In September 2016, a salvage excavation was conducted south of Tel Yarmut (Permit No. A-7793; map ref. 19670–793/62265–360), prior to construction of a reservoir in Ramat Bet Shemesh D3. The excavation, on behalf of the Israel Antiquities Authority and funded by the Emek Ayalon Infrastructure and Project Management of the Ministry of Housing, was directed by Z. ‘Adawi, with the assistance of A. Hajian, Y. Shmidov and M. Kahan (surveying and drafting), M. Dinstein (field photography), S. Gendler (metal detection), I. Lidsky-Reznikov (pottery drawing), D. Levi and M. Birkenfeld (GPS), S. Kisilevitz, Y. Paz and A. de Groot (pottery identification), I. Reznitzky (metallurgy laboratory) and D.T. Ariel (numismatics).
The excavation took place in two areas (A, B; Fig. 1), on two hillocks c. 1.2 km apart on a hill west of Horbat Bet Natif and north of Horbat Qolad. In Area A, three field towers were excavated, as well as two winepresses, two clearance heaps and a limekiln. In Area B one limekiln was excavated.
The hill overlooks key sites in the Ramat Bet Shemesh area: Tel Yarmut to the north, Tel Socho to the south, Kh. Qeiyafa to the west and beyond it—Tel ‘Azeqa. The area was surveyed in the past, when numerous and varied remains were documented, including structures, field towers, winepresses, limekilns, cisterns, basins, cupmarks and clearance heaps (Dagan 2010:149–151, 279, 288–294, Sites 189, 353, 358–361, 370.1).
Field towers. Three field towers (A–C; Fig. 2) were excavated at the top of the hillock, overlooking distant slopes and agricultural plots. The towers were almost square in plan (3.5 × 3.5 m on average, preserved height c. 1 m), and were built of large fieldstones set in one row on the bedrock and surviving to a maximum height of two courses. The collapsed stones around them attest that they were originally much higher. Some of the stones may have been taken for secondary use in construction or for the lime industry in the nearby kilns. The three towers contained layers of fill (thickness 0.5–0.6 m) above the bedrock.
Tower A was filled with a layer of fieldstones, soil and quarry debris (L114; thickness 0.3–0.5 m; Figs. 3, 4). Tower B was filled with a layer of small fieldstones (L112; stone size 3–5 × 3–5 cm, thickness of layer 0.5–0.6 m; Figs. 5, 6). A body fragment of a jug from the Iron Age II was found in the fill (Fig. 7:3). Tower C was filled with a layer of soil and medium-size and large fieldstones (L120; stone size 0.2–0.4 × 0.3–0.5 m; Figs. 8, 9). The excavation around the towers (L100, L108, L113, L117) uncovered their foundations, but no entrance was found. The towers may have been entered by means of wooden ladders.
Winepresses. Two winepresses were discovered, one simple (A; Fig. 10) and the other complex (B; Figs. 5, 11). Winepress A, c. 20 m northeast of Tower A, was hewn in a bedrock outcrop; it comprised a rectangular treading floor (L105; depth c. 0.1 m) and a collecting vat (L102; width c. 0.65 m, depth 0.5 m).
Winepress B comprised a treading floor (L109; depth c. 0.1 m) and a collecting vat to its north (L106; depth c. 0.8 m). Cracks in the treading floor and the dislocation of a block of rock that separated the treading floor from the collecting vat were apparently the result of earth movement. To the northeast and the northwest of the collecting vat were two small vats (L104, L107; 0.40–0.70 × 0.45 m, depth c. 0.5 m) that were linked to Vat 106 by small channels (1, 2). An additional channel (3) linked Vat 107 to a slightly sunken bedrock surface to its west, where grapes could have been set. If so, the must that dripped from the grapes prior to treading flowed through Channel 3 to Vat 107. Another possibility is that the sunken bedrock surface was installed after the larger treading floor (L109) had cracked. The bedrock outcrop included two natural depressions (1, 2) and four cupmarks (1–4), which seem to predate the winepress. The components of the winepress and the bedrock surfaces were covered with alluvium. A body fragment of a jar from the Iron Age II (Fig. 7:4) was found in Vat 106, and Vat 107 yielded a coin from the fourth century CE (perhaps from 351–361 CE; IAA 149364). The potsherd and the coin were probably swept into the vats after the winepress went out of use, and thus do not help in dating the installment of winepress or its use.
