Area F1
A simple winepress and adjacent installations were hewn in a bedrock surface (Figs. 3, 4). The winepress comprised a square treading floor (L107) and a rectangular collecting vat (L101) linked by a short channel (depth 0.1 m). An arched niche (L124; 0.35 × 0.50 m, height 0.6 m) hewn in the middle of the treading floor’s east wall was probably designed to secure the end of a wooden beam for secondary pressing of the grapes (Frankel 2009:4). A square rock-cutting (0.3 × 0.3 m, depth 0.2 m) discovered in the upper part of the treading floor’s west wall, opposite Niche 124, may have served to allow the lowering of the end of the wooden beam to increase the pressure while extracting the grape juice. The walls of the collecting vat retained traces of light-colored, hard plaster prepared from lime, small pieces of gravel and crushed pottery. An elliptical sump (L106) was hewn in the collecting vat’s southwest corner. A shallow cupmark (diam. 0.2 m, depth 3 cm) connected to the pit via a channel (depth 3 cm) was hewn c. 0.2 m away from the collecting vat’s northwest corner. Based on the depth of the cupmark, it was probably not intended to hold a jar or any other vessel to collect the grape juice, but to grind additives for clarifying the must or improving the taste of the wine (Frankel and Ayalon 1988:29). About half a meter west of the treading floor was a rock-hewn circular basin (L105; diam. 0.45 m, depth 0.4 m) with an elliptical cupmark beside it (0.15 × 0.30 m, depth 0.15 m). To the south of the collecting vat was a hewn circular basin (L100; diam. 0.5 m, depth 0.4 m) with a narrow groove cut around the opening, probably to hold a cover. A circular surface hewn around the north side of the basin contained a small cupmark (diam. 0.15 m, depth 0.1 m). The installation was probably a bodeda for the extraction of olive oil.
No finds were discovered in the soil deposited above the winepress and the installations, apart from a stone tool fragment (L107, B1015; Fig. 5), which was probably used as a grinding or whetting stone; the stone is a dark, metamorphic rock, probably belonging to the pyroxene group of minerals that probably originates from the vicinity of Antioch. The tool was fashioned on a partially worked pebble; its working edges are broad and flat and had obviously been smoothed from prolonged use, whereas its narrow sides were not used and were therefore left rough. The stone is very hard and may have been used to sharpen metal implements.
 
Area F2
An elliptical stone-clearance heap (5 × 8 m, max. height 0.6 m; Figs. 6, 7) was delimited by a wall (W101) built of a single row of large fieldstones set directly on the bedrock. Two perpendicular probes were excavated in the heap. The inner face of W101 was exposed down to bedrock. The wall’s stones were laid on a slight eastward incline created by means of a soil fill and small fieldstones that were inserted beneath the west side of the lower course. A section of a wall (W108) built of two rows of small and medium-sized fieldstones and preserved to the height of two courses (0.4 m) was unearthed to the north of the stone heap. The construction of W108 differs from that of W101. The south end of W108 was built against W101, suggesting that W108 probably postdates W101, at least technically. The stone heap yielded a folded rim of a krater made of a pinkish fabric (L102, B1002/1; Fig. 8), characteristic of the Judean region during the eighth–seventh centuries BCE (Freud 2007: Fig. 3.30:3).
 
Area F3
The excavation unearthed the remains of a rectangular building and two rock-hewn columbaria, one in the west and one in the east (Figs. 9, 10). The building is connected to the eastern columbarium via a rock-hewn shaft and passage. The eastern columbarium was quarried within an underground cavity that, which may have previously served for storage for the building. The two columbaria are linked together by a rock-hewn underground passage.
 
