The HadÎqatul-haqÎqat (the Enclosed Garden of the Truth)

INTRODUCTION

I.     LIFE OF THE AUTHOR

II.   MANUSCRIPTS AND LITHOGRAPHS

III.  HISTORY OF THE TEXT

IV.  THE COMMENTATORS

V.   THE Hadîqatu'l-Haqîqat

VI.  SANÂ'Î'S PREFACE

I- LIFE OF THE AUTHOR

Abû'l-Majd Majdûd b. Adam Sanâ'î1 was born at Ghazna, and lived in the reign of Bahrâmshâh (A.H. 512-548, A.D. 1118-1152). Ouseley says of him that he "while yet young became one of the most learned, devout, and excellent men of the age which he adorned. His praise was on every tongue; for, in addition to his accomplishments in the Sufi philosophy, he possessed a kind and benevolent heart, delightful manners, and a fine taste for poetry . . Sanâî in early life retired from the world and its enjoyments, and the reason for his doing so is supposed to have arisen from the following circumstance.

"He had frequented the courts of kings and princes, and celebrated their virtue and generous actions. When Sultan Ibrahim of Ghazni determined upon attacking the infidel idolaters of India, Hakim Sanâî composed a poem in his praise, and was hurrying to the court to present it before that monarch's departure. There was at that time in Ghazni a madman known as Lâi Khûr (the ox-eater), who often in his incoherent wanderings uttered sentiments and observations worthy of a sounder head-piece; he was addicted to drinking wine, and frequented the bath. It so happened that Sanâî, in passing a garden, heard the notes of a song, and stopped to listen. After some time the singer, who was Lâi Khûr, addressing the cup-bearer, said, 'Saki, fill a bumper, that I may drink to the blindness of our Sultan, Ibrahim.' The Saki remonstrated and said it was wrong to wish that so just a king should become blind. The madman answered that he deserved blindness for his folly in leaving so fine a city as Ghazni, which required his presence and care, to go on a fool's errand in such a severe winter. Lâi Khûr then ordered the Saki to fill another cup, that he might drink to the blindness of Hakim Sanaî. The cup-bearer still more strongly remonstrated against this, urging the universally esteemed character of the poet, whom everyone loved and respected. The madman contended that Sanâî merited the malediction even more than the king, for with all his science and learning, he yet appeared ignorant of the purposes for which the Almighty had created him; and when he shortly came before his Maker, and was asked what he brought with him, he could only produce panegyrics on kings and princes,--mortals like himself. These words made so deep an impression on the sensitive mind of the pious philosopher, that he secluded himself from the world forthwith, and gave up all the luxuries and vanities of courts.

"Sirâjuddin Ali, in his 'Memoirs of the Poets,' says, that in consequence of the sudden impression occasioned by Lâi Khûr's remarks, Sanâî sought instruction from the celebrated Sheikh Yusef Hamdani, whose cell was called the 'Kaabah of Khorâsân.'

"It was about this time that Behrâm Shah offered him his sister in marriage, which honour, however, he gratefully declined, and almost immediately set out on a pilgrimage to Mecca and Medinah. It is to the refusal of the royal bride that he alludes in his Hedîkeh, as an apology to the king, in the following lines:--'I am not a person desirous of gold or of a wife, or of exalted station; by my God, I neither seek them nor wish them. If through thy grace and favour thou wouldest even offer me thy crown, I swear by thy head I should not accept it.'" The account of Sanâ'î's conversion contained in the foregoing extract is probably, as Browne says, of little historical value.

Sanâ'î composed the present work after his return from the pilgrimage; according to most copies he completed it in A.H. 525 (A.D. 1131), though some MSS. have A.H. 534 or 535 (A.D. 11391141).

Sanâ'î was attacked during his lifetime on account of his alleged unorthodoxy; but a fatwa was published by the Khalîfa's court at Baghdâd, vindicating his orthodoxy against his calumniators. His commentator `Abdu'l-Latîf, if in his Preface (v.post .) mentions the suspicions of the various sects on the subject of the Hakîm's heresies.

Several dates are given for the Hakîm's death. His disciple Muhammad b. `Ali al-Raffâ (Raqqâm), in a preface to the work preserved in one of the Bodleian MSS., gives Sunday, the 11th Sha`bân A.H. 525 (A.D. 1131). This date, however, fell on a Thursday; the 11th Sha`bân of the year A.H. 545 (A.D. 1150), which is the date given by Taqî Kâshî and theÂtashkada , was, however, a Sunday. Daulatshâdh and Hâjî Khalfa give A.H. 576 (A.D. 1180, 1181). Since the poet completed hisTarîqu't-Tahqîq in A.H. 528, the earliest of the three dates is impossible; the second would appear -to be the most probable.

Besides theHadîqatu'l-Haqîqat , the first chapter of which is here presented, Sanâ'î wrote theTarîqu't-Tahqîq ("Path of Verification"),Gharîb-nâma ("Book of the Stranger"),Sairu'l-`ibâd ila'l-Ma`âd ("Pilgrimage of [God's] servants to the Hereafter"),Kâr-nâma Book of Deeds "),`Ishq-nâma (" Book of Love "), and*`Aql-nâma* ("Book of Reason"), as well as aDîwân , or collection of shorter poems in various metres. All these works, with the exception of theHaqîqa and theDîwân , are said by Prof. Browne, from whom the above list is taken, to be very rare.

Footnote

1-  For the facts contained in the following sketch I am indebted to Sir Gore Ouseley's "Biographical Notices of the Persian Poets," Lond., Or. Trans. Fund, 1846; Rieu's and Ethé's Catalogues; and Prof. Browne's "A Literary History of Persia," Vol. II.

II- MANUSCRIPTS AND LITHOGRAPHS

I have used the following manuscripts and lithographs in the preparation of the text:--

(1) Br. Mus. Add. 25329. Foll. 298, 7 ¾" x 4 ¾", 15 ll. 2 3/8" long, in small Nestalik, with gold headings, dated Safar A.H. 890 (A.D. 1485) [Adam Clarke].

There are marginal additions by two other hands; f. 1 is on different paper, by a different and later hand. The letters #, #, #, # are often not distinguished, # never; # and # are often not distinguished from # and #; the small letters are often without dots; the scribe usually writes the modern undotted # with three dots below. There are large omissions as compared with later MSS. and the lithographs.

I denote this MS. by C.

(2) Br. Mus. Or. 358. Foll. 317, 6 ¾" x 3 ¾", 17 ll. 2" long, in small Nestalik, in two gold-ruled columns, with two `unvâns, apparently written in the 16th cent. [Geo. Wm. Hamilton].

