Occurrence and Habits of Asian Cervidae and Bovidae

Cervidae in Asia

Some of the more primitive, such as the Musk Deer, Tufted Deer and Water Deer, are partly or chiefly confined to China and Mongolia, where they seem to have found a congenial environment and have persisted to the present time. In these same areas are larger and more progressive species, of which one, Pere David's Deer, is a curiously aberrant type, no longer known in the wild state. Deer agree in their slenderness of limb and relatively light build; the small lateral toes of the fore foot are present with their phalanges, but the corresponding metacarpals are reduced and lack either the upper or the lower end.

The Musk Deer is a somewhat aberrant member of the family, peculiar in lacking a suborbital gland and antlers in either sex, while the presence of a gall bladder is unique among existing species; these characters are believed to be reminiscent of a relation with the Bovidae, perhaps retained from a common ancestry. The lower incisors are alike in having spatulate crowns. For these reasons the genus Moschus is ranked as a separate subfamily in most classifications, and by some authors is even given family rank. In a study of the external characters of Hydropotes, Pocock has drawn attention to the presence of two small inguinal glands, and on account of its lack of antlers and the presence of large tusk-like canines in the male, would make this genus a subfamily, Hydropotinee, considering it one of the most primitive of the deer. Probably Alces should also be given subfamily rank on account of its laterally spread antlers, enlarged muffle, shortened nasals, lack of canines in the upper jaw, short neck, high withers, and lack of spotted pattern in the young.

Antler
In most of the genera antlers are present in the male and usually absent in the female. These are deciduous structures, grown in the summer and carried for most of the remaining part of the year before being cast, after which a new set is grown. In the more primitive deer antlers may be wanting, and their place as defensive or offensive weapons taken by the upper canines which become long, blade-like tusks. In others both these tusks and small antlers are carried, but in the larger species the canines are small or absent, and the antlers of the male may reach large proportions and great complexity of branching. The upper incisors are wanting, but the lower ones are all present, and the lower canine becomes lying alongside the third incisor. The premolars of the upper jaw consist of a double crescent of enamel each, while the molars are of four crescents. The third lower molar has a third or accessory enamel lobe at the posterior end. Usually all the cheek teeth are shortcrowned with definite short roots. Foot glands and tarsal glands are often present, and usually a prominent suborbital gland. Deer are usually fond of forest or other thick cover, and in feeding habits are largely browsers.

MUSK DEER

Externally the Musk Deer are small with rather heavy limbs, the pelage coarse, consisting of brittle, quifly hairs, minutely crinkly; the tail is very short; the lateral digits have the metacarpals represented by their lower ends; there are no tarsal or metatarsal glands and no suborbital gland, but the males have a large preputial gland medially situated, which at the mating season is well developed and gives forth a substance having a strong musky odor from which musk is obtained by the Chinese. A caudal gland is also present in males. The mammas are two pairs, inguinal. Neither sex possesses antlers, but in the male the upper canines are long and laterally compressed, projecting as tusks from the upper lips. As an exception among the members of the family, a gall bladder is present in the liver of this species. In the skull there is no depression in front of the orbit, such as often is found in deer for the reception of the suborbital gland, and this point is useful in distinguishing skulls from those of the Water Deer, Hydropotes. The teeth, except for the canines, are much as usual in the family, short-crowned, with three upper and three lower premolars and three molars, the lower incisors have oval, spatulate crowns. The dental formula is therefore: i.' c.' Pm.' m.'=34. Extemally the Musk Deer much resembles the Chinese Water Deer, but may usually be told by its dark feet, whereas those of the latter are pale mixed buffy gray like the rest of the pelage; moreover, the tusks of the male are much more narrow and compressed.
Musk deer are found in the forested parts of eastern Asia from the limit of tree growth slightly north of the arctic circle southward to the northern China.

