Area I (Fig. 2)
This was the southernmost of the excavated areas, wherein a tomb from the Roman period was discovered, but not excavated. It is not clear if the tomb was disturbed, although its entrance was found still sealed since its last use.

A rock-cut shaft (c. 1.25 × 2.00 m; max. depth 1.75 m), which was filled with earth and had an intentional boulder fill blocking its lower portion, accessed the tomb. The latter’s doorway, where a stone doorsill was placed, consisted of a plain basalt door (0.8 × 1.0 m) that had a simple carved vertical handle in its center (0.25 m long) and was still in place on its hinges. To the north of the door was a constructed doorjamb that had been repaired through the careless insertion of a limestone block as its upper segment. The jamb repair suggests a later, intrusive entrance into the tomb.

Four large rectangular loculi (c. 2 m long; 1.1 m high) were hewn in the soft limestone of the north wall, with an additional single large loculus in the northwest corner of the west wall. Immediately to its south, a seam of hard stone was encountered, which may explain the absence of other loculi in the tomb, leaving the remainder of the western and southern walls coarse and unworked. An arcosolium was cut in the east wall, north of the door.

An unusual rectangular opening (0.6 × 1.5 m; Fig. 3) was cut in the ceiling and sealed with three ashlar blocks. The opening slightly tapered, thereby allowing the blocks to fit tightly while not falling into the tomb cavity.
The façade of the tomb was well smoothed. Unusual, however, was the flat hewn surface (3.5 m long) above the tomb that ended both in the north and south with a rectangular cornice. The bedrock slope to the south and to the north of the shaft into which it was hewn had been severely damaged; thereby any evidence of treatment or ornamentation was removed.

The only finds attributable to the unexcavated tomb were two common Roman-period (second–third centuries CE) disc lamps with broken discs (Fig. 4) and glass fragments datable to the same time span. A small fragment from a first century CE stone vessel was found in the shaft fill.


Area II
The area (c. 5 × 9 m; Fig. 5) was cleared down to bedrock, revealing the remains of a quarry that dated to either the Roman or Byzantine periods and extended over the entire exposed area. The quarry was limited to an outcropping of soft limestone, while the hard limestone surrounding it was not quarried. The measurements of quarried blocks, which could be determined by the few partly hewn blocks that remained in the quarry, ranged between 0.35 × 0.50 m to 0.6 × 0.9 m. Embedded in the hard soil were mostly early Roman potsherds and fewer Byzantine-period fragments. It was, however, impossible to determine the source of these sherds and whether they should be associated with the period of the quarrying activity or shortly thereafter, or were washed down from the upper slopes of the settled area over the course of time.

Area III (Fig. 6)
This area is located c. 12 m northwest of Area II on an outcropping of bedrock covered with a thin accumulation of hard-packed soil. The clearing of the hard limestone bedrock revealed a partially preserved winepress in the southern portion of the area (Fig. 7) and a stepped-quarry adjacent to it in the north.

The treading floor (L306, L317; 4.00 × 4.50 m) was poorly preserved. Only one third of its southern wall had survived but, more consequentially, its eastern wall that connected it to the collecting vats was missing. The extant hewn walls in the west and southwest exhibit at least three layers of plaster, indicative of periodic repairs. The hewn floor was paved with a coarse white mosaic, which was preserved in a few patches and was evidenced in the large quantity of tesserae, ubiquitously present throughout the excavation.

A layer of ash (L316), containing burned sherds from the late Byzantine/Early Islamic periods (Fig. 8:9, 13) and bones, covered the northern half of the floor and postdated the final use of the installation.
Two collecting vats were located to the east of the treading floor. The southern vat (L312; c. 1.6 × 1.7 m; 1.5 m deep; 4.08 cu m) was hewn parallel to the treading floor. However, the drainage technique utilized between the vat and treading floor could not be determined due to the destruction of the dividing wall. The vat’s hewn walls were coated with a number of thick plaster layers that contained a large amount of Byzantine-period sherds. The floor was paved with white mosaic and had a circular settling pit (0.2 m deep) in its northwest corner, also paved with white tesserae. Along the eastern wall three descending steps were built (c. 0.4 m high). The two upper ones were plastered, while the lowest step was paved with white mosaic.

A small trapezoidal draining basin (L314) was north of the vat. It had an exterior layer of grayish plaster (0.12 m thick) that overlaid a layer of pinkish plaster (2 cm thick). Barely perceptible remnants of a channel, connecting it to the treading floor, were discerned.

