The translation of all this literature falls into three periods, (i) from the seventh century until the reign of Ralpachan in the ninth, (ii) the reign of Ralpachan, and (in) some decades following the arrival of Atisa in 1038.  In the first period work was sporadic and the translations made were not always those preserved in the Kanjur.  Thonmi Sanbhota, the envoy sent to India in 616 is said to have made renderings of the Karanda Vyuha and other works (but not those now extant) and three items in the Tanjur are attributed to him.[997] The existence of early translations has been confirmed by Stein who discovered at Endere a Tibetan manuscript of the Salistambhasutra which is said not to be later than about 740 A.D.[998] The version now found in the Kanjur appears to be a revision and expansion of this earlier text.

A few translations from Chinese texts are attributed to the reign of Khri-gtsug-lde-btsan (705-755) and Rockhill calls attention to the interesting statement that he sent envoys to India who learned Sanskrit books by heart and on their return reproduced them in Tibetan.  If this was a common habit, it may be one of the reasons why Tibetan translations sometimes show differences in length, arrangement and even subject matter when compared with Sanskrit and Chinese versions bearing the same name.  During the reign of Khri-sron-lde-btsan and the visit of Padma-Sambhava (which began in A.D. 747 according to the traditional chronology) the number of translations began to increase.  Two works ascribed to the king and one to the saint are included in the canon, but the most prolific writer and translator of this period was Kamalasila.  Seventeen of his original works are preserved in the Tanjur and he translated part of the Ratnakuta.  The great period of translation—­the Augustan age of Tibet as it is often called—­was beginning and a solid foundation was laid by composing two dictionaries containing a collection of Sanskrit Buddhist terms.[999]

The Augustus of Tibet was Ralpachan who ruled in the ninth century, though Tibetan and Chinese chronicles are not in accord as to his exact date.  He summoned from Kashmir and India many celebrated doctors who with the help of native assistants took seriously in hand the business of rendering the canon into Tibetan.  They revised the existing translations and added many more of their own.  It is probable that at least half of the works now contained in the Kanjur and Tanjur were translated or revised at this time and that the additions made later were chiefly Tantras (rGyud).  On the other hand it is also probable that many tantric translations ascribed to this epoch are really later.[1000] The most prolific of Ralpachan’s translators was Jinamitra, a pandit of Kashmir described as belonging to the Vaibhashika school, who translated a large part of the Vinaya and many sutras.[1001] Among the many Tibetan assistants Ye’ses-sde and Dpal-brTsegs are perhaps those most frequently mentioned.  These Tibetan translators are commonly described by the title of Lo-tsa-va.  As in China the usual procedure seems to have been that an Indian pandit explained the sacred text to a native.  The latter then wrote it down, but whereas in China he generally paraphrased whatever he understood, in Tibet he endeavoured to reproduce it with laborious fidelity.