1862.] JTJEES— RITEE-VAIKETS, " ;401 



Martigny the Ehone "Valley suddenly tiiriis at right angles to its 

 former course, and runs N.N.E. for 38 miles into the Lake of Geneva. 



Above Martigny the valley runs along the strike of the rocks, and 

 seems to have been chiefly excavated in some comparatively soft 

 black slates lying between the hard crystalline rocks of the Oberland 

 on the north and those of the Main Chp'n on the south. But these 

 comparatively soft slates appear to continue in the same straight 

 line to the E.S.E. from Martigny through the hills crossed by the 

 Col de Balme, and down the Yalley of Chamouny, and into the hills 

 at the further end of that valley. Why, then, was not the valley of 

 the Ehone continued along this straight line into the Yalley of Cha- 

 -mouny, instead of turning at right angles at Martigny, and running 

 for nearly 40 miles across the strike of all the hard rocks on the north 

 of it ? Because, as I believe, a lateral valley formed by the Dranse 

 and Durance and their tributaries, running directly from the main 

 watershed down the original northern slope of the chain, was com- 

 menced before the excavation of the longitudinal valley of the Upper 

 Rhone was begun ; and the waters of those rapid rivers had always 

 sufficient power to cut that lateral valley deep enough to deflect the 

 waters that ultimately came into it from the west and carry them 

 them along it. 



Suppose all the valleys and hollows of the Alps to be fiUed up, 

 so that the present peaks shall be merely the summits of gently 

 swelling hills rising but little above the slightly undulating, smooth, 

 general surface of the mountain-chain. This smooth swelling surface 

 will be the Hmit of marire denudation. The first rains that fall upon 

 it will run directly off from the main watershed to the right and 

 left down the easiest and steepest slopes they can find, and com- 

 mence to form a number of lateral brooks along the bottoms of 

 these slopes, those rivulets running across the strike of the rocks. 

 As these lateral rivers deepen their channels, the waters running 

 into them on either side will also deepen and enlarge theirs ; and 

 thus will be commenced a number of valleys running along the 

 strike of the rocks parallel to the length of the chain, and therefore 

 called longitudinal valleys. When one of these coincides with a 

 band of soft rock, or rock more easily eroded by water than those on 

 each side of it, it may readily happen that the longitudinal valley 

 may ultimately become much larger than the lateral one by which 

 its contents are carried off. It cannot, however, become deeper, 

 because it is the drain originally caused by the excavation of the 

 lateral valley which is the only motive power for excavating the 

 longitudinal one. When once a lateral valley has succeeded in 

 cutting a sufficiently deep channel, the waters of the longitudinal 

 valleys that are afterwards poured into it cannot cross it, or overmount 

 the walls opposite to their own mouths, because they are inevitably de- 

 . fleeted down the lateral valley, and help to excavate it deeper ard 

 deeper below their junction with it. Hence the longitudinal valley 

 , and the part of the lateral one below their junction may be equally 

 wide and deep, and appear to be the result of one action, whilst the 

 lateral valley above the junction may be a narrow and broken.ravine 



