Area A
The Excavation. At first, 18 squares were opened over an area of 40 × 70 m; later, after ancient remains were uncovered in only the southern part of the area, three additional squares were excavated (Fig. 2). In 14 of the squares, which were initially excavated manually and then with mechanical equipment, no ancient remains were exposed. An upper layer of red hamra and lower layer of brown hamra overlay the kurkar bedrock, discovered at a depth of c. 3 m. In seven of the squares ancient remains were exposed at the interface between the two hamra layers.
 
In Sqs J8, J9 and K9, a foundation wall of a square building (W700; Figs. 3, 4), founded on sterile hamra soil and built of partially dressed kurkar stones, was exposed. The wall was preserved to a height of two courses (c. 0.4 m). A poorly preserved floor bed (L210) built of small- and medium-sized kurkar stones abutted the wall. Ceramic finds dating to the first–second centuries CE were discovered beside the wall. West of the building, near the excavation’s western balk, was a refuse pit that contained small pottery sherds, also dating to the first–second centuries CE.
In Sqs G10 and G11, a pavement built of kurkar slabs (L208; c. 2 sq m; size of slabs c. 0.3 × 0.6 m; Fig. 5) was exposed at a depth of 0.45 m below the surface; it was founded directly on top of the hamra soil, and is probably the poorly preserved remains of a street pavement. A stone socket, apparently preserved from a building threshold that was situated near this street, was discovered next to the stone surface. The remains in the two squares were damaged by erosion and might also have been robbed. The scant amount of non-diagnostic ceramic finds could not date the stone surface. Nevertheless, the finds elsewhere at the site substantiate the supposition that these remains are from the Roman period.
In Sq I2, the foundation of a corner of a building (Fig. 6) founded on sterile soil and built of well-dressed kurkar stones bonded with plaster was exposed. The methods of construction and stone dressing in this structure were different from those of the building in Sqs J8, J9 and K9; thus, it is reasonable to assume that the two structures are not contemporary. Apart from a coin dating to the British Mandate discovered between the foundations, no artifacts were found near the building remains. This coin seems to date the structure to the modern era. Several stones that may represent the continuation of the building foundation revealed in Sq I2 were exposed in the southeastern corner of Sq J4.
 
The finds in Area A include pottery sherds, mainly jars, a stone stopper and several fragments of glass vessels. Most of the finds were discovered on the floor of the square building in Sqs J8, J9 and K9, on its floor bed and alongside it. The following ceramic artifacts are noteworthy: a jar (Fig. 7:1) with a slightly everted cut rim, thickened at the point where it connects to the neck, and a slightly curved neck lacking a ridge where it joins the body; a jar (Fig. 7:2) with a triangular rim, long neck and a ridge where it connects to the body, a type that was widespread throughout the country from the first until the mid-second century CE; and a casserole lid (Fig. 7:3) with a cut rim, ornamented with delicate combing, typical of the Roman and Byzantine periods. A stone stopper (Figs. 7:4; 8) dating to the end of the first century–early second century CE was discovered next to W700 in the square building, along with the pottery vessels. Similar stoppers were discovered in Jerusalem, Judah and the hiding refuges in the Judean Shephelah (Kloner and Tepper 1987:44). Glassware from the Roman period includes bowls with a curved rim and a linear decoration below it (Fig. 7:5), bowls with a folded rim (Fig. 7:6, 7), a flat bowl base (Fig. 7:8), a bottle or goblet (Fig. 7:9) and a goblet base (Fig. 7:10). The finds date the building to the Roman period.
 
