no 



CLAUDE W.. MITCHELL 



In the early generations of E, the saccate type was constant 

 and will be discussed later on. In the second generation of F, 

 however, a transition occurred which was followed by three 

 generations of the humped form. But this soon reverted to the 

 saccate type. This change had occurred in the earlier genera- 

 tions of a series and it seemed well to start other series to ascer- 

 tain whether the causal factor we were seeking rested in the 

 first generation as such. Hence the following experiment. 



TABLE 8 



History of Series E 

 S = saccate type; H = humped type; I = intermediate stage 



' One individual died. 



