284 DESTRUCTION OF THE GEEEK EMPIEE machines was going on at the landward walls and while feints were being made which kept the defenders always on the alert, to resist attacks or effect repairs, a portion of their forces had to be told off to defend the north-western walls facing the Golden Horn. Many attempts were made from these walls on the Horn, and from the Christian ships to destroy the Turkish vessels. Nearly every day as long as the siege lasted, some of the Greek or Venetian ships were told off to watch or attack them. Sometimes the Turks were chased to the shore : at other times the pursuers became the pursued. 1 Building To enable his troops to pass readily across the Golden overUpper Horn, Mahomet commenced and carried through with his Horn. usual energy the construction of a bridge over the upper part of it, near the place where the landward walls join those on the side of the Horn. This district was then known as Cynegion, and now as Aivan Serai. 2 The bridge was formed of upwards of a thousand wine barrels, all securely fastened together with ropes. Two of the barrels placed lengthways made the width of the bridge. Upon them beams were fixed, and over the beams a planking sufficiently wide to enable five soldiers to walk abreast with ease. 3 The object in placing the bridge so near the walls was, not merely to facilitate communications between the troops behind Pera and the army before the walls, but to attach to it pontoons upon which cannon could be placed for attack- ing the harbour walls. The paucity of the number of the defenders greatly alarmed 1 Crit. xliv. 2 Dr. Mordtmann places the bridge between Cumberhana and Defterdar Scala. 3 Ducas gives the above dimensions. Assuming the width from centre of each barrel, including a space between them, to be four feet, this would give the length of the bridge as 2,000 feet, which is about the width of the Horn at the place mentioned. Phrantzes gives its length at a hundred fathoms and the breadth fifty fathoms. These dimensions are clearly wrong if applied to the bridge, since the length falls far short of the width of the gulf. Leonard says it was thirty stadia long. Here, as elsewhere, I suspect that he uses stadium for some measure about one ninth of a furlong in length. If this conjecture is right, his estimate of the length of the bridge is about 2,000 feet.
