LACUSTRINE AND FLUVIATILE BRICK-EARTHS. The great variability of brick-earths deposited in river valleys is reflected to some extent in those laid down in lakes, though the size of the latter is frequently a controlling factor. The chief difference consists in the broader expanse of the sediment laid down--especially in large lakes--and variation in structure is not so noticeable horizontally. Let us consider a simple case in which a lake is fed by a large river bringing down abundant sediment. The lake acts as a species of settling tank, and the method of deposition of the sediment by the river is mainly guided by the velocity of the stream. The tendency under normal conditions is for the river to commence parting with its sediment immediately on entering the lake. The detritus alluded to is only held in suspension by the velocity of the water; when the latter is checked, as on entering the lake, the grosser pieces subside, and as its rapidity becomes progressively curbed, medium-sized fragments are compelled to give way, until at last only very minute particles are left in the water. In due time most of these also are deposited. Thus gravel is laid down before grit, grit before sand, and sand before clay. If the velocity of the river always remained the same, we should be presented with thick accumulations of the same character in sharply defined areas. But it is always changing. With every storm and every steady rain the motion of the river becomes greatly accelerated, with the result that the deposits for the time being are deposited farther out in the lake than in more quiescent periods. In this way we may have a gravel thrown down on sand, sand on clay, and so on. From the foregoing observations it will be gleaned that, in general, deposits in large lakes are more persistent in character than are river deposits; indeed, in very large sheets of water, as Lake Superior, Lake Erie, &c., they are in this sense more comparable with sediment of marine origin.