THE CASPIAN SEA AND ITS NEIGHBORS 343
there is now no place near the Sea of Aral where snow stays so long.
The next important author, Edrisi, A. D. 1154, speaks of Aral as " a well-known lake," and confirms most of what Istakhri says, including the snow. He gives the distance, however, between the mouths of the Oxus and Jaxartes in Lake Aral as only ten miles. This apparently means that the Jaxartes had then changed its course to the old channel already referred to in connection with Herodotus, a course which it appears to have followed intermittently. It flowed there as late as 1816. Of the Caspian Sea, Edrisi says that " it is elongated from north to south less than from east to west. [The italics are mine.] The two axes have the ratio of four to three." His map, however, shows it otherwise. It seems as if we had here a revival of the Scythian gulf, either because Edrisi had read Ptolemy, or more probably because the rising water of the Caspian had once more broadened the southern end of the sea.
According to Rawlinson, " The Arab geographer Yacut ... about A. D. 1225 furnishes the earliest record of the Oxus having found its way into the Caspian, after it had been turned into its old bed by the Moghuls at the siege of Urgenj [ancient Khiva] in 1221." Rawlinson, it should be said, believed that in ancient Greek times the Oxus flowed to the Caspian. Hence his reference to the " old bed " into which the Mongols diverted the river. Abdulfeda, seventy or eighty years after Yacut, describes the Oxus and Aral in essentially the same way as Istakhri and Edrisi, with no reference to any connection with the Caspian. Possibly the overflow of the Oxus to the Caspian lasted only a few years.
IIIF Color HighRes
