﻿INSECTS OF EASTERN TROPICAL AFRICA. 303 



there is a moon and little or no wind. When it is most active it would certainly 

 seem to be more attracted by moving men or animals than by stationary ones, as 

 other observers have already pointed out. 



The limit of elevation at which G. morsitans occurs in Nyasaland would appear 

 to be about 3,000 feet. In Northern Rhodesia I have never met with it myself 

 at over about 4,200 feet, though it is said to have been taken at somewhat greater 

 elevations. This difference may perhaps be accounted for by the fact that 

 Northern Rhodesia has on the whole a somewhat drier climate than Nyasaland. 



Compared with most species of Glossina, the complete independence of water 

 exhibited by this species is remarkable. In the Luangwa Valley I have seen it 

 swarming, in intensely hot weather, at least five miles from any known water. 

 Indeed, the drier the atmosphere, the greater seems to be the activity of this fly. 



With regard to the relations of G. morsitans with game I must unhesitatingly 

 group myself with those who consider that the presence or absence of big game 

 is not the primary factor in determining the distribution of the fly. That the 

 presence of a large quantity of big game within a morsitans area might influence 

 the numbers of the fly by increasing the food supply is probably true, but 

 that the game materially affects the distribution of the insect I do not for one 

 moment believe. The majority of those who hold the opposite view have 

 acquired their experience in Northern Rhodesia and Nyasaland and, as I shall 

 endeavour to show, have probably been misled by peculiar and abnormal circum- 

 stances. It is now I think, pretty generally admitted that, north of the Zambezi 

 at any rate, there is no special relation between G. morsitans and buffalo. Cer- 

 tainly in this part of Africa their distribution by no means coincides. Where it 

 does so, these animals are certainly very attractive to the fly, probably because 

 they are slow-moving and are usually in herds of some size. 



Perhaps the most important point in connection with this insect is the extension 

 in its distribution which is believed to have taken place in Northern Rhodesia 

 and Nyasaland within recent years. There is now a considerable accumulation 

 of evidence, and my own experience is certainly in accordance with it, that a 

 well-marked spreading into apparently new areas has really taken place. I say 

 apparently new because, as I have already pointed out,* I believe that the fly is 

 really recovering the ground lost at the time of the rinderpest. By what means 

 the death of large quantities of game could have caused the disappearance of the 

 insect is now impossible to say positively. I think however the suggestion which 

 has been put forward that the blood of the infected animals was in some way 

 noxious to the fly is at least worthy of consideration. It is certainly not easy to 

 understand how the fly could have died of starvation during the rinderpest 

 epidemic since no actual extermination of the game as a whole seems to have 

 taken place. The zebra, for instance, which swarm in the Luangwa Valley, were 

 unaffected, and it is very doubtful if antelopes such as impala, waterbuck, puku 

 and roan, all of whicn are very common there, were much reduced in numbers. 

 It will also be seen from Major Stevenson Hamilton's interesting account! that a 

 similar disappearance of the fly, which cannot be explained on the ground of 



* Bull. Ent. Res. I, p. 306. 

 f Bull. Ent. Res. II, p. 114. 



