These subterranean workers do not allow their proceedings to be easily observed. Réaumur was fortunate enough to be able to notice their skill in the construction of their shells or cocoons. The _Cucullia verbasci_ (Fig. 120) makes itself a thick and very compact cocoon of the form of an egg (Fig. 121). Réaumur took one of these out of the ground before it is fortified. He tore it partially open, and placed it in a glass vase containing sand, but the poor insect was not long in repairing the disorder caused by the rough hand of our naturalist. It only took four hours to restore its cocoon to its former state. "It began," says Réaumur, "by coming almost entirely out, and left only its hinder part within. It moved its head forwards as far as was necessary to enable it to seize a particle of earth. As soon as it had got its load, it re-entered the interior of the cocoon. It deposited the grain of earth, and came out again immediately, as it did at first, to pick up another grain, which it carried likewise into the interior of the cocoon. This operation it continued for more than an hour.... The provision of materials being got together, the caterpillar now devoted his whole attention to working them up. It began by spinning over one part of the edges of the opening. After having put over this a small band of very loose web, the caterpillar's head left the opening, the insect went right back again into its cocoon, and the head returned to the opening loaded with a little grain of earth, which it entangled in the silky threads. It then entangled in them two or three, or a greater number of grains, according to the quantity of threads it had spun. It bound them into these with other threads, after which it drew threads over the edges of another part. By thus going round the whole rim of the opening, and by carrying and fixing the grains of earth in the threads which were the last stretched over the opening, it rendered its diameter smaller and smaller." It was by working with its head that our mason gave to the new wall of its cocoon the necessary curvature. It was interesting to know how, as it could no longer put out its head, it could stop up the orifice.