Rock-cut and built agricultural installations were identified and excavated—winepresses, pressing installations, cupmarks, farming terraces, fences and field walls, stone clearance heaps, watchman’s huts, cisterns, a limekiln, a compound and a building—that attest to intensive agricultural activity in two main periods, the Late Hellenistic and the beginning of the Early Roman periods (first century BCE–first century CE), and the end of the Ottoman period. At least one installation is earlier, but this cannot be determined with certainty.

The Excavations
Five excavation areas (A–E) were opened in locations where ancient remains had been identified during the preliminary survey or after the removal of vegetation. Areas A–C were adjacent to each other, on the western slopes of the spur that descends to the Telem Valley. Area D was opened c. 150 m further up the spur, toward the northeast, on the slopes of the hidden hilltop, west of TP 377 and Area E was on the western side of the Telem Valley, c. 100 m west of Area B.
Area A (Fig. 2). Three rock-hewn installations, described below from south to north, were discovered.
Installation 1 (Fig. 3) was probably a winepress. A hewn channel (L2; length 1.75 m, width 0.15 m, depth 2–3 cm) extended westward from a round treading floor (L1; diam. 1.4 m, depth 0.2 m) to a round collecting vat (L22; diam. 0.77 m, depth 0.8 m) that was hewn into a natural crack in the split bedrock. The collecting vat was coated with bright white plaster (thickness 1–6 cm), composed of lime mixed with small gravel. These features are similar to installations from the Iron Age, but the finds do not provide any other evidence to corroborate this dating. The meager remains of a field wall (W2; length 1.2 m, width 0.9 m, height 1.2 m), oriented north–south, were preserved three courses high. The wall was built of medium fieldstones (0.2 × 0.3 × 0.4 m), unevenly set in place. It was founded on a layer of terra rossa soil (thickness 0.4 m) that had accumulated inside the collecting vat.
Installation 2. A shallow channel (L40; length 1.7 m, width 3 cm, depth 2–3 cm), oriented southeast-northwest, was hewn west of Collecting Vat 22, further down the bedrock surface. The channel led toward a shallow elliptical rock-cutting that became wider at the end of the channel and then, the liquid flowed into a large cupmark (L43; diam. 0.42 m, depth 0.3 m), hewn at a lower level. The cupmark was carefully fashioned and its rim was encircled by a hewn recess (diam. 0.38 m, depth 5 cm), which could accommodate a lid or used to secure a grating for squeezing.
Installation 3 (Fig. 4). A conical cupmark (L3; diam. 0.2 m, depth 0.12 m) was hewn c. 1.4 m east of Treading Floor 1. Four narrow, elongated cupmarks in a row were hewn c. 2.6 m northeast of Treading Floor 1. The cupmarks were separated by intervals of c. 0.2 m (L5—0.2 × 0.7 m, depth 1–10 cm; L6—0.23 × 0.70 m, depth 1–8 cm; L7—0.25 × 0.70 m, depth 1–10 cm). Cupmark 5 was split in two, apparently due to the shifting of the nari bedrock in this region. The uniformity of the cupmarks’ dimensions and the spaces between them show that they were used together. Although similar installations are known in the Judean Shephelah, especially in the area of Ramat Bet Shemesh, their function remains obscure. An irregular-shaped depression (L4; 0.25 × 0.30 m, depth 0.2 m) is hewn in bedrock further along the row of cupmarks, next to Cupmark 5. It is difficult to determine if it was natural or man-made, but it could have been used together with one of the adjacent installations. A rock-cut groove (L39; length 2.2 m, width 3 cm, depth 2–3 cm), which extended west from the cupmarks, was probably the initial outline of an installation that was ultimately never quarried.
Area B (Figs. 5, 6). A complex winepress, which consisted of two treading floors (L11, L13), two settling pits (L34, L35), a shared collecting vat (L14) and a rectangular recess (L26) that was surrounded by a bench (L9) and may have been used for storage, was discovered. A cave that was hewn inside Collecting Vat 14 evidently postdated the usage phase of the winepress. Sections of a hard light colored plaster surface, installed in Treading Floor 11 (2.9 × 3.2 m, depth 0.4 m), which sloped to the west, had survived next to its sides. A niche (L33, 0.6 × 0.8 m, height 0.3 m) was cut in the middle of the treading floor’s eastern side and a large, partially dressed fieldstone, was placed inside the niche, completing this side of the surface. The niche could have been utilized for securing a beam that was used in a secondary pressing process.
