Aspects of the physical transmission of the Qur'ánin r 9th-century Sudan:Script, deco tatron, bindirg and paperbv Adrian Brockett

The University of Leeds possesses two complete copiesof the Qur'án from l9th-century Sudan in excellentcondition - Leeds Arabic ms. 301 and Leeds Univer-sity ms. 619. A wealth of information about the trans-mission of the Qur'án in the Sudan can be gatheredfrom them, but the latter has never been written about,and only a brief notice of the former has beenpubl ished 1.

Ms. 301 has no colophon, date, nor name of scr ibe,but the script is that used largely for Qur'án copies inwestern Sudan and West Africa. loosely calied SilrlanrMaghribt. This manuscript will be referred to as 'the

West African Copy'. Its script will be called 'lfriqi'.

and as would be expected. the text is in the transmis-sion of Warsh from Náfi'2. According to the pricewritten on the top page. it was acquired by (what isnow) Leeds University Modern Arabic Studies Depart-ment for 25 shill ings. This was in 1957, probably froma sale in London.

By contrast, ms. 619 has a clear, and poignant,history. I ts colophon is dated 1299 (: 1881 A.D.) andit was taken 18 years later from the massacre at UmmDubaykrat on24 November 1899. This was in south-ern Kordofan about 200 miles down the White Nilefrom Omdurman, near present-day Kosti. The Mahdihad declared holy war against the Turco-Egyptianoccupiers in 1881, so the useful life of this copy of theQur'àn exactly spans that remarkable period of power-ful religious revival. It will therefore be referred to as'the Mahdiyya Copy'. It is in the transmission of al-Durr from abu 'Amr3.

On the death of the Mahdi in

Mahdiyya. Wingate took over from Kitchener asGovernor-general in December and remained so untill9 l l4. The photograph probably shows the owner ofthis Mahdiyya Copy of the Qur'án. because accordingto a note in the hand of a British major (bintbashiT.E.N. Lewis) the copy was 'found in the saddle-bag ofan Emir who was killed near the Khalifa ... at UmDebrekat'. Major Lewis presumably brought the copyback to Britain, and it was subsequently presented toLeeds University Library in June 1929s.

This art ic le looks at some physical aspects of thesetwo manuscripts in turn their st l ' les of scr ipt . theirdecorat ions. their bindings. the paper thev u'ere u r i t tenon. and i t s vua termarksó. Ms. Arab . d . l4 l o f theBodleran Librarl is reÍèrred to here and there forcomparison ".

I Tsp wEST AFRICAN copy(letos ARABTc us. :or)8

Description oJ' the manuscript :

fol. 332 (163 bifolios, 6 folios); 220-230 x 160-1 6 7 . 5 m m . : w r i t t e n a r e a 1 5 0 - 1 6 0 x 1 0 0 - l l 0 m m . e :16-20 l ines per page; laid paper; bold I f r iqr hand inshiny black inkro. with diacritics in black, vocalisa-t ion in red. and hant:at al-qat ' in yel low; sÍra-t i t lesin the same hand but in red, with diacritics andvocal isat ion in black; marginal decorat ions in red,brown, yel low and black; 4larger decorat ions in'earthy' yellow, reddish brown and black (ff. lrb,8lb, 163a, 246a); strong, leather loose-cover bin-ding, stained reddish brown, with dark brown(almost black) associated with the tooling, ending inan envelope-flap and strap for fastening; the wholecontained in a rigid suede-leather satchel, with atriple flap, thongs and straps; no date.

l . l Tue scRrpr :

The script has a bold appearance, echoing the majes-tic western Kufi of the mushaf al-Hadina, from whichthe whole west Sudani tradition could well derive r 1.

The 'somewhat clumsy look' i2 of the I fnqr scr ipt can

1885 the leadership passed to his Khalifa Abdallahiwho maintained the Islamic state for fourteen years. In1896 the British and Egyptian forces under Kitchenerbegan their conquest, culminating in the massive victo-ry over the Khalifa on I September 1899 at Karari(often called the battle of Omdurman). What remainedof the Mahdiyya retreated for their lives into Kordo-fan, and were followed by the army under Wingate toUmm Dubaykrat, where they were gunned down asthey were praying. There is a macabre photograph ofthe Khalifa lying dead on his prayer-mat with hisgeneral the Emir Ahmad Fadrl, and others, also dead,close beside him. This was the final end of the

Manuscr ip ts o f the Midd le Easr 2 (1987) ! Ter Lugt Press . Donkersreeg 19 2312 HA Le iden, Nether lands . 1987 ISSN 0920-0401

46

largely be put down to the freedom of stroke that the

script shares with the northern Maghribi tradition, in

that it is the entire word that is formed rather than

individual letters 13. But whereas the strokes and

diacritics of the northern tradition are usually drawn

with a sharp nib or fine brush, those of the Leeds West

African copy are drawn with a blunter nib, and on

unpolished paper. Where a fresh, or newly sharpened'

nib has been used, the script appears more elegantra.

The letters are upright and vary little in height 1 s. A

ía'. for instance is often as tall as a kaf or lam, and the

tail of a nun or sad sweeps shallowly. The resultant

squat, dumpy appearance when compared with the fine

lines, variety in height of letters and curvaceous tail-

sweeps of Fasr Maghribí again might lead one at first

to think it clumsy. The long-term impression, however,

is one of primeval vigour and ingenuous unselfcon-sciousness. Nor does this overall boldness of scriptpreclude the presence of detailed indications regarding

tajv'íd like the double lengthening of ali.f ntaqst1ro and

the 2nd person plurai possessive before lutnt:ctÍ ol'qa! ' t6 and assimi lat ion of ni ln lo nl ln i before bA'\^ '

The best analysis of the or igins and history of the

Ifr ïqr scr ipt is b-v Bivart8. He argues that i t was in

general use in North Africa up until the Almohads in

the l2th century. after which it was preserved in

western Sudan, even though the extant Qur'án copiesfrom Bornu may only date from the early 17th cen-

tury. By the early 19th century (and indeed right into

the 20th) it was stil l the prevailing tradition over the

whole of Bornu, and into neighbouring areas iike the

ci ty of Kanole. Without known i l lustrat ions from

Sokoto and further west. i t is not possible to be sure

that Qur'án copies were written there also in the Ifrlqrhand, but judging by the conservat ism in Qur 'án cal l i -graphy it is likely. This does not contradict the fact

that the correspondence of the Fulani Sultans of

Sokoto was in a different, more secular script, JihAdí2o.Lack of firm knowledge of the extent of use of theIfrrqr script is the reason for calling the Leeds copy'West African' rather than by a precise title like 'the

Bornu copy' .Where the script of the Leeds West African copy

differs from Fasl MaghribT it sometimes resembles theolder Kufi, which lends support to the theory that theIfrrqr script is an earlier branch of the overall Maghribitradition. Other differences of the Leeds West Africancopy's script from Fasí Maghribí are: the lower strokeof the dal and dhal is more inclined; the teeth of the sínand shtn are tall and written by lifting the pen for each;medial

'ayn and ghayn are rounded rather than flat-topped2l ; fa' and qaf are on tall stalks and the initialand medial forms are the samel initial mrm is semicir-cular, and the medial and final forms join at thebottom; final ya'can be little different to final nun, andwhen it curves backwards. as in the word f it is usuallvverv short.

MIDDLE EAST 2 (1987)

All this is typical of the Bornu Ifriqr hand. Of those

examples published it is perhaps closest to the hand in

the correspondence of the Kanemi rulers of Bornu

whose reigns spanned 1812-9322.

1.2 DeconartoN

Verses are separated by trefoils in red ink fil led in

(and often over) in yellow23. The two bottom lobes are

round, the top is usually pointed. Every fifth verse is

separated by a ha' (khamsa) in black ink fil led, apart

from the very centre, with red. Every tenth verse is

separated by a roundel ('ashara) of two concentric

circles in black. the outer one with a thicker line, the

centre one fil led in with yellow and the space between

it and the outer circle having pairs of red strokes or

single dots top. bottom, and sides2a.Marginal decorations are circular or rectilinear.

