The streambed of Nahal ‘En Kerem begins at the watershed line near Bayit Va-Gan neighborhood and runs westward between Mount Herzl to its north and Qiryat Ha-Yovel neighborhood to its south. The stream stretches over c. 2.5 km before pouring into Nahal Soreq. It is dammed with agricultural terraces, and the remains of orchards, fence walls and agricultural cultivation steps have been preserved on both banks, some of which are located in the territories of the former Arab villages of Bet Mazmil (Storchan and Dolinka 2014; ‘Adawi and Dolinka 2021) and ‘En Kerem.
Seven find spots with early remains were detected along the stream (Fig. 1:1–7), and areas containing agricultural terraces were mapped. Most of the points are associated with an agricultural system that was in use in the region until the late Ottoman period and the British Mandate era. The facade of a typical Early Roman (late first century BCE–first century CE) burial cave (6) was also surveyed.
1. Cistern (map ref. 217037/630944; Fig. 2). The cistern was hewn into the stream’s northern bank. It has an oval aperture (1.0 × 1.3 m) and was found covered with an iron grate fastened to the rock. Its sides retained traces of light-gray plaster (thickness 2 cm). Approximately 1.9 m southwest of the cistern’s mouth, a square rock-cut feature (0.65 × 0.70 m, depth 0.2 m) was documented; its purpose is unclear.
2. Agricultural Terrace (Fig. 3). An agricultural terrace retaining wall was built across the streambed in a general north–south orientation. The wall was built of large fieldstones (0.3–0.4 × 0.5–0.6 m) placed directly on the bedrock and interspersed with smaller stones to support and stabilize it. Stones that had fallen from the wall’s upper courses are strewn at its feet. Many agricultural terraces have been sited along the stream’s main part, manifesting features identical to those described above; they are located at varying distances from each other.
3. Stone-Clearance Heap (map ref. 216916/630903; Fig. 4). The stone-clearance heap (4 × 5 m, height 0.1–0.2 m) is located within the stream channel, between two agricultural terraces, and consists of small fieldstones piled directly on bedrock. A few worn, non-diagnostic potsherds were observed between the stones.
4. Agricultural Unit. A modern agricultural unit was incorporated into the ancient setting of agricultural terraces. The unit is surrounded by a stone wall. In its center, one observes the remains of a recent building (Figs. 5, 6) and a rock-hewn cistern with a settling pit linked by a channel (Fig. 7). Modern concrete was used to construct the building and the cistern’s settling pit. Potsherds from the Late Ottoman period and the British Mandate era are scattered on the surface, and pieces of recent metal barrels are strewn around the building; the absence of pottery from earlier periods reinforces the assumption that the unit is recent. The unit’s features and the finds dispersed around it suggest its attribution to farmers from the nearby Arab villages.
This agricultural unit was previously documented in the Jerusalem Survey (Kloner 2003:107; Site 120) when Byzantine and Early Islamic pottery was identified. The current survey detected no pottery from these periods.
5. Field Tower (map ref. 216814/630952; Fig. 8). Near the southern boundary of the agricultural unit (Point 4), the foundations of a small oval field tower were documented (diam. 2.2 m), leaning on a bedrock step to its north. The walls (thickness c. 0.5 m) were built of medium-sized fieldstones and preserved 1–2 courses high, and the tower’s western side was buried under collapsed stone rubble. The upper preserved course leans inward from the foundation’s line, suggesting that the building had a domed roof. The entrance to the tower was not identified; it may have been on the western side, buried beneath the rubble.
6. Burial Cave (map ref. 215812/630850; Fig. 9). The courtyard (1.1 × 2.1 m) and facade (height 2.1 m) of an incomplete burial cave were discovered. They were hewn into a hard limestone rock terrace on the stream’s southern side. The cave facade contains a rectangular, northwest-facing, rock-hewn niche (2.1 × 2.2 m, depth 0.8–1.1 m), probably the beginning of an unfinished entrance. The niche was surrounded by an asymmetrical rock-hewn rectangular frame (width 0.43 m on the eastern side, 0.26 m on the western side). Except for a leveled section in its southeastern part, the rock face inside the niche was undressed and remained arched. The cave courtyard is demarcated on the east and west by rock-cut walls (length c. 1.4 m and 1.2 m, respectively), along which benches were hewn (width 0.15–0.20 m, height 0.4 m).
7. Built Channel (Fig. 10). The channel is c. 100 m long, built along the stream’s lower reaches until its confluence with Nahal Soreq. The channel consists of two sections of wide parallel walls (width 1.2–1.5 m, average height 1.5 m), made of different sized fieldstones and built directly on the bedrock. The average distance between the walls is 2 m on average. The channel constituted an integral part of the agricultural terrace system in the streambed, used to drain the runoff water toward the Nahal Soreq valley. This channeling method is known from many streams in the Jerusalem Hills, such as Nahal Soreq, Upper Nahal Kesalon and Nahal Zova. Damage to these channels in cultivated areas leads to flooding of the land along the streambed.
Most of the remains surveyed (2–4, 6, 7) reflect the latest phase of traditional agricultural operations practiced in the streambed and along its banks. The hewn cistern, the agricultural terraces and the channel built to divert runoff water were constructed to ensure maximum water utilization during the rainy seasons. The channel’s construction beside the stream required considerable investment and organizational ability, underscoring its necessity. Without channeling the excess runoff during the winter months, the cultivated plots along the stream were at risk of losing their soil to erosion. The stream water drained into Nahal Soreq, flowed westward and was probably used to irrigate agricultural terraces and plots downstream that belonged to inhabitants of the nearby village of ‘En Kerem. The surveyed stream’s proximity to the Mamluk and Ottoman remains of Khirbat Bet Mazmil (Walker and Dolinka 2020) suggests that inhabitants of this site may have been the farmers that tended to these plots. In either case, the series of agricultural terraces along the stream and on its two banks is probably the product of a slow process that began long before the periods represented in the current survey. This hypothesis is reinforced by our knowledge of ancient settlements in and around ‘En Kerem (Saller 1946) and Bet Mazmil (Walker and Dolinka 2020) and in light of excavated agricultural terraces throughout the Jerusalem Hills, some dating from Iron Age II and some from the Roman and Byzantine periods (Edelstein, Milevski and Aurant 1998:6–14). The unfinished burial cave is typical of burial caves from the Early Roman period and may be associated with a nearby farmstead.