little doubt that the name is a pure synonym. 


Ianthocincla cinereiceps styani (Oustalet) 

One adult male: Malipa, Burma border, 3600 feet altitude, March 
16, 1917. This specimen represents true styan? with a black cap and 
yellowish-brown ear-coverts. The one skin taken February 6 and prob- 
ably a migrant, recorded by Phillips and myself from near Mengtsz, 
Yunnan, under the name styanz proves on comparison to be J. c. 
cinereiceps (Styan). The black cap, I have lately been told by La 
Touche, is only a sign of maturity, and is acquired by old birds of both 
forms. The two subspecies can, however, be distinguished easily by the 
colors of the ear-coverts and of the under parts. 


Pomatorhinus macclellandi odicus Bangs and Phillips 
Two adults, male and female: Mu-cheng, Salwin drainage, Yunnan, 
February 10 and 13, 1917. These agree exactly with the Mengtsze speci- 
mens upon which the subspecies was based. 


Garrulax leucolophus leucolophus (Hardwick) 
Two males: Malipa, Burma, March 14, 1917. These seem to be 
quite typical, showing no characters even approaching those of G. 1. 
belangeri Lesson. 


Garrulax pectoralis pectoralis (Gould) 
One adult female: Malipa, Burma, March 14, 1917. This speci- 
men appears to be nearer the bird of the Himalayas than it is to the 
southern G. p. meridionalis Stuart-Baker. 


Dryonastes chinensis chinensis (Scopoli) 


Two adults, male and female: Chang-lung, Salwin River, Yunnan, 
2000 feet altitude, March 18 and 21, 1917. 


588 Bulletin American Museum of Natural History [Vol. XLIV 


Dryonastes sannio albosuperciliaris (Godwin-Austen) 
Two adult females: Wan-tien and Mu-cheng, Yunnan, February 
13 and May 14, 1917. With ample material I now have no hesitation in 
recognizing two forms of D. sannio: D. sannio sannio (Swinhoe), central 
and southeastern China, more reddish olive above with more rusty tail, 
and D. s. albosuperciliaris (Godwin-Austen), Manipur to Yunnan, olive 
above with much less rusty tail. 


Pellorneum ruficeps minus Hume 
Two adults, male and female: Chang-lung, Salwin River, Yunnan, 
and Malipa, Burina border, March 15 and 23, 1917. These two skins 
are, I think, best called minus, though they appear to be somewhat inter- 
mediate between that form and P. r. mandellii Blanford. 


Drymocataphus tickelli tickelli (Blyth) 

One male: Namting River, Burma border, 1700 feet altitude, 
February 25, 1917; ‘‘caught in a rat trap set in the forest.’”’ I have seen 
no other skins of this species and have identified the specimen as best I 
can by Harington’s ‘Notes on the Indian Timelliides and their Allies’ 
(1915, Journ. Bomb. Nat. Hist. Soc., XXIII, p. 485). He thinks true 
tickelli and assamensis Sharpe will prove to be one and the same. Cer- 
tainly this skin is not “‘rufescent’”’ olive-brown above. 


Alcippe pheocephala magnirostris Walden 
One adult male: Namting River, Burma border, 1700 feet altitude, 
February 21, 1917. This specimen, with no whitish eye-ring and very 
well-marked black stripes on the side of the head, I refer without much 
doubt to this form, of which, however, I have seen no other examples. 


Shoeniparus genestieri (Oustalet) 


Three specimens, two males and a female: Ho-mu-shu Pass, 8000 
feet altitude, and Mu-cheng, Salwin drainage, 7000 feet altitude, 
Yunnan, February 10 and April 4, 1917. 


Mixornis rubricapilla sulphurea (Rippon) 


Two adults, male and female: Meng-ting, Burma border, and 
Chang-lung Salwin drainage, Yunnan, February 18 and March 18, 1917. 


Myiophoneus eugenei eugenei Hume 
Four adults, both sexes: Namting River, Burma border, and 
Yung-chang, Yunnan, January 28, February 18 and 28, and March 2, 
1917. 


1921] Bangs, Birds of the Asiatic Expedition 589 


Rothschild is wholly right in his suspicion that M. eugenez and M. 
tibetanus Madarfsz are in reality one and the same. The skins of the 
‘so-called tibetanus collected by Zappey in high western Szechwan are in 
every way like Yunnan specimens of eugener. ‘This I detected long ago, 
when we first received a series of eugenet. 


Lioptila annectens annectens Blyth 
One adult male: Mu-cheng, Salwin drainage, Yunnan, 5000 feet 
altitude, February 15, 1917. The flanks and undertail coverts in this 
specimen are rather paler, otherwise it agrees with birds in the Mu- 
seum of Comparative Zoédlogy from Manipur with the back Sanford’s 
brown. L. annectens satuwratus, which I have not seen, is said to have 
the back “rich, deep chestnut.” 


Lioptila desgodinsi (David and Oustalet) 


Two adults, male and female: Tai-ping-pu and Yao-kuan, 
Yunnan, 6000 to 7000 feet altitude, January 31 and April 12, 1917. 


Staphidia striata (Blyth) 


One adult male: Chang-lung, Salwin River, Yunnan, 2000 feet 
altitude, March 20, 1917. 


Siva cyanuroptera wingatei Ogilvie-Grant 
Two-adults, male and female: Hui-yao and My-cheng, Yunnan, 
February 10 and May 1, 1917. 


‘ 


Yuhina diademata ampelina Rippon 
Two adults, male and female: Li-chiang, Snow Mountains, 10,000 
feet altitude, November 11 and 12, 1917. These specimens are no darker 
than examples from Hupeh and Szechwan—in fact, one of them is lighter 
in color than true diademata. I might add that, in comparing examples 
of this species, it is well to compare only those taken at approximately 
the same season of the year, spring and summer specimens being much 
lighter in color than autumn and winter ones. 
I follow Rothschild in the name used for the bird of western Yunnan, 
as I am afraid I do not know the real characters of ampelina. 


Yuhina occipitalis obscurior Rothschild 
Two adults, male and female: Lung-ling, Yunnan, March 27, 1917. 


Ixulus flavicollis rouxi Oustalet 
One adult male: Tai-ping-pu, Yunnan, April 2, 1917. 


590 Bulletin American Museum of Natural History [Vol. XLIV 


Cutia nepalensis nepalensis Hodgson 


One adult male: Ho-mu-shu Pass, Yunnan, 8000 feet altitude, 
April 5, 1917. 


Pterythius eralatus ricketti Ogilvie Grant 
One adult female: Ho-mu-shu Pass, Yunnan, 8000 feet altitude, 
April 7, 1917. There is a note on the label of this specimen reading 
“contained eggs.” 


Mesia argentauris argentauris Hodgson 


One adult male: 20 miles south of Chen-kang, Salwin drainage, 
Yunnan, February 7, 1917. 


TROGLODYTID& 
Spelzornis souliei Oustalet 
One male: Tai-ping-pu, Yunnan, 7000 feet altitude, April 12, 1917; 
“caught in a rat trap set in the forest.”’. I have compared this skin with 
Oustalet’s description and believe it to belong to the species he described 
as souliet. It is, however, the only specimen of the species that I have 
seen. 


Pnoepyga pusilla pusilla Hodgson 


Two specimens, male and female: Ho-mu-shu Pass, Yunnan, 7000 
feet altitude, and Namting River, Burma border, 1700 feet altitude, 
February 25 and April 8, 1917. Both examples were caught in traps set 
for small mammals in the forest. 


TURDIDA 
Turdus merula mandarinus Bonaparte 


Two adults, male and female: Yung-chang, Yunnan, January 24 
and 29, 1917. 


Turdus castaneus gouldi ( Verreaux) 

Four specimens, three adults, both sexes, and an immature female: 
Li-chiang, Snow Mountains, 10,000 feet altitude, Yoa-kuan, and Tai- 
ping-pu, Yunnan, November 10 and 14, 1916, January 31 and April 
10) 1997. 


Turdus dissimilis Blyth 


Two adults, male and female: Chang-lung, Salwin River, Yunnan, 
March 18, 1917. 


1921] Bangs, Birds of the Asiatic Expedition 591 


Turdus auritus conquisitus, new subspecies 


Type and only specimen.—No. 143452, Amer. Mus. of Nat. Hist.; adult female; 
Li-chiang, Snow Mountains, 10,000 feet, Yunnan; November 16, 1916; R. C. Andrews 
and E. Heller. 

CHARACTERS.—Similar to Turdus auritus auritus Verreaux, but under parts 
much more heavily spotted with black, especially on sides and flanks, all the spots 
larger and more intensely black less brownish black. Wing, 124 mm.; tail, 91 mm.; 
tarsus, 36 mm.; culmen, to base of forehead, 23. 

When I first compared the Yunnan skin with one adult of 7’. a. awritus, collected 
by Zappey in western Szechwan, both killed in November, I was at once struck by 
the great difference in the spotting of the under parts and made a note to that effect 
but did not name the Yunnan bird having only one skin from each region. Since then 
Rothschild has called attention to exactly the same difference shown by his one adult 
from Yunnan, compared with his one adult from the Tsin Ling Mountains, and now 
no reason remains for not giving the Yunnan form a name. 


Turdus mollissimus Blyth 


One adult female: Li-chiang, Snow Mountains, Yunnan, 10,000 
feet altitude, November 9, 1916; ‘‘caught in a steel trap.” 


Monticola solitarius pandoo (Sykes) 
One adult male: Tung-chang Fu, Yunnan, January 26, 1917. 


Monticola erythrogaster Vigors 


One adult male: Ho-mu-shu Pass, Yunnan, 8000 feet altitude, 
April 9, 1917. 


Enicurus schistaceus Hodgson 
One female: Namting River, Burma border, February 25, 1917. 


Chimarrhornis leucocephala (Vigors) 


Two adults, male and female: Mu-cheng, Salwin drainage, and 
Yuan-chiang-Chou, Yunnan, January 26 and February 16, 1917. 


Phenicurus hodgsoni (Moore) 


Two specimens, male and female: Yung-chiang-chou, Yunnan, 
January 27, 1917. 


Calliope calliope calliope (Pallas) 
Two adults, male and female: Namting River, Burma border and 
Chang-lung, Yunnan, March 2 and 21, 1917. 


592 Bulletin American Museum of Natural History [Vol. XLIV 


Ianthia rufilata practica Bangs and Phillips 
Two specimens, male and female: Mu-cheng, Yunnan, February 
10 and 14, 1917. The male, a fine adult in full plumage, has the blue 
parts of its plumage a little deeper and slightly more purplish than in 
the type of practica. There is also some slight white at the bases of the 
superciliaries. 


Notodela leucura (Hodgson) 
Two adult males: Namting River, Burma border, 1700 feet alti- 
tude, February 20 and 21, 1917. These two specimens show a wing 
measurement of 95 mm. in one and 97 mm. in the other. 


Copsychus saularis saularis (Linnzus) 
One female: Meng-ting, Burma border, February 18, 1917. 


Saxicola torquata przewalskii Pleske 


Three specimens, two males and a female: Yung-chang Fu, Yun- 
nan, January 27 and 28, 1917. 


Oreicola jerdoni Blyth 
One adult male: Namting River, Burma border, February 22, 1917. 


Oreicola ferrea haringtoni Hartert 


Three specimens, a male and two females: Malipa, Burma border, 
and Wan-tien, Yunnan, March 14 and May 14, 1917. 


SYLVIIDZ 


Megalurus palustris andrewsi,! new subspecies 
Two adult males: Malipa, Burma, and Meng-ting, Burma border, 
February 18 and March 14, 1917. 


Typre.—No. 143478, Amer. Mus. of Nat. Hist.; adult male; Meng-ting, Burma 
border; February 18, 1917; R. C. Andrews and E. Heller. 

CHARACTERS.—Similar to M. palustris palustris Horsford from Java and of 
about the same size, differing in the black striping of the upper parts being wider and 
. more intensely black; the brown of upper parts deeper, brighter, more reddish brown, 
especially on the crown. The general color of upper parts in M. palustris palusiris 
is clay-color to buck-thorn brown; general color of upper parts in the new form is 
ochraceous tawny, almost tawny on the crown. 


1Named in honor of Roy Chapman Andrews. 


1921] Bangs, Birds of the Asiatic Expedition 593 


M®eASUREMENTS 
Exposed 
Sex Locality Wing Tail Tarsus Culmen 
A. M.N.H. fof Burma border 107 mm. 135 mm. 41mm. 17.5mm. 
No. 143478 Meng-ting 
A.M.N.H. rot Burma 106 133 42 18.0 
No. 143477 Malipa 
M.C.Z.No. 92 India 94 124 SY 16.0 
34207 Buxa Doars 


I cannot find that the Indian and Burmese form has ever been sepa- , 
rated from the typical Javanese bird. Gray (1848, ‘Gen. Birds,’ I, 
p. 169, Pl. xtvmt1) figured a young bird in the very yellowish plumage 
and named it Megalurus citrinus. He did not state where his specimen 
was from, and Sharpe in Vol. VII of the catalogue of birds does not claim 
the type in the British Museum. I believe the young as figured by Gray 
could not be positively identified as that of either one or the other sub- 
species. 

Both the Javanese and Burmese forms are much browner, less gray- 
ish, than the Philippine bird, 7. palustris forbest Bangs. 


Phylloscopus fuscatus (Blyth) 
One female: Yuan-chiang-Chou, Yunnan, January 27, 1917. 


Phylloscopus davisoni (Oates) 
One male: Wan-tien, Yunnan, May 15, 1917. 


Phylloscopus proregulus forresti Rothschild 


One male: Yung-chang Fu, Yunnan, January 27, 1917. 

This specimen, in winter plumage, fits the description of the lately 
described forresti well, except that it has the extreme base of the lower 
mandible of a pale color. 


Phylloscopus humei premium Mathews and Iredale 
One male: Chang-lung, Salwin River, Yunnan, March 21, 1917. 
This is, of course, the bird we used to know as P. superciliosus super- 
ciliosus (Gmelin). 
Phylloscopus lugubris (Blyth) 
One male: Wan-tien, Yunnan, May 15, 1917. 


Horornis canturians (Swinhoe) 
Two males: Yung-chang Fu, Yunnan, January 28, 1917. 


594 Bulletin American Museum of Natural History [Vol. XLIV 


Prinia inornata exter Thayer and Bangs 
Two females: Yung-chang Fu, Yunnan, January 27, 1917. 


PRIONOPIDZ 
Hemipus picatus capitalis (McClellan) 
One male: Chang-lung, Salwin River, Yunnan, March 20, 1917. 


LANIIDZ 
Lanius schach tephronotus (Vigors) 


Two specimens, adult male and immature male: Yung-chang Fu, 
Yunnan, January 27 and 28, 1917. 


Lanius nigriceps nigriceps Franklin 
One adult female: Meng-ting, Burma border, February 18, 1917. 


Lanius colluroides Lesson 


Two adult females: Chang-lung and Yung-chang Fu, Yunnan, 
January 28 and March 21, 1917. 


PARIDZ 
Parus major commixtus Swinhoe 


Two males: Yung-chang Fu, Yunnan, 5500 feet altitude, January 
27 and 28, 1917. These skins, like many taken by Zappey in parts of 
western Szechwan, are not extreme commixtus but are nearer to it than 
they are to any of the other races. The wing measures 70 mm. in one of 
these and 74 mm. in the other. 


SITTIDE 
Sitta europea nagaensis Godwin Austen 
One adult female: Ho-mu-shu Pass, Yunnan, 8000 feet altitude, 
April 4, 1917. This specimen is much grayer, less rusty below, than in 
any of our skins of S. ewropea montium LaTouche and, if the two are 
distinct, should, I believe, be referred to magensis. 


Sitta frontalis corallina Hodgson 
Three adult females: Malipa, Burma and Namting River, Burma 
border and Chang-lung, Salwin River, Yunnan, February 23 and March 
13 and 20, 1917. 
Indian and Burmese birds are slightly different from true S. frontalis 
of Java. They are somewhat paler below with a vinaceous rather than 
a lilaceous tinge and have a more extended white throat-patch. 


1921] Bangs, Birds of the Asiatic Expedition 595 


I follow Hellmayr and unite all the true nuthatches in one genus, 
being loath to accept the excessive subdivision of the genus proposed by 
Buturlin in his recent (1911) review. 


CERTHIIDE 
Certhia disco‘or discolor Blyth | 
One adult male: Tai-ping-pu, Yunnan, 7000 feet altitude, April 
9, 1917. 
ZOSTEROPIDZ 
Zosterops palpebrosa simplex Swinhoe 


Two adults, male and female: Chang-lung, Salwin River, Yunnan, 
and Malipa, Burma border, March 15 and 21, 1917. 


DICHIDA 
Diceum minullum olivaceum Walden 
One female: Chang-lung, Salwin River, Yunnan, March 20, 1917. 


NECTARINIIDZ 
AEthopyga ignicauda (Hodgson) 


One immature male (in change from young to adult plumage): 
Yoakuan, Yunnan, 6000 feet altitude, January 31, 1917. 


#thopyga nipalensis (Hodgson) 


Two adults, male and female: Mu-cheng and Chang-lung, Yunnan, 
February 16 and March 18, 1917. 


#Ethopyga dabryi (J. Verreaux) 


Four adult males: Wan-tien, Ta-shui-tang and Mu-cheng, Yunnan, 
February 2, 3 and 16, and May 14, 1917. 


MOTACILLIDA 
Motacilla alba hodgsoni Blyth 
One adult female: Yung-chang Fu, Yunnan, January 27, 1917. 


Motacilla alba ocularis Swinhoe 
One adult male: Yung-chang Fu, Yunnan, January 27, 1917. 


Motacilla cinerea melanope Pallas 
One immature male: Yung-chang Fu, Yunnan, January 27, 1917. 


596 Bulletin American Museum of Natural History [Vol. XLIV 


Anthus hodgsoni Richmond 


_ Twospecimens, male and female: Yung-chang Fu, Yunnan, January 
27 and 28, 1917. 

Lately Uchida and Kuroda (1916, Annotationes Zoologicze Japonen- 
ses, [X, p. 184) have named a form from Yunnan Anthus maculatus 
yunnanensis, apparently based upon migrant birds. The only char- 
acter ascribed to the new form is a smaller bill than in the typical bird. 
The two specimens listed above have rather small bills, but five winter 
birds from Mengtsz are quite like examples from anywhere else in this 
respect. In a good series of breeding birds from the high mountains of 
Hupeh and, Szechwan I find a good deal of individual variation in the 
size of the bill, as also in breeding birds from Sachalin Island, and I do 
not believe yunnanensis is a recognizable form. 

Sarundy, in 1909, named the breeding bird of south-western Kansu 
Anthus maculatus berzowskiz, on the character of a grayer back with 
blacker shaft markings. I have seen no specimens from Kansu, but all 
mid-summer examples that I have examined show these two charac- 


teristics to a marked degree when compared with winter or spring killed: 


individuals from the same regions. 

T am inclined to believe that Rothschild, in using berezowskii as the 
name of the species, overlooked the fact that A. hodgsoni Richmond, 
as a substitute for the preoccupied Anthus maculatus Jerdon, dates from 
1907, and A. berezowskiit Sarundy dates from 1909; but perhaps he did 
not and meant to treat berezowskii as a species distinct from hodgsont. 


FRINGILLIDA 
Eophona migratoria migratoria! Hartert 
One female: Yung-chang Fu, Yunnan, January 28, 1917. 


Spinus ambiguus Oustalet 


Two males: Yung-chang Fu, Yunnan, 5500 feet altitude, January 
28, 1917. 


Passer montanus montanus (Linnzeus) 


One adult male: Yung-chang Fu, Yunnan, January 28, 1917. I 
cannot distinguish this specimen from European birds. It apparently 
does not approach P. montanus malaccensis Dubois of tropical India, 
Malaya, ete. 


1For the change of the specific name from melanura to migratoria, see Penard, 1919, Proc. New 
Eng. Zool. Club, VII, p. 22, October 31. 


1921] Bangs, Birds of the Asiatic Expedition 597 


Passer rutilans cinnamomea (Gould) 

One adult male: Lung-ling, Yunnan, 5000 feet altitude, March 27, 
1917. 

Rothschild records the specimens taken by. Forrest as Passer ruti- 
lans assimilis Walden. I cannot reconcile any Yunnan skin examined by 
me with Walden’s description which calls for a bird with “the cheeks 
and sides of the neck pure white, and the breast, flanks and ventral 
region ashy grey.”’ All specimens from Yunnan as well as those from 
western Szechwan that I have seen have yellow cheeks and sides of the 
neck, and are strongly washed with yellow all over the under parts, and 
appear to me indistinguishable from birds from the eastern Himalayas. 

Zappey, however, took in Hupeh and eastern Szechwan seven spar- 
rows that Thayer and I referred to P. rutilans rutilans (Temminck). 
These skins agree well with Walden’s description, but I cannot see 
much difference between them and Japanese birds, except that some, not 
all, of them are intermediate between rudilans and cinnamomea. 


Carpodacus edwardsii Verreaux 


One male in plumage similar to that of the female: Tai-ping-pu, 
Yunnan, April 9, 1917. 


Pyrrhula erythaca altera Rippon 


One adult male: Li-chiang, Snow Mountains, Yunnan, 10,000 
feet altitude, November 11, 1916. 


Emberiza pusilla Pallas 


Three specimens, two males and a female: Malipa, Burma and 
Yung-chang Fu, Yunnan, January 28 and March 13, 1917. : 


Emberiza spodocephala melanops Blyth 
Two males: Chang-lung, Salwin River, Yunnan, March 21 and 24, 
Bol. 
Melophus melanicterus (Gmelin) 
One immature male: Namting River, Burma border, February 23, 
1917. 
PLOCEIDE 
Munia punctata topela Swinhoe 


One immature female: Namting River, Burma border, February 
28, 1917. 


598 Bulletin American Museum of Natural History [Vol. XLIV 


STURNIDZA 
Sturnia nemoricola Jerdon 


Two females: Namting River, Burma border, and Chang-lung, 
Salwin River, Yunnan, February 25 and March 20, 1917. 


Gracupica nigricollis (Paykull) 


Three adults, two males and a female: Meng-ting, Burma border, 
February 28, 1917. 


Acridotheres tristis (Linnzeus) 
Two adults, male and female: Shih-tien, Yunnan, January 30, 1917. 


Athiopsar cristatellus cristatellus (Gmelin) 


Seven adults, both sexes: Malipa, Burma, Yoa-kuan and Hsiao,! 
Salwin drainage, Yunnan, January 30, February 3, March 12 and 13, 
1917; “often seen feeding on the backs of buffalo.” 


ZEthiopsar albocinctus Godwin-Austen and Walden 
One adult female: Malipa, Burma, March 13, 1917. 


ORIOLIDE 
Oriolus indicus tenuirostris Blyth 


7 


One specimen (marked ‘ <,’’ apparently a female): Yung-chang, 


Yunnan, January 26, 1917. 


DICRURIDZE 
Chibia hottentotta hottentotta (Linnzus) 


One adult male: Chang-lung Salwin River, Yunnan, March 21, 
1917. This is a large billed bird; the bill measured as Stuart Baker 
(1919, Nov. Zool., XX VI, p. 44) measures his series gives 29 mm. 


Dicrurus leucophzus nigrescens Oates 


Three adult males: Yung-chang and Chang-lung, Salwin River, 
Yunnan, January 27 and 28, March 22, 1917. These are all large birds 
(wing: 145, 146, and 140 mm.) 


CORVIDZ 


Corvus coronoides levaillantii Lesson 


One adult female: Li-chiang, Snow Mountains, Yunnan, 10,000 
feet altitude, November 12, 1916. 


1“ Hsiao’? means in English ‘‘small, little.’’ 


1921] Bangs, Birds of the Asiatic Expedition 599 


Corvus insolens Hume 
One adult male: Meng-ting, Burma border, February 18, 1917. 
Nucifraga caryocatactes macella Thayer and Bangs 

One adult male: Li-chiang, Snow Mountains, Yunnan, 10,000 feet 
altitude, November 6, 1916. On comparing this skin with the type of 
macella from the mountains of Hupeh and with one skin from Tachienlu, 
I can detect no differences that would seem to be subspecific. The type 
is a little paler brown than in either the Tachienlu or Yunnan specimens, 
but I cannot believe that this slight difference would prove to be con- 
stant. Also, the white spotting in the type of macella extends quite down 
the middle of the belly to the vent, whereas in the two other skins the 
whole belly is unspotted. If long series should show that Nucifraga 
yunnanensis Ingram (1910) is different from N. macella (19C9) of Hupeh, 
then the Tachienlu bird must be referred to ywnnanensis. For the pres- 
ent, I unite the two under the older name. 


Pica pica sericea Gould 
Two adult females: Yung-chang Fu, Yunnan, January 27, 1917. 


Urocissa erythrorhyncha erythrorhyncha (Gmelin) 


Four adults, both sexes: Hui-yao, 5500 feet altitude, and Li-chiang, 
Snow Mountains, 10,000 feet altitude, Yunnan, November 7, 9, and 12, 
1916, and May 7, 1917. 


Dendrocitta himalayensis Blyth 


Two adults, male and female: Wantien and Taipingpu, Yunnan, 
April 12 and May 14, 1917. 


Garrulus leucotis Hume 
One adult male: Malipa, Burma, March 14, 1917. 


Garrulus bispecularis sinensis Swinhoe 


One adult male: Lichiang, Snow Mountains, Yunnan, 10,000 feet 
altitude, November 11, 1916. This specimen has a grayish back and a 
whitish throat, wing 195 mm., and thus represents the variant named 
rufescens by Reichenow. 

Rothschild has relegated the supposed subspecies rufescens to 
synonymy. J had written a long account of our large series of Chinese 
jays showing that ‘‘rufescens”’ has no region of its own, but occasionally 
turns up anywhere within the range of the variable sinensis. Briefly 
stated, our material wholly supports what Rothschild has said. 


600 Bulletin American Museum of Natural History [Vol. XLIV 


BirpDs FROM FOKIEN 
PHASIANIDZE 
Francolinus pintadeanus pintadeanus (Scopoli)! 


Four specimens, two males, two sex undetermined: Futsing, Fokien, 
June 1911 and 1912, and July 10, 1916. I think Scopoli’s name must be 
used for the large bird of southern China, probably introduced from 
thence into Mauritius, with a wing ranging from 153 mm. to 157 mm., 
and the smaller form of Burma, Cochin China, Siam and Yunnan be 
known as F. pintadeanus phayrei (Blyth). 


Arboricola ricketti Ogilvie-Grant 


Two specimens: a male, Yenping, Fokien, June 13, 1916; and one, 
sex undetermined, Futsing, Fokien, 1912. 


Bambusicola thoracica (Temminck) 


Three adults, a male, and two females: Futsing, Fokien, April 1911 
and March 17, 1912. 


COLUMBIDZ 
Streptopelia orientalis orientalis (Latham) 
Three males: Futsing, Fokien, July 25 and August 1, 1916. 


Streptopelia chinensis chinensis (Scopoli) 
One adult female: Futsing, Fokien, July 27, 1916. 


RALLIDZ 
Porzana pusilla pusilla (Pallas) 


Two specimens, male and female: Futsing, Fokien, April and 
October 1912. 


Amaurornis akool coccineipes Slater 


Four specimens, male, two females and one with sex not determined : 
Futsing, Fokien, May 1911, September 1912, and 1912. 


Gallicrex cinerea (Latham) 


Three specimens, two adult males, one immature, sex not deter- 
mined: Futsing, Fokien, June 1911. 


1F¥or change of name from Francolinus chinensis to F. pintadeanus cf. Oberholser, 1919, Proc. Biol. 
Soc. Wash., XXXII, April, p. 21. 


> 


1921] Bangs, Birds of the Asiatic Expedition 601 


LARIDZ 
Sterna albifrons sinensis Gmelin 
Two adults, male and female: Futsing, Fokien, May 1911. 


Larus argentatus vege Palmén 
One adult female: Futsing, Fokien, December 1912. 


Charadriide 
Arenaria interpres interpres (Linnzus) 
Two specimens in winter plumage, sex undetermined: Futsing, 
Fokien, 1912. 
Vanellus vanellus (Linnzus) 
One adult, sex not determined: Futsing, Fokien, 1912. 


Pluvialis dominicus fulvus (Gmelin) 
One male, in winter plumage: Futsing, Fokien, October 1912. 


Charadrius leschenaultii Lesson 


One male: Futsing, Fokien, June 1911. 


Charadrius alexandrinus dealbatus (Swinhoe) 


One specimen, sex not determined: Futsing, Fokien, 1912. 


Numenius arquatus lineatus Cuvier 
One adult female: Futsing, Fokien, December. 


Numenius pheopus variegatus (Scopoli) 
Two specimens, one male, one sex not determined: Futsing; 
Fokien, 1912. 
Tringa ochropus Linnzeus 
One adult female: Futsing, Fokien, July 21, 1916. 


Heteractitis brevipes (Vieillot) 
One female, in winter plumage: Futsing, Fokien, September 15, 
1911. 
Erolia ruficollis (Pallas) 


Two females: Futsing, Fokien, May and June 1911. 


Erolia acuminata (Horsford) 


Three adults, male and two females, all in spring plumage: Futsing, 
Fokien, May 1911. 


G02 Bulletin American Museum of Natural History [Vol. XLIV 


Erolia alpina sakhalina (Viecillot) 


Two specimens, sex not determined, in winter plumage:. Futsing, 
Fokien, 1912. 


Gallinago gallinago gallinago (Linnzus) 
One female: Futsing, Fokien, May 1911. 


Rostratula bengalensis bengalensis (Linn:eus) 
Two adults, male and female: Futsing, Fokien, May and June 1911. 


GLAREOLID 
Glareola maldivarum Forster 


Three adults, two males and a female: Futsing, Fokien, May and 
June 1911. 


ARDEIDE 
Nycticorax nycticorax nycticorax (Linn:eus) 


One specimen, immature, sex not determined: Futsing, Fokien, 
1912. 


Ardeola bacchus (Bonaparte) 


Two specimens, an adult male and a somewhat immature female: 
Futsing, Fokien, July 25, 1916. 


Ixobrychus sinensis sinensis (Gmelin) 
Two females: Futsing, Fokien, May 1911 and 1912. 


Ixobrychus cinnamomea (Gmelin) 


Two adults, male and female: Futsing, Fokien, July 26 and 31, 
1916. 


ANATIDE 
Melanonyx segetum serrirostris (Swinhoe) 
One adult, sex not determined. Futsing, Fokien, 1912. 


Mergus serrator Linnzus 


Four specimens, adult male and three females: Futsing, Fokien, 
November and December 19, 1912. 


1921] Bangs, Birds of the Asiatic Expedition 603 


FALCONIDZ 
Astur soloensis (Horsfield) 


Two specimens, adult male, and female immature: Futsing, Fokien, 
August 1913 and June 1914. 


Accipiter gularis (Temminck and Schlegel) 
One adult male: Futsing, Fokien, May 1911. 


Buteo buteo japonicus (Temminck and Schlegel) 
One male: Futsing, Fokien, October 1912. 


Butastur indicus (Gmelin) 


One adult, sex not determined: Futsing, Fokien, 1912. 


Falco columbarius insignis (Clark) 
One female: Futsing, Fokien, November 1912. 


Cerchneis tinnunculus japonicus (Temminck and Schlegel) 
Two females: Futsing, Fokien, January 1912 and 1912. 


BUBONIDA: 
Otus bakkameena glabripes (Swinhoe) 
One immature female: Futsing, Fokien, July 24, 1916. 


Ninox scutulata scutulata (Raffles) 


Two adults, sex not determined: Futsing, Fokien, March 1912. 


Glaucidium brodiei (Barton) 


One adult, sex not determined (=male): Futsing, Fokien, 1912. 


Glaucidium cuculoides whitelyi (Swinhce) 


Four specimens, one male, two females and one sex not determined : 
Futsing, Fokien, July 24 and 28 and August 1, 1916 and 1912. 


CORACIIDE 
Eurystomus orientalis calonyx Sharpe 


Three specimens, an adult female, and two young, male and female: 
Futsing, Fokien, April 1911 and August 1912. See Stresemann, 1913, 
Noy. Zool., XX, p. 299 for discussion of the geographical races and the 
points where intergradation takes place. 


604 Bulletin American Museum of Natural History [Vol. XLIV 


ALCEDINIDZE 
Ceryle lugubris guttulata Stejneger 
Two males: Futsing, Fokien, July 27 and August 1, 1916. 


Alcedo atthis bengalensis Gmelin 


Three specimens, two males and a female, all immature: Futsing, 
Fokien, July 12, 27 and 28, 1916. 


Halcyon smyrnensis fusca (Boddaert) . 
Five specimens, both sexes: Futsing, Fokien, July 26 and 28, 
1916 and 1912. 
Halcyon pileata (Boddaert) 
One female: Futsing, Fokien, July 31, 1916. 


CAPRIMULGIDZ 
Caprimulgus indicus jotaka Temminck and Schlegel 
Two females: Futsing, Fokien, 1912. 


MICROPODIDZE 
Micropus pacificus pacificus (Latham) 


Four specimens, three males, one with sex not determined: Futsing, 
Fokien, May and June 1911 and 1912. 


TROGONIDE 
Pyrotrogon erythrocephalus yamakanensis Rickett 
One adult male: Yen-ping, Fokien, June 12, 1916. 


CUCULIDE 
Clamator coromandus (Linnzus) 


Three adults, two males and a female: Yen-ping and Futsing, 
Fokien, June 12, 1916 and 1912. 


Cuculus optatus Gould 
Three females: Futsing, Fokien, May 1911 and March and April 
1912. 
Eudynamis orientalis chinensis Cabanis and Heine 
Two adults, male and female: Futsing, Fokien, 1912. 


1921] Bangs, Birds of the Asiatic Expedition 605 


Centropus bengalensis lignator Swinhoe 


Three adults, male and female and one sex not determined: Futsing, 
Fokien, July 20, 23 and 24, 1916. 


CAPITONIDZ 
Megalaema virens virens (Boddaert) 
Four adults, both sexes: Futsing, Fokien, July 27, 1916 and March 
1912. 
PICIDA& 
Picus canus ricketti Stuart-Baker 


Two adults, male and female: Futsing, Fokien, July 1916 and 
August 1912. 


Dryobates cabanisi cabanisi (Malherbe) 


Two males, one adult, one immature: Futsing, Fokien, October 
1912 and July 26, 1916. 


Micropternus brachyurus fokiensis (Swinhoe) 
One adult male: Yenping, Fokien, June 12, 1916. 


Jynx torquilla japonica Bonaparte 


Three specimens, male and two with sex not determined: Futsing, 
Fokien, March 1912 and 1912. 


HIRUNDINIDZ 
Hirundo rustica gutturalis Scopoli 
One immature male: Futsing, Fokien, July 3, 1916. 


MUSCICAPIDE 
Hemichelidon sibirica sibirica (Gmelin) 
One adult, sex not determined and without date of capture: Fut- 
sing, Fokien. 
Poliomyias mugimaki (Temminck) 
Four specimens, both sexes: Futsing, Fokien, 1912. 


' Cyanoptila cyanomelana Temminck 
Seven specimens, both sexes: Futsing, Fokien, March 1912, Sep- 
tember 1912, and 1912. The three adult males in this series are all 
black-throated birds with dark blue backs lined with black, and all be- 


606 Bulletin American Museum of Natural History [Vol. XLIV 


long to this form which seems to be specifically distinct from C. cumatilis 
Thayer and Bangs, the breeding bird of Central China. 
Hypothymis azurea styani (Hartlaub) 
One (female): Futsing, Fokien, 1912. 


Tchitrea paradisi incii Gould 
Three specimens, two males and a female, all in the brown phase of 
plumage: Futsing and Ling Sioh, Fokien, March 1912 and August 2, 
1916. 
Tchitrea princeps princeps (Temminck) 
One male: Futsing, Fokien, March 1912. 


CAMPEPHAGIDE 
Volvocivora melanoptora (Riippell) 
Two males: Futsing, Fokien, April 1912. 


Pericrocotus griseigularis Gould 
Two adult, male and female: Futsing, Fokien, June 1912 and 1912. 


Pericrocotus cantonenis Swinhce 
Four specimens, two adult males, an adult female and immature 
sex not determined: Futsing, Fokien, March 1912 and July 10, 1916. I 
do not use the genus Motacilloides Buturlin for the black, white and gray 
Minivets, as it does not seem to me worth while to subdivide the group. 


PYCNONOTIDA 
Hypsipetes leucocephalus (Gmelin) 
One immature (without white in the head) male: Futsing, Fokien, 
June 12, 1916€. 
Hemixos canipennis Seebohm 
Four adults, both sexes: Futsing and Ling Sioh, Fokien, July 27, 28, 
and 29, 1916, and April. 1912. 
Iole maclellandi holti (Swinhoe) 
Three specimens, one male, two females: Futsing and Ling Sioh, 
Fokien, July 28, 1916, and April 1912. 
Pycnonotus sinensis (Gmelin) 


Twelve specimens, adult and immature of both sexes: Futsing, 
Fokien, July 3, 4, 22,.23, 24, 27, 28, and 31, 1916, and March 1912. 


1921] Bangs, Birds of the Asiatic Expedition 607 


Spizixos semitorques Swinhoe 
One adult, sex not determined: Futsing, Fokien, 1912. 


TIMELIIDZ 
Ianthocincla canora (Linnzus) 
Eight specimens, adult and immature, both sexes: Futsing, Fokien, 
July 10, 24, 27, and 28, 1916, and March 1912. 
Pomatorhinus ruficollis stridulus Swinhoe 


Eleven specimens, both sexes: Futsing, Fokien, July 1, 24, 26, and 
31, 1916, and April 1912. These all have short bills, but the color of the 
back is variable in this series, some specimens being much less reddish 
than others. 


Pomatorhinus swinhoei David 
One adult male: Futsing, Fokien, 1912. 


Garrulax picticollis Swinhoe 


One male: Yenping, Fokien, June 15, 1916. 


Dryonastes perspicillatus perspicillatus (Gmelin) 


Four specimens, one adult, three immature: Futsing, Fokien, July 
13 and 26, 1916, and March 1912. 


Dryonastes sannio sannio Swinhoe) 


Four specimens, two adult, two immature: Futsing, Fokien, July 
26 and 27, 1916, and April 1912. 


Alcippe nipalensis hueti (David) 


Seven specimens, adults and immature of both sexes: Futsing, 
Fokien, July 26 and 31, 1916. 

Hartert (Vog. Pal. fauna, p. 616) says he cannot substantiate the 
differences claimed by Styan to separate his davidi of Szechwan from 
true hueti of Fokien. In Harington’s review of the genus published in 
1915, however, both forms are kept. I have compared the present series 
with a large one in Mus. Comp. Zool. from Szechwan and find that the 
Fokien birds are decidedly paler below, the chest more pinkish, less 
grayish, the sides more buffy, less olivaceous and, therefore, consider 
A. nipalensis davidi Styan of Szechwan a good form. 


Stachyrhidopsis ruficeps davidi Oustalet 


Six specimens, adults and immature, both sexes: Futsing, Fokien, 
July 10, 26, and 29, 1916, and January 1912. 


608 Bulletin American Museum of Natural History [Vol. XLIV 


Myiophoneus ceruleus (Scopoli) 
Four adult males: Futsing, Fokien, July 28, 1916, August and Sep- 
tember 1912. 
Staphidia torquola Swinhoe 
One adult male: Yen-ping, Fokien, June 21, 1916. 


TURDIDE 
Turdus merula mandarinus Bonaparte 
Three females: Futsing, Fokien, March 1912. 


Turdus cardis lateus Thayer and Bangs 
Three adults, two males and a female: Futsing, Fokien, March 
1912 and 1912. 
Turdus eunomus Temminck 
Four specimens, all unmarked as to sex: Futsing, Fokien, 1912. 


Turdus hortulorum Sclater 
One adult male: Futsing, Fokien, 1912. 


Turdus chrysolaus (Temminck) 
Two females: Futsing, Fokien, April 1912. 


Turdus aureus aureus Holander 


Two specimens, a female and one with sex not determined: Futsing, 
Fokien, February 1912 and 1912. 


Monticola solitarius pandoo (Sykes) 
One adult male: Futsing, Fokien, March 1912. 


Monticola solitarius philippensis (Miiller) 
Two males: Futsing, Fokien, 1912. 


Monticola solitarius magna La Touche 
Two specimens a male and a female: Futsing, Fokien, May 1911 
and September 1912. These are large birds, apparently migrants of the 
big race that breeds in Northeast Siberia and Japan. The wing in the 
male gives 128 and in the female, 120 mm. 


Enicurus sinensis Gould 


Two males: Futsing and Ling Sioh, Fokien, July 27, 1916, and April 
1912. 


1921] Bangs, Birds of the Asiatic Expedition 609 


Enicurus schistaceus Hodgson 


Three adults, two males and a female: Futsing and Ling Sioh, 
Fokien, July 31, 1916, June 21, 1914, and June 1914. 


Chimarrhornis fuliginosa fuliginosa (Vigors) 
One female: Futsing, Fokien, September 1912. 


Phenicurus auroreus auroreus (Gmelin) 
One male: Futsing, Fokien, 1912. 


Ianthia cyanura (Pallas) 
One male: Futsing, Fokien; no date of capture. 


Copsychus saularis saularis (Linnzus) 
Eight specimens, adults of both sexes and one immature: Futsing, 
Fokien, July 4, 10, 23, and 24, 1916, and March 1912. 


Saxicola torquata stejnegeri (Parrot) 
One female: Futsing, Fokien, 1912. 


SYLVIIDZ 
Locustella ochotensis (Middendorff) 
One specimen: Futsing, Fokien, May 1911. 


Sutoria sutoria phyllorrhaphea Swinhoe 
Two males: Ling Sioh, Fokien, July 1 and 30, 1916. 


Cisticola cisticola tintinnabulans (Swinhoe) 
Two specimens, male and female: Futsing, Fokien, May 1911. 


LANIIDZ 
Lanius tigrinus Drapiez 
One adult male: Futsing, Fokien, July 10, 1916. 


Lanius schach schach (Linnzus) 


Six specimens, adults of both sexes and one immature female: Fut- 
sing and Ling Sioh, Fokien, July 22, 24, 27, and 28, 1916, and December 
1912. 


Lanius cristatus lucionensis Linnzeus 


Three specimens, an adult male and two immature: Futsing, Fokien, 
May 1911, September 1912, and 1912. 


610 Bulletin American Museum of Natural History [Vol. XLIV 


PARIDZE 
Parus major commixtus Swinhoe 


Five specimens, adults and immature: Futsing, Fokien, July 1, 10, 
21, 23, and 24, 1916. 


ZOSTEROPIDA 
Zosterops palpebrosa simplex Swinhoe 


Eleven specimens, both sexes: Futsing, Fokien, July 1, 3, 4, and 
10, 1916, and October 1911 and 1912. 


DICHZIDZ 
Diceum ignipectus ignipectus (Hodgson) 


Three specimens, a male and two females: Futsing, Fokien, 1911 
and 1912. 


MOTACILLIDZ 
Motacilla alba leucopsis Gould 
Two specimens, an adult male and an immature female: Futsing, 
Fokien, July 24, 1916, and September 1912. 
Anthus hodgsoni Richmond 
Two males: Futsing, Fokien, January and March 1912. 


ALAUDIDE 
Alauda gulgula celivox Swinhoe 
One adult male: Futsing, Fokien, April 1912. 


FRINGILLIDZ 
Eophona migratoria migratoria Hartert 
Four adults, three males and a female: Futsing, Fokien, April 
1912 and 1912. 
Chloris sinica sinica (Linnzeus) 
Two adult males: Futsing, Fokien, Apri 1912. 


Fringilla montifringilla Linnzeus 


Two specimens, sex not determined: Futsing, Fokien, 1912. 

Possibly there may be an eastern subspecies, Fringilla montifringilla 
subcuneolata Kleinschmidt. But after carefully comparing long series of 
specimens from the far east with European birds, I find the only char- 


1021]; - Bangs, Birds of the Asiatic Expedition 611 


acters claimed by Kleinschmidt—the size and distinctness of the paler 
marking of the outer tail feather—are very variable in both series and, 
in our material at least, do not indicate the existence of such a race. 


Passer montanus taivanensis Hartert 


Four specimens, two adult males, two immature males: Futsing and 
Ling Sioh, Fokien, July 28 and 31, 1916. I can refer these skins to no 
other form than taivanensis originally described from Formosa. 'The 
adults have bills much larger than in P. montanus saturatus Stejn. of 
Japan and the Liu Kiu Islands. The culmen affording respectively 12.5 
and 13 mm. 


Passer rutilans rutilans (Temminck) 


Ten specimens, both sexes. Futsing, Fokien, July 1, 3, 4, 10, and 
23, 1916, and January 1912. 


Emberiza spodocephala spodocephala Pallas 
Two females: Futsing, Fokien, May 1911 and March 1912. 


Melophus melanicterus (Gmelin) 


Three specimens, two males and a female: Futsing, Fokien, April 
and March 1912. 


PLOCEIDZE 
Munia punctata topela Swinhoe 


Three adults, male and two females: Futsing and Ling Sioh, Fokien, 
July 26, 1916, January 20, 1914, and March 1912. 


Uroloncha squamicollis Sharpe 


Ten adults, both sexes: Futsing and Ling Sioh, Fokien, July 26, 
28, 29, 30, and 31, 1916. 


STURNIDZE 
Sturnia cineracea (Temminck) 
Three specimens, both sexes: Futsing, Fokien, January 21, 1914, 
and 1912. 
Sturnia sinensis (Gmelin) 
Two males: Futsing, Fokien, 1912. 


612 Bulletin American Museum of Natural History [Vol. XLIV 


Sturnia violacea (Bodcaert) 

One adult male: Futsing, Fokien, May 1911. This example is a 
fine adult male in full spring plumage, and the record proves beyond 
doubt that the species does occasionally, at least, occur in China on 
migration. 

#thiopsar cristatellus cristatellus (Gmelin) 


Five specimens, adults of both sexes and one immature female: 
Futsing and Ling Sioh, Fokien, July 27 and 31, 1916, and March 1912. 


ORIOLIDZ 
Oriolus indicus indicus Jerdon 


Seven specimens, both sexes: Futsing, Fokien, July 4, 20, and 24, 
1916, and March 1912. 


DICRURIDZ 
Chibia hottentotta brevirostris (Cabanis) 
One immature male: Ling Sioh, Fokien, August 2, 1916. 


Dicrurus leucogenys cerrussatus (Bangs and Phillips) 
Four adults, both sexes: Futsing, Fokien, March and April 1912. 


CORVIDE 
Corvus coronoides levaillantii Lesson 
One adult male: Futsing, Fokien, April 1912. 


Urocissa erythrorhyncha erythrorhyncha (Gmelin) 
Nine specimens, adults and young of both sexes: Futsing and Ling 
Sioh, Fokien, July 24, 27, and 28, August 1, 1916, and March 1912. 
Dendrocitta formose sinica Stresemann 
Two adult males: Futsing, Fokien, April 1912 and June 1914. 


Garrulus bispecularis sinensis Swinhoe 


Four specimens, three adults, one immature, two marked as males, 
two without sex mark: Futsing, Fokien, January and April 1912. 


eae Pe “fia nia hate ie 
ie ie tis + es 


‘ i, ~? 


poe “H a 


: ‘| Fy 3 
Baths Coes a Aue Rice: ee 
age an Fei Reel r hg BB 


‘ 


a DESCRIPTION. OF A NEW LOACH FROM 
_ NORTH-EASTERN CHINA 


% %, f = 4 } op ela t been 


pr tad 


Oho ONO By FEN RYO ROWLER © SON ya ena oe 


Issued May 25, 1922 

iy A é 
; FE}, < 

By. ORDER oF THE TRUSTEES i 


t 


“THE AMERICAN ‘MUSEUM OF NATURAL HISTORY oe . 
New Yor Atv 


AMERICAN MUSEUM NOVITATES 


Number 38 May 25, 1922 


59.7,55L (51.1) 


DESCRIPTION OF A NEW LOACH FROM NORTH-EASTERN 
CHINA 


By Henry W. Fow ier 


Lefua andrewsi,! new species 


Head 4%; depth 7; D.u, 6; A.1,6; P.1, 12; V.I, 6; scales about 104 in a 
median lateral series; head width about 134 in its length; head depth 2; snout 344; 
eye 4%; maxillary 3%; interorbital 340; depressed dorsal 134; depressed anal 1%; 
least depth of caudal peduncle 2; caudal length 14; pectoral 144; ventral 14. 

Body elongate, moderately slender, considerably depressed forward and becom- 
ing compressed posteriorly, edges all convex except slight keel forward above and 
below on caudal peduncle by rudimentary caudal rays to caudal base. Caudal 
peduncle strongly compressed, least depth little less than its length. 

Head moderate, robust, broadly depressed, especially behind. Snout broad, 
obtuse, length % its width. Eye small, hind edge about midway in head length. 
Maxillary small, about half-way to eye. Jawseven. Lipsrather thin. Nasal barbel 
reaches eye center. Maxillary barbel to hind eye edge. Upper lateral barbel to eye 
center. Interorbital broadly though slightly convex. 

Gill-opening lateral, long as snout. 

Scales all small, not overlapping, in rather irregular distribution though close-set 
and with imbedded appearance; marginal radiating strie 31 to 37; circuli moderately 
fine. No developed lateral line. 

Dorsal origin little nearer that of pectoral than to caudal base, depressed fin 
slightly less than caudal base. Anal inserted little behind dorsal base, though little 
before depressed dorsal tip, depressed fin three-fourths to caudal base. Caudal 
rounded, median rays longest. Pectoral about half-way to ventral. Ventral reaches 
about three-fourths to anal. Vent close before anal. 

Color in alcohol nearly sepia above, dusted very obscurely with darker. Dusky 
lateral band, rather obscurely defined, from each side of snout tip to eye, though 
below and over infraorbitals, back to caudal base. Posteriorly band much darker 
to blackish. Dark vertebral line on predorsal, slightly so behind dorsal. Barbel 
edges and lip margins dusky. Iris paleslaty. Dorsal and caudal grayish, both finely - 
and obscurely spotted with dull dusky, only a distinct median black blotch on latter, 
reflected out on median rays basally. Other fins all pale, pectorals with few shadings. 

Length, 52 mm. 

Type, No: 7974, American Museum of Natural History. Shing Lung Shan, 
Eastern Tombs, China. August 7, 1921. Collected by The Third Asiatic Expedition 
of The American Museum of Natural History. 


1In recognition of Mr. Roy Chapman Andrews, leader of the Third Asiatic Expedition of The 
American Museum of Natural History. 


2 AMERICAN MUSEUM NOVITATES [No. 38 


This species is closely related to Lefua costata (Kessler) and appears 
to differ only in the color-pattern. Apparently the two forms occur asso- 
ciated, as they were received in the same lot. In Lefua andrewsi the 
broad and well-defined lateral band of dusky to blackish, which is re- 
flected out on the median caudal rays, is diagnostic. In Lefua costata 
the scarcely evident lateral streak is replaced at the caudal base by a 
definite small rounded black spot, clearly defined and not reflected out 
on the median fin-rays. 


Bus 


nae By Weiliee Guicis As Cuarzes P, Beane 


Issued August 7, 1922 


By site oF THE TRUSTEES 


OF, 


THE AMERICAN - MusHUM OF ‘NATURAL HISTORY 
EE ‘Naw. Yore Crrx = RES eee 


oo 


my cat : 


AMERICAN MUSEUM NOVITATES 


Number 42 August 7, 1922 


56.(117:51.7) 
DISCOVERY OF CRETACEOUS AND OLDER TERTIARY 
STRATA IN MONGOLIA! 


By WALTER GRANGER AND CHARLES P. BERKEY 


The American Museum commenced its natural history explorations 
in Asia in 1916. The First and Second Asiatic Expeditions, in charge of 
Roy Chapman Andrews in 1916-1917 and 1918-1919, were engaged in 
zoological exploration and in laying the foundations for broader work. 
The Third Asiatic Expedition, sent out by The American Museum of 
Natural History, the American Asiatic Association, and Asia Magazine, 
has included zoélogy, paleontology, geology, and geography under the 
leadership of Mr. Andrews, with Walter Granger as paleontologist, 
Charles P. Berkey as geologist and Frederick K. Morris as topographer, 
and other cognate lines of research may be taken up when the results of 
reconnaissance warrant it. 

The scientific results of these expeditions will be Saab: in 
numbered sequence as indicated below. The following reports or con- 
tributions have already been published: 

(No. 1) ‘New Chinese Fishes.’ By John Treadwell Nichols. Proc. Biol. Soe. 
Washington, XX XI, pp. 15-20, May 16, 1918. 

(No. 2) ‘Description of a New Species of Serow from Yiin-nan Province, China.’ 
By Roy Chapman Andrews. American Museum Novitates, No. 6, March 24, 
1921. 

(No. 3) ‘The Birds of The American Museum of Natural History’s Asiatic Zodélogical 
Expedition of 1916-1917.’ By Outram Bangs. Bull. Amer. Mus. Nat. Hist., 
XLIV, Art. 20, pp. 575-612, December 30, 1921. 

(No. 4) ‘Description of a New Loach from North-eastern China.’ By Henry W. 
Fowler, American Museum Novitates, No. 38, May 25, 1922. 


RECONNAISSANCE EXPEDITION IN MONGOLIA 


On April 21, 1922, the Third Asiatic Expedition left Kalgan, North 
China, for its announced reconnaissance trip into Mongolia. It is 
planned to devote the first three weeks of the season to observations along 
the regular caravan route between Kalgan and Urga, the capital of 
Mongolia, and the rest of the season to points scattered far to the west, 
perhaps even as far west as Ulyosutai and Kobdo and the eastward 
extension of the Altai mountains. 


1Contribution No. 5. Asiatic Expeditions of The American Museum of Natural History. 


2 AMERICAN MUSEUM NOVITATES — [No. 42 


The chief effort of the present season is to be devoted ‘o geology, 
paleontology, geography, and zoélogy, but other scientific interests will 
be cared for in subsequent seasons if the reconnaissance warrants such 
expansion. It is believed that fields inviting more extended and detailed 
work will be discovered and that the reconnaissance will furnish a basis 
for final plans and indicate the nature of the problems that promise best 
results. 

At the close of the season President Henry Fairfield Osborn is ex- 
pected to join the expedition staff at the headquarters in Peking and will 
take an important part in the conferences in which plans for the next three 
years w ll be formulated. . 

The scientific staff on the present reconnaissance includes Roy 
Chapman Andrews, zodlogist, Walter Granger, palzontologist, Charles 
P. Berkey, geologist, and Frederick K. Morris, physiographer. 

It is hoped that there may be opportunity to send short notes of 
observations or discoveries of special interest directly from the field but, 
in any case, a summary of the season’s results will be issued with little 
delay on the return of the Expedition. 

CRETACEOUS STRATA IN EasTERN Asia.—The Third Asiatic Expedi- 
tion announces, under date of May 3, 1922, that strata of Cretaceous 
age, overlain by two distinct Tertiary formations, have been discovered 
in the Gobi region of southeastern Mongolia. 

They were found on the outbound trip from Kalgan to Urga at a 
point about 260 miles northwest of Kalgan. Strata of Cretaceous age are 
wholly unknown in Eastern Asia, as far as the writers are aware, and be- 
cause of the apparent importance of the find, it was decided to leave the 
geologists in camp at this place while the rest of the party moved on. 
Accordingly Messrs. Berkey, Granger, and Morris spent a week in addi- 
tional inspection of the ground and furnish the notes for this memoran- 
dum. 

Th _ best exposures of the Cretaceous formation are in the vicinity of 
the small salt marsh Iren Dabasu, where a total thickness of about 150 
feet of nearly horizontal strata is judged to be of thisage. Tertiary beds 
not older than the Miocene lie on top of the Cretaceous strata and are 
best exposed about five miles south of Iren. Twenty miles farther south 
early Tertiary beds were found in essentially the same relation. 

In each occurrence of the Tertiary beds only a single horizon has 
furnished determinative fossils, but in the Cretaceous formation below, 
‘ there are at least two fossil bone-bearing horizons. Fortunately the 
faunal evidence is unmistakable. Otherwise the widely different age 


1922] CRETACEOUS AND TERTIARY IN MONGOLIA 3 


relations of the strata would not be suspected, for the corresponding 
physical breaks are inconspicuous and the beds are almost perfectly con- 
formable. 

The structural basin in which these strata lie measures forty miles 
across from north to south and is floored with ancient slates and lime- 
stones of extremely complicated deformation structure. This is only one 
of six basins of similar form and relation between Kalgan and Iren but it 
is much the longest one and the only one in which, thus far, the presence 
of strata of Cretaceous age has been proven. 

In the vicinity of the small salt lake Iren Dabasu, the Cretaceous 
beds lie immediately on the slate floor of the basin and between this base 
and the first determinable beds of later age, in this case late Tertiary, 
about 150 feet of strata are exposed. The bottom members are domi- 
nantly sands and sandstones, prevailingly thin-bedded, some of which 
are strongly cross-bedded and well cemented. The middle members be- 
come finer grained, more mixed with clay and more variable in color. 
The upper beds are dominantly claysand sandy clays and very fine sands, 
varying in color from white to dark red and drab and yellowish green. 
No less than twenty distinct beds or layers can thus be distinguished, 
all of which are regarded as belonging to a single geologic formation. 

Only the lower members of this formation have been found to be 
fossiliferous. The list includes: 


1.—Predentate dinosaurs, probably of the bipedal type. 
2.—Carnivorous dinosaurs of at least two genera, the smaller one 
being of the Ornithomimus type. 
-  3.—Crocodiles. 
4.—Turtles of the Trionyz type. 
5.—A few pelecypod shells. 


Obretcheff, the Russian geologist, who gives an account of a recon- 
naissance trip over this same route from Ude to Kalgan, describes sedi- 
mentary beds at many places, always referring to them as representa- 
tives of the Gobi formation. His only age determination, however, was 
made on the basis of a few fragments of Rhinoceros, found at the escarp- 
ment five miles south of Iren. These remains were judged by Eduard 
Suess, to whom they were referred, to indicate an age not earlier than the 
_ Miocene. The Tertiary age of the rest of the occurrences mentioned by 
him seems to have been taken for granted and apparently that is in 
general correct, but it is evident that the Gobi formation cannot properly 


+ AMERICAN MUSEUM NOVITATES*. © [No. 42 


include strata of both Tertiary and Cretaceous ages. It is clear also that 
the term Gobi formation or Gobi series is properly applied to the Tertiary 
beds instead of to those of Cretaceous age. The finding of a Cretaceous 
formation below makes a new designation necessary. For this purpose 
nothing seems to be as appropriate as the name of this locality. We 
therefore propose the term IREN DaBasu FORMATION for these beds. | 


° ° 
og 116° 120,40 


Scale of Miles 
100 150 9 250 300 


5 


Fig. 1. Sketch map showing location of type sections of Iren Dabasu Formation (Cretaceous)1; 
Irdin Manha Formation (Eocene) 2; Houldjin Formation (Miocene) 3. 


Tue Hovuupsin Beps (Mippie Trertiary).—For the late Tertiary 
beds found five miles farther south and belonging to the Gobi Series of 
Obretcheff we propose the term Houups1n Beps, taken from the local 
name of the upland formed by these beds. They are characterized by the 
following fossil content: 


1.—A rhinocerid. 

2.—A large carnivore. 

3.—An artiodactyl of the size of a Virginia deer. 

4.—An enormous mammal, probably a perissodactyl and possibly 
related to or identical with Baluchitherium, discovered by Forster Cooper 
in Baluchistan. 

5.—A tortoise of large size. 


1922} CRETACEOUS AND TERTIARY IN MONGOLIA 5 


There is a sharp physical change immediately below this formation 
and only the coarse sandy conglomeratic member at the very base has been 
found to be fossiliferous. The fossil remains are unusually fragmentary. 


IrpiIn Manua Formation (Harty Trrtiary).—For the early 
Tertiary beds found twenty-five miles farther south, also assumed 
properly to belong to Obretcheff’s Gobi Series, we propose to use the 
term IrpIn MANHA FORMATION. It appears to lie immediately on Cre- 
taceous beds, the Iren Dabasu formation, and again there is a sharp 
change in type of rock. The beds are cross-bedded sandstones, limy 
sands and pebbly sandstones. Only the lower member has been found to 
be fossil-bearing. It is characterized by the following forms: 


1.—Small Lophiodonta of at least two species in great abundance. 

2.—A perissodacty] about the size of the Upper Eocene titanotheres 
and possibly related to this family. 

3.—A small artiodactyl. 

4.—A small creodont. 

5.—An abundance of turtles of both the hard-shelled and soft- 
shelled groups. 
' 6.—Teleost fishes. 


The geologic column for the Iren Dabasu basin therefore is essen- 
tially as in the following table. 


Recent Uplift and Erosion 
Peneplanation 
Miocene or Upper barren sands 25’+|The Houldjin 
Later Rhinoceros gravels 5’ |Formation : 
Tertiar : : ; : The Gobi 
ertlary Oligocene or Upper barren sand- 25’+|The Irdin Dorie 
Eocene stones Manha 
The Lophiodont bed 4’ |Formation 
Physical and Faunal Break 
Upper barren members, chiefly clays, 
marls and fine sands 90’ |The Iren 
Cretaceous Dabasu 
Lower or Dinosaur beds, chiefly sands} 60’ | Formation 
and sandstones 
ny Great unconformity 
Pre- The old-rock floor, chiefly slates, Probably 
Cretaceous limestones and igneous rocks The Nank’- 
on Series 


6 AMERICAN MUSEUM NOVITATES [No. 42 


VERTEBRATE Fossits, ADDITIONAL Deraiis.'—Remains in all three 
beds are fragmentary, decidedly so in the Houldjin grave's, but they 
are of unusual interest apparently and we have taken everything which 
has any character. Dinosaurs are represented by one complete tibia, 
ends of femora and humeri, presacral and caudal centra, many good foot 
bones, including claws of fore and hind feet, portions of a small carnivor- 
ous dinosaur skull with two or three teeth, and two teeth of a predentate, 
as well as two portions of jaw with the alveoli of some teeth, also 
predentate. Remains of the small Ornithomimus-like creature are 
particularly abundant and the last day at Iren Dabasu we picked up 
probably fifty good foot bones and centra from two or three knolls. We 
could find no teeth of the little fellow though—wonder if he was edenta- 
tate like Struthiomimus ? The Cretaceous exposures are very lim ted so far 
as we could see but may, of course, outcrop in other basins to the east 
or west of the road. We did not have time to extend our work in either 
direction. The outcrops we did see will stand a more careful going over. 

The Houldjin gravels are exposed as a rather thin capping to a low 
bench of Cretaceous which we followed for several miles. Things are 
badly broken up here—even such massive bones as the heads of femora 
and humeri were usually cracked into several pieces before deposition. 
There is one fine bone—a calcaneum of the big beast which would be a 
match for the astragalus of Baluchitherium? (?). I can think of nothing 
else to which it might belong. It is as long as the great Megatherium’ 
calcaneum from Long Branch, N. J., but is not edentate. A head of a 
femur is the size of one’s head and other limb bone ends correspond. 
Some enormous rhinoceros teeth (broken) may belong with this animal. 
Smaller teeth are surely Rhinoceros. We did not explore the full length of 
the exposure and there are possibilities in excavation at one or two points 
of the bluff where we did explore. 

The Irdin Manha beds offer the greatest opportunity for future work. 
Mammal'an remains are abundant though fragmentary and we examined 
less than two miles of a line of exposures extending many miles both east 
and west of the trail. A small lophiodont (Helaletes-like) is most abund- 
ant and we got numerous teeth besides two maxille (one with premaxilla 
and orbital region) and a few lower jaws, also numerous oot bones, 
limb bones and vertebree. Next in abundance is a perissodactyl, looking 
much like our late Eocene titanotheres. We have several premolars, 
many incomplete molars and one lower jaw with p;—m; in fair condition. 


1Communicated in a letter by Mr. Walter Granger, dated May 10, 1922. 
2A gigantic perissodactyl described by C. Forster Cooper from Baluchistan. 


1922] CRETACEOUS AND TERTIARY IN MONGOLIA e 


Other forms are curiously rare, a creodont lower jaw and an artiodactyl 
astragalus or two being the only things noted. Trionychids are common 
and we saw a complete though badly broken carapace which we were 
hurrying to get to our car before a storm overtook us the last day we 
were there. We made three trips down from Iren Dabasu camp but 
could not do more as our food was getting short and we had to join the 
rest of the party here. 

. Much additional detail is in possession of the Expedition which 
will appear in due course, and it is expected also that further investiga- 
tion of this area and related ones will be made at a later time. 

Tur TuHirp Asiatic EXPEDITION. 


A eee 


Te 3 


‘ er Sebo ai : 


+ 


oak >. 


atone: be) will Ratings Bb bree ciel 
te ua weal eta aes Ligtae’ antl sa iesaaerirr os 


raietesas RES LBA EUR MMe eT Ee a Oat 


a) 


—s 
4 
y 


ce ae ey a 
. — ar 
o.2 
, 7 
ia) 
> 


y 


AMERICAN MUSEUM NOVITATES 


Number 72 May 4, 1923 


56.81,9P(117:51.7) 
~PROTOCERATOPS ANDREWSI, A PRE-CERATOPSIAN DINO- 
SAUR FROM MONGOLIA! 


By WALTER GRANGER AND WILLIAM K,. GREGORY 


With an Appendix on the 
STRUCTURAL RELATIONS OF THE PROTOCERATOPS BEDS 


By Cuarues P. BERKEY 


The type of Protoceratops andrewsi, new genus, new species, was 
discovered on September 2, 1922, by Granger and party on the Kwei- 
wa-ting trail, east of Artsa Bogdo, Mongolia, during a_ preliminary 
geological and paleontological survey conducted by the Third Asiatic 
Expedition of The American Museum of Natural History. The speci- 
men consists of a skull, lacking the occiput. It was found by Mr. 
Shackelford in exposures of red shale in a formation which has been 
provisionally referred to the Cretaceous by Professor Berkey.? 

The skull (A. M. N. H. No. 6251) is hornless and far smaller than 
that of any known ceratopsian or ankylosaur, being only about 160 mm. 
in length from the anterior end of the premaxilla to the posterior border 
of the jugal. As seen from above, it is broadly triangular, with a pointed 
apex and wide lateral crests, the latter composed chiefly of the backward- 
and-downwardly expanded jugals. The greatest width of the skull 
across the posterior borders of the jugals is about 190 mm., while the 
depth of the jugal below the middle of the orbit is 43 mm. The orbits 
are very large (50 mm. in anteroposterior length), not surmounted by 
supraorbital bones or horns. The postorbital-squamosal bar is narrow. 
Parts of the anterior and lateral borders of the supratemporal fenestra 
as preserved indicate that the fenestra was large and that the occipital 
roof was very delicate and not produced as far backward as in later 
Ceratopsia. The squamosal broadly overlapped the enlarged jugal and 
was produced posterosuperiorly but was not greatly enlarged. The 
pineal foramen is small or absent. The single preorbital fossee are far 
larger than in other predentates. The premaxille were very large and 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Publication 


0. 6. 
2See Appendix, p. 7, below. 


(No. 72 


AMERICAN MUSEUM NOVITATES 


Position of mandible 


Protoceratops andrewsi, type skull, side view. 


Fig. 1. 
corrected. x 


PD. 


Ses, 
SSSSV_zs~» 


Protoceratops andrewsi, type skull, top view. 


2. 


Fig. 


1923] A PRE-CERATOPSIAN DINOSAUR 3 


probably supported a large rostral bone, which is broken off; the pre- 
maxillz and nasals approach the ceratopsian type and the same is true 
of the pterygoids, the internal nares, and the quadrates. The quadrato- 
jugal lies on the posterior surface of the quadrate. 

The mandible has on each side a straight row of about nine relatively 
large and long-crowned teeth, worn on their buccal sides and set far 
inward toward the midline. The remains of the lower molar crowns 
suggest the three-pointed lower molars of ceratopsians, rather than the 
spatulate, many-cusped teeth of ankylosaurs and of European Acan- 
thopholide.! The anteroposterior measurement of the four teeth shown 
in Fig. 1 is 28 mm. The last tooth preserved has its tip about 13 mm. 
above,the alveolus. The first four teeth are represented by their alveoli. 
The diastema from the first alveolus to the predentary bone was about 
14mm. in length. The strong coronoid process rises from the dentary at 
a gentle slope. The predentary bone is well developed and has a pair of 
long inferior processes, one on either side of the midline. 

At first sight the specimen suggested the Procolophonia in the very 
large size and backward prolongation of the orbits and in the presence of 
a lateral crest below and behind the orbit; but reference to that group is 
excluded, especially by the absence of a large pineal foramen, by the 
fact that the lateral crest is composed of the jugal instead of the quadrato- 
jugal, by the presence of a predentary bone and by the characters of 
the dentition. Mezolania, Elginia, the pariasaurs and other reptiles 
with flaring lateral crests all differ from Protoceratops in fundamental 
characters. 

The presence of a predentary bone and the characters of the 
mandible and dentition positively determine the specimen as an orni- 
thischian (orthopod) dinosaur. Of the Ornithopoda none of the known 
skulls have expanded lateral crests and there is a general tendency toward 
dorsoventral flattening of the beak. The squamosal is reduced and 
widely separate from the jugal, the latter not greatly expanded pos- 
teriorly. The Jurassic Hypsilophodon has a relatively short and primi- 
tive type of skull (Marsh, 1896, Pl. txxxiv), which might well be the 
starting-point for the far more specialized conditions of Protoceratops. 

Of the Stegosauria the most primitive is the Liassic Scelidosaurus, a 
longer skull, the details of which seem to point toward Stegosaurus. In 
the latter the squamosal is small and widely separated from the small 
jugal, the coronoid process of the dentary is reduced or wanting and 


1For figures of all these see Nopesa, F. B., 1918, ‘Leipsanosaurus, n. gen. ein neuer Thyreo- 
phore aus der Gosau,’ Sep. Féldstani Kézlony, XLVIII,Taf. ur. 


d AMERICAN MUSEUM NOVITATES [No. 72 


the beak is somewhat flattened dorsoventrally. The Acanthopholide 
have small heads and spatulate, many-cusped teeth. The ankylosaurs 
agree with Protoceratops, rather than with Stegosaurus, in the characters 
of the temporal region, but have acquired a heavily armored skull roof 
and expanded muzzles. 

The true Ceratopsia, hitherto unknown below the Upper Cretaceous 
of America, are all far larger than Protoceratops; all of them have horns; 
the crest is much expanded above and behind the occiput; there are 
epoccipital and supraorbital bones; and the orbit is small, placed high 
up and bounded by a wide postorbital bar. The preorbital fossa is 
reduced to a small slit. 

As Protoceratops presents the opposite of these characters, it may 
prove necessary to erect for it a new suborder (Protoceratopsia) but we 
prefer at present to regard it only as the type of a new and probably 
primitive family, the Protoceratopside, characterized by the lack of 
horns, the very large size of the orbits, and the narrowness of the post- 
orbital-squamosal bar. 

Protoceratops thus stands far below the Upper Cretaceous ceratop- 
sians and structurally it tends to bridge the long gap between the latter 
‘and such primitive Jurassic Ornithopoda as Hypsilophodon. 

The Protoceratops skull tends also to settle the relationships of the 
ankylosaurs. The latter differ widely from Stegosaurus and resemble the 
Ceratopsia in the temporal region of the skull, in the reduction of the 
pubis to a vestige and in the outward growth of the dorsal border of the 
ilium. Abel! groups them with the Ceratopsia, and Protoceratops may 
prove to be near the common ancestor of the two groups. 

In conclusion, the discovery of Protoceratops constitutes one of the 
foremost items of direct evidence in support of the view advocated 
especially by Osborn and Matthew, namely that, as, the palzeontologic 
record of Asia is more fully explored, it will fill many gaps in our knowl- 
edge of the origin, evolution, and migrations of the late Mesozoic and 
Tertiary faunz of western North America and Europe. 

We therefore take pleasure in dedicating this important type to 
Mr. Roy C. Andrews in recognition of his splendid qualities as the 
organizer and leader of the American Museum Third Asiatic Expedition. 


1Abel, O., 1919, ‘Die Stimme der Wirbeltiere,’ p. 653. 


¢ 


‘azIs Jeanyeu SparTygy-OMy yNoqy “MeLA apIs enbiTqo “YNys edAq ‘2smaupuy sdoyw.iav0j01g *g “SI 


Fig. 4. Protoceratops andrewsi, type skull, top view. Two-thirds natural size. 


1923] A PRE-CERATOPSIAN DINOSAUR ‘ 


APPENDIX 
SrrRUCTURAL RELATIONS OF THE PROTOCERATOPS BEDS 


By Cuarues P. BERKEY 


The type Protoceratops andrewsi, described by Professor Gregory, 
was found on the return journey by Mr. Shackleford in making a rapid 
inspection of some ground a short distance from the trail while waiting 
for the rest of the party. A few minutes later all came up and joined in a 
search of the locality. The finds made in the first few minutes of the 
stop netted some fine specimens, although none surpassed the first one; 
and all proved to be so unusual in character that it was decided to spend 
the remaining two hours of daylight in fossil hunting. The next morning 
the expedition moved on. 

The ground would undoubtedly richly reward a more extended 
investigation. We touched only one spot and each one of the party 
carried off a load of specimens, leaving behind in our hurry many others 
either too fragmentary or too heavy or too much imbedded in the rock 
for recovery. 

The spot is on the north side of the Kwei-wa-ting trail, 50 miles east 
of Artsa Bogdo. The rocks are red, friable sandstones and shaly sand- 
stones which are very well exposed at this point by erosion. Badland 
cliffs and remnants, more than 200 feet in total relief, form an escarpment 
here and mark the beginnings of a considerably dissected country extend- 
ing for many miles northward and eastward, quite in contrast to the 
smooth peneplane surface over which the trail had led to this point. 
In fact, at a distance of less than half a mile the escarpment is not notice- 
able from the upper plain, although one can see that there is a belt of 
lower ground off to the side. 

Beds of the same series and of apparently the same physical rela- 
tions were crossed by Morris and Berkey on their side trip with camels 
from Artsa Bogdo to a large mountain group known as the Gurban 
Saikhan. Several hundred feet of red, sandy beds were seen on the north 
margin of the Gurban Saikhan, but where the examination was made the 
beds were barren. The Kwei-wa-ting trail, where the fossils were found, 
passes north of the Gurban Saikhan at a distance of 20 or 30 miles out on 
the open plain. 

Although there was no time for local side study of the stratigraphic 
relations in this vicinity, enough of the geology was determined by this 
earlier trip to the Gurban Saikhan and by the continuous route-cross- 
section work kept up by the geologic staff to fix these strata structurally 


8 AMERICAN MUSEUM NOVITATES [No. 72 


within certain well-prescribed limits. They lie well above the great 


Jurassic or post-Jurassic unconformity, which is the most marked struc- 
tural break in central Mongolia. They also lie beneath an early Tertiary 
or pre-Tertiary unconformity of much less physical prominence, these 
strata thus partaking of a deformation that antedates all Tertiary 
sediments. 

They are to be regarded, therefore, as belonging to the same series 
that has been referred to in our reports as of Cretaceous age, using the 
term in its large sense to cover everything thus far found between the 
Jurassic strata on one side and Tertiary beds on the other. 

This series doubtless does cover a very wide range. Some of the 
beds may correlate with the Comanchic of America. In Mongolia the 
series must for the present be kept flexible enough and broad enough to 
include the dinosaur-bearing beds of Iren Dabasu (already described in 
Amer. Mus. Novitates No. 42), the Ondai Sair dinosaur-bearing forma- 
tion of the Hsanda Gol region, and the dinosaur-bearing Ashile formation 
of the basins north of Artsa Bogdo. 

The relative positions in the time scale of these different local devel- 
opments are yet to be determined, but they probably can be fixed 
definitely with the material already collected or to be collected this year. 
A tabulation of locality formational terms, without insistence on the 
significance of the order, is as follows: 


) 


Cenozoic | Tertiary 
ener Ghee helooa a Re [z-.... Umeonformity....,.-.)b0' > 6... eer 
(Iren Dabasu Formation 
Later Cretaceous Shamo ae ae a 
Mesozoic Series Ashe Sora 
i | | Dja-doch-ta Formation 
(Protoceratops Beds) 
Great Unconformity 
5 Dai Sabres acre afar oce eae aE os eaten bee aa cau sli sae Mee 
Earlier | Tipass a | 
. | Jurassic | 
Mesozoic | | 


The beds seen at the Gurban Saikhan, together with these at Dja- 
doch-ta furnishing the Protoceratops remains along the Kwei-wa-ting 
trail, doubtless are identical with those seen by Chernov, the geologist 


ee ee ee en 


oS Pues 


1923] A PRE-CERATOPSIAN DINOSAUR : 


of the Kosloff expedition and referred to by him as the Red Khan-Khai 
beds. Khan-Khai is a well-established term introduced by von Richth- 
ofen and has been widely used, apparently rather indiscriminately, for 
any or all of the later sedimentary beds supposed by the earlier ob- 
servers to have been formed in the disappearing or evaporating sea. But 
it is loosely used and undoubtedly has served to cover strata of a large 
range of age relations. Perhaps it is inadvisable now to attempt any 
narrower limitation. Jt is proposed therefore to introduce the term 
Saamo Series for all of the later Mesozoic strata above the Great 
Unconformity. : 


AMERICAN MUSEUM NOVITATES 


. Published by 
Number qa. Tue AmeRIcAN Museum or NatTuraL History May 25, 1923 
New York City 


55.1,77(51.7) 
LATER SEDIMENTS OF THE DESERT BASINS OF CENTRAL 
MONGOLIA! 


By CHarues P. BERKEY AND WALTER GRANGER 


A preliminary notice of the discovery of Cretaceous and Tertiary 
strata, based on studies of a small sedimentary basin at Iren Dabasu 
within the first 265 miles of exploratory travel of the Third Asiatic 
Expedition, was published as No. 42, American Museum Novitates, 
August 7, 1922. The expedition subsequently covered approximately 
3000 miles of scientifically unknown territory in Mongolia, crossing the 
Gobi Desert twice, reaching as far north as Urga, the capital, near the 
Siberian frontier, as far west as Sain Noin Khan and Tsagan Nor, and 
- finally crossing the easterly range of the Altai Mountains with its 
bordering desert basins, in a side journey, as far south as the Gurban 
Saikhan. From these advance positions the return course was by a new 
route to Kalgan, nearly 1000 miles distant from the last general camp at 
Tsagan Nor. (See accompanying map for itinerary.) 

Special and first attention was given to the later sedimentary 
strata which seemed to promise reptilian and mammalian fossils. But 
the continuous route study that was conducted by the geologic staff, no 
matter how rapidly or erratically the expedition moved, has given basis 
also for an outline both of the older geologic features, the structural 
and deformational detail of the later fossil-bearing sedimentary forma- 
tions of the desert basins and of the still later physiographic history. 

An advance summary of the stratigraphy of the later sediments is 
the purpose of this present statement. Other numbers are to be pre- 
pared, one for the older geologic features and one for the physiography. 

The observations forming the basis of this study were made between 
April 19 and September 19, 1922 and cover the whole field season, includ- 
ing the work at Iren Dabasu already announced.? 


I.— GENERAL GEOLOGIC FEATURES 


Central Mongolia is structurally a series of later sedimentary basins 
underlain by a floor of more ancient rocks. These basins of later sedi- 


_ Publications of the Asiatic Expeditions of The American Museum of Natural History. Publi- 
cation No. 7. Based on the field observations of Fiederick K. Morris, Physiographer, Walter 
Granger, Paleontologist and Charles P. Berkey, Geologist, of The Third Asiatic Expedition of The 
American Museum of Natural History. 

2No, 42, American Museum Novitates, August 7, 1922, 


© 


“‘souleu [osuoy a3¥1mdoudde Aq peyeorput 
a1B Spley [Issof JuBJOduIT se seded ay} UI OF palIejod SaIzI[BOO] BY} JO [TY “UMBAP 1B s[1B14 UBABIBO YSNOIY] [BdioulId 944 AJUO pus ‘payreUr o1B sa.inyBey Jor[er JOfeur 914 
ATWO WOISNJUOD PIOAB 07 Iapio UT “UBS[BY 4B DurauIdseq oul] AAVOY PI[OS Ul poyBOIpUL SI YOIYM UOoTZIpedxe oy} Jo AresOUTAT BY} S19AOD 4Y ‘UOTVRTIduIOD w st deur sty, 


UOTIIPOdx OYRVISY patty, oy} Jo dvyy oynoy [Btoudy "f “SyT 
00% 00! o os ool 
31vos 


"a 


AON” 

uniyso5—, 
5 / 
— 


4ON-P 
~ Kon 3) 


ae 


5S 
o 


1923] LATER SEDIMENTS OF CENTRAL MONGOLIA 3 


mentary strata are separated one from another by stretches of open 
ground where these same ancient floor-rocks form the surface. This 
floor has a very complicated structure, its members ranging in age 
from Archzan to Jurassic time. In mid-Mesozoic time, after repeated 
earlier mountain folding and extensive igneous activity, the region was 
peneplaned, and it is this old peneplane surface that has been warped 
and faulted to make the basins which hold the sediments of later age, 
now proven to be exceptionally good fossil ground. 

Since that time the region has never been subjected to mountain 
folding again, but it has been deformed by warping and faulting more 
than once, and it is these successive disturbances that give the clearest 
breaks and the sharpest changes in the several series of later strata. 
Wherever the deformation is confined to gentle warping the strata lie 
almost flat. Here and there, however, deformation has been more 
violent, so that the earlier strata stand steeply tilted and beds of suec- 
ceeding formations lie on their upturned edges. 

In such places the latest beds are formed in part from the waste of 
the exposed portions of beds still lying immediately beneath, and the 
break thus made between two physically distinct series helps materially 
in marking and following important formational limits. Thus there are 
oceasional sharp unconformities within the later series of sediments them- 
selves, but the evidence of it is not general. In the same region there are 
places that show only obscure disconformity or no apparent. strati- 
graphic break at all, so that one can not readily tell where the dividing 
line should be. This obscurity is all the more troublesome because a 
majority of the beds of the later series seem to be almost barren of fossils. 
Only here and there are fossils abundant or characteristic enough to 
serve as basis for age discrimination. 

Thus, in the very best structural areas, some members of the strati- 
graphic succession are missing, whereas at the places least disturbed, 
which ought to exhibit the most complete stratigraphic column, the 
earlier strata are usually covered so deeply with successive sedimentary 
accumulations that it is quite impossible to find an exposed section of the 
underlying beds. 

It was found also that sedimentation did not continue to the very 
present time, but was broken by a deformational epoch judged to be at 
about the close of the Pliocene. The Pleistocene, therefore, has been a 
period of erosion in practically all of the region traversed. Denudation 
processes of that time have not only stripped off all sedimentary cover 
from large areas of the old floor, but, incidentally, long-continued ero- 


4 AMERICAN MUSEUM NOVITATES [No. 77 


sion developed a new peneplane on the later sediments, and, within some 
of the basins, where deformation still continued, a new and compara- 
tively deep dissection has been accomplished. This has given a typical 
badland topography in certain areas, where extensive exposures of 
basin strata can be seen. 

After determining these facts about the geologic structure of the 
region, it was a comparatively simple matter to discriminate between the 
more promising and the less promising areas. The edges of active basins 
or the interiors of basins large enough to be affected by internal deforma- 
tion are particularly suitable places for detailed study. The vicinity of 
areas of high relief, indicating as they do extensive deformation of the 
ancient rock floor, are also places with some of the desirable geologic 
elements. The simpler small warp-basins, not much disturbed, may carry 
just as important beds, but, unless these beds now form the surface or 
lie very near to the surface, there is little chance of seeing them, and only 
a small stratigraphic range is likely to be exposed. 

The itinerary of the expedition led across many basins of this sort. 
These will be indicated in the forthcoming reconnaissance report. Some 
are only a few miles across from rim to rim and so little deformed that 
there is no apparent break in the monotonous relief of the slightly rolling 
country. Some also are not dissected at all. A few large basins, however, 
were entered, and one nearly a hundred miles across proved to carry 
later strata several thousand feet in total thickness. This lies just north 
of Baga Bogdo of the Altai mountain range. It proved to be especially 
prolific and helpful in both structural and paleontological returns. Here 
the party spent several weeks in working out critical points in the 
structure and stratigraphy of the later sediments and in making exten- 
sive collections. 

In this basin a plainly marked unconformity separates a lower series 
of dinosaur-bearing sediments from younger, overlying, mammal-bearing 
strata. In addition to this, the successive warpings and internal deforma- 
tions of the basins were such as to give most suggestive help in further 
subdivision and definition of the successive series of associated strata 
between the mid-Mesozoic peneplanation and Pleistocene conditions. 
These are represented in the accompanying tabulation. Corresponding 
or complementary occurrences at other places are included also under 
their locality designations. There is still too little certainty about their 
inter-relations to attempt closer correlation at the present time. As the 
fossil collections are worked over, and especially as larger collections are 
secured, it will undoubtedly be possible to make much closer correlation 
of these locality finds. 


1923] LATER SEDIMENTS OF CENTRAL MONGOLIA 5 


II—STRATIGRAPHY OF THE LATER SEDIMENTS 


By “‘later sediments” is meant all of the sedimentary formations 
that have accumulated in the deformation basins since the post-Jurassic 
peneplanation. The lower members total at least 2000 feet in thickness 
and carry fossil remains indicating Mesozoic age. 

In certain localities there is above this Mesozoic series a distinct un- 
conformity which separates these formations from overlying ones whose 
fossil content indicates Tertiary age. The total thickness of this upper- 
most series is at least 5000 feet in the Tsagan Nor Basin. 

Although the Tsagan Nor Basin proved to be the most favorable of 
all for structural subdivision, altogether five distinct basins have con- 
tributed to the general summary of stratigraphic relations. The chief 
items of this subdivision and structural relation may be briefly indicated 
as follows: 


1.—Tue Iren Dasasv BasINn 


As indicated in the first communication, this basin is shallow, the 
beds lie practically flat, and the breaks between formations whose fossil 
content indicates wide difference in age are exceedingly obscure. 

The geologic column covering this Iren Dabasu relation is as follows: 


Uplift and Erosion 


Peneplanation 
Miocene Upper barren sands |The Houldjin 
or later Rhinoceros gravels | Formation 
TERTIARY er 
Upper barren Irdin Manha 
Pre-Miocene | sandstones Formation | The “Gobi 
Lophiodont-bed Series”’ of 
!—_————}  Obruchev 
Physical and Faunal Break 
CRETACEOUS Upper barren members Iren Dabasu 
Dinosaur beds Formation 
Great Unconformity . 
Old-rock floor, deformed, metamor- Probably the 
PreE-CRETACEOUS phosed and peneplaned Nan-k’ou 


Series of von 
Richthofen 


6 AMERICAN MUSEUM NOVITATES [No. 77 


The lowest beds furnish abundant dinosaur remains and are cer- 
tainly Mesozoic in age. Not more than 100 feet above the beds furnish- 
ing these dinosaurs lie a series of looser sandy strata and shales, in one 
member of which rhinoceros remains and the first bones of Baluchitherium 
were found. Twenty miles distant, but still within the same general 
structural basin, beds are exposed which furnish remains indicating that 
they are lower and of somewhat greater age than those furnishing 
Baluchitherium and rhinoceros, although they belong also to the Tertiary 
series. 


2.—TuHE TsaGAn Nor BasIn 


North of Baga Bogdo, a very prominent mountain group forming a 
part of the eastward extension of the Altai mountain range, there is a 
very extensive sedimentary basin not less than 75 miles across from north 
to south and of still greater extent east and west. It is in part a down- 
warped and in part a down-faulted basin which has suffered additional 
subsequent internal deformation. These later deformations have fur- 
nished particularly favorable structural conditions for tracing the major 
breaks and exposing nearly the whole of the stratigraphic succession. 

For example, in the midst of the basin itself there are two smaller 
mountain blocks that have suffered uplift and have been involved in the 
distortion since the earliest beds were laid down. One of these is the Mt. 
Uskuk block, and on its borders a distinct unconformity can be traced 
between the beds forming the earliest series and all those of later age. 

At a still later time, after more than 3000 feet of Tertiary 
sediments had been laid down on the eroded surface of the preceding 
series, this interior block suffered another deformation adjustment which 
turned these beds almost up on end along the chief flexure margins. 
Deposition continued, however, in the same basin somewhat farther 
to the south, between Uskuk and the Baga Bogdo border. It appears 
also that Baga Bogdo itself was raised to its maximum height during 
this later time, contributing much of itself to these latest sedimentary 
accumulations. 

Subsequent peneplanation beveled off the upturned edges of the 
strata along the flexure lines and reduced practically the whole basin to a 
monotonous erosion level, remnants of which are still preserved with 
residual gravel cover so smooth that one can drive anywhere on them. 
But, still later, this peneplane was warped enough to cause complete 
rejuvenation of the streams and initiate a new cycle of erosion, which has 
extensively developed ravines, gorges, and typical badland topography 


1923] LATER SEDIMENTS OF CENTRAL MONGOLIA v 


in the most favored localities, thus exposing all of these structural fea- 
tures that have been described. 

The major formational units distinguished in this basin are typically 
developed in localities somewhat separated from each other. Thus a 
certain formation is best exposed in the Ondai Sair, another along the 
Hsanda Gol, and yet another in the hills of Hung Kureh. For convenience 
of reference these locality names are attached to the corresponding forma- 
tional units, the chief characteristics of which are given below: 

(a). Tur OnpaAr Sater FormMation.—This is a series of thin-bedded, 
slightly deformed sandstones and black paper shales. The sandstones 
range in color from white to buff or yellowish, but are rarely red. Al- 
though there are variations from this description, they are of little 
importance, because at every point in this area the rather coarse-grained, 
moderately indurated, cross-bedded and thin-bedded and plain sandstone 
beds are very prominent indeed, and alternating with them in much 
smaller amount are black, carbonaceous shales, some of which are of 
remarkably fine paper-shale quality. An occasional bed of shale has 
such a perfect structure that the lamine split apart almost as true as the 
leaves of a book and have preserved in them remarkably delicate fossil 
forms. 

The formation, therefore, is to be regarded as a sandstone-paper 
shale formation. It was noticed that in this locality the basal members 
are more persistently sandstones and that the uppermost members carry 
a larger proportion of paper shales. It is not at all certain, however, 
that these relations would hold for other areas. 

Formations believed to belong to the same series, though probably 
not to the same horizon, found at several other places, were largely 
reddish sandstones. At one other place, however, nearly a thousand 
miles distant on the road between Iren Dabasu and Mt. Tuerin, a similar 
series of sandstones and paper shales was discovered. Although in that 
case there was not sufficient fossil evidence found to prove their age, they 
are now believed to belong to the same series and perhaps to the Ondai 
Sair horizon. Their petrographic similarity, considering the rarity of 
this type of sediment, is suggestive at least. 

The fossil content of the Ondai Sair is as follows: a dinosaur of 
Camptosaurus type; rib of a large dinosaur; insects, including mosquitoes 
and butterflies; a small form of fish, probably Lycoptera; fragments of 
crustacea; and a thin-walled form of phyllopod crustacean, probably 
Estheria. 

This indicates age at least as old as the Cretaceous, perhaps the 
Lower Cretaceous or Comanchian, rather than the Upper Cretaceous. In 


8 AMERICAN MUSEUM NOVITATES [No. 77 


this locality, at Mt. Uskuk, the series is sharply delimited by an un- 
conformity, above which no fossils of this type are to be found; and it is 
limited also below by a much greater unconformity, beneath which lies 
folded strata with coal beds, conglomerates, and intrusives, all judged 
to be of Jurassic age, as well as many still older formations. 

(b). Tur Hsanpa Gout Formation.—Along the Hsanda Gol, a 
dry, sandy stream-course leading from Mt. Uskuk through the Ondai 
Sair locality southward past Loh to the bottom of the basin of Tsagan 
Nor at the foot of Baga Bogdo, the finest series of Tertiary deposits thus 
far examined in the whole of Mongolia are exposed. They lie uncon- 
formably upon the uneven erosion surface of the Mesozoic Ondai Sair 
formation just described, and are deeply cut into by recent erosion, which 
has developed a fine badland topography for several miles along the 
course of the stream. This stream-course crosses also a strong flexure, 
along which the basin itself has been deformed, and there the earlier 
beds are turned up nearly on edge. It is possible, therefore, to examine 
and measure this formation in detail. 

This whole succession of beds along the Hsanda Gol, from the base 
at Ondai Sair to the topmost beds at Loh, a distance of about fifteen 
miles, is here included in the Hsanda Gol formation. They are prevail- 
ingly yellowish conglomerates and pebbly sands at the base, and vary 
greatly in quality. This type constitutes at least 800 feet of the lower- 
most portion of the formation. The succeeding members above include 
alternating beds of sands, marls, clays and clayey sands of variegated 
colors, chiefly red, yellow, and white, with no apparent system or uni- 
formity of succession. The high colors are strikingly shown in the middle 
portion of the formation, where the badland structure is best developed, 
and where the most clayey and least hardened members occur. The 
uppermost beds are prevailingly sands and clayey sands inclined to be 
reddish in color, some beds of which are fairly well indurated. 

Altogether, a thickness of approximately 3000 feet has been meas- 
ured and estimated from measurements, a large proportion of which 
has been inspected in detail. 

The upper portion of the formation carries fossils. In certain layers 
fossils are abundant with many individuals but not very great range of 
forms. The middle and lower members are largely barren. In the lower 
third, the yellowish conglomerates, no fossils whatever were found. 
Higher in the series an occasional find was made, but only certain of the 
upper members were found to be prolific, and the age determination is 
based on them. The formation is judged to be of Miocene age. Whether 


1923] LATER SEDIMENTS OF CENTRAL MONGOLIA 9 


the whole thickness of 3000 feet to the very bottom is also Miocene is, of 
course, not known; but in the failure to find any physical break or any 
faunal evidence on which to base any other determination, we have 
chosen to regard the whole succession as a unit. 

Thousands of fossil specimens were secured in this ground, particu- 
larly in the vicinity of Loh, ten miles downstream from Ondai Sair, 
and in the vicinity of the’so-called Grand Gorge, a corresponding erosion 
exposure ten miles farther to the west. The forms recognized in the field, 
taken from these beds, are as follows: 

Baluchitherium, a fine skull nearly five feet long, and other rhino- 
cerids; rodents by the hundreds; artiodactyls; insectivores; and 
carnivores. 

(c.) Tue Hune Kuren Formation.—In the very bottom of the 
basin of Tsagan Nor, between the lake itself and the foot of Baga Bogdo, 
still younger beds overlie those of the Hsanda Gol. They constitute a 
series of whitish and yellowish sands and clayey sands at the base, which 
are overlain by coarser, gravelly sands and conglomerates of great 
variety and thickness. These beds are not as well exposed as are those 
of the Hsanda Gol; and, on this account, the complete succession cannot 
be so well determined. It has been possible, however, to measure more 
than 1000 feet, and a reasonable estimate would be approximately 2000 
feet in the type locality, the hills of Hung Kureh. 

These hills stand above the average peneplane level and formed 
originally a low monadnock group which has been redissected by later 
erosion. On the extreme margin next to the mountain of Baga Bogdo 
the upper beds become coarsely conglomeratic. An exposure of 800 feet 
in total thickness occurs in the hills immediately facing the central 
canyons of the Baga Bogdo front. 

As a whole, this series of beds which we regard as a single formation 
is almost barren. An occasional bed, however, does carry fossil remains. 
This is particularly true of the lowermost portion of the formation, 
within 200 or 300 feet of the base, where yellow, iron-stained sands and 
white sand beds are the prevailing type. There is apparently no real 
break in the succession, the conglomerates in this case marking the in- 
fluence of the adjacent rising mountain mass of Baga Bogdo, which 
furnished alluvial fan type of materials to the adjacent basin sediments. 
These uppermost beds, therefore, may be appropriately referred to as 
alluvial fan conglomerates. They doubtless interlock with the finer sedi- 
ments washed down from the opposite direction, derived from the erosion 
of simpler sediments within the deformed basin itself. It is this inter- 
mixture that constitutes the Hung Kureh formation. 


10 AMERICAN MUSEUM NOVITATES [No. 77 


Although the beds of this formation are not so prolific in fossils as 
are those of the Hsanda Gol, they are, nevertheless, sufficient to indicate 
a Pliocene age and include a wide variety of forms. We have, therefore, 
set off this member as the Hung Kureh formation, limited at the base 
by a conformable contact with the Miocene Hsanda Gol formation, and 
at the top by the present erosion surface. 

In this region no deposits of consequence’ are preserved above the 
Hung Kureh formation. The Pleistocene, or at least the period follow- 
ing the deposition of the Hung Kureh formation, was a period of erosion; 
and no sediments are preserved except very insignificant accumulations 
of comparatively recent time along the stream courses, where the dis- 
appearing streams drop their load or where newer fans are reaching out 
from the mountain blocks. 

The fossil content of the Hung Kureh, as identified in the field, is as 
follows: afew fragments of horse; afew bones and egg-shell fragments 
of a very large bird, probably Struthiolithus; a large cervid; mastodon; 
and rhinoceros. 


3.—TuHE ULAN Nor AND NEIGHBORING BASINS 


Eastward from Tsagan Nor for several hundred miles there is no 
mountain barrier. The Altai mountain range continues along on the 
south side, and at first glance the whole region to the north appears to 
be a single basin. If one follows it, however, with enough care to unravel 
its structural detail, it is found that the basin sediments are not continu- 
ous but that the ancient rock floor comes to the surface at several places, 
separating the region into individual, smaller basins. Thus it happens 
that on the north side of Artsa Bogdo, a hundred miles east of Tsagan 
Nor, there is another large sedimentary basin; and, one hundred miles 
still farther east, yet another. The dividing barriers are scarcely more 
conspicuous than the monotonous basin surface itself, but the sediments 
are not continuous. As would be expected from such circumstances, the 
sediments that do occur in adjacent basins are not necessarily of the same 
age or type. In the shallower basins there is good chance of finding the 
earlier formations, representing Mesozoic strata, and in the deeper ones 
one is almost certain to find both Mesozoic and Tertiary. 

(a). Tue AsnitE FormMation.—Such conditions are found in 
both of the basins to the east. In that one lying to the north of Artsa 
Bogdo, at the locality known as Ashile, a series of reddish and buff, 
fairly well indurated sands and sandstones were found. Certain members 
of this series carry dinosaur remains, indicating Mesozoic age. This is 
the Ashile formation of the accompanying table. 


1923] LATER SEDIMENTS OF CENTRAL MONGOLIA 11 


(b) THe Dsa-pocu-ra Formation.—In the vicinity of Ulan Nor, 
at the locality Dja-doch-ta, along the Kwei-wa-ting Trail, a series of 
somewhat similar red sands and clayey sands were found, which again 
carried a rare species of dinosaur and a variety of undetermined fossil 
material. One entirely new form has been described recently by Granger 
and Gregory under the name Protoceratops andrewsi. Because of the 
strikingly different fossil content, these beds are also given a separate 
designation, and because of the very primitive character of the forms it is 
believed that they may well be lower and older than even those of Ondai 
Sair. This unit is, therefore, given the locality name, the Dja-doch-ta 
formation. 

In the trip across the desert basin south of the Altais, between Artsa 
Bogdo and the Gurban Saikhan, a very similar series of reddish sand- 
stones was encountered, and on the northern margin of the Gurban 
Saikhan uplift itself at least 600 feet of upturned conglomeratic and coarse 
cross-bedded sands are well exposed. No fossils were found at the 
Gurban Saikhan locality; and, on this account, there is no way to 
determine just how these beds may be related to others just described. 
They may deserve a descriptive name, such as the Gurban Saikhan con- 
glomerates; but there is no way of placing them at any particular horizon 
in the column. Their physical character and general structural relations 
indicate close relationship to the Ashile and Dja-doch-ta, and they 
doubtless belong to the same group of Mesozoic strata referred to in this 
tabulation as the Shamo Series. . 

Most of the beds of this whole complicated basin region are barren, 
but here and there certain ones are prolific and tell essentially the same 
story. On the basis of these few finds these formations are all regarded 
as Mesozoic in age; and because of the difference between this material, 
both as to fossil content and rock quality, they are judged to be the 
equivalent of neither the Ondai Sair beds encountered to the west, in 
the Tsagan Nor Basin, nor the Iren Dabasu, farther to the east. The 
Dja-doch-ta beds at least are probably Lower Cretaceous in age. 

In the Ashile formation were found a sauropod of very large size, 
and a dinosaur of Camptosaurus type. 

In the Dja-doch-ta formation were found a few fragments of a bird, 
and an entirely new form of dinosaur, since described under the name 
Protoceratops andrewsi. 

There are doubtless higher beds, some of which must be of Tertiary 
age, in these easterly basins; but at the places crossed by the Expedition 
none of these is well exposed and no collections could be made from them. 


12 AMERICAN MUSEUM NOVITATES [No. 77 


As the Expedition passed northward, keeping on the east side of Ulan 
Nor, sediments continued for more than seventy-five miles; but in this 
whole distance, which was covered in less than a day, no specific determina- 
tions could be made beyond the fact that the underlying rocks represent 
the later sedimentary series and must correspond closely with some of 
those already seen in more favorable localities. 


4.—ARDYN Ono 


South of Sair Usu, on the main Kalgan-Uliasutai Trail, several 
basins carry later sediments. The general nature of the formations can 
generally be seen, but only in two or three localities are there extensive 
exposures of these later sediments, where exploration would be certain 
to furnish good fossil evidence. Both the earlier, Mesozoic, and the 
later, Tertiary, series are represented and a few new forms were found 
whose significance is not yet determined. : 

One of the better localities is at Ardyn Obo. Here a great escarp- 
ment, surmounted by that inevitable guidepost of the desert, an Obo, 
stands 300 feet above the general level of the plain over which the trail 
passes; and for many miles the edges of the flat-lying strata are exposed. 
They are rather loose and slightly indurated sandstones and clayey sands 
of considerable variety of texture, quality, and color. In the upper 
members of this series of beds fossils were secured, the chief items of 
which are several fine specimens of rhinoceros and numerous fragments of 
turtles. 

This content is judged to indicate mid-Tertiary age, not very differ- 
ent from that of the Miocene beds of Hsanda Gol, or the Houldjin forma- 
tion of Iren Dabasu. It is not possible with the present evidence, how- 
ever, to correlate the Ardyn Obo beds with either of the others, and it 
may very well be, indeed, that they belong to an entirely different horizon 
in the same series. It seems best, therefore, to preserve the locality name 
in the ARDYN OBO FORMATION. 


5.—SHARA Murun 


A hundred miles farther south another basin carries fossiliferous 
strata. The sedimentary area is very large and the best exposures occur 
along the borders of a 200-foot escarpment. At this place were found 
titanothere remains very like those found at Irdin Manha early in the 
summer, in the Iren Dabasu area. These seem to be, therefore, early 
Tertiary in age, perhaps as early as Eocene. 


1923] LATER SEDIMENTS OF CENTRAL MONGOLIA 13 


Still farther south, on the very borders of Kalgan, a great series of 
later sedimentary strata furnished a few reptile fragments. Too little 
work was done at these places to warrant further description or classifica- 
tion of beds. They deserve careful inspection. 


SUMMARY 


It appears from this listing of the formational units distinguished by 
locality and structural features that there are at least four formations of 
locality significance of late Mesozoic age, presumably Upper and Lower 
Cretaceous, and four formations belonging to the Upper Tertiary, 
including beds of Miocene and Pliocene age. The early Tertiary is much 
less fully represented, or else is represented by formations not yet fur- 
nishing adequate fossil criteria. Only the Irdin Manha formation, in the 
Iren Dabasu region, and the beds of Shara Murun, one of the basins on the 
Uliasutai Trail, have been placed earlier than the Miocene on fossil 
evidence. It may be, of course, that the lower barren beds of the 
Hsanda Gol are pre-Miocene also; but for this there is no direct evidence. 

This is an additional reason for grouping these scattered formational 
units into definite series. There are no used or published terms that have 
suitable limitations for this purpose. Obruchev’s ‘‘Gobi Series’? was 
used by him to include all sorts of later sedimentary beds, without age 
distinction. Von Richthofen’s term “‘Khan-Khai Beds” has been used 
in the same way for all of the later sediments, without more intimate 
discrimination. 

Perhaps it is not advisable to disturb usages of this sort, which have 
at least the merit of being fixed. But, since these terms were not origi- 
nally based on fossil evidence and have been used only as a convenient 
field term for general purposes, it may be permissible to continue to use 
them as good general terms for all of the later sediments as defined in this 
communication, and, when referring to smaller groups of formations or 
specific portions of the later sedimentary column, to make use of more 
accurately defined and delimited terms. 

If this is done, the preference, perhaps, should be given to the term 
Khan-Khai, which is much the older one. But the error of interpreta- 
tion involved in the name makes it less appropriate than Obruchev’s 
Gobi Series. Von Richthofen thought that the name Khan Khai 
meant deposits of an evaporating sea and used it with that significance. 
None of these sediments, however, have such an origin. All are 
strictly continental in origin. On this account Obruchev’s Gobi Series 


14 AMERICAN MUSEUM NOVITATES [No. 77 


is a much better term. Gobi means desert basin, and Gobi Series 
therefore is a very suitable name for the desert basin deposits. 

It would be particularly appropriate if these older terms could be 
used for the larger divisions, corresponding to systems, leaving all other 
terms to be used for smaller groups and individual formations. In any 
case, there will be little cause for confusion if the significance of each is 
kept clear. With this object in view, the following terms are proposed for 
the smaller groups: 

1.—The Tsagan Nor Series (Upper Tertiary). 

2.—An unnamed series (Lower Tertiary). 

3.—The Shamo Series (Upper Mesozoic). 

The Tsagan Nor Series is intended to include formations of Pliocene 
and Miocene age, or the Upper Tertiary. To it belong the units already 
described as the Hung Kureh, Hsanda Gol, Ardyn Obo and Houldjin 
formations. 

Another series ought to include the Lower Tertiary, or the pre- 
Miocene Tertiary. Thus far, the only formations classified in this series 
are the Irdin Manha and the beds of another basin, the Shara Murun, 
farther west. Others, however, are certain to be added as the work pro- 
ceeds. Perhaps a better name than has yet been proposed for this 
series will be suggested by the coming season’s work. 

The Shamo Series is pre-Tertiary. It is intended to include the 
Cretaceous beds overlying the great unconformity above the folded and 
eroded Jurassic. The units distinguished in the field are the Iren Dabasu, 
Ondai Sair, Ashile, and Dja-doch-ta formations. 

The geologic column showing these groupings and relations is 
given in the table, p. 15. 

It is entirely likely that some of these separately designated forma- 
tions overlap in time, so that the sum of all the thicknesses given is not 
an accurate statement of the total column. Making all due allowance, 
however, for such probability, it appears that no less than 6000 to 8000 
feet of later sedimentary strata are accounted for in the complex series 
of later sediments lying above the great post-Jurassic unconformity in 
the Gobi region of Mongolia. 

All of the formations named are distinguished by characteristic fossil 
content, but great thicknesses of strata that are barren are included with 
them. The region has proved to carry not only a prolific fossil fauna but 
one that is also of unusual scientific interest and significance; and the 
field is capable of furnishing immensely greater returns in this direction. 
All are of strictly continental type. No evidence of marine invasion was 
found anywhere. 


1923] LATER SEDIMENTS OF CENTRAL MONGOLIA 


Quaternary 


Uplift and Erosion 


Peneplanation 


| a 
| ‘Pliocene 


The 
[Upper | 
‘Tertiary | 


CENOZzOIC 


Nor 


Miocene 


TERTIARY 


Tsagan (Baluchitherium and 


Series Ardyn Obo Formation 


‘Hung Kureh Formation | 
(Mastodon and Cervus) 


| 


‘Hsanda Gol Formation 


| 


rodents) 


| (Rhinoceros) 
el: Formation 

(Rhinoceros) and 
| Baluchitherium 


Oligocene 
Eocene 


‘Lower 
Tertiary 


Irdin Manha Formation 
(Lophiodont beds) 
| Shara Murun Formation 


20007 


3000/ 


500’ 


50’+ 


50’+ 


| (Titanothere beds) 


’ Unconformity 


n 
Upper 
Cretaceous 


The Shamo Series 
Lower 
Cretaceous 


Larr Mesozoic 


| Ashile Formation 


Tren Dabasu Formation 
(Dinosaurs) 
Ondai Sair Formation 
(Dinosaurs and insects) 


(Dinosaurs) 
Dja-doch-ta Formation 
(Dinosaurs) 


150’= 
500+ 
1000’ 


300’ + 


The ‘‘Khan Khai Beds” of von Richthofen 


Great Unconformity 


The “Gobi Series” of Obruchev 


Jurassic 


EARLY 
Mesozoic 


/ A great thickness of conglomerates, 
/ sandstones and volcanics 


The Expedition has found that later sediments of this general range 


are abundant in the Gobi region. 


They all occur in basin-like depres- 


sions, but not all are‘well enough exposed by later erosion to be open to 


examination. 


It has been our good fortune, howeyer, to find excellent 
exposures of certain parts of the series. 


The determined formations are so well distributed through the 
Upper Mesozoic and Tertiary column, and are related to the general 


16 AMERICAN MUSEUM NOVITATES [No. 77 


sedimentation and deformation history of the Gobi region in such a way 
as to promise ultimately a practically complete and continuous succes- 
sion from Lower Cretaceous to Pleistocene time. Even the Pleistocene, 
which is an erosion blank in the Northern Gobi, ought to be represented 
in certain adjacent regions farther south and east, where deposition 
continued while these northerly basins were being peneplaned and re- 
dissected. 


AMERICAN MUSEUM NOVITATES 


Published by 


‘Number 78 Tue AMERICAN MusEUM OF NaTURAL History May ALah 1923 
New York City ‘ 


56.9,72B (51.7) 
BALUCHITHERIUM GRANGERI, A GIANT HORNLESS 
RHINOCEROS FROM MONGOLIA 


By Henry FAIRFIELD OsBorRN! 


In previous communications on the rhinoceroses (Rhinoceros Con- 
tributions 1 to 11), Osborn separated six distinct phyla or subfamilies. 
The remarkable discoveries by Clive Forster Cooper in Baluchistan, by 
A. Borissiak in north Turkestan, and by Walter Granger of the Third 
Asiatic Expedition in southeastern and -central Mongolia, indicate the 
existence of a seventh subfamily which we may term Baluchitheriine, 
if the generic name proves valid. At present our knowledge rests on the 
following materials: 


Buetrt Hits, Chur-lando, Baluchistan. Cooper Collection, British Museum. 
Paraceratherium bugtiense Cooper, December 1911. Fairly complete skulls and 
lower jaws of about the size of a large rhinoceros, simple aceratherine 

molars, abnormal lower incisors. 

Thaumastotherium osborni Cooper, October 1913, changed to Baluchitherium 
osborni Cooper, November 1913. Fragmentary skeletal remains found 
in close proximity to Paraceratherium, including neck vertebre, foot and 
limb bones of elephantine size. 

TuRGAI, a province of north Turkestan. Discoveries by A. Borissiak, published 
1915-1918. 

Indricotherium asiaticum Borissiak, 1916. Teeth, skull, and skeletal remains, 
occurring in situ and resembling both Paraceratherium and Baluchitheriwm. 

Epiaceratherium turgaicum Borissiak, 1918.2 

Lou, central Mongolia, Third Asiatic Expedition Collection, 1922. Associated skull 
and skeletal remains similar in size to the type of Baluchitherium osborni. 

Baluchitherium grangeri, new species. Type, nearly complete skull and jaws 
(Amer. Mus. 18650) associated with parts of vertebra and of limb bones, as 
described in the present bulletin. 

IrEN Dazasu, southeastern Mongolia. Baluchitheriwm ref., caleaneum and other 
fragments of skeleton. 


DISCOVERY OF SKULL AND SKELETON IN Moneo tia, 1922 


The party left Kalgan on April 21, 1922. (1) The first Baluchi- 
thercum material was discovered by Dr. Charles P. Berkey on the journey 


1Contribution No. 8, Asiatic Expeditions of The American Museum of Natural History. Twelfth 
Contribution on the Evolution of the Rhinoceroses. See Bibliography of H. F. Osborn, 1916, p. 21. 
The genus Epiaceratherium was founded on the species Epiaceratherium bolcense. See Abel, 
ence ure cckenaey iiber die paliogenen Rhinocerotiden Europas,’ Abh. k. k. geol. Reichsanst., 
XX, Heft 3, p. 20. 


2 AMERICAN MUSEUM NOVITATES [No. 78 


north towards Urga, near Iren Dabasu; this consists of a large caleaneum 
and other tarsal or carpal bones, more or less fragmentary, bearing the 
' field number ‘Third Asiatic Exped. 37,” to which has been assigned 
the catalogue number ‘‘Amer. Mus. 18651.” (2) The second and most 
important find was made on August 5, 1922, near Loh, in the Tsagan 
Nor Basin, Hsanda Gol Beds; this includes a skull, portions of the jaws, 
and the distal end of a humerus, bearing the field number “Third 


of 
TURGAL 7 MANCHURIA 
1A 


2 ere. 


/ 
CHINESE 


TURKESTAN 


AFGHANISTAN 


Fig. 1. Map of central and southwestern Asia showing the type localities of (1) 
Baluchitherium osborni type, eastern Baluchistan; (2) Indricotherium asiaticum type, 
near Turgai, northern Turkestan; (3) Baluchitherium grangeri ref., near Iren Dabasu, 
southeastern Mongolia; (4) Baluchitherium grangeri type, near Loh, central Mongolia. 


Asiatic Exped. 90,” to which has been assigned the catalogue numbers 
“Amer. Mus. 18650, 18652.”’ The skull and mandible were about fifty 
feet apart and probably belong to the same individual. This specimen is 
made the type of Baluchitherium grangeri described below. (8) The 


1923] BALUCHITHERIUM GRANGERI 3 


humerus was found a quarter of a mile distant from the type and may or 
may not belong to the same individual, but probably to another indi- 
vidual. The American Museum catalogue references to the three speci- 
mens are as follows: 


Amer. Mus. 18651 _ Amer. Mus. 18650 Amer. Mus. 18652 
Tuirp AsraTic ExprEpb. Tuarrp Asratic Expep. Turrp AsratTic EXpEp. 
Calcaneum and other frag- Type skull and part of Distal end of humerus 

ments of skeleton mandible 


Fig. 2. Type skull of Baluchitheriwm grangeri as restored up to March 16, 1923. 
A portion of the jaw and of the right side of the skull is now being added for the skull 
and jaw cast. One-twelfth natural size. 


Grotocic Acre.—The Baluchitherium caleaneum (Amer. Mus. 
18651, Fig. 8 of this article) is from the Houldjin formation near Iren 
Dabasu, an upper member of the Gobi series containing at the base 
rhinoceros gravels five feet in thickness; the Houldjin beds were corre- 
lated by Granger and Berkey (Amer. Mus. Novitates No. 42, August 7, 
1922, p. 4) as of Miocene or more recent age, more recent than the 
underlying Irdin Manha formation, which is regarded as of Oligocene age. 
Of the same age is the equivalent formation in the Tsagan Nor district 
near Loh, several hundred miles to the west, from which the type of 
Baluchitherium grangeri was obtained. This formation has been named 
the Hsanda Gol formation. Borissiak placed the Indricotherium asiaticum 
zone as far down as the Middle Oligocene; it is regarded by Granger and 
Berkey as more probably of middle or late Miocene age, since it contains 
(op. cit., p. 4) a rhinocerid, a large carnivore, an artiodactyl of the size 
of a Virginia deer, and a tortoise of large size, in addition to the enormous 
perissodactyl known as Baluchitherium. 


4 AMERICAN MUSEUM NOVITATES [No. 78 


Baluchitherium grangeri, new species 


This remarkable type specimen, named in honor of Walter Granger, 
head of the paleeontologic division of the Third Asiatic Expedition under 
Roy Chapman Andrews, was found as indicated in the map (Fig. 1, 4) 
and described above. 


DESCRIPTION OF THE TYPE SKULL 


The skull of B. grangeri, while of enormous size, as shown by com- 
parison (Fig. 3) with that of the type of Acerathertum incisivum Kaup and 
the skull of the large white rhinoceros, is relatively primitive in structure; 
the central portion of the forehead is prominently arched or convex, and 
the bones preserved (Fig. 3A) show that there is absolutely no indication 
of either a frontal or nasal horn. The proportions are decidedly dolicho- 
cephalic, as indicated by the following chief measurements: 


Fig. 3. Comparative view of the skulls of: A, Baluchitherium grangeri type; 
B, Aceratherium incisivum type, from the Darmstadt collection; C, Ceratotherium 
simum (Amer. Mus. 1142). One-sixteenth natural size. 


Length of Aceratherium incisivwm type skull, from premaxil- 

MATIC HONCODE VIES rece oy Sure ok tea: okt weet eek tone ener Tae 694 mm. 
Length of Ceratotherium simum skull, from premaxillaries to 

COMICS Es rane alsitiars oh elo ors aie bap AO ee oO Ce Eas 728 


Sou” 


Fig. 4. A, Top view of skull of Baluchitherium grangeri type, to show the origi- 
nal and restored portions; the original portion including parts of the frontals, 
the middle portion of the nasals and the tip of the nasals. B, Side view of the skull 
and jaws of the same; part of the fragments of the jaw, reversed from the other side; 
anterior portion of the jaw restored from Paraceratherium bugtiense, after Forster 
Cooper, PI. x, fig. 1, highly conjectural. C, Palatal view of same showing the simple 
and primitive characters of the premolar teeth. All figures one-twelfth natural size. 


5 


6 AMERICAN MUSEUM NOVITATES [No. 78 


Length over all of Baluchitherium grangeri type, from front of 


premaxillary symphysis to back of occipital condyles...... 1286 (4 ft. 3 in.) 
Greatest zygomatic width, of B. grangeri type, anterior to 
slenord stoned: era oes at eee Gee Raa ee he see 614 (2 ft. % in.) 


Consequently, the zygomatic-cephalic index of the type of Baluchi- 
therium grangeri is 47.7; the zygomatic-cephalic index of the much 
smaller Aceratherium incisivum type skull is 54.3; the zygomatic- 
cephalic index of the massive Ceratotherium simum leu is 46.7. 

The proportions, length, and breadth of the B. grangeri type skull 
are extremely close to those of the Cenopus (Aceratherium) occidentalis 
skull of the American Oligocene, namely: 

Cxnopus occidentalis, zygomatic-cephalic index =4 
Baluchitherium grangeri, zygomatic-cephalic index =4 
Rhinoceros indicus, zygomatic-cephalic index =H 
Ceratotherium simum, zygomatic-cephalic index =4 
Aceratherium incisivum type, zygomatic-cephalic index = 54. 3 

This is a primitive Eocene and Lower Oligocene form of skull 
grown large. In profile it nearly resembles the primitive skulls of several 
Eocene perissodactyls in several 
independent lines of perissodactyl 
descent, also of several Lower Oligo- 
cene aceratheres. The reasons are 
obvious: (1) the greatly enlarged 
second incisor, I?, functions as a de- 
fensive tusk or canine; (2) the ani- 
mal was completely protected by 
this as well as by its rapidly increas- 

Fig. 5. Occipital view of the jing height, so that it probably tow- 
skull of Baluchitherium grangeri type gored over all the mammals of the 
(Amer. Mus. 18650), to be compared : : . i ‘ 
with occipital views of the Oligocene period; with its very thick rhinoc- 
aceratheres. One-twelfth natural size. erotic skin, it was safe from attack; 

(3) with an entire absence of horns 
there is no secondary cranial modification to support the powerful nasal 
and frontal horns, as in the ceratorhines and African rhinoceroses. 

The amazing size and power of the skull is none the less manifested 
in the condyles, which are enormous as compared with those of C. sxomum 
and as large as in any species of proboscidean. The occipital condyles 
of Baluchitherium grangeri exceed in both diameters the anterior facets 
of the atlas in Forster Cooper’s specimen; it follows that the skull of B. 
grangeri belongs to an animal with even larger and more massive cervicals 
than those described and figured by Forster Cooper in B. osborni. The 


1923] BALUCHITHERIUM GRANGERI . 7 


seven superior grinding teeth, P'-M?, are of the brachyodont, short- 
crowned, browsing type and parallel, in the very retarded development 
of the superior crests or metalophs of the premolar teeth, those in the 
American aceratheres described by Osborn as Aceratherium platycephalum 
(see Osborn, 1898, p. 140, Pl. x11, figs. 9, 10). 


CoMPARATIVE MEASUREMENTS IN MILLIMETERS 


s 5 Sache s 
= 2 Ba] 34 
Ss 3 = = Q S & 
Sim rea Be a 
= TN ee aoa aes 
isa) ~ fo) Ns 
SKULL, length condyles to symphysis ae Matias Bee 1286. 
Zygomatic width Ee APES ha aaa 614 
Condylar width a ie aa ths hehe a 306 
Houmervs, length between articular surfaces 840 930 roe: ae 
Proximal width 376(?) dere a aa gees 
Distal width 230 ine Dae 287 
Una eee 1200 eden ig Mig 
Frmor, length between articular surfaces 1200 1230 
Proximal width 300(?) 
Distal width 190 
Tri, length between articular surfaces 790 860 
Proximal width 250 
FIBULA : Sauk 
CALCANEUM, max. length Sante BAY 324 
TUBER CALCIS max. width Sok antal ee Memeeae 148 
ASTRAGALUS en 
Max. height 144est. dy 
Max. width 195 212 
Merararsat IIT, max. length Beas 510 
Meracarpat III, max. length 440 545 
Meracarpat IV, max. length 375 Y 
Ris conte 660 
ATLas (1st CERVICAL) Luatead Ree 
Max. width 475 
Width of ant. condylar surfaces 280 
Extreme length of wing 240 
3RD CERVICAL Renee 
Extreme width 400 
Extreme length 420 
6TH CERVICAL DOP 
Max. width 460 


Max. length 300 ie er teeta 


8 AMERICAN MUSEUM NOVITATES [No. 78: 


Speciric DenTAL CHARACTERS OF Baluchitherium grangeri.—Pre-- 
molar crests retarded in development: First premolar a small simple: 
tooth. Second premolar, small, protoloph distinct; also with large 
postero-internal tetartocone and rudimentary metaloph. Third pre- 


Fig. 7 


Fig. 6. Occipital view of: A, Baluchitherium grangeri type; B, Ceratotherium 
simum, an unusually large skull in the American Museum (Amer. Mus. 1142). 
Both figures one-sixteenth natural size. 

The Baluchitheriwm occiput is very similar to that of the Oligocene aceratheres 
of America, as figured by Osborn, 1898, Pl. x1x, namely, A. trigonodum, copei, occi- 
dentale, tridactylum. 

Fig. 7. Lower jaws of: A, Baluchitherium grangeri type (Amer. Mus. 18650), 
as compared with B, Paraceratheriwm bugtiense, after Forster Cooper. The dotted 
restoration of the anterior portion of the Baluchitheriwm grangeri jaw is conjectural. 
Both figures one-sixteenth natural size. 


molar with protoloph forming a long hook-like crest continuous with 
tetartocone; metaloph slender, not connected with tetartocone. Fourth 
premolar still larger, with very prominent metaloph curving (hook-like) 
into internal tetartocone; within this a slender metaloph. The premolar 
transformation in B. granger7is shown in the type specimen (Fig. 4C). The 
fourth premolar is relatively wider than in J. aszaticum. 

This kind of premolar transformation is different from that observed 
in the American Oligocene aceratheres, A. occzdentale and D. tridactylum; 
it approaches A. platycephalum (compare Osborn, 1898, Pl. x11, figs. 9, 
10), but is more advanced. As compared with the premolar transforma- 
tion in the European aceratheres (compare Osborn, 1900, p. 242, Fig. 8, 
“Evolution of the grinding teeth in the Aceratheriine’’), it resembles 
most closely the condition observed in P*, P4in Aceratherium filholi type, 
Lower Oligocene, Phosphorites, Quercy. 


1923] BALUCHITHERIUM GRANGERI 9 


Fig. 8. Foot bones of: B, Baluchitherium osborni, and A, Baluchitherium 
grangeri, as compared with the corresponding foot bones of C, Ceratotherium simum. 
A, External, and A1, anterior views of the left caleaneum (Amer. Mus. 18651) found 
near Iren Dabasu. B, Internal, and B1, anterior views of astragalus, also B2, meta- 
tarsal ITV, drawn after casts presented by the British Museum; cast Amer. Mus. 
5210 (?metacarpal), and 5209 (astragalus). Corresponding astragalus and calcaneum 
united of Ceratotheriwm simum (Amer. Mus. 51862). C, External view; C1, anterior 
view; C2, anterior view of right fore foot. All figures one-eighth natural size. 


SeconD ResToRATION OF Baluchitherium grangeri' 


The skeleton of Baluchitherium has been restored through careful 
comparison of four sets of measurements (p. 7), two given by Forster 
Cooper (1923) and Borissiak (1918), and two from specimens in the 
American Museum, one from Iren Dabasu, and one the type from Loh. 
These measurements are astonishingly uniform and indicate that the 


1This restoration has been prepared under the direction of the writer with the codperation of Dr. 
W. K. Gregory and Mrs. L. M. Sterling. The first restoration was drawn chiefly under the direction of 
ne W.D. Matthew by Mrs. E. Rungius Fulda and has been published in the magazine Asia, April 
1923. 


. OL 
‘O[BOS OUTS 04} OF UMUIp ‘(suiooUN “z) SOLBDOUTY UBIPUT oY} Jo ozIs [[NJ 07 UMEIP asoyy Y}IM 


poredurod ‘smojzu0s Apog “pourquios zu10gso “g pus aabunib wnrseynyonog JO UoJaJOYs JO UOT}eIOJSeI [eIVAVG “6 ‘SIA 


2 


) 
' 
' 
1 
! 
1 
! 
t 
| 
I 
1 


wee ee HK we 


1923] BALUCHITHERIUM GRANGERI 11 


specific forms known as osborni, asiaticum, and grangeri are closely 
similar in size, the differences indicated in the accompanying compara- 
tive measurements being partly attributable to age, or growth, and to sex, 
male or female. The proportions of the skull are fortunately exactly 
determinable from the superb American Museum type specimen No. 
18650. The skull belongs to an animal larger even than the individual 
which possessed the gigantic neck vertebre figured by Cooper; yet the 
skull, gigantic as it is, appears proportionately small. The length of 
the neck has been ascertained by a comparison of cervical 1 (atlas), 
cervical 3 and cervical 6, figured and measured by Cooper; the propor- 
tions of the neck and head are very similar to those in a giant specimen 
of the horse from Kansas preserved in the American Museum. 

The height of the fore limb has been determined by the humerus, 
ulna, carpals and metapodials, figured and measured by Cooper, but it is 
apparent that this fore limb, like the neck, belongs to an individual in- 
ferior in size to the American Museum skull. The scapula is restored in 
dotted lines by careful comparison of the scapule of C. somum, R. 
unicornis, and especially of the American Oligocene aceratheres A. occiden- 
tale and Cenopus tridactylus. The Baluchitherium scapula will probably 
be found to have: (1) a large glenoid region for the massive head of the 
humerus; (2) a strong coracoid for attachment of the biceps; (3) a 
relatively high, narrow blade adapted to the elongate muscles of the 
elongate limbs; (4) inasmuch as the humerus is very distinctly rhino- 
cerotoid in form (fide Cooper), the scapula must have the form of cur- 
sorial rhinoceroses; (5) the posterior border is more projecting than in 
C. simum. 

‘The hind limb may be estimated from comparison of the measure- 
ments given by Cooper and Borissiak with the proportions of the limbs 
and calcaneum preserved in the American Museum collection. The pel- 
vis is given a vertical rather than a horizontal character in adaptation 
to weight; the pelvis and the hind limbs, and especially the abbreviate 
proportions of the tibia as compared with the femur, correspond with the 
graviportal quadrupedal type of Osborn, the ratio between tibia and 
femur being: Tibia, 790100-~femur, 1200=66. The corresponding 
ratios in the following quadrupeds are: 


TrBI0- 

TIBIA FEMUR FEMORAL 

RatTIo 
Indricotherium asiaticum Bor. 860 mm. 1230 mm. 70 
Baluchitherium osborni F. C. 790 1200 66 
White Rhinoceros (C. simwm) 343 523 66 
Indian Elephant (E. indicus) 618 1020 60 


Loxodonta africana (“ Jumbo’’) 755 1050 72 


12 AMERICAN MUSEUM NOVITATES [No. 78 


Consequently, we infer that the locomotion in the hind limb of Baluchi- 
therium was of Osborn’s graviportal type, namely, that of the rhinoceroses 
and elephants. The closest correspondence in the actual length of the 
femur is with Loxodonta africana (in the skeleton of ‘‘Jumbo’’) which 
measures 1050 mm.; the hind limb of L. africana is shorter as a whole 
than that of Baluchitherium, but the tibia is relatively longer in 
Loxodonta. 

The radio-humeral ratios, that is, the proportion between length of 
radius and that of the humerus, are also on the whole rhinocerotine rather 
than elephantine. The corresponding ratios in the following quadrupeds 
are: 


Rapio- 
Rapius HuUMERUS HUMERAL 

' Ratio 

Indricotherium asiaticum 976 mm. 930 mm. 105 
Indian Rhinoceros (R. unicornis) 385 385 100 
White Rhinoceros (C. simwm) 364 381 96 
Indian Elephant (EH. indicus) 685 810 85 
African Elephant (Loxodonta africana) 870 1000 87 


These measurements show that, while the actual combined length 
of the radius and humerus in [ndricotherium equals 1906 mm. and that in 
Loxodonta africana equals 1870 mm., the radius is relatively longer in 
Indricotherium (ratio, 105) and resembles the ratio of Rhinoceros uni- 
cornis (100) more closely than the low radial ratios of the Indian elephant 
(85) and of the African elephant (87). The radio-humerus is as elevated in 
Baluchitherium as in the elephants, while the metapodials are much more 
elevated, which accounts for the greater height at the shoulder in 


Baluchitherium. 
Meracarpo- METATARSO- 


HUMERAL FEMORAL 
Ratio Ratio 

Baluchitherium osborni 52 

Indricotherium asiaticum 59 41 
Rhinoceros unicornis 48 36 
Ceratotheriwm simum 45 31 
Elephas indicus 2 13 
Loxodonta africana 20 14 


The above ratios show that the baluchithere podials (41) are rela- 
tively higher than in the two rhinoceroses (86 and 31); they are very 
much higher than in the two elephants (13 and 14). In other words, the 
metapodials are relatively three times as tall in the baluchitheres as 
they are in the elephants; this accounts for the much greater height of 
the baluchitheres at the shoulder and at the hip. 


1923] BALUCHITHERIUM GRANGERI 13 


The astragalus and caleaneum exceed in size and in length those of the 
Pleistocene mammoths, e.g., Hlephas jeffersoni type, Amer. Mus. 9950, 
but the structure of both the astragalus and caleaneum is distinctively 
rhinocerotoid (teste Forster Cooper, 1923). 

RELATIVE Bopy AND Limp Proportions.—The height of the 
B. grangeri dorsal spines is nearer 13 feet than 12 feet, as originally esti- 
mated. The fore limbs, the trunk and the hind limbs are all within a 
square 4270 mm. long by 3965 mm. high; the height-length ratio is 92; 
as compared with a giant horse (Hquus caballus) in which the same ratio 
is 94; or with the Indian rhinoceros (2. uwnicornis) in which the same 
ratio is 82; or with the extremely short-limbed teleocerine rhinoceros 
which falls within a square 1200 mm. high by 2060 mm. long, the same 
ratio being only 58. 


CHARACTERS OF THE Baluchitheriine, New SUBFAMILY 


There is no doubt that Baluchitherium, while most closely parallel 
to the Aceratheriine, belongs in an entirely distinct line of descent or 
phylum of its own, to which the name Baluchitheriine may be applied, 
or Paraceratheriine in case by any chance this animal proves to be 
synonymous with Paraceratheriwm bugtiense. 

The polyphyletic evolution of the rhinoceroses was pointed out by 
Osborn in 1898 in his memoir on ‘The Extinct Rhinoceroses,’ Preface, 
p. 77; it was carried to a further point in 1900 in his ‘ Phylogeny of the 
Rhinoceroses of Europe,’ in which the subfamilies (I.) Diceratheriine, 
(II.) Aceratheriine, (?IIa.) Elasmotheriine, (III.) Brachypodine, (IV.) 
Ceratorhine, (V.) Atelodine, and (VI.) Rhinocerotine were clearly de- 
fined as indicating early separation, absolute distinctness, and great 
geologic age. 

The new subfamily, Baluchithertine, is most clearly distinguished 
as follows. (1) The caniniform adaptation of the second upper incisor 
tooth, which will probably be found to be correlated with a second lower 
incisor tooth quite dissimilar to that in any of the Aceratheriine; con- 
sequently both the upper and lower second incisors of the Baluchi- 
theriine are distinctive. (2) Thesecond feature of great importance is the 
horse-like elongation of the entire cervical region which is also a unique 
subfamily character. (3) The third feature, although not quite so dis- 
tinctive, is the marked elongation and lateral compression into a func- 
tional-tridactyl-monodactyl type. (4) A subsidiary subfamily feature is 
the relatively small size of the head which with the greatly elongated 
neck we may correlate adaptively with tree-browsing habits, that is, 


14 AMERICAN MUSEUM NOVITATES [No. 78 


browsing on the higher branches and leaves of trees, as compared with 
the shrub-browsing of the D. bicornis of Africa or the R. sondaicus of 
India in which the head is carried very low and near to the ground. 
(5) An additional subfamily character is the peculiar proportions of the 
head, neck, fore limb, hind limb, and trunk, which, while more rhinoc- 
erotine than equine, are nevertheless unique among rhinoceroses. 

Hapsits oF BALUCHITHERES.—The writer anticipates that the balu- 
chitheres will be found to be unique, large animals of the particular 
geologic period, either Upper Oligocene or Miocene, in which their re- 
mains occur; they were typical browsers and probably browsed on the 
herbage of the lofty branches of trees, as do the elephants and giraffes 
and to a less extent certain of the chalicotheres. They were amply de- 
fended by their powerful tusks. The type skull of Baluchitherium grangeri 
indicates that these animals attained a greater height, when the neck was 
elevated and stretched, than 14 feet, nearer 15 and possibly 16. The 
writer further anticipates that, when the complete fore limb and scapula 
of B. grangert become known, it will be found that the shoulders were 
elevated above the hips, as in the chalicothere Moropus, in the okapi, 
and in the moose (Alces), because it is generally found that the anterior 
part of the body of tree-browsers is elevated in correlation with the 
elongation of the neck. It is obvious that tree-browsing animals of 
increasing height, of length of neck, of height of shoulder, and of stretch 
of prehensile lips would be constantly selected, as in the giraffes. 


ABEL, OTHENIO. 1910. ‘Kritische Untersuchungen tiber die paléogenen Rhino- 
’ cerotiden Europas.’ Abh. k. k. geol. Reichsanst., XX, Heft 3, pp. 1-52, 
Taf. 1, 1. Wien. 
Borissiak, A. 1915. ‘Sur le genre Jndricotherium (Indricotherium n. gen.).’ Geol. 
Vestn., No. 3, pp. 131-134, text figs. 1, 2. 
1916. ‘Sur l’appareil dentaire du genre Indricotherium.’ Bull. Acad. Imp. 
Sci., (VI) No..5, March 15, pp. 343-348, text figs. 1-4. Petrograd. 
1916a. ‘L’Indricotherium n. gen., Rhinocéros gigantesque du Paléogéne 
d’Asie.’ Compt. Rendus, Paris, Tome 162, pp. 520-522. 
1917. ‘Sur l’ostéologie du genre Indricotherium.’ Bull. Acad. Imp. Sci., 
(VI) No. 4, March 1, pp. 287-299, text figs. 1-18. Petrograd. 
1918. ‘Sur lostéologie de lEpiaceratherium turgaicum nov. sp.’ Mém. 
Soc. Paléontol. Russie, I, pp. 1-82, Pls. 1-111, text figs. 1-17. Petrograd. 
Dr. Borissiak refers on p. 69 (French translation) to Indricotherium 
asiaticum as having been originally described in Comptes Rendus, 
CLXII, No. 14, April 3, 1916; Mém. Ac. Sci. Petrograd, XXXVI. 
The author, however, fails to find the specific name given in the Comptes 
Rendus and the latter memoir is not available. 


1923] - BALUCHITHERIUM GRANGERI 15 


Cooper, CuiivE Forster. 1911. ‘ Paraceratherium bugtiense, a new Genus of Rhinoc- 
erotide from the Bugti Hills of Baluchistan. Preliminary Notice.’ 
Ann. Mag. Nat. Hist., (8) VIII, December, pp. 711-716, Pl. x. 

1913. ‘Thaumastotherium osborni, a new Genus of Perissodactyles from the 
Upper Oligocene Deposits of the Bugti Hills of Baluchistan. Prelimi- 
nary Notice.’ Ann. Mag. Nat. Hist., (8) XII, October, pp. 376-381, 
text figs. 1-7. 

1913. ‘Correction of Generic Name [Thaumastotherium to ee gas US 
Ann. Mag. Nat. Hist., (8) XII, November, p. 504. 

1923. ‘Baluchitherium oeiaria (? syn. Indricotherium turgaicum, Borris- 
syak).’ Phil. Trans. Roy. Soe. London, Ser. B, CCXII, pp. 35-66, 


text figs. 1-31. 
GRANGER, WALTER, AND BERKEY, CHARLES P. 1922. ‘Discovery of Cretaceous and 
: Older Tertiary Strata in Mongolia.’ Amer. Mus. Novitates No. 42, 
P August 7, pp. 1-7, one text fig. 


Ossporn, Henry Farrrietp. 1916. ‘Bibliography of the Published Writings of 
Henry Fairfield Osborn for the Years 1877-1915.’ See p. 21 for Con- 
tributions on the Evolution of the Rhinoceroses Nos. 1 (1882.9) to 11 
(1905.269). Of these note especially 1898.143, ‘The Extinct Rhinoc- 
eroses,’ Mem. Amer. Mus. Nat. Hist., I, part 3, pp. 75-164, Pls. 
Xila-xx, and 1900.192, ‘Phylogeny of the Rhinoceroses of Europe,’ 
Bull. Amer. Mus. Nat. Hist., XIII, Art. 19, pp. 229-267. 


DD Bt: £ sows ALE NAR Atak pe Pos TENS 
iE weAHINE, ages wet Pte is SEE. Phd 734 ey 
aH eee. Re ene a wee % Sines: a gb vethne OE 
Beet a eee ee os 

; erie Ser a ae Wiebe SES: met OPar ici; ae 


: ae y ares y ; 
7 % Pas 2 AbESes Rite aes reese a. ani ieee 
* bs "2a : | 
et We. aa Tht pita? coe Leah otis we eT te! eT ae RAT, 
7 ae c =s MS Ti to fe a a “ ee ye ‘ as 
7 AA | ah |W oe 
oa aN be ; ae . Pees elys & " a! 
7% ‘ Et 
3 le ee ‘ fOFr "i gio ho 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 83 Tur AMERICAN MusEuM OF NATURAL History July 25; 1923 
New York City 


59.7.55C (51) 


DESCRIPTION OF A NEW CYPRINOID FISH FROM CHINA! 
By Henry W. FOWLER? 


Recent examination of material sent to The American Museum of 
Natural History by its Third Asiatic Expedition has brought to light the 
following new cyprinoid fish with Indian affinities. 


Chela nicholsi,’? new species 


Head, 5; depth, 444; Dn, te Aur, 250 7 P ld Went 8) senles.64iin 
lateral line to caudal base, and 4 more on latter; 10 scales above lateral line to dorsal 
origin, 3 below to anal origin; 52 predorsal scales; head width, 3 in its length; head 
depth, 1%; first branched dorsal ray, 144; first branched anal ray, 2)4; least depth of 
caudal peduncle, 274; caudal length, 1; pectoral, 1; ventral, 1%; snout, 3!4 in head 
measured from upper jaw tip; eye, 4; maxillary, 30; interorbital, 4. 

Body elongately compressed, dorsal and caudal edges convex, abdominal edge 
strongly trenchant from head to vent with scales not passing over ridge, and greatest 
depth about tips of depressed pectoral. Caudal peduncle well compressed, least 
depth 1) in its length. 

Head attenuate, strongly compressed, flattened sides little more approximated 
below and profiles nearly alike, except lower little more inclined. Snout conic, long 
as wide. Eye advanced, little inferior in depth of head, center at first *4 in length of 
head, long as snout, equals interorbital. Moderate adipose eyelid all around eye 
marginally. Mouth oblique, lower jaw slightly protruding and with slight symphyseal 
knob fitting into slight cavity in front of upper jaw. Mazxillary slips largely below 
preorbital, reaches opposite eye, and expansion 44% in eye-diameter. Nostrils to- 
gether; front one at least 74 in snout, simple pore; hind one exposed as narrow 
crescent close behind. Interorbital convexly elevated. Suborbital chain narrow; 
preorbital width about 1) in eye. 

Gill-opening forward midway in head. Gill-rakers 5 + 8, lanceolate, half length 
of gill-filaments, which 17 in eye! Pseudobranchiz 114 in gill-filaments. Pharyngeal 
teeth 4, 4, 2—2, 3, 5, hooked, slender and most of larger at least with slight grinding 
surfaces. 

Scales closely adherent, thin, papery, in even longitudinal series largely parallel 
with lateral line, little smaller on predorsal and caudal base. Row of medium- 
sized scales along anal base. Pointed scaly flap in ventral axil about % length of fin. 
Head naked. Scales with 9 to 11 basal radiating strie and 3 or 4 short marginal 
auxiliaries; apical circuli 55 to 62. Lateral line complete, strongly decurved, passes 
ventrals about lowest fourth in body depth and ascends midway along side of caudal 


pp nbBestions of the Asiatic Expeditions of The American Museum of Natural History. Publica- 
tion No. 9. 

2Of the Academy of Natural Sciences of Philadelphia. 

3Named for Mr. John T. Nichols, of The American Museum of Natural History, in recognition of his 
paper on Chinese fishes (1918). 


2 AMERICAN MUSEUM NOVITATES [No. 88 


peduncle to caudal base. Tubes in lateral line well exposed, slender, simple, extend 
over first half to 7% of scale exposures and each with slight terminal branch below. 

Dorsal origin midway between gill-opening and caudal base, depressed fin reach- 
ing 4 to caudal base. Anal begins little before end of depressed dorsal, first branched 
ray highest and forms apex of slight anterior lobe. Caudal strongly forked, lobes 
slender and sharply pointed, lower slightly longer. Pectoral long, acuminate, 1% to 
ventral. Ventral fin inserted midway between snout tip and caudal base, fin reaching 
1% to anal origin. Vent close before anal. 

Color in alcohol, back dull olivaceous, sides and lower surface silvery white. 
Dorsal and caudal slightly grayish, other fins whitish. Iris silvery white. 

Length, 177 mm. 
Typr.—No. 8254, Amer. Mus. Nat. Hist., Ningkuo, An-hwei Province, China, 
September 15 to October 15, 1921, collection Third Asiatic Expedition. 

Only one Chinese species of Chela has been described. Otherwise 
the genus is confined to the Indian region. The imperfectly known Chela 
melanopus Bleeker, based on a Chinese drawing (evidently a poor draw- 
ing), does not seem to be the same as the present species. Its fundamen- 
tal characters, such as the very large head, posterior position of its dorsal, 
striate opercle and greatly fewer anal rays (could we accept them as 
accurate) would clearly define it as distinct. Moreover, the position of 
the eye and the lips, according to the artist’s representation apparently 
fringed, convey the impression of a fish with an absolutely different 
physiognomy from the present specimen. Possibly the unusual black 
basal regions of the ventrals are the result of stain or other accident to 
the specimen from which the artist made his sketch. 


AMERICAN MUSEUM NOVITATES 


4 Published by 
Number 85 Tue AmeERICAN Mustum or NatTurat History August 28, 1923 
New York City 


59.9,4(51) 
NEW CHINESE BATS! 
By Guover M. ALLEN 


Over five hundred bats have been thus far sent back by the Asiatic 
Expeditions under the leadership of Mr. Roy C. Andrews. These are 
chiefly from Fukien, Szechwan, southeastern Yunnan, and North China. 
The series of skins is well supplemented by specimens in alcohol, and in 
many cases wide-ranging species are represented by a number of skins 
from the coastal regions as well as by others from the higher or more 
inland provinces, so that an unusual opportunity is afforded for a com- 
parison of the lowland and the upland members of a species. This is 
undoubtedly the largest single collection of bats yet obtained by any 
one expedition in China, and my thanks are due The American Museum 
of Natural History for the opportunity to study this important series. 
The following are recognized as new. 


Rhinolophidze 


Rhinolophus blythi calidus, new subspecies 


? Rhinolophus cornutus pumilus ANDERSEN, 1905, Proc. Zool. Soe. London, II, p. 127 

(in part, as to specimen from Foo-chow). 

Typr.—Adult female, skin and skull, No. 44692, American Museum of Natural 
History, from Yenping, Fukien Province, China. Roy C. Andrews, collector. June 
17, 1916. 

DescripTion.—Similar to R. blythi szechwanus Andersen, but much brighter, 
more cinnamon throughout. The bases of the hairs above are everywhere whitish, 
with a faint buffy tint, their tips dull cinnamon, near “sayal brown” of Ridgway 
(1912); below pale pinkish buff, the hairs becoming whitish near their bases. 

MEASUREMENTS.—The skull is a very little larger than in szechwanus; its total 
length, occiput to front of canine, 16 mm.; palatal bridge, 1.8; maxillary toothrow, 
5.7; mandibular toothrow, exclusive of incisors, 6.3. The forearm measures 38 mm.; 
tibia, 15; foot, 7. 

This is a bright-colored lowland representative of R. blythi of India, 
and is very different in color from the series of smoky-gray specimens 
from Szechwan representing Andersen’s R. b. szechwanus. The species 
was included as Rhinolophus minor in this author’s preliminary revision 
of the small bats of the lepidus (=pusillus) group in 1905. A fragmen- 


_ Publications of the Asiatic Expeditions of The American Museum of Natural History. Publi- 
cation No. 10. 


2 AMERICAN MUSEUM NOVITATES [No. 85 


tary skin from Foo-chow, included by him as possibly representing R. 
cornutus pumilus, may have been the present form. The nose-leaf has 
the characteristic form of the species, with the sella slightly constricted 
in the middle, and its rounded tip narrower than its base. The horn on 
the connecting process is nearly an isosceles triangle in side view. The 
dark-gray immature pelage is practically alike in both races. Twelve 

skins from Fukien Province represent this lowland form and fourteen 
from Szechwan are assumed to be typical of szechwanus. 


Rhinolophus episcopus, new species 


Tyrre.—Adult male, skin and skull, No. 56895, American Museum of Natural 
History, from Wanhsien, Szechwan Province, China. Third Asiatic Expedition. 
October 9, 1921. 

DeEscrIpTIon.—A small species of the macrotis group. Ears large and broad, with 
a wide trapezoidal antitragus, about 7 mm. across, well marked off by a deep notch. 
Horseshoe of the nose-leaf broad, quite covering the muzzle, with a narrow median 
anterior notch. On each side below the horseshoe is a small supplementary out- 
growth extending back nearly to the level of the nostril. Sella parallel-sided, with 
broadly rounded vertex, its base if anything narrower; the nasal lappets thin, their 
outer margins continuous with the sides of the sella, and vertically raised, to form a 
shallow cup at the base of the sella. The connecting process is well developed, com- 
mencing about a millimeter below the summit of the sella, prominent and convex. 
The terminal leaf instead of being pointed is ovate, with convex sides and rounded 
tip, recalling a bishop’s mitre. It is thin and leaf-lke, about as high as the sella itself 
(5 mm.) and stands erect between the ears. Front face of the sella as well as the 
terminal leaf with numerous minute hairs, smaller than those rising from the con- 
necting process. Two pairs of much longer hairs rise from the sides of the latter and 
exceed the leaves in height. 

Wings from the metatarsus. Third metacarpal shortest, fourth and fifth of equal 
length. Second phalanx of third finger one and one-half times the first. Tail about 
as long as combined tibia and hind foot. 

Color above, smoke gray, the hairs dull whitish at their bases. Below, the chin, 
throat and middle of the abdomen are pale, almost white, the hairs at the sides of 
the body smoky (near ‘‘avellaneous’’), paler at their bases. 

The skull is characteristic of the group, with very low sagittal ridge, prominent 
globular nasal swellings, narrow zygomata, and a broad palatal bridge, equalling 
one-half the maxillary toothrow. The small p? is fully in the toothrow and the third 
upper molar has its W-pattern nearly complete, the posterior commissure about half 
the length of the anterior. In the lower jaw the small premolar is external to the 
row on one side, and partly so on the other. 

Munasvrements.—Collector’s measurements of the type are: head and body, 
51 mm.; tail, 24; foot, 10; ear, 26; spread, 275. The forearm measures 47.5 mm.; 
the third metacarpal, 34.5; fourth and fifth metacarpals, 36; tibia, 18. 

-.The skull measures: occiput to front.of canine, 19 mm.; occipital condyle to 
front of canine, 17; palatal bridge, 3.7; mastoid width, 9.2; zygomatic width, 8.2; 


1923] NEW CHINESE BATS 3 


maxillar width, 6.2; width outside canines, 4.4; maxillary toothrow, 7; length of 
mandible, 12; mandibular toothrow to front of canine, 7.2. 


_ This is a larger species than macrotis with a peculiar terminal nose- 
leaf, which is rounded rather than pointed. Like pearsoni, it is repre- 
sented in the low coastal area by a smaller, brighter-colored race which 
may be named as follows. 


Rhinolophus episcopus caldwelli, new subspecies 


Typr.—Adult female, skin and skull, No. 44771, American Museum of Natural 
History, from Yuki, Fukien Province, China. H. R. Caldwell, collector. October 
31, 1916. 

DescripTion.—Similar to the typical form but smaller (forearm 43 mm., 
against 48), the pelage above with a warmer, cinnamon tint, near ‘“‘sayal brown,” 
instead of smoky. The difference in tint is similar to that separating the upland and 
lowland races of R. blythi in China, and the differences in measurements are obvious 
from the following. The teeth are strikingly smaller. The small p?is full in the tooth- 
row and. has a long sharp cusp. The lower premolars are not crowded. 

MEASUREMENTS.—Forearm, 43 mm.; tibia, 17; foot, 9; third metacarpal, 
31.5; fourth and fifth metacarpals, 33. Skull: occiput to front of canine, 18 mm.; 
occipital condyle to front of canine, 15.5; palatal bridge, 3; mastoid width, 8.5; 
zygomatic width, 7.8; maxillar width, 6.7; width outside canines, 3.7; maxillary 
toothrow, 6; length of mandible, 11.4; mandibular toothrow to front of canine, 6.5. 

The single specimen on which this form is based differs so strikingly 
from the Szechwan series that there can be no doubt of its distinctness. 
It was found in a cave at the summit of a mountain by Mr. H. R. 
Caldwell, whose codperation has resulted in the addition of many inter- 
esting species and in whose honor the form is named. : 


Rhinolophus rex, new species 


Typre.—Adult female, skin and skull, No. 56890, American Museum of Natural 
History, from Wanhsien, Szechwan Province, China. October 12, 1921. 

Descrietion.—A large member of the macrotis group, with ears, horseshoe and 
sella enormously enlarged, but the terminal lancet much reduced. 

