LYCAENA ARION. 341 



T 3 gin. across the middle, the head rounded, thorax convex rising into a 

 slight dorsal ridge, metathorax and first abdominal segment sunken, 

 abdomen swollen at the middle and curving to posterior segments 

 which are rounded ; and segment is compressed to ventral surface, 

 cremastral hooks absent, wing ample, sw r ollen, and rounded across the 

 middle and extending to the fifth abdominal segment. Pupa found 

 wild, July 12th, 1905 (Frohawk). 



The pupa is very similar to other pupa? of Lyca?nids, but of rather 

 paler and lighter texture than those of most others. There is no pocket for 

 ends of maxilla?. Hairs are few, but may be found near the spiracles, 

 but small. They are terminally spiculated, with suggestions of being 

 allied to the umbrella hairs of Chysophanids, their length is only 

 O06mm. The lenticles are also very small, (see pi. xliv., xlv.). These 

 plates w 7 ill also show the character of the network of the skin 

 sculpturing; in this respect pi. xliv., fig. 2, is excellent. 



A portion of the cremastral area is shown pi. xliii., fig. 2, x 100. 

 It shows -no hairs, lenticles, hooks, or rosettes. Close by this area in 

 the photograph are two very small symmetrical areas with a few 

 amorphous rough points that may be vestiges of cremastral hooks. No 

 definite rosettes, in fact, occur anywhere in the pupa. The scar of the 

 honey gland (pi. xliii., fig. 1) is very distinct, perhaps rather more 

 so than is frequent in pupa? of blues. 



The larva given me by Mr. Eayward w 7 as so close to pupation that 

 there existed within the loose skin an actual pupa, but w r ith the w T ings, 

 legs and antenna? in crumpled condition. PL xlii., fig. 2, shows the 

 cover of the prothoracic spiracle in this pupa, it differs from that in 

 the actual pupa only in being very slightly clearer for photographing. 

 In both pupa? the actual spiracle beneath this is quite circular. PI. xlv. 

 is from the immature pupa ; the structure of the spiracle is, perhaps, 

 a little clearer than in pi. xliv. from the mature pupa (Chapman). 



Changes in colour of pupa of Lycaena arion. — When newly- 

 formed the pupa is of a very clear, pale, apricot yellow, gradually 

 deepening to a dark amber-colour, with light ochreous wings. When 

 mature (four days before emergence) it was uniformly ochreous with the 

 eyes dark leaden-green ; it gradually turned darker on the head, thorax, 

 and abdomen ; the wings remained ochreous, but showed leaden-grey 

 hind margins ; then the median wing-spots appeared, and soon the 

 whole pupa began to deepen more uniformly, until it assumed a deep 

 leaden-grey all over, and remained unchanged for over thirty hours. 

 A $ emerged at 8*30 a.m. on July 3 6th (Frohawk). 



Habits. — Lycaena arion is, as its powerful form indicates, a strong 

 flier for a " blue." An example observed on August 21st, 1907, on a 

 flowery bank at the termination of the Via Mala, flew across the road, 

 up the steep bank, and aw r ay out of sight over the ridge above almost 

 before one could recognise it. For speed it was more like a Coliad 

 than a Lyca?nid. Similarly, on the slopes between Airolo and Brugnasco, 

 when it is really busy on the wing in the hot sun, one finds the greatest 

 difficulty in following it, and it will fly a considerable time without 

 resting, and at such times almost the only hope of taking it is that it will 

 frequently double back, or otherwise be tempted for a moment by the 

 nectar of some flower and thus give one time to overtake it. We 

 observed one on the morning of August 8th, 1907, flying briskly about 9 

 .a.m., up the steep slopes at Brugnasco ; it then zigzagged along the foot- 



