mass of water took an hour and a half in rushing from the glacier to Martigny. The same event took place from the same cause, and with nearly similar results, in 1595. Torrents may therefore scoop out ravines in certain formations, and produce effects which appear considerable, because we judge of them by comparison with our own feeble means. But how diminutive and circumscribed are these changes produced in the configuration of the globe, compared with the long and broad valleys which furrow in vast numbers the immense surface of the earth, and to the formation of which neither the torrents nor great rivers which exist at the present day have in any way contributed, as we shall presently demonstrate. 2. _Action of Rivers._ The action of _rivers_ must be examined under two very different circumstances, or at two different parts of their course. _First_, When they are compressed between mountains, whether at no great distance from their source, or even at the middle of their course. _Secondly_, When they have reached broad valleys, whose declivity is slight, or plains which commonly surround their mouth. In the first case, these rivers partake of the impetuosity and power of torrents. They often run with rapidity, and in great quantity, at the bottom of narrow and deep valleys: they are as it were inclosed in channels, whose vertical walls appear as if cut by art. The first idea which presents itself to all who have seen these appearances for the first time, and who are satisfied with first impressions, is, that these streams, which are pretty powerful and always very impetuous, have dug these deep grooves; and if sometimes the hardness of the rocks and the height of the precipices which form their sides, appear too great for those small streams that meander at their feet, what cannot be attributed to their immediate power is attributed to the continued action of time. Without examining how long a series of ages it would be necessary to admit, before the rivers which we have mentioned above, and the water-courses encased in the deep valleys of the Alps, Pyrenees, Jura, Grampians, &c. could have scooped their valleys, on which their present action is so slow that no one has yet been able to estimate it; without examining if this long series of ages agrees with the phenomena, which preclude our attributing so remote an antiquity to the actual state of the earth’s surface, a question of too much importance to be treated indirectly; it will be sufficient to mention here four sorts of observations, in order to be persuaded, or at least to suspect, that the present rivers, even supposing them ten times the size that they are, could not have scooped out the deep channels at the bottom of which they run. 1. We must recur to the period when the ranges of hills which border the present valleys were not as yet scooped out, but were united in such a manner as not to leave any hollow between them, or merely a slight original depression.