40 e:6sume of suhell's study of 



surface of the globe ; when we observe that the bottom of these valleys is 

 flat, levelled by the waters, and entirely formed by their alluvial deposits ; 

 when, going back to the most ancient historic times, we see in Egypt, in 

 China, in India, &c., the first societies of men, descending little by little from 

 the heights, occupied in struggling against the inconstancy, and the 

 tremendous overflowings of these rivers, — may we not believs that all these 

 courses have had, during a long course of centuries, changes in their 

 channels such as those which the Durance exhibits now ? But, gradually, 

 the field of these devagations has been confined, as is seen so distinctly in 

 the case of torrents, and like these they have ended in being confined within 

 their present banks. The Durance, on the contrary, is still existing in its 

 second stage — that of instability — which has succeeded to the first, charac- 

 terized by a succession of lakes, and to which in course of time a period of 

 stability will gradually succeed." 



And inferring that the most stable rivers of to-day have passed through 

 an epoch of change, of course corresponding to the second period of torrents, 

 he goes on to say, — " In the study of these same rivers there have been 

 collected a multitude of observations which show that they have had in a 

 former age to open their thalweg, and to create their slopes the same as we 

 have said has been done by the Durance, and the same as we see being 

 done under our eyes by the torrents in the interior of the continents ; they 

 furrowed coutineuts, they fiUed up basins, and the traces of these phenomena 

 are still very apparent. In approaching these as they cast there immense 

 deltas — ever enlarging deltas — on which sites for entire kingdoms have been 

 found, which deltas constitute true beds of dejection. Thus have these 

 rivers at a certain epoch of their existence acted as the torrents have done 

 in the first period of their history." 



And he goes on to say, — " Resuming this discussion, I wiU undertake to 

 show in the action of torrents a faithful and miniature image of that which 

 passed or will pass in all rivers in general. 



" In all I see three consecutive stages, succeeding each other in the same 

 order, and dividing their existence into three distinct periods — First, a 

 period of corrosion and elevation, which prepares the bottom of the thalweg 

 and puts throughout its course the slopes in equilibrium with the resistance 

 of the soil and the friction of the waters. It has for its end to determine 

 the longitudinal profile of the water-coui'se. 



" Second, a period of devagation, when the rivers seek that form and those 

 bondings of the course which correspond to the greatest stability (for the 

 rectilineal course is not the most stable, since it does not necessarily lead 

 the current over those points where the bank is most solid and least likely 

 to be changed). In this the action of the waters is confined to going hither 

 and thither on the same level without perceptibly carrying away or eleva- 

 ting the bottom ; it is the liquid mass which displaces itself rather than the 

 soil. The result of this stage is to fix the laying out of the line of the 

 course, or, if the expression be preferred, to determine the plan of it 



" Third, in fine, a period of permanence, when the waters may overflow 

 their banks but ever return again to their place in an unchanging bed 



" The violence of torrents in the first period has been seen. There oueht 

 to be the same m the first period of rivers; and this analogy may sei-ve to 

 explam the formation of those aUuvial deposits spread out in such a mass 

 m the greater part of extensive vaUeys. If it be true that the mountains 

 have been elevated successively in the midst of convulsions of which nothing 



