THE TSETSE PROBLEM IN NORTH MOSSURISE. 353 



Grass Fires, Grass Growth and Soil. 



The fires scorch the leaves and undoubtedly drive out the tsetses for the time 

 being from the area burned, but on the granite-gneiss, on 27th July, I found 

 tnorsitans had returned to a patch of ground that had been fired on 5th July. 

 The grass was already 6-8 inches high even in the drier vleis and the trees on 

 their edges, stripped by the fire, were already coming back into leaf. 



Of brevipalpis it must be said that of the two types of forest that it inhabits, one 

 (prunary forest and its extensions) never burns save along its edges, except in 

 special circumstances which I have referred to already. 



With one possible exception, I should say that the tsetse pupa has no necessities 

 such as are not found incidentally in the environment necessary to the adult fly. 

 The nature of the surface soil is, I judge, a matter of nearly complete indifference 

 as a direct factor, for some slight accumulation of detritus is almost always present 

 in such spots as receive the larvae, and I have found puparia on all soils. As an 

 indirect factor it may be far more important, for it is quite certain that at most 

 only an occasional pupa could survive the extraordinarily fierce fires of the heavily 

 grassed dolerite outside the denser woods, when they do not take place too early. 



Food Supply. * 



I have already referred to the role of the big game in annually re-distributing 

 the fly. I found no country in which largish mammals (pigs, bluebucks, etc.) were 

 entirely absent, so that I am unable to say from my own observation that the fly 

 can or cannot exist without them. That man must contribute much to the seasonal 

 dispersals of G. morsitans I had abundant evidence, and I had plenty of evidence 

 also that it does not avoid native kraals if trees are present. 



Two apparent Fly-barriers in Mossurise. 



(1). The Sitatonga Hills. These (except possibly at their extreme southern end) 

 form the eastern boundary of the morsitans area. That this boundary is permanent 

 there can be no doubt. It was the boundary when I visited the area in 1900 and it 

 stni is to-day. I failed to find the fly or its pupa west of it, and no morsitans has 

 been amongst the numerous flies {brevipalpis and pallidipes) that have been collected 

 for me during many months by Dr. Lawrence. Mr. Jack, who passed from the 

 Gogoyo (Dysart) station to Mtobe's, imder the Sitatongas, in 1917, also took no 

 Qnorsitans. 



The Sitatonga Hills are a lofty and uninterrupted " knife-edge " passing right 

 from the Lusitu to the Buzi. There is relatively little back and forth movement 

 of game except at the southern end (where the buflalos have a passage), and it might 

 appear therefore that the ridge acts as a barrier, either direct or (through the game) 

 indirect. On the other hand, at Umtomana's pass and about it the hills are very 

 passable by game and have probably constantly been crossed by them formerly 

 and are crossed to some extent now, and a main native path goes through here. 

 My temperature experiment (above) suggests that the mere height of the hiUs will 

 not constitute a direct barrier; and (a conclusive argument) the ridge evidently 

 acts in no way as a barrier to the distribution of pallidipes and brevipalpis, which 

 occur freely on both sides of it ; why then should it to morsitans ? 



