1 
198 [February, 
top of its cocoon covering, and inspects tte strength, durability, and 
finish of its work, and may be seen spinning additional threads to such 
parts as require strengthening. "Whilst thus occupied, we may observe 
that a vacant space is left in the centre of the cocoon covering, which 
space had been occupied by the body of the larva whilst at work. 
I apprehend this covering is only spun with the object of further 
indenting the leaf, so as to make it more suitable for the reception of 
the oval cocoon, as I have observed that when the larva makes its 
cocoon in the angle formed by the top and side of a box-lid, it often 
forms it without any covering whatever. 
The larva now retires backwards through this identical opening on 
the cocoon canopy, which it does not trouble itself to close, and at once 
turning on its back, begins the construction of its cocoon proper directly 
beneath, but not attached to, the upper canopy; in this position 
it works away, spinning and carrying silken threads from side to 
side till it has formed the cocoon ; it then turns over on to its feet and 
spins a white silken band all round the inner margin of its partially 
completed cocoon. 
"When I first perceived the silken band I conjectured thab its 
object was to keep the larva snug and warm during the winter months, 
but, as I find that the summer broods of the larvae construct similar 
silken bands, this can harly be the object. 
In the subterranean cocoons I do not perceive any silken band, or 
anything analogous. When the band is complete the larva sets to 
work to carpet the floor of the cocoon with white silk, as a soft couch 
on which to recline ; it then again turns on its back, and continues 
actively employed in spinning additions to its cocoon, until this 
at last becomes so opaque that the movements within are no longer 
perceptible. 
The time occupied by the larva in the construction of its cocoon 
is between nine and ten hours — hours of almost incessant labour. If 
turned out of its cocoon directly it is finished, the larva sets to work 
and constructs another cocoon complete in every resjject like the first, 
but if it be again ejected after fabricating its second cocoon, its supply 
of silk being quite exhausted, it lies perfectly helpless, and a larva so 
treated very often dies, not having even strength to assume the pupa 
state. 
If the larva be left undisturbed in its cocoon, its body soon becomes 
quite white, and a death-like trance steals over it (it may then be 
handled or turned about without its showing any power of movement), 