The northern collecting vat (L309; c. 1.4 × 1.5 m; 1.3 m preserved depth; min. 2.73 cu m) was a more complex element. It was positioned at a somewhat acute angle to the treading floor, which can probably be explained by its being a secondary use of a large (1.6 × 3.0 m), previously hewn shaft. The walls of the vat were lined with 3–4 layers of thick plaster and its floor was simply plastered, in contrast to the white mosaic floor in Vat 312. A single step (0.4 m high) occupied the southeast corner and a quarter circle settling pit was set in the northeast corner. A clear indication of the vat’s earlier arrangement was found during its dismantling.


Remains of two mosaic-tessellated steps were discovered underneath the plastered wall in the southeast and the existing corner step. In addition, many black tesserae, along with the common white ones, occurred only here. The southern and western walls were bedrock-hewn while the northern (W321) and eastern (W322) walls were put up to fit the vat into the already existing shaft. Wall 321, supporting the upper part of the thickly plastered face of the vat, filled in an area between the inner facing of the vat’s plaster and bedrock. Wall 322 (0.5 m thick), coarsely constructed from boulders, small-stone fill and plaster, divided the original shaft in half (see Fig. 7). To its west was the vat and to its east was a loose fill of earth and stones that reached up to the original hewn east face of the shaft (L323).

The plastered floor of Vat 309 (L325; 8–10 cm thick) superposed a layer of binding material, consisting of plaster, ash and tuff (10–12 cm thick), which overlaid a yellowish pounded-earth fill on a layer of uncut boulders arranged in rows (Fig. 9) and partially placed upon an ashlar block stone ledge (1.35 m long) and partially on a loose fill. The composition of the vat walls and floor resulted in its being as hard as cement and extremely difficult to dismantle.

A channel (L324; c. 0.5 m wide; Fig. 10) was hewn from the northeast corner of the original shaft, sloping away and down to the east. The channel had a narrow lip on each side to enable the placing of flat cover stones, some of which were still found in situ. The channel led water away from the original shaft to a natural depression in bedrock from where it appears to have dispersed to the east and south. The western extension of the channel, inside the shaft, was a series of coarsely cut boulders, creating a staircase that descended into the shaft below the dismantled Vat 309. The series of boulders abruptly ceased as it led into a narrow corridor (L327), which was bounded on its south by a wall supporting a loose fill (L328) that occupied the entire southern half of the shaft and on its north by a wall constructed along a hollow in bedrock. The corridor entered a cavity below the stone ledge underneath Vat 309’s floor. The ledge was apparently the lintel of an entrance to the cavity that was not excavated (Fig. 11). The stair-corridor-cavity complex remains enigmatic as no evidence of its purpose had emerged.

The ceramic evidence indicated a Byzantine date for the winepress (Fig. 8:6–8, 10–12) with a relatively large presence of imported wares, such as LRC 3, LRC 10A and ARS 104 (Fig. 8:1–5). A single coin (IAA 106052) minted in Antioch between the years 364–375 CE was found in the soil accumulation of Vat 309.

Immediately to the north of the winepress complex, below a layer of hard-packed dark red soil, a quarry descending down the slope in a series of steps was partially excavated (dimensions of the quarried blocks from 0.7 × 0.8 to 0.55 × 1.00 m). The overlying earth contained a relatively large amount of Hellenistic potsherds, including numerous GCW fragments, and fewer Roman- and Byzantine-period sherds. As in Area II, it was impossible to determine the source of these potsherds. One may assume that the proximity of the quarry to the winepress may imply the quarry post-dated the winepress. This assumption is somewhat problematic because if the press went out of use late in the Byzantine or in the Early Islamic period, as evidenced by the burned area in the treading floor (L316), then the quarry would necessarily be later than the Byzantine period, possibly Mamluk or Ottoman. The dark red packed earth covering the quarry required a much longer period to accumulate than the time span from the Mamluk or Ottoman periods to the present provides. Furthermore, one would expect even minimal finds dating to these later periods in the quarry and such was not the case. It is thus posited that the quarry predated the winepress.

Area IV (Fig. 12)
This area, situated higher up the slope and to the northwest of Area III, revealed a large plastered installation that functioned as part of a winepress, or for water storage. Only the eastern portion of the structure was excavated (Fig. 13).

