The site is located on a terrace below Mezoqe On (the On cliffs), near Kibbutz Mevo Hamma in the southern Golan Heights, overlooking the Sea of Galilee (Fig. 1). The site was described for the first time during the Golan Survey (Hartal and Ben Efraim 2012: Site 28) as a large structure, apparently comprising several rooms and an outer wall built of large fieldstones. A plastered installation (1.2 sq m) was identified to the south of the structure. The finds included pottery dating from the Early Roman, Middle Roman and Byzantine periods, roof tiles and a fragment of red-painted plaster.

Drone imagery taken prior to the excavation showed the site to comprise a roughly rectangular enclosure (16.0–17.5 × 22.0–33.0 m; Fig. 2) with a central structure, probably a tower, and a set of rooms. Three excavation areas were opened: inside the tower (c. 2.0 × 4.8 m); in Room 1 (c. 2.5 × 2.9 m), adjacent to the tower on the southeast; and in Room 2 (c. 6 × 6 m) at the northeastern part of the enclosure.

 
The outer walls of the enclosure are built of two rows of fieldstones. The plastered installation noted in the survey is located southwest of the tower. The entrance to the enclosure was tentatively located in the northern corner, above the cliff. Pieces of Pompeian millstones were found lying in the central part of the enclosure. Roman and Byzantine pottery, together with fragments of roof tiles were observed. An oil press was identified approximately 190 m to the northeast of the enclosure (map ref. 261752/738928).
 
The tower (c. 5.6 × 7.6 m). The walls (width 0.8–1.2 m) are built of large, roughly hewn fieldstones. Ashlars are used mainly in the corners and for the doorjambs; some of these bear drafted margins with prominent bosses. The lower courses of the walls are constructed of smaller fieldstones. Of the northwestern wall (W172), above the edge of the cliff, only the corners survive. Wall 148 (width 0.9 m) divides the structure into two rooms. The entrance (width 0.9 m) is indicated by a doorjamb and a threshold in the northwestern wall (W124; Fig. 3) of the eastern of the two rooms (between W122 and W148). The upper levels in this room contained bricks and roof tiles. Below this, a floor (F135) was uncovered on the same level as the threshold. A probe (1.5 × 2.0 m) opened just inside the threshold revealed the pressing platform of a small screw-type oil press positioned upright and set within the floor. Only two side arms of the platform originally protruded above the floor surface (Fig. 3). Below this was a cellar (L146; Fig. 4) covered by a roof made of basalt beams that rested on corbels (F144) fixed in W122 and W148. The cellar, as well as the walls of the structure, were founded on the bedrock, approximately 1.2 m below the basalt beams (F135). Apart from pottery and lamp fragments, two bronze coins attributable to the late fourth–fifth centuries CE, and a fragment of a bronze fibula were found in the fill.
There is apparently a small courtyard (c. 55 sq m) located between the tower, Room 2 to the north and Room 1 on the southwest. To the northwest it is probably defined by the outer enclosure wall of the complex (W139). Walls 125 and 130–132 (all built of one or two faces of fieldstones), probably mark a series of rooms on the southern side of the courtyard (including Room 1).
 
Room 1 (c. 2.5 × 4.5). Its walls (width 0.4–0.8 m) are constructed of one or two rows of fieldstones. A probe opened in the southwestern half the room reached a sandy yellow floor (F133; Fig. 5) below a layer of collapsed stones. A sequence of fills below this floor yielded large numbers of pithos sherds and brick fragments. Below this sequence, remains of a large cylindrical clay oven (diam 1.8–2.0 m) were discovered. Within, collapsed stones and terracotta fragments, probably pieces of the oven itself, were excavated above a layer of ash (L150; Fig. 6).
 
Room 2 (outer dimensions c. 7.7 × 7.7 m, inner dimensions c. 6 × 6 m; Fig. 7) was accessed by two entrances: from the northwest through W128 (Fig. 8) and through an entrance in W167, oriented toward the what seems to have been an inner courtyard. Remains of a plaster floor were observed on the outer face of the threshold. The walls (width 0.7–0.8 m) are built of a single row of large, roughly worked blocks with small fieldstones in the gaps; the doorjambs and thresholds were constructed of ashlars. The northeastern wall (W169) is largely conjectural. Two phases were observed in the room.
The first phase floor was laid with a mosaic carpet. In the northwestern half of the room (F147; Fig. 9) it is composed of simple geometric designs in white, blue and red. A two-line Syriac inscription (length 4.4 m) bordered by two blue bands runs along the line separating the two halves of the room. The inscription begins with a cross, located just inside the entrance in W167; a schematic bird can be recognized above the inscription in the middle of the room. In the southeastern half of the room, the mosaic pavement is 7–8 cm higher than in the northwestern half, leaving a gap of 0.3–0.4 m between the two; this suggests that a partition wall divided the room. The mosaic in the southeastern part is apparently composed of rectangular geometric designs with inscribed crosses. The border of the mosaic in the southern corner suggests that there had been a bench here, later replaced by the surviving bench (see below). Fragments of white plaster were found in patches on the walls of the room; nevertheless, some painted fragments were also recovered.
In the later phase, a pilaster was built against W167, immediately next to the entrance. A stone bench (width 0.45 m, height 0.31 m) ran between the pilaster and the southern corner of the room and partly along W168. The edges of this bench encroach on the borders of the mosaic floor. A sandy yellow plaster floor (F141) was laid over the mosaic floor.
 
Most of the pottery from the excavations dates generally from the sixth–seventh centuries CE. The assemblage includes local wares—cooking vessels, red and black storage jars and basins—and a large number of pithoi with flat rims and a wavy decoration. Imported wares include Late Roman Red Wares (ARS, CRS and PRS). Only the fill of the cellar yielded mixed pottery of Early Roman and Byzantine date.
 
Given the evidence, we can distinguish several building phases in the complex:
(1) The erection of the complex in the Byzantine period, consisting of the tower and Room 2. Although the mixed Early Roman material found in the foundations of the tower may suggest an earlier date for its construction, the results of the excavations are not clear in this regard. Room 2 was probably divided by a partition wall into two halves.
(2) A new set of rooms was built on the southeastern side of the inner courtyard and against the tower: Room 1, including a large oven, and possibly other rooms. In Room 2 the partition wall was removed, and a pilaster and a new bench were added.
(3) After the oven in Room 1 went out of use, a new floor (F133) was laid above it, and the rooms southeast of the tower were re-arranged by a new dividing wall (W130). A similar floor was laid over the mosaic floor in Room 2, effectively levelling the area of the entire room.
We suggest that the complex represents a Byzantine rural monastery comprising a fortified central tower—possibly built already in the Early Roman period—with rooms attached, an outer courtyard to the southwest and an inner courtyard between the tower and Room 2. Room 2 is most probably a small chapel, indicated by the mosaic and the inscription. However, the room is not oriented to the east, has no apse and no indications of an altar or decorations typical of an ecclesiastical building were recovered, except for few tiny marble fragments. The Syriac inscription, still being studied, might provide more information on the date and function of the complex and perhaps also on the people behind its construction.