356 The Glacial Theory of Prof. Agassiz. 



not permit us to suppose that any cascade could ever have existed. In 

 other cases, a long, sinuous, dry, water-worn gutter or channel is ob- 

 served, the course of which runs across, instead of along, the natural 

 declivity of the ground. The study of the glaciers has enabled Agas- 

 siz to find a key to these enigmatical phenomena, which had perplexed 

 previous inquirers. Streams of water flow along the surface of a gla- 

 cier, and when one of these falls into a fissure which is open to the bot- 

 tom, it often forms a cascade, and cuts a round cavity in the rock with the 

 gravel and sand which it either finds there, or carries down with it, as 

 some of our rivulets work out the hollows termed cauldrons. When no 

 fissure exists, the stream sometimes cuts a funnel or shaft {couloir, en- 

 tonnoir) through the ice by the action of gravel. If the glacier is trav- 

 elling downwards, the cascade will travel with it, and convert the round 

 cavity in the rock into a long gutter ; or, supposing the water to reach 

 the bottom without falling in a cascade, still, in finding an issue below 

 the glacier, it will be compelled to follow the sinuous openings left by 

 inequalities in the bottom of the ice, and thus take a course at variance 

 with the natural inclination of the surface. We have here an explana- 

 tion of the creux, or holes, and the long water-worn gutters found in 

 such unlikely situations, which bear the local names of lapiaz or karren. 

 These are chiefly observed where the rock is soft, and are seldom vis- 

 ible on the granite. 



Stratified Gravel on sides of Valleys. — When a small portion of 

 stratified gravel or sand is found adhering to the side of a valley, high 

 above its bottom, the conclusion usually come to is, that a lake or arm 

 of the sea had once filled the whole up to that level, and that the de- 

 posit is merely a remnant of one much more extensive. Agassiz has 

 shown that this conclusion may be erroneous. When the streamlets 

 flowing on or under a glacier, cannot find an escape below, they often 

 form small lakes at the surface on its flanks, and, as in other lakes, the 

 gravel and sand carried into these, arrange themselves in strata. This 

 stratified deposit may be continuous with, and form as it were a portion 

 of, a lateral moraine, which is not only unstratified, but which follows 

 a line probably far from level. Here again the study of existing gla- 

 ciers enables us to explain very anomalous appearances. 



Glacier Barriers. — A glacier descending a valley opening into an- 

 other, sometimes pushes forward till it forms a dike or barrier across 

 the latter. Behind this the water collects and constitutes a lake, which 

 augments till it breaks the icy barrier, or flows over it, producing fright- 

 ful inundations. In 1815 the glacier of Getroz formed a dike across 

 the valley of Bagnes. This dike went on increasing till 1818, when 

 it was 500 feet high and 800 long. It was then burst by the pressure 

 of the waters of the Drance, which committed terrible ravages as far 



