244 DARWINISM chap. 



the former and reject the hitter. The Pieridse would, however, 

 usually he less numerous, hecause their larvae are often pro- 

 tectively coloured and therefore edi1)le, while the larvae of the 

 Heliconidae are adorned with warning colours, spines, or 

 tubercles, and are uneatable. It seems probable that the 

 larvae and pupae of the Heliconidae were the first to acquire 

 the protective distastefulness, both because in this stage they 

 are more defenceless and more liable to fatal injury, and also 

 because Ave now find many instances in which the larvae are 

 distasteful Avhile the perfect insects are eatal^le, but I believe 

 none in which the reverse is the case. The larvae of the 

 Pieridae are now beginning to acquire offensive juices, but 

 have not yet obtained the corresponding conspicuous colours ; 

 while the perfect insects remain eatable, except perhaps in 

 some Eastern groups, the under sides of Avhose Avings are 

 brilliantly coloured although this is the part Avhich is exposed 

 when at rest. 



It is clear that if a large majority of the larvae of Lepido- 

 ptera, as well as the perfect insects, acquired these distaste- 

 ful properties, so as seriously to diminish the food supply of 

 insecti Amorous and nestling birds, these latter Avould be forced 

 by necessity to acquire corresponding tastes, and to eat Avith 

 pleasure AAdiat some of them noAV eat only under pressure of 

 hunger ; and A^ariation and natural selection Avould soon bring 

 about this change. 



Many Avriters haA^e denied the possibility of such Avonderful 

 resemblances being produced by the accumulation of fortuitous 

 variations, but if the reader Avill call to mind the large amount 

 of variability that has been shoAvn to exist in all organisms, 

 the exceptional poAver of rapid increase possessed by insects, and 

 the tremendous struggle for existence always going on, the 

 difficulty Avill A^anish, especially AA^hen Ave remember that 

 nature has the same fundamental groundAA^ork to act upon in 

 the tAvo groups, general similarity of forms, AAdngs of similar 

 texture and outline, and probably some original similarity of 

 colour and marking. Yet there is evidently considerable 

 difficulty in the process, or AAdth these great resources at her 

 command nature would have produced more of these mimicking 

 forms than she has done. One reason of this deficiency prob- 

 ably is, that the imitators, being ahvays feAver in number, have 



