THE ORTHOGRAPHY OF THE SAMARQAND CODEX

A. JEFFERY and I. MENDELSOHN

THE LIBRARY of Columbia University has recently acquired a copy of the Pissareff photographic reproduction of the famous Samarqand Codex of the Qur'an1, which has made it possible for the writers to take up the long overdue task of a re-examination of the text of this unusually important Codex.

This Codex was introduced to the attention to the learned world in 1870 by a notice in Petzholdt's Neuer Anzeiger für Bibliographie und Bibliothekswissenschaft, where, in the number for that year, on p. 372 we read -

818 Aus St Petersburg

hat die kaiserliche öffentliche Bibliothek von dem General-Gouverneur von Turkestan, Generaladjutanten v. Kaufmann in Samarkand, ein sehr werthvolles Geschenk, nämlich einen alten bisher in der Moschee Chodscha - Achrar aufbewahrten Koran in kufischer Schrift ohne Punkte und Vocalzeichen, erhalten, der über 1200 Jahre alt und von Osman selbst geschrieben sein soll.

The local legend regarding this Codex is that it was brought to Samarqand by Khoja Akhrar himself­ when he removed there from Tashkent, and when his Mosque was built there this venerable Codex was placed therein. Khoja Akhrar, whose real name, it seems, was 'Ubaidallah, lived in Tashkent in the latter half of the fifteenth centu­ry and was, towards the end of his life, the local Pir of the Nakshbandiyya Order of Dervishes. His possession of the Codex was due to a disciple of his Order, who, after accomplishing the duties of the Pilgrimage to Mecca, decided to extend his journey to Constantinople and return home from there. While in Constantinople it so happened that by the use of a prayer taught him by his Pir, he was instrumental in curing the Caliph of that day of a dread disease. In gratitude for the cure the Caliph offered him anything he might choose to take from the Treasury. He chose the ancient Qur'an said to have belonged to the third Caliph Uthman, and indeed to have been the copy which he was reading when he was murdered, the stains of his blood being visible on the pages which were open at the moment the murderers attacked him. This precious volume he brought with him to Tashkent, where for many years it was exhibited as an object for the veneration of pious Muslims, but when the head of the Fraternity removed to Samarqand the Codex accompanied him there. In the Khoja's Mosque at Samarqand it lay in public and was stroked and kissed by the pious as a source of blessing2.

Von Kaufmann's letter which accompanied the gift to the Imperial Library in St. Petersburg has been preserved, and gives an account of the ac­quisition of the MS. It is addressed to the Minister of Public Instruction, and is accompanied by two depositions made by the Ulama' at the Khoja Alkrar Mosque.

Chancellery of the Governor General of Turkestan,
Division-24th October 1869.
Journal: No.182. City of Samarqand.

His Excellency, the Minister of Public Instruction.

The Commander of the Zariavshansky District has handed over to me a Qur'an, written on parchment in Kufic characters without diacritical points or vowels, which previously was in the possession of the Mosque of Khoja Akhrar in Samarqand. Being aware of the great value of this Qur'an, and its sacredness in the eyes of the Muslims, Major General Abramov commissioned the Commander of the Samarqand District, Lieutenant Colonel Sierov, to investigate whether the acquisition by us of that manuscript would in any way violate the religious susceptibilities of the community. The 'Ulama' of the Mosque and certain honourable citizens testified:

1) that this Qur'an, though it was permanently deposited in the Mosque of Khoja Akhrar, did not really belong to it, but was regarded as the possession of the Crown, being the property of the Emir of Bokhara.

2) that this Qur'an is at present of no importance either to the Muslim community or to the Mosque. Formerly (indeed, very long ago) it used to attract many worshipers, but lately only the Emirs ar­riving at Samarqand have worshiped before it.

3) that nobody is able to read it, and that for many years it has been lying around without any use.

Thereupon Major General Abramov received the book, and in return for it donated from his own money 500 kokans (100 roubles), with which the clergy of Samar­qand were completely satisfied.

In view of the fact that such a book may, from a bibliographical point of view, be of great value to the scholarly world, I hasten to send the Qur'an thus ac­quired to your Excellency, together with the depositions of the two 'Ulema' of the Mosque of Khoja Akhrar, Mullah Abdul Jalil and Mullah Mughin Mufti, as ar­ranged by me, describing the origin of the Qur'an and how it came to the Mosque of Akhrar, and I humbly ask you, Sir, to deliver the book with the enclosed depositions, in my name, as a gift to the Imperial Library.

Signed-Adjutant General von Kaufmann.
Countersigned-Director of Chancellery
Major General Gomzin.
Correct: Secretary Diakov.

The accompanying depositions of the two Mullahs give the story as outlined above, and repeat the claim that it was the Qur'an of 'Uthman. As a goodly number of other Qur'ans, however, have at various times turned up in different parts of the Islamic world, all purporting to show the traces of the blood of the third Caliph 'Uthman upon certain pages, and thus be the genuine 'Uthmanic Codex, the Imam, which he was reading at the time of his death, this may only be pious legend first invented for this particular Codex at Samarqand itself. That it came there from Constantinople, is not however, unlikely.

In 1891 in Vol. VI of the Zapiski Vostochnago Otdieleniia Imperatorskago Russkago Archeologicheskago Obshchestva (St. Petersburg, 1892) ­pp. 63-133, A. Shebunin gave an account of the examination of the peculiarities of its orthography. The publication of this article gave rise to a great deal of discussion as to the relationship of the text represented in this Codex to that in the ordinary lithographed editions in use throughout the Muslim world. So great indeed was the interest excited that in 1905 S. Pissareff was encouraged to publish a facsimile edition, which he did by photographic process after having carefully inked in those places on some of the folios where the writing had been almost obliterated by the greasy hands of the faithful stroking the pages to secure blessing. Only fifty copies of this facsimile seem to have been made, of which only twenty-five were offered for sale (Chauvin, Bibliographie, X, No. 94). The reproduction is printed on papier d'ivoire 50 cm x 67 cm with all the decorations of the original reproduced in colour. The title-page reads-

Coran Coufique de Samarcand érit d'après la tradition de la propre main du troisième calife Osman (644-656), qui se trouve dans la Bibliothèque Impèriale publique de St Petersbourg. Edition faite avec l'autorisation de l'Institut archéologique de St Petersbourg, par S. Pissareff. St Petersbourg, 1905. (Facsimile.)

Muslim savants have frequently asserted that Pissareff in his reinking of the dulled folios deliberately made alterations in the text, but an examination of the facsimile shows that while some mistakes due to ignorance have been made here and there in the process of reinking, there are no adequate grounds for this charge of deliberate alteration.

When Shebunin made his study of the orthography of the Codex he used Flügel's Quran as his standard for the text, and the first edition of Nöldeke's Geschichte des Qorans as his authority for the older Kufic form of text. There was, of course, nothing else for him to do, for the Muqni of ad-Dani had not then been printed, even Musa Jarullah's text of the 'Aqila of ash-Shatibi was only published at Kazan in 1903, and the oriental lithographs of the Qur'an available to him differed so much among themselves in matters of orthography, that though they might have been preferable to Flügel, in that they did offer some consistent form of Oriental tradition, Shebunin had nothing to guide him in his choice among them. At present however, we are in a better position, and so in 1926, when publishing the first fasciculus of the third part of the new edition of Nöldeke's Geschichte des Qorans, Bergsträsser announced his intention of taking up a fresh examination of the Samarqand Codex in the light of our more advanced knowledge of the early Qur'anic orthography3. His untimely death left this, as so many other promised studies, uncompleted, and it is a peculiar pleasure to be able in this present study to carry to completion a plan that he had long had in mind.

Unfortunately we are now dependent entirely on the Pissareff facsimile, as the original Codex has disappeared. Indian Muslims had been much an­gered at the attention given to this Codex and is supposed deviations from the standard text, and had made many attempts to have the Codex given back to the Muslim community. In 1917 they succeeded in getting the consent of the Bolshevik leaders to the return of the Codex to its former resting place, and somewhat later it was handed over to the Muslim leaders in Petrograd for transmission to Samarqand. In Vol. LI of the Revue du Monde mussulman (1922) p. 10 we find the notice:-

Restitution aux musulmanes du Coran d'Osman.

