An area (6 × 22 m, max. depth 1.3 m) was opened and a long wall, oriented northeast-southwest (W13; length 21.5 m, width 1.3 m; Fig. 2), was discovered within heavy alluvial soil in its midst, close to the surface. Its northeastern end had been destroyed by earthmoving equipment and its southwestern end continued into the balk. Both sides of the wall were built of large fieldstones with straight outer surfaces (0.35 × 0.50 × 0.60 m) and smaller stones in-between (Fig. 3). The northwestern side of the wall was partially destroyed. The southeastern side of the wall was better preserved (max. height 0.6 m) and its stones were arranged with their narrow side facing out. Another wall (W17; length 7 m; Fig. 4), built of large fieldstones, was discovered next to the western side of and parallel to W13. Collapsed stones were discovered in its northern section, but it is clear that they were part of the wall (see Fig. 3). A large amount of alluvium together with small irregular shaped fieldstones had accumulated southeast of W13 (L11, L14; Fig. 5). The alluvium contained worn potsherds that dated to the Early and Late Roman periods. A few Roman potsherds were also recovered from the dismantling of the wall.
Some 443 flint artifacts were found, most of them flint nodules (Table 1) that were not knapped or were naturally broken, and chips that were mostly natural. Most of the flint artifacts have a single or double layer of brown patina, and an overwhelming number of them are worn, indicating they were transported there from another site. These features are characteristic of lithic finds, regardless of their location or the depth of their loci in the excavation. It should be noted that the flint occurs together with the Roman-period ceramics throughout excavation. 
Table 1. Frequency of the Flint Artifacts
Primary flake
Natural nodules
The flint finds from the excavation are typified by a multitude of natural nodules and chips and a low occurrence of tools and cores. The cores and flakes are characteristic of a Levallois industry common to sites of the Middle Palaeolithic period. This is a flake industry and its purpose was to produce broad thin flakes. The dorsal scars on the flakes are mostly converging radial scars and a triangular-shaped scar is in the center of the flake. The patina and weathering on their ends indicate that the flint artifacts were not found in situ. The excavation was carried out in an ancient streambed and it is therefore likely that the artifacts were transported by the stream from another location, probably from the Sollelim Hills, east of the site. This phenomenon is typical of most of the alluvial plains in the Lower Galilee. The non-worm artifacts are but a few. They mostly come from the surface layer in the area and are a commodity of an industry that specialized in the production of blades, whose source was apparently the nearby site that is dated to the Pre-Pottery Neolithic B period.
The wall is apparently a diverting dam that was built in the channel of a small winter tributary. It was properly constructed but not plastered and it meant to stop the erosion and regulate the flow of water, preventing damage to agriculture. At the spot where the current was strongest, probably in the middle of the tributary, another wall was erected next to the western side of the dam, to reinforce and support it. The Roman-period potsherds found in the alluvium were probably swept along by the water from a Roman site that is located to the east. The Roman sherds recovered from the wall of the dam date the time of its construction to the Roman period. Most of the retouched flint artifacts are from the Middle Palaeolithic period and were transported from their original provenance by the flow in the stream.