Et-Tuweiri is a known Byzantine site in the northern ‘Akko plain, slightly over 1 km west of Qibbuz Kabri and 2.5 km east of Nahariya (Fig. 1). It is situated on the cultivated plain floor, barely attaining a height of 30 m above sea level. The site has been severely damaged by intensive agricultural activity that turned up numerous artifacts and remains, many of which are stored and exhibited in various local settlements. Guérin (Description géographique, historique et archéologique de la Palestine; Troisième Partie: Galilée, 1-2. 1880. Paris), who visited the site in the nineteenth century CE, observed the occurrence of architectural elements and conjectured the existence of a church. The site was later surveyed by Frankel and Getzov (Map of Akhziv [1], Map of Hanita [2]: Site 2.184). The present excavation exposed the remains of a structure, assumed to be a church, dating to the Byzantine period and a later Muslim grave. This church is possibly the second one attested to on the site.


Evidence of intense destruction, which is characterized by a number of elements, appears at 0.15–0.40 m below surface. Thousands of roof-tile fragments were found strewn throughout the two squares, as well as collapsed hewn kurkar and limestone masonry stones. Many of the stones had a thick plaster application and some were architectural elements, such as pillars, pillar bases, capitals and small decorative pillars (Fig. 2). Fragments of red-painted plaster were also ubiquitous. A large number of stones and elements were scarred by fire and intense heat, attesting to the destruction.

The section of a single, well-built wall (W112) that belonged to a large structure was found along the southern edge of the excavation (Fig. 3). This wall (width 0.6 m, exposed length 3.5 m) was constructed from two rows of hewn stones and a small-stone core; only two courses of its northern face were exposed, as its base was not reached (Fig. 4). The western end of W112 continues into the western balk and its eastern end is badly disturbed.

A rectangular, stone-built grave (L111; 0.8 × 1.6 m) abutted the northern face of W112 at its eastern end. The grave was clearly later than the wall it abutted, since it had four walls of upright stones and did not utilize the existing W112 as one of its sides. Stones and fragments of marble elements from the earlier destroyed structure were incorporated in the walls of the grave. A large stone used apparently to cover the grave was found pried up above the opening. The grave contained the poorly preserved remains of a 30–40 year-old male, lying on his right side, facing south and not accompanied by offerings. It may be assumed that this is a later Muslim burial of indeterminable date. A similar, but smaller, stone-built unit (L110; 0.40 × 1.07 m) that did not contain any human remains, was located 1.75 m further west and also abutted W112. An additional, partially exposed stone construction (L113) was directly south of W112. To facilitate the construction of L110 and L111, an area (L108) was prepared to the north of W112. It was cleared of collapsed stones, which were partly retained. Moreover, the eastern end of L110 was partially built on the collapsed stones.

The sterile alluvial soil directly below the collapse and destruction (Fig. 5) yielded no finds other than a few random, small and extremely weathered potsherds.

Aside from W112, many indicative finds enable us to define the structure as a church. A large quantity of marble architectural elements was present, including numerous profiled marble slabs (thicknesses 3–5 cm), probably remains of chancel screens, as well as a fragment of a channeled square chancel pillar.

A partial inscription, dating probably to the sixth century CE or possibly the late fifth century CE, was discovered on a broken profiled marble slab (Fig. 6). It is an invocation to St. Sergius who was commonly referred to during the Byzantine period. The remaining inscription is too short to enable its unequivocal identification. Di Segni suggests that it may be reconstructed as Boethi agie Sergi, amen (St. Sergius, help! Amen), as found on an inscription from Nessana.

A portion of a marble slab with a carved out triangle pattern, possessing pecked and roughly surfaced areas that were possibly prepared for inlay applications, can be partially reconstructed (Fig. 7). The acute and obtuse external angles of the slab suggest it had served as a screen for a gallery staircase banister. The smoothed edges surrounding its protruding obverse imply that it was itself inset, possibly in a frame. Fragments of additional slabs worked in a similar manner were also found (Fig. 8).


Marble was utilized for the columns, bases and capitals and for wall and floor applications, as evidenced by hundreds of marble tile fragments of varying thicknesses (1–3 cm; Fig. 9). The edges of the tiles are rounded, squared or beveled, while their shape may be square, rectangular or irregular. Remains of two spindle-shaped balusters, one of which could be partially reconstructed to a height of at least 0.75 m, were unearthed (Fig. 10). They may have served in a balustrade or as table legs. A section of a marble tabletop (Fig. 11), which has a concave profile and its corner is styled as a fold, was recovered. These properties appear to be more distinctive of ritual tables than chancel screens that usually possess a flat profile. Similar styling is visible on a ritual table uncovered in the North Church of Shivta (S. Margalit, PEQ 118–119:106). Another fine piece of white marble shows a relief of either a floral or faunal representation (Fig. 12). Numerous rectangular pieces of cut marble, perhaps intended for inlay, were among the finds.

