396 BRITISH LEPIDOPTERA. 



thoracic or first two abdominal segments, but the abdominal segments 

 3-7 have each, along their dorsal margin, and terminating laterally 

 against the marginal flange, a row of spines. These are dark brown 

 in colour, almost black, are about 50 in number on each segment, and 

 are directed backwards from the top of a somewhat raised ridge, which 

 may be regarded, however, as the bases of the spines. The spines 

 vary a little in size, but without any very definite order, and 

 are rather larger midway between the dorsum and the side. On 

 the 8th abdominal segment is a similar row, but smaller, owing 

 to the smallness of the segment, and terminating at about the 

 same point in the segment, though the lateral flange terminates 

 on the previous segment ; a still.srnaller row is found on the 9th ab- 

 dominal segment. The remainder of the 8th, 9th and 10th abdominal 

 segments is smooth, except the median cicatrices on 9 and 10. 

 These three segments in the pupa described are fused together, the 

 individual being a male. It is noticeable that the bases of the 

 maxilke, instead of forming a nearly transverse line, extend upwards 

 in the middle line, so that the grooves between them and the cheeks 

 are in line (inwards) with the groove on the other side, between the 

 cheeks and clypeus, forming a St. Andrew's cross, the centre just 

 below the labium. Similarly the external angle extends upwards as 

 well as outwards, nearly reaching the antennae, and is a distinct trace 

 of the base of the maxillary palpi (Chapman, in lift., June 8th, 1897). 



DEHISCENCE. In dehiscence, the antenna-cases remain attached 

 to the head, but free throughout their length. The third pair of legs 

 are quite separate, and the wings are separate from the other appen- 

 dages, but the head, eyes and other front appendage-cases remain 

 soldered together. The dorsal head-piece remains attached to the 

 1st thoracic segment, but free from the eye-covers. The dorsal slit 

 extends backwards nearly to the posterior margin of the metathorax. 

 The appendages are now obviously free from the first abdominal 

 segment (Chapman). 



FOOD-PLANT. Humex acetosa. [Reputed also, by Kaltenbach, to 

 feed on Helianthewum vtdyare (= food-plant of A. yeryon), lluwe.i- 

 acetosella, Centaurea scabiosa (food-plant of //. globvlariae), Globiilaria 

 vulyaris. Milliere says "many low plants," and gives Oistus salri- 

 foliits as food of the larva of the var. micam. We found this form 

 near Cannes (on Milliere's own ground), in April, 1898, obtained eggs, 

 larvse from which fed in most approved fashion on li. acetosa. The 

 imago sits on Cistus flowers, and probably this misled Milliere.] 



HABITS AND HABITAT. Meadows and pastures, especially on the 

 outskirts, of woods, are the favourite haunts of this species. Forty 

 years ago it was abundant in Hammersmith marshes (Taylor). On 

 the Sussex Downs it is found in a sheltered valley known as Oxsteddle 

 Bottom, but this locality is very unlike its usual meadow haunts, and 

 in Guernsey it is confined to patches a few yards square on the southern 

 cliffs ; near Brighton it is found in Hollingbury Combe, a place over- 

 grown with long tangled grass, whilst about three miles from Gloucester 

 it occurs in a damp vale abounding in sorrel. In the New Forest it 

 occurs on the railway bank, and at Chingford it frequents a low-lying 

 damp field at the back of the " Woodman," and settles here (as else- 

 where) on the flowers of Lychnis Jios-cuculi. At Hemsby, near Yar- 

 mouth, it is found on the sand-hills, which are covered with marram 



