The greater or less specific gravity of the different constituents of rock doubtless aids in separating them into distinct masses when once disintegrated, though there are veined and stratified beds of sand where the difference between the upper and lower layers, in this respect, is too slight to be supposed capable of effecting a complete separation.[409] In cases where rock has been reduced to sandy fragments by heat, or by obscure chemical and other molecular forces, the sandbeds may remain undisturbed, and represent, in the series of geological strata, the solid formations from which they were derived. The large masses of sand not found in place have been transported and accumulated by water or by wind, the former being generally considered the most important of these agencies; for the extensive deposits of the Sahara, of the deserts of Persia, and of that of Gobi, are commonly supposed to have been swept together or distributed by marine currents, and to have been elevated above the ocean by the same means as other upheaved strata. Meteoric and mechanical influences are still active in the reduction of rocks to a fragmentary state; but the quantity of sand now transported to the sea seems to be comparatively inconsiderable, because--not to speak of the absence of diluvial action--the number of torrents emptying directly into the sea is much less than it was at earlier periods. The formation of alluvial plains in maritime bays, by the sedimentary matter brought down from the mountains, has lengthened the flow of such streams and converted them very generally into rivers, or rather affluents of rivers much younger than themselves. The filling up of the estuaries has so reduced the slope of all large and many small rivers, and, consequently, so checked the current of what the Germans call their _Unterlauf_, or lower course, that they are much less able to transport heavy material than at earlier epochs. The slime deposited by rivers at their junction with the sea, is usually found to be composed of material too finely ground and too light to be denominated sand, and it can be abundantly shown that the sandbanks at the outlet of large streams are of tidal, not of fluviatile origin, or, in lakes and tideless seas, a result of the concurrent action of waves and of wind.