Three circular devices are found on /. I b 2 s, each

different. as usual in West African copies, and each

with a double circumference. as also usual. Their inner

circles are fil led with simply drawn diamond-shapes,squares and/or segments in alternating brown andyellow. On this first page of the text their function is

decorative2Ó, but in the remainder they serve to indi-

cate the sixty hizb divisions, occurring every 10-12

pages. These sixty are each usually divided into eight

sections, mostly indicated in the margin by rectilinear

devices containing one of the let ters tha' . ha' , tha' , nun.

ha' , tha' , ba' . tha' . in that order2r. They are most

commonly contained within an outlined square whose

two sides extend up and down to at least twice, and

sometimes man,v. times the length of the sides. Trefoils

with tips extending into iong wavy finials centre on

these openended extensions. Lines are in red ink,

letters are in black. and the trefoils are again fil led in

with yel lou'28. There is a number of var iat ions on this

pattern. like diagonalled squares in place of the wavy

finials. as well as ciifferent ones like triangles2e.There is no frame around the written area, nor are

there either marginal decorations or frames to indicate

new srTras. Instead sÍra titles, locations and verse-

counts are rubricated, and along with their basmalqusually take up two lines or the best part of them. The

first verse of a sura always begins a new line. The

basmala of sura I is stretched to filI a line30.Larger decorat ions precede suras 2,7, 19, and 383r,

that is approximately at quarterly intervals in the text.Again all the patterns are purely geometric, eitherweave or diagonal, and in no way vegetative. The onesar suras 2 and 38 are single-panelled and occupy less

than half the written area, the other two are triple-panelled and take up almost the whole of the writtenarea. They are set within two, three or four frames,sometimes coloured in, sometimes not. They are allsimply executed and have an ingenuous charm. Thelines are drawn with a finer nib, or lighter touch, than

MANUSCRIPTS OF THF

ASPECTS OF THE PHYSICAL TRANSMISSION OF THE QUR'ÀN

those of the script. The weave-pattern present ln them

all is composed of three transverse sets of a crenella-

ting strapwork32 with their mirror-image, overlapping

with vertical sets of the same. There are between three

and four sets of each. The strapwork itself is unpain-

ted, but the small spaces between straps are fil led with

a reddish brown, and the larger central spaces with an'earthy' yellow. The decoration at sura 38 has circles

added in the larger central spaces, reminiscent of an

Umayyad design33. In the one at sura 19 the weave-

pattern is in a central panel flanked by two smaller

panels with linear, diagonal strapwork' The one al sura

7 is the converse, with the weave design making up the

two smaller flanking Panels.A similar, but tighter, weave pattern creating an

overall diamond design is found in the Bodleian ms'

Arab. d. 14134 and the lTth-century Maiduguri

copy3s. A different weave-pattern is found in the full-

page decoration preceding sura 19 in the Bodleian ms.

Arab. d. l4 l3ó. I t is absent from the Leeds copy but

common in the mainstream of Qur'án illumination

from Kufi copies right down to Safavid, Mughal and

Ottoman ones, especially in interlace borders. A

modified form of it is found in the decoration of the

printed West African copy in the British Library3?.Diagonal strapwork is also a recurring feature of

West Afr ican Qur'án decorat ion. On the decorat ionpreceding sura 7 of the Leeds copl there are seven

rows, four with crossed diagonals al ternat ing * i th

three with single diagonals3s. On the one preceding

sura 19 the diagonals are within squares. three rows of

nine squares in the top panel, and three of ten in the

bottom panel3q. In the top panel the diagonals are

single with bobbles on either side of the middle. In the

bottom panel the diagonals are crossed. Similar

examples of crossed diagonal strapwork, with or with-

out bobbles, and within rows of squares, can be seen in

the West African copies in Oxford, Dublin and

Chicagoa0. A detailed comparison of all these designs

with those on West African textiles might well provide

a means of locating these Qur'án copies more exactly.

The disregard for exactness in line and angle in most

of these decorations seems to be an indigenous feature

of West African culture, apparent also in the avoid-

ance of rectangles in buildings. It often seems to be

deliberate non-symmetry or anti-symmetry. The major

dislocation in the strapwork on the right of the open-

ing decoration of the Leeds copy41, where special

attention to symmetry would be expected elsewhere in

the Islamic world, is a good example. The full-page

design preceding surat Maryam in the Bodleian ms'

Arab. d. l4la2, where the otherwise regularly executedwoven design is squashed in the middle, is another, and

the dislocation of the pattern at the bottom of thedesign illustrated by Safadi is a third. If this feature is

deliberate, then it could be put down to fetish and the

avoidance of the evil eyea3.

1.3 Tur LEATHER

The West African copy's leather is made up of three

components. The first is a rigid outer satchel, well over

a century and probably 130 to 140, and perhaps even

150 years old. Such satchels for Qur'án copies

and religious works were common in Muslim West

Africaaa. They were primarily to safeguard the pre-

cious contents from coming into contact with anyone

or anything uncleanas. This one has a plaited and

thonged strap for attaching round the waist or

shoulder, and an elaborate triple system of flaps to go

over the inserted copy. After the three flaps of a first,

soft, layer are folded over, a middle, rigid. flap is held

down by a leather strap that comes over it and ties on

to another strap attached to the front of the satchel.

Finally an outermost flap with a four-tailed thong and

leather stitching around and across folds over unsecu-red. The outer satchel, measuring 235 x 180mm.. is

well-made, with the suede nap, that is the flesh, outer-most. It is of hair-sheep or goat. Words have beenbranded in a crude West African hand on to the front.

Of those that are legible, 'Fás' suggests that an ownermay at some time have been on pilgrimage to Fez.

The second component is a cover or wallet to secure

the loose bifoliosaó. It is original - the paper fits

perfectl,v. It also fits the satchel snugly. so that also

looks or iginal . The grain of the wal let shows i t to be

aimost certainl l sheepskina-. Just as with the outer

satchel this ' ,val let ts also probabl,v 130 to 140, and

perhaps even I 50 years old. The leather has been

folded over homemade cardboard. or pasteboard,

made up of laminated paper. Endpapers have then

been pasted over for final neatness, and strips of

leather have been stuck down the inside of the folds. In

pencil near the inside of the envelope-flap is written:'Cat 79162 r l r ly 10 : . This is probably the catalogue

number and price for an eariier sale in London.

Straight lines have been tooled along the outer edges of

the wallet in groups of three. four and five, with a

wider band of seven along the middle. The whole has

been tanned with vegetable material and stained reddish

brown. while these bands of lines along the edges and

the middle have a l5-20 mm. dark brown stain super-

imposed. lnside the rectangles thus formed is more,

simple hand tooling of straight lines and dots. The

lines are mostly in threes, combining to make four

different motifs of diagonals and straight lines. The

dots have been made with a circular ended punch with

a cut-out cross shape, producing an imprint not unlike

a small Maltese cross. These are placed either atjunctions of the lines or in tight juxtaposition to form

thick lines themselves. Attached to the envelope-flap,or tongue, is a thin suede thong a metre long ending in

the base of a small white Cowrie shell used as afastener once the thong has been wound round the

cover. In this case the envelope-flap was clearly not

designed, as usual, to tuck inside the covera8.

^ 1

48 MANUSCRIPTS OF THE MIDDLE EAST 2 ( I9IJ7)

The third component is an upper and lower end-pieceae. They appear to be akin to the daffatan of

very early Qur'án copies referred to in the celebrated

Collection hadrths. A session studying the manuscriptsoon reveals their function. Being loose-leaved thecopy needs to be tapped back into shape after use, aswith a pack of cards. The leather endpieces enable thereader to grasp the copy firmly to do this withoutendangering any of the pages. These endpieces arethick pieces of cow or camel skin, crudely worked.Similar ones can be found with other unstitched bind-ings, for instance of al-JaztlT's Dala'il, SOAS ms.I 9620 s 0 .

1.4 Tsr PAPER

The texture of the paper of the West African copycan be compared to a heavy cartridge paper of today,al though some pages can be thinner. I t consists of 163bifolios folded in half to make 326 folios or leaves.which, with 6 other single folios here and there, makeup the to ta l o f 332 s ' .

