Part III

Dînâ-î Maînôg-î Khirad, Sikand-gûmânîk Vigâr, and the Sad Dar

Translated by E.W. West

Introduction

I. The Dînâ-î Maînôg-î Khirad

THE Pahlavi phrase Dînâ-î Maînôg-î Khirad, ‘Opinions of the Spirit of Wisdom,’ is a name applied to sixty-two enquiries, or series of enquiries, on subjects connected with the religion of the Mazda-worshippers, made by an anonymous wise man and answered by the Spirit of Wisdom. But, as this name is only found prefixed to a manuscript, written in A.D. 1569, in which the first part of the work is missing, it is doubtful whether it be the original name of the book, or not, although it is very suitable to the general character of the work.

Regarding the reading of this naive, here adopted, it must be observed that the correct pronunciation of the Pahlavi word maînôg, ‘spirit,’ is uncertain; the traditional reading is m ad ô n a d, which is a possible pronunciation of its letters, but is otherwise inexplicable; Haug proposed to read maînivad or mînavad, but, in that case, the word ought to end with d = t, or with nd; some of the present Dastûrs read mînôê, but this would be written minôêk in Pahlavi; the Pâzand writers have mainyô, but this is evidently an imitation of Av. mainyavô, and does not correspond with the Pahlavi letters. As the word is manû or minô in the Sasanian inscriptions, and mînû in Persian, to which words a final k would be added in Pahlavi, it seems probable that the final letter of the Pahlavi word is not d or ê, but g, a corruption of k, and that we ought to read mînôg or maînôg. At the same time it should be noticed that a very old copy of the Pahlavi Farhang, in the library of Dastûr Jâmâspji Minochiharji in Bombay, has the word written with an extra medial stroke, so that it might be read mînavand, as required by Haug’s hypothesis, although this copy of the Farhang gives madonend as the traditional reading.

The subjects discussed by the Spirit of Wisdom are of a very miscellaneous character, and their discussion is evidently intended to furnish an outline of the tenets, legends, and morality of the religion with which they deal; but it forms by no means a complete, or systematic, treatise on these subjects, and it is remarkably silent with regard to all details of religious rites and ceremonies, which are only occasionally. mentioned. This silence may, perhaps, be due to the fact that the author was a layman, as seems clear from the account he gives of his doubts and enquiries in Chap. I, 14-56. Any incompleteness of the treatise may also be explained by the apparent loss of the latter end of the work, as the sixty-second reply (Chap. LXIII) terminates the extant text of the treatise abruptly, and without any trace of peroration.

By the Spirit of Wisdom the author means the innate wisdom of Aûharmazd (Chap. LVII, 4), the âsna khratu of Yas. XXII, 29, XXV, 18, through which the spiritual and worldly creations were produced (Chaps. I, 49, 51, LVII, 5). It was originally created by Aûharmazd (Chap. VIII, 3, 8), and is superior to the archangels (Chap. I, 53); it can appear in a personal form, and undertake to be an instructor (Chap. I, 57, 60, 61); and it can likewise be used as a defence (Chap. XLIII, 6).

With regard to the author of this treatise, and the age. in which he lived, we have no further information than can be gathered from the contents of the book itself. The author was evidently a devoted Mazda-worshipper, and probably a layman, as has been already remarked, but he has given us no further hints about himself. Whether he wrote before or after the Arab conquest of Persia is doubtful. There are only two passages that might be strained into allusions to Muhammadanism: one in Chap. I, 18, which alludes to some heterodox religion injuring the property of the orthodox faith, but the author has just been talking of many sects, and the grievance here mentioned is much too common to be considered as applicable only to the Arabs; the other passage is Chap. XVI, 37-48, which describes the advantages of ‘the moderate drinking of wine,’ and might be supposed to be written in indirect opposition to the Muhammadan prohibition of such indulgence. In either case the allusion is certainly far too obscure to form a fair basis for argument. On the other hand, Chap. XIII, 13, 14, speaks of the sovereignty of Vistâsp existing in connection with the most powerful sect or form of devotion, which statement might be strained to imply that the government was still orthodox; and the definitions of good and bad government in Chap. XV, 12-39 could hardly have been written after the Arab conquest. The allusion to the continued conflict of the Arûmans and Tûrânians with the Irânians, in Chap. XXI, 23-26, may possibly refer to some troublesome wars carried on by the Greeks and Turks against the Persians in the time of the author, and the late Dr. A. D. Mordtmann has suggested A.D. 580-590 as a probable period for such remarks, but, here again, the allusion is too obscure to be relied on.

