1 Introduction

Looking back over the past decade, we can say that the study of estates in the Islamic West has intensified as a consequence of a growing interest in the urban periphery, rural spaces, and the symbiosis between architecture, gardening, and agriculture. This increased activity has yielded numerous works and monographs that reveal the sumptuous nature of almunias devoted to recreation and productivity on the one hand, and to setting the stage for a display of power on the other.¹ Outstanding examples include al-Rummāniyya in Cordoba, Generalife in Granada, Agdāl in Marrakesh and Castillejo in Murcia, all of which were designed by some of the most renowned dynasties in al-Andalus and the Maghreb and remain the best preserved to this day.² However, these represent just a very small fraction of the many estates created by rulers in this context, and are indeed an insignificant sample if we bear in mind the abundance of this type of estate among the elites and other wealthy social strata.

Remarkably, while it has been recognized that almost all dynasties that governed al-Maġrib al-Aqṣā since the eleventh century (the Almoravids, Almohads, Saadids, and Alaouites) founded estates, nothing has been presented regarding the Marinid dynasty that ruled this same territory while the Nasrids reigned in Granada, between the thirteenth and fifteenth centuries. The purpose of this paper is therefore to present what was most likely the Marinid formal estate, although its memory and remains have been more badly damaged by the passage of time and historical vicissitudes than other such properties.

It is only thanks to the testimonies of certain chroniclers that we know about the existence of a famous estate connected to the palatial city of al-Madīna al-Bayḍāʾ (The White City), or Fās al-Ǧadīd (New Fez), the seat of the Marinid Sultanate. This was nothing but an almunia located extramuros to the north of the city, with several pavilions and landscaped spaces. However, its location has not always been clear. Initially, Louis Massignon (1906, pp. 219–236), aware of its existence, proposed that it sat in the Boujloud or Sbil garden, next to a small noria (Café La Noria), and it was not until Bressolette and Delarozière that the true site was identified.

Between 1938 and 1939, on several sites in the area and with the support of aerial photographs, the latter authors performed a surface exploration, as well as small clearing works, leading to the discovery of a set of remains inside the Bāb al-Sākima³ cemetery that could be clearly assigned to the ancient Marinid estate. The most significant vestiges that remain from the renowned al-Muṣāra, however, are not at that site: we refer to a great noria on the Fez River and a long aqueduct that hauled water from there all the way to the estate. Both structures are in a reasonably good state of conservation, which attracted Bressolette and Delarozière’s attention (Bressolette and Delarozière, 1978–79; Delarozière and Bressolette, 1939). Although they did not study the find in sufficient depth, their work resulted in two articles that are the only attempts to address this topic to date. One is devoted to the estate and the other to the noria and aqueduct.⁴ It is indeed shocking that, after these works shed light on the remains, nobody decided to pick up their work, with the exception of some indirect but valuable approaches (El Idrissi el Omari, 2018, pp. 117–119).

However, although the above works are valuable as they offer information on features that to some extent have disappeared, they are rather limited. On the one hand, except for a few maps produced by Bressolette and Delarozière (1978–79, plans II and III) of two ruined pavilions, in general their works lack the planimetric support that would enable correct documentation and analysis of the remains. In this respect, it is striking that they never created plans of the noria and the aqueduct. On the other hand, despite their remarkable visual survey of the structures, there is no accurate diachronic analysis of the whole site in relation to the city of New Fez. Therefore, as part of our research project on the palatial city, it proved essential to undertake a new study of the vestiges of the estate, a task that was tackled from a new perspective using contemporary tools. This is the only way in which to reinforce the documentation and analysis of the remains. Firstly, the architectural survey of the existing structures was carried out using photogrammetry, a highly precise tool that considerably facilitates the work in highly complex cases. Nevertheless, this task has been also complemented with traditional in situ measurements.⁵ Secondly, the analysis of the waterwheel and the aqueduct has been undertaken following principles specific to the archeology of architecture, which has allowed us to propose a relative sequence of phases in the evolution of the ensemble, contrast some observations previously made by Delarozière and Bressolette, and advance the functional interpretation of the structures. Finally, with regard to written sources, a deep search and analytical translation has revealed new details and increased the dialog between documentary and material evidence.

2 Urban and Natural Landscape

The city of Fez represents one of the most important urban centres of the Islamic West and has therefore, since its foundation between the eighth and ninth centuries, undergone a very intense evolutionary process. In general terms, and according to the archaeological evidence and some descriptions by Medieval chroniclers, the historical city (Old Fez) was configured on two different settlements (al-Andalusiyyīn and al-Qarawiyyīn), which were separated by the river Fez. However, during the twelfth century the city was presumably united within one single enclosure by the Almoravids or Almohads. Later, in 1276, the Marinid ruler Abū Yūsuf Yaʿqūb (r. 1259–86) decided to carry out the construction of New Fez to the southwest, as an isolated city equipped to fulfill three main functions: administration, military management, and residence for the sovereign and courtiers.⁶ Accordingly, the location of New Fez was strategic for many reasons, but particularly because it facilitated the withdrawal of water from the river Fez with priority over Old Fez, which was essential to ensure the supply of water to the royal city and the estate. In this context, it is worth noting that by the end of the Middle Ages, Fez already reached a significant size, and this necessitated a rigorous and complex management of the water provided by the river and other, complementary sources located in the surrounding area.⁷

The Marinid estate was to the north of New Fez and, although its boundaries are still uncertain, we know with some certainty that it spanned an area with a slight slope that extends from north to south, all the way to the Fez River (Fig. 1). It seems as though it was an east-west facing area, the geographical boundaries of which were the river to the south and a change in the topographical conditions of the terrain to the north, where the slope becomes somewhat steeper.

View Full Size

Fez. To the east, Old Fez; to the southwest, New Fez; and to the northwest, the cemetery of Bāb al-Sākima where Ǧannat al-Muṣāra once existed

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

Author’s elaboration of a GoogleEarth map

• Download Figure
• Download figure as PowerPoint slide

It must be noted that the intense traffic up to the present in this zone, near the major gates of the two cities (New Fez and Old Fez), as well as the various changes in the Fez River, impede the study of the estate significantly; in particular with regard to aspects such as the characteristics of the environment where it was founded, the operations needed to set it up, and its physical connection with the palatial city. Interventions in the surroundings of the estate, which are still ongoing, have been very aggressive, such as the Fez River being canalized many times and following different courses, the creation of an industrial area, the expansion of cemeteries, road projects, and, ultimately, urbanization.

Regarding the river, it is very likely that it would have run very close to the north side of the early palatial city (thirteenth century), or more specifically along its walls (Fig. 2), since the beginning. Currently, we do not know whether the outer wall had already been planned in the early phase. In any case, testimonies and evidence indicate that along some sections of the north of the city, the river ran between the inner wall and the outer wall. Also, there is no doubt that its course under the bridge preceding the north gate may have remained relatively unchanged (Bāb al-Qanṭara or Bāb al-Wādī).

View Full Size

Ǧannat al-Muṣāra and New Fez (at the end of the Marinid period). The archaeological remains of the estate are integrated into this hypothetical plan of the city, produced according to current and ancient cartographies as well as written sources.