In nearly every case what is now the top edge ofeach bifolio was formed by a straight cut through the

middle of an original sheet, a little more than doublethe size of the present bifolio. So, allowing at least l0mm.52 for the trimming of what is now the bottomedge of each bifolio, the longer side of the originalsheet would probably have measured about 480 mm.The side edges of the bifolios hare also been hand-tr immed. Al lowing about 10 mm. here as wel l points toan or iginal shorter s ide of 350+ mm. These dimen-sions, 480 x 350 + mm.. were the average size made byVenetian papermakers from l l l4s3. The bottom andouter edges of the bifolios are not trimmed as straightas the top - the bottom edges can curve into or awayfrom the fold and the outer edges can also curve invarious ways. The continuity of these curves frombifolio to bifolio up to an average of 20 in sequencesuggests that the trimming was done to gatherings ofalready individually folded bifolios.

The final trimming is at times careless, and so maywell not have been done by the papermakers them-selves, but perhaps by a middleman, like a warehouse-keeper for an exportersa, or maybe at its West Africandestination. Further. although the whole manuscriptseparates out into three self-contained lots of paper,and although the overall proportion of watermark tocountermark within each lot is as near to l:l as makesno difference, the two bifolios from each original sheetoften do not follow each other directlyss. Thisshuffiing also suggests a middleman.

Judging by the size of the lower margin in relation tothe upper one, the main trimming was probably donebefore any writing. From the catchwords to the bot-tom edge measures 20-30 mm. and from the top edgeto the top of the first written line measures 15-20 mm.The size of the written area balances with the size of

the page. Had the pages been longer more lines wouldprobably have been written. There could have been atrim after the text was written to marry the manuscriptto the cover, and this could account for its lack ofstraightness, but the similarity of the page-size to thatof the Mahdiyya Copytu suggests that if so it wouldhave been minimal. The paper would thus most proba-bly have reached the scribe in stacks of folded bifolios,more or less in the form and in many cases in the orderthat they are in stil l.

That the bifolios are neither stitched together norattached to the cover enables the watermarks to beexamined easily. Indeed a number of them were photo-graphed using a beta-autoradiograph techniquesT. Thelack of stitching and binding also makes it possible(rare with European books) to examine the entiremanuscript, and so ascertain whole sequences of paper.

There are three different sets of watermarks dividingthe manuscript into three lots. All display Íhe tre lune,the three crescents decreasing in size. The first andthird lots have the simple te lune, and the second thetre lune with human profile (moonface). On the origi-nal sheets they would have been in the centre of the left

half with the largest crescent on the left. Chain-linesvary between 25 and 30 mm. apart.

The f i rst lot comprises 113 bi fol ios and 5 fol ios ( f f . l -230). I ts watermark is a l re lune and the let ters cl inout l ine capitals ' '8. The tre lune occurs 56 t imes, the cr-55 t imes and there are 7 apparent blanks. Beforecutt ing and folding the iot would have comprised 59orrginal sheets. or their equivalent. I t has been men-tioned that shuÍfling occurred. but if 7 or 8 originalsheets made up a gatheringse, then this first lot couldhave consisted largely of eight gatherings, allowing fora damaged sheet, or perhaps two - none of the 5single folios match.

The second lot comprises 28 bi fol ios ( f f .231-86). I tswatermark is 3 moonfaces. and the countermark isAndrea Galvani Pordenone60. The Íre lune here occurs14 times, as does the countermark. The 14 originalsheets could have made up two gatherings.

The third lot comprises 22 bifolios and I folio(ff.287-331). Its watermark is a simple tre lune andthere is no countermarkór. The tre lune occurs l1times, the blank 12 times. These 12 original sheetscould have been a gathering and the bulk of another,the remaining few cut and folded parts of which werenot needed once the manuscripts was completed.

1.5 Tur wATERMARKS AND CoUNTERMARKS

The tre lune of the first lot of paper, from the centreback of the largest crescent to the mid-point betweenthe cusps of the smallest, measures between 7l and 78mm. The cusps of the largest crescent range from 30 to38 mm. apart, and those of the smallest from 20 to 23mm. These variations can partly be put down to twinmoulds in the paper's manufactureó2, but mainly to

ASPECTS OF THE PHYSICAL TRANSMISSION OF THE QUR'ÁN

the mobility of the marks in the mould. On the onehand, attaching semicircular shapes like crescents (asopposed to angular ones like letters) to the mouldseems to have left the shapes more open to beingjostled and pressed by the stuff. And on the other,during the manufacture of the paper many crescentshave slipped along the underlying transverse wires wellaway from the chain-lines. Moreover, a number of thelargest have swivelled through 60 or so degreesó3.

The shape of the letters of the countermark cl onthe other hand is far more stableóa. The I would havebeen sewn on at both its top and bottom left tips, andthe c, although also probably only attached at itscentre back. is smaller than the smallest crescent of therre lune and more compact. The letters consistentlymeasure 17 mm. in height, but for the other dimen-sions there are two distinct sets of measurements.clearly il lustrating that the paper was made from twinmoulds. From centre back to the point between itscusps the c spans 10 or l l .5 mm., and from i ts back toits belly measures 2 or 3.5 mm. Similarly, the base ofthe r spans 13 or 14 mm., and its upright measures 2.5or 3 mm. across. This countermark has not beenidentifled ó s.

Without going into further detail, the 28 bifolios ofthe second lot of paper also fali into two equal andseparate groups. and the measurements of the crescentsof the third lot of paper are also var iable in a simi larway. These are both indicat ions of twin moulds in themanufacture of the paper of these other two lots also.

1.6 DlrrNc

There are three issues here. the date of the manufac-ture of the paper. the date of its arrival in West Africa,and the date of the writing of the text. None can bedetermined precisely, but the general evidenceconverges with the estimation of the age of the leatherbinding, point ing to a probable mid- l9th centurydating for the writing of the text.

While the watermarks of the first and third lots ofthe West African copy's paper provide no precise helpin dating its manufacture, the watermark, and espe-cially the countermark, of the second lot place usalmost certainly between 1836-80. Further evidencesuggests that it is closer to the earlier rather than to thelater date. As for the watermark, three moonfaces, thisdevice apparently superseded the simple tre lune inEgypt by the early 1840'sóó, and could therefore alsohave done so around this time in western Sudan. Asfor the countermark, Andrea Galvani Pordenone, ittakes us into the world of a famous l9th-centuryfamily of Venetian papermakers.

The original firm was known at the beginning of thel8th century as Fratelli Galvani, or more usuallyValentino Galvani. Valentino was the founder of thefirm. and died 1800-1067 leaving four paper-mills inthe north-eastern province of Friuli, one at Pordenone.

His son Carlo was the next director, until at least 1835.After the death in 1824 of Antonio, Carlo's brotherand partner, the firm was known as Carlo Galvani.Andrea, a grandson of Valentino, was already active inthe firm by 1835 and became its director in 1836. Hehad been born in 1797 and died in 1855ó8. We havetherefore in 1836 an earliest date for the paper'smanufacture, a terminus a quo, but what about a latestdate, a terminus ad quem?

The direction of the Galvani firm passed from An-drea in 1855 to the brothers Giorgo and Giuseppeóe,but it is not known for how long after this the firmcontinued to use Andrea's countermarkT0. Texts ofmanuscripts now in Jos (some 200 miles south ofKano) with Andrea's countermark date from as late asthe 1890s. 1900s . l9 l3 and even 192871. However ,whiie in 19th-century Europe it was usual for paper tobe used fair ly soon after i ts manufactureT2, inWest Alrica it could be kept unused for considerablylonger?3. That many of these Jos manuscripts containamalgams of different papers (up to eight)74, makes itmore than likely that they underwent a long period ofstorage before use. Conversely, the lack of amalgama-tion of papers in the Leeds West African copy, andwhat is more, the preservation of long runs of sheets intheir original order, point to minimal storage. Despitetherefore not being able to ascertain a latest date forthe Galvani f i rm's use of Andrea's countermark. a datefor the export of the paper. and indeed for the writingof the West Afr ican cop) later than 1880 can almostcertainly be ruled out. Moreover. given the evidencethat the leather could wel l be 140 or so years old, adate nearer the middle of the century is more feasible.

