1This paper presents an overview of the results of the investigations conducted at the site of Meninx (Jerba) in 2017 and 2018. These activities were part of a Tunisian-German project (a cooperation between the Institut National du Patrimoine, Tunis, and the Institut für Klassische Archäologie der Ludwig-Maximilians-Universität München), the aim of which is to shed light on the urban history of the most important city on the island of Jerba in Antiquity (fig. 1).
Fig. 1: Meninx in its North African context
(P. Scheding)
- 1 For a short overview of the topography and the research history of Meninx see Ritter et alii 2018, (...)
- 2 An Island 2009.
2Meninx, situated on the SE shore of the island (fig. 2), was the capital of Jerba during the Roman Empire and eponymous for the island’s name in Antiquity1. The outstanding importance of this seaport derived from the fact that it was one of the main production centres of purple dye in the Mediterranean (Plin., nat, 9, 127). Until recently, the layout of the city was only partially known from a few isolated buildings. Systematic archaeological research began only very late, with a Tunisian-American survey project conducted on the island under the direction of Ali Drine and Elisabeth Fentress from 1996 to 2000, the results of which, relating to the Punic and Roman period, were published in a monograph which is the indispensable basis for anyone studying Meninx2.
Fig. 2: Meninx, aerial photograph
(Deutsches Zentrum für Luft-und Raumfahrt – Zentrum für Satellitengestützte Kriseninformation, April 9, 2018; © DLR/ZKI 2018)
- 3 For the results of the magnetometer survey in 2015 see Ritter et alii 2018.
3Based on the important results of these investigations, we started our project at Meninx in 2015 with a large-scale geophysical prospection which gave us a first insight into the internal urban organization. Within 10 days we surveyed the full strip between the present coastline and the modern coastal road, which encloses the nucleus of the Roman city, from the SW periphery of the forum to the theatre in the northeast. The magnetometry revealed that the central area of the city is crossed by long main roads which run roughly parallel to the coastline, and between them several irregular crossroads and numerous buildings of different size and shape became visible3.
4The new city plan provided us with a reliable topographic frame of reference for extending our explorations by focusing now on the urban history of Meninx.
- 4 These campaigns were conducted in autumn 2017 and autumn 2018 (each for a period of six weeks). Ea (...)
5During two fieldwork campaigns in 2017 and 2018, we conducted various investigations which were closely interrelated4. Firstly we carried out excavations at some selected places. The different groups of finds were studied individually by specialists (ceramics, small finds, coins, architecture, sculpture, mosaics, wall-painting, animal bones, and faunal remains). We then extended our magnetometry survey to the city’s peripheral quarters. And finally, we conducted targeted underwater exploration, primarily to get information on the harbour structures of Meninx.
- 5 The results will be published in a monograph, supplemented by the presentation of the individual s (...)
6We are currently preparing the various results for publication5. In this paper, we give only a short overview of some of the most important results.
- 6 Fentress et alii 2009, p. 133-135.
7In considering where to conduct excavations within the large area surveyed by magnetometry in 2015, our decision was driven by two key objectives. First of all, we wanted to gain exemplary insight into all periods of the city’s long history, from the Hellenistic period until the abandonment of Meninx in the 7th c. A.D. We hence began our excavations in the wider forum area where, according to previous investigations, the earliest settlement was located6. Our second intention was to gain insight into different spheres of urban life at Meninx, which is why we intended to study residential buildings as well as sanctuaries and commercial buildings.
8We conducted excavations at nine different places. Four trenches were placed to the south and west of the forum area (Trenches 1, 2, 3, and 5), three at the edges of the forum square (Trenches 4, 6, and 7), and two at a distance from here to the northeast, towards the theatre (Trenches 8 and 9) (fig. 3).
Fig. 3: Meninx, site plan, with results of the geophysical prospection 2015-2018 and the localisation of the excavation trenches 2017-2018 in red
(J. Fassbinder, A. Weinhuber)
9Trench 1 (10×10 m, excavation in 2017) was located at the northern corner of the macellum, with the intention to gain information on its history and infrastructural connectivity (fig. 3). The large building (c. 60×60 m) presents the structural characteristics of a canonical Roman macellum: a rectangular courtyard with a tholos in the centre and surrounded by porticoes giving access to a series of shops distributed along the exterior walls (fig. 17).
10The stratigraphic exploration showed that the construction of the macellum was part of major building activities that took place in this area in the late 1st or early 2nd c. A.D. At about this time, a large adjacent building was erected to the northeast, probably for storage purposes, and the southwest-northeast running street, which passes the front of both buildings and was one of the main traffic routes of the city, was paved with large, flat stones including a stone-built drainage channel running alongside the macellum and leading towards the sea (fig. 4). Some minor alterations were made in the two buildings in the 3rd c. A.D., such as the construction of a cistern in the northeast building, and the partial rearrangement of the channel crossing the street. Already in the late 4th or early 5th c. A.D., the walls of the macellum as well as of its neighbouring building were despoiled, and the stones removed. From then on, the paved street started to be successively covered by compact sandy layers and remained in use as a provisional route up to the 7th c. A.D.
Fig. 4: Trench 1, towards west
(photo N. Sheldrick)
- 7 For the macella in Roman North Africa and their chronology see De Ruyt 1983, esp. p. 259-263 ; Ham (...)
- 8 The only larger complex is the macellum at Lepcis Magna (c. 70×42 m) which was dedicated already u (...)
- 9 De Ruyt 1983, p. 150-158 with fig. 57 (size c. 58×75 m). The close similarity to the macella at Po (...)
11Most of the c. 20 macella which are known from Roman North Africa so far were built during the Antonine and then the Severan period7, so that the complex at Meninx is quite early. It stands out among the North African macella also due to its enormous size, as well as its strictly square and regular structure8. The best parallel in size and structure is the well-preserved macellum at Puteoli built in the Flavian period9, so that the macellum at Meninx obviously followed a model which had been established in Italy shortly before. Another striking difference to other North African macella is its location within the cityscape: in coastal cities, such as Hippo Regius, Gigthis, or Lepcis Magna, the macellum is placed inland, whereas the spacious food market at Meninx, with its great number of shops, was erected near the coast, in proximity to the harbour facilities (see below).
12Trench 2 (10×10 m, excavation in 2017) was placed west of a high mound southwest of the forum (‘South hill’), which consists of debris from earlier excavations in the forum area (fig. 3). Because the magnetometry showed only some faint, roughly rectangular lines here, we decided to check the building structures beneath. The excavation brought to light a residential building consisting of a courtyard (c. 5.50×6.50 m) and surrounding rooms (fig. 5). The courtyard was covered with an opus signinum floor, and its northeast side was completely occupied by a subterranean cistern accessible from above.
Fig. 5: Trench 2, towards northeast
(photo N. Lamare)
- 10 Such structures could be interpreted as fishermen postholes: they resemble those discovered on the (...)
13In the small room immediately behind the northeast wall of the court, the lack of any preservable flooring allowed us to penetrate the deeper layers, down to the virgin soil. The earliest layers contained a group of seven post-holes and a small ditch, the arrangement of which, however, did not reveal what sort of structure they might have belonged to10.
14Whereas the directly related layers did not contain any finds, a terminus ante quem was provided by the filling layers immediately above, dating to the first half
of the 3rd c. B.C.
15The first solid building activities took place in the middle of the 1st c. B.C. At that time, the massive foundation walls of the courtyard, all of them made of substantial ashlars of local limestone, were erected, as was the well-built cistern covered with carefully cut stone slabs.
16This house maintained its general structure and its function over several centuries. It underwent only some limited changes which consisted of the installation of new floors (such as the opus signinum floor in the courtyard and mosaic floors in some rooms in Augustan times), and in the reconstruction or addition of some interior walls. The latest structural interventions took place in the Byzantine period, and the ceramics from the destruction layers above show that the house was inhabited until the final abandonment of the city in the 7th c. A.D.
- 11 This courtyard house type is especially well attested at Kerkouane, see Fantar 1985, passim.
- 12 See Bullo, Ghedini 2003, p. 85-91 (Gigthis, 1 and 2), p. 193 (Pupput, 3). This comprehensive monog (...)
- 13 Fentress et alii 2009, p. 163-166 ; Aït Kaci et alii 2009, p. 213-217.
