on some Carnivorous Insects. 321 



animal on which to experiment. Two or three broods of 

 small warblers that I tried to rear failed, and I could not 

 secure a fully-grown bird of a suitable species. Finally 

 I had to use driver-ants, unsuitable in so far as they possess 

 no sight, yet suitable in their general apparent readiness 

 to eat any animal substance, and in the fact that they 

 must very commonly indeed meet with lepidopterous eggs 

 when searching the herbs and lower shrubs on their forag- 

 ing expeditions. I also tried other insects — cockroaches 

 (suitable mainly in the fact that they are credited with 

 the readiness to eat or, at any rate, try any food of an 

 animal nature), a carnivorous ladybird {Alesia hidentata) 

 and a cricket {Arytropteris sp.), which has often been a 

 great nuisance to me, devouring insects of many kinds 

 that I have left on the verandah table. 



Obviously the animals were not perfect for my purpose ; 

 yet I felt that rejections by the drivers, the cockroaches 

 and the cricket, also any preferences any of them might 

 show, would at least, for the reasons I have indicated, be 

 suggestive, though of course by no means conclusive. I 

 therefore carried out on them the experiments which I 

 shall describe. 



A few remarks on the eggs used are comprised in the 

 concluding section of this paper. 



Note. — I have mentioned the fine grading of prey that 

 occurred in my experiments on insectivorous birds, wild 

 and tame, and the suiting of the capture or acceptance 

 to the exact state of appetite of the moment. I show 

 a diagram to illustrate these " layers " of appetite. It 

 is also true (and this too has an explanation bearing on 

 Dr. G. D. H. Carpenter's highly-interesting observations, 

 read to-night *) that a rapidly digesting animal may go on 

 eating a fairly low-grade insect (such as I have found many 

 Lycaenids to be), or even very low-grade species, indefi- 

 nitely, with occasional short intervals, if higher-grade 

 prey is not available in sufficient quantity to carry it to 

 a more advanced stage in the process of filling up. Thus 

 a swallow of mine ate more than 80 Neptis and a small 

 hornbill (Lophoceros) more than 50 Danaida chrysippus, in 

 each case in quite a short space of time ; a rest of a few 

 minutes after each refusal, accompanied doubtless by 

 subsidence, rendering the bird's digestive apparatus cap- 

 able of dealing with three or four more. The swallow 

 * Proc. Ent. Soc, 1916, p. Ixiv. 

 TRANS. ENT. SOC. LOND. 1915. — PARTS III, IV. (JUNE) Y 



