66 Mr. E. B. Poulton's notes upo7i the colours 



entirely for the protection of the exposed winter pupse, 

 and, as has been shown, the predominance is only mani- 

 fested by that i^roportion of the second generation of 

 larvae (in the year), which will form the winter pupse. It 

 is strange that the imagines of E. omicronaria should 

 show distinct seasonal dimorphism, and yet that only a 

 small proportion should emerge as the summer form. 

 On the other hand E. yendularia (imago) shows in- 

 distinct seasonal dimorphism, and has two summer 

 broods. It is probable that there is a considerable 

 summer brood of E. omicronaria in more favourable 

 years. 



Phyletic parallelism in metamorphic species. — Prof. 

 Weismann's valuable essay on this subject (in the w^ork 

 previously referred to) proves that the different stages of 

 two great groups may not show an equal divergence or 

 affinity. He show^s that the grouping would in many 

 cases be entirely different according to the stage which 

 is selected as the criterion of divergence, and further, 

 that this varying divergence depends upon a corre- 

 sponding difference or agreement in conditions. He 

 proves these propositions from Diptera, Hymenoptera, 

 and^ Lepidoptera. "With regard to the latter he shows 

 that the division into butterflies and moths depends u23on 

 imaginal characters, while the larvae do not manifest 

 an equal divergence. It is not possible to arrange the 

 larvae of lihopalocera together in one great group as it 

 is ])0ssible to arrange the imagines. And this imaginal 

 divergence is accompanied by a difference of conditions, 

 for butterflies as a whole live under very different con- 

 ditions from moths as a ivhole. On the other hand, the 

 want of divergence between the corresponding groups of 

 larvae is accompanied by a similarity of conditions. (On 

 the other hand he points out that the larvae of PJiopalo- 

 cera can be divided into smaller groups corresponding 

 with imaginal classification). 



It seemed to me that it is worth adding a note about 

 the pupae to this most important comparison, for which 

 we are indebted to Professor Weismann. The pupae of 

 Pihopalocera can be formed into a large group corre- 

 sponding to the union of the imagines into one of the 

 two chief divisions of Lepidoptera. The characters by 

 which these pupa can be identified as a whole are bright, 

 or at any rate varied, colours ; angularity of outline, 



