The rock-hewn complex (Fig. 1) comprises subterranean cavities that were first used in the Hellenistic period as quarries and later as storerooms; one cavity may have been a water reservoir. Toward the Bar Kokhba Revolt they were connected by long and narrow tunnels, forming an underground hiding complex.
The complex is currently entered through a shaft (A; Fig. 2) that was breached by antiquity robbers in 2012. The shaft descends c. 2 m from the surface to a narrow tunnel with a curved ceiling (a-d; length 13 m, width c. 0.8 m, height c. 0.8 m). A rectangular room (B; 2.2 × 3.0 m) was hewn at the northern end of the tunnel, next to Shaft A. The ceiling of Room B has partially collapsed and most of the room was filled with stones from the surface. It seems that Room B was utilized as a storeroom, and that Shaft A served as its original entrance. Another room (C; c. 2 × 2 m), reached through a short tunnel that branches out of Tunnel a-d, is hewn c. 2.5 m east of Room B. Its floor is covered with alluvium that penetrated through Shaft A. In the middle of Tunnel a-d is a low, narrow, arched opening. At the southern end of Tunnel a-d is a T-intersection (D), from which two tunnels (d-e, d-f) continue to the east and west respectively. Tunnel d-e (length c. 11 m) leads to a bell-shaped elliptical cavity (E) hewn at its end; two steps hewn near the middle of the cavity descend to a lower level. The tunnel breaches the southern wall of Cavity E at an elevation of c. 5 m above the floor level (dimensions of the cavity at the level of the tunnel: c. 3.0 × 4.5 m). A rock-cut spiral staircase descends to the floor of Cavity E from an opening (width c. 0.8 m) hewn high in the cavity’s northern wall. A wide, elliptical shaft (depth c. 8 m) opens in the ceiling of the cavity; its upper part of reaches the surface. It seems that Cavity E was first used as a quarry and later adapted for either storage or as a cistern. Similar cavities for water storage that have rock-cut spiral steps, which are dated to the Hellenistic period, were discovered at nearby Maresha (Kloner 2003:35–39; Kloner and Zissu 2013). If Cavity E was indeed a cistern, it is likely that Tunnel d-e was used to secretly draw water when the hiding refuge was in use. Similar hiding tunnels that facilitated the clandestine drawing of water were discovered in underground complexes in the Judean Shephelah.
In Tunnel d-f (length c. 8 m) an opening that leads to a secret room (G; c. 2 × 2 m, height c. 1 m) was hewn c. 4 m from the tunnel intersection (D). On the doorjambs are two dressed recesses, one above the other, which were used for a door bolt of a locking mechanism. The floor is covered with soil, on which fragments of jars characteristic of the period between the two revolts against Romans were found.
At the western end of Tunnel d-f is a circular cavity (F) connected to a larger cavity (H) by means of a wide rock-cut step. Cavity F severed the upper part of the northern wall of Cavity H, and it therefore seems that the quarrying of Tunnel d-f and Cavity F postdated Cavity H. Cavity H is round (diam. c. 6 m, height 5–6 m) and widens at its bottom. At a low level in the southwest of Cavity H, antiquity robbers exposed the opening of a short tunnel (h-i; length c. 2.5 m) that leads to a large elliptical chamber (I; c. 5 × 9 m, elevation above the accumulated soil on the floor of the cavity c. 6 m). The opening of the tunnel from Cavity I is delineated by a shallow frame (Fig. 3). A large opening (width c. 1 m, height c. 2.5 m) that was intentionally blocked with big ashlars, including a stone weight of an olive press in secondary use, was documented in the western wall of Cavity H. The opening leads to a corridor (L) that was blocked with a layer of earth and large stones. Corridor L leads to a well-designed entrance that opens onto Cavity I. The opening, hewn in the eastern wall of Cavity I, is broad and high (width more than 1 m, height c. 2.2 m; Fig. 4). The opening’s doorjambs are fashioned to resemble engaged pillars with Doric capitals that have a diagonal triangular echinus and a short vertical abacus. A stepped, cornice-like lintel is carved in the bedrock above the doorjambs. The combination of engaged pillars and a stepped lintel (a type of entablature) gives the opening the appearance of a classical façade. Similar decorated openings of underground cavities were discovered at Maresha and date to the Hellenistic period. A Greek inscription alongside an engraving of a sailing vessel with a schematic sail (below) was discovered on the lower part of the lintel, facing Corridor L; the blockage of Corridor L with earth and stones made it difficult to document and photograph the opening and the inscription and engraving. Hewn steps that have a low railing descend from the ornamented opening toward the cavity’s floor. Next to one of the bottom steps is a hewn opening to Tunnel h-i (above), which leads to Cavity H. The location of the ornamented opening in Cavity I opposite the opening in Cavity H indicates that the two openings were planned together and that Cavities H and I were probably a pair of adjacent subterranean storerooms. Corridors similar to Corridor L are characteristic of the underground complexes at Maresha and its environs, and they were hewn from the surface level (Kloner and Zissu 2013; see a similar cavity complex that was discovered at Khirbat ‘Araq Khala north of Bet Guvrin; Zissu, Langford, Acker and Eshel 2013:112). It therefore seems that Corridor L was originally hewn from the surface, but only its lower opening was exposed here. Apparently, Corridor L was deliberately blocked, probably when the hiding complex was being quarried and Cavities H and I were connected by Tunnel h-i. Concentrations of pottery sherds and fragments of glass vessels, probably gathered by the antiquity robbers, were discovered on the floor of Cavity I. Noteworthy among the finds are fragments of candlestick-like glass bottles typical of the period between the two revolts against the Romans (Fig. 5); as well as numerous fragments of ceramic jars that have a ridge on their rim (Fig. 6) and fragments of cooking pots with a triangular rim and a groove on the upper part, both of the types characteristic to time period between the last third of the first century CE and the Bar Kokhba Revolt.
