1887.] VALUE OF COLOUR AND MARKINGS IN INSECTS. 225 



insect-eaters. Again, only one species out of the 9, viz. Nomada 

 marshamella, always remained untouched by very hungry animals 

 when other food was withheld ; and we can only surmise as to what 

 would have been the results if the other 8 had been similarly tested. 

 It is, however, quite certain that many of them would have failed. 

 This is, indeed, proved by the following figures: — Out of the 37 

 cases 15 were exposed to this rigid test, i. e. the species in the columns 

 under my name in Tables I., II., and IV. (excluding S.fuciformis and 

 C. neustrid), and of the 15 only three remained uiitasted, and of 

 these two have been shown by Miss Cundell to be eaten under cer- 

 tain circumstances. 



Looking at all these figures, and especially the last, we can well 

 understand the following objection being urged against Wallace's sug- 

 gestion. It may he said, the tables, indeed, show that Wallace was 

 right in predicting that an unpleasant attribute would be found to be 

 associated with a conspicuous appearance. That has certainly been 

 proved by the results of a vast majority of the experiments ; but of 

 what value is this association when insects are seized, tasted, and 

 rejected in spite of the warning colours which, ex hypothesi, are 

 assumed to prevent this very contingency ? In the first place, an 

 answer to the objection is found in the very fact that the insects 

 were tasted and rejected to a much greater extent when the verte- 

 brates were thoroughly hungry, for we see that when other food is 

 present the conspicuous insects are, as a rule, untasted. We shall 

 presently see that an inconspicuous but nauseous insect is approached 

 by Lizards in a very different way from one which is conspicuous 

 and nauseous. There was, in fact, strong evidence in the details of 

 some of my experiments, that the vertebrate enemies were well 

 aware that the insects were distasteful, and yet, when exceedingly 

 hungry, did their best, in some cases successfully, to eat them. 

 This was especially seen in the behaviour of the Lizards towards 

 the larvse of O. potatoria, P. bucephula, and Croesus septentrionalis 

 in my experiments, and towards the larvse of G. neustria, A, gros- 

 sulariafa, and P. auriflua in Jenner Weir's experiments ; for in 

 all these cases the Lizards made repeated attempts to eat the larvae, 

 again and again rejecting their prey with every sign of disgust 

 (rubbing the mouth against the cage) ; and yet in the end the 

 larvae were reluctantly eaten. I believe, however, that the suspicion 

 with which conspicuous insects are approached results from the 

 strongly impressed experiences of early youth and not from a habit 

 which has become hereditary. In many cases, however, the warning 

 experience may have been gained without tasting the insects ; for 

 we have seen that the latter are often protected by smell, which 

 can be perceived from a distance. Excluding these instances, 

 however, the experience of conspicuous nauseous forms must have 

 been gained by actual trial of a large number. I hope to be able 

 to show that it is not necessarj' for the young insect-eating 

 Vertebrate to actually make trial of every species of unpleasant- 

 tasting insect in its locality, in order to be equipped with an effi- 

 cient stock of experiences with which to conduct its later life. 



