364 C. F. M. SWYNNERTON. 



to the borders of which and the old game paths traversing them the male flies were 

 obviously betaking themselves on emerging. The soil was sandy, the formation 

 was granite-gneiss, and the wooding was the Brachystegia-dommahted savannah 

 forest of that formation. 



The females when not feeding certainly spent their time in hiding away — ^from 

 the males (as Lamborn has suggested) and, I am certain myself, more especially 

 from birds and the sun. The indirect evidence showed clearly that it was mostly 

 Avhen thus hiding — in a spot which suited their own convenience but by no means 

 necessarily that of the pupa (cf. also Lloyd) — ^that they dropped their young, sometimes 

 (as my indirect evidence again most strongly suggested) from several feet from the 

 ground. Most of my puparia were found under prostrate trunks and branches, 

 raised little or much from the ground, but batches were also found in the angles of 

 root-buttresses, under mere leaning trees, under fallen palm leaves, and below and 

 between the outleaning dry leaf stems at the bases of palms (Hyphaene ventricosa), 

 in holes in trees and in holes made by animals in banks, and where the hoof -marks of 

 passing animals had broken the slight crust of a sandy stream-bed immediately 

 under light overhanging grasses. I failed, in spite of much search, to find tsetse 

 pupae in the leaf-sheaths of palms, though I took Lepidopterous and often other 

 Dipterous pupae there. The situation was usually dry, but some of my sand-stream 

 pupae were in a moist situation, and I took a pupa-case that had emerged success- 

 fully under a fallen Eugenia log in very damp ground on the edge of a vlei. 



I also carried out an experiment, lasting many weeks, with the pupae taken. 

 Some were exposed to conditions as normal as I could make them, some were in 

 soil exposed to day-long sunlight and others were in ground kept always moist and 

 either warm or cold. It was wonderful how well those exposed to the extremes 

 (perhaps less extreme than the conditions used by Koubaud) emerged, but less so 

 when one remembers the continuous state of the ground in such a wet season as 

 1917-18 or so dry a one as 1912-13, Pupae placed in a calabash of sand exposed 

 continuously to the sun failed, but in travelling they worked right down under the 

 sand in any case and many of them showed definite crushing by it. It is worth 

 noting that in the moister situations (in the vleis) practically no trees grow except 

 on the ant-heaps, consequently logs and other dark hiding places are relatively 

 seldom found there. Otherwise it is probable that pupae would have been taken 

 there more frequently. 



One other condition should be mentioned. The vast majority of the pupae were 

 in the nearest suitable hiding place to a spot in which big animals had been lying 

 down — buffalos (mainly), wart-hogs, hartebeests, etc. This seemed to accord with 

 the fact that we ourselves were attacked by feeding flies (as opposed to males 

 utilising us as vehicles) mainly when we were stationary and especially when we 

 were squatting do\vn searching for pupae. Again, of two sand-streams, otherwise 

 not dissimilar, one that was used as an occasional game path yielded a number of 

 pupae ; the other, used only by small carnivora, yielded only one pupa — in a highly 

 placed hole where a duiker had lain on the edge of the bank just above. It was 

 frequent for the puparia to be near ant-heaps, but this was the result of the fact 

 that the buffalos had chosen the sheltered sides of ant-heaps to lie on during the rains. 



