During February–March 2005 a salvage excavation was conducted on the southeastern fringes of Horbat ‘Illit (Permit No. A-4385*; map ref. NIG 2030/6187; OIG 1530/1187), following an archaeological survey in the region carried out by A. Nagorski and H. Stark and prior to the construction of the separation fence (Fig. 1). The excavation, on behalf of the Antiquities Authority and funded by the Ministry of Defense, was directed by Z. Greenhut, with the assistance of V. Essman, V. Pirsky and T. Kornfeld (surveying), E. Belashov (drafting), A. Pikovski (pottery drawing), C. Amit (studio photography), D.T. Ariel (numismatics), A. Nagorski and H. Stark (logistic and professional support) and K. Sozab (fieldwork coordination).
The site is located in the upper Shephelah, c. 490 m above sea level (Fig. 2), c. 3 km northeast of the Zurif village and c. 10 km south of Bet Shemesh. The ruin affords a view over the Hebron Hills to the east (Fig. 3), the hills of the upper Shephelah to the north and south and the lower Shephelah and coastal plain to the west. Nahal Gadur, which begins at Khirbat Jadur in Gush ‘Ezyon, flows to the north and east below the ruin.
The site had been surveyed in the past by Y. Dagan who described a large ashlar-built structure, other buildings, pits and industrial installations, as well as potsherds from the Hellenistic, Roman and Byzantine periods. N. Tal discovered a hiding complex at the site in 1996. A. Ganor and A. Klein discerned another hiding complex at the site in 2000, or the continuation of the one found by Tal, which included a long tunnel, caverns and a water cistern. B. Zissu noted that the potsherds recovered from the complex dated to the Early Roman period and the time of the Bar-Kokhba Revolt. The current excavation did not locate the hiding complex (or complexes); yet, the finds indicate that a settlement did exist at the site from the latter part of the Second Temple period until the Bar-Kokhba Revolt.
The excavation area on the southeastern slope of Horbat ‘Illit (Fig. 4) included two farming terraces that were separated by a wall (W1) built of a single row of large fieldstones (Figs. 5, 6). The construction method of the wall closely resembled the housing remains of the Arab village at the top of the hill and the wall should, therefore, be dated to this period.
Four underground cavities, identified in the survey (Fig. 1:16, 17, 19, 20; Fig. 4: 108, 103, 102, 106), were excavated, as well as other spots discerned during the excavation (Fig. 4:101, 101A, 104, 105). Fieldwork began in Points 101, 101A, 102–106, which mostly turned out to be burial caves that formed part of the settlement’s cemetery. A new alternate route was selected for the fence, designed to run along a topographically lower terrace than the original route, through Point 16 of the survey (Fig. 1)––Point 108 of the excavation (Fig. 4) and through the water cistern (Point 107), which is used by the local residents to date. Hence, the excavation began in the lower area where the new route is scheduled to pass.
Cave 101 (Fig. 7)
This kokh-type burial cave is characteristic of the latter part of the Second Temple period (first century BCE–first century CE). The cave was documented and not excavated. It was hewn in qirton bedrock of poor quality, especially in the ceiling and the eastern wall. The cave included a rectangular main chamber (3.5 × 4.0 m), which was breached in its northeastern corner that joined the cave with Cave 101A. The entry to the cave was probably located here. Six kokhim (I–VI; 0.4–0.5 × 1.9–2.0 m), two in the southern wall, three in the western wall and one in the northern wall, were discovered. The opening to Kokh VI had a shaped frame (Figs. 7: Section 2-2; 8) and its ceiling was broken prior to the excavation (Fig. 9). The absence of ossuaries in the cave could be due to: (1) the cave was plundered in the past and the ossuaries were removed, (2) People who could not afford ossuaries were interred in the cave and (3) The cave slightly predated the initial use of ossuaries.
It is noteworthy that following the completion of the documentation, the breach connecting Caves 101 and 101A was sealed.
