82 BRITISH BUTTERFLIES. 



Comparison with plate iii of the preceding volume will show, not only- 

 characteristic differences from those of the Chrysophanids, but also 

 distinctly less uniformity in type ; indeed, the differences between the 

 Callophryid (C. rubi), Bithynid (B. quercus), and Ruralid (R. betulae) 

 eggs, both inter se, and, compared with the Strymonid eggs (S. iv- album-, 

 8. pruni, etc.), are most marked, and those of the latter group are so 

 peculiarly specialised that they have no very close similarity with 

 those of any other butterflies, so that we see there is much less 

 uniformity in the eggs of the " hairstreaks," than in those of the 

 "coppers" and "blues." Scudder says that they are "echinoid- 

 shaped, studded with numerous projections, connected by radiating 

 ridges, and are laid singly." Doherty, who bestowed much attention 

 to the study of the eggs of the Lycaenids of India, subdivided (Journal 

 As. Soc. B., iv., p. 110) the Ruralidae or Lycaenidae into six groups, on 

 the structure of the egg, the third subdivision, or Theclinae, being 

 noted as having — 



" The egg fully half as high as wide, convex above, widest close to the base, with 

 coarse, minutely vesicular, reticulations, forming large irregular pits over the 

 surface, and bearing broad depressed tubercles at their intersection." 



General statements of this kind, however, give little idea of the 

 remarkable structure exhibited by the eggs of Strymon w-album, S. pruni, 

 Laeosopis robor is, etc. (see plate ii), and Chapman observes (in litt.) that 

 the sculpture of the Ruralid eggs appears to be (at least in many, if not 

 all, cases) formed by an outer foamy layer, the true egg within being 

 smooth. This, he thinks, will prove to be the case also with Limenitis. 

 Scudder's comparison of the eggs of the three main subfamilies 

 of the Ruralids has already been given in the preceding volume, 

 p. 317, and we may here draw attention to Chapman's statement 

 that the Theclids have their elements more widely distinct than 

 the Lycaenids and Chrysophanids, the coppers coming much nearer 

 to the "blues" than to the "hairstreaks;" Callophrys rubi and 

 Thestor ballus, he says, are unquestionable Theclids ; the separation 

 between them and the other " hairstreaks " is an arbitrary boundary, 

 that between them and the "coppers" is a broad neutral boundary, 

 whose inhabitants we do not know. As bearing on this, we may note 

 that, whilst most (if not all) of the Ruralid and Strymonid species go 

 over the winter in the egg-stage, the Callophryids and Thestorids 

 hybernate as pupae. 



The "hairstreak" larvae have already been diagnosed and compared 

 with those of the Lycaenids and Chrysophanids (op. cit., pp. 315 and 

 317), both in their first and last stadia. They are, perhaps, rather 

 flatter than their nearest allies, and are said to feed chiefly on the 

 leaves of trees and bushes, rarely on low herbaceous plants, but this is 

 only true in part, for the Callophryid and Thestorid larvae appear to 

 prefer such. Some, however, bore into, and devour, the interior of 

 fruits, for which purpose their marvellously developed neck (interseg- 

 mental membrane between head and prothorax) is particularly fitted ; 

 one of the most remarkable of these is the well-known " pomegranate 

 butterfly " (Virachola isocrates), described by Westwood (Trans. Ent. 

 Soc. Lond., ii., pp. 1-8, fig. 1), and discussed at length by Niceville 

 (Butts, of India, etc., iii., pp. 477-481). We have already stated (supra) 

 that most of the " hairstreaks " hybernate in the egg-stage, but the 

 larvae are quite fully-formed within the egg, and it would be as accurate 



