On leaving the Ishtar Gate we cross the substructure of the threshold, which rested on many layers of brick and must itself have been of stone. On the south of the gate some later insignificant buildings, perhaps Parthian, have clustered round it. These leave the entrance free, and Nebuchadnezzar’s great paving-blocks of the upper roadway, over which Nebuchadnezzar, Daniel, and Darius must frequently have passed, are still in position. Farther on only the lower pavement remains. It extends parallel with the east front of the Southern Citadel as far as the end of the mound, where it surrounds an altar (?) of mud brick. A branch of the street leads to the principal entrance of the Southern Citadel. A great number of limestone and turminabanda paving-stones found in the southern portion originally formed part of the destroyed upper pavement. It appears that during the Greek or Parthian periods balls for projectiles were made out of this limestone, as many have been found here. They divide into groups of various weights (Fig. 34). Some measure 27.5 centimetres in diameter, and weigh 20.20–20.25 kilos; others 19 centimetres, and 7–7.75 kilos; and others again 16 centimetres, and 4–4.5 kilos. [Illustration: FIG. 34.—Limestone projectiles. ] South of the Citadel the street crosses a watercourse, which apparently varied at different periods both in width and in name. In the time of Nebuchadnezzar it was perhaps the canal “Libil-ḫigalla,” while in Persian and Greek times it was the Euphrates itself that flowed here. We dug a ditch here that extended from the mound to the recommencement of the street, and which clearly showed the stratum to have been formed by the deposit of water. The strata contain no ruins with the exception of a canal, which in places is barely 3 metres broad. This canal is constructed in later fashion with the ancient bricks of Nebuchadnezzar, the best outside, the fragments inside, and all laid in mud. To the east it soon comes to an end and disappears in the banked-up watercourse. To the west it first widens out into a basin of three times its breadth, where narrow steps lead down the embankments to the level of the water (Fig. 35), and then once more narrows to its ordinary width. Farther to the west we know nothing of it. At the narrow portions, at about the height of the ancient water-level, courses of squared limestone of considerable size were laid. In the western part the northern bank contained a square opening many brick courses deep. The whole conveys the impression of a kind of sluice, which perhaps served to connect a watercourse in the east, of high water-level, with another in the west of lower level. This construction may date from the time of Neriglissar, when throwing a bridge across the canal to carry the Procession Street presented no difficulty. In earlier times the street appears to have been carried on a dam with walled embankments, which latter still exist below the walls of the canal. [Illustration: