THE MULTIPLE ORIGIN OF HORSES AND PONIES. 451 



sqiiinvl. (lu* l):ulir(>i\ pine iiuirtin, niid wild boar, (lir stajr, rot'decr, 

 unis, horse, and other recoj^nizod iiuMubors of a triio forest fauna. 



In the case of the P2qui(hv it is often extremely dinicult to deternrmc 

 to which species any given bones belong, and hence it is impossible to 

 state definitely that the horses found along with the hamsters and 

 other steppe forms essentially diii'ered from those Avhich were con- 

 temporaries of the stag and wild boar and other typical forest forms. 

 It may. however, be assumed that even in post-Glacial times the 

 majority of the inhabitants of the steppes would when mature be 

 quite or nearly whole colored, while frequenters of the forests would 

 as often be either striped or spotted; that, e. g., the horse which fre- 

 quented the Khine valley along with the kiang and woolly rhinoceros 

 would resemble the wild horse {E. c. prjevalskii) which, with the 

 kiang, now lives in the vicinity of the Great Altai Mountains, while 

 the horse which at a subsequent period was a contemporary of the 

 Avild boar, stag, and roedeer would be more or less richly striped, and 

 in its limbs and general conformation adapted for a life in or near 

 forests. 



That there is some ground for this assumption will, I think, be 

 admitted when due consideration is given to results obtained by cross- 

 ing various kinds of horses Avith a Burchell zebra. ^Y[len ponies of 

 the Celtic type — i. e., ponies which in their color are identical with 

 Prjevalsky's horse, almost certainly the lineal descendant of the 

 steppe horse of PahTolithic times— are crossed with a male Burchell 

 zebra, hybrids are obtained which, while in build strongly suggesting 

 a Burchell zebra, are as profusely striped as the great zebra of Soma- 

 liland, and have at least i\\c times as many transverse stripes across 

 the trunk as occur in their zebra sire. When, hoAvever, pony mares of 

 the Norwegian type are crossed with a Burchell zebra the hybrids 

 resemble in make their Norse dams, and in their markings closely 

 approximate the common or mountain zebra. The explanation of 

 these remarkable differences seems to be that in the case of the Celtic 

 pony hybrids the remote (Grevy like) ancestors of the Burchell zebra 

 control the development and determine the plan of the decoration, 

 while in the case of the Norse pony hybi'ids the remote striped-horse 

 ancestors contribute the more obvious characters— the Norse ponies 

 having more influence in determining the plan of striping than the 

 highly specialized Celtic ponies, from which stripes had probably all 

 l)ut completely disappeared countless generations before they began 

 to fade on the horses which belonged to the forest fauna. 



It is probable that the highly si)ecialized Celtic pony, as well as 

 the primitive Gol)i wild horse, belong to the steppe fauna, and it is 

 equally probable that the yellow-dun (Fjord) horse, in which a 

 striped coat may be said to be latent, belongs to the forest fauna. If 

 this be admitted, it follows that the environment of the Norse race has 