The liquid from Treading Floor 11 drained toward the northwestern corner of the surface and flowed via a narrow hewn tunnel (L32; diam. 0.1–0.2 m, length 0.2 m) to the southeastern corner of the rectangular Settling Pit 34 (0.75 × 1.25 m, depth 0.9 m); the tunnel entered the pit 0.2 m below its rim (Fig. 5: Section 2-2). A conical cupmark (L38; diam. 0.2 m, depth 0.12 m; Fig. 5: Section 3-3), in which a jar could be placed, was hewn on the edge of the pit’s western side. 
Settling Pit 34 was coated with white plaster, composed of lime mixed with fragments of flint, broken quartz crystals and small lumps of charcoal that served as bonding material. A rounded sump (L44; diam. 0.22 m, depth 0.1 m) was hewn in the southwestern corner of the pit’s floor. At the bottom of the pit and west of Sump 44, was a hewn tunnel (L15; diam. 0.10 m, length 0.25 m) that drained the must from the settling pit, probably into a jar (Fig. 5: Sections 3-3, 4-4); a groove was cut above the tunnel. The opening of the tunnel was found blocked with a small stone (diam. 9 cm, thickness 4 cm) that apparently served as a plug, which was forced into it. Although the stone was not worked, its dimensions matched the diameter of the opening. At the eastern end of the pit’s northern side was another hewn tunnel (L35; diam. 0.1–0.2 m, length 0.2 m; Fig. 5: Section 2-2) that opened slightly higher up, c. 0.15 m below the edge of the pit, and led to Collection Vat 14.
Treading Floor 13 (2.6 × 3.0 m, depth 0.4–0.6 m) was separated from Floor 11 by a thick bedrock wall (thickness 1 m). A plaster surface (L37; thickness 0.1 m; Fig. 7) was well preserved in the eastern and northern parts of the floor and its composition was identical to that of the plaster in Settling Pit 34. A short tunnel (L48; diam. 0.1 m, length 0.3 m) was hewn near the northern edge of the treading floor’s western side. The must flowed through this tunnel into the settling pit (L49) that was hewn above the northeastern corner of Collecting Vat 14 and had survived by a narrow section (0.1 × 0.8 m, depth 0.9 m). It seems that the original dimensions of this settling pit resembled those of Settling Pit 34. Most of the pit was removed after the winepress was no longer in use to facilitate the breaching of an opening into the cave on the eastern side of Collecting Vat 14 (below).
Collecting Vat 14 (2.20 × 2.45 m, estimated depth 2 m, estimated volume c. 12.75 cu m; Fig. 8) was shared by the two treading floors. An especially hard layer of plaster (thickness 2.5 cm), preserved on the vat’s southern wall, was identical in composition to the plaster on Treading Floor 13 and in Settling Pit 34. The vat, which was not excavated down to its bottom, contained rims of jars that were common to sites in Judah, dating from the end of the Hellenistic and the beginning of the Early Roman periods (Fig. 9).
A niche (L46; length 1.5 m, min. depth 1.5 m; Fig. 8) was hewn c. 0.2 m below the vat’s eastern edge and extended beneath Treading Floor 13; it was not excavated to its full depth. The outline of the niche was stepped so that its width was uneven; it was 0.5 m wide to a depth of 1 m from the ceiling and became c. 0.7 m wide below it. The sides of the niche were coated with plaster identical to that found on the southern side of Collecting Vat 14. After the installation was no longer in use, the eastern side of the collecting vat was breached c. 0.2 m below the opening to Niche 46. The carved irregular opening (0.6 × 0.7 m) led to a cave (L47; 2.5–4.0 × 3.0–6.0 m, height in excess of 1.8 m) shaped like an asymmetric trapezoid. The cave, which extended beneath Treading Floors 11 and 13, had a very straight ceiling and signs of diagonal rock-cutting along its sides.
