LACHNEIDES. 459 



ritbi in one genus, and derives the whole from Cosmotriche, is not likely 

 to commend itself to earnest students. As a friendly criticism of our 

 own views, Dyar added (Ent. Rec, xi., pp. 141-2) an important 

 contribution to the subject. He writes : " The English species divide 

 into four phyla : 



Phylum A. — The larva is cylindrical, primary warts not altogether obscured, 

 secondary hairs simple ; no special structures. Moth with ordinary venation, 

 veins 6 and 7 of hindwings from the end of the cell, vein 8 forming a small inter- 

 costal cell at base by anastomosis with the discal cell — Jlalacoxoma. 



Phylum B. — Larva flattened, primary warts visible only as far as the largest 

 ones i and iii ; secondary hairs modified with white hairs subventrally : lateral 

 lappets and coloured thoracic bands. Moth with extraordinary venation, inter- 

 costal cell of hindwings greatly expanded, confluent with discal cell nearly to apex, 

 and finally forming a connection with vein 7 — Epicnaptera and Eutricha. 



Phylum C. — Larva degenerating, less flattened, but still with white tufted 

 hairs subventrally ; primary warts obscured, the other special structures lost. Moth 

 with the intercostal cell of hindwings less expanded, but carrying vein 7 back with 

 it from the apex of discal cell — Cosmotriche. 



Phylum D. — Larva again cylindrical, densely hairy, some of the secondary 

 hairs specially modified into irritating hairs ; white subventral hairs lost. Moth 

 with intercostal cell very small, but vein 7 arises from its apex with vein 8, and is 

 not connected with the discal cell — Macrothylacia, Ladocampa, Eriogaster, Achno- 

 campa (Trichiura), and Poccilocampa. 



Dyar then observes : ''Phylum A, I regard as distinctly the most 

 generalised. Phylum B, I was at first inclined to place much higher, 

 but the venation of the moth seems to preclude any other position. 

 Phylum C is closely allied, but the larva is decidedly less specialised. 

 I take this to be due to degeneration, rather than to generalisation, 

 since the venation of the genus, Cosmotriche, seems derivable from 

 Epicnaptera and not the reverse. Phylum D is the highest. A degenera- 

 tion from the flattened larva of phylum B having been once assumed, 

 it is easy to imagine this proceeding further to the round hairy larvae 

 of Easiocawjia and Macrothylacia. Eriogaster has the hairs less 

 developed, but evidently belongs here, and I presume that Aehnocawpa 

 (Trichiura) and Voccilocampa do also, though their larvae are not at 

 present before me. The venation is very singular. Vein 7 instead of 

 arising from the discal cell is quite separate from it, and arises from 

 the basal loop of vein 8. I think this is only explicable by supposing 

 a contraction of the large intercostal cell of phyla B and C, which has 

 carried vein 7 with it away from the discal cell. 



" So much for the British species. The accompanying tree (pi. vii), 

 in large part explains itself. I have added several European and 

 American genera, and one Australian genus, which are distinguished 

 from the English ones by the absence of shading on their respective 

 branches. Five special points may be noted : — 



Phylum E is the Australian Colussa. It is the most generalised larva of the 

 whole group, closely resembling that of Eustaudingeria. Indeed the genus scarcely 

 belongs here, as the moth still retains the frenulum. 



Phylum E is the most generalised phylum known of the true Laslocampids. 

 The females are wingless, but the male venation closely resembles Malacbsoma. 



Phylum a is a very curious form. The larva has much the structure of Mala- 

 cosoma, but the adaptation to 1he flattened form has begun, and the subventral 

 shadow is neutralised by a white hand, not by white hairs as higher in the scale. 

 The moth lias already the large intercostal cell of the hindwings. 



Phylum II is an American group, at first sight quite contradictory with the 

 larva of Eutricha or even more specialised, and the venation almost of Malacosoma. 

 I interpret the absence of the large intercostal cell to degeneration, by a process 

 different from that of phylum D. Here it seems that the lower border of the cell 



