22 BRITISH LEPIDOPTERA. 



most commonly found on fine grass, eating from the top downwards, 

 but when they have reached about two-thirds their full size they 

 leave the grass, and feed almost exclusively on Ornithopus perpusilhis^ 

 and one is always more fortunate in breeding imagines from bird's 

 foot trefoil than any other plant ; they can, he asserts, be starved 

 on the grasses, but thrive on clover of any sort. Mitford notes 

 that on Romney Marsh the larvae feed in May on the tufts of a 

 very wiry grass, growing on the shingle above high-water mark. 

 Gregson says the young larvae are very small, and very yellow in 

 colour in March, when they are abundant on the Crosby sandhills. 

 The larvae are best obtained after 7 p.m. on the star-grass, but on 

 hot close days they may be found stretched out on the bare sand 

 in the afternoon. Walker says that the larvae should be collected 

 just before dusk, when they are very conspicuous, but they soon take 

 alarm and roll themselves into a ring, but do not, as a rule, drop from 

 the plant on which they rest. Harker says that to breed the larvae suc- 

 cessfully plenty of fresh food and light are required. Turner notes that 

 "at Bolt Head and Starehole Bottom the larvae are common in April, and 

 appear to be very general feeders, nor was there any clover of any 

 kind growing in the locality in which they occurred. Several were 

 found eating different species of grass, one was seen to nibble bracken, 

 others were found, though not actually feeding, on heather, bramble 

 and violet ; one was found eating gorse-blossom, which had apparently 

 been its food for some time, as it ejected several pieces of yellowish- 

 brown frass. The larvae are very shy, and on one's approach they 

 curl up, often writhing to and fro, and remain thus for a considerable 

 time ; several were found sitting on rocks, but by far the greater 

 number were stretched out on dead bracken. After the severe winter 

 of 1894-5, the larvae did not appear at Salcombe until April 26th 

 (compared with April 14th in 1894), and the specimens found were 

 very small indeed ; these, and others found later, were at rest on dead 

 bracken or bare stumps of blackthorn, and it was not until May 1st 

 that they were actually observed feeding." Lane observed half-fed 

 larvae basking in the sun about 4 p.m. on Whitemoor, Lyndhurst, in May, 

 1898. Newman says that the larva (after hybernation) is first observed 

 in spring when it is three-tenths of an inch in length ; it then rests 

 extended on a blade or stalk of grass in a straight position, and, when 

 it feeds, embraces the food with its feet, and devours from the top 

 downwards, but if disturbed it immediately falls to the ground and 

 rolls itself into a compact, but not very perfect, ring, the two 

 extremities not meeting with precision, but passing each other, and 

 thus giving a one-sided appearance to the ring, and the same characters 

 are observable during the entire period of its growth, whilst after the 

 end of April it feeds up very rapidly, and is fullfed at the end of May 

 or beginning of June. Fowler says that larva? occur on the "cridrum " 

 ground at Ringwood ; he finds them in June, when they are usually 

 about an inch long. The larvae positively swarm in certain seasons in 

 some districts. Sepp collected 85 larvae on Calluna iwlgaris in early 

 I line, and bred imagines from the beginning of August onwards. 

 Prescotl observes that on June 8th, [857, he obtained 86 larvae at Black- 

 pool, and saw 400 taken the same day. Bishop notes that on June 

 1 6th, 1857, he had collected a large quantity in the Plymouth district, 

 and from the same locality Lethbridge obtained a great number on 



