relate,
we must admire her precautions in exposing certain individuals to a
mortal danger.

The detail on which I have just entered clearly indicates the final
cause of the opening left by the royal worms in their coccoons; but it
does not shew whether it is in consequence of a particular instinct that
they leave this opening, or whether the wideness of their cells prevents
them from stretching the thread up to the top. This question interested
me very much; the only method of deciding it was to observe the worms
while spinning, which cannot be done in their opaque cells. It then
occurred to me to dislodge them from their own habitations, and
introduce them into glass tubes, blown in exact imitation of the
different kind of cells. The most difficult part of the operation
consisted in extracting worms and introducing them here; but my
assistant accomplished it with much address. He opened several sealed
royal cells, where we knew the larvae were about to begin their coccoons,
and, taking them gently out, introduced one into each of my glass cells
without the smallest injury.

They soon prepared to work; and commenced by stretching the anterior
part of the body in a straight line, while the other was bent in a
curve. This formed a curve of which the longitudinal sides of the cells
were tangents, and afforded two points of support. The head was next
conducted to the different parts of the cell which it could reach, and
it carpeted the surface with a thick bed of silk. We remarked that the
threads were not carried from one side to another, and that this would
have been impracticable, for the worms being obliged to support
themselves, and to keep the posterior rings curved, the free and
moveable part of the body was not long enough for the mouth to reach the
sides diametrically opposite, and fix the threads to them. You will
remember, Sir, that the royal cells are of a pyramidal form, with a wide
base, and a long contracted top. These cells hang perpendicularly in the
hive, the point downwards, from which position the royal worm can be
supported in the cell, only when the curvature of the posterior part
forms two points of support; and that it cannot obtain this support
without resting on the lower part, or towards the extremity. Therefore
if it attempted to stretch out and spin towards the wide end of the
cell, it could not reach both sides from being too distant. One part
would be touched by its extremity, the other by