274 History op Art in Antiquity. Pic. 13 The T«kht>i<Soldinau. Plan. Flandin aud CosTEt tieru ameitntttt Plate CCI. ture. It is a parallelogram with only three sides ; the fourth facr, or rather where it should have been, is formed by a curve in the slope of the hill, in strong contrast with the geometrical regularity of the other parts of the tracing. ResaultSi in the form of advanced works, occur at the four corners of this enormous die of masonry (Fig. 139). All who have seen it are agreed that it is a substructure. Of its construction we have already spoken;' it remains to determine what it was intended to carry. ^. Some have spoken of a citadel.^ If so, the situation was singu- larly ill chosen ; the right place for a cas- /f ////;//, whose function was to cover the plain from the enemy threatening it on the north, would have been at the entrance of the defiles whence emerges, along with a stream, the road that comes from Me- dia. Had this been a purely defensive work, would the stone have been so bcautitul Would they have taken the trouble to ' Hist, of Art, torn. v. pp. 469, 470. '•' Flandin and Coste, Perse ancienne, pp. 161, 162 ; TtxiER, Description^ toni. li. P- 149. Fiti. 140.— The TakhL Llevaliun uf angle of wall. JHd, Plate ecu. uiyiii^LU by Google
