From October 2017 to March 2018, a salvage excavation was conducted north of Khirbat Khaur el-Bak, northeast of Ashqelon (Permit No. A-8127; map ref. 163820/623830), prior to the construction of a new neighborhood and an access road. The excavation, undertaken on behalf of the Israel Antiquities Authority and financed by the Israel Land Authority and the Ashkelon Economic Company, was directed by I. Taxel and N.S. Paran, with the assistance of E. Kogan-Zehavi, A. Fraiberg, S. Weiss and I. Weissbein (area supervision and field photography), Y. Huster, A. Nevo, L. Shilov, N.D. Michael, and H. Zilberklang (assistant area supervision), Y. Alamur, A. Levi-Hevroni and E. Hemo (administration), M. Kahan and Y. Shmidov (surveying), E. Aladjem (drone photography), O. Tamir, D. Pukshanski, S. Bloch, S. Kasa, A. Azoulay and N. Argov (educational program), as well as D.T. Ariel and G. Bijovsky (numismatics).
Hellenistic Period (third–second centuries BCE)
The excavation unearthed buildings and a pottery kiln, which were constructed either directly on or into the virgin soil layer, or on a dark soil layer containing pottery sherds from EB II–III, MB II, LB II and Iron Age II. This soil may have been brought from nearby Tel Poran, where habitation layers from these periods were uncovered, in order to level the ground and/or improve the agricultural soil in the Hellenistic period.
A few adjacent structures that were probably part of a single complex, were identified; only the thick wall foundations (c. 1 m wide), built of fieldstones, were preserved. Some of the walls’ upper courses may have been built of mud bricks, as in many contemporaneous buildings discovered in the Ashqelon region.
The southern building, probably constructed on an east–west alignment, contained two or three rows of pillars supported by square stone foundations, but the building’s outer perimeters were not exposed. To its north, a few adjacent, north–south aligned architectural units were composed of several rooms, some of which were quite narrow. These units were either independent buildings or parts of a single large building; they were clearly contemporaneous and had a similar function. Judging by the thickness of the walls, it is probable that the walls reached a considerable height, and the complex served some kind of military or defensive purpose. No clearly identifiable floors were extant in the buildings, as the later Roman structure had apparently severely damaged the Hellenistic remains. Nonetheless, numerous Hellenistic pottery fragments were found associated with these remains.
About 15 m west of the pillared building, a small mud brick pottery kiln (c. 2.5 m diam., c. 1.5 m preserved depth; Fig. 3) was discovered. The kiln had a retaining wall inside the firebox to support the firing chamber floor. The bricks were soft and friable, and the kiln contained no slag or pottery wasters, indicating that it only operated for a very short time; it may even have been abandoned shortly after its construction, before being used. The large quantity of Hellenistic potsherds and other finds recovered from inside the kiln, suggests that it was abandoned in the course of the Hellenistic period, and converted into a rubbish pit.
In the Roman period, a large architectural complex was built at the site, including a winepress, a bathhouse and a building that was probably used for residential and storage purposes. It was apparently a farm, or an agricultural estate that was settled from the second to the late third or early fourth century CE.
The winepress (c. 6 × 8 m; Fig. 4), constructed in the northwestern part of the complex, comprised a square treading floor, a rectangular settling pit and a square collecting vat. The walls were built partly of fieldstones bonded in mortar, and partly of stones combined with fired clay bricks, the latter characteristic of bathhouses; the bricks may have been surplus bricks from the adjacent bathhouse (see below). Some of the winepress walls were founded directly on top of walls of the Hellenistic buildings.
The treading floor had a light gray mortar bedding lain on a fieldstone foundation. The mortar bedding was covered with a layer of fine mortar, into which many closely spaced jar body sherds were inserted vertically with their straight edge uppermost. The potsherds were arranged in five or six parallel bands, alternately aligned in a north–south and an east–west direction (see Fig. 4). The potsherd surface was probably coated with a thin layer of fine plaster so that it could be used for treading grapes. This paving technique is characteristic of sites in the southern coastal plain during the Roman and Byzantine periods. In the southwest corner of the treading floor, a narrow band of large white mosaic tesserae was preserved. This may have been part of an earlier floor; alternatively, the treading floor may have been paved with a combination of different materials.
