MAMMALS COLLECTED BY THE SMITHSONUN-HARVARDEXPEDITION TO THE ALTAI MOUNTAINS, 1912.By N. HOLLISTER,Assistant Curator, Division of Mammals, United States National Museum.INTRODUCTION.The fauna of the Altai Mountains, Siberia and Mongolia, has beenvirtually unrepresented in American museums. The United StatesNational Museum, therefore, gladly welcomed the generous invitationof Dr. Theodore Lyman to participate, with the Museum of Compara-tive Zoology at Harvard, in a zoological expedition to that part ofAsia during the summer of 1912, and I was detailed as its representa-tive for field work.Doctor Lyman's plans were to foUow the routes of the Demidoff ^and the Swayne ^ sporting expeditions to the country of the famousAltai wild sheep, on the Mongolian side of the Little Altai, and, whilesecuring good specimens of the sheep, ibex, and other large game ofthat region, to make the most of the opportunity afforded to coUectthe smaller vertebrate animals for the two museums. The resultingtrip extended to nearly four months, but so much time was takenby the long journey to and from the scene of operations that onlyabout thirty-five days of actual collecting were enjoyed. We werefortunate, however, in our selections of camp sites in three differentphysiographical regions, and by working at high pressure when theopportunity offered we were enabled to make much greater andmore varied collections than we had anticipated. Doctor Lyman,whose time was chiefly devoted to the collecting of large game, hadengaged as my assistant in the work of collecting the smaller verte-brates Conrad Kain, of Vienna. Without the assistance of thisindefatigable mountaineer, whose resourcefulness, ability, and goodcompanionship are known to Alpine Club people of many countries,the collection would be much smaller than it is. The collection of
' E. Demidoff, After Wild Sheep in the Altai and Mongolia, pp. i-xii; 1-324; numerous illustrations andmap, London, 1900.2 H. G. 0. Swayne, Through the Highlands of Siberia, pp. i-xiv; 1-259; 60 illustrations and map, London,1904. Proceedings U. S. National Museum, Vol. 45?No. 1930. 507
508 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.45.mammals, including about 350 specimens, lias been turned over tothe National Museum for determination. An account of the birds,with descriptions of three new forms, has been published by Mr.Outram Bangs.^ The collections have now been divided betweenthe two institutions. ITINERARY.Having been joined by Kain in London the latter part of May, 1912,we proceeded to St. Petersburg, completed the arrangements forpermits with the Russian officials, engaged the services of an inter-preter, and departed over the Siberian Railway, June 8, for the East-On the morning of June 12 we left the train at Novonikolsevsk, onthe Obi River, and before evening were embarked on a steamer forthe upper river. Biisk, the head of navigation for steamers, wasreached on the night of June 15. Arrangements for horses andtarantasses were made for the first stage of the trip over the postroad to the southward and on the morning of June 18 we began thelong trip by team to the last Russian outpost, Kosh-Agatch. AtUst-Inya, somethmg over half the distance, we negotiated withKalmuks for horses for our pack trip beyond Kosh-Agatch, and thesereached Kosh-Agatch with us June 28. A delay of two days atUst-Inya gave me the opportunity for setting a few traps, and thefirst mammals were collected there.Kosh-Agatch, on the Chuisaya (or Tchouia) Steppe, is the end ofthe road, and from here on to the Mongolian border the trip was bypack and saddle horses. Our party now consisted of Doctor Lyman,Kain, and myself, and the Russian interpreter and four native Tartarsand Kalmuks, with 15 horses. Two days' journey across the desertsteppe and up the valley of the headwaters of the Chuya (or Tchouia)River brought us to the "last timber," a few larch trees {Larixsihirica) near Tschornia Creek, a branch of the Chuya near the Mon-golian border, m Tchegan-Burgazi Pass. Here a suitable camp waspitched and I settled down, July 1, with Kain and one Tartar campman, to make a collection of the mammals and birds of the LittleAltai. Doctor Lyman, with the rest of the outfit, pushed on throughthe pass for a circle of the range in Mongolia after big game, particu-larly sheep and ibex.Except for a few larches in some of the valleys on the Siberian sidethe region is without timber, though there is evidence that othersheltered slopes have been deforested within comparatively recenttimes. Trees were, however, never plenty, and the constant drainon the limited growth by Elirghiz, Kalmuk, and Tartar has almostcleaned up the remnant. The Mongolian side is absolutely withoutbush or tree, and as on the Chuisaya Steppe, "tezelc," or dry horse-
? BuU. Mus. Comp. Zool., vol. 54, No. 16, pp. 463-474, January, 1913.
NO. 1990. MAMMALS FROM THE ALTAI MOUNTAINS?HOLLL^TER. 509dung is the principal fuel for cooking. The country about the siteof our Tchegan-Burgazi camp is steep foUing mountains, with rocky-slides and beautiful alpine tops. Close by, to the southwest, is thesnow-capped border range, and from the higher points the countryas far as can be seen is rolling grassy alps, interspersed with rocky,desert areas, and snowy ranges. The altitude of our camp in thevaUey bottom Avas 8,300 feet, and the highest j^oint near camp for aday's collecting, about 12,000 feet. The rolling alplands averageabout 9,000 to 10,000 feet.While we were engaged m work in the Tchegan-Burgazi, DoctorLyman had a most interesting and successful trip into Mongolia, byway of the Tchegan-Burgazi Pass, returning by the Tarkuta Pass andover the range to our camp. His hunting was chiefly in the SuokPlain and Tayliike Valley country, desert mountains and plateaux,and his notes on this trip will be found beyond with the text on thesheep, ibex, and other game animals.The Tchegan-Burgazi camp was occupied, and active collectingcarried on until July 27, when. Doctor Lyman having returned, webroke camp and moved northward to the center of the ChuisayaSteppe. ? One night's trapping and a day of shooting and preparingspecimens gave us a small collection of steppe species and we reachedKosh-Agatch on the return trip July 29. Two days were occupiedwith packing and preparing for the homcAvard journey, the plans forwhich included a week's stay in the heavily forested mountains northof Ongudai, and on July 31 we departed over the post road to the north.August 5 we reached the point we had selected as a good place towork in the forested mountains, and pitched camp at 6,875 feet, justin the edge of the forest, at the north end of Tapucha Pass, about 5miles south of the village of Tapucha. This locality is about 125 milessoutheast of Biisk. The camp was m a beautiful forest of nut pine(Pinus cemhra), heavy with moisture and thick with underbrush andwindfalls, with alps and rocky tops near by, and snow peaks in thedistance. Here collecting was carried on until August 11, when, theweather having warned us of the severe mountain winter approach-ing, and our time being nearly up, we left on the last part of thetarantass trip, and reached Biisk late on the night of August 14.The trip by steamer down the Obi was made without incidentworthy of note, and on August 21 we took the westbound train atNovonikolsevsk for St. Petersburg. We were fortunate in many ways,and though the weather throughout our stay in the Altais wasunspeakably bad, and rain, sleet, hail, and snow the rule rather thanthe exception, yet the trip passed off without serious trouble. Wewere treated with great kindness and attention by the Russian officialsat every point, and the settlers along the post roads always gave usthe best to be had.
510 PROCEEDINGS OP THE NATIONAL MUSEUM. VOL. 45.Thirteen new species and subspecies of mammals, collected on thisexpedition, have been described in preliminary papers. These are:Sorex rohoratus.Myotis petax.Mustela hjmani.Mustela Jineiventer.PJiodopus crepidatus.Myopus morulus.Apodemus nigritalus.
