EFFECTS OF INBREEDING AND CROSSBREEDING. 53 



cessive periods was too low to be certainly significant. The low cor- 

 relation in regard to litters per year (+0.25) is, as already noted, 

 largely due to the remarkable change from one extreme to the other 

 in Family 2, and the only slightly less remarkable change in Family 

 35. But in view of the likelihood that a real genetic change took 

 place in these families, through segregation in the early generations 

 and expansion of one line of descent at the expense of the others, the 

 changes in these families must not be weighed too heavily against the 

 absence of important changes in rank in 17 other families. Similarly, 

 in the case of the percentage raised of those born alive, and the per- 

 centage raised of all young, the changes in Families 35, 2, and 24 play 

 a large part in making the correlations low ( + 0.32, +0.36). 



If we attempt to arrange the families in order with respect to 

 general vigor, there would be little hesitation in picking out Family 13 

 as the best. It is the only family which was average or better than 

 average in every character in both periods. It was easily among the 

 best families in weight and fertility, and changed from medium to 

 good in the ability to raise the young. At the other extreme come 

 Families 1 and 15. Family 1 was among the poorest families in the 

 majority of characters in both periods. Family 15 was similarly 

 poor during the first period and it is not surprising that it was the 

 first family to become extinct, in spite of all efforts to maintain it, 

 and that it was followed to extinction by Family 1. 



Even in these families, however, we are not dealing merely with 

 differences in general vigor. Family 13 is relatively lower in its 

 ability to raise its young than in growth and fertility, and Families 

 1 and 15 each have a redeeming trait. In both the earlier and later 

 period slightly more than the average percentage of the young were 

 born alive in Family 1, an advantage lost through inability to rear 

 them successfully. The situation was reversed in Family 15, which 

 lost a larger percentage of the young at birth than any other family, 

 but was well above the average in the percentage raised of those born 

 alive. 



When we consider the remaining families, the impossibility of 

 ranking them in general vigor becomes at once apparent. Family 38 

 would be placed second to Family 13 in general vigor but for the fact 

 that it produced litters less frequently than any other family. Owing 

 probably to this defect, it was always a small family. Families 11 

 and 9 are two similar families which have a remarkable combination 

 of vigor and weakness in different characters. They were among 

 the best three families in both size of litter and weight, yet both of 

 them produced litters irregularly and were unsuccessful in raising the 

 young. The contrast was especially marked in Family 11, which led 

 all of the families during the second period in size of litter, gain, and 

 weight at 33 days, but was one of the three poorest families in the 



