The Upper Stratum. Wall remains (W2, W4–W6, W10, W11, W13) that formed three architectural units (1–3) were exposed right below the top soil. These were probably parts of small rooms belonging to a large building; the northern part of these units was located beneath a modern structure situated outside the excavation area. The walls were built of two rows of roughly hewn fieldstones with a fill of soil and small stones in between; ashlar stones were incorporated in the construction of some of the walls. They were preserved to a height of one or two courses. The walls were founded on bedrock, and in places where the bedrock was low the area was leveled with fill and small fieldstones. Architectural Units 1 and 2 were delimited on the south by W5 (width 0.8 m; Figs. 5, 6), which continued intermittently for most of the length of the excavation area; it seems that W13 was the continuation of W5 on the eastern side. A surface of small fieldstones and soil (L207), part of which abutted W5, was exposed in Architectural Unit 2. It is possible that Surface 207 was intended to create level bedding for a pavement that was not preserved. A similar surface was exposed south of W5 (L206). Architectural Unit 3 included Walls 10, 11 and 13 (Figs. 7, 8), which apparently enclosed several rooms. Wall 10 extended north and south beyond the excavation area. Wall 11 was severed at the eastern end of the square. It seems that W13 was attached to W10 in a later phase, possibly a secondary phase or a repair.
Ceramic artifacts dating to the Byzantine period and the beginning of the Early Islamic period (sixth–seventh century CE) were discovered in the excavation of the upper stratum. The pottery discovered in the excavation of Unit 1 included a jug (Fig. 9:22). The pottery in Surface 206 of Unit 2 included bowls (Fig. 9:3, 4), a cooking pot (Fig. 9:15) and a jar (Fig. 9:19). The finds discovered in the soil fill that covered Unit 2 included bowls (Fig. 9:1, 2), kraters (Fig. 9:8, 9), a lid (Fig.9:11), a cooking pot (Fig. 9:16), a jar (Fig. 9:18) and a juglet (Fig. 9:24). Other artifacts discovered in these soil fills included numerous roof tiles (Fig. 10) and many white tesserae, indicative of the structure’s roof and pavement. The ceramic finds discovered in the excavation of Unit 3 included bowls (Fig. 9:6, 7), a krater (Fig. 9:10), a lid (Fig. 9:12), cooking pots (Fig. 9:13, 17) and jars (Fig. 9:20, 21).
 
The Lower Stratum. Three sections of walls (W3, W7, W12) were exposed in Sqs 1, 3 and 4. They were constructed in a north–south direction and were not connected to the rooms of the upper stratum. Wall 3 (length c. 0.8 m, width 0.4 m) was built of various-sized fieldstones. A thin layer of gray cement attached to the bedrock was revealed around the wall. Wall 12 (length c. 1.3 m, width 0.6 m) was built of one row of large fieldstones. A layer of tamped white chalk (L410) was exposed alongside the wall. Wall 7 (length c. 1.5 m, width 0.8 m) was built of a single row of various-sized fieldstones; it was not connected to the other remains in the stratum. A rock-cutting (L302; 0.3 × 0.6 m, depth c. 0.4 m; Figs. 11, 12) was exposed south of the wall. In the soil accumulation overlying the stratum’s remains were ceramic finds similar to those of the upper stratum; they were dated to the Late Byzantine period and the beginning of the Early Islamic period (sixth–seventh century CE) as well. The finds included a cup (Fig. 9:5), cooking pot (Fig. 9:14) and a jug (Fig. 9:23).
 
In addition to the pottery and the large quantity of roof tiles and tesserae, finds from the excavation included glass vessels (Winter, below), which are dated like the pottery to the Late Byzantine period and the beginning of the Early Islamic period, coins, two unidentified metallic objects and a ring (Fig. 13), a crushing stone of an oil press (Fig 14) and a stone crushing basin (Fig. 15). These artifacts were discovered in the soil accumulation that covered the remains from the wall tops down to bedrock. Five coins were discovered: a coin from the reign of Antiochus III (minted in Tyre; 198–189 BCE; IAA 143524), a coin from the time of Alexander Jannaeus or an imitation of such a coin (IAA 143528), a coin dating to the fourth century CE (IAA 143529), a half follis from the time of Justinian II (minted in Thessaloniki; 574/5 CE; IAA 143525) and a follis dating to the reign of Constans II (minted in Constantinople; 641/2–642/3 CE; IAA 143526).
 