The channel or pipe that led from the treading floor to the settling pit was not preserved. The settling pit and collecting vat were coated with a few layers of white hydraulic plaster, attesting to prolonged use of the winepress. A semi-circular niche was unearthed on the northern edge of the settling pit, and a lead pipe was inserted in the wall above it to direct the liquids from the settling pit into the collecting vat. A round depression paved with mosaic was uncovered in the center of the collecting vat. The settling pit and collecting vat were found full of many pottery sherds, animal bones and other finds, and they were evidently converted into refuse pits after the winepress fell out of use. The overwhelming majority of the pottery dates from the Late Roman period (late third or early fourth centuries CE). Fragmentary remains exposed to the west of the winepress included a plastered pit and a surface paved with white mosaic, which were either part of the winepress or of some other adjacent structure.
Bathhouse. About 8 m south of the winepress, a small bathhouse (c. 8 × 16 m; Fig. 5) was built, its plan was almost entirely preserved. The bathhouse was constructed on an east–west alignment, with a slight deviation to the southeast in relation to the winepress. The walls were built of fieldstone foundations that probably originally supported courses of dressed stones, almost all of which had been robbed. The building had an entrance room, a changing room (apodyterium), a small cold pool (frigidarium), a warm pool (tepidarium) and a hot pool (caldarium); the latter two rooms were furnished with a hypocaust system. The furnace area (praefurnium), the reservoir and parts of the drainage-pipe system were also discovered.
The bathers would enter the bathhouse from the entrance area that was located at the eastern end of the bathhouse, southeast of the tepidarium. The white mortar foundation of its ruined floor was preserved, on which several mosaic stones as well as a few marble slabs, some of which were cut in various geometric shapes, were discovered.
An opening in the west wall of the entrance room led into the apodyterium. The threshold and the changing room itself were paved with a mosaic floor composed of medium-sized tesserae, most of which were white, apart from a central panel composed of black, white, red and yellow tesserae. Most of the mosaic floor was destroyed, but a guilloche that probably surrounded an inner motif or inscription, was extant.
From the apodyterium, the bathers continued westward to a rectangular pool with a step down to the floor, which functioned as a frigidarium. The pool walls and the step were coated with hydraulic plaster, and the floor was paved with large white tesserae. A small opening located above the southwestern corner of the floor drained the dirty bathwater into a long pipe that curved to the southeast, probably to an open area outside the building. The piping segments that were installed in the pool wall were made of lead, and the outdoor pipe was composed of clay segments. A small inspection hole was built between the second and third segments of the pipe.
After bathing in the frigidarium, the bathers apparently returned to the apodyterium and continued northeast to the tepidarium. The floors of the tepidarium and the caldarium both incorporated a hypocaust system that was laid over a raised foundation surface made of fieldstones bonded in mortar. The tepidarium’s hypocaust contained 12 colonnettes built of fired clay bricks, most of which were square or rectangular and a few were round. The floor overlaying the hypocaust colonettes was not preserved.
The western wall of the tepidarium had an entrance, not preserved, that led to the caldarium, the largest unit in the building. A short, narrow passage linked the tepidarium’s hypocaust to that of the caldarium. The caldarium’s hypocaust contained 20 colonnettes, constructed similarly to the colonnettes of the tepidarium. A pipe made of clay segments, whose northern end was not preserved, drained the used bathwater from the caldarium to an outdoor area south of the building.
At the western end of the hypocaust was a passage whose L-shaped walls were built of fired bricks. This area was full of ash and soot, and was the bathhouse’s furnace, beyond which was the stoker’s praefurnium. To the southwest of the furnace, remains of a plastered pool were discovered that was probably the reservoir for the bathhouse.
