350 Aaron Franklin Shidl 



Since all the positive results upon which earlier conclusions 

 were based may readily be explained as due to substances in the 

 water, let us see whether the negative results offer any obstacles. 

 There is but one point of any considerable importance on which 

 my results are seemingly at variance with those of previous 

 workers. This relates to the question of starvation. The experi- 

 ments of Punnett and Whitney went to show that quantity of 

 food, or any concomitant factor, had no influence upon the pro- 

 portion of male-producers. Though I have arrived at an opposite 

 conclusion, my results are not, it seems to me, opposed to their 

 results. Starvation in the experiments of Punnett and Whitney 

 was limited to a period of hours after hatching, whereas in mine 

 it continued throughout life. It was supposed that the female 

 would be more susceptible early in life than afterwards. If the 

 sex of the immediate offspring of a female were to be affected, 

 probably only those factors which operated early in life would 

 be of any avail. If the effect of (apparent) starvation is not notice- 

 able until the second generation, there may be late stages in the 

 development of oogonia and eggs which are more susceptible to 

 the influence of starvation (really the small amount of certain 

 substances in the water) than are the very early stages shortly 

 after hatching. In Nussbaum's experiments and in mine, this 

 influence occurred in late stages as well as early, and it is impossi- 

 ble to state whether the critical period, if such exist, occurs at one 

 stage or at another. Whether my starvation experiments differ 

 essentially, therefore, from those of Punnett and Whitney, seems to 

 depend on whether the influence of the substances dissolved in 

 the water is felt in the first generation or the second. Both Maupas 

 and Nussbaum discussed this point with regard to certain 

 external conditions, but disagreed in their conclusions. Some 

 light is thrown on this question by the experiments with the fil- 

 trate from old food cultures. Among the females transferred 

 from spring water to the filtrate, some were male-producers; 

 but of the females of the next generation, none were male- 

 producers. This shows that the full effect of the filtrate is not 

 apparent until the second generation. Whether the proportion of 

 male-producers among the females transferred to the filtrate 



