L’HISTOIRE DE L’EAU ET

DES INSTALLATIONS HYDRAULIQUES

DANS LE BASSIN DE .

(pages 59-64)

TEXTE DE FAOUZI MAHFOUDH SAMIR BACCOUCH BECHIR YAZIDI

TUNIS NOVEMBRE 2004

(Retrieved on October 23, 2008, from http://www.iwmi.cgiar.org/assessment/FILES/word/ProjectDocuments/Merguellil/Histoire%20eau%20Kairouan.pdf ) 6- PROPAGATION OF THE KAIROUAN MODEL

We saw through the preceding developments that a particular kind of circular cistern was developed in Kairouan and its zone of influence as far as the northern borders and more precisely those with the Proconsulaire . This type had been identified as Aghlabide. Regarding them, Solignac has written: “On this subject, one can pose the question of the origin of this type, of which the basins of Dahmani seem to have been among the first applications. Research in Persian, Arabic, Syrian, Egyptian and Byzantine literature and iconography, has been made and did not provide any frame of comparison. It thus indicates, probably, a new and original process … a conclusion therefore is essential, that it is indeed a Moslem technique and that it is specifically African” 77 .

It is now well established that the circular basins are not specifically African. The work of Saad Al-Rashid on the pilgrimage route between Kûfa and Mecca, named Darb Zubaydah, revealed several tanks similar to those which had been believed to be typically African 78 . Among the best known examples, one quotes those of Birkat Ath Thulaymah (site n°6), Birkat Al Rashidiya (site n° 10), Birkatt Tuzzi and Birkat Uraynah (site n° 25 et° 26), Dlay' ash Shaqaq (site n°31) and Birkat Al - Kharabah (site n° 38). From the existen ce of the same type of tanks in these regions of the East one supposes the technique and form was known outside of . No doubt a good knowledge of the Muslim world, in particular of the central , would allow us to identify other examples.

Moreover, and in the light of the latest archaeological explorations carried out around the country, one can affirm that a good number of these circular cisterns known as Aghlabid, are in fact of Roman origin. This is the testimony of the investigations carried out by Kh. Annabi, in an area between the towns of Kairouan and El-Djem. In this zone of the low Tunisian steppe, the author of the excavation was able to bring to light a great number of sites (176), in an eminently traditional archaeological context, of which the antique origin cannot be questioned. These are primarily from Ain , Bîr Maghanma and Kondar. These sites conceal a multitude of remains characteristic of the Roman period: thermal baths, oil mills, mills, pottery kilns, necropoles, mausoleums, elements of architecture and decoration, pottery (amphoras), a Roman road, etc… But what interests us here it is the very great number of circular cisterns with or without buttresses. Having been generally discovered in a Roman context, the existence of these tanks can only be explained only by presence of an ancient town. This context had encouraged Kh. Annabi to wonder if Mr. Solignac had visited the few buildings which he described; even if the dating that he allotted to them could be retained, based on the analysis of the mortar 79 , it is not out of the question that the ancient buildings were restored later on.

It is essential to state, with considerable significance for our topic, that the datings established by Solignac take account only of the nature and the composition of the mortar. However, it has been shown that medieval and modern Ifriqiya remained, to a great extent, faithful to the ancient techniques and construction materials. The proportions of the various components of the mortar are almost unchanged. In any case, they vary less in time than in space and one undoubtedly observes regional variations due to differences in the nature of the materials available. It will thus be necessary to take care to avoid drawing hasty conclusions.

Lastly, it is now necessary to to take account of other discoveries which enrich the chart by Solignac, all of which correct the geographical distribution that he proposed for the installations of the circular type whose southernmost limit was fixed by the French scientist to the north of the 200 mm isohyet line.

Among the monuments discovered in the zone delimited by the author of the hydraulic installations of Kairouan and the low steppes, one must underline the Aghlabid basins of . One of these basins was already known without Solignac describing it. We studied it in our thesis on the monuments of Sfax. This basin, today in the area of the cemetery of Sfax, is composed of a small decantation reservoir equipped with external and internal buttresses, a reserve cistern a score of meters in diameter, equipped with cylindrical buttresses of the same type as those of Kairouan, and of a group of vaulted supply cisterns 80 . But the most recent discovery and one which deserves more interest is that which was made by A. Louhichi. It is Fesqiyet el Fendri, also called Fesqiyet Borj el Ksar and is located at the foot of the south-western ramparts of the medina of Sfax [Ed. perhaps ‘north-western’ would be more accurate, though the medina is not actually aligned N-S, E-W, but about 23 degrees off that, in the direction of Mecca]. This basin was announced by many researchers and writers 81 . Today, we know its form and the details of its construction 82 .

