458 
POPULAR SCIENCE REVIEW. 
of time, would at length have the effect of turning a plain into 
a succession of hills and steep-sided valleys. (See fig. 1.) 
All countries present similar instances of plains, or plateaux, 
which have been formed into mountain chains by the channel- 
ling out of valleys. A noble example is Scandinavia ; and in 
our own country we have the Cotteswold Hills, the Chalk 
Downs, Cannock Chase, and the Pennine Chain. That the 
valleys by which they are intersected have to a great extent 
been scooped out by streams, rains, &c., is undeniable ; but that 
many of the larger ones are not due to these agents is, I think, 
equally certain, for several reasons, which I shall now proceed 
to explain. In the first place, it is evident that during the 
slow process of elevation the sea would have had opportunities 
of wearing channels in the softer strata, which it would not 
be slow to take advantage of, and which afterwards became 
the chief river valleys ; in the second place, many of these 
valleys were formed before the Drift-period, as they are par- 
tially filled with boulder clay ; and in the third place, many of 
them are of such extent and width as to be altogether dispro- 
portioned to the rivers or brooks which flow through them. 
Now I hold that all true river valleys bear some proportion to 
the size of their streams. Not only is this the case, but there 
are some well-formed valleys which have no brooks whatever 
for a certain distance, forasmuch as they cross the watersheds. 
The valley of Todmorden,on the borders of Lancashire and York- 
shire, is a remarkable example of this kind (fig. 2.) At Little- 
borough it contracts into a narrow, steep- sided glen, bounded 
by lofty cliffs several hundred feet in height, with a smooth 
and almost imperceptible slope. At two miles north of Tod- 
morden the watershed is crossed, and for some distance there 
is no perceptible stream. Yet the valley is as well and sharply 
defined here as in any other part of its course. The valley of the 
Calder, between Todmorden and Burnley, affords another illus- 
tration ; as do those of Whitworth, near Rochdale ; Sabden, 
near Clitheroe ; and that along the course of the Caledonian 
Railway, near Moffat. Indeed, most hilly districts afford 
similar examples. Nothing can be more certain than that 
such valleys are not due to river action, as in some parts they 
contain no rivers. 
There are also in many of the uplands of England double 
valleys, or valleys within valleys. I have for a long time felt 
persuaded that in such cases we have before us examples both 
of marine and river denudation, the limits of the latter being 
clearly defined. Such instances of double valleys occur pro- 
bably in every hilly district. They are to be found in the 
Pennine Chain and in the Cotteswold Hills, and their general 
form is represented in the annexed wood-cut (fig. 3). The 