The Building (exterior dimensions 3 × 6 m; Figs. 11–13). Most of the building’s walls were founded on bedrock and built of large and medium-sized fieldstones (average size 0.4 × 0.5 × 0.7 m) bonded together with earth. The south wall was laid on a fill of small fieldstones and soil (0.1 m thick). The north, east and west walls were preserved to a height of 1–3 courses, whereas the south wall was preserved to the height of a single course — the foundation course. The building’s floor utilizes the bedrock, which was leveled by hewing; cracks in the rock were filled in with tamped soil to even the surface. An accumulation of grayish soil (L110) discovered above the floor contained finds, which date the building’s last phase of use to the late Iron Age II. Collapsed stones (L108) discovered above Soil Deposit 110 had fallen from the walls of the building and sealed the underlying accumulation. Alluvial soil (L103) was deposited on top of the architectural remains. A large stone (L116) set in the foundation in the northeast corner of the building’s walls contained a rock-cutting; it may have been the threshold stone of a narrow entrance opening (width c. 0.35 m). A section of another wall (W107) built of a single row of fieldstones laid directly on top of the bedrock was uncovered to the west of the building’s west wall (W103). As the wall was flimsily built, it probably carried no more than one other course and served as a low partition wall. The southeast end of the wall was only 0.1 m away from W103, and therefore W103 may have cut across W107. In the north of the building, two flat slabs of hard limestone (0.15 × 0.40 × 0.90 m) placed on the bedrock floor were unearthed beneath Soil Deposit 110. The two slabs blocked a natural crack in the bedrock (0.7 × 1.7 m, depth 1.2 m), which had been widened by quarrying its lower section to form a shaft leading to a rock-hewn passage; the passage led northward to the eastern columbarium. The crack and the shaft contained a soil deposit (L112).
 
Rock-Hewn Shaft and Passage. Two steps facilitating access were hewn in the upper part of the shaft (0.6 × 1.5 m, depth 0.6 m; Figs. 14, 15) leading from the building to the passage. The south section of the passage (length c. 3 m, depth c. 2 m) had five steps of various dimensions hewn in an angle of about 150 degrees. The middle step (tread 0.5 m, rise 0.3 m) stands out prominently in the passage, making the passage smaller, possibly to allow the blocking of the passage with a slab. The north section of the passage is low and horizontal (length 2.6 m, height at least 0.9 m) and can be negotiated only by crouching or crawling through it. The passage is narrow at both ends (0.5–0.7 m) and slightly wider in the middle (width 1 m). Its walls bear diagonal chisel marks (Fig. 16). Based on the direction of the marks, slanting to the south at their upper end, the passage was probably quarried from south to north. This assumption is substantiated by the fact that the natural crack in the floor of the building was widened into a shaft in order to connect the building with the underground cavity to its north.
 
Eastern Columbarium. The north end of the rock-hewn passage terminated in a rectangular opening (width 0.5–0.7 m, height 0.8 m) that led to the eastern columbarium (3 × 6 m, height 3.7–4.0 m; Fig. 17). The columbarium was quarried in an underground cavity, which was previously used by the occupants of the building, probably for storage. The columbarium’s walls were straight and bore visible diagonal chisel marks made by both the wide edge (2 cm) and the narrow edge (3–4 mm) of a rock-cutting chisel. As the columbarium’s ceiling was convex and contained no visible quarrying marks, it was probably part of the originally natural cavity. A rectangular hewn shaft (B; 1.0 × 2.2 m, depth 1.0–1.3 m; Fig. 18) leading down from the surface was necessary for removing the quarrying debris, providing access for the doves, and allowing light and air to penetrate the columbarium. A passage (L136; 0.7–1.7 × 1.5 m, length 1.3 m; Fig. 19) cut in its west wall connected it with the western columbarium. A hewn bedrock shelf (width 0.4 m) ran along the walls of the columbarium, 0.8 m above the floor. Above this shelf, small niches in various dimensions and shapes (depth 0.17–0.20 m) were cut into the four walls. Most of the niches were arranged in regular horizontal rows, but some deviated from these rows. Most of the niche openings were rounded (diam. 0.10–0.17 m) or square (0.10 × 0.11 m). Some were damaged due to the erosion of the rock, and part of the rock cuttings seem to have never been completed. Forty-five niches arranged in five rows were preserved in the south wall. Above the upper row of niches, signs of two more niches were identified; these probably indicate the presence of a sixth row, which did not survive. The niches in the south wall were also neatly arranged around the opening leading from the rock-hewn passage. There were no signs that any of the niches had been damaged by the formation of this opening, and no cut niches were visible inside the passage itself. Forty niches arranged in seven rows were preserved in the east wall. Large parts of the bedrock in the center of the east wall collapsed, making it impossible to estimate the number of niches that were cut in the wall. Judging by the number of niches preserved in the west wall, each row probably had 14–15 niches, making a total of 98 or 105 niches. Above the upper row of niches were eroded signs of hewing, possibly representing the remains of two more rows of niches in this wall. Fifty-two niches arranged in five rows were preserved in the west wall; seven of these niches were hewn in a sixth row, in the upper part of the wall. Some of the niches were destroyed or damaged as a result of breaching the opening leading to the western columbarium in this wall. The west wall probably originally contained 100–105 niches. Thirty niches arranged in five rows were preserved in the north wall, some of which had not been completed. The limited number of niches cut into this wall and the incomplete state of some of them may indicate that this was the last of the installation’s walls to be worked. A total of 174 niches were preserved in the columbarium, but there were probably about 145 additional niches.
Four superimposed accumulation layers (L104A–D) discovered in the columbarium reached a height of 0.5 m below the ceiling. Layer 104D (c. 2 m thick; Fig. 20), the bottom layer, was an accumulation of terra-rossa soil, building stones of various sizes and small fieldstones on top of the level bedrock floor (L140). The earth and stones were probably washed into the columbarium. Layer 104C (0.35 m thick) is a deposit of eroded rock mixed with a large quantity of charcoal and ash. Once this accumulation had been excavated, traces of soot were visible on the east, west and south walls of the columbarium. Layer 104B (0.3 m thick) is composed of naturally eroded rock from the ceiling and walls of the cave. Layer 104A (0.1 m thick) is a thin deposit of alluvium mixed with small fieldstones.
 
