46 SPH.ING1D.S2. 



Key to the Subfamilies. 



Imagines. v 



1. First segment of palpus without patch of short 



sensory hairs on the inner surface near base. 



(Asemanophorse) 2. 



First segment of palpus with a patch of short 

 sensory hairs on the inner surface near base. 

 (Semanophoree) 3. 



2. Terminal segment of antenna long and thin, with 



long hairs and bristles (short in Dolbinopsis 

 and Dolbina, but these two genera may be 

 distinguished from the Ambulicinje by the [p. 48. 



fore wing not being sinuate) Acherontiinae, 



Terminal segment of antenna short (long in 

 Compsogene, Oxyambulyx and Cypa, but these 

 three genera may be distinguished from the 



Acherontiiislze by the fore wing being [p. 98. 



sinuate) Ambulicinae, 



3. Pilifer normal 4. 



Pilifer consisting of a terminal part bearing 



sh^rt or vestigial bristles and a proximal part [p. 397. 



bearing long bristles Choerocampinae, 



!. Abdomen with the tenth segment mesially 



divided Sesiinae, p. 238. 



Abdomen with the tenth segment not mesially [p. 259. 



divided Philampelinae, 



It is difficult to make keys to the larvae and pupae by 

 subfamilies, as the larval and pupal characters cut across the 

 divisions in which the imagines have been placed. 



In the subfamily Acherontiinae. for instance, the larval head 

 is usually round, but in Dolbina inexacta it is triangular, as 

 in many of the Ambulicinae. In the subfamily Ambulicin^e 

 the head may be round, as it is in most of the AcheronthnyE, 

 or triangular. In Ehodoprasina callantha, of the Ambiilicinje, 

 the spiracle of segment 5 is surrounded by an ocellus-like 

 marking, a character which reappears in some larvae of the 

 tribe Nephelini, subfamily Philampelin.3E. 



In the tribe Sesiini, subfamily Sesiini, the larvae of one 

 genus (Hsemorrhagia) are Philampeline, of another genus 

 (Cephonodes) Acherontiine, and of the third (Sataspes) 

 Ambulicine. In the tribe Nephelini, subfamily Philam- 

 pelinse, many of the larvae closely resemble those of the 

 Chcerocampin^e. The horn and other characters are very 

 variable in all the subfamilies. 



In dealing with the pupae we meet with similar difficulties. 

 Pupae of the Asemanophorse can usually be distinguished 

 from those of the Semanophorse by the uniform colouring 

 of the former, and by their being formed in a cell underground, 

 those of the Semanophorze being seldom of uniform colouring 

 except in the Sesiini, and being formed in a rough cocoon 

 on the surface of the ground or among leaves, but it is difficult 

 to find any constant characters distinguishing the pupae 



