THE SELECTION PROBLEM IN ANIMAL BREEDING 



497 



TABLE LXIV. THE TEN GREATEST PRODUCERS OF SPEED UP TO AND INCLUDING 1901 



(After E. Davenport) 



a few generations, therefore, this famous family of racing horses has 

 produced a remarkable series of performers, horses which have been 

 able to trot or pace a mile within 2 : 30. There seems to be little question, 

 therefore, that this family of fast horses had its foundation in the care- 

 ful fostering of the favorable genotypic material of Hambletonian 10; 

 and a transmission of it through a relatively small number of exceptional 

 sires which may have possessed a genotypic arrangement somewhat 

 superior to that of Hambletonian 10, as the record of Geo. Wilkes 519 

 in particular might indicate. Davenport has made a very careful study 

 of the records in the Register and Yearbook, a study which should be 

 continued and extended. Without considering in any detail the ex- 

 tensive data which have been collected, it appears fairly certain that 

 selection in the improvement of trotting and pacing horses has operated 

 by detecting and multiplying the most favorable genotypes; and that 

 training, in so far as it has had influence, has served as a means of developing 

 inborn potentialities to the full, and, therefore, of detecting most favorable 

 lines of descent. 



Fecundity in Fowls. Pearl's investigations on the inheritance 

 of fecundity in fowls have already been touched upon, but they deserve 

 more extended treatment at this point, for in them the relative effective- 

 ness of phenotypic and genotypic selection is strikingly contrasted. 

 For if performance has anything to do with development of more favorable 

 hereditary material, or if selection has a creative effect in a given direc- 

 tion, then it would appear to be a conclusion unavoidable that mass 

 selection must result in increased average winter egg production. Yet 

 as a matter of fact, as shown graphically in Fig. 185 there was actually 

 a slight decrease in average winter egg production during a 9-year period 

 of such selection. 



That this selection was rigid and a fair demonstration of the ineffect- 



32 



