782 



PROCEEDINGS OP THE AMERICAN ACADEMY. 



three units in the X 1 culture, which was not inbred beyond a single 

 generation. This difference probably represents, in part at least, an 

 effect of inbreeding, which shows itself between the sixth and the thirty- 

 first generations. Compare the variation in spines of XI, A 6- (7 6, 

 J/31, and A 61, all of which were reared under optimum conditions. 

 Number of spines is correlated with tibial length, as we shall see, and 



TABLE XXV. 



Correlation between Number of Spines in Sex-combs and Length 

 of Tibia, Series A, Generation 61. 



x = spines (sum of both legs). 



^ = 10.55 

 A y = 38.72 



r = .4687 J- .0526 



(ri = 1.8574 

 o-y = 2.031 



the number of spines in A Q-C 6 was slightly higher than in A 61. It 

 is fair to suppose that the tibial length also was slightly greater, and so 

 the size. These differences may have been due simply to differences in 

 nutrition, but the fact that X 1, which was reared under the same condi- 

 tions as A 61, yet exceeds it in number of spines and tibial length, 

 suggests that A 6 may exceed A 61 for the same reason, that is, because 

 less inbred. 



