DIURNAL LEPIDOPTERA. 71 



ordinarily small size, and to this we may add the tenuity and general 

 delicacy of their structure; their head between the eyes is usually 

 very narrow or twice as high as broad, so that the eyes are approxi- 

 mated, and infringe to such an extent upon the antennal scrobes, as 

 to excise to a greater or less extent their own inner margin ; the meta- 

 thorax is less distinctly separated from the meso-thorax than usual ; 

 the front wings are pretty uniformly broad, rarely as elongate as in 

 other iamilies, and both wings are entire, excepting when the hind 

 pair is tailed ; in the heteromorphous character of the fore legs in 

 the two sexes they may be known from all other butterflies, excepting 

 the very lowest Nymphales (Libytheidae), which, on that account, have 

 been placed with them by Bates and some recent authors. 



The transformations of so few of the higher group of this family 

 are known, that it is impossible to make any general statement con- 

 cerning them. But the eggs of the Lycaenids or lower sub-family are 

 peculiar for their echinoid or turban-shaped, heavily pitted form, in 

 which respect only Parnassius appears to agree with them. The 

 caterpillars are remarkable for their onisciform shape and gliding 

 motion, their nearly aborted pro-legs, the minuteness of their head, 

 and its power of complete concealment within the first body-segment; 

 they are hairy, but never bear spines or filaments. The chrysalids 

 are short and compact, completely rounded and closely attached to 

 the surface by a girt; the cremaster is wanting, and the hooks seated 

 directly upon the last abdominal segment, which, like the head, is 

 completely carried over to the under surface of the body. 



Doubtless the early stages of the Erycinids agree to a certain- extent 

 with those of Lycaenids, but not altogether, for the only egg known 

 (that of the European species), is described as almost globular and 

 smooth ; the head of the caterpillar cannot be so completely with- 

 drawn, and the body is furnished to some extent with filaments or 

 possibly spines and only appears subordinately onisciform. Still, so 

 far as known, the early stages of Erycinids agree better with those 

 of Lycaenids than with any other butterflies, and these features, with 

 the compact form of the chrysalis and its closely girt attachment to 

 its support, "*" together with the opposite development of the fore legs 

 in the two sexes of the imago and the close siiiiilarity of all other 

 points in the structure of the perfect ibrm, including tlie absence of 

 the nervule attached in all other butterflies, excepting the Pierids, to 



♦According to Bates, the pupa of Stalachtis, one of the highest Erycinids, i 

 " secured rigidly by the tail in an inclined position, without a girdle ;" this mod 

 of suspension forms a natural passage to the freely hanging Nymjihales. 



