NORTH AMERICAN LKPIDOPTERA. 213 



ceptiou ; 9 IVom tlie subcostal to the eosta near tij), soiuetiuies fork- 

 ing near the end of its course giving off thus 10, which, when present, 

 is always very short ; 1 1 from the subcostal usually about the middle ; 

 12 is the costal vein. Secondaries with the cell not extending to the 

 middle of the wing, cross-vein variable; four always from the inferior 

 angle ; two usually from the middle of the median ; three variable ; 

 five from the cross-vein at varying distances from four, usually closer 

 to four than six ; six and seven together from the end of the sul)- 

 eostal ; costal vein variable, depending somewhat on the wing shape. 



This subfamily is very distinctly separated in its entirety from the 

 rest of the Sphinges, forming a division equal in value to all the rest. 

 In its division I have met with considerable difficulty, less because 

 there were no characters, but because there wei'e so many that a 

 genus might almost be based on every s|)ecies. Indeed, this has been 

 done, except for >nyop^, which I believe is not a generic type. Adylus 

 and myops, in fact, are the only species strictly identical in structure, 

 except in the genitalia. The other extreme — lumping everything 

 under iSmerudhus seemed e(jually umlesirable, and it was not an easy 

 task to strike the correct mean. Prior writers seem to have had no 

 definite basis for their svdjdivisions, a number of species being first 

 associated in one way then in another, and finally sej)arated alto- 

 gether. The genital structure of the male has here again given 

 valuable hints, and I believe the present arrangement as satisfactory 

 as can be easily made. I have excluded the species of Ellema from 

 the subfamily contrary to my previously expressed opinions, having 

 convinced myself that Mr. Grote was more nearly right in consider- 

 ing them degraded Sphlngliue. Though having an aboi'ted tongue, 

 tlie wing form and habitus, as well as the genitalia, are more Sj)hingid 

 than otherwise. 



I head the species with Trij/fogoii, which contains our lai'gest spe- 

 cies, and has the fore tibiae armed with a long, stout claw. The 

 genital structure is unique and descril)ed in its proper j)lace. There 

 is but a single species. I follow this by Smerlnthw^, which contains 

 three species agreeing in type of genital structure, in the general body 

 form, in habitus and in pattern of maculation. In all the dor- 

 sum of thorax between \\w patagia? is deep brown, the primaries 

 have an outwardly angulated l)and or line at basal third, and the 

 secondaries have the ocellus black centred and connected by a black 

 spur, with the anal angle ; the costal margin is straight. In an- 

 tennal structure these sj)ecies differ greatly, and to a less extent in 



