Area A
Thirteen squares, divided into two sub-areas (A1, A2) were opened in a zone where the natural chalk bedrock is in close proximity to the surface.
Area A1
Nine squares were opened and immediately below topsoil, a paved surface (7.5 × 23 m) with neatly arranged limestone fieldstones, interspersed with partly hewn stones that could be in secondary use from an earlier building, was exposed (Fig. 2); it was
oriented north-south. It was clearly delineated only on the eastern side, by a straight borderline, consisting of a single row of medium-sized fieldstones (W10; length 13.5 m). The function of this pavement is not clear; it may have formed part of an ancient road or was part of an extensive platform of unknown purpose. Only a few potsherds were found on and in-between the stones of this pavement. Two cross-sections, perpendicular to each other, were cut through the stone surface in an attempt to find in situ potsherds below it. A thin layer of soil fill that contained potsherds, none postdating the first–second centuries CE, was discovered in both sections, directly set on top of the natural bedrock. It thus provided an ad quem date for this stone pavement. Outside the confines of the actual stone surface, Early Bronze Age IB potsherds were found wherever deeper lying bedrock pockets were exposed.
Area A2
Soft chalky bedrock was exposed in all three squares opened in this sub-area, just a few centimeters below surface. Except for a north-south deep cut in the bedrock (length 5.5 m, width 1.5 m, depth c. 1 m), no other features were discerned.
This apparently man-made cut yielded, besides a few potsherds dating to EB IB, three large, round and heavy stones, similar in shape to scores of smaller round (sling?) stones that were discovered in nearby Area B (see below). The function of this partially exposed cut remains enigmatic.
 
Area B
Nine squares were excavated down to natural bedrock, which consisted of Eocene chalks. It had a gentle northwest-southeast gradient (29.8 m asl in northwest; 28.7 m asl in southeast). Today, this soft and pocketed bedrock surface is covered with a layer of colluvial sediments (thickness c. 0.25–0.50 m). Three features dating from two distinct periods were excavated.
A local spread of small lime stones (thickness c. 0.25 m) just below surface was exposed in three adjoining squares. Numerous potsherds on and between the stones, dating to Early Bronze IB (c. 3000 BCE), were found. Below this stone spread more EB IB potsherds, directly lying on top of bedrock, were discovered. Building remains were nonexistent. Similar, discontinuous spreads of small stones, between various structural remains, were exposed over larger areas in Areas C and D (see below). Still unclear is the presence and nature of scores of mainly small (diam. c. 5 cm), heavy, round stones (quartzite?), found in the fill just above and immediately resting on the natural bedrock. Because of their rather uniform size, weight, and sometimes visible signs of pecking, it is clear that these natural stones had been selected by ancient man. They are too small to function as pounding stones, but they could have been used as sling stones for hunting or warfare.
A narrow bedrock cavity (depth c. 3 m), whose original entry was from the southwest, was excavated in one square. A cache of ritual ceramic vessels (Fig. 3), dating to Late Bronze II (LB II), based on the locally-produced and the imported ceramic vessels, was uncovered. Evidence of EB I occupation was on top of this cavity’s roof, which had collapsed in antiquity, thus creating a bedrock depression that was filled in and sealed by fine, colluvial sediments over time.
A small bedrock depression in an adjoining square, which contained additional LB II ceramic vessels, including numerous disc-based bowls, a pilgrim flask and a dipper juglet, was excavated. This natural depression is located just 4 m to the north of the cache cavity and the finds therein should, most likely, be associated with the cache. The finds are possibly leftovers of some ritual or ceremony performed during the time when the cache of ritual vessels was deposited inside the cavity.
 
The LB II Cache
This assemblage of cultic ceramic artifacts comprises more than 200 complete specimens, including many different kinds of pottery vessels, locally produced and imported from the eastern Mediterranean, i.e., Myceneae and Cyprus. It apparently belonged to a small local temple or shrine, as yet unidentified. The local pottery vessels included a human head-shaped cup or face-goblet (Fig. 4), four tall, cylindrical stands (braziers), with round or rectangular fenestrations and scores of chalices and small, plain stands, incense burners and oil lamps, disc-based bowls, goblets, plain jugs, small dipper juglets and a few cooking pots.
The imported pottery vessels included several Cypriot White Slip bowls, White Shaved juglets, and Base Ring bowls, as well as seven Mycenaean vessels: four stirrup jars, a piriform jar and two flasks which provide a clear date within LB II for the cache. They also point to the regional interconnections prevalent at the time. The only non-ceramic object in the cache is an imported, hemispherical, flat-based, faience bowl. Metal or stone finds are conspicuously absent from the assemblage.
 