Sicista napsea.Allactaga grisescens.Eutamias asiaticus altaicus.Ochotona nitida.Lepus quercerus.Procapra altaica.ANNOTATED LIST OF SPECIES.SOREX ARANEUS BOREALIS Kastschenko.1905. Sorex araneus horealis Kastschenko, Oda. MjieKonnT. 3an. CHdnpn mTypKecTaHa [Synopsis mammals western Siberia and Turkestan], I, Chiropteraand Insectivora, p. 85. Tomsk.Six specimens from alpine meadows at 9,000 to 9,500 feet inTchegan-Burgazi Pass, and tliree specimens from the forests nearTapucha at 6,400 feet. The series, collected in July and August,contains specimens in both the summer and winter pelages. Thisrace averages somewhat smaller than any European form of' araneus,except the Spanish granarius. It is otherwise very like true araneus.No definite type-localit}^ is given in the original description, but theaccount is largely based on specimens collected by Middendorff inthe "far north." The range, as given b}^ Kastschenko, includes allof extreme northern Siberia, all the mountain ranges from the Altaieastward, and possibly all of eastern Siberia.Measurements of Sorex araneus horealisfrom the Altai Mountains, Siberia.
Locality.
NO. 1990. MAMMALS FROM THE ALTAI MOUNTAIN8?R0LLISTER. 511lorealis, common in the same forest, by its large size, dark color, andlarge hind foot. In general size it approximates Sorex araneus tetra-gonurus, but seems not to be closely related to any known form.MYOTIS PETAX Hollister.1912. Myotis petax Hollister, Smiths. Misc. Coll., vol. 60, No. 14, p. 6, No-vember 29.No bats were seen in the desert mountains along the Mongolianfrontier. The single specimen collected on the trip, the type ofthis new species, flew into the window of our room in the post stationat Kosh-Agatch on the evening of July 30. Bats seem to be rarein the Altai region and only two or three were seen during the tripson the upper Obi and on the long drive from Biisk to the southward.Two bats recorded by Kastschenko^ from Cherga, as Vespertiliodauhentonii, are doubtless of this species.CANIS ALTAICUS (Noack).1910. Lupus altaicus Noack, Zool. Anz., vol. 35, p. 465.There is reason to believe that wolves are fairly plentiful in theAltai; but, strange to say, only one was seen alive by members of ourparty. The perfect skull of a wolf, found dead near our Tchegan-Burgazi Pass camp, was preserved. Compared with skulls of Canislupus from Sweden, this specimen has considerably smaller teeth, alonger, more slender rostrum, and higher, rounded audital bullae.The differences seem great enough to warrant the separation of theform. Without specimens of the earlier named varieties from Tibetand China for comparison, I use the name altaicus of Noack as un-questionably referring to this animal. We saw a number of woKskiQs, including some of small pups, in the trading post at Kosh-Agatch.That the wolves destroy many wild sheep, especially during thewinter season, is hardly to be doubted. The great numbers of sheepskulls and skeletons found in the mountains show that the gamesuffers considerably from some cause. Referring to conditions on theMongolian side of the range, Lyman writes in his journal:The bottom of the valley is a perfect bone yard. On an average there must be asheep head for every 200 yards over the 10-mile range of which I am writing. Myhunter found a large ram freshly killed by a wolf. Wolves may be the destructiveagency. I saw but one in the valley, at a considerable distance. Among the greatnumber of sheep heads in the valley bottom I saw none of very large size.Demidoff,^in 1897, whilehunting in what seems to be the exact valleyof which Lyman writes, had some of his ponies attacked one night bywolves.The woK skull from Tchegan-Burgazi Pass is fully adult, with basalsuture closed, and is probably a male. It measures: Condylobasal
' Pes. AjiTaficKoa sooji. Skcd. 1898. [Results Altai Zool. Exped. 1898], Vertebrates,p. 12, Tomsk, 1S09.3 After Wild Sheep in the Altai and Mongolia, 1900, p. 143.
512 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 45.length, 244 mm.; zygomatic breadth, 140; mastoid breadth, 82.5; post-orbital constriction, 41; interorbital constriction, 45.5; rostralbreadth over canine, 47; depth of braincase, crest excluded, 64.5;palatal depth behind tooth row, 66; rostral depth behind canine, 36.1;mandible, 190; maxillary tooth row, 107.5; mandibular tooth row,122.5; upper carnassial, 25 by 14.2; first upper molar, 16 by 20.5;lower carnassial, 27.9 by 10.7.VULPES VULPES TARIMENSIS Matschie.1908. Vulpes tarimensis Matschie, Wiss. Exp. Filcliner China uud Tibet, vol.10, 1 teil, p. 166.The fox is said to be common in the desert mountains in thevicinity of Tchegan-Burgazi Pass. Lyman saw one on the Mongolianside, and we found two skeletons, from which the skulls were saved,on the Siberian slopes. In the trading post at Kosh-Agatch were anumber of skins. We purchased a skin from Kalmuks which, fromtheir account, must have been collected near the site of our TschorniaCreek camp, where we obtained the skullsThe form represented is separable from the European Vulpesvulpes by its peculiar color and the large, high audital buUse. Thegeneral color of skins is a yello\vish-buff ; the face, top of head, andan irregular dorsal stripe tawny-ochraceous; ears black outside.The tail is quite grayish, in marked contrast to the color of the body.The throat, upper breast, and stripe the entire length of ventral side,slate color, mixed mth white and buff. This fox is apparently theVulpes tarimensis of Matschie, based on the Jlavescens of Blanford *(not 'i Jlavescens Gray). Noack has already referred the Altai fox^to Vulpes Jlavescens Gray. Owing to the general confusion regardingthe foxes of central Asia, and to the uncertainty as to the origin ofthe type oi Jlavescens, it seems best, at present, to follow Matschieand treat the northern fox of the Jlavescens type as distinct.The Kalmuks call this fox '^dil-koo.'"Cranial and dental measurements of Vulpes vulpes tarimensis from the Tchegan-BurgaziPass.
175175.Male ? ad. 175176.Male ? ad.
Condylobasal lengthZygomatic breadthMastoid lireadthInterorbital breadthRostral breadth over caninesPalatal depth behind tooth row.Rostral depth behind canineMandibleMaxillary tooth rowMandiliular tooth ro'^'Upper carnassialFirst upper molarLower carnassial
mm.1377347.126.322.934.716.71066571.76 by 7.8by 12. 6by 5.9
1 Scientific Results of the Second Yarkand Mission, 1879, pp. 23-24.
? Zool. Anz., vol. 35, 1910, p. 457.
