158 HISTORICAL NOTICES OF KHOTAN [Chap. VII
him. All that can safely be asserted in regard to this piece of folklore is that the ` popular etymology ' upon which it rests presupposes an early Sanskritization of the true old name of Khotan into Kustana. And, as we have seen above, this Sanskritized form is proved by documentary evidence to have been in current use in the third century of our era.
The connexion of this mythic eponymous king with the god Vaigravana or Kubera possesses historical interest only in as far as it shows that the worship of this divinity of the Indian pantheon must have been popular in Khotan from an early period. Kubera, the Hindu god of wealth, king of the demons or Yaksas, is a figure familiar to Buddhist mythology of all countries from Gandhara to Japan. As he is worshipped pre-eminently among the four ` Lokapalas ' or guardians of the world as the ruler of the North, his adoption by Khotan as the genius loci of the kingdom seems thoroughly appropriate 6. We shall see that Vaigravana claims this position yet more conspicuously in the Tibetan version of the legend, and we have Hsüan-tsang's explicit testimony for the popularity and wealth of his shrine in the Khotan capital.
That the same tradition which credited the Khotan dynasty ruling in the seventh century of our era with a divine origin, should have assumed it to have ruled without a break and in direct descent from the time of its founder, is natural enough. We need scarcely insist on the fact that the Annals of the Later Hans record a succession of usurpers on the throne of Khotan for the years 55-75 A. D., in order to justify critical doubts as to this unbroken continuity of the reigning house. But these doubts cannot impair the significance of that salient feature of the tradition which ascribes the first occupation of Khotan to the joint settlement of two colonies drawn from the extreme north-west of India and from China.
Several considerations combine to invest this part of Hsüan-tsang's story with a value quite distinct from that of the rest. In the first place, it deserves attention that the tradition on this point has no direct connexion with the eponymous legend of Kustana. It is further noteworthy that, though the immigration of the Indian colony is placed in the time of Moka, the most renowned of the patrons of Buddhism, the introduction of Buddhist worship into Khotan is distinctly attributed, as we shall see, by both Hsüan-tsang and the Tibetan ` Annals of Li-yul ' to a later period. It is thus evident that the introduction of an Indian element into the story of the genesis of Khotan could not have been a specifically Buddhistic fiction intended merely, as otherwise might be supposed, to account for the flourishing condition of the Buddhist church in Khotan. Finally, it is characteristic that, while according to the story as heard by Hsüan-tsang the union of the two colonies in Khotan was brought about by the victory of a prince exiled from China, yet there is nothing in the tradition indicating a tendency to flatter Chinese notions. Considering the close relations which had already united Khotan to China for centuries and the rapid reassertion of Chinese influence throughout Eastern Turkestan at the time of Hsüan-tsang's visit, the absence of any such tendency or of any attempt to connect some event of Chinese history with the Khotan tradition seems to encourage the belief that the latter possessed in part at least a genuine foundation.
In the story told of the mysterious way in which the site for the capital of Khotan was chosen we can, of course, recognize only a piece of folklore ; but this, too, is of interest,
c Compare, regarding Kubera : Vaigravana, also called Yakarâja, Dhanapati, &c. Grünwedel-Burgess, Buddhist art in India, pp. 45, 136 sq. For his representation in an interesting sculpture excavated in one of the Dandân-Uiliq shrines, D. II., see below chap. ix. sec. iii.
[I may point out here in passing that the distinctly
non-`Aryan' features, which have caused a famous Gandhara statue of Kubera in the Lahore Museum (see fig. 88 in Grünwedel-Burgess, loc. cit.) to be mistaken for that of an Indo-Scythian king ', would find their best explanation in this special connexion of the god with the Scythian North].