' Completed in great joy and full of gratification this last portion of the manuscript of the incomparable, priceless, and unequalled Dfokari/, at the place where >/ was found and happily disinterred ^ by us in Asi^ristdn, within the happily prosperous, odoriferous, precious, well-thriving, and glorious Bakd&^ of Good Recti- tude'; from a copy which, as regards the religion, is Jus/ as the leaders of the saintly and orthodox, who were of the family of the saintly Mttrp^, son of Miraspend, (who re-explained know- ledge, by five or six well-destined ones, from the pure revelation
^ Khiij-kand might be the name of a place here, but camiot be so in the
next pamgraph.
' The angel Ahaiirrang (Av. ashij vangnhi).
XXXIV PAHLAVI TEXTS.
which is the all-embellished learning of learnings) and the suc- cessive leaders of the orthodox {who again provided at different times [ahamv&r] for its restoration, through manuscripts ai various places, to maintain reading and investigation therein) had written.
^ The remainder of this colophon, so far as it is here translated, is also quoted in the second colophon.
' Here written ar'^5, bat it is an-ar*^S in the second colophon.
' Reading den farnkhS zand shtr-h^>&5-t, bnt this is donbtfhL From this point the whole of the rest of this colophon, including the aphorisms, is also found in K«
Introduction. Xxxv
(This is followed by a long succession of aphorisms, and
the colophon winds up with some threats against those
who shall misuse the manuscript.)
As this colophon mentions only the ' last portion ' of the Dinkar//, and is appended to the text of Dk. III-IX, it is probable that the first portion of the work, Dk. I, II, had already become separated from the rest within 150 years of its revision and completion. And if MSAi-vindSji did not copy from the original manuscript of AtdrpAd, he must certainly have done so from a very early transcript.
^ Observe the use of the phrase ' P&rs! year ' in the third colophon and in the manuscript K (see p. zzxviii).
XXXVl PAHLAVI TEXTS.
by a third colophon, written by the actual writer of B, and preserved in copies transcribed from B since its arrival in
India.
This manuscript B, thus defective, afterwards came into
the possession of Mulia Firfiz, who was high-priest of the
Kadmi Parsis in Bombay; and, after his death in 1830,
it descended to his successor. In 1875 it belonged to
DastOr SohrAbji Rustamji, high-priest of the Kadmis,
through whose courtesy, and that of DastAr Dr. J4m4spji
Minochiharji, it was then lent to me long enough to
Introduction. Xxx Vi I
enable me to copy and collate two- thirds of Dk. Ill and
to collate Dk. IV-IX; and Dastd^r J&m^pji, afterwards,
kindly supplied me with a copy of the remainder of
Dk. III.
The manuscript has been bound in its defective state, and contains 322 folios, originally fourteen inches high and ten inches wide, written 20 to 22 lines to the page* When complete it appears to have consisted of 392 folios, all numbered in Persian words, but with several blunders, in- cluding one of fifly folios, so that the last folio was really numbered 442. Of the 70 folios not bound with the rest of the manuscript, fourteen were lying loose in the volume ; forty-three belonged to DastOr Rustamji Kaikobidji of Naus&ri, with a copy of which I was kindly supplied by Dastdr Dr. Peshotanji Behramji of Bombay, who also enabled me to collate it with the original folios; and seven folios were lent to me by Dasttkr Dr. Hoshangji J&m^pji of Poona, for the purpose of copying. The remaining six folios have not been discovered ; they com- prise the first folio of the manuscript, containing the com- mencement of Dk. Ill, which was probably lost before the manuscript arrived in India ; also one folio in Dk. VII, two in Dk. IX (see pp. 262, 270 of this volume), and the last two folios of the manuscript, containing the third colophon and final approvals (see p. xxxvi).
I am likewise much indebted to the kindness of Professor Kielhom, who gave me a modem copy of Dk. I V-IX (with the text in its defective state) which had been prepared at Poona, so that it was only necessary to collate this copy with the original text of the manuscript B. With the aid of all this liberal assistance I was enabled to obtain the whole text of the Dinkar^/, known to exist, in the course of a few months ; that it has since taken as much as sixteen years to find opportunities for translating and publishing rather more than one-fourth of its contents, will not surprise any one who is acquainted with the nature of the work that had to be done.
The only known manuscript, independent of B, that contains any portion of the Dinkar^, is the old codex K
XXXVlll PAHLAVI TEXTS.
Regarding the authorship of the summary account of the Nasks, contained in Dk. VIII, IX, it may be reasonably assumed, in default of any positive information, that the compiler was Atftfp^, son of H£m&/, the last editor of the Dinkar^. And, as nothing is said about any previous treatise being consulted, it may be safely supposed that he had access to the Avesta texts and Pahlavi versions of all the Nasks he describes, fully three centuries after the Muhammadan conquest of Persia. The only Nask he could not obtain was the Vaftag, and the Pahlavi version
Introduction. Xxxix
of the NlU/ar was also missing ; under which circumstances, the fully detailed accounts of these two Nasks, given in the Persian Riv&yats, must be viewed with suspicion, until better evidence of their authenticity has been discovered than is at present available.
The survival of so much of the sacred Zoroastrian litera- ture, during three centuries of Muhammadan rule, indicates that the final loss of nearly all this literature was not so directly attributable to the Arabs as the Parsis suppose. So long as a considerable number of the Persians adhered to their ancient religion, they were able to preserve its literature almost intact, even for centuries ; but when, through conversion and extermination, the Maer^/a-wor- shippers had become a mere remnant, and then fell under the more barbarous rule of the Tartars, they rapidly lost all their old literature that was not in daily religious use. And the loss may have been as much due to their neg- lecting the necessary copying of manuscripts, as to any destructiveness on the part of their conquerors ; because the durability of a manuscript written on paper seldom exceeds five or six centuries.
