raised platform, which consequently places the foundations of the temple nearly on a level with the roofs of the surrounding houses.  The raised platform of the temple of Nannar at Uru still measures 20 feet in height, and its four angles are orientated exactly to the four cardinal points.  Its facade was approached by an inclined plane, or by a flight of low steps, and the summit, which was surrounded by a low balustrade, was paved with enormous burnt bricks.  On this terrace, processions at solemn festivals would have ample space to perform their evolutions.  The lower story of the temple occupies a parallelogram of 198 feet in length by 173 feet in width, and rises about 27 feet in height.

* Perrot-Ohipiez admit that between the first and second story there was a sort of plinth seven feet in height which corresponded to the foundation platform below the first story.  It appears to me, as it did to Loftus, that the slope which now separates the two vertical masses of brickwork “is accidental, and owes its existence to the destruction of the upper portion of the second story.”  Taylor mentions only two stories, and evidently considers the slope in question to be a bank of rubbish.

** Perrot-Chipiez place the staircase leading from the ground-level to the terrace inside the building—­“an arrangement which would have the advantage of not interfering with the outline of this immense platform, and would not detract from the strength and solidity of its appearance;” Reber proposes a different combination.  At Uru, the whole staircase projects in front of the platform and “loads up to the edge of the basement of the second story,” then continues as an inclined plane from the edge of the first story to the terrace of the second, forming one single staircase, perhaps of the same width as this second story, leading from the base to the summit of the building.

[Illustration:  134.jpg THE TEMPLE OF NANNAR AT URU, APPROXIMATELY RESTORED.]

     Drawn by Faucher-Gudin.  The restoration differs from that
     proposed by Perrot-Chipiez.  I have made it by working out
     the description taken down on the spot by Taylor.

The central mass of crude brick has preserved its casing of red tiles, cemented with bitumen, almost intact up to the top; it is strengthened by buttresses—­nine on the longer and six on the shorter sides—­projecting about a foot, which relieve its rather bare surface.  The second story rises to the height of only 20 feet above, the first, and when intact could not have been more than 26 to 30 feet high.* Many bricks bearing the stamp of Dungi are found among the materials used in the latest restoration, which took place about the VIth century before our era; they have a smooth surface, are broken here and there by air-holes, and their very simplicity seems to bear witness to the fact that Nabonidos confined himself to the task of merely restoring things to the state in which the earlier kings of Uru had left them.**