514 



THE AMERICAN NATURALIST 



[Vol. LII 



safe in assuming that these were of the same line. From 

 one of them the sixth (?) and succeeding generations of 

 Table VI were obtained. 



Attention is directed in Tables IV, V, and VI to the 

 following points: 



1. The wingless viviparous females, more abundant 

 early in the cycle, are gradually replaced by winged 

 females. This is especially clear in Tables IV and V. It 

 is obscured in Table VI by the fact that this line is not 

 really a continuous one. A catastrophe in the fifth gen- 

 eration made it necessary to resume this line by means 

 of a female from the same stock. Up to the fifth genera- 

 tion there is an irregular increase in the proportion of 

 winged females, which reaches its climax in the fourth 



