Area A (Fig. 1). In the northern part of the excavation area, a smooth rock surface (L113; Fig. 2) was exposed. Light brown soil above this surface yielded fragments of pottery vessels and industrial waste from lathe-made chalk vessels. The pottery sherds, dating to the first century CE, include cooking pots (Fig. 3:6–8; Tchekhanovets 2013:131–132). The chalk vessel industrial waste (Fig 3:9–12; Magen 2002: Pls. 2.24, 2.25, 3.29) dates to the Early Roman period. A stone lid discovered in the assemblage (Fig. 3:9) bore deep- cut lathe marks, indicative of poor craftsmanship. Apparently, a production center for chalk vessels was located at the site.
A wall foundation (W120) built of medium-sized fieldstones (c. 0.17 × 0.28 × 0.36 m) set on the bedrock and aligned north–south, was discovered in the center of and north of the excavation area. East of W120, a smoothed bedrock surface (L119) was exposed; the soil above (L110) yielded fragments of pottery vessels from the fifth–sixth centuries CE, including bowls (Fig. 3:1, 2; Hayes 1972:323–337) and kraters (Fig. 3:3–5), coarse white tesserae (2 × 2 cm) and fragments of roof tiles.
In the south of the excavation area were remains of a wall (W102; Fig. 4) built of two rows of roughly hewn fieldstones with a small-stone fill, and oriented north–south. Only the small stone fill was preserved in a 2 m long section in the center of the wall. Byzantine pottery sherds, fragments of roof tiles and coarse white tesserae were discovered in the stone collapse west of the wall.
Area B (Fig. 5). A rock-hewn ritual bath (miqveh) was discovered, consisting of five steps and an immersion pool. It was coated with two layers of plaster—a gray bottom layer and a white top layer (Fig. 7)—characteristic of the first century CE. A step that served as a bench was hewn at the bottom of the western side of the immersion pool. At the top of the steps was a hewn and plastered bedrock surface, divided by a plastered bedrock partition. Slightly south of the bedrock partition was a wall (W212) built of two courses of large, roughly hewn stones (c. 0.50 × 0.60 × 0.66 m), aligned east–west; it was founded on the layer of plaster that coated the ritual bath. This wall apparently canceled the use of the bath and served as a retaining wall for the reservoir (below). Gray soil (L205; thickness 7 cm) exposed in the immersion pool and on the steps that had accumulated there after the bath ceased use contained fragments of pottery vessels mixed with 33 glass fragments dating to the early second century CE. The ceramic finds include bowls (Fig. 8:1–3), ETS bowls (Fig. 8:4, 5), a cooking casserole (Fig. 8:6), cooking pots (Fig 8:7–9), jars (Fig. 8:10–12), jugs (Fig. 8:13–15), a lid (Fig. 8:16) and a discus lamp (Fig. 8:17, 18). The glass fragments, light blue in shade and covered with a thick layer of black weathering, include five bowl fragments with a thickened rim, two fragments of low ring bases, six rim fragments of cups and two bases of cups (Gorin-Rosen 1999:85–90). Based on the ceramic and glass artifacts, the ritual bath apparently went out of use at the beginning of the second century CE.
South and west of the ritual bath was an elliptic hewn and plastered water reservoir (L210; Fig. 9), in which two phases could be discerned. In the early phase, a small reservoir was hewn that served as an otzar for the bath. A plastered bedrock floor ascribed to this early phase exposed south of the bath led west to the opening of the reservoir. In a later phase, after use of the bath ceased, the reservoir was widened by quarrying. During this phase, a stone pavement (L214; Fig. 10) was constructed above the plastered bedrock floor in front of the reservoir. A stone threshold in secondary use was discovered between the paving stones. A rock column was left in the center of the reservoir to support the ceiling. A plastered vertical rock-cut shaft (L213) that led to the surface was discovered in the western side of the reservoir’s ceiling; it was used to convey runoff into the reservoir. The northern side of the shaft was lined with medium-size stones up to the surface. An inscription engraved in English (Fig. 11) in the ceiling of the reservoir consisted of the names CPL Scarlett and Walsh, the initials RAE, the numbers NX9168 and NX7792 and the date, 30.05.1940. This inscription shows that the reservoir was visible until the 1940s. An investigation of the content of the inscription revealed that the initials CPL stand for the rank of corporal, the numbers are the serial numbers of Australian soldiers and RAE stands for Royal Australian Engineers. According to the Australian National Archives, Corporal Scarlett (Philip William Scarlett), Serial No. NX7792, was born in Melbourne on 09.02.1918, was drafted into the army in 1939, survived the Second World War and died on 01.01.1970, while Walsh (Patrick Raphael Walsh), Serial No. NX9168, was born in Cowra on 05.01.1910, was drafted into the army on 12.12.1939, survived the war and passed away on 12.07.2005. It seems that these soldiers served with the forces of the Australian Sixth Division that trained in Israel during the time of the British Mandate, prior to being sent to France. However, since France capitulated before the troops were ready, they were deployed to Egypt in October 1940 and fought on the front in the Western Desert. The finds discovered in the reservoir included 26 rifle cartridges, one 9 mm sub-machine gun cartridge, 2 fin assemblies of a Type 36 3-inch mortar shell, 2 fin assemblies of a 2-inch mortar shell, an item resembling the base of a detonator, a decorated copper pendant from the Ottoman period and several pieces of iron that could not be identified. The ammunition discovered in the reservoir shows that a British Army training ground or military base was located nearby.
Three walls (W208, W217, W218) and a hewn shaft (L204) were exposed above the reservoir. Wall 208 was set on the bedrock and built of large dressed stones, including a broken threshold in secondary use; it was preserved to a height of one course. Wall 217 was built of medium-sized stones above a rock-hewn foundation trench; it was preserved to a height of two courses. Apparently, W217 originally adjoined W208, but the stones where the two walls presumably met were not preserved. Wall 218 (width 0.7 m, preserved height 0.75 m) was only discovered in the area’s western section; east of the section, only the wall’s foundation trench was apparent, hewn in the bedrock (length 2.4 m). Evidently, the three walls delimited a space in the center of which hewn Shaft 213 conveyed water into the reservoir. Alluvium accumulated inside Shaft 204. An opening (diam. 1.3 m; Fig. 12) that led to the reservoir was hewn in the southern wall of Shaft 204, at a depth of 5 m below the surface. For safety reasons, the passage to the reservoir was not completely excavated.