412 BRITISH LEPlDOPTEfcA. 



lines, composed outside of grains of earth, very firmly united to a few 

 fibres of grass-roots. On removing the earthy particles, the inner 

 cocoon of opaque greenish-white soft silk is found. This is strong and 

 elastic, resembling that of Cosmotriche Rotatoria, very soft and closely 

 woven anteriorly. Zeller says the brownish-grey cocoon is placed on 

 the surface of the earth, well-concealed, close to the roots of plants. 

 Buckler's description of " greenish-white," and Zeller's of "brownish- 

 grey," for the silk, suggests some variation in this respect. The pupa 

 emerges from the cocoon, and travels to a little distance from it before 

 the imago appears. Buckler found the pupa-skins lying on the bare 

 earth, and only the old shrivelled larval skin in the otherwise empty 

 cocoon. 



Pupa. — The pupa is about 13 mm. in length, of moderate sub- 

 stance throughout, with prominent thorax, the wing-covers short, but, 

 towards their ends, projecting a little free from the body, the long 

 antenna- and leg-cases are all free from the body, and seem to 

 be suggestive of locomotion even before the disclosure of the moth ; 

 the deeply-divided abdominal rings have each, on the back, near their 

 beginning, a transverse ridge, thickly set with hooks pointing back- 

 wards, the top of the abdomen rounded off in a blunt point ; the 

 colour of the head, thorax and wing-covers is dark olive-green, and 

 very glossy, the leg-cases and abdomen are of lighter shining 

 green, and the hooks black (Buckler). Chapman gives many further 

 details. He says that the pupa is straw-coloured, narrower and 

 straighter, but reminding one a good deal, at first sight, of Cochlidion 

 avellana {testudo). A line down the middle of the venter would be 

 straight, the back is rounder, the waist is very slightly marked, flat- 

 tened from back to front, so that the width is nearly twice the antero- 

 posterior diameter. There is a slightly developed lateral flange carry- 

 ing the spiracles. The first two abdominal segments, and possibly the 

 third, are apparently fixed, but are free, so far that the appendages 

 are not firmly attached to them, though closely appressed. The first 

 abdominal spiracle is sheltered by the hind-wing, as by a flange lying 

 over it, but is not covered, and, having a ridge just dorsal to it, is, as 

 it were, in a pit, and is quite free and functionally open ; the second 

 only just escapes being covered by the hind-wing. There are dorsal rows 

 of spines along the anterior margins of the abdominal segments 2, 3, 4, 

 5, 6, 7, 8 and 9, very close to the anterior margins, consisting of 

 14-18 very fine brown spinous points on either side, terminating out- 

 wardly at the lateral flange. The anal armature is nil, nor have any 

 other hairs or spines been detected. The posterior margin of each 

 abdominal segment, and even of the third thoracic, is a raised ridge, 

 and overhangs the next segment. The head very definitely extends 

 outside the general level, and forms a knob or boss, whether viewed 

 laterally or ventrally. The eyes and eye-plates extend very low down 

 laterally, and the mouth -parts are placed unusually forward between 

 them. The legs, antenna, etc., are rounded, and stand away from 

 the body of the pupa. The first pair of legs and the antennas are 

 most prominent ; the second pair and the maxillaa dipping beneath 

 them, the maxillse partially, the legs entirely, the tarsi appearing again 

 beyond them ; the third pair of legs extend several segments beyond 

 the wings, the tarsi of the second pair and the antenme slightly so ; 

 the posterior cephalic plate is very distinct, but there is no definite 



