260 MR. C. F. M. SWYNNERTON ON THE 



Atella phalantha, accepted with disinclination but ate a Precis cebrene, refused 

 without tasting a ffamanuTnida dmdalus, took, held for a few seconds, and finally 

 crushed and swallowed a, Precis cebrene, refused without tasting a Eurytela hiarhas, 

 accepted and ate a Hammiumida dcedcdus, and now, probably fairly replete, refused 

 without tasting Atella phcdantha, Hmnambinida dcedalus, and Precis cebrene. 



Five minutes later she again refused these three, as also a small brown grass- 

 hopper, of which she is very fond. 



Half an hour later she once more refused all four. 



[Summary, etc. : — (1 ) I quote from my previous conclusions in this experiment : — 



" The rejections of Acrcea caldarena, doubledayi^ and areca were vei-y decided and 

 emphatic, particularly the latter. On the other hand, all the Acrcea serena, as also 

 the Acrma asema, were subjected to rather more prolonged tasting. Baniaida was 

 evidently not distasteful in the same sense as the Acreeas, but the bird, probably 

 largely as the result of some past experience, was afraid to swallow it. 



" In the second stage the bird tried each Acrcea before rejecting it, but would 

 not touch NychAtona medusa, which was evidently intensely disliked. It seemed 

 probable, too, that the less nauseous Acrseas come in acceptability almost up to 

 the level of Mylothris agathina, which was itself less disliked to-day than 

 Mylothris yulei. The latter had its peculiar smell strongly developed, none of 

 the former had. 



" Henotesia perspicua is evidently more palatable than Neptis saclava. 



'■'• Prob. Order. — 1. Precis nataleoisis Q, Atella phalantha, Hamanumida 

 dmdalus, and Junonia cebrene, all eaten at the very end, are evidently 

 exceedingly pleasant. To these might be added, Leuceronia argia $ 

 (in preference to Eurytela dryop>e), which was once more accepted 

 practically at repletion-point, though a Teracolus was refused, being 

 confused perhaps with Belenois. 



2. Henotesia perspicua. 



3. Neptis agatha. 



4. Neptis saclava. 



5. Neptis goochi. 



6. Belenois mesentina. 



7. Terias senegalensis. 



8. Mylothris agathina. 



9. {Mylothris yulei '7) 



10. Acrcea serena anfl asema. 



11. {Nychitona''1 and Banaida chry- 



sippus ?) 



12. Acrcecc doubledayi and Acrcea 



caldarena. 

 18. Acr(xa areca." 



(2) Revising, I would not go so far, although I have reason to know that the 

 above list does represent the roller's usual preferences. The following list only, 

 I think, can be really at all justified from the first section of the experiment and 

 the other lists that I shall give fi-om the remainder of it : — 

 (a) 1. Belenois mesentina. 

 T, senegalensis. J 2. M. agathina d & $ . 



1^3. Danaida chrysippus, pi-obably (on taste, not necessarily 

 after-efi'ect). 

 . Nychitona medusa, Acrcea asema, A. serena (rubbed preferred 



to fresh). 

 . A. ccddarena, A. doubledayi, probably (on manner as against 

 A. serencib), A. areca $ . 



M. yulei 5 . 



