give it a polygonal outline, the projections of which will serve for the central ridges or back-bones of the first series of regular flakes that he strikes off. The removal of this first series of flakes leaves a number of projecting ridges, which serve as guides for the formation of a second series of flakes, and so on until the block is used up.
But where a flake 10 or 12 inches in length is required, a different process becomes necessary. For it is nearly impossible with a rough mass of flint, to produce by single blows plane surfaces 10 or 12 inches in length, and arranged at such an angle as to produce a straight ridge, such as would serve to form the back-bone, as it were, of a long flake; and without such a back-bone, the production of a long flake is impossible. It is indeed this ridge (which need not, of course, be angular, but may be more or less rounded or polygonal) that regulates the course of the fissure by which the flake is dislodged from the matrix or parent flint; there being a slight degree of elasticity in the stone, which enables a fissure once properly commenced in a homogeneous flint to proceed at right angles to the line of least resistance in the dislodged flake, while at the same time exerting a nearly uniform strain, so that the inner surface of the flake becomes nearly parallel to the outer ridge. It was to obtain this outer ridge that the Pressigny cores were chipped into the form in which we find them; and it appears as if the workmen who fashioned them adopted the readiest means of obtaining the desired result of producing along the block of flint a central ridge whenever it became necessary, until the block was so much reduced in size as to be no longer serviceable. For, the process of chipping the block into the boat-like form could be repeated from time to time, until it became too small for further use. The same process of cross-chipping was practised in Scandinavia in early times, and the obsidian cores from the Greek island of Melos, Crete, and other ancient Greek sites prove that it was also known there. The blocks are found in various stages, rarely with the central ridge still left on, as Fig. 3, and more commonly with one or more long flakes removed from them, like Figs. 4 and 5. The sections of each block are shown beneath them. Two of the flakes are represented in Figs. 6 and 7. All the figures are on the scale of one-half linear measure.
The causes why the nuclei were rejected as useless are still susceptible of being traced. In some cases they had become so thin that they would not bear re-shaping; in others a want of unifor-
