The excavation was located west of Tel Iztabba and Hanot Beshan (Khan al-Ahmar) on a plain that gently slopes southwest toward Tel Temes. The local soil is replete with travertine, indicating that the area was once a floodplain. Previous excavations nearby revealed architectural remains and installations dating from the Hellenistic to the Early Islamic periods (Har’el 2016; Fig. 1: A-6918, Permit Nos. A-5727, A-6002, A-7648), and hewn and built tombs from the Roman and Byzantine periods (Zori 1962:190) were discovered 500 m west of the excavation.
Three excavation areas were opened (A–C; Figs. 2–4), exposing four chrono-stratigraphic units: a Hellenistic period soil surface and potsherds (Stratum IV), an Early Roman period mausoleum (Stratum III), a Middle–Late Roman periods tomb (Stratum II), and Mamluk period architectural remains (Stratum I). A few finds from the end of the Byzantine period were found outside any architectural context.
Stratum IV. In Area A, a layer of soil containing abundant Hellenistic pottery was uncovered (L221; Fig. 5), including a Rhodian seal-stamped handle. This soil layer is probably alluvial, deriving from an ancient stream that flowed nearby. Preliminary tests northwest of the excavation area showed that the alluvial fan extends a few dozen meters in this direction. The site’s proximity to the Hellenistic city at Tel Iztabba suggests that a small village or farmstead was located there.
Stratum III. The remains of an underground mausoleum were uncovered in Area B (Figs. 3, 6). It was built into sterile soil consisting of compact clay and numerous travertine deposits. The mausoleum follows the plan of rock-hewn kokhim caves (Feig 1999: Plan 1; Ovadiah 1999: Plan 1; Tatcher and Gal 2009: Plan 3), including two levels of burial niches, one on top of the other. Only the remains of the upper level were excavated, including a central chamber (L238) and six wall-enclosed kokhim (i.e., burial niches; L235, L236, L237, L239, L251, L254): two on each of its north, east and west sides. The walls enclosing the niches consisted of a single row of dressed basalt stones: Their inner face was smooth while their outer face was partially dressed, probably to fasten them to the surrounding soil. Only the lower courses were preserved; the upper courses collapsed, producing a well-defined rubble layer (L235; Fig. 7) of stones and ossuary fragments (Fig. 8).
The excavation in Chamber 238 uncovered parts of the northern and eastern niches’ facades (W248, W249; Figs. 9, 10, respectively). They were built with care, but only the side facing the main chamber seems to have been well-dressed. Burial Niche 237 was well-preserved (0.8 × 2.1 m, preserved height 0.5 m) and is probably indicative of the other niches’ dimensions too. Niches 236, 237 and 239 included a travertine-based plaster floor (thickness 0.5 cm). The goal may have been to provide the niches a limestone-like appearance, imitating rock-cut burial caves. Unfortunately, whether the niches’ facades were also plastered could not be determined due to the many vegetation roots in the niches.
Two horned limestone sarcophagi lids were discovered in Chamber 238, c. 1 m below the tops of Walls 248 and 249 (Fig. 11); they were not excavated. The northern of the two lids may have broken due to the collapse of the upper level. The chamber probably held a row of stone sarcophagi that faced W248.
Although the mausoleum was only partially excavated, a tentative reconstruction can be proposed. Assuming that both levels shared the same plan, the tomb would have had at least 12 burial niches. Combined with two stone sarcophagi and one ossuary, the number of interred individuals reaches fifteen. Furthermore, given that the sarcophagi lids seem to have faced north, the entrance must have been from the mausoleum’s southern side.
Accumulations on the mausoleum floors yielded potsherds dating from the second half of the first century BCE to the mid-first century CE (Sandhaus, below), glass fragments from the late first–third centuries CE (Gorin-Rosen, below) and some poorly preserved iron nails, perhaps indicating the use of wood, possibly for coffins, closing niches or building benches.