Et pour mieux attirer les bonnes dispositions du monde musulman, le Gouvernement bolsheviste, par decret du 9 décembre 1917 (No 6 art. 103), ordonnait de restituer au Congres regionale des Musulmanes de Petrograd qui en avait fait la demande le Coran très sacré d'Osman, qui avait autrefois été deposé à la Bibliothèque nationale de Pétrograd apres son transfert do Samarkande.

Since then nothing has been heard of it. Musa Jarullah in a private letter informs us that he has heard of it being seen at Tashkent, its original home, but it has been impossible to get confirmation of this rumour. 'Abdallah az-Zandjani, on the other hand, in his Tarikh al-Quran, p. 46 (Cairo, 1936), asserts that it was taken from Petrograd to England and stored there, but one may suspect that this is merely a confusion with the transfer to the British Museum of the Codex Sinaiticus4.

From the Pissareff facsimile and the article of Shebunin in the Zapiski, however, since Shebu­nin's study, which was made from the actual text of the Codex before it had been retouched, enables us to correct in places mistaken reinkings, we can make with fair success the needed re-examination.

Originally the Codex was a complete Qur'an, written on thick, strong parchment folios averaging 68 x 53 cm in size, with the written portion averaging 50 x 44 cm. Only 353 folios were left, how­ever, when General von Kaufmann secured the Codex, and of these only fifteen were quite whole without any paper mending, viz. folios 210, 214, 215, 218-220, 232-235, 237, 238, 240, 243 and 246. Many folios had been damaged by dampness, and others were worn, and had been mended with paper, a thick, soft cotton-paper, which looks very much like parchment. Sixty-nine folios, which were missing entirely, had been replaced by folios made of this paper. The portions of the text which survive in the Codex are -

The paper leaves, which were later additions and may be neglected for out purpose of comparison are the following-

Folio 90 containing IV 33/37 'aqadat to 36/40 seems to have been added still more recently than the foregoing paper leaves. Folios 2, 6-7, 46-58, 89, 92, 112, 183, 315, and 316 consist of approxi­mately half paper and half original parchment, according to Shebunin. Unfortunately it is not possible to distinguish in the facsimile where this paper mending begins or ends, so that it is always possible that some of the peculiarities of orthog­raphy which we note, are due to the later hand which did the mending, and not to the original scribe of the Codex, for where we cannot check from Shebunin's remarks or from the style of the writing, we are at a loss.

THE SCRIPT

As can be seen from the facsimiles, the script is large, straight, and well-proportioned Kufic (i.e. the style of writing which became specialized in Kufa for the writing of Qurans), and is fairly uniform, On some folios, particularly in the early part of the Second Surah and the beginning of the Seventeenth (ff. 231 ff.), the writing is in a smaller more rapid hand, but Shebunin is doubtless right in thinking that they are even so by the hand of the original scribe, and not the work of another hand. The scribe has his own peculiarities. At times his Kaf is hardly to be distinguished from a Ta. His 'Aim at times has an open head both medially and finally, and sometimes is confusingly like when it occurs initially or medially. In the case of Hamza he is quite uncertain about the kursi, sometimes providing one and at others not, as e.g. in and , and often using an Alif - as kursi where we should expect a Waw or a Ya, while final Hamza is very commonly neglected altogether. A final Ya may turn to the left as we normally expect or may turn in under to the right in the fashion that has become common in writing Urdu .

On the whole, diacritical points are few, though every now and again for a few folios they become relatively numerous. These diacritical points where they occur are certainly contemporary with the original writing, at least in the great majority of cases. They are not in form of dots but strokes ( for ), but this is more in appearance than in reality for with the broad cut reed pen necessary to produce letters the thickness employed in this Codex, the pressure above a letter which with our pens would produce a dot naturally produced a thin line. Thus there will be one thin stroke above for a Nun or Fa' or Ghain two above for a Ta' or Qaf, and three above for a Tha' or Shin, and so on, with corresponding strokes below for the Ba', Ya' and Jim. Occasionally there are mistakes in the putting of these strokes, as when on fol. 23v a medial ba' is marked with a stroke above instead of below, or on fol. 26v where a ta' is marked with one stroke only instead of two, or on 22r where a tha' has but two strokes instead of three, or 32v where a lam is wrongly marked with a stroke beneath. These, however, are purely scribal mistakes, and have no significance. Indeed it is just possible that they are due to Pissareff's inking in, though so far as one can judge they seem to go back to the original scribe. By far the commonest letter to be marked is Nun, seldom when it occurs initially, but very commonly when it occurs medi­ally or finally or in the ending -na. The next most commonly marked letter is Ta'. Some letters, such as Dhal, Za', Dad and Ghain are very rarely pointed.

Other signs such as the shadda, sukun, wasla, etc., and the hamza where there is no kursi indi­cating its presence, are entirely lacking, and there are no signs to indicate the vowels. The scribe has no scruples about breaking up words, filling out his line with as many letters of a word as he needs, and finishing the word on the next line that he begins.

VERSE DIVISION

The verse division of this Codex is in general that of the Kufan School but the scribe was some­what careless. Where he does mark the verse end­ings his ending is usually that of the Kufans, but he will frequently run on for verse after verse without remembering to put in any sign of verse ending. The sign he uses is the commonly known series of oblique parallel strokes , sometimes more, sometimes fewer, and in one or two places only a single oblique stroke; but their number has no significance. At the conclusion of roughly ten verses he places a coloured rosette, sometimes ac­companied by the strokes indicating a verse end­ing (as at II,81/75, III/105; IV, 111, etc.), but more frequently not. Sometimes he forgets his rosette altogether (c.f. II, 109/103), and occa­sionally he has a blank space left for a rosette but nothing has been filled in (e.g. III, 180/176), and at II, 171/170 where a rosette would normally appear he has just drawn a black circle around his strokes . At XXVII, 68/70, XXXVII, 60/58, 182 the strokes are at the end of the verse and the rosette appears in the margin against that line of writing. These coloured rosettes seem to have been put in later than the original writing of the Codex.

The following peculiarities of his verse-marking may be noted -

II, 102/96. The normal Kufan ending of the verse at ya'lamuna is not marked, but one is marked within the verse after khalaqin where has a pausal sign , and the lithographs with the Sajawandi system of pausal signs have .

109/103. A rosette comes after lahum where no verse ending, or even pause, comes in any of the known systems. After the next word ul-haqq, however, there is a pausal mark in , and a in the Sajawandi system.

III, 145/139. Besides the normal verse ending at ash-shakirina, there is here another at mu'ajjalan where has a pausal sign , and the Sajawandi system a .

IV, 12/13. There is a verse ending marked after the first dain, where Flügel ends v. 13 and where has a pausal sign , and Sajawandi a . There is also a verse ending marked after the second dain, where Flügel ends v. 14, and where has and Saj. .

IV, 81/83. In the middle of the verse there is an ending marked after yubayyituna, where has only a pausal and Saj. a . The normal sign for verse ending probably came after wakilan as it should, but the page is defective here.

141/140. After having noted no verse endings at the end of 139 or 140 or at the normal ending of 141, there is one placed after al-muminina in the middle of v. 141/140, where has only a pausal and Saj. a .

VI, 73/72. Besides the mark at the normal verse ending after ul-khabiru, one is marked after fa yakunu where Flügel ends his v. 72, but where has only a and Saj. a .

VII, 89/87. There is no mark at the normal ending after al-fatihina, but within the verse one is placed after 'ilman, where has only a pausal , and Saj. a .

XI, 86/87, 88. There is no mark at the normal ending of the verse after bi hafizin, nor at the normal ending of v. 87/89, i.e. after ur-rashidu, but within 86 one is marked after mu'inina where Flügel ends v. 87, but where has only a and Saj. a .