Commonly found were large white tesserae, but concentrated in the northeast quadrant (L102) were hundreds of very small glass, colored stones and ceramic tesserae (average size 5 × 5 × 7 mm); some of the glass tesserae contained a gold leaf (see Gorin-Rosen, below).



The ceramic assemblage is typical of the western Galilee in the sixth and seventh centuries CE. It includes imported bowls (Fig. 13:1–5), many ribbed storage jars that were the most common vessel-type present (Fig. 13:6–9), including a Gaza storage jar (Fig. 13:10) and over twenty jar covers (Fig. 13:11–14). The ten rims and several large ridged handles of pithoi and doliya (Fig. 14:1–5) and the one imported amphora (Fig. 14:6) are also common elements in the western Galilee repertoire at this time. Fragments of only three cooking vessels and a casserole cover were identified (Fig. 14:7–9).

Other than ceramic finds, many iron nails were retrieved from the destruction layer.


The remains uncovered in the excavation point to the existence of a church at the site. This discovery is well incorporated into the known picture from the western Galilee, particularly the northern coastline, which was the center of an extensive Christian settlement during the Byzantine period, as reflected in the various researches of M. Aviam (Five Ecclestiastical Sites in Western Upper galilee. In Eretz Zafon: Studies in Galilean Archaeology. Jerusalem 2002, pp. 165–218; Jews, Pagans and Christians in the Galilee: 25 Years of Archaeological Excavations and Surveys, Hellenistic to Byzantine Periods. Rochester 2004, pp. 181–204).

The lack of finds datable to later than the mid-seventh century CE attests to the destruction of the church at this time, either due to the Persian conquest of 613 CE or the later Muslim conquest in the mid-seventh century CE. Many of the church-related structures in the region were similarly abandoned at this time. Nonetheless, it is not a universal occurrence, since a number of sites continued to exist into the Early Islamic period. Continuing occupation is evident at Kh. esh-Shubeika, Shelomi, Kh. el-Ghureiyib and Horbat Bata.

The final utilization of the site was for Muslim burials, as evinced by the built grave with a single interment. The two additional similarly built units possibly served the same function, although no proof or verification for this proposal is evident. Similar opportunistic use of existing walls for Muslim burials occurs frequently at ancient ruins, e.g., at nearby Kh. esh-Shubeika (D. Avshalom-Gorni and A. Tatcher. 2002. In Eretz Zafon, pp. 220–254).


The present limited excavation revealed the existence of an unknown, yet quite ornate church in the midst of the fertile agricultural soil of the northern coastal plain, an area that was a center of Christian habitation in the Byzantine period. The church, located on the southern periphery of a large Byzantine site (et-Tuweiri), was destroyed and abandoned in the first half of the seventh century CE.


The Glass Finds

Yael Gorin-Rosen


The majority of glass finds from the excavation were glass tesserae that included an impressive quantity of gold-glass tesserae (Fig. 15), as well as some broken vessels, window panes and glass debris. The glass vessels, dating to the Byzantine period, include bowls with hollow, out-folded rims, beakers and wine glasses with rounded rims, hollow ring bases and stems (Fig. 16), bottles with a neck decorated with wound wavy trails, various rims of bottles and fragments of bottle bases, bowl-shaped oil lamps with hollow out-folded rims and three handles (Fig. 17) and oil lamps with hollow, conical stems (Fig. 18).

Most of the c. 40 fragments of window panes belonged to the rectangular thick window type (Fig. 19) that has at least one rounded side (fragment on right). Other window pane fragments have traces of plaster left from the original window frames.

The c. 500 pieces of glass tesserae were made of bluish green, green, dark blue, cobalt blue, yellow, yellowish green, emerald green, brown and red glass, as well as colorless glass that sometimes has a light yellowish hue (Fig. 20). Most tesserae were of translucent glass and a minority was of opaque glass. A significant component of the tesserae were gold-glass tesserae, composed of two glass layers, a thick bottom layer and a very thin upper one, usually colorless or yellowish, with a thin golden sheet inserted between them (Fig. 21). Many of the gold-glass tesserae came from the margins of the original tile, as evidenced by the edges of the upper glass layer that were not fully spread over the lower glass layer. A large number of gold-glass tesserae were found split; sometimes the powdery gold traces were visible upon both layers and at other times, no traces of gold were discerned (Fig. 22).

The majority of tesserae were relatively small (0.4–0.8 cm) of low quality and irregular in shape; they were broken asymmetrically and had at least one rounded side. Other tesserae, which were elongated with one rounded side, reflected a particularly economical usage by the craftsmen who exploited the margins of the original tile, so that some of the tesserae were left with a rounded side that was not straightened as a precise cube.

Several of the vessels and window panes were found distorted by heat (Fig. 23) that may have resulted from the conflagration that consumed the building. At the same time, some of the distorted artifacts may have been part of debris from glass production that was carried out at the site. However, the melted glass debris, the chunks (Fig. 24) and the distorted vessels are insufficient to determine their source, whether industrial or from a destructive raging fire. The occurrence of the glass items together with the numerous glass tesserae may indicate that all derived from the destroyed church building.