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

Plan by the author

• Download Figure
• Download figure as PowerPoint slide

Considering the topographical and hydrological conditions, there were two possible solutions to irrigate the estate using water from the river. One would be to create a canal diverting water upstream, which would involve building a long ditch that could prove insufficient, since the source is not very high. The other would be to build a mechanical contraption to elevate the water from the river and guide it through an infrastructure to the estate. According to written sources that will be presented below and to the preserved remains, the second option was without a doubt the one finally implemented. However, it is possible that other water resources such as the now disappeared natural spring called Aïn el Gorna, which was depicted on the 1933 Plan de Fès (Service géographique du Maroc, 1933), were used additionally. Of course, the appropriation of water for the almunia may have been a controversial subject, as it meant a considerable reduction of the river’s flow rate and the seizure of the sources, which might have raised concerns among the landowners and beneficiaries who had used that water previously – but about this we have no information.

On another note, one of the most remarkable urban elements in this environment is the historic westward road departing from the Bāb al-Maḥrūq gate, in the old town. It must have existed since ancient times, and it was one of the busiest routes, as it connected Old Fez with neighboring Meknès, as well as with Rabat on the Atlantic coast. In fact, this route was already consolidated before the foundation of New Fez, so the layout of Bāb al-Qanṭara or Bāb al-Wādī, the north gate of the early palatial city, south of the river, may have been intended to connect with this road, which would explain why it was always equipped with some structure to overcome the obstacle of the river. It could even be argued that, thanks to its integration in the road system, this access became the main one in New Fez, and it evolved throughout history with an external design that ended up connecting directly with the road. Initially, it was complemented during the late Marinid period with the construction of an advanced gate (Bāb al-Sabʿ) and an intermediate plaza-bridge (Figs. 1 and 2). Then in the nineteenth century, during the Alawite period, the course of this road between New Fez and the Kasbah Cherarda was the rationale for the complex gate system of Bāb al-Sākima and Bāb al-Makīna/Bāb al-Saʿāda, some of which has since disappeared.

For this reason, if the river flowed along next to the walls, it would be logical that the road would have run to the north of it since Medieval times, perhaps in parallel, thus avoiding unnecessary crossings and taking the current direction of the Route de Meknès or Abu Bakr Ibn al-Arabi Avenue. With these conditions and preexisting elements, we can assume that the limits of the estate must have been constrained, at least to the north of the route, in such a way that a strip would have been devoted for to route between the almunia and the palatial city, or rather, between the almunia and the river. Had the estate been adjacent to the walls of New Fez, as proposed by Bressolette and Delarozière, that would have created an anomalous situation, as both the road and the river would have run inside it. Factors in favor of our thesis include: first, this type of orchard was always defended, often with walls, to protect the crops and private spaces. Second, the two colossal octagonal towers that are part of the aqueduct flank the stretch that crosses the route, as if attempting to highlight it over the general course of the aqueduct. Unfortunately, the construction of the Kasbah Cherarda, the creation of the Bāb al-Sākima – Bāb al-Makīna/Bāb al-Saʿāda gate, and the building of the new road involved the destruction of this singular stretch of the aqueduct, which was the only section that might have been crossed by a monumental arched gate.

3 Historical Context

Although the archaeological remains of the almunia are limited, based on the written sources it seems to have become a widely known space and the scene of numerous memorable events, with the result that Marinid chroniclers left ample records of its construction and various changes relating to the estate. Let us start by referring to the work Rawḍ al-Qirṭās by Ibn Abī Zarʿ (d. 1326), the only one to provide dates about the possible origin of the estate as a Marinid project:

That year [685 H / 1286] the great noria (al-nāʿūra al-kubrā) of the Fez River was built, starting in the month of raǧab, and it began turning in the month of ṣafar in the year 686 H (1287) […]. In that year, al-Muṣāra was filled with plants, and the White House (al-Dār al-Bayḍāʾ) of the New City was built. (Ibn Abī Zarʿ, 1964, p. 737; Ibn Abī Zarʿ, 1972–73, p. 407)

Those dates had witnessed recent political changes, as Abū Yūsuf Yaʿqūb passed away in the first month of the year 685 H (1258–1286), while it was during the second month that Abū Yaʿqūb Yūsuf, his successor, was sworn in (Ibn al-Aḥmar, 1962, pp. 30 and 38–39). So this text, that we may consider as closest to the facts, makes it clear that the plantation and beginning of operations of the noria correspond to the rule of Abū Yaʿqūb Yūsuf.

Having said this, the creation of the almunia, with irrigated plantations, would not have been possible without ensuring a sufficient water supply. This is why it seems as though the Fez River noria was always connected to its origin. Moreover, this hydraulic system is frequently mentioned by chroniclers and travelers, to the point that al-ʿUmarī (1301–1349) himself refers to it as follows in his Masālik al-abṣār:

On the river lies the famous noria (nāʿūra) that hauls water to the sultan’s orchard (bustān), known as al-Muṣāra. It is a magnificent orchard that contains a beautiful pavilion (qubba). This orchard is outside the New City, and the noria, which is remembered with celebration, has become proverbial and people talk about it. (Al-ʿUmarī, 1927, pp. 153–158; al-ʿUmarī, 2010, IV, pp. 89–90)

We also have information about the sultan who commissioned it and the identity of the engineer who designed it, as the renowned Ibn al-Ḫaṭīb included in his work al-Iḥāṭa the biography of the engineer, Muḥammad b. ʿAlī b. ʿAbd Allāh b. Muḥammad Ibn al-Ḥāǧǧ:

His father was a Mudéjar (mudaǧǧan) carpenter from Seville, of the ones that know how to build mechanical contraptions. Well versed in manufacturing and using war machines. He moved to Fez in the times of Abū Yūsuf al-Manṣūr b. ʿAbd al-Ḥaqq and built for him the wheel (dawlāb) of wide diameter and long reach, smoothed in the center and the contours, with numerous scoops (akwāb) and hidden movement, now located in the New City, the Royal House of the city of Fez, and that constitutes one of the monuments that travelers want to see. He was also responsible for building the shipyard (Dār al-Ṣanaʿa) in Salé. After his father’s death, he presented himself before the second Nasrid sultan. (Ibn al-Ḫaṭīb, 2003, II, p. 81)

This reinforces the hypothesis that the almunia was Yaʿqūb al-Manṣūr’s endeavor after the palatial city had been finished, although it was his successor who concluded it. It is indeed logical to think that the creation and development of the estate would be subject to the successive contribution of different rulers. Therefore, as it happened in the palatial city, each ruler would have had the chance to include his own additions and constructions.

