330 SKIN, COLOUR, AND HAIR. 



are bred under artificial conditions. As a rule, these mark- 

 ings take the form of a dark stripe down the back (dorsal 

 stripe, Fig. 398), which is often seen in bay horses ; 

 horizontal stripes about the knees and hocks (Figs. 398 

 and 399) ; an indistinctly defined dark patch on the side 

 of the withers (Figs. 398 and 400) ; and indistinct stripes 

 across the neck (Figs. 398 and 400). On rare occasions, 

 stripes appear on the forehead (Fig. 401). As these stripes 

 are generally plainer in the foetus and foal than in the 

 mature animal, we have von Baer's law (p. 326) to support 

 the assumption that the wild ancestors of the horse were 

 striped animals, though not to the same extent as zebras 

 (P- ^57)- I^ zebras, the markings, as we have seen 

 (p. 327), are particularly useful in protective coloration ; 

 but the markings in the Nubian wild ass (Fig. 402) and 

 in the Somaliland wild ass (Fig. 615), for instance, are too 

 insignificant to take any prominent part in this important 

 function. If we turn to mankind, in order to obtain in- 

 formation respecting the habits of the lower animals, we 

 shall find that "like" always seeks to associate with 

 '' Uke," and in that endeavour, adopts pecuhar marks 

 for recognition. Thus, in Southern India, Hindoos put 

 on their foreheads special coloured stripes to indicate the 

 particular caste to which they belong ; and in England, 

 the respective apparel of society people and coster- 

 mongers is equally distinctive. This desire for class 

 segregation is evidently an inherited instinct, which no 

 doubt was formerly more useful for obtaining mutual 

 help against enemies than at present, and is an important 

 factor in the origin of species ; the recognition marks 

 having been furnished by variation. We also know that 

 almost any herd of wild animals will evince suspicion, 

 enmity, or alarm on the first approach of even one of 

 their own species which is not garbed like unto them- 

 selves. I think we are therefore justified in assuming 

 that the comparatively insignificant stripes of the Nubian 

 wild ass and of the Somaliland wild ass are merely recog- 