Very few of the author’s quotations can be identified, but this is no argument for a greater age than eight or ten centuries, as we know, from passages quoted in the Shâyast Lâ-shâyast, Dâdistân-î Dînîk, and other works, that some of the lost Nasks must have been still extant as recently as that. The Avesta is quoted only twice by name, in Chaps. I, 27, XVI, 15; the former passage has not been identified, but the latter may perhaps be from the Pâzag Nask. Several quotations, however, are made from the dînô or ‘revelation,’ a term which, when it refers to writings, is often applied by Pahlavi writers to the Avesta only. Of these passages Chap. XLIV, 18-23 is from the Vendîdâd, Chap. XXI, 24-26 may be from the Kidrast Nask, and six other quotations have not been identified. In other cases the quotations are merely prefaced by the phrase ‘it is declared.’ And of these the passage in Chap. LVII, 24-28 appears to be derived from the Vendîdâd, and that in Chap. II, 155, 156 from the so-called Hâdôkht Nask, while eight other passages are unidentified. In this last class the quotations seem to be rather paraphrases than accurate translations of the original texts.

Of the original Pahlavi text of the Dînâ-î Maînôg-î Khirad only two manuscripts are yet known to exist; one of these (K43) is contained in No. 43 of the Irânian manuscripts in the University Library at Kopenhagen, and the other (TD2) belongs to Mr. Tehmuras Dinshawji Anklesaria of Bombay.

The manuscript K43 is a small quarto volume of 178 folios, of which the Dina occupies fols. 2-37, written fifteen lines to the page. The first and second folios also contain the conclusion of the larger Bundahis, of which the first 129 folios are missing from this codex, as described in SBE, vol. v, introd. pp. xxxix-xli. And the latter part of the codex contains about one-fifth of the Dînkard, in several detached fragments, and four-fifths of the Bahman Yast. This manuscript was brought from Persia by the late Professor Westergaard in 1843, and the Pahlavi text of the Dînâ, which it contains, was published in facsimile by Andreas in 1882.

In this codex the text of the Dînâ-î Maînôg-î Khirad begins in the middle of Chap. I, 28; but, as the copyist has prefixed an introductory heading to this imperfect text, it is evident that he, or some predecessor of his, must have copied the work, in this imperfect state, from some manuscript whose first folio had been lost. Besides this deficiency, ten folios of the text have been lost from this particular codex; nine of these were occupied by Chaps. XIV, 1-XXVII, 49, and the tenth contained Chaps. XXXIX, 31-XL, 17. At the end of the work, Chap. LXIII is followed by a colophon to the following effect: Completed in peace and pleasure and joy on the day Shatvaîrô of the month Âvân of the year 938 of Yazdakard, king of kings, [26th May 1569]. I, Mitrô-âpân Anôshak-rûbân Rûstâm Shatrô-îyâr, wrote it for my own possession. From the copy of Dastûr Gadman-pîrûg Aspendiyâr Gadman-pîrûg, and that from the copy of Dastûr Shatrô-aîyyâr Vêgan Khûsrôîshah, and that, as regards these several sayings, was written from the copy of the heavenly-destined Mâh-vindâd 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 acknowledgment 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 Pâzand-Sanskrit version of Nêryôsang, 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 Pâzand 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 khvêsinîdârîh, barâgûmân for avîgûmân, hamê for hamâî, avas for aûbas, lâ and mâ for al, denman instead of hanâ for Pâz. e, the constant use of the adjective suffix -îk for the abstract suffix -îh, and the frequent omission of the final k in such words as dânâk, avistâk. It can be seen at once, by any one really acquainted with Pahlavi, that a text of this description is merely a modern transliteration of the Pâzand version by some one whose knowledge of Pahlavi was rather limited and artificial.

Most of the Indian manuscripts of this work contain only the Pâzand version written in short sentences, alternating with a word-for-word Sanskrit translation of each sentence; the Sanskrit being written upside down, for the sake of forming a continuous line with the reversely-written Avesta characters of the Pâzand. This Pâzand-Sanskrit version of the Mainyô-i Khard (as it is called in Pâzand) was compiled by Nêryôsang, son of Dhaval, a Parsi priest who is supposed to have lived some time in the fifteenth century, and evidently possessed a very good knowledge of Pahlavi, though not sufficient to avoid some few mistakes, especially in reading foreign names. His authorship is attested by a Sanskrit introduction, prefixed to most manuscripts of this version, to the following effect: — ‘Through the name and almighty power and assistance of the lord Ahura-mazda, the greatly wise, may the achievement be auspicious, and be the progress and success of the good Mazda-worshipping religion, and energy in body and long life for all the good and right-minded. This Pahlavi heavenly wisdom, called the Mainyô-i Khard, is translated by me, Nêryôsang son of Dhaval, from the Pahlavi language into the Sanskrit language, and written from the difficult Parsi letters with the Avesta letters, for the joyful understanding of the good listeners to instruction, the true-minded. Salutation to the good, the pure-thinking, the true-speaking, the just-acting.’