There are many marginal additions, mostly by one, a later, hand,: the MS. as a whole has been subjected to a great many erasures and corrections. The writing is good, the pointing of the letters fairly complete; the scribe usually writes and the # and #, the # rarely appears with three dots below. The MS. contains the prefaces of Raqqâm and of Sanâ'î himself, but, like the preceding, shows omissions as compared with later MSS. and the lithographs.

I denote this MS. by H.

(3) Br. Mus. Add. 16777. Foll. 386, 10 ¾" x 6 ¼", 15 ll., 3 ½" long, in fair Nestalik, with gold-ruled margins, dated A.H. 1076 (A.D. 1665) [Win. Yule].

This is a clearly written MS., the pointing of the letters usually full, #, and, # are frequently distinguished by their dots, and the pure # usually written with three dots below. Erasures are not frequent; the marginal corrections usually by the original hand. This MS. gives a very large number of divergent readings as compared with the others; its order is very different from that of the others; it is, as regards its extent, not so much defective as redundant, long passages appearing twice, and some passages not to be found in any of my other sources are also included. Some of these latter I have found in subsequent chapters of the Hadîqa, and it is possible that a more thorough search might have shown that they are all contained there.

This MS. is denoted by M.

(4) Ind. Off. 918. Ff. 395, 2 coll. each ll. 15; Nasta`lîk; the last four pp. written by another hand; 9 ½" by 5 ½". Written at Isfahân A.H. 1027 (A.D. 1618); occasional short glosses on the margin.

A clearly written and well-preserved MS., closely related to the following. The letters # and #, are frequently distinguished; the signmadda , is usually omitted.

I denote this MS. by I.

(5) Ind. Off. 923. The description given in the Catalogue is as follows --"Sharh-Hadîkah. The revised and collated edition of Sanâ'î's Hadîkah with a commentary and marginal glosses by `Abd-allatîf bin `Abdallâh al-`Abbâsî, who is best known by his revised and annotated edition of Jalâl-aldîn Rûmi's Mathnawî, his commentaries on the same poem, and a special glossary, Latâ`'f-allughât (lithogr. Lucknow under title Farhang-i-Mathnawî 1877). He died 1048 or 1049 (A.D. 1638, 1639) in Shahjahân's reign. The present copy, which is the author's autograph, was finished by him 20th Jumâdâ alawwal A.H. 1044 (=Nov. 11th, 1634), and represents an abridgement from a larger commentary of his, the Latâ'if al Hadâ'ik, from which also the glosses are taken (marked #). According to the dîbâca he began the larger work 1040 and completed it 1042 (1630-33) supported by his friend Mîr `Imâd-aldîn Mahmûd al Hamadânî, with the takhallus Ilâhî, the author of the well-known tadhkirah of Persian poets the Khazîna-i-Ganj.'

The following is an account of the contents of this MS. First comes a short preface by `Abdu'l-Latîf, introducing Sanâ'î's own preface, which is stated to have been written to the complete collection of his writings; it is frequently, states `Abdu'l-Latîf, not to be found in copies of his works. After Sanâ'î's preface comes another, calledRâsta-i khiyâbân , by `Abdu'l-Latîf, described as a short preface to this writer's commentary; this concludes with a reference to Ilâhî and his share in the work, and twotârîkhs by Ilâhî, giving A.H. 1040 as the date of its commencement, and 1042 as that of its completion. A few more lines by `Abdu'l-Latîf introduce the work itself. The original numbering of the folia commences with the text; there is also a pencil numbering, in English characters, beginning with the first preface The poem closes with 59 verses, in the same metre, which form an address to Abû'l-Hasan `Alî b. Nâsir al Ghaznawî, named Biryângar, sent to him at Baghdâd, because of the accusations of the traducers of the book. The date of completion of the text is given as A.H. 535; and, in a triangular enclosure of gold lines, it is stated that "this honoured copy was completed 20th Jumâdâ al-awwal, 1044 A.H." A few pages at the end, written by the same hand, give an account of how the book was sent to Biryângar at Baghdâd, on account of the accusations that were brought against it; how it was found to be orthodox, and a reply sent to Ghaznî.

This MS. I denote by A.

(6) The Lucknow lithograph published by the Newal Kishore Press, dated A.H. 1304 (A.D. 1886). This is an edition of the whole, work, including prefaces and `Abdu'l-Latîf's commentary. It comprises 860 pp., of 15 verses to a page; the paper, as usual, is somewhat inferior; the text is on the whole easily legible, but the same cannot always be said for the commentary, written in the margins and in a much smaller hand. It contains first a list of the titles of all the sections of all the chapters, followed by some verses setting forth the subjects of the ten chapters each as a whole. The ornamental title-page follows, stating that theHadîqa of Sanâ'î is here accompanied by the commentaryLatâ'ifu'l-Hadâ'iq of `Abdu'l-Latîf al-`Abbâsî. On p. 2 begins the 'First Preface', calledMirâtu'l-Hadâ'iq , by 'Abdu'l-Latîf, dated 1038 A. H.; this is not included in A; an abstract of it is given later (v. p. xxi). After this comes Sanâ'î's preface with `Abdu'l-Latîf's introductory words, as in A; this is called the 'Second Preface'. The 'Third Preface', which is `Abdu'l-Latîf'sRâsta-i khiyâbân , is here written in the margins of the ' Second Preface'. Then comes the text with marginal commentary, introduced as in A by a few more words from `Abdu'l-Latîf. At the conclusion of the work is the address to Biryângar; and finally someqit`as on the dates of commencement and completion of the printing of the book.

I denote this lithograph by L.

