The next peculiarity which I have to notice, is that of the tapering sides of the celt being curved inwards, as if for the purpose of being more securely fixed either to a handle or in a socket. In the last implement described, the reduction in width towards the middle of the blade would appear to have been intended to assist in fastening it at the end of a handle, as an adze cutting at each end. In Fig. 80 the reduction in width is more abrupt, and the blade would appear to have been mounted as an axe. It is formed of a compact light grey metamorphic rock, and was formerly in the collection of the Rev. S. Banks, of Cottenham, Cambridgeshire. I have a greenstone celt found at Carnac, Brittany, with shoulders of the same character about the middle of the blade. A form of celt expanding into a kind of knob at the butt-end is peculiar to the Lower Loire.[433] It is known as the “_hâche à bouton_,” or “_hâche à tête_.” [Illustration: Fig. 80.—Near Cottenham. 1∕2] [Illustration: Fig. 81.—Near Malton. 1∕2] The original of Fig. 81 was found in a gravel-pit near Malton, Yorkshire. It was at first supposed to have been found in undisturbed |136| drift, and some correspondence upon the subject appeared in the Times newspaper.[434] The gravel, however, in which it was found seems to belong to the series of Glacial deposits, and if so, is of considerably greater antiquity than any of the old River-gravels, in which the unpolished flint implements have been discovered. This celt is of greenstone, carefully polished at the edge, and towards the butt slightly roughened by being picked with a sharp pointed tool. This roughening is in character similar to that which has been observed on many of the celts from the Swiss Lake-dwellings and from France,[435] and was no doubt intended in their case to make the stone adhere more firmly in the socket of stag’s horn in which it was inserted. The object in this case would appear to be the same; and, like other polished celts, it belongs to the Neolithic Period. The expansion of the blade towards the edge is very remarkable. A celt of the same type as that from Malton, but somewhat oblique at the edge, and formed of quartz containing pyrites, found at Soden, is in the Museum at Bonn. A flat form of stone hatchet, expanding rapidly from a slightly tapering butt about half the entire length of the blade, so as to form a semicircular cutting-edge, has been found in South Carolina.[436] There is a small perforation in the centre, as if for a pin, to assist in securing it in its handle. Another form, with the blade reduced for about half its length, so as to form a sort of tang, is engraved by Squier and Davis.[437] [Illustration: Fig. 82.—Mennithorpe, Yorkshire. 1∕2]