These valleys are in all probability ancient sea-bays, and I have sometimes speculated whether this sudden steepening of the sides does not mark the edges of vertical cliffs formed when these valleys were filled with sea-water, as would naturally happen in strata such as the chalk.
In most countries the roads and footpaths ascend along the bottoms of valleys, but here this is scarcely ever the case.

All the villages and most of the ancient houses are on the platforms or narrow strips of flat land between the parallel valleys.

Is this owing to the summits having existed from the most ancient times as open downs and the valleys having been filled up with brushwood?
I have no evidence of this, but it is certain that most of the farmhouses on the flat land are very ancient.
There is one peculiarity which would help to determine the footpaths to run along the summits instead of the bottom of the valleys, in that these latter in the middle are generally covered, even far more thickly than the general surface, with broken flints.

This bed of flints, which gradually thins away on each side, can be seen from a long distance in a newly ploughed or fallow field as a whitish band.

Every stone which ever rolls after heavy rain or from the kick of an animal, ever so little, all tend to the bottom of the valleys; but whether this is sufficient to account for their number I have sometimes doubted, and have been inclined to apply to the case Lyell's theory of solution by rain-water, etc., etc.
The flat summit-land is covered with a bed of stiff red clay, from a few feet in thickness to as much, I believe, as twenty feet: this [bed], though lying immediately on the chalk, and abounding with great, irregularly shaped, unrolled flints, often with the colour and appearance of huge bones, which were originally embedded in the chalk, contains not a particle of carbonate of lime.