Stratum II. Above the collapsed upper level of the mausoleum in Area B, the remains of a cist grave (L220; Figs. 3, 12, 13) were uncovered. It was built of small and medium-sized fieldstones probably acquired from the western part of Stratum III W247. The grave yielded five bronze rings (Fig. 14) to which a piece of metal was welded. Apparently, these rings were the handles of a wooden coffin that was not preserved. The grave’s proximity to the mausoleum suggests a connection between them. Since the grave lacked cover slabs and contained no bones or grave offerings, it was probably robbed in antiquity, rendering its date unclear. However, on stratigraphic grounds, it is safe to say that the grave was built after the mausoleum fell into disuse and, therefore, should be dated to the second–third centuries CE or later.
Stratum I. In Area C, alluvial layers were removed to reveal construction remains with no clear plan (L223; Figs. 4, 15): a level of ashlars, fieldstones and even gravestones lying directly on sterile soil. Among them, Mamluk pottery (Sandhaus, below) and several stone items (Tzin, below) were found. These remains may be part of a temporary structure or an agricultural installation.
The Pottery
Debora Sandhaus
The pottery from the excavation dates from four periods: second century BCE (Fig. 16), first century BCE–first century CE (Fig. 17:1–13), second–third century CE (Fig. 17:14–16), and the late twelfth–thirteenth century CE (Fig. 18).
Second Century BCE. The excavation yielded cooking pots with a straight neck and a simple rim (Fig. 16:1, 2), a juglet with a cup-like rim (Fig. 16:3), jars with a short neck and a folded, slightly everted rim (Fig. 16:4–6), jars with a rounded rim (Fig. 16:7, 8), and Rhodian amphora handles (Fig. 16:9, 10). Visually, it seems that Cooking Pot 1 and Jars 3 and 4, made of friable orange clay that leaves powder on the hands, were made at the site. In contrast, Jars 7 and 8, made of white clay with a few red inclusions, were brought to the site from the northern coastal region (Berlin, Monnickendam-Givon and Stone 2022).
First Century BCE–70 CE. Most of the pottery from the area of the mausoleum—particularly that discovered on the floors—dates from this period, thus dating the final use of the structure. Since no building phases were identified in the mausoleum, it may also have been built in this period. The assemblage includes tableware and storage vessels common in the Jordan Valley and Galilee. These vessels include an Eastern Terra Sigillata red-slipped bowl (Fig. 17:1) that was probably made in Cilicia (Rosenthal-Heginbottom 2014:387; Lund 2015; Berlin et al. 2022), the base of a cup made of local clay (Fig. 17:2), cooking bowls with a rounded wall and a grooved rim (Fig. 17:3, 4), two closed cooking pots, one with a groove on the inside of the rim (Fig. 17:5) and another with a slightly flattened rim (Fig. 17:6), jugs with a variety of rims—rounded (Fig. 17:7), triangular (Fig. 17:8) and folded (Fig. 17:9)—locally made jars with a groove on the inside of the rim (Fig. 17:10), jars with a simple rim (Fig. 17:11, 12) and an amphora ( Fig. 17:13).
Second–Third Centuries CE. The pottery from the collapsed stones on top of the mausoleum comprises local cooking ware, including a cooking pot with a slightly concave ledge rim (Fig. 17:14) and a cooking pot with two grooves on its rim (Fig. 17:15), and an Aegean amphora (Fig. 17:16; Peacock and Williams 1986:193–194, Class 47) dating from the end of the second century to the fourth century CE.
Mid-Twelfth–Thirteenth Centuries CE. Area C yielded little pottery that consisted of glazed bowls with a ledge rim and yellow slip (Fig. 18:1, 2)—a type that began to appear in the mid-twelfth century CE and continued until the Ottoman period (Avissar and Stern 2005:19)—a simple bowl with a ledge rim (Fig. 18:3), and a closed cooking pot with a thickened rim (not illustrated), typical of the third quarter of the twelfth century CE (Avissar and Stern 2005:91).
