Introductory Remarks:

The Mosque of Ibn Tulun was built in 879, the third congregational mosque in Fustat. The first was built in 641, by ‘Amr ibn al-‘As, the Muslim conqueror of Egypt, and the second at al-‘Askar, the Abbasid capital founded in 750. All trace of al-Askar has now completely vanished, and the original form of the mosque of ‘Amr has been all but obliterated over centuries of extension and restoration.Therefore Ibn Tulun’s mosque is the oldest in Egypt to survive in anything like its original form. All of its many restorations have taken into consideration the original plan of the mosque, and have thus not altered its size or unique architectural elements. It is therefore possible to assess the mosque as it was originally intended to be received: the most common reaction has been to think the mosque odd, out of context, not fitting into the architectural landscape of Egypt.

Stories were even told of the mosque to explain its unusual features: for example, Ibn Duqmaq relates that the people wanted to know why the qibla did not follow the orientation of nearby mosques; Ibn Tulun explained that his engineers had been disagreeing over the orientation until the Prophet Muhammad appeared to Ibn Tulun in a dream, and ordered him to place the qibla in the position in which Ibn Tulun had built it. The Prophet also drew a design of the mihrab on the mosque floor for him to trace. Ibn Tulun went straight to the place where the Prophet had been and found the design still on the floor.

What such stories clearly show is the unfamiliarity of the people of Fustat with the enormous caliphal city of Samarra with its two great mosques, the biggest in the world. Many elements in Ibn Tulun, even the misorientation of the qibla, are imitated from the Samarra mosques. These huge structures were symbols of the power and grandeur of the ‘Abbasid caliphs, and as we have seen, according to Bloom’s thesis they were conspicuously imitated or omitted by their clients or enemies, a religious symbol adopted for political means.

So why these imitations at this mosque in Cairo? Is it a symbol of political affiliation to the Abbasids? I hope to show that in fact it means just the opposite, but to begin with I think a brief survey of the life and times of Ibn Tulun is important.

Ahmad ibn Tulun: life and works

Ahmad ibn Tulun’s life exactly spans the period of caliphal occupation at Samarra (835-883). He was born in Baghdad, the son of a Turk of Mongol origin who was a slave of the Caliph al-Ma’mun. Tulun converted to Islam in order to escape slavery, and rose to command the Caliph’s Guard at a time when the number of Turks with direct influence over the Caliph was becoming controversial, resulting in the full-scale transportation of the court and troops to Samarra in 836. This is where the young Ibn Tulun grew up, receiving an exceptional education for his status.

His stepfather, another Turk, was awarded the governorship of Egypt and sent Ahmad as deputy in his place. Thus Ahmad, now aged 33, arrived in Fustat in Ramadan 868, with an army numbering 100,000 men. The political situation in Egypt had been unstable for some time: it was over 1,000 km from Samarra which nevertheless drained Egypt’s resources by demanding annual tribute. This situation was exacerbated by public disturbances and the lack of continuity in government: since the conquest of Egypt in 641, only one man had been governor for longer than a year, and after the Tulunids Egypt saw 13 governors in 30 years.

In 870 the governorship was fully conferred on Ibn Tulun on the death of his stepfather. In 872 the Abbasids attempted to bring the province of Egypt back under direct control, but Ibn Tulun managed to hold onto his governorship by sending rich gifts to Samarra. However, his relationship with the ‘Abbasids began to deteriorate: he refused to send tribute in 875 and assumed control of Egypt’s finances; troops were sent against him to extract the money, but the attempt was abandoned. By 880, Ahmad was striking dinars in his own name, which is one of the prerogatives of the Caliph. He stimulated commerce by reinvesting Egypt’s revenues into agriculture and industries instead of sending them to Samarra; his long stay in government allowed administrative reforms, and a series of important public works.

Most significant of these was the establishment in 870 of a totally new capital city called al-Qata’i’, “The Allotments” or “Districts”, roughly 2 km by 2 km to the north of Fustat. Here his congregational mosque had a prime location in the centre of the city, on top of the Jabal Yashkur which was the most prominent hill amidst the plains and depressions of the surrounding area. From this mosque his name was proclaimed on Fridays, another prerogative of the Caliph.

