CYANIRIS SEMIARGUS. 293 



dark, hardly black ; no yellow except that deeply sunk in larva. At 

 rest it shows the upper surface of prothorax nearly level or parallel with 

 the lower surface, and well below the slightly overhanging mesothorax, 

 the head being quite retracted. The lower flange shows quite a fan 

 of hairs (? those of tubercle vi) above prolegs, on 3rd- 6th abdominal 

 segments, less marked on other segments. The transparency of skin 

 is very notable. Dorsa of segments rounded at flange. Larva looks 

 natter than it is, but is so in latter segments. The honey-gland 

 is an oval patch without hairs or lenticles, ?".<?., as if it were permanently 

 a little open. The fans on the 8fch abdominal are difficult to find on the 

 older larva, and show only an impressed point, hardly visible ; on the 

 younger they are each a pale circle, somewhat conspicuous. In the 

 older larva this conspicuous area is retracted. As part of its change 

 for spinning up (?) the larva is very transparent, especially laterally, 

 and looks as if it contained a yellowish mass centrally, with lateral 

 processes (the oblique lines) (Chapman). 



Foodplaxts. — Trifolium pratensis (Brabant, Chapman), Anthyllis 

 vulneraria (Assmuss, Frey, Eothke. Paul and Plotz, Loffler, etc.), 

 flower-heads of Armeria vulgaris (Zeller, Frey), Melilotus officinalis 

 (Schneider, Kranz), prefers the seedpods of Melilotus and Anthyllis, 

 probably also on Cytisus sagittalis and Lotus comiculatus (Hofner). 

 [Statice armeria (Laplace) ? miscopied for Armeria vulgaris.] TJlex 

 europaeus, in confinement (Frohawk). [The remark by Breit (Soc. 

 Ent., xiv., p. 99), that the eggs are laid on Scabiosa succisa and 

 other low plants, is perhaps hardly worthy of notice.] 



PupariUxM. — In confinement the larva, noted as the more forward (antea 

 pp. 281-5), spun a few vague threads, that were certainty not a girth, and 

 were less than an apology for a cocoon; the larval skin fell free, and the 

 pupa remained in situ, but became loose by merely moving the box in 

 which it was; still it had a slight hold by some cremastral structure to 

 which some threads still adhere, and there are some very definite hooks 

 at the end of the 9th and 10th abdominal segments. The larva noted 

 as the less forward (antea pp. 284-5), which was apparently a stronger 

 and more healthy one, being larger and pupating a little later (both pro- 

 duced 2 imagines), fastened itself within a few vague threads by a girth 

 and cremaster to a portion of filter paper (used to keep food moist); the 

 girth consists of very few threads, being very slight, and passes from 

 points opposite the middle of mesothorax and somewhat backwards so 

 as to cross the front margin of the 6th abdominal segment. After 

 emergence, this girth retained the pupa in position, although the hold 

 of the cremaster had given way (Chapman). The full-fed larvae of 

 the summer brood kept in confinement in a flower-pot filled with 

 earth, in which red-clover plants were growing, at the commence- 

 ment of July, sought out a suitable spot where the surface of the 

 earth met the upright wall of the flowerpot, and here spun some silken 

 threads, from the side of the pot to the nearest clover leaves, 

 holding down the latter, and changing to pupa? in the little cocoons 

 thus formed. The pupal stage, in 18^6, lasted 17 days, July 

 7th-2lth, at Cambrai (Brabant). The larva? spin up, in confine- 

 ment, on different parts of the plant, on stems, leaves, and flowers ; 

 in each case a very slight cocoon was formed by a few strands of 

 silk, the larva being attached by the anal claspers to a pad of silk, 

 and a cincture passed round the middle ; the pupa is, therefore, 



