The site (c. 130 × 300 m) extends over the incline of a rocky spur on the southern slopes of the Menashe Hills and the northern bank of Nahal ‘Iron. Previous archaeological excavations at the site unearthed many ancient remains (Massarwa 2019, and see background and references therein). Four excavation areas were opened (A–D; 150 sq m; Fig. 2), encompassing the remains of about twenty quarries, five winepresses, two tombs, a limekiln, a cupmark, a cistern, a round rock-cut feature, a natural pit and natural fissures in the bedrock.
Area A
Quarries. Remains of quarries were uncovered over an area of about two dunams, where it was principally the nari rock overlying the soft chalk that was extracted. Hewn stones that remained in the area (0.2–0.4 × 0.4–0.6 × 0.5–0.8 m) were either of poor quality or cracked and broke during quarrying and, therefore, left on site. The quarries’ sizes varied with local conditions and an effort to maximize nari exploitation. Therefore, they included large quarries (max. dimensions 6 × 8 m; 100–102, 103A, 103B, 108, 110, 111; Fig. 3) and small quarries (105, 106, 109, 112, 116, 117; Fig. 4). The quarries are square, rectangular or polygonal, their corners are straight and they belong to the stepped-quarry type. First, the stone’s outer edge was marked; then, a severance channel was cut around it. The final detachment of the block was achieved from the lower part of the stone’s facade, thus producing the quarry’s stepped form. The quarrying marks show that various tools were used: A hammer and chisel that leave straight, upright sides; a pickax or pick (blade width 3–4 cm; Fig. 5) that leave rounded grooves and corners in the rock, an indication of circular motion caused by the swinging of the tool; and detachment spikes—rectangular iron rods inserted into regularly spaced holes along the stone block, exerting pressure until the block cracks horizontally (Fig. 6).
Only a few finds were recovered during the cleaning of the quarries. They consisted, predominantly, of body sherds and handles of bag-shaped jars dating from the third–fourth centuries CE.
Winepress 107. The winepress was hewn into the nari rock surface and bordered on the south and north with natural fissures. The winepress was adjusted to fit the rock face: on its eastern side, a square treading floor (c. 2 × 2 m, depth c. 0.2 m) slopes westward toward a collecting vat (0.6 × 1.2 m, depth c. 0.4 m) with a leveled floor and settling pit in its center (diam. c. 0.2 m, depth c. 0.2 m; Fig. 7). Two diagonal holes were drilled (c. 0.5 m apart), descending from the treading floor to the bottom of the collecting vat (Fig. 8). The holes were conical (outer hole diam. 0.2 m, inner hole diam. 7–8 cm), indicating that they were hewn from one side. On the treading floor near the southern hole, a kind of secondary, elliptical treading floor was recorded (hewn depth 3–5 cm), either for second treading of grape skins or for treading small quantities of grapes.
Natural Rock Fissures. Long, deep, east–west oriented fissures (width 0.7–1.2 m, max. depth 1.3 m; Fig. 9), 80–90 m long, were found at a uniform elevation on the southern margins of Area A on the spur’s margins. The fissures follow a winding route, and secondary north–south oriented fissures were seen to branch off it. At the base of the fissures, a dissolution notch was formed between the fissured nari crust and the calcareous chalk. The cracks were filled with alluvium, calcite encrusted pebbles and quarry debris. Since no evidence of human quarrying was found, the fissures may be due to slipping and fissuring of the nari crust, which formed on a thin layer of marl and soft calcareous chalk. The cause for this may have been an earthquake; the site is not far from the Barqai fault line.
Area B
Quarries. Four quarries were found here (118, 119, 121B, 123). They were similar to those discovered in Area A but more widely dispersed because the cracked rock surface renders quarrying difficult. The quarries were stepped, and the nari crust was hewn down to the chalk layer (Fig. 10).
