Two squares were opened in the first season, revealing two large walls, apparently from the same building (Fig. 1). The southern wall (Sq A, W3; width 1.5 m; Fig. 2) was preserved more than 2 m high. A vault, comprising the upper 30 cm of the northern face of W3, arched toward Wall 1 (Sq B) in the north. Only the vault of W1 that used the upper 65 cm of the wall was exposed. Both walls were built of rough boulders and smaller stones, standing 7.1 m apart; they were, most likely, abutted by a beaten-earth floor (L10), which was exposed at the bottom of a 2 m wide probe against W3. 


On top of the floor were two layers of stone collapse, with a layer of loose soil between them, in a matrix of compact brown soil. The lower collapse layer (L9) probably derived from the vault itself, as was clearly visible in the section of the probe (Fig. 3). The loose soil above it (L7) contained a tile fragment with a stamped Hospitaler's cross (Knights of the Order of the Hospital of St. John of Jerusalem; Fig. 4). The Master of the Order of St. John, Fra Raymund du Puy, introduced the eight-pointed cross in c. 1128, as an alternative to the traditional Latin cross. An enigmatic ostracon, bearing a Second Temple Hebrew inscription, incised before firing, was also found (Fig. 5). The upper collapse (L6, L8) was possibly from a second storey; it contained two coins (Figs. 6–8). A natural fill and topsoil covered the upper collapse, including, among other finds, a tile with a circular stamped impression and an unidentifiable coin.


None of these finds can accurately date the building. However, the nature and size of the walls, as well as comparisons to other structural remains in the vicinity indicate that the walls can probably be dated to the Crusader Period.


During the second season, the probe in Sq A was deepened 1.30 m. Locus 10, renamed L13, was thought to be a floor, yet turned out to be part of the third layer of stone collapse, containing Mamluk pottery. The fourth layer of stone collapse, carried medieval pottery and the deepest fifth layer of stone collapse yielded Ottoman pottery. While none of these loci were clean, they showed that the building stood until recent times and was occupied as late as the Ottoman period.


The removal of the general collapse of the Crusader building in Sq B showed that the lines of collapse were perpendicular to the walls of the building, which may indicate that this was the collapse of the exterior western wall. Square C, which was opened to the east of Sq B (not marked on Fig. 1), revealed lines of collapse that were parallel to the walls, attesting to the roof collapse. Among the potsherds in Sq C was a distinctive group of Ottoman fragments that contained a lot of sand; they were probably manufactured along the coast.