34 sisuMfi OP stjbelij's study op 



ment being that which renders exploitation most easy. Now, wherever the 

 woods have been cleared in this manner, at the place of each clearance the 

 vegetable soil has been carried away by the waters ; a furrow is formed 

 there, of little depth at first, but which digs away more and more, extends 

 itself upwards, enlarges itself, and soon constitutes a complete torrent. In 

 the intermediate stripes, where the trees have been spared, it is seen to be 

 altogether different. There — with the same soil, under the same exposure, 

 under the same slope, and this often very steep, — the ground has been held 

 firm, and the contour has been respected by the waters. In going over the 

 forest we often traverse thus a succession of zones, the differences of which 

 are striking. We may even catch sight of intermediate shades, which fill 

 up the contrast. We see nascent ravines in parts where the stumps thickly 

 standing bear testimony to a recent destruction of trees. We see completed 

 torrents in other parts, where the indications of the ground, and the inform- 

 ation given by the inhabitants, bear testimony to trees having been destroyed 

 in times more remote. We are thus well assured that we are not taking 

 the effect for the cause, when we affirm that it is the destruction of trees in 

 the clearance which has formed the ravine, and not the ravine which has 

 formed the clearance." 



As is the case with the gorges, so does it appear to be the case with the 

 hasdns de reception. There is no question in regard to the fact that the 

 effect of such a conformation of the basin drained by a torrent, as has 

 been described, is to bring a large body of water, falling over a great extent 

 of surface, to concentre in the orifice of the gorge ; but the allegation of 

 Surell is that the form of the basin is itself the product of the long-continued 

 violent action of waters, collected first in a recess of the mountains, and 

 flowing over a soil of little compactness and cohesion ; and he accounts for 

 the absence elsewhere of certain characteristics of the torrents of the High 

 Alps by stating that, where the ground presents more resistance, and where 

 the climate is less rigorous, there may be formed only brooklets and moun- 

 tain streams. Similar torrents are not met with in the Vosges, in the 

 C6vennes, or in the Auvergne. In the Lozfere there are vallats which are 

 not without characteristic features of torrents of the third class, such as are 

 of frequent occurrence between Briangon and the Monestier, and along the 

 Guissanne ; but these, through their weakness, scarcely resemble true tor- 

 rents, though, compared with vallats, they are torrents of great energy. 



The torrents of the Pyrenees, generally called Gaves in the district, are 

 very rapid water-courses in deep cuttings, often losing themselves in sub- 

 terraneous canals, but they should be classed with mountain streams or 

 torrential rivers. And no torrents are met with in the mountains of La 

 Corse, or in those of the Jura. But torrents similar to those of the High 

 Alps are found in a portion of the mountains of the department of the Is6re 

 of the Dr6me, and of the Lower Alps, which belong to the same formation! 



A chapter is devoted to the consideration of climatal or atmospheric 

 influence, and another to the effects attributable to the character of the 

 geological formations of the locality. 



In regard to climate, he shows that the elevation of the High Alps brings 

 them into the region of snow. When this accumulates all winter over an 

 extensive area, and under the powerful rays of spring melts in great quanti- 

 ties all at once, the process being often accelerated by the arrival of warm 