Regarding the uses of this almunia, in addition to being a place to relax and grow produce, it was also an ideal environment for receptions and for housing distinguished guests amongst the irrigated crops. This is evidenced in a note written by Ibn Ḫaldūn about the visit of the Nasrid sultan Muḥammad IV to negotiate an alliance against Alfonso XI:

In the year 732 H (1331–32), the lord of al-Andalus [Muḥammad IV of Granada] was appalled by the advance of the Christian king, so he crossed the sea and appeared before the sultan Abū al-Ḥasan, who was at the time in the royal house of Fez. The Sultan received him with great honors, sent a retinue to meet him, and housed him at Rawḍ al-Muṣāra, adjacent to his palace. (Ibn Ḫaldūn, 1971, VII, p. 255; Ibn Ḫaldūn, 1969, IV, p. 217)

In effect, the palatial buildings where both the sultan and his guests were accommodated must have been one of the most important elements in the almunia. Al-ʿUmarī himself mentioned a qubba, and even al-Numayrī (d. 1383), secretary and panegyrist of the sultan Abū ʿInān, informs us about the role played by its palaces and compares them, exaggerating, with mythical examples of the East:

Our lord left for al-Muṣāra and stayed in its palace (qaṣr), before which Ḫawarnaq and Sadīr pale in comparison. Also, it makes one forget al-Ǧaʿfariyya, that was inhabited by the first al-Mutawakkil, as this [palace] is inhabited by the last al-Mutawakkil. Al-Muṣāra housed the state elites, notable servants, people of high rank due to their obedience and honor, as well as the heads of the troops. (Al-Numayrī, 1990, p. 173)

However, of all the notable structures that must have existed in al-Muṣāra, written sources only mention the Qubbat al-ʿArḍ (troop inspection pavilion). More specifically, the two references that we have come from Ibn al-Ḫaṭīb, who narrates the return of Muḥammad V to Granada after his exile in the Marinid court. The first forms part of al-Iḥāṭa, and reads:

On the 7th of the month of šawwāl [of the year 762 H / 1361], and after arranging everything necessary to this end with instruments and fineries, the sultan [Abū Sālim Ibrāhīm] sat in the Qubbat al-ʿArḍ, outside Ǧannat al-Muṣāra, to bid him farewell [Muḥammad V]. Then, the crowds arrived to witness the parade for the lamented one. (Ibn al-Ḫaṭīb, 2003, II, p. 13)

The second is included in Nufāḍat and, although it does not differ much, it contains relevant nuances:

On the morning of Saturday, 17th of šawwāl of the year 762 H (1360), the Sultan, prince of Muslims of al-Maġrib [Abū Sālim Ibrāhīm], sat in the Qubbat al-ʿArḍ, that lies in the Ǧannat al-Muṣāra, whilst the announcement took place among the cheering crowd in the open space and while the banners, kettledrums and money recipients were gathered. Then, the Sultan [Muḥammad V of Granada] was summoned, so he walked up to the qubba and then descended, wearing the king’s vestment (ḫilʿa). (Ibn al-Ḫaṭīb, 1985, p. 184)

As both fragments reveal, the qubba was an outstanding construction that could have had a certain height and that was connected to the almunia. However, it would be interesting to establish its position regarding the estate, as Ibn al-Ḫaṭīb refers to it in two different ways. On one occasion he uses the expression “bi-Qubbat al-ʿArḍ bi-ẓāhir Ǧannat al-Muṣāra,” making it clear that it is on the outside, but then, in Nufāḍat he says “fī Qubbat al-ʿArḍ al-muttaḫaḏa bi-Ǧannat al-Muṣāra,” a form that seems to connect the two places physically. This way, if we consider the spatial context described for this pavilion, we could present the hypothesis of a structure attached to a perimeter wall and projected outwards. This expression would make sense, as would the function of a pavilion opening out to the adjacent esplanade.

In fact, the meaning of the word “ʿArḍ” refers to exhibition, and more precisely within a military context, it alludes to the inspection, review or parading of the troops. Therefore, it could be understood as a pavilion from which to supervise the troops gathered in the open adjacent space. A similar example would be that of Madīnat al-Zahrāʾ (Cordoba), the greatest exponent of a palatial city in the Islamic West. The Parade Ground of the ensemble is a large square space that precedes the administrative-palatial spaces and that hosted military parades, public audiences and court proceedings. Notably, the west side of this plaza was closed by a great monumental portico with an upper platform and an elevated pavilion in the main symmetry axis; the latter serving as a chamber from which the caliph could watch the events (Almagro, 1996b, pp. 214–216; Arnold, 2017, p. 76).

Going back to the norias, the work Fayḍ al-ʿUbāb by the secretary al-Numayrī is quite revealing, informing us of the existence of several wheels and of the interventions of the sultan Abū ʿInān (r. 1348–58). Although it is presented in a somewhat ambiguous way, we understand that there were in the area at least one large noria (al-nāʿūra al-ʿuẓmā) that irrigated the garden (rawḍ), two new wheels commissioned by Abū ʿInān at the top of the garden, a group of three small norias inside al-Muṣāra, a water wheel next to the zāwiya al-Mutawakkiliyya that could be the previously mentioned large noria and, finally, a noria with an aqueduct (qanāt) commissioned by Abū ʿInān that replaced the water wheel to increase its reach and flow rate (al-Numayrī, 1990, pp. 174–181 and 211–213). These textual fragments about these structures are indeed quite confusing, as the author does not provide many details about their location and uses, but rather focuses on highlighting his feelings and uses metaphors and celestial comparisons that refer repeatedly to the scoops, water, noise, movement, elevation, and the fruit production. Therefore, thanks to this source, we surmise that there were several contraptions that served the almunia, although it is impossible to identify which of them corresponds to the well-known noria whose remains have survived to our day.

Also, as reflected by some sources, the noria and the orchard created an attractive landscape for the poets of that time, as described in a fragment of a qaṣīda composed by the poet Aḥmad b. Yaḥyā b. Aḥmad b. ʿAbd al-Mannān al-Anṣārī al-Ḫazraǧī for the sultan Abū ʿInān, which was later included in a compilation by Abū al-Walīd Ismāʿīl Ibn al-Aḥmar (Ibn al-Aḥmar, 1976, pp. 343–353; al-Mannūnī, 1979, p. 45). More specifically, we refer to verses 65–72, which mention the great noria and some aspects of the garden. However, we do not know what became of this formidable construction. Based on Mármol’s description, it may have still been in operation at the beginning of the Saadid period (sixteenth century), although possibly by then the almunia was already in decline and its structures were starting to decay (Mármol, 1573, fol. 93).

Lastly, we want to comment on the name given to the estate in written sources. We opted to use the form “Ǧannat al-Muṣāra,” the most widely used denomination by the Marinid authors that knew it, such as Ibn Abī Zarʿ or Ibn al-Ḫaṭīb, although it is sometimes referred to as “Rawḍ al-Muṣāra.” However, this name is an old Arabic term derived from the root “mṣr” and the form “maṣara” which, according to the al-Muʿǧam al-Wasīṭ dictionary, means something like “train a horse to show its running skills.” Furthermore, this work includes the term “al-Muṣāra” as “place where horses train to race” (Al-Muʿǧam al-Wasīṭ, p. 873), while Dozy indicates a similar meaning, as an equestrian space or stadium (Dozy, 1881, I, p. 597). Other dictionaries offer the same reading (Ibn Manẓūr, 1984, V, p. 175; Fayrūzābādī, 2005, p. 475). Thus, the meaning would not have anything to do with the definition offered by Bressolette and Delarozière as “marvel” or “place for rejoicing” (Bressolette and Delarozière, 1978–79, p. 52).

All things considered, we are lucky to have the work by Torres Balbás that analyzes this term in an illustrative way for various cases in al-Andalus (Torres Balbás, 1959, pp. 425–433). This architect reflects on the characteristics of large esplanades, which may have been versatile, multipurpose spaces; first, for leisure, military uses, playing, racing or training. Second, the area was also suited for military marches, parades and reviews, like the documented example of the muṣāra of Córdoba that stretched along the southwest of the city, next to the road that lined the river.

