PACHYTHELIA VILLOSELLA. 409 



margin slightly bilobed ; maxillae outside this and fully as long, 

 rounded, but broadly triangular, Avith a very definite palpal prolonga- 

 tion, that, in one specimen, is continuous with the palpus on one side 

 and has a joint between on the other, showing that the deep differing 

 proximal joints are not absolutely obsolete ; antenna? very broad, as 

 also base of first legs, first femur long and broad ; all these parts adhere 

 to pupa-case on dehiscence by the third legs which are attached to 2nd 

 and 3rd abdominal segments by their tips, the latter hardly visible 

 between the apices of the wings. In one specimen, the second legs are 

 well seen, in another, the antenna? cover all but a very narrow margin 

 that needs looking for. The tubercles as in the larva (Chapman). ? . 

 The female pupa is about 18mm. in length, when fully stretched and 

 about 4mm. in diameter, fairly cylindrical, rather narrowed in front 

 and with the head bent forwards, the anal extremity more bluntly 

 rounded. The colour is a rather light brown, without any darker 

 areas except slightly where the chitin of each segment is medially and 

 dorsally a little denser. The spiracles are well marked, and the larval 

 prolegs leave very pronounced oval or nearly circular scars. Dorsally 

 there are some roughnesses or wrinkles on abdominal segments 3, 4, 5 

 and 6, not very marked and less apparent on the other segments, but 

 beyond this there is no trace of the anterior dorsal row of hooks. The 

 intersegmental hooks, however, are well-developed, they occur on the 

 divisions 2-3, 3-4, 4-5, and 5-6, being especially very large and strong 

 on the latter incision, where they are definitely in one straight level 

 row ; on the anterior incisions they are less regular and might even be 

 regarded as showing indication of a second row. When the segments 

 are stretched they point directly forwards and though really minute, 

 look under a lens terrible and formidable implements. As there are 

 no hooks for forward progression these cannot be functional for 

 regression and must subserve the purpose of a cremaster, in retaining 

 the pupa-case in its place, on the partial emergence of the imago, 

 when pairing, &c. The anterior trapezoidal (i) and supraspiracular 

 (iii) hairs are small but very distinct, the posterior trapezoidal hair 

 (ii) is comparatively large and less than half the distance from the 

 middle line of the dorsum that i is, but measured antero -posteriorly, 

 i is about the middle of the segment, ii halfway from this to posterior 

 border ; at some distance below and behind the spiracles are two small 

 hairs (iv, v),anda solitary one, rather larger, further down and immedi- 

 ately in line (transverse) of spiracle ; two others are just above and close 

 to scar of proleg, the posterior being the upper, if there be a third, it can- 

 not be detected ; there is also one on either side beneath the prolegs. On 8 

 and 9 one or two of these hairs have distinctly recurved tips. The thoracic 

 and 1st abdominal segments are narrower than the others behind, and 

 the 1st abdominal spiracle is rather small, the wings and legs are 

 represented by certain folds and wrinkles, without definite structure, but 

 which, by their position, are recognisable as representing these parts. 

 The face and leg region is delicate and pale ; the venter is also paler 

 than the dorsum, Avhich is often deep brown, but could hardly in any 

 case be described as black. 



Dehiscence. — $ . The head and face-parts separate in one piece, 

 with the legs, &c, but are held in place by the internal dissepiments, so 

 that they remain close to their natural position ; the antennte may 

 slightly separate from the first legs ; the thorax splits dorsally to not 



