THE TSETSE PROBLEM IN NORTH MOSSURISE. 371 



tsetses there in great numbers and very troublesome, but could find nothing to shoot. 

 The concentration (in this case merely the arrival of a large herd of buffalos) at the 

 more easterly Ma papa vleis in 1918 took place when the flies were (according to the 

 natives) already abundant there, and the small parties of buffalos that were appar- 

 ently keeping down the numbers of the visible fly at the Masando in 1918 could 

 hardly be described as a concentration. It was certain from the detailed distribution 

 of the large number of nmddy puparia taken at Kanyezi's vleis that the fly con- 

 centration was already in being here in the late rains, when also a great buffalo herd 

 was temporarily present ; and as my flies and pupae were indubitably largely the 

 descendants of the depositors of these puparia, it might be said that fly concentration 

 found in the dry season may be the result of game concentration either then or 

 previously. Even this (as accounting for the presence of a fly centre, apart from 

 the mere question of its population) was negatived by the native statement that this 

 fly concentration is permanent though variable in numbers, that it had been there 

 before the buft'alos came (as it had also survived their departure), and that the 

 animals had stayed a very few days only. In testimony to the great size of this 

 herd my guide, Kanyezi, who lives near and had seen them, pointed out strong paths 

 that had been beaten out and vleis that had been heavily pitted with spoor in a single 

 day. The natives said that the other centres seen or heard of by me were also present 

 each dry season, though again with a varying population. The conclusions that I 

 would draw from all I saw and heard, including the fact that away from these centres 

 we failed to find any but very stray pupae, even when buffalos and other game had 

 been abundantly present, are that what Shircore has called '' primary centres " 

 undoubtedly exist ; that their actual (not their apparent) population no doubt 

 varies with the immediately preceding food supply; and that in their situation 

 beside vleis they have hit off the best obtainable combination of the three desiderata 

 of food, moisture and shade (for vleis, generally speaking, and their neighbourhood 

 are attractive to game and other animals, and the wooding bordering them is 

 quickly back in laaf). It is even possible (in an area in which morsitans happens to 

 be dependent on the larger mammals) that a kind of selection is continually taking 

 place by which those vleis that for any reason have become neglected by the game 

 for a considerable period cease to be fly centres. It was certainly strange that 

 Gundoda's vleis, well watered, midway between Kanyezi's and the Masando and 

 quite near each, and in the same wooding, uncleared, but carrying a human popula- 

 tion and a more permanent shortage of game therefore than the others, should have 

 been shown by a fairly exhaustive examination to be practically devoid of pupae 

 and to have produced few flies. Some of these few, latterly, were marked flies 

 released by me in the Kanyezi vleis, and it is likely that the great number of flies 

 said to be at Gundoda's in the rains come first from this and other vlei series. 



Village Concentrations. 



A concentration of another kind than the two referred to above, and presumably 

 one of a very shifting nature, is that which appears to take place at native villages, 

 even in the rains, and particularly whenever the fly is abundant. Gundoda (whose 

 village, besides affording good shade and being situated at the edge of a weU-watered 

 vlei, is a tempting dismounting-place for fly)j^described to me well how in the 



