XOT CLOSELY ASSOCIATED. 293 



one remembers that both inhabit densely-wooded, damp, secluded 

 districts, but the arguments advanced to prove that the Bos caffer 

 alone is the host of the Tsetse, and that the extermination of 

 the former leads to the disappearance of the latter do not appear 

 to me to be convincing. From my own observation I am inclined 

 to the opinion that hosts other than the one species mentioned 

 (or some other factors) are necessaiy for the propagation of the 

 riy, and that the hjematozoon is in all probability to be found in 

 many species of diptera. 



During my residence in Jubaland, East Africa Protectorate, 

 which extended over two years, I had ample opportunity of 

 studying the habits and distribution of the Tsetse-fly and the 

 effect it produced on domestic animals. 



Speaking generally, the fly belt is confined to two distinct 

 areas in that province. 



(1) The valley of the Juba River within the forest belt. 

 Here fly abounds for a distance of upwards of 400 miles — in 

 fact, so general is it in places that it is a source of annoyance to 

 Europeans and natives. Yet within the whole of that great 

 tract of country Buffalo are few and far between. The late 

 Mr. Jenner and I came across them in one place only (Lake 

 Galey and district). Within this belt is the district of Gosha, 

 80 to 100 miles in length, where the fly is peculiarly abundant, 

 yet Buffalo are not found ; notwithstanding this, the hsematozoon 

 is very virulent, and on the occasion of the late Mr. Jenner's 

 expedition to Lugh in 1899, he lost every camel and pack-ox 

 that went through (vide my Report, May 16th, 1899, forwarded 

 to the Secretary of State for Foreign Affairs, No. 53, May 31st, 

 1899). 



(2) The region of Lake Kumbi to the north-east of Desek 

 Wama (Lake Hardinge), where the distribution of the fly is 

 restricted to the dense forest belt in the neighbourhood ; yet 

 Buffalo are not known to frequent this district. It appears to 

 me reasonable to assume that where fly is abundantly found 

 extending over a large tract of country that its natural host (if 

 one only) should be in large iiumbers also ; or else the fly must 

 be possessed of extraordinary migratory powers. 



If that host be the Buffalo, it is strange that it is particularly 

 conspicuous by its absence in the extensive districts mentioned, 

 while waterbuck and bushbuck are common in most parts, and 

 yet again in others no animals of any sort are found save 

 monkeys and rodents. Amongst other pests that affect camels 

 very seriously in Jubaland is a species of gad-fly which is 

 restricted in its distribution to the open plains and sparsely- 

 bushed country in the districts of Desek Wama and Derib. 

 This fly, unlike the Tsetse, attacks animals during the day at all 

 hours, and the symptoms produced by it (in camels) are identical 

 in every particular with those produced by Tsetse, but I was 