Winepress 121A. This winepress was hewn into a southward sloping boulder. The rock surface was heavily pitted and rough, probably due to karstic activity. The winepress comprises a polygonal treading floor (c. 2.2 × 2.5 m, depth c. 0.2 m) with rounded corners and, at a lower elevation, a rectangular collecting vat (0.7 × 1.2 m, depth c. 0.5 m) hewn at a north–south axis. The collecting vat’s floor was leveled, and its southern wall was broken and missing. The treading floor was connected to the collecting vat via two diagonal bi-conical holes, located roughly 0.4 m apart (outer diam. c. 0.2 m, inner diam. c. 8–10 cm). An oval, probably natural pit (diam. c. 0.6–0.8 m, depth c. 0.6 m) was uncovered on the western side of the treading floor; apparently, it was adapted for use as a collecting vat at a later phase. Approximately 0.3 m south of the oval collecting vat, a krater/cupmark was exposed (diam. c. 0.3 m, depth c. 0.3 m); its sides gradually converge to the center, ending in a deep depression (diam. c. 0.05 m).
Winepress 122. Winepress 122 was hewn in the southward inclined nari rock and consisted of a treading floor and a collecting vat. The topography was well adapted for the treading floor (c. 3 × 3 m, depth c. 0.15 m), and relatively little work was invested in it. The floor’s eastern side and some of its northern side were clearly preserved, while its western side was destroyed. The collecting vat was square (1.4 × 1.5 m, depth c. 1.2 m) and, compared with the treading floor, very well-hewn: its walls are vertical, straight and smooth, and the floor is level. A deep pit (c. 0.5 × 0.5 m) on the collecting vat’s northwestern side may have been a settling pit. Another irregularly hewn pit was uncovered on the southwestern side of the treading floor; it may have been used as a settling pit for the must flowing toward the collecting vat, or it may have been a later collecting vat, installed after the square collecting vat became obsolete.
Burial Remains. During topsoil removal, a deep pit was mechanically dug at the foot of Winepress 122. The pit was full of soil, and a few human bones were encountered at a depth of approximately 1.6 m below the surface (122A). In the pit’s eastern section, a pocket or penetration of gray soil, typical of the soil in burial caves, was recorded; it yielded a few fragments of bag-shaped jars dating from the third–fourth centuries CE. Subsequent manual excavation yielded no further remains.
Rounded Rock-Cut Feature 123. A steep southwardly inclined nari boulder with arch-shaped quarrying marks was uncovered. These marks were identified during topsoil removal, raising the possibility that this might be an entrance to a burial cave. The manual excavation yielded several jar fragments dated to the fourth–seventh centuries CE.
Cupmark 120. An oval rock-hewn cupmark (diam. 0.6 m, depth c. 0.3 m) was uncovered. In its center, it had a sort of small settling pit (diam. c. 0.1 m, depth c. 0.1 m). Its sides are vertical and slightly concave.
Area C
Quarries. Area C encompassed quarries similar to those recorded in Area A but more widely dispersed. Two quarries were excavated: 126B and 129.
Winepress 124. Winepress 124 is located outside and north of Area C. It comprises a westerly inclined treading floor (1.1 × 1.2 m, depth c. 0.2 m) in the east and a round collecting vat in the west. The collecting vat has concave walls and an inverted rim, producing a spherical shape. A deliberate breach in the vat’s eastern side allowed the liquid to flow from the treading floor directly into the vat. Artificially hewn surfaces on the vat’s southern and western sides probably served to affix a wooden cover to close the vat. Winepresses of this type were used by local villagers at the beginning of the twentieth century, mainly in the capacity of private family winepresses for small-scale production of grape juice (or oil) and grape honey. Similar winepresses were very common in viticultural regions, and they resemble ones found in an archaeological survey conducted in the fields of Zur Natan (Ayalon, Kidron and Sharvit 1990:97–99, Fig. 7a).
Winepresses. Located in the north of Area C, Winepress 125A consisted of a southward-sloping treading floor (1.20 × 1.05 m, depth c. 0.1 m) that drained into an irregularly shaped collecting vat (1.2 × 1.5 m). The sides of the vat were concave, and a shallow settling pit (diam. c. 0.3 m, depth 0.15 m) was hewn into its floor. The soil fill in the vat yielded body sherds of bag-shaped jars dating from the third–fifth centuries CE. Another rock-cut installation (127) was uncovered in the south of the excavation area, probably a winepress’s treading floor.
