________________
INTRODUCTION.
xix
of Dastûr Gadman-pîrug Aspendiyâr Gadman-pirug, and that from the copy of Dastur Shatrô-aiyyar Vêgan Khúsrôishah, and that, as regards these several sayings, was written from the copy of the heavenly-destined Mah-vindad Naremâhân with the righteous soul, and comes unto us from the realm of the Hindûs. May even our writing be in accordance with the will of the sacred beings.' In addition to the date, the chief matter of interest in this colophon is its acknowledg. ment of the fact that the work had come from India, where the original Pahlavi text appears to have since become extinct. We have, therefore, in this text, merely so much of the work as had reached India, on which the PazandSanskrit version of Nêryðsang, described below, was undoubtedly based; and the possibility of hereafter finding the latter part of the work in Persia should not be overlooked. It is, however, upon the text contained in K43, so far as it is preserved, that the translation of the Dînâ-î Maînôg-î Khirad in this volume is founded.
Of the other Pahlavi manuscript, TD2, nothing further is known to the translator than a copy of the passages corresponding to those contained in the ten folios lost from K43, upon which copy the translation of those passages has been based.
Besides these manuscripts of the original Pahlavi text, there exist other copies, in which the text has been merely reproduced from the Pazand version described below; and, of these copies, K22 (No. 22 in the University Library at Kopenhagen) may be cited as a typical example. This manuscript is a large octavo volume of 56 folios of glazed Indian paper, probably about a century old, but without a date. The first 48 folios contain a corrupt Pahlavi text of the Dînâ-î Maînôg-î Khirad, alternating with the usual Sanskrit version described below, written nineteen lines to the page, and extending as far as Chap. XXVII, 41. The corruptions in the text consist of misuse of Huzvâris equivalents, and errors in orthography which no old writer of Pahlavi would be likely to commit, such as writing kolâvist for harvist, nafsmanîdârîk for kh vêsinîdârîh, baragūmân for a vîgûmân, hamê for hamâî, avas for a û bas,
b 2
Digitized by Google
