EUTRICHIDiE. 113 



intermediate, Cosmotrichid in general character, but tending to the 

 stumpiness of those of cajani and marginepunctata. So far as can 

 be judged from the pupa and cocoon of Bhima (undulosa), there 

 is some alliance indicated with the species just mentioned, but the 

 tufted abdomen of the imago indicates considerable diversity. 

 Taragma (siva), from its cocoon and shrunken larval skin, would 

 appear to be Cosmotrichid, whilst the egg is almost Pcecilocampid 

 in colour, but not in structure. Pachypasa (otus) is distinctly Eutrichid 

 from the larval structures, but Lebeda (nobilis), of which there are 

 no eggs, larvae, cocoons, or pupae in the collection, appears to be 

 rather Metanastriid than Eutrichid. Trabala vishnou, Lef., although a 

 tufted (?) species, has larva and cocoon which appear to show 

 Eutrichid characters, but the adult larva of this is very specialised, 

 and there is a certain suggestion of Pachygastria about the dorsal 

 hairs, whilst the pupa is very rounded and smooth. As a matter 

 of fact, this species appears to combine the characters of separate 

 Lachneid groups in a remarkable manner; the imago has suggestions 

 of Cosmotriche (potato ria), Macrothylacia (rubi), and Lasioca?npa 

 (quercus) ; the cocoon suggests Cosmotriche, but is very specialised ; 

 the larva, also very specialised, suggests Pachygastria (trifolii) and 

 Etitricha ( quercifolia), but one cannot say, on the material, which 

 resemblances are due to convergence and which to real relation- 

 ship ; one requires young larvae to settle some of these points. 

 Arguda jiavovittata, Moore, placed in Radhica by Kirby {Cat., p. 810), 

 is apparently Eutrichid, whilst berhoba, Moore, and vita, Moore, 

 placed in Odonestis by Kirby (Cat., p. 811), and rosea, Humps., 

 in Radhica {teste Kirby, Cat., p. 811), are all certainly related to 

 Odonestis (pruni), and show a distinct affinity thereto. One might 

 fairly, perhaps, consider that the Metanastriids exhibit sufficient 

 general characters to suggest a closer alliance than any other Eutri- 

 chids with the Pachygastriid stem, but still with the main (egg, cocoon, 

 and some imaginal) characters of the present specific forms distinctly 

 Eutrichid and not Pachygastriid. Nor does Dyar's tree (anted, 

 vol. ii., pi. vii) help us here. This author certainly brings Gloveria 

 and Macrothylacia somewhat near together, but still follows the old 

 views that the alliance is such as to bring them both close to 

 Lasiocampa and separate them widely from the Eutrichids, the various 

 subfamilies of which he distributes in such a manner as to suggest 

 for them no real connection. He widely separates, as we have 

 already pointed out (Joe. cit., pp. 460 — 462), Eutrichids on both 

 sides of his tree, and we have to unite his phyla, B, C, and J 

 (as well as part of D), for our Eutrichid stem. There is no need 

 to repeat here the critical arguments adduced 1 loc. cit., pp. 458 — 

 463) to show that the Dendrolimids, Cosmotrichids, and Eutrichids 

 {sens, stricto) are distinctly related, and we have already expressed in 

 tabular form (anted, p. in) our views of the main subdivisions of the 

 family Eutrichidae so far as we have been able to study them. Bacot 

 insists (in litti) that Odonestis {pruni), usually closely associated by the 

 authorities with Dendrolit?ius (pint), has a young (prehybernating) larva 

 that he can barely separate specifically from that of Eutricha (querci- 

 folia) in the same stage, whilst from the structural characters presented 

 by the newly-hatched larva of Dendrolimus, the latter, he says, appears 

 to be rather nearer Cosmotriche than Eutricha. In the Journal of the 



H 



