348 C. r. M. SWYNNERTON. 



bush might be moderately open (as it was in Makwiana's Makubvii through the 

 trampling of elephants and buffalos). Very frequent — almost constant — con- 

 stituents of it were the small trees or large shrubs, Brachylaena rhodesiana (ipahla), 

 Maxkhamia lanata (mubfeya) and Vernonia podocoma (mudambasese), the latter 

 being replaced near the Sitatongas and across them by another woody Vernonia. 

 Bersama niassae (muyahawa), Vitex eylesii (mukubvu) and Conopharyngia elegans 

 were not infrequent, but a great mixture of other species, both of the savannah 

 forest and the primary forest outskirts, were commonly present, and where the tree 

 element overtopped the lower layer the wooding became fit for brevipalpis. 



Generally speaking, brevipalpis needs coppice with overwood, while pallidipes 

 prefers coppice without it, but in the right types — ^such as rather poor Brachystegia 

 —it can apparently dispense with yet lower growth. Clearing experiment No. 1 

 (below, p. 373) is of possible interest in this connection. Here jmllidipes attacked 

 the cattle somewhat freely in tall primary wooding in which the undergro\Hh had 

 been cleared, though before the clearing only brevipalpis had been present ; but it 

 may have come direct to the cattle from the wooding (PL xvii, fig. 2) outside. 

 It is curious that in all my special trials on the Buzi for brevipalpis in wooding 

 of the types shown in PI. xiii, fig. 1, I never took pallidipes either, though 

 I did take it here in places with low growth. It may possibly best be described as 

 a "low- wooding fly," the low wooding varying from mere coppice and bush 

 savannah to poor Brachystegia, etc., though better wooding is to some considerable 

 extent utilised. 



All vleis that I came across west of the Sitatongas were searched for male clusters 

 of this fly and the neighbouring — and other — wooding for breeding places, but 

 unsuccessfully. Though stated to be abundant at times in the rains it is generally 

 by far the scarcer of the two flies where it occurs with brevipalpis. The latter 

 often appears to be scarcer through its reluctance to show itself to man. A pomt 

 to be referred to also is the fact that pallidipes did not occur in all suitable wooding. 

 Much would be passed through without encountering more than a stray fly, if any ; 

 then suddenly a simultaneous attack by three, four, half a dozen or more might 

 take place, as already described. Water (streams or vleis) was usually present 

 not far away, but this part of the country generally is well watered, and similar 

 Avater in sunilar woods more often failed to produce pallidipes. A small dry season 

 population in a large area, and gregarious habits, probably sums the matter up 

 west of the hills, but a study of my notes shows also that even stray pallidipes were 

 relatively seldom taken anywhere at any great distance from vleis, glades or streams. 



Of Glossina austeni I have in my possession a male fly and about sixty puparia 

 all taken in two localities on the Mtshanedzi, near its junction with the Buzi. Fifty- 

 four of the puparia, empty and (judging by their muddiness) dating from the late 

 rains, were taken in the (roughly) primary-type forest that lines the Rupisi thermal 

 stream, from under logs. The Rupisi at this point, immediately above its junction 

 with the Mtshanedzi, was still hot enough to give me a pleasant warm bath, and 

 its temperature doubtless influences that of the fringing forest. The latter is of 

 appreciable width, and is here dominated by Kliaya nyasica and such trees as 

 Trichilia. It is, to be exact, secondary forest under invasion and still contains 

 fine tamarinds (T. indicus), Kigelia pinnata and other secondary constituents. 