Many aspects of his capital city evoked elements of Samarra, both in layout and in architectural form, so that the environment thus created would be familiar to him and his followers from Iraq. But I am not going to discuss them and you’ll have to take my word for it!

In 878, Ahmad openly rebelled against ‘Abbasid government by leaving Egypt to occupy Syria, but had to return to deal with a revolt by his eldest son, who was thereafter excluded from the succession: Ahmad’s second son was designated heir in 882, aged 18, and Ahmad died the next year, aged only 48, from dysentery contracted on campaign in Syria, apparently leaving 10 million dinars in the Treasury. The dynasty lasted until 905 when the ‘Abbasids sent troops to bring Egypt back under direct control: members of the Tulunid family were taken as prisoners to Baghdad, and the city of al-Qata’i’ was burned and looted. The mosque was left standing because of its holy status. Governors of Egypt once more took up residence in al-‘Askar, allowing the old Tulunid city to fall into ruin, until the arrival of the Fatimids in 969 when the mosque of Ibn Tulun lay on the main processional route linking al-Qahira to Fustat. The Tulunid Dar al-Imara remained in use until Ayyubid times, with the consequence that the mosque was kept in good repair, and was one of the four congregational mosques used by the Fatimids.

What this overview shows is that Ibn Tulun was a figure who constantly rebelled against the Abbasid regime, going to the extent of establishing his own capital city and dynastic rule in Egypt, and even adopting caliphal prerogatives such as minting coins and having his name included in the khutba at his mosque. Such a man would hardly have been likely to assert a political message of homage to the Abbasids, so the architecture of his mosque must be telling us something different.

The Mosque of Ibn Tulun:

Now we will turn to look at the mosque, and before discussing two particular features which are emblematic of the Abbasid mosques at Samarra – namely the minaret and the stucco decoration – I want to take you on a short tour around Ibn Tulun, picking out the small details which evoke features from the Samarra mosques. So on the left hand side we will have pictures of the Ibn Tulun features, contrasting with the comparative features from Samarra on the right hand side.

a. Construction:

The mosque is built entirely of baked bricks, which was not the usual building material for Egypt, but was at Samarra. The only feature which is made of limestone is the minaret, which is one reason why it was thought to be later than the mosque. The courtyard is roughly square with an area of 92 m², but its four sides are not quite equal [91.75 x 91.90 x 92.10 x 92.35] which must reflect the speed with which the mosque was built (3 years), or possibly the lack of familiarity of the local craftsmen with the materials used. It is surrounded by arcades with 13 arches opening onto it on each side.

The arches stand on square piers which were an innovation in Egypt, but an essential component of the Samarra mosques. What the population of Egypt was accustomed to were the columns of the mosque of ‘Amr, and stories were told to explain away this oddity: for example, Maqrizi (265) relates: “When [Ibn Tulun] undertook the construction of the mosque, they estimated that 300 columns would be required, and they told him that he could not find them except by sending to churches in the country and deserted lands, whence they could be transported. He disapproved of that and would not adopt this means.”

At Samarra the piers were square and plastered over to give a marbled effect which matched the marble columns attached at each corner: this effect is repeated at Ibn Tulun but the columns are brick, engaged at the four corners of the pier, and the decorative capitals and bases were modelled into the wet plaster. High up in the spandrels of the arches are small windows which enhance the flow of air through the mosque, but also help to illuminate the arcades, and lighten the weighty effect of the piers around the courtyard façade. These windows did not occur in the piers at the Great Mosque in Samarra, but at Abu Dulaf the piers carried narrow recessed arches, so perhaps Ibn Tulun converted these to windows as a lesson learnt.