Limekilns. Two limekilns were excavated (A, B). Limekiln A (Figs. 12–14) was ovoid; its lower part was built in a karstic cavity (L116; diam. c. 3.5 m, preserved depth c. 3.9 m), and its upper part consisted of two parallel rows of large fieldstones. A layer of small fieldstones was found inside the kiln (L121, L123; stone size c. 0.2–0.3 × 0.2 × 0.4 m, thickness of layer c. 2.5 m). On the southwestern side of the kiln was a channel (L132; length c. 3.5 m, width c. 0.7 m, height 1.2 m), which may have served to allow air to enter the kiln. It was built of two rows of fieldstones covered with large fieldstones (c. 0.4 × 0.9 m); its outer wall was set on the bedrock.
Limekiln B (L131; Figs. 15, 16), like Limekiln A, was ovoid; its lower part was built in a karstic cavity (L121; diam. and depth c. 3.5 m), and its upper part consisted of two parallel rows of large fieldstones. In the lower part of the kiln, along the walls of the cavity was a layer of small fieldstones (L130; stone size c. 0.2–0.3 × 0.2–0.4 m, thickness of layer c 2 m), apparently intended to be burned to produce lime. A channel on the northeast side of the kiln a channel was discovered (L133; length c. 2.5 m, width 0.6 m, height 0.9 m) built of two rows of fieldstones set in burned soil. No datable finds were discovered.
Clearance heaps. Five clearance heaps were identified (A–E; Fig. 2; Clearance heaps C and D do not appear on the plan), in various sizes, some surrounded by fieldstones.
Clearance Heap A (length c. 7 m, width c. 5 m, height c. 0.8 m; Fig. 3) was found in the northwestern part of the excavation area. It consists of small fieldstones (c. 5–20 cm) and was surrounded by a row of large fieldstones, some only partially preserved. The heap was covered with alluvial soil (L126). In the center of the heap a section was excavated (L128, width c. 1 m; length c. 7 m, height 0.8 m), which revealed no datable material. On the western side of the heap were two rows of fieldstones partly surrounding the heap. The inner row is apparently more ancient (W127). Pottery found in the accumulation above W127 dates mainly from the Iron Age II; it includes bowls from the ninth century BCE (Fig. 7:1, 2) and an unidentified bowl (Persian? Fig. 7:5).
Clearance Heap B (length c. 5 m, width c. 3 m, height 0.5 m; Fig. 17), in the southeastern part of the excavation area, is ovoid and surrounded by large and medium fieldstones. Two layers of fill were identified: the lower layer (L125; thickness 0.3 m) consisted of alluvium with a few fieldstones; the upper layer (L124; thickness 0.2 m) consisted of small fieldstones (stone size 0.1–0.2 m). Clearance Heap E (length c. 5 m, width 2–3 m; Fig. 17), like Heap B, was surrounded by large and medium fieldstones.
Clearance Heaps C and D (length c. 8 m, width 3–5 m, height 0.5 m; neither was excavated or drawn) were situated in the southeastern part of the excavation area, c. 20 m southwest of Tower C. Like Clearance Heap A, they were surrounded by large fieldstones.
The excavated remains are agricultural-industrial in nature. The lack of datable finds, except for one sherd in Tower B, one sherd in Collecting Vat 106 and several sherds in Clearance Heap A—mostly from the Iron Age—preclude any conclusions regarding the date and spatial context of the remains. However, it is possible that the finds from the Iron Age II reflect an early phase of activity in this area.
The nature and function of the towers are not sufficiently clear. Many such towers are known throughout the country. They have been dated to various periods, such as the Iron Age and the Hellenistic, Roman and Byzantine periods, based either on pottery discovered in them or the date of nearby sites (Shadman 2016:216–223). While all the towers at this site were similarly built, the fill within them is not uniform, and therefore it cannot be ruled out that they date from different periods. Their use is unclear, but they may have served to store agricultural products. Their proximity to the winepresses suggests that they were used to store jars for fermenting the must. However, their location on the hilltop, overlooking the slopes and the surrounding area—and except for one, with a clear view of central multi-period sites, such as Tel Yarmut, Kh. Qeiyafa, Tel ‘Azeqa and Tel Socho, as well as of nearby sites, such as Kh. Shumeila, Kh. edh-Dhiab and Horbat Qolad (Dagan 2010:235–285, Nos. 306, 326, 354)—may point to an association with one of these sites.