NORTH AMERICAN LEPIDOPTERA. 0/ 



tapering in front and prolonged in front of the primaries. Abdomen 

 somewhat flattened beneath, with a broad, fan-like anal tuft. The 

 posterior edge of the segments above is armed w'ith minute, flattened 

 spinules. The primaries are eleven veined, the outer margin rounded, 

 entire, the centre of the wings transjxirent and crossed by the dark 

 veins. Secondaries also transi)arent in the middle, the outer margin 

 somewhat excavated between veins lb and 2. The colors are black, 

 with yellow and ferruginous maculation, and the species, when flying 

 in the bright sunshine, bear a deceptive resemblance to humming 

 birds, for which they are not infrequently mistaken. 



In this genus I regret that the material at my command has been 

 too incomplete to accord that careful study which it requires. There 

 are many species described, and all sorts of opinions as to their va- 

 lidity. Boisduval's species arc not .satisfactorily identified in any 

 collection, and are probably .synonyms of others. Mr. Strecker's 

 species are not certainly known to any but him.self, while of the 

 other species described, many of them are based upon characters 

 which all analogy seems to prove evanescent. 



In the few species of which I have been able to examine the male 

 genitalia, I find an unexpected similarity amounting jjractically to 

 identity of structure. This removes one important guide to the dis- 

 tinction of species and leaves us to more superficial characters. Of 

 these, Mr. Henry Edw-ards has remarked that differences of color 

 are here given specific value, which would not be accorded them in 

 other genera. It is this fact which makes long series of bred speci- 

 mens necessary to decide upon the range of variation. Many of the 

 species have been bred, but apparently only in small numbers, and 

 the results have never been carefully analyzed. I have, therefore, 

 given descriptions here of all the described forms, mostly reproducing 

 the original descriptions and supplementing them with such notes as 

 I have made. 



It may be added that the genitalia are of the typical Sphingid 

 type, the supra-anal hook showing a decided tendency to divide — in 

 fact it consists of two hooks closely joined along the median line. 



The transformation of the larvse takes place on the surface of the 

 ground in an imperfect cocoon among the leaves. 



As the best way to give an adequate idea of the characters used 

 to separate the more common forms I will give a review of the work 

 of Messrs. Grote & Robinson and of Mr. Grote singly upon the genus. 

 The first paper was by Messrs. Grote & Robinson in the Ann. Lye. 