17This modest house differs from the well-known peristyle houses which dominate our idea of residential building in Roman North Africa. With its small courtyard surrounded by narrow rooms that have off-centered entrances and communicate with each other rather than with the court, it corresponds to a house type which is rooted in Punic tradition11, and can be found in Roman Imperial times in small and modestly equipped houses where the limited space did not allow for a larger peristyle or spacious representation rooms12. At Meninx, a courtyard house of similar size and structure was partly excavated in the southwest quarters of the Roman city during the Tunisian-American Jerba project in 1998 and 1999 (house ‘Meninx II’)13. This building was erected between 70 and 80 A.D., and it has two cisterns above ground and a slightly larger courtyard (c. 10×10 m), which was transformed into a peristyle at the beginning of the 2nd c. A.D. In contrast to this domicile, which was likewise inhabited up to the Late Antique period, the house in Trench 2 is significantly earlier, has a solid subterranean cistern, and a smaller courtyard, which did not allow for even a small peristyle.
- 14 See https://www.klass-archaeologie.uni-muenchen.de/forschung/d-projekte-laufend/meninx/giz/index.h (...)
18Another domestic building came to light in Trench 5 (27×13 m, excavation in June 2018), situated near the modern coastal route at the place where we started to construct an entrance building for our Archaeological Park14, the foundation of which required us first to investigate the ancient structures beneath (fig. 3). The particular circumstances forced us to stop excavating at a depth of c. 1 m, which is why we only penetrated the Late Antique layers. This rescue excavation revealed a group of six small rooms with simple lime plaster floors, which are arranged in two rows on the north side of a large courtyard. Two rooms each gave access to a large subterranean cistern, and a water basin and a fireplace in one room indicate that it was used as a kitchen in the final phase. The earliest building activities we could verify date to the 3rd c. A.D., and the latest ones to the 7th c. A.D. The size and structural interconnection of the rooms, as well as the remains of their equipment, strongly suggest that they should be interpreted as part of the domestic wing of a larger residential house.
19Trench 3 (10×10 m, excavation in 2017; extended by 3×10 m towards the East, and by 3×5 m towards the South in 2018) was placed immediately northeast of the above-mentioned debris hill (‘South hill’) (fig. 3). Since the magnetometry was not very clear here, we wanted to check the stratigraphic sequence and the character of the ancient structures in this area, situated not far from the forum square.
20Under the modern debris layers and a homogeneous layer of sand beneath, a thick and widespread homogeneous layer of Late Antique debris came to light which covered a quite well-preserved bath complex (fig. 6).
Fig. 6: Trench 3, towards north
(2017, photo c. Lehnert)
21The building consists of a large distribution hall (c. 11.00 m×4.50 m) and a series of small square rooms symmetrically arranged at its long sides, some of which have mosaic pavements with simple geometrical patterns. One of these square rooms was laid out as a pool, accessible by two waterproof steps. A second basin at the northwestern narrow side of the hallway has the shape of a three-quarter circle and is sunken into the ground, presenting a well-preserved water pipe in the rear wall and a drain channel at the bottom. Since the two pools, directly accessible from the hallway, did not show any traces of a heating system, the hallway is obviously to be interpreted as a frigidarium.
22By extending the trench in 2018 to the East and the South (away from the ‘South hill’), three more rooms came to light (fig. 7). Two of them are passage rooms, equipped with small marble basins. The room in the southeast corner of the trench, separated from the hallway by a passage room, has a small water basin in the northeastern corner. Since the walls contained clay pipes belonging to a wall heating, and hypocausts appeared under the floor, this room was a caldarium. Outside these rooms to the northeast, we gained information on the water supply of the bath. A large reservoir is situated high above ground and fed the basin in the square pool through a sloping water channel. In the northeastern corner of the trench, a large and solidly built cistern emerged.
Fig. 7: The bath complex in Trench 3, phase plan
(results of the excavations in 2017 and 2018. Plan c. Lehnert)
- 15 Hewitt 2000 ; De Haan 2010, esp. p. 48-51, 248-275, 339-356 (North Africa).
- 16 Cuicul (Djemila), Maison de l‘Âne: Hewitt 2000, p. 274-275 no. C.11 fig. 59 (plan) ; De Haan 2010, (...)
23The modest dimensions of the complex and its rooms might indicate a domestic bath. Such baths, integrated into an urban dwelling, were quite common in ancient North Africa from the 2nd c. A.D. onwards15. Private baths which are quite similar to the complex at Meninx in their size as well as in the number and the arrangement of their rooms can be found, for example, in the Maison de l‘Âne at Cuicul (also situated in the vicinity of the forum), and in the Maison au Péristyle figuré at Pupput, each of them also with two small pools in the frigidarium, one rectangular and one semi-circular in shape16. The axial arrangement of the main rooms is quite common, but the building at Meninx differs from other private baths in North Africa in the symmetrical arrangement of the smaller rooms flanking the frigidarium, as the largest room, on both long sides.
- 17 See Hewitt 2000, p. 157-189.
24Since we did not want to destroy the mosaic pavements of the rooms, we did not excavate further down into the foundation levels, which is why we do not have any direct stratigraphic evidence for the dating of the building yet. The only hint at a date available so far is provided by several large fragments of wall-painting found in the destruction layers immediately above the floors of the frigidarium and of the northeast room which, given the usual architectonic sequence of apodyterium, frigidarium, tepidarium, and caldarium17, might have been an entry room of the bath. The largest painted fragments show elaborated architectural and figural compositions, the stylistic features of which point to the Severan period or a little later, so that these rooms seem to have received their decoration with wall-paintings and mosaics in the first half of the 3rd c. A.D.
- 18 Hewitt 2000, p. 205-206 ; De Haan 2010, p. 85-86. The statuettes from the bath at Meninx have been (...)
- 19 Tölle-Kastenbein 1986, p. 57-62 (“Typus Hydrophore”).
25We also found remains of the former sculptural decoration of the bath, which is worth mentioning since the evidence for the statuary in Roman private baths is very scarce18. Within the circular pool at the northwest side of the frigidarium, the deepest debris layer directly on the floor revealed six fragmentary statuettes, which, according to our marble analysis, are made of Pentelic marble. All of them were missing their heads. One of the best-preserved pieces is a Peplophore with a wine jug in her lowered right hand, a figural type which goes back to an Early Classical Greek prototype, and was quite often adopted in Imperial Roman times for small sculptures used to decorate rich houses19 (fig. 8). Another quite well-preserved statuette shows Jupiter seated on a throne, and two other statuettes are prepared as fountain figures by having a water channel drilled through their lower body. All these sculptures date stylistically to the middle of the 2nd c. A.D. but the heterogeneous mixture of figures, including Jupiter, makes it unlikely that they were created together for a common visual context. They were obviously later brought together from different places to decorate the bath.
Fig. 8: Statuette of a Peplophore, from the bath in Trench 3
(photo B Schumann)
26Since these statuettes were found without their heads in the deepest debris layer in the pool, it seems that they were deliberately deposited there at some time in the course of the 4th c. A.D., as indicated by the ceramics from this layer. The bath underwent a major remodelling at that time before it finally collapsed in the 5th c. A.D.
- 20 For the state of research at the end of the Tunisian-American project in 2000 see Th. Morton in Fe (...)
27A core issue concerning the urban centre of Meninx is the size and architectural design of the forum square20. Since the magnetometry data in the wider forum area are heavily disturbed by débris from undocumented earlier excavations, we conducted stratigraphic excavations at some selected places: in- and outside the basilica (Trench 6), at the north corner of the forum square (Trench 4), and at its western periphery (Trench 7) (fig. 3).
- 21 These investigations were conducted by a team from the University of Tübingen in the framework of (...)
- 22 The basilica was studied by Linda Stoeßel who slightly modified the reconstruction of the basilica (...)
- 23 Morton in Fentress et alii 2009, p. 144.
28The starting point for exploring the forum is the basilica at the southeast edge of the square because it is the only building that had already been unearthed by previous excavations. The basilica was carefully restudied by cleaning the visible remains, and then conducting some targeted excavations (Trench 6) (fig. 9)21. The building (c. 48×24,70 m) had a three-aisled, two-storey interior colonnade (11×4 columns) and a square exedra on the northeast side.22 The stylistic analysis of the architectural fragments revealed that the building was erected in the middle of the 2nd c. A.D. The basilica of Meninx stands out by virtue of its size as well as the extraordinary use of imported marble in the interior decoration, and was, as Th. Morton had already put it, “one of the largest and most elaborate in N Africa”.23
Fig. 9: Plan of the forum, magnetogram overlaid by the phase plans of Trenches 3, 4, 6, 7, and 13
The dashed lines mark the presumed contours of the forum square, the ‘North temple’ and the presumed temple at the southwestern side of the forum
(K. Wolf)
29The excavations outside the basilica revealed some modest remains of the two porticoes which bordered the forum square at the southeast and northeast sides. The southeast portico, attached to the front of basilica, has a depth of only 3.80 m, whereas the northeast portico is 6.30 m deep.
