In describing the trireme it will be convenient to deal first with the disposition of the rowers and subsequently with the construction of the vessel itself. The object of arranging the oars in banks was to economise horizontal space and to obtain an increase in the number of oars without having to lengthen the vessel. We know from Vitruvius that the 'interscalmium,' or space horizontally measured from oar to oar, was two cubits. This is exactly borne out by the proportions of an Attic aphract trireme, as shown on a fragment of a bas-relief found in the Acropolis. The rowers in all classes of banked vessels sat in the same vertical plane, the seats ascending in a line obliquely towards the stern of the vessel. Thus in a trireme the thranite, or oarsman of the highest bank, was nearest the stern of the set of three to which he belonged. Next behind him and somewhat below him sat his zygite, or oarsman of the second bank; and next below and behind the zygite sat the thalamite, or oarsman of the lowest bank. The vertical distance between these seats was 2 feet, the horizontal distance about 1 foot. The horizontal distance, it is well to repeat, between each seat in the same bank was 3 feet (the seat itself about 9 inches broad). Each man had a resting-place for his feet, somewhat wide apart, fixed to the bench of the man on the row next below and in front of him. In rowing, the upper hand, as is shown in most of the representations which remain, was held with the palm turned inwards towards the body. This is accounted for by the angle at which the oar was worked. The lowest rank used the shortest oars, and the difference of the length of the oars on board was caused by the curvature of the ship's side. Thus, looked at from within, the rowers amidship seemed to be using the longest oars, but outside the vessel, as we are expressly told, all the oar-blades of the same bank took the water in the same longitudinal line. The lowest or thalamite oar-ports were 3 feet, the zygite 4-1/4 feet, the thranite 5-1/2 feet above the water. Each oar-port was protected by an _ascoma_ or leather bag, which fitted over the oar, closing the aperture against the wash of the sea without impeding the action of the oar. The oar was tied by a thong, against which it was probably rowed, which itself was attached to a thowl ([skalmhos]). The port-hole was probably oval in shape (the Egyptian and Assyrian pictures show an oblong). We know that it was large enough for a man's head to be thrust through it. [Illustration: ANCIENT GALLEY FIGHT, FROM POMPEII.]