sea, leaving its former circuit as a great pool, or moat, as it is called by the common folk along the banks of the Connecticut River. It often happens in the lower Mississippi that the course of the river around the promontory of the ox-bow is ten or more miles in length while the space across the neck is less than a mile in distance. When the river finally breaks across the neck, the whole system of rebounds of its currents against the banks, from the point of change downward to the mouth, may become altered. The points which before were in process of erosion may become the seats of deposition, and those which previously were gaining may begin to wear away. In this manner a river, in time, wanders to and fro across its whole valley, taking material from

*This term "moat " deserves a place in our geological language. for the reason that it is a brief and expressive word

for the topographic feature, ill-described in our present sys

tem of naming. Moreover it preserves, in an interesting way, a memory of mediaeval conditions. The name was

doubtless given because of the likeness which the early set

tlers saw between these circular ditch-like pools and the defences which, in the seventeenth century, were still familiar objects about many of the country houses in Great Britain. I shall therefore use the term in the present writing and hereafter in the sense above indicated.

one side, sorting it over, removing that part which is fine enough to be borne away by the current, and rebuilding the remainder into the alluvial plains.

We are now prepared to consider a very peculiar and most important function which these alluvial plains perform in the physical life of the earth. In such a valley as the Mississippi, we have probably not less than fifty thousand square miles of alluvial plains which have been formed of the waste removed from the rocks in the torrential portions of the streams in the mountain and hill districts of the valley. This alluvial material is, on the average, not less than fifty feet thick. It is therefore equivalent to about five hundred cubic miles of matter. Now, this great river carries out to sea about one-twentieth of a cubic mile of sediment each year. This sediment which goes into the sea is in small part directly derived from the action of the mountain torrents; in larger part, it is composed of waste taken from the alluvial plains by the wanderings of the various streams which constitute the Mississippi system of waters. It therefore

follows that the average time required for the sediment discharged from the mouth of the Mississippi to make its way from the head-waters to the sea is not less than ten thousand years. As soon as a pebble or other bit of rock is laid away in the alluvial terrace, it begins to decay; the vegetable acids which penetrate the mass in which it finds lodgement favor its disintegration. When it is turned over by the stream at the time of encroachment on its resting place, it probably falls to pieces, the finer bits are hurried onward by the stream, those too coarse for the current to control are again stored away in the bank to await further decay. In this manner the alluvial material lying on either side of rivers is a great storehouse, or

storage and decay, the seas could not be supplied with the débris essential for the maintenance of the life which they contain; for that life, unlike the life of the land, does not depend on the soil of the ocean floors, but upon the dissolved matter contained in the water, from which the marine animals and plants take all their store of nutrition. This nutrition comes mainly from the land-waste brought to the sea in the state of solution by the streams, and, as we have just seen, the comminution and solution of this waste depends upon the work which goes on in the laboratories of the alluvial plains.

We have now seen the way in which the water operates upon the surface of the stream-beds. At the source of the

rather we should say laboratory, in which sediments are divided and brought into a chemical condition which permits them to be taken into the control of the waters and borne away to the ocean, in order to become rebuilt into strata, which are in time, with the growth of the continents, to become dry land and be again subjected to this erosive work. Were it not for this system of alluvial

mountain-torrents, a pound of water has in it, by virtue of its height above the level of the sea, a great store of energy, which it may apply to the erosion of the earth's surface. Let us suppose that when it comes to the earth it is three thousand feet above the ocean's level. It has then as much force to expend as would be required to lift it to that height above the sea. At first the stream plays

the part of spendthrift with this energy, the greater portion of the force is expended in brawling with the stones and in beating against the limits which confine it. In the first five miles or so of its path to the sea it uses up in its descent perhaps one-third of its dynamic resources, and so, for the last thousand miles, it may not have more power at its command than it gave out in the first five miles of its journey.

Thus our streams, though always growing larger, are continually becoming less and less powerful in proportion to the weight of water which flows over their beds. In the lower portion of their courses they have very little capacity for eroding the rocks over which they flow, except where that power is due to some peculiar circumstances. They deepen their beds slowly, and the greater portion of this deepening is accomplished by the corrosion or chemical decay of the rocks over which they flow. Still, certain peculiar circumstances may give them a chance to cut down the floors of their lower channels. This work is done in either of the following ways: When the lateral swinging of the river-beds to and fro through the alluvial plain dislodges great forest-trees from the bank,

these trees often have great quantities of stones entangled in their roots. These roots are thus held against the bottom while the trees are swept onward by the current, and so the entangled stones rasp upon the bed and serve to wear the channel deeper. Again, it often happens in cold countries that the rivers are deeply frozen, and during the winter season, in the shallow water, the loosened stones of the bottom may be entangled in the ice. When the time of "breaking up" comes, the sheets of ice, as they float downward in great fields, strike against the banks of the river where there is a sharp bend in the channel, and, owing to their great momentum, are heaped up in a wall of fragments, which may in a few minutes dam the river quite across. Owing to the pressure to which these cakes of ice are subjected, they freeze together and the whole of one of these ice dams or gorges becomes a solid mass. When this happens, as is easily conceived, the stream rises rapidly, forming a great lake above the dam, while it drains away below, and thus, as in the Ohio River, these dams may have a difference of twenty or thirty feet of water above and below their obstructions. In a brief time the pressure of the water above the dam pushes

the whole mass forward, grinding it upon the bottom and the sides, and so powerfully eroding the rock-bed in which the stream flows.

