694 REPORT — 1886. 



4. Some Experiments iipon the Protection of Insects from their Unemics by 

 means of an- tmpleasant taste or smell. By E. B. Poulton, AI.A. 



When ■working up the historical side of this question my attention was directed 

 towards the necessity for further experiments. Wallace had predicted that bril- 

 liantly coloured aud conspicuous insects would be refused by the ordinary verte- 

 brate enemies of their class — that, in fact, the gaudy colouring acts as a warning 

 of the existence of something unpleasant about its possessor. Conversely, Wallace 

 argued that insects which were protectively coloured, resembling their surround- 

 ings, would be eaten when detected. It appeared that experiments (conducted by 

 Mr. Jeuner Weir aud others) yielded the most complete proof of the existence of 

 these sharp distinctions. But on thinking the whole subject over it seemed to me 

 that the acquisition of an unpleasant taste or smeU, together with a conspicuous 

 appearance, was so simple a mode of protection and yet, ej: hypothesi, so absolutely 

 complete, that it is remarkable that more species have not availed themselves of 

 this means of defence. What could be the principle which worked in antagonism 

 to this mode of protection? For in Wallace's theory no suggestion of a true 

 counterbalancing limit appeared, i.e., one which increased with the increasing 

 application of this method of defence until the latter was checked, or for the time 

 being rendered of no avail, or even turned into an absolute danger. But if a very 

 common insect, constituting the chief food of one or more vertebrates, gained an 

 unpleasant taste, the latter animals might be forced to devour the disagreeable 

 objects in order To avoid starvation. And the same thing might readily happen if 

 a scarce and hard-pressed form adopted the same line and so became dominant, 

 after ousting many species which were much eaten by vertebrates. If once the 

 vertebrate enemies were driven to eat such an insect in spite of the unpleasant 

 taste, they would certainly soon acquire a relish for what was previously disagree- 

 able, and then the insects would be in great danger of extermination if in the mean- 

 time they had become conspicuous by gaining warning colours. If this reasoning be 

 correct it follows that this mode of defence is not necessarily perfect, and that it 

 depends for its apparently complete success upon the existence of relatively abundant 

 palatable forms. In other words, its employment must be strictly limited. In order 

 to test my argument I determined to experiment with a view to ascertain whether 

 hunger would drive a vertebrate to eat an insect which was evidently unpleasant to it. 

 I obtained a few different species of Italian lizards and some tree-frogs, and very soon 

 found that I had reasoned correctly. The lizards would often refuse a conspicuous 

 insect at first, with all the signs of repugnance, but would afterwards make the 

 best of it, when they were not supplied with other food which they liked better. 

 I sometimes found this to be the case with species (e.g., larva P. iiiicephnlus) to 

 which other observers have ascribed the most complete immunitv. I should add 

 that it has been always recognised that an insect may be distasteful to one 

 vertebrate enemy but palatable to another, and to this extent Wallace admits a 

 limit to the application of his principle of defence. But the limit which I have 

 proved is of course entirely different, for I have shown that the vertebrate may be 

 forced to eat the insect, although tinpalatable to it. Although the latter limit is 

 thus quite distinct, it would certainly in time become identical with the former, for, 

 as I have argued above, the unpalatable forms when eaten would soon become 

 palatable. I have, in fact, shown how the limit which Wallace himself admits has 

 grown up, and how it may become a counterbalancing principle, working against 

 and perhaps reversing the principle which he was the first to point out as of 

 general application in insects. Some quite new modes of defence came out during 

 the inquiry. Thus size alone seems to act as a protection : a large moth (»S'. ligustri), 

 evidently palatable and quite harmless, was untouched by small lizards, but eaten 

 by larger ones. Again, it is probable that a species may acquire protection by 

 causing digestive troubles after being eaten, rather than by having an unpleasant 

 taste, and it may be that the former effect would leave an even more indelible 

 effect than the latter upon the memory of the captor. Some experiments with the 

 frogs rendered this conclusion probable (E. Jacohmm imago being used). 



Another result at first surprised me very much ; it was a limitation to the 



» 