4 The Estate and Its Archaeological Remains

4.2 Ensemble 1

As mentioned in the Introduction, thanks to Bressolette and Delarozière, we know that the Bāb al-Sākima cemetery hosts the remains of at least two ensembles each of which include their own basin and a pavilion presiding over it in the form of a manzah (a pavilion that overlooks the surrounding landscape). However, time and weather have deteriorated these structures significantly, a situation that was terribly aggravated by the fact that the area became a burial site in an undetermined time. Graves and, in some cases, family pantheons have been placed amongst the archaeological remains, hampering any type of archaeological activity.

The first pavilion is on the highest point of the current cemetery; that is, close to the north boundary and slightly towards the west. It is a basin whose inner side was 22 m long, of uncertain depth and demarcated by 2 m thick walls over which lay a narrow gutter that discharged its contents via the middle point of each side, inside the basin (Bressolette and Delarozière, 1978–79, p. 59). Since the basin is on a sloped area, building it required some leveling work, thus creating a particular relationship with the topography (Figs. 3 and 4). As a result, the top of the northern front approximately matches the level of the surrounding ground, while the southern front is 1.5 meters above it. However, the difference between levels may once have been greater, as the area may have been filled with deposits.

View Full Size

Ǧannat al-Muṣāra. Ensemble 1. Plan

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author, based on the map published by Bressolette and Delarozière in 1978–79 (plan II) with adjustments after own survey

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Ǧannat al-Muṣāra. Ensemble 1. Remains of the basin

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

Photograph by the author

• Download Figure
• Download figure as PowerPoint slide

The southern side moreover, where water pressure was greater, presents a further containment structure that could have provided additional wall strength. This is a bit striking, since despite the pressure on this side, the size of the structure does not demand such a solution. In any case, for some reason, an additional 0.73 m wide wall with eight narrow 4.4 m thick buttresses was erected. Their excessive length and their distribution in an irregular sequence may indicate that they had other uses, such as stairs to the edge of the basin or open rooms like iwānāt (sing. iwān) facing the orchard. In general, the basin and buttresses were built with lime-rich rammed mortar and some parts of them still feature the remains of white plastering (Fig. 5).

View Full Size

Ǧannat al-Muṣāra. Ensemble 1. View of lower buttresses

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

The edge of the basin was wide enough to walk along it, something that seems totally appropriate considering that it served not only as a place to store water for the adjoining crops but as a recreational area. In fact, the basin is not an isolated element, as a sophisticated pavilion sat next to the north side of it and overlooked both the reservoir and the connected crops, with a panoramic view of New Fez in the background. This construction is in a terrible state of repair and it is very difficult to delineate its layout. Based on the remains and the plan published by Delarozière and Bressolette, we know it had a maximum length of 14 m and was 13.5 m wide, and it is possible to identify a partitioned structure within. On one side, there is a main, T-shaped space that may have been composed of a large, three-span portico opening to the basin in the forefront and a small adjoining alcove at the back. On the other, we find three ancillary chambers of various sizes around the alcove, two of which can be accessed from the portico itself. The third room seems to have complemented the pavilion’s quadrangular plan, although its outer walls are badly deteriorated. Regarding the walls of the pavilion, today it is only possible to track lime mortar beds or foundations where brick walls once stood. As is usual in historical constructions of the Islamic West, bricks are always an easily looted material, such that sometimes only the parts built with rammed earth remain.

To sum up, we find ourselves before a model that is not foreign to the palatial architecture of the Islamic West, particularly during the thirteenth–fourteenth century period. If we look at Nasrid Granada, we can find the main example of this model in the Partal Bajo palace built by the Sultan Muḥammad III (r. 1302–9).⁹ However, we must note that the landscape context in both cases is very different. The Nasrid example is perched on the side of the wall, so its portico opens up to the basin and gardens, while its qubba tower overlooks the city in the opposite direction. By contrast, the Marinid instance is designed to connect with the landscape in just one direction, where all the components are integrated: the basin and gardens in the foreground, the road to Meknés on a second level, and, finally, the capital city in the background.

4.3 Ensemble 2

The second pavilion found in the almunia is to the east of the first one and is also equipped with a basin; both structures are larger than those in the first ensemble (Fig. 6). Regarding its location, from a topographical perspective, it lies on a small hill that is flatter than the one where the first pavilion was built, so the remains are in a poorer state of repair, which severely limited their documentation. A few parts can still be recognized thanks to small fragments of lime mortar that can be associated with the remains of walls. This is why we must rely heavily on the data provided by Bressolette and Delarozière (Bressolette and Delarozière, 1978–79, p. 60).

View Full Size

Ǧannat al-Muṣāra. Ensemble 2. Plan

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author: based on the map published by Bressolette and Delarozière in 1978–79 (plan III) with adjustments

• Download Figure
• Download figure as PowerPoint slide

Firstly, the interior measurements of the basin are approximately 39 × 44 m, with 1.5 m wide walls except to the south, where the wall is 2 m wide. However, this ensemble was designed in a unique way, as a platform centered on the symmetry axis is projected 7 meters into the reservoir over the north side, serving as the basis for part of a pavilion that overlooks the basin and the adjacent gardens. The pavilion whose remains emerge from this front was a 19.6 m wide, tripartite structure with a central space noticeably larger than the sides, which are narrow and appear to be transitional spaces for accessing the ancillary rooms in the back. This architectural solution seems to seek a visual effect, as if the pavilion was floating, a sophisticated design with no known parallel throughout al-Andalus and al-Maġrib, where basins tend to follow a regular shape (for example, the upper and lower gardens in Madīnat al-Zahrāʾ, the Agdāl and Menara gardens in Marrakech, or the Nasrid palaces in the Alhambra). The only similar example is found in the old Abu Fihr estate in Tunis, where the remains of a large rectangular basin and a pavilion presiding over its northwest side are partially preserved. In this case the qubba is projected on a platform towards the interior of the basin, while behind it there is a set of smaller rooms (Soulignac, 1936; Hadda, 2018).

This said, the ensemble is not limited to the structures described, but rather were included in a larger compound surrounded by an enclosing wall. This 0.48–0.78 m wide wall, now in a very poor state of repair, enclosed the compound 22 m away from the walls of the basin. The separation distance on the northern side is unknown, and it may have corresponded to the side of the pavilion. This would mean there was a perimeter wall protecting the ensemble, including the free interior space, that may have been landscaped. This layout preserved the interior of the ensemble as a maximum privacy area, and probably more select species were grown here, away from the rest of the estate where crops were likely less refined and more exposed (Fig. 7). It is therefore a notable part of the almunia that already appears in Andalusi agricultural treatises (Akef and Almela, 2021, pp. 12–14). As for irrigation of this area and the surrounding land, it seems two canals on the southern side of the basin hauled water westward (Bressolette and Delarozière, 1978–79, p. 60).

View Full Size

Ǧannat al-Muṣāra. Ensemble 2. Hypothetical restitution of the pavilion with the surrounding landscape

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

Photograph by Lucía Martínez Bernal, Almenara Blanca

• Download Figure
• Download figure as PowerPoint slide

4.4 Tower-Pavilion

Unlike the structures presented above, the third known element inside the estate has never been documented before. This is a tower attached to Ensemble 2; more specifically, to the west side of the southern closing wall (Fig. 8). Regarding the stratigraphic sequence of both structures, it is evident that the tower was attached to the wall, and they were probably not part of the same project. For one thing, the composition of the wall is significantly different. The tower was made with lime-rich rammed mortar, which is a more durable mix, while the wall is much earthier. This explains why the perimeter wall of Ensemble 2 deteriorated more severely, revealing how the wall of the tower rested partly on it (Fig. 9). Furthermore, this showed the potential height of the perimeter wall of Ensemble 2, which could have reached 4.5 m over the most prevalent ground level.