- 24 Eingartner 2005, p. 132-137.
30To check the size of the forum square, we placed a trench (Trench 4) at its presumed north corner (fig. 3). Opposite the basilica and parallel to it, the magnetometry had shown a square sanctuary (c. 30×25 m), which obviously marked the northwestern edge of the square. It consists of an enclosed courtyard and a small temple set against its rear wall, thus corresponding to the “templum cum porticibus” type which was very common in Roman North Africa, and can be found in several variations, for example, in the four temples surrounding the forum at Sabratha24. To explore the building situation, we excavated a greater area around the east corner of the sanctuary (10×10 m, excavation in 2017), and then extended it towards the forum (by 4×10 m towards the South, and by 32 sq m towards the South and West, extensions in 2018), also in the hope of finding a suitable place for exploring the earliest settlement layers (fig. 9).
31The excavation in 2017 revealed that the ‘North temple’, as well as the adjoining northeast portico of the forum square, were erected in the Augustan period, before later being redecorated with marble (fig. 10). In the northeastern part of the trench, beyond these buildings, moderate and small-scale remains of domestic housing came to light, which testify to intensive building activities starting already from the 3rd c. B.C. onwards.
Fig. 10: Trench 4, overview, towards east
(2017, photo R. Peitsch)
- 25 This dating was confirmed also by the radiocarbon dating of some botanical samples.
- 26 Cf. Fentress 2009, p. 75-76.
32By extending this trench towards the forum in 2018, we reached an area of 1×2 m where no traces of the forum plaster were preserved, which allowed us to excavate down to the deepest layers (fig. 11). The earliest structures, directly above the virgin soil, are two successive walking levels of beaten earth, the beddings of which provided a lot of ceramic fragments dating back to the 4th c. B.C.25. The deepest bedding layer is a thick package of crushed murex shells, and the reuse of this waste provides the earliest evidence for the start of the production of purple dye at Meninx. Above all, these layers are the earliest stratigraphic evidence so far for human activities at Meninx, and confirm the assumption that such activities did not start until the 4th c. B.C.26.
Fig. 11: Sondage in Trench 4, towards south
(2018, photo R Arndt)
33By extending the trench towards the west, we expected to find a portico edging the forum at its northwestern long side, equivalent to the portico attached to the basilica at the opposite side of the square. It turned out, however, that the forum had no colonnade here. What we found instead was the pavement of a forecourt of the ‘North temple’, which, slightly higher than the forum, was directly accessible from the square.
34All these pavements - of the forum square, its northeast portico, and the ‘North temple’ - date to the first half of the 2nd c. A.D. From the early 2nd c. A.D. onwards, the forum area was receiving a new and splendid appearance, not only by installing new pavements but also by decorating the pre-existing buildings with marble, and this process of luxurious remodelling then culminated in the erection of the basilica in the middle of the century.
- 27 Later we added two smaller trenches (Trench 13, south: 2×6 m ; north: 2×5 m) halfway between Trenc (...)
35In order to investigate the northwest and southwest limits of the forum, we placed trench 7 in the presumed western part of the square, and excavated here two long, narrow sections (south section: 2×30 m, plus three small east-west extensions; north section, situated 10 m to the north: 2×10 m; excavation in 2018) (fig. 3)27. We did not find any trace of colonnades framing these sides of the square. The excavations instead revealed a complex cross-section of Meninx’s urban core beginning in the 1st c. A.D., with multiple walls, pavements, and also streets (fig. 9).
36In the central and northern parts of the south section of the trench, we found several small rooms testifying to a sequence of modest houses, all of which are oriented northwest-southeast, separated by some streets running towards the forum. One of these houses turned out to have been abandoned already around 230-240 A.D., which is surprisingly early considering the proximity to the forum, and the fact that all other buildings persisted well into Late Antiquity.
- 28 Among the numerous architectural fragments scattered across the wider forum area, the best candida (...)
- 29 Th. Morton (in Fentress et alii 2009, p. 137) had suggested a size of c. 36×70 m, based on the dat (...)
- 30 See Ben Baaziz 1987. The forum at Meninx is even slightly smaller than the forum at nearby Gigthis(...)
37Of particular interest concerning the size of the forum square are two massive ashlar walls in the very south of Trench 7. The thicker wall (1.30 m wide) runs southwest-northeast towards the forum, and is met by the smaller one (86 cm wide) at its southeast side at a 90-degree angle (fig. 12). A wide street paved with large stone slabs runs along the exterior side of the thicker wall so that the two of them form the western corner of a monumental building oriented towards the forum, parallel to the basilica and the ‘North temple’. The unusually massive walls, sitting on top of solid earlier structures, are most likely the foundation of a monumental temple which dominated the southwestern side of the forum28. The material from the building layers dates to the first half of the 2nd c. A.D. The building itself, as well as the adjoining street, clearly indicate that the forum cannot have been extended further to the southwest, and provide the first substantial evidence for defining the size of the square (fig. 9). They show that the open space of the forum (without the porticoes at its northeast and southeast sides), with a width of c. 36 m, can hardly have been longer than c. 55 m. This means that the forum of Meninx was smaller than previously assumed29, and of modest size compared to other North African fora30.
Fig. 12: Wall of a monumental temple at the southeastern edge of the forum, Trench 7, towards north
(photo N Sheldrick)
- 31 For a summary see Eingartner 2005, p. 132-137 (with further literature).
38The forum at Meninx (fig. 9) has remarkable similarities to that at Sabratha, the nearest important coastal city of Tripolitania. The forum of Sabratha started to take its architectural shape in the Augustan period and then underwent a long process of monumentalization, which was completed in the late 2nd c. A.D.31.
- 32 For the basilica at Sabratha see Kenrick 1986, p. 68-87 with figs. 28 (plan of Period I: c. A.D. 7 (...)
- 33 For the forum at Sabratha and its buildings see Kenrick 1986, p. 7-117. The open area of the forum (...)
39In both cities, the forum square is oriented southwest-northeast. Each of them is surrounded by several temples, and the basilica is situated along the southeast side of the square. The basilica at Meninx has almost the same dimensions as the basilica at Sabratha, which was erected already in the Flavian period32. Even the two forum squares, as indicated by our excavation results, had nearly the same width, and probably roughly the same length33. And in the same central position at the southwestern side of the square, where at Meninx the foundations of a monumental temple are situated, the forum at Sabratha has the Capitolium dominating the square.
- 34 Kenrick 1986, p. 24-29.
40The architectural design of the forum, however, is quite different in both cities. The forum at Sabratha is framed by parallel porticoes on both long sides, which were erected only in the late Antonine period34, and the two temples dominating the small sides are axially related to one another. The forum at Meninx has a more irregular design, with porticoes only on the southeastern long side and the northeastern small side, which are also of different depths. These irregularities have probably to be explained by the early formation of the general design of the forum at Meninx. With the construction of the two porticoes and of the ‘North temple’ already in the Augustan period (and maybe with the presence of more early buildings on the other sides, which did not leave enough space for adding colonnades), the forum seems to have received a consistent structure in Early Imperial Roman times, which then served as a frame of reference for the placement of the prestigious marble buildings, which were erected in the 2nd c. A.D., such as the monumental temple in the southwest around 130 A.D., and the basilica a little later, around 150 A.D.
41Two other trenches were placed in the northeastern parts of the area investigated by magnetometry in 2015 (fig. 3).
- 35 The layout is similar to the sanctuary in Trench 4 (see above).
42Trench 8 was placed in the level area towards the north where the magnetometry had revealed in great clarity a large sanctuary oriented towards the sea, consisting of an almost square courtyard enclosed by porticoes and a temple with three cellae, set against the rear wall35, as well as a round structure set precisely in front of the main cella that seemed to be an altar (fig. 17). In order to verify these structures and explore their history, we excavated a long narrow, north-south oriented trench within the courtyard in such a way that we could check the front of the cellae, the supposed altar, as well as the southeastern entrance portico (5×23 m, with an east extension of 4×2 m, excavation in 2018). The excavations, conducted down to the virgin soil in some places, revealed that the earliest building activities date to the middle of the 1st c. A.D. The temple, standing on a podium, was built first, followed then by the altar, and finally the portico including a cistern. All these structures are made of the local limestone. The building underwent several modifications in the 2nd and 3rd c. A.D., before it was abandoned already in the later 4th or 5th c. A.D.
