It is always extremely difficult to restore any building from mere verbal description, and still more so when erected by a people of whose architecture we know so little as we do of that of the Jews. Still, the woodcut on the opposite page is probably not very far from representing the Temple as it was after the last restoration by Herod, barring of course the screen bearing the Vine mentioned above, which is omitted. Without attempting to justify every detail, it seems such a mixture of Roman with Phœnician forms as might be expected and is warranted by Josephus’s description. There is no feature for which authority could not be quoted, but the difficulty is to know whether or not the example adduced is the right one, or the one which bears most directly on the subject. After all, perhaps, its principal defect is that it does not (how can a modern restoration?) do justice to the grandeur and beauty of the whole. As it has been necessary to anticipate the chronological sequence of events in order not to separate the temples of the Jews from one another, it may be as well before proceeding further to allude to several temples similarly situated which apparently were originally Semitic shrines but rebuilt in Roman times. That at Palmyra, for instance, is a building very closely resembling that at Jerusalem, in so far at least as the outer enclosure is concerned.[112] It consists of a cloistered enclosure of somewhat larger dimensions, measuring externally 730 ft. by 715, with a small temple of an anomalous form in the centre. It wants, however, all the inner enclosures and curious substructures of the Jewish fane; but this may have arisen from its having been rebuilt in late Roman times, and consequently shorn of these peculiarities. It is so similar, however, that it must be regarded as a cognate temple to that at Jerusalem, though re-erected by a people of another race. A third temple, apparently very similar to these, is that of Kangovar in Persia.[113] Only a portion now remains of the great court in which it stood, and which was nearly of the same dimensions as those of Jerusalem and Palmyra, being 660 ft. by 568. In the centre are the vestiges of a small temple. At Aizaini in Asia Minor[114] is a fourth, with a similar court; but here the temple is more important, and assumes more distinctly the forms of a regular Roman peristylar temple of the usual form, though still small and insignificant for so considerable an enclosure. The mosque of Damascus was once one of these great square temple-enclosures, with a small temple, properly so called, in the centre. It may have been as magnificent, perhaps more so, than any of these just enumerated, but it has been so altered by Christian and Moslem rebuildings, that it is almost impossible now to make out what its original form may have been.