any other purpose, in front of the gateway, where it might easily have fallen. On the other hand, the clear space between the jambs is nearly twelve feet in length, and it may be thought that the jambs are not designed of sufficient relative proportions to support a mass and weight so great as would have been required of a single stone that spanned them. In the preserved upper part of the left-hand monolith also (B) it may be seen that the horizontal portion declines a little as though to form the offspring of an arch, while the vertical stop is inclined slightly outwards, as though designed to receive the direct thrust of an arch, whether of two large stones or of many small ones. The thrust would be further taken by a backing of masonry behind the monoliths, which may be seen from the plan to have been included in the original design. Notwithstanding these considerations, the material evidence in this case is in favour of a great stone lintel, of strength and size proportionate to the width of its span. A third point is the reconstruction of the upper courses of the building, and this involves a consideration of the mutual relation of the upper and lower walls of the frontage (EF and XY). With regard to the upper wall (EF), a dressed block of the second course remains seemingly not much out of its original position.[592] This prepares us for the restoration of the whole course in stone, and possibly another fallen block (lying just in front of that marked _f_) might be fitted into place to the left hand of the former. This creates for itself a precedent, for in other sculptured walls that are known, as at Sinjerli and Sakje-Geuzi, the wall was carried up in brick. In those cases, however, the sculptures were carved on facing slabs merely, not on cubical building blocks, so that for architectural analogy we must look rather to the palace at Boghaz-Keui. The latter, however, furnishes no direct evidence on the point, though M. Perrot suspected that it was carried up on a timber frame. We must, then, accept the suggestion of this single stone _in situ_; but we must hesitate to accept without clear proof the theory that there were sculptures also on the second tier,[593] a feature for which we have no analogy in Hittite works. The two sculptured blocks (_x_, _y_) recently found at a lower level, in front of the lower wall (XY), represent subjects entirely different to those _in situ_, and seem from the published photographs[594] to illustrate a phase of art as different as the early and later reliefs of Sinjerli. Measurements taken of stones irregular in outline are apt to be deceptive, and not until these, when tried, are found to fit accurately into the position suggested, alongside that which is _in situ_, on the second course of the upper wall, can we believe that such was their original position. It has been noted that these blocks were found in the excavation which disclosed the lower wall. Now there is clear suggestion, in plan and