(7) I obtained from Bombay, from the bookshop of Mirzâ Muhammad Shîrâzî, another lithograph, which comprises only the first chapter of the work accompanied by a copious marginal commentary. Pp. 15 + 4 + 31 + 188, 15 ll. to a page; published at Lûhârû (near Hiss ar, Punjab) 1290 A.H. (1873 A.D.). The title-page states that this is the commentary on Sanâ'î's Hadîqa by Nawâb Mirzâ `Alâu'd-Dîn Ahmad, Khân Bahâdur, chief (###) of Lûhârû, called `Alâ'î, the scribe being Maulavî Muhammad Ruknu'd-Dîn of Hiss ar. Ruknu'd-Dîn states (p. 2) that he himself was doubtful of many words, and did not understand a number of the verses; he took his difficulties to `Alâ'î, who explained all; and "Praise be to God, there never has been such a commentator of the Hadîqa, nor will be; or if there is, it will be an imitation or a theft from this king of commentators." This reads rather curiously when considered in connection with the fact, to be mentioned hereafter, that the authors have incorporated in their commentary the whole of that of `Abdu'l-Latîf, and that their original contributions to the elucidation of the text are of slight value. Ruknu'd-Dîn was asked one day by the printers (###) to bring them his copy (###) of theHadîqa on its completion, for printing and publication. Pp. 4-10 are occupied by an Arabic preface by Ruknu'd-Dîn, again in extravagant praise of `Alâ'î and his accomplishments as a commentator. There follows (pp. 11-14) another title-page, and a short poem by `Alâ'î; and then (p. 15) aqit`a , giving the dates of commencement and completion of the work. Four pages of introduction (pp. 1-4) follow, and again with separate paging, 31 pp. of commentary on the first 28 pp. of the text, the reason apparently being that the whole of the commentary on these pages could not conveniently be written in the margins. The text comprises 186 pp., and includes (though I cannot find this stated anywhere) only the first book of the completeHadîqa ; the volume is concluded by some lines of `Alâ'î in praise of Muhammad, and a benediction. At the end of the marginal notes on every page is written "`Alâ'î sallamahu," or "Maulânâ `Alâ'î sallamahu Allâhu ta`âla ."

III- HISTORY OF THE TEXT

Muhammad b. `Alî Raqqâm informs us, in his preface to theHadîqa , that while Sanâ'î was yet engaged in its composition, some portions were abstracted and divulged by certain ill-disposed persons. Further, `Abdu'l-Latîf in his preface, theMirâtu'l-Hadâ'iq , states that the disciples of Sanâ'î made many different arrangements of the text, each one arranging the matter for himself and making his own copy; and that thus there came into existence many and various arrangements, and two copies agreeing together could not be found.

The confusion into which the text thus fell is illustrated to some extent by the MSS. which I have examined for the purpose of this edition. C shows many omissions as compared with later MSS.; at the same time there is a lengthy passage, 38 verses, which is not found in any other; H, though also defective, is fuller than C but evidently belongs to the same family. M contains almost all the matter comprised in `Abdu'l-Latîf's recension, much of it twice over as has already been mentioned; and in addition about 300 verses, or altogether 10 folia, which apparently do not of right belong to this first chapter at all; the first chapter, too, is here divided into two chapters. The remaining MSS. and lithographs agree closely with each other and are evidently all nearly related.

The same story, of an early confusion of the text, is even more strikingly brought out if, instead of the omissions and varying extent of the text in the several MSS., we compare the order of the text. Here M startles us by giving us an order totally at variance with that of any other of our sources. There seems to be no reason for this: the arrangement of the subject is not, certainly, more logical; and it would appear that the confusion has simply been due to carelessness at some early stage of the history of the text; the repetitions, and the inclusions of later parts of the work, point to the same explanation. I need only mention the consequent labour and expenditure of time on the collation of this manuscript. C and H agree mostly between themselves in the order of the text, and broadly speaking the general order is the same as that of the later MSS.; the divergences would no doubt have appeared considerable, but that they are entirely overshadowed by the confusion exhibited by M. IALB agree closely with each other, as before.

The same confusion is again seen in the titles of the various sections as given in the several MSS. I am inclined to doubt how far any of the titles are to be considered as original; and it seems to me very possible that all are later additions, and that the original poem was written as one continuous whole, not divided up into short sections as we have it now. At any rate, the titles vary very much in the different MSS.; some, I should say, were obviously marginal glosses transferred to serve as headings; in other cases the title has reference only to the first few lines of the section, and is quite inapplicable to the subject-matter of the bulk of the section; in other cases again it is difficult to see any applicability whatever. It appears to have been the habit of the copyists to leave spaces for the titles, which were filled in later; in some cases this has never been done in others, through some omission in the series, each one of a number of sections will be denoted by a title which corresponds to that of the text following section in other MSS.

It is then obvious that `Abdu'l-Latîf is right in saying that in the centuries following Sanâ'î's death great confusion existed in the text of theHadîqa . This text he claims to have purified and restored, as well as explained by means of his commentary; and it is his recension which is given in A, as well as in the Indian lithographs Land B. He says that he heard that the Nawâb Mirzâ Muhammad `Azîz Kaukiltâsh, styled the Great Khân, had, while governor of Gujrât in the year 1000 A.H., sent to the town of Ghaznîn a large sum of money in order to obtain from the tomb of Sanâ'î a correct copy of theHadîqa , written in an ancient hand; this copy the Nawâb, on his departure on the pilgrimage, had bestowed on the Amîr `Abdu'r-Razzâq Ma`mûri, styled Muz affar Khan, at that time viceroy of that country. `Abdu'l-Latîf, however, being then occupied in journeys in various parts of India, could not for some time present himself before the Amîr; till in A.H. 1035 this chief came to Agra, where `Abdu'l-Latîf presented himself before him and obtained the desire of so many years. This MS. of theHadîqa had been written only 80 years after the original composition, but the text did not satisfy the editor, and it was besides deficient, both in verses here and there, and also as regards twenty leaves in the middle of the work.

In the year A.H. 1037 `Abdu'l-Latîf came to Lahore, where having some freedom from the counterfeit affairs of the world and the deceitful cares of this life, he entered again on the task of editing the text, with the help of numerous copies supplied to him by learned and critical friends; He adopted the order of the ancient MS. before-mentioned, and added thereto such other verses as he found in the later MSS. which appeared to be of common origin, and to harmonize in style and dignity and doctrine, with the text. As to what `Abdu'l-Latîf attempted in his commentary, v. p. xxiipost .

So far `Abdu'l-Latîf's own account of his work. We can, however, supplement this by a number of conclusions derived from the MSS. themselves.

In the first place, it appears that A is not, as stated in the India Office Catalogue, `Abdu'l-Latîf's autograph copy. The statement that it is so is apparently based on the fact of the occurrence of the words "harrarahu wa sawwadahu `Abdu'l-Latîf. b. 'Abdu'llâhi'l-`Abbâsî ," at the end of the editor's few words of introduction to Sanâ'î's preface and again of the occurrence of the words "harrarahu `Abdu'l-Latîf . ki shârih wa niusahhih-i în kitâb-i maimunat ni sâb ast ," at the end of the few lines of introduction immediately preceding the text. But both these sentences are found in the Lucknow lithograph, and therefore must have been copied in all the intermediate MSS. from `Abdu'l-Latîf's autograph downwards the words in each case refer only to the paragraph to which they are appended, and were added solely to distinguish these from Sanâ'î's own writings.

