The Transmission Of Sanskrit Manuscripts 111
thugs dam. The last one, on the Aṣṭasāhasrikā manuscript, is obviously a later
addition by a Tibetan hand. In the cases of the Hevajrapañjikā and *Parikathā,
the notes are written at the bottom of the manuscripts, and this again sug-
gests that they were added later, but the question about the person who added
them remains open (i.e. we do not know whether they were added by Atiśa
or a direct disciple of his or by someone else at a later time). Accordingly, the
materials of group (b) are relatively reliable indicators that Atiśa did indeed
own the manuscripts, but they cannot be called rock-solid ones.
The last group (c) comprises colophons that establish Atiśa’s possession.
They are the Sanskrit manuscript of the Mahāsūtrasamuccaya (section 3, [5])
and that of an unknown work preserved in the Potala (section 5, Box A, [3],
12 fols.). We may add the Sanskrit manuscript of Ratnākaraśānti’s Śuddhimatī,
whose colophon has the name of Dīpaṃkaraśrījñāna (see section 4, [9] and
footnote 43).84 Compared with materials of the other two groups ([a] and [b]),
these colophons provide the most reliable evidence for establishing the histo-
ricity of Atiśa’s ownership.85
7 Conclusion
We have seen the first fruitless attempts to access the Sanskrit manuscripts
of Retreng by Ekai Kawaguchi in 1914 and Sāṅkṛtyāyana and Gendün chöpel
in 1934. Gendün chöpel nevertheless drew attention to some important facts: in
Retreng, Tāranātha saw manuscripts that had belonged to Atiśa, some of which
were sent back to India after Atiśa’s death. The first fact is stated in Tāranātha’s
autobiography, and the second in biographies of Atiśa. Furthermore, we also
find documents telling of Dromtön’s acquisition of Atiśa’s Sanskrit manu-
scripts in 1054 and their being deposited in Retreng in 1056.
84 I am grateful to Dr. Luo Hong for having supplied this information.
85 Of course, we cannot completely exclude the possibility that some unknown person later
added the colophon in order to attribute the ownership of the manuscripts to Atiśa, but it
would be highly unnatural for someone to add completely new sentences to the colophon
of some of Atiśa’s most precious religious possessions. In general, authorship mentioned
in a colophon of a work is sometimes a later attribution, whereas ownership of the manu-
script appearing in a colophon is more reliable. It is necessary to question critically the
historicity of a wide range of statements in Tibetan historical literature, but at the same
time one should avoid excessive skepticism as offering its own pitfalls.