It is an ensemble aligned according to a north-south axis and composed of a circular basin of decantation, a large circular basin of reserve and an arched supply cistern. The first two basins are equipped with cylindrical semi buttresses recalling those of Kairouan. This unit obviously appears more important than that already known 83 , as its dimensions testify. Thus, the decantation tank measures 10 meters (external diameter), the reserve basin 33.30 meters and the supply cistern is 10 meters long and 2.5 meters wide. But what draws our attention more particularly, is that part of the supply cistern is found below the ramparts, thus inviting us to re- examine the chronology of this complex hydraulics.

The ramparts of Sfax were built, as the sources testify, during the Aghlabid period. Thus Labîdî dates them in the year 234/849 84 , whereas Ibn Idhârî and Ibn Nâjî, recalling Tujîbî, place their construction in the year 245/860 85 . The epigraphy and the nature of construction support the texts 86 . From its position, at the lower part of ramparts, it is necessary to wonder whether Fesqiyet Al-Fendri should be regarded as being earlier than this period or if this state results from a change in the layout of the ramparts? A study of the maritime walls of the same city could show that its circuit, having been moved towards the south, involved the enlargement of a few meters urban space 87 . In the case which occupies us, a survey carried out by A. Louhichi had enabled the discovery of a treasure in the bastion built over the drains. It is composed of pieces of money dating from the end of 6th (12 th ) century, or the first half of 7th (13 th ) 88 . This chronology is interesting especially if one remembers that, at the time of a short stay that he made in Sfax, the Almohad caliph Al-Nâsir had had built a great hydraulic complex made up of a few hundred cisterns called, since then, the “Nâsriya” 89 . It is thus extremely probable that at that time the enclosure was further strengthened, by raising new fortifications which span the large Aghlabid basin, this one no longer being, by then, the only source of water.

The rediscovery of this Aghlabid basin confirms the idea of the adoption of the urban layout of Kairouan by the city of Sfax. Similarly it shows the density of the basins which became more widespread in the 3rd and 4th (9th and 10 th ) centuries.

Not far from Sfax, in the archipelago of Kerkéna, we discovered a series of basins, all located in the small island of Sefnou which is isolated today. The state of the place, as well as the current trend of rising sea levels, leave us persuaded that this complex was built in a different context. A geomorphological observation could show that during antiquity and into the , the small island of Sefnou was attached to the large island of Kerkéna: Chargui. These cisterns of which there is hardly a mention in the ancient sources, nor in the authors of the medieval period, occupy the south-western end of the small island, and are, thus, a few meters across from Chergui.

Four basins in a row were identified. The first, reinforced by external buttresses, draws a circle approximately 10 meters in diameter. A small sump of 1 m broad accomodates the water of a nearby well. The second fesqiya, located 20 meters from the first, is quite damaged; nothing remains but the vestiges of a single circular basin 14 meters in diameter, with semi-cylindrical external buttresses. 32 m to the south of this one, is the third group of cisterns; it is composed of two circular basins with external buttresses, of 10 and 17 m diameter. This arrangement which recalls the hydraulic installations of Kairouan, attests to the existence of a decantation chamber and a reserve basin. Lastly, the fourth group of basins is within a few meters of B3. They comprise two circular reserve basins 14 m in diameter approximately, paired one against the other and equipped with semi- cylindrical buttresses. There too, there is a decantation unit and another for reserve.

One is astonished by this great concentration of basins in a zone so small as that of the small island of Sefnou. It does not seem, indeed, that this ensemble alone, quite complex as it remains, was intended to supply the only small island of Sefnou, but water which was there collected was useful, undoubtedly, for Chergui, in the period when both formed a single island. In addition, such a concentration of cisterns caused us to ask ouselves whether such important reserves were intended to provide drinking water to the inhabitants of the island, or if they were used in the irrigation of the gardens. The second assumption seems to us the most probable, in as much as the old sources evoke, strongly, the agricultural activity of archipelago 90 .

However, it is not easy to be precise with the dating of these basins. First of all the archaeological context pleads in favour of a later chronology: the ceramics collected in the immediate surroundings of the basins, is comprised of fragments of Punic amphoras (Mana C and D), by a great quantity of ceramics with black varnish and by a great deal of clear African sigillée of the 3 rd and 4th centuries AD. The vestiges met with go up almost to the Roman period: such is the section of wall which is drawn up against the western cliff or the pottery kiln which is in the center of the small island. It would, however, be imprudent to hastily accept an ancient origin for the hydraulic installations of Sefnou, even though Bakrî identifies Kerkéna as the island of the cisterns. Thus, let us be us brought to suggest, at least on a purely conjectural basis, the re-use of the cisterns during medieval times. [HERE]

To close the dossier on hydraulics, let us mention finally the basins of Gightis. The site is well to the south of the current village of Maharès, regarded by Solignac as southernmost limit of the distribution of basins of the Aghlabid type. This discovery takes on a particular importance because it shows that the limits are not too rigid as one would believe. It seems to us, on one side, that it is a further proof of the unity of Ifriqiya, and on the other, the value of an exploration which would cover the whole of the territory. Located at the bottom of the Gulf of Bou Grara which marks the southernmost limit of Small Syrte [Gulf of Gabes], Gightis receives and average of 150 mm of rain annually.

