INHERITANCE OF COLOR IN PERCHERON HORSES. 267 



registration must be considered as a factor. Thus the 179 indi- 

 viduals are not in all probability a representative population, 

 owing to the elimination of grays from the register. 



It should be added that this is the surface explanation of the 

 facts, which might be accounted for partly at least in other ways, 

 as by an actual prepotency of the black color, or by a melanic 

 tendency due to effect of change of climate on the breed on re- 

 moval from its original home in France to America. All of the 

 offspring and parents and almost all of the grandparents are 

 American registered animals. It is chiefly in the great-grand- 

 parental generation that the French records have been drawn 

 upon. 



The following description of the Percheron breed is taken from 

 a work on "The Horse" by Roberts (Macmillan, 1905): 

 "About 1820 two noted gray Oriental stallions, Godolphin and 

 Gallipoli, were introduced into the Government stables at Pin. 

 These two prepotent stallions fixed the style of color and fast- 

 ened it on an already susceptible breed (p. 159). . . . The color 

 of most Percherons is gray of varied shades. Sometimes it is 

 quite light, becoming nearly pure white in old age. Again, the 

 striking light and dark dapples are seen, and dark grays, almost 

 black, with a few white hairs. Comparatively few blacks have, 

 as yet, been bred, although dark colors are sought and are more 

 common than formerly. The American purchaser prefers darker 

 rather than lighter colors ; hence the effort in France is to pro- 

 duce darker colored animals than formerly. ... It will take 

 many generations to entirely eliminate the light colors, so long 

 one of the characteristics of the breed ; but this will be accom- 

 plished in time if Americans persist in preferring dark- rather 

 than light-colored draft-horses. This preference is not founded 

 on a fad, for, other things being equal, dark-colored horses are to 

 be preferred to light-colored ones" (p. 162 ff). 



From all that we know of the history of the breed it is quite 

 obvious that the increase in black and decrease in gray through 

 four generations, shown in Table I., is at least partly a result of 

 breeders' selection, and would not warrant the assumption of a 

 melanic tendency or prepotency of black. Indeed, it is quite 

 equally obvious that we should expect the long predominant 

 color, gray, to be the prepotent color. Independent data have 



