182 BRITIHS BUTTERFLIES. 



for a few hairs. The 3rd abdominal (l-3mm. broad) is more like the 

 1st, with a flatter anterior half, and the posterior presenting three 

 ridges, the hairs are more numerous on the posterior half. The 4th 

 abdominal is (extended) about 2mm. long, and has about four equal 

 portions, viz., two anterior fairly flat, then an intersegmental subseg- 

 ment, with the fine sculpture of intersegmental membrane, but 

 possessing hairs, etc., and then the true intersegmental membrane. 

 The hairs are numerous, and, whilst giving the impression of transverse 

 alignment, do not, in fact, anywhere afford very definite lines. The 

 5th abdominal segment is much like the 4th, 2mm. wide (extended) ; 

 there are many hairs, of which two rows on the intersegmental subseg- 

 ment and the row in front of this are fairly in alignment. The 6th 

 abdominal is slightly narrower, but otherwise much the same. 

 The 7th abdominal is about l-3mm. wide, the intersegmental sub- 

 segment narrow, but what is most remarkable is that (in this 

 dehisced pupa, a $ ) it easily extends from the 8th and displays a 

 narrow border of real intersegmental membrane. The 8th abdominal 

 is about 1mm., fairly uniform in surface, but with the posterior inter- 

 segmental subsegment distinguishable ; the hairs are here rather 

 longer than in front and more slender, 02mm.-024mm. long, but 

 always ordinary hairs, nowhere any trace of the vase (or trumpet) 

 hairs of the larva. The 9th abdominal segment is a dark, wrinkled, 

 narrow portion, not easily distinguished from the 10th, which is repre- 

 sented by the cremastral spine, a conical piece l'5mm. broad at base, 

 nearly 2mm. long, dark brown in tint, longitudinally fluted or ridged, 

 but so that there is, on either side of the dorsum, a larger and higher 

 ridge underlying the others. It carries a good many hairs, roughly 

 and inaccurately (in dorsal view), say four rows of four each. The 

 terminal bundle of hooks (cremaster proper) is a little obscured by silk, 

 the hooks are closely packed together about 0-2mm. or 0*25mm. long, 

 slightly curved, and ending, each, in a recurved portion, nearly a 

 complete circle, but so closely curved that there is no central opening. 

 The sculpture of the 1st abdominal segment was described as in trans- 

 verse ridges, but this obtains less and less on each following segment. 

 The front margin (one-fifth of segment) of the 3rd abdominal segment has 

 a series of little pits, not spherical but rather cylindrical hollows, with 

 a plane surface between them ; this obtains on the first element of the 

 4th abdominal segment and largely on the 2nd. On the 6th and 7th ab- 

 dominals there is still a little tendency at the posterior margin of the seg- 

 ment for the plane surface to run into ridge-like form between the pits. 

 With this in mind, it is easy to persuade oneself that the ridges on the 

 1st segment are really a special modification of the pitted surface. As 

 a matter of colour, the first few segments have, trapezoidally placed, 

 small, brown marks, that look slightly hollow, but probably due to the 

 false perspective caused by their coloration, and being free from hairs ; 

 there are one or two similar marks more laterally placed. On the 2nd 

 and 3rd abdominal segments are certain marks, probably to be classed 

 as lenticles, and traces of them exist on other segments ; they are a 

 little further out than the presumed position of tubercle ii, one on the 

 front and one on the back portion of the segment ; the posterior is the 

 most typical, and is a raised brown mark with raised dark sharp margin, 

 narrow from front to back, longer in line across the segment. Taking 

 a lateral view, the spiracles are not dissimilar to the mark last 