1 cannot find any other facts in favour of the statement that A is the editor's autograph; there are, however, many against it. Thus A is beautifully written, and is evidently the work of a skilled professional scribe, not of a man of affairs and a traveller, which `Abdu'l-Latîf represents himself as having been. Again, there are occasional explanatory glosses to the commentary, in the original hand; these would have been unnecessary had the scribe been himself the author of the commentary. The handwriting is quite modern in character and the pointing is according to modern standards throughout; the late date of A is immediately brought out clearly by comparing it with I (of date 1027 A.H.) or M (of date 1076 A.H.); though the supposed date of A is 1044 A.H. it is obviously much later than either of the others. But perhaps the most curious bit of evidence is the following; at the top of fol. 11b of the text of A there is an erasure, in which is written ### in place of an original reading ###, and as it happens this line is one which has been commented on by the editor; in the margin is a note in a recent hand,--###, which is true,--the commentary certainly presumes a reading ###, but this MS. had originally ###; the scribe could not therefore have been the commentator himself, i.e., `Abdu'l-Latîf

Further, not only is A not `Abdu'l-Latîf's autograph, but it does not accurately reproduce that autograph. I refer to 34 short passages of Sanâ'î's text, which in A are found as additions in the margin; these, though obviously written in the same hand, I regard as subsequent additions from another source by the same scribe, not as careless omissions filled in afterwards on comparing the copy with the original. In the first place, the scribe was on the whole a careful writer; and the mistakes he has made in transcribing the commentary, apart from the text, are few. The omissions of words or passages of commentary, which have been filled in afterwards, are altogether 10; of these, two are of single words only; two are on the first page, when perhaps the copyist had not thoroughly settled down to his work; five are short passages, no doubt due to carelessness; and one is a longer passage, the whole of a comment on a certain verse,--an example of carelessness certainly, but explicable by supposing that the scribe had overlooked the reference number in the text indicating that the comment was to be introduced in relation to that particular verse. Roughly speaking, the commentary is of about equal bulk with the text; yet the omissions of portions of commentary by the copyist are thus many fewer in number and much less in their united extent than the omissions of the text,--supposing, that is, that the marginal additions to the text in A are merely the consequence of careless copying. The reverse would be expected, since owing to the manner of writing, it is easier to catch up the place where one has got to in a verse composition; it would seem therefore. as said above, that the comparatively numerous marginal additions to the text are rather additions introduced afterwards from another source than merely careless omissions in copying. In the second place, none of these 34 passages are annotated by `Abdu'l-Latîf; in all likelihood, if they had formed part of his text, some one or more of the lines would have received a comment. The passages comprise, together, 63 verses; there is only one instance in the First chapter of theHadîqa of a longer consecutive passage without annotation, and in general it is rare (eleven instances only) to find more than 30 consecutive verses without annotation; usually the editor's comments occur to the number of two, three or more on each page of 15 lines. I think, therefore. it must be admitted that the chances would be much against a number of casual omissions aggregating 63 lines falling out so as not to include a single comment of the editor. Thirdly, it is a remarkable fact that of these 34 passages the great majority are also omitted in both C and H, while they are present in both M and 1; to particularize, C omits 30 ½, H omits 28, both C and H omit 25 ½, and either C or H or both omit every one of these 34 passages; while I and M each have all the 34 with one exception in each case; further, while many of these 34 marginally added passages in A correspond exactly to omissions in H, the corresponding omissions in C may be more extensive, i.e., may include more, in each case, of the neighbouring text.

We must therefore, I think, conclude that after completing the transcription of A the scribe obtained a copy of theHadîqa of the type of I or M, and filled in certain additions therefrom; and that '`Abdu'l-Latîf's edition did not originally contain these passages.

Let us turn to a consideration of I and its relation to `Abdu'l-Latîf's edition. I is dated A.H. 1027; it is, therefore, earlier than `Abdu'l-Latîf's edition of A.H. 1044. As we have seen, A is not `Abdu'l-Latîf's autograph; but we have, I think, no reason to doubt that it was either copied from that autograph, or at any rate stands in the direct line of descent; so much seems to be attested by the occurrence of the words --"harrarahu `Abdu'l-Latîf . . . " and by the inscription at the end as to the completion of the book in A.H. 1044, the actual date of the completion of `Abdu'l-Latîf's work. Regarding, then, A as presenting us (with the exception of the marginally added passages) with a practically faithful copy of `Abdu'l-Latîf's own text, we notice a striking correspondence between this text and that of I. As to the general agreement of the readings of the two texts, a glance at the list of variants will be sufficient; and it is not impossible to find whole pages without a single difference of any importance. The titles also, which as a rule vary even so much in the different MSS., correspond closely throughout. The order of the sections is the same throughout; and the order of the lines within each section, which, ' is also very variable in the various M88., corresponds in I and A with startling closeness. The actual spellings of individual words also, which vary even in the same MS., are frequently the same in I and A; for example, at the bottom of p. ### of the present text the word ### or ### occurs three times within a few lines. The word may also be written ###, ###; thus while C and M have ###, H has first ### and then twice ###; I however has first ### and then twice ###; and this is exactly repeated in A. Another example occurs a few lines afterwards (p. ###, l. ###); the reading is ###,mâr-i shikanj ,mâr being followed by the izâdfat; this I writes as ###; in A an erasure occurs between ### and ###, doubtless due to the removal of a ### originally written there as in I.

The above will serve to show the close relation between I and A, or between I and `Abdu'l-Latîf's autograph, of which A is a copy or descendant. But, however close this relationship, `Abdu'l-Latîf cannot actually have used I in the preparation of his revision of the text, or he would certainly have incorporated many of the 34 passages before alluded to, which were all, with one exception, contained in 1. These, we have seen, were only added by the scribe of A, and by him only subsequently, from another source, after he had completed his transcription from `Abdu'l-Latîf's autograph.

The facts, then, are these. There was in existence, before `Abdu'l-Latîf's time, a tradition, probably Persian, of the order of the text, which he adopted even in detail. This is represented for us by I, written A.H. 1027 at Is fahân; but I itself is somewhat fuller than the copy of which `Abdu'l-Latîf if made such great use. This copy may be called P. Such use, indeed, did `Abdu'l-Latîf make of P, that, so far as can be seen, it is only necessary that he should have had P before him, with one or two other copies from which he derived a certain number of variant readings, which he substituted here and there in his own edition for those of P.

