OBSERVATIONS ON THE BIONOMICS OF G. PALPALIS. 265 



strain. On Nsazi the flies were less numerous, but the average daily number obtained 

 by the 6 boys together when all were working was a little over 700. These flies were 

 caught in the height of the dry season, of which Carpenter (1) writes : "A decrease 

 in the relative humidity of the atmosphere is therefore responsible for a large drop 

 in the number of flies ... it appears that flies can live several months, but are killed 

 off in large numbers at the onset of the dry weather." 



The diminution in the numbers of fly on Kimmi is very marked. One of my original 

 fly boys spent 6 hours on the island which he knows well and caught 100 flies, and this 

 represents a fair average day's catch in July 1918. The pupa grounds are also 

 diminished, though on this island there was no single extensive breeding ground. 

 On Nsazi fly-beach in July 1918, 12 boys searching for 20 minutes under my super- 

 vision obtained 33 shells and 8 full pupae. Fly were rare, the most expert fly boy 

 catching only 60 in the course of a whole day's search ; of these, 40 were from this 

 beach and 20 from neighbouring parts of the shore. The sex proportions of the flies 

 caught on these two islands during this 1918 visit are : Nsazi 63 flies, males 84-13 

 per cent., females 15-87 per cent. : Kimmi 1,289 flies, males 82'32 per cent., females 

 17-68 per cent. 



Damba Island. 



Carpenter (1) in 1911 described the Damba breeding ground as follows : " Formed 

 of small pebbles mixed with coarse sand left by subsidence of the lake 4 or 5 feet above 

 its present level, and about the same number of yards away from the present water's 

 edge." The high level here referred to was that of 1906. This famous beach, 

 owing to its very gradual slope, is peculiarly subject to the encroaching effects of 

 any rise in the lake level. From end to end of the beach the highest high-water 

 mark is clearly indicated by a tangled accumulation of branches of trees, sticks, reed 

 stems, and all kinds of vegetable debris. This line represents a lake level about 

 18-20 inches above the water line of July 1918. The distance of this high- water 

 line from the water's edge (July 1918) at that part of the beach where most pupae 

 are found is about 8-10 yards. In steeper parts of the beach only a few yards or 

 feet separate the two levels. 



This high-water line, as will be seen, marks the level of June 1917. Up to July 1918 

 the lake level has sunk some 17-18 inches. This 1917 line also coincides roughly with 

 the lowest level of forest vegetation. In former days, as is well shown by Carpenter's 

 photographs (1), there was a certain amount of Triumfetta and Acalypha growth 

 outside the actual forest zone along this Damba beach. Under these bushes large 

 numbers of pupae were found in the dry white sand. It was possible to walk prac- 

 tically dry-shod from end to end along the beach. In July 1918 this was no longer 

 possible, except in the southern half, and much of the Triumfetta had disappeared. 

 The fringe of grass and trailing plants which, in Carpenter's photograph, extends 

 beyond the forest nearly to the zone of coarser pebbles, has been pushed back until 

 it coincides roughly with the line of the big trees. 



At lower lake levels, such as those of 1911-12, part of the space between the edge 

 of the forest and the water line was often occupied by the two shrubs mentioned 

 above, which flourish on a sandy soil. Under these shrubs the fly like to deposit 

 their pupae. Any fallen tree-trunks are especially patronised, likewise the angles 



