50 JOHN B. SMITH. 



Ill the present paper I have changed, considerably, my views as 

 given in the synopsis above cited, and in the direction suggested by 

 Prof. Fernald. I have no apology to offer for this, since the prin- 

 cipal changes are the result, less of corrected observation, than of 

 the advisability of adopting generic terms for certain fairly well 

 marked divisions or sections of the genus Sphinx. The genera 

 adopted in the Sphhigidce are very largely opinionative. It rests 

 merely on individual opinion as to what value is to be given them — 

 whether we will with Boisduval have a large genus with numerous 

 sections, or with Butler have small, compact genera, each name ex- 

 pressing a definite idea, less of structure than of appearance oi' 

 habitus. I have adopted the latter view in this paper, though I have 

 not carried it to the extent of some of the recent authors. From 

 the other point of view the subfamily Sphingidce. would contain only 

 the genera Argeus, Dilophonota, Sphinx, Cautethia and Ellema, the 

 genera Ceratomia, Daremma, Proioparce, Diludia, Hyloicus and Dolba, 

 being comprised under the generic term Sphinx. The assemblage 

 thus constituted is unwieldy, and the generic term loses definiteness. 

 The smaller genera are susceptible of sufficiently accurate definition 

 to enable them to be readily recognized. 



The fiimily Sphingidrr, as here ti-eated, corresponds with the Sp)hinx 

 /egiUnuc of Linnaeus, and the family characteristics are as follows : 



Body i-obust, yet generally graceful ; abdomen long, conic, usually 

 cylindric, rarely depressed, often tufted at the sides and at tip. Head 

 usually large, rarely retracted ; eyes often prominent, naked, some- 

 times lashed, never hairy. Palpi usually well developed, never ab- 

 normally so, reaching the middle of front, its clothing often forming 

 a short, conic, obtuse snout. Tongue usually long and strong, often 

 much exceeding the length of the insect itself, sometimes obsolete. 

 The thorax is robust, oval, convex, rarely somewhat depressed, often 

 considerably produced in front of the base of primaries. The an- 

 tennae are peculiar ; they are prismatic, the joints, as a rule, with 

 hexagonal sides, the top and bottom flat ; sometimes they are almost 

 clavate, with an abrupt little hook at ti]) ; more usually, in the 

 typical genera they are thickest in the middle with a similar tip ; 

 quite commonly they are fusiform, the terminal hook long and 

 gradual, usually recurved, but often simply curved. The wings are 

 usually comparatively small and narrow, the primaries with usuallv 

 acute apices and oblique outer margin, the inner margin always 

 much shorter than the costal, outer margin sometimes angulated and 

 dentate. 



