The surface layer (1.0–1.8 m) was removed with mechanical equipment. Five strata were identified in the excavation (Strata V–I; Fig. 4). Strata V and IV, identified in a mechanically dug probe following the completion of the excavation, yielded a plastered rock floor and a wall of indeterminate date. Strata III–I were manually excavated: Stratum III comprised the remains of a building or an installation from the Late Roman or Byzantine period, which included a room decorated with a mosaic floor and may have belonged to farmstead or a villa; Stratum II comprises the robbery trenches for removing the ashlar-built walls of the Stratum III building; and a field wall and layers of fieldstones were unearthed in Stratum I.
Strata V and IV. Part of a smoothed bedrock floor coated with gray plaster was ascribed to Stratum V (c. 3 m below surface level; Fig. 5). A fill that covered the Stratum V floor and a northeast–southwest wall (exposed length 1.5 m, width 0.4 m; Fig. 6) built on the fill were ascribed to Stratum IV; the wall was constructed of a row of medium–large fieldstones set on their narrower side, which was preserved to the height of a single course.

Stratum III yielded a room (A1), which seems to have had adjacent rooms, and a courtyard (A2) belonging to a building or an installation (excavated length 6 m, width 5.75 m; Figs. 7–9). Five walls (W19–W21, W27, W28; c. 0.4 m wide) built of limestone blocks bonded with lime mortar and founded on clayey soil were preserved. The western wall (W28) was semicircular, resembling an apse. The room had a colorful mosaic pavement that lay on two layers of plaster. The upper layer (L18; thickness 3 cm) was probably a bedding or a repair for embedding the tesserae, and the lower layer (L22) was part of an earlier bedding which predated the mosaic or was annulled when it was laid.
The mosaic (L3; 1.5 × 3.1 m, thickness 5 cm, 56 tesserae per sq decimeter; Figs. 10–12) comprises a roundel (diam. 1 m) at the western end and second geometric motif, which may have included an inscription which was later plundered, to its east. The roundel’s outer frame composes four rows of reddish tesserae. It demarcates alternate red, black and white lotus flowers surrounding a schematic flower whose four petals are alternately red and black; each lotus flower is depicted in a single color. The space between the lotus flowers and the central flower, serving as background, was filled in with red, black and white tesserae. The geometric motif to the east of the roundel is framed by two bands, black and white, each composed of three rows of tesserae. The frame surrounds a panel that has been almost entirely destroyed: only the corner of a rectangular black frame filled with red tesserae remains.
The motifs in the pavement are set against a background white tesserae. The borders along the eastern panel delineate its rectangular shape and are arranged in horizontal and vertical rows. Horizontal rows of unequal length continue and surround the eastern part of the roundel. Continuing westward, the outer tesserae follow the circular outline of the ‘apse’ as well as the roundel. The mosaic artist used tesserae of uniform size in three colors, but the designs are schematic and not precise. The lotus flowers, each rendered in a single color, create a two-dimensional image. The use of three colors instead of plain white for the background of the flower motif in the roundel is puzzling, as it makes the motif hardly recognizable at first glance. Nevertheless, the artist attempted to make the composition conform to the room’s structure, with a rectangular panel in the east, where the walls meet at right angles, and a roundel in the western, apse-like area.
A wall or vault (L29; exposed dimensions 1.7 × 3.2 m, over 0.88 m wide) abutted the room’s northern wall (W20); the wall continues northward, beyond the limits of the excavation. It was constructed of a combination of stream pebbles, flat stones and chunks of limestone in brown soil.
An opening in W19, in which a threshold stone was preserved, led eastward from the room to a courtyard (length 6 m, width 2 m) with preserved sections of bedding for a plaster floor (L11, L30; thickness 0.10–0.22 m; Fig. 13); the courtyard extended northward, beyond the limits of the excavation. On the south, the floor abutted a square built pit (L7; 1.1 × 1.2 m, over 1 m deep; Fig. 14) coated with hydraulic plaster; the excavation within the pit was not completed due to groundwater seeping out of its northwest wall (Fig. 15).
Most of the finds recovered from this stratum were worn fragments of pottery, glass, metal and stone dating from various periods, which came from disturbed contexts. The pottery included a cooking pot (Fig. 16:1) and holemouth jars (Fig. 16:2, 3) from the Iron Age (eighth century BCE); jars (Fig. 16:4) from the Early Roman period (first century BCE – first century CE); and a simple bowl (Fig. 16:5), a rouletted bowl (Fig. 16:6) and pipes (not drawn) from the Late Roman period (fourth–fifth centuries CE). Only a few fragments of glass vessels were found; they were mostly non-diagnostic and poorly preserved. Nevertheless, several fragments were identified and dated to the Roman period (not drawn). The metal items included nails, several worn artifacts and what is probably part of a bracelet, which had penetrated the stratum in a later period (Fig. 17). The stone vessels included a grinding stone, pounding stones (Fig. 18:1) and a polishing stone (Fig. 18:2). Four coins were identified. One coin, recovered from the plastered wall of pit 7, is a copy of a coin from the reign of Alexander Jannaeus (80/79 BCE onward; IAA 152756). Three coins were found inside the pit: one coin of Antiochus IV (173–175/172 BCE; minted in ‘Akko; IAA 152758) and two identical coins from the reign of Severus Alexander (minted in Caesarea; IAA 152757, 152759).