Although the cache as a whole is instructive with regard to actual cult rituals practiced in a local temple or shrine, as yet unidentified, it neither reveals any specifics about the deity or deities worshipped, nor about the worshippers’ identity. The existence of a possible consecrated structure can be inferred from some of the objects in the cache, particularly the large cult stands and those concerned with libation and the burning of oil and incense.
Samples from many vessels in the cache have been taken for residue, petrographic and NAA analyses, to establish their original contents and provenance.
 
Area C
Forty-two squares in a row were opened (length 130 m; Fig. 5).
Natural bedrock was exposed c. 2 m below the upper clayey deposits of grumic/alluvial soils. A single, nearly continuous level (depth c. 0.5 m below present surface) of small stone spreads, associated with a few stone-built walls, installations or structures, as well as high densities of mainly potsherds, was exposed in each square. Flints are few and animal bones are almost entirely absent. An appreciable amount of ground stone tools was uncovered, foremost basalt grinding slabs/stones, but also the upper part of a tournette in secondary deposition.
These stony surfaces are slightly undulating, due to post-dispositional sub-soil warping (see below). Structural remains included segments of two parallel stone walls, each built of a double row of fieldstones and preserved to a maximum of two courses high, perhaps the remains of a partially preserved building. These wall remains were apparently later integrated into a rectangular stone-built structure or installation, with a carefully laid, stone-paved floor (Fig. 6).
Scant remains of two additional, likely installations, which also had a stone-paved floor and bordered on the eastern perimeter of the first structure, were exposed. This particular area is delineated toward the north by an east–west oriented, slightly curvilinear wall, built of a single row of stones and preserved two courses high. The wall extended over three adjoining squares and protruded from their extreme opposite balks. While the interior face of this wall that faced the three installations is fairly clear and pronounced, its external face has numerous small and large fieldstones that were seemingly piled up in disarray alongside it. A single, cross-section through this wall shows that it is resting directly on sterile, grumic soils, whereas the stone debris piled up against it is overlying an apparent depression (channel? drain? pit?) filled-in with small fieldstones. Another, circular, stone-paved installation or structure, whose western part protruded from the balk, was partially exposed in the eastern extreme and outside of this lengthy wall. To the north and south of this area, additional uneven small-stone surfaces, some continuous and some intermittent, were exposed. Small segments of stone walls were occasionally discerned in association with the stone surfaces and at least in two instances, in-situ basalt grinding stones/slabs and non-stationary limestone mortars were discovered, indicating an area of food processing.
Based on associated pottery finds, the majority of which consisted only of two types of storage jars, i.e., holemouth and bow-rim jars, this level is dated exclusively to EB IB.
It is interesting to note the nearly complete absence of ground stone tools from the contemporary habitation levels on the tell; a single basalt grinding stone fragment was found in EB IB settlement remains. This could perhaps imply that household activities, such as the grinding of grain and the preparation of flour, was carried out on the outskirts of the actual habitation zone, rather than on the tell itself.
Post-dispositional motions, possibly due to sub-soil warping and/or earthquakes, observed also at other nearby, contemporary archaeological sites (e.g., ‘En Shaddud), resulted in severe distortion of these poorly preserved structures, which hinders an overall interpretation of the remains.
 
Area D
Six squares were opened. Below a top layer of grumic soils and various, horizontal spreads of recent, chalky stone debris, with a few embedded modern iron rods and glass, as well as a few Middle Paleolithic flints (Levallois techniques) in obvious secondary deposition, a single level of small-stone surfaces and wall remains, similar to that in Area C, was exposed, resting directly on sterile grumic soils (Fig. 7). As in Area C, the pottery associated with this layer is dated to EB IB. Although excavation was less intensive in this area than in Area C, due to lack of time, it seems that the EB IB level exposed in this area is a further, southward extension of the activities areas uncovered in Area C.
 
The data from Areas C and D provide a glimpse into the so-called activity areas for processing and storing food produce away from and at the fringes of the contemporary EB I settlement, situated on Tel Qashish. Since part of this settlement was excavated in the 1970s (Ben-Tor 2003), the new data will enable to compare the published tell material concerning the village where the people lived, with the area where the very same people had worked.
The unexpected discovery of a LB II cultic cache in Area B is an extra and welcome boon to the research of this period. Apart from the obvious conclusion that a local LB II temple or shrine must have existed in the near vicinity of the present site, perhaps near the unexcavated apex of the nearby Tel Qashish, a careful study of the wealth of finds from this cachewill eventually provide a deeper understanding of the actual cult practices in a provincial LB II temple or shrine.