Ears large, almost funnel-like, the antitragus fully half their height and almost 
an isosceles triangle with broadly rounded point, marked off by a deep notch from 
the rest of the conch. Horseshoe very broad extending far (at least 3-4 mm.) beyond 
the sides of the muzzle, with a deep narrow median cleft anteriorly. There is no 
accessory outgrowth external to it. The lappets covering the nostrils at the base of 
the sella are much enlarged, thin and membranous, about the width of the muzzle, their 
edges slightly raised to form a cup and their posterior wings enclosing the base of the 
sella. The latter is large, about 9 mm. high in the dried specimen (11 as measured by 
the collector), tongue-shaped, narrowest at base, and gradually expanding to the 
broadly rounded summit, which stands upright between the ears. Its front face is 
thickly beset with microscopic whitish hairs. The connecting process commences 


4 AMERICAN MUSEUM NOVITATES [No. 85 


about 4 mm. below the summit and is relatively low with a convex outline. The 
terminal lancet is very small, almost concealed by the fur of the occiput, and does not 
quite reach the summit of the connecting process in height. Its summit is broadly 
rounded instead of pointed. 

The wings arise from the metatarsus slightly below the toes. Calcaneum slender, 
about 1 times the length of the foot. Tail with its tip projecting. Third meta- 
carpal slightly the shortest, the fourth and fifth practically of equal length. The 
second phalanx of the third digit barely exceeds 114 times the first; that of the fourth 
is less than 14 times the first phalanx. 

Fur rather long, about 16 mm. on the back, 10 mm. on the chest. In color it is 
light “cinnamon-buff”’ above, paler below except at the extreme sides under the arm- 
pits. A thin fringe of hairs borders the inner edge of the ear conch and the rib 
parallel to it on the lower three-fifths of the ear. 

The skull is peculiar in having the surface of the braincase above the ear cancellar 
or spongy in appearance with numerous small fenestre as far forward as the orbit. 
The nasal swellings are elliptical, with a deep cavity behind, but the sagittal 
crest is very low. The small p? is fully in the toothrow, or even with a hair-space 
behind it and has a well developed cusp. In the lower jaw the minute middle pre- 
molar (ps) is fully in the row in all four specimens. 

MEASUREMENTS.—The collector’s measurements of the type are: head and body, 
55 mm.; tail, 38; foot, 10; ear, 33; stretch of wings, 356. The forearm measures 58 
mm.; third metacarpal, 41.5; its first phalanx, 17; its second, 26; fourth metacarpal, 43; 
its first phalanx, 12.7; its second, 17; fifth metacarpal, 43; tibia, 21. 

The skull measures: occiput to front of canine, 22 mm.; occipital condyle to 
front of canine, 19.8; palatal bridge, 4.5; mastoid width, 11; zygomatic width, 10; 
maxillar width, 7; width outside canines, 4.8; maxillary toothrow, 8; length of 
mandible, 13.5; mandibular toothrow to front of canine, 8. 


This extraordinary bat is a most interesting discovery. The exag- 
gerated development of its anterior nose-leaves is in contrast to its 
otherwise primitive structure in the unspecialized wing and the un- 
modified position of the small premolars. The addition of this and the 
preceding species to the known Asiatic members of the macrotis series is 
of importance as pointing to the origin of the group from the central 
Asian land mass. The four specimens were captured in the Yen-ching- 
kao cave, whence also was obtained the Szechwan series of Rhinolophus 
episcopus. 


Hipposideride 
Hipposideros armiger swinhoii (Peters) 
Phyllorhina swinhoii Peters, 1870, Proc. Zool. Soc. London, p. 616. 

In his review of the horseshoe bats of the armiger group, Andersen 
(1906, Ann. Mag. Nat. Hist., (7) XVII, p. 37), working mostly with 
alcoholics, placed Peters’ name swinhozi in the synonymy of this species. 
The fine series of skins obtained in Fukien Province, Szechwan and 


1923] NEW CHINESE BATS 5 


Yunnan seems to show clearly that those from the coast are uniformly 
more brightly colored, with a strong buffy suffusion. The name may 
therefore be revived in a subspecific sense for the eastern, lowland form, 
type locality Amoy, Fukien Province. 


Vespertilionide 
Myotis chinensis luctuosus, new subspecies 


Typr.—Adult male, skin and skull, No. 56867, American Museum of Natural 
History, from Wanhsien, Szechwan Province, China. October 12, 1921. Third 
Asiatic Expedition. 

Description.—A large, dark Myotis, differing from typical chinensis of the low- 
lands in having the under surface evenly gray, instead of with black sides. 

Color above a uniform grayish brown, nearly ‘buffy brown’ of Ridgway 
(1912), the top of the head a trifle grayer; below uniformly gray, the hairs nearly 
fuscous at their bases with a minute whitish tip, which gives an evenly frosted appear- 
ance, darkened by the bases of the hairs showing through. 

The calcar is long and slender, without keel. At the base of the fifth metacarpal 
a prominent membranous slip extends from the lower side of the carpus to the inner 
base of the digit. 

The skull is obviously larger than that of a specimen of chinensis from Yunnan, 
though the forearm is no longer. In both, the second small upper premolar (p*) is 
drawn slightly in from the toothrow. There is no protoconule nor hypocone on the 
molars. ; 

MEASUREMENTS.—The collector’s measurements of the type are: head and body, 
80 mm.; tail, 65; foot, 16; ear, 21; spread of wings, 456. The forearm measures, 65 
mm.; third metacarpal, 64; fourth metacarpal, 62; fifth metacarpal, 59. 

In the following measurements of the skull, those of a specimen of chinensis from 
Yung-chang, Yunnan Province, are added in parenthesis; greatest length, 24 (22.5); 
basal length, 22.5 (21); palatal length, 13.4 (12.4); mastoid width, 11.5 (11); 
zygomatic width, 15.5 (14.9); maxillary width, 9.7 (9.5); upper toothrow, 11.5 (11.3); 
lower toothrow, 12.2 (11.5). 


A series of these large bats was secured in the same cave, Yen-ching- 
kao, from which the two new Rhinolophi came. They agree closely in 
the characters given. A specimen of typical chinensis from Fukien 
Province is a much richer brown than any of the Szechwan series, and it is 
likely that this is a further difference separating the two races. Tomes, 
in his original description, gave no locality for his specimen beyond 
“China.” It was received from ‘Mr. Fortune,’ a botanist who col- 
lected in southeastern China, hence undoubtedly came from somewhere 
along that coast, perhaps from Shanghai, where he obtained other bats. 
The contrasting black sides are mentioned as a distinguishing character. 


6. 4AM ERICAN MUSEUM NOVITATES [No. 85 


Myotis frater, new species 


Typr.—Adult male, alcoholic, No. 48039, American Museum of Natural History, 
from Yenping, Fukien Province, China. August 10, 1920. H. R. Caldwell, collector. 

DescripTion.—A small species, structurally similar to VW. volans, the long-legged 
bat of western North America, but differing in details, as follows. 

Tail long, as in volans, about 50 per cent of total length; tibia very long, exceeding 
those of volans; foot much less than half its length equalling that of volans. Like the 
latter, the calear has a low but evident keel about the length of the tarsus from the 
ankle. Wings ample, the metacarpals graduated, that of the third digit nearly reach- 
ing the elbow but falling short of it by 1.5 mm. Ears short, barely reaching the muzzle 
when laid forward, their tips less abruptly rounded off than in volans. Tragus 
similar in both, short, its anterior edge slightly concave, its lower half broad, the 
posterior upper margin slightly crenulate and abruptly bevelled off to the tip. 

Below, the fur extends thinly on the under surface of the wing to a line from the 
middle of the femur quite to the elbow, as characteristic also of the American species. 
Its color is not evident in the alcoholic specimens, but is doubtless dark reddish brown 
as in volans of the Pacifie coast of California. 

The skull resembles closely that of the American species in its short, upturned 
rostrum, elevated forehead (in profile) and slightly inflated braincase. An important 
but minute detail of agreement is the conformation of the sagittal crest. Among the 
American species this is characteristic in that the temporal ridges, after uniting an- 
terior to the occiput are continued back to meet the Jambdoid crests not as concave 
but as convex lines. 

The teeth in both are small and weak, but in the Asiatic species the second small 
upper premolar is much crowded inward from the toothrow instead of standing prac- 
tically in the row, and it is proportionally as well as absolutely smaller than in the © 
American bat. Similarly in the lower jaw the second premolar is more reduced in 
size and crowded a very little inward from the row. 

MEASUREMENTS.—Total length, 94 mm.; tail, 47; foot, 8; ear from meatus, 11; 
forearm, 39; tibia, 20; combined length, knee to end of claw, 29. Skull: greatest length, 
13.5; basal length, 13.2; palatal length, 6.6; maxillary width, 5.9; zygomatic width, 
9.2; mastoid width, 8; maxillary toothrow, 5; mandibular toothrow exclusive of 
incisors, 5.4. 


A most interesting discovery is this Asiatic counterpart of M. 
volans (long known as M. longicrus) of western North America, with 
which it agrees in all important structural details, though with even more 
elongated tibiez and more progressive dentition in that the minute pre- 
molar 3 in both jaws has gone farther on its way to entire suppression. 
The more primitive condition of the American species indicates that it 
was derived from the Asiatic bat. The three specimens were found in 
holes of live bamboos on mountains at 2500 feet elevation. The bats of 
this group are distinguished by the combination of short ears, long tibiz, 
keeled calcar, fur extending to elbow ventrally, inflated skull with short 
rostrum, elevated occiput, and convex outline of temporal ridges at 
occiput. 


~I 


1923] NEW CHINESE BATS 


Nyctalus velutinus, new species 


Typr.—Adult male, skin and skull, No. 44649, American Museum of Natural 
History, from Futsing, Fukien Province, China. Edmund Heller and R. C. Andrews, 
collectors. July 29, 1916. 

DESCRIPTION eee above Prout’s brown; below paler, near Dresden brown of 
Ridgway, slightly grayer on the chest. The bases of the hairs are darker, fuscous, 
both above and below. 

On the dorsal surface, the fur of the body extends out as far as a line joining the 
proximal half of the humerus and the knee, and back on the interfemoral membrane 
nearly to a line joining the middle of the tibia. Below, the wing membrane is thickly 
furred from the knee to the basal third of the fifth digit and on the base of the fourth 
digit, as well as on the propatagium and the under side of the humerus itself. The 
extent of fur on the interfemoral membrane is much like that on its dorsal side. 

MerasurREMENTS.—The collectors’ measurements of the type are: head and body, 
75; tail, 52; foot, 11; ear, 15. The forearm measures 49 mm.; third metacarpal, 
49.5; first phalanx of same, 18; second phalanx, 21.5; fourth metacarpal, 48; first 
phalanx of same, 18; second phalanx, 8.3; fifth metacarpal, 39.5; first phalanx of 
same, 9.3; second satere: 5.3.! 

The skull measures: greatest length, 18; basal length, 18. 3. palatal length, 9; 
mastoid width, 11.2; zygomatic width, —; maxillary width, 8.5; upper toothrow 
to front of canine, 7; mandible, 13; lower toothrow to front of canine, 7.3. 


I have been unable to reconcile the characters of this Chinese noctule 
bat with those of any of the described species. In a previous paper (1912, 
Mem. Mus. Comp. Zodél., XL, p. 243), I referred specimens from 
Ichang and eastern Szechwan to Hodgson’s labiatus, but this seems to be 
a larger animal, though the two may eventually prove to be but sub- 
specifically related. Until a more thorough study of their distribution 
and relationships can be made, the eastern bat may stand as a full 
species. 


Miniopterus schreibersi parvipes, new subspecies 


Typp.—Adult male, skin and skull, No. 44656, American Museum of Natural 
History, from Yenping, Fukien Province. Roy C. Andrews, collector. June 16, 1916. 

DescripTion.—lIn. general similar to M. schreibersi chinensis Thomas from Chili 
Province, but the coloring much richer, a deep brownish instead of smoky gray, and 
the hind foot shorter and narrower. 

Color of the type, a uniform dark cinnamon-brown above, nearly “Verona brown” 
of Ridgway; below slightly paler, about “snuff brown.”’ On the lower parts the roots 
of the hairs are darker, but on the back they are nearly uniform, only a shade deeper 
in color at, their bases. 

The females, as usual in this genus, are darker than the males, from chestnut- 
brown to blackish. 

MerasurEeMEN'ts.—The forearm and finger measurements are the same as in the 
more northern form, but the foot is smaller and more slender. The type measures: 
forearm, 48; third metacarpal, 43.5; first phalanx, 10.5; second phalanx, 39; fourth 


8 AMERICAN MUSEUM NOVITATES [No. 85 


metacarpal, 42; first phalanx, 8.5; second phalanx, 19; fifth metacarpal, 39; tibia, 
17; foot, 9.5 (10.5 in chinensis). 

The skull measures: greatest length, 16; basal length, 15.5; palatal length, 8; 
mastoid width, 8.5; zygomatic width, 8.7; maxillary width, 6.5; upper toothrow 
(exclusive of incisors), 6; mandible, 12; lower toothrow (exclusive of incisors), 
6.8. 

A small series of these bats from Fukien Province is obviously 
different from others obtained in Chili Province, representing the sub- 
species chinensis. They are much browner and lack the dark smoky 
color of the latter which, in combination with the smaller and slenderer 
foot, will at once distinguish the more southern race. Its relation to 
the Indian Miniopterus remains to be studied further. In other respects 
the two seem sufficiently alike to make it probable that the South China 
form is only subspecifically distinct. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 91 Tur American Museum or Naruran History October 17, 1923 
New York City ; 


56.9,72(118:51.7) 


TITANOTHERES AND LOPHIODONTS IN MONGOLIA! 
By Henry FAIRFIELD OSBORN 


The object of this issue is to describe two of the Upper Eocene 
mammals first reported by the Third Asiatic Expedition in the Irdin 
Manha Formation (early Tertiary). It is stated (Granger and Berkey, 
1922, p. 5)?: “1—Small Lophiodonta of at least two species in great 
abundance. 2.—A perissodactyl about the size of the Upper Eocene 
titanotheres and possibly related to this family.” 

This constituted the first notice of the existence of the perisso- 
dactyl families Lophiodontide and Brontotheriide in Mongolia. Later 
there was reported by Berkey and Granger (1923, p. 12)? another for- 
mation known as the Shara Murun: 

“5—SHaRA Murun. A hundred miles farther south another 
basin carries fossiliferous strata. The sedimentary area is very large 
‘and the best exposures occur along the borders of a 200-foot escarp- 
ment. At this place were found titanothere remains very like those 
found at Irdin Manha early in the summer, in the Iren Dabasu area. 
These seem to be, therefore, early Tertiary in age, perhaps as early 
as Eocene.” 

The two new species of mammals described herewith appear to 
indicate that the Irdin Manha formation and the Shara Murun for- 
mation are alike of Upper Eocene age and comparable to the Uinta C 
formation, northern Utah. In 1883 Scott* described Desmatotherium 
and Dilophodon as two new Eocene lophiodonts; Osborn later united 
them with the Middle Eocene Helaletes of Marsh; Osborn and Matthew 
in 1909 (U. S. G. S. Bull. 361, pp. 52, 98) and Peterson in 1919, 
p. 127,° separated them, and Troxell, 1922, p. 367, also separated them. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Publi- 
cation No. 11. 

2Granger, Walter, and Berkey, Charles P., 1922, ‘Discovery of Cretaceous and Older Tertiary 
Strata in Mongolia,’ Amer. Mus. Novitates No. "42, pp. 1-7, 1 text-fig. 

‘Berkey, Charles P., and Granger, Walter, 1923, ‘Later Sediments of the Desert Basins of Central 
Mongolia,’ Amer. Mus. Novitates No. 77 , pp. 1-16, map. 

4Scott, W. B., 1883, ‘On Desmatotherium and Dilophodon, Two New Eocene Lophiodonts. P 
Contrib. E. M. Mus. Geol. and Arch. Prince. Mus., Bull. III, pp. 46, 47, Pls. v—vitr. 

5Peterson, O. A., 1919, ‘Report upon the Material Discovered in the Upper Eocene of the Uinta 
Basin by Earl Douglas i in the Years 1908-1909, and by O. A. Peterson in 1912,’ Ann. Carnegie Mus., 
XII, Nos. 2-4, pp. 40-168, 14 plates. 

®Troxell, Edward L., 1922, ‘ Helaletes Redefined’ (Contributions from the Paleontological Labora- 
tory, Peabody Museum, Yale University), Amer. Journ. Sci., III, pp. 365-370, text-figs. 1-3. 


2 AMERICAN MUSEUM NOVITATES [No. 91 


We find that the Irdin Manha lophiodont, now reported as extremely 
numerous, is very close to Desmatotherium guyotii, a genus and species 
distinguished by double internal cones in the upper premolars; in 
Helaletes these cones are single. 


Desmatotherium mongoliense, new species 
Fam. Lophiodontidz, Subfam. Helaletinz 


In the Irdin Manha beds, twenty-three miles south of Iren Dabasu, 
there were discovered on April 26, 1922, parts of ten individuals of a 
small lophiodont which were described in field letters of Granger as 
representing an animal very similar to Helaletes of the Bridger and Uinta 
beds of Wyoming and Utah. Now that these materials have reached 
the Museum and been closely compared with the Lower Eocene Hepto- 
don and with the Middle Eocene Helaletes and Desmatotherium, it is 
found that Granger’s identification is entirely correct. These animals 
certainly belong to the family Lophiodontide, subfamily Helaletine 
of Osborn, and the best preserved specimens, namely, a right maxilla 
(Amer. Mus. 19161) and a left lower jaw (Amer. Mus. 19162) are selected 
as the type and paratype. The comparative measurements are as 
follows: 


Indices 
Desmatotherium guyotii (Prince. Mus.) P-M?=79 Mi3=44 
M? diams. a—p. Xtr.=14X15 107 
ae mongoliense, Type PL_M?=75e M!3=37 
M? diams. a—p. Xtr.=15 X15 100 
eS te Paratype P:-M; =67 
Dilophodon minusculus (Prince. Mus.) P.-M; =42 


It thus appears that the species Desmatotherium mongoliense is 
intermediate in size between D. minusculus and D. guyotii of the Middle 
Eocene Bridger beds of Wyoming, as described by Scott in 1883. In 
revising these Eocene species of the Helaletine, Osborn (1892.67, p. 
131)! pointed out that Desmatotheriwm is defined as follows: ‘ Premo- 
lars 4. Third and fourth upper premolars with two internal lobes. 
Third lobe of the last lower molar variable. Paracone conic, and meta- 
cone flattened, symmetrical, of equal length.”’ 

The new species from Mongolia is defined as follows: Vestigial P! 
indicated by a small alveolus; P? and P* with duplicate internal cusps, 
i.e., tritocone and tetartocone; P* with single internal tritocone only, 
and rudimentary conules; M3 with a third lobe. 


1Osborn, H. F., and Wortman, J. L., 1892, ‘Fossil Mammals of the Wahsatch and Wind River 
Beds. Collection of. 1891,’ Bull. Amer. Mus. Nat. Hist., IV, Art. 11, pp. 81-147, 1 plate and 18 text- 
figs. The definition is under the name of Helaletes but is based upon Desmatotherium. 


1923] TITANOTHERES AND LOPHIODONTS IN MONGOLIA 3 


Thus this animal is specifically but not generically distinguishable 
from Desmatotherium. Parts of the metapodials, complete calcanea, 
and complete isolated superior and inferior molar teeth found under 
the same field numbers support this reference. 


Protitanotherium mongoliense, new species 


The type (Amer. Mus. 18653) consists of lower jaw with well 
preserved series of lower grinding teeth, P:-Ms, of the right side, also 
foot bones, belonging to an animal superior in size to Protitanotherrum 
emarginatum of the Uinta C beds, Utah, and somewhat inferior in size 
to P. superbum the largest and most robust animal thus far found in 
the Upper Eocene of Utah and of which only the jaw is known, but 
the skull of which doubtless presented osseous horns of considerable 
size. The relative size and proportions of the teeth in these three 
specimens may be indicated by the following comparative measurements. 


Indices 
P.. superbum (Amer. Mus. 2501) P.-M3;=296 M)3=214 
Mi diams. a-p. Xtr.=51 X31 61 
M “ “ =62 X32 52 
Wis s* Ne ae OS AL 43 
P. mongoliense (Amer. Mus. 18653) P.-M3=293 M4)-3=200 
‘ Mi diams. a-p. Xtr.=50X30 60 
M “ fo = GIOCSe 56 
Naw es Oe = OD X32 35 
P. emarginatum (Prince. Mus.) P.-M3;=276 Mi3=199 
Mi diams. a-p. Xtr. =46 X29 63 
M “ eres 59 
M3 OC“ “94K 34 36 


The above figures support the statement that Protitanotherium 
mongoliense is intermediate in size between the two Utah species in 
the total length of its grinding series. The grinding teeth are also inter- 
mediate in proportions, as shown by the indices of the three inferior 
molar teeth; this index tends to establish the generic relationship of 
the species mongoliense to Protitanotherium rather than to Diplacodon, 
in which the grinding teeth are relatively narrower. The internal and 
external aspects of the crowns are also closely similar in P. superbum 
and in P. mongoliense; the molarization or transformation of the pre- 
molars into the molar pattern appears to be somewhat more advanced 
in P. mongoliense than in P. superbum. The premolar measurements 
and indices are as follows. 


4 AMERICAN MUSEUM NOVITATES [No. 91 


Indices 
P.” superbum P, diams. a-p. Xtr.=28 X16 57 
See Bae. fh on ba Cee Oe aoe 71 
or Panny “ & =34 X22 65 
P. mongoliense Pye eS e  =28.5X16.5 58 
Peace os eae | bak nea J 68 
Poses “= 34 X23 68 


Fig. 1. Protitanotherium mongoliense, new species. 


Type lower grinding teeth, P2-Ms, of the right side (Amer. Mus. 18653). After a photograph 
reduced to one-third natural size. From the Upper Eocene of Iren Dabasu, southeastern Mongolia. 
Amer. Mus. Exped. 1922. 


Lnner vLrew- 


Fig. 2. Type of Protitanotherium superbum (Amer. Mus. 2501). 


_ Lower grinding teeth, Pi-Ms, of the right side. After a photograph reduced to one-third natural 
size. From the Upper Eocene of Utah, Uinta C formation. Amer. Mus. Exped. 1895. 


1923] TITANOTHERES AND LOPHIODONTS IN MONGOLIA 5 


It remains to be seen whether these two formations are of similar 
geologic age. The two species above described give a most interesting 
American aspect to the fauna of this region of Mongolia in Upper 
Eocene time. 

The Expedition of 1923 adds a large number and variety of titano- 
theres from Mongolia apparently of Upper Eocene and Lower Oligocene 
age. This material represents three or four species and probably three 
genera; the animals are very large, medium, and small. The titano- 
theres must have been exceedingly abundant. These animals will be 
described by the present writer in a succeeding number of American 
Museum Novitates. 


aa 
re 


J Bae Shera! * 


er - 


acess eih Sel ic 


t Pas 
‘ae 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 92 Tue AMERICAN Museum or Natura History October 19, 1923 
New York City 3 


56.73,5(118:51.7) 


CADURCOTHERIUM FROM MONGOLIA! 


By Henry FAIRFIELD OSBORN 


Six Tertiary formations have thus far been determined in Mon- 
golia, as follows: 


Horizon 
NAMES Firetp Notes Museum Norss 
Hung Kureh Mastodon and Cervus—2000’ Hipparion, Castor 


Hsanda Gol Baluchitheriwm and rodents—3000’ Baluchitherium grangeri type 
Houldjin Rhinoceros and Baluchitherium—50’ 


Ardyn Obo —Rhinoceros—500’ Cadurcotherium, Schizotherium 
Shara Murun Titanothere beds—50’ Protitanotheritum mongoliensetype 
Irdin Manha Lophiodont beds—50’ Desmatotherium mongoliense type 


The Ardyn Obo formation was named and defined by Berkey 
and Granger in American Museum Novitates No. 77, p. 12. The 
fauna was judged in the field to be of mid-Tertiary age, not very dif- 
ferent from the Hsanda Gol and Houldjin formations in other Tertiary 
basins in Mongolia. The preliminary examination in the Museum 
of the material collected in 1922 confirms this judgment and likewise 
confirms the suspicion intimated by Berkey and Granger that they 
are either of different facies or of somewhat different geologic age. 
A preliminary list of the fauna follows: 


Canin, cf. Cynodictis; part of lower jaw 
AMYNODONTIDZ 

Cadurcotherium ardynense, new species; skulls, jaws, ete. 
?RHINOCEROTIDZ 

New genus and species; upper jaw, etc. 
CHALICOTHERIID 

Schizotherium, new species; M3. 
CrERVID® 

New genus and species; lower jaws 
CHELONIA 

Testudo, large species; parts of shell 

Emydid, gen. indet.; fragments of shell 

Trionychid, gen. indet.; plates of carapace 


_ Publications of the Asiatic Expeditions of The American Museum of Natural History. Publi- 
cation No. 12. 


2 AMERICAN MUSEUM NOVITATES [No. 92 


The little cervid is the only element in common with the Hsanda 
Gol fauna except for the doubtful Cynodictis and perhaps the chelo- 
nians (not yet studied). Cadurcotherium, the extreme evolutionary 
stage in the family Amynodontide, appears to be represented by teeth 
and other fragments in the Houldjin fauna. Larger collections from 
the three Oligocene formations (Ardyn Obo, Hsanda Gol, and Hould- 
jin) will presumably increase the common elements and means of 
comparison. At present, while it appears that all three are of Oligo- 
cene age, a more exact correlation would be premature. 


Cadurcotherium ardynense, new species 


Type.—Amer. Mus. 19154; skull with lower jaw, fore and hind limbs and feet 
probably associated. With this are a young lower jaw and several limb and foot 
bones of other individuals. 

Paratyprs.—Amer. Mus. 19155-19158; upper and lower jaw fragments and teeth. 

Horizon AND Locauiry.—Ardyn Obo formation, Promontory Bluff, Ardyn 
Obo basin, about 350 miles west of Kalgan on the Sair Usu caravan trail. 

FamMILy AND GENERIC CHARACTERS.—Dentition 2:+:3'3 (the incisors are not 
known in the Mongolian animal). Teeth moderately hypsodont, premolars reduced 
and crowded, molars compressed transversely, the transverse crests short and, 
strongly oblique. Canines enlarged, vertical, of subcircular cross-section save at 
the unworn tip, the tips usually worn to a flattened oblique shear, anterior on the 
upper, posterior on the lower canines. Digits 4-3, the fifth metacarpal only a little 
smaller than Mc II or IV; the metacarpals considerably longer than the metatarsals. 

Distinctions FRoM Metamynodon.—(1) Upper premolars with the inner crests 
completely joined into an inner crescent, the wings extending to the external angles, 
and enclosing a deep round median fossa at.almost all stages of wear. In Metamy- 
ncodon the inner crests are more or less distinct and transverse. In European species 
of Cadurcotherium the condition is more or less intermediate. (2) Lower premolars 
like those of Cadurcotherium cayluxi and minus; P3 very small and simple with two 
small subparallel crests extending backward from apex of main cusp; Pi sub- 
molariform with obliquely transverse inner crests, anterior, median, and posterior, 
simulating the transverse lophs of the molars. In Metamynedon both P3 and P, are 
of this submolariform type. (3) Limbs and feet small and slender in comparison 
with Metamynodon, proportioned much as in Aphelops megalodus except that the 
metacarpals are considerably more elongate. The metapodials are both smaller and 
more slender than in Amynodon, and lack the peculiar broad flattening of the shafts 
that distinguishes both Amynodon and Metamynodon. 

Speciric Cuaracters.—About the size of Cadurcotherium cayluxi, but the 
molars appear to be shorter crowned, the upper premolars have the median fossa 
more fully developed. Other characters may appear when the type is completely 
prepared. As the skull and feet of Cadurcotherium have not hitherto been known, 
the new species will add materially to what is known of the genus as well as extend 
its distribution to a region hitherto unknown palzontologically. 

Cadurcotherium indicum Pilgrim of the Gaj series of India is of much larger 
size and different in other particulars enumerated by Doctor Pilgrim. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 95 Tur Amertcan Museum or Naturat History October 19, 1923 
New York City 


56.81,9(117:51.7) 


TWO LOWER CRETACEOUS DINOSAURS OF MONGOLIA! 
By Henry FAIRFIELD OSBORN 


The Third Asiatic Expedition of 1922 discovered four formations 
which, from preliminary definition and examination of the fossils they 
contain, were referred to the Cretaceous, namely: 


HUNG KURE 
HSANDA GOL 
SX ONDAI SAIR 


SCALE 
190 50 ° roo zoo 


Fig. 1. Route of Third Asiatic Expedition in Mongolia, 1922, showing the 
localities of the Cretaceous and Tertiary formations discovered. 
*Cretaceous formations; °Tertiary formations, 


Iren Dapasu Formation, southeast Mongolia, containing (1) predentate dinosaurs, 
probably of the bipedal type; (2) carnivorous dinosaurs of at least two genera, 
the smaller one being of the Ornithomimus type; (3) crocodiles; (4) turtles of 
the Trionyx type; (5) a few pelecypod shells. 

Onpar Sarr Formation, Ussuk, Tsagan Nor basin, western Mongolia, containing 
skeleton (Amer. Mus. 6253) articulated and nearly complete, except skull 
which is mostly weathered out. 


é <a laa of the Asiatic Expeditions of The American Museum of Natural History. Publica- 
ion No 


2 AMERICAN MUSEUM NOVITATES [No. 95 


AsHILE Formation, Artsa Bogdo basin, western Mongolia, containing skeleton 
(Amer. Mus. 6254) in sandstone, tail and pelvis mostly weathered out, skull 
and jaws, backbone and large part of limbs in block. 

Dsa-pocH-TA ForMATION, Kwei-wa-ting trail, east of Artsa Bogdo, Mongolia, con- 
taining type skull of Protoceratops andrewsi; nearly complete skull (Amer. Mus. 
6251) and lower jaws, hornless and far smaller than that of any known 
ceratopsian or ankylosaur, being only about 160 mm. in length, as fully de- 
scribed by Granger and Gregory. 

The geographic location of these four Cretaceous formations is 
shown in the accompanying sketch map (Fig. 1). The relative age of 
each can only be determined by careful comparison of the fauna found. 
The only specimens carefully examined up to the present time are the 
Protoceratops andrewsi type and the two new types from the Ondai Sair 
and Ashile formations to be described in the present bulletin. 


ASHILE FORMATION 


Psittacosaurus mongoliensis, new genus and species 


The type of this new genus and species (Amer. Mus. 6254) is an 
almost perfect skull and jaws of an adult animal, with the skeleton, 
including the backbone, tail and pelvis and most of the limbs, in the 
block. This superb specimen, locally known as the Red Mesa (Ohshih) 
skeleton, was discovered by the Third Asiatic Expedition in the Artsa 
Bogdo basin through the bright observation of Wong, the Mongolian 
chauffeur engaged in Urga. At the time of the present description its 
characters are not fully known, because the partly exposed teeth (Fig. 
2A) are not yet worked out and the skeleton still lies in the matrix. On 
the revelation of these still concealed characters the question of the 
affinity of Psittacosaurus to Protiguanodon will be decided. 

The proportions of this skull are relatively short, deep, narrow in the 
facial region, broad in the cranial region. Teeth apparently confined to 
the maxillaries and dentary, crowns closely compacted; dentition not 
fully exposed. Functions of edentulous premaxillary and predentary 
element below suggested by powerful cutting and crushing rostrum or 
beak, for which the name Psittacosaurus, or parrot-beaked saurian, has 
been suggested by Dr. Gregory. 

Lert Larerat Aspect (Fig. 2A).—Premaxillary distinct, uniting 
by suture with nasals. Edentulous borders of premaxillary and anterior 
portion of maxillary. Small narial opening, large orbital, large latero- 
temporal, small auditory fenestre. Sutures of premaxillaries, prefrontals, 
supraorbitals, postorbitals, squamosal, partly separated. Quadrate 


1923] CRETACEOUS DINOSAURS OF MONGOLIA 3 


deep, forming anterior border of auditory fenestra, codssified with 
quadratojugal element below. Prominent osseous horns on sides of 
jugals. Broad and powerful jugal and infraorbital bar, indicating power- 
ful muscles to control mastication, beak sheathed in horn. In the 
powerful jaw, surangular, angular, dentary and predentary regions 
fairly indicated; predentary coalesced with dentary in type. 

Ricut Latrerau Aspect (Fig. 2B).—In this aspect the tubercular 
dermal armature of the side of the throat is revealed, consisting of 
numerous rounded dermal cones, which represent minute ostoses. 
Whether this armature belongs to the side of the jaw or is a throat 
armature slipped out of place is difficult to say; it corresponds in general 
appearance with some of the throat armature observed in the Ankylo- 
sauria. Similar rudimentary armature may have extended over the 
surface of the skull preliminary to the formation of dermal plates, as 
observed in the armored Stegoceras and Ankylosaurus. In other features 
the right lateral aspect agrees with the left, except that the supraorbital 
is less clearly exposed and the premaxillary is more firmly united with 
the maxillary. 

Superior Aspect (Fig. 2C).—In this aspect the general triangular 
form of the skull, the broadened cranial, the narrow facial and rostral 
region are clearly shown, reminding us of the Protoceratops andrewsi 
skull. The orbital, laterotemporal and supratemporal fenestre are clearly 
shown, placing this skull with the Diapsida group. The nasals, supra- 
orbitals, postorbitals, squamosals and quadrates are shown. Beyond the 
auditory fenestra lies the paroccipital process of the opisthotic element. 
The sutures of the mid-cranial region are closed and the boundaries of 
the cranial bones not clearly defined. 

Posterior Aspect (Fig. 2D).—In this aspect the skull is seen to be 
still partially enclosed in matrix. The borders of the occipital foramen, 
of the auditory and of the post-temporal fenestree may be seen. The 
powerful quadrate exhibits its sutural line of union with the quadrato- 
jugal. The cranium itself is shallow but is given a depressed aspect by 
the downward extension of the quadrates and lateral osseous jugal horns. 

ANTERIOR Aspect (Fig. 2#).—In front aspect the parrot-beaked 
rostrum of Psittacosaurus is well shown, also the relation of the dermal 
armature of the right side of the skull (left in this figure) to the osseous 
jugal horns. It is the correlation of these horns with the dermal armature 
which first led the writer to the opinion that this animal may be ancestral 
to some member of the armored suborder Ankylosauria. In this front 
aspect the sutures of the maxillaries, nasals, prefrontals, supraorbitals, 


AM.6254 


Ze 


WY I WY): 
Yy fy 
VME YY Za 
Yj "yyy 


““ 


LOTT 
We 


Fig. 2. Psittacosaurus mongoliensis, type skull (Amer. Mus. 6254). 


A. Left lateral aspect. One-half natural size. 


B. Right lateral aspect, showing impressions of dermal armature beneath the osseous jugal horns, 
One-half natural size. 


Superior aspect. One-half natural size. 
Posterior aspect. One-half natural size. 


E. Anterior aspect, showing dermal armature of the right side below quadratojugal horns. One= 
half natural size. 


6 AMERICAN MUSEUM NOVITATES 


[No. 95 


may be partly detected, as indicated in the drawing. It appears that 
the narial and orbital fenestra face partly forward. 


Psittacosaurus mongoliensis 

Herbivorous diapsid reptile with pre- 
dentary bone and horny beak. Maxillary 
teeth compressed, not fully known. Skull 
short and deep, narrow anteriorly, broad 
posteriorly. Rostrum prominent, parrot- 
like, edentulous. Nostrils small, orbits 
large. Infraorbital region and jaw heavy, 
with attachment for powerful muscles. 
Primitive dermal armature in head region; 
lateral osseous horns on jugals. 


Genotype of Psittacosauride, new 
family. Skeleton and teeth only partly 
known; supposed primitive armored 
dinosaurs, possibly related to the fully 
armored Upper Cretaceous types. 


Protiguanodon mongoliense 

Probably a diapsid reptile with pre- 
dentary bone and horny beak. Nine 
teeth on the dentary, of iguanodont pat- 
tern. Rostrum short, jaw very deep but 
relatively slender. Teeth less compacted 
than in Psittacosaurus. Skull probably 
deep and short, without rostral prolonga- 
tion. No evidence of dermal armature. 
Skeleton of Pro-Iguanodontia type; 
limbs of prebipedal type. Manus func- 
tionally tridactyl with vestigial IV. Pes 
functionally tetradactyl with vestigial V. 
Prominent prepubie process. Ilium, 
scapule and limbs of Pro-Iguanodontia 
type. Osseous tendons and tail indicat- 
ing bipedal locomotion. 


Genotype species of subfamily Prot- 
iguanodontine. 


ONDAI SarR FORMATION 
Protiguanodon mongoliense, new genus and species 


The type of this new genus and species is a nearly complete, articu- 
lated skeleton (Amer. Mus. 6253) lying on its ventral face, with the four 
limbs prone, the left jaw and portions of the skull weathered out, as 
represented in the diagram (Fig. 3). This figure, drawn by Mr. John 
Germann, exhibits only the parts from which the matrix had been re- 
moved at the time of writing and serves to supplement this preliminary 
description. 

The animal is 53 inches (1350 mm.) in length; the outstretched 
fore limb measures about 12 inches (340 mm.); the outstretched hind 
limb measures about 480 mm., 140 mm. longer than the fore limb. These 
proportions are of a prebipedal type, in which the fore limb occasionally 
touches the ground. The balancing tail measures 710 mm. from the 
center of the ilium to the tip. The thoracic vertebree measure 340 mm. 
from the center of the ilium to the base of the scapula. The neck has not 
yet been exposed and its length cannot now be established. 

Peculiar iguanodont features are the ossified tendons lying above 
the transverse processes of the thoracic vertebre; another iguanodont 


: ‘OZI8 [BIN}VU YZUIU-IUGQ “puNo; 
se Ayeyeurxoidde posed ere [[Nys ay} Jo yuIUINS oy} puB MEl OY, “paAtosqo SBM AINYBUIIE [VULIOp JO 9oVI} OU UOZO[AYs sIyy Sutsodxe uy 


‘OAOQE WOIF Woas SB UOT}RUIIO eg tepug WoT (EGZO ‘SHI “Ioury) asuayobuow uopouvnbyourg Jo uoyapays eddy, “Ee “SI 


eSZzo ny 


NNWHYSS 


8 AMERICAN MUSEUM NOVITATES [No. 95 


‘ character is the presence of 11 stomach-stones opposite the superior 
border of the scapula. 

The contour and proportions of the scapula and of the ilium are 
well shown in the diagram. Below the anterior border of the ilium is a 
prepubic extension of the pubis; the postpubis has not yet been exposed 
from the matrix. The manus is tetradactyl: I, Mtc, short, with two 
phalanges; II, intermediate, with three phalanges; III, elongate, with 
four phalanges; IV, vestigial, with single phalanx. Three equal-sized 
carpals. The pes is pentadactyl: I, Mts, short, with two phalanges; II, 
elongate, three phalanges; III, most elongate, four phalanges; IV, 
same length as Mts II, five phalanges; V, vestigial, one phalanx. Manus 


Fig. 4. Type jaw of Protiguanodon mongoliense (Amer. Mus. 6253). 


A. External view of dentary, showing in the outer aspect the nine dentary teeth. Natural size. 
B. Inner aspect of the upper portion of the dentary, showing same nine teeth from the sculptured 
internal surface. 
and pes, although differing in size, are homodynamous. Four tarsals 
are observed, including an astragalus closely united with the tibia. Both 
manus and pes are in the same stage of evolution as those of Hypsi- 
lophodon, except that the Wealden genus has four digits in the manus. 
Affinity to the suborder Iguanodontia is revealed in these skeletal 
characters, which indicate a prebipedal stage. The affinity shows still 
more clearly in the jaw (Fig. 4) and teeth, as worked out of the matrix by 


1923] CRETACEOUS DINOSAURS OF MONGOLIA 9 


Mr. Falkenbach and carefully drawn by Mr. R. Weber. Of the nine teeth 
in the dentary, the unworn ninth tooth of the left side of the mandible © 
shows a most clearly typical structure, consisting of a prominent median 
ridge, with lesser anterior and posterior ridges serrated along the superior 
border. Less worn teeth are No. 4 and No. 6; in more advanced wear are 
the apparently trilobed Nos. 3 and 8; completely worn are Nos. 1 and 7. 
In external aspect the triserrate interior view of No. 9 appears in front of 
the anterior border of the coronoid. The jaw itself is short and deep, 
apparently edentulous anteriorly, a thin border rounding inwards, as 
shown in Fig. 4. Thus the animal is short-jawed. 

Comparing this animal with Hypsilophodon of the Wealden, with 
Thescelosaurus of the Lance, with the three genera composing the family 
Laosauride of Marsh (Nanosaurus, Laosaurus, Dryosaurus), and with 
the families Iguanodontide and Trachodontide, we may define it as 
belonging to a new genus, Protiguanodon, typified by the genotypic 


(9 


A.M.6253 


Fig. 5. Greatly enlarged (X3) anterior, exterior and interior aspects of a 
single tooth, which may belong in the maxillary series. 


species Protiguanodon mongoliense. When more fully known it may prove 
to belong to a distinct subfamily of the Iguanodontide, for which the 
subfamily name Protiguanodontinz would be appropriate. ; 
The writer is of the opinion that this animal is distinct from the 
reptile above described, but Dr. W. K. Gregory, to whom the writer is 
deeply indebted for valuable notes and comparative observations, con- 
siders that it represents a juvenile stage of an animal similar to the very 
adult stage which constitutes the type of the genus Psittacosaurus. In 
favor of the view that the animals are the same are (1) close similarity in 


* 


10 AMERICAN MUSEUM NOVITATES {No. 95 


size; (2) close resemblances in the parietals of the skull, as far as observed 

‘in this type, with that of the skull described above ; (3) general similarity 
in the quadrate. Against the idea that these animals belong together 
are the following: (1) absence of dermal armature; (2) comparative 
unlikeness of the skeleton and skull, as far as observed; (3) wide geologic 
separation of the two types. The question will be positively settled by 
the following means: (1) by the complete exposure of the dental series in 
Psittacosaurus; (2) by the exposure of the skeleton in the same; (3) 
by minute comparison of the tooth structure of the two types. In case 
the animals prove to be the same, Psittacosaurus will have precedence. 


I 3 
ae Brees st 
LR ry Oe eA ne rs 


* 4 ° 
ey 
eS os 


mS a ee 
i ftir athe. 
; hid Nyte 


59.81.4:14.71,5 


Article XVI—SKULL CHARACTERS OF ALLIGATOR SINENSE 
FAUVEL!} ? 


By CuHaryues C. Moox 


When the skull characters of the recent species of Crocodilia were 
described by the writer in 19214 no skull of Alligator sinense was available 
for study, in consequence of which this species was dismissed with a 
brief summary of characters taken from Boulenger’s ‘Catalogue of the 
Recent Reptilia in the British Museum.’ 

In 1922 the American Museum received a number of skins, skulls, 
and skeletons of the Chinese alligator, collected by the Third Asiatic 
Expedition. This material sheds light upon previously unknown char- 
acters of this species and forms the basis of the present communication. 
This description is based particularly upon one skull (A. M. N. H. No. 
23898) and many of the characters have been verified upon three other 
skulls (A. M. N. H. Nos. 23899, 23900, 23901). 


GENERAL ForRM 


The skull is short and broad. It is relatively shorter than that of A. 
mississippiensis and resembles in this respect the: Miocene A. thomson. 
The height is greater than in the Florida species. 

The snout is moderately broad at its anterior end and expands 
rapidly to the level of the fourth maxillary teeth, then contracts slightly 
to the level of the sixth maxillary teeth, back of which it expands again. 
The length of the snout is one and one-sixth times as long as its breadth 
at the base. 

The interorbital plate is situated at a distinctly higher level than the 
snout and descends abruptly at its anterior end. A pair of prominent 
ridges extend forward and outward along the anterior borders of the 
orbits to points slightly in front of the semi-detached supraorbitals and 
then extend directly forward over the base of the snout. In this char- 
acter the skull resembles those of the various species of Jacare and that of 
A. thomsoni but differs from that of A. mississippiensis. The inter- 
orbital plate is relatively narrow and is uprolled at its edges. It is also 
slightly convex in antero-posterior profile. 

The cranial table is short antero-posteriorly but rather broad 
laterally. Its lateral borders converge in the anterior direction at such 


1Contributions to the Osteology, Affinities and Distribution of the Crorodilia, No. 1 

hate aaa of the Asiatic Expeditions of The American Museum of Natural History. Publi- 
cation No 

3Bull. Amer. Mus. Nat. Hist., XLIV, Art. 13, pp. 123-268. 


553 


Am.Mus.No.23898 


Fig. 1. Alligator sinense Fauvel. 


‘ Skull (Amer. Mus. No. 23898). One-half natural size. A, superior view; B, lateral view, left 
side; C, inferior view. 


1923] Mook, Skull Characters of Alligator Sinense 5DO 


an angle that, if produced forward, they would meet slightly beyond the 
tip of the snout. The plate between the supratemporal fenestre is of 
moderate breadth and is distinctly uprolled at its edges. The lateral 
borders of the skull are more distinctly wavy than in the southern 
alligator. 


THE CAVITIES OF THE SKULL 


SUPRATEMPORAL FENESTR2&.—The supratemporal fenestre are 
large at the surface of the cranial table but diminish rapidly in size in 
depth. Viewed from above, less than half of the area of each fenestra 
penetrates to the base of the brain-case. The larger part, chiefly pos- 
terior, is very shallow and is floored by portions of the parietal and squa- 
mosal bones. The small internal cavities which extend from the supra- 
temporal fenestra to the aural passages are in most crocodilians nearly or 
quite invisible when the skull is viewed from above; in this species they 
are entirely visible from above. The edges of the fenestre are distinctly 
uprolled, especially on the postero-internal borders. In this character 
they differ from those of A. mississippiensis and A. thomsoni. In 
shape the fenestre are irregular and not smoothly rounded. Their out- 
lines end posteriorly in rather blunt points and anteriorly, or rather 
antero-externally, in sharp points. There is a certain amount of varia- 
tion among the various individuals in regard to this character. The 
space between the fenestre is relatively broad. 


INFRATEMPORAL FENESTR#.—These fenestra are of moderate size. 
Their chief point of interest is that their triangular outline is more 
rounded than in most crocodilians. 


Orsits.—The orbits are large. They are irregular in outline; their 
posterior and postero-internal borders are broadly rounded; _ their 
antero-internal borders are nearly straight, as are their external borders. 
They end anteriorly in rather sharp points, which are nearer their ex- 
ternal than their internal boundaries. The length of the orbits is con- 
siderably greater than their breadth. The interorbital plate is of mod- 
erate breadth; its borders are sharply uprolled. In all of the specimens 
the anterior ends of the orbits lie over several maxillary teeth. This 
character indicates immaturity in all of them. 

EXTERNAL Nares.—The external narial aperture is broader than it 
islong. The bony bar which divides it into right and left elements at the 
surface is moderately stout. It is composed chiefly of the anterior proc- 
esses of the nasals, but partly of processes of the premaxillaries. In 
outline the cavity is a somewhat rounded quadrilateral, whose anterior 


596 Bulletin American Museum of Natural History [Vol. XLVIII 


border is slightly curved and whose lateral borders converge slightly in 
the posterior direction. The posterior border is very irregular. The 
lateral borders are distinctly elevated above the general level of the 
anterior end of the snout. The median bar is also elevated and the pos- 
terior border slightly so. 

PREMAXILLARY ForRAMEN.—This foramen is very small; on the 
median line it occupies less than one-third of the distance between the 
tip of the snout and the premaxillo-maxillary suture. It is acutely 
pointed at both anterior and posterior ends and its lateral borders are 
simple curves. 

PALATINE FENESTR&.—The palatine fenestrae are small and are 
very irregular in shape. In this the species differs from A. mzssiss7p- 
piensis. The internal borders are nearly straight parallel lines through- 
out the posterior two-thirds of their lengths; the anterior thirds diverge 
rapidly. The external borders are composed of two components of 
about equal length, which make pronounced angles with each other 
slightly posterior to the level of the last maxillary teeth. About one- 
fifth of each external border consists of maxillary bone, the ectoptery- 
goid portion comprising nearly four-fifths, with a minute portion at the 
posterior end consisting of pterygoid; the entire internal border is com- 
posed of palatine. The maximum breadth is considerable in proportion 
to the maximum length. The anterior end is broadly rounded but the 
posterior end is acute. The anterior end is situated at the level of the 
space between the tenth and eleventh maxillary teeth. 

INTERNAL NARIAL APERTURE.—This aperture is completely divided 
by a median vertical plate. On its anterior border at the median 
line is situated a shght but distinct elevation. Posterior to the aperture 
is a prominent vertical ridge, approximately semicircular in outline, 
which completely separates the depression, of which the aperture is the 
center, from the postero-external elevated flanges of the pterygoid bone. 


THE BONES OF THE SKULL 

PREMAXILLARIES.—The premaxillaries are considerably broader in 
proportion to their length on their superior surfaces than in the Florida 
alligator. Their posterior processes are of moderate length, extending 
backward from the level of the second to that of the fourth maxillary teeth. 
A small process of both premaxillaries extends backward from the 
anterior end of the narial aperture to meet the anterior process of the 
nasals, asin A. mississippiensis. The edges of the premaxillaries forming 
the lateral borders of the aperture are turned sharply upward however, 


1923] Mook, Skull Characiers of Alligator Sinense a5 


differing in this character from those of the Florida species. The pos- 
terior end of the narial aperture is situated farther back than in the latter 
species with respect to the level of the premaxillo-maxillary sutures at the 
lateral borders of the skull. 

On the palate the proportion of length to breadth is the same as in 
the Florida alligator. There are two deep pits on each side, one posterior 
to the space between the first and second teeth, the other at the pre- 
maxillo-maxillary suture. The first of these lodged the first mandibular 
tooth, the second, the fourth mandibular tooth. On the median line the 
distance from the anterior end of the premaxillary foramen to the 
anterior border of the skull is equal to the distance from the posterior end 
of the foramen to the premaxillo-maxillary suture. 

The suture between the premaxillaries and the maxillaries differs 
from that of A. mississippiensis. It extends inward and backward from 
the external border, across the pit for the fourth mandibular tooth, to a 
point half-way between the external border and the median line and at 
the level of the space between the first and second maxillary teeth; 
from this point it extends inward and slightly forward half-way to the 
median line, then turns inward and backward and meets the median line 
at the level of the spaces between the first and second maxillary teeth. 
From the median line it extends in a symmetrical direction to the 
opposite border of the skull. The entire suture is therefore wavy in 
outline. 

Each premaxillary contains five teeth, of which the fourth are the 
largest, the third second in size, the fifth third in size and the first and 
second very small. The teeth are evenly spaced. 

MAaxiILLaries.—These bones are short and broad. They are espe- 
cially short along the sutures with the nasals; these sutures are less 
than one-fifth as long as the skull, in contrast to over one-fourth in a 
Florida alligator of slightly larger size. The sutures with the prefrontals 
are exceedingly short, being about one-fourth the length of the maxillo- 
lacrymal sutures. In the Florida alligator they are over one-half the 
length of the maxillo-lacrymal sutures. The sutures with the jugals are 
long, especially in their transverse portions. On the palate the maxil- 
laries are especially short and broad. Comparison with a small skull of 
Alligator mississippiensis may be expressed as follows. 


558 Bulletin American Museum of Natural History (Vol. XLVIII 


A. sinense A. M.N.H. (A. mississippiensis 
No. 23898 A. M. N. H. No. 12572 
Length maxillaries, median line 


215 295 
Length skull 
Lengt illari ian li 
sal h maxillaries, pede ine AMA 500 
Maximum length, maxillaries 
Length maxillaries, median line 
; ae 451 .720 
Maximum breadth, maxillaries 
Length maxillaries, median line 
Se 682 .968 


Length palatines, median line 


Each maxillary contains thirteen teeth, of which the fourth is the 
largest and the third is second in size. The first and second maxillary 
teeth are moderately stout and moderately sharp; the crowns of the 
posterior teeth are all small. The first six teeth are spaced moderately 
and evenly; the last seven are close together. In the largest skull 
studied (A. M. N. H. No. 23898) the first six maxillary teeth have 
separate alveoli and the last seven are lodged in a common alveolar 
groove. In smaller skulls, however, not at present thoroughly cleaned, 
the posterior teeth, or at any rate, some of the posterior teeth, appear to 
have separate alveoli. 

Nasats.—These bones are very short; on the median line they 
occupy considerably less than one-half the length of the skull. _ The 
anterior process, extending forward into the narial aperture, is moder- 
ately broad at its base and very narrow where it joins the process of the 
premaxillaries, extending back from the anterior border of the narial 
aperture. From the posterior border of the narial aperture the nasals 
broaden rapidly to the posterior extremities of the premaxillaries, which 
is the point of their greatest breadth. From this point back they narrow 
gradually to the anterior ends of the naso-prefrontal sutures, back of 
which they narrow rapidly to their blunt posterior extremities, which 
are situated slightly forward from the level of the anterior ends of the 
orbits. The sutures with the maxillaries are relatively short; those 
with the prefrontals are relatively long. 

PREFRONTALS.—The prefrontals are moderately long. Their 
anterior ends extend forward as narrow processes, wedging apart the 
nasals and lacrymals and to a slight exterit the nasals and maxillaries. 
Their contacts with the lacrymals are somewhat elevated into ridges and 
their orbital borders are greatly elevated, to a slight extent even over- 
hanging. Their posterior processes are short and their contacts with the 
frontal are moderately so. 


1923]. . Mook, Skull Characters of Alligator Sinense 559 


LacryMaLs.—The lacrymal bones are relatively short and broad. 
Each suture with the prefrontal extends in a line which is almost parallel 
with the median line but which is inclined very slightly toward the 
median line in the anterior direction. The sutures with the maxillaries 
extend forward and outward irregularly from the anterior ends of the 
prefronto-lacrymal sutures to points near the external boundaries of the 
bones, then slightly outward and backward to join the straight lacrymo- 
jugal sutures. A small but conspicuous oblique ridge extends outward 
and forward from the postero-internal corner of each lacrymal to the 
center of the lacrymo-jugal suture, separating a small, low, smooth 
orbital surface from the more elevated pitted surface. The anterior 
extremities of the lacrymals are situated farther forward than those 
of the prefrontals. In this character the species differs from A. 
MASSISSLPPIENSIS. 

Frontau.—The anterior process of the frontal is smooth and is rel- 
atively short. It does not wedge apart the nasals at their posterior 
extremities but ends abruptly. The interorbital portion is narrow and is 
decidedly uprolled at the edges. The pitting is arranged in such a 
manner in this portion of the bone as to leave a distinct but slight median 
ridge. The posterior portion is relatively broad. The sutures with the 
postorbitals are very short; the suture with the parietal is a simple curve, 
the convexity being directed backward. The frontal is entirely removed 
from the supratemporal fenestra, thus differing somewhat from that of 
the Florida alligator. . 


PostorBITALs.—The postorbitals are slightly longer than broad; 
in A. mississippiensis they are considerably longer than broad. They 
also differ from those of the latter species in occupying a larger proportion 
of the external border of the cranial table. 

SquamosaLs.—The proportion of the breadth to the length of the 
squamosals is greater than in the Florida alligator, and the sutures of 
these bones with the parietal are less symmetrically curved than in the 

latter species. 

ParieTaL.—This bone occupies a considerable portion of the sur- 
face of the cranial table anterior to the supratemporal fenestre as well 
as posterior to them. It occupies nearly a third of the posterior border 
of the cranial table. Its interfenestral plate is relatively broad, com- 
pared with A. mississippiensis. It is somewhat uprolled at its external, 
or fenestral, borders. A low median ridge serves to distinguish it readily 
from that of the American species. 


560 Bulletin American Museum of Natural History [Vol. XLVIII 


Am.Mus.No. 23898 


Fig. 2. Alligator sinense Fauvel. 
Mandible (Amer. Mus. No. 23898). One-half natural size. A, lateral view, left side; B, superior 


view. 

SuPRAOCCIPITAL.—As in A. mississippiensis this bone occupies no 
portion of the border of the cranial table. On the posterior surface of 
the skull it is narrow laterally and deep vertically. It extends down- 
ward to a point about four-fifths of the total distance from the foramen 
magnum to the cranial table, below the latter. 

ExoccipiraLs.—These bones are relatively deep in the vertical 
direction. They comprise small portions of the occipital condyle. 

Bastoccip1rTaL.—This bone comprises most of the occipital condyle 
but not all of it. 

BasISPHENOID.—This bone is not especially characteristic. 

QuapDRATES.—The quadrates are stout in proportion to their length 
and their articular surfaces are unusually broad. 

QUADRATO-IUGALS.—These bones do not differ from those of A. 
mississippiensis in any characteristic way. 

JUGALS.—The jugals are relatively larger than in the Florida species 
and differ from those of the latter in form. Their length is over half the 


1923] ‘Mook, Skull Characters of Alligator Sinense 561 


total length of the skull and they occupy a large portion of the lateral 
surface of the skull. Their posterior processes are slender, as in the 
American form, but the anterior processes are much greater than in the 
latter. The maximum height of each jugal in A. mississippiensis is 
considerably behind the posterior end of the maxillary, near the posterior 
end of the orbit; in A. sinense it is over the posterior process of the maxil- 
lary, near the level of the anterior end of the orbit. The suture with the 
maxillary differs considerably from that in the Florida alligator. 

Paatines.—The palatines of this species differ considerably from 
those of A. mississippiensis and in fact of all other crocodilians. Their 
sutures with the maxillaries extend inward and slightly forward from the 
anterior ends of the palatine fenestrae about half-way to the median line, 
then turn directly forward or with a gentle curve whose concavity faces 
outward to the level of the eighth maxillary teeth, then either directly 
inward (A. M. N. H. No. 23899) or inward and forward (A. M. N. H. 
No. 23898) to meet at the median line. The interfenestral portions are 
very broad and their fenestral borders are nearly parallel. Their ptery- 
goid borders form a nearly straight line. 

Prerycoips.—The pterygoids are not especially distinctive. They 
are considerably arched and the ridge back of the internal narial aperture 
is unusually high. The aperture is divided, as in the Florida species. 

EcropTERYGOIDS.—These bones are very short antero-posteriorly 
and are very stout. 

Tue Manprsite.—The mandible is relatively broader in proportion 
to its length than in the Florida alligator, the rami diverge at a relatively 
sharper angle, and the individual elements are all stouter. The symphysis 
extends back to the level of the fifth mandibular teeth, contrasting with 
the short symphysis reaching only slightly back of the level of the third. 
in the American species. The splenials do not enter the symphysis but 
extend forward within a relatively minute distance of it; this is in con- 
trast with the wide spaces between the anterior ends of the splenials and 
the symphysis in A. mississippiensis. 

The dental borders occupy about half the length of the mandible, 
an unusual condition among the living crocodilians, probably approached 
only by Jacare latirostris. There are nineteen teeth in each ramus. The 
fourth is the largest of these and the thirteenth is second in size. The 
first teeth are moderately large and are widely separated from the smaller 
second, which are widely separated from the equal-sized third, which are 
moderately far from the fourth. Back of the fourth are seven small 
teeth close together. Back of the thirteenth the short blunt teeth are 


562 Bulletin American Museum of Natural History [Vol. XLVIII 


all close together. The variation in the size and the arrangement of the 
teeth is quite different from that in A. mississippiensis. 

The external mandibular foramen is relatively high in proportion to 
its length. Its surangular border is longer than in the American form. 
The internal mandibular foramen is relatively small. 


DISCUSSION 


This species resembles the living Florida alligator very closely in 
some respects but differs from it quite markedly in others. 

It has certain resemblances to the various species of Jacare which are 
absent in A. mississippiensis. These resemblances are offset by other 
more fundamental characters, however, and there is no evidence of close 
relationship. 

The nearest approach to the structure of this species is to be seen 
in the Miocene Alligator thomsoni, recently described by the writer.! 
The species may therefore be considered more primitive than A. mis- 
stosippiensis, though it has some specializations absent in the latter. A. 
thomsoni approaches A llognathosuchus polyodon from the Bridger, which in 
turn approaches Allognathosuchus heterodon of the Wasatch in a number 
of characters, so that we may consider the following as a logical mor- 
phological sequence: Allognathosuchus heterodon (Eocene); Allognatho- 
suchus polyodon (Eocene); Alligator thomsoni (Miocene); Alligator 
sinense (Recent); A. mississippiensis (Recent). This does not neces- 
sarily indicate a line of descent, but it does indicate that Alligator sinense 
serves partially to bridge the wide structural gap between A. misszssip- 
piensis and the earlier Tertiary crocodilians. A. mississippiensis has 
been found in Pleistocene deposits. We await the discovery of A. sznense, 
or a very closely related form, in Pleistocene, or even Pliocene, deposits. 


1Amer. Mus. Novitates, No. 73. 


56.9(1183:51.3) 
Article XVII.—NEW FOSSIL MAMMALS FROM THE PLIOCENE 
OF SZE-CHUAN, CHINA! 


By W. D. MarruHew AnD WALTER GRANGER 


The following is a preliminary notice of a collection secured during 
the winter of 1920-1921 by Mr. Walter Granger, paleontologist of the 
Third Asiatic Expedition sent out by the American Museum, the Ameri- 
can Asiatic Association, and Asia Magazine, under leadership of Mr. Roy 
C. Andrews. The locality is a series of pits or fissures at the village of 
Yen-ching-kao in the vicinity of Wan-hsien, province of Sze-chuan. 
These pits have been worked by the natives for many years for the 
Chinese drug trade. Mr. Granger’s account of their occurrence and 
stratigraphic observations and more extended descriptions of the fauna 
will be published later, this notice serving to place certain interesting 
novelties upon record. 

The age of the fauna is provisionally placed as Upper Pliocene on 
account of the abundance of Stegodon remains and absence of any 
higher type of Proboscidean, but its final correlation is left open for the 
present. 

The Chinese fossil mammals described by Owen in 1870? came from 
“a cave near the city of Chung-king-foo in the province of Sze-chuan.”’ 
Chung-king is on the Yang-tse-kiang above Wan-hsien, about one 
hundred and forty miles distant in an air line. If correctly reported by 
the finders, Owen’s specimens could hardly have come from the Yen- 
ching-kao pits; but it is apparently the same fauna, or at all events of 
similar facies and doubtless the same conditions of preservation. Pos- 
sibly the Chinese informants of Consul Swinhoe, who sent the fossils to 
Owen, misled him, unintentionally or deliberately, as to the locality. 
This point will be further discussed at a later date. Owen regarded the 
fauna as Pliocene and described the following species: 


Stegodon orientalis. Parts of molars. 

Rhinoceros sinensis. Parts of 4 upper and 4 lower molars. 
Tapirus sinensis. Parts of 3 upper and 4 lower teeth. 
Chalicotherium sinense. Part of an upper molar. 

Hyena sinensis. Canine, 2 premolars. 


> eyey shane of the Asiatic Expeditions of The American Museum of Natural History, Publica- 
tion No. 15. 
2Quar. Journ. Geol. Soc., London, XXVI, pp. 417-436, Pls. xxv11-xxrx, 


563 


9S 


‘OZIS [BINJEN ‘[[NYS edAjoou jo ajvlVq “SZOST “ON “SN “AOU “WOMGO sisuawis soladomyy *T “BIT 


1923] Matthew-Granger, New Fossil Mammals from Sze-Chuan, China 565 


The first three genera are among the most abundant types of large 
animals in the Yen-ching-kao pits. Chalicothertwm and Hyena are rare. 
Owen’s descriptions and figures accord very well with some of the species 
in our collection, so that we have referred them to his species, whether 
or not later investigation proves them to-be exact topotypes. 

Koken in 1885! described a collection secured by von Richthofen, 
apparently from the trading junks of the Yang-tse-kiang and under- 
stood by him to have come from far up the river in “caves in Yun-nan.”’ 
Whether this was the real locality remains to be verified; one has the 
impression from the reading of von Richthofen’s letter, quoted by Koken, 
that the traveller himself suspected that the locality might not have been 
correctly stated. It is certain at all events that the major part of 
Koken’s collections, like Owen’s, represent substantially the same faunal 
facies, and they seem to agree as to species, in part at least, with our 
collections. Koken also distinguishes an older fauna of supposed Lower 
Pliocene age, including Hipparion, Camelopardalis, Palezomeryz, etc., 
which is more extensively represented in Schlosser’s later collections, and 
is probably substantially the same fauna as the fine collections secured 
recently by J. G. Andersson? and now being studied by Professor Wiman.: 

Schlosser in 1903% described a large collection secured by Dr. 
Haberer for the Munich museum, and revised the work of Owen, Koken 
and other previous writers. He concluded that Owen’s fauna, except 
Stegodon, and most of Koken’s material, was of Pleistocene age. There is 
no doubt, however, that the Stegodon is coeval with the rest of the fauna 
in Granger’s collection, and one may assume that it was probably so in 
the Owen and Koken collections. Schlosser’s material belonged mostly 
to the older Pliocene fauna distinguished by Koken and came from 
localities farther to the north. 

The collections from Sze-chuan described by Professor Matsumoto 
in 19154 may have come in large part or all from the Yen-ching-kao 
pits; he does not state any exact localities, but the correspondence of the 
fauna is evident. Matsumoto divides the material studied by him into 
two faunas, one found in brown clay and more strongly petrified, the 
other in cave-loam, feebly fossilized and the teeth strongly colored. The 
former includes Stegodon, Acerathérium hipparionum, Proboselaphus 
watasei and liodon, Bibos geron and two unnamed species of Buffelus. 
The latter includes Hyena ultima, Rhinoceros sinensis and R. plicidens. 

1Koken, E. 1885. ‘Fossile Saiigethiere aus Chinas.’ Pal. Abh., III, Heft 2. 
2Andersson, J.G. 1922. Bull Amer. Mus. Nat. Hist., XLVI, pp. 727-737. 


%Schlosser, Max. 1903. ‘Fossile Satigethians Chinas.’ Abh. k. bayer. Akad. Wiss., XXII, Abt. 1 


*Matsumoto, H. 1915. Science Reports, Tohoku Imp. Univ., Second Ser. (Geol.) III, No. 1, pp 
1-28, Pls. 1-x. 


ae 


individual with milk dentition, the last molar not yet emerged and the second 
unworn. Natural size. 


Fig. 2. Rhinoceros sinensis. Front of skull and jaws, No. 18626, young 


566 


1923] Matthew-Granger, New Fossil Mammals from Sze-Chuan, China 567 
He regards the first as Upper Pliocene, the second as Lower Pleistocene. 
In our collections the Stegodon material was limited to certain pits 
occurring low down on the slopes of the mountain valley, and did not 
occur in pits higher up on the mountain; but we are provisionally dis- 
posed to regard this as a matter of limitation of range, and to consider 
the material from all the pits as of substantially the same geologic age. 

Professor Matsumoto’s researches! upon this and related faunas 
have been of peculiar value to us as a guide in searching for the probable 
affinities and identifications of our material. 

The Granger collection includes skulls, jaws and numerous parts of 
jaws of Stegodon and Tapir, incomplete skulls and many jaws of Fhi- 
noceros, a tooth of Chalicotherium and a large series of other animals, 
including a large bovid, smaller antelopes and two or three deer, a pig, 
various carnivora and a very abundant rodent. 

The following notes upon certain described species are a necessary 
preliminary to the discussion of our new collections. 


Stegodon orientalis Owen 


Schlosser regards this species as identical with S. cnszgnis of India, . 
basing the reference upon the fragmentary teeth described by Owen. 
Matsumoto regards it as distinct, upon the evidence of the referred 
material which he describes and figures. The Yen-ching-kao material 
includes a fairly complete adult skull, two young skulls, a series of palates 
and lower jaws and many teeth. It should enable us to estimate the affi- 
nities of the species more exactly when it has been cleaned up and studied. 


Rhinoceros sinensis Owen 


Schlosser in his masterly review recognizes this species as valid 
and considers it most nearly related to platyrhinus of the Siwaliks and to 
the Atelodine group of Pleistocene and modern times. It was, however, 
practically absent from his Chinese collections at Munich, which seem 
to have come mostly from the “red clays” of Shan-si, Shen-si and Sze- 
chuan, but as they were not observed in place the real character and age 
of the formations remain doubtful. Probably they are chiefly Lower 
Pliocene. Most of the rhinoceros teeth he refers to R. habereri, related in 
his opinion to R. palxindicus and thus to the typical modern Rhinoceros 
of India. 

Matsumoto appears disposed to assign R. sinensis to the Teleoceras 
group; but if our material be correctly referred, the affinities of Owen’s 


1Matsumoto, H. 1915, loc. cit.; 1921, idem, V, No. 3, pp. 75-91, Pls. x111—-xtv. 


Fig. 3. Stegodon orientalis Owen. No. 18630, adult skull, laterally crushed. 
One-fifth natural size. 


568 


Fig. 4. Stegodon orientalis. No. 18640. Young skull and lower jaw, uncrushed. 
Three-tenths natural size. 


569 


18630 


Fig. 5. Stegodon orientalis. No. 18630. Palate of adult skull. One-half 
natural size. 


570 


[AS 


‘OZIS [VANJVU J/RY-IUG “][NNS SUNOA JO MOIA [LABILY ‘OFOST ON ‘syPzuatuo uopobagy *9 “BIW 


Or9OsI WV 


572 Bulletin American Museum of Natural History [Vol. XLVIII 


species must be with R. unicornis. At all events, the Yen-ching-kao 

rhinoceros is a near relative of the typical modern Indian rhinoceros. 

: The type of R. sinensis consists of parts of upper and lower teeth, 
probably of different individuals. We designate the following as a 
neotype. 

Neotype.—A crushed skull, Amer. Mus. No. 18628. 

CHARACTERS.—A large nasal horn. No clear indications of a second horn. 
Occiput apparently rather posterior in position. Teeth moderately hypsodont, 
slightly less so than in R. indicus. Premolars 130; length of molars, 160; p! small, 
deciduous. Both external ribs prominent on p 2“, posterior rib weak on m!, wholly 
absent on m 2%, the anterior rib prominent on all three molars. Crochet prominent 
on p?-m3, doubled on p*-m!; crista rudimentary except on p?, where it is prominent. 
No antecrochet save as an obscure swelling. Postfossette on p*-m! only when con- 
siderably worn. The two inner cones of p? strongly twinned, slight twinning on p*. 

The above characters are shown on the neotype and in Owen’s 
type, so far as it goes. A number of incomplete skulls, palates and upper 
jaws and teeth show more or less variation in the external ribs, details 
of the crochet, crista and posterior fossette, but in all it may be said 
that the crochet is strong and more or less reduplicate, the crista and 
antecrochet weak or absent, the postfossette moderately developed, the 
external ribs variable, the teeth subhypsodont, premolars considerably 
smaller than molars, but p? unreduced and only p! vestigial, molars sub- 
equal, m’ smallest of the three. 

The characters of the teeth in the neotype are strongly suggestive of 
affinity to the Indian and Javan rhinoceroses, combining peculiarities of 
the two; the referred specimens bring it on the whole nearer to the 
Indian species. None of our specimens has the premaxilla preserved 
sufficiently to demonstrate the presence or absence of upper incisors; 
but the cheek teeth are nearer to the true rhinoceroses than to Atelodus 
and the proportions of the anterior end of the lower jaw agree best with 
R. indicus. The neotype skull is too badly crushed to be decisive as to 
the characters of the occiput, and no other specimens show this region. 
The position of the horn, on the nasals but not quite terminal, is like .R. 
indicus and unlike Atelodus. 

In the skeleton, including especially the length and proportions of 
the limb bones and feet, all the Yen-ching-kao rhinoceros material agrees 
fairly closely with R. indicus. 

Among the numerous fossil species described we find certain Indian 
and western Eurasian forms that may be nearly red, especially R. 

platyrhinus, palxindicus, sivalensis. 

The species described by Koken and Schlosser are founded upon 
tooth distinctions, of which the constancy is doubtful, to judge from our 


1923] Matthew-Granger, New Fossil Mammals from Sze-Chuan, China 573 


collections. Both Schlosser and Matsumoto, in sorting out the material 
described and assigning it to various species and horizons, have attached 
great importance to the degree of fossilization and the quality of the 
matrix. Wide variation is shown in our collections in this respect, from 
almost unaltered and recent-appearing to thoroughly fossilized teeth and 
bones in hard clay matrix. But we cannot associate these differences at 
present with any faunal distinctions and it is probable that they are due 
chiefly or wholly to the accidents of location of the specimen, whether in 
the path of mineralizing waters or protected from their action. The 
present species can be satisfactorily placed as to its relationships, but not 
as to its nomenclature and synonymy. 

AFFINITIES.—R. sinensis is clearly excluded from Aceratherium, 
Teleoceras or Celodonta and apparently from Opsiceros. Affinity with 
Ceratotherium is not especially indicated. All the positive evidence goes 
to show that it is a near relative of the true Rhinoceros, but specifically 
distinct from either the Indian or the Javan species, nearer perhaps to 
the former. | 


Tapirus sinensis Owen 

Besides the teeth described by Owen, Koken figured a number of 
teeth and Schlosser records two from the Haberer collections. The 
latter were obtained at I-chang, a hundred miles down-river from Yen- 
ching-kao; Koken’s material is said to be mostly from caves in Yun-nan 
or other southern provinces. They are all referred to the Pleistocene by 
Schlosser. It would appear that Owen’s species is closely related to T. 
indicus, perhaps doubtfully distinct. Our tapir material consists of 
skulls, jaws, etc., of a very much larger species described below. It is 
not close to the modern Malayan tapir; whether the genus is distinct 
remains to be determined. 7’. sinensis is not represented in the Granger 
collection. 


' Chalicotherium sinense Owen 
Owen’s collection contained one upper molar. Koken described a 
supposed p, (m, according to Schlosser). Our collection contains a single 
lower molar, No. 18453, probably m;. It affords no especial light upon 
the relations to the Indian C. sivalense. 


Hyena sinensis Owen 
No. 18392, upper and lower jaws, is referred to this species; also 
Nos. 18395-7, isolated parts of jaws. 


574 Bulletin American Museum of Natural History [Vol. XLVIII 


Owen distinguishes the species as larger and more robust than the 
modern H. crocuta, much larger than the Asiatic striped hyzena, and as 
allied to the African and not to the Asiatic species, “unlike the European 
cave hyena.”’ 


NEW GENERA AND SPECIES IN THE YEN-CHING-KAO COLLECTION 
Spalacide 
Rhizomys troglodytes, new species 


Typr.—No. 18408, skull and jaws. 

ParatyPes.—Nos. 18401-18417, a series of skulls and jaws, some with parts of 
skeleton associated. 

DIsTINCTIVE CHARACTERS.—(a) Size large, length of skull incisors to condyles = 
77-85 mm.; (b) skull rather long and narrow, postorbital crests contracting sharply 
behind the orbits to a long and well-marked sagittal crest; (c) infra-orbital foramen 
sub-triangular, the maxillary crest in front of it and plate beneath extended upward 
on the side of the muzzle almost to its upper surface; (d) nasals long, narrow, wedge- 
shaped, tapering backwards almost to a point; (e) squamosals fail to reach the sagittal 
crest superiorly or the postorbital constriction anteriorly; (f) occiput strongly sloped 
forward; (g) posterior nares narrow and contracted transversely; (h) bulla somewhat 
flattened inferiorly, strongly convex anteriorly, culminating in a ridge directly behind 
the posterior lacerate foramen, slightly reflexed on the anterior inner border against 
the basisphenoid; (7) carotid foramen lying close behind the basisphenoid-basi- 
occipital suture and the bulla extending a considerable distance in front of it; (7) 
bulla strongly reflexed posteriorly upon the surface of the paroccipital process; (hk) 
inferior surface of auditory meatus strongly concave both ways, without any longi- 
tudinal ridge, and the opening large and flaring; (J) incisors strongly convex, the 
points of the upper pair directed somewhat backward, the anterior faces of the lower 
pair strongly flattened, of the upper pair, moderately so; (m) first upper molar some- 
what, and first lower molar considerably reduced and m! wearing to a lower grinding 
plane than the others; (m) third upper molar unreduced, approximately equal to m? in 
size, the posterior portion of the third lower molar correspondingly enlarged and 
broadened. 


Of the above characters, Nos. c, d, e, f, g, h, 7, k, 1, m, and n appear to 
be characteristic of Rhizomys proper as against Nyctocleptes. Nos. a and 
j resemble the latter, while 6 is peculiar. The affinities of the species are 
thus clearly shown to be with the much smaller Rhizomys of China, 
although in size and one or two characters associated with size it is sug- 
gestive of the large Malayan bamboo-rat, Nyctocleptes sumatrensis, 
ete., which it fully equals in size. 

The above comparisons were made with modern skulls from South 
China collected by Mr. Andrews and a series of Malayan skulls in the 
National Museum loaned through the courtesy of Dr. Gerrit 8. Miller 
and Mr. J. W. Gidley. 


A.M. 184086. TYPE 


Fig. 7. Rhizomys troglodytes. Type skull, No. 18408. Natural size. 


575 


A.M.184i6 


Fig. 8. Rhizomys troglodytes. Side view of skull and jaws, No. 18416. Outer 
and top views of lower jaw, No. 18413. Both natural size. 


576 


Fig. 9. luropus fovealis. No. 18385, type. Lower jaw, outer and crown view 
of teeth. Natural size. 


577 


M.18588 


A 


view: 


outer and top 


jaw, 


Lower 


388. 


18 


Yo 


i 
4 


ws 


s foveal 


Aluropu 


Fig. 10. 
Natural size. 


578 


1923] Matthew-Granger, New Fossil Mammals from Sze-Chuan, China 579 


Ursidze 


#luropus! fovealis, new species 


Typr.—No. 18385, right lower jaw with p, to m3, also left m3 of the same in- 
dividual. 

DisTINCTIVE CHARACTERS.—The teeth resemble those of 4. melanoleucus as 
figured by Lankester, 1901, except in the following particulars: the protocone of p4 
is distinctly higher than the anterior and posterior cusps; m; retains more of the 
normal canassial construction, the anterior end being less quadrate, protoconid 
larger, paraconid more advanced and the whole tooth is relatively larger; m2 and mg 
are broader, though not longer. Bardenfleth’s figure in 1913 of the teeth of a speci- 
men also in the British Museum agrees much more closely in proportions with our 
specimen and, if both are accurate, would suggest that the differences noted above 
are individual rather than specific. However, as it seems unlikely that a species of 
the Carnivora would persist unchanged from the Pliocene to the present day, it 
appears better to regard the species provisionally as distinct. Three other speci- 
mens, Nos. 18386-8, are referred to the species. Two of them show the unworn m; 
in broken lower jaws. The third is a lower jaw with mi. complete, so much larger and 
more robust than the type that we hesitate to include it under the same species. 

The affinities of Hluropus appear to be with Hyznarctos, as has been 
observed by Lydekker,? Winge* and other writers. Its systematic posi- 
tion appears to be clearly in the family Urside,‘ although of a distinct 
subfamily from the true bears. Bardenfleth* has presented the evidence 
for this view very clearly. The occurrence of #luropus almost com- 
pletely modernized in the Pliocene, if these deposits are in fact Pliocene, 
contemporary, or nearly so, with Hyznarctos, shows that it cannot be a 
direct descendant, although Hyznarctos seems to be in general structur- 
ally ancestral. 

Lydekker‘ has reported a species of Hyznarctos from the collection 
of Chinese fossils described by Owen. Schlosser? gives reasons (not very 
convincing) for regarding it as Pleistocene and notes an incisor and m; in 
the Haberer collection at Munich, but doubts their pertinence to this 
genus. They approach the amphicyons, differing from Hyznarctos in 
quite an opposite sense from the present species. 


1#Mluropus = Zluropoda, for the purists. 

*Lydekker, R. 1896. ‘Geographical History of Mammals,’ p. 321. 

3Winge, H. 1896. ‘Jordf. og. nulev. Rovdyr (Carnivora) fra Lagoa Santa,’ p. 62. These are 
probably by no means the earliest authorities, for the comparison is too obvious to have escaped notice. 
It is at least implied in Flower’s arrangement of the genera in the ‘Catalogue of Mammals, Mus. Roy. 
Coll. Surgeons.’ 

4As placed by most authors. Osborn in the ‘Age of Mammals,’ following Lankester’s authority, 
places it in the Procyonide. 

5Bardenfleth, K. S. 1913. ‘On the Systematic Position of Hluropus melanoleucus.’ Mindesk. f. 
Japetas Steenstrup, Kobenhavn. 

6Lydekker, R. 1885. ‘Cat. Foss. Mam. Brit. Mus.,’ Part I, p. 157, fig. 23. 

7Schlosser, M. 1903. ‘Fossile Saiigethiere Chinas,’ p. 23. 


ry 


580 Bulletin American Museum of Natural History [Vol. XLVIII 


Fig. 11. Ursus kokeni. No. 18384, type. Lower jaw, outer and top views. 
Natural size. 


Ursus kokeni, new species 


Type.—No. 18384, a lower jaw with mj and adjacent alveoli. 

DIsTINCTIVE CHARACTERS.—Jaw very short and deep as in the sun-bear U. 
malayanus, but size large, comparable with U. arctos; m; narrow and long, lacking the 
metastylid cusp of U.malayanus; m» rather short and wide, wider posteriorly than 
anteriorly. 


It is very likely that the molar figured by Koken as U. aff. japonicus 
is of this species. 


Arctonyx rostratus, new species 
Type.—No. 18393, skull lacking the zygomatic arches and with damaged 
teeth. 

PaRATYPES.—Nos. 18394, skull, and 18382, 18383, lower jaws. 

Distinctive CHaractrers.—Length of skull, premaxille to condyles, 148 mm.; 
sagittal crest narrow, distinct; pt absent, p$ larger than in A. collaris and more 
clearly two-rooted, the diastema behind p? greater than length of p?; p* larger with 
inner cusp better developed and more antero-internal; m! larger, broader and more 
quadrate in form; auditory meatus and posttympanic process broad, massive and 


T8¢ 


‘OZIS [RINJVNT “MOTTA doy [Nys edAT, “GEST “ON “sNp0UZ804 Thuopsy “ST “Si 


S6“e8l WV 
adAt 


68 


‘OZIS [BANJEN “AMOTA [wyLyEd “EGERTON ‘WHYS odAy, ‘snpo.ysor rhuojsp “ET “SUT 


CBeCRI WV 
3dAL 


1923] Matthew-Granger, New Fossil Mammals from Sze-Chuan, China 583 


A.M.18381 


Fig. 14. Arctonyx rostratus. No. 18381° Lower jaw, outer and top views. 
Natural size. — 


flattened, occiput broader at the base. P3 and p4are more robust than in A. collaris 
and there is no diastema between them; m; and m, are considerably larger and more 
robust, with the cusps more conical in form. 


This species differs but little from Milne Edwards’ drawing of A. 
collaris. The differences from a specimen obtained in the mountains 
of Shensi (with which the above comparisons are made) are more con- 
siderable but may also be reduced in essence to the greater size and robust- 
ness of the fossil species and the somewhat higher degree of specializa- 
tion of its modern relative. 

The construction of the teeth in Arctonyz is essentially the same as 
in Meles, to which it is rather nearly related, in spite of the wide difference 
in proportions. 


584 Bulletin American Museum of Natural History [Vol. XVLIII 


Cyon antiquus, new species 


Typr.—No. 18389, a pair of lower jaws. No. 18583, parts of crania, limb bones 
and vertebre of a canid of appropriate size and characters are provisionally re- 
ferred to the species. 

Distinctive CuaractTers.—Metaconid distinct upon m; and m2. Teeth slightly 
more robust than in our specimens of C. alpinus, more decidedly larger and heavier 
than in C. javanicus. 

There is some question as to the validity of this species, as Mivart 
in his ‘Monograph of the Canide’ figures the metaconid as present on 
m; of both species of Cyon, although it is absent in our specimens referred 
to them. It may therefore be a variable character. 


Fig. 15. Cyon antiyuus. Lower jaw, No. 18389, type specimen, top and outer 
views. Natural size. 


Felis aff. tigris Linnzeus 


No. 18624, a complete skull and jaws; also a part of skull with lower 
jaws associated, and a number of jaws and limb bones more or less 
associated, are referred here. In comparison with a series of skulls of 
the modern tiger we have been unable to recognize any constant distine- 
tions for the fossil form, and therefore refer it to F. tigris, although a more 
minute and exhaustive comparison might very well show valid specific 
distinctions. 

There is no doubt, at any rate, that it belongs nearer to the tiger 
than to the lion and that it is quite distinct from F. cristata of, the 
Siwalik Pliocene. 


¢c8¢ 


aZIS [BANYVU FTeY-ouQ ‘saul puv [NYAS ‘FZOST “ON “sb “Be syay “OT “SLT 


yz9Bl wv 


Fig.17. Viverrasp. Lower 
jaw, No. 18390, top and outer 
. views. Natural size. 


Fig. 18.  Bunopithecus  sericus. 
No. 18534, type, lower jaw frag- 
ment, top and outer views. Nat- TYPE 
ural. size. 


A.M. 18466. 


Fig. 19. Rhinopithecus tingianus.. No. 18466. Type skull, side view. 
Natural size. 


586 


Fig. 20. Rhinopithecus tingianus. No. 18466. Type skull, top and palatal 
views. Natural size. 


587 


588 Bulletin American Museum of Natural History —— [Vol. XLVIII 


Bunopithecus sericus, new genus and species 


Typr.—No. 18534, a lower jaw with mo-; on the left side. 
GENERIC DistTINcTIons.—Jaw and teeth much as in Hylobates except for greater 
width of molar and large size of hypoconulid on m2 and m;. 


The heels are slightly broader than the anterior half of the teeth and 
the hypoconulid is as large as the entoconid on both teeth. In the gib- 
bon it is small on mz, and absent on m3; m3; is narrower and smaller than 
mz in the gibbon but broader in Bunopithecus. 

The species is about the size of the hoolock. 


Rhinopithecus tingianus, new species 


Typr.—No. 18466, a skull, immature, retaining the milk premolars, and the 
last molar not yet emerged. 

PaRATYPES.—Nos. 18467—-9, upper and lower jaws. 

DIsTINCTIVE CHARACTERS.—Larger and more robust throughout than R. rozel- 
lane. Size about as in R. bieti but with much smaller teeth. 

The modern langhur monkeys of this genus have a somewhat ill- 
defined range in northwestern and southwestern China and eastern 
Thibet. This species is typical of the genus, not in any marked degree 
primitive or synthetic in generic position. It is named in honor of Dr. 
V. K. Ting, the able and progressive director of the Geological Survey 
of China. 


Tapirus (Megatapirus) augustus, new species 


Typr.—No. 18433, skull and jaws. 

PaRATYPES.—Nos. 18428, 18431, and 18432, skulls, the latter two with lower jaws. 

DistincTIVE CHARACTERS.—Teeth and skull about one-fourth larger lineally 
than 7. indicus or terrestris and almost as much exceeding 7’. sinensis in size. Anterior 
premolars more molariform than in 7’. indicus, the inner cusp and cingulum much 
more developed, especially in p! which in 7. augustus is wider than long (?). Skull 
very short and deep, the vomer higher and thicker than in 7’. indicus, much more 
so than in T. terrestris. 

This species far exceeds in size any living tapir of which we can find 
record and differs so considerably in proportions of skull and details of 
tooth construction that we consider it provisionally as representing a 
distinct subgenus. All of our tapir specimens appear to be referable to 
this gigantic species. T’. sinensis is not present here, although the speci- 
mens provisionally referred to it by Schlosser may be T. augustus. Al- 
though resembling 7’. terrestris in the relative complexity of the anterior 
premolars, it appears in the skull to be an exaggerated type of 7’. indicus, 
deeper and shorter with more massive vomer, high-set nasals, ete. 


68° 


‘OZIS [RINYVU SYIY-OM], “MOIA Opts qnys odkq ‘eePST ‘ON ‘snysnbnv (snudvwbayy) snudvy “1% ‘St 


Sevsl WV 
adKy 


06S 


“OZIS TRINJVU SYIY-OMT, “[[NYyS odAy Jo MOIA [RYLTRG “EEFRT ‘ON ‘snpsninv snudny °z7zZ ‘By 


ESr7sl WV 
adkt 


T6S 


‘ozs [BINyeU syysy-OMy, “Mota doy ‘eddy Jo Mel IAMOT “EEFST ON ‘snysnbnv snurdv], EZ “SLT 


SOVBI NV 
adhky 


66S 


“O2I8 [BAINJVU SY}JG-OMT, “META optsqno ‘adAy Jo Mul IOMOTT “EEFRT ‘ON ‘sngsnbnvsnudvy ‘FZ “By 


PadAd Seve WY 


> 


£69 


“soIn3y §,UeMG wos gy 


61X92 IZX 8% 12X88 82X88 

61X&z 1ZX 8% TES 1ZX 6Z 9%X G8 LZX¥E Gs # 

FICS SIXFo 61 X 1z Lexie CZXE PEXZE ‘a ,, ” 

61 X1Z LIXGZ 61 X&Z 9ZXTE GZ XIE rd, ” 

61 X0z 61 XFZ. 61 XZ 61X9Z POXIE CeXUS sa, ed ” 

GT X0Z SLXFZ LIX&Z 6ILXFE td jo siojouleiq 
OL 68 LOT : FOL eT SIBjOUr —,, 
981 . 2ST 191 661 tur-2d 300} JOMO'T 

92 X SZ 0g X9z 11EX LZ OrXFE SexrE OFX FE eUl ,, ” 

9ZX FG 0fX 2z 0$ X9z 1gX6z OFX FE er xge Or x9g gu ,, ” 

FOX6I GZXFG SZ XFS 18X1€ orxT¢ sexs yur ,, i 

LZX61 93X22 88X08 68X 6Z SEX 6z yd ,, ” 

GZX61 LZX&Z GZ XZ 1gX&Z 1EX6Z 88 X 6% 98X08 ed ,, fs 

TOX8T ZX 02 GZX1Z FE X8Z BExXz TeX 22 2d ,, 9 

LEE FIX8I 61 X0z €TX9Z GZX LZ "13 Xd-8 ,d jo siojyoueig 
OL 6L 101 S01 101 g-1Ul SIBjOU ,, 
CPI 1ST “a PLT . 902 SIZ FIZ © eul-d ‘Y}00} todd) 

08% “A WPM, 
06 LIP ‘ 0&8$ “A 08S “A sejApuoo-xuid y4Su9z [NYG 


S619E “WW *V | 9OTFT “IN “V | “P®°V “FIM 
Polchse “[\snupivjpow +7} wshoy 7 sisuauis ", | CEP8LT “ON | S8ZPr8T “ON SEP8T “ON 


SLNANAYOASVA] 


594 Bulletin American Museum of Natural History [Vol. XLVIII 


A.M.18445 | 


Fig. 25. Sus sp. ef. hyotherioides Schl. No. 18445, skull and jaws. One-third 
natural size. 


Sus compare S. hyotherioides Schlosser 


This is a species about the size of the modern Potamocherus but p* 
is larger and more complex. 


Proboselaphus watasei Matsumoto 


Several incomplete skulls, numerous jaws and skeletal bones are 
provisionally referred to this species. If the reference be correct, it would 
appear to be rather nearly related to the nilghai (Boselaphus) of India. 


Bibos geron Matsumoto 


This species was based upon parts of upper and lower jaw. We refer 
to it a series of skulls, skeletons, upper and lower jaws, etc., of which No. 
18465, a fairly complete skull, is selected as neotype. The affinity to the 
gaur and other species of this group is shown especially in the character 
of the horns, flattened, angulate, arising from the vertex of the skull and 
sweeping downward and upward, but not backward. 

Koken and Matsumoto record Buffelus and Bison upon the evidence 
of teeth. While there is a very large series of Bovine skulls, jaws, ete., 
and a considerable variation in the characters of the teeth, we have not 
seen among the skulls and horns any evidence of any other true bovine 
type than Bibos. Whether the supposed distinctions among the teeth 


Fig. 26. Bibos ? geron Matsumoto. No. 18465, skull, top and side views. 
One-sixth natural size. 


595 


596 Bullelin American Museum of Natural History [Vol. XLVIII 


A.M.18465 


Fig. 27. Bibos? geron. No. 18465, palatal view of skull. One-sixth natural size. 


are really constant characteristics of the several genera of Bovine re- 
mains to be verified by more careful comparative study of the materials. 

It is quite clear, however, that there are two distinct types of 
Bovine represented in the foot material; one with extremely short meta- 
podials, the other of larger size and with metapodials somewhat longer 
than in the American bison. 


AFFINITIES OF THE YEN-CHING-KAO FAUNA 
The following faunal list is a preliminary one and may be considerably 
modified and better ‘defined by further study. It will serve, however, 
to show the general character of the fauna. 


1923] Matthew-Granger, New Fossil Mammals from Sze-Chuan, China 597 


PRIMATES 
Bunopithecus sericus ef. Hylobates Malaysia 
Rhinopithecus tingianu “ Rhinopithecus | W. China 
FERE 
Ursus kokeni U. malayanus Malaysia 
' Aluropus fovealis “ A. melanoleucus W. China, Thibet 
Arctonyzx rostratus A. collaris fs zs 


Cyon antiquus “ C. alpinus v a 
Viverra sp. “ Viverra sp. div. ov * 
Hyzna sinensis ** H. crocuta Africa 
Felis aff. tigris “ F, tigris India, E. Asia 
GLIRES 
Rhizomys troglodytes “ R. sinensis S.-W. China 
Lepus sp. 
PROBOSCIDEA ; 
Stegodon orientalis ‘““ EHlephas India 
PERISSODACTYLA 
Tapirus augustus “ Tapirus Malaysia, tropica' America 
Chalicotheriwm sinense 
Rhinoceros sinensis “ R. indicus India 
ARTIODACTYLA 
Bibos geron “ B. gaurus India 
?Bos (cf. grunniens) “ B. grunniens W. China, Thibet 
? Antilope 
?Proboselaphus watasei  B. nilghai India 
Gazella “ G. gutturosa Thibet 
Cervus sp. “ C. wapiti, ete. Central Asia 
Sus sp. ef. hyotherioides ‘“* Sus sp. div. Malaysia 


The above list is remarkable, as a cave or fissure fauna, for the 
seareity of rodents (other than Rhizomys) and small carnivora. While 
the remains of large animals are abundant and varied, the bamboo-rat 
is the only rodent, except for a single hare jaw, and no small mustelids 
or viverrids appear.' It is no less remarkable that no trace of Equide 
is found in it, nor of camels, giraffes, typical Canidz or macherodonts. 
This, coupled with the abundance of tapirs and deer, may point to a 
heavily forested condition. The abundance of Stegodon and entire 
absence of Hlephas and the presence of Chalicotherium are the only ob- 
served indications of Pliocene age; for the most part the fauna appears 
to be quite closely related to modern species and might well be con- 
sidered Pleistocene. The faunal affinities appear to be principally Chinese, 
partly Malayan, not much Indian; there is nothing especially suggestive 
of North American or of Siberian affinity. A more careful comparison 


a rag second season (1922-3) Mr. Granger reports finding good material of small carnivora.— 
. -« AVL. 


598 Bulletin American Museum of Natural History [Vol. XLVIIT 


and identification of the whole fauna, especially of the smaller ruminants, 
might show a clearer differentiation from the modern species than we 
have observed in this preliminary study, but could hardly alter materially 
the geographic and environmental affinities of the fauna. It is such a 
fauna as one might expect to find in the valleys of southwestern China ét 
any time before the appearance of civilized man, and under climatic con- 
ditions similar to those now prevalent. The effect of the clearing and 
cultivation of the valleys and the lower slopes of the hills by man has been, 
broadly speaking, to drive the smaller animals to the mountains and to 
exterminate the larger ones. Some of the extinct types have left rela- 
tives, more or less distant, in the jungles of southeastern Asia, more re- 
sistant to human encroachment than the Chinese hills. But the tapir, 
rhinoceros, gaur and Stegodon of the Yen-ching-kao fauna, although their 
nearest existing relatives are of tropical habitat, do not necessarily in- 
dicate a warmer Pliocene climate in China. They may quite well have 
been species adapted to a temperate climate, such as is more definitely 
indicated by the geographic affinities of the remainder of the fauna. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number Q7 THE AMERICAN MusEUM OF NATURAL HIstTory Dee. 18, 1923 
New York City 


56.(1181:51.7) 


THE FAUNA OF THE HOULDJIN GRAVELS! 
By W. D. MatruHew anp WALTER GRANGER 


The first discovery of Baluchitherium in Mongolia, as reported from 
the field by Andrews and Granger and published by Osborn in Asia 
Magazine and elsewhere, was in the Houldjin gravels near Iren Dabasu 
on the Kalgan-Urga caravan trail. The remains were very fragmentary 
and the identification wholly provisional. The geologic relations were 
described by Berkey and Granger in Novitates No. 42, p. 4. Granger’s 
field identifications have now been revised by comparison of the material 
in the Museum, with the following results: 


Field Identifications Museum Identifications 
1. A rhinocerid ? Cenopus or Preaceratherium sp. 
? Cadurcothertum sp. 
2. A large carnivore Entelodon dirus 


3. An artiodactyl of size of Virginia deer ? Entelodon sp. 
4. An enormous mammal, probably a 
perissodactyl and possibly related to 
or identical with Baluchitheriwm ? Baluchitherium 
5. <A tortoise of large size ? Testudo 
This fauna is of Oligocene age but cannot be more exactly correlated 
until more completely known. It may be coéval with the Hsanda Gol 
fauna but as there is nothing in common save the very doubtfully identi- 
fied Baluchitherium, it would not be safe to correlate it at present. The 
localities are more than a thousand miles apart and the character of the 
two formations quite different. It is more like the Ardyn Obo fauna; 
Cadurcotherium is probably the same. It compares also with the fauna 
described by Borissiak from the Turgai Oligocene. 


Entelodon dirus, new species 
Typre.—No. 19181, last upper molar from Houldjin gravels, Iren Dabasu, 


Mongolia. 

Driaenosis.—Size of Dinohyus hollandi but cusp construction more as in the 
gigantic Oligocene entelodonts; the posterior part of the molar much reduced, with 
small subequal metacone, metaconule and hypocone, the last set upon a wide but 


eee eons of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 16. ‘ 


~ AM.I9181 


Fig. 1.  Entelodon dirus, upper 
molar, m3, right side, crown view. No. 
19181, type. Houldjin gravels, Expe- 
dition of 1922. Natural size. 


Fig. 2. Cadurcotherium sp., 
left lower molar, anterior end 
broken, crown and_ external 


A.M.19183 views. No. 19183, Houldjin 


; gravels, 1922. Natural size. 


1923} THE FAUNA OF THE HOULDJIN GRAVELS 3 


Fig. 3. Cadurcotherium ? right upper molar, external and crown views. Hould- 
jin gravels, 1922. Natural size. : 


obscure posterior cingulum. Exterior cingulum narrow but distinct; anterior cingu- 
lum broad and well defined; paracone and protocone large, equal, the paracone with 
obscure anterior and posterior ridges and a rather more distinct ridge connecting it with 
the protocone. Protoconule quite vestigial. 

The reference to Entelodon is in a broad sense, used as including the 
entire group of closely related genera or subgenera: Archzotherium, 
Dezodon, Pelonax, Dinohyus, Megacherus, etc. It is not close to the 
typical Entelodon magnus of the Ronzon Oligocene nor to Dinohyus of the 
Lower Miocene of Nebraska. It agrees well enough with Archxotherium 
of the American Lower Oligocene except for its larger size, and with 
Megacherus of the Protoceras beds (Upper. White River) of South 
Dakota. It is not so close to the John Day entelodont ‘Cherodon.” 


4 AMERICAN MUSEUM NOVITATES [No. 97 


Fig. 4 Fig. 5 


Fig. 4. Cadurcotherium, upper premolar, crown view. Houldjin gravels, 1922. 
Natural size. 

Fig. 5. Cenopus or Preaceratherium, upper molar, m'’, right side, crown view. 
Houldjin gravels, 1922. Natural size. 


The remaining ‘‘genera’”’ of this group are not comparable, as their 
upper molars are unknown (Pelonax, Deodon, Bodécherus) 1 

In addition to the type species we refer provisionally to this genus a 
canine tooth and an astragalus, which, if correctly referred, represent a 
smaller species; but there is little to distinguish the astragalus from that 
of the anthracotheres, and the canine from that of the large amphicyonids, 
etc. 


Cadurcotherium species 
No. 19183, an incomplete lower molar, is referred to this genus. It 
is about the size of C. cayluxi, with which it accords in characters. A 
premolar also probably belongs here, and an anterior upper molar, 
m! or m?. 


Cenopus or Preaceratherium species 
No. 19184, upper molar tooth, m’, may be referred to this genus 
provisionally. It indicates a species of about the size of the larger individ- 
uals of C. occidentalis, and in about the same stage of molar evolution, 


_ The entire range of variation among these typical entelodonts is not wider than among many 
single genera of bunodont mammals, e. g., Phenacodus. As Sinclair has recently shown, there is a wide 
range of individual variation in teeth of this type. 


1923] THE FAUNA OF THE HOULDJIN GRAVELS 5 


AM.A9IBZ 


Fig. 6. Baluchitherium ? lower molar, crown and external views. No. 19182, 
Houldjin gravels, 1922. Natural size. 


6 AMERICAN MUSEUM NOVITATES [No. 97 


but is somewhat more quadrate in form. No certain reference can be 
made in absence of the anterior teeth. Other species of Cenopus and 
Trigonias among the American rhinoceroses, as also Aceratherium filholi 
Osborn and Ronzotherium reichenaui Deninger, show this subquadrate 
form in m%, which appears to be a rather variable species character. 
Both these European species are placed by Abel under Preaceratherium. 
Borissiak! has described under the name of Epiaceratherium turgaicum 
a small rhinoceros with which the Houldjin species may prove to be 
identical, at least generically. 


?Baluchitherium 


No. 19182, a lower molar and fragments of other teeth indicate a 
gigantic rhinoceros of size appropriate to Cooper’s genus. A few very 
fragmentary remains of the skeleton, No. 19180, are of comparable size. 
These were found scattered in the gravels, along with equally fragmen- 
tary remains of smaller perissodactyls, presumably including the 
Cadurcotherium and Cenopus noticed above; also some fragments of a 
large tortoise. 


rr 


1Borissiak, 1918, Mem. Soc. Pal. Russie, I, pp. 1-82, Pl. 1-111. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 98 Tur AMERICAN Museum or Naturat History Dec. 18, 1923 
New York City 


56.(1181:51.7) 


THE FAUNA OF THE ARDYN OBO FORMATION! 
By W. D. MatrHew and WALTER GRANGER 


The Ardyn Obo formation was named and defined by Berkey and 
Granger in American Museum Novitates No. 77. A small collection of 
mammals was secured in 1922 at ‘‘Promontory Bluff”’ on the Sair-Usu- 
Kalgan trail, about 150 miles from Sair-Usu and 350 from Kalgan. It 
consists chiefly of an amynodont, apparently a species of Cadurcotherium, 
of which finely preserved skulls, jaws, limb-bones and feet were secured. 
These are described by Professor Osborn in another number of Novitates. 
In addition to these, fragmentary remains of a few other animals were 
obtained, affording the following faunal list: 


Carnivora 

PLOT EG GE Peau oS ERS Sai ha ean ae anterior part of lower jaw (no molars) 
Perissodactyla 

GOdurcothertum ardyNense,. Mor. 2i sc ie eens skulls, jaws, limbs and feet 

Ardynia precox, new genus and species..................... upper jaw, ete. 
Artiodactyla 

1 INET ATTACHES Wee, MCN RA rehire Rey A aa oe a a eae lower jaws 

Schizotherium avitum, new species..................020000 eee lower molar 

Anthracotheriid, gen. indet................ AS MN etA Wy Coyepat Nene, lower molar 
Chelonia 

Testudo insolitus, new species...... parts of carapace and plastron; lower jaw 


This appears to be an Oligocene fauna. Cadurcotherium, Schizo- 
thertcum and Cynodictis are characteristic Phosphorites genera; the giant 
tortoise, while suggesting the giant tortoises of the Miocene and Pliocene 
in its size, is apparently a rather primitive stage in its costal plates. It 
may or may not represent the Baluchitherium fauna but cannot be very 
different in geologic age. 

Its nearest geographic affinities, it will be observed, are with West- 
ern Europe, not with the United States. The bearing of this and other 
evidence upon Tertiary zoédgeographic divisions and upon the centers of 
dispersal of different groups will be considered in later papers after the 
pertinent facts have been ascertained and placed on record. 


fs a leer of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
ion No. 17. 


2 AMERICAN MUSEUM NOVITATES [No. 98 


A.M.i9156 


Fig. 1. Ardynia precox, upper jaw, p?-m! and m§ of left side, external and crown 
views. No. 19156, type specimen, Ardyn Obo formation, 1922. Natural size. 


Carnivore cf. Cynodictis 


The anterior part of lower jaw with c-p2 represents a slender-jawed, 
fox-like animal, whose exact relations are indeterminate. It agrees with 
Cynodictis so far as it goes. 


Cadurcotherium ardynense Osborn 
(See Amer. Mus. Novitates, No. 92) 


Ardynia precox, new genus and species 


Typr.—No. 19156, upper jaw with p*m!; m? doubtfully associated. Ardyn 
Obo formation, Promontory Bluff, Expedition of 1922. 

CHARACTERS :—Size of Hyracodon and crowns of teeth of about the same height 
and general aspect but premolars considerably reduced and last molar of normal 
rhinocerotid type, the ectoloph not extended behind the metaloph. The first pre- 
molar is absent, the second about half the transverse diameter of m!, the third and 
fourth smaller than m!. Ectoloph of p? somewhat convex, of p*‘ flat externally; 


1923] THE FAUNA OF THE ARDYN OBO FORMATION 3 


metaloph and protoloph distinct on p?, apparently so on p*4. Ectoloph of m* and 
probably of other molars flat or nearly so, the antero-external rib not conspicuous. 

A number of fragments of lower jaws and teeth of a small rhinoceros 
and a premaxilla without teeth are of size appropriate to Ardynia but 
there is no proof of their association. If correctly referred, the lower 
premolars have somewhat the same construction as in Cadurcotherium, the 
molars are more like those of Hyracodon, and the premaxilla bore three 
equal incisors of larger proportionate size than in Hyracodon, but of 
similar arrangement. In absence of more definite: proof of association 
the genus is provisionally referred to the Hyracodontide on the ground of 
the general appearance of the type and the provisionally referred pre- 
maxilla, but the premolar construction would suggest Amynodontide, 
and the third molar Rhinocerotide. 


Fig. 2. Schizotherium avitum, third lower molar, crown and external views, 
Ardyn Obo, 1922. Natural size. 


Schizotherium avitum, new species 


Typr.—No. 19157, lower m3. Ardyn Obo formation, Promontory Bluff, Expedi- 
tion of 1922. 

CHARACTERS.—Size a little less than that of S. priscum (=modicum) of the 
Phosphorites; heel of ms narrow and more reduced, lacking the transverse cingular 
crest on each side of it that characterizes S. priscum. Trigonid and talonid of sub- 
equal width, while in S. priscwm the trigonid is wider than talonid; length of tooth more 
than twice the width; in S. priscum it is somewhat less. 


4 AMERICAN MUSEUM NOVITATES [No. 98 


Exact comparison with S. turgaicum Borissiak! is not possible, as his 
material does not include the last molar. The Turgai species appears, 
however, to be very closely related to the Ardyn Obo animal. 

An unnamed species from the Gaj formation of India, doubtfully 
referred by Pilgrim to Schizotherium,? appears to be of larger size than the 
three species of the Holarctic Oligocene. 


Measurements of m3; 


S. avitum S. priscum 
No. 19157 No. 11077 No. 10495 
(cast) 
Length, of total, ms 25.3 30.0 28.2 
“of heel (hypoconulid) 2.2 3.2 yew ign 
Width of trigonid 11.9 ae! 15.0 
Ca" balourd. 12.1 13.9 14.0 
iH “ heel (hypoconulid) 4.0 9.5 8.9 
Index, length to trigonid width 2.1321 1.99:1 1.88:1 


Anthracotheriid, gen indet. 
An incomplete lower molar indicates a small species of this family. 
It is comparable in size with Microbunodon of the Upper Oligocene of 
Europe, with the smaller Ancodons of the Upper Eocene and Oligocene. 
It compares more nearly with Ancodon in height and structure of the 
crown but is hardly determinable as to genus. 


Eumeryx species 


The type of this new genus is #. culminis of the Hsanda Gol forma- 
tion in the Tsagan Nor basin. It is a small ruminant, scarcely larger 
than Leptomeryx of the American Oligocene and represented in the 
Hsanda Gol by numerous parts of upper and lower jaws, limb and foot- 
bones, which indicate a dentition intermediate between Leptomeryx and 
Blastomeryx, somewhat more advanced than in Prodremotherium and 
Amphitragulus, while the foot-bones show both metacarpals and meta- 
tarsals united into cannon-bones of characteristic Pecoran type, much 
like the earlier species of Blastomeryz. 

This interesting form will be described more fully in a later article. 
The jaws from the Ardyn Obo formation are referable to the genus but 
it is doubtful whether they belong to the same species, as the construc- 
tion of the premolars seems to be simpler and nearer to the tragulid type. 


1Borissiak, A., 1921, Ann. Soc. Pal. Russie, III, p. 
2Pilgrim, G. E., 1912, Mem. Geol. Sur. Ind., ree Indica, N. S8., IV, No. 2, p. 36. 


1923] THE FAUNA OF THE ARDYN OBO FORMATION | 5 


Testudo insolitus, new species 


Typr.—No. 6275, parts of carapace and plastron of several individuals (co- 
types) from Promontory Bluff, Ardyn Obo basin, Mongolia. 

. CHARACTERS.—Middle costal plates not perceptibly wedge-shaped, their sides 
approximately parallel. Shell moderately thick, sutures slightly open in half-grown 
specimen, and surface rather heavily undulate. Upon the middle marginal plates 
the borders of the shields are marked by a somewhat raised ridge on the plate with a 
deep median furrow; elsewhere the furrows are normal; a precursor, apparently, of 
the condition seen in 7’. cubensis of the Pleistocene, where the ridge has become much 
more prominent than the furrow. I have not been able to find any description of this 
curious character in other species of Testudo either living or fossil; but it appears to 
be indicated in a drawing by Gervais of a costal plate of a tortoise from St. Gerand- 
le-Puy.! The species attains the size of the large tortoises so common in the Ameri- 
can Miocene, but apparently retains the more primitive character of the Eocene 
and some Oligocene species in the uniform costal plates. The material is too frag- 
mentary to render profitable any detailed description of the various pieces or dis- 
cussion of its exact relationships. The curious resemblance to the extinct Cuban 
species may be interpreted as indicating some affinity between the two, but much 
more evidence would be needed to prove it. The shell is not thinned and incomplete 
to any noticeable degree as compared with other large Tertiary continental species. 


_ An incomplete lower jaw, lacking the articular ends of either ramus 
is referred to this species. 


Emydid, gen. indet. 
A single fragment of a small emydid with the rugose surface of 
Trachemys. 


Trionychid, gen. indet. 
Two small incomplete plates represent a small trionychid. 


1Gervais, P., 1858, Zool. et Paleont. Franc., Pl. 1, fig. 7. 


j ‘ a i 
T Ses hSt ae 
nee 


th let én ie 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 100 Tue American Museum or Naturat History Dec. 28, 1923 
New York City 


59.9,32(51) 


NEW CHINESE INSECTIVORES! 
By GuLover M. ALLEN 


The mammals obtained by the Asiatic Expeditions of the American 
Museum of Natural History, under the leadership of Mr. Roy Chapman 
Andrews, include over 460 specimens of Insectivora, many of them of 
unusual rarity, such as the genera Neotetracus, Scapanulus and Scap- 
tonyx. Most important are the excellent series of skins from western 
and southwestern Yunnan, which afford ample basis for demonstrating 
that distinct representative subspecies in many groups are to be found in 
the extreme southwest of that Province, and that these are readily dis- 
tinguishable from those of the Li-chiang highlands whose summits run to 
13,000 feet and over. These in turn are in some cases represented by 
distinct races farther north in Szechuan. Some of the western Yunnan 
races have lately been described by Mr. Oldfield Thomas from collections 
made by Mr. George Forrest. The intensive field work of Messrs. 
Andrews and Heller has resulted in the discovery of additional new races 
in this region, while their smaller collections from eastern China, 
Szechuan and southeastern Mongolia have been of much help in affording 
comparable material of other forms. The following appear to be new. 


Talpide 


Of the long-nosed, shrew-like moles of the genus Rhynchonaz, the 
collection contains a fine series of fourteen skins from Mu-cheng, on the 
Salween drainage, southwestern Yunnan, taken at altitudes of from 6000 
to 7000 feet; and a series of eight from the Snow Mountain (Ssu-shan), 
Li-chiang, Yunnan, from near timber-line at 10,000 to 12,000 feet. As 
might be expected the former series is strikingly darker than that from 
the higher level, with conspicuously blackish rumps. Both series differ 
again from typical R. andersoni of Szechuan of which a series is available 
in the Museum of Comparative Zodlogy, some of which have been iden- 
tified as of this species by Mr. Oldfield Thomas of the British Museum. 
The two Yunnan forms may stand as subspecies. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
bution No. 18. 


2 AMERICAN MUSEUM NOVITATES [No. 100 


Rhynchonax andersoni atronates, new subspecies 


Typr.—Female, skin and skull, No. 44848, American Museum of Natural His- 
tory, from Mu-cheng, Salween drainage, southwestern Yunnan, 7000 feet altitude. 
February 13, 1917. R.C. Andrews and E. Heller. 

Description.—A dark form, the rump nearly unmixed slaty black; skull 
smaller, with less reduced third upper premolar, than in typical andersoni. 

General color above nearly “Prout brown’ (Ridgway). The pelage consists of 
wholly black shining hairs mixed with others that are blackish slate basally, tipped 
with hazel. On the rump, the latter sort of hairs are fewer or absent, giving a strik- 
ingly blackish appearance to this region. The lower surfaces of body and limbs are 
uniform blackish slate. Backs of feet and the entire tail, scaly, with scattered minute 
blackish hairs; the tail usually not paler underneath. 

The skull is smaller than in typical andersoni of Szechuan and the teeth are 
smaller throughout except that the third upper premolar, which in the latter is minute, 
about half the crown-area of the first,.and barely reaches the level of the cingulum of 
the two adjoining premolars, 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, while in all, the small 
second premolar (pz) is well developed, though slightly smaller than the canine; 
whereas in the Szechuan series it is either very minute or even altogether absent. 

MeraASUREMENTS.—The collectors’ measurements are: head and body, 67 mm.; 
tail, 57; hind foot, 14; ear, 10; these are practically as in typical andersoni. The 
skull measures as follows (with corresponding dimensions of a Szechuan specimen in 
parenthesis): greatest length, 20.5 (22.0); basal length, 16.4 (17.6); palatal length, 
9.2 (10.2); mastoid width, 11 (11.4); zygomatic width, 10.2 (10.6); width outside 
first molars, 6.0 (6.7); upper tooth row, 8.8 (10); lower tooth row, 8.0 (9.0). 


Remarks.—In addition to the series from Mu-cheng, a single skin 
from Pei-tai-ping, Mekong drainage, appears to be this form. 


Rhynchonax andersoni nivatus, new subspecies 


Typr.—Male, skin and skull, No. 44352, American Museum of Natural History, 
from Ssu-shan (Snow Mountain), Li-chiang, western Yunnan, 12,000 feet altitude. 
October 22, 1916. R.C. Andrews and E. Heller. 

Descrietion.—Similar to typical andersoni but much paler with a smaller skull 
and larger third upper premolar. In size and cranial characters it resembles the 
preceding form but is much paler, lacking the blackish rump, and the tail is indis- 
tinetly bicolor. 

General color above light cinnamon-brown, almost the same tone as in our Sorex 
personatus; fore legs and under parts “deep neutral gray” (Ridgway, 1912), much 
paler than in andersoni from Szechuan. In certain lights the tips of the hairs both 
above and below reflect the light in little glints. Tail fuscous above, pale beneath. 

The skull is smaller than in the typical form, and closely resembles that of 
atronates. The third upper premolar is as large as the first, stands well in the tooth 
row with its cingulum on a level with those of the adjacent teeth, and is provided with 
a distinct crown about as high as the width of its cingulum. The second lower incisor 


1923] NEW CHINESE INSECTIVORES 3 


is larger than the canine, hence much larger than in Szechuan examples of andersoni. 

MEASUREMENTS.—The collectors’ measurements of the type are: head and body, 
68 mm.; tail, 60; hind foot, 15. The skull measures: greatest length, 20; basal 
length, 16; palatal length, 9.7; mastoid width, 11; zygomatic width, 10; width out- 
side first molars, 6.5; upper tooth row, 9; lower tooth row, 8.2. 

Remarks.—The series of nine skins from the Snow Mountain is 
very uniform in its paler brown tint, as compared with typical andersoni 
and the subspecies atronates. Indeed, so striking was the difference that 
the entire series was picked out at once in unpacking, without reference 
to the labels for locality. In both the Yunnan races the third upper pre- 
molar and second lower incisor are less reduced than in the Szechuan 
series, indicating the retention of a more primitive condition by the 
southern forms. The discovery of the genus in southwestern Yunnan at 
high altitudes is an interesting extension of its known range. 


Soricidz 

Seventeen skins of small striped-backed shrews prove very interest- 
ing. Nine of these, from Li-chiang, Yunnan, seem to correspond closely 
with Milne-Edwards’s description of Sorex cylindricauda. The dark line 
down the back is rather poorly defined, often amounting only to an in- 
distinct darkening of the mid-dorsal area, as represented in Muilne- 
Edwards’s plates (‘Recherches pour servir 4 l’ Hist. Nat. des Mammiféres,’ 
1868-74, Pl. xxxvu A, fig. 3; xxxvul B, fig. 3). The teeth also 
agree in that the fourth unicuspid is very little smaller than the three 
preceding, so that the entire four are nearly equal. In Sorex bedfordiz, 
the three anterior unicuspids are subequal, the fourth much smaller in 
side view. Two specimens from Tai-pei-shan, southern Shensi, seem to 
be this, with much better-marked dorsal stripe. Five skins from western 
Yunnan, of this general type, seem to correspond with Thomas’s Sorex 
wardi fumeolus (type locality Wei-chow, western Szechuan). They are 
much darker in color than bedfordiz but with pale gray instead of brown- 
ish bellies. All these specimens were obtained at altitudes of from 10,000 
to 12,000 feet. At 7000 feet on the Salween drainage, a single example 
was secured which is so much darker and smaller than all the rest that it 
evidently represents a distinct low-altitude race, probably related to 
bedfordiz as its lower side is of the same dark color as the back instead of 
contrastingly pale. It may be known as follows. 


Sorex bedfordie gomphus, new subspecies 
Typr.—Male, skin and skull, No. 44320, American Museum of Natural History, 
from Mu-cheng, Salween drainage, western Yunnan, 7000 feet altitude. February 
11, 1917. R.C. Andrews and E. Heller. 


4 AMERICAN MUSEUM NOVITATES [No. 100 


DescripTion.—A very small dark-brown shrew with black dorsal stripe. 

General color above a rich dark cinnamon-brown, near Mars brown of Ridgway 
but clouded. A narrow blackish stripe runs from the nape to the base of the tail, its 
edges not sharply defined but merging gradually with the color of the back. Below, 
the chest and belly are washed with cinnamon-brown and the throat is clearer gray 
with a silvery sheen. Backs of feet and the tail all around clothed with minute dark 
hairs of about the same cinnamon-brown as the body. 

In side view the three anterior unicuspids are subequal, the fourth decidedly 
smaller than the fifth, but in crown view the cross-section of the last is greater than 
that of the fourth, agreeing in these respects with bedfordiz. 

MrasurEMENTS.—The collectors’ measurements follow, with those of the type of 
bedfordiz in parenthesis: head and body, 55 mm. (55); tail, 39 (55); hind foot, 13 
(11). The skull is small and delicate; it measures: greatest length, 16.6; palatal 
length, 7.3; mastoid width, 8.0; width outside upper molars, 4.4; upper tooth row, 
7.0; lower tooth row, 6.3. 


Sorex excelsus, new species 


Typr.—Adult male, skin and skull, No. 44859, American Museum of Natural 
History, from summit of Ho-shan, Pei-tai, thirty miles south of Chung-tien, Yunnan, 
China, altitude 13,000 feet. November 29, 1916. R.C. Andrews and E. Heller. 

DescriptTion.—A medium-sized shrew, of a general grayish brown above, silvery 
below, with sharply bicolor tail. Skull with long attenuated rostrum, anterior two 
unicuspids equal, the third and fourth smaller, the former slightly the larger of the 
two. 

Entire dorsal surface of head and body nearly “ Prout brown” (Ridgway, 1912), 
slightly grayer on the head, and faintly darker on the middle of the back. Below, 
silvery gray, the bases of the hairs “blackish plumbeous”’ but concealed by their pale 
tips; chin whitish. Backs of the feet clothed with short whitish hairs. Tail sharply 
bicolor, dusky above, whitish below, with a dark-brown pencil of longer hairs. 

The skull shows no especial peculiarities except for its long slender rostrum. The 
teeth are conspicuously pigmented, and the upper unicuspids recall those of certain 
American species in that the anterior two are equal and larger than the posterior 
two which are nearly equal though the third is obviously larger than the fourth in 
side view. 

MEASUREMENTS.—The collectors’ measurements of the three specimens, all 
from the same locality, are: , 


No. Head and Tail Hind Foot 
Body 
44357 62 50 14 
44358 60 50 1335 
44359 (Type) 60 51 13 


The skull of the type measures: greatest length, 18.7 mm.; basal length, 16.2; 
palatal length, 8.5; breadth of brain case, 8.6; breadth outside last molars, 4.6; 
upper tooth row, 8.2; mandible, 10.2; lower tooth row, 7.8; depth of brain case, 5.1. 

Remarks.—This alpine shrew does not closely resemble any of the 
Chinese species yet described. In color and general external appearance, 
it is indistinguishable from S. longicaudus of the humid northwest coast 


1923] NEW CHINESE INSECTIVORES 5 


of America. Its skull, however, is longer, and in this respect, too, it 
exceeds its neighbors, S. cylindricauda, S. bedfordiz and S. wardi fumeolus 
(the last perhaps best considered a race of S. bedfordix). These three 
have a more or less well-marked black dorsal line, no trace of which is 
present in S. excelsus, while their lower surfaces and the backs of their feet 
are darkened with a brownish wash. There is no suggestion of the tri- 
color pattern, with buffy sides, characteristic of the S. araneus group. It 
is not unlikely that the relationship between the new species and some of 
the American shrews is fairly close. 


Soriculus caudatus umbrinus, new subspecies 

Typr.—Adult male, skin and skull, No. 44388, American Museum of Natural 
History, from Mu-cheng, Yunnan Province, China, Salween River drainage, 7000 
feet. February 11,1917. R.C. Andrews and E. Heller. 

DescriptTion.—*A small subspecies, most like S. sacratus Thomas, the Szechuan 
representative of caudatus, but differing in its much darker-brown coloring and in 
having the tail dark all around instead of bicolor. 

General color above a uniform ‘‘seal brown” (Ridgway, 1912) instead of the 
“slate gray” of sacratus; below slightly paler brown (nearly ‘‘ Brussels brown’’) 
tinged with gray. Backs of the feet and the tail all around light seal brown like the 
back. : 

Skull similar to that of sacratus in measurements, hence slightly smaller than that 
of caudatus of Nepal. The tips of all but the minute unicuspid and the last molar 
above are pigmented chestnut. 

M®&ASUREMENTS.—The collectors’ field measurements of the type are: head and 
body, 60 mm.; tail, 55; hind foot, 12. Average of nine specimens: head and body, 
62 mm.; tail, 53.5; foot, 12.8. 

The skull measures: greatest length, 18.1 mm.; palatal length, 8.1; mastoid 
width, 9.0; width outside molars, 5.0; upper tooth row, 8.1; mandible (exclusive of 
incisors), 9.5; lower tooth row, 7.3. 


Remarks.—This small and very dark-brown Soriculus is interesting 
not only as extending the known range of the species into the middle 
altitudes of southwestern Yunnan, but also as offering another instance 
of the response of a wide-ranging species to the climatic conditions of this 
area through the development of a smaller and darker race as compared 
with the representatives farther north. 


Chodsigoa hypsibia parva, new subspecies 

Typre.—Male, skin and skull, No. 44890, American Museum of Natural History, 
from Li-chiang, Ssu-san-chong, Yunnan, 9000 feet altitude. October 12, 1916. R.C. 
Andrews and E. Heller. 

Description.—A small dark representative of the more northern species, 
hypsibia. 

General color a uniform ‘“‘deep mouse gray”? (Ridgway, 1912), scarcely lighter 
below, and with a faint brownish tinge in some lights. Feet pale, with scattered dark 


6 AMERICAN MUSEUM NOVITATES [No. 100 


hairs; tail dark above, pale whitish below, with minute short hairs not dense enough 
to obscure the scales. The general color is nearly the same as that of Blarinella wardi, 
an associated species. 
M#ASUREMENTS.—The smaller size of this little pita: is its most notable feature. 
The collectors’ measurements of the four specimens are (compare with head and body, 
84 mm., tail, 65, foot, 15 of C. hypsibia): 
No. Head and Tail Hind Foot 


Body 
44390 (Type) 54 44 11.5 
44391 55 43 11 
44395 56 45 1] 
44396 55 45 11 


The skulls of all four were badly crushed in the traps. None of the molars is 
pigmented, and the other teeth have merely the tip chestnut. The rostrum of the 
type has a width outside second molars of 4.2 mm. (6.1); length of upper tooth row, 
6.6; of lower tooth row, 6.1; tip of lower incisor to condyle of jaw, 9.1 (12.3); to 
angle, 9 (11.5) (dimensions in parenthesis are of the type of C. hypsibia). 


Remarks.—Of this genus of shrews characterized by $=28 teeth, 
four specimens were captured at Li-chiang, which is apparently the most 
southern locality yet recorded for it. It is probably best considered a 
geographical variety of C. hypsibia of Szechuan, from which it differs in 
its notably smaller size and darker color. It is the smallest of the 
forms yet discovered. 


Chodsigoa smithii parca, new subspecies 


TypE.—Male, skin and skull, No. 44409, American Museum of Natural History, 
from Ho-mu-shu Pass, western Yunnan, China, 8000 feet. April 6, 1917. R. C. 
Andrews and E. Heller. 

Description.—A long-tailed species with nearly the external dimensions of C. 
smith but with a much smaller skull and less attenuate rostrum. The ears and feet 
are slightly smaller and the latter are brownish instead of whitish. 

General color above and below a “dark mouse gray”? (Ridgway, 1912) with a 
slight brownish wash; the color is darkest on the rump, palest on the belly. Tail 
dusky brown above, scarcely lighter below, clothed with minute dark-brown hairs 
which do not exceed the scaly rings. Backs of the feet distinctly brownish, digits 
whitish. 

The skull, though nearly the size of that of C. hypsibia and much less than that 
of C. smithii, is very different from either in having the rostrum relatively shorter 
and more gradually tapering from brain case to tip instead of having the premaxillary 
region abruptly narrowed. The teeth are smaller and more slender than in either. 

MEASUREMENTS.—The tail is decidedly longer than head and body, more so 
than in typical smithii, but not so long as in the long-tailed salenskii. The collectors’ 
measurements of the type follow, with those of a C. smithii from southern Shensi in 
parenthesis: head and body, 70 mm. (80); tail, 91 (83), foot, 17.5 (18). The skull 
measurements follow with those of C’. smithii in parenthesis: greatest length, 19.3 mm. 
(22.5); palatal length, 9.0 (10); greatest width, 9.0 (10); width outside molars, 5.6 


1923] NEW CHINESE INSECTIVORES 7 


(6.6); upper tooth row, 8.5 (10); mandible, 12 (13); lower tooth row, 7.6 (9). Upper 
unicuspids subequal. In the type the tips of the large anterior incisors alone 
are pigmented, though in the’other specimens the tips of the five anterior teeth are 
chestnut. 

Remarks.—Notwithstanding the peculiar formation of the rostrum 
and the slightly greater proportionate length of tail, it seems best to 
consider these Yunnan shrews a race of the larger species, smithiz, of 
which the collection contains two specimens from 10,000 feet in Tai-pei- 
shan, in the Tsing-ling Range of southern Shensi. A large series of C. 
hypsibia from the base of the same mountain has also been useful for 
comparison. The nearly equal size of the three upper unicuspids seems 
to be a character allying the new race to smithii rather than to hypszbia, 
in which the anteriormost seems to be the largest. In addition to the 
type and a second specimen from the same locality, the collection contains 
two other examples from Ssu-shan-chong, Li-chiang, Yunnan, 9000 feet, 
which are practically identical with them. 


Crocidura ilensis phaeopus, new subspecies 


Typr.—Adult female, skin and skull, No. 56013, American Museum of Natural 
History, from Wanhsien, Szechuan Province, China. November 2, 1921. Third 
Asiatic Expedition. 

Description.—A very small, dark-brown species of the ilensis group, distin- 
guished by its dark-brown instead of whitish feet. 

The entire dorsal surface of the body, the backs of the fore and had feet, and the 
upper surface of the tail are dark brown, nearly ‘“‘mummy brown”’ (Ridgway, 1912). 
At the lateral border the under side of chin, throat, forearms and belly becomes rather 
abruptly whitish, the hairs except at the chin white-tipped with “dark gull gray” 
bases. The base of the tail is gray beneath but terminally 's much the same on both 
surfaces. The long bristles are scattered evenly throughout its length. 

The skull has about the same dimensions as in its neighbors shantungensis and 
coree. It is small and delicately formed, the rostrum not especially elongated. The 
first upper unicuspid is double the height of the rounded posterior cusp of the an- 
terior incisor. The second and third unicuspids are practically equal in height and 
cross-section, and reach the level of the anterior cusp (paracone) of the carnassial. 
The lacrymal foramen is exactly over the point of contact between first and second 
molars. 

MEASUREMENTS.—The collectors’ measurements of the type are: head and body, 
62 mm.; tail, 37; foot, 12; ear, 9. The skull measures: greatest length, 16.9 mm.; 
basal length, 15; palatal length, 7.6; width of brain case, 7.6; occipital depth, 4.0; 
mandible, 10.7; upper tooth row, 7.3; width outside molars, 5.0; lower tooth row, 6.6. 


ReMarks.—This very small brown species seems to be widely 
spread in Asia from Corea (corex) to Palestine (portali). From the typical 
subspecies, zlensis, and from the lowland form of eastern China, appar- 
ently identical with shantwngensis, the. race from the Szechuan highlands 


8 AMERICAN MUSEUM NOVITATES [No. 100 


differs conspicuously in its dark feet and tail. The Third Asiatic Expedi- 
tion secured a fine series at Wanhsien and a single specimen at the base 
of Tai-pei-shan, Tsing-ling Mountains, on the southern border of Shensi, 
which are very uniform in this respect. A skin from Weihsien, in Shan- 
tung, and another from Shensi, 45 miles south of Fengsiangfu are paler 
with obviously whitish feet and lower side of tail. They evidently repre- 
sent the race shantungensis, from which corex is apparently not very 
different. A skin from Ichang, which I had provisionally referred to 
coree (Mem. Mus. Comp. Zodél., XL, p. 242, 1912), is, as might be ex- 
pected, somewhat intermediate between the Shantung and Szechuan 
races, but on the whole, with its paler feet, is best referred to the lowland 
race (shantungensis). This species is not listed among the many small 
mammals secured for the British Museum by Anderson and other collec- 
tors in Szechuan and may be of local or rare occurrence. 


Crocidura vorax, new species 


Type.—Adult male, skin and skull, No. 44383, American Museum of Natural 
History, from Li-chiang, Yunnan, China, taken in timber-line forest on Ssu-shan 
(Snow Mountain), 12,000 feet altitude. October 15, 1916. R.C. Andrews and E. 
Heller. 

DESCRIPTION. 
C. russula group. 

Color of head and body above a very pale grayish brown, nearly ‘‘ wood brown” 
of Ridgway, and very closely like that of Sorex personatus, with a similar pepper-and- 
salt appearance, due to the presence of a narrow gray band below the brown tips of 
the hairs. This-color gradually pales into the gray of the belly, with a faint wash of 
buffy on the chest. The bases of the hairs everywhere are slaty, becoming paler, 
almost “slate gray’’ below, where they show through slightly. Tail distinctly bicolor, 
darker than the back above (near ‘‘clove brown’); clear gray below. The hair of the 
tail is thick enough to conceal the scales, while the long bristle-like hairs are rather 
scattered, and are more numerous near the base of the tail. Ear thin and rather 
small, less conspicuous than usual, but not entirely concealed. 

The skull is of about the same size as in C. russula of Europe, and has a low but 
well-defined sagittal ridge and more prominent lambdoid ridges. The first wpper 
unicuspid is about twice the height of the cusp of the first incisor. The second and 
third unicuspids are practically equal in both vertical extent and cross-section; their 
tips are practically on the same line as the tip of the anterior cusp (paracone) of the 
succeeding (carnassial) tooth (instead of exceeding it as in russula). The height of 
the paracone of the carnassial above the anterior root is about equal to one-half the 
height of the main cusp, and equals the distance from the tip of the anterior cusp to the 
summit of the main cusp along the front face (in russula it is much less). The profile 
of the posterior edge of the carnassial is a wide, backwardly directed crescent, and the 
summit of the main cusp is about over the center of the base (a much longer tooth 
than in russula). The main cusp of the anterior incisor is slender and turned strongly 
downward. 


A medium-sized, pale-brownish species, apparently allied to the 


1923] NEW CHINESE INSECTIVORES 9 


MerasurEMENtTS.—The collectors’ measurements of the type are: head and body, 
72mm.; tail, 51; foot, 13. Five other specimens from the type locality average: 
head and body, 64.2 mm.; tail, 40.6. 

The skull of the type measures: greatest length, 19.8 mm.; basal length, 17.4; 
palatal length, 8.6; greatest width of brain case, 9.0; width outside molars, 5.7; 
mandible (excluding incisor), 10.4; upper tooth row, 8.3; lower tooth row, 7.6. 

Remarks.—Its rather long fur, brownish coloration and thin small 
ears combine to give this mountain shrew much the appearance of a 
Sorex. The collectors’ note states that the type specimen was seen 
devouring a mouse (Apodemus) in a trap. Two additional examples 
from lower altitudes to the eastward (Yangtse River, at Chih-tien and 
Taku Ferry, 6400 and 6000 feet respectively) are similar but differ slightly 
in having the paracone of the carnassial shorter than the unicuspids 
adjoining it. 

What is obviously a representative of this same shrew farther south- 
ward and at a lower altitude is a single skin from the Mekong River. 
It is so very much richer in its brown coloration that it cannot be included 
with the Li-chiang specimens, and in common with other southern repre- 
sentatives of more northern species in this part of Yunnan seems worthy 
of distinction. It may be known as follows. 


Crocidura rapax, new species 


Typr.—Adult male, skin and skull, No. 44321, American Museum of Natural 
History, from Ying-pan-kai, Mekong River, southern Yunnan, at 9000 feet. Decem- 
ber 25, 1916. R.C. Andrews and E. Heller. 

Description.—In size and proportions resembling C. vorax but the entire dorsal 
surface a richer brown, nearly “bister” of Ridgway (1912), slightly peppered with 
gray on the head and shoulders. Below a light “mouse gray.” Feet clothed with 
minute gray hairs; tail bicolor like the body, the bristle-hairs rather few and scattered. 

Skull like that of C. vorax but very slightly more delicate. 

MEASUREMENTS.—The collectors’ measurements are: head and body, 64 mm.; 
tail, 42; foot, 12.5. The skull measures: greatest length, 18 mm.; basal length, 
16.3; palatal length, 8; greatest width, 8.2; mandible, 11.4; upper tooth row, 8; 
width outside molars, 5.3; lower tooth row, 7.4. 


Remarxs.—This and the previous species are of about the same size 
apparently as C. indochinensis of Siam and I have referred to the latter a 
single specimen from Ho-mu-shu Pass, Yunnan, which agrees closely 
with the description given by Robinson and Kloss in its slaty under parts 
and dark mouse-gray instead of brown color above. Although treated as 
distinct species, it is likely that the two new forms are but geographical 
representatives of a single species. 


10 AMERICAN MUSEUM NOVITATES [No. 100 


A handsome series of twenty-two skins of Anourosorex proves to be 
very interesting, as it represents the two species sguamipes and assamen- 
sis. Two skins from Tai-pei-shan, Tsing-ling range, Shensi, are appar- 
ently typical of squamipes, and extend its known range to this Province. 
They were caught at an altitude of 10,000 feet. Thirteen other specimens 
from various localities in southern Yunnan appear to represent a smaller 
geographic race, here described as new. A third lot from the Salween 
drainage, southwestern Yunnan, is again quite different, with larger, 
heavier skulls, less shiny fur, a greenish wash below, and brownish rumps. 
They are evidently referable to assamensis, but differ from it in size and 
color sufficiently to be worthy of recognition as an eastern form. The 
dentition in the two species is very similar, but in squamzpes there is 
usually a shallow notch on the anterior edge of the second upper incisor, 
whereas in assamensis this is usually lacking. The descriptions follow. 


Anourosorex squamipes capnias, new subspecies 


Typr.—Male, skin and skull, No. 44506, American Museum of Natural History, 
from To-mu-lang, Chung-tien district, Yunnan, 10,000 feet altitude. December 
3, 1916. R.C. Andrews and E. Heller. 

DeESCRIPTION.—Similar to typical squamipes but smaller, the fur slightly grayer 
and less shining. 

Entire upper surface a “dark mouse gray”’ with a finely peppered appearance on 
close inspection; lower surface of body with a light wash of ‘tawny olive.” The entire 
pelage in certain lights is silvery, but less so than in typical sqguamipes. Both fore 
and hind feet are smaller and more delicate. 

The skull is shorter and of slightly smaller proportions than in the latter, with 
which it shares the character of having a slight notch on the anterior border of the 
second upper incisors. 

M®EASUREMENTS.—The corresponding dimensions of a typical squamipes from 
Szechuan are given in parenthesis following those of the type. Length, 90 mm. (100); 
tail, 14 (14); hind foot, 13 (16). Skull: greatest length, 23 mm. (25); basal length, 
20.8 (23); palatal length, 11.5 (12.5); upper tooth row, 11 (12); mastoid width, 
11.6 (13); mandible, 11 (12.2); lower tooth row, 10 (11). 


ReMARKS.—In addition to the type, specimens were taken at the 
following localities in southwestern Yunnan: Mekong River, Siao-ke-la, 
8000 feet; Mekong drainage, Chiang-wei, 8000 feet; Mekong River, 
Yin-pan, 9000 feet; Mekong River, La-chu-mi, 9000 feet; Ha-pa, 20 
miles north of Taku, 10,000 feet; Peh-tai Mountain, 30 miles south of 
Chung-tien, 10,000 feet. It will be seen that all these places are at 
elevations of 8000 feet or over. The range of A. assamensis appears to 
be at a lower altitude, for the series from Mu-cheng, Salween drainage, 
was secured at elevations between 6000 and 70°0 feet. The rarity of this 


1923] NEW CHINESE INSECTIVORES 11 


species in collections (the type has apparently remained unique) seems to 
have caused some doubt as to its distinctness from squamipes, and 
Wroughton has even proposed that the two be considered inseparable. 
But Anderson’s figures and measurements, as well as his careful descrip- 
tion, seem to leave no doubt that assamensis is a distinct species, with 
skull proportionally very much larger than in squam7pes, equaling one- 
half the length of the vertebral column to tlte end of the sacrum. That of 
his type specimen is relatively enormous, much larger than any of the 
Mu-cheng series, which again are larger and heavier than those of 
squamipes. The skins have a peculiar greenish wash on the belly, quite 
different from any coloration found in the latter. The lack of a notch 
on the anterior edge of the second upper incisor seems also to be distinc- 
tive of assamensis. The type locality of assamensis is Assam, between 
Seebsaugor and Jeypur. The western Yunnan race may be described as 
follows. 


Anourosorex assamensis capito, new subspecies 


Type.—Adult male, skin and skull, No. 44510, American Museum of Natural 
History, from Mu-cheng, Salween drainage, Yunnan, 7000 feet altitude. February 
10,1917. R.C. Andrews and E. Heller. 

DeEscriptTion.—Size less than in typical assamensis, the skull smaller; belly with 
a distinct greenish or yellowish-gray wash. 

The dorsal surface of body is a dark mouse gray, a little darker than in the above 
race of squamipes, with a minutely peppered appearance. The long hairs on the rump 
just above the root of the tail are rusty-tipped, apparently the result of staining 
through the effect of some glandular secretion. Chin, throat, chest and belly gray 
tinged with greenish or yellowish. Pelage both above and below silvery when viewed 
from in front. Hind feet longer than tail. 

MEASUREMENTS.—The type measured by the collectors: length, 93 mm.; tail, 
16; hind foot, 16.5. The skull measures: greatest length, 25.8 mm.; basal length, 
23; palatal length, 12.5; upper tooth row, 11.7; mastoid width, 13.4; mandible, 13; 
lower tooth row, 10.8. 

ReMARKS.—Compared with A. squamipes capnias the striking differ- 
ence in color of the under surface, the larger size of the feet and skull, 
and the unnotched second upper incisor separate this form easily. The 
skull especially is more massive, with a much more prominent sagittal 
ridge. The seven specimens from the type locality are very uniform in 


the characters described. 


Sethe wat sik a este ae us : ee 
-y a all : easiest eas wn? a : bah , 
ih eat aah eae seach Neb aig si 


se ae nee Tati? 


> 


SE srces 
aes the oF Piet ae Fie he ae : 3 
. PAA bias: ak Ree ae Pion ae Se tot sear a 
as SPREE Sa ae ea ARIE Ae oe rea sant tae eas , 


a, y 
‘ 


‘ : * 
4 } net ee ASGor ots ; : 
ae 7 igs Wyn tie, Rare! ae a Ee ea op 2 ieee “, _s is 


tb 0 ; ; 
_ 7 By a bat h 2 
yy re re ‘Bag e Pee + ao ha? \ x oh 
Cy ; 
mint; - bi z eur e ¥ sag mee Oy va 
ES As 4 civeteth 3 “5 
i minh . - 
-_ rt a ; 
; . 
a, % 
-. , 
7 ae ye pie Sak r3. Ss a2 r 
5 ¥ ‘7 - 
a i ys 
’ spay, ex : 
, ; Gee Sm 
ry ce : Ary ; ee 
: , F ’ ‘ 
: # - be : 
wer 
a: . s : 3 : 
1 
: 4 
£ 4 
« 7 ie 
a , 
’ 
; *- 
; | 
| aa 
m 
r 
‘ 
" 
f 
{ ' 
i 
. ’ 
st 
‘ 
a 
' . 
en 
. it, 
: é 
i 1 a 
¢ ‘ 
he f = 
= 
Zs . 
roy, | 
ay 
- 
¥ 
. 
ait « | 
\ 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 101 Tre American Museum or Natura History Dee. 28, 1923 
, New York City : 


56.9,32(1181:51.7) 


NEW BATHYERGIDA FROM THE OLIGOCENE OF MONGOLIA! 
By W. D. Matrruew ano WALTER GRANGER 


The genera here described are from the Hsanda Gol formation in the 
Tsagan Nor basin of outer Mongolia, named and defined by Berkey and 
Granger in American Museum Novitates No. 77, May 25, 1923. They 
are from the red strata in the lower part of the variegated beds of that 
formation, the same horizon as the skull and other parts of Baluchitherium 
described by Osborn in Novitates No. 78, May 25, 1923, and from the 
same or neighboring localities. 

The age of the Hsanda Gol formation. was provisionally given as 
Miocene by Berkey and Granger, but the correlation of the fauna 
described in this and following articles shows that the variegated beds 
containing the Baluchitheritwm fauna are Oligocene, probably not later 
than Stampian. 

Baluchitherium was first discovered by Forster Cooper in the Bugti 
Hills, Baluchistan, in beds correlated by him as Upper Oligocene.2 The 
related or identical genus Indricotheritum was described by Borissiak in 
1915, from a formation which he correlated as Oligocene in the Kirghiz 
steppes north of Russian Turkestan. The third discovery, made by 
the Third Asiatic Expedition of the American Museum, was at Iren 
Dabasu on the Kalgan-Urga caravan trail in eastern Mongolia, in the 
Houldjin formation‘ associated with a few very fragmentary remains of 
other animals. The fourth discovery in the Hsanda Gol formation is in 
association with a large and varied fauna mostly of small mammals. The 
correlation of this fauna will be discussed in a later contribution. 

The rodents are the most abundant element of the collection. Many 
skulls, more or less complete, a few skeletons or parts of skeletons, and 
some thousands of jaws and fragments of jaws, were obtained by the 
Expedition in 1922 in the Tsagan Nor basin. They represent eleven 
species of nine genera, as follows: 


‘ ig tae of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
ution No 
2Cooper, C. Forster, 1911, Ann. Mag. Nat. Hist., (8) VIII, p. 711, and later articles. 
3Borissiak, A., 1915, Geol. Vestnik; 1917, Bull. Acad. Imp. Bei. Petrograd, (6) XI, p. 287. 
. eh and Berkey, 1922, Amer. Mus. Novit. No. 42, Aug. 7; Matthew and Granger, 1923, 
idem, No. 97. 


2 AMERICAN MUSEUM NOVITATES [No. 101 


Simplicidentata (Glires, sensu stricto) 
Tsaganomys altaicus, new genus and species 
Cyclomylus lohensis, new genus and species 
? Prosciurus lohiculus, new species 
Cricetops dormitor, new genus and species 
Tataromys plicidens, new genus and species 
Tataromys sigmodon, new species 
Karakoromys decessus, new genus and species 
Selenomys mimicus, new genus and species 
Eumys asiaticus, new species 

Duplicidentata (Lagomorpha) 
Desmatolagus gobiensis, new genus and species 
Desmatolagus robustus, new species 


The first two genera are the subject of this paper; the remaining 


rodents of this fauna will be described in the next contribution. 


An — 
As 

iN 
As 


ie a 


- 
-_ 


-<=s° 


= “te 
eS | 


Sa a= 


<7: 


etd 


Jp 
A.S4. 19019 = FF nat. size 


Fig. 1. Tsaganomys altaicus, skull and jaws, side view, natural size. No. 19019 
type specimen, characters partly supplemented from Nos. 19021, 19030 and 19033. 


Bathyergidze 


Tsaganomys altaicus, new genus and species 
Typr.—Amer. Mus. No. 19019, skull and lower jaws. 


ParatyPEs.—Nos. 19020, 19029, 19030, 19033, 19037, 19038, skulls, some with 
lower jaws associated. 


Horizon AnD Locauiry.—Oligocene, Hsanda Gol formation, red strata, Loh, 
Tsagan Nor basin, outer Mongolia. 


Dracnosis.—Skull short, wide, robust, the occiput very broad and pitched 


heavily forward with prominent occipital and sagittal crests. Infraorbital foramen 


1923] BATHYERGIDH FROM OLIGOCENE OF MONGOLIA 3 


small, oval, situated low down, apparently not traversed by any slip from the mas- 
seter, which has a scar for attachment on inferior face of spring of zygomatic arch as 
in Bathyergus (and Paramys, etc.). Above the root of the zygoma the orbits are 
built out laterally to a marked degree with a prominent sharp-edged crest in front, 
separating them from the muzzle. Although this crest in some ways resembles that of 
the sciuromorphs, it does not appear to have lodged a division of the masseter on its 
anterior face, but is more nearly analogous to the much less conspicuous preorbital 
crest of Bathyergus. Tympanic bulla broad and flattened anteriorly, extending 
postero-externally into a short ? meatus. Sagittal crest high, somewhat convex in 


Z 
WY 


A.M. 19019 TY, 


nat. S7ze | 


Fig. 2. Tsaganomys altaicus, skull, top view, natural size, type specimen. 


outline, short, owing to the extreme forward pitch of occiput; postorbital constriction 
pronounced, abrupt. Interorbital space nearly twice as wide as in Bathyergus; 
nasals narrow with straight sides, increasing slowly but uniformly in width from back 
to front. 

Incisors large, broad, with flat anterior faces ornamented with obscure longitu- 
dinal ridges. The socket for the upper pair forms a prominent bulge on the maxilla 
above the cheek teeth and the curvature is low in both upper and lower pair, so that 
they project strongly forward at the points. The socket of the lower pair is directly 
under the condyle of the lower jaw. Cheek teeth 4, very hypsodont, of persistent 


4 : AMERICAN MUSEUM NOVITATES [No. 101 


growth, without closed roots at any observed stage of wear, round or oval in cross- 
section, decreasing somewhat in diameter from p{ to m3. The crown pattern dis- 
appears at a very early stage of wear, as in Bathyergus, leaving only an encircling 
band of enamel. The crown pattern of the unworn lower molars appears to be one 
external and three internal inflections, the latter rapidly converted by wear into 
three shallow pockets and then disappearing. The structure of the unworn upper 
molars consists essentially of three shallow transversely-extended pockets, of which 
the center one is formed by ridges connecting protocone with paracone and meta- 
cone, the anterior and posterior ridges enclosed by anterior and posterior cingular 
crests, the first comparatively short, connecting protocone with parastyle, the second. 
much longer, sweeping around the margin from behind the protocone to meet the 
metacone. The cones, however, do not exist as such, the whole tooth being reduced 
to a series of narrow crests and shallow basins. Jaw massive, corresponding with 


Fig. 3. Tsaganomys altaicus, front view of skull, showing small antorbital forae 
men with wide crest external to it. No. 19021, natural size. 


Bathyergus in the relations of angle and coronoid so far as they are preserved. 

The milk dentition is well shown in No. 19023; dp? is a small peg-like tooth with 
simple blunted crown; dp‘ has almost exactly the form and pattern of m! but is of 
smaller size and presumably shorter crowned; dp, is similarly like m; in pattern, but 
somewhat smaller, narrower, and with the anterior and posterior fosse slightly im- 
pressed, the median inflections, both internal and external, more persistent and the 
external inflection directed more backward. : 

ReEMARKS.—This appears, if properly referable to the family, to be 
the first fossil record of Bathyergide, hitherto known from the recent 
Ethiopian fauna. It is by no means close to the living genera and should 
perhaps be distinguished as a separate subfamily, Tsaganomyine, on the 
short massive proportions of skull with heavy forward pitch of occiput, 
wide differences in otic region and some rather minor differences in teeth. 
It suggests the Asiatie ancestry of the family, although it cannot be 
considered as even approximately ancestral to the living genera. 


1923] BATHYERGIDA FROM OLIGOCENE OF MONGOLIA 5 


es, 


A.M. 19031 SS-_ 7 
oe GY, 


WOLt ~ 


Fig. 4. Tsaganomys altaicus, upper and lower teeth of left side, crown views, 
e twice natural size. Nos. 19033, 19031. 


Cyclomylus lohensis, new genus and species 


Type.—No. 19096, skull, badly preserved. 

ParatypPres.—No. 19098, upper and lower jaw; Nos. 19095, 19097, palates; No. 
19099, incomplete skull, badly preserved. All from same horizon and locality as the 
preceding species. 