As long as the river flows onward over rocks of uniform hardness, especially where the strata lie in horizontal attitudes, the course of the stream generally exhibits a uniform descent. Various accidents in the attitude of the rocks may, however, give rise to rapids or waterfalls. These features in the course of a river are so important in its mechanism, especially with reference to the interests of man, that they deserve a careful consideration, which we shall now give to them.

Waterfalls and rapids owe their existence in the main to one of three conditions of the bed rock. These conditions are as follows: First, the path of the stream may be crossed by a dike or a vein, which are rifts in the rocks, filled with some deposit brought into them by the action of water or forced to its place in the condition of a lava. Where these dike- or vein-materials are softer than the neighboring rock over which the stream flows, the river easily cuts them down and they create no interruption to its course. Where, however, as is often the case, the rocks which fill the fissures are harder than the materials which formed its walls, the river is obstructed, and we generally have a cataract, that is, an irregular fall, in which the stream takes no one conspicuous plunge. Another case in which a local hardening of the stream-bed produces a waterfall is where a stream, flowing over rocks which may be horizontal in their attitude, encounters a coral reef, formed on the old sea-floors in which the strata were deposited. In this case the crowding together of the fossil corals may make the rock much firmer than the neighboring portions of the strata, and so produce a decided interruption in the uniform descent of the stream. Only one important case of reef-cataract is known to me, that which occurs in the Ohio at Louisville, where coral-reef in the Devonian period has so far interrupted a gentle descent of the river as to create a formidable obstruction, only passable, save during the flood-times of the river, by means of a canal extending

VOL. IV.-16

from the head to the base of the rapid. The most common condition which leads to the formation of a waterfall, the condition which gives us the greater part of the fine falls of the world, is where a river flows across strata which dip or sink downward in the earth toward the head-waters of the stream. In this condition, wherever a hard bed of the strata overlies a soft deposit, the stream inevitably forms a waterfall.

The first two of the above named classes of waterfalls demand no very extensive consideration. Those produced by dikes and veins are generally conspicuous only in the torrential portion of a river-system. The veins and dikes account for a very large part of the little cataracts which diversify our mountain torrents. Coral-reefs are so rare in our older rocks that they are seldom cut by the streams, and are therefore not often seen, even by the professional student of geology. The third group, in which each plunge of the fall is due to the upstream slope of strata, alone demands some special consideration.

Falls due to inclined strata can best be represented by Niagara, perhaps the noblest of all such geological accidents. As is shown in the diagram, [p. 148] we have at Niagara Falls a tolerably hard layer of limestone, belonging to a division of the Silurian age, which has indirectly received its name from this great cataract. This Niagara limestone is underlaid by a considerable thickness of softer shaly rocks known as the Clinton group. The waters of the Niagara River plunge over the hard rim afforded by the limestone and descend about a hundred and seventy feet, acquiring in this movement a very great velocity. At the base of the fall, the water strikes against a mass of hard fragments which in succession have tumbled down from the resisting upper layer. These fragments, set violently in motion, cut out the soft material, the erosion of which is also aided by the violent whirls of water and of spray driven against the shaly beds in the space behind the fall. From this wearing action, the soft materials are constantly working backward more rapidly than the hard upper layer is worn away, and so, from time to time, the projecting shelf over the waterfall is deprived of sup

--

though the retreat of the fall is slow, it will in a very brief time, in the geological sense of that word, lead to certain momentous consequences. When the hard layer of Niagara limestone passes

port and tumbles to the base in fragments, which, in turn, are used for the further erosion of the soft deposits. In Niagara, as in all other waterfalls of this description, the border of rock over which the plunge takes place is constantly and pretty rapidly working up stream. The fall is progressively decreasing in height, as is shown in the diagram, and in the end, when the hard layer has descended to the general level of the stream-bed, especially when the softened limestone rocks have passed altogether below that level, the fall will disappear; first passing into the stage of a cataract and afterward vanishing altogether.

།་

In the case of Niagara Falls the rate of retreat is about three feet in a century; this rate is very variable. It was probably more rapid in the past than at pres

below the bed of the river, the stream will then cut upon rocks of another constitution, making for a time certain small falls at a higher geological level; but in the course of ages, much less long than those which have elapsed since the birth

of this waterfall, the gorge of the river will extend up into the basin of Lake Erie, draining away a considerable portion of that fresh-water sea. We shall then, if the continent retains its present height above the level of the sea, have another system of cataracts, in the passage between Lake Erie and Lake Huron, which will also in time be worn away. Other cataracts will then form at the exit of Lake Michigan; and thus the lower lakes of our great American system would be diminished in area, or perhaps even disappear. At a yet later stage, we may look for diminution in the size of Lake Superior, though that basin, owing to the strong wall which separates it from the lower lakes, is destined to endure long after the last-named basins have been diminished or entirely drained away.

ent, for the reason that the undercutting power of the falling water diminishes with the decrease in the height of the precipice over which it plunges, and this height has been growing less and less ever since the fall began to be. Al