Area A
Winepress. A rock-hewn winepress was exposed along the western part of a hard nari terrace. It consisted of a large plastered treading floor, which sloped to the south toward a settling pit, and a collecting vat. It was also equipped with a surface for extracting the must by means of a beam and rope type press and another collecting vat for the extracted output (Figs. 3, 4).

The winepress had an almost square treading floor (L116; 3.8 x 4.5 m, c. 17.7 sq m) with hewn sides (height 0.1–1.6 m). A plastered surface (L133) discovered to the north of the treading floor was intended for placing baskets with grape waste. The surface was surrounded on the north, west and east by rock-hewn sides (max. height 1.6 m) and it sloped south toward a small plastered collecting vat (L105) that stored the must produced from pressing the grape pomace (0.58 x 0.60 m, depth 0.7 m, capacity 224 liters). A hewn trapezoidal niche in the northern side of the plastered surface was intended for anchoring a horizontal beam that exerted pressure on the baskets (Figs. 3: Section 1-1, 5). A wooden plank was installed in a hewn rectangular socket (L112; see Fig. 3: Section 1-1) in the middle of the treading floor for fastening a rope that was tied to the free end of the press beam and was tightened to exert extra pressure on the baskets. A winepress with a similar system for extracting must was discovered at Arbel, dating to the Roman and Byzantine periods (E. Ayalon 1988. Agricultural and Work Installations at Ancient Arbel. In Z. Ilan and A. Izdarechet (eds.) Arbel – An Ancient Jewish Settlement in the Eastern Galilee, pp. 172–177; R. Frankel and E. Ayalon 1988. Vineyards, Wine Presses and Wine in Antiquity. Tel Aviv, p. 45). A large stone that may have been used in an additional extraction stage was discovered above Collecting Vat 105.

A channel leading to a settling pit (L109; 0.75 x 0.85 m, depth 0.9 m, capacity c. 333 liters) was hewn in the southern part of the treading floor. A round sump was hewn in the floor of the settling pit, into which the grape pomace was meant to sink, facilitating the collection of the remainder of the must. A narrow gutter hewn in the bedrock wall, separating the settling pit from the collecting vat, conveyed the must flow to the collecting vat (Fig. 3: Section 3-3).

The collecting vat (L110; 0.95 x 1.37 m, depth 1.45 m, capacity c. 1,887 liters) was hewn in nari bedrock and its walls were lined with a thin layer of plaster. Two gray plaster floors, upper (L119) and lower (L131), were discovered at the bottom of the vat, indicating at least two usage phases. Floor 131 was placed directly on top of the bedrock and hard gray soil was deposited above it as foundation of Floor 119; potsherds dating to the Roman period were found in the soil.

Several depressions were hewn above the edge of the collecting vat; they were probably used to raise a shelter above the installation, and one depression was suitable for a standing jar. Above the southwestern corner of the vat was a hewn installation (L111), composed of a surface (diam. 0.2 m, depth 0.23 m), a small collecting vat (diam. 0.3 m) and a narrow channel that connected them. The installation was probably intended for extracting liquid (spice or oil) that was mixed with the must for the purpose of adding flavor. A similar installation (L128) was hewn in the bedrock surface south of the vat.

Other depressions were hewn west of the winepress. These were probably used for storing marl or crushed chalk that were spread on the grape clusters in order to absorb the dirt in the must, to neutralize its acidity and improve its clarity.

The eastern wall of the treading floor was deliberately broken, indicating that rain water was probably stored in the collecting vat after the winepress was no longer in use;  it also increased the drainage basin area around the collecting vat.


Reconstruction of the Winepress Usage. Initially, the grapes were trodden in the floor and the must flowed into the rectangular collecting vat via the hewn channel. The grape pomace—grape pulp and pips—was separated in the settling pit from the must by means of precipitation and floatation. The must was collected in the collecting vat for primary fermentation. The grape pomace was transferred to baskets that were placed on the high press floor and were squeezed once more by the beam and rope press. The beam exerted pressure when a stick with the rope wrapped around it was turned next to the beam. The must from the secondary pressing was collected in the square collecting vat (L105) and its overflow drained into the settling pit (L109). Presumably, the large stone found above Vat 105 was placed on the remainder of the grape skins for tertiary pressing. After the primary fermentation in the collecting vat, the must was probably transferred in jars to Cave 2 (below) for further fermentation.


Cave 2. A large natural cave, enlarged by quarrying, was exposed on the bedrock terrace, c. 8 m west of the winepress. Its upper part was hewn in the nari bedrock and its lower part—in the qirton. A hewn pit was discovered in the floor of the cave. Both cave and pit were not entirely excavated due to cave fever that infected the workmen during the excavation. Four superposed lime and tamped-earth floors (L120–L123; Figs. 6: Section 1-1, 7) were exposed in a deep trench that was excavated perpendicular to the cave’s entrance. Horizontal layers of black ash, fragments of clay ovens, sheep/goat bones and a few potsherds were found on the floors. The floors were set above a large amount of collapse that resulted from the crumbling ceiling of the cave. It seems that these floor levels were used by shepherds and farmers who tilled their land in the valley south of the site.