View Full Size

Ǧannat al-Muṣāra. Tower-Pavilion. Plan, elevations and sections

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author and L. Martínez Bernal

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Ǧannat al-Muṣāra. Tower-Pavilion. View of the northern side where the tower was attached to the former enclosure of Ensemble 2

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author and L. Martínez Bernal

• Download Figure
• Download figure as PowerPoint slide

Turning to the tower-pavilion, it was an 11.2 × 10.5 m square structure, solidly built with lime-mortar rammed earth walls on its four sides, although only a limited portion of the southern side remains (Fig. 10). The walls have been preserved to a height of 5.9 m across the entire construction; that is, up to the top line of a layer of rammed earth. As to thickness, the walls are 0.8 m wide except for the northern wall, measuring just 0.6 m, which can be explained by the fact that it is attached to the wall of Ensemble 2.

View Full Size

Ǧannat al-Muṣāra. Tower-Pavilion. View from the south

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

As for the internal arrangement, very few elements remain, as the ground has been filled and contains several burials. However, a few details can still be made out, particularly on the side walls, where traces of three parallel built-in brick arcades remain (Fig. 11). Based on the thinness of the arch and the clear span of the tower (8.9 m), together with the curvature of the imposts that remain, it seems unfeasible that every arcade would consist of one single arch crossing the tower’s width; more likely is an arcade with at least two intermediate supports. It is impossible to establish this with certainty without exploring the underground, which is not possible at this time due to the use of the grounds as a cemetery. Also, the southern side may have had a portico opening to the south, which would explain why only the sides were preserved.

View Full Size

Ǧannat al-Muṣāra. Tower-Pavilion. Inside view of the missing arcades and their remains on the wall

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

This said, the use of such a solid structure with such close arcades (1.5 m) points to the possibility that there was a second level that housed the structure’s main use. Therefore, it could have been a tower-pavilion from which to enjoy a view of the whole estate, particularly to the south, as this is the only side that is not blind, and that is where the openings would be found. Although it is hard to know the level of the second floor, we believe it may have corresponded to the horizontal groove preserved in the north and south sides, where traces of a wood ceiling with square joists can be discerned. In fact, the level of this wood ceiling would match the height of the impost of the lower arches. Thus, the upper level may have served as a rooftop with a 1.37 m parapet; or, if it was a closed, roofed space it would have had brick walls that have since been lost and despoiled. Regarding the vertical connection, no traces of a staircase have been found on the walls; however, an upper level must have existed, since the perimeter walls are especially thick and there is a structure of internal arches. A single-level hall of these dimensions would be covered with a simple roof without the need for supports. Only an archaeological exploration of the ground could prove this feature.

So, given these characteristics and based on the testimony of Ibn al-Ḫaṭīb, it seems reasonable to wonder whether this structure could be the Qubbat al-ʿArḍ (Ibn al-Ḫaṭīb, 2003, II, p. 13; Ibn al-Ḫaṭīb, 1985, p. 184). Although for the time being it is difficult to answer this question, it is interesting to at least suggest the possibility. It is an elevated pavilion, next to the ancient road to Meknès (100 m) and open to it on its two levels. Yet it does not seem as though the tower bordered the exterior of the estate directly, as there is a large opening precisely on its southern side, meaning it would be inside the enclosure. To sum up, we would be speaking of an estate composed of different enclosures and with different levels of accessibility.

5 The Waterwheel and the Aqueduct

Although the estate is 250 m away from the Fez River and from the city, its dependence on the water course and its connection to the court scene linked it directly to New Fez. More precisely, the most noticeable physical contact point is congruent with its water supply system, without which the estate would not have been possible. In this way, on the river bank and facing the city walls, a water elevation system was built that lifted water 20 m above the level of the river; that is, to the highest point of the estate, where it was transported through an aqueduct. The dimensions of this structure are quite substantial, to the point that it became one of the most remarkable highlights in Fez during the Marinid period. However, the passage of time and abandonment of the site turned it into a colossal ruin that remains almost intact today.

Regarding the urban context into which it was incorporated, we should mention that the noria was built, initially isolated, next to the primitive bridge that crossed the Fez River from Bāb al-Qanṭara/Bāb al-Wādī, connecting with the road to Meknès. This location was later transformed when the advanced Bāb al-Sabʿ gate and the plaza-bridge were constructed. In fact, the structures mentioned make up a heritage ensemble that is very valuable to the city and offer great potential for performing a more profound stratigraphic analysis than what has traditionally been carried out in this urban environment (Fig. 12). The ensemble includes remarkably significant structures that played a crucial role in some of the most prominent phases of its Medieval (gates and hydraulic structures) and Contemporary (watermills and nineteenth-century arms factory) periods.

View Full Size

New Fez. General plan of the noria and the aqueduct. At the south end is Bāb al-Qanṭara and at the north end, Bāb al-Makīna/Bāb al-Saʿāda and Bāb al-Sākima.

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

With regard to the hydraulic structures that are the focus of this section, the ensemble includes two very different elements. On the one hand we have the delicate noria, comprising both a solid bearing support, which has been preserved, and a somewhat fragile hydraulic mechanism that required constant maintenance. On the other hand, we have the far sturdier aqueduct, a costly and ambitious project.

The most characteristic element of the noria was the wooden wheel that was turned by the flowing water, which has since disappeared. What remains is a solid structure that supported and guided the wheel (Figs. 13 and 14). This construction is a rectangular box (27.2 × 7.1 m) defined mainly by its two dominant walls, made of high-quality lime mortar formwork that was later plastered. Bricks were also used, in edges, openings and the crown.

View Full Size

New Fez. Noria. View from the river

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

View Full Size

New Fez. Noria and aqueduct. Plan at different levels

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

To be able to house the wheel, the two walls are arranged with a 2-meter channel between them for water to flow, which explains why the short sides are nothing but perpendicular arched bracings (Fig. 15). Its current height is 11 m over the current riverbed, but this might have changed due to the accumulation of sediments and alterations in recent centuries. Indeed, it has been estimated that the walls may continue 3.5 meters below the current ground level.

View Full Size

New Fez. Noria. Inside view of the noria’s canal

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

Regarding the southern wall, the façade that opens to the river is 2.7 m thick and is composed of a series of blind and full openings that lighten the top of the structure (Fig. 16). These openings are, however, closer to the sides, such that the central part remains robust enough to support the wheel. Some of the formwork was erected in a phased way, composing a load triangle that transmits the weight of the wheel to a larger support base. The arched openings inside it come in pairs, where one is 1.5 m deep and blind, and the other is complete. The easternmost couple is two rows further down and has almost fully lost one of the rings and the corner buttress. Lastly, there seems to have been a small corbel or projection in the center of the top element of the southern elevation. It is evident that this element already existed when the wooden wheel worked, since below it the wall did not show the typical lime accretions generated by the water spilled on the rest of the wall surface.

View Full Size

New Fez. Noria. South elevation

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

The north wall features a different layout. It is 2.4 m wide and the openings in the upper levels consist of large 1.15 m deep arches preceding smaller openings that are a projection of those in an equivalent location on the southern wall. Furthermore, this north wall is totally different from the southern one, as it houses the connection with the aqueduct, explaining why this wall is four courses higher in its northern edge with a 0.7 m thick wall, and connects diagonally to a central solid tower projected to the north (Fig. 17). Indeed, this solid tower is the start of the aqueduct, and continues rising 4.7 m, thus reaching the same height as the aqueduct. Beyond the crown of the box, the tower is no longer solid and accommodates a vertical U-shaped chamber open to the south (Fig. 18).

