162 A. FRANKLIN SHULL 



chemical substances in the medium may influence the appearance 

 of sexual females among rotifers leads us to ask whether various 

 substances may not have a corresponding effect upon daphnids and 

 aphids. It is not necessary to suppose that the same substances 

 affect all these groups as affect one of them; nor will the effect 

 of any one substance necessarily be found to be the same in all 

 three groups. But it is certainly a pertinent question whether 

 chemical substances of some kind do. not have an effect in one 

 direction or the other, either in producing sexual aphids and daph- 

 nids or in preventing their occurrence. Schmankewitsch ('75) 

 early suggested that the concentration of salts in the water might 

 influence cyclical phenomena among daphnians in nature. Kurz 

 (74) likewise named chemical change as one of several agents that 

 might have this effect. Their suggestions were rejected by Weis- 

 mann (79), but Langhans ('09) has given evidence which he 

 believes shows that the excretions of the daphnians themselves 

 influence the proportion of sexual individuals. Papanicolau 

 ('10) rejects the excretions as an agent affecting the cycle, citing 

 a short experiment of his own in support of his conclusion. One 

 of his colonies was reproducing only by parthenogenesis, another 

 was starting into a sexual period. He exchanged the water in the 

 two cultures, but at the end of four days the former colony was 

 still wholly parthenogenetic, the latter still partly sexual. Even 

 if we assume that the entrance of the sexual period in one of 

 these cultures was due to the accumulated excretions, it could 

 hardly be expected, it seems to me, that four days would suffice to 

 change the mode of reproduction in an animal in which a period 

 of two weeks or more elapses between one generation and the next. 

 Moreover, another phenomenon reported by Papanicolau may 

 perhaps be looked upon as supporting the view that chemical 

 substances affect the cycle. He finds that late broods of a given 

 female are more likely to contain sexual individuals than are early 

 broods of the same female. This reminds us of the similar phe- 

 nomenon in rotifers, where, however, the facts are just the reverse. 

 The late females of a family of Hydatina are more likely to pro- 

 duce parthenogenetic daughters (ShuU, '10b) than are the early 

 females of the same family. The explanation suggested in my 



