The 2022 excavation (5–16 September) in the vicinity of the church of Santa Maria della Strada (Matrice, Campobasso), led by the British School at Rome, the Ashmolean Museum and King's College London, was the second season in a new phase of targeted excavations following earlier work between 1980 and 1984 commenced by the Universities of Sheffield and Aberdeen (Lloyd and Rathbone, Reference Lloyd and Rathbone1984). Following the discovery of the site during the construction of a new road, at the invitation of the Soprintendenza an excavation was conducted, directed by John Lloyd, which revealed a multiphase settlement. The earliest phase was identified through evidence of occupation in the Late Bronze Age, after which a small farmstead in the Samnite period was constructed (Lloyd, Reference Lloyd, Baker and Lloyd1991b: 184). Roman control over the area, following the Social War, led to the restructuring and enlargement of the farm into a villa which was located on a key route of transhumance on the western edge of the Biferno valley.

The new phase of research, the main purpose of which is the publication of the site, aims to refine the site's overall chronology as well as assess whether the previous excavations had fully identified the extent of the settlement. A campaign of geophysical prospection in 2017 revealed a further extension to the west which was subsequently excavated in 2018 (Kay, Roberts and Rathbone, Reference Kay, Roberts and Rathbone2019). The short 2022 season focused on three key areas identified within the villa through small-scale targeted excavation: the first trench extended a sondage opened in the 1980s where a deposit of late antique material was recovered, providing crucial evidence for the later occupation of the site; a second trench was opened in an area of production identified by the earlier excavations, and aimed to test theories as to the usage of the facility; the third trench, similar to the 2018 excavation, focused upon a potential internal area of the Samnite structure.

In 1983 a small trench had been opened on the northern side of a long wall, dating to the Republican period (Lloyd, Reference Lloyd, Baker and Lloyd1991b: 185), which was interpreted by Lloyd as defining the northern limit of the villa. Against the wall the excavation had revealed a significant deposit of fifth-century AD material attesting to a significant phase of occupation as well as connectivity beyond the local sphere due to the presence of imported ceramics. The 2022 season therefore sought to investigate further the deposit north of the wall that had remained partially excavated. The excavation recorded similar classes of late antique material, but at a depth beyond the previous investigations a mortared tile structure was discovered. The tiles, which in part continued underneath the unexcavated southern section, appear to have been exposed to extreme levels of heat, with the surrounding context 90 per cent ash with 10 per cent burnt limestone. The tiles were laid flat and consisted of three layers, bordered by unworked blocks of limestone. Whilst the structure was not fully excavated, a potential hypothesis is that this may have been part of a limekiln. The high levels of heat to which the tiles were exposed exclude the possibility of a simple hearth, whilst few indications of production have been found during excavations at Matrice, excluding the possibility of a pottery kiln. The earlier excavations had recorded dumps of lime, principally inside a dolium, indicating the production of lime. Soil and tile samples were collected for further chemical and archaeometric analysis.

In the eastern range of the villa, dated by Lloyd to the early first century AD (Lloyd, Reference Lloyd, Capini and Di Niro1991a: 262), the previous excavation had discovered several sunken dolia, seemingly associated with low rectangular stone structures interpreted as workbenches used in the production of wine. The new excavations sought to assess whether the purposes of the benches could be securely identified and whether these could be confirmed as belonging to the early Imperial phase. A structure previously identified only in section was therefore fully excavated and samples collected for residue analysis, as well as a complete plan being drawn of the production facility (Fig. 1). In order to assess the chronology given by Lloyd to the eastern range, a small trench was excavated alongside the internal face of the eastern wall. The upper layers were partially contaminated by a later medieval barn that was built over the structure; however, the fill alongside the wall could be dated to the first century AD. At a depth of 0.90 m an earlier wall on a differing alignment, that continued beneath the eastern wall, was recorded, potentially belonging to the Samnite phase of the settlement and therefore extending the size of the settlement.

Fig. 1. A rectangular stone structure associated to the production facility (photo: S. Kay).

A small trench was excavated in the northeast corner of the structure built from roughly hewn large limestone blocks dated by the earlier excavation to the second century BC (Lloyd, Reference Lloyd, Baker and Lloyd1991b: 184). A similar sondage in 2018, excavated a short distance to the south, had recorded an internal wall associated with material including black-gloss ware and dolia, confirming the chronology (Kay, Roberts and Rathbone, Reference Kay, Roberts and Rathbone2019). The 2022 sondage recovered similar material, but also revealed that the structure had an internal facing with smaller blocks bonded with a coarse white mortar that served to straighten the wall face where spaces were left by the large limestone blocks (Fig. 2). A compact clay layer at the foot of these blocks may indicate the remnants of a beaten earth surface, and material recovered from this layer suggests a date of the second century BC.

Fig. 2. The internal wall face of the Samnite building (photo: S. Kay).

The two seasons of excavation, together with the geophysical surveys around both the villa and the nearby church of Santa Maria della Strada, have provided a more detailed overview of the immediate topographical context of the site. The refined chronology, based upon the newly documented stratigraphy and ceramics, will assist in appraising the results of the earlier excavations, with the datasets drawn together in the final monograph.