HYPEEALONIA. 179 



2. Wings very dark brown, but always with a 



single limited clear space 3. 



Wings very dark brown, violet-tinged, no clear 



spaces whatever ; in one species wings mo- 

 derately dark smoke-brown 4. 



Wings pale or dark grey 6. 



•3. Hind margin of wing narrowly clear ; remain- 

 der of wing wholly dark dives, Walk., p. 180. 



Wing-tip rather broadly clear ; remainder of 



wing wholly dark ; 2nd posterior cell much 



broader at base than tip trisds, Wulp, p. 180. 



4. Wings uniformly very dark blackish brown, 



posterior part hardly lighter ; 2nd posterior 



cell much broader at base than tip (except in 



semifascata) 5. 



Wings very dark blackish brown, fading to dark 



grey towards hind border ; 2ud posterior cell [p. 180. 



not greatly wider at base than tip , aurantiaca, Guer., 



Wings moderately dark smoke-brown (vari- 

 able); 2nd posterior cell much broader at 



base than tip sphinx, F., p. 187. 



•5, Median stripe of pubescence on venter white ; 



antennte tall black tantalus, F., p. 181. 



Median stripe of pubescence on venter yellow ; [p. 182. 



1st antennal joint reddish brown chrysolampls, Jaen., 



'6. The 2nd posterior cell slightly narrower 



at base than tip; wing colour fading gradu- [p. 188. 



ally from costa to hind margin semifuscnta, Brun., 



The 2ud posterior cell nearly twice as wide 



at base as at tip ; wing colour pale grey, with [p. 188. 



small darker suffusions snffuslpennis, Brun., 



7. Median stripe of pubescence on venter bright [p. 183. 



reddish yellow ; antennae red-brown Jiaviventris, DoL, 



Median ventral stripe of pubescence snow-white: [p. 185. 



antennae all black anotnaus, Rond., 



Though the characters adopted ia the foregoing table may be 

 somewliat difficult to follow, especially by the inexperienced 

 student of this family, they are actually more reliable than an}^ 

 •drawn from the coloration of the body or pubescence, and more- 

 over are much more lasting. The following further notes may 

 prove of assistance in separating these closely allied but perfectly 

 valid species, of nearly all of which I have seen more than one 

 example. Of the ten species in the table, only the last two, j^rti-j- 

 ventrls and (enomaus, have any considerable extent of clear or 

 practically clear surface of the wing, this being confined to the 

 apical and posterior portions, and without any clear-cut line of 

 demarcation, the dark part of the wing being the baso-costal band 

 so common in Atsttecuacix.^. Secondly, of the remainder, tristis 

 and dives have wholly dark wings, except for a very definite and 

 clearly-cut pale space, in the former at the wing-tip, in the latter 

 on the hind border.* Of the remaining species, tantalus, chri/so- 



* As regards dives, Walker is my authority, because if the clear part is not 

 so sharply defined as he says, the species will fall into the group containiiigr 

 Jiaviventris and cenomaus, and a doubt as to its exact identity may then an'se^ 



X 2 



