You are here:Home/Research/ Byzantine Studies/ Project Grants/ Preliminary Report on Excavations of a Catacomb Church at Lamta, Tunisia, 8 May–16 June 2006: Dharet Slama Site 304, Leptiminus Archaeological Project

Preliminary Report on Excavations of a Catacomb Church at Lamta, Tunisia, 8 May–16 June 2006: Dharet Slama Site 304, Leptiminus Archaeological Project

Susan T. Stevens, Randolph College, Project Grant 2006–2007


Fig. 1: Sketch plan of the subterranean early Christian burial complex (Stevens 2006–2007)
Fig. 1: Sketch plan of the subterranean early Christian burial complex.

A subterranean early Christian burial complex, discovered in 2002 by Nejib Ben Lazreg, lies adjacent to the late second–fourth–century East Cemetery of ancient Leptiminus (modern Lamta). With walls and ceilings partly rock-cut and partly constructed in opus africanum masonry and concrete, this underground building includes at least four rooms (fig. 1). A stairway descends from the surface, from the lower landing of which Tunnel 2 runs south under the Roman necropolis some 4.5 m to the northeast corner of the Roman cemetery’s Hypogeum 1. Immediately north of the stairway landing is a vestibule that provides access to the Small Vaulted Room (SVR) through a door in its west wall and to the Large Vaulted Room (LVR) though a doorway that occupies its whole north end. Both the vestibule and the Large Vaulted Room contain unusual and extraordinarily well-preserved tomb mosaics (Ben Lazreg, “Roman and Early Christian Burial Complex at Leptiminus: First Notice,” JRA 15 [2002] 336–45). Small segments of catacomb tunnel had been uncovered by excavations over the years in the vicinity of the East Cemetery, but their connections to each other and to the underground building were unknown. Furthermore, two large and well-preserved underground features, aligned roughly northwest–southeast, were identified by ground penetrating radar (GPR) in 2005 just north of the underground building and tentatively interpreted as tunnels. The goals of the 2006 excavation were (1) to explore the function of the Small Vaulted Room located in 2002, (2) to determine the extent of the adjacent Large Vaulted Room, and (3) to investigate the connection of the underground rooms to adjacent tunnels.

The Small Vaulted Room

Fig. 2: Overview of the Small Vaulted Room during excavation (Stevens 2006–2007)
Fig. 2: Overview of the Small Vaulted Room during excavation.

Excavation of the Small Vaulted Room in 2006 revealed a second doorway in its southwest corner that provides access to Tunnel 1. The visible portion of this passage, oriented northwest–southeast and measuring some 1.8 m wide and 6.25 m long, has a rock-cut south wall and a north wall constructed of opus africanum, like the east wall of the vestibule. Widely spaced rows of burial niches (loculi) were found cut into the friable bedrock of the south wall and subsequent burials were found in cists stacked along the south wall and in simple pits on top of each other in the fill of the passage.

Fig. 3: Tombs along the north wall of the Small Vaulted Room (Stevens 2006–2007)
Fig. 3: Tombs along the north wall of the Small Vaulted Room.

The original floor of the SVR was cut for five tombs, oriented west–east. Three individuals, a young adult female in a tomb against the south wall of the room (see fig. 2) and male and female adults in tombs in its northwest corner, were buried in cists, the top 10–15 cm of which are above floor level and covered with a layer of white plaster or mortar (fig. 3). The only preserved tomb marker is a Chi-Rho and the inscription “Eolius dor/mit in pace in blue and green glass mosaic tesserae. Two other graves in the room are below floor level: one for an adult in a pitched tile tomb in the center of the SVR, and the other for an infant, buried in an amphora in the northeast corner.

The Large Vaulted Room

Fig. 4: Overview of the floor of the Large Vaulted Room (Stevens 2006–2007)
Fig. 4: Overview of the floor of the Large Vaulted Room.

The segment of the Large Vaulted Room excavated in 2006, measuring 4.9 m wide and 3.5 m long, extends almost to the room’s western terminus. Although the LRV lies immediately north of the Small Vaulted Room, it is not directly connected to it; access was probably provided at its eastern end through a doorway in the north wall of the vestibule. The floor of the LVR was cut for at least seventeen tombs, placed individually and ad hoc, but in an identifiable general sequence. Eleven preserved tomb markers are fully exposed while the other six partially underlie the east and west baulks. In the course of the 2006 season, the tomb markers of the LVR were cleaned and recorded but not lifted.

All but one of the tomb markers lies parallel to the north and south walls of the room, and twelve indicate the underlying burials are oriented west–east. Six of the markers are anepigraphic and made of marble, and from their size five were for children. Three markers are single slabs of marble, a fourth consists of a marble slab and a panel made of small pieces of marble, and the fifth and sixth are made of two to four different marble slabs. Eleven tomb markers in the LVR are mosaics; many of these employ glass tesserae and a variety of techniques for shading and perspective comparable to floor mosaics of the North African realistic tradition. These mosaics also contrast in size, format, content, and execution with known tomb mosaics from Leptiminus and its immediate environs.

