The risk of making very serious errors in computing the amount of sediment discharged, unless proper precautions are taken, is well illustrated in the case of the determinations made by Messrs. Brown and Dickson, to which reference has already been made. Although their report shows that they took great pains in order to arrive at correct results—in fact, they computed the total annual quantity of sediment discharged to within a cubic foot—after all, instead of being correct to this minute quantity, they gave a total more than fourfold what it ought to be. A somewhat similar discrepancy exists in reference to the denudation of the basin of the Ganges. The time required to lower its surface by one foot is, according to one estimate, 2,358 years; according to another, 1,751; and according to a third, only 1,146 years. The first figure is probably nearest the truth. Still, these differences show both the difficulty of the problem and the necessity of caution in adopting any of these results as correct.
By far the most trustworthy determinations of the whole are those of the Mississippi by Messrs. Humphreys and Abbot, which may be relied upon as not far from the truth. But, supposing the estimates in the foregoing table to be perfectly correct, can we assume that their mean may be safely taken as probably representing the average rate of denudation of the whole earth? I would most unhesitatingly reply, Certainly not. The Rhone and Po are full of glacier mud from the Alps; and the amount of sediment which they carry down may give us the rate of denudation of Switzerland, but certainly not that of the whole earth, or even of Europe. The same may be said of the Ganges, which is charged with the mud which it brings down from the Himalaya Mountains. The Hoang-Ho, or Yellow River, is an exceptionally muddy river; in fact, it derives its name from the vast quantity of yellow mud held by its waters in a state of solution. It was probably the exceptionally muddy character of the Po, the Rhone, the Ganges, and the Yellow River which attracted attention, and led to observations being made of the sediment they contain. Rivers more unsuitable than these to give us the average denudation of the earth’s surface could not well be selected. Among the seven rivers in the table, leaving out of account the small Scottish stream, the Nith, with its basin of only 200 square miles, there are only two, the Mississippi and the Danube, that drain countries which may be regarded as in every way resembling the average condition of the earth’s surface. I would choose the Mississippi as being superior to the Danube, for two reasons: (1) because the rate of denudation of its basin has been more accurately determined; and (2) because the area of its basin not only exceeds that of the Danube as five to one, but better fulfils the necessary conditions, as Sir Charles Lyell has so clearly shown. “That river,” says Sir Charles, “drains a country equal to more than half the continent of Europe, extends through twenty degrees of latitude, and therefore through regions enjoying a great variety of climate, and some of its tributaries descend from mountains of great height. The Mississippi is also more likely to afford us a fair test of ordinary denudation, because, unlike the St. Lawrence and its tributaries, there are no great lakes in which the fluviatile sediment is thrown down and arrested on its way to the sea.”[[27]] There is no other river in the globe which to my mind better fulfils the required conditions. It is no doubt true that the rate of denudation of the basin of the Mississippi is probably less than that of Switzerland, Norway, and the Himalayas, where glaciers abound, and certainly less than that of Greenland and the Antarctic continent; but, on the other hand, this rate is certainly much greater than that of the whole continent of Africa, Australia, and large tracts of Asia, where the rainfall is much smaller. One foot in 6,000 years may, therefore, I think, be safely taken as the average rate of denudation of the whole surface of the globe.
The average rate of denudation in the past probably not much greater than in the present.—The belief has long prevailed that the rate of denudation was much greater in past ages than it is now; but I am unable to perceive any good grounds for concluding that such was the case at any time since the beginning of the Palæozoic period. Various reasons have, however, been assigned for this supposed greater rate; and to the consideration of these I shall now very briefly refer.