You will have noticed that most of the arches are pointed! In fact, Ibn Tulun is the third earliest known example of the pointed arch used continuously throughout a building. The earliest is the ‘Abbasid cistern at Ramalla (d.789), and it will not surprise us to learn that the second earliest use of the pointed arch is at Samarra. Pointed arched windows were used above the doorways into the Great Mosque, and at Abu Dulaf, pointed arched recesses within rectangular niches were set in the spandrels above the piers. At Ibn Tulun, an emphasised pointed arch with a slight return at the spring, the precursor of the horse-shoe arch, becomes the fundamental architectonic element, and is used throughout the building in the arcades, the small arches in the spandrels, and the windows. The same feature is found in the remains of his aqueduct.

b. Dimensions and Ziyadas:

The mosque covers a rectangular area roughly 122 m [wide] by 146 m [deep]. It is surrounded on three sides by ziyadas, corridors 19 m wide which isolate the sanctified space from the public space, in this case the surrounding markets, streets and houses. The added space of the ziyadas bring the mosque to almost an exact square, giving it a total area of 26,318 m², or 6.5 acres, making it the third largest mosque in the world after the two Samarra mosques. [the Great Mosque has an area of 38,000 m² and Abu Dulaf of 28,750 m²] Both the Samarra mosques’ enclosing walls had corner towers, with semicircular bastions at regular intervals along them: do the semicircular flights of stairs in the ziyadas at Ibn Tulun recall these bastions.

At the top of these walls was a frieze of recessed squares with bevelled edges, within each of which was a shallow saucer 25 cm deep, 1 m in diameter, and coated with stucco: this motif, likened by Gertrude Bell to “rows of shields set within rectangular frames on Assyrian fortifications”, is repeated at Ibn Tulun as small squares pierced with circles, underneath the crenellations at the top of the mosque walls. Above this frieze is the distinctive cresting which have been likened to rows of “paper cut-outs of human figures linking arms”. There is no trace of anything similar surviving from Samarra, though it is possible that battlements of some kind surmounted the fortress-like walls. However, these are perhaps the most identifiable Tulunid feature to be found on later buildings.

Creswell suggests that the inner ziyada at the Great Mosque may have been decorated with blind arcades, and as restored the mosque walls have three small arched windows above each doorway. At Ibn Tulun the external decoration of the ziyada is entirely plain, except for the merlon crenellations that cap both the ziyada and mosque walls. The external decoration of the mosque wall, however, consists of the exteriors of the windows alternating with small blind niches with fluted hoods. These are set higher in the walls than the windows, and recall the general pattern of the interior, where the spandrels of the piers are pierced by small arches.

The only wall that does not have a ziyada attached is the south-eastern qibla wall: the Samarra mosques also have ziyadas on three sides and not on the qibla wall, but they are enclosed by a second ziyada on all four sides: it is this outermost wall that gives them their huge dimensions. Behind the qibla wall at Ibn Tulun was the Dar al-Imara, three rooms that connect to the mosque by doors on either side of the mihrab. This building was used until Fatimid times for all kinds of administration, and perhaps housed a library, but it also gave access to Ibn Tulun’s maqsura, the private area where he and his family and close associates would pray on a Friday. A similar structure has been discovered attached to the exterior of the qibla wall at the mosque of Abu Dulaf in Samarra: two wide doors flanked the mihrab and opened into a building behind.

c. The Fountain:

The domed fountain at the centre of the courtyard is not contemporary with the mosque and was built by the Mamluk Sultan Lajin. However, Ibn Duqmaq describes the original Tulunid structure as follows: “the fawwara which was in the middle of the sahn had windows on all sides, and over it was a gilt dome on ten marble columns, and round it were sixteen marble columns with a marble pavement. And under the dome was a great basin of marble, 4 cubits in diameter with a jet of water in the centre...and on the roof was a sun-dial. The roof had a railing round it of teakwood (saj).”

