In reading the above observations, the practical reader can hardly fail to have been struck with the remarkable similarity in physical properties between stiff clay and peaty soils. Both retain much of the water that falls in rain, and both part with it slowly by evaporation. Both contract much in drying; and both absorb moisture readily from the air in the absence of the sun. In this similarity of properties, we see not only why the first steps in improving both kinds of soil must be very nearly the same; but why, also, a mixture either of clay or of vegetable matter will equally impart to a sandy soil many of those aids to, or elements of, fertility—of which they are alike possessed. SECTION IV.—OF THE CHEMICAL CONSTITUTION OF SOILS. Soils perform at least three functions, in reference to vegetation. They serve as a basis in which plants may fix their roots and sustain themselves in their erect position,—they supply inorganic food to vegetables at every period of their growth,—and they are the medium in which many chemical changes take place, that are essential to a right preparation of the various kinds of food which the soil is destined to yield to the growing plant. We have spoken of soils as consisting chiefly of sand, lime, and clay, with certain saline and organic substances in smaller and variable proportions. But the study of the ash of plants (see chap. iv.) shews us, that a fertile soil must of necessity contain an appreciable quantity of at least eleven different substances, which in most cases exist in greater or less relative abundance in the ash both of wild and of cultivated plants. Two well known geological facts lead to precisely the same conclusion. We have seen that the soils formed from the unstratified rocks,—the granites and the traps,—while they each contain certain earthy substances in proportions peculiar to themselves, yet contain also in general a _trace_ of most of those different kinds of matter which are found in the ash of plants. And when to this fact is added the other, that the stratified rocks appear to be only the long accumulating fragments and ruins of more ancient unstratified masses—which, under various agencies, have gradually crumbled to dust, been strewed over the surface in alternate layers, and afterwards again consolidated,—the reader will readily grant, that in all rocks, and consequently in all soils, _traces_ of every one of these substances may generally be presumed to exist.