We have now brought down the history of the text to A.H. 1044. Not much remains to be said; A, as we have seen, is quite possibly a direct copy of `Abdu'l-Latîf's autograph, with, however, marginal additions from another source. This other source might be at once assumed to be 1, but for the fact that only 33 out of the 34 marginally added passages occur in I; and it still seems to me at least possible that I was thus used. 1, though written at Is fahân, was probably by this time in India, where A, the so-called "Tippu MS.," was certainly written; at least, that I did come to India may be assumed from its presence in the India Office Library. Again, though it is, I think, impossible that the whole of the 34 passages added marginally in A should have been careless omissions of the copyist, one or two might possibly be so, and it is possible that the single line now under discussion may be such an omission, filled in from the scribe's original, not from another source. Finally it is, of course, always possible that the additions were taken from two sources, not one only; i.e., that while perhaps even 33 were filled in after comparison with I, the single remaining line may have been derived from elsewhere. Though absent in C, it is present in both H and M.

As to the lithographs, both are obviously descendants of A. The above conclusions may be summarized in the followingstemma codicum .

The present text is founded on that of the Lucknow lithograph L, with which have been collated the other texts mentioned above. L is practically a verbatim copy of A, the value of which has been discussed above. Though MSS. of theHadîqa are not rare, at least in European libraries, I have not met with any in India; and a considerable portion of the first draft of the translation and notes was done on the basis of L and B alone. TheHadîqa is not in any case an easy book, with the exception, perhaps, of a number of the anecdotes which are scattered through it; and it was rendered far more difficult by the fact, which I did not recognize for some time, that a very great amount of confusion exists even in the text as it is published to-day, in the lithographs descended from `Abdu'l-Latîf's recension. There appeared to be frequently no logical connection whatever between successive verses; whole pages appeared to consist of detached sayings, the very meaning of which was frequently obscure; a subject would be taken up only to be dropped immediately.

I ultimately became convinced that the whole work had fallen into confusion, and that the only way of producing any result of value would be to rearrange it. This I had done, tentatively, for part of the work, before collating the British Museum and India office MSS. cited above.

When I came to examine the MSS., the wide variations, not only in the general order of the sections to which allusion has already been made, but in the order of the verses within each section, showed me that probably no MS. at the present day, or at any rate none of those examined by me, retains the original order of the author: and I felt justified in proceeding as I had begun, altering the order of the lines, and even of the sections, if by so doing a. meaning or a logical connection could be brought out. I need not say that the present edition has no claims to represent Sanâ'î's original; probably it does not represent it even approximately. In some cases there is, I think, no doubt that I have been able to restore the original order of the lines, and so to make sense where before it was wanting; in other cases this is possible, but I feel less confident; while in still others the reconstruction, preferable though I believe it to be to the order as found in any single MS., is nevertheless almost certainly a, makeshift, and far from the original order. Lastly it will be seen that I have quite failed, in a number of instances, to find the context of short passages or single lines; it seemed impossible to allow them to stand in the places they occupied in any of the MSS., and I have, therefore, simply collected them together, or in the ease of single lines given them in the notes.

IV- THE COMMENTATORS

Khwâja `Abdu'l-Latîf b. `Abdullâh al-`Abbâsî, already so frequently mentioned, explains to us in his Preface, theMirâtu'l-Hadâ'iq , what he has attempted in his commentary on theHadîqa . He states that he was writing in A.H. 1038, in the second year of the reign of the Emperor Shahjahân, that he had already completed his work on Jalâlu'd-Dîn Rûmî's Mathnawî, and that he had in A.H. 1037 settled down to work on theHadîqa . 'What he professes to have done for the text of that work has been mentioned in the last section; the objects he has aimed at in the way of commentary and explanation are the following:-- Firstly, he has followed up the references to passages in the Qur'ân, has given these passages with their translations, and a statement of the sûra in which they are to be found. Secondly, the traditions referred to are also quoted. Thirdly, obscure passages have been annotated, and strange or curious Arabic and Persian words have been explained, after an investigation into their meanings in trustworthy books. Fourthly, certain signs have been used in transcribing the text, in order to fix the signification of various letters; thus theyâ'i kitâbî is denoted by ### subscript, theyâ'i majhûl similarly by ###, theyâ'i ma'rûf by ###, the Persian # (#) by #, the Arabic # by #, and so on. Again the vocalization has been attended to in words which are often mispronounced; thus ignorant people often substitutefatha forkasra in such words as*`khizâna'* , of which theQâmûs says "Khizâna is never pronounced withfatha "; 'Shamâl ', meaning the North wind, should be pronounced withfatha , notkasra , as is often done. Theizâfat ,jazm , and other orthographical signs have often been written in the text; and finally a glossary of the less known words has been added in the margin. Since it is inconvenient to have text and commentary separate, "in this copy the whole stability of the text has been dissolved, and the text bears the commentary along with it (###), i.e., text and commentary are intermingled, the commentary not being written in the margin, but each annotation immediately after the word or line to which it applies. These researches the author has also written out separately, and called them "La tâ'ifu'l-Hadâ'iq min Nafâ'isi'l-Daqâ'iq ." The date is again given as A.H. 1038.

It appears then that the original form of the commentary was not that of marginal notes, as it is presented in A and L; that it was completed in 1038 A.H., and, in its separate form, was called theLa tâ'ifu'l-Hadâ'iq . That this is the name of the commentary we know and possess, seems to have been the opinion of the scholar who prepared the Lucknow lithograph, which is entitled "Sanâ'î'sHadîqa , with the commentaryLa tâ'ifu'l-Hadâ'iq ."

Besides the preface just considered, there is also another, found in both A and L, called theRâsta-i Khiyâbân , written especially, it would seem, as an introduction to the commentaryLa tâ'ifu'l-Hadâ'iq . After dwelling on the unworthiness of the writer, `Abdu'l-Latîf states that the interpretations given by him are not mere expressions of private opinion, but are derived from the best Arabic and Persian books; the emendations of the text are all derived from authentic MSS., and are in accordance with the judgment of discerning men; everything has been weighed and discussed by the learned. He does not, however, say that these explanations are the only ones, nor that he has commented on every line that to some people would seem to require it. Though his work may seem poor now while he is alive, it may grow in the esteem of men after his death. The work has been done in the intervals of worldly business, while occupied with affairs of government. There follows a lengthy eulogy of his friend Mîr `Imâdu'd-Dîn Mahmûd al-Hamadâni, called Ilâhî, twotârîkhs by whom close this preface. The firsttârîkh says that the work having been begun in the year 1040. all the correction and revision was completed in 1042 (###); the second simply gives the date 1040.

These dates evidently cannot refer to the edition and commentary as first written; since we have seen that the text and theLa tâ'ifu'l-Hadâ'iq are referred to by `Abdu'l-Latîf in 1038 as having been completed. It would seem that the editor had either been at work on another, revised and improved edition; or, as is assumed in the India Office Catalogue (No. 923), on an abridgment of his earlier work. Lastly, we have the date 1041 for the completed work of which A is a copy (see description of contents of A, in Section II, p. xi); and this seems to represent the final form of the work. in which the annotations are written in the margin, not, as at first, intermingled in the text.

