are commonly, without further question, supposed to be indicative of the site of ancient lakes, which have been slowly but completely filled: the supposition is often correct, but it is sometimes erroneous. Rapid rivers, which, in times of inundation, drift coarse materials down their rough beds, and deposit them in the expansions of their valleys, are thus partly choked in their courses, and turned into new channels. Thus they wander irregularly over a large area, every where filling it, to about the same height, with a mass of partial deposits, related to the successive positions of the channel, which, when unconfined by man, seeks always the lowest passage. On a cross section of such a valley, these many distinct streams of gravel and sands appear nearly as in the annexed diagram.
But such a distribution of materials appears not to occur in lakes; whether they receive sediments from gentle streams, rapid rivers, or sudden inundations. The reason of this is the great lateral diffusion of motion in water. Where any great depth of quiet water is interposed on the path of a river, the lacustrine sediments assume various modes of arrangement, depending on their own fineness, and the velocity of the water by which they are hurried along.
Deep Lakes on the Course of a River.—On entering a deep lake, the mingled sediment of a river is subjected to a new influence, the descending force of gravity, in addition to the direct horizontal force imparted by the current, and the lateral movements which it occasions. Each particle, in conse-