D1acnosis.—Dentition apparently as in T'saganomys, molars of similar apical 
pattern early lost, but less hypsodont and of much smaller size, the roots closing when 
full grown. Skull much narrower and smaller, the proportions more as in Bathyergus 
but the cranium not so flat and broad as in that genus; zygomatic arch of moderate 
proportions, the orbits not built out asin Tsaganomys, infraorbital foramen apparently 
small and low set, masseter attachment on under face of zygomata limited by a scar. 
The angle of the lower jaw springs from the side as in J’saganomys, so far as the speci- 
mens show, but its exact relations are not so fully demonstrated. 

The permanent premolar is preceded by a minute, simple dp? and a large round 
molariform dp‘, as in Tsaganomys. 


i oi 
ne 


Fl pa 
es to 


mah ah he 


Nae eee 
BE en 


rh “/ mae o- Tarra. 
were! tigrt-altte. co Obs 5 iy “ORE einhecofoty 
: Dy wae (cy moe 
ie ” Owe 
4 > :, . a J ™ 
‘ at 
. < ‘ ¥ 7 BP vite » 
‘ 
4 i ; 
~ 
: " 
Lory E 
> 
j 
i, e i ‘ ‘ ~ ¥ 
‘\ j 
‘ j . . 
ore 
Ly . 
. 
‘ 
‘ 
’ / 
i . 
” - > 
4 — 
a f 
. - i 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 102 Tuer AMERICAN MusEUM OF NATURAL History Dec. ol, 1923 
New York City 


56.9,32(1181; 51.7) 


NINE NEW RODENTS FROM THE OLIGOCENE OF MONGOLIA! 
By W. D. MatrHew AND WALTER GRANGER 


In the preceding number of Novitates eleven new species of rodents 
from the Hsanda Gol formation of Mongolia were listed and two of them 
described. The remaining rodents are described in this article. 


Cricetopide 


Cricetops dormitor, new genus and species 


Typre.—No. 19054, skull, lower jaw and fore foot. 

PARATYPES.—Several more or less complete skulls and numerous upper and lower 
jaws. 

Horizon anp Locariry.—Oligocene, Hsanda Gol formation, near Loh in the 
Tsagan Nor basin, outer Mongolia. 

Dracnosis.—Cheek teeth 3, the upper series decreasing in length and width 
from first to third, the lower series subequal and of nearly square outline. Brachy- 
dont crowns, the cusps arranged in pairs, two pair each on m$ and m3, but a well- 
developed anterior pair on m! and a rudimentary anterior heel on m:. In the lower 
teeth the outer cusps tend to be crescentic, the inner are nearly round; the upper 
teeth reverse this arrangement. The skull is cricetoid in proportions, rather long and 
narrow, arches well preserved on the type and several other specimens show that the 
infraorbital foramen was round and of rather large size, resembling some of the 
dormice, e. g., Graphiurus, also Pseudosciurus and other Oligocene genera and, to a 
less extent, the Dipodide, the masseteric scar on the zygoma wholly beneath it and 
defined by a clear-cut margin. 


The teeth are very like those of Cricetus in proportions and pattern, 
to such a degree that a true affinity rather than parallelism may be in- 
dicated. The front of the zygoma, however, is unlike any true myo- 
morphs but approaches the primitive construction which is universal in 
the Eocene, prevalent in the Oligocene, and preserved among the dor- 
mice (not in Myozxus) in Anomalurus, Haplodontia and Bathyergus with 
less alteration than in other modern rodents. The Dipodide have the 
i. o. f. greatly enlarged; cricetids and other myomorphs have the mas- 
seteric attachment on the zygoma extended forward and upward in a 
plate. 


_\Publications of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
bution No. 20. 


Fig. 1. Cricetops dormitor, skull and lower jaw, side view, twice natural size. 
Type specimen, No. 19054. 


A.M. 19054 
7ype 


Fig. 2. Cricetops dormitor, 
skull, top view, twice natural 
size. Type specimen, No. 19054, 
the occiput supplemented from 
No. 19051. 


1923] NEW RODENTS FROM OLIGOCENE OF MONGOLIA 3 


This genus would apparently fall into the Dipodoide of Miller and 
Gidley’s classification and, if we understand correctly the assumptions 
underlying their arrangement, could have nothing to do with the Cri- 
cetide. It appears inadvisable to accept these assumptions until their 
validity has been more conclusively proven; in some respects they do 


A.M. /905/ 


Fig. 3. Cricetops dormitor, anterior and posterior views of skull, twice natural 
size; A, anterior view, showing character of antorbital foramen, from the type; 
B, occiput from No. 19051. 


Fig. 4. Cricetops dormitor, upper and lower teeth, crown views, enlarged to four 
diameters: A, upper teeth and roof of zygomatic arch from the type specimen; B, 
lower teeth, No. 19059. 


not seem to us conformant to the general tenor of the evidence of fossil 
rodents, and have compelled these authors—as they admit at the begin- 
ning of their classification—to deny, practically, that any of the extinct 
types of rodents are either directly or approximately ancestral to any of 
the existing types, and to assign all of the numerous resemblances in 


4 AMERICAN MUSEUM NOVITATES [No. 102 


dentition, skull and skeleton which would suggest a more or less ancestral 
relationship, to parallel—or rather, convergent—evolution. No interpre- 
tation of the affinities of existing and extinct rodents can avoid the 
assumption of a large amount of parallelism, but it would seem that 
Messrs. Miller and Gidley have carried it to improbable extremes in 
support of certain preconceived theories of what can or cannot occur in 
the modification of the zygomatic and dental construction, and that a 
reasonable application of the law of probabilities to what we know of 
fossil rodents would lead to some modification of these theories and a 
resultant simplifying of their otherwise admirable revision, which we 
fully recognize as based upon a most thorough and complete review of the 
order, particularly as including the extinct as well as the existing genera. 
It is proper to emphasize, however, that they have not yet published the 
evidence in support of their conclusions, and this may prove to overcome 
the difficulties which we see in accepting them without certain modifica- 
tions. The fauna herein described will add materially to the fossil 
evidence and in our opinion may make it necessary to reconsider to some 
extent the very complex and difficult problem of the true affinities and 
evolution of the major groups within the order. That, however, is an 
undertaking much beyond the scope of the present contribution. 

It does not appear advisable to assign this genus to any recognized 
family, as its systematic position turns upon the above problem. It 
might be placed with the dormice save that this family includes, auct. 
Miller and Gidley, two groups of quite diverse affinities, and Cricetops 
appears to be quite as diverse from either as they are from each other. 
It might be referred to Pseudosciuride or Komyide on the zygomatic 
characters, but the teeth are wholly unlike any of the genera of those 
families. The same objection applies to its reference to the Theridomyi- 
dee. It is still less possible to associate it with Ischyromys or Paramys 
or with the Dipodide. Anomaluride in the broad scope given to the 
family by Winge and Schlosser would perhaps include this and the fol- 
lowing genera, but it would require at least subfamily distinction. With 
it, in the same broad sense, might be placed a number of new genera from 
the Hsanda Gol, which are known only from upper and lower jaws and 
appear to have some resemblance in masseter attachments, although, 
except for Selenomys, they are quite diverse in dentition. Pending a 
reconsideration of the relations of these anomaluroid or dipodoid genera, 
it appears convenient to place Cricetops and Selenomys in a separate 
family and refer Karakoromys and Tataromys to the Eomyide. 


1923] NEW RODENTS FROM OLIGOCENE OF MONGOLIA 5 


Selenomys mimicus, new genus and species 


Typre.—No. 19085, an upper jaw. 

ParatyPes.—Nos. 19086-19093, a series of upper and lower jaws. 

Horizon AND Locauiry.—Hsanda Gol formation, near Loh, Mongolia. 

DiaGcnosis.—Three subequal molar teeth, no premolars. (There is some doubt 
about the absence of the upper premolar.) Crowns of molars moderately high, each 
composed of four inward-facing crescents, an anterior and a posterior pair, as in 
ruminant molars. The lower jaw in front of the molars is rather thick, not deep, 
moderately long; incisor not preserved. Angle only partly preserved, appears to be 
straight, as in Myomorpha generally. 


Fig. 5. Selenomys mimicus, upper and lower teeth four diameters: A, upper 
teeth, right side; B, lower teeth, left side, crown view; B1, external view of left ramus 
of lower jaw. All from the type specimen, No. 19085. 


We do not know of anything near to this peculiar genus. It is 
provisionally associated with Cricetops, which has some suggestion of 
approach in pattern, but the jaw proportions are quite different. Cteno- 
dactylus has a pattern which sugggests derivation from something of this 
type but it retains the premolar in upper and lower jaw which Selenomys 
has lost. So far as preserved, the character of the zygomatic arch accords 
with Cricetops; the genus is clearly not a myomorph and probably be- 
longs in the same group as Cricetops. 


? Eomyide 


Tataromys plicidens, new genus and species 


Typr.—No. 19082, a palate with p*m‘, r. and 1. 

ParatyPres.—Nos. 19081, 19083, 19084, upper and lower jaws. 

Horizon AND Locatiry.—Hsanda Gol formation, Loh, Mongolia. 

DraGnosis.—Premolar smaller than the molars, trigonal with three submarginal 
crests, not at all molariform in pattern but of fair size. Molars with two principal 


6 AMERICAN MUSEUM NOVITATES [No. 102 


Fig. 6. Tataromys plicidens, upper teeth, right side, enlarged to four diameters. 
From type specimen, No. 19082. 


transverse crests connected by an external commissure; on m? and m* supplementary 
anterior and posterior crests obliquely inward from the main crests opposite com- 
missure. The lower molars reverse this pattern in the usual manner but the arrange- 
ment is less regular. 


Tataromys sigmodon, new species 


Typr.—No. 19079, a palate. 

Horizon AND Locatiry.—Hsanda Gol formation, Loh, Mongolia. 

Dragenosis.—Dentition and details of construction of teeth very close to T. 
plicidens but of smaller size, length of p*m?=8.8 mm. 


Karakoromys decessus, new genus and species 


Tyre.—No. 19070, lower jaw, both rami with cheek teeth and left incisor 
complete. 

Horizon AND Locauity.—Hsanda Gol formation, red beds, Loh, Tsagan Nor 
basin, Mongolia. 


Fig. 7. Karakoromys decessus, lower jaw, type specimen, four times natural 
size, external view of left ramus and crown view of cheek teeth. 


DiacGnosis.—P, present, much smaller and simpler than molars. Molars in- 
creasing slightly in size from first to third, moderately brachyodont, longer than wide, 
the crowns with high transverse crests, a principal anterior (trigonid) and posterior 
(talonid) crest connected by a commissure, and a hypoconulid crest extending postero- 
internally from a point on the outer half of the talonid crest. P, with a single trans- 
verse crest and a wide but short posterior heel. 


1923] NEW RODENTS FROM OLIGOCENE OF MONGOLIA fj 


This genus appears to be nearly related to Tataromys in molar 
construction. 


Paramyide 
Prosciurus lohiculus, new species 


Type.—No. 19100, upper jaw with p*-m?. 

Horizon anv Locauiry.—Oligocene, Hsanda Gol formation, Loh, Mongolia. 

Diacnosis.—Upper molars and p‘ with crests arranged much as in P. vetustus, 
but higher and lacking any trace of mesostyle on external margin between the crests. 
Size about a fourth larger than P. vetustus. 


AM. 19100 Type 


Fig. 8. Prosciurus ? lohiculus, upper jaw, four times natural size,type specimen, 
No. 19100. 


This species is referred to Prosciurus provisionally. It is too im- 
perfectly known for satisfactory allocation. It equally resembles in 
dentition several modern Sciurine genera or subgenera, but the masseteric 
scar appears to be confined to the inferior face of the zygomatic process 
of the maxilla, much as in Prosciurus, Paramys and related genera. 


Eumys asiaticus, new species 


Typre.—No. 19094, upper jaw with m!-*. 

Horizon and Locauiry.—Oligocene, Hsanda Gol formation, near Loh, 
Mongolia. 

Diaenosis.—Tooth pattern much as in FE. elegans, masseteric plate of zygoma 
typically myomorph and closely resembling that of Humys and Cricetodon. Size: 
m!-3=5,.5 mm.; mi-3=6 mm. 

This appears to be the only true myomorph rodent in the Hsanda 
Gol fauna. It is closely allied in tooth pattern to Humys of the American 
and Cricetodon of the European Oligocene. The Miocene species of 
Cricetodon are, so far as I have examined, decidedly more advanced 


towards Cricetus. 


8 AMERICAN MUSEUM NOVITATES [No. 102 


Fig. 9. Humys asiaticus, upper and lower jaw fragments, four times natural 
size: A, upper jaw, left side, with m !-3, showing also the obliquely-pitched masseter 
plate in front of zygomatic arch; B, lower jaw, right side. The anterior end of the 
upper jaw faces to left, of lower jaw to right. Type specimen, No. 19094. 


Leporide 


Desmatolagus gobiensis, new genus and species 


Type.—No. 19103, upper jaw with p?- m3. 
PaRATYPES.—Various upper and lower jaws. 


Fig. 10. Desmatolagus gobiensis, upper and lower jaw, four times natural size: 
A, crown view of upper cheek teeth; B, crown view of lower teeth; B1 external view 
of lower jaw, right side. Type specimen, No. 19102. 


1923] NEW RODENTS FROM OLIGOCENE OF MONGOLIA a 


A. 7.19/16 Type 


_# 
/ 


Tig 11. Desmatolagus robustus, upper jaw, crown view, four times natural 
size, showing p* - m? and roots of p? and m’. Type specimen, No. 19116. 


ATL SILT: 


Fig. 12. Desmatolagus robustus, lower jaw, superior and external views, four 
times natural size. No. 19117. 


10 AMERICAN MUSEUM NOVITATES [No. 102 


Horizon AND Locauiry.—Oligocene, Hsanda Gol formation, red beds, Loh, 
outer Mongolia. 

DiaGcnosis.—Cheek teeth £ as in Leporide but the first and last teeth of the 
series (ps, m3) greatly reduced. Although it retains the formula of the Leporide, the 
genus undoubtedly is related to the Ochotonide and may be considered ancestral to 
some of the genera (not to Titanomys, which is stated to have rooted molars'). It is 
placed in the Leporide provisionally upon the formal distinction of the number of 
cheek teeth. 


Desmatolagus robustus, new species 


’ Typr.—No. 19116, lower jaw with p3-mo. 
ParatyrPe.—No. 19116a, lower jaw with ps-m3. 
Horizon AND Locatiry.—Oligocene, Hsanda Gol formation, red beds, Loh, 
Outer Mongolia. 
DraGnosis.—Size, one-half greater than the preceding, to which it is in other 
respects nearly alike. 


'The published descriptions and figures of Titanomys do not seem to prove, however, that the 
permanent teeth have roots. In Palxolagus the milk molars have roots but the permanent teeth 
are rootless. 


4 


& 


- 


itt we * 


k 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 104 Tur American Museum or Natura History Jan. 15, 1924 
New York City 


56.9,74(118:51.7) 


NEW CARNIVORA FROM THE TERTIARY OF MONGOLIA! 


By W. D. MatrHew AND WALTER GRANGER 


Most of the species here described are from the Oligocene Baluchi- 
theritum zone, Hsanda Gol formation, in the Tsagan Nor basin, in outer 
Mongolia. One is from the Eocene Irdin Manha Protitanotherium zone 
in the Iren Dabasu basin, eastern Mongolia. 


Paracynohyzenodon morrisi, new species 


Typrre.—No. 19160, lower jaw, immature, with m; unworn, p3 and m; preformed, 
Irdin Manha beds. Found by F. K. Morris, 23 miles south of Iren Dabasu. 

CHARACTERS.—The molars, as in Dr. Martin’s type,” are distinguished from those 
of Cynohyenodon and Tritemnodon by the more compressed and secant character of 
the talonids. The species appears to be distinguished from the genotype, P. schlosseri 
of the Phosphorites, by the reduced and crowded premolars. 


Hyznodon pervagus, new species 

Typr.—No. 19005, part of lower jaw. 

PaRATYPES.—Nos. 19006, 19015, 19125, 1926, parts of jaws; No. 19002, hind 
limbs and feet. 

Horizon AND Locautiry.—Oligocene, Hsanda Gol formation, Loh, Mongolia. 

D1acnosis.—Species of moderate size in the genus, about equalling H. heberti 
and cruentus. So far as comparisons can be made, it belongs among the shorter- 
jawed species. Distinguished from heberti by entire lack of anterior accessory cusps 
on premolars, by larger relative size of mi, ete. The hind limb and foot bones are 
finely preserved and agree very closely with H. cruentus in size and in all details of 
construction. 


RemMarks.—Hyexnodon is widespread in the Upper Eocene and older 
Oligocene of Europe (Débruge, Phosphorites, older Bohnerzen, Ronzon), 
in the Lower and Middle Oligocene of America (Tvtanotheriwm and 
Oreodon zones of the White River) and in the Lower Oligocene of Egypt 
(Fluvio-marine beds). Its occurrence in the Hsanda Gol points to a 
rather early Oligocene age for this formation. 


Didymoconus colgatei, new genus and species 


Typr.—No. 19124, skull and jaws. 
ParatyPres.—Nos. 19003, 19004, lower jaws. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 21. 

2Martin, Rudolf, 1906, ‘Revision obereoc. u. unterolig. Creodonten Europas.’ Rev. Suisse de 
Zool., XIV, p. 424. Paracynohyenodon schlosseri. 


2 AMERICAN MUSEUM NOVITATES [No. 104 


Horizon AND Locatiry.—Oligocene, Hsanda Gol formation. 

Generic Dracgnosis.—Dentition 4. Incisors small, crowded, ? reduced; 
canines of normal carnivore type, p¢ molariform; p? and p»-3 simple, two-rooted, com- 
pressed, with sharp cusp and small posterior heel, p* with two external cusps. Molars 
of leptictid type, the trigonid of lower molars composed of two high round, twinned 
cusps and a small low paraconid, the heel rising sharply at posterior margin to a trans- 
verse crest, incofnpletely separated into hypoconid and entoconid. Mz slightly 
larger than mi, heel narrower than trigonid; m; with heel and trigonid of equal 
width; ps quite molariform (but the crown less worn and the tooth less fully emerged 


=> 


==. : Ss 


\s 

1 
\\ RR 
(i 


chant /4. 1/9/24 Type 
ree WES = 2 
ey e 
nat. size AMM. 19124 Type 


—_ 


Ge GRACE ear 


é Fig. 2 
Ay ~ yp he 
ZS ‘ 


Fig. 1 ‘ 
Fig. 1. Didymoconus colgatei, skull and jaws, type specimen, side and top views. 
The original is somewhat broken and crushed and the distortion has been corrected 
in the drawing. Oligocene, Hsanda Gol formation, Mongolia. Natural size. 
Fig. 2. Didymoconus colgatei, upper dentition, cm. Principally from the 
type skull, supplemented by comparison of several referred specimens. Twice 
natural size. 


from the jaw than the one behind it, and therefore assigned to the premolar series), 
the heel wider than trigonid, and paraconid a little stronger than on the true molars. 
Posterior mental foramen under p;. Jaw short and deep with strongly sutured 
symphysis. 

Upper molars transversely extended, m! consisting of a pair of separate subequal 
outer cusps, an inner conical protocone opposite the paracone and a strongly 
developed posterior cingular crest extending somewhat further inward than the 
protocone. On p?‘ this posterior crest is represented chiefly by a posterointernal cusp 
with a rather rudimentary cingular crest extending from it toward the base of the 
metacone; the external stylar cusps, very rudimentaryon m!, are quite distinct and 
the external cingulum, distinct on m!, is obsolete. P* has two external cusps, the 
metacone much smaller. 

Spreciric Dracnosis.—Size of Spilogale, c-m,=28 mm.; lower jaw, shallower, 
depth below m:=9 mm.; lower canine, comparatively small and slender. 


1924] NEW CARNIVORA FROM MONGOLIA 3 


Didymoconus berkeyi, new species 


Typre.—No. 19001, lower jaws, from the same horizon and locality as the 
preceding. 

Speciric DraGnosis.—Size of MW Hes c-m2;=3.5 mm.; lower jaw, deep, robust; 
lower canines, very large and stout. 

Remarks.—This genus is unlike any known Oxyenide, and in some 
respects it has a marked resemblance to Mesonychide, in others to 
Leptictide. The formula is not so positively determined as one would 
like, nevertheless it seems almost sure that the first molarifor m tooth is a 
premolar (as in Leptictide). The true molars have almost lost their 
carnivore construction, nevertheless one can see in the pattern that it is a 
derivative of the oxyzenid type of carnassiform teeth with the carnassial 
socket between m' and m’; the great posterior crest of m! is evidently an 
exaggerated cingulum and without it the tooth would be much of the 
Limnocyon pattern. 


A./4./900/ Type 


Fig. 3. Didymoconus berkeyi, lower jaw, type specimen, external view and 
crown view of teeth, p3-m.. Twice natural size. 


In the lower molars the twinned conical cusps of the trigonid are 
leptictid in type, but the same construction is approached in Dissacus 
and Hapalodectes of the Mesonychide and in Apterodon among the 
hyenodonts. The short, heavy jaw, stout canines and massive compact 
symphysis are typically oxyznid, and the two subequal molars are as in 
the limnocyonines. 


4 AMERICAN MUSEUM NOVITATES [No. 104 


Amphicticeps shackelfordi, new genus and species 


Type.—No. 19010, a skull. 
Horizon AND Locatiry.—Oligocene, Hsanda Gol formation, Loh, Mongolia. 


DiacGNnosis.—Dentition “5. Canines of moderate size. Premolar region rather 
short, premolars somewhat reduced, simple, stout, much as in Cynodon. P# fully 
carnassiform, protocone (deuterocone) anterointernal, well developed, forming a low, 
broad inner heel; no parastyle. M? large, much extended transverely, paracone close 
to antero-external margin, metacone only slightly smaller, more internal in position; 
protocone low and a heavy inner cingulum. M2? quite small, aligned with inner margin 
of m!, apparently not extending beyond inner half of the preceding tooth, the roots 
connate or single. Cranium wide and rather short, with heavy sagittal and occipital 
crests; basicranial region wide and short, tympanic bulla incomplete or loosely at- 
tached, paroccipital processes free, directed backward. Mastoid processes prominent, 
flattened and projecting laterally. 


The lower jaw is very like that of Cynodon, except that the carnassial has a 
narrower and shorter heel with more distinct hypoconid crest and m; is absent. Me 
is not preserved in our specimens; its alveolus indicates a tooth of about the same size 
and proportions as in Cynodon, with connate roots. 


A./7. 190/10 Type 


Fig. 4. Amphicticeps shackelfordi, upper jaw with the first, second and fourth 
premolars and first true molar preserved. From the type skull, No. 19010. 
Twice natural size. 


This genus is intermediate between the cynodontoid and steno- 
plesictoid groups of the Phosphorite fauna. It has the sharply reduced 
post-carnassial dentition of the latter with the short, heavy precarnassial 
dentition of the former. It is not close to any one genus with which I 
have made comparisons and might be regarded as a highly progressive 
miacid rather than as a member of any of the existing families of fissipede 
carnivora. 


ten) 


NEW CARNIVORA FROM MONGOLIA 


1924] 


De 
janees \ = Yi 


‘ 


WA 
AW 
\ \ 


"i We 
Up. Be 
=) 


\ Re I = 


/7.190/0 Type 


eae 


mat. size 


top and palatal views. 


side, 


Amphicticeps shackelfordi, type skull, 


Fig. 5. 
Natural size. 


6 AMERICAN MUSEUM NOVITATES [No. 104 


A../90/3 Type 


19004 Type 


Fig. 6. Jaws of Carnivora from Hsanda Gol formation, external and crown views. 
A, ?Cynodictis elegans; B, Bunelurus parvulus; C, Bunelurus ulysses; D, Cynodon 
(Pachycynodon) teilhardi; E, Paleoprionodon gracilis; F, Viverravus constans. All 
twice natural size. 


1924] NEW CARNIVORA FROM MONGOLIA é 


Paleoprionodon gracilis, new species 


Type.—No. 19123, lower teeth and parts of skeleton. 
_ Horizon anv Locatiry.—Oligocene, Hsanda Gol beds, Loh, Mongolia. 
Dracnosis.—Carnassial compressed, cat-like, with metaconid much reduced and 
heel vestigial. Mes very small, narrow and elongate, with flattened trigonid of three 
low cusps and a trenchant heel. P, large, compressed much as in Felis domestica. 
Upper and lower canines subequal, very much alike, of moderate size, long, sharp- 
pointed, not compressed. Limb bones long and slender, humerus expanded trans- 


A.NI./9/235 Type 


nat. Size 


Fig. 7. Palzoprionodon gracilis, limb bones and metatarsal of the type speci- 
men (teeth shown in Fig. 6); outer views of humerus, radius and ulna, anterior view 
of third metatarsal. Natural size. 


versely at distal end with strong epicondylar bridge. Radius slender, ulna wide and 
flattened at proximal half of shaft, the distal half triangular, considerably less than 
radius in sectional] area. 


Astragalus with narrow deep trochlea, the inner crest well developed. No fibular 
facet on caleaneum. Metatarsals long and slender; mt. I, vestigial or absent. 


8 AMERICAN MUSEUM NOVITATES [No. 104 


This species agrees with the Phosphorite genus in dentition, so far as 
known, and in the character and proportions of the limbs and feet as 
figured and described by Schlosser; its reference, however, is provisional 
until the dentition is better known. 


Bunezlurus ulysses, new species 


Typr.—No. 19004, left ramus of lower jaw with ps-m: 1 complete. 

Horizon AND Locatity.—Hsanda Gol formation, Loh, Mongolia. 

D1aGcnosis.—Dentition ¢1, pa, M2. First premolar one-rooted, others two-rooted, 
the fourth with small accessory cusp. Carnassial without metaconid, heel narrow, 
trenchant; m, small, two-rooted with narrow trenchant crown. Length c-m», 
estimated, 25 mm.; ps-m» 12.5 mm. 


Bunelurus parvulus, new species 

Typr.—No. 19013, part of lower jaw from Hsanda Gol formation. 

DiacGnosis.—Very like the preceding species, but smaller; ps-m2=9.5 mm. 

We have referred these species to the American genus Bunezlurus, 
which is separable from Palxogale by retention of a minute m? in the 
upper jaw. As the upper dentition of the Mongolian species is unknown, 
they might be referred to Palxogale, but the reduction of mg is relatively 
greater than in P. felina, conforming somewhat better with Bunzlurus. 


Cynodon (Pachycynodon) teilhardi, new species 


Typr.—No. 19007, lower jaw fragment with m1: and alveolus of ms. 

Horizon AND Locatiry.—Oligocene, Hsanda Gol formation, Loh, Mongolia. 

Dracnosis.—Size of Amphicticeps shackelfordi. Carnassial somewhat less 
robust and with larger and longer heel, the heel as wide as the trigonid surface, a 
shallow basin with wrinkles radiating from anterointernal notch to the marginal 
crests. | M. subquadrate with proto- and metaconid cusp, hypoconid cusps some- 
what smaller, and an internal and posterior marginal crest enclosing a small basin. 
M; smaller than m., the crown not preserved, two closely approximate roots. 


REMARKS.—This species can be referred only provisionally until 
better specimens are available. It appears to fall within Pachycynodon 
rather than the typical Cynodon, by Teilhard’s key to the Phosphorite 
genera. 


Cynodictis ? elegans, new species 


Typr.—No. 19016, anterior part of the lower jaws with the canine and premolars 
preserved. 

Horizon AnD Locatiry.—Oligocene, Hsanda Gol formation, Loh, Mongolia. 

Dracnosis.—Size about that of the smaller individuals of C. compressidens, but 
distinguished from this species, as also from the American “ Cynodictis,”’ by the simple 
compressed ps without accessory cusps. The accessory cusp of ps, is strong, well 


1924] NEW CARNIVORA FROM MONGOLIA 9 


separated, and somewhat external in position. P; is single-rooted with compressed 
crown, anteriorly pitched and recurved at the tip; ps2 is two-rooted, nearly as large 
as ps, both being compressed simple crowns, p, with some forward pitching and 
recurving tip, ps nearly upright. The canine is quite small, slender, the jaw shallow 
and thin with loose symphysis extending back to the middle of ps. 

ReMaArkKs.—This species is provisionally referred in absence of the 
molar teeth. See below under Viverravus. 


Viverravus constans, new species 


Typr.—No. 19130, part of lower jaw with m-2 preserved. 

Horizon anp Locauiry.—Oligocene, Hsanda Gol formation, Loh, Mongolia. 

D1acnosis.—Size somewhat less than V. sicarius; the teeth show the generic 
characters in the high, somewhat compressed trigonid with angulate cusps, pr¢ over- 
topping the others, heel small, sharply trenchant; m, tuberculosectorial with rather 
high trigonal trigonid of three subequal cusps and narrow trenchant heel; ms absent. 
Considerably smaller than V. antiquus of the Phosphorites. 

Remarks.—The reference of this species to Viverravus is necessarily 
provisional, but it agrees quite closely so far as it goes. It is quite pos- 
sible, however, that the anterior portion of lower jaws provisionally re- 
ferred to Cynodictis is the same species as No. 19130; in which case it 
certainly is not Viverravus, as p3 of that genus has always a strong acces- 
sory cusp. 


yy as 


Say a fe ; 
ots oni 


i€ 
{ 
o< ¥ b: 
rs . ‘ . 
; 
4 \ 
‘ 
,, 
* i 
’ . 
, 
4 
& 
, 
{ 
‘ 
® 
5 
' 
, ' 
‘ 
* 
5 
4 
bi ~ 
1 
' 
2 ay 
. a) 
7 a 
° 4 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 105 Tue AMERICAN MusEum or Natura History Jan. 18, 1924 
New York City 


56.9(118:51.7) 
NEW INSECTIVORES AND RUMINANTS FROM THE TER- 
TIARY OF MONGOLIA, WITH REMARKS ON THE 
CORRELATION! 


By W. D. MatrHew anp WALTER GRANGER 


Tupaiodon morrisi, new genus and species 


Type.—No. 19134, upper jaw and part of skull, parts of lower jaw with mo_3. 

Horizon AnD Locariry.—Hsanda Gol formation, Red beds, Loh: 

Diacnosis.—Upper molars and premolars somewhat resembling those of 
Ptilocercus, but with larger and more separate hypocones on m'.  Incisors unknown. 
Canine and p'? small, two-rooted, with short stout cusp and rudimentary heel. 
P*4 much larger with strong internal cusps, p* with strong metastyle blade and rudi- 
mentary tetartocone. M!? sub-quadrate with high, rather angular cusps, meta- 
style distinct, hypocone more developed and separated than in Ptilocercus, mé tri- 
angular with no metastyle or hypocone, proportioned about as in Ptilocercus. Lower 


Fig. 1.  Tuwpaiodon morrisi, right upper jaw, c-m’, palatal and external views. 
Type specimen, No. 19134, Hsanda Gol formation, Mongolia. Three times natural 
size. 
molars with rather short wide trigonid of three angular cusps, pr“ slightly highest, 
talonid as wide as trigonid, deeply basined with cusps at posterior angles, not as high 
as trigonid cusps. Infraorbital foramen above p?; root of zygoma above m?. 

Size about that of the European mole or the short-tailed shrews. C-m?=13 mm. 

While provisionally referred to the Tupaiide, the true affinities of 
this genus are uncertain. It is excluded from pegs by the single 


‘Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 22. 


2 AMERICAN MUSEUM NOVITATES [No. 105 


outer cusp of p*, prominence of paraconids and other characters, from 
the Soricoidea by the comparatively large and complex p**. 


? Tupaiodon minutus, new species 


Typr.—No. 19135, lower jaw fragment, ps-m, right side. 

Horizon AND Locatiry.—Same as preceding species. 

Dracenosis. Somewhat smaller than 7. morrisi, the molars of nearly the same 
size, but narrower, with distinct external cingulum; anterior part of lower jaw much 
more slender and weak; p3 two-rooted, simple, with single acute uncompressed cusp, 
minute anterior basal cusp and well-developed posterior basal cingulum; p4 sub- 
molariform but smaller than molars, having two well-separated central cusps, a 
lower anterior cusp, and a broad low cingulum encircling posterior end of tooth. 
Length p3-m,=9 mm. 


Paleoscaptor acridens, new genus and species 


Typr.—No. 19138, a lower jaw with ps-ms, associated with two others. 

Horizon AND Locauiry.—Hsanda Gol formation, Red beds, 15 miles east of 
Loh, Mongolia. 

Dragnosis.—Dentition probably 27.33. First incisor (? i) enlarged, a long 
slender, simple procumbent tooth, enamelled, without heel or serrations; second in- 
cisor not preserved, the alveolus quite small, round oval; the canine small, root semi- 
double, crown premolariform and extended forward, followed by a similar but some- 


A.M 19/38 Typ 


QE 


se 
Z 


Fig. 2. Palzoscaptor acridens, lower jaw, left ramus, external view and crown 
view of lower teeth, ps-m3. Type specimen, No. 19138. Three times natural size. 


what smaller and less forwardly extended tooth with semi-double root; the last pre- 
molar two-rooted, small, sub-molariform, with trigonid-like main cusp and small 
transversely-crested heel; m; much larger, with trigonid longer than wide, of three 
sharply angulate cusps, and deeply basined heel with acute cusps at posterior angles. 
M» considerably smaller than m,, similar trigonid relatively lower and shorter; mg 
much reduced, with low single-cusped heel. The heel cusp is of considerable size on 
the type, but in other specimens it is vestigial; there is also a considerable range in 
size in the twenty or more fragmentary jaws that represent the genus; but we 
are unable to fix associated constant distinctions to divide the material into species. 


1924] MONGOLIAN INSECTIVORES AND RUMINANTS 3 


Paleoscaptor rectus, new species 


Typr.—No. 19146, lower jaw with mo-s. 
~Horizon AND Locauiry.—Oligocene, Hsanda Gol formation, Loh, Mongolia. 
Dracnosis.—Teeth, so far as shown, similar to those of P. acridens, except for 
considerably larger size and greater reduction of ms, which has no heel. Lower jaw 
angle long, flat, projecting backwards in a rather slender process upturned towards 


tip. 

Only one specimen of this species has been discovered. 

AFFINITIES OF Palxoscaptor.—This genus is provisionally placed 
with the Soricidse, but it is decidedly more generalized than any of the 
existing genera or of the better known extinct genera. The reduction of 
m3 is greater than in any of the modern genera, either of moles or shrews; 
the primitive brachyodont trituberculy also distinguishes it from most 
modern Soricoidea, likewise from Proscalops; and the reduction of the 
anterior teeth is greater than in the moles, less than in the shrews, while 
the enlarged incisor shows no trace of the heel or serrated edge developed 
in the typical Soricid. It is possible that some of the imperfectly known 
Oligocene Soricoidea may approach it; none of the Bridger soricoids 
come very near to it. On the whole, it might very well represent an 
ancestral type from which the soricid genera could be derived. 


Eumeryx culminis, new genus and species 


Typr.—No. 19147, fragments of upper and lower jaws and foot bones, probably 
of several individuals found together. 

Horizon AND Locatiry.—Hsanda Gol, 10 miles west of Loh, ‘‘Grand Canon,” 
probably below lava. 

Dracnosis.—General characters of the primitive Cervide. Upper canine, a 
large compressed, slender tusk. Molars brachyodont, the upper molars with promi- 
nent styles and anterior rib but no trace of posterior rib, the lower molars with slight 
traces of a “paleomeryx-fold.”’ Four lower premolars in series, p: small, one-rooted 
and simple, ps-4 compressed, two-rooted, with inner crests nearly as in Blastomeryz, 
more developed than in Prodremotherium. Navicular and cuboid united, inner cunei- 
form separate. Median metacarpals and metatarsals united into cannon-bones, but 
the distal keels not extended over the dorsal surface. Fifth metacarpal and meta- 
tarsal coéssified proximally, second separate. 

Size about that of Blastomeryx advena; but there are a few specimens of con- 
siderably larger size that may prove to be a distinct species. 


The above diagnosis is based upon a large number of fragments of 
jaws and bones of the skeleton of many individuals. There can be little 
doubt, however, that they belong all to one genus and mostly to one 
species, which is the only artiodactyl present. It is an interesting type, 
as representing a stage of ruminant evolution intermediate between 


4 AMERICAN MUSEUM NOVITATES [No. 105 


Leptomeryx and Blastomeryzx in the structure of teeth and feet, somewhat 
more progressive in premolar construction than Amphitragulus and 
Prodremotherium of the Phosphorites but of smaller size and differing 
too much to refer it to any of the described genera. In the Phosphorites 
genera, according to Schlosser, the metacarpals are not united into a 
cannon-bone and the distal keels are less developed than in Humeryz. 


In the Lower Miocene species of Blastomeryx the teeth are very like 
those of the new genus except for loss of pi, and the median metacarpals 


A./9. 19147 Type 


nat. Size 


Fig. 3. Humeryx culminis, fragment of upper jaw with canine tusk, lower jaw 
with dps-ms, proximal and distal ends of metatarsus. Type specimen, No. 19147, 
Hsanda Gol formation, Mongolia. All natural size. 


and metatarsals are similarly united into cannon-bones, me ii remaining 
separate and complete; but the distal keels of the cannon-bones are 
extended over the front of the bone and the form and proportions of the 
cannon-bones have assumed more the type of the fully developed 
Cervide. 

The new genus is of interest as fulfilling more nearly than any 
hitherto described the required characters for an early Oligocene direct 
ancestor of the Cervide. 


1924] MONGOLIAN INSECTIVORES AND RUMINANTS 5 


List oF THE HsanpAa Gout FAUNA 


Carnivora 
Hyenodon pervagus, new species 
Didymoconus colgatei, new genus and species of Cee 
_ Didymoconus berkeyi, new species 
Amphicticeps shackelfordi, new genus and species 
Bunelurus ulysses, new species 
Bunelurus parvulus, new species 
Palzoprionodon gracilis, new species 
?Cynodictis elegans, new species 
?Cynodon (Pachycynodon) teilhardi, new species 
?Viverravus constans, new species 
Glires 
Tsaganomys altaicus, new genus and species of Bathyergidse 
Cyclomylus lohensis, new genus of Bathyergidz 
Cricetops dormitor, new genus and species of Cricetopide 
Selenomys mimicus, new genus and species of Cricetopidz 
Tataromys plicidens, new genus and species of ?EKomyide 
Tataromys sigmodon, new species of ?Komyidee 
Karakoromys decessus, new: genus and species of ?Eomyide 
?Prosciurus lohiculus, new species of Paramyidee 
Humys asiaticus, new species of Cricetide 
Desmatolagus gobiensis, new genus and species of Leporidz 
Desmatolagus robustus, new species of Leporidee 
Insectivora 
Tupaiodon morrisi, new species of tupaioid Insectivora 
?Tupaiodon minutus, new species 
Palzoscaptor acridens, new genus and species of Soricoidea 
Palzoscaptor rectus, new species of Soricoidea 
Perissodactyla 
Baluchitherium 
?E piaceratherium 
Artiodactyla 
Eumeryzx culminis, new genus and species, primitive Cervidee 


This fauna is about half made up of new genera; the remainder are 
referred, positively or provisionally, to known genera on the present 
evidence. Hyznodon ranges through the Upper Eocene, Lower and Mid- 
dle Oligocene of Europe; in North America it is limited to Middle and 
Lower Oligocene; in North Africa it occurs in the Lower Oligocene. The 
Oxyeenide range through the Eocene of North America and have one repre- 
sentative, Thereutherium, in the Phosphorites.1. The remaining carnivora 
all belong to that primitive group of Fissipedia best represented in the 
Phosphorite fauna of France and very difficult to place in the accepted fam- 


10xyena is also recorded from the Phosphorites but upon quite insufficient evidence. 


6 AMERICAN MUSEUM NOVITATES [No. 105 


ilies of Fissipedia, as they are essentially transitional between the Miacide 
ofthe Eocene and the Mustelide, Viverride, and Canide of the later 
Tertiary and Recent. Palxoprionodon, Palxogale, Cynodictis, Viver- 
ravus and Cynodon are characteristic of the Phosphorites; Bunelurus 
of the Middle Oligocene of North America is almost identical with 
Palzogale of the Phosphorites, and Viverravus, although occurring in the 
Phosphorites, is typical of the Lower and Middle Eocene of North 
America. Amphicticeps is comparable with the cynodont and steno- 
plesictid groups of the Phosphorites. All of these carnivora are clearly 
in an Oligocene stage of evolution and appear to be rather early Oligocene. 

The rodents are mostly new. One species is comparable with the 
Oligocene species of Cricetodon in Europe and with Humys of the Middle 
Oligocene of North America; another with Prosciurus of the Lower (and 
?Middle) Oligocene of North America. The remainder represent an 
Oligocene stage of evolution in the writers’ view, but their true relations 
to the later Tertiary and existing rodents require further study. 

The two insectivore genera are not closely comparable but would be, 
in our judgment, in an Oligocene stage of evolution. Hwmeryx compares 
with Prodremotherium and Amphitragulus of the French Oligocene, and is 
decidedly more primitive than Blastomeryx of the Lower Miocene, but 
more modernized than any of the American Oligocene White River or 
John Day ruminants. 

The character of the Baluchitherium fauna is peculiar as compared 
with most Tertiary mammal faunas, in the great abundance and variety 
of rodents and small carnivora, and scarcity of ungulates, especially artio- 
dactyla. It represents probably a somewhat different facies from the — 
badland faunas of Western America, or the fissure and quarry faunas of 
Western Europe. It may perhaps be a desert basin fauna, The associa- 
tion of true though primitive Cervide with a fauna rather closely cor- 
related with the older Oligocene of Europe and America is of importance 
as indicating the Asiatic origin of this group, if the principles of evolution 
and dispersal be adopted which were outlined by Matthew in ‘Climate 
and Evolution.’ Hasty conclusions, however, are to be deprecated, as the 
evidence is still scanty and imperfectly studied and there is excellent 
prospect of obtaining more of it in the near future. 

The Irdin Manha beds contain the following fauna: 

Carnivora 
Paracynohyenodon morrisi, new species of Hysenodontide 
Mesonychid, undescribed 


Insectivora 
?Pantolestes of Pantolestide 


1924] MONGOLIAN INSECTIVORES AND RUMINANTS 7 


Perissodactyla 

Desmatotherium mongoliense Osborn 

Protitanotherium mongolicum Osborn 
Artiodactyla 

Undetermined genus, Anthracotheriidie 

The age of the formation appears to be fixed as Upper Eocene by the 
occurrence of Protitanotherium. Desmatotherium, a Bridger Middle 
‘Eocene genus, might suggest somewhat earlier age, but the Mongolian 
species which Osborn has referred to this genus appears to be consider- 
ably more hypsodont and otherwise more advanced than its American 
relative. The absence of ancestral types for the Oligocene invasion of 
Europe and America is unexpected. The Irdin Manha may be correla- 
ted with the Pondaung fauna of Burma, which underlies a marine 
Upper Eocene formation. This would suggest that both may be Mid- 
dle rather than Upper Eocene, and the same would probably hold for 
the Uinta of Utah. 


Sor aaer 
ceomeen 4 

, ee Sell ey 
ocala Gein whe 


aprons 


anes 
. hin oeneene Prec te - 
Ble so IRA ety cape TE tS h totes 
RS ai a i a ed 
Bes Saige! hice Iomlid spent Pe are 
a beiecsiae SPE ee tht igs ca 


a 


rt ea rt ee : Sy ae eh 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 117 Tue American Museum or NaturAt History June 7, 1924 
New York City 4 


56.81,45(51.7) 


A NEW CROCODILIAN FROM MONGOLIA! 


By CwHarues C. Mook 


Among the Mesozoic Reptilia collected by the Third Asiatic Expedi- 
tion in 1923 there is a small crocodilian which represents a new genus and 
species, possibly a new family. It was found by Mr. Walter Granger at 
Shabarakh Usu, on the Kwei Hua Chung—Uliassutai trail. The hori- 
zon is the Djadochta Beds, of lower Cretaceous, or possibly Comanchean, 
age. The specimen consists of skull and jaws, moderately well preserved. 


Shamosuchus djadochtaensis, new genus and species 


Typr.—Skull and jaws, incomplete. Amer. Mus. No. 6412. 

GENERIC CHARACTERS.—Absence of mandibular foramen, prominent postero- 
external process of squamosal, exoccipital comprising a considerable portion of the 
condyle. 

Speciric CHARACTERS.—Median ridge on frontal bone, prominent ridges on 
lachrymals, medium size of supratemporal fenestrae, and their position close to the 
median line and far from the posterior and external borders of the cranial table. 

The skull is comparatively broad for its length. The tip of the snout 
is not preserved, but the convergence of the lateral borders of the skull 
and of the two rami of the mandibles indicates clearly that the skull was 
short. The cranial table is relatively large, and its posterior border is 
strongly curved. It is flat in the portions preserved. The facial region 
of the skull is broad and flat. The supratemporal fenestra are of mod- 
erate size. They are situated on the anterior portion of the cranial 
table, relatively near the mid-line of the skull. They are comparatively 
far from the external and posterior borders of the table. Theirlength is 
somewhat greater than their breadth. The orbits are large. The aper- 
tures of the anterior portion of the snout are not preserved. The internal 
narial aperture on the palate has been pushed out of position with respect 
to its surrounding bones by crushing, but it was apparently similar in 
form and position to that of the eusuchian crocodiles of to-day and did 
not resemble the aperture of the Mesosuchia. 

The premaxillaries are not preserved. The maxillaries occupy more 
than two-thirds of the total breadth of the snout. The anterior teeth are 


'Publications of the Asiatic Expeditions of The American Museum of Natural History. Contri- 


bution No. 23. F 
Contributions to the Osteology, Affinities and Distribution of the Crocodilia. No. 13. 


AMERICAN MUSEUM NOVITATES [No. 117 


ms ne | 
Wy SSO 9 M642 


(Ga 


Fig. 1. Shamosuchus djadochtaensis, new species. 
Type skull, Amer. Mus. No. 6412. Natural size. A, 
superior view; B, posterior view. 


lacking. One tooth of the middle region is of average proportions and is 
moderately sharp. Four small teeth near the posterior end of the left 
side are short-crowned and sharp. They are considerably longer in their 


1924] A NEW CROCODILIAN FROM MONGOLIA 3 


antero-posterior than in their transverse diameters. Several teeth 
slightly farther forward on the right side are similar, except that they are 
slightly sharper. No button-like, crushing teeth are present. The 
maxillo-nasal sutures converge sharply forward. The borders of the 
maxillaries, near these sutures, are slightly elevated. 

, The nasals are very narrow at their anterior ends, where they are 
broken off. They broaden rapidly in the posterior direction, reaching 
their maximum breadth at their junctions with both lachrymals and 
prefrontals. From these points backward they narrow rapidly. The 
posterior end of each nasal is somewhat abruptly pointed. The two 
posterior points are separated by the blunt anterior wedge of the frontal. 
The superior surfaces of the nasals are slightly concave from side to side. 
Along the median line there is a deep excavation. 

The lachrymal is a distinctive bone. It is completely preserved on 
the left side only. Its suture with the maxillary is very irregular. The 
inferior process, extending backward below the anterior border of the 
orbit, is relatively long and slender. The anterior process is both long 
and broad. On it is situated an elevation which is prominent near the 
center of the maxillo-lachrymal suture. This elevation extends inward 
and slightly forward. Near the center of the lachrymo-prefrontal suture 
it turns sharply forward and extends parallel with the nasal border of the 
lachrymal, joining the ridge on the maxillary mentioned above. The 
anterior branch of the elevation is low and narrow and not prominent. 
The lateral branch is very prominent. It slopes forward and outward 
with a gentle concavity, and backward and inward with an abrupt de- 
pression. There is a small flat area between the elevation and the 
suture with the prefrontal. The lachrymal occupies a considerably 
greater portion of the orbital border than does the prefrontal. 

The prefrontal bone is roughly triangular, the long base of the tri- 
angle being the inner border of the bone, composed of the naso-prefrontal 
and prefronto-frontal sutures. This border is somewhat convex toward 
the median line. The other two sides of the triangle are the orbital border 
of the bone and the lachrymo-prefrontal suture, both of about the 
same length and both concave outward. The bone is crossed by a low 
‘ridge which extends outward and forward from the center of the pre- 
fronto-frontal suture to the angle of the ridge on the lachrymal. 

The anterior portion of the frontal is preserved. It is finely pitted 
and is elevated along the median line into a keel. Consequently the 
portion preserved resembles one of the nuchal scutes of the exo- 
skeleton. The breadth of the interorbital portion is considerable. The 


4 AMERICAN MUSEUM NOVITATES [No. 117 


anterior process is broad and blunt, wedging apart the posterior termina- 
tions of the two nasals. Near the borders of the anterior wedge there are 
two deep pits opposite the ends of the prefrontal ridges. 

A portion of the left postorbital is preserved. Its orbital border is 
incomplete. It occupies only about one-fourth of the external border 
of the cranial table,—an unusually small proportion. ; 


A a aS eS Zl" ) 


Z 


y'\\ 


it 


Fig. 2. Shamosuchus djadochtaensis, new species. Type skull, Amer. Mus. 
No. 6412. Natural size. A, lateral view, left side; B, inferior view. 


The squamosal of the left side is preserved. It is relatively large, 
occupying about three-fourths of the external border of the cranial 
table. This unusual proportion is due to the backward extension of the 
postero-external corner of the bone backward in the form of a prominent 


1924] A NEW CROCODILIAN FROM MONGOLIA 5 


posterior process. This process is thickened at its end, and there is a 
shallow concavity anterior to the thickening, extending as a groove over 
the side of the cranial table and separated from the anterior portion of 
the bone by a shallow oblique ridge. 

The parietal is not preserved, but the squamosal ends in the speci- 
men at what is apparently the squamoso-parietal suture. This extends 
forward and slightly inward. The parietal must have occupied not more 
than 40 per cent. of the posterior border of the cranial table, and per- 
haps much less. 

A portion of the supraoccipital is preserved. This bone evidently 
occupied a small portion of the posterior of the cranial table, also a very 
small portion of the surface of the table. The postero-inferior plate of 
the bone extends downward more than one-half the distance from the 
cranial table to the foramen magnum. 

The exoccipitals almost entirely surround the foramen magnum, only 
a small portion of the condyle being composed of basioccipital. The 
region of the pterygoids is so badly crushed that its characters are not 
positively distinguishable. 

Portions of both mandibular rami are preserved. They converge 
sharply toward each other in the anterior direction. Their most notable 
characteristics are a considerable lateral thickness, especially of the 
splenial bones, and the absence of a mandibular foramen. 


aw 


Loh 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 119 Tue American Mussum or NaTuray History June 20, 1924 
New York City 


55.2,41(51.7) 


THE GREAT BATHYLITH OF CENTRAL MONGOLIA! 
By Cuarues P. BERKEY AND FREDERICK K. Morris 


INTRODUCTION 

This paper is in the nature of an announcement rather than a dis- 
cussion. It does not attempt a review of the literature on this subject. 
Pumpelly, von Richthofen, Loczy, Obruchey and others have noted 
granite in the Gobi or in adjacent regions. At some other time, an 
attempt will be made to summarize the geological observations of those 
who preceded us; and then, we are confident, it will be found that 
many additional occurrences, some of them far beyond the limits of our 
own territory, should be regarded as belonging to the same great geo- 
logic unit that is emphasized here. 


The Third Asiatic Expedition, in the early stages of its traverse 
across Mongolia, repeatedly noted occurrences of granite. Some are 
small intrusives, whereas others cover extensive areas of undetermined 
boundaries. They are associated with so great a variety of other rock 
formations of widely different ages, that at first there was little to suggest 
their possible unity. Because of the fact also that the granites them- 
selves show considerable variety of minor character, it was assumed that 
they were essentially independent intrusion phenomena that might have 
as great age differences as the hosts with which they are associated. 

As the traverse was extended northward from Kalgan, however, a 
certain similarity of rock type and of field relation was noted, suggestive 
of a possible common origin. This became all the more impressive in the 
north and west, where, in many places, the rock floor for tens of miles 
together is made up wholly of granite. 

It is the purpose of this paper to indicate some of the evidence 
bearing on the structural and genetic relations of these granites, and to 
support, in more definite terms than has hitherto been done in the 
publications of the Expedition, the conclusion that they represent a 
great granite bathylith. 


eee eetions of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 2 


2 AMERICAN MUSEUM NOVITATES [No. 119 


LY EY) 


YA), 
U4), 


7 Yi" se 
] : 


ee Iren Dabasile = 
kArdyn Obo 


; 
S/N 


Mountain Areas 


OS 


ao Granite Areas Pe 


S SS 
= S 


anil 


200 Miles Peking 


100 100 


Fig. 1. Location map of the central portion of Mongolia. 


This map covers a total area of about 475,600 square miles, showing the principal visited localities 
where granites of the type believed to belong to the Great Bathylith are extensively exposed. There 
are doubtless many other exposures lying beyond the reach of the traverse of the Expedition. These 
already mapped, however, serve the present purpose. A lateral extent of more than 720 miles is indi- 
cated, and the total area is probably much greater than the map itself. 


SPECIAL LOCALITIES 

Many localities exhibit granite outcrops, but particularly promi- 
nent and suggestive are those of the granite hills along the Urga trail 
from 100 to 150 miles out from Kalgan, the vicinity of Ude, 230 miles 
out, and another stretch of ten miles beginning at 430 miles out. At 
Mount Tuerin, 500 miles out from Kalgan, granites of several varieties 
form the floor for 35 miles, the mountain itself being only an erosion 
remnant wholly composed of granite of a single type. Along the south 
side of the Tola River on the way from Urga southwestward to Tsetsen- 
wan, intrusive granites in the form of great dikes and bosses become more 
and more abundant and stand out as prominent elements in the topog- 
raphy. Here the contrast between the erratic igneous bodies and the 
simple uniform features developed by erosion of the regularly folded 
graywacke series is very striking indeed. The interrelations of the intru- 
sive bodies and the graywackes, as well as the effects produced, are re- 
markably well displayed. . 


1924] GREAT BATHYLITH OF CENTRAL MONGOLIA 3 


The Tsetsenwan district itself exhibits some of the most illuminating 
relations, for at this place, about 150 miles west of Urga, erosion has cut 
down through the graywacke roof, not only exposing the massive granite 
beneath, but also furnishing illustration of the phenomena that might be 
expected near a great igneous contact. There for a hundred miles, 100 
to 200 miles southwest of Urga, the granite either forms the surface or 
lies so close below that it has broken through in numerous places and has 
affected the overlying rock in the characteristic manner of contact meta- 
morphism. The same granites outcrop again at the Ongin Gol, thirty 
miles farther west, and at Arishan, the holy mountain of Sain Noin, thirty 


+— Course S70W = = 
890 9 8 7 i 6 5 4 3 2 «cf “'660""cs 8 7 6 5 4 3 2 1 870 
he : 2 e 
Graywacke Stries Jurassic 8 8 Ancient crystalline schists and gneisses 
Siege Conglomerates Ss = 


invaded by granites 


Underlying Mongolian Granite Bathylith 
Fig. 2. Geologic cross section of 20 miles between Tsetsenwan and the 
Ongin Gol. 


_ The underlying granites are repeatedly encountered,and where critical structural relations between 
the granite bathylith and the ancient gneisses and schists can be seen. Here ancient metamorphic series, 
separated by conglomerates, are invaded by the granite in a complex way, following the structural weak- 
nesses of the older rock and giving rise to more intimate mixture than is noted where graywacke forms 
the roof. The contact line is, therefore, more sharply defined with the graywacke than with the older 
metamorphics. The section includes a small synclinal remnant of the much younger Jurassic 
conglomerates. 


miles farther in the same direction. At Gorida on the Uliassutai trail, 
midway between Sair Usu and Uliassutai, and at Baga Bogdo, of the 
Altai system, even the mountains are of granite which has been lifted into 
prominence by faulting in later time. These are a thousand miles distant 
from the first occurrences. At many other places between Baga Bogdo 
and Kalgan there are equally good examples, such as that on the Uliassutai 
trail, west of Sair Usu, and the mountain area Golobai’n Ola, as well as 
the large area of granite along the same trail, 400 miles southeast of Sair 
Usu, within a hundred miles of Kalgan. 

Some of these are places where granites have been developed in 
great prominence, and where their relations to the older and the younger 
rocks are clearly shown. The structural relations of other long stretches, 
as well as the connection of one outcrop with another, must of course be 
inferred. Such of these relations as can be determined make it abundantly 
evident that most of the granites of Mongolia must be intimately related 
to one another; and the most satisfactory explanation for them is that 


\ 
4 AMERICAN MUSEUM NOVITATES [No. 119 


they represent one great underlying granitic mass of enormous dimen- 
sions, exposed in patches over the entire area touched by the travels of the 
Expedition, wherever erosion has stripped off the formations that con- 
stituted its roof. 

Evidence has accumulated also to show that such mineralogic and 
textural variety as the granites themselves exhibit probably originated 
in the processes of differentiation within the magma, and in the process of 
absorption or syntexis belonging to the normal life history of this ancient 
magma. 


CHARACTERISTIC FEATURES 

Facigs oF THE Rocx.—Typically, the granites of Mongolia are 
light pink or somewhat reddish in color, and are comparatively coarse- 
grained. In both color and texture, however, they vary greatly. Ortho- 
clase feldspars of light tints are dominant. Quartz is abundant, but even 
this mineral varies considerably. The dark constituents vary greatly in 
amount and range from dominant biotite to dominant hornblende, 
biotite being much the more common. Thus there is a good deal of 
mineral variation in different localities or individual instances, but the 
surprising thing after all is the similarity that one finds in them, even in 
widely separated localities. 

Although the common structure is massive and comparatively 
coarse-grained, there are varieties of strongly porphyritic habit, occa- 
sionally an obscure gneissoid arrangement of constituents, and varieties 
also that range from fine to extremely coarse grain. The larger areas 
exhibit the most massive habit and the most uniform quality. In many 
places, these granites are slightly miarolitic and show some pegmatitic 
tendency, but this is not a striking feature. In the Mount Tuerin area, 
a hornblende granite, judged to be a facies of the same rock, develops black 
stains which we believe indicate a small manganese content. 

All of these varieties we judge to be understandable as facies of the 
same magmatic type. 

Forms.—By far the larger number of individual examples appear in 
the form of dikes, irregular intrusions or lit-par-lit injections. Besides 
these, masses of very great extent are found, which are best understood 
as bosses or cupola-like upward extensions of the bathylithic mass, which 
have been truncated by erosion. Some of the larger areas are twenty 
miles or more across, with only minor remnants of the former roof to break 
the monotony. More rarely, one finds very much more intimate rela- 
tion between the granite and the older country rocks. The granitic 


1924] GREAT BATHYLITH OF CENTRAL MONGOLIA 5 


material forms fine injection-bands, and sometimes it is even more in- 
timately mixed, as though it had literally soaked the formation until the 
whole has become a veritable complex of original rock and introduced 
granite. All gradations are found, from extensive masses many miles 
across, that must be part of the major bathylith itself, to dikes of all 
sizes, and to the most delicate, penetrating stringers and impregnations 
that seem to die out as they penetrate and mingle with the enclosing host 
rock. 

In other places, especially where the granite cuts the younger forma- 
tions, there is a sharper contact, though still a very complex one. For 
instance, at the holy mountain Arishan, the edges of the graywacke are 
abruptly truncated by the granite. Angular masses of granite rise into 
the graywacke and send upward long dikes, some of which, branching, 
completely enclose blocks of graywacke. Pendants of the graywacke 
extend down into the granite, and xenoliths lie isolated near the con- 
tact. These relations suggest that the invading magma reached its 
present position partly, at least, by stoping. 

Thus the mode of emplacement of this great body has probably 
varied from place to place, involving at least three methods of attack 
upon the roof: 

1. By injection, soaking and even solution and replacement, 
involving metamorphism of the strata so invaded. 

2. By magmatic stoping of xenolithic blocks whose place the 
magma takes. 

3. By mechanical pushing and displacement of overlying rocks. 

The three methods are not mutually exclusive, except that simple 
stoping does not include much of the more complex and intimate mode 
‘of attack named as method 1. The third method, that of mechanical 
displacement, involves the question of the extent to which a large upris- 
ing mass of magma may deform rocks and actuate orogenic movements. 
It is probable that the bathylith caused more or less movement, and, on a 
relatively small scale, we can see that this has been the case. It is 
difficult, however, to evaluate the part in deformation actually played 
as a whole by the bathylith. | 

Although the common representative is a biotite granite, the fact 
that there is no distinguishable difference of meaning or structural rela- 
tion in the hornblende granites or in any of the other varieties makes it 
look reasonable that they are all simply facies of the same magma, due to 
a moderate amount of differentiation. There are, however, other masses 
of exactly the same relations whose compositions depart much more 


6 | AMERICAN MUSEUM NOVITATES [No. 119 


from the average. Included in these are occasional syenites and diorites, 
and, rarely, still more basic types. If one includes the ‘‘serpent-form”’ 
dikes to be described in a later paragraph, and other small masses asso- 
ciated with the granite as host, which are most reasonably explained as 
closing-stage representatives of internal solidification, the range is still 
greater. Thus even dolerites and gabbros, found at a few places, may be 
included among the differentiates of the bathylithic magma. 

STRATIGRAPHIC RELATIONS.—These granites cut a great variety of 
rock formations. No member of the ancient: crystalline, metamorphic 
floor has wholly escaped either actual intrusion or some of the contact 
effects of this invading magmatic mass. It is clearly younger than any 
of the pre-Cambrian formations, because it cuts even the great graywacke 
series of the Tola River region, which has been referred to in previous 
communications as the Khangai series of graywackes and slates and is 
regarded as probably the equivalent of the Nan K’ou series of China, the 
latest of the pre-Cambrian formations.!. Wherever the relations are 
clear, it is evident that the folded Khangai graywacke series and the 
whole confused complex of still earlier series together constitute the roof 
beneath which this magma solidified. 

Subsequent erosion has exposed remnants of this roof that must 
have extended downward into the magma itself like great pendant pro- 
jections. Thus, one frequently crosses such surrounded bodies of older 
rock from granite on one side to granite again on the other, and, because 
of the insignificant soil cover, one can observe in as much detail as could 
be desired every transitional and transforming step with all the field rela- 
tions and effects characteristic of such a history. Some of these roof 
pendants are most impressive field exhibits, those seen in the Tsetsenwan 
district being on a particularly large scale. At occasional points, the’ 
intimate relations of granite and older rock can be seen in still greater 
detail, attended by typical contact phenomena. 

It is clear, therefore, that the granites cut through formations up 
to and including the Khangai graywackes. But, because of the scarcity of 
rocks of determined Paleozoic age, it is much less certain what the rela- 
tions of the bathylith to the Permo-Carboniferous series of sediments 
are. No direct associations of bathylithic granite and these upper 
Paleozoic strata have been observed thus far. 

Erosion has uncovered also the relation between the granites and 
the still younger strata in very many places. At Tsetsenwan the lower 

lyon Richthofen, Ferdinand. 1887. ‘‘China,’’ II, p. 


306. 
Willis, Bailey, Blackwelder, Eliot, and Sargent, R.H. 1907. ‘Research in China,” Part I, p. 1238, 
Carnegie Institution of Washington, Publication 54, 


1924]) GREAT BATHYLITH OF CENTRAL MONGOLIA t 


Jurassic conglomerates have this relation, and it is clear that the granites 
had been exposed by deep erosion before that period. (Fig. 3.) In many 
places, also, Cretaceous and Tertiary strata lie on a granite floor. This 
fact indicates that, at the time of general post-Jurassic peneplanation 
and before the development of the later basin sediments, the granites of 
Mongolia were exposed over immensely larger areas than now. 

Contact AND MiNERALIZATION Errects.—At most places, contact 
phenomena are not very marked; but typical effects are observed at 
many places, the sum of which gives the usual list of transformations due 
to the influence of such a unit. In the Tsetsenwan region, for example, 
W aE 


Contact metamorphic effects Erosion Unconformity 


The Graywacke Series 


ne ate Spr SEs: — 
NW STS ent 
Orem 


Underlying granite bathylith cut by great numbers of dikes 


Jurassic Conglomerates 


Gr aywacke roof pendant 


Fig. 3. Diagrammatic cross section of a critical structural relation at 
Tsetsenwan. 

This section lies about 200 miles west of Urga, where erosion had exposed the granite bathylith and 
carried away most of its roof of graywacke before the Jurassic sediments were laid down. A synclinal 
remnant of Jurassic conglomerate is preserved here, although several thousand feet of these strata are 
to be found only a few miles away. Two great denudation epochs, therefore, are represented by these 
profile lines, and the section illustrates well one of the important unconformities of the region. This 
particular area exhibits a very fine development of the serpent-form dikes, which cut both the granite 
and the roof in great numbers, but which are most prominent in the granite areas. 
the graywacke-slate series is very heavily tourmalinized, and much 
epidote is developed. Sometimes there is excessive induration and sil- 
icification, as well as silication of the adjacent or overlying rock. 
Doubtless, also, some of the variations in crystallinity of the overlying 
metamorphic rocks are due to the greater or smaller influence of the 
underlying bathylith. For example, the graywacke series, which nor- 

. . . . . . . 
mally exhibits abundant evidence of its clastic character and is not 
markedly crystalline, becomes in some places quite strongly schistose with 
well developed metaphenocrysts. In certain places, it was possible to 
demonstrate that this metamorphism was due to the presence of the 
granite. In others, where no granite was visible, its presence close below 
has been inferred. The only satisfactory working hypothesis is that all 
excessive metamorphic effects were due to the influence of the granite 
bathylith. , 

Thus, in crossing certain country, there are surprising changes in 
the quality or condition of formations that are elsewhere uniform and 


simple. Such changes become intelligible in the light of such relations as 


8 AMERICAN MUSEUM NOVITATES [No. 119 


are here implied; but they seem anomalous, until one has discovered 
that, immediately beneath, lies the granite bathylith, and that the 
effects produced are such as are reasonably to be expected from its in- 
fluence, even though the granite itself does not reach the present surface. 

These metamorphic areas of obscure relation have been checked up 
repeatedly and compared with the effects produced where the relations. 
are perfectly plain, and there is, in our opinion, no doubt whatever of the 
major facts. We consider the underlying granite, therefore, to be the 
dominant cause of all the more pronounced contact metamorphism. 

The striking thing, of course, is that metallic mineralization is. 
absent. In the whole of central Mongolia, from the Kalgan border to the 
Tola River on the north and from the meridian of Baga Bogdo on the 
west to the Kalgan-Urga trail, surprisingly few traces of metallic minerals 
have been seen. Even at points where contact phenomena are developed 
on a magnificent scale, as at Tsetsenwan and at Sain Noin, no important 
metallic mineralization was noted. Quartz veins, on the other hand, are 
common, and in a few places such products are developed in great abun- 
dance; but none of the veins thus far inspected has noticeable metallic 
content. In spite of the observed poverty in metallic mineralization, it 
may well be that elsewhere, particularly on the margins of the bathylith, 
beyond the reach of our traverse, more favorable conditions for its: 
occurrance exist. 

THE SERPENT-FORM D1krEs.—One of the most striking features of 
certain areas is the great number of dikes that cut the granite. They 
occur literally in thousands and take prevailingly most erratic serpent- 
like courses. At places such as Tsetsenwan, where, from any good van- 
tage point, it is possible to look out and down upon a large stretch of such 
country, the whole landscape looks as if it were a tangle of serpent-like 
forms. In composition, they almost exhaust the range of porphyries. 
They vary from quartz porphyry and trachyte porphyry at the acid ex- 
treme to comparatively basic types,—at least to andesite porphyry and 
even, though more rarely, to a basaltic type. In size are they equally 
variable, some of the larger intrusions reaching such dimensions as to 
form prominent ridges; and they cut each other in a most confusing 
way. The tangle of tortuous forms is, however, the most impressive 
feature. Similar structural relations were found at the holy mountain 
of Sain Noin, but the country there does not lend itself so well to surface 
display. These dikes stand up under weathering better than does the 
granite host, and this accentuates the peculiar physiographic effect 
referred to in the term serpent-form dikes. 


1924] GREAT BATHYLITH OF CENTRAL MONGOLIA 9 


What the genetic relation of these dikes is to the granite bathylith is 
a matter much more difficult to determine, but they have not been noted 
in such profusion in any place, except in the districts where, judging 
by the abundance of roof pendants and xenoliths, the present surface is 
very near the roof contact. They are believed to have originated as an 
end-product. of the deeper interior cooling of the magma mass itself, 
and thus they actually represent facies of the magmatie differentiation 
of the bathylithic magma. 

OTHER INTRUSIVES.—There are older granites, such as those that 
form an integral part of the ancient gneisses, which we have tentatively 
called Archzeozoic and have correlated with the T’ai Shan complex in 
China.! Possibly some granites, younger than the T’ai Shan, are yet 
older than the great bathylith. Such granites are found injecting schists 
which we believe to be correlated with the Wu T’ai Shan series as de- 
scribed by Willis and Blackwelder for China.? 

A few andesite dikes were found cutting the marine limestones of 
Permian age. Intrusions also of granite porphyry, syenite porphyry and 
porphyrites of various kinds have invaded the Jurassic conglomerates, 
sandstones and shales, and there are abundant surface flows with ash 
and tuff beds associated with these strata in some localities. Some of 
these porphyry masses are large, and, in many places, the different 
porphyry units cut each other in a most confusing way, giving a veritable 
complex of intrusives that completely dominates the geologic structure. 
The flows include many rhyolitic, fewer trachytic and some andesitic lavas. 

After the late Jurassic or post-Jurassic revolution, basaltic lavas and 
dikes are found in the Lower Cretaceous (Comanchean) formation of 
Oshih (Ashile), again in the early Eocene basin of Gashato, and in the 
Oligocene of Hsanda Gol. Similar lavas and volcanic plugs are associated 
with the sedimentary strata of Wan Chuan Pass above Kalgan, on the 
eroded edge of the plateau. Most of these flows are basalts, although 
trachytes and rhyolites also are met with. 

How many of these other intrusives, some much older and some 
much younger, have had a genetic dependence on the underlying granite 
may be impossible to determine; but it is reasonable to believe that both 
some of the antecedent and some of the subsequent igneous outbreaks 
were connected with active stages of the life history of the Great Mon- 
golian Bathylith. 


1Willis, Bailey, Blackwelder, Eliot, and Sargent, R. H. 1907. ‘Research in China,’ Part I, 
p. 19, Carnegie Institution of Washington, Publication 54. 
20p. cit., p. 109. 


(=| 
EE 
= 
> 
=} 
co 
= 
Ww 
= 
ke 
iste 
Li 
= 
= 
== 
es 
= 
a 


RECOGNIZED IN PART 
BY EARLIER EXPLORERS 


ARCHEOZOIC 
ARCHEAN 


POWER PORTION Gh tae 
GENERALIZED GEOLOGIC COLUMN 


FOR CENTRAL MONGOLIA 
COVERING THE STRUCTURAL UNITS OF THE BASIN FLOOR 


BUR et Aya lt) BUN. BO. DEF AG RP isle 


ALL ROCKS BELOW THIS LINE ARE FOLDED 


A GREAT SERIES OF CONSLOMERATES, SAND- 
STONES,AND SHALES,WITH ASSOCIATED 
LAVA FLOWS, TUFFS AND ASHES;CARRYING 
OBSCURE PLANT REMAINS AND LOCALLY, 
COAL; THE WHOLE ABOUT 20,000 FEET THICK 
APPARENTLY CORRESPONDS TO LOWER JU- 
RASSIC OF NORTHERN CHINA 


TSETSENWAN 
SERIES 
JURASSIC 


MESOZOIC 
EARLY MESOZOIC 


LN ON FE vO el ae a 


LIMESTONES 


SHALES 
A SERIES OF 


Perea 


SANDSTONES 


SLATES 


SERIES 


FOSSILS 


SAIR USU 


QUARTZITES 


CONGLOMERATES 


PALEOZOIC 
LATE PALEOZOIC 


CARBON- 
IFEROUS 


UNCONFORMITY COVERING EARLY PALEOZOIC TIME 


MONGOLIAN 


GREAT BATHYLITHIC INVASION GRANITE 


THE THE TOLA RIVER 
NANK OU SYSTEM KHANGA| | GRAYWACKES 


FIRST OY oe Ci SERIES AND SLATES 


WITHOUT FOSSILS 


' SCHISTS 
PHYLLITES 


WU-TAl SYSTEM 


AS USED BY WILLIS IN CHINA QUARTZITES 
GREENSTONES 


PROUT ERUAOE 
PEARLY |[_LATE 


, CRYSTALLINE © 
THE T’AI-SHAN COMPLEX LIMESTONES, 
SCHISTS, 
AND COMPLEX 
AS USED BY WILLIS IN CHINA INJECTION 
GNEISSES 


1924] GREAT BATHYLITH OF CENTRAL MONGOLIA 11 


GENERAL INFERENCES 

From the distribution and character of these granites and from their 
structural relations to the recognizable forms of the region, it is a fair 
inference that a great granite bathylith underlies all the country trav- 
ersed by the Third Asiatic Expedition and even extends considerably. 
beyond it in every direction. To the bathylith belong all of the large 
areas of granite in the Gobi region, including that which comprises Baga 
Bogdo of the Altai mountain system, which has been pushed up thousands 
of feet by later faulting. It is a fair inference, also, that most of the small 
occurrences, thousands of dikes and other forms of this general type 
preceding Jurassic time, have come from the same source and only repre- 
sent the extreme outward penetration of this mass as it invaded the over- 
lying formations. ' 

From the facts given, it is clear that in age the bathylith is younger 
than any of the pre-Cambrian series, and is clearly younger than the 
Khangai graywackes, of whatever age the latter maybe. Itisclear, also, 
that the bathylith is so much older than the great conglomerate series 
judged to be of lower Jurassic age, that there was time entirely to remove 
the roof by erosion over large stretches of country before these conglom- 
erates were laid down. Somewhere between the latest pre-Cambrian, 
therefore, on the one hand, and the early Mesozoic represented by the 
Jurassic conglomerates on the other, the granites invaded the overlying 
terrane of Mongolia. But this is a big gap,—the whole of the Paleozoic 
is left out of the calculation. 

It is certain, however, that the graywacke-slate series was folded 
before the maximum invasion of granite, so that a mountain-making 
epoch intervened between the making of the graywacke formation and 
the full development of the bathylith. If, therefore, the graywackes are 
really latest pre-Cambrian, as now believed, time must be allowed in the 
Paleozoic era for this igneous invasion. 

It is difficult to determine the relation to authentic Paleozoic strata, 
because of the very slight development of rocks of this age. Some ques- 
tion also arises as to the uncertainties of the age of the graywacke-slate 
series. The best that can be said at present is that the Paleozoic strata 
were not affected by the granites in those places coming under observa- 
tion, and the similarity of habit in this respect to that of the Jurassic 
conglomerates, which are clearly very much later, leads to the tentative 
conclusion that the granite bathylith is itself of early Paleozoic age. 

For this bathylith, which in dimensions seems to compare favorably 
with the greatest bathyliths thus far known in other parts of the world, 
we propose the name ‘“‘the Great Mongolian Bathylith.” 


ceAS 


Ee eee ae ; a 
eT ‘Shy “0 aes yt oe abe aaa Paco aks Ps vat he pout! 
os Sn Gk eM iy 0. hy, abel ead el en ae © 

* PE WEE IEe, Pe, ORK FF Pritts Rite eats kM 
: PRE MS TAD PRR PTI SL 12, frais ne an eWvibi bash: 
a) s Av: wa id 3 yf e ~~ a ‘) hy > pus 
Pie te ATs ee te rere pation Z iii eee Emin 
Hea ame Furr es satu da teenie tus Bum Tp RABEL eee he 
4 4 f ie , : 
it ky, Bay bg ila ee thn « Cary) 
FiydD eRe Whe ay APN PM Mh OE Aa Ne 
s i. f 
j 
ad oA t ; ‘i bie 
ats b aa! Bt + Yet? && eit Oks | ‘ae ee § a ¥ ¥ 
¢ 
7 e re. . » « ry 
Ses Nhs ie 5 j oR righ pb et tee > 
aan , 7 Ri ‘ 
yar Pee pee Att tai ® Say Ke PPM ARDS soe oil 
wy vithan hes * F ; eRe Bit 
i yy be i Hn te VP ei hy x ~ ‘ fi I toi 
‘# » ¥ a y 
: RE Veter i aboe 6 TOLER Ret Pe eas hog eM 
r ¢ ’ a> } 
Me sh cache ty] tie ee be hait 
a he af Pye sal + d M % ' wi 
. ) i ‘ 
BY ; f ' f 
AS * wie Hayy ny viel By taf < “ . 4 afi 
ere TS TR  Soae) ¢ uk (Si RSE es 
Gre’ og a i ‘ «f 
rs +-y. r alt 2. o* > 
ears e 4 ¢ onl v2 ‘ ¥ ‘ 7 
te ; ay re ay : bs x iY : 
2 ; he ‘ 
Meet tle RA a ie Ss ; 
. 4 , 
es eae 
erlerae? , I. ©: thes teh & Pat aan ¥ fe ef “a 
- vv . Mad an mia 
ray Cathe ey, PENS : 
APT ac rar knee sbPasees 1s eeeay 


ie 


> ay Prey 4 


Sees 
Rea ete REE mapas) Sateen ue Siva, 7 ence date sden ays a 
ia ott Saotiiar tt, 3 
easy ae Guten rhs es ddd ae a HHO rae aay: 


a i SAW ie Dee st 
plane F ates ae Ce Yi raed eget aa Re 


ts 


ei 
n " 


ft gaged Toth a AB. ti ae 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 127 Tue American Museum or NaTurat History Sept. 4, 1924 
t New Yor« City 


56.81,9P(117:51.7) 
PSITTACOSAURUS AND PROTIGUANODON: TWO LOWER 
CRETACEOUS IGUANODONTS FROM MONGOLIA! 


By Henry FAIRFIELD OSBORN , 


The preliminary description of these iguanodonts? was, prior to the 
complete exposure and restoration of the two type skeletons, an extremely 
long, difficult and delicate process, followed by detailed drawings and 
restorations which give us an exceptionally complete knowledge of these 
animals. The types are: 

Osuin (ASHILE) ForMATION.—Psittacosaurus mongoliensis (Amer. Mus. 6254), an 
almost perfect skull and jaws with greater part of skeleton. 

Onpar Sarr Formation.—Protiguanodon mongoliense (Amer. Mus. 6253), an im- 
perfect skull and left jaw with a practically perfect skeleton. 

These two types resemble each other in so many characters that they 
obviously belong to a distinct family of iguanodonts to which the 
name Psittacosauride has been applied. These short-skulled iguano- 
donts derive their family name Psittacosauride from the very deep 
parrot-like beak, with small nostrils located at the top of the very deep 
maxilla. There is still some question as to the validity of the subfamily 
name Protiguanodontine proposed at the same time. 

The characters which Pszttacosaurus and Protiguanodon exhibit in 
common are: (1) Cranium relatively short and broad, premaxillaries and 
anterior portion of dentaries edentulous; maxillary teeth of iguanid 
type. Functional teeth in a single row. Nine dentary teeth in Proti- 
guanodon; 7+ maxillary teeth in Psittacosaurus. (2) Neck short; 
cervicals, 6 with 5 free ribs. (3) Thoracics: 16 in Psittacosaurus and 15 
in Protiguanodon. (4) Sacrals: 5 in Psittacosaurus, 6 in Protiguanodon. 
(5) Caudals: 43 estimated in Psittacosaurus, 43 actual in Protiguanodon. 
(6) Cervicals, thoracics and sacrals: 27 in both Psittacosaurus and 
Protiguanodon. (7) Shoulder girdle with free clavicle, coracoid and 
coracoid foramen, and distally expanded, elongate scapula. This is the first 
record of the occurrence of a clavicle in ornithischian dinosaurs. (8) 


; Sate lea of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. ‘ 

“Osborn, H. F. 1923. “Two Lower Cretaceous Dinosaurs of Mongolia.’’ Amer. Mus. Novitates, 
No. oT October 19, pp. 1-10. 

®Ide 26. 


m, 


2 AMERICAN MUSEUM NOVITATES [No. 127 


Fore limb, total length from humerus to extremity of phalanges, 250 mm. 
Psittacosaurus; 245 mm. Protiguanodon. (9) Pelvic girdle of character- 
istic iguanodont type, elongate ilium, pre- and postpubic extension, 
elongate ischium. (10) Hind limb, total length, 470mm. Pszttacosaurus; 
435 mm. Protiguanodon. (11) Manus in both genera with 4 free carpalia, 
Digit I, 2 ph., D.II, 3 ph., D-III, 4 ph., D.IV, reduced, 1 vestigial ph. 
(12) Pes in both genera enlarged, elongate, 4+ free tarsalia in Protiguan- 


PSITTACOSAURUS MONGOLIENS/IS TYPE SO, i 


AMER, MUS. 6254 ae 
WS 


Fig. 1. Type skeleton of Psittacosaurus mongoliensis Osborn (Amer. Mus. 
6254). 


Since the original description the skeleton has been completely exposed and is very accu- 
rately represented in this figure as it lies in the matrix. The original pencil drawing is full size; 
os eee figure is reproduced one-eighth natural size. The missing parts are represented in 

otted lines. 


odon, tibio-tarsus closely united but not coalesced. Digit I, 2 ph., D.II, 
3 ph., D.III, 4 ph., D.IV, 5 ph., D.V, vestigial mts.,0 ph. (13) Osseous 
tendons extending from fourth or fifth thoracic through the sacral 
series to the anterior caudal. 

The above practically common characters, which indicate a marked 
affinity between these two animals, are accompanied by certain differ- 
ences in proportion of the fore and hind limbs as shown in the following 
table of measurements: 


€ 


*BZurjop Aq pozeorpul aie sjred pasojsed osayT, “[BANZooTuod atv Y}90} Areyuep Zurpuodses109 ay} : (199 “SNIT 
“IOULY) [[NYAS paliajor ayy ULOAJ PI10zsot JIB Y99} ALB[[IXVUL JYDIGT "“WOTPBWULLOF OIHOTOIS iUuBs oY} WOT (1929 ‘SN, “loury) usuteds pasiojor 
ay} ul Os[e “(FGZO ‘SHYT “Loury) edd} oy} UL peAtesyo SdopOBIBYO [BpUOP PUB BUM OFOLO[OS SOUTGUIOD UOTJONAZSUOIAL SIG} UT [[NYS AY TL, 
‘OZIS [BIN}JVU Y}YSId-9UQC 
‘qoodst [B.10}V] S}I UL (FGZO ‘SNJA] “Loury) WoJoToyYs ad. Ay sysuarxobuow sninvsoovjNs_ oY} JO UOLN.YSUO9Y °Z “ST 


#529 SIW wIWV 
TAAL SISNITODINOW SNYWSOWLLISA 


4 AMERICAN MUSEUM NOVITATES [No. 127 


Psittacosaurus Protiguanodon 


Axial length from premaxillaries to 43d caudal ‘. 2310 mm. 1350 mm. 
te of fore limb extended 250 245 
‘“‘ hind limb extended 470 435 
e ‘“« humerus 119 123 
. ‘““ ulno-radius 90 90 
i “* manus 85 92 
Co ~ € fener 162 158 
“© tibio-fibula 179 167 
es “* pes 158 
16 Oe Digi wir 93 89 
Ratio, femur to tibia, femoro-tibial 90% 89% 
‘“‘ fore limb to hind limb, brachio-crural 53% 56% 


Comparison.—The above linear measurements as well as the femoro- 
tibial and brachio-crural ratios are very similar, demonstrating that 
both animals were: (1) Essentially bipedal in locomotion, with fore limbs 
well raised above the ground, a brachio-crural ratio of from 53% to 
56%. (2) Manus functionally tridactyl, since D.IV is greatly reduced. 
(3) Pes also functionally tridactyl-subtetradactyl, because D.I is of con- 
siderable size although well raised off the ground, while D.V is vestigial. 
(4) Osseous tendons connecting the fourth thoracic with posterior sacral 
or first caudal vertebra, indicating adaptation to a bipedal gait. (5) 
The normal walking position was probably semi-erect, as indicated in 
figure 2 (Psittacosaurus), figure 5 (Protiguanodon). (6) The ilio-sacral 
articulation of Psittacosaurus includes five vertebre, while the ilio- 
sacral articulation of Protiguanodon includes six vertebre. (7) Limb 
and foot bones of Pszttacosaurus are somewhat more massive, while the 
limb and foot bones of Protiguanodon are somewhat more slender. (8) 
Scapular arches, including clavicle, coracoid and scapula, are of about 
the same proportions in both species. (9) Pelvic girdle of Psittacosaurus, 
including the ilium, ischium and pubis, somewhat more massive than the 
pelvic girdle of Protiguanodon, in which the iliac crest is slender, the 
prepubic process much more slender and the ischium more slender and 
elongate than in Psittacosaurus. (10) There are 16 thoracic ribs in the 
Psittacosaurus thorax, which are slightly more robust than the 15 ribs in 
Protiguanodon. (11) From a comparison of the ten adaptations to a 
bipedal locomotion, we conclude that Protiguanodon was somewhat more 
cursorial in habit than Pszttacosaurus. 

A number of important additional characters in the pelvis should be 
noticed: (a) The absence of the ‘‘obturator processes’’ on the ischium, 
both in Psittacosaurus and Protiguanodon, a conspicuous difference from 
Thescelosaurus, Camptosaurus, Trachodon, Iguanodon, in which these 


1924] TWO LOWER CRETACEOUS IGUANODONTS 5 


Amer. /7us. 626/ ref 


IO 


A:M.6253 


Fig. 3 i 


Fig. 3B 
Fig. 3. Referred skull of Psittacosawrus mongoliensis (Amer. Mus. 6261). 


Found in the same formation (Oshih) as the type. This skull contains seven maxillary teeth in situ. 
Skull, one-third natural size. Seven maxillary teeth, natural size. Two of the same teeth are shown 
in Fig. 3A enlarged three diameters. 


Fig. 3A. Two superior teeth in referred skull of Psittacosaurus mongoliensis 
(Amer. Mus. 6261) enlarged three diameters. 


For comparison with restored tooth of Protiguanodon mongoliense type (Amer. Mus. 6253), enlarged 
three diameters. 


Fig. 3B. Type maxillary tooth of Protiguanodon mongoliense (Amer. Mus. 
6253) enlarged three diameters; anterior, exterior and interior aspects. After Osborn, 
1923, Fig. 5. 


processes are present. (b) Superior border of ilium not reflected laterally, 
—an important character separating Psittacosaurus and Protiguanodon 
from Iguanodon, Trachodon, Troédon. (c) Prepubic process shorter than 
anterior process of ilium, differentiating Pszttacosaurus and Protiguanodon 
markedly from Iguanodon, Trachodon, Thescelosaurus, Camptosaurus. 
(d) Ischia much flattened dorsoventrally. (e) Ischia not curved down- 
ward toward the posterior ends, distinct from Igwanodon, Camptosaurus, 
Troédon. (f) Postpubie processes short and slender,—probably a 
reduction character. 


THE PsITTACOSAURUS MONGOLIENSIS SKELETON 


Sxutu.—The perfectly preserved skull was described and figured in 
great detail in the type description! in which the skull characters are 


10p. cit., pp. 2-6. 


6 AMERICAN MUSEUM NOVITATES No. 127 


summarized as follows: ‘‘Psittacosaurus mongoliensis. Herbivorous 
diapsid reptile with predentary bone and horny beak. Maxillary teeth 
compressed, not fully known. Skull short and deep, narrow anteriorly, 
broad posteriorly. Rostrum prominent, parrot-like, edentulous. Nostrils 
small, orbits large. Infraorbital region and jaw heavy, with attachment 
for powerful muscles. Primitive dermal armature in head region; lateral 
osseous horns on jugals.”’ 

To the above description should be added the osseous sclerotic ring 
in the orbit represented in figure 2 and which is now exposed in the orbits 
of the skull. The osseous horns below the orbits and the impressions of 
the epidermal armature at the side of the jaw and throat led Osborn to 
the following conjecture:! ‘Genotype of Psittacosauride, new family. 
Skeleton and teeth only partly known; supposed primitive armored 
dinosaurs, possibly related to the fully armored Upper Cretaceous types.” 

Gregory remarks that Osborn’s term “osseous horns” on the jugal 
appears to be misleading; he believes that they correspond with a strong 
downwardly directed process on the jugal correlated with a development 
of the masseter muscle. Osborn is still disposed, however, to maintain 
that the osseous protuberances of the jugals are defensive bony spines 
(compare Ankylosaurus) and not muscular adaptations, and that the 
dermal impressions of the throat in Psittacosaurus are part of a dermal 
defensive system. 

TreretTH.—The teeth in the type skull (Amer. Mus. 6254) are still 
deeply buried in the matrix. Fortunate, therefore, is the discovery in 
the same Oshih formation of a second specimen of skull and skeletal 
parts (Amer. Mus. 6261) apparently referable to the species Psittaco- 
saurus mongoliensis, in which seven of the maxillary teeth are preserved. 
These referred teeth are shown natural size in figure 3 and enlarged three 
diameters in figure 3A; the sculpturing of these teeth is apparently differ- 
ent from that of the type tooth of Protiguanodon mongoliense (Fig. 3B). 

Family Psittacosauride: Skull abbreviate; rostrum edentulous, 
prominent, parrot-like; jaws deep; teeth 7-9, trilobed, brachyodont 
in dentaries and maxillaries; nostrils small; orbits large; cervicals 6; 
thoracics 16-15; sacrals 5-6; caudals 43+; clavicles reduced; brachio- 
crural ratio 58-56%; bipedal locomotion; manus and pes tridactyl-sub- 
tetradactyl; gait cursorial; ischia flattened, a broad ischiac symphysis. 

Subfamily Protiguanodontine Osborn, 1923, p.6: Type Protiguan- 
odon. Cannot be further defined until it is known what the complete 
separation is between Psittacosaurus and Protiguanodon. 


10p. cit., p. 6. 


‘9 SIq Ose aes /snunpsoopyisg YIM UOSUIvduI0D UT ‘(SouT] p9zJOP) patoyser Ayes rR] St [NYS oy, “Seul] pezop ur pezueseider oie sjied SuIssIul OY, “9821S 
jeinzeu yj} yz10-0u0 peonpoidas st ainsy yuaseid ay} ‘ezts [[Nj Sf suLMBAp [loued [eUIZIIO ay, “[[N AS 94} Suleq uoty10d paioyser A]UO 9y} ‘XI1jeBUI 9Y} UL pUNO;s SBM 
jt se oinsy quasaid 9y} Ul payueseddas Ajayeinoo" AJaA st puw pasodxe Ajoyo[/du09 useaq sey UOZe[ays edA}4 SIq} ‘UOTydtosep [BULSIIO ay} JO UOT}BoT[qQnd ay} VOUT 


“BIOSUOPY ‘UOTPEUIIO] Teg Tepug “(EGZ9 ‘SN “OULY) EZ6T ‘ULOGSO asuayobuow wopouvnbyoLq JO WOZP¥s adAy, ‘FS 


£529 ‘SAW ‘YIN 
FAL FSNITOINOW NOTONWMILOYS 


8 AMERICAN MUSEUM NOVITATES [No. 127 


GENERIC DistincTrions.—These two animals prove to be so similar 
in general size, proportions, gait, skeleton and limb segments, that while 
specific distinctions are very obvious indeed, and generic distinctions 
are more difficult than was at first supposed by Osborn (op. cit., pp. 9, 10), 
yet the genera may be distinguished as follows: 


Psittacosaurus Protiguanodon 
Cranium solid, with suborbital horns, Cranium slender, bones light, small 
large occipital condyles, epidermal tuber- occipital condyles (suborbital region 
cular armature on throat and side of face. and epidermal armature unknown at 


(See remarks, Gregory, Granger.) present). 
Neural arch of atlas vertebra elongate. Atlas vertebral elements apparently 
- abbreviate. 

Superior maxillary teeth flattened, with Superior maxillary teeth convex, 
asymmetrical trilobate sculpturing (Figs. | with Symmetrical trilobate sculpturing 
3, 3A). (Fig. 3B). 

Maxillary teeth: 7+. Thoracic verte- Dentary teeth: 9. Thoracic verte- 
bre: 16. Sacral vertebree: 5. Pubislarge, bre: 15. Sacral vertebree: 6. Pubis 
postpubis slender. and postpubis slender. 


COMPARISON OF PsITTACOSAURUS AND PROTIGUANODON 


DentitTion.—The striking difference in the external sculpture of 
the extremely short-crowned or brachyodont teeth is well displayed in | 
Psittacosaurus (Figs. 8, 3A), and in Protiguanodon (Fig. 3B); see also 
Fig. 4 of Osborn, op. cit., p. 8. The Psittacosaurus teeth, viewed 
externally, are relatively broad, flat, and the median ridge is on the 
posterior half of the tooth, whereas in Protiguanodon (Fig. 3B) the con- 
tour is a deep oval, the median ridge very prominent and directly in the 
center of the tooth. On wear these teeth become trilobate. 

Sxuuu.—The fully preserved skull of Pszttacosaurus, described and 
figured in detail in the type description (Osborn, op. cit., Figs. 2A, 2B, 
2C), differs from that of all iguanodonts previously described in its solid, 
massive characters, the sutures being partly closed, excepting the pre- 
maxillo-maxillary; this skull certainly had a powerful horny beak like that 
of a chelonian and was adapted to feeding upon very resistant plants. 
In the referred skull (Amer. Mus. 6261) the sclerotic ring and the same 
characters are observed; the dentaries are relatively short and massive. 
The sclerotic ring is observed in the type skull of Pszttacosaurus. Granger 
believes that the supposed ‘“‘epidermal tubercles” on the side of the skull 
of Psittacosaurus, regarded as epidermal impressions by Osborn, more 
‘strongly resemble concretions such as may be seen where no organic 
remains are present. 


"SNUNDSoIn}7}tS J ULOI} SOUL, PI}JOP UL UMBIp [[NYS oq} JO 9Ul]INO 
‘aZIS [BIN] CU YYS19-9UC 
‘yoodse [e1o}B] SPI UL (EGZO “SHY “Aoury) uUoxeYs addy asuayobuow uopounnbryjolig oY} JO UOTPNAYSUODIY "G “BIT 


£929 ‘saw YIWY 
3dkL FISNITOONOW NODONWDILOYS 


10 AMERICAN MUSEUM NOVITATES [No. 127 


In the type skull of Protiguanodon (Amer. Mus. 6253), as shown in 
figure 6, there is a marked contrast in the relatively slender character of 
the bones, the apparent sutural separation, the relatively small and 
broad occipital condyles, the very deep depression of the dentary below 
the nine dentary teeth. The inference is that this skull was adapted to 
less resistant food. 

Gregory further observes: “‘All the fragments of the Protiguanodon 
skull, when compared with Pszttacosaurus, Camptosaurus, Iguanodon, 
etc., show that it is unmistakably much closer to Pszttacosaurus. The 
small nostril, located at the top of the deep maxilla, the form of the 
frontoparietal, of the quadrate, squamosal, etc., differ from Pszttacosaurus 
chiefly in greater slenderness.”’ That Protiguanodon is on the whole far 
more primitive than ZJguanodon one can hardly doubt, but that it is 
geologically older, he would question for the following reasons: ‘“ (a) 
Hypsilophodon is certainly far more primitive than Iguanodon, but both 
occur in the Wealden. (b) Troédon of the Lance (Upper Cretaceous) is 
far more primitive than the ankylosaurs of older formations. (c) 
Thescelosaurus is a survivor in the Lance, retaining many primitive 
features recalling Hypsilophodon of the Wealden.” 

VERTEBR&Z.—Correlated with the massive skull of the Psittacosaurus 
type is the unusual elongate neural arch of the atlas (C 1) that may be 
seen in figures 1 and 2. The axis (C 2) and remaining cervicals (C 3-7) 
appear to be of the same length in the two genera. The thoracics, in- 
cluding neuropophyses, pre- and postzygopophyses and parapophyses, 
exhibit the same general characters in both genera. Ten bicipital ribs 
are observed in Psittacosaurus; 9+ bicipital ribs in Protiguanodon. 
Five sacrals unite with the iliac crest in Pszttacosaurus; 6 sacrals 
unite with the iliac crest in Protiguanodon. Eight free caudal ribs are 
observed in Protiguanodon; the anterior caudals of Pszttacosaurus are 
imperfectly preserved; the 9th to the 20th caudals in Protiguanodon 
have consolidated ribs or pleurapophyses; the 15th to the 34th caudals in 
Protiguanodon bear chevrons. 

Limp CHaArAcTeRS.—The limb characters in Psvttacosaurus and 
Protiguanodon are clearly displayed in figures 1, 2, 4 and 5; resemblances 
and contrasts are as enumerated above. The ulna, radius and manus 
resemble in proportions the manus of Hypsilophodon foxii; Digit IV is 
somewhat more reduced than that in Hypsilophodon, and there is no 
trace of D.V; large D.I is closely applied to D.II and shows no sign 
of the abduction characteristic of Igwanodon. The manus is still of 
locomotor type and shows little if any prehensile adaptation. The 


1924] TWO LOWER CRETACEOUS IGUANODONTS 11 


H . AM 6253 


teen 


Fig. 6. Type skull, largely restored, of 
Protiguanodon mongoliense (Amer. Mus. 
6253). 


The type jaw of the same skull is shown in Os- 
born, 1928, p. 8, Fig. 4. 


pes of Protiguanodon (Fig. 4) is double the size of the manus, from which 
we may infer that the pes performed twice as much work as the manus. 
Locomotion.—The proportions of the organs of the axial and appen- 
dicular skeletons are approximately similar to those of Iguanodon bernis- 
sartensis. In Iguanodon the vertebral column on the whole is relatively 
more massive. The ossified tendons of Jgwanodon extend down along the 
sides of the elongated spines of sixteen to twenty of the anterior caudals, 


12 AMERICAN MUSEUM NOVITATES [No. 127 


whereas in Protiguanodon the ossified tendons stop at the first or second 
caudal. The neural spines of the anterior caudals are not elongate. 

Consequently, we infer that adaptation to bipedal locomotion, to a 
suberect position, to arboreal feeding habits, was much further advanced 
in Iguanodon than in Protiguanodon. This comparison supports the idea 
that Proteguanodon of Mongolia is far more primitive in structure and 
may belong to a much older geologic stage than Iguanodon of the Wealden 
of England and Belgium. 


PELVES OF PsITTACOSAURUS AND PROTIGUANODON 


Comparison of the pelves of these Mongolian species with the pelves 
of five other iguanodonts seen in the diagrammatic drawing (Fig. 7) 
shows the dominant ornithischian type. We observe that the prepubis 
of Psittacosaurus and of Protiguanodon is more abbreviate than in any 
of the other genera, while the postpubis has the same proportions as in 
Trachodon, somewhat less developed than in Iguanodon, far less developed 
than in Camptosaurus, Thescelosaurus or Hypsilophodon. The ischia of 
Protiguanodon and Psittacosaurus are relatively well developed; a 
very distinctive feature of the ischium is the ischiac symphysis, namely, 
the broad plate-like union of the ischia posteriorly, as observed in Thescelo- 
saurus. The ilia are relatively longer than in any of the other iguanodonts 
figured. 

Consequently, we may sum up the comparative characters of the 
pelvis in Psittacosaurus and in Protiguanodon as follows: (1) Prepubis 
slender; small pubic foramen. (2) Postpubis very slender, closely 
apposed to ischium. (3) Ischia relatively elongate, flattened, produced 
into a very broad ischiaec symphysis. (4) Ilia relatively elongate and 
depressed, extending anteriorly beyond the extremity of the prepubis. 

CoMPARISON OF PELVvES.—In figures 8 and 9 we are afforded a de- 
tailed study of the pelvis in Psittacosaurus and in Protiguanodon made 
after complete removal from the matrix, that of Protiguanodon being 
especially perfect in preservation. It is. shown that: (1) The postpubis 


Fig. 7. Pelvic characters of the Iguanodontia. Diagrammatic. 


The seven pelves here figured are reduced for purposes of comparison to the same absolute size, 
regardless of actual wide differences in scale. The pelves should be examined in descending geo- 
logic order as follows: 

Montana, Upper Cretaceous. Trachodon mirabilis. After Brown, 1913. 

Montana, Upper Cretaceous. Thescelosaurus neglectus (Amer. Mus. 5889). After Brown. 

Belgium, Lower Cretaceous. Iguanodon bernissartensis. After Dollo, 1883, slightly modified. 

Wyoming, Upper Jurassic. Camptosaurus medius. After Gilmore, 1909. 

Mongolia, Oshih Formation. Psittacosaurus mongoliensis (Amer. Mus. 6254). After Osborn, 1923. 
eh to Ondai Sair Formation. Protiguanodon mongoliense (Amer. Mus. 6253). After Osborn, 

England, Lower Cretaceous, Wealden. Hypsilophodon foxii. After Hulke, 1882. 


Thescelosa WLS SATE 


Fig. 7. See legend on opposite page. 
13 


Prot iguan odon 


Amer /us. 6253 


Psi1llacosaurus 
Amer /us. 6254. 


Fig. 8. Pelves of Psittacosaurus and Protiguanodon. 


Psittacosaurus mongoliensis, type skeleton (Amer. Mus. 6254), drawn from both sides. 
Protiguanodon mongoliense, type skeleton (Amer. Mus. 6253), right lateral aspect partly recon- 


Lower. 


Upper. 
structed from left side. 
Both figures one-half natural size. 


14 


Frotig wanodon 


Amer. /Tus. 6253 


Proviguanodon 


Amer. /Jus. 6253 


Fig. 9. Pelvis and sacrum of Protiguanodon mongoliense, type skeleton (Amer. 
Mus. 6253). 


Upper. Internal aspect of right os innominatum showing attachment of 6 sacral vertebre. 
_ ., Lower. Superior aspect of pelvis showing 6 sacral vertebrx, also coalescence of ischia at the 
ischiac symphysis. 


16 AMERICAN MUSEUM NOVITATES [No. 127 


is elongate, slender and closely appressed to the flattened under surface 
of the ischium; (2) the prepubis (Fig. 8) is much more robust in Psztta- 
cosaurus than in Protiguanodon; the pubic foramen is apparently an 
enclosure between the postpubis and the peduncle or acetabular border 
of the pubis. As these pelves are primitive, the postpubis appears as of 
secondary origin, or part of.the extension of the primitive pubis. In 
figure 9 there is clearly shown the internal and superior aspects of the 
Protiguanodon innominate bone, namely: (1) Ilium with rugose attach- 
ment of six sacral vertebre; slender anterior or pubic peduncle. (2) © 
Short ischiae peduncle. (3) Slender prepubis when seen from above. 
(4) Flattened ischia, when seen from above, conjoined posteriorly into 
the ischiac symphysis. 

Comparison (Fig. 7) of the pubic components in these iguanodonts 
would support the view that the prepubis is the primary element 
(=pubis), the postpubis a secondary element. 

Prepusis.—Gregory remarks that the Protiguanodon skeleton affords 
convincing evidence for his view that the prepubic processes of Orni- 
thischia diverge widely on each side toward the last rib. He doubts 
whether the pelves of Psittacosaurus and Protiguanodon afford any 
support of the view (see Osborn above) that the postpubis is a new 
process. These animals are very far removed in skull and other struc- 
tures from the primitive Triassic Pseudosuchia which appear to be their 
nearest relatives. The postpubic process lies immediately below the 
pubic foramen in the position of the true pubis of embryo birds and adult 
Triassic Erythrosuchus; consequently the postpubis = pubis. 


ACKNOWLEDGMENTS 


The author is greatly indebted to the following persons in the prep- 
aration of this article: First, to Otto Falkenbach for many weeks of 
laborious and skilful work in exposing and restoring these type skel- 
etons. Second, to Mrs. L. M. Sterling for her skill in preparing life- 
size illustrations of these two skeletons, which will subsequently be 
published, the present reduced illustrations doing scant justice to 
the original pencil drawings: Third, to Professor William K. Gregory 
for valuable notes and suggestions throughout the study of the 
skeletons and for the critical comments which are inserted in the present 
text, together with the comments of Mr. Walter Granger on the “epi- 
dermal tubercles.” 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 128 Tue AMERICAN Museum or Natura History Sept. 22, 1924 
New York City 2 


56.81,9(117:51.7) 
SAUROPODA AND: THEROPODA OF THE LOWER CRETA- 
CEOUS OF MONGOLIA! 


By Henry FAIRFIELD OSBORN 


The discovery of Sauropoda was one of the most interesting results 
of the Expedition. From the Psittacosaurus mongoliensis life zone, 
Oshih (Ashile) formation, Red Mesa, Oshih Basin, Mongolia, were 
collected by the Expedition of 1922 two characteristic sauropod teeth 
which bear considerable resemblance to those of Camarasaurus Cope and 
of Caulodon Cope, and are widely different from those of Diplodocus 


Fig. 1. Key to Camarasaurus skull (Amer. Mus. Cope Coll. 5761), shaded; 
outlines restored from referred skull (Amer. Mus. 467). One-tenth natural size. 
After Osborn and Mook, 1921, p. 286, fig. 29. 


Marsh. Both the characters of the teeth and their occurrence in a new 
locality and life zone prompt their reference to a new genus of Sauropoda, 
family Camarasauride. 

A comparison of the type and paratype teeth with those of Camara- 
saurus (see Osborn and Mook, 1921, pp. 286, 287, Figs. 29 and 30, Fig. 


_|Publications of the Asiatic Expeditions of The American Museum of Natural] History. Contri- 
bution No. 26. 


2 AMERICAN MUSEUM NOVITATES [No. 128 


1 of this article) indicates that the type teeth of Aszatosaurus belong in 
the lower jaw, the smaller type tooth being posterior, the larger paratype 
tooth being anterior in position; the family resemblance to Camarasaurus 
is quite strong. The generic characters are very distinct; all the teeth 
of Camarasaurus (op. cit., Fig. 30) and of the type of Caulodon diversidens 
(op. cit., Pl. Lx) have subspatulate crowns, expanding at the base, con- 
tracting toward the summit, with a prominent median and symmetrical 
internal ridge. Compare Asiatosaurus type (Fig. 2) with Caulodon 
diversidens type (op. cit., Pl. Lx, figs. 1a, 1b, 5a, 5b). A second important 
difference is the asymmetry of the summit of the crown in Asiatosaurus 
as compared with the symmetry of the crown summit in Camarasaurus. 
A third difference is the more robust fang anteroposteriorly seen in the 
paratype (Fig. 3) of Asiatosaurus, as compared with the much less robust 
fang in the large anterior and central teeth of Camarasaurus. 


= re ae 5732. 
A.M. 6264 TYPE 
Fig.2. Type of Asiatosaurus mongoliensis (Amer. Mus. 6264). Supposed 
posterior tooth in the dentary series. Natural size. 


Asiatosaurus mongoliensis, new genus and species 


Typr.—The tooth chosen is the most characteristic (Amer. Mus. 6264), repro- 
duced herewith natural size (Fig. 2); this is believed to be a posterior tooth of the 
lower jaw or dentary bone. The paratype (Amer. Mus. 6294) is of much larger size 
and is believed to be an anterior tooth of the dentary. The type and paratype teeth 
have certain general characters in common but differ greatly in size. 

Hor1izon.—Oshih (Ashile) formation, Psittacosaurus mongoliensis life zone, 
Mongolia. 

Locatiry.—According to Granger these teeth were found approximately on the 
same level about one hundred feet apart; we are certain that they do not belong to 
the same individual; it is probable that they belong to the same species. We cannot 
be positive, however, that they belong to the same species, or as to their position in 
the jaw, until more complete material is found with the teeth in situ. 


1924] SAUROPODA AND THEROPODA OF MONGOLIA 3 


GENERIC CHARACTERS.—Asiatosaurus, derived from the Greek ’Agva, Asia, and 
gavypa, lizard. Type: (?Posterior) maxillary or dentary tooth, small, concavo- 
convex; subspatulate crown, slightly expanding superiorly; two lateral external and 
one median internal vertical grooves. Paratype: (?Anterior) enlarged tooth with 
transverse equaling anteroposterior (ap. 126 mm.=tr. diameter) diameter of fang, 
expanded summit, convexo-concave (internal section), shallow lateral and internal 
vertical grooves. 

Type Description.—(1) The type tooth (Fig. 2) measures 41 mm. in height as 
preserved, 47 mm. in circumference, 18.2 mm. in anteroposterior direction (.e., 
parallel with the direction of the tooth-row) and 11.7 mm. perpendicular to this direc- 
tion; the external surface is convex in both vertical and horizontal directions; near 


nat. szze 
A. 1. 6294 PARATYPE 


Fig. 3. Paratype of Asiatosaurus mongoliensis (Amer. Mus. 6294). Supposed 
to be an anterior tooth of the dentary. Natural size. ; 


each edge is a broad, shallow vertical groove. (2) The internal surface is flat at the 
base, but concave in both vertical and horizontal directions near the apex of the crown. 
The surface therefore roughly resembles the inner or upper surface of a spoon. A 
small median groove extends vertically through the base of the bowlshaped depres- 
sion. (3) The edges are straight and parallel near the base of the tooth, but near the 
crown are beveled by worn surfaces; the latter are asymmetric in development, one 
being rather convex in profile and extending down only 15 mm. from the apex, while 
the other is concave and extends downward about 30 mm. The two worn surfaces 
unite at the apex into a single surface. The unequal degree of wear on opposite edges 
gives an asymmetric appearance to the entire tooth. The worn surfaces face some- 


4 AMERICAN MUSEUM NOVITATES [No. 128 


what inward, being entirely visible on the internal aspect of the tooth, and not at all 
visible on the external aspect. This indicates that the tooth is near the posterior end 
of the dental series. 


Comparison of its asymmetrical form with that exhibited by the 
teeth in the mounted skull of Morosaurus in the American Museum 
(Amer. Mus. 969) indicates that it belongs to the right side. 


Paratypr.—(1) The paratype tooth (Amer. Mus. 6294) is about 74 mm. in 
length as preserved, and about 80 mm. in circumference; at the base its anteroposte- 
rior diameter is equal to its transverse, 27 mm. The crown is damaged, the expanded 
upper portion being broken; near the tip the anteroposterior diameter exceeds the 
transverse.! This contrasts with the broader and more flat horizontal sections of 
Cope’s Camarasaurus (=Caulodon) teeth. (2) The external surface of the tooth is 
convex, both in vertical and horizontal directions. The internal surface is slightly 
concave vertically, and near the tip horizontally as well. For two-thirds of its length, 
however, the concavity of the internal surface is interrupted by a low vertical ridge 
which disappears near the tip of the crown; this ridge is not as prominent as similar 
ridges in the type teeth of Cope’s Camarasaurus (=Caulodon); it is faintly striated 
vertically. Anteroposteriorly the crown is somewhat asymmetrical, but less so than in 
Camarasaurus. The summit of the tooth is imperfect, but the portion preserved ex- 
hibits a considerable amount of wear surface on the external side of the tooth; a 
small, smooth surface on each inner edge may indicate wear. The great extent of worn 
surface on the external side of the tooth indicates that this is a lower tooth, and its 
large size and robust proportions indicate that it was situated near the anterior end 
of the jaw. 


Amer. Mus. 6532. Another tooth referable to Aszatosaurus was 
found in the Oshih Basin, August, 1923 (Field No. 378). A sauropod 
tooth fragment also from the Oshih Basin bears the same field number, 
Le., 378. 


SAUROPODS OF THE OSHIH AND ONDAI SAIR FORMATIONS 


In all, the remains of eight individual sauropods were found in the 
Oshih and Ondai Sair formations, of slightly different geologic ages and at 
different levels, indicating that this was an important center of sauropod 
distribution, as follows: 


Osuin ForRMATION, PSITTACOSAURUS MONGOLIENSIS LIFE ZONE, EAST END OF 
Rep Masa. 
1. Type of Asiatosaurus mongoliensis (Amer. Mus. 6264). 
2. Paratype of Asiatosaurus mongoliensis (Amer. Mus. 6294); same level as 
type, 200 yards distant. 
3. Several limb bones, weathered, broken, and not collected; same level as 
type and paratype teeth, not far distant. 


*The term ‘‘anteroposterior’’ here refers to the direction parallel to the toothrow, and ‘‘trans- 
verse’ at right angles to this direction. These directions may or may not correspond with the 
general directions of the skull, depending upon the position of the tooth. 


1924} SAUROPODA AND THEROPODA OF MONGOLIA 5 


4. Referred tooth of Asiatosawrus (Amer. Mus. 6532), Field No. 378. 
5. A sauropod tooth fragment (Field No. 378). 
THREE MILES EAST OF RED MESA, FROM A HIGHER LEVEL, ABOUT 600 FERT. 

6. Amer. Mus. 6533, two anterior dorsal vertebre, badly weathered, several 
ribs, and one chevron; three ribs and chevron collected (Field No. 106); 
see photograph, Fig. 4. 

7. Single dorsal vertebra, badly weathered, not collected. Found 100 feet 
from vertebre (Amer. Mus. 6533) possibly belonging to the same in- 
dividual. 


OnpDAI Sarr Formation, Mr. Uskuk. 
8. One rib (Amer. Mus. 6258). 


Fig. 4. Two anterior dorsal vertebre, several ribs, and a chevron of a sauropod 
dinosaur in situ. Oshih (Ashile) formation, Oshih Basin, August, 1923; Albert F. 
Johnson, collector. Owing to their much weathered and disintegrated condition, the 
two vertebrze were not taken. 


ConcLusion.—Awaiting the evidence afforded by more perfect 
material, Asiatosaurus seems to resemble Camarasaurus in the subspatu- 
late form of the summit of the crowns, in its large anterior teeth diminish- 
ing gradually to the smaller posterior teeth of similar general spatulate 
pattern. We await with interest comparison of these teeth with those of 
the analogous African genera. 


A/7, 6265 
(gears 


nat. Size 
Pesiinies com 
» 
» 


Loe 


Ein ee” 


Fig. 5. Type 
of Prodeinodon 
mongoliensis 
(Amer. Mus. 
6265), Oshih 
(Ashile) form- 
ation, Oshih 
Basin, Mon- 
golia; collect- 


AMERICAN MUSEUM NOVITATES [No. 128 


Prodeinodon mongoliense, new genus and species 


Tyrr.—Amer. Mus. 6265. Upper section of a carnivorous 
dinosaur tooth, Oshih (Ashile) formation, Mongolia (Fig. 5). Col- 
lected by Walter Granger. 

PARATYPE.—Complete carnivorous dinosaur tooth (Amer. 
Mus. 6531), Oshih (Ashile) formation, Mongolia. Collected by 
Walter Granger. . 

; Horizon AND Locatiry.—Oshih (Ashile) formation, Psittaco- 
saurus mongoliensis life zone, Red Mesa, Oshih Basin, Mongolia. 

GENERIC CHARACTERS.—A theropod or carnivorous dinosaur, 
tooth crown with flattened sides, rounded anterior border, com- 
pressed posterior border terminating in a serrate ridge. Diameters: 
ap. 134 mm., tr. 82 mm. Closely minute serrations of posterior 
edge. 

The Theropoda or carnivorous dinosaurs of the 
period, contemporaries of Pszttacosaurus and A siatosaurus, 
are represented by a single tooth fragment (Amer. Mus. 
6265) which is shown natural size in figure 5, as collect- 


ed by Walter 
Granger. Nat- 
ural size. 


ed by Granger in 1922. 
As Asiatosaurus may 
prove to be close to 
Brachiosaurus or to Camarasaurus, so 
this carnivorous dinosaur tooth may 
prove to belong to an Old World or 
New World genus such as Megalosaurus 
Buckland, Dryptosaurus or Allosaurus 
Marsh. Meanwhile it is provisionally 
given the name Prodeinodon, referring to 
its greater geologic age than Deinodon 
Leidy of the Judith River, Upper Cre- 
taceous. 

In the type of Megalosaurus (Buck- 
land, 1824, Owen, 1860, p. 260, fig. 75) 
both the anterior and posterior edges 
of the teeth are serrate; the teeth so 
far as preserved are similar or homo- 
dont.) The teeth of Deinodon are also 


AM653/ 
PARATYPE 


Fig. 6. Paratype of Prodei- 
nodon mongoliensis (Amer. Mus. 
6531), Oshih (Ashile) formation, 
Oshih Basin, Mongolia; collected 
by Walter Granger, August, 1923 
(Field No. 378). Natural size. 


sharpened on both edges (Leidy, 1860, Pl. rx, figs. 21-48); they are 
heterodont, the anterior teeth (named Aublysodon by Leidy) being 
rounded in front and flattened behind with double serrations on the 


two posterior borders. 


The present type of Prodeinodon agrees most 


closely with the tooth figured by Leidy (Leidy, 1860, Pl. rx, figs. 33 


1924] SAUROPODA AND THEROPODA OF MONGOLIA 7 


and 34, described on p. 144). Leidy observed 
that Deinodon was probably heterodont; the 
same may be true of Prodeinodon. A _ closer 
resemblance to the Prodeinodon type is seen in 
Lambe’s figure of Deinodon explanatus Cope 
(Lambe, 1902, p. 49, Pl. xv, figs. 11 and 12, 
Fig. 7 of this article). Consequently the generic 
name Prodeinodon is selected for this Oshih 
type of Mongolia. The tooth differs in its more 
rounded and less laterally compressed section 
from the tooth referred to Deinodon explanatus 
Cope. 


ag 
Fig. 7. Referred 
tooth of Deinodon ex- 
planatus Cope. After 
Lambe, 1902, Pl. xv, 
Figs. lland12. Twice 
natural size. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 132 Tue AMERICAN Museum or Narturat HIsTorY Sept. 29, 1924 
New York City 


56.78M (1181:51.7) 
A NEW SPADEFOOT TOAD FROM THE OLIGOCENE OF 
MONGOLIA WITH A SUMMARY OF THE EVOLUTION 
OF THE PELOBATIDé! 


By G. K. Nose 


The Third Asiatic Expedition of the American Museum discovered 
in the Hsanda Gol formation in the Tsagan Nor basin of outer Mongolia 
a beautifully preserved pelobatid. This single specimen is of unusual 
interest as representing the group from which the modern spadefoot 
toads arose, to spread on one side across Europe and on the other into 
North America. It is the oldest known fossil which belongs unquestion- 
ably to the Pelobatide. 

The Hsanda Gol formation has been assigned to the Oligocene.” 
The specimen described below was found associated with a varied mam- 
malian fauna, mostly rodents. Eleven species and nine genera of the 
latter have already been described by Matthew and Granger. The 
formation consists mostly of sandy clays. The terrain during Oligocene 
times was therefore similar to that to which modern spadefoot toads are 
restricted, except that it may have contained more clay and less sand. 
The climate was apparently semiarid. 


PELOBATID 
Macropelobates osborni,‘ new genus and species 


Typr.—No. 6252; an incomplete skeleton, crushed anteriorly, but well preserved. 

Horizon AND Locatity.—Hsanda Gol formation, Tsagan Nor basin, Mongolia. 

Dracnosis.— Undoubtedly a pelobatid in that it exhibits the following characters: 
Anomoccelous®, with the coceyx not ankylosed to sacrum; coccyx with a single 
condyle; teeth present on upper jaw; coracoids suggesting an arciferal condition; 
an enormous prehallux (spade); epiphyses absent (cartilaginous); a bony encrusta- 
tion on the frontoparietal, nasal and squamosal. 

Agrees with Pelobates in: Maxillary teeth guarded mesially by a ridge, but this 
more pronounced than in Pelobates; neural processes of anterior vertebre long and 


: [ae of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 27. 
pune W.D., and Granger, Walter. 1923. Amer. Mus. Novitates, No. 101, p. 1. 
p. cit. 
_ ‘Named for Professor Henry Fairfield Osborn, who has been so largely the inspiration of the Third 
Asiatic Expedition. 
5For this term see Noble, 1922. 


Fig. 1. Macropelobates osborni Noble, type specimen. 
A, dorsal, and B, ventral aspect. 


2 


Mx. 


Fr p. 


Ser 
Vert.2 


T fib. A 


Ast. 
Cal, 


Coc 


Hum. 


Rad. ul. 
/1. 


Sac. 

Coc. Fem. 
Prh. 

fem: 


Cal, 


ree Ast. 


Explanation of Fig. 1: Ast., Astragalus; Cal., Caleaneum; Coc., Coccyx; Eth., Ethmoid; Fem., 
Femur; Fr. p., Frontoparietal; Hum., Humerus; II., Ilium; Mx., Maxilla (fragment); Na., Nasal (frag- 
ment); Pr.h., Prehallux; Pro., Prodtic; Pub. isch., Puboischium; Rad. ul., Radioulna; Sac. Sacral 
diapophysis; Sc., Scapula (part); T. fib., Tibiofibula; Vert. 2., Vertebra (apparently the 2d); Vert. 
3., Vertebra (apparently the 3d). 


Fig. 2. Vertebral columns ventral aspect. A, Scaphiopus couchii Baird; B, Macropelo- 
bates osborni Noble; C, Pelobates fuscus (Laurenti). 


1924] A NEW SPADEFOOT TOAD OF MONGOLIA 5 


pointed (possibly exaggerated through crushing); coracoid with an expanded me- 
sial end; squamosal with a large sheath of bony encrustations upon its lateral sur- 
face; presacral vertebra with transverse processes directed very obliquely forward 
(but not quite as much as in P. fuscus). 

Agees with Scaphiopus in: Puboischial plate as long as at least three presacral 
vertebree; ilium longer than the femur; coccyx longer than the sacral wings; only 
one segment in the prehallux (as in S. couchii and S. hammondii, but not Scaphiopus 
holbrookii and S. dugesiz). 

Distinguished from the other fossorial pelobatids by: Size very large; coccyx and 
sacrum not fused (this condition appears as a variation in Pelobates); transverse 
processes of vertebre immediately anterior to the sacrum, narrow; sacral diapo- 
physes as wide as the ventral surfaces of three and one-half presacral vertebrae 
(three in Scaphiopus, five in Pelobates); transverse diameter of sacral wings contained 
in the longitudinal diameter a trifle over two times; coccyx slender; radioulna 
rather wide and slightly curved (Fig. 3E); astragalus and calcaneum very slender 
(Fig. 3G); puboischial plate solid, the pubis apparently ossified and forming part of 
this plate. 

DescrIPTION OF Type SPECIMEN.—Skull strongly ossified and sheathed as in 
Pelobates with a secondary deposit of bone, the frontoparietal completely covered 
with this pitted sheathing (except anteriorly where it has obviously been broken away) ; 
a fragment of a nasal present and covered with this ossification; the squamosal 
destroyed anteriorly, but the remaining part sheathed laterally as in Pelobates, this 
sheathing more than twice as long and twice as wide as the anteroposterior diameter 
of the unsheathed dorsal surface of the squamosal. Ethmoid entirely ossified, lateral 
wings fairly well preserved and similar to the ethmoid of Scaphiopus except that they 
appear more massive; left nasal represented by two fragments which would indicate 
that the nasals had originally the same form as in Scaphiopus and Pelobates, at least 
they were in broad contact in the mid-line; frontoparietal single, its bony sheathing 
fractured along the sides posteriorly and hence no lateral processes as in Scaphiopus 
or Pelobates; both proétic and dorsal part of the squamosal exposed and without 
bony encrustations; squamosal and pterygoid represented only by fragments which 
may be matched with parts of the homologous bones in Scaphiopus; a fragment of 
the right maxilla present; the teeth walled in mesially by a pronounced ridge 
(Fig. 4A). 

Vertebral column typically anomoccelous, the coccyx articulating by a single 
condyle with the sacrum; the dorsal surfaces of only three of the vertebre exposed 
and these surfaces very much broken. Apparently the anterior vertebra were provided 
with long transverse processes and the neural spines were long and slender; the long- 
est neural spine retained is slightly longer than the ventral surfaces of any of the 
presacral vertebre, and about equal to the greatest anteroposterior diameter of the 
exposed dorsal surface of the proétic. The transverse processes of the two vertebra 
immediately anterior to the sacrum well preserved, these proportionately more slender 
than in Pelobates and directed not so obliquely forward as in that genus. The longest 
of these transverse processes is a little longer than the greatest length of any of the 
vertebrae, possibly just equal to these vertebre if they were disarticulated. Sacral 
vertebre with strongly dilated diapophyses, the transverse diameter of these wings 
slightly shorter than half the longitudinal diameter; longitudinal diameter of these 
diapophyses equals the longitudinal diameter of three and one-half presacral verte- 


6 AMERICAN MUSEUM NOVITATES [No. 132 


bre; coccyx broken but longer than the longitudinal diameter of the sacral wings, 
nearly equal in-length to five of the vertebre together; coccyx in no way contributing 
to the sacrum. ; 

Coracoids both broken but very similar to Pelobates; the mesial end expanded, 
this expansion over twice as wide as the narrow middle portion of the coracoid; 
glenoid part of the coracoid expanded about once and a half or more the width of the 
narrow portion of the coracoid. Both humeri broken but the fragments in no way ~ 
different from those of Pelobates; radioulna broken but apparently wider than in 
Pelobates and with a shallower distal articulation (Fig. 3E); part of the right scapula 
present, massive; two fragments, which I take to be the right and left suprascapule, 
differ from similar structures in Pelobates and Scaphiopus in having the posterior 
process narrower, more massive (Fig. 4B). 

Pelvis apparently more complete than in Pelobates and Scaphiopus; pubis may 
have been ossified, at least there is no space for a pubic cartilage. Longitudinal 
diameter of the acetabulum equals the distance between the posterior margin of the 
acetabulum and the posterior margin of the ischium; longitudinal diameter of the : 
puboischial plate equal to the length of three and one-third vertebre; ilium, as 
measured from the apparent suture between the ilium and ischium, distinctly longer 
than the femur; femur very slightly longer than the tibiofibula; femur and tibio- 
fibula without epiphyses (these cartilaginous in life). Caleaneum and astragalus much 
slenderer than in either Pelobates or Scaphiopus (Fig. 3F and G); caleaneum slightly 
longer than astragalus; these elements free from each other and less in contact than in 
Scaphiopus; caleaneum slightly less than half the length of the tibiofibula. Pre- 
hallux enormous and formed of a single bone, which apparently is in contact with the 
astragalus as in S. couchii and S. hammondii and not separated by a bone (which 
has been variously named) as in S. holbrookii and P. fuscus; apparently a single bone 
lateral to the prehallux and also a space for a cartilage asin S. couchii. Prehallux less 
than half as long as the third metatarsal; the metatarsals more or less complete; the 
third metatarsal a trifle shorter than the astragalus. 


MEASUREMENTS 

Length of frontoparietal... few iloadalind ss Side jethes Oe 
Distance from middle of aceiptd to paleo. at squarmenal Ra yahe icy bel aie a, 21 He 
Longitudinal diameter of the exposed dorsal surface of proétic........... Gah 
Transverse diameter of the squamosal encrustation. . fe a SG 
Distance from the posterior edge of the base of the Steead diapophyss te ve 

the anterior margin of the fourth vertebra anterior to the sacrum.... 26.5 “ 
Longitudinal diameter of the sacral diapophysis......................... 22 © 
Transverse diameter of the sacral diapophysis.................-+.+++5- 9.5 “ 
EGY occ Ate ce Se ple LR eR acts aad CONE UA nhc Meichara se be Dice re chee ice yi pen 
J EEEG IES UST S SPce tee ehONe | bot ROOMS et hart SM Se oN er yr URN Reem ES eae ee PAC 193 ive 
Aitmsinca, vioiseliall jewture.;, P25 a0. Loves d a isiace dee See we wed hated op alae ea 
HST E GA Nile ctiebeks SRB aN. cabo <i Aa hd ae Re Meee a Mi Avid aos ee tees Alvis 
ARIES AON i Asc cys sae ach ile ale Bia SLA omg sa we SIR Ea oe Aad ache REN 40 s 
RG GREW Eee Eat Whe ath sinc, wel ee Mae ole BRAUER hind CR ASE 19,5 98 
Longitudinal diameter of prelalhur.:.o 2: pte Sects Gass se aed 6. eee fale oS GB 


Berit papme elo. Vo. ord kuin Gow bee AMMEA yee, eo teeehet Storaients epee Oe eG pare 


Fig. 3. A comparison of various skeletal elements of Pelobates, Scaphiopus and 
Macropelobates. A, left coracoid of Pelobates fuscus (Laurenti); B, left coracoid of 
Macropelobates osborni Noble; C, right coracoid of Macropelobates osborni Noble; 
D, left radioulna of Scaphiopus couchii Baird, median aspect; E, left radioulna of 
Macropelobates osborni Noble, median aspect; F, left foot of Scaphiopus couchii Baird, 
ventral view; G, left foot of Macropelobates osborni Noble, ventral view. 


7 


Fig. 4. A, fractured piece of maxilla of Macropelobates osborni Noble; B, 
a portion of the suprascapula of Macropelobates osborni Noble; C, pelvic girdle of 
Scaphiopus couchii Baird; D, pelvic girdle of Macropelobates osborni Noble. 


8 


1924] A NEW SPADEFOOT TOAD OF MONGOLIA 9 


THE ORIGIN AND DISPERSAL OF THE PELOBATIDA 

The species described above exhibits the characteristic features of 
the fossorial pelobatids. These diggers are not the most primitive pelo- 
batids. They form a rather uniform group, and some, such as Fejérvary 
(1921, 1923), who delight in naming every series of related genera, would 
distinguish this group as a subfamily. I have recently had occasion 
to sketch the broad lines of evolution within the Salientia (Noble, 1922). 
Since then it has been my privilege to study more intensively the primi- 
tive discoglossids. This work has served to strengthen the theses ex- 
pressed and has lent no support to the recently devised ‘‘systems”’ of 
Bolkay (1919, 1922) and Fejérvary (1917, 1921, 1923). Asa matter of 
fact, the first of these systems is based. upon a misconception as clearly 
shown by Fejérvary (1921), while the second has little in its support, if 
enough forms are considered. The double sacrum which Fejérvary dis- 
cusses in such detail cannot be considered primitive or very distinctive. 
Atelopus, for example, may or may not exhibit such a condition. 

The pelobatids arose from the more primitive discoglossids of which 
LIiopelma and Ascaphus are the only surviving representatives. This 
group was characterized by such fundamental features as amphiccelous 
vertebre, a coccyx with a single condyle, and ten vertebre, exclusive of 
coccyx which is itself of a primitive type. The pectoral girdle is exceed- 
ingly primitive in lacking an acromion and in having the scapula extend- 
ing down into the “procoracoid region”’ as a single element. These 
features will be discussed in more detail elsewhere, but in the present 
connection, I desire to emphasize the primitive nature of these two 
genera, and hence will distinguish them as a family distinct from their 
discoglossid relatives. For this family the name LiOPELMIDZ! seems 
appropriate. 

The more generalized pelobatids as represented by Megalophrys 
have advanced beyond the Liopelmide in: (1) the ossification of the 
intervertebral cartilages (centra) which usually adhere to the more 
anterior vertebre forming a proccelous vertebral column; (2) reduction 
of the number of presacral vertebre to eight; (3) loss of ribs apparently 
by fusion with the vertebre; (4) restriction of the ossification in the 
ventral part of scapulocoracoid and the development of an acromion; 
(5) reduction of pubis and expansion of sacral diapophyses. There were 
other less obvious changes in the hyoid and appendages. The muscula- 
ture, however, was not greatly modified, except that certain elements 
such as the caudalipuboischiotibialis were lost. The pelobatids appar- 


1T follow present-day custom in using the oldest generic name in forming the family name. 


10 AMERICAN MUSEUM NOVITATES [No. 132 


ently early developed excessive bony or calcareous deposits in their 
derm, for we find such encrustations not only.in the integument of the 
head but also along the back of several species of Megalophrys. 

The second stage in the progressive modification of the pelobatids is 
represented by Macropelobates. The prehallux has become enlarged, 
the bony encrustations ankylosed to the skull and the sacral diapophyses 
greatly expanded. The two former modifications are apparently adap- 
tive and correlated with a fossorial life. The modern spadefoot toads 
differ only slightly from Macropelobates. The coccyx has become anky- 
losed to the sacrum (remaining free as a variation in some specimens of 
Pelobates).!_ No modern spadefoot toad attains the size of Macrope- 
lobates. There are also differences between these genera in proportion 
which seem to make the recent forms better diggers. 

The final modification in the evolution of the pelobatids was one 
which has occurred many times in the specialization of the Salientia. 
The maxillary teeth were lost. This change accompanied a dwarfing 
in size and a modification in structure. The tympanum was lost and the 
eustachian tubes became vestigial. The two genera Ophryophryne and 
Scutiger which exhibit this change may have evolved directly from Megalo- 
phrys, for their osteology is very similar in spite of their different out- 
ward appearance. They exhibit, however, the enormously expanded 
sacral diapophyses of Pelobates and similarly curved procoracoids (Bou- 
lenger, 1919; Procter, 1921). A determination of the exact relationships 
of these genera will have to await a more detailed study of their osteology 
and myology. 

It may be asked, when did the changes outlined above actually 
occur? The Liopelmide are found to-day only in the northwestern United 
States and in New Zealand. It is probable, therefore, that they existed 
in southeastern Asia during late Mesozoic or early Tertiary times. By 
the Oligocene the spadefoots had evolved and were living in Mongolia. 
Since then they have succeeded in migrating westward across favorable 
sandy areas of Asia to western Europe and eastward across the Bering 
Strait connection, southward to southern Mexico. The toothless forms 
are found to-day only at high altitudes in certain mountains of Sik- 
kim, Kashmir, Tonkin, Tibet and the Chinese province of Sze- 
chuan. They apparently arose from stocks existing in this region. The 
distribution of the other genera cannot be so easily accounted for. 
Pelodytes has been considered by Boulenger (1899) as close to Pelobates, 


1It is fused in 10 specimens of Scaphiopus holbrookii, 2 of S. couchii,1 of S. dugesi, 2 of S. hammondii, 
and 2 of Pelobates fuscus in the American Museum. 


1924] A NEW SPADEFOOT TOAD OF MONGOLIA if 


but it exhibits many unique features and it does not seem likely that this 
genus arose directly from Pelobates. Megalophrys, although primitive 
in many respects, has extended its range only from southern China and 
the Philippines to India. Although found in the Malay Archipelago and 
in Borneo, it apparently never succeeded in pushing south and east- 
ward into those East Indies which are not typically oriental in their 
faunal relations. No pelobatid occurs in New Guinea, as usually stated 
in textbooks (Nieden, 1923). Asterophrys is a brevicipitid as pointed out 
by Van Kampen (1923). I suggested (1922, p. 73) that Lechriodus (= Ba- 
trachopsis) was probably a toothed bufonid. Thanks to the kindness of 
Dr. P. N. Van Kampen, I have recently examined a specimen of L. 
melanopyga and may now confirm that opinion. 

In Figs. 5 to 7,1 have shown the pectoral musculature and thigh 
musculature of this specimen. The following features stamp Lechriodus 
as a bufonid in spite of its expanded sacral diapophyses: 

(1) Sartorius distinct from the semitendinosus. 

(2) Semitendinosus deep, not visible ventrally. 

(3) Tendon of semitendinosus piercing gracilis major and minor (as 

in various Australian bufonids). 

(4) Episternocleidohumeralis longus! distinct from supracoracoi- 

deus. 

(5) Supracoracoideus profundus distinct from supracoracoideus 

superficialis. 

None of these features is found in any pelobatid (Scaphiopus, Pelo- 
bates, Pelodytes and Megalophrys). 

Lechriodus also has a double condyle on the coceyx which is found 
among pelebatids only in the specialized Pelodytes. I have discussed 
(Noble, 1922, p. 11) the several exceptions to the form of the sacrum as 
being diagnostic of family relations. Lechriodus may now be definitely 
relegated to the Bufonide. 

If no pelobatids are found in the Papuan or Australian region, has 
the family arisen since the Indo-Australian connection was broken? 
This seems to be the best interpretation. If, therefore, we place the 
origin of the Pelobatide at the beginning of the Tertiary, we have still 
to account for the somewhat restricted range of the apparently primitive 
Megalophrys. Many factors, such as desert barriers or temperature, may 
have prevented their spread, but of these we have no definite knowledge. 
All evidence points toward an origin of the pelobatids in southern Asia 
during early Tertiary times, and to their differentiation, during the 
Oligocene or earlier, into spadefoots which have succeeded in spreading 


1] have followed the nomenclature of Anthony and Vallois, 1914, Bibliog. Anat., XXIV, p. 271. 


ee a einiereaineal 
Se 
eee 
— | 
ee SS See 


_———— a 
SSS" 
—S 
= 
ee 


y 


—————— 
—— 


\\ 


A. cl. h, Acromio cleidohumeralis; C. b., Coracobrachialis brevis; C. l., Coracobrachialis longus; 
C. p., Coracoradialis proprius; E. cl. h. 1., Episternocleidohumeralis longus; P. a., Pectoralis abdomi- 
nalis; P. s., Pectoralis sternalis; R. a., Rectus abdominis; S. p., Supracoracoideus profundus; S. s. 


Pig. Rec 
Supracoracoideus superficialis. 


Ch. Be Z i =. 
up Bs. 3G 


Gy 


R.a. 


A 
= 
Z| 

—Y 

= 


“SG 


SI 
SJ 
I 
LS 
J 
\ { 
| ‘ 
y) a as 
reviations 


f Megalophrys hasseltii (Tschudi). Abb 


Fig. 6. Pectoral m ] 
s in Fig. 5, ; 


Tens. fase. lat 


Pect. all iE 


Grac. may. 


Grac. min. 


>—Tond semitend, 


Crur. 


Add. mag. 


Sart. 


Tend. semitend. 


Semimbr: 


Semitend. 


Grac. min. 


Grac. May. 


B 
Fig. 7. Thigh muscles of Lechriodus melanopyga (Doria), ventral aspect. 


the superficial muscles; B, the distal muscle complex. 
Crur., Cruralis; Grae. maj., Gracilis major; Grac. min., Gracilis 
Semitend., Semitendinosus; 


A, 


Add. mag., Adductor, magnus; 1 
Pect.—Pectineus Sart., Sartorius; Semimbr., Semimembranosus; 


minor; 
Tens. fase. lat., Tensor fascie late; Tend. semitend., Tendon of semitendinosus. 
14 


1924] A NEW SPADEFOOT TOAD OF MONGOLIA 15 


westward into Europe and eastward to Mexico and eastern United States. 
Finally, part of the primitive stock residing in the mountains of southern 
Asia was modified by a loss of their dentition. 


LITERATURE CITED 


Bouxay, Sr. J. 1919. ‘Elements of the Comparative Osteology of the Tailless 
Batrachians.”’ Glas. zemal. muz. u Bosni i Hercegovini, XXXI, pp. 277- 
358. - 
1922. “Answer upon Dr. Baron G. J. von Fejérvdry’s Kritische Bemerkungen 
zur Osteologie, Phylogenie und Systematik der Anuren.’”’ Glas. zemal. 
muz. u Bosnii Hercegovini, XX XIII (for 1921), pp. 197-206. 
Boutencer, G. A. 1899. ‘On the American Spade-foot (Scaphiopus solitarius 
Holbrook).”” Proce. Zool. Soc. London, pp. 790-793, Pl. x11. 
1919. “On Alurophryne mammata, Gthr., an addition to the Batrachian fauna 
of Kashmir.” Rec. Indian Museum, XVI, pp. 469-470. 
Fresérviry. G. J. pp. 1917. ‘“Anoures fossiles des couches préglaciaires de Piis- 
pokfiirds en Hongrie en considération spéciale du développement phylé- 
tique du sacrum chez les anoures.” Féldtani Kézlény, XLVII, pp. 5-38, 
Pls. 1-111. 
1921. “Kritische Bemerkungen zur Osteologie, Phylogenie und Systematik der 
Anuren.” Arch. f. Naturg., LXX XVII, Abth. A, pp. 1-80. 
1923. ‘‘Ascaphide, a New Family of the Tailless Batrachians.” Ann. Mus. 
Nat. Hung., XX, pp. 178-181. 
Kampen, P. N. van. 1923. ‘The Amphibia of the Indo-Australian Archipelago.” 
. Leiden. 
Niepen, F. 1923. ‘Amphibia, Anura I, Subordo Aglossa und Phaneroglossa,” 
Sectio 1, Arcifera. Das Tierreich (Preus. Akad. Wiss. Berlin). 
Nostz, G. K. 1922. “The Phylogeny of the Salientia I The Osteology and the 
Thigh “Musculature; their Bearing on Classification and Phylogeny.” 
Bull. Amer. Mus. Nat. Hist., XLVI, pp. 1-87, Pls. 1-xxm1. 
Procrer, Joan B. 1921. “Ona new Toad, Cophophryne alticola, collected on the 
Mt. Everest Reconnaissance Expedition, 1921.” Ann. Mag. Nat. Hist., 
(9) IX, pp. 583-587. 


Cree a Pele hbers: 
pa? ne toe 


~ et bl 
c rd 
f 


rh ys 
£ +. 
Silay t 
uy wr 


Ls 


AMERICAN MUSEUM NOVITATES 


Mirber 138 Tue A Goede or ane Sent, 30, 1924 
a HE AMERICAN USEUM OF ATURAL ISTORY e} : 
umber | New York City Pp “ 


59.9,32C(51) 


MICROTINES COLLECTED BY THE ASIATIC EXPEDITIONS! 
By Guover M. ALLEN 


The field work carried on by the Second and Third Asiatic Expedi- 
tions of The American Museum of Natural History under the able 
leadership of Mr. Roy Chapman Andrews has resulted in the accumula- 
tion of a splendid series of voles and other microtines including a 
large number from the highlands of western Yunnan (a locality that has 
lately yielded many novelties), series from near Peking, from localities 
in Shensi Province, and from central Mongolia. In working out the 
identifications of these specimens, one is impressed by the number of 
microtine types occurring in eastern Asia. These are, many of them, 
superficially much alike, although their dental characters at once 
mark them off into groups which are of greater or less superspecific 
value. These have usually been considered subgenera, but, partly for 
convenience, have by some authors been accorded full generic standing. 
The occurrence of various intermediate steps in details of tooth-pattern, 
together with their general external resemblances, leads me to prefer for 
the present the more conservative course of regarding most of these 
groups as subgenera, pending some future and comprehensive review. 

Among the many interesting facts of distribution brought out by 
these collections, a few may be particularly noted. The moist uplands 
of Szechuan have a characteristic series of species whose range extends 
southward into the high country of Yunnan, where local forms may de- 
velop. The eastward extension of many types common in western Asia 
is brought out by the occurrence in the Mongolian high plateau of 
species of EHllobius and Lagurus, and representatives of Muicrotus 
obscurus, and a form of the subgenus Alticola. The great tongue of 
high desert extending across Mongolia forms an effective barrier to the 
further southward progress of certain species common farther north 
where there is a certain amount of sheltering forest—such, for example, 
as Evotomys and Myopus. 

At the suggestion of Mr. Andrews, the complete list of species ob- 
tained is here given, together with a brief diagnosis as a help to those 


.. }Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 28. 


2 - AMERICAN MUSEUM NOVITATES [No. 133 


who may have opportunity to do field work in the same country. High 
credit is due him for the abundance and excellent quality of the material 
he has brought back from this little-known area. 


Myopus saianicus Hinton 


A short-tailed lemming with a dull-reddish back and slaty head, 
sides and belly. Specimens of this genus have been very rare in collec- 
tions. Middendorff reported it nearly seventy-five years ago from the 
Okhotsk Sea, but his record was long believed to be erroneous since it 
had previously been known from northern Europe only. In 1912 Hol- 
lister described M. morulus on the basis of a single specimen from the 
Altai Mountains; while another taken in the Syansk Mountains, one 
hundred miles west of Lake Baikal, has lately been made the basis of M. 
satanicus by Hinton. To the latter I have referred the fine series of four- 
teen taken at two different localities, fifteen miles north and forty-five 
miles northeast, respectively, of Urga, and four others from Sain Noin 
Khan. It seems likely that the Altai and the Syansk forms must be very 
closely related, and both perhaps hardly more than subspecies of M. 
schisticolor of Europe. The present series from the forest area bordering 
the northern Gobi establishes the occurrence of the genus in Mongolia 
and perhaps marks its southern limit of distribution in that country. 


Evotomys rutilus russatus (Radde) 


This red-backed mouse was*abundant in the wooded country forty- 
five miles northeast of Urga, Mongolia, where a large series was obtained. 
These differ from a series from the Altai referred to rutzlus in their slightly 
brighter colors and in having the tails usually more or less reddish like 
the back instead of blackish as in these latter. Although no specimens of 
rutilus from the Obi region are available as topotypes, it is assumed that 
the Altai specimens are the same, and I have therefore applied Radde’s 
name to the Mongolian series. His figures and description (1862, ‘Reise 
Siberia,’ I, p. 186, Pl. vu, figs. 2, 2 a-e) seem to indicate this animal 
clearly enough, even to the reddish tail, in spite of doubts that have been 
expressed on account of his rough outlines of the enamel pattern of the 
teeth. 


Evotomys rufocanus (Sundevall) 


In addition to its cranial characters, this species can be distinguished 
from the preceding by its duller red back and the slaty muzzle, cheeks, 


1924] MICROTINES FROM THE ASIATIC EXPEDITIONS 3 


and sides. A good series was secured along the southern border of the 
forest at stations fifteen miles north and forty-five miles northeast of 
Urga, Mongolia, as well as at Sain Noin Khan to the westward. It seems 
to occur, therefore, in the same general localities as the smaller red-backed 
mouse. I have carefully compared this series with topotypes of rufocanus 
from Sweden and can find no tangible differences. Apparently, however, 
the species becomes more richly colored with more ochraceous flanks in 
Korea, a form to which the name regulus has been given. This should 
evidently be considered a subspecies, Evotomys rufocanus regulus. Mr. 
Andrews obtained a small series of this form in Korea during a previous 
expedition. 


Evotomys rufocanus shanseius (Thomas) 


In this race the back is less rufous, the sides more ochraceous, with a 
resulting paler, yellowish-brown appearance. It was originally described 
as a full species, Craseomys shanseius, type from the spruce forest, one 
hundred miles northwest of Tai-yuan-fu, Shansi, but the series now at 
hand including ten from Kwei-hwa-ting, Shansi, and five from one 
hundred miles northeast of Peking, indicates that its range passes into 
that of rufocanus on the eastward, although separated from it on the 
north by the Gobi plateau and Ordos Desert in Mongolia. Indeed, 
Thomas has referred a single specimen from sixty miles east of Peking to 
regulus, but the additional skins from that region secured by the Asiatic 
Expeditions of The American Museum of Natural History are quite 
indistinguishable from the pale form, shanseius. These two forms should 
be regarded as geographic subspecies of EL. rufocanus. 


Microtus (Eothenomys) melanogaster confinii (Hinton) 


Externally this is a small dark vole with blackish slaty belly; 
many of the long hairs on sides and belly have burnished tips. The type 
locality is the Kiuchiang-Salween divide in latitude 28° north, altitude 
11,000 feet. This is the southern representative of melanogaster of Sze- 
chuan and occurs on the mountains of southern Yunnan at altitudes from 
6000 feet (Salween drainage) to over 10,000 feet (Pei-tai Mountain, 
south of Chung-tien). A series secured by the Second Asiatic Expedition 
includes two which, as noted by the collector, contained on February 8, 
1917, one and two embryos respectively, a number correlated no doubt 
with the reduced number of mamme. 


4 AMERICAN MUSEUM NOVITATES INo. 133 


Microtus (Eothenomys) fidelis (Hinton) 


The largest, known member of the subgenus, a very reddish-looking 
vole with large skull and hind foot. The type came from the Li-chiang 
range, Yunnan, in latitude 27° 30’ north. A fine series was secured from 
the same area by the Second Asiatic Expedition, as well as others from 
Chunglu, Siao-ke-la, Chiangwei and Yangtsien, on the Mekong River 
between altitudes of 6000 and 9000 feet; Ha-pa (north of Taku), 10,000 
feet; Yangpi River (Tengyueh road) at 5000 feet; Tali Lake, 6500 feet; 
and other localities. -The entire series is very uniform in its characters. 
In two cases, three embryos were found in specimens captured the first 
week in October, and two in a female taken October 30, 1916. These are 
not only late litters but, as in the case of the preceding species, indicate a 
correlation between the small number of young per litter and the reduced 
number (four) of mamme. It is possible also that the breeding season 
extends over a longer period than in some other microtines. 


Microtus (Eothenomys) proditor (Hinton) 


Externally like M. (E.) fidelis but the tail and hind foot smaller; 
the tooth pattern is slightly more complicated (especially m*) and some- 
what transitional between that of fidelis and the species of the subgenus 
Anteliomys. The type came from the Li-chiang range, Yunnan, at 
13,000 feet altitude, and the series of ten specimens brought back by the 
Second Asiatic Expedition is from the same range, at the following locali- 
ties: Ssu-shan-chong, 9000. feet; Peswi, 10,000 feet; Ssu-shan (Snow 
Mountain), 12,000 feet; timber-line, 13,000 feet. Evidently it is a species 
of high levels. 


Microtus (Eothenomys) olitor Thomas 


A small dark grayish-brown species with a very small skull. Three 
specimens were secured at an elevation of 7000 feet on the Mucheng- 
Salween drainage, western Yunnan. The type locality is Chao-tung-fu, 
Yunnan. This species is apparently the least common of the members of 
the subgenus in Yunnan. Its small size is at once distinctive, while the 
tooth pattern is interesting in its resemblance to that of Anteliomys in the 
increasing complexity and form of the third upper molar and the elimina- 
tion of the postero-internal angle of the first upper molar. 


1924] MICROTINES FROM THE ASIATIC EXPEDITIONS 5 


Microtus (Anteliomys) custos Thomas 


This species bears much resemblance to M. (E.) fidelis and proditor 
in its general appearance and reddish-brown coloration, but the belly 
is usually much grayer, and the tooth characters are distinctive. The 
posterior edge of the palate usually shows two slight projections, one on 
either side of the mid-line, which in M. (A.) chinensis meet to form a 
short median spine. It is an alpine species and was described from 
specimens taken at A-tun-tsu, northwestern Yunnan, at over 11,000 
feet altitude. The Second Asiatic Expedition secured a series at To-mu- 
lang, 10,000 feet, and at the same altitude at Tu-gan-sha, both localities 
near Chung-tien. Others were taken at Ying-pan-kai and La-chu-mi, 
9000 feet. One specimen (No. 44141) has an incipient postero-internal! 
cusp on the two anterior upper molars, reminiscent of their fuller devel- 
opment In certain species of the subgenus Hothenomys. On the Li-chiang 
range this species is represented by an allied form, rather sharply marked 
off by the peculiarities of the nasal and intermaxillary bones, and may be 
described as follows. 


Microtus (Anteliomys) custos rubellus, new subspecies 


Typr.—Adult female, skin and skull, No. 44001, American Museum of Natural 
History, from Ssu-shan (Snow Mountain), Li-chiang range, at timber line, 13,000 
feet. October 13, 1916. R.C. Andrews and Edmund Heller. 

DescrIpTion.—Similar in general to M. custos but slightly more reddish above, 
the belly clearer gray, lacking the decided brownish wash. The skull differs notably 
in that the nasals are exceeded by the premaxillaries in their backward extension. 

General coloration above, a dull rusty brown, nearly uniform, though slightly 
clearer along the flanks. The individual hairs are slaty at base, with an ochraceous 
terminal portion that becomes more rufous near the tip which is black. At the sides 
this minute black tip is often lacking, producing the clearer effect. Entirely black 
hairs are scattered throughout the pelage. Entire lower surface of the body gray or 
bluish gray, sometimes with a very faint wash of brownish on the chest. Feet and tail 
dusky above, the latter paler (grayish) below. 