47.72624'34.816.8102.562.563.713. 5 by 8.3A 5 by 12. 516.4 by 6.5
NO. 1990. MAMMALS FROM THE ALTAI MOUNTAINS?UOLLISTER. 513MUSTELA LYMANI HoUister.1912. Mustela lymani Hollister, Smiths. Misc. Coll., vol. 60, No. 14, p. 5,November 29.A stoat was seen running over the rocks at our Tschornia Creekcamp, early in the morning, July 1. At the camp in the timber 5miles south of Tapucha Ave were fortunate enough to shoot the adultmale which was afterwards made the type of this new species. Noweasels of the M. nivalis group were seen; several skins in full winterpelage, noted in the trading post at Kosh-Agatch, were of the largerspecies and doubtless lymani. Kastschenko has, however, recordedM. nivalis from the River Komoorlu in his report on the collectionmade by the Tomsk University expedition of 1901.^MUSTELA (PUTORIUS) LINEIVENTER Hollister.1913. Mustela lineiventer Hollister, Proc. Biol. Soc. Washington, vol. 26, p. 2,January 18.Three specimens, all fully adult, from about 9,000 feet in the openalps at Tchegan-Burgazi Pass, July 8, 10, and 23. All were trappedat marmot burrows, and the habits of the animals seem to be muchlike those of the American black-footed ferret. We saw a numberof skins of this polecat in the trading post at Kosh-Agatch, and ourKalmuck and Tartar camp men knew the animal well, using for itthe Russian name '^Icar-yoke' ,^' which doubtless includes the relatedMustela eversmanni, an animal not obtained by us. The polecatswere perhaps even more infested with fleas than were the groundsquirrels and marmots.MELES AMURENSIS ALTAICUS Kastschenko.1901. Meles amurensis altaicus Kastschenko, Ann. Mus. Zool. Acad. St.-P6ters-bourg, vol. G, p. 613.We purchased a single skin of this badger. It was collected northof Kosh-Agatsh in the Baskkaous Valley country. No signs of badg-ers were noted in the border mountain range.PHODOPUS CREPIDATUS Hollister.1912. Phodopus crepidatus Hollister, Smiths. Misc. Coll., vol. 60, No. 14, p. 3,November 29.Our single night's trapping in the center of the Chuisaya Steppe,July 28, added 16 of these little hamsters to our collection, andseveral other specimens not preserved were caught. The burrows areclosed by the animals during the day, and there is little evidence ofthe abundance of the species until the visit to the traps in the morning.A single specimen was caught by Kain, in his hands, on the mountainabove Tschornia Creek, at 10,000 feet altitude, July 26. Our per-
1 Ann. Mus. Zool. St.-P^tersbourg, vol. 7, 1902, p. 291.80459??Proc.N.M.vol.45?13 33
514 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.45.
sistent trapping on these higher alplands of the frontier range hadfailed to produce a specimen of this mouse. The mountain specimenagrees in all details with the series from the Chuisaya Steppe. Thewonderfully developed cheek pouches of nearly all the specimensexamined were well filled with tiny seeds.MYOSPALAX MYOSPALAX (Laxmann).1773. Mus myospalax Laxmann, Kongl. vet. Acad. Handl., vol. 34, p. 134.One specimen from the border of alps and forest in the mountainsnear Tapucha. It would seem from the great number of earth moundsin this vicinity that this species is an abundant one, but our trapswere not suited for capturing it, and we had difficulty in obtaining evena single specimen. MYOPUS MORULUS Hollister.1912. Myopus morulus Hollister, Smiths. Misc. Coll., vol. 60, No. 14, p. 1,November 29.The unique type-specimen of this lemming was trapped, August 6,under a log in the thick nut pine forest 5 miles south of Tapucha, at6,875 feet altitude. Only a single specimen of the genus has hereto-fore been recorded from Asia. Middendorff^ mentions a skin andskull from the west coast of Okhotsk Sea, which he identified, by acomparison with Liljeborg's colored plate, as the Scandinavian species,M. scJiisticolor. At the same time Middendorff prophesied that theselemmings would eventually be found to range across the continent,and the capture of a specimen in the Altai goes far toward the fulfill-ment of his prediction. Middendorff's record is without doubtauthentic, and is entitled to more consideration than has hithertousually been shown it.EVOTOMYS RUTILUS (Pallas).1778. Mus rutilus Pallas, Nov. Spec. Quad. Glir. Ord., p. 246.1874. Evotomys rutilus Coues, Proc. Acad. Nat. Sci. Philadelphia, p. 187.We found this species abundant in the damp mountain forestsnear Tapucha, at about 6,900 feet. Sixteen specimens were pre-served, collected from August 6 to 11.Pallas described his Mus rutilus from the wooded regions and alpseast of the Obi River. From this, and the fact that in the sameaccount he refers specimens from the mouth of the Obi to his species,but remarks upon a difference in color, it is reasonable to argue thatthe '' type-locality" of rutilus is near the head of the Obi, close to theAltai, and not at a point near the Arctic coast. Our specimens may,then, be considered as fairly typical. Thomas has recently recordedrutilus from the Syansk Mountains, 100 miles west of Lake Baikal.^
? Sibirische Reise, vol. 2, 1S53, p. 108.? Ann. and Mag. Nat. Hist., ser. 8, vol. 9, p. 398, April, 1912.
NO. 1990. MAMMALS FROM THE ALTAI MOUNTAINS?HOLLISTER. 515?In color our specimens from the Altai differ conspicuously from aseries of E. wosnessensHi, from Kamchatka, in the decidedly darkertone of the red, and the much grayer sides. In all the 16 specimens(except one which has the teeth so worn that the enamel pattern isobliterated) the last upper molar has on the inside four salient anglesand three re-entrant angles, the posterior loop being deeply cut byan inner re-entrant angle. Ten adults measure, averages andextremes: Head and body, 101 mm. (95-107); tail vertebrae, 34(30-38); hind foot, without claw, 16.7 (16-17.5).EVOTOMYS RUFOCANUS (Sundevall).1846. Hypudseus rufocanus Sundevall, Ofversigt af Kongl. Vet. Akad. Forhandl.,vol. 3, p. 122.1897. Evotomys rufocanus Bailey, Proc. Biol. Soc. Washington, vol. 11, p. 122,May 13.I can not distinguish the six specimens of this mouse collected inthe Tapucha forests, August 6 to 11, from Scandinavian examples.A single immature specimen collected at Tchegan-Burgazi Pass issomewhat paler, and has grayer sides and a restricted dorsal stripe;but with the variation in both rufocanus and latastei m mind is seemsbest to refer this specimen to the same form as the small series fromTapucha. Thomas has recorded Evotomys rufocanus latastei from apoint 100 miles west of Lake Baikal, and remarks that the specimensagree closely with examples from Kamchatka.^ Dr. J. A. Allen haskindly lent me a series of seven specimens of his latastei, from Gichiga,northeast Siberia, but as his specimens are all immature my com-parisons with latastei are of little consequence. My three immaturespecitnens from Tapucha agree better, however, Avith immatureexamples from Norway than \Adth the young of latastei.The single Tchegan-Burgazi Pass specimen was trapped in openalps at about 9,000 feet elevation. As there are indications of tim-bered draws within comparatively recent times in the immediatevicinity, it is probable that before the almost complete deforestationof such places the species was more abundant in the frontier rangethan it is at the present time.MICROTUS OBSCURUS (Eversmann).1841. Hypudaeus ohscurus Eversmann, Y^en. 3aimcKH, KasancKHM-b jhhb.,(VIII), I, p. 156. [Mem. sav. univ. Kasan.]1899. [Microtus arvalis] var. ohscurus Trouessart, Cat. Mamm., I, p. 558.Thirty specimens, from Tchegan-Burgazi Pass and the forests nearTapucha, seem to represent M. ohscurus, described by Eversmannfrom the Altai. The species is allied to Microtus arvalis, but islarger, with a much larger skull, actually smaller teeth, and is muchdarker and richer colored. In the 30 specimens the true arvalis
1 Ann. and Mag. Nat. Hist., ser. 8, vol. 9, p. 39", April, 1912.
516 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 45.enamel pattern is present in 28. Two have m^ abnormal. Inone, No. 175333, the last loop has an additional inner reentrantangle posteriorly; and another, No. 175325, has the second outersalient angle of m ^ so reduced, and the first inner triangle so length-ened transversely, that it has the appearance of a second loop, muchlike the anterior loop. It seems quite certain that this is an inde-pendent species, and not a race of Microtus arvalis. Poliakoff evi-dently confused other forms with it. It is uncertam from his accountwhether he considered the original example the Altai or the KirgliizSteppe specimen, listed by him. Eversmann in his original descrip-tion, however, mentions only the Altai and gives measurements ofonly one specimen.Measurements of selected adult specimens of Microtus obscurusfrom the Altai Mountains,Siberia.
