„ ' >P. cardui, H. dcedalus, and T. flesus. 

 dla, ) ^ J 



Explanation ot foeM and colourInq. 277 



of which was tasted and promptly rejected. The Leuceronia argia S was, however, 

 eaten after very thorough crushing ; Eurytela hiarhas with two wings was crushed 

 and rejected, and Papilio lyceus § was very thoroughly crushed and swallowed. The 

 bird then tasted and rejected Eurytela hiarbas with only a hindwing, tasted and at 

 once readily swallowed a Precis cehrene, crushed well and rejected the Eurytela, 

 this time without wings, and tasted and very emphatically rejected the dark 

 Byhlia. 



The bird certainly gave everything a much fairer tasting than usual in this last 

 experiment, being probably unable to distinguish the colours well by lantern- 

 light. 



[Summary, etc. : — (1) Note at the time : " The object of the whole experiment had 

 been to place Crenis, and to obtain a further opinion from the roller on the 

 relative unpalatability of GatopsiUa fiorella, Leuceronia argia, Papilio lyceus, and 

 Eurytela hiarhas. In this I partly failed, but the bird was throughout in excellent 

 form and the results, so far as they went, appeared to me to be particularly trust- 

 worthy. Atella was unluckily lost, so could not be reoffered." 



(2) Grades:— 



1 . *S', anacardii and perhaps C neanthes. 



2. P. cehrene, 



3. C. fiorella, 



4. L. argia S and P. lyceus $ . 



5. E. hiarhas. 



6. N. saclava and Byhlia. 



7. Ypthima and Terias senegalensis. 



8. M. rueppelli cS and $ . 



9 M. agathina $ and M. yulei $ . 

 10. N. medusa and A. serena. 



One link in the above was weak, and to mend it I had to use my knowledge 

 (from other experiments) that Byhlia and N. saclava are approximately equal in 

 the roller's estimation. 



Actually Ypthima was eaten in apparent preference to Terias senegalensis, and 

 were it not for the possibility of special stimulation would constitute an additional 

 grade between Neptis (definitely preferred to Terias) and Terias. Were it not for 

 the same complication R. liheon and R. pisistratus, the Crenis, and Macroglossa 

 trochilus might be inserted on a level with at any rate Pyrameis. 



But even without all this the results are very full, and I have not hesitated 

 to deduce from differences in manner which were particularly good in this 

 experiment.] 



Exp. 60. — July 4. Crushed and rejected Mylothris yulei, but readily ate Neptis 

 saclava, and after it a few small grasshoppers and another Neptis saclava. 



I now offered two large grasshoppers which were readily eaten, the bird then 

 tasting and rejecting a Byhlia but crushing and readily eating a Neptis agatha. 

 She next refused for some time but finally crushed well and swallowed a Neptis 



