106 [May, 



mark that may be the scar of one is visible on 8th segment ; and the 

 10th segment is hardly made out with certainty. In most views there 

 appear to be only nine abdominal segments. The wings arise some 

 way in front of the hind margin of the mesothorax, so that their 

 inner borders instead of being continuous (as in Lepidopfera) with 

 the posterior margin of segment, run straight up over it, just as they 

 do over the other segments, the suture ending abruptly about one- 

 third up the segment. From the melathorax there extends downwards 

 on each side along and showing from beneath inner margin of upper 

 wing for nearly the width of first abdominal segment, a narrow margin 

 of what can only be called hind-wing, and exactly like the hind-wing 

 as seen in many Lepidopterous pupa3, but shorter and narrower. 



The pupa is of the brown chitinous aspect of ordinary Lepido- 

 pterous pupae. All the segments appear to be soldered solidly, ad- 

 mitting of no movement, though at each incision the transparency of 

 the pupa allows to be seen an intersegn.ental membrane extending 

 inwards and upwards. There is a central prominence of the meso- 

 thorax, and the metathorax and first four abdominal segments have 

 this repeated as a dorsal ridge along their anterior halves, rather more 

 raised at the anterior margins. No other sculpturing or any hairs 

 are detected, beyond the minute surface texture. 



Seen laterally the outline is less that of a cylinder, the ventral 

 aspect is quite straight, except for a slight prominence of the head. 

 The truncate front is formed by the head, but chiefly by the meso- 

 thorax, which being rounded above looks like a ball with the head 

 stuck on in front. The prothorax a very narrow inconspicuous stripe. 

 There is a slight waist behind the mesothorax dorsally, and the rest 

 of the dorsal line sweeps in a regular curve from this point to the 

 posterior extremity, first rising a little, then nearly (but being curved 

 not quite) parallel to the venter for several segments, and then 

 bending forwards. The wings on the lateral aspect appear large, and 

 extend down to middle of 4th abdominal segment, and are soldered 

 down, just as in a Noctua or Geometer pupa ; down the side in front 

 of spiracles is, rather a depression, as between dorsal and ventral 

 plates, than a flange. The wing origin is marked by a good deal of 

 complicated grooving and elevation. The 10th abdominal segment 

 can be seen better in this view, but even so doubtfully. The first 

 spiracle, like the others, is small, on the summit of a conical elevation, 

 it seems to have a cribriform structure. No hairs can be made out 

 in this lateral view on either the dorsal or ventral surface. The wing 

 neuration is well marked by raised ridges. 



