470 GENETICS IN RELATION TO AGRICULTURE 



served facts is in the relation of bay, black, and chestnut to one another. 

 Wilson's formulation does not account for the production of bay foals from 

 matings of chestnut and black, although such matings produce a large 

 proportion of bay foals, whereas the first formulation accounts for them 

 very simply. Moreover the first analysis is more nearly in harmony with 

 our knowledge of the inheritance of coat color in rodents, which is so 

 well understood that there is no question of this kind as to factor relations. 

 Wilson's formulation, however, is of this service: it points out clearly 

 how uncertain are analyses based on herd book records, and thereby at 

 the same time indicates the need of actual experimental investigation. 

 It would be a very simple matter to demonstrate experimentally which 

 of these analyses accounts for the actual factor interrelations. 



We cannot refrain here from indicating some of the consequences in 

 breeding practice of a formulation such as the one we have favored in 

 this discussion. The really important feature of the analysis, of course, 

 lies in the emphasis it gives to the definiteness of phenomena of coat color 

 inheritance in the horse. In every case there is a definite reason why a 

 horse should be of a certain color, and the reason is comparatively 

 simple. Moreover, since the phenomena are so definitely predetermined, 

 it is possible within certain limits to control them. 



To take a definite instance, the government has set itself the task at 

 the Iowa station of creating a gray breed of draft horses. Since gray is 

 dominant to all the common horse colors save roan, it is impossible to get 

 gray from matings of other colors. Moreover, grays when mated to- 

 gether produce gray, bay, brown, black, and chestnut foals, according 

 to the particular gray genotypes which are involved. Grays of the 

 genetic constitution HhBbGg mated inter se produce the entire series of 

 colors in the ratio 



48 gray: 9 bay and brown: 3 black: 4 chestnut. 



The bay, brown, black, and chestnut offspring of such matings, or 

 any other for that matter, might be mated together ever so often, yet 

 they would never produce gray foals, although themselves the offspring 

 of gray horses. 



Accordingly the method which should be followed in establishing a 

 gray breed of draft horses is not difficult to map out. An effort should be 

 made to get homozygous gray horses for breeding stock. Such horses, 

 of course, will be met with only in breeds in which there is no prejudice 

 against gray, as for example in the Percheron breed; but, if it should be 

 thought desirable to utilize some of the good qualities of other breeds in 

 which gray is not a favored color that may be done at the sacrifice of 

 uniformity of color in the first generations. Matings should, however, 

 always be of gray to gray individuals, and all animals of other colors should 