This description clearly recalls what we know of the fountain at the Great Mosque of Samarra. Al-Ya’qubi describes a fountain in the centre of the courtyard which “played without ceasing”. Al-Mustawfi says it was known as “Pharaoh’s Cup” (Kas-i-Fir’awn), and that its basin was formed from one block of stone 23 cubits in circumference, standing to a height of 7 cubits, and half a cubit in thickness. Excavation has revealed fragments of marble columns and capitals, painted stucco, gilding and glass mosaics, showing it was richly decorated.

d. The Roof:

As at the Samarra mosques, Ibn Tulun’s roof is flat and forms a terrace, resting directly on the tops of the walls. Immediately below the roof is the continuous wooden Kufic inscription of verses from the Qur’an, which Corbet and Creswell have both calculated to comprise only a seventeenth or fifteenth (respectively) of the whole Qur’an. Folktale, however, ascribes the whole Qur’an to this frieze which was also believed to have been carved onto planks from Noah’s Ark, and Corbet and Creswell never make clear whether their calculations are done on the basis of the total surviving length of inscription, or the total potential length for the inscription, which is just over 2 km.

e. The Mihrab:

The main mihrab is mostly a Mamluk reconstruction but the frame of the mihrab is original and it follows the design of the Samarra mihrabs: a blind stilted semicircular recess, set inside a second one, and flanked on each side by two marble columns with perforated capitals (the inner one a basket capital, and the outer one decorated with vine leaves and bunches of grapes detached from the background). The whole is set in a rectangular frame, with thick stucco mouldings in circular and almond shapes. Above this is a wooden band of Kufic inscription in the same style as the ceiling inscription, containing the shahada.

Again there is much similarity with Samarra: at the Great Mosque, the mihrab was like a great doorway in the centre of the south wall, reaching nearly to the ceiling; the niche was flanked by 2 pairs of columns of rose-coloured marble. Two concentric arches rested on these columns, and the whole was set in a rectangular frame which did not project beyond the wall face. There are traces of gold mosaic in the spandrels, and several fragments of an ornamental border in stucco. At Abu Dulaf, the mihrab niche was more elaborately carved, but was again framed by a rectangle and has bases for two flanking columns on either side.

f. Lighting effects:

Finally, a quick word about the lighting effects. The mosque’s external walls are pierced by 128 windows which are decorated with plaster grilles in geometric patterns, each one different. The windows are spaced so as to correspond with the axis of the arcades; thus, looking from the courtyard, every third window is centred in an arch. These windows provide special lighting effects: one scholar [Swelim] writes: “In the early morning, the sun shines through the grilles into the qibla riwaq; as the sun moves, it shines onto the piers of the arches and the floor of the riwaq, revealing the outlines of the windows and the different geometric designs, giving a remarkable effect. Later, the sun shines on the piers and the floor between the piers alternately, and the entire floor is decorated visually along the whole riwaq creating a magnificent carpet of light. It is at that time one can read the inscription, as the sun is reflected upwards onto the ceiling and illuminates this area.” This must have been an intentional effect by the architect, showing that every element in the mosque has been carefully designed and controlled.

Discussion of features copied from Samarra

I have drawn your attention to small details in the mosque which are closely copied from the Samarra mosques. I would now like to concentrate in a bit more detail on some of the more obvious features that are copied from Samarra. Perhaps the most emblematic features of Samarra are the spiral minarets and the style of stucco decoration, both of which have been taken to show the spread of Abbasid influence around the Islamic world.

Firstly, a brief look at al-Malwiyya, or the Spiral, the distinctive minaret type of both Samarra mosques. At the Great Mosque this stands at a distance of 27 m from the north wall of the mosque, but is connected to the mosque by a ramp which led to the top of the square base on which the minaret stood. The base was decorated on either side with rectangular niches. From the base rises a helicoidal tower, starting in the centre of the south side and winding round in an anti-clockwise direction until it has made 5 complete turns. The staircase ramp is 2.30 m wide and there are holes at the edge for a wooden balustrade. At the summit is a cylindrical storey decorated with 8 pointed-arched recesses; the southern niche is a doorway from which a steep internal staircase leads to the top platform, which is exactly 50 m above the top of the socle. On the top platform, holes in the edge suggest columns for a wooden pavilion.

