fiXPLANATiON OF FORM AND COLOURING. 273 



to accept Neptis saclava and its like or to reject them, but the Fapilio had once 

 more carried her on too far, for even ten minutes later she i-ef used without tasting 

 PseudacrcBa lucretia ; then barely tasted if at all Pseudacrcea trimeni and refused to 

 take it from the forceps, leant forward to take a Neptis saclava but thinking bettei" 

 of it at the last moment withdrew without having actually tasted it. The decisive 

 moment was evidently however very near at hand, so giving the bird another two 

 minutes — all I dared to risk — I oflered the Byblia. She tasted it and then went 

 on to crush it well and eat it, refused with a shake of the head, then, relenting, 

 accepted, crushed very thoroughly and swallowed Pseudacrcsa trimeni, similarly 

 refused Pseudacrcea htcretia, but again relenting, tried it, ci'ushed it thoroughly, 

 and swallowed it. She next refused Neptis saclava too, but, changing her mind 

 once more, tasted it very thoroughly indeed and finally flung it away; refused but, 

 on tasting, ate without further hesitation a blackish skipper, after it tasting very 

 thoroughly and eating with equal readiness a Baoris. She then tasted well and 

 swallowed the Neptis which she had just rejected, and readily ate a Gegenes. 



[Summary, etc. : — (1) Note at the time : " The difficulty of satisfactorily placing a 

 butterfly of which one possesses only a single example is well illustrated by these 

 experiments on Pseudacrcea trimeni. It was not until late in the fifth experiment 

 that I was at last successful, but the decisive and unequivocal nature of the result 

 then obtained would have amply repaid a greater amount of trouble. I had before 

 suspected, chiefly from the behaviour of the shrike, that Pseudacrcea trime^ii was 

 decidedly pleasanter than Pseudacrcea lucretia, but no doubt could possibly remain 

 in the mind of anyone witnessing this concluding experiment that to the roller there 

 is not a particle of difference between them in this respect, and that that between 

 the Pseudacrceas and Neptis saclava is, if it exists at all, infinitesimal. 1 had the 

 good fortune to bring the bird twice to the exact point at which she wavers over 

 Neptis and other butterflies of the same calibre. On the first of these occasions 

 she showed unmistakably that neither Pseudacrce.a is inore acceptable than Neptis 

 saclava, on the second that neither is Zess acceptable. The fact of having just eaten 

 the two large Pseudacrseas would be sufficient to alter to a trifling extent the state 

 of its appetite and to secure the rejection of the not more unpalatable Neptis, which 

 was otherwise treated exactly similarly and in any case eaten [after stimulation] 

 immediately afterwards. 



The Pseudacrcea trimeni, I may here mention, had been kept on fresh damp moss 

 and remained alive up to the moment of its final disappearance down the throat of 

 the roller." 

 (2) Grades :— 



1. E. dryope, E. hiarhas, blackish skipper. 

 H. p)ers'picua (or, more likely, M. r 1. N. saclava, P. trimeni, P. lucretia, all about 

 campiiia, with which JI. per.'!, equal. 



spicua seems to get confused). L 3. Ypthima. 



4. Probably M. yulei and larva of A. serena ; 

 for, though the first Ypthima may have been 

 eaten under the stimulation of the grass- 

 hoppers, it is unlikely that any stimulation 

 would at that stage have induced the roller 

 to eat the Mylothris ov larva.] 