- 36 Cf. Wilson 2009, p. 183-184 (“Cistern complex C1”).
43Trench 9 was placed near the coast, at a distance of c. 120 m southwest of the theatre, to explore the surroundings of a huge complex comprising six long parallel cisterns that had previously been excavated36. We decided to excavate two sections immediately outside this building, and parallel to it (north section: 4.50×2.70 m, south section: 12×5 m, excavation in 2018). The exploration of a passageway that runs between the northwestern front of this building and a smaller one, comprising a group of much smaller cisterns, revealed that the huge cistern complex was built in the later 1st c. A.D. In the southern section, located in front of its exterior wall, a group of several adjoining rooms came to light, which was not connected to the cistern complex. These rooms, which had some sort of economic function, were erected already in the mid-1st c. A.D., and thoroughly despoiled in Late Antiquity, so that neither the different and very disturbed layers nor the findings gave any substantial hint as to their specific function.
44Our excavations have provided us with some quite selective but instructive insights into different aspects of urban life through all periods of the city’s long history.
45In some trenches, we found traces of construction activities already in Punic times, the earliest datable evidence going back to the 4th c. B.C. The early activities seem to have been primarily focused on residential buildings (Trenches 2 and 4), and already involved the production of purple dye.
- 37 Aït Kaci et alii 2009, p. 222-229.
46More substantial building activities took place in the Early Imperial Roman period, and seem to have been related mainly to sanctuaries: in Augustan times, a temple was built at the northern edge of the forum square (Trench 4), and around the middle of the 1st c. A.D. another one was erected in the North of the city (Trench 8). Extensive monumental interventions then took place in the later 1st / early 2nd c. A.D. when investments were made especially in commercial infrastructure on the coastline: to the south of the Forum, the macellum and the previously excavated horrea37 were built, and further northeast a monumental complex with six large cisterns and surrounding commercial buildings was erected (Trench 9). During the Hadrianic and Antonine periods, when many other North African cities likewise experienced a construction boom, the forum area saw a thorough remodelling. First, a huge marble temple was built at the southwestern edge of the square (Trench 7), and then, shortly afterward, the splendid basilica at its southeastern side (Trench 6). At the same time, several existing buildings were redecorated with new pavements and with wall panels made of imported marble. Further building activities are attested for the 3rd c. A.D. when, for example, a domestic bath was decorated with wall paintings, mosaic floors, and marble statuettes (Trench 3).
- 38 Fentress et alii 2009, p. 159-163.
47Remarkably, the building activities in several areas ended already in the late 4th / early 5th c. A.D. At about that time, not only the collapse of the domestic bath complex took place (Trench 3) but also the demolition of several other buildings, such as the macellum (Trench 1) and the temple at the north corner of the forum square (Trench 4). Greater areas within the city seem to have been abandoned in the first half of the 5th c. A.D., in particular in the area around the forum. Nevertheless, the city was inhabited during the 6th and 7th c. A.D. as our excavations have shown in some areas, especially in two habitations (Trenches 2 and 5). Construction activities obviously shifted from the former city centre to the periphery, as is indicated by two Christian basilicas at a certain distance from the city38. The city existed until the late 7th c. when it was finally abandoned and started serving only as a source of building material for the following centuries.
48The excavations carried out during the 2017 and 2018 campaigns have yielded enormous quantities of ceramics whose study allows us to distinguish eleven major contexts.
Ceramic context 1: ca. 350 – 325/300 B.C.
- 39 A very similar rim shape is attested in the shipwreck of Tagomago 2, dated to the second or early (...)
- 40 The cargo of many ships which were loaded in the late 5th and during the 4th c. B.C. contained pit (...)
- 41 Bechtold 2010, p. 10 ; Ben Tahar 2016-2017, p. 26-27.
- 42 This hypothesis was put forward by one of the authors (SBT) for the interpretation of the material (...)
49This very first phase starts around the middle of the 4th c. B.C. This level has been brought to light in Trenches 2 and 4. The material consists mainly of imported ceramics, namely amphorae from Carthage of type Ramón T-4.2.1.539 (fig. 13, 1), from the Ionio-Adriatic area, the so-called Corinthian B of type Gassner Randform 5 (fig. 13, 2), from eastern Sicily, from Velia, from Reggio Calabria, from Corinth (Corinthian A’), and from Malta type Ramón T- 2.2.1.4 (pithos?)40 (fig. 13, 4). The Fine Ware found consists of Attic Black Glazed Pottery, namely drinking vessels such as skyphoi (fig. 13, 5) and bolsals (fig. 13, 6) which were used in the consumption of wine. Such items, which belong clearly to the “Greek Symposium”41, reflect the desire of the Meningitani to behave like Carthaginians42. Among the rare Attic lamps attested among the material, we can point out the type Howland 25B (fig. 13, 3).
Fig. 13: Ceramics from Meninx, phases I–II
Phase I.1: Amphora of type Ramón T-4.2.1.5 (1); “Corinthian B” amphora type Gassner Randform 5 (2); Lamp type Howland 25B (3); Maltese amphora type Ramón T-2.2.1.4 (4); Attic skyphos (5); Attic bolsal (6); handmade bowl (7); Handmade pan (8). – Phase I.2: Outturned rim bowl (9); Carthaginian Black Glazed bowl (10); Rilled Rim plate (11); Bowl of type Vegas 10 (12); Atelier des petites estampilles: Lamboglia 27b (13); Punic amphora of type Sabratha 7f (14). – Phase I.3: Campanian A: Lamboglia 6 (15). – Phase II: Local amphora of type Sabratha 8A (16); Lid of type Sabratha 118 (17); Local black glazed plate of type Lamboglia 5/7 (18); Campanian B plate of type Lamboglia 5 (19); Campanian C bowl of type Lamboglia 1 (20); Local black glazed bowl of type Lamboglia 1 (21); Byzacena Black Top of type Vegas 69 (22); Byzacena Black Top of type Vegas 68.3 (23).
- 43 Bechtold 1999, p. 144, pl. XXVII, 242 ; Docter et alii 2006, p. 61 and 62, fig. 32.
- 44 Guerrero 1999. The fact that Byzacena Black Top Red Cooking Ware was not imported at Carthage befo (...)
50As far as the Cooking Ware is concerned, these vessels were mainly imported from Carthage during this phase such as chytrai Vegas F. 6743; only some few casseroles were imported from Byzacena, in particular the lopas of type Guerrero II.1.b44. It is worth noting that only handmade pottery was produced locally at that time. Among the most representative forms, one can point to splayed rim bowls, red slipped inside and outside (fig. 13, 7), and pans/tajine (fig. 13, 8).
Ceramic context 2: ca. 300 – 200 B.C.
51The earliest italic potoria, found in Trench 2, were produced by a Roman workshop known as the “atelier des petites estampilles” in the first half of the 3rd c. B.C. Among the most common forms, one can point to Lamboglia 27b (fig. 13, 13).
- 45 This form is attested in Carthage in the context of the LPI from the first half of the 3rd c. B.C. (...)
52In the middle or during the second half of the 3rd c. B.C., the first imports of Carthaginian and Western Sicilian Black Glazed Pottery (Byrsa 401) appeared, which are clearly inspired by an Attic repertoire such as outturned rim bowls (fig. 13, 9)45, hemispherical bowls with plain rim and straight wall (fig. 13, 10) and rilled rim plates (fig. 13, 11). Then there is some Carthaginian coarse ware such as bowls type Vegas 10.1 (fig. 13, 12).
53From the last decades of the 3rd c. B.C. onwards, Meninx imported its first amphorae produced in Tripolitania and Byzacena of types Sabratha 7f (fig. 13, 14) and Ramón 7.2.1.1. The Western Greek amphorae were still imported from south Italy (MGS V).
Ceramic context 3: beginning of the 2nd c. –
mid-2nd c. B.C.
54From the second quarter of the 2nd c. B.C. onward Campanian A is largely attested in Meninx (fig. 13, 15), along with Carthaginian amphorae type Ramón 7.4.3.3 and African Cooking Ware from Byzacena, especially types Vegas 69 and 68.3.
Ceramic context 4: mid-2nd c. B.C. -
end of the 1st c. B.C.