The city is attested from the 6 th century AD. During the Roman period, it formed part of the Proconsulaire for at least three centuries, from the end of 3rd century AD, slong with the province of Tripolitaine. In the 5 th century, it was the seat of a Bishopric represented at the conference of in 411. In the 6th century the Byzantines built an imposing fortress and restored some buildings in the south of the forum. On this ancient site, Ali Drine uncovered five circular basins of which the chronology does not seem to pose any problem. Re-use of the stones recovered on the ancient site, the absence of ancient mortar and the abundance of Islamic ceramics are as some of the indicators which permit us to date the construction to these installations in the beginning of the Aghlabid period, whereas the date of abandonment remains uncertain 91 . This attribution is strengthened by Al-Tijânî who accompanied the Hafsid prince to the Tunisian south into 1306-1309 92 . In effect, Al-Tijânî describes Tajjeght as: “a spacious plain where there are ruins of various constructions and hydraulic works of all kinds”. As has been well demonstrated by A. Drine, the abundance of hydraulic installations comes from the number that explains the qualifier of “all kinds” 93 .

Through the examples of these hydraulic installations we wanted to insist on the importance of the work already accomplished by Solignac, but especially on the task assigned to us. In the past, there was a tendency to " to age the monuments" and to assign all that is built to endure to the Romans, whereas brick or cob constructions were considered to be from the Arab time. Thereafter, and in a kind of reaction to this tendency, many buildings were wrongly attached to the Middle Ages while being based on quite controversial examples, and without account being taken of the archaeological context. Indeed, on one side, a certain negligence will be noticed concerning the instruction which ceramics can provide, and on the other, of the phenomenon of permanence in materials and the techniques of construction lasting of long periods. They, just as their predecessors the Romans did not revolutionize the art of construction; they were generally satisfied to perpetuate the same methods and the same forms, their builders, at least during the first period, undoubtedly being natives. The work on hydraulics offers, in our eyes, a field which will be able to bring still more instruction on life in the Middle Ages. However, it is not necessary to be hasty in drawing borders and tight limits between Ifriqiya and Africa, between western and eastern Moslems and between the various zones of Ifriqiya.

Nowadays one can affirm, without risk of error, that the circular basins were well-known in Roman Africa; they are also known in Arabia and Iraq. In Ifriqiya, one finds them in the area of Kairouan, in the Sahel, on the islands of Kerkéna and even in the south of contemporary . Exploration is still capable of enriching our knowledge even further.

References

77 SOLIGNAC, 1952, p. 28. 78 Darb Zubaydah , Riyadh University Bookshops, 1980; See also other examples of circular cisterns in the study of T.J. WILKINSON, “Darb Zubaydah: The Water resources”, ATLAL , 1980, p. 51-67. 79 Mr. KH. ANNABI, “Archaeology of region of Wadi Cherita” Bulletin of the work of the National institute of archaeology and art , Reports, Fasc. April 4, June, 1991. p. 7 - 28. 80 cf F. MAHFOUDH, “Hydraulic installations of Sfax”, IBLA, 1991, n° 167, p. 13-29. 81 The quote here is from MONLEZUN, “The site of Sfax”, Bulletin de la Géographie Historique et Descriptive , , 1896, ABDALKÉFI, tarikh safaqus , Sfax, 1960; as well as the map of the capture of Sfax in 1881. See our thesis. 82 A. LOUHICHI, “Excavations of an Islamic site in the médina of Sfax: Basins of Borj Al-Qasr”, Africa XV, 1997, p. 91-112. 83 F. MAHFOUDH, IBLA, 1991, p. 13-29. 84 LABÎDÎ, Manâqib , p. 197-198. 85 Bayân , I, p. 111. Ibn NÂJÎ, II, p. 147. 86 F. MAHFOUDH, “Aspects of the fortification of the coast of Ifriqiya: the defenses of Sfax” BAC , 1987-1988, p. 235-250. Paris, 1992. 87 F. MAHFOUDH, BAC , 1987-1988, p. 235-250. Paris, 1992. 88 L. RAHMOUNI, “A treasure of Almohad coins coming from Sfax”, Africa , XV, 1997, p. 113-116. 89 MAGDISH, Nuzhat , T. II, p. 179. It concerns Al-Nasir son of Yaqûb son of Abd el Moumin. 90 KOLENDO J. “The economic role of the Kerkéna islands in the first century before our era”, BAC , 1981, p. 241-249. See also our article “Kerkéna in the accounts of the ancient and Arab geographers” to appear in Africa Romana 13. 91 DRINE A., “Hydraulic installations of Gigthi”, Africa Romana , XI, 1994, p. 683 - 692. 92 On the visit to the South by Tijani, see A. GRAGUEB, “Notes on the travels of Tijani in the Tunisian south”, CT , 1976, p. 25-44. 93 DRINE A., Africa Romana , XI, p. 683-692.