ROE DEER

The Roe are small, delicately built deer, standing about 2-5 feet high at the shoulder, and agree with Hydropotes, Alces, and the American Odocoileus in retaining the distal ends of the lateral metacarpals, but the significance of this character loses some of its importance in view of the fact that unlike the last, the vomer shows no tendency to extend backward dividing the posterior nares; furthermore, these genera are widely different in other ways. Extemally, the Roe Deer in winter coat much resembles Hydropotes, but has a contrasted white rump-patch, a conspicuous metatarsal gland high up and marked by a tuft of stiffer hairs; the tail is even more reduced, being a mere stump, and the canines are wanting in the upper jaw except in abnormal cases. Foot glands open on the anterior side of the hind pastems, and there is a suborbital gland- of medium size, which makes a shallow pit in the skull before the eye.

The antlers are normally present in the male only, and are of characteristic and primitive form, with short, stout pedicels, from which the main beam ascends some three or four inches before dividing into an anterior and a posterior branch, the latter forking once more, making three points in all, of which the posteriortnost hooks strongly inward. The basal part of the antler in older animals is much roughened ("pearly"). The lower incisors are as usual in the subfamily, the central pair largest and with the outer anterior comer slightly angular. The upper molars show practically no trace of the small accessory column of enamel in the valley between the two inner cusps, but slight suggestion of it may appear in the two anterior molars of the lower jaw. The young have the coat spotted and striped with white.

The range of the Roe is continuous from the British Isles and the Mec.1terranean coast eastward north of the Himalayas to the Pacific coast of north temperate China and Korea, as far north as northern Mongolia. According to Lydekker, it does not reach the mouth of the Amur River.
This author recognizes three "species"-C. capreolus of Europe an western Asia (the genotype); C. pygargus from western Siberia to the Altai region; and C. bedfordi of China and Korea. In a later review of the subject, Flerov (1928) follows the same course, but in studying the specimens brought back by the American Museum Asiatic Expeditions it seemed to me that t supposed differences separating these deer are, after all, of a quantitative rather than a qualitative nature, and I proposed to regard the eastern forms as subspecies of the European (G. M. Allen, 1930b). Still more recently Barclay (1933) independently comes to the same conclusion, and shows that "the roe of Central and Eastern Asia more closely resemble the roe of Sweden (the type-locality of C. capreolus) than do the roe of Scotland, Spain, France, and Armenia" and that "there is but one species of roe-deer." While the differences between the Chinese and Mongolian anirnal and that of northern Europe are probably small, in view of the lack of topotypes of C. capreolus for comparison, I am for the present assuming that the former is racially different.

Roe Deer occur commonly in China in almost any situation where there is a reasonable amount of cover and a small human population, for they are much hunted for their antlers when these are in the velvet. In ' the northern part of Mongolia where the forest edge comes down to the outer borders of the Gobi, Dr. Roy C. Andrews found them in some numbers as near Urga as fifteen, forty-five and sixty miles to the northeast of that city. Farther westward, he writes that they were abundant in the patches of forest about Sainnoin Khan (8,000 feet), but as they were shedding badly, he was unable to preserve specimens. None was heard of on Baga Bogdo or on Artsa Bogdo, and "indeed, none should be expected because neither of these mountains is forested even to a very limited extent." Northward of the northern Mongolian border, Radde (1862) traced them in the forested area in greater or less abundance from the Khingan Range and the Onon valley to the Syansk Mountains, on both north and south slopes of which they are common. He mentions their herding together in small companies in the valleys in winter, and notes that they avoid thick forest, preferring the lightly wooded valleys and lower slopes of the mountains. In the summer, however, they seek even the subalpine levels and delight in poplar thickets and growths of pine and fir, where they hide and where the does bring forth their young.
To the eastward of the Gobi, its range continues into parts of Hopei, where suitable conditions occur, as in the 'wild country of Laotsatzu north of Peiping, where Dr. Andrews obtained a series. Rhoads (1898) records it from the Imperial Hunting Park near Peiping. No doubt over-hunting has greatly reduced its numbers and deforestation has circumscribed its range in this area. To the south of the Gobi it is found on the Mongolian plateau at Dolon Nor and again over northern Shansi and Shensi to Kansu. In the western highlands it is found about Sungpan, northern Szechwan, for Jacobi records a skull from there, and Mr. Brooke Dolan in 1931 and in 1934-36 obtained it in that vicinity and learned of its presence between Tatsienlu and Dawo. Still farther south it probably reaches its southern limit in the vicinity of Tatsienlu, where, according to Pousargues, it has several times been taken by the French missionaries who have sent the specimens to the Paris Museum. It also occurs to the northeast of the Sungpan district as far at least as the Nan Shan.

