326 SKIN, COLOUR, AND HAIR. 



we have chestnut Suffolk Punches, Cleveland Bays, 

 brown Exmoors, and dun Norwegians and Connemaras, 

 as distinct breeds. It is generally believed that a foal 

 from a chestnut dam by a" chestnut sire, is always of a 

 chestnut colour ; but variation plays such an important 

 part in heredity, that we cannot implicitly accept this 

 statement, although it is undoubtedly correct in the vast 

 majority of cases. Grey and white have but shght per- 

 sistency, as we see by the fact that not unusually a bay, 

 brown, or chestnut foal is produced by a dam and sire 

 which are grey or white ; and a grey or white foal 

 hardly ever appears except when one or both of its 

 parents are grey or white. In accepting von Baer's law, 

 that the history of an individual is the history of the 

 species, we have a strong proof that the wild ancestors 

 of the horse were neither grey nor white ; because all 

 grey and white horses are much darker in colour before 

 birth or when foaled, than subsequently. In fact, we all 

 know that every grey horse gets whiter with age. Mr. 

 Wilfrid Scawen Blunt, who is probably the best English 

 authority on Arab horses, tells me that although Arabian 

 breeders, from religious motives, like their stock to be 

 grey or white; the majority of their horses are bays or 

 chestnuts. Black appears to be a much less persistent 

 hereditary colour than chestnut or ba}^, because in the 

 produce of black dams and black sires, red pigment has 

 a strong tendency to manifest its presence. As a 

 characteristic of a species has a far greater hereditary 

 persistency than a Variation, and as dun is a characteristic 

 of many old, semi-wild breeds, I think we are justified in 

 saying that our horses' wild ancestors were probably of 

 a dun colour, which domestication has enriched into bay. 

 Before horses were domesticated, they required, in the 

 struggle for existence, the aid of colour as a means of 

 concealment from their enemies, and of recognition by 

 their fellows. Protective coloration, to be efficient, must 

 of course be suitable to the special surroundings. We 



