G92 REPORT— 1897. 



18. Report on the Necessity for the Immediate Investigution oj the Biohgif 

 of Oceanic Islands. — See Reports, p. 352, 



SATURDAY, AUGUST 21. 

 The Section did not meet.. 



MONDAY, AUGUST 22. 

 The following Papers were read : — 



Protective Mimicry as Evidence for the Validiti/ of the Theory of Natural 

 Selection. By Edward B. Poultox, M.A., F.R.S., Hope Professor of 

 Zoology, Oxford. 



Several suggestions have been put forward to account for the superficial 

 resemblances between animals, especially insects, occupying the same geographical 

 area. It has been suggested, and indeed strongly maintained, that food, climate, 

 or some other chemical or physical influence of the locali+v may have supplied 

 the cause. On the other hand, many naturalists consider that the facts cannot be 

 interpreted by any of these suggested causes, and only receive an intelligible and 

 probalile explanation in the theory of natural selection. This theory supposes 

 that the resemblance is advantageous in the struggle for existence, the weaker 

 forms being shielded by their resemblance to the stroug and well-defended species 

 (mimicry of H. W. Bates), or the latter gaining by a resemblance which enables 

 their local enemies more easily — and thus with a smaller waste of life — to recognise 

 and avoid them (mimicry of Fritz Midler). The present paper directs attention 

 to certain facts commonly associated with mimetic resemblance which receive a 

 ready explanation upon the theory of natural selection as the efficient cause, but, 

 on the other hand, constitute a serious difficulty in the way of any other theories 

 as yet brought forward. 



Natural selection, as is well tnown, acts upon «?^^/ variations, whatever they 

 may be, which are in the advantageous direction, and are at the same time not 

 injurious in themselves. When the end to be gained (in this case the attainment 

 of a superficial resemblance) is common to a variety of distantly related species 

 possessing entirely different constitutional tendencies, we may feel confident that 

 an approach brought about by natural selection will be by extremely diverse paths 

 of variation. Under natural selection we might predict that such a common end 

 would be reached by great diversity of means, while under the other hypotheses 

 mentioned above a result of the kind is inexplicable. Hence the facts of the case 

 should act as a convenient test between these rival suggestions. 



First as to colour. We know but little of the chemical nature of the pigments 

 made use of in mimetic resemblance. One case, however, has been investigated 

 by Gowland Hopkins — viz., the bright tints by which certain S. American Pivrince 

 have come to resemble Heliconince and Ithomiince in the same locality. Gowland 

 Hopkins has shown that these close resemblances in colour and pattern are 

 produced by pigments which are characteristic of the PieritKP, and of an entirely 

 different chemical nature from those of their models. 



Another very interesting case is that of resemblance to ants. Ants arc 

 mimicked more or less closely by a great variety of insects and by spiders. In 



