The current excavation involved cleaning and documenting ancient remains that had been damaged, including a mosaic floor (Phase 1) and the remains of two later construction phases (2, 3). House 18 is an elongated structure built along a north–south axis with a ceiling of cross vaults supported by pilasters built along the walls. Beneath the building’s floor is wide cistern (L103). It seems that its upper part was built, whereas the lower part of the installation was rock-hewn. The cistern’s walls, revealed in a small excavation probe (L102), were treated with a thin application of gray plaster that contained neither grog nor lime. The cistern connected on the north to a small cubicle located beneath the building’s floor level.
The infrastructure work that was carried out up to the line of the structure’s northern facade destroyed the built remains and damaged the hewn bedrock. The cistern’s ceiling was dismantled, and building remains were discarded into the cistern, after which it was covered with concrete. The infrastructure work created a broad subterranean cavity. The cavity’s sections were covered with concrete, except for the northwestern corner of the cavity which remained protruding, with two exposed sections (a, b; Fig. 2). Both of those sections were cleaned, and the remains identified in them along with the bedrock floor in the northern part of the cavity were documented.
The excavation revealed that the bedrock surface (L101) was high, and rose toward the northwest. Remains of at least three construction phases were discerned in the sections. The relation between the phases and the cistern is unclear.
Phase 1. A white mosaic pavement (L105; Figs. 3, 4) was composed of small tesserae (c. 0.3 × 0.3 cm) set on a bedding (L106) of small stones and earth mixed with plaster. The bedding was placed on the bedrock, which was leveled for that purpose. Three body fragments of jars were found in the bedding.
Phase 2. An entrance to a building was constructed above the mosaic floor. In order to carry out the construction, the mosaic was covered with a layer of fill (L104) consisting of soil and small- and medium-sized fieldstones. The western doorjamb (W201) was exposed in the western part of Section b (Figs. 5, 6); a thin layer of white plaster still adhered to the surface of the doorjamb. A fragment of a large, carefully dressed threshold stone (width c. 0.7 m) of the entranceway was visible in Section a. The threshold stone broke when the entire eastern side of the entranceway was damaged during the infrastructure work. Another fragment of the threshold stone was discovered on the floor of the cavity.
Phase 3. A staircase consisting of at least three steps (L107; Fig. 5) that ascended northward, toward the street, was built next to the doorjamb on the inside of the entranceway. The steps were constructed of building stones in secondary use; these included an ashlar bearing remains of plaster and two threshold stones (Fig. 7). The construction of the staircase was probably related to the rising street level.
No remains that could assist in dating the three construction phases were found. Judging by the construction style, it seems that the mosaic floor should be dated to the Roman or Byzantine period. It was impossible to date Phases 2 and 3, and they are apparently not connected to the existing building which was probably erected in the Ottoman period.