XVI, 23/24, 25. Besides the normal verse end­ing after al-mustakbirina one is marked after yu'linuna where Flügel ends v. 24, but where has only a and Saj. a .

91/93. Besides the normal verse ending after raf'aluna one is marked after kafilan, where has only a causal sign and Saj. a .

XVII, 35/37. Besides the normal verse ending after ta'wilan, one is marked after il-mustakimi where has only a . and Saj. a .

82/84. Besides the normal verse ending after khasaran, one is marked also after lil-mu'minina where expressly marks it to show no pause is to be made, and Saj. does likewise.

XVIII, 2. There is no mark at the normal verse ending after hasanan, but there is one after al-mu'minina where neither nor Saj. mark any pause.

98. Besides the normal pause at verse ending after haqqan) one is marked after dakka'a where has only and Saj. a .

XIX, 17. There is no mark at the normal verse ending after sawiyyan, but one is marked in the middle of the verse after hijaban, where no pausal sign is given in or Saj.

41/42. There is no mark at the normal ending of the verse nor after the next (v. 42/43) but in the middle of v. 41/42 one is marked after Ibrahima, where has only the sign and Saj. .

XX, 53/55. Besides marking the normal verse ending after shatta there is one marked also after subulan, where no pausal sign at all is given in either or Saj.

86/88. Besides marking the usual verse ending after mau'idi, two other verse endings are marked within the verse; one after 'asifan where Flügel marks the end of his v. 88, and where has a and Saj. a ; and another after hasanan where has a and Saj. a . But in compensation no mark occurs here at the normal ending of v. 87/90 after as-Samiri, where Flügel also marks no verse ending.

88/90. The verse ending is marked after Musa, whereas it ought normally to come after the next word fa nasiya. This is doubtless merely a scribal error.

123/121, 122. Besides marking the normal verse ending after yashqa, one is also marked after hudan where Flügel ends his v. 121, but where neither nor the Sajawandi system has any mark of pause.

XXXIX, 7/9, 10. Besides the mark for the normal verse ending after is-suduri there is one marked also after ta'maluna, where Flügel ends his v. 9, but where has only a and Saj. a .

XLII, 48/47. Besides marking the normal verse ending after kafurun, one is also marked after hafizan where has only a pausal and Saj. a .

The most striking fact in this list is the number of coincidences of verse endings in the Codex with those adopted by Flügel in his text. There are further coincidences in that II, 40/38, 67/63, 78/73; III, 38/33, 131/126, 196; IV, 3, 27/32, 41/45,118; V,101; VI, 66; VII, 105/103; XI, 74/77,82/84,118/120; XIV,43/44; XVII,104/106; XVIII, 2, 23, 32/31, 84/83; XX, 33/34, 72/75, 78/81, 87/90, 92/94, 106, 116/115; XL, 53/56, 73/74; XLII, 32/31, this Codex has no verse endings marked where Flügel marks none but does6. Since we are entirely in the dark as to the source from which Flügel drew his verse divisions, these coincidences are significant. Flügel's verse endings agree with none of the known systems whose tradition has come down to us, nor with any that we have been able to trace in the Masoretic literature under the section Ru'us al-Ayy, and it has been generally assumed that he selected his verse endings on an arbitrary system of his own. The number of agreements between his system and that followed in this Codex, however, suggest that he may have been following the system of some MS in his possession which may have followed some divergent Oriental tradition. It must be ad­mitted, however, that the table Shebunin constructs of the divergences between this Samarqand Codex and the Flügel text in the matter of verse endings, is equally long and imposing, so that it is obvious that the question of Flügel's system of verse divi­sion awaits further elucidation.

In placing the rosettes at roughly ten verses apart, the scribe was following the ancient practice of indicating the 'ushr or tenths, which is possibly the earliest of the various systems of verse group­ing. It is clear that he was following a system and not just counting verses, for in many places his rosette comes where it ought to come on the Kufan system of marking the tens, whereas between two rosettes he himself has marked more or less than ten verse endings. For example in Sura XVIII there is a rosette at v. 10 after rashadan and an­other at v. 20 after abadan, where they would normally come according to the Kufan system, but in the Codex only seven verses are marked between them. As these rosettes are witness to an early 'ushr system it is worth while listing them, though with the remark that the witness is not as perfect as it might have been, for the scribe has often forg­otten to put in a rosette in places where, even on his own counting, he has gone well beyond ten verses. The list following is corrected from that of Shebunin.

Fol. 4v after sadiqina (II,31/29); 7v after zalimuna (51/48); 9v after ya'taduna (61/58); 11r after yaf'aluna (71/86); 12v after khaliduna (81/75); 16v after ya'lamuna (101/95); 18v after lahum (109/103); 19r after sadiqina (111/105); 21r after al-khasiruna (121/115); 23r after al-alamina (131/ 125); 28r after ta'amuna (151/146); 29v after ajma'ina (161/156); 47r after yasha'u (III,40/35); 53r after ta'lamuna (71/64); 56r after ash-shahidina (81/75); 58v after kafirina (100/95); 66r after az-zalimina (140/134); 69v after al-mu'minin (160/154); 74r after khabirun (180/176); 79r after tuflihuna (200); 86v after mubinan (IV,20/24); 103v after hakiman (111); 109r after hamidan (131/130); 113v after mu'minuna (V,88/90); 117r after rahimun (98); 123v after al-hakimu (118); 126r after yalbisuna (VI, 9); 128r after tushrikuna (19); 133r after mustakimin (39); 137v mubinin (59); 140r after yakfuruna (70/69); 144v after bi kafirina (89); 167r after as-sajidina (VII, 11/10); 171v after ya'lamuna (32/30); 174r after az-zalimuna (41/39); 177r after yajhaduna (51/49); 184r after musrifuna (81/79); 187r after jathimina (91/89); 189r after al-kafirina (101/99); 190v after muftaruna (XI, 50/52); 193r after mujibun (61/64); 195r after Ya'quba (71/74); 197r after bi qaribin (81/83); 199r after wadudun (90/92); 201r after hasidun (100/102); 203r after muribin (110/112); 205r after az-zahidina (XII, 20); 207r after al-'awwalina (XV, 10); 208r after bi-raziqina (20); 209r after 'ajma'una (30); 210r after al-mukhlasina (40); 211v after al-ghabirina (60); 212r after al-alamina (70); 213r after al-mursalina (80); 214r after tusimuna (XVI, 10); 215r after yukhlaquna (20); 217v after al-muttaqina (30/32); 219v after fa yakunu (40/42); 222r after al-hakimu (60/62); 224r after qadirun (70/ 72); 226v after hinin (80/82); 228r after tadhak­karuna (90/92); 229v after mushrikuna (100/102); 232r after 'aliman (XVII, 10/11); 233v after mahz­uran (20/21); 234v after basiran (30/32); 235v after 'aziman (40/42); 237v after kabiran (60/62); 239r after tafdilan (70/72); 240r after nasiran (80/82); 241v after yanbu'an (90/92); 243r after qaturan (100/102); 244v after takbiran (111); 246r after rashadan (XVIII, 10/9); 248r after 'abadan (20/19); 250v after 'amalan (30/29). 252r after talaban (41/39); 254r after 'adudan (51/49); 256r after saraban (61/60); 257r after 'imran (71/70); 259r after saddan (94/93); 260v after waznan (105); 263r after maqdiyyan (XIX, 21 ; 264v after hayyan (31/32); 265v after Ibrahima (41/42); 267r after shai'an (60/61); 268r after siliyyan (70/71); 269v after maddan (79/82); 270r after 'iddan (89/91); 271r after rikzan (98); 272r after ya Musa (XX,11); 273r after al-'ula (21/22); 273v after 'azri (31/32); 274v after ya Musa (40/42); 275v after al-'ula (51/53); 276v after iftara ­(61/64); 278r after 'abqa (71/74); 279v after ihtada (82/84); 281r after Musa (91/93); 282v after zurqan (102); 283v after dhikran (113/112); 284v after hudan (123/121); 286r after 'abqa (131); 287v after ta'buduna (XXVI, 70); 289r after lil-muttaqina (90); 289v after shafi'ina (100); 291v after ar-rahimu (140); 293r after 'ajma'ina (170); 294r after il-'alamina (180); 295r after mu'minina (190)' 295v after al-mujrimina (200); 297v after al mursaluna (XXVII, 10) 299v after al-gha'ibina (20); 300r after ir-rahimi (30) 301v after yuslihuna (48/49) 303r after ul mundharina (58/59); 305v after ul 'awwalina (68/70) 306v after ul-'alimu (78/80) 308r after muhtaduna (XXXVI,21/20); 309v after yarji'una (31); 310v after il-mashhuni (41); 312r after yansiluna (51); 313r after mustaqimun (61); 314v after malikuna (71); 3l7v after ud-dini (XXXVII,20); 318v after taghina (30/29); 319v after al-mukhlasina (40/39); 320r after yatas 'aluna (50/48); 320v after ul-­azimu (60/58); 321v after yuhra'una (70/68); 322y after as-salihina (100/98); 323v after il-­muhsinina (110); 324r after Haruna (120); 324v after 'il Yasina (130); 325r after il-mashhuni (140); 326r after shahiduna (150); 326v after il-mukhlisina a after il-mukhlisina (160); 327r after ya'lamuna (170); 328r after yasifuna (180); 329v after al-ahzabi (XXXVIII, 11/10); 330v after ul-mihraba (21/20); 336v after ya'lamuna (XL, 70/72); 338v after tunkiruna (81); 340r after ta'i'ina (XLI 11/10); 342r after ul-jabiru after shakurin (33/31); 350r after turja'una (21/20); 346v after ul-kabiru (XLII, 22/21); 348v after shakurin (33/31); 350r after il-'umuri (43/41); 352v after ul-'umuru (53).

