During October 2004, an excavation was conducted next to Khirbat en-Nusrani (Permit No. A-4256; map ref. NIG 19120–378/59652–900; OIG 14120–378/09652–900), along the route slated for the construction of the separation fence. The excavation, undertaken on behalf of the Israel Antiquities Authority and underwritten by the Ministry of Defense, was directed by M. Haiman (photography), with the assistance of N.S. Paran, S. Talis and I. Peretz (area supervision) and A. Hajian (surveying).
The excavation was carried out in three areas, spread over c. 400 m in the wadi and on the slope to its north. Winepresses, agricultural installations, terrace walls, stone clearance heaps and a rock-hewn cave were discovered. It seems that the sites excavated along the course of the road did not constitute a single agricultural system on the edge of a particular settlement, but were rather a random collection of installations located on the fringes of adjacent ruins, such as Khirbat en-Nusrani, Giv‘at Mirsham and Horbat Benaya. The potsherds collected along the surface and from the excavation dated to the Iron Age and the Byzantine period.
Area A was located in the wadi channel and a number of terrace walls were identified in it. Two segments of one of the walls (length c. 80 m, width 0.80–1.15 m, height 2.4 m) were excavated. The first segment (W102; length 8.0 m, height 0.4 m; Figs. 1, 2) was in the middle of the wadi, where the terrace wall was breached. It became clear that the wall was built of a row of large stones (up to 0.8 m long) on the down-hill side of the terrace and a second row of smaller stones on the terrace side facing uphill. The space between the two rows was filled with small stones and soil. Overlying the original wall was a fill (height c. 0.5 m) of various size stones that was meant to dam the breach and prevent the erosion of soil. A layer of loess (thickness c. 1 m) was beneath the base of the terrace. The second segment (W105; Fig. 3) was excavated at the point where the terrace wall met another wall (W106) that was built along the bank of the wadi. Wall 105 was also built of two rows of stones similar to those in W102. Wall 106 had an early phase (height c. 0.5 m) on which other stones were heaped in a later period. Non-diagnostic pottery fragments were found in the excavation.
Area B was located c. 400 m north of Area A. A rock-hewn cave, a winepress and two stone clearance heaps were discovered. The bedrock-hewn cave (L301; 4 × 6 m, depth c. 1.7 m; Figs. 4–6), whose ceiling had partly collapsed, had no building remains or steps in its entrance (width c. 0.5 m). Alluvium and modern debris had accumulated above bedrock in the vicinity of the entrance (thickness c. 0.4 m) and inside the cave (L304; thickness c. 0.7 m). During the excavation, the area above the caved-in ceiling (L302) was also cleaned. Non-diagnostic potsherds that had been swept into the cave were found, as well as potsherds from the Iron Age and the Byzantine period. Caves of this kind are common to the region and have been used as dwellings until about 60 years ago; thereafter, they were utilized as sheepfolds.
The bedrock-hewn winepress included a treading floor (L401; 2.4 × 2.4 m, depth c. 0.2 m; Figs. 7, 8) and two hewn channels (length c. 0.3 m, width c. 8 cm) that led to a collecting vat (L409; 1.0 × 1.4 m, depth c. 1.3 m). An opening in the bedrock floor of the collecting vat accessed a hewn, underground cavity (L411; 1.8 × 2.6 m, depth 1.4 m) that may possibly be a cistern. Above L411 and next to the collecting vat was a hewn cupmark (diam. 0.7 m, depth 0.2 m). A stone clearance heap (3 × 7, height c. 0.5 m) was discovered next to Winepress 401 and a square (c. 2 × 2 m) was excavated in its midst. The heap was built of fieldstones (max. dimensions c. 0.2 × 0.2 m), placed the bedrock.
A square (c. 2 × 2 m) in the center of another stone clearance heap (3 × 9 m, height c. 0.5 m; Fig. 9) was excavated down to bedrock level.
Area C was located between Area A and Area B, c. 200 m north of W102 and halfway up the slope. A rock-hewn winepress and trough were exposed. The winepress consisted of a treading floor (L202; c. 4 × 4 m, depth c. 0.5 m; Figs. 10–12) that was paved with mosaic. The bedrock south and east of the treading floor was leveled into a smooth surface (L201; c. 3 × 4 m) that sloped gently in the direction of the treading floor. A channel (length 0.5 m, width 0.1 m) cut into the eastern wall of the treading floor led to a filtration pit (L207; c. 0.5 × 0.5 × 0.5 m) at whose bottom was a rock-hewn sump (diam. 0.1 m, depth 8 cm). A hewn channel (length 0.3 m, width 8 cm) led from the filtration pit into a collecting vat (L203; c. 1 × 1 m, depth c. 1.5 m), which was paved with mosaic and had a sump in the corner (diam. 0.15 m, depth 0.1 m). Two sections of a stone column (diam. 0.3 m; Fig. 13) were found in the collecting vat, one of which was embedded in the mosaic floor. It seems that this was a piece of a roller stone that was discarded in the collecting vat after the winepress was no longer in use. Winepresses equipped with a roller stone are dated mainly to the Byzantine period; however, they probably appeared as early as the Roman period (D. Amit and Y. Baruch, 2007, Winepresses with Roller Stones in the Southern Hebron Hill Country, Judea and Samaria Research Studies 16:299–322).
A rock-hewn trough (L205; length 1.3 m, width 0.3 m, depth c. 0.1 m; Figs. 14, 15), which was incorporated in a curved terrace wall (length 5.0 m, width 0.5 m), was found near the winepress. Potsherds were not recovered from the excavation, but a scattering of potsherds that dated to the Byzantine period was discerned in the vicinity of the winepress and the trough.