The skull appears to average a little larger than in typical M. custos. The most 
noticeable difference is in the relations of the nasals and the premaxille. In M. 
custos the posterior extension of the nasals exceeds that of the premaxille, whereas in 
M. c. rubellus they are uniformly shorter and do not (or rarely) extend even as far 
back as the terminal border of the premaxille. The palate, instead of ending in a 
median spine (as in M. chinensis), has a slightly projecting edge with usually two 
small blunt points. The teeth are essentially as in M. custos. 

MrASUREMENTS.—The collector’s measurements of the type are: head and body, 
100 mm.; tail, 41; hind foot, 18.5. Four other adults (topotypes) measured as 


6 AMERICAN MUSEUM NOVITATES [No. 133 


follows: head and body, 100, 100, 102, 105; tail, 44, 39, 40, 32; hind foot, 18.5, 18, 
19, 17. The average of five adults of typical M. custos from the Chung-tien district 
is: head and body, 93.6; tail, 44; hind foot, 17.9. 

The skull of the type measures: greatest length, 2%.5 mm.; condylobasal length, 
24.4; palatal length, 13.1; interorbital width, 3.7; zygomatic width, 14.2; mastoid 
width, 11.0; occipital depth including bulla, 8.0; upper molars (alveoli), 6.1; lower 
molars (alveoli), 5.9; tip of mandible to condyle, 15.2. These measurements are a’ 
very little larger than the average of adult M. custos. 

This representative of M. custos (type from A-tun-tsu, 11,500 to 
12,500 feet), like many other of the small mammals from the snow peak 
of the Li-chiang range, is sufficiently different to be worth recognizing as a 
distinct local race. It is much paler underneath, practically lacking the 
brownish wash so conspicuous in M. custos, and the body is a very little 
redder. In series, M. custos shows a slight contrast between the dull 
reddish tint of the head and the brownish of the back, whereas in M. c. 
rubellus there is less contrast, and the general tone is redder. The differ- 
ence in the relations of the nasals and the premaxille is striking and fairly 
uniform in a large series. Apparently the larger species, M. (Anteliomys) 
chinensis, is not represented here, but reaches its southern limit on the 
divides northwest of A-tun-tsu, whence, under the name wardi, a form 
has been described by Thomas and again recorded by Hinton. 


Microtus (Caryomys) aquilus (G. M. Allen) 


A series of fifteen skins of this dark-brown, long-tailed vole from 
Tai-pei-shan, Tsing-ling range, southern Shensi, 10,000 feet, is apparently 
the same as the species I described in 1912 as Craseomys aquilus, which 
proves to be one of the Caryomys group with closed triangles in the first 
lower molar. This in turn may be found inseparable from M. (C.) 
alcinous Thomas, from Wei-chow, western Szechuan, but in the absence 
of typical specimens for comparison may at present stand. Probably 
M. (C.) eva (from Kansu) is one of the same group, and the others should 
.be treated as subspecies of it. : 


Microtus (Alticola) worthingtoni semicanus, new subspecies 


Typr.—Adult male, skin and skull, No. 57805, American Museum of Natural 
History, from Sain Noin Khan, Mongolia. June 5, 1922. Third Asiatic Expedition. 

DeEscrIPTION.—Similar to worthingtoni but with larger skull, longer tooth rows 
and with a buff lateral line and buff wash over the lower surfaces. 

General color above a buffy gray slightly darkened with scattered black hairs. 
The individual hairs are slaty at base with a broad subterminal band of white which 


1924] MICROTINES FROM THE ASIATIC EXPEDITIONS 7 


passes into pale buff for a very short distance below the minute blackish tip. Sides 
of nose, front and back of ears nearly ‘pinkish buff.’’? Entire under parts, including 
the upper lips, lower cheeks, sides of neck, as well as the feet, legs and tail all around, 
buffy white, the hairs on legs and body with dark bases. The mixed gray of the back 
is sharply defined at the sides of head and body and the buff is here clearer and 
brighter, forming an indistinct lateral line (about ‘‘pinkish buff’’). Vibrissee very 
long, the longest about 50 mm., some black, some white. 

The skull is larger fanancnout with longer tooth rows than in typical Maite 
toni, with which it shares the well-defined first outer re-entrant angle of the last upper 
Golan The first lower molar, in addition to the posterior transverse loop, has two 
completely closed triangles on each side and an anterior trefoil, the inner loop of 
which is slightly larger than the outer. 

MEASUREMENTS.—The collector’s measurements of the type are: head and body, 
110mm.; tail, 30; foot, 19; ear, 20. Theskull measures: greatest length, 29.1 mm.; 
condylobasal length, 28.7; palatal length, 15.0; zygomatic width, 16.7; mastoid 
width, 14.4; diastema, 9.0; tip of mandible to condyle, 18.0; upper cheek teeth 
(alveoli), 6.7; lower cheek teeth (alveoli), 6.5. 

This is undoubtedly a close relative of Alticola worthingtoni of the 
Tian-shan, nearly a thousand miles to the westward. In color, the two 
are closely similar except that the Mongolian animal is distinctly buffy 
instead of clear white below, and the ears and sides of the flanks are 
even clearer buff. The much larger size of the skull and teeth are further 
distinctive, and distinguish it from the subspecies subluteus of Yarkand 
as well. From albicauda, the species is apparently quite distinct, as 
pointed out by Miller, while the Altai species, strelzovz, is smaller and 
grayer with a much more flattened skull. 

The fine series of this large gray microtine extends the known range 
of the subgenus northeastward into central Mongolia, which may be 
near the limit of its distribution in this direction. In addition to some 
twenty-seven specimens from the type locality, others were secured at 
Hurum-tu, 7000 feet, Gun Burte, 6800-feet, and forty miles south of 
Tzetzenwan. 


Microtus (Lasiopodomys) brandti (Radde) 


A pale sandy-colored vole with long claws on both fore and hind feet. 
A large series was secured by the expeditions of 1919 and 1922 in the 
northern part of Mongolia at distances of from eighty to one hundred and 
forty miles southeast of Urga and at Hurum-tu (7000 feet), Tzetzenwan 
and vicinity, Sain Noin Khan and on the Ongin River. Radde’s type 
series came from localities on the Mongolian high steppe and near Tarei 
Nor, so that these may be considered typical. So far as may be judged 
from the description, Microtus warringtoni Miller from Tabool, Mon- 
golia, is quite the same. The latter was supposed to be larger than 


8 AMERICAN MUSEUM NOVITATES [No. 133 


typical brandti, but the comparisons were made with Bichner’s figures 
and description of specimens obtained by Prjevalski in northeastern 
Tibet. Many of the specimens from southeast of Urga exceed M. 
warringtoni in size of skull. Possibly there is a decrease in size westward. 
A brood of four very young mice was found June 24 near Urga. 


Microtus mandarinus (A. Milne-Edwards) 


A buffy-brown short-tailed vole with unusually long silky fur which 
almost hides the small round ears. The original specimen was secured by 
Pére David in “la Mongolie chinoise avec les espéces précédentes,”’ 
that is, in Shansi, probably near Saratsi, whence came several other 
species described from the same collections. A series of eleven specimens 
collected near Kwei-hwa-ting in the same province, are therefore practi- 
cally topotypes, as the localities are hardly fifteen miles apart. These 
specimens agree well with Milne-Edwards’s description and figure, even 
to the white of the throat. The whitish belly is usually slightly tinged 
with buffy. The rediscovery of this species in northern Shansi is note- 
worthy in connection with Thomas’s suggestion that the original speci- 
mens may have come from southern Shensi. The Kwei-hwa-ting series 
is apparently somewhat larger but otherwise very similar to M. johannes 
(from IKo-lan-chow, Shansi), which is at most but a subspecies of 
mandarinus. In addition to this lot, a second series of fourteen was 
secured in Chili Province, one hundred miles northeast of Peking, which 
differ in the generally darker color, especially of the under side as com- 
pared with those of the drier interior. They seem to be a well-marked 
subspecies, and may be described as follows. 


Microtus mandarinus feceus, new subspecies 


Typr.—Adult male, skin skull, No. 56358, American Museum of Natural 
History, from Chili Province, 100 miles northeast of Peking, China. March, 1922. 
Third Asiatic Expedition. 

DescripTion.—Similar to M. mandarinus but darker above, with a clear slaty . 
throat, and slaty belly heavily washed with buffy. 

General color above a nearly uniform ‘‘hair brown’’ rather than the brighter 
“wood brown” of typical mandarinus, aresult of the greater admixture of black hairs 
and of the narrower subterminal ochraceous rings of the parti-colored hairs. The 
general effect is of a much darker, less buffy pelage. The flanks are a slightly clearer 
buff, deeper in tone than in the typical form, “‘ warm buff”’ instead of ‘‘light ochrace- 
ous buff.”’ The ventral surface is very different in the two. In typical mandarinus 
it is grayish or whitish throughout, with a light wash of ‘pale ochraceous buff”’ over 
the chest and belly. In fxceus, the whitish is lacking so that the slaty bases of the 
hairs are not concealed; the chin and throat are therefore dark slaty gray, and the 
chest and belly similar, strongly washed with ‘warm buff,” which, being confined to 


1924] MICROTINES FROM THE ASIATIC EXPEDITIONS 9 


the extreme tips of the hairs, does not conceal their dark bases. Feet covered with 
short dusky and silvery hairs, the former more in evidence on the basal portion and 
metapodial area. Tail bicolor, its upper surface like the back, lower surface buffy. 
Ears small, concealed in the fur. Feet averaging slightly larger and longer than in 
the typical form, 19 to 21 mm., against 17 or 18 in the latter. 

No field measurements or skulls accompany the Chili series. 

This darker form of mandarinus is readily distinguishable on 
account of its lack of whitish-tipped hairs below, resulting in a contrast- 
ing dark slaty throat and a slaty belly washed heavily with buffy, 
whereas the typical form is gray-bellied, with or without a light buffy 
tint over the chest and abdomen, and is less dark above. 


Microtus (Neodon) irene Thomas 


Externally a rather grayish-brown vole, with white hind feet, a 
bicolor tail, and blue-gray belly. Some individuals have a slight tinge of 
buff below. The first lower molar has but four closed triangles. A small 
series was secured in northern Yunnan near Pei-tai, a locality some thirty 
miles south of Chung-tien. Here, at 13,000 feet, about the summit of 
Ho-shan, it was evidently common, and perhaps at its southern limit, 
for none was obtained in spite of careful trapping on the Li-chiang range 
to the southward. Its chief distribution, so far as known, is in Szechuan 
Province, whence came the original specimens (at Ta-tsien-lu). It now 
appears, also, that the series secured by Zappey at Ramala Pass (16,000 
feet) and Shuowlow (13-—15,000 feet) on the Tibetan border of Szechuan 
are this species and not M. mandarinus as I recorded in 1912. 


Microtus (Stenocranius) angustus Thomas 


A pale, almost buffy vole, with rather small ears and short tail; 
the skull is of the long, narrow type (Stenocranius). This is evidently an 
abundant species of the grass-lands over much of the Mongolian plateau. 
It was described from specimens obtained by Anderson at its extreme 
southeastern border, about one hundred miles northwest of Kalgan, 
Chili Province, after his expedition had topped the southern escarpment 
and descended slightly to the tableland. The Asiatic Expeditions under 
Mr. Roy C. Andrews, continuing on the same caravan route, secured 
a large series on the way to Urga. The localities are: 140 miles southeast 
of Urga; 120 miles, 60 miles, and 40 miles southeast of the same place in 
successive stages of the journey, and again some 45 miles northeast of 
that center. To the westward, it was found in numbers at Sain Noin 
Khan and a single one was secured at the Ongin River. Partly grown 
young were caught during July. 


10 AMERICAN MUSEUM NOVITATES [No. 133 


Microtus poljakowi Kastchenko 


A very small dark vole, with small hind foot (16 mm.), short bicolor 
tail, and whitish belly. The even admixture of black and buffy-tipped 
hairs above gives a uniform ‘pepper-and-salt”’ effect. The skull is 
barely 23 mm. in greatest length. This species is based on the Arvicola 
gregalis of Radde who found it in the Apple Mountains, Dauria, 
Baikal region. The Asiatic Expeditions obtained it near Urga, at stations 
15 and 45 miles northeast and 40 miles south of that center. Elsewhere 
on the Mongolian plateau it was not met with. 


Microtus mongolicus (Radde) 


A medium-sized species with coarse dark brownish fur, and a general 
ochraceous or buffy tint that appears especially on the lower edge of the 
cheeks, the end of the muzzle and faintly over the belly. This was 
described by Radde ‘‘aus den daurischen Hochsteppen, von dem Um- 
gegenden des Tarei-nor,”’ and is evidently a common and characteristic 
vole of the northern part of Mongolia. A large series was secured at 
localities 35 and 45 miles northeast, and 15 miles north of Urga, but it 
was not found to the westward. 


Microtus obscurus (EHversmann) 


A species closely resembling M. mongolicus but distinguished by the 
blue-gray belly, pale instead of dark-brown feet, and by the absence of 
the buffy tint of the under surface and the patch about the vibrissz and 
cheeks, which are blue-gray instead. It apparently occurs with mongolicus 
in certain localities, possibly, however, with some difference in habits or 
choice of situation. A large series was secured at points 15 miles north 
and 45 miles northeast of Urga as well as at Sain Noin Khan, to the west- 
ward. These have been compared with a series from the Altai Moun- 
tains, the type locality, but no tangible differences were made out, 
though the skulls seem a little larger in the Mongolian specimens. It is 
apparently another of the Altai species that has extended eastward into 
northern Mongolia. 


Microtus calamorum superus Thomas 
A large, long-tailed species with the dark coloring of the common 
field-vole. Four specimens from 45 miles south of Fengsiangfu are 
practically topotypes of this slightly smaller race of the Yangtze Valley 
meadow-mouse. As suggested by Thomas, it resembles Arvicola in 
having but five instead of six plantar tubercles, and in the long tail 


1924] MICROTINES FROM THE ASIATIC EXPEDITIONS 11 


(slightly more than half the length of head and body). It lacks the 
quality of fur found in the water-voles of Europe, however, and so more 
nearly resembles the American species of this subgenus. 


Lagurus przewalskii (Biichner) 


A pale yellowish mouse with very small ears, minute tail, and well- 
clawed feet with hairy soles. This microtine has become so modified 
both exteriorly and in the shape of the skull with its enlarged auditory 
region, that it seems deserving of full generic standing as Thomas has 
advocated. Its adaptations seem to fit it for a life in sandy deserts. The 
Third Asiatic Expedition secured specimens at Tsagan Nor, Loh, Ussuk, 
and Artsa Bogdo, on the Mongolian Plateau. These, although from 
stations nearly 800 miles from the type locality (Gass, Zaidam, Tibet), 
seem to agree closely with Biichner’s description and plate, thereby serv- 
ing to extend considerably the known range of the species. 


Ellobius larvatus, new species 


Type.—Adult male, skin and skull, No. 57886, American Museum of Natural 
History, from Artsa Bogdo, Sain Noin, Mongolia, altitude 6500 feet. August 21, 
1922. Third Asiatic Expedition. 

DescripTion.—Face from the upper lips to a line half-way between the eye and 
the ear, and on the forehead to a transverse line at the same level, contrastingly dark, 
“clove brown,’ passing rather abruptly into the color of the back, which from the 
lower half of the cheeks and the occiput to the root of the tail is clear ‘‘ cinnamon buff,”’, 
through which the slaty bases of the hairs appear faintly. Feet thinly covered with 
short white hairs which form a stiff fringe at the sides. The entire under surface, the 
limbs and flanks are dull gray, the individual hairs with short white tips and conspicu- 
ous slaty bases (near ‘blackish mouse gray’’). The very short tail is clothed with 
cinnamon-buff and clove-brown hairs, with a very small terminal tuft of white. A 
few whitish hairs are also present about the ears. 

The skull is smaller than in E. tancrei and E. albicatus of the Altai and Hami 
Mountains respectively, which it resembles in its narrow strap-shaped interparietal; 
but it differs in having the supraoccipital ridge bowed forward instead of directly 
transverse along the posterior border of the interparietal. The upper incisors are 
strongly inclined forward. The crown outline of the posteriormost upper molar is 
like that of a figure 8 with the anterior and posterior edges flattened so as to be nearly 
transverse, while each side is strongly concave. 

M®EASUREMENTS.—The collector’s measurements of the type and three other 
specimens from the same locality are: 


No. Head and Body Tail Hind Foot Sex 
57881 110 mm. 13 23 Q 
57883 110 10 24 g 
57885 JOB. 47 12 24 9 
57886 (type) 105 11 23 a 


12 AMERICAN MUSEUM NOVITATES [No. 133 


The skull of the type measures: greatest length, 32 mm.; condylobasal length, 
29.8; palatal length, 8.0; diastema, 11.8; zygomatic breadth, 21; mastoid breadth, 
14; interorbital breadth, 5.6; upper tooth row (alveoli), 8.0; tip of mandible to 
condyle, 22; lower tooth row (alveoli), 7.5; interparietal, 6.6 2.5. 


A series of seven adults from the type locality is fairly uniform in its 
pale cinnamon-buff color above with dark facial mask and white-tipped 
belly-hairs. A young one taken August 22 is a uniform gray all over with 
darker muzzle. 

This is a smaller, paler species than E. tancrei of the Altai, or EZ. 
albicatus of Chinese Turkestan, some 600 miles to the westward. From 
the latter it differs further in its shorter tail and in having the supra- 
occipital ridge bowed forward at the center. H. caenosus of the Tian-shan 
is a larger, darker species with squared, instead of strap-shaped, inter- 
parietal, and with a buffy belly. The relationships of this new species 
may prove to be with E. fusciceps and its dark subspecies ursulus of 
northwestern Dzungaria, which it resembles in size. The discovery of 
this species considerably extends the known range of the group into the 
Mongolian region. 


Ellobius orientalis, new species 


Typr.—Adult male, skin and skull, No. 57893, American Museum of Natural 
History, from Iren Dabasu, eastern Mongolia. April 25, 1922. Third Asiatic 
Expedition. 

DescripTION.—Smallest of the known species of the genus, with bright cinna-. 
mon back, white belly, and reduced dark facial area. 

The dark face-patch is much less extensive than in the previous species, and 
covers the forehead from the muzzle to or including the eye, at the level of which it 
merges with the general dorsal coloration. The face-patch itself is nearly “fuscous”’ 
(Ridgway, 1912). The rest of the upper surface, from the eyes to and including the 
tail, as well as the sides of the face from the muzzle backward as high as the level of 
the eye, is uniform clear “pinkish cinnamon’’; this color, however, hardly reaches 
the flanks, which with the entire lower surfaces are dull white, with the dark slaty 
bases of the hairs showing through everywhere. The feet are sparsely clothed with 
short whitish or silvery hair. The character of the pelage is quite different from that 
of E. larvatus, not loose and fluffy showing the dark bases above, but shorter and 
_closer, helping to produce a more uniform clear coloration. There is a minute tuft. 
of whitish hair springing from the front of each ear and one at the extreme tip of the 
tail. The latter has a few coarse dark hairs interspersed with the cinnamon. 

The skull compared with that of larvatus is much smaller but of the same general 
shape, with the white upper incisors strongly thrown forward. The interparietal, 
however, is less narrow though not square, the audital bulle are obviously more flat- 
tened, and the mesopterygoid fossa is relatively much shorter. The squamosal and 
the zygomatic process of the maxillary in E. larvatus do not quite meet, but a small 
portion of the jugal intervenes, whereas in E. orientalis the jugal lies wholly dorsal to 
these two bones which meet below its posterior end. 


19241 MICROTINES FROM THE ASIATIC EXPEDITIONS 13 


The last upper molar instead of being shaped like a flattened figure 8 is nearly a 
triangle with concave sides and blunted angles, the inner anterior of which is produced 
conspicuously inward, obviously exceeding the outer anterior angle in length. 

M®EASUREMENTS.—The collector’s measurements of the type and three other 
specimens are respectively: head and body, 108, 100, 95, 95 mm.; tail, 10, 9, 9, 8; 
foot, 21, 20, 19, 19. 

The skull of the type measures: greatest length, 28.5 mm.; condylobasal length, 
26.5; palatal length, 16; diastema, 10; zygomatic breadth, 19; mastoid breadth, 
13; interorbital breadth, 5.2; upper tooth row (alveoli), 6.4; tip of mandible to con- 
dyle, 20; lower tooth row (alveoli), 6.8; interparietal, 6.5 x4. 

This appears to be a perfectly distinct species from EF. larvatus, and, 
occurring at a locality some 500 miles farther to the eastward, not only 
extends the known range of the genus to the eastern part of the Mongo- 
lian plateau but constitutes its easternmost record. Compared with the 
latter species it is not only much smaller and brighter-colored, with 
noticeably shorter and weaker feet, but the last upper molar is much re- 
duced and the jugal is excluded from the lower margin of the zygoma by 
the meeting of the maxillary process with the squamosal. 


‘ 73 F ieseteiaodaes > 
tants Si ee snl a 


. a ves . 


Eee { 


So otal a 
a’ bak, *~ 
‘ sah | t y “ ‘ ql 
* ‘ ‘ 
a ¥ — 
‘ 
. 7 i 
. ‘ 
5 ; 
‘ 
‘ 
. » 
5 er, 
‘ 
’ 
. 
‘ 
‘ 
' ' 
’ ‘ 
‘ , mf 
. 
tel ve 
. 
£ é 
. \ 
» f 
. 
~ . 
- 
re 
. - 
ave 
-: 
ae 
i 
< * 
* 


Basin Structures in Mongolia 


By CHARLES P. BERKEY AND FREDERICK K. Morris 


‘BULLETIN 
OF 
THE AMERICAN MUSEUM OF NATURAL HISTORY 
Vou. LI, Arr. V, pp. 103-127 
New York 
Issued October 7, 1924. 


BULLETIN, NOVITATES, = MEMOIRS _ 
| OF” iets 
The Pecntate ee Museum of Natural History. 
FRANK E. Lurz, Editor 


The following are the more recent papers on aha pi PaLa- 


~ontotocy. Orders should be addressed, Library, The American 


Museum of Natural aaa 77th St. and Central Park West, New, an 


York ae 


VERTEBRATE PALAONTOLOGY 


‘Reptilian and Stegocephalian Remains from the Triassic of Pennsylvania 3 


in the Cope Collection. By F. von Huene, 1921, Bulletin, XLIV, Art. 
19, pp. 561-574, 20 text figures. 


_ New Fossil Rodent from Ecuador. By H. E. Anthony, 1922, Novitates, / ; 


No. 35, pp. 1-4. - 


Hesperopithecus, the First Anthropoid Primate Found in America. By 


Henry Fairfield Osborn,.1922, Novitates, No. 37, pp. 1-5. 

Bird Remains from the Caves of Porto Rico. By Alexander Wetmore, 
1922, Bulletin, XLVI, Art. 4, pp. 297-333, 25 text figures. 

The Family Deinodontide, with Notice of a New Genus from the Cretace- 
ous of Alberta.’ By W. D. Matthew and Barnum Brown, 1922, 
Bulletin, XLVI, rer 6, pp. 367-385, 1 text figure. 

Species of American Pleistocene Mammoths. Elephas jeffersonii, New 
Species. By Henry Fairfield Osborn, 1922, Novitates, No. 41, pp. 1-16, 
12 text figures. 

The Locomotor Apparatus of Certain Peiniitive and Mammal-like Reptiles. 
By Alfred 8S. Romer, 1922, Bulletin, XLVI, Art. 10, pp. 517-606, Pls. 

XXVII-XLVI, 7 text figures. 

Dibelodon edensis (Frick) of Southern California, Miomastodon of. the Middle 


Miocene, New Genus.. By Henry. Fairfield Osborn, 1922, Novitates, © 


No, 49, pp. 1-4, 1 text, figure. 

Revision of Palzomastodon and M eritherium. Palzomastodon intermedius, 
and Phiomia osborni, new species. _By H. Matsumoto, 1922, Novi- 
tates, No. 51, pp. 1-6, 3 text figures. 

Notes on. the Type of Hesperopithecus haroldcookii Osborn. By W. K. 


Gregory and M. Hellman, 1923, Novitates, No. 53, pp. 1-16, 6 text’ 


figures. 

‘The Sapnased Plumage of the Eocene Bird Diatryma. By T. D. A. Cock- 
erell, 1923, Novitates, No. 62, pp. 1-4, 1 text figure. 

Protoceratops andrewsi, a Pre-ceratopsian Dinosaur from Mongolia. By W. 
Granger and W. K. Gregory, 1928, Noyitates, No. 72, pp. 1-9, 4 
text figures. 

A New Species of Alligator from the Snake Creek Beds. By C, C. Mook, 
1923, Novitates, No. 73, pp. 1-18, 5 text figures. 

Baluchitherium grangert, a Giant Hornless Rhinoceros from Mongolia. By 
Henry Fairfield Osborn, 1923, Novitates, No. 78, PP- 1-15, 9° text 


gures 
Some Skulls of Percherus [Thinohyus} from the White River and John Day 


Formations. By Helga S. Pearson, 1923, Bulletin, XLVII, Art. 3, pp. 


61-96, 17 text figures. 
A Contribution to the Knowledge of Meritherium. By H.. Matsumoto, 
1923, Bulletin, XLVIII, Art.-4; pp. 97-139, 11 text figures. 


Y 


e 


nae oS 


is ae 
eK 


A. 
. Say 


2 


4 
4 
H 


Cc 
SUESS , WILLIS, HOB8S, STAMP, EGORY 
UMPELLY , HUNTINGTON , KONSHIN , BURRARD, HAYDEN 
GRABAU, DE LAUNAY, LEUCHS, OBRUCHEV, GERASIMOFF 
Fig. 1. Map of Asia, showing the trend of mountain ranges, and the position of 
5] ) 
the great basins. 


Large positive elements are shaded in slanting lines. The great structural basins are stippled. 


55.18(51.7) 


Article V.tBASIN STRUCTURES IN MONGOLIA! 


By Cuarues P. BerKEY AND FREDERICK K. Morris 


CONTENTS 
EERIE eer aM (oii are eat 87 x owe g 2 kk Ue Oe RR Bae eth tag 104 
RUT 0 71S (2 Re SR We A gr) AE eT 106 
ReMIOMANTICLIL DIR Cee y So RE NO ec is ae. aha c ipeey AL pe eae IRIN AS kee 109 
wep NS FEY ey c's Sn 8 AD A ROM RA GDS Wt 110 
SG POUL WANES. 2! e eames Pee hans cid cain clon ee eh eee 112 
POOUIECH OF PIRAINOMbA Mera. ick. ced OSL be Dele eA oe Sees 114 
UWA er (esa th 1g mens, oo) TL AE oe SERRE On RN ZR Ng en an Oem ea 118 
RNIN ete LATIF ANI METLC ER Grs'o os, siete 0 «fe cocioke echinacea seks bec Rh a A seule ee 122 


Publications of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
bution No. 29. ; 


103 


INTRODUCTION 

Between the vast areas of Asia enjoying ocean drainage lies a series 
of great interior basins of diverse type, extending from the Amur head- 
waters in the northeast to the Caspian in the southwest; and from the 
Arctic divide to the Himalayas. These basins, as indicated on the 
accompanying map (Fig. 1), are the Gobi, the Dzungaria (or Sinkiang), 
the Lop (Tarim or Taklamakan), the Balkash and Aral-Caspian, besides 
high intermontane basins like Tibet and Iran. All these are semiarid 
steppe-countries, including desert ranges, broad, open minor basins, 
and occasional depressions with lakes and salt pans. 


A ) ARCTIC TUERIN IREN NG KALGAN 
eo VIDE | ee | KiANG PACIFIC DIVIDE 4 | 
° 100 200 300 500 600 MILES 
g proriie A 
Arctic Sain Noin _ Gorida Uskuk Peak Tsagan Nor Baga Bogdo 
divide (7000) (6800) (8500) (5200) ote 
loooo' 
5000’ 
SEA L. 
fe) 20 40 60 80 100 120 140 MILES ~~ 
Da a ee 
PROFILE B 


Fig. 2A. Profile across the Gobi region between Kalgan and Urga, showing a 
broad, very shallow downwarp between the Arctic and Pacific divides. Vertical 
scale < 10. 

B. Profile across the northern part of the Gobi region, between the Khangai and 
Altai mountain ranges, showing stronger warping and block-faulting in this part of the 
great basin than in the eastern section. Vertical scale X 10. 


The Gobi, easternmost of the great basins, has a width of, roughly, 
500 miles north and south, and a length of 1,000 miles east and west. 
The entire country from the southerly margin to the Arctic divide is 
warped into a gently sloping concavity, or broad open syncline, whose 
central portion is 3,000 feet lower than the outer margins. Thus the rims 
of the basin stand from 5,000 to 7,000 feet above the sea, and the broad 
downwarped expanse between forms a basin-shaped plateau, parts of 
which are real desert. Generalized profiles are plotted in figure 2. We 
judge that the eastern profile represents almost pure warping, while the 
western profile is the result of warping plus extensive faulting in the 
midst of the original basin. 


‘104 


1924] Berkey and Morris, Basin Structures in Mongolia 105 


CAG 


5 aa is < 3 : 
" oS - i | aS ae Sea fet “ag iy Ry y/) ¥ 
soit m7 a oie Vg Zee "4 1/13 2 


@ A 
0 50 100 MMW, 
<= 


hr PERINE 1500 


Fig. 3. Map of the Gobi region. 


The mountainous areas are shaded with slanting lines. The lowlands are white, while the deeper 
depressions are stippled. This map serves as a location map for the talas named in the text, and also 
shows the routes of the Expedition. The index-letters have the following meanings: AO, Ardyn Obo; 
D, Djadochta; G, Golobai; GS, Gurbun Saikhan; ID, Iren Dabasu; IM, Irdin Manha; K (northwest 
of Peking), Kalgan; K (in the west, just north of the initial A of ALTAI), Kobdo; MT, Murukh Tchu; 
O, Oshih (Ashile); OS, Ondai Sair; PK, Pang Kiang; SM, Shara Murun; U (south of the K of 
KHANGALD), Uliassutai. 


The great basin of the Gobi contains many minor basins, which we 
are calling ‘‘talas,”’ from a Mongol word for an open steppe-country 
(Fig. 3). The following talas may be demonstrated: the Dalai Nor tala, 
now draining through the Argun river to the Amur; the Iren tala; the 
Gashuin Nor, or Edsin Gol tala; the Kisin or Shargin tala; the Khara 
and Dzapkhin, or Kirghiz Nor tala, in which are the cities of Kobdo and 
Uhassutai; the Tez, or Ubsa Nor tala. Each tala has its own local interior 
drainage and is bounded by inconspicuous warp divides or by mountain 
ranges, or both, separating it from neighboring areas of similar habit. 

Again, within each tala there are still smaller basins, which contain 
sediments of late Mesozoic or Tertiary age, or both. These smaller units 
appear as broad level spaces whose surface is beveled by the Gobi pene- 
plane, the remarkably smooth flat surface of which is one of the most 
striking features of Mongolia. These basins of the third order of magni- 
tude are the units of special interest to this investigation. Inquiring of 
the Mongols as to the derivation of the word “gobi,” it was found that 
such open level-surfaced basins are called ‘‘gobis,” and we can think of 
no better term by which to call them in science. A cross section taken 
from one of the field books will serve to show the typical structure of such 
a minor basin, or gobi, in a restricted sense (Fig. 4). 


106 Bulletin American Museum of Natural History | [Vol. LI 


NATURE OF THE ROCK FLOOR 

The oldrock floor is a complex of ancient sedimentary strata, meta- 
morphic rocks and intrusive igneous masses, both large and small. The 
oldest rocks recognized by us are complex injection-gneisses, crystalline 
limestones and related rock types exhibiting the most complex structural 
conditions and mixtures of composition found in the whole region. They 
appear to be more deformed, more modified, and richer in injected igneous 
material than any of the other series, and they are the most confusing to 
interpret. They should be the oldest of all and may indeed correspond 
to the Archean of other lands. On these grounds they are judged to be 
Archean, and to represent the T’ai Shan complex described by Willis and 
Blackwelder as occurring in China proper. 


Ne 1 2 3 4 Ss 6 ee, 8 9 1040 «61 2 3 4 5 6 7 


Bathylithic granite 
Fig. 4. Cross section from the geologist’s field notebook, showing the basin 
north of Uskuk Mountain. 


Ancient crystalline rocks—gneisses, schists and interbedded limestones, invaded by granites repre- 
senting the Great Mongolian Bathylith, form the warped floor on which later sediments have been 
accumulated. Further deformation with faulting has affected the basin, pushing up the Mt. Uskuk 
block and making possible development of alluvial fans. Erosion has stripped some of the floor and has 
carved a broad shallow valley in the sediment, in the almost abandoned depressions of which a salt 
pan is located. This section is almost continuous with Fig. 10, which lies south of the Uskuk block. 


The next younger series includes schists and crystalline limestones 
which are found in the Tsetsenwan hills. Between Shara Murun and 
Ardyn Obo, there is a great series of greenstones and phyllites which we 
consider to be older than the Khangai graywackes next to be described. 
The fact that these series have not been observed in close contact with 
one another makes it difficult to determine their exact relations. They 
are very clearly more intensely modified by metamorphic processes and 
constitute a more varied series than does the graywacke series referred to 
as the Khangai, and they evidently belong to a more ancient geologic 
time. Perhaps the early Proterozoic system, the Wu T’ai Shan, as used 
in China, is large enough to hold them. 


IWillis, Bailey, Blackwelder, Eliot, and Sargent, R. H. 1907. ‘Research in China,’ I, pp. 19, 59, 
99, 157; II, p. 1; Carnegie Institution of Washington, Pub. 54. 


1924] Berkey and Morris, Basin Structures in Mongolia 107 


The most widespread sedimentary unit was given the field name 
“‘Khangai series,”’ from a range of mountains of that name on the Arctic 
divide. The total thickness of the series is not less than 20,000 feet. It 
consists almost, wholly of graywackes, siliceous argillites and slates. 
Locally, red j jasper or blue siliceous limestone is found with these strata, 
but these types do not figure heavily in the series as a whole. The rocks 
are not very highly metamorphosed, and yet they are surprisingly un- 
fossiliferous. The immense extent, great thickness and uniform character 
of this formation are most impressive. The strata are everywhere 
strongly folded, and because of the lack of fossils, together with the gen- 
eral structural relations, are judged to be late pre-Cambrian—possibly 
to be correlated with the Nan K’ou of Richthofen and Bailey Willis, 
or the Sinian of Grabau,? as described for China. 

Besides many minor intrusive bodies, all these series of rocks are cut 
by an immense bathylith, dominantly granite, the great stocks of which 
are exposed over broad areas. Apophyses from it were found as far north 
as the Khangai range and the Gangin Daba and Olon Obo mountains, 
and it is our belief that the granite of Nan K’ou Pass above Peking be- 
longs to the same unit. Possibly this bathylith is genetically responsible 
for the mineralization of the gold veins north of Urga and in the Ulias- 
sutai country. 

Overlying these formations unconformably are marine limestones, 
limy shales and sandstones with fossils of Mississippian, possibly of Penn- 
sylvanian, and certainly of Permian age. All are very complexly folded 
and faulted and have been almost completely swept away in the course 
of later erosions. Only two comparatively small infolded synclinal and 
graben-fault remnants of the sedimentary formations of this short- 
lived, but significant, Paleozoic geosyncline were found in Mongolia. A 
few dikes cut these beds, but no large intrusives seem to have reached 
them. 

Unconformable, also, upon the pre-Paleozoic rocks, is an extensive 
and persistent series of continental conglomerates, sandstones and minor 
shales, with interbedded tuffs and surface flows, the aggregate thickness 
of which is in places more than 10,000 feet. Obscure woody plant- 
remains are found in these coarse clastics, and at one locality several thin 
seams of coal were seen. All are strongly folded and the question may be 
raised whether the same disturbance folded these that plicated the marine 


lyon Richthofen, Ferdinand Freiherr. 1882. * “China,” IT, pp. 316-317, Berlin. Willis, Bailey, Black- 
welder, Eliot, and Sargent, R. H. 1907. ‘Research in China,” a eyey 123, Carnegie Institution of 
Washington, Pub. 5 

2Grabau, A. Ww 1922. “The Sinian System,” Bull. Geol. Soc. China, I, p. 44. 


801 


‘ “41 WOIf poddi1ys ueeq Sey }VY} FUOUIIpES Jo JUNOUIB SHOtIOUS 9Y} PUB 
IOP plo 24} Jo ssouuaAoun oy} saziseydurs YoIOYS OY L, *punoI#aIOF OUIAI}XO OY} UI UIBJUNOTY YIYSO Ut sivaddvai Jooys BAL] Po}[NV} 94} Jo espe oy, *10}e}00ds oY4 
pismo} prvmMuMop A[qvzOU Spueq BAB 94} JO aZpa oy} pues ‘BSaUT SITY} JO YUOI LOAM OY} BuOTe sun zNBF Y “YOROYS VY} JO 10}U99 OY} UT BSoUL pROig dy} Ul SIveddvar 
BART OY, “yoeys Baw] B Aq paddvo suo usYyNOS oY} ‘gpue[peq JO Stare OMY o1B SABPLI osoy] JO JOOS ay, VW “SOSplI YoOI-prey oy} IOAO popuojxo 90u0 78Y} S}UeUIIpes 
sjuaseided ‘ninyy tffnag payee ‘FYystr ey} 7B Toprnoys yay ey, ‘eyes Aq 4nd ‘ouoZsSoUIT] aUTT[e}SAID JO IoY4}O ay} pus ‘KaAkydiod ay AyoV1} papfoj Jo auo oy} ‘qs1Su09 
aoUBYSIP V[pplur ay} UT ST[TY Jo gosUBI MOTOM], “UOISUATUIP SIY} Ul esuURI Tez, VY OY} JO SSoUMOIIEU OY} SMOYS ‘yynos 07 YYLOU WOT; BTyYord ul uses ‘opsog eseg 


“pAvAySoM SUPYOO] “UISeG YIYSO oy} Ul YopYG “¢ “SM 


“nuNlY Y24S0 U2 SyuaWIpes saybiy pub yaays DARTP 
“DIDY uorng Jo 700f 7D auo + dun Jp evo ‘spsow eaddln2- pany owe 
ANC PUL PIDMOZ P4OMIsem SanuyyuoI ssDdU 
Aumyuaunpas Siy/- $42}/U7 2049/0 YY burddoyseno nny iljnsy fo sepjnoys— juewpe 
‘sarskydiod = DIDYy uoIDE PUD DIDYYy BEUNT 


4 


apluaks + aUOsSaul) = DIDYY UNS] ~ UISDY YIYsSO UI S4aI/UfE f 
I 3 opclatael ~ Soy wos apfosdssois v ~ opbogy obog Jo atfosd funjsig” 


smoys p/Dmysem buryoo; uipyunody Y)IYSO J? epay wsajsam cuosf lL 


J 
2 an 
z a 
a7 
My 


ra 
i 


/ 
4 


TNT Zp 
r f 7 
6 Rice is 


nuny y? Sf pe f Le Se Z 2 Sao 


—— 


ERA R 


be yy, vase 
niny Wns ophoy vbog nunyy eeung pany yy uosogy duin> 


1924] Berkey and Morris, Basin Structures in Mongolia 109 


Paleozoic beds. At another time, we may submit evidence for the view 
that the two series were folded at different times and that the conglomerate 
series is approximately of the same age as the folded coal-bearing con- 
glomerates and shales of northern China, which are conceded to be of 
Lower Jurassic age. The latest mountain-folding, then, follows the 
development of these Jurassic strata and, if this correlation proves ad- 
missible, must be of Middle or late Jurassic date. 

At this point in the geologic column, a complete change of structural 
behavior takes place. All earlier rocks are mountain-folded and exten- 
sively deformed, whereas those strata that lie above them are not folded 
at all, and, although they are tilted and warped, or broken, they are 
only locally much deformed even in this way. The Jurassic unconformity 
marks, therefore, the most significant break in the whole geologic column 
for central Mongolia. It marks the change from repeated deformative 
revolutions to a state of much greater stability and the establishment of a 
continental history that has persisted unbroken to the present day. 


BASIN SEDIMENTS 

All these older disturbed rocks were worn down to a mature, rolling, 
locally baseleveled surface before the first of the nearly horizontal sedi- 
ments in which we find vertebrate fossils was laid down. A’sketch (Fig. 
5) made in the Oshih (Ashile) basin, oldest of the gobis, shows, besides 
sediments and an interbedded lava flow, two inliers, one of the pre- 
Cambrian limestone, the other of possible Jurassic porphyry flows; 
both have been brought into view by the stripping away of the sediments 
rather than by local uplift or by faulting. The later sediments, such as 
those of Urulji Nuru, doubtless once covered the tops of these inliers and 
probably even such mountain groups also as the more distant Baga 
Bogdo, Artsa, Bogdo, and the Gurbun Saikhan,—the easterly represen- 
tatives of the Altai mountain range. 

The Oshih beds contain sauropods and primitive armored dinosaur 
bones and may prove to be Lower Cretaceous (Comanchean) in age.! 
The sediments are entirely of inland continental type—sands, gravels, 
clays, paper shales, gypsum-bearing clays—and of moist to semiarid 
climatic association. The entire fauna is non-marine and very diffi- 
cult to correlate with the faunas of other regions. It clearly represents 
an inland, relatively isolated basin.2, At this remote time, then, Mon- 


1Osborn, Henry sien es 1923. ‘‘Two Lower Cretaceous Dinosaurs of Mongolia.’’ Amer. Mus, 
Novitates, No. 95, pp 

Reis, O. 1910. Die Binnenfauna der Fischschiefer i in Transbaikalien.’’ Explor. Géolog. Chemin 
de Fer, Sibérie, Petrograd. Cockerell, T. D. A. 1924. ‘‘ Fossils in the Ondai Sair Formation, Mongolia.” 
Amer. Mus. Bull. (in press). 


110 Bulletin American Museum of Natural History [Vol. LI 


golia and Angara were dry land, warping into inland basins at first of 
swampy deposition, with lacustrine and flood-plain types predominating, 
but later and intermittently developing more pronounced aridity. 


ORIGIN OF DEPRESSIONS 

In a paper now in preparation, we discuss the origin of depressions 
in the Gobi desert of to-day, and attempt a review of the literature on 
this subject. A brief statement must suffice for the present. 

The text of this paper will make it clear that the basins cannot all be 
of the same age; it is very improbable that just the same climates pre- 
vailed during all the periods recorded by basin sediments, although we 
shall offer evidence in another paper to support the thesis that the 
climate of Mongolia in the past has varied between relatively narrow 
limits,—between semiarid and desert conditions. It is improbable, 
furthermore, that the relief was similar in all the periods represented, 
but it must have ranged from a peneplane to a surface rugged enough to 
provide coarse rubble. The part played by warping has been touched 
upon. The basin or depression which now contains the sediments is 
not by any means coextensive with the original area over which those 
sediments were deposited, yet we must deal primarily with the basins 
that have retained sediments to the present day. 

We believe that the basins of deposition were not of simple origin, 
but were the resultant of a number of interacting agencies. To choose a 
single general case as an illustration, let us suppose, in Tertiary time, a 
region of moderate relief, including areas of oldrock and of basin sedi- 
ments, dissected in part by stream work and in part by deflation. The 
following variable quantities will enter into the problem :— 

1. Greater and lesser depth of eroded hollows. 

2. Greater and lesser areas of hollows. 

3. Hollows or depressions in oldrock areas, where all the sediments 
deposited in the hollow must be derived from the slow decay of the hard 
rock. This introduces in turn a new variable, the amount and character 
of loose material available in the hard rock area. A long quiescent 
period, and especially a moist climatic rhythm, would have developed a 
good mantle of soil. 

4. Hollows in or near sediment basins, where easily eroded material 
may be washed out or blown out, offering a great supply of sediment. 

5. Disposition, no less than amount, of water supply. 

6. Relation of prevailing winds to source of sediments. 


1924] Berkey and Morris, Basin Structures in Mongolia et 


Independently of all these variables, it seems necessary to introduce 
the hypothesis of periodic warping, as will be made clear in the pages that 
follow. 

In figure 6 we show a region of low, mature relief, broad lowlands or 
hollows alternating with higher regions, In figure 7 these are warped 
into very gentle swells and depressions, which, save by chance, bear no 
relation to the topography. Where a downwarp coincides with a low- 
land, as in figure 7, A-B, the condition will be most favorable to the 
retention of sediment, and the basin will be filled. Where an upwarp 
coincides with high ground, as in figure 7, C-D, a broad divide will be 
formed, from which debris will be washed into surrounding regions. If an 
erosion-hollow is upwarped, its resultant behavior will depend roughly 


Fig. 6. Lay figure of a maturely dissected region of low relief, before warping. 
Fig. 7. The same region, after warping,—showing how basins and uplands may 
be the result of original relief and deformation combined. 


Let both higher relief and upward movement be denoted by the + sign, while lowland and down- 
warp are denoted by the — sign. Then in A-B, both relief and movement are negative, so that a basin 
is formed. In C-D, both influences are positive, while in E-F and G-H, positive and negative elements 
combine, making either basin or highland according as the relief factor is greater or less than the warp- 
ing. Thus in the Oshih ba asin, Fig. 5, ridges of hard-rock have formed part of the lowland, and have 
— covered by sediments. 


upon the algebraic sum of two opposite influences, the depth of the hollow 
and the amount of uplift. Thus, if the upwarp is less than the depth 
of the hollow, as in figure 7, E-F, the hollow still remains, though dimin- 
ished as a possible repository of sediments. Where a divide is down- 
warped, it will be included in a basin and will be covered over with sedi- 
ments, if the downwarping considerably exceeds the height of the divide; 
otherwise it may still remain a divide. 

If we suppose that, in general, the basins formed where all the con- 
ditions were most favorable, the amount of warping that need be 
assumed would be reduced to a minimum. The relative part played by 
wind and water in the erosion of primary hollows or valleys forms an 
important subject in itself, and a report on the evidence is in preparation. 


112 Bulletin American Museum of Natural History [Vol. LI 


TYPES OF GOBI BASINS 

Two types of gobi basin are distinguished: first, the faulted or 
piedmont gobi basin, found along the base of the fault-block mountains 
of the Altai; and second, the warped or plains type, in the less disturbed 
areas. The types are not sharply demarked from one another, but the 
distinctions are instructive. 

1. The faulted, or piedmont, gobi basins are distinguished by: (a) 
greater thickness of sediment; (b) great range in texture of sediment,— 
including paper shale, fresh-and-brackish-water limestones, clay, sand, 
gravel and coarse rubble, even with great bowlders; (c) wide range of 
time-periods represented in the sediments,—which implies. longer life 
of the basin as a locus of warping; (d) igneous flows, which are common 


Gurbun Saikhan 
650 


Fig. 8. Cross section of the Djadochta region. 


This is a simple faulted type of gobi basin. The beds near the mountain front dip 56° north. 
Younger beds are encountered as one goes toward the mountain front. In this respect the basin is 
comparable to the Baga Bogdo piedmont basin, Fig. 10. It is not positively known that Oshih beds 
underlie the Djadochta; their presence is inferred. Compare with Fig. 11. 


in these basins. Volcanism seems to have been associated with faulting, 
and all the more disturbed basins observed by us contain lava flows 
(Fig. 8). 

2. The warped, or plains, gobi basins are shallower and contain 
thinner sediments, of but few periods, so that the gaps in the geologic 
record are even larger than in the sediments of the piedmont basins. The 
range in coarseness of sediment is less wide—sandy clay, sands and small-. 
pebble gravels predominate. In addition, there are a few occurrences 
of fresh-water limestones, dense limy marl, marly sandstone, and thin 
beds of gypsum. None of the warped basins has been markedly disturbed 
by faulting or tilting. 

None of the basins of either type contains a complete record of the 
sedimentation; none even contains all the horizons represented in 
Mongolia. In each basin, the sedimentary formations are separated by 
disconformities or by angular unconformities, representing long gaps or 
erosion-intervals in the geologic record. 


1924] Berkey and Morris, Basin Structures in Mongolia 113 


These facts will be considered in some detail, and will be marshaled 
in support of three theses, as follows: 

1. Warping took place in a series of successive increments of move- 
ment, separated by pauses or intervals of quiescence long enough in some 
instances to permit the peneplanation of faulted and warped sediments, 
and to allow for a complete change of fauna. 

2. The locus of warping shifted from place to place; that is, after 
deposition of one formation in a given basin, warping did not as a rule 
continue in that basin, but rather in another place, which, in turn, having 
received a shallow filling of sediment, passed into a quiescent period, and 


angat Camp « 1 MS “@ ey SS 
m7 chackhurle, a & N\A ci TS 
eee pre iy iw 
7 


USKUK Mts 2 Se on ‘&>Frin Nor 
smne< Ht nda Sai ir 
ates t Jtisanda Gol, Loh 
Faptlung Kureh 
aot ene SS 


(iva *S 
ran 5 


\ 
“A lic, 
N Y ZW) 
Salt oes ae 
. Go. 
. al ES Sparsh shill uta 


pup WN 


Ba 2 EDO aay juss sane 


ony ws >" B 
Dy 
£ mn Ine 0,8 D 3, 
if ni 


x SL = Sse = a 
Ne we Coes x My “4 d ‘ 
(Siig, rn Rr ~~ SI 
ar, ta) apn k =. 


ial “1 nwsanyy 
oY, mos 


QBor Shandi p 
) 


Se Whi Micsisliiyy,, ANAT 1G 

Wij: } Zagat YU yy ANY 

etn, nO tO ng. Sono, u\\IIfyy 
7) 


‘oDzamain Tek ming 
J 
' 
1 


Fig. 9. Sketch map of the eastern Altai region, showing location of stations 
mentioned in the text. 


The limits of the sediment basins are not sufficiently known to justify an attempt to draw them. 


114 Bulletin American Museum of Natural History [Vol. LI 


the next locus of deposition was in still another basin, or possibly back in 
the first. Each basin had its rhythm of alternating deposition and 
quiescence, which probably implied removal of part of the earlier deposit. 
Some basins, however, continued to warp intermittently during longer 
periods of time than others. 

3. Small units, such as the Mongolian gobi basins, carrying shallow 
sedimentary fills, probably cannot of themselves set in motion the deep- 
seated shift of material required by the theory of isostasy. It is not 
improbable that we sometimes overestimate the effect of the positive 
weight of the sediments, as well as of the negative load due to stripping 
of up-arching areas, as the actuating causes of warping. This thesis 
does not imply that the authors do not accept the principle of isostasy 
as applying to earth movements of the first magnitude. 

We will describe two basins briefly, a faulted or piedmont basin, and 
a warped or plains basin. 


FAULTED OR PiepMontT TyPE oF Gost BASIN 


Each of three ranges of the eastern Altai, the Baga Bogdo, Artsa 
Bogdo and Gurbun Saikhan, presents essentially a fault-front along its 
northern margin (Fig. 9). 


-Peak of Baron Shiliuta 
(97 Uskuk Mb (west of this section) BAGA BOGDO 


certs Ondai Sair 


pa Restored profile of eroded lava flow Alluvial fans 
Sirk Bey Bee Hsanda Gol Ze 
Higher gravel benches 


aE 
were d ===> ——————— 
B.Shiliuta 
E.} Bava 


Fig. 10. Cross section of the Baga Bogdo faulted gobi basin. 


Though on a much smaller scale, this section is almost continuous with Fig. 4, lying south of the 
latter. The Uskuk fault-mountain lies just west of the section and so is drawn in dotted lines. Part 
at least of the sediments once rested upon the Uskuk block, because remnants of sediments have been 
found upon it, and apparently the same lava flow that is seen in the sediments caps the granite of Uskuk 
peak. These relations are indicated in the dotted lines. The lava-capped butte, Baron Shiliuta, is 
seen projected against the higher peak of Uskuk. A short cross section, oriented east-west, is intro- 
duced to show the relation between these two peaks. The Cretaceous Ondai Sair beds are much more 
deformed than the Tertiary formations, which overlie them unconformably. Tertiary beds are also 
deformed by both faulting and warping. The southern part of the section, between Hung Kureh and Baga 
Bogdo, is much more disturbed than is represented in this drawing, and carries the latest sediments of 
the region. The Gobi peneplane is highly developed in this district, and is itself deformed. Since the 
PEneplene bevels Pliocene beds, this section is one of our best means of determining the date of its 

ormation. 


In the Baga Bogdo piedmont basin (Fig. 10) the following section 
is recorded, from oldest to youngest :! 


1Berkey, Charles P., and Granger, Walter. 1923. ‘‘ Later Sediments of the Desert Basins of Central 
Mongolia.’?’ Amer. Mus. Novitates, No. 77. 


1924] Berkey and Morris, Basin Structures in Mongolia 115 


1. The oldrock floor, consisting of schists, marbles, graywackes 
and slates, all of diverse pre-Cambrian age invaded by granite. Infolded 
in this complex are conglomerates and sandstones, locally containing 
seams of coal, which are judged to be Lower Jurassic in age. These 
form part of the rock floor, not part of the basin sediments. 

2. The Ondai Sair sands and paper shales, of Lower Cretaceous 
(Comanchean) age, resting upon the old floor. They are at least 500 
feet thick, and are faulted and tilted. 

3. About 3,500 feet of early Tertiary gravels, sands and sandy clays 
resting unconformably upon the Ondai Sair. They include at least one 
lava flow, and are in places uptilted and faulted, but they do not share 
all the disturbances of the Ondai Sair. The higher beds carry the Lower 
Oligocene Baluchitherium fauna, and are called the Hsanda Gol formation, 
a name which is provisionally extended to the base of the conglomerates, 
though the lower beds may yet prove to be Eocene. 

4. The Lower Miocene clays of Loh, less than 100 feet thick, 
resting upon the Hsanda Gol clays, without any obvious physical 
disconformity. Going southward along their dip (Fig. 10), we found 
that they were succeeded by an undetermined thickness, probably 
as much as 1,000 feet, of ‘sandy clays and sands, in which as yet no 
fossils have been found. 

5. The Hung Kureh sands and clays, about 2,000 feet thick, 
gravelly toward the base of the exposure. They carry a fauna of late 
Pliocene age, according to the opinion of Mr. Walter Granger, paleon- 
tologist of the Expedition. They have been disturbed by tilting and 
faulting, which may be of the’same age as that which deformed the 
older Tertiary beds, but which are probably of later date. 

6. A mantle of coarse rubble, clearly derived from Baga Bogdo. 
This coarse rubble is at least 2,000 feet thick, is partly consolidated, 
and is unfossiliferous, so far as we now know. Its age may be latest 
Pliocene or Pleistocene. 

The thicknesses of all the formations exposed in the basin add up to 
8,500 feet, but, since the whole thickness of some of the members is not 
known, it may be as much as 10,000 feet. Only a fraction of these strata 
was accumulated in any one period, yet this basin is the longest-lived 
and most active basin yet observed in Mongolia, containing the oldest 
and the youngest basin sediments of which we have record. 

Among the features that bear upon the problem of basin-structure 
may be mentioned the following: 

1. The sedimentary formations dip toward the Altai, and younger 
beds are encountered as we go southward toward the Altai front. 


116 Bulletin American Museum of Natural History [Vol. LI 


2. The conglomerate at the base of the Tertiary consists of well- 
rounded pebbles that were derived almost certainly from the Jurassic 
conglomerate of the general Uskuk region. It would seem that these 
beds were washed into the basin from the north, rather than from the 
south, implying that the present Altai was not a notable range in the 
early Tertiary. 

3. The Hung Kureh formation consists of fine materials everywhere 
except toward the base, where it grades into gravel, and therefore in- 
dicates no very marked relief in the Altai region in Pliocene time. The 
first evidence of rugged relief is the deluge of coarse rubble overlying the 
Hung Kureh. 

We infer from these facts that deformation has been secular and 
progressive, with long periods of quiescence alternating with periods of 
disturbance. Thus faulting and tilting took place after the Lower | 
Cretaceous (Comanchean) and after the Pliocene, at least, and possibly 
after the Miocene as well. 

An even more suggestive example is found in the Oshih and Dja- 
dochta divisions of the Altai piedmont basin. The Oshih is the pied- 
mont basin of the Artsa Bogdo range. Here are about 2,000 feet of very 
considerably faulted and tilted sediments. The faults are of small throw, 
ranging from five feet to 200 feet, and in one instance possibly even more; 
the dips, apart from drag in the fault zones, do not exceed 24°, and are 
for the most part 10° or less. The Oshih contains fossils of about the 
same age as the Ondai Sair and is almost undoubtedly continuous with it. 

South of the Artsa Bogdo range lies a basin whose sediments con- 
tain reptile bones, which Granger judges to be of Oshih age,—indicat- 
ing either a former extension of the Oshih basin over the site of the present 
Altai, or an entirely different basin separated from the northern Oshih by 
a Lower Cretaceous divide. 

North of the Gurbun Saikhan, about forty miles east of Oshih, 
lies a piedmont basin whose floor is not exposed, so far as observed. 
This basin contains the Djadochta sands, of Cretaceous age, with the 
Protoceratops fauna and the dinosaur eggs. The sand of the Djadochta 
is very fine, of uniform red color, and about 500 feet thick. The beds 
dip gently southward into the Gurbun Saikhan piedmont basin, and have 
been very faintly tilted and arched, but not faulted. 

These sands cannot have been derived from the direct weathering of 
the Gurbun Saikhan, which is composed of such rocks as graywackes, 
argillites, phyllites, limestones, serpentines and diorites. They must have 
been derived instead from a preéxisting sand that had already undergone 


1924] Berkey and Morris, Basin Structures in Mongolia 117 


Site of the Gurbun- 
Saikhan range Sandy sediments, possibly Sairim or Oshih 


Kb 


F 


LAW 


A_Nl Ne 
EN 


\\ C. DEPOSITION OF POST-DJADOCHTA GRAVEL 


Reg and brown clays, 


= sands and gravel 


Y) | SSA 
LUNE ZENA ce 


\\, D. DEPOSITION OF GASHATO (Paleocene?) 


Fig. 11. Four stages of warping indicated in the Djadochta basin. 


A shows an old basin of sandy sediments deposited upon a floor of complex old rocks. B shows 
initial upwarp that caused the Djadochta sands to be washed down into the newly formed basin, without, 
however, exposing crystalline rocks of the present Gurbun Saikhan range. C shows the deposition of a 
gravel of angular pebbles, derived from the Gurbun Saikhan rocks, and hence implies that those rocks 
have been laid bare. D shows the deposition of the Eocene Gashato beds, the peneplaned remnant of 
which is at least 300 feet thick, and may be much more. The structure of the rock floor of the basin is 
inferred from the granite exposed to the north of the basin, and the rocks observed in the Gurbun 
Saikhan range. The successive warpings of relatively small throw suggest the method by which this 
wing of the Altai was made. 


considerable assorting. The only earlier sand of which we have knowledge 
in the region is the Oshih. Now the highest member of the Oshih that 


we have seen is a red-and-white sand, called the Sairim member, which 
is strikingly similar to the Djadochta sand in composition and texture. 


118 Bulletin American Museum of Natural History [Vol. LI 


Its present easternmost outcrops lie less than forty miles from the present 
westernmost outcrops of the Djadochta, and both formations must have 
been more extensive in the late Mesozoic. We suggest that the Oshih, 
at least in its upper members, may once have overlapped the present site 
of the Gurbun Saikhan (Fig. 11A). It is not improbable that when up- 
lift began in early Djadochta time, the present Gurbun Saikhan rocks * 
were not exposed, but that the overlying Oshih sands were eroded and 
washed into the newly formed basin, to become the Djadochta deposits 
(Fig. 11B). As evidence for this view, we cite the facts that a gravel of 
angular pebbles overlies the Djadochta, and that these pebbles are all of 
Gurbun Saikhan origin, which suggests that the next, post-Djadochta, 
increment of war ping brought the Gurbun Saikhan rocks to light (Fig. 
NC): 

The age of these gravels is known approximately, since they lie 
underneath the Gashato clays and gravels which carry an early Eocene 
fauna, according to identifications by Granger (Fig. 11D). 

Therefore, the chief dates of warping in this basin fall: (1) in Oshih 
(Lower Cretaceous) time, when the pre-Djadochta sand-terrane was_ 
being deposited over a large part of the eastern Altai region; (2) in 
Djadochta time (early Cretaceous ?); (3) in post-Djadochta time, when 
the Gurbun Saikhan rocks emerged; (4) in Gashato (early Eocene) 
time. 


WarPED Tyrer or Gost Basin 

The Irdin Manha basin includes the Iren Dabasu, limited on the 
north by slate hills, and the Shara Murun, limited on the west by com- 
plex old crystalline rocks. The southern boundary consists of hills of 
granite, schist and graywacke. We have not seen its eastern boundary, 
but the basin covers an area of at least 10,000 square miles (Fig. 12). 

Within the basin is a succession of sediments which do not extend 
throughout the basin, but which overlap in a complex manner, each 
occupying a special area (Fig. 13). 

The oldest sediment is the Iren Dabasu, of late Cr etaceous age, not 
more than 180 feet thick, resting directly upon the crystallines. It is 
succeeded by lake beds which, though nearly barren of fossils, yielded a 
tooth of a small lophiodont, which should prove them Eocene. Upon 
these barren rocks rest the Lower Oligocene Houldjin gravel, here fifteen 
feet thick, but apparently thickening northwestward to fifty feet or 
more.! 


1Matthew, W. D., and Granger, Walter. 1923. ‘‘The Fauna of the Houldjin Gravels.’’ Amer. 
Mus. Novitates, No. 97. 


1924] Berkey and Morris, Basin Structures in Mongolia 119 


a gSatuea. See 


MAW > 


==A limestone 
schist 


ele ure 5 


ily oy 


Fig. 12. Sketch map of the Irdin Manha region. 


This map serves to locate the stations mentioned in the text, but not to delimit the basin. 


At Irdin Manha, twenty miles southeast of Iren Dabasu, the Hould- 
jin is not found, and the section exposed consists of 40 to 100 feet of 
gray sands, with a rich titanothere fauna, which may be late Middle or 
even Upper Eocene.' Beneath the titanothere beds there are red clays, 
provisionally called Arshanto, and probably to be correlated with the 
barren beds above the Iren Dabasu. The Arshanto may prove to be 
only the lower Irdin Manha, or it may be separated from the Irdin 
Manha by a disconformity. The base of these beds has not been seen. 

The Pang Kiang beds, 60 miles farther south, are about 500 feet ° 
thick. In some places at least they rest directly upon the old crystalline 
rocks. Only one fossil has been found in the Pang Kiang, a fragment of a 
rodent jaw. Dr. Matthew identifies this as an ochotonid, which is in- 

1Granger, W., and Berkey, C. P. 1922. ‘Discovery of Cretaceous and older Tertiary strata in 
Mongolia.’”’ Amer. Mus. Novitates, No. 42. 

Berkey, C. P., and Granger, W. 1923. ‘‘Later Sediments of the Desert Basins of Central Mon- 
golia.’” Amer. Mus. Novitates, No. 77. 

Osborn, H. F. 1923. ‘‘ Titanotheres and Lophiodonts in Mongolia.”” Amer. Mus. Novitates, No. 91. 


Matthew, W. D., and Granger, W. 1924. ‘‘ New Carnivora from the Tertiary of Mongolia.’”’ Amer. 
Mus. Novitates, No. 104. 


120 Bulletin American Museum of Natural History [Vol. LI 


sufficiently diagnostic to serve as an index of the exact age of the forma- 
tion. He says, however, that this jaw could hardly be older than 
Miocene.! The exact relations of the Pang Kiang to the Irdin Manha 
must await further field study. 

About 75 miles south of Irdin Manha, and 50 miles southwest of 
Pang Kiang, at the temple Murukh Tchu, a group of clays and sands 
lies directly upon the crystallines. About 150 feet are exposed, but the 
thickness may be 200 feet. No fossils were found in these beds. 

Twenty-five miles south of Murukh Tchu, at Boltai Urtu, a great 
mass of conglomerates, about 1,000 feet thick, rests upon the oldrock 
floor and dips westward and northward away from the oldrock rim of the 


5 Irdin Manha escarpment 
p Houldjin escarpmenty oo. papasu N 
Gobi peneplane| undrained lowland Gob: peneplane | undrained| lowland 
IRDIN MANHA FORMATION “>. BEDS. 


ARSHANTO REDBEDS » ¥ vv vv ve vv vy ee vy vv 
? IREN DABASU? vvvyvv~~yyv PROBABLY GRANITE 


5 10 15 20 F Neos MILES 


Fig. 13. Cross section of a repeatedly warped basin, from Iren Dabasu to Irdin 
Manha. 


The section shows at least three periods of basin-making, and three unconformities. The uncon- 
formities lie (1) at the base of the Cretaceous Iren Dabasu beds, upon the uneven floor of crystalline 
rocks; (2) at the base of the Middle or Upper Eocene Arshanto (lower Irdin Manha) formation, upon the 
slightly disturbed Cretaceous beds, and even upon knobs of the old erystallines; (3) at the base of the 
Lower Oligocene Houldjin gravels, upon the eroded Eocene. Each unconformity implies a long period of 
quiescence before the deposition of the next succeeding sedimentary formation. The fact that the area 
became alternately a locus of deposition and of erosion seems to imply a periodic warping of the region. 
All thicknesses are twice exaggerated. 


basin. These conglomerates pass northwestward under and interdigitate 
with sand and clay sediments which at Shara Murun bear a rich fauna of 
titanotheres and lophiodonts (Fig. 14). Professor Osborn regards this 
fauna provisionally as being somewhat younger than the Irdin Manha.? 

These complex relations are represented diagrammatically in figure 
15. No one place in this broad basin shows a complete section, even of 
the formations represented in the basin as a whole. All the formations 
are thin, their combined thickness probably nowhere exceeding 1,000 
feet. In most parts of the basin it must fall short of this amount. 

There are at least three gaps or breaks in the record, representing 
non-deposition or even removal of sediment: (1) pre-Cretaceous, a 
great unconformity, between the Iren Dabasuand the pre-Cambrian slates; 
(2) Cretaceous-Eocene, between the Iren Dabasu and the “barren 


1Personal communication. 
2Personal communication. 


1924] Berkey and Morris, Basin Structures in Mongclia . 121 


‘beds”’ (Arshanto?); (3) Eocene-Oligocene, between the barren beds at 
Iren Dabasu and the Houldjin. Two other disconformities, of very 
minor value, may possibly be present,—one between the Arshanto and 
the Irdin Manha, and the other between the Irdin Manha and the Shara 
Murun. 

The basin is not walled in by definite mountain uplifts, nor have we 
seen evidence to indicate that the hard-rock boundaries are faulted 
against the basin, except for very minor faulting at the southern bound- 
ary, 55 miles south of Iren Dabasu (Fig. 15). It represents, on the contrary, 
a broad, very gentle warping. The sediments now found in it have 
less than their former thickness, for the region has been peneplaned. 


BARON SOG IN SUMU SHARA MURUN (ULA USU) 


Boltai Urtu Gobi penepl Goli Urtu| Gobi peneplane 


SE NW 


MT. Th 60 


Fig. 14. Cross section of a warped gobi basin, from Boltai Urtu to Shara Murun 
(see map, Fig. 12). 


The section shows conglomerates at Boltai Urtu resting directly upon the complex rock floor, 
These conglomerates die out westward, passing underneath, and grading into the sands and clays which 
at Shara Murun carry a rich fauna of Middle or Upper Eocene age. 


The relation, too, of the sediments themselves suggests the former 
greater extent and thickness of the Houldjin at least, if not of the Irdin 
Manha also (Fig. 13). 

The physical history of this great basin must consist of a series of 
gentle warpings followed by sedimentation, and the epochs of sedimenta- 
tion were separated by longer periods in which sediments either were not 
laid down or have since been removed. In the epochs of filling, the 
deposition exceeded the removal of material from the basin, and this 
excess of deposition over removal is essentially the measure of warping. 
In the far longer quiescent periods, removal of material either balanced 
or exceeded the amount of sediment made available for the streams of 
that day, so that either no sediments were permanently retained in the 
region, or, more probably, part of what had been laid down in the preced- 
ing epochs of deposition was carried away. The absence of deep channels 
in the underlying sediments implies that the older beds were reduced to a 
smooth surface—a local peneplane—before the newer beds were laid 
down. 


P22 Bulletin American Museum of Natural History [Vol. LI 


The vast lengths of time represented by the periods of quiescence 
or stability, the thinness of the sediments, the wide area covered and the 
absence of large faulting or great mountain uplift bordering these basins 
are all cited to support the thesis that the weight of sediments alone 
could not have caused the warping; first, because the thickness is too 
slight to have disturbed the isostatic equilibrium of the earth’s crust, 
second, because the long periods of quiescence indicate that the crust 
sustained not only the positive weight of the sedimentary load without 


Se 


SHARA MURUN 


PANG KIANG 


Fig. 15. Block diagram to show the general relations of the formations. 


The left-hand or front edge runs northwestward and shows the basins between Pang Kiang and 
Iren Dabasu (compare with Fig. 13). The rear diagonal cut shows the section between Boltai Urtu 
and Shara Murun (compare with Fig. 14). The cut from front to rear shows one interpretation of 
the relations between the Irdin Manha and Shara Murun formations. It is possible that further studies 
by Osborn, Matthew and Granger may result in correlating these two formations. 


being depressed by it, but also the negative load due to peneplanation 
without being uplifted. Instead, the next deposit of sediment, resting 
disconformably upon the older deposit, marks a new increment of down- 
warp following the peneplanation. 


GENERAL RELATIONS AND INFERENCES 


Many other examples could be offered from every gobi basin we 
have studied, and all will be described in the larger report now in prepara- 
tion. But we believe that the examples we have given are typical, and 
are enough to support the thesis that warping was a slow and discon- 
tinuous process, taking place by a series of small increments of move- 
ment; and that the sediments are to be correlated with the warping, 
recording, if the records could be read, both the upwarp that determined 
the removal of sediment from the upland, and the downwarp of the basin 
receiving it. 


1924] Berkey and Morris, Basin Structures in Mongolia 123 


In figure 16 are plotted the thickness of sediments actually observed 
in the most important basins. Distances and directions between stations 
are plotted from the westernmost basin at the left to those along the 
Kalgan-Urga trail at the right. The column for China is generalized 
from J. G. Andersson’s “Essays on the Cenozoic of Northern China.’”! 


“2 Ue UR sol ec 


PLIOCENE 


OLIGOCENE 


EOCENE 


CRETACEOUS 


| LOWER oad: Cre rae ae 
| CRETACEOUS ioe apie ated 


‘ ‘ 


TYPE OF BASIN— ‘S-FAULTED— FAULT “——————WARPED  BASINS———— FAULT 


MILES ——> 10 eahee 70 4200 oss 70 109 20 4 70 4 80 4 70 t 
DIRECTION— Ss |SE| SE] E— | SE] © | SE | E | N | SE| SE | SE] SE 


ONDAI HUNG DJA- ARDYN | MURUKH | IRDIN TABUL CHINA 
SAIR KUREH | DOCHTA OBO TCHU MANHA 
STATIONS HSANDA OSHIH GOLO- SHARA _ IREN PANG WAN 
GOL BAI MURUN  DABASU __ KIANG CHUAN 


Fig. 16. Columnar diagram of the formations thus far observed in Mongolia. 


Thicknesses are plotted toscale. It will be seen that the faulted basins contain far thicker deposits 
than do the warped basins. The diagram brings out clearly the shifting of the locus of deposition from 
place to place. The distances indicated at the bottom of the diagram are measured between stations or 
camps, not between the limits of the basins. The section for northern China is compiled from J. G. 
Andersson’s “‘ Essays on the Cenozoic of Northern China.” 


The diagram shows clearly the thinness of the sediments actually observed 
in contrast with the vast lapses of time involved. In figure 17 the 
columns for Mongolia are condensed into one for comparison as to thick- 
ness with those of the Rocky Mountain region, and with the American 
marine column, emphasizing the relative thinness of the deposits thus 
far observed in Mongolia. 

Among the most interesting inferences which the facts support is 
the shifting of the locus of deposition from place to place. The Lower 


11923. Mem. Geol. Surv. China, Ser. A, No. 3. 


EBB conti- MONGOLIA 


nental sedim 


== marine 


sediments 


OSHIH 
DJADOCHTA 
WARPED 


BASINS 
SOUTH DAKOTA 


HSANDA GOL 
WAN CHUAN 
JOHN DAY 
BASIN ORE 
IBIGHORN BA- 
[SIN MONT 
| NORTHWESTERN 
HUERFANO 
BASIN COLO 
COLORADO 
SPRINGS 
S, JUAN BASIN 
N.MEX 
Scale of feet 


PLIOCENE 


MIOCENE 


OLIGOCENE 


EOCENE 


Fig. 17. The columns in Mongolia are shown in condensed form at the ex- 
treme left. The numbered columns represent typical sections in the western 
United States. Formations chiefly of marine origin are represented by parallel 
lines, while those chiefly of continental origin are in black. References are as 
follows: 


1, Sinclair, W. J. 1909, quoted from Osborn, H. F. 1910. ‘Age of Mammals,”’ p. 359.—2, 
Hills, Richard C. 1899. ‘‘Elmoro Folio, Colorado,”’ U. 8. Geol. Survey, Geol. Atlas, folio 58.—3, 
Fisher, Cassius A. 1906. ‘‘Geology and Water Resources of the Bighorn Basin,’’ U.S. Geol. Survey, 
Prof. Paper 53.—4, Winchester, Dean E., Hares, C. J., Russell, Lloyd, and Parks, E. M. 1916. 
“The Lignite Fields of Northwestern South Dakota, U. S. Geol. Survey, Bull. No. 627, p. 16.—5, 
Darton, N. H. 1905. ‘Preliminary Report on the Geology and Underground Water Resources of 
the Central Great Plains,’’ U. 8S. Geol. Survey, Prof. Paper 32. Wanless, Harold R. 1923. ‘‘The Stra- 
tigraphy of the White River Beds of South Dakota,” Proc. Am. Phil. Soc., LXII, No. 4.—6, Osborn, 
H. F. 1897. ‘‘The Huerfano Lake Basin, Southern Colorado, and its Wind River and Bridger Fauna,”’ 
Bull. Am. Mus. Nat. Hist., IX, Art. 21.—7, Finley, George I. 1916. ‘‘Colorado Springs Folio, Colo- 
rado,”’ U. 8. Geol. Survey, Geol. Atlas, folio 203.—8, Sinclair, W. J., and Granger, Walter. 1914. 
“Paleocene Deposits of the San Juan Basin, New Mexico,” Bull. Am. Mus. Nat. Hist., XX XIII, Art. 
22. Bauer, C. M., and Reeside, J. B., Jr. 1920. ‘‘Coalin the middle and eastern parts of San Juan 
County, N. Mex.,”’ U. 8. Geol. Survey, Bull. No. 716 (g.).—9, Lawson, Andrew C. 1914. ‘‘ San Fran- 
cisco Folio, California,’’ U. S. Geol. Survey, Geol. Atlas, folio 193.—10, Branner, J.C., Newsom, J. F., 
and Arnold, Ralph. 1909. ‘‘Santa Cruz Folio, California,’ U. 8. Geol. Survey Geol. Atlas, folio 163. 


124 


1924] Berkey and Morris, Basin Structures in Mongolia 125 


Cretaceous Oshih and Ondai Sair gobi basins are almost certainly con- 
nected. Deposition must have begun in the Oshih earlier than in the - 
Ondai Sair, since the Oshih beds are thicker, coarser and carry a fauna 
of rather more primitive aspect, and since the paper shales and large 
sauropods are found only in thé upper beds of the Oshih, whereas they 
occur in the lower beds of the Ondai Sair. After deposition of the Oshih 
and Ondai Sair, sedimentation in this region ceased, or, if beds were 
deposited, they have been removed by erosion; but at Djadochta, forty 
miles east of Oshih, a gobi basin received 500 feet of early Cretaceous 
fine red sands probably derived from the destruction of Oshih beds. The 
next Mesozoic deposits of which we have knowledge are at Iren Dabasu, 
380 miles farther east, where late Cretaceous beds rest directly upon pre- 
Cambrian slates. The locus of deposition had shifted, therefore, during 
Lower Cretaceous (Comanchean) and Cretaceous time. 

Returning to the Altai and referring again to figure 16, the Gashato 
beds, which are the lowest Eocene strata we have seen, rest on the angular 
gravel which covers the Djadochta red sands (Fig. 11). The gap between 
the Djadochta and the Gashato represents the time recorded in the Iren 
Dabasu trachodont beds, plus an interval of Cretaceous and Paleocene or 
Eocene time that has left no sedimentary record so far as we now know. 
No younger formation was seen in this region, but both to the west and 
to the east there are later sediments. The Irdin Manha and the Shara 
Murun formations represent the highest Eocene yet found in Mongolia. 
They rest in some places upon the oldrock floor, but in other places they 
may lie upon Cretaceous or older Eocene sediments. 

Thus, although we do not know the age of the Murukh Tchu or the 
Golobai, the facts demonstrate that there has been a shifting of the 
locus of deposition during the Eocene from the Altai region in earlier 
Eocene toward the Iren Tala in later Eocene time. 

The oldest Oligocene beds yet seen in Mongolia are, apparently, 
those of Ardyn Obo, which rest directly upon the crystalline oldrock 
floor. The Oligocene of the Houldjin bench at Iren Dabasu should be 
rather younger, since it contains Baluchithercum bones. The Baluchi- 
therrum beds of Hsanda Gol have been called Lower Oligocene by Dr. 
Matthew. 

Whether or not these three formations prove to be of the same age, 
the evidence of shifting of the locus of deposition is convincing. In at 
least three widely separated regions, warping recommenced in Lower 
Oligocene time, so that beds of this age were laid down, in one locality 
upon eroded Lower Cretaceous (Comanchean) beds, in the second upon 


126 Bulletin American Museum of Natural History [Vol. LI 


the bare crystalline rocks, and in the third upon Eocene beds of doubtful 
correlation, somewhat older than the typical Irdin Manha (Fig. 16). 

The only Miocene beds yet seen in Mongolia are those of Loh, 
which rest upon the Hsanda Gol without any notable appearance of a 
break in sedimentation. Thus far, this is similar to the findings in China 
where Dr. Andersson! refers a few beds to the ““Lower Pliocene or Upper 
Miocene.’”? 

This almost complete absence of the Miocene over vast areas of 
basin-lands constitutes one of the major problems of the region. Some 
of the elements of the problems are: 

1. No marine Miocene beds are known in northern Asia, though 
Miocene lignites are reported as far north as the New Siberian Islands? 
hence the sea was remote and the Angara-Gobia continent was broad 
during that period. 

2. The suggestion made by the patchy distribution of the earlier 
formations, as already recited in this paper, is that the deposits are all 
inland deposits. Despite the intermittent warping to form basins and 
receive sediments, therefore, the continent as a whole was stable and was 
undergoing removal rather than deposition of sediments. 

This should imply either that we should find inland sediments of 

‘Miocene age more extensive than we have yet found, or that northern 
Asia was undergoing marked erosion during the Miocene. 

The gaps in the record may represent periods of relative quiescence 
in which sediment was not being deposited, or was being eroded. If an 
upwarp was in progress, it is hard to see why a corresponding deposition 
should not be found as a record of it, unless the entire region was subject 
to erosion at the time. Such general erosion would be recorded in a pene- 
plane, like that which beveled the Mesozoic beds prior to the deposition 
of the Tertiary. The preservation of vast areas of relatively thin, soft 
sediments, which represent many diverse horizons, and which are slightly 
or not at all disturbed, leads one to believe that the country was not 
greatly uplifted, and that it was suffering but little warping or denuda- 
tion during the long intervals represented by the gaps in the sedimentary 
record. 


1Andersson, J. G. 1923. ‘‘Essays on the Cenozoic of Northern China,’’ Table II. Mem. Geol. 
Surv. China, Ser. A, No. 3. 

2A note recently published, entitled ‘‘Biological and Palwontological Collecting in Northern 
China,” in the China Journal of Science and Arts, II, 1924, pp. 72-73, states, ‘‘The most recent 
expedition carried out by the two scientists [refering to Pére Emile Licent and Pére Teilhard de Chardin] 
resulted in the discovery in the valley of Shara Ossa Gol, a river in Mongolia to the north of Kansu, 
of a rich find of fossils and archeological specimens in deposits running in a complete series from the 
Miocene of the Upper Tertiary to recent strata of Neolithic age.” 

3Toll, Baron E. W. 1900. ‘‘Carte géologique du Nord de la Sibérie Orientale,’’ reproduced in 
Petermann’s geographische Mitteilungen, XLVI, Pl. xt, quoted from Suess, 1902. ‘‘La Face de la 
Merres:” LUD lip. 29) 


1924] Berkey and Morris, Basin Structures in Mongolia 127 


It seems to follow, independently of the inferences just offered, that 
the weight of the sediment could have had no causal relation to the 
depression of the earth’s crust. The fact that areas as large as 10,000 
square miles in extent are floored by thin formations, as in the case of the 
Irdin Manha-Shara Murun basin, seems to indicate that these beds 
were laid down upon a relatively flat surface that did not tend to sink 
under the load. The crust evidently supported these and other formations 
while they were peneplaned. The peneplanations are taken to imply 
great stability and rigidity of the bedrock floor,—that is, the crust 
neither sank under the load sufficiently to deform the sediments and 
encourage thicker deposits, nor rose under the negative load, when the 
region was eroded to a peneplane. In a paper to be offered later, a re- 
view of the diastrophic periods in Mongolia will be attempted, but at 
present it is enough to say that we believe that isostatic balance can be 
overset only by very large positive and negative loads, and that the 
stripping and loading observed in Mongolia have been of an order of 
magnitude too small to overcome the inertia of the earth’s crust, even 
where fault lines exist. In the dynamics of the region, the sediments 
played a passive part. Basin-warping and mountain-building were 
complementary parts of a great orogenic movement or succession of 
movements, and were controlled, we believe, by deep-seated causes. 


+ 
tye 
* 
¥ 
. 


pe Aan 
a> < 


ee s 
ue F 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 135 Tue AmertcaN Museum or Naturat History QOectober 14, 1924 


New York City 
5a. 1,8 (51.0) 
STRUCTURAL ELEMENTS OF THE OLDROCK FLOOR OF THE 
GOBI REGION! 


By CHARLES P. BERKEY AND FREDERICK K. Morris 


INTRODUCTORY STATEMENT 

Reference has been made in several preceding articles on the Gobi 
region to the ancient rock floor, on which the so-called later sediments — 
were laid down. The break between these two groups of formations 
separates comparatively simple strata above from very complexly inter- 
related series of much more ancient and obscure formations below. In 
the structural units of this complex floor is recorded all of the pre-Creta- 
ceous history that is now readable in this portion of the Asiatic continent. 
The major structural elements of it have been indicated without special 
comment, in connection with other problems, and it has already been 
pointed out that a wide range of geologic time is represented, stretching 
from the Jurassic back to very early pre-Cambrian time. Nowhere, 
however, has there been any adequate description defining the chief field 
units. This paper is directed particularly to these elements of the ancient 
floor. 


TWO MAJOR DIVISIONS OF ROCK FORMATIONS 

Very large areas in the Gobi region are covered with younger sedi- 
ments that lie nearly flat. The strata themselves are simple and, 
wherever they are disturbed, the deformation is of comparatively simple 
type also,—either gentle warping or, somewhat more rarely, sharp flex- 
ure and actual normal faulting. 

In all other areas, much more complex rock formations are exposed, 
representing a more ancient floor which is doubtless continuous beneath 
all of the sediments. Wherever the old-floor rocks are encountered, the 
type of deformation and the degree of internal modification exhibited 
by them are very different from those of the simpler overlying strata. 
Everywhere these floor rocks are folded, often are cut by igneous intru- 
sives, and to a marked degree are metamorphosed. These features are, 
of course, more pronounced in the older members. 

Wherever the rocks of these two very different types of formations, 
the sedimentary cover and the floor, are seen in contact, or where their 


1Publications of the Asiatic Expeditions of the American Museum of Natural History, Contribu- 
tion No. 20Ts» « 


e> 


2 AMERICAN MUSEUM NOVITATES [No. 135 
structural relations can be determined, a very great unconformity is 
found between them. The hiatus is so great that mountain-folding and - 
erosion of thousands of feet of material were accomplished before the 
first basin sediments were laid down. Furthermore, it appears that 
during this interval an entire change in the diastrophic habit of north 
central Asia came about. Mountain-folding characterized the deforma- 
tions that took place before that time, whereas warping and block fault- 
ing, without mountain-folding, characterized subsequent epochs. 

Late Mesozoic and Tertiary continental sediments carrying a re- 
markable new fauna constitute the formations developed above the 
unconformity. The rocks below, representing together all the ages pre- 
ceding the Lower Cretaceous, form the floor immediately beneath these 
sediments wherever they occur, and form the present surface in other 
parts of the region. The overlying sediments are at best but a thin 
veneer with many interruptions, and the dominant structural foundation 
for the whole of Mongolia is the ancient rock series below the great 
Mesozoic unconformity. 

Traces of the peneplane developed at that time still form major 
elements of the topography, and surprisingly large tracts of this old 
floor are to-day entirely bare. Many of these bare areas have been 
covered at one time or another by later sediments, only to be denuded 
subsequently to some of the minor warpings of Tertiary time. In the 
more elevated areas, of course, agents of erosion working toward a new 
level have dissected this old surface, leaving only upland remnants of 
the former peneplane, whereas in well-protected areas little change has 
been effected in all the intervening time. 


SUBDIVISION OF THE ANCIENT FLOOR 

From the great variety of rocks noted as belonging to this ancient 
floor, and the very great differences of physical condition represented by 
them, it is evident that this floor is of compound make-up. It has been 
possible to distinguish several sharply defined series of sedimentary 
strata, other more obscure metamorphosed formations and definite 
igneous units. These have, in some cases, prominent structural breaks 
between them, or have structural relations characteristic of important 
differences either in age or in origin. Some are strictly igneous types of 
large extent and evident structural importance; some, on the other hand, 
are profoundly metamorphosed and have taken on all the complexities 
usually characterizing the crystalline gneisses and schists of very ancient 
time; still others are only moderately affected by such internal and modi- 


1924] ELEMENTS OF OLDROCK FLOOR OF GOBI REGION 3 


My a 


7; ss : Sage ; 
ay en ; 
Ca ies | ae 
Bodo A wd “lt 
as ae 
; fi 


we ass 
; Iren Dabast'o 
\ 
d anal 
eee Obo Fw Rcd 
Tar Sa Pang Kiang 
P, \ 
Mountainous areas 
Jurassic porphyry 
and conglomer ate! 


Late Paleozoic 


eat on olian nye “nm 
Bafyith y lt el 
Late Proterozoic 
Ginian of Grabau) 


Fenty, Proterozoic 


HMMM, -, Me 
d 


and ‘Archaeozoic 
100 so oO 100 200 Miles 300 


Fig. 1. Sketch map of central Mongolia, showing general geology of the rock 
floor along the route traversed by the Expedition. 


Mountain areas are shaded by coarse slanting lines. An attempt has been made to distinguish the 
major groups of floor rocks along the route by different patterns. The younger basin sediments along 
the same lines of march are left white. 


fying processes, and consequently are regarded as of much later age, 
corresponding in some degree to their greater simplicity. 

At least six great groups are thus distinguished, some of which are 
capable of additional subdivision (Fig. 1). This is true particularly of the 
lowest, most ancient one, where for present purposes all the strongly 
metamorphosed units are grouped together. These major groups, 
in descending order, are as follows: 

6. Mesozoic porphyry intrusives, cutting all formations up to and 
including the sedimentary series involved in the last folding of the region 
previous to the development of the great Mesozoic unconformity. 

5. A great series of folded conglomerates and sandstones of con- 
tinental type, considered to be of Jurassic age. 

4. Strongly folded, fossiliferous Paleozoic strata of marine origin. 

3. An extensive underlying and invading mass of granite, described 
as the Great Mongolian Bathylith. 


4 AMERICAN MUSEUM NOVITATES [No. 1385 


2. A very thick and widely extended series of folded, unfossilifer- 
ous graywackes and slates, older then the granite bathylith and only 
moderately metamorphosed. We have called this the Khangai series, 
and consider it to be of late pre-Cambrian age. 

1. Still more ancient, underlying complex groups of quartzites, 
slates, phyllites, schists, gneisses, crystalline limestones and other 
associated metamorphic rocks. This complex is made up undoubtedly of 
more than one series. The upper members are judged to belong to the 
division distinguished in China as the Wu T’ai system, and the oldest 
members are regarded as local representatives of the T’ai Shan complex. 


Mesozoic INTRUSIVES 
A very great variety of porphyries, occurring both in dikes and in 

irregular forms of much larger extent, have been seen at many places. 
Typically, they are associated with and cut the latest sedimentary series 
preceding the great unconformity. This clearly establishes the fact that 
they are the youngest of the formations of the ancient floor. The strata 
above the unconformity are now regarded as of earliest Lower Cretaceous 
age, whereas those immediately below are judged on rather obscure 
grounds to be Jurassic. These intrusives, therefore, must also be of 
Jurassic age. 

SE NW. 


4085 4 Ss 2 ad 400 9 8 fe 6 Sos 
A4550'AT. MILES 


4380° 


(PRG: Gi orgs) tava gare, 
Bs Rs ae 


Complex of eruptive porphyries Bathylithic granite 


Fig. 2. Reproduction of ten miles of structure-section from the geologist’s field 
notebook, 70 to 80 miles southeast of Sair Usu on the Uliassutai trail. 


Chiefly Mesozoic rocks. Folded Jurassic conglomerates are seen *t the left, cut by dikes. A 
complex of eruptive porphyries, probably also Jurassic, occupies the center. Both rest unconforma! ly 
upon the Great Mongolian Bathylith, which occupies the right-hand end of the diagram. The extreme 
simplicity of the desert peneplane is well shown, contrasting with the complex underground structure. 


Representatives of this group are very widely distributed, and in 
places exhibit a formidable complexity of relations. So many different 
units are represented, and they cut one another in so irregular a way, 
that in certain areas they form a veritable igneous complex (Fig. 2). 

The commonest type is an acid porphyry, ranging in minor character 
from that of a simple quartz porphyry to granophyre and granite por- 
phyry of comparatively massive habit. Intermediate composition ig 
common also, and occasionally there are more basic types, so that the 


1924] ELEMENTS OF OLDROCK FLOOR OF GOBI REGION 5 


compositional and structural range is very great indeed. The most 
constant characters are fine grain, dense texture, only moderately 
porphyritic habit. These rocks are brittle and exhibit a very broken 
condition, due apparently to their deformation. This physical condition, 
together with the great irregularity of form and occurrence as part of a 
confused complex, is not so strikingly exhibited by any other series of 
rocks. 

Wherever such an igneous complex intrudes the Jurassic strata, the 
original sediments are entirely displaced and none of the original struc- 
tural trend is preserved. Areas represented by such rocks, observed at 
several points, cover many square miles. The best examples are those 
seen at Tsetsenwan, at Sain Noin, in the Mt. Uskuk district, in the 
Artsa Bogdo range, on the Sair Usu trail east of the Ongin Gol, and on 
the trail southeast of Sair Usu. It is worth noting that igneous activity 
of a somewhat similar sort is prominent in China also, in association with 
exactly the same sorts of sedimentary formations. The type is constant 
enoughin character, no matter how widely the occurrences are separated, to 
warrant the belief that some very widespread general source for these intru- 
sions must have existed, operating under regional rather than very local 
control. We are inclined to the belief that the active history of the great 
granite bathylith, in spite of its much greater age, is In some way tied up 
with the genesis of the Mesozoic intrusions. There is a certain difference 
of habit, one age after another, that makes the whole lot look like a 
genetic succession, as if they all, from beginning to end, represented only 
stages in the active history of a single great, slowly differentiating and 
repeatedly rejuvenated bathylithic mass. Perhaps these peculiar por- 
phyries are only the normal product of a particular stage from this master 
source (Fig. 7). 


JcuRAssIc SEDIMENTS 

The youngest of the sedimentary groups forming the old floor is a 
great series of conglomerates and sandstones of continental type, simply 
folded or sometimes block-faulted, and quite free from important meta- 
morphism. A great proportion. of the material is coarse-grained, and 
considerable thicknesses are strictly conglomerates. Other great thick- 
nesses are simple, well-bedded sandstones. Occasionally interbedded 
finer sandstones occur in considerable prominence, but as far as noted 
there are no large developments of shale or limestones in this series. 
Nowhere is there any evidence of marine conditions. The entire series" 
consists of stream deposits. The only fossils seen are plant remains, 


6 AMERICAN MUSEUM NOVITATES [No. 135 


chiefly stems, very poorly preserved. In certain portions of the series, 
however, there are thin beds of coal of very low grade. Even in these 
beds, the original fossil forms are poorly, preserved, and are nearly de- 
stroyed by deformation, so that the fossil content has proved thus far 
to be quite inadequate to determine the age of the beds (Fig. 3). 

The material of these sediments is largely quartzose or at least very 
siliceous, and the forms of the fragments indicate much wear. The three 
striking features are the enormous thickness of the series, its wide distri- 
bution and the abundance of quartz pebbles and grains. 

These rocks were found at several widely separated localities, the 
principal ones being at ‘‘Camp Jurassic,’’ 50 miles north of Ude; at 
Tsetsenwan, 125 miles west of Urga; at Sain Noin, 300 miles west of 
Urga; in the Mt. Uskuk region, 40 miles north of the Altai Mountains; 
SE 


ess s 2 1 250 =) 8 cs Ss 24s 
MILES 


Jurassic conglomerates §100' 


4800'AT. 
7 D 


PS 
Eruptive porphyries Bathylithic granite 
Fig. 8. Typical section of rock floor. 


One of the important rock-floor formations is the Jurassic conglomerates. This series itself lies 
unconformably on other older groups, and is commonly cut by a great complex of intrusive porphyries. 
The section used here is a reproduction of ten miles of field traverse on the Uliassutai trail, 80 miles 
northwest of Sair Usu. 


in the Artsa Bogdo range; on the trail midway between Sair Usu and 
Ardyn Obo, and at a few other spots where field relations and evidence 
were too obscure to determine the extent and local importance. As a 
matter of fact, there is evidence that strata of this series formerly ex- 
tended over a much greater area than that covered by the Expedition. 
Almost everywhere these strata stand on edge, or at least are 
strongly folded,—more seldom they are mashed and faulted, while occa- 
sional synclinal remnants show little disturbance. The total thickness in 
the district where this point could be best determined, is no less than 
25,000 feet, and at several other places great thickness is indicated, 
although estimates were not made. Doubtless these strata form the 
floor beneath the covering of simple sediments at many places, but it 
appears that erosion has cut so deep into the geologic structure of that 
time that only the lowest portions of the synclinal folds and the bases of 
fault blocks are preserved. South of Tsetsenwan, we observed a vast 
series of surface voleanic rocks,—flows, ash and tuff beds, all of which 
shared the deformation of the Jurassic conglomerates. They range, like 


1924] ELEMENTS OF OLDROCK FLOOR OF GOBI REGI ON 7 


the intrusive rocks, from rhyolitic through andesitic types, though basalts 
are present in minor proportion. They are regarded as a surface-flow 
expression of the same magmas that penetrated the Jurassic, and are 
now found as intrusive bodies cutting the conglomerate series. 

In age the whole series is pre-Cretaceous and precedes the general 
peneplanation. On the other hand, the series lies above another uncon- 
formity, the exact position of which in the geologic scale is undetermined, 
except that it lies above the latest Paleozoic sediments. These strata, 
therefore, are apparently mid-Mesozoic, and are in all essential respects 
analogous to and in many important features similar to the Lower 
Jurassic strata of China proper. The fossil evidence for age determina- 
tion is Inadequate, but there are enough points of similarity in type of 
strata, character of content and deformation history to warrant tenta- 
tive assignment to the same age. No fossils other than plant remains are 
found in either. On this basis we are referring to this series as wholly of 
Jurassic age, although there is no good evidence against the presence of 
representatives of the Triassic also. In any case, the series must be re- 
garded as a unit, in which the only breaks of consequence are those 
marked by the great igneous intrusions, described under the preceding 
heading. 

These intrusions occur at so many places, where they are associated 
with the Jurassic sediments directly, that one is impressed with the neces- 
sity of accounting in some way for this close association. It may very 
well be that the deformation that accomplished the foldings and faultings 
of Jurassic time was connected with and occasioned by igneous activities 
in the depths beneath, one expression of which is marked by these intru- 
sives. Down-faulting blocks, therefore, or very deep down-foldings may 
mark the places of weakness which guided the outbreak, and thus 
both the sedimentary remnants and the associated intrusives are now 
preserved together. Other higher portions were more successfully re- 
moved by erosion. 


PaLEozorc STRATA 

It is a most striking fact that all the sediments thus far found in 
central Mongolia, from the present back to the break at the close of 
Paleozoic time, are of continental type. But beneath the unconformity, 
at the base of the so-called Jurassic sediments, there is a great series of 
strata of marine origin, carrying abundant and characteristic fossils. 
These beds include basal sandstones of only moderate development, with 
much greater thicknesses of limestones and shales preserved in down- 


8 AMERICAN MUSEUM NOVITATES [No. 135 


folded remnants. Several thousand feet in thickness have been seen, but 
the actual total or maximum thickness is unknown. 

Strata of this age have been found in only two areas, both southeast 
of Sair Usu. Undoubtedly they constitute an eastward extension of the 
ancient folded Altaides of Suess, but the Tertiary faulting that has 
raised the modern Altai ranges did not extend into this portion of the 
desert; so that the representatives of the folded Paleozoic strata have 
not been uplifted, and are still simply part of the peneplaned floor. 

The larger proportion of the strata is judged by Dr. A. W. Grabau, 
from fossil collections made during the first season of the Expedition’s 
work, to be of Permian age, but there is a continuation downward into a 
still earlier period, at least into the Pennsylvanian. Curiously enough, no 
representatives of these strata were seen in the Altai Mountains them- 


SE NW 


YU 


. Shale +limest 
oO rT 2 3 a MILES 5 


Fig. 4. Cross section of an area of marine Permian beds, noted on the Ulias- 


sutai trail, about 120 miles southwest of Iren Dabasu. 


The complexly folded and faulted sandstones, shales and limestones now occupy a graben between 
a broad area of granite hills on the southeast and a broad area of the Khangai graywackes and other 
old rocks on the northeast. 


selves, either in the Artsa Bogdo or the Baga Bogdo districts, although a 
single hand specimen carrying fossils of Paleozoic age was found loose on 
the northerly flanks of the Gurbun Saikhan. From this it seems probable 
that representatives of Paleozoic age are to be found somewhere to the 
south in that region. Probably marine Paleozoic strata were formerly 
extensively distributed in this central Asiatic region, but the early Meso- 
zoic or mid-Mesozoic epochs of diastrophism and erosion wrought such 
havoc that now only a few remnants are preserved (Fig. 4). 

These rocks are all closely folded, and, although considerably de- 
formed, the fossil content is fairly well preserved. The strike is nearly 
east and west, conforming in this respect to the average structural trend 
of the other elements of the ancient floor. Nowhere have we seen the 
exact relations between this series and the Jurassic above or the gray- 
wackes below. But the relative position in the scale can be inferred, and 


1924] ELEMENTS OF OLDROCK FLOOR OF GOBI REGION 9 


the general nature of the relation is reasonably well determined by differ- 
ences of structural habit and physical condition. There is clearly an 
important break indicated between the Paleozoic sediments and the 
Jurassic series, since these strata are marine, whereas the Jurassic beds 
are strictly continental. 

It is particularly disappointing that these Paleozoic beds have not 
been seen in direct sedimentary contact with the graywacke-slate series 
beneath. This leaves some uncertainty about structural relation and 
relative age. Itis clear that the graywackes are older, and, in view of the 
fact that they are unfossiliferous, somewhat more metamorphosed and of 
an entirely different petrographic habit, we are inclined to believe that. 
the graywackes are much older and are probably separated from the 
Paleozoic strata by an unconformity as pronounced as either of those 
above. There is abundant evidence that the Great Mongolian Bathylith, 
described as the next unit under the following heading, is later than the 
graywacke series, and is very much older than the Jurassic sediments, 
which in some places lie on an erosion floor of granite (Fig. 3). Butitisnot 
entirely clear, from any relation yet observed, whether the Paleozoic. 
sediments are younger or older than the maximum invasion stage of the 
bathylith. All the early and mid-Paleozoic strata are missing, so that 
there are no representatives yet found from Cambrian to Mississippian 
time. Apparently the Paleozoic era is the most defective one, as indi- 
cated by the few sedimentary remnants still preserved. The Paleozoic 
rocks mark a transient marine history between two very long epochs of 
continental control. 


Tue Great MonGouian BaTHYLITH 

Between the sedimentary series just described and the older ones to 
follow, there developed in central Asia a great granite bathylith, expos- 
ures of which can be seen at various places over a very large territory. 
The formations existing at that time, including the Archean crystalline 
rocks and the graywacke-slate series, are invaded by the granite, and at 
many places where subsequent erosion has been deep enough, remnants 
of these earlier formations are preserved as roof pendants. The granite 
appears as large areas of massive rock, and also as smaller intrusive 
masses, even dikes, which cut all the formations up to and including the 
graywackes. It is not so clear what its relations are to the Paleozoic 
series, but in one place the granite appeared to be faulted against the 
Paleozoic strata (Fig. 4, Oshigo Ola). As the Permian beds near the 
contact are not metamorphosed, and as no granite dikes are seen 


10 ; AMERICAN MUSEUM NOVITATES [No. 135 


cutting them, it seems fair to infer that these Permian sediments are 
younger than the granite. 

It is possible, of course, that even the later intrusives, such as those 
which cut the Jurassic series, are products of the same great bathylithic 
magma, but if so, they belong to a much later stage in its own develop- 
ment than that represented by the great areas of true granite. That 
stage, the stage of massive granite solidification which was also the stage 
of maximum invasion, is probably pre-Pennsylvanian, and certainly 
later than the Khangai graywacke. It is entirely possible that every. 
igneous unit in the region, no matter what its age, is genetically connected 
with this immense bathylith. Its early developmental stages may have 
been responsible for the injection phenomena of the ancient gneisses; 
its maximum encroachment was attained in Paleozoic time, and its old 
age rejuvenations may in this view be recorded in the outbreaks of later 
periods (Fig. 7). 

Tice 8 7 6 5 4 3 2 1 770 Rh 
SW Folded Graywacke-slate series . 5445’ NE 
LLY 


1: epee WIR 
iC NV OO 


ill eal alae 


Great Mongolian Granite Bathylith 


Fig. 5. Typical section of the Khangai graywacke series. This section is from 
Five-Antelope Camp, 100 miles southwest of Urga. 


The simply folded graywacke-slate series is undercut and penetrated by the Great Mongolian 
Bathylith, and locally metamorphosed by contact influence over a very wide territory. This section 
reproduces a ten-mile stretch along the route followed by the Expedition within the northern mountain 
area. 


These granites show considerable variety of composition and minor 
habit, but the dominant type is a biotite granite of medium coarse tex- 
ture and massive structure. It has produced an extraordinary variety of 
end-product effects, and considerable contact metamorphism. Its re- 
lations and distribution and special features are made the subject of’a 
separate paper already published.! 


THe KHANGAI GRAYWACKE SERIES 
A very extensive series of graywacke sandstones and interbedded 
shale or slate rocks is widely distributed in Mongolia. This series forms 
the major composition of the mountains of the Arctic divide, and con- 
stitutes the country rock of the Urga-Tola River-Tsetsenwan region, as 
well as the Khangai mountain range through the province of Sain Noin 


1Berkey, Charles P., and Morris, Frederick K. 1924. ‘‘The Great Bathylith of Central Mongolia.” 
Amer. Mus. Novitates, No. 119. 


1924] ELEMENTS OF OLDROCK FLOOR OF GOBI REGION 11 


toward Uliassutai. These rocks are particularly well exhibited in the 
Gurbun Saikhan and in the small mountain tract 20 miles east of the Mt. 
Uskuk block, as well as in the Ude region along the Urga trail. Repre- 
sentatives of the same formation are found in many other areas. 

In the Tola River region and at Tsetsenwan and westward, gray- 
wackes dominate, whereas southward and eastward a greater proportion 
of shales is interbedded. Neither the top nor the bottom of the series 
has been determined, but it is certain that it is of very ereat thickness,— 
probably at least 20,000 feet. 

As far as observed, these strata are unfossiliferous throughout. 
Diligent search was made for possible fossil content, and, in the general 
Uskuk region, slates were found with obscure markings that are believed 
to represent imprints of some simple organic form, probably alge. A 
striking thing, of course, is the unfossiliferous nature of these rocks in 
spite of their simple sedimentation structure, in strong contrast to the 
richly fossiliferous habit of the Upper Paleozoic strata. It is not believed 
possible that these graywackes and slates, with their splendid bedding 
structure and abundance of original shales, could be of Paleozoic age 
without bearing better evidence in the form of fossil content. 

Siliceous limestones, of dark gray to blue color; are found associated 
with slates in some localities, and these also are quite without fossils so 
far as we have seen. Probably these limestones represent incursions of a 
shallow sea, and the very fact that they lack fossils so completely suggests 
that they are pre-Cambrian. The graywackes are probably non-marine 
and of the same age, and, especially in the Khangai Mountains and the 
region about Urga, where limestones are quite lacking, it is believed that 
this great series is essentially continental. 

Everywhere the series is folded. In places where it is made up 
largely of original shales, there is much internal deformation, so that 
ty,pical slates have been formed; but wherever the much more massive 
graywackes make up the formation, simple folding is more common, 
with very little internal deformation or meta-structure. 

The series was invaded by the granite bathylith subsequent to its 
folding, and in many places over extensive areas this rock now forms the 
only roof. Great numbers of dikes cut through the series, and the 
bathylith itself is uneven enough, so that with later erosion patches of 
granite are exposed, alternating with patches of graywacke. In places 
where the granite lies close beneath, there is considerable contact meta- 
morphic effect produced on certain qualities of the graywacke-slate 
series. In some places a crystalline condition and moderately schistose 


12 AMERICAN MUSEUM NOVITATES [No. 135 


structure are thus produced, whereas normally the rock is eminently 
granular and not schistose at all (Fig. 5). 

Other observers have noted graywacke series in regions beyond the 
reach of the traverse of the Third Asiatic Expedition, especially in districts 
to the north. Some Russian geologists have classified graywackes in 
Siberia as of Devonian age. It does not appear, however, that there is 
sufficient. reason to follow this classification for the Khangai series of 
Mongolia. A graywacke found north of Urga by J. Morgan Clements! 
is regarded by him as of pre-Cambrian age. This may correlate with the 
Khangai series of the Third Asiatic Expedition. Graywackes and slates 
also are mentioned in the region very far to the south by other observers, 
and again with suggestion of different age, but it is not certain, of course, 
that the same formation is referred to. 


SE 
seo 9 8 7 6 NS) 4 3 2 
MILES 
AA20'AT Granite r ir r TE 
TYR PUT eke MRO THI 
Wie tse bcd seebicdy bites mee ney Vy i} 
' Me f y la ; 4). i 4 A XV 4 (i WHE tees 


AELLEL ELH 


A complex of schists, injection gneisses and crystalline limestones cut by granite 


Fig. 6. Reproduction of ten miles of structure section from the geologist’s 
field notebook, 9 to 19 miles northwest of Sair Usu on the Uliassutai trail. 


The figure shows pre-Cambrian rocks of very complex structure. These have been peneplaned and 
two shallow basins of the later sediments overlie them. The peneplane is very well shown. 


Tue ANCIENT CRYSTALLINE COMPLEX 

Clearly older than the graywacke series, as indicated not only by 
their structural relations but also by the much greater metamorphic 
modification, is a great group of formations which doubtless includes 
several separable series, but which together may be conveniently referred 
to as the ancient crystalline complex (Fig. 6). The simplest of the rocks 
of this class are slates, phyllites, schists, limestones and conglomerates 
that are clearly derived frorh some ancient sedimentary series, of much 
more variable habit and somewhat different origin from that of the over- 
lying graywacke series. They stand everywhere on edge; they are re- 
peatedly exposed in many places as widely separated as are the observa- 
tions of the Expedition. The tracts are extensive, however, at but few 
places, and there are but few of these where many of the members are 
exposed together. 


1Clements, J. Morgan. 1922. ‘‘Gold Placer Area in Mongolia, China.’”’ Department of Com- 
merce, Trade Information Bulletin No. 4, Far Eastern Division. 


Basin sediments, Lower Cretaceous to 
Recent, unconformable upon the 


OLDROCK FLOOR_ 
“Mezozoic porphyries and surface 


Slows, tuffs ané ashes 


Jurassic series of continéntal sedi- 


ments cut by porphyries 


MEZOZOIC 


SSSS 
AU $ 


cS 


Paleozoic series of marine sediments, 
shales, [imestones,and sandstones, cut by 
dikes. Alf Later intrusive rocks may pos- 
sibly be derived from the bathylith, by re- 
juvenation and differentiation in ils deep- 
er parts 


AAS 
—s : 
IN2 AE 


Af 
= — a 
a ae 
ee 
. 


a 7 
NAD, 


PALEOZOIC 


Ike Great Mongolian Bathylith — 

a subjacent body, whose deplh end dif 
ferentiation are suggested by the right- 
hand portion of the diagram 


PALEOZOIC 


‘Khangai series of graywackes and slates, V7 


Jaspers, quartzites 


S 3 
ey ig 
es ia 
ae 
hy © 
ae 
ees 
eS 
ee 
fos 
<2 
y) 


The Wu Sai system, phyllites, schists, green- 


stones and crystalline limestones 


PROTEROZOIC 


The Jai Shan complex — gneisses, schists 
and crystalline Limeslones, cut and satur- 
ated and even replaced by intrusives 


ARCHAOZOIC 


14 AMERICAN MUSEUM NOVITATES [No. 135 


The Wu T’ai System 


Because of the fact that in certain places conglomerates are found 
in the midst of these series, and on account also of an observed difference 
in the degree of metamorphic complexity of certain members, we think 
the evidence favors the recognition of at least two great systems of pre- 
Cambrian rocks, older than the rocks of the Khangai graywacke series. 

If we are right in regarding the graywackes as pre-Cambrian, they 
must be late pre-Cambrian, and as such they should correspond approxi- 
mately to the Nan K’ou series of von Richthofen and Bailey Willis, in 
China, or the Sinian system of Grabau. Our next older series in Mongolia 
may include the greenish chloritic schists in the Artsa Bogdo range of the 
Altai Mountains. There is also a vast series of greenstones and chloritic 
phyllites in the mountains east of Ardyn Obo, which may be of igneous 
origin, and perhaps represent ash beds and surface flows that have 
undergone thorough reorganization. More data bearing on this problem 
may be expected when these rocks are studied with the microscope. Thin 
beds of limestone were found in the greenstone area. Mica schists and 
mica phyllites were seen on the Kalgan-Urga trail north of P’ang Kiang, 
and crystalline limestones are associated with these. The phyllite- 
schist-limestone-greenstone group occurs at many places, yet is nowhere 
as thoroughly pierced and saturated by invading igneous material as is 
the group of rocks next to be described. No better guide for these still 
older rocks is available than that of Bailey Willis as given in his “Re- 
search in China,” where he classifies a series of phyllites, limestones, 
-quartzites, schists and greenstones under the Wu T’ai system, and places 
this system as Early Proterozoic. We see no better classification for 
these very similar rocks in the Gobi region. 


The T’ai Shan Complex 


Still more complexly modified rocks are found in the Gobi region. 
They are largely gneisses and associated schists and crystalline limestones. 
The gneisses range from granitic to dioritic in general composition, but 
they are not simply granites and diorites. They represent a complex in 
which the original rock probably was a schist. This original rock has 
been invaded by igneous material which has penetrated and saturated 
the original, following in the main the structural lines of the schist. 
The magma has replaced as well as penetrated the host rock, so that now 
the igneous matter is a streaked gneiss because it has inherited the struc- 
ture of the schist which it has largely destroyed and replaced. Where the 


ST 


sasstous-uorjoofur xefduroo eulyy S. E 
puB s}sItos ‘souoysouNT] OUTT[eISAID, UI STTTIAA AQ pasn se xofduroy ueyg Ie, 7, 8 8 
ie Bote io 
= ie) 

- souoysusers ‘soqyizjaeNnb ‘soyrTuL | tS 

: ; & A & 
-ofop ‘SoU0}SOULI] ‘sop Ayd ‘SysTyog BUIYL) UI STIM AQ posn SB Woysag Tey 1M = 7 
Eee, 5] 
iS) 
S[ISSOJ JNOYJIM Sallag NVqBiy Jo STTILM pur Usjoyyyory WOA sy = 
So}BIS puv SoyoVMALIS IOAN BIOL, Iesuvy yy Wd}SAG UBIUIG jO Wla}SAG NO, YI URN cra & 

yyypAqyeq oyluvis UBTOSUOTT UWOISBAUT oIypAYweE year 
GWLy, OOZOMIVG ATAVY ONIMAAOD ALINYOANOON /) = 
ia) 
S9}BIOUIO[SUOD ‘SazIZJIVNK ‘Sa}BIS SnosopMOqgaey BS 5 
S[ISSOF 9}VAIQOJAVAUL OTSTIOJOVIVYO YIM Somos e Sales 
-pues puv soeys ‘seuojseull] JO solies VW souoyspurg ‘soreyg ‘souoqsourry | uemM9g et ase 
a 
ALINUOANOON() 

“eulyy uroy}.100 jO dISSBINF IAMOT 0} Spuodsoii09 ATJUOIVddy “yory} Yoo} = = 
000‘0z Hoge ajoy oy} ‘[eoo AT[ROO]T puB sUIBUTEI JULT[d sINOSqo SuULAIIBO ‘soyse puB SYNy oIssBIne 4 ® 
‘SMOP{ BART Po}VIOOsse YIM ‘SoTeYys pu SaUOJspUBS ‘SO}BIOUIO[SUOD JO Soltas Vols WV 6= = 

lo) DR 

o 

N 

aqadloyq GUY ANIT SIHL MOTAG SHOOY TIV g. 


ALINUOANOON() OLOZOSA LVAND AH, 


‘100 JUBTIOUR oY} UO JUSUTIps oY} UI po}Uesoidad o1v ‘OUITY 9UBI0STO[ J 0} SNOIIVJOID JOMO'T WOT] OSB UT 
Suisuevl puv ‘spunois d1s0[o}uogTed 10 [Banjons}s UO aTqRYysMsuTysip ‘suUONBVULIOJ ATeJUUIIpES o}TUYap VATOMT, 


AYVNINOAS 


NOLIGUdXY OLLVISY GUI], THT Ad LAO GAMUOM SV NWOTO) OIDOTORY) 


po}v[aI109 IO paprArp 


-Qns OJ1oy}IY JOU ‘SUOTZVUTIOF JUOTOUR JO IOOH VW 


Poplarpqns 07104 31Y 


you ‘s}ueuTpes SuLApIIAG 


s1910][dxa Jota Aq poZzTUSOIaI 919M SUOTYVULIOF YOOI JO SUOTSLATP OMY BSOYT, 


16 AMERICAN MUSEUM NOVITATES [No. 135 


igneous streaks are less predominant, the rocks may still be classed as 
schists, and among these the commonest type is a coarse-grained musco- 
vite-biotite schist, streaked with lenses and thin sheets of granite or 
pegmatite. . The limestones are white to blue or gray crystalline marble. 

Such. rocks were especially noted in the block mountain south of 
Tsetsenwan; in another block mountain 40 miles southwest of Tsetsen- 
wan, and north of Kalgan on the road to the pass, as well as at many 
other places. 

Because of the greater complexity of structural habit, these rocks are 
regarded as still older than those we have referred to the Wu T’ai 
system, and, again following the usage established by Willis in China, 
we have chosen to regard these oldest of the formations yet seen in the 
Gobi as Archzeozoie and equivalent to the T’ai Shan of China. 


TECTONIC LINES 

All these crystalline rocks, both those of the Wu T’ai system and 
those of the T’ai Shan complex, are folded and sheared, and usually are 
found standing almost on edge with the major trend or structure running 
nearly east and west. Very rarely have there been great deviations from 
this trend, although in a minor way there is a good deal of variation. 
The average is about south 80° east and north 80° west. All the very 
ancient structural lines belong to this trend, and this is so fundamental 
in the floor structure that it affects even the later mountain ranges down 
to and including those of Jurassic time. 


MISSING PARTS OF THE COLUMN 

Undoubtedly there are great breaks between the principal members 
of the pre-Cambrian series. It is believed that the conglomerates seen 
on the Urga trail in the midst of these formations, and the conglomerates 
seen also on the Tsetsenwan-Sain Noin trail 300 miles west of Urga, 
mark some of these important breaks, but even without these there is 
sufficient evidence in the differences of the rocks themselves to warrant 
such subdivision as has been made, and such age differences as are 
indicated in the tabulation. Doubtless very much greater detail of 
formational make-up is actually exhibited than has been determined as 
yet, but the major elements of the ancient rock floor and the major 
characters of the individual unit series are reasonably satisfactorily 
represented as a working basis by the accompanying table (page 15). 


AMERICAN. MUSEUM NOVITATES 


Published by 


Number 136 Tur American Museum or Naturau History (October 16, 1924 
New York City “ 


THE PENEPLANES OF MONGOLIA! 


By Cuarues P. BerKeEY AND FREDERICK K. Morris 


This paper is another of the brief announcements that have been 
issued from time to time by the geologists of the Third Asiatic Expedi- 
tion. A larger chapter on the Physiography of Mongolia is in preparation; 
but, as references are being made in other connections to the “Gobi 
Peneplane,” the ‘“‘ Mongolian Peneplane,” etc., it seems desirable at this 
time to define these baselevel surfaces, even though adequate discussion, 


UP 


Fig. 1. Generalized map of Mongolia. 


Mountainous areas are shaded in slanting lines. Lowlands are white, and the deeper depressions 
arestippled. The Arctic divide is shown by a line of round dots. The routes of the Expedition are shown 
by heavy broken lines. Index letters along the route of the Expedition indicate the following places 
mentioned in the text: K (northwest of Peking), Kalgan; PK, P’ang Kiang; IM, Irdin Manha; AO, 
Ardyn Obo; OS, Ondai Sair. Sain Noin Khan is in the Khangai Mountains, on a river that lies south 
of the Arctic divide, about midway between the Ondai Sair and M of Mts. Uskuk Mountain, not 
shown on this map, is just west of Ondai Sair. 


especially of the problems of their mode of origin, must await the 
publication of the larger report. 

Tue MoncGouian PENEPLANE.—A clearly defined ancient erosion 
surface bevels all the mountainous areas of Mongolia. Remnants of an 


‘Publications of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No«30 


a A 
— 


2 AMERICAN MUSEUM NOVITATES [No. 136 


older unreduced upland rise above it as monadnocks, and the valleys 
carved out below the peneplane represent so mature a dissection that by 
far the major part of the old erosion plane has been destroyed. Provi- 
sionally, this old surface of mature erosion has been called the Mongolian 
peneplane, because of its very widespread distribution in Mongolia’ 


(Fig. 2). 


Fig. 2. The maturely dissected Mongolian peneplane in southern Mongolia, 
on the Kalgan-Urga trail. 


The picture is drawn from a panoramic photograph, taken at the upland level, looking south and 
southwest. 


It is possible that this peneplane is the surface which passes under- 
neath the Cretaceous sediments, which undoubtedly rest upon a mature 
surface of erosion.!. This matter will be discussed in another paragraph. 

The peneplane was first observed in the granite mountains of south- 
ern Mongolia, at an altitude of 5,300 feet (1,600 meters). But as 
we climbed toward the Arctic divide, at about 6,000 feet (1,830 meters) 
south of Urga, an old mature surface was again seen beveling the schists 


Mongolian peneplane : 
Monadnocks N 


Inner Valley 
Rock terrace Ce River 


(4350') 


SN 
IN 
Fig. 3. Diagrammatic section across the Tola River at Urga. 


The figure shows: (1) monadnocks rising above the peneplane; (2) the peneplane; (3) the broad 
rock terrace; (4) the inner valley cut below the rock terrace. 


and the younger graywackes. Northwestward, other higher ranges could 
be seen, and probably the faulted ranges of Transbaikalia are beveled by a 
peneplane, which in the faulting has been lifted to unequal heights and 
tilted at somewhat unlike angles in the several fault blocks. 

At the Tola River, near Urga, the following topographic elements 
were observed: (1) monadnocks rising above the peneplane; (2) the 


_ Berkey, C. P., and Morris, F. K. 1924. ‘‘Basin Structures in Mongolia.’”’ Bull. Amer. Mus. Nat. 
Hist., LI, p. 109. 


1924] THE PENEPLANES OF MONGOLIA 3 


peneplane; (3) a broad rock shelf or terrace within the valleys of the 
Tola and its tributaries; (4) an inner valley, sunk gorgelike below the 
rock terrace yet having a mature floor upon which the river meanders. 
These relations are graphically expressed in figure 3. 

THe KHAanGcal PENEPLANE.—When we climbed into the Khangai 
Mountains north of the great lamasery of Sain Noin Khan, we saw again 
an ancient mature upland, carved by streams and, in the highest parts, 


ST. eee = 


Fig. 4. Distant view of the Khangai peneplane. 


The figure is part of a large field sketch made near Sain Noin Khan, looking northward over two 
intervening mountain ridges to the Arctie divide, which forms the skyline. The peneplane is a broad, 
gently rolling surface, above which rise low monadnocks. Severa! glacial cirques are seen, but the 
glaciation was not severe. ’ 


Arctic 


S divide N 


{ane 
en e Pp _ 
= TTT 


Fig. 5. Diagram showing the Khangai peneplane at the north, warped down- 
ward so as to form the Mongolian peneplane in the south. 


Arcti 
Ss divide N 


Fig. 6. Diagram showing the Khangai peneplane as an independent, older 
surface than the Mongolian peneplane. 


by former glaciers. Above this peneplane rise notable monadnocks. The 
physiographic unconformity between the valleys and the ancient upland 
surface is very clearly seen in this region (Fig. 4). We were not wholly 
sure of the identity of this upland with the Mongolian peneplane, and so 
called it provisionally the Khangai peneplane. It stands at about 10,000 
feet (3,000 meters), at the place where we observed it, and so should be 
older than the Mongolian peneplane, unless these two erosion surfaces 


4 ‘AMERICAN MUSEUM NOVITATES [No. 136 


can be shown to be identical. There is a shoulder or high rock terrace 
bordering the valleys. 

Descending from the Khangai Mountains and coming southward, we 
followed the peneplane as carefully as possible, to test its relations with 
the Mongolian peneplane. The valleys broaden southward, and the 
rock benches within the valleys tend to coalesce, so that remnant hills or 
outliers of the Khangai Mountains are cut off from the long spurs and lie 


Fig. 7. Field sketch showing the dissected Gobi peneplane at a point about 30 
miles north of Baga Bogdo. 


The skyline represents the peneplane beveling upturned Tertiary sediments. The upland is ex- 
traordinarily smooth, but in the badlands formed by dissection of the Gobi peneplane, the tilted beds 
develop varying topographic expression, according to their hardness. Many of the gullies have asym- 
metrical cross section, the southward sloping wall being the longer. A lens of cemented white sandstone 
forms a small hogback, near the center of the foreground. ; 


Old rock hills PangKiang lowland 


Fig. 8. Field sketch of the Gobi peneplane at Ardyn Obo, looking eastward. 


The sketch shows: (1) the remarkably level surface of the Gobi peneplane, beveling strata that are 
sensibly horizontal; (2) the lowland of the P’ang Kiang stage at the left; (3) the badland bluffs descend- 
ing about 300 feet from the Gobi upland to the P’ang Kiang level: (4) the remarkable shortness of the 
gullies of the dissected zone, in contrast to the great area of the P’ang Kiang lowland (see figure 11). 


isolated, surrounded by the erosion lowland. Looking southward from 
such border fringes of the Khangai hills, it seemed that the new, lower 
beveling continues southward over the tops of the ranges there. Lacking 
adequate maps or surveys, it is very difficult to be wholly sure of such a 
correlation; but if our observation was not at fault, it would support the 
inference that there are two peneplane levels beveling the hard-rock 
structures of Mongolia: an older baselevel, the Khangai peneplane, and 
the younger and lower Mongolian peneplane (Fig. 6). This matter will 
be discussed more fully in a later paragraph of this paper. 


1924] THE PENEPLANES OF MONGOLIA 5 


Tue Gost PenEPLANE.—At altitudes lower than the Mongolian 
peneplane, we observed a surface of extraordinary smoothness developed 
upon the relatively soft basin sediments. It is a peneplane, not a deposi- 
tion surface, as shown by the following considerations: (1) it bevels tilted 
and faulted strata; this was notably observed in the Altai piedmont 
regions at Eastern Badlands (Fig. 7); (2) it is underlain by sediments of 
widely different age; for example, by the Eocene rocks at Irdin Manha, 
and by the Oligocene sediments of Houldjin, less than 10 miles north of 
Irdin Manha; (3) even where the strata were sensibly horizontal (Fig. 8), 
the plateau upland cannot be a depositional surface, for it is impossible 
that surfaces of Eocene, Oligocene and Pliocene deposition should to-day 
all be in the same stage of incipient dissection. 

It seems to us to be demonstrated that the Gobi upland surface is 
indeed a baselevel of erosion, or true peneplane; but whether it is 
of zolian origin or is the work of water in a past cycle of more 
humid conditions, is one of the most difficult and unsettled problems 
of the region. 

Tue P’ane Kiana Lowianp.—The Gobi peneplane is interrupted 
by innumerable undrained hollows, which range in size from about 200. 
yards to tens of miles in length, and from 20 feet to 400 feet deep. The 
larger hollows have relatively flat floors, though they are never so per- 
fectly level as the Gobi upland, and in almost all cases contain several 
shallow playas. Because the important telegraph station at P’ang 
Kiang lies in a large hollow of this type, we have assigned all these 
hollows to the “ P’ang Kiang stage”’ of dissection (Fig. 8). 

We have thus four levels to consider—the Khangai, the Mongolian, 
the Gobi and the P’ang Kiang levels. At least three of these levels are 
clearly separate, but it is not certain that all four are separate or are of 
wholly different age. The brief space allotted to this paper permits us to 
do little more than state the problems. 

RELATIONS OF THE KHANGAI AND MonGoLiAN PENEPLANES.—The 
question of the possible identity of the Khangai and Mongolian pene- 
planes is given graphically in the diagrams, figures 5 and 6. 

In figure 5 the significant events may -be summarized as follows: 
(1) there was a folded and complex mountainous oldland; (2) this moun- 
tainous country was peneplaned,—leaving residual monadnocks; (3) 
the peneplane was warped and locally faulted, especially in the Altai and 
Zabaikal regions, and was deeply dissected. In the Khangai Mountains 
the uplift was a very broad, gently sloped anticlinal warp, which was 
much higher than some of the other upward areas farther south. 


6 AMERICAN MUSEUM NOVITATES [No. 136 


Figure 6 records the following stages: (1) there was a complex and 
folded mountainous oldland; (2) this oldland was reduced to a baselevel 
or peneplane, above which stood low residual elevations or monadnocks: 
this is the Khangai peneplane; (3) the region was uplifted; (4) it was 
then subjected to erosion so prolonged that the Khangai peneplane 
was wholly destroyed over a broad area, and a new baselevel was 
achieved, south of the Khangai region. This is the Mongolian peneplane, 
above which the Khangai is itself a monadnock unit. 

The crux of this problem les in the question of the age of the 
Mongolian peneplane. The oldest basin sediments are of late Mesozoic, 
probably Lower Cretaceous age. They rest upon a peneplane which 
has been carved upon the oldrocks since the last mountain-folding. 
If this folding took place, as we believe, in Middle Jurassic time, these 
mountains must have been reduced virtually to baselevel by the begin- 
ning of the Lower Cretaceous (Comanchean). It seems very improb- 
able that two major peneplanes, the Khangai and the Mongolian, could 


Be CO Ee ee PSB Pp / a 


Fig. 9. Diagram illustrating the thesis that the warped Mongolian peneplaneis 
the surface upon which the basin sediments rest. 


have been developed in this interval. If then the Mongolian pene- 
plane is the surface on which the oldest Gobi sediments rest (Fig. 9), 
we should consider this an argument in favor of regarding the Khangai 
and Mongolian peneplanes as one and the same warped surface. But if 
the Mongolian be a much younger surface than the pre-Cretaceous pene- 
plane, as suggested in figure 10, the question of the identity of the Khan- 
gai and Mongolian peneplanes would be reopened. Even in that case, 
the following considerations are opposed to their being two separate 
stages: (1) the Mongolian peneplane is as elaborately dissected as is the 
Khangai upland; if it were so much younger, it should be less dissected ; 
(2) it can be shown that very much warping and faulting have taken 
place in Mongolia, so that an old upland peneplane might be expected to 
lie at very different levels in different parts of the country; (3) it seems 
to us that it is not logical to expect a great upland peneplane like the 
Khangai to survive the removal of the enormous quantities of hard rock, 
together with the long period of very slow decay and removal that must 


1924] THE PENEPLANES OF MONGOLIA 7 


come in the later stages of post-mature dissection, involved in the carving 
of a new peneplane, several thousand feet lower than the Khangai, over a 
large part of Mongolia. 

At first sight, this interpretation might seem comparable to the con- 
clusions of Marius R. Campbell! in the Front Range, where he finds the 
Flat Top peneplane as an older, and the Rocky Mountain peneplane as a 
younger and lower level, both beveling the same hard rocks. But the 
Flat Top peneplane is almost wholly destroyed by erosion, while the 
Rocky Mountain level is the most striking and widespread landform of 
the Front Range; this contrast in degree of dissection and destruction of 
the two peneplanes seems to be lacking in the case of the Khangai and 
Mongolian peneplanes. It would be stretching Dr. Campbell’s interpre- 
tation beyond reason, we think, to cite it as a case quite parallel to that 
of the Khangai and Mongolian peneplanes. A further discussion of the 
problem involves the relation of the Mongolian to the Gobi peneplanes. 

RELATIONS OF THE GOBI AND MONGOLIAN PENEPLANES.—The 
Mongolian peneplane always lies higher than the Gobi peneplane that 
bevels the basin-sediments. There is no doubt that the Mongolian base- 
level is much the older of the two. The questions to be solved are the 
following: Is the surface that underlies the gobi basins the downwarped 
Mongolian peneplane (Fig. 9)? If not, at what date was the Mongolian 
peneplane completed? 

In some instances the peneplane now exposed in the field can be 
shown to be a morvan, that is, a once-buried peneplane which has been 
laid bare by the stripping of the sediments which formerly covered it. 
This is the case in the peneplane at Uskuk Mountain, where early 
Tertiary sediments, and possibly Cretaceous beds as well, once covered 
what is now the top of the mountain block.? 

It is a possible interpretation, therefore, that the Mongolian pene- 
plane is the surface upon which the earliest basin sediments were laid 
down, and this was indeed the theory with which we at first worked in 
the field. 

As we obtained increasing evidence of the widely different periods of 
basin-warping, however, it was judged that the entire history was more 
complex than it seemed at first. The following considerations are some 
of the elements of the problem: Buried peneplanes or unconformities and 
disconformities are numerous and imply long gaps in the sedimentary 


poe M.R. 1922. “The Peneplanes of the Rocky Mountain Region.”’ U.S. Geol. Survey, 


Bull. Ro 
Pik G. 1 aad Neciied F.K. 1924. “Basin Structures in Mongolia.”” Bull. Amer. Mus. Nat. 
Hist., LI, p. 114, Fig. 1 


8 AMERICAN MUSEUM NOVITATES [No. 136 


record. Thus in the Altai region the section includes: (1) early Lower 
Cretaceous shales and sands, disturbed by tilting and faulting; (2) early 
Tertiary conglomerates, sands and clays, whose upper beds carry a Lower 
Oligocene fauna; (3) Lower Miocene clays, and (4) Upper Pliocene sands 
and gravels. Clearly there are gaps in the sedimentary record, and at 
least two of these gaps represent peneplanations—one at the base of the 
Cretaceous, and the other at the base of the Tertiary, when the faulted 
and tilted Mesozoic rocks were reduced to baselevel. 

The relatively thin fills of sediment, deposited at great intervals of 
time, seem to indicate that during each of these long periods in which no 
deposit was made, erosion might well have baseleveled the very mod- 
erately uplifted land from which the preceding sedimentary fill had been 

aii 


Fig. 10. Diagram illustrating the thesis that the basin sediments rest upon an 
older surface than the Mongolian peneplane. 


The stages of development indicated in this diagram are: (1) an ancient peneplane, probably pre 
Cretaceous. This was warped to form basins, and sediments were deposited in these basins; (2) the 
warped region was peneplaned, forming the Mongolian peneplane; (3) a new level surface, the Gobi 
peneplane, was beveled across the soft rocks of the basins, and, to some extent, the upstanding hard- 
rock masses have been girdled by a rock bench which is continuous with the Gobi peneplane. 


4 


Mongolian ~~~~~.. $e _peneplane 


washed. It seems to us that, in a region where one alternation of slight 
warping and quiescence followed another all through the late Mesozoic 
and Tertiary, a peneplane like the Mongolian might be made pari- 
passu with the deposition of sediment. Hence, the Mongolian peneplane 
may not be of pre-Cretaceous age, but may have been finished at a much 
later time, say in the Middle Tertiary, and may have been finished at 
different times in different parts of the country. 

The absence of Miocene sediments over most of the region studied 
by the Expedition! and over all or nearly all of Northern China? as well, 
rather suggests the Miocene as a period of very widespread erosion, in 
which the closing stages of peneplanation were completed. The great 
post-Oligocene disconformity, which is very striking everywhere except 
at the Hsanda Gol, where the break between beds of Lower Oligocene and 


el 


1Berkey, C. P., and Morris, F. K., loc. cit., p. 126. 
2Andersson, J. G. 1923. ‘‘Essays on the Cenozoic of Northern China.’’ Mem. Geol. Surv. 
China, Ser. A, No. 3. 


1924] THE PENEPLANES OF MONGOLIA Y 


those of Lower Miocene age is inconspicuous, may correspond to the 
Mongolian peneplane developed on the hard rocks. 

Summarizing these relations, it seems possible that the Mongolian 
peneplane may prove to be the surface upon which the oldest Lower 
Cretaceous sediments were laid down (Fig. 9); but if so, it could not be 
the surface upon which were deposited such sediments as the Ardyn Obo 
formation (Oligocene) and the P’ang Kiang formation (Miocene or 
later). Butit seems more likely to us that the pre-Cretaceous peneplane 
has been destroyed everywhere, except (1) where it is still covered by 
basin sediments (Fig. 10), or (2) where it has been re-exposed, in com- 
paratively recent time, by the removal of its sediments. In this case the 
Mongolian peneplane would be much younger than the pre-Cretaceous, 
and might have been completed in Tertiary time, perhaps in the Miocene. 

The conditions under which the Gobi peneplane was formed, and the 
date of its completion are questions of some difficulty. Since late Pliocene 
beds are beveled by this surface at Hung Kureh, the peneplane must have 
been formed during the Pleistocene. Its recency is also attested by its 
extraordinarily smooth surface, undissected save for the hollows that 
represent the P’ang Kiang stage of erosion. Part of the difficulty lies in 
the question, how such widespread peneplanation could have been accom- 
plished in a region as uplifted as the interior of Asia must have been dur- 
ing the Pleistocene, which was undoubtedly a period of mountain growth. 
East of Uskuk Mountain in the piedmont region north of the Altai, the 
Gobi peneplane is domed or arched, and is carved into an intricate net- 
work of gullies that in some places are well-developed badlands. Clearly 
the Gobi peneplane is here being destroyed, not formed, at the present 
time. The fresh badland bluffs bordering the P’ang Kiang lowlands in 
so many parts of Mongolia show that everywhere the Gobi peneplane is 
being destroyed rather than made in the present cycle, and that both 
wind and running water are agents of its destruction. 

RELATION BETWEEN THE GOBI PENEPLANE AND THE LOWLANDS OF 
THe P’ang Kiana Stace.—As there is no doubt that the P’ang Kiang 
hollows are younger than the Gobi peneplane, the only problem is their 
mode of origin. A paper on this subject will be offered shortly, and it 
must suffice here to give only a brief description. 

The hollows never were the beds of large lakes, though small shallow 
playa basins lie Scattered on their broad floors. There is commonly a 
shelf or terrace in the hollow, but no sign of beaches, bars, wave-cut 
cliffs or delta terraces that might indicate the former presence of large 
lakes. The bluffs that form the descent from Gobi upland to P’ang Kiang 


10 AMERICAN MUSEUM NOVITATES [No. 136 


floor may be fairly smooth, or may be carved into badlands by innumer- 
able short gullies, and these two contrasting conditions may be present 
on the same bluff within a few miles. We believe that the hollows are the 
result of the combined action of wind and running water. According to 
this view, the first stage of development was a deflation hollow, or blow- 
out, made by the wind. But when it became deep enough to reach mois- 
ture that might encourage vegetation, the rate of deepening by wind was 
decreased, while erosion of the surrounding bluffs that walled in the 
hollow was carried on, intermittently, in the rainier seasons, by short 


G 6'b9 OP Gn exp da. Pang Kiang lowland 


Playa lakes 


Fig. 11. Two diagrams illustrating the distribution of ground water in the 
sediment basins. 


The vast gravelly surface of the Gobi peneplane absorbs water readily into the porous sediments. 
In time of high groundwater level, a series of springs or seepages, or both, appears at the cliff front to 
reinforce whatever rainwash there i is. Thus a series of short, steen gullies i is developed, as shown in 
figure 8. The sediment washed out of these gullies is spread over the | bottom of the hollow and, when 
dried, is largely carried away by the wind. 


streams supported by springs and seepages. The vast area of the Gobi 
upland, with its gravel surface, readily absorbs the rain and melting 
snow, while the broad shallow fill of porous sediments acts as a reservoir 
of groundwater. When seasons of exceptional rainfall raise the water- 
table, dissection of the cliff-front becomes active through the escape of 
spring waters (Fig. 11). The washed-out material is spread in thin 
sheets upon the floor of the hollow, where, through the long dry season, it 
is subject to the work of powerful winds. The upland near the hollows, 
especially on the southern and eastern sides, is in many instances covered 
with sheets of dune sand. We believe that water is the chief agent of 
erosion of the cliff, while wind is the chief agent of complete removal of 
loose material from the floor of the hollows. 

CoMPARISON WITH OTHER ReEaions.—As yet, data are not available 
to us to show whether the Russian geologists have published papers 


‘ 


1924] THE PENEPLANES OF MONGOLIA 11 


dealing with the peneplanes of the northern country. In China, the 
classic work of Bailey Willis' and the more recent studies by Andersson? 
offer means of comparison, though as yet it is too early to seek an actual 
correlation in physiographic stages. We offer for brevity’s sake the 
stages recognized by Bailey Willis as revised in part by Andersson: 

1. The Pei T’aistage.—Early Tertiary—“ We take this broad flat form to repre 
sent a stage of erosion to advanced old age, the nearest approximation to a peneplain 
which we have found in the course of our journey.’’s 

2. “T’ang Hsien stage (Pliocene) —Deposition of gravels and clays with the 
Hipparion fauna. Landforms of advanced maturity. 

3. ‘Fen Ho stage (Early Pleistocene).—Earth movements and subsequent re- 


vival of vertical erosion. 

4. ‘Ma Lan stage (Middle Pleistocene).—Cold arid climate. Deposition of 
valley gravels and xolian loess with Elephas sp. 

5. ‘Pan Chiao stage (Late Pleistocene).—Climate semiarid, abundant summer 
rains. Dissection of the valley gravels and primary loess. Formation of redeposited 
gravels and loess, with Bos sp., Ovis sp. and Cervus sp.’ 

These may tentatively be compared with the stages recognized in 
Mongolia, as follows, placing the oldest stages at the bottom of the 


column: 


China f Mongolia 
8 Modern dissection Modern dissection 
7 Pan Chiao dissection P’ang Kiang dissection 
6 Ma Lan, loess stage 
5 San Men,® pluvial epoch aa Gobi peneplane 

gravels) 
4 Fen Ho uplift Warping and uplift . 
3 . T’ang Hsien, partial peneplanation Rock terraces 
2 Pei T’ai peneplane Mongolian peneplane 
Khangai peneplane ? 
1 Monadnocks Monadnocks 
ADDENDUM 


After this paper had been sent to press, the authors read V. A. 
Obruchev’s essay “The Gateway to China,’’* in which the great Russian 
geologist describes a very perfect upland peneplane beveling the moun- 
tains of western Dzungaria. Several photographs of this ancient sur- 
face are figured. 


1Willis, Bauey: Jeecivelies, Eliot, and Sargent, R. H. 1907. ‘‘ Research in China.” Carnegie 
Instn. Wash., 


Pe iter wp ¢ "1920, quoted by Yih, L. F. 1920. ‘‘Geology of the Western Hills of Peking.” 
Mem. Geol. Surv. China, I, pp. 65-77. 

sWillis, B., loc. cit., p. 237, and Pls. XXXII, XXXIII, XXXV and Atlas sheet E I. 

‘Andersson, eG. 1920, quoted by Yih, L. F. 1920. “Geology of the Western Hills of Peking.’’ 
Mem. Geol. Surv. China, I, p 

5Ting, V. K. 1923, eye re Andersson, J. G. 1923. ‘‘Essays on the Cenozoic of Northern 
China.’”’ Mem. cay Surv. China, Ser. A, No. 3, p. 117. 

6Obruchev, V. A. 1915. ‘‘Votora v’Kitai. Geographicheski i geologicheski ocherk pogranichnoi 
Dzungari.”’ (“The Gateway to China. Geographical oe geological sketch of the boundary of 
Dzungaria). Izvestia Imp. Russ. Geogr. Soc., LI, No. 5, pp. 277-322, with teotonic map and 5 


plates. (St. Petersburg.) 


| we ui iG, ean ne 


sgt 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 144 Tue American Museum or Natura History November 7, 1924 
New York City 


56.81,9T (117:51.7) 
THREE NEW THEROPODA, PROTOCERATOPS ZONE, 
CENTRAL MONGOLIA! 


By Henry FAIRFIELD OSBORN 


In 1923, in the red Djadochta sandstone at Shabarakh Usu, were dis- 
covered three remarkable new types of small dinosaurs related to the 
Theropoda. All three types are approximately of the same geologic age, 
namely, the life zone which we now regard as near the beginning of 
Upper Cretaceous time. The skulls are entirely dissimilar and extra- 
ordinarily interesting. 

The first (Fig. 1) of the typical megalosaurian type, although of 
small size, seems to have been an alert, swift-moving carnivorous dino- 
saur to which the generic name Velociraptor is applied. 

The second (Figs. 3, 4), although megalosaurian and provided with 
a row of teeth, was at first mistaken for the skull of a bird, owing to its 
long slender rostrum; it may prove to have avian relationships; hence 
we name it Saurornithoides, the “birdlike theropod.”’ 

The third (Figs. 6, 7) is a short skull, entirely toothless like the 
Ornithomimide, which was found lying directly over a nest of dinosaur 
eggs, separated only by four inches of friable sandstone; hence we name 
it Oviraptor, the ‘‘egg seizer.”” The fore limb found with this skull is 
clearly related to the Ornithomimide. 

The actual proportions of these three skulls are well displayed in the 
accompanying figures (Figs. 1, 3, 6 and 7) to a one-half scale. 


Velociraptor mongoliensis, new genus and species 


Type.—Amer. Mus. 6515, skull and jaws, one front claw and adjoining phalanges. 
Djadochta beds at Shabarakh Usu, August 11, 1923, Field No. 377, collector P. C. 
Kaisen. 

GENERIC CHARACTERS.—Skull and jaws of diminutive megalosaurian type. 
Cranium abbreviated; orbits greatly enlarged; face elongated; four fenestra in 
the side of the cranium, one fenestra in the jaw. Teeth recurved, serrate on one or both 
borders, alternating in replacement; 3 ? + in premaxillaries, 9 ? + in maxillaries, 
14 in dentaries. 


; Rr eabons of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 32. 


2 AMERICAN MUSEUM NOVITATES [No. 144 


Sprcrric CHARAcTERS.—Ungual phalanges very large, laterally compressed, 
strongly recurved, super-raptorial in type. 

This skull was found in the soft sandstone matrix lying along- 
side a skull of Protoceratops andrewsi. Although slender and diminutive, 
it is none the less of typical megalosaurian or theropod type, 
suggesting a family relationship to the Megalosauride. The ten or 
twelve sharply recurved teeth are strongly serrate on the posterior border, 
less serrate on the anterior; thirteen to fourteen in number in each jaw, 
somewhat homodont or similar in form, although differing in size, recur- 
vature and serration; irregular or alternating in replacement, so that the 
wide gaps between the recurved crowns are perfectly adapted to the 
sudden seizure of light and swift-moving prey; sharply compressed, 
strongly recurved phalanges facilitating the holding of the prey; the 
long rostrum and wide gape of the jaws indicate that the prey was not 
only living but of considerable size. 


3 1 todth tn 7x. 1etaoth tr 17? RX. A. 14.6515 TYPE 


Pp 
= SLL OP LO Ds 


=== 


) r. last lower tooth 3 +s 


Fig. 1. Type skull and jaws of Velociraptor mongoliensis (Amer. Mus. 6515). 


From the Djadochta formation, Protoceratops zone, central Mongolia. One-half natural size. 

Enlarged maxillary and mandibular teeth: (Above) first tooth in maxilla, double serration; 
third tooth in maxilla, single serration.. (Below) posterior tooth in dentary, single serration. 
$+ natural size. 


MEASUREMENTS 
Skull, length over all, occiput to premaxillaries............... . . 176 mm,, est. 
Facial length, anterior border of orbits to a eeilarias: Aaa bape epee lat 
Cranial length, anterior border of orbits to occiput............ ah ean oh 0134 
Miaeetoner aivial eimttercs sti hee 2 oe tart erate te exe eS a ons ious she 170% 
Mandibular length, tip of dentaries to back of surangular............ 175 


These measurements indicate that the face is relatively much longer 
(index 1.70) than the cranium, that the postorbital region of the cranium 
is extremely abbreviated, that the jaws are excessively long and slender. 
The post-temporal region is of the typical diapsid type with two fen- 
estre, supratemporal (s.t.f.) and laterotemporal (l.t.f.). There is a 


1924] THREE NEW THEROPODA, CENTRAL MONGOLIA 3 


very large antorbital fenestra (a.f.1) and a smaller antorbital fenestra 
(a.f.2). With the aid of Prof. W. K. Gregory, the following cranial 
elements have been made out: quadrate, quadratojugal, squamosal, 
parietal, postorbital, jugal, prefrontal, lachrymal, palatine, nasal, 
maxillary,. premaxillary; in the jaw, surangular, prearticular, angular, 
splenial, dentary. 


AM65IS. TYPE 


I 
7 


$ 


Fig. 2. Phalanges of Velociraptor mongoliensis associated with type skull and 


jaws (Amer. Mus. 6515). Natural size. 


By comparison with Allosawrus (cf. Gilmore, 1915), these phalanges probably belong to the enlarged 
Digit I of the manus, namely, 1 ph.,2 ph. (Upper) viewed from above. (Lower) viewed from the side. 


Saurornithoides mongoliensis, new genus and species 


Typr.—Amer. Mus. 6516, skull and jaws found in a concretion lying on the sur- 
face with bone and teeth exposed and much weatherworn; nearby in another con- 
cretion a series of vertebree with pelvis, parts of hind limb and pes which may belong 
to the same individual as the skull. Collected by the Chinese assistant, Chih, July 
9, 1923, Field No. 256, Djadochta beds at Shabarakh Usu. 

GENERIC CHARACTERS.—A diapsid and probably theropod reptile with five 
craniofacial fenestrations and one mandibular fenestration; nineteen maxillo- 
premaxillary teeth, practically homodont, fairly uniform in dental replacement; not 
of active raptorial type; teeth serrate only on posterior borders. 

Speciric CHaracrers.—Teeth flattened, serrate on posterior borders only, 
diminishing in size from the first to the fourth premaxillary tooth, increasing in size 
from the first to the tenth maxillary tooth; uniform in replacement, closely com- 
pacted below; crowns recurved, subacutely pointed. 


When first discovered, the type skull (Fig. 3) was partly covered 
with sand and largely worn by the action of the drifting sand. Its long 
pointed rostrum strongly suggested the skull of one of the toothed birds, 
but it has since proved to be on the reptilian or megalosaurian side. The 
name Saurornithoides is assigned, signifying the “‘saurian with birdlike 
rostrum.” For avian character, one naturally examines the quadrate 


4 AMERICAN MUSEUM NOVITATES [No. 144 


Al 6516 TYFE 


Fig. 3. Type cranium and jaws of Sawrornithoides mongoliensis (Amer. Mus. 
6516), Protoceratops andrewsi zone, Djadochta formation, collection of 1923. al 


figures one-half natural size. 


(Upper) right lateral view of cranium and jaws as embedded in the Djadochta sandstone 


indicated by dots. 

(Middle) palatal view of cranium, showing inferior view of dentaries, palatines, pterygoids, 
posterior nares and basioccipitals. 

(Lower) top view of cranium as embedded in the sandstone, The walls of the brain case, includ- 
ing the cerebrum and cerebellum region, are clearly indicated. 


which appears to have movable articular surfaces with the parietal above 
and quadratojugal below; but this point awaits confirmation. The 
fenestration of the skull is of the diapsid type, resembling in its five 
openings on each side the typical megalosaurian Velociraptor described 
above, namely, the supratemporal, laterotemporal (1.t.f.), antorbital 


1924] THREE NEW THEROPODA, CENTRAL MONGOLIA 5 


(a.f.1), second antorbital (a.f.2), narial (nar.); it has also apparently a 
large mandibular fenestra (m.f.) behind the dentaries. With the excep- 
tion of the fairly free borders of the quadrate (q.), the bony sutures seem 
to be chiefly closed and the following elements are made out mainly by 
their anatomical position: parietal (pa.), quadrate (q.), basioccipital 
(b.oc.), pterygoid (pt.), jugal (ju.), lachrymal (la.), palatine (pl.?), 
maxillaries (mx.), premaxillaries (pmx.), dentaries (d.). Above the 
parietals is observed an imperfect cast of the cerebrum (cb.) and of the 
cerebellum (cbl.). 


AM 6516 TYPE 


Fig. 4. Diagram of seven anterior maxillary teeth of the left side of Saurorni- 
thoides mongoliensis type, beneath the anterior nares (nar.) and the antorbital 
fenestra (a.f.2), showing deep serrations on the posterior border. Three times 
natural size. 


Denrition.—The teeth are subhomodont or of similar shape but differ widely 
from those of Velociraptor in their uniform replacement, the summits of the fifteen 
crowns of the maxillary teeth being approximately on the same level. It is difficult 
to determine the exact number or characters of the teeth, owing to the severe attri- 
tion of the dental series. There are apparently four premaxillary and fifteen maxil- 
lary teeth, nineteen superior teeth in all. This indicates that Sawrornithoides had 
different feeding habits from Velociraptor; it was less adapted to seizing alert, swift- 
moving prey; it may possibly have been an egg feeder, namely, ovivorous. We must 
await the evidence afforded by the limbs. 