Hewn openings in the southern wall of Cavity I lead to two square rooms (J—2.0 × 2.2 m, height above soil accumulation 1.3 m; K—1.5 × 1.7 m, height above soil accumulation 1.3 m). The opening of Room J is arched; rock-cut triangular recesses (width 0.6 m, height 0.8 m) for placing lamps were hewn in the doorjambs. A low, narrow rock-hewn tunnel (length 1.6 m) leads to Room K. Recesses for oil lamps are hewn in the walls of the room. The dimensions of the rooms and their openings, the plan of the rooms and the tunnel leading to Room K—all suggest they were hewn as part of a hiding complex.
In the northern wall of Cavity I is a hewn opening of a small square room (M; 1.0 × 1.2 m, height 0.8 m), which is set at a higher level than the floor of Cavity I. Triangular recesses meant for lamps are hewn on both sides of the opening in the northern wall of Cavity I. An Aramaic inscription engraved above the western recess was discovered c. 2 m above the soil accumulation on the bottom of Cavity I (below). A narrow tunnel (m-n; length c. 8 m) leads north from Room M. At the northern end of the tunnel is a hewn step going up toward a shaft cut in the ceiling of the tunnel (N). Shaft N is blocked with soil and rocks, but it evidently leads to the surface and served as the opening to the complex. Two rectangular rooms (O, P) are hewn along Tunnel m-n. Pottery sherds of vessels characteristic of the period from the last third of the first century CE to the Bar Kokhba Revolt were discovered in a layer of soil in both rooms.
Recesses for lamps are hewn along the walls of all of the tunnels in the complex.
 
An Aramaic Inscription (Figs. 7, 8). The inscription, discovered on the northern wall of Cavity I, is written in cursive Aramaic script consisting of eight letters (width of letters 0.6–0.7 cm, average height of letters c. 1 cm). The letters resemble Aramaic letters that appear on ostraca from Maresha and its surroundings, which are dated to the fourth–second centuries BCE (Eph‘al and Naveh 1996; Eshel 2010). One of the characteristics of the letters is that the letters extend upward. The area with the inscription is small and was smoothed with a broad chisel. The following is a proposed reading of the inscription: ‘Abdalahi Zaba(n). The name ‘Abdalahi is a known Idumean name; it appears on an ostracon from the fourth century BCE (Eph‘al and Naveh 1996:80–81). The prefix ‘Abd is common among Idumean names of this period (Eph‘al and Naveh 1996:100; Eshel 2010:232). The meaning of the Aramaic root zbn is related to commerce, and the words zabun (buying) and zabina Idumean ostraca of the Hellenistic period (Eph‘al and Naveh 1996:46–47, 90–91). It seems, then, that the inscription is a record of a transaction conducted between ‘Abdalahi and another person, and that the merchandise was kept in the underground storerooms, where the inscription was discovered. (merchandise) are derived from it; the words zabuna and zabina appear on
 
A Greek Inscription alongside an Engraving of a Sailing Vessel (Figs. 9, 10). The inscription and engraving were discovered on the lintel of a ornamented opening connecting Cavity H with Cavity I. The Greek inscription was written in thin lines by a shaky hand (average width and height of the letters c. 4 cm). The following is a proposed reading of the inscription: Φ (Α/Λ) ΠΑTΡΙΚΟΣ. The first letter, Φ, probably denotes an abbreviation of a name, possibly the private name of the engraver. This is followed by the name Patrikos which is apparently a Greek form of the common Latin name Patricius. The meaning of the Latin name is “noble” and “honorable”, and it is known from the patrician class in the Early Roman period. The meaning of the Greek name Patrikos is “belonging to father” or “it comes from father”. The Greek form of the name is not common in Israel, but it appears in other inscriptions that were discovered in the Mediterranean basin, mostly from the first–second centuries CE (Fraser and Matthews 1994: 36). It seems that this is the name of the owner of the underground storeroom, probably in the Early Roman period.
