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



with a very small piece of it, but readily ate a P. cehrene and again refused without 

 tasting the small piece of abdomen. She then refused for a second, but changing 

 her mind accepted, though with marked disinclination, a common grasshopper 

 that she usually accepts amongst the last, crushed it and threw it away, and refused 

 even to taste several others of different species which I now offered her, yet she 

 very readily ate a P. cehrene. 



More than half an hour later she readily ate a P. angolanus and the Gynanisa 

 abdomen full of eggs, but then refused to taste a second P. angolanus or the 

 P. demodocus. 



[^Apparent order : — 

 A. hasuta J, and, if not merely f^ p ^gj^,,,,^ ^rown Hepialid (and C. cloantha'i). 

 through stimulation, P. ceryne^^ 6'. /oreW« (if not eaten through stimulation). 

 Qi.^Siwdi Gynanisa ethr a {thovAx). {_ 



3. P. demodocus. 



4. P. angola7ius. 



5. Terias. 



The abdomen at any rate of the Gynanisa was placed heloiv P. cehrene.] 



Exp. 129. — Decemher 28, Ate very readily indeed several smallish pieces of the 

 whitish meat of a leopard and refused the last, but readily ate a P. angolanus and 

 a Cymaiura hifasciata. She then refused without tasting a portion of a frog with 

 skin attached. 



Later I noticed that she had brought up the meat quite undigested in the form 

 of two large pellets, also the Cymatura. The beetle was lying by itself untainted 

 by the meat and may of course have been brought up on its own merits. I could 

 find no trace of the P. angolanus. If brought up it may have been picked out and 

 swallowed again [cf. observation on Bucorax cafer). 



Several hours after the above experiment I fed the bird up on grasshoppers until 

 she obstinately refused to eat any more, though several favourite species were 

 offered. She then accepted, crushed and ate a P. angolanus, wings and all, with a 

 certain amount of disinclination and persistently refused the next without tasting. 

 Yesterday's P. demodocus was dead and the thorax slightly dry, abdomen quite 

 damp and supple. I therefore removed the latter and holding the butterfly's 

 hindwing against it offered it, but the bird at first refused it, then leant forward 

 with disinclination, took and dropped it, crushed and at once threw away a 

 Gymatttra hifasciata, crushed thoroughly and swallowed a Polyhirma cenigma, 

 crushed well and rejected Bostrichop)lites cornutus. She was now shaking her head 

 and clapping her bill, possibly, I thought, in regret for having eaten the Carabid 

 (which was smelling appreciably, though not its strongest, when swallowed). I 

 therefore waited until she had once more recovered her usual calm demeanour — a 

 matter of a few minutes — and offeied a Lycoid Lygseid bug. 



This was tasted and at once flung away as were also a red millipede {Spirolohus sp.) 

 and once more the Bostrychid. A water-bug, Laccotrephes ater, L., was then 

 accepted with an air of puzzledom, turned in all directions, and crushed in every 

 conceivable position. Its leathery hide — for that is the only term to desci-ibe it — 



