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



Exp. 22. — May 21. Ate readily in succession five Charaxes cithceron and six 

 Charaxes candiope, refusing a seventh (probably replete). Some of these had a wing, 

 some wei^e wingless. Five minutes later she accepted and ate the seventh candiope 

 with apparent enjoyment, refused JVeptis saclava without tasting, tasted and rejected 

 Neptis agatha, crushed and very readily ate Precis archesia, tasted and rejected a 

 Papilio dardanus cS with the usual smell somewhat strongly developed, and refused 

 without tfifiting Salamis anacardii and Papilio dardanus $ f. hippocoon. All these 

 were wingless. 



Twenty minutes later I offered a wingless Eurytela dryope. This was at first 

 refused without tasting, but the roller, which had possibly at first still had Neptis 

 in mind, changed the latter and, having tasted the butterfly, crushed and swallowed 

 it. Wingless Neptis saclava was then also tasted but rejected, while a S Gatop- 

 silia Jiorella, wingless, and two freshly dead (and siipple) Atella phalantha with one 

 wing each were well tasted and eaten, as also a Eurytela hiarhas (no wings) and a 

 wingless 6 Leuceronia thalassina. A wingless Neptis agatha was now tried, but 

 the bird only got one antenna in its mouth as in the case of Papilio eclierioides 

 yesterday and then drew back in apparent suspicion and refused it, but a Pyrartieis, 

 with one wing, was tasted and eaten most readily. The bird now tasted and 

 rejected a wingless Terias senegalensis and twice crushed and rejected with evident 

 dislike a Mycalesis campina with one wing, afterwards refusing persistently to 

 touch a Belenois mesentina, again with one hindwing attached. She readily, 

 however, accepted and ate a Pyrameis cai-dui, a Charaxes, a Precis cebrene, two 

 Precis cerytie, a Melanitis leda, and another Pyrameis, refusing the next Pyrameis, 

 probably owing to repletion. Except the Charaxes, which was wingless, each of 

 these last seven insects had one wing attached. 



[Assuming, what at the time seemed likely, that Charaxes candiope was eaten to 

 near repletion-point, P. archesis © f . is also likely to have been eaten fairly near 

 that point. P. dardanus S was at any rate probably not liked better than these 

 two. The 20-minutes' interval renders the placings of the later insects relatively to 

 C. candiope more uncertain, but the following preferences seem deducible from the 

 experiment as a whole : — 



?■- P. archesia preferred to N. agatha and N. saclava, E. dryope and \C. candiope, at 

 any rate not appreciably worse than the latter. 



C.florella and perhaps A. phalantha, E. hiarhas, and L. thalassina preferred to 

 N. saclava, and the Leuceronia at any rate was probably not liked less than 

 N. agatha. 



P. cardui, and perhaps Charaxes, P. cebrene, P. ceryne, M. leda each preferred to 

 T. senegalensis, M. campina, B. mesentina, N. saclava, and N. agatha, and, if the 

 Pyrameis was finally refused through repletion, probably in or near Grade 1. 



P. dardanus S was probably at any rate not preferred to P. archesia ; N. agatha 

 and N. saclava to G. candiope ; N. saclava. to E. dryope ; or N. agatha to L. tha- 

 lassina. 



The bird's recognition and refusal of Belenois may be just worth commenting on. 

 It will be remembered that the last — and that a single — example of this genus with 

 wing attached that the bird had seen was on May 12, nine days before (readily 

 eaten), and that previously to that she had seen none since April 10, when she 



