EXPLANATION OF FORM AND COLOURING. .^05 



tin a Papilio angolanus. The liird was distinctly surprised at finding this in its 

 bill instead of the expected grasshopper, but as it was there [or because of 

 stimulation by the grasshopper] she proceeded to crush and swallow it. The next 

 was a JS^eptis agatha, which was well cruslied and rejected, and Etirytela hiarbas, 

 which was similarl}' crushed but eaten, as was also a Papilio angolanus. 



After three more small grasshopj^ers, Papilio angolanus, at first refused, was, 

 after a minute grasshopper, crushed and definitely rejected, Leuceronia argia 5 

 similarly crushed and thrown away, Papilio demodocus crushed and readily eaten, 

 Papilio dardanus readily eaten, and Papilio angolanus and Etirytela dryope 

 persistently refused without tasting. The bird then refused to touch a grasshopper 

 nor would she look at any of a varied series of grasshoppers that I now ofiered her 

 in turn. She also persistently refused without tasting Papilio demodocus S fi^nd 

 Papilio antheus ; crushed and liking it readily ate a Charaxes ethalion, refused 

 Papilio demodocus, ate after crushing it well Papilio antheus, and even more readily 

 a second Charaxes ethalion ; again refused Papilio demodocus, crushed well and ate 

 Papilio antheus, and declined once more to have any dealings with demodocus, also 

 however refusing without tasting Precis archesia, Pi-ecis antilojje, a grasshopper, and 

 Precis cehrene, being evidently neai-ly replete. 



[Summary : — 



1. G. ethalion. 



2. P. antheus. 



3. P. demodocus. 



r . [ 4:. P. angolanus, H. perspicua. 



L. argia a i i i 



5. E. hiarbas. ") E. dryope and 



$ - 



I 6. N. agatha and Terias brigitta. J P. echerioides S 



P. saba was probably placed at any rate not appreciably below N. agatha and 

 P. echerioides c5' . It was probably stimulation by C ethalion that caused antheus to 

 be eaten right up to ceftrene-refusing point, but antheus was in any case definitely 

 preferred to demodocus. 



A. rabbaice is likely to be amongst the pleasanter Acrseas.] 



Exp. 84. — September 9. There was food in the cage, but I fed the bird by hand 

 on various Orthoptera till 'she "absolutely refused to touch another. She then 

 crushed thoroughly and rejected Papilio angolanus, refused several diflferent 

 palatable grasshoppers, readily ate two Charaxes candiope, tasted and rejected 

 Papilio demodocus, readily ate a Charaxes pollux, refused, then tasted slightly and 

 rejected Papilio demodocus, refused without tasting Salamis cacta and Pyrameis 

 cardui, then simply took in the point of her bill and dropped Hypolimnas 

 ivahlbergi and Salamis cacta, first looking at the former with some suspicion 

 (its body and forewing no doubt reminded it of things unpleasant), and refused 

 most obstinately to taste Pyrameis cardui. Three minutes later she refused 

 persistently without, tasting Pyrameis cardui, Salamis cacta, Salamis anacardii, 

 Hypolimnas wahlbergi, and Precis cebrene. 



Later in the day a second small experiment failed in its object. I fed the roller 

 just too much, and all the above butterflies were once more persistently refused 

 without tasting. 



