236 MR. C. F. M. SWYNNEETON ON THE 



but suddenly (as on previous occasions) thinking better of it, withdrew and thence- 

 forth simply ignored it. She then refused 2 -^crcea areca with a sharp shake of 

 the head. 



[This Pseudacrcea was, I noted, distinctly more like Acrcea acm^a than that used 

 in experiment 12, having a whitish area on the upper surface of each hindwing 

 and a particularly strongly-marked subapical fulvous patch to the forewing. The 

 bird, nevertheless, had seemed to discriminate between it and a very similarly- 

 coloured specimen of Acrcea acara, possibly partly as a result of the difierence 

 in size. She may, of course, have distinguished it equally from Acrcea areca $ . 

 At any rate, its general likeness to one or both Acrseas was probably sufficient to 

 deter the bird, on second thoughts, from tasting it. 



The last time the roller had seen either of these Acrjeas or any other butterfly 

 with at all similar coloration was exactly fifteen days before, and before that 

 again she had seen nothing at all resembling them for at least twenty-five days — 

 that is to say, since March 7, the day on which she was captured. On April 1 she 

 had tasted and rejected each, apparently distinguishing one from the other so far 

 as appearance was concerned. 



It is a pity that the roller's state of appetite was not ascertained here, as without 

 that the suspicion must necessarily remain that she may have refused the insects 

 through sheer repletion.] 



Exp. 18. — Apr. 28. Coracias gcorrulus ate readily four Pyrcmieis carclui a.nd 

 refused the next, but five minutes later accepted equally readily and ate with 

 evident relish two more and refused a third, and, five minutes later, again five 

 more and again refused. All had wings. 



[The refusals were probably due to the bird's having reached the repletion-stage 

 for Pyrameis. Whether this was actual, absolute repletion-stage or not is unfor- 

 tunately not indicated by the experiment as recorded.] 



Exp. 19. — May 12. Crushed well and readily swallowed a Belenois mesentinct (an 

 individual already tasted and rejected by Phyllostrophihs jiavistriatus and Emheriza 

 orientalis and refused without tasting by Lanius lateralis). No sign of dislike was 

 shown, but a 6 Mylothris agaihina with decided sweet-brier smell had barely been 

 tasted when it was emphatically flung away. 



[It seems not unlikely that the Mylothris was found distinctly more unpleasant 

 than the Belenois, though this conclusion must be regarded as based purely on 

 the great contrast in manner. It would have been a safer conclusion had the 

 Belenois been eaten after the Mylothris refusal.] 



The conclusions to the first thirteen preference experiments on C. garrulus, B, 

 were worked out at a date when I had not begun fully to appreciate the influence 

 of " the preceding ofiering." This will have made so little difierence, however, 

 that, being pressed for time, I have not thought it worth while to recast them. 

 Perhaps even the otherwise rather unnecessary detail in which they are stated is 

 useful in emphasizing the vast difference between a low-grade insect and one that 

 is eaten to repletion-point. Contrast, for example, Melanitis leda and Mylothris 

 yulei in experiment 27, 



