352 C. F. M. SWYNNERTON. 



itself carried some brevipalpis ; but in Mtobe's area the intervening spaces were 

 largely mere grass or bare Pterocarpus as on tlie Inyamadzi, yet the scattered 

 darker thickets carried unusual numbers of the big fly. The clue is doubtless 

 given in part by the fact that an actual detached primary-forest *clump, diminutive 

 enough, near the Inyamadzi-Chiredza junction did produce a few brevipalpis and 

 that the brevipalpis-'pioduciRg clumps in Mtobe's, etc., were of the shadier dense 

 secondary or more or less primary types. The temperature conditions in continuous 

 bush are more equable than in isolated patches, and the conditions for the existence 

 of the fly may have been lost in such Brachystegia patches, though retained by quite 

 diminutive clumps of the denser types. The Brachystegia bush on the Inyamadzi 

 is rather poorer in any case in thickets and in fly than that across the Buzi, and 

 it is right on the edge of the fly area with no passage to points beyond. Even 

 if the explanation suggested is the correct one — ^and it must be noted that the 

 observations it is based on were scanty — splitting up the shelter, as a substitute for 

 complete clearing, on the margin of such an area would (by itself) exclude the fly only 

 during a smaU portion of the year. 



Interesting also was the fact that kloofs ascending towards Spungabera from 

 the infested Brachystegia wooding on the Buzi showed no brevipalpis appreciably 

 above the general margin of that wooding, though much of the wooding in the 

 kloofs was apparently still suitable for it, and I had taken this fly up to a greater 

 elevation near the Inyamadzi. The explanation that this is due to the necessity 

 for a compact breeding population (possibly in part applicable to the occurrence of 

 7norsitans centres in some vleis but not others) seems discounted by what I have 

 just related of Mtobe's country ; but this principle would doubtless show best 

 on the margin of a fly area and if there were no passage of game to carry the flies 

 to points outside. It was perhaps significant here that below the Spungabera 

 forest a line of precipices divided the general fly-bush below from a wood of Brachy- 

 stegia above in which there was much leaf and undergrowth but no fly, and that 

 further along northwards the limit of the fly at that moment was not quite that of 

 the Uapaca- Brachystegia bush itself but a little inside it at the foot of a precipitous 

 slope up which the Gogoyo-Spungabera path goes, and on which the above bush 

 terminates. The absence of fly from the Brachystegia bush above the precipice 

 might be explained in any case on the lines that I have suggested for the isolated 

 patches on the Inyamadzi, but the precipice very definitely and the steep slopes 

 to some extent do seem to mark the limit here of the freer wandering of the pigs 

 and other large mammals ; though pigs and baboons are said to come up in numbers 

 in October and November to eat the ripening manzhanzhe {Uapaca kirkiana) fruits, 

 and buflalos come up in the rains (especially from south and north of the more 

 precipitous barrier) and, as the indirect evidence shows, bring up tsetses. I have 

 been interested to hear from Mr. A. J. Orner, who has been along the path there 

 recently in the rains, that even at that season, though he saw brevipalpis all the 

 way from the Umtwadza to the foot of the steep ascent (mostly male queues in the 

 path after sunset), after commencing to climb he saw not one more. In the rainy 

 season there is much leafy wooding with thickets above the Brachystegia also, 

 thus adding to the interest of an observation that is, however, incomplete through 

 tlie absence of cattle bait to draw the fly out. 