A collapsed cavern to the north of Cave 101, probably a cave, was discovered (Fig. 10). Beneath the collapsed qirton ceiling of the cave was an accumulation of brown soil that contained human bones, indicating this cavern functioned as a burial cave. The original plan of the cave is unclear and its excavation was suspended when the brown accumulation was uncovered.
Cave 102 (No. 19 in survey)
A circular breach at the top of the cave postdated the period of the cave’s original use. The cave was only partially cleared because as soon as human bones appeared, the excavation was suspended. Thus, the plan of the cave is incomplete and segmented.
A carefully qirton-bedrock hewn corner was preserved in the southeastern section of the main chamber, which was excavated to a depth of c. 0.3 m (Figs. 11, 12). An arcosolium (Fig. 11: I; length 1.25 m, depth 0.7 m, max. height 1 m) in the northern part of the burial chamber was coarsely finished, probably due to the inferior quality of bedrock. The arcosolium contained human bones that were left in situ.
The entrance to the main chamber, probably in its southern side, was found blocked with large stones. A bottle-shaped recess in the southwestern corner of the chamber had a wide bottom, tapering toward the narrow top (Fig. 11: II: diam. 0.7 m, depth 0.3 m); its nature remains unclear. Next to it was another niche (Fig. 11: III). The western side of the chamber sloped from top to bottom, toward the interior of the burial chamber.
The excavation was suspended following the exposure of human bones.
Ceramic Finds (Fig. 13)
The upper part of the accumulation that filled the cave contained animal bones and numerous fragments of pottery vessels (Fig. 13:3–13) from the Hasmonean period, as well as a few potsherds from Iron Age II.
Fragments of two bowls (Fig. 13:1, 2) attributed to Iron Age II, have a folded out rim. These fragments had probably penetrated the cave from the outside through the breach in the ceiling and did not belong to the cave’s original use. Their appearance attests to the existence of a settlement from this period at Horbat ‘Illit.
The overwhelming amount of the ceramic assemblage dated to the Hasmonean period and included jars with a flared rim (Fig. 13:3–5), characteristic of the second–first centuries BCE and jars with a thickened, folded out rim (Fig. 13:7–10) that are dated to the second century BCE. It seems that the jug (Fig. 13:11) whose rim resembles that of the jars should also be ascribed to this type of vessel. Another jar (Fig. 13:6) with a wide mouth, an everted rim and a curved shoulder is either a geniza or an archival jar.
Two lamps, a wheel-made pinched lamp (Fig. 13:12) characteristic of the Hasmonean period and an intact carinated lamp (Fig. 13:13) next to the cave’s wall, with a convex ring base and traces of soot, were discovered. The wick-hole of the latter is surrounded by a sunken band and on its upper rear part the beginning of a lug handle was preserved. A shallow semicircular depression on one of its sides was possibly intended for placing a finger when the lamp was held. The lamp is typical of the second century BCE.
It seems that most of the ceramic assemblage is dated to the second half of the second century BCE, which may possibly be the date of Cave 102. It is noteworthy that no evidence of kokhim was discerned in the cave.
Cave 103 (No. 17 in the survey; Fig. 4)
The front of the cave was excavated to bedrock and no finds were discovered.
Cave 104 (Fig. 14)
This kokhim cave was filled with alluvium to the top of the kokhim (Fig. 15). The cave was only measured and not excavated. It consisted of a square chamber (3.4 × 3.4 m) and an entrance (0.85 m wide) in the east that was mostly destroyed. Three kokhim (0.5 × 1.7–2.1 m) were hewn in each of its western and northern walls. In the northeastern corner of the chamber was a small recess (0.3 × 0.5 m) that may have served as a bone repository. The southern wall had no kokhim and in the eastern wall was a breach. The reasons for the absence of ossuaries in Cave 101 are probably valid with regard to this cave as well. Upon completing the documentation, the entrance was sealed and covered with stones and earth.
