244 



CLAUDE W. MITCHELL 



TABLE 8 

 Starvation at birth of high potential females from Series D 



we contrast the results of starvation immediately following birth 

 as in this experiment with the results of starvation after the 

 initial ovarian development we get the contrast of 76 per cent to 

 30 per cent. 



These results at least confirm, if not absolutely prove, the 

 conclusions before reached, that male production follows upon 

 the summation of favorable external and internal conditions, 

 plus a sudden interruption by a nutritive check. The results of 

 the last two experiments also show that this nutritive check is 

 highly effective only when applied during a brief period of the 

 young female's existence. 



There still remains a residuum of unexplained results, both 

 negative and positive: why do 25 per cent of the individuals 

 most favorably treated for male production still remain female 

 producers? And, on the other hand, if starvation is a vital course 

 of male production, why do 15 per cent of highly nourished indi- 

 viduals become male producers without this apparent stimulus? 

 We can only answer these questions theoretically. But to one 

 who has noted minutely the phenomenon of variation in this 

 species and the unusual amount of variation which involves struc- 



