150 



lar sheet of white silk, which, by its contraction, draws from the 

 circumference towards the centre, and causes a circular bulge 

 or projection on the under side of the leaf, the diameter of 

 which is a little greater than the length of the larva. Turning 

 again upon its back, it spins, in like manner, a white and thin, 

 thouo-h tough, transparent sheet above it, which, by its contrac- 

 tion, causes the nidus to project still more on the under side of 

 the leaf, and the cocoon is complete. 



In the larvae of this group the upper surface of the integu- 

 ment is shagreened, and, as the feet are of little use in crawling, 

 this roughness is no doubt useful in aiding progress, by contact 

 with the separated cuticle ; in some of the species the hind 

 margin of the penultimate segment beneath is serrated, which 

 affords similar aid. The larvse of this group never really crawl, 

 they only wriggle. 



In L. orviatella the surface is not roughened, nor the penulti- 

 mate segment serrated, and the legs are no better developed 

 than in the flat larvae, to which, in all of its stages but the last, 

 it is closely allied. It differs from those larvee, however, in 

 having the segments more distinctly mammillated at their sides, 

 and its outline in transverse section is more nearly elliptical, as 

 shown by fig. 8b. The larva is white, and the maculte are only 

 distinct in the second, third and fourth stages. At the fifth 

 stage it is tinged with green, which becomes deeper in each 

 succeedino; stage. Until the seventh moult the mouth parts are 

 as in fig. 2. At that moult, however, the tropin, like those of 

 the larvae of the flat group, become as in fig. 4, and the legs be- 

 come larger and more perfect than before, or than ever in the 

 cylindrical group. The form also is so altered that, immediately 

 after the moult, it is almost semicircular in transverse section 

 (fig. 8c). Like the larvjfi of the flat group, L. ornatella does not 

 eat in its eighth stage, but, unlike them, it is far from inactive. It 

 is the only larva of the genus (except L. helianthemella, one of the 

 cylindrical group in Europe) which is capable of crawling when 

 removed from the mine, and in a few hours after its seventh 

 moult it voluntarily cuts a semicircular opening in the cuticle, 

 by which it leaves the mine, and crawls away to spin its cocoon 

 and pupate. Its cocoon is not flat, but is so tightly drawn 