Pit 125B. A natural pit in the bedrock was exposed in close proximity to Winepress 125A. It was probably incorporated into the winepress complex and accommodated to serve as a collecting vat.
Burial Cave 126A. A vertical square shaft (1.3 × 1.3 m; Fig. 11), leading to a burial cave, was hewn in the bedrock. The shaft was cleared to a depth of 2 m and yielded several Intermidiate Bronze Age sherds. The cave, which was not excavated, was hewn into the chalk layer and roofed by the hard nari crust. Many burial caves of this type have been found in the region (Gorzalczany and Sharvit 2010).
Rock-Hewn Cistern 128. The cistern (0.4 × 0.5 m, depth 0.8 m) was broken on its southern side.
Area D
Limekiln (Figs. 12, 13). The kiln builders attached the installation to a high bedrock terrace. The kiln’s northern part was quarried, while its southern part was excavated and supported by a wall. The kiln’s interior was lined with a wall of fieldstones (W02). The natural bedrock economized on building stones and served to strengthen and stabilize the kiln. The roof may have been domed, as in other kilns discovered in the region. Burnt limestone was found on the surface beside the kiln (L132). A detritus of fieldstones and some roughly dressed blocks was uncovered inside the kiln. While the upper stones in the rubble were charred on their lower side, stones placed deeper in the rubble were often completely charred, cracked and crumbling. Below the rubble, a layer of charcoal and black ash (L148; thickness c. 0.15 m) above a thick layer of burnt stones and limestone (L149) were uncovered. Beneath Layer 149, at a depth of about 2 m from the surface, a thin layer of black ash was found. This ash layer superimposed the kiln’s floor (L151) that sloped toward the center, where a 0.1 m deep step was hewn.
The stokehole was installed in the kiln’s southwestern side, where a blockage of medium-sized fieldstones, a thick layer of ash and signs of a conflagration were found. On the other side and adjacent to the stokehole, an elliptical wall (W04; Fig. 14) delimited the firing chamber (L144; max. height 0.9 m). Its southern part had collapsed, and, beneath the rubble, a well-preserved bronze coin (diam. 1.7 cm; Fig. 15) was found. The coin’s obverse carries the portrait of Constantine II, laureated, cuirassed and draped, facing right. The coin’s reverse portrays two standing soldiers in frontal view, holding a lance in one hand and a sword in the other, hemming two standards. Initial dating places the coin within 335–340 CE (D. Syon, pers. comm.).
Stepped quarries, similar to those unearthed in the current excavation, have been recorded throughout the Roman empire and dated to the Roman and Byzantine periods (Gorzalczany 2007). Based on these parallels, the resemblance to quarries found by A. Yannai to the south of Nahal ‘Iron (Yannai 2010:9294), and the pottery and the coin found during our excavation, the quarries should probably be dated to the fourth–seventh centuries CE. Given the quarries’ irregular distribution, their varied dimensions and their shallow depths, it is apparent that the quarrymen focused on the nari—a thin layer of semi-hard limestone—that is suitable for construction purposes. The nari layer overlies a layer of softer chalk; once it was removed, the quarrymen evidently moved to a new location and did not dig into the soft chalk rock. This procedure explains the large number of quarries that can be typologically divided into small quarries, suitable for the work of one or two workers, and large quarries, where several quarrymen worked (Safrai and Sasson 2001:18–23). The dressed stones and the lime prepared in the kiln may have been transported to Horbat Gilan.
The winepresses found in the current excavation, as well as other excavations and surveys in the region, have local characteristics. They are distinguished by the pairs of conical or bi-conical holes that may have facilitated the introduction of a filtering substance, perhaps thorny burnet (Ayalon, Frankel and Kloner 2013:28).
The site profited from its location—on the road connecting the coastal plain with the Jezreel Valley and the north of the country—affording the opportunity to export building blocks, lime, wine and agricultural produce as and when required.