In the India Office Catalogue the series of events is interpreted somewhat differently. The commentary as it appears in A (and L, the only form, apparently, in which we possess it) is stated to be an abridgement from a larger commentary, theLa tâ'ifu'l-Hadâ'iq ; according to the preface (the Catalogue states) the larger work wa-. begun in 1040 and completed in 1042. It is with diffidence that I venture to question this presentation of the facts; but A, in the description of which the above statements occur, does not contain the preface calledMirâtu'l-Hadâ'iq , and therefore presents no indication that the text andLa tâ'ifu'l-Hadâ'iq had already been completed in 1038. That the work done between 1040 and 1042 consisted in the preparation of the originalLa tâ'ifu'l-Hadâ'iq is, from the statement of theMirâtu'l-Hadâ'iq , impossible. We have seen, moreover, that the tradition in India is that the commentary as we have it, as it appears in A and L, is theLa tâ'ifu'l-Hadâ'iq itself., and not an abridgement. I do not gather from the India Office Catalogue or elsewhere that two-commentaries, a larger and a smaller, are actually in existence; there may be other evidences of their former existence of which I am ignorant, but so far merely as my own knowledge goes, I can see no reason for assuming two commentaries, and would look on the labours of 1040-1042 in the light of revision and rearrangement, a work which was perhaps only finally completed in 1044, the date given in A for the completion of the work. Besides his work on theHadîqa , `Abdu'l-Latîf had previously, as has been mentioned, published a revised and annotated edition of Jalâlu'd-Dîn Rûmî's Mathnawî, commentaries on the same poem, and a special glossary, the lithographed at Lucknow in A.D. 1877 under the titleFarhang-i Mathnawî . He died in 1048 or 1049 A.H. (A.D. 1638, 1639).

A general description of the volume containing the other commentary which I have used in the preparation of the notes appended to the present translation, has already been given. Of the authors, or author and scribe, Mirzâ `Alâu'd-Dîn Ahmad of Lûhârû, called `Alâ'î, and Maulavî Muhammad Ruknu'd-Dîn of Hiss ar, I know no more, than is to be gathered from their prefaces.

Their commentary is of slight value as compared with that of `Abdu'l-Latîf: that is to say, that part of it which is original. The commentary is considerably more bulky than `Abdu'l-Latîf's, perhaps between two and three times as extensive; but it includes, without one word of acknowledgment, the whole of `Abdu'l-Latîf's work. This is, in the great majority of cases, reproduced verbatim; in some instances a paraphrase of `Abdu'l-Latîf's commentary has been attempted, and in certain of these it is plain that the authors did not understand the sense of what they paraphrased. Of their own work, a certain amount is superfluous, the sense of the text being immediately obvious; a certain amount is mere paraphrase of Sanâ'î's words: and another portion consists in an attempt to read mystical meanings into the original in passages which, as it seems, were never intended by the author to bear them. Notwithstanding these facts, I have, as will be seen, quoted freely in my notes from their commentary; for a certain portion of their work is helpful, and moreover, it seemed to me to be of interest to give in this way a specimen of present-day Indian thought and criticism in the field ofS ûfîistic philosophy. I cannot, however, leave the subject of Sanâ'î's commentators without expressing my sorrow that scholars should have existed who were not only capable of such wholesale theft, but even lauded themselves on the results of it; witness the extravagant praise of `Alâ'î in Ruknu'd-Dîn's preface; and again the words

Praise be to God! There has never been. such a commentator of theHadîqa , nor will be; or if there is. it will be an imitation or a theft from this king of commentators!" There is also no indication that the volume comprises only one out of ten chapters of theHadîqa ; it is everywhere implied that the. completeHadîqa is presented.

V-THE HADÎQATU'L-HAQÎQAT

TheHadîqatu'l-Haqîqat , or the "Enclosed Garden of the Truth", commonly called theHadîqa , is a poem of about 11,500 lines; each line consists of two hemistichs, each of ten or eleven syllables; the bulk, therefore, is equal to about 23,000 lines of English ten-syllabled verse. It is composed in the metre ### which may be represented thus:

The two hemistichs of each verse rhyme; and the effect may therefore roughly be compared to that of English rhymed couplets with the accent falling on the first (instead of the second) syllable of the line, and, occasionally, an additional short syllable introduced in the last foot.

The chapter,; of which theHadîqa consists treat, according to a few lines of verse at the end of the table of contents in the Lucknow edition, of the following subjects; the First, on the Praise of God, and especially on His Unity; the Second, in praise of Muhammad; the Third, on the Understanding; the Fourth, on Knowledge; the Fifth, on Love, the Lover, and the Beloved; the Sixth, on Heedlessness; the Seventh, on Friends and Enemies, the Eighth, on the Revolution of the Heavens; the Ninth, in praise of the Emperor Shâhjahân; the Tenth, on the characters or qualities of the whole work. This, however, is not the actual arrangement of the work as presented in the volume itself; the first five chapters are as already given, but the Sixth concerns the Universal Soul; the Seventh is on Heedlessness; the Eighth on the Stars; the Ninth on Friends and Enemies; the Tenth on many matters, including the praise of the Emperor. Prof. Browne (Lit. Hist. Persia, vol. ii., p. 318) gives still another order, apparently that of an edition lithographed at Bombay in A.H. 1275 (A.D. 1859).

Sanâ'î's fame has always rested on hisHadîqa ; it is the best known and in the East by far the most esteemed of his works; it is in virtue of this work that he forms one of the great trio ofS ûfî teachers,--Sanâ'î, `Att âr, Jalâlu'd-Dîn Rûmî. It will be of interest to compare some of the estimates that have been formed of him and of the present work in particular.

In time he was the first of the three, and perhaps the most cordial acknowledgment of his merits conies from his successor Jalâlu'd-Dîn Rûmî. He says:--

I left off boiling while still half cooked;

Hear the full account from the Sage of Ghazna."

And again--

"`Att âr was the Spirit, Sanâ'î the two eyes:

We walk in the wake of Sanâ'î and `Att âr."

`Abdu'l-Latîf, in his preface called theMirâtu'l-Hadâ'iq , enters into a somewhat lengthy comparison between Sanâ'î and Rûmî, in which he is hard put to it to avoid giving any preference to one or other. It is interesting to observe how he endeavours to keep the scales even. He begins by adverting to the greater length of theMathnawî as compared with theHadîqa , and compares theHadîqa to an abridgement, theMathnawî to a fully detailed account. Sanâ'î's work is the more compressed; he expresses in two or three verses what theMathnawî expresses in twenty or thirty, `Abdu'l-Latîf therefore, as it would seem reluctantly, and merely on the ground of his greater prolixity, gives the palm for eloquence to Jalâlu'd-Dîn.

