212 BRITISH LEPIDOPTERA. 



margin ; apex strongly purple : cilia unicolorous dark grey. Posterior 

 wings pale grey, rather darker towards apex ; cilia dark grey. 



Egg.— The egg is placed by the side of the midrib beneath a leaf. It 

 forms a perfect oval in outline, and is almost of the same shape and pro- 

 portion (length, breadth, thickness) as a hen's egg, the shell transparent, 

 slightly iridescent, and packed inside with a ring of black f rass. Under 

 a two-thirds lens the surface of the shell appears to be quite smooth. 



Mine. — The mine in Rosa arvensis commences as a faint thread, so 

 nearly of the same colour as the leaf as to be almost indistinguishable. 

 In the first part of the mine the frass forms a pretty continuous thread 

 along the centre of the mine, but after the mine is about half an inch 

 long the larva moults, and the minute frass pellets are distributed 

 over the whole width of the mine, which is in this portion particularly 

 inconspicuous ; this portion of the mine is possibly an inch in length. 

 The larva evidently moults again, and immediately the gallery widens 

 considerably ; the frass is arranged in regular lines following the 

 direction of the mine, and a broad whitish-green edge is left on either 

 side of the frass-track. This portion of the track, including the 

 windings, is at least two to two and a-half inches in length. The 

 full-fed larva leaves the mine by the upper side of the leaf. [Described 

 from mines sent by Mr. W. H. B. Fletcher, June 21st, 1898.] 



Comparison of mines of N. fletcheri with N. anomalella. — 

 We are indebted to Mr. Fletcher for mines of the " red-headed " form, in 

 leaves of Rosa sempervirens var. , for mines of the same form in leaves of 

 Rosa canina, and for mines of the " black-headed " form in leaves of 

 Rosa arvensis. We may say at once that the mines in R. sempervirens 

 and R. canina are identical in every respect, long sweeping mines with 

 broad curves, and with the frass exceedingly densely packed in the 

 centre of the gallery throughout the whole length of the mine, the frass 

 brownish at first, but afterwards black. The mine in the leaf of 

 Rosa arvensis differs greatly, not only in the appearance of the pale 

 (almost whitish) track itself, but in the inconspicuousness of the early 

 part, of the mine, in its sharply defined boundary against the 

 parenchyma in the later portion ; its more sudden widening in the 

 middle part of its course, and in the enormous difference in the 

 disposition of the frass. Instead of the dense black line which 

 characterises the mines in R. sempervirens and R. canina, the mine in 

 R. arvensis shows no trace of frass to the naked eye, only a rather 

 darker green shade, where one knows the frass should be. Under a two- 

 thirds lens the minute black frass pellets form a distinct central broken 

 line ; in the second part they are brownish and scattered over the whole 

 width of the mine ; in the third part the mine suddenly widens, and the 

 pellets, although still scattered and separate, form a distinct central 

 path, a character that is now maintained to the end, for, instead of 

 the thick central black line of the mine of the " red-headed " form, the 

 individual pellets are here scattered separately over the central part of 

 the gallery, it being a rare occurrence for even two frass pellets to 

 touch each other, and their regularity in lines following the direction 

 of the gallery is sometimes remarkable. 



Cocoon. — The cocoons examined average 2-2 mm. in length and 

 1-9 mm. in width. They vary much in shape, according to the 

 position in which they have been spun up. Some, constructed in the 

 narrow hollow of a contracted stipule at the base of the petiole, form 