Based on the pottery, most of the activity at the site took place in the Hellenistic and Early Roman periods. Some activity may have occurred to the north of the site in the second century BCE, and the finds were swept down the streambed. The mausoleum was built in the second half of the first century BCE and was used until the first century CE. The meager pottery from the second–third centuries CE and from the mid-twelfth–thirteenth centuries CE probably shows that only limited use was made of the site in these periods.
The Glass Finds
Yael Gorin-Rosen
The excavation yielded fragments of 22 glass vessels, 16 of which were identified and dated to two periods: the Roman period (late first–third centuries CE), associated with the mausoleum, and the Byzantine period, associated with loci closer to the surface.
Roman Period. The glass fragments of this period include a delicate bowl with fire-rounded rim (not illustrated), a body fragment of a bowl with a ledge rim and a hollow double fold near the rim's edge (L236; not illustrated), a type that was common between the Great Revolt and the Bar Kokhba Revolt (Jackson-Tal 2016:40, 42, 44, Fig. 10:5, 6, and references therein), an infolded and flattened bottle rim (Fig. 19:1) reminiscent of candlestick bottle-type rims, although it may also belong to a bottle with a globular body, and a neck and a shoulder of candlestick bottle (L237; not illustrated). Candlestick bottles are characteristic of Roman-period burials, from the late first to the early third centuries CE.
A slightly everted and cut rim with an open inner fold below it (Fig. 19:2) derives from a bottle or jug and is attributed to the Late Roman period. A similar rim, with a pointed edge, was found in a Late Roman period assemblage at Qiryat Sefer (Magen, Tzionit and Sirkis 2004:236–237, Pl. 9:30). Another similar bottle rim with an open fold or step below its edge was found at Jalame, albeit shaped slightly differently (Weinberg and Goldstein 1988:72–73, Fig. 4-34:292).
Byzantine Period. A number of Byzantine vessels were found (not illustrated): a rim of a wine glass; a delicate bead-like stem belonging to a wine glass or a stemmed oil lamp; and a bottle with an upright fire-rounded rim. These vessels are very similar to vessels found in excavations at Bet She’an and the surrounding area, as well as throughout Israel and neighboring countries in the Byzantine period.
Area C yielded a fragment of a mold-blown base with a rosette design (Fig. 19:3). Remains of a mold-blown pattern also appeared on its side, judging by a small preserved fragment. The base’s shape suggests that the vessel was multifaceted and may have been hexagonal. One possible reconstruction is of a mold-blown jug with six facets bearing various designs and a rosette on the base. Perhaps, one can reconstruct the lower part of the design of a branch on one of the facets. A metal mold with a rosette on the base discovered in Samaria was probably used to produce vessels of this kind. A blessing jug with a molded design like the one described above was found in a tomb behind the apse wall of the church at Kursi (Katsnelson 2014:199–202, Fig. 2, and see extensive discussion and references to other fragments discovered in the country). Two additional juglets of this type are located in the Dobkin collection in the Israel Museum and have designs on their base: One has six lines forming a kind of schematic flower, and the other has a rosette with twelve petals (Israeli 2003:278–279, Cat. Nos. 371, 372). These jugs are blessing vessels called eulogiae and date from the sixth–seventh centuries CE; they were used to contain oil, water or sacred earth from the Holy Land. The symbols depicted are attributed to Christians or Jews, and some are decorated solely with geometric patterns (for a discussion of this group, see Israeli 2003:270–271, 277 and references therein).
Most comparable vessels are kept in private collections and are rarely found in excavations, hence the great significance of the present fragment. The vessel from Kursi is one of the few found in situ. Fragments of this vessel type were also found in a glassware workshop near the street leading to the northern gate in Bet She’an at the foot of Tel Iztabba (License No. G-12/1994; unpublished). The discovery of this vessel may indicate the existence of a late Byzantine burial at the site or, alternatively, a religious structure of some sort. In any case, this vessel is important; it must have fetched a high price compared with common tableware or oil lamps, and its owner probably purchased it for its symbolic value and content.
