you have to move your position, an effort which has been avoided
until now. With this object, the caddis worm cuts its moorings, that
is to say, the rootlets which keep the cylinder fixed, or else the
half-severed leaf of pond weed on which the cone-shaped bag has come
into being.

The worm is now free. The smallness of the artificial pond, the tumbler,
soon brings it into touch with what it is seeking. This is a little
faggot of dry twigs, which I have selected of equal length and of slight
thickness. Displaying greater care than it did when treating the slender
roots, the carpenter measures out the requisite length on the joist. The
distance to which it has to extend its body in order to reach the point
where the break will be made tells it pretty accurately what length of
stick it wants.

The piece is patiently sawn off with the mandibles; it is next taken in
the fore legs and held crosswise below the neck. The backward movement
which brings the caddis worm home also brings the bit of twig to the
edge of the tube. Thereupon, the methods employed in working with the
scraps of root are renewed in precisely the same manner. The sticks are
scaffolded to the regulation height, all alike in length, amply soldered
in the middle and free at either end.

With the picked materials provided, the carpenter has turned out a work
of some elegance. The joists are all arranged crosswise, because
this way is the handiest for carrying the sticks and putting them in
position; they are fixed by the middle, because the two arms that hold
the stick while the spinneret does its work require an equal grasp
on either side; each soldering covers a length which is seen to be
practically invariable, because it is equal to the width described by
the head in bending first to this side and then to that when the silk
is emitted; the whole assumes a polygonal shape, not far removed from
a rectilinear pentagon, because, between laying one piece and the next,
the caddis worm turns by the width of an arc corresponding with the
length of a soldering. The regularity of the method produces the
regularity of the work; but it is essential, of course, that the
materials should lend themselves to precise coordination.

In its natural pond, the caddis worm does not often have at its disposal
the picked joists which I give it in the tumbler. It comes across
something of everything; and that something of everything it employs as
it finds it. Bits of wood, l