244 BRITISH LEPIDOPTERA. 



looked upon as the main Sphingid stem, with Hemarids, 

 Sesiids, and even the Sphingids and Amorphids as branches, 

 that is, giving the Sphingids a monophyletic instead of diphy- 

 letic grouping, although our further work shows that Bacot's 

 suggested grouping (anted, vol; hi., pp. 365-366) has much in its 

 support, and, unless one is prepared to grant a diphyletic origin 

 to the oblique-striped larvae — Amorphids, Sphingids — a mode of 

 grouping based largely on that suggested by Bacot must be granted, 

 and, as certain genera — Ceratomia, Daremma — possibly to be referred 

 to the Sphingidae (sens, strict.), have unspecialised pupae, and of a 

 character not far different from that of the more generalised 

 Eumorphids— Darapsa, Gurelca, &c. — possibly such an arrangement 

 is not so difficult as at first appears. 



Although it may be considered as outside the scope of our work, 

 it may be well to consider the position of the Daphnids with regard 

 to the rest of the Eumorphids (sens. lat.). For this purpose Chapman 

 and Kaye give the following grouping, starting with a basal 

 Eumorphid not very definitely separable from the Sesiids: 



1. Darapsinae* — Larva with subdorsal lines without tendency to break up 

 into ocellated spots. Pupa with labium ventral. Puparium superficial. 

 Imago with nervure 8 of hindwing sharply curved upwards at starting, 

 approximating at end of cell and receding well beyond cell. 



2. Philampelinae — Larva with subdorsal line simple, but with tendency to 

 reversed oblique stripes and to breaking up of spiracular line. Pupa with 

 labrum anterior. Puparium subterranean. Imago with nervure 8 ot 

 hindwing only slightly upcurved at starting, approaching 7 before the 

 end of the cell and continuing almost parallel for a long way beyond ; 

 antenna with a long produced tip. 



3. Daphnidinae — Larva with eye-spot on thorax. Pupa with labrum dorsal. 

 Puparium superficial. Imago with nervure 8 of hindwing only slightly 

 upcurved at starting, approaching 7 before the end of the cell and con- 

 tinuing almost parallel for a long way beyond ; antenna with a very 

 short strongly recurved tip. 



4. Eumorphinae — Larva with abdominal eye- or ring-spots. Pupa with 

 labrum anterior or dorsal. Puparium superficial. Imago with markings 

 dominated by streaks from apex ; nervure 8 of hindwing approximating 

 to 7 well beyond cell, and receding again after a very short distance. 



a. Phryxidi — Larva with thoracic and abdominal ring-spots. Pupa 

 with labrum anterior. Imago with large head, &c. 



b. Eumorphidi — Larva with abdominal eye-spots aud retractile 

 thorax. Pupa with various specialisations. Imago more 

 primitive, small head, &c. 



Groups such as the Acosmerygids, Pachyliids, &c, are not 

 considered here, since they are outside the proposed range of our 

 work. 



* We are fully aware that the terminology of these is not in accordance with 

 our previous work in this group, in which we have treated the whole of the 

 Eumorphids as a subfamily of the Sphingidae. As our work has progressed, however, 

 our ideas have become much modified, and we are inclined to look upon the 

 EumorpJiidae as a subfamily of quite equal importance with the Sphingidae. Besides, 

 to have maintained these as tribes, it would have been necessary, owing to our 

 limited terminology, to make Da raps id i\ Philampelidi,Daphnididi equal, phylogeneti- 

 cally, to Eumorphidi, Phryxidi^ &c,, a view that would have been entirely wrong, 

 for a branch systematically is of importance, not in proportion to its number of 

 species, nor to the number of subbranches into which it divides, but to the amount 

 o! its differences from the remaining branches. Therefore, if we retain Phryxidi, 

 Hippotionidi, Eumorphidi and Elibiidi as tribes of Eumorphinae, we must make the 

 other great Eumorpnid branches subfamilies equal in value to Eumorphinae [sens, 

 rest.) — i.e., Darapsinae, Philampelinae, &c, — to show their phylogenetic value, 