The Glass Finds
Tamar Winter
 
The glass finds from the excavation included some 60 small fragments, of which 23 are diagnostic pieces. The finds included bowls, wineglasses, bottles, oil lamps, windowpanes, a bracelet and a gold-glass tile, all characteristic of the late Byzantine and early Umayyad periods.
 
Bowls are represented by several rounded rims (L404, L405, not illustrated).
Wineglasses. Several rims (L405, L409, Fig. 16:1, 2) probably belonged to wineglasses, and a base with a single-bead stem (L 302, Fig. 16:3) supported a wineglass of a type commonplace in Jerusalem and its vicinity.
Bottles are represented by body fragments and rounded rims (L201, L403, L406, L409, not illustrated).
Oil lamps. Fragments of at least seven oil lamps were recorded, some with an outfolded rim (L405, Fig. 16:4; L104, not illustrated), others with a large wick tube (L406, L411, Fig. 16:9, 10; L301, not illustrated), resembling examples from the Western Wall Tunnels (Permit No. A-5124). The various oil-lamp handles included simple trail handles (L404–L406, Fig. 16:5–7; L103, L210, L301, L304, not illustrated), four of which probably belonged to two different lamps, two to each lamp (for lamps that originally has three handles each; L405, L406). One of the handles has a tooled trail extending below it (L210, Fig. 16:8).
Windowpanes. Two tiny windowpane fragments (not illustrated) were recovered, one from a surface level (L201), the other from an accumulation east of W10 (L403).
Bracelet. A blue twisted bracelet, possibly dating to the Byzantine period, was recovered in an accumulation over W4 (L210, 16:11).
Gold-glass tile. Only a small piece from a corner and an edge of a gold-glass tile was preserved (L302, Fig. 16:12). It was made of two glass layers (the bottom: 5 mm thick, the top less than 1 mm thick) with a gold leaf between them. Gold-glass tiles, alongside gold-glass tesserae, adorned architectural elements and wall mosaics in ecclesiastical structures in the region during the sixth–seventh centuries CE. Gold-glass tiles were found in small numbers in Jerusalem and other cities, as well as in smaller settlements throughout the country (Gorin-Rosen 2006). Moreover, the gold-glass tile from Umm Tuba is an exceptional find, and as only very few have been documented so far from Jerusalem and its environs, it constitutes a significant addition to the known glass corpus of the period.
 
The glass vessels and other glass artifacts from Umm Tuba are typical of assemblages from complexes in Jerusalem that date from the sixth–seventh centuries CE, as the one from the excavation at the grounds of the Jerusalem International Convention Center (Gorin-Rosen 2005: Figs. 2, 3).
Furthermore, the choice of glass vessel types, the relatively large number of oil lamps, the occurrence of windowpanes and the gold-glass tile may all suggest that the structure/s excavated at Umm Tuba was/were of an ecclesiastical nature. This coincides with other evidence from the excavation, such as a white mosaic pavement and fragments of roof tiles. Previous excavations at the site yielded a glass assemblage dated mostly from the seventh–eighth centuries CE, including a quadrangular windowpane, as well as numerous bottles, a few which may have also appeared in a later period, i.e., the ninth–tenth centuries CE (Permit No. A-4397).
 
The architectural remains of the upper stratum in the excavation were apparently part of a large public building from the sixth–seventh century CE. This structure might have belonged to one of the wings of a monastery whose remains were exposed at the site in 2005 (Eirikh-Rose 2007). Other parts of this monastery apparently extend southwest of the remains excavated in 2005, as well as in a private plot located c. 20 m northeast of the current excavation area (Fig. 17). Architectural remains damaged several years ago in this lot included mosaic floors, one of which was made of colored tesserae, wall sections and a cave whose ceiling was supported by a rock-hewn pillar (not excavated; ‘Adawi 2010:115–117). The pottery assemblage from the excavation is missing elements that clearly date to the Abbasid period, such as glazed vessels or Mafjar types; thus, a probable conclusion is that the building went out of use in the late seventh or early eighth century CE, i.e., before the Abbasid period. This conclusion is corroborated by the glass assemblage. The coin from 641/2–642/3 CE seems to indicate that the building went out of use around this time, perhaps as a result of the earthquake that struck in 660 CE (Amiran 1950–51; Russell 1985). The structure’s building stones were removed from the site or taken for construction in secondary use at the site and its environs. It seems that the remains discovered in and around the site should be ascribed to the Byzantine settlement of Metufa, or a contemporary monastery which was built by St. Marinus and is mentioned by St. Cyril of Scythopolis in one of his bibliographies about the monks of the Judean Desert in the sixth century CE (‘Adawi 2010:115, 137–138).