OBSERVATIONS ON THE BIONOMICS OI< G. PALPALIS. 269 



sandy breeding grounds by the lake has exerted a powerful adverse effect on the fly 

 of the northern shores and islands of Victoria Nyanza. As the lake falls, on this 

 hypothesis, the fly will again increase in numbers. The level is already some 18 

 inches lower than it was in June 1917, and the fall will probably continue steadily. 

 Much valuable time has thus already been lost. I strongly recommend, therefore, 

 that an endeavour be made immediately to accentuate the work of the lake by 

 systematically clearing away the sheltering Acalypha and Triumfetta shrubs from 

 certain known breeding areas. In this way the sun will succeed where the lake has 

 failed, and there will be a still further reduction of suitable breeding accommodation 

 at a time when the fly are still much reduced in numbers. Such measures will entail 

 a relatively small expenditure of labour, and should be applied primarily, indeed 

 solely, to certain carefully selected islands. Nsazi, Lwagi and Bulago Islands carry 

 few or no situtunga, and their fly are confined to limited stretches. All these 

 islands would be suitable for cattle. By judicious treatment of the fly shores and of 

 the breeding grounds on the islets Kimmi, Tavu, and Ngamba, I consider that the 

 three first mentioned islands, all of which are of fair size, could be reopened to a 

 properly controlled native population. The fly on Nsazi and Lwagi were tested by 

 Fraser and myself in 1911, and proved negative to T. gambiense. Carpenter, in 

 an unpublished report which I regret is not available, found the Nsazi fly infective 

 to monkeys in 1914, by which time a few situtunga had established themselves on 

 the island. As regards the trypanosome reservoir, however, none of these islands 

 present any serious difficulty. Bulago will be the most difficult of these islands 

 to deal with, but its treatment is, none the less, quite a feasible undertaking. Trained 

 fly boys should be included in the population, and their work will be diligently to 

 catch the persisting fly, and to search for their pupae. A simple arrangement of 

 bonuses will ensure the efficiency of this measure. Islands such as Damba, Kome, 

 and Bugalla, thickly forested and teeming with situtunga, must be deemed beyond 

 the pale, and every administrative effort must be used to prevent indiscriminate 

 visiting of their shores should the above-mentioned islands be re-opened. 



There are, doubtless, other islands of the great and fertile Sesse and Buvuma groups 

 which could, with relatively little expense and labour, be opened once again. The con- 

 trol of these distant islands would, however, be a very much more difficult proposition. 

 There is good reason to hope that properly organised measures on the fines suggested 

 and applied immediately, before the falling lake permits of an extensive recovery on 

 the part of the fly, will result in the permanent recovery of a valuable tick-free grazing 

 area within 20 miles of headquarters. At a rough estimate the area of the three 

 islands concerned is 8 square miles ; Lwagi 3, Bulago 2, and Nsazi 3 square miles 

 approximately. 



The above recommendations are based upon the assumption that the variations 

 in the level of the lake are phenomena over which we have no control. Were it 

 feasible to secure a rise in level by some engineering device, it might be possible 

 to eliminate G. palpalis from Victoria Nyanza. A rise of 3 feet above the level 

 of July 1918 would ensure the submergence of the greater part of the present 

 breeding area of the fly, without affording it any corresponding compensation. 

 I believe that the fly would be reduced to such an extent by this procedure that, 

 even if the rise in level were maintained for only 6 months, the remnant could be 



