STATES OF INSECTS. (Larva.) 215 
rounded itself with a slight and loosely spun netting. 
In the centre of this, when contracted into a space suf- 
ficiently small, it lays the foundation of the znterior co- 
coon. Fixing itself by its prolegs to some of the sur- 
rounding threads, it bends its body, and by successive 
motions of its head from side to side spins a layer of silk 
on the side opposite to it: when this is of the requisite 
thickness, the larva shifts its position, and repeats the 
same process in another quarter, covering each layer in 
turn with a new one until the interior cavity is reduced 
to the size desired. Thus, the silken thread which forms 
this new cocoon is not, as might have been supposed, 
wound circularly as we wind the thread of a ball of cot- 
ton; but backwards and forwards in a series of zigzags, 
so as to compose a number of distinct layers. Malpighi 
could distinguish six of these layers*, and Reaumur 
suspects there is often a greater number’. The former 
found the length of the thread of silk composing them 
when wound off, without including the exterior case, to 
be not less than 930 feet*; but others have computed it 
at more than a thousand‘: consequently the threads of 
five cocoons united would be a mile in length. Esti- 
mating by the weight,—the thread of a pound of co- 
coons, each of which weighs about two grains and a 
half, would extend more than 600 miles*; and such is 
its tenuity, that the threads of five or six cocoons re- 
quire to be joined to form one of the thickness requisite 
_ * De Bombyc. 24. b 7, 498. 
© De Bomoyc. 43. * N. Dict. d Hist. Nat, vi. 294. 
© Lesser, L. ii. 150, note 22. Boyle says an English lady found 
that the silk of a single cocoon would extend 300 English leagues or 
$00 miles. But this must be a mistake. 
