THE AMERICAN NATURALIST [Vol. XLVII 



The parents themselves showed a correlation to the off- 

 spring as follows : 



The grandparents are much more closely correlated 

 with the pigs (.2962 ± .057). Two explanations may be 

 offered for this. Among the fifteen dams there were five 

 pairs of sisters, so that there were only ten grand-dams 

 to correlate. This would throw some of the litters into 

 one class that were separated when correlated with the 

 dams and so would modify the coefficient. An interpre- 

 tation more satisfactory to the writer is that the higher 

 coefficient between grand-dam and offspring represents 

 the segregation of some mammary character the limits of 

 which can not, at present, be defined. 



Relation of Asymmetrical to Symmetrical Pattern. — 

 An interesting point because of the constancy of the ratio 

 lies in the relation of the symmetrical and asymmetrical 

 patterns. In the pigs they occur in the ratio of 1:2, or 

 172 asymmetrical to 337 symmetrical. In the writer's 

 paper of last year, already cited, the same ratio held and 

 seemed unaffected by patterns in the parent. If divided 

 according to sex, we find no difference, the boars showing 

 89 asymmetrical to 174 symmetrical and the sows showing 

 83 of the first to 163 of the second. The fact that both 

 parents bear the asymmetrical pattern may influence the 

 per cent, of asymmetry in the pigs, but this influence is 

 very slight. "Where boar and sows were asymmetrical the 

 ratio was 50 : 85, while when the boar only was asymmet- 

 rical the ratio was 122 : 252. Putting them on the same 