Shebunin notes rosettes also on fol. 3r after II, 21/19 fol. 288r after XXVI, 80; fol. 344v after XLI, 31, where none are visible on the pages of the Pissareff facsimile. On the other hand he omits to catalogue those on fol. 18v, fol. 58v, fol. 254r, which are there quite plainly in the facsimile. It is possible that these omissions on his part are purely mistakes due to oversight, and it is also probable that the rosettes that are missing in the Pissareff facsimile may be due to further deterioration of the Codex between the time when Shebunin examined it and Pissareff's work in reproducing it.

Within the rosettes are crudely formed Arabic letters used as numerals, which once doubtless noted the numbers of the a'shar in the original Codex in its complete form, but which now, with so many folios of the original missing, present no sort of sequence. In the Columbia copy of the facsimile folios 138, 139, 141, 143, 194, 197, 207, 209, 210, 218, 231, 238, 239, 276, 291 and 295 have been bound in back to front, i.e. the recto is the verso and vice versa, which only adds to the confusion.

A comparison with the Kufan and Basran 'ushr marks in the lithographed Qur'ans of Cairo and Stambul reveals that the system in this Codex does not agree with the later systems of either of these cities, coinciding sometimes with a Basran 'ushr mark, sometimes with a Kufan, and sometimes, perhaps more often with neither. The litho­graphs, it is true, do not always agree with one another on the matter, but none of the several examined showed any marked connection with the system here.

At the end of a Surah there is a coloured band of decoration stretching right across the page to separate the end of one Surah from the beginning of another. Such are preserved on fol. 79r; 244v 271r; 316r; 328r; and 352v. If the last words of the Surah do not fill out a complete line, the scribe fills in what remains of the line with this coloured decoration, so that his Bismillah for the next Surah will start at the beginning of a line. There is no rubric of any kind at the head of a new Surah. Each new Surah begins directly with the Bismillah, and there is no pause mark or space between it and the first words of the verse to follow. The so-called Mystic Letters are found, and again there is no break between the Bismillah and these letters or between them and the beginning of the verse. Needless to say, since there is no rubric at the head of a Surah there is no name or number attached to the Surah.

No pausal signs are used in the Codex, and there are no marginal indications of 'ajza' or liturgical divisions.

ORTHOGRAPHY

Shebunin in his account of the orthography compared this text with that of Flügel, but as Flügel followed no known Oriental tradition of Rasm in preparing his text, a comparison on that basis is almost valueless. Compared with Flügel's text this Samarqand Codex presented a great number of peculiarities, but when we compare it with the rules for Qur'anic orthography given in the Muqni of ad-Dani,7 and the Egyptian Government text (), which attempt to follow consistently the Kufan masoretic tradition of Rasm al-Masahif in Qur'anic orthography, we find that this Codex, while presenting numerous deviations, yet follows the general Kufan system with fair consistency. Where it deviates it presents numerous points of interest, so that a detailed comparison is of a cer­tain importance.

Only what remains of the original folios can be used in this comparison, and unfortunately our comparison will be to some extent vitiated by an element of uncertainty not present in Shebunin's case. He, as already mentioned, was able to make his comparison directly from the Codex, where it was possible to distinguish the places where the folios had been mended, but this distinction is not possible on the facsimile which is all we now have available. The folios which Shebunin notes as being entirely paper may be rejected and indeed they are generally distinguishable by the difference in handwriting, but where the parchment leaves have been mended to a greater or lesser extent with paper patches, the patching does not show up in the facsimile, and though it is sometimes possible to distinguish the later hand which filled in the writing on these patches, there yet remains an element of uncertainty which is unfortunate but inescapable.

In a later number of the Zapiski (Vol. XIV, 1901, pp. 119-154) Shebunin published an exami­nation of another ancient Codex of the Qur'an, No.534 of the Collection in the Khedivial Library at Cairo, which in many respects was closely similar to the Samarqand Codex in matters of orthography. This Codex had also been restored at various times, and not always by skilful hands, but it still has 248 original parchment leaves, besides 34 imitation parchment leaves, 61 leaves taken from another Codex and inserted to fill in missing passages, and 219 paper leaves supplied to complete the volume. We may label this Codex (C) and from the evidence of the original portions ef the 248 parchment leaves as they were scrutinized by Shebunin, use its evidence to check with the peculiarities we have before us in the Samarqand Codex.

II, 22/20. is written without the medial alif, i.e. . So in 165/160 and XLI, 9/8.
24/22. without the medial alif. So in 74/69; XI, 82/84; XV, 74.

25/23. is written with the final alif, i.e. contrary to Muqni' 23-25, which says that this word should have the final alif only in XLII, 22/21. In this Codex, however, it is written with the final alif in III, 136/130, 195/194,198/197; IV, 122/121; V, 119; XV, 45; XVI, 31/33; XVIII, 31/30; XIX,61/62; XX, 76/78; XXVI, 134; XXXVI, 34; XXXVII, 43/42, but elsewhere without the alif, even in XLII, 22/21 though the defec­tive state of the folio at this point makes the reading a little uncertain. C has it always with the alif. is written , but possibly this is on the paper portion, and in any case would be a scribal error without textual significance.

26/24. without the medial alif.

28/26. is written with the alif i.e. c.f. XLI, 36. In the alif of is omitted before pronominal suffixes in this Surah only, the alif being written in in V, 35; XLI, 39 and in in XXII, 66/65. The only one of these passages extant in this Codex is XLI, 39, where, however, it is written without the alif. In C the alif is generally written before suffixes.

30/28. without the alif. So normally in C.

35/33. without the hamza, i.e.

38/36. without the medial alif. nor­mally writes this without the alif before pronominal suffixes, but here and at XX, 123/ 122 it has the alif, obviously to distinguish from but the more primitive writing may well have been as here without the alif (see Muqni' 68), though this Codex has it in XX, 123/122, as does C.

41/38. without the alif.

49/46. has an alif for the hamsa, i.e. quite irregularly.

53/50. without the medial alif. So in C.