Potsherds found next to the building’s foundations date its construction to the second century CE, probably also the date when the other structures of the farmhouse or estate were built. The hypocaust chambers of the tepidarium and the caldarium, and the furnace area were found full of a collapsed debris layer, including fragments of colonnettes, segments of square-sectioned clay pipes (tubuli), through which the hot air passed, and parts of the floors of the upper units, including marble slab fragments. The debris layer also yielded potsherds, an intact pottery juglet and other finds that date the bathhouse’s abandonment and destruction to the late third or early fourth centuries CE, contemporaneous to the abandonment of the nearby winepress. Many more potsherds and an intact Spanish amphora from the same period were found outside the furnace area (Fig. 6).
The area between the bathhouse and the winepress yielded wall sections that can be attributed to the Roman period, but their poor preservation made it impossible to determine their function.
Residential and Storage Building. The fragmentary remains of a large building (at least 15 × 25 m) were discovered east of the winepress and the bathhouse. Its walls were built of ashlar stones placed on foundations of fieldstones bonded in mortar. Parts of several rooms and units were uncovered in the building, but none was entirely exposed. The building’s southern wing incorporated an area with ashlar stone square pillars (c. 1 × 1 m) to support a roof, some free-standing and others engaged in the walls. No floors could be categorically attributed to the original phases of the building’s use, but finds retrieved in the foundations and fills that may have been deposited in the floor’s make-up layer, date the building’s construction and its main occupation period to the second and third/early fourth centuries CE.
Byzantine Period (fifth–seventh centuries CE; possibly until early Umayyad period)
Meager and fragmentary remains attributed to the Byzantine period were discovered throughout most of the excavation area. At the area’s northern fringes, at a spot without Hellenistic and Roman structures, a Byzantine refuse pit was uncovered. Another refuse pit was found inside one of the rooms in the Roman residential and storage building. Other parts of the Roman building exhibited secondary floors, some of tamped earth and others of stones; these floors were dated to the Byzantine period by the pottery and numismatic finds recovered on the floors and in their foundations.
Adjacent to the west of the winepress, vestiges of a Byzantine building were preserved, consisting of a wall abutted by a tamped earth floor. Two smashed Gaza jars were found on the floor, and the upper half of another jar was sunk shoulder-high in the floor (Fig. 7). North of these remains, another Byzantine refuse pit was uncovered.
Byzantine architectural remains were also encountered outside the bathhouse. A small structure found next to and north of the reservoir, had walls abutted by a coarse floor made of stones bonded in mortar. A small oven built of fired clay bricks dismantled from the bathhouse was discovered on the floor. The building’s construction probably damaged parts of the furnace area and reservoir on the bathhouse’s western fringes.
A cist grave encountered on the northwest side of the Roman residential and storage building should probably be attributed to this period, or later. The grave was built on a north–south axis and covered with five flat stone slabs; its contents were not excavated.
A few potsherds and other finds from the Early Islamic and Ottoman periods retrieved from the surface layer indicate a temporary presence at the site, probably related to agricultural activities, and possibly also to the plundering of building stones from the ancient buildings.
At this stage, it is difficult to determine whether the excavated site was at some time part of the settlement that existed at Khirbat Khaur el-Bak, or whether it was an independent settlement. The Hellenistic settlement layer points to the site’s antiquity, and also contributes important information regarding the settlement pattern in the Ashqelon region during this period. There was apparently a settlement gap at the site from the end of the Hellenistic (Hasmonean) to the Early Roman period. During the Roman period, the site was occupied by a farmstead or agricultural estate, whose economy was also based on wine production. The farm was abandoned toward the end of the Roman period, for a reason that has not yet been established. Following another settlement gap from the fourth century to sometime in the fifth century CE, the site was resettled in the Byzantine period and continued to be occupied until the end of the Byzantine or early Umayyad period. It appears that during this time, the remains of the Roman buildings and their surroundings attracted sparse settlement and were used to dump refuse, and possibly also for burial; during this phase, the site may have been part of the northern fringes of Khirbat Khaur el-Bak.