FEEDING-HABITS OF FISH AND BIRDS. 59 



These considerations are of some importance, because it is difficult to 

 conceive Low mimicry can have had a beginning, except by a variation of 

 considerable magnitude ; further, it has been thought that birds would be 

 able to see through a disguise, unless it were good. However, the obser- 

 vations which have been recorded indicate that both birds and fish are easily 

 deceived, and that a crude resemblance would be likely to give a mimicking 

 insect some protection, and especially because birds often begin their attack 

 from such a distance that many details of pattern and coloration are 

 invisible. On the one hand, a bird may allow to pass at twenty feet a mimic 

 which is only a poor imitation of a protected species : on the other hand, at 

 two feet, a good imitation may be necessary for immunity from attack. It 

 would follow that, although a poor imitation may gain some protection, a 

 good mimic would gain more ;. so that from a rough resemblance, a good 

 resemblance could conceivably be built up by the agency of selection by 

 birds. There is ample evidence that butterflies, as a whole, are less palatable 

 to birds than many other groups of insects. An examination of their form, 

 pattern, coloration, and habits shows that they exhibit several characters 

 which must render them, relatively to other insects, conspicuous in Nature. 

 Nevertheless, if butterflies, as a whole, be considered to present some warning 

 characters, it obviously does not follow that, within the group, there may 

 not be degrees of palatability associated with variations in the amount of 

 warning coloration, and that therefore there may not be also within the 

 group mimicry. 



Summary. 



Attention is especially directed to two aspects of the relations between 

 preyer and preyed upon. 



First, that there are many factors which determine whether or not, at 

 any particular time, one animal will prey upon another, of which relative 

 palatability is only one. In order, therefore, to demonstrate a difference in 

 palatability between various foods, all these determining factors must be 

 taken into account. 



Second, that observations indicate that both birds and fish are deluded by 

 rough resemblances to the insects upon which they are at the time feeding. 



Bibliographic References.- 



(1) Marshall, G. A. K. : Trans. Ent. Soc. Lond., Sept. 1909, p. 329. 



(2) Swynnerton, C. J. M. : Proc. Ent. Soc. Lond. 1915, p. xxxii ; 



Ibis, 1916, vol. iv. p. 264, and 1917, vol. v. p. 529. 



(3) Pouxton, E. B. : Proc. Zool. Soc. 1887, p. 191. 



(4) Carpenter, — : Proc. Ent. Soc. Lond. 1915, pp. lxix-lxxii, lxxv. 



(5) Mottram, J. 0. : Proc. Zool. Soc. 1916, p. 383. 



