§ 5. Let us return to the type in Fig. 2, of the fully accomplished summer's work: the rod with its bare buds. Plate 51, opposite, represents, of about half its real size, an outer spray of oak in winter. It is not growing strongly, and is as simple as possible in ramification. You may easily see, in each branch, the continuous piece of shoot produced last year. The wrinkles which make these shoots look like old branches are caused by drying, as the stalk of a bunch of raisins is furrowed (the oak-shoot fresh gathered is round as a grape-stalk). I draw them thus, because the furrows are important clues to structure. Fig. 4 is the top of one of these oak sprays magnified for reference. The little brackets, _x_, _y_, &c., which project beneath each bud and sustain it, are the remains of the leaf-stalks. Those stalks were jointed at that place, and the leaves fell without leaving a scar, only a crescent-shaped, somewhat blank-looking flat space, which you may study at your ease on a horse-chestnut stem, where these spaces are very large. § 6. Now if you cut your oak spray neatly through, just above a bud, as at A, Fig. 4, and look at it with a not very powerful magnifier, you will find it present the pretty section, Fig. 5. [Illustration: FIG. 5.] That is the proper or normal section of an oak spray. Never quite regular. Sure to have one of the projections a little larger than the rest, and to have its bark (the black line) not quite regularly put round it, but exquisitely finished, down to a little white star in the very centre, which I have not drawn, because it would look in the woodcut black, not white; and be too conspicuous. [Illustration: FIG. 6.] The oak spray, however, will not keep this form unchanged for an instant. Cut it through a little way above your first section, and you will find the largest projection is increasing till, just where it opens[1] at last into the leaf-stalk, its section is Fig. 6. If, therefore, you choose to consider every interval between bud and bud as one story of your tower or pinnacle, you find that there is literally not a hair's-breadth of the work in which the _plan_ of the tower does not change. You may see in Plate 51 that every shoot is suffused by a subtle (in nature an _infinitely_ subtle) change of contour between bud and bud. [Illustration: FIG. 7.] [Illustration: FIG. 8.]