A REVIEW WITH HYPOTHETICAL CONCLUSIONS. 101 



may presume that the proportions of the sexes in a sufficiently large series of 

 caught flies offer a very good index of the proportions actually prevailing in the 

 locality. 



The possible connection between Sex Disparity and Specialised Migratory 

 Movement. 



Obviously the female flies on an island like Damba, where both sexes are 

 known to be produced freely and in approximately equal numbers, must either be 

 present and not caught, or they must be absent, and as shown above, such 

 evidence as is at hand indicates that they are not present. Partial sex segrega- 

 tion must be assumed to be actual and not merely apparent. If actual, the 

 writer is unable to see how it is possible to explain it except through — 



(1) Some agency which effects a greater rate of mortality upon males in one 



locality, and which is complemented by some other agency similarly 

 and discriminatively prejudical to the opposite sex in another locality ; 

 or else 



(2) the migration of males to, or of females from, localities where males pre- 



dominate and vice versa. 



Arguments in favour of (1) have been advanced by Kleine (as mentioned), who 

 thinks that a diet of crocodile blood reacts more disadvantageously on females 

 than males. His arguments are not at all convincing. They infer that the 

 islands, where males very largely predominate are unfavourable places for the fly, 

 on account of bad food. Nevertheless the flies remain abundant there. 

 Suppose the average percentage of females to be 20 ; then upon this 20 per cent, 

 must devolve the responsibility of maintaining the race, which is to infer that 

 each 20 individuals must produce 160 fully developed offspring, 80 males and 80 

 females. Of these, 60 females (or 75 per cent.) disappear, leaving the same 

 abundance and the same disparity in sex. Thus each female produces 8 young, 

 or 4 pairs, and the prevalent rate of reproductivity indicated is 400 per cent, each 

 generation. 



On the mainland, on the other hand, where Kleine assumes that conditions are 

 more favourable, 50 females would be only obliged to produce 100 progeny, or 

 one pair each ; this indicates a prevalent reproductivity of only 100 per cent., 

 and it would appear that instead of the islands being unfavourable for the fly, 

 they are (putting it in one wa.y) four times more favourable than the mainland. 



Kleine's conclusions not only that the flies died when confined to a diet of 

 reptilian blood, but that even when they survived reproduction practically ceased, 

 were later confirmed by himself and by Roubaud, and were supplemented by 

 experiments conducted at the laboratory of the Royal Society's Commission in 

 Uganda. The inference that islands where shore birds and reptiles only are 

 available as hosts for Glossina would be unfavourable breeding localities is, 

 therefore, wholly natural. 



But, aside from the results of these experiments, all evidence points toward the 

 islands as the most favourable localities for rapid increase, and one is forced to 

 the conclusion that in this instance, as in many others, laboratory results cannot 

 be depended upon implicitly as indicative of field conditions. Either the flies 