Sambar deer

Sambar are found in small numbers throughout much of Yunnan and southern Szechwan as far north as the country about Tatsienlu, the type locality of R. dejeani, and eastward along the southern Chinese border to the island of Hainan. It remains to be shown how different this eastern animal is from C. unicolor, type locality Ceylon, for the name was given provisionally without much comparison. Nevertheless, the presumption is that it would be racially distinct. In Szechwan the sambar occurs west of Tatsienlu around Litang as far westward as Batang (Brooke Dolan); and northward to the high mountains west of Lifanting, according to E. H. Wilson. The same author quotes Major M'Neill as having seen a hind and calf west of Tatsienlu. The hind looked so black among the scrub that at first it was mistaken for a bear, and Wilson calls it the "Black Deer," giving as its Chinese name, "hei lu-tsze." Antlers said to come from Yunnan may be seen, he writes, in any large medicine shop in Chungking, Suifu, and other cities. The extent of the trade in deer antlers in the velvet for medicinal purposes is very large, and Wilson has given some surprising figures of the annual export of these. He states that a large part comes from the littleknown region lying between the upper Min River and the Tibetan frontier, where virgin forests still remain in these high ranges. The upper limits of these forests are the home of deer, but owing to the difficulty of access are seldom visited by foreigners. Dr. R. C. Andrews found sambar in the forests near Mucheng, Yunnan, where they are much hunted by the natives who use dogs io drive them. A female that was killed weighed "at least five hundred pounds." He found sambar again near the village of Watien in southern Yunnan, and his companion, Edmund Heller, succeeded in securing a fine male with antlers of 27 inches. The animals are shy, feeding by night and spending the day in the thick cover of ravine bottoms.

Pere David's Deer

Occurrence and Habits.-The history of this remarkable deer is well known. It seems first to have been brought to the notice of Europeans by Pere Armand David who in 1866 sent a specimen to the Paris Museum, where it was promptly described by Milne-Edwards and named in honor of that indefatigable collector, who, himself, later referred to it under a new name of his own, Cervus tarandoides, in reference to its peculiarly branched antlers in which the anterior fork is sub-branched as in reindeer. The name is, however, antedated by that of Milne-Edwards. David's animal, like all others from China, was one of the herd which for an unknown period had lived in the enclosed Imperial Hunting Park south of Peiping. According to information received later, David (1898) in a letter to Oustalet notes that the species had not apparently been known in a wild state since several centuries before the Christian era! Shortly after P6re David's specimen was secured, attempts were made by Sir Rutherford Alcock and Consul Robert Swinhoe to obtain living animals for the London Zoological Gardens (Proc. Zool. Soc. London, 1868). Attempts to procure young to be raised until they could be shipped alive were at first unsuccessful, but a sldn and skeleton of an adult that had died at Peiping were the first specimens to be received in Great Britain, in 1868. Swinhoe (1870) recounts that Sir Rutherford prevailed upon a mandarin to send five fawns in early July to the British Legation, but three of them died due to improper feeding, while two survived on a diet of elm twigs. When about six weeks old they were described as yellowish brown spotted all over with white. By October scarcely a trace of the spotting remained. These two successfully survived a long ocean voyage and eventually reached London alive and well (Proc. Zool. Soc. London, 1869, P. 468). The further history of this pair is unknown to me, but in November, 1883, the same Gardens received another "young pair" purchased from the Socisis d'Acclimatation of Paris. These had been bred in the Zoblogical Gardens at Berlin from imported parents, so that at that time David's Deer was represented by living animals in at least three European collections. P. L. Sclater in 1895 (Proc. Zool. Soc. London, 1895, p. 688) reports that in a recent visit to the jardin d'Acclimatation at Paris he had seen there a fine herd, "males, females, and young, six in all." In addition, the Duke of Bedford has for some years maintained a small herd on his estate at Woburn Abbey, England, descendants from the stock of the Imperial Hunting Park near Peiping.

