Table 1. Excavation areas
Area Size (sq m)
Common Finds
Late Natufian
Lunates with an abrupt back, microburins
Blades with retouched back, sickle blades, microburins
Mixed finds
Microlithic tools, ad-hoc tools, pottery
Mixed finds
Microlithic tools
Early Natufian
(14C dating)
Ostrich-egg shells
Late Natufian
Lunates with an abrupt back, a pit containing debitage
Harif points, small lunates
Geometric Kebaran
Geometric microlithic tools
Harif points, small lunates
Harif points, small lunates
Area A yielded red chalcedony flint items from the Late Natufian culture, with debitage that included bladelet cores and bladelets, microburins (Fig. 2:1–3) and lunates with an abrupt, retouched back (Fig. 2:4–7). Also recovered were two Helwan lunates (lunates with bifacially retouched backs) from the Early Natufian culture; they may have come from the adjacent area (Area B, below). Further finds included round flint hammerstones (Fig. 3), burnt stones and a few Dentalium beads (Fig. 4).
Area B yielded sparsely dispersed light-brown or brown-and-gray flint items. The finds include knapping debitage containing blades and bladelets, microburins, and blades with an abrupt, retouched back, some of which bear sickle gloss (Fig. 5:2–5) and may be sickle blades. Blades of this type are the most common tools in this area. Also recovered were sickle blades with a worn, curved cutting edge and sporadic sickle gloss. The blades with an abrupt, retouched back and the sickle blades are not truncated, and their cutting edge is not retouched; they probably date from the Epipaleolithic period. Three Helwan lunates (Fig. 5:1) from the Early Natufian culture were also discovered, reinforcing the suggested date of the blades. However, the Helwan lunates are made of red chalcedony flint, which is not characteristic of Area B.
Areas C and D. The gentle slope of a hill damaged by IDF activity yielded flint items from various periods. They include a tabular scraper, lunates made of gray flint with an abrupt back, and brown flint bladelets with a retouched back. A base of a Chalcolithic or Early Bronze Age jar (Fig. 6) was found beside the scraper.
Area E. A large area (over 100 sq m) of scattered fragments of ostrich-egg shells was examined (Fig. 7). Two eggshell samples were taken for radiocarbon dating at the Weizmann Institute, yielding the following results, after calibrating two standard deviations: 14,850–14,190 and 14,530–14,100 BP, i.e., the Early Natufian culture. It is not yet clear if the shells are from an abandoned nest or if they are the result of the deliberate collection of ostrich eggs. A few flint items were found scattered in the area, the most common being lunates with an abrupt back resembling those found in Area A.
Area F yielded a pit containing dozens of debitage items (Fig. 8). A fragment of a core was found not far from the pit. The surface finds include items produced from flint that differs from the type of flint discovered in the pit; the latter include debitage, lunates with an abrupt back and other tools associated with the Late Natufian culture.
Area G. Three flint concentrations (G1–G3) were discovered on a gentle slope in the north bend of a long, narrow, hill. Harif points and small Harifian lunates (Fig. 9) were found in concentrations G1 and G3. Also found were debitage, including bladelets and bladelet cores (Fig. 10), a grinding stone, and a large and wide Dentalium bead (Fig. 11). Concentration G2 yielded a number of Geometric Kebaran tools, among them geometric microlithic tools such as rectangles and trapezes (Fig. 12:1–5), scrapers (Fig. 12:6, 8) and a drill (Fig. 12:7). A modern copper pot that had been buried in the ground with its mouth facing downward (Fig. 13) was discovered in a probe dug on the lower slope of the hill. An excavated trench (below) yielded a limestone pestle (Fig. 14) and a hammer stone.
On the hillside north of concentration G1, a trench (0.5–3.5 m deep) was dug mechanically for the purposes of a geomorphological study. Layers of clay exposed in a section of the trench and in other probes indicate the pooling of water. The Nahal Besor streambed was probably blocked by a migrating sand dune, forming a seasonal lake.
Area H. Harif points and tiny Harifian lunates (9–12 mm long) were discovered at the foot of the hill where Area G was excavated. Also found here were bladelet cores, bladelets, a pebble incised on both sides (Fig. 15), a concentration of burnt pieces of limestone, possibly part of a hearth, and Dentalium beads. This concentration of items was found near the end of the excavation, and therefore only part of its was collected.
The excavation recovered flint artifacts dating from the late Epipaleolithic period—the Late Natufian and Harifian cultures—the majority of which are microliths which were used for hunting. The tools were apparently knapped at the site, as all the areas, except for Area E, contained debitage from all production stages. The raw material was brought from flint outcrops at the nearby Yeroham Ridge (Rekhes Yeroham). The mapping of the ancient lake has indicated that the flint concentrations were found along its banks. Human activity most probably took place around the lake following the winter season, when the lake accumulated sufficient water to support the surrounding vegetation and attract wildlife; such a lakeside environment would have attracted bands of hunter-gatherers. The link between seasonal lakes in the desert and transitory camps of hunter-gatherers has been studied in the past (Goring-Morris and Goldberg 1990; Roskin et al. 2014). Another Late Natufian site that attests to this association was recently discovered in Nahal Sekher (Barzilai et al. 2015).