54/51. without the medial alif. So (bis) in V, 89/91, in XVI, 61/63, in XVIII, 58/57 and in XVIII, 73/72.

61/58. without the alif. So , in V, 96/97, but with the alif in V, 95/96, XVIII, without the medial alif.
19/18, as invariably in and .

64/61, is omitted before but this is probably merely a scribal error, if not due to a mistake on a paper patch in the folio.

65/61. has the alif but no kursi for the hamza.

68/63. without the alif.

69/64. without the alif.

74/69. without the alif. See under 24/22.

78/73. without the alif. So in 111/105 and IV, 123/122. C agrees.

83/77. without the alif. So C. has it here but is without it elsewhere.

99/93. with the alif of the fem. plural, and so in IV, 140/139; XVII, 101/103; XVIII, 17/16; XIX, 58/59; XXVII, 1, 12; XXXVI, 46; XL, 56/58, 69/71, but elsewhere without, as normally in . Shebunin noted the same inconsistency in C.

102/96. with the alif, though the corresponding is without, as in . Muqni 23 says that there was variation among the Codices as to this alif.

without the alif, as in C, and as it should be according to Muqni' 18.

is written , and so in IV, 15/19, 18/22; VII, 38/36, 40/38; XVII, 34/36; XVIII, 60/59, 60/69, 86/84; XXVII, 18, 32; XXXVI, 39; XLI, 20/19, though elsewhere it is written normally, as is in­variably the case in C.
without the medial alif.
with the alif, but in III, 77/71 without the alif, as normally in .

108/102. is written There were word, so that this may be an ancient variant numerous variant readings recorded for this in the text.

109/103. without the alif. So in III, 91/85, but with the alif in II, 161/156 and IV, 18/22. C has it without the alif, though according to Muqni' 13, 16 it ought to be written with the alif save in XIII, 42.

111/105. without the medial alif. See 78/73.

119/113. The text has here some word ending in but the edge of the folio is lost. There is no known variant here, so possi­bly this is to be taken as a mistake.

124/118. without the alif, as in 30/28. So in III, 55/48.

without the alif. So in XV, 79 and XXXVI, 12/11. writes it with the alif save in XVII, 71/73 where it agrees with this text. C writes without alif.

126/120. What is written seems to be but doubtless it is merely an error in the rewriting.

128/122. without the first alif. has it here, but writes it without in 200/196.

133/127. without the second alif, i.e. . So in XXVII, 67/69; XXXVII, 17, though elsewhere with the alif, as in . C also varies in this matter.

136/130. without the medial alif. So in 140/134 and III, 84/78. C agrees.

139/133. without the medial alif. C agrees.

145/140. without the alif (bis).

156/151. with the alif. So in IV, 72/74, 73/75; III, 172/166; XI, 81/83; III, 146/140; XLII, 30/29; XI, 89/91; V, 106/105; III, 165/159, 166/160. differs in its treatment of the alif in this word. In and it always has it (as does this Codex in XI, 89/91, the only place we can test); has it in (where we cannot test it in this text), but omits it in (which in this text has it in III, 166/160; IV, 73/75; XLII, 30/29); has it in (as does this text), but omits it in and (where this text has it in III, 165/159; IV, 72/74; V, 106/105; II, 156/151).

158/153. without the alif. So C, and so save in this verse. without the alif, but with it in V, 93/94; IV, 23/27, 128/127. Both and C write it with the alif throughout.

166/161. without the medial alif. So C, and so save in this verse.

171/166. is written with no room for the . This looks very much like the original writing, but it must be merely a scribal error.

172/167. . The nun has been omitted by scribal error.

176/171. without the medial alif, but is written with it in 137/131; IV, 35/39; XI, 89/91, XXXVIII, 2/1. Both C and have it throughout.

177/172. is written , but it may be due to a reinking.

III, 39/33. without the medial alif.

40/35. without the medial hamza, i.e. , though elsewhere it is written regularly. C always has the alif as kursi for the hamza.

without the alif. So in XIX, 5, 8/9.

50/44. written with two kursis for the ya. See 58/51. Shebunin notes that cases of this superfluous kursi occur in C in various forms of the word .

52/45. without the medial alif, though it is written with it in 192/189, and in in this verse. C writes it with the alif when it is without pronominal suffixes, but without it when it has the suffixes.

58/51. is written with two kursis for the ya'. So also in 70/63. See on v. 50/44, where the superfluous kursi occurs in the singular form.

61/54. without the alif. So also in 64/57, though it is written with it in 167/160 and V, 104/103. C always has the alif

65/58. without the alif So in 68/59 and in 73/66.

69/62. without the alif, i.e. ,though written with it in 72/65; IV, 113; and III, 154/148.

75/68. without the medial alif. So in IV, 20/24. without the medial alif. C would normally be without.

75/69. is written with two ya's i.e. as in many old Codices.

78/72. The words are omitted by the scribe, obviously by error, thinking he had already written them.

79/73. written without the alif, i.e.. So in VII, 32/30; XVII, 17/18, 30/32, 96/ 98; XIX, 61/62, 82/85; XXVII, 15, 59/60; XXXVIJ, 111, 122, 132, 171; XXXIX, 7/9; XLII, 23,/22, 25/24, 27/26 bis, 52, but else­where with the alif, as at XXXVI, 30/29; XXXVII, 40/39 etc. has the alif written throughout save in XLIII, 19/18, LXXXIX, 29; XXXVIII, 45 and in XIX, 65/66. C normally writes with the alif.
is written with the alif. C commonly has the alif in words of the form .

83/77. with the first long alif. So 133/127, 180/176; IV, 131/130, 132/131; VI, 12.75; XI, 107/109; XVIII, 14/13, 26/25, 51/49; XX, 4/3; XXVII, 65/66; XXXVI, 81; XLI, 12/11; XLII, 49/48, 53, though elsewhere he writes as with no long alif. C always omits the long alif. has second long in XLI, 12/11.

87/81. with the medial alif. So in XV, 30; XVII, 92/94; XX, 116/115; XLI, 30, but elsewhere without as . C generally has it without the alif.

88/82. The final has been omitted by error.

92/86. without the medial alif.

133/127. written without the medial alif.

134/128. without the alif.

136/130. is written without any waw, i.e.. So in IV, 93/95, but see XX, 76/78.

138/132. written without the alif, which would be the normal writing in C.

140/134. without the alif.

144/138. is written which is pecu­liar, but see Muqni' 50.

146/140. without the medial alif.

147/141. without the medial alif. So in IV, 6/5. C would normally omit it.
without the medial alif. So in XLI, 29. So normally in C.

154/148. without the medial alif. So in IV, 34/38. So normally in C.

155/149. without the alif, agreeing with Muqni' 18. In 166/160, however, it is with the alif as in .

156/150. is written with the kursi for a ya' before the final . So in XV, 23 and XXXVI, 11/12, 78, though elsewhere it follows with only a final . C is incon­sistent in writing this word.

158/152. is written , which may be a mistake of the scribe, though C has it written thus in XXXVII, 68/66.

159/153. is written contrary to Muqni' 82. without the alif.

160/154. written without the alif.

is written . So in IV, 17/21, 85/87; V, 92/93, 99, 117; VI, 93; XVIII, 15/14 and in III, 179/173. This form is never found in C.

167/160. is written without the alif.
without the alif. C and always have the alif.
is written with an extra alif; [see Muqni' 18.]

168/162. is written without the alif, i.e.

172/166. without the medial alif. So in XLII, 38/36, but with it in III, 195/193.

174/168. is written probably a scribal error, though Shebunin notes that this superfluous alif occurred four times in C, in XL, 15; LIII, 6; LXV, 7; and LXXXV, 15 in the case of this word .

175/169. is written without the alif. So in IV, 34/38.

178/172. is without the alif. So in in IV, 137/136; XVIII, 25/24.

183/179. is written without the alif on the analogy of etc. above. It would normally be written without in C.

195/193. is with the medial alif, where has it without. In this word has the alif only in VI, 135, .

