166 THE entomologist's kecord. 



The wings came clown to their full extent during the moult, and the 

 maxillae, etc., with them. The moult did not occupy many minutes. 

 The maxilhe projected nearly a millimetre (guessed, not measured) 

 beyond the wings. 



The 5th abdominal and following segments were very movable and 

 active, more capable ot various contortions exercised m pushing back 

 the larval skin than they ever appear to be during larval existence. 

 In only a few minutes after the moult was completed thej^ drew 

 up and assumed very nearly their final pupal positions. Whilst these 

 segments were still very movable, the intersegmental membranes were 

 wide and flexible to allow of those movements. As soon as the 

 descending skin had left the 4th to 5th abdominal incisions the 

 membrane here was alternately stretched and closed up as the peri- 

 staltic movements of moulting went on. At first this was, in fact, 

 simply the ordinary action of an incision when the segments on either 

 side are in relative motion, and the projecting maxilhe, extending, 

 saj', a millimetre beyond the wings reached halfway across it when 

 it was extended, when it was closed they were between its folds. 

 Then the extent of the incision on either side in front seemed to 

 diminish, as if by shrinking of the membrane, but in front over the 

 maxilire it continued to open, at each movement, the margin of the 

 fifth segment getting gradually sharper and sharper in the middle 

 line. Then when the incision closed at each movement, this margin 

 came over the ends of the maxill;D, which seemed a little bent back- 

 wards. Next, they were not completely uncovered when the incision 

 Avas most open, but elsewhere less opening was accompanied by 

 gradual shrinking of tlie intersegmental membrane. Finally, the 

 maxillae were permanently covered. 



The pocket was thus seen to consist of the intersegmental mem- 

 brane kept stretched as it was during the moult, whilst throughout 

 the rest of its circle it contracted together, so as to be little more 

 than an inappreciable cement, joining the margins of the 4th and 

 5th abdominal segments together. 



The photographs from the Jlritis/i Lepidoptcra are here repro- 

 duced. 



Explanation of Platks. 



Plate XI. — Fig. 1 — Front of 4th and 5th abdominal segments of pupa of Flebeui-t 

 argiis, showing projection ot maxillffi into pupal interior through the 

 4-5 abdominal incisions, x (JO. 

 Fig. 2. — Front of 4th and 5th abdominal segments of pupa of Plebeius 

 argits, with portion of 5th segment cutaway to show maxilliE project- 

 ing into interior of pupa, through 4-5 abdominal incision, x 60. 



Plate XII. — Diagram of condition of portion of pupa when involved; the inter- 

 segmental membranes are still uncontracted and admit of free 

 movements. 



4. Fourth abdominal segment. 



5. Fifth abdominal segment. 



W, wings. A, antennas. M, maxillffi. 



P^ P^. Posterior border of fourth abdominal segment. 



Pg Pg. Posterior border of fifth abdominal segment. 



Ar, Ag. Anterior border of fifth abdominal segment. 

 These are the borders as seen in the mature pupa, but at the moment illustrated 

 they are separated by an expanse of nearly structureless membrane, I. 4-5, I. 5-6, 

 and which admits of the pupal movements. Very rapidly, however, these mem- 

 branes shrink up, so that in the mature pupa P^ coincides with Ag, and the inter- 



