Let us watch it at work when it is unable to use its favorite joist. There is no point in giving it clumsy building stones; that would only bring us back to the uncouth sheaths. Its propensity to make use of soaked seeds, those of the iris, for instance, suggests that I might try grains. I select rice, which, because of its hardness, will be tantamount to wood and, because of its clean whiteness and its oval shape, will lend itself to artistic masonry. Obviously, my denuded caddis worms cannot start their work with bricks of this kind. Where would they fix their first layer? They must have a foundation, quick and easy to build. This is once more supplied by a temporary cylinder of watercress roots. On this support follow the grains of rice, which, grouped one atop the other, straight or slanting, end by giving a magnificent turret of ivory. Next to the sheaths made of tiny snail shells, this is the prettiest thing with which the caddis worm's industry has furnished me. A fine sense of order has returned, because the materials, regular and of identical character, have cooperated with the correct method of the worker. The two demonstrations are enough. Sticks and grains of rice make it plain that the caddis worm is not the bungler that one would expect from the monstrous buildings in the pond. Those Cyclopean piles, those mad conglomerations, are the inevitable results of chance finds, which are used for the best because there is no choice. The water carpenter has an art of its own, has method and rules of symmetry. When well served by fortune, it is quite able to turn out good work; when ill-served, it acts like others: the work which it turns out is bad. Poverty makes for ugliness. There is another matter wherein the caddis worm deserves our attention. With a perseverance which repeated trials do not tire, it makes itself a new tube when I strip it. This is opposed to the habits of the generality of insects, which do not recommence the thing once done, but simply continue it according to the usual rules, taking no account of the ruined or vanished portions. The caddis worm is a striking exception: it starts again. Whence does it derive this capacity? I begin by learning that, given a sudden alarm, it readily leaves its scabbard. When I go fishing for caddis worms, I put them in tin boxes, containing no other moisture than that wherewith my catches are soaked. I heap them up loosely, to avoid any grievous tumult and to fill the space at my disposal as best I may. I take no further precaution. This is enough to keep the caddis worms in good condition during the two or three hours which I devote to fishing and to walking home.