EUMORPHA ELPENOR. 81 



difference in their disposition, sometimes the lips with the central 

 line as depression occupy a circular area, more often it is narrower, 

 i.e., elliptical, and sometimes, by the sides being nearly parallel, it 

 is rather quadrangular. The small angular portions of the ioth 

 segment not occupied by the scar and spike are strongly pitted, 

 but usually without special feature ; rarely, close to the posterior angle 

 of the anal scar, there is an irregularity that may be the scar of the 

 larval claspers. The male tubercles are full and rounded, occupy an 

 exactly circular area, with the median crevice between them ; this is a 

 little wider posteriorly. Though in the middle of abdominal segment 

 9, these male tubercles have the appearance of belonging to 10, as 

 the posterior border of 9 is wanting opposite them, i.e., appears 

 to bend up to them, and may more easily pass round them in front 

 than behind. Precisely similarly the segmental divisions are wanting 

 in the same parts in the female pupa between 8, 9, and 10 ; 9 

 presents a few wrinkles, which pass back to the anal scar, and 

 the margin of 8 — 9 passes forwards to the posterior pore. This 

 is very similar to the male pore, but on a much smaller scale ; 

 in front of this is, less marked but quite distinct, another longitudinal 

 impression ; the area on 8 occupied by these two pores and their 

 accessories is a square with its diagonals longitudinal and trans- 

 verse, and from its lateral angles a line passes outwards and fades 

 into the posterior margin of the segment (Chapman). 



Pupal habits. — The pupa of E. elpenor is remarkable in that 

 it is able to work its way out of its cocoon before emergence in 

 the same way as does the pupa of Dimorpha versicolora. This is 

 effected by means of the spines already described (anted,, p. 78) 

 as existing along certain of the abdominal segments. Bacot notes 

 that he had larvae, in 1895, that pupated in a flower-pot filled with 

 moss, and formed long and rather narrow cocoons by spinning the 

 moss stems together with a small amount of silk. These cocoons 

 were all vertical, or nearly so, and were from half-an-inch to an 

 inch longer than the pupa, the top being usually left open; in the 

 spring the pupae were observed to move up and down in their, 

 cocoons, and one, in particular, used to push itself half-way out on 

 sunny days, going down again in cold and dull weather, and it was 

 half-way out of its cocoon when the moth was disclosed ; some moths, 

 however, emerged from pupae that had their heads only protruding, 

 while, in other instances, no movement on the part of the pupa 

 had apparently been made before emergence. Ransom observes 

 that, about a month before emerging, some pupae work their way 

 out of the cocoons and travel several inches away from the cocoon ; 

 the movement is accomplished by means of strong spines on the 

 abdominal segments ; the abdomen is moved from side to side, 

 and the spines, catching hold of any substance with which they come 

 in contact, enable the pupa to progress. Russell notes that as 

 early as May 14th, the pupae (which remained in the moss in 

 which the larvae had pupated) prepared for emergence. They 

 raised themselves on end, and, in many instances, stood out of the 

 moss practically upright, although about one-third ot the pupae 

 appeared to make no movement at all. A note in the Ent. 

 Wk. Int., vii., p. 109, states that a pupa worked itself out of its 

 loose cocoon, and lay quite bare on the moss for 5 or 7 weeks before 



