i895. THE CANONS OF SOUTHERN FRANCE. 383 



Cantal, and there was a time when they meandered over the surface 

 of the limestone which is now so deeply trenched by their valleys. 



Many people admit that ordinary valleys have been excavated by 

 rain and rivers, but when they see a river passing between lofty 

 walls of solid rock, they ask how liquid water can saw out such a 

 passage. The geologist replies that a river in flood is not only water, it 

 is full of sand and mud, and, moreover, it sweeps stones along the 

 bottom of its channel. It is the silt, sand, and stones that erode ; the 

 stream is only the motive power which works and guides the machinery. 



Still, it is objected that few rivers, either in England or on the 

 Continent, seem to be deepening their channels at the present time, 

 even when they are in flood. This is perfectly true. A river above 

 the influence of the tides cannot sink its channel below the level of the 

 sea, and, when it has cut its valley down to a certain slope, it ceases 

 to have any power of deepening so long as the land remains stationary. 

 If, however, the land is raised, the erosive power of the rivers is 

 renewed. Upheaval increases the height through which the water 

 has to fall before it reaches the sea, and enables the rivers to cut their 

 way deeper into the land. Subsidence lessens the relative height of 

 the land, and not only stops the deepening process, but causes the 

 streams to deposit the materials they carry. 



It is because geologists for a long time failed to perceive that 

 the erosive power of a river is controlled and limited by the move- 

 ments of the land that certain sceptics have refused to believe that 

 valleys have been made by rivers alone. The anonymous author of 

 a book entitled " Scepticism in Geology," referring to the rocky barrier 

 known as the Iron Gate just below the great gorge of the Danube, 

 remarks that 'geologists have failed to explain to us how water- 

 erosion, having (as they assert) cut through cliffs 2,000 feet high, 

 should have stopped short at this petty barrier-reef." The answer, I 

 should think, is that the river had reached its base level of erosion when 

 it had cut down through the 2,000 feet of rock, and it will never 

 remove the Iron Gate unless that part of Europe is once more raised 

 to a higher level above the sea than that at which it now stands. The 

 rivers of Lozere are not now deepening their channels, because their work 

 is done ; but their present incapacity for erosion is no proof of their 

 having been incapable of making the canons during a period of upheaval. 



Those who imagine that the gorges which trench these limestone 

 districts have been formed by the enlargement of fissures or by the 

 falling in of caverns, forget that they are only part of a river-channel 

 and that this river-channel is part of a great system of drainage. 

 They forget, too, that in other countries similar ravines have been cut 

 through rocks in which caverns do not exist. 



The fact that the main rivers of the causses have their ultimate 

 sources beyond and outside the limits of the causses country is alone 

 sufficient to point out the manner in which the canons have been 

 formed. This has not escaped the notice of some of the French 



