154 MR. R. MELDOLA ON VARIABLE PROTECTIVE [Feb. 4, 



" survival of the fittest," it is perhaps scarcely necessary to add. 

 Notwithstanding this identity of origin, I would venture to suggest 

 the propriety of confining the application of the word " mimicry " 

 to such cases as those to which it was first applied by Mr. Bates — 

 to those, viz., in which the object imitated is animate ; while the 

 expression " protective resemblance " should be restricted to those 

 cases in which the object simulated is inanimate or part of a vege- 

 table structure. This distinction is, I am persuaded, well adapted 

 to prevent that confusion of ideas which is apt to arise when the 

 term " mimicry " is used in the sense in which it has been recently 

 used by Dr. Hagen, whose paper, designated " Mimicry in the Colour 

 of Insects"*, contains in reality no cases of mimicry at allf . The 

 distinction here enforced was adopted by Mr. A. R. Wallace in his 

 well-known essay published in the * Westminster Review ' for July 

 1867, but it appears to have been neglected by most subsequent 

 writers on the subject. 



Classification of the cases of protective resemblance. 



In every case of protective resemblance the disguised species 

 simulates some object in the environment; and as the object thus 

 imitated may be of a nature either constant or variable, we are ob- 

 viously provided with a means of classifying the cases of protective 

 resemblance, though but imperfectly, according to the stability of 

 the characters of the imitated object. As the result of an attempt 

 to arrange the cases of protective resemblance according to this 

 system, 1 have found it necessary to erect the four following classes, 

 which include, so far as I know, all the known cases, with the 

 exception of a certain small group which will be considered here- 

 after : — 



I. Cases in which both the characters of the imitated object and 

 the disguising characters of the species remain constant during the 

 lifetime of each individual. 



This class includes a very large proportion of the known cases of 

 protective resemblance, and passes by small gradations into Class II. 

 Most of those instances in which there is mere harmony of colour- 

 ing between a species and its environment belong to this class. 



II. Cases in which the imitated object varies within certain small 



* 'American Naturalist' for July 1872. 



t It will be well here again to insist upon the fact, previously insisted upon 

 by Mr. Wallace, that the word mimicry is not to be understood in the sense of 

 voluntary imitation. It is true that in many cases the mimicker copies its 

 model in mode of flight or other habits ; but even here the imitation cannot 

 be considered voluntary, since such modification of habit has most probably 

 arisen by the natural selection, through many generations, of individuals whose 

 manner of flight resembled in any way the manner of flight of the imitated 

 species — just in the same way as this agency, selecting through many genera- 

 tions those individuals whose colour, form, pattern, &c. approached in any way 

 the colour, form, or pattern of the imitated species, lias at length brought about 

 the close resemblance in external characters which we now behold : in other 

 words, along with the structural there has gone on a psychical mimetic adap- 

 tation. 