View Full Size

New Fez. Noria and aqueduct. View of the northern tower attached to the noria. In the upper part: brick turret

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

View Full Size

New Fez. Noria and aqueduct. Central northern tower with preserved holes for wooden beams

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

The crown of the box presents a brick platform that seems to have housed a series of supports and compartments. On the southern wall, this crown is badly deteriorated and encroached upon by vegetation, which complicates its analysis, but in the north wall a small corridor and a sequence of eight pillars can still be made out (Fig. 19). Moreover, parapets or partitions were raised between these supports, except for the central section across from the tower chamber, which featured a thinned mortise in the base of the wall. An analogous mortise can be discerned on the southern wall, although considerably altered by the industrial adaptation during the 19th century.

View Full Size

New Fez. Noria. Top view of the noria

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

Lastly, a 2.8 m high brick turret closes one of the sides of the solid tower; it is accessed through a staircase going up from the crown of the aqueduct (Fig. 20). Inside it lies a small catch basin, the depth of which we were unable to determine, although, in connection to this same component, several qādūs (fired clay piping embedded in sturdy brickwork) were found. One runs vertically inside the solid tower and shows marks of having been repaired several times, while the other, as explained below, runs along one of the top parapets of the aqueduct.

View Full Size

New Fez. Noria and aqueduct. View of the first section of the aqueduct from the east

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

Let us move to the aqueduct. It is the most outstanding element of the hydraulic infrastructure and consists of a great, 170 m long wall that connects Bāb al-Sabʿ and Bāb al-Sākima (Fig. 21) and features buttresses and arcades.¹⁰ The north side of it (Bāb al-Sākima) culminates in one of the two large octagonal towers, although it can be presumed that it did not end there originally, but continued to the north to reach the highest part of the property. As mentioned, its course included the two octagonal towers and flanked the road to Meknès; most likely it ran north, following the same orientation that would later be adopted by the Kasbah Cherarda. Also, given that it was erected in an area sloping toward the river, the height of the aqueduct is reduced progressively as it approaches the north, so a difference of up to 5 meters is found between the two remaining sides.

View Full Size

New Fez. Noria and aqueduct. East elevation of the aqueduct

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

Yet, regardless of its robustness and good quality, the work does not follow a straight line, but is composed of two parts following different directions and separated by a turning point that is quite striking, as it marks a change of pattern that brings about several divergences. The first part starts at the noria and runs northward, is 1.4 m thick, and contains seven sections separated by six buttresses. Generally, each of the sections has a great arch that does not seem to reach the ground, as it always rests on the same mortar row that, in some points, shows continuity. All of them are pointed horseshoe arches with alfiz and voussoir lintels, all made with brickwork that is integrate with slight serration within the mortar rows. The most widely used technique in the aqueduct is indeed rammed earth made with good quality lime mortar, although the ratio of lime used is lower than in the noria, and a percentage of earth can be found in the mix. Exceptionally, this first section of the aqueduct does not present any arch similar to those previously described, and the stretch is solid, although it is possible to discern a rather noticeably tiered joint (Fig. 22). In any case, this does not seem to indicate a major change of phase, but rather two steps in the construction process, as the composition of the wall and the continuity of the rows remains unchanged.

View Full Size

New Fez. Noria and aqueduct. View of the first section of the aqueduct from the west

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

View Full Size

New Fez. Aqueduct. Top view of the channel towards Bāb al-Sākima

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

As for the upper canal, it is on the top between two side parapets made of brickwork and mortar (Fig. 23). This canal becomes slightly wider at its southern end, precisely next to the solid tower attached to the noria, where it veers toward its western arm. Additionally, over the western canal parapet, the level was raised with bricks solidly laid with mortar containing the remains of a fired clay pipe (qādūs). This alternative way of channeling water is directly related to the brick turret over the central solid tower attached to the noria which, thanks to the catch basin, created a piezometer that generated a flow of pressurized water. This type of contraption is well known in Morocco, and often appears in connection with fountains or supply points in palaces or emblematic buildings, allowing water to be taken to upper levels or even create effects in fountains.¹¹ It is not intended for irrigation purposes as, unlike ditches and channels, this system does not provide an abundant flow (Navarro, Garrido and Almela, 2018, pp. 10–16; Almagro, 2022, pp. 131 and 146–148).

The second part of the aqueduct, after the point of inflexion, has four sections, the last of which is much shorter. The buttresses are similar here, even though several differences regarding the first section can be discerned. On the one hand, the wall thickness is reduced to 1 m at the bottom, although it remains unaltered at the top, making it necessary to use a continuous transition corbel. On the other hand, the buttresses are slightly higher for the purpose of consolidating the top part. Lastly, the openings here are smaller arches.

Delarozière and Bressolette already acknowledged some of these differences and expressed concern about them, even suggesting that they may point to two different construction phases. Their theory was that an initial stage could be found to the north of the point of inflexion, featuring a lower, narrower wall with small arched openings. Then, at a later period, the south section of the aqueduct would be built, with wider, higher walls, large arches and thick buttresses; an intervention that tried to leverage the previous work and raised it 5 meters so that it could be reused. This would explain the fact that the buttresses in this part do not consistently line up with the arches, and why it was necessary to broaden the wall at the top to house the canal (Delarozière and Bressolette, 1939, pp. 631–632).

They also spotted some details that may be significant in analyzing the structure and which, unfortunately, can no longer be seen. One was that, to the north of the point of inflexion, the wall presented a horizontal interruption two thirds of its height from the ground that was close to the point where the thickness changes. There, a small 0.22 × 0.27 m canal within the thickness of the wall was identified. The other is that it seems as though some buttresses were attached to the wall and blinded the arches, which would clearly indicate that they were planned at a later stage. Finally, looking at the octagonal tower that closes the aqueduct to the north, they detected an attachment that was not connected to the tower, a relevant detail considering that the scale of the towers better matches the raised height (Delarozière and Bressolette, 1939, pp. 631–632).

Other than the preserved aqueduct, the only elements that we know today are the two octagonal towers that flanked the road to Meknès and that were later integrated into the late Alawite Bāb al-Sākima – Bāb al-Makīna/Bāb al-Saʿāda project, a set of gates that organized the traffic between Old Fez, New Fez, the arms factory and the road to Meknès (Fig. 24). However, in the twentieth century, the whole area was cleared to build a large road replacing the old one, and only the two Marinid towers remained, isolated, as well as the Bāb al-Saʿāda gate. The section that connected the two towers and probably included a large passage was demolished and the towers were left disconnected.

View Full Size

New Fez. Aqueduct. View of the two octagonal towers from the east. On the left, Bāb al-Makīna/Bāb al-Saʿāda

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

However, traces on the tower walls and some historical photographs reveal that a 2.5 m thick wall used to link them and continued northwards (Fig. 25). In fact, the towers and walls were one single project, as shown by the fact that the formwork was connected. Also, the marks on the southern tower, which seems to have survived almost intact, leave no margin for doubt about its connection with the aqueduct, as it retains a section hole and a height that are similar to those of the aqueduct’s canal. There is evidence pointing to the presence of an alure or wall walk on top of it, probably roofed (Fig. 26). As for the northern tower, it was not preserved in all its height, and the top of it was rebuilt at some point in the twentieth century. However, many details can still be observed, such as the presence of a very thin and smooth layer of lime that was later covered with whitish plaster.

