§ 2. I am not sure with what absolute stringency this law is observed, or what slight lengthening of substance may be traceable by close measurement among inferior branches. For practical purposes, we may assume that the law is final, and that if we represent the state of a plant, or extremity of branch, in any given year under the simplest possible type, Fig. 44, _a_, of two shoots, with terminal buds, springing from one stem, its growth next year may be expressed by the type, Fig. 44, _b_, in which, the original stems not changing or increasing, the terminal buds have built up each another story of plant, or repetition of the original form; and, in order to support this new edifice, have sent down roots all the way to the ground, so as to enclose and thicken the inferior stem. [Illustration: FIG. 44.] But if this is so, how does the original stem, which never lengthens, ever become the tall trunk of a tree? The arrangement just stated provides very satisfactorily for making it stout, but not for making it tall. If the ramification proceeds in this way, the tree must assuredly become a round compact ball of short sticks, attached to the ground by a very stout, almost invisible, stem, like a puff-ball. For if we take the form above, on a small scale, merely to see what comes of it, and carry its branching three steps farther, we get the successive conditions in Fig. 45, of which the last comes already round to the ground. [Illustration: FIG. 45.] "But those forms really look something like trees!" Yes, if they were on a large scale. But each of the little shoots is only six or seven inches long; the whole cluster would but be three or four feet over, and touches the ground already at its extremity. It would enlarge if it went on growing, but never rise from the ground.