An elliptical pit (diam. 2.6 m; Fig. 8) hewn in the qirton bedrock was exposed 4 m below the ceiling. The pit was deliberately blocked with soil fill (L124, L129, L130), large fieldstones and large dressed stones that were discarded into it. The fill was cleared to a depth of 1.5 m; however, the excavation was not completed. Several potsherds were found in the stone collapse. These dated to the end of the first and the beginning of the second centuries CE, similar to those found on the floor of the collecting vat, indicating the time when the pit ceased to be used. The hewn pit was probably utilized as a winery for the adjacent winepress. Pits or caves near many winepresses were used for storing jars with must for further fermentation and their proximity to the winepresses prevented the shaking of the must before it turns into wine damages its quality.

Area B

Cave 3 is a small natural cave that was exposed on the bedrock terrace, east of Cave 2 (Fig. 9). The cave was completely excavated. A layer of crushed earth (L202) was exposed on its bedrock floor and above the soil were ash remains and a few potsherds (L203). It seems that the cave served as an encampment for shepherds and farmers, taking shelter inside, from the Roman period until modern times.

The ceramic finds were discovered mostly in Area A and included mortaria bowls (Fig. 10:1, 2), cooking pots (Fig. 10:3, 4) and an unguentarium (Fig. 10:5).

The mortaria bowls are shallow and have a thick, curved wall. Mortaria fragments were found in the foundation of the collecting vat (L119) and in the fill of the hewn pit in Cave 2 (L129); all the mortaria have a broad ledge rim that is grooved on top and a tapered ridge above the groove is pointed upward. Based on comparative mortaria, a flat base and a wide spout that slants downward from the rim can be reconstructed. The clay is brown and rich in quartz and basalt temper, and the surface is greenish and light colored. These mortaria are designated as ‘Early Roman Mortarium Type B’ based on Riley’s typology (Riley J.A. 1979. Coarse Pottery. In J.A. Lloyd [ed.], Excavations at Sidi Khrebish, Benghazi [Berenice], Vol. II, Tripoli , Figs. 112:669–671; 113:678–683). Similar mortaria were found in Stratum V of the Crowne Plaza Hotel compound in Jerusalem, attributed to the end of the first and the beginning of the second centuries CE, and in Stratum IV, which parallels the Tenth Legion’s pottery kilns in Binyene Ha-Umma (Levy D. and Beeri R. 2010. Excavation at the Crowne Plaza Hotel [Binyanei Ha-Uma]. Innovations in the Archaeology of Jerusalem and its Environs, Vol. IV. Jerusalem, pp 72–74; Magness, J. 2005. The Roman Legionary Pottery. In B. Arubas and H. Goldfus [eds.], Excavations on the Site of the Jerusalem International Convention Center [Binyene Ha-Umma]. A Settlement of the Late First to Second Temple Period, the Tenth Legion's Kilnworks, and a Byzantine Monastic Complex. The Pottery and Other Small Finds [Journal of Roman Archaeology 60]. Portsmouth. P. 96, Fig. 30). Magness identified this mortaria type with Roman sites in Italy and in Western Europe, dating from the first and second centuries CE; she cited comparisons from Tel Dor, Caesarea and Jerusalem in Israel (Magness 2005, p. 97).

The two cooking-pot rims were found in the fill of the hewn pit in Cave 2. Both have a triangular everted rim, a convex neck and a pair of strap handles drawn from rim to shoulder. Such cooking pots have a round body and a wide round base; their clay is pale red and well-fired. They were very common to the Jerusalem region at the end of the first century BCE and in the first century CE; they were found in Strata VI and VII of the Crowne Plaza Hotel excavation and in layers of the Binyene Ha-Umma excavation, which dated from the end of the first century BCE to the year 70 CE (Berlin A. 2005. Pottery and Pottery Production in the Second Temple Period. In B. Arubas and H. Goldfus [eds.] Excavations on the Site of the Jerusalem International Convention Center [Binyene Ha-umma]. A Settlement of the Late First to Second Temple Period, the Tenth Legion's Kilnworks, and a Byzantine Monastic Complex. The Pottery and Other Small Finds [Journal of Roman Archaeology 60]. Portsmouth. Pp. 36–38, Fig. 4).
A flat base of a delicate thin-walled unguentarium was retrieved from the pit’s fill in Cave 2.
The clay is light colored, well-levigated and fired. Based on comparisons, it possibly had a bulbous-shaped body. This type is classified as Shape 2 by Stojanović and is mostly common to Hellenistic and Early Roman burial assemblages from the second and first centuries BCE (Anderson-Stojanović V. 1987. The Chronology and Function of Ceramic Unguentaria. AJA 91: 105–122, Figs. 1: f; 7:2).

The appearance of mortaria that is associated with activities of the Tenth Legion in Jerusalem in the floor foundation of the second phase of the winepress indicates that the latter was last used after 70 CE. The potsherds discovered in the fill of the pit in Cave 2 are also ascribed to the end of the first–beginning of the second centuries CE, indicating that the pit was no longer used in this period. After the winepress ceased to be used, its eastern wall was deliberately broken and the treading floor was apparently exploited for collecting rainfall in the collecting vat. The latter’s function had terminated when an earthquake struck and split the winepress’ treading floor, raising its southeastern half above its northwestern half.