View Full Size

New Fez. Aqueduct. Northern octagonal tower. Protruding fragments of the aqueduct’s demolished section

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

View Full Size

New Fez. Aqueduct. North elevation of the southern octagonal tower

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

Beyond this point the aqueduct is almost imperceptible, as the whole area has been intensely urbanized. Delarozière and Bressolette spotted the remains of a 0.9 m wall 75 meters away, flanked by two buttresses on each side and with an orientation that veered slightly towards the west, in the direction of a rammed earth mass 120 meters away from the northeast corner of the Kasbah Cherarda (Delarozière and Bressolette, 1939, pp. 631–632). They interpreted this vestige as a possible reservoir or supply pond, and then identified the traces of a secondary aqueduct that led to the south and then turned slightly to the west. It appeared that this last section consisted of a rammed earth wall, 1 m thick and 0.2 m high, on a 1.4 m high foundation.

All things considered, although it may initially seem as though the works took place in one single phase and that the typological differences on both sides of the point of inflection could be explained by the decrease in the altitude the water had to reach, we believe that given the details observed on site and the remarks by Delarozière and Bressolette, it is plausible to think that the aqueduct was built in two stages, the second of which would be intended to resize both the flow rate and the scope of the infrastructure. That said, this hypothesis is subject to further studies, since as of today the number of fragments to be analyzed is reduced.

Another testimony that would support Delarozière and Bressolette’s hypothesis are the words of al-Numayrī (fourteenth century) reported above. Even though this evidence is ambiguous to a certain extent, it would accord with this assumption. According to this chronicler, Abū ʿInān had several norias erected in connection with the estate, including a noria with an aqueduct (qanāt) built to replace a previous water wheel to expand its reach and flow rate (Al-Numayrī, 1990, pp. 174–181 and 211–213). He could indeed be referring to the great noria, but this chronology conflicts with the potential chronology that we are considering for Bāb al-Sabʿ, which might have already existed during the first half of the fourteenth century in the times of Abū al-Ḥasan, which was without a doubt built after the noria-aqueduct. In any event, although he might have been referring to a different noria, there is a chance that the celebrated noria commissioned in the thirteenth century was subject to refurbishment later in the Marinid period (fourteenth century). Nevertheless, there is another possibility that would call for deeper analysis, as it may be a redesign that took place during the same construction process, either due to the need to irrigate higher grounds or to errors when calculating the altitude of the basins to be provided. In fact, it was a complex and colossal work, so it was probably subject to checks and modifications during the construction process. An additional fact in favor of this supposition is that the structures do not present irregularities between the two stages, and were integrated through very clean construction solutions.

5.1 Later Stages

It is also possible to establish how the aqueduct structure was a preexisting element for other, later interventions in the area. Of all of them, it seems that the most immediate one was the Bāb al-Sabʿ advanced gate, which used the aqueduct to form its western side and leveraged the southernmost arch in the aqueduct (Fig. 27). This way, the heterogeneous blinding of this arch with two rows of rammed earth, a mixed masonry wall with rows of bricks and a brick crown (Fig. 28) dates from the time when the gate was built, making it necessary to seal the opening to make the interior of the bastion safer.¹² It must be noted, however, that during this stage no activities affecting the functioning of the hydraulic system or evidencing its lack of use were detected. To the contrary, it seems as though the two infrastructures coexisted. What is clear is that the gate was built after the noria. Therefore, if the noria was built between the thirteenth and the fourteenth century, then Bāb al-Sabʿ could be an addition from a later time during the Marinid period, whether it was built immediately after the noria or later.

View Full Size

New Fez. Noria and aqueduct. Elevation of the first two sections of the aqueduct from the west. The first arch was walled up as part of the Bāb al-Sabʿ project.

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

View Full Size

New Fez. Aqueduct. View from the southwest

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

Next, the arms factory (Dār al-Makīna) was built by Ḥasan I in 1888–1890 to the north of New Fez, in a location that favored control of the industry while ensuring its connection with the river, to guarantee the water supply and enable the discharge of wastewater. More precisely, the new factory occupied the grounds to the west of the aqueduct, and used this structure to establish its boundaries.¹³ The other sides are composed of an earthy, whitish rammed earth wall with low rows of lime-pointed formwork.

In the noria’s immediate surroundings and in the first section of the aqueduct, we can also appreciate how the Medieval structures have been deeply transformed as they integrated new industrial facilities. In fact, it is obvious that when this intervention took place in the nineteenth century, the noria was no longer functioning, as the two uses would clearly be incompatible. Firstly, the structures built to the north of the noria, both to the west of the aqueduct and in the grounds between the noria and Bāb al-Sabʿ, reduced the thickness of the Medieval structures to increase the useful surface. Secondly, multiple walls were attached, and grooves were opened to insert their roofing system with arches and sloping roofs. Thirdly, in spite of the narrowness of the noria’s box, it was encroached with rooms for which partitions were erected, arches were blinded, one of the short façades was sealed, walls were thinned to increase the surface, openings were made, and grooves were opened to insert slabs and roofs (Fig. 15).

Additionally, to the south of the noria, a building that may have had an industrial use – although it appears unrelated to the arms factory – was attached at an unspecified time. It was 4 meters wide and was complemented with a bridge attached to the wall of the plaza-bridge; that is, it was connected to the wall’s barbican. As shown by the traces seen in the noria’s elevation, the wall was perforated to create a small cupboard, a deep niche and six small (1 m high) rectangular openings. These last, the depth of which has not been determined, are located below ground level; that is, in a chamber between the river bed and the building’s floor, leading us to believe that it was a modern mill equipped with a set of wheels.¹⁴

5.2 Interpreting Its Functioning

After analyzing, to the extent possible, the evolution of the hydraulic complex and identifying its original features to a certain degree, we can try to discern how the noria worked, which is one of the most fascinating aspects about it. To this end, in addition to the vestiges that remain, there are studies about similar mechanisms that serve as a support to the proposal of some interpretations. Logically the outcome of this endeavor is much more reliable when it comes to the functions of the vestiges that remain, and becomes more hypothetical regarding the lost carpentry mechanism.

First, we must consider the connection with the river and understand that the noria used part of its flow as a source of pressure to make it turn, while some of that water was lifted and diverted to the almunia. This required a minimum volume flowing along the noria’s canal, and to this end, it is very likely that the system included an ancillary structure in the river-bed, such as a diversion dam to retain water and ensure a steady level in the canal.¹⁵ However, the water that was not lifted would return to the river after exiting the noria’s canal, although once the plaza-bridge was built, it did so through one of its vaults. In short, the importance of an ancillary structure in the river lies in the fact that if the volume went down, the wheel could lose the necessary drive. In any case, these ancillary structures remain unknown; they either disappeared or are hidden by sediments.

As explained below, the noria had two dominant walls that were the main bearing support over which the wheel rested and turned. These two walls were highly solid constructions, although the sides were lightened thanks to brick arches. Moreover, thanks to traces left by the wheel on the lime accretions on the inner faces, it is possible to know its diameter (approximately 25 m¹⁶ ), the location of its axis and how wide it was, taking up almost the width of the canal (Fig. 29). As for the axis, it rested on two 2.4 m wide and 0.95 m high mortises on the sides of both walls right below the crown of the box. It can be assumed that they held a concave carved wood bearing on which the shaft laid and turned. Likewise, thanks to the position of the axis, we can establish that the upper half of the wheel was completely visible, while the lower half could be glimpsed through the arches.

