462 BRITISH LEPIDOPTERA. 



as egg-producing organisms. Bacot notes : The ova of Eutricha 

 (quercifolia) are much smaller than those of Cosmotriche (potatoria), 

 even smaller than those of Lasiocampa (querciis), and consequently the 

 larva in the first instar is also much smaller than those of Cosmotriche, 

 Lasiocampa or Macrothylacia, and one is strongly inclined to suspect 

 that the small size of the eggs of these species compared with 

 the size of the moth represents a specialised condition. In con- 

 clusion we may add that we should certainly be inclined to 

 disagree with any phylogeny that did not separate Dyar's phyla B, C 

 and J from the main stem at the same point, that did not find a much 

 less specialised position for Lachneis and Trichiura, and an entirely inde- 

 pendent one for Poecilocampa (vide, pi. viii). The peculiarities in the 

 neuration noticed by Dyar (supra) certainly appear to us, as we have 

 attempted to show, to be explicable on other grounds than those offered 

 by him. We suspect that he will grant that the " large intercostal cell" 

 of the hindwings in phyla B and C is a specialisation, and that it is 

 much more in accordance with fact that those genera that have never 

 had the large expansion of the area at the base of the costa of the 

 hindwings, have never required this large cell, nor the nervures that 

 arise therefrom to carry the enlargement. We suspect that the isolation 

 of nervure 7 from the discal cell is a generalised and not a specialised 

 feature of the neuration, and that this has been maintained in the 

 Lachneidae, both in otherwise generalised and specialised forms, and 

 lost more or less, in the Eutrichidae, under the stress of costal 

 development. 



Some of Bacot 's suggestions as to the phylogeny of the group, have 

 already been embodied in the preceding pages (ante, pp. 460-461), but 

 he adds that he is not sure that "the highly specialised larva?, offer 

 such good characters for the purposes of pbylogeny as the relatively 

 more generalised imagines. The larval life is usually long, and the 

 imaginal short and often retired. Taking the ensemble of the characters 

 offered, the mode of making the cocoon and the structure of the pupa? 

 are very likely to give a correct view — although the similarity of 

 cocoons of the ' eggar ' type (even in quite unrelated orders) , which 

 must have been independently evolved, may easily be pushed too far, 

 and one cannot think that the similarity of the cocoons of L. querciis to 

 those of the Cochlidids is really one of close relationship, for it is 

 impossible not to place such forms as rubi, quereifolia, &c, with their 

 divergent cocoons, much nearer to L. querciis than one would the 

 Cochlidids, i.e., it is impossible to conceive that I+. querciis and L. 

 lanestris have an unbroken descent from the Cochlidids to the exclusion 

 of such forms as Cosmotriche, Malacosoma, Macrothylacia, &c." He 

 further adds: "The species trifolii and querciis, potatoria and querei- 

 folia, castrensis and neustria, represent tbree well-marked and distinct 

 groups, supported by all the characters yet obtained from the oval, 

 larval and pupal stages as well as the imaginal habits, but when it 

 comes to placing any of the other species with these groups, difficulties 

 at once crop up. Lachneis lanestris might be placed on the evolutionary 

 line with the querciis group, there is nothing in the larva, pupa, or 

 cocoon of which one knows that contradicts this position, but the egg 

 and egg-laying habit suggest a position nearer Trichiura and perhaps 

 between Trichiura and Malacosoma. Macrothylacia rubi has a larva 

 closely approximating to that of L. querciis in structure and develop- 



