Through all the periods the actual rocks formed are very similar—shales, limestones, sandstones, clays. When any rocks happen to have preserved neither plant nor animal remains it is almost impossible to tell to which epoch they belong, except from a comparative study of their position as regards other rocks which do retain fossils. This depends on the fact that the physical processes of rock building have gone on throughout the history of the globe on very much the same lines as they are following at present. By the sifting power of water, fine mud, sand, pebbles, and other débris are separated from each other and collected in masses like to like. The fine mud will harden into shales, sandgrains massed together harden into sandstones, and so on, and when, after being raised once more to form dry land, they are broken up by wind and rain and brought down again to the sea, they settle out once again in a similar way and form new shales and sandstones; and so on indefinitely. But meantime the living things, both plant and animal, have been changing, growing, evolving, and the leafy twig brought down with the sandgrains in the flooded river of one epoch differs from that brought down by the river of a succeeding epoch—though it might chance that the sandgrains were the same identical ones. And hence it is by the remains of the plants and animals in a rock that we can tell to which epoch it belonged. Unless, of course, ready-formed fossils from an earlier epoch get mixed with it, coming as pebbles in the river in flood—but that is a subtle point of geological importance which we cannot consider here. Such cases are almost always recognizable, and do not affect the main proposition. From the various epochs, the plants which have been preserved as fossils are in nearly all cases those which had lived on the land, or at least on swamps and marshes by the land. Of water plants in the wide sense, including both those growing in fresh water and those in the sea, we have comparatively few. This lack is particularly remarkable in the case of the seaweeds, because they were actually growing in the very medium in which the bulk of the rocks were formed, and which we know from recent experiments acts as a preservative for the tissues of land plants submerged in it. It must be remembered, however, that almost all the plants growing in water have very soft tissues, and are usually of small size and delicate structure as compared with land plants, and thus would stand less chance of being preserved, and would also stand less chance of being recognized to-day were they preserved. The mark on a stone of the impression of a soft film of a waterweed would be very slight as compared with that left by a leathery leaf or the woody twig of a land plant.