DAVIS: GLACIAL EROSION. 
297 
cover, rent on the steep valley sides where weathering comes to 
have a greatly increased value, and thickened where the ice streams 
have established their courses. This change corresponds to that 
between the ill-defined initial drainage in the early youth, and the 
well-defined drainage in the maturity, of the river cycle. 
It is probable that variations in rock structure will have permitted 
a more rapid development of the graded condition in one part of 
the glacial valley than in another, as is the case with rivers of water. 
Steady-flowing reaches and broken rapids will thus be produced in 
the ice stream during its youth; and the glacial channel may then 
be described as “ broken-bedded.” But all the rapids must be worn 
down and all the reaches must become confluent in maturity. It is 
eminently possible that the reaches on the weaker or more jointed 
rocks may be eroded during youth to a somewhat greater depth 
than the sill of more resistant or less jointed rock next down stream; 
and if the glacier should vanish by climatic change while in this 
condition, a lake would occupy the deepened reach, while the lake 
outlet would flow forward over rocky ledges to the next lower reach 
orjake. Many Norwegian valleys today seem to be in this con¬ 
dition. Indeed some observers have described broken-bedded val¬ 
leys as the normal product of glacial erosion, without reference to the 
early stage in the glacial cycle of which broken-bedded glacial chan¬ 
nels seem to be characteristic. Truly, it is not always explicitly 
stated that the resistance of the rock bed varies appropriately to the 
change of form in a broken-bedded channel; but the variations of 
structural resistance or firmness that the searching pressure and fric¬ 
tion of a heavy glacier could detect might be hardly recognizable to 
our superficial observations; and on the other hand the analogy of 
young ungraded glaciers with young ungraded rivers seems so nat¬ 
ural and reasonable that broken-bedded glacial channels ought to be 
regarded only as features of young glacial action, not as persistent 
features always to be associated with glacial erosion. If the glaciers 
had endured longer in channels of this kind, the “ rapids ” and 
other inequalities by which the bed may be interrupted must have 
been worn back and lowered, and in time destroyed. 
If a young glacier erodes its valley across rocks of distinctly dif¬ 
ferent resistances, a strong inequality of channel bed may be devel¬ 
oped. Basins of a considerable depth may be excavated in the 
weaker strata, while the harder rocks are less eroded and cross the 
valleys in rugged sills. Forms of this kind are known in Alpine 
valleys; for example, in the valley of the Aar above Meiringen 
