Hill B––Cupmarks and Rock-Cut Installations
Separated from Hill ‘A’ (see HA-ESI 117) by a narrow east-west oriented shallow wadi, Hill ‘B’ was noted for similar barren outcrops of limestone bedrock, some bearing flint nodules. Sixty-four features, hewn in or built on bedrock outcrops and spread over various descending terraces on the east slope of Hill ‘B’, were identified. 
The features can be divided into the following subtypes:
(1) Oblong-oval, shallow cupmarks (Fig. 1), appearing usually in clusters and very similar to those exposed at Nevallat (HA-ESI 117) and Kh. el-‘Alya (ESI 17: 94–104). As in the case of Hill ‘A’, these constituted the majority of cupmarks that were probably used for grinding activities, extending over a long period of time.
(2) Round, deep cupmarks, found in association with the oblong-oval cupmarks (Subtype 1). They may represent mortars used for pounding and crushing, especially when found together with the Subtype 1 cupmarks.
(3) Twenty dispersed large circular vats or basins were recorded (diam. over 1 m), in addition to 35 similar large vats and basins documented previously on Hill ‘A’ and another 8 that were noted on Hill ‘C’ during the current season. The shallow vats (diam. c. 1.2 m) served perhaps as threshing floors. Deeper, and somewhat tapering ones (basins) were smaller (diam. c. 0.8–1.0 m). At least some of these vats predated the Subtype 1 cupmarks that had cut them. If we were to assume that Subtypes 1 and 3 were contemporary, one might expect them to always appear together at the same site, which is not the case. To date, twelve sites are known to have oblong-oval cupmarks, similar to Subtype 1, but only the Modi‘in area (including Mevo Horon) is known to exhibit large vats/basins and oblong cupmarks.
(4) Three man-made, hewn shaft-like features that were not explored to full depth due to time constraints.
Based on pottery and flint tools found in association with some of the features on Hills ‘A’ and ‘B’ and corroborated, in retrospect, by results of the excavations at Nevallat, the majority of rock-cut features did not postdate the Late Chalcolithic period, i.e., they could not be later than the early to mid-fourth millennium BCE and some might be earlier, in particular the Subtype 3 basins. A Chalcolithic date for the majority of Subtypes 1 and 2 cupmarks appears to harmonize with the dating of two copper adzes in excellent condition (Fig. 2), found hidden away together in a small crevice outside one of the rock-cut installations on hill ‘B’.

Hill ‘B’––Stratified Deposits
Patches of what appeared to be ancient walls and floor remains were exposed by rainfall along the depression between Hills ‘A’ and ‘B’ and noted during the 2003 excavations. It was conjectured that this depression was likely to contain in-situ remains of habitations because it was low-lying and protected. A similar topographic situation existed between Hills ‘A’ and ‘C’, where mechanical probing revealed thick anthropogenic deposits, dating from the Chalcolithic and Early Bronze I periods (E. Yannai pers. comm.). Mechanical exposure of Hill ‘B’  revealed in one trial trench over 2.5 m of anthropogenic deposits in a depression that were subsequently, manually excavated. The deposits presented a remarkably well-preserved superimposed series of occupations that evidenced a sequence of seven strata (height over 3 m), each containing remains of stone architecture and associated fills. The strata are described from below upward.


Stratum 7
Bedrock, gently sloping down from northwest to southeast and from east-northeast to west-southwest, was exposed in most of the area at c. 3 m below surface. The highest part of bedrock in the northwest area of the excavation was at 248.95 m asl. The lowest parts in the southeast and southwest areas of the excavation, ranged from 247.26 to c. 246.50 m asl. The northeast–southwest bedrock gradient of the depression was much steeper than the general northwest–southeast gradient of the surrounding Hills ‘A’ and ‘B’. It is highly possible that the original bedrock in the depression was part of a collapsed (?) karstic feature. At the deepest point within the excavated area (c. 1.5 × 2.5 m), bedrock was covered with a layer of sterile, reddish, hamra-like soil (thickness c. 0.5 m). A small storage pit (diam. 0.8 m, depth c. 0.5 m), dug into this sterile soil, protruded from the west balk of the excavation. It was lined at the top with two rows of cobblestones. Its bottom was paved with a few flat, heat-cracked fieldstones that provided a stable and leveled surface. A small fireplace, yielding carbonized wood, was uncovered close by. The only potsherd in the pit was a large jar rim fragment of undefined Chalcolithic date. The pit and the limited area excavated around it were sealed by successive layers of carbonized organic materials, grayish soil, a lime layer and again, carbonized organic materials, all sampled for analyses. Two Chalcolithic ceramic bowls, a smaller one atop a larger one, placed upside down, were found resting on the topmost layer of carbonized organic materials. The smaller bowl contained carbonized material, including several olive pits.

