SEX RATIO IN GAMBUSIA. 40! 



males killed much more readily in " high temperatures, high H-ion 

 concentration, oxygen deficiency and concentrations of KCN." 

 He also found a much higher death rate among males than fe- 

 males in Gambusia held in containers for shipment, as well as 

 when confined in aquaria. Further along in the same discussion 

 this author writes, '' The males are smaller and hence are more 

 liable to be devoured by small predaceous fish than the much 

 larger female. Gravid female Gambusia in aquaria, also, attack 

 and frequently kill the males." The present writer is in entire 

 accord with Doctor Geiser's views and the data presented herein 

 appear to lend support to his contentions. 



The present writer in connection with his work dealing with 

 fishes in relation to mosquito control has had occasion to make 

 shipments of hundreds of thousand of Gambusia. It was noticed 

 from time to time, that among the dead the males usually appeared 

 to be proportionately much more numerous than the females. It, 

 of course, was impracticable to sex fish that were being trans- 

 ferred from one place to another for breeding purposes, hence the 

 percentage of deaths of each sex cannot be given for such ship- 

 ments. The writer is certain, however, that at least usually the 

 death rate among the males was higher than among the females. 



In a large series of experiments performed for the purpose of 

 determining the best method of catching and handling Gambusia, 

 the adult fish were sexed and for these the results follow. The 

 effects of high temperatures, also the effects of crowding- the fish, 

 as well as the results to be derived from the use of varying 

 amounts of water in containers holding fish, were tested. Fur- 

 thermore, an effort was made to determine, by the number of fish 

 that survived after several days of confinement under identical 

 conditions whether catching by seine or by dipnet was preferable. 1 

 The results with respect to the death rate for each set of experi- 

 ments will not be given separately, as it was higher for males than 

 for females for every condition tested. In some individual lots 

 proportionately fewer deaths occurred among the males than 

 among the females. Generally the reverse, however, was true, as 

 already stated and as shown by the data that follow. 



1 The practical results derived from the experiments referred to were 

 reported by Hildebrand (1925). 