Sika Deer

The South China or Kopsch's Sika in former times doubtless had a more or less continuous range in eastern China from the Yangtze basin south into northern Kwangtung, and perhaps still farther south. In the eighties and early nineties, it was not uncommon in a district of the lower Yangtze including the basin of Poyang Lake and the country about Kiente. From this area he named over a dozen different "species" based on individual variations in the intimate structure of the teeth and in the form or degree of divergence of the antlers. While this multiplicity of names gives some idea of the variability of the characters of these parts, it is clear that all must refer to a single species in the usual sense of the word. Sika are still found in Anhwei in spite of constant persecution by native hunters for the antlers in the velvet, which are prized for supposed medicinal virtues. Swinhoe in 1872 wrote that sportsmen at ShAnghai occasionally reported seeing antlered deer on the hills at some distance from that city, doubtless the mountainous country of the Anhwei border, where as recently as 1925 A. N. Steward reports starting "a big white-tailed deer" in open pine forest. In 1842 Cantor recorded the presence of these deer on the island of Chusan off the mouth of river.

Red Deer

This group includes the more typical elaphine deer, exemplified by the Red Deer of Europe and the Wapiti of North America. In these the "bez" or second tine of the antlers comes off close above the brow tine, and, like it, hooks down and forward. Pocock, in his paper in the Proceedings of the Zoological Society of London on the homologies between the branches of the antlers in the Cervidoe, has very beautifully shown that the branching is on a fundamentally dichotomous system. It seems further probable that one tendency in evolutionary progress is to develop additional forkings in the terminal part of the main beam (usually the posterior fork) and at the same time to reduce the length of the more proximal intemodes or portions of the stem of each Y. Thus in the Roe Deer there is a long basal stem before the first fork, after which the posterior tine of the fork divides again into two. In Cervus the part corresponding to this basal stem is very much shortened so that the first anterior fork comes almost at the base of the beam to form a "brow" tine and a main posterior branch. In the Sika Deer as in the White-lipped Deer, the posterior branch of the first fork does not again divide until after a considerable interval so that the next tine is high up on the beam, or, as usually stated, there is no "bez" tine. The more typical members of the subgenus Cervus have, however, shortened this interval, bringing the "bez" tine close to the brow tine as a progressive character, a condition also seen in the genus Rangifer. Pocock has suggested that the "bez" tine is fortuitously added in this case, but the true explanation would seem to be rather that the intemode between the two tines is shortened.

The Red Deer of western Europe tends to develop a terminal "cup," formed by a cluster of three or four short points coming off nearly together, and the same may be seen in large antlers of, for example, the Persian Red Deer. The eastern Asiatic races seem to lack this trait, as do also the American Wapiti, and for this reason the former have been by some regarded as derivatives of the American Wapiti, through a reinvasion of eastern Asia by the latter in late geological times. The American animal, however, is differently colored and larger of skull than the eastern races of Asia, and it seems more logical to regard at all events the Chinese and Mongolian Red Deer as but larger races of the same species as found in Europe.

Elk or Moose

The Elk is a species of the boreal evergreen and mixed hardwood forests, with a' more or less continuous range across northern Europe and Asia to eastern Siberia. It seems to be absent, however, from the region of the lower Amur Piver to the eastward of the Bureja Mountains. In former days it very likely was found as far south in eastern China as the northern parts of Hopei, for Milne-Edwards writes that, according to notes sent him by Pere Armand David, the Elk was said to be found in the northern part of China. There appears to be no recent record of its presence there, however, while the nearest that it comes to Hopei at the present time is possibly northern Manchuria in the province of Hilung Kiang (Sowerby, 1923). Radde writing in 1862 traced its occurrence in southern Siberia along the borders of Mongolia from the middle Oka River, west of Lake Baikal, skipping the Syansk basin, to the Baikal and Apple Mountains, and east to the broad valleys of the Onon River and the lakes of this region as far at least as Tarei Nor in Transbaikalia on the Mongolian border. In the autumn of 1851, he writes, six Elk coming from the northwest appeared at the latter locality, mingling with a herd of domestic cattle, much to the surprise of the natives, to whom this deer was unfamiliar. Evidently it ranges south in Transbaikalia across the Mongolian border to within some sixty miles of Urga, where Dr. R. C. Andrews in August, secured an immature female. This is nearly the southern limit also of various other northern species associated with the larch and birch forests which here dwindle away at the edge of the Gobi.