At Abu Dulaf, the minaret is again a spiral standing on a square base, which stands 9.60 m from the north wall of the mosque. Its entrance is from a ramp in the south side which begins its spiral ascent through the solid brickwork of the plinth. The stairway is 1 m wide and gives 3 anti-clockwise turns.

The Minaret:

The minaret is the most distinctive feature of Ibn Tulun, as it is of the Samarra mosques. It was the first minaret to be built in Egypt, and directly evoked the ‘Abbasids and their mosques in Samarra. In fact the same folk-tale is related of both minarets, over a thousand kilometers apart: according to al-Yaqubi, Ibn Tulun was in a meeting with his advisors discussing how to build the minaret of his new mosque, when he started trifling with a piece of paper by rolling it up and pulling it out to form a spiral. It is said that either he wanted to cover his embarrassment for absent-mindedness, so said “Let’s build it like this”; or he was deliberately trifling in order to explain to his architect the concept of the Samarran helicoidal tower.

This same story was reported to Herzfeld in 1948 by the commissioner of excavations at Samarra: “One day, during the reign of one of the ‘Abbasid caliphs, the caliph met with his viziers and counsellors. While they were gathering, the caliph began to reflect to himself, having a piece of paper in his hand. So he played with it while he thought. When he awoke from his reverie he saw that he was seated amidst his viziers and had taken the sheet of paper and toyed with it. He realised that this was inappropriate behaviour for a caliph, so he made up an excuse and began to lecture his companions: ‘I have just been thinking that I should have a tall towering place made for me; at its top a special place for sitting so that I may sit and enjoy the fresh air and the view far and wide. And this building should resemble this paper. And this day I order it to be made.’ The people thought it a good idea and rushed to build it. And this is the reason for building the Malwiyya in Samarra.”

Firstly, we have to deal with the issue of its contemporaneity with the mosque. Since Creswell, Ibn Tulun’s minaret had been taken unquestioningly to be a later reconstruction by the Mamluk sultan Lajin who did extensive restoration work on this mosque. Creswell writes of soundings into the square base of the brickwork, seeming to think that this could only be explained by encasement of the round tower within a square bastion, as at the mosque of al-Hakim : since the soundings showed that “the lower part…forms one with the whole minaret” he concludes “[it]…can only have been built by Lajin.” Other suggested reasons for the unoriginality of the minaret are that it is made of limestone quarried from the Muqattam hills, not brick like the rest of the building; the style of the bridge and the horse-shoe arches that decorate it and the square base were only introduced into Egypt by Andalusi and Maghrebi craftsmen in the late C13th/early C14th; also the junction of the bridge and the wall of the mosque occurs in the middle of a window.

Only Doris Behrens-Abouseif has sensibly argued that these suggestions do not prove the whole thing is a Mamluk construction. While the uppermost storey and its mabkhara dome are clearly Lajin’s work, the use of stone means nothing as al-Muqaddasi saw the original minaret in the C10th and describes it as made of stone, with the staircase on the outside. Also the bridge and the horse-shoe arches should not be taken to date the whole structure: they may be part of a later embellishment, and the bridge is clearly later because it blocks a window. In fact the minaret was probably not even connected to the mosque at all originally, just as the minarets at Samarra were free-standing inside the ziyadas. Furthermore, there is the story related by Nasir-i-Khusraw who travelled through Egypt in 1047: the descendants of Ibn Tulun sold the mosque to the Fatimid Caliph al-Hakim for 30,000 dinars, and then they began to demolish the minaret. When he rushed to the site and demanded what they were doing, they replied they had sold him the mosque, not the minaret. He promptly paid them another 5,000 dinars for the minaret. This surely means that the minaret was not attached to the mosque.

Lastly, if Lajin had reconstructed the minaret, he would have done so in an entirely Mamluk style; after all, he constructed other features such as the mabkhara storey, the domed fawwara, the dome above the mihrab, all in a wholly Mamluk style. In any case neither the historical sources, nor even the waqf document of Sultan Lajin itself, mention the rebuilding of the minaret, nor does the minaret bear any historical inscriptions declaring that it was rebuilt.

