Sec. iv] REMAINS ON THE HILL-TOP 923
of a holy man known as ` Pir Chilli ', who is believed to have been a Miràsi or strolling player from India.
The whole area around the Ziàrats, as Fig. 475 shows, is covered with graveyards. The tombs are in most cases, but not always, placed above ground, a mode of burial which here, owing to the abundance of rough stone material, was, perhaps, more convenient than the digging of graves in rocky soil. There can be no doubt that the vast majority of the tombs here, as on the slopes close to Ghàgha-shahr, are Muhammadan ; but I noticed some which did not show the orthodox orientation. Most of the tombs I saw had been opened. This was explained by my local informants as due to extensive rifling operations said to have been carried out by Sistàn villagers ` three or four generations ago '. Finds of trinkets, jewellery, and the like were alleged to have been made then, and the trouble taken to open these hundreds of tombs suggested that there was some foundation for the statement.
A very curious feature on the plateau are the numerous large excavations to be found on more or less level ground between Kok-i-Zal and the Ziàrats. They undoubtedly mark old quarries, and in view of their position it seems difficult .to believe that they could have served any other purpose but that of furnishing materials for the multitude of surface tombs. Within or near the pits I could see only the live rock or small pieces of stone useless for the purpose of rough masonry. Large spoil heaps of such pieces litter the ground near the pits. The excavations were said to retain water for a short time after rain, and this chance is duly appreciated by the pilgrims. But that the pits were constructed to serve as reservoirs is very unlikely.
That the local worship of Kôh-i-Khwàja attested by Ziàrats, cemeteries, and name is old does not stand in need of special demonstration. It is obvious that this hill with its cap of igneous rock, rising in impressive isolation fully 400 feet in the middle of the level expanse of marsh and alluvial plain of the Sistàn basin, was bound from early times to attract the veneration of those dwelling in its vicinity and to become for them, to use the Indian hieratic term, a svayambhûtirtha, ` a self-created place of worship '. Fortunately it is possible for us to prove this ancient fame of the hill from the earliest religious texts of Irin, the Avesta. I have already referred above to the passage Yasht xix. 66 in connexion with the importance attaching in the Avesta to the lake of Sistàn.3 Vivid recollection of that passage, due to its having been the subject of my first effort in print, made me realize, while still at Kôh-i-Khwàja, that it contains not only the names of the rivers which flow into the Sistàn lake but also the name of the hill which rises from its midst. It was subsequently a special satisfaction to me to find that the correctness of the interpretation of the passage which leads us to identify Kôh-i-Khwàja with Mount Ushidhâo of the Avesta had already been recognized by my old and much respected friend, the late Professor James Darmesteter.4
The passage Yasht xix. 66-7 runs thus : 5 ughrem ah'aretem h°arenô ... yat upanhacaiti yô avadhâi frakhshayêitê, yathâ zrayô yat Kâfaêm Haêtumatem, yathâ gairis yô Ushidhâo yim aiwitô paoiris âpô ham gairishâcô jafentô. (67) avi tem avi-hantacaiti avi tem avi-ham-vazaitê Hvâçtraca Hvafpaca Fradatha H°arenuhaitica yd frîra Ustavaitica yd Ora Urvadhaca pouru-vâçtra Erezica Zarenumatica. avi tem avi-hantacaiti avi tem avi-ham-vazaitê Haê[tumdo] 6 raêvâo hvarenuhâo . . . ` [We worship] the mighty unattainable [kingly] glory which attaches itself to him who rules there where is the lake Kâçaoya formed by the Helmand, where Mount Ushidhâo is, around which many mountain streams come together. Towards this [mount] flows and unites the H°àçtra and the
3 See above, ii. pp. 906 sq.
4 Cf. Darmesteter, Zend-Avesta, ii. p. 634.
5 For the sake of typographical convenience the transcription follows the system once rendered familiar by Justi's
Handbuch, replacing q by h°.
6 Professor Geldner's emendation of Haê - - - - of the text into the name Haêtunuto is certain and generally accepted ; see Bartholomae, Altiran. Wörterbuch, p. 1729.