360 C. F. M. SWYNNERTON. 



piece of evidence whicli, coming from high- veld kaffirs previously unacquainted 

 with tsetse, seemed rather convincing. Whether they then " homed " or died 

 it is impossible to say. Old natives who herded cattle in the earlier days of Umzila's 

 experiment had previously given me similar instances of long stays by the fiy. Of 

 course there was an attraction, but the seasonal spread of the fly into other deciduous 

 areas {v. Sections ix and xvii) will itself, if fully confirmed; constitute strong evidence 

 against the universality of homing, as the well known recent spread of many fly 

 areas already does against its completeness. 



Hours of Activity. 



G. brevipalpis, as I found, attacks in the shade of the bush at any hour, but does 

 not fly appreciably outside its own thicket to do so except in the early morning, 

 after sunset and on dull or rainy days. Having attacked, even in sunny weather,, 

 it completes its object of feeding or riding, unless its victim emerges into the open. 

 In this event it usually leaves. In a special experiment that I carried out with the 

 aid of a lantern occasionally lit and turned on I found that both this fly and 'pallidifes 

 attacked right up to dark. Then every fly suddenly disappeared. On moonlight 

 nights they continue their attacks, and all three flies come to light, including fire- 

 light, and, once there, bite freely. G. morsitans 1 found, as have other observers, to 

 be most active in the rather warmer hours and least on cold or dewy mornings, but 

 it too attacked up to dark. It is probably dependent on shade to retire to occasion- 

 aUy when blood and water are not available, for I captured several male flies that 

 had been resting under logs in the wooding adjoining vlei edges frequented by male 

 clusters. The frayed wings of some showed that they had not just emerged. 



The Bite, and the Attack; Methods of Marking. 



Nothing struck me more than the frequency with which a fly would steal a feed 

 from the back or (especially) leg of a kafir walking in front of me and fly ofl replete 

 before the man noticed it. Very frequently the first intimation of the fly's presence 

 comes with the final withdrawal of tne proboscis ; sometimes even that is not 

 noticed. At other times a bite is noticed instantly and felt acutely. The natives 

 who converse on the phenomenon hold that a bite over a bone hurts and in a fleshy p'^rt 

 does not, but it is obviously a matter of local presence or absence of the specific 

 sensory nerve-terminations. I have myself had both painful and painless bites 

 quite close together on my forearm. 



The practical bearing of the fact is in relation to the criticism (based on the success 

 of observers' pet animals) that flies will not make good their feeds on baboons, etc., 

 and that such animals need not therefore be seriously considered in relation to the 

 efiect of game destruction. A second fact to be remembered here is that captive 

 animals have abundant leisure for fly-catching. A wild baboon is often absorbed in 

 raiding, hastily cramming his pouches, quarrelling, keeping up with the troop, or 

 participating in shouting contests. Especially where tsetses are present all the time,, 

 he can hardly attend to them except incidentally. The black man in these circum- 

 stances brushes the fly he notices away — commonly, if walking, by a wrist move- 

 ment that brings the butt of his assegai sharply to the point on his back that was 

 bitten — and the fly at once returns ; he makes an eft'ort to capture it chiefly when 



