The installation was hewn at an elevation of 367 m asl, outside the inhabited limits of the ancient site. It was completely plastered and its plan indicates that it was constructed for use as a ritual bath (miqveh): it included a shallow basin, a stepped entrance and an immersion chamber, all rock-hewn and arranged along a north–south axis (Fig. 1). The stepped entrance was entirely cleared, and a trench was excavated along the immersion chamber, from the western lintel to the wall opposite the opening. Thus, more than one third of the immersion chamber was exposed, and the installation’s architectural plan and operating phases were understood. The bottom of the installation was found covered with a layer of alluvium and collapsed stones. On the walls of the immersion chamber were remains of whitish pink plaster applied to a bedding of ribbed jar fragments characteristic of the Late Roman and Byzantine periods; no earlier layer of plaster was identified. A graffito with Christian motifs was engraved on the eastern doorjamb. It therefore seems that the ritual bath, which was installed at the time of the Second Temple, was used as a water cistern in the Byzantine period.
The shallow basin was rectangular in shape (L104; 0.60 × 0.85 m, depth 0.15 m; Fig. 2) and was not plastered; it may have served for cleaning feet or filtering the water that drained into the installation. A low rock-cut partition (width 0.2 m) separated the basin from the stepped entrance (L101; length up to the doorjamb 2.6 m, width 1.2 m). Eight steps (rise 0.3–0.4 m, average run 0.3 m) were hewn the full width of the entrance and led to the opening of the immersion chamber. Part of a hewn threshold that fitted a door (length 0.85 m) was preserved at the top step (width 0.4 m): a depression in the threshold served as a hinge socket and a slot served to accommodate the other side of the door. The width of the door opening matched the width of the hewn basin in front of it, suggesting that the two elements were planned together. Chisel and leveling marks were visible on the upper parts of the entrance’s long walls that probably supported a roof, possibly a barrel vault. The run of the lowest step (c. 0.6 m), which abutted the doorjambs of the immersion chamber (opening dimensions: width 0.8 m, height c. 1.8 m, threshold run 0.5 m) was sufficiently deep to allow one to stand comfortably. The eastern doorjamb in the opening of the immersion chamber was engraved with an intricate graffito (Fig. 3): two large birds, probably doves or partridges, depicted in great detail, an arched installation (Fig. 4), probably a barred cage, slightly to their right. Two smaller birds are visible in the rear. The tail of the upper on is spread out in a fan-like manner, probably to portray peacocks (Fig. 5). Birds beside a cage or inside one symbolized in ancient Christian art the Holy Spirit or the spirit of the faithful imprisoned within the body. The bird motif appears in several Byzantine-period mosaics in Israel, such as one at the Armenian chapel north of Damascus Gate in Jerusalem (Bliss and Dickie 1898:253–259) and the mosaic pavement in the Byzantine synagogue at Ma‘on (Avi-Yonah 1961). Above the birds was an engraved Christogram—a monogram consisting of a cross combined with the Greek letters chi and rho designating the first two letters of the name Christos—hence it is obvious that the engraver was Christian (Fig 6). Below the monogram and to its right is an intricate graffito composed of many lines, the nature of which is unclear (a fish?). On the western doorjamb is another graffito of a chi-rho monogram (Figs. 7, 8). Although the engravings were most likely done by a Christian and the installation was meant to serve as a ritual bath, the graffiti probably do not depict the baptism of Jesus by John the Baptist in the Jordan River despite the common theme of doves as a symbol of the Holy Spirit (Matthew 3:16; Ziffer 1998:105–110), since it depicts several birds rather than a single bird. It therefore seems that the birds were engraved alongside the Christogram to intensify its protective power and sustenance or as a depiction of paradise, like many portrayals of this combination in early Christian art (Ziffer 1998:111). The graffiti were damaged by pitting with chiseled over the doorjamb, in preparation of a rough surface before applying a layer of plaster. The plaster was preserved in several place near the graffiti, indicating that the latter predates the plastering of the doorjamb.