Of this Pâzand-Sanskrit version the oldest manuscript that has been examined is L19, No. 19 of the Avesta and Pahlavi manuscripts in the India Office Library in London, one of the manuscripts brought from India by Dr. Samuel Guise who was head surgeon of the general hospital at Surat from 1788 to 1795, and obtained several manuscripts from the widow of Dastûr Dârâbji, the instructor of Anquetil Duperron. It is a small octavo volume, containing 148 folios of old Indian paper, of which the first 132 are occupied by the Pâzand-Sanskrit Mainyô-i Khard, written fifteen lines to the page. At the beginning of the text the folio containing Nêryôsang’s Sanskrit introduction (described above) has been lost, but the text itself is complete. At the end of the work is a Pâzand-Sanskrit postscript which may be reasonably attributed to Nêryôsang himself, and car be translated as follows: — ‘Completed for the peace and pleasure, happiness and dominion of all the good who are virtuous. To him for whom it is written may it be well-resulting and well-omened, and, after a hundred and fifty years, may he be a transmitter of it to his own religious children’s children, through the will of the sacred beings. Of whomsoever the best ability is not wisdom, that best ability of his is even then owing to it. Wisdom which is without learning is poor, and learning which is without wisdom is helpless.’ After this postscript a Pahlavi colophon has been copied from some older manuscript to the following effect — ‘Completed in peace, pleasure, and joy, and ended; written by me, a servant of the religion, the priest Shatrô-aîyyâr, contemporary (?) of Nêryôsang.’ And this is followed by a colophon in very corrupt Sanskrit, which states that this manuscript was completed, in the district of Nâga-mandala, at a date corresponding to Friday, the 19th October 1520, by the teacher Mihrvân, son of Mahyâr and grandson of Padama, for the priest Bahrâm, son of Pâlhan. This manuscript of the Pâzand text is, therefore, nearly 49 years older than that of the original Pahlavi text (K43) upon which the present translation is based. It corresponds very closely with that Pahlavi text, and where it differs the variation is nearly always due to some mistake, or attempt at improvement, on the part of Nêryôsang. It must, however, be acknowledged that very few translators adhere so closely to their original texts as this learned Parsi priest has done to his.

Other manuscripts of the Pâzand-Sanskrit version are PA10 and PB6. The former is No. 10 of the Anquetil Collection in the National Library at Paris, and was brought from Surat by Anquetil Duperron in 1761. It is an octavo volume, in which the Mainyô-i Khard occupies the first 211 folios, and commences with Nêryôsang’s Sanskrit introduction, translated above, but does not contain the postscript. The date of its colophon appears to correspond to the 7th December 1649, new style. The latter manuscript, PB6, is No. 6 of the Burnouf Collection in the same library, and is probably about a century old.

The Pâzand version also occurs alternating with a Gugarâti translation in K23, No. 23 of the Irânian manuscripts in the University Library at Kopenhagen. It is an octavo volume of 168 folios of glazed Indian paper, of which the first 162 contain the Pâzand-Gugarâti text, written fifteen lines to the page, and the remaining six folios contain an index stating the contents of each chapter. A colophon, at the end of the text, has a date corresponding to the 25th August 1663, new style; and another, at the end of the index, states that the manuscript was written by the priest Yazad-yâr, son of Vikaji, of Sangân, and finished at a date corresponding to the 17th October of the same year.

In another class of Pâzand manuscripts of the Mainyô-i Khard the Pâzand text is written in the Perso-Arabic character, and accompanied by a Persian translation, forming what may be conveniently termed a Pârsî-Persian version. One example of this version is contained in MH7, No. 7 of the Haug Collection in the State Library at Munich, of which it occupies the first 70 folios, written fifteen lines to the page. Most of the Persian translation is written in sentences alternating with those of the Pârsî text, in which case the translation is merely a paraphrase of the Pârsî; but some of the translation is interlined, and this is much more literal, each Parsi word having its Persian equivalent written below it. This manuscript contains several other texts, and from two colophons, one near the middle, and the other near the end of the volume, it appears that it was written by Dârâshâh, son of Mihrbânji, and the first half of the volume was completed at a date corresponding to Wednesday the 9th August 1809.

Another example of the Pârsî-Persian version is found in No. 2769 of the Persian manuscripts in the India Office Library in London, in which manuscript it occupies 75 folios, written eleven lines to the page, and is not dated, though probably written early this century. In this copy the Parsi text is tolerably complete, but long passages of the Persian translation are omitted; when given, the Persian is usually identical with that in MH7, though some instances of Independent translation occur.

In addition to the Pahlavi, Pâzand, Sanskrit, Gugarâti, Pârsî, and Persian texts of the prose Dînâ-î Maînôg-î Khirad, the popularity of the work is further evinced by the existence of two versions in Persian verse. One of these was described by Professor Sachau in the Journal of the Royal Asiatic Society, new series, vol. iv, pp. 229-283, from a manuscript in the library of that Society in London, written probably near the end of last century. The author of this metrical Persian paraphrase appears to have been a native of Râvar in Sindh, named Marzubân, who composed it from a Pârsî version of the original text, bequeathed to him by his teacher while he was studying the old traditions at Yazd; and the date of his composition seems to have been A.D. 1612. His verses contain only fifty-four questions and answers, but these contain the substance of the greater part of the Mînôkhirad, as the work is called in Persian, with some few additions from other sources.

