118. This very essential point is introduced with great accuracy. The lake of Geneva was one of those which I had in view in my Letters to Dr. Hutton, and the mention of it was indispensable. The other instances adduced by Mr. Playfair are some lakes of North America, while at the same time he remarks, that the annual increase is there much less considerable than in those of the Alps; "as the rivers "which supply these vast reservoirs are none of them very great." He then returns to his general subject.

"In order to give uniform declivities to the rivers, "the lakes must not only be filled up or drained, "but the cataract, wherever there is one, must be

worn away. The latter is an operation in all cases "visible. The stream, as it precipitates itself over "the rocks, hurries along with it, not only sand and gravel, but occasionally large stones, which grind " and wear down the rock with a force proportioned "to their magnitude and acceleration.

119. Thus far I perfectly agree with Mr. Playfair; and I shall dwell for a moment on the consequence of this operation. It is certain, and I had myself remarked it, that when the rain waters began to assemble in torrents among mountains, they found in their course abrupt sections of rocks, whence they precipitated themselves in cataracts; as also cavities, which they must have filled up with such materials as they were carrying along with them, before they could convey farther any portion of these materials. This

indeed Mr. Playfair acknowledges; but here he forgets what I remarked (§ 97. a.) that he had founded his theory of the hollowing out of vallies, by the slow but long" continued action" of the running waters, on the circumstance that all of them together form a system of vallies communicating with one anether, and having such a nice adjustment of their "declivities, that none of them join the principal

valley, either on too high or too low a level." After quoting this passage, I observed that Mr. Playfair had very correctly stated what ought to be the case, in order to make it appear probable that "all vallies * were the work of the streams that flowed in them; but that he had forgotten what ought to be, when he came to support his opinion by actual facts; for all the cataracts, which have fallen, and are still falling, -down the sides of the principal vallies, manifestly show that some lateral valley had joined them on too high a level; and all the cavities, which are filled up by water, or alluvial earth, make it no less evident that streams had there joined them on too low a level. Here then we have original inequalities still existing in vallies, which running waters, far from having produced the vallies themselves, have had a tendency to efface: but, even supposing that these inequalities should have altogether vanished, these effects, from the traces left behind them, would be perfectly, distinct from the formation of rallies. Mr. Playfair's opinion, therefore, that these inequalities are tending to be effaced, has no connection with the system that the vallies have been thus produced; and it is under this point of view that I shall examine the following passage, § 322.

4

"The

"The smooth surface of the rocks in all waterfalls, "their rounded surface, and curious excavations,

are the most satisfactory proofs of the constant "attrition which they endure; and, where the rocks are deeply intersected, these marks often reach to a great height above the level on which the water now flows. The phenomena, in such instances, are among the arguments best calculated to remove all incredulity respecting the waste which "rivers have produced, and are continuing to pro"duce. They suffer no doubt to remain, that the height and asperity of every waterfall are continually diminishing; that innumerable cataracts ❝are entirely obliterated; that those which remain are verging toward the same end; and that the "falls of Montmorenci and Niagara must ultimately disappear."

120. It is evident that this is quite unconnected with any proof that running waters have produced the vallies by erosion of the land; we are told only that "constant attrition" wears away the original "asperity" of the rock. But is this operation really carried on, in the manner described by Mr. Playfair? By no means; and he might have seen, in my second Letter to Dr. Hutton, that numerous waterfalls have been obliterated, and are tending to obliteration, not by the erosion of the rocks, but by the accumulation of the materials brought down by torrents, under the form of semi-cones. And with regard to waterfalls, which do not propel large fragments of stone, after they have once rounded the edge of the rock, they are 12

so far from wearing it down, that they flow only over the moss, which they permit to grow upon it; a case which I have observed in numerous instances, and particularly in the fall of the Elbe, near its source, in the Giant's Mountains. In the letter above-mentioned, I introduced the description given by M. RAMOND DE CHARBONNIERE of the fall of the Adour at Trames-Aigues, in the Pyrenees, which, as he observes, has "laid aside its ancient fury," and 、suffers vegetation to spread nearer and nearer to its course in every part. After this digression on the subject of waterfalls, which has evidently no favour. able connection with the hypothesis of the production of vallies by running waters, Mr. Playfair returns to the lakes, and thus proceeds

"Though there can be no doubt of the justness "of the preceding conclusions, when applied to "lakes in general, some apparent exceptions occur, "in which the progress of draining and filling up

seems to have been suspended, or even to have goue "in a contrary direction. These exceptions consist "of the lakes which appear to have received a greater quantity of materials than was sufficient to have filled them up. Such, for example, is the lake of

4.6

Geneva, which receives the Rhone descending "from the Fallais, one of the deepest and longest "vallies on the surface of the earth. Now, if this "valley, or even a large proportion of it, had been "excavated by the Rhone itself, as our theory leads c us to suppose, the lake ought to have been entirely filled up; because the materials brought down by

•

"the

← the river seem to be much greater than the lake, on any reasonable supposition concerning its original magnitude can possibly have received. What,

then, it may be said, has become of all that the "Rhone has brought down and deposited in it? "The lake, at this moment, retains, in some places, "the depth of more than 1000 feet; and yet, of all "that the Rhone carries into it, nothing but the pure "water issues. If it has been continuing to dimi

nish, both in superficial extent, and in depth, from "the time when the Rhone began to run into it, what must have been its original dimensions? I cannot pretend to remove entirely the difficulty which is here stated; yet I think the following remarks may go some length in doing so.'

33

121. We come at length to the critical point of the question, which, as far as relates to its nature, is very correctly stated by Mr. Playfair; although, in some respects, the "difficulty" is not brought forward in its full force, as I propose hereafter to show: but here a more essential remark presents itself. It is merely as an "exception" that he speaks of the lake of Geneva; it was, therefore, incumbent on him to prove in what way it formed an exception to every other phenomenon that was observed of the kind. He himself acknowledges that he "cannot pretend "to remove entirely the difficulty;" but is this a case, where we ought to rest satisfied with attempts alone? a case, where it is sufficient to make a few "remarks," which may go some length only towards removing the *bjection? This is an important question, on which I

must