319. If, following the light derived from these indications, we go back to the time when the river ran above the highest of those levels at which it has left any traces of its operations, we shall see it composed of a series of lakes and cataracts, from which, by the filling up of the one, and the wearing down of the other, the waters have at length worked out to themselves a quiet and uninterrupted passage to the ocean. We may, indeed, on good evidence, go back still farther than the succession of such meadows or terraces, as are above mentioned, will carry us, and may consider the whole valley, or trough of the river, as produced by its own operations. The original inequalities of the surface, and the disposition of the strata, must no doubt have determined the water courses at first; but this does not hinder us from considering the rivers as having modified and changed those inequalities, and as the proximate causes of the shape and configuration which the surface has now assumed. 320. From this gradual change of lakes into rivers, it follows, that a lake is but a temporary and accidental condition of a river, which is every day approaching to its termination; and the truth of this is attested, not only by the lakes that have existed, but also by those that continue to exist. Where any considerable stream enters a lake, a flat meadow is usually observed increasing from year to year. The soil of this meadow is disposed in horizontal strata: the meadow is terminated by a marsh; which marsh is acquiring solidity, and is soon to be converted into a meadow, as the meadow will be into an arable field. All this while the sediment of the river makes its way slowly into the lake, forming a mound or bank under the surface of the water, with a pretty rapid slope toward the lake. This mound increases by the addition of new earth, sand, and gravel, poured in over the slope; and thus the progress of filling up continually advances.