46 



THE AMERICAN NATURALIST 



[Vol. LV 



not strictly comparable to those presented in Tables VII, 

 VIII and IX for the animals actually tested. However, 

 it is safe to conclude from a comparison of these tables 

 that the animals tested were a fairly random selection 

 from their respective families and experiments. 



It may be noted in passing that the ligures in Table 



TABLE XVI 



Families (2, 13, 32, 35 and 39) ANb Certain Crossbreeding Ex- 



■everse oi CA. B i 



85.2 74.2 I 88.4 



146 90.3 78.8 71.5 

 258 80.2 \ 84.7 



XVI give a good illustration of the statements made in 

 regard to the effects of crossing, with the exception of a 

 few records based on inadequate numbers. 



From this table it will be seen that family 35, the most 

 resistant to tuberculosis, held a rather low position 

 among the inbred families in most other respects. It 

 is actually the poorest in mortality among the young, 

 and fourth in size of litter and first in nothing. There 

 was thus no close relation between high resistance to 

 tuberculosis and vigor in other respects in 1919. 



However, the rank of a family in a single year is not 

 always a safe indication of its true position genetically 