Locality.
Tchegan-Burgazi PassDoDoDoTapuchaDoDoDoD6DoDoDoDoDoDoDoDoDo
Number.
NO. 1000. MAM}rA Lfi rnOM TTU: ALTAI MOUyTAIXF^?HOLLISTER. 517toward distinctly grooved incisors in several specimens. The nameeversmannii of Poliakoff has been misapplied by Biichner to an arvalis-like Microtus from further south. This has caused considerable con-fusion and Mr. Thomas has recently remarked that "judging fromBiichner's figure of the skull, M. eversmanni Poliakoff, would seemto be either this [M. arvalis] or a related form, and not a Stenocra-niusJ' ^ I believe that Professor Kastschenko, in his paper on theforms of Stenocranius ^ fixes the true status of eversmannii, as amountain form of the wide ranging Stenocranius type. In this paperProfessor Kastschenko quite properly raises the question if the formsof the subgenus Stenocranius linovni to him are not possibly all racesof one wide ranging species. These forms include M. gregalis (Pallas)not Radde,^ not Poliakoff, of eastern Siberia; M. eversmannii (Polia-koff'), in the mountains of south-central Siberia: M. slowzouni(Poliakoff'), of the west Siberian steppes; and M. raddei (Poliakoff),of the Transbaikal country. To these Kastschenko adds, in the samepaper, three new forms of M. sloivzowii; lutea from Tomsk, hrevicaudafrom the Yakootsk District, and trldenticulata; the last described asa "variety," not a race, with abnormal enamel pattern, and comingfrom no special territory.The Microtus eversmanni of Trouessart,* is likewise not the M.eversmannii of Poliakoff, but apparently that of Biichner, and there-fore relates to some form of arvalis.MICROTUS (STENOCRANroS) TIANSCHANICUS Bttchner.1889. Microtus tianschanicus Buchner, Wiss. Res. Przewalski Central-Asien Reis.Zool. theil, vol. 1, p. 107.Thirty-one specimens from the desert mountains near the Mon-golian frontier in Tchegan-Burgazi Pass. The skins in this excellentseries are uniformly much lighter colored than the lot from the forestedmountains near Tapucha, which I have referred to M. eversmannii.Aside from the color differences I can detect no characters to separatethem. The enormous difference in size of adult skulls of this speciesis most exceptional but it seems to represent true mdividual varia-tion. Examination of fresh specimens showed seven plantar tuber-cles and, in old males, large side glands well forward of the hips. Oftwo pregnant females examined July IS, one contained 7 embryos,4 right, 3 left, 12 mm. in diameter; and one 8 large fetuses, 4 rightand 4 left. These mice seemed most abundant in the open alplandsat from 9,000 to 10,000 feet altitude.
1 Ann. and Mag. Nat. Hist., ser. 8, vol. 9, p. 398, April, 1912.
= Ann. Mus. Zool. Acad. St.-P(?teisbourg, vol. e, 1901, pp. 16.5-206.s The Microtus gregalis of Radde and of Poliakoff (not of Pallas), from Transbaikalia, is renamed Microtuspoljakowi by Kastschenko in his paper in the Ann. Mus. Acad. St.-P6tersbourg.
? Cat. Mamm., vol. 1, p. 699.
518 PROCEEDINGS OF THE NATIONAL AJUSEUM. VOL. 45.
Measurements of selected adult specimens of Microtus (Stenocranius) from the AltaiMountains.
Locality.
no. 1990. MAMMALS FROM THE ALTAI MOVNTAIN8?H0LLISTER. 510formulas are as in Alticola, and the two distinguishing characters,the large ears and peculiarly flattened skull, seem best considered assubgeneric.Measurements of selected adult specimens of Alticola (Platycranius) strelzovi from theTchegan-Burgazi Pass, Altai Mountains, Siberia.
520 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 45.Three specimens. Two of these were trapped under thorn busheson the dry, semi-desert hills near Ust-Inya, at 3,850 feet, June 24,and one was captured at the Tapucha canip in heavy forest border-ing open alps at 6,875 feet, August 8. The species was evidentlya common one at Ust-Inya, as several examples, either ruined bythe trap or very immature, were caught in the one night's trapping,in addition to those saved. At Tapucha this little mouse occurswith the much larger and differently colored Apodemus nigritdlus,but it seemed by no means as abundant as the latter species. Withthese specimens, and a good series of the larger species from theAltai before me, it is plain that Kastschenko's original descrip-tion of Mus tsclierga relates wholly to this, the smaller, species.Later * Kastschenko received an additional collection from theAltai which contamed specimens of the larger species, and assum-ing that these were the adults of the species previously namedtsclierga, he proceeded to rediagnose his form. The young of the largerspecies, of a size about equal to the adults of the smaller tsclierga,have the hind foot always 22 mm. or more, or about as in the fullygrown (hind foot of adults of A. nigHtalus measures 22 to 25 mm.),while the adults and immature of tsclierga have a hind foot measure-ment of 19 or 20 mm. The form described by Miller from RussianTurkestan ^ as Apodemus microtis seems indistinguishable fromApodemus sylvaticus tsclierga.
NO. 1990. MAMMALS FROM THE ALTAI MO UNTAINS?HOLLISTER. 521SICISTA NAPiEA Hollister.1912. Sicista napaea Hollister, Smiths. Misc. Coll., vol. 60, No. 14, p. 2,November 29.A single specimen of this jumping mouse was trapped in the damppine forest 5 miles south of Tapucha, August 6. It was caught in adensely grown part of the woods near the border of extensive openalplands, at 6,875 feet elevation.ALLACTAGA GRISESCENS Hollister.1912. Allactaga grisescens Hollister, Smiths. Misc. Coll., vol. 60, No. 14, p. 2,November 29.Two specimens from the center of the Chuisaya Steppe, near thebanks of the upper Chuya River, 8 mUes south of Kosh-Agatch,July 28. One of these was trapped; the other was shot by Kain onthe open steppe just before dark. A few burrows, too large forPhodopus, which were seen on the steppe were doubtless the densof these jerboas. As stated in the original description, this newspecies seems related only to Allactaga mongolica and its subspecieslongior. It is possibly confined to the Chuisaya Steppe, the faunaof ? which is decidedly Mongolian, as might be expected from itsisolation from the great Siberian plains and its connection with thenorthern extensions of the Gobi, the Kobdo and Suok Plains, by theTarkuta, Tchegan-Burgazi, and Bain-Chagan Passes.The Kalmuks call this jerboa "yeV-maJin."MARMOTA CENTRALIS (Thomas).1909. Arctomys centralis Thomas, Ann. and Mag. Nat. Hist., ser. 8, vol. 3, p. 260,March.Four skins and skulls and four odd skulls from the Tchegan-Burgazi Pass. The skin of this marmot is one of the chief articlesof trade from the natives. We saw great bales, containing hun-dreds of skins each, in the trading post at Kosh-Agatch, and thetrader informed us that he had only a short time before sent a largeshipment to Biisk. The animal is known to all the natives by theRussian nam.e ^'su-roclc'."The Altai marmot was first separated by Brandt, but his namefor the animal appears to be a nomen nudum, in every instance ofits use, down to Kastschenko's paper on the Results of the TomskAltai Expedition in 1899. In the meantime the name has becomevalid as a synonym of Marmota hohaTc in Trouessart's Catalogue ofMammals, 1897. The Marmota centralis of Thomas, from the north-ern Tian-Schan seems to be identical with the Altai form, however,so the species is provided with a name. The relationships seem closewith the Transbaikal Marmota sihirica, of which centralis will doubt-less prove to be a subspecies.
