dinavia,' when speaking of the antiquities referable to the Palæolithic age, says "that they are usually found in beds of gravel and loess, extending along our valleys, and reaching to a height of 200 feet above the present water-level;" and he adds, "that these beds were deposited by the existing rivers, which then ran in the same direction as at present, and drained the same areas."
As long as it was believed that the implement-bearing gravels were never found except on or very near to the banks of rivers, it was reasonable to attribute to those rivers the transport of the gravel in which they were imbedded; but from more recent observations, both in England and France, it seems evident that the implement-bearing gravels, as well as others of the same character, which are not yet known to contain implements, do occur in localities so far removed from existing rivers, and, when found near rivers, at such elevations as almost to preclude the belief that those rivers, however swollen by excessive rainfall or melting snow, or otherwise, could have at all affected their condition.
Thus, for instance, as we see at Brandon, the implement-deposit occurs at an elevation of from 80 to 90 feet above that at Broomhill (which is two miles higher up the stream), and about the same above Shrub Hill, which is several miles lower down. Yet, notwithstanding this great difference in the levels, we have strong, if not unmistakeable indications, derived in some measure from the implements themselves, that all these deposits were (geologically) contemporaneous. The implements in each are substantially of the same age and character; the matrix of red gravel, in which they rest, is of the same composition; the beds rest directly upon the eroded surface of the chalk or the gault, and are more or less overlain by sands of the same description. But it is incredible that such deposits (if of the same age), should owe their origin to one and the same river; for if so, in order to reach the higher level, it must have been swollen to the height of 100 feet above the level at Broomhill and Shrub, and extended three miles to the south. This would require a volume of water of dimensions and power several thousand fold greater than those of the present river; and to supply such a stream, the basin from which the river is fed (occupying as it does an area of not more than 300 square miles) is altogether insufficient; nor, indeed, would the present contour of the country allow such a river to flow in that direction.
Nor would the difficulty be removed or lessened if we were to disregard those proofs of contemporaneity to which I have adverted, and to assume that these deposits were separated from each other by an interval of time sufficient to allow of the excavation of the existing river-valley to the depth of the 80 or 90 feet which now divide it from the Brandon bed. We have no reason whatever to believe that, in these districts, the relative levels of the surface have undergone any material change during the Quaternary period[1]; and, if not, in order to account for the Brandon and Lakenheath deposits, we must suppose that this river once flowed at
- ↑ Phil. Trans. 1864, pp. 286–290.