The southern wall (W404) was erected on a thin layer of soil deposited on bedrock and its base descended with the slope. The angle of descent was quite acute as the wall’s base descended c. 1.25 m over its exposed length of 4.5 m. Wall 404 was constructed from an outer and inner face of medium-sized fieldstones with a small-stone core. The inner face was coated with a thick layer of plaster (0.25–0.30 m thick), which covered the inner faces of all the installation’s walls. It appears that W404, originally built as a 0.60–0.65 m wide wall, predated this complex and was adapted to serve as its southern wall. Its inner, northern, stone face was an added fill between the earlier wall and the thick layers of plaster inside the installation. Wall 404 appears to have been abruptly cut at its eastern end to create a corner with W407. Yet, it still extended slightly past the corner, which was poorly preserved and the precise connection between W404 and W407 could not be definitely ascertained. The partially excavated W415 may represent the original eastern continuation of W404.

Wall 407 was built of a single row of ashlar blocks on a fill of small stones down to bedrock. Its outer, eastern, face was covered with a thin layer of plaster. A complete cooking pot (Kefar Hananya Type 4C; Fig. 14:1) was found in close proximity to the center of the outer face of W407. 

The northern wall (W408) was constructed from two sections (Fig. 15). It bonded with W407 near the northeast corner where it too retained remnants of its plastered outer face, identical to W407. One and a half meters west of the corner was the outer face of W408, well-constructed from large cut stones. It met up with and cut the earlier W413, running perpendicular to it. A gap (0.15–0.20 m wide) was left between the two walls in the lowest courses. From the intersection of W408 and W413 up to the western balk, W408 was built of small stones randomly piled on the remains of W413, creating a wall that was somewhat out of line with its eastern section and had no orderly outer, northern, face, no horizontal stone courses and no evidence of being plastered. The westernmost section of W408 was, in similarity to W404, built on a descending layer of earth. W413, in contrast, was set directly on bedrock. Adjacent to the western balk, the inner-face plaster layer extended over the top width of W408 but did not continue down its outer face. The horizontally applied plaster indicated that the actual height of W408 was 1.35 m inside the installation.

The three walls enclosed an area (4.7 m from north to south and a min. of 3.5 m from east to west; volume greater than 22.2 cu m), which was entirely covered with thick layers of high-quality plaster; on W407 was an extensive thickening of the plaster. The floor plaster contained a large amount of ash, charred material and small pebbles and was spread directly on bedrock. Its thickness (5–8 cm) depended upon the smoothness of the bedrock surface. The two exposed inner corners were rounded, as was the meeting of the floor with the walls.

The ceramic evidence (Fig. 14:1–10) pointed to the middle to late Roman construction of the installation and its activity. Especially outstanding was the quantity of Kefar Hananya type vessels in this area. Byzantine-period sherds were discovered in the material that accumulated within the installation after it was no longer utilized. The many Hellenistic sherds from the installation were nearly all fragments of GCW jars and pithoi (Fig. 14:11).

A coin (IAA 106054) minted in Antioch and dated to 383 CE was found in the earth that accumulated over W415.

Area V (Fig. 16)

This area was situated above Area III, c. 7 m higher up the slope. Near the edge of the steep slope and below the accumulated dark soil was a boulder terrace wall, one stone thick. The wall was excavated in two sections. Three courses were exposed in the northern section (W503) while only one course was preserved in the southern section (W508). Wall 508 was buried deep in the accumulated soil and could be dated to the Byzantine period. Wall 503 appears to be an Ottoman-period reconstruction of W508. Although few in number, only Byzantine-period ceramic finds were identified below W508 down to bedrock, nearly 3 m below surface. In the fill associated with the Ottoman wall was a coin (IAA 106053) attributed to Alexander Jannaeus (78–76 BCE).  


Area VI (Fig. 17)

Area VI, a single square, was the northernmost excavated area. The only element found was a wall (W603), constructed from two stone faces with a core of small stones. Wall 603, oriented east–west, was built directly on the soil accumulation on the slope, as it descended eastward. It abruptly ended 1.5 m east of the western balk. No connecting elements were found due to the limited size of the area. Wall 603 was dated to the Byzantine period on the basis of ceramic evidence. On the exposed bedrock, in the southwestern corner of the square, were Roman-period potsherds that indicated a Roman-period use of this area, and certainly part of the Roman-period activity in the adjacent Area IV.