A copy of the other metrical Persian Mînôkhirad occupied fols. 527-550 in the second volume of B29, a two-volume quarto Rivâyat, No. 29 in the Bombay University Library. It is doubtful whether the original number of folios were twenty-four or twenty-six, but only twenty-two now remain. These contain 497 couplets of introductory matter, 1060 representing the text of the work, and 190 of epilogue; and from 160 to 330 further couplets of the text are missing. According to statements in the introduction and epilogue the verses appear to have been composed, from Nêryôsang’s Pâzand-Sanskrit text, by the priest Hormazyâr and his son Dârâb, the latter being the actual writer, and the former being a son of Farâmruz, son of Qavâmu-d-dîn, son of Kaî-Qubâd, son of Hamkârapadam of Sangân, of the family of the priest Nêryôsang Dhaval. The work was commenced on the 7th November 1676, new style, and completed in thirty-five days; and the copy in B29 was finished on the 21st November 1679, new style. The order of the subjects discussed in this metrical version differs, in some respects, from that followed in the prose texts, and the 1060 couplets of extant text represent only forty chapters of the work, though several of the others were, no doubt, represented in the missing couplets. Another copy of this later metrical version appears to exist in pp. 231-248 of No. 12 of Anquetil’s Collection in the National Library at Paris.

Of the Pâzand text of the Mainyô-i Khard, Chaps. LVII, XXVII, LXII, I, 51-61, VII, 9-12 have been published, with German translations, by Professor Spiegel, in his ‘Grammatik der pârsi Sprache,’ pp. 128-155, 161-173, 185, 186, 188, 189. He has also published German translations of Chaps. II, 110-193, VIII, XXXVII, XLII in his ‘Traditionelle Literatur der Parsen,’ pp. 138-144, 147-150. And the complete Pâzand-Sanskrit texts, with an English translation, Pâzand glossary and grammar, were published by the present translator in 1871. Since that date the original Pahlavi text of the Dînâ-î Maînôg-î Khirad has been discovered, from which the present translation has been made.

In connection with this account of the various versions of the ‘Opinions of the Spirit of Wisdom,’ it should be noticed that an abridgement of the work also exists in Persian prose, and is called the ‘Other Mînôkhirad. ‘A copy of this abridgement is contained in fols. 71-78 of MH7 (described above), and consists of a very free Persian translation of the Pâzand texts of Chaps. I, 14-II, 64, III-VII, XIV, XV, XXI, XXV, followed by a variety of short statements about thankfulness towards the sacred beings, the supreme heaven, male and female angels and demons, wealth and poverty, &c., and concluding with the names of the first sovereigns of the world, the descent of mankind from Gâyômard, and of the 292 species of animals from the primeval ox. Another copy of this abridgement appears to be contained in fols. 80-84 of No. 15 of Anquetil’s Collection in the National Library at Paris.

2. The Sikand-gûmânîk Vigâr.

The term Sikand-gûmânîk Vigâr, ‘doubt-dispelling explanation,’ is the Pahlavi name applied to a controversial work by its author. The chief object of the work is to prove the correctness of the fundamental doctrine of the Mazda-worshipping religion, that good and evil do not proceed from the same source, and to show that other religions, while professing to believe in the unity of creation, can only account for the origin of evil, either by degrading the character of the sacred being, or by attributing evil to a corrupting influence which is really a second being. In other words, the author’s object is to show that all people, who believe in an all-good and omnipotent creator, must logically admit the existence of an independent origin of evil, whatever they may say to the contrary. In the course of his arguments, he naturally finds it easier to attack the inconsistencies of other beliefs than to defend his own, and much of his attention is, therefore, given to pointing out apparent inconsistencies and seemingly delusive statements in the scriptures of the Muhammadans, Jews, Christians, and Mânîchaeans.

The author’s name was Mardân-farukh, son of Aûharmazd-dâd (Chap. I, 35), and his account of his enquiries (§§ 36, 37) bears much resemblance to what is said of the wise man’s proceedings in Mkh. I, 34-36. He determines to write a treatise for removing religious doubt, and calls it the Sikand-gûmânîk Vigâr (Chap. I, 38). He is also careful in stating that he has selected many of his facts and arguments from older writings, such as those of Âtûr-pâdîyâvand, which he had found in the Dînkard compiled by Âtûr-frôbag, son of Farukh-zâd. In this statement he must be referring to the first two books of the Dînkard, which have not yet been discovered, as the other seven books, which are extant, do not contain the matters to which he alludes. He also mentions the Rôshan manuscript compiled by Rôshan, son of Âtûr-frôbag, a writer who is often quoted in the Pahlavi commentaries on the Avesta. And he begins his religious discussion by replying to some difficulties that had been suggested to him, in a friendly manner, for solution by Mitrô-aîyyâr, son of Mahmâd, of Ispahân.

