Towns and their Defences. correct the testimony of Herodotus relating to a mode of decora- tion likely to have impressed him in that it was utterly opposed to the taste and habits of Greece, whilst his minute details could not be readily trumped up. He had them from a voucher worthy of belief. If the Median kings tinted the summits of their walls it was on the model of Assyrian edifices, for at that time the only influence wdiich could reach them was the grand civilization, many centuries old, of the Euphrates valley. In the variously coloured crenelations Herodotus enumerates, we recognize an arrangement of frequent occurrence on the enamels of Nineveh, Babylon, and Susa. That the tradition has not died out in the country, the mosques of modern Persia attest, to embellish which the enamellist has occasionally introduced tiles of brilliant pris- matic hues, refle.es of gold and silver, similar to those which glinted about the loftier walls at Ecbatana ; that is to say, those next to the palace.* No isolated mound has been reported, either from llamadan or the immediate neighbourhood, answering to the idea which the account of Polybius relating to the site of the old castle at Ecba- tana is apt to conjure up in our mind. The aspect of the ground appears to have completely changed since antiquity, so that no traveller, at first sight, has been able to identify it. Sir Henry Rawlinson is inclined to seek the fortress with the sevenfold wall, not in the vicinity of Hamadan, but in Media Atropatene, at a place called Takht-i-Suleiman.'^ Its situation is certainly remarkably strong, and art has greatly added thereto and increased the number of its natural lines of defence. The whole question resolves itself in this : Were there actually two Ecbatanas in Media — as Moses Chorenus seems to imply — which Greek historians, ignorant of the fact, confounded one with the other. Did Herodotus's description apply to northern or southern Ecbatana, e.g. the city of Polybius. and the historians of Alexander's campaign ? This is not the place to discuss a point bearing upon history and topography. It is enough to have pointed it out to the curiosity of future explorers. > Flandin and Coite point oat a hillock whidi rises in the middle of the plain east of Hamadan, as the most likely site for the fortress under consideration. Traces of ancient remains are certainly seen here, but in other respects nothing about the site corresponds with the data furnished by Herodotus {First ancienne, p. i8). » G. Rawi.inson, The Five Great Monarchies^ torn. ii. p. 268. Digitized by Google
