78 History op Art in Antiquity. We are reduced to one hypothesis, but which has the merit of being highly probable: as at Nineveh and Babylon, the walls were constructed with crude bricks, laid out whilst still moist. Burnt bricks were reserved for the casing.* The information to be gathered from the state and arrangement of the preserved parts of the building confirm the above conjecture. All seems to have been calculated in view of establishing a perfect correspondence between the independent pieces that still encumber Fio. Ruins of palace, PMtg^dx. Flanoin and CosTB, flemaiuimaef PUte CLVII. the ground, and the softer material that was to fill up the intervals between the openings. Thus the rugosity at the sides of the stone frames facilitated adhesion, the sheer weight of the clay mass causing it to penetrate the slight uneven nesses of the field ; whereas had this been as finely polislied as the other apparent faces, the two elements must have parted during the desiccating process and consequent shrinking of the bricks. Nor is this all. Both at
- Ttfxier adduces valid reasons to show the unsonndness of any otiicr hypothesis,
yet does not care to commit himself to the conclasion which he f<ne8aw {Description, torn. ii. pp. 169-1S7). Flandin confines himself to the statement that small units were u^cd {Rtlation, torn. ii. p. 169). Diculafoy hns the merit of being the first who iniiisls thai the problem admits of no other solution {L'Art antique, torn. ii. p. 3 ; torn. t. p- 3 > ; torn. iii. p. 1 1). . . Digitized by Google
