726 



THE AMERICAN NATURALIST 



[Vol. L 



whole series of experiments, namely, that the first selec- 

 tion brought about an immediate and abrupt establish- 

 ment of two races with means 3.05 grades apart. The 

 greatest divergence between the two races due to a single 

 selection in all the following generations was 0.64 grade. 

 This followed the third selection. In the second genera- 

 tion there was a reduction of the average of the plus race. 

 Castle explains this as follows: 



To obtain larger numbers of offspring, several new pairs were added 

 to the experiment in this generation which did not appear in Table I 

 either as offspring or parents, but which were derived from the same 

 general stock as the parents of generation one (Castle and Phillips, :14, 

 p. 9). 