522 PROCEEDINGS OF THE NATIONAL MVSEUil. vol. 45.CITELLUS EVERSMANNI (Brandt).1841. Spermophilus CTcrsmanni Brandt, Bull. Sci. Acad. St.-P6tersbourg, vol. 9,p. 43.1903. Citellus eversmannii Allen, Bull. Amer. Mus. Nat. Hist., vol. 19, p. 142,March 31.We first saw this spermophile at Clierga, two days' journey southfrom Biisk. Two young specimens were killed in the house where wespent the night at that place. The people called them rats andseemed little surprised to find them m the buildings. The next day,June 20, we noted many of the animals between Shebalina andTapucha. Every grassy fiat was inhabited by them, and though acold rain was falling the spermophiles paid no attention to it andwere apparently all out feeding. Ground squirrels were abundant inthe Alpine flats between Tapucha and Ongudai, but only a few wereseen beyond Ongudai until we reached the desert frontier range. Theanimals were exceedingly abundant about, our Tchegan-Burgazicamp, and during our stay in this neighborhood we collected a seriesof 50 specimens. During the early part of July most of the animals,and especially the nursing females, were in a faded and ragged stateof pelage, but before the end of the month we secured many speci-mens in the bright, fresh coat. The great difference in age betweennumerous young examples indicates a long breeding season. Wefound spermophiles in all sorts of places, and the burrows were placedboth in rocky cliifs and far out in the open alplands. In addition toinnumerable fleas, the spermophiles are terribly infested with botts{Cuterebra, sp.) and it is rare to obtain an example without half adozen or more in some part of the skin.Our Tartar and Kalmuk camp men called this species "you-mali-runk' .'' EUTAMIAS ASIATICUS ALTAICUS Hollister.1912. Eutamias asiaticus cUaicus Hollister, Proc. Biol. Soc. Washington, vol. 25,p. 183, December 24.A single cliipmunk was seen at Cliibit, June 26. Later, in theheavier forested region south of Tapucha, we found the species abun-dant and collected a series of 12 specimens. Not one was seen be-yond a point about 15 miles south of Tapucha. The animal is re-markably quiet for a chipmunk 'and might readily be passed manytimes unobserved. Only rarely did we hear a low "chip" note.OCHOTONA NITIDA Hollister.1912. Ochotona nitida Hollister, Smiths. Misc. Coll., vol. 60, No. 14, p. 4, No-vember 29.Thirty-one specimens; 17 from near the Mongolian border inTchegan-Burgazi Pass, and 14 from the Tapucha camp. Along theupper Chuya River, in Tchegan-Burgazi Pass, we found pikas chieflyin the lower rocky cliffs along the main river and its tributaiy, Tschor-nia Creek, at about 8,500 feet elevation. The call note is decidedlydifferent from that of the American pikas, and was not at first recog-
NO. 1990. MAMMALS FROM THE ALTAI MOUNTAINS?EOLLISTER. 523nized. It is much more like the high pitched note of some small birdand lacks all the nasal tone so characteristic of our species. The oldanimals, from July 1 to 25, were chiefly in the left-over winter pelage,or in a ragged state of molt. The young were already nearly grown,some of them molting the post-juvenal pelage and renewing into thebrown fall coat of the adult. Numbers were caught in rat trapsbaited with rolled oats, a method I have never found at all successfulwith various America-ti species.In the heavily forested mountains near Tapucha we found pikason the rocky tops, and even in dry rocky creek beds far below timber-line. In these latter localities they seemed entirely out of place tous, but were nevertheless fully at home. The specimens taken herefrom August 7 to 10 are in full fresh fall coat. Our Tartar andKalmuk camp men called the pika ^ ' seen-d-staft' ."I beUeve that true 0. alpina, from which I have separated thisspecies, will be found to inhabit the Bia-Altai, the region north ofLake Teletzkoi, and between the Bia and the Katun Rivers. Thereis a specimen of alpina in the United States National Museum labeled"Barnaul."Measurements of Ochotona nitidafrom the Altai Mountains, Siberia.
Locality.
524 PROCIJEDiyOS of the national museum. vol. -15.Juvenal pelage, head and skull of a male in post-juvenal pelage, andan odd skull of an adult. These specimens are nearly topotypes oflugubris, which was described primarily from a specimen taken atOngudai. Neither of the skins has the outer side of the ear entirelyblack, as described of the species by Kastschenko and by Barrett-Hamilton ; this may possibly be due to their immaturity. The adultskull, compared wdth skulls of timidus, agrees perfectly with theaccount given by Kastschenko. It is evident from the differences inthe age of the two young animals secured that this hare breeds atleast twice during the season.LEPUS ZAISANICUS Satunin.1907. Lepus zaisanicus Satunin, Ann. Mus. Zool. St.-P6tersbourg, vol. 11 (1906),p. 161.Four hares from Tchegan-Burgazi Pass, near the Mongolian border,differ fi-om all other Asiatic species, and agree very well with the colordescription and measurements of Lepus zaisanicus, and are thereforeprovisionally referred to that species. The upper incisors of thesespecimens, both juvenal and adult, present a remarkable appearancein being projected far forward. The skull of zaisanicus was notdescribed by Satunin, and it is impossible without direct comparisonto judge if these specimens actually represent his species, but fromthe external characters and on geographical grounds it seems mostprobable. The specimens were collected from July 9 to 25, on thehigh open alplands above Tschornia Creek, at from 8,700 to 10,000feet elevation. Hares were by no means abundant in these moun-tains and comparatively few were seen. Lyman reported seeing aconsiderable number while sheep hunting on the Mongolian side of therange. LEPUS QUERCERUS Hollister.1912. Lepus quercerus Hollister, Proc. Biol. Soc. Washington, vol. 25, p. 182,December 24.The hares of the ^'tolai group" collected on the Chuisaya Steppe,near Kosh-Agatch, differ conspicuousl}' from all the previouslydescribed species from northern Mongolia and Siberia, and are specifi-cally distmct from the hares of the border mountams to the south-ward, which I have referred to Lepus zaisanicus Satunin. Hares aremuch more numerous on the desert steppe than in the higher moun-tains, and six were killed by members of our party during the 24hours coUectmg m the center of the plain. Fom- specimens werepreserved. It is not at all unlikely that this species is confined tothe Chuisaya Steppe, as we failed to find it in the desert mountainsto the southward, and the 'heavy forest on the north must prove aneffective barrier to such a desert species. The Kalmuks call thishare "Ico-yun'J'
NO. 1900. 2JAMMALS FROM THE ALTAI MOVNTAINH?HOLLISTER. 525CAPREOLUS PYGARGUS (Pallas).1771. Ccrvus pygargus Pallas, Reise prov. Russ. Reichs, vol. 1, p. 453.1843. Capreolus pygargus Gray, List mamm. British Mus., p. 176.Roedeer are fairly abundant in the timbered parts of the Altai.We saw many horns, attached to small parts of the frontals, in thetrading posts, and at Kosh-Agatch the trader had a great many pairs.He told us that the animal is very abundant at times north of Kosh-Agatch, and in the ^^dnter many come down in the low foothills neartown. It is then that most of these heads are taken. We obtainedfive representative sets of antlers from this place. In the Tapuchaforest in August, in the early morning or late evening, we frequentlyheard roebucks barking. Lyman's journal, written at this place,contains the following notes
:
The roedeer near the pass to the south of Tapucha appear to spend the day in thethick cover in the valleys. Here they are protected by the rank grass and the bushes.In the late evening they come up on the high rocky outcrops which rise on both sidesof the valley bottom. I saw two in the timber, probably females, and one on thehillside. The calling of the males, which I heard several times, is much like thebarking of an angry dog. At first I thought there were very few roedeer in this region,but later I came to believe they were fairly plenty. If the weather be fine, it shouldnot be difficult to get a shot if one concealed oneself in these outcrops in the evening.Two varieties of the Siberian roedeer have been described by Mehely.^A slender horned lowland form Irom Tomsk, Ekaterinburg, and otherplains localities he calls Capreolus pygargus leptocerus, and a heavyhomed upland form from Minusinsk he calls O. p. pachycerus. Thesesubspecies are based on hunters' frontals and antlers, such as arecollected and traded by natives throughout the southern Siberianmountains m great numbers, Kastschenko ^ states that no roedeerarc known on the Siberian plains, and he beheves that Mehely's hornsaro from unknown locahties. He states further:Thero are no roedeer near Tomsk, which Mehely repeatedly indicated as one sourceof his material; but by sportsmen and by commerce many horns are brought to Tomskfrom the Altai. Probably the horns were obtained by Count Zichy in Tomsk.Li an earlier paper, Kastschenko ^ has shown the immense variationexhibited by roedeer antlers irom the Altai. Our observations in. thetrading posts, and the series of antlers brought back by us, agreeWell with his remarks; and these names for races of roedeer, basedon antlers of unkno^^^l origm, may both be placed in the synonymyof pygargus. OVIS AMMON (Linnaeus).1758. Capi'a amnion Linnaeus, Syst. Nat., ed. 10, vol. 1, p. 70.1777. Ovisammon Erxleben, Syst. Regn. Anim., vol. 1, p. 250.After crossing the Chuisaya Steppe, the second day south fromKosh-Agatch, we saw weathered horns, and an occasional skull, of
1 Dritte asiatische Forschungsreise des Grafen Eugen Zichy, vol. 2, Zoolog. Ergeb. I, Saugethiere, 1901,p. 18, pi. 4.? Ann. Mus. Zool. AoLid. St.-P^tersbourg, vol. 7, 1902, pp. 294-296.3 Results Altai Zool. Exp., 1S9S, Toms'i, 1899, pp. 25-40, pis. 3, 4.
