THE MULTIPLE ORIGIN OF HORSES AND PONIES. 443 



In all three xarictics (he mane is short and npfiii'ht in th(^ antnnni, 

 l)ut h)ni;' enonii'h in spring- to ni'ch to one side ol" the neck; in sunnner 

 the uppei- two-thirds of the (h)ek of the tail carries short hair, the 

 distal third lon<i- hair, which continues to ^row until it reaches the 

 ground: in winter the uppi^r two-thirds of the tail carries hair from 

 '2 to 4 inches in l(>n<>th. 'Vho hair of tlie body and limbs is short in 

 sunnner, but under the jaw and o\er the o-rcater })art of the body 

 and limbs it is from o to 4 inches in lent^'th in winter. 



The hoofs are narrow(M" and have loni>'ei- '' heels" than in the com- 

 mon horse, but, as in the common hoi'se, each limb is pi'ovided with 

 a chestnut and with an ei"i>ot, the Iiind chestnut (hock callosity) 

 being very long and narrow. 



In the variety (fig. 1, })1. i) occurring in (he Al(ai south of Kobdo, 

 probably the most primitive of the three, the liead is large and coarse, 

 and, compared with the length of the body, longer than in any do- 

 mestic breed. In a side view it is noticed that the forehead is promi- 

 nent (bumpy), th(^ lower part of the face straight or slightly con- 

 vex, the under li}) long, and that the head forms nearly a right angle 

 with the short neck. The eyes are lateral in position, and appear 

 unusually close to the ears owing to the great length of the space 

 between the eye and the nostril. The ears are long and generally 

 project obliquely outward (fig. 1, pi. i), as in many cart horses. The 

 croup is nearly level, but the hocks are usually bent and turned in. 

 Judging by the behavior during the last two years of the wnld horse 

 in my possession, I am inclined to think his less remote ancestors, 

 though in all probal)ility members of the steppe fauna, lived for a 

 time (perhaps during the Ice age) in the vicinit}' of forests. As is 

 the case with other gregarious animals, he strongly objects to be 

 separated from his companions, and he also objects to have his 

 movements circumscribed by fences. It has often been said '' nothing 

 jumps better than a cart colt." The cart colt jumps because he 

 has sprung from big-jointed, broad-hoofed, forest-haunting ances- 

 tors whose existence often dejiended on their being able at a bound 

 to clear fallen trees and othei" obstacles. The wild horse when shut 

 up in a loose box by himself is very restless, and keeps reai-ing u]) 

 against the door until set at liberty; if placed in a paddock away 

 from his special comrades he generally succeeds in either scraml)ling 

 over or breaking down the fence. 



The wild horse never encounters fences in (lu' (Jobi Desert, but, 

 probably because he had forest-bred ancestors which had often to 

 cross fallen trees, he endeavors without a moment's hesitation to 

 clear all obstacles that come in his way, while true desert forms 

 endeavor to break through them. 



It has been suggested that the wild horse of the Gobi Desert is 

 not a true wild animal, but onl}" a domesticated breed that has re- 