Rock-Hewn Underground Passage. The floor of the passage (L136; width 0.9 m) that was quarried out between the two columbaria was 1.2–1.8 m above the level of both their floors. Four niches were hewn in the passage’s south wall, indicating that the eastern columbarium continued to be used even after the passage had been breached into it.
 
Western Columbarium. This was a square columbarium (3.5 × 4.2 m, height 2.6–2.7 m; Fig. 21) with a slightly arched ceiling and a hewn shaft (A; 1.0 × 1.2 m, depth 1 m). The columbarium was excavated via Shaft A, which was found devoid of accumulations. A perimeter bedrock shelf hewn around the walls resembled in its dimensions the shelf in the eastern columbarium. Niches hewn in the columbarium’s four walls were similar to those in the eastern columbarium. Four to seven rows of niches were cut into each wall, and each row contained 12–14 niches, some deviating from the rows. Forty niches arranged in seven rows were preserved in the south wall. This wall was apparently only partly utilized, perhaps due to a natural crack that made it difficult to cut additional niches. The east wall was breached by the opening leading to the underground passage connecting the two columbaria. Fifteen niches arranged in five rows were preserved in the north part of this wall. The absence of niche-chiseling marks around the opening in the east wall indicates that this part was probably set aside from the start as the location for the rock-cut passage. Some of the niches in the northeast corner of the columbarium were hewn in the natural bedrock wall, possibly indicating that here too, as in the eastern columbarium, a natural underground cavity was enlarged. Fifty-two niches arranged in five rows were preserved in the west wall. Above the top row, elongated depressions are visible in the rock (see Fig. 21). It is difficult to determine whether they are natural or man-made, but they may have been formed by a collapse in the wall. Approximately 70 niches arranged in seven rows were preserved in the north wall, which slants slightly inward. A total of about 177 niches was preserved in the western columbarium.
An accumulation of terra-rossa soil (L131; 1.6–1.7 m thick) mixed with fieldstones of various sizes was unearthed on the columbarium’s bedrock floor. Above Accumulation 131 was a layer of small fieldstones (L127; c. 1 m thick) that had been discarded into the installation during stone clearance; the stones reached a height of about one meter below Shaft A.
 
