space of the two long sides of the ship, with the exception of the two ends; they were arranged over one another in rows of different heights, not separated by partitions, but only by the open structure of wood. In each row each rower sat immediately in front of the next man in a straight line, but there is a difference of opinion as to the manner in which the rowing benches were arranged. According to Graser, they were immediately under one another, but the rowers did not sit perpendicularly above each other; but in order to save space as much as possible, and partly to facilitate their movements, they were arranged in such a way that the seat of the next highest was in the same direction and height as the head of the man on the next seat below, so that each man, instead of sitting directly under the man above, sat a little towards the back, and, in moving, kept his arms immediately under the seat of the man above. Lemaître, on the other hand, assumes that only the lowest benches were close to the edge, and those [Illustration: FIG. 195.] above were removed by the breadth of the thwart, the third by two breadths, in which case the height must have been so arranged that the oar of the man above always passed over the head of the one immediately below. It is impossible to attain any certainty about this matter; both hypotheses are open to objection. For the length of the oars naturally increased in proportion to the distance of the rowers from the water, and those of the highest row must have been longest; according to Graser’s arrangement, the length of the oars increased 1 yard for each row, so that in a ship of five rows the lowest rank had oars 2½ yards long, the highest 6½; according to the arrangement of Lemaître, the length was even greater, but there was this advantage, that the longer oars had also longer leverage, and could consequently be more easily controlled. The larger the number of rows, the greater in consequence was the length of the oars, but still they were able to build and control ships of fifteen or sixteen rows. The splendid ship of Ptolemy Philopater is said to have had no less than forty rows, and the length of the highest oars was 18½ yards; but this was not a ship of war, and was only used in calm water--in fact, a modern authority on seafaring regards the whole description of this forty-decker as a satire. Of course, the larger the ships the greater the number of oarsmen required, since the number of rows would be greater; a “trireme” was rowed by 174 men, a “quinquereme” by 310, the arrangement being that each higher row had two men more than the one below, because the bulk of the ship was broadened towards the top. In rowing the greatest regularity of movement was indispensable; this was attained by the command of a special captain, and also by marking time with flutes, so that all the oars might strike the water at the same moment. Here we meet with a problem, hitherto unsolved: how was it possible for the long