Alpine Valleys and Alpine Lakes. 85 



the alpine world its present form and aspect, to attribute to 

 each its own share, and to find fair evidence in support of his 

 conclusions, is the formidably difficult task of the alpine geolo- 

 gist ; and even if he be fortunate enough to gain the help of the 

 ablest workers in the field of physical science, it will be long 

 before that task can approach to a conclusion. In considering 

 whether or no it is probable that the agent which has excavated 

 valleys whose vertical depth below the ridges that enclose them 

 often exceeds an English mile, and in two of the valleys referred 

 to by Professor Tyndall is at least a mile and a half, the first 

 step is to examine the mode of action of existing glaciers. 



The whole mechanical effect of a glacier upon its bed is 

 directed towards the removal of inequalities, whether in the bot- 

 tom or sides. It is quite understating the case, to say that if 

 valleys were excavated by ice through strata at all approaching 

 to uniform hardness, they would all tend to the same type of 

 equal slope in the bed, and absence of projecting masses in the 

 containing walls, of which the valley of the Rhone between Mar- 

 tigny and Sion gives the best known example. Any one who 

 has watched the manner in which the bottom of a glacier slides 

 over the concavities without touching them, and applies all its 

 immense grinding power to the convex portions of its bed, will 

 admit that a great degree of inequality in the resistance of dif- 

 ferent portions of the rock would not prevent a valley scooped 

 out exclusively by glacier-action from approaching nearly to the 

 same uniform model. Fortunately for the lover of natural 

 scenery, the fact is widely different from what it would be if the 

 theory were generally well founded. The most common type of 

 an alpine valley is that which is formed by a succession of level 

 basins rising in steps one above the other as they approach the 

 head of the valley, and connected by gorges whose opposite sides 

 often approach near together, and which are always much nar- 

 rower than the basins that they link together. There is often 

 distinct evidence, and usually good reason to suspect, that these 

 basins were originally lakes ; and in the walls of the gorges in- 

 termediate it is common to see proof that at some former period 

 the depth of these lakes must have been greatly increased by 

 barriers of rock, that once held them in and at a later period 

 were cut through by streams of ice or water. I do not pretend 

 that the type of valley above described is universal; but it is 

 too common, both in the Alps and other mountain countries, not 

 to present a formidable difficulty in the way of Professor Tyn- 

 dalVs bold hypothesis. Even though 1 should admit, as I 

 cannot do, that such lakes as still exist, or have existed, could 

 be formed by a glacier, the existence of a succession of such ap- 

 proximately level steps, separated by steep slopes and narrow 



