A circular installation (diam. 5.5–6.0 m, preserved depth 0.3–0.5 m; Fig. 2), hewn in the chalk bedrock, was excavated in the middle of the slope. Its sides were coated with two layers of plaster (upper layer thickness 3–7 cm, bottom layer thickness 2–5 cm; Figs. 3, 4). Similar installations were excavated on the southeastern slope of the tell (HA-ESI 111:62*–64*; 114:72*–73*).
The northern bottom part of the installation (L105) was damaged by probe trenches that were dug prior to the excavation. A surface (L103; 1.7 × 2.2 m) of small fieldstones, some of which were coated with layers of plaster, identical to those on the sides of the installation, was discovered in the southern bottom part of the installation. An examination of the plaster revealed that it was not hydraulic and therefore, the installation was not used for liquids but probably to stockpile grain, like similar installations that had previously been excavated at Tell el-Ful (N. Lapp, 1992. Tell el-Ful. In E. Stern [ed.] The New Encyclopedia of Archeological Excavations in the Holy Land, 4, pp. 1307–1308, Jerusalem).
The installation was filled with potsherd that mostly dated to Middle Bronze II, with a few dating to Iron II. The potsherds were probably swept into it from the upper part of the hill where remains of these periods had been exposed in the past, such as the cave, excavated in 1996 by S. Gibson and Z. Greenhut (Permit No. A-2506). The later potsherds probably indicate the end of the installation’s use.