So we have a minaret free-standing behind the north wall of the mosque and detached from it, as at Samarra. The minaret must be still in its original form, with some later embellishments, and when al-Quda’i says that the minaret was built copying (‘ala bina) the Samarra style, it is this very minaret that he is describing. The upper storeys presumably continued their anti-clockwise turns to end in a pavilion, like at Samarra. Its square socle (21.35 m high) is much more emphasised than the Samarra minarets, which were only 2 - 3 m high: it rises another 6 m above the top of the mosque wall, before the circular storey began, taking the whole structure to a height of 40 m. Circular storeys would be more extensive in their use of bricks than a rectangular socle and as the lower storeys were less visible above the mosque and ziyada walls, it may have been judged more economical to emphasise the square base. The minaret’s massive proportions dominate the mosque interior, and would have dominated the markets and plains of al-Qata’i’, being visible perhaps as far away as the old quarter of Fustat.

It was not intended to be used for the call to prayer, as we know the fountain building was used for this; and in the story that Herzfeld was told about the Malwiyya, the adhan is not mentioned as one of the functions of that minaret [it would be rather impractical as he’d be out of breath by the top!]. So the presence of this unique minaret, the first in Egypt, and the conspicuous nature of its design, made it directly symbolic of the ‘Abbasid towers in Samarra, even though this relationship may only have been apparent to those who had lived in or visited that city. Why, when we know that Ibn Tulun rebelled against the authority of the Abbasid regime?

Stucco:

I would like lastly to cover the subject of the stucco and woodwork decoration, which is the other feature of this mosque which conspicuously imitates Samarra. In contrast with the plain space of the whitewashed brick walls, many facets in this mosque are covered with intricate geometric and floral patterns which have been incised while the plaster was still wet, not produced by a mould. This implies skilled craftsmen who knew how to produce virtuoso effects, and as the style of this stucco decoration is now universally called “Samarran” perhaps we can infer that Ibn Tulun actually had Iraqi craftsmen with him, or had sent for them. Samarra stucco techniques can be divided into three basic styles (A, B and C) which existed simultaneously throughout the C9th, developing in a trend towards abstraction. Creswell thinks Ibn Tulun’s mosque displays a mixture of Styles A and B because there is very little use of moulding, but the woodwork in the lintels of the mosque doors employs C, also called the bevelled style.

Such decoration fills the soffits of all the large arches in the riwaqs, the column capitals at the corners of the piers, the bands that surround the archivolts of the arches and run across the tops of the piers, other bands that surround the windows, or run along in repetitive friezes. Each design is entirely unique. The soffits of the windows are also decorated, and a band of Kufic inscription in stucco surrounds them: each of the 128 windows contains a grille carved from stucco, each with a unique geometric design. Stucco is also used to frame the main mihrab, for the secondary Tulunid mihrabs, and to decorate the façades: the arches in the spandrels are flanked by rosettes of different designs, and along the very top of the wall is a continuous rosette frieze.

The presence of this technique of stucco decoration again evokes the buildings of Samarra, though again, it may only be a reference recognisable to those who had seen Samarra, which was not the majority of the population of Egypt. The people, then, were perhaps not the primary intended audience of this building: who were? The answer must surely be the Abbasids, who finally in 905 destroyed all trace of al-Qata’i’, leaving only the mosque because of its holy status. Perhaps they decided to reinterpret it as a faithful attempt to copy those two great mosques back home, to which it owed so many architectural elements.

But bearing in mind Ibn Tulun’s dynastic hold on this region, his refusal to send tribute to his overlords, his adoption of caliphal prerogatives such as minting coins and including his name in the khutba, we should instead interpret his mosque as a political symbol: a building in the style of Samarra but incorporating lessons learnt from it, producing something more graceful and accessible, though innovative and suspicious to the Egyptians. His architectural projects bore his name as patron, thus he made a competitive statement to the ‘Abbasids that he too could build like a king.