Finds from the Building. The excavation of the building yielded pottery finds from the late Iron Age II, including fragments of bowls (Fig. 22:1–8), a krater (Fig. 22:9), cooking pots (Fig. 22:10, 11) and a holemouth jar (Fig. 22:12). The bowls are of various types, and most are made of a reddish or light brown fabric with small white inclusions. Some have traces of wheel-burnishing on the interior, sometimes also on the rim, and an inner red slip. They are common in contemporary assemblages in the Judean Shephelah. The cooking pots have thin walls, and most of them are made of well-fired reddish fabric that has a beige core and small white inclusions. An intact pot of this type discovered on the floor of the eastern columbarium (see below, Fig. 24:4) may have come from the building or from the earliest phase of the underground cavity’s use. A cruciform mark was incised on the handle of one of the cooking pots (Fig. 22:11). The rim of the holemouth jar (Fig. 22:12) was discovered in the soil deposited above the building and dates from the eighth to the early sixth century BCE. Loci 110 and 122 (an accumulation sealed below the collapsed walls and above the building’s floor) as well as L138 (a probe in the foundation of the south wall) were sealed loci, and therefore the pottery finds from them date the building (see pottery tables describing loci from which potsherds were retrieved).
Similar late Iron Age II assemblages have been discovered throughout Judea, for example in Stratum II at Tel Lakhish, dated to the seventh–sixth centuries BCE (up to 586 BCE; Zimhoni 2004:1871–1888, Figs. 25.56–26.44), in Stratum I at Tel Bet Shemesh (Bunimovitz and Lederman 2016a:132–149, Figs. 5.72–5.79), in Stratum V at Ramat Rahel (Freud 2016:254–265, Figs. 16.1–16.3) and in Strata 11–10 at the City of David in Jerusalem (Gitin 2015).
The building also yielded two clay loom weights, a zoomorphic clay figurine, two basalt grinding stones and a fragment of a decorated stone artifact. The two clay loom weights (Fig. 23:1, 2) were retrieved from the building’s bedrock floor and are characteristic of the Iron Age. The weights are round and were obviously shaped with a knife while the clay was leather-hard. Grooves were made with a thin, sharp tool at both ends of the hole perforated through the center of the weights. Similar weights were found at Tel Lakhish (most of them attributed to Stratum III; see reference in table) and at Tel Bet Shemesh (most discovered in Stratum II, dated to the eighth century BCE; see reference in table). The fragment of the clay figurine was discovered in the soil accumulation in Shaft 112 in the floor of the building. The fragment is conically shaped and probably belongs to the foot of a zoomorphic figurine (a horse? Fig. 23:3). Similar fragments have been found at Tel Bet Shemesh, where they were dated to the Iron Age IIB–C (see reference in table). The two basalt grinding stones were found on the floor of the building: one in the center of the building, near the loom weights (L110; Fig. 23:4); and the other on the quarried stone in the building’s foundation, which may have served as the entrance threshold (L116; Fig. 23:5). The two grinding stones are flat and broken and their working faces were smoothed by prolonged use. Many similar grinding stones have been found in Iron Age IIB–C strata at Tel Bet Shemesh (see reference in table). The fragment of the decorated stone artifact (c. 4 × 4 cm, 1 cm thick; Fig. 23:6) was discovered in the soil accumulation in Shaft 112. The stone is ornamented on one side with incised circles and its narrower side is smoothed. The fragment seems to be made of marble, possibly from Asia Minor or Crete. The fragment may belong to a grinding tablet or bowl.
 
Finds from the Columbaria. On the floor of the eastern columbarium (L140), beneath a thick accumulation of terra-rossa soil and stones (L104D), were three complete pottery vessels alongside each other—a bowl, a cooking pot and a juglet—which were broken into large fragments but could be fully mended (Fig. 24:1, 4, 5), as well as several other fragments of pottery vessels dating from the late Iron Age IIB. A few Hellenistic and Roman potsherds were also discovered on the floor of the eastern columbarium. The finds from the late Iron Age II include a complete bowl (Fig. 24:1) similar to several recovered from Stratum II at Tel Lakhish; a fragment of another bowl (Fig. 24:2:) similar to a red-slipped one recovered from the building (see Fig. 22:3); the rim of a krater (Fig. 24:3) resembling one found in the stone-clearance heap from Area F2; an intact cooking pot (Fig. 24:4) similar to those found in the building (see Fig. 22:10, 11); and an intact juglet (Fig. 24:5). The pottery finds from the Hellenistic period include the rim of a cooking-pot (Fig. 24:6), characteristic of Judean sites in the latter half of the second century BCE. Pots of this type were also found at Tel Gezer (see reference in table). The Roman period was represented by a cooking-pot rim (Fig. 24:7) and the rim of a jar (Fig. 24:8) dating from the late first to the second centuries CE.
On the bedrock floor in the center of the underground passage cut between the two columbaria was an intact Fine Byzantine Ware cup (L136; Fig. 24:9). The cup has straight walls and a flat base and dates from the latter half of the sixth to the first half of the eighth centuries CE.
Soil Accumulation 131 above the floor of the western columbarium yielded a cooking-pot rim (Fig. 25:5) characteristic of the second half of the seventh and the early sixth centuries BCE; a holemouth jar (Fig. 25:6), for which parallels have been found in Stratum III at Lakhish, dated to the eighth–seventh centuries CE; and a cooking-pot rim from the Hellenistic period (Fig. 25:7). Among the small fieldstones (L127) thrown into the western columbarium, an Early Bronze Age thumb-impressed shelf handle was recovered (Fig. 25:1), as were the rims of bowls from the Iron Age IIB (25:2–4) similar to those from the building (see Fig. 22:1–3).
 
