Rock-hewn agricultural installations from the Roman–Byzantine periods were identified along the northern outskirts of the modern settlement, c. 1 km northeast of the settlement-remains from these periods, and c. 1 km south of the current excavation area.
Several rock-hewn installations were exposed (Fig. 2), including two undated quarries that produced building stones, and a winepress that is ascribed, on the basis of the pottery finds, to the Late Roman period (fourth century CE). Several cupmarks (Fig. 3), which could not be dated, were also observed.
Quarry L104 (Fig. 4). The severance channels (width c. 0.1 m) indicate that rectangular building stones (0.3 × 0.7 m) of unknown height were extracted from this quarry. The southern part of the quarry, which was not preserved, may have been damaged recently by mechanical equipment.
Quarry L105. This was a stepped quarry (c. 16 sq m; depth in center c. 0.6 m, depth along the edges 0.3 m; Fig 5) for building stones. Its southeastern part was not excavated. Severance channels (width 5–10 cm) that were identified in several places indicate that the size of the stones that were produced varied.
The sediment that filled the quarries yielded several worn fragments of pottery, dating mainly to the Roman–Byzantine periods. Two sherds date to the Mamluk period.
Winepress L108. A winepress (Fig. 6) on a bedrock surface south of the stone quarries was excavated, and was dated on the basis of pottery to the Late Roman period (fourth century CE). The installation had three components:
Treading Floor. The edges of the treading floor could not be detected, and probably were not preserved. In the middle of the floor was a square platform (A; 0.9 × 0. 9 m, height 0.4 m) with a depression (depth c. 0.3 m) in its center. This was probably the base for a screw press that extracted liquid from agricultural produce. It was surrounded by a hewn channel (min. depth 0.3 m) that conveyed the liquid to a rock-cut collecting vat to the north. The upper part of the channel was covered with flat stones, apparently to shield the liquid flowing through it.
Settling Pit. The settling pit was square (B; c. 0.7 × 0.7 m, depth c. 0.4 m) and coated with white plaster (thickness c. 2 cm). A hewn channel with a round cross-section, at the bottom north, led to the collecting vat.
Collecting Vat. The square collecting vat (C; c. 0.26 × 2.6 m) was mostly cut into the bedrock. In the northern and western sides the bedrock was too shallow, and construction was completed with rectangular, dressed stones of different sizes, possibly from the quarries that were exposed to the north (Fig. 7). Plaster remains were identified, mainly in the northwestern corner and on the southern wall of the vat. Two steps were preserved (Fig. 8) from a staircase adjacent to the eastern wall of the vat, that descended from the top in northeast to the bottom in the southeast. Chisel marks on the eastern wall of the vat indicate that at least two more steps did not survive. Outlets of two channels that conveyed liquid to the vat (Fig 9) were found in the southern wall; one, with a rectangular cross-section, carried liquid from the treading floor, and the other, round in cross-section, from the settling pit.
The excavation exposed remains of rock-hewn installations that testify to a building industry, and to production of wine or olive oil. Considering that most of the ceramic from the quarries and the winepress date to the Roman–Byzantine period and the fourth century respectively, these facilities may be part of the agricultural hinterland of the contemporary settlement known as Karm er-Ras that was identified farther south, on a small hill in the western part of Kafr Kanna (Alexandre 2008). Several sherds dating to the Mamluk rule (1260–1517 CE) raise the possibility that these installations were also used during this period.