Stratum II. A robbers’ trench identified in a thin layer of clayey soil remained from the robbery of the ashlar-built walls of the Stratum III building (Fig. 19). The assorted assemblage recovered from the trench contributes nothing toward dating the stratum: worn potsherds, mostly non-diagnostic; some are ribbed and others belong to Iron Age cooking kraters (Fig. 20:1–5) and holemouth jars (Fig. 20:6, 7).

Stratum I. Approximately 0.7 m below surface level, a layer of hard clayey soil with a fieldstone fill was found (L10; thickness 0.3–0.8 m; Figs. 21, 22). Above the stones was a section of a north–south field wall (W2; exposed length 2.25 m, width 0.4–0.6 m, preserved height 0.4 m; Fig. 23), built of medium–large fieldstones (average size 0.4 × 0.4 m) set on their wider sides and interspersed with a fill of small fieldstones.
This stratum yielded a fourth-century CE coin (IAA 152755) and abraded and worn potsherds from various periods, including Iron Age bowls (Fig. 24:1, 2), a cooking pot (Fig. 24:3) and a jug (Fig. 24:4) from the Roman period, and a Crusader or Mamluk-period bowl (Fig. 24:5). This stratum most likely represents later farming in this plot.
Flint Items
Hamudi Khalaily
In all, 44 flint items were collected, including four cores, 25 flakes and eight tools. The absence of some of the reduction components, such as chunks and chips, prevented us from identifying the flint industry represented in this collection and from dating of the site. Three of the four cores are exhausted flake cores, and only one bears bladelet scars (Fig. 25:1, 2). The core was made of purple raw material, characteristic of the flint industry of the Pre-Pottery Neolithic B period; it bears transparent patina with abrasion marks and may have come from a nearby site farther up the slope.
Seven of the eight implements were made of local raw material, which is light brown and finely textured; the eighth was made of light gray flint. The tools consist of four retouched flakes, one retouched blade, two awls and one axe. The flakes are small, and the retouch is confined to one side. One of the awls was made on a large, retouched flake; two notches accentuated the pointed tip of the awl on the tool’s right edge (Fig. 25:3). The axe was probably prepared from an elongated pebble; it was first shaped with rough flaking, and then the edges were refined with bifacial thinning until none of the cortex remained. The axe was evidently damaged during its manufacture, causing it to be discarded before it was completed. Its size and form resemble a bifacial tool discovered near the excavation, which has been dated to the early Pre-Pottery Neolithic B period (Khalaily et al. 2007).
The nature of this flint collection and the absence of diagnostic elements limit the extent to which it can be typologically and technologically classified. It thus could not be used for dating the excavation. Nevertheless, it can be assumed that some of the flint artifacts, especially the axe, came from a nearby, previously excavated Neolithic site, whereas the rest were randomly scattered flint items.
The excavation unearthed remains from an unknown period (Strata V, IV), above which lay the remains of a building or an installation that was probably built during the Late Roman or Byzantine periods (Stratum III). The building includes a western room with a colored mosaic floor and adjacent rooms arranged around an open courtyard with a plastered pit with groundwater. The building’s purpose is unclear, but the courtyard’s plastered pit with groundwater may suggest a use connected to the storage of water. However, one cannot rule the possibility that it was an agricultural or industrial pressing or treading installation for the production of oil or wine. The lack of any link between the Stratum III remains and the water channeling complex at ‘En Maqlef (Greenhut and De Groot 2009; Fig 2: Area F) seem to corroborate the second hypothesis. It is possible that the remains belong to a farmstead or a villa, which was partially unearthed c. 50 m to the northeast; it was founded during the Lage Roman period and was abounded late in the Byzantine period (Kisilevitz et al. 2014; Greenhut and De Groot 2009:24; Fig. 2: Area A, Stratum II).
Dating the building remains unearthed in the current excavation to the Late Roman or Byzantine period is based on the mosaic pattern in the western room and on its possible link to the nearby homestead/villa remains. The finds from the robber’s trench (Stratum II) as well as from inside and outside the building could not assist in dating the building, as they comprise mixed and eroded material that was probably washed down from the adjacent tell. Similarly, plaster samples collected from the building did not provide information regarding the date in which the building was constructed or abandoned. It seems that at some later date, the building remains and the evidence of ancient plundering (Stratum II) were leveled, as the area was converted for use as farmland; a single field wall (Stratum I) from this phase was possibly connected to the Arab village of Qalunya.