action, where it may be covered up with sand or mud. When the abraded land has been reduced to this level, it

FIG. 28. Section of submarine plain.^/. Land cut into caves, tunnels, sea-stacks, reefs and skerries by the waves, and reduced to a platform below the level of the sea (ss) on which the gravel, sand, and mud (d) produced by the waste of the coast may accumulate.

reaches a limit where erosion ceases, and where the sea, no longer able to wear it down further, protects it from injury by other agents of demolition.

We see, then, that the goal toward which all the wear and tear of a coast-line tends, is the formation of a more or less level platform cut out of the land. Yet an attentive study of the process will convince us that in the production of such a platform the sea has really had less to do than the atmospheric agents of destruction. An ordinary seacliff is not a vertical wall. In the great majority of cases it slopes seaward at a steep angle; but if it had been formed, and were now being cut away, mainly by the sea, it ought obviously to have receded fastest where the waves attack it-that is, at its base. In other words, if sea-cliffs retired chiefly because they are demolished by the sea, they ought to be most eroded at the bottom, and should therefore be usually overhanging precipices. That this is not the case shows that some other agency is concerned which causes the higher parts of a cliff to recede faster than those below. This agency can be no other than that of the atmospheric

forces-air, frost, rain, and springs. These cause the face of the cliff to crumble down, detaching mass after mass, which, piled up below, serve as a breakwater, and must be broken up and removed by the waves before the solid cliff behind them can be attacked.

II. Accumulations formed by the Sea. It is not its erosive action that constitutes the most important claim of the sea to the careful study of the geologist. After all, the mere marginal belt or fringe within which this action is confined forms such a small fraction of the whole terrestrial area of the globe, that its importance dwindles down when we compare it with the enormously vaster surface over which the operations of the air, rain, rivers, springs, and glaciers are displayed. But when we regard the sea as the receptacle into which all the materials worn off the land ultimately find their way, we see what a large part it must play in geological history.

During the last fifteen years great additions have been made to our knowledge of the sea-bottom all over the world. Portions of the deposits accumulating there have been dredged up even from the deepest abysses, so that it is now possible to construct charts, showing the general distribution of materials over the floor of the ocean.

Beginning at the shore, let us trace the various types of marine deposits outward to the floors of the great oceanabysses. In many places, the sea is more or less barred back by the accumulation of sediment worn away from the land. In estuaries, for example, there is often such an amount of mud in the water that the bottom on either side is gradually raised above the level of tide-mark, and forms eventually a series of meadows which the sea can no longer overflow. At the mouths of rivers with a considerable current, a check is given to the flow of the water when it reaches the sea, and there is a consequent arrest of its detritus. Hence, a bar is formed across the outflow of a river, which during floods is swept seawards, and during on-shore gales is driven again inland. Even where there is no large river, the smaller streams flowing off the surface of a country may carry down sediment enough to be arrested by the sea, and thrown up as a long bank or bar running parallel with the coast. Behind this bar, the drainage of the interior accumulates in long lagoons, which find an outflow through some breach in the bar, or by soaking through the porous materials of the bar itself. A large part of the eastern coast of the United States is fringed with such bars and lagoons. A space several hundred miles long on the east coast of India is similarly bordered.

But the most remarkable kind of accumulation of terrestrial detritus in the sea is undoubtedly that of river-deltas. Where the tidal scour is not too great, the sediment brought down by a large river into a marine bay or gulf gradually sinks to the bottom as the fresh spreads over and mingles with the salt water. During floods, coarse sediment is swept along, while during low states of the river nothing but fine mud may be transported. Alternating sheets of different kinds of sediment are thus laid down one upon another on the sea-floor, until by degrees they reach the surface, and thus gradually increase the breadth of the land. Some deltas are of enormous size and depth. That of the Ganges and Brahmaputra covers an area of between 50,000 and 60,000 square miles-that is, about as large as England and Wales. It has been bored through to a depth of 481 feet, and has been found to consist of numerous alternations of fine clays, marls, and sands or sandstones, with occasional layers of gravel. In all this great thickness of sediment, no trace of marine organisms was found, but land-plants and bones of terrestrial and fluviatile animals occurred.

Turning now to the deposits that are more distinctively those of the sea itself, we find that ridges of coarse shingle, gravel, and sand are piled up along the extreme upper limit reached by the waves. The coarsest materials are for the most part thrown highest, especially in bays and narrow

FIG. 29.-Storm-beach ponding back a stream and forming a lake; west coast of Sutherlandshire.

creeks where the breakers are confined within converging shores. In such situations, during heavy gales, storm-beaches of coarse rounded shingle are formed sometimes several yards above ordinary high-tide mark (Fig. 29). Where a barrier of this kind is thrown across the mouth of a brook, the fresh water may be ponded back to form a small lake, of which the outflow usually escapes by percolation through the shingle. In sheltered bays, behind headlands, or on parts of a coast-line where tidal currents meet, detritus may At

accumulate in spits or bars. Islands may in this way be gradually united to each other or to the mainland, while the mainland itself may gain considerably in breadth. Romney Marsh, on the south-east coast of England, for instance, a tract of more than 80 square miles, which in Roman times was in great part covered by the sea at high water, is now dry land, having been gained partly by the natural increase of shingle thrown up by the waves and partly by the barriers artificially erected to exclude the sea.

While the coarsest shingle usually accumulates towards the upper part of the beach, the materials generally arrange themselves according to size and weight, becoming on the whole finer as they are traced towards low-water mark. But patches of coarse gravel may be noticed on any part of a beach, and large boulders may be seen even below the limits of the lowest tides. As a rule, the deposits formed along a beach, and in the sea immediately beyond, include the coarsest kinds of marine sediment. They are also marked by frequent alternations of coarse and fine detritus, these rapid interchanges pointing to the varying action of the waves and strong shore-currents. Towards the lower limit of breaker-action, fine gravel and sand are allowed to settle down, and beyond these, in quiet depths where the bottom is not disturbed, fine sand and mud washed away from the land slowly accumulate.

The distance to which the finer detritus of the land is carried by ocean-currents, before it finds its way to the bottom, varies up to about 200 miles or more. Within this belt of sea, the land-derived materials are distributed over the ocean-floor. Coarse and fine gravel and sand are the most common materials in the tracts nearest the land. Beyond these, lie tracts of fine sand and silt with occasional