Eleven squares were excavated in an area of c. 1,800 sq m and three units, each consisting of two phases, were partly exposed. The finds from the early phase included a winepress, a plastered installation and a habitation level. Two buildings and a refuse pit of a pottery workshop were ascribed to the late phase. Both phases dated to the Byzantine period; however, the area between the units was not excavated and therefore the stratigraphic ties between them could not be defined.
Winepress. The relatively small winepress had a complex design. Its walls (width 0.40–0.75 m, height c. 0.9 m; Figs. 3, 4), built of fieldstones and bonded with hydraulic cement, were founded into a hamra level on top of the natural bedrock.
Part of a treading floor (6.6 × 7.2 m) that was disturbed by later construction (Loci 125, 156) was exposed in the center of the winepress. The floor, coated with white hydraulic plaster (thickness 5–12 cm), was paved with white industrial mosaic that sloped gently to the east, toward the collecting vats.
A narrow channel (diam. c. 0.12 m) was discovered in the eastern wall (W16) of the treading floor. It led to a square collecting vat (L159), whose northwestern corner was only uncovered. The northern wall of the collecting vat (W20) was plastered on both sides (thickness of plaster c. 0.1 m), suggesting that another collecting vat, not exposed, was located to its north.
Three rectangular cells, connected to the treading floor by plastered channels, were revealed south and west of the floor. The interior surface of the cell walls was coated with white hydraulic plaster. The cells in the south (L133) and in the west (L142) were paved with a white industrial mosaic. The tesserae in the floor of Cell 142 were larger than those in the rest of the winepress, evincing repairs that had been carried out on the floor of this cell. The cell in the southwestern corner (L141; 1.0 × 2.5 m) was paved with potsherds, arranged on their side in a herringbone pattern and bonded in hydraulic plaster. A shallow depression (depth 5 cm) in the middle of the cell may have been created due to the settling of the floor bedding. The thick walls of the cells may indicate that they were used as containers for the initial fermentation and treading floors, which had not survived, were built above them.
Building 155. The southeastern part of the treading floor was found sealed beneath a collapse of large stones, plaster fragments and floor sections of potsherds in herringbone pattern that were apparently removed from the winepress. On top of the collapse was part of a later building, in which a single room was exposed (L155; Fig. 5). The room was delimited by wall sections in the north (W12; width 0.4 m, height 0.4 m), south (W13; width 0.55 m, height 0.3 m) and east (W14; width 0.6 m, height 0.5 m). The walls, built of small fieldstones and not plastered, were founded on top of the treading floor’s mosaic pavement. The floor of the room was not identified as it had been destroyed by the plowing.
Building 139. Two phases were discerned in this building, which was exposed west of the winepress. The remains ascribed to the early phase included a level that was not excavated and was overlain with fragments of pottery vessels, several kurkar stones and the rim of a round installation, built of fired mud-brick material. Two walls that formed a corner, which was probably part of a room, were associated with the late phase. The walls (W5, W9; width 0.6–0.7 m), built of medium-sized fieldstones and founded on top of the hamra level, were preserved two courses high (0.09–0.20 m). A tamped earth floor (Loci 137, 139; thickness 0.15–0.20 m) abutted the base of the walls and another floor (L147; thickness 0.2 m), into which a clay tabun (L158; diam. 0.75 m, wall thickness 2 cm; Fig. 6) was embedded, was exposed west of W5. The tabun was lined with a row of small kurkar stones on its exterior and its upper part was found covered with small stones and many potsherds.
Plastered Installation 157. The northwestern corner of this installation was revealed c. 5 m west of Building 139. Its walls (W4, W18; width 0.40–0.56 m, preserved height 0.85 m; Fig. 7), built of fieldstones and bonded with mortar, were coated with several layers of hydraulic plaster (thickness c. 0.12 m). The walls of the installation, as in the case of the winepress, were sunk into the hamra level (depth c. 0.67 m). Below the stone collapse that covered the installation was a floor of white hydraulic plaster that abutted its walls (L157); hence, the installation may have been another winepress.
Two fieldstone walls (W10, W11; width 0.4–0.6 m, height 0.2–0.4 m), built on top of the hamra level, were exposed c. 6 m south of the installation’s corner. The alignment of the walls and their method of construction were similar to those of the adjacent plastered installation and therefore, one may assume that they were part of the same unit. It was not possible to identify the floor that abutted the walls due to the later disturbance of a pottery workshop refuse pit (see below).
Pottery Workshop Debris. Layers of tamped fill (up to 1.8 m deep), which contained workshop debris that consisted of numerous pottery vessel fragments, ceramic refuse and pieces of fired mud bricks, were exposed in the row of Squares A2–A12, west of the plastered installation. The eastern part of a pottery workshop refuse pit was exposed in Square A11 and levels of similar material had accumulated inside it (L150; diam. c. 4 m, depth 1.8 m; Fig. 8).
Two excavation squares were opened in a 20 × 25 m area, south of Area A. Layers of pottery workshop debris, which consisted of fragments of pottery vessels, ceramic refuse waste, mud-brick material and fragments of fired mud bricks (Loci 502, 503; 0.20–0.25 m below surface, depth 1.6–1.8 m), were discovered.
The pottery assemblage from the phases and the different areas in the site were not different from each other. It was characteristic of the southern coastal region in the fifth–sixth centuries CE and consisted of bowls (Fig. 9:1–3) and imported bowls, including Late Roman C bowls (Fig. 9:4–6) and African Red Slip Ware bowls (Fig. 9:7, 8), a combed krater (Fig. 9:9), cooking vessels (Fig. 10:1–6), numerous Gaza jars that are the most common vessel type in the region (Fig. 10:7–11), as well as jugs (Fig. 10:12, 13) a juglet (Fig. 10:14) and a flask (Fig. 10:15).
The glass finds are very meager. Of the twenty-three discovered fragments, eighteen vessel fragments were identified.
Most of the vessels date to the Late Roman and the beginning of the Byzantine periods and include fragments of cups with a solid base (Loci 123, 125) that date to the fourth century CE and a tooled-out and pushed in base fragment of a cup or a juglet (L139), dating to the same period. Fragments of bottle rims and bases that date to the Byzantine period (Loci 120, 131, 148), juglet handles (L153) and part of a single handle of a double kohl tube (L148) were also found. All the vessels belong to local types that are prevalent in the Land of Israel in the period under discussion.
Eight coins were found, five of which were identified. Four of them are minimi, small bronze coins that date to the end of the fourth century CE (IAA 109778–81; Loci 116, 118, 121). The latest coin is a five numini bronze coin of the Byzantine emperor Justin I (517–528 CE) that was minted in Constantinople (IAA 109782; L141). These coins are very common to all sites of the Byzantine period.