BIONOMICS OP TSETSE, ETC., IN THE GOLD COAST. 203 



From the foregoing table it is quite evident that there is a much greater disparity 

 in the proportion of the sexes of G. submorsitans than in the case of G. tachinoides. 

 The general average ranges from 90-95 per cent, of males, but there are certain 

 discrepancies ; e.g., at Murugu in June the percentage of males was only 46, while 

 in August it was 85. This is probably accounted for by the greater rainfall between 

 June and August than in the preceding months. The same may be the case in 

 October and November at Yapi, and it would certainly be accounted for in the case 

 of Kofaba in February. It is worthy of note that at Kapoto the percentage of males 

 in June was also low, viz., 60. This is in distinct contradiction to the percentages 

 found by Newstead in corresponding seasons in Nyasaland. It is worthy of note, 

 however, in connection with Newstead's figures, that the two last numbers given by 

 him (see page 200) are only for a period of over two weeks. Now, although there 

 were very heavy rains during the latter fortnight, it would not materially affect the 

 number of G. morsitans in so short a time. Had he taken the whole month's collection 

 together, the percentage would have been approximately the same as in the corres- 

 ponding two months. This is a clear case, I think, of false deduction from over- 

 subdividing. 



In May, however, when I visited the two Bush Camps at Larabanga the percentage 

 of males was even higher than at Larabanga in July and August. This is anomalous, 

 and I will venture no solution, but would only note that these districts were greatly 

 overrun by game. It may be that the food supply was very plentiful and that the 

 females were well fed and therefore more retiring, whereas the greater number of males 

 caught would be accounted for by their being on the look-out for females. This seems 

 to me a more feasible explanation than that given by Bequaert. 



V. Flight Experiments with Glossina. 



In order to ascertain whether Glossina tachinoides liberated at some distance from 

 the river would return to it, I performed a series of experiments by marking a number 

 of specimens by means of cutting off the two terminal tarsi of the various legs. 

 Altogether in these experiments about 3,000 specimens were used. On the road 

 from Yapi to Tamale there are two small streams, about two miles from Yapi, which 

 are dry in the dry season but contain a considerable amount of water in the wet season. 

 Notwithstanding this latter fact;, very few G. tachinoides are to be found on the banks. 

 They seem to prefer the larger river. 



In the first experiment in the dry season I liberated 150 specimens at each of the 

 following distances, 1 mile, 1J miles, 2 miles, 3 miles, 4 miles and 5 miles along the 

 main Yapi-Tamale Road. Of these, 59 per cent, liberated at 1 mile were again 

 recaptured, 15 per cent, liberated at 1 \ miles were again caught, while of those liberated 

 at 2 miles only 5 per cent, were retaken on the river bank. Of those liberated at 

 3 miles and 5 miles, none were taken ; while only one liberated at 4 miles was re- 

 captured. None were taken further away from the river than where they were 

 liberated. 



When this experiment was again repeated in the wet season, 47 per cent, of those 

 liberated at 1 mile were retaken, 25 per cent, of those liberated at \\ miles, and 10 

 per cent, of those liberated at 2 miles. On the other hand, an examination of the banks 

 of the stream at the different seasons showed a marked disparity. 



