452 BRITISH BUTTERFLIES. 



the dorsal line, the 2nd abdominal segment is about l-2mm. long, the 

 3rd I'Omm., the 4th and 5th about the same, the 6th rather less, the 7th 

 07mm., the 8th 05mm., whilst the 9th and 10th together form approxi- 

 mately a circle of about l*6mm. The head-piece has some 20 hairs in 

 front, two at the base of the mandibles. The labrum and mandibles are 

 both marked off by sutures. The mandibles meet in the middle line for 

 about 0-08mm., and there is a minute square of labium, between them 

 and the maxillae. The labrum ends in a point, the apex of an angle of about 

 90°. The face projects with a more than usually sharp and pronounced 

 angle, between the bases of maxillae and first legs. The glazed eye is 

 a dark line, with a broad, smooth, space outside it (the eye proper being- 

 marked by the points of the eye facets), and centrally (against antennas 

 and first leg) is a portion of ordinary surface carrying reticulations and 

 8 hairs, and three minute lenticles. The maxillae are about 3 -6mm. 

 long (this pupa is rather a small one, and all dimensions are possibly 

 below average); the basal, broad, portion is short, and the ends dis- 

 appear by the antennae meeting over them in middle line ; the antennae 

 proceeding another 2*7mm. between the wings. The 1st legs are very 

 broad (and curved) basally, 2-3mm. long, 0-6mm. broad (where broad), 

 ending in a point between maxillae and 2nd legs, and against antennae 

 for about 0-4mm. The 2nd legs are slender, about same length, and 

 end in points at each end. The sculpturing of the surface is a network of 

 raised ribs, generally darker than the area between them ; the ribbing 

 is, however, rarely closed into a network over any considerable area, 

 but the loose ends terminate in a somewhat dendritic wa) T , branching 

 into paler (and also smaller) ribs before fading into the general surface. 

 In many places, however, a further structural element of the ribs is 

 evident, these dendritic endings being continuous with a very small- 

 celled netting that fills up the spaces between the larger ribs ; this 

 small netting is very much on the same scale as the network on the 

 larval surface, and the inference is strong that it is the same structure. 

 The large ribbing would then be certain lines of the larval network 

 preserved and emphasised, the remainder disappearing except where, 

 as here, some remnants of them are preserved. The points on the 

 ribbing would correspond with the points at the angles of the network 

 on the larva. The points on this pupa are comparatively few, and, 

 though always on the network, do not occur solely at points of inter- 

 section, unless one carefully notes the minor fading branches ; they 

 are about three times as broad as the ribs, are round, and have a 

 central bright point, and some indication of radial structure ; as a 

 special peculiarity, they appear to be like buttons laid on the top of 

 the ribs, instead of thickenings in the course of them, as they do in 

 most Theclid pupae. The hairs are everywhere of almost identical 

 pattern and length, about O-lSmm. long, tapering regularly till near 

 the end, then more rapidly, spiculated, but very faintly, till near the 

 end, when the rapidly-thinning portion seems to dissipate itself in 

 larger and more numerous spicules, nearly at right angles to the stem and 

 often nearly as long as the stem is thick where they arise. Lenticles are 

 sparsely scattered everywhere, and are -very like (in size, etc.)hair-bases, but 

 the finely dotted closing membrane at once distinguishes them. Hairs 

 and lenticles, as in other Lycaenid pupa?, occupy interspaces of the 

 network, and are never on it. The wings and appendages are, as 

 usual in allied pupae, netted, but have no hairs, lenticles, or network 

 points. There is the exception, however, that I now recognise, viz., in 