55After the destruction of Carthage in 146 B.C., the first amphorae of local production appear (fig. 13, 16) and were associated with local common Ware (fig. 13, 17), Campanian Black Glazed A (fig. 13, 18), B (fig. 13, 19) and C (fig. 13, 20) Wares and local Black Glazed Pottery, the forms of which are clearly inspired by the Campanian morphological repertoire (fig. 13, 21). Italic ceramics and amphorae, such as Dressel 1A and 1B continue to arrive in large quantities.
56Regarding the Cooking Ware, it is mainly imported from Byzacena (Byzacena Black Top Red Cooking Ware). Among the most common forms, one can point to Vegas F. 69 (fig. 13, 22) and Vegas F. 68.3 (fig. 13, 23).
57The material relating to the following two contexts is particularly abundant in Trench 8 (the temple).
Ceramic context 5: first half of the 1st c. A.D.
(Augustan and Julio-Claudian)
- 46 Hayes 1994, p. 139-141.
58This context is characterised by the steady and increasing arrival of some luxurious Table Ware, such as Italic Sigillata from Central Italy (types Conspectus 12, 18, 19 et 22) and Eastern Sigillata A from northern Syria (types 3 et 4). The production of local ceramics exploded with new forms that combined Punic and Tripolitanian traditions such as Red Slipped plates similar to those of Sabratha (fig. 14, 1)46.
Fig. 14: Ceramics from Meninx, phases III-VII
Phase III.1: Red Slipped plate (1). – Phase III.2: African amphora of type Dressel 2/4 (2); South Gaulish sigillata bowl of type Dragendorff 27B. (3); Eastern Sigillata A Hayes 36 (4); African Cooking Ware CA Hayes 194 (5). – Phase IV: African Cooking Ware CB Hayes 183/Sabratha 59 (6). – Phase V: African Red Slip Ware E Hayes 68 (7); Continental Red Slip Ware Hayes 1972, fig. 58.b (8); Tripolitanian Red Slip Ware Hayes 3 (9). – Phase VI: African Red Slip Ware C/D Sidi Jdidi 3 (10). – Phase VII: African amphora Keay 61 (11); Late Red Slip Ware Hayes 105B (12).
Ceramic context 6: mid-1st c. A.D. –
end of the Flavian period
59Compared with the previous context, this new one is
distinguished by the emergence of the Dressel 2/4 amphorae of local production (fig. 14, 2), the Tripolitanian I oil amphorae, and the first appearance of Cooking Ware imported from Carthage (Hayes 194 (fig. 14, 5), Ostia II 306). The Fine Ware is mainly represented by south Gaulish items (Dragendorf 27B (fig. 14, 3), 15B, 17A) and Eastern Sigillata A (Hayes 37 (fig. 14, 4).
Ceramic context 7: 2nd-3rd c. A.D.
- 47 This variant is very specific to the 3rd c. A.D.: Bonifay 2004, p. 219 ; 227-229.
60The biggest new feature from the most recent excavation campaign in 2018 is the discovery of some contexts relating to the second quarter of the 3rd c. A.D., especially in Trench 7. Among the most characteristic features, one can point out: the African amphoras type II A, Dr. 2/4 late, the African Sigillata C type Hayes 45, the African Cooking Ware type 23B, Hayes 181C, 182B, 197B, 183/variant Sabratha 5947 (fig. 14, 6) and 184. Regarding the lamps, they are represented by Deneauve 8.1.
Ceramic context 8: 4th-5th c. A.D.
General abandonment between the end of the 4th and the first half of the 5th c.
- 48 See Fentress et alii 2009, p. 174-176.
- 49 We are dealing here with one of the characteristic forms of the border area between Algeria and Tu (...)
61This context corresponds to the biggest abandonment level in the ancient city of Meninx, which was called Girba since the second quarter of 3rd c. A.D.48 This context is attested in all the trenches. Among the main forms attested in this context, we can point out: Hayes 68 (ARS E) (fig. 14, 7), Hayes 61B (ARS D), Hayes 1972, fig. 58.b (ARS Z49) (fig. 14, 8), and Hayes 3 (Tripolitanian ARS) (fig. 14, 9).
62As far as the Red Slipped Cooking Ware is concerned, the plate Hayes 181D is the most common find. It is no longer covered with the lid Hayes 182 which was replaced by another lid without slip imported from south Byzacena.
63During this phase, the Pantellerian handmade Ware is well represented, and along with the plates imported from Byzacena, constitutes the bulk of the Cooking Ware attested in Meninx.
Ceramic context 9:
second half of the 5th – first third of the 6th c. A.D.
- 50 This form is well attested in the context of the destruction of 484 at Sidi Jdidi: Mukai 2016, p. (...)
64This context corresponds to the Vandal occupation of Africa. From a ceramic point of view, we have noted some changes both in imported and local ceramics. Compared with the previous context, the new one is distinguished by the emergence of some examples characteristic of the second half of the 5th c. A.D., such as the category of African Red Slip Ware C5 (Hayes 84 and 85) and some productions of the workshop of Pheradi Maius (ARS C/D) located in north Byzacena (Hayes 87B, 88 and Sidi Jdidi 350 (fig. 14, 10). In the same phase, we have a lot of amphoras from Nabeul (Northeastern Tunisia) filled with garum and wine (Keay 35B, 57 and 25.2). We should also note the appearance of new bowls of local production called Meninx 2 with a fabric whose surface is speckled with white spots.
Ceramic context 10: end of the 6th c. A.D.
65In the current state of the documentation, the single well-conserved archaeological layer dating back to the end of the 6th c. A.D. comes from the last redevelopment of the dwelling excavated in Trench 2 in 2017 by Nicolas Lamare. This phase is characterised by the first variants of the last series of the African Red Slip, namely Hayes 87/109 and 105A and some over-molded lamps from central and northern Tunisia.
Ceramic context 11: second half of the 7th c. A.D.
66This context corresponds to the abandonment of the town of Girba. Unfortunately, we currently lack any sealed destruction or abandonment layers with variegated material.
67The latest archaeological layers have yielded ceramics such as African amphora type Keay 61 (fig. 14, 11) and African Red Slip plate type Hayes 105B (fig. 14, 12) dating back to the second half of the 7th c. A.D. with a lot of residual material.
68Apart from the ceramics and the above-mentioned marble statuettes from the bath in Trench 3, our excavations brought to light a great variety of artefacts made of various materials, such as: fragments of marble statues and reliefs; many fragments of wall panellings of imported marble from various regions of the Mediterranean; figurines and antefixes of clay; fragments of white as well as of painted stucco; fragments of mosaic floors; coins; various objects made of different metals, including gold; needles and other objects of bone; and glass vessels. From among these finds, we will here mention only the coins and a remarkable marble head of Antoninus Pius.
- 51 Our coins have been studied by Saskia Kerschbaum.
- 52 Viola 2010, p. 205 no. 194.
- 53 See above.
69Most of the c. 400 coins from the different trenches were recovered in a very corroded state of preservation, caused by the wet and salty soil at the site, but the careful restoration of at least 70 specimens made identification possible51. The earliest legible bronze coins have Tanit on the obverse and a horse in front of a palm tree on the reverse, and were produced in a Carthaginian mint in the later 4th c. B.C. (fig. 15)52; even though all these specimens were found in Late Antique layers (two in Trenches 7, and one in Trench 9), their dating fits well with the other datable material from the very earliest layers from the site53. Several other bronze coins date to the Roman Imperial period and were issued under Hadrian, Marcus Aurelius, Severus Alexander, and Gordian III. The majority of the coins, however, were minted under Constantine and his successors in the 4th c. A.D., and the latest specimen under Theodosius II. As far as the mints are concerned, the coins from the Roman Imperial and the Late Antique periods were produced at mints such as Rome, Siscia, or Lugdunum. An interesting phenomenon is a greater number of counterfeit coins from the later 3rd c. A.D., which imitate original coins of Claudius Gothicus, and many of which were minted in North Africa.
Fig. 15: AE coin, Carthage, 4th c. B.C.
(from Trench 7, photo B. Schumann)
- 54 This portrait head has been studied by Lena Gabler in her Master thesis finished in July 2019.
- 55 Fittschen, Zanker 1994, p. 64 (cat. no. 59) with n. 8-29.
- 56 For portraits of Roman emperors from North Africa see Zanker 1983, p. 30-37.
