External Factors of Sex Determination 379 



provided, of course, that it is not in reality already determined, 

 since the absence of differentiation of the sex cefls does not neces- 

 sarily mean that the sex may not have been previously settled. 

 Lowne had found that flies from large larvae are almost always 

 females, and those from small larvae males ; but Lowne himself 

 doubted the value of his observations, and Cuenot says his con- 

 clusion from them is erroneous. Weismann fed one lot of Cal- 

 liphora vomitoria abundantly, and starved another lot from time 

 to time. Both lots produced males and females. Cuenot's ex- 

 periments were as follows : 



Experiment I. The larvae after hatching were fed abundantly 

 on putrefying flesh, and produced large flies. The proportion of 

 females to males is given in the next table. PER CENT OF 



FEMALES 



Lucilia caesar 273 females 281 males 49.0 



Calliphora vomitoria 224 females 215 males 51.0 



Sarcophaga carnaria 96 females 90 males 51.6 



Experiment II. The larvae were given as little nourishment 

 as possible ; they often were entirely without food, and the mor- 

 tality was very great. The flies were small. The proportion 

 of females to males was as follows : PER CENT OF 



FEMALES 



Lucilia caesar 95 females 69 males 57-92 



Calliphora vomitoria 93 females 91 males 50.05 



Curtonevra pabulorum 26 females 17 males 50.50 



Experiment III. The larvae were kept under varied condi- 

 tions, being fed on brain, suet, and a little meat. They were 

 in a starved condition at first and were then abundantly fed or 

 were fed at first and then starved. Several species were reared 

 together. 



Calliphora vomitoria (April 19) .... 185 females 166 males 

 Lucilia caesar (Aug.-Sept.) 143 females 141 males 



The normal proportion of males to females in these species 

 of flies is approximately equal. This relation is also found in 

 those kept under the different conditions of the other experi- 