The inferior view of the skull apparently displays the surface of the 
partly damaged jaws, the dentaries and the wide closure of the posterior 
nares by the union of the pterygoid plates. The backward extension of 
the posterior nares and the union of the pterygoid are frequently an 
adaptation to aquatic life; it is possible, therefore, that Sawrornithoides 
may have had an aquatic habitat, a point to which the limbs might bear 
testimony. The fragment of hind limb and left pes (Fig. 5) found in a 
concretion near the type skull is of cursorial and raptorial type. Al- 
though much fractured and weathered, the four digits exhibit the follow- 
ing characters: D.I, 2 ph., partly compressed; D.II, 3 ph., partly com- 
pressed; terminal claw 30 mm., recurved; D.III enlarged, 3+ ph., 


5 ‘|BNPIAIpUL ouTes oy} 0} Sursuopeq ATqe 
00 UOTA1OU0D OY} AvoU SULAT UOodoUOD B UT QUIT pury puw stAjod 171M ; UNOYT 


-qord pu sisuarjobuow ‘gy Jo [[nAs adAq oy} Sule} U 
‘OZIS [BANPRN 


‘outdns SutAy sod yyoT JO WOTWOg ‘(QTGQ ‘SHI, “ouLy) sisuarpobuow saproypusoinng yo od, “G “OT 


19241 THREE NEW THEROPODA, CENTRAL MONGOLIA 7 


terminal wanting; D.IV, 2+ ph., two terminals wanting; D.III appar- 
ently the largest. These appear to be of typical megalosaurian type. 
Weatherworn vertebre in concretion are of cursorial not aquatic adap- 
tation. 


MEASUREMENTS 
Length of skull, basioccipital to premaxillaries....................... 192mm. 
Length of skull, occipital to anterior border of orbits................ 67 
Length of face, anterior border of orbits to tip of premaxillaries...... 125 
este concern Lalo UIC Gare ds, tcesheise Sra Mich ta caine Stee eh aoe ae ee i eS al ou a alia espero a 187% 
Space occupied by dental series....... Re ate RE ee Roa co A ST en 73 


From the above measurements it appears that the proportions of 
the face to cranium (187%) are approximately similar to those of Velo- 
ciraptor (170%). The cranium as a whole is greatly elongated and the 
rostrum depressed toward the extremity. The apparent tapering and 
shallowness of the anterior part of the cranium are partly due to the fact 
that the lower jaws are thrust in between the upper, thus concealing the 
dental border and all the inferior teeth. 

ConcLusion.—Our conclusion, from the imperfectly preserved type 
skull and associated type hind limb, is that Sawrornithoides mongoliensis, 
despite its elongate rostrum and flattened teeth, was a small cursorial 
theropod, more sluggish than Velociraptor, which was swift and raptorial 
in habit, but remotely related to it. 


Oviraptor philoceratops, new genus and species 


Type AND LocaLity.—Djadochta beds at Shabarakh Usu, skull, jaws, cervical 
vertebre and one fore limb, collected by George Olsen, July 13, 1923, Field No. 
268 (Amer. Mus. 6517). 

GENERIC CHARACTERS.—A diapsid reptile with eight craniofacial and mandib- 
ular fenestrations. Cranium exceeding facial region in length; entirely edentulous; 
probably related to the edentulous Ornithomimidz. Skull extremely abbreviate; 
orbit and fenestrations exceptionally large. A large interclavicle. Manus tridactyl; 
metacarpals abbreviated; digits irregularly elongated as in the Ornithomimide; 
elements of digits not compressed laterally. 

Speciric CHaracters.—Faciocranial index 70%. Lower jaw with greatly 
elevated mandibular border and two mandibular fenestre; prominent bony eminence 
above the rostrum. Scapula extremely elongate. Humerus, ulna and radius sub- 
equal. Digit II extremely elongate, D.I and III relatively abbreviate. 


The unique toothless cranium, which forms the type of Oviraptor 
philoceratops, was found by Mr. George Olsen embedded in a nodule of 
reddish sand, which after careful preparation in the American Museum 
laboratory yielded the shattered appearance illustrated in figure 6. The 
upper figure shows the left side of the shattered type cranium of 
Oviraptor; the lower figure shows the right side of. this shattered 


left side 


Fig. 6. Right and left aspects of type skull of Oviraptor philoceratops (Amer. 
Mus. 6517) exactly as it lay in the matrix, the shattered and widely separated condi- 
tion of the bones as they appear embedded in the sandstone indicated by dots. One- 
half natural size. 


1924] THREE NEW THEROPODA, CENTRAL MONGOLIA 9 


cranium. Both figures correspond exactly with the type as it now appears 
in the matrix. Through very careful comparison of the bony elements 
represented on the right and left sides respectively, it proved possible to 
make the complete restoration of the right side of the cranium shown in 
figure 7. This cranium exhibits the following openings or fenestre: 
(1) Supratemporal fenestra, exposing the parietal (pa.). (2) Latero- 
temporal fenestra, between the squamosal and jugal arches (1.t.f.). 
(3) Orbit, of large size, in mid-cranial region. (4) Antorbital fenestra 
(a.o.f.1); antorbital fenestra (a.o.f.2). (5) Anterior nares (nar.). 
(6) Mandibular fenestra (md.f.1); mandibular fenestra (md.f.2). 

CHARACTERS.—This eight-fenestrated cranium is a marvel of lightly 
arched skull structure; the sutures are mostly closed; consequently the 
determination of the following cranial components by Professor Gregory 
is largely according to position and not according to sutures, namely: 

‘parietal, squamosal, opisthotic, quadratojugal, jugal, postorbital, frontal, 
maxillary, nasal. Sutural separation is indicated between the typical 
components of the lower jaw. The most remarkable feature, next to 
fenestration, lightness of the cranial arches and absence of cranial 
sutures, is the form of the skull and jaw. The cranium, vertically deep 
posteriorly, is shallow anteriorly to admit the elevated dentary portion 
of the jaw; the jaw, very shallow posteriorly, is vertically deep anteriorly, 
thus rising to oppose the shallow portion of the cranium. 

Hasirs.—The generic and specific names of this animal, Oviraptor, 
signifying the ‘“‘egg seizer,”’ philoceratops, signifying ‘fondness for cera- 
topsian eggs,’’ may entirely mislead us as to its feeding habits and belie 
its character. The names are given because the type skull (Amer. Mus. 
6517) was found lying directly over a nest of dinosaur eggs, the one photo- 
graphed being actually separated from the eggs by only four inches of 
matrix. This immediately put the animal under suspicion of having 
been overtaken by a sandstorm in the very act of robbing the dinosaur 
egg nest. 

This animal differs as widely as possible from Velociraptor and Saur- 
ornithoides, first, in the extreme abbreviation and depth of its skull, 
second, in its relatively long cranial and facial region, third, in the extra- 
ordinary fenestration of the side of the skull and jaws, there being eight 
fenestrae or openings in all, and, fourth and most important, the entire 
absence of teeth. The latter character removes it from the Megalo- 
sauridz, to which we might refer the two skulls previously described, and 
relates it to the Ornithomimide, the toothless dinosaurs, although it 
differs generically from either Ornithomimus or Struthiomimus. 


10 AMERICAN MUSEUM NOVITATES [No. 144 


MEASUREMENTS 
Length of skull, opisthotics to maxillaries....................-.00005 179 mm. 
Length of cranial region, opisthoties to antorbital bar................. 111 
Length of facial region, antorbital bar to premaxillaries............... 78 
WR eioeeamAl Mec ae tiy oe butch, Sas eles Sotto ee, ame rahe re RCE eas 70% 
Depth of skull, summit of parietals to angulars........................ 112 


These measurements and sees ieee the widest possible 
contrast to those of the typical megalosaurian skull. The abbreviation 
of the facial region is especially distinctive. It is probable that the 


Amer Mus. 65/7 TYPE 


Fig. 7. Oviraptor philoceratops skull (Amer. Mus. 6517) reconstructed from 
‘right and left sides of original specimen as shown in Fig. 6. One-half natural size. 


rostrum was sheathed in a horny beak, but there is no evidence of a 
predentary. Were it not for the evidence afforded by the fore limb 
and vertebre, we should be completely at a loss to determine the 
relationships of this animal. 

Fore Livs.—Fortunately there was found associated with the skull 
the anterior portion of the skeleton (Fig. 8), including cervical verte- 
bree (C.V.), thoracic ribs (R), interclavicle (Ic.), the greater part of the 
left fore limb and supine right manus and portion of the prone left manus. 
So far as at present excavated from the rock, these are shown in figure 8. 
The interclavicle (Ic.) lies near its union with the extremely long and 
slender scapula (Sc.). The right manus lies supine, showing the palmar 
surface of Mte. I (1) with the elongate first phalanx (1) and extremely 
long laterally compressed phalangeal claw (2); next lies Mte. IT (II) 
with three elongated phalanges (1, 2, 3), the terminal claw showing just 


Il 


‘ouoid sasuvyeyd yyIM snueur yjoy ‘ourdns sosuvyeyd yyIM snuvut yYysu= ‘TIT ‘TT ‘T ‘sniauny= pF “vpndvos= ‘ag ‘OPPTAB[OIOJUL= “OT 
9ZIS [BIN}JVU YINOFJ-9UC) *sdoyosav0pryd 10, DLC) jO snueur Vy st1 PUB Jfoy OS[B ‘quay o1OF JJoT Jo suoT1og “g “Sy 


SdAL LSINV 


SANVWA LAST a8 SANA ce 


12 AMERICAN MUSEUM NOVITATES [No. 144 


beneath the ulna; next is the more slender Mte. III (III) with four phal- 
anges (1, 2, 3,4). Fortunately in the more scattered left manus, which 
lies prone showing the dorsal surface, the second enlarged metacarpal 
(II) is preserved, also the reduced third metacarpal (III); what is pre- 
sumed to be the third phalanx of the third digit (III, 3) was exposed at 
the edge of the block; the isolated claw (II, 3) was entirely weathered. 
out and lying among the eggshell fragments on the surface. 

Comparing this manus with that of Ornitholestes Osborn, of Stru- 
thiomimus altus Osborn, and of Chirostenotes Gilmore, as figured by Gil- 
more. (1924, pp. 4 and 5), the extremely elongate second digit (II) of 
Oviraptor is analogous to that of Ornztholestes and of Chirostenotes rather 
than to that of the more symmetrical digits of Struthiomimus. On the 
contrary, the metacarpals and. phalanges of Oviraptor are relatively broad, 
as in Struthiomimus, rather than extremely compressed laterally, as in 
Ornitholestes and Chirostenotes. 


CONCLUSIONS 


The discovery of these three new Theropoda in central Mongolia is 
extremely interesting and important. It tends to establish the theory 
that Mongolia was a highly fertile center of terrestrial dinosaur life in 
Lower Cretaceous times, as well as of mammalian life in Lower Tertiary 
times. 

The three small carnivorous dinosaurs above described may be 
classified as follows: 

Family Megalosauride: 

Velociraptor mongoliensis, super-raptorial, certainly carnivorous; maxillodentary 

teeth +4; teeth serrate on both borders. 

Saurornithoides mongoliensis, raptorial, carnivorous or ovivorous; rostrum 
elongate; maxillodentary teeth +2; teeth compressed, serrate on posterior 
borders only. 

Family Ornithomimide: 

Oviraptor philoceratops, herbivorous or ovivorous; cranium abbreviate; maxillo- 

dentary borders edentulous; manus tridactyl, attenuate. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 145 Tue American Museum or NaTuraL History Nov. 10, 1924 
New York City e 


56.9,66E(1181:51.7) 


EUDIN OCERAS, UPPER EOCENE AMBLYPOD OF MONGOLIA! 
By Henry FArIrFIELD OsBoRN 


One of the most surprising and welcome discoveries of the season of 
1923 was that of two superior premolar teeth which demonstrate the pres- 
ence of archaic ungulates of the order Amblypoda in Upper Eocene time 
in Mongolia, serving as a new link with America. No Amblypoda have 
hitherto been found in Eurasia, excepting Coryphodon of the Lower 
Eocene of France and England. 

In the Uinta B beds of northern Utah, the last surviving Upper 
Eocene members of the Amblypoda referable to the genus Uintatherium 
Leidy or Eobasileus (= Loxolophodon) Cope occur in the same beds with 
titanotheres of the genera Dolichorhinus, Manteoceras and Metarhinus. 
The Irdin Manha formation of Mongolia is more recent in age than the 
Uinta B beds of Utah; it is uppermost Eocene, or Uinta C. 

In the Irdin Manha formation the presence of Amblypoda in the 
Mongolian fauna is demonstrated by the two superior premolar teeth. 
They belong to a new genus distinct from any of the known American or 
European Eocene forms hitherto described. Consequently we apply the 
new generic name Hudinoceras, the prefix “eu” signifying that the 
superior premolar teeth are more progressive than those of Marsh’s 
Dinoceras (= Uintatherium), or of Cope’s Loxolophodon, in the posses- 
sion of a prominent internal cone, no trace of which is observed in the 
Upper Eocene American genera. 

Discovery.—The first tooth (Fig. 2A, Amer. Mus. 20101) was dis- 
covered by Leader Andrews June 1, 1923, on a bench of the Irdin 
Manha formation, about two miles south of camp. On September 15, 
1923, the author joined the party and, expressing the most lively interest 
in this tooth, agreed with the other paleontologists that it was un- 
doubtedly one of the Amblypoda and that it established the presence of 
this order in Mongolia. It was at first compared with the Lower Eocene 
Coryphodon, but its closer resemblance to the Upper Eocene amblypods 
soon became apparent. The discovery was so important that Leader 


Tene of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No 


2 AMERICAN MUSEUM NOVITATES [No. 145 


Andrews was photographed on September 17 on the spot where the first 
tooth was picked up (Fig. 1). 

The author had a strong presentiment that he must stop, at a point 
eight miles to the south of Andrews’s find, and examine a small exposure 
of the Irdin Manha. He said to Mr. Andrews, “I am going to find another 
Coryphodon tooth’’; he walked to the bluff a hundred yards distant, 
traversed about seventy-five feet of the middle of the exposure, and 
there at his feet lay a second tooth of exactly the same character (Fig. 
2B, Amer. Mus. 20102), belonging to an animal of the same size but 
from the opposite side of the upper jaw! This led to the author’s being 
photographed (Fig. 1). Thus double confirmation was obtained of the 
presence of the Amblypoda in Mongolia; the explanation of this re- 
markable telepathic coincidence is left to the psychologist. 


Eudinoceras mongoliensis, new genus and species 
Typr.—A third or fourth superior premolar of the right side (Amer. Mus. 20101), 
collected by Roy C. Andrews, June 1, 1923. 
Paratype.—A third or fourth superior premolar of the left side (Amer. Mus, 
20102), collected by Henry F. Osborn, September 17, 1923. 
Horizon.—Irdin Manha formation, Protitanotherium mongoliense zone, south- 
eastern Mongolia. 


Both these specimens are from the Irdin Manha formation, Upper 
Eocene. 

Generic Cuaracters.—Eudinoceras, from the Greek ev, intensive, dew os, 
terrible, K€pas, horn. The superior premolars are intermediate in structure between 
the Coryphodon and the Dinoceras (= Uintatherium) premolars, namely, with a promi- 
nent internal cone which is present in Coryphodon and entirely lacking in Dinoceras. 
Main portion of crown yoke-crested, as in Coryphodon and Uintatherium; broad 
anterior and posterior cingula. 

Sprciric CHARACTERS.—Type (Amer. Mus. 20101), ap. 26mm., tr. 36 mm., trans- 
verse exceeding anteroposterior diameter. Length-breadth index .72. Paratype 
(Amer. Mus. 20102), ap. 27 mm. est, tr.41mm. Length-breadth index .66est. This 
may be expressed as follows: 

Eudinoceras mongoliensis type: r.p?, ap. 26 mm., tr. 36 mm., index .72. 

Eudinoceras mongoliensis paratype: l.p*, ap. 27 mm. est., tr.41 mm., index .66 est. 


The length-breadth indices of these premolar crowns, namely, .72 and 
.66, as compared with the index of p* of Dinoceras mirabile type, ap. 
23 mm., tr. 30mm., index .77, indicate that in Eudinoceras the premolars 
are relatively shorter and broader than in Dinoceras mirabile. Accord- 
ingly we anticipate that in Hudinoceras the skull will be found to be 
shorter than in Dinoceras. This suggests the possible relationship of 
Eudinoceras to the genus Bathyopsis Cope in which the jaw is relatively 


Fig. 1. Discovery of Eudinoceras mongoliensis teeth. 

(Above) Roy C. Andrews standing on the spot where the type was found, June 
1, 1923. 

(Below) Henry F. Osborn kneeling on the spot where the paratype tooth was 
found, September 17, 1923. 


4 AMERICAN MUSEUM NOVITATES [No. 145 


right upper premolar 


: i A.M. 20/0! TYPE 
anterior view 


PARA- 

17. 70. 

left upper premolar , A. 1. 20/02 7 VBE 
anterior view 


Fig. 2. A, Type of Eudinoceras mongoliensis (Amer. Mus. 20101), a third or 
fourth superior premolar of the right side: crown view (right), anterior view (left). 
Natural size. ; 

B, Paratype (Amer. Mus. 20102), a third or fourth left superior premolar: crown 
view (right), anterior view (left). Natural size. 


short and deep; the name Bathyopsis is derived from the Greek B%6us, 
deep, oyus, face, signifying deep-faced. We must, however, await the 
discovery of a cranium of Hudinoceras before we can determine whether 
it has a short-faced, bathyopic, cranium, harmonic with its relatively 
short and broad teeth. 

Meanwhile we may observe that none of the Upper Eocene Dino- 
cerata we have hitherto known, namely, Uintatheriwm, Dinoceras, Eoba- 


1924] EUDINOCERAS OF MONGOLIA 5 


sileus (= Loxolophodon), possesses the prominent internal cone of the pre- 
molars characteristic of Hudinoceras. The homologue of this cone is 
apparently a tritocone or anterointernal premolar cusp, as is observed, 
for example, in the Lower Eocene Coryphodon testis p+ (Osborn, 1898, p. 
204, Fig. 22). 

The above type and paratype figures, reproduced precisely natural 
size, represent the characters of this opposite pair of teeth. The paratype 
(tr. 41mm.) is little broader than the type (tr. 36 mm.) and may represent 
either a larger individual, or the two teeth may be successive instead of 
directly opposite, namely, the type may represent a p? of the right side, 
while the paratype may represent a p‘ of the left side. 


Pe 
Roky 
pits ray} 


tas 
eae 

SA te 
AL 


uh 


sed 
i / 


yew } 


sMeRi et 
arn ae 


ee 


Pat ay 
Gay 
Sale 


o” ay 
~ 21 4a eh 
Bre: Bai) os : 
im 
(ay 


ge 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 146 Tue American Museum or Naturat History Nov. 11. 1924 
New York City 4 


56.9,743A (1181:51.7) 


ANDREWSARCHUS, GIANT MESONYCHID OF MONGOLIA! 
By Henry FarrFieELD OSBORN 


Dedicated to the leader of the Third Asiatic Expedition, Mr. Roy 
Chapman Andrews, are the name and description of this giant omnivor- 
ous carnivore of the Upper Eocene of the Irdin Manha formation of 
Mongolia. When first discovered by Mr. George Olsen, it was hailed 
by Mr. Andrews as a carnivore, a supposition which proved to be 
correct. 

Later, its surpassing size led to the view that it was a member of 
the giant pig family represented by Entelodon in Europe and by the 
Entelodontidz in North America,—giant omnivorous pigs with elongate 
skulls; in fact, the cranial and facial proportions of Andrewsarchus are 
remarkably similar to those of Entelodon of the Oligocene and of Dinohyus 
of the Lower Miocene of North America, doubtless because of similar 
omnivorous feeding habits. 

An outline sketch of the skull was sent in a letter to the Museum, 
from which Dr. W. D. Matthew immediately observed its real affinity 
to the primitive Creodonta of the family Mesonychide. When the 
specimen reached the laboratory, it was compared with the giant 
Mesonyx (Harpagolestes) wuintensis of the Upper Eocene of Wyoming 
(Osborn, 1895.98, p. 79, Fig. 4). 


ANDREWSARCHUS, new genus 
Derived from surname ‘‘ Andrews” and from apxés, a leader, chief or 
commander. Giant mesonychid with greatly elongated facial region. 
Zygomata broadly expanded; face contracted behind muzzle; full 
eutherian dentition, thus differing from Mesonyx (Harpagolestes) in 
which m3 is absent; fourth premolar, p*, molariform, triangular, with 
prominent. tritocone. 


Andrewsarchus mongoliensis, new species 


Genotypic species of Andrewsarchus. Type cranium (Amer. Mus. 
20135), Irdin Manha formation, Mongolia. Breadth-length index of 
cranium, .67; faciocranial index, 1.50; basicranial length, condyles to 


_ 1Publications of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 34. 


2 AMERICAN MUSEUM NOVITATES iNo. 146 


tips of premaxillaries, 834 mm.; zygomatic width, 560 mm.; m? with 
metacone greatly reduced, ectoloph consisting mainly of paracone. 

This is the largest terrestrial carnivore which has thus far been dis- 
covered in any part of the world. The cranium far surpasses in size that 
of the Alaskan brown bear (Ursus gyas), the largest living carnivore, 
which, when full-grown, weighs 1,500 lbs.; in length and breadth of skull, 
A. mongoliensis is double Ursus gyas and treble the American wolf 
(Canis occidentalis). It is also treble the size of its American relative 


Amer, (us. 20/35 TYPE 


ms, 


Fig. 1. Type skull of Andrewsarchus mongoliensis (Amer. Mus. 20135), Irdin 
Manha formation of Mongolia. Lateral and occipital views. One-tenth natural 
size. , 


Mesonyx obtusidens from the Middle Eocene of Wyoming and double that 
of Mesonyx (Harpagolestes) uintensis (Fig. 3) from the Upper Eocene of 
northern Utah, Uinta B. 


CRANIAL CHARACTERS.—A comparison of the palate (Fig. 2) and occiput (Fig. 
1) of A. mongoliensis (one-tenth scale) with the palate of Harpagolestes wintensis 
(Fig. 3), reproduced to one-fourth scale, brings out: (1) the great difference in size, 
and (2) the difference in generic and specific characters. The comparative measure- 
ments are as follows: 


Harpagolestes Andrewsarchus 
uintensis mongoliensis 
Total basal length of cranium... ............... 429 mm. 834 mm. 
Width of zygomatic arches. ...:....25 0.000 052 270 560 
Facial length, distance from 1! tom’......... 206 500 


Cranial length, distance from m* to condyles. . . 223 334 


1924] ANDREWSARCHUS OF MONGOLIA 3 


Harpagolestes Andrewsarchus 


uintensis mongoliensis 
Zygomatic width across occipital condyles... . . 71mm. 117 mm. 
Bretocratital IGE. i. i eae 92% 150% 
Cephalic, or breadth-length index............. 63% 67% 


ALASKAN BROWN BEAR 


Ursus gyas 
Amer. Mus. 21802 


Amer (Vhs. 20/35 TYPE 


ss 


ANDREWSARCHUS MONCOLIENSIS 
(ALL TO THE SAME SCALE ) 

Fig. 2. Comparison of Andrewsarchus with Mesonyx, Ursus and Canis. Type 
cranium of Andrewsarchus mongoliensis (Amer. Mus. 20135). Palatal and superior 
views. All figures one-tenth natural size. 

Cranium of Mesonyzx obtusidens (Amer. Mus. 126438). 

Cranium of Alaskan Brown Bear (Ursus gyas) (Amer. Mus. 21802). 

Cranium of Wolf (Canis occidentalis) (Amer. Mus. 31624). 


DENTAL CHARACTERS.—Type characters of A. mongoliensis which may prove to 
be of generic or specific value are the following (Figs. 1, 2) :» 

I? greatly enlarged; i much smaller. 

Canines, superior, not preserved, of enormous size. 

P! small, with single fang and cone. 

P* bifanged, with single cone; ap. 46 mm., tr. 21 mm. 

P’ with rudimentary third fang, double cone, prominent protocone, no deutero- 

cone; ap. 59 mm., tr. 37 mm. 

P* enlarged, molariform, with prominent tritocone, three fangs; ap. 51 mm., tr. 

43 mm. 


4 AMERICAN MUSEUM NOVITATES [No. 146 


M! somewhat smaller in size, three fangs; ap. 39 mm., tr. 42 mm. 
M? broadly triangular, large internal fang, with prominent proto- and tritocones, 
deuterocone worn off; ap. 44 mm., tr. 61 mm. 
M$ prominent proto- and tritocones, rudimentary deuterocone; ap. 50 mm., 
tr. 66 mm. 


Fig. 3. Palate of Mesonyx wintensis 8S. & O. ref. (Amer. Mus. 1892), Uinta B, 
Utah, Dolichorhinus cornutus level. One-fourth natural size. For comparison with 
palate of Andrewsarchus mongoliensis (Fig. 2). 


These measurements show that the cranium of Andrewsarchus (834 
mm.) is nearly twice as long as that of Harpagolestes (429 mm.), that the 
zygomatic arches are twice as broad and that the facial length of Andrews- 
archus (500 mm.) is 150% as compared with that of Harpagolestes (206 
mm.),92%. The face of Andrewsarchus is 150% of the cranium, whereas 
the face of Harpagolestes is 92% of the cranium, supporting the state- 
ment that in Andrewsarchus the face is relatively elongate. 


1924] ANDREWSARCHUS OF MONGOLIA 5 


CoMPARISON WITH HARPAGOLESTES.— While in nearly all dimensions 
double the size of its American contemporary Harpagolestes, Andrews- 
archus differs also in other characters: 


Harpagolestes Andrewsarchus 
BEING BETLER oes Se ang sen then Transverse in posi- Anteroposterior in 
tion. position. 
Faciocranial proportions............... Equal. Face greatly exceeds 
cranium. 
RCETION ATOR. «yOu wie de 2 aaee Pahoa Tubular alisphenoids. Widely open ali- 
sphenoids. 
MMBC THOVAR Ss. ins 2 oetee we hee Absent. Present. 
RESTORATION 


According to Scott (1918, pp. 558-560), the Mesonychide, prevail- 
ingly a North American family, ranged from the Paleocene to the Upper 
Eocene, in adaptation an analogue of the Hyznodontide among the 
Carnivora. Skeletal restorations are those of Scott (1887, Journ. Acad. 
Nat. Sci. Phila., IX, Pl. v, Mesonyx obtusidens) and of Wortman (1901, 
Amer. Journ. Sci., XII, Pl. viz, Dromocyon vorax). From these resto- 

rations the following comparisons may be made: 


Dromocyon vorax Mesonyx obtusidens 
Incisive teeth to back of pelvis........1272 mm. (4ft.2in.) 1278mm. (4ft. 4in.) 
Height of shoulder above the ground.. 672mm. (2ft.2%in.) 576mm. (1ft.104in.) 
Basal length of cranium............... 318 mm. 279 mm. 


LenetH oF Bopy.—As the length of the cranium of Mesonyx ob- 
tusidens (279 mm.) is to the total body length (1278 mm.), so is the length 
of the cranium of Andrewsarchus (834 mm.) to its total body length, 
namely, 3820 mm. (=3 m. 82 cm. or 12 ft. 6% in.). 

Heicut oF Bopy.—As the length of the skull of W. obtustdens (279 
mm.) is to the height of the vertebre (576 mm.), so is the length of the 
skull of Andrewsarchus (834 mm.) to the height of the vertebre, namcly, 
1890 mm. (=1m. 89 cm. or 6 ft. 2 in.). 

If Andrewsarchus mongoliensis was proportioned in the same manner 
as Mesonyz obtusidens, it had a length from the snout to the back of the 
pelvis of 12 ft. 6% in. and a height from the ground to the shoulder or 
middle of the back of 6 ft. 2in. Thus in round numbers it was three times 
the size of Dromocyon vorax or of Mesonyx obtusidens of the Middle 
Eocene of Wyoming, Bridger formation. 


‘ 
ALLS Rr 
Dp it hr ok i 


4 ge eae 
Bp a 


‘ ane 


f ee NC eat oe 
, 5 1 
Ag ORI okt SBA Tyree aii 
ley Esivaaatn ent io? 
i tS pe cee f 
we ¢ 


al 
Ti 
ia 


tae ia om rs, 
a: foD Nana sat ts 


4 
an 
ih 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 147 Tur American Museum or Natura. History Nov. 11, 1924 
New York City é 


56.9,735C (1181:51.7) 


CADURCOTHERIUM ARDYNENSE, OLIGOCENE, MONGOLIA! 


By Henry FAIRFIELD OSBORN 


Cadurcotherium ardynense appears in the Ardyn Obo formation? 
associated with animals of Lower Oligocene age, referred to Cynodictis, 
Schizotherium and Aceratherium. In the preliminary description of 
Cadurcotherium ardynense,’ written before the cranial characters were re- 
vealed as shown in the present paper, the generic and specific characters 
were vaguely stated. A restatement is now made of both the generic 
and specific characters of this animal. 


Cadurcotherium ardynense Osborn, 1923 

Typr.—Amer. Mus. 19154, aged skull with complete upper dentition. 

ParatyPes.—(1) Amer. Mus. 20441, adult lower jaw of slightly younger age 
than the type. (2) Amer. Mus. 19155, cranium with complete premolar-molar denti- 
tion, of younger age, m* not yet in place. (3) Amer. Mus. 20444, palate with molars, 
old individual. (4) Amer. Mus. 20448, juvenile jaw and milk dentition. (5) Amer. 
Mus. 20442, limb and foot bones of other individuals; all found close together 
in the same quarry. According to Granger, the six or seven individual specimens, 
with remains of limbs, were found on the same level, in one pit, and undoubtedly 
belong to the same species. 

Horizon.—Ardyn Obo formation, Lower Oligocene, Mongolia. 

Speciric CHARACTERS OF C. ardynense.—Animals of small size, long and powerful 
upper and lower tusks, suboval in section; canine tusks suboval in section; incisors 
rudimentary or absent; lower premolars reduced to 3; inferior molars with longi- 
tudinal greatly exceeding transverse diameter; superior molars with transverse 
exceeding or equaling longitudinal diameter. Facial region deep, greatly abbreviated. 


As shown in the aged type (Figs. 1, 2), the skull is abbreviated, the 
face especially so; facial and premolar abbreviation together with great 
enlargement of superior and inferior tusks, a distinctive family feature of 
the Amynodontide; principal measurements of the type skull (Amer. 
Mus. 19154) and of the paratype skull (Amer. Mus. 19155) as follows: 


ae of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 35. 
2Named and defined by Berkey and Granger in American Museum Novitates, No. 77, p. 12. 
83Osborn, H. F. 1923. ‘‘Cadurcotherium from Mongolia.’ Amer. Mus. Novitates, No. 92, pp. 1,2. 


: Bias Mus. 19/54 TV PE 


ZA 


« NYS 


i NO 


Fig. 1. (Upper figure) type skull of Cadurcotherium ardynense (Amer. Mus. 
19154). (Lower figure) paratype jaw of same (Amer. Mus. 20441). Both figures 
one-fourth natural size. 


bo 


1924] CADURCOTHERIUM ARDYNENSE, MONGOLIA 3 


A. M.200/8 


Fig. 2. Palate of type cranium of Cadurcotherium ardynense (Amer. Mus. 
19154). One-fourth natural size. 

Fig. 3. Cadurcotherium ? sp. (Amer. Mus. 20018),].m,,m3. One-fourth natural 
size. These teeth are in the Barnum Brown collection of 1922, found on the Irra- 
waddy River; Burma, and they correspond with the species described by Pilgrim. 


Type Paratype 
Amer. Mus. Amer. Mus. 

19154 19155 

mm. mm. 
Total length of cranium and face............... Behe 370! 
Length from orbit to premaxilla.. ............... 1231 1531 
Total depth of cranium, third molar tooth...... 175 165! 
Length of lower jaw, angle to symphysis........ 363 Ree 
Superior premolar-molar series............... 183 182} 
Inferior premolar-molar series (A. M. 20441)... ipso 164 
Superior premolars, p> ~ ooo ve ccs. oh ave ee 56 56 
Srpeloncaolans, ms homer auch notte ec) a 137 132 
Pleioht, ot ectolapnscM ss... ake ae. aes an ode pk 59 
Pine lieret CGuOLOp Ms Mia ox a. a s.zee ate aos tw 551 


Fortunately the younger cranium (Amer. Mus. 19155) supplements 
the aged type (Amer. Mus. 19154) and gives us the complete characters 
of the superior and inferior dentition. In the deeply worn premolars of 
the type (Fig. 2), the proto- and ectoloph unite internally; in the younger 
cranium, when unworn, p* shows a separate proto- and ectoloph. 


Dentit10on.—In the juvenile paratype jaw (Amer. Mus. 20443) the dental 
formula is: Diz Dey Dpz Myz+. In the adult jaw (Amer. Mus. 20441) and type 
skull (Amer. Mus. 19154) the formula is: 132 C1 P3 M38. 


1Estimated. 


4 - AMERICAN MUSEUM NOVITATES [No. 147 


Thus in the young there is one pair of lower milk incisors. In the adult the 
evidence is doubtful; there may have been a temporary pair of incisors. The superior 
and inferior canines equally large, the lower tusks being rounder, the upper tusks oval 
anteroposteriorly; ap. 31_mm., tr. 22 mm., length 81mm. The premolars are 
reduced to two lower and three upper, the formula being: P23; pg isa simple, laterally 
compressed crown; py is submolariform; p?~? have rudimentary internal crests. 

The laterally compressed, extremely hypsodont crowns of the inferior molars, 
m,-3, are generic characters of Cadurcotherium. In m! the transverse diameter 
(46 mm.) greatly exceeds the anteroposterior (32 mm.). In m? the anteroposterior 
diameter (46 mm.) equals approximately the transverse (56 mm.). In m® the 
anteroposterior diameter (56 mm.) equals the transverse (56 mm.). We are there- 
fore able to define this Mongolian species with considerable accuracy. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 148 Tue American Musrum or NaTurat History Nov. 11, 1924 
New York City 


56.9(1182:51.7) 
SERRIDENTINUS AND BALUCHITHERIUM,. LOH 
FORMATION, MONGOLIA! 


By Henry FAIRFIELD OSBORN 


In a thin deposit of olive-colored clays and light gray sandstone 
resting on the red-banded beds of the Hsanda Gol formation at Loh, and 
believed to be of Lower Miocene age, were found two highly character- 
istic fossils: 

1) Proboscidean. A fragmentary series of lower mastodont teeth 
(Amer. Mus. 19152) which first reveals the presence of an undoubted 
Serridentinus in Mongolia, which we name Serridentinus mongoliensis. 

Serridentinus probably marks the arrival of a proboscidean related 
to the M. angustidens of the Lower Miocene of Europe. Osborn has 
recently (1923) separated the generic phylum Serridentinus as a 
medium-jawed trilophodont readily distinguished from the true Trilopho- 
don angustidens phylum by the trefoil conules arising from the side 
of the external cones (protoconids) in the lower molars and from the 
side of the internal cones (protocones) in the upper molars. This trefoil 
characteristic is clearly displayed in the crown view of r.m,_, (Fig. 1). 
In the true Trilophodon angustidens, as in Phiomia, the trefoil conules are 
directly in the center of the crown. Thus distinguished, Serridentinus 
forms a phylum parallel with Trilophodon, which together migrated from 
Eurasia into North America. Species of *Serridentinus appear to 
characterize forested and swampy habitats; they are never found in 
exactly the same areas as species of Trilophodon. We conclude that 
Serridentinus had a different adaptive radiation from Trilophodon. 

2) Rhinocerotine. The facial portion of a skull (Amer. Mus. 19185) 
containing three grinding teeth and perfectly preserved ndsals, which we 
name Baluchitherium mongoliense. 


Serridentinus mongoliensis, new species 
Type.—Amer. Mus. 19152. A series of right inferior grinding teeth, p, (dp,), 
M,, mg; also one left grinder, m, (reversed in drawing); in juvenile condition, 
dp, greatly worn, m, partly worn, m,., embedded in the jaw. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 36. 


2 AMERICAN MUSEUM NOVITATES [No. 148 


outer view 


Serridentinus mongoliensis 
Amer. Mus. 19152 Type 


Fig. 1. Type of Serridentinus mongoliensis (Amer. Mus. 19152), four lower 
grinding teeth, p, (dp,)-m,, Loh formation, ? Lower Miocene, Mongolia. 


(Upper) external view. (Middle) crown view. (Lower) internal view. All one-fourth natural size. 
The teeth are from the right side, except the first lower molar, |. mi (rev.), which is from the left side. 


Horizon.—Loh formation, ? Lower Miocene, Mongolia. 

Locatiry.—Loh, near camp. Upper olive clays. -Found by J. B. Shackelford, 
July 15, 1922, Field No. 71 (Amer, Mus. 19152). 

This is one of the most welcome discoveries of the 1922 expedition, 
because it serves to demonstrate the arrival of what appears to be a true 
species of trilophodont mastodon of the phylum Serridentinus. 


The type specimen was found in an extremely fractured condition 
and was restored, as shown in figure 1. It required great skill in recon- 
struction; the restored parts are indicated by dotted lines. The princi- 
pal measurements in millimeters, all estimated, and indices (I.) are: 


p, (dp,)-mg ine.= 420e 
p, (dp,), much worn, ap. 65, tr. 51, I. .78 


m,, left, ap. 99, tr. 58, I. .59 
Mo, right, ap. 108, tr. 59, I. .55 
Ms, right, ap. 188, tr. 63, I. .46 


Comparing these measurements and indices with those of the type 
of Serridentinus (Mastodon) productus Cope from the Upper Miocene 
of Texas, we observe the following: 


1924] SHERRIDENTINUS, BALUCHITHERIUM, MONGOLIA 3 


P4 (dp,4) my My M3 
Spe thes wilse ape te. by spewing Tocaps) theo, 
Upper Miocene 
Serridentinus productus 

type, and referred m® 55 43 78 93 63 68 138 75+ 54 163! 76! 46! 

?Lower Miocene 
Serridentinus mongoliensis : 

type, 65! 51! 78! 99! 58! 591108! 59! 551138! 63! 46! 

As compared with S. productus, the type of S. mongoliensisis asmaller 
animal, m,., measuring 340 mm. S. mongoliensis is also less progressive, 
each loph consisting of four conelets (by reduplication of the two primary 
cones, corresponding with the protoconid and metaconid in each loph), 
and two large trefoil conules on each side of the external cones (cor- 
responding with the protoconid). More in detail: the crown of dp, is 
entirely worn off; the crown of left m, is largely worn off; in the crown of 
m, the protolophid is wanting; the metalophid has the typical structure 
described above, namely, four conelets, two trefoil conules; in the trito- 
lophid the cones are less distinctly paired into conelets, there is a small 
anterior trefoil conule and a large posterior. In the right m,, the proto- 
lophid has a very large anterior trefoil conule, a small posterior; the 
metalophid is typical, consisting of four symmetrical conelets and two 
symmetrical trefoil conules; the tritolophid is less progressive; the 
tetartolophid is depressed and consists of three main conelets. These 
specific characters may be verified by applying the magnifying lens to 
figure 1. 

As shown in the comparative table above, the tooth proportions of 
S. mongoliensis are very similar to those of S. productus; we may thus 
readily designate the specific stage. . 


Spreciric CHAarActTEeRS.—Ridgecrest formula of S. mongoliensis: Dp4¢g M1z M2z 
M3z. In each typical ridgecrest (e.g., the metalophid) three conelets and two trefoil 
conules attached to the external cones. Molar proportions as in S. productus. 


Baluchitherium mongoliense, new species 


Typr.—Amer. Mus. 19185. Anterior portion of skull including right zygoma, 
complete orbit, frontals, nasals and portions of four grinding teeth, p*-m?. The type 
fortunately preserves the long, smooth and beautifully arched nasal bones, which, 
with the relatively complete smooth frontals, indicate the entire absence of horns. In 
profile view the facial region is similar to that of the type of Baluchitherium grangeri? 
except that the nasals are more prominently arched and exhibit a lateral flange not 


1Estimated. , 

2Osborn, H. F. 1923. ‘ Baluchitherium grangeri, a Giant Hornless Rhinoceros from Mongolia.”’ 
Amer. Mus. Novitates, No. 78. Contribution No. 8, Asiatic Expeditions of The American Museum 
of Natural History. 


mt AMERICAN MUSEUM NOVITATES [No. 148 


observed in the former species. The type skull of B. mongoliense indicates an animal 
little more than half the size of the type skull of B. grangeri, the actual proportions 
being clearly indicated in the following comparative measurements: 


Baluchitherium Baluchitherium 


grangeri mongoliense 
mm. mm. 
Length, fourth superior premolar to third 
Molar, saw Mae cere ye RR ses ee 288 163! 
Fourth superior premolar, p*, anteroposterior . 52 32 
Fourth superior premolar, p*, transverse ...... 92 42 
Nasals, length from tip of nasals to maxillary 
SUCULE iG sa ee es Re ators exiedne & tMhe erasate 334! 224 
Facial-hetehtabove orbit... 23... i ese. Yo os 430 220 


AMER. MUS. 19185 TYPE 


al- 


Q|— 


Fig. 2. Type of Baluchitherium mongoliense (Amer. Mus. 19185), anterior 
portion of skull, Loh formation, ? Lower Miocene, Mongolia. Skull one-sixth natural 
size. Crown, p*-m?, one-third natural size. 


Horizon.—Loh formation, ? Lower Miocene, Mongolia. 

Locauiry.—Loh, upper light gray sandstone, approximately the same level as 
hind foot of ?Rhinoceros (Amer. Mus. 19151). Found by R. C. Andrews, July 1, 
1922, Field No. 67. 

DescrIPpTION.—These measurements indicate that B. grangeri is not far from 
double the size of B. mongoliense. Size, however, is not a reliable specific character. 
A very marked specific distinction is found in the progressive condition of the proto- 
loph and metaloph in the fourth superior premolar, p*, of B. mongoliense; the proto- 
cone is constricted asin m!;_ the metaloph is elongate and develops a crista; thus the 
fourth premolar of B. mongoliense is much more progressive than the fourth premolar 


ifistimated. 


1924] SHERRIDENTINUS, BALUCHITHERIUM, MONGOLIA 5 


of B. grangeri (compare Osborn, op. cit., Fig. 4C) in which the protoloph extends into 
a broad simple loop around the inner side of the abbreviated metaloph, as in Oligo- 
cene aceratheres generally. An additional character is the apparent hypsodonty of 
the molar teeth, the crown of m? showing the following measurements: ap. 45 mm., 
tr. 44mm., index .98; height of ectoloph 39 mm.; protocones strongly constricted 
in m+~?; prominent postcrochet on the metaloph. 

Sprciric Cuaracters.—Nasals strongly arched with lateral flange. P* sub- 
molariform with constricted protocone, a metaloph and crochet. Molars subhypso- 
dont with constricted protocone, antecrochet and crochet. Type cranium about half 
the size of the type skull of B. grangeri. 


REFERRED RHINOCEROTINE PES 

In the same horizon with the type skull of B. mongoliense and in the 
upper olive clays, Loh formation, was found the pes labeled “? Rhinoceros, 
hind foot, found by J. B. Shackelford, Field No. 50, Amer. Mus. 19151, 
June 27, 1922.” This pes certainly belongs to a different individual 
and probably to a different species. It includes the astragalus, cal- 
caneum, cuboid, navicular, ectocuneiform, mesocuneiform and metatar-_ 
salia II, III, 1V. Unfortunately, the metatarsals are incomplete below, 
so that we cannot estimate their length, although it would not appear 
from Mts. IV that the metapodial was extremely elongate as in B. 
grangert. The foot is undoubtedly rhinocerotine. 


Notes by W. D. Matthew (October 3), “I regard the above de- 
scriptions as erroneous in two points: 

1. The distinction that you draw between teeth of your Serriden- 
tinus and Trilophodon phyla appears to me non-existent, as both T'rilo- 
phodon and Phiomia have the cusp construction that you ascribe to 
Serridentinus alone. 

2. The molars of “‘Baluchitherium”’ mongoliense are so widely dif- 
ferent from B. grangeri that I cannot believe there is any near relation- 
ship. On the other hand, both molars and premolars show a marked 
approach to the Ccelodonta group in the hypsodont wavy ectoloph and 
in many other details. Of course I am fully aware of the superficial 
resemblance to Baluchitherium in the nasals, but would rather expect 
to see some such construction in a hornless member of the Ccelodonta 
group.” 


Ta os 


bared ee That 


ee ie 
ane arie 


i 
ian 
ee 
AY 
* 
on 


LSichiosk si aaa as 


bh, lied Ye i he bat ceeeeuh 
Sane ean tt he Costa oneiba arn a 
} c yea n 
‘ ' Hy \ om iy , f if ui, a i 
Ser al NR PAY Ketan 
* ' ig 
‘ 
het 4 . 
anh 7 
ft 
198 da he heh | 
wos ot Ne aa 
‘ » 
’ si hy 
i ‘ 3 r 
, ll VR, fa Ae tier 
‘ hag ey y 
" r ; = 4 rr . 
Ase iy eh eS 
ik | y - 
ae os 
ry 4 } 
oe thea arae ds ’ | 
i i 
Phot eatery shy eit » 
j 
‘ 7 
‘\ 
oe ! 
‘i ‘ 
i 4¥ 
‘ I * 
v 
‘ Ly 
‘ . ys ef 
? i 
ats 8 
i 4 


Note) 
read ahi 
' , 
ir 
A se 
1 otal 
py > 
H tale t 
Vee oe 
‘ 
wae 
Key 
a ns 
i . 
‘ 4 - 
‘ ei 


oe 


Fossils in the Ondai Sair Formation, Mongolia 


By T. D. A. CockerEenn 


BULLETIN 
| OF 
THE AMERICAN MUSEUM OF NATURAL HISTORY | 
, Vou. LI, Arr VI, pp. 129-144 | 
New York 
_ Issued December 30, 1924 


u ox yy) oo 
pr a cc 3 Uy Vj ay a 


Nae pss : eV 
ES ‘ 


Fig. 1. Map of the Gobi region. 


The mountainous areas are shaded with slanting lines; the lowlands are white; the deeper de- 
pressions are stippled, 

Ondai Sair is indicated by the index letters OS, south of the letters MTS of Kaancat Mts. Ust 
Balei is not marked on the map, but it lies on the large river that leads northwestward out of Lake 
Baikal, near the prong of mountain-country that projects toward this river from the Sayan ranges. 


56(117:51.7) 
Article VI—FOSSILS IN THE ONDAI SAIR FORMATION, 
MONGOLIA! 


By T. D. A. CockERELL 


INTRODUCTION 


I am indebted to The American Museum of Natural History for the 
opportunity to study a most interesting collection of fossils in the paper 
shales of the Ondai Sair formation of the Gobi Desert. These were 
collected by Messrs. Charles P. Berkey and F. K. Morris, on the Third 
Asiatic Expedition of the Museum. Messrs. Berkey and Granger (Ameri- 
can Museum Novitates, No. 77) discuss the Ondai Sair formation, and 
state that it is ‘‘at least as old as the Cretaceous, perhaps the Lower 
Cretaceous or Comanchean, rather than the Upper Cretaceous. In this 
locality, at Mt. Uskuk, the series is sharply delimited by an uncon- 
formity, above which no fossils of this type are to be found; and it is 
limited also below by a much greater unconformity, beneath which lie 
folded strata with coal beds, conglomerates and intrusives, all judged to 
be of Jurassic age, as well as many still older formations.” 


eee of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 37. 


129 


130 Bulletin American Museum of Natural History [Vol. LI 


Mr. Frederick K. Morris has been good enough to write me at con- 
siderable length, summarizing the observations and conclusions of the 
investigators. I venture to quote the following: 

‘““We found the fish-bearing paper shales associated with sandstones 
in which were sauropod reptiles at Ondai Sair. At the Oshih (Ashile) basin 
we searched the paper shales this year again for fossils, but found only 
the remains of a few plants, too badly macerated prior to deposition to 
be identified. The paper shales of the Oshih basin carry gypsum crys- 
tals, which perhaps points to bitter waters in which the fauna of Ondai 
Sair could not live. The Oshih and Ondai Sair basins are very probably 
continuous and form one basin. Sauropods and primitive predentates 
are found in each locality, although perhaps there is a slight difference 
between the two faunas. At Oshih the paper shales lie far higher in the 
series, not less than five hundred feet above the chief sauropod beds; 
but a few large sauropod vertebre were seen only one hundred feet below 
a horizon which, on other grounds, we believe is correlated with the 
paper shales. At Ondai Sair, as I have said, the sauropods are found 
close to the paper shales and essentially of the same age. 

“‘Fish-bearing shales are found in North China and have been called 
Jurassic, on the assumption that the fishes were indeed the Lycoptera 
middendorffi described by Reis. Paper shales were seen by us on the 
main road to Urga, between Ude and Tuerin, but our brief search of them 
did not reveal any fossils. We think this basin may be of the same age as 
the fish shales of Ondai Sair. 

“The Oshih-Ondai Sair basin is the oldest true basin found by us in 
Mongolia; its sandstones rest directly upon the strongly disturbed old 
rocks that have passed through one or more mountain-bearing revolu- 
tions. At Ondai Sair the basement was chiefly pre-Cambrian rocks— 
graywackes and slates invaded by granites. At Oshih less is seen of the 
basement, but we found an inlier of crystalline limestone, cut by syenite 
and granite; also other inliers of the old graywacke series of which the 
Altai range to the south is principally composed; and a younger chain of 
strongly folded porphyries, which may be what Doctor Berkey and I 
judged to be Jurassic in other localities last year. Clearly, there is a 
great unconformity beneath the Oshih and Ondai Sair and, so far as we 
know, these basins, which carry the paper shales, date the beginning of 
the basin or gobi sedimentation; that is, the beginning of deposition of 
shallow masses of continental sediments—alluvial fans, flood-plains, 
deltas, playa lake deposits and, locally, deposits in shallow but more 
permanent lakes. These deposits are but little disturbed and it is in 
them that all the vertebrate fossils are found. 


1924] Cockerell, Fossils from Mongolia 131 


“The accurate dating of the oldest of these basin deposits is of 
great importance, not only for paleontology but for the dating of the 
last great disturbance prior to the Altai uplift. 

“We found marine beds of Permian age involved in the older moun- 
tain-folding. Apparently younger than these, is a vast series of shales, 
sandstones, conglomerates and surface volcanic rocks—ash, tuffs and 
flows of rhyolite, trachyte and basalt which are strongly folded. After 
folding, these rocks had been eroded to a peneplane. This latter series is 
wholly non-marine and carries obscure plant remains and, locally, coal. 
We named these rocks the Tsetsenwan series. 

“South of Mongolia the Lower Jurassic in northern China carries 
coal and is strongly folded, even slaty in places, and has porphyries very 
like those in the Mongolian series just mentioned. In northern China, 
therefore, a folding took place sometime between Middle Jurassic and 
late Jurassic time. In view of the physical similarity between the Lower 
Jurassic rocks of the Western Hills and the Tsetsenwan seriesin Mongolia, 
and because in China also the fish-bearing shales are not severely dis- 
turbed, it has seemed to us that for reasons of structural geology it is 
likely that the earliest basin lands are possibly of Comanchean age. The 
“Upper Jurassic or Lower Cretaceous” of Reis may well be what we call 
in America, Comanchean. 

“The significance of the Ondai Sair fauna is increased when we 
consider that there is apparently a chain of basins extending from Siberia 
across Mongolia into China; if the fauna proves to be the same or of 
nearly the same age in all, this fact would correlate the entire column of 
post-mountain-folding, continental sediments in the three regions.” 


ONDAI SAIR BIOTA 
The following species were obtained by the Expedition in the Ondai 
Sair formation: 

Reptilia: Protiguanodon mongoliense Osborn, Amer. Mus. Novi- 
tates, No. 95, 1928. 

Pisces: Lycoptera middendorfii Johannes Miiller, 1848. Ex- 
ceedingly abundant in the paper shales (PI. II, fig. 1). 
The specimens agree with L. middendorffi in the - 
proportions of the fins, rather than with L. sznensis 
A. 8S. Woodward, from the Lower Jurassic (?) of 
the Province of Shantung, China. These fishes 
will be discussed more fully in a later contribution; 
they form a new family, Lycopteride, apparently 


132 Bulletin American Museum of Natural History [Vol. LI 


ancestral to the Cyprinide, and have scales scarcely 
differing from those of the European minnow. 
Crustacea: Estheria middendorfii T. R. Jones, 1862. Exceedingly 
abundant in the paper shales (PI. II, fig. 2). 
Insecta: Ephemeropsis trisetalis Kichwald, 1864 
Ephemeropsis melanurus, new species 
Cymatophlebia (?) mongolica, new species 
Trichopterella torta, new species 
Indusia reisi, new species 
Chironomopsis gobiensis, new species 
Coleoptera, spp. incert. 
Plants: Baiera furcata (Lindley and Hutton) 
Phyllocladites (?) morrisi, new species 
Czekanowskza sp. 
Equisetacez (?) fragments 


Concerning this biota, we have to ask, (a) does it appear, taken as it 
stands, to be Upper Jurassic or Lower Cretaceous? (b) is it identical, 
or nearly identical, with that of the supposed Upper Jurassic fish shales of 
the Transbaikal country directly north of the Gobi Desert? 

With regard to the first question: 

(1) The dinosaur, I learn from Dr. W. D. Matthew, is provisionally 
referred to the Cretaceous, but it is not of itself decisive as to age. 

(2) The fish is singularly modern in aspect and affinities, but it be- 
longs to a type always considered Jurassic. 

(3) The Estheria, as an organism, is of no value in determining age, 
except within very wide limits. Similar species are living today. 

(4) The insects do not clearly indicate age; they are on the whole of 
modern type, but such forms are known to occur in the Jurassic. 

(5) The plants appear to represent a Jurassic flora, but similar forms 
have been found in the Cretaceous. Thus the contemplation of this 
biota apparently leaves us where we were before, in uncertainty as to 
whether it is Upper Jurassic or Lower Cretaceous. 

Turning now to the second question, it must evidently be answered 
- in the affirmative. The Transbaikal fish shales are found in several 
places, for instance: Turga, about 60 miles north of Mongolia, directly 
north of the Gobi Desert; Konduevskoe or Konduyewskaya, about 90 
miles due east of Turga; Nertchinsk or Nertstschinsk, 100 miles north- 
northwest of Turga. 

These shales are, just as those from Ondai Sair, full of Lycoptera 
middendor ffi, Estheria middendorffi and Ephemeropsis trisetalis; they also 


1924] Cockerell, Fossils from Mongolia 133 


carry plants which appear to be Bazera and Czekanowskia. It does not 
appear possible that such an association could occur in two deposits of 
very different, age. I have not had access to the important paper by O. 
Reis, “Die Binnenfauna der Fischschiefer in Transbaikalien,” Explor. 
Géolog. chem. de fer Sibérie (Petrograd, 1910), but Mr. Morris has sent 
me photographs of the plates, with explanations, and a brief statement 
of Reis’s position. The beds have always been considered Upper Jurassic 
but Reis confessed to finding in the fauna puzzling affinities ranging from 
early Tertiary to Jurassic, although the strongest bonds appeared to be 
with the Upper Jurassic or Lower Cretaceous. 

As Reis records a number of organisms not yet found in the Ondai 
Sair, it will be well to review the more significant of them: 

(1) Some additional fish remains are too fragmentary to be of much 
value. One of these, based on some fins, is called Stzchopterus woodwardt. 
The generic name is new and has been omitted from Jordan’s Classifica- 
tion of Fishes and Genera of Fishes. The genus is, however, probably 
unrecognizable. 

(2) There is a series of Mollusca, referred to Lymnza, Paludina, 
Cerithium and Cyrena. Lymnea obrutschewi Reis is closely similar to 
L. accelerata White, from the Comanchean near Cafion City, Colorado, 
except that itis smaller. Paludina pura Eichwald very closely resembles 
Lioplacodes veternus Meek and Hayden, from the American Jurassic, 
but is also smaller. It is a shell with much the aspect of the modern 
Bythinia tentaculata Linné, but I think it is safe to call it Lzoplacodes 
purus.! Cyrena pusilla? Reis is a very small species but characteristic of 
the genus and having the general form of Cretaceous species. Cerithium 
gerassimowi Reis is not closely related to anything known to me, but 
the approximately similar types belong to later epochs. 

(3) The additional insects are of no consequence. One beetle is 
named Carabites latecostatus, but its family position is uncertain. 

(4) Numerous ostracods are recorded by Joseph Georg Egger, with 
a plate of figures. They are referred to described species, but distinctive 
characters are few and I do not know what reliance can be placed on them. 
Our confidence in the determinations is not increased by finding one of 
the species referred to Cypris faba, the common ostracod of the Miocene 
at Wangen ((Eningen) in Baden. 

(5) The additional fragmentary plants convey little information. 
Pinus witimi, based on a winged seed, suggests a much later epoch; but 


1Comparison may also be made with the genus Bazkalia Martens, living today i in Lake Baikal. 

2The name Cyrena pusilla is preoccupied by Cyrena pusilla * ‘Parreyss, ” Phil., Abbild. II. 78. 
Cyrena, Tab. 1, f. 7 (February 1846), from the Nile. The fossil species described. by Reis may be 
known as C. reisi, new name. 


134 Bulletin American Museum of Natural History [Vol. LI 


there is no reasom why this should not be called Pityospermum witimi, 
The genus Pityospermum of Nathorst includes winged seeds resembling 
Pinus, found in the Jurassic. Another Siberian species, Pityospermum 
maackianum, was described by Heer as Pinus maackiana.* 


UST BALEI BIOTA 

There remains one other method of attacking our problem, by com- 
parison with the biota of the Ust Balei beds, north of Irkutsk and west of 
Lake Baikal. These beds were first mentioned by Trautschold in 1867, 
and since then have proved fruitful of plants and insects. They are uni- 
versally recognized as Jurassic and generally ascribed to the Middle or 
“Brown Jura,” more or less equivalent to the Lower Odlite. Since those 
who consider the Lycoptera shales Jurassic, agree in placing them in the 
Upper Jurassic, it is not to be expected that they should carry the same 
biota as the Ust Balei; but there might be some similarity. 

The Ust Balei insects, as recorded, number about 25 species, as 

follows: 

Blattoidea: Ophismoblatta sibirica (B., R. and G.)?; O. maculata 
(B., R. and G.). The first is a wing, the second 
a nymph. 

Mantoidea: Pseudohumbertiella grandis (B., R. and G.). Based 
on half a wing; was supposed to be a mantid, 
but Handlirsch thinks it may better be referred 
to the locustoids. 

Dermaptera: Baseopsis (?) sibirica B., R. and G. Based on the 
anterior half of an insect; was supposed to be an 
earwig, but Handlirsch regards it as unrecogniz- 
able, perhaps a beetle larva. 

Orthoptera: Parapleurites gracilis B., R. and G. A good teg- 
men, referred by Handlirsch to his family 
Locustopside. 

Panorpate: Mesopanorpa hartungi (B., R. and G.). Referred 
to the family Orthophlebide. 

Plecoptera: Mesonemura maacki B., R. and G. (imago); 
Mesoleuctra gracilis B., R. and G. (larva); 
Platyperla platypoda B., R. and G (larva), 

Ephemeroidea: Mesobaétis sibirica B. R. and G.; Mesoneta antiqua 
B., R. and G. These are May-fly nymphs of 

1This is spelled maakiana in the books, but it was named after Maack, the celebrated Siberian 


explorer. 
2Brauer, Redtenbacher and Ganglbauer (1889). 


1924] Cockerell, Fossils from Mongolia 135 


ordinary size, quite unlike the giant forms of 
the fish shales. 

Odonata: Palzophlebia synlestoides B., R. and G. (an imper- 
fect wing). Samarura gigantea, minor, pulla, 
angustata and rotundata, all of B., R. and G. 
(nymphs). 

Coleoptera: Carabocera prisca B., R. and G.; Timarchopsis 
czekanowskit B., R. and G.; Doggeria sibirica 
Handl.; Memptus brawert Handl.; M. redten- 
bacheri Handl. Five beetles of rather doubtful 
affinities. laterites sibiricus Heer is said to 
come from Irkutsk. 

Diptera: Mesopsychoda dasyptera B., R. and G. 

Lepidoptera or Homoptera: Phragmatecites damesit Oppenheim; 
Palzocossus jurassicus Oppenheim; two remark- 
able insects considered by Handlirsch to be 
primitive Lepidoptera, but Tillyard has re- 
cently given apparently conclusive reasons for 
referring the Palzontinide to the Homoptera. 
They are recorded from East Siberia, and were, 
I presume, from Ust Balei. 


These insects show no affinity with those of the fish shales, and 
since they constitute an analogous fresh-water and lake-side fauna, the 
absence of the large ephemerids is especially noteworthy. The number 
of species (especially in view of the fragmentary condition of many) iis 
too small to permit any very decisive conclusions, and it is much to be 
desired that the locality should be explored again for additional material. 
We can, however, state that the Ust Balei fauna is not that of the fish 
shales. 

Heer described a number of plants from Ust Balei, but as Seward 
has shown, his work needs radical revision. So far as we know it, the 
flora does not seem to correspond with that of the fish shales. 

We must then conclude that the fish shales of Ondai Sair and the 
Transbaikal contain a fairly homogeneous biota, which must be approxi- 
mately of one age. This biota is quite distinct from that of the Ust Balei 
beds, and presumably much later. It may be Upper Jurassic, but there 
is no proof that it is not Lower Cretaceous. Actually, it presumably 
belongs to a period near the beginning of the Lower Cretaceous and its 
classification as Jurassic or Cretaceous may be merely a matter of arbitrary 
definition. In any event, this biota, which represents an early stage in 


136 Bulletin American Museum of Natural History [Vol. LI 


the development of a number of groups still surviving, is of great interest 
and well merits further investigation. 


INSECTS 
EPHEMERIDA 
(PLECTOPTERA) 


The May flies were well developed as early as the Permian, with 
many genera. Numerous species are known from the lithographic 
stone of Bavaria (Jurassic), and a series of nymphs or larve has been 
obtained from the Jurassic rocks of Siberia. The Permian species, 
constituting a family, Protereismatide, are remarkable for having the 
hind wings as large, or nearly as large, as the upper. Nevertheless the 
venation is singularly modern, as can be seen by comparing the Permian 
Prodromites! rectus (Sellards) with such a genus as the living Ameletus. 
Thus it may be said that at least some of the Permian wings approach 
the modern Siphlonuride in structure and it is therefore perhaps less 
surprising to find apparently genuine siphlonurids in the Jurassic. The 
genus Pedephemera of Handlirsch, with four species in the lithographic 
stone of Bavaria, scarcely differs at all in the wings from Szphlonurus, 
except that the hind wings are appreciably larger. The gigantic May- 
fly nymphs from the fish shales of Siberia have also the characters of the 
Siphlonurid, but differ from Szphlonurus in having more double ab- 
dominal gills, herein agreeing better with Bengtsson’s genus Szphlurella. 
The venation of the adult, as shown by the Mongolian material, is more 
like that of Ephemerella, but also shows distinct affinity with the Permian 
Prodromites. The modern siphlonurids will then constitute a subfamily 
Siphlonurine, in contrast with the gigantic Mesozoic Ephemeropsine. 


EPHEMEROPSIS Hichwald 


Eichwald in 1864 founded the genus on an unfigured nymph from 
Siberia, which he called EF. trisetalis. It came from the shales of the 
Transbaikal country, in the vicinity of Nertchinsk, north of the Gobi 
Desert. In 1868 the same author obtained a larger specimen from the same 
formation and vicinity and set it forth, with a figure, as H. orientalis. 
As Handlirsch remarks, this is very probably the same as F., trisetalis. 
Considerably earlier, in 1848, Miiller had figured the tail of one of these 
nymphs in Middendorff’s ‘Reise,’ the locality being Byrka, Siberia. 
Handlirsch bases his H. middendorffi on this figure but, except that the 
caudal appendages are said to be only about 15 mm. long (instead of 20 


1Prodromites, new name for Prodromus Sellards, 1907 (not Distant, 1904). 


1924] Cockerell, Fossils from Mongolia 137 


mm. or over), there is nothing to distinguish it. In 1889, Brauer, Redten- 
bacher and Ganglbauer figured a nymph from Turga, Siberia, which they 
considered to be Ephemeropsis orientalis. Handlirsch, on their account, 
founds a new generic name, Phacelobranchus, calling the species P. 
brauert. The genus is founded on the fact that the tracheal gills are 
beset with fine hairs, as in the siphlonurid genus Chirotonetes Walker, 
which, however, has quite differently shaped gills. These hairs are seen 
only with difficulty and I cannot doubt that they exist in true Ephe- 
meropsis as well. O. Reis (1910), in describing the fauna of the fish shales 
of the Transbaikal, figures Ephemeropsis orientalis with six pairs of double 
gills, the outer portion ciliated. The figure represents a ‘‘restoration” 
and in some details is not very accurate. Without having the original 
types, it is difficult to be positive about the actual characters of the four 
supposed species, but my present opinion is that they are all synonymous, 
and accordingly to be designated Hphemeropsis trisetalis Eichwald. 
Furthermore, I am not at present able to discern that the Ondai Sair 
species is separable from Eichwald’s. 


Ephemeropsis trisetalis Kichwald 
late I, Figures 1 to 9; Text Figures 2, 3, 4 


Nymrpu.—The one nearly complete Ondai Sair specimen is about 40 mm. long 
from front of thorax to base of caudal appendages (PI. I, fig. 1). The thorax is about 
13 mm. wide, thus considerably broader than in the Reis restoration, the body dis- 
tinctly tapering posteriorly as in the figure of Phacelobranchus. I believe the last two 
pairs of gills are not double, but it is difficult to be sure that the inner lobe is not con- 
cealed beneath the abdomen (PI. I, figs. 1,2). The outer tapering lobe, resembling a 
slender knife blade in form, is about 5 mm. long, 
as described for FE. trisetalis (Pl. I, fig. 3). The 
hind margins of the last two abdominal segments 
are laterally produced and pointed, as in the 
modern Siphlonurus. This feature is not brought 
out by Reis, but it is better indicated in the fig- 
ure of Phacelobranchus, though even here the pro- 
jections of the last segment are not nearly long Fig. 2. Nymphal wing of 
enough. The lateral caudal appendages are 23.5 Ephemeropsis. 

mm. long; the middle one is evidently shorter, 

but my best specimen does not show the extreme tip (Pl. I, figs. 4, 5). I have been 
fortunate in finding nymphal wing pads, with venation as shown in the figures (Fig. 
2 and Pl. I, figs. 6, 7). The drawing, figure 2, shows the branching of the radius 
(radial sector), about 6 mm. from base and 5 from apex of wing, and what looks like 
a faint cross vein from the lower branch to the vein below. ; 

Wincs.—Part of the adult wing has been preserved in one instance (Figs. 3, 4, 
and Pl. I, figs. 8,9). It has the following characters: costa strongly arched at base, 
the strong oblique vein ending in the highest part (basad of this in Ephemera); sub- 


138 Bulletin American Museum of Natural History [Vol. LI 


costa normal, but at oblique vein very close to radius; radius (Ri) stout, practically 
parallel with subcosta; radial sector branching early, but not widely spreading, the 
upper branch forking about 7.5 mm. from its origin, the fork thus formed very narrow, 
its lower branch not elbowed, but doubling of cells beginning after the fourth; Rs 
eventually widely divergent from R, (following Tillyard’s nomenclature),! leaving a 


large open reticulated area, with some 


é oe RI supplementary longitudinal veins; R445 
ee ar (M3;+,4 of Miss Anne Morgan) forking 

R x Bee about 7.7 mm. from base of wing, the fork 
Ra very wide, the lower branch most diver- 

gent but not elbowed at base, several 

3 ais, supplementary longitudinal veins be- 
Rs tween the branches; first cubitus close 

m3i4 M1+2 to second, with no place for oblique veins 


to margin. 


The above description is based 
on the supposition that the wide 
fork so conspicuous in the specimen 
is actually the last radial fork of 
Tillyard, the medial fork of other 
authors. Difficulty arises, how-. 
ever, when we attempt an analysis 
of the radial branches, unfortu- 
nately somewhat disturbed by 
breakage. Leaving the wide fork 
out of the question, the radial 
branches lie close together, fork- 
ing at extremely acute angles. 
The main branch of the sector is 
thicker and darker than the others 
and divides early, its upper branch 
again forking as described above. 
From the base of the main sector below appear to arise two branches, the 
one at the extreme base being the stem of the wide fork (supposed above to 
be R445); the other, a little (1.3 mm.) farther along, is resolved on close 
inspection into two rather weak veins, practically contiguous, but after 
about 3 mm. gradually diverging, as shown in the figure. The uppermost 
of these veins again divides 4 mm. farther on, but the lower section is 
angulated basally, meeting a cross vein. The second or lower of these. 
veins diverges widely apicad from the supposed Ry and about 8 mm. from 
its origin branches again with a wide fork, the lower portion being so 


ee 3. Venation of the basal part 
of adult wing of Ephemeropsis: an analy- 
sis of the photograph, Plate I, Fig. 10. 


C, costa; Se, subcosta; R, radius; Rl, 2, 
OAs, branches of radius; Media, with branches 
mi+2, 344; Cu, cubitus. 


Fig. 4. Venation of part of the 
wing of Ephemeropsis. Compare with 
Plate I, Fig. 11. 


1Tillyard, 1923. Trans. New Zealand Institute, LIV, p. 227. 


1924] Cockerell, Fossils from Mongolia 139 


delicate as to appear like a supplementary vein. Now, if we assume that 
the conspicuous wide fork represents Tillyard’s media (cubitus of Miss 
Morgan), then the weak fork above it and more apicad, although ap- 
pearing doubtfully to belong to the primary venation, is the last fork 
of the radial sector, usually conspicuous in May-fly wings. On this basis, 
the veins above are all branches or divisions of Rez, except the lowermost, 
whichis R3. It then appears that R; arises from the sector at practically 
the same point as R4+;, contrary to what we might expect, and con- 
trary also to the condition in the Permian Prodromites. It is possible, 
however, to maintain that the closely adjacent veins arising from near 
the base of the sector are actually Ry and R;, separate to base, the vein 
R3; being placed as usual. 

Very competent authors who have intensively studied the venation 
of May flies have already given us three quite different interpretations, so 
perhaps we may be excused if we hesitate to dogmatize about the 
homologies in Ephemeropsis.: If the vein Re can present so many rami- 
fications, other veins may also show unexpected complexity, and the 
determination of the basic scheme of venation becomes difficult. Cer- 
tainly the wing pad of Ephemeropsis does not suggest that the com- 
plexities of the adult wing are directly derived from the primary veins. 

It is proper to state that we have no absolute proof that the wing 
described belongs to the same species as the nymphs; but as it occurs 
with them and is of suitable size, I cannot doubt that it is identical. 

Horizon and Locauity.—Ondai Sair formation (paper shales), 
Ondai Sair, Mongolia. The wing and wing pads were in the field parcel 
No. 79. The nearly entire nymph is in field parcel No. 64, but pieces 
of the same nymphs are also from parcel No. 79. 


Ephemeropsis melanurus, new species 
Plate I, Figure 10 

Speciric Cuaracters.—Abdominal gills 4.3mm. long, double, both parts sharply 
pointed; the outer slender, curved below the middle; the inner broad, its width about 
1.4 mm., caudal appendages fringed as in E. trisetalis, but much smaller, probably 
about 10 mm. long (about 8 mm. preserved), the appendages and their fringes black 
(fringes pale in £. trisetalis). The median tail at base is only about half a millimeter 
wide, while that of H. trisetalis is a full millimeter; but nevertheless at 7 mm. from 
base the tail with fringes is 2.5 mm. across, the fringes being very long. 

Horizon and Locatiry.—Ondai Sair formation (paper shales), Ondai Sair, 
Mongolia. 


Although this specimen shows only the structures described, it is 
certainly a distinct species. Both the caudal appendages and the gills 
are sufficiently characteristic for recognition. 


/ 


140 Bulletin American Museum of Natural History [Vol. LI 


ODONATA 
ZEshnide 
Cymatophlebiine (Cymatophlebiina Handlirsch) 
CYMATOPHLEBIA Deichmiiller 


This genus was based on a species from the lithographic stone of 
Solnhofen in Bavaria. A new figure was given by Needham (1907) in 
Bull. Amer. Mus. Nat. Hist., XXIII, p. 141. The Mongolian fossil is 
too incomplete for certain generic reference, but what there is agrees 
sufficiently with Cymatophlebia. 


Cymatophlebia (?) mongolica, new species 
Plate II, Figures 3, 4, 5 

Costa and radius very stout; cells between costa and subcosta before nodus 
broader than high, thus much broader than in C. longialata (Germar); cells between 
subcosta and radius in same region similarly broad; radius in this region nearer to 
media than to subcosta; nodus as in Cordulegaster; costal cells beyond nodus broader 
than high; media branching at lower end of the very oblique subnodus; Mp2 bent 
near base as if by a cross vein to Rs, but I think this is merely a distortion. These 
characters are from the type, in parcel No. 64. 

Horizon aND Locauiry.—Ondai Sair formation (paper shales), Ondai Sair, 
Mongolia. 

Another specimen (also with No. 64) shows the nodus and part of 
the region just beyond. It isso much distorted that it looks like another 
species, M, being under the radius, and the end of the subcosta thrust up- 
ward. Itis of value only as showing an early duplication of cells between 
M, and Mg, a character of Cymatophlebia, only whereas in the latter the 
paired cells are about equal, in our fossil the first few paired cells show the 
upper one very large and the lower very small, hardly a fifth the size of 
the upper. More apicad, the cells rapidly become subequal. Still 
another specimen (field label No. 79) shows part of the sub-basal region 
of wing and indicates the presence of high narrow cells in the hind wing, 
the lower portion folded over and much confused. 


TRICHOPTERA 
TRICHOPTERELLA, new genus 


GENERIC CHARACTERS.—Size medium; wings broad; costal margin very gently 
arched, slightly concave near base; radius (Ri) straight to Ri-Re cross vein, then 
arched upward, resuming a direct course at a higher level; radial sector arising 
near base of wing, branching to form a long discoidal cell, the base of which is more 
acute than the base of the first fork; cross vein at end of discoidal cell somewhat 
oblique; separation of R, from Rs; a short distance before end of discoidal cell, of 


1924] Cockerell, Fossils from Mongolia 141 


R, and R; just beyond discoidal cell; no R-M cross vein; media apparently with 
only two branches, but the ends of these are obliterated and it is not impossible 
that they forked near the apical margin; fork of media at about same level as fork 
of R4 and R;; M-Cu cross vein obliquely arising from media a little before the fork 
and joining the cubitus at the fork or very slightly before it; first anal straight to 
end, parallel with and remote from second; postcosta or second anal forming with 
the others the usual pointed cell, enclosing a smaller cell toward base. Only the 
anterior wing is known. 


Trichopterella torta, new species 
Figure 5 


Speciric CuaractTers.—Anterior wing 15.5 mm. long, 6 mm. wide; veins pale 
brown, no markings preserved; discoidal cell 3.8 mm. long; Rs branching 5.7 mm. 
from base of wing; long anal cell ending 5 mm. from base of wing; length of cubital 
fork 4mm. 

Horizon AND Locatity.—Ondai Sair formation (paper shales), Ondai Sair, 
Mongolia. In parcel No. 64. 


Fig. 5. Discoidal cell of wing of Ti- 
chopterella torta. 


The numerous Trichoptera from the Lias are all small, the anterior 
wings ranging from 3.2 to 9 mm. long, only one species exceeding 5 mm. 
Two of the Lias genera, also represented by small species, occur in the 
Jurassic. The entirely separate and parallel-running first anal of T’77- 
chopterella is equally seen among the Lias genera, such as Necrotaulius, 
Mesotrichopteridium and Pseudorthophlebia. All these insects, however, 
have the characteristic anal cells, which are lacking in Tillyard’s Meso- 
psychide (four genera) from the Upper Triassic of Queensland. 

On the other hand, Trichopterella seems to be specialized, after the 
manner of the Tertiary and modern Setodes, in the reduction of the 
media to two simple branches; but it is not impossible that these were 
forked near the margin. The media and cubitus of Trzchopterella sep- 
arate very early, at the same level as the origin of the radial sector, hence 
the cellula thyridii is very long, nearly twice as long as the discoidal. 

There is a rather wide crack in the rock across the middle of the 
wing, but this does not obscure the venation. 

There is one Jurassic genus, Mesotaulius of Handlirsch, from the 
lithographic stone of Bavaria, which has the anterior wings 31 mm. long. 
It has little affinity with Trichopterella. 


142 Bulletin American Museum of Natural History [Vol. LI 


The living genus Alepomyia Banks from Newfoundland has a two- 
branched media, but is otherwise quite distinct from T'richopterella. 

Reis recorded a trichopterous larva case from the fish shales of the 
Transbaikal. A similar or identical species occurs in the Ondai Sair 
shales, as follows: 


Indusia reisi, new species 
Plate II, Figure 10, on the stem of Phyllocladites 

Sprciric Cuaracters.—A larva case was found contiguous with the base of a 
specimen of Phyllocladites morrist. It is cylindrical, straight, very broad, composed 
of grains of sand, and similar to that of the modern genus Stenophylax. On the east 
side of the main draw, half way between Uskuk and Hsanda Gol, F. K. Morris 
found a heavy slab, about 15 mm. thick, showing numerous specimens of Hstheria 
middendor ffi and a few caddis cases on one side, and on the other very numerous caddis 
cases. From this additional material it can be determined that Indusia reisi cases are 
sometimes as much as 22 mm. long, very slightly curved, 2 mm. diameter at the little 
end, 3.7 at the larger. The distinct pebbles are mainly at the larger end, the choice 
of materials varying with age, asin Stenophylax. The case figured by Reis was orna- 
mented with ostracods. 

Horizon anp Locatiry.—Ondai Sair formation (paper shales), Ondai Sair, 
Mongolia. 


DIPTERA 
Chironomid 
Cutronomopsis Handlirsch 


This genus was based on Chironomus arrogans Giebel, a name founded 
on Brodie’s figure of a supposed Chironomus from the English Purbeck. 
It shows no distinct generic characters and I have used the name only in 
a general sense, as applicable to fossil Chironomide of uncertain status. 


Chironomopsis gobiensis, new species 
Plate II, Figures 6, 7 


Sprciric CHARACTERS.—Female. Length about 7 mm.; head small, with large 
eyes; antennz long and slender, with cylindrical joints, longer than broad, not 
visibly hairy; palpi extended, about half as long as antenne; thorax elevated, gibbous 
anteriorly; legs long, tibise apparently longer than femora, hind tibiz about or nearly 
3mm. long. Color as preserved, brown, evidently dark brown or blackish in life. 

Horizon AND Locatiry.—Ondai Sair formation (paper shales), Ondai Sair, 
Mongolia. In parcel No. 79. 


Several specimens, apparently referable to a single species, have 
been found. So far as visible, the antenne and mouth parts strongly 
suggest the bloodsucking Culicoides group, but it would be going too far 
to affirm distinctly that this is a bloodsucking form. It is possible that 


1924] Cockerell, Fossils from Mongolia 143 


the bloodsucking habit among Diptera arose without reference to the 
Mammalia, as Simulium is known to attack young fish and Culicide 
have been seen to prey an cold-blooded vertebrates, reptiles and amphib- 
ians. Thus the opportunities for the development of such habits pre- 
sumably existed before there were any Diptera. 

C. gobiensis is evidently the species reported as ‘“‘mosquitoes,” but 
I think it is practically certain that ‘it does not belong to the Culicide. 
It is very different from Mesopsychoda dasyptera Brauer, Redtenbacher 
and Ganglbauer, from the Jurassic of Ust Balei, Siberia. 


COLEOPTERA 
There are two or three species of beetles in the Ondai Sair collection, 
but their preservation is so poor that I do not venture to describe them 
(Pl. II, fig. 8). The Carabites latecostatus of Reis is certainly not repre- 
sented. 


PLANTS 
GINKGOALES? 
Baiera furcata (Lindley and Hutton) 
Plate II, Figure 9 

The plant remains in the Ondai Sair shales are fragmentary and not 
readily referable to particular species. There is, however, one specimen 
which, so far as it goes, appears to agree exactly with Baiera furcata 
(Lindley and Hutton) Braun, which Seward calls B. lindleyana (Schim- 
per), disregarding priority. This B. furcata, or fossils indistinguishable 
from it, may be either Jurassic or Lower Cretaceous. Seward states that 
it is known from the Middle Jurassic of Chinese Dzungaria; ‘he is also 
disposed to refer to it the plant from the Lower Cretaceous of the Black 
Hills, which Fontaine recorded as Czekanowskia nervosa Heer. Baiera 
is in general highly characteristic of the Jurassic but rare in the Lower 
Cretaceous. 

From the Siberian Jurassic, Heer recorded a Trichopitys setacea. 
The genus Trichopitys of Saporta, with Bazera-like foliage, was originally 
based on a plant from the Permian, which appears to be related to the 
Ginkgoales. Seward has suggested that 7’. setacea does not belong to the 
Permian genus, but is to be compared with Baiera lindleyana. 

In the Ondai Sair shales are robust shoots bearing stalked obpyri- 
form bodies more or less comparable with the fruits of Trichopitys hetero- 
morpha as figured by Zeiller. In one specimen, at least, there is a distinct 
seedlike impression at the base, and on closer analysis there seems to be 


144 Bulletin American Museum of Natural History [Vol. LI 


nothing to separate the species from Phyllocladites Heer (Drepanolepis 
Nathorst), which occurs in the Jurassic of Spitzbergen. Seward defines 
the genus as consisting of strobili of open habit with single-seeded sporo- 
phylls, of uncertain systematic position but probably gymnospermous. 
The general resemblance to the fruiting shoots of the Ginkgoales is 
sufficiently close to suggest a possibility that these are after all the 
fruits of Bazera furcata, or at least belong with the foliage which cannot 
be separated from the latter species. There is, however, some indication 
of possible narrow bract scales, suggesting remote affinity with Lariz. 
It seems desirable to give this organism a name, so it may be introduced 
as follows: 


Phyllocladites (?) morrisi, new species 
Plate II, Figures 10, 11; Text Figure 6 

Speciric CHARACTERS.—Stout erect shoots, at least two inches long, the straight 
stiff stem about 2mm. wide about 35 mm. from apex; sporophylls on short stout stalks 
about 2 mm. long. Sporophylls 6 to 7 mm. long, about 3.5 
wide near base, obpyriform, with very obtuse apex. Ovules 
or seeds circular. 

Horizon AND Locatrry.—Ondai Sair formation (paper 
shales), Ondai Sair, Mongolia. In parcel No. 64. 

Named after Mr. Frederick K. Morris, who collected the 
material. 


I sent a sketch of this plant to Dr. E. W. Berry; 
Fig. 6. Phyllo- he kindly replied that he was not able to place it with 
cladites morrisi. any species known to him. 


Czekanowskia species 

Four contiguous broad-linear (1.5 mm. diameter) leaves, the portion 
preserved 56 mm. long, resemble pine needles, but apparently belong to 
Czekanowskia. It is impossible to refer them to any particular species. | 

Ondai Sair formation, ‘‘No. 64, Ondai Sair.”’ 

Reis has figured similar leaves, but apparently more slender, from 
Turga. He referred them to Czekanowskia sp. The genusis characteristic 
of Jurassic strata, but is reported from the North American Cretaceous. 


Puate I 


Figs. 1-9. Ephemeropsis trisetalis Eichwald. 

Fig. 1. Nearly complete nymph about 40 mm. long. 

Fig. 2. Part of a nymph, showing legs and abdominal gills. 

Fig. 3. Abdominal gills, detached from the nymph. 

Figs. 4, 5. Caudal appendages of nymph. 

Figs. 6, 7. Nymphal wing pads. 

Figs. 8, 9. Adult wing, basal portion. 

Fig. 10. Ephemeropsis melanurus, new species. Detached gills and caudal 
appendages. 


Vol. LJ, Plate I 


Bull. A. M. N. H. 


PuateE II 


Fig. 1. Lycoptera middendorffi Johannes Miiller. 

Fig. 2. Hstheria middendorffi T. R. Jones. 

Figs. 3, 4. 5. Cymatophlebia (?) mongolica, new species. Details of wing. 

Figs. 6, 7. Chironomopsis gobiensis, new species. 

Fig. 8. Beetle, undetermined. 

Fig. 9. Baiera. 

Figs. 10, 11. Phyllocladites (?) morrisi, new species. Near the base of Fig. 10 
specimen is the caddis case of Indusia reisi, new species, touching the stem of Phyllo- 
cladites. 

Fig. 12. Plant, undetermined. 


Plate II 


LI, 


Vol 


. A. 


N 


Bull. A. M. 


Some Fishes’ Collected by the Third Asiatic Expedition in 
China : 


By Henry W. Fow.Ler 


BULLETIN 
3 : : OF 
THE AMERICAN MUSEUM OF NATURAL HISTORY 
Vou. L, Art VIL, pp. 373-405 
New York — 
Issued December 31, 1924 


59.7(51.7) 
Article VII—SOME FISHES COLLECTED BY THE THIRD 
ASIATIC EXPEDITION IN CHINA! 


By Henry W. Fow.er? 


The fishes reported in this paper were sent to The American Mu- 
seum of Natural History by its Third Asiatic Expedition. They are 
represented by sevéral lots, chiefly from Hsing Lung Shan (Eastern 
Tombs region, Chihli Province, North China) and vicinity in August 1921 
and Ningkwo in the Province of An-hwei Central China during Sep- 
tember and October of the same year. All from these two localities 
were collected by Mr. Clifford H. Pope. Preliminary notices have been 
given of two new species.’ 

The specimens embraced in this study of the Third Asiatic Expedi- 
tion’s material number 1103, of which the majority are cyprinids. 
Descriptions are given of seventeen species of special interest, as imper- 
fectly or little known. The other species are noticed with condensed 
computations of their variation, often supplemented with discussion. I 
have also included several notes on types or rare species in the Academy 
of Natural Sciences of Philadelphia and used several series in their col- 
lections of both Chinese and Japanese specimens in comparison. Figures 
illustrating two new cyprinids that I have previously described, a little- 
known loach, and some of the salient color-variants of the Chinese Cobitis 
tenia are given. 

Acknowledgment is due to Mr. John T. Nichols and The American 
Museum of Natural History for the loan of this material for study and 
the reservation of a set of the duplicate specimens for the museum of the 
Academy of Natural Sciences of Philadelphia. 


ENGRAULIDE 
Mystus nasus (Schlegel) 


Head contained 6% to 6% times in length to base of caudal; depth, 6 to 6%. 
Dorsal m1, 11,1; anal 11, 92 to 107; pectoral v1, 11 or 12; ventral1, 6. Scales 72 to 76 
in lateral line to caudal base, and 3 more on latter; 12 or 13 scales transversely 
between dorsal and ventral origins; 18 to 23 predorsal scales; abdominal scutes 19 or 
20 to ventral origin, 28 to 33 following. Head width 2% to 3 times in its length; 
snout 4% to 54; eye 54 to 6; maxillary % to 14; interorbital 3% to 3%; first branched 
dorsal ray 1% to 14; length of caudal 1% to 1%; pectoral 1% to 1%; ventral 2% to 24. 


a ed of the Asiatic Expeditions of The American Mugeum of Natural History. Contribu- 
ion No. 38. 

2Of the Academy of Natural Sciences of Philadelphia. 

%American Museum Novitates, No. 38, May 25, 1922, pp. 1-2; and No. 83, July 25, 1923, pp. 1-2. 


373 


374 Bulletin American Museum of Natural History [Vol. L 


Body strongly compressed, deepest at dorsal and ventral origins, predorsal edge 
convex and postdorsal with only slight median keel. 

Head acuminate, strongly compressed, flattened sides but slightly approximate 
below. Snout protrudes rather obtusely, length %4 its width. Eye without distinct lid, 
adipose tissue covering over; center about first fourth in head, slightly more backward 
in smaller example. Mouth large, narrow, mandible tip midway in snout length. 
Maxillary very long, extends beyond angle of preopercle, though not quite to gill- 
opening in younger example and to below gill-opening and slightly beyond pectoral 
origin in larger example. Teeth in jaws or along maxillary its entire length, and in 
mandible edges uniserial; small patch each side of vomer and another moderate uni- 
serial row on each palatine. Tongue small free knob, advanced, edentulous. Nostrils 
together, front one at last third in snout, and hind one much the larger. Interorbital 
broadly convex. Opercles smooth. 

Gill-opening extends forward opposite hind eye edge, connecting membranes 
moderate. Gill-rakers 18 or 19+-22, fine, lanceolate, equal in length to gill-filaments, 
which are as long as eye. Pseudobranchiz \ the length of gill-filaments. Isthmus a 
long, slender, trenchant keel. 

Scales caducous, thin, rather narrowly imbricated, cycloid. Row of broad scales 
extends over greater part of anal basally, also some small scales on caudal base. 
Scaly flap in ventral axil 7% length of fin. Scales with 1 or 2 basal radiating strie; 
circuli basally 114 to 120; apical striz 9 or 10, fine, largely reticulate. Abdominal 
scutes sharply pointed. 

Dorsal inserted slightly behind ventral origin, first branched ray longest and 
depressed backward farther than others. Anal low, mostly uniform, begins at first 
*, in combined head and trunk, excluding caudal, slightly behind middle in smaller 
example. Caudal continuous with anal, median upper rays longest, forming a point. 
Pectoral without filaments, reaches slightly beyond ventral; longest filaments reach 
midway in combined head and trunk, without caudal, to a distance contained 2% 
times in head and trunk, in smaller example, though in all cases at least to anal origin. 
Ventral reaches 4 the distance to anal, nearly in young. Vent close before anal. 

Color in alcohol, faded pale brown, sides and lower surface very pale. Slight 
neutral-brown tint at predorsal. Iris pale slaty (doubtless silvery white in life). Fins 
pale, slightly dotted with pale dusky on dorsal basally. 

Length 283 to 310 mm. 


Five examples from Ningkwo. 

Kreyenberg and Pappenheim! discuss the variation of the length of 
the maxillary, contending for Cozlia brachygnathos that ‘unsere simt- 
lichen Exemplare grosse wie kleine, zeigen nur das obige Verhalten.” It 
seems quite likely, therefore, that smaller or younger examples usually, 
if not always, have shorter maxillaries. 


FLUTIDZ 


Fluta alba (Ziuew) 


Eleven from Ningkwo, 335 to 533mm. All gray-brown, paler below. 
Sides and above variously though finely reticulate, spotted or marbled 


11909, Abh. Ber. Mus. Magdeburg, II, Heft 1, p. 10. 


1924] Fowler, Some Fishes Collected in China 375 


with dusky-brown. Often a dark median lateral line more or less com- 
plete. In some examples dark mottling extends more or less completely 
over belly and under surface, sometimes only for short space behind 
gill-opening. Again, dark specks on back obsolete or absent and usually 
three dark median parallel lines down back. Some examples may have 
spots very fine or absent so back uniform grayish in appearance and belly 
pale. 
MASTACEMBELID 


Mastacembelus sinensis (Bleeker) 


Head contained 6 to 644 times in length to base of caudal; depth 10% to 12. 
Dorsal spines XXXIV, vary XXXI to XXXV; anal spines III. Snout 3! to 3% 
times in head; eye 8 to 84; mouth cleft 345 to 444; interorbital 8 to 9. 

Body well compressed, deepest just before vent. Head width 3 to 37 times in its 
length. Snout conic, compressed, ends in a fleshy tip, width 2 to 2 in its length. 
Eye center at about first third in head; diarneter 24 to 245 in snout, little greater 
than interorbital. Jaws slender, lower shorter, inferior. Maxillary nearly reaches 
opposite eye. Teeth small, conic, in narrow bands in jaws. Lips fleshy, moderate. 
Nostrils appear as a pair of short fleshy tubes each side of snout tip; hind nostril a 
simple pore at last sixthin snout. Interorbital nearly level. Preorbital spine strong, 
below front of eye. Tail 134 to 1°4 times in combined head and trunk. Pectoral 4 
times in head; caudal 2) to 3. 

Color in aleohol brownish generally, under surface of head and belly paler to 
creamy. Broad dark or dusky-brown lateral band, starting as line from snout tip 
across side of head, then broadening over greater part of side to caudal base. This 
band sometimes entirely uniform, speckled, dotted irregularly, or with 40 or more 
narrow pale vertical lines. Back with dusky reticulations, also lower side of abdomen. 
Some may extend well over basal portions of soft vertical fins. Often basal portions 
of last more or less dark or neutral and edges narrowly pale. Belly often immaculate 
white, sometimes finely reticulated with brownish. Some examples largely uniform 
or with markings very faint. Often a distinct pale streak runs along upper side of 
back delimiting upper border of dark lateral band and this with more or less regular 
indentures to form pale transverse lines. 

Length 126 to 228 mm. 


Seventeen from Ningkwo. 


CATOSTOMIDZ 


Myxocyprinus asiaticus (Bleeker) 


Head contained 4% times in length to base of caudal; depth 24. Dorsal v, 
48,1; analry, 10,1; pectoral 1, 17; ventral 1, 10. Scales 50 in lateral line to caudal 
base and 3 more on latter; 13 scales above lateral line, 8 below; 19 predorsal scales. 
Head width 1} times in its length; snout 20; eye 6; least depth of caudal peduncle 
2: first branched dorsal ray 24%; first branched anal ray 149; lower caudal lobe 34; 
pectoral 444; ventral 444. 

Body strongly compressed; back greatly compressed, elevated and its front 
profile strongly inclined, edge but slightly keeled; greatest depth at origin of dorsal. 
Caudal peduncle strongly compressed, long as deep. 


376 Bulletin American Museum of Natural History [Vol. L 


Head small, moderately compressed sides but little flattened and scarcely 
approximated below. Snout obtuse, convex over surface, length % its width. Eye 
small, midway in head length; diameter 24 in snout, 3 in interorbital. Mouth 
small, inferior. Premaxillaries protractile downward. Lips moderately small, fleshy, 
plaited. Nostrils small, together; front one pore-like, with cutaneous rim, at last 
fourth in snout; hind one a crescentic slit. Interorbital broadly convex. Suborbitals 
rather narrow. 

Gill-opening extends forward to opposite last fourth of head. Gill-rakers 13+19, 
flexible, lanceolate, contained 24 times in gill-filaments, which equal eye in length. 
Pseudobranchiz rather small, about % the length of gill-filaments. 

Seales in even longitudinal series; breast and belly covered with small scales; 
row of smaller basal scales on anal and several rows on caudal; head naked and a 
narrow naked median predorsal strip; no scaly axillary flaps. Scales with 22 to 33 
basal radiating striz; basal circuli 61 to 66; apical radiating strie 13 to 15. Lateral 
line complete, midway along side from suprascapula to caudal base; tubes slender, 
simple and well exposed over each scale exposure to its hind edge. 

Dorsal very long basally, inserted slightly nearer pectoral origin than ventral; 
front lobe of fin 1% its length, rounded. Anal begins well POPE ye much nearer 
caudal base than ventral origin; depressed fin reaching about 4 in caudal; first 
branched ray longest. Caudal deeply forked, lower lobe a little the longer. Pectoral 
broad, not quite reaching ventral. Ventral like pectoral, reaches % to anal. Vent 
close before anal. 

Color in alcohol, deep dusky-brown, center of each scale with darker shade to 
form slightly darker longitudinal streaks. Neutral-slaty on cheek and below eye, 
also upper hind side of head, and on all fins. Upper caudal lobe, lips, chin and 
under surface of head pale or whitish. 

Length 288 mm. 


One example, bought twenty miles from Ningkwo, in Ching Tsui 
Ho on way to Wuhu. 


CYPRINIDZ 
Cyprininze 
Carassius auratus (Linnzeus) 

Head contained 24 to 3% times in length to caudal base; depth 2% to 2. 
Dorsal 11, I, 16,1 to 18,1; analu, I, 5,1. Scales 26 to 28 in lateral line to caudal base 
and 2 or 3 more on latter; 7 scales above lateral line, 6 below; 13 predorsal scales. 
Snout 3% to 3% times in head; eye 3% to 5; maxillary 34 to 3%; interorbital 24 
to 256. 

Color in alcohol dull brownish generally above, below little paler. Fins grayish, 
sometimes ends of paired ones a little darker. 

Length 115 to 153 mm. 

From Hsing Lung Shan August 7, 1921, 93 examples; from twenty- 
six miles south of Hsing Lung Shan, August 12, 1921, 83 examples; 
Ningkwo, 3 examples. 


1924] Fowler, Some Fishes Collected in China SOee 


Barbinz 
Hemibarbus labeo (Pallas) 


Head contained 3% to 3% times in length to caudal base; depth 4 to 5. Dorsal 
III, 7, 1; anal 1, 6,1. Scales 40 to 48 in lateral line to caudal base, and 3 more on 
latter; 6 or 7 scales above lateral line, 5 or 6 below; 11 to 13 predorsal scales. Snout 
2 to 2% times in head; eye 3% to 4%; maxillary 2% to 3; interorbital 3% to 4; 
first branched dorsal ray 1% to 1%; first branched anal ray 1% to 1%4; caudal 17g to 
1%; pectoral 1% to 14; ventral 1% to 2. Pharyngeal teeth 1, 3, 5—5, 3, 1, hooked, 
compressed, with well-developed grinding surfaces. Scales with 15 to 19 obsolete 
weak apical striz; circuli moderate, finer basally. 

Color in aleohol with row of 8 or 9 deep brown spots close along and above lateral 
line, most distinct or contrasted in young. On back and sides base of each scale with 
dark spot. Dorsal and caudal with several rows of dusky spots on outer portions. 

Length of largest 244 mm. 

Two from Hsing Lung Shan, August 7, 1921; six from Ningkwo, 
September 15 to October 15, 1921. 

For comparison only a few young examples in the Academy from 
the Iwaii River at Ichinoseki, Japan, representing the nominal Gobzo 
barbus Schlegel, which appears in every way the same, as admitted by 
Berg. These examples show: 

Head contained 3% to 3% times in length to base of caudal; depth 4 to 434. 
Dorsal 1, I, 7,1; anal 1m, 6, 1, rarely m1, 7,1. Scales 40 to 42 in lateral line, occa- 
sionally 39 or 44, to caudal base and 2 or 3 more on latter; 8 scales, often 7, above 
lateral line; 6 scales below lateral line, seldom 5; 13 to 17 predorsal scales. Snout 
2% to 24% times in head; eye 2% to 34%; maxillary 3% to 34; interorbital 4 to 4/4. 
Pharyngeal teeth 1, 3, 5—5, 3, 1, seldom 1, 2, 5—5, 2, 1 or 1, 2, 5—5, 3, 1. 

Length 54 to 62 mm. 

Oshima says that Hemibarbus maculatus! from China differs only in 
color. In alcohol the nominal Hemibarbus joitent Jordan and Starks is 
said to be “pinkish yellow, with a longitudinal series of eight large spots 
above the lateral line; smaller spots irregularly placed on back and sides; 
dorsal and caudal with similar black spots; other fins without markings. 
Although faint dark spots are present in the young specimen of Hemz- 
barbus labeo, they are not permanent; the color of the adult is always 
uniform grayish-brown.”’ My examples show traces of all these mark- 
ings, though now apparently greatly faded. 


Gobionins 
Gobio gobio (Linnzeus) 


Head contained 3% to 3% times in length to base of caudal; depth 4% to 4%. 
Dorsal u, 7,1; anal 1, 6,1. Scales 38 in lateral line to caudal base and 2 more on 
latter; 6 scales above lateral line, 5 below; 14 to 18 predorsal scales. Snout 234 


11919, Ann. Carnegie Mus., XII, Nos. 2-4, p. 212. 


378 Bulletin American Museum of Natural History [Vol. L 


to 2% in head length; eye 34 to 54; maxillary 344 to 34; interorbital 3% to 4. 
Pharyngeal teeth, 3, 5—5, 3, hooked, with well-developed grinding-surfaces. Scales 
with 13 to 18 apical radiating stri#; superior and inferior circuli coarse, basal very 
fine. 

Color in alcohol brownish-olive above, with row of 6 to 10 median lateral black- 
ish blotches, usually closer behind, and lower whitish surface strongly contrasted. 
On back, following scale courses longitudinally, there are rows of dusky spots, mostly 
appearing as longitudinal lines. Dusky streak from side of snout below nostrils to 
eyes. Dorsal and caudal grayish with 5 to 7 transverse rows of blackish spots, on 
latter more aslines. Blackish blotch in pectoral axil. Lower fins and barbels whitish. 

Length 48 to 100 mm. 

Forty-three from Hsing Lung Shan, August 7, 1921; one from Ning- 
kwo with the depth of the body contained 6% in the length to the caudal 
base, and lips and lower front surfaces of the barbels strongly papillose, 
sides also with five dark blotches. 

A comparison of the European material in the Academy shows: 

Head contained 34 to 4 times in length to base of caudal; depth 334 to 44. 
Dorsal m1, 7,1; anal m1, 6,1. Scales 34 to 40 in lateral line to caudal base and 2 or 
3 more on latter; 5 to 7 scales above lateral line, usually 6; 4 or 5 scales below lateral 
line; 14 to 17 predorsal seales. Snout 2 to 24 times in length of head; eye 3 
to 4%; maxillary 2% to 344; interorbital 3% to 444. Pharyngeal teeth 3, 5—5, 3. 

Length 70 to 172 mm. 

These examples from Niirnberg, Germany; Leyden, Holland; Lake 
Lucerne, Switzerland; Arno River, Italy. 

Several little-known species have been described from China. Gobzo 
argentatus Sauvage and De Thiersant! differs in many respects from Gobio 
gobio, as the presence of seven branched anal rays, two rows of scales 
between the lateral line and the base of the ventral fin; and the color- 
pattern and fins without spots, though a bluish lateral band is present. 
Kreyenberg and Pappenheim identify examples with Gobzo argentatus,? 
but they give the head as contained 4 to 444 times in the body and 3% to 
4 seales below the lateral line. In most every other way their specimens 
appear to agree with mine. 

Gobio nitens Giinther? is described with five scales below the lateral 
line, though with but two rows between the lateral line and the base of 
the ventral fin. It is, however, without barbels, in which it agrees with 
Gobio imberbis Sauvage and De Thiersant,* which in turn differs. 

fobio nigripinnis Giinther® is another species without barbels but, 
as its specific name indicates, is with blackish fins. 


11875, Ann. Sci. Nat., (6) I, Zool., p.9. Yang-tze Kiang. 

21908, Sitzs. Ges. Naturf. Freund. ee 1908, p. 97. Tuntingsee, Hankau. 
31873, Ann. Mag. Nat. Hist., (4) XII, p. 246. ‘Shanghai. 

41875, Ann. Sci. Nat., (6) IJ, Zool., p. 9. Yenkiatsoun (S. Shen-si). 

51873, Ann. Mag. Nat. Hist., (4) NUT, p. 246. Shanghai. 


1924] Fowler, Some Fishes Collected in China 379 


Gobio nummifer Boulenger! does not seem to differ from Gobzo gobio. 
It is described with six round black spots along the body and tail above 
the lateral line. 


Gobio wolterstorfii Regan 


Head contained 2*4 to 3% times in length to base of caudal; depth 4 to 5. Dorsal 
mi, 7,1; anal i, 6,1. Scales 34 in lateral line to caudal base and 2 more on latter; 
5 scales above lateral line, 4 below; 12 predorsal scales. Snout 3! to 3! times in 
length of head; eye 34 to 34; maxillary 3 to 34; interorbital 3 to 34. Pharyngeal 
teeth 2, 5—5, 3, hooked, some of upper broadly compressed, with broad grinding 
surfaces. Scales with 15 to 18 apical radiating strie; circuli above and below mod- 
erate, converging finely basally. 

Color in alcohol pale brownish above, with olive tinge on back, which mottled 
coarsely with blackish spots very irregularly. Sides and under surface whitish, with 
silvery reflections on sides of head. Each tube of lateral line with black spot close 
above and below. Iris silvery. Fins all pale, lower whitish. 

Length 20 to 93 mm. 


From Hsing Lung Shan, August 7, 1921, 2 examples; from twenty- 
six miles south of Hsing Lung Shan, August 12, 1921, 25 examples. 


Pseudogobio rivularis (Basilewsky) 


Head contained 3% times in length to base of caudal; depth 444 to 54. Dorsal 1, 
6, lor, 7,1; anal, 5,1. Seales 31 to 35 in lateral line to caudal base and 2 more on 
latter; 5 or 6 scales above lateral line, 4 below; 11 or 12 predorsal scales. Snout 2) 
to 2% times in length of head; eye 3’ to 449; maxillary 3% to 334; interorbital 34 
to 3%. Pharyngeal teeth 2, 4—4, 2, hooked, with narrow grinding-surfaces. Scales 
with 7 to 10 apical radiating strix; circuli coarse above and below, converging finely 
basally. 

Color in aleohol pale brownish on back, mottled obscurely with darker.. Back 
with 5 broad dark saddles, each a little narrower than interspaces. Sides with about 
7 very pale or obscure round blotches made up of minute grayish dots, more or less 
uniform in size and about equal to eye-diameter, distributed along middle of side or 
on line of vertebral axis. Dusky line from side of snout to eye and narrower less 
defined line from snout tip to above nostrils. Fins pale. Dorsal with 4 series of dusky 
spots, each dark blotch on rays only. Caudal only with 4 transverse dark bands, each 
inclined backward, and a small conspicuous median black spot, much less in diameter 
than half of pupil, at caudal base. Other fins all pale to whitish, though each with 
pale to obscure dusky median blotch or shade. 

Length 43 to 56 mm. 


From twenty-six miles south of Hsing Lung Shan, August 7, 1921, 
3 examples. 

Giinther notes under Pseudogobio sinensis that the black caudal 
spot disappears with age.” 


2 


€ 


11901, Proc. Zoél. Soc. London, I, p. 269, 
p. 


Pl. xx111, fig.2. Ningpo. 
21873, Ann. Mag. Nat. Hist., (4) XII, 47. 


380 Bulletin American Museum of Natural History [Vol. L 


Saurogobio dabryi Bleeker 


Head contained 3% to 4% times in length to base of caudal; depth 4%4 to i. 
Dorsal 111, 8, 1, once 111, 7,1; anal 11, 6,1. Scales 36 to 47 in lateral line to caudal 
base and 2 more on latter; 4 to 6 scales above lateral line, 3 or 4 below; 11 to 13 
predorsal scales. Snout denenmned 2 to 3 times in length of head; eye 3% to 5; maxil- 
lary 3% to 3745 interorbital 34 to 44. 

Body moderately compressed, slender, long, deepest at dorsal origin. Caudal 
peduncle well compressed; its least depth contained 2 to 2% times in its length, 2% 
to 4% in head. 

Head saree slightly compressed, with flattened sides scarcely constricted 
below, width 1°4 to 14 in its length. Snout conic, upper profile with a slight depres- 
sion just behind its tip, as long as broad; with age its width is contained 1% times 
initslength. Eye elevated, its center midway in length of head; its length contained 
1% to 2 times in length of snout, 144 in interorbital, except that the latter is less than 
the eye in young. Mouth small, inferior, mandible depressed. Maxillary concealed, 
reaches front nostril, and with a fleshy barbel from subterminal outer face, 1% to 
1% in eye. Lips narrow, soft, fleshy, coarsely papillose. Nostrils together, within 
last fourth of snout; hind one much the larger and mostly concealed by posterior 
cutaneous flap of front one. Interorbital broadly concave. Suborbitals narrow, pre- 
orbital longer than eye. Occipital fontanel as long as eye. 

Gill-opening extends forward opposite hind eye edge. Gill-rakers 2+14, 
short, low, compressed papilla, greatly less than gill-filaments, which contained 1%4 
in eye. Pseudobranchie % of gill-filaments. Inside of pharynx coarsely papillose. 
Pharyngeal teeth small, compressed, or pointed without terminal hooks or grinding- 
surfaces. 

Seales with 8 to 56 slightly waved apical strize, about 8 to 34 in young; apical 
circuli 35 to 38, imperfect. Scales firmly adherent, in even longitudinal rows, well 
exposed; few scales on caudal base; breast, Scan part of belly medially and head 
naked. Scaly pointed flap in ventral axil, 2% in fin. Lateral line slopes a little low 
at first, then runs midway over ventral and along side of caudal peduncle to caudal 
base. Tubes in lateral line slender, simple, half way over each scale exposure. 

Dorsal origin midway between snout tip and anal origin, upper fin edge slightly 
concave; depressed fin contained 1°4 to 134 in the distance to caudal base; first 
branched ray 1% to 144in head. Anal origin a little nearer caudal base than last dorsal 
ray base; depressed fin contained 1% to 1%4 times in the distance to caudal base; first 
anal ray 144 to 2{ in head. Caudal deeply forked, pointed lobes equal, 4% to 4% 
in combined head and trunk. Pectoral reaches 1% to 1% to ventral, length 1¥%o to - 
1% in head. Ventral inserted below third to fifth dorsal ray bases, dened fin con- 
tained 2 to 2% times in distance to anal origin; fin contained 1% to 1% in head. 

Color in AIeHE pale olivaceous-brown, edge of each scale broadly and slightly 
darker. Grayish median lateral streak close along and above lateral line, with about 
10 to 14 obscure darker spots. Dorsal and caudal grayish terminally, other fins 
whitish. Iris brownish. 

Length 26 to 252 mm. 


From Hsing Lung Shan, August 7, 1921, 6 examples; twenty-six 
miles south of Hsing Lung Shan, August 12, 1921, 8 examples; Ningkwo, 
An-hwei Province, 14 examples. 


1924] Fowler, Some Fishes Collected in China 381 


Longurio athymius Jordan and Starks! is synonymous. It is\-hown, 
apparently erroneously, with the breast and chest scaled. 


Paraleucogobio notacanthus Berg 


Head contained 3% to 3% times in length to base of caudal; depth 3% to 3%. 
Dorsal 11, 7,1; anal m1, 6,1. Scales 35 to 36 in lateral line to caudal base and 2 or 
3 more on latter; 5 scales above lateral line, 4 below; 12 or 13 predorsal scales. 
Snout 3% to 3% times in its length; eye 4 to 4%; maxillary 3% to 34; interorbital 
2% to 2%. Pharyngeal teeth, 3, 4—5, 3, hooked, with grinding-surfaces. Scales with 
10 to 20 basal radiating striz; circuli moderate, finely convergent basally. 

Color in alcohol, back olivaceous, each row of scales marked with a longitudinal 
dusky band, made up of blackish dots. Along middle of side longitudinal blackish 
band, little less in width than eye. Dorsal grayish, with blackish blotch above for- 
ward. Caudal grayish, with dusky clouding basally in center. Other fins all more or 
less whitish. 

Length 68 to 90 mm. 


From Hsing Lung Shan, August 7, 1921, 27 examples. 


Sarcocheilichthys lacustris (Dybowski) 


Head contained 4% to 4% times in length to base of caudal; depth 3% to 3%. 
Dorsal 111, 7,1; anal 1, 6,1. Scales 38 to 40 in lateral line to caudal base and 3 or 4 
more on latter; 6 or 7 scales above lateral line, 5 or 6 below; 14 to 16 predorsal scales. 
Snout 236 to 2% in head length; eye 4% to 44; maxillary 3°4 to 4%; interorbital 2% 
to 2h. 

Body compressed, moderately robust, deepest at dorsal origin and edges all 
convexly rounded. Caudal peduncle well compressed, its least depth 1% to 14 in 
its length or 1% to 1% in length of head. 

Head small, compressed, flattened sides not convergent especially above or 
below, lower profile a little more inclined; width 1% to 1%4 in its length. Snout 
obtusely convex, length % to 44its width. Eye with hind pupil edge midway in length 
of head; 1} to 1% in snout, 2 to 2 in interorbital. Mouth inferior, small. Maxillary 
concealed, reaches opposite hind nostril. Small barbel on maxillary above near end 
of expansion. Lips smooth, fleshy, not extending across mandible, folded deeply at 
each side. Mandible spatulate, horny, firm and firm bony median surface opposite 
in middle of upper jaw. Nostrils together, in last third of snout; front one a simple 
pore with broad marginal posterior flap, exposing hind one in a crescent. Interorbital 
broadly convex. 

Gill-opening extends forward opposite last third of head, not to eye. Gill-rakers 
2+4, short rudimentary tubercles. Gill-filaments as long as eye. Pseudobranchize 
%f the length of gill-filaments. Pharyngeal teeth, 2, 5—5, 2, compressed, hook at end 
of each and grinding-surfaces moderate. 

Scales with 38 to 40 apical radiating strie; 48 basal circuli. Scales firmly ad- 
herent in even longitudinal rows, all well exposed; caudal base scaled; small scales 
cover breast; axillary ventral scale % the length of fin. Lateral line slopes slightly 


11905, Proc. U.S. Nat. Mus., XXVIII, p. 197, Fig. 3. Chemulpo. 


382 Bulletin American Museum of Natural History [Vol. L 


at first until midway in depth; tubes simple, small, each one exposed over basal third 
of scale. Row of few tubercle scars around border of upper jaw above lip, also few 
below nostril and front of eye below. 

Dorsal origin much nearer snout tip than caudal base, upper edge of fin concave; 
length of fin contained 2 to 2 times in the distance to caudal base; first branched 
dorsal ray 4 to 443 in combined head and trunk. Anal inserted close behind depressed 
dorsal tip, or a little nearer caudal base than dorsal-origin; length of depressed fin 
1% to 1% in the distance to caudal base; first branched anal ray 1} to 1% in head. 
Caudal deeply forked, lobes broad, even; length of fin contained 3°5 to 3% times in 
combined head and trunk. Pectoral reaches *4 to 44 to ventral; length of fin 1 to 1% 
in head. Ventral inserted slightly behind dorsal origin, depressed fin reaches %4 to 
4 to anal; length of fin 144 to 1}4 in head. 

Color in aleohol deep brown generally, each row of scales longitudinally with 
deeper or median dusky streak. Fins all with neutral-slaty or dusky, especially 
terminally and edges narrowly whitish. Bases of all fins narrowly whitish. Under 
surface of head and belly somewhat pale. 

Length 150 to 179 mm. 


Fifteen examples from Ningkwo. 


Pseudorasbora parva (Schlegel) 


Head contained 3% to 3/% times in length to base of caudal; depth 34 to 436. 
Dorsal 11, 7,1; anal 1, 6,1. Scales 33 to 35 in lateral line to caudal base and 2 more 
on latter; 5 or 6 scales above lateral line, 4 to 6 below; 13 to 15 predorsal scales. 
Snout 3 to 3% in head; eye 3 to 5; maxillary 34 to 4; interorbital 2% to 234. Pharyn- 
geal teeth 5—5, hooked, with grinding-surfaces; bones quite small. Scales with 2, 3 
or 4 apical radiating striz; circuli moderate. 

Color in aleohol brownish generally, on back and upper sides each scale as if 
with median brownish blotch. All fins with more or less grayish to dusky terminally. 
Sometimes dark pigment spots concentrated along lateral line to form median lateral 
dark streak, resolving in dark spot just before caudal base. One example shows 
terminal portions of all fins, caudal broadly bordered so behind, with dusky to black- 
ish, also sort of transverse or horizontal basal blackish band, as if made up of black 
pigment dots. Length 28 to 91 mm. 

Hsing Lung Shan, August 7, 1921, 110 examples, and south of 
Hsing Lung Shan, August 12, 27 examples. 

In the very young the front dorsal edge is blackish and there is a 
narrow blackish lateral band from side of snout to caudal base. It also 
has the lateral line very incomplete, only on the first few anterior scales. 
One young example is without mandibles, its jaws abortive, though both 
nostrils still in front of the eyes; or in profile what remains of the snout 
is less than half the eye. 

For comparison with the Japanese specimens in the Academy, there 
is a series represented by many specimens from Yodo River in Osaka, 
Lake Biwa at Matsubara, Iwai River at Ichinoseki and near Nagoya in 
Owara. These show the following. 


1924] Fowler, Some Fishes Collected in China 383 


Head contained 3 to 44 times in length to base of caudal; depth 3% to 4. Dor- 
sal 11, 7,1; anal 1, 6,1. Scales 29 to 36 in lateral line to caudal base and 2 or 3 more 
on latter; 5 or 6 scales above lateral line, 4 below; 12 to 15 predorsal scales. Snout 
3M to 3% in head; eye 3% to 434; maxillary 3% to 4%; interorbital 2% to 244. Pharyn- 
geal teeth 5—5, rarely 1, 5—6. 

Length 38 to 90 mm. 


Rasborine 
Aphyocypris chinensis Giinther 
Figure 1 


Head contained 3% times in length to base of caudal; depth 3/g. Dorsal m1, 
7,1; anal in, 6,1. Scales 33 in lateral line to caudal base and 2 more on latter; 6 
scales above lateral line, 4 below; 13 predorsal scales. Snout 3 times in head measured 
from upper jaw tip; eye 4; mouth width 3}4; interorbital 2). 

Body compressed, fusiform, deepest at dorsal origin, edges convexly rounded. 
Caudal peduncle strongly compressed, least depth 1%3 times in its length or 2 in total 
length of head. 


Fig. 1. Aphyocypris chinensis Giinther. Ningkwo. 


Head conic, upper profile slightly concave, flattened sides but little constricted 
below; width 1% in its total length. Snout conic, little depressed above, length 74 its 
width. Eye with hind pupil edge midway in total head length; 14 times in snout, 1°4 
in interorbital. Mouth superiorly terminal, greatly inclined, with short gape, man- 
dible well protruded. Maxillary concealed, not reaching opposite nostril. Lips firmly 
coriaceous, lower broad and trenchant. Nostrils together, within last third of snout; 
front one a pore, with broad hind cutaneous flap exposing posterior nostril in crescent. 
Interorbital broadly convex. Suborbitals moderate, invade little over half of cheek 
to preopercle ridge. Opercle smooth. 

Gill-opening extends forward to opposite last third of head. Gill-rakers 4+12, 
short points, rudimentary, greatly less than gill-filaments, which nearly equal eye. 
Pseudobranchiz 7% of gill-fillaments. Pharyngeal teeth 3, 5—5, 3, hooked and with 
broad grinding-surfaces. 


384 Bulletin American Museum of Natural History [Vol. L 


Scales with 33 to 36 waved radiating strie apically; 32 to 36 basal circuli. 
Scales firmly adherent, in even longitudinal series, all well exposed, scarcely smaller 
on breast and caudal peduncle; axillary ventral scale 4 the length of fin. Head naked. 
Lateral line extends along middle of side, complete; tubes extend over half of each 
scale exposure, slender, simple. 

Dorsal origin midway between snout tip and caudal base, upper edge of fin con- 
vex; depressed fin reaches half way to caudal base. Anal inserted a little behind 
dorsal base, lower edge of fin convex; depressed fin reaches a distance contained 1% 
times in that to caudal base. Caudal deeply emarginate; lobes broad and pointed. 
Pectoral low, reaches % to ventral. Ventral inserted slightly before dorsal origin, 
reaches % to anal. Vent close before anal. 

Color in alcohol, back dull olive-brown, each scale with darker submarginal 
blotch. Dusky streak from snout tip back from eye and along middle of side to 
caudal base, where expanded, much broader. Its entire course made up of dusky 
specks or dots. Lower surface of head and trunk whitish. Dorsal pale, a little below 
its middle, a short dusky bar close behind each ray. Median portion of each anal ray 
with slight gray-dusky shade. Caudal pale gray. Paired fins whitish. Iris silvery- 
white. 

Length 100 mm. 


One example from Ningkwo. 


Xenocypridine 
Xenocypris davidi Bleeker 


Head contained 4% to 4% times in length to base of caudal; depth 3% to 3%. 
Dorsal III, 7,1; anal mr, 10,1 or 11,1; pectoral 1, 16 or 17; ventral1, 8. Scales 60 
to 64 in lateral line to caudal base and 3 more on latter; 12 scales above lateral line, 
7 below; 26 to 29 predorsal scales. Head width 2 to 2 in its length; head depth 1% 
to 134; snout 3% to 34; eye 34 to 4; maxillary 4%; interorbital 2% to 3. 

Body elongately ovoid, strongly compressed, deepest at dorsal origin, and edges 
all convexly rounded. Caudal peduncle strongly compressed, least depth 1% in its 
length, or 2 to 2% in head. 

Head well compressed, flattened sides slightly approximated below, upper profile 
nearly straight to slightly convex, and lower profile similar to slightly more inclined. 
Snout convex, rather obtuse, length 4 to % its width. Eye moderate, hind edge 
slightly nearer gill-opening than snout tip; 1} in snout, 134 in interorbital. Mouth 
small, inferiorly terminal, width 1% to 14 in eye. Maxillary largely concealed by 
preorbital, at least its expansion, which 4 in eye, reaches opposite front nostril. Lips 
moderate, entire. Edge of upper jaw firmly coriaceous. Nostrils together, front one at 
about last third in snout, pore-like, and hind one a crescentic slit close behind, formed 
by cutaneous marginal flap of anterior. Interorbital broadly and evenly convex. 
Suborbital chain moderate; preorbital depth about 4% its length which 134 in eye. 
Opercle finely striate. 

Gill-opening extends forward to opposite hind eye edge. Gill-rakers 12+-40, 
short, thin, compressed, triangular, 14 the length of gill-filaments, which 1% in eye. 
No pseudobranchiz. Pharyngeal teeth 2, 4, 6—6, 4, 2, strongly compressed, without 
hooks, and all with grinding-surfaces, outer quite broad. Some teeth deciduous, 
as several loose ones of outer set to each base. 


1924] Fowler, Some Fishes Collected in China 385 


Seales with 12 apical radiating striz, often 0 to 3 apical marginal auxiliaries; 
basal circuli 55 to 60. Scales in even longitudinal series parallel with lateral line, 
scarcely smaller on predorsal; slightly smaller scales on breast, belly and caudal 
base; pointed scaly flap in ventral axil % the length of fin. Head naked. Lateral line 
complete, well decurved, passes lower % in depth at ventrals and ascends along side 
of caudal peduncle to caudal base medially; tubes simple, extend over each scale 
about midway to * of exposure. 

Dorsal origin midway between snout tip and caudal base, first 3 rays spine-like, 
osseous, compressed and third greatly longest; first branched dorsal ray 14 to 1% 
in head; depressed fin reaches a distance contained 2% to 2% times in that to caudal 
base. Anal begins well behind depressed dorsal tip, first branched ray highest and 
forms apex of slight anterior lobe; first branched anal ray 1% to 2 in head. Caudal 
strongly forked, lobes sharply pointed, slender, equal, little longer than head. Pec- 
toral reaches a distance contained 144 to 1% times in that to ventral, contained 14 in 
head. Ventral inserted opposite dorsal origin, reaches a distance contained 1% to 
1% times in that to anal; 1% to 144in head. Vent close before anal. 

Color in alcohol with back tinged with pale olive, sides and lower surface pale to 
whitish. Dorsal and caudal pale olivaceous, lower fins whitish. Iris slaty and whitish. 

Length 173 to 228 mm. 


Six examples from Ningkwo. 


Rhodeine 
Acanthorhodeus guichenoti Bleeker 


Head contained 3°4 to 4 times in length to base of caudal; depth 2 to 2%. Dorsal 
III, 16,1 to 18,1; anal II, 11,1 to 14,1. Scales 34 or 35 in lateral line to caudal base 
and 3 or 4 more on latter; 6 or 7 scales above lateral line, 5 or 6 below; 14 or 15 pre- 
dorsal scales. Snout 34 to 34 in head; eye 3% to 34; maxillary 34 to 4; interorbital 3. 

Body strongly compressed, deeply ovoid, predorsal scarcely trenchant and all 
other edges rounded convexly; greatest depth at dorsal origin. Caudal peduncle 
well compressed, its least depth 1 to 14 in its length or 2 to 2 Mo in head. 

Head small, strongly compressed, sides flattened and scarcely approximated 
below; profiles similar; head width 2 to 2 inits length. Snout conic, short, obtuse, 
length % its width. Eye advanced, hind pupil edge slightly before center in head 
length; diameter equals snout, 1% to 1}4 in interorbital; very narrow border of adi- 
pose tissue all around eye. Mouth slightly inclined, lower jaw a little the shorter. 
Maxillary largely concealed, reaches opposite hind nostril, and with short terminal 
barbel, not over ¥ of pupil in length. Edges of jaw firmly coriaceous and lips narrow. 
Nostrils together, within last fourth of snout; front one a simple pore, its hind cutane- 
ous flap exposes hind nostril in crescent. Interorbital widely convex. Suborbital 
chain moderate; posterior infraorbital largely covering cheek. 

Gill-opening extends forward to opposite hind edge of eye. Gill-rakers 2+5 
short points, about % the length of gill-filaments, which 14 in eye. Pseudobranchiz 
1% in gill-filaments. Pharyngeal teeth 5—5, compressed, each with slight terminal 
hook and lower face with strong crenulations, these more or less obsolete in smaller 
examples; each also with broad entire inner grinding-surface. 

Scales with 55 to 61 fine waved weak apical marginal striz; circuli fine, all basal. 
Scales closely adherent on trunk, thin, in longitudinal series, deepest medianly and 


386 Bulletin American Museum of Natural History [Vol. L 


rows slightly convergent anteriorly and posteriorly; scales much smaller on breast. 
Basal scaly sheaths to dorsal and anal moderate and caudal base covered with mode- 
erate scales. Pointed axillary ventral scale 2!4 in fin. Lateral line complete, curves 
slightly inferiorly along side to caudal base medianly; tubes slender, simple, well 
exposed or at least half way in scale exposure. Male with front half of snout before 
nostrils with thick-set pits, evidently tubercle scars and not present in female. 

Dorsal origin little nearer snout tip than caudal base; first 3 rays strongly osseous, 
compressed, though third with flexible tip; length of first branched ray 1% to 1% 
in head. Anal origin midway between gill-opening and caudal base, in younger 
specimens, a little advanced or midway between hind eye edge and caudal base; 
first 2 rays similarly osseous to those of dorsal and last soft rays greatly shorter in 
both, than front ones; first branched anal ray 144,to 1% in head. Caudal deeply 
forked, lobes pointed, equal; lower lobe 3% to 3°45 in combined head and trunk. 
Pectoral low, nearly or quite reaching ventral, 144 to 1% in head. Ventral inserted 
little before dorsal origin; depressed fin reaching % to % to anal; fin 1% to 1% in 
head. Vent opposite middle of depressed anal and anal papilla moderate in female, 
not extending beyond end of depressed ventral tips. 

Color in alcohol pale brownish generally, back with slight grayish tinge. Close 
above lateral line about 4 to 6 scales each with an obsolete pale dusky spot, and along 
side of caudal peduncle a narrow slaty streak, also above lateral line and horizontally 
median in position. Fins all pale, dorsal, anal and caudal with 3 or 4 rows of obscure 
grayish spots. Pectoral and ventral uniform. Iris whitish. 

Length 105 to 170 mm. 


Fourteen examples from Ningkwo. 
Acanthorhodeus hypselonotus Bleeker! apparently differs in its much 
deeper body, larger scales, fewer dorsal rays and the absence of barbels. 


Abramidins 


Chanodichthys bramula (Valenciennes) 


Head contained 4 times in body to caudal base; depth 24. Dorsal ITI, 7, 1; 
anal 11, 28,1. Scales 50 in lateral line to caudal base and 3 more on latter; 12 scales 
above lateral line, 9 below; 28 predorsal scales. Snout 34in head; eye 44; maxil- 
lary 373; interorbital 2}4. 

Body strongly compressed, deeply ovate, predorsal slightly trenchant and post- 
ventral with distinct keel over which scales not passing; greatest depth at dorsal 
origin. Caudal peduncle strongly compressed, its length *4 its least depth, which 2 in 
head. 

Head small, profiles alike, that of occiput curving up suddenly and convexly to 
dorsal fin; flattened sides but slightly constricted below; width of head 2 in its length. 
Snout broad, conic, its length % its width. Eye with center slightly behind first third 
of head; its diameter 1'4 in snout, 2 in interorbital. Mouth broad, with short gape, 
and closed jawseven. Maxillary concealed, reaches opposite hind nostril. No barbels. 
Both jaws with strong, trenchant, cartilaginous edges. Lips firm. Nostrils together, 
within last third of snout; hind one twice size of front one and exposed in crescent, 
due to hind cutaneous flap of front one. Interorbital broadly convex. Suborbital 


11871, Verhandel. Kon. Akad. Wet. (Amsterdam), XII, (No. 2), p. 43, Pl. x1, fig. 2. 


1924] Fowler, Some Fishes Collected in China 387 


chain narrow, extends only over upper fourth of cheek. Opercle with finely radiating 
striz. 

Ponce extends forward opposite last ?4in head. Gill-rakers 5+10, slender 
points, 4 the length of gill-filaments, which 1}, in that of eye. Pseudobranchiz half 
length of gill-filaments. Pharyngeal teeth, 2, 4, 4—5, 4, 2, compressed, tips slightly 
hooked, grinding-surfaces slight and entire. 

Scales with 21 to 29 apical striae; basal circuli fine, 67. Scales closely adherent, 
thin, in even longitudinal series all more or less well exposed, but slightly larger about 
body edges and on breast; caudal base with little smaller scales. Ventral with pointed 
axillary scale, 1; the length of fin. Lateral line slightly decurved, becomes median at 
caudal base; tubes simple, extend over half of each scale exposure. 

Dorsal origin a little nearer caudal base than snout tip, first 3 rays osseous, entire, 
and third with flexible tip; a fin reaches a distance 1’ in that to caudal base; 
first branched dorsal ray 1%o in head. Ratal origin opposite base of last dorsal ray, 
3 times length of last; first branched ray 1%9 in head. Caudal deeply forked, lower 
lobe slightly ee or 34 in combined head and trunk. Pectoral low, not quite 
reaching ventral, 144 in head. venine inserted well before dorsal origin, reaches vent, 
which close before anal; ventral 1° in head. 

Color in alcohol with back pale olivaceous, sides and below paler. Each row of 
scales with median white streak. Fins pale, all shghtly grayish submarginally, ver- 
tical ones with slight dusky edges. 

Length 185 mm. 


One from Ningkwo. 


Culter erythropterus Basilewsky 


Head contained 3% to 3% times in body to caudal base; depth 3% to 4 %4 
Dorsal III, 7,1; anal 1m, 20, 1 to 25, 1; pectoral 1, 14 or 15; reriteal 1,8. Scales 62 to 
70 in lateral line to caudal base and 4 or 5 more on latter; 14 to 16 scales above lateral 
line, 8 or 9 below; 44 or 45 predorsal scales. Snout 3% to 4 in head. measured from 
upper jaw tip; eye 44 to 5%; maxillary 3 to 3; interorbital 344 to 444 

Body elongately ovoid, strongly compressed, deepest at dorsal origin, edges all 
convexly rounded, except slight moten poston Beet Caudal peduncle strongly 
compressed, least depih 1% to 144in its length or 24 to 244 in head. 

Head well compressed, flattened sides sloping Shae above and below, upper 
profile sebily concave from snout to occiput; width of head 2% to 2%4 in its length, 
depth 144 to 145. Snout convex, profile slightly convex in front, its length 7 its width. 
Eye moderate, ong edge about midway in total head length; diameter 1 to 14 in 
snout, 1g to 1% in interorbital. Mouth moderate, well inclined, lower jaw well 
pe eae ine. Maxillary largely concealed above by preorbital, reaches eye, expansion 
1% to 1% in eye. Lips firm, fleshy, lower much broader laterally. Nostrils together; 
front one about at last third of snout, a simple pore; hind one a crescentic slit close 
behind, bounded by cutaneous marginal flap of front one. Interorbital broadly 
convex. Suborbital chain moderate; preorbital little deeper than long and its length 
slightly greater than eye. 

Gill-opening extends forward opposite eye center. Gill-rakers 5+16, lanceolate, 
equal gill-filaments in length, or 1% in that of eye. Pseudobranchie half length of 
gill-filaments. Pharyngeal teeth, 2, 4, 5—4, 4, 2, hooked and with slight narrow grind- 
ing-surfaces. 


388 Bulletin American Museum of Natural History [Vol. L 


Seales with 6 to 8 apical radiating strie and 10 or 11 marginal auxiliaries; basal 
circuli 91 to 96. Scales in even longitudinal series parallel with lateral line, slightly 
crowded and smaller on predorsal, belly, breast and caudal base. Pointed scaly flap 
in ventral axil about 4 the length of fin. Lateral line complete, decurved moderately, 
passes ventral at lowest third in body depth and ascends caudal base medianly; 
tubes simple, slender, extend over each scale about midway in its exposure. 

Dorsal origin a little nearer caudal base than snout tip, first 2 rays spine-like, 
osseous, and third greatly longest; depressed fin 24 in distance to caudal base; first 
branched ray 1% to 1% in head. Anal begins slightly before depressed dorsal tip, 
first branched ray highest and forms apex of moderate anterior lobe; 2o to 2% in 
head. Caudal strongly forked, lobes sharply pointed, slender and equal; lower lobe 
equals or little more than head. Pectoral reaches *% to ventral, % in smaller ex- 
amples; length 14 to 1% in head. Ventral inserted slightly before dorsal origin, 
reaches % to % to anal; length 1% to 1% in head. Vent close before anal. 

Color in alcohol, back tinged with pale olive, sides and lower surface pale to 
whitish. Dorsal and caudal pale olivaceous, lower fins whitish. Iris slaty to whitish. 

Length 226 to 266 mm. 


Two examples from Ningkwo. 


Leuciscine 
Leuciscus brandti (Dybowsk1) 


Head contained 3% to 34 times in length to base of caudal; depth 4% to 54. 
Dorsal u1, 7,1; anal 1, 7,1. Scales 80 to 90 in lateral line to caudal base and 1 to 4 
more on latter; 16 to 20 scales above lateral line, 10 to 16 below; 52 to 60 predorsal 
scales. Snout 3% to 3% in head; eye 34 to 5; maxillary 2% to 3%; interorbital 3 to 
3%. Pharyngeal teeth 2, 5—4, 2, latter rarely 5, 2, slender, slightly hooked, with 
grinding-surfaces. 

Color in alcohol with back deep dusky brownish generally, with lower and under 
surfaces pale to whitish. An inconspicuous blackish horizontal line from just above 
origin of lateral line backward, but obsolete behind dorsal and marking dark color of 
back. Along middle of side, from about level with eye, broad band little narrower 
and becoming conspicuously black after dorsal and ventral to middle of caudal. 
Caudal base with conspicuous and prominent black blotch size of pupil, midway at 
caudal base and distinctly separated from dark lateral band. Whole side of body 
sprinkled with variably dusky to black dots, specks and spots. Dorsal and caudal 
pale brownish, also pectoral with slight tinge of same. Ventral and anal whitish. 
On head dark band as obsolete or slightly reflected. 

Length 34 to 163 mm. 


Sixty-five examples from Hsing Lung Shan, August 7, 1921. In 
three very young examples the lateral line is present only on a few 
anterior scales. Two also have the eye slightly greater in diameter than 
the snout length, while one has the breast finely scaled. 

The short diagnosis of Telestes brandtii! Dybowski gives the devel- 
oped anal rays as 8, a condition I find in only one specimen in my series. 


11912, Faune Russie, Pisc., III (1), p. 155, Fig. 6. 


1924] Fowler, Some Fishes Collected in China 389 


Berg shows an example from the lower Amur region. In his diagnosis of 
the species he repeats Dybowski’s formula, though his figure shows cer- 
tainly nine branched anal rays. It also does not show the dark median 
lateral band, the conspicuous detached black basal caudal spot and the 
speckled or dusted pattern of dark dots, so conspicuous in my examples. 


Phoxinus lagowskii Dybowski 


Head contained 3}4 to 3% times in length to base of caudal; depth 4% to 4%. 

- Dorsal 11, 7,1; anal 1m, 7,1. Scales 71 to 73 in lateral line to caudal base and 3 or 4 

more on latter; 16 to 19 scales above lateral line, 11 or 12 below; 46 or 47 predorsal 

scales. Snout 3/6 to 34 in head; eye 5 to 5%; maxillary 3 to 3%; interorbital 3Mo to 

3%. Pharyngeal teeth 2, 5—4, 1, compressed, ends hooked and grinding-surfaces 

‘present. Scales with 380 radiating marginal striz, of which 18 or 19 basal; circuli 
coarse. 

Color in alcohol dusky-brown on back and upper surfaces, finely speckled or 
mottled irregularly with dusky to blackish, though in no place forming a dark 
lateral band. Vertical blackish line, less than eye, at caudal base. Upper inner 
border of gill-opening blackish. Lips pale. Iris dull slaty. Dorsal and caudal 
grayish-brown, other fins paler to whitish. 

Length 106 mm. 

Two. examples from Hsing Lung Shan, August 7, 1921. 

My examples agree largely with the figure and account by Berg! as 
Phoxinus lagowskii variegatus (Ginther). This Berg admitted to sub- 
specific rank as the southern and Chinese representative of Phoxinus 
lagowskit. I find nothing, aside from the alleged deeper caudal peduncle 
of variegatus, to distinguish it. 

A comparison of the four types of Leuciscus costatus Fowler,? 
shows them to be synonymous. A further study of them shows the fol- 
lowing details. 

Head contained 374 to 3% times in length to base of caudal ; depth 4% to 4%. 
Dorsal 11, 7,1; anal 1, 7,1. Scales 71 to 73 in lateral line to caudal base and 2 to 4 
more on latter; 14 to 16 scales above lateral line, 8 or 9 below; 44 to 48 predorsal 
scales. Snout 3 to 3% in head; eye 3% to 44; maxillary 3 to 34; interorbital 2% 
to 2%. Pharyngeal teeth 2, 5—5, 2, grinding-surfaces little developed. In all, asin 
the Hsing Lung Shan specimens, the lateral line is complete. There is also a faint 
trace of the dark or blackish vertical basal caudal bar as in Berg’s figure of 
Phoxinus lagowskii.* 


Idus waleckii Dybowski 


Head contained 3% to 4/40 times in length to base of caudal; depth 3% to 4. 
Dorsal 111, 8,1; anal 111, 10,1. Scales 44 to 47 in lateral line to caudal base and 2 or 3 


11912, Faune Russie, Pisc., III, (1), p. 231, Fig. 13. 

71899, Proc. Acad. Nat. Sci. Phila., 1899, p. 180. Tan Lan Ho, tributary Shu Lan Ho, Sungari 
Basin in Eastern Mongolia. 

31912, Faune Russ., Pisc., IIT, (1) p. 228, Pl.1, fig. 8. 


390 Bulletin American Museum of Natural History [Vol. L 


more on latter; 9 or 10 aersi above lateral line, 5 below; 23? to 26 predorsal scales. 
Snout 3% in head; eye 444 to 5; maxillary 3 to 3°4; interorbital 3 to 344. 

Body elongate, Uae slender. Head width 2 in its length, scarcely constricted 
below. Snout rather short, convex, length *4 its width. Eye at first 34 of head. 
Mouth very oblique. Maxillary greatly inclined, reaches about to eye. Interorbital 
broadly crescentic. Gill-rakers 3+6 short points, about \ of gill-filaments, which 1% 
in eye. Pharyngeal teeth 3, 5—5, 3, hooked, with narrow grinding-surfaces. Scales 
with 4 or 5 apical radiating strise; circuli fine, though doubly so basally. Scales on 
breast and preventral but little smaller than others. Lateral line complete, well 
decurved. Dorsal origin midway between eye center and caudal base. Anal origin 
well behind base of last dorsal ray. Caudal emarginate. Pectoral contained 1% 
timesin head. Ventral inserted a trifle before dorsal origin; fin 1°4in head. 

Color in alcohol faded dull dusky-brown on back, with olive tinge. Sides and 
lower surface dull brassy. Under surface more or less uniform. Iris brownish, also 
fins. 

Length 168 to 215 mm. 


Described from the type of Leuciscus farnumi Fowler, in the Acad. 
Nat. Sci. Philadelphia from the Tore River, and two paratypes from 
Dalai Nor, all obtained in 1897.! 


Opsariichthys uncirostris (Schlegel) 


Head contained 3% to 3} times in length to caudal base; depth 4 to 4%. Dorsal 
ut, 7,1; anal i, 9,1. Scales 41 to 45 in lateral line to caudal base and 3 more on 
latter; 9 scales above lateral line, 5 below; 20 or 21 predorsal scales. Snout 3)3 
to 3% in head; eye 3 to 5; maxillary 2 to 244; interorbital 3 to 3%. Pharyngeal teeth 
with broad grinding-surfaces and small terminal hooks. Scales with 5 to 12 apical 
radiating strize; circuli moderate, fine apically. Lateral line only on about 7 anterior 
seales In young. 

Color in alcohol silvery white. Dorsal and caudal grayish, lower fins whitish. 
Jaws and lower side of head pale, like belly. 

Length 32 to 134 mm. 

Fifteen examples, from twenty-six miles aioe of the Hsing Lung 
Shan, August 12 , 1921. 

For comparison there are in the Academy two from the Tore River 
in Eastern Mongolia, and a series of Japanese specimens from the Yodo 
River in Osaka. They show: 

Head contained 3}4 to 3}4 times in length to caudal base; depth 4 to 5. Dorsal 
1, 7,1; anal, 9,1. Scales 40 to 48 in lateral line to caudal base and 2 or 3 more on 
latter; 8 to 10 scales above lateral line, 4 or 5 below; snout 3% to 4in head; eye 3% 
to 4; maxillary 2 to 2%; interorbital 3% to 4%; pharyngeal teeth 2, 3, 5—5, 3, 2, 
vary 2, 3, 5—4, 3, 2 or 1, 4, 4—5, 4, 1 or 1, 4, 44, 4, 1. 

Length 50 to 135 mm. 


11899, Proc. Acad. Nat. Sci., Phila., 1899, p. 179. 


1924] Fowler, Some Fishes Collected in China 391 


Zacco platypus (Schlegel) 


Head contained 34 to 3% times in length to caudal base; depth 4 to 5. Dorsal 
m1, 7,1; anal 1, 9,1. Scales 41 to 43 in lateral line to caudal base and 3 more on 
latter; 7 or 8 scales above lateral line, 3 or 4 below; 16 to 18 predorsal scales. Snout 
3 to 3% in head; eye 3 to 4; maxillary 24% to 3; interorbital 24 to 3. Pharyngeal 
teeth 2, 4, 4—4, 4, 2, hooked, with grinding-surfaces. Scales with 11 to 20 apical 
radiating striz; circuli largely coarse apically, become fine basally. Pearl organs of 
male, a row of small ones from nostrils to antero-supraorbital; row around front edge 
of snout, slightly doubled on each side; then a continuous series of 4 horizontally on 
preorbital; on suborbital chain, a row close below eye to postorbital inferiorly; a 
series of 7 on lower ridge of preopercle; scattered tubercles on upper part of opercle 
and along its front edge vertically; a double row along lower side of each mandibular; 
second, third and fourth anal rays each with row of small tubercles along outer face. 

Color in alcohol largely dull olivaceous above, edge of each scale slightly darker. 
Muzzle, upper surface of head and chin largely with dusky-slate to blackish, pale in 
female or non-ornamented male. Male with dusky-slate broad lateral band, inter- 
rupted by a number of pale lines or bars extending up and intersecting with whitish 
color of belly. Membranes of dorsal and anal dusky to blackish. Caudal largely with 
dusky tinge, especially on outer median portion. Other examples, tuberculate, and 
with but slight dusky on fins, show pale or orange tinge on middle of front dorsal edge 
and upper part of pectoral medianly. Same on ventral fin and front half of anal. 

Length 34 to 130 mm. 


From Hsing Lung Shan, August 7, 1921, 114 examples; twenty-six 
miles to south of Hsing Lung Shan, August 12, 1921, 105 examples. 

For comparison I examined the following series of Japanese ex- 
amples in the Academy: Kinu River at Utsonomiya, Yodo River in 
Osaka, Chilongo River in Kurume, Yabe River at Funayado, Kawatana 
near Nagasaki and Tsuruga. They show: 

Head contained 3% to 4 times in length to caudal base; depth 34 to 44%. Dorsal 
m1, 7,1; anal m1, 8,1 t0 10,1. Scales 36 to 46 in lateral line to caudal base, and 
2 or more on latter; 8 scales above lateral line, rarely 7, 4 below; 15 to 17 predorsal 
scales. Snout 3 to 34 in head; eye 2%; to 44; maxillary 2% to 2%: interorbital 2%; 
to 3%. Pharyngeal teeth 2, 4, 5—5, 4, 2, vary 1, 4, 5—5, 4, 1 or 2, 4,45, 4, 2 or 2, 
4, 5—4, 4, 2or 1, 4, 5—4, 4, 1 or 2, 4, 44, 3, 1. 

Length 57 to 144 mm. 

Zacco platypus has not before been reported from Northeast Mon- 
golia. In the American Museum series another more progressively 
ornamented male shows the series of lateral snout pearl-organs and those 
on the lower preopercle ridge fused as a ridge. The latter are greatly 
similar to Boulenger’s figure of Opsariichthys acanthogenys. Further, 
there are small scattered pearl-organs on the scales along the anal base 
and under surface of the caudal peduncle. The lower rudimentary caudal 
rays are much covered with adipose tissue, abruptly ending forward and 
like basal portions of lower caudal rays also finely studded with tubercles. 


392 Bulletin American Museum of Natural History [Vol. L 


Tubercles on the anal fin are present on the anterior or rudimentary 
rays, basal tubercle of fourth, fifth and sixth anal rays enlarged and at 
tips of fifth, sixth and seventh two or three small terminal tubercles on 
each side. 


Pseudaspius leptocephalus (Pallas) 


Head contained 3% times in length to caudal base; depth 434. Dorsal m1, 7, 1; 
anal 111, 9,1. Scales 91 in lateral line to caudal base and 6 more on latter; 16 scales 
above lateral line, 10 below; 52 predorsal scales. Snout 3% in head measured from 
upper jaw tip; eye 64; maxillary 3%; interorbital 4%. 

Body elongate, well compressed, edges rounded, deepest at dorsal origin. Caudal 
peduncle compressed, least depth 3 in total head length. Head attenuate, compressed, 
flattened sides not convergent; width 2°; in its length. Snout broadly convex, de- 
pressed, a trifle longer than wide. Hye slightly before first third in head length. 
Premaxillaries protractile forward. Maxillary not quite reaching to eye or but a 
trifle behind posterior nostril. Mandible depressed; protrudes, 2% in total head 
length. Nostrils at last third of snout, superior. Interorbital broadly convex. Pre- 
orbital 1% in snout, width 1% its length. Postorbital equals eye. Gill-rakers 3+6, 
short strong firm points, 2*4 in length of gill-filaments, latter 144 in eye. Pharyngeal 
teeth 2, 5—?, ?, with slight grinding-surfaces. Scales small, reduced and crowded on 
median preventral and breast, little smaller on median predorsal and caudal base. 
Pointed adnate scaly flap in ventral axil. Lateral line well decurved. 

Dorsal origin midway between hind eye edge and caudal base, fin reaches half 
way to caudal base, first branched dorsal ray longest, 174in head. Anal a little behind © 
dorsal base, first branched ray 2 in head. Caudal deeply forked, pointed lobes evi- 
dently subequal, upper lobe 14in head. Pectoral3in head. Ventral inserted a little 
before dorsal, nearer anal origin than pectoral origin, 2 in head. 

Color in alcohol, dull uniform brownish, paler below. Fins and iris plain brown. 

Length 247 mm. 


Described above from an example in the Academy obtained in the 
Tore River, a tributary of the Sungari. 


Chela nicholsi Fowler 
Figure 2 

Head contained 4% times in length to caudal base; depth 4%. Dorsal 1m, 7, 1; 
anal 111, 22,1; pectoral1, 14; ventral1, 8. Scales 58 in lateral line to caudal base and 
4 more on latter; 10 scales above lateral line, 3 below; 40 predorsal scales. Snout 
3% in head measured from upper jaw tip; eye 3°4; maxillary 3; interorbital 4%. - 

Length 152 mm. 

Type in the American Museum and paratype noted above in the 
Acad. Nat. Sci. Philadelphia. These are the only examples known. 


Pseudobrama dumerili Bleeker 


Head contained 4% times in length to caudal base; depth 3%. Dorsal III, 7, 1; 
anal nt, 9,1; pectoral1, 14; ventral, 8. Scales 44 in lateral line to caudal base and 
3 more on latter; 9 scales above lateral line, 4 below to ventral origin, 6 below to anal 
origin; 20 predorsal scales. Snout 4in head; eye 34; maxillary 54; interorbital 2%. 


1924] Fowler, Some Fishes Collected in China 393 


Body moderately ovoid, strongly compressed, deepest at dorsal origin, post- 
ventral with slight median keel over which scales not passing and other edges all 
convexly rounded. Caudal peduncle strongly compressed, least depth 1% in its 
length or 1% in head. 

Head well compressed, flattened sides slightly approximated below, upper profile 
slightly convex, lower more so; width 2 inits length, depth 14. Snout convex, rather 
obtuse, length %4 its width. Eye rather large, hind edge midway in head length, 
diameter 1 in snout, 1% in interorbital. Mouth small, inferiorly terminal, width 134 
in eye. Maxillary largely concealed by preorbital, at least its expansion, which 3% 
in eye, or reaches opposite front nostril. Lips moderate, entire. Edges of jaws 
firmly coriaceous. Nostrils together, front one at last third in snout, pore-like, and 
hind one a crescentic slit close behind, formed by cutaneous marginal flap of anterior. 
Interorbital broadly and evenly convex. Suborbital chain moderate, preorbital 
depth 44 its length, which 1in eye. Opercle smooth. 


Fig. 2. Chela nicholsi Fowler. Type. 


Gill-opening extends forward about opposite hind edge of eye. Gill-rakers 38+ 
91, short, fine, slender, close-set, 2/4 in length of gill-filaments, which 144 in eye. No 
pseudobranchie. Pharyngeal teeth 6—7, strongly compressed, all with broad grind- 
ing-surfaces, and pointed but without hooks. 

Scales with 16 to 19 apical radiating striz, also 7 or 8 auxiliary marginals; basal 
circuli 45 to 60. Scales in even longitudinal series parallel with lateral line, scarcely 
smaller on breast and caudal base. Pointed scaly flap in ventral axil, about % length 
of fin. Lateral line complete, deeply decurved, passes lower % in body length at 
ventrals and ascends along side of caudal peduncle to caudal base medianly. Tubes 
simple, extend over each scale about midway of exposure. 

Dorsal origin midway between snout tip and caudal base, first 3 rays spine-like, 
osseous, compressed, third greatly longest; first branched dorsal ray 1 in head. Anal 
begins well behind depressed dorsal, first branched ray highest and forms apex of 
slight anterior lobe, 2in head. Caudal strongly forked, lobes sharply pointed, slender, 
equal, little longer than head. Pectoral reaches a distance contained 1% in that to 
ventral, 14 in head. Ventral inserted slightly before dorsal origin, reaches a distance 
contained 1% in that to anal, 1% in head. Vent close before anal. 

Color in alcohol with back tinged pale olive, sides and lower surface pale to whit- 
ish. Dorsal and caudal pale olivaceous, lower fins whitish. Iris slaty and whitish. 

Length 170 mm. 


One from Ningkwo. 


394 Bulletin American Museum of Natural History [Vol. L 


Ochetobius elongatus (Kner) 


Head contained 4% to 4% times in length to caudal base; depth 54 to 6%. Dorsal 
1, 9, 1; anal mt, 9,1. Scales 63 to 65 in lateral line to caudal base and 3 more on 
latter; 10 or 11 scales above lateral line, 5 or 6 below; 28 to 30 predorsal scales. 
Snout 3% in head; eye 54 to 6; maxillary 3% to 334; interorbital 34% to 3%. 

Body greatly elongate, fusiform, compressed, deepest at dorsal origin, edges all 
convexly rounded. Caudal peduncle well compressed, its least depth 2 to 2% in its 
length or 2% to 3Mo in head. 

