1226 THE BUTTERFLIES OF NEW ENGLAND. 



rally a distinct mesonotal tubercle or prominence of some sort. The mode of 

 support is like that of the otiicr members of the family, with the exception 

 already alluded to of the anomalous mode of pupation of Thais and the 

 other instances, Doritis and Parnassius, where tiic chrysalis is confined in 

 a cocoon, and it is not clear from what has been stated, in precisely what 

 manner the body of the chrysalis itself is supported. To judge, however, 

 from specimens received from Dr. Staudinger, it would appear that they 

 find no support beyond the cocoon itself, since there is no sign of cremastral 

 hooks ; and though tiie middle of the mesonotum, at least in Doritis, lias a 

 slight, transverse depression such as a girth might make at this point while 

 the body was still soft, there is no mark whatever at the appropriate place 

 on the ridge which follows down the wing from the basal wing tubercle ; 

 and in none is there any real mark of the thread itself. 



It has lately been suggested by Mr. W. H. Edwards, without offering, 

 however, any reason for his position , that Parnassius should be removed to the 

 neighborhood of the Lycaenidae. This opinion would seem to be formed 

 largely by regarding the structure of the egg, which certainly bears a certain 

 resemblance to those of that group. But not a single additional point can 

 be brought forward. Even the form of the egg does not agree ; and a 

 different story is told by the structure of the caterpillar at birth, including 

 the form of the head, its clothing of hairs to the top, the mere thread- 

 like ring which forms the second antennal joint, the excessive brevity of 

 the third antennal joint, tlie clotliing of the body with bristles not greatly 

 longer than the segments, their apparently hollow nature, their clustering 

 on ranged tubercles, the absence of distinct series of crateriform disks, the 

 long legs, both thoracic and abdominal, and the low position of the eighth 

 abdominal spiracle, not to mention the osmateria which are present iiere 

 as in the mature larva. The same is the case without question in the 

 mature caterpillar which one finds it hard to separate by any common 

 characteristics from tiiat of the Papilionidi and which in no single char- 

 acteristic of its structure resembles the Lycaenidae. Precisely the same 

 may be said of the chrysalis, with the single exception as far as I know of 

 the genus Doritis, where in the rounded compact form with its few abrupt 

 elevations, one sees a certain resemblance to the Lycaenidae, a resemblance 

 which is heightened by the abrupt falling of the curve in front and behind, 

 which causes the head to face almost completely forward and the last seg- 

 ment almost completely backward. But the intimate relation of this 

 genus to Luehdorfia, in which none of these characteristics are found, ren- 

 ders such a view of its affinities impossible. Moreover, the structure of 

 the prothoracic stigmata is distinctly that of the remaining members of 

 the group and not at all tiiat of the Lycaenidae. 



Gr-eneric divisions. As iiiis been stated, the bulk of the species of 

 this subfamily have been placed by authors in a single magazine genus, 



