established therein the first community of Tibetan Buddhist lāmas, in the year 749.
During his sojourn in Tibet at that time, and during subsequent visits, Padma Sambhava had many Tantric books translated into Tibetan out of Indian Sanskrit originals—some of which have been preserved in the monasteries of Tibet—and hidden away with appropriate mystic ceremonies in various secret places. He also endowed certain of his disciples with the yogīc power of reincarnating at the proper time, as determined by astrology, in order to take them out, along with the treasures hidden away with them and the requisites needed for properly performing the rites described in the texts. This is the generally accepted tradition; but according to another tradition the Tertons are to be regarded as various incarnations of the Great Guru himself. According to a rough estimate, the religious texts already taken out by such Tertons, from century to century, would form an encyclopædia of about sixty-five volumes of block-prints, each, on an average, consisting of about four hundred ordinary-sized folios.
Our text, the Bardo Thödol, being one of these recovered apocryphal books, should, therefore, be regarded as having been compiled (for the internal evidence suggests that it was a Tibetan compilation rather than a direct translation from some unknown Sanskrit original) during the first centuries of Lāmaism, either—as it purports to have been—in the time of Padma Sambhava or soon afterwards. Its present general use all over Tibet as a funeral ritual and its acceptance by the different sects, in varying versions, could not have been the outcome of a few generations; it testifies rather convincingly to its antiquity, bears out the pre-Buddhistic and at least partially Bön origin which we attribute to it, and suggests some validity in the claims made for the Tertons.
We are well aware of the adverse criticisms passed by European critics on the Terton tradition. There is not lacking, nevertheless, sound reason for suspecting that the European critics are not altogether right. Therefore, it seems to us that the only sound attitude to assume towards the Terton problem is to keep an open mind until sufficient data accumulate to
