94 Journal of Agricultural Research voi. x. no. 2 



The method of selecting by family perfonnance is a natural corollary 

 of the principle that offspring do not "inherit from their parents" but 

 that offspring are derived from some of the same kind of germinal stuff 

 as that from which those parents and their brothers and sisters were 

 derived. And the best knowledge of the varied qualities of the germ 

 plasm is obtained by a comprehensive view of its performance in a, num- 

 ber of closely related, fully developed somas. The studies of a score of 

 geneticists (among whom it may not be invidious to mention Pearl, 

 working with the fecundity of poultry) are in agreement upon this point. 



Since most of the ewe lambs, but, on the other hand, only a very few 

 ram lambs, are preserved as breeders, the selection of males is the more 

 rigorous. We may illustrate the general method of selection by an ex- 

 ample. In the 1 91 6 selection the available ram lambs belonged to 12 

 "families." A "family" comprised brothers and sisters and the two 

 parents. In selecting the trait "body weight" the average weight of all 

 the members of each family group at a fixed age is calculated. The 

 family having the highest average weight is graded i ; the next highest 

 average is 2 ; and so on. If the average is the same in two families, they 

 receive the same grading rank; thus, in one selection two males grade 

 No. 4 in body weight and four grade No. 2 in shoulder height. Naturally, 

 one family will grade high in body weight, but low in weight of fleece and 

 perhaps will be medium in ratio of head width to head length. The 

 rank of every family with respect to every quantitative trait is thus 

 determined; the rank is multiplied by its appropriate weight factor as 

 in ordinary scoring. The family which gives the lowest sum of products 

 grades highest, and the best ram from that family (or the better if there 

 be two) is ordinarily chosen. However, if the individual belonging to 

 the "best" family is sickly or has any physiological quality that would 

 interfere with its success as a breeder, the male from the next higher 

 family may be preferred — that is, selection is made primarily on the basis 

 of family performance, but the somatic insufficiency of the individual is 

 permitted to veto the choice based on amily alone. To facilitate such 

 veto, the individual males from which selection is to be made are graded 

 on the basis of their quantitative traits. It is practically found that the 

 best individuals usually come from the families that stand high in the 

 scale. If the representative of the best family should be a ram grading 

 at the bottom of the scale individually, he would probably be rejected 

 and the representative of the second-best family selected. But prac- 

 tically, as stated, the question of relative weight to be given to the in- 

 dividual (as contrasted with family) does not cause much hesitation so 

 long as the principle of selecting permanently on the basis of family is 

 kept in view. 



To illustrate the foregoing the procedure in a particular case is given 

 in Tables I and IL 



