628 F. Finn — Experiments with various Birds. [No. 4, 



I then put in a number of butterflies, mostly Papilio polites and 

 P. demoleus, with several Euplceas, one or two D. genutia and 

 D. chrysippus, and one P. eurypylus and two or three other non-warningly- 

 coloured butterflies. 



A Yellow-vented Bulbul attacked first one of the last-named. A 

 White-crested Bulbul took and rejected an Euploea, but I saw one of 

 these birds swallow a P. polites, and one eagerly pursue a Junonia, 

 which was ultimately taken by a common Bulbul, I think because it 

 was nearest; at any rate the bird left it, and took a P. demoleus. Then 

 the Liothrix, which had had a P. polites, took this Junonia. 



I saw one of the White-crested Bulbuls flick away a P. demoleus 

 as if distasteful, but I also saw a specimen of this butterfly manipu- 

 lated by a bird of this species. 



While P. polites and P. demoleus were still available, a male Elym- 

 nias undularis, which had escaped from a White-crested Bulbul early 

 in the progress of this experiment, was alive in a corner apparently 

 unnoticed. I blew it out, and it settled further up, when after a 

 little time a Red-whiskered Bulbul took and ate it. 



One White-crested Bulbul then after long manipulation swallowed 

 a P. demoleus, not appearing to relish it. 



The P. eurypylus was attacked by a Yellow-vented Bulbul, but 

 I found it later, apparently unhurt save for the loss of a wing. 

 Yet a little while afterwards it had disappeared, while a P. polites and 

 P. demoleus were still left alive when the birds roosted. There was 

 fruit in the cage. 



V. Next day, both these butterflies left overnight had dis- 

 appeared. 



At the end of the afternoon (there being fruit in the cage) I put 

 in eight Danais chrysippus, and a similar number of non-warningly- 

 coloured butterflies, and a hawk-moth. I placed these on the ground, 

 decapitated, instead of offering them alive as usual. 



I saw Bulbuls of the Yellow-vented, White-crested, and common 

 species eat D. chrysippus, and these were all gone before the other 

 butterflies, which were inconspicuous on the sand. But the Liothrix 

 saw and selected the latter, and did not eat any Danais, though he 

 had been eating their heads with those of other butterflies as I picked 

 them off. I saw one White-crested Bulbul eat a Danais and then a 

 Junonia, while its fellow was engaged with a non-warningly-coloured 

 species. The Yellow- vented Bulbul I saw eat a Danais did not seem to 

 like it much. I then put in a D. genutia, an Euplcea (dry and with- 

 out abdomen) and a male Nepheronia hippia, dead. A Common Bulbul 

 ate the D. genutia, and a White-crested one the Nepheronia. I put in 



