THE SPHINGO-MICROPTERYGIDES. 123 



to look upon the Lasiocampid ancestor as resembling the pro-legged 

 Megalopygid, rather than, in this particular, the more specialised 

 Eucleid stock. 



We have already referred to Dyar's statement that the Saturniicl 

 larva, by the union of tubercles iv and v into a sub-spiracular wart, 

 resembles the generalised members of this stirps. Packard says that 

 the larva of the Megalopygid, Laijoa crispata, is, in some respects, inter- 

 mediate between the Satumiidae (especially the higher Attacinae) and 

 the Cocldiopodidae (Eucleidae). The characters of the clypeus, the 

 setiferous tubercles, the distinct separation of the segments from each 

 other (i.e., the well-marked segmental incisions) are the points, 

 Packard says, in which the larva of Lar/oa approach the Attacinae 

 (Proc. Amer. Phil. Soc, xxxii., p. 292). He further states (Ibid., 1893, 

 p. 83) that the shape of the young larva? of Adoneta and Empretia, with 

 their large tubercles bearing three radiating setae or bristles, is such as 

 to remind one of the larva? of the Satumiidae, and to suggest one of two 

 alternatives : (1) The Cochliopodids (Megalopygids and Eucleids) have 

 originated from the Saturniids, or from forms allied to them. 

 (2) Both the Saturniids and Cochliopodids have descended from a 

 common stem form, The first alternative of deriving a generalised 

 from a specialised superfamily is out of the question. With the second, 

 of course, we agree, since we are treating them as members of the 

 same stirps. 



Packard further draws attention to the fact that in some of the 

 Saturniids the seta? are modified into urticating spines, as in the 

 Eucleids. This similarity, Dyar considers, does not imply relationship 

 since there is here only a similarity of function, whilst the structure 

 of the bases of the tubercles is essentially different. It is somewhat 

 remarkable also that in the Anthrocerids and Saturniids there is a 

 tendency for the tubercles to be arranged in a single transverse line on 

 each segment. In Anthrocerid larvse, tubercles i and ii coalesce ; in 

 those of the Saturniids, Dyar says that ii disappears after the first 

 moult, both developments (very different in actual value) ending in the 

 same result, the production of a single wart in the place of i and ii. 

 In both superfamilies iv and v are consolidated into a single wart. 



Another superfamily that has been placed in this stirps is the 

 Lacosomides. It was grouped by Comstock with the Saturniides, but 

 Dyar says that it belongs to the generalised Frenata?, and remarksf 

 that the Lacosomid larva? have retained a generalised condition on 

 account of their secluded life .... and present a case the converse of 

 that of the Eucleids, where specialisation has taken place in the larva 

 and where the generalised adult gives the best indication of the 

 relationship of these curious insects. He statesj later, that the Laco- 

 somids are in the same line of descent with the Saturniids. This super- 

 family, therefore, probably belongs to the stirps under consideration. 



The Eupterotides, which Hampson first separated from the Noto- 

 donts, are very closely allied to the Lasiocampids. Their larvse show 

 the generalised condition of tubercles iv and v, both being sub-spiracular 

 and ill-developed. They show no tendency to the conversion of iv 



* "Classification of certain Lepidopterous Larvse," Ann. Neic. York. Acad. 

 Sci., viii., p. 201. 



f " Classification of Lepidopterous Larvae," Ann. New York Acad. Sci., viii., p. 202. 

 { "Classification of Lepidopterous LarvaB," Trans. ]\ew York Acad. Sci., xiv., p. 51. 



