WARNING COLOEATION. 169 



and exliibited warning colours, or exhibited warning colours 

 without becoming distasteful, birds and other insect-eating 

 animals would be compelled to feed upon them ; and it has 

 been proved by Mr. Poulton and by others that if pressed by 

 hunger animals will eat such distasteful insects. 



It appears to me, however (as I have already said on p. 133), 

 that the same objection might be applied to other characters : 

 protectively-coloured caterpillars are suffered to escape their 

 enemies by their powers of concealment ; but if all, or the 

 majority of insects became thus difficult to find, some improve- 

 ment in their foes — additional keenness of sight or smell — 

 would tend to render these disguises easier to detect. 



Objections to the Current Theory of Warning Coloration. 



The larva of the Swallow-tail {Papilio Mac/taon) is coloured 

 with green, black, and orange ; it possesses, in Prof, Weismann's 

 opinion, "a striking appearance," and when offered by this 

 naturalist to Lacei-ta f/ridis, was untouched by the lizard. 

 This caterpillar becomes specially conspicuous when touched 

 or interfered with in any way ; it then everts from the neck a 

 glandular apparatus of a red colour. 



When at rest upon its food plant the colours are by no 

 means conspicuous. These invisible glands referred to have 

 been termed " osmateria " or " stench throwers " ; they secrete 

 an offensive-smelling substance ; the caterpillars are no doubt 

 to some extent protected by them. 



Mr. Wallace thinks that their function is chiefly to ward off 

 the attacks of ichneiimons ; but they may also, he considers, 

 frighten small birds. 



We may consider, perhaps, that this larva is inconspicuous ^ 

 but iv/ie)i found it possesses a means of defence in the osma- 