526 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 45.the wild sheep in the valley of the upper Chuya; and as we nearedthe Mongolian line these reUcs of the former occurrence of sheep, tothe north of their present distribution, became more common. Inthe bed of Tschornia Creek were numbers of old weathered horns andparts of skulls, and though sheep do not range in these hills at thepresent time, in summer at least, they must have commonly grazedhere not many years ago. In the border range, at the head of Tschor-nia Creek, Kain found signs of sheep which he beUeved had beenmade that same spring, and also much evidence that bands %vinter onthose slopes. Several heads and parts of carcasses, probably ofanimals destroyed by wolves, were comparatively recent kills. Dur-ing the summer, of late years, most of the sheep range on Mon-goHan territory. We were assured by aU the natives that no sheepare known in the country north of Kosh-Agatch. We saw numbersof skins in the trading post, all of which were said to come from theneighborhood of Tchegan-Burgazi Pass.Lyman collected four fine rams on the Mongolian slopes, in theTayliike Valley, not far from Tarkuta Pass, July 6 to 20. Thomas^has fixed the t3rpe-locality of Ovis ammon as the upper Irtisch River,Siberia, so our specimens may be assmned to be fairly typical.These four summer skins are almost free of the old pelage and arerenewing the coat. The hair is very short, mixed white and brown,and the general color varies greatly, from quite white to brownwith the rump patch not at aU conspicuous. Among the winter skinsseen in Kosh-Agatch, all said to come from Tchegan-Burgazi Pass,the range of color was equally great, and we purchased examples ofthe extremes. One is rather light brown with grayish neck andshoulders, and the underside of body pure white. The other is verydark blackish-brown, almost black, with whitish dorsal stripe,broadly pure white on withers and mixed with brown hairs on neckand lower back. All the winter skins show the large white rumppatch in shai-p contrast to the color of back and hips. The hornsof our fom* rams are good average size, by no means "record heads,"and exhibit the normal circumference and spread. The skulls andhorns measure as follows
:
Skulls of Ovis ammon. 175180
NO. 1990. MAMMALS FROM THE ALTAI MOUNTAINS?HOLLISTER. 527Heads of greater size may be liad if one is eager for extreme meas-urements. Two sportsmen, besides our party, visited the range ofthis sheep earlier in the season of 1912, and obtained large heads.The horns of one of these heads, which I afterwards examined in theestablishment of Rowland Ward, London, measured over 60 inches(1524 mm.) around the curve, and is, I was told, the largest sheephead received at Ward's in several years.The following extract, taken from Lyman's manuscript journal ofhis sheep hunt across the Mongohan border, begins with July 1, theday he left our camp in Tchegan-Burgazi Pass. It is interesting,not only as an account of this wonderful sheep, but as a descriptionof its local habitat and the methods of hunting.Having left the Black River [Tschomia Creek] and with it the last timber, we fol-lowed up the main stream [headwaters of the Chuya] to a point where it forks, and con-tinued up the left or east branch. At or near these forks I began to observe a goodnumber of sheep heads. It was difficult to guess their, age, for even on a live animalthe horns often have a cracked and weathered look. I remember but one head in thisdistrict where the lower jaw was with the skull. At no very great distance from theforks we turned up a smaller stream to the left and made camp ; 3h hours' march, whichat 2^ miles per hour would be about 8| miles. From this camp, which must be nearDemidoff's "Happy Valley," I hunted the afternoon of this day and the whole of thenext. The method of hunting is as follows: The guide, Yam-bai, the assistant guide,Mattai, and I, all ride. The country being generally rolling and without sharp ridgesit is necessary to keep one's eyes fixed on slopes at a great distance , but it is not necessaryto dismount often. When something is sighted, then it is necessary to examine witha good telescope, or when ridges occur one must go up on foot lest he should comesuddenly on game. It is extremely uncommon to sight game very close, and whenthis does occur the country is so open that a successful stalk is almost impossible. Theresult is that one rides nearly all the time, a method which seemed strange to me atfirst, but which is the correct thing I am persuaded if one has a far-sighted hunter.The whole matter rests with the keen sight of the hunter. If left to himself the averagegood American sportsman would loose at least half his chances because he could notsee the game before the game saw him.The country on which sheep are found is of much the same character on both theRussian and Mongolian side of the frontier. It is everywhere absolutely devoid oftrees and bushes and covered only with short grass. This grass gives place to shaleslopes from time to time, and often one side of the mountain will be washed andweathered, leaving an expanse of broken rocks exposed. Near the tops of many ofthe hills are often found rocky outcrops of fantastic shapes. The higher moimtainswhich seem to mark the boundary are completely washed, free from grass, their sidescovered with rock slides, and extremely rugged.The sheep are generally found on the open grass-covered slopes, pretty well towardthe tops. When in such a position it is obvious that they are very difficult to approach.It is first necessary to get into a position where one can get a good view with a tele-scope ; field glasses are of but little use, the distances are too great. It is then generallypossible, by making a sufficient detour, to ride to the foot of the particular hill onwhich the band of sheep were seen. Sometimes it is even possible to ride a good waytoward the desired position on the hill itself. The final stalk itself is always verydifficult because of lack of cover, except in those cases where the sheep are foundamong rocks or near a rock outcrop. The result of the stalk is either a long shot, 300yards or more, at animals at rest or undisturbed; or a short shot, 100 to 150 yards, atk
528 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 45.animals on the point of flight. If the sheep are on a smooth slope near the top onegets a shot of the second type by coming up hill from the other side. The momentone's head appears above the sky line the sheep are off. The fact that the hills areround topped and without a ridge makes the last moments of the stalk very uncom-fortable. To compare this kind of sheep hunting with the pursuit of the bighorn inNorth America one may say that the physical effort is far less in Mongolia than in NorthAmerica, but as a rule the final stalk is more difficult. The contrast between thenature of the ground in the two countries may be brought out by the fact that it isnearly always possible to bring horses directly up to the kill in Mongolia.To return to the narrative of the journey. On the first afternoon as we returned tocamp two rams came out on the sky line, but seeing us, departed. On the secondday going to the southwest saw a wolf at a great distance but nothing else. Weathervery cold with snow. The next day we broke camp and continued up the smallvalley and over a low and inconspicuous pass, direction east of south. The Mongolianline is probably near this pass; there is a great pile of stones but no stake. The countryahead is high and rolling; to the west are a range of rugged peaks. After crossing aheight of land, which I take to be the Bain-Chagan Pass, we proceeded down a valleyeast of south, and camped on a stream coming in from