________________
INTRODUCTION.
xli
from the fragment translated above. Regarding its original extent, it is possible to make an approximate estimate, by calculating the quantity of text which the 129 lost folios must have contained, from the quantity actually existing on folio 130. According to this calculation, the original extent of the text of the Bundahis in this MS. must have been very nearly 30,000 words; and it is remarkable that a similar calculation of the extent of the text in TD, based upon the actual contents of ten folios out of 103, gives precisely the same result. This coincidence is a strong argument in favour of the absolute identity of the text lost from Westergaard's MS. with that actually existing in TD; it shows, further, that the original extent of the Bundahis may now be safely estimated at 30,000 words, instead of the 13,000 contained in K20 when that MS. was complete.
That this fragment belonged to a separate MS., and is not the folio missing from the end of TD, is shown not only by its containing more of the text than is said to be missing, but also by the first folio of the fragment being numbered 130, instead of 103, and by its containing fifteen lines to the page, instead of seventeen, as would be necessary in order to correspond with TD.
1
Regarding the age of the Bundahis many opinions have been hazarded, but as they have been chiefly based upon minute details of supposed internal evidence evolved from each writer's special misinterpretation of the text, it is unnecessary to detail them. The only indication of its age that can be fairly obtained from internal evidence, is that the text of the Bundahis could not have been completed, in its present form, until after the Muhammadan conquest of Persia (A. D. 651). This is shown not only by the statements that the sovereignty 'went to the Arabs' (Chap. XXXIV, 9), that 'now, through the invasion of the Arabs, they (the negroes) are again diffused through the country of Iran' (Chap. XXIII, 3), and that 'whoever keeps the year by the revolution of the moon mingles summer with winter and winter with summer' (Chap. XXV, 19, referring probably to the Muhammadan year not corresponding with the seasons), but also, more positively
Digitized by Google