THE PRACTICE OF BREEDING 295 



that they are descended from some famous sire or dam 

 ten or even twenty generations back. If Galton's law 

 of ancestral heredity is a fair estimate of the potential 

 strength of a breeding animal, we should expect that 

 the individual animal would inherit one-half from his 

 two immediate parents, one-fourth from his four grand- 

 parents, and one-eighth from his eight great-grandparents. 

 From one great-grandparent, therefore, we should expect 

 the inheritance in the descendant to be represented by 

 the fraction -^. It is evident that a single ancestor six 

 or eight generations removed would contribute a very 

 small fraction of the sum total of qualities of the individual. 

 In the application of Galton's principle, it is necessary 

 to assume that all the ancestors are equally prepotent. 

 This assumption is contrary to the experience of breeders. 

 Certain individuals are known to be more prepotent 

 in certain characters, or, more properly speaking, certain 

 characters are dominant and others recessive. The 

 dominant characters will largely determine the character 

 of the offspring. It is conceivable that a dominant char- 

 acter which it is greatly desired to perpetuate in a strain 

 might reappear in the offspring through many genera- 

 tions. It might be argued that because of this fact a 

 study of the characters of an ancestor far removed hav- 

 ing this character would be valuable. To this it must 

 be said that if the character is dominant it will be clearly 

 apparent in the immediate ancestors. If it is not apparent 

 it may have been lost or overshadowed by the develop- 

 ment of other characters, and if so it is good evidence that 

 the real character of the strain is being determined by 

 nearer and stronger ancestors. Every modern biological 

 conception gives added weight to the principle that it is 

 the recent ancestors that should be most carefully in- 



