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



threw away a JSfejytis saclava. She went on to crush and eat in quick succession 

 another Neptis agatha, a Neptis saclava, and a Neptis goochi, then seven more 

 Neptis saclava, four Nep)tis goochi, and yet another Neptis agatha, also a Belenois 

 and a Leuceronia argia $ . All these were eaten with apparently the greatest 

 relish. She then tasted and rejected with evident dislike Mylothris agathina, 

 Nyctemera leuconoe, and Danaida chrysij^j^us, but readily ate a Pyrameis cardui, a 

 Precis clelia, and a Precis cebrene. All the above had one hindwing attached. 



I now gave her five Charaxes brutus, all of which were eaten with apparent 

 relish. She then tasted and rejected Neptis agatha with evident dislike and refused 

 without tasting Neptis saclava, each of these with one wing, afterwards tasting and 

 rejecting, without wings, Neptis saclava and Eurytela hiarhas. She then refused 

 Precis cebrene, wingless, but immediately afterwards tasted it and, evidently 

 reassured, swallowed it at once with the gi-eatest readiness, as also & Pyrameis cardui. 

 She now tasted and rejected a one-winged Eicrytela hiarbas, tasted suspiciously and, 

 after a little wavering, dropped Catopsilia florella, but readily ate, after crushing 

 it, a Charaxes neanthes, refused persistently even to taste Terias senegalensis, 

 Glutophrissa saba 2 ? and Belenois severina, each with one wing, hesitated con- 

 siderably over taking a Charaxes zoolina, very possibly, I think, as a result of its 

 whitish coloi-ation, and finally tasted it very warily and with evident suspicion and 

 dropped it. On my reofiering it a second and third time, she each time again went 

 through this performance, apparently not finding enough wrong with the taste to 

 decide her finally against the insect, yet perhaps unable to make up her mind to 

 eat a bvitterfly which in appearance possibly reminded her of some Pierine that 

 she was not hungry enough to eat. 



She now readily ate a Charaxes cithcBvon, two Charaxes candiope, and three 

 Charaxes bruti(,s, the last two or three with decided difiiculty, evidently through 

 repletion, refused without tasting Pyrameis cardui and Precis cebrene, crushed 

 perfunctorily and threw aside a Precis archesia (wingless), refused without tasting 

 a wingless Charaxes zoolina, crushed and listlessly threw aside Charaxes neanthes, 

 and refused without tasting Pseudacroia lucretia. 



Now and during the next forty minvites I several times offered her the pleasanter 

 grasshoppers and such butterflies as the large species of Charaxes, Pyrawjeis, and 

 Junonia, bvit they were always refused without tasting. At about the end of that 

 time, however, she roused herself somewhat, tasted and readily ate two Precis 

 cebrene and a Precis clelia, tasted and rejected Neptis saclava and Glutophrissa 

 saba 5 J but readily ate after tasting each a Precis cebrene, a Pyrameis cardui, the 

 Charaxes neanthes previously refused, and Precis archesia, tasted and rejected 

 Papilio angolanus and Pscudacrcea lucretia, but readily ate a Pyrameis cardui. She 

 then tasted and rejected a Byblia, readily ate a Precis cebrene, tasted and rejected 

 Acrcea encedon, but readily ate a Pyrameis cardui and a Precis cebrene. 



A little later, evidently slightly hungrier, she ate readily enough, after crushing 

 each, a Eurytela hiarbas and a Neptis saclava, then (but with some slight hesitation) 

 a Neptis goochi. 



[In the first part of the experiment the roller ate Pyrameis cardui (2), Precis 

 cebrene (2), Precis clelia, Charaxes brtctus (8), G. neanthes, C. cithceron, and C can- 

 diope (2) after refusing to eat Mylothris agathina, Nyctemera leuconoe, and Danaida 



