plethra.[699] That the number of towers was so small in comparison with the circuit of the wall is no reason for wonder--so Diodorus or his authority adds--for the city was surrounded by a wide belt of marshes, and it did not appear to be necessary to build towers where the marshes afforded sufficient protection.[700] The accounts which have been preserved of the stay of Alexander of Macedon in Babylon also prove the existence of two royal citadels, one on each bank of the Euphrates. In the last days of his life Alexander lived in the king's palace, from which the house of Bagoas, with whom on one occasion he banqueted, was distant ten stades.[701] From the banquet-hall in this palace, where he had given his commands to his generals, and rested till the dusk of the evening, he was carried in a litter to the river, and conducted on board ship to the park on the other side of the river, where he bathed and rested. After spending three days there in his chamber--on the first day he played at dice with Medius; on the second he listened to the account of Nearchus about the voyage from the Indus through the great sea; on the third he bade his generals enter to receive instructions for setting out in three days--he caused himself to be brought into the house near "the large bath;" he gave orders for the generals to keep watch in the portico, and the Chiliarchs and the Pentacosiarchs before the doors. When more seriously ill he was conveyed from the garden into the more distant royal palace, where the generals entered, and the soldiers forced their way into his presence.[702] Leaving out of sight what may have remained, and did remain uninjured, of the outer walls and towers of the city, when Herodotus and Ctesias were in Babylon, and when the Macedonians of Alexander saw the city, it is clear that in the fifth and fourth century B.C. so much remained standing that the line of the trenches and the wall could be clearly traced. If the circuit of 365 stades, given by Cleitarchus, is clearly a fiction derived from the number of days in the Babylonian year, we shall still be able to give the preference to the 360 stades of Ctesias over the 480 stades of Herodotus, though Aristotle remarks, "Babylon reached the extent of a nation, not of a city."[703] Since, as Ctesias also tells us, the two walls on the Euphrates were nearly 160 stades in length, the wall on each bank would be nearly 80 stades in length, _i.e._ about 10 miles. Supposing that the Euphrates passed diametrically through the city, which was not the case, the city wall, if we also suppose that the city, as Herodotus says, was an exact square, would at the utmost have a circuit of 320 stades, _i.e._ of about 40 miles. We cannot therefore avoid the conclusion that in the 480 stades of Herodotus the 160 stades of the two walls on the river were included; if he inquired about the total length of the city walls the answer may very well have included the walls by the river.