304 Dr. T. A. Chapman's Contribution to 



day the whole pupa darkened, but one could not guess 

 whether owing to the extension of ailment or to the ordinary 

 progress of development before emergence, the latter proved 

 to be the case. The original black patch must therefore 

 have been a precocious maturing of the one portion of 

 wing, owing to some cause that also produced the variation 

 noted in the imago. 



The presence in the pupa of a dorsal headpiece is evidence 

 that the pupa is not highly evolved, but the opening for 

 emergence of the abdominal incisions before the 4th shows 

 a more primitive condition than occurs in any obtect pupa 

 of the Heterocera. 



The pupa of A. ther sites presents at least as great an elaboration 

 of the maxillary pocket as I have noticed in any other pupa, which 

 has led me to observe certain facts in connection with it that I 

 ought to have seen before but overlooked. (PI. LII. and LIII.) 



I first called attention to this " pocket " in Tutt's " British 

 Lepidoptera," vol. x, p. 226, and in the Ent. Rec, vol. xxv, p. 165, 

 I related how it is formed as observed in the pupal moult of Agriades 

 coridon. 



The overlooked circumstances arc, that the intersegmental 

 membrane of the two following incisions presents certain involu- 

 tions and persistences in some species, that do not seem directly 

 associated with the present development of this pocket; these are 

 well marked, for example, in the pupa of A. heUargus, in which 

 there is a very definite and symmetrical fold in the next (5-6) 

 incision just below the pocket and a longer but shallower one in 

 the following incision (6-7, abdominal segments), and similar 

 involutions, on a slightly smaller scale in the same three incisions 

 half-way between the mid-ventral line and the spiracles. It is 

 not easy to suppose these latter have any direct relationship to the 

 pocket. 



In my original figure of the pocket in P. argus {cegon),* the 

 extremity of the pocket and two obvious folds in the next incision, 

 suggest a strong approach to the condition in A. ther sites. 



In a4. thersites the involutions of the membrane, as what may be 

 called subsidiary pockets, are almost exactly the same as described 

 above in reference to A. bellargus, where the development of the 

 pockets between segments 5 and 6 is very easily seen. 



Though I call these pockets they are really merely folds, i. e. the 

 two walls touch one another and contain no cavity. This second 

 pocket in A. thersites is very large and looks at first glance as if 



* Tutt's Br. Lep., vol. x, PI. XXXIII, reproduced Ent. Rec. /. c. 



