§ 9. LAW II. THE LAW OF SUCCESSION.--From what we saw of the position of buds, it follows that in every tree the leaves at the end of the spray, taking the direction given them by the uppermost cycle or spiral of the buds, will fall naturally into a starry group, expressive of the order of their growth. In an oak we shall have a cluster of five leaves, in a horse-chestnut of four, in a rhododendron of six, and so on. But observe, if we draw the oak-leaves all equal, as at _a_, Fig. 25, or the chestnut's (_b_), or the rhododendron's (_c_), you instantly will feel, or ought to feel, that something is wrong; that those are not foliage forms--not even normally or typically so--but dead forms, like crystals of snow. Considering this, and looking back to last chapter, you will see that the buds which throw out these leaves do not grow side by side, but one above another. In the oak and rhododendron, all five and all six buds are at different heights; in the chestnut, one couple is above the other couple. [Illustration: FIG. 25.] § 10. Now so surely as one bud is above another, it must be stronger or weaker than that other. The shoot may either be increasing in strength as it advances, or declining; in either case, the buds must vary in power, and the leaves in size. At the top of the shoot, the last or uppermost leaves are mostly the smallest; of course always so in spring as they develope. [Illustration: FIG. 26.] Let us then apply these conditions to our formal figure above, and suppose each leaf to be weaker in its order of succession. The oak becomes as _a_, Fig. 26, the chestnut shoot as _b_, the rhododendron, _c_. These, I should think, it can hardly be necessary to tell the reader, are true normal forms;--respecting which one or two points must be noticed in detail. § 11. The magnitude of the leaves in the oak star diminishes, of course, in alternate order. The largest leaf is the lowest, 1 in Figure 8, p. 14. While the largest leaf forms the bottom, next it, opposite each other, come the third and fourth, in order and magnitude, and the fifth and second form the top. An oak star is, therefore, always an oblique star; but in the chestnut and other quatrefoil trees, though the uppermost couple of leaves must always be smaller than the lowermost couple, there appears no geometrical reason why the opposite leaves of each couple should vary in size. Nevertheless, they always do, so that the quatrefoil becomes oblique as well as the cinqfoil, as you see it is in Fig. 26. [Illustration: FIG. 27.] [Illustration: FIG. 28.] The normal of four-foils is therefore as in Fig. 27, A (maple): with magnitudes, in order numbered; but it often happens that an opposite pair agree to become largest and smallest; thus giving the pretty symmetry, Fig. 27, B (spotted aucuba). Of course the quatrefoil in reality is always less formal, one pair of leaves more or less hiding or preceding the other. Fig. 28 is the outline of a young one in the maple.