LACHNEIDES. 461 



for it has retained many more ancestral characters common to the 

 Eutrichid and Lachneid (in sensu strict.) branches, than has Eutricha. 

 Again Dyar's separation of the phyla 15 and J is scarcely justified on 

 the grounds given, for so great is the resemblance between the larvae of 

 Eutricha (quercifolia) and Odonestis (pruni) up to the hybernating stages, 

 that one doubts whether they can belong to different genera. Our 

 interpretation of the neuration appears to bring the phylogeny as 

 shown by this character into harmony with that presented by the 

 structural features offered by the larvae and pupae. 



We conclude from the facts presented by newly-hatched larvae of 

 the Lasiocampidi and Eutrichidi, that both have specialised almost to 

 the same extent on their respective stems, having i, ii, iii, and iv + v 

 large many-haired warts when they leave the egg, the earlier 

 generalised stages evidently having been pushed back into the egg ; 

 but Bacot considers that Eutricha is, as a larva, rather more specialised 

 than Lasiocampa. The ancestral Lachneid larva appears to have had 

 generalised tubercles, i and ii trapezoidal, iii supraspiracular, iv and v 

 subspiracular (iv tending to be somewhat posterior), vi below these, 

 and vii marginal at base of legs. These carried setae with chitinous 

 bases. Development proceeded by the formation of warts (somewhat 

 as in Anthrocerids) each individual seta having a chitinous base, i, ii, 

 iii, iv and v being especially well-developed. In some genera later 

 necessities for the perfection of protective resemblances necessitated 

 the flattening of the warts, so as not to disturb the larval contour, and 

 these ultimately merged in the general skin clothing, driving back the 

 preceding evolutionary stages into the egg. Thus E. quercifolia, D. 

 jrini, L. qaercus, and P. trifolii hatch with i, ii, iii and iv + v well- 

 develope:!. In the reduction of the warts, i, ii and iii, as most disturbing 

 the outline, were the first and most completely eliminated. A small 

 ii and iii may, therefore, suggest either a generalised or highly 

 specialised form, and care must be taken in forming any conclusion 

 based on the character of these tubercles. The small size of ii compared 

 with i, is, of course, not necessarily on the same line of development 

 as the flattening of the warts and diffusion of the setae of the dorsal 

 and supraspiracular tubercles. The tendency to lose i is already well- 

 marked in some cases in which the tubercles are quite ancestral and 

 scarcely wart-like in appearance, e.y., Malacosoma (castrensis) , &c. The 

 wart-like cushions, seen in the larvae of Eutricha (quercifolia), Lasio- 

 campa (querciis), &c, are probably a further development of the 

 Malacosoma form. In the flattening and diffusion of the warts, these 

 processes affect i and ii simultaneously, there being a general tendency 

 to the loss of all dorsal and subdorsal warts in certain genera. The 

 newly-hatched larva 1 of Eutricha and Lasiocampa are quite a moult in 

 advance of Pachyyastria and distinctly in advance of Dendrolimus. It 

 is necessary then to clearly recognise that the tendency to have ii weak 

 is a distinctly earlier character than the tendency to the flattening 

 and dispersion of all the dorsal and subdorsal warts. 



The pupae have undergone but slight modification in the group, 

 but the imagines appear also to have specially developed along different 

 lines, Eutricha with a complicated branching of the costal nervure of 

 the hindwing, Lasiocampa in the strong costal and median nervures 

 which enable it to maintain its rapid flight. On the other hand, the 

 females of Eutricha, Macrothylacia and Lasiocampa, are very specialised 



