flakes to the more highly-wrought implements appears also to be far greater in the caves than that in the gravels. This apparent greater abundance may, however, to some extent be due to the flakes in the gravel escaping the notice of the workmen, or to their having been broken to pieces during the formation of the gravel.
3. Flat flakes are more common, but these are usually shorter, thicker, and broader than those of the Surface Period. They frequently exhibit that minute chipping at the edge, which is probably the result of wear from scraping some hard substance, such as bone or even wood. Occasionally a notch has been worn in the edge of the flake, as if the object scraped had been cylindrical.
4. Polygonal flakes are those most abundant in the River-drift; but the large, broad flakes of this character, such as are common in the valley of the Somme, and especially in its lower deposits, as at Montiers, near Amiens, are much rarer in England. Fig. 461, from Reculver, is a flake of this character, but I am not sure whether it does not, more properly speaking, come under the head of a wrought flake, as it appears to have been somewhat trimmed at the edges. It is worth while remarking, that many of the French specimens have the edge worn away by use, just on one side of the bulb of percussion, at a place where there is generally a clean sharp edge in a newly-made flake of this form. Occasionally similar marks of use are apparent on English specimens of the same character.
Taken as a whole, the simple flakes of the River-drift Period may be described as larger, coarser, thicker, and broader than those of the Surface Period, or of caves of later date than Le Moustier. Their use appears to have been for cutting and scraping whatever required to be cut or scraped.
I formerly regarded some of them as having possibly been arrow-heads, but the extreme rarity of any light, sharp-pointed flakes, and the absence of any evidence that those who fashioned them were acquainted with the use of the bow, render this assumption almost untenable. It is, however, barely possible that some may have served to tip spears or lances.
TRIMMED FLAKES.
One of the commonest forms into which flint flakes were fashioned in Neolithic times, is that produced by trimming the end of the flake to a semicircular bevelled edge. To this form the
