Four excavation areas were opened (Fig. 1), revealing stone quarries (A, D), terraces (B) and two winepresses that were later adapted for use as a limekiln (C).
Area A (8.5 × 13.0 m; Figs. 2, 3). Two phases were discerned. In the first phase, two adjacent hewn cavities of identical size (Loci 102, 103; 1.2 × 1.2 m, depth 1.5 m) were linked to each other by a passage; L103 was plastered.  The cavities, whose function is unclear, were devoid of artifacts. Based on the plaster that survived in Cavity 103, they were probably used for storing liquids. In the second phase, the area was used as a stone quarry (Loci 100, 101, 104) and even though it was quite shallow (c. 0.5 m), the quarrying destroyed the ceilings and facades of the cavities of the first phase.
Area B (5 × 7 m; Figs. 4, 5). A sounding was excavated in the area opposite a terrace wall (W203). Brown soil fill (L200), which yielded potsherds mostly dating to the Hellenistic period that included bowls (Fig. 6:1, 2), storage jars (Fig. 6:3–5) and a lamp (Fig. 6:8), was exposed. A storage jar fragment from the Early Roman period (Fig. 6:7) was also found. Four coins were in the fill, three of which were identified: a coin from the time of Constantine II (330–335 CE, IAA 102267), an early Byzantine coin (IAA 102268) and an anonymous Umayyad follis from the Damascus mint that dates to the middle of the eighth century CE (IAA 102269). It seems that the terrace was built of soil fill that was brought to the site from an adjacent ruin and the contents of the terrace fill reflects the periods of occupation of that ruin. The excavation reached bedrock, descending from north to south, on which the terrace wall (W203) was founded, against which the soil fill and small stones (L201) rested. A considerable effort was invested in the construction of the terraces (height 2–3 m, width 10 m).
Area C (10 × 20 m). Four phases were discerned in this, the largest of the excavation areas.
Phase 1. A square cave, documented but not excavated, was exposed (L320; 4.5 × 5.0 m; Fig. 7). It was found half filled with alluvium and its walls showed evidence of hewing, but no kokhim or arcosolia were discerned. Its entrance was in the south and an ashlar-built wall (W322) abutted the eastern side of the entrance and most probably enclosed a built passage that led to the entry. A rectangular aperutre (0.3 × 0.8 m, height 1.0 m) was cut in the southeastern corner of the ceiling in the cave. The entrance to the cave was blocked by a fieldstone wall. A modern iron plate and a terrace soil fill (L313) were placed over this blockage in the last phase of the cave’s use.
Phase 2. Two adjacent winepresses, separated by a bedrock partition (width 1.10 m; Fig. 8), were hewn. The western part of the cave extended below the treading floor of the eastern winepress.
The eastern winepress (Fig. 9) included a square treading floor (L305; 4.5 × 4.5 m), which was paved with white mosaic and a collecting vat (L312; 1.2 × 1.2 m, depth 1.2 m) that had a rectangular settling pit (0.3 × 0.7 m, depth 0.3 m) in its northeastern corner. A secondary pressing installation was discovered to the north of the collecting vat. This was bedrock hewn (L319; 0.7 × 1.2 m, depth 0.35 m) and included a niche, cut in the southern wall of the winepress and a sedttling pit in its bottom. The northern wall of the winepress was hewn as a step (width 0.5 m) and a hewn channel in the wall (width 0.1 m) led to the ceiling of the cave to the north. The purpose of the channel and the stone step is unclear. The south side of the winepress was not preserved due to later damage. 
The western winepress was only survived by a square treading floor (L302; 5 × 5 m). It seems that its construction was never completed due to the quality of bedrock in the eastern part.
The winepress complex and the ceiling of the cave were close to surface and were covered with the soil of later terraces; therefore the ceramic finds were mixed, which made it impossible to date Phases 1 and 2. It seems that the winepresses were built in the Early Roman or Byzantine periods, but this is not certain.
Phase 3. The winepresses were canceled in this phase by the construction of a square limekiln (L318; 1.5 × 1.5 m) in the center of the eastern winepress’ treading floor whose mosaic floor was destroyed in its eastern part. Only the base of the limekiln survived. The winepress’ collecting vat, which was found filled with lime up to its rim, was used as the collecting pit for the lime. At the end of this phase the lime pit was sealed with a layer of small stones, which included fragments of a large jar.
In three other places (Loci 303, 309, 317), concentrations of ash, jar fragments and the remains of bones were found. The finds from this phase included fragments of bowls (Fig. 10:1–5), jars (Fig. 10:6–15) and a teapot spout (Fig. 10:16) that dated the end of the installation’s use to the Early Islamic period (seventh–eighth centuries CE). A single lamp fragment (Fig. 10:17) that dated to the Byzantine period was found. Glass vessels that also dated to the Early Islamic period (Fig. 11, see below) were found as well. A fragment of a tile recovered from L300 (?) was decorated with a stamped impression of a table amphora with raised S-shaped handles (Fig. 12) that dated to the Roman or Byzantine periods.
Phase 4. Farming terraces that were generally oriented north–south direction were built in this phase. The walls of the terraces stood 3–4 m high and the soil fill against them seems to have been similar in composition to that in Area B. A small retaining wall (W324) of small fieldstones was built perpendicular to the lines of the terraces, and lower walls were built in places where bedrock was high (W323). A fieldstone-built installation, most likely an oven (L317), was built next to W323 and was also ascribed to this phase.
Area D. Next to the terrace and and slightly west of Area B, a shallow stone quarry (L400; 4.5 × 4.5 m; Fig. 13) was revealed, covered by a soil fill similar to the terrace fill in Areas B and C. The quarry resembled the one in Area A and was located along the same terrace line. On the slope between Areas B and D, a tractor exposed the remains of an oval-shaped cave (L402; 2.0 × 2.5 m, depth 1.5 m), which contained soil fill that yielded numerous potsherds from the Early Roman period. It seems that this cave was intended for burial, but its quarrying was never completed. With the construction of the terraces the cave was blocked with soil.
Glass Artifacts
Natalya Katsnelson
A number of glass fragments were found in Area C (Loci 304, 309, 311, 319). The assemblage is extremely limited and includes two bottles or cups (Fig. 11:1, 2), a wine goblet (Fig. 11:3) and jars (Fig. 11:4–10). The fragments are small and poorly preserved; due to the calciferous coating their color is almost invisible. Most of the fragments are decorated with trails of the same color as the vessel (1, 6), yellow (2) and turquoise (7, 10). The decorations are arranged horizontally (1, 2), forming incised triangles (6–8) or a zigzag pattern (9, 10). The types of vessels and their decorations are characteristic of the Umayyad period, thus providing a tentative date for the limekiln phase (Phase 3).
Fragments 4–10 constitute an interesting group. While the decoration on No. 6 is made of translucent glass, Nos. 7–10 are decorated with a turquoise trail on the body of the vessel, which is translucent. Nos. 4 and 5 are fragments of either one or two different vessels. Jars of this type have a globular body and are decorated with a strip of zigzag threads around the upper part of the vessel and another strip below it forms irregular triangles that face up and down in an alternating pattern. These vessels appear in the region at the end of the seventh and the beginning of the eighth centuries CE. They are well known from their wide distribution in museums but are rarely found in archaeological excavations.