right, or west, side. March 4Jhours, say 11 miles. Saw two or three small sheep on hilltops after we had crossedthe pass. Our camp ground here had been much used by Kirghiz with their flocks.The next day I tried for sheep both on the west and east sides of camp, but saw none.There are a good number of heads in the valleys, on both sides, showing that ramshave been here, and recently.We continued the journey on July 5, and turned sharp to right around hills on whichI saw a gazelle. The view from these hills to the south and west is striking. Thecountry rapidly flattens out to the south and descends into a valley in which seemsto flow a stream. This is perhaps part of the Suok River system. On the other, south,side of this valley high hills, or mountains, rise and extend as far as the eye can see.To the west and south the valley is filled with large buttes of very characteristic andpronounced shapes. Looking down on this part of the valley, the view reminded meof a model of a great mountain system done on a small scale, or of a picture in a schoolgeography. Farther to the west and south, and a good way off, rose a range of highmountains, some completely covered with snow. Later in the morning we came tosome water, apparently a branch of the Suok system, and there saw at least 100 ewesand lambs, which came down off a hill and ran over the valley bottom. Saw too a smallram. Having crossed a ridge, we turned due west, into what I am told is the TayliikeValley. We proceeded west up this valley about 3 miles and camped. Total dis-tance since morning say 15 miles.The Tayliike Valley, in which we spent 16 days, runs at this point nearly eastand west. The valley bottom is perhaps 8,000 to 9,000 feet above sea. On the southside it is bordered by grass-covered hills, which roll with lessening height into abroad valley of a part of the Suok system. The hills on the north side of the Tay-liike Valley roll back with increasing height to the boundary range. They are oftenbroken on one side and have a good deal of rock outcrop in places. Sheep werefound on both sides of the valley. They showed a preference, however, for the northside. They were not plenty to the east of the place of our first camp, the hills beingstony and without much grass; and beyond a point 10 miles to the west they alsobecame scarce. Their range to the north is limited by the barrier mountains and to thesouth by the Suok Plain. The extent of the range is thus very limited, for the dis-tance from the Tayluke to the high barrier mountains is not more than 2 or 3 milesat the place where the sheep abound. On this range, however, rams are very numer-ous. Not a day passed that we did not see some. They go in bands of from 6 or 8to 30 or 40. I counted one band of 40 rams. The ewes and lambs seem to be in thelower country to the east. Besides one band seen on the march and already men-tioned, I saw but one lot of perhaps 30 on the east of the range.
NO. 1990. MAMMALS FROM THE ALTAI MOUNTAINS?HOLLISTER. 529An average adult ram will carry horns about 19 to 19^ inches at the butt, and 46 to48 inches round the curl. Sheep of this size are very common. A few larger headsare to be found in nearly every band; very large heads are of course rarer. I regretthat I took no body measiirements of sheep. It is safe to say that the animal hasa larger body than the average bighorn ram; but it is also certain that the increasein body is not in proportion to the increase in head. In the Ovis ammon the headis out of all proportion to the body. At a distance at this season (July) the ramsappear to be of two rather distinct colors?yellowish and black. Near to, the colordifference is not so noticeable. It is a characteristic perhaps due to the limitedextent of the range that the rams will return time after time to a particular mountaintop after they have been driven from it. When alarmed they never hesitate, butdash away until out of rifle shot. Their sight is keener than that of any animal Iknow. A black servant who accompanied an English sportsman into this regionthis summer told our men that he had been with the original Demidoff party in 1897,and that then the sheep were tame and comparatively easy to kill. An idea of thenumber of sheep may be got from the fact that in 15 days' hunting I shot four rams,wounded two, and missed three shots. The number of shots in proportion to thesheep seen was due to the great skill of my head hunter, Yambai, in stalking. It isto be noted that these rams are extremely tough. In my experience they will gofarther when mortally hurt than almost any animal I ever saw.The weather in the valley was bad. About July 8 there were three days when itwas unpleasantly warm at noontime; for the rest of the time it was impleasantly cold.There were few days of continual storm, but there was hardly a day that it did nothail, rain, or snow. The very sudden changes from hot to cold were very trying.If the altitude of the valley is taken at 8,500 feet, that of the hills on which the sheepare found may vary from 9,000 to 11,000 feet. On the hills, often near the tops, Ifound maral horns, but never a skull. The horns seemed very old. It is to be remem-bered that there is not a trace of timber in this whole region. Throughout the valleythere are signs that it is visited by natives with considerable flocks at some season,but we saw only one party of Kirghiz or Kalmuks, who had got lost trying to crossthe frontier, and another party of horsemen, seen at a distance. The number ofsheepskins on sale at Kosh-Agatch makes it probable that the natives kill a goodmany in winter, perhaps in^this valley bottom.I made only two camps in the valley, the last some 6 mUes west of the first. On^July 22 moved west of north. Three hours' march took us over a steep pass, theTarkuta, and down into the valley on the other side. On the north side of the passsheep heads suddenly ceased. It seems pretty certain that the animals in the Tay-liike Valley do not cross this divide. The hills in the valley where we camped aremore broken than those on the Mongolian side, being covered with rock slides onmany of their faces, but their tops are grassy and suitable for sheep. In fact, thereis nothing to account for the absence of sheep.CAPRA SIBnUCA FASCIATA Noack.1902. Capra fasciata Noack, Zool. Anz., vol. 25, p. 623.1905. Capra sihirica fasciata Trouessart, Cat. Mamm., Suppl., vol. 3, p. 738.The type-locality of this ibex is on the Bia Kiver, near LakeTeletzkoi, about 70 miles north of Kosh-Agatch. The animal isreported as common throughout the mountains in that vicinity, andnumbers of heads and hides are traded at Kosh-Agatch. We weretold of the presence of ibex in the mountains near Chibit, on the postroad between Kosh-Agatch and Ust-Inya ; and one of ourmen reportedseeing one in the mountains near our Tapucha camp, August 7. A80459??Proc.N.M.vol.4&?13 34
