2. But how do they acquire this new character, and why is it the work of so much time? When the young grass throws up its leaves into the air, from which it derives so much of its nourishment, it throws down its roots into the soil in quest of food of another kind. The leaves may be mown or cropped by animals, and carried off the field, but the roots remain in the soil, and, as they die, gradually fill its upper part with vegetable matter. It is not known what average proportion the roots of the natural grasses bear to the leaves; no doubt it varies much, both with the kind of grass and with the kind of soil. When wheat is cut down, the quantity of straw left in the field, in the form of stubble and roots, is sometimes greater than the quantity carried off in the sheaf. Upon a grass field two or three tons of hay may be reaped from an acre; and if we suppose only a tenth part of this quantity to die every year in the form of roots or parts of roots, or of excretions from roots, we can easily understand how the vegetable matter in the soil thus gradually accumulating, should at length become very considerable in quantity. In arable land this accumulation is prevented by the constant turning up of the soil, by which the vegetable fibres being exposed to the free access of air and moisture, are made to undergo a more rapid decomposition. But the roots and leaves of the grasses contain inorganic earthy and saline matter also. Dry hay leaves from an eighth to a tenth part of its weight of ash when burned. Along with the dead vegetable matter of the soil, this inorganic matter accumulates also on the surface, in the form of an exceedingly fine earthy powder; hence _one_ cause of the universal fineness of the surface mould of old grass fields. And the earthy portion of this inorganic matter consists chiefly of silica and lime, with scarcely a trace of alumina, so that, even on the stiffest clays, a surface soil may be ultimately formed, in which the quantity of alumina will be comparatively small. But there are still other agencies at work by which the surface of stiff soils is made to undergo a change. As the roots penetrate into the clay, they more or less open up a way into it for the rains. Now the rains in nearly all lands, when they have a passage downwards, have a tendency to carry down the clay along with them. They do so, it has been observed, on sandy and peaty soils, and more quickly when these soils are laid down to grass. Hence the mechanical action of the rains,—slowly in many localities, yet surely,—has a tendency to lighten the soil, by removing a portion of its clay. They constitute one of those natural agencies by which, as elsewhere explained, important differences are ultimately established, almost everywhere, between the surface crop-bearing soil and the subsoil on which it rests.