The uneven limestone bedrock surface gently sloping down to the northeast was exposed directly below the ground surface (L101; c. 140 sq m; Figs. 2–4). Several shallow, curving cracks in the bedrock surface and a larger hollow (L103, L105; c. 2 × 2 m) at its southern edge that penetrated the nari crust down to the softer chalky limestone layer were probably created by surface runoff. At the northern edge of the excavation area, a steep rock-hewn step (L108) may have been part of a rock-cut installation located mostly beyond the excavated area. A single course of boulders and various-sized stones set in a row directly on the bedrock may have been part of a boundary wall associated with this conjectured installation (W106; Fig. 5).
 
Quarry. The bedrock surface at the western edge of the excavation bore evidence of stone quarrying (L110; Fig. 6): two rock-cut steps (2.6 m long, c. 0.4 m wide, c. 0.3 m deep) separated by a narrow rock-cut fissure (1.5 m long, 3–4 cm wide, 8–10 cm deep), as well as a couple of small, perpendicular cuts along the fissure that marked the divisions between the individual blocks. These cutting lines allow us to estimate the dimensions of the stone blocks that were hewn (c. 0.7 m long, 0.4 m wide, 0.3 m deep).
 
Burial Cave. A sunken rectangular forecourt or courtyard (L102, L107; 1.6 × 2.6 m, more than 1.9 m deep; Fig. 7) of a burial cave was hewn in the bedrock immediately adjacent to, and aligning with, the quarry lines. The courtyard was not fully excavated since its floor was overlain with large limestone blocks with an earthen accumulation in between them. These blocks, of various dimensions (0.4 × 0.4–0.5 × 0.7–0.8 m), must have been quarried within the courtyard or in the immediate vicinity and placed there after the burial cave and the courtyard fell out of use. Although most of the blocks were not removed in the excavation, it was possible to discern that the courtyard walls were carefully hewn and that no entrances were cut into its northern, western and southern walls.
An entrance was cut in the eastern wall of the courtyard. It consisted of a double arch, an outer arch with sloping-in sides (1.2 m wide, 1.3 m exposed height, 0.4 m deep) and within it a smaller, inner arch (0.8 m wide, 0.6 m exposed height, 0.3 m deep; Fig. 3: Section 1–1; Fig. 8). The entrance passage was partially blocked by a large, rectangular limestone block (0.4 × 0.4 × 0.7 m) that was removed in the course of the excavation in order to allow access into the burial chamber.
The burial chamber was squarish in shape (L109; 2.2 × 2.4 m, 0.9 m height above the accumulated earth layer). It was partially blocked with earth that had filtered in via a small hole in the eroded hollow in the overlying bedrock surface described above (L103, L105) that penetrated the ceiling of the southern loculus (Fig. 3: Section 3–3). Its straight walls curved in at the top to form a nearly flat ceiling. Six vaulted loculi (kokhim; c. 2.0 m long, 0.5 m wide, 0.6 m high; Fig. 9) were cut into the walls: three in the northern, two in the eastern and one in the southern wall. The openings of four of the loculi were partially blocked by plain limestone slabs (0.7 m long, 0.6 m wide, 0.2 m thick), leaning somewhat obliquely against the wall. A glance into the loculi revealed that although they seem to have contained only a shallow accumulation of soil, no artifacts or skeletal remains were visible. The burial cave was probably looted in antiquity, before the courtyard was filled in, when the robbers could still enter via the main entrance, as the hole in the depression (L105) penetrating the roof of the southern loculus was too small to allow human access. Since the burial chamber was not excavated, we neither know whether it had a standing pit or other additional features, nor whether it contained any extant grave goods or skeletal remains.
 
The Finds (Fig. 10). No pottery was collected from the central chamber or from the loculi as they were not excavated; approximately 32 diagnostic rims were collected in the sunken courtyard and on the bedrock surface around it. A few restorable potsherds of a Kefar Hananya cooking pot Form 4C (Fig. 10:1) and of a ridged-neck storage jar lacking a rim (not illustrated), together with a single coin (see below), came from the lowest level excavated in the courtyard bedside the arched entrance (L107). A restorable Kefar Hananya bowl Form 1E (Fig. 10:2) and some potsherds of a Kefar Hananya open cooking pot Form 3B (Fig. 10:3) were found in the hollow in the bedrock surface (L103). A few small, worn Late Hellenistic, Roman and Early Byzantine potsherds and a few small glass fragments were scattered over the bedrock surface (L101) and in the upper layer of fill within the courtyard (L102).
The three illustrated pottery sherds range in date from the second to the fourth centuries CE (Adan-Bayewitz 1993:103–109, 119–124, 128–130). The single coin is a Roman provincial bronze, apparently minted in Caesarea, from the reign of Hadrian or Antoninus Pius (117–161 CE; IAA 143478).
 
The rock-hewn burial cave consists of a sunken rectangular forecourt or courtyard that led into a central burial chamber with six loculi. Since the burial chamber itself was not excavated, the only dating evidence was the limited pottery and the single coin found in the courtyard and on the bedrock surface. On that basis, the cave may have been in use in the Middle and possibly in the Late Roman period (second–fourth centuries CE).
Similar burial caves were excavated at Kafr Kanna, both on the hill of Jebel Khuwweikha (Najjar 1997; see Fig. 1:2) and in the nucleus area of the old village (Abu-Uqsa 2002), which was not settled in the Roman period but served as a burial ground (Alexandre, forthcoming). Similar burial caves, usually with a standing pit hewn into the floor of the central chamber, and often exhibiting distinct Jewish characteristics such as ossuaries and Herodian lamps, are characteristic of the Roman period in the Jewish Galilee.
The proximity of the burial cave and the stone quarry suggests that the rock-hewing for the burial cave and the quarrying of stone blocks may well have been carried out contemporaneously, the two processes being interdependent. The presence of the burial cave, the quarry and what seems like a rock-hewn installation, as well as several possible installations and caves that were observed in the immediate vicinity, reflect the utilization of the area on the outer perimeter of the settled Roman-period Jewish village. This cemetery must have served the hamlet or small village that lay c. 250 m to the west (Fig. 1:4, 5), and may also have served a population from additional Jewish settlements located farther afield, such as ‘En Kivshan, an unexcavated multi-period site located c. 2 km to the east (Gal 1998:25*, Site 9).