Building and Quarrying Phases. In the absence of stratified accumulation layers in the remains of Area F3, the relative dates of the remains are based on the identified phases and the relationship between them. The building was probably built first and joined to a natural underground cavity, which was enlarged by quarrying and probably served for storage. The building and this cavity were contemporaneous, since they are clearly linked. They were connected by a rock-hewn shaft and passage, and the dimensions of the building (3 × 6 m) are identical to those of the cavity. After the building was abandoned, the cavity was converted into a columbarium. The identical elevation of the passage floor and the perimeter ledge in the eastern columbarium suggest that this was the natural level of the bedrock floor, whereas the floor of the columbarium itself was deepened by quarrying. Shaft B, leading from the surface to the ceiling of the eastern columbarium, was also hewn in this phase.
As the two columbaria have similar features, they were probably quarried at the same time or during a same period. Nevertheless, the eastern columbarium appears to have been hewn earlier than the western columbarium. Both the damage suffered by the niches in the west wall of the eastern columbarium, when the entrance to the underground passage was breached, and the absence of niches in the east wall of the western columbarium, to which this passage leads, indicate that the eastern columbarium was hewn first and that the passage was designed to connect the two columbaria at some later stage. It is also possible that while hewing the east wall of the western columbarium, sections of the bedrock wall collapsed, halting any further quarrying of niches in the wall and linking the two columbaria. Shafts A and B, which provide separate access to each columbarium, may indicate that they were not in used simultaneously.
 
Area F4
A rectangular stone-clearance heap (max. height 1.5 m; Fig. 26), consisting mainly of small fieldstones, was unearthed. A probe on the west side of the heap uncovered the west face of a wall (W20) that delimited the heap on the west. The wall was built of medium-sized fieldstones and was preserved to a height of 1–3 courses (0.3–0.8 m). A probe on the east side of the heap revealed part of a natural hollow in the bedrock, where the stones were piled up. No archaeological finds were recovered.
 
Area F5
A circular stone-clearance heap (max. height 1 m; Fig. 27) was enclosed on the south by a wall (W21). The wall, founded directly on the bedrock, was built of medium-sized fieldstones preserved to a height of five courses (1.8 m). The four lower courses were built of flat fieldstones, whereas the upper course was built of large stones. No archaeological finds were recovered.
 
Area F6
A winepress (Figs. 28, 29) hewn in a rock outcrop comprised a treading floor (L221; 2.5 × 3.1 m) and a collecting vat (L206; 1.15 × 1.20 m, depth 1.6 m) joined together by a short channel (length 0.25 m, width 8 cm). A step (height 1.1 m) in the southeast corner of the collecting vat facilitated access. A circular sump (L206; diam. 0.4 m, depth 0.1 m) was hewn in the southwest corner of the vat. A layer of gray plaster was preserved on the north wall of the collecting vat up to a height of 0.9 m from its floor. A hewn groove (depth 8 cm) in the bedrock around the south and east sides of the vat’s opening may have served to hold in place a wooden cover over the vat. No archaeological finds were recovered.
 
Area F7
Rock-hewn installations were discovered (Fig. 30): two rock surfaces (L201, L206); a pit (L200); a cave (L256) accessed by a rectangular shaft (L205) and a staircase (L208); and a rectangular shaft (L209) leading into a cistern. Rock Surfaces 201 and 206 were divided by a rock-hewn wall. Pit 200 is adjacent to Rock Surface 201, and the north part of the surface had collapsed into the pit as a result of rock erosion. Surface 201 and Pit 200 may have been part of a winepress. In view of its proximity, Surface 206 may also be part of the winepress, although it is not connected to it by a channel, and its floor is a few centimeters lower than that of Surface 201. Cave 256 is a natural cave which was enlarged by hewing (3.5–7.0 × 4.5 m, height 1.2–2.0 m). Shaft 205 (1.1 × 1.3 m, depth 0.5–0.8 m) was cut through the roof of the cave and led to it from the surface. Staircase 208 was cut southeast of the cave and contains five steps that lead into it (Fig. 31); the staircase was probably part of a closed passageway. A groove hewn around the entrance of Shaft 209 (0.65 × 0.85 m, depth 0.6 m) was probably intended to hold a wooden board to close it. Shaft 209 leads to a round water cistern (diam. 2.6 m, depth 3.3 m) with straight walls and a convex ceiling. Traces of light gray plaster (4 cm thick) were preserved on the cistern’s walls. Rounded footholds were cut in the cistern’s east wall to enable access. The cistern was filled with a deposit of soil and modern refuse (c. 2 m thick). An accumulation of terra-rossa soil mixed with small fieldstones and several building stones (0.9 m thick) was unearthed on the cistern’s floor and beneath this deposit; no diagnostic finds were retrieved. It is impossible to date the installations. They did not cancel each other, as they may have been part of a single complex that included a winepress, a storage cave and a water cistern. The installations may be linked to a Hellenistic-period farmstead discovered nearby (Yogev 1984).
 
