AGRIADES THETIS. 375 



species we are describing, this pocket is usually everted in the 

 disturbance of emergence, or in mounting the pupa afterwards, so that, 

 being internal, it is not easily discoverable in the unmounted pupa, 

 and is apt to be so displaced as to escape detection in the mounted 

 one ; some other pupa?, however, such as that of Plebeius argus (aegon), 

 in which it has a more solid structure, enable one to trace it here and 

 elsewhere. Until this is grasped, the small, semi-chitinised end of 

 the maxillae is puzzling. The 1st legs have a region with lenticles, 

 with a hair or two about l*3mm. from top, and again an odd one 

 about 1mm. lower (i.e., 1mm. from where it disappears); basal to the 

 lenticular region, the network on this leg has alternate bands of rather 

 square and of elongated cells. The 2nd leg carries several lenticles close 

 to its lower end; it also shows some indications of the general arrange- 

 ment of netting. The antennae are narrowed where they pass the side 

 of the head, and the basal enlargement is small; they widen just 

 where the maxillae go under them (a good way short of the club) and 

 then narrow again ; the network runs into a rather transverse arrange- 

 ment at intervals (joints of antenna?) (Chapman). The pupa is about 

 12mm. in length, dumpy in figure, the profile of the back swelling 

 out at the thorax and dropping again at the waist, the abdomen 

 sloping off in a curve to the rounded anal end ; the ventral profile 

 much straighter, though still with a slight curve ; the wing-cases 

 reach more than two-thirds of the whole length, and the widest part 

 is just where they end. The thorax and wing-cases are slightly glossy, 

 the abdomen granulated ; some very small hairs scattered all over 

 surface ; the colour at first greenish on the wing-cases, greenish-brown 

 on the rest of body ; afterwards ochreous all over, finally turning very 

 dark the day before the imago emerges (Buckler). 



Sex of PUP.E.— The determination of the sex in the pupal stage 

 appears to be somewhat difficult, or, indeed, impossible. Chapman 

 reports that, of 20 pupae examined, 8 were supposed to be $ s, 6 $ s, 

 and 6 others doubtful, judged by form and size ; these produced 

 respectively 4 $ s and 4 $ s, 2 $ s and 4 $ s, 3 $ s and 3 $ s. A 

 careful examination of the pupae, proved by breeding to be respectively 

 $ and 2 , shows no differentiation of end segments, owing to the 

 incurving ventrally. In both sexes the 8th abdominal segment is just 

 evanescent ventrally, the 7th and 9th (or 10th) abdominals meeting 

 here, and the two sides of the 8th abdominal ending in a point on 

 either side ; the 9th and 10th abdominals appear to be fused, unless 

 the 9th abdominal is entirely hidden ; a slight duplication of the 

 incision at either side, does not settle the point, but is the only indica- 

 tion that here are two segments instead of one. 



Colour-change of pupa of Agriades thetis.— A pupa, appar- 

 ently just changed (9 a.m., October 8th, 1909) is of a translucent pale 

 green colour; the anah end withdrawn entirely from cast skin, so 

 that there are no attachments whatever. About 11 a.m., the wings 

 appear to be rather darker green towards base, paler towards outer 

 margin, the prothorax and frons very pale green, the rest of the body 

 already much greyer, a darker mediodorsal line from mesothorax to 

 anus. October 15th (pupa now a week old) : The meso- and metathorax 

 green, wings pale green, the base rather fuller coloured, as if body of 

 meso- and metathorax was showing through ; also more definitely 

 green marginally (i.e., against " Poulton's line"), the central area of 



