40 f"^"^^' 



attained twice that length, and about the end of the month they, hke mine, all went 

 to earth : the moths appeared September 2nd — 11th. "With me, however, the eggs 

 did not hatch before the 11th of April, 1870. 



The shape of the egg is circular, but flattened, and greatly depressed in the 

 centre, slightly ribbed and reticulated, having withal a shrivelled, empty appearance, 

 as though its contents had been squeezed or dried out of it : the colour at first is 

 pale yellow, soon after turning to a slightly pinkish leaden hue, and the shell is very 

 glistening, in fact, the flat mass of eggs seemed smeared over with a coating of trans- 

 parent varnish or gum, which I fancy would in nature attach them to some cover on 

 the upper as well as the lower surface ; I think they would be laid in cracks, or 

 under loose bits of bark ; as the time of hatching approached the colour seemed but 

 Tery little heightened, and when the young larvae had emerged from them, the 

 empty shells looked brilliantly crystalline. 



The newly-hatched larvae were pinkish-grey in colour, with the dorsal vessel ap- 

 pearing as a dark grey, leaden stripe, the pale brown head large in proportion ; 

 within a fortnight they became of a dull flesh colour, after the second moult they 

 were pale greyish-green, and soon turned darker, then showing the usual lines pale 

 and distinct, their length being now three-eighths of an inch, and their figure pro- 

 portionately stout ; by the 6th of May they were five-eighths of an inch long, of a 

 yellowish-green colour, the lines rather paler green, the tubercular dots exceedingly, 

 small and blackish ; by the loth, they were three-quarters of an inch long, of a pale 

 yellow-green, the lines all present but unobtrusive ; up to this time they had fed 

 almost entirely on the common garden montlily rose, which they preferred from the 

 first to all other food given them in great variety from time to time ; but at this 

 period they were tried with bramble, and, after tasting it, they no longer cared for 

 rose, and thenceforward fed up chiefly on brambles of different species. 



The full-grown larva was one inch and a half in length, moderately stout, cylin- 

 drical, and unifoiTn in bidk, with tolerably well defined segmental divisions ; the 

 coloiu" on the back and sides, as far as the spiracles, green, somewhat inclining to 

 olive, and freckled with a little darker green, and, on this fi-eckled surface, the dorsal 

 and sub-dorsal lines could be distinctly traced, a little paler than the ground, but 

 edged with interrupted, freckly, almost blackish, lines, which, in some instances, es- 

 pecially with the dorsal line, seemed almost to obscure the pale line they enclosed : 

 the tubercular dots were also paler than the ground, and very finely ringed with 

 darker green ; the boundary of this green colouring along the side was completed by 

 a black line, interrupted only where the spiracles (white, outlined with black) were 

 placed upon it ; immediately beneath the spiracles the contrast of whitish-yellow 

 deepened a little by degrees into a pale yellowish- gi-een, which was the colour of the 

 belly and legs, these last were tipped with brownish : the head was brownish-green, 

 freckled with darker ; the second segment was not very different in texture from the 

 rest of the body, it was in most examples edged in front with very dark brown, and 

 the pale lines that appeared on it were without any dark edging : the whole brood 

 presented scarcely the least variety, either in colom-ing or detail, but were as constant 

 as possible in their uniformity. 



Some of the larvae, which were kept in a flower-pot with sand for soil, formed 

 very neat, compact cocoons of silk, covei'cd thinly but uniformly with the sand, 

 rather more than five-eighths of an inch long, and about five-sixteenths broad ; pro- 

 bably, in a coarser soil, they would have been less regular in outline. 



