562 



The Journal of Heredity 



esis of independent factors instead of 

 complete linkage. There are first cer- 

 tain considerations based merely on 

 appearances. Wilson himself grants 

 that roan is inherited independ- 

 ently of the others, on the ground that 

 the roan pattern obviously can be 

 combined with any of them in effect. 

 Bay, blue and strawberry roans are all 

 common, and Wilson mentions roans 

 which later turn gray. Doubtless he 

 would consider piebald as due to an 

 independent factor for a similar reason. 



But it is also well known that grays 

 may be any color at birth, and also in 

 duns, varieties are recognized which 

 appear to be the dilute forms of bays, 

 blacks and chestnuts respectively. 

 These are the dark-maned duns, the 

 mouse-colored horses and the creams. 

 Exen among chestnuts, differences homo- 

 logous to that between bay and black 

 may be visible in the red and liver- 

 colored chestnuts, though this does not 

 appear to have been tested genetically. 



The genetic evidence for the theory 

 of independence is thoroughly satisfac- 

 tory in regard to the factor which 

 changes bay to black (A,a), and that 

 by which chestnut, differs from either 

 (B,b). In breeds in which bay is more 

 common than black more chestnuts 

 should possess factor A than a and 

 chestnut by black, both recessive to 

 bay, should yet produce many bays, 

 each parent supplying the dominant 

 factor which the other lacks. This turns 

 out to be the case. Thus Anderson 

 found 111 chestnuts, 83 blacks, 20 

 browns (generally equivalent to dark 

 bay) and 124 bays from chestnut by 

 black among Kentucky saddle horses. 

 In these quite extensive results 42% 

 of the offspring are inexplicable on Wil- 

 son's theory under which neither blacks 

 nor chestnuts could ever possess the 

 dominant bay factor. 



In the case of gray or dun it is more 

 difficult to get conclusive evidence, as it 

 is necessary to find individual gray 

 or dun sires which showed themselves 

 to be double heterozygotes in ])articular 

 kinds of matings. Gray and dun are not 

 among the commonest colors in most 



breeds and stud books, moreover, are 

 known to be none too accurate in regard 

 to colors, so that individual cases must 

 be used with caution. Wilson'*' gives 

 the record of twenty gra}'' Thorough- 

 breds which had at least one chestnut 

 foal from a chestnut mare and were 

 therefore GC on his hypothesis. These 

 stallions produced from chestnut mares 

 94 grays, 15 bays, 1 black and 61 chest- 

 nuts. The presence of so many bays 

 and one black can hardly be ascribed 

 to errors although it must be admitted 

 that they are not as numerous as one 

 would expect on the hypothesis of in- 

 dependence. The sire which seems 

 most certainh^ to be a double hetero- 

 zygote, GgBb, is Gray Frairs, with 15 

 grays, 3 bays and 4 chestnuts from 

 chestnut dams. Possibly partial link- 

 age between G and H may exist. In 

 this case, most grays which produce 

 chestnuts at all would be of formula 

 GBgb and so produce an excess of 

 chestnuts. At any rate it seems clear 

 that gray can transmit both chestnut 

 and bay. 



GRAY, ROAN AND PIEBALD 



Turning to the list of factors, it may 

 be said that the factor for gray has been 

 certainly demonstrated. A number of 

 gray stallions are known which are 

 homozygous, producing only grays from 

 mares of all colors. Most grays are 

 naturally heterozygous and produce 

 very close to the expected 50% gray, 

 50% of other colors, when mated with 

 other colors. Factor G has been placed 

 in class la i as a dominant factor which 

 inhibits color regardless of its quality, 

 but its mode of action is peculiar. 

 Grays generally show little or no white 

 when born, but gradually become whiter 

 as they grow older, white ai)pcaring 

 early especially on the head. 



The dominant factor for roan has 

 also been clearly demonstrated, both 

 homozygotes and heterozygotes being 

 known. It is more typical of class lai 

 than is the factor for gray. The mixing 

 of white hairs with colored ones occurs 

 regardless of the color of the latter. 

 The head and lower limbs arc usually 



'"Wilson, j. l')16. Loc. cil. 



