204 uAnnals Entomological Society of America [Vol. II, 



The study of Mimicry is of the highest value in relation to both 

 •evolution itself and the motive causes of evolution. 



Apart from all question of the means by which ]\Iimicry has 

 been produced, it will be generally admitted that the mimetic 

 species has in some way evolved a superficial resemblance to the 

 pattern of one or more species, more or less remote from it in the 

 scale of classification. Looking on the changes by w^hich the 

 resemblance has been produced as a piece of evolutionary history, 

 and, as I have said, disregarding for the moment their causes, we 

 have one of the very simplest and sharpest pictures of organic 

 transformation presented for our investigation. An effect — 

 generally a strongly marked and conspicuous effect — has been 

 brought about in those elements which make up the superficial 

 appearance of a species, and this important change is manifestly 

 in the direction of only a minute fraction of the infinitely complex 

 organic environment, viz., that fraction contributed by the super- 

 ficial appearance of one or more very different species, commonly 

 indeed of but a single one. When, as in North America, a recent 

 invader becomes the model determining the direction of evolu- 

 tion in some constituent of the ancient butterfly fauna, the case 

 becomes especially striking. 



The effects produced on the mimic are generally sharper 

 and more distinct than those seen in the concealing resem- 

 blances to bark, lichen, earth, &c., — the difference corres- 

 ponding to the more definite and individual appearance usu- 

 ally presented by the pattern of the model as compared with such 

 elements in the vegetable and mineral surroundings. There are 

 also other important differences. The models of Mimicry are 

 generally more restricted in their range, and differ more widely 

 in different areas and in different parts of the same area than the 

 models of ciyptic resemblance. Dift'erences between the local 

 forms of the same model imply that the mimicked species has 

 itself been subject to rapid change, while the models of cryptic 

 resemblance appear by comparison to be stereotyped and per- 

 manent. Furthermore, the models as well as their mimics within 

 the same area are liable to changes of distribution, whereas the 

 models of cryptic resemblance are as a rule comparatively fixed. 

 A mimetic species may often be found passing into an area where 

 its model exists in a different form or does not exist at all, and 

 highly instiTictive conclusions may be drawn from the study of 

 the corresponding changes. 