The severe effect on trans-Saharan trade routescaused by the encroachment of European powers onWest Africa from at least the 1880s lends furthersupport to a pre-1880 date, at least for the transport ofthe paper to the Bornu areal s. Indeed, further con-sideration of the route the paper would have takenfrom Venice readily accommodates a mid-centurydating. Paper was certainly reexported from lgth-cen-tury Egypt to both eastern and western Sudan moreespecially to the former after Muhammad 'Ah's

conquest of that area in 1821 - but the full AndreaGalvani Pordenone countermark is not apparentlyattested in Egypt before 186876. So were there notmore direct routes to western Sudan than throughEgypt? Moving westward we hit upon more likely(trans-Saharan) trails, if not in the Garamantian routefrom Tripoli through Fezzan to Hausaland and Bor-nu?7, then in the route from Tripoli through Ghada-mes and Air to Kano78. Tripoli is almost certainlywhere the consignment would have arrived for theVenetians had dominated paper imports there from thelTth century through the 19th7e. Galvani paper inparticular was dominant in this market. It was one ofthe commonest papers in 1851 in Kuka (the capital ofBornu, just west of Lake Chad) and in Kano, and had

50 MANUSCRIPTS OF THE

been in use further down in Nigeria from 1824 atleast8o. So one or other of these routes would havebeen the most likely way the paper for the WestAfrican copy came down, and whichever of the routesit was. it could well have been in the Bornu area withina year or two of its manufacture. The earliest possible

date of its text could therefore be about 1837. Thesuggestion that paper bearing the Andrea Galvani Por-denone countermark was probably not used in westernSudan before 187081, seems to be based on an assump-tion that it would always have come through Egypt.That this particular lot of paper could have comethrough Egypt cannot of course be totally ruled out,and in that case 1870 would be about the text's earliestdate, but all evidence is against this.

In conclusion, the countermark provides 1836 as theearliest date of manufacture. The lively trans-Saharantrade in Galvani paper at this time means that it couldhave been transported to the Bornu area very soonafter manufacture. The well preserved originalsequence of bifolios points to a short period of storage.The evidence of the leather corroborates all this b"vstrongly indicating a mid-centur-v date for the r"'ritingof the text.

2 THe MAHDTyyA copy ( leeos UNIVERSITy MS. 619)8 ,

Description of the manuscript:

fol. 346 Uï. 247, 341, 342 cancels); 234-238 x 160-l64mm. ; wr i t ten area 170-175 x 100-102mm. ; 13l ines per page; laid paper;east Sudani naskh hand inblack ink, finely vocalised also in black. with recita-tive notation. verse-dividers and .ir7rc-titles in red;frequent marginal notes again in red: no decora-t ions: strong. leather loose-cover binding art ist ical lytooled, ending in an envelope-f lap: dated 1299 (:

1 8 8 1 A . D . )

2 . 1

There is no frame around the text. The surroundingmargins contain frequent notes in red ink. These areusually in the side margins, sometimes in the top one,and a few times ali around83. They are placed atdifferent angles to the text to prevent any possibility ofbeing confused with it8a. The colophon, after praise to

God and blessings on Mubammad, tapers to a point,beside the last word of which (al-ianna), is the date1299. The last three lines say, 'God, great of genero-

sity, give to its copyist and its owner paradise, para-

dise.' Below that in red are five lines containing ahadíth about the Prophet. Writing has been washed off

to the right hand side of the colophon, but attempts toread it under ultra-violet and infra-red light revealednothing. On the opening page there is a six linedevotional poem by a certain Shaykh Mulrammad al-Amrn. Each hemistich rhymes, but there is a newrhyme with each line.

MIDDLE EAST 2 (1987)

The script is a somewhat rounded naskh but, as with

the Maghribi tradition, regard is paid to the whole

word rather than to individual letters, whose shape can

vary considerably. It is completely unlike the clearly

Maghribi hand of a l7th-century manuscript from

Darfur 8 s, yet on the other hand it has very little

similarity with the rounded naskh of an 18th or l9th

century East African copysÓ, which appears to be

more akin to the Indian tradition. Three features of the

Mahdiyya copy's script may be isolated that can also

be seen to some extent in the autograph of the

Mahdi87. Final ra 'can sometimes be long88. Final

mïnt fteqtently impinges on the line below. It usually

drops down inclined backwards slightly, but can also

sweep forwardse. Final la 'of ten sweeps backwardse0.

2.2

There is no decoration to speak of. Verse-dividersare indicated by a red comma-like symbol. Every fifthverse is marked by' a small red hcl'. and every tenthverse b.v t$ 'o concentr ic red circ lesel. The f i rst word orphrase of each eighth of a hi-b is in larger scr iptshaded in red. and thtnm. rub'or nisf hizh is wri t ten inred in the marginq2. Certain other important words inthe text are similarly shaded, e.g. much of the /'atiha.

2 . 3

The well preserved tooled leather binding is a centur-y old and without doubt original. It is of hand-tannedhair-sheep- or goat-skin. In the middle of both coversts a turuni. or st.vlised ovoid citrus design 48 mm. longterminat ing in tuo bel l - l ike shapes. This tr ipart i te pat-

tern is contained r" tthin an outlined rectangle(1275 x 420mm.) jo ined a t the cent re o f i t s four s idesb,v l ines to the inner frame (1875 x 1200mm.). Thecentral design is reflected by Íour corner-pieces 50mm.high and 30 at base with the third side curving. Theoverall pattern is perhaps the dominant decoration on

Qur'án bindingse3. The exact correlat ion of measure-ments of the various components of the design on thetwo covers but the absence of correlation of measure-ments between them suggests the lack of an overallblock-stamping techniquen4. The stiffness of the coversis produced by a homemade cardboard attached to theleather, that is, layers of paper stuck together with agum, such as cereal starch.

2.4

The paper is in single folios - the bifolios have beencut in half, and very neatly. It is of more consistentthickness throughout the manuscript than the paper ofthe (in all l ikelihood) earlier West African copy. Whileshowing signs of having had active use, this would onlyhave spanned eighteen years, so the paper is in overallvery good condition. As with the West African copythe leaves are neither stitched together nor bound to

lASPECTS OF THE PHYSÍCAL TRANSMISS ION OF THE QUR 'ÁN 51

the cover, placing it in the Maghribi binding traditionrather than that of Egypt/Syria or South Arabiaes.

As with the West African copy a number of pageswere photographed using a beta-autoradiograph tech-nique. Apart lrom the three cancel-pages there is onlyone watermark and its countermark occurringthroughout the manuscript. The watermark is the two-headed eagle of the Austro-Hungarian Empire withswordeó. Being in the centre of the original bifolio itstop or bottom half occurs at the edge of the respectivefolio. The countermark. Andrea Galvani Pordenonewith a shield containing a moonface, is well attested inEgypt in the 1880s, and so the paper would most likelyhave come that wayq7. Cutting the bifolios in half hasresulted in this case in the top or bottom of the shieldoccurring at the edge of the respective folio. The foliosand then the bifolios put together indicate an originalsheet size similar to that of the West African copy.

The mere fact that the whole manuscript (bar threefolios) is made up of one type of paper points toa probably undisturbed export from Pordenone toeastern Sudan. The cutting of the bifolios, and thus ofthe marks. means there are four different type of folio,one with the lower part of the eagle. one with the top,one with the top of the shield. and one with Galvani'sname and the bottom of the shield. The overall equaloccurrence of these four over the uhole manuscriptesstrenghtens the possibi l i ty that the exported package ofpaper may not even har"e been opened between Porde-none and eastern Sudan. The neat cutt ing of thebifolios further indicates this. as does the order ofoccurrence of the four types of folio. While there hasbeen a good deal of shuffiingee, more in some parts ofthe manuscript than others, there are distinct runswhere the four types of folio occur in groups of four,with one or more rogues intervening here and there.The best example is between the seventy-five folios103-177 where there are sixteen groups of four. i.e.original sheets, with eleven rogues. Of the sixteengroups nine fall in exactly the order naturallyoccurring from the sheet being folded into four andthen cut100, four of them one after another, while afurther three have the first two folio types exchan-ged t o t. The rogues taken in their order of occurrencefall into a further two groups of four plus threerandom102. In other words what shuffiing has occuredhas almost certainly been around the time of theoriginal sheet being cut and stacked, at or near thesource.

All this creates a picture, similar in some ways tothat for the West African copy. The original sheetswere cut into four in a mill in Pordenone, or awarehouse, probably in the 1870s. Many of the folioswere piled straightaway in the order occuring frombeing cut, but most got shuffied a little, for some

reason. The pile, or ream, was then tied up into

a package, exported to Egypt and thence down to

eastern Sudan, where it was sold to a buyer probably

without having been previously opened. After comple-ting each folio, the scribe probably simply took thenext one from the top of the package and continuedwriting. Having written the colophon there stil l wouldhave been some paper left in the package.