A fragment of a thick glass plaque or disk was also recovered (L236; Fig. 19:4). This object was made of hot glass flattened on a flat surface. Similar plaques/disks sometimes bear spirally marks of the hot glass’s rotation. Comparable finds were recovered in Nazareth (Bagatti 1969:313, Fig. 237:37) and at Zippori, where two examples of different sizes were found beneath the synagogue floor and dated to the Late Roman and early Byzantine periods (Gorin-Rosen 2005:299, Pl. D.I:9, 10). Their purpose is unknown, but they may have been used as weights.
Area C also yielded a single glass mosaic tessera.
The glass finds from the current excavation augment our knowledge of burial practices in Bet She’an in the Early Roman period and provide a significant addition to the distribution of late Byzantine eulogiae in Israel.
Stone Items
Barak Tzin
Stone items found in the excavation include a shattered ossuary, three stone artifacts and a fragment of a limestone capital or cornice. The ossuary (Fig. 8) is made of soft limestone, probably chalk. Its exterior is undecorated and has multiple chisel marks, while the interior is carefully smoothed. The stone artifacts include a marble stopper and a smoothed basalt grinding stone (L221; Fig. 20:1, 2) that may be associated with funerary practices, like preparing various materials or sealing a bottle/juglet. Another stone artifact was a whetstone of green slate (Fig. 20:3) found in an accumulation near the Mamluk architectural remains. The item is polished, and several fine lines visible on it are probably sharpening marks. Comparable finds are common at sites of the Early Islamic, Crusader and Mamluk periods (Kletter and Stern 2006: Fig. 26:8; Barbé and Shapiro 2012: Fig. 11:4–5; Barbé 2015: Fig. 2:27). The fragment of a capital or cornice (not illustrated) is poorly-preserved. Nevertheless, it suggests that the mausoleum was splendidly decorated and, together with the travertine plaster, probably provided the structure with the appearance of a limestone cave.
The excavation uncovered remains and finds from the Hellenistic, Roman and Mamluk periods. Presumably, during the Hellenistic period, there was a settlement nearby associated with the city at Tel Iztabba (Bar-Natan and Mazor 1994).
The Early Roman mausoleum may have been built to imitate a burial cave hewn in limestone rock. The construction of a two-story mausoleum requires considerable effort and attests to the high status of those interred in it. The highest status individuals may have been placed in the northern niches on the upper level, which constituted the most prominent position, facing the entrance. Most of the ceramics from the mausoleum date from the first century BCE–first century CE, and some are from the second–third centuries CE, whereas the glass finds date from the first–third centuries CE. Most of these finds come from the upper level, which may have been used later than the lower story, which was not excavated.
Thus, the mausoleum may have been a family tomb used over an extended period, possibly associated with the burial ground at Tel Temes (Zori 1962:190), as was the later cist grave. The mausoleum and grave were located on the northern boundary of this extensive burial ground, which surrounded the northern and western fringes of Bet She’an. Conversely, the tombs may have been part of another cemetery located c. 500 m west of the current excavation, where Roman and Byzantine sarcophagi and built tombs have been identified. Mausolea were mostly built after the Early Roman period. Middle and Late Roman-period mausolea were uncovered in Tiberias (Stepansky 1999; Vitto 2008), Iksal (Mokary 2014) and Susita (Eisenberg 2016:16–17), while Late Roman and Byzantine mausolea were uncovered in Bet She’an (Avshalom-Gorni 2000; Tepper 2008; Sion 2014). Although magnificent mausolea from the Roman period are found near major cities such as Tiberias, Bet She’an and Susita, these structures seem non-uniform; they were probably built according to local traditions.
Although no Byzantine architectural remains were found, the glass vessels, including the eulogia fragment, indicate a nearby occupation during this period.
The few Mamluk finds suggest a temporary presence in the region, perhaps associated with Hanot Beshan (Khan al-Ahmar) to the east of the site.