195/194. is written without the alif So in IV, 89/91; XVI, 41/43.

198/197. is written without the alif, i.e. , though it has the alif in 193/191 as and C have throughout.
200. without the alif.
without the alif.

IV, 1. is written without the alif, i.e. . So in VII, 48/46; XVI, 43/45, but elsewhere with the alif as and C.
without the alif.

3. is written , which was said to be the writing in 'Uthman's Codex. Muqni', 71.

4/3. is written as it was written in the Codex of Ibn Mas'ud (see Muqni' 45). So in VI, 38, 91, 93; XI, 57/60; 101/103; XV, 21; XVI, 35/37, 75/77, 89/91; XVIII, 70/69; XX, 50/52. £ has it thus only in XVIII, 23 but C has this form in VI, 38, 93, XVIII, 23; 70/69; XX, 50/52. Elsewhere in this Codex as in C it is written normally, as in . [In VI, 93 it is everywhere else .]

6/5. is written without the alif.
is written without the alif.

7/8. (bis) is written without either alif, i.e. . See Muqni' 18.

15/19. written . See II, 102/96.

16/20. is written without the alif.

19/23. is mistakenly written but this may he due to the reinking.
is written without the alif.

20/24. is without the alif.

23/27. is written with the alif, i.e.

25/29. is followed by an that does not belong there. It is quite obviously a scribal error and not a textual variant.
is written without the alif.

36/40. is written , which Muqni' 110 notes from some Kufan Codices.
is written without the alif. is written without the alif, i.e.

75/77. without the alif. So in 97/99; XVI, 28/30; XVIII, 35/33; XXXVII, 113.

92/94. without the alif. So in V, 95/96,
89/91 and C always write with alif.

93/95 as in III, 136/130 is written without any waw. 94/96. is written without the medial alif.

107. is written without the medial alif.
is without the alif. So C would normally write it without alif.

115. is written without the medialalif, has the alif in this word save in XVI, 27/29 where in this Codex also it is written without the alif.

119/118. is written without the medial alif. So in XVII, 46/48, XVIII, 11/10, but in XVIII, 57/55 it has the medial alif as has all through.

125/124. is written without the ya'. So in XIX, 41/42, 58/59. has it thus in Surah II only, but elsewhere with the ya' (see Muqni' 36). In C it is apparently always written with the ya'.

128/127. is written without the alif. So in VI, 35. C would normally omit alif.

137/136. is written with two alifs, but perhaps by mistake. is wrongly written with for , but this is possibly due to the reinking.

140/139. is written with waw, i.e. . This for the kursi of the hamza is an understandable writing, but Muqni' 60 expressly lists this verse as one of the places where there was an alif as kursi in all the early Codices.
is written without the alif.

142/141. is written without the alif, which is possibly the original writing since there are textual variants and (see Abu Hayyan Bahr III, 337, Ibn Halawaih 26) which assume an original form without alif from which these could be derived.

V, 89/91. (bis) without the alif. So in XVI, 61/63-; XVIII, 58/57, 73/72. is written without the alif. is inconsistent in writing this word, writing it with alif here and in LVIII, 4/5, but with­out alif in XC, 14. As C also would seem to have written it without the alif this may well have been the original writing.

90/92. without the alif, as would be the normal writing in C also.

94/95. is written without the medial alif, i.e. .
without the alif, as would be normal in C also.

95/96. There is only an unintelli­gible writing here, which looks like , but is possibly due to the reinking process, where the original was too faint to be properly traced.

96/97. is written without the alif, i.e.

106/105. is written with the medial alif, but as this alif comes at the end of a line, that may be the explanation. See on II, 156/151.
without the alif here and in the next verse, contrary to but in agreement with the regulation in Muqni' 18.

107/106. are both without the alif, agreeing in this with Muqni' 18.

111. is without the alif here and in v.112, though it had it in III, 52/45 as .

116. is written with the alif, so in C.

VI, 6. is written without the alif. So in XI, 52/54. This would be normal in C.

7. without the alif here and in v.91. So C would normally write.

14. is without the alif.

17. is without the alif.

18. without the alif here and in v. 61.

36. wrongly written by the scribe as though this again may be a case of wrong reinking of faded letters.

37. is without the alif, So in v.65 and in XVII, 99/101. is inconsistent in writing this word. Generally writes it as here with the alif, but without it in XXXVI, 81; XLVI, 33/32; LXXV, 40. See Muqni' 14.

38. is without the alif, as it is in C.

38. is without the medial alif. is inconsistent, writing it generally as here with the alif, but without it in XXIV, 35; XXV, 9/10, 39/41; XXIX, 43/42; XLVII, 3,10/11, 35/40; LVI, 23/22, 61; LIX, 21: LXXVI, 28. This Codex has it without the alif in XVI, 74/76 and XVII, 48/51 also, in both of which passages has the alif.

50. is written without the medial alif. So in XV, 21; XVII, 100/102; and XXXVIII, 9/8.

52. is written without the medial alif, but has it as . C is without the alif in LXV, 8.

59. is written without the medial alif. So C would normally omit the alif.
is written with only one kursi but doubtless by error.

71/70. is written without the alif. without the alif. So probably in C.

74. is without the medial alif. So in XXVI, 71. C would normally omit it.

85. is written without the alif writes it with the alif here but without XXXXVII, 123, where this text has the alif as .

91. without the medial alif as in the singular form in v. 7.

92. with out the alif, So in XIX, 31/32. This would agree with Muqni' 19, 82. The writing in is inconsistent, having it with alif here and at v.155/156; XIX, 32; XXI, 50/51; XXIII, 29/30, but without it in XXIV, 35.61; XXVIII, 30; XXXVIII, 29/28; XLIV, 3/2; and L, 9.

92. is without the alif.

93. is without the alif, i.e. .

94. is written here, as normally in IV, 12/15; VI, 100. has the waw ending only here and at XLII, 21/20. See Muqni' 61.

96. is written without the alif, which is perhaps the original writing, for there are textual variants and (Abu Hayyin Bahr IV, 185, Kirmani Bayan 79), which could only have arisen from a form without alif.
is without the alif. So C would normally have written.
is without the alif. So in XVIII, 40/38.

99. is written without the alif, though it has the alif in XVIII, 45/43; XX, 53/55. C is without it in LXXVIII, 15. has the alif throughout.

99. is written without the medial alif, is without the alif. C would normally omit it also.
is without the alif, which is perhaps the original writing, for has it without the alif in all passages gave this and II, 266/268. So C would normally write it without the alif.

100. is written with the alif, which again is possibly original, for has it with the alif everywhere save here and at XVI, 57/59; LII, 39 (see Muqni' 23). C always has it with the alif, and in this Codex it is with it in XVI, 57/59.

is written with the alif, and so in XVII, 43/45 also, though without alif as in XXVII, 63/64; XX, 114/113. Muqni' 19 says it should be without alif. 104. is written without the alif.
So in XVII, 102/104. C would agree.

VII, 10/9. is written with the alif.
17/16. is written without the alif, though in XVI, 48/50 it has it.

32/30. is without the alif. So in XVI, 66/68.

38/36. is written with the ele­ments separated. There was some dispute as to how this should be written. In the Muqni' 79 and Mansar al-Huda, the rule is given that this separated form should be written only in XIV, 34/37, but has it separated in IV, 91/93; and XXIII, 44/46, though joined everywhere else. C has it joined in V, 64/69 but separated in LXVII, 8.
is written without the alif, which may be original, for save in this pas­sage omits the alif Muqni' 13), and there were textual variants here (Ibn Halawaih 44, Kirmani Bayan, 85) which could only have arisen from a form without the alif.

40/38. is written without the alif, which may be original for there were textual variants (Abu Hayyan Bahr IV, 297, 298) which assume an original form without an alif.

43/41. (bis) is written with but one kursi

46/44. is written without the alif here and in 48/46. So C would have omitted alif,

54/52. is without the alif, agreeing with Munqi' 19 and with C. writes it without in LV, 78 and LXVII, 1. See on XLI, 10/9.

