Area A (Fig. 1)
Meager remains of two buildings were exposed. Only the northern part (3.0×9.5 m) of the southern building (L346), which consisted of just the foundations of two rooms (max. height 0.2 m), was exposed. The finds were too scant for dating the structure.
Part of another building (L316; 6.5×12.5 m) was excavated north of Building 346. Its walls survived mostly in the east, where two phases of use were discerned. Sections of walls, scant remains of a stone floor and built surfaces of kurkar, bonded with mortar, were found in the western part of the building, and it is unclear if the surfaces are the remains of walls or floors.
The finds from Building 316 included fragments of pottery vessels, among them jars, amphorae, cooking pots, bowls, kraters and a few fragments of lamps, fragments of glass vessels and coins, some from the Early Roman period (first–second centuries CE), but mostly from the Late Roman–Early Byzantine periods (third–fifth centuries CE). Potsherds, not found in an architectural context, dated to Iron Age II, the Late Byzantine period (fifth–seventh centuries CE) and the Middle Ages. Several objects of stone, marble and bone were also found.
Area B (Fig. 2)
Three construction phases (1–3) were exposed. Phases 2 and 3 were ascribed to a winepress composed of settling pits and a collecting vat, north and south of which were two treading or work surfaces, and east and west of which were several channels and fermentation cells. Meager architectural remains were ascribed to Phase 1. The phases described below follow the remains from early to late.
Phase 3. The settling pits and collecting vat were in the middle of the winepress. Their sides were coated with at least two layers of plaster and their floors were built of potsherds arranged in a herringbone pattern. At the eastern end of the settling pit’s floor (0.5×1.4×2.3 m) was a sump consisting of a ceramic krater (diam. 0.35 m); next to the sump was a pipe in a shallow depression, which led east to the collecting vat. In the center of the collecting vat (1.36×2.30×2.40 m) was a sump, part of which was made of a jar (diam. 0.8 m).
A rectangular treading floor (4.0×4.9 m) was discovered south of the settling pits and collecting vat. In the northwestern part of the floor, in a wall between it and the settling pit, was a round opening blocked with thick plaster that covered the wall. The treading floor was built of potsherds bonded in mortar, in groups arranged perpendicular to each other and covered with light gray plaster. Potsherds, glass and coins from the Late Roman–Early Byzantine periods were found on the treading floor, which may be indicative of its usage date.
Three rectangular fermentation cells (1.2×2.7 m; Fig. 3) arranged in a row were built in the eastern part of the winepress. The floors of the cells consisted of several layers of gray plaster applied to a bedding of potsherds laid on their side and forming a diagonal pattern; ceramic sumps (diam. 0.3 m) were embedded in their floors.
A lead pipe that led to a narrow channel (length c. 3.2 m; Fig. 4), running north–south, was discovered west of each sump. The channel conveyed the must to a collecting installation that did not survive and was probably located west or south of the cells. An opening blocked with plaster was at the southern end of the channel. Opposite the middle fermentation cell was another channel (length 1.1m) coming from the west that merged with this channel. This second channel had two openings: an eastern opening that faced south, which was blocked with plaster, and a western opening that consisted of a lead pipe, which emerged from the eastern side of the northern treading floor at a level lower than that of the treading floor’s surface. A third channel (length 4.8 m), which was connected to the first one, was partly exposed north of the two channels.
A fermentation cell, or another pressing surface whose floor was almost completely destroyed, was partly exposed west of the settling pit. An opening in the northern side of the surface was connected to a built channel (Fig. 5). The channel and the work area continued beyond the limits of the excavation. The eastern end of the channel terminated in the western side of the settling pit. The channel was probably negated by the plaster of the western wall that delimited the settling pit. No opening connected to this channel was found on the inside wall of the settling pit. The channel was built similar to the channels in the eastern wing and therefore they might belong to the same phase.
Dark brown soil fill, containing fragments of pottery and glass vessels from the Late Roman period, was excavated in several places in the eastern wing and near the southern surface. This is the earliest period when the winepress could have been built.
Phase 2. Two rectangular fermentation cells, one in the south (1.4×1.9 m) and one in the north (1.2×1.9 m), were built west of the fermentation cells of the previous phase and their construction was different than that of the earlier cells. The floors of these cells were covered with gray plaster and no sumps were installed in them. A narrow opening in the western side of the southern cell led to the northern treading floor (5.4×5.4 m), which was built west of the fermentation cells. The surface of the treading floor was coated with light gray plaster, applied to a foundation of potsherds laid next to each other and forming a herringbone pattern. A screw base (2.35×2.50 m) with a rectangular press bed (0.6×0.7 m) was discovered in the middle of the floor. An opening leading to a settling pit was discovered in the south of the treading floor.
The channels from Phase 3 were no longer used when the western fermentation cells and the northern work surface in Phase 2 were constructed, hence the Phase 3 cells were canceled altogether or their function was modified. During the last phase of the winepress’ operation, the southern surface was probably no longer used in its original function. At this time, only the northern surface, the cells from the second phase of the eastern wing and the collecting and settling vats were apparently used because their openings were not found blocked with plaster.
The finds from the winepress included fragments of ceramic and glass vessels from the Late Roman and Byzantine periods (third/fourth–sixth centuries CE), coins from the same periods, not later than the end of the fifth century CE, and a few animal bones. The winepress went out of use and was turned into a refuse pit still in the Byzantine period (end of the fifth–sixth century CE), based on the fragments of pottery and glass vessels and the coins that were found on the floor of the collecting vat. The latest identified coin is dated to the years 457–474 CE.
Phase 1. A wall was built on the refuse in the collecting vat, next to the vat’s northern side in the Late Byzantine period (the late fifth–early sixth century CE). Incorporated in the wall were lumps of cement with potsherds that were taken from the northern work surface and roughly hewn stones that were probably dismantled from the surrounding buildings or from Area A. Fragments of pottery vessels from the Late Byzantine period (sixth century–early seventh century CE) that were found above the level of stones that covered most of the winepress’ components probably belonged to this phase.
Area C (Fig. 6)
The foundation of a structure’s southwestern corner (2.5×4.0 m) was exposed in the north of the area. The finds were meager and included a few fragments of pottery vessels from the Roman and Byzantine periods and two coins that date to the fourth century CE. The foundations of two parallel walls, built similar to the walls of the structure, were exposed 8 m south of the corner. Next to them, an intact juglet and a bronze arrowhead from Iron Age II, a pit that contained fragments of pottery vessels from the Early and Late Roman periods and a coin from the fourth century CE, were found. A single coin from the Early Islamic period (ninth–tenth centuries CE) was discovered in fill that had no architectural context.
The poor state of preservation of the buildings in Areas A and C probably stemmed from the robbery of stones that might have been accompanied by damage to the floors. The paucity of finds from clear assemblages and the modern plowing at the site allow us to make a cautious proposal that the remains mostly date to the Late Roman–Byzantine periods. The finds might be part of an agricultural estate or village, and some of the buildings could have been used as storehouses that were connected to the winepress.
The winepress in Area B constitutes additional evidence of the complexity of the wine industry in the southern coastal plain, which reached its zenith in the Byzantine period. The construction phases, the differences in the construction of the winepress’ wings and the material finds seem to be indicative of prolonged activity. One sees in the winepress several different models of pressing surfaces alongside each other, starting with the simple model, without side cells, the use of fermentation cells or pressing with channels, and finally a transition to the model whereby the cells are perpendicular to the work surfaces in the middle of which is a screw press. The last model characterizes the complex winepresses that appear toward the end of the Byzantine period.