SEX-RATIO IN DROSOPHILA AMPELOPHILA. 365 



the opposite direction the "male" strain. This terminology 

 will be used throughout this discussion. As the starting point 

 of his female strain, he selected from nature a pair of flies which 

 produced 52 males and 135 females and in his male strain he 

 started with a pair which produced 84 males and 75 females. 



Considering first the effect of selection upon his female strain 

 we find here no conclusive evidence of any progressive change in 

 the direction of the selection. (His data for this strain are given 

 in Table XI.) To be sure, there is a slight increase in the 

 relative number of females in his second and third generations 

 over his first generation, but this fluctuation could easily be 

 attributed to the small number of individuals examined. The 

 greatest difficulty in assuming the effectiveness of selection here, 

 is the fact that his last two generations are the ones in which he 

 obtained the lowest relative number of females. Moenkhaus 

 explains these low ratios of the last two generations in that he 

 possibly made a poor selection in the preceding generation, but 

 if we consider selection as having a cumulative effect (as Moenk- 

 haus seems to consider it) it is difficult to see how one could lose 

 by one poor selection all that he had accomplished in the previous 

 selections! It must be admitted that by using as parents of 

 each succeeding generation, pairs which threw the most extreme 

 ratios in favor of the females, Moenkhaus was able to maintain 

 a strain which on an average gave a relatively high number of 

 females. But this was probably not due to any cumulative 

 effect of selection but to the isolation of a peculiar type of female 

 which will be discussed more fully later. 



As to the male strain (Table XII)., Moenkhaus admits that 

 the effect of his selection here has been slight and after examining 

 his data carefully it is difficult to see how it could be assumed 

 that there has been even a slight effect of selection. He started 

 with a ratio (84 males to 75 females) which was only slightly 

 different from normal and this slight difference was not trans- 

 mitted to the first generation nor any succeeding one. So it 

 seems that he had here a normal strain which would have given 

 the same ratio regardless of the direction of the selection. 



The female strain is the unusual one and is in need of an ex- 

 planation. It throws some very ^exceptional ratios in favor of 



