doors are found; two column-shafts consisting in palm-trunks, sunk into the ground and surrounded at the foot, by a circular brick walling strengthened with asphalt and lime, were however actually found in one of the courts, but Koldewey assigns the restored building of which they form a part to the Persian period. In the Amran[77] mound at Babylon Koldewey discovered the truncated remains of twenty-two brick columns, which evidently formed part of a columned building, but the date of this building seems to be uncertain. It is here however that the bas-reliefs come to our aid; in Pl. XIV we have a reproduction of the famous Sun-god Tablet which was made by Nabû-aplu-iddina, king of Babylonia in the first half of the ninth century B.C., in which there is a shrine, the roof of which is supported by a column in the form of a palm-trunk which was probably overlaid with plates of metal, for plain unadorned wood would hardly be suitable for the shrine of Shamash, and moreover the capital and base, both of which are much the same, could only have assumed this form in metal, the one material that would easily adapt itself to such motifs. Similarly the curved back wall and roof were probably made of metal, for wood of the kind procurable in Babylonia would not readily bend in this manner. But this notwithstanding, the column always appears to have occupied a subordinate position in Babylonian architecture. Such also appears to have been the case in Assyria: there too the excavations have done little in the way of recovering the actual columns used by the Assyrian monarchs, and for our knowledge of the general form and appearance of Assyrian columns we are in the main dependent on the information afforded us by the wall bas-reliefs. Another source of great fruitfulness would be the series of ivories found in the north-west palace of Nimrûd (Calah), but as these are the work of either Egyptian or Phœnician artists, the columns therein represented can hardly be regarded as illustrations of Assyrian columns. [Illustration: FIG. 13, _a._—Capital of large Column. (Place, _Nineve_, Pl. 35.)] [Illustration: FIG. 13, _b._—Capital of small Column. (Brit. Mus.)] Of remains of actual columns, the best-preserved is probably that discovered by Victor Place at Khorsabad; it comprises the capital and a portion of the shaft (cf. Fig. 13, _a_) both in one piece; it is made of limestone, and the surviving fragment is some forty inches high. The decoration of the capital proper is a variety of the volute, a device which probably originated in a more or less accurate imitation of the horns of the goat, and which is a characteristic feature of Babylonian and Assyrian decoration. [Illustration: FIG. 14.—A, cf. Layard _Discoveries_, p. 590. B, cf. Layard, _Mon, Ser._ I, Pl. 95. C, cf. Botta, _Ruines de Ninive_, II, p. 114. D, E, Bas-reliefs from Kouyunjik.]