That the refacing was not part of the original plan is shown by the fact that in some of the mud-brick towers the cavities of walled-up gutters still remain, such as we find in the city walls and the temples. The brick casings of the gutters were taken away, and in their place the brickwork of the facing was jointed in. With the exception of being widened the wall was little altered by the new building. The towers correspond in no way with those of the principal wall, at any rate it is only in the eastern portion that the same principle has been adopted, and a tower placed crossways is always succeeded by a smaller one placed lengthways. Here too, however, the western end consists of an especially large tower exactly in a straight line with that of the principal wall. The gateway in the west forms, in its position, its facings, and alterations a fairly exact counterpart of that in the principal wall. But, besides this, the mud wall had also four other gateways, of which only the one in the 5th mesopyrgion was retained in the rebuilding. A drain with two inlet shafts carried off the surface water from here with a sharp fall to the south, probably to the main conduit behind the principal wall. In front of the two wall heads at the west lay a building with the usual arrangement of a court and surrounding chambers. It was built over the ancient fosse wall, which by that time was destroyed, and might well represent the dwelling of the commandant of the walls. [Illustration: FIG. 92.—Space between the two mud walls. ] There were also two wall lengths of mud brick of a similar kind on the east of the Ishtar Gate. They are not long. The thicker one breaks off in the 2nd mesopyrgion, and is there supported by a later sloping embankment wall, which turns off in a south-easterly direction, where we have already followed it for 25 metres. The northern wall length is still shorter. The excavations, which at this point were carried considerably below the base of the mud wall, yielded mud and river sediment that apparently came from the Euphrates, which during the Persian period washed the eastern side of the Acropolis. In Nebuchadnezzar’s time these walls certainly extended farther east, and united themselves in some way, which is not yet entirely explained, with the inner city wall, which according to the inscriptions found there is to be recognised as the Nimitti-Bel of Sardanapalus. This is the more certain because the Ishtar Gate is also named in inscriptions as belonging to Imgur-Bel and Nimitti-Bel. Thus it is imperative that we should make a slight digression of 1000 metres to the east in order to observe this fortification. After that we will return to the Kasr. XXV THE INNER CITY WALL