Three shallow elliptical cupmarks (L16—0.25 × 0.30 m, depth 3 cm; L24—0.20 × 0.23 m, depth 2 cm; L25—0.16 × 0.20 m, depth 1–2 cm) were hewn in the bedrock surface next the western side of Collecting Vat 14; another cupmark (L12; 0.14 × 0.20 m, depth 0.16 m) was cut c. 2 m south of them. Numerous cupmarks of this type were identified throughout the site.
An installation whose use is unclear was hewn south of Treading Floor 11. It consisted of a rectangular depression (L26; 1.40 × 1.75, depth 0.3 m) that was cut inside an almost square hewn surface (L9; 2.6 × 2.9 m, depth 0.3–0.8 m). Bedrock surfaces or benches (width 0.5–0.7 m) were left on the eastern, northern and western sides of this installation, which could have been used to store agricultural produce prior to processing it.
A rectangular opening (0.7 × 0.8 m) that led into an underground cavity, was uncovered during the course of development work in a bedrock surface, c. 200 m east of Areas A and B. The cavity turned out to be natural and yielded no archaeological remains.
Area C (Figs. 10, 11). A winepress was exposed. It consisted of a square treading floor (L18; 2.8 × 2.9 m, depth 0.2–0.8 m), which was drained via a hewn channel (L19; diam. 3–12 cm, length 0.38 m), located c. 0.75 m from the southwestern corner of the treading floor, into a rectangular collecting vat (L20; 1.1 × 1.3 m, depth 1.52 m). An irregular-shaped sump (L27; 0.30 × 0.34 m, depth 0.35 m) was cut at the bottom northeastern corner of the collecting vat.
Area D (Figs. 12, 13). A winepress was discovered. The plan of its treading floor (L28) matched the outline of bedrock, hence it was asymmetric (max. length 3.4 m, width in center 2 m, depth 0.2 m). The treading floor was drained via a hewn channel into a rectangular collecting vat (L29; 0.5 × 0.6 m, depth 0.8 m), located next to the northwestern side of the floor.
Area E. A probe was manually excavated in a heap of stones, which turned out to be the collapse of a farming terrace stone wall that yielded no datable archaeological finds.
The Survey (Fig. 1)
Contemporary with the excavation, a partial survey of its region was conducted. Fifteen agricultural installations were discovered and together with those exposed in the excavation, they form one complex that can be divided into three clusters: on the lower western slope of the spur that descends from the hidden hilltop west of TP 377 (Areas A, B, E; Survey Site 1); at the top of the hidden hill (Areas C, D; Survey Sites 2–10); and on the hill of TP 377 (Survey Sites 11–15). Hereafter is the description of the survey sites:
(1) Winepress (Fig. 14) that was found filled with alluvium and included a rectangular treading floor (2.75 × 2.90 m) and an almost square collecting vat (0.95 × 1.00 m) to its west. The composition of the plaster in the collecting vat is identical to that of Treading Floor 13 and Collecting Vat 14 in Area B.
(2) A cupmark (diam. 0.4 m, depth 0.2 m) hewn on top of a leveled bedrock surface.
(3) Remains of a rock-hewn winepress that was split due to the shifting of the nari bedrock in this region. The winepress included a rectangular treading floor and a collecting vat, located to its south. The winepress’ poor state of preservation did not enable its measurement.
(4) A rectangular building (2.5 × 4.7 m, height 2.7 m; Figs. 15, 16) in whose construction both fieldstones and ashlars were used. A rectangular opening (width 0.5 m, height 1.3 m) built of ashlars, which were marked with short diagonal stone dressing, was set in the western side of the building. A support arch of well-dressed stones survived in the middle of the longitudinal axis of the building. The walls were built of two rows of stones, bound with light brown mortar. The corners of the building were especially well constructed. The building’s characteristics suggest it should be dated to the Ottoman period.
(5) A small bedrock-hewn pressing installation (Fig. 17), which consists of a small recess (0.2 × 0.2 m, depth 0.2 m) for anchoring a beam that is hewn c. 0.3 m above a rectangular basin (0.26 × 0.46 m, depth 5–50 cm).
