A REVIEW WITH HYPOTHETICAL CONCLUSIONS. 99 



Bagshawe in a note appended to the above-mentioned paper of Degen's, which 

 was reviewed by him in manuscript, states that the total of several catches made 

 in 1906-07 on Lakes George and Edward and tributary rivers, where no 

 crocodiles were to be found, was : — 



£ 1166, Q 1662 ; d 41 %, Q 59 % 



There were some localities in this region where males were in excess. 



On Lake Albert and the Victoria Nile, where " crocodiles were always present 

 and doubtless were often fed upon," the catch was : — 



S 1951 ; Q 808 ; S 68 % Q 32 % 



McConnell* was " impressed with the apparent marked superiority of the males 

 along the large rivers and of the females on the smaller streams." He caught on 

 the Nile, in October, 5 females and 35 males ; in July, 10 females and 67 males ; 

 in September, near the mouth of small tributaries of the Nile between Nimulc 

 and Wadelai, 32 females and 24 males ; and was sure that further upstream the 

 proportion of females would be greater. He notes that in February (dry season) 

 Hies were rare along small streams, while in June (wet) they were common ; on 

 the Nile itself there was no marked diminution in numbers. 



Evidence of this general character might be further adduced, and it all tends 

 to indicate rather plainly (if the disparity of the sexes is actual and not merely 

 apparent) that : — 



(1) Local disparity between the sexes is characteristic of G. palpalis. 



(2) That it is not dependant on the season (vide Carpenter and Degen). 



(3) That it is not due to production of the sexes in unequal numbers (vide 



Carpenter). 



(4) That in some manner it is associated with conditions characteristic of the 



shore of lakes and larger rivers. 

 All these assumptions, however, depend upon the further assumption that the 

 proportions of the sexes in a lot of caught flies is indicative of the proportions 

 prevailing in that locality. 



Are the proportions of the sexes of caught flies a correct index of the sexes 

 prevailing in that locality ? 



The consensus of opinion seems to be that they are not. Theoretically, they 

 ought not to be, if laboratory results indicating that the males feed more fre- 

 quently and are generally more active than the females are properly applicable to 

 the field. If, for example, males feed once in every two days on the average 

 and females once in three, then, the sexes actually being equally divided, the 

 males ought to be to the females as 60 is to 40 in the catch. Fortunately there 

 are available precise figures, which, so far as conclusions may be drawn from a 

 single experiment, seem to answer the question in the affirmative. Carpenter, in 

 his Jinja marking experiment, to which reference has already been made, publishes 

 figures from which this is deduced, as follows. 



In the first place, out of a total of 6100 flies caught and marked between 

 27th July and 19th September the sexes were : — 

 2952^, 3148Q = c?48-5#, Q5V5%. 



* Bull. Ent. Research, lii, May 1912, p. 58. 



