RECORD: S123. Wallace, A. R. 1866. Natural selection. Athenaeum No. 2040: 716-717.

REVISION HISTORY: Body text helpfully provided by Charles H. Smith from his Alfred Russel Wallace Page

NOTE: See a clipping of this item in Darwin's private papers here.

[page] 716

Natural Selection

At the last meeting of the Entomological Society, the subject of "Mimicry," or protective resemblances, was brought forward by Prof. Westwoodand myself, and during the discussion that ensued some objections were made to the explanation of the phenomena first given by Mr. Bates, and adopted by me as the only sufficient one. Time did not permit of an answer being made to those objections, and as they are at first sight very plausible, and were brought forward not by one of Mr. Darwin's opponents, but by a gentleman who fully admits the great principles of evolution and development in organic nature, they may probably have weight with some persons. Believing, however, that they are entirely unsound, may I beg a little of your space to give my reasons for rejecting them.

Mr. Sharp stated that four different causes might be sufficient to produce the phenomena of "mimicry" more or less completely, viz., first, accidental resemblances; secondly, similar conditions of life; thirdly, heredity, or reversion to a common ancestral type; and, fourthly, the preservation of useful variations.

To the first, or accidental resemblances, it was admitted that very few, if any, of the cases adduced by Mr. Bates or myself could be due. The last is the one we adopt. The second and third only remain, and these, Mr. Sharp argued, would account for most, or perhaps all, of our cases without the agency of natural selection at all. Now, all I can admit is, that in some cases of very closely allied species of the same or of closely allied genera, an accurate external resemblance, such as we term "mimicry," might possibly be produced either by "heredity," or by the action of like conditions. But in all the cases in which the insects resembling each other belong to distinct orders, or distinct families, or to genera not intimately allied, or even to well-marked sections of extensive genera, I entirely deny

[page] 717

that either or both of these causes could have produced the whole series of phenomena presented by mimicking insects, and for the following reasons, which appear to me sufficiently conclusive:—

1. In all cases of mimicry, the resemblance of the one species to another in a different group is entirely superficial, and is always strictly confined to those characters which cause the one to look like the other. The structure, the habits, the form of inconspicuous parts, the colour of inconspicuous parts, the nature of the food, or the character of the larva and pupa, are not, as far as we know, ever modified in a similar manner. But if such general causes as "heredity" or "similar conditions" produced resemblances, these resemblances should affect various parts of the organization, not those conspicuous to the eye only. The effect being limited with strict reference to external resemblance, seems to me a fatal objection to referring it to any cause or causes of a general nature.

2. There are no grounds for believing that minute details of colouration and marking are due to climatal conditions at all, still less that they can be produced so identically alike in species of groups widely differing in organization; neither is there any evidence that such details are ever continued by heredity to one species only in each of two distinct family groups which contain hundreds of other variously-coloured species.

3. It is only a very few groups of insects which are the subjects of imitation by many other groups. But "heredity" should affect nearly all groups not too remotely allied; and "common conditions" should affect all species inhabiting the same forests with some approach to an average frequency. The fact that there is no such miscellaneous character in the resemblances (the group of Danaioid butterflies being the mimicked in the great majority of cases) tells us plainly that no causes affecting all insects alike can be at the bottom of this curious phenomenon.

4. Protective resemblance to a species of a distinct order sometimes occurs, as in the curious Orthopterous insect adduced by Prof. Westwood, which had been always taken for a Coleopterous insect that inhabits the same country (Tricondyla sp.). Neither "heredity" nor "like conditions" can be called in here; yet the phenomenon is so similar to that of the mimicking butterflies, that the idea of a similar cause in both instances is irresistibly forced upon us.

5. Resemblances of the most perfect kind occur between insects and inanimate objects. Phasmidæ imitate sticks, leaves, or moss most wonderfully. The larvæ of Geometræ also imitate sticks. Thousands of tropical Coleoptera imitate bark (and it is always those that cling to bark); others that sit motionless on leaves cannot be distinguished from the dung of birds dropped on a leaf. These are most clearly protective imitations, and they can none of them possibly be produced by "heredity" or "similarity of conditions," but, if produced at all by natural causes, seem clearly due to the continued preservation of useful variations. The mimicry of other insects is equally protective, and there is every probability that both were produced in a similar manner.

6. This is rendered still more certain by the fact that in both classes of resemblance it is the female only that is most frequently protected, for reasons which I have already explained, but cannot now enter into. It is only the female "leaf-insect" that is so wonderfully like a leaf; and in many species of Pieris and Diadema it is the females only that mimic Heliconias and Euplæas. This fact alone renders it in the highest degree improbable that the two groups of phenomena should have been due to distinct causes, even if the preceding arguments had not shown us how impossible it was to explain any of the main features of "mimicry" by such causes as "heredity" or "the action of like external conditions."

For these reasons it appears to me indisputable that "natural selection," or the continued survival of variations useful to the possessor, is the only theory yet before us which is capable of explaining the whole of the facts presented by "mimicking insects."

This document has been accessed 2122 times

Return to homepage

Citation: John van Wyhe, ed. 2012-. Wallace Online. (

File last updated 26 September, 2012