BIBLIOGRAPHY

Only works actually referred to in the article are l isted here.

Arberry. A.1 .. The Koran illuminated: a handlist of theKorans in the Chester Beattv Librarv. Dublin 1967.

Abbot t . N. . 'Maghr ib i Koran manuscr ipts of the seven-teenth to ei-ehteenth centuries'. American Journal oJ Semi-t ic languages and l i terct rures 55 (1938). pp.6 l -5 & Plate.

Benedetti. A.. Storía di Porclenone. Edizioni de ' i l Noncello'Pordenone 1964.

Bivar . A.D.H. . 'Arabic documents of nor thern Niger ia ' .Bulletin of the Sclnol oÍ Oríentol and A./|ican Studies 22(1959 ) . pp .324 -49 .

-. 'A dated Kuran from Bornu'. Nigeria nnga:ine

11960 ) . pp . 199 -205 .-, lThe w^athrqat ahl al-Sudan: a manifesto of the Fulani

j ihad'. Journal of African Hisrory If 2 (1961). pp. 235-43.-, 'The Arabic call igraphy of West Africa'. AJiican lan-

guage revíev '7 (1968), pp. 3-15 & Plates.Boahen. A.A.. Britain, the Sahara, and vrestern Sudan 1788-

1861. Oxford 1964.Briquet. C.M.. Ie.r f i l igrane.s. ed. A. Stevenson. 4 vols.

Amsterdam 1968.Brocket t . A.A. . 'St . Andrervs Univers i ty Or. ms. 16.

Fragment of a Qur 'an on paper in Maghr ib i scr ip t ,Probably Spanish. 15th century' ' . Codices Manuscriptí 10( 1 9 8 4 ) . p p . 4 1 - 5 1 & P l a t e s .

Churchil l, W.A., Waternturks ín paper ín the XVII & XWilcenturies. Amsterdam I 935.

Digby, S., 'A Qur'an from the East African coast'. Art and

archaeologl'research papers 1 (1915), pp. 49-55.Ebied, R.Y. , and Young. M.J.L. . 'Some Maghr ib i manu-

scripts in the Leeds University collection', Journal o.fSemític Studies 2l (1976). pp. 109-l 19.

Eid, O.M.. The kholv'a as an Islamit' etlucational institutíonin the Sudan. Unpublished Ph. D. thesis. Edinburgh 1985.

Eineder. G., The ancient paper-mills ofthe lbrmer AusÍro-Hungarían Entpire and theír r.'atermarks. MonumentaChartae Papyraceae, vol. 8. Hilversum (Holland) 1960.

Ell is, A.G.. and Fulton, A.5.. Supplementar!- catalogue ofprinted Arabic hooks in the British Museum. London1926.

Fage, J.D., An atlas of African histort '. London 1963.Fedrigoni, A., L'Industria venera della carta tlallo seconda

dominazione ausÍríaca all 'unita d'Italia. Torino 1966.Gacek, A., Catalogue of the Arabic manuscripÍs in the library

of the School o.f Oriental and A.frican Studíes, Universítt'o f London. London 1981.

-, Catalogue o.f Arabic manu.scrípts ín the líbrart' oí theInsÍiÍute of Ismaili Studies. vol.l. London 1984.

Gaskell, P., A nex' introduction to bíbliograplr_t'. Oxfordt972 .

52 MANUSCRIPTS OF THE

Haldane, D., Islan'tic hookbindings in the Victoria and Albert

Museum. London 1983.Heawood. E., Watermarks, mainlr ol the ITth Qnd ISth

centuríes. Monumenta Chartae Papyraceae, vol. 1. Hil-

versum (Holland) 1950.Holt, P.M., The Mahdist state in the Sudan 1881-98. Oxford

University Press (Kenya) 1977.-, and W.M. Daly, The history o/'the Sudan. London

1979.Houdas, O., 'Essai sur l 'écriture maghrébine' . Publicatíons

de I'Ecole des Langues Orientales Viruntes, pp. 85-115.

Paris 1 886.Ibn al-Jazari. M.. al-Nashr Jil-qira'at al- 'ashr. Editor A.M.

al -Dabbá' . 2 vols . Cairo (n.d. )

James, D.. Qur'ans and hindíngs from the Chester Beattl '

Lihrart '. World of Islam Festival Trust 1980.-. Das arqhische Buch. The Arab book'. and exhibit ion of

Arabic manuscripts from the Chester Beatty Library in

Hamburg. 1983.Jeffery. A.. 'Progress in the study of the Qur'an text' '

Moslem Wor ld 25 (1935), pP. 4-16.Labarre. E.J. (ed.), The Nostit; papers. Monumenta Chartae

Papyraceae. vol . 5 . Hi lversum (Hol land) 1956.Last, M.. The Sokoto t 'ctl iphute. London 1967.

Lings. M.. The Qtlaníc art o/'ctt l l igroplt.t untl i l lnrtinttt ion.

Wor ld of Is lam Fest iva l Trust 1976.Luchetta. M.. 160 anni di .storitt t lell t 'eruntit 'rt Gulvuni di

Pordenone. Pordenone 1971.Marchetti. G.. I l Friuli uotrtine e tentpi. Udine 1959.Masafti/ ' SanZ'. Catalogue of an exhibit ion at Dár al-Àthár

al - Is lámiyyah. Kuwai t Nat ional Museum. 1985.Mi tchel l . T.F. . Wrí t ing Arabi t ' . London 1953.Nikolaev, Y.. Watermarks on medieval docuntents Jrom Bul-

garian urchit,es. Vodnite znatsina otomanskata imperiya.vol. l. Bulgarian Academy of Sciences. Sophia 1954.

Nachtigall, G., Sahara and Suclan. vol. 4 (trans. A.G.B. andH.J . F i she r ) , London 1971 .

Safadi, Y.H., Islamic callígrapht'. London 1968.Sa'id, Labrb al-., Al-. iam' al-sawrr al-at'val lí l-Qur't1n, ur ctl-

Mushaf al-murattal. Cairo 2nd. ed. 1978. (translated and

adapted by B. Weiss, M.A. Rauf and M. Berger às Ttv

ret' i ted Koron. Princeton 1975).Simmons. J .S.G. , 'The Leningrad method of watermark

reproduct ion ' . The hook col lector l0 (1961 ) . pp. 329-30 &Plates.

Stevenson, A.H., 'Watermarks are twins'. Studies in Bíblio'g raphy 4 (1951 -2 ) , pp . 57 -91 .

-. The prohlem of the Missctle Speciale. London 1967.Vajda, G.. Álbum de paléographie arabe. Paris 1958.Walz, T.. Trade betx'een Eg1'pt and Bilad as-Sudan ( 1700-

1820 ) . Ca i ro 1978 .-. 'The paper trade of Egypt and the Sudan in the

eighteenth and nineteenth genturies'. M.W. Daly. Moder-nization in the Sudan. pp.29-48. New York 1985.

Witkam, J.J., Arabic' manusc'ripÍs ín the l ibrart ' of the Uníver-sitt' of Leíden and other collecÍions in the l{etherlands. Ageneral intoduction ro the t 'atalogae. Leiden 1982.

Norss

I As part of a l ist of manuscripts in Ebied and Young,

pp. 109. 10. 'Sudan' is used here in the t i t le of th is ar t ic le in

MIDDLE EAST 2 ( I 987 )

i ts comprehensive lgth-century meaning of the belt of sub-

Saharan Africa between the desert and the tropical forest,

stretching from Futa (present day Senegal) in the west to the

foothil ls of north-west Abyssinia, see Fage, maps 13. 38. In

what follows 'eastern Sudan' refers roughly to the area of

modern-day Sudan (also called 'the Sudan'), 'western

Sudan' to the area west of that and along the north of

modern-day Nigeria.2 As witnessed by the red dot over the aliJ'to signify

li.vahaba (not, as might be expected, l i 'ahaba'). see i l lustration

No. 4. second full l ine of text; and by the red dot under the

ligature between nun and alif ntaqsura in anna to signify

intala (i.e. annay) in the next l ine of the same il lustration.

