The excavation took place in the western part of the neighborhood of Gillo, on a slope descending west toward Naal Gillo and north toward Naal Refa’im. Five areas were opened (A–E; Fig. 2); three secondary areas were opened in Areas A, C and E. Four rock-cut caves (1–4) were discovered, as well as two limekilns, a winepress, a quarry, a field of shaft tombs, an agricultural terrace and a field tower. In past surveys in the vicinity of orbat Bet Arza, east of the current excavation, dozens of agricultural sites were documented, featuring terraces, winepresses, field towers, agricultural installations and cisterns, burial caves and structures (Fig. 1.1; Kloner 2000:53–77). Numerous excavations have been undertaken over recent decades in this area. A few tombs from an Intermediate Bronze Age burial ground found near the current excavation were probably associated with the arry of agricultural settlements in the area during this period and the Middle Bronze Age II in and around the Naḥal Refa’im basin (Edelstein, Milevski and Aurant 1998; Zilberbod and Be’eri 2017). On the hill opposite the excavation, in the southern part of the Mala (Manaat) neighborhood, remains of Middle Bronze Age rock-cut shaft tombs were uncovered in the early 1990s (Fig. 1:2; Zehavi 1994). A more recent excavation at the site revealed additional shaft tombs from this period (Kisilevitz and Turgeman-Yaffe 2018, and see therein details of surveys and excavations carried out in the vicinity). In 2000, a burial cave (Cave 900) from the Intermediate and Middle Bronze Ages was excavated on the northern bank of Nahal Refa’im (Fig. 1:3; Weksler-Bdolah 2017). Settlement remains from the Intermediate–Middle Bronze Ages were uncovered on a hill near the Biblical Zoo (Fig. 1:4; Eisenberg 1993; Eirikh-Rose 2015). The remains of another Intermediate–Middle Bronze Age settlement were uncovered at Manahat, near the Malḥa Mall (Fig. 1:5; Edelstein 1993a; Edelstein, Milevski and Aurant 1998). A Middle Bronze Age temple was uncovered on the slope of the village of Walajeh and additional shaft tombs were identified (Fig. 1:6; Ein Mor 2011). Additional Intermediate–Middle Bronze Ages cemeteries were uncovered in the Holyland compound, on hills near the central settlements whose remains were uncovered at Manaḥat and in Naḥal Refa’im (Fig. 1:7; Zelinger and Golani 2005; Greenhut, Milevski and Agha 2008). A recent excavation in Holyland Park, near the Holyland compound, uncovered an underground dwelling complex from the Chalcolithic period and quarries and rock-cuttings from the Roman period, as well as a building from the Ottoman period which lay near the remains of the village of Malha (Milevski and ‘Ad 2019). At Khirbat er-Ras, north of the site, settlements that existed from Iron Age II to the Hellenistic period were discovered (Fig. 1:8; Edelstein 2000; Gadot 2011). A villa from the Roman period was uncovered at ‘En Ya‘al and remains of a settlement from the Hellenistic to the Early Islamic periods (Fig. 1:9; Edelstein 1993b; Avner 2015).

