^{page 3 note 1} In the level I building this is destroyed, but it can safely be reconstructed by analogy with the building of level II of which level I is largely a reproduction; in level II the street door is a very marked feature.

^{page 4 note 1} The statement is broadly but not literally true. We found one sherd of fifth-century Attic pottery, the foot of a late Roman glass goblet, a piece of Arab glazed ware, c. thirteenth century, several Arab tobacco-pipes and a nineteenth-century Turkish coin. Such ‘exceptions’ only emphasize the general truth.

^{page 8 note 1} The narrow cupboard-like room 5 of house C was full of baked clay sling-bolts and ballista-balls; the fact of the house having been something of an armoury would account for the hard fighting in its neighbourhood.

^{page 8 note 2} Layard, , Monuments of Nineveh, 96. 10Google Scholar. A somewhat similar dagger has been found at Ras Shamra.

^{page 8 note 3} With the evidence of Carchemish, Salcje-geuzi, Tal Ahmar, and Hama we are here on fairly safe ground.

^{page 9 note 1} Speiser, E. A., ‘The Pottery of Tell Billa’, Museum Journal (Philadelphia), vol. xxiii, 3, pl. lxi.Google Scholar

^{page 10 note 1} Iraq, III, part i, fig. 27, no. 20, for one Chagar Bazar fragment; for other information regarding the Habur finds I am indebted to Mr. M. E. L. Mallowan.

^{page 10 note 2} McEwan, , Syrian Expedition, back cover.Google Scholar

^{page 10 note 3} That is, immediately below level III; of course there are remains of older buildings in deeper levels to which we have not yet dug down.

^{page 12 note 1} Beneath the floor of rooms 6 and 7 there came to light the NW. wall of a room belonging to an earlier building whose floor lay about 0·50 m. lower down.

^{page 13 note 1} Sometimes instead of the hand the under side of the cup is decorated with an elaborate palmette design; Syria, vol. xi, pl. xxiv.

^{page 13 note 2} Cf. British Museum Catalogue A 32, A 51.

^{page 17 note 1} At the point chosen for our trench this rubbish took the form of pure ashes. It is not to be supposed that this would be the case round the whole circuit, and we were fortunate in hitting upon a spot where the uniform character of the later filling made the distinction between the two works so very obvious.

^{page 17 note 2} This would be quite effective and could easily be renewed when, as would necessarily occur, it gradually got washed down the slope. Actually it cannot be traced on the top third of the embankment, then becomes clear, and thickens as it goes downhill; see the section on pl. XII.

^{page 18 note 1} The arrangement of the double wall at Atchana might be compared with that at Sinjirli.

^{page 18 note 2} The revetment foundations went down to a surface of beaten clay and sand 0·80 m. below the cobble floor; the house foundations were 0·35 m. higher than this, too low to be strictly contemporary with the cobble floor and probably corresponding to an intermediate surface. The foundations of the town wall revetment would naturally be laid deeper than those of a private house, and it is also to be remembered that the house walls rested on the stumps of the walls of the earlier building on the same site, and their depth would depend on the height to which those stumps were standing.

^{page 19 note 1} A single cross-section cannot be held to have settled the whole question; a further study of the wall at other points is necessary, and any conclusions reached at this stage are provisional only.

^{page 19 note 2} Cf. Carchemish, ii, pl. 27; the vases there belong to the first half of the second millennium B.C.; cf. that on pl. XVI, 2, b.

^{page 19 note 3} McEwan, C. W., Syrian Expedition of the Oriental Institute.Google Scholar

^{page 20 note 1} I was at first in some doubt as to the date of the pit in sq. s. 10, but it is certain that the longer drain did empty into it, though the connexion is now lost. The drains are not cut through by the stone wall, as might appear from the ground-plan, but lie at a higher level.

^{page 21 note 1} There seems to be the start of one against the SW. limit of the excavation; to the NE. work has not gone far enough to produce evidence on the point.

^{page 22 note 1} Perhaps from a room above the chamber to the east, perhaps from above the entrance-hall.

^{page 23 note 1} There is no mark of a column-base on the surviving threshold and it is therefore impossible to restore two columns; moreover, enough of the wall remains at the north corner to show that the opening was not central.

^{page 23 note 2} The centre of the stone where the column rested is smooth but unpolished; the edges which showed beyond the shaft are polished.

^{page 25 note 1} See MacEwan, , The Syrian Expedition of the Oriental Institute.Google Scholar

^{page 25 note 2} Carchemish, ii, p. 149.

^{page 25 note 3} On the face of the mud-brick which projects so as to be flush with the face of the horizontal beams there is no sign of any plaster, nor is the brickwork even pointed; the conclusion to be drawn from this is that the brickwork was not visible.

^{page 27 note 1} The technique is interesting. In the sheet gold the pattern was defined by sunken channels in the base of which were impressed hollows, one for each gold pellet: the latter were fixed by a gold solder sprinkled over them and heated from above, presumably by means of a blow-lamp. The solder has joined the pellets together but has not always sunk low enough to secure them to the base.

^{page 27 note 2} To be published by Mr. Sidney Smith in Iraq.

^{page 27 note 3} The date of them is uncertain; they were found in a late burial, possibly of the eighth century B.C., but their close resemblance to the Yasilikaya stone reliefs inclines one to the belief that they were heirlooms at the time when they were buried. With this bead may be compared the cloisonné dagger-hilt from Shaft Grave IV, Mycenae, and a staff-head now in the Cyprus Museum; v. Buxton, L. H. D., Casson, Stanley, and Myres, J. L. in Man, vol. xxxii, 1Google Scholar.

_{page 28 note 1} One would have expected ivory to be damaged by the fire which destroyed the building; it is indeed burnt to a deep brown colour, but is otherwise perfect and has not even lost its original polish.