266 H. LYNDHURST DUKE. 



among the roots of trees near the forest edge. With the rise of the lake in May 

 and June 1917 a large area of sand was submerged, and of that remaining much 

 was kept constantly damp by percolation or direct exposure to the spray. 



The Triumfetta shrubs have suffered severely or totally disappeared in these invaded 

 areas. As a result, with the fall of the lake, some of the reappearing sand is exposed 

 to the full glare of the sun and is useless, for the time at any rate, as a pupa ground. 

 Among other changes the old tree-trunk photographed by Carpenter, from under 

 which so many pupae were formerly obtained for the Mpumu laboratory, has 

 disappeared. In November 1911, the year of Carpenter's visit to Damba, the lake 

 reached its lowest recorded level. The highest record was touched during June 1917. 

 The difference between these two extreme levels is 5 '19 feet vertical measure- 

 ment. As already stated, the 17-18 inches fall which has occurred between June 

 1917 and July 1918 has laid bare in places along the beach some 10 yards of ground, 

 comprising an upper zone of fine white sand blending below with coarser pebbles. 



As regards the number of pupae found on this Damba beach in 1910-11, several 

 thousands were obtained monthly for the laboratory at Mpumu. All the old records 

 are not available, but the following extracts give a good idea of the quantity found : — 

 March 1911, 5,500 ; June 1911, 3,700 ; July 1911, 7,380 ; August 1911, 2,600 ; 

 September 1911, 8,000 ; February 1912, 4,620. I also recollect one month receiving 

 11,000. These numbers were obtained by 10 or 12 men working unsupervised during 

 a stay of a day, or perhaps 2 days, on the island, at intervals of a month or 5 weeks. 



In July 1918, 10 men working for 4 hours under my supervision obtained 1,680 

 pupae from the southern part of the beach where pupae were always most numerous 

 in former years. Of these, 514 came from around the roots of a single large tree 

 growing at the edge of the forest belt, its trunk being just above the 1917 high-water 

 level and at the edge of the grass and trailing-plant growth. On either side of this 

 tree were scraggy Acalypha bushes, and among their roots the bulk of the remaining 

 pupae were found. The whole area is in the shade of a large tree practically the whole 

 day long. No other sector of the beach produced pupae in anything like such numbers. 

 This particular pupa tree, together with the old trunk which has now disappeared, 

 was always specially patronised by the fly for larval deposition. On the eastern part 

 of the beach pupae were found only in small numbers and in isolated situations, 

 almost always in connection with Acalypha bushes or the roots of larger trees. For 

 several years no one has collected pupae from this beach. Two features common to 

 all places where pupae were found in July 1918 are dry, finely divided soil, generally 

 white sand or a varying mixture of sand and dry humus, and shade. 



The roots of the big tree which yielded the 514 pupae are surrounded by finely 

 divided, perfectly dry earth with a small admixture of sand. Wherever dry sand 

 and shade were found pupae also occurred. Dry, finely divided earth at tree roots 

 did not necessarily carry pupae, even though ant-lion larvae had established 

 themselves hard by. The vast majority of the pupae were within a few yards of the 

 high-water mark. Two pupa-cases were found at the base of a tree some 30 feet 

 within the forest zone and 60 feet from the water's edge. At the part of the beach 

 where most of the pupae are found, the 1917 high- water mark is at its farthest from 

 the present water level. The forest zone at this point is narrow, scanty, and backed 



— ■■' 