View Full Size

New Fez. Noria. Top of the box with remains of pillars, central recess to support the wheel axis, and rotation traces

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

We cannot establish any details about the composition and design of the driving structure, although considering studies about other norias in al-Andalus, we can imagine that it must have been a very hierarchical and orderly assembling.¹⁷ For one thing, the axis must have been equipped with metal or wood stops on the sides to prevent it from moving on the bearing, and its shaft would have been equipped with a series of mortise holes where the interlocked arms were inserted. These arms were the key pieces over which the wheel drum was mounted. Thus, the interlocked arms spanned the wheel’s diameter and were the base for other pieces, like the spokes. Additionally, to fasten the previously mentioned elements, rails making up concentric circumferences were attached, as well as a type of board bracing that partially closed the side of the drum and consolidated the connection with the shaft inside the hub.

Up until that time, both sides of the wheel presented a similar design and left enough space between them to house the blades and the scoops in the hub. However, for a noria of this scope, ceramic scoops do not seem suitable as they would have to be large, thus adding a considerable amount of weight. A solution that seemingly would have been adequate would be using small wooden buckets whose shape and position facilitate them being filled and emptied, thus guaranteeing water spilling only at the highest point of the rotation.¹⁸ This solution would correspond to a certain extent with that described by Luis del Mármol in the sixteenth century as “concavities made of wood” (Mármol, 1573, fol. 93).

Returning to the static structure, we must pay special attention to the system of collecting the water spilled by the wheel at its upper rotation arch. The north wall supported this function and the diversion of water towards the aqueduct. However, considering the remains, it seems as though the water was taken up at more than one level (Fig. 30). First, we have the main level set by the aqueduct that matches the height of the northern tower (7.8 m over the axis). Then, there are three sets of fired clay pipes. Two rest directly on the upper catch basin (10.4 m), although one runs over the aqueduct and the other runs vertically along a shell attached to the inside of the solid tower. The third connects with the aforementioned vertical shell 2.7 m above ground level, although this may be the result of a later repair.

This means that on the top of the north wall of the box we find a set of structures that reveal two highly significant functions. On the one hand, it may have served as maintenance infrastructure and, on the other, it supported the canals that captured the water spilled by the wheel. It includes a thinner rammed earth wall, a parallel structure of ruined pillars, and an intermediate corridor. Based on its layout, it seems to have created two levels of circulation next to the wheel, the highest of which was an architraved system, as shown by the squared openings (0.2 × 0.3 m) found on the elevation of the wall, consistent with the pillars. This higher gallery was continuous, although the central section, corresponding to the solid tower and the axis of the noria, had a slightly more elevated floor supported by a pair of higher beams.

However, the structure presents some limitations to reaching the level of the canal, as the rammed earth wall finishes 3 m over the axis, and is 2 m away from the wheel. Two explanations seem possible for this. In the first of these, this set of structures would have been the bearing for a carpentry ancillary structure that was integrated with the catchment canals. Yet that solution is barely feasible as, considering the sophistication of the rest of the structure, it would be strange to choose an unstable material for the most crucial part of the hydraulic system.

So, the most convincing option would be for the structure that emerges over the northern side of the wall to have continued with an increased height, although this would have only been possible over the line of pillars and not over the whole gallery, as we can barely see any traces of bricks on the top of the rammed earth wall that borders the northern face. Therefore, we might consider a brick archway, slim but solidly braced, on the lower half, utilizing the wooden slabbing that connects it to the rammed earth wall (Figs. 30, 31 and 32). Furthermore, we consider that this archway must have included two short perpendicular sections that were raised over the gallery to connect with the two arms of the solid tower at different levels. Only that could explain why these two arms present a brick mortise and why the whole southern elevation of the catch basin turret is bare, as if part of the structure had collapsed. Also, the traces of wood pieces inserted inside the solid tower seem to indicate much lighter slab systems not related to the support of the pipes, but to two levels of wooden ceilings that would have had a logical function, allowing maintenance operators to access the canals, piping and drains. This way, on the open side of the tower, these wooden ceilings were supported by beams and cross braces, the insertions of which are clearly recognizable on the wall.

View Full Size

New Fez. Noria and aqueduct. Cross section of the noria. Current stay and hypothetical restitution

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

View Full Size

New Fez. Noria and aqueduct. Hypothetical restitution of the noria; south elevation

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

View Full Size

New Fez. Noria and aqueduct. Hypothetical restitution

Citation: Journal of Material Cultures in the Muslim World 5, 1-2 (2024) ; 10.1163/26666286-12340051

By the author

• Download Figure
• Download figure as PowerPoint slide

Indeed the solid tower, which is part of the same project as the aqueduct and the noria, is a key aspect in the system, as it connects catchment and canalization elements. The catchment basin or piezometer may have been a last-minute addition to have pipes with pressurized water. At any rate, this element was already planned when the solid tower was built, as it influenced its design with two specifically sized arms. Accordingly, the western arm is narrower, while the eastern one houses a wider-section canal appropriate for the aqueduct.

Lastly, we do not know the purpose of the fired clay piping running along the interior of the solid tower, but it seems to be a later addition. In fact, it may have been built at the same time as the plaza-bridge so its wall could be used as support for pipelines. This type of solution is not unconventional, and is indeed mentioned by Leo Africanus when referring to the norias in Fez (Leo Africanus, 1956, pp. 232–235; Leo Africanus, 2004, pp. 276–278).

6 Final Remarks

In conclusion, in view of the set of remains presented and analyzed, we can establish that, as was customary in the Islamic West, the Marinid dynasty also undertook the construction of its own almunia for formal uses, as had been done previously by the Umayyads in Cordoba, the Almohads in Marrakesh and Seville, and, contemporaneously, the Nasrids in Granada. In this way, the landscaped space created was directly connected to the palatial city, establishing an architectural dialogue and fulfilling a scenic and ceremonial purpose, as it was the setting for military parades, welcoming retinues, or serving as temporary accommodation for the ruler.

Furthermore, regarding the interior of the estate, although it is very dilapidated, we can recognize that it was equipped with court constructions. Even if we have only come to know two very battered pavilions, these elements are very revealing, as they present a sophisticated design and enable us to define, to some extent, how the almunia was organized, with small autonomous enclosures within it. Unlike the Almohad tradition of establishing one main core area,¹⁹ this planning seems to be more dispersed and multinuclear, with smaller units, although at the same time the Almohad custom of a dominant basin with a palace or pavilion perched on one of the sides is partly maintained. So far, we cannot present anything about how the crops and inner sectors were organized.

Finally, the relevance of Ǧannat al-Muṣāra is made clear by its most significant material evidence: the great noria and the aqueduct. This hydraulic system is unparalleled in the Islamic West. What is more, there are very few examples of monumental hydraulic systems in general, of which the more remarkable is the noria-aqueduct on the Oued Mikkès close to Moulay Idriss (Bazzana, Guichard and Montmessin, 1987). It is indeed not hard to understand why no other similar norias exist, as projects of such a scope could only be undertaken by the state or an agent with enough financial capacity to mobilize the necessary resources.