EXPLANATION OF FORM AND COLOURING. 283 



Exp. 66. — July 11. Hungry, ate two small pieces of meat, crushed and tasted 

 well and swallowed without any hesitation a pupa of Acrc^a serena, then recognized 

 and refused its larva without tasting and with violent shakes of the head. She 

 then crushed and readily ate a Terias and Belenois severina, but tasted and threw 

 right away a S Mylothris agathina. 



[At the time I was convinced that the serena pupa was far better liked than the 

 larva, even perhaps fchan M. agathina J , and commented — " The roller's behaviour 

 on being shown the larva to-day was indicative of nothing less than horror." 

 Reading the experiment now, and contrasting the great difference in manner, I 

 still think that I may have been right, but it does not necessainly follow fi-om the 

 experiment. The possibility of special stimulation by the meat and inhibition by 

 the pupa, and the fact that the acceptance and refusals came in the wrong order, 

 necessitates caution. Terias was preferred to the serena larva, and probably 

 (5 M. agathina was liked at any rate not better than £. severina.^ 



Exp. 67. — August 1. Hungry; tasted and at once rejected Mylothris agathina^ 

 readily accepted, crushed, and ate two Terias senegalensis, accepted, crushed and 

 battered very thoroughly, and finally swallowed a large dark green Phymateus 

 with a strong unpleasant smell reminding one of Oascara sagrada and the leaves 

 of Datura Stramonium. She now refused without tasting Terias senegalensis, 

 tasted slightly and rejected the common Coccinellid, Epilachna hirta (exuding a 

 yellow juice), refused persistently without tasting and with violent and continued 

 shakes of the head a large, brilliant-metallic, black-spotted bug, Gryptacrus comes, 

 common on Croton sylvaticum, readily accepted, crushed, and ate Neptis saclava, 

 and refused obstinately without tasting it a second dark green Phymateus. 



The refusal of the second Terias was evidently simply due to the Phymateus 

 having filled the roller's stomach to a point in advance of the Therms-eating stage. 

 All that can be deduced with regard to the Phymateus (which was exuding bubbles 

 from behind its posterior pair of legs) is that it is likely to be unpleasanter than 

 Neptis saclava and less so than Mylothris agathina. It is probably on a level with 

 Terias. The refusal of the Cryptacrus comes (which I have found in the stomach 

 of a wild bulbul, Pycnonotus layardi, v. ' Ibis,' 1908, p. 66) may have been due (1) 

 to a previous knowledge of it or (2) to a mere distrust of its gaudy coloration. 



I now proceeded to feed the bird up on grasshoppers, etc., until she absolutely 

 refused to touch another. She then tasted and rejected Eurytela dryope, leaned 

 forward evidently much tempted by a huge green Pentatomid bug, Encosternum 

 delegorguei, which I now passed through the bars, but withdrew with a shake of 

 the head without having tasted it, and refused without tasting Antanartia schmieia 

 and Precis cebrene. 



Ten minutes later she refused without tasting Eurytda dryope, Antanartia 

 schceneia, and Precis cebrene, but accepted, crushed very thoroughly, and swallowed 

 head first the big green Encosternum (it smelt very strongly on being crushed, 

 not much before), and once more refused Precis cebrene and a common small grey 

 Pentatomid bug. She kept on opening her mouth, doubtless in connection with 

 what she had just swallowed, but whether this was to be interpreted as a sign of 

 pleasure or the reverse I could not say. 



LINN. JOURN. — ZOOLOGY, VOL. XXXIII. 21 



