[Illustration: A Caterpillar's tent upon a leaf of the elm.--_a_, _a_, The part of the leaf from which the tent has been cut out; _b_, the tent itself.] For this purpose, it began to eat through one of the two outer membranes which compose the leaf and enclose the pulp (_parenchyma_), some of which, also, it devoured, and then thrust the hinder part of its body into the perforation. The cavity, however, which it had formed, being yet too small for its reception, it immediately resumed the task of making it larger. By continuing to gnaw into the pulp _between_ the membranes of the leaf (for it took the greatest care not to puncture or injure the membranes themselves), it soon succeeded in mining out a gallery rather larger than was sufficient to contain its body. We perceived that it did not throw out as rubbish the pulp it dug into, but devoured it as food--a circumstance not the least remarkable in its proceedings. As the two membranes of leaf thus deprived of the enclosed pulp appeared white and transparent, every movement of the insect within could be distinctly seen; and it was not a little interesting to watch its ingenious operations while it was making its tent from the membranes prepared as we have just described. These, as Réaumur has remarked, are in fact to the insect like a piece of cloth in the hands of a tailor; and no tailor could cut out a shape with more neatness and dexterity than this little workman does. As the caterpillar is furnished in its mandibles with an excellent pair of scissors, this may not appear to be a difficult task; yet, when we examine the matter more minutely, we find that the peculiar shape of the two extremities requires different curvatures, and this, of course, renders the operation no less complex, as Réaumur subjoins, than the shaping of the pieces of cloth for a coat.[BW] The insect, in fact, shapes the membranes slightly convex on one side and concave on the other, and at one end twice as large as the other. In the instance which we observed, beginning at the larger end, it bent them gently on each side by pressing them with its body thrown into a curve. We have not said it _cuts_, but _shapes_ its materials; for it must be obvious that if the insect had cut both the membranes at this stage of its operations, the pieces would have fallen and carried it along with them.