^{page 360 note 1} The archaeological evidence given in the following paragraph would exclude the possibility of the change taking place at Ur later than at Tello and therefore of bringing the date of the graves down into the Sargonid period; the fact that at Tello the change comes in the middle of one king's reign and not e.g. at the beginning of a new dynasty would rather suggest that the improved type of brick was borrowed from elsewhere and not a spontaneous local invention.

^{page 361 note 1} Gadd, History and Monuments of Ur, p. 70. Sidney Smith, Early History of Assyria, p. 40, would put Ur-Nina later, c. 2700 B.C., and would make the downfall of the Second Dynasty of Ur come just about then at the hands of Lagash, whereas Gadd had rather favoured the First Dynasty of Ur as the victims of the Lagash conquest. Weidner and Christian, Archiv für Orientforschung, Bd. v, p. 141, put Entemena at about 2550 B.C. and therefore immediately before Sargon of Akkad (2528 B.C).

^{page 361 note 2} Of the forty-five types of clay vessels found in the shaft graves and in PG/1422 twenty-two are found elsewhere in the cemetery area only in Sargonid graves, eleven are common to the Sargonid and to the early cemetery, four occur in the later graves of the predynastic cemetery and are not found in Sargonid graves, and eight are peculiar to the shaft graves and can therefore be called specifically ‘Second Dynasty’ types. These figures result from the study of the 1,850 graves in the cemetery area; an enlarged basis of study might modify the figures slightly but would not, I imagine, seriously change them.

^{page 361 note 3} In Archiv für Orientforschung, Bd. vii, p. 108.

^{page 362 note 1} In the old cemetery, PG/1237 was a death-pit separated from its tombchamber; in PG/1051 the death-pit lay under the chamber.

^{page 362 note 2} For a section of this see Antiq. Journ. ix, pl. xxvi

^{page 364 note 1} For this cf. the Royal tomb shaft PG/1054; Antiq. Journ. ix, pl. xxvi.

^{page 368 note 1} The depth below the modern surface was 6.00 m.

^{page 372 note 1} The core of the second stage is thus constructed; the same is true of the Eanna Ziggurat of Ur-Engur at Warka, where the burnt-brick facing has disappeared and only the mud-brick core survives.

^{page 373 note 1} In the Neo-Babylonian work of the top stage we find plaster on the mud brickwork behind the burnt brick.

^{page 373 note 2} It was the failure to identify some of these that accounted for mistakes in our first attempt at restoration.

^{page 378 note 1} v. Jordan, Abhandlungen der preussischen Akademie der Wissenschaften, 1930, iv, p. 49.

^{page 383 note 1} I am aware that Dr. Jordan regards these vases as serving a constructional and not an ornamental purpose, but to me the intention of the original builders seems certainly to have been decoration.

^{page 388 note 1} For the most part the new work was superimposed on the old with at most a slight overlap; additions to the original design are shown in the plan in outline only.

^{page 389 note 1} The fort, which enclosed that of Warad-Sin, probably served as a gateway to the Ziggurat terrace and corresponded to E-dublal-mah on its other side.

^{page 390 note 1} The impression is considerably smaller than the original coin. Probably this is due to a clay mould having been made from the coin and a new positive cast from this: the present clay impression is from the cast. The repeated shrinkage of the clay in drying would account for the small size of our impression. In the Calene phialae decorated with central medallions taken from coins the shrinkage is approximately the same, owing to the die having been of clay as well as the vessel itself. For the casting of impressions from silver coins, cf. G. F. Hill in Hermes, vol. xxxvi, p. 317. Probably in this case a cast (in gold?) was taken from the coin to make the bezel of a finger-ring: this would account for the presence of the apparent coin-impression in a collection of impressions from gems.