CELASTRINA ARGIOLUS. 437 



thickest at one-third of its length and narrow towards the end, hut 

 would be nearly as long as the middle hair, if, instead of being 

 rounded off, it continued to a point. This different structure 

 possibly separates this hair, as an accessory, from the other two, 

 which would then be iv + v - Below these are two slender 

 hairs, the longest about Olmm. a little towards front of segment 

 (vi ?), and the smaller one, 0-04mm. or so, also very sharp and 

 slender, a little above and in front of it. These are also present on 

 the three thoracic segments at bases of legs. Two much smaller and 

 finer hairs are found just outside prolegs. The lenticles on abdo- 

 minal segments are two between tubercles i and iii, nearly in line 

 with them, and about equally spaced, but the upper one a little in 

 front of the exact line ; they are nearly 0-02mm. in diameter with 

 dotted lumen ; the 1st abdominal has a rather smaller lenticle at front 

 margin below spiracle, a little above iv -J- v ; on the 2nd abdominal 

 there is one between iv -f v and vi in one specimen (but not in 

 another), the 7th and 8th abdominal segments each have one in the 

 same situation. The disposition on the more specialised segments 

 differs, in the way usual in Lycsenid larvae, from that on the abdominal 

 segments. The prothorax has a plate nearly square (but placed crosswise) 

 with the front angle rounded (pi. xxv., fig. 1) ; it has a pair of hairs in 

 the middle in front, outside each of which is a large lenticle ; in the 

 middle another pair of hairs a little further apart than those in front ; 

 these are very long (about O'Bmm.), and, near the posterior border, 

 another pair, again further apart, so that these three hairs on each 

 side are in lines diverging backwards ; these posterior hairs are rather 

 shorter than the middle hairs but larger than those in front ; there 

 is a shorter hair in each outer angle ; on the segment, on each side, 

 along the front border of the plate, are the usual three hairs, the 

 front one rather short ; a fourth, almost in line with these, is further 

 back ; in the front of the spiracle is a lenticle and three hairs ; 

 two fine ones are at base of legs. On the mesothorax are, on 

 each side, two hairs (apparently) of tubercle i, one curved back- 

 wards, one forwards ; a lenticle below the first one ; then two hairs 

 in line with tubercle ii, but a good deal larger than on the follow- 

 ing segments (02mm.) ; then a long space, corresponding with 

 the lenticles and tubercle iii of the following segments, before the four 

 flange-hairs (012mm. to 014mm. long), the 2nd and 4th level, the 

 1st higher, and the 3rd lower, are reached. The metathorax is more 

 like an abdominal segment— tubercles i and ii the same, only one 

 lenticle, only one hair at iii, and four hairs on flange as on mesothorax. 

 The 7th abdominal segment has no seta on ii ; the upper lenticle rises 

 up into its place ; it has only one hair on iii, nearly in front of the 

 other lenticle. The 8th abdominal has a hair on i, and a lenticle just 

 outside it, then the spiracle, and three marginal hairs ; the lenticle seems 

 to be incorporated with the base of i, the same occurs (or all but occurs) 

 on the 7th abdominal segment. The 9th abdominal seems unarmed 

 above the three flange-hairs, but it is difficult to say precisely what is 

 the 8th, 9th, or 10th abdominal, especially along the flange. The 

 10th abdominal has a large anal plate, about 008mm., without hair or 

 lenticle, and a small plate on each side, about in line with spiracle, 

 and hardly larger, faintly tinted with darker, like the anal plate ; it is 

 on the front margin of the 10th, or may, indeed, belong to the 9th ; 



