EXCAVATION OF THE SAN FELICE VILLA

The ninth season at the Roman villa at San Felice continued the more limited and focused excavation that began last year, as the Project seeks to draw the excavation phase to a close, while placing more emphasis on the field survey covering the territory west of the Basentello river. As noted last year (McCallum and vanderLeest, Reference McCallum and vanderLeest2013), this survey is attempting to examine the broader context of this imperial estate and the region in which the villa existed, as well as trying to understand the process of Romanization and cultural hybridization in this region of southern Italy.

Last year, in an area southeast of the peristyle of the villa, three basins set into an opus signinum floor were discovered. At that time we were not able to investigate this area thoroughly, and although it seemed clear that they were related to the processing of agricultural produce, the nature of that activity was not clear. Thus we returned to this area of the site and expanded the area of excavation around it.

The room containing the basins is about 4.5 m east–west by 3.5 m north–south, with lower courses of the east, north and west walls well preserved (Fig. 1). Because of the modern drainage ditch between the field and the modern road, the south side is largely lost, but a small surviving section of the south wall confirms the overall dimensions. Two rectangular basins (c. 0.85 × 1.0 × 0.77 m) were set close to the eastern wall. A circular basin (diameter 1.3 m; depth 0.9 m), constructed using part of a large dolium, lay near the middle of the south wall of the room. Clear indications were found of a third rectangular basin that had been set north of the circular one, but the basin itself had been completely robbed. An opus signinum floor surrounded and connected these four basins, forming an apron against the adjacent walls. This floor ended just west of the circular basin, with a clear edge. The area between this edge and the western wall contained only rubble down to the natural sand layer, and it was not possible to determine how this area might have been used. The floor between the two rectangular basins in the east was damaged, but the shape of the disturbed area seems roughly circular. It has been suggested (Jeremy Rossiter, pers. comm.) that this might have been the location of a small screw press and that this room would have been used for wine and/or oil making.

Fig. 1. San Felice. Excavation of the villa.

We also excavated the area to the east of the room with the basins, hoping to discover more extensive evidence of the industrial activities in this part of the villa. However, the adjacent room, despite having signs of post-occupation use, had such an extensive layer of rubble that we were not able to complete the excavation.

Our other excavations in 2013 explored the limits of the villa. Since the southernmost areas excavated each had north–south walls that extended into the modern drainage ditch, we wondered whether the remains of the villa extended below the modern road and into the field on the other side. The hill beyond the road (to the southeast) begins to slope up rapidly, and so fill carried down the hill and onto the terrace would have covered more deeply any remains on that side. Two areas for excavation were identified south of the road: in each of these we used a backhoe to excavate down over 3 m. The fill removed was remarkably clean, with almost no finds, and the soil layers soon changed to layers of sterile sand: we concluded that the southern edge of the villa must lie under the modern road.

In previous years we had uncovered several rooms to the north of the peristyle. Not much further north of these rooms the terrace begins to drop away. Last summer we excavated one more area on this northern edge to see if we would trace a northern limit for the villa. This also allowed us to follow for a longer distance the large and well-made drain that started at the northwest corner of the peristyle and cut through four other rooms as it led north and west. Where best preserved this drain had a channel about 50 cm deep and large rectangular limestone cover slabs. The new excavations revealed a length of about 4 m of the drain as it moved north, as well as another curve as the drain changed direction and began to lead west. The drain itself appears to have been filled during the late first or early second century ce, and contained a small number of almost complete fine-ware jugs as well as half of a marble tabletop.

As this area was so close to the modern surface, most of the remains of the associated rooms were greatly disturbed. None the less, patches of an east–west wall with an opus signinum floor on its north side and many fragments of painted wall plaster gave clear indications that we had not yet reached the northern limits of the villa, and suggest that this part of the structure had a residential or reception function at the time of its construction. However, it is likely that any remains of the northern outer wall of the structure are completely lost through ploughing and erosion of the edge of the terrace.

No new excavation will take place at San Felice in 2014, as we shall turn our attention to studying the material that has been recovered over the last nine years, with the goal of producing a book-length final excavation report.

SURVEY

As part of our investigation into the imperial estate at San Felice/Vagnari and its environs, since 2012 we have been engaged in an ongoing field-survey project. In the summer and autumn of 2013 a team of sixteen field walkers covered an area of approximately 35 square kilometres and identified a total of 109 sites, of which 90 can be dated to periods between the ninth century bce and the seventh century ce (Fig. 2). Further work on the slopes of Monte Serico, where in 2012 we discovered that the late iron age and Archaic/Classical-period sites extended well beyond the boundaries of the archaeologically protected area established in the late 1990s, has brought to light data indicating that the northeastern slopes, including a natural terrace near Masseria Chio, were occupied throughout the Hellenistic, Republican and Imperial periods.

Fig. 2. Basentello Valley. All sites discovered in 2012 and 2013.

To the south and east of Monte Serico, on a plateau overlooking the Basentello river, a densely packed mass of 36 sites was discovered, many of whose surface scatters were within 50 m of adjacent scatters. The scatters generally included heavily abraded roof-tiles, utilitarian pottery made in a fine-bodied paste, occasional loom weights, and, at almost every site, small fragments of volcanic millstones or grinding-stones; all of these finds suggest the presence of residential structures. Although very few datable sherds were recovered from these sites, those that were are overwhelmingly black gloss (Fig. 3). The majority of the diagnostic sherds can be dated to the fourth through late third centuries bce, although there are some forms that appear to be from the second and first centuries bce. Only three of the sites, situated at the southernmost limit of Serra Fontana Vetere, had evidence for continued occupation into the first century ce.

Fig. 3. Basentello Valley. Sites with black gloss pottery, and inset showing the sites on Serra Fontana Vetere/Piana Zaccari.

Another important finding of the 2013 season is a series of Imperial-period sites that appear to grow in size during the late antique period (principally from the mid-fourth century ce, based on the presence of ARS C and D as well as large numbers of late Roman painted ware sherds), some of which may be agricultural villages or vici (Fig. 4). This also may indicate a change in the system of landownership and agricultural exploitation.

Fig. 4. Basentello Valley. Locations of possible late antique villages or vici in the study area.