Family BOVIDAE: ANTELOPES, GOATS, SHEEP, CATNE


Members of the Bovidae are typically grazing animals of plains or mountainous country, less often of forests, and include many diverse types. As a feature in common, the males and frequently the females have true hones, which are permanent outgrowths of the frontal bones covered with a hardened sheath , and, if the American Antilocapridee be excepted as worthy of family distinction, are not shed nor are they branched, although their form is very varied. The incisor teeth, as in the Cervidae, are lacldng in the upper jaw, but in addition the upper canines are never present, while the cheek teeth are usually hypsodont, or with high crowns, in correlation with their grazing habits. The lower incisors are always six, the central ones with their outer comer angular, while the lower canines are, as in the Cervidee, incisiform and lie against the third incisors. The lower cheek teeth are narrow, high-crowned, and the last has an additional terminal lobe; in other respects they much resemble those of the deer, and consist of three premolars and three molars, giving thus the formula: i."g C."r Pm."N m.-'T=32. The lachrymal bone usually is complete without a vacuity between it and the nasal or the frontal. The lateral toes may be absent, but are usually represented by small hoofs ("dew claws") with traces of the phalangeal bones, although the distal ends of the metapodials are not present. The gall bladder is characteristically present. The placenta is cotyledonous as in the Cervidee, but with usually many more cotyledons. The family is characteristic of the Old World except Australia, with a few North American species, mostly of relatively late arrival. Four chief types may be recognized among the living species: the lightly built antelopes, more common in open plains or desert country; the goat-antelopes, represented by a few stocky species of rough mountainous country, serving in some respects to bridge the gap between the antelopes and the third section, the goats and sheep, which are again species of mountain areas; and finally the fourth type which includes cattle, heavier in body, and for the most part of temperate and tropical distribution. All four of these groups are represented

Zeren

Zeren is no longer found in that area nor in Manchuria, but reaches its eastern limit in the Khingan Mountains at the eastern end of Mongolia. It abounded on the plains between the Onon and Kerulen Rivers, but became rare in the Selenga basin and did not pass northward of the Syansk Mountains in southern Siberia. To the westward it occurs across northern Mongolia in the grasslands to the Altai region, merging no doubt with the Altai race, off that is really distinguishable. In winter there was some evidence that it migrated southward a short distance from the northern portion of this range.

It is a characteristic and abundant animal of the great grassy plains of northern Mongolia, in contrast to the Goitred Gazelle, which is a more desert-loving species and is typical of the barren deserts of the Gobi to the south and west of the grasslands, although at times both animals are found together. This distinction in habits and habitat was found by Dr. R. C. Andrews to be frequently very sharp, so that in passing from Uskuk ten miles south to Loh, this species was abundant at first, but

Gazelle


The necks of the bucks are considerably swollen. "When in pursuit of a female the buck holds his tail straight up in the air" (Wallace). He found these gazelles in mid-January in small bands of a male and three to half a dozen does on very bare open plains country, sometimes "on absolutely level, stony ground dotted with miserable, stunted bushes, or again in hollows, half hidden by clumps of long yellow grass, but always wary and difficult to approach, seldom allowing the hunter to come within two hundred yards. Dr. Roy C. Andrews found this species to be typically a desert-lover, preferring and country with scant vegetation rather than the grassy plains which are the especial home of Prodorcas gutturosa. This preference results in the practical absence of the Goitred Gazelle from the northern parts of Mongolia, but in the southern parts of this country they follow the desert eastward to about the region of Kalgan on the southeastern edge of the Mongolian plateau. In his march across from Kalgan to Urga, he saw many between Pangkiang and Tuerin, but none between the latter point and Urga, nor between Urga and Sainnoin Khan to the west and southwest, for their place is taken in this grassland area by the other species (Prodorcas). Rarely the two may be found in the same herds where these types of country meet, but since the present species is decidedly a desert animal, it does not gather into such large bands. He writes: "there were hundreds of gazelle all of this species except two of the other and mostly females. On the great plain just north of Tsagan Nor it was present in extraordinary numbers, and here the males far more than the females. The bucks were in herds of four or five to 30 or 35, and of all ages from oldest to two years old. We saw a few herds of two or three females with as many young and one or two bucks. As a rule, however, the latter were together and the females either alone or in couples with their young. A few times we saw one female with two young but not often, usually with but one young. At Loh a young one about ten days old was caught on June 27th, and all the adult females were in milk." The country about Loh he describes as a plain about five miles wide in a north-south direction and extending for many miles east and west. To the north is a range of low but very rough mountains. The plain is of rather fine gravel mixed with a little sand and sparsely covered with bunches of grass about three inches high and with here and there a clump of desert bushes. There are many typical "bad-land" ravines, buttes and knolls of absolutely clean sand and clay, very red and without a spear of vegetation, forming an absolutely desert region. In the dry river bottom are a lot of sand "nigger heads" topped with a "sage-brush" growth. Dr. Andrews remarks on the speed and endurance of these gazelles. "Their gait is strikingly different in running from that of the other species. When not going at full speed they progress in a series of bounds which make the animal appear to be on springs; but when going at top speed they seldom