Fig. 5: Tomb mosaics of Florentia, Elia Theodora, and Agapia (Stevens 2006–2007)
Fig. 5: Tomb mosaics of Florentia, Elia Theodora, and Agapia.

Human figures, usually rare on tomb mosaics, appear on nine of the LVR tomb mosaics, and the figures themselves are unusual: four are Good Shepherds and six are portraits of the deceased. Perhaps the liveliest portraits commemorate three infant girls buried side by side on the south side of the LVR (fig. 5). All hold a bird in one hand and a bunch of grapes in the other, but their hairstyles and postures seem age specific. The eight-month-old Florentia has short blond hair and is shown seated on a cushion while Elia Theodora, at fourteen months, has curly brown hair over her ears and walks with her cloak billowing out behind her. The format of the figural LVR tomb mosaics is tripartite: an inscription is over the head of the deceased, a figure over the torso, and symbols (especially roses and birds) over the feet, but they lack the otherwise normative Chi-Rho. Brief inscriptions record the name of the deceased, the formula “in pace” with or without the verb vixit or dormit, sometimes followed by the age at death and/or date of death. The attribute fidelis, otherwise ubiquitous on Christian epitaphs, is notably absent from these inscriptions, although two children are identified as “innocens.” Unified by a limited number of conventions the tomb mosaics of the LVR give the impression of a distinctive collective identity.

In the absence of liturgical arrangements or clerical titles on tomb inscriptions that might be expected in a church, we interpret the LVR as an elite and privileged funerary chapel, associated with the adjacent catacomb tunnels but not directly connected to them.

The Catacomb System

Both presumed tunnels located by the 2005 GPR survey were sealed under bedrock. However, the 2006 excavation revealed a foundation wall, built on bedrock and aligned east–west, parallel to the north wall of the LVR but 2 m north of it. Associated with this wall is a vaulted stairway cut through bedrock and constructed in much the same way as the stairway into the underground building. The stairway starts at ground level close to the western terminus of the LVR; it ends in a rock-cut tunnel running northeast–southwest, varying in width from about 2 m to more than 4 m and with a visible length of approximately 13 m. This tunnel is the center of an extensive system of rock-cut subterranean passages: it continues north from the staircase for at least 7 m, perhaps intersecting with the possible tunnels identified by GPR, and some 8 m to the south where it probably meets the perpendicular Tunnel 1 that originates in the SVR. Two parallel tunnels, which lead off in a westerly direction, are visible from the stairway across the main tunnel (see fig. 1). The main tunnel is filled with waterborne sediment and boulders of fallen bedrock to within 80 cm of its ceiling; excavation was not attempted because of the dangers of falling bedrock and noxious air. The absence of visible tombs in the main tunnel is consistent with the situation in Tunnels 1 and 2, where burials appeared only after 1–1.5 m of fill was removed.


A study of the construction and configuration of the rooms of the underground building indicates that it was planned and constructed all at one time, for which preliminary ceramic analysis suggests a date in the mid-fourth century. The style of the mosaics of the LVR, and the presence of the Africana III/Keay 25 amphora burial in the SVR suggests that the tombs in both rooms were inserted between the late fourth and mid-fifth century. Since the rubble fill of the LVR includes large pieces of vault that overlies a complete orthostat fallen from the north wall onto a layer of wall plaster covering the floor, the large room appears to have gone out of use immediately after its vault collapsed, perhaps in the third quarter of the fifth century. By contrast, the earliest stratified fills of the SVR, rich in pottery and other finds, suggest that the small room continued to function as an entrance to the catacombs into the sixth century.

At the end of the 2006 season, the LVR and the main catacomb tunnel access were backfilled to modern ground level in the interest of public safety and site conservation. Analysis of the finds and data from the 2006 excavation is ongoing while plans are made for future campaigns of excavation and conservation in the Small and Large Vaulted Rooms.


Funding for the 2006 excavation was provided by grants from Dumbarton Oaks, the Shohet Scholars Program of the International Catacomb Foundation, and the Social Sciences and Humanities Research Council of Canada. I would like to thank my colleagues and codirectors Nejib Ben Lazreg (Institut national du Patrimoine, Tunis) and Lea Stirling (University of Manitoba) for welcoming me into the long-running and meticulously published Leptiminus Archaeological Project, which has provided invaluable context for the current excavation. I am very grateful to the efforts of the congenially international 2006 team of twelve field supervisors and specialists, fifteen students, and seventeen local support staff.