EXPLANATION OF FOEM AND COLOURING. 235 



I now continued to offer it, four times in all, at short intervals, but the bird refused 

 to have anything to do with it — in fact, simply ignored it. On the fifth occasion I 

 offered Eypolimnas ivahlbergi instead. The bird appeared at once to recognize 

 that there was a difference, and, as in the case of the A. acara of the other 

 day, leant forward to take it. At the last moment, however, she thought better 

 of it and withdrew, without having actually touched it, and, though I continued 

 to hold it to her, would take no further notice of it. Papilio dardanus $ f. hippo- 

 coon was now offered, but simply ignored. 



In each case I had first fully displayed the colours of the butterfly just outside 

 the bars before actually offering it. I had then held it through the bars by two 

 wings, in the forceps, just in front of the bird, so as to allow it to flap. 



She now ate with the greatest readiness six Pyrameis ccnxlui, each with one 

 wing attached, and, after the usual probably futile five minutes' interval for 

 possible after-effects, a seventh ; then accepted, crushed well, and ate with 

 apparent relish the Hypolimnas loahlbergi with all wings removed, and after it 

 the Papilio dardanus 5 f. hippocoon similarly stripped. 



Probably both these butterflies, as also P. cardtd, were preferred to the 

 Amauris. 



[I apparently^felt little doubt at the time that the bii'd at once detected the 

 difference between Amauris ochlea and Hypolimnas toahlhergi. That she should 

 have hesitated, nevertheless, to attack the latter might indicate (1) that so far as 

 she was concerned Amauris ochlea is to some extent capable, even at a close view, 

 of acting as protector of the mimics of A. dominicanus, (2) that the bird had 

 already met with A. dominicaiius and remembered it in time to refrain from 

 attacking the Hypolimnas (and Papilio), (3) a general suspicion of large black and 

 white butterflies, or (4) repletion to the point of not caring to swallow such large- 

 winged insects. She has not usually hesitated to swallow even large species of 

 Charaxes, wings and all, up to practically repletion-point, so that this fourth 

 explanation does not seem likely to be the correct one, especially as she went on 

 to eat with readiness and relish six Pyrameis with one wing and a H. wahlhergi 

 and Papilio with none.] 



Exp. 1%.—Apr. 10. Had not long before retired to her perch after feeding on 

 migratory locusts, etc. She, nevertheless, accepted and ate with apparent readiness 

 and in quick succession, each after crushing it, a Neptis agatha, a cJ Hypolimnas 

 misippus, two Belenois mesentina, three Belenois severina, a Mycalesis campina, 

 a Precis natalensis 0, and a S Leuceronia thalassina. 



[This is quite a good instance of the fact that one should never assume a bird to 

 be replete, or nearly so, unless one has actually fed it up to repletion by hand. 



Several of the above butterflies have since proved to be acceptable to her by 

 no means up to repletion -point.] 



Exp. 17. — Apr. 15. A Mimicry Experiment. — Coracias garrulus persistently 

 refused Acrcea acara without tasting and with emphatic shakes of the head. 

 On my now offering Pseudacrcea trim.eoii, a specimen that, except in size, was 

 distinctly more like acara than areca, the bird at once leaned forward to take it, 



LINN. JOURN. — ZOOLOGY, VOL. XXXIII. 18 



