kalular-khana Ula auchub shunaar >**} ^ %**T Thfroy'al that the walled-town of Akhsi is «£**^£je&1. wiTSs dwellings are along a ravine. The WW*.J*J%* ., died) . pigeons and their house from the ravine, became a ialcon (t.e. aieu,.
A few particulars about Akhsi will shew that, in the translations just quoted, certain small changes of wording are dictated by what, amongst other writers, Kostenko and von Schwarz have written about the oases of Turkistan.
The name Akhsi, as used by Ibn Haukal, Yaqiit and Babur describes an oasis township, U a walled-town with its adjacent cultivated lands. In YSqufs time Akhsi had a second circumvallation, presumably less for defence than for the protection of crops against wild animals. The oasis was created by the Kasan-water,* upon the riverain loess of the right and higher bank of the Saihun (Sir), on level ground west of the junction of the Narin and the Qara-darya, west too of spurs from the northern hills which now abut upon the river. Yaqut locates it in the 12th century, at one farsakh {circa 4 m.) north ot the river/ Depending as it did solely on the Kasjin-water, nothing dictated its location close to the Sir, along which there is now, and there seems to have been in the 12th century, a strip of waste land. Babur says of Akhsi what Kostenko says (i, 321) of modern Tashkint, that it stood above ravines (jarlar). These were natural or artificial channels of the Kasan-water.
To turn now to the translations ;-Mr. Erskine imaged Akhsi as a castle, high on a precipice in process of erosion by the Sir. But Babur's word, qurghan means the walled-town; his for a castle is ark, citadel; and his jar, a cleft, is not rendered by ' precipice.' Again ;-it is no more necessary to understand that
» Until the Yangi-ariq was taken off the Sir late in *^£%%>gt Namangan, the oasis land of Farghana was fertilized, not from the river Dux by its intercepted tributaries.
, ,, , the
2 Ujfalvy's translation of Yaqut (ii,_ 179) reads one farsakh from tne mountains instead of ' north of the river.' .
3 Kostenko describes a division of Tashkint, one in *"* MJ^t?£: ijar-kucha), as divided by a deep ravine ; of another he says that it is cut oy deep ravines (Babur's 'utnlq jarlar).