It has been thought that at some remote epoch of the earth’s history, when the moon was much nearer and the day much shorter than now, the rate of denudation would, owing to the erosive power of the enormous tides which would then prevail, be much greater than at the present day. This, however, is very doubtful. There is nothing in the stratified rocks which affords any support to the idea of great tidal waves having swept over the land, at least since the time when life began on our globe. Such a state of things would have destroyed all animal life. “The Palæozoic sediments,” as Professor A. Winchell remarks, “have been deposited, for the chief part, in quiet seas. The deep beds of limestones and shales are spread out in sheets continent-wide, which testify unmistakably to placid waters and slow deposition.”[[28]] But high tides, not sweeping over the land, would not increase the rate of denudation to the extent supposed. High tides silt up a river channel more readily than they deepen it. A higher tide would probably produce a greater destruction of sea-coast: it would tend to increase the rate of marine denudation, but this would not materially affect the general rate of denudation. For, as the present rate of marine denudation is to that of subaërial denudation only as 1 to about 1,700,[[29]] it would take a very large increase in the rate of marine denudation to affect sensibly the general result. Suppose the rate of marine denudation to have been, for example, ten times as great during the Palæozoic age as it is now (which it certainly was not), it would only have shortened the time required to effect a given amount of denudation of the whole earth by 9 years in 1,700, i.e. by little more than one-half per cent.
Again, it is assumed that the greater rate of terrestrial rotation in the early ages would produce certain influences which would in turn bring about a greater amount of denudation. The rate of rotation has been slowly decreasing for ages, and in Palæozoic times it must, of course, have been greater than at present. A more rapid rotation would increase the velocity of the trade and anti-trade winds, and would thus tend to augment the action of those meteorological agents chiefly effective in the work of subaërial denudation. Here again the testimony of geology is negative. We have no geological grounds to conclude that the winds of Palæozoic times were stronger than those at the present day. The heat was no doubt greater, and perhaps there was more rain; but, on the other hand, there would be less frost, snow, ice, and other denuding agents.
There is one cause which would, perhaps, be more effective than any of the foregoing: viz. the periodic occurrence of glacial epochs. When a country is buried under ice, the erosion of the surface is great. But it must be borne in mind that the influence of rain, rivers, and other denuding agents now in operation would then, in the glaciated regions, be almost nil. Besides, the greater part of the materials ground off the rocks would be left on the land; and, when the ice disappeared, it would be found in the form of a thick mantle of boulder clay—a mantle which would protect the rocky surface of the country for thousands and tens of thousands of years from further denudation. This is shown by the fine striæ on the rocky surface, made perhaps more than 50,000 years ago, remaining under the boulder clay as perfect as the day on which they were engraved. But, more than all this, a very considerable part of the 1 foot presently being removed off the country in 6,000 years consists of the loose materials belonging to the glacial epoch, such as sands, gravels, and boulder clay, which are being swept off the surface by rain and river action. Were it not for this, the present rate of subaërial denudation would not be so high as it actually is. Taking all things into consideration, it is, I think, obvious that the average rate of denudation since the beginning of Palæozoic times was probably not much greater than at the present day.
How the method has been applied.—Having determined what appears to be the probable average rate of subaërial denudation, we may now proceed to consider the way in which this rate has been applied to measure past geological time. There are two ways in which it may be applied for this purpose. It may (1) be applied directly: knowing the thickness of strata which may have been removed by denudation, we can easily tell, from that rate, the time it required to effect their removal. If we have evidence, for example, that at some epoch 1,000 feet of stratified rock were carried away, then, on the assumption that the rate of denudation was the same at that epoch as now, we have 1,000 × 6,000 = 6,000,000 years as the required time. (2) It may be applied indirectly: knowing the thickness of the strata, we may estimate the time required for their formation. This is the way in which it has usually been applied, but, as we shall see, it is the less satisfactory way of the two.
Dr. A. Geikie gives the land area of the globe as 52,000,000 square miles, and that of water as 144,712,000 square miles.[[30]] We may thus take the proportion of land to water roughly as 1 to 3; about one-quarter of the earth’s surface being land, and three-quarters water. One foot, therefore, removed off the surface of the land would cover the whole globe with a layer 3 inches thick, or the entire sea-bottom with a layer 4 inches thick.
If we knew the total quantity of stratified rock on the globe, we could easily tell the time that would be required for its formation. Most geologists would, I believe, be inclined to admit that, if spread uniformly over the entire globe, it would form a layer of at least 1,000 feet in thickness. In such a case the time required for its deposition would be as follows:
1,000 × 6,000 × 4 = 24,000,000 years.