The Sudanese call this script simply 'Kufi ' . and the Hausapeople 'Ajami ' (Bivar . ALR.p. l0) . Because there have been

several varieties of Maghribi used in western Sudan' the

script of the West African copy' is termed 'Ifriqi'. following

Bivar and probably Ibn Khaldun. see Bivar NM, p. 204'

A L R , p . 9 .3 As witnessed b1' the readings v'a'ar.iulikum (5:6), see

il lustration No. 8. l ine 7. and narsa'ht1 (2:106). see i l lustra-

t ion No.7. l tne 1. No copl in th is t ransmiss ion has been

descr ibed in Western $orks. see Jef fery, p.6. and compareLab rb a l -Sa ' i d . p .9 l n . I (Eng l i sh t r ans la t i on , p . 143 n . l 6 ) .

While there is no shortage of manuscripts from earlier

centuries. especially' lrom Egypt. in the transmission of Ab['Amr. it was steadily displaced from early last century by theprinting-press, and has never itself been printed. A copy in

the Hafs transmission is said to have recently been doctoredin the Sudan with Ab['Amr readings. but without approvalfrom Omdurman (oral communication from M. Turabi).

The small number ol Qur'án copies existing from eastern

Sudan, and similarly of manuscripts on Qur'án sciences. can

mainly be explained b;' ' trvo factors. one the prevail ing

Sudanese custom of education by' oral transmission (cf. Eid,p. 453) . and two. the h igh pr ice of a wr i t ten mu$hqf(upn'ards of t 300 today) . I t is p lanned to make a study ofthe textual transmission of the Mahdiyya copy elsewhere.

a For these events. see Holt, p. 243, and Holt and Daly.pp. 112. I19, and for the macabre photograph (opposi te one

of Wingate) Holt and Daly, Plates 24 and 25.t By a Mrs. G.R. Lancaster on the death of her husband.

a Barnsley (West Riding) council lor and long-serving mem-ber of Leeds University's governing council. How he acqui-red the copy is not known.

ó Greater space is given to the West African copy. This is

because it has more decoration and more leather. Also its

date and place are not stated, so a detailed discussion of itspaper and watermarks is necessary. For much useful assis-tance with this article I would especially l ike to thank Dr.Adam Gacek. David James. Dr. Y.H. Saladi and Dr.A.D.H. Bivar. The cost of publication of the colour platesfor this article has been generously met by the University ofLeeds. and I am indebted to Mr. R.P. Carr, Librarian of theBrotherton Library, for his support. I am also very grateful

both to Mr. P.S. Morrish, Sub-Librarian in charge ofManuscripts and Special Collections, and to the Photogra-phic Section of the University of Leeds.

r Many thanks to Colin Wakefield, Senior AssistantLibrarian in the Bodleian for his kind assistance with themanuscripts there, and also to the Bodleian Library forpermission to publish Plates of Bodleian ms. Arab. d. l4l.

I See il lustration No. 1. centre and left.

ASPECTS OF THE PHYSICAL TRANSMISSION OF THE OL R'ÀN 5 3

q The variation in the size of the written area arises in themain from the varying width of the pen-strokes according tothe wear on the nib of the reed-pen. see Mitchell, p. 18.

r0 In eastern Sudan made from soot, gum Arabic andwater (Eid, p. 480).

11 Wr i t t en i n 1020 (L ings , P l . 10 ; Sa fad i . pp .23 ,78 ) . Iowe this observation to Dr. Safadi. and it f i ts the views ofHoudas (pp. 102, 111) and Bivar ( l /M, p.204, ALR. p. 9) .

12 Safadi. p. 24; 'grossier', Houdas, p. 109. The culture'slack of need for symmetry should also not be ignored.

'3 Cf . Houdas, p. 101 (quot ing Ibn Khald[n) and p. 107for the possibil i ty of f inding the same letter on the samepage in three or four different forms.

ta . f . 122a . f o r i ns tance .

's Unl ike Houdas'example (h is i l lust rat ion No. 11) . Theslant of Shetimo Kawo's Maiduguri copy (Bivar. ALR, PI.I)and that of Safadi's i l lustration (see note 17 below) appearsnot to be the norm.

tu e.g. f. 162b ult.. hattaaa ' idho and ilaahLrkwnuuu ilaa-

hun.t ' e .g. / ' .167a.4. rantbughr. Note a lso the same ni th the

tanwtn before each basntula. whether in. or on (see il lustra-t ion No. 2) . In fact . vocal isat ion shous the cop) ' to be in thetransmission of Warsh by wal of al-Azraq. This is seen bylack of ass imi lat ion of the br7 ' to the nunt in í rkah mu'no(l l:42). and the presence of .fatha over .r 'a' in ikhv'atiya(12:100) and v'ali.t 'a (20: l8) (see |bn al-Jazari. vol. 2,pp. 168-9. 173.16) . The same appl ies to the pr inted Nigenancopies in the Brit ish Library, one printed in 1905 (Ell is andFul ton. no. 14507. cc. 11. a lso in the Bodle ian, shel fmarkArab. d.408) . and the other a handsome 1960s facs imi le ofa late 19th-century manuscript (Brit ish Library Select BookOr.74. d. 23) . i l lust rated in Safadi . p . 24.

tB BSOAS, and especia l ly ALR.te ALR. p. 10. Kano \ \ 'as an inrpor tunt centre of Qur ' t1n

scholarship. and a very possib le provenance of the WestAfrican copy. written there perhaps by a Hausa scribe.

20 B i va r . BSOAS. P l . I I I . ALR. p .6 .21 Unlike the pointed ones in Abbott (Pl. I). Very occa-

sionally the sud is pointed in the Leeds West African copy asalso in Abbot t .

22 Bivar , BSOAS, Pls. I , I I .23 This trefoil motif, preserved far longer in the Islamic

West than East (see, for example. Brockett. pp. 44. 5). goesright back to Kufi Qur'án copies (e.g. Lings. Pl. l) andseems to have emerged from the triangular group of dots tobe seen in many ma'i l copies. The yellow pigment has ametall ic element that does not allow beta rays to passthrough (see note 57 below, and the white blobs on thei l lust rat ions Nos 18, 19,20,21 - the larger ones are verse-dividers and the smaller ones hantzas). It is most l ikelyorpiment. or King's yellow, which is an arsenic sulphide thatwas used in north Maghribi Qur'án manuscripts for thesame purpose, verse-dividers. The red wil l most l ikely bevermill ion. Thanks to Don Baker for information aboutpigments.

2a For all these, see i l lustration No. 5. This scheme. withminor variations. is found in all i l lustrations I have seen.

2s See il lustration No. 2.2ó Similarly with the five roundels on the opening folios

of Oxford ms. Arab. d. 141, and of Chester Beatty Libraryms. 1599 il lustrated in James, Qur'áns and Bíndíngs, no. 95(misplaced on p. 116 for l l7) , and wi th the four on Chester

Beatty Library ms. 1601 il lustrated in Arberry, Pl. 69. In the

17th-century Maiduguri copy a ',shu.ia.vra' function is found,

in that a roundel is connected to the frame of a design. In

this case it is semicircular and reminiscent of a mihrab on amosque floor plan (Bivar, ALR, Pl. l). Again reflecting acloser adherence to the mainstream tradition of Qur'ánil lumination. the other roundel in this Maiduguri copyil lustration is akin Í,o the'ashara symbol. Another differentfeature of this copy is the strapping together of the fiveroundels in the margin of sura 1 (photograph in Dr. Bivar'spossession) .

21 tha' for thumn. ha' for rub' and nun for nisf, as inAbbott, p. 64.

28 See i l lust rat ion No. 4. For the device around 1800. seeBivar. ÁLR, Pl. II. and compare the more elaborate one inOxford ms. Arab. d. l4 l ( i l lust rat ion No. l6) .

tn . .9 . í Í . 89b (d iagonals) . I I I b . 129b. 131a. 244b ( t r ian-g le ) . See i l l us t ra t i ons Nos . l l . 12 . 13 .

30 See i l lust rat ion No. l . as in Abbot t , p . 63 a/1. , but notin Chester Beat t l L ibrar l rnss. 1599, 1601.

3 1 í . l b . 8 1 b . 1 6 3 a . 1 4 6 a . s e e i l l u s t r a t i o n s N o s . 2 . 3 , 6

and 5 respectivell ' . In the Sudan these are sometimes called' d t tbh t t l t t l

32 Strapwork is a long established Maghribi feature, cf.James, Qur 'ans and Bindings, p. 109b. and p. l l l for asimilar crenellation pattern (interlacing rather than overlap-ping) from the 13th-century Maghrib.