70Among the marble objects, the most remarkable piece of sculpture is a portrait head of Antoninus Pius (28 cm in height) made of white Greek, probably Parian marble54 (fig. 16). The head was broken off a statue, and came to light in a Late Antique debris layer in the forum area (in the southern part of Trench 7), together with several other fragments of marble sculpture and architecture which had obviously been stored here for later lime-burning. The quite well-preserved head, only partly damaged at its left side, corresponds to the main type of the portraits of Antoninus Pius (type ‘Formia/Croce-Greca 595’), and its stylistic features indicate a date between 138 and 150 A.D.55. Whereas the portraits of Antoninus Pius from North Africa vary widely in detail and quality, and are often quite large in size, the head from Meninx is only slightly larger than life, and stands out due to its particularly high quality as well as the fact that it was produced in a – probably local – workshop which quite accurately copied an urban Roman prototype56. The place of recovery indicates that the statue of the emperor was once displayed somewhere in the forum area, and the untreated rear side shows that it was set up in front of a wall, whether in the basilica, a temple, or one of the porticoes of the forum.
Fig. 16: Marble head of Antoninus Pius
(from Trench 7, photo B. Schumann)
- 57 These studies have been conducted by Joris Peters and Simon Trixl. See Trixl et alii 2020.
71Our archaeozoological studies were focused on a series of closed layers of different dating, to get insight into all periods of urban history57. The analysis of more than 7.000 faunal remains revealed that the proportions of sheep/goat, cattle, and pig are remarkably constant from the 4th c. B.C. to the later 7th c. A.D. The proportion of pigs is significantly lower than it is in other North African cities in Roman Imperial times, but this is simply due to the extraordinarily high proportion of sheep across time. The only plausible explanation for the striking predominance of sheep over the centuries is the prominent role of the production of purple dye which was most probably not exported directly but used to dye wool goods for export so that this industry required extensive pastoralism in the region. Apart from that, more than 50 species of animals have been identified so far, among them various domestic and wild mammals, birds, and – as we might expect in a coastal town – a wide range of fish, molluscs, land and sea snails, the latter ones clearly dominated by innumerable fragments of murex shell (Trunculariopsis trunculus).
- 58 These investigations have been conducted by Michèle Dinies (German Archaeological Institute, Berli (...)
72As for the archaeobotanical studies, 60 samples from different closed contexts were floated for seeds, fruits, and charcoals, again to get a representative diachronic overview58. The cereals are dominated by bread wheat, barley, and emmer, and among the other cultivars, there are several seeds of olive, the cultivation of which was most probably introduced into North Africa by the Phoenicians. Most remarkable is the distinct dominance of olive wood in the charcoal spectra: almost all determinable charcoal fragments are olive wood, and the radiocarbon dating of some samples revealed that the extensive use of olive wood for heating started already in the 4th c. B.C, and lasted over the centuries.
- 59 For the results of the magnetometer survey in 2015 see Ritter et alii 2018.
- 60 The magnetometry was again conducted by Jörg Fassbinder.
73Starting from the results of the magnetometry conducted in 201559, we extended our geophysical prospection to the city’s peripheral quarters: in 2017 to the coastal zone north of the theatre (35 areas of 40×40 m), and in 2018 to the inland area on the northwest side of the modern coastal road (43 areas of 40×40 m)60. We will mention only the most important results (fig. 17).
Fig. 17: Meninx, magnetogram with the results of 2015, 2017 and 2018, and with preliminary results of the archaeological interpretation
(J. Fassbinder, S. Ritter)
- 61 Kleinwächter 2001, esp. p. 335-339. Two different public squares can be found, for example, at Cui (...)
- 62 See above.
74In the area north of the theatre, the main roads coming from the city centre continue heading further northeast, slightly bending towards the coastline to keep running roughly parallel to it. At a distance of some 50-60 m north of the theatre, the magnetometry then revealed a larger area which, framed by two main roads, has no visible building structures. This space might indicate a second public square in addition to the forum, as is the case with several Roman cities in North Africa which have two separate public squares located in different quarters of the city, and connected by main roads61. The existence of a second public square at Meninx, situated a certain distance from the forum, is quite probable, especially considering how far the city extends along the coast62. And finally, it turned out that the row of monumental buildings that occupies the present coastline from the macellum to the theatre continues further to the northeast.
- 63 Wilson 2009, p. 177-181 ; Aït Kaci et alii 2009, 233-235 (“Aqueduct 1”).
- 64 See above (Trench 9).
- 65 Wilson 2009, p. 181, 183.
75In the flat area northwest of the modern coastal road, the magnetometry showed another main road running roughly parallel to the coastline, and several crossroads heading inland, in continuation of the streets coming from the city centre. The magnetometry then revealed the course of an aqueduct, a small part of which is visible above ground northwest of the surveyed area63. This aqueduct heads towards the above-mentioned cistern complex on the foreshore64, which, as A. Wilson had already assumed, obviously formed its terminal reservoir65. The most striking features in this area are some large sections that can easily be distinguished in the magnetometry by their extremely high positive anomalies. These anomalies point to workshops, the activities of which involved heavy firing. This, in turn, might indicate that the production centres of purple dye, a smelly industry which was hence generally located a certain distance from habitation quarters, were situated here, at the northwestern outskirts of the city.
- 66 These investigations were conducted by a team of the Bavarian Society for Underwater Archaeology u (...)
- 67 Ritter et alii 2018, p. 371-372 with fig. 12.
76To supplement the picture provided by the geophysical prospection on land, we started to conduct underwater exploration in 2017 and 2018, mainly in order to answer the question where the port of Meninx that is mentioned in written sources was situated66. These investigations were based on the insight that the only opportunity for larger ships to approach the city was by some broad and deep submarine channels that lead from the open sea into the shallow waters of the Bou Ghrara Gulf (fig. 2)67. The most interesting one is a long channel which, splitting off from the main channel and then turning towards Meninx, runs for a distance of c.2 km parallel to the coast, some 300 to 400 m from the modern coastline.
77Since the sea level, as elsewhere on the Tunisian coast, has risen by between 50 and 75 cm since Antiquity68, the offshore situation at Meninx has so far been unclear. This is why our divers began to explore the area between the modern coastline and the submarine channel next to the city by making use of side-scan sonar, core drilling, and targeted underwater excavation (fig. 18).
Fig. 18: Satellite image with the location of the core drillings off the modern coastline in yellow, and the location of Trenches 10-12
(M. Fiederling, T. Pflederer)
78Systematic core drilling revealed that the ancient shoreline was situated at a distance of between 70 and 80 metres from the modern shore. The ancient coastal strip did not reveal any traces of buildings, so that the dense row of monumental complexes, from the horrea up to the theatre and further northeast, was preceded in Antiquity by a wide beach zone.
79The beach was followed by a shallow water zone that extended down to the submarine channel next to the city. The channel itself is up to 7 metres deep and up to 50 metres wide, and the finds from two excavation trenches at its bottom (Trenches 10 and 11) revealed that it was intensively used for shipping traffic from the Punic period until the Middle Ages.
- 69 Stone 2016, with fig. 2 (plan of the jetties at Ras Segala, Leptiminus, Acholla, Lepcis Magna, and (...)
80The densely placed sediment core drillings in the shallow water zone in front of the macellum exposed a striking concentration of objects at a distance of c. 80 metres off the modern shoreline, i.e. c. 20 metres from the ancient one. An underwater excavation in this area (Trench 12) brought to light a thick layer of debris which – apart from many Roman vessels, charcoal, and botanical remains – contained several massive mortar fragments with traces of demolition. These remains point to a solid jetty built with ashlar blocks, dated by the earliest finds from the debris layer to the Flavian period. As some traces of stone structures on the modern shoreline show, the jetty started from the ancient beach, then extended up to a total length of 80 metres into the shallow water, and terminated in a platform (fig. 19). At its end, the water had a depth of 1,30-1,40 m in Antiquity, which was sufficient for landing Roman cargo ships. This type of “platform jetty” is well known from several ancient ports at the shallow eastern shore of Tunisia, especially from Gigthis and Ras Segala, both situated on the Bou Ghrara Gulf as well69. At another place, we also found traces of a wooden pier which, as radiocarbon dating revealed, dates to the time between 256 and 394 A.D.
Fig. 19: Reconstruction of the coastal zone with the jetty between the macellum (left) and the horrea (right)
(drawing M. Fiederling, T. Bitterer)
81It is now clear how the intensive shipping traffic was handled at Meninx throughout Antiquity: cargo ships coming from the open sea entered the Gulf, approached the city via the deep submarine channel running parallel to the shore at a greater distance, and then made a turn into the shallow water zone to reach the piers leading to the mainland.
