Montfaucon[656] also has given the representation of two ancient saws taken from Gruter. One of them seems to be only the blade of a saw without any frame; but the other figure I consider as a cross-cut saw; and I think I can distinguish all the parts, though it is imperfectly delineated. One may however perceive both the handles between which the blade is fastened; the wooden bar that binds them together, though the blade is delineated too near it; and about the middle of this bar, the piece of wood that tightens the cord which keeps the handles as well as the whole instrument firm. Saws which were not placed in a frame, but fastened to a handle, are thus described by Palladius[657]:--“Serrulæ manubriatæ minores majoresque ad mensuram cubiti, quibus facile est, quod per serram fieri non potest, resecando trunco arboris, aut vitis interseri.” The most beneficial and ingenious improvement of this instrument was, without doubt, the invention of saw-mills, which are driven either by water, wind, [or by steam]. Mills of the first kind were erected so early as the fourth century, in Germany, on the small river Roer or Ruer[658]; for though Ausonius speaks properly of water-mills for cutting stone, and not timber, it cannot be doubted that these were invented later than mills for manufacturing deals, or that both kinds were erected at the same time. The art however of cutting marble with a saw is very old. Pliny[659] conjectures that it was invented in Caria; at least he knew no building incrusted with marble of greater antiquity than the palace of king Mausolus, at Halicarnassus. This edifice is celebrated by Vitruvius[660], for the beauty of its marble; and Pliny gives an account of the different kinds of sand used for cutting it; for it is the sand properly, says he, and not the saw, which produces that effect. The latter presses down the former, and rubs it against the marble; and the coarser the sand is, the longer will be the time required to polish the marble which has been cut by it. Stones of the soap-rock kind, which are indeed softer than marble, and which would require less force than wood, were sawn at that period[661]: but it appears that the far harder glassy kinds of stone were sawn then also; for we are told of the discovery of a building which was encrusted with cut agate, cornelian, lapis lazuli, and amethysts[662]. I have, however, found no account in any of the Greek or Roman writers of a mill for sawing wood; and as the writers of modern times speak of saw-mills as new and uncommon, it would seem that the oldest construction of them has been forgotten, or that some important improvement has made them appear entirely new.