EXPLANATION OF FORM AND COLOURING. 229 



second wingless P. lyceus was bitten two or three times and rejected. The bird 

 soon afterwards, however, descended and ate it, still neglecting that with wings 

 and the Neptun%des\ both were lying on the floor of the cage. A Charaxes hrutus 

 natalensis without wings was then well crushed in the bird's bill, but in the end 

 rejected. Half an hour later (I do not know whether the bird had fed in the 

 meantime) I took it out from amongst the other food on the cage-floor and 

 reoffered it. It was again well masticated and thrown back into the throat to 

 be swallowed, but brought out again and rejected, the bird shaking his head 

 afterwards. 



[Nothing is said of the state of the bird's appetite. The fact that there was 

 food in the cage throughout does not necessarily in this species prove repletion, 

 but if by any chance Roller A's sentiments towards Charaxes hrutus (and other 

 large Charaxes) resembled that of my other two rollers, then its rejection was 

 practically tantamount to a declaration of nearly complete disinclination for food, 

 whether through repletion or some other cause. In that case the Tenebrionid and 

 the wingless P. lyceus, both eaten up to the point of complete disinclination, might 

 fairly be regarded as having been probably found highly acceptable, while the 

 rejection of the Neptunides and of the Avinged P. lyceus need not necessarily 

 indicate any very great M?iacceptability for them. A preference may perhaps 

 have been indicated in the fact that the Neptimides was rejected with promptitude, 

 the P. lycBus with wings more waveringly ; no preference necessarily attaches to 

 the Tenebrionid, seeing that it was already eaten before the other insects of the 

 experiment were offered. It may, however, be infei-red that wingless P. lyceus 

 was preferred both to the JVejJtuiiides and to P. lyceus with wings : a good instance 

 of what I have found throughout my experiments — that a butterfly withovit wings 

 tends to be slightly preferred to the same butterfly with those obstacles still 

 attached.] 



Exp. 2. — Feb. 25. Tasted with caution and rejected with signs of disgust a larva 

 of Atella phalantha. 



[Experiment useless, seeing that nothing is stated which might even approxi- 

 mately indicate the state of the bird's appetite.] 



Later in the day, noticing that yesterday's Papilio lyceus and Charaxes hrutus 

 were still, with a good deal of other food, lying uneaten in the cage, I offered 

 a S Papilio lyceus without wings. It was well crushed and readily eaten. On my 

 now offering a second, however, the bird shook his head and at first took no more 

 notice of it, finally, however, accepting, crushing, and swallowing it. I now offered 

 him a Precis natalensis f. with the brilliant upper surface of the wings displayed. 

 The bird refused obstinately to have anything to do with it, shaking his head and 

 clapping his bill violently whenever I brought it close up to it. I now offered two 

 butterflies of the same species with all wings removed. Each was crushed at once 

 and swallowed with the greatest apparent relish. On my presenting a yellow 5 of 

 Catopsilia florella with all wings attached, the roller hesitated, then took it in his 

 bill in a half-hearted way, the butterfly at once escaping and fluttering about the 

 cage while the bird settled down and took very little further notice of it. 



