( ^1 ) 



formed the subject of much experiment by well-kuown inves- 

 tigators, and all are agreed that insect-eating animals refuse 

 those larvae which possess conspicuous coloration. 



Professor Packard next proceeds to emphasize the paucity 

 of evidence on which rests the theory that insectivorous birds 

 are the principal agents of natural selection in Lepidoptera. 

 Here 1 feel myself bound to agree that the recorded instances 

 of butterflies forming the staple food of birds are few and far 

 between. It would at first sight appear that if birds are really 

 the principal cause of modification in the markings of butter- 

 flies' wings there ought to be no necessity for the supporters 

 of the Bates-Muller theories to be obliged to seek out instances 

 of birds eating butterflies. It should be generally recognized 

 that birds do actually destroy butterflies in large numbers, but 

 on the other hand most experienced collectors seem agreed 

 that the sight of a bird even chasing a butterfly is of rare 

 occurrence. One point, however, is frequently overlooked 

 alike by the supporters and opponents of the Bates-Muller 

 theories, viz : — that it is not to be expected that any of the 

 full-grown and experienced birds will be frequently observed 

 catching butterflies, since they have already learned to avoid 

 the distasteful species and incidentally their imitators. Only 

 the young birds will attack and experiment with distasteful 

 and edible kinds alike, and since on the one hand, young birds 

 are notably retiring and inconspicuous and furthermore will 

 speedily learn that butterflies are unprofitable prey, it is not 

 to be wondered at that birds are not universally observed to 

 be butterfly eaters. 



Many common occurrences in nature are exceedingly diffi- 

 cult to observe, and it would seem at least possible that the 

 destruction of butterflies by young birds is carried on to a far 

 greater extent than would appear from casual observation. 

 The evidence adduced by Professor Packard is to some extent 

 contradictor}^, and whilst some of the observers assert that 

 they have never seen butterflies attacked by birds, others 

 again mention numeious cases of this form of destruction. 

 Instances are given by butterflies being eaten by birds in 

 North America, India, Ceylon, Burmah, Natal, etc., but it 

 must be acknowledged that the evidence on this head, and 