Area A
Cave 1 (A1; L100; Fig. 3) was hewn on a bedrock terrace (height c. 1.7 m). The cave had a square, rock-cut entrance (c. 0.8 × 0.8 m, depth 1.1–1.3 m); it seems that its hewing had not been completed (Fig. 4). On the inner side of the entrance was a hewn niche (width c. 0.3 m, depth c. 0.2 m; height 0.75 m). Other than a fragment of the back of a bronze dagger or awl (Fig. 5) retrieved in the topsoil, there were no other finds.
Cave 2 (A2; L126; c. 2.50 × 2.65 m, height c. 1.7 m; Fig. 6) was cut into the abovementioned terrace. A rectangular entrance (0.8 × 1.0 m; Fig. 7) facing northwest was hewn in its façade. Its floor (L126) was made of a layer of crushed limestone; a probe (L128; 1.0 × 1.5 m, depth 0.4 m; Fig. 8) cut into the floor revealed no datable finds.
Terrace wall (A2; W127; length 6 m, width 0.7 m, height 0.8 m; Fig. 6). A retaining wall of an agricultural terrace was found south of Cave 2; it was built of two rows of large to medium-sized fieldstones interspersed with small stones and preserved to a height of three courses. The wall ran along a southwest–northeast and then turns to the southeast.
Winepress (A3; Figs. 9. 10). The winepress was cut into hard limestone. Its western part was damaged by development work prior to the excavation. It featured a square treading floor (L152; c. 3.9–4.0 × 4.0 m, depth 0.2–0.5 m), three collecting vats (L151—c. 1 × 1 m, depth c. 0.5 m; L153—diam. And depth c. 1 m; L154—diam. c. 1.5 m, depth 1.7 m), a small cupmark (L157) and two channels (L156, L160) east and west of the treading floor. There were patches of plaster along the northern, southern and eastern walls of the treading floor, indicating that the surface was once plastered. A shallow channel connected the treading floor to collecting Vat 153, at the bottom which was a sump. Collecting Vat 153 was connected to Collecting Vats 151 and 154 by small channels cut into the upper thirds of its sides. A protruding small step was hewn in the upper part of the wall of Collecting Vat 154 (Fig. 11), and a sump (L155; diam. c. 0.7 m, depth 0.3 m) was hewn in the vat’s floor. Cupmark 157 was found north of the three collecting vats and was probably used to set jars filled with must. Channel 156 (exposed length c. 7 m, width c. 0.5 m, max. depth 0.65 m) was hewn west of the winepress; its use and its association with the winepress are not clear. Channel 160 (length c. 7.5 m, width c. 0.2–0.6 m, depth c. 0.5 m) was hewn east of the winepress, apparently utilizing a natural split in the bedrock. This channel was connected to the treading surface by a small channel in the northeastern corner of the surface, and it may have been used to drain rainwater from the treading floor. A single fragment of a jar from the Early Roman period (not drawn) was found in the silt that covered collecting Vat 153.
Area B (Figs. 12, 13)
Remains of a rock-cut limekiln (diam. c. 3.5 m) were unearthed. It was lined with medium-sized fieldstones (W206), and an enclosing wall (W207; diam. c. 7 m), built of one course of large fieldstones set on the surface, delineated it. The southern half of the limekiln was excavated (L212). A layer of burnt stones and patches of pure white limestone was found at the bottom of the limekiln, near its floor; no datable finds were recovered. A rounded area (diam. c. 1 m) unearthed west of the kiln contained a yellowish-white layer of soil (L211), the nature of which is unclear.
Area C
Limekiln (C1; L242, L243; Fig. 14). The kiln was built inside a rock-cut rounded space that was probably used earlier as a plastered cistern (Fig. 15). Only the northern half was preserved, while its southern half had been completely demolished by mechanical equipment. The kiln was lined with stones (W227; inner diam. c. 3.5 m, outer diam. c. 4 m). A ventilation channel on the western side of the limekiln allowed the flow of air into it; the channel was dismantled during the excavation. The kiln revealed no datable finds, but sherds from the Ottoman period (not drawn) were found on the surface to its south.
Field tower (C2; Fig. 16). A partially preserved rectangular field tower was documented near the kiln. It was built of three walls of large fieldstones and a fourth wall, the northern one, utilized the bedrock of the slope. A rectangular, arched entrance was set into the northeastern corner of the tower (Fig. 17).
Cave 3 (C2; Figs. 18, 19). This irregular space was hewn in the hard limestone (L235; 3.0 × 3.3 m, height c. 1.3 m). The floor of the cave was leveled bedrock; no datable finds were discovered.
Area D
Quarry (L376; c. 3.0 × 3.3 m, height c. 1.3 m; Figs. 20, 21). This was a two-level quarry. It appears that large chunks of quarried rocks (c. 0.5 × 2.0 m) were divided into smaller pieces and worked into building blocks on the spot.
Cave 4 (Fig. 22). This irregular rock-cut space was hewn into hard limestone. Most the cave was damaged by mechanical equipment during work prior to the excavation; no datable finds were discovered.
Area E

