What may be meant by _spina piscis_ it is perhaps difficult to conjecture; but I can by no means make _spina dorsi_ of it, as Dion. Salvagnius has done, in his observations on the passage quoted from Ovid’s Ibis. The small bony processes which project from the spine of a fish have some similitude to a saw; but it would be hardly possible to saw through with them small pieces of wood. These bones are too long, as well as too far distant from each other; and the joints of the back-bone are liable to be dislocated by the smallest force. I am not acquainted with the spine of any fish which would be sufficiently strong for that purpose. The jaw-bone of a fish furnished with teeth would be more proper; but the words _spina in medio pisce_ prevent us from adopting that alteration. I should be inclined rather to explain this difficulty by the bone which projects from the snout of the saw-fish, called by the Romans _serra_, and by the Greeks _pristis_. That bone, indeed, might not be altogether unfit for such a use: the teeth are strongly united to the broad bone in the middle, and are capable of resisting a great force; but they are placed at rather too great a distance. The old inhabitants of Madeira, however, we are told, really used this bone instead of a saw[653]. That Talus found the jaw-bone of a snake with teeth like a saw is extremely probable, for there are many snakes which have teeth of that kind. The saws of the Grecian carpenters had the same form, and were made in the like ingenious manner as ours are at present. This is fully shown by a painting still preserved among the antiquities of Herculaneum[654]. Two genii are represented at the end of a bench, which consists of a long table that rests upon two four-footed stools. The piece of wood which is to be sawn through is secured by cramps. The saw with which the genii are at work has a perfect resemblance to our frame-saw. It consists of a square frame, having in the middle a blade, the teeth of which stand perpendicular to the plane of the frame. The piece of wood which is to be sawn extends beyond the end of the bench, and one of the workmen appears standing and the other sitting on the ground. The arms, in which the blade is fastened, have the same form as that given to them at present. In the bench are seen holes, in which the cramps that hold the timber are stuck. They are shaped like the figure seven; and the ends of them reach below the boards that form the top of it. The French call a cramp of this kind _un valet_[655].