33 ms . IN : 15 -27 .1 . Masah i / San 'A ' . p .49 . C i r c l es a re a l soin the spaces between the weave-pat tern on the openingfolio of a Bornu Qur'an copl' front Barna (photograph inDr. Bi r "ar 's possession) .

3+ See i l l us t ra t i on \ o . 1 ,1 .35 B i ra r . . l lR . P l . I . \ Í a i du -qu r i i s j L r s t SE o f Lake Chad .36 See i l l us t ra t i ons No . 15 . and No . l 7 f o r c l ose -up .3- Safadi . p . 2,1.- r8 See i l lust rat ion No. 3.3e See il lustration No. 6 and front cover of this journal.a0 See i l lust rat ion No. l5 ;Arberry. PI .69; Abbot t . P l . I .a l See i l lust rat ion No. 2.a2 See i l lust rat ion No. 15.a3 Thanks for these and other points about West African

culture to Dr. S. Brett-Smith.aa James, Qur'ons ancl Bindings, p. 138. See also James.

Das arabisc'he Buc'h, entry 23. for an elegant example. Forhelp with the leather and the judgement about its age. I amindebted to S. Wolstenholme. formerly of the Leather Dept..University of Leeds. The judgement was made without anyknowledge whatsoever of the manuscript. or any promptingfrom the author.

as See i l lust rat ion No. l . le f t . Such a satchel can becalled'kehek' in the Sudan. The fetish for hiding anythinggood from the evil eye may also be relevant here. The outersatchel of Oxford ms. Arab. d. 141 is soft. and similar tothat i l lustrated in James. Qur'ans and Bindings, pl. l15.

aó Loose folios or bifolios, rather than stitched gather-ings, were a tradition in N. and W. Africa, see Haldane,p.66 re Pl. 74. Although by the early l8th-century bookssuch as al-Jazuh's Dala'i l were often being bound in N.W.Africa (Haldane. Plr.74, but for an unbound one see James,Dqs arabísche Buch, Pl. 23). Qur'án copies maintained theolder tradition, especially in west and east Sudan.

ar For a sheepskin cover from N. Africa see Haldane, Pl.53, and pp. 14, 15.

MANUSCRIPTS OF THE MIDDLE EAST 2 (1987)

fo l lows:1 1 6 - 1 2 3124-1291 3 0 - 1I 32-5136-7

I 38-49l 50-9

a8 Haldane, p. 14.aq The top of one of them is just visible inside the opened

cover, see i l lustration No. 1, centre.so Gacek, SOIS Cata logue, no. 51.s r Following the custom with Arabic manuscripts, catch-

words appear at the bottom of verso folios (Witkam, p. l3).That early copies of the Qur'án would have been loose-leaved provides a possible reason for this practice, not self-evident in stitched manuscripts. Foliation has been added inpencil in European numerals top left of each folio. with theomission of the fourth. which has now been numbered 2.2.Although therefore there are 332 folios in all, they arenumbered 1-331. The s ix s ingle fo l ios are nos. 14,85, 104,1 1 0 . i l 3 , 1 5 3 .

s2 Gaskel l , p .84. For d iagrams of how a sheet was cutand folded into quarto. and the position of the watermarks,see Gaskel l . f igs. 48. 49, pp. 90-1.

s3 Walz, 'Paper Trade' . p . 45 n.22.sa Gaskel l . p . 142.ss The run . /J ' . l16-165. for instance. breaks down as

sheets. While closer, this again does not match the apparent

gathering of 7 of the West African copy. Gatherings of 10

were also sold (Walz, 'Paper trade', p. 39) Wear and tear on

the outer bifolios, depending on distance travelled and

handling, perhaps explains these variations.ó0 See il lustrations Nos. 21 and22. Oxford ms. Arab. d.

141 has the same, but less finely designed, suggesting that itspaper may predate that of the Leeds West African copy.

ór See i l lust rat ions Nos. l8 and 20.ó2 Compare Gaskell, pp. 58, 62; and especially Steven-

son. ' twins ' .

ó3 See il lustration No. 20. Compare Stevenson, Missale.pp. 248-52 (excursus 4 - The movement of the mark on the

mould). For jostl ing see the smallest moonface, i l lustrationN o . 2 1 .

óa See i l lust rat ion No. 19.ós Eineder has only one cL entry (no. I 105) as counter-

mark to a mitre and grapes of Austrian origin from 1738(paper-mill unknown). The letters are shaped similarly to

those of the West African copy's countermark, althoughlarger and with the r superimposed on the c. Briquet has

four entries (see index. p. 129. * no. 9339) but they are

from the l5th and 16th centuries and are not shaped

similarly. Heawood also has only one. dissimilar cL entry,

no.216. dated to Madrid 1749. Again, Labarre found only

one in the German Imperial archives of the lTth and lSth

centur ies. l lost i tz , no.357, dat ing to Breslau 1641.óó Walz. 'Paper t rade' . p . 35.ó - E inede r . p . 169 . and pp . 166 -175 fo r maps o f t he a rea

6s Luchet ta. p.12. r ihere there is a lso a p ic ture of h im. as

also Fedr igoni . pp. 182. 3 (u i th colour por t ra i t ) ; Marchet t i .

p . 758. I am indebted to Dr. D.E. Rhodes for these and

other Italian references. and to Brian Richardson for

translating them for me.óe Luchetta, p. 13. It is not clear who their father was.?0 Not mentioned in Benedetti. nor in a number of other

Italian works consulted.11 Walz. 'Paper t rade' . p .47.?2 In the case of printing paper rarely more than two

years, although writ ing paper is l ikely to have had a longer

inactive l ife (Gaskell. p. 319: Stevenson. Missale. pp' 90-99).Thanks to G.D. Hargreaves for help here.

73 Las t . p . 194 n . 62 .'a Walz. 'Paper t rade' . p . 47.?s Fage, maps 40 and 43.1ó Walz. 'Paper t rade' , pp. 36, 41.r? Nachtigall, p. 28n.; Last, p. 194. But for its decline

mid 19th century, Boahen, p. 115. Arab paper, i f not

Venetian, was sti l l going down the Garamantian route even

in the late lgth and early 20th centuries. This is seen from

Íhe Beniamino Arhihlya nqsíb watetmarked paper, probably

made west of Tripoli at that t ime, used extensively in

Nigeria but not attested in Egypt (Walz, 'Paper trade', p'

42).The Qur'án copy bought in Lagos in 1928 and dated by

Abbott in 1938 to the lTth or 18th centuries was written on

this paper, and should therefore be redated l9th or early

20th century (Abbot t , pp. 61, 2) .?8 Boahen, p. 108.1e Walz, 'Paper trade', p. 40f. This makes it unlikely that

the paper came down the most important trans-Saharanroute west of Egypt of the later l9th-century, the Sanlst-

two original sheets one after the other.3 countermarked bifolios.I watermarked bifolio.an original sheet.1 countermarked bifolio (not the other halfo fe i t he r 130 -1 o r 138 -9 ) .6 watermarked bifolios.5 countermarked b i fo l ios (150- l not the

I 58_9 )162-5 an original sheet.

s6 234+ x 160* mm. see below. The fo l io-s ize of thefour West African copies in the Chester Beatty Library arealso similar. varying only between 220-230 x 163-170 mm.(Arberry. pp. 76. 77), as is that of Oxford ms. Arab. d. l4l(235 x 176 mm.) The folios of the three manuscripts in theOriental Institute in Chicago are a l itt le larger - 240 x 182mm., presumably having been trimmed less (Abbott, pp. 61,65). On the other hand. a couple of (more northerly?) WestAfrican Qur'án fragments in a thinner. more elegant, Ifríqrhand (from a collection in the Gulf. and kindly shown meby Dr. Safadi) were on folios half the size and less.

s? See il lustrations Nos. l8-24. In practice. the manu-script is sandwiched between an activated C-14 polymethylmethacrylate sheet and Kodak CX Industrex Íi lm to produ-ce a radiograph of the manuscript. The radiograph is thenphotographed and a positive paper print made. This showsthe watermark as black against a white background. Fordetails of this technique, see Simmons; and Stevenson,Missale pp. 66ff., 298.

s8 See i l lust rat ions Nos. 18, 19,20.sq The word 'quire' has been avoided because it is not

certain how many sheets would have made up a ream inl9th-century Sudan. Reams of tre lune paper made by theGalvani f irm in 1820 numbered 500 sheets (Eineder, p. 170;see also Gaskell, p. 59), so in that case the 20 quires wouldhave had 25 sheets each. None of the lots of the WestAfrican copy's paper falls into such gatherings. There isevidence of reams in eastern Sudan, and the Hijaz, number-i ng on l y l l 5 and 120 shee ts (Wa lz , 'Pape r t r ade ' , p .46n. 43), in which case a gathering may only have been 6

other hal f o f 148-9) . (Fr iu l i ) and other in format ion regarding the f i rm. See a lso

160- l I watermarked b i fo l io (not the other hal f o f Fedr igoni .