82The results of our underwater exploration demonstrate the extent to which the urban layout of Meninx was due to its geographical situation.
- 70 Mattingly 1995, p. 116-122 (Lepcis Magna), p. 125-127 (Sabratha) ; Di Vita, Di Vita Évrard, Bacchi (...)
- 71 See Ritter et alii 2018, p. 369-371 with fig. 12 and 13 (plans of Meninx, Sabratha, and Lepcis Mag (...)
83The particularities of Meninx’ urban layout become evident when compared to Sabratha and Lepcis Magna, the most prominent coastal cities of Tripolitania70. Despite several similarities, Meninx differs from the two other port cities in its overall layout as well as in the distribution of its monumental buildings71. Since Meninx’s immediate hinterland was covered by saline mudflats, unusable for agricultural purposes, the city nowhere extended further than 500 m inland, and instead expanded for more than 1.5 km along the coast. The extension of the city along the coast is reflected in the arrangement of the road system as well as in the placement of a number of monumental buildings: in contrast to other port cities such as Sabratha or Lepcis Magna, the main roads at Meninx do not lead towards the sea but run parallel to the coast, and the theatre as well as the macellum and other economic buildings are not located inland but close to the shore. The axial alignment of several monumental buildings with the sea must have given the city a particularly impressive skyline when viewed from the sea.
84The particularities of the urban layout of Meninx resulted from its exceptional position on a large sheltered Gulf, which had a direct impact on the economic character of the city. The shallow coastal waters provided fish and, above all, murex for the purple dye industries, the most important economic sectors which needed an adequate commercial infrastructure near the coast.
85The new results on the port facilities reveal that the elongated urban layout was first of all determined by the use of the submarine channel next to Meninx, which offered the only possibility for larger boats to approach the city. This channel, which ran parallel to the coast for a great distance, was the main traffic artery of Meninx. The detection of solid jetties and wooden piers, distributed along the ancient coastline, show that this channel was extensively used in its length for handling the intensive shipping traffic.
86The main reason for founding a settlement at the site in the 4th c. B.C., and then for the rapid urban development of Meninx in Roman Imperial times was obviously its favourable geographic position on the Bou Ghrara Gulf. The shallow coastal waters, shielded from the open sea by wide sandbanks and accessible via underwater channels, formed a kind of natural harbour, so that – in contrast to other coastal cities such as Carthage, Sabratha, or Lepcis Magna – there was no need to build an artificial harbour basin. The organisation of the shipping traffic required only that the shallow water zone between the nearby underwater channel and the long coast in front of the city be bridged.
87The exceptional situation of the port facilities which mirror the extension of the city along the coast demonstrates the unusual extent to which Meninx, whose economy was based on purple dye and fish industries, was oriented towards the Mediterranean trade.
- 72 Emporia means the coastal strip between Thaenae and Lepcis Magna.
88Based on the current evidence, the earliest ceramics from our excavations at Meninx do not date before the middle of the 4th c. B.C., which is quite late in comparison with other coastal settlements in the region of the emporia.
- 73 Mattingly 1995, p. 117.
- 74 See Gill 1986, p. 275. This chronology was established on the basis of the study of some Castulo C (...)
- 75 Ben Tahar, Fersi 2009, p. 76, cat. 2, p. 81, fig. 4 (cantharos type “Saint Valentin“) ; p. 76, cat (...)
- 76 These results are based on the material from an archaeological sondage carried out by one of the a (...)
- 77 The early data on Punic Thaenae are limited to a few Attic ceramic fragments of the 4th c. B.C. wh (...)
- 78 Souk el Guébli is located 7 kilometres north of Meninx ; see Ben Tahar 2010, p. 100, p. 98, fig. 4 (...)
89Apart from Lepcis Magna, where the archaeological evidence seems to go back to the 7th c. B.C.73, most of the Libyphoenician emporia were obviously founded in the second half of the 5th c. B.C.: at Sabratha, the earliest ceramics date to around 440 B.C.74, and on the mainland near Jerba, the earliest imported ceramics at Gightis date back to the second half of the 5th c. B.C.75, and at Zitha to the end of the 5th c. B.C.76. For the cities in the northern part of the emporia, such as Thaenae and Tacapae, there is not enough information to determine the date of foundation77. On the island of Jerba itself, the settlements at Souk el Guébli, a site near Meninx and probably part of its territory, and at Guellala, the ancient Haribus, were probably founded in the second half of the 5th c. B.C. as well78.
- 79 Ben Tahar 2016-2017, p. 13, cat. 6 et p. 14, fig. 4, 5.
- 80 Ben Tahar 2016-2017, p. 15, cat. 13 et p. 16, fig. 5, 2.
- 81 Ben Tahar 2016-2017, p. 15, cat. 12 ; p. 16, fig. 5, 1.
- 82 Ben Tahar 2016-2017, p. 16, fig. 5, 3.
- 83 Bechtold 2016-2017, p. 95.
- 84 Ben Tahar 2018, p. 328-330. For the site of Henchir Bourgou and its importance in Antiquity see be (...)
90The only coastal site on Jerba where excavations revealed even earlier material is Ghizène on Jerba’s north coast: imported ceramics from Carthage79, Sardinia80, Western Sicily81, Southern Etruria82, Eastern and Western Greece83 date to the late 6th c. B.C., which corresponds to the Late Archaic / transitional period between the Early Punic and Middle Punic periods at Carthage (c. 530-480 B.C.). The importance of Ghizène is most likely based on the fact that it served as the port of Henchir Bourgou, the most important town of the island in Punic times84.
- 85 Ben Tahar 2016-2017, p. 69.
- 86 For the maritime trade of the Phoenicians, see Schmitz et alii 2007, p. 65.
91These are the indicators of an intensive maritime trade with Greek and Etruscan partners, based on the coastal markets in North Africa from the 6th c. B.C. onwards85. In this context, it is most likely that Ghizène was an early Punic emporion, involved in the long-distance trade conducted by the Phoenicians86.
- 87 For the expansionist policy of the Magonides in the Western Mediterranean and in Africa see Iust. (...)
- 88 See Herodotus 5, 42 ; such a dossier was restudied by some scholars, such as Lancel 1999, p. 356, (...)
- 89 The “punicized territories” under the control of Carthage were situated, according to Polybius (3, (...)
- 90 Manfredi 2003, p. 331. For the political and cultural meaning of the term ‘Libyphoenician’ see Man (...)
92The importance of Ghizène at that time can be explained in the historical context of the expansionist ambitions of Carthage under the Magonides87. Since the end of the 6th c. B.C., Carthage interfered more and more in the Libyphoenician emporia for political88 as well as economic reasons, the latter ones driven by Carthage’s interest in exploiting the various resources of this region which was under its administrative control at that time89 and deeply influenced by Punic culture90. Ghizène was probably part of a first wave of Phoenician foundations in the region of the emporia, just like Lepcis Magna.
- 91 Ras Segala was probably the port of Zitha: Slim et alii 2004, p. 18.
93In the second half of the 5th c. B.C., a second wave of “fondazioni secondarie” followed, including Gigthis, Haribus, Zitha/Ras Segala91, and Sabratha. The appearance of these new coastal settlements can be explained by three historical reasons.
- 92 Lancel 1999, p. 354.
- 93 Ben Tahar 2016-2017, p. 75.
94Firstly, there is what was called by S. Lancel “l’ancrage africain”92. Carthage tried increasingly to exploit the agricultural and maritime resources in Africa. One of the manifestations of these attempts was the foundation or the development of coastal settlements closely linked to their hinterlands. The appearance of Carthaginian common and cooking wares at the coastal sites might indicate the presence of Carthaginian communities, which tried to create the best possible conditions for the exploitation of the maritime resources and for the control of the maritime trade with Eastern Tripolitania93.
- 94 Ben Jerbania 2011, p. 80-84 ; Bechtold 2016-2017, p. 97-98.
- 95 We would like to take this opportunity to thank Yamen Sghaier (INP, Tunis) for having provided us (...)
- 96 Data provided by excavations carried out by one of the authors (SBT) in 2011, 2014, 2017 and 2018 (...)
- 97 Bechtold 2016-2017, p. 97.
- 98 Bechtold 2016-2017, p. 106.
