THE PRJEVALSKY HORSE. 57 



animal — are given the approximate limits of its distribution, 

 from the Eiver Atschik-Su eastward to Sydshira. Not far from 

 this district, a considerable number of Prjevalsky horses were 

 captured by D. N. Clemenz. No. 3,089 came from the Dzungarian 

 Gobi, behind the Baityg-Bogdo Mountains, and Nos. 3,087 and 

 3,088 are likewise from Dzungarian Gobi (between Nvirsu and 

 Seligendze). Two foals, received from D. N. Clemenz, were killed 

 in the Gobi Desert near the Ebi well, in the neighbourhood of the 

 road leading from Kobdo to Barkul (Nos. 3,072 and 3,073). A 

 stallion (No. 3,094), obtained by the Roborovsky-Kozlov Expedi- 

 tion, was killed near Lake Gutschen. Definite statements regard- 

 ing the other horses are wanting : these being the specimens added 

 to the Museum by N. M. Prjevalsky (No. 1,523), Schischmareff 

 (No. 3,090, from the Kobdo district) and D. M. Clemenz (No. 

 3,071, from the River Bulunga). 



In regard to living specimens, there are only the detached 

 accounts of the animals in the Moscow Zoological Gardens. These 

 were presented to the Gardens by Assanoff and came from the 

 Kobdo district of the Otschiginsk part of the Gobi Desert (48° N., 

 90° 35' E.). The horses now at Tsarskoe-Selo, as I gather from a 

 letter of Prince A. A. Oochtomsky, were obtained from the herd of 

 the Van's of Torgoutsk in Western Mongolia. 



From the above it is evident that the distribution of E. jyrjeval- 

 skii, so far as the present data go, is fairly limited. Taking into 

 consideration that the Prjevalsky horse is found northwards as far 

 as the River Urungu and inhabits the Kobdo Steppe ; and that 

 M^estward its distribution is limited by the Manasu, close by meri- 

 dian 84° (as evidenced by the statement of Pewtzoff) ; it may be 

 concluded that its area of distribution is confined to a rectangular 

 disti'ict, bounded on the north by 48°, on the south by 46°, on 

 the west by 84° E. and on the east by 90°-91° E. 



Of the habits of E. lorjevahhii comparatively little is known. 

 The brothers Grum-Grjimailo, the only travellers who have had an 

 opportunity of observing the horses in the wild state, give the 

 following account of them : " The wild horse is an inhabitant of 

 the level districts, and goes at night to the pasture-lands and 

 drinking-places. At break of day he returns to the desert, where 

 he rests until sunset. In the Spring, when there are foals in the 

 herd, the animals always rest in the same place. The fact that 

 I found a place of about 1,609 metres square, covered with a 

 thick layer of foals' dung, affords proof of this habit. From the 

 almost entire absence of coarser dung, it may be assumed that 

 the resting-place of the herd is changed as soon as the foals grow 

 up ; whereas it is constantly the same so long as the foals are small. 



