THE AMERICAN NATURALIST [Vol. LV 



mission of resistance from each parent to offspring of 

 each sex. The total figures — 62.6 days where family 35 

 is on the sire's side, 62.4 where on the dam's side — indi- 

 cate that there is probably eqnal inheritance from each 

 sex. 



Degeee of Determination of Resistance by Heeedity 

 We have shown that the length of life after inoculation 

 was determined less than 2 per cent, by weight, age and 

 rate of gain combined, in lots 7-9 and less than 7 per 

 cent, by these factors in lots 10-14. It is interesting to 

 compare the degree of determination by heredity with 

 these figures. Such a comparison can be made for the 

 crossbreds by finding the correlation between length of 

 life and the amount of blood of family 35. The correla- 

 tion in lots 3-5 comes out + .319 ± .082, in lots 7-9, 

 + .539 ± .069 and in lots 10-14 + .572 + .048. For rea- 

 sons w^hich have been mentioned the latter two groups 

 are much more satisfactory than the group of three mis- 

 cellaneous lots, 3, 4 and 5. The average in lots 7-9 and 

 10-14 is +.560 ±.039. The square of this coefficient, 

 .314 indicates that over 30 per cent, of the variation in 

 length of life is caused by heredity, neglecting such 

 heredity as may be due to differences among families 2, 

 13, 32 and 39. 



If we assume that some 10 per cent, of the variation 

 is due to differences in condition, weight and age and 

 over 30 per cent, to heredity, we still have over 50 per 

 cent, of the variation due to unknown causes. From 

 the nature of the case, however, a large amount of acci- 

 dental variation is to be expected. 



Relations between Resistance and Other / 

 Characteristics 

 It is clear that family 35 is markedly more resistant 

 to tuberculosis than the other inbred families, and that 

 among these, family 2 is somewhat more resistant than 

 families 13, 32 and probably 39. The question arises 



