time. Still he only estimated the deposition of sediment under water at the known rate of accumulation all over the valley already dry land, except at the periodical floods.

It is, I think, a well-established fact, that the delta of this river has encroached on the Gulf "several leagues" since Louisiana was settled. The annual increase is estimated at three hundred feet, or say five miles in a century. At this rate the alluvial plain from Natchez or Port Hudson down cannot be over six thousand or seven thousand years old. At Natchez the valley is only twelve miles wide, with bluffs two hundred feet high, so there is no evidence that the gulf ever extended above this place.

Sir C. Lyell gives us an estimate of the age of a number of the river deltas in Europe, and the highest figure he gives us is seven thousand five hundred years; still it is not clear that he intended in any case to include the whole delta, although the rate of increase in a number of instances was well authenticated. Was he afraid of their testimony? It would

seem so.

Another evidence of the existence of an eroding force now no more is the lakes found excavated out of the solid sedimentary rocky strata, such as Lake Pepin on the Mississippi, twenty-five miles long, five miles wide, and four hundred feet deep, with the strata of rock on the opposite sides exactly corresponding. Chains of this kind of lakes are now found in other river valleys, with no conceivable force now in existence to excavate them.

Professor Hind, of the Geological Survey of Canada and the Hudson's Bay Territory, gives a particular description of chains of lakes in the river valleys of the north-west. These lakes, nine in number, and from three or four to forty-five miles long, and uniformly of about one mile in breadth, occupying chasms in the sedimentary rocky strata of three hundred feet in depth, could not possibly have been formed by the small stream that meanders through the valley. In his Report, page 57, he says: "The first view of the south branch of the Saskatchewan, fully six hundred miles from the point where the main river disembogues into Lake Winnipeg, filled me with astonishment and admiration. We stood on the banks of a river of the first class, near half a mile broad, and flowing with a swift current

not more than three hundred and fifty miles from the Rocky Mountains, where it takes its rise. We had reached this river by traversing either within it or on its banks, for a distance of two hundred and seventy miles, a narrow deep excavation continuous from the valley of one great river to that of another, and exhibiting in many features evidences of an excavating force far greater than the little Qu' Appelle, which meandered through it, was at first blush thought capable of creating. How were the deep lakes hollowed out? lakes filling the breadth of the valley, but during the lapse of ages not having increased in breadth, preserving too for many miles such remarkable depths, and although in some instances far removed from one another, yet maintaining those depths with striking uniformity. What could be the nature of the eroding force which dug out narrow basins fifty-four to sixty feet deep at the bottom of a valley already three hundred feet below the slightly undulating prairies, and rarely exceeding one mile in breadth? It was easy to understand how a small river like the Qu'Appelle could gradually excavate a valley a mile broad and three hundred feet deep. The vast prairies of the north-west offer many such instances: the Little Louis river, for example, in passing through the Blue Hills; the Assiniboine, for a hundred and fifty miles, flows through a broad deep valley, evidently excavated by its waters; the rivers in Western Canada often flow in deep eroded valleys; but in no instance to my knowledge are deep and long lakes known to occupy a river valley where the altitude and character of the rocks preclude the assumption that they have been occasioned by falls, without having increased its width by the action of their waves on their banks, or without leaving some traces of the force which had excavated them."

On page 118, speaking of an immense granite bowlder in the bed of the Qu'Appelle, he says: "This would involve the assumption that the Qu'Appelle valley dates the epoch of its erosion anterior to the last submergence of the continent, affording an illustration of a river valley before the epoch of the bowlder drift. The physical aspect of the country is by no means opposed to this view."

The occurrence of ancient river valleys on this continent has already attracted attention. In his illustrations of Surface

Geology, Dr. Hitchcock says: "Some of the erosions that have been described in this paper are clearly beds of antediluvian rivers; that is, of rivers existing upon this continent before its last submergence beneath the ocean, which beds were deserted when the surrounding surface emerged from the water, although essentially the same rivers as existed previously must have been the result of drainage." Then follows an enumeration of ten ancient river beds in Canada, (Niagara,) New England, and the State of New York. Thus far Professor Hind.