Serow

The mammas are two pairs, inguinal, as in the Cervidoe, and there are foot glands present. The skull shows a number of interesting characters in comparison with the sheep and goats. There is a shallow depression or pit occupying most of the lachrymal bone, for the reception of the suborbital gland as in the deer, but the bone is large and articulates closely with the frontal and nasal instead of being separated from them by a vacuity as in deer and goats. The nasal bones are wide basally, tapering to a median point distally; they abut squarely against the anterior ends of the frontals and continue their lateral contour. Laterally they are in close union with the anterodorsal border of the lachrymals, and partly with the upper border of the maxillaries, but distally lose their contact with the latter. The premaxillaries, instead of touching the nasals at their upper ends, do not quite reach these bones, but the interval is much nearer being closed. The basicranial axis forms a wide angle with the palate, but even so is less bent than in the latter genus. The bony ridge below the orbit which marks the dorsal extent of the facial muscles does not reach the level of the orbit, whereas in the latter genus it extends forward half-way up in front of the orbit. The lower incisors have short, nearly spatulate crowns with the anteroexternal angle of the central ones rather squared than produced outward. The upper premolars have well-developed vertical ridges at the front and back corners exteriorly, while the molars have similar ridges at the front and middle of the outer side, marking the outer comer of the two crescents. There are no accessory enamel columns between the transverse crests of the molars.
Pocock has described the preorbital gland as a thick-walled, nearly spherical sack, "absolutely packed with long hairs growing nearly vertically from its walls and protruding as a tuft from the small, circular, non-valvular orifice." He found pedal glands on both fore and hind feet, opening by a small circular orifice anteriorly, at the summit of the interdigital cleft, "exactly as in Ovis and Ncemorkedus." In the skeleton of the limbs, the radius and ulna, though closely appressed, remain separate at least until the animal is fully adult.
This genus is represented by but one or two living species, found from Sumatra and the Malay Peninsula to Kashmir and the mountains of western and eastern China. It has been divided into a number of local races of which Lydekker (1913) regarded no less than nine as valid, with a second species from eastern China, which,'as Pocock shows, is hardly more than another race. The type species is Antilope thar Hodgson, a race of Capricornis sumatraensis Bechstein. Three races are probably to be distinguished among the Chinese serows, but the color differences are variable so that their discrimination requires a comparison of representative series. A dwarf species is found in Japan.

