The proportion of extinct species of quadrupeds is more numerous in the fossil remains of the gravel No. 1 than in that indicated as No. 2; and in No. 3 they agree more closely, sometimes entirely, with those of the existing fauna. The usual absence or rarity of organic remains in beds of loose gravel and sand is partly owing to the friction which originally ground down the rocks into small fragments, and partly to the porous nature of alluvium, which allows the free percolation through it of rain-water, and promotes the decomposition and removal of fossil remains. The loose transported matter on the surface of a large part of the land now existing in the temperate and arctic regions of the northern hemisphere, must be regarded as being in a somewhat exceptional state, in consequence of the important part which ice has played in comparatively modern geological times. This subject will be more specially alluded to when we describe, in the eleventh chapter, the deposits called “glacial.” Denuding Power of Rivers affected by Rise or Fall of Land.—It has long been a matter of common observation that most rivers are now cutting their channels through alluvial deposits of greater depth and extent than could ever have been formed by the present streams. From this fact it has been inferred that rivers in general have grown smaller, or become less liable to be flooded than formerly. It may be true that in the history of almost every country the rivers have been both larger and smaller than they are at the present moment. For the rainfall in particular regions varies according to climate and physical geography, and is especially governed by the elevation of the land above the sea, or its distance from it and other conditions equally fluctuating in the course of time. But the phenomenon alluded to may sometimes be accounted for by oscillations in the level of the land, experienced since the existing valleys originated, even where no marked diminution in the quantity of rain and in the size of the rivers has occurred. We know that many large areas of land are rising and others sinking, and unless it could be assumed that both the upward and downward movements are everywhere uniform, many of the existing hydrographical basins ought to have the appearance of having been temporary lakes first filled with fluviatile strata and then partially re-excavated.