57/55. is written without the alif.
is without the alif, as would be normal also in C.

66/64. is without the alif, So in the next verse.

77/75. has been mistakenly written twice by the scribe.

79/77. is written without the alif, C would normally omit the alif also.

85/83. is without the alif. So also in XI, 84/85, 85/86. C also without alif,

XI, 53/56. is written without the alif, and so in XXXVII, 36/35.
56/59. is without the alif.

63/66. is written with the alif, i.e. . C would normally write as here.

75/77. is written with the alif, i.e. . C often kept the alif in such forms.

77/79. is written apparently on the analogy of for C writes it as in .

82/84 is written with the alif, i.e. and so in XV, 74.

84/85. is without the alif here and in the next verse.

87/89. is mistakenly written . Doubtless a scribal error.

88/90. is without the alif.

XV, 23. is written with the extra letter . See on III, 156/150.

79. is written without the alif, i.e. . See on II, 124/118.

86. is written with the alif, i.e. against the rule of Muqni' 18, but on the analogy of in V, 116.

XVI, 14. is written without the alif.

25/27. is without the alif, as is , here and in XX, 87/90.
is without the alif.

26/28 is written without the medial alif, though it has it in II, 127/121. is with alif here and at II, 127/121, but without it at XXIV, 60/59. C is normally without the alif in such forms.

61/63. is written with alif as kursi for the hamza, i.e. .

65/67. is written . C agrees with

66/68. is written without the alif.

80/82. is without the alif.
is without the alif.

80/82. is written without the alif. So normally C.
is without the alif. So nor­mally C.

89/91. is written . See IV, 4/3.

90/92. is written without the alif, i.e. . See Muqni' 50.

92/94. is written

117/l18. is written with the alif, though it is without it in III, 185/182, 197/ 196; V, 97/96; XVI, 80/82; XXXVI, 44; XLII, 36/3-4. This is not a word where She­bunin notes any inconsistency in C, which would normally write it without the alif.

VII, 5. is written without the alif, This is perhaps the original writing for writes it without alif everywhere save in this verse, and this Codex and C have it consistently without the alif.

15/16. is written without the alif. So in XXXIX, 7/9.

18/19. is without the alif.

23/24. is written without the alif, Muqni' 100, 101 notes that the early Codices differed on the spelling here.

24/25. written without the alif.

35/37. is without the alif. So in XXVI, 182, doubtless in the analogy of such forms as , VI, 7.

38/40. is written , probably by scribal error.

40/42. is written with the alif.

64/66. is without the alif. So normally would C write.

68/70. is written without the alif. So in XXXVII,8 and here in v. 83/85.
is written without the alif.

69/71. is written without the alif.

79/81. is written without the alif.

82/84. is written without the alif.

84/86. is without the alif.

93/95. is written without the alif. This is possibly original for has the alif here alone. See Muqni' 18, 101. C omits the alif.

100/102. is without the alif.

107/108. is written without the alif here and in the next verse.

110. is written without the alif, which is perhaps the original writing, for there were textual variants here (see Zamakh­shari on the verse, and as-Suyuti's Durr IV, 208), and shows much inconsistency in writing the word when it has attached pronouns. C seems usually to have had the alif.
is written without the alif.

XVIII, 16/15. is written . See Muqni' 54.

18/17. is written without the medial alif.
is without the medial alif.
is written with the alif. has the alif in V, 28/31 but writes the word without it elsewhere. C would normally omit the alif.
is written without the alif.

22/21. and and are all without the alif.
is written without the alif.
This would be normal in C.

29/28. is written without the alif.

31/30. is written without the alif.
is without the alif.
is written without the alif, here and in XXXVI, 56.

34/32. is without the alif, here and in v.37/35.

37/35. is written without the alif.

46/44. is without the alif here, though with it in other occurrences, as in and in C.

47/45. is written without the alif.
is without the alif; and so in the verse, 49/47.

49/47. is written without the alif.

53/51. is without the medial alif, which is perhaps original, as is without it in LVI, 75/74.

62/61. written without the first alif. is inconsistent, having the alif here and in II, 249/250; XLVI, 16/15, but without in VII, 138/134; X, 90.

64/63. is written without the alif. So in XL, 82, though elsewhere with the alif as in XXXVI, 12/11, where writes without any alif.

69/68. is written without the alif,
96/95. is written without the alif, as would be normal with C.

XIX, 10/11. is written without the alif. So would C normally write.

11/12. is without the alif here and in XXXVIII, 21/20, but with it in III, 39/33.

13/14. is written without the alif.

14. is written without the alif here and in v. 32/33, but with it in XI, 59/62.

17. is written without the alif here, but with it in VII, 46/44; XVII, 45/47 and XLII, 51/50.

29/30. is written without the alif.

37/38. is without the alif here and in XXXVIII, 11/10, 13/12; XL, 5.

54/55. is written without the alif.
59/60. is written without the medial alif.

71/72. is without the alif.

73/74. is written without the alif, but has it in II, 125/119; XVII, 79/81; XXXVII, 164.

75/77. is written without the alif, though elsewhere it has the alif as . C is inconsistent in its treatment of this word.

83/86. is written but this seems a scribal error.

97. is written without the alif, and so in XXVI, 195, but elsewhere with the alif as in . C is inconsistent, sometimes having the alif and sometimes not.

XX, 12. is written with the alif, which is perhaps original, for there were textual variants and (Marandi, 141; Abu Hayyan, VI, 231; Ibn Jinni, 49; Kirmani, 150) which assume an original alif in the text.

39. is written without the alif.

47/49. is written though probably only by error.

63/66. is written with the second alif, i.e. which is perhaps original, for has the alif in XXII, 19/20. See Muqni' 18.

63/66. is written with the second alif. It is so written in C also, and is probably th original form. See Muqni' 18.

66/69. is written without the alif.

69/72. is written without the alif, possibly the original form, for is without it in all places save here. See Muqni' 21.

71/74. is written with the alif, i.e. though Muqni' 12 says it should be written without the alif, and it is thus without it in XVII, 76/78.

76/78. is written with the waw, i.e.. See III, 136/130. Shebunin notes that C sometimes wrote with the waw and sometimes without. See Mugni' 61, 106.

85/87. is written without the alif here and in v. 87/90. This is perhaps original, though writes it with alif in v. ­85/87, 87/90, but in v. 95/96 without, as ­does this Codex.

94/95. is written See Muqni' 81. There were textual variants here (see Marandi and Kirmani 154) which show that there was confusion even in the earliest Codices. C writes as here.

97. is written without the alif.

106. is written without the alif.

108/107. is written without the alif of the fem. plu. ending, which is probably original for has the alif here contrary to the rule of Muqni' 23, though elsewhere, as at XXXI, 19/18, it follows the rule.

121/119. is without the alif, in acordance with the rule in Muqni' 18.

129. is written without the alif.

130. is written without the alif

XXVI, 92. is written joined, i.e. which is probably correct, for Manar 11, and Muqni' 77 are against here. C separates here and at XL, 73/74; LVII, 4; LVIII, 7/8, but joins in XXXIII, 61.

114. is written without the alif.

165. is without the alif. So in XLII, 50/49.

XXVII, 19. is written without the alif.

32. is without the alif,i.e. .

44. is written without the alif.

45/46. is without the alif, probably correctly; see Muqni' 18.

49/50. without the alif.

57/58. is written but possibly by mistake.

60/61. is written without the medial alif.

61/62. is written without the alif. C is inconsistent, writing it with the alif in XXXVIII, 60 and without it in XL, 39/42.
is written without the alif.
C normally omits the alif.

67/69. is written with the alif, which is perhaps the original form. is inconsistent, having the alif everywhere save in XIII, 5; XXVII, 67/69 and LXXVIII, 40/41. See Muqni'. 20. C always has the alif.
is written without the alif, but in the next verse with it as . It is without it also in XXXVII, 17. See on II, 13-3/127.

