314 ME. C. F. M. SWYNNERTON ON THE 



more exhibited, the bird edged towards it with a somewhat doubtful air, took 

 it from the forceps but then at once threw it right away in disgust, afterwards 

 ignoring a second Charaxes hrutus, offered with the upper surface towards her. 



[It was evident from this experiment that for a bird in the appropriate stages of 

 repletion Papilio echerioides is capable of acting most effectively as a model for 

 Gha7^axes hrutus. 



Preferences: T. hi'igitta was preferred to A. monteironis, A. satis d , and Argina 

 amanda, and at any rate probably not less than A. asemaJ] 



Exp. 98. — October 21. After several grasshoppers the bird ate, though with 

 disinclination, a Alyccdesis cmnpina. After three more small gi-asshoppers she 

 persistently refused to have anything to do with a second Mycalesis campina, but 

 accepted and ate, again with an air of disinclination, a Nejjtis agatha, afterwards 

 refusing persistently once more to touch the Mycalesis campina. I then offered 

 a Coprid which I had found rolling a millipede. She eagerly seized it, but directly 

 pressure was applied to its glossy surface the beetle shot away to the other side of 

 the cage. Three or four times the roller obtained an excellent grip, but quite 

 failed to crush the beetle, though she appeared to exert her greatest force, and the 

 end was always the same — the beetle shot out violently and rebounded from the 

 side of the cage. At last the bird lost her temper and taking it from the forceps 

 dashed it to the ground, after that edging away every time I offered it. I then 

 offered the millipede {^Spirostreptus sp.). She crushed it three or foui' times with 

 her bill and rejected it; then refused to accept Neptis agatha or a Mycalesis 

 campina. 



[JYeptis agatha was preferred to Mycalesis campina, and at this stage the roller 

 was deterred by Sceliages adamastor from continuing her attempts to crush it 

 after about four attempts, hardness and gloss being the qualities that led to this 

 result. A broken-up millipede, probably dead for some time, was rejected at the 

 same stage.] 



Exp. 99.' — October 23. Had been feeding on grasshoppers. I offered her a wing- 

 less Pyrameis, the top of whose thorax I had smeared thinly with honey. The 

 roller accepted it readily, but at once flung it away in disgust and wiped her bill 

 vigorously on the perch. She then refused without tasting Neptis agatha, but 

 readily ate some working termites, of which insects she is usually not very fond. 



The experiment suggested that a new, strong or unexpected taste not in itself 

 unpleasant may cause suspicion in an experienced bird and bring about rejection. 

 It was carried out to test this view. 



Exp. 100. — October 24. Readily ate a Mycalesis campina, then accepted a Coprid 

 found rolling dung yesterday in the forest, crushed it for some time without 

 however breaking it up greatly, and swallowed it. She did not experience with 

 this beetle the difficulty of the other day. She then tasted and rejected an 

 Aletis monteironis, readily ate a Terias, and tasted and at once rejected a Lycoid 

 beetle. 



