The benches for the oarsmen were always perpendicular to the longitudinal axis of the boat, this being necessary in view of the special construction of the vessels. Under the Middle Empire, small extra elevations supplied with a hand rail were erected on the forecastle and poop decks. They were, respectively, the posts of the captain and helmsman. [Sidenote: II 20] [Sidenote: II 20] The mast, erected near the middle of the ship, was movable on all vessels. The double mast under the Old Empire rested in two shoes set on either side of the longitudinal axis. The single mast (under the Middle and New Empires) went down to the bottom of the hold and rested against the beams which held up the rowers’ benches; it was also fixed by ropes (this appears clearly on several reliefs), either directly or by means of a step, as is shown on the model at Berlin. (See ~Belger~, pp. 27-29.) In this latter case, the mast was fastened to the step itself, a manner of consolidating which is still applied even now. In this respect, a relief coming from a mortuary chamber and now at the museum of Gizeh offers a great deal of interest. This relief shows a mast being let down, and Belger, the author already quoted, calls attention, in this connection, to the fact that the sculptor let the extreme part of the mast fall behind the coat of the man who is occupied with the task, probably because he did not know what to do with it. Then too only two of the five oarsmen of this style of boat appear in this reproduction, whereas, if the drawing were well made, it would follow therefrom that the brackets shown on sundry models abaft of the benches of the rowers, were only to serve as a support to the backs of these last. The great length of the boats and their relatively small immersed length, required special precautions against the hogging of the ship. A rope was stretched from bow to stern along the longitudinal axis of the vessel. This rope was supported on forks and was fastened bow and stern to a cable which ran around the boat. (Dr. ~Moritz Rühlmann~, p. 32.) [Sidenote: II 12]