M. Agassiz on Glaciers, Moraines, and Erratic Blocks. 367 



moraines from the dykes which are formed by the overflow of 

 water, and from the masses of debris, more or less extensive, 

 which are produced by avalanches. The dykes are very irre- 

 gular, and extend only to small distances, when they gradually 

 become level ; the masses of debris from avalanches again as- 

 sume the shape of very flattened cones, debouching from the 

 valleys, and gradually losing themselves in the plains ; whilst 

 the moraines are continuous and parallel dykes or walls, which 

 extend along the two sides of the valleys, and are formed of 

 rounded blocks, that have been evidently triturated, so to 

 speak, 171 situ, one against the other, as actually happens 

 at the margin of the modern glaciers, which occur in long 

 and narrow valleys. The blocks of avalanches, on the con- 

 trary, are angular ; those of the dykes formed by torrents 

 may, indeed be rounded, when, for example, they are derived 

 from broken up moraines, but then they extend themselves in 

 irregular plane surfaces or sheets ; and when they are derived 

 from recent avalanches, they are equally angular, unless in their 

 progress they meet with ancient moraines, which they carry 

 along with them, and mix with their own constituent parts. 



That we may be convinced of the accuracy of these facts, 

 all that is necessary is to traverse the valley of Chamouni, there 

 tracing the moraines which approach nearest to the glaciers, or 

 to mount jjerpendicularly upon the sides of the valley of the 

 Rhone, between Saint-Maurice and Martigny, on the left bank, 

 above the Pissevache, near the hamlet of Chaux-Fleurie (Tsau- 

 fria), or on the opposite bank, ascending from the village of 

 Morcles to the baths at I^avey. The rubbish of the late de- 

 bacles of the Dent du Midi, the immense avalanches whose 

 traces are every where seen, and the numerous dykes formed by 

 the Rhone, will, moreover, enable us justly to appreciate the 

 difference which exists in those various phenomena produced 

 by such different causes. 



Lateral valleys, it should bo remarked, exhibit the same phe- 

 nomena, which may be seen by ascending the course of tb.e 

 Aven^on, as far as the glacier of Paneyrossaz. 



In perambulating these valleys, nothing has exited my at- 

 tention more than a peculiar polished appearance which those 

 rocks present on which glaciers have moved, an appearance 



H b 2 



