During August 2005, a salvage excavation was conducted at Nahal Kefira (Permit No. A-4583*; map ref. NIG 2094/6377; OIG 1594/1377). The excavation, carried out on behalf of the Antiquities Authority and financed by the Ministry of Defense, was directed by I. Zilberbod, with the assistance of R. Abu-Halaf (administration), T. Kornfeld (surveying), T. Sagiv (photography), E. Belashov (drafting) and I. Lidski (pottery drawing) .
Two areas, c. 300 m apart, were opened at the bottom of the ridge slope, to the north of the channel.
A winepress hewn in hard limestone bedrock was exposed (Figs. 1, 2). The treading floor (L2; 2.0 × 3.5 m) was rectangular and the height of its walls varied according to the gradient of the slope (northern wall 1.5 m; southern wall c. 0.05 m). On the floor near the northern wall was an elliptical depression (length 0.4 m, max. width 0.2 m, depth 0.2 m), which seems to have been a natural karstic hollow that may have functioned as a ‘tapuah’ where the grape skins were collected during the pressing process. No plaster was traced on the surface of the treading floor. Three niches, 0.2–0.3 m apart (Loci 3–5; width 0.35–0.40, height 0.55–0.60 m, depth 0.10–0.15 m), were hewn c. 0.5 above the floor in the middle of the northern wall. The middle niche was cut deeper (L4; depth 0.4 m) in a later phase. A cupmark (L9; diam. 0.15 m, depth 0.06 m) that may also be a karstic depression was located outside of and adjacent to the southwestern corner of the treading floor. It is not certain if the cupmark was used for setting a jar. A shallow channel (L11; length c. 0.25 m, width c. 0.15 m, depth c. 0.1 m) conveyed the must from the southwestern corner of the treading floor to a rectangular settling vat that had rounded corners (L8; 0.9 × 1.3 m, depth 1 m). An oval settling pit (L14; 0.55 × 0.90 m, depth 0.25 m) was hewn in the western side of the vat’s floor. The must flowed from the settling vat to the collecting vat (L16) via a perforation (diam. 7 cm, length 15 cm) that was hewn 0.4 m below the top of the settling vat’s northern wall. The two vats were also linked by way of an open channel (L10; length c. 0.1 m, width c. 0.1 m, depth c. 0.15 m). It is impossible to know whether the perforation and the channel belonged to different phases or were operated at the same time on an alternating basis. A shallow round depression that joined the settling vat by a gutter (L13; diam. 0.25 m, depth 0.15 m) was hewn in bedrock surface above the northwestern corner of the vat. The depression was probably intended for placing a jar. A narrow shallow channel (L12; length c. 0.2 m, width c. 0.1 m, depth 0.1 m) was used to convey the must from the treading floor directly to the collecting vat. It is presumed that this process occurred during the pressing stage, following the treading, when the grape skins were wrapped in sheets of cloth that filtered the must and thus, it was unnecessary to pass it through the settling vat.
The collecting vat, which was hewn north of the settling vat, had a bell-shaped outline (L16; 1.2 × 1.4 m, depth 2 m). Sections of yellowish white plaster were traced on its walls and floor and an elliptical settling pit (L18; 0.30 × 0.75 m, depth 0.2 m) was hewn in the southern part of its floor.
The bottom of the vat was reached by way of an arched recess, where one could place one’s foot (L22; width 0.3, height 0.2 m, depth 0.15 m) and a column that stood 0.6 m below it in the western half of the vat (L23; trapezoidal cross-section 0.25 × 0.40 m, height 0.6 m). One could also stand on the column while drawing the must from the vat below.
Potsherds from the Early Roman period were discovered in the fill that had accumulated in the collecting vat, including cooking pots (Fig. 3:1–3) and jugs (Fig. 3:4–9); several body fragments that may date to Iron II were found in the lower part of the vat, close to the floor.
Due to typological considerations, it is suggested that during Iron II, the winepress was originally part of the agricultural activity that was practiced in the environs of the biblical city of Kefira whose tell is located east of the excavation. It seems that the winepress was preserved in its entirety and was reused during the Second Temple period, as indicated by the ceramic finds.
A small quarry, c. 7 m east of the winepress (L1; 1.5 × 4.0 m), was cleaned. Remains of rock-cut channels and the negative of a row of six rectangular stones were discerned. Part of another quarry (L6; c. 3 × 4 m), which had remains of quarrying channels and the negatives of stones that had been detached (depth c. 0.25 m), was exposed close to the northwestern side of the winepress. It is impossible to know whether the winepress predated the quarries or visa versa and it is also likely that the installations were contemporary.
A road, aligned east–west and a c. 25 m long section of it (width 2.5–5.0 m; Figs. 4, 5) that was preserved along the surface, were discovered. A probe trench (Loci 19–21; 1.5 × 6.5 m) excavated across the road exposed the road bedding, which consisted of tamped-soil fill (max. thickness 1.2 m) that was deposited on a natural bedrock terrace. The fill was supported by a southern (W1) and a northern (W2) walls that were built of large fieldstones with small stones among them and together, formed a terrace. The tops of the retaining walls served as curbstones and protruded c. 0.5–0.9 m above surface. Worn potsherds from the Early Roman period were recovered from the fill in the probe trench.