254 SUBTERRANEAN EROSION, 
One and all, however, issue from underground charged with 
matter derived from the subterranean erosion of the clays, sands, 
and gravels under or through which they have passed, Waste 
so constant cannot have continued for a long period of time 
without leaving evidences of its effects. 
For instance, in Ithell’s sandpit (see fig. 1), the sands and 
gravels beneath the Upper Boulder Clay have been proved for 
30 feet without reaching the rock (Bunter Pebble-beds.) If the 
clay, sand, and gravel were of a uniform thickness across the 
whole line of section from S.S.E. to N.N.W., the clay would be 
upon the horizon A-A, and the sand and gravel should crop 
out at Z-Z instead of at B-B. It should be noticed that the 
height of the clay from ridge to ridge is uniform throughout, so 
that an observer at either end has an uninterrupted view across. 
If A~A represents the former line of junction of the clay with 
the sands and gravels, the latter must have been removed to the 
extent of the difference between A and B, the present outcrop 
in each valley. In valley No. 1 the clay has subsided 18 feet, 
and in valley No. 2, 30 feet. The total length of the section is 
only 480 yards. 
It is characteristic of subterranean erosion that its action is 
lateral; it is greatest at the point of escape and least at the 
farthest distance from it, while the interval between these two 
extremes becomes an inclined plane of subsidence. I believe 
this fact accounts for the occurrence of the present outcrops of 
the sands and gravels in the section at B—B instead of at Z-Z, 
the difference being the measure of the subterranean erosion of 
the Drift. 
Subterranean erosion is frequently intermittent, and especially 
so if an impervious stratum rests upon a pervious one. ‘Thus, the 
accompanying diagram (fig. 2) illustrates a very familiar example. 
Fig. 2. 
X. Clay. Y. Sand and Gravel. Z. Rock. 
The overlying clay, by the removal of the sands and gravels 
between A and B, has brought down the clay X upon the rock Z, 
effectually stopping any further subterranean erosion until the 
barrier of clay between A and B has been removed by subaérial 
denudation. 
The sands and gravels beneath the clay are not eroded ina 
uniform manner, the finer sands and gravels being most easily 
removed. ‘The stratification of these beds is most variable, and 
cuirent-bedding very frequent, this giving rise to constantly 
changing deposits of light sands intermingled with beds of 
coarse sands and gravels on the same horizon. In sandpits it 
