^ ledu ) botJi in (J lestu ^ Gatha vata
vaitulya (alongwith vaipulya)
(in Ch. I—IV)
(but in Gatha graddadhisyanti)
(but in Gatha = O.)
As to the orthography of the Kashgar text it will suffice to remark that in the correct use of b and v, as well as of the three sibilants, it excels the Nepalese MSS. In general the distinction of those sounds is the same as in the Bombay editions of Sanskrit texts. A notable exception is Baranasï for Varanasï, like in Pali. Worth mentioning is the regular use of the Jihvamülïya in duhkha and similar cases. Barring in the instance of Baranasï we have not scrupled to follow the spelling of the Kashgar text in respect to the aforenamed sounds, without paying regard to the other MSS.
The characteristics, such as described, warrant, in my opinion, the conclusion that, in formal respect, the Kashgar version is more original than the Nepalese texts, and that the more regular Sanskrit of the latter is the outcome of subsequent innovations due to continued revision. It is a quite different question whether the Kashgar version is materially a more faithful copy of the original composition of the Saddharmapundarïka than the other texts we know of. An attempt at solving this question will only be possible after a comparison of the existing Chinese translations with the texts of the Nepalese MSS. and with the Kashgar text or maybe texts, a complete edition of which is much to be desired.
Whatever may be the results of further investigations, it can not be doubted that the Kashgar writings precede some centuries the date of the oldest MSS. from Nepal, though we are unable to define the time of those writings with something like precision. This much may be observed that the kind of Brahmï of our Kashgar text exactly resembles the writing of the two pages containing a fragment of the Sanskrit Dharmapada with the