68 NATURE SKETCHES IN TEMPERATE AMERICA 



such as green or decayed leaves, dead twigs, bits of lichen, 

 flowers, spines, excrement of birds, and living insects; . . . 

 The resemblance is often wonderfully close and is not confined 

 to color, but extends to form and even to the manner in which 

 the insects hold themselves. The caterpillars which project 

 motionless like dead twigs from the bushes on which they feed 

 offer an excellent instance of a resemblance of this kind. The 

 cases of the imitation of such objects as the excrement of birds 

 are rare and exceptional. . . . But in all the foregoing cases 

 the insects in their original state no doubt presented some 

 rude and accidental resemblance to an object commonly found 

 in the stations frequented by them. Xor is this at all improb- 

 able, considering the almost infinite number of surrounding 

 objects and the diversity in form and color of the hosts of 

 insects which exist. As .some rude resemblance is necessary 

 for the first start, we can understand how it is that the large 

 and higher animals do not (with the cxcej)tion, as far as I 

 know, of one fish) resemble for the sake of protection special 

 objects, but only the surfaces which commonly surround 

 them, and this chiefly in color. Assuming that an insect 

 originally happened to reseml)le in some degree a dead twig, or 

 a decayed leaf, and that it varied slightly in many ways, then 

 all the variations which rendered the insect at all more like 

 any such ()l)ject and thus favored its escaj)e would be preserved, 

 while other variations would be neglected and ultimately 

 lost; or if they rendered tlu' insect at all less like the imitated 

 object they would be elimiuatod." 



Wallace placed both protective resemblance and mimicry 

 in the same category, claiming l)oth to have resulted from 

 natural selection. But there are a certain number of English 

 naturalists, including Poulton, who have elaborated on this 

 subject and have contributed a quantity of very interesting 

 data in su])port of the desirability of drawing a distinction 

 between protective resemblance and mimicry. The supposed 

 difference between them is that in mimicry one animal mimics 

 another living animal, in contradistinction to an animal that 

 resembles the surrounding medium in which it lives. Of late, 

 moreover, there has arisen another group of physiological ex- 

 perimenters who place the coloration of animals on a chem- 



