Finally, all our doubts are removed by a bas-relief from the palace of Assurbanipal, which is now in the British Museum (Fig. 86). The upper part of this carved picture is destroyed, but enough remains to show that it reproduced the facade of some richly decorated building. Four columns supported on the backs of so many lions, and two flat pilasters upheld in the same fashion by winged griffins, may readily be distinguished. That these griffins are not repeated on the left of the relief, is due perhaps to the haste or laziness of the sculptor. He may have thought he had done enough when he had shown once for all how these pedestals were composed. However this may have been, the lions in this relief play exactly the same _role_ as that attributed by us to the little model found by George Smith, and to the winged sphinx discovered by Sir Henry Layard before one of the doors at Nimroud. A base in the form of a vase or cushion is inserted between the back of the animal and the bottom of the shaft. In the pilaster--if we may believe that the artist took no liberties with fact--the junction is direct without the interposition of any ornamental motive. [Illustration: FIG. 85.--Winged Sphinx carrying the base of a column; from Layard.] [Illustration: FIG. 86.--Facade of an Assyrian building; from a bas-relief in the British Museum. Height 10 inches.] In what M. Place calls the state doorways (_portes ornees_) of Khorsabad, the arches spring from the backs of the great mitred bulls that guard the entrance.[270] But, whether the columns rose from the backs of animals real or fantastic, they always seem to have had a base. Almost the only instance of its absence is in the open gallery in Fig. 76, and there, perhaps, they are hidden by a balustrade. Everywhere else we find a more or less ornamental member interposed between the shaft and the ground. At Khorsabad (Fig. 41) it is a simple torus (Fig. 87), at Kouyundjik (Fig. 42) it is a kind of cushion (Fig. 88), which we find represented in not a few of the bas-reliefs. The curves bear a distant resemblance to the volutes of a capital; above this base appears a ring or astragal, the origin of which may be easily guessed. The original timber column, the newly felled tree that was set up to support the roof of a tent or a house, must have been placed upon a block of stone or wood, to which it was joined, in some degree, by hollowing out the latter and setting the foot of the timber beam in the hollow, and then hiding the junction by those reed bands that, as travellers tell us, were still used for the same purpose in the last years of Babylon.[271] In time a ring of metal would take the place of the reeds, and when stone columns came to be used, a feature which was at first a necessity, or, at least, a useful expedient and a guarantee of duration and solidity, came at last to be simply an ornament. [Illustration: FIGS. 87, 88.--Bases of columns; from the bas-reliefs.]