His allusions to Muhammadanism are of a very guarded character, though sufficiently clear to leave no doubt as to the religion he means. Like all Pahlavi writers, he never mentions that religion by name, but when, in the position of a Zoroastrian in Persia, he states that he did not admire the religion that was then in supremacy, there can be little doubt that he refers to Muhammadanism. And any such doubt would be dispelled, not only by such vague references to passages in the Qur’ân as occur in Chap. XI, 4, 5, 269-271, but also by the distinct quotation of a striking legend, from the same source, regarding the fallen angel in §§ 52-60, 248 of the same chapter, and by the use of the term Mûtazalîk (Ar. mu’htazil) with reference to a certain sect in § 280.

With regard, therefore, to the age of the Sikand-gûmânîk Vigâr, we may be quite certain that it was written long after the Arab conquest of Persia; and from the names mentioned by the author, as stated above, it is evident that he lived after the time of Rôshan, son of Âtûr-frôbag, son of Farukh-zâd. Now, according to a Pahlavi tale, the accursed Abâlis, the Zandîk, had a religious disputation with Âtûr-frôbag, son of Farukh-zâd, in the presence of the Khalîfah Al-Mâmûn who reigned A.D. 813-833; it is, therefore, hardly possible that Rôshan, son of Âtûr-frôbag, could have written his commentary before the middle of the ninth century. We also know, from the last chapter of the third book of the Dînkard, that Âtûr-frôbag was not the last editor of that work, but was succeeded by his son Zaratûst, and, later still, by Âtûrpâd, son of Hêmîd, who appears to have given the book its final revision. Of Âtûrpâd’s work the author of the Sikand-gûmânîk Vigâr does not speak, and it is, therefore, reasonable to suppose that it had not been completed in his time. But, according to Bd. XXXIII, 10, 11, this Âtûrpâd was a contemporary of Zâd-sparam who was living in A.D. 881, and his revision of the Dînkard was, therefore, probably in progress by the end of the ninth century. From these facts we may conclude that the Sikand-gûmânîk Vigâr was written after the middle, but before the end, of the ninth century; unless we were to suppose that, although its author consulted only the first two books of the Dînkard (as mentioned above), the remaining seven books may have existed as a separate work unknown to him. Considering, however, that Âtûrpâd, son of Hêmîd, was so important a personage as ‘the leader of those of the good religion’ of his time, this supposition would not be very probable.

There is probably nothing new to defenders of Christianity in Mardân-farukh’s attacks upon the apparent inconsistencies of their scriptures, with regard to the origin of evil and the existence of unity in trinity, subjects that are more usually admitted without investigation than seriously discussed. This is not, however, the mode in which such subjects are likely to be treated by outsiders, and missionaries will no doubt find among Mardân-farukh’s arguments many that they must become accustomed to hear from educated men of other faiths. And, if they engage in controversy, as they ought to do, they must expect to hear them stated in less considerate language than this author uses.

An interesting question, for any one who possesses special information on the subject, would be to ascertain from what version of the Old and New Testaments Mardân-farukh drew his quotations. There seems every probability that his translation of the passages, though it may not be altogether literal, is yet sufficiently so to admit of the particular version being identified, if its peculiarities of wording were carefully considered. The peculiar Pahlavi spelling of the name Isaac in Chap. XIV, 42, as deduced from its corruption in Pâzand, points to a Syriac version of one of the legendary works consulted by the author.

In his discussion of the tenets of the Mânîchaeans Mardân-farukh is dealing with a subject that is far less known than the other faiths he attacks, and the information he gives may be valuable. Unfortunately the latter part of this discussion is missing, although the loss of text is probably not very extensive.

The original Pahlavi text of the Sikand-gûmânîk Vigâr has not yet been discovered, although there are several existing copies of a Pahlavi version of the earlier part of the work, which are evidently reproductions from the Pâzand text. These pseudo-Pahlavi manuscripts usually end with the fifth chapter, and are certainly superior to the similar reproductions of the Dînâ-î Maînôg-î Khirad, represented by K22 (see p. xix). Yet they generally use the adjective suffix-îk for the abstract suffix-îh, because both these suffixes become-î in Pâzand; they often have kabed, ‘much,’ for afas, and ‘by him,’ when the Pâzand has vas by mistake for vas; they also substitute the Pâzand misreading ainâ for the true Pahlavi adînas; besides adopting other occasional miswritings for which the Pâzand version alone is responsible. Such manuscripts could be of no critical value, unless they had descended from some family of Pâzand manuscripts which had left no surviving representatives in Pâzand, and this does not appear to be the case.

A specimen of these Pahlavi reproductions is contained in the last 36 folios of L15, No. 15 of the Avesta and Pahlavi manuscripts in the India Office Library in London. It commences with the words ‘all the angels’ in Chap. I, 4, and ends with Chap. V, 71; the handwriting being the same as that in L26, a manuscript that contains a date corresponding to A.D. 1737.

In fols. 9-16 of BM. No. 22,378 of the Additional Oriental manuscripts in the British Museum Library, there is a modern fragment of this reproduced Pahlavi text, interlined with a transliteration in the Persian character, and alternating with a Persian paraphrase. This fragment contains only Chap. I, 1-31.