ASPECTS OF THE PHYSICAL TRANSMISSION OF THE QUR'ÀN 55

controlled Benghazi-Kufra-Wadai route (Boahen, p. l l l).That the Leeds West African copy had connections with theSanlsiyya is also unlikely since the transmission they tendedto use was that of Qál[n from Náfi '. The routes are clearlyshown in Boahen, map 3 (p. 102) and Walz, Trade, map I(p s).

80 Walz. 'Paper trade', p. 36. The Galvani f irm had alsocornered one of the largest shares of the Levant papermarket by the mid- l9th century (Walz. 'Paper trade'. pp. 35,t6 \

8 1

8 2

8 4

ó 5

8 ó

8 8

8 g

same word in line 2. Cf. also lakum and am in line 3. Thetail of the final nttm of the basmala can even run risht downthree l ines of text, f. 346a.

eo See il lustration No. 7, l ines 3, 8.er See il lustration No. 7. l ine 2 and oenult.o : S e e i l l u s r r a t i o n N o . 8 .

Walz, 'Paper t rade' . p . 41.See il lustration No. 1, right.e.g. f. 284b. The notes mostly concern qira'aÍ.See il lustration No. 7.Vajda. Pl. 60.Digby.Ho l t , P l . 19 .See i l lust rat ion No. 7. e.g. l ines l . 2 . 3.See il lustration No. 7, as in ta'lam in l ine I, but see the

e3 See Haldane throughout. and glossary, p. 203. col. l.ea Compare Haldane, fig. 19.es Haldane, pp. 20, 200 n.22.eó See il lustration No. 24.q? Walz, 'Paper trade', p. 36 and n. 34. The pair span

1810-67 in Nikolaev. An exact correlation to the marks inthe Mahdiyya copy has not been found, but one dated 1851(Nikolaev. Pl. l l32) is very similar. but with a differentcountermark. For the double headed eagle see also Hea-wood, Pls . 168-93, and wi th a sword especia l ly nos. 1282and 4306.

nt Top of eagle 84 times, lower part of eagle 89. top ofshield 84. Andrea Galvani Pordenone * bottom of shield 86.

nn e.g. one occurrence of four of one type of foliotogether (ÍÍ.236-9). three ofthree togerher (23-5, 231-3.26j-9). and often of two of one type together.

r00 Lower part of eagle, Ándrea Gah.ani Pordenone +bot tom of sh ie ld. top of sh ie ld. top of eagle. These are. Í Í .121-4. 129-32. 135-8. I 39-42. 143-6. 147_50. 153_6. 166_9.t ] 4 -7 .

tot Andrea Galt 'oní Pordenone + bottom of shield. andlower part of eagle - .í l I I I -4. I 57-60. l6 t -4.

t o z f f . 1 1 9 , 1 2 0 . 1 2 5 . 1 2 6 ; 1 2 7 . 1 2 8 . 1 3 3 . 1 3 4 : 1 5 1 " 1 5 2 .

I 65 .

56 MANUSCRIPTS OF THE MIDDLE EAST 2 ( I987)

L Leeds Arabic ms. 301 ( le f t and centre) and Leeds ms.619 ( r ight)

ASPECTS OF THE PHYSICAL TRANSMISSION OF THE QLIR 'ÀN 51

@ ,@,

t

<íÍoqltiÉ{b-

,i;r:,;,

L Leeds Arabic rns l 0 l l . l b ( r r r l r r 1 : l l : l )

3. Leeds Arab ic ms. 301 f .8 lb (sura 6 :165)

58 MANUSCRIPTS OF THE MIDDLE EAST 2 (1987)

{ " ' q

4. Leeds Arabic ms. -101 l ' . l61a (deta i l : .sr7rr r l9 : l7-15)

5. Leeds Arabic ms. 301 f .246a (sura 37:168 - beginning of 38)

ASPECTS OF THE PHYSICAL TRANSMISSION OF THE QUR'AN

6. Leeds Arabic ms. 301, f .163a. (sura 18:110 - beginningof sura 19)

t:

59

í,:',,:t , . ,,,t,t:r 'Y::v::t:t::'.'t .a:.'4l:..Yll .

U:,,,

7. Leeds University ms. 6l9l l}a (sura 2:106-12)

60 MANUSCRIPTS OF THE MIDDLE EAST 2 (1987)

8. Leeds Univers i t l n- rs . 619 / . 6 la ( . i i l ra ,s :5-7)

...7:.:' .ti '&tu .' ,,.. ,,r:, .. ,,.,, . . .,... : .' :. -,:,,i*,..:..-t.rat;a.:::',',.,:...

9. Leeds University ms. 619 colophon (í.346b)

\

Wó lASPECTS OF THE PHYSICAL TRANSMISSION OF THE QUR'ÀN

30r, í. l07b (detail)

F,lt6t

' l :

?a,a:.,

1. ':.::::'.

*F iffiilt}{-iffiiffi

11. Leeds Arabic ms. 301, í 89b (detai l )

. r 4

Yr't a,/"; ..

iilt =3q:gwru* _

a g } "

sP@drww

- *@,

ffi

301. f . I I 1b ídeta i l ' )

(írl'F-áryp

12. Leeds Arabic ms. 301, f.244b (detail)

o l MANUSCRIPTS OF THE MIDDLE EAST 2 (1987)

14. Oxford, Bodle ian L ibrary, ms. Arab. d. l4 l . f .2r

15. Oxford, Bodleian Library. ms. Arab. d. Al, f. 202v-203r (sura l8:109 - beginning of sura 19)

63ASPECTS OF THE PHYSICAL TRANSMISSION OF THE QUR'ÀN

l4l,.f. 6v (detail)

à

ffiiÉtlt +{t#ltËl'!g:l$1$a,rrtsfáitffif.13gï-jj-:t@F$

tf:ffiffiÍ16. Oxford, Bodleian Library, ms. Arab. d.$wfi

èi$

i,., Nffr**,'i':Ëifff i17. Oxford, Bodleian Library, ms. Arab. d. 141,f .203r (detai l )

'r $:j

r i

.sÀrW

'*

;sï

9**

sï.$.ïs

r l

q &

s.* S.s*Kx+B w* * W {

M A N U S C R I P T S O F T H E M I D D L E E A S T 2 ( I 9 8 7 )

i * S

bsË{s

{. $ss.i

. $è

18. Leeds Arabic ms. 301, watermark tre lttne

19. Leeds Arabic ms. 301. countermark cL.

65A S P E C ' T S O F T H E P H Y S I C A L T R A N S M I S S I O N O F T H E Q U R ' À N

q.wÊ'q

rw4 F "a.át4* , W e ê & l

+t :*%rffir- *"'YWW'M,#"w,**

&r&#qÉ'&

#r"4F 1 ï &*4" i " 'G í"9. ::',.,#. "q

*"c{Yr . , ; *

&4

H

20. Leeds Arabic ms. 301. watermark su ' i r ,e l led t re lune.

21. Leeds Arabic ms. 301. watermark 3 moon faces.

MANUSCRIPTS OF THE MIDDLE EAST 2 (1987 )

22. Leeds Arabic ms. 301. countermark Andrea Goh'aní Pordenone

23. Leeds University ms. 619, countermark shield with moonface (top half only)

ASPECTS OF THE PHYSICAL TRANSMISSION OF THE QL R ÁN

24. Leeds University ms.6l9, watermark eagle (composition of two halves)