188 Dr. A. Jefferis Turner's Observations on 



characters within the various families have sometimes 

 considerable phylogenetic value, but little value in indi- 

 cating the true relationship between the families. To 

 attempt a classification of the Lepidoptera from the relative 

 position of the tubercles on the larval skin, as has been 

 done by Mr. Dyar, seems to me as profitless as it would 

 be to classify the whole order by the form of the palpi, 

 or the shape of the outlines of the wings in the perfect 

 insect. 



The pupal and oval characters advanced by Dr. Chapman 

 appear to me to stand on a better foundation. I am quite 

 ready to admit that forms with a greater number of mov- 

 able pupal rings, his Incompletae, are more primitive than 

 those with fewer movable rings, his Obtectae; but this docs 

 not carry us far. It seems probable also that the dis- 

 tinction between forms with vertical eggs having a central 

 micropyle from those with flat eggs in which the micropyle 

 is on the side may have important phylogenetic signifi- 

 cance. In the former group are the four families Arctiadae 

 (from which arose the Syntomidae), Nocluidae, Liparidae 

 (with which I associate the Hypsidae) and Notodontidae, 

 families which, it will be observed, I have associated on 

 neurational characters. The only other family with upright 

 eggs, excepting those usually known as Rhopalocera, whose 

 origin I am not at present prepared to discuss, are the 

 Castniadae, and these I regard as an isolated group. In 

 the Cossidae the oval characters appear to have remained 

 in a fluid state, for Dr. Chapman states (Trans. Ent. Soc. 

 1896, p. 579) that while Cossus has a vertical egg, that of 

 Zeuzera is flat. If this is so it is admissible' to assume 

 that the same held true of the Protocossidae, and that these 

 have given origin not only to the above-mentioned families 

 with vertical eggs, but to other families in which the eggs 

 are always flat. 



The neuration is by far the best guide to lepidopterous 

 phylogeny that w r e possess; it is a hidden structure pro- 

 tected from the direct action of outside influences to a 

 large extent, rarely affected by directly adaptational 

 changes, and, in short, the best field in which to search 

 for homogeny nnobscured by homopiasy. Yet even here 

 homoplastic influences are at work, and cautions interpre- 

 tation is necessary, for there is a widespread tendency in 

 many different families towards a simplification of the 

 neuration, which often proceeds along parallel lines. 