Head small, compressed, its flattened sides slightly approximated below, upper 
profile nearly straight and much less inclined than lower; width of head 2% to 2% 
in its length. Snout conic, its length / to 1 in its width. Eye center at first third in 
head; eye diameter 1% to 134 in snout, 144 in interorbital; adipose-lid moderate, in- 
vades a little more of eye behind than in front. Mouth oblique, closed jaws even in 
front. Maxillary reaches opposite hind nostril, largely concealed. Jaws firm and 
their edges rather trenchant. Nostrils together, within last fourth of snout; front 
one a pore, with hind cutaneous flap exposing posterior nostril in crescent. Inter- 


orbital broadly convex. Suborbitals narrow, cover but little of cheek, less than 44 


to preopercle ridge. 

Gill-opening extends forward opposite hind eye edge. Gill-rakers 8+25, lanceo- 
late, slender, contained 1% in the length of gill-filaments, or 174 in eye. Pseudo- 
branchize about half the length of gill-filaments. Pharyngeal teeth 2, 3, 5—5, 3, 2, 
with broad smooth grinding-surfaces but without terminal hooks. 

Scales with 23 to 30 apical slightly waved radiating strize; basal circuli fine. 
Seales largely adherent, thin, in even longitudinal series, largely uniform; scales a 
little smaller on breast and but slightly smaller on caudal base than on body. Ventral 
axillary scale 2% in the length of the fin. Lateral line complete, decurved slightly 
along side, and becomes median at caudal base. Tubes simple and extend about half 
way over scale exposure. 

Dorsal origin slightly nearer snout tip than caudal base, first branched ray de- 
pressed reaches a little beyond tip of others, or is contained 1% to 14 in head; length 
of depressed fin contained 3), to 3)4 times in length to caudal base. Anal origin a 
little nearer last dorsal ray base than caudal base, first branched ray tip shorter than 
last when depressed, or 2 to 2% in head; depressed fin 144 to 1% times in distance to 
caudal base. Caudal deeply forked, lobes sharply pointed, alike, 4% to 4°4in combined 
head and trunk. Pectoral reaches a distance contained 1%9 to 2% in that to ventral, 
144 to 1°4in head. Ventral inserted slightly before dorsal origin, 24 to 2)4 in distance 
to anal origin, 175 to 14 in head. 

Color in alcohol with back olivaceous, sides and below pale brownish, with traces 
of whitish. Eye and under surface of head silvery-white. Dorsal and caudal pale 
brownish, paired fins and anal whitish. 

Length 235 to 270 mm. 


Five from Ningkwo. 


COBITIDZ 


Misgurnus anguillicaudatus (Cantor) 


Head contained 5 to 6/4 times in length to caudal base; depth 7% to 9. Dorsal 
m1, 7,1; anal um, 6,1. Scales 138 to 170 in lateral line to caudal base. Snout 25 to 


1924] Fowler, Some Fishes Collected in China 395 


2% in head; eye 54% to 74; maxillary 3% to 4; interorbital 47 to 544. Scales with 36 
to 38 radiating striz in young, adult with 68 to 70; circuli fine. 

Color in aleohol dull brownish, back darker and color of upper surface rather 
abruptly distinct from that of lower surface. Back and upper surface more or less 
distinctly spotted with darker everywhere, though spots more pronounced pos- 
teriorly, or above anal and on caudal peduncle. Spots on back and sides often 
arranged as 2 longitudinal dark lines. Dorsal pale brown, with about 4 horizontal 
bands of dusky spots. Caudal with about 10 transverse vertical dusky spots and 
black blotch at base above nearly large as eye. Anal pale, also with few dusky spots. 

Length 66 to 206 mm. 

Eighty examples from Hsing Lung Shan and two from twenty-six 
miles south of Hsing Lung Shan, latter August 12, 1921. Besides these 
I have also included three from Tan lan Ho, in the Academy. 

The color is variable, though always with a black spot at the bases 
of the upper caudal rays. Head and trunk always with scattered spots, 
specks or dots, variously close-set or more or less scattered. Sometimes 
only large scattered blotches to several times size of eye present, or 
larger blotches may occur only above lateral line. None show so great 
development of adipose-like rudimentary caudal rays as in the next 


species. 


Misgurnus decemcirrosus (Basilewsky) 
Figure 3 
Head contained 5 to 5% times in length to base of caudal; depth 6 to 6%. 
Dorsal 11, 7, 1, sometimes 8, 1; anal 11, 6, 1, sometimes 5, 1. Scales 95 to 127 in lateral 
line to caudal base. Snout 2! to 2°4in head; eye 5 to 74; maxillary 4 to 444; inter- 
orbital 3°4 to 674. Scales with 36 to 44 radiating striz in young, with circuli about 70 
rows; adults with 77 to 80 radiating striz. 


Fig. 3. Misgurnus decemcirrosus (Basilewsky). Tien Tsen. 


Color in alcohol dull brown, finely specked or dotted with dusky on back and 
upper surface. Obscure dusky or blackish spot at bases of upper caudal rays not 
much larger than pupil. 

Length 132 to 196 mm. 


Six from Ningkwo. Also twenty-one from the Pietto River at Tien 
Tsen collected by N. F. Drake, in the Academy. 


396 Bulletin American Museum of Natural History [Vol. L 


The Ningkwo examples are all with very great development of the 
rudimentary caudal rays, starting close behind the dorsal base in most 
cases and joining behind with the caudal. Lower rudimentary caudal 
rays likewise greatly developed, begin close behind anal base and fused 
posteriorly with caudal. 

Color pattern also very variable, as some with large scattered dusky 
to blackish spots on head and trunk, though none much larger than eye. 
Body similarly finely sprinkled with dark dots, specks or small spots as 
in Misgurnus anguillicaudatus. 

Misgurnus crossochilus Sauvage is doubtless a synonym. The de- 
tails set forth in the original description do not show any characters 
worthy of specific value and agree so far as given with my materials of 
the present species. 


Nemacheilus toni (Dybowski) 

Head contained 4 to 4% times in length to caudal base; depth 6% to 84. Dorsal 
u, 7,1 or 8,1; analiz, 5. Scales quite minute, about half a millimeter in size and 128 
to 140 estimated in lateral count; each with 30 radiating marginal strize; circuli 
moderately fine. Snout 2% to 2% in head; eye 5% to 6%; maxillary 3% to 4M; 
interorbital 444 to 5%. 

Color in alcohol pale olive-gray on back to paler or whitish below, side and back 
with many large dull dusky blotches, rather irregular and most much larger than eye. 
Obscure dusky streak from side of snout to nostrils. Dorsal with about 4 and caudal 
with 5 or 6 cross bands made up of dusky blotches, fins otherwise pale to whitish. 
Several grayish blotches on pectoral medianly and anal terminally. In young 
example color-pattern much more contrasted. The markings appear variable in 
preserved examples as in the one in best condition they are as a row of large blackish 
blotches above and another below lateral line. 

Length 60 to 99 mm. 

Three from Hsing Lung Shan. 

The type of Nemachilus pechiliensis Fowler, in the Academy, is now 
in such poor preservation that it is useless for examination. The original 
description, however, shows that it is clearly a synonym. 

Orthrias oreas Jordan and Fowler’ is also in agreement and is another 
Synonym. 


Lefua andrewsi Fowler 


Figure 4 
Known only from the type in the American Museum. 


11903, Proc. U. 8. Nat. Mus., X XVI, p. 796, Fig. 2. 


1924] Fowler, Some Fishes Collected in China 397 


Fig. 4. Lefua andrewsi Fowler. Type. 


Lefua costata (Kessler) 

Head contained 4 to 4% times in length to caudal base; depth 6% to 6%. Dorsal 
1, 6, 1; anal u, 5,1. Scales 93 to 102 in median lateral series. Snout 21% to 3% in 
head; eye 4% to 64; maxillary 3% to 444; interorbital 3 to 3/5. Scales with 38 to 43 
radiating marginal strie (30 to 36 in type of Nemachilus dixoni); circuli rather fine, 
moderate. 

Color in alcohol brownish generally, with slight olive tint. Back and sides with 
very obscure scattered moderate spots of small size. Dorsal and caudal finely and 
obscurely spotted with gray-brown, though latter fin always with distinct round black 
spot at middle of base, not larger than pupil. Iris dull slaty, lips and jaws all pale or 
dull. Dark median lateral streak not very distinct and largely evident only after 
ventral. 

Length 40 to 73 mm. 

Six from Hsing Lung Shan. 

In the Academy the type of Nemachilus dixoni, a synonym of the 
present species, is included above. 

Elvis nikkonis Jordan and Fowler’ is certainly closely related, though 
the alleged larger scales, given as about fity-six, would seem to warrant 
specific distinction. It is, however, otherwise closely related, as the 
black basal caudal spot suggests. 

Lefua echigonia Jordan and Richardson? is based on three young 
examples only 38 to 45 mm. long. It is represented with four barbels in 
profile, showing it doubtless had eight. Its coloration greatly suggests 
Misgurnus decemcirrosus (Basilewsky). 

Nemachilus variegatus (Dabry) Sauvage and de Thiersant® is 
probably a Lefua. It is described as having large brownish cloudings 
formed into a series of somewhat undulate bands and a black band at 
caudal and dorsal base. 


11903, Proc. U.S. Nat. Mus., XXVI, p. 768, Fig. 1. 
21908, Proc. U. 8. Nat. Mus., XX XIII, p. 263, Fig. 1. 
31874, Ann. Sci. Nat. Zool., I (2), p. 14. 


4.4, etree cel rs Ets i) 


B@Sasces 


< 
7% . 
erate oe £56 


=e > Oe aap Kee PE Se : 
= ‘. ~ ae 2: hom EAS ats 
» & to “Ee Se Sey CD Sap 
oo on Vi oed- PR at yn y tote Ea ay WO ? ee oY, 
: Lape aay es * 6 
> We Se ae 


Fig. 5. Cobitis tenia Linneus. Variation in color-pattern. 


398 


1924] Fowler, Some Fishes Collected in China 399 


Cobitis tenia Linnzeus 
Figure 5 

Head contained 4°4 to 5% times in length to onde base; depih 644 to 734. Dorsal 
i, 6, 1 to 8, 1; anal im, 4,1 to 6,1. Snout 2% to 2%4in head; eye 5! to 64; maxillary 
344 to 444; interorbital 8 to 87%. 

Length 55 to 98 mm. 

From Hsing Lung Shan, August 7, 1921, 83 examples. 

In the Academy a series of 20 examples from Europe, the Italian. 
Lakes and Sweden. These show the following. 

Head contained 4%4 to 5! times in puke to caudal base; depth 5% to 7%. ses 
1, 7,1; anal mi, 5,1 or 6,1. Snout 2% to 2% in head; eye 4%4 to 6; maxillary 3% 
to 44; interorbital 7%4 to 9. 

Length 48 to 95 mm. 

The scales are most striate in the largest European examples, the 
radiating marginal strie showing 38 to 48 and circuli in the largest in: 
about 25 series. In a Japanese example, representative of the nominal 
Cobitis biwe Jordan and Snyder, from Kiroshina, and in all the above 
Chinese examples examined, besides one I reported from the Academy 
collection as the nominal Cobitis sinensis Sauvage and De Thiersant 
taken in the Tan Lan Ho, the marginal radiating striz are 28 to 31. In 
no way do they differ from the other Chinese examples examined. 

Considerable variation in the color-pattern is noticed in this species, 
and for comparison I have placed a drawing of the largest European 
example at the top of the accompanying figure, the others representing 
the extremes found in the Hsing Lung Shan series. 

From the description of Cobitis dolichorhynchus Nichols! I am in- 
clined to consider it a variant. It is said to be more elongate than 
Cobitis tenia, though its depth is given as only 5.8. From the above 
details this depth is within the range of my European examples. The 
alleged longer snout also appears variable. I fail to find any other char- 
acters on which to separate the Futsing specimens. 

Four examples also obtained at Ningkwo, 137 to 165 mm. 


SILURIDEA 
Parasilurus asotus (Pallas) 
Head contained 3% to 4°4 times in length to caudal base; depth 545 to 645. Dorsal 
I, 5; anal 1 to v, 68 to Les Snout 2% to 3% in Beas measured from tip of upper jaw; 
eye 416 to 9; maxillary 2!4 to 3%; interorbital 2!4 to 24%. Lower jaw well protruded. 
In young raetiilaae y barbel extends back to middle of deg essed dorsal, or well beyond 
depressed pectoral. 


11916, Proc. Biol. Soc. Wash., XX XI, p.16. Futsing, Fu-kien. 


400 Bulletin American Museum of Natural History [Vol. L 


Color in alcohol nearly deep mouse-gray above, becoming deep olive-buff on 
under surface of head and belly, sides below all with pale smutty dusted appearance. 
About 17 or 18 vertical rows of very inconspicuous pale small spots on back, trans- 
versely down till level with lateral line. Also few similar pale spots on upper surface 
of head. Pectoral and ventral pale. Iris slaty. Maxillary barbel dusky above, pale 
below. Mental barbels pale, like chin. 

Length 58 to 332 mm. 


Eleven from Hsing Lung Shan, August 7, 1921 and two from Ningkwo. 


PORCIDE 
Pseudobagrus macropterus (Bleeker) 


Head contained 4% times in length to caudal base; depth 8%. Dorsal I, 7; 
anal v, 10; pectoral I, 9; ventral1, 5. Snout 234 i in head; eye 7; maxillary Bi inter- 
orbital 32%. ; mouth width 2. 

Body slender, depressed forward, strongly compressed behind, deepest about 
middle of depressed pectoral. Caudal peduncle little free, strongly compressed, least 
depth about 2% in its length or 2% in head. 

Head broadly depressed, flattened medianly above, sides convex, width 1% 
in its length, depth 2%. Snout broadly depressed, length 34 its width as measured 
across at front of eyes. Eye center falls about first % in length of head, elevated, lids 
free; 2% in snout, 2 in interorbital. Mouth broadly transverse, lower jaw much the 
shorter. Teeth in broad villiform bands in jaws and continuous across vomer and 
palatines. Lips thick, fleshy, plicated. Maxillary barbel long, reaches dorsal origin; 
hind nasal barbel longer than eye, 14 in snout; outer mental barbel reaches gill- 
opening or %4 to pectoral origin; inner mental 3 in head. Nostrils separated; front 
one with low cutaneous rim, closer to hind one than to snout edge; hind nostril about 
*/ in snout profile. Interorbital level. 

Gill-opening extends forward opposite eye center. Gill-rakers 8+-13, lanceolate, 
1% in gill-filaments, which 1 in eye. Isthmus wide, depressed, broadly triangular. 

Skin smooth. Lateral line distinct, midway along side. 

Dorsal origin about first third in combined head and trunk, third ray longest 
and edges of membranes slightly emarginate. Dorsal spine 2in head. Adipose dorsal 
very long 2% in combined head and trunk. Anal small, first branched ray highest, 
inserted a little nearer caudal base than pectoral origin; first branched ray 2% in 
head. Caudal deeply forked, lobes rounded, hind edge deeply emarginate and lower 
lobe 1% in upper; length 14% in head. Pectoral rounded, reaches a distance contained 
1% in that to ventral; spine 1% in head, flattened, 14 teeth along hind edge. Ventral 
insertion nearer snout tip than caudal base or close behind base of last dorsal ray, fin 
reaches a distance contained 1% in that to anal, length 1% in head. 

Color in alcohol pale slaty-gray above, with obscure slightly darker slaty dots 
scattered about, also on dorsals and caudal. Under surface of body pale to whitish. 
Maxillary and nasal barbels slaty, others whitish. Iris slaty. 

Length 345 mm. 


One from Ningkwo. 


1924.] Fowler, Some Fishes Collected in China 401 


Pseudobagrus fulvi-draco (Richardson) 


Head contained 2% to 3% times in length to caudal base; depth 3 to 3%. Dorsal 
I, 7,1; anal, v, 150r16. Snout3in head; eye4to7; maxillary 2°4 to 3; interorbital 
ov to ‘M, Maxillary barbel extends back but little hevand head, imal sometimes 
quite a little shorter. Serrze on hind edge of pectoral spine 14 to 16 and much more 
developed than the fine and more numerous even small denticles along front edge of 
spine. Dorsal spine often with very obsolete similar armature, or the denticles or 
spinules absent. 

Color in alcohol slaty-brown above, with about 3 large obsolete blotches on side, 
well contrasted from same color of back above. In many color faded more or less 
brownish. All fins with more or less blackish blotch terminally, though on caudal in 
middle of each lobe. In young color-pattern greatly contrasted as blackish and very 
pale brownish. 

Length 27 to 170 mm. 


Ten from Hsing Lung Shan, August 7. 

Pseudobagrus emarginatus Sowerby! is based on an example 413 mm. 
long without caudal. Sowerby says it is very much more elongate than 
Pseudobagrus ussuriensis (Dybowski), and closely resembles it except 
in its emarginate caudal. Pseudobagrus ussuriensis is said in the original 
description to reach 1000 mm., and to have a rounded caudal. Possibly 
this may be a condition with advanced age. 


OPHICEPHALIDA 
Ophicephalus argus Cantor 


Head contained 2% to 3 times in length to caudal base; depth 5 to 5%. Dorsal 
48 to 51; anal 32 or 33. Scales 59 to 62 in lateral line to caudal base and 3 or 4 more 
on latter; 8 to 10 scales above lateral line, 13 to 15 below; 29 to 30 predorsal scales. 
Snout Bik to 6 in head measured from upper jaw tip; eye 5% to 7%; maxillary 26 
to 2%; interorbital 4% to 54. 

Body moderately long, compressed. Least depth of caudal peduncle 3% to 3% 
in total head length. 

Head width 2% to 2% initslength. Hind pupil edge at first fourth in head; hind 
eye edge about first third in head in young; diameter 1 to 14 in snout, 1% to 1% 
in interorbital. Mouth large, lower jaw protruding. Maxillary extends well beyond 
eye, but slightly beyond in young. Teeth finely conic, in narrow bands in jaws; 
inner row along each side of meee and row on vomer and palatines enlarged. 
Front nostril in short tube at first 74 in snout; hind one a simple pore close above and 
before front eye edge. Tatceorbital depressed, level. 

Gill-rakers 3+6, short low tubercles, largest slightly less than gill-filaments, 
which 2 in eye. 

Scales with 9 to 18 basal marginal radiating strie, 12 to 15 apical circuli, other 
circuli fine. Head with muzzle and jaws naked; 12 to 15 scales on cheek to preopercle 
edge; occipital scales scarcely larger than those on sides of head; small scales on 


11921, Proc. U. 8S. Nat. Mus., LX, p. 1, Yalu River, Southern Manchuria. 


402 Bulletin American Museum of Natural History [Vol. L 


breast and caudal basally. Lateral line a little high at first, drops 2 scales a little 
behind pectoral, then mee to caudal base; tubes narrow and simple. 

Median dorsal rays 3%5 to 373 in head; es anal rays 3% to 3°4; caudal round- 
ed, 134 to 134; pectoral 2 to 3% (?); ventral 34 to 3. 

Color in alcohol brown, paler below to whitish on under surface of head. Pree 
streak from side of snout to eye, then back along upper side of head to shoulder; 
another parallel from lower eye edge to pectoral base. Jaws and lower side of head 
with some pale brown spots. Pores along preopercle flange and mandible dusky. 
Trunk with row of 11 dark large rings, beginning at shoulder and continued above 
median axis to caudal base. Below median axis nearly equal number of less regular 
dark blotches, some angular or giving off slight oblique bars below, but all similarly 
with more or less dark bordering line. Vertical fins dusky terminally, with dark 
blotches. Pectoral pale brown, with dusky spot less than eye at bases of upper rays. 
Ventral whitish. Iris brown. 

Length 98 to 222 mm. 


Seven from Ningkwo. 


Channa ocellata Peters 


Head contained 3% to 3% times in length to caudal base; depth 4% to 54g. Dorsal 
43 to 48; anal 27 to 32. Scales 55 to 60 in lateral line to caudal bape, if on 8 scales 
above lateral line, 14 or 15 below; 20 to 28 predorsal scales. Snout 3% to 4%4 in hee 
measured from upper jaw tip; eye 54 to 7; maxillary 2% to 24: interorbital 334 
to 3%4. 

Body elongate, compressed. Least depth of caudal peduncle 2°4 to 3 in total 
head Jength. 

Head width 1°4 to 144 in its length. Snout depressed, its length 7% to 14 its width. 
Eye ae at first fourth in head, a little backward in young; diameter 1 to 144 in 
snout, 14 to 244 in interorbital. Mouth large, lower jaw slightly protruding. Mouth 
extends a little beyond eye, to hind eye edge in young. Teeth fine, simple, conic, in 
bands in jaws and on vomer and palatine; inner series of teeth in lower jaw enlarged, 
widely spaced. Lips fleshy, rather narrow. Front nostril in short tube on side of 
snout, 145 in eye; hind nostril elevated, opposite upper front eye edge. Interorbital 
broadly convex. 

Gill-rakers as 5 tubercles, lower longer, all much shorter than gill-filaments, 
which 173 in eye. 

Seales with 19 to 30 basal parallel striee; apical circuli 10, giving place with age 
to about 35 irregular marginal strie. Front half of snout, mandible and branchio- 
stegal region naked, and scales on top of head largest, much larger than those on 
cheek; 9 rows of scales across cheek to preopercle edge. Most of pectoral and caudal 
finely scaled. Lateral line high anteriorly, after seventeenth scale dropping 2 scales, 
when midway to caudal base; eles simple, small and with small pore. 

Median dorsal rays 2 to 244 in total head lage median anal rays 2% to 344; 
se es behind, 144 to 14; pectoral 1%5 to 134, reaches a distance contained 
144 to 144 in that to vent. 

Color in aleohol brown on back, paler to livid whitish on under surface of head 
and trunk. Dark streak from upper hind eye edge back toward middle of opercle, 
lower one from lower hind eye edge back toward pectoral base. Lower jaw neutral 
tint, with dusky line on and along upper maxillary edge and another on side of 


1924] Fowler, Some Fishes Collected in China 403 
» 


mandible close along edge of lower lip. In young, lower side of head more or less 
distinctly spotted with brown. Trunk with 10 dusky to blackish large median 
blotches. First as large black round spot above pectoral axil and last as larger round 
black spot at caudal base, which mostly above lateral line and its edge narrowly pale. 
With age most of entire body finely spotted with pale or whitish, sometimes 2 spots 
on 1 scale. Dorsal and anal neutral-dusky, former usually with scattered small 
whitish spots. Pectoral and caudal brownish. 
Length 93 to 255 mm. 


Thirteen from Ningkwo. 


SERRANIDA 


Siniperca chuatsi (Basilewsky) 


Head contained 2% to 2% times in length to caudal base; depth 2% to 373. Dorsal 
XII, 12 to 14; anal III, 8to 10. Scales 128 to 130 counted along lateral line to caudal 
base; pores 115 in lateral line to caudal base; 20 scales above highest arch of lateral 
line to spinous dorsal base, 38 below to spinous anal origin; 33 ? predorsal scales. 
Snout 3}, to 4 in head measured from upper jaw tip; eye 5 to 54; maxillary 2 to 24; 
interorbital 574 to 7. 

Body elongately fusiform, well compressed. Least depth of caudal peduncle 14 
to 134 in its length to 3% to 4% in total head length. 

Head width 2) to 3% in its length, much more attenuated in young. Snout conic, 
width 1 to 1% in its length. Hind edge of eye midway in length of head, diameter 14 
to 174 in snout, *4 to 1in interorbital. Mouth moderate, mandible protruding, greatly 
so in young. Maxillary extends slightly beyond hind eye edge, opposite hind pupil 
edge in young; expansion 1% in eye. Teeth moderately fine, conic, 4 or 5 series in 
front above narrowing in 1 series on side posteriorly, 2 or 3 innermost each side of 
median line largest and directed inward; mandibulars 5 or 6 rows in front, narrowing 
to single outer low series with large inner row of 4 on each side; patch of small conic 
teeth on vomer and each palatine, none on tongue. Nostrils together, within last 

fourth of snout. Interorbital broadly depressed, level in young. Hind preopercle 
edge denticulate and 3 large denticles at angle, much larger in young. 

Gill-rakers 1+3, lanceolate, 14 in gill-filaments, which 1% in eye. 

Scales with 9 basal parallel strie marginally; circuli 7 to 14 apically. 

Fifth dorsal spine 2% to 4 in total length of head; eighth dorsal ray 2% to 3; second 
anal spine 3% to 3/4; second anal ray 2% to 3; caudal rounded behind, 1% to 2; pec- 
toral 2) to 2%; ventral 2! to 244. 

Color in alcohol brown generally, little paler below. Trunk and head with more 
or less rounded close-set darker to dusky blotches or spots, some ring-like and small 
spots, dots, bars or short lines in paler areas. On head usually dark underlaid streak 
from behind eye back across postocular and another from lower hind eye edge. Jaws 
usually with obscure brownish blotches. In young markings on trunk usually re- 
duced to large dark blotches, in very young forming about 6 dark vertical transverse 
bands, much wider than pale interspaces. Vertical fins pale, finely spotted with dark 
brown, spots fewer and larger in young, absent or little evident in very young. Pec- 
toral pale. Ventral pale basally, little dusky sub-terminally. 

Length 72 to 225 mm. 


Nine from Ningkwo. 


404 Bulletin American Museum of Natural History [Vol. L 


GOBIIDZ 


Eleotris potamophila Giinther 

Head contained 2% times in length to caudal base; depth 414 to 4%. Dorsal VI 
to VIII—I, 8, ror 9,1; anal I, 7,1. Scales 40 to 45 in median lateral series to caudal 
base and 4 or 5 more on latter; transversely at soft dorsal origin 13 to 15 scales; 30 
to 35 predorsal scales. Snout 3% to 4 in head measured from upper jaw tip; eye 
536 to 7%; maxillary 2)4 to 2%: interorbital 3% to 54. 

Body robust, subcylindrical, trunk more or less compressed posteriorly, deepest 
about dorsal origin. Caudal peduncle well compressed, least depth 1% to 1% its 
length or 34 to 3% in total head length. 

Head robust, depressed, width 1% to 2initslength. Snout broad, surface convex, 
length % to %4 its width. Eye slightly impinging on upper profile, center at first third 
in head; front pupil edge about first third in young; diameter 144 to 2 in snout, 1 to 
2% in interorbital. Mouth large, wide, mandible well protruded. Maxillary reaches 
opposite hind pupil edge, about to eye center in young; expansion 1) to 2 in eye. 
Lips firm. Teeth simple, conic, moderate, large, in bands of 3 or 4 irregular series in 
jaws; none on vomer, palatines or tongue. Tongue broad, slightly convex along 
entire front edge. Front nostril in small short tube, at last %in snout; hind nostril 
a simple pore, midway between front nostril and eye. Interorbital broadly and 
slightly depressed concavely. Preopercle entire, without spine. 

Gill-rakers 3+8 or 9, spinescent tubercles, greatly shorter than gill-filaments, 
which equal eye. 

Scales with 10 to 25 basal radiating strie; apical denticles 65 to 70, and 2 or 3 
series transversely; circuli fine. Jaws, snout, preorbital and under surface of head 
naked. Supraorbital with row of fine papille and above short series opposite nostrils; 
another row from close below eye up over postocular and suprascapula; double row 
along preopercle edge and continued forward along lower face of mandible to sym- 
physis; small cluster of papilla above maxillary on snout edge, also above front of 
supraorbital row; another cluster just below nostrils; infraorbital row begins just 
below hind nostril, slopes down toward maxillary, then abruptly up toward hind eye 
edge, where forking sends horizontal row back across cheek; below and parallel 
another midway on cheek, also small bar between posteriorly and anteriorly downward 
extension above upper maxillary edge; row down along front part of opercle closely 
parallel with preopercle edge. Breast and belly covered with small cycloid scales; about 
17 scales across cheek to preopercle edge; caudal and pectoral bases finely scaly. 

Third dorsal spine 2% to 3 in total length of head; third dorsal ray 2% to 2%; 
second anal ray 243 to 2%; caudal rounded behind, 14 to 1%; pectoral 144 to 1%; 
ventral 24% to 24%. Anal papilla moderate. 

Color in alcohol with back largely burnt umber, with 3 dark to blackish saddles, 
broadening as blotches above median lateral axis. First saddle includes most of 
spinous dorsal, second from hind half of soft dorsal-base and third not crossing caudal 
peduncle, but invading caudal base. Dusky loup across occiput. Under surface of 
head and belly pale, with dusky or smutty-brown spots. Dusky blotch before, below 
and behind eye. Soft dorsal and caudal gray-brown, with 6 or more blackish cross- 
lines. Pectoral pale brownish, with 6.transverse lines of deep brown. Two dusky- 
black spots at pectoral base, both externally andin axil. Ventral whitish, with median 
dusky blotches. 

Length 72 to 165 mm. 


Twelve from Ningkwo. 


1924] Fowler, Some Fishes Collected in China 405 


TETRODONTIDA 
Spheroides rubripes (Schlegel) 


_ Head contained 3% times in length to caudal base; depth 2%4. Dorsal rv, 12 to 
15; anal rv, 8 to 10; pectoral 1, 17 or 18. Snout 2% to 244 in head; eye 6 to 7; 
mouth width 3% to 4; interorbital 1% to 2%. 

Body robust, moderately long, deepest about pectoral base and back broadly 
convex. Caudal peduncle conic, least depth 1% its length or 3% in head. 

Head about as wide as deep, upper profile evenly convex; width 1 to 1% in its 
length. Snout broad, convex over surface and in profile, length % to % its width at 
front of eyes. Eye small, elevated, hind edge midway in head; diameter 24 in snout, 
3% to 4% in interorbital. Mouth moderately wide, terminally inferior. Lips thick, 
fleshy, with striate papille, which also extend over most of lower jaw or mandible. 
Teeth with entire edges, median groove well defined above and below. Nostrils 
lateral in oval cutaneous sac, outer slit much the larger; falls just before last fourth in 
snout profile. Interorbital widely convex. 

Gill-opening 4 to 44 in head. Line of lateral mucous system encircles eyes, then 
follows along upper side of back well above pectoral back to middle of caudal base; 
branch below nasal sac, another above pectoral origin upward and finally one from 
junction behind eye downward. Predorsal, interorbital, postorbital, breast and ab- 
domen with prickles, slightly larger on abdomen, variably more or less meeting behind 
depressed pectoral. 

Dorsal begins a little nearer caudal base than pectoral origin, rounded or median 
rays longest; fourth ray 2) to 244 in head. Anal similar and opposite dorsal; fourth 
ray 2 to 24 in head. Caudal convex behind, 134 to 1%4in head. Pectoral broad, hind 
edge convex, reaches % to % to anal, 2 to 2% in head. 

Color in alcohol dark neutral-gray on back, below whitish. Blackish blotch 
opposite hind half of depressed pectoral and another at dorsal base both with line 
of pale to whitish boundary. Four or 5 similar pale lines cross back, often broken or 
variably incomplete. Fins pale, often dark spot at pectoral base. Iris pale slaty. 

Length 116 to 143 mm. 


Eleven from Ningkwo. 
A very interesting species, quite strikingly marked and not rare in 
China and Japan. Not previously reported from Chinese fresh-waters. 


‘The Affinities of the Fish Lycoptera M iddendor ffi 


- By T. D. A. CockeRELn 


BULLETIN 
| OF : 
THE AMERICAN MUSEUM OF NATURAL HISTORY 
"Vou. LI, Ann. VIIL, pp. 313-317 | 
New Fork 


Issued. February 11, 1926 


56.7,41L 


Article VIII—THE AFFINITIES OF THE FISH LYCOPTERA 
MIDDENDORFFP 


By TD: AL CockEREers 
Plate III; Text Figure 1 


Among the interesting fossils obtained by the Third Asiatic Expedi- 
tion of The American Museum of Natural History in 1922 were numer- 
ous fishes from the Ondai Sair formation of Mongolia, ascribed to the 
Lower Cretaceous. These appear on examination to be old and young 
of a single species, Lycoptera middendorffi, described by Johannes 
Miiller in 1848 from shales supposed to be of Jurassic age, occurring in 
the Transbaikal region of Siberia. The problem of stratigraphy and 
correlation has been discussed as fully as circumstances permit in another 
paper, and it remains only to give a fuller account of the fish remains, 
which are of quite unusual interest. 

Lycoptera is referred by Doctor A. 8. Woodward to the family 
Leptolepide, of which he writes 2 


The Leptolepide differ from the two preceding families [Pholidophoride and 
Oligopleurids] in the absence of fulcra on the fins [whereby they agree with all living 
Isospondylous fishes], and are remarkable as being the earliest family in which inter- 
muscular bones occur. These elements, forming so conspicuous a feature among 
modern fishes, appear to be arranged here only in a single series above the vertebral 
column in the abdominal region; though there may perhaps be traces of them some- 
times also in the lower half of the caudal region. The vertebral centra of Leptolepis 
itself exhibit interesting gradations in the degree of development according to the 
geological age of the species, these centra never being more than delicate constricted 
rings or cylinders in the Upper Lias, and always strengthened by secondary periph- 
eral calcifications in the Oxfordian [Middle Jurassic] and upwards. In Thrissops, 
* which ranges as far at least as the Lower Cretaceous, the vertebral centra are still 
more robust. These fishes, it will be noticed, approach very closely the Clupeide, 
among which they are sometimes included; but they are distinguished by the meeting 
of the parietal bones in the median line, by the non-fusion of the haemal spines at 
the base of the tail, and by the presence of a thin film of ganoine on the scales. [Re- 
marks in brackets are mine.] 


Jordan and Branner* say of the Leptolepide: 


This family stands almost intermediate between the Ganoids and the Isospondyli. 
It has the general fin arrangement of the latter, but the scales are more or less diamond- 
shaped and ganoid on their exposed parts, and the last vertebra are more or less 
turned upward, although the tail is usually or always forked. The orbital plates 
cover the cheek as in the Elopide, but there is no gular plate, so far as known. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 39. 

21895, ‘Catalogue of Fossil Fishes in the British Museum (Natural History),’ III, p. xxi. 

31908, Smithsonian Miscellaneous Collections, LII, Part 1, p. 13. 


313 


314 Bulletin American Museum of Natural History (Vol. LI 


However, a gular plate is clearly shown in Reis’s figure of Lycop- 
tera middendor ffi. 

Boulenger! separates the Leptolepidz from other malacopterygian — 
families only by the perforation of the vertebral centra, at least so far 
as his table shows. He recognizes that the coating of the scales with 
ganoin is not a family character. 

We may readily gather from all this that the Leptolepide, beginning 
in the Upper Lias, represent the line of .development which first points 
distinctly to the modern dominant isospondylous and related fishes. It 
is a group which therefore deserves minute analysis to determine what 
characters were evolving and in what manner. Gregory? has recently 
remarked: ‘‘The Jurassic Leptolepide are the earliest known true 
teleosts, with their cycloid scales, vertebral centra nearly complete, 
no fin fulera, intermuscular bones present, and head and jaws remark- 
ably like those of primitive Clupeide. The homocercal tail sometimes 
develops hypural bones of primitive teleost type.’’ He here refers to 
Woodward’s figure of tail of Leptolepis dubius. 

An analysis of the above statements and of the known characters 
of the fishes serves only to bring out more clearly the relationship to 
existing families. The supposed family character of more or less ganoid 
scales breaks down entirely, as Gregory has indicated. As regards the 
vertebral centra, in some forms at least they appear to be well ossified, 
and Lycoptera during its lifetime goes through about the same stages 
described by Woodward as existing in successive species of Leptolepis. 
The distinct elevation of the end of the vertebral column corresponds to 
the condition found in the young of Salmo and other fishes. 

I am particularly fortunate in being able to present some enlarged 
figures of the scales of Leptolepis and Thrissops, photographed by Mr. « 
Herring under the direction of Doctor A. S. Woodward at the British 
Museum. The type of Leptolepis is L. coryphenoides (Bronn)? from the 
Upper Lias. It may be that ZL. dubcus (Blainville) from the Lithographic 
Stone of Bavaria is not strictly congeneric, in which case Ascalabos 
Munster may be available in a subgeneric sense. A photograph of the 
scales of L. dubiws (which may possibly differ somewhat from those of the 
type of the genus and family) shows the following characters: scales 
cycloid, broader than long, apex very broadly rounded, base truncate, 
laterobasal angles obtuse; nucleus central; circuli fine and distinct, 


11904, ‘Cambridge Natural History,’ VII, p. 544. 
21923, Bull. Amer. Mus. Nat. Hist., XLVIII, p. 241. 
3’Cyprinus coryphenoides Bronn, 1830; Leptolepis bronnii Agassiz, 1832. 


1925] The Affinities of the Fish Lycoptera Middendor ffi 315 


concentric; no radii. The laterobasal circuli cut the more central ones 
somewhat as indicated in Geinitz’s figure of Kymatolepis, but about half- 
way between nucleus and margin. A distinct ridge on each side passes 
from the region of the nucleus to the laterobasal corner. 

Now it is singular that the features of this scale, even to the inter- 
ference of the circuli laterally and the indication of ridges from the 
nucleus to the laterobasal corners, are very nearly those of the living 
Caranz hippos (Linneus), or Carangus hippos as Jordan has it. The 
ridges, however, are just as well seen in certain salmonoid scales, such as 
Leucichthys nigripinnis (Gill). The fish Leptolepis dubius is of course 
quite distinct from the Salmonide, and still more so from the Carangi- 
de, but I believe that the scale-structure may be regarded as prophetic. 
That is to say, the scale type of the modern fishes was acquired earlier 
than many of the other structures, and hence when found in a fossil 
may be used to indicate the line of development initiated by the latter. | 
If this generalization is valid, it affords us a very important clue to the 
evolution of fishes. In the case of Thrissops, I am fortunate in having a 
photograph of the scales of the type species, 7. formosus Agassiz, from 
the Lithographic Stone of Bavaria. The scales,are extremely broad and 
very different from those of Leptolepis. They may be described as 
follows: scales transverse, very much broader than long; apical margin 
simple, broadly rounded or slightly angular; some distance below the 
margin, but well above the nucleus, is usually a more or less irregular or 
wavy band, interpreted as an annulus; occasional feeble rudiments of 
apical radii, but distinct basal radii of the type seen in Ichthyodectes; 
circuli extremely fine, concentric. 

These scales much resemble those of the Cretaceous Hypsodon, lack- 
ing, however, the prominent tuberculation of the apical area. There isa 
superficial, but probably not significant, resemblance to the scales of 
the modern Hemirhamphide and Exoccetide. 

Coming now to Lycoptera middendorffi, we find still another quite 
diverse type of scale (Fig. 1). The scales are minute, about 1.2 mm. 
across, cycloid, transversely broad oval, without corners; nucleus central; 
circuli fine and concentric; apical and basal radii, the basal (about 14) 
extending up the sides nearly to meet the apical; apical radii about 13 
but weaker and shorter, irregular. In some scales the radii are less 
developed or evanescent. 

This scale is practically identical with that of the European min- 
now, Phoxinus phoxinus. I am unable to point out any difference 
which could be called generic. Lycoptera has so much the aspect of a 


316 Bulletin American Museum of Natural History [Vol. LI 


cyprinid that Egerton long ago proposed to place it in the cyprinid genus 
Aspius. It isa freshwater fish, and in it think we must see the ancestor 
of the Cyprinide and their allies. 

That it is not a cyprinid is shown by the following characters. 


1.—Teeth are present on the premaxilla, maxilla, and dentary, 
though extremely small. I could not demonstrate these teeth in my 
material and for a time believed them absent, but in one specimen I was 
able to see small sockets, and a specimen from the original (Siberian) 
lot in the British Museum ¢learly shows the teeth. Through the kindness 
of Doctor A. 8. Woodward, I am able to present an enlarged figure of 
this specimen. 


2.—As shown by Reis, there is a gular plate resembling that of 
-Amia. 


MMe 
//. 


Ny G3 

i “ijn G 43 = 
UU fe 

Hiri 


Mf 
ni 


Fig. 1. Two scales of Lycoptera from the Ondai Sair shales, showing their 
general character. 


3.—The end of the vertebral column turns upward, though the 
caudal fin is bifurcated and entirely like that of a cyprinid in appearance. 

4.—The anterior vertebre are not modified. 

Characters 2 and 3 suggest a certain affinity with Ama, but the 
scales are wholly different. The dorsal fin is far posterior, about opposite 
the anal, as in the cyprinid genus Engraulicypris. In the adult the 
vertebral centra are solid, yet apparently always with a central canal for 
the notochord. 

Doctor A. S. Woodward has described a second Lycoptera, L. 
sinensis, from the Lower Jurassic (?) of the Province of Shantung, 
China. Our specimens are, however, referable to L. middendorffi. 


1925] The Affinities of the Fish Lycoptera Middendor ffi 317 


I think it is now evident that the Leptolepidee of authors must be 
subdivided, at least to the following extent. 


1.—Leptolepide. Marine (or sometimes brackish water ?) fishes, 
definable as indicated above, 
a.—Leptolepinz, new subfamily. Type Leptolepis, with scales 
as described above. (For other characters of these 
genera see Woodward.) 
b.—Thrissopsine, new subfamily. Type Thrissops, with 
broad scales as described above. From the position 
of the dorsal fin opposite the anal, we might infer 
affinity with Lycoptera, but the scales are entirely 
different; so also the Coryphzna-like caudal fin. 
2.—Lycopteride. Freshwater fishes, with scales resembling those 
of Phoxinus. Lycoptera, two species, or three if we include 
Prolebias davidi of Sauvage, 1880, ae om the supposed 
Tertiary of northern China. 

Other genera referred to Leptolepide will have to be left in a rather 
uncertain position until their scales can be critically examined, but 
there is no indication that any one of them belongs to Lycopteride. In . 
the Kimmeridge shale at Ringstead, Dorset, England, I collected circular 
or subcircular scales, 3 to 4 mm. diameter, with concentric circuli, 
strong annuli, and noradul. Doctor A. S. Woodward believes they must 
belong to Leptolepidz, but they seem to represent a genus distinct from 
Leptolepis dubius. 


Puate III 

Fig. 1. Enlarged figure of head of Lycoptera middendorffi. British Museum, 
P. 1841. 

Fig. 2. Median dorsal scales of Thrissops formosus. Lithographic Stone, Soln- 
hofen, Bavaria. British Museum, P. 9138a. 

Fig. 3. Anterior dorsal scales of Leptolepis dubius. Lithographic Stone, Soln- 
hofen, Bavaria. British Museum, P. 924. 

Figs. 4-8. Lycoptera middendorffi, from Ondai Sair. 


Fig. 4. General view of a small specimen. 

Fig. 5. A larger specimen that preserves the head and pelvic fins fairly well. 

Fig. 6. A large fish, in which the pectoral pelvic and anal fins are well preserved. The scales, 
shown in text figure 1, are drawn from the specimen. 

Fig. 7. Well-preserved anal and part of a dorsal fin. 

Fig. 8. A good caudal fin. 


Bulletin A. M. N. H. Vol. LI, Puats III 


fi y 


ni \ 
im ue 
: rf 


ao . oe 


Re g ty oe 
te. 


ie 


aT the nee, 
| ee ‘- 
a = = 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 156 Tur AMERICAN Museum or Naturat History Feb. 11, 1925 
New York City b 


56.81,9P(117:51.7) 
ON PROTOCERATOPS, A PRIMITIVE CERATOPSIAN 
DINOSAUR FROM THE LOWER CRETACEOUS 
OF MONGOLIA! 


By WiuuraAm K. GREGoRY AND CHARLES C. Moox 


One of the most interesting of the many remarkable discoveries of 
the American Museum Third Asiatic Expedition is a small predentate 
dinosaur from the lower Cretaceous of Mongolia, which has been named 
Protoceratops andrewsi by Granger and Gregory. The type skull, which 
was received and worked out of the matrix some time before the rest of 
the seventy-odd specimens of the form arrived at the Museum, proves 
to be that of a young animal with an estimated extreme skull length of 
less than ten inches, while the largest skull of Protoceratops measures 
twenty-three inches in length. 

Except for the absence of horns, the whole configuration of the skull 
conforms to the general ceratopsian type: the parietals and squamosals 
are produced into a perfectly formed, fenestrated occipital frill; the 
enlarged squamosal enters the side of the frill and is in contact below 
with the jugal; the mandible and the crowns of the teeth are much like 
those of later ceratopsians. On the other hand, many primitive features 
are retained which were lost or disguised in the typical ceratopsians. 
Thus, in half-grown specimens the “‘frill’”’ is seen to be merely an enlarged 
scaffolding for the powerful jaw and neck muscles; there are no epoccipi- 
tal bones, and the lateral and superior temporal fenestre are instantly 
recognizable as such, whereas in later Ceratopsia, through the continued 
growth of the frill, their original character is largely concealed. The 
opposite prefrontals and ‘‘postfrontals’” (postorbitals) remain in their 
primitive positions and do not form a secondary roof above the frontals 
as they doin the typical Ceratopsia; hence there is no median “ pseudo- 
pineal”’ or postfrontal foramen. Freely articulating palpebral bones are 
attached to the anterosuperior corner of the orbits, as in Psittacosaurus. 
The premaxille each bear two fairly long, cylindrical teeth, instead of 
being edentulous as in the later Ceratopsia, and the very deep beak, in 
contrast to the anteroposteriorly elongate beak of Triceratops, also re- 


‘Publications of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
bution No. 40, 


2 AMERICAN MUSEUM NOVITATES [No, 156 


calls that of such relatively primitive predentates as Pszttacosaurus. The 
anterior nares are simple ovals and the premaxille lack all the peculiar 
specializations of the later ceratopsians. The cheek teeth succeed each 
other in a closely appressed vertical series, including not more than two 
rows, in contrast with the numerous rows of the later Ceratopsia. The 
roots are single, not cleft buccolingually. The orbits are relatively 
much larger than in typical ceratopsians, especially in the young skulls. 
The preorbital fossze are very large depressions, instead of being narrow 
slits, and the lachrymal bones are of considerable size. The well-developed 
parietal fontanelles are transversely oval, instead of being produced 
anteroposteriorly as in typical ceratopsians. 

Many very primitive features are also retained in the postcranial 
skeleton. The first three cervical vertebre, although appressed, are not 
coalesced, and the spine of the axis is not produced backward as it is in 
the typical Ceratopsia. The scapula is already elongate to aid in sup- 
porting the large head, but the fore limb as a whole is not nearly so large 
in proportion to the hind limb as it is in typical Ceratopsia; the small 
manus also is much smaller than the pes, the middle digit of the manus 
being much shorter than the middle digit of the pes, whereas in typical 
ceratopsians it is as large or larger. 

The pelvis is in many respects remarkably primitive: the dorsal 
border of the ilium is vertical in position, whereas in later Ceratopsia 
(except Leptoceratops) it is more or less reflected outward and finally 
forms a wide shelf above the femur. The prepubic process of the pubis, 
instead of being a very large, vertically-extended process, as in later 
Ceratopsia, is relatively small and but little extended vertically; the 
postpubic process is relatively much less reduced than in the later types. 
The sacral complex includes seven to eight vertebre, in contrast with the 
ten of Triceratops. The femur retains a large fourth trochanter and is 
slightly shorter than the tibia, while in typical Ceratopsia the fourth 
trochanter is reduced and the femur is longer than the tibia. The hind 
foot is remarkably long and slender for a ceratopsian, and is in fact more 
like that of Psittacosaurus than like that of Triceratops or Monoclonius. 
In short, the whole proportions and configuration of the skeleton indi- 
cate that Protoceratops was not far removed from the ancestral bipedal 
ornithischian, the former existence of which was long since inferred by 
Dollo (1905) after a brilliant analysis of the various types of dinosaur- 
ian pelves. The tail vertebree have very long neural spines, while in 
Triceratops the neural spines are much shortened. The feet and tail 
of Protoceratops possibly indicate partly aquatic habits. 


A.M.6408 CO oe 


Fig. 1. Protoceratops andrewsi. Small, young adult skull, possibly a female. 
A. M. No. 6408. Viewed from above. Two-fifths natural size. 


The occipital frill is composed exclusively of the expanded parietals. There is no good evidence of a 
separate interparietal or fused dermosupraoccipitals. The squamosals are limited to the anteroexter- 
nal border of the frill. There being no horns on the postorbitals, there is no secondary skull roof 
above the frontals and consequently no ‘‘ pseudopineal”’ opening. . 


3 


4 AMERICAN MUSEUM NOVITATES [No. 156 


The definition of the family Protoceratopside, proposed by Granger 
and Gregory, may now be extended as follows: 

Primitive small ceratopsians, with a hornless skull, without either secondary 
skull roof or pseudopineal foramen above the frontals, no epoccipital bones; with 
simple oval anterior nares and unspecialized premaxille. A well-developed occipital 
frill, with large transversely oval parietal fontanelles. Freely articulating palpebral 
bones (supraorbitals) attached to the anterosuperior corner of the orbits. Pre- 
maxillaries with teeth. Cheek teeth arranged in a vertical series of not more than 
two developed at one time; roots simple (not bifid). Fore limb slender, manus much 
smaller and shorter than pes, the latter elongate, compressed. Sacral complex of 
seven or eight vertebrae. Ilium with blade but slightly inclined outward to the sagittal 
plane, not reflected or produced laterally above the femur. Prepubic process rela- 
tively small, not expanded vertically; postpubic process but little reduced. Femur 
with large fourth trochanter, femur shorter than tibia. Midcaudal vertebre with 
very long spines. 

The genus Leptoceratops of Brown (1914) is a little-modified survivor 
in the Edmonton formation in North America, of the Protoceratopside. 
It agrees with Protoceratops in the following characters. 

(1.) The roots of the cheek teeth are simple, not bifid. 

(2.) The nasals are hornless. 

(3.) The parietal frill has a long, high, sagittal crest. 

(4.) The mandible is short and deep, in contrast with the elongate 
proportions of the mandible in Triceratops. 

(5.) The femur has a large fourth trochanter. 

(6.) The tibia is longer than the femur. 

(7.) The medial surface of the ilium is not reflected or produced 
outwardly over the femur, the whole ilium being extremely like that of 
Protoceratops. 

(8.) The ischium is long, not shortened as in the typical ceratopsians. 

(9.) The neural spines of the caudal vertebre are extremely high in 
proportion to the anteroposterior extent of the centrum, in contrast with 
the very low spines of Triceratops. 

On the other hand, Leptoceratops has progressed toward the typical 
ceratopsian in (1) the reduction of the parietal fontanelles, (2) the 
coalescence of the anterior three cervical vertebra, (8) the backward 
prolongation of the spine of the axis above the third cervical vertebra, 
(4) the shortening of the metacarpals and especially of the metatarsals. 

The genus Brachyceratops Gilmore, from the Belly River formation 
of the upper Cretaceous, clearly belongs with the Ceratopside, of which 
it represents a young individual and a very primitive stage. Thus it 
shows: (1) an early stage in the development of the horns, (2) the begin- 
ning of the lateral reflection of the iliac blade, (3) the incipient shortening 


1925} A PRIMITIVE CERATOPSIAN DINOSAUR 5 


of the tibia as compared with the femur, (4) the lack of coalescence of 
the two halves of the secondary skull roof above the frontals. 

Of the typical Ceratopside, the genus Ceratops (Chasmosaurus) 
represents a stage in which the frill has become produced backwardly, 


A.M.6414 


Fig. 2. Protoceratops andrewsi. Side view of supposed old male skull. A. M. 


No. 6414. One-fourth natural size. 


Although considerably distorted by the lens, this view shows well the great depth of the beak, which 
recalls that of Psittacosaurus, the shortness and depth of the lower jaw, the height and prominence of 
the sagittal crest of the parietal, the relatively small size of the orbit as compared with that of the 
younger skull, etc. The great development of the occipital frill and of the jugal is obviously not pri- 
marily for the protection of the neck but for the support of the robust muscles necessary to operate the 
great Heal and grinding apparatus. The cheek teeth are not shown in this specimen. One-fourth 
natural size. 


elongating the parietal fontanelles so that their longitudinal much exceeds 
their transverse diameter. The nasal and supraorbital horns are still 
small. Torosaurus, with its highly fenestrated frill, would be the logical 
outcome of this line of development. 


' 


6 AMERICAN MUSEUM NOVITATES [No. 156 

Styracosaurus would seem to be an aberrant offshoot of a primitive 
Ceratops. In all these the parietal crest is produced far behind the 
squamosals. 

Monoclonius may well be a direct descendant of Protoceratops, in 
which the nasal convexity has grown up into a long horn, and the frill, 
acquiring epoccipital bones, has been produced at the upper end into the 
erratic processes of Centrosaurus. 

Anchiceratops may be regarded as one of the Monoclonius group, 
with a secondary tendency for the closure of the fontanelles. 

The Triceratops-Diceratops group may also be derived from Proto- 
ceratops, perhaps by way of a form related to Leptoceratops, which already 
shows a strong tendency toward the secondary closure of the parietal 
fontanelles and a transverse widening of the crest. 

Although most of the Protoceratops material has still to be cleaned 
up, a series of skulls has been worked out, starting with an extremely 
young stage not long out of the egg, and ending with a very old stage 
with a wide frill 511 mm. wide. In one of the younger stages, witha total 
skull length of 283 mm., the frill is not much wider than the skull itself, 
but is already produced behind the occipital condyle. The parietal 
fontanelles are large, broad ovals. The orbits are relatively very large 
and the snout is short. As growth proceeds the crest becomes relatively 
larger and much wider, and the orbits become relatively smaller, the 
snout more compressed and the lateral temporal fenestre smaller. There 
seem to be two kinds of skulls, a long and a very broad kind, possibly 
representing females and males. Specific differences have not yet been 
worked out. 

In conclusion, Protoceratops affords decisive evidence for Dollo’s 
inference that the gigantic quadrupedal Ceratopsia have been derived 
from some small bipedal predentates. In skull characters it is already 
in a primitive ceratopsian stage, but its postcranial skeleton retains 
much of the bipedal heritage, especially in the pelvis and hind limbs, 
which was lost by its gigantic graviportal descendants. 

When compared with the small bipedal predentates Pszttacosaurus 
and Protiguanodon, also of the Cretaceous of Mongolia, Protoceratops 
exhibits such a great number of significant agreements in the skull, 
dentition, vertebre, and limbs, that the existence of an earlier common 
ancestral stock is virtually demonstrated, Psittacosaurus retaining much 
the greater number of primitive characters. More precisely, some pre- 
Wealden bipedal predentate closely allied to Hypsilophodon appears to 
be indicated as the common ancestral stock, not only for the camptosaurs, 


AM.6417 aA 


Fig. 3. Protoceratops andrewsi. Under side of a nearly complete skeleton. A. 
M. No. 6417. One-tenth natural size. 


Note the large size of the hind limb and foot, as compared with the forearm and hand; the elonga- 
tion of the metatarsus as in bipedal Ornithischia. The dorsal border of the ilium is not reflected outward 
above the acetabulum. The prepubic process of the pubis, while stout, is not nearly as long or as much 
expanded vertically as it is in later Ceratopsia, and the postpubic process is not so much reduced. The 
long ischia are concave on the lower border but lack an obturator process. The spines of the caudal 
vertebre are very long and slender. 


ba | 


8 AMERICAN MUSEUM NOVITATES [No. 156 


iguanodons, trachodons, and corythosaurs, but also for the variously 
specialized psittacosaurs, troddons, acanthopholids, nodosaurs or anky- 
losaurs, and ceratopsians. The stegosaurs of the Comanchean, retain- 
ing a relatively unspecialized skull and pelvis, are on the whole an older 
branch usually recognized as derived from the Liassic Scelidosaurus. 

The Protoceratops material is also of interest as affording strong 
evidence for the older view that the middle part of the ceratopsian frill is 
formed from the parietals, as maintained by Marsh, Hatcher, Lull, and 
Lambe, in opposition to the newer view of Hay, von Huene, and Gilmore 
that it is formed from the enlarged dermosupraoccipitals. These origi- 
nally paired elements are unknown in any other group of dinosaurs and, 
aside from doubtful vestiges in Crocodilia, appear to be limited to the 
Permian orders of reptiles. 


LITERATURE CONSULTED 


Apams, L. A., 1919, ‘A Memoir on the Phylogeny of the Jaw Muscles in Recent and 
Fossil Vertebrates,’ Ann. N. Y. Acad. Sci., XXVIII, pp. 51-166, Pls. 
I-XIII. 


Brown, B., 1908, ‘The Ankylosauride, a New Family of Armored Dinosaurs from 
the Upper Cretaceous,’ Bull. Amer. Mus. Nat. Hist., XXIV, pp. 187- 
202. 
1914, ‘ Anchiceratops, a New Genus of Horned Dinosaurs from the Edmon- 
ton Cretaceous of Alberta,’ Bull. Amer. Mus. Nat. Hist., XX XIII, 
pp. 539-548, Pls. xxrx—xxXXvVII. 
1914, ‘Leptoceratops, a New Genus of Ceratopsia from the Edmonton 
Cretaceous of Alberta,’ Idem, pp. 567-580, Pl. x1. 
Dotto, L., 1905, ‘Les Dinosauriens adaptés a la Vie Quadrupéde Secondaire,’ Bull. 
d. |. Soc. Belge d. Géol., etc., XIX, pp. 441-448, Pls. x1, xm. 
GitmorE, C. W., 1917, ‘Brachyceratops, a Ceratopsian Dinosaur from the Two Medi- 
cine Formation of Montana,’ U. 8. Geol. Surv., Prof. Paper 103. 
1919, ‘A New Restoration of Triceratops, with Notes on the Osteology of 
the Genus,’ Proc. U.S. Nat. Mus., 55, pp. 97-112, Pls. 11-1x. 
1922, ‘The Smallest Known Horned Dinosaur, Brachyceratops,’ Idem, 61, 
pp. 1-4, Pls. r-tv. 
1924, ‘On Troédon validus, an Orthopodous Dinosaur from the Belly River 
Cretaceous of Alberta, Canada,’ Univ. of Alberta Press, Bull. No. 1, 
pp. 1-48, Pls. xv. 
GRANGER, W., AND Grecory, W. K., 1923, ‘Protoceratops andrewsi, a Pre-Ceratop- 
sian Dinosaur from Mongolia,’ Amer. Mus. Novitates, 72, pp. 1-6. 
Gkecory, W. K., 1920, ‘A Review of the Evolution of the Lacrymal Bone of Verte- 
brates, with Special Reference to that of Mammals,’ Bull. Amer. Mus. 
Nat. Hist., XLII, pp. 95-263, Pl. xvir. (Homology of the antorbital 
fenestra of Ceratopsia, pp. 127, 128.) 


1925] A PRIMITIVE CERATOPSIAN DINOSAUR — 9 


Hartcuer, J. B., Marsu, O. C., Luu, R. §., 1907, ‘The Ceratopsia,’ Monogr. U. 8. 
Geol. Surv.,. XLIX, pp. 1-198, Pls. 1-11. 

Hay, O. P., 1909, ‘On the Skull and the Brain of Triceratops,’ Proc. U. 8. Nat. Mus., 
XXXVI, pp. 95-108, Pls. 1-11. 

Hveng, F. von, 1912, ‘Beitriige zur Kenntnis des Ceratopsidenschiidels,’ N. Jahrb. 
f. Min., Geol. u. Pal., 2, pp. 146-162. 

Lamse, L. M., 1915, ‘On Eoceratops canadensis, gen. nov., with remarks on other 


genera of Cretaceous horned dinosaurs,’ Canada Geol. Surv. Mus. Bull., 
12; Geol. Ser. 24. 


Lut, R. S., 1903, ‘The Skull of Triceratops serratus,’ Bull. Amer. Mus. Nat. Hist., 
XIX, pp. 685-695. 
1905, ‘A Restoration of the Horned Dinosaur Diceratops,’ Amer. Jour. 
Sci., XX, pp. 420-422. 
1908, ‘The Cranial Musculature and the Origin of the Frill in the Ceratop- 
sian Dinosaurs,’ Amer. Jour. Sci., XXV, pp. 387-399. 
1921, ‘The Cretaceous Armored Dinosaur Nodosaurus textilis,’ Amer. 
Jour. Sci., Ser. 5, I, 2, pp. 97-126. 
Meyer, H. von, 1859, ‘Stenopelix valdensis, ein Reptil aus der Walden-Formation 
Deutschland’s,’ Palzeontographica, VII, 1, pp. 25-34, Pls. 1v, v 
Norcsa, F., 1915, ‘Die Dinosaurier der Siebenbiirgischen Landesteile Ungarns,’ 
Jahrb. d. Kgl. Ungar. Geol. Reichs, X XIII, 1, pp. 1-24, Pls. r-1v. 
1918, ‘Leipsanosaurus, n. g., ein neuer Thyreophore aus der Gosau,’ Féld- 
tani Kézlony, XLVIII, 7-9, pp. 324-328, Pl. m1. 
1923, ‘The Primitive Reptilia of the Upper Cretaceous in Hungary,’ Quart. 
Jour. Geol. Soc., LX XIX, 1, pp. 100-116. 
1923, ‘Notes on British Dinosaurs. Part VI: Acanthopholis,’ Geol. Mag., 
LX, pp. 193-199. 
Ossorn, H. F., 1924, ‘Psittacosaurus and Protiquanodon: two Lower Cretaceous 
Iguanodonts from Mongolia,’ Amer. Mus. Novitates, 127, pp. 1-16. 
Pompecks, J. F., 1920, ‘Das Angebliche Vorkommen und Wandern des Parietal- 
foramens bei Dinosauriern,’ Sitzungsber. d. Gesellsch. Natiirforsch. 
Fr., Berlin, 3, pp. 109-129. 
1921, ‘Besass der Dinosaurier TJ'riceratops ein Parietal-foramen?,’ Idem, 
1-3, pp. 1-18. 


mol! es Ai oe 


= cong 
rats aan, An Rhy nice 
SN ke i‘ Se | Nis 


sbi iW font) awe Nd NE 
A aa 
* ; i I ae vt arte AP oa ene 


7 7 
by Morn: or aie! is wit 


a « a } ¥ 
‘ 
Lp ; ; ee é » 1 ite) Oy a 
F = 
' j 
AA ry 3 | i : ee 
” ‘ +o 1 rhe the 
i eae | 
‘ i 
Ah ’ P ap 
¥ + J 
t +) , 
* 
\ ’ i 
\ it 
, 7 Tae! cy ay 
' . 
i A 
” eae | 
i 
i ‘ 
bie 
‘ 
a ‘ 


A 
epee Tem F - 
ah Ga aa or j “< if = i y = 

a, | oii a Tita a ra wil? Oe" haw Ty ipa 98. : ~~ 
7 Ad fF 


: pan Cott 0” Veo : iy 
=—— } La? i? — : %, * 


» : 
ee “> a uP Beate 
Yon sf ; ‘ tt 


itech 

oy ’ 
Sai ey 
Pa 


os wa 
a 


‘ ie - * 1g) ‘f 


: : legit ag 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 157 Tue American Museum or Naturau History Feb. 13, 1925 
New York City 


59.81(51.7) 


NEW REPTILES AND A NEW SALAMANDER FROM CHINA! 
By Kari PATTERSON SCHMIDT? 


The collections of the several expeditions to China of The American 
Museum of Natural History contain a number of new species of reptiles. 
In addition to these I include here a preliminary notice of a new sala- 
mander presented to the Field Museum of Natural History by Mr. 
Robert B. Ekvall. More extended descriptions of the following species 
will be included in a forthcoming detailed report on the collections of 
reptiles thus far accumulated by Mr. Clifford H. Pope and other members 
of the Third Asiatic Expedition under the leadership of Mr. Roy Chap- 
man Andrews. 


TESTUDINATA 


Geoclemys grangeri,® new species 


Typr.—A. M. N. H. No. 23481; Yenchingkau, Wanhsien, Szechwan, 1500 feet 
altitude; November, 1921; Walter Granger. 

Diaenosis.— Differs from Geoclemys reevesii in having the axillary shield larger 
than the inguinal; the small occipital shields much smaller than in reevesii; and the 
spots of the plastral shields much smaller and more sharply defined. The gular suture 
more than twice that of the humerals; first marginal broadest; bridge a little longer 
than the posterior lobe of the plastron. 


SAURIA 


Sphenomorphus leveretti,+ new species 


Typre.—A. M. N. H. No. 30201; o; mountains south of Nodoa, Hainan; July 
30, 1923; Clifford H. Pope. 

Dracnosis.—Allied to Sphenomorphus indicus, from which it may be distin- 
guished by its more elongate snout; the longer and more pointed frontal; the greater 
extension of the rostral shield on the upper surface of the snout; the longer dorsal and 
nuchal scales; twenty-two sharply keeled lamellze beneath the fourth toe; and a more 
spotted pattern, the lateral black band of indicus being represented only by more 
numerous black spots, while the mid-dorsal area is heavily spotted with black, without 
the median black line occasionally found in indicus. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
bution No. 41. 

2Of the Field Museum of Natural History. 

‘Named for Mr. Walter Granger, Chief Paleontologist of the Third Asiatic Expedition. 

4Named for Rey. William J. Leverett of the American Presbyterian Mission, Nodoa, Hainan, in 
recognition of his invaluable aid to the work of the Expedition. 


2 AMERICAN MUSEUM NOVITATES [Ne. 157 


Leiolopisma septentrionalis, new species 


Typr.—A. M. N. H. No. 21451; Hsing Lung Shan, Eastern Tombs, Chihli 
Province. China; August 1 to 15, 1921; Clifford H. Pope. 

Diacnosts.—Body elongate, limbs failing to meet by half the length of the arm; 
digits five; head wider than the neck; an undivided transparent shield in the lower 
eyelid; no supranasals; ear opening without denticles; dorsal scales smooth, in 
twenty-eight rows around the body; a pair of enlarged preanals; anterior loreal 
smaller than the second; fifteen lamellee beneath the fourth toe; sides dark brown, 
back light metallic brown, the dorso-lateral line where the two colors meet not 
straight, regularly scalloped; throat with brown spots. 


Leiolopisma monticola, new species 


Typn.—A. M. N. H. No. 20998; Snow Mountain Village, 9000 feet altitude, 
Likiang, Province of Yunnan, China; November, 1916; R. C. ae and Edmund 
Heller. 

D1aGNnosis.—Body elongate, limbs weak, separated when adpressed by the length 
of the arm; digits 5; head wider than the neck; an undivided transparent shield in 
the lower eyelid; no supranasals; ear opening without denticles; dorsal scales 
smooth, slightly larger than the ventrals, in twenty-four rows around the body; a 
pair of enlarged preanals; anterior loreal longest; frontal in contact with the anterior 
two supraoculars; fronto-parietals larger than the interparietal; 12 lamelle beneath 
the fourth toe; back light brown with rows of darker spots; sides dark brown; venter 
very dark gray. 


Lygosaurus salsburyi,! new species 

Typre.—A. M. N. H. No. 30198; o; Nodoa, Hainan, China; January-July, 
1923; Clifford H. Pope. 

DriacNosis.—Very closely allied to Lygosaurus sowerbyi Stejneger, recently 
described from Fukien Province, from which it may be distinguished by the greater 
number of subdigital lamelle. ZL. salsburyi has from nineteen to twenty-one smooth 
lamelle beneath the fourth toe, as compared with fifteen or sixteen in sowerbyt. 


SERPENTES 
Sibynophis hainanensis, new species 
Typr.—A. M. N. H. No. 27788; o; Nodoa, Hainan, China; December 1922- 
July 1923; Clifford H. Pope. ; 
DriacGnosis.—Closely allied to Sibynophis collaris; maxillary teeth 40; rostral 


just visible from above; upper labials 8; parietals in contact with the lower post- 
ocular on each side; ventral plates 167; caudals 115. 


Natrix andrewsi,” new species 


Typr.—A. M. N.H. No. 28255; o; mountains south of Nodoa, Hainan, China; 
ee 30, 1923; Clifford H. Pope. 


1Named for Dr. Clarence G. Salsbury of the American Presbyterian Mission, Nodoa, Hainan, 
who also rendered important aid to the work of the Expedition. 
2Named for Mr. Roy Chapman Andrews, Leader of the Third Asiatic Expedition. 


1925] NEW REPTILES AND SALAMANDER FROM CHINA 3 


DiaGnosis.—Subgenus Macropophis Boulenger; maxillary teeth 37, uniform in 
front, the last gradually enlarged; body very slender; eye large; all the scales 
strongly keeled, in nineteen rows; ventral plates 164; caudals 118; anal divided; 
prominent vertical white bars in front of and behind the eyes. 


Natrix helleri,! new species 


Typre.—A. M. N. H. No. 20149; 9; Tengyueh, 5500 feet altitude, Province of 
Yunnan, China; April 24, 1917; R. C. Andrews and Edmund Heller. 

Diaenosis.—Closely allied to Natrix subminiata of Java and southeastern Asia, 
from which it is distinguished by a higher number of ventral scales, 163-172, compared 
with 132-157 in subminiata as here restricted. 

Dorsal scale rows 19, the outer smooth, the median rows sharply keeled; ven- 
trals 163-172; anal divided; caudals 75-86; upper labials 7-9, three entering the 
eye; a single preocular; three postoculars; temporals 2-2; general color uniform 
olive, with reddish markings on the neck, chiefly confined to the skin between the 
scales. 


Natrix nivalis, new species 


Type.—A. M. N. H. No. 21021; 9; Snow Mountain Village, 9000 feet alti- 
tude, Likiang, Province of Yunnan, China; November, 1916; R. C. Andrews and 
Edmund Heller. 

Diaanosis.— Directly derived from Natrix nuchalis, from which it is distinguished 
by the lower number of ventral plates, and a proportionately broad and short frontal. 
Dorsal scales weakly keeled, in 17 rows; ventrals 150-152; anal divided; caudals 
43-54; upper labials 6, the fifth very large; one preocular; postoculars 1-3; tem- 
porals 1-1 or 1-2; general color dark olive-brown, without markings at the base of 
the scales; venter dark gray, the median part black. 


Natrix popei,? new species 

Typr.—A. M.N. H. No. R 27763; o&; Nodoa, Hainan, China; December 1922- 
July 1923; Clifford H. Pope. 

Diacnosis.—Closely allied to Natrix vibakari and to Natrix sauteri of Formosa; 
maxillary teeth 20, the last gradually enlarged; anal divided; scales in nineteen rows; 
one or two anterior temporals; eight upper labials; fourth and fifth entering the eye; 
ventral plates 130-137; subcaudals 78-86; apical pits very faint, small, absent on 
most scales. 


Elaphe osborni,* new species 
Type.—A. M. N. H. No. 21073; 9; Tengyueh, Province of Yunnan, China; 
May 10, 1916; R. C. Andrews and Edmund Heller. 
DriaGnosis.—Most closely allied to Elaphe hodgsoni (Giinther) and Elaphe 
teniura Cope. 


rs 1Named for Mr. Edmund Heller, Assistant Curator of Mammals, Field Museum of Natural 
istory. 
2Named for Mr. Clifford H. Pope, Assistant in Zoélogy, Third Asiatic Expedition. 
3Named for Professor Henry Fairfield Osborn, President of The American Museum of Natural 
History, whose personal interest in the Museum’s Asiatic Expeditions has greatly furthered their 
> work. 


4 AMERICAN MUSEUM NOVITATES [No. 157 


Body form not specialized, ventrals not angulate, head distinct from neck; dorsal 
scales faintly but sharply keeled, in twenty-one rows; ventrals 215-225; anal divided; 
caudals 77-79; supralabials 8, 4th and 5th entering the eye; preoculars 2; post- 
oculars 2; temporals 2-3; color fawn, with black transverse bars anteriorly and longi- 
tudinal lines posteriorly; no black line through the eye. 


Gonyosoma caldwelli,' new species 


Typr.—A. M. N. H. No. 21010; o; Yenping, Fukien Province, China; 1916; 
H. R. Caldwell. 

_ Draenosis.—Very closely allied to Gonyosoma melli (Vogt) from Kwangtung and 
to Gonyosoma frenata Giinther of the Khasi Hills. 

Head and body elongate, body compressed; ventrals sharply angulate; snout 
obliquely truncate, projecting; dorsal scales very faintly keeled, in nineteen rows; 
ventrals 223; anal divided; caudals 108 (? +); supralabials 8, 3rd, 4th, and 5th 
entering the eye; no loreal; one preocular; two postoculars; temporals 1-2; uniform 
green above and below, with a black stripe through the eye. 


Boiga sinensis, new species 

Typr.—A. M. N. H. No. 23495; o; Fukien Province, China; 1921; H. R. 
Caldwell. 

DracGnosis.—Head short and broad, the snout longer than the diameter of the 
eye; body compressed, tail long; ventrals not angulate; anterior palatine teeth 
slightly enlarged; posterior pair of chin shields much smaller than the anterior; 
dorsal scales smooth, oblique, in twenty-one rows, the mid-dorsal row not enlarged ; 
ventrals 230; anal divided; caudals 127; preoculars 3; postoculars 2-3; temporals 
very small, 4 to 6 in the first row, 6 to 7 in the second, not regularly arranged; ground 
color light reddish brown, with three series of darker brown spots. 


Trimeresurus stejnegeri,? new species 


Typr.—A. M. N. H. No. 21054; &; Shaowu, Fukien Province, China; 1916; 
R. C. Andrews and Edmund Heller. 

Diagnosis.— Differs from the widespread Trimeresurus gramineus, with which it 
has hitherto been confounded, in the very small shields between the chin shields and 
the first ventral plate, the smaller and more widely separated supranasals, the dis- 
tinct first labial (which in south Chinese gramineus is frequently fused with the nasal), 
and the usual uniform green coloration of the side of the head. 


Trimeresurus yunnanensis, new species 
Typr.—A. M. N. H. No. 21058; o; Tengyueh, Yunnan Province, China; 
May 18, 1917; R. C. Andrews and Edmund Heller. 
Diagnosis.—Closely allied to the preceding species, and distinguished from 7’. 
gramineus by the same characters. It is distinguished by having only nineteen rows of 


f 1Named for Mr. Harry R. Caldwell, who is largely responsible for the Expedition’s Fukien 
collections. 

2Named for Dr. Leonhard Stejneger, Head Curator of Biology, United States National Museum, 
as a small tribute to his invaluable contributions to Oriental herpetology. 


1925] NEW REPTILES AND SALAMANDER FROM CHINA 5 


dorsal scales at mid-body, and twenty-one on the neck, compared with twenty-one at 
mid-body and 23-25 on the neck in 7’. stejnegeri. The average number of ventral 
plates, 155-160, is perhaps somewhat lower in T.. yunnanensis. 


CAUDATA 


Batrachuperus tibetanus, new species 


Typr.—F. M. N. H. No. 5900; adult female; near the Tibetan border of Kansu, 
southwest of Titao, 9000 feet altitude, in Hwang Ho drainage; December, 1923; 
Robert B. Ekvall. 

DiaGnosis.—Closely allied to Batrachuperus sinensis, from which it may be 
distinguished by the more posteriorly situated vomerine teeth; the more depressed 
head; the fourteen costal grooves; the absence of horny covering on the palms and 
soles, only the tips of the digits having a horny epidermis; the somewhat longer tail, 
.49-.52 of the total length; and the much lighter coloration. 


j re 
et d- de het. 


eye ae —) <4 a > he PY REP Hed ae gety.7 Lh 
fr te oe pent a iF alee nee Ping pyle Reale wah) 


‘eu et re eee a ay opts 4 he W Pi 
Fe mare ee oi oe apoed oy 


dent tod fa we 
pee : : ve 


bade 


a ee ee 

, : of 

a a Wn *: 2 a ei 9 ee S Peay Fe WE I are Al v 
a i sis i 
= I 
i co 
1 wy 
’ oe eo 5 P< ~ 1 
3 ae 
ie 

; 7 ers F ‘ a? a 
> mi es . i 

a he =i a a a 
: @ @ 

. 
i < — 
- 
{ ; ; 

7 - r 7 

‘ 

' > 
‘ 
‘ 
i + 
t 


- 


7 a Ao a ‘e i 2 a 
WA ue it eee . 
Been 


a 
ak as One 
b gaat J) 


a ye), Nene 


‘ iv fo< 
aed atari en ae gs 
Enith icine 


“om ee Weve Se! eee 


or ‘ fi | 
; . 4 4 7 y 7 
' HP a mks evn Ah 


t 
; We oe s 


a x 


A Aes 6 Naa 
Peas eee Wi aa 


a nN Bot Bee Tyleptrredses | iow tiers © | 
Cw 4 bai? iy iyi uers ¢ iv | 
in ros ee iP, t “ij a; y 

i aa aL a) ee 
ee 
= AO bfinn.« 
wr) ie ee 


Ps 
Ven 7 


ae fa ane oy ph, Tein Ai ba 
BE Dt ee ee if 4c'ly 
ee ee ee ee jada) 
ee* oie ig pls woolen alien) ho WE 
fc seae sae 
he ee on, re wey ne Nd aah uy 
nee ave ce . or ae 
ae 


t 
: Pa rh 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 161 Tue AmERIcAN Museum or Naturau History March 31, 1925 
New York City 


59.9,32(51.7) 


JERBOAS FROM MONGOLIA! 
By Guover M. ALLEN 


A very beautiful series of jerboas from several localities in the Gobi 
Desert was secured by the Asiatic Expeditions of The American Museum 
of Natural History under the leadership of Mr. Roy Chapman Andrews. 
Although the species represented are few, they are of unusual interest, 
for they include a very interesting new Allactaga, the five-toed jerboa, 
and a striking new genus related to Dipus, the three-toed jerboa. The 
series of Allactaga mongolica may be considered as typical and shows, 
on comparison with specimens of the species from Chili, China, that the 
latter is subspecifically distinct and may bear the name annulata given 
many years ago by Milne-Edwards, but latterly placed in synonymy. 


Allactaga mongolica (Radde) 

Dipus jaculus mongolica RappE, 1862, ‘Reisen im Suden von Ost-Sibirien,’ 
I, p. 170, Pl. vr11, figs. 3a-3b. 

A five-toed, long-eared jerboa, buffy gray above, clear white below; 
the tail-tip with a flattened tuft, white at its base, black in its middle 
three-fourths, and white terminally. . 

This characteristic desert species was found in abundance by the 
Expeditions of 1922 and 1923, and a fine series was secured at localities 
in the central Gobi, namely, Turin, Artsa Bogdo, Tsagan Nor, Hurum 
Tu, Gun Burte, Sain Noin Khan, Ussuk, Loh, near Tze-Tsen “Wang, 
twenty miles southwest of Urga, and on the Tola River, eighty miles 
west of that city, as-well as in the vicinity of Erhlien, Sair Usu, and east 
as far as Iren Dabasu. Among the specimens taken during May and 
early June there is a very striking preponderance of males. Thus, of the 
one hundred and seventeen skins taken in May, all but fifteen were males, 
indicating some difference in habits in the earlier part of the season; 
for in August the proportion is just reversed, with only five males to 
fourteen females. The type locality of this species is Tarei Nor, whence 
Radde had fourteen specimens on which his description was based. 


Publications of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
bution No. 42. 


2 AMERICAN MUSEUM NOVITATES [No. 161 


Allactaga mongolica annulata (Milne-Edwards) 

Dipus annulatus MILNE-Epwarps, 1867, Ann. des. Sci. Nat., (5) VII, p. 376. 

Milne-Edwards’ name is currently regarded as a synonym of A. 
mongolica. The original specimens were obtained by Pére A. David 
at or near the southwestern border of the Mongolian plateau and others 
were noted by him at Swenhoafu, Chili Province. Comparison of a skin 
from two hundred kilometers northeast of Shehol in that province reveals 
the fact that the latter is different from true mongolica in its darker 
muzzle and forehead, contrasting sharply with the white cheeks; the 
much darker back, which is nearly solid brown in the specimen; and the 
dark outer sides of the tibize. The black band of the tail-tuft is also 
somewhat shorter than in true mongolica of the Gobi Desert. The skull 
of the Chili animal differs in its narrower interpterygoid fossa and in 
having the small palatal vacuities narrowed to mere slits instead of being 
broadly elliptical. It also averages slightly larger, its greatest length 
being 40-40.5 instead of 37-39 mm., as in the series from the Gobi. 
These differences come out in Milne-Edwards’ plates (‘Recherches,’ 
1868-1874, Pls. x, xa, fig. 3-3f) of Dipus annulatus, so that his name 
may be considered applicable in a subspecific sense to the southeastern 
darker form of this jerboa. The exact type locality is not stated but was 
probably not very far west of Peking. 


Allactaga bullata, new species 


Typr.—Adult male, skin and skull, No. 58723, A. M. N. H., from Tsagan Nor, 
Mongolia. July 5, 1922. Third Asiatic Expedition. 

DescripTion.—Externally similar to A. mongolica but slightly smaller. Skull 
smaller, with more nearly vertical upper incisors and very much larger bull that 
nearly meet at their anterior ends. 

Color and pattern indistinguishable from those of A. mongolica. Dorsal surface 
of head, ears, body, and outer sides of thighs, grayish buff, brighter on the sides of 
the face below the ears and on the back and thighs; grayer on the muzzle, forehead, 
and sides of body. The individual hairs are slaty at base with a short buffy, sub- 
terminal ring, and a black tip, sometimes minute, sometimes half the length of the 
buffy ring. Mixed with these on the back are scattered longer black hairs. An in- 
distinct spot above the eye, paler gray, and at the posterior base of each ear, a spot 
of pure white. Lower side of body, forearms, feet, and a prominent hip-stripe pure 
white to the roots of the hairs. Tail white all around for about 8 mm. at base, then 
pale ochraceous above to the beginning of the terminal tuft which is white for about 
10 mm. at its base, slightly darkened by black hairs among the white; then follows 
the main part of the tuft which is distichous, blackish for about 45 mm. in the median 
line, with short white tip measuring 20 mm. in the median line. The ventral side of 
the tail is white except that the black portion of the tuft is penetrated by a narrow 
extension of the white medially. The hind feet are dark brownish in a narrow line 


1925] JERBOAS FROM MONGOLIA 3 


along the sole, with a larger blackish area beneath the basal phalanges. Median toe 
of hind foot exceeding the lateral toes by about 5mm. Incisors white. - 

Sxutu.—The skull differs notably from that of A. mongolica in its smaller size, 
in having its upper incisors very much less thrown forward, in its more abruptly ex- 
panded braincase behind the interorbital constriction, and most strikingly in the 
very much larger audital bulle. These are about three times the volume of those in 
the larger species and are so closely approximated medially that the basioccipital is one- 
half as broad as in the latter, while their extreme tips anteriorly are nearly in contact. 
The interparietal differs in having its anterior edge bracket-shaped, with the median 
point anteriormost. The cheek teeth show no difference except in size. 

MEASUREMENTS.—The type was measured by the collector.as follows: head and 
body, 105 mm.; tail, 188; hind foot, 70; ear, 40. In the dry skin the hind foot now 
measures 65 mm. The skull measures: occipito-nasal length, 34 mm.; condylobasal 
length, 31; palatal length, 21; diastema, 10.5; incisive foramina, 6; nasals, 12.6; 
interorbital constriction, 10.5; zygomatic width, 24; mastoid width, 22.3; greatest 
diameters of bullz, 10.37.6; upper cheek teeth, 6.5; lower cheek teeth, 6.3. 

So similar is this species to A. mongolica in details of external 
appearance that the two are not distinguishable except by the slightly 
larger size of the latter, as seen especially in its somewhat longer ears. 
The skull, however, with its very much larger bull, is at once strikingly 
different. The new species does not seem to show relationship with any 
of those so far described, and forms a fine addition to the known fauna of 
the Gobi. It is apparently less common than A. mongolica. The series 
at hand is from the following localities in.the central Gobi: Tsagan Nor, 
Sair Usu, fifty miles west and one hundred and sixty miles southeast of 
Sair Usu, and twenty-three miles south of Erhlien. 


Dipus sowerbyi Thomas 

Dipus sowerbyi THomas, 1908, Ann. Mag. Nat. Hist., (8) II, p. 307. 

A jerboa with but three toes on the hind foot, the upper incisors 
orange and grooved; color, buffy above, white below, the flattened tuft 
at the tip of the tail black, with the end white. 

A splendid series of this bright-buffy jerboa was secured from various 
localities in the Gobi Desert, including Iren Dabasu on the east, Erhlien, 
Loh, Turin, Artsa Bogdo to Tsagan Nor on the west. The type locality 
is Yulinfu, in Shensi province, at the southern edge of the Ordos desert, 
but although the only topotype available is much brighter buff or ochra- 
ceous than the average of the Mongolian series, there are occasional 
specimens that match it in every detail, so that for the present these 
latter are best considered D. sowerbyi. It seems likely from a careful 
comparison of descriptions that this should be regarded as a subspecies 
of D. sagitta typical in eastern Russia. According to Thomas its muzzle 
is broader and color slightly brighter, but there seem to be few other 
appreciable differences. 


a AMERICAN MUSEUM NOVITATES [No. 161 


STYLODIPUS, new genus 


Structurally resembling Dipus in external characters, but the hind foot propor- 
tionally shorter, and the flattened tail tuft not so strictly terminal, but beginning at 
about half-way on the length of the tail and gradually increasing in width distally. 
Hind foot with but three toes. Skull in general like that of Dipus, but the bull 
much larger, their mastoid portion produced some 2 mm. behind the occipital bone, 
whereas in Dipus they do not quite reach that level. Palate with a pair of small oval 
foramina as in Dipus, and in addition a second minute pair just back of the last 
molars. Incisors white, the upper pair grooved. The minute upper premolar more 
reduced than in Dipus, a mere spicule, not reaching the crown level of the molars; 
the latter with the reéntrant enamel folds much more nearly of equal depth on the 
inner and the outer sides, resulting in an approximately figure-eight pattern, with 
slight wear. 

The genotype and only referred species is the following. 


Stylodipus andrewsi, new species 


Typr.—Adult male, skin and skull, No. 58549, A. M. N. H., from Ussuk, 
Mongolia. June 22, 1922. Third Asiatic Expedition. 


Descriprion.—Externally like Dipus, but the hind feet proportionally shorter, 
the tail pen-shaped, thick, and with the terminal half distichous, gradually increasing 
in width to the tip (Fig. 1). 

Entire dorsal surface, including head, body, outer side of ears, the arms and 
thighs as far as wrist and ankle, sandy colored or pale buff, evenly darkened by black 
hairs or black-tipped hairs. The hairs of the back are ‘‘deep neutral gray” (Ridgway, 
1912) at base, with a tip of ‘light ochraceous buff’? and a minute black point. An 
indistinct whitish spot is present above the eye and a slightly paler area below it. A 
small tuft of silky white hairs clothes the inside of the ear at its base and there is a 
distinct postauricular white patch. The fore feet, backs of hind feet, a prominent 
stripe running forward across the back of the hip, and the entire ventral surface of 
the body are pure white to the roots of the hairs. The soles of the hind feet are 
“dusky drab,” the long hairs under the toes much shorter and softer than in Dipus 
sowerbyt. The tail is white all around in a narrow ring at the extreme base. Its 
middle third is apparently somewhat thicker than the base or the tip; in color it is 
like the back except that toward the end the dusky tips of the hairs are long, with 
pale bases, giving a smoky tint to the expanded, pen-shaped terminal portion. Be- 
low, the short hairs are very pale buffy along the median line, the terminal half of 
the tail with a buffy lateral border and dusky edge. Just below the tip the whitish 
bases of the long hairs form an indistinct pale area. 

SKULL (Figs. 2, 3).—The great enlargement of the upper portion of the audital 
bullze results in a corresponding compression of the interparietal, so that it becomes 
nearly triangular in outline as compared with the faintly pentagonal one of Dipus 
(sowerbyt). The lacrymal is reduced also as compared with the latter, but the 
breadth of the vertical portion of the zygomatic plate is much the same in both. 
Interesting is the great reduction of the small anterior upper premolar, which is a 
minute spicule reaching about half-way to the crown of the first molar. The re- 
éntrant enamel folds of the upper molars are nearly equal on both inner and outer 
sides, and nearly or quite touch at the center of the tooth, the outer fold, however, 


1925] JERBOAS FROM MONGOLIA 5 


slightly more posterior than the inner. The lower posterior molar lacks a reéntrant 
from the inner side. The incisors are white instead of yellow, the upper ones with a 
median groove. 

MEASUREMENTS.—In proportions the ears, tail, and hind foot seem rather shorter 
than in the species of Allactaga and Dipus found living in the same region. The col- 
lector’s measurements are: head and body, 128 mm.; tail, 150; hind foot, 55; ear, 
16. The skull measures: greatest length, 33 mm.; condylobasal length, 30.5; 


LAE ELE 


Fig. 2. ne Fig. 3. 


Fig. 1. Stylodipus andrewsi, new species. Dorsal aspect of the distichous tail. 
x4/7. 

Fig. 2. Stylodipus andrewsi, new species. Outline of cranium from above. 
X3/2. The type, No. 58549. 

Fig. 3. The same in ventral aspect. X 3/ 2. (Outlines with camera lucida.) 


palatal length, 19; diastema, 8.2; incisive foramina, 6; zygomatic breadth, 22.8; 
interorbital breadth, 10.0; breadth across auditory openings, 22; diameters of bulla 
(ventral aspect), 10.29.8; vertical depth of bulla, 13; width across first upper 
molars, 7.4; upper cheek teeth, 6; length of mandible, 22; lower cheek teeth, 6. 
This very interesting new jerboa is obviously related to Dipus, but 
shows a greater inflation of the auditory region and a consequent change 
in the shape of the surrounding bones. The tendency to reduction of 
the small first upper premolar is carried farther, and there is an approach 
to equality in the development of the enamel folds on inner and outer 


6 AMERICAN MUSEUM NOVITATES [No. 161 


sides of the molars. The shape of the tail is more like that of Pygeret- 
mus, pen-like instead of having merely a terminal tuft or expansion. 

This is a species of the Gobi Desert. The small series obtained is 
from Ussuk, Tsagan Nor, Loh, and near Erhlien. It seems fitting that 
the name of the able leader of the Asiatic Expeditions, Mr. Roy Chapman 
Andrews, should be associated with this handsome jerboa. 


Hay : 


v> | << 


: 


i Ye _ he ct 
seheiha? bits i 


vel Hapa 
a . &\ ; 

¥ » j 
Pi hart ae? Ya 
me aie CREAR Ye ae 


£8 


BT he i aR 


ate ah. of i hes , 
we. er fete : 
ri 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 163 Tur American Museum or NATURAL History April 2, 1925 
New York City ‘ 


59.9,32S(51.7) 
SQUIRRELS COLLECTED BY THE AMERICAN MUSEUM 
ASIATIC EXPEDITIONS? 


By Guover M. ALLEN 


The squirrel-like rodents hitherto brought back by the Asiatic 
Expeditions of The American Museum of Natural History amount to 
some 750 specimens from a wide range of territory, chiefly southwestern 
Yunnan, the southeast coast of China including the island of Hainan, 
and from the region of Peking and central Mongolia. This fine series of 
well-prepared skins has helped to make clearer the distribution and rela- 
tionships of many species. Noteworthy are: the extension of range of 
certain spermophiles described from Kansu and Ala Shan and now found 
by Mr. Roy Chapman Andrews in central Mongolia; the tracing of the 
coniferous-forest types in their southward extension into Mongolia; 
the discovery of two flying squirrels in Hainan where none had pre- 
viously been reported; and the evidences of intergradation between 
various subspecies of the widely distributed red-bellied squirrel (Callo- 
sciurus erythreus). The list of species collected is here given. 


Sciuride 
- Citellus dauricus (Brandt) 
Spermophilus dauricus Branpt, 1848, Bull. Acad. Imp. des Sci., St. Pétersbourg, 
el. phys.-math., II, p. 379. 

A single specimen, taken twenty miles southwest of Urga, is un- 
doubtedly referable to this species. It was first brought to the attention 
of naturalists by Pallas who obtained it near Tarei Nor. Radde in 1856 
secured additional examples and, in his report published in 1862, gave a 
careful account with colored figures of the animal. Like his, the present 
specimen, though taken May 19, is still in winter pelage. Its lips, sides 
of neck, flanks, feet, and entire under side of body are white, and a dull 
white line runs from the muzzle to the base of the ear. The top of the 
nose is pinkish (vinaceous-buff), the rest of the upper side pale buff 
finely lined with black hairs evenly distributed. The tail is short, 59 


1Publications of the Asiatié Expeditions of The American Museum of Natural History. Con- 
tribution No. 43. 


2 AMERICAN MUSEUM NOVITATES [No. 163 


mm., buff above and below, its terminal half above and its terminal 
third below bordered with black and fringed with white. This is evi- 
dently closely related to C. mongolicus, differing only in its slightly more 
pallid coloring, and perhaps in slightly larger skull. The latter should 
therefore stand as a subspecies. 


Citellus dauricus mongolicus (A. Milne-Edwards) 

Spermophilus mongolicus A. Mitnn-Epwarps, 1867, Ann. des Sci. Nat., Zool., 
(5) VII, p. 376. 

The type-locality of this spermophile is “la Mongolie chinoise et 
dans le voisinage de Pékin.”’ Specimens obtained by Mr. Andrews near 
Peking are therefore topotypes, and a careful comparison of these with a 
series from Taboul, 100 miles northwest of Kalgan (on the edge of the 
Mongolian plateau), fails to reveal any tangible difference. Evidently, 
therefore, the name wmbratus, given by Mr. Oldfield Thomas to Taboul 
specimens, must be regarded as a synonym of mongolicus. He compared 
the latter with pale specimens from northwestern Shensi and Ordos desert 
but was misled by supposing the latter were typical of mongolicus. As 
compared with dauricus, this is less pallid and decidedly more pinkish 
above, with a yellowish belly in winter pelage. The summer coat is not 
assumed until fairly late, about the first of July. Specimens from Tsi- 
nanfu are grizzled pinkish buff and black above, washed with pale 
yellow on sides, limbs, and belly. The very short tail, 80 mm. or less, is 

‘distinctive. 


Citellus obscurus (Buechner) : 

Spermophilus obscurus BUECHNER, 1888, ‘Przewalski Exp., Mamm.,’ p. 17. 
The type locality is Kansu, north of Tchagryn-gol, whence Buech- 
ner had two August skins and a third taken in May that still retained 
part of the paler winter pelage. A series taken by Mr. Andrews at 
Artsa Bogdo, Mongolia, in August, 1922, agrees closely with his descrip- 
tion and figure and is provisionally referred to the same species. The 
general color above is an evenly grizzled mixture of dull buffy and black, 
producing a dark sandy appearance; sides and belly dull whitish, washed 
with buffy. It lacks the decidedly ruddy tinge of mongolicus and, even 
in a specimen still partly in winter dress, seems to lack the pinkish-buff 
patch onthe nose of the latter. A distinguishing point noticed by Buech- 
ner is that the black subterminal border of the tail is more or less obscured 
on the ventral side by long rusty hairs. The dark brown bases of the 
_ hairs on back and belly are also more conspicuous, and the basal half of 


1925] ASIATIC SQUIRRELS 3 


the tail is less distichous, more cylindrical. In the dry plains country of 
Shansi this species is represented by a paler, more pinkish form, two 
specimens of which are in the Museum of Comparative Zoélogy. It may 
be described as follows. 


Citellus obscurus siccus, new subspecies 


Typr.—Adult female, skin and skull, No. 19924, Museum of Comparative 
Zodlogy, from ten miles west of Taiyuanfu, Shansi, China. August, 1921. F. R. 
Wulsin. 

DeEscripTIon.—In summer pelage, paler, more vinaceous buff above than C. 
obscurus; slightly paler and with longer tail than C. d. mongolicus. 

Forehead, sides of face, entire back and base of tail “‘ vinaceous-cinnamon”’ evenly 
grizzled with fine black hairs. End of the muzzle clearer, ‘ pinkish-cinnamon”’; 
white eye-ring, indistinctly continued to the end of the nose and to the ear as a whitish 
line. Sides of neck, the fore limbs and feet, the flanks, front of hind leg, and the hind 
feet, pale buffy (“pinkish buff’’). Below, white washed with buffy. On the chin and 
upper throat, fore arms and lower part of hind leg, the hair is practically white to the 
base, but elsewhere, on throat and belly the dark tuscous bases show through every- 
where. The tail is distichous in its terminal half, which is pale rusty (‘‘cinnamon’’) 
in the middle, bordered by black and fringed with buff. The lower side is entirely pale 
rusty, with the subterminal black border partly obscured by rusty hairs. The soles 
of the hind feet are hairy as far as the palmar tubercles. 

The skull is not distinguishable from that of specimens from Artsa Bogdo, Mon- 
golia, referred to C. obscurus except that the nasals equal the intermaxillaries in back- 
ward extension instead of falling short of them. Compared with that of mongolicus 
the skull is a trifling amount smaller. 

MEASUREMENTS.—The type skin is about 210 mm. in length of head and body, 
the tail 76, hind foot 38. The skull measures: greatest length, 45.5 mm.; basal 
length, 41.7; palatal length, 24; diastema, 10; nasals, 16.5; zygomatic width, 29; 
width across molar rows, 13; upper cheek teeth, 10.5; mandible, 29; lower cheek 
teeth, 9.4. 


This is apparently the representative of C. obscurus occurring in 
Shansi, and may at once be distinguished from mongolicus by its longer 
tail and paler, more pinkish, tint. I have seen no specimens of the 
shorter-tailed animal from Shansi, although Thomas has recorded mon- 
golicus from that province, contrasting it, however, with darker Taboul 
specimens. 


Citellus pallidicauda (Satunin) 
Spermophilus pallidicauda Satuntn, 1903, Annuaire Mus. Zool, St. Pétersbourg, 
Vip, ool. 
A rather large.species, at once distinguished by its large feet with 
naked palms and the tail, which is dull ochraceous at the base above, 
paling elsewhere to whitish or pale buffy, and so quite lacking the sub- 


4 AMERICAN MUSEUM NOVITATES [No. 163 


terminal black border of other Chinese species. This was apparently 
common on the Gobi Desert, whence series were secured at Ude and Us- 
suk as well as one each from Loh and Gun Burte. Although referred to 
pallidicauda on geographical grounds (type locality, Lake Chulmu Nor, 
Gobi-Altai), this is unquestionably very closely related to C. alaschani- 
cus Buechner from southern Mongolia, if not identical with it. 


Citellus eversmanni jacutensis (Brandt) 

Spermophilus jacutensis Branpt, 18438, Bull. Acad. Imp. des Sci., St. Péters- 
bourg, cl. phys.-math., II, p. 379. 

A large, long-tailed species with the back thickly speckled with 
whitish. In winter pelage the sides and belly are white with a faint 
buffy wash, and the back is dull buffy, slightly darkened with scattered 
black hairs and marked with small white spots through the close jux- 
taposition of several of the white bands of the longer hairs. In summer 
coat, the back is very much darker, while the forehead, sides of the head 
and neck, the flanks, limbs, and belly are ochraceous. Compared with a 
series from the Altai Mountains representing typical eversmanni, the 
back in the present series from Mongolia is slightly duller, less clear 
black-and-white speckled, while the ochraceous tint of sides and belly is 
more intense. Pallas long ago remarked these differences in Yakutsk 
specimens and Brandt in 1843 gave the provisional name jacuiensis, 
which is probably worth recognition in a subspecific sense. 

This is the common spermophile of the open, wooded country of 
northern Mongolia, where its southern limit is probably conterminous 
with that of the evergreen forest. A large series from Sain Noin Khan in 
the early part of June is just beginning to change from winter to summer 
dress. In one taken June 4, the new and contrastingly ochraceous hair 
of the muzzle, eye, and ear region has just appeared, while in others 
taken June 10 and 11, the new coat is coming in on the shoulders as 
well. A large series taken in mid-July forty-five miles northeast of Urga 
has fully assumed the summer pelage. 


Marmota bobak sibirica (Radde) 

Arctomys bobacvar. sibirica RADDE, 1862, ‘ Reise im Siiden v. Ost-sibirien,’ I, p. 159. 

A pale-buffy marmot, with short dark-brown tail. A large series was 
secured from localities between eighty miles southeast of Urga and forty- 
five miles northeast of the same center; others were taken in the region 
about Tzetsenwan, Mongolia. There is some variation in color, tend- 
ing to a darkening of the buffy hair-tips to brown. Several very small 
young were captured in mid-June. 


1925] ASIATIC SQUIRRELS 5 


Eutamias asiaticus (Gmelin) 

Sciurus asiaticus GMELIN, 1788, Linné’s ‘Syst. Nat.,’ 13th Ed., I, pt. 1, p. 150. 

This small ground squirrel with its five sharply defined, black stripes 
on an olive or khaki-colored ground, and with a pure white belly, seems 
to show very little geographic variation in color. It was met’ with in 
sparsely wooded country near Urga (fifteen and forty-five miles north- 
east) and to the southwest at Sain Noin Khan, but seemingly it does not 
penetrate into the grass-lands or the more desert parts of the Gobi. 
Specimens from the above localities seem identical with others from 
Gichiga on the Sea of Okhotsk; nor am I able to distinguish them 
from EF. a. altaicus, described as being less brightly colored than the 
typical form. 


Eutamias senescens Miller 

Eutamias senescens M1iiEr, 1898, Proc. Acad. Nat. Sci., Philadelphia, p. 330. 

Jacobi is undoubtedly correct in regarding this as a species distinct 
from E. asiaticus, as originally described. It differs in its slightly larger 
hind foot, shaggier, coarser fur and in having the rump rusty instead of 
olivaceous; in addition, the two lateral pairs of black stripes are much 
mixed with ochraceous and merge with the reddish of the rump, whereas 
in £. asiaticus the five black stripes are all clear and distinct, the two 
inner lateral ones continuing to the root of the tail. The belly of £. 
senescens is faintly washed with yellowish, but the throat is contrastingly 
white, while in H. asiaticus the belly is clear white. The skull of the 
former is more slender, the rostrum longer. By these criteria, orientalis 
and albogularis are subspecies of senescens, rather than of asiaticus as 
suggested by Jacobi. 

Four skins from Tungling and the Eastern Tombs in Chili Province 
were secured by Mr. Andrews. 


- Eutamias senescens intercessor Thomas 

Eutamias asiaticus intercessor Tuomas, 1908, Abstr. Proc. Zodl. Soc. London, 
December, p. 44. 

Shansi specimens are a very little paler than the typical form, with 
less dark hairs on the forehead, and with the rusty and ochraceous of 
rump and flanks less intense. Skins from Kweihwating and from forty- 
five miles east of Paotow are referred to this subspecies. 


Tamiops macclellandi swinhoei (A. Milne-Edwards) 


Sciurus macclellandii var. swinhoei A. Mitnn-Epwarps, 1868-1874, ‘Recherches 
Hist. Nat. Mamm.,’ p. 308. 


6 AMERICAN MUSEUM NOVITATES [No. 163 


Of these striped tree-squirrels with white-tufted ears a series of four 
representing this subspecies was taken at Mucheng on the Salween 
drainage, 6500 to 7000 feet altitude. The light lateral stripes are clear 
ochraceous, the top of the head is strongly suffused with the same, the 
belly lightly so. February skins have the three black dorsal stripes well 
defined. This series probably marks nearly the southern limit of the 
subspecies, described originally from Moupin. 


Tamiops macclellandi forresti Thomas 

Tamiops maritimus forresti THomas, 1920, Ann. Mag. Nat. Hist., (9) V, p. 305. 

General color (winter) olive, the median black stripe alone distinct, 
the four other dark stripes faintly rusty, the white stripes washed 
with buffy; cheek stripes and belly white. On the isolated Lichiang 
range this local form has developed, and is well marked through its 
nearly uniform buffy-olive ground color. Four specimens were secured 
at an elevation of 10,000 feet near the summit of Pei Shui. 


* 


, Tamiops macclellandi vestitus Miller 

Tamiops vestitus M1LumR, 1915, Proc. Biol. Soc. Washington, XXVIII, p. 155. 

This is a much more grayish race than any of the others. The 
summer and winter pelages differ slightly. In summer three black dorsal 
stripes are clearly shown, whereas in winter the two lateral of the three 
are brown, and the intermediate area is more buffy. The crown is 
sometimes distinctly russet. This was first described as a species but 
Jacobi is doubtless correct in regarding it as a subspecies of macclellandt. 
Mr. Andrews secured specimens from Tungling and from the vicinity of 
the Eastern Tombs, Chili Province. 


Tamiops macclellandi monticola (Bonhote) 

Sciurus macclellandi monticolus Bonnotsr, 1900, Anni. Mag. Nat. Hist., (7) V, 
p. 52. 

A darker race of the Fukien highlands. Summer skins show five 
prominent dorsal stripes on a grayish-brown ground, the two outermost 
enclosing a white stripe with a decided ochraceous tinge. In winter the 
four lateral black stripes are rusty brown. A series from Yenping in the 
mountains of Fukien comes from very near the type locality. The bright 
buffy facial stripes are conspicuous but cannot be traced continuously 
across the shoulder to the light body-stripe as they can, for example, in 
vestitus. ) 


1925] ASIATIC SQUIRRELS 7 


Tamiops macclellandi maritimus (Bonhote) 

Sciurus macclellandi maritimus Bonnore, 1900, Ann. Mag. Nat. Hist., (7) V, 
Paco; 

A somewhat grayer race with narrower and paler whitish lateral 
stripe than in the preceding race. This is a slightly marked form of the 
low coastal area of Fukien. A small series was secured at Yuki by Rev. 
Harry R. Caldwell. 


Tamiops macclellandi hainanus J. A. Allen 

Tamiops macclellandi hainanus J. A. ALLEN, 1906, Bull. Amer. Mus. Nat. Hist., 
XXII, p. 476. 

Very similar to maritimus but the light side-stripes are more buffy 
and the foot is smaller, more delicate. A fine series was secured by Mr. 
Clifford Pope at Nodoa, on the island of Hainan. He writes that it 
is common there, living in patches of jungle, prickly bamboos, in large 
bushes, or in isolated groups of trees. A favorite place is a mass of 
vines covering a dead tree. It is extremely quick and agile in its move- 
ments and easily alarmed. He never saw it descend to the ground. 


Tamiops barbei (Blyth) 

Sciurus barbei Buyru, 1847, Journ. Asiatic Soc. Bengal, XVI, p. 875. 

This is undoubtedly a species distinct from macclellandi, which it 
replaces in the lower country of Burma and the Malay peninsula. It is 
distinguished: (1) by the more conspicuous outer pale stripe, which is 
broadly continuous across the shoulder with the cheek stripe; (2) by the 
smaller hind foot, usually less than 30 mm.; (8) by the much narrower 
tail; and (4) by the ochraceous instead of pale yellow or whitish under 
surface. The skull is very slightly smaller, the cheek teeth noticeably so, 
with the outer anterior corner of p* less developed so that the anterior 
outline of the tooth is slightly convex instead of concave. A series of 
twelve specimens was taken near Mengting at the Burmese border of 
Yunnan, and two others on the Salween drainage at Mucheng (5000 feet). 
There it must be close to its northern limit in Yunnan, and meets the 
range of 7. macclellandi swinhoei, specimens of which were secured at 
the slightly higher altitude of 7000 feet near the same place. Careful 
comparison with specimens of typical barbed from Tenasserim, kindly 
loaned by the U. 8. National Museum, fails to reveal any important 
difference between the two series, although in the latter the hair of the 
lower surface is shorter and with slightly less conspicuous dark bases. 


8 AMERICAN MUSEUM NOVITATES [No. 163 


Rupestes forresti ‘Thomas 

Rupestes forresti THomas, 1922, Ann. Mag. Nat. Hist., (9) X, p. 399. 

A medium-sized squirrel, grizzled black and ochraceous above, 
merging into nearly clear ochraceous on sides of head, throat and flanks. 
There is a narrow whitish stripe on each side of the body, and the belly is 
pale ochraceous. Of this rare squirrel a single flat skin was obtained at 
Lichiang. The type was from the Mekong-Yangtze divide at 7000- 
9000 feet. 


Funambulus tristriatus (Waterhouse) 
Sciurus tristriatus WATERHOUSE, 1837, Charlesworth’s Mag. Nat. Hist., I, p. 
499. : 
Three skins from the Faunthorpe-Vernay Expedition to India show 
stages in molt from the russet winter pelage to the black-backed summer 
coat. 


Sciurotamias davidanus (A. Milne-Edwards) 
Sciurus davidanus A. M1tNE-Epwarps, 1867, Rev. et Mag. de Zool., (2) XIX, 
p. 196. 

A grizzled black, gray and buff squirrel, with a white eve-ring and 
white streak at the posterior base of the ear. The long hair of the tail is 
white-tipped, the belly white, washed with buffy. The pelage is rather 
coarse, almost shaggy. A series from Tungling and the Eastern Tombs 
is typical and three skins from. He-shuin, Shansi, seem to represent the 
same squirrel, here perhaps near its western limit. 


Sciurotamias davidanus owstoni J. A. Allen 

Sciurotamias owstoni J. A. ALLEN, 1909, Bull. Amer. Mus. Nat. Hist., XX VI, 
p. 428. 

This is a much more richly colored race, in wh’ch the buffy of the 
typical form is replaced by ochraceous, with a heavy wash of the same 
on the lower side. The eye-ring and the postauricular patches, however, 
remain whitish. The type locality is Tai Pei Shan, the mountain range 
of central Shensi. The Third Asiatic Expedition secured specimens 
along the base of these mountains, as well as forty-five miles southwest 
of Fengsiangfu in the same Province. 


Dremomys pyrrhomerus (Thomas) 
Sciurus pyrrhomerus THomas, 1895, Ann. Mag. Nat. Hist., (6) XVI, p. 242. 
A medium-sized squirrel, grizzled buffy and black above, white 
below, and at once distinguished by its ochraceous cheeks, and the bright 


1925] ASIATIC SQUIRRELS 9 


ferruginous thigh-patch and under side of the tail. Its range seems some- 
what circumscribed in the middle Yangtze basin. The type was from 
Ichang and the Third Asiatic Expedition secured & series at Wanhsien, 
in eastern Szechwan, collected by Mr. Walter Granger. 


Dremomys pernyi howelli Thomas 
Dremomys pernyt howelli Tuomas, 1922, Ann. Mag. Nat. Hist., (9) X, p. 401. 
Squirrels of this species are marked by a bright chestnut patch in 
the anal region and by having the backs of the ears ochraceous, contrast- 
ing with the dark olivaceous, black and buff mixture of the back. The 
present subspecies has a rich dark tone with a faintly marked black line 
in the middle area of the back. Two specimens from Taipingpu, Shweli 
River, Yunnan, are referred to it. The type locality is Machangkai, 
twenty-five miles southwest of Tengyueh, Yunnan. 


Dremomys pernyi flavior G. M. Allen 


Dremomys pernyi flavior G. M. ALLEN, 1912, Proc. Biol. Soc. Washington, XXV, 
p. 178. 


This is a smaller race, slightly paler in color. The type is from 
Mengtsz, southeastern Yunnan. Two skins from the Litien and Wie-shu 
Pass, 11,000 feet, and a third from Chungtu, Mekong River, 6000 feet, 
seem to be this race. 


Dremomys pernyi lichiensis Thomas 
Dremomys pernyi lichiensis Tuomas, 1922, Ann. Mag. Nat. Hist., (9) X, p. 403. 
This race of the Lichiang Range is barely distinguishable. Com- 
pared with a series from western Szechwan, representing D. p. griselda, 
it is a very little less gray, the skull possibly smaller. Specimens were 
obtained by Messrs. Andrews and Heller between 8200 and 10,000 feet 
near Lichiang. 


Dremomys pernyi calidior Thomas 
Dremomys pernyi calidior Tuomas, 1916, Ann. Mag. Nat. Hist., (8) XVII, p. 394. 
A darker, ruddier form from the mountains of Fukien, whence a 
single skin is in the collection. 


Callosciurus erythreus gordoni (Anderson) 
Sciurus gordoni ANDERSON, 1871, Proc. Zoél. Soc. London, p. 140. 
As a species, Callosciurus erythreus has a wide range, extending 
from Burma eastward to the coast of central China. Over this great 


10 AMERICAN MUSEUM NOVITATES [No. 163 


territory it is represented by a number of local races. The general color- 
ing is grizzled black and buffy above, with the ventral surface of body 
and limbs sharply contrasted chestnut. In the subspecies gordoni, a 
narrow, sharply defined grizzled line similar in color to the back passes 
medially from the throat down the chest. The long hairs of the tip of 
the tail are black, tipped with chestnut, forming contrasting areas. The 
type locality is Bahmé, eastern Burma. Specimens from southwestern 
Yunnan (Homushu Pass, Taipingpu, Huiyao, Watien) are quite like 
Anderson’s plate and doubtless represent typical gordoni. A series from 
the Salween drainage, however, is exactly intermediate between these 
and the paler mzchianus of the Lichiang range. Above, they are quite as 
dark as gordoni; but, below, they show great variation. Thus, in two 
the grizzled median stripe is absent, while in the others it is mixed with 
ochraceous in varying degree; but on the whole the series is perhaps 
nearer gordoni. 


Callosciurus erythreus pranis Kloss 


Sciurus erythreus pranis Kuioss, 1916, Journ. Bombay Nat. Hist. Soc., XXIV’ 
p. 178. 

The under side is pale ochraceous, the median grizzled line indistinct, 
the tail fringed with pale ochraceous or whitish. Specimens secured by 
the Asiatic Expeditions from the Burma border on the Namting River 
agree exactly with the description of this form. the type locality of which 
is Pran, southwestern Siam. Its range must meet that of gordon? some- 
where in southwestern Yunnan. Two embryos were found in a specimen 
killed March 1,a number correlated with the reduced number of mammz 
(four) in this group of squirrels. 


Callosciurus erythreus michianus (Robinson and Wroughton) 


Sciurus castaneiventris michianus RoBINSON AND WrovuautTon, 1911, Journ. 
Fed. Malay States Mus., IV, p. 234. 

This is a paler race of the Lichiang highlands. Compared with 
gordoni, with which it intergrades, it is much paler, especially along the 
sides, and the tail is yellow-tipped instead of rusty. The mid-ventral 
grizzled stripe is typically lacking, or it may be represented by a con- 
trasted line of ochraceous extending from the throat for a varying dis- 
tance posteriorly. Specimens from Yunnanfu, to the east of Lichiang 
though referred here, are intermediate between this and the smaller- 
footed hemobaphes. 


1925] ASIATIC SQUIRRELS Hi 


Callosciurus erythreus castaneoventris (Gray) 
Sciurus castaneoventris GRAY, 1842, Ann. Mag. Nat. Hist., (1) X, p. 263. 

The type of this race is a skin in the British Museum collected by 
Reeves, but with no other locality than ‘‘China.’’ Bonhote, in his review 
of these squirrels, says that specimens from Fukien are typical. I have 
therefore referred to it a series from Futsing and Yenping in that Prov- 
ince, obtained by the Asiatic Expeditions. In this race the belly lacks 
the median gray line, and the tips of the long fringing hairs of the tail 
are pale ochraceous. The throat is usually gray but in two specimens the 
chestnut of the belly is continued nearly to the lips. A female containing 
two embryos was killed at Futsing on August 2, 1916. 


Callosciurus erythreus ningpoensis (Bonhote) | 

Sciurus castaneoventris ningpoensis Bonuore, 1901, Ann. Mag. Nat. Hist., (7) 
VII, p. 163. 

This is a poorly marked and perhaps untenable subspecies. A series 
of eleven topotypes brought back by Mr. Andrews from Ningpo, Che- 
kiang Province, is barely distinguishable from the Fukien series, but 
averages a little grayer along the sides and slightly paler below. One or 
two of the series are practically identical with castaneoventris from Fukien. 


Callosciurus erythreeus styani (Thomas) 

Sciurus styani Tuomas, 1894, Ann. Mag. Nat. Hist., (6) XIII, p. 363. 

A buffy-bellied race of the lower Yangtze valley. Two skins from 
Tunglu, Chekiang Province, taken in March, are probably best re- 
garded as intermediates between ningpoensis and styani, with the latter 
of which they seem to agree in general coloring. One has a gray mid- 
ventral line, as normally in the more southern representatives of:the 
species, a character usually lacking, however, in the more northern races. 


Callosciurus erythreus insularis (J. A. Allen) 

Sciurus erythreus insularis J. A. ALLEN, 1906, Bull. Amer. Mus. Nat. Hist., 
XXII, p. 473. 

This Hainan race is at once distinguished by its white-fringed tail. 
In the large series obtained at Nodoa and Nam Fong on the island of 
Hainan, there is every gradation from a clear chestnut belly to the con- 
dition in which a sharply defined narrow line of grizzled gray divides 
this area medially. The usual coloring, however, is with gray throat 
and chestnut belly, the latter shade sometimes extending forward nearly 
to the lips. 


12+ AMERICAN MUSEUM NOVITATES [No. 163 


Sciurus vulgaris chiliensis Sowerby 

Sciurus vulgaris chiliensis SowmRBY, 1921, Ann. Mag. Nat. Hist., (9) VII, p. 253. 

This is the representative in eastern China of the tufted-eared 
squirrel of Europe. The type locality is the Tungling area, seventy-five 
miles northeast of Peking. In a series of eleven skins from near there, 
there is much variation in the amount of rufous mixed with the pre- 
dominating black of the sides and upper surface. The cheeks are usually 
blue-gray, the blackish tail with a tinge of brown in the central portion. 
In three specimens the lower side of the forearms and hind leg is clear 
rufous as well as a narrow flank stripe, while in two of these the rufous 
extends to the backs of the hind feet. There is no white eye-ring. 


Sciurus vulgaris (near mantchuricus Thomas) 

Sciurus vulgaris mantchuricus THomas, Ann. Mag. Nat. Hist., (8) IV, p. 501. 

A series of ten skins from localities fifteen miles north and forty- 
five and sixty miles northeast of Urga, Mongolia, seems to approach 
mantchuricus in having the dorsal surface much more mixed with gray 
than in chiliensis which is of a more intense black. The skulls are, as in 
the latter, 54-55 mm. long, hence slightly smaller than in mantchuricus. 
As usual in these melanistic forms of the species, occasional individuals 
show more or less rufous, one for example, having chin, throat and belly 
white bordered by a rufous line, and with rufous feet and inner surfaces 
of legs. The range in Mongolia is doubtless conterminous with that of 
the coniferous forest. 


Ratufa gigantea (MacClelland) 

Sciurus giganteus MacCLeLLaND, 1839, Proc. Zoél. Soc. London, p. 150. 

A giant squirrel, shining black above, light ochraceous below includ- 
ing inner side of legs. This large species reaches the borders of western 
Yunnan. One was secured in 1917 on the Namting River, 1700 feet 
altitude, on the Burma border, and two other skins were bought at Wa- 
tien, Yunnan. 


Ratufa gigantea hainana J. A. Allen 

Ratufa gigantea hainana J. A. ALLEN, 1906, Bull. Amer. Mus. Nat. Hist., XXII, 
p. 472. 

Similar to gigantea, but the belly darker. This is apparently a rare 
species on Hainan. In addition to the type secured nearly twenty 
years ago, a female was obtained by Mr. Clifford Pope at Nam Fong. 
According to the native hunters, it lives only in heavy forest and is 
extremely agile, hence difficult to kill. 


1925] ASIATIC SQUIRRELS 13 


Ratufa indica maxima (Schreber) 
‘Sciurus maximus SCHREBER, 1784, ‘Sdugethiere,’ Pl. coxvu B. 


Ratufa indica centralis Ryley 


Ratufa indica centralis RyLEy, 1913, Journ. Bombay Nat. Hist. Soc., XXII, 
p. 437. 


Skins representing these two races were obtained by the Faunthorpe- 
Vernay Expedition in India, 1923. 


PETAURISTIDZ 
Petaurista yunnanensis (Anderson) 
Pteromys yunnanensis ANDERSON, 1875, Ann. Mag. Nat. Hist., (4) XVI, p. 282. 
General color above maroon evenly ticked with white; membrane 
maroon; feet and adjacent borders of membrane, and terminal three- 
fourths of tail black; under surface of body white, extending out on the 
membrane. Two skins were purchased at Wei-shie, and two specimens 
collected at Taipingpu, 7000 feet, both localities in southwestern Yunnan. 


Petaurista alborufus castaneus Thomas 

Petaurista alborufus castaneus Tuomas, 1923, Ann. Mag. Nat. Hist., (9) XII, 
pe bi2: 

A specimen from the Szechwan-Hupeh border is practically a topo- 
type (type locality, Ichang). This large flying squirrel is very handsome: 
back, except the dull buffy lower half, deep chestnut; the tail chestnut; 
feet black; lower sides of body and membrane rufous; head and throat, 
including a patch on upper surface of shoulder, white. As a species it 
ranges westward to the borders of Tibet and southward into the Yunnan 
highlands. 


Petaurista alborufus ochraspis Thomas 

Petaurista alborufus ochraspis THomas, 1923, Ann. Mag. Nat. Hist., (9) XII, 
ps 172: 

An imperfect skin was purchased at Lichiang, Yunnan, and evi- 
dently represents this race. The buffy area of the back is a little paler 
than in P. a. castaneus and is narrowly continuous with the buffy area 
of the base of the tail; otherwise the two are much alike. 


Petaurista petaurista rufipes, new subspecies 
_ Typr.—Adult male, skin only, No. 58224, American Museum of Natural History, 
from Yungan, Fukien Province, China. September 26, 1921. H.R. Caldwell. 
DeEscription.—Entire dorsal surface of the body, including the backs of the feet, 
the fingers and entire tail, rich “tawny”’ or ferruginous, glossy, the tips of some of the 
hairs of the nape and mid-dorsal area minutely tipped with black which causes a 
slight darkening. Vibrissee, and a narrow eye-ring black; a minute dull brown spot 


14 AMERICAN MUSEUM NOVITATES [No. 163 


at the chin. Entire lower surface of the body pinkish rufous, nearly “ochraceous- 
salmon,” deepening to ‘‘tawny”’ at the border of the membrane. 

MEASUREMENTS.—The skin measures approximately: head and body, 375 mm.; 
tail, 330. The hind foot is 74 mm. 


The typical form of this squirrel occurs in Java. Thomas, in review- 
ing the species in 1908 (under the specific name nitzdus), recognized the 
animal of the Malay peninsula as a distinct subspecies for which Gray’s 
name melanotus was revived. The discovery of the species in the wooded 
mountains of Fukien is apparently an extension northward of its pre- 
viously known range. The two adults secured by Rev. H. R. Caldwell 
agree in lacking all trace of black on the feet, ears, and tip of tail, and in 
these respects are apparently different from the darker forms to the 
southward. A single young, about two-thirds grown, from Kweihwa, 
Fukien, is similar but, as usual, is less bright than the adults, with a 
darkening of black hairs at the sides of the toes and along the outer 
margin of the hind foot; below, the tail shows a narrow median line of 
black. 

Petaurista hainana, new species 


Typr.—Adult female, skin and skull, No. 58200, American Museum of Natural 
History, from Nam Fong, island of Hainan, China. February 19, 1923. Third 
Asiatic Expedition; Clifford Pope, collector. 

Description.—Occiput, nape, and entire dorsal surface of body, including the 
arm to the elbow and leg to the knee, as well as the base of the tail and the inter- 
femoral membrane, a grizzled rusty and black, the individual hairs black-tipped with 
a subterminal ring of ochraceous buff to tawny, and long blackish bases. Forehead, 
sides of head and a broad area behind each ear along the side of the neck, shining 
black. Ears black narrowly bordered with whitish. Lips white, but the chin black. 
Hair of the throat brownish black, white-tipped, this coloring extending in a triangular 
point on the side of each cheek between the eye and the ear, and in a broader area at 
the side of the neck to the upper side of the humerus. Forearm, lower leg, and most 
of the parachute blackish brown, tipped with ferruginous above; the feet, anterior 
and posterior edges of the membrane, forearm below, and the tail all around (except 
upper side of base) deep shining black. Ventral surface of body, humerus, and that 
part of the parachute between elbow and middle of tibia white, the hairs in the median 
region of the body with gray bases. A narrow border at the lateral edge of the para- 
chute consists of black-tipped hairs with broad cinnamon-rufous bases showing 
through. These become wholly black on that portion of the membrane outside the 
forearm, but white-tipped for about 40 mm. at the extreme outer edge just back of 
the long rod that spreads it from the wrist. The interfemoral portion of the mem- 
brane and the lower surface of the tibia are covered with white-tipped hairs, whose 
bases are gray, becoming tawny in the central portion of each patagium. _ 

MrASUREMENTS.— The collector’s measurements of the type are: head and body, 
445 mm.; tail, 595; hind foot, 84; ear, 45. 

The skull measures: greatest length, 75 mm.; condylo-basal length, 68; palatal 
length, 37.5; diastema, 14.5; zygomatic width, 49; interorbital width, 17; mastoid 
width, 36; upper cheek teeth, 17; mandible, 44; lower cheek teeth, 18. 


1925] ASIATIC SQUIRRELS 15 


This is a large and richly colored flying squirrel, seemingly one of 
the oral group, with the tips of the dorsal hairs rusty. In a considerable 
series obtained on Hainan by Mr. Clifford Pope, there is but little varia- 
tion, though one specimen is much more rufous than usual, with a tuft 
of this color behind the ear and with the upper surface of the parachute 
and the tibize completely rufous. The tail also is much mixed with this 
color. In only one specimen in the series of twenty is the lower side of the 
parachute buffy. 

This genus has not previously been reported from Hainan. Mr. 
Pope writes that it was found only in the ‘‘big woods more than fifteen 
miles to the south of Nodoa. There it must be abundant, for the Miao 
hunters shot as many as four in one day. The patches of jungle about 
Nodoa are entirely devoid of them and they do not seem to be found 
even in the wooded mountains eight miles to the west. There is no 
market for them at Nam Fong and their continued existence seems to be 
due to that condition. They ae a very strong odor and perhaps the 
flesh is too strong to be eaten.’ 


Petaurista clarkei Thomas 

Petaurista clarkei THomMas, 1922, Ann. Mag. Nat. Hist., (9) X, p. 396. 

A large, grizzled gray-and-buff species with ochraceous feet and 
lower surfaces, as well as a spot of the same behind each ear. Two im- 
perfect skins were bought at Lichiang and Talifu, Yunnan, extending 
the known range southward from northern Yunnan, Mekong valley 


Petaurista philippensis (Elliot) 

Pteromys philippensis Exutot, 1839, Madras Journ. Lit. and Sci., X, p. 217. 

A skin was brought back from India by the Faunthorpe-Vernay 
Expedition. 

Petaurista melanopterus (Milne-Edwards) 

Pteromys melanopterus MitNn-Epwarps, 1867, Ann. des Sci. Nat., Zool., (5) 
VIII, p. 375. 

A long-haired gray species, with black-edged hind feet and a horder 
of ochraceous along both surfaces of the parachute. This flying squirrel 
was obtained in Chili Province at Tungling and the Eastern Tombs. 


Pteromys (Hylopetes) alboniger (Hodgson) 

Sciuropterus alboniger Hopason, 1836, Journ. Asiatic Soc. Bengal, V, p. 231. 

A medium-sized, gray-and-buff flying squirrel, with dark feet, white 
hind toes, and whitish belly with the bases of the hairs everywhere 
pale gray. A single skin from Lichiang, Yunnan, is apparently this 
species; and Thomas has recorded it from the same range at 11,000 feet. 


16 AMERICAN MUSEUM NOVITATES [No. 163 


Pteromys (Petinomys) electilis, new species 


Tyrr.—Adult female, skin and skull, No. 58177, American Museum of Natural 
History, from Nam Fong, island of Hainan, China. April, 1923. Third Asiatic 
Expedition; Clifford Pope, collector. 

DESCRIPTION.—A medium-sized species with pale russet back grading into 
fuscous on the upper part of the membrane. Tail tapering in width from the basal 
third to the dark tip. 

Dorsal surface from the nose to the base of the tail uniform pale cinnamon, the 
basal four-fifths of the hairs fuscous. On the limbs and flanks the tips of the hairs 
are pale (grayish or whitish), becoming obsolete on the membrane which is blackish 
brown distally with a narrow white edging, most prominent along the posterior half 
and passing more to the ventral side in the anterior half. Backs of the feet covered 
with short hairs, whitish and fuscous, the latter predominating on the hind feet; 
terminal half or more of hind toes white. Side of the head from eye to ear dusky, 
below which the white of the lips is continued backward and upward, forming a white 
streak behind the ear on the side of the upper neck. A narrow dusky ring surrounds 
the eye. Below, the chin, throat and upper arm are pure white to the bases of the 
hairs; elsewhere the hairs have slaty bases, those of the membrane paler. Along the 
flanks a wash of cinnamon extends from axilla to knee. The tail is distichous, broad- 
est at about its basal third, whence it tapers regularly to the tip. It is slightly darker 
than the back, pale cinnamon washed with dusky, the latter tint deepening distally 
to produce a distinctly dark tip. Ears naked. 

Individuals vary in the relative amount of dusky and cinnamon. Some have the 
parachute membrane rusty instead of fuscous, and the tail may be less cinnamon. In 
some the pure white areas of throat and axilla may be continuous and even extend as 
a narrow line down the middle of the chest, or there may be a pure white area at the 
groin. Immature examples are grayer above than adults through the prevalence of 
white-tipped hair. 

Sxutu.—The skull shows the short rostrum and low uninflated bulle typical of 
the subgenus. The cheek teeth are roughened on their crowns and show a small 
lateral cusp between the two main cusps of the molars. 

MEASUREMENTS.—The dimensions of the type as noted by the collector are: 
head and body, 172 mm.; tail, 159; hind foot, 35; ear, 26. The skull measures: 
greatest length, 41.5; basal length, 35; palatal length, 21.7; diastema, 8.7; nasals, 
11; zygomatic width, 25; mastoid width, 18.7; upper cheek teeth, 8; mandible, 24; 
lower cheek teeth, 7.6. 

The discovery of this small flying squirrel on Hainan is an interest- 
ing extension of the range of the subgenus into southeastern China, but 
it seems specifically distinct from any of the other forms described from 
India and southeastern Asia, although agreeing well subgenerically with 
Petinomys. Mr. Pope, who secured nearly forty specimens, writes that 
many were obtained by the native Miao hunters while searching for the 
large Petaurista in the “big woods,”’ but they appear not to inhabit the 
jungle areas north of the central mountains. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 167 Tur American Museum or Narurau History April 22, 1925 
New York City ? 


59,7,55H (51.2) 


A NEW HOMALOPTERIN LOACH FROM FUKIEN! 
By J. IT. NicHo.s 


Regan, 1911, proposed the genus Hemimyzon for Homaloptera formo- 
sana Boulenger, 1894, Formosa, a species more or less intermediate in 
form between Homaloptera and Gastromyzon. We have a similar un- 
described species from the province of Fukien, China, which is here 
placed in Hemimyzon. Certain differences which it shows from H. 
formosana, namely ventral rays 9 to 11 (versus 15) and caudal obliquely 
truncate (versus forked), seem to require that it be subgenerically 
distinguished as Pseudogastromyzon, new subgenus. 


Hemimyzon zebroidus, new species 


Body depressed, disc-shaped anteriorly, flattened beneath, compressed behind; 
1.5 times as broad as high. Rostral membrane crenulate; edge of the lower jaw 
narrow and firm; two pairs of minute inferior barbels on the snout, at the corners of 
and towards the center of the rostral membrane, each of the latter pair in a notch in 
its border; a pair of slightly larger barbels at the corners of the mouth. Width of 
head equal to its length; head in length of pectoral, 1.6. Origin of the ventral slightly 
in advance of that of the dorsal; pectorals subhorizontal and ventrals in a horizontal 
plane, second to fourth outer rays of ventral and second to eighth of pectoral bifid; 
anal well developed reaching lower caudal base; ventrals appreciably shorter than 
pectorals, pointed, their tips passing the vent; dorsal origin equidistant from snout 
and anal axil; caudal obliquely truncate. Pectoral with about 20 rays, ventral with 
9 to 11. 

DESCRIPTION OF Typr.—No. 8392, American Museum of Natural History, col- 
lected near Yenping, Fukien, by H. R. Caldwell, co-operating with the Third Asiatic 
Expedition of The American Museum of Natural History. 

Length to base of caudal 63 mm. Depth in length 6; head 4.4; pectoral in 
length 2.7; ventral 3.5; width between pectoral axils 5. Eye in head 5; snout 1.7; 
interorbital 2; width of gill-cleft 3.5; depth of peduncle 2; its length 2.5; longest 
dorsal ray 1.3; height of anal 1.4; caudal 1.1. Width of mouth in snout 2.5; mouth 
to snout 3. 

Dorsal 914; anal 8; pectoral 21; ventral 10. Scales about 90. 

Head below, breast, and belly flat; pectorals and ventrals expanded to resemble 
the condition in Gastromyzon, free ends of pectorals overlapping ventrals and appressed 
to sides in the same manner, but ventrals well separated, pointed behind; head 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 44. 


2 AMERICAN MUSEUM NOVITATES [No. 167 


Fig. 1. Hemimyzon zebroidus, type. 


depressed, the profile sloping; tail compressed; snout from above broad subtruncate. 
slightly rounded; vent appreciably nearer origin of anal than axil of ventral. Inter- 
orbital flat; eye with a free rim; pectoral origin under center of eye; thence a narrow 
membranous ridge borders the flattened lower surface of the head forward to the 
sides of the snout; free edge of opercle well curved; mouth inferior, semicircular, 
transverse; lips full, membranous, smooth, the upper overhanging the mouth, and in 
turn overhung by a snout membrane with crenulate edge; a small barbel at the end 
of the maxillary; small, scattered horny tubercles on the sides and tip of the snout. 
Dorsal origin equidistant from tip of snout and anal axil; slightly behind ventral 
origin; anal reaches to caudal base; more than the posterior half of the pectoral 
free; caudal obliquely truncate. Head, breast and belly to axils of ventrals scaleless; 
a large membranous ventral axillary flap with a rounded end; lateral line complete, 
in the middle of side, straight except for a slight double flexure behind the head. 

Dark grayish brown; belly pinkish; dorsal with a black tip and imperfect 
crossbars; caudal with about four blackish bars; faint bars on pectoral and ventral; 
narrow, pale, somewhat oblique bars on the flanks. Smaller specimens are somewhat 
more sharply marked. 


REFERENCES 


BouLENGER, 1894, Ann. and Mag. Nat. Hist., XIV, p. 463. 
Reaan, 1911, Ann. and Mag. Nat. Hist., VIII, p. 32. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 169 Tue American Museum or NaTuRAL History May 23, 1925 
New York City 


59,7.55M (51) 
AN ANALYSIS OF CHINESE LOACHES OF THE 
GENUS MISGURNUS! 


By J. T. NicHoLs 


Chinese loaches of the genus Misgurnus are exceedingly variable. 
A number of species have been described in the past, but recent authors 
are inclined to look upon their differences as individual variation, and to 
assign all or almost all to Misgurnus anguillicaudatus (Cantor), type 
locality Chusan, a coastal island south of the Yang-tze River. Berg, 
1916, makes this a race of European Misgurnus fossilis. The variations 
are so great, however, that it is difficult to believe that more than one 
species is not involved. Thus, Jordan and Snyder, 1906, have revived VM. 
decemcirrosus Basilewski, 1855, which name is also used by Fowler, 
1924, for a fish which the present writer refers to M. mizolepis Giinther, 
1888. 

In considerable Misgurnus material on hand from different parts of 
China, several distinct forms are apparently recognizable, complicated by 
individual variation so as to appear to be races rather than species. 
However, both in Anhwei and Fukien two such forms occur together, and 
in Tungting Lake of the Yang-tze River, Hunan, three forms. Nomen- 
clature may best be adapted to the situation by recognizing three species, 
and two or three races of each. Unfortunately, most of the species of 
early authors are not described with sufficient accuracy for their names to 
be available. For the present at least, decemcirrosus Basilewski is un- 
certain from available data; anqguillicaudatus Cantor is pretty definitely 
determinable by reason of its definite type locality; mizolepis Giinther is 
clear enough. 

. Referring one of the three species here recognized to anguillicaudatus 
and a second to mizolepis, the third is represented in our collections by a 
plain-colored form from Yunnan and a heavily-spotted form from Tung- 
ting, to which, among our material, Berg’s, 1916, figure of Misgurnus 
fossilis anguillicaudatus from the north Corean boundary, which may be 
taken as representative of Misgurnus mohoity (Dybowski), 1869, from 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 45. 


2 AMERICAN MUSEUM NOVITATES [No. 169 


the Onon and Ingoda rivers, east of Lake Baikal, seems to be closest. 
Without having seen material of this, its typical subspecies, our third 
Chinese species is provisionally referred to mohoity. 

Consistency with recognition of three species of Misgurnus in China — 
would seem to dictate consideration of European WM. fossilis as a distinct: 
species rather than as a race of one of these. 

The following analysis of Misgurnus in China is published with 
some uncertainty at this time, although our material is probably more 
extensive than that of earlier writers. It may be considered as a pre- 
liminary treatment of the subject, a basis for discussion and for further 
collecting now in progress, in the light of which it will perhaps be advan- 
tageously revised at a later date. 


SPECIES AND Races or Misgurnus 


Misgurnus fossilis (Linnzus) 
Cobitis fossilis LINNxUS, 1758, ‘Syst. Nat.,’ 10th Ed., I, p. 303. Europe. 
Misgurnus fossilis Brera, 1916, ‘Poiss. Eaux Douces Russ.,’ p. 353, Fig. 
Differs from Asiatic forms in being marked with bold, lengthwise, 
dark stripes. Europe. 


Misgurnus anguillicaudatus (Cantor) 
Cobitis anguillicaudata CANTOR, 1842, Ann. Mag. Nat. Hist., IX, p. 485. Chusan. 
Striz on scales little radiating; the scales more or less embedded or 
fully exposed. Head large, less than 6 in the length to base of caudal. 
Slender (depth between 7 and 8). Caudal long, equal to or slightly 
greater than head. Markings not very bold and sharp. East central 
China. 


Misgurnus anguillicaudatus anguillicaudatus (Cantor) 


This race occurs to the eastward, south of the mouth of the Yang-tze, 
Anhwei south into Fukien. Its limits to the north are uncertain. 


Description of a specimen from Ningkwo, Anhwei, September 15 to October 15, 
1921, C. H. Pope.—Length to base of caudal 130 mm. Depth in length 7.5; head 5.8. 
Snout in head 2.7; width of body 1.8; depth of peduncle 1.6; its length 1.3; 
pectoral 1.4; ventral 1.8; longest dorsal ray 1.5; longest anal ray 2; caudal 0.9. 
Eye in snout 2; interorbital 1.7; maxillary 1.5; anterior barbel (the longest) 1. 

Dorsal 9; anal 7%. Scales about 155. 

Elongate and a little compressed; vent at 4 the distance from anal origin to 
ventral axil. Snout narrow and blunt; interorbital slightly convex; mouth inferior, 
horizontal, horse-shoe-shaped, with thick lips; 3 subequal barbels above mouth, of 
which the anterior is slightly the longest on one side and middle one on the other; 


1925] CHINESE LOACHES, GENUS MISGURNUS 3 


lower lip cleft, with one moderate (posterior) and one short barbel. Dorsal origin 
equidistant from base of caudal and gill-cleft; ventral origin slightly behind that of 
dorsal; pectoral reaching % to ventral; ventral 4; to anal; caudal narrow, rounded 
at the end. Scales rather irregular, more or less embedded in thick skin; with numer- 
ous slightly radiating striz. 

Dark above, paler below, with scattered faint dark marks of irregular size and 
placement, some along back as large as eye; caudal lightly barred; dorsal faintly 
marked; lower fins pale. 


Misgurnus anguillicaudatus tungting, new subspecies 


A race from the central Yang-tze, dark above, pale below, without 
large or pronounced markings on the body. 


DESCRIPTION OF THE Type.—Number 8393, American Museum of Natural 
History. Huping, Tungting Lake, Hunan, C. H. Pope. Length to base of caudal 
89 mm. Depth in length 7.7; head 5.6. Snout in head 2.6; width of body 2; pec- 
toral 1.4; ventral 1.6; longest dorsal ray 1.5; longest anal ray 1.7: caudal 1; depth 
of peduncle 2; its length 1.2. Eye in snout 2; maxillary 1.6; interorbital 2.4; 
posterior upper barbel 1.4. 

Dorsal 9; anal 8. Scales 143. 

A little compressed; vent at 4 the distance from anal origin to ventral axil; 
snout pointed; interorbital a little convex; eye somewhat superolateral; mouth in- 
ferior horizontal semi-circular; 3 barbels above it, the posterior the longest and 
central one the shortest, posterior 2 closest together; 2 small barbels below, the 
posterior the longest. Dorsal origin equidistant from base of caudal and center of 
opercle; ventral origin just appreciably behind that of dorsal; pectoral pointed, 
reaching 14 the distance to ventral; ventral }4 to anal, approximately in a horizontal 
plane; caudal narrow, somewhat rounded behind. Scales small, regular, exposed, 
with close-spaced only slightly radiating striz. 

Dark above and on upper sides; pale, unmarked below; the dark color in fine 
obscure freckling on sides, with slight tendency to concentrate into a broad dark 
lateral band anteriorly; a short dark bar at upper caudal base; caudal with fine 
blackish irregular bars; dorsal a little spotted, and lower fins slightly marked with 
dusky. 


Misgurnus anguillicaudatus erikssoni Rendahl 
Misgurnus erikssoni RENDABL, 1922, Ark. Zool., XV, No. 4, p. 3. Mongolia. 


This race has a very long peduncle, depth of same 2.9 to 2.5 in its 
length. 


Misgurnus mizolepis Giinther 

Misgurnus mizolepis GUNTHER, 1888, Ann. Mag. Nat. Hist., I, p. 434. WKiu-kiang 
on the Yang-tze River. 

Striz on scales little radiating, scales embedded, peduncular keels 
always well developed, sometimes excessively so. Markings sharply 
contrasted, usually small. Yang-tze River valley and southward along 
the coast. 


4 AMERICAN MUSEUM NOVITATES [No. 169 


Misgurnus mizolepis mizolepis Giinther 


Deep-bodied (depth 5 to 6), the peduncular keels excessively devel- 
oped; markings few, small; caudal short. Reaches a large size. Yang-tze 
Valley. 


Description of a specimen from Tungting Lake, Hunan, January 3, 1921, C. H. 
Pope.—Length to base of caudal 167 mm. Depth in length 5.5; head 5.5. Eye in 
head 7.5; snout 2.5; interorbital 4; maxillary 3.5; depth of peduncle 1.3; dorsal 
height 1.5; anal 1.9; pectoral 1.3; ventral 1.7. Ventral under last third of dorsal, 
not reaching anal. Rudimentary rays of caudal above and below precurrent as 
fleshy keels. 

Dorsal 7; anal 7, Scales about 135. 

Color dark, fins and belly paler, brownish, lightly speckled. 


Misgurnus mizolepis fukien, new subspecies 


More elongate, sides well covered with scattered small black 
marks. Fukien. 


DESCRIPTION OF TypE.—Number 8394, American Museum of Natural History, 
Yenping, Fukien, H. R. Caldwell. Length to base of caudal 127 mm. Depth in 
length 7.1; head 6.6. Snout in head 2.6; width of body 1.5; least depth of peduncle 
(with keel) 1.4; its length 0.8; pectoral 1; ventral 1.6; longest dorsal ray 1.4; long- 
est anal ray 1.4; caudal 1. Eye in snout 2.5; interorbital 1.6; maxillary 1.4; pos- 
terior barbel 0.9. 

Dorsal 8; anal 7. Scales about 140. 

Compressed ; vent at }4 the distance to ventral axil from anal. Snout broad and 
blunt; interorbital broadly convex; eye slightly superolateral; mouth inferior, 
horizontal, horse-shoe-shaped, with thick lips, the lower cleft; 3 barbels above (only 
2 on one side) the posterior decidedly the longest, 2 small ones below, the posterior 
the larger.' Dorsal origin equidistant from base of caudal and gill-cleft or slightly 
nearer the latter; ventral origin slightly behind that of dorsal; pectoral pointed; 
extending slightly more than }4 the distance to ventral; ventral slightly more than ve 
to anal; caudal rounded, the precurrent keels prominent and fleshy. Scales irregular, , 
more or less embedded in thick skin with numerous close-set, sub-parallel, slightly 
radiating striz. 

Dark above to about mid-line of side, paler below. Sides with evenly scattered 
fine black spots. Caudal speckled; dorsal faintly barred; lower fins plain. Upper 
basal caudal black spot, squarish, distinct. 


Misgurnus mizolepis hainan Nichols and Pope 
Misgurnus mizolepis hainan NicHoLs AND Pops, (in press), Bull. Amer. Mus. 
Nat. Hist. 
An elongate form approaching typical anguillicaudatus, with bold 
black markings of irregular size. Hainan Island. 


1925] CHINESE LOACHES, GENUS MISGURNUS 5 


Misgurnus mizolepis grangeri, new subspecies 


DeEscrIPTION OF TyPpE.—Number 8395, American Museum of Natural History, 
Yen-ching-kao, Szechwan, November, 1921, W. Granger. Length to base of caudal 
117mm. Depth in length 6.8; head 5.9. Snout in head 3; interorbital 4.5; width of 
body 1.8; depth of peduncle 1.6; its length 1.2; pectoral 1.4; ventral 2; longest 
dorsal ray 1.7; longest anal ray 1.8; caudal 1. Eye in snout 2.6; longest barbel 2. 

Dorsal 9; anal 7. Scales about 140. 

Little compressed except behind the dorsal; peduncular keels strongly developed, 
thick, the lower extending forward almost to anal axil, the upper stopping short of 
dorsal axil by a considerable space. Mouth inferior, horse-shoe-shaped, surrounded 
by 10 barbels as usual; eye small, deep beneath the skin. Dorsal origin equidistant 
from base of caudal and edge of preopercle; ventral origin appreciably behind that of 
- dorsal; pectoral reaching 34; the distance to ventral; ventral }4 the distance to anal; 
caudal pointed. Scales imbedded in thick skin, with conspicuous close-spaced sub- 
parallel striz, the strie above radiating somewhat from those below. 

Dark gray above, paler, pinkish on the belly; sides with small, irregular scattered 
black spots. Dorsal faintly marked; caudal with a few irregular blackish bars and 
spots, no especially distinct one on its upper base. 


Misgurnus mohoity (Dybowski) 


Cobitis fossilis var. mohoity DyBowskt, 1869, Verh. Zool.-Bot. Gesell., Wien, 
XIX, p. 957. Onon and Ingoda rivers, east of Lake Baikal. 

Cobitis fossilis anguillicaudatus BeRG, 1916, ‘Poiss. Eaux Douces Russ.,’ p. 354, 
Fig. Lake tributary to Tumen-Ula River (north Corean boundary). 


Eastern Asia north of the Gobi. The following two forms from 
western China east to the central Yang-tze are provisionally considered 
conspecific with Misgurnus mohoity mohoity (Dybowski), which differs 
from them in being more slender, dorsal origin nearer base of caudal 
than gill-cleft, rather small dark marks evenly scattered on sides. 


Misgurnus mohoity yunnan, new subspecies 


DESCRIPTION OF TypE.—Number 8396, American Museum of Natural History, 
Yunnanfu, Yunnan, October 20, 1920, John Graham. Length to base of caudal 123 
mm. Depth in length 6.6; head 5.4. Eye in head 6; snout 2.7; interorbital 5; 
maxillary 4; width of mouth 44; posterior maxillary barbel 2.5; width of body 2; 
depth of peduncle 1.7 ; its length 1; pectoral 1.5; ventral 2.3; longest dorsal ray 
1.5; longest anal ray 2; caudal 1. 

Dorsal 9; anal 7. Scales about 130. 

. Elongate; moderately compressed; the back and belly rounded; vent at 4 the 
distance from anal origin to ventral axil. Interorbital convex; snout somewhat 
pointed and compressed; mouth inferior, horse-shoe-shaped; a barbel at the side of 
the tip of the snout, and 2 close together near the end of the maxillary, the posterior 
the longer; upper lip loose, lower expanded, free behind, divided in the center; with 
two small barbels, the outer the longer. Dorsal origin equidistant from base of 
caudal and gill-cleft; ventral placed slightly before the center of dorsal base; pectoral 


6 AMERICAN MUSEUM NOVITATES [No. 169 


‘extending }4 the distance to ventral; ventral }4 to anal; caudal blyntly pointed, its 
rudimentary rays precurrent in low keels. Scales small, firm, with conspicuous close- 
spaced radiating strie. 

Olive, finely freckled with pale above; pale below. A small oblique black bar 
on caudal base near the top; caudal faintly and irregularly barred with dark; dorsal 
with a few fainter bars; other fins plain. 


Misgurnus mohoity leopardus, new subspecies 


DEscrIPTION OF TypE.—Number 8397, American Museum of Natural History, 
Tungting Lake, Hunan, C. H. Pope. Length to base of caudal 105 mm. Depth in 
length 6.5; head 5.6. Snout in head 2.7; width of body 1.6; depth of peduncle (in- 
cluding keels) 1.6; its length 1.2; pectoral 1.2; ventral 1.6; longest dorsal ray 1.6; 
longest anal ray 1.7; caudal 1.2. Eye in snout 2; interorbital 2.2; maxillary 1.9; 
posterior barbel 1. 

Dorsal 8; anal 8. Scales about 135. 

Scarcely compressed; vent almost immediately before anal origin, at not more 
than the distance to ventral axil. Snout rather pointed, narrow; interorbital 
slightly convex; the eye slightly superolateral; mouth small, horizontal, inferior, 
semi-circular; 3 barbels above it of almost equal length and equally spaced, two small 
ones below, the posterior decidedly the longer; *lower lip deeply cleft. Dorsal origin 
equidistant from base of caudal and margin of opercle; ventral one slightly behind 
that of dorsal; pectoral reaching 4 distance to ventral; ventral 74 to anal; caudal 
rounded or slightly pointed, with conspicuous precurrent keels, not fleshy. Scales 
small, rather regular, exposed; with comparatively few radiating striz. 

Back and sides with contrasting dusky spots, rather regular, and increasing in 
size upward to diameter of eye, the largest along back; caudal finely and irregularly 
barred with blackish; a black triangular spot at its upper base; dorsal spotted; 
lower fins with slight, faint marking. 


Key TO THE E1gHT CHINESE Races oF Misgurnus 


1.—Skin not thickened, scales rather regular and fully exposed................ 2 
Skin more or less thickened and scales more or less embedded.............. 4, 
2.—Strie on scales well radiating. Scales moderate (about 130); depth moderate 
(about6.5); head large (about 5.5); peduncle moderate (about equal to head) ; 
dorsal origin equidistant from base of caudaland gill-cleft. Misgurnus mohoity. 
Yunnan; TPonetmes. {ee SP ea OR a Se ee 3. 
Strize on aoales little radiating. Scales moderate (about 145); elongate (depth over 
7.5); head large (about 5.5); peduncle moderate (slightly less than head); 
dorsal origin equidistant from base of caudal and middle of opercle. Sides 
finely marked with dark, tending to form a broad band anteriorly. 

Misgurnus anguillicaudatus tungting. Tungting. 
3.—Color, freckled olive above, pale below, without dark marking on body. 
Compressed, width of body 2 in head..Misgurnus mohoity yunnan. Yunnan. 
Color, back and sides with contrasting dark spots, rather regular and increasing 
in size upward to the diameter of eye along the back. Very little compressed 
CuaelthaeG) yee ae etek Misgurnus mohoity leopardus. Tungting. 


1925] CHINESE LOACHES, GENUS MISGURNUS 7 


4.—Moderate or elongate, peduncular keels about precurrent caudal rays sometimes 

Heshy, not excessively developed: 9.5 S 2 aecw.n Siew ce eter ewes sey sleeve 5 

Deep (depth less than 6). Skin very thick, peduncular keels greatly developed, 
fleshy. Black spot at upper caudal base, faint or wanting. 

Misgurnus mizolepis mizolepis. Yang-tze Valley. 

5.—Head large (less than 6). Dusky spotting on sides vague and irregular, the 

arrangement and size of spots variable; peduncle short, less than head; 

compressed (width about 2); scales fine and irregular. 

