360. Again, if we consider a valley as a space, which perhaps with many windings and irregularities, has been hollowed out of the solid rock, it is plain, that no force of water, suddenly applied, could loosen and remove the great mass of stone which has actually disappeared. The greatest column of water that could be brought to act against such a mass, whatever be the velocity we ascribe to it, could not break asunder and displace beds of rock many leagues in length, and in continuity with the rock on either side of them. The slow working of water, on the other hand, or the powers that we see every day in action, are quite sufficient for this effect, if time only is allowed them. 361. Some valleys are so particularly constructed, as to carry with them a still stronger refutation of the existence of a _debacle_. These are the longitudinal valleys, which have the openings by which the water is discharged, not at one extremity, but at the broadside Such is that on the east side of Mont Blanc, deeply excavated on the confines of the granite and schistus rock, and extending parallel to the beds of the latter, from the Col de la Segne to the Col de Ferret; its opening is nearly in the middle, from which the Dora issues, and takes its course through a great valley, nearly at right angles to the chain of the Alps, and to the valley just mentioned. From the structure of these valleys, Saussure has argued very justly against Buffon's hypothesis, concerning the formation of valleys by currents at the bottom of the sea.[182] It affords indeed a complete refutation of that hypothesis: and it affords one no less complete of the system which Saussure himself seems on some occasions so much inclined to support. For if it be said, that this valley was cut out by the current of a _debacle_, that current must either have run in the direction of the valley of Ferret, or in that of the Dora, which issues from it. If it had the direction of the first, it could not cut out the second; and if it had the direction of the second, it could not cut out the first. Besides, the force which excavated this valley must have been _nothing_ at the two extreme points, viz. at the Col de la Segne and the Col de Ferret, and must have increased with the distance from each. It can have been produced, therefore, only by the running of two streams in opposite directions, on a surface that was but slightly uneven, these streams at meeting taking a new direction, nearly at right angles to the former. A clearer proof could hardly be required than is afforded in this case, that what is now a deep valley was formerly solid rock, which the running of the waters has gradually worn away; and that the waters, when they began to run, were on a level as high, at least, as the tops of those mountains by which the valley is bounded toward the lower side. [Footnote 182: Voyage aux Alpes, tom. ii. § 920.]