330 C. F. M. SWYNNERTON. 



There are tkree great permanent fly-areas in the country investigated : — (1) The 

 morsitans-pallidipes area of the granite-gneiss ; (2) the brevipalpis area of the 

 Lusitu rubber forests ; (3) the brevipalpis area of the " Oblong." 



Between the last two is sandwiched an area of more deciduous foliage. Out of 

 the higher parts of this such fly as has got there is ousted in the cold weather bv the 

 leaf -fall, and in its lower parts the fly becomes scarcer and more localised before the 

 rains, reappearing with the leaf in October. These last remarks refer to G. brevipal- 

 pis and pallidipes, the latter fly occurring throughout the area. 



Finally, there is a narrow strip along some of the frontier, widest at Spungabera 

 and on the hill-mass enclosed by the Puizisi-Muchamba confluence, that may be 

 regarded as quite fly-free, excepting in so far as travelling parties of male flies may 

 on occasions be carried into it during the rains and early dry season by buffalos, etc. 

 Nearly all the country on the British side of the frontier may be regarded as falling 

 within this category. 



G. austeni was taken only near the junction of the Buzi and Mtshanedzi. The 

 type of wooding in which it occurred is not uncommon on lowland streams, and 

 there can be little doubt that further investigation will show it to have a more 

 extended local range. 



More generally it may be concluded that each fly is confined closely to particular 

 woodland types, which, in turn, are confined to particular geological formations, 

 the conditions that suit j;a^^^^^j9e5 being the most widely distributed. But the 

 distribution of the permanent fly is by no means thus conterminous with that of 

 either the buffalos, the big game generally, or any particular species of mammal, 

 though abundant mammalian life of some kind is present in each of the areas, flv- 

 carrying or otherwise. Man is relatively scarce (except on the Lusitu), and very 

 small mammals are abundant everywhere in the fly area. The dominant bird 

 elements of the brevipalpis areas are insectivorous and gallinaceous. The morsitans 

 area has in addition a stronger bustard-plover element and, at some of the 

 permanent fly centres, a very limited water-bird clement, as well (apparently) as a 

 somewhat larger reptilian population than the other areas. I may add that 

 various moth larvae, numerous, gregarious and often immense, are much more 

 in evidence in Brachystegia areas than elsewhere ; but, for reasons to be stated, 

 it is doubtful in spite of Roubaud's results whether they are of great use to 

 the fly. The fact that brevipalpis does not come readily to man makes its 

 presence difficult to detect unless one is lucky enough to come across a male queue 

 some evening. However, my conclusion that the " Oblong " was well infested was 

 come to after I had seen only two flies. Their hum in flying struck me as being 

 identical with a sound I had been hearing continually since I came there. I then 

 made a point of disturbing all large flies of any species I saw and made my natives 

 bring me for a day or two aU they could catch for release close to me, but I failed 

 completely to match the sound, which I therefore concluded to be that of G. brevi- 

 jmlpis. I continued to hear it, often on the part of large flies which I could not then 

 identify visually that came out to us and turned back, evidently rejecting us, and of 

 flies that flew away from the shady sides of tree-trunks when these were beaten. 

 An observation on pigs of a day or two later and subsequent results from cattle 

 abundantly justified the conclusion. 



