498 GENETICS IN RELATION TO AGRICULTURE 



iveness of purely phenotypic selection for fecundity in the Barred 

 Plymouth Rock is indicated by the plan which was followed during 

 this portion of the investigations. Only pullets were used for breeding 

 stock which had laid 150 or more eggs during their first laying year, 

 and cockerels were selected from among the progeny of 200 egg hens. 

 The type of selection practised was, however, strictly mass selection, for 

 the selected birds were bred together without respect to genetic relation- 

 ship, and no tests were made of the laying capacities of progenies from 

 particular matings. This last point is of particular importance, because 

 it definitely distinguishes the method of breeding used as one typically 

 of mass selection. 



Obviously the reason for the ineffectiveness of selection during this 

 period of mass selection lies in the fact that modifiability in fecundity 

 is very great. This particular fact has been discussed fully in a pre- 

 ceding chapter, but here it must be considered again as the reason for 

 the fact that this system of selection failed to result in improvement in 

 egg-laying capacity, for were performance and genotypic constitution 

 closely correlated, then this system of mass selection should have been 

 effective. But as a matter of fact the criterion of selection used in this 

 portion of the investigations, namely total yearly egg production, was 

 evidently not a good index of genotypic constitution, for apparently 

 it failed to distiguish between individuals belonging to a number of 

 intergrading genotypes. Consequently, whenever, by chance a female 

 was selected which by phenotypic variation represented the upper 

 limits of her genotypic class, the population was thereby thrown back 

 by that much to the level representing the mean phenotypic performance 

 of her particular genotypic class. A wide range of modifiability for each 

 genotype, therefore, continually held the average yearly production 

 down to the original value for the population. 



But beginning with the year 1908 a radical change was made in the 

 method of selection. During the first portion of the second period, 

 the object was merely to ascertain the actual mode of inheritance of 

 fecundity, a subject which is discussed more fully elsewhere; but during 

 the second portion, from 1912 to the present time, selection was only 

 carried out for high egg production. Essentially, however, the mode of 

 selection during these two portions of the second period was the same so 

 that we may consider this as a single period. The performance index 

 during this period was winter egg production rather than total egg 

 production. But in the selection of high winter producers for breeding 

 purposes, a progeny performance test was employed as well as an actual 

 individual performance test. Every female which was selected during 

 this period came from a high-producing mother, the female progeny of 

 which were all high producers. In case such a female failed to give 



