( xlii ) 



as Bugalla and Damba, large in area and the latter separated 

 by twenty-five miles from the mainland ? And even if 

 a certain amount of interchange should take place (as is 

 likely), or the mainly local birds that form the island parties 

 be reinforced for a few months in the year by migratory 

 species (as they will), will not the new arrivals very rapidly dis- 

 cover (as the result of a fact I have referred to under " Alleged 

 indiscriminateness ") that the " Planemas " they attack when 

 hungry enough for Acraeinae are in this place mostly Pseud- 

 acraeas and acceptable therefore (to rely on my experimental 

 results with P. lucretia and P. trimeni) right away up to 

 A^e^/?s-ref using point ? 



In relation to any bird that has made this discovery the 

 fact that it wears the livery of a Planema will no longer be 

 of the full use to a Bugalla Pseudacraea that it probably was 

 to those of the mainland, and the very interesting phenomenon 

 described by Dr. Carpenter might be expected to come about 

 in the course of time, other things being equal, if the Selec- 

 tionist view be correct. I fully endorse Col. Manders' warn- 

 ing as to the necessity for watchfulness in respect of the 

 possible occurrence of complicating or alternative factors. 



Replying to the question, ''Are birds deceived by resemblances 

 in their prey ? " 1 may say that I have tested this in very 

 numerous and varied experiments, and that the answer is 

 in the affirmative. I have some evidence even in the case 

 of wild birds. As for the experimental tastings of young 

 birds, I have obtained plenty of evidence that a vastly more 

 powerful potential factor is at work. 



The visual memory of birds proved to be inadequate to the 

 task it has to perform, and constant mistaken attacks followed 

 by rejection were found to take place — though less so, on the 

 whole, in relation to prey that had been frequently and recently 

 experienced. It appeared that no bird could be too old or 

 too experienced to make continual mistakes of this kind. 



The factor was shown to be probably an extraordinarily 

 powerful one and far more than capable, from this point of 

 view, of replacing that suggested by Fritz Miiller. In fact, 

 it is doubtful whether the experimental tastings of young 

 birds contribute, relatively speaking, specially heavily towards 