Area B
The Excavation. At first, ten squares were opened over an area of c. 25 × 55 m; later, after ancient remains were discovered, three additional squares were excavated (Fig. 9). Seven squares were devoid of any finds from the surface level down to the kurkar bedrock (depth c. 1 m).
A segment of a wall foundation (width 0.6 m) was exposed in Sq S33. It was built of coarsely dressed kurkar stones (c. 0.2 × 0.5 m) placed directly on the bedrock in a northeast–southwest direction, and was preserved to a height of one course. A dark level, apparently indicative of a fire that occurred nearby in the past, was discerned in the southeastern balks of Sqs R33, S33 and T33.
In Sqs T30, T31, U30 and U31, remains of the bottom part—the firebox—of a round potter’s kiln (L113; diam. 2.9 m; Figs. 10, 11) were exposed. The kiln wall (width 0.5 m) was built of kurkar stones preserved c. 1.5 m high. The firing chamber, which was usually built above the firebox, did not survive, except for the bottommost part of the wall; it was treated with light colored plaster. In the middle of the firebox were remains of a column built of plastered, coarsely dressed kurkar stones, which were preserved to a height of c. 0.7 m; the column evidently supported the floor of the firing chamber. A stokehole (width 0.4 m, height 0.6 m; Figs. 12, 13), whose upper part was triangular, was set in the eastern side of the firebox. It seems that the firebox was subterranean and the stokehole was accessed through a pit dug in the ground.
Refuse pits (L114, L125, L128; Fig. 14) belonging to the potter’s workshop were discovered to the east and west of the kiln. They were found filled with a large amount of broken pottery vessels and ash. A pavement of kurkar fieldstones (L122; stone dimensions 0.1–0.2 m) treated with a thick, coarse layer of plaster abutted the kiln on the south; the pavement was exposed at a depth of 0.1–0.2 m below the surface. It seems that the pavement was at the level of firing chamber’s floor, probably leading up to it so as to load it with vessels prior to firing and to empty it afterwards. Several coarse tesserae were discovered not in situ on the paved surface, and the plaster layer above the stones may have originally served as a foundation for a coarse, white mosaic floor.
The excavation of the kiln enabled us to reconstruct its history. It seems that at first a large pit (Fig. 15:b) was dug in sterile soil (Fig. 15:a). A circular firebox (Fig. 15:c), supported by crushed kurkar fill deposited north and south of the kiln (Fig. 15:d), was built at the bottom of the pit. The areas east and west of the kiln remained empty in order to insert the fuel in one side (Fig. 15:e) and remove the debris from the other side (Fig. 15:f). In the next stage, the area south of the kiln was paved (Fig. 15:g); now the installation was ready for operation.
Interestingly, a floor abutting a kiln’s firing chamber was also described in Kiln K6 in the pottery workshop of the Tenth Roman Legion at Binyene Ha-Umma in Jerusalem (Goldfus and Arubas 2002:116, upper right illustration). Although the kiln in Jerusalem produced mostly roof tiles and bricks, it seems the work there was conducted in the same manner as in other ceramic kilns. While the kiln was used and probably afterwards as well, industrial debris consisting of a large quantity of wasters (Fig. 16) was discarded into refuse pits behind the kiln, to its west (Fig. 15:h) and later in the area near the stokehole as well (Fig. 15:e). After the installation was destroyed, a layer of soil (Fig.15:i) accumulated over the kiln and throughout the area.
 
The Finds in Area B include pottery sherds, discovered mainly in the refuse pits near the kiln (L114, L125, L128); some of the vessels were found in the kiln’s firebox (L113), outside the kiln (L121) and on the plaster floor next to the kiln (L122). Jars constituted the overwhelming majority of ceramic finds in Area B (over 90%), thus it seems that the workshop specialized in producing them. The jars belong to three main types: jars with a triangular rim, jars with a straight rim, and those with an everted rim. Jars with triangular rims (Fig. 17) are characterized by a long, upright or sometimes wavy neck, an out-folded rim that sometimes forms an everted triangular shape, and a ridge where the neck connects to the body (Fig. 18). These jars date to the Roman period, particularly to the first and early second centuries CE. The second group of jars (Fig. 19) are characterized by a straight rim that is sometimes slightly thickened at the end as a result of an outwards fold of the rim. It also has a ridge where the base of the neck joins the body (Fig. 20). Like the first group, these jars date to the Roman period, mainly to the first and early second centuries CE. The jars of the third group (Fig. 21) are characterized by a slightly everted rim or neck and also date to the Roman period, mainly to the first and early second centuries CE.
The ceramic finds from the area include several other vessels, notably a bowl with a flared rim and a carinated body (Fig. 22:1), dating to the second half of the first century CE; the base of an amphora (Fig. 22:2) imported from Italy and dating to the first century CE; and a single sāqiye jug with a perforated base (Figs. 22:3). Sāqiye jugs were used to draw water from wells beginning in the Roman period. Since these vessels were used for other purposes as well, we can conclude they were used in a sāqiye water-wheel only if found in large numbers (Ayalon 2000:216). A coin dating to the third century CE was discovered on the surface in Sq M33.
 
The excavation revealed the remains of a settlement from the Roman period (first–second centuries CE) and meager remains of a building that is difficult to date yet seems to be modern. The Roman-period settlement was divided into a residential area (Area A) and an industrial area (Area B). The remains of the residential area comprise part of a building and a section of a paved street that apparently ran between buildings; in the industrial area a pottery kiln was uncovered. The ceramic finds indicate that the pottery workshop specialized in producing jars. Thus, it seems that there was a special demand for such vessels in and around the site, perhaps due to the need to store the local agriculture produce.
Despite the consensus that the coastal plain suffered a settlement crisis in the late first and second centuries CE, perhaps as a result of the Great Revolt (Roll and Ayalon 1989:141), surveys and excavations, including this one, have revealed that several settlements continued to exist. Some of these settlements expanded, and new ones were even established along the national roads that passed through the coastal plain (Roll and Ayalon 1989).