The immersion chamber was almost square (3.50 × 3.75 m, height from the floor of the pool to the ceiling c. 3.5 m). Three steps (run 0.4–0.5 m, average rise 0.3 m, rise of the bottom step from the pool’s floor 1 m) were hewn in its floor. Hydraulic plaster (thickness 0.05 m) was applied to the walls of the chamber, its floor and the approach steps leading up to the opening. It included an undercoat of white-colored mortar (thickness c. 1.5 cm) embedded with ribbed body fragments of jars, characteristic of the Late Roman and Byzantine periods (Fig. 9), and above them – an upper layer of white plaster mixed with small stones (Fig. 10). A depression (L103; diameter 0.3 m, depth 0.24 m) that opens to the south, toward the lower step (see Fig. 9), was hewn into the western part of the upper step. The depression was not treated with plaster, and its upper part cut through the layer of plaster on the top of the step. Thus, it is clear that the depression postdates the plaster. Judging by its shape, the depression probably served to hold a jar for drawing water from the chamber.
None of the excavation finds are likely to aid in dating the installation; its date can be thus determined only according to its architectural plan, the texture of the plaster and the sherds in its bedding, and the motifs appearing in the graffiti. The architectural plan of the staircase leading to a stepped underground cavity is typical of ritual baths in the Judean hills and the Judean Shephelah during the Second Temple period (Zissu 2001; Reich 2013). At least three additional ritual baths typical to this period were documented at the site, one in a previous survey and at least two others during recent inspection work by members of the Unit for the Prevention of Antiquities Robbery. However, whereas the plaster on the walls of these three ritual baths has a texture typical to the Second Temple period—it is gray and includes small gravel—the texture of the plaster in the aforementioned installation is completely different and includes a fairly uniform layer of body sherds of ribbed jars characteristic of the periods postdating the Second Temple period. The Christogram engravings on the doorjambs of the opening, the likes of which are generally dated to the Constantinian dynasty and later, and at the earliest to the late third century (Jensen 2000:138), also indicate a date later than the Second Temple period and even later than the Bar Kokhba uprising. The graffiti are covered with plaster, and it is therefore clear that the plaster is later than the late third century CE and should probably be dated to the Byzantine period. The fact that the graffiti were not visible during the final use of the installation may imply there was no connection between its use and the graffiti. Thus, it seems that the graffiti is of secondary importance in the function of the installation. In fact, it is the depression hewn in the top step of the immersion chamber, where a jar was placed to aid in drawing water from the installation, that alludes to the function of the installation during its last phase of use.
It is therefore possible to discern several phases in the function of the installation. A ritual bath was hewn sometime between the end of the Second Temple period and the Bar Kokhba uprising, like other baths discovered at the site, and it was meant to be used by the Jewish population that resided there. The size of the installation and its location outside the limits of the settlement reinforce the theory that the bath was public and was meant to be used toward the end of the summer, when it was either less convenient or impossible to use the private baths (Zissu and Amit 2008). However, since no remnant of the bath’s original plaster, exhibiting a texture characteristic of the period, was discovered, it seems that its construction was never completed; thus, it did not fulfill its intended purpose at that time. Later, most likely during the Byzantine period and at the very earliest in the late third century CE, a Christian engraved the graffiti on the installation’s doorjambs. Sometime afterwards, the installation was plastered, making it watertight so it could be used to store water. The engraver was probably a laborer engaged in plastering the installation, who let his imagination run wild using a sharp implement on the smooth surface. We assume that during this period the installation was no longer being used for its original purpose, although there is no certainty regarding this. The entire installation was utilized for water storage, as evidenced by the completely plastered outer staircase; the volume of the cistern was therefore estimated at c. 50 cu m. It seems that at some point during the installation’s use as a cistern, perhaps during the summer, the water level dropped into the inner chamber; this required and allowed the hewing of the depression in the top step of the immersion chamber, where those drawing water could place a ceramic jar.