73/75. is written . Shebunin notes that C often had this superfluous alif.

75/77. is written without the alif.

XXXVI, 14/13. is without the alif, i.e. which is perhaps original.

22/21. is mistakenly written .

39. is written without the alif.

40. is written without the alif.

51. is without the alif

73. is without the alif.

XXXVII, 6. is written without the alif So normally in C.

7. is written without the alif, as would normally be the case in C.

10. is without the alif.

11. is written without the alif. So normally in C.

16. is written with the alif here and in v. 53/51, C always has the alif. is inconsistent, though it spells it generally without the alif.

25. is without the alif.

30/29. is written with the alif, which is perhaps correct. (has the alif in XXXVIII,55 and LXXVIII, 22 against Muqni' 23 and with alif in LI, 53; LII, 32 in accord­ance with Muqni' 24.

32/31. is written with the alif, possibly correctly, as has the alif in all passages save this.

36/35. is written without the alif.

38/37. is without the alif.

57/55. is written in accordance with the rule in Ibn Abi Dawud's Kitab al-Masahif, p. 214, but against Muqni' 82, 83 and Manar 12. C also writes it with here as also in XXXI, 31/30 and LII, 29.

75/73 written without kursi for the alif, i.e. So in C.

99/97. is written without the alif.

102/101. is written without the alif.

106. is written which is perhaps original. See Ibn Abi Dawud, p. 214 and Muqni' 62.

176. is written without the alif.

177. is written without the alif. So C would normally omit the alif

XXXVIII, 12,11 is written without the alif.

18/17. is without the alif.

22/21. is without the alif, probably correctly. See Muqni' 18.

is written .

XL, 4. is written with the alif i.e. . It is not one of the words which Shebunin marks as taking the alif in C.

XLI, 10/9 is written with the alif, contrary to C and and the rule in Muqni' 19. See on VII, 54/52.

16/15. is written without the alif.

29. is written without the alif, agreeing with Muqni' 18.

31. is written probably correctly: see Muqni' 40.

XLII, 22/21. is written without the alif. is inconsistent, writing it with alif here at VII, 171, 170; LXX, 1, but without the alif elsewhere.
is without the alif, which is probably original.

33/31. is written without the alif.

40/38. is written without the final alif, i.e. which is possibly the original form; but see Muqni' 61.

48/47. is written with alif, as it is in C.

XLIII, 8/7. is written .

The general principles of the orthography are thus those of the Kufan School as set forth by ad-Dani in the Muqni', and followed for the most part by the text. It is noticeable, however, that this Samarqand Codex agrees with the Muqni' not a few times where departs from its instructions, notably in not expressing the alif of the dual ending. In other respects, as Shebunin noted for the C text, it has correspondences with what we know of the old Basran Codex, so that we may safely say that the text is Iraqi in type, as against the Syrian or Hijazi tradition. Where we are able to check its readings against those listed in the Muqni', p. 106 ff. as characteristic of the great Metropolitan Codices, we find that it is almost always in agreement with those of Iraq as against those of the other centres, even in such readings as for in IV, 36/40 and for in VI, (33, where the Kufan reading is opposed by all the other Codices. The one exception is in XXXVI, 35, where it reads with and Basra and the majority of Codices, against the which Muqni' 113 gives as the reading in the Kufan Codex. An Iraqi origin is also indicated by the writing which was characteristic of the Codex of Ibn Masud, so long influential at Kufa. It agrees with the 'Iraqi Codices in often omitting an alif when hamza follows (Muqni' 21), in read­ing in XLI, 31 (Muqni' 40), in writing (Muqni' 40), in reading in XX, 130 (Muqni' 51) and in XX, 76/78 even against (Muqni' 61, 106), and in reading in XLII, 21/20 (Muqni' 61), though it disagrees with 'Iraq on this word in VI, 94.

On the other hand it disagrees with the 'Iraqi Codices in reading for in XVII, 93/ 95 (Muqni' 18), in reading without the alif in XLIII, 3/2 (Muqni' 20), perhaps in reading for in XLII, 22/21 (Muqni' 25), though the page of the MS is defective here, in reading with and the Madinan Codex (Muqni' 51), and in reading instead of in XVIII, 87/88 (Muqni' 61) to agree again with and the Madinan Codex. It disagrees with all the Codices in reading in XXXVII 106 (Muqni' 62), and has one curious agreement with 'Uthman's Codex, the Imam, in reading for in IV, 3 (Muqni' 71), though it definitely disagrees with what is recorded of the readings of Uthman's Imam in other passages, e.g. in writing the second alif in in XX,63/66 (Muqni' 16).

Other peculiarities, in so far as they are not pure errors, whether of the original scribe, or made during the process of one or other of the renova­tions of the text, seem to be nothing more than the natural peculiarities of a scribe working at a time when the minutiae of orthography were not so firmly fixed as they later became. One can thus safely date the Codex earlier than the time of ad Dani (444 A.H.), by whose time most of these minutiae had become fixed. The fact that its peculiarities are of the Basra-Kufa circle sug­gests that it must date from a time when the tradition of those Schools was beginning to take its characteristic form, and this would point the third Islamic century. Shebunin wanted to date both this Codex and C in the second or late first Islamic century, but this would seem too early. The fact that it has no vowel points does not necessarily point to a very early date, for Ibn Abi Dawud (316 A.H.) in his Kitab al-Masahif, p. 141 ff, records the prejudice there was in many circles against putting any marks, whether to dis­tinguish consonants or vowels, in Codices of the Qur'an. In this Codex, as already mentioned, the marks to distinguish consonants are by the original scribe, and not, as in some of the fragments of early Codices known to us, inserted by later hands, and though the scribe of this Codex is spasmodic in his marking of them, the fact that he marks any would point to a date later than that of the earliest Codices. Thus we shall be safe in assigning the Codex to some centre in 'Iraq, probably Kufa, early in the third Islamic century.

Footnotes

1 See I. Mendelsohn, "The Columbia University Copy of the Samarqand Kufic Qur'an," in Moslem World for October, 1940.

It was purchased from a local Russian bookseller, and is apparently one of the twenty-five copies of Pissareff's work which were put on the market, the other twenty-fi­ve having been presented to important Libraries and Institutions. It is an excellent copy, clean and complete. The reproduction, it may be mentioned, is the exact size of the original, is on heavy paper, and has an additional ornamental title-page, with the inscription Kalam Sharif, besides the title-page in Russian and French.

2 Eugene Schuyler, Turkistan (1876), Vol. I, pp. 256, 257; Landsell, Russian Central Asia (1885), Vol. I, p. 582.

3 Geschichte des Qorantexts, p. 8, n.1.

4 In Moscow News for June 12th, 1941, V. Nagel has an article on rare MSS in the Leningrad Library, and mentions this Codex as though it were still there (p. 20).

5 The verse numbering is given both according to the Kufan tradition of verse numbering as represented in Egyptian Standard Edition of 1344 A.H = 1925 A.D. hereafter referred to as the text, and 209 in that of Flügel, which is the verse num­bering quoted in most European works and almost all European translations of the Qur'an. Where both texts agree only one number is written.

6 The number of cases may even be more numerous than this, for in many cases the end of a verse comes the edge of a folio, and as the edges are badly broken it is often not possible to ascertain whether any mark for the end of the verse was there or not. In such cases we have assumed that it was, but this may be wrong.

7 Orthographic und Punktierung des Koran: zwei Schriften von Abu 'Amr Utmam ibn Sa'id ad-Dani, herausgegeben von Otto Pretzl. Istanbul, 1932.


Journal of the American Oriental Soceity, Vol. 63, 1943, pp. 175-195 (New Haven [etc.] American Oriental Society).
Further reading: A ‘Perfect’ Qur'an?, Chapter 9: Samarqand vs. 1924 Edition is a detailed comparison of the Samarqand Codex with the Cairo edition of the Qur'an, not only regarding orthographical questions.

Writings by Arthur Jeffery
Answering Islam Home Page