Misgurnus anguillicaudatus anguillicaudatus. Anhwei, Fukien, Chusan. 

Head small (6 or more). Dark markings on side contrasted; peduncle long 

(slightly longer than head), or else little compressed (width of body less 

than 2), black spot at upper caudal base faint or wanting. Muisgurnus 


mizolepis. Hainan, Fukien, Szechwan, Tungting...................... 6. 
6.—Spots on sides irregular, varying in size, some as large as eye. Canmnesed 
Keyl ili) Se vies opeum ape psa ti cael cactsiewencene Misgurnus mizolepis hainan. Hainan. 


Spots on sides small, blackish. Very little compressed (width less than 2)..... ies 
7._Spots on sides irregular, that on upper caudal base ill-defined or absent. 

Misgurnus mizolepis grangeri. Szechwan. 

Spots on sides fine, regular, blackish, that on upper caudal base sharply marked. 

Misgurnus mizolepis fukien. Fukien. 


REFERENCES 


BAsILEWSKI, 1855, Nouv. Mem. Moscou, X, p. 239, Pl. vm, fig. 2. 

Bera, 1916, ‘Poiss. Eaux Douces Russ.,’ p. 354, Fig. 

Cantor, 1842, Ann. Mag. Nat. Hist., IX, p. 485. 

DysowskI, 1869, Verh. Z.-B. Gesell., Wien XIX, p. 957. 

Fow ter, 1924, Bull. Amer. Mus. Nat. Hist., L, p. 395, Fig. 3. 

GinTuHER, 1888, Ann. Mag. Nat. Hist., I, p. 484. 

JORDAN AND SNYDER, 1906, Proc. U. S. Nat. Mus., XXX, p. 833. 

. Linnzus, 1758, ‘Syst. Nat.,’ 10th Ed., I, p. 303. 

NICHOLS AND Pops (in press), ‘Fishes of Hainan,’ Bull. Amer. Mus. Nat. Hist. 
RENDABL, 1922, Ark. Zool., XV, No. 4, p. 3. 


be A an aN : 
Pee 
i il das 

* mek tidy 
a 7 ; (ge hoe Wits ae - ufo (i "4 ide, ep 
\ Mbpaiatce it Vath ovis obseanrt isan Sint data tataht ia re dagen 


“sh | 


me ders hess: Coes pe | ; ne 
- A Y Ss . al ve Gee rity mit A Li io Bh! 
raed ¥ | P i? 2 j wn 45 Win, a were iy ae : “ yey 
. ea) ; 1 wert at “a als iy ope 
d . Fin . CS Ya ane ae 
oye ~*~ P Ms mi 
at n m Kigek zn ri wil rt peat 
i { . _ 
* ‘ ks “nes i Me 
| | BG ULES rape AB yh eneben l ae y 
if \ 
iu it i 4 + ek 
* >? wel nh eit 
} a ae . \ 7, oe) ¢ a 
‘) igh 1 ae ifr a Pe 4 ( hoht ee 
a é 
t yn ib 
, ' a 
Hh) cols Ae, i) seh “inn, 
wares ty A ; 
| i 
: in i re, i ony 
Ny ft ie 
’ a ue e i ‘ 
a4 i , rf ; w a 
} eis, ig 
¥ Sa? Meola es 
. (naoe : 
- e o 24 P I he ‘ ‘ 
r 
> vid geen 


. 
vi 

mM 

—s 

Lt - ao 

a? « 

i 

As ‘ 

| i Kaa 

fy 

- 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 170 Tue AMERICAN Museum or NaTurRAL History May 25, 1925 
New York City 


59,7.55C(51) 


THE TWO CHINESE LOACHES OF THE GENUS COBITIS! 
By J. T. NIicHoLs 


Berg, 1916, recognizes a Chinese Cobitis, Cobitis sinensis Sauvage 
and Dabry de Thiersant, 1874, as subspecifically distinct from C. tenia 
which he credits to northern Asia as well as to Europe. Fowler, 1924, 
synonymizes it with tenia; but Nichols and Pope (in press), recognize 
this and another race of C. tenia from China. 

Kreyenberg and Pappenheim, 1908, identify all their Yangtze | 
material with Cobrtis [Lepidocephalichthys| macrostigma Dabry de 
Thiersant, 1872, from lakes of central China. However, Mr. Clifford H. 
Pope’s collections from Tungting Lake, Hunan, adjacent to the Yangtze, 
contain two distinct loaches. One of these is apparently a race of C. 
tenia and is here identified as sinensis; the other is clearly referable to 
macrostigma. A drawing by Miss Olive Otis of a specimen of each from 
Tungting of approximately the same size (sinensis 110 mm., macrostigma 
115 mm., to base of caudal) is here reproduced to show the differences. 
C. macrostigma has a longer peduncle; smaller scales, about 135 versus 
100 to 110 before the dorsal; dorsal origin midway between end of snout 
and middle of peduncle, versus base of caudal. There seems to be a 
difference in the lips, the free lip between maxillary and mental barbels 
pointed versus broad, squarish. 


Cobitis tenia sinensis Sauvage and Dabry de Thiersant 


Cobitis sinensis SAUVAGE AND Dasry DE THIERSANT, 1874, Ann. Sci. Nat., 
Zool., I, Art. 5, p. 16. ‘Ruisseaux du Se-tchuan occidental (A. David).’ 

Description of aspecimen from Tungting Lake, Hunan, Clifford H. Pope.—Length 
to base of caudal 110 mm. Depth in length 7.3; head 5.4. Eye in head 6; snout 2.3; 
maxillary 4.2; depth of peduncle 2.1; its length 1.3; pectoral 1.4; ventral 1.6; 
longest dorsal ray 1.3; longest anal ray 1.5; caudal 1.2. Interorbital in eye 2; 
posterior barbel 1. 

Dorsal 9; anal 8. Scales before dorsal about 110. 

Conspicuous low adipose keels from the upper and lower caudal origins, not quite 
to dorsal and anal axils respectively. Mouth flanked by 3 barbels, the posterior the 
longest, adnate to a broad free lip which is adnate at its inner edge to the single 


Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No, 4 


2 AMERICAN MUSEUM NOVITATES [No. 170 


(paired) mental barbel, the lip between the two squarish, Dorsal origin equidistant 
- from tip of snout and base of caudal; pectoral extending a little more than 4 the 
distance to ventral; ventral 44 to anal; anal more than } to caudal base; caudal 
narrow, truncate. 

About a dozen faint blotches along the mid-line of the back; nine more or less 
narrow, oval, dark blotches along the side; darkish freckles above thisseries tending to 
form two lengthwise stripes; pale below it; a narrow stripe from below eye to snout. 
Dorsal with faint cross-marks; lower fins pale; caudal with a narrow, oblique, oval 
black spot on the base of its upper rays, and 3 or 4 broad V-shaped dusky cross-bands. 


Fig. 1. Cobitis tenia sinensis (upper) and Cobitis macrostigma (lower) compared. 


It is convenient for the present to consider this race of Cobztis tenia 
from the central Yangtze the same as Sauvage and Dabry de Thiersant 
had from its upper reaches, a point which cannot be settled without 
seeing material from nearer their locality. As to its range in eastern 
China, it occurs little changed farther down the river at Ningkwo, Anhwei. 
A specimen of 112 mm. from there has dorsal equidistant from end of 
snout and base of caudal; 9 cuboid blotches on side, those in front narrow; 
black mark on upper caudal base sharp and bold; lower lip squarish, but 
somewhat pointed at inner edge, which point extends well beyond 
mandibular barbel. Depth in length 7.5; depth of peduncle in head 1.8; 
its length 1.2. Scales before dorsal about 110. Another large specimen 
has about 12 blotches on the side. Eight small specimens, 55 to 74 mm. 
in length to base of caudal, from the same locality, are more variable than 
material to hand from any other one point. Depth in length 6.9 to 8.0. 
Snout in head 2.0 to 2.1; depth of peduncle 2.1 to 2.6; its length 1.3 to 
1.5. Dorsal equidistant from base of caudal and end or front part of 
snout; peduncular keels variable, moderate or little developed. Colors 


1925] CHINESE LOACHES, GENUS COBITIS 3 


rather sharp, 10 to 15 blotches along side; mark on upper caudal base 
sharp and strong. This variability may be explained by supposing that 
Cobitis tenia sinensis is here mixed to some extent with Cobztis [tenia] 
dolichorhynchus Nichols, 1918, a smaller race described from Fukien. A 
specimen of 43 mm. would pass for the young of dolichorhynchus. Depth 
5.8; snout 1.9; depth of peduncle 2.3; its length 2. Dorsal equidistant 
from base of caudal and middle of eye. 

Small specimens from Kwei-hwa, Shansi, situate on a small river 
entering the Yellow River from the northeast, near where that master 
stream turns south, and others from Hsing-lung-shan, Chihli, are still 
referable to C. t. sinensis, though not typical. A number of specimens 50 
to 63 mm. long to base of caudal from the first locality have depth in 
length 6.7 to 7.5; depth of peduncle in head 2.5 to 2.7; its length 1.0 to 
1.4. Dorsal origin equidistant from base of caudal and front of eve, or 
a point before middle of snout; peduncular keels moderately developed. 
Blotches on sides narrow, with a tendency to join in a narrow lengthwise 
stripe; caudal sharply barred, spot on its upper base well defined. Four 
specimens 59 to 66 mm. long from the last mentioned (Chihli) locality 
resemble these others in color, one individual having a dark longitudinal 
streak without blotches on the side and lacking the spot on upper caudal 
base. They have depth in length 6.6 to 7; depth of peduncle in head 
2.1 to 2.3; its length 1.2 to 1.3. Peduncular keels moderately developed ; 
dorsal origin equidistant from base of caudal and fore part of snout. 
The above specimens are too small for conclusions based on them to be 
satisfactory. It will be noticed that they vary towards dolichorhynchus 
more or, properly speaking, away from sinensis in more posterior position 
of dorsal origin, but the development of peduncular keels characteristic 
of sinensis holds better than with small Anhwei material. 

Three small specimens, 65 to 72 mm. in length, from Chin-ssu, Shansi, 
in the hills east of the southward-flowing limb of the Yellow River, 
have a uniform and rather striking color pattern. The locality is not very 
distant from Kwei-hwa in a south-southeasterly direction, but in what 
appears to be a rather distinct faunal area, very likely inhabited by a 
recognizable race of C. tenia, towards which sinensis from Kwei-hwa 
varies. 


Cobitis tenia melanoleuca, new subspecies 


Slender like sinensis, but more boldly marked, with posterior dorsal like doli- 
chorhynchus. Peduncular keels little developed, especially above. Lip between 
maxillary and mental barbels broad, squarish, not pointed, the mental barbel little 
projecting. Depth in length to base of caudal 6.8 to 7. Depth of peduncle in head 


4 AMERICAN MUSEUM NOVITATES [No. 170 


2.4; its length 1.3 to 1.5. Dorsal origin equidistant from base of caudal and front of 
eye. Color sharply marked; 12 to 16 lengthwise blotches on side; spot on upper 
caudal base inconspicuous or absent; a more or less perfect, more extensive dark bar 
across caudal base. 

The type is No. 8403, American Museum of Natural History; Chin-ssu, Shansi; 
Clifford H: Pope. 


Cobitis macrostigma Dabry de Thiersant 


Cobitis macrostigma DaBry DE THIERSANT, 1872, ‘Pisciculture en Chine,’ Pl. 
XLIx, fig. 4. Lakes of central China. 

Description of a specimen from Huping, Tungting Lake, Hunan, Clifford H. 
Pope.—Length to base of caudal 115 mm. Depth in length 8; head 5.5. Eye in 
head 5.5; snout 2.5; maxillary 4.7; depth of peduncle 2.3; its length 0.9; pectoral 
1.4; ventral 1.5; longest dorsal ray 1.2; longest anal ray 1.4; caudal 1.2. Inter- 
orbital in eye 2; posterior barbel 1. 

Dorsal 944; anal 8. Scales before dorsal about 135. 

Conspicuous low adipose keels from the upper and lower caudal origins not quite 
to dorsal and anal axils respectively. Mouth flanked by 3 barbels, the posterior the 
longest, adnate to a broad free lip which is adnate at its inner edge to the single 
(paired) mental barbel, and forms a pointed flap with an angle of slightly less than 90° 
between these two. Dorsal origin nearly equidistant from tip of snout and middle 
of peduncle; pectoral extending a little more than one-third distance to ventral; 
ventral less than }4 to anal; anal }4 to caudal base; caudal narrow, truncate. 

About 15 rectagonal dark blotches along mid-line of back; about 6 large rec- 
tagonal blotches along side; a narrow stripe from below eye to snout; caudal with V- 
shaped cross-marks and dorsal speckled; lower fins plain. A short, oblique black 
mark on the upper caudal base. 


REFERENCES 
Bera, 1916, ‘Poiss. Eaux Douces Russ.,’ p. 356. 
Dasry DE THIERSANT, 1872, ‘Pisciculture en Chine,’ Pl. xurx, fig. 4. 
Fow er, 1924, Bull. Amer. Mus. Nat. Hist. L, p. 399. 
KREYENBERG AND PAPPENHEIM, 1908, Sitzb. Ges. Naturf. Fr. Berl., p. 106. 
Nicuots, 1918, Proc Biol. Soc. Wash., XXXI, p. 16. 
NicHoLs AND Pops, (in press), ‘Fishes of Hainan,’ Bull. Amer. Mus. Nat. Hist. 
SAUVAGE AND Dasry DE THIERSANT, 1874, Ann. Sci. Nat. Zool., Art. 5, p. 16. 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 171 Tue AMERICAN Museum oF NatTurAL History May 26, 1925 
New York City 


59.7,55N (51) 


NEMACHEILUS AND RELATED LOACHES IN CHINA! 
By J. T. NicHoLs 


REVIEW OF THE GENERA AND SUBGENERA 


There are a great number and variety of small loaches in Asia with 
no suborbital spine and three pairs of barbels, more or less generally 
referred to the genus Nemacheilus, and The American Museum of 
Natural History’s recent Chinese collections contain several species of 
these, representing four distinct groups. 

First, there is a very prettily marked species from Hainan Island, 
Nemacheilus pulcher Nicholsand Pope (in press). Thisisa moderately short- 
bodied, symmetrical fish; finely scaled; lateral line complete; the caudal 
margin slightly forked; nostrils close together, well before the eye, the 
anterior in an even margined tube. It is evidently not distantly related 
to Nemacheilus fasciatus (Valenciennes), the type of Nemacheilus. | 

Secondly, there arespecimens of Regan’s Nemacheilus nigromaculatus 
and Nemacheilus pleurotenia from Yunnan. This latter fish is short- 
bodied, well compressed, the head slightly so; scales small, non-imbricate 
but distinct, lateral line imperfect; caudal margin slightly forked; 
nostrils separated by a distance greater than that of the posterior from 
eye, the anterior in a flap-like tube. The new subgenus Yunnanilus is 
proposed for Nemacheilus pleurotenia Regan, 1904, to include also 
Nemacheilus nigromaculatus Regan, 1904, and Nemacheilus salmonides 
Chaudhuri, 1911, doubtfully distinct from it, all from Yunnan. 

Thirdly, there are three or four species from northern China which 
are more elongate, little compressed; without evident scales but the lat- 
eral line complete or essentially so, distinct; nostrils narrowly separated 
from eye, close together, the anterior with a flap behind, its rim little 
raised in front; caudal variously truncate or slightly indentate. One of 
these is Nemacheilus toni (Dybowski) Fowler, 1924. Berg, 1916, makes 
Cobitis toni Dybowski, 1869, a race of European Nemacheilus barbatulus 
(Cobitis barbatula Linneeus, 1758), which is the type of the genus Barba- 
tula. Barbatula is, by the present writer, considered worthy of full 
generic rank for loaches of this third group. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 47. 


2 AMERICAN MUSEUM NOVITATES [No. 171 


The genus Diplophysa is an old one for barbatuloid loaches, mostly 
from high central Asia, having the posterior part of the air-bladder free 
and only the anterior part enclosed in a bony capsule. Herzenstein, 
1888, in an analytical study of numerous forms of Diplophysa and Barba- 
tula in Turkestan, Tibet, and elsewhere, concludes that this air-bladder 
character has here little systematic value and does not recognize Diplo- 
physa as a genus or natural group. Berg, 1916, however, gives it full 
generic rank, and Hora, 1922, does so with emphasis. The present writer 
is inclined to agree with Herzenstein, but is not familiar with any member 
of the genus. . 

The classification of barbatuloid loaches with center of abundance 
and variety in high central Asia presents a problem of extreme intricacy 
and difficulty. Hora, 1922, bases his conclusions on a careful considera- 
tion of anatomical details, and a reasonable idea of the importance of 
such details, which, however, may or may not apply in the present case. 
In attacking the same problem, Herzenstein, 1888, had more material, 
or material from a wider area, and was more influenced in its arrange- 
ment by the apparent closeness of relationship of one form to another in 
view of his knowledge of the whole. It is not unreasonable to suppose, 
bearing in mind the inherent taxonomic difficulties these loaches present, 
that Herzenstein’s findings may be emended to advantage, but the 
present writer agrees with his opinions where they differ from those of 
Hora as to the distinctness of the genus Lefua, as to the probable wide 
range, perhaps with geographic modification, of Barbatula stoliczkat and 
other species, and as to the comparative slight importance of certain 
structural characters here, and to the relatively great importance of other 
characters, usually trivial. 

Fourthly, there is in the Museum’s collections a singular small loach 
from Szechwan which is identified with Nemacheilus potanini Giinther, 
1896, from the River Ya. On the basis of this identification, N. potanini 
Giinther is here made the type of Homatula, new (from Homaloptera, 
which it suggests, and Barbatula; for the present considered a subgenus 
of Barbatula), to include also Barbatula berezowskii (Giinther, 1896) from 
Kansu and Barbatula oxygnathus (Regan, 1908) from Yunnan. Our 
specimens of B. (Homatula) potanini are moderately elongate, head well 
depressed, tail strongly compressed, approximately as deep as the great- 
est depth of body; caudal truncate; no evident scales, lateral line com- 
plete, well marked; nostrils close together at a moderate distance before 
eye, the anterior in a short tube with pointed flap behind. The jaws are 
peculiar, premaxillaries fused in a rounded point above, and each man- 
dible firm, curved, prominent, the two separated by a notch. 


; 


1925] NEMACHEILUS FROM CHINA 3 


NEW RACES OF BARBATULA IN NORTH CHINA 
Barbatula toni fowleri, new subspecies 

Nemacheilus toni Fow.eEr, 1924, Bull. Amer. Mus. Nat. Hist., L, p. 396. 

Barbatula toni (Dybowski), Cobitis toni Dybowski, 1869, was 
described from the Onon and Ingoda Rivers in Siberia, in the area north 
of eastern Mongolia and west of northern Manchuria. Berg, 1916, 
recognizes it as a widely distributed Asiatic subspecies of Barbatula 
barbatula of Europe and gives a very excellent figure of the form which he 
had to hand (Nemacheilus barbatulus toni, p. 341, Fig. 263, Lake Hanka, 
Ussuri drainage). -Fowler, 1924, lists American Museum of Natural His- 
tory material from Eastern Tombs, Chihl Province, as Nemacheilus toni 
(Dybowski), and considers Nemacheilus pechiliensis Fowler, 1899, the 
same. The present writer, however, cannot agree that the Eastern 
Tombs material is identical with that figured by Berg, or with pechiliensis 
as described by Fowler, 1899, from northeast of Dolon-nor close to the 
northern boundary of Chihli, and which resembles Berg’s figure. He here 
proposes the name Barbatula toni fowleri for the Eastern Tombs fish. 


DESCRIPTION OF TypE.—No. 8409, American Museum of Natural History, 
Eastern Tombs, Chihli, August 7, 1921, Clifford H. Pope. 

Length to base of caudal 85 mm. Depth in length 6.2; head 4.6. Eye in head 
5.7; snout 2.4; interorbital 4; maxillary 3; width of mouth 3.8; barbel 3.6; width 
of head and of body 1.8; depth of peduncle 2.3; its length 1.3; pectoral 1.3; ventral 
1.7; longest dorsal ray, 1.4; longest anal ray 1.8; caudal 1.3. 

Dorsal 9; anal 7%. Scales very fine, only evident on peduncle. 

Elongate, sub-cuboid; very little compressed except behind; snout bluntly 
pointed, vent almost immediately before anal origin in a low, backwardly directed 
tube. Interorbital slightly concave; orbital rim free above and in front adnate below 
and behind; mouth inferior curved transverse, appreciably behind tip of snout; lips 
thick, smooth, free, the lower cleft in the center with keel-like flaps on either side of 
the cleft; nostrils close together, the anterior in a low tube with pointed flap behind; 
eye slightly superolateral; maxillary not nearly to under front of eye; with a termi- 
nal barbel equalling in length the outer and longer of 2 barbels on the snout; gill- 
cleft vertical, gill-membranes broadly joined to side of breast before the lower pectoral 
axil. Dorsal origin equidistant from tip of snout and base of caudal; ventral placed 
below the front part of dorsal; pectoral broad, extending %.the distance to ventral; 
ventral % to anal; caudal oblong, subtruncate, its outer rays a little the longer; 
base of caudal slightly oblique forward and downward, with slight precurrent keels 
’ above and below. Lateral line complete, in the center, rising slightly to meet opercle. 

Color dull yellowish, darker above; numerous vague broken dark bars on sides; 
faint ones on dorsal and caudal. 


Prominent keel-like flaps on either side of the cleft of the lower lip, 
and appreciable precurrent caudal keels on the peduncle appear to be 
significant characters in Asiatic fishes of the genus and subgenus Barba- 


4 AMERICAN MUSEUM NOVITATES [No. 171 


tula, characteristic of the races of B. tont. Compared with B. toni as 
figured by Berg, 1916, B. t. fowlerz has a longer pectoral, slightly concave 
versus truncate caudal, ventral origin slightly behind that of dorsal, 
and double versus single row of dark marks on the side. 


Barbatula toni posteroventralis, new subspecies 


DESCRIPTION OF TypE.—No. 8410, American Museum of Natural History, 
Chin Ssu, Taiyuanfu, Shansi, August 15, 1922, Clifford H. Pope. 

Length to base of caudal 66mm. Depth in length 7.3; head 4. Eye in head 5.3; 
snout 2.5; interorbital 3.3; width of mouth 2.5; maxillary barbel 3; width of head 
1.5; depth of peduncle 3; length of peduncle 1.4; pectoral 1.3; ventral 1.6; longest 
dorsal ray 1.6; height of anal 1.7; caudal 1.4. 

Dorsal with 9 developed rays; anal with 744. No evident scales. 

Slender, little compressed, deepest at shoulder; head large; belly flat. Mouth 
inferior, transverse, with thick soft lips loosely surrounding the smooth firm jaws; 
two membranous keels at the chin; two barbels on each side of the snout overhanging 
the front of the mouth; and one at the end of the maxillary; no spines about the eye; 
eye superolateral, with a free rim best developed above and in front; nostrils close 
together in front of the eye, the anterior with a shallow tube and flap behind; gill- 
membranes confluent with the breast, the clefts separated by about % snout. Paired 
fins in a horizontal plane; origin of dorsal about equidistant from tip of snout and 
base of caudal; ventral under hind part of dorsal, their axils apposed; pectoral not 
reaching ventral, ventral not quite reaching anal, anal not reaching caudal; caudal 
subtruncate, very slightly concave, its rudimentary rays precurrent above and below 
on the peduncle in rather conspicuous low keels. Lateral line ceasing on the peduncle 
shortly before caudal base, straight to its end. 

Back and caudal freckled, the latter with a narrow pale tip; a series of small 
‘obscure dark blotches more or less confluent along the lateral line; dorsal speckled; 
lower parts and lower fins pale. 


This race differs markedly from B. t. toni and B. t. fowleri by the 
more posterior position of the ventrals, placed under the hind part of the 
dorsal. 


Barbatula yarkandensis sellefer, new subspecies 

Barbatula yarkandensis, Nemacheilus yarkandensis Day, 1876, was 
described from Turkestan. Herzenstein, 1888, credits it with a wide 
distribution in high Asia, where he divides it into several subspecies, but 
it has not been listed from even the western Chinese provinces. Never- 
theless, we have a small loach from Shansi which agrees sufficiently with 
Herzenstein’s diagnososis of yarkandensis to lead the writer to describe 
it as a race of that form. Its most noticeable difference is one of color, 
sharply marked, dark, cross-saddles on the back, whereas yarkandensis is 
said to have less tendency to dark cross markings than related forms. On 


1925] NEMACHEILUS FROM CHINA 5 


the other hand, it differs from the figure of B. y- brevibarbus (Herzenstein, 
1888, p. 78, Pl. 1, fig. 1), the subspecies to which it stands nearest, in 
having a decidedly longer peduncle, 1.4 versus 2 in the head. 


DeEscrieTION oF Type.—No. 8411, American Museum of Natural History, 
Chin Ssu, Taiyuanfu, Shansi, August 20, 1922, Clifford H. Pope. 

Length to base of caudal 73 mm. Depth in length 6.8; head 3.9; peduncle 5.5; 
dorsal base 7.5; length of pectoral 4.7; ventral 6.4; anal base about 10. Eye in 
head 6.2; snout 2.3; interorbital 5; width of head 1.7; width of mouth 3; maxillary 
barbel 3.1; pectoral 1.2; ventral 1.7; longest dorsal ray 1.5; height of anal 1.8; 
caudal lobe 1.3; depth of peduncle 3.3; length of peduncle 1.4. Postdorsal depth in 
the greatest depth 1.3; depth of peduncle 1.9; depth of peduncle in its length 2. 
Eye in interorbital 1.4. Dorsal base in its longest ray 1.4; ventral in pectoral 1.4; 
anal base in its longest ray 2. 

Dorsal with 9 developed rays; anal with 6. No evident scales. 

Body elongate, little compressed; head large; belly flat; depth of peduncle 
considerably greater than its thickness. Mouth inferior, curved, with thick grooved 
lips; maxillary extending % the distance to eye from snout; 2 pairs of barbels art 
snout overhanging front of mouth, one at end of maxillaries; nostrils close together 
in front of the eye, the anterior with a shallow tube and flap behind; eye superolateral, 
with a free rim best developed above and in front; gill-membranes broadly joined to 
isthmus. Pectorals and ventrals in a horizontal plane, the former broad; ventral 
base somewhat behind dorsal origin; dorsal origin equidistant from nostril and base 
of caudal; last simple ray of dorsal stiffened for somewhat less than half its length; 
pectoral bluntly pointed, not reaching ventral; ventral *4 to vent; anal not reaching 
caudal; caudal shallowy lunate. Lateral line well developed, complete, ending in 
an abrupt drop and rise at base of caudal. Distance from the base of caudal to the 
vent equal to that from the vent to the tip of the pectoral; distance from the vent 
to the ventral axil greater than the snout by the diameter of the eye. 

Broad dark saddles along the back, and irregular blotches on sides; pectoral and 
dorsal marked with dark; caudal with a broad dark V on its base opening backward, 
the limbs of the V brokenly produced as a streak on each lobe, also dark streaks be- 
tween the limbs on center of fin. 


Barbatula stoliczkai (Steindachner) 

Colitis stohékai SreEINDACHNER, 1866, Verh. Zool.-Bot. Ges., Wien, XVI, p. 
793, Pl. xiv; fig. 2. 

This is an abundant and very widely distributed form in high central 
Asia, also listed from the western provinces of China by Giinther, 1896. 
As Herzenstein, 1888, differentiates several subspecies to the westward, 
it is reasonable to suppose that our material from Shans! is differentiable. 
On the other hand, the writer finds no marked peculiarities therein and, 
such being the case, hesitates to recognize it as subspecifically distinct, 
lacking other material for comparison, more especially as Berg, 1916, 
synonymizes with stoliczkai forms recognized as full species by Herzen- 


6 AMERICAN MUSEUM NOVITATES [No. 171 


stein. For the benefit of those who would follow a different course, 
Barbatula stoliczkai shansi, new name, is available. 


DESCRIPTION OF A SPEcIMEN.—No. 8412, American Museum of Natural History, 
from Mai-tai-chao, Shansi (42 miles east of Paotow, Mongolia), May 27, 1922, 
Clifford H. Pope. 

Length to base of caudal 78 mm. Head in length 4.5. Depth in head 2; head 
width 1.6; eye 6; snout 2.6; interorbital 3.6; width of mouth 3; maxillary barbel 
3.6; depth of peduncle 2.9; length of peduncle 0.8; pectoral 1.3; ventral 1.5; longest 
dorsal ray 1.5; longest anal ray 1.9; caudal 1. Maxillary in snout 1.4. 

Dorsal 10; anal 7. No evident scales. 

Elongate, cylindrical, snout bluntly pointed; depressed before dorsal origin, 
compressed behind; belly flattened, the paired fins in a horizontal plane. Mouth 
inferior, curved, with thick grooved lips; two barbels on each side from snout over- 
hanging the mouth in front, and one from near end of maxillary; eye superolateral 
with rim imperfectly free; nostrils close together in front of the eye, the anterior with 
a shallow tube, and flap behind; gill-membranes confluent with breast in front of 
pectoral base. Dorsal origin equidistant from front of nostrils and base of caudal; 
ventral origin slightly behind that of dorsal; pectoral falling far short of ventral, 
ventral almost to anal, passing vent; anal far short of caudal; caudal with margins 
converging, but notched shallowly behind, the lower “lobe” longest. A cutaneous roll 
or flap in ventral axil; lateral line well developed, straight, stopping just short of 
caudal base. 

Freckles on the back; irregularly cross marked and spotted on the side; an in- 
complete dark bar across the base of caudal. None of the markings bold. 


In addition to the above, the following species listed from China 
appear to belong to the genus and subgenus Barbatula. Nemacheilus 
bleekeri Sauvage and Dabry de Thiersant, 1874, western Shensi, Nema- 
cheilus grahami Regan, 1906, Yunnan. Nemacheilus robustus Kessler, 
1876, Kansu, a doubtful species. Nemacheilus scleropterus Herzenstein, 
1888, extending eastward into Kansu. Nemacheilus mongolicus Bleeker?, 
listed from Yunnan by Regan, 1914. Incidentally, Cobztis spiloptera 
Cuvier and Valenciennes, 1846, from Cochin-china, looks like a race of 
Barbatula toni. 


REFERENCES 


Berg, 1916, ‘Poiss. Eaux Douces Russ.,’ p. 340. 
Cuavupuunl, 1911, Rec. Ind. Mus., VI, p. 18, Pl. 1, fig. 3, 3a. 
CUVIER AND VALENCIENNES, 1846, ‘Hist. Nat. Poiss.,’ XVIII, p. 27, Pl. pxx11. 
Day, 1876, Proc. Zoél. Soc., p. 796. 
DysowskI, 1869, Verh. Zool.-bot. Gesell., Wien, XIX, p. 957, Pl. xvi, fig. 10. 
Fow ter, 1899, Proc. Acad. Phila., p. 181. 
1924, Bull. Amer. Mus. Nat. Hist., L, p. 396. 
GiunTHER, 1896, Ann. Mus. Zool. St. P., I, p. 217-218. 
HERZENSTEIN, 1888, Wiss. Res. Przewalski Reis., Zool., III, pt. 2. 
Hora, 1922, Rec. Ind. Mus., XXIV, p. 63, ete. 


1925] NEMACHEILUS FROM CHINA 


Kesster, 1876, in Prejevalski, ‘Mong. Stran. Tang.,’ IT, pt. 4, p. 32. 
Linnavs, 1758, ‘Syst. Nat.,’ 10th Ed., I, p. 303. 
NICHOLS AND Pops (in press), ‘Fishes of Hainan,’ Bull. Amer. Mus. Nat. Hist. 
Reean, 1904, Ann. and Mag. Nat. Hist., (7) XIII, p. 192. 

1906, Ann. and Mag. Nat. Hist., (7) XVII, p. 333. 

1908, Ann. and Mag. Nat. Hist., (8) II, p. 357. 

1914, Ann. and Mag. Nat. Hist., (8) XIII, p. 261. 
SAUVAGE AND Dapry DE THIERSANT, 1874, Ann. Sci. Nat., I, No. 5, p. 15. 
STEINDACHNER, 1866, Verh. Zool.-bot. Gesell. Wien, XVI, p. 793, Pl. xtv, fig. 2. 


oe) See 


ase te ae ie ae 


x — © fi 
omtay aretaii 


ee irey anu) 
shal 


A esherid* Cap Plait 
Mai 


% 
° 


( 
A 
1 es é 
, ut 
EY tek ad : 
} t i 
4 
LAN 
4 b i 
| 
ove Fj ‘ 
A y AN 
Fi | 
, . 
fe 
' 


ree ‘A TAs 


‘é ¢ 


a, gh: aes ef 
SG) A 


& ~ is 
4 , ty 
, Ae aa 
¥ : a bs 
igs . a 
abet wee lef 
ey td ’ 
yict <7 ee : 
A ry * 
Py, + oy 1. 1% 
i a 
el ; Fi 
‘ 4 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 172 Tue American Museum or NaTuRAL History May 26, 1925 
New York City 2 


59.7,55H (51.2) 


_HOMALOPTERA CALDWELLI, A NEW CHINESE LOACH! 
By J. T. NicHois 


Homaloptera caldwelli, new species 


Description oF Type.—Number 8413, American Museum of Natural History; 
from near Yenping, Fukien; H. R. Caldwell. 

Length to base of caudal, 48 mm. Depth in length, 5; head, 4.6. Eye in head, 5; 
snout, 2; interorbital, 2.4; width of mouth, 3.5; posterior barbel, 5; greatest width 
of body (behind head), 1.5; depth of peduncle, 2; its length, 1.7; pectoral, 0 8; 
ventral, 1; longest dorsal ray, 1.1; longest anal ray, 1.3; caudal broken. 

Dorsal, 10; anal, 7. Scales very fine, about 150. 

Head depressed; body compressed; vent behind a thick papilla, a little nearer 
ventral axil than anal origin; lower surface of head and breast flattish. Interorbital 
very slightly convex; snout rounded above, moderately tapering to a rounded point; 
eye slightly superolateral; orbital rim free; mouth inferior, on the lower surface of 
snout, curved, transverse; overhung in front by a fleshy, slightly curved membrane, 
immediately before which on the snout are four small barbels, the outer the longer; 
a spongy lip on the lower surface of mandible, at each corner of which is a longer 
barbel with a little knob at its base on the inside; gill-membranes joined to breast 
below the front of the pectoral base. Dorsal origin equidistant from base of caudal 
and front of eye; ventral origin slightly behind it; pectoral reaching about *4 the 
distance to ventral; ventral, 34 to anal; caudal broken, was probably truncate or 
subtruncate. Lateral line complete, slightly curved upward, in the center of peduncle. 

A black stripe from shoulder onto base of caudal; dorsal crossed by two or three 
broken streaks; back behind dorsal with four cross blotches. 


We have only this one specimen. The species is related to Homalo- 
soma stenosoma Boulenger, 1901, Ningpo. Homalosoma is probably best 
considered a subgenus of Homaloptera, which genus is typically East 
Indian. The former is very divergent from Lepturichthys Regan, 1911, 
of which Homaloptera fimbriata Giinther, 1888, from the Yangtze, is the 
type. Homaloptera abbreviata Giinther, 1892, mountain streams running 
into the Min (tributary to the western Yangtze in Szechwan), approaches 
fimbriata. 

REFERENCES 
BovuLENGER, 1901, Proc. Zoél. Soc., part 1, p. 270, Pl. xxm1, fig. 3. 


GUNTHER, 1888, Ann. and Mag. Nat. Hist., I, p. 483; 1892, in Pratt, ‘Snows of 
Tibet,’ p. 248, Pl. 1, fig. B. 


Reaan, 1911, Ann. and Mag. Nat. Hist., VIII, p. 31. 


‘Publications of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 48. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 173 Tue American Museum or NaTurAt History May 27, 1925 
New York City . : 


56.81,9:13.1 
THE MICROSTRUCTURE OF THE DINOSAURIAN 
EGG-SHELLS FROM THE CRETACEOUS BEDS 
OF MONGOLIA! 


By Victor VAN STRAELEN? 


Through the courtesy of President Henry Fairfield Osborn of The 
American Museum of Natural History, I have been given the opportunity 
of examining the structure of several of the dinosaurian egg-shells re- 
cently collected in Mongolia by the Third Asiatic Expedition under the 
leadership of Mr. Roy Chapman Andrews. ; 

In view of an exhaustive study of this material, to be carried on 
together with the study of other fossil eggs of various origins, I am giving 
here a preliminary accounteof my observations. 


I. PROBABLE EGGS OF PROTOCERATOPS ANDREWSI, FROM THE 
DJADOCHTA BEDS (CRETACEOUS) OF SHABARAKH USU, MONGOLIA 


1. Samples of two eggs, out of a group of fifteen, Amer. Mus. No. 6508. 

The parts of the shell which have been fossilized are the mamillar 
zone, i.e., the zone in contact with the chorion, and the prismatic zone. 
The shell, formed of calcite, mixed with phosphate of lime, is about 1 mm. 
thick. Its color is red brown, due to the infiltered iron oxide. High 
subparallel and meandriform hillocks, separated by very large and deep 
valleys, give a vermiculate aspect to the outer face. On the contrary, 
the inner face is nearly smooth, the mamillz being very small and closely 
crowded together. The fibro-radious structure is therefore finely packed 
and the prisms of the prismatic zone are small. The pores are extremely 
reduced, both in number and size, and consequently the aériferous canals 
have a very small diameter. Throughout their length, the canals main- 
tain their diameter and are not ramified. The prisms of calcite are 
crossed by thin black layers of organic matter. 

Secondary mineralization of the shell took place during the fossiliza- 
tion, so that the pores and canals are entirely choked with calcite. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
bution No. 49. : 
2Université Libre, Brussels, Belgium. 


— 
——— 

a ee eS —- —- ——_—____ 
en —— 
—— 
ee ——————— 


| 
—— 


Co man AVY 


Fig. 1.. Section through a portion of an egg-shell found at Shabarakh Usu, 
Mongolia. (Amer. Mus. No. 6508.) 


M.Z.=mamillar zone; P.Z.=prismatic zone; m.=mamillz; a.c.=aériferous canals; p.=pores. 
x48. 


i : 
; 


| 
( 
* 


PZ 


i 


Fig. 2. Section through a portion of an egg-shell found at Iren Dabasu, Mon- 
golia. (Amer. Mus. No. 6552.) 


eee zone; P.Z.=prismatic zone; m.=mamille; a.c.=aériferous canals; p.=pores. 
X48. 


1925] DINOSAURIAN EGG-SHELLS FROM MONGOLIA 3 


2. Sample of one egg, out of a group of five, Amer. Mus. No. 6511. 

The nature and state of preservation of this eg@ are just the same as 
in the samples of the group of fifteen described above. 

3. Sample of one egg, out of a group of three, Amer. Mus. No. 6510. 

The outer and inner faces of the shell are much corroded, only the 
layers showing the prismatic structure being left. The concentric layers, 
perpendicular to the crystalline fibres, are well shown and lined by thin 
black sheets of organic matter. The structure of the part left is identical 
with that of the eggs of the two other groups. 


II. EGGS OF A DINOSAUR (? DUCKBILLED TYPE), FROM THE IREN 
DABASU BEDS (UPPER CRETACEOUS) AT IREN DABASU, MONGOLIA 


Sample from a specimen, Amer. Mus. No. 6552. 

The two faces, outer and inner, are worn; the shell formed by calcite, 
containing phosphate of lime, is reduced to the mamillar zone and a 
great part of the prismatic zone. Nevertheless, the thickness is still 
about 1.2 mm., which makes about 2 mm. for the complete shell. 

The outer face is whitish and shagreened, due to small sinu- 
ous hillocks, separated by depressions. Numerous pores perforate that 
surface, many of these pores being visible with the naked eye and com- 
parable in size with those existing in the eggs of certain crocodiles. These 
pores are the orifices of the aériferous canals, which open indifferently 
on the ridges as well as in the depressions. The shape of the pores is 
apparently irregular. 

The grayish brown inner face has a more irregular shape, the mamil- 
le being fairly isolated at their origin. The mamille are lining more or 
less circular alveole, at the bottom of which the aériferous canals take 
their origin. The fibro-radious structure is particularly well developed 
in the most internal zones, where the crystalline fibers are very minute. 
In the more external parts, the fibers gradually thicken. 

The aériferous canals are largely expanded in their middle part. 

Thus, there are striking differences between the eggs of Shabarakh 
Usu and those of Iren Dabasu. The first-named cannot be correlated 
with any of the actually known eggs, either living or fossil. The second- 
named have a structure similar to that of the supposed eggs of Hypselo- 
saurus priscus, the dinosaurian of Rognac, which themselves have a 
structure that partakes of the characters of both the paleognathie and 
the neognathic birds. But the Iren Dabasu eggs differ essentially from 
the Rognac eggs in the shape of the aériferous canals. 


1The only reptilian egg-shells of secondary times, whose microstructure is known, are those 
found in the lacustrine marls of Rognae (southern France), at the top of the Upper Cretaceous. 


1 AMERICAN MUSEUM NOVITATES [No. 173 


The eggs of Protoceratops andrewsi are of the utmost interest. From 
the rugosities of the outer surface together with the rare and extremely 
small pores, it is right to infer that the eggs had no outer cuticle. This is 
a character shown to-day by birds and turtles which lay their eggs in 
very dry regions. We may find herein a confirmation of the desert 
conditions prevailing in Mongolia during the formation of the Djadochta 
beds. ; 

I wish to express to President Osborn my appreciation of the rare 
opportunity given me of examining these dinosaurian egg-shells. I am 
also indebted to Mr. Walter Granger, Palzeontologist of the Third Asiatic 
Expedition, who has given me valuable information regarding them and 
who made the selection of the fragments submitted to me. 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 175 Tue AmericaN Museum or Natura History May 28, 1925 
New York City 


59.76(51) 


NEW CHINESE AMPHIBIANS AND REPTILES! 
By Karu Patrerson SCHMIDT? 


The Chinese collections of The American Museum of Natural 
History contain four new forms of frogs and toads. In the course of the 
preparation of more extended reports on the Chinese amphibians and 
reptiles of the Third Asiatic Expedition (now ready for the press), seven 
additional new reptiles have been brought to light. Diagnosis of these 
new forms supplement the series described in American Museum Novi- 
tates No. 157, bringing the total number of new forms, in the collections 
reported upon, to twenty-six. 


SALIENTIA 


Bufo andrewsi,’ new species 


Type.—A. M.N. H. No. 5769; ; Likiang, 8500 feet altitude, Yunnan; 
October 4, 1916; R. C. Andrews and Edmund Heller. 

Diaenosis.—Closely allied to Bufo bufo, from which it is distinguished by the. 
presence of a tarsal fold, its finer and more uniform tuberculation, the less divergent 
parotids, and the tuberculate top of the head. 


5) 


Rana nigromaculata mongolia, new subspecies 


Type.—A. M. N. H. No. 18149; <; Mai Tai Chao, northern Shansi; May 1922; 
Clifford H. Pope. 

D1aGnosis.—Derived from Rana nigromaculata nigromaculata, with which it 
agrees in having a very elongate metatarsal tubercle. Distinguished by a much more 
rugose dorsal skin, with very short longitudinal folds, shorter legs, a broader head, 
and the absence of the light line on the dorsolateral fold. 


Rana noblei,* new species 
Typr.—A. M. N. H. No. 5285; 9; Yunnanfu, Yunnan; John Graham. 
Draenosis.—Allied to Rana nigromaculata, from which it is distinguished by its 
more rounded snout, absence of dorsal folds between the dorsolateral folds, smaller 
metatarsal tubercle, and very different coloration, which is largely reddish brown. 


oe of the Asiatic Expeditions of the American Museum of Natural History. Contribu- 
tion No. 50. 
2Of the Field Museum of Natural History. 
’Named for Mr. Roy Chapman Andrews, Leader of the Third Asiatic Expedition. ; 
4Named for Dr. G. Kingsley Noble, Curator of Herpetology, American Museum of Natural History. 


2 AMERICAN MUSEUM NOVITATES [No. 175 


Rana caldwelli,'! new species 


Type.—aA. M. N. H. No. 18485; o; Fukien Province (probably near Yenping); 
H. R. Caldwell. ; 

Diaenosis.—Allied to Rana adenopleura, of which it is the continental repre- 
sentative. Distinguished by having a more projecting snout, rougher skin, and the 
dorsolateral glandular folds broken up posteriorly. 


SAURIA 


Calotes alticristatus, new species 


Type.—A. M.N. H. No. 17395; 9; Yunnanfu, Yunnan; 1919; John Graham. 

Diacnosis.—Closely allied to Calotes emma, from which it is distinguished by a 
greater number of scales around the body, smaller postcanthal and nuchal spines, a 
longer nuchal crest, larger tympanum, and the absence of a dorsolateral light line. 
Differs from Calotes yunnanensis in having a better developed fold in front of the 
shoulder, and in the higher number of scales around the body. 


Eremias barbouri,”? new species 


Typr.—A. M. N. H. No. 24045; o; Mai Tai Chao, northern Shansi; May 1922; 
Clifford H. Pope. 

Draenosis.—Directly allied to Hremias argus, from which it may be distin- 
guished by its larger dorsal scales and a color pattern of light longitudinal lines or 
rows of spots, combined with transverse black bars. 


SERPENTES 


Natrix septemlineata, new species 


Typr.—A. M. N. H. No. 21051; o&; Tengyueh, Yunnan; May 1917; R. C. 
Andrews and Edmund Heller. 

Draenosis.—Dorsal scales, 19; weakly keeled, outer row smooth; ventrals, 
159-171; anal divided; caudals, 82-89; upper labials, 8; preoculars, 1; postoculars, 
3; temporals, 2-1-2; venter uniform light, without spots at the ends of the ventrals; 
back with seven dark longitudinal stripes. 


Dinodon rufozonatum williamsi,* new subspecies 


Type.—A. M. N. H. No. 17453; 9; Changsha, Hunan; July 1920; J. W. 
Williams. 

DraGenosis.— Distinguished from the typical subspecies by a greater number of 
ventrals, 207-213; of caudals, 77-86; and of transverse light markings, 59+21 to 
87 +26. 


1Named for Mr. Harry R. Caldwell, Yenping, Fukien. 
2Named for Dr. Thomas Barbour, of the Museum of Comparative Zodlogy. 
‘Named for Mr. J. W. Williams, of the College of Yale in China, Changsha, Hunan. 


1925] CHINESE AMPHIBIANS AND REPTILES 3 


Elaphe bimaculata, new species 


Type.—A. M. N. H. No. 24640; 92; Ningkwo, Anhwei, September-October 
1921; Clifford H. Pope. 

Diacnosis.—Closely allied to Elaphe dione, from which it is distinguished chiefly 
by color characters. Dorsal scales, 25; ventrals, 188-207; subcaudals, 67-74; 
transverse dorsal spots dumb-bell-shaped, often separated as a pair of spots; several 
of these spots unite on the neck and are confluent with the head marking; tail witha 
light median and dark dorso-lateral stripes. 


Elaphe porphyracea pulchra, new subspecies 


Type.—A. M. N. H. No. 17705; o; 20 miles north of Yunnanfu, Yunnan; 
July 6, 1920; John Graham. 

Diacnosis.—Closely allied to Elaphe porphyracea porphyracea in pattern; 
distinguished by having fewer ventrals, 177-185, and subcaudals, 51-56. 


Trimeresurus orientalis, new species 


Typre.—A. M. N. H. No. 21028; 9; Shaowu, Min River, Fukien; R. C. 
Andrews and Edmund Heller. 

Diacnosis.— Distinguished from its near relative, the Himalayan T. monticola, 
by having ten upper labials instead of eight or nine; a second near ally, 7. okinavensis, 
has seven or eight labials; ventrals, 138; caudals, 37; dorsal scale rows, 27-25-21. 


Nina sts _ ease: canis 
nc > kare BP ei eh, Raat a eae 
my hg ee afin. endl Roce aethast tii Witdknai? teased 
sale vy iy oo Noni inant ee ea * heen 
at swath ere as da Y 


i 
, al » 7 


ciate Wate hale 8a 
cessor dam 


Bi) Wiener Sat Pv eget MAS, jy ay Se Nien ia Bite ie 
Fi: bent god bn ue oh oanha asta ar om f 
. a Feel akin rome hE” Tactile & fi ath aaa VR dae bs 
on belt Niniinr keh on 
ee Melbianies Taye Sas pat Ay ‘aot a em btske gat inde 
' Witenes | Thy ay ahah orn bain tes wives hphandiat sanity 
be: tats Crd Vy Fa bits “heen Mie miata x a i iy 2 olen tas aaa 


wee | RNs, ae 


* TBs ui 1} at 
* ‘ 5 : 
Ce ped al ; ve es ae yard) Le ae? ‘eS SOR 
i ; ' 
Suu iy. 40% r #1 Vers a 4 i ry c. a) as 
, 
7 + 
j 
‘ j A a 
. 
a 
my 4 
‘ p 
' 
‘ 
’ 
. 
- x \ 
. 
a 
. | 
vf 
’ wry 
> 
mth NP 
“he be 
al v 
i \ se N 
wt - _ 
a « 1 i 
fo 4 rm . 
é 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 177 Tue American Museum or NaTuRAL History June 20, 1925 
New York City ? 


59.7(51) 


SOME CHINESE FRESH-WATER FISHES! 
By J. T. NicHOoLs 


I—LOACHES OF THE GENUS BOTIA IN THE YANGTZE BASIN 
Botia rubrilabris (Dabry de Thiersant) 

?Cobitis variegata DABRY DE THIERSANT, 1872, ‘Pisciculture en Chine,’ Pl. xix, 
fig. 5. Unidentifiable. Yetchuen. 

Parabotia rubrilabris DABRY DE THIERSANT, 1872, ‘Pisciculture en Chine,’ PI. 
XLIX, fig. 8. Yangtze (referred to Guichenot). 

Botia variegata GUNTHER, 1889, Ann. and Mag. Nat. Hist., IV, p. 228. Ichang, 
Yangtze. 


Fig. 1. Botia rubrilabris from Tungting Lake. Length, without caudal, 65 mm. 


The synonymy of this, the best-known species of the genus in China, 
is somewhat uncertain and involved. Our two small specimens from 
Tungting Lake, Hunan, are clearly referable to Dabry’s inaccurate but 
characteristic figure of rubrilabris, and also agree with Giinther’s type 
description of variegata. Giinther aptly says in Ann. and Mag. Nat. 
Hist., 1888, I, p. 429: ‘I regret not to be able to make use of the notes on 
Chinese fishes in ‘La Pisciculture et La Péche en Chine par P. Dabry de 
Thiersant,’ as the figures as well as the accompanying notes are the work 
of persons not conversant with the rudiments of descriptive ichthyology, 
and as likely to lead to misconceptions as to assist in the determination of 
the species.”” Nevertheless, some of the species in this work are perfectly 
identifiable. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 51. . 


2 AMERICAN MUSEUM NOVITATES [No. 177 


A specimen from Huping, Tungting Lake, Hunan, collected by 
Clifford H. Pope, may be described as follows. 


Length to base of caudal, 50 mm. Depth in length, 4.5; head, 3.3. Snout in 
head, 2.5; width of body, 2.4; depth of peduncle, 2.2; its length, 1.8; pectoral, 1.8; 
ventral, 1.8; longest dorsal ray, 1.4; longest anal ray, 1.6; caudal lobe, 1. Eye in 
snout, 4; interorbital, 2.1; maxillary, 2; maxillary barbel, 2.3. 

Dorsal, 11; anal, 7. Scales very fine, evident only on peduncle. 

Head long and pointed; peduncle moderately constricted; body compressed; 
peduncle strongly compressed; ventrals contiguous, in a horizontal plane; vent’ 
equidistant from ventral axil and anal origin. A row of four contiguous barbels across 
the tip of the snout, the inner slightly the longer and slightly shorter than a barbel on 
the end of the maxillary; mouth slightly oblique, inferior, horseshoe-shaped; the 
maxillary not reaching half the distance from tip of snout to under front of eye; lips 
thickish; the lower cleft in the middle; snout long and narrow; interorbital strongly 
convex; eye with a slight free rim, except behind; subtended by asimple, concealed, 
hackwardly directed curved spine, a little longer than diameter of eye; gill-membranes 
broadly joined to side of breast beneath posterior third of opercle. Dorsal origin 
equidistant from base of caudal and middle of eye; ventral origin slightly behind that 
of dorsal; pectoral reaching 47 the distance to ventral; ventral % to anal; caudal 
deeply forked with pointed lobe (one broken). Lateral line complete, straight, in 
the center; short, bluntly pointed fleshy flaps in pectoral and ventral axils. 

Pale; six dark saddles along mid-dorsal line; the first between the front of the 
eyes, the fifth on the dorsal base; a narrow stripe across the caudal base; a large 
blotch on the peduncle, two faint blotches anterior to this one on one side. <A less 
perfect specimen of 63 mm. with seven dorsal blotches. 


Botia pratti Giinther 
?Parabotia teniops SAUVAGE, 1878, Bull. Soc. Philom. Paris, p. 90. Yangtze. 

Botia pratti GUNTHER, 1892, in Pratt, ‘The Snows of Tibet,’ p. 250, Pl. rv, fig. A. 

This is a species with an elongate snout, but the very small eye rather 
nearer the end of the snout than the edge of the operculum. Depth 
rather more than ¥% the length to base of caudal; head in same, 3.5; 
interorbital narrow, convex, 3 or 4 times the diameter of eye. Dorsal 
origin midway between base of caudal and orbit; caudal deeply forked, 
its lobes as long as the head. Dorsal with 11 rays; anal with 8. Brown- 
ish olive, without distinct markings on the body; dorsal with two black- 
ish bands parallel to the upper margin; caudal rays with numerous linear 
black markings or without spots; lower fins with indistinct blackish 
markings. The figure shows snout in head, 2.3; depth of peduncle, 2.4; 
its length, 1.6; longest anal ray, 1.5; eye in snout, 8.5. 

Giinther had three specimens 8 inches long, from Kia-tiang-fu (eleva- 
tion 1070 ft.) at the foot of Omieshan. The locality is central Szechwan 
near long. 104° East, and where the Tung and Ya rivers join to enter the 
Min, a northern affluent of the Yangtze. 


1925] SOME CHINESE FRESH-WATER FISHES 3 


PLATE Iv. 


Fig. 2. Botia pratti (upper) and Botia swperciliaris (lower) from Giinther, by 
courtesy of the publishers, Longmans, Green and Co. 


Botia superciliaris Giinther 

Botia superciliaris GUNTHER, 1892, in Pratt, ‘The Snows of Tibet,’ p. 250, Pl. rv, 
fig. B. 
This is a species allied to B. rubrilabris from the same locality as the 
above, with an elongate snout, eye behind the center of the head, and a 
short deep peduncle. Depth in length to base of caudal, 4.5; head, 3.5. 
Interorbital narrow, convex, 2% times the diameter of eye. Dorsal 
origin midway between base of caudal and eye; caudal forked, its lobes 
shorter than the head. Dorsal with 11 rays; anal with 8; light olive, 
with broad, brownish bars across the back, five in front of and five behind 
the dorsal; a yellowish streak on the side of snout running back through 
superciliary region; three yellowish longitudinal lines on the crown of the 
head; dorsal and generally pectoral with a broad, dark cross-band well 
within the margin; each caudal lobe with three or four dark oblique 
bands. The figure shows snout in head, 1.9; depth of peduncle, 1.9; 
its length, 2.4; longest anal ray, 1.9. Eye in snout, 5. 

Giinther had five specimens, six inches long, from Kia-tiang-fu, 
Szechwan. 


4 AMERICAN MUSEUM NOVITATES [No. 177 


Botia purpurea, new species 
DEscrIPTION OF TyprE.—Number 8401, American Museum of Natural History, 
from Huping, Tungting Lake, Hunan; collected by Clifford H. Pope. 
Length to base of caudal, 132 mm. Depth in length, 4.2; head, 4. Snout in 
head, 2.5; width of body, 2; depth of peduncle, 1.6; its length, 1.4; pectoral, 1.4; 


Fig. 3. “Botia purpurea of 63 mm. without caudal, from the 
side and from above. 


ventral, 1.5; longest dorsal ray, 1.4; longest anal ray, 1.5; upper caudal lobe, 0.9. 
Kye in snout, 5.5; interorbital, 1.7; width of mouth, 2; maxillary, 1.6; maxillary 
barbel, 2.5; distance between gill clefts, 1.5. 

Dorsal, 10; anal, 8. Scales about 150. 

Compressed; peduncle broad and strongly compressed; snout pointed; vent 
equidistant from ventral axil and anal origin; ventrals contiguous, in a horizontal 


1925] SOME CHINESE FRESH-WATER FISHES 5 


plane. Mouth small, slightly oblique, inferior, transverse, horseshoe-shaped, with 
moderately thick lips, the upper free behind, the lower, which is split in the center, 
free in front; maxillary reaching half-way to under front of eye from tip of snout; 
three small barbels on each side, two at the side of the tip of the snout, and one at the 
end of the maxillary a little the longest; eye with a slight free rim, except behind; a 
strong, simple, backwardly directed spine beneath the eye, half as long again as same; 
gill-slit curving around pectoral base, gill-membranes broadly joined to side of breast 
before lower pectoral axil. Dorsal origin equidistant from tip of snout and base of 
caudal; ventral origin slightly behind that of dorsal; pectoral reaching % the distance 
to ventral; ventral 74 to anal; caudal deeply forked, with pointed lobes, the upper 
slightly the longer. Scales small, oval, imbricated, with concentric and parallel 
horizontal striz; lateral line complete, straight, in the center; pectoral and ventral 
with fleshy axillary flaps joined by membrane to the fin, that of the ventral elongate, 
pointed. 

Rather dark purplish, the sides vermiculated; midline of back from nape to 
dorsal with about five irregular dark saddle-marks, narrowly separated; fins with dark 
bases, each crossed by a blackish bar; mouth and barbels pale. In a sketch from life 
the ground color is dull purplish rose; broad dark cross-bars before the dorsal leave 
narrower interspaces of the ground color; behind the dorsal light and dark bars are 
more numerous and of about equal width; the sides are irregularly mottled. 


A number of specimens of this form were collected at Tungting Lake. 


Botia citrauratea, new species 


Description oF TypE.—Number 8402, American Museum of Natural History, 
Tungting Lake, Hunan; December 29, 1921; collected by Clifford H. Pope. 

Length to base of caudal, 50 mm. Depth in length, 5.4; head, 3.8. Snout in 
head, 2.6; depth of peduncle, 1.8; its length, 1.9; pectoral, 1.7; ventral, 1.8; longest 
dorsal ray, 1.5; longest anal ray, 1.8; caudal lobe, 0.9. Eye in snout, 3; interorbital, 
2; maxillary, 1.4; maxillary barbel, 1.4. 

Dorsal, 10; anal, 7. Scales minute, imbedded, scarcely evident. 

Body strongly compressed, particularly the peduncle. Mouth inferior, strongly 
curved; two barbels on snout and one on maxillary on each side, which latter reaches 
back to opposite center of eye, none on chin. Small eye with a free rim; a strong, 
erectile, backwardly directed spine below eye; gill-membranes broadly joined to 
isthmus, gill cleft extending to opposite lowest ray of pectoral. Dorsal origin midway 
between base of caudal and middle of snout; ventral origin under first rays of dorsal. 
Lateral line complete. 

Color purplish brown; yellowish below; a dark band across base of caudal; a 
few pale flecks on sides and dark flecks on dorsal; midline of back rather dark with a 
row of numerous small, close-spaced, rounded, pale blotches. Caudal faintly but 
distinctly mottled with dark. A sketch from life is orange in color, the center of the 
back dark gray; an orange stripe extends between the eyes and onto the back, behind 
which there are eight rounded orange spots in the midline of the back before the dorsal 
and nine behind it. 


These five Chinese species of Botia may be conveniently differen- 
tiated as follows: 


6 AMERICAN MUSEUM NOVITATES [No. 177 


1.— Dorsal origin midway between base of caudal and eye..................005- 2. 

Dorsal origin midway between base of caudal and middle of snout. Center of 

back dark with a series of close-spaced, rounded, pale marks. Slender, 

depthyabouto.aa. cee ee ee ae Tee ee ate B. citrauratea. 

Dorsal origin midway between base of caudal and end of snout. Dark saddles 
across the back, sides finely marked. Deeper, depth less than 4.5. 

B. purpurea. 

2.—Eye very small, more than 3 times in the interorbital, and about 8 times in the 

snout. No dark cross-marks on back or sides. Eye in about the center of 

end" Deptdbout bree. 5 cco uoas ote ee ela ee nein eee oe B. pratti. 

Eye somewhat larger, less than 3 times in the interorbital, about 4 or 5 times in 

the snout. Dark cross-marks on back which may be continuous downward, 

represented by blotches or absent across the sides. Eye behind the center of 

head> Depth about/4.5.02. acct. aoe c ne certin eles os ecient ee ee 3. 

3.—About 10 cross-marks on the back. Peduncle short and deep. .B. superciliaris. 

About 5 to 7 cross-marks on the back. Peduncle moderate, its’ length 1.8 in 

HOWG Seif 5 soe taere hs Peed Ore ae Rite es a ee a an ers eee B. rubrilabris. 


II.—A NEW MINNOW-LIKE CARP FROM SZECHWAN 
CARASPIUS, new genus 


A small, active cyprinid, with the chubby blunt-headed appearance of the gold- 
fish (Carassius), but with a long, slender peduncle. Dorsal and anal short, without 
spinous rays. Mouth very oblique, lower jaw projecting. No barbels. A short keel 
on the belly before anal. Scales moderately large; the lateral line incomplete. Name 
from Carassius and Aspius. Type: Caraspius agilis, new species. 


This genus resembles European Leucaspius, which has a longer anal, 
smaller scales, etc., and may be more or less closely related thereto. 


Caraspius agilis, new species 

Description or Typre.—No. 8414, American Museum of Natural History, 
Yen-ching-kao, near Wanhsien, Szechwan, January 1923, collected by Walter Granger. 

Length to base of caudal, 41 mm. Depth in length, 3.3; head, 3.6. Eye in head, 
3.5; snout, 3.5; interorbital, 2.5; maxillary, 3; depth of peduncle, 1.9; length of 
peduncle, 1.4; width of body, 1.5; pectoral, 1.3; ventral, 1.6; longest dorsal ray, 
1.4; longest anal ray, 1.5; caudal, 0.8. 

Dorsal, 9; anal, 9. Scales, 31. Teeth, 4.2 (in one of the cotypes), their tips 
narrowed and curved. ? 

Vent immediately before anal origin, with a short, naked keel before it almost to 
ventral axil. Lower jaw projecting; mouth very oblique; maxillary barely to front 
of eye; no barbels; gill-membranes narrowly joined to isthmus under edge of pre- 
opercle. Dorsal and anal without spinous rays; dorsal origin equidistant from base of 
caudal and edge of preopercle, between ventral axil and anal origin, which latter is 
very slightly behind dorsal axil; pectoral reaching’ to over ventral origin; ventral 
not quite to anal origin; caudal forked, the upper lobe pointed. Scales with radiating 
strie; lateral line on about 3 anterior scales only. 


. 


1925] SOME CHINESE FRESH-WATER FISHES 7 


Color dark above; paler, yellowish or reddish [rust from tins?], below and on fins; 
an indistinct blackish stripe from eye to caudal. 

Two smaller specimens were taken with the type—“‘in a spring on 
the edge of a paddy field; cold water; very shy and active” (W. G.). 


I1I].—THE CHINESE SUCKER, MYXOCYPRINUS 

The peculiar Chinese sucker with big sail-like dorsal fin appears to 
be widely distributed in central China, but nowhere common. The type 
of its genus (Myzocyprinus Gill, 1878, Johnson’s ‘Cyclopedia,’ p. 1574) is 
Carpiodes asiaticus Bleeker, 1865, (?probably northern) China. A 
second species, Carpiodes chinensis Dabry de Thiersant, 1872, Yangtze; 
Sclerognathus chinensis Giinther, 1889, Ichang, Yangtze (named in- 
dependently), has since been described. A specimen from Anhwei is 
identified with the former form, which very likely actually came from the 
vicinity of Shanghai, and two specimens from Tungting, with the latter. 
The two forms appear to be racially, though not specifically, distinct. A 
tangible difference is to be found in the number of dorsal fin rays. A 
fourth individual to hand from Fukien does not agree with either of the 
above, and is here made the type of a third race. Unfortunately, it is too 
small for very satisfactory comparison. 


Myxocyprinus asiaticus asiaticus (Bleeker) 
Carpiodes asiaticus BLEEKER, 1865, Ned. Tijdschr. Dierk., II, p. 19. 
Description of a specimen from Anhwei, bought 20 miles from Ning- 
kwo, on a clear-water river on the way to Wuhu, by C. H. Pope, follows. 


Length to base of caudal, 220 mm. Depth in length, 2.4; head, 4.6. Eye in 
head, 6; snout, 2.3; interorbital, 2.2; maxillary, 2.3; width of mouth, 3; greatest 
width (the back of head), 1.5; depth of peduncle, 2.6; its length, 3; pectoral, 0.9; 
ventral, 0.9; longest dorsal ray, 0.6; longest anal ray, 1; lower caudal lobe, 0.8. 

Dorsal, 52; anal, 12. Scales, 53. 

Deep, compressed; lower surfaces broad, very slightly convex; pectorals and 
ventrals in a horizontal plane; head short and blunt; back before dorsal steep and 
very narrow. Interorbita] slightly convex, mouth inferior, transverse, surrounded 
with thick striate lips; maxillary not reaching to below front of eye; eye lateral, with 
a free rim; gill-membranes joined to side of breast behind the edge of the preopercle. 
Dorsal and anal without spinous rays, the former elongate, much elevated in front; 
dorsal origin equidistant from tip of snout and first third of depressed ventral; ventral 
origin well behind that of dorsal; pectoral about reaching ventral origin; ventral 
reaching three-fourths the distance to anal; caudal forked, with narrow, pointed lobes, 
the lower decidedly the longer. Scales with conspicuous, rather close-spaced radiating 
strie; lateral line complete, straight, in the center; pectoral and ventral without 
modified free axillary flap or scale. 


8 AMERICAN MUSEUM NOVITATES [No. 177 


Body and fins blackish; mouth to breast, a vague band from origin of dorsal to 
between pectoral and ventral bases, upper caudal lobe, and narrow inner margin of 
pectoral, pale; snout and opercle grayish. 


This race differs from M. a. chinensis (Giinther) from the Yangtze 
in having fewer dorsal rays, slightly lower body and dorsal lobe, and more 
striate scales. 


Myxocyprinus asiaticus chinensis (Dabry de Thiersant) 


Carpiodes chinensis DABRY DE THIERSANT, 1872, ‘Pisciculture en Chine,’ p. 182, 
Jeb ab, shh, Ale 
Sclerognathus chinensis GUNTHER, 1889, Ann. and Mag. Nat. Hist., IV, p. 223. 


Description of a specimen from Huping, Tungting Lake, Hunan, 
collected by C. H. Pope, follows. 


Length to base of caudal, 200 mm. Depth in length, 2.3; head, 4.2. Eye in 
head, 5.5; snout, 2.5; interorbital, 2; maxillary, 2.6; width of mouth, 3.9; greatest 
width (the back of head), 1.6; depth of peduncle, 2.6; its length, 2.8; pectoral, 
0.9; ventral, 0.9; longest dorsal ray, 0.6; longest anal ray, 1; lower caudal lobe, 
0.8. 

Dorsal, 57; anal, 14. Scales, 55. 

Deep, compressed; lower surfaces broad, very slightly convex; pectorals and 
ventrals in a horizontal plane; head short and blunt; hack before dorsal steep and 
very narrow. Interorbital convex; mouth inferior, transverse, surrounded with thick 
striate lips; maxillary not reaching to below front of eye; eye lateral, with a free rim; 
gill-membranes joined to side of breast behind edge of preopercle. Dorsal and anal 
without spinous rays; the former elongate, much elevated in front; dorsal origin 
equidistant from tip of snout and ventral origin; ventral origin well behind that of 
dorsal; pectoral passing ventral origin slightly; ventral extends % the distance to 
anal origin; caudal moderately forked, the lower lobe slightly the longer and the less 
pointed. Scales with a few radiating striz, less conspicuous than in typical asiaticus; 
lateral line complete, straight, in the center; pectoral and ventral without modified 
free axillary flap or scale. 

Body and fins blackish; mouth to pectoral bases, upper caudal lobe, narrow 
inner margin of pectoral, and opercle, pale; snout, and origin of dorsal to belly in 
front of ventrals, grayish. 


Myxocyprinus asiaticus fukiensis, new subspecies 


DESCRIPTION OF TypE.—No. 8415, American Museum of Natural History, col- 
lected near Yenping, Fukien, by H. R. Caldwell. 

Length to base of caudal, 36 mm. Depth in length, 2.6; head, 3.6. Tye in head, 
3.6; snout, 2.7; interorbital, 2.1; maxillary, 3.5; depth of peduncle, 3; its length, 
3.3; pectoral, 1.2; ventral, 1; longest dorsal ray, 0.7; longest anal ray, 1.4; lower 
caudal lobe, 0.9. 

Dorsal, 55; anal, 14. Scales, 47. 

Head rather broad and blunt; back steeply elevated to dorsal origin; body 
strongly compressed; breast flattish, so that the bases of pectorals (placed low) and 


1925] SOME CHINESE FRESH-WATER FISHES 9 


ventrals are in a horizontal plane; belly rounded. Top of head slightly convex; 
front of snout deep, vertical; mouth small, horizontal, transverse, curved, inferior, 
protractile downward; entirely surrounded with thick ridged-papillose lips; no 
barbels; maxillary not reaching under front of eye; orbit with a free rim, lateral 
placed high; gill-membranes adnate to breast near its center, about half of the 
diameter of eye apart. Dorsal high, not falcate (in this small specimen), occupying 
most of the back; its origin nearer tip of snout than to anal origin; pectorals reaching 
ventral origin; ventral passing anal origin; anal passing caudal base; caudal deeply 
forked, with narrow pointed lobes, the se er a little the longer. Lateral line complete, 
straight, in the center. 

Color black; anterior and posterior ends of head, a broad band slanting down and 
back from origin of dorsal and involving the inner edge of pectoral, a narrow, less 
perfect parallel band from middle of dorsal base, and caudal, pale. 


Pe tn Ws 
ye M i y ‘Nite At 


i Tro Ak 
Aha) A PES ter a oh 
iS a y ’ I Fai v re i ? 
te, a Paki : vn! 
7 yh 4 hoe as 5 - a 
iy sae : i 
Seah ay : One LP a cores 
Vt ige oF, OF are ee 
oi Ia aS ’ F se AU a ryan ratte 
mt _ ' mm R iy xi tM \ 
; : is , a < 2 ra 
4 ee | ae 4 2 an 


ty, : 4 


fd 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 179 Tue AmMERicAN Museum or NaTuRAL History June 23, 1925 
New York City d 


59.9,32C (51.7) 


HAMSTERS COLLECTED BY THE AMERICAN MUSEUM 
ASIATIC EXPEDITIONS! 


By Guover M. ALLEN 


The rodent family Cricetide is represented in eastern Asia by a 
number of small species collectively known as hamsters. All of these are 
burrowers, more or less modified for ground living, in contrast therefore 
to the long-tailed forest-living Peromyscus of North America (belonging 
to the same family), but recalling the American genus Onychomys, which 
again is an open-country or desert type. Hamsters are apparently absent 
in China south of about latitude 32° where, presumably, conditions have 
not in recent times been suitable for them. 

The Asiatic Expeditions under the leadership of Mr. Roy Chapman 
Andrews have now assembled a fine series of these handsome little mam- 
mals from various points in North China and the Gobi Desert, making 
possible a comparison of specimens from various parts of a species’ range 
and an estimate of the amount of geographic variation shown. As in the 
case of many other mammals, the dry interior of the country is inhabited 
by paler representatives, while nearer the coast in Chili, or along the 
damper borders of southern Shensi, darker coloration prevails. Some 
seem to be wholly confined to the Gobi. 


Cricetulus andersoni Thomas 

Cricetulus andersoni Tuomas, 1908, Proc. Zoél. Soc. London, p. 642. 

A small gray hamster with prominent ears, a tail of medium length, 
and white belly with slaty bases to the hairs. This was considered by 
Anderson to be the commonest small mammal in Shensi and Shansi, 
and it was found in abundance by Andrews at Kwei-hwa-ting and He- 
shuin in the latter province. In Mongolia a large series was also 
secured by Andrews at Artsa Bogdo, but not elsewhere except for a single 
skin at Ussuk and another forty miles south of Tsetsenwan. I am unable 
to see tangible differences between the Shansi and the Mongolian series, 


pean ese of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 52 


2 AMERICAN MUSEUM NOVITATES [No. 179 


though the latter are the least bit paler. A small series from Chili 
Province, northeast of Peking, is, however, much darker with a greater 
amount of black in the upper surfaces. It may be distinguished as follows. 


Cricetulus andersoni nigrescens, new subspecies 


Typre.—Adult male, skin and skull, No. 56307, American Museum of Natural 
History, from Province of Chili, 100 miles northeast of Peking, China. March 1922. 
Third Asiatic Expedition. 

DEscrRIPTION.—Similar to C. andersoni but the dorsal surface of head and body 
much darkened by black hairs. 

General color above, a buffy gray heavily lined with black particularly over the 
lower part of the back, with a tendency in some specimens to form an indistinct dark 
line medially. This general color is the result of a mixture of long, black-tipped hairs 
with the more numerous hairs that are slaty for their basal three-fourths, with a pale 
* buff tip. At the sides of the body between fore and hind legs, the latter hairs pre- 
dominate giving a nearly clear “light ochraceous buff.”” A small area behind each 
ear is similar. Upper side of tail dusky, with a few scattered white hairs. Ears 
blackish brown (‘‘ Natal brown”’), their tips narrowly edged with white. Feet and 
under side of tail pure white to the roots of the hairs; upper lips and entire under 
surface of the body and limbs white with a distinct buffy wash on the chest, the bases 
of the hairs slaty. 

The skull is not appreciably different from that of typical andersoni. 

MEASUREMENTS.—None of the series of topotypes is accompanied by measure- 
ments, but the specimens do not seem to differ from typical andersoni in size except 
that the tail appears to be slightly shorter, about 30 instead of 88 mm. 

The skull of the type measures: greatest length, 26 mm.; basal length, 22.5; 
palatal length, 12.5; diastema, 7.0; width of braincase, 11.7; interorbital width, 
4.0; upper cheek teeth, 4.1; mandible, 14.3; lower cheek teeth, 4.0. 


The series of twenty-three skins from 100 miles northeast of Peking 
is so very much darker than typical andersoni from near Taiyuenfu, 
central Shansi, as to be easily distinguishable on comparison. A series 
from northern Shansi, at Kwei-hwa-ting, is exactly intermediate between 
the pale typical animal and the Chili form, so that the subspecific stand- 
ing of the latter is unquestionable. In addition to its darker back, the 
buffy wash on the chest seems distinctive. 


Cricetulus griseus (A. Milne-Edwards) 
Cricetus (Cricetulus) griseus A. M1ttnE-Epwarps, 1868-1874, ‘Recherches Hist. 
Nat. Mammiféres,’ p. 133; Pl. x11, fig. 1; Pl. x11, fig. 1-1h. 

A short-tailed buffy species, with a narrow black stripe in the middle 
of the back. This is common in parts of northeastern China. A series 
from Peking are topotypes and a large number were secured at Shan-hai- 
kwan in the same province (Chili). The collection also includes a 


1925] ASIATIC HAMSTERS 3 


number from Chimo, on the north coast of the Shantung peninsula 
collected by Paul D. Bergen. These are all very uniform in color, and 
the young are but little darker. 


Cricetulus griseus obscurus (A. Milne-Edwards) 
Cricetus (Crécetulus) obscurus A. MILNE-Epwarps, 1868-1874, ‘Recherches Hist. 
Nat. Mammiféres,’ p. 136; Pl. x11, fig. 2; Pl. x11, fig. 2-2c. 

This is similar to the preceding but paler. Milne-Edwards’ type 
came from Saratsi, northern Shansi, so that Andrews’ series from Kwei- 
hwa-ting in that province are practically topotypes. These already 
show a sandier, paler tint than those to the eastward, though occasional 
specimens are hardly distinguishable. Following Thomas’s suggestion, 
however, the name may be retained in a subspecific sense for the pale 
western animal of Shansi and Mongolia. In the latter region this 
hamster was obtained at Turin, Loh, Ussuk, Gun Burta, Tsetsenwan, 
Tsagan Nor, and Sain Noin Khan. Compared with the Kwei-hwa-ting 
series these are a perceptible shade paler, with in some specimens pro- 
nounced white tufts at the exterior base of the ears, and the dark line 
on the back sometimes less sharply defined. Possibly this race will be 
found to merge with kozlovi from Sa-chou. 


Cricetulus longicaudatus (A. Milne-Edwards) 

Cricetus (Cricetulus) longicaudatus A. M1tNE-Epwarps, 1868-1874,‘ Recherches 
Hist. Nat. Mammiféres,’ p. 136; Pl. x11, fig. 3; Pl. x111, fig. 3-3a. 

Resembles C. andersoni but the hairs of much of the lower surfaces 
are white to their bases. This seems to be rare or local, and was not 
secured except at Artsa Bogdo, Mongolia, where a single male was taken 
that seems referable to the species. Its vibrissz are noticeably longer and 
more abundant than in C. andersoni, which apparently greatly out- 
numbers it .at the same locality; the feet also are larger, and the skull 
has a more elongate rostrum. 


Cricetulus migratorius curtatus, new subspecies 


Type.—Adult male, skin and skull, No. 57873, American Museum of Natural 
History, from Iren Dabasu, Mongolia. May 2, 1922. Third Asiatic Expedition. 

Description.—A medium-sized hamster, uniform buffy gray above, feet and 
tail white, the latter very short, not exceeding the extended hind foot; belly white, 
the hairs with slaty bases except on skin, throat, forearms, and tail. 

General color of the upper parts from the nose to root of tail, and laterally as 
far as the vibrisse, cheeks, shoulder and lower thigh nearly “cinnamon buff” (Ridg- 
way, 1912) faintly and evenly lined with fine black-tipped hairs. Half-way between 


4 AMERICAN MUSEUM NOVITATES [No. 179 


eye and ear an indistinct grayish-white bar extends upward from the white of the 
throat across the cheek on each side. A small tuft of white hairs is present at the 
anterior base of the ear, and an ill-defined pale buffy patch marks the posterior base. 
Outer surface of ears scarcely darker than the back, thinly covered with short whitish 
and dusky brown hairs. Lips (including the bases of the vibrissze), the entire forearm 
and foot, and the hind leg from the lower part of the thigh, the tail, the sides and belly 
white, the hairs with slaty bases except on chin, upper throat, a narrow median area 
between the forelegs, the entire forearms, fore and hind feet, and the tail, which are 
white to the bases of the hairs. Vibrissee and a narrow eye-ring black. 

Immature specimens are “drab-gray”’ to ‘light drab’’ above with less of the 
pale cinnamon tint. The tail, though usually entirely white, may have a narrow 
line of scattered dark or blackish hairs mid-dorsally. 

MEASUREMENTS.—The type was measured by the collector as follows: head 
and body, 115 mm.; tail, 18; hind foot [s. u.], 17; ear, 16. The hind foot with claw 
measures on the dry skin, 19mm. The largest specimen has a head-and-body length 
of 128 mm. 

The skull of the type measures: greatest length, 32.6 mm.; basal length, 31; 
palatal length, 17; diastema, 9.5; zygomatic width, 18; mastoid width, 13.5; width 
outside anterior molars, 7; upper cheek teeth, 4.9; lower cheek teeth, 4.8; mandible 
to condyle, 20. , 

This appears to be one of the C. migratorius group, the center of whose 
range is western Asia to the borders of western Europe, and is apparently 
another of the species of that area to have made its way through the 
Altai region into the eastern Gobi Desert. In addition to being appar- 
ently the most eastern member of the mzgratorius group yet discovered, 
it is also the shortest-tailed.. The tail is stumpy and conical, barely 
reaching the tip of the extended hind foot, the hairs at its base notice- 
ably longer than those at its tip. The pale buffy or cinnamon color is 
indicative of the dry desert habitat. In addition to specimens from the 
type locality, others apparently indistinguishable were taken at the 
following places in the Gobi Desert, Mongolia: Gun Burte (6800 feet: 
altitude), Pang Kiang, Tsagan Nor, Ussuk, Loh, Turin. 


Cricetulus triton (de Winton) 

Cricetus (Cricetulus) triton DE WINTON, 1899, Proc. Zoél. Soc. London, p. 575. 

Ten skins from Chimo, near the coast of the Shantung peninsula, 
are practically topotypes of the large hamster of northeastern China, 
and are very uniform in their general buffy-gray appearance, slightly 
darker in the middle of the back. The largest of the series, though un- 
accompanied by collector’s measurements, are apparently of nearly the 
same size as C. nestor Thomas from northeast of Seoul, Korea, and the 
two are doubtless closely related. Thomas has lately described a third 
race, a pale gray subspecies, C. ¢t. encanus, from the dry inland country 


1925] ASIATIC HAMSTERS 5 


along the edge of the Ordos Desert. The collections obtained by the 
Third Asiatic Expedition include a small series from near Peking, which 
differ uniformly from the Shantung specimens in having the base of the 
tarsus dark instead of white like the rest of the foot, while still another 
lot frcm the hills of southern Shansi and Shensi are the opposite of C. t. 
incanus in their darkened coloration correlated with the moister climate 
of this area. The descriptions of these two races follow. 


Cricetulus triton fuscipes, new subspecies 


Typr.—Adult male, skin and skull, No. 56792, American Museum of Natural 
History, from Peking, Chili Province, China. 1921. Third Asiatic Expedition. 

DescrieTion.—Similar in general appearance to C. triton, but the ankles and 
basal part of the metatarsals dusky, instead of white like the distal part of the foot. 

Entire dorsal surface of head and body a nearly uniform buffy, slightly darkened 
on the lower back by a greater admixture of long black hairs. Along the sides of the 
cheeks and flanks these black hairs are few or absent so that the color is here clearer, 
nearly “light ochraceous buff.” The individual hairs of the back are of two sorts: 
some entirely black, others with a fine black tip, then a broad subterminal ring of 
“light ochraceous buff’? and a slaty hase. The ears are clothed with short brownish 
(“fuscous’’) hairs, with which are mingled a few grayish hairs, these latter more 
abundant at the extreme tip of the ear. The chin and a varying median area extend- 
ing back from it, the hands and wrists are pure white to the roots of the hairs. The 
under side of the body and limbs is elsewhere white with the slaty-gray bases of the 
hairs showing through. The hind feet are white on the distal part of the metatarsals 
and on the toes, but the proximal third or half of the metatarsal area is contrastingly 
brownish (‘‘fuscous’’). The tail is thinly haired, dusky above and whitish below. 

The skull is stoutly built, the orbits square-edged above, giving a less rounded 
appearance than in the smaller species of the genus. 

M®raAsuREMENTS.—The skins are unaccompanied by measurements, but the hind 
foot of the type measures 23 mm., or practically as in typical C. triton. As made up, 
the tails of the Peking series seem longer than those from the Shantung peninsula. 

The following skull measurements are of the type and a topotype (No. 56335): 
greatest length, 35+, 38 mm.; basal length, 35, 34.5; palatal length, 18.5, 18.8; 
diastema, 10.5, 10.7; zygomatic width,—, 19.4; width of braincase, 15.5, 15.4; 
upper cheek teeth, 5.5, 5.5; lower cheek teeth, 5.4, 5.5; mandible, 22, 21.8. 


Seven specimens from Peking agree in having the basal portion of 
the hind foot dusky brown, whereas in the other races at hand it is 
entirely white. In other respects this race is not obviously different 
from typical C. triton unless the longer tails of the skins as prepared are 
not a result of difference in preparing the specimens. 


Cricetulus triton collinus, new subspecies 
Tyrr.—Adult female, skin and skull, No. 56389, American Museum of Natural 
History, from the base of Tai-pei-shan, Tsing-ling Mountains, Shensi Province, 
China. October 10, 1921. Third Asiatic Expedition. 


6 AMERICAN MUSEUM NOVITATES [No. 179 


Description.—A large hamster, similar to C. triton but much darker and with 
slightly longer tail. 

General color of the upper parts from nose to tail between “‘drab”’ and ‘‘mouse- 
gray,” the individual hairs either entirely black, or with minute black tip and a 
broad subterminal band of “warm buff,” the latter predominating at the sides of 
the head and body, so that these portions are slightly brighter. Inner and outer 
sides of ears thinly covered with short blackish-brown hairs, except that the extreme 
edge is white. Feet and wrists, the chin, and a small median spot on the throat, clear 
white to the bases of the hairs; the lower surfaces elsewhere are clothed with whitish 
hairs whose slaty-gray bases everywhere show through and darken the general grayish 
tone. Tail thinly covered with short appressed hairs, blackish brown (near ‘‘sepia”’), 
with many whitish hairs on the lower side. Vibrissee blackish brown or whitish, short. 

The skull hardly differs from that of C. t. fuscipes, except that the incisive fora- 
mina in adults tend to be longer, reaching about to the level of the cheek teeth. The 
interparietal in the latter also differs in that the anterior corners are produced forward 
so that the border is brace-shaped instead of V-shaped. 

MrasoreMENTS.—The collector’s measurements of the type are: head and body, 
155 mm.; tail, 72; ear, 22; the hind foot measures 24 mm. ~ 

The skull of a topotype (No. 56388) measures: greatest length, 37.5 mm.; basal 
length, 34.6; palatal length, 17.6; diastema, 10.2; zygomatic width, 19; width of 
brainease, 15.3; interorbital width, 4.7; upper cheek teeth, 5.2; mandible, 21.6; 
lower cheek teeth, 5.2. 


The series of nine specimens from the base of Tai-pei-shan, Tsing- 
ling Mountains, comes from the northwestern border of the damp forested 
area of which Szechwan may be considered the center. The saturate or 
darkened appearance of this form of C. triton is therefore what might be 
expected in contrast to the very pale C. triton incanus from the edge of 
the Ordos Desert, or the pale-buffy C. triton from Shantung. A single 
specimen from the northern edge of the Tsing-ling range, about 45 miles 
south of Fengsiangfu, Shensi, is also referable to C. t. collinus, and prob- 
ably represents nearly its northern limit. The altitude of this point is 
about 3600 feet. Two other similar specimens were taken at He-shuin, 
in southern Shansi. The discovery of a representative of C. triton in 
this southern part of Shensi is apparently a considerable extension of 
its known range, and it seems unlikely that it will be found to penetrate 
much farther southwestward. 


Cricetiscus campbelli (Thomas) 
Cricetulus campbelli THomas, 1905, Ann. Mag. Nat. Hist., (7) XV, p. 322. 
For these dwarf hamsters with shortened tail, black dorsal line, 
and with the white of the lower side extending up on the flanks to form a 
convex area on neck, body, and hip, Thomas has lately proposed the 
generic name Cricetiscus. The skull differs from that of Cricetulus in its 


1925] ASIATIC HAMSTERS 7 


more spreading zygomata and in having the dorsal profile distinctly more 
bowed. The type of campbelli was from northern Chili near the southern 
edge of the Mongolian plateau. The Second and Third Asiatic Expedi- 
tions, under Mr. Roy Chapman Andrews, secured specimens at various 
places on the great plateau, namely, at Urga, and at localities along the 
caravan route 60, 80, 120, and 140 miles to the southeast, as well as at 
Ussuk, Turin, 30 miles to northeast of Tsetsenwan, and 40 miles south- 
west of there, as well as a single one on the Tola River 80 miles west of 
Urga. It seems to be a characteristic species of the tableland, but is 
represented by a similar and darker form in the Altai region, ‘‘ Phodopus”’ 
(=Cricetiscus) ‘crepidatus Hollister. 


Phodopus bedfordie (Thomas) 

Cricetulus bedfordix THomas, 1908, December 22, Abstract Proc. Zodl. Soe. 
London, p. 45. 

A short-tailed dwarf hamster, pale buff above, pure white below. 
This was described from Yulinfu, Shensi, on the edge of the Ordos Desert. 
Three specimens were secured by the Third Asiatic Expedition at 
Tsagan Nor, and a fourth 160 miles southeast of Sain Usu, Mongolia, 
thus considerably extending its known range. These four do not differ 
appreciably from a topotype in the Museum of Comparative Zodlogy. 


a oo ‘oye 
apie i 


ry wo i 
sain 


id ny" sae 


wesare 


; volt u i “ yee iu, 7 
; | : dpe. \ ie ae 
EG >, "i ii aah a os tee’ si os ney A ‘ ‘‘ 4 vray sealb ary 
. ree, he Pe ih Oa Ua AS c “ae 
/ f has (ve 
A h ¥ » ish 
" ‘ena f iy Sri at ct a P| 1 f vi 
; , Ay ‘ 
f, a i) pra ; 7 ey , } ‘ n'y 
’ Pi. ; a 

gripe ast aL Vieiyd it 


Pi oe rm ; ay Niy® ie 
yt wb) Us a a fy) vO hap ' or 
q \ os tes * wi ats +4) ve 73 bigna os At sable 


¢ ; e ‘ Mie kh ait _ a A ye enc 
O 


J 
is FT 
7 2Vety 
. Y . gS ag 
r a 
* 
‘ a 
ia L 
: 
iS 5) 
‘ vi 
Xe 
= 
i " 
; 
‘ 
’ 
' : 
8s we 
i i 
, i 
” 
? - 
wey 
Sar ‘ 
r . 
: 4 
i 
ft 
. 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 181 Tue AMERICAN Museum or Natura History July 16, 1925 
. New Yorx City 


59.7 (51) 


SOME CHINESE FRESH-WATER FISHES! 


By J. T. NicHots 


IV.—GUDGEONS OF THE GENUS CORIPAREIUS 


The genus Coripareius is based by Garman, 1912, on Labeo cetopsis 
Kner, 1867, from Shanghai. Some of Garman’s material from the 
Yang-tze which he identified with cetopsis was probably referable to 
Pseudogobio styani Giinther, 1889. We have no material referable to 
cetopsis, which apparently differs from styani in having the vent half- 
way between ventral and anal (from description), the dorsal origin 
equidistant from end of snout and front of anal (from figure). 

This genus is characterized by a single pair of exceptionally long 
barbels, small eye a little before the center of the head, long compressed 
peduncle, rather small scales, and a single row of anvil-shaped teeth. 
Its distinguishing characters are perhaps less tangible than those of such 
specialized gudgeons as Pseudogobio, Rhinogobio, and Saurogobio; but 
the writer was of the impression, before finding that it had already been 
named by Garman, that it is actually less, rather than more, closely re- 
lated to Gobzo than are these. Gobio uranoscopus caucasicus as figured by 
Berg, 1916, is very suggestive of Coripareius. Recent collections of the 
Third Asiatic Expedition contain two species as follows. 


Coripareius styani (Giinther) 
Pseudogobio styani GUNTHER, 1889, Ann. and Mag. Nat. Hist., (6) IV, p. 224. 


The following is a description of a specimen from Huping, Tungting 


Lake, Hunan, collected by C. H. Pope. 

Length to base of caudal, 74mm. Depthin length, 5; head, 4.1. Eye in head, 6; 
snout, 2.5; interorbital, 2.6; maxillary, 3.5; barbel, 2; width of body, 1.6; depth of 
peduncle, 2.2; its length, 1.2; pectoral, 1.1; ventral, 1.4; longest dorsal ray, 1; 
longest anal ray, 1.5; caudal lobe, 0.9; distance between gill-slits, 5.5. 

Dorsal, 9; anal, 8. Scales, 56. 

Little compressed before the dorsal, compressed behind it; lower surface of head 
flattish, the upper evenly convex; front of back narrow, breast broad and gently 
rounded; vent a short distance before anal origin. Snout bluntly pointed, extending 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 53. 


2 AMERICAN MUSEUM NOVITATES [No. 181 


a little beyond the small inferior horizontal curved mouth; lips rather thick and 
smooth; maxillary not nearly to eye; with a long cylindrical tapering roughened 
subterminal barbel; gill-membranes rather broadly joined to isthmus under edge of 
preopercle. Dorsal origin equidistant from tip of snout and middle of last anal ray; 
dorsal and anal without spinous rays; ventral under middle of dorsal; pectoral not 
quite reaching to ventral, ventral not nearly to anal; caudal well forked with pointed 
lobes; its rudimentary rays precurrent in a short low keel above and below. Scales 
with rather wide-spaced horizontal parallel striz; lateral line complete, straight, in 
the center; an elongate pointed axillary ventral flap with 1 or 2 covert scales. 
Color dull, uniform; the front and tip of the dorsal dusky. 


Coripareius septentrionalis, new species 


DEscRIPTION OF TyPpE.—Number 8416, American Museum of Natural History, 
from Paotow, Mongolia, on the Yellow River a little west of the Shansi boundary; 
collected by C.:-H. Pope. 

Length to base of caudal 240 mm. Depth in length, 4.5; head, 4.8. Eye in 
head, 8.6; snout, 2.8; interorbital, 3; maxillary, 3; barbel, 1.8; distance between 
gill-clefts, 8.6; greatest width (at shoulder), 1.4; depth of peduncle, 2; its length, 1; 
pectoral, 1; ventral, 1.4; longest dorsal ray, 1.1; longest anal ray, 1.4; upper caudal 
lobe, 0.9. 

Dorsal, 9; anal, 8. Scales, 55 (last three on caudal base). 

Somewhat compressed; snout and anterior part of head depressed; the head 
narrowing forward; front of back elevated; breast broad; vent five times as far from 
ventral axil as from anal origin. Mouth inferior, transverse, horizontal, horseshoe- 
shaped; the bluntly pointed snout extending a little beyond it; lower jaw squarely 
truncate at the end; with free lip only at the side, the same confluent with maxillary 
and lip of upper jaw; lips and chin granular with non-prominent papille; maxillary 
not nearly to below eye; a subterminal, long, tapering, thick-based barbel which 
reaches to slightly beyond edge of preopercle; interorbital flattish across the top, 
dropping at the sides to the small eye which is appreciably below the profile and 
almost strictly lateral; gill-membranes joined to isthmus before edge of preopercle. 
Dorsal and anal without spinous rays; dorsal origin equidistant from tip of snout 
and middle of anal base; pectoral not reaching ventral; ventral about half-way to 
anal; free edges of dorsal and anal concave; caudal forked, the upper lobe a little 
the longer. Body fully scaled; scales with conspicuous close-set parallel horizontal 
striz; lateral line straight, in the center, complete; a bluntly pointed ventral axillary 
flap scaled on its base. 

Color rather dark; the border of dorsal, margin of caudal and center of pectoral 
more or less blackish. 


This species is close to C. styant, than which it probably reaches a 
larger size, and, if it were not for the narrower head and more anterior 
dorsal origin, might be taken for an overgrown example of styani. 


V.—GUDGEONS RELATED TO THE EUROPEAN GOBIO GOBIO 


Certain of the numerous tribe of Chinese gudgeons are closely re- 
lated to the typical species, Gobio gobio of Europe. In fact, Fowler, 1924, 


1925] SOME CHINESE FRESH-WATER FISHES 3 


synonymizes a form from Chihli with it. In the writer’s opinion this 
fish is slightly different, though very likely identical with Gobio gobio 
soldatovi Berg, 1914 and 1916. A small fish from Shansi, again, would 
seem to be subspecifically separable from it, and another with exception- 
ally long barbel from the same province represents a well-marked 
species. Descriptions of these three forms follow. 


_Gobio soldatovi Berg 


Gobio gobio var. soldatovi Bera, 1914, ‘Faune Russ.,’ Poiss. III, part 2, p. 461, 
Fig. 63. 


The following is a description of a specimen from Hsing-lung-shan, 
Eastern Tombs, Chihli, August 7, 1921; collected by C. H. Pope. 


Length to base of caudal, 76 mm. Depth in length, 4.6; head, 3.4. Eye in 
head, 4; snout, 3; interorbital, 3.8; maxillary, 3.3; width of mouth, 3.8; barbel, 
3.8; greatest width of body (the back of head), 2.2; depth of peduncle, 2.7; its 
length, 1.4; pectoral, 1.5; ventral, 1.8; longest dorsal ray, 1.5; longest anal ray, 
2; caudal lobe, 1.2. 

Dorsal, 9; anal, 8. Scales, 40. 

Little compressed; snout bluntly pointed; vent at one-quarter the distance 
from anal origin to ventral axil. Eye very slightly superolateral; interorbital 
slightly concave; a slight groove across snout; mouth slightly oblique, inferior, 
horseshoe-shaped; maxillary not quite to under front of eye; with a well-developed 
subterminal barbel; lips moderately thick, not free across chin; gill-membranes joined 
to side of breast under edge of preopercle. Dorsal and anal without spinous rays; 
dorsal origin slightly nearer tip of snout than base of caudal; ventral origin beneath 
about the center of dorsal base; pectoral rounded, reaching three-quarters the dis- 
tance to ventral; ventral not quite to anal; caudal moderately forked, the upper 
lobe the more pointed and slightly the longer (another specimen with lower lobe 
the longer). Scales with conspicuous close-set radiating strize; scales absent from 
breast before pectoral axil; lateral line complete, in the center, rising a little to 
meet opercle. 

Top of head dark; lower surfaces pale; a blackish longitudinal stripe, faint and 
ill-defined in front; also ill-defined dark longitudinal streaks on back and sides; a 
black blotch over upper pectoral axil; a faint blackish blotch in the stripe near end 
of peduncle; dorsal and caudal lightly barred; pectoral with a couple of faint bars. 


Gobio soldatovi minulus, new subspecies 


DESCRIPTION OF THE TypE.—Number 8417, American Museum of Natural 
History, from Kweihwa, Shangi; collected by C. H. Pope. 

Length to base of caudal, 57 mm. Depth in length, 4.6; head, 3.5. Eye in head, 
3.8; snout, 3; interorbital, 3.5; maxillary, 3.5; depth of peduncle, 3; length o 
peduncle, 1.4; pectoral, 1.3; ventral, 1.6; longest dorsal ray, 1.5; longest anal ray, 
1.7; caudal lobe, 1. Barbel in eye, 1.3. 

Dorsal, 94; anal, 8. Scales, 41. Teeth (from a cotype), 5,2. 


4 AMERICAN MUSEUM NOVITATES [No. 181 


Slender, moderately compressed, broadest at the back of the head; eyeplaced 
high, almost strictly lateral, very little superolateral; the interorbital slightly concave 
due to orbital rim being slightly raised; profile rounding downward before the eye, 
with a shallow dent in front of nostril; breast and belly flattish but not so much so 
that either pectorals or ventrals lie in a horizontal plane; vent a little nearer anal 
origin than to ventral axil. Mouth horizontal, slightly inferior, semicircular; with 
thick, slightly and finely papillose lips, grooves behind sides of lower lip not quite 
meeting at the chin; maxillary not quite to front of eye, with a well-developed sub- 
terminal barbel; gill-membranes adnate to breast under edge of preopercle. Dorsal 
and anal without spinuous rays; center of dorsal base equidistant from front of eye 
and base of caudal; ventral base under center of that of dorsal; pectoral and ventral 
rounded; pectoral not quite reaching ventral, ventral not quite reaching anal; 
caudal shallowly forked; rudimentary caudal rays precurrent as short low keels above 
and below on peduncle. Lateral line complete, dipping slightly in front, straight and 
in center from behind dorsal axil; breast scaleless. 

About seven blotches along side with a tendency to be confluent; somewhat 
dark above, a dark stripe from eye toward snout; lower surfaces and lower fins pale; 
dorsal and caudal faintly barred. 


Gobio coriparoides, new species 


DESCRIPTION OF THE TypE.—Number 8418, American Museum of Natural 
History, from the vicinity of Ningwu, Kolan and Tsinglo, Shansi; collected by C. H. 
Pope. 

Length to base of caudal, 77 mm. Depth in length, 4.6; head, 3.5. Eye in 
head, 4.5; snout, 3; interorbital, 3; maxillary, 3.3; barbel, 2.3; width of mouth, 
3.4; greatest width (the back of head), 1.8; depth of peduncle, 2.5; its length, 1.2; 
pectoral, 1.2; ventral, 1.5; longest dorsal ray, 1.2; longest anal ray, 1.4; lower 
caudal lobe, 1. 

Dorsal, 944; anal, 8. Scales, 42. Teeth, 5, 2. 

Moderately compressed; head below, breast and belly flattened; ventrals (and 
pectorals to a less extent) in a horizontal plane; vent at one-quater the distance from 
anal origin to ventral axil. Interorbital flattish; mouth inferior, horizontal, horseshoe- 
shaped; lower jaw included; maxillary not quite to under front of eye, with a long 
terminal barbel; eye almost strictly lateral; gill-membranes joined to side of breast 
below edge of preopercle. Dorsal and anal without spinous rays; dorsal origin equi- 
distant from tip of snout and tip of depressed anal; ventral origin behind that of 
dorsal; pectoral passing ventral origin; ventral reaching to vent; anal falcate; - 
caudal forked, the lower lobe the longer. Scales with conspicuous rather close-spaced 
radiating striew; breast scaleless anterior to pectoral axil; lateral line complete, 
straight, in the center, rising a little to meet opercle at extreme front end. 

Color rather dark, paler below; a faint dark longitudinal band on the peduncle. 


From general appearance, the writer would be tempted to place this 
species in Coripareius, were its teeth not typical of Gobzo. 

There are two other somewhat aberrant, apparently undescribed 
species of the genus Gobo. One from Tungting suggests certain species 
of Rhinogobio; the other, from Shansi, has a form suggestive of Sauro- 


1925] SOME CHINESE FRESH-WATER FISHES 5 


gobio, color pattern of Pseudogobio but lips restricted, versus unusually 
developed as in these two specialized genera. 


Gobio longipinnis, new species 

?Rhinogobio ventralis SAtuvaGE AND DaBry DE THIERSANT, 1874, Ann. Sci. 
Nat., (6) I, Art. 5, p. 11. 

DEscripTION OF THE TypE.—Number 8419, American Museum of Natural 
History, from Huping, Tungting Lake, Hunan; collected by C. H. Pope. 

Length to base of caudal, 95 mm. Depth in length, 4.7; head, 3.9. Eye in 
head, 7; snout, 2.1; interorbital, 3; maxillary, 2.4; width of mouth, 3.5; barbel, 7; 
greatest width of body (at back of head), 1.7; depth of peduncle, 2.2; its length, 1.4; 
pectoral, 1.1; ventral, 1.2; longest dorsal ray, 0.9; longest anal ray, 1.2; lower 
caudal lobe, 0.8; distance between gill-clefts, 3. 

Dorsal, 10; anal, 9. Scales, 52. Teeth, 5, 2. 4 

Moderately compressed; snout pointed; vent at three-eighths the distance from 
anal origin to ventral axil. Interorbital convex, striate with rows of small tough 
prominences; mouth at a distance behind tip of snout, one and one-half times eye, 
inferior, transverse, horizontal, forming a broad subtruncate curve; maxillary reach- 
ing a little short of front border of eye; with a terminal barbel; eye somewhat supero- 
lateral; gill-membranes broadly joined to side of breast behind edge of preopercle. 
Dorsal and anal without spinous rays; dorsal origin equidistant from tip of snout 
and tip of last anal ray; ventral origin slightly behind that of dorsal; pectoral 
passing ventral origin; ventral reaching to anal origin; anal three-fourths the dis- 
tance to base of caudal; caudal well forked, lower lobe a little the longer. Scales 
with well-spaced horizontal striz#; small scales present on breast to line between 
gill-clefts; lateral line complete, straight, in the center, rising a very little to meet 
opercle at its extreme front end. 

Color uniform. 


Gobio rivuloides, new species 


DESCRIPTION OF THE TyprE.—Number 8420, American Museum of Natural 
History, from Niang-tze-kwan, Shansi; collected by C. H. Pope. 

Length to base of caudal, 133 mm. Depth in length, 5.5; head, 3.7. Eye in 
head, 6; snout, 2.3; interorbital, 3.8; maxillary, 2.6; width of mouth, 2.7; barbel, 
2.6; width between gill-openings, 6; depth of peduncle, 2.8; its length, 1.4; pec- 
toral, 1.3; ventral, 1.4; longest dorsal ray, 1.4; longest anal ray, 1.8; upper caudal 
lobe, 1.2. 

Dorsal, 915; anal, 9. Scales, 42. Teeth (from a cotype), 5, 3. 

Head flat-topped; snout steep; body cylindrical; peduncle compressed; the 
back not elevated; belly flattish, ventrals in a horizontal plane; vent about equi- 
distant from ventral axil and anal origin. Tip of snout soft, somewhat swollen, 
papillose; mouth inferior, semicircular, horizontal, with thick papillose lips produced 
backward at the corners and stopping at the sides of the lower jaw; maxillary not 
reaching eye; with a single well-developed subterminal barbel; eye superolateral; 
gill-membranes joined to the isthmus under edge of preopercle. Middle of dorsal base 
equidistant from nostril and base of caudal; the ventral origin slightly in advance 
thereof; pectoral reaching more than two-thirds the distance to ventral; ventral 


6 AMERICAN MUSEUM NOVITATES [No. 181 


more than three-quarters distance to anal; caudal forked, upper lobe the longer (in 
another specimen it is the shorter) and more pointed. Scales with rather close-spaced 
radiating strie; breast naked, the scaleless area ending in a blunt point nearer 
isthmus than ventrals. 

Color dark, paler along the belly, about ten vague squarish unequal blackish 
blotches along lateral line; two blackish stripes radiating from the lower border of 
the eye to the mouth; caudal crossed by rows of fine black spots; dorsal with a few 
such rows, and pectoral with two or three faint ones. 


VL—NEW GUDGEONS OF THE GENERA GNATHOPOGON AND 
LEUCOGOBIO 

The recognition of Gnathopogon Bleeker, 1860, and Leucogobio 
Gunther, 1896, genera of minnow-like and chublike gudgeons, as distinct 
from Gobio, is largely a matter of convenience. There seem to be several 
well-marked species of each in the Chinese fish-fauna. Recent authors 
sometimes synonymize these genera with Gobio (Leucogobio chankaénsis 
and txniatus,—Berg, 1916: Gobio [Gnathopogon] argentatus and wolter- 
storffi,—Fowler, 1924) while recognizing Paraleucogobio Berg, 1907 (in 
the writers’ opinion only subgenerically distinct from Leucogobio, in a 
single relative character of the last simple dorsal ray) as a full genus. 
Jordan, 1920, unites Leucogobio with Gnathopogon. They seem to be 
about equally different from one another and from Gobio, but the type of 
Gnathopogon (Capoéta elongata, Temminck and Schlegel, 1844) approaches 
Leucogobio more closely than do most of the species referable to Gnathopo- 
gon. Squalidus chankaénsis Dybowski, 1872, may be an aberrant species 
of Leucogobio as here understood, in which case Leucogobio becomes a 
synonym of Squalidus Dybowski. 


Leucogobio polytenia, new species 


DESCRIPTION OF THE TypE.—Number 8421, American Museum of Natural 
History, from Niang-tze-kwan, Shansi; collected by C. H. Pope. 

Length to base of caudal, 76 mm. Depth in length, 3.7; head, 3.7. Eye in 
head, 4; snout, 3.5; interorbital, 2.8; maxillary, 3.5; depth of peduncle, 1.9; its 
length, 1.3; pectoral, 1.3; ventral, 1.5; longest dorsal ray, 1.5; longest anal ray, 1.6; 
caudal lobe, 1.3. Barbel in eye, 2. 

Dorsal, 914; anal, 8. Scales, 39. Teeth (from a cotype), 5, 3. 

Head and snout blunt; body little compressed; peduncle compressed; breast 
broad and rather flat. Jaws equal; mouth small, oblique, transverse; maxillary not 
reaching eye; with a small, slender subterminal barbel, the base of which is concealed 
behind the broad upper lip; tip of lower jaw squarish; gill-membranes narrowly 
joined in a fold across the isthmus. No spinous dorsal or anal rays; dorsal origin 
over that of ventral, slightly nearer tip of snout than base of caudal; pectoral 
rounded, not reaching ventral; ventral not quite to anal; caudal moderately forked, 


1925] SOME CHINESE FRESH-WATER FISHES 7 


the upper lobe the more pointed and the two equal. Scales rather rough, with close- 
set subparallel strie; lateral line complete, straight, in the middle of side. Vent at 
tips of ventrals, three-quarters the distance from their axils to origin of anal. 
Blackish above; the color most solid along mid-side; with indistinct pale streaks 
along back; side below lateral line with alternate dark and silvery streaks of about 
equal width, and belly paler; lower half of face silvery. Dorsal with a black spot on 
its front rays, continued backward as a faint shade across the fin; other fins plain. 


Leucogobio tzniellus, new species 


DESCRIPTION OF THE TypE.—Number 8422, American Museum of Natural 
History, from streams in the Min River Basin, near Yenping, Fukien, 1920; collected 
by H. R. Caldwell. 

Length to base of caudal, 55 mm. Depth in length, 3.7; head, 3.8. Eye in 
head, 3.5; snout, 3.5; interorbital, 3; maxillary, 3.3; depth of peduncle, 1.8; 
length of peduncle, 1.7; pectoral, 1.4; ventral, 1.6; longest dorsal ray, 1.6; longest 
anal ray, 1.6; caudal lobe, 1.3. Barbel in eye, 1.5. 

Dorsal, 9; anal, 8. Scales, 36. 

Moderately compressed; the head rather blunt; vent a little in advance of anal 
origin. Lower jaw slightly included; mouth moderately oblique; maxillary not reach- 
ing front of eye; with a moderate subterminal barbel; gill-membranes narrowly joined 
to isthmus under edge of preopercle. No spinous dorsal or anal rays; dorsal origin 
equidistant from tip of snout and base of caudal; ventral origin under that of dorsal; 
pectoral and ventral rounded; pectoral not quite reaching to ventral, ventral not 
quite to anal; caudal little forked. Scales with close-set slightly radiating strix; 
lateral line complete, in the middle of the side except for a slight rise in front. 

Irregularly dark along middle of side; along back; and in one or two lines 
between. Front of dorsal with a black mark, which is faintly indicated across the fin. 


Gnathopogon punctatus, new species 


DESCRIPTION OF THE TypE.—Number 8423, American Museum of Natural 
History, from near Yenping, Fukien; collected by H. R. Caldwell. Close to Gnatho- 
pogon wolterstorffii (Regan). 

Length to base of caudal, 46 mm. Depth in length, 4.5; head, 3.9. Eye in 
head, 3; snout 3.4; interorbital, 4; maxillary, 3.6; barbel, 4; greatest width of body 
(at back of head), 2; depth of peduncle, 2.7; its length, 1.6; pectoral, 1.4; ventral, 
1.4; longest dorsal ray, 1.3; longest anal ray, 1.7; caudal lobe, 1. 

Dorsal, 10; anal, 8. Scales, 35. 

Symmetrical, fusiform, not much compressed; the vent at about one-third the 
distance from anal origin to ventral axil; breast and belly broadly rounded. Snout 
bluntly pointed; mouth very slightly oblique; maxillary to almost under front of eye; 
lower jaw very slightly included; along slender terminal maxillary barbel; interorbital 
almost flat; eye large, oval; gill-membranes narrowly joined to isthmus under pos- 
terior margin of eye; suborbital crossed by fine subvertical lines of pores. Dorsal 
and anal without spinous rays; dorsal origin equidistant from the tip of snout and 
middle of peduncle; very slightly in advance of ventral origin; pectoral not reaching 
ventral, ventral not reaching anal; caudal narrow, well-forked, with pointed lobes. 


8 AMERICAN MUSEUM NOVITATES [No. 181 


Scales with conspicuous wide-spaced radiating striz; lateral line complete, a little 
bent down, running in center of peduncle. 

A series of small dark spots above the lateral line, and a faint plumbeous streak 
in the center of peduncle. 


REFERENCES 


Bere, 1907, Ann. Mag. Nat. Hist., (7) XIX, p. 163; 1914, ‘Faune Russ.,’ Poiss., 
III, part 2, p. 461, Fig. 63; 1916, ‘Poiss. Eaux Douces Russ.,’ pp. 223-225, 
Figs. 165-167; pp. 228-229, Fig. 170. 

BLEEKER, 1860, Nat. Tijd Ned. Ind., XX, p. 435. 

DysowskI, 1872, Verh. Zool.-Bot., Ges., Wien, XXII, p. 215. 

Fow ter, 1924, Bull. Amer. Mus. Nat. Hist., L, pp. 377-379. 

GarRMAN, 1912, Mem. Mus. Comp. Zodl., XL, p. 120. 

GunTHER, 1889, Ann. Mag. Nat. Hist., (6) IV, p. 224; 1896, Ann. Mus. Zoél. Acad. 
St. P., p. 212. 

JORDAN, 1920, ‘Genera Fishes,’ IV, p. 471. 

Kner, 1867, Novara Fische, p. 351, Pl. xv, fig. 2. 

SAUVAGE AND DaBRY DE Tarceen 1874, Ann. Sci. Nat., 6) 1 [Arts 5, p- ate 

TEMMINCK AND SCHLEGEL, 1844, ‘Fauna Japonica,’ p. 200, Pive; figs 1: 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 182 Tur AMERICAN Museum OF NATURAL History | uly il if 1925 
New York City 


59.7,55C (51) 


SOME CHINESE FRESH-WATER FISHES! 
By J. T. NicHous 


VII—NEW CARPS OF THE GENERA VARICORHINUS AND XENOCYPRIS 


There are four species of Varicorhinus with very small barbels in 
the collections of the Third Asiatic Expedition. One of these from Fukien 
seems not to be differentiable from Varicorhinus tamusuiensis (Oshima, 
1919) described from Formosa. The relationship of these forms is prob- 
ably close to species figured by Berg, 1916, but it is improbable that 
any will prove to be identical with same. 


Varicorhinus tamusuiensis (Oshima) 


Scaphesthes tamusuiensis Osuima, 1919, Ann. Carn. Mus., XII, p. 209, Pl. 1, fig. 1. 

Description of a specimen from streams in the general Min River Basin, near 
Yenping, Fukien, August, 1920; collected by H. R. Caldwell. 

Length to base of caudal, 95 mm. Depth in length, 4.3; head, 4. Eye in head, 
3.4; snout, 3.4; interorbital, 2.9; maxillary, 2.7; width of mouth, 2.8; depth of 
peduncle, 2.8; its length, 1.4; pectoral, 1.2; ventral, 1.3; longest dorsal ray, 1.4; 
longest anal ray, 1.7; caudal lobe, 1. 

Dorsal, 10; anal, 74. Scales, 48. Teeth, 4 or 5, 3, 2. 

Elongate, compressed; the head broad and blunt, rounded above and in front; 
the lower jaw flat, concave in profile. Mouth inferior, transverse, almost straight, 
the lower jaw very broad, with sharp, brown, horny edge, and no free lip; two pairs 
of minute barbels; maxillary horizontal, to under front of eye; front and sides of 
snout with large, horny warts; gill-membranes narrowly joined to isthmus under 
posterior margin of eye. Dorsal and anal without spinous rays; dorsal origin equi- 
distant from tip of snout and base of caudal; ventral origin under first third of dorsal; 
pectoral not reaching ventral, ventral not reaching anal; caudal deeply forked, 
upper lobe a little the longer. Scales rough, with subparallel slightly radiating strie; 
lateral line complete; dropping along the upper part of the opercular margin, little if 
any curved and near center of side from the tips of the pectorals back. 

Color dark, paler beneath; margin of dorsal with black on the membrane. 


Such differences as this species shows from tamusuzensis are mostly 
attributable to Oshima’s Formosan specimen having been of larger size, 
230 mm: The pectoral of our fish is longer, extending two-thirds versus 
three-fifths the distance to ventral, but this difference is insignificant. 


A leatons of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 54. 


2 AMERICAN MUSEUM NOVITATES [No. 182 


Varicorhinus robustus, new species 


DESCRIPTION OF THE TyPE.—No. 8424, American Museum of Natural History, 

from near Yenping, Fukien; collected by H. R. Caldwell. 

Length to base of caudal, 103 mm. Depth in length, 3.7; head, 4. Eye in head, 
3.5; snout, 3; interorbital, 2.6; maxillary, 2.5; width of mouth, 2; greatest width of 
body (at shoulder), 1.5; depth of peduncle, 2.4; its length, 1.4; pectoral, 1.2; ventral, 
1.3; longest dorsal ray, 1; longest anal ray, 1.2; caudal lobe, 0.8. Posterior barbel 
in eye, 6. 

Dorsal, 10; anal, 8. Scales, 47. Teeth (in a cotype), 4, 3,2; hooked. 

Not greatly compressed; the head rounded and blunt; breast broad, rounded; 
ventrals approximately in a horizontal plane. Mouth broad, almost straight across, a 
little behind tip of the rounded snout; upper lip over-hanging front of mouth and 
overhung in turn by transverse snout flap; lower jaw with a sharp, brown, horny edge; 
lower lip with slight line of demarkation behind; maxillary to under front of eye; 
with a small terminal barbel, a trifle longer than a similar barbel on each side of tip of 
snout flap; interorbital somewhat convex; gill-membranes narrowly adnate to isthmus 
below corner of preopercle. Dorsal and anal without spinous rays; dorsal origin 
equidistant from tip of snout and middle of peduncle; ventral base slightly behind 
middle of that of dorsal; pectoral not nearly reaching ventral; ventral not reaching 
anal; anal almost to origin of caudal; caudal forked with equal pointed lobes. 
Scales rough, with somewhat radiating striez; lateral line complete, straight, in the 
center, rising a little to meet opercle. 

Color dark, especially the tips and bases of the scales, paler below the eye and on 
belly. 


This is a more robust fish than Varicorhinus tamusuiensis which 
occurs in the same locality. 


Varicorhinus shansiensis, new species 


DESCRIPTION OF THE TyPE.—No. 8425, American Museum of Natural History, 
from Niang-tze-kwan, Shansi; collected by C. H. Pope. 

Length to base of caudal, 174 mm. Depth in length, 3.9; head, 4.5. Eye in 
head, 5; snout, 2.6; interorbital, 2.6; maxillary, 2.9; width of mouth, 2.9; greatest 
width (at front of dorsal), 1.5; depth of peduncle, 2.4; its length, 1.4; pectoral, 1.2; 
ventral, 1.2; longest dorsal ray, 1.3; longest anal ray, 1.3; caudal lobe, 0.8. Barbel 
in eye, 3; distance between gill clefts, 1.2. 

Dorsal, 10; anal, 7. Scales, 51. Teeth (in a cotype), stout, slightly hooked, 
5, 3, 2, first tooth of the main row small. 

Little compressed; breast and belly broad. Mouth inferior transverse almost 
straight across, a little behind the blunt snout; lower jaw with a sharp, cartilaginous 
edge, pale in color, without free lip; maxillary almost to under front margin of eye; 
with a small subterminal barbel; a smaller second barbel at corner of snout flap on 
one side only; front and sides of snout with a few small horny points; gill-membranes 
joined to sides of isthmus under edge of preopercle. Dorsal and anal without spinous 
rays; dorsal origin equidistant from tip of snout and middle of peduncle; ventral 
under middle of dorsal base; pectoral reaching two-thirds the distance to ventral; 
ventral three-fourths the distance to anal; caudal well forked with equal pointed 


= J 
1925] SOME CHINESE FRESH-WATER FISHES 3 


lobes. Scales with rather close-spaced, slightly radiating strize; scales small on the 
breast; lateral line complete, rising slightly to meet opercle, otherwise straight in 
the center. 

Dark above, paler below, fins plain. 


This species has smaller scales than the two preceding and differs 
from each of them in other details. We have three specimens from the 
base of Tai-pei-shan in the Tsingling Mountains, Shensi, which we also 
refer to it. The position of the dorsal origin is variable, sometimes 
nearer base of caudal than tip of snout. 


Varicorhinus tungting, new species 


DESCRIPTION OF THE TyPE.—No. 8426, American Museum of Natural History, 
from Huping, Tungting Lake, Hunan; collected by C. H. Pope. 

Length to base of caudal, 126 mm. Depth in length, 4.5; head, 4.4. Eye in 
head, 5; snout, 2.4; interorbital, 2.2; maxillary, 2.5; width of mouth, 2.5; width of 
body, 1.5; depth of peduncle, 1.6; its length, 1.5; pectoral, 1.1; ventral, 1.3; longest 
dorsal ray, 1.1; height of anal, 1.4; caudal lobe, 0.7; across breast, gill slit to gill 
slit, 3.6. Barbel in eye, 2.5. 

Dorsal, 12; anal, 7. Scales, 45. Teeth (in a cotype), 5, 3, 3, or 5, 4,2; com- 
pressed, bluntly pointed, loosely attached, crowded. 

Head somewhat depressed; body increasingly compressed backward. Inter- 
orbital gently convex; snout bluntly pointed; extending well beyond the broad in- 
ferior, slightly curved, transverse mouth; mouth overhung by a thick membrane 
with free notched-fluted edge; lower lip free in front, with a similarly notched edge; 
maxillary not to under front of eye; with a small terminal barbel; an exceedingly 
minute barbel on the side of snout in advance of corner of snout membrane; gill- 
membranes broadly joined to side of breast under edge of preopercle. Dorsal and anal 
without spinous rays; dorsal origin equidistant from tip of snout and base of caudal; 
ventrals under middle of dorsal; pectorals reaching two-thirds the distance to ventral; 
ventral to anal origin; caudal large, well forked. Scales rough,-with close-spaced 
subparallel as well as the more marked concentric striz; lateral line complete, straight 
in the center; body scaling extending onto caudal base. 

Color dusky. 


This species is well differentiated from the three preceding by the 
notched free membranes surrounding the mouth. 

A number of species of Xenocypris have been described from China, 
and to draw up a satisfactory synonymy of this genus will be difficult 
and require more time than is now available. X. microlepis, macro- 
lepis and davidi Bleeker, 1871, and lamperti Popta, 1908, are probably 
good species. X. tapeinosoma Bleeker, 1871, argentea Giinther, 1868, 
mitidus Garman, 1912, enea and gunthert Sauvage and Dabry de Thier- 
sant, 1874, argenteus (Basilewski, 1855) and simoni (Bleeker, 1865) 
which may not belong in this genus, and jesella (Cuvier and Valenciennes, 


t AMERICAN MUSEUM NOVITATES [No. 182 


1844) which is unidentifiable, are probably not good species. Mr. Cald- 
well has collected a species in Fukien which is quite distinct from all 
these, and is here described as new. It resembles Distachodon tumi- 
rostris Peters, 1880, from Ningpo, Oshima, 1919, from Formosa, but is 
deeper, with a longer pectoral, etc. Distechodon Peters, 1880, is said to 
beseparated from Xenocypris by the character of the teeth, 2 rowed,—7,3. 


Xenocypris compressus, new species 

DESCRIPTION OF THE TypE.—No. 8427, American Museum of Natural History, 
from near Yenping, Fukien; collected by H. R. Caldwell. 

Length to base of caudal, 103 mm. Depth in length, 3.8; head, 3.9. Eye in 
head, 3.5; snout, 3.1; interorbital, 3.3; maxillary, 4.3; width of mouth, 3.6; great- 
est width of body (at back of head), 2; depth of peduncle, 2.6; its length, 2; pectoral, 
1.3; ventral, 1.6; longest dorsal ray, 1.3; longest anal ray, 1.8. Mouth to tip of 
snout in eye, 2.7. 

Dorsal, II, 7; anal, 12. Scales, 74. Teeth broken. 

Compressed, pointed before and behind; top and bottom of head flat, slanting. 
Interorbital flattish; mouth inferior, transverse, almost straight across; gill-mem- 
branes narrowly joined to isthmus behind posterior margin of eye. Last simple 
dorsal ray aslender spine, its tip soft; dorsal origin equidistant from base of caudal 
and front of eye; ventral origin under that of dorsal; pectoral narrow and pointed, 
reaching two-thirds the distance to ventral; ventral two-thirds to anal; caudal forked. 
Scales with conspicuous concentric and wide-spaced slightly radiating striz; lateral 
line complete, somewhat bent down behind head, running in center of peduncle. 

Pale below and on sides, darker on back; sides with more or less distinct, fine 
dark streaks following the rows of scales, in one of which the lateral line is situated; 
fins plain. 


VIII.—CARPS REFERRED TO THE GENUS PSEUDORASBORA 


The type of Pseudorasbora Bleeker, 1859, is Leuciscus parvus Tem- 
minck and Schlegel, 1844, from Japan. A small fish, widely distributed 
on the mainland of eastern Asia, is generally considered identical there- 
with. We have also a closely related form which is distinct, collected 
by Mr. Granger in Szechwan. 

Secondly there is to hand a fish from Anhwel, described and figured 
by Fowler, 1924, and referred by him to Aphyocypris chinensis Giinther. 
An authentic specimen of Aphyocypris, courteously sent us more 
recently by Mr. J. R. Norman of the British Museum, shows this to be a 
quite different genus apparently related to the Abramidinz. Fowler’s 
species may be provisionally referred to Pseudorasbora, as may also a 
form resembling it from Shansi. 

Thirdly, an undescribed fish from Fukien, is here included in 
Pseudorasbora. 


1925] SOME CHINESE FRESH-WATER FISHES 5 


Pseudorasbora parva (Temminck and Schlegel) 


Leuciscus parvus TEMMINCK AND SCHLEGEL, 1844, ‘Fauna Japonica,’ p. 215, 
Pl. cu, fig. 3. Japan. 
Pseudorasbora parva Fow ER, 1924, Bull. Amer. Mus. Nat. Hist., L, p. 382, Chihli. 


Pseudorasbora altipinna, new species 


DESCRIPTION OF THE TyPE.—No. 8428, American Museum of Natural History 
from Yen-ching-kao, Szechwan; collected by Walter Granger. 

Length to base of caudal, 55 mm. Depth in length, 4; head, 3.8. Eye in head, 
4; snout, 3; interorbital, 2.6; maxillary, 4.6; width of head and of body, 1.9; depth 
of peduncle, 2; its length, 1.1; pectoral, 1.3; ventral, 1.3; longest dorsal ray, 1; 
caudal lobe, 0.8; longest anal ray, 1.8. 

Dorsal, 9; anal, 8. Scales, 38. Teeth (in a cotype), 5, small, hooked, in one 
row. 

Compressed; nape slightly elevated; breast broad and belly rounded. Mouth 
small, transverse, vertical; lower jaw projecting; maxillary not reaching nearly to 
front of eye; no barbels; interorbital flattish; eye slightly infralateral; gill-membranes 
joined to side of isthmus well behind edge of preopercle. Dorsal and anal without 
spinous rays; dorsal origin equidistant from tip of snout and middle of peduncle; 
ventral origin under that of dorsal; pectoral not quite reaching ventral, and ventral 
not quite reaching anal; caudal forked, with equal lobes. Scales with faint, slightly 
radiating striz; lateral line complete, in or slightly below the center, rising gently to 
meet opercle. 

Color uniform. Several specimens were taken, of which this is one of the largest. 


Pseudorasbora fowleri, new species 


Aphyocypris chinensis FowLER, 1924, Bull. Amer. Mus. Nat. Hist., L, p. 383, 
Fig. 1. Anhwei. Not of Giinther. | 


Pseudorasbora depressirostris, new species 


DESCRIPTION OF THE TypE.—No. 8429, American Museum of Natural History, 
from Chin-Ssu, Shansi; collected by C. H. Pope. 

Length to base of caudal, 49 mm. Depth in length, 3.8; head, 3.6; eye, 4.2; 
snout, 3.3; interorbital, 2.5; maxillary, 4; depth of peduncle, 2; its length, 1.4; 
pectoral, 1.5; ventral, 1.5; longest dorsal ray, 1.3; longest anal ray, 1.6; caudal 
lobe, 1.2. 

Dorsal, 9; anal, 8. Scales, 38. Teeth (in a cotype), 5, small, hooked, in one row. 

Body compressed; interorbital broad, very slightly convex; nape elevated, snout 
depressed. Snout pointed; mouth small, transverse, almost vertical; lower jaw 
projecting; maxillary reaching about half the distance to eye; no barbels; opercle 
with a membranous edge; gill-membranes joined to breast behind edge of preopercle; 
2 pairs of horny points on the chin, 2 such points in line back of angle of mouth, and 
one near maxillary higher up. Dorsal and anal without spinous rays; dorsal origin 
equidistant from tip of snout and base of caudal; very slightly in advance of ventral 
origin; pectoral short, reaching two-thirds the distance to ventral, ventral not reach- 
ing anal; caudal moderately forked. Scales with close-spaced, conspicuous, radiating 


6 ' AMERICAN MUSEUM NOVITATES [No. 182 


strie; lateral line complete, dropping slightly behind opercle, thence straight to 
caudal base in center of side. 

Dark above, pale below; scales of back and sides bordered with dark; fins plain, 
ventrals somewhat yellowish; an obscure dark band from eye to base of caudal, 
scarcely indicated in the type, but better marked in some individuals, of which we 
have several of about the same size. 


Pseudorasbora monstrosa, new species 


DESCRIPTION OF TyPE.—No. 8480, American Museum of Natural History, from 
near Yenping, Fukien; collected by H. R. Caldwell. 

Length to base of caudal, 62mm. Depth in length, 4; head, 4. Eye in head, 4; 
snout, 3.2; interorbital, 2.7; maxillary, 4; depth of peduncle, 2.2; its length, 1.2; 
pectoral, 1.5; ventral, 1.5; longest dorsal ray, 1.4; longest anal ray, 1.8; caudal 
lobe, 1.1. 

Dorsal, 9; anal, 8. Scales, 33. Teeth, 5, raised, hooked, in one row. 

Moderately compressed; head blunt, its top broad and flattish, the nape slightly 
elevated; breast and belly rounded. Maxillary vertical; mouth small, almost 
strictly transverse; lower jaw very little projecting; no barbels; tip of snout before 
nostrils slightly swollen and set off by a groove; 2 large horny warts on side of snout, 
2 or 3 smaller ones below eye; also a large two-pronged wart on one side of mandible; 
gill-membranes adnate to breast behind edge of preopercle. Dorsal without spinous 
rays, the 2nd (last) simple ray stiffened at base; dorsal origin equidistant from tip of 
snout and center of peduncle, directly over ventral origin; pectoral not reaching 
ventral, ventral not reaching anal; caudal forked with equal pointed lobes. Scales 
with conspicuous radiating strie; lateral line straight to over anal origin, absent on 
peduncle. 

Scales with dusky borders; the fins bordered with blackish. 


This may be an abnormal, large individual of a variable species. 
Smaller ones which we refer to it, with some hesitation, are less elongate, 
especially the peduncle, have the lateral line complete, and the ventral 
origin before that of the dorsal. Mr. Pope, who is now in the field, is 
likely to obtain additional material bearing on this point. 


IX.—THREE NEW ABRAMIDIN CARPS 
Rasborinus fukiensis, new species 


DESCRIPTION OF THE TypE.—No. 8431, American Museum of Natural History, 
from near Yenping, Fukien; collected by H. R. Caldwell. 

Length to base of caudal, 69 mm. Depth in length, 3.4; head, 4. Eye in head, 
3.5; snout, 4; interorbital, 3.4; maxillary, 3.6; depth of peduncle, 2.5; its length, 2; 
pectoral, 1.2; ventral, 1.5; longest dorsal ray, 1.4; longest anal ray, 1.5; caudal 
lobe, 1. 

Dorsal, 10; anal, 17. Scales, 39. Teeth (in a cotype), 5, 4, 1, slightly hooked. 

Compressed, the head rather blunt; breast broadly rounded; a low, sharp, naked 
keel between ventrals and anal. Mouth oblique; lower jaw broad, slightly projecting, 
with a raised inner edge; maxillary to under front margin of eye; no barbels; inter- 


1925] SOME CHINESE FRESH-WATER FISHES vi 


orbital somewhat convex; gill-membranes narrowly united and free from isthmus at 
base. Dorsal and anal without spinous rays; dorsal origin equidistant from base of 
caudal and hind margin of eye, nearer anal origin than ventral axil; pectoral reaching 
ventral origin; ventral not reaching anal; caudal shallowly forked with equal lobes; 
upper ray of pectoral thickened, with fine tooth-like papille above on the inside. 
Scales with a few widely radiating strie; lateral line complete, bent down, rising on 
peduncle to terminate in its center; anal with a broad sheath of scales. 
Color uniform. 


This species is closely related to one found in the island of Hainan, 
but is less deep, lower jaw slightly projecting, versus included, ete. Both 
species are close to Rasborinus takakit Oshima, 1920, from Formosa, with 
scales, 36, interorbital broader, 2.5, etc. 


Hemiculterella engraulis, new species 


DESCRIPTION OF THE TypE.—No. 8432, American Museum of Natural History, 
from Huping, Tungting Lake, Hunan; collected by C. H. Pope. 

Length to base of caudal, 148 mm. Depthin length, 5; head, 3.7. Eye in head, 
4.4; snout, 3.4; interorbital, 3.6; maxillary, 2.5; depth of peduncle, 3.5; its length, 
2.5; width of body (the back of head), 2; pectoral, 1.4; ventral, 2.1; longest dorsal 
ray, 2.3; longest anal ray, 2.8; lower caudal lobe, 1.2. 

Dorsal, 9; anal, 22. Scales, about 50. Teeth, 4, 4, 2, hooked, pharyngeal bone 
heavy. 

Broad, moderately compressed only; snout blunt, profile low; a low naked Keel 
between ventrals and anal, traces of which (crossed by scales) continue forward to 
the pectoral axil. Interorbital convex; snout blunt; jaws equal; the pointed tip of 
mandible fitting into a notch in the middle of snout; front of jaws with an appreciable 
S-shaped curve; upper jaw protractile; mouth strongly oblique; maxillary to under 
front of eye; no barbels; angle of preopercle bluntly projecting; gill-membranes con- 
fluent with one another and center of isthmus behind.edge of preopercle. Dorsal and 
anal without spinous rays; dorsal origin equidistant from base of caudal and posterior 
margin of eye; well behind ventral base; anal origin slightly behind dorsal axil; 
pectoral not reaching ventral, ventral not nearly to anal; caudal well forked, the lower 
lobe decidedly the longer. Scales deciduous; with radiating strie; lateral line com- 
plete, slanting down to over tip of pectoral, thence turning back horizontally, rising 
in a steep slant over anal axil to run in the center of peduncle. 

Dark on the back; pale below; fins plain. 


This well-marked species, of which but a single example is to hand, 
is provisionally referred to Hemiculterella Warpachowski, 1888. 


Hemicultur clupeoides, new species 


DESCRIPTION OF THE Type.—No. 8433, American Museum of Natural History, 
from Tungting Lake, Hunan, December 24, 1921; collected by C. H. Pope. 

Length to base of caudal, 127 mm. Depth in length, 4.3; head, 4.6. Eye in 
head, 3.7; snout, 3.9; interorbital, 3.2; maxillary, 3.9; depth of peduncle, 2; pec- 
toral, 1; ventral, 1.6; height of dorsal, 1.4; height of anal, 2.5; caudal lobe, 0.7. 


8 AMERICAN MUSEUM NOVITATES [No. 182 


Dorsal, II, 7; anal, 14. Scales, about 55, the last well on caudal base. Teeth, 
4, 4, 2, slightly hooked. 

Body moderately compressed; jaws equal; maxillary concealed under pre- 
orbital, not quite reaching to below front of eye; gill-membranes joined, forming a 
narrow fold across isthmus. Last dorsal spine slender; ventral origin slightly in 
advance of that of dorsal; pectoral not reaching ventral, ventral falling far short of 
anal. A sharp, naked keel from ventrals to anal; scales deciduous; lateral line com- 
plete, running low, rising abruptly to center of peduncle over anal‘axil. Scales on 
breast pointed, with an appreciable slight keel running forward from ventrals two- 
thirds the distance to isthmus. 


This species differs from others of the genus Hemiculter in being less 
compressed, with more deciduous scales, the anterior flexure in the lateral 
line less marked. Hemiculter kneri Kreyenberg and Pappenheim, 1908, 
is probably referable to it at least in part, but this is preoccupied by 
Hemiculter kneri Warpachowski, 1888. 


REFERENCES 


BasILEwskI, 1855, Mem. Soc. Nat. Mosc., X, p. 232. 

Bera, 1916, ‘Poiss. Eaux Douces Russ.,’ Figs. 189-192. 

BLEEKER, 1859, Nat. Tijd. Ned.-Ind., XX, p. 485; 1865, Ned. Tijd. Dierk., II, p. 
25; 1871, ‘Cyp. Chine,’ Verh. Akad. Wet. Amst., Nat., XII, pp. 52-59. 

CUVIER AND VALENCIENNES, 1844, ‘Hist. Nat. Poiss.,’ XVII, p. 360. 

GaRMAN, 1912, Mem. Mus. Comp. Zoél., XL, p. 117. 

GtnrTueER, 1868, ‘Cat. Fishes,’ VII, p. 205. 

KREYENBERG AND PAPPENHEIM, 1908, Sitzb. Ges. Nat. Berl., p. 105. 

Osuima, 1920, Proc. Acad. Phila., LX XII, p. 130, Pl. 111, fig. 3. 

Peters, 1880, Monatb. Akad. Wiss. Berl., p. 924. 

Popra, 1908, Zool. Anz., XXXII, p. 243. 

SAUVAGE AND Dasry DE THIERSANT, 1874, Ann. Sci. Nat., (6) I, Art. 5, p. 13. 

WarPacHowskI, 1888, Bull. Acad. Sci. St. P., XXXII, pp. 17 and 23. 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 185 Tue AMERICAN Museum oF NATURAL HIsToRY Sept. 25, 1925 
New York City 


59.7(51) 


SOME CHINESE FRESH-WATER FISHES! 


By J. T. NicHoLs 


X. SUBGENERA OF BAGRIN CATFISHES 


There are a number of related catfishes in China variously assigned 
to the genera Pseudobagrus Bleeker (type, Bagrus aurantiacus Tem- 
minck and Schlegel, Japan) and Lezocassis Bleeker (type, L. micropogon 
Bleeker, Sumatra, Borneo, etc.). Those referred to Lezocassis have the 
adipose rather long, and are supposed to have the orbital rim adnate. 
As a matter of fact they fall readily into two quite different series, 
neither of which is referable with any certainty to.the type represented 
by East Indian mcropogon, and it is best for the present to separate 
them from the same as subgenera, as follows. 

NASOCASSIS, new subgenus. Snout witha tendency to be elongate or swollen. 
Orbital rim much as in Pseudobagrus, usually partially free, but eye much smaller, 
so that its rim appears adnate. Caudal well forked. Type: Lezocassis longirostris 
Giinther, China.? 

DERMOCASSIS, new subgenus. Orbital rim completely or almost completely 
adnate. Caudal emarginate, truncate, or rounded. Type: Lezocassis ussuriensis 
Berg (Bagrus ussuriensis Dybowski), Manchuria, China, ete. 

On the other hand, Flwidraco Jordan and Fowler (type Pseudo- 
bagrus ransonnett Steindachner, Japan) for those species of Pseudo- 
bagrus wherein bones on the top of the head are exposed, rugose, is 
certainly not entitled to more than subgeneric rank. 

A related but quite distinct form of catfish common in China, with 
elongate body, flat head and very long adipose, was described as Hemz- 
bagrus macropterus Bleeker, and is thus far recognized as a single species. 
The writer doubts the close relationship of this fish to Bagrus. nemurus 
Cuvier and Valenciennes, Java, Malacca, Siam, etc., in fresh and brackish 
water, which is the type of Hemzbagrus Bleeker. He here proposes for 
Hemibagrus macropterus Bleeker the subgenus Macropterobagrus, 
distinguished by elongate body, very long adipose, depressed head and 


ar ee ar of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 55. 

*Giinther, 1888, Ann. Mag. Nat. Hist., (6) I, p. 430, says: ‘‘ This is not a Japanese species, as I was 
incorrectly informed when I described it.”’ 


2 AMERICAN MUSEUM NOVITATES [No. 185 


weekly forked caudal; to stand as a full genus if Hemibagrus be sub- 
ordinated to Macrones Dumeril (equals Aorta Jordan) or to Pseudo- 
bagrus Bleeker. ; 


XI.—CERTAIN APPARENTLY UNDESCRIBED CARPS FROM FUKIEN 
Barbus caldwelli, new species 


DESCRIPTION OF THE TypE.—No. 8434, American Museum of Natural History, 
from near Yenping, Fukien; collected by H. R. Caldwell. 

Length to base of caudal, 100 mm. Depth in length, 3.6; head, 3.5. Eye in 
head, 4; snout, 3; interorbital, 2.7; maxillary, 3.1; width of mouth, 3.6; width of 
snout, 3; posterior barbel, 3.2; width of head and of body, 1.8; depth of peduncle, 
2.8; its length, 2.5; pectoral, 1.5; ventral, 1.6; longest dorsal ray, 1.9; longest anal 
ray, 1.7; lower caudal lobe, 1.1. 

Dorsal, 11; anal, 744. Scales, 24. 

Moderately compressed, the head broad, snout pointed. Mouth very slightly 
oblique; lower jaw included; two well-developed slender barbels, the posterior at the 
end of the maxillary decidedly the longer; maxillary to under front margin of eye; 
left side of snout only with a band of small, crowded, warty points above the maxillary; 
eye lateral; interorbital very slightly concave; gill-membranes rather narrowly 
joined to side of isthmus just before the edge of the preopercle. Dorsal and anal with- 
out spinous rays; dorsal origin equidistant from tip of snout and base of caudal; 
ventral placed slightly behind center of dorsal base; pectoral reaches two-thirds 
the distance to ventral; ventral three-fourths to anal; caudal moderately forked with 
pointed lobes, the lower slightly the longer. Scales rough, with close-spaced radiating 
strie; lateral line complete, slightly bent down, in the center of peduncle. 

Dorsal with a black free margin; scales faintly outlined in dark; a dark streak 
in the outer basal margin of each caudal lobe. 


Three larger specimens also were collected of this handsome barb, 
which is closely related to a similarly colored form we have from the 
island of Hainan. Its differences from the latter seem to lie in a less 
chubby body, narrower, more pointed snout, and slightly higher scale 
count. . 


Barbus (Lissochilichthys) hemispinus, new species 


DESCRIPTION OF THE TyPE.—No. 8435, American Museum of Natural History, 
from near Yenping, Fukien; collected by H. R. Caldwell. 

Length to base of caudal, 64 mm. Depth in length, 3.1; head, 3. Eye in head, 
3.5; snout, 2.6; interorbitat, 4; maxillary, 3.8; posterior barbel, 4.5; depth of 
peduncle, 2.4; its length, 2; pectoral, 1.4; ventral, 1.5; longest dorsal ray, 1.5; 
longest anal ray, 1.6; caudal lobe, 1.1. 

Dorsal II, 84; anal, 8. Scales, 39. 

Compressed, back somewhat elevated. Snout long and narrow; lower jaw in- 
cluded; lower lip deeply grooved in the middle, the jaw projecting forward beyond it; 
two rather long maxillary barbels, the posterior the longer; mouth horizontal; 
maxillary not reaching to below front border of eye; interorbital little convex; gill- 


1925] SOME CHINESE FRESH-WATER FISHES 3 


membranes broadly joined to side of breast below edge of preopercle. Last simple ray 
of dorsal stiffened and serrate behind, but soft distally; dorsal origin equidistant from 
base of caudal and middle of snout, over that of ventral; pectoral not reaching ventral, 
ventral not reaching anal; caudal forked, with pointed lobes. Scales rough, with 
well-marked, somewhat radiating striz; scales somewhat smaller posteriorly; lateral 
line complete, almost straight in the center, very slightly bent down in front. 

A vague dark mark on peduncle. . 


The presence and character of a dorsal spine does not appear to be 
of much significance in barboid carps, hence this species, on the basis of 
the structure of its mouth, is placed close to Lissochilichthys matsudai 
Oshima, 1920, Formosa, which would apparently stand as a synonym 
of Gymnostomus labiatus Regan, 1908,? were labiatus not preoccupied by 
Barbus (Labeobarbus) labiatus Boulenger, 1902,° if Lissochilichthys be 
considered a subgenus of Barbus. Barbus (Lissochilichthys) matsudaz, 
by the way, is common in Fukien, apparently not separable from the 
Formosan fish. Also there is a fish closely related to B. hemispinus in 
the island of Hainan, apparently more slender, with shorter barbels, 
and reaching a larger size. 


Sarcocheilichthys sinensis fukiensis, new subspecies 


DESCRIPTION OF THE Typp.—No. 8436, American Museum of Natural History, 
from near Yenping, Fukien; collected by H. R. Caldwell. 

Length to base of caudal, 90 mm. Depth in length, 3.6; head, 4.4. Eye in head, 
3.5; snout, 3; interorbital, 2.5; maxillary, 3; width of mouth, 3.4; width of body, 
1.5; depth of peduncle, 1.7; its length, 1.2; pectoral, 1; ventral, 1.2; longest dorsal 
ray, 0.9; longest anal ray, 1.2; caudal lobe, 0.9. Barbel in eye, 6. 

Dorsal, 915; anal, 8%. Scales, 41. 4 rows between lateral line and a 16 
around caudal podanel: ; 11 between the lateral lines across nape. 

Moderately compressed, snout somewhat pointed, nape little if at all elevated; 
vent at one-third the distance from anal origin to ventral axil, with a long ovipositor. 
Mouth slightly behind the tip of snout, horizontal, inferior, heart-shaped; with mod- 
erate lips, confined to the sides of lower jaw, which is, with an exposed shaft between 
them, somewhat expanded to a curved sharp horny tip; length of mouth 0.8 in its 
width; maxillary to under front of eye; a small barbel above its end; interorbital 
very slightly convex; gill-membranes broadly joined to side of isthmus behind edge of 
preopercle. Dorsal origin equidistant from tip of snout and anal axil; ventrals under 
center of dorsal base; pectoral reaching five-sixths the distance to ventral; ventral 
five-sixths to anal; caudal moderately forked. Scales rough, with subparallel horizon- 
tal or very slightly radiating striz; lateral line complete straight in the center. 

Four rather well-defined broad blackish cross-bands, one from shoulder to pec- 
toral axil, one from dorsal to ventral narrowly crossing belly behind ventral axil, one 
opposite the anal, and one on peduncle; anterior and central portion of dorsal, 


1Proc. Acad. Nat. Sci. Phila., LX XII, p. 124, Pl. 11, fig. 2. 
2Ann. and Mag. Nat. Hist., (8) II, p.358. 
3Proc. Zo6l. Soc., part 2, p. 223, Pl. xv 11, fig. 1. 


4 AMERICAN MUSEUM NOVITATES [No. 185 


anterior (upper) portion of pectoral, central portions of ventral anal and caudal, 
blackish, the margins of the fins pale. 

Compared with a specimen of the same size (S. s. sinensis) from Tungting Lake 
on the central Yangtze which has depth, 3.7; nape more gibbous, snout blunter; 
length of mouth 1.4 in its width; barbel 8 in eye; 5 rows of scales between lateral line 
and ventral, 18 around caudal peduncle; cross-bands less defined and an ill-defined 
lateral band; more dusky on the bases of the fins, dorsal with less of a pale tip. — 


Berg, 1916,! considers Barbodon lacustris, type of the subgenus 
Barbodon Dybowski, to be a race of Sarcocheilichthys sinensis Bleeker, 
1871,? doubtfully credited to the Yangtze by Bleeker. Our Barbodon from 
Tungting is certainly very close to lacustris as figured by Berg, and we 
follow him in identifying it with sznenszs Bleeker; although a species of 
the subgenus Chilogobio inhabiting the same waters might perhaps as 
well have been so identified. 

Sarcocheilichthys (Barbodon) sinensis from Ningkwo, Anhwei, may be 
taken as representing this species in the lower Yangtze Basin. The differ- 
ences which it shows from the Tungting fish are too slight to receive 
taxonomic notice. 


Garra orientalis, new species 


DESCRIPTION OF THE TyPE.—No. 8437, American Museum of Natural History, 
from near Yenping, Fukien; collected by H. R. Caldwell. 

Length to base of caudal, 75mm. Depth in length, 4.4; head, 3.9. Eye in head, 
5.5; snout, 2.2; interorbital, 2.7; maxillary, 2.3; width of mouth (and of disk), 2; 
mouth to posterior edge of disk, 2.7; distance between gill clefts, 3; greatest width of 
body (at back of head), 1.5; depth of peduncle, 2; its length, 1.7; pectoral, 1; 
ventral, 1.2; longest dorsal ray, 1; longest anal ray, 1.4; caudal lobe, 0.8. Barbels 
in eye, 1.5. 

Dorsal, 1044; anal, 7%. Scales, 33. 

Moderately compressed; the head broad and blunt, somewhat cuboid; lower 
surface of head and breast flattened so that bases of pectorals and ventrals are in a 
horizontal plane; vent at slightly more than one-third the distance from anal origin 
to ventral axil. Eye placed high; slightly superolateral; top of head slightly convex; 
marked off before the nostrils from the lower front part of snout by a groove; the tip 
of snout again slightly raised; mouth inferior, transverse, very little curved; over- 
hung in front by the broad snout membrane which is finely pimply near its margin 
and with comblike serrations on its edge; a rounded disk on chin, smooth in the 
center and finely pimply near front and hind borders; a small barbel at the tip of the 
maxillary and another on the side of the front of the snout; forehead and top of 
snout with warty points; top of head and to some extent side of snout with fine, little 
prominent pimples; gill-membranes broadly joined to side of breast. Dorsal and anal 
without spinous rays; dorsal origin equidistant from tip of snout and anal axil; ventral 
placed under center of dorsal; pectoral not reaching ventral, ventral not reaching anal; 


1 Poiss. Eaux Douces Russ.,’ p. 234, Fig. 175. 
2Verh. Akad. Wet. Amst., Nat., XII, p. 31, Pl. rv, fig. 2. 


1925] SOME CHINESE FRESH-WATER FISHES 5 


caudal forked, with equal lobes. Scales rough, with fine irregular parallel strie; 
lateral line complete, straight, in the center, rising slightly to meet opercle. 

Color uniform, rather dark, paler on belly; a vague dark blotch in center of 
peduncle near base of caudal. 

This species, with a groove before the nostrils, approaches one found 
in the island of Hainan, wherein the forehead projects as in Schisma- 


torhynchos. 


XII.—A SMALL GOBY FROM THE CENTRAL YANGTZE 
Gobius cliffordpopei, new species 


DESCRIPTION OF THE TyPE.—No. 8438, American Museum of Natural History, 
from Tungting Lake, Hunan, December 16, 1921; collected by C. H. Pope. 

Length to base of caudal, 34 mm. Depth in length, 4.7; head, 3.4. Eye in 
head, 5; snout, 3.5; maxillary, 2; depth of peduncle, 3; its length, 1.3; pectoral, 1.3; 
ventral, 2; longest dorsal spine, 1.7; longest dorsal ray, 1.8; longest anal ray, 2; 
caudal, 1. 5. Interorbital, 4 in eye. 

Dorsal VI or VII—9; anal, 8. Head ae before dorsal; scales, 28. No canine 
teeth. 

Eyes close together, superolateral; mouth somewhat oblique, with thick lips, 
the maxillary extending to under front of eye; jawsequal or lower slightly projecting; 
gill-membranes joined to sides of breast under or behind edge of preopercle; pectorals 
reaching beyond ventrals, but not to front of anal; caudal bluntly pointed. 

Broad black cross-bands posteriorly, somewhat variable in width. 

