in roots, proportioned to the advance of their outer fibres. How far interchange of elements takes place between root and leaf, it is not our work here to examine; the general and broad idea is this, that the whole tree is fed partly by the earth, partly by the air;--strengthened and sustained by the one, agitated and educated by the other;--all of it which is best, in substance, life, and beauty, being drawn more from the dew of heaven than the fatness of the earth. The results of this nourishment of the bough by the leaf in external aspect, are the object of our immediate inquiry. § 5. Hitherto we have considered the shoot as an ascending body, throwing off buds at intervals. This it is indeed; but the part of it which ascends is not seen externally. Look back to Plate 51. You will observe that each shoot is furrowed, and that the ridges between the furrows rise in slightly spiral lines, terminating in the armlets under the buds which bore last year's leaves. These ridges, which rib the shoot so distinctly, are not on the ascending part of it. They are the contributions of each successive leaf thrown out as it ascended. Every leaf sent down a slender cord, covering and clinging to the shoot beneath, and increasing its thickness. Each, according to his size and strength, wove his little strand of cable, as a spider his thread; and cast it down the side of the springing tower by a marvellous magic--irresistible! The fall of a granite pyramid from an Alp may perhaps be stayed; the descending force of that silver thread shall not be stayed. It will split the rocks themselves at its roots, if need be, rather than fail in its work. So many leaves, so many silver cords. Count--for by just the thickness of one cord, beneath each leaf, let fall in fivefold order round and round, the shoot increases in thickness to its root:--a spire built downwards from the heaven. [Illustration: FIG. 36.] And now we see why the leaves dislike being above each other. Each seeks a vacant place, where he may freely let fall the cord. The turning aside of the cable to avoid the buds beneath, is one of the main causes of spiral curvature, as the shoot increases. It required all the care I could give to the drawing, and all Mr. Armytage's skill in engraving Plate 51, to express, though drawing them nearly of their full size, the principal courses of curvature in even this least graceful of trees.