Point 105 (Fig. 4)
An opening (0.8 m wide) in the cliff of the upper bedrock terrace, c. 6 m south of Cave 104, led to a blocked cavern that was neither examined nor surveyed.
Cave 106 (No. 20 in the survey; Fig. 16)
The cave, only partly excavated, was at the top of the slope, south of the hilltop. It included a corridor (1.3 m wide), blocked in its northern part with dark brown alluvium, 3 m of which were excavated and proved to be devoid of finds. The corridor was created by narrowing the bedrock benches on either of its sides, which were 0.3 m higher than the floor (Fig. 17). The nature of the corridor is unclear; the sloping of its floor from north to south seems to indicate that the entrance to the underground chamber was from the north. To its south was an open space without a ceiling, possibly a collapsed cave (Fig. 18).
Water Cistern 107 (Fig. 19)
A bell-shaped water cistern located east of the hillside. The lower part of the cistern’s opening (diam. 0.7 m) was bedrock-hewn (depth 0.8 m) and the upper part consisted of three fieldstone-built courses (height 0.8 m). The water in the cistern, which is still used to date by the local residents, reached a level of 6.7 m below the top. A square iron cover (0.5 × 0.5 m) overlies the top of the opening and a concrete surface (2.4 × 3.5 m) around it has two troughs one at either end.
Water Cistern 108 (No. 16 in the survey)
A cave was identified at this spot in the survey; however, in the wake of a backhoe-dug probe it was ascertained to be a plastered water cistern (diam. 4.6 m, depth c. 6 m; Fig. 20), whose upper part was not preserved. At 3–6 m below the opening, which was visible on surface, a plastered water cistern whose ceiling and aperture were not preserved, was discerned (Fig. 21). Its sides were coated with a gray plaster layer, overlaid with a smooth plaster layer. On the lower part of the cistern’s walls a third gray plaster layer that contained numerous small inclusions was applied. Small depressions and cavities were made in the smooth plaster layer so that it adhered to the gray layer with inclusions, which was common to the Second Temple period and enables us to date the cistern to the period when the burial caves were used.
Two coins were discovered on surface with the aid of a metal detector. One coin was of Constantius II (351–361 CE) and the other, a Mamluk coin.
The ceramic artifacts recovered from Cave 102 indicate a settlement dating to Iron Age II existed at the site. The occurrence of an arcosolium shows that this cave predated caves of a similar type in Jerusalem. Arcosolia caves, common to the Shephelah, were initially described in research as a Phoenician or Egyptian type, dating to the Seleucid period. It is now known that Arcosolia caves dated to the third century CE. Arcosolia caves in Jerusalem were first dated to the Hasmonean period, e.g., the Eshkolot Cave, but it is currently suggested that their appearance in Jerusalem was a phenomenon of the first century CE and scholars assume that this burial type began to be used only in the last decades of the Second Temple period. It is further presumed that the origin and essence of arcosolia in Second Temple period Jerusalem can be traced to the internal development of burial forms, which provided answers to functional and halachic problems (see The Necropolis of Jerusalem in the Time of the Second Temple :39–40).
Cave 102 seems to contradict this assumption. Although the arcosolium in this cave is certain, it is irregular, has an inferior finish perhaps due to the poor quality of bedrock and it differs from the well-hewn arcosolia caves that are dated to the beginning of the second century CE and particularly to the Byzantine period. The ceramic finds from the cave are mostly dated to the second half of the second century BCE and the provenance of the intact clay lamp seems to be from the cave itself rather than penetrating from outside in a phase postdating the use of the cave. One can therefore contend that the finds at Horbat ‘Illit corroborate the claim that arcosolia caves existed in the Shephelah during the second century BCE.
Burial Caves 101, 104 point to a settlement from the first century BCE to the first century CE at the site. Furthermore, the evidence of a hiding refuge complex at the site alludes to the settlement’s continuation until the time of the Bar-Kokhba Revolt, similar to most of the sites in the region. In addition to using this area as a burial site, the water cistern from the Second Temple period evinces agricultural activities in the region.