 

Stratum 6
Stone wall segments that formed part of an apparently substantial, east–west oriented (broad-room?) building (6.6 × 14.6 m, height 0.8 m), which was internally subdivided into at least two separate rooms, were founded directly on bedrock. One segment rested on the upper layer of carbonized organic materials that sealed the pit of Stratum 7 and must be, therefore, later than these deposits.
A small cache of eleven, mostly intact, pottery vessels rested on a floor of small fieldstones that created a leveled space over bedrock and abutted one of the stone walls, was associated with this building. The composition of this cache––small and medium-sized bowls and a single jar with strainer (Fig. 3)––is reminiscent of a somewhat larger cache of 16 medium-sized bowls and a single jar with ellipsoid mouth, found in a contemporary layer at Nevallat.

The pottery and flint tools, including several adzes, retrieved from this building point to an early phase within the Late Chalcolithic, Ghassulian IV/Beer-Sheva cultural horizon. The general character of the pottery assemblage from this stratum appears to differ from the assemblages of the subsequent strata (5–1), irrespective of their date within the Chalcolithic or EB IA periods, in that small, open forms seem to dominate this assemblage, in contrast to large, closed forms, particularly large storage jars or pithoi, which are dominant in the later strata. This distinction may be indicative of the functional difference between the strata. 

 

Stratum 5
A straight wall segment, partly overlying and physically separated from the building remains of Stratum 6 by a thick deposit of carbonized wood and more than a thousand carbonized olive stones (Olea europaea; apparently domesticated, N. Liphschitz pers. comm.), was uncovered. Due to the relatively limited extent of the excavated area it was uncertain whether this wall and two additional wall segments in the same stratum formed part of a single building. They may be superimposed on the Stratum 6 building, whose orientation they share. The few diagnostic potsherds deriving from associated fills dated these remains to a phase within the Late Chalcolithic period.

 

The carbonized materials, which were ‘sandwiched’ between wall segments of Stratum 6 and Stratum 5, provide an absolute ante quem date for Stratum 6, and a post quem date for the Stratum 5 building. The stratigraphic and horizontal closeness of the two superimposed buildings indicates that a relatively short time elapsed between the disintegration of the Stratum 6 building and the erection of the Stratum 5 structure.

 

The repeated presence of olive pits and other carbonized organic materials in Strata 6 and 5, together with a date somewhere in the Late Chalcolithic period for both strata, suggests a functional link with the Chalcolithic rock-cut installations and cupmarks that surrounded the site to its north (Hill ‘A’) and south (Hill ‘B’).


 
Stratum 4
Stone foundations of a single broad-room (5.0 × 13.5 m, height 0.8 m), partly overlying the building remains of Stratum 6 and physically separated from them by c. 0.5 m of fill, were uncovered. The entrance to the building, still unidentified, was probably in an unexcavated segment of the long, northern wall. The northeast–southwest orientation of this building was at c. 45 degrees variance with that of the underlying architecture of Strata 5 and 6. This change in building orientation possibly reflects a lapse of time between Strata 5 and 4, which was considerably longer than that between Strata 6 and 5.
The partly excavated remains of a tightly packed and circular cobblestone floor or installation, a few meters to the south of the building, were probably associated with it.
An intact, decorated jar was found in the northwest corner of the building and apparently indicated a floor level, a few centimeters above a thick lime layer that ran toward the southeast, beneath the stone foundations of this building and above the building remains of Stratum 6. Another notable find, indirectly associated with this building, was a small, piriform and incomplete hematite mace-head.

Two stelae/orthostats
Two carefully shaped, polished and abraded stelae or orthostats (mazzevot), found in overlying EB IA deposits of strata 3 and 2, perhaps derived originally from the Stratum 4 building. A large and deep pit, dug from an EB IA level higher up in the stratigraphy, partly destroyed the stone foundations in the west part of the Stratum 4 broad-room building. One of the stelae was found broken, lying on top of the fill in this pit, while the other, dug up perhaps through the same pit, yet found at a distance of c. 8 m away, was apparently re-used as a visible marker flanking the entrance to an oval enclosure of Stratum 2.
 If the two stelae came, indeed, from the Stratum 4 building, it should be asked whether the people who erected the Stratum 2 enclosure, dating from EB IA, were aware of the sacredness attached to these stelae by the former Chalcolithic occupants of the site. A different outlook would be to consider the stelae in primary position at the entrance to the Stratum 2 enclosure and thus, a possible Late Chalcolithic date for the enclosure wall itself. This assumption is contradicted by the many pottery finds from Stratum 2, which are dated to an early phase of Early Bronze IA. The morphology of the two stelae is strikingly similar to two other stelae from a Chalcolithic burial cave near Shoham (IAA Reports 27: Figs. 9.19, 9.20) and a third similar stela that was recently discovered in a Chalcolithic burial cave at Horbat Qarqar South (P. Fabian, pers. comm.).

