If we look at the point under discussion as purely a question of fact, to be determined by positive evidence and not by argument, the observations of Boussingault are, both in the circumstances they detail, and in the weight of authority to be attached to the testimony, among the most important yet recorded. They are embodied in the fourth section of the twentieth chapter of that writer's _Economie Rurale_, and I have already referred to them on page 191 for another purpose. The interest of the question will justify me in giving, in Boussingault's own words, the facts and some of the remarks with which he accompanies the details of them: "In many localities," he observes,[196] "it has been thought that, within a certain number of years, a sensible diminution has been perceived in the volume of water of streams utilized as a motive power; at other points, there are grounds for believing that rivers have become shallower, and the increasing breadth of the belt of pebbles along their banks seems to prove the loss of a part of their water; and, finally, abundant springs have almost dried up. These observations have been principally made in valleys bounded by high mountains, and it is thought to have been noticed that this diminution of the waters has immediately followed the epoch when the inhabitants have begun to destroy, unsparingly, the woods which were spread over the face of the land. "These facts would indicate that, where clearings have been made, it rains less than formerly, and this is the generally received opinion. * * * But while the facts I have stated have been established, it has been observed, at the same time, that, since the clearing of the mountains, the rivers and the torrents, which seemed to have lost a part of their water, sometimes suddenly swell, and that, occasionally, to a degree which causes great disasters. Besides, after violent storms, springs which had become almost exhausted have been observed to burst out with impetuosity, and soon after to dry up again. These latter observations, it will be easily conceived, warn us not to admit hastily the common opinion that the felling of the woods lessens the quantity of rain; for not only is it very possible that the quantity of rain has not changed, but the mean volume of running water may have remained the same, in spite of the appearance of drought presented by the rivers and springs, at certain periods of the year. Perhaps the only difference would be that the flow of the same quantity of water becomes more irregular in consequence of clearing. For instance: if the low water of the Rhone during one part of the year were exactly compensated by a sufficient number of floods, it would follow that this river would convey to the Mediterranean the same volume of water which it carried to that sea in ancient times, before the period when the countries near its source were stripped of their woods, and when, probably, its mean depth