There is the most perfect accord between Sanâ'î and Rûmî; tile substance of their works, indeed, is in part identical. Shall it therefore be said that Rûmî stole from Sanâ'î? He asks pardon from God for expressing the thought; with regard to beggars in the spiritual world, who own a stock-in-trade of trifles, bankrupts of the road of virtue and accomplishments, this might be suspected; but to accuse the treasurers of the stores of wisdom and knowledge, the able natures of the kingdom of truth and allegory, of. plagiarism and borrowing is the height of folly and unwisdom.

With regard to style, some suppose that the verse of theHadîqa is more elevated and dignified than the elegantly ordered language of theMathnawî . TheHadîqa does indeed contain poetry of which one verse is a knapsack of a hundreddîwâns ; nor, on account of its great height, can the hand of any intelligent being's ability reach the pinnacles of its rampart; and the saying--

"I have spoken a saying which is a whole work; I have uttered a sentence which is a (complete) dîwân,"

is true of theHadîqa . But if the sense and style of the Maulavî be considered, there is no room for discrimination and distinction; and, since "Thou shalt not make a distinction between any of His prophets ," to distinguish between the positions of these two masters, who may unquestionably be called prophets of religion, has infidelity and error as its fruit. Who possesses the power of dividing and discriminating between milk and sugar intermingled in one vessel? `Abdu'l-Latîf sums up thus "in fine, thus much one may say, that in sobriety the Hakîm is pre-eminent, and in intoxication our lord the Maulavî is superior; and that sobriety is in truth the essence of intoxication, and this intoxication the essence of sobriety."

Prof. Browne, however, places theHadîqa on a far lower level than the Eastern authors quoted above. He says1 :--"The poem is written in a halting and unattractive metre, and is in my opinion one of the dullest books in Persian, seldom rising to the level of Martin Tupper'sProverbial Philosophy , filled with fatuous truisms and pointless anecdotes, and as far inferior to theMathnawî of Jalâlu'd-Dîn Rûmî as is Robert Montgomery'sSatan to Milton'sParadise Lost ."

It is of course true that to us, at least, the interest of theHadîqa is largely historical, as being one of the early Persian text-books of theS ûfî philosophy, and as having so largely influenced subsequent writers, especially, as we have seen, the Maulavî Jalâlu'd-Dîn Rûmî. Yet I cannot butt think that Prof. Browne's opinion, which is doubtless shared by other scholars, as well as the neglect to which theHadîqa has been exposed in the West, is due not to the demerits of the original text so much as to the repellent and confused state into which the text has fallen; and I would venture to hope that the present attempt at a restoration of the form and meaning of a portion of the work, imperfect in the highest degree as I cannot but acknowledge it to be, may still be of some slight service to its author's reputation among European Orientalists.

The first Chapter or Book of theHadîqa , which is here presented, comprises a little more than one-sixth of the entire work. The subjects of which it treats may be briefly resumed as follows:--

After an introductory section in praise of God the author speaks of the impotence of reason for the attaining a knowledge of God; of God's Unity, of God as First Cause and Creator and delivers more than one attack against anthropomorphic conceptions of God (pp. 1-10). After speaking of the first steps of the ascent towards God, for which worldly wisdom is not a bad thing, with work and serenity (pp. 10-11), he devotes the next portion of the book to God as Provider, to His care for man through life, the uselessness of earthly possessions, and to God as guide on the road, but self must first be abandoned (pp. 11-46). A fine section on God's incomprehensibility to man might perhaps come more fittingly at an earlier stage instead of here (pp. 16-18). After overcoming self, God's special favour is granted to the traveller on the path: but we see crookedly, and He alone knows what is best for us: He has ordered all things well, and what seems evil is so only in appearance (pp. 18-25).

The greater part of the book is really concerned with the life and experiences of theS ûfî, and especially with continually repeated injunctions as to abandonment of the world and of self; to be dead to this world is to live in the other. Pp. 25-30 are thus concerned with poverty in this world, with loss of the, self, humility, man's insignificance and God's omnipotence; pp. 30-34 with the necessity of continual remembrance of God, of never living apart from Him, and again of dying to the world; death to the world leads to high position with God. There follows (pp. 34-41) a series of passages on the duty of thanksgiving for God's mercies; His mercy however has its counterpart in His anger, and examples of His wrath are given; then returning again to the subject of His mercies, the author speaks of God's omniscience, and His knowledge of the wants of His servants; we must therefore trust in God for all the necessaries of life, they will be given as long as life is destined to last. Two later pages (48-50), which are similarly devoted to the subject of trust in God, should probably come here. Pp. 41-48 deal with theS ûfîs desire for God, and his zeal in pursuing the path; various directions for the road are given, especially as rewards the abandonment of the world and of self, and fixing the desires on God only; union with God is the. goal. The abandonment of self is again the theme of pp. 50-51.

A portion of the book (pp. 51-56) is, curiously, here devoted to the interpretation of dreams; after which the author treats of the incompatibility of the two worlds, again of the abandonment of earth and self, and of the attainment of the utmost degree of annihilation (pp. 56-58). There follows a passage on the treatment of schoolboys, a comparison with the learner on theS ûfî path, and an exhortation to strive in pursuing it (pp. 58-60). The next portion of the book (pp., 60-67) treats of charity and gifts as a form of renunciation, of relinquishing riches for God's sake; prosperity is injurious to the soul, and the world must be abandoned; possessions and friends are useless, and each must trust to himself; each will find his deserts hereafter, and receive the reward of what he has worked for here.

Pp. 67-80 treat of prayer, the preparation for which consists in purity of heart, humility, and dependence upon God. Prayer must come from the heart; the believer must be entirely absorbed in his devotions. Prayer must be humble; the believer must come in poverty and perplexity, and only so can receive God's kindness. A number of addresses to God follow, prayers for help, and humble supplications to God on the part of the author. A few pages (80-92) treat of God's kindness in drawing men towards himself, though His ways may appear harsh at first. The progress of the believer is described in a strain of hyperbole (pp. 82-83); and this portion closes with a few sections (pp. 83-86) on God's majesty and omnipotence somewhat after the manner of those in the earlier part of the book.

In pp. 86-97 the author speaks of the Qur'ân, and its excellence and sweetness. The letter however is not the essential: its true meaning is not to be discovered by reason alone. The Qur'ân is often dishonoured, especially by theologians, and by professional readers, who read it carelessly and without understanding it. A short section (pp. 97-98) on humility and self-effacement follows, and the book is brought to a close by a description of the godlessness of the world before the advent of Muhammad (pp. 98-100), which serves to introduce the subject of the Second Chapter.