Small gobies of the subgenus Rhinogobius, perhaps referable to G. 
hadropterus (Jordan and Snyder) of Japan, appear to be common in the 
coastal rivers of China. It is not unlikely that careful study of an ade- 
quate series will make it possible to recognize two or three species coast- 
wise. At least the present species from Tungting Lake in the middle 
Yangtze is distinct from hadropterus, differentiated by the, if anything, 
slightly projecting lower jaw, and broad, conspicuous vertical bands on 
the sides. Absence of scales on the nape may be due to the small size 
of the type, but the species would seem to be a small one. 

Named for Mr. C. H. Pope, the thoroughness of whose field work in 
China has brought to light many species of fishes previously overlooked. 


XIII.—A NEW MINNOW REFERRED TO LEUCOGOBIO 
‘There is to hand a single specimen of a small carp from Anhwel, 
the exact position of which was at first something of a puzzle. Careful 
study leaves no question that it is close to Lewcogobio, in which genus it is 
provisionally placed. It bears a strong superficial resemblance to Pseudo- 
rasbora fowleri Nichols, 1925, Amer. Mus. Novitates, No. 182, p. 5 


6 AMERICAN MUSEUM NOVITATES [No. 185 


Leucogobio imberbis, new species 


? Aphyocypris chinensis Fowmr, 1924, Bull. Amer. Mus. Nat. Hist., L, p. 383, 
in part, not of Giinther. 

DESCRIPTION OF THE TyPE.—No. 8439, American Museum of Natural History, 
from Ningkwo, Anhwei, September 15 to October 15, 1921; collected by C. H. Pope. 

Length to base of caudal, 68 mm. Depth in length, 3.7; head, 3.4. Eye in head, 
4.4; snout, 3.7; interorbital, 3.1; maxillary, 3.4; depth of peduncle, 2.8; its length, 
1.5; pectoral, 1.4; ventral, 1.6; longest dorsal ray, 1.6; longest anal ray, 1.8; caudal, 
1.3. 

Dorsal, 9; anal, 8 (counting 2 simple rays). Scales, about 42. Teeth, 5, 2 or3, 
rather stout, slightly hooked. 

Moderately compressed; back elevated; nape slightly gibbous; top of head flat. 
Jaws equal; mouth oblique; maxillary not quite reaching to under front of eye, witha 
minute thick barbel concealed near its tip, on one side only; gill-membranes joined, 
free from isthmus, at extreme base; opercle with a membranous edge reaching pectoral 
base. Dorsal without spinous ray; its origin equidistant from end of snout and base 
of caudal; ventral origin directly under that of dorsal; pectoral not quite reaching to 
ventral, ventral not quite to anal, caudal moderately forked. Lateral line apparently 
complete (scales lost on peduncle), running almost straight and about in the middle 
of the side; scales rather rough, with slightly radiating striz; an elongate folded mem- 
branous scale in ventral axil, with a second, pointed scale as covert. 

A blackish band from shoulder to base of caudal, and dark mark from back of eye 
towards shoulder; a pale area above the lateral band, divided by a darkish stripe; 
back mostly dark; fins plain. 


XIV.—TWO APPARENTLY UNDESCRIBED FISHES FROM YUNNAN 


A considerable amount of scattered systematic work has recently 
been done on the fishes of Yunnan, which province appears to possess a 
large fish fauna. With only a small series of forms from there for com- 
parison, the writer has some hesitation in proposing the following two 
new species, which, nevertheless, he is unable to find anywhere described 
in the literature. 


Xenocypris yunnanensis, new species 


DESCRIPTION OF THE TyPE.—No. 8440, American Museum of Natural History, 
from Yunnan-fu Lake, Yunnan, February 20, 1919 ; collected by John Graham. 

Length to base of caudal, 123 mm. Depth in length, 4.2; head, 3.9. Eye in 
head, 3.5; snout, 4; interorbital, 3; maxillary, 4.5; width of mouth, 4.6; width of 
body, 2.6; depth of peduncle, 2.6; its length, 1.7; longest dorsal ray, 1.3; longest 
anal ray, 2.6; pectoral, 1.5; ventral, 1.6; caudal lobe (broken), about 1. 

Dorsal. II, 744; anal, 134. Scales, 72. Teeth, 6, 4, 2. 

Moderately compressed; a very slight keel on belly reaching perhaps halfway 
forward to ventrals. Topofhead very slightly convex; snout blunt; lower jawslightly 
included, the almost terminal mouth curved, transverse, with narrow-edged jaws, 
and little or no free lip; gill-membranes joined to isthmus under preopercle; maxillary 
only slightly oblique, not reaching front of eye; lower jaw oblique, its lower margin 


1925] SOME CHINESE FRESH-WATER FISHES 7 


straight; no barbels; opercle with a conspicuous membranous edge. Last simple dor- 
_ sal ray a slender spine, articulate towards the end and with a short soft tip; dorsal 
origin equidistant from base of caudal and middle of snout; ventral origin under that 
of dorsal; pectoral extending about two-thirds the distance to ventral, ventral about 
two-thirds to anal; caudal well forked, with narrow pointed lobes. Scales with con- 
spicuous radiating striz, lateral line complete, almost straight, running rather low, 
ending in the center of peduncle; irregularly broken and doubled in two or three 
places in this specimen. 
Silvery white, a little darker along the back. 


Ischikauia transmontana, new species 


DESCRIPTION OF THE TyPpE.—No. 8441, American Museum of Natural History, 
from Yunnan-fu, Yunnan, October 20, 1920; collected by John Graham. 

Length to base of caudal, 100 mm. Depth in length, 4.4; head, 3.7. Eye in 
head, 4; snout, 3.6; interorbital, 3.5; maxillary, 3; width of body (at shoulder), 2; 
depth of peduncle, 2.8; its length, 1.6; pectoral, 1.4; ventral, 1.7; longest dorsal ray, 
1.3; longest anal ray, 1.9; lower caudal lobe, 1. 

Dorsal, 915; anal, 11. Scales, 66. Teeth (in a cotype), 4, 4, 2, hooked. 

Compressed; a low keel between ventrals and anal, scaled to the edge and 
crossed by scales except its posterior part. Snout pointed; lower jaw projecting; 
interorbital flattish; mouth oblique; maxillary barely or not reaching to below 
front margin of eye; no barbels; gill-membranes narrowly joined to center of 
isthmus behind posterior margin of eye. Dorsal and anal without spinous rays; 
dorsal origin equidistant from base of caudal and middle of snout; pectoral narrow 
and pointed, reaching two-thirds the distance to ventral; ventral two-thirds to anal; 
caudal well forked with narrow pointed lobes, the lower the longer. Scales with con- 
spicuous radiating striz; lateral line complete, slanting down over pectoral, then 
running rather low to anal axil, rising on front part of peduncle to its center, but with- 
out sharp changes in direction as in Hemiculter. 

Pale, silvery, a little darker along back. 


ts eh 
Mee 
ye fitale= = yf ahve 
A (ai a te 7 


pein 5 
Be Bao ad am 7 ie sil 
ies Fe pion eee 


Pe > a a ee Y = La? = a i Pot on at oe 

i oe ® be ad v : 
a ea on a SA OC. 
; , '2 sega Vee Apatnenwine 


“< 


wa oat V4 7 Ree 8 tes 


SE pet apee hs uma ecntper ARAM ike ame 
. ens a e © aj i eho ; , - : aPa\s >, Atte uated gk 8 aay 
ieee by x haat ele bn) ake OR situ) Tapers, 
fobs trite RCaiite (oarteany ay myles SO 
a 4 i) ae DEAT “ee ary Z ba p< noteal Tis b abs diss ents § 
a 77 ee phe. _ Mids ehh pay? ai 
"4 f ~ F ta een” ¥ A ay 
, t ; Py L 
faa = 
. te “y ay 
2 ’ . 
- ae 
F ma 
” 
. 
« ' 


AMERICAN MUSEUM NOVITATES 


Published by 7 
Number 189 Tue American Museum or Naturat History October 7, 1925 
New Yorx City 


56.9(1181:51.7) 
FAUNA AND CORRELATION OF THE GASHATO FORMATION 
OF MONGOLIA! 


By W. D. MatruHew AND WALTER GRANGER 


In the Gurbun Saikhan piedmont basin, just north of the eastern 
end of the Altai Mountains, there are two distinct’ formations. The 
lower of these is the Djadochta formatior of Cretaceous age and con- 
tains abundant remains of the primitive horned dinosaur Protoceratops 
and its eggs. Resting unconformably upon these Cretaceous beds is a 
series of not less than two hundred feet of reddish and drab sediments to 
which the name Gashato has been assigned and which yielded the inter- 
esting primitive mammalian fauna herein described. 

The exposures of the Gashato beds are of rather limited extent in 
the type locality, which is near Shabarakh Usu on the Kweihwating— 
Uliassutai trail, and are only very sparingly fossiliferous. Mr. F. K. 
Morris, while studying the stratigraphy of the beds, discovered the first 
mammalian remains, and later Mr. George Olsen, with an assistant, 
spent several days there. The fragmentary and weathered remains of the 
largest form, Phenacolophus, were all found within a small area, and the 
rest of the collection, all diminutive forms, came from two small knolls 
not far distant. A most careful examination of the entire exposure re- 
sulted in no further discoveries. 

The fauna is later than Lower Cretaceous, and there can be scarcely 
any question that it is older than the Irdin Manha, Upper Eocene. The 
faunal list follows. 


Palzostylops iturus Order Notoungulata 
Prionessus lucifer “ — Multituberculata 
Baénomys ambiguus ‘“‘ ~ Glires 
Eurymylus laticeps ‘2 Menotyphla 
Phenacolophus fallax “2 Condylarthra 
Hyracolestes ermineus ‘““ Creodonta 


uncertain (?ereodont or carniv- 
orous marsupial) 


Sarcodon pygmexus 


The presence of a multituberculate suggests Paleocene or late 
Cretaceous age; the ancestral relationship of Palxostylops to the Wasatch 


{Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 56. 


2 AMERICAN MUSEUM NOVITATES [No. 189 


genus Arctostylops is indicative of Paleocene, Torrejon or possibly older 
age. The remaining genera throw no light upon the correlation, as they 
are but distantly related to any known types and three are of very doubt- 
ful ordinal position; but they are not incongruous with a Paleocene fauna, 
although not representing the ancestral relations to the Eocene faunz of 
Europe and America that had been anticipated. A more extended faunal 
list might modify this conclusion, but the absence of primitive perisso- 
dactyls, artiodactyls, ete., and the rather aberrantly specialized character 
of the several genera are unexpected, and fail thus far to confirm the 
hypothesis that the Eocene invaders of Europe and North America 
came from Central Asia. 


On the other hand, the minute and primitive notoungulate Palzo- 
stylops confirms the view that the South American Tertiary hoofed mam- 
mals were originally derived from the north, although undergoing a 
great secondary evolution in the Neotropical region. 


NOTOUNGULATA 
Arctostylopide Schlosser! 


Palzostylops” iturus, new genus and species 
Typr.—No. 20414, a lower jaw. 


ParRATYPE.—No. 20415, an upper jaw. Both from the Gashato formation. A 
considerable series of more or less fragmentary upper and lower jaws is also referred. 


CHARACTERS.— Dentition poked, Canines not differentiated, all anterior teeth 
of lower jaw short-pointed, sharp-crested, much compressed, and of nearly uniform 
size and character, changing gradually into the simple compressed premolars of some- 
what larger size, with prominent anterior and posterior accessory cusps. P2-4 two- 
rooted. No diastemata. No molariform premolars. Molar construction as in 
Arctostylops, the pattern much like those of the Entelonychia and toxodo nts, but the 
trigonid relatively shorter than in any of the South American genera. The crest, 
extending from middle of outer wall of talonid inward into center of basin, is well dis- 
played, as in Arctostylaps. Premolars all simple and trenchant, none molariform, but 
ps shows an incipient molariform structure. 


Upper molars with a high, straight, ectoloph crest, and prominent anteroexternal 
pillar, much as in'rhinoceroses, toxodonts, etc.; protoloph extending obliquely back 
from anteroexternal angle, and metaloph extending transversely inward from near 
middle of ectoloph; both protoloph and metaloph are short, and apparently tend to 
send out a wing from the inner end posteroexternad; but the heavy wear of the sur- 
face between ectoloph and inner end of protoloph and metaloph precludes an exact 


1In Zittel’s Griindzuge der Palzontologie, 4e Aufl., 1923, p. 614. 

2The name is intended to be suggestive of the relationship of the genus to the Wasatch Arctostylops 
and to the numerous South American genera (Notostylops, Pleurostylops, Trigonostylops, etc.) of primi- 
tive notoungulates. The species name alludes to the subsequent migration to North and thence to 
South America, of the Notoungulata. 


1925] FAUNA OF GASHATO FORMATION, MONGOLIA 3 


Fig. 1. Palzostylops iturus, new genus and species. Lower jaw with complete 
dentition, except crowns of first two incisors; superior, external and internal views. 
Type specimen, No. 20414. Four times natural size. 


description from any of our specimens. On the inner side of the tooth is a sharp 
internal cingulum, extending around the bases of the two inner lophs. Metaloph 
absent on m*, which is of trigonal outline and smaller than m?; m! smaller than m?, 
of similar pattern but less extended anteroposteriorly. The upper premolars are 
much smaller than the molars, of simple construction, sharp, high external crest on all 
and a basal internal pillar on p*“, none on p!. The canine is larger than p! but only 
its alveolus is preserved. The infraorbital foramen is above the anterior end of m!; 
the zygomatic process is stout and springs from a point between m? and m‘, 


The affinities of this genus to the notoungulates can hardly be 
questioned; but it is more primitive than any described form in the 


4 AMERICAN MUSEUM NOVITATES [No. 189 


AM.20428 


Fig. 3 
Fig. 2. Palzostylops itwrus, new genus and species. Upper jaw with premolar- 
molar series, external and palatal views. 
No. 20415, topotype. Four times natural size. 
Fig. 3. Palzostylops iturus, new genus and species. Fragment of lower jaw 
with three premolar teeth, unworn. External view (left) and internal view (right). 
No. 20428, topotype. Four times natural size. 


Fig. 4. Palzostylops sp. Anterior tooth, perhaps from upper dentition; or pi 
of a larger species. 
No. 20426 (20422 by errorin figure), topotype. Four times natural size. 


perfectly simple premolars, although quite hypsodont in comparison 
with many of the South American genera. It may be regarded as ances- 
tral to Arctostylops and through that genus to some of the South American 
Eocene Notoungulata (e.g., Leontinia, Notostylops, etc), but to the latter 
only in a broad way, as no one of the genera of the Deseado fauna 


1925] FAUNA OF GASHATO FORMATION, MONGOLIA 5 


can be cited as clearly following out the line indicated by Palzostylops- 
Arctostylops. 

An isolated tooth, No. 20422, appears to agree more nearly with the 
anterior teeth of Palxostylops than with any other known type, but is 
of much larger size, about twice the lineal dimensions of p; in P. zturus. 
The upper front teeth of P. zturus are not known, but it is to be expected 
that they would conform more or less’ to the lower teeth, which decrease 
forward regularly in size. The single, backward-pitched root shows that 
the present tooth is well forward, and it is provisionally identified as a 
first premolar, on account of the relative complexity of the crown. 

We therefore distinguish Palxostylops from any of the South Ameri- 
can families of Notoungulata by the extreme reduction of the trigonid 
and simple premolars, associating Arctostylops with it in the family 
Arctostylopide. 


GLIRES 
(Fam. indet.) 
Baénomys!' ambiguus, new genus and species 
Type.—No. 20424, a lower jaw fragment with two teeth preserved, alveoli of 
the remaining teeth, from the Gashato formation. 
CHARACTERS.—Dentition ->-z3-. Incisor rootless, stout, diastema short, ps 
much reduced, py, of size of molars, mi-2 rather high-crowned, quadrate, pattern of 


crown two transverse crests or pillars united at the base, of equal width, but the 
anterior one a little higher; m; probably similar but small. 


AM.20424 Type LSS 


Fig. 5. Baénomys ambiguus, new genus and species. Lower jaw fragment, 
superior and external views. 
Type specimen, No, 20424. Four times natural size. 


18a, intensive; évos, old; pus, mouse. 


6, AMERICAN MUSEUM NOVITATES [No. 189 


The pattern of the molars suggests the more hypsodont genera of 
pocket-mice, but has more definite suggestion of the Lagomorpha. The 
retention of p3 distinguishes the genus from all simplicidentate rodents, 
but does not exclude it from their ancestry; however, the hypsodont 
molars of heteromyid pattern are by no means what one would expect in 
a prosimplicidentate rodent. Their pattern is suggestive of an ancestral 
relationship to the Lagomorpha, but the reduction in number has gone 
further than in the rabbits, and the subequal crests, although shorter than 
in Desmatolagus of the Oligocene, do not explain the marked reduction of 
posterior moiety of tooth in the most primitive known Lagomorpha; 
nor has the jaw the slender proportions of the lagomorph jaw. It seems 
probable that the genus represents some archaic specialization of 
the rodent order, but whether lago- 
morph or simplicidentate is doubtful 
until better specimens are found. 


MULTITUBERCULATA 
Plagiaulacide 
Catopsalinze 
Prionessus! lucifer, new genus and 
species 

Type.—No. 20423, a lower jaw without 
teeth, from the Gashato formation. 

Diacnosis. — Dentition ypu. Incisor 
enlarged, scalpriform, long-rooted, stout; 
diastema short, ‘premolar’ much reduced, 
with two connate roots, followed by two 
rather small subequal two-rooted molars, 
each apparently about as wide as long. 
Lower jaw short and deep, flattened beneath 
as in other multituberculates, with prominent 
external crest beneath the masseteric fossa, 
and sharp internal crest beginning abruptly 
at a point directly behind the root of the 
incisor and extending backward to the in- 
flected angular process of the Jaw. 


AM.20423 Type 


Fig. 6. Prionessus lucifer, new 
genus and species. Lower jaw with 
alveoli of teeth; superior, internal The genus agrees with Catopsalis, 
and inferior views. Meniscoéssus and Tzniolabis (Polymas- 


T imen, No. 20423. . 
natu eee etimem No. 20423. Twice  todon) in the greatly reduced premolar, 


lrplwy, saw; joowv, inferior or weaker. The name is intended also to be suggestive of Dipriodon, 
Tripriodon and Meniscoéssus, three names which have been applied to a related genus from the Lance 
formation. The species name alludes to the light cast upon the probable age of this fauna by the 
recognition in it of a multituberculate. 


1925] FAUNA OF GASHATO FORMATION, MONGOLIA 7 


stout incisor and short, deep jaw, but differs in the smaller size and rela- 
tively smaller, shorter and probably simpler, molars. In Pézlodus and its 
allies the premolar is much enlarged, the incisor compressed and slender, 
the jaw much more elongate and narrow, the first molar probably 
more elongate. The present genus is possibly an ancestral type of the 
catopsaline subfamily (‘‘Polymastodontine’’) but is too imperfectly 
known for any precise determination of its affinities. 


MENOTYPHLA 
?Plesiadapide 
Eurymylus laticeps,! new genus and species 


Typr.—No. 20422, an upper jaw with the five cheek teeth preserved, from the 
Gashato formation. 

CHARACTERS.—Two premolars and three molars in the cheek-tooth series, 
anterior teeth unknown. Teeth low, bunodont, wide transversely, paracone and 
metacone round conical, external, no outer 
cingulum, protocone crescentic, internal, no 
additional cusps observed. Premolars smaller 
than molars, submolariform, with strong 
selenoid inner cusp, two conical outer cusps, 
the posterior one much smaller and imper- 
fectly separate, no other cusps observed. 

First molar the largest, second one-fifth 
smaller, third two-fifths smaller lineally. 
Fourth premolar equalling m? in size, third 
smaller than m°. 

The wide molars and simplicity of 
the dentition, with conical external 
para- and metacone, are unlike any Fig. 7. Hurymylus laticeps, new 
Insectivora except Leptictide, and genus and species. Upper jaw, 
suggestive of Mioclenide, the smaller P?¢™olar-molar series, p’-m’, palatal 

; : view. 

Periptychidee, Onychodectes, etc., types Type specimen, No. 20422. Four times 
distributed at present among several  »atural size. 

orders. But they suggest more than 

any other group the Eocene tarsioids and plesiadapids, to which group 
they might be referred except for the complexity of the premolars. 
The genus is provisionally placed in the Menotyphla, but its affinities 
to any known mammal, living or extinct, are not close enough to be 
decisive as to the ordinal position. 


_ ., leupis, broad; uid, mill (i.e., molar tooth). The species name alludes to the wide, short skull 
indicated by the proportions of the palate. 


AMERICAN MUSEUM NOVITATES [No. 189 


?CONDYLARTHRA 
Phenacolophus! fallax, new genus and species 


Typre.—No. 20411, a lower jaw with ps-ms, r., mi-3, 1, associated with upper 
jaw with m3 damaged by corrosion of surface, from the Gashato formation. 


; eae >) 
if ie Ay 
AN W/V LY) 
] SSR AM 2.041! Type {2 Wr I 
oS NRRL I 


Veo 
f 


ey) h +7 

ha NAM IP if ye 

el, ae ADE 
a 


A.M.204\) Type 


molar teeth. 
Type specimen, No. 20411. 


Fig. 8. Phenacolophus fallax, new genus and species. Upper jaw fragment with 


Natural size. 


Fig. 9. Phenacolophus fallax, new genus and species. Lower jaw with molar 

teeth and two posterior premolars; superior and external views. 
Type specimen, No. 20411. Natural size. ( 

Fig. 10. Phenacolophus fallax, new genus and species. Occiput, superior view. 


No. 20412, topotype, perhaps a part of the type specimen. Natural size. 


1From devaét, a deceiver; Aodos, crest, in allusion to the pseudolophiodont character of the molar and 
premolar teeth. ; 


1925] FAUNA OF GASHATO FORMATION, MONGOLIA 9 


ParatyPEs.—No. 20430, symphyseal parts of three lower jaws. A number of 
skeleton fragments are also referred here. 


CuaArRAcTEeRS.—Molars lophodont, rather low-crowned. Upper molars with six 
cusps, three in the anterior crest, three in the posterior, and a prominent mesostyle 
continuous with the posterior crest. Basal cingulum continuous around the tooth. 
(This structure is best shown in m’, which is slightly larger than the others.) Lower 
molars composed of two slightly oblique cross-crests and a small, low heel; the crests 
consisting of an external and an internal cusp, the former having an oblique wing 
running inward and forward, much as in EKohippus. Premolars narrower and smaller 
than molars, the size and relative width of the teeth increasing uniformly from p; to 
m;. Anterior premolars not known, except for the roots in the paratypes, which indi- 
cate a continuous unreduced series without diastema. Canines considerably enlarged, 
incisors not preserved, presumably small and very likely reduced in number or wholly 
absent. 


AM 20412 


/ wn, 
oe {\ 
4 
ont = Soe mn es iN 
/ a ee oF *S as ; 
\ Cem SN / \ oS ee 
—— Vv A 
\ =e Noe >< / \ a“ Y 
~ Py. ‘ I NT, ' - 
x, x x y / 
% N { Vv pee De 
See x \ ' / Wa 
EN DEA \ = i es a 
- 
Se + SS / > Le bee 
~ Vv \é pan 
ies Leet aes 


Fig. 11. Phenacolophus fallax, new genus and species. Occiput, posterior view. 
No. 20412. Natural size. 


Among the skeleton fragments doubtfully referred to Phenacolophus 
is a caleaneum of very singular type, comparable to some degree with 
Periptychinz and Ectoconus, more clesely with Isotemnide of the South 
American Eocene. It is short and massive, with broad fibular facet at . 
a very low angle with the astragalar facet, which is nearly flat and in the 
same plane with the sustentacular facet; and the cuboid facet is very 
strongly oblique, facing more internad than distad, and is moderately 
convex dorsoventrally and not concave laterally. 


The femur of a young individual is short and stocky with a fairly 
well developed third trochanter; the distal end is wide and the trochlea 
broad, short and shallow. The head of the ulna is short, notably deep 


10 AMERICAN MUSEUM NOVITATES [No. 189 


and compressed, and almost in line with the shaft. Some fragments of 
humerus and scapula, condylarthran or taligrade in general appearance, 
may belong to a larger species of this genus. 

As the name implies, the lower teeth have a deceptive resemblance 
to the lophiodont perissodactyls, but the construction of the upper molars 
can hardly be reconciled with that order, nor is there anything perisso- 
dactylic about the anterior part of the lower jaw, or the very characteristic 
calcaneum. 

The genus may be related to some of the South American Eocene 
ungulates, but probably rather to those that are provisionally referred 
to Condylarthra than to the Entelonychia or other notoungulate groups. 


Fig. 12. Phenacolophus fallax, new genus and species. Calcaneum doubtfully 
referred to this genus and species. Superior view (left) and internal view (right). 
No. 20413, topotype, perhaps part of type specimen. Natural size. 


There is a certain degree of suggestion of litoptern affinities, but not 
much. Comparisons with the hyracoids, arsinoitheres, barytheres, 
primitive proboscideans and various other groups are equally unsatis- 
factory. The genus differs so much from any known mammal, living or 
extinct, that its ordinal position is provisional, and it is impossible to 
assign it to any described family. 


?CREODONTA 
?Oxyclenidee 


Hyracolestes' ermineus, new genus and species 
Typr.—No. 20425, a lower jaw with pi-m, and roots of other teeth, from the 
Gashato formation. 


l bpat, shrew; Anorns, robber,—i.e., a carnivore of shrew-like size. 


1925] FAUNA OF GASHATO FORMATION, MONGOLIA 11 


Cuaracters.—Molars with high three-cusped trigonid and small low-crested 
heel. Protoconid highest, paraconid lowest of the trigonid cusps, metaconid distinct 
on m;, More prominent on m2. M; appears to have been as large as mp. P, small, 
high, sharp, moderately compressed, with small posterior basal cusp, rudimentary 
anterior basal cusp, no accessory cusps. Anterior premolars appear to have been 
rather small, crowded, perhaps reduced, and canine of moderate size for a creodont. 
Posterior mental foramen beneath ps. Jaw of moderate depth and rather convex 


inferior outline. 
MU YY ey Tx 


aA 


AM.20425 
Fig. 13. Hyracolestes ermineus, new genus and species. Lower jaw, superior 


and external views. 
Type specimen, No. 20425. Four times natural size. 


This is a primitive type of creodont in many respects, but apparently 
somewhat specialized in a direction unfamiliar to us, not closely paralleled 
in the Cernaysian or American Paleocene faunas. It is still less like any 
Eocene types. The minute size is suggestive of Insectivora, but the 
dentition is not like any known insectivores. 


Order Uncertain (?Creodont or Carnivorous Marsupial) 


Sarcodon pygmeus, new genus and species 


Type.—No. 20427, an upper molar. 

CuaractEers.—Molar construction as in certain creodonts and carnivorous 
marsupials (cf. Limnocyon, Sinopa and related oxyenid hyenodont genera, also 
Cladosictis and Amphiproviverra or the modern Thylacinus and Sarcophilus), but with 
a heavy posterointernal flange like that of mustelid carnassials. Size small, compar- 
able with Putorius. Protocone large, strongly compressed and extended anteroin- 
ternally, paracone and metacone strongly connate, parastyle rudimentary, meta- 
style considerably extended as a shearing crest, no external cingulum. Hypocone a 
large, prominent, anterointernal flange. 


1From capi, flesh; ’odéus, tooth. 


12 


AMERICAN MUSEUM NOVITATES [No. 189 


This tooth evidently represents a specialization paralleling the 
Mustelidz but superposed upon a predaceous adaptation with carnassi- 
form molars instead of eucreodine type with carnassial fourth premolar. 
Three known groups might give rise to such a specialization, the pseudo- 


Fig. 14. 
Sarcodon 
Pygmeus, 
new genus 
and _ species. 
Upper molar, 
crown view. 

Type speci- 
men, No. 29427. 


Four times nat- 
ural size. 


creodont Carnivora Oxyznidze and Hyzenodontide of the 
Eocene and later Tertiary; the Leptictide, Paleocene to 
Oligocene; or the carnivorous marsupials, especially the 
Borhyzenide of the South American Tertiary and Cimo- 
lestide of the northern Cretaceous. It is impossible to 
assign it to any one of these groups without some light on 
the number of upper molars and position occupied by this 
tooth in the series. The creodonts offer the closest anal- 
ogies, but the faunal association with a notoungulate is 
very suggestive of carnivorous marsupials which take 
the place of true carnivores in the South American Ter- 
tiary faunas, as they probably did in the Upper Creta- 
ceous faunas of North America. 

A certain degree of correspondence may be noted to 


Didymoconus (fam. Oxyzenide) of the Hsanda Gol fauna, as also to 
the Leptictide, but it is less suggestive of relationship. 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 193 Tue AMERICAN Museum oF NaTurAL History October 27, 1925 
New York City 


56.9(1181:51.7) 
NEW CREODONTS AND RODENTS FROM THE ARDYN OBO 
FORMATION OF MONGOLIA! 


By W. D. Matruew AND WALTER GRANGER 


The Third Asiatic Expedition secured in 1923:a considerable collec- 
tion of fossil mammals from the Ardyn Obo formation. It adds materially 
to the small faunal list published by us in 1923,” and affords a somewhat 
more exact correlation with other Mongolian Tertiaries and with the 
Tertiary succession in North America and western Europe. The 
Cadurcotherium skulls have been described by Osborn?; the smaller 
Perissodactyla and Artiodactyla and the Brontops skull will be described 
in forthcoming numbers of Novitates. 


zn 


my oI oe 
ae 


») 


Fig. 1. Hyznodon eminus. Lower jaw, external and crown views. Type speci- 
men No. 20362. Natural size. 


PE nalications of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
tion No. 57. 

21923, The Fauna of the Ardyn Obo Formation, Amer. Mus. Novitates, No. 98, December 18. 
a econ Henry Fairfield, 1923, Amer. Mus. Novitates, No. 95, October19; 1924, idem., No. 147, 
November 11. 


2 AMERICAN MUSEUM NOVITATES [No. 193 


Se 
" Dea Wy Z 
- eZ Pg aS < 
i aa 


UO 


Fig. 2 Fig, 3 
Fig. 2. Hyzenodon eminus. Part of lower jaw, external and crown views. No. 


20363. Natural size. 
Fig. 3. Hyzenodon eminus. Last upper molar, external, internal, and crown 


views. No. 20364. Natural size. 


CARNIVORA 
Hyenodontide 
Hyznodon eminus,! new species 


Typr.—No. 20362, a lower jaw with po-m; and roots of ¢-pi. 

ParatyPrEs.—No. 20363, lower jaw, ps-m2; No. 20364, upper jaw fragment, m’. 

CHARACTERS.—Size of H. minor Gervais, but teeth and lower jaw more slender 
and compressed than in specimens of that species figured by Depéret. Anterior pre- 
molars not spaced, pi: small, one-rooted; ps; moderately pitched backwardly, with 
minute anterior basal cusp as well as the large posterior one. 

This species agrees in teeth and jaw proportions rather nearly with referred 
specimens of H. minor from Euzet-les-Bains, identified by Professor Depéret. It 
belongs clearly to the brachyrhynchine group of the genus, not known in America.? 


Oxyzenide 
Ardynictis furunculus, new genus and species 


Typr.—No. 20366, front of skull and lower jaw. 

ParatypPy.—No. 20365, upper and lower jaw. (Both found interlocked.) 

CuHARACTERS.—Dentition 73-5; size of Didymoconus colgatei, but teeth less 
aberrant throughout, the protocones of upper molars higher and more compressed, 
without hypocone crest, the paracone and metacone somewhat more connate, para- 
styles of molars and metastyle of pt more prominent, p‘ smaller than m! with metacone 


1F minus =at a distance or from afar. 
y 2H. paucidens has been referred to this group, but has none of the distinctive group characters, and 
is probably a mere individual variant of H. crucians, with which it agrees in proportions of skull and 
jaws and detail construction of all the teeth, differing onlv 1: absence of p! and partially transverse 
setting of p* and p3 characters often seen as individual) variants among Carnivora and not specific 
where not associated with confirmatory evidence of their constancy. 


1925] NEW CREODONTS AND RODENTS FROM MONGOLIA 3 


rudimentary and closely connate, no parastyle, and small low protocone (lingual cusp). 
Heel of p* rudimentary. Lower molars with high, well separated, rounded, paired 
main cusps (protoconid and metaconid), low small paraconid and heel. M» somewhat 


AM.20365 


Ffg.4. Ardynictis furunculus. 
Upper jaw of type specimen. 
No. 20365, external and crown 
views. Three times natural size. 


larger than m;; p, with high main cusp slightly 
twinned, vestigial heel; p; similar but without 
heel; p2 small, one-rooted, spaced. Canines 
large, stout; incisors small, not preserved. 

This genus is nearly related to 
Didymoconus and structurally ancestral 
throughout, connecting that very aberrant 
and peculiar genus with the more typical 
Oxyzenide of the Eocene. It differs from 
any Oxyenide except Didymoconus in 
the well separated, rounded protoconid 
and metaconid and vestigial paraconid, 
as also in the corresponding but less 
obvious specializations in the upper teeth. 
It is a specialization more or less parallel 
to Dissacus among the Mesonychide, 
Apterodon among the Hyznodontide, 
and the Leptictidze among Insectivora, 


but unmistakably a derivative of the oxyzenid stock. Didymoconus is 
a further development of the same specialization, more closely parallel 


to the Leptictide. 


The adaptive significance of such a specialization would seem to be 
the piercing of somewhat hard tough shells of small prey, soft within 


== — AM.20365 


Fig. 5. Ardynictis furunculus. Lower jaw of type specimen. No. 20365, 
external and crown views. Three times natural size. 


' AM.20366 


Fig. 6. Ardynictis furunculus. Front of skull and lower jaws, young individual. 
No. 20366. Three times natural size. 


WOCs 


dpa ps p2 de 


PUP REO 


Fig. 7. Ardynictis furunculus. Crown views of upper and lower teeth of young 
individual. No. 20366. Three times natural size. 


- 
- 
- 
~ 


ESE} 
2 
o 


\ 
\ 
\ 


1925] NEW CREODONTS AND RODENTS FROM MONGOLIA 5 


the shell, requiring only a moderate amount of cutting and little crush- 
ing. It is not adapted to vertebrate prey, save possibly small mailed 
fishes; nor is it adapted to very small insect prey, such as ants and 
termites, which are handled by crushing apparatus. It is too delicate 
for the heavy-shelled mollusks but might be adapted to the smaller and 
rather thin-shelled types, and to the larger mailed insects, crickets, 
beetles and such. No very close modern parallel can be cited; some of 
the smaller Mustelide are as near as anything. 


Ischyromyide 
Ardynomys olseni, new genus and species 

Typre.—No. 20368, a lower jaw with ps-ms, from the Ardyn Obo formation of 
Mongolia, found by George Olsen in 1923. 

CHARACTERS.—Angle of jaw, so far as indicated, a deep vertical plate with the 
lower border inflected; the masseteric fossa posterior in position, its anterior border 
beneath the posterior part of mz. Jaw rather short and deep anteriorly, incisor stout 
with anterior face flattened and slightly concave. Molars short-crowned, subequal in 
width, ps slightly shorter anteroposteriorly, and m, a little longer anteroposteriorly 
than the intermediate molars. Protomere of molars, a shallow inner basin with low 
protocone external to it, low anterior and internal marginal crests, and a commissure 
behind the protocone connecting with the metamere. Metamere, a stout hypocone 
crest considerably more external than the protocone, extending internally into two 
closely parallel transverse crests with a narrow valley between, the anterior crest of 
less height and width but extending further internally, and separated at its internal 
end by a notch from the internal crest of the protomere. 

The premolar has a metamere similar in proportions to that of 
the molars but with only a single transverse crest, and the protomere 
consists of two stout cusps, the inner one extending into a crest curving 
around the anterointernal angle of the tooth, separated posterointernally 
by a notch from the metamere; the commissure behind the protoconid 
is weak. 

In terms current for more hypsodont rodent teeth, the pattern con- 
sists essentially of a main external inflection, a principal central internal 
inflection, a narrow compressed posterior internal inflection, while the 
anterior internal inflection is quite rudimentary; the premolar is well 
developed but only its posterior half is molariform and that not fully so. 

The teeth retain the partly subcircular outlines of Tillomys and 
have not taken on the rectangular proportions of Ischyromys, but the 
relationships of the genus are probably with this family, as Troxell has 
also urged for Tillomys. The posterior position of the masseteric scar 
probably indicates that the masseter was not extended forward on the 


—>~ 


Fig. 8. Ardynomys olseni. Lower jaw, internal, superior, and external views. 
Type specimen. No. 20368. Twice natural size. 


6 


1925] NEW CREODONTS AND RODENTS FROM MONGOLIA 7 


muzzle, but limited as in the more primitive rodents to the zygomatic 
arch. The flat vertical angle is as in the Paramyide, Ischyromyide, 
Theridomyide, Eomyide, retained also in many myomorphs, pedetids, 
ete. 

The genus is provisionally referred to the Ischyromyide, but without 
more complete material its true relationships can hardly be determined. 
It might represent a group ancestral to Palxocastor of the Upper Oli- 
gocene, but lacks the specialized construction of teeth and jaws of the 
beaver group. 


A.M.20370 


Fig. 9. Ardynomys chihi. Lower jaw, external view, and crown view of lower 
cheek teeth. From the type specimen. No. 20370. Twice natural size. 


Ardynomys chihi, new species 
Typr.—No. 20370, lower jaws, 1-m; r., pa-m; 1. 
ParatypEs.—No. 20371, lower jaw, dps-m; r., ps preformed in the jaw; and No. 
20372, lower jaw with heavily worn teeth, ps-m3. 
Sizz one-tenth less than the preceding, molars less robust, central inner valley of 
ms; closed by a marginal inner crest, posterior inner valley more widely open. 


Ochotonidez 
Desmatolagus robustus Matthew and Granger 

Two lower jaws, Nos. 20373-4, referable to this species, are in the 
Ardyn Obo collection. The better one shows a minute vestigial stump 
of a tooth which may be the last remnant of the lost p2 Until this is 
shown to be a constant character of the Ardyn Obo species, it is inade- 
quate to distinguish it from the Hsanda Gol D. robustus, with which the 
rest of the teeth agree in structural details. 


‘AMERICAN MUSEUM NOVITATES 


Published by 


Number 195 Tur AMERICAN Museum or NaTuRAL History Nov. 19, 1925 
New York City 2 


56.9,6(1181:51.7) 
NEW UNGULATES FROM THE ARDYN OBO FORMATION OF 
MONGOLIA! 


WITH FAUNAL LIST AND REMARKS ON CORRELATION 
By W. D. MatrrHew AND WALTER GRANGER 


In a preceding article? we described two creodonts and three rodents 
from the Ardyn Obo formation, collected in 1923 by the Third Asiatic 
Expedition. The ungulate remains from this horizon include skulls of 
an amynodont rhinoceros already described by Professor Osborn and a 
complete titanothere which he will describe in a forthcoming number of 
Novitates. The smaller and more fragmentary ungulate material, in- 
cluding four perissodactyls and three traguloid ruminants, is described 
in the following pages. 


PERISSODACTYLA 


Schizotherium avitum 
Matthew and Granger, 1923 


This species is based on a last lower molar. 
No. 20384, a poorly preserved piece of the lower 
jaw with pe and dp;-mi, probably represents the 
same species. The second premolar is unworn 
and only partly emerged. It has a compressed 
protocone with sharp and prominent anterior 
and posterior ridges, the anterior curving around 
to the antero-internal corner of the tooth, the = Fig. 1. Schizotherium 
posterior extending back to become continuous tum, last lower molar, 
with the sharp-crested, nearly median heel, ‘Superior and external 
There is a small, low-set, but quite sharply de- ~~ ”” MAGMEAL HIRE: 


; i : Type specimen, No. 19157, 
fined little cusp on the inner face of the tooth, Atdy2 Obo formation, Mon- 


lia. 
a little above the base and posterior to the point ni 
of the protocone. It occupies about the position one would expect for 
a rudimentary metastylid, corresponding substantially in its relations 


to the base of the metastylid of the molars. We have not found this 


(a 


Ly 


. qrobieatinpy of the Asiatic Expeditions of The American Museum of Natural History. Contribu- 
ion No. 58. 
2Amer. Museum Novitates, No. 193. 


2 AMERICAN MUSEUM NOVITATES [No. 195 


AM,20384 


Fig. 2. Schizotherium avitum, part of lower jaw, immature, superior and internal 
views, natural size. 
No. 20384, Ardyn Obo formation. 


cusp in p»2 of any other chalicothere, but it might be expected to 
occur in the family in view of the characteristic development of the 
metastylid in the molars. It is not present in p2of Moropus. The 
supposed milk teeth behind pz are so worn and shattered that nothing 
except the size and proportions of the teeth can be learned from them. 
The first molar is characteristically chalicothere in construction, although 
considerably worn. It compares rather closely with the type m3; save 
that it is a little broader and lacks any trace of heel. It is a little larger 
in both dimensions than m, of S. modicum. 


? Schizotherium species 


No. 20385, a small jaw fragment, has an unworn molariform tooth, 
the alveolus of a larger two-rooted tooth behind it, and behind the alveolus 
a part of a pocket for a preformed tooth. This tooth may perhaps be 
dps, although no exact comparison can be made with the shattered dps 
of No. 20384, and the length and verticality of the roots suggest a perma- 
nent molar. The tooth has the construction of the Schizothertum molar, 
with a small transverse hypoconulid heel, not unlike the m; of S. modi- 
cum; but it is considerably smaller than the molars of S. modicum and 
S. avitum, and the alveoli behind it indicate a considerably larger tooth. 
If a permanent tooth, it would necessarily be mi, and the proportions of 
m, to me would differ considerably from those of the known species. 


£ 
"G80 ‘ON ‘uourIoeds edAT, 


‘OZ1S [BINYCU ‘SMITA UMOIO PUL 194NO “Y}99} AvpouroId YIM vel Joddn jo 4rd ‘soroeds puv snus Mau ‘snzino uopojoowwg *G “BIT 
“U01}BULIO} OGG UAPIY 

‘ous [BANYUU “Y}99} JO MOTA MMOD ‘7EE0Z “ON ‘uoUTIDeds ody ‘wel soddn ‘saoods mou ‘smdoour vopojog “p “Buy 
*UOI}BUIIOJ OGD UApry ‘egeoz ‘ON 

“OZ]S [BANICU “SMOTA [BULOJUT pus 1oModns ‘Mel IaMo] Jo JuoWTeIy ‘soroods wnwayjozyag “g “Bq 


*g “317 


\ 


uo--- eee --- -- 


———-—-—-— — 


ie 


ni Us es 


NG 


Wed 
= 


\ 
| 


| 


= \ 
hese ale ee 
— 4ns 
Z =u =x 


1 AMERICAN MUSEUM NOVITATES [No. 195 


Colodon inceptus, new species 


Typr.—No. 20357, upper jaw with p*-m? of the left side. 

CHARACTERS.—Size of C. occidentalis (Leidy) (as represented by the palate 
figured by Osborn and Wortman), molars similar in construction, except that the 
posterior crest of m* runs nearly straight from paracone to hypocone, while in C. 
occidentalis it curves outward just behind the paracone. M2? shows a similar difference 
but less marked. Inner cusp of p‘ single, on p* partly divided but still closely con- 
nate, as in p* and p‘ of C. occidentalis. 


This species is a little more primitive than the C. occidentalis of the 
Oreodon beds. The type of Leidy’s species is recorded as from the 
Titanotherium beds but is known only from the last lower molar. It 
might be equivalent to C. znceptus, although probably not cospecific. 

Among the fragmentary lower jaws and teeth, there are two or three 
specimens that may belong to this or the following genus, but we can- 
not identify them with certainty. 


Paracolodon curtus, new genus and species 

Typr.—No. 20355, upper jaw with p®-! and parts of alveoli of canine and first 
molar. 

_ Generic CHAracters.—Premolar construction as in Colodon but the premolars 
progressively reduced forward, p! absent, postcanine diastema short and molars rela- 
tively large. P? is proportioned like p' of Colodon, but the construction is more as in 
p?, with two outer cusps, two inner cusps, the anterior one vestigial, the posterior 
connecting crest incomplete. P? constructed much as in Colodon but nearly quadrate, 
the anteroposterior and transverse diameters subequal. P* like that of Colodon. 
Narial notch extending far back, at least to a point above p*. Premaxilla extending 
backward to a point above p’, probably farther. 

Specrric CHARACTERS.—About the size of C. inceptus and p‘ of nearly identical 
size and proportions. A slight groove obscurely separates the inner cusps; the postero- 
external cusp less flattened than in C. inceptus, in which it has begun to resemble the 
corresponding part of the molars. Inner cusps of p* well separated, of subequal size 
and prominence, whereas in C. inceptus the posterior cusp is larger. Postcanine 
diastema less than the length of p?*. Transverse diameter of molar roots fifteen per 
cent. greater than that of p*; in Colodon they are nearly equal. 


AFFINITIES OF Colodon AND Paracolodon 


The relationship of the Mongolian and of Borissiak’s species to the 
American genus appears to be beyond reasonable doubt. Colodon in 
turn appears to be rather nearly related to “Desmatotherium” mongo- 
liense, which is certainly near to D. guyotiit of the ? Washakie;! 


1Recorded as Bridger, but it may have come from the Washakie. It has not been recognized in any 
of the subsequent collections from the Bridger and appears to be decidedly later in type than the 
Bridger lophiodonts. 


1925] NEW UNGULATES FROM MONGOLIA 5 


although congeneric only in a broadly inclusive sense. We are unable to 
agree with Peterson that Desmatothertum should be removed from the 
Helaletide to the Hyracodontide, but a general discussion of the affini- 
ties of this group of lophiodonts may better be postponed until the 
Irdin Manha material is more fully described. Among the European 
genera, Chasmotherium comes nearest in molar structure, but the pre- 
molars and anterior teeth differ widely. Lophiodon differs in the molar 
construction, as pointed out by Stehlin and Depéret, and is decidedly 


Fig. 6. Ardynia precox Matthew and Granger, lower jaw with premolar and 
molar dentition, crown and outer views, natural size. 
No. 20358, Ardyn Obo formation. 


more primitive in its premolars. Both Desmatotherium mongolicum and 
Paracolodon curtus had apparently the same peculiar recession of the 
narial notch that is seen in skulls of Helaletes collected by one of us in 
1904, and has been observed by Peterson, 1919, and Troxell, 1922, in the 
type of this genus. The character is not known in the remaining species 
of this group, but may well have been distinctive of all of them, and 
would separate them from Lophiodon, Chasmotherium and Hyrachyus, 
as well as from Systemodon and Isectolophus. The beginnings of this 
recession are doubtfully seen in Heptodon. 


6 AMERICAN MUSEUM NOVITATES [No. 195 


Ardynia preecox, Matthew and Granger, 1923 


This small rhinocerotoid is represented by a number of fragmentary 
jaws and teeth. No. 20358, a lower jaw with ps-m; and alveolus of ps, 
shows the marked reduction in the lower premolars corresponding to 
that in the upper premolars of the type. The combined length of the 
three premolars was about half that of the three molars. A small round 
alveolus indicates that ps was reduced to a vestigial tooth. Ps; is reduced 


AM.19156 


Fig. 7. Ardynia precoxr Matthew and Granger, last upper molar, external and 
crown views, natural size. 

Paratype, No. 19156. 

Fig. 8. Ardynia precox Matthew and Granger, part of upper jaw, outer and 
crown views of premolars and first true molar, natural size. 

Type specimen, No. 19156. 


anteriorly somewhat as in p2of Hyracodon. The second and third molars 
are subequal, the first considerably smaller and about equal in size to 
ps. No. 20386, lower jaw fragment with p3-s, has alveoli of pe, c1, and 
part of an incisor alveolus. The front teeth correspond in size and 
arrangement to Hyracodon. 


1925] NEW UNGULATES FROM MONGOLIA ij 


““Prothyracodon”’ uintense Peterson resembles Ardynia to some 
extent, but appears to be more brachydont, and the premolars, so far 
as one may judge from the milk teeth, are not reduced. Peterson’s 
species can hardly be congeneric with P. obliquidens (Scott and Osborn), 
in which m retains the free posterior flange of the ectoloph, but may be 
comparable with Prohyracodon Koch of the Eocene of Hungary. 


ARTIODACTYLA 
LOPHIOMERYX Pomel 


Typr.—L. chalaniati from the Oligocene of France. 

CHARACTERS.— Distal end of fibula separate from tibia. Navicular and cuboid 
united, median pair of metacarpals and metatarsals separate, not closely appressed, 
lateral metacarpals (and probably metatarsals) complete with slender shafts, distal 


Vr 


AM.20387 I 


Fig. 9. Lophiomeryx angarz, new species, upper jaw, outer and crown views, 
natural size. 
No. 20387. Last molar reversed from opposite side of jaw. Ardyn Obo formation. 


keels very sharp posteriorly but not extended in any degree over anterior face of 
metapodials. Molars brachydont, with mostly crescentic cusps, hypocone of m3 
reduced or vestigial. Upper molar crescents in obliquely set pairs, outer styles 
prominent, anterior rib sharp and prominent, no trace of posterior rib. Fourth upper 
molar of two crescents, third non-crescentic, composed of three outer cusps in a row 
and a median internal cusp, the outer cusps partly united into a crest but more or less 
clearly distinguishable, middle cusp the highest. Third upper molar similar but no 
inner crests and anterior and posterior outer crests weaker. 

Lower molars with imperfect inner crescents, the anterior flange of metaconid 
and posterior flange of entoconid in varying degree imperfect or absent. Heel of 
ms; with two subparallel crests running forward from apex, the inner one imperfect, 
no inner cusp on heel. 

Premolar composition of anterior, median, and posterior inner transverse crests, 
the last cingular (distinction from Prodremotherium) in position and imperfect, so 
that no closed pocket is formed between it and the median crest (distinction from 
Eumeryx); the anterior crest also little developed. 


7 A.M.20376 


YAW 


1! 
AMX. 
—. 


OTA 4 : 
Det 
= 4) Sti am aa i IN 

S \ = i TO Hae 2 


Fig. 10. Lophiomeryx angarz, new species, lower jaw, outer and crown views, 


natural size. 
Type specimen, No. 20376; the front teeth supplied from No. 20387. 


A.M. 20387 


MMU: 


= 
<UL 


Fig. 11. Lophiomeryx angare, new species, foot bones associated with upper and 


lower jaws. 


A, metacarpal iii, proximal, dorsal, and distal views. B, corresponding views of metatarsal iv. 


C, metacarpals articulated, incomplete, showing completeness and relative size of lateral digits. 
natural size. No. 20387. 


8 


All 


1925] NEW UNGULATES FROM MONGOLIA 9 


The characters specified above show, first, that we are dealing with 
a member of the pecoran-traguline division of the Selenodontia; second, 
that it is a traguloid, not one of the higher groups; third, that its affini- 
ties lie with Prodremotherium, Gelocus, Leptomeryx, and other early 
Oligocene genera of that group, but that it is more primitive than any 
of them in the complete separation of both metacarpals and metatarsals, 
as well as in the imperfection of the inner crescents of the molars, and 
various other details of composition of the teeth. 

Gelocus retains a more primitive construction in the simple heel of 
ms; and the simple trenchant premolars of traguloid type, but is much 
more progressive in reduction of lateral digits, union of metatarsals into 
a cannon bone, and compression of metacarpals, and in the transverse 
set of the upper molar crescent pairs. Prodremotherium has the molars of 
more distinctly cervid construction, premolars more progressive, meta- 
tarsals united and metacarpals compressed. 

The above diagnosis is based upon the Ardyn Obo species, chiefly 
upon L. angarx, of which we have numerous parts of jaws and skeleton 
bones, some of the foot bones partly articulated, found at two or three 
localities along the Chinese Post-road near Ardyn Obo. As there is no 
admixture of other artiodactyl remains, the association is reasonably 
certain. These specimens agree closely in dentition with Lophiomeryx 
from the Phosphorites received from the Munich Museum in 1896 and 
identified by Doctor Schlosser. As to the foot-construction, Schlosser 
remarks in the last edition of Zittel’s Grundziige that the ‘“‘metapodials 
are separate,’”’ and places the genus in the traguline group, but we have 
failed to find any more detailed description or figures of the foot bones 
which might verify the generic reference of these Asiatic species. The 
close correspondence in teeth, however, renders it a highly probable one. 

Lophiomeryz is typical from the Cournon (Stampian) but our com- 
parisons are with the Phosphorites species referred to it. We have not 
personally verified the accuracy of this reference, but it rests upon the 
high authority of Schlosser and Stehlin. 


Lophiomeryx angare, new species 
Type.—No. 20376, lower jaw, associated with No. 20387, a series of jaws and 
skeleton bones found together. 
Species Distincrion.—P;-m3=60; mi-»=32. M® with vestigial hypocone; 
p’-m?=49; m!~=29. P, single-rooted, a short diastema behind it, a long one in 
front of it. 


10 AMERICAN MUSEUM NOVITATES [No. 195 


Lophiomeryx gobiz, new species 

Type.—No. 20381, lower jaw, associated with various fragmentary lower and 
upper jaws found together; Nos. 20379-80, upper jaws. 

Seecies Distinctions.—M!3=24 mm. Hypocone of m’ well developed, al- 
though smaller than protocone. Three lower premolars. Metaconids of lower 
molars crescentic, the anterior crest complete. Premolar and molar construction 
otherwise resembling that of the larger species. 


i i 


AM.20379 ! 
mi! . ? 


A.M.20383 
Fig. 12. Fig. 13. 


Fig. 12. Lophiomeryx gobix, new species, upper jaw, outer and crown views, 
natural size. No. 20379. 

Fig. 13. Miomeryz altaicus, new genus and species, upper jaw, outer and crown 
views, natural size. 

Type specimen, No. 20383. 


There is some doubt about the reference of this species to Lophio- 
meryx, as it lacks p; and also one characteristic feature of the genus, the 
imperfect antero-internal crescent of the molars. 


Miomeryx altaicus, new genus and species 

Typr.—An upper jaw, No. 20383, with p?-m?. 

Nearly related to Lophiomeryx, but distinguished by more brachydont molars, 
less development of inner cusps of premolars, stronger external ribs on anterior halves 
of molars, an external rib on p‘, more rugose enamel. Size smaller than L. gobizx; 
the length of p?-m’=34 mm. 

P2 has neither internal root nor cusp. Dimensions of p? are tr., 4.0 Xa-p., 7.5; 
of p, tr., 6.4Xa-p., 5.2, with medial anteroposterior width of 3.5. Corresponding 
dimensions of p? in L. angarz are 6.08.5; of p*, 7.66.1, with medial width of 4.9. 


Ratios M. altaicus L. angarz 
P3 a-p 100 100 
PS tr. 53% 71% 
P4 a-p 100 100 
PP ix, 123% 124% 


Pt med. width 67% 80% 


1925] NEW UNGULATES FROM MONGOLIA it 


The molars of this genus compare nearly with specimens from the 
Phosphorites identified by Doctor Schlosser as Bachitherium. We have 
no premolars for comparison, and Filhol’s figures of the upper dentition 
of the type species B. znsigne do not agree well in the molars and have 
obviously a distinct type of premolars, the inner crescents well developed 
on both p* and p*. Filhol’s figures are obviously not very accurate and 
represent referred specimens which may well belong to another genus; 
but we have not found any published redescription of the genus or any 
accurate figures. Pending an adequate revision of the Phosphorite 
artiodactyls, it appears better to record the Gobi genus as new. It is, 
however, nearly related to certain material from the Phosphorites, 
whatever name properly applies to that material. 

Leptomeryx, the nearest American genus, differs in the better devel- 
opment of inner cusps on the premolars, in higher crowned molars with 
weaker anteroexternal rib and less rugose enamel. 


List or ARDYN Oso Fauna As IDENTIFIED Marcu 1,1925 
CARNIVORA 


Oxyenide 
Ardynictis furunculus, n. g.,N.Sp........... front of skull and lower jaws 
Hyenodontide 
Hyznodon eminus, 0. Sp......... 00.00. eee lower jaws, upper molar 
? Canide 
DGG AHAIS Boece ok ioe s,s aaa a 6 oo s)0 + +5 1 RONG O1 LOWER TW: 
RopDENTIA 
? Ischyromyide 
Ardynomys olsen, De @.,. DSP. 2.0/6 uisin 3 oo lower jaws 
Ardynomys chihi, n. sp.......................lower jaws 
Ochotonidze 
Desmatolagus robustus M.and G............ lower jaws 
PERISSODACTYLA 
Titanotheriide 
Ba pS WOOT AOSDOIT: o..c55 + =< 54:8 «tka heey skull and jaws, ete. 
Chalicotheriide 
Schizotherium avitum M. and G.............. .™s3; part of lower jaw 
Helaletide 
Coladan Teentus NASD. «<5. < voce, » edad eae Be upper jaw 
Paracolodon curtus, 0. g., ND. SP.........6.608- .part of upper jaw 
? Hyracodontide 
Ardynia precor M.andG....................parts of upper and lower jaws 
Amynodontidze 
Cadurcotherium mongoliense Osborn......... . Skulls, jaws, part of skeleton 


? AMBLYPODA 
Gene Wes, vol715 ae ek ace ee fragment of skull with horn 


12 AMERICAN MUSEUM NOVITATES [No. 195 | 


ARTIODACTYLA 
? Anthracotheriidz 
(Sen eD S381, Set here lates eters wneud lower molar 
Hypertragulide : 
LO DILOMETYL GNYOTL, Di: SPs ee so. 2 ac ese ee numerous jaws, parts of feet 
Lophiomerya gobie, pospli ck. Flees upper and lower jaws 
Miomeryzx altaicus, n. g.,n. Sp...............upper jaw 
CHELONIA 
Teshude wmeombis Mi. sid Go no. ate ws ws parts of carapace and plastron; 


lower jaw. 


CORRELATION OF THE ARDYN OBO 


Present evidence indicates that this fauna is older than the Hsanda 
Gol, approximately equivalent to the Titanotherium zone in the American 
succession and probably to the Sannoisian or Ludian faunas of Europe. 
There is, however, little or nothing on which to base any exact correla- 
tion with the European succession. The absence of any higher types of 
Artiodactyla, of true rhinoceroses, of higher Carnivora, with one doubt- 
ful exception, and of any Oligocene rodents, gives the fauna a broadly 
Eocene aspect, while the specialized stages reached in the various Eocene 
phyla point to its being at least at the end of the Eocene. It appears 
distinctly older than the Hsanda Gol in the survival of titanotheres and 
perhaps of amblypods, absence of the Rhinocerotide, of the higher 
Carnivora and more specialized rodents of that fauna, and especially 
in the tragulid ruminants in contrast to the cervid Humeryx of the 
Hsanda Gol. The creodonts are more primitive stages in the same phyla. 
It is decidedly more advanced than the Irdin Manha and Shara Murun 
faunas as shown by the more progressive titanotheres, tragulids, lophio- 
donts and hyznodonts. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 196 Tur AmericaAN Museum or Natura History Nov. 20, 1925 
New York City : 


56.9(1181:51.7) 
NEW MAMMALS FROM THE SHARA MURUN EOCENE OF 
MONGOLIA! 


By W. D. MatTHEew AND WALTER GRANGER 


The Shara Murun formation is typically exposed about 100 miles 
south of the Irdin Manha Eocene and carries a fauna of corresponding 
type, including the very characteristic titanotheres which enabled the 
expedition to refer it in the field to the end of the Eocene. It was at 
first regarded as a probable equivalent of the Irdin Manha, but, as first 
observed by Professor Osborn in his studies upon the titanotheriids of 
the two faunas, and fully confirmed by the present review of the smaller 
mammals, it is quite distinct and represents a later phase of the Upper 
Eocene. 


CARNIVORA (Creodonta) 
Hyenodontide 


Pterodon hyznoides, new species 


Type.—No. 20307, complete skull from Shara Murun formation at Ula Usu, 
Mongolia. Third Asiatic Expedition, 1923. 

CHARACTERS.—Dentition 2.1.4.3. Premolars enlarged and robust as in 
Hyznodon. Last molar transverse, as in P. dasyuroides, but smaller and more 
vestigial in cusp construction. Protocones of molars 1-2 probably present, but less 
prominent than in P. dasyuroides, and worn off in the type and only known specimen. 
P! one-rooted as in Pterodon and the Eocene hyenodons. Molar series extended on 
maxillary portion of zygomatic arch behind the back of the palate, but not so much so 
as in Hyxnodon. Skull broad and robust, comparable in proportions with Pterodon; 
basicranial construction as in Hyzxnodon except for greater width, this portion of the 
Pterodon skull being undescribed so far as we can find. 


This species is in all observed particulars an intermediate between 
Pterodon and Hyzxnodon. It is very well distinguished from the European 
species of either genus; its precise relations to the several more or less 
intermediate species from the Fayim ‘fluviomarine’ fauna are not 
certain. These species as distinguished by Osborn? are based upon the 
lower jaws. Schlosser? comments upon their intermediate position 
vee ra goin of the Asiatic Expeditions of the American Museum of Natural History. Contribu- 


59 
2Osborn, H. F., 1909. Bull. Amer. Mus. Nat. Hist., XXVI, pp. 415-424 
3Schlosser, M., 1911. Beitr. z. Pal. u. Geol. Osterreich-Ungarns u. d. Orients., XXIV, pp. 87-88. 


NK ean 
STAB e 


MY 
‘| 
\\ 


“i 
ae 
AM 20307 
1 H 
. J! 
c i3 


Fig. 1. Pterodon hyznoides. Skull, side view, one-third natural size. Type 
specimen No. 20307. 


AM 20307 


x {\ N 


\ 


Fig. 2. Pterodon hyenoides. Type skull, palatal view, one-third natural size. 


s 


1925) NEW MAMMALS FROM .EOCENE OF MONGOLIA 3 


between the two genera. None of the species, however, appears to be 
very near to P. hyxnoides. x 


? Mesonychide 
Olsenia mira, new genus and species 


Typr.—No. 20303, left astragalus. 

CHARACTERS.—Size comparable to the tapir. Trochlea broad and rather shallow, 
both keels prominent, a prominent external flange on outer face of outer keel. Neck 
short, distinct, head about as deep as wide, intermediate in type between perissodactyls 
and mesonychid creodonts, with a considerable cuboid facet, separated by a marked 
keel from the narrow but deep navicular facet, the keel running parallel with the 
facet for distal end of caleaneum. Astragalo-calcanear facet long, narrow, deeply 
concave, with parallel margins, sustentacular facet set far back on the body of the 
astragalus, so that its posterior border is strongly rolled down under the back of the 


Fig. 3. Olsenia mira. Astragalus, internal, distal and dorsal views, one-half 
natural size. Type specimen, No. 20303. 


trochlea. The body of the astragalus is singularly shallow on its external side, the 
space between trochlear and astragalo-calcanear facets very small; on the inner side 
the depth of the body is more nearly normal. 

We are unable to refer this astragalus to any known genus, and are 
uncertain as to its family relationships. It is not a perissodactyl or 
artiodactyl, although it has some points of resemblance to both orders. 
It is unlike any normal creodont or carnivore, but it conforms to the 
constructive features of Mesonychide in most respects. It is distin- 
guished from most mammals by the peculiar type of cuboid facet, which 
is characteristic of Mesonychide and of Artiodactyla. ‘The distal gin- 
glymus of artiodactyls is lacking, and the bone is in other ways much 
nearer the mesonychid type, although the astragalar foramen is absent 
and various other differences appear. The proportions are not unlike 
those of the Urside, but the bear astragalus never shows the sharply 
keeled separation between navicular and cuboid facets. 

To this genus is possibly referable No. 20319, an upper premolar of 
mesonychid type, recalling in its size and proportions the fourth premolar 


4 AMERICAN MUSEUM NOVITATES [No. 196 


of Dissacus or of Pachyena. It appears small in proportion to the 
astragalus; it may, however, be a milk tooth. These two specimens 
sufficiently indicate the survival of aberrant Mesonychide in the Shara 
Murun. 


GLIRES (Rodentia) 


The only rodent specimens in the collection are a jaw fragment 
with one molar, and two isolated incisors. 


Lagomorph, indet. 

A fragment of lower jaw with one molar, No. 20326, from the Shara 
Murun beds, is comparable with Desmatolagus, but is too incomplete for 
definite identification. It differs from that genus in a much less hypso- 
dont crown and more divergent roots, and probably represents a primi- 
tive undescribed genus of Lagomorpha. 


PERISSODACTYLA 
Helaletide (Colodontide) 
Deperetella, new genus 


We name this genus in honor of Professor Charles Depéret, whose 
discoveries and researches have added so greatly to our knowledge and 
understanding of the Tertiary faunas of Europe. 


Typre.—D. cristata infra. 

Cuaracters.—Molar pattern related to that of Desmatotherium and Colodon 
but more specialized in the direction of sharp transverse crests; the premolars more 
molariform. Limbs and feet long and slender, manus and pes tridactyl. The upper 
molars have high, sharply compressed, nearly transverse crests curving around to 
join externally. Parastyle and free portion of ectoloph much reduced, joined by a 
continuous basal cingulum which passes around the anterior, inner and posterior 
margins of the tooth. Upper premolars with sharp, well separated, transverse crests, 
external crest higher, somewhat convex externally, distinct parastyle and encircling 
basal cingulum. P! small, ? one-rooted. Lower molars with wide, sharp transverse 
crests, no connecting ridge, a low, transverse, crested cingulum behind the posterior 
crests of all three, but no heel on ms. External and anterior basal cingula, inner 
cingulum imperfect to absent. Fourth lower premolar fully molariform, third partly 
so, with strong connecting crest between the transverse crests and well-developed 
paraconid. First and second premolars compressed, two-rooted, the second submolari- 
form and longer than the posterior premolars. Diastema short, canine of moderate 
size, incisors smaller. ; 

Humerus of moderate proportions, radius much elongate, one-third longer than 
humerus, manus tridactyl, the fifth digit reduced to a small, short, strongly divergent, 
vestigial nodule. Pes long and slender, all the tarsals relatively long as well as the 
metatarsals. Lateral digits reduced much as in Colodon. Phalanges short. 


1925] NEW MAMMALS FROM EOCENE OF MONGOLIA 5 


AM20290 


Fig. 4. Deperetella cristata. Upper jaw, external and crown views, two-thirds 
natural size. -Type specimen, No. 20290. 
The outlined portions of teeth restored from No. 20293. 


; ——= = ae — ~ te 
: Seer sis <i 
AM 20291 ae py 


Fig. 5. Deperetella cristata. Lower jaw, external and crown views, one-third 
natural size. No. 20291. 


This remarkable genus is clearly related to Colodon, with which it 
agrees in teeth and feet more nearly than with any other described peris- 
sodactyl. It is, however, very well distinguished by the sharply crested, 
peculiar pattern of the molars, entire absence of heel on mz, ete. It is 
the largest and most specialized of the colodont group. The character 
of the entocuneiform is tapiroid and not rhinocerotoid or equid; it 


6 AMERICAN MUSEUM NOVITATES [No. 196 


would probably be more nearly approached by Colodon if the bone were 
known in that genus. 


Deperetella cristata, new species 


Typr.—No. 20290, parts of upper jaws and fragments of skull, young individual, 
with unworn molars and premolars. 

ParatyPes.—Nos. 20291, lower jaw; 20292, lower jaws; 20295 and 20305, hind 
feet; 20294, humerus, radius, part of forefoot. 

Horizon AnD Locatity.—Ula Usu beds of the Shara Murun formation. 

Size comparable to modern Brazilian tapir, but limbs much longer. Length of 
Pi-mM3= 141 mm. 


Hyracodontide 
Cznolophus, new genus 


Upper teeth rhinocerotoid in structure. Upper molars with very oblique proto- 
loph and metaloph, the metaloph relatively reduced. Posterior flange of ectoloph of 
m? reduced to varying degree. P* with two strongly oblique transverse crests, the 
protoloph higher and directed toward the parastyle, the metaloph directed toward the 
paracone or anterior external rib; no distinct postero-external rib. P? with the trans- 
verse crests united internally to protocone but similar relations externad. 

Lower molars with crests somewhat more oblique than in Prothyracodon, no 
heel on mz. 


This genus appears to be rather nearly related to Prothyracodon of 
the Uinta, but the construction ofthe premolars does not agree entirely 
with that shown in Osborn’s figures (‘‘ T'riplopus’’), and the proportions 
of m? are considerably different. The lower teeth are characteristically 
like those figured by Osborn. Skull and feet are unknown, but if, as 
seems probable, they are like those figured by Peterson as Prothyracodon 
obliquidens, the nares are not extended backward and the manus is 
tridactyl. - 


Cenolophus promissus, new species 

Typr.—No. 20297, left upper jaw with p?-m? moderately worn. 

Horizon anp Locatiry.—Shara Murun formation, Ula Usu, Expedition 1923. 

SpEciEs CHARACTERS.—Size medium, p?-m?=61, m= =43 mm. M§ trapezoidal, 
with moderate posterior flange or ectoloph. Metaloph of p‘ shorter, metaloph of p* 
imperfectly separated. No. 20304, mi-2 in a fragment of jaw from the Shara Murun 
beds, is probably of this species. There is no trace of paraconid cusp on the lower 
molars, the anterior crest sweeping down uniformly to the base of the tooth at the 
antero-internal angle. 


1925] NEW MAMMALS FROM EOCENE OF MONGOLIA 7 


Fig. 6. Ceznolophus promissus. Upper jaw, external and crown views, natural 
size. Type specimen, No. 20297. 


Cenolophus obliquus, new species 


Typr.—No. 20296, upper jaw with dp?-m!, part of m2, p** preformed and chamber 
of m’, Shara Murun formation, Ula Usu, Mongolia. ? 

CuHARACTERS.—Larger than the two preceding species; p*-? with strongly oblique 
transverse crests united at protocone on p*. P3-m? approximately 66 mm.; estimated 
length of p*-m?=83 mm. 


Cznolophus progressus, new species 


TypEe.—No. 20298, left upper jaw with m!-* moderately worn. Same horizon 
and locality as C. promissus. 
Species CHARACTERS.—Size smaller, m3 =35 mm. M? subtrigonal, the posterior 
flange of ectoloph vestigial. 
* No. 20309, lower jaw fragment with mi-2 (possibly dps-mi), may belong to this 
species, although smaller than the type if the teeth are both permanent molars. It 
has the same characters as the jaw fragment referred to C. promissus. 


Cenolophus ? minimus, new species 


No. 20310, a jaw fragment with m-2, represents a species provisionally referable 
to this genus, but decidedly smaller than any of the preceding. The molars have the 
same characters as those referred to the preceding species, but the length of mi- is 
15.5 mm., of mi-3, estimated at 24 mm. 


1925] NEW MAMMALS FROM EOCENE OF MONGOLIA 9 


ARTIODACTYLA 
Hypertragulide 
Archeomeryx optatus, new genus and species 


TypE—No. 20311, skull and jaws; paratypes, Nos. 20320-20322, articulated 
skeleton, skulls and jaws, all found associated in one pocket in the Ula Usu locality, 
Shara Murun formation, later Eocene. 


AM2031| 


TAIN 

: % uy! 
MN, fi: uw Vif \ 
er (2) : 


fyi J Uy 
S\N XG \ 


= Zz 


SO Cy 


ae 


Fig. 10. Archxomeryx optatus. Upper and lower dentition, external and crown 
views, twice natural size. Type specimen, No. 20311. 
Front teeth supplied from No. 20322. 


Cuaracters.—Dentition +35. Upper molars tetracuspid, brachyselenodont, 
the pairs of cusps somewhat obliquely set, the posterior wings of the inner crescents 
low and incomplete. Parastyles prominent, continuous with anterior wing of antero- 
internal crescent; mesostyles well developed; metastyles very weak or absent on 


10" AMERICAN MUSEUM NOVITATES [No. 196 


m2, moderately developed on m’. Anterior external rib prominent, posterior rib 
absent. Postero-internal crescent of molars smaller than antero-internal; inner 
cingula well developed; molars increasing somewhat in size from first to third. 

‘Upper premolars three, pt with complete inner and outer crescent, p? with inner 
cusp and three-cusped outer crest, p? with outer crest imperfectly three-cusped, no 
inner cusp. Postcanine diastema rather short, canine small, pointed, recurved, one- 
rooted, three very small, pointed, simple incisors in front of it, a little spaced, but no 
considerable diastemata. 

Lower incisors three, very small and like the upper incisors; lower canine close 
behind them and but little larger. Pi with short diastemata before and in some in- 
dividuals slightly spaced behind, small, simple caniniform, but with more flattened 
crown. P: two-rooted, with three strongly compressed cusps in line, the central one 
highest. P3 similar, but with anterior crested cusp inturned to a transverse ridge, and 
a suggestion of posterior transverse ridge behind the posterior cusp. Ps with anterior 
crest partly inflected, a distinct cusp internal to the central and the transverse crest 
at posterior border further developed and sweeping around to become connected with 
the median inner cusp; a small oblique crest also extending posterointernad from the 
posterior of the three original cusps. 

Lower molars tetracuspid, brachyselenodont, with the pairs of cusps set somewhat 
obliquely, third lobe of last molar looped, no internal gusp. The crests of the loop 
correspond to the wings of the external crescents of the molars, the posterior wing 
sweeping around to an internal basal crescent. 

Skull about as large as that of Tragulus javanicus, but much more primitive and 
generalized. The braincase is much smaller and the orbits are smaller; the orbits 
appear to be enclosed behind; from the postorbital process the postorbital crests run 
obliquely inward and backward to meet at the anterior ends of the parietals and form 
a low continuous sagittal crest. The occipital crest appears also to be prominent, as 
usually in Eocene mammals. The skulls are so crushed and incomplete that little 
more has been determined from them than the above. The lower jaws are rather 
heavier than in Tragulus and shorter anteriorly, the muzzle in front of pz being little 
more than half as long. 

Skeleton somewhat larger than that of Tragulus, the femur an eighth longer and 
heavier in the shaft, the tibia of the same length but considerably heavier shaft. 
Radius and ulna separate, the ulnar shaft with less breadth but greater thickness 
in its distal portion than Tragulus or Hyxmoschus, the shaft of the radius more slender 
than in Tragulus but the proximal and distal ends about the same size. Fibula re- 
duced to proximal and distal vestigia, the latter with long, slender splint, but shaft 
apparently not complete. Navicular and cuboid codssified, carpals all separate. 
Metacarpals and metatarsals separate, the lateral digits complete and moderately 
reduced, less relatively than in Hyzxmoschus, much less than in Tragulus. Distal 
keels of metapodials not extended over dorsal face. Phalanges proportioned much as 
in Tragulus, but both fore and hind feet are of considerably larger size. 

Ischia constructed as in Hyxmoschus and the pees in contradistinction to 
Tragulus. Tail long (cf. anoplotheres). 


This genus is of exceptional interest, as it appears to be an approxi- 
mate ancestral type for the pecora, It has assumed the characteristic 
pecoran-traguline character of the united naviculo-cuboid, but it still 


1925] NEW MAMMALS FROM EOCENE OF MONGOLIA 11 


retains the separate median pair and complete lateral pair of digits, the 
ulnar and fibular shafts are more primitive than in any pecora, the upper 
incisors are still retained, the premolars are of primitive pattern, the 
molars brachyselenodont. It lacks any of the various aberrant speciali- 
zations which exclude all of the Eocene genera hitherto described from 
direct ancestry to the pecora and, as fortunately the principal osteo- 
logical and dental characters are determinable from the exceptionally 
complete material, the affinities of the genus can be very satisfactorily 
appraised. So far as the higher ruminants are concerned, it affords 
tangible and very convincing proof of the theory of an Asiatic dispersal 
center. 


» i arg no fe ti hy 
he on ARG? 
cel | (a Ey ipphena 


a | oa 


we, ie ‘ ; 
7 wi L Pere Yagi i tien wy nie 
el rie oiribita GD way sacl by 


> ; A sire al a fii& ; 2) ty ah ali 6 7 ty 


; 
: lmapiedly inthe ey reart) j Bi Sot ; Pa a a 
: TN is It bes 
om . d 
: 
is 
Y " 
! 
» i] 
‘ 
- | 
"4 


oae~ | é 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 198 Tue AMERICAN Museum or NaTuRAL HisTorY Nov. 21, 1925 
New York City 


56.9(1181:51.7) 
NEW MAMMALS FROM THE IRDIN MANHA EOCENE OF 
MONGOLIA! 


By W. D. MatrHew AND WALTER GRANGER 


The following descriptions are based upon collections secured by the 
Third Asiatic Expedition during 1923. The stratigraphy of the Irdin 
Manha formation has been briefly described and notices of some of the 
fossil mammals secured in 1922 and 1923 have been published in preced- 
ing articles in Novitates.27 The most important finds of the expedition in 
this formation, skulls of a series of titanotheriids, and skeleton, skulls, 
etc., of a large amynodont rhinoceros, are described by Professor Osborn 
in articles now in press or forthcoming. The smaller perissodactyls are 
described by us in a following number of Novvtates. 


CARNIVORA (Creodonta) 
Mesonychidze 
In addition to the giant Andrewsarchus® skull, this family is repre- 
sented by scattered teeth and jaw fragments of at least four smaller 
animals. 
Mesonychid, gen. indet. 


One species, No. 20132, is of about the size of Harpagolestes winten- 
sis of the Uinta, but we are unable to determine the molar formula, and 
it cannot be even provisionally assigned to any genus. 


Mesonychid, gen. indet. 


A second, smaller species, about the size of Synoplotherium lanius, 
is likewise represented by isolated teeth, No 20133, and cannot be placed 
in any one of the mesonychid genera. 


Hapalodectes serus, new species 
Typre.—No. 20172. Lower tooth, molar or posterior premolar. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 60. 

2Amer. Mus. Novitates, Nos. 42, 77, 91, 104, 145, 146. 

83Osborn, Henry Fairfield, 1924, Amer. Mus. Novitates, No. 146, November 11. 


Fig. 1. Hapalodectes 
serus. Lower molar, type 
specimen, No. 20172, twice 
natural size, crown and 
side views. 


A.M.20128 


Fig. 2. ? Hapalodectes 
auctus. Upper molar, type 
specimen, No. 20130, twice 
natural size, crown and 
side views. 


Fig. 3. Propterodon irdinensis. Lower jaw, type specimen, No. 20128, natural 


size, external and crown views. 


AM.20128 


Fig. 4. Propterodon 
irdinensis. Upper jaw 
with p* and roots of mo- 
lars, somewhat doubt- 
fully associated with the 
type lower jaw. N atural 
size, crown view. 


Li] 


Fig. 5. Miacis invic- 
tus. Upper molar, type 
specimen, No. 20137, 
three times natural size, 
externaland crown views. 


1925] NEW EOCENE MAMMALS FROM MONGOLIA 3 


CHARACTERS.—Cusps high and sharply compressed, the principal cusp (proto- 
conid) showing no trace of metaconid; the heel large, sharply crested in the char- 
acteristic mesonychid style, a minute anterior basal cusp (paraconid). Length of 
tooth 5.6 millimetres, smaller than either of the described American species. Much 
smaller than ?H. auctus. 


?Hapalodectes auctus, new species 

Type.—No. 20130. An upper molar (possibly a premolar). 

CHARACTERS.—Tooth much wider than long, protocone high, round, metacone and 
paracone moderately connate into a high, rounded cusp, with basal heels representing 
parastyle and metastyle. Valley between inner and outer portions of tooth deep 
and wide, this portion of the tooth rather strongly constricted. No crests from 
protocone. Size of the smaller Dissacus species, but very different in proportions of 
the tooth. 

In its general character this peculiar tooth somewhat suggests the 
Leptictide. It is unlike any known mesonychid upper molars, and it 
may not in fact belong to the Mesonychide. Theupper teeth of Hapalo- 
dectes (Wasatch and Wind River) are not known, but they should be much 
of this type, and the peculiar roundness of the cusps and the absence of 
connecting crests are suggestive of mesonychid and not leptictid rela- 
tionships. The tooth, therefore, is referred provisionally to Hapalodectes. 


Hyznodontide 


This family is represented in the Irdin Manha fauna by a lower jaw 
described in 1923 as Paracynohyenodon morrisi, and by a number of jaw 
fragments of a species referable to Propterodon. 


Propterodon Martin, 1906 

GrEnotTyPE.—A lower jaw from the Egerkingen beds, probably identical with 
Hyznodon schlosseri Rutimeyer. 

Martin bases this genus! upon a lower jaw, figured by Rutimeyer? 
under the name of Pterodon, but not given any specific name. Ona later 
page of the same memoir (op. cit., p. 461), Martin discusses two lower 
molars which Rutimeyer figured (op. cit., figs. 18, 14) and described as 
Hyznodon schlosseri. Rutimeyer regarded these teeth as m2; Martin 
identifies them as m; and observes that they must represent a genus 
distinct from Hyznodon and retaining the third upper molar; but he does 
not say whether he refers them to Propterodon or make any comparison 


1Martin, Rudolf, ae Revision der obereocinen und unteroligocinen Creodonten Europas. Rev. 

Suisse de Zool., XIV, 4 

oy *Rutimeyer, es 1891, Eociine Siugethierwelt von Egerkingen. Abh. schw. pal. Ges., p. 99, 
vu, fig. 15 


4 AMERICAN MUSEUM NOVITATES [No. 198 


with m; of the genotype jaw. So far as one can judge from Rutimeyer’s 
figures, they might well belong to the same species, and we have provis- 
ionally so referred them. 

Depéret! questions Martin’s identification of the type teeth of H. 
schlosseri as m3, but without considering the evidence afforded by the 
lower jaw above mentioned, or the peculiar proportions of the teeth, 
which are not like the mz of Hyznodon. 

The above record is cited to show the status of Propterodon in 
nomenclature of the hyzenodonts. It does not affect the validity of the 
genus or the reference to it of the species here described. 


Propterodon irdinensis, new species 


Typr.—No. 20128, a number of jaw fragments, probably in part associated 
(field No. 156), from the Irdin Manha Eocene of Mongolia. 

CHaRACTERS.—M; with minute talonid, no metaconid, protoconid broad, with 
flat posterior face (in place of the narrowed ridge of Hyznodon); m, with rather large, 
high-crested heel nearly as broad as the body of the tooth, a minute metaconid, para- 
conid considerably smaller and lower than metaconid. Four rather large, close-set 
premolars, not pitched backward, ps lacking anterobasal cusp, p: one-rooted, crowded. 
' Three upper molars, the third about as in Pterodon, transverse, two-rooted, crown un- 
known. First and second with well-separated antero-internal roots, crowns unknown. 
P* lacks antero-external cusp, the inner root is well developed, antero-internal in 
position (not median as in H yeenodon and Pterodon), and carries a shelf and cingulum 
but no distinct cusp. Principal cusp and postero-external cusp much as in Hyznodon 
crucians. In size the species about equals Hyznodon mustelinus. 

The heel of m; is decidedly more reduced than in Pterodon, but in 
other particulars the teeth are more like the primitive hyznodonts 
Sinopa, Tritemnodon and Cynohyxnodon, especially the constitution of 
ps, p? and mj, and the genus—or at least this species— might be regarded 
as a connecting link and structurally ancestral to Hyznodon. 

Whether its stratigraphic position would admit of this is another 
question. Hyzenodon proper makes its first appearance in Europe in the 
lower Ludian? (Euzet-les-Bains), in America in the Chadron, Titano- 
therium beds, commonly regarded as Sannoisian. The Irdin Manha is 
correlated with the lower Uinta of America, but its European equivalent 
is by no means certain; it may be Ludian, but is possibly older. There is 
little or nothing on which to base a direct comparison. 


x Deperete Charles, 1917, Mammiféres fossiles d’Euzet-les-Bains, Ann. Univ. Lyon, N. S., fase. 
VD: 5 

“Hyznodon’’ nouleti, Bartonian of Castrais. is excluded as generically indeterminate; ‘H.” 
schlosseri, Lutetian of Egerkingen, as probably not Hyxnodon (vide supra). 


1925] NEW EOCENE MAMMALS FROM MONGOLIA 5 


Miacide 
Miacis invictus, new species 

Type.—No. 20137. An isolated upper molar, m!. 

CHARACTERS.—NSize nearly as in M. medius of the Washakie, but construction of 
molar more as in M. parvivorus of the Bridger. Anteroposterior width considerably 
greater than in m! of Procynodictis vulpiceps. Paracone larger than metacone, proto- 
cone with both wings complete, conules distinct, parastyle well developed as a simple 
crest, no meta- or mesostyle, a moderately broad shelf external to the paracone and 
metacone. Cingulum encircling outer half of tooth, joining conules at its inner ends, 
a separate Basal cingulum encircling inner half of tooth broadened into a ‘‘bourrelet”’ 
at the postero-internal part of the protocone. 

The reference of a single tooth is necessarily doubtful, but this 
specimen agrees very closely with the genotype of Miacis, M. parvivorus. 
It is readily distinguished from Cynodictis and other genera by the in- 
equality of para- and metacone, from Procynodictis by the less compressed 
proportions of the tooth, from Viverravus and Uintacyon by the equal 
development of the wings of the protocone. The proportions of the tooth 
indicate that it belongs in the typical section of the genus, represented in 
the Uinta by Wiacis (‘‘ Mimocyon’’) longipes Peterson, rather than to the 
subgenera Lycarion, Harpalodon and Prodaphxnus, in which the tubercu- 
lar dentition is relatively reduced. A species of Miacis (M. exilis Filhol) 
is recorded from the Phosphorites, but it is much smaller than this one; 
Viverravus angustidens is distinguishable by absence of metaconule and 
of “bourrelet”? on m!. 


AMBLYPODA 
Eudinoceras mongoliense Osborn 


This genus was based upon two upper premolars identified by Osborn 
as p® or p*. A third isolated upper premolar, No. 20134, is probably p? 
of the same genus. It is considerably smaller and less expanded trans- 
versely, the outer crescent is not so deeply inflected, the inner cusp more 
rudimentary (?vestigial), and a strong encircling cingulum expanded 
into a considerable shelf on the antero-internal face. 

Osborn regards these teeth as representing a higher stage of specializa- 
tion than ‘‘ Dinoceras,”’ and specifies that he applies the prefix ‘‘eu”’ in 
this sense [i.e.=meta]. It might equally well be regarded as more primi- 
tive, the small internal cusp being intermediate between the uintatheres, in 
which it is absent, and Coryphodon and Pantolambda, in which it is well 
developed. The new tooth is suggestively like the Bathyopsis premolars, 
although the type teeth are very unlike p‘ of that genus. 


6 AMERICAN MUSEUM NOVITATES [No. 198 


V4 


“ny My, , 
Vit Wns 
UTI TS y 


Fig. 6. Eudinoceras. Upper 
'4 premolar, No. 20134, natural 
) size, external and crown 
views. 


Fig. 7. Achzenodont. 
Supposed upper milk 
molar, No. 20136, nat- 
ural size, crown view. 


ARTIODACTYLA 
Achenodontidz 


No. 20136, an incomplete upper tooth, is doubtfully identified as 
dp‘ of an achenodont. It is too incomplete for positive reference. The 
outer border of the tooth is missing, but it appears to have been of 
rounded subtrigonal outline with two major outer cusps (paracone, 
metacone) rounded and well separated, a smaller internal cusp (proto- 
cone) and a much smaller postero-internal cusp (hypocone) which stands 
a little apart but not far from intermediate between protocone and 
metacone. 

The tooth has some resemblance to dp‘ of the entelodonts, differing 
in absence of the conules, which are well developed in the deciduous as 
also in the permanent molars of entelodonts. In the achenodonts they 
are absent in the permanent molars; dp* is not known, but presumably 
would conform to the permanent molar construction. The Mesonychidze 
have a similarly simple construction of the molars, but lack the hypo- 
cone; dp? appears to be unknown in Mesonychide but presumably would 
also conform to the permanent molar pattern and lack a hypocone. 

The above reasons warrant a provisional reference of No. 20136 to 
the Achzenodontide. We agree with Peterson that this group is not 
ancestral to the entelodonts, and probably only distantly related. 


1This view was indicated by Matthew in 1899, Achenodon and Protelotherium being referred to 
a separate family from Entelodon. 


1925] NEW EOCENE MAMMALS FROM MONGOLIA 4 


The relationship to Helohyus may be much closer, and, if brought into 
the same family, the name Helohyide Marsh, 1877, would have 
precedence. 


Helohyide 
GOBIOHYUs, new genus 

Tyrre.—Gobiohyus orientalis infra. 

Dracnosis.—Teeth bunodont, low-crowned, upper molars five-cusped, without 
hypocone or mesostyle, with small parastyles and encircling cingula. Third and 
fourth upper premolars trigonal, with deuterocone and inner root, large on p‘, rather 
small on p*; p* simple-crowned, compressed, two-rooted. Lower molars with very 
small internal paraconids, minute hypoconulids on m;-2 and large conical third lobe on 
m3. P, with distinct deuteroconid. Anterior premolars spaced, p; one-rooted, ¢ 
large, procumbent, incisors small. 

A few specimens in the Irdin Manha collection represent a bunodont 
artiodactyl resembling the Bridger genus Helohyus, but distinguished 
by the double cusp of ps (upper premolars unknown in Helohyus). It is 
well distinguished from any of the genera of the European Eocene, nor 
do any of the known bunodont artiodactyls of the Uinta approach it 
closely. The absence of thypocone distinguishes the new genus from 
Dichobunide and Entelodontide, the proportions of teeth and jaw, size 
of third lobe of ms, ete., from Cebocherus and Chaeromorus, and from the 
earlier genera of Suidz and Tagassuidee. The conical cusps and lack of 
external stylar cusps on the upper molars sufficiently distinguish it from 
the Anthracotheriide, to which, however, it may be rather nearly related. 
From Cheropotamus it is distinguished by absence of either mesostyle or 
central cuspule on the upper molars, simple third lobe of m3; and other 
details that suggest the closer affinity of the Parisian genus to the pigs. 

Cheropotamus and Cebocherus are generally regarded as primitive 
suillines, and grouped by Depéret under the family Hyotheriide. On 
the other hand, Schlosser makes Cheropotamus the starting-point of the 
Anthracotheriide. Whether Hyotheriwm, which is much more clearly 
of Suid affinities, should be included in a separate Eocene family may be 
open to question; in any event the family name Chceeropotamidz Owen, 
1840, long antedates Hyotheriide Cope, 1888.' This family might be 
regarded as the common ancestral stock of suillines and anthracotheres, 
and it may well prove that Helohyus and its allies, separated as a distinct 
family by Marsh in 1877, united with the Dichobunide in Sinclair’s 
revision, are in reality a more primitive group of the same stock, and 


\Hyotheriine, Amer. Nat., XXII, p. 1087. Hyotherida Heckel, 1895. 


: 


8 AMERICAN MUSEUM NOVITATES [No. 198 


should be referred to Chceeropotamide.! The American genera are very 
imperfectly known at present, and no distinctively suilline or anthra- 
cothere characters appear in the parts described, so that it seems better 
to retain the family for the present, pending determination of its real 
affinities. Near relationship between Gobiohyus and Helohyus is in- 
dicated by the close resemblance throughout of the molars and premolars, 
apart from the more progressive character of the latter (indicated in pu, 
inferred in upper premolars), a difference to be expected in an Upper 
Eocene genus compared with its Middle Eocene relatives. It should 
also be observed that Stehlin’s interpretation of the anterior teeth in 
Cheropotamus would exclude it from any typical position in the common 
ancestral stock of Suide and Anthracotheriide. The Helohyide may 
likewise prove to belong rather with the dichobunids than with the 
chceropotamids, in spite of the absence of hypocone. 


—— 


AM.20249 


Fig. 8 Fig. 9 
Fig. 8. Gobiohyus orientalis. Upper jaw, type specimen, No. 20249, natural 
size, external and crown views. 
Fig. 9. Gobiohyus orientalis. Lower jaw, No. 20250, natural size, crown and 
external views. 


Gobiohyus orientalis, new species 
Typrr.—No. 20249, upper jaw with p?-m*. Paratypes: No. 20248, lower jaw, 
m)-3, and two upper molars probably accidentally associated; No. 20250, lower jaw, 
pa-m3. All from Irdin Manha formation. 
Species CHARACTERS.—P?-m*? =50; ‘m+-3=26.5; mi-3=30; m2-3;=22. Lower 
molar teeth more robust, trigonid of m3; wider and heel smaller; mglength 12.8 
width 8.2 


= 1.56 


1Stehlin, H.-G., 1906, however, regards Helohyws_as more probably of dichobunid Affinities. 
See his discussion in Siugethiere des schweizerischen Eocins, part iv, 1906, p. 672. 
r 


‘OZIS [BINJEN “OFZOST “ON “uaumt9eds 
ody, ‘ourueo Joddn Jo MarA [BUI}Xo YIM ‘SMOTA UMOIO pue [BUI0} x0 ‘Mul IAMOT “snpsngousnfiiyorgoy “QT “St 


9b202 WY 


10 AMERICAN MUSEUM NOVITATES [No. 198 


Gobiohyus pressidens, new species 


Typr.—No. 20247, lower jaw fragment with m.-3. _Irdin Manha formation. 
Species CHARACTERS.—Size smaller than G. orientalis, teeth narrower, m; with 


narrower trigonid and larger heel. M2-3=18.5; ms length= 11. a 
width 5.2 


Gobiohyus robustus, new species 

Typre.—No. 20246, lower jaw, Ps-m3, broken canine and premolars, and alveoli 
of incisor teeth. 

CHARACTERS.—Teeth eighteen per cent. larger than in G. orientalis, relatively 
more robust, the trigonids of lower molars equaling talonids in breadth, whereas 
in G. orientalis they are narrower. 

The relations of the anterior lower teeth as shown in the type speci- 
men are presumably typical of the Helohyide. The canine is enlarged, 
procumbent, the incisors small, three in number, as indicated by their 
alveoli. The first premolar is rather small, one-rooted, with rather long 
diastemata before and behind it. There was apparently a considerable 
diastema behind the two-rooted, compressed second premolar. The third 
premolar was compressed, apparently simple, as long as ps, or slightly 
longer. P, has the same construction as in the genotype, but is more 
robust, as are also the three molars. 

This construction of the anterior teeth suggests the primitive Anthra- 
cotheriidz! and Suide—not Cheropotamus if Stehlin’s interpretation of 
the front teeth in that genus be correct. 


TRAGULINA 


cf. Archeomeryx, gen. indet. 


No. 20173. A fragment of lower jaw, with p3-, badly damaged, belongs to the 
ruminant group, but can hardly be definitely placed or compared. It is about the size 
of Archxomeryx, and the construction of the premolars, so far as.preserved, appears 
to be similar. 

The principal interest of this specimen is that it shows the presence 
of pro-Pecora in the Irdin Manha fauna as well as in the Shara Murun. 


1The arrangement of the lower ‘premolars i is exactly as in the type of Lophiohyus Sinclair of the 
Bridger Eocene; a complete skull and jaws of a small anthracothere from the Eocene of Burma 
(shortly to be described by Barnum Brown) also agrees in the relations of the anterior teeth. 


sd Fah at ~ ld 
a pe bs 1 2 ' i LDA, 
) en Abas 
apa Ted, duit linn bis 

; ‘ey 17 pore ee we } ( P ae “at a 
“ - 6 \ 

ry ; 5 : ; Li P = “4 

& Od - 1 FOC 3 <8 rw ar! he 


ariy she ‘4 ine "PL Pol eerie ry iste 


Dee aa Pee ee 


AMERICAN MUSEUM NOVITATES 


Published by 
Number 199 Tue AMERICAN Museum or NaTuRAL History Nov. 23, 1925 
New York City 


56.9,72(1181:51.7) 
THE SMALLER PERISSODACTYLS OF THE IRDIN MANHA 
FORMATION, EOCENE OF MONGOLIA! 


By W. D. MatrHew AND WALTER GRANGER 


A series of skulls of titanotheres, complete skeleton of an amynodont 
rhinoceros, and various more fragmentary specimens of these two 
families of Perissodactyla, were secured from the Irdin Manha forma- 
tion by the Third Asiatic Expedition in 1928. They are described by 
Professor Osborn in forthcoming numbers of American Museum Novitates. 
Besides these larger animals, there were numerous fragmentary remains 
of small perissodactyls assigned to the present writers for study and 
description. They appear to be referable to the Helaletidze, Lophio- 
dontide, and Hyracodontide, the entire absence of horses and palzo- 
theres being a feature of this and other early Tertiary faunas of Mongolia. 


Helaletide (?Colodontide) 
Desmatotherium mongoliense Osborn 

Type.—No. 19161, a right maxilla with p?-m?. 

This species was based upon ‘‘parts of ten individuals of a small 
lophiodont,”” among which the specimen selected as type belongs to a 
different genus and family from the rest. The species is here restricted 
to the type maxilla among the 1922 collections, the other specimens 
being referred to the lophiodont genus Lophialetes. The description and 
measurements of D. mongoliense are based only in part upon the type, 
and are rather misleading. It is about the same size as D. guyotianum, 
not smaller; the metacone is not flat but deeply concave, with its free 
flange quite short; referred specimens show that in m? it is further 
reduced. The lower molars are sharply cross-crested, and m3 has no 
heel. The premolars are of the helaletid type, very different throughout 
from the hyracodont premolars, but resembling those of D. guyotii and 
of Colodon; more primitive than in the latter, and more progressive than 
in Helaletes. 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
bution No. 61. 


2 AMERICAN MUSEUM NOVITATES [No. 199 


AM.I9I61 Type 


m2 1 p4 3 2 1 CC  Paae 


Fig. 1. Desmaiotherium mongoliense Osborn, 1922. Upper jaw, type specimen, 
No. 19161. 


Natural size, external and occlusal views. 


1925] SMALLER PERISSODACTYLS OF MONGOLIA 3 


NN 


AY 


ii 
| BM 


& 


WP 
fl x 


py) 
fp 


a 
OMS 


AY 


AM.20155 


Fig. 2. Desmatotherium mongoliense. Lower jaw, No. 20155. 
Natural size, external and superior views. 


Desmatotherium fissum, new species 


Typr.—No. 20161, upper jaw fragment, p*-4. 
CHARACTERS.—Somewhat smaller than D. mongoli- 


ense, the internal cusps of p? and p* wide apart, that of p* Fig. 3. Desmato- 
indicated by a sharp groove on the inner face and a broad- therium fissum. Up- 
ening of the apex. Paracone and metacone on p7* well per premolars, type 
separated, convex externally, parastyles lower and some- specimen, No. 20161. 
what smaller cusps, no distinct metastyles. Natural size, crown view. 


Teleolophus' medius, new genus and species 


Typr.—No. 20166, lower jaw with p:-m3. 

ParatypPes.—No. 20163, lower jaw, ps-m3; No. 21064, upper jaw, dp*-m?; 
No. 20165, miscellaneous upper and lower teeth and jaw fragments. 

All from the Irdin Manha formation, Telegraph Line Camp. 

CHARACTERS.—Lower molars sharply cross-crested, increasing in size from first 
to third, no heel on m;, premolars all two-rooted, non-molariform, a transverse crest 
in front and an anteroposteriorly-crested hypoconid behind, entoconid rudimentary. 
On py» the transverse crest is imperfect and on p; barely suggested. Upper molars 
with high, crested protoloph and metaloph curving around externally to join the 
flattened paracone at its anterior and posterior ends; the metacone has practically 
disappeared, the only vestiges of it being small crests (=free posterior flange of 


1Greek réX\ecos, complete, perfect; Addos, crest. 


¥ 
‘S918 IvjOW-IU[OUIeId JO META UMOIO PUB MOIA [VUIOYXe ‘azZIs [BIW Ny 


‘99T0Z ‘ON ‘ueurtoeds odAy ‘mel omMoT ‘snypau snydopoajay “FS 


| addy g99I0Z WV 


1925] SMALLER PERISSODACTYLS OF MONGOLIA 5 


ectoloph) at the postero-external angles of m!-? which rise only half-way up to the 
3 


vertex of the metaloph and are entirely absent on m*. 

Length of p;-m3=52 mm. Dimensions of m’, a.-p. <tr. =21 26 mm. 

This genus differs from ‘‘ Desmatotheritum” mongoliense and fissum 
as well as from the type of Desmatotherium in the decidedly higher and 
more sharply-crested crowns, reduction of the metacone on the upper 
molars, absence of heel on m3, of entoconid on ps, etc. The molars are 
very like those of Deperetella but the premolars are simpler, resembling 


those of Desmatotherium. It may well be ancestral to Deperetella. 


AM.20165 AM.20168 A.M.20164 


Fig. 5. Teleolophus medius. Upper teeth. Composite of three specimens, 
Nos. 20164, 20165, 20168, as indicated in the drawing. 


Natural size, external and crown views. 


Lophiodontidze 
Lophialetes' expeditus, new genus and species 

Typr.—No. 19163, upper jaw, with p*-m! r. 

ParatyPes.—Nos. 19162, 20144-20160. 

Horizon anp Locauiry.—Irdin Manha, type locality. 

CHARACTERS.—Metacones of upper molars flat externally, with long, free flange. 
Lower molars sharply lophodont, with connecting crests moderately developed; a 
reduced third lobe on m3, with looped crest as in Lophiodon. Upper premolars with 
crested metacone, nearly flat externally, inner crescents undivided, but both wings 


1Lophiodon +Helaletes—as combining characters of the two genera. 


6 AMERICAN MUSEUM NOVITATES [No. 199 


AMI9I63 | 
| 


A.M19162 i a, are oe 
ohigen 


Fig. 7 
Fig. 6. Lophialetes expeditus. Upper jaw, type specimen, No. 19163. 
Natural size, external view and crown view of teeth; anterior premolars from No. 20160. 


Fig. 7. Lophialetes expeditus. Lower jaw, No. 19162. 


Natural size, superior and external views. 


complete on p*-4. Size one-sixth smaller than Desmatotherium guyotii, m-’ =37 mm.; 
Pe-m;=67 mm. (Osborn, 1923, ‘‘D. mongolicum’’). 

The upper molars have the primitive rhinocerotoid structure much 
as in Triplopus, but the last lower molar has a distinctly Lophiodon- 


1925] SMALLER PERISSODACTYLS OF MONGOLIA 7 


like heel. The construction of the premolars distinguishes it from the 
type of Prothyracodon as well as from that of Triplopus. 

A great number of fragmentary specimens, teeth and parts of Jaws 
from the type locality of the Irdin Manha formation belong to this 
genus, and represent probably one rather widely varying species. The 
skull is sufficiently known (No. 20144) to show that it has somewhat 
the helaletid type of recession of the nares. The feet are not known. 

The present genus resembles the Hyracodontide in the rhinocerotoid 
type of the upper molars, sharply crested, with the metaloph joining the 
ectoloph far foward, and the external face of the ectoloph flat behind the 
anterior pillar. The premolars are rather helaletid than hyracodont in 


Fig. 8. Lophialetes expeditus and L. minutus. Upper molars, Nos. 20160, and 
20139 (type). 


Twice natural size, crown view. 


type, and the heel of mz; is of the looped type of Lophiodon, differing 
from the small transverse heel of Helaletes or the absence of heel in Tr7- 
plopus or in Hyracodon. It appears to occupy a somewhat intermediate 
position, not closely related to any of these groups, and probably will have 
to be placed in a separate subfamily, Lophialetine, distinguished by the 
above combination of characters. 


Lophialetes minutus, new species 
Typre.—No. 20139, upper molar, Irdin Manha beds, Telegraph Line Camp, 
Mongolia. 
CHARACTERS.—Size little more than half the preceding; diameters of molar a.-p. 
Xtr.=8xX9mm. Nocrista at head of median valley. 


Hyracodontide 
Cenolophus proficiens, new species 


Tyrre.—No. 20141, lower jaw, pi-m;, from Irdin Manha formation, Mongolia. 
Species CHARACTERS.—Size larger than C. promissus of the Shara Murun (type 
of the genus), pi-m; =89, mi-;=54 mm. Lower molars moderately increasing from 


ee ————————————aEa—————E———————— en SSS ee ee EE ee _—- — > 7: | a 


8 
‘OZIS [BANJEN ‘TFIOS ‘ON ‘uourroeds ody, “Yy}90} JO MOIA UMOIO pu ‘MoIA [eUIO}Xe ‘Mel JoMoT <‘suatooud snydojouw) *6 “SIT 


\ ad |plozwv 


1925] SMALLER PERISSODACTYLS OF MONGOLIA 9 


first to third, posterior crest of m3; pitched forward and inward, with a narrow cingular 
ledge behind it, coming to a blunt point posteromedially. Third and fourth premolars 
submolariform, first and second nearly simple, with a double posterior crest; p; in 
type two-rooted and crowded so as to be mainly internal to p, instead of anterior to it 
(probably abnormal—in No. 20140 it is one-rooted and not displaced). 

Two isolated upper teeth, probably dp’ and m?, from the Irdin Manha 
beds are referred to this species. They are hardly distinguishable from 
C. obliquus of the Shara Murun. Other specimens referred to C. profi- 
ciens show the characters of the front teeth. They retain the primitive 
characters from which typical amynodonts have departed by enlargement 
of the canines to powerful tusks, and typical hyracodonts by reduction 
of the canines to completely incisiform status. In one specimen of C. 
proficiens, however, the second incisor appears to be moderately enlarged, 
thus suggesting a beginning of the true rhinoceros specialization in 
which ig becomes a powerful tusk. Unfortunately the tooth is broken 
off, and of the adjacent teeth only the roots remain. 

The genus is placed provisionally in the Hyracodontide. It is not 
nearly related to Lophialetes, or to Desmatotherium. 


~ bat (1 ee 


Peed 


bdatons “he rye a 
a) Serre iin a a 
ee a w dtenik yee 
pereird (hited il se 


; 


ee as 


eid 
7, 


“aha “og fans sn ely" 
eee) cana ba 

iy aitsh: iy ne ia ans 

aie Ca hate Semi 


, 
* e 
7 A rr a 1 aay 


_e’* 


rE Gr. wu Wei sae 


"Ry maltrabeia ads! chen aid err ton 
. 1 J oe iM 
, hide Z eee « Lares ac) 
‘ 4 rat ty mt, : 7h is ia ~ 
] ’ iain’ Ws ny years i 
> ee - i 
“ef 2 OE es 
. vs Ja Sn 
ied | | eee file Le 
4 ‘ 713 Prk} if 
- ‘af i : oe oh 
— 
i . 
* e Ry 
. 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 201 Tue AMERICAN Museum oF Natura. HIsToRY Nov. 24, 1925 
‘ New York City 


56.9 (116 :51.7) 


A MESOZOIC MAMMAL SKULL FROM MONGOLIA! 


By GEORGE GAYLORD SIMPSON 


Among the many remarkable discoveries of the Third Asiatic 
Expedition of The American Museum of Natural History is a large part 
of the skull with associated lower jaws of a Mesozoic mammal. This 
unique specimen was found by A. F. Johnson in the Djadochta forma- 
tion at Shabarakh Usu in Mongolia. Mesozoic mammals are among the 
rarest of the fossils, and, while it might well have been inferred that 
there were multituberculates in Mongolia in Cretaceous times, this 
splendid confirmation is no less startling than it is welcome. Only one 
other partial Mesozoic mammal skull has been found, that of Tritylodon 
Owen (1884) from the Lower Jurassic of South Africa, and the lower 
jaw of the latter form is quite unknown. No Mesozoic mammal has 
hitherto been found in Asia, and, although mammals occur at earlier 
and later horizons, none has yet been found in a formation of age equiva- 
lent to the Djadochta. Consequently, this specimen is noteworthy, 
not only in being the second known skull, and the first with associated 
lower jaws, and as adding new morphological knowledge, but it also adds 
a new horizon to the stratigraphic range and a new continent to the 
geographic distribution. The writer is deeply indebted to Professor 
Osborn and Doctor Matthew for the great privilege of naming and 
describing this important form. 

The preparation, a task calling for the greatest delicacy and address, 
owing to the softness of the bone and the nature of the sandstone matrix, 
has been undertaken with great skill by Mr. Albert Thomson. This 
preparation is not yet complete, but enough has been done to reveal all 
the essential features and to make possible this preliminary description. 


Djadochtatherium matthewi, new genus and species 
Typre.—Amer. Mus. No. 20440. Anterior part of skull with associated lower jaw. 
Hor1zon.—Djadochta formation, early Upper Cretaceous, Mongolia. 

GrNneERIc CHaracters.—A multituberculate of moderate size; the third upper 
incisors widely separated from the second ones and antero-internal to the first pre- 


1Publications of the Asiatic Expeditions of The American Museum of Natural History. Con- 
tribution No. 62. 


2 AMERICAN MUSEUM NOVITATES __[No. 201 


molars; the lower premolars reduced to one; the lower molars with six cusps. (The 
specific characters are included in the following description.) 

The anterior half of the skull is present and is well preserved save 
for the extreme tip and for a little distortion due to crushing. The 
palate anterior to P? is present and exposed. The nasals and pre- 


AM.20440 


Fig. 1. Djadochtatherium matthewi, superior view of the type skull. X2. 
The distortion has been corrected and the parts cross-lined restored. 


maxille are present and complete, as are also the maxillze save for the 
posterior parts of their alveolar and palatal portions. The sutures are 
well defined, and the only possibility of misinterpretation lies in the 
posterior upper part, where crushing has been most severe. 

As seen from above, the facial region is triangular, resembling 
Ptilodus closely in this respect. Most of the upper surface is formed by 
the very large nasals, which are broad at all points but slightly more so 


1925] A MESOZOIC MAMMAL SKULL FROM MONGOLIA 3 


posteriorly. They reach back almost to the posterior border of the 
orbit and form a part of the superior border of the latter. The anterior 
parts of the frontals seem to be present; they come forward to a median 
point as in Polymastodon or Ornithorhynchus, and each was probably 
much reduced and triangular as in those genera. No part of the parietals 
is preserved, but from the shape of the nasals and frontals it seems fairly 
certain that they had a well-developed nasal contact at about the point 
where the nasals are broken off. The parietals did not reach the maxillez. 

The premaxilla is well developed with long maxillary and nasal 
sutural contacts, and is unusually elongated. It forms a larger part of 
the lateral aspect of the face than does the maxilla. The latter element 


AM20440 


Fig. 2. Djadochtatherium matthewt, side view of type skull and jaws. X 2. 


Only the parts not present on either side are cross-lined. 


is also large, but with its strictly facial portion much reduced. It 
probably formed at least half of the zygomatic arch, which is uncommonly 
stout and arises opposite the third premolar. The infra-orbital canal 
is short, and the single foramen is situated not far from the alveolar 
border just anterior to the root of the zygoma. 

There is no trace of a lacrymal bone, and this element is probably 
absent. Neither is there any trace of a jugal, and this bone, too, may 
have been totally absent. If present, it was very small and was confined 
to the upper part of the zygomatic arch. 

A somewhat questionable but interesting feature is the apparent 
presence near the middle of each of the large nasals of a small elliptical 
foramen. These were present before burial, and their presence in both 


4 AMERICAN MUSEUM NOVITATES [No. 201 _ 


nasals in symmetrical positions seems to indicate that they are original 
features of the animal. If so, the only condition which might be com- 
pared is that seen in Ornithorhynchus: the supra-orbital foramina of the 
latter are nearly but not quite inclosed within the nasals—if they became 
fully so they would agree in position with the present doubtful foramina. 
Anterior to these openings are two longitudinal cracks which are as well 
marked as sutures but which are probably due to pressure on the buried 
skull. 

The palatal aspect, although incomplete posteriorly, is very in- 
structive. The palatine processes of the premaxillz are well developed 
and somewhat longer and broader than is usual. They meet in a slight 


AM.20440 


Fig. 3. Djadochtatherium matthewi, palatal view of type skull. X 2. 


Cross-lined parts restored. 


groove along the midline. Consonant with the elongation of the pre- 
maxille, the anterior palatine foramina appear to have been shifted 
posteriorly, but they retain their relationships with the third incisors, 
being immediately internal to them as in other multituberculates. 

The palatine portions of the maxille are also large, and meet in a 
low longitudinal ridge at the midline. Opposite the third premolar 
may be seen the anterior end of a small palatal vacuity. 

The mandible is of the familiar multituberculate type, differing 
only in detail (and in some features of the dentition) from that, for 
instance, of Ctenacodon, and the resemblance to Pézlodus is even closer. 
Above the symphysis, where the two rami were loosely united by carti- 


1925] A MESOZOIC MAMMAL SKULL FROM MONGOLIA 5 


lage, is a small median bony protuberance such as is seen, for example, 
in the rat. The coronoid process arises with its anterior border outside 
the last molar. The process is small and is separated from the condyle 
by a large semicircular notch. The masseteric and pterygoid fosse are 
not well defined anteriorly or superiorly; but the masseteric crest, al- 
though now broken off, was strong, while the pterygoid crest opposite it, 
on the inner side of the posterior portion of the mandible, was very promi- 


AM.20440 


Fig. 4. Djadochtatherium matthewi, photograph of type skull from above. X 2. 


nent. As seen from above or below, its edge is a gentle, unbroken curve 
departing from near the condyle, and there is not the slightest reason to 
speak of an angle, inflected or otherwise. 

The condyle is sessile, not marked off by a neck or constriction. 
The articular facet is roughly oval with its longest diameter in the same 
plane as the longitudinal axis of the jaw. It is convex in all directions, 
but more so along the longer diameter. 

DeEnTITION.—The teeth, unfortunately, are on the whole very 
poorly preserved, and they leave a number of very important points to 


6 AMERICAN MUSEUM NOVITATES [No. 201 


be cleared up by hoped-for further finds. The dental formula, so far as 
indicated, is 2“?-*-2+" and it seems most probable that it was?-°-*2, 
One cannot be sure that small first upper incisors did not occur as in 
Tritylodon and Ctenacodon (Allodon), but there is no indication of them. 
The second incisors are the largest and are implanted nearly vertically 
near the anterior end of the premaxille. Their tips did not meet in the 
midline. The apparent mode of wear is peculiar, for the worn facet, 
instead of facing backward, faces obliquely outward and forward. 
The third incisors, instead of closely following the second ones, are some 
distance directly back of the latter. Since the width of the palate ex- 


A.M.20440 


Fig.5. Djadochtatherium matthewi, photograph of type skullfrom the side. X 2. 


pands considerably in this distance, while these teeth are but little 
farther apart than the preceding ones, they are internal to the border of 
the palate and give the impression of actually standing on the latter 
itself. This very unusual character, together with the reduction of the 
lower premolars and simplicity of the lower molars, may be considered 
diagnostic of the genus. These incisors were conical and pointed. 

The upper premolars are so worn that their crowns are nearly 
shapeless. Two are preserved on the right side and two and the roots 
of a third on the left. On the left side, where they are a little less worn, 
it is seen that the crown of P! is nearly equilaterally triangular, with an 
angle external. On P? of this side two external cusps can be seen, one a 
little more external than the other, but the inner part of the crown seems 
to have been broken or worn off. The third premolar had two roots 
closely approximated, as had also the first and second, and was not 
larger than the second. There appears to be a diastema back of this 


1925] A MESOZOIC MAMMAL SKULL FROM MONGOLIA ‘if 


tooth, but in view of the condition of the specimen this remarkable 
impression may well be erroneous. Nothing remains of a possible fourth 
premolar or of the molars. Judging by analogy with other multitubercu- 
lates, the large lower premolar would require a mate, the fourth premolar, 
in the upper series, and the two lower molars would each occlude with 
an upper molar, making the formula that indicated as probable at the 
beginning of this section. It is very improbable, in view of the small 


Z 
| 


AM.20440 


Fig. 6. Djadochtatherium matthewi, photograph of palate of type skull. X 2. 


amount of space between the lower molars and the incisors and in view 
of the reduction of the lower premolars, that there were five upper pre- 
molars as in Ctenacodon and Plagiaulax. 

The lower dentition is much better known, although still hardly 
satisfactory. The incisors are remarkably stout, and their long curved 
fangs extended back at least to beneath P; or M; and undoubtedly were 
derived from a persistent pulp. The chief wear is on the superior (pos- 
terior) face, and here is developed a facet which faces backward and 


8 AMERICAN MUSEUM NOVITATES [No. 201 


inward. The premolar series has been reduced to a single tooth, Pu. 
As preserved, this tooth is large, blunt, rather formless, and somewhat 
compressed laterally. It seems probable that it was a sharp-edged 
shearing tooth, but this is not absolutely demonstrated. The first 


Fig. 7. Djadochtatherium matthew, photographs of mandible of type. X 2. 


A, from above; B, from the side. 


molar again is poorly preserved, not only badly worn but also broken. 
The crown had two longitudinal rows of cusps with a small number of 
cusps, quite surely three, in each. The posterior molar is unknown. 
When the jaws were closed, the lower incisors had their tips in- 
serted between the small posterior upper incisors. It seems certain, 


1925] A MESOZOIC MAMMAL SKULL FROM MONGOLIA 9 


however, that sufficient anteroposterior motion was possible for the 
lower incisors to work against the anterior upper ones—the facets on 
the two are complementary. 


AFFINITIES 


That Djadochtatherium is a multituberculate is, of course, obvious. 
That part of the dentition preserved is absolutely diagnostic, as are also 
the mandibular and cranial characters. The order Multituberculata 
may be divided into two great groups, one including Tritylodon and its 
allies and the other including all the upper Jurassic, Cretaceous, and 
Paleocene forms, with probably some earlier ones also. The Tritylodon 
group is characterized by almost undifferentiated upper cheek-teeth 
with three rows of two to four cusps each, and by the retention of a large 
lacrymal and jugal and of unreduced frontals. Djadochtatherium is 
plainly very distantly related to this group, if at all. 


The other group, or suborder, which may be called the Plagiaulac- 
oidea in contrast to the Tritylodontoidea, is much in need of revision 
and of careful redefinition of the several families which compose it, 
but it is quite outside the scope and aim of’ Novitates to present this 
taxonomic revision here, and it will suffice to point out merely the 
course of evolution as it is now known. The uppermost Jurassic multi- 
tuberculates have simple molars with six cusps each in two rows 
both above and below. There are four or three trenchant lower pre- 
molars, and five upper premolars, the last two of which are trenchant. 
In the Upper Cretaceous and Paleocene, the multituberculates have 
more complicated molars with two rows of cusps below and three above, 
and the first molar is larger and more complex than the second, instead 
of being nearly equal as in the Jurassic. In the more usual of these later 
types there is a trenchant lower premolar and sometimes another func- 
tionless one in front of this, and there are four upper premolars, only one 
of which, the last, is trenchant. In another type (Polymastodon), the 
premolars are reduced to a single conical one above and below. The 
cranial osteology is known only in two forms, Ptilodus and Polymastodon, 
but the Jurassic forms seem already to agree in the most striking speciali- 
zations, 1.e., the reduction or loss of the lacrymal and jugal. 

On comparing the subject of this paper with these plagiaulacoids, 
the first impression is the remarkable agreement. If it were found in 
beds of age equivalent to the Djadochta formation, but in England or 
America, the Mongolian form would, indeed, be a new and marked type, 


10 AMERICAN MUSEUM NOVITATES [No. 201 


but not a surprising or anomalous one, and certainly a hypothesis of 
Asiatic or other exotic origin would be entirely unnecessary. 

Djadochtatherium is in a very definitely Upper Cretaceous or, indeed, 
even Paleocene state of evolution. The reduction of the lower premolars 
is complete, for but one remains. The upper premolars, it must be in- 
ferred, were reduced to four, and of these only the last one can have 
been shearing. The size also is larger than that of any Jurassic forms 
but well within the range of the more common later genera. The simple 
character of the lower molars, if correctly interpreted, is a remarkably 
persistent feature in view of the specialization elsewhere. The equal 
size of the two lower molars, as also the small number of cusps, shows 
little or no advance beyond Plagiaulax or Ctenacodon of the Purbeck and 

*Morrison. It is interesting that no known feature of Djadochtatherium 
precludes direct descent from these English and American Upper 
Jurassic forms. 

The Mongolian mammal is, of course, very distinctive. The action 
of the lower incisors, the peculiar specialization of the upper ones, the 
early reduction of the lower premolar series, the prolongation of the 
premaxille, and the retention, in the face of these advances, of primitive 
molars, all place this form in a unique position. It could not possibly 
have been ancestral to any later known North American or European 
multituberculate, but is a member, perhaps the terminal member, of a 
hitherto unknown sideline of descent. Whether that line was close to 
that of the Ptilodus-like forms, the numerous Cretaceous and Paleocene 
ptilodontids, or whether its departure was in an entirely different direc- 
tion, cannot be definitely asserted until the crucial upper molars are 
known. If these are simple, with only two rows of cusps and no indi- 
cations of a third, Djadochtatheri'um would undoubtedly have to be 
placed in a different family, as will also be the case if, as is very possible, 
the unknown parts exhibit unpredictable peculiarities in keeping with 
those now known. If the upper molars are three-rowed, with twelve 
cusps or more, only a minor branch of the ptilodont stock would be 
indicated. The latter view may be adopted provisionally in view of the 
fact that the contour and size of the skull, shape of the lower jaw, pre- 
molar reduction and (so far as can be deduced) form, all agree remark- 
ably with the later ptilodonts. 

The cranial osteology is very similar to that of Polymastodon. The 
probable reduction of the frontals and the reduction or total loss of the 
lacrymal and jugal are significant, especially in view of the age of 
Djadochtathertum, which is much greater than that of Polymastodon, and 


1925] A MESOZOIC MAMMAL SKULL FROM MONGOLIA 11 


also in view of the great differences from the only other known Mesozoic 
mammal skull, T'ritylodon. 

So far as regards the bearing of this specimen on the affinities of 
the Multituberculata, a full discussion is not possible here. Many of the 
characters of Polymastodon and Ptilodus cited by Broom as indicating 
monotreme relationship are substantiated and carried back to a greater 
antiquity. In two features, the parietonasal contact without a parieto- 
maxillary one and the apparent presence of supraorbital foramina, 
Djadochtatherium resembles Ornithorhynchus even more than does Poly- 
mastodon, but these facts should not be overemphasized, not only be- 
cause they are not very binding evidence of affinities but especially be- 
cause their reality is not considered as established beyond all doubt. 


CONCLUSIONS 


The finding of a mammal in a later Mesozoic uplands formation of 
central Asia is an event which has-been awaited by paleontologists with 
passionate interest, for it has long been felt by many that in such a forma- 
tion the ancestors of the mammals which appear so suddenly in Europe 
and North America at the close of the Mesozoic might be found. Dja- 
dochtatherium goes far toward answering such anticipations with a firm 
negative. It is a Mesozoic mammal in every sense of the word, belong- 
ing to a previously known Mesozoic order, suborder, and probably 
family, and finding a possible structural ancestor in the Upper Jurassic 
either of England or of America. In conjunction with what we know of 
other faunas, it seems to indicate quite positively that the earlier Cre- 
taceous mammals of Asia, Europe and North America were from a com- 
‘mon source and not very different. Although the possibility is not to be 
flatly denied, it would be very surprising to find the ancestors of the 
Tertiary, or even of the Paleocene, placental mammals living with 
Djadochtatherium. 

In any event, this fine specimen has permitted the description of a 
second Mesozoic mammal skull, the addition to the known multituber- 
culates of a new and interesting adaptive type, the extension of the 
peculiar sort of cranial osteology seen in Polymastodon well down ‘into 
the Cretaceous, and the addition of a new continent to the known range 
of the Mesozoic Mammalia. 


oy fits le 7 ¥ 
pats “tt ely ve 
te nid 


iA as ma bab « a 
PTT io : ‘ Beir Jay are a 
ms eri ny mah eit r 


, 
% 


AMERICAN MUSEUM NOVITATES 


Published by 


Number 202 Tue AMERICAN Museum or NaTuRAL History Nov. 24, 1925 
’ j New York City 


56.9,72T (1181:51.7) 
UPPER EOCENE AND LOWER OLIGOCENE 
TITANOTHERES OF MONGOLIA! 


By Henry FAIRFIELD OSBORN 


The Eocene and Oligocene titanotheres of Mongolia, discovered by 
members of the Third Asiatic Expedition in the years 1922 and 1923, 
will be described and figured in detail by the author in Monograph 55 
of the U.S. Geological Survey, now going to press. Subsequent and still 
more complete description will appear in “‘The Natural History of 
Central Asia,” the series of volumes dealing with the Central Asiatic 
explorations of the American Museum. The object of the present issue 
is to name the various titanotheres discovered and to briefly designate 
the genera and species to which they belong as well as the geologic 
life zones in which they occur. Fortunately for purposes of correlation 
and bearing on zodgeographic relations in Upper Eocene and Lower 
Oligocene time, seven of the genera closely correspond in the two 
countries, namely, Mongolia and the Rocky Mountain region of North 
America. The geologic distribution of these genera and species in 
descending order is as follows: 

Lower OLIGOCENE, ARDYN OBO FORMATION, BRONTOPS GOBIENSIS LIFE ZONE. 

Brontops gobiensis, the broad-skulled ‘thunder-eyed’ titanothere of the Gobi desert, 
in a similar stage of evolution to Brontops brachycephalus of the Chadron A 
formation, Lower Oligocene, South Dakota. 

Menodus mongoliensis, the ‘long-skulled’ titanothere of the Gobi desert, in a 
‘similar stage of evolution to Menodus giganteus of the Chadron B formation, 
Lower Oligocene, South Dakota. 

UrrerMmost Eocenr, SHARA Murun Formation, PROTITANOTHERIUM MONGOLIENSE 
Lire ZONE. 

Protitanotherium andrewsi, named in honor of Roy Chapman Andrews, leader of 
the Third Asiatic Expedition; in a somewhat more advanced stage of evolu- 
tion than Protitanotherium superbum of the Uinta C formation, northern 
Utah. 

Protitanotherium mongoliense, the titanothere first named from Mongolia, believed 
to be in a somewhat similar stage of evolution to the Protitanotheriwm super- 
bum of the Uinta C formation, northern Utah. 

Dolichorhinus kaiseni, named in honor of Peter Kaisen, a member of the Third 
Asiatic Expedition; believed to be in a more progressive stage of evolution 


eens of the Asiatic Expeditions of The American Museum of Natural History. Contri- 
bution No. 


2 AMERICAN MUSEUM NOVITATES ' [No. 202 


than Dolichorhinus (cornutus) hyognathus of the Uinta B formation, northern 
Utah. 

Genus?, species?, a long, slender-limbed titanothere, possibly ancestral to Menodus 

mongoliensis, represented by the fore limb bones (Amer. Mus. 20253). 
Uprer Eocenst, Irpin Manua FORMATION, PROTITANOTHERIUM GRANGERI LIFE 
ZONE. 

Protitanotherium grangeri, the first titanothere found in Mongolia, named in 
honor of Walter Granger, chief paleontologist of the expedition; believed 
to be in a stage of evolution similar to that of Protitanotherium emarginatum, 
lower levels of the Uinta C formation, northern Utah. 

Dolichorhinus olseni, named in honor of George Olsen, a member of the Third 
Asiatic Expedition. 

Telmatherium berkeyi, named in honor of Charles P. Berkey, chief goolenes of 
the expedition, who collected the first titanothere jaw. 

Manteoceras? irdinensis, of doubtful generic affinity to the M ania of the 
Uinta B, Utah, named specifically in reference to its discovery in the Irdin 
Manha formation of Mongolia. ; 

Metarhinus? mongoliensis, of doubtful generic reference to the fae Meta- 
rhinus of the Uinta B of Utah, named specifically in reference to its dis- 
covery in Mongolia. 


FIELD CATALOGUE AND Notes BY WALTER GRANGER, 1922 and 1923 

Of great importance is the precise geographic and geologic distribu- 
tion of titanothere remains, skulls, jaws, and portions of the skeleton, 
collected from the three chief localities. Treated in geologic descend- 
ing order, the records are as follows. 


Ardyn Obo Formation, Lower Oligocene. Teeth, Jaws, and Skull col- 
lected on the Kalgan-Uliassutai Trail, southeastern Gobi, Mongolia 


Amer. Mus. Genus and Species Material 
No. 
20354 Brentons gobiensis, new species. Crushed skull. July 2, 1923. 
20353 X M referred. Sympbhysis of jaw. June 30, 1923. 


20351 Menodus mongoliensis, new species. Lower grinding tooth. June 29, 1923. 


Shara Murun Formation, uppermost Eocene. Remains of twenty-seven 
Titanotheres collected on the Kalgan-Uliassutai Trail, southeastern 
Gobi, Mongolia, chiefly during the season of 1923. Type of Proti- 

tanothertum mongoliense collected in 1922 

Amer. Mus. Genus and Species Material 

No. 


20252 Dolichorhinus kaiseni, new species, Male skull and jaws, nearly com- 
plete dentition. ae 


1925] 


Amer. Mus. 
No. 
20257 
20255 
20260 
18653 
20263 
20256 
20261 
20270 
20273 
20271 


20251 


20254 
20262 


20269 
20280 


20272 
20253 


20265 
20268 
20258 


20259 
20267 


20274 
20275-20276 


20277 
20327 


TITANOTHERES OF MONGOLIA 3 


Genus and Species : Material 


Dolichorhinus kaiseni, paratype. Palate and basicranium, complete 


dentition. : 

Dolichorhinus kaiseni, referred. Female skull and jaws, nearly complete 
dentition. 

Dolichorhinus kaiseni, referred. Male skull and jaws, complete den- 
tition. 


Protitanotherium mongoliense Osborn, type. Right ramus, fragment, 
with six grinders. First titanothere to reach Museum from 
Mongolia. 

Protitanotherium mongoliense, neotype. Palate and dentition complete. 

Protitanotherium mongoliense, referred, Palate, canine, superior grinders, 
lower jaws. 

Protitanotherium mongoliense, referred. Female skull complete (crushed), 
incisors, canine, grinders. 

Protitanotherium mongoliense, referred. Right maxilla and zygoma, 
superior grinders. 

Protitanotherium mongoliense, referred. Fine pair jaws, lacking symphy- 
sis, inferior grinders. 

Protitanotherium andrewsi, new species. Complete male skull, denti- 
tion, nasals wanting. 

Protitanotherium andrewsi, paratype. Male? Left ramus, inferior 
grinders. 

Protitanotherium andrewsi, referred. Male skull, superior dentition. 

Protitanotherium andrewsi, referred. Jaws, symphysis, upper incisors, 
grinders. 

Protitanotherium andrewsi, referred. Symphysis of jaws, with canines. 

Protitanotherium andrewsi, referred. Juvenile palate, deciduous 
premolars, first, second molars. 

Protitanotherium andrewsi, referred. Fine jaw, male, inferior dentition. 

Genus? and species? not determined. Premaxilla, incisor; ulna, femur, 
vertebra, foot bones, pro-Menodus phylum? 

Genus? and species? not determined. Maxilla fragment. 

Genus? and species? not determined. Left ramus of lower jaws. 

Genus? and species? not determined. Protitanotherium? Fore foot, 
with distal end ulna and radius. 

Protitanotherium andrewsi?, referred. Part of fore foot (metapodials). 

Genus? and species? not determined, Protitanotherium. Associated foot 
bones. 

Protitanotherium mongoliense, referred. Incomplete hind foot. 

Genus? and species? not determined. Protitanotherium? Fore and hind 
foot bones. . 

Protitanotherium mongoliense, referred. Complete hind limb and foot. 

Protitanotherium andrewsi, referred. Right humerus, 2 right tibiz, 
left tibia, foot bones. 


| 


AMERICAN MUSEUM NOVITATES [No. 202 


Irdin Manha Formation, Upper Eocene. Remains of twenty-six 
Titanotheres collected north and south of the Kalgan-Urga Tele- 
graph Line, southeastern Gobi, Mongolia. Chiefly collected during 


Amer. Mus. 


No. 
20111 
20117 
20167 
20103 
19179 
20120 
20122 
20123 
20126 


20105 
20108 


20113 
20114 
20119 
20112 
20104 
20110 
20106 
20107 
20121 
20127 
20115 
20125 
20124 


20109 
20171 


the season of 1923. First specimen found in 1922 


Genus and Species Material 


Manteoceras? irdinensis, new species. & right lower jaw. 

(Misplaced.) Field label, ‘‘titanothere lower jaw.” 

Metarhinus? mongoliensis, new species. Small lower jaw, two grinders, 
milk superior grinders. 

Protitanotheriwm grangeri Osborn, type. Female skull, jaws, incisors, 
canines, fragmentary molars. 

Protitanotherium grangeri, referred. Left Jaw and symphysis, with six 
grinders. First. titanothere found in Mongolia. 

Protitanotherium grangeri, referred. Perfect portion of right maxilla. 

Protitanotherium grangeri, referred. Juvenile palate and grinding teeth. 

Protitanotherium grangeri, referred. Fragment of left maxilla and 
grinding teeth. 

Protitanotherium grangeri, referred. Fragment of left ramus with 
grinding teeth. 

Protitanotherium grangeri, referred. Large adult male jaw with teeth. 

Protitanotherium grangeri, referred. Small female right maxilla and 
grinding teeth. 

Protitanotherium grangeri, referred. Facial portion of female cranium, 
anterior grinders. 

Protitanotherium grangeri, referred. Anterior portion female cranium 
with dentition. 

Protitanotherium grangeri, referred. Left jaw fragment with four grind- 
ing teeth. 

Protitanotherium grangeri, referred. Young male right ramus of jaw 
with six grinders. 

Protitanotherium grangeri, referred. Very large male jaw with nearly 
complete dentition. 

Protitanotherium grangeri?, referred. Large male jaws with complete 
dentition. 

Telmatherium berkeyi, new species. Adult female lower jaws, right 
maxilla, canines, grinders. 

Telmatherium berkeyi, paratype. Anterior half of jaw, canine, inferior 
grinders. 

Telmatherium berkeyi, referred. Perfect left maxilla and grinding teeth. 

Telmatherium berkeyi, referred. Palate and superior grinders. 

Telmatherium berkeyi, referred. Juvenile lower jaw with incisors and 
grinder. 

Telmatherium berkeyi, referred. Left maxilla with premolar teeth. 

Telmatherium berkeyi, referred. Right jaw with well-preserved grinders. 

Dolichorhinus olseni, new species. Fine pair of jaws. 

Protitanotherium grangeri, referred. Right humerus, 2 right ulna, left 
ulna and radius, axis, right tibia. 


1925] TITANOTHERES OF MONGOLIA 


Or 


BRONTOTHERIIDE 
Brontopine 
Brontops gobiensis, new species 

. Inthe Ardyn Obo formation were found remains of three titanotheres 
in a distinctly Oligocene stage of evolution. The type skull (Amer. Mus. 
20354), named herewith Brontops gobiensis, contains three superior in- 
cisors with distinct tetartocones in the three molars, no hypocone in M?®. 
The massive symphysis (Amer. Mus. 20353) is referred to the same 
species. 

The type of Brontops gobiens7s (Amer. Mus. 20354) consists of the cra- 
nium reconstructed after comparison with that of Brontops brachycephalus. 
The horns partake of the short, broad character of the cranium; they 
are obtusely prominent. The nasals are elongated, broad, and shovel- 
shaped as in Protitanotherium and Menodus (=Titanotherium). The 
symphysis of a referred jaw (Amer. Mus. 20353) is extremely massive 
like that of Brontops. The dental: formula I 3=3 C7=++ P#:¢ M#:3 
agrees with that of Teleodus Marsh in the presence of three superior 
incisors, and differs from that of Brontops brachycephalus Osborn, in 
which the incisive formula is 1323. The cranium is as brachycephalic 
as that of B. brachycephalus. The strong, well-developed tetartocones 
of the premolars or double internal premolar cones mark the chief 
progression beyond Protitanotherium. 


Menodontine 
Menodus mongoliensis, new species | 


The type of Menodus mongoliensis (Amer. Mus. 20351), from the 
Ardyn Obo formation, is a second right inferior molar tooth in which the 
second lobe is perfect, the first lobe is broken in front. The generic and 
specific characters, as compared with the Brontops gobiensis type, are: 

An animal equal in size to the large male individuals of Menodus 
giganteus (=ingens Marsh) of South Dakota; length of two lobes of 
second inferior molar estimated at 93 mm., actual breadth. of anterior 
lobe 40 mm., height of relatively unworn second lobe 53 mm.; promi- 
nent postero-external cingulum; prominent posterior crests on metaconid 
and hypoconid; internal valleys of trigonid and talonid widely open; 
inferior molar proportions similar to those of Menodus, wholly dissimilar 
to those of Brontops, indicating that M. mongoliensis was a long-headed 
rather than a broad-headed animal. 


6 AMERICAN MUSEUM NOVITATES [No. 202 


Manteoceratine (Brontopine) 
Protitanotherium andrewsi, new species 


The seven specimens, including the type of Protitanotherium andrews 
(Amer. Mus. 20271), the paratype (Amer. Mus. 20251), and the referred 
skulls and jaws of this species, as listed above, while recorded from the 
same formation as Protitanotherium mongoliense, exhibit decidedly 
progressive mutationsand rectigradations warranting a specific separation. 
This advance is indicated not only by the greater measurements of the 
cranium and jaws throughout, but by several progressive characters 
more or less clearly observed in the grinding teeth, although the grinders 
do not greatly exceed in length those of Protitanotherium grangert. 

The specific distinctions are as follows: (1) Tetartocones prominent 
in P*3; (2) rudiments of a hypocone in M°; (3) grinders otherwise 
similar to those of P. mongoliense; (4) male canines larger and more 
robust; (5) relative reduction of lateral superior incisors; (6) bluntly 
rounded summits of the second superior incisors. 


The comparative measurements in three of the species of Protztano- 
therium are as follows: 


Basilar Superior Inferior 


Chief Comparative Measurements ’ length of grinding grinding 
skull teeth, teeth, 
P2-Mé P.-M; 
Protitanotherium andrewsi 757 283 308 
Type (Amer. Mus. 20271), paratype -(Amer. 
Mus. 20251) 
Protitanotherium mongoliense 677 269 289 


Type (Amer. Mus. 18653), neotype (Amer. 

Mus. 20263), referred (Amer. Mus. 20261) 

Protitanotherium grangert 695 242 174 
Type (Amer. Mus. 20103) 


Protitanotherium mongoliense Osborn, 1923 

ORIGINAL REFERENCE.—Osborn, H. F., 1923, Titanotheres and Lophiodonts in 
Mongolia. Amer. Mus. Novitates, No. 91, October 17, p. 3. 

Supplementing the type of Protitanotherium mongoliense (Amer. 
Mus. 18653), a right ramus with a well-preserved series of grinding teeth, 
discovered in 1922, is a very fine series of specimens, fully listed above, 
discovered in 1923. Although recorded from the same Shara Murun 
formation as the more progressive species Protitanotherrum andrewst, 
all the above specimens agree in their relatively less advanced stage of 


1925] TITANOTHERES OF MONGOLIA 7 


evolution, as shown chiefly in the rectigradations of the grinding teeth 
and in the inferior dimensions throughout. As pointed out in the original 
description (Osborn, 1923, p. 3), P. mongoliense is intermediate in its 
actual measurements and dental indices between Protitanotherium emar- 
ginatum and P. superbum of the Uinta C formation, Utah. 

SPECIFIC CHARACTERS.—The skull, the two palates, the right maxilla, 

and the two pairs of jaws agree sufficiently with the type jaw in size and 
in progressive evolution of the incisors, canines, and grinding teeth to 
clearly demarcate this species of the Shara Murun from the more primi- 
tive Protitanotherium grangeri of the Irdin Manha formation. At the 
same time they indicate a phase of Protitanotherium evolution somewhat 
more ancient geologically than P. andrewsi, the giant species of this 
genus. Unlike P. grangeri, the inferior premolars (type Amer. Mus. 
18653, referred Amer. Mus. 20256, 20273) exhibit very distinct ento- 
conids in P»-4, with distinct entoconid shelf and deep talonid basin partly 
closed internally by the metastylid. Similarly the superior premolars 
(skull Amer. Mus. 20261 referred, palate Amer. Mus. 20256 referred, 
neotype palate Amer. Mus. 20263, right maxilla Amer. Mus. 20270 re- 
ferred) all exhibit progressive tetartocones in the second and third pre- 
molars, P?’, and a more or less well-developed tetartocone rudiment in 
the fourth premolar, P*. The several maxille and premolar grinding 
series show progressive steps in the development of this tetartocone of 
P*; in fact, the palatal series (Amer. Mus. 20263) exhibits the tetarto- 
cones of P*-* in their most rudimentary stage, whereas the tetartocones 
in the skull (Amer. Mus, 20261) are in their most progressive stage. 


Protitanotherium grangeri, new species 


Occurring in the more ancient geologic formation of the Irdin Manha, 
the type female skull and jaw of Protctanotherium grangeri (Amer. Mus. 
20103) and the numerous referred specimens listed above, found north 
and south of the Kalgan-Urga telegraph line, exhibit very uniform cranial 
and dental characters supplemented by the remains of fore and hind 
limbs of both the right and left sides (Amer. Mus. 20171, belonging to 
several individuals). 

SPECIFIC CHARACTERS.— The exceptional features of the type cranium 
(Amer. Mus. 20103), such as (a) the horseshoe-shaped concavity border- 
ing the posterior nares and (b) the deep pits on either side of the pre- 
sphenoid, characters observed also in specimens of (a) Dolichorhinus and 
of (b) Sphenocelus, led us at first to regard this type as related to the 
above genera. 


8 AMERICAN MUSEUM NOVITATES [No. 202 


The real propinquity of the species grangeri to Protitanotherium is, 
however, firmly established by: (1) The elongate horns, the broad, 
shovel-shaped nasals, (2) the broad sweep of the zygomatic arches, and 
(3) the saddle-shaped cranial top; also (4), in the dentition, by (5) the 
sharply incurved canines, (6) the enlarged lateral superior and inferior 
third incisors, I3, (7) the diminishing second and first incisors, which in 
this species retain their pointed, postcingulate form. 

Dentiti0on.—The ancestral or primitive character of the grinding 
teeth of Protitanotherium granger?, as compared with the more progres- 
sive grinders of P. mongoliense and of P. andrews?, is seen in the abso- 
lutely simple internal deuterocones without even a rudiment of the 
tetartocones. The ancestral character is observed also in the absence 
of any trace of a hypocone in M$ (compare type Amer. Mus. 20103, also 
referred Amer. Mus. 20114, 20108, and 20120). These non-progressive 
superior grinders indicate that P. grangerz is geologically far more ancient 
than the P. mongoliense and the P. andrews? of the Shara Murun. 


7 


Telmatheriine 
Telmatherium berkeyi, new species 


Remains of this very large and deep-jawed titanothere were dis- 
covered in the Irdin Manha formation in two localities intermingled 
with those of Protitanotheriwm grangeri and in one locality intermingled 
with those of P. grangeri and of Dolichorhinus olsen. 

The specific characters are derived solely from the gigantic type jaw 
and maxilla (Amer. Mus. 20106) and frem six other specimens of 
superior and inferior maxille, none of which yields a knowledge of the 
cranium, yet we may be confident that the skull was relatively high and 
deep as well as elongate, that the bony horns were somewhat rudimen- 
tary, while the canine tusks were sharp and powerful. The jaws contrast 
widely with those of the Mongolian Dolichorhinus and Protitanotherium, 
not only in their prodigious size but in the great depth of the mandibular 
rami, of the chin processes, and of the lower borders. The incisors are the 
largest known among the titanotheres, the second pair being greatly 
enlarged. The canines are very large, pointed, sublanceolate. The pre- 
molars have prominent crowns and highly sculptured internal cingula. 
The posterior molars are exceptionally long and relatively narrow, the 
breadth-length index being 82 as compared with 98 in Dolichorhinus and 
115 in Brontops brachycephalus. The elongate jaw of Telmatherium 
berkeyz, measuring 30 inches from condyle to symphysis, nearly 9 inches 


1925] TITANOTHERES OF MONGOLIA 5 


in its deepest portion below Me, M3, and 7% inches at the symphysis, 
is quite unique; it betokens a long deep cranium of prodigious size. 
Unlike its contemporaries, the coronoid is very broad, the condyle is 
only moderately elevated above the line of the molar teeth, the angle 
is small, not prominent; the lower border is deeply depressed below the 
grinding series, then rises slightly below the premolars and sinks into the 
deep, powerful symphysis which rises like the inclined straight prow of a 
ship to the elevated border of the cutting teeth. 

The cutting teeth are all extremely prominent. The prominent 
superior canines resemble those of Telmatheriwm and of Menodus 
(=Titanotheritum). The lateral incisors are smaller than the second 
incisors, although I’? are very prominent, flattened anteroposteriorly, not 
deeply cupped behind. High, laterally compressed inferior premolars 
with sculptured crowns; superior premolars with prominent sculptured 
ectolophs. Molars elongate, compressed, with prominent ectolophs, 
fairly prominent protocones and hypocones. In M? a very’ prominent 
hypocone; this tooth is extremely long and narrow. 


Dolichorhinine 
Dolichorhinus kaiseni, new species 


This species is based upon the type male skull and jaws (Amer. 
Mus. 20252) and the paratype palate and basicranium (Amer. Mus. 
20257) from the Shara Murun formation, of uppermost Eocene age, 
which also yields Protitanotherium mongoliense and P. andrewsi. Strik- 
ingly uniform in character and measurements are the type, paratype, 
and referred specimens, as fully listed above. 

Speciric CHARACTERS.—As shown in a comparative series of 
measurements, Dolichorhinus kaiseni is superior in length of jaw and in 
all its dental measurements to Dolichorhinus olseni; the breadth-length 
indices of the fourth inferior premolar, Ps, and of the third inferior 
molar, M;, are approximately the same in the two species. These rela- 
tively long, narrow indices are harmonic with the extremely dolicho- 
cephalic, elongate cranium and jaws. As in several of the Wyoming 
and Utah species of this genus, the small canines, even in the males, are 
in compensation for the precociously developed horn swellings; these 
bony horns are relatively much more prominent than in Dolichorhinus 
hyognathus (=cornutus) of the Washakie and Uinta formations. Al- 
though the canines are relatively small as compared with the robust 
canines of Protitanotherium, we judge that the four crania (Amer. Mus. 


10 AMERICAN MUSEUM NOVITATES [No. 202 


20252, 20257, 20260, 20255) belong to male individuals because of the 
uniform size of the canines and of the uniform development of the 
protuberant bony horns. These horns project outward rather than up- 
ward; the fore-and-aft diameter of the horn-base is short as compared 
with the elongate horns of Protitanothertum. Below them are the flaring 
sides of the premaxillo-nasal junction, which are extended forward into 
the hooded nasals that resemble the inverted, rounded prow of a boat; 
this exaggerated evolution of the long and narrow nasals suggested the 
name Dolichorhinus cornutus for the ‘horned-long-nosed’ titanothere of 
Wyoming and Utah. Back of this extension is the long tubular cranium, 
which in section is quite hollow, including the large tubular air chambers 
which completely conceal the small brain-case below. 

The cranial proportions are correspondingly elongate, laterally 
compressed, arched superiorly, with slender, flattened zygomatic arches 
correlated with relatively feeble powers of mastication; the breadth- 
length index, length 695 mm., breadth 330 mm., index 47, contrasts with 
an index of 69 in a skull of Protitanotherium grangeri of the same basilar 
length, namely, 695 mm. The horns do not partake of this elongated 
character but are rounded, short, obtuse, and much more prominent than 
in any known American species. 

The jaws are of elongate, slender, angulate character, with dentin 
coronoid processes. The dentition harmonizes with the relatively feeble 
masticating and offensive powers, the canines being relatively small, 
incisors of medium size with posterior cingula, second lower incisors 
slightly enlarged. As in the American species, the premolars are simple 
with large median internal deuterocones and rudimentary tetartocones. 

It is this intermingling of the characters observed in the American 
specimens with those discovered in this Mongolian: species of Dolicho- 
rhinus that gives these animals such exceptional interest. 


Dolichorhinus olseni, new species 


The type of Dolichorhinus olsenz is a fine pair of jaws (Amer. Mus. 
20109) found one-half mile north of the Kalgan-Urga telegraph line. 
This is the only specimen referable to Dolichorhinus discovered in the 
Irdin Manha formation, whereas the more progressive species Dolicho- 
rhinus kaisent is very abundant in the overlying Shara Murun formation. 

The type jaw of Dolichorhinus olseni is inferior in size throughout 
to the referred jaw of D. kaiseni (Amer. Mus. 20260). Besides its smaller 
size we observe the following less progressive characters: (1) Canines 
narrow and sharply pointed; (2) lateral first and second incisors more 


1925] TITANOTHERES OF MONGOLIA 11 


pointed, less broadly cupped or cingulate posteriorly; (3) first premolars 
small, single-fanged, pointed; second and third premolars, P»-s, lacking 
cup-shaped concavity of talonid which is present only in P,; that is, 
all rudiments of entoconid and all evidence of broadening of talonid are 
lacking in D. olseni in P2-s; these teeth are far more primitive than the 
corresponding teeth of D. kaiseni; (4) the fourth inferior premolar still 
very primitive, exhibiting a shallow concavity of the trigonid, a relatively 
broad, slightly concave talon. 


Manteoceratine (Brontopine) 
Manteoceras? irdinensis, new species 


The type of Manteoceras? irdinensis (Amer. Mus. 20111) is a right 
mandibular ramus with symphysis, containing M;-; and alveoli of the 
premolar and cutting teeth. The locality is the Irdin Manha forma- 
tion, Upper Eocene, two miles north of the Kalgan-Urga telegraph line. 
The generic reference to Manteoceras is doubtful and provisional, be- 
cause the dentition does not agree with that of any known American 
species of this genus. _ 

The distinctive characters are: (1) Lower canines enlarged, approxi- 
mated, and strongly procumbent; (2) incisive border of alveoli corre- 
spondingly narrow and laterally compressed; (3) apparently three pre- 
molars, anterior alveolus of Ps? seemingly double; entire premolar 
alveolar border short (80 mm.); molar border relatively elongate (190 
mm.); molars typically titanotheroid; anteroposterior measurement of 
M,=42 mm., of M2=60 mm., of M;=83 mm., transverse measurement 
of M;=33 mm.; (4) symphysis broad, extremely shallow, flattened on 
infericr surface. ; 

On further knowledge, this may represent a new genus of 
titanotheres. 


Dolichorhinine? 
Metarhinus? mongoliensis, new species 

The type of Metarhinus? mongoliensis, consisting of the anterior 
portion of a right mandible (Amer. Mus. 20167), is also from the Upper 
Eocene Irdin Manha formation, the exact locality being unrecorded. 
This jaw fragment, containing Ps, M,, is inferior in size both to Mante- 
oceras? irdinensis and to the smallest individuals of Protitanotherium 
grangert. 


12 AMERICAN MUSEUM NOVITATES [No. 202 


The distinctive characters of this species are: (1) Linear measureé- 
ment of Ps, Mi=43 mm., anteroposterior measurement of M,;=24 mm., 
transverse measurement 14 mm.; anteroposterior measurement of Ps= 
19 mm., transverse measurement 11 mm.; (2) in Protitanotherium gran- 
gert the linear measurement of Pz, Mi1=45 mm., anteroposterior measure- 
ment of M:=24 mm., transverse measurement 17 mm.; (8) valleys of 
metalophid cupped or deeply concave; (4) valleys of protolophid open 
with prominent paraconids. 

While the resemblances of the type teeth to those of both Meta- 
rhinus and Mesatirhinus of Utah are fairly close, the generic reference 
of this svecies is decidedly doubtful. 


bec tants 


vay 


f 
ae 


>> 


= 


ae 


b 
i 

) 
4 


ait a 
igs 


2a) 
shige 
3 Fy ED) 
badicaih 


pps 


Hees 


His 
pF py 


bier 


ss es 
His 


asia elie 
f 


ei: i