BOMEYLIID^. 177 



the first usually open, tliis cliaracter not being of generic value. 

 IJpper branch of 5th A^ein foriiiiug a considerable portion of lower 

 side of discal cell; the contact occasionally only punctiform. Anal 

 cell always open to wing-border. 



This subfamily is easily recognised from all others by the com- 

 paratively long prsefurca and the sharp angle at the base of the 

 2nd longitudinal vein. The body is flattened and broad, the eyes 

 well separated in both sexes, and the antennae comparatively wide 

 apart at the base. 



Table of Qenera. 



1. Four submarginal cells IIypekalonia, Rond., 



Three submarginal cells 2. [p. 178. 



Two submargiDal cells 3. 



2. Third antennal joint elongate conical, 



with terminal stvle bearing microscopic [p. 189. 



bristle Exoprosopa, Macq., 



Third .antennal joint short onion-shaped, 

 with styliform prolongation bearing a [p. 213. 



distinct pencil of hairs at tip Spogostylum, Macq., 



3. Third antennal joint short onion-shaped 



(sometimes more conical), with stjdiform 



prolongation alwaj's bisected near tip, [p. 214. 



and bearing an apical pencil of hairs . . Argyrawceba, Sch., 

 Third antennal joint forming a style-like 

 cone with microscopic apical bristle ; no 

 pencil of hairs 4. 



4. Contact of discal cell with 3rd posterior 



not greatly longer than with 4th pos- [p. 232. 



terior. Lepidanthrax, O.S., 



Contact of discal cell with 3rd posterior at 

 least twice as long as with 4th, the latter 

 contact sometimes merely punctiform . . Anthrax, Scop., p. 234. 



Several authors have endeavoured to split up some of the above 

 genera, but such groups must be adopted with much caution. 



When Osteu-Sacken wrote on Hyperalonia in the ' Biologia 

 Centrali-Americana,' his knowledge was admittedly confined 

 practically to North and Central American species, and though 

 Verrall, possessing the late Mr. Bigot's collection, in which 

 numerous exotic species of both this genus and BxojirosojHi are 

 represented, includes the character " claws not toothed near the 

 base" as generic in Hyperalonia, and " claws toothed near the 

 base " in Exoprosopa, I am compelled to omit it as being incorrect. 

 In three species of the former genus (chri/solamjyis, sphinx, and 

 suffusipennis) a pair of perfectly distinct teeth are present on the 

 posterior tarsi, and they are present, though small, in tantalus. In 

 some other species they are absent, though I have not examined 

 specimens of all the species available. It may be observed that 

 Loew, in his treatise on the South African species, noted that 

 some species of Exoprosopa had this tooth at the base of the ungues, 

 but declined to regard it as a generic character. In those species of 

 Exoprosopa that I have specially examined for this character the 



