226 MR. E. B. I'OULTON ON THE PUOTECTIVE [Miir. 1, 



Such an education would be somewhat dearly bought ; it would be 

 unpleasant to the insect-eater and destructive to the insect. 

 But if, as I shall endeavour to show, there is a superficial resem- 

 blance between the colours employed by very different insects, 

 and frequently even a similarity of pattern, we spe that a com- 

 paratively few unpleasant experiences would be sufficient to create 

 a prejudice against any insect with colours and patterns at all 

 resembling the nauseous forms which have already produced so 

 indelible an impression upon the memory. And thus it is most 

 probable that the conspicuous appearance which astonishes one 

 sense becomes associated in the mind of the Vertebrate insect-eater 

 with the well-remembered effect of other qualities upon other senses. 

 Different Vertebrates vary much in their rates of education. Thus 

 my Frogs were much more stupid in this respect than the Lizards ; 

 but then the imperfect memory or limited power of generalization in 

 Frogs is less fatal to insects than it would be in the case of the other 

 Vertebrates ; for I do not think that the larvee were ever injured in 

 the least after having been tasted by these animals. It is therefore 

 probable that the gradual development of warning colours by natural 

 selection was due to the fatalities which followed the experimental 

 tasting of other Vertebrate enemies (especially Birds and Lizards), 

 which inflict incidental injuries during the process of tasting. But 

 the warning appearance having been acquired by such means, the 

 Frogs have certainly taken the opportunity (thus offered to them 

 read^'-made and without having themselves contributed towards its 

 existence) to acquire a somewhat limited education. This was seen in 

 the case of the queen wasp (see Appendix II.), which on being placed 

 in the case was tasted by three Frogs out of twelve, but afterwards 

 was untouched for rr.any hours (as far as I was able to observe). A 

 proof of the limited extent of the education is given by Butler, who 

 speaks as if his Frogs repeatedly tried to eat the two species of 

 nauseous larvae (A. grossulariata and H. wavaria), seeming only to 

 become suspicious when thej' had actually made a trial of the insects 

 on each separate occasion. It also seemed to me that my Frogs 

 generally, if not always, ate bees from want of memory or deficient 

 discrimination; for in nearly all cases they were finally rejected. 

 But the experience did not seem to make any difference to the readi- 

 ness with which the next bee would be seized and again rejected. 

 On the other hand I did not see a ladybird tasted on any occasion. 

 The slight power of discrimination possessed by Frogs was also 

 shown by the fact that they frequently jumped at and seized the 

 dark-coloured ends of the forceps with which I used to introduce 

 insects into their case. 



For the tolerably complete demonstration of the principle which I 

 believe has been at work, a far larger number of observations are 

 necessary, while complete confirmation requires experimental evidence 

 with young Vertebrates which have been reared in confinement, so that 

 the whole c<f their education is under observation. As conducing 

 towards this end, I publish the suggestion with its foundation on the 

 resemblances indicated by the tables given below, which have this 