The statements of the Dinkar^/, about the classification and subdivisions of the Nasks, are corroborated and sup* plemented by those of ZS^/-sparam (see pp. 401-405). The division of all literature into three classes of knowledge, religious, worldly, and intermediate, is one that would naturally suggest itself to any classifier^, but the names employed (which are transcribed from the Avesta, and do not exactly correspond with these three meanings) must have originated at a period when the Avesta language was still spoken. That such a classification cannot be very strictly carried out in practice is already admitted in Dk. VIII, Chap. I, 13.
^ Professor Darmesteter has suggested, to me the very similar apportionment of the old Hebrew literature, mentioned in Jeremiah xviii. 18, thus :—* For the law shall not perish from the priest, nor counsel from the wise, nor the word from the prophet.* And in Ezekiel vii. a6, thus : — ' Then shall they seek a vision of the prophet : but the law shall perish from the priest, and counsel from the ancients.'
xl PAHLAVI TEXTS.
The further division of the literature into twenty-one books, seven in each of the three classes, is a much more artificial arrangement, and can, perhaps, be best explained as an attempt to make the twenty-one words of the Ahunavair serve the purpose of a reminder for enumerating the Nasks in their proper order. This arrangement was probably made at some period when the scattered Avesta literature was being collected and re-arranged, the Pahlavi versions being then supplied, and the present Pahlavi names of the Nasks appointed. This may possibly have been the work of * composition and preservation ' attributed to Atftr- p&jd, son of MAraspend, in Dk. VIII, Chap. I, 2a, when *the Nasks were enumerated' (see Dk. IV, 27, in p. 415), which occurred in the fourth century.
^ So long as it was preserved.
INTRODUCTION. xH
and N}ki^(im, may indicate some similar resemblance of contents ; especially as the contents of the ATitrad^ and Bak&n-yajt were so far from being strictly legal that these Nasks were placed in a sub-class by themselves, and the connection of the Spend with the Glthas appears to have been merely historical. The Persian Riv4yats place the Spend next after the Vajtag, thereby bringing the two imperfectly Gothic Nasks together, as well as the two im- perfectly legal ones; but then they also transpose the Ganabi-sar-ni^^a^/ and the Ht^sp&ram, for which there seems to be no justification.
With regard to the names of the Nasks, it is evident that several of the Persian names, used in the Riv^yats, are more or less irreconcileable with the Pahlavi names in the Dinkar^, and some others are improbable readings of the Pahlavi forms. In this translation the Pahlavi forms have been followed, as clearly more authentic than the Persian corruptions, and some few of the names have been read differently; while in other cases the most probable readings have been merely suggested in foot-notes, not on account of the Persian reading being justifiable, but because the evidence for the suggested reading is less complete than would be desirable.
In dealing with this account of the Nasks it is always necessary to remember that the compiler of the Dinkarrf relies entirely upon their Pahlavi versions, as he states distinctly in Dk. VIII, Chap. 1, 3 ; he occasionally mentions the Avesta texts, as in Chaps. VI, i, XII, i, and it is abundantly evident, to the practised translator, that Avesta phrases often underlie the Pahlavi passages which seem to be quoted at length from the original Nasks, especially in Dk. IX ; but, for some of the details mentioned, there may be no older authority than a Pahlavi commentary, and this should ever be borne in mind by the sceptical critic in search of anachronisms.
Owing to his complete reliance upon the Pahlavi versions, it is impossible to ascertain with certainty whether any particular statement, made by the compiler of the Dinkar^, was contained in the Avesta text; his summary, there-
xlii PAHLAVI TEXTS.
fore, throws little or no satisfactory light upon the origin of that text. A few of the details he mentions (such as those contained in Dk. VIII, Chaps. XIII, 17-20, XLIII, 24 and Dk. IX, Chaps. XXXII, 17, XXXIX, 13-16, LIII, 3) evidently refer to Sasanian times, and may be reason- ably supposed to have originated in the Pahlavi versions of those times. But vj^er prophecies of good or evil, such as are common in all religions at all times, may have often occurred in the Avesta texts themselves.
It is evident, however, that all the Nasks have accumu- lated around the G5itha centre of the St6rf-ya^, and that this Gdtha centre in the earliest Sasanian times was neither more nor less extensive than it is at present. The age of G«lthic composition had so long passed away in the time of the earliest Sasanian monarchs, that the sages whom they appointed to collect and re-arrange the sacred litera- ture, were unable to fully understand many of the stanzas they had to translate into Pahlavi, much less could they have added to their number. How far they may have been able to write ordinary Avesta text is more uncertain, but any such writing was probably confined to a few phrases for uniting the fragments of old Avesta which they dis- covered, or for interpolating opinions of their own. All such compositions, however, would have been hazardous, as forming no part of their duties, which seem to have been confined to the arrangement of the fragmentary Avesta texts, and their translation into Pahlavi with explanatory comments in that language. It appears from the traditional statements, mentioned in p. 415, that this work was com- pleted, and the Nasks were fully arranged, by AtiirpS^/, son of MAraspend, in the reign of Shahpfihar II (a. D. 309- 379) ; but the Pahlavi versions were certainly revised, and some further commentaries added, after the suppression of the heresy of Mazdak, as late as the reign of Kh&srdi I (A. D. 531-579).
That the Avesta texts themselves were not written, to
any great extent, in Sasanian times, is shown by the
quantity of Pahlavi commentary necessary to adapt them
to the altered circumstances of those times. The GAthic
