Phylogeny and Evolution of the Le'pidoptera. 575 



very precise theories. I am less puzzled by the question 

 of frenuin and jugum. The jugum is certainly a remains 

 of a wing lobe well- developed in many Neuroptera, and 

 appears to have no such function as is attributed to it of 

 combining the wings in flight, whilst there are distinct 

 traces of a commencing frenulum in some stronger hairs 

 in Eriocephalids. 



With regard to the Macro-lepidoptera, I have not been 

 able to find any definite pupal characters on which they 

 may be classified, still less distributed in accordance to 

 their phylogeny. A closer examination of a larger series 

 of species may throw more light on the matter, but this 

 is very doubtful, since, for example, I find in the Geome- 

 trides, pupae as thoroughly obtect as the highest Noctuas, 

 whilst in some groups there are pupge with dorsal head 

 plates, and in some instances these even carry the 

 eye-covers, a very parallel condition to that in the 

 Skippers. Dorsal head plates also occur in Lasiocampids, 

 Thyatyrids, Notodontids and some others. Obviously 

 these remains of earlier structure appear in the lower 

 families of each division. 



It occurred to me to recur to some old ideas I had 

 about the eggs of Lepidoptera, in hope of finding some 

 solution of this difficulty. 



It seems very probable that the eggs of Lepidoptera 

 should give characters useful for classification, not only 

 for the simple reason that every structure and habit does 

 so, but because it is not merely a structure, but the whole 

 individual at one stage of its existence, and because the 

 circumstances of its environment being different from 

 those of the later stages, it cannot have responded in the 

 same manner as they have, and may, therefore, afibrd us 

 differences and resemblances when they do not do so, or 

 have them overlaid by various characters of more modern 

 acquirement and less fundamental significance. Especially 

 being the most ancestral stage, it may probably have 

 some earlier characteristics, in spite of having varied in 

 many particulars like the other stages. 



We all know that in many genera the eggs of the 

 different species are easily recognisable, as in Vanessa, 

 Pieris, Ennomos, Acronycta, Cerura, &c. It is less easy 

 to find characters from eggs of families, yet most of us 

 can recognise a Noctua or a Geometra egg, one of a butter- 

 fly or of a sphinge. 



