Two abutting walls (W108, W110) whose direction indicated this building continued west and south beyond the excavated area, were exposed. The walls (width 0.9 m) were preserved a single course high and were constructed from both large field and hewn stones. No floors or living surfaces were associated with this structure.


Another structure was 3.4 m north of the first one. It consisted of two abutting walls (W106, W107) that extended west and north beyond the excavated area. The walls (width 0.9 m), preserved two courses high, were constructed from alternating courses of large field or hewn stones and small flat stones. No floors or living surfaces were associated with this structure. A probe north of W106 indicated that the walls were erected above a fill layer (L105), overlying sterile soil and local bedrock. 
A hewn cornice was found in the area separating the two structures. This cornice and additional collapsed stones probably originated from the south structure.

The two easternmost squares of the excavation were void of architectural remains and the probes (L101, L103) reached sterile soil.
The sparse finds recovered from the excavated area consisted of potsherds and glass fragments. The ceramic finds included a storage jar (Fig. 3:1) dating to the Early Roman period (first century CE), although the bulk of the potsherds were dated to the Late Roman period (second–fourth centuries CE). These were composed of Kefar Hananya Types 1B, 1C bowls (Fig. 3:2, 3), Kefar Hananya Type 4C cooking pot (Fig. 3:4), a krater (Fig. 3:5) and storage jars (Fig. 3:6–8). Pottery fragments from the Byzantine period were not found in the excavation.


Previous excavations of the Hebrew University expedition to Bet She’an exposed a branched water system of channels and ceramic pipes that extended in a southwest direction, beyond the city wall, toward the location of the current excavation. Due to the high topography it was postulated that a castellum aquae might be situated at the present location, which was presumably supplied by an aqueduct, originating at the Gilboa‘ Mountain springs. A probe trench (Fig. 1, Area 2) was excavated to investigate this hypothesis.   

In the northwest end of the probe trench (length 35 m, width 0.6 m, depth 2 m), adjacent to the IAA excavation area, a brick-built water channel was revealed in the sections. The channel (width at base 0.3 m, max. width of channel walls 0.4 m, total width 1.1 m; Fig. 4), oriented southwest–northeast, was exposed for a length of 1.5 m; it was cut in its southwestern end. The channel walls were constructed from alternate light brown (3.5 cm thick) and olive green (5.5 cm thick) sun-dried bricks. The interior sides of the channel walls were originally plastered; however, only a negative impression of this plaster was preserved on the travertine buildup (thickness at the base 1 cm to 3.5 cm at the center). The floor was composed of well-fired bricks (thickness 2.5–3.0 cm, width 52 cm). The foundation below the floor was built of successive layers of light colored clay, which overlaid a layer of well-fired bricks and an additional two layers of clay bricks. The clay bricks of the upper layer were fired an orange hue, whereas the dried clay bricks of the lower layer were olive-green. The proximity of the channel to the modern surface resulted in its partial destruction, primarily the scraping of the upper courses. Thus, it was preserved to a maximum of 0.32 m high from its floor to the modern surface and its original depth could not be reconstructed (Fig. 5).


The channel was found filled with muddy sediment, stone cobbles, shells, potsherds and glass fragments. The fill below the channel contained potsherds and flint chips dating to the Chalcolithic/Early Bronze Age I; finds from these periods were discerned in most probes carried out at Bet She’an. The potsherds in and around the channel were dated to the Roman and Byzantine periods, although the brick construction of the channel floor was more characteristic of the Byzantine period at Bet She’an.


A single square was opened in the excavation of Area 3 (Fig. 1), revealing the continuation of the water channel discovered by the Hebrew University team. The curving water channel traversed the square from the southwest corner toward the northeast (interior width 0.3 m; exterior width, including walls 1.1–1.2 m; Fig. 6). The upper part of the channel was destroyed, precluding the estimation of its original depth. Nevertheless, its eastern cross section indicated that at least part of the channel was 0.45 m high. The channel’s walls were built of mud bricks (max. width 0.4 m, thickness 0.35–0.55 m), with travertine accumulating on their interior side (thickness 1–3 cm; Fig. 7).

The floor of the channel consisted of dark brown fired bricks with large grits (thickness 3 cm, width 52 cm). The floor was set on multi-layered bedding. The first below the floor was an unbaked, light brick mortar layer (thickness 7 cm); the second was a layer of baked mud bricks. The third layer was again unbaked brick mortar, which overlaid a fourth layer of baked mud bricks and a fifth and ultimate layer of unbaked mud bricks.

The sediments into which the channel was cut contained many potsherds from the Late Roman period, indicating the channel could not be earlier than this period. Judging by its construction method, it seems likely that the channel was from the Byzantine period.