 

Stratum 3
Stone foundations of probably another broad-room structure built perpendicularly and slightly to the west of the Stratum 4 building and overlying part of the Stratum 5 building, were separated from Stratum 4 by fills (thickness c. 0.3 m). The structure’s dimensions (5 × 10 m, height 0.4 m) were somewhat smaller than those of Stratum 4; the width of the walls was narrower and the fieldstones used were generally smaller. Remains of a lime-washed, earthen floor were exposed in the larger of the two rooms where at least five restorable vessels were found, in situ. Among them were a medium-sized jar with well-developed, wavy ledge handles and a large, coarse pithos with applied rope decoration, the likes of which are known at Yiftah’el, Stratum II (IAA Reports 2: 9.15–9.18), dating from an early phase of Early Bronze Age I (EB IA). Another find on this floor was a small clay figurine of a bull (Fig. 4) with horns curved backward that originally probably stood in the center of a ceramic bowl or plate, like a specimen unearthed by de Vaux in Tomb 14 at Tell Far‘a North (RB 59, 1952: Pl. 14) which has definite EB IA affinities. The floor and the pottery remains above it were not associated with a curvilinear building, but with an apparently rectangular broad-room building, reminiscent of Chalcolithic precursors. As far as architectural traditions are concerned, the transition from rectilinear buildings to curvilinear ones may have been slightly less abrupt than generally assumed. Other ways of looking at the apparent dichotomy, i.e., a rectangular broad room dating to the early EB IA, are either to consider this building as of a Late Chalcolithic date, which seems to be contradicted by the pottery vessels found on its floor, or to assume that these pottery vessels began earlier, in the Late Chalcolithic, than is usually accepted, i.e., in the EB IA. Whatever the merits of the present issue, it certainly highlights the intricacies involved in the interpretation of the data collected from this transitional site. Morphologically, the pottery from this stratum fits the EB IA period, although technologically, it has strong affinities to the preceding Chalcolithic potting traditions, a situation observed also at a few other sites in the country, e.g., Ashqelon, Afridar – Area G (‘Atiqot 45:228).

Stratum 2
Stone foundation remains of a long, curvilinear enclosure wall, which damaged the rectangular, southwest corner of the EB IA Stratum 3 building and was slightly overlying part of the Stratum 4 Chalcolithic broad-room, were discovered. The enclosure wall had a well-defined entrance in its southeast part, carefully paved with flat stone slabs. Presumably, a wooden door was originally located to the left of the entrance, as indicated by the stone door-socket found in situ, which is one of the earliest examples known to date. A tall and carefully shaped stone stela, lying flat-faced, probably stood erect on the stone pavement at the exterior right side of the entrance. As noted above, this and another broken stela may have been dug up from the underlying Stratum 4 building.

 

A circular structure on stone foundations was attached to the interior face of the enclosure, at a distance of c. 2 m from its entrance. Remains of a pavement that consisted of a row of carefully arranged flag stones, along the outer face of the enclosure wall, were exposed. Although it is hard to establish a direct stratigraphic relationship between the enclosure and another, freestanding circular structure built on stone foundations toward the northwest, we assume they were contemporary, given that both the curvilinear enclosure wall and the circular wall cut into or damaged the underlying rectangular Stratum 3 building. At least three successive floor levels were preserved within the circular structure; a nearly complete pottery jar rested on the topmost one. A segment of a second curvilinear wall, close to and possibly touching upon the circular structure, represented, in all likelihood, the northwest corner of the foresaid enclosure, although at this point it is difficult to prove a direct stratigraphic link between the various segments. If, however, this assumption is correct, the enclosure would have occupied an area of c. 18 × 22 m. Sparse, earthen floor remains, as well as remains of pebble stone beddings of possibly superimposed earthen floors, were revealed within the enclosure, with some restorable pottery vessels, mainly storage jars, in situ. The pottery dated these floor levels, and by extrapolation the enclosure and the two circular structures within it, to an early stage in the Southern EB I.


Stratum 1
This topmost stratum at the site was, most seriously affected by recent natural and human agents. The northern part of the excavation area gradually merged into a path leading down-hill that had been used over the years by local people and animals. Moreover, Hill ‘B’ had been used for many years as an IDF military training ground and shooting range. Several shallow pits and a few, disjointed but parallel segments of stone-wall foundations, were uncovered only tens of centimeters below surface. They directly overlaid a Stratum 3 wall segment while foundations of another wall that seem to have cut it overlaid the Stratum 4 circular wall. A stone pavement that sealed part of the Stratum 2 circular wall was attributed to this stratum. Pottery finds from fills within and outside the various pits indicated a date within the EB IA.

 

Hill ‘C’
Due to time constraints, Hill ‘C was only hastily examined for the presence of cupmarks and rock-cut installations. Thirty-five features, hewn in bedrock outcrops and spread over the west slope of Hill ‘C’ were identified and recorded.