Hsüan Chuang broke his return journey at Khotan in 644. He mentions the fondness of the people for music and says that their language differed from that of other countries. The Mahâyâna was the prevalent sect but the pilgrim stopped in a monastery of the Sarvâstivâdins[515]. He describes several sites in the neighbourhood, particularly the Go'sringa or Cow-horn mountain[516], supposed to have been visited by the Buddha. Though he does not mention Zoroastrians, he notices that the people of P'i-mo near Khotan were not Buddhists. About 674 the king of Khotan did personal homage at the Chinese Court. The Emperor constituted his territory into a government called P'i-sha after the deity P'i-sha-mên or Vai'sravana and made him responsible for its administration. Another king did homage between 742 and 755 and received an imperial princess as his consort. Chinese political influence was effective until the last decade of the eighth century but after 790 the conquests of the Tibetans put an end to it and there is no mention of Khotan in the Chinese Annals for about 150 years. Numerous Tibetan manuscripts and inscriptions found at Endere testify to these conquests. The rule of the Uigurs who replaced Tibet as the dominant power in Turfan and the northern Tarim basin does not appear to have extended to Khotan. It is not till 938 that we hear of renewed diplomatic relations with China. The Imperial Court received an embassy from Khotan and deemed it of sufficient importance to despatch a special mission in return. Eight other embassies were sent to China in the tenth century and at least three of them were accompanied by Buddhist priests. Their object was probably to solicit help against the attacks of Mohammedans. No details are known as to the Mohammedan conquest but it apparently took place between 970 and 1009 after a long struggle. Another cultural centre of the Tarim basin must have existed in the oases near Lob-nor where Miran and a nameless site to the north of the lake have been investigated by Stein. They have yielded numerous Tibetan documents, but also fine remains of Gandharan art and Prakrit documents written in the Kharoshthî character. Probably the use of this language and alphabet was not common further east, for though a Kharoshthî fragment was found by Stein in an old Chinese frontier post[517] the library of Tun-huang yielded no specimens of them. That library, however, dating apparently from the epoch of the T'ang, contained some Sanskrit Buddhist literature and was rich in Sogdian, Turkish, and Tibetan manuscripts. 4 Ample as are the materials for the study of Buddhism in Central Asia those hitherto published throw little light on the time and manner of its introduction. At present much is hypothetical for we have few historical data--such as the career of Kumârajîva and the inscription on the Temple of Maitreya at Turfan--but a great mass of literary and artistic evidence from which various deductions can be drawn.