Closely allied to the process of grinding is that of sawing stone. It is however rarely, if ever, that in this country any of the stone implements show signs of having been reduced into shape by this process. Among the small hatchets in fibrolite, so common in the Auvergne and in the south of France, and among the greenstone, and especially the nephrite celts found in the |44| Swiss Pfahlbauten,[135] many show evident traces of having been partially fashioned by means of sawing. I have also remarked it on a specimen from Portugal, and on many fibrolite hatchets from Spain.[136] Dr. Keller has noticed the process, and suggests that the incisions on the flat surface of the stone chosen for the purpose of being converted into a celt were made sometimes on one side, and sometimes on both, by means of a sharp saw-like tool. He has since[137] gone more deeply into the question, and has suggested that the stone to be sawn was placed on the ground near a tree, and then sawn by means of a splinter of flint fixed in the end of a staff, which at its other end was forked, and as it were hinged under one of the boughs of the tree sufficiently flexible to give pressure to the flint when a weight was suspended from it. The staff was, he supposed, to have been grasped in the hand, and moved backwards and forwards while water was applied to the flint to facilitate the sawing. The objection to this suggestion is, that in case of the flint being brought to the edge of the stone it would be liable to be driven into the ground by the weight on the bough, and thus constantly hinder the operation; nevertheless some such mechanical aids in sawing may have been in use.