The excavation unearthed nineteen walls or wall segments. These comprise the partial remains of several, apparently one-room rectilinear dwellings with rounded corners (Figs. 2, 3). The orientation of the structures is similar to those excavated in the past on the tell (Ben-Tor, Bonfil and Zuckerman 2003
). Not one of the structures was found intact, and only the first course of their stone foundations was preserved; mud-brick debris found inside three of the better-preserved structures clearly indicates that the stone foundations carried mud-brick superstructure. The only preserved floor was paved of carefully placed large flagstones (Fig. 4).
Though far from complex, the internal stratigraphy of the site could not be clarified in its entirety, due to time constraints. The architectural remains are contained within a single stratum, but there are at least two separate building phases, as can be seen where a wall or walls of one building cut into those of another. As far as it was possible to ascertain, the earlier building remains rest directly on the sloping, virgin soil consisting of weathered basalt, whereas the later remains rest on a thin layer of dark colluvial soil mixed with sherds. In the westernmost part of the excavation area, toward the hilltop, several narrow wall segments were unearthed. They were embedded in extensive dispersals of unworked chalk and nari stones mixed with large concentrations of clearly in-situ pottery. These walls might belong to a third phase or possibly a second stratum that is later than the two previous phases. The massive stone debris and the pottery accumulations that consist of large, layered fragments might have been caused by an earthquake.
Approximately 85% of the almost 30,000 potsherds collected during the excavation are body sherds. Of the remaining 15% indicative pottery fragments, a large majority comprises holemouth jars and bow-rim storage jars, with only a small representation of smaller vessels, such as Gray Burnished Ware bowls, a few knobbed, hemispherical bowls and red-polished ‘tea-pots’ with bent spouts. One red-polished neck or shoulder fragment of a medium-sized jar bears a cylinder seal impression (Fig. 5), showing two four-legged animals, one with two long horns. An identical cylinder-seal impression was found (ex situ) more than eighty years ago at Megiddo. It would seem that both examples were rolled from the same cylinder seal. The groundstone tools consist mainly of basalt grinding stones, as well as a single complete basalt bowl. All these finds clearly point to an EB IB date. Other, sparser finds include flint tools and animal bones, indicating a mixed subsistence of farming and herding at the site.
The excavation has shown that the rural EB IB settlement previously unearthed on the tell (Stratum XV; Ben-Tor, Bonfil and Zuckerman 2003
) extended northward, well beyond the tell’s immediate confines. The settlement remains uncovered in the excavation described here are characterized by dwellings with rounded corners, stone-built foundations and mud-brick walls. While the custom of building houses with rounded corners is said to be the last echo of a curvilinear building tradition introduced during EB IA, it is of interest to note that EB IA remains are absent from both the present excavation and the excavation on the tell itself. However, whereas the EB IB settlement on the tell continued into the succeeding EB II–III periods, the EB IB settlement on the eastern slope of the chalk hill was apparently abandoned. The indications of a massive destruction, possibly by an earthquake—maybe the same that destroyed the nearby, contemporary settlement at Tel Megiddo—may hold the clue to this abandonment.