50 Mr. E. B. Poulton's fiotes upon the colours 



But this is very different from the special resemblance 

 of S. ocellatus, and the former larva would be very im- 

 perfectly protected were it not for the additional general 

 resemblance. 



The number of larvge protected by general resemblance 

 is very large. A very small class (comparatively) is 

 protected by taste or smell, a still smaller class by the 

 possession of terrifying markings, structures, attitudes, or 

 movements. Those specially protected form a large class, 

 and I imagine that all larvaB unprotected in one or more 

 of these three ways are protected by general resemblance. 

 The latter will, 1 believe, prove to include by far the 

 largest number of instances. In many cases a larva 

 may be specially protected upon one food-plant, and 

 generalli/ upon others. So also a larva may have been 

 very specially protected upon its original food-plant, 

 which may be now unknown. 



These same terms also apply to other cases of pro- 

 tection, such as the mimicry of distasteful forms, or of 

 forms otherwise protected (special protective mimicry). 

 The same terms also apply to all organisms which avail 

 themselves of protective shapes, colours, attitudes, &c. 



The significance of larval (and pupal) dimorphism. — 

 Professor Weismann has proved that in many cases 

 dimorphism is simply a phase of transition into mono- 

 morphism of a different kind from that which the species 

 assumed before the commencement of changes which 

 led to dimorphism. The whole transition from the first 

 monomorpMsm is, he says, first variahility. ; then poly- 

 Diojphisiii, produced by the comparative permanence of 

 the favourable varieties ; then dimorphism, by the pre- 

 dominance of the two most favourable forms ; finally 

 monomorphism, by the ultimate permanence of the one 

 most favourable form. One of the forms in the stage 

 of dimorphism is the old monomorphism, and the other 

 is that which will become the new monomorphism. This 

 theory is proved for manj' larvae, but I believe that there 

 are instances in which such an explanation does not 

 hold. It is indeed probable that there are several 

 explanations for as many forms of dimorphism. I will 

 now allude to one instance, and will show some grounds 

 for not accepting the above explanation of its cause, 

 afterwards attempting to account for it in another way. 

 The larvse of some of the Ephyridce, after the last 

 ecdysis, are dimorphic, appearing in the two usual 



