354 C. F. M. SWYNNERTON. 



We must look for an explanation to a difference in the conditions. They are as 

 follows : — 



G. morsitans present. 6^. morsitavs absent. 



1. Smaller rainfall. 1. Larger rainfall. 



2. Surface soil loose coarse sand on gneiss, 2. Soils compact, much humus in denser 

 compact on basalt. wooding. 



3. On gneiss impervious kaolin sub- 3. Vleis fewer, mostly small and more 

 stratum bottoming numerous vleis, the often with long grass ; permanent streams 

 latter fringed merely with the Brachystegia numerous, largely fringed with big trees. 

 of the general wooding. 



4. Poor clean-stemmed savannah forest 4. Brachystegia savannah forest on 

 {Brachystegia) on gneiss ; scattered bush sedimentary rock far finer and with, in 

 savannah on basalt ; each type continuous general, much more undergrowth ; on 

 over a great area. dolerite in widely scattered groves. Here 



the main wooding is (a) dense secondary 

 types, also primary ; [h) highly deciduous 

 tree savannah. 



5. Grass mostly short, fires not severe. 5. Grass of tree savannah immense, 



fires fierce if late, milder in the denser 

 wooding. 



6. Much big game. 6. Except in " Oblong " less big game, 



but many pigs, etc. 



2 and 6 may be dismissed. I found successfully emerged morsitans puparia on the 

 compactest of soil on the basalt and the other two tsetses seem able to subsist here 

 in numbers in the absence of much " big " game, as morsitans also did after the 

 rinderpest. 



4 and 5. If (as is likely) it should be that only the tree savannah of the dolerite 

 is really, as a whole, suited to morsitans, then its high deciduousness, with its 

 fierce fires, should suffice to exclude this fly except as a possible summer visitor 

 carried in on game or man. The scattered Brachystegia groves of the dolerite are 

 utilised by pallidipes, and it is likely enough that it retires to these when the dense 

 secondary wooding loses leaf towards the end of the dry season ; but to morsitans, 

 unable to use the bulk of the bush, these Brachystegia groves would stand in the 

 same relation as the outliers of the Brachystegia of the higher Buzi in which I failed 

 to find brevipalpis do to the latter — only they are much farther away from the main 

 block and separated from it by a high hill range and the two narrow and broken strips 

 of dense, unsuitable wooding that fringe its eastern and western foot. To this extent 

 the Sitatongas are a barrier, and doubtless also if they were more freely crossed by game 

 moi'sitans would appear more frequently on their west side in summer. The apparent 

 absence of morsitans from the generality of the continuous Brachystegia wooding 

 of the " Oblong " (neither Dr. Lawrence nor I have obtained any thence) seems 

 to suggest that the diSerence in shade and undergrowth between this type of wood- 

 land and the poorer type on the gneiss is sufficient to exclude this fly. 



To be able to assess the influence of the rainfall here one would require to have 

 figures from many morsitans areas. It certainly seemed to me that morsitans, 

 on the gneiss, was very dependent on vleis of a particular type for its dry-season 

 breeding. Vleis quite of this type are rare, if not absent, on the west side of the 

 hills, but it is again difficult, without an acquaintance with other morsitans areas, 

 to know if these vleis do really represent a vital condition. It may be this factor, 

 rather than that of the wooding, which causes the demarcation, and at present 

 that is rather strongly my opinion. 