The reproduced Pahlavi text also occurs, in parallel columns with the usual Pâzand and Sanskrit versions and a Persian paraphrase, in R, an imperfect polyglot manuscript given to the late Mr. J. Romer by a Dastûr in Surat. Of this foolscap-folio manuscript Mr. Romer sent pp. 16-31 (with the first fifteen pages of a Pahlavi-Persian Bundahis) to the late Professor M. J. Müller, through Mr. Poley; he also sent pp. 32-63, 82-93 to the late Professor H. H. Wilson on 3rd December 1836, who afterwards transferred them to Professor Max Müller; and he gave pp. 64-81, 99-143 to the late Mr. Norris. The first of these fragments, together with that of the Bundahis now constitute No. 10 of the Müller Collection in the State Library at Munich; the next two fragments were presented to the India Office Library, and the two last mentioned were acquired by it, in 1876. It is most probable that the first fifteen pages of this polyglot manuscript were not given to Mr. Romer, but the first fifteen pages of the Bundahis were substituted for them. The portion extant (pp. 16-143) contains all four versions of Chaps. I, 28-V, 57, with the Sanskrit and Persian versions of Chap. I, 25-27, and the Pahlavi and Pâzand versions of Chap. V, 58-62; and the latter two versions are everywhere interlined with a transliteration in Persian characters. This manuscript is modern and of no particular critical value; but, as the combination of the four versions is rare, if not unique, it would be very desirable to discover the rest of the manuscript.

In another manuscript, No. 18 of the Anquetil Collection in the National Library at Paris, the reproduced Pahlavi text has the usual Pâzand version written above it. This manuscript, which is in the form of a roll, begins at the same point as L15 and ends with Chap. V, 95, which is said to be the usual extent of other manuscripts of this class in India. A copy of this manuscript is No. 23 of the Müller Collection in the State Library at Munich.

An extension of the same reproduced Pahlavi text, with the Pâzand version written above it, and alternating with the Sanskrit version, is contained in K28, No. 28 of the Irânian manuscripts in the University Library at Kopenhagen. It is an imperfect octavo manuscript, of which only 66 folios remain, written eleven lines to the page, and, in its present state, it is undated, but seems to be fully 150 years old. The portions of the text that it still contains are only Chaps. I, 1-II, 8; III, 1-25; III, 36-IV, 106; VIII, 103-IX, 16; IX, 30-X, 13; X, 71-XI, 28; XI, 55-61; so that more than half the text that ought to be included within its extreme limits is missing; but its original extent, within the same limits, was more than double the usual length of the reproduced Pahlavi text, as stated above. In this particular, of unusual length, only one other manuscript of that text seems to be known in India that resembles it, in addition to the imperfect copy next described. K28 contains Nêryôsang’s usual Sanskrit introduction (see p. xxxiii), and differs from the oldest Pâzand manuscript AK in only two or three instances, and these variations can be explained as corrections made on the authority of the Sanskrit version.

An imperfect and modern copy of the Pahlavi-Pâzand-Sanskrit texts is also contained in twenty-two folios prefixed to AK (described below). This copy commences with Nêryôsang’s Sanskrit introduction, and includes only Chaps. I, 1-IV, 100 and X, 71-XI, 47. Its writer has intended to give the three versions in successive sentences, but, after Chap. I, 23, the Pâzand and Sanskrit sentences are less and less frequently written, till they cease altogether after I, 43, with the exception of one or two isolated sections. In several cases he has also substituted the correct abstract suffix -îh for the usual incorrect-îk, but this correction is generally confined to abstract nouns in common use.

As none of these Pahlavi manuscripts can be considered otherwise than as reproductions from the Pâzand, it is to the Pâzand-Sanskrit version of Nêryôsang that we must still look for the nearest approach to the original text of the work. It is in this version, too, that we find the greatest extent of text still extant, although the Sikand-gûmânîk Vigâr seems to possess the peculiarity of wearying out all its copyists at some point or other, so that not only is there no complete copy of the work known, but also nearly every copyist has stopped his work at a different place.

The oldest known manuscript of the Pâzand-Sanskrit version belongs to Dastûr Hôshangji Jâmâspji of Poona, and is called AK, because it is supposed to have been written by Âsadîn, son of Kâkâ. In its present state this manuscript consists of seventy-seven small quarto folios of very old, discoloured, Indian paper, written sixteen lines to the page, and containing the Pâzand version in short sentences, alternating with a word-for-word Sanskrit translation of each sentence; the Sanskrit being written upside down, for the sake of forming a continuous line with the reversely-written Avesta characters of the Pâzand. From other manuscripts it is known that this Pâzand-Sanskrit version was compiled by Nêryôsang, son of Dhaval, but in this manuscript his usual Sanskrit introduction is lost with the first three folios of the text, and the existing seventy-seven folios contain only Chaps. I, 16-XI, 145. As this extends only one folio beyond the middle of the whole of the text that is extant, it is supposed that this old manuscript was divided into two nearly equal moieties on the occasion of some division of property, of which the earlier moiety has been preserved, and the later one either lost, or destroyed, or buried in some inaccessible library.