95Secondly, a restructuring or reorganization of the Mediterranean trade network started to take place at the end of the 5th c. B.C., with Malta as a new hub for the redistribution of amphorae towards the Byzacena, the Little Syrtis, and Pantelleria94, omitting the port of Carthage. One of the most important new African seaports for the storage and further distribution of amphorae towards the emporia was probably Leptiminus. A strong indication is the fact that amphorae as well as Black Top Red Cooking Ware produced in the Byzacena, the fabric of which is quite similar to that of Leptiminus95, were found together with Carthaginian and Maltese ceramics in several stratigraphic contexts of the 4th and 3rd c. B.C. at Zitha, Ghizène96, and Meninx as well. The Byzacena was apparently an important economic player in this new maritime network. This is also indicated by the fact that amphorae of type Gassner 6, produced in Southern Calabria, are completely lacking at Carthage97 but well attested in the Sahel zone and in the Tripolitanian emporia. On the other hand, the Ionio-Adriatic Corinthian B amphorae which circulated more or less in the same period are present at Carthage as well as in the Sahel98 and in Tripolitania. This contrast illustrates the complexity and multiplicity of the Mediterranean networks at that time, but causes problems with interpretation for those who are interested in economic history.
- 99 Ben Tahar 2018, p. 336. The hypothesis of the existence of trading contacts between Jerba and Gara (...)
- 100 Ben Tahar 2016-2017, p. 72.
96Thirdly, the extension of the trans-Saharan trade under the aegis of the Garamantes required an increased access to seaports, such as Gightis and Tacapae, in addition to the famous port of Lepcis Magna99. It is quite probable that the island of Jerba played an important role as a mediator for the trans-Saharan trade, as has already been assumed100.
- 101 See Fontana, Ben Tahar, Capelli 2009, p. 254.
97As far as Meninx is concerned, the question is how to interpret the fact that the earliest ceramics from our excavations do not date before the middle of the 4th c. B.C. There are two possible explanations. One possibility is, of course, that we simply have not found earlier material yet; at the present state of research, we cannot rule out that the settlement was founded already in the 5th c. B.C.101. On the other hand, the few sondages where we reached the very earliest cultural layers, provided us with quite a lot of datable findings, none of which is earlier than the middle of the 4th c. B.C. If the settlement activities at Meninx started only at about that time, this might indicate that, until then, the seaports of Ghizène, closely linked to Henchir Bourgou, and Guellala/Haribus were sufficient to meet the island’s needs to participate in the maritime trade of the region. This question can only be answered by conducting further excavations.
98Based on these results, we plan to continue our interdisciplinary investigations by exploring the rôle that Meninx played in its regional context in the different periods of urban history.
- 102 Fentress 2009, p. 76-80 ; Fentress et alii 2009, p. 131-133.
- 103 For the present state of research on Bourgou see Ben Tahar 2018 and Ben Tahar et alii 2020.
99Most promising results in this respect are to be expected from a systematic comparison with the site of Bourgou, located in the northeastern part of Jerba at a distance of only around 15 kilometres from Meninx (fig. 1). Bourgou (it is controversially debated whether or not the site is to be identified with ancient Thoar/Phoar, mentioned in the written sources) was the most important town of the island in Punic times before it started to decline already in the 2nd c. A.D., losing its leading position to Meninx102. As recent archaeological investigations have demonstrated, settlement activities started here as early as in the 8th c. B.C, and the site was inhabited until the 7th c. A.D.103.
- 104 For Ghizène see above.
100Despite the close proximity between Meninx and Bourgou, the two cities are remarkably different from each other in several respects. Bourgou is situated inland, had a fertile hinterland, and its economy was mainly based on agricultural production, whereas the next seaport at Ghizène was situated at a distance of some 7 kilometres from the town104. Bourgou’s urban layout in Roman times was centripetal, with the main routes oriented from the outside towards an elevation in the centre of the town. Recently, the first comparative studies of stratified findings from Bourgou and Meninx have revealed notable differences even in the urban lifestyle, as indicated, for example, by the use of different forms of drinking vessels, or by very different preferences in meat consumption in Roman times. These differences are significant enough to suggest that Bourgou was inhabited by a largely indigenous population which persisted over the centuries and maintained its local traditions that stretched back to the Punic period, in contrast to Meninx where the Roman way of urban life was adopted in all its aspects, as is the familiar picture for many cities in Roman North Africa.
101The striking differences between the two neighbouring cities are a phenomenon of exemplary interest and require detailed study and attention to their economic, cultural and social implications. A systematic comparison of the different rôles that Bourgou and Meninx played within the same geographical and economic microregion is a promising way of getting a better understanding of the rôle that the island of Jerba played as an epicentre between Mediterranean trading networks and the nearby mainland with its trading routes from and into the Sahara.
Acknowledgements
102We wish to thank all the people who have participated in the campaigns in 2017 and 2018. Trench directors: Marina Apatsidis (Köln, Trench 8), Robert c. Arndt (Buochs / Switzerland, Trenches 4 [in 2018] and 5), Nicolas Lamare (Paris, Trenches 2 and 9), Christoph Lehnert (Bonn, Trench 3), Robert Peitsch (Bonn, Trench 4 [in 2017]), Nichole Sheldrick (Oxford, Trenches 1 and 7), Linda Stoeßel (Tübingen, Trench 6). – Measurements: Agnes Weinhuber, Max Hofacker, Beatrice Waldner (Technische Universität München). – Excavation techniques: Julian Schierenbeck, Tobias Prange, Karl Wachsmann (Hochschule für Technik und Wirtschaft, Berlin). – Building research: Kilian Wolf (TU München). – Information technology: Paul Scheding (LMU München). – Survey of the quarries: Dennis Beck (Freie Universität Berlin). – Director of the finds department: Karin Mansel (München). – Ceramics: Tomoo Mukai (CNRS, Aix-en-Provence), Francesca Assirelli, Bianca Maria Mancini (Università di Bologna). – Coins: Saskia Kerschbaum (Kommission für Alte Geschichte und Epigraphik, München). – Small finds: Salvatore Ortisi, Katharina Berz, Stephanie Schmittner (LMU München). – Archaeozoology: Joris Peters, Simon Trixl, Lisa Bauer (LMU München). – Archaeobotany: Michèle Dinies (Deutsches Archäologisches Institut, Berlin). – Restoration of finds: Elisabeth Lehr-Stempel (München). – Photography: Stefanie Holzem (Bonn), Björn Schumann (Hochschule für Technik und Wirtschaft, Berlin). – Mekki Aoudi (Université de Sfax). – Students and other members of staff: Armin Bektic, Lorenzo Canals, Lena Gabler, Rosa Galusic, Alexander Köppe, Sebastian Kranz, Marisol Lang Navarro, Marco Rossini, Maria Rüegg, Ernst Bartlin Schöpflin, Julia Thois (LMU München); Fatma M’Barek (TU München); Annika Kirscheneder, Tobias Krug, Nathalie Weilbächer (FU Berlin); Ferdinand Wulfmeier (Universität Bonn); Saloua Abidi, Yousra Bouabid, Riadh Ben Brahim, Siwar Esseghir (Université de Tunis); Samir Ben Hmouda, Houda Ben Hamida, Kouloud Ghram, Thabet Ghabri (Université de Sfax); Imen Askri, Hasna Ben Hadada, Rym Jrad (INP, Jerba); Ali Mansouri (INP, Medenine); Lassad Boumellassa (INP, Gigthis); Zied Msellem (INP, Tunis); Hajer Zaghdoud (Musée de Zarzis); Marouen Ben Brahim; Amira Jrad, Olfa Souissi. – Architectural studies: Johannes Lipps, Dennis Joch, Elisabeth Kammerer, Daniel Richter, Annika Skolik, Julien Vogel (Universität Tübingen), Katharina Sahm (Technische Universität Berlin). – Geology: Vilma Ruppiene (Universität Würzburg). – Geophysical prospection: Jörg Fassbinder (LMU München and Bayerisches Amt für Denkmalpflege), Marion Scheiblecker (LMU München). – Underwater Archaeology: Tobias Pflederer (Bayerische Gesellschaft für Unterwasserarchäologie), Max Fiederling, Maximilian Ahl, Michael Heinzlmeier, Eric Kreßner, Marko Runjajić (LMU München), Mladen Pešić (International Centre for Underwater Archaeology, Zadar / Croatia), Abdallah Mateur, Hamza Chouikhi (Ajim, Jerba). – Logistics and transport: Rhouma Abjlil (INP Jerba). – The project was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation – Projektnummer 264557697) and the Bayerische Akademie der Wissenschaften (Bavarian Academy of Sciences and Humanities). – Further information and more images are available on the project’s website: https://www.klass-archaeologie.uni-muenchen.de/forschung/d-projekte-laufend/meninx/index.html.