Sec. ill] UDYANA IN CHINESE RECORDS OF TANG TIMES 15
with Sivaitic magic, and the Tibetans still regard that locality as the classic ground of sorcery and witchcraft.' 6
Hsüang-tsang then proceeds to give information which plainly shows that Buddhism, though still the predominant form of worship, was in a state of far-advanced decay. The people valued the law of Buddha, and believed reverentially in the Mahayana doctrine. Along the two banks of the Su p'o-su-tu or Swat River,' there had once existed 1,400 monasteries, but most of them were now in ruins. Once they had contained 18,o brethren, but now their number was greatly diminished. All these monks studied the Mahayana, and specially devoted themselves to the practice of meditation. They were fond of reading their texts, but were incapable of penetrating their meaning. They conducted themselves according to their rules, and specially cultivated the science of magical formulas. The pilgrim then enumerates the five redactions of the Vinaya taught, curiously enough all belonging to the ' Little Vehicle ', and adds that there were also about ten Deva temples, in which lived a medley of various sectarians, a distinct reference to Hindu worship.
According to him, Udyana contained four or five strong cities, and most of its rulers had taken
for their capital the town of Mêng- chieh-li ft4 3, which was sixteen or seventeen li in circuit,
and supported a flourishing population. The identification of Mêng-chieh-li with the present Manglaur, first apparently proposed by V. de Saint-Martin, is strongly supported by what is known of the natural advantages of the site and the extensive ruins about it.8 Hsüan-tsang uses the town as the starting-point for the topographical indications he furnishes as regards the various sacred localities detailed in his Memoirs, and the relative ease with which it has been possible to trace these within the area so far accessible for an archaeological survey helps to give confidence in that identification. Unfortunately the only opportunity which has so far brought Europeans to Manglaur and its neighbourhood was the rapid punitive expedition to Upper Swat in the course of the great tribal rising of 1897, and this could not be utilized for archaeological work. In consequence the various Stupas which Hstian-tsang's Memoirs specify at short distances to the south-west, west, and north-east of Mêng-chieh-li, cannot be identified at present with any certainty.° It seems, however, probable that the Stûpa which Uttarasena, an early king of Udyana, was believed to have erected over relics of Buddha's body, is marked by the great ruined dome near Barikôt on the left bank of the Swat River, as the distance and bearing agree closely with the sixty to seventy li to the south-west of the capital which the pilgrim mentions.10
The ground for identification is safer, even though the distance from the surveyed area be greater, as regards the spring of the Naga Apalala already discussed above, in which a legend related at length by the pilgrim placed the source of the Swat River." With this once located it
Topographia sacra of Udyana.
6 Cf. Yule, Marco Polo, i. pp. 164 sq.
7 This form of the river name, as found in the old text A of the Hsi-yii-chi (cf. Waiters, Yuan Chwang, i. p. 226), is distinctly preferable to the form Su-p`o an ?) fa-su-tu found in other texts, as it represents a correct phonetic transcription of Suvaslu, the original Skr. form of the name. Cf. Marquart, Zur Geschichte von Iran, ii. p. 248, who justly protests against Lassen's reconstruction of a Skr. *.§ubhavastu out of the reading Su p`o fa-su-lu which, if correct at all, probably is due to the influence of a learned etymology. Cf. Deane, Notes on Udyana, J.R.A.S., 1896, pp. 655 sqq.; V. de Saint-Martin, in Julien, ill/moires, ii. pp.
9 For conjectures regarding the sacred sites thus men
tioned by Hsiiang-tsang (see Julien, Mdmoires, i. pp. 133, 139 sq. ; Watters, Yuan Chwang, i. pp. 227, 236 sqq.), cf. Deane, J.R.A.S., 1896, pp. 659 sq.
10 For a photograph of the Barikôt Stûpa, which is just visible from the Landaké ridge, the eastern limit of Lower Swat, sce Foucher, L'art du Gandhâra, i. p. 67.
11 See above, p. 8 ; Deane, J.R.A.S., 1896, p. 656. Watters, Yuan Chwang, i. pp. 229 sq., rectifies several obvious mistakes in Julien's translation of the passage describing the origin of the Swat River. The wicked Naga before his conversion used to destroy the crops of the country by the ' white water', i. e. the floods he sent down from his spring. After his submission he was allowed for his maintenance to carry off the crops only once in every twelve