In consequence of the imperfect state of this manuscript it bears no date, but an old Sanskrit colophon has been copied by the writer of JE (one of the more modern manuscripts that are evidently derived from AK through one or more intermediate copies), and this may be fairly assumed to be the colophon of AK. This colophon may be translated as follows: — ‘In the Samvat year 1625, in the current Sâka year 1490, on the present day (?), the fourth day Shahrîvar of the eleventh month Bahman, in the district of Nâga-mandala, in the royal reign of king Sultân Muthaffar-shâh, the book named Sikand-gûmânîk Vigâr is written, for the use of Amalshâh Kangashâh, by the priest Âsadîn, son of the priest Kâkâ. May it become auspicious! may it be beneficial!’

The date indicated by this colophon seems to correspond to the 23rd September 1568, but it may, of course, be doubted whether it originally belonged to AK, because the text to which it is appended in JE is incomplete. If it were attached to AK, the text in that manuscript must either have been originally incomplete, or some of the later folios must have been lost, while the last one, containing the colophon, was still preserved. If it did not belong to AK, it must have belonged to some later manuscript, because there is no doubt that JE has descended from AK, and could not, therefore, contain the colophon of an older manuscript than AK, unless it had been written in AK itself, or obtained in an irregular manner from some unrecorded source. For these reasons there seems little doubt that AK was written either in 1568, or earlier; and the general appearance of its folios favours this assumption. So far as it extends this is the best manuscript of the Sikand-gûmânîk Vigâr that is known to exist, and the present translation has, therefore, been based upon its texts, which are, no doubt, very nearly in the same state as when edited by Nêryôsang; the Sanskrit version, especially, is far more correct than in the later copies. Many of the Pâzand sections in Chaps. V-VIII are written in Pahlavi only, or in Pahlavi with the Pâzand written above it; but, in all cases, this Pahlavi is as corrupt as that of the reproduced Pahlavi manuscripts.

The most complete manuscripts of the Pâzand-Sanskrit version are JJ and JE, of which JJ is the oldest and best, but it has not yet been thoroughly examined. It is a small quarto volume of 182 folios of Indian paper, written fifteen to seventeen lines to the page, and belongs to Dastûr Khurshêdji Jamshêdji of Nausârî. From certain blunders and peculiarities, which its writer has copied, it is certain that this manuscript has descended from AK, and, also, that it has derived a few variations from some other source. Its Sanskrit text is not written inverted, as it is in AK, and it commences with Nêryôsang’s usual Sanskrit introduction, as translated in p. xx, but with the clause containing the names altered to the following effect: This book, named Sikand-gûmânîk Vigâr, is translated by me, Nêryôsang son of Dhaval, from the Pahlavi language into the Sanskrit language, and written from the difficult Parsi letters with the Avesta letters, for the joyful understanding of the good listeners to instruction, the true-minded.’ The texts in JJ are of the same extent as the translation in this volume, and are followed by a colophon in Persian, Sanskrit, and imperfect Pahlavi, which states that the manuscript was written by Dastûr Jamshêd, son of Jâmâsp, son of Âsâ, son of Frêdûn, inhabitants of Nausârî, and completed on the day Srôsh of the month Vohûman, A.Y. 1137 (corresponding to the 28th August 1768).

The other manuscript, JE, which is as complete as the translation in this volume, is a foolscap-folio volume of 132 folios, written eighteen lines to the page, and belongs to Dastûr Hôshangji Jâmâspji of Poona. It corresponds very closely with JJ, but its Sanskrit (which is not written inverted) is rather more corrupt; and it contains the same indications of descent from AK as that manuscript does, with the same variations derived from some other source. It commences with Nêryôsang’s usual Sanskrit introduction, and at the end of the text it has the old Sanskrit colophon translated above, and supposed to belong to AK. And this is followed by a Persian colophon, written on the day Hôrmazd of the month Bahman, A.Y. 1211 (corresponding to the 26th July 1842), and stating that this manuscript was copied from that of Âsadîn, son of Kâkâ, in Bombay, by Jamshêd, son of Edalji, son of Bahmanji, son of the writer of JJ. From this it might be too hastily assumed that the old manuscript AK was still complete as recently as 1842; but, if such were the case, it would be difficult to understand why Dastûr Hôshangji could learn nothing about it’s missing moiety some twenty-five years afterwards, when he made searching enquiries on the subject; and it would be still more difficult to explain the variations in J E, already mentioned as derived from some other source than AK. It is more probable that the writer of JE found the old colophon of AK copied at the end of a more recent manuscript, which led him to believe that the latter was written by Âsadîn, son of Kâkâ.

