The present excavation (54 sq m) exposed a stone-paved courtyard with a small winepress and a plastered water cistern that stored run-off rain water collected from an adjacent building. These installations were built over and into the sloping bedrock surface that had previously been exploited as a stone quarry (Fig. 2). On the basis of the limited potsherds, it is suggested that these activities took place in the Middle–Late Roman period, after which time they fell out of use.
The bedrock slope was originally used as a quarry, the sloping nari bedrock surface exhibiting some regularly cut-out steps, as well as faint negative markings of large ashlar blocks. The size of the cut-out blocks was fairly standard (length 0.6–0.9 m, width 0.4–0.5 m, height 0.4 m), with some variation. The quarrying was done in a regular fashion and the cutting lines were orientated northwest-southeast and northeast-southwest (Fig. 3) Two parallel cut-out rock step lines and a perpendicular rock-cut step line connecting them (L101), may have been intentionally cut out to serve secondarily as foundation courses for the stone walls of a room.
Subsequent to the quarrying activity, the excavated area served as a courtyard, which incorporated various installations, including a rain-water collection system and a winepress. The cut-out area between the two parallel bedrock steps (L101) was filled with rows of large stones that may have been waste stones from the quarrying activities; only one stone appears almost complete (0.4×0.5×0.9 m). These stone rows seem to be simply backfill of the quarried steps rather than a floor surface. However, the area further north (L102) was paved with variously-shaped, flat stone slabs that were carefully laid with hardly any intervening spaces (Fig. 4). Two large square stones lying directly on the stone-slab floor, 0.7 m apart, may have been pillar bases for supporting a roof. A small plastered square basin (L106; 0.2×0.4 m, depth 0.4 m; Fig. 5) was built embedded into the floor, probably to serve as a rain-water collecting installation that would have received water from the roof via a pipe running down a no longer extant wall, to the floor. The collecting basin had two small exits: one leading into a narrow, square-profiled channel (L107; length 2.5 m, width 0.2 m, depth 0.15 m) that was plastered integrally with the collecting basin, and another leading into a terracotta pipe (length 2 m, diam. 0.1 m, width of walls 0.5 cm). The pipe, installed at a slightly higher level than the channel, was encased in a covering of small stones and mortar (L105; Fig. 6). The channel and the pipe were parallel and both led the water off in a westerly direction, the pipe leading into a plastered cistern, while the channel may have diverted the water to an adjacent water installation located further north, beyond the excavation limits (see below). It seems that some of the sediment would have been trapped in the small collecting basin, prior to the water flowing on into the open channel, whereas the higher elevation of the pipe would ensure that water flowing in it was clean. The rectangular rock-hewn water cistern (L109; internal measurements 1.4×1.9 m, height 0.9 m; Fig. 7), located in the northwest corner of the excavated square, was internally lined with a row of large ashlar nari blocks whose upper surface was partially exposed (c. 0.4×0.6 m); the stones probably came from the local quarry. A small step (0.38×0.54 m, height 0.12 m) in the northeast corner of the cistern’s floor and the inner surfaces of the entire installation were integrally plastered with a thick (3 cm) layer of soft white plaster, covered with a thin (0.5 cm) layer of hard pink plaster that was finished with carefully contoured corners. The two plaster layers formed a waterproof coating. The contour of a plastered wall traced in the upper part of the northern cistern wall (c. 0.4 m above floor) indicated that another channel (width 0.44 m) was leading off into an additional water installation, which was possibly large and bell-shaped, and must have lain beyond the excavation limits. This installation would have collected overflow water from the rectangularcistern, as well as from the small open channel (Fig. 8). Cistern 109 was found full of large stones that must have been thrown in when it was no longer in use. Among these stones were a large ashlar block (0.4×0.4×0.6 m) and a cylindrical pillar segment (diam. 0.4 m, height 0.62 m), which were not removed from the cistern in the course of the excavation, due to their size and weight. The few potsherds recovered from the cistern dated to the Late Roman or Early Byzantine periods; these probably reflect the cessation date of the cistern’s use. A single body sherd extracted from the wall plaster dated to the Roman period.
In the southwestern part of the square, a small winepress was exposed. It consisted of a slightly sloping bedrock treading surface (L110; 1.8×3.0 m), in which the negative markings of the quarried stones were faintly visible (Figs. 9, 10). The treading surface led the pressed juice northward into an L-shaped collecting basin (L104; 0.90×0.93 m) with a shallow circular sump (diam. 0.38 m, depth 0.18 m) in its northeastern corner for collecting residue. The collecting basin was lined with a soft white plaster layer, covered with a thin hard pinkish layer, in the same fashion as the water cistern, and here too, as in the cistern, the corners were contoured. Two small plastered surfaces around the collecting basin, on its western and northern sides, were damaged by the preliminary backhoe trenches. The winepress was contemporary to the stone-paved floor (L102), as the collecting basin was carefully set into it.
The relatively small amount of potsherds retrieved from the excavation dated the courtyard and its installations to the Middle to Late Roman periods, including cooking pots of Kefar Hananya Form 3B (Fig. 11:1) and Form 4C (Fig. 11:2), a mortarium basin of a buff-colored hard fabric (Fig. 11:3), a few cooking bowls of Kefar Hananya Form 1E (Fig. 11:4, 5), a few Early Byzantine potsherds, probably dating to the period when the courtyard was abandoned, including a Late Roman Red Ware Cypriote bowl, Hayes Form 9 (Fig. 11:6) and a Byzantine open cooking casserole with horizontal handles (Fig. 11:7).