Two quarries were fully excavated (Quarries 13 and 15) and two were partially excavated (Quarries 10 and 86). They comprised quarrying steps (height 0.25–0.35 m), several stones whose quarrying had not been completed (e.g., Quarry 13: stone dimensions 0.30 × 0.30 × 0.45 m; Fig. 2), quarrying marks corresponding to two different sizes of stones (0.4 × 0.7–0.8 m, 0.60–0.75 × 1.00–1.20 m) and severance channels (max. width 9–11 cm; Fig. 3) that became narrower as they deepened. The severance channels attest to a uniform working method. The quarries are not deep and seem to have served for extracting blocks from the layer of nari—a semi-hard calcareous rock crust—that is particularly suited for building. The nari layer covers a layer of softer chalk, and rather than continuing to hew down into the soft chalk once the nari layer was exhausted, the quarrying work apparently moved to a new location. This method explains the numerous quarrying sites visible on the hill and on nearby hills. Once the quarries were no longer used, soil containing a few potsherds accumulated in them. Most of the pottery found in the quarries and scattered over the hill dates from the Roman and Byzantine periods (first century BCE – seventh century CE). However, since the deposition of these sherds occurred only once the quarries were no longer in use, they do not necessarily date the quarrying activity.
Nevertheless, the pottery does point to the site’s main periods of activity. Stepped stone quarries attributed to the Roman and Byzantine periods are known throughout the empire (Gorzalczany 2007:42). It is also highly probable that the ashlars’ main consumers were the inhabitants of Yafi‘a, a village which was occupied during these periods.
Quarry 10 (L1010; max. depth 0.8 m; Fig. 4) has two quarrying steps.
(L1003; max. depth 1 m; Figs. 2, 5) is a courtyard quarry. It consists of four quarrying steps and is suitable for one or two laborers (Gorzalczany 2007
Quarry 15 (L1004; Figs. 6, 7) is a courtyard quarry with five quarrying steps. The quarry contained a deposit of dark brown soil (L1007) mixed with stones, some of which were roughly hewn. A concentration of stone flakes and a few large stones were discovered in the north part of the quarry (L1008). One of the stones was dressed (0.20 × 0.40 × 0.45 m) and had probably been prepared inside the quarry but was left behind for some reason. The stone flakes, which are probably dressing debris, indicate that the stones were finished on site.
Quarry 86 (L1001; Fig. 8) contains several shallow quarrying areas (max. depth 0.55 m) interspersed with unworked bedrock. These quarrying areas consisted each of a single step, except for one area, which had two steps.
Limekiln (Site 69; L2001; Fig. 9)
The kiln (diam. 3 m) was hewn in the bedrock near the entrance to a cave. Only the lower part of the limekiln’s firebox was preserved, and it was partly excavated. The kiln was partially lined with stones. On a step encircling the bottom part of the kiln (diam. c. 2 m) were traces of soot from the firing of the lime. The kiln contained a deposit of soil mixed with stones. A layer of ash (0.8 m thick) mixed with partially burnt stones intended for lime production was unearthed at the bottom of the kiln. No potsherds were recovered. Since limekilns were built in a similar manner over three millennia, up to the late Ottoman period (early twentieth century CE; Sasson 2002:38), it is impossible to date the kiln.
Flint Quarrying and Initial Knapping Site (Site 74)
A cluster of knapped flint items and natural flint mixed with small calcareous stones was visible on the surface. Natural flint nodules of various sizes are scattered nearby, and the scars identified in the bedrock indicate that flint nodules were extracted there. A probe (2.5 × 2.5 m, depth 1.3 m), excavated down to bedrock, uncovered three layers: a top layer (L1002), which contained a high density of flint items, limestones and a small amount of sediment; a middle layer (L1006), which resembled the upper one, but contained only half the amount of flint items; and a bottom layer (L1011), which contained few flint items and a relatively large amount of sediment compared with the upper layers. Several Roman- and Byzantine-period potsherds were found in all the layers, but these had apparently penetrated into them and were unrelated to the knapping in this area.
The flint items were made of good-quality, homogeneous dark brown flint, and although most of the items have one or even two layers of patina (Figs. 10, 11) they are evidently not worn. These observations indicate that the flint was knapped in the vicinity, and that the knapped items remained exposed to weathering. Over half the assemblage in each layer consisted of unidentified knapped items (debris; Table 1) and natural flint chunks. The industrial debitage (Table 2) includes flakes, primary items, blades, knives with a natural back, core debitage and cores. Over half of the industrial debitage in each layer consisted of flakes. These are followed by primary items—flakes and blades with a calcareous cortex covering at least half of the dorsal side. There was a low percentage of blades, which were more common in Layer 1002. Some of these had bipolar flaking, a technology characteristic of the Pre-Pottery Neolithic B and the Early Pottery Neolithic periods. Most of the cores (Table 3) were used for flake production, and they had one or two striking platforms. Some cores had two opposing striking platforms (bipolar; Figs. 12; 13:2, 3), corresponding to the presence of blades exhibiting this technology. Levallois and Levallois-like cores were also recovered (Fig. 14), along with cores bearing a scar pattern resembling that of discoidal cores (Fig. 15), a bladelet core, amorphous cores, broken cores, a core on a flake and cores in initial stages of preparation (preform cores; Figs. 10, 16). Core debitage items were common, but only a few of them can be defined as crested blades or core tablets, which are a bi-product of blade production. This is consistent with the low incidence of blade cores. In Layer 1002, core debitage was commoner than cores, whereas in Layer 1006 the ratio was reversed. The frequency of tools (Table 4) was very low; in Layers 1002 and 1006 they comprised approximately one percent of all the items and consisted mainly of retouched flakes. They also included two retouched blades, one which bore bipolar scars (Fig. 13:1), and bifacial tools, some of which were in initial stages of preparation (roughouts; Figs. 11, 17) and were probably destined to be adzes. One of the bifacial tools was prepared on a core with two opposing striking platforms, such as the one in Fig. 12:3.
The flint nodules found in the vicinity of the site, along with the high incidence of industrial waste, particularly primary items and cores, the paucity of tools and the tools found in an initial stage of preparation—all indicate that this was a site where flint was quarried and underwent initial knapping. The Levallois and Levallois-like cores show that the site began to be used in the Middle Palaeolithic period. The production of blades using two opposing striking platforms (bipolar), which is evidently the predominant technology at the site, shows that it was used mainly during the PPNB and/or the Early Pottery Neolithic period. Adzes occur in the Neolithic period and become even more common in the Chalcolithic period, suggesting that the site may have also been in use during the Chalcolithic period. An opportunity for further research lies in examining the connection between the quarrying and knapping site at Yafi’a and nearby Neolithic sites, such as Kefar Ha-Horesh, ‘En Zippori and Yiftah’el.