288 LEPIDOPTERA. 



slopes of the cliffs ; the insect sparingly distributed over an 

 extent of ground of about three-quarters of a mile in length, 

 but in one or two places it was abundant. The place in 

 which I found it in greatest plenty was a level plateau 

 about three hundred feet in length, and fifty in breadth, 

 apparently formed by a landslip at no very remote period, at 

 the side of a cliff about a hundred and thirty feet above the 

 sea. The ground was extremely rough, with masses of rock 

 lying about in all directions, and collecting there was not 

 without a spice of danger; the vegetation was of a very 

 varied character, including several species of coarse grass, 

 but apparently no Calamagrostix epif/cjos. The butterflies 

 apjieared particularly fond of the flowers of Ononis arvcnsis 

 (Eest-harrow), but I rarely sa«- them alight on any other 

 flower. I watched hundreds of specimens but was unable 

 to detect a female in the act of oviposition." 



This species has been stated to be double-brooded, but all 

 the evidence which I can collect appears to prove that there 

 is but one brood in the year, though emergence extends 

 sometimes over a very long period. Mr. E. E. Bankes says, 

 " Common in certain very restricted spots, where its food 

 plant grows plentifully, along the south-east coast of Dorset, 

 on both sides of West Lulworth, and also along the coast 

 line of the Isle of Purbeck. Its food plant is Brachypodmm 

 finnatum, and, in nature, with us, it never feeds on any 

 other grass. It is not, as generally supposed, confined to the 

 coast. I have taken it inland, in two or three spots along 

 the chalk range of the Purbeck Hills, at a distance of four or 

 five miles from the sea. It is certainly only single brooded, 

 and the best time for it is from the beginning of July to 

 the middle of August. I have taken it as early as June 16th 

 in an early year in a very warm and sheltered spot — a regular 

 sun-trap — and a specimen in good condition as late as Sep- 

 tember 8th in a colder spot, and in a late and wet year — 

 1888.'" Mr. C. W. Dale says that it has been taken as early 

 as May 31st, and as late as September 13th, but this is the 