This area is a kind of strip (length c. 180 m, width 6–10 m) below the top of the slope descending northward to Naal Refa’im, and it is a continuation of the Intermediate Bronze Age cemetery in which more than 20 shaft tombs were discovered in 2014 (Zilberbod and Be’eri 2017). Three tomb clusters were sampled in the excavation (Areas E1–E3; Fig. 2), at a distance from each other, revealing at least 24 shaft tombs cut in soft chalk.

In Area E1 (c. 5 × 25 m, c. 150 sq m; Fig. 23, 24) remains were uncovered of at least seven tombs (L427, L428, L430, L434, L435, L454, L456). Tomb` 430 yielded remains of two amphoriskoi from the Intermediate Bronze Age (Fig. 25:1 [B4027], 2 [B4026]).

Twelve excavation squares (4 × 4 m; Figs. 26–28) were opened in Area E2, revealing at least 15 tombs (L432, L437–L450). Four intact pottery vessels were found in Tomb 446—two jars (Fig. 29:1 [B4460], 2 [B4461]), an amphoriskos (Fig. 29:3 [B4462]) and a four-spouted lamp (Fig. 29:4 [B4463]), dating to the Intermediate Bronze Age.

Two shaft graves were discovered in Area E3, (Figs. 30, 31). One had a round opening (L451), probably the entry shaft; the other, unearthed in the 2014 excavation, had an entry shaft and an adjacent room whose ceiling was not preserved (L452, L453).

The excavation of the three grave clusters revealed an almost identical grave plan: an ovoid or round vertical entry shaft, at the bottom of which was a small horizontal opening (0.6 × 0.6 m) leading to an ovoid burial chamber (1.5–2.0 × 1.5–2.0 m). The upper part of the shaft was cut through nari (to a depth of c. 1.5 m) and its lower part penetrated soft chalk. The opening in the lower part of the shaft was sealed with a stone that blocked the passage between the shaft and the burial chamber. Only a few of the tombs were fully excavated; these revealed no human bones. Similarly, no human skeletal remains were discovered when the topsoil layer and vegetation were removed from the area. The paucity of such remains may be due to the limited excavation of the tombs, but it could be that the bones they contained did not survive. The excavation in Area E was stopped due to the opposition of religious elements.

In December 2015, surface debris and vegetation were removed from the entire strip of Area E and beyond it (length c. 225 m, width 6–12 m; Fig. 32) in preparation of a salvage excavation that did not take place (Permit No. A-7597). The work was carried out using mechanical equipment and with the help of laborers. Some 114 round or ovoid rock-cuttings that were documented were parts of shaft tombs. The upper part of the rock-cuttings was razed, probably by work carried out by the Jewish National Fund work in the past. Due to the density of the tombs, it was difficult to identify their parts—shafts or burial chambers; however, it may be assumed that any two adjacent rounded rock-cuttings belong to the shaft and burial chamber of the same tomb.


The remains uncovered in the excavation were part of the agricultural hinterland of Jerusalem over the generations. The shaft tombs from the Intermediate Bronze Age are the earliest remains unearthed, and they appear to have been part of a partially uncovered large cemetery of a settlement from that period in the area. The latest remains, including the limekiln and the field tower, were likely used by the inhabitants of the nearby villages (Beit afafa, Sharafat, Mala, Walajeh and Beit Jalla) during the Ottoman period and were part of extensive agricultural undertakings in the area (for more about the agricultural activities in the area, see Kisilevitz and Turgeman-Yaffe 2018; Shor 2019).