To the left of the inscription is a schematic engraving of a sailing vessel (length 26 cm, height c. 12 cm). The sailing vessel has an elongated deck with a mast in its center (height 6 cm) and a yard at its top (length 18 cm); it seems to be sailing to the left. A kind of rectangular sail is hanging from the yard. Two diagonal lines drawn from the ends of the yard to the center of the deck probably represent two braces. A vertical line is depicted above the bow of the vessel. At the top of the line is an elliptical object that might be the curved end of the boat’s prow, as portrayed on sailing vessels from the classical world, or a schematic depiction of the a foresail (artemon). A line on the right side beneath the stern which connects to an L-shaped object covered with hatching marks is probably the sailing vessel’s rudder. Below the deck are engraved vertical lines; at their end is a square pattern, probably representing the oars. Judging by the proximity of the sailing vessel to the Greek inscription, it seems that it should be dated to the Early Roman period as well.
Depictions of similar sailing vessels were discovered in the past at Maresha and in its environs. The most notable of these was engraved on a small stone that was exposed in excavations on behalf of the Palestine Exploration Fund more than one hundred years ago, and is dated to the Hellenistic period (Gibson 1992). Two sailing vessels were engraved on walls of a rock-hewn burial complex from the Hellenistic and Roman periods north of Tel Maresha (Haddad and Artzy 2011). In this context we should also mention the sailing vessel drawn in charcoal on the wall of the anteroom in Jason’s Tomb in Rehavia, Jerusalem, dated to the Hasmonean period and to the Early Roman period (Rahmani 1964:8–9). As for the significance of the engraving of the sailing vessel, there are many interpretations in the research literature. According to one interpretation, depictions of sailing vessels have a protective role, as they were meant to guard against evil spirits (Kashtan 2001).
 
The documented underground complex is composed of numerous cavities and rooms, which were used for a variety of purposes during several periods. Several phases were identified in the complex. In the first phase, probably in the Hellenistic period, Cavities E, H and I were hewn and utilized as quarries where rectangular blocks of chalk were produced. The rock in the region is composed of an upper layer of hard nari overlying soft chalk (qirton). Apparently, the underground quarries were meant to exploit the chalk, which is easily hewn, without quarrying through the hard layer of nari. Later, but still during the Hellenistic period, the quarries were adapted for use as underground storage space; Cavity E was probably used as a cistern. The architectural characteristics of Cavities E, H and I resemble those of underground cavities from the Hellenistic period, dozens of which were discovered at Maresha; these cavities, arranged in a clover-leaf plan, include a staircase with a low, hewn railing and an entrance with a magnificent façade. The Aramaic inscription discovered on the wall of Cavity I, in which a commercial transaction is apparently recorded with a person of Idumean origin, may indicate that Cavities H and I were used in the Hellenistic period by Idumeam residents from the region. These cavities are located on a hill west of the lower city of Maresha, and it seems that they were part of the industrial–agricultural complex that existed around the city in the Hellenistic period.
Cavities H and I continued to be used as storerooms in the second phase, probably in the Early Roman period. Judging by the Greek inscription on the lintel of the ornamented opening connecting Cavity I with Cavity H, it seems that the storerooms were taken over by new owners. By this time, the city of Maresha had declined from its former greatness, and was probably replaced by a rural, whose inhabitants utilized the underground cavities in its vicinity. It was presumably during this period that Cavity B, which was used as an underground storeroom, and Shaft A, which connected this storeroom to the building located above it, were also hewn.
In the third phase, Cavities B, E, H and I were linked together by a system of low narrow tunnels for the purpose of creating a subterranean hiding complex. In this phase, the original openings of the cavities were deliberately sealed, and small rooms for hiding and storage were hewn throughout the complex. The pottery and glass artifacts discovered in the complex date to the period between the two Jewish revolts against the Romans (70–136 CE) and corroborate the assumption that the hiding complex was established as part of the preparations by the local inhabitants for the approaching Bar Kokhba uprising. The features of the complex belonging to this phase are similar to those of hundreds of hiding complexes from the time of the Bar Kokhba uprising that were discovered in the Judean Shephelah. These include the use of earlier subterranean installations; tunnels that connected the installations; a tunnel for the clandestine drawing of water; and tunnels with arched ceilings, uneven levels, sharp changes in their direction and hewn recesses for oil lamps in their walls.