224 MR. C. F. M. SWYNNERTON ON THE 



one the reality and probable importance of the defence that is known as 

 mnpalatability. 



To illustrate, however, my remarks on "edibility" (and incidentally to 

 supply the reply to a question raised by Mr. McAtee), I may say that m}^ 

 shrikes not only ate when sufficiently huDgry such very low-grade insects as 

 Mylahris and Zonoeerus (often, it is true, with many signs of dislike), but 

 would even, when insufficiently hungry for them, impale them sometimes 

 for future use. If one were taken off the spike and re-offered they would 

 reject or re-impale it, and yet immediately after doing so they would, given 

 the chance, make a large meal off certain other insects. If they were now 

 offered one of these when becoming too replete for it and no spike were vacant, 

 they would often remove and throw away the Zonoeerus or Mylahris and 

 replace it by the higher-grade insect, and this in turn might be replaced 

 later by a still more u elcome species. I have never seen the order reversed. 

 My cat, even when hungry, would similarly continue for some time to play 

 with or often re-visit a "nauseous " bird or small mammal for which he had 

 no present use and would eventually perhaps eat it — his appetite seeming to 

 be gradually stimulated by the playing with it. But he would in the mean- 

 time at once relinquish it if offered more acceptable prey and would make an 

 immediate meal off the latter. The last remark also applies, roughly, to my 

 owl. She, too, would otherwise hold for a long time prey too low-grade to be 

 of use to her in her then state of appetite, thus seeming to show, like the 

 shrikes and the cat, that she realized that it would become perfectly edible 

 under certain gastric conditions that had yet to arrive. Not that either birds 

 of prey or butcher-birds are continually killing animals that they are not at 

 the moment hungry enough for, however provident they may sometimes 

 show themselves over them when they have killed them. So at least I would 

 conclude from the very numerous refusals that I have witnessed on the part 

 of hawks and Lanius in the wild state and from the fact that with my captive 

 birds of the latter genus rejections were far more numerous than impalements. 



Whether the term " unpalatability " sufficiently describes such protection 

 as is possessed by an Acrcea or Zonoeerus I will discuss fully in finally 

 summing up the results of my experiments. I have, I think, a good deal of 

 evidence in favour of the view that smells and tastes, like appearance, are of 

 use in averting the eating of their possessor mainly when they have become 

 associated in the enemy's mind with unpleasantness of a more fundamental 

 character. This is probably, in most cases, indigestibility, greater or less, 

 conquerable by a greater or less degree of gastric activity. 



Possible errors and past mistakes : The working out (and in most cases 

 re-casting) of the conclusions from the different preference experiments has, 

 owing to pressure of other work, been spread over a very considerable period 

 and has at all times been greatly interrupted. It is very possible, therefore, 

 that, iu spite of care, mistakes may be present. Should any such be detected 



