20 E^tJMi OF Sueell's study of 



There may be reckoned up on the Durance very distinct forms of five of 

 these ancient lakes, extending from the neck of Mont-Genfeve, where its 

 source is, to the boundary of the department. Vestiges of the same 

 phenomenon are to be seen in the valleys of the Grand Bueoh, and of the 

 Petit Bueoh. They are to be seen, again, in the vaUey of the Drac, and in 

 that of the Eomanche. In general, all the great valleys of the department 

 present similar traces. Some of these lakes existed within historical times, 

 and we may remark, in fine, that the same appearances have been observed 

 in a great many other places, and on all sorts of rivers. 



In this, then, we find a general mode of action, of which traces are 

 constantly reproduced in a certain kind of valleys, to which may be attri- 

 buted not only the formation of the valleys but also the formation of their 

 thalwegs, which two things are, he states, distinct and different. 



There are, he remarks, valleys which seem to have been created solely by 

 the erosion effected by waters flowing at first in a simple depression in the 

 soil ; other valleys seem to have originated in dislocations of the soil opening 

 clefts into which the waters have afterwards precipitated themselves. But 

 in valleys of both formations the action of the waters has invariably been 

 the same, and it has produced the same results. Thrown upon an irregular 

 surface of soil, they have followed at first the line of the greatest inclina- 

 tion ; then they have modified this. Whilst this was going on there has 

 been thus formed the most stable curve of the bed; under the double 

 influence of the friction of the waters tending to reduction to a minimum, 

 and the resistance offered by the soil tending to a maximum : this curve, 

 thus formed, is the thalweg. 



Thus are brought together and harmonized a great many facts, the 



explanation of all of which are embodied in one formula — vague it may be 



but unique, general, and of universal applicability. 



If the valleys be studied in their topographical aspect several laws may 

 be discovered, covered by this regulated appearance, which seem to be 

 entirely the result of chance. Amongst these are two beautiful laws evolved 

 by Brisson, which may be verified here in most of the necks of the moun- 

 tains. I adduce only one illustration of each. The first is supplied by the 

 col of the Laiiteret, situated between two water-courses, parallel and flowing 

 in opposite directions — La Romanche and La Guisanne. The other is supplied 

 by the col of the Bayard, situated in the district where the Drac and the 

 Durance, after they have both flowed from east to west, separate, — the one 

 directing its course towards the north, the other towards the south. A high- 

 way which passes from the second basin into the first shows distinctly the 

 thalweg passing by the col from the one into the other. 



By this notice of the action of water in flood we are prepared for entering 

 upon the more special study of torrents. 



In the torrent, or what, in accordance with the English application of 

 that term, may be called the torrent-bed, there are noticeable these three 

 distinct parts,— the basin drained by the torrent or funnel-shaped hollow 

 from which the waters are collected, called the hasdn de reception ; the 

 gorge and channel by which the waters are carried off from this funnel- 

 shaped basin, called the canal d'ecoidement ; and the deposit of detritus at 

 the lower extremity of this, called the lit de ddjection. 



To this last great importance is attached, as by detritus borne down by 

 torrents many fruitful fields have been buried under a layer of debris under 