Area F8
The area contained a rock-hewn installation consisting of a treading floor (L210; 1.1 × 1.5 m; Fig. 32) that sloped slightly toward a cupmark resembling a truncated cone (L211; diam. 0.75 m, depth 0.65 m). Another rock-hewn cupmark (L212; diam. 0.3 m, depth 0.2 m), uncovered c. 0.8 m to the north, was square in section. The installation was probably a bodeda for extracting olive oil.
 
Summary
The excavation unearthed the remains of a building from the late Iron Age II, two columbaria, rock-hewn winepresses, a rock-hewn bodeda, stone-clearance heaps, a cistern and a cave. No diagnostic finds were retrieved from the installations, the stone-clearance heaps or the cave. These remains may be related to the previously discovered farmstead. This farmstead was dated to the Hellenistic period, although the pottery recorded in the excavation file suggests that it may have been established during the Iron Age II. It is also possible that the uncovered remains are related to the nearby site of Kh. Umm Tunis, which has been severely damaged by modern development work.
The finds discovered within the building in Area F3 indicate that it was abandoned at the end of the Iron Age. The building’s contents, which were sealed beneath the collapse of its walls, attest to daily activities, such as grinding grain, weaving and cooking. The absence of any additional architectural remains in the vicinity of the building may suggest that this was an isolated building that was in use during seasonal agricultural work, for example during the grape or cereal harvests. The winepress unearthed near the building in Area F1 reinforces this assumption. The underground cavity to which the building was connected was probably used as a storage room for agricultural produce, as underground temperatures are significantly lower than those on the surface. It was also possible to block the passage leading to the cavity to keep rodents out. Access to a cool storage area for agricultural produce and byproducts, such as grapes and milk, was extremely important, and this explains the great effort invested in quarrying out such a cavity. Although it is difficult to date the two columbaria, they may also be associated with the nearby Hellenistic farm. This, however, raises questions regarding the three late Iron Age vessels that lay—broken into large sherds that were completely mended—on the floor of the eastern columbarium: were they found in situ and could they date the columbaria?
It is now generally accepted by scholars that columbaria first appeared in the third century BCE (Tepper 1986; 2000). In 1986, Tepper estimated that there are about 1000 columbaria in Israel, approximately 900 of which are located in the Judean Shephelah (Tepper 1986:177). In a later study, Kloner counted about 600 columbaria at dozens of sites in the Judean Shephelah. About 85 installations were found at Maresha, and partial inspections and excavations conducted in roughly one-fifth of these yielded finds from the third–second centuries BCE (Kloner 2003:45). It is therefore probable that during the third–second centuries BCE, columbaria were hewn mainly in the Judean Shephelah and the southern coastal plain (Tepper 1986:180; Kloner 1996:29; Tal 2006:133). Most of the columbaria found elsewhere in the country have been dated principally to the late Second Temple period (first century BCE–first century CE; Kloner 1996:29). Columbaria have recently been excavated in the Jerusalem region (Regev and Shapira 2013), near ‘Amman (Kakish 2012) and at ‘Atlit (Sari 2010), where their excavators suggested that they may initially have been used in the Iron Age or the Persian/Hellenistic period. Several excavations in the vicinity of Jerusalem have uncovered columbaria containing numerous Iron Age finds (Zilberbod 2015; Adawi 2016; Shor 2016). Due to the difficulty of dating installations based on their pottery contents, the dating of columbaria remains problematic at this stage in the research.
I am of the opinion that the three complete vessels found on the eastern columbarium floor were originally in the building or the underground cavity during its early use, before it was converted into a columbarium. The vessels’ state of preservation makes it difficult to assume that they were washed from the building through the shaft or passage to the underground cavity, in which case they would have broken into small fragments. However, these vessels may have reached the shaft or the passage while the building was still in use and after the cavity was converted into a columbarium and Shaft B was quarried and—when it was possible to access the passage and the shaft from below—they were retrieved from there and brought into the columbarium.The intact cup found in the underground passage between the two columbaria and dating from the second half of the sixth to the first half of the eighth centuries CE attests to some form of activity in the underground complex during this period.