§ 3. "Down the stalk?" yes, and down a great way farther. For, as the leaves, if they did not thus contribute to their own support, would soon be too heavy for the spray, so if the spray, with its family of leaves, contributed nothing to the thickness of the branch, the leaf-families would soon break down their sustaining branches. And, similarly, if the branches gave nothing to the stem, the stem would soon fall under its boughs. Therefore, by a power of which I believe no sufficient account exists,[2] as each leaf adds to the thickness of the shoot, so each shoot to the branch, so each branch to the stem, and that with so perfect an order and regularity of duty, that from every leaf in all the countless crowd at the tree's summit, one slender fibre, or at least fibre's thickness of wood, descends through shoot, through spray, through branch, and through stem; and having thus added, in its due proportion, to form the strength of the tree, labors yet farther and more painfully to provide for its security; and thrusting forward into the root, loses nothing of its mighty energy, until, mining through the darkness, it has taken hold in cleft of rock or depth of earth, as extended as the sweep of its green crest in the free air. § 4. Such, at least, is the mechanical aspect of the tree. The work of its construction, considered as a branch tower, partly propped by buttresses, partly lashed by cables, is thus shared in by every leaf. But considering it as a living body to be nourished, it is probably an inaccurate analogy to speak of the leaves being taxed for the enlargement of the trunk. Strictly speaking, the trunk enlarges by sustaining them. For each leaf, however far removed from the ground, stands in need of nourishment derived from the ground, as well as of that which it finds in the air; and it simply sends its root down along the stem of the tree, until it reaches the ground and obtains the necessary mineral elements. The trunk has been therefore called by some botanists a "bundle of roots," but I think inaccurately. It is rather a messenger to the roots.[3] A root, properly so called, is a fibre, spongy or absorbent at the extremity, which secretes certain elements from the earth. The stem is by this definition no more a cluster of roots than a cluster of leaves, but a channel of intercourse between the roots and the leaves. It can gather no nourishment. It only carries nourishment, being, in fact, a group of canals for the conveyance of marketable commodities, with an electric telegraph attached to each, transmitting messages from leaf to root, and root to leaf, up and down the tree. But whatever view we take of the operative causes, the external and visible fact is simply that every leaf does send down from its stalk a slender thread of woody matter along the sides of the shoot it grows upon; and that the increase of thickness in stem, proportioned to the advance of the leaves, corresponds with an increase of thickness