426 CARTER : GLACIATION OF DON AND DEARNE VALLEYS. 
watershed. They were mostly strike valleys, and therefore 
liable to suspicion, but, seeing that the glacial advance, as in- 
dicated by the drift, appears to have been parallel to the strike, 
it would seem certain by the geological structure of the country 
that any lake overflows would necessarily be found along the 
lines of shales between prominent sandstone beds. The cols 
naturally made by sub -aerial denudation would form natural 
channels of escape, and the overflow streams would quickly 
cut down such cols into long strike valleys. The valleys are not, 
and I do not think they can be expected to be, of the Cleveland 
type. The latter are probably the last relics of the retreating 
ice- sheets from Yorkshire, and are at sufficient elevations to 
be fairly preserved from rapid subsequent denudation. Hence 
they retain their sharpness of slope and definition of outline. 
On the other hand the Don-Dearne valleys must have belonged 
to a much earlier phase of glaciation, and the contours being low, 
and the whole area having again and again been invaded by lakes 
and inundations of water, have suffered so considerably as to 
modify in a great measure their glacial outlines. The streamless 
parts of these valleys, however, do not appear to have been much 
lowered, although the sides have been worn back considerably by 
subaerial denudation, and hence the cols will indicate pretty 
nearly the levels of glacial overflow. The courses of neighbouring 
streams being independent of these valleys, seem to imply that 
their formation has been quite apart from the present drainage 
system. 
In order to check these points with all possible care, the 
drift patches were taken as indicating lines of moraine, and 
therefore of glacial rest, and the results of the obstruction of the 
drainage by a glacier along each of these lines was plotted on 
the six-inch ordnance contour map, and the lakes outlined that 
would be produced by such obstructions of the drainage. As 
this work was done an overflow channel was found, corresponding 
to each lake. 
The first invasion of the ice would probably be through 
Frickley gorge into the low-contoured area between Darfield 
and Hickelton, the present flatness of which may be greatly 
due to the scouring of a long period of glacial occupation. 