Though it must he admitted that the author is occasionally obscure, sometimes dull, and not infrequently prosaic, some fine sections and a larger number of short passages of great beauty are contained in this chapter; I may perhaps be permitted especially to refer to the sections "In His Magnification," pp. 16-18. and "On Poverty and Perplexity," p. 74; while as characteristic and on the whole favourable passages may be mentioned "On His Omniscience, and His Knowledge of the Minds of Men," pp. 37-39; " On the Incompatibility of the Two Abodes," pp. 56-58; "On intimate Friendship and Attachment," pp. 62-63; and certain of the addresses to God contained in pp. 74-77.

Footnote

1 A Literary History of Persia, Vol. II., p. [^319]:

VI- SANÂ'Î'S PREFACE

The author's Preface to the work, given in A and L, and occupying in the latter nearly thirteen closely printed pages, is here given in abstract. It was not, as will appear, written specially as an introduction to theHadîqa , but to his collected works.

After an opening section in praise of God, the author introduces the tradition, "When a son of Adam dies, his activity ceases, except in three things; a permanent bequest, and knowledge by which men are benefited, and pious sons who invoke blessings on him after his death ." Considering these words one day, and reflecting that none of the three conditions was applicable to himself, he became sorrowful, and continued for some time in a state of grief and depression. one, day while in this condition, he was visited by his friend Ahmad b. Mas`ûd, who inquired the cause of his sorrow. The author told him that, not fulfilling any one of the above conditions, he was afraid to die; possessing not one of these three advocates at court, he would stand without possessions or adornment in the Presence of the Unity. His friend then began to comfort him, saying, "First let me tell you a story." Sanâ'î replied, "Do so."

Ahmad b. Mas`ûd then related how one day a company of women wished to have audience with Fâtima, Muhammad's daughter. Muhammad gave permission; but Fâtima, weeping, said, "O Father, how long is it since I have had even a little shawl for my head? and that mantle that I had pieced together in so many places with date-leaves is in pledge with Simeon the Jew. How can I receive them?" But Muhammad said, "There is no help; you must go." Fâtima went ashamed to the interview, and came back in sorrow to her father; who was comforting her when the rustle of Gabriel's wings was heard. Gabriel looked at Fâtima and asked, "What is this sorrow? Ask the women, then, what garments they had on, and what thou." Muhammad sent a messenger to the women, who returned, and said, "It was so, at the time when the Mistress of Creation bestowed beauty on that assembly, that the onlookers were astounded; though clothed, they seemed to themselves naked; and among themselves they were asking 'Whence came this fine linen, and from which shop this embroidery? What skilful artificers, what nimble-fingered craftsmen!'" Fâtima said, "O my father, why didst thou not tell me, that I might have been glad?" He answered, "O dear one, thy beauty consisted in that which was concealed inside thyself."

"By my life," continued Ahmad, "such modesty was allowable in Fâtima, brought up in seclusion; but here we have a strong and able man of happy fortune, one who is known as a pattern to others in both practice and theory! Though thou hast considered thyself naked, yet they have clothed thee in a robe from the wardrobe of Eternity. Is it proper for this robe to be concealed, instead of being displayed for the enlightenment of others? " And adverting to the saying, "When a son of Adam dies, his work is cut short, except in three things ," he takes the three one by one. First,a continuing alms ; but 'Every kindness is an alms; and it is a kindness that thou meet thy brother with a cheerful countenance, and that thou empty thy bucket into the pots of thy brother ;' that is, alms does not wholly consist in spreading food before a glutton, or giving some worthless thing to a pauper; it is a truer alms and a more imperishable hospitality to wear a cheerful countenance before one's friends, " and if others have the outward semblance of alms, thou hast its inward essence; and if they have set forth a table of food before men, thou hast set forth a table of life before their souls; so much for what thou sayest, 'I am excluded from a continuing alms!'"

Ahmad b. Mas`ûd then takes up the second point, knowledge that benefits; and quotes, "We take refuge with God from knowledge which does not benefit " and "Many a wise man is destroyed by his ignorance and his knowledge which does not advantage him ." As examples of knowledge that does not benefit he takes the science of metaphysics, a science tied by the leg to desire and notoriety, lying under the opprobrium of "He who learns the science of metaphysics is a heretic, and flys in circles in the air ;" as well as of the saying "A science newly born, weak in its credentials "--"I have perfected it for the sake of heresy, and so peace." Then similarly the science of calculation, a veil which diverts attention from the Truth, a curtain in front of the subtilties of religion; and the science of the stars, a science of conjectures and the seed of irreligion, for "Whoso credits a soothsayer has become an infidel ." After a tirade against the ordinary type of learned man, he proceeds, "All their falsifyings and terrorizings and imaginings and conjecturings are limited by their own defects; that philosophy of the law is cherished which is notorious over all the quarters and regions of the world; there is your 'knowledge that men benefit by '! From earth to Pleiades who is there sees any benefit in our doctors?" He then tells Sanâ'î that he is master of a more excellent wisdom; "the poets are the chiefs of speech ;" "the gift of the poets comes from the piety of the parents ;" "verily from poetry comes wisdom ;" and will have none of such sayings as "poetry is of the affairs of Satan ."

As to the third part of the tradition,and pious descendants to invoke blessings on him after his death , Ahmad says, "The sons which suffice are thy sons; what son born in the way of generation and begetting is dearer than thy sons, or more honoured? Who has ever seen children like thine, all safe from the vicissitudes of time? The sons of poets are the poets' words, as a former master has said--

'A learned man never desires son or wife Should the offspring of both these fail, the scholar's offspring would not be cut off.'

A son according to the flesh may be a defilement to a family; but the son of intelligence and wisdom is an ornament to the household. These sons of yours you cannot disown."

He then asks Sanâ'î why he has thus become a recluse, and indolent and languid. This languidness is indeed preferable to a total heedlessness and forgetfulness of God, though Mutanabbi has said--

"I have not seen anything of the faults of men like the failure of those who are able to reach the end ."

He asks Sanâ'î not to bring forward the saying, "Laziness is sweeter than honey ," but to bestir himself and collect and complete his poetical works.

Sanâ'î tells us that he submitted himself to the advice of his friend, but brought forward the difficulties of house and food, since the work could not be performed friendless and homeless. Ahmad b. Mas`ûd thereupon built him a house, gave him an allowance for his maintenance for one year, and sent also a supply of clothing. He was therefore enabled to complete and arrange his writing's free from all care and anxiety. The preface ends with the praise of his generous friend.