If we look at the matter from another point of view we shall meet with equally absurd results. The oars in the upper banks of Athenian triremes are known to have been about 14 ft. in length. Underneath them, were, of course, two other banks. If, now, we assume that the upper bank tholes were 5 ft. 6 in.[10] above the water-line, and that one-quarter of the length of the upper bank oars was inboard, and if we add thirty-seven additional banks parallel to the first bank, so as to make forty in all, simple proportion will show us that the outboard portion of the oars of the uppermost bank must have been just under 99 ft. long and the total length of each, if we assume, as before, that one quarter of it was inboard, would be 132 ft., instead of the 57 ft. given by Callixenos. Any variations in the above assumptions, consistent with possibilities, would only have the effect of bringing the oars out still longer. We are therefore driven to conclude, either that the account given by Callixenos was grossly inaccurate, or else that the Greek word, τεσσαρακοντἠρης, which we translate by "forty-banked ship," did not imply that there were forty horizontal super-imposed tiers of oars.
The exact arrangement of the oars in the larger classes of galleys has always been a puzzle, and has formed the subject of much controversy amongst modern writers on naval architecture. The vessels were distinguished, according to the numbers of the banks of oars, as uniremes, biremes, triremes, quadriremes, etc., up to ships like the great galley of Ptolemy Philopater, which was said to have had forty banks. Now, the difficulty is to know what is meant by a bank of oars. It was formerly assumed that the term referred to the horizontal tiers of oars placed one above the other; but it can easily be proved, by attempting to draw the galleys with the oars and rowers in place, that it would be very difficult to accommodate as many as five horizontal banks and absolutely impossible to find room for more than seven. Not only would the space within the hull of the ship be totally insufficient for the rowers, but the length of the upper tiers of oars would be so great that they would be unmanageable, and that of the lower tiers so small that they would be inefficient. The details given by ancient writers throw very little light upon this difficult subject. Some authors have stated that there was only one man to each oar, and we now know that this was the case with the smaller classes of vessels, say, up to those provided with three, or four, to five banks of oars; but it is extremely improbable that the oars of the larger classes could have been so worked. The oars of modern Venetian galleys were each manned by five rowers. It is impossible in this work to examine closely into all the rival theories as to what constituted a bank of oars. It seems improbable, for reasons before stated, that any vessel could have had more than five horizontal tiers. It is certain also that, in order to find room for the rowers to work above each other in these tiers, the oar-ports must have been placed, not vertically above each other, but in oblique rows, as represented in Fig. [14]. It is considered by Mr. W. S. Lindsay, in his "History of Merchant Shipping and Ancient Commerce," that each of the oblique rows of oars, thus arranged, may have formed the tier referred to in the designation of the class of the vessel, for vessels larger than quinqueremes. If this were so, there would then be no difficulty in conceiving the possibility of constructing galleys with even as many as forty tiers of oars like the huge alleged galley of Ptolemy Philopater. Fig. [15] represents the disposition of the oar-ports according to this theory for an octoreme.
Fig. 14.—Probable arrangement of oar-ports in ancient galleys.
Fig. 15.—Suggested arrangement of oar-ports in an octoreme.
It appears to be certain that the oars were not very advantageously arranged, or proportioned, in the old Greek galleys, or even in the Roman galleys, till the time of the early Cæsars, for we read that the average speed of the Athenian triremes was 200 stadia in the day. If the stadium were equal in length to a furlong, and the working day supposed to be limited to ten hours, this would correspond to a speed of only two and a half miles an hour. The lengths of the oars in the Athenian triremes have been already given (p. [42]); even those of the upper banks were extremely short—only, in fact, about a foot longer than those used in modern 8-oared racing boats. On account of their shortness and the height above the water at which they were worked, the angle which the oars made with the water was very steep and consequently disadvantageous. In the case of the Athenian triremes, this angle must have been about 23.5°. This statement is confirmed by all the paintings and sculptures which have come down to us. It is proved equally by the painting of an Athenian bireme of 500 b.c. shown in Fig. [9], and by the Roman trireme, founded on the sculptures of Trajan's Column of about 110 a.d., shown in Fig. [16].[11] In fact, it is evident that the ancients, before the time of the introduction of the Liburnian galley, did not understand the art of rowing as we do to-day. The celebrated Liburnian galleys, which were first used by the Romans, for war purposes, at the battle of Actium under Augustus Cæsar, were said to have had a speed of four times that of the old triremes. The modern galleys used in the Mediterranean in the seventeenth century are said to have occasionally made the passage from Naples to Palermo in seventeen hours. This is equivalent to an average speed of between 11 and 12 miles per hour.