It is probably to this Gandharan Buddhism that the Chinese pilgrims refer when they speak of the Sarvâstivâdin school of the Hînayâna as prevalent. It is known that this school was closely connected with the Council of Kanishka. Its metaphysics were decidedly not Mahâyânist but there is no reason why it should have objected to the veneration of such Bodhisattvas as are portrayed in the Gandhara sculptures. An interesting passage in the life of Hsüan Chuang relates that he had a dispute in Kucha with a Mahâyânist doctor who maintained that the books called Tsa-hsin, Chü-shê, and P'i-sha were sufficient for salvation, and denounced the Yogaśâstra as heretical, to the great indignation of the pilgrim[523] whose practical definition of Mahâyânism seems to have been the acceptance of this work, reputed to have been revealed by Maitreya to Asanga. Such a definition and division might leave in the Hînayâna much that we should not expect to find there. The Mahâyânist Buddhism of Khotan was a separate stream and Hsüan Chuang says that it came from Kashmir. Though Kashmir is not known as a centre of Mahâyânism, yet it would be a natural route for men and ideas passing from any part of India to Khotan. 5 The Tarim basin and the lands of the Oxus[524] were a region where different religions and cultures mingled and there is no difficulty in supposing that Buddhism might have amalgamated there with Zoroastrianism or Christianity. The question is whether there is any evidence for such amalgamation. It is above all in its relations with China that Central Asia appears as an exchange of religions. It passed on to China the art and thought of India, perhaps adding something of its own on the way and then received them back from China with further additions[525]. It certainly received a great deal from Persia: the number of manuscripts in different Iranian languages puts this beyond doubt. Equally undoubted is its debt to India, but it would be of even greater interest to determine whether Indian Buddhism owes a debt to Central Asia and to define that debt. For Tibet the relation was mutual. The Tibetans occupied the Tarim basin during a century and according to their traditions monks went from Khotan to instruct Tibet. The Buddhist literature discovered in Central Asia represents, like its architecture, several periods. We have first of all the fragments of the Sanskrit Agamas, found at Turfan, Tun-huang, and in the Khotan district: fragments of the dramas and poems of Aśvaghosha from Turfan: the Prâtimoksha of the Sarvastivâdins from Kucha and numerous versions of the anthology called Dharmapada or Udâna. The most interesting of these is the Prakrit version found in the neighbourhood of Khotan, but fragments in Tokharian and Sanskrit have also been discovered. All this literature probably represents the canon as it existed in the epoch of Kanishka and of the Gandharan sculptures, or at least the older stratum in that canon.