574 PEOF. E. B. POTJLTON : NATUEAL SELECTION 



But wbicliever interpretation be ultimately adopted, tlie fact 

 remains the same — that the resemblances iu the Coleoptera are 

 of the same character as those in the Lepidoptera, and are, like 

 the latter, independent of affinity. They are readily to be 

 explained by the operation of a theory of selection, but present 

 the same difficulties as those presented by the Lepidoptera to an 

 interpretation by any other theory as yet brought forward. 



(6) The Besemhlances in question are not accomfanied hy any 

 changes in the direction of another Species except such as 

 assist in the production or strengthening of a Superficial 

 Likeness. 



This argument is fatal to any theory as yet advanced except 

 one based upon the principle of selection directed to a definite 

 end, viz. the production of resemblance. It is impossible to 

 explain why external forces or internal forces should thus act 

 upon a certain set of characters whose only relationship is that 

 they tend to produce a superficial likeness to another species — 

 that they should act upon these alone to the exclusion of all 

 other sets. No assistance can be obtained from the conclusion 

 that the results are recent and therefore superficial, and that a 

 resemblance in deeper characters will follow in time. In the 

 first place, the examples of more perfect (and presumably older) 

 resemblance show no more tendency towards approximation in 

 characters which do not help to produce likeness, than the 

 examples in which the resemblance is comparatively rude (and 

 presumably recent in origin). In the second place, deep-seated 

 parts of the organism are affected when the superficial resemblance 

 is thereby increased, but not otherwise. To take a single 

 example, the common British 'Long\covn. Clytiis arietis strongly 

 suggests the appearance of a wasp, partly because of its black and 

 yellow banding, but even more because of its alert and wasp-like 

 movements. This implies, of course, appropriate changes in its 

 nervous and muscular systems. 



There are many cases, like that of Clyfus, in which the changes 

 in deep-seated structures are of more importance than any- 

 thing else in determining the resemblance. I know of no more 

 striking example of this than the movements and attitudes of the 

 young (Lepidopterous) larvae of Endroviis versicolor, which 

 render them extremely like the larvae of saw-flies (Phytophagous 

 Hymenoptera). Numerous experiments have convinced me that 