530 PROCEEDINGS OF THE NATIONAL MVSEVM. vol. 45.fine male was said to have been brought into Tapucha during thetime of our stay in the vicinity, but we failed to secure the specimen.I believe ibex are much more common and generally distributed inthe mountains between Altaiskoe and Kosh-Agatch than is generallyknown to the sportsmen who have visited the Altai. This formis probably confined to the wooded country north of the ChuisayaSteppe, on the Siberian side of the ranges, where it inhabits thenumerous timberline mountains. Kastschenko reports several speci-mens of the ibex from the Katoun Mountains and Buchtarme Kiver,southwest of Ongudai and Ust-Inya. In Kosh-Agatch we purchasedthe skull and horns of a fuie old male, killed on the Baskkaous River,which empties into Lake Teletzkoi. This specimen is virtually atopotype of fasciata. The horns are long, gracefully curved, andwide spreading, and the knobs are small and low. This skull, No.175188, U.S.N.M., measures: Condylobasal length, 270 mm.; greatestbreadth, 149; upper tooth row, alveoli, 66; length of horns overcurve, 1013 and 1045; circumference at basal knob, 240; distancebetween horn tips, 810.A pet ibex, a young male, which, in company with a young domesticgoat of about the same age, enjoyed the entire freedom of Kosh-Agatch, made friends with us at once on our arrival in town. It fol-lowed us into the buildmg given us as camp quarters, and was withconsiderable difficulty induced to leave, and then only after it hadinspected the whole apartments and bounced from the bunk to thetable, window sills, and on to the cement stove, or oven.CAPRA SIBIRICA HAGENBECKI Noack.1903. Capra sibirica var. hagenbechi Noack, Zool. Anz., vol. 26, p. 381.1907. Capra sibirica hagenbecki Lorenz, Denksckr. Kais. Acad. Wiss., vol. 80,p. 89.This ibex, originally described from some point near Kobdo,Mongolia, is common in favorable places on both the Siberian andMongolian slopes of the border ranges. Our two specimens, killedby Lyman on the Siberian side of Tarkuta Pass (about 35 miles westof Tchegan-Burgazi Pass), July 23 and 24, are both adult males, andagree in every particular with the description of TiagenhecJci. It is,indeed, likely that the original specimens, on which the form wasbased, came from much nearer our locality than the town of Kobdo,and the specimens may be assumed to be typical. The headwatersof the Suok River, one of the chief sources of the Kobdo, are not farfrom Tarkuta Pass.This is one of the forms without the light-colored "saddle." Ourskins, in the short summer coat, have the upper parts of uniform fawncolor, blending to almost pure white on the belly. There is a faintnarrow dorsal stripe from head to tail. Both specimens have thewell-developed callosities on the knee joints, as described of this race,
NO. 1990. MAMMALS FROM THE ALTAI MOUNTAIN8?H0LLISTER. 531and the horns are very heavily knobbed. The two skulls measure:Condylobasal length, 251, 250 mm.; greatest breadth, 145, 146; uppertooth row, alveoh, 71, 72; lower tooth row, 70, 73.5; length of hornsover curve, 558, 563; 506, 475; chcumference of horn over first basalknob, 260, 251; distance between horn tips, 310, 204.Lyman's notes contain only one reference to ibex in Mongoha. Inthe Tayliike Valley, not far from Tarkuta Pass, he ''frequently sawibex on the north side of the valley, among the rocks. They were allfemales and young, of a rather light yellowish color. They did notconfine themselves to broken ground, but often fed on the grassy hilltops." Later on, July 22, after crossing back into Siberia by theTarkuta Pass, he wrote:I hunted the mountains to the east and west for ibex. Saw no females, but on themountain to the west saw two bands of male ibex on two succeeding days, one eachday. The first lot consisted of 5 or 6, from which I shot one. Two or three ofthese carried horns much larger than the one shot. The second lot I could not countaccurately; there were 12 to 18, perhaps, and of these 6 were very large, with horns aalargo as the head afterwards purchased in Kosh-Agatch. Shot one from this lot. Ihad a very good look at the large ones as they ran off, and there is no doubt as to thesize of their horns. Both bands were feeding on grassy tops, not among rocks. Theseibex were not so farsighted as the sheep, and when shot into the bands ran a little wayand then stopped to look back before making off. In this they resemble our mountaingoat, though not stupid to the same degree. They carry shot remarkably; the secondone I killed was hit clean through the body, yet ran more than a mile. The meat ofthe first ibex was very good eating.PROCAPRA ALTAICA HoUister.1913. Procapra altaica Hollister, Smiths. Misc. Coll., vol. 60, No. 19, p. 1, Feb-ruary 8.We sighted four gazelles the first day south from Kosh-Agatch,while crossing the Chuisaya Steppe. They started at a good distanceand ran directly into a low sweepmg cloud of wind, rain, and sand,disappearmg as if by magic. The many skins seen in the tradingpost at Kosh-Agatch were said to come chiefly from the Mongolianside; and though formerly abundant on the Chuisaya Steppe, gazellesare by no means common there at the present day. We had one dayin the center of the steppe on our return journey, and I find the fol-lowing in Lyman's journal regarding the gazelles:Gazelles on the Kosh-Agatch steppe are scarce and wild. They keep to the openplain and do not get into the foothills, or at least I could find none in that type ofcountry. I saw one band at a great distance on oiu* way southward. On the retm-njourney I devoted five or six hoiu-s to hunting and saw but one lot of four males. Onthe plains it is impossible to get near them. I missed a shot at about 300 yards. Sawalso two lots of three or four females and young. My skillful hunter Yam-bai went outby himself for the whole afternoon, but could not get one. This is excellent proofthat they are rare and shy in this region. The character of the feed on the Kosh-Agatch steppe is much the same as that in the foothills near the Suok Plain, Mongolia,thin grass growing on stony, desert-like ground.
532 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.45.The type-specimen of this gazelle, which is nearly related only toProcapra gutturosa, was shot by Lyman at the point where the trailsouth from the Bain-Chagan Pass turns westerly, on the edge of theSuok Plain, Mongolia, July 5. In his notes on the game of the tripover the Mongolian border I find a few references to gazelles. Hementions seeing a female soon after crossing the frontier, south ofTchegan-Burgazi Pass ; four on the Suok Plain, July 5 ; and later inthe same day, well up among foothills, three or four bunches offive or six each. Later notes mention a few in the Tayliike Valley.EXPLANATION OF PLATES.Plate 37.Collecting stations in the Altai Mountains.Upper. Camp on the Chuisaya Steppe.Middle.' Camp on the Tchegan-Burgazi River.Lower. Camp in the forests of Pinus cemhra near Tapucha.Plate 38.Natives of the Altai and Habitat of Platycranius.Fig. 1. Kirghiz marmot hunter and horse.2. Camp on Tschomia Creek, showing rock slide inhabited by AUicola {Platy-cranius) strelzovi.3. Kalmuk horse owner. Plate 39.Scenes in the Altai Mountains.Upper. Yaks (Poiphagus grunniens) in Alpine meadow.Middle. Young male ibex (Capra sikirica fasdata) at Kosh-Agatch.Lower. The "last timber" to the south, on the border of the desert Altais. Larches{Larix sibirica) in valley of Tschomia Creek.Plate 40.Skull of Ovis ammon, U.S.N.M. Cat. No. 175180, male, from Altai Moimtains,Mongolia. (Greatly reduced).Fig. 1. Dorsal view.2. Ventral view. Plate 41.Skull of type-specimen of Procapra altaica, U.S.N.M. Cat. No. 175179, Suok Plains,Mongolia (one-fourth natural size).Fig. 1. Lateral view.2. Dorsal view. Plate 42.Skull and teeth of type-specimen of Procapra altaica, U.S.N.M. Cat. No. 175179,Suok Plains, Mongolia.Fig. 1. Left upper tooth row, natural size.2. Left lower tooth row, natural size.3. Skull, ventral view (one-fourth natural size).
U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 45 PL. 37
Collecting Stations in the Altai Mountains.For explanation of plate see page 532.

1_ "

U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 45 PL. 39
Scenes in the Altai Mountains.For explanation of plate see page 532.

U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 45 PL. 40

U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 45 PL. 41
Skull of Procapra altaica from Suok Plains.For explanation of plate see page 532.

U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 45 PL. 42
Skull and Teeth of Procapra altaica from Suok Plains.For explanation of plate see page 532.