That the first folio of AK had already been lost, considerably more than a century ago, appears from PB3, No. 3 of the Burnouf Collection in the National Library at Paris, which was evidently copied from a copy of AK, and is certainly more than a century old, judging from the general appearance of the paper on which it is written. This manuscript, which was given to Burnouf by Mr. Mânekji Khurshêdji of Bombay, is a small octavo volume of 125 folios of Indian paper, written twelve to sixteen lines to the page, and contains the Pâzand-Sanskrit text of Chaps. I, 5-53, and II, 5-X, 66: the Sanskrit being written upside down, as in AK. The loss of Nêryôsang’s Sanskrit introduction and Chap. I, 1-4 of the text indicates that the first folio of AK was already missing when the original of PB3 was copied, and several lacunae in the earlier folios, which have been filled up in red ink from some other source, indicate the torn condition of the earlier folios of AK. The loss of Chaps. I, 54-II, 4 is due to two folios being absent between folios 11 and 12 of PB3; and after Chap. X, 66 all further folios have been lost. In some sections in Chaps. VI and VIII, where the Pâzand text is written above its Pahlavi equivalent in AK, much confusion has been occasioned in PB3 by reading the Pâzand and Pahlavi versions as two successive lines of text; and it is evident that this confusion originated in some manuscript intermediate between AK and PB3, though it has been increased by further blundering on the part of the writer of PB3 itself.

The Pâzand version of Nêryôsang also occurs in short sentences alternating with a Gugarâti translation in MH19, No. 19 of the Haug Collection in the State Library at Munich. This manuscript, which was given to Haug by Dastûr Kaî-Khusrô at Surat in 1864, is a small quarto of 124 folios of old Indian paper, of which the first Ira folios contain the Pâzand-Gugarâti version of Chaps. I, 1-XI, 201, written thirteen to nineteen lines to the page. Towards the latter end of the manuscript blank spaces are left for the Gugarâti version of many of the sections; and several of the passages that are written only in Pahlavi in AK are similarly written in MH19. From this and other peculiarities it is evident that MH19 has descended from AK, but probably through some intermediate manuscript that must have been written when AK was more complete than it is now. Judging from the appearance of the paper of MH19 it can hardly be less than 150 years old, but it contains no date or colophon of any description.

Another manuscript, which contains a large portion of the Pâzand version of Nêryôsang, without his Sanskrit translation, is L23, No. 23 in the India Office Library in London. It is an octavo volume of eighty folios of Indian paper, written ten to twelve lines to the page, in the same handwriting as L15 and L26 (see p. xxix), which last manuscript contains a date corresponding to A.D. 1737. L23 contains the Pâzand text of Chaps. I, 34-VIII, 23, and many of the passages written in Pahlavi in AK are similarly written in L23, which indicates the descent of the latter manuscript from the former; an indication which is confirmed by the repetition of other peculiarities.

From this account of all the manuscripts of the Sikand-gûmânîk Vigâr, that have been examined by the translator, it appears probable that no manuscript independent of AK has yet been discovered. The few variations which indicate another source can easily be explained as emendations by some later copyist, who had noticed, or imagined, some deficiencies in the text of that manuscript.

The Sikand-gûmânîk Vigâr has not been hitherto translated into any European language, but an edition of its Pâzand and Pahlavi texts was prepared by Dastûr Hôshangji about fifteen years ago, and arrangements have been made for the publication of these texts, with the Sanskrit version, at an early date.

3. The Sad Dar.

As its name implies the Sad Dar is a treatise on ‘a hundred subjects’ connected with the Zoroastrian religion. The word dar, literally ‘door, or gate,’ being also applied to the ‘chapters’ of a book, and to the ‘matters, or subjects,’ of which it treats. This work is not a Pahlavi text, being written in Persian with an admixture of about four per cent. of Arabic words; it is, however, more quoted than any other work by the Parsi compilers of the Persian Rivâyats, or religious ‘traditions,’ in the seventeenth century. In one of its recensions it is also found written in Avesta characters, and the Avesta-Persian sentences alternate with an old Gugarâti translation, in imitation of the Pâzand-Sanskrit versions of Pahlavi texts compiled by Nêryôsang. In consideration of the existence of this pseudo-Pâzand recension, together with the general acceptance of the work as an important authority, and its being a convenient summary of many of the religious customs handed down by Pahlavi writers, this work may be offered as a suitable appendix to the true Pahlavi texts, connecting them with the Persian writings that are too modern to be accepted as authorities in religious matters.

The Sad Dar NaTHr, or prose Sad Dar, which is here translated, appears to be first mentioned in the introduction to the Sad Dar-i Bahr-i Tavîl, or long-metre Sad Dar, in which the versifier states that the prose Sad Dar was compiled by three celebrated high-priests, named Mêdyômâh, Vardast, and Siyâvakhsh, near the time of the Arab conquest of Persia. This, however, really means little more than that the prose Sad Dar was considered a very old work at the time when the long-metre Sad Dar was composed from it. It appears, from Dastûr Jâmâspji’s preface to his Gugarâti translation of the long-metre Sad Dar, that this metrical version was composed in A.D. 1531 by Mullâ Rustam Isfendiyâr of Khurâsân and Mullâ Behzâd Rustam. It may, therefore, be concluded that the prose Sad Dar had the reputation of being a very old work in the early part of the sixteenth century.