OBSERVATIONS ON THE BIONOMICS OP G. PALPALIS. 267 



by an abrupt low cliff some 40 yards inland. The soil here is poor in humus and tends 

 to be finely divided, as the sandy elements of an ancient lake level persist to an 

 unusual extent. The fly are thus still able, in this particular locality, to find the 

 requisite conditions. 



Discussing the breeding grounds of 67. palpalis on Victoria Nyanza, Fiske (2) notes 

 that the vast majority of the regular pupa grounds of the fly occur between the high- 

 water level of 1906 and the then (1913-14) lake level. Since his visit the lake has 

 risen to a level "79 feet above that reached in 1906. The highest record was touched 

 in June 1917, but on 20th November 1917 the lake was only lj inches below this 

 record level, and it remained high all through December. Thus for a relatively 

 long period the greater part of all the principal pupa grounds of the fly was under 

 water. This rise in the lake level did not, however, lay bare proportional new areas 

 of sand to make good those submerged. The level of June 1917 corresponds roughly 

 with the limit of forest along the lake shore, which itself probably indicates a more 

 permanent and still higher level of long years ago. All along the shores of the Lake 

 there is ample evidence of ancient levels many feet above those of more recent times. 



The soil of the forest zone consists of tough humus, often hard on the surface, and 

 apparently unsuited to the requirements of the larvae. 



Conclusions. 



There has been, without doubt, a marked reduction in the numbers of fly on the 

 islands visited. The only factor that offers a reasonable explanation of this decrease 

 is the destruction of breeding grounds by the rising water. The rainfall and humidity 

 figures are given below, but do not suggest any probable explanation. 



On this subject Fiske (2) writes, as a deduction from his studies of 67. palpalis 

 on the islands, " Eestrict the breeding grounds and there is a corresponding decrease 

 in abundance of fly " ; again, " I believe it is a very conservative estimate that 

 90 per cent, of the flies on the islands are from puparia deposited on these beaches " 

 (i.e., of sand and gravel). He recognised, however, other types of breeding places 

 which the fly appeared to frequent by deliberate choice rather than by force of 

 necessity : " On four separate occasions considerable deposits of puparia have been 

 found in vegetable debris, fairly compact, not fine and, of course, well shaded." 



These tentative conclusions put forward by Fiske in April 1914 are strikingly 

 borne out by the events of 1917-18. It is of interest to note that, in considering the 

 effect of another big rise of the lake to the then record level of 1906, Fiske anticipates 

 that " The recurrence of the high- water will open up the sand and gravel deposits, 

 newly washed and clean of vegetable debris, to reoccupation." This will, 

 undoubtedly, be the eventual result of the 1917 rise. But, meantime, and until 

 a further fall in the lake level, any such beneficial effect is vastly counterbalanced by 

 the wholesale dampening of the favourite and almost exclusive breeding areas ; though 

 it is conceivable that in certain cases abandoned breeding grounds at the 1906 level 

 may be again rendered useful to the fly. 



I will not presume, on so short an investigation into a difficult entomological 

 problem, to dogmatise as to the cause and effect of the present marked diminution 

 of 67. palpalis on the islands visited. But, as a working hypothesis, it is justifiable, 

 after due consideration of the facts, to assume that the invasion of the favourite 



