POLYOMMATUS ICARUS. 201 



only four instars in the summer brood, this is possible, but I think 

 my having omitted to observe properly is more likely (Chapman). 

 Final instar: Head very small, glabrous, oblong, porrected in 

 crawling, otherwise withdrawn into the 2nd segment and totally 

 concealed ; body convex above, flattened below, rounded at both 

 extremities, dilated and lobed at the sides; the divisions of the 

 segments conspicuously marked ; spiracles situated much above the 

 lateral margin, the posterior pair dorsal ; whole surface covered with 

 extremely minute warts, each of which emits a hair. Colour of the 

 head intensely black, of the body green, sometimes bright apple-green, • 

 at other times dull olive- green ; a mediodorsal stripe rather darker ; 

 a narrow lateral stripe below the spiracles, but above the lateral lobes 

 much paler, almost white ; between the dorsal and lateral stripes 

 there are, on each side of each segment, three pale oblique lines, 

 their inclination being from the anterior to the posterior margin of 

 the segment ; the minute warts black ; legs, claspers, and ventral 

 surface of the same green hue as the body (Newman, Ent., iii., p. 15). 



Quiescent stage preceding pupation. — A larva was laid up for 

 pupation several days before it pupated, on March 24th, 1908. Seen 

 against the light during this period, it has a curious appearance ; it 

 seems filled with blue fluid, almost the blue of sulphate of copper, 

 but not very dark, with a denser body floating centrally in it, dorsally 

 and ventrally, the clear fluid seems nearly 1mm. thick, the dense 

 interior about 3'5mm. thick. This may be really fluid, or an exagger- 

 ation of the Lycasnid larval skin, which often seems to be a very thick 

 gelatinous material. The specimen has pupated this morning, March 

 24th (Chapman). 



Foodplants. — Onobrychis hedysarum (Engramelle), Ononis spinosa, 

 Astragalus glycyphyllos (Ochsenheimer), Trifolium, Melilotus, Genista 

 vulgaris (Neustadt teste Treitschke), Lotus corniculatus, Ononis arvensis 

 (Hellins), Ornithopus verjmsillus (Buckler), Trifolium, (Caradja), T. 

 pratense (St. John), T. repens (Bird), Medicago sativa (Curtis). M. 

 minima (Miihlig), M. falcata (Favre), flowers of Ulex euroijaens 

 (Chapman), Spartium (Riihl), [Fragaria vesca (Borkhausen, Ochsen- 

 heimer, Viret, Frionnet, etc.), F. collina (Assmuss), strawberry leaves 

 (Curtis). In spite of the oft-repeated Fragaria, of authors this 

 requires confirmation.] [Grasses . (Kirby), is certainly an error; 

 probably assumed from Lewin, Brit. Ins., p. 80; dahlia (Goossens teste 

 Ragonot) also wants confirmation] ; wild liquorice (Curtis). 



Puparium. — This species, as those of Agviades, pupates without 

 forming a definite girth or pad ; indeed, the absence of cremastral 

 hooks has long since shown that the pupa could not possibly be 

 supported by its. anal extremity. Chapman, however, notes (in litt.) 

 that he observed in a mounted pupa of P. icarus, close to tbe cremastral 

 area, a scrap of larval skin, which serves to illustrate how, in this and 

 the allied species without cremastral hooks, the pupa, nevertheless, 

 often remains suspended by attachment to the larval skin, which in 

 turn still adheres to the few threads spun by the larva. Speaking of 

 some larvae that had pupated on April 3rd, 1909, Chapman says (in 

 litt.) that four larvae " went down " amongst the loose flowers of JJlex 

 in the bottom of the tin and spun four or five of these together as a 

 cocoon ; two accepted little bits of paper rolled up into a quill-like 

 tube. It is a little difficult to see what there may be to call a girth, 