All rivers that are now cutting their own channels through the solid rock present the appearance of Niagara, the Falls of the St. Anthony, and other rivers of this class we have seen, a gorge between perpendicular cliffs. Lake Pepin could not have been formed in this way as it is five miles wide. Were all the waters of the Mississippi to be precipitated over a fall of such a height, one third or one fourth of a mile is ample for all its waters. The Falls of St. Anthony are not over one eighth of a mile wide. If Lake Pepin was cut in this way, say one fourth of a mile wide at first, and then widened by the slow abrasion of the elements on the mural precipices, how long would it take? This is a point of the utmost importance to decide, for the waves have not undermined the perpendicular cliffs of solid rock. It has been established, by experiments instituted at the Smithsonian Institute, in Washington, that common marble in that latitude will abrade one inch in ten thousand years. Common limestone will wear away much faster than marble. For the sake of obtaining something like data to guide us as to the time required, we will admit that limestone will abrade ten times as fast as marble; this doubtless is far above the mark. At this rate it would require Niagara, St. Anthony, and Lake Pepin one hundred and fifty millions of years to widen five miles. If it has taken all this time to widen Lake Pepin to its present breadth, how deep must it have been at first, seeing all the abrasion of the materials five miles wide and four hundred feet in thickness, and all the sediment brought down by the Mississippi in all these years have not sufficed to fill up this tranquil lake? Were the Missouri turned through it, it would fill it up in twenty years.

But if it has taken so long to widen these gorges to their present extent, and the delta of the Mississippi has all the time

been encroaching on the gulf at the rate of five miles in a century, we should now have a delta at the month of this river that would extend one million five hundred thousand miles, or sixty times around the globe. We are not yet through our difficulties growing out of this theory; for every mile of increased length in the delta necessitates a rise of about three inches all over the alluvial valley, or embankments to that extent, to keep the waters within the banks. If embanking had been resorted to, they would now have been seventy five miles high; at the same time Lake Itasca, the head and source of the Mississippi, is only one thousand eight hundred feet above the Gulf of Mexico, making an average descent of less than seven inches in a mile for its whole length. At the present rate of filling up, the bluffs along the lower half of the Mississippi then, will all be covered with river mud in much less than twenty thousand years. The changes going on in this river are not peculiar to it. The Yellow River in China has embankments forty feet high, to keep the waters in the channel. We cannot suppose embanking was resorted to over three thousand years ago, which would make a rise of fifteen inches in a century in the embankments, some of course in excess of the actual rise in the bed of the stream. The river Po in Italy has embankments of twenty feet in about two thousand years, or one foot in a century.

At the present rate of the increase of all the river deltas known, the Nile, Ganges, Danube, Po, Yellow River, etc., we are carried back not over a period of from six to ten thousand years. The gorge at Niagara, according to the present rates of recession and the testimony of our most eminent geologists, has not been in existence over that time. At the rate of three hundred feet a year, or five miles in a century, the Mississippi cannot be over six or seven thousand years since its tide of mud first rolled to the ocean.

Another fact in proof that lakes like Lake Pepin, and those mentioned by Professor Hind, were not cut through the solid rock by the forces now in existence, and then widened by the slow action of the weather on their mural sides, is that in various places in the old silurian formation there are chasms in the solid limestone rock, some of them occupied by foxes and ground-hogs, and some of them channels for water in times of

flood. These chasms have been fairly exposed to the action of the elements on their sides, but have not, in all past time, widened but six, eight, or ten inches, for few of them exceed ten inches in breadth. The Falls of Niagara and St. Anthony show no symptoms of widening. Thus, from many sources, we come to the same conclusion, that there formerly existed an eroding force that has denuded vast regions of the surface sediment, and in all probability cut out the channels of our large sedimentary rivers about as we now see them. This eroding force ceased to act before the drift period, as all drift rivers have manifestly cut their own channels with about the present amount of force found in existence.

There is another point regarding rivers worthy of notice, though bearing more remotely on the subject in hand. All rivers that convey sediment are pushing out deltas into the sea, ocean, bay, or gulf, as the case may be. In all their lower courses they are filling up, while in all their upper courses they are wearing down. If there are permanent barriers or obstructions to the regular flow of a stream, as indestructible rocks or lakes, each separate portion of the river is independent as to this law. As the delta pushes out, so the sediment accumulates over all the alluvial plain still higher and higher up the stream. This makes a movable midway point in all streams, all below this point is filling up, while all above is wearing down. This midway point is near the mouth of the Platte on the Missouri, and not far from Keokuk on the Mississippi. Above the mouth of the Platte there are high bottom lands not now reached by the greatest flood, because there the channel is wearing down; but all below is subject to flood from bluff to bluff each freshet. We saw again these high bottom lands a little above Quincy, Illinois, with great accuinulations of bowlders on them, showing conclusively that this midway point is not below Quincy, and that this great valley between the bluffs existed as now, before the drift period.

•

It then results from these laws that all rivers are slackening their flow, and straightening out their sinuosities, for in proportion to the velocity of the current in these rivers, so is the sinuosity increased. In descending the Missouri as we approach the Mississippi, the reaches in the river become longer