Among the Gramineae, the distinction between the towering, woody bamboos, and the herbaceous grasses and cereals, is so striking that an effort of the mind is needed before we can think of them together, as members of a single and seemingly natural family. In Chapter iv we treated the tree habit of the bamboo merely as an observed fact; in this chapter we shall make an attempt to understand something of its significance. It cannot, however, be interpreted at all, if it is considered as an isolated phenomenon; in order to get any glimmering comprehension of it, we shall be obliged now to turn to the problem of the dendroid and herbaceous type in the flowering plants as a whole. From the earliest times at which a desire to pigeon-hole the facts of the world led man to attempt some kind of classification of the plants which he saw around him, the difference between the tree and the herb has seemed to him an important one. Theophrastus (born circa 370 .c.), the earliest scientific botanist of whose opinions we have a record, divided plants into the tree, the shrub, the undershrub and the herb. His classification thus makes a fundamental distinction between the arboreal and herbaceous habits, though his shrubs and undershrubs to a certain extent form a transition between the two.