Takin

W.N.Fergusson(Ig,j),andinKansubyWallace(I913). Fergussondescribes the haunts of the takin as he saw them in the Wassu country and the Changmin district, slightly to the northeast of Muping. He writes that in the winter season the animals come down off the upper levels to sheltered valleys, and in the spring migrate upward again- to the grassy plains where they spend the summer near and above tree-line. In the exceedingly rough country of these high mountains, takin are to be found at elevations of between 8,ooo and I4,000 feet in dense thickets of rhododendron and dwarf bamboos near the upper limit of tree growth. The animals break out narrow trails through this thick undergrowth and use them regularly in their passage to and from grazing areas and salt licks. E. H. Wilson (1913) gives an excellent account of the native methods of its capture by the use of deadfalls or the spear-trap, which is tripped off by the takin in passing along the path, releasing a spear or knife fixed to the end of a heavy beam which comes down upon the victim from above with such force as to drive the blade almost through the animal's body behind the shoulder. "Around Wa Shan the Takin is lolled by an arrow shot from a cross-bow fixed by hunters alongside the run or by an ingenious gun device. It is also captured by cunningly arranged foot-snares" (E. H. Wilson, 1913). The natives hunt the takin for its flesh which is much esteemed by them, although Wilson regards it as decidedly inferior, though better than that of the serow. He writes that the native name is everywhere "yeh niu," equivalent to Wild Cattle or Wild Ox.

The range of the takin in western China includes the high country from south of Tatsienlu northward more or less continuously to the borders of southern Kansu and thence eastward to the Tsingling Range. In the latter area it is slightly paler, lacking the black eye-ring, and acquiring a more golden tint, the race B. t. bedfordi, but exactly where or how intergradation takes place has not been worked out. Probably typical B. t. tibetana extends to Kansu and northeastern Szechwan, and for the present the race B. t. bedfordi is asstuned to be more or less restricted to Shensi. "In certain places, like the wild country between Lungan Fu and Sungpan, the Panlan range, and in the petty state of Yutung, it may be said to be common. Anywhere in these regions where there are 'salt-licks' this animal is to be found. In western Szechwan its eastern limits are the high ranges forming the western boundary of the Red Basin" (E. H. Wilson, 1913). There seems to be no evidence of its presence in Yunnan, although it may yet be found in the high mountains of the southwestern part. It would seem as if the great bend of the Yangtze formed nearly the southern boundary of the area to which this race is confined at the present time, and that its range is not quite continuous with that of the race in the mountains of Nepal and Bhutan to the westward.

Sheep

There is, however, a slight individual variation in this respect, for in one skull examined the inner faces of the middle third are practically parallel.No doubt the local range of Mountain Sheep has been somewhat restricted during the last century in the eastern part of its range, through human interference and other causes. Radde, writing about i86o, says that they are not to be found in the Kentai, Khingan and Bureja Mountains of southeastern Siberia along the northern, border of Mongolia, but reach the upper course of the Selenga River and the mountains of extreme eastern Mongolia. In Pallas's time they were apparently founding other localities. Twenty years after Radde, Przewalski found them in his first journey into the mountains of Suma Khada, northern Shansi (northwest of Peiping), and probably the specimen described by Peters as O. jubata came from near this same region. Przewalski found them in small bands of from five to fifteen, keeping pretty much to the upper slopes of the mountains, coming at intervals to the flatter country) out the base to drink at springs, and sometimes even mingling with the cows and sheep of the Mongols, pastured there. He supposed they moved but little from chosen localities, but it seems more likely that from time to time they seek new grounds as they are known to do elsewhere. In the waterless country of the Khurkhu Mountains, Przewalski believed the sheep obtained their moisture from succulent vegetation, such as the plant called "budargana" and occasional wild onions. He found argali in the mountains bordering the northbend of the river Khuanke (Chuanche?) and in the Shug Range, while Kozlov also reported them from the neighboring Burchan-Budda Mountains. In his first journey Przewalski had reports of sheep in the mountains of Kansu, perhaps in the western parts, and he later noted their presence in the northern Ala Shan, where Kozlov also found them, as well as in the Humboldt Range of Nan Shan. In the extreme northwest of Mongolia, sheep are present in small numbers, as reported by Demidoff (igoo), along the River Suok and the upper Kobdo. Apparently the ranges of hills and mountains stretching more or less brokenly across the central Gobi from northwest to southeast have been a factor in enabling sheep to find their way across to northern Shansi. Beyond the mountain plateaus of Shansi and the mountains north of Peiping they seem not to have penetrated within historic times, although Sowerby (1924) mentions a report of one in southern Shansi. It is interesting, however, that Matsumoto (1926) has found fossil remains of a similar or identical sheep which he names Ovis ammon shantungensis, from near Chinchow in Shantung, indicating a wider range for the species at not very remote date.