o 



782 THE BUTTERFLIES OF NEW ENGLAND. 



cell without reaching tlicin. The mieropyle rosette, which I have seen 

 only in nais, is a deep inf'undiijulif'orm pit at the very centre of which is a 

 secondary minute puncture. That of palmerii is figured by P^dwards as 

 forming a very broad, cratcriform, abrupt pit. The caterpillars at birth 

 have been excellently figured by Mr. P^dwai-ds, but as I have already 

 ofiven the characteristics observed in specimens they need not be re])i'ated 

 licre. Mr. Edwards has kindly sent me a caterpillar of Chrysobia ver- 

 gulti in its second stage, which is the only one of the later stages of the 

 caterpillar 1 have seen in this tribe : but Mr. Edwards figures that of 

 Chrysobia nais in an excellent manner. In this it appears that the cater- 

 pillar is strictly cylindrical, nearly equal, tapers very gently from the 

 middle of the body backward, more rapidly at the extreme tip, scarcely 

 tapers forwards, and has somewhat abbreviated legs and prologs. The 

 head, which is constructed precisely as I have described that of Nemeo- 

 bius, is relatively lai'ger, being about three-fourths the greatest width of 

 the body, and therefore not greatly narrower than the segment imme- 

 diately posterior to it; according to Edwards, it is "partly covered" 

 by this but not retractile ; the posterior border is dorsally emarginate 

 as in Nemeobius, and long hairs are found upon it up to the summit. 

 The body is covered with longitudinal rows of spreading fascicles of 

 short hairs, two upon either side besides a stigmatal series of longer 

 hairs ; some long arching hairs are also found at each extremity of the 

 body, and the first thoracic segment is covered with a divided dorsal shield 

 not conspicuously more corneous than the other parts of the body, bristling 

 with hairs. The chrysalis has a shape precisely like that of Nemeobius, 

 excepting in being slightly longer, and like it is attached both bj^ the tail 

 and by a girdle round the middle ; the ventral surface is nearly flat, with a 

 sinuate curve ; the body is sparsely covered with not very long hairs and 

 the cremaster, while completely independent and protruding, is not pointed 

 but apically truncate. 



We think that it will appear very clearly from this review that the 

 structure of the early stages and the transformations of the Lemoniinae are 

 not widely different from those of the Lycaeninae, by no means sufiiciently 

 so to support the separation of the two groups as distinct families. There is, 

 indeed, no more difficulty in harmonizing the somewhat curious differences 

 one sees in the larval and pupal forms of the Lemoniinae than there is in 

 those of Lycaeninae ; there is no more difficulty in retaining any one of them 

 within the group than there is in retaining among the Lycaeninae such 

 forms as Sjjalgis, figured liy Moore, or Curetis, figured by Horsfield, or our 

 own Feniseca. On the other hand the subdivision of the subfamily suggested 

 by Godman and Salvin would seem to be borne out by the characteristics 

 of the early stages. For it would appear that we may even be able to 

 separate these two groups by characters drawn from the early stages. It 



