116 JOHN B. SMITH. 



iiig it clear and easy of definition. The addition of these genera to 

 the Cheer ocampince, add a little discordant material to the already 

 heterogeneous mass. I believe the group capable of further subdi- 

 vision, but cannot attempt it here from paucity of genera and species 

 in our fauna, belonging to those very groups which here seem aber- 

 rant. 



The genera added from the Macror/lossince, are readily distinguished 

 l)y having the al)domen tufted at the sides or at tip, and excejit in 

 jEUopos the wings are angulated. I place them at the head of the 

 series here, though in a classification of the s})ecies of the world other 

 situations would probably be assigned. 



^•Ellopos is readily distinguished by the short, rather slender an- 

 tennae, nearly subequal and furnished with a little hook at tip ; by 

 the broad depressed body, the fan-like tuft at ti}) of abdomen and 

 the entire margin of wings. 



This genus is a close relative of the European Macroglosm siella- 

 fariim, and connects that form with the Choeroeampince. Macroglossa 

 steUatarum, by-the-bye, offers too many points of diflference from 

 croatlca to be considered congeneric with it, while on the other hand, 

 despite the opaque wings, I believe croatlca to be generically identical 

 with the clear winged forms of Hemaris. 



Emjo is not very closely related to the preceding. It is aberrant 

 and easily recognized in our fauna by the prominently crested thorax, 

 the fan-tufted abdomen and small angulated wings. This may form 

 a subfamily type when its tropical relatives are considered. 



A)nphion contains a bright little species with long, fusiform an- 

 tennae recurved at tip, and the wings are angulated. 



Thyreus is a peculiar form, — plum}), heavy, somewhat depressed. 

 Head retracted. iVbdomen with lateral, truncate tufts to the ab- 

 domen ; primaries angulated. Altogether, a distinguished form. 



Deldamia is still more aberrant, with a decidedly Smerlathid 

 habitus. The head is retracted, crested, the vestiture thin. Abdo- 

 men slender, tufted at tip ; the primaries angulated. This form in 

 cataloguing should be placed at the foot of the series. Its tenden- 

 cies are all toward the Sinerinihids. 



The remainder of the genera are those usually referred to this 

 subfamily, and they are all untufted, with wings not angulated, and 

 with no striking structural features. 



Deiiephila has the antenna thickened towai'd tip, but not forming 

 a distinct club. The outer side of the first tarsal joint is armed with 

 long spines, a feature not found in any other genus of the subfamily. 