(6) A bell-shaped rock-hewn cistern, accessed via a shaft with a round opening (diam. c. 1 m, depth c. 1.5 m). Rock-hewn footholds are visible at the lower part of the shaft’s side. The cistern itself was not examined.
(7) A rock-hewn winepress (Fig. 18) that consists of a rectangular treading floor (0.7 × 1.3 m, depth in excess of 1.2 m) and a collecting vat (0.7 × 1.3 m, depth in excess of 1.2 m) that is coated with white plaster. Hewn on the edge of the treading floor’s eastern side are two recesses, 0.2 m apart, which have a concave cross-section (width 0.15 m, depth 0.1 m).
(8) A rectangular enclosure (8.0 × 12.5 m; Fig. 19) that is partly built and its northern corners are rock-hewn. Four steps were cut in the northeastern corner. A square socket was carved at the level of the top step (0.8 m above the compound’s floor). Rock-cut leveled areas were visible west and east of the staircase. A pressing installation was hewn in the bedrock north of the northwestern corner. It included a cupmark (diam. 0.4 m) and a rectangular collecting basin (0.5 × 0.7 m) that was found filled with alluvium. Along the bedrock-hewn southern part of the western side was a field wall (length 3 m, width 0.5 m, preserved height 1.3 m), which most likely was meant to elevate this side. Meager building remains were identified in the eastern part of the compound, but thick vegetation precluded their documentation. The function of the apparently agricultural compound is unclear.
(9) A rock-hewn winepress on a bedrock surface that slopes from east to west, filled with silt. The treading floor (2.0 × 2.5 m), which was partially preserved because bedrock had split, drained into a collecting vat (0.8 × 1.6 m), located to its west.
(10) Two round watchman’s huts (diam. 1–3 m, average height 1.5 m; Fig. 20), which were built of a circumferential wall of small and medium fieldstones that was slightly inclined outward, creating a cone-like structure. A fill of indigenous soil mixed with small fieldstones was placed in the huts. Several non-diagnostic potsherds were recovered from the fill. Seven other watchman’s huts—four circular and three square—were scattered between fences and farming terraces near the top of the hill.
(11) Two long walls (total length c. 45 m), built of a row of large, partially dressed fieldstones, were identified at the top of the hill. They were probably used to delimit a square enclosure where a cistern (No. 12) and hewn basins were located; a potsherd scattering was found inside the enclosure.
(12) Opening (diam. 0.5 m) of a rock-hewn cistern.
(13) A large ruinous limekiln (diam. c. 4 m) with a stoke-hole that faced west.
(14) A rock-hewn winepress that consisted of a rectangular treading floor and a collecting vat to its west (not measured).
(15) A rock-hewn stepped corridor (1.0 × 3.5 m; Fig. 21), leading to a hewn opening in a large bedrock outcrop (min. height 2 m, width 1.5 m); the opening, which led into a hewn installation (a miqwe? a tomb?) was found filled with silt.
Throughout the survey, potsherds dating to the end of the Hellenistic and the beginning of the Early Roman periods (first century BCE–first century CE) were gathered from surface. These included a cooking pot (Fig. 22:1) and jars (22:2–7), similar to the pottery fragments from Collecting Vat 14 of the winepress in Area B. These ceramic finds, together with nine winepresses (only one of which may have predated the Hellenistic period), pressing installations and cupmarks, attest to the very intensive agricultural activity, whose essence was apparently wine production, during the Late Hellenistic and Early Roman periods. The installations may possibly have belonged to a farmstead that existed in the area, although no archaeological site or location with antiquities in the examined area is mentioned on the map of the British survey (Vol. XVII) or the Mandatory map (Map of Sha’ar Ha-Gāy [Sheet 13-15], 1:20,000).
Agricultural activity in the area was resumed in later periods, as evidenced by the farming terraces, fences and field walls, stone clearance heaps, watchman’s huts, limekiln and building. Indeed, an adjacent area denoted as an olive grove appears on the Mandatory map. It is possible that these cultivated plots belonged to one of the two Arab villages nearby: Deir Ayyub, southwest of the site, which existed until 1948 and Yalu to the north, which existed until 1967.