Stratum IX. The excavation reached bedrock in two small areas (L359, L360), at a depth of c. 2 m below surface, after removal of top soil. The outline of a circular, bedrock-hewn pit (L360) was traced, but the pit was not excavated and therefore, could not be dated. The potsherds on bedrock were mainly from the Hellenistic period, although several could be dated to Iron IIB.


Strata VI, V. A packed-earth floor (L350, L351), exposed in the two squares on either side of the balk, 0.3 m above bedrock, was probably associated with a stone wall (W352) and with the lower courses of two additional stone walls (W338b, W326b). No coherent plan could be obtained as the early remains were overlain by the later walls and it is clear that the later Roman strata reused the earlier Hellenistic walls, with changes. A thick burnt layer with much ash and a considerable quantity of store jar and cooking pot fragments from the Early Hellenistic period overlaid the floors, as well as many animal bones, showing signs of butchering. These finds probably indicate that the Early Hellenistic house suffered a violent destruction. An additional packed earth floor (L349) superposed Floor L351 and was also strewn with a large quantity of Hellenistic potsherds. The dating of the Hellenistic stratum’s destruction will be determined after the pottery analysis.


Strata IV, III. Several phases may be observed within these strata. A single large building with several rooms was constructed in this period. A major long wall (W326a/W329; Fig. 2), running the whole length of the excavation (10 m), was exposed. Several walls were built perpendicular to it (W337, W338, W362), as well as a covered water channel (L348), built of dressed upstanding nari blocks, alongside and at one point crossing W328a. An additional channel (L357), built in an identical fashion, ran along the western side of W328/W328a and joined up with Channel L348, leading water in an easterly direction, beyond the excavation limits. An impressive, almost completely preserved miqwe (ritual bath) was attached to this building. This large miqwe (length 4.3 m, width 2.3 m, preserved max. height 2.2 m; Fig. 3), partially rock hewn and partially built of stone, had six broad steps and was completely coated with a thick layer of plaster. The extant walls indicated that the miqwe had originally a vaulted ceiling, which had caved in. Despite its large size, entrance into the miqwe from inside the house suggests that it belonged to a family, possibly an extended one. Alterations to the outer walls of the miqwe were noted and at some stage during the Middle Roman period, the miqwe was no longer in use and intentionally filled with considerable quantities of stones that partly came from the collapsed ceiling. Two superimposed floors that abutted the same walls were found in most of the excavated rooms, indicating at least two phases of use in the house. Unfortunately, the limits of the excavation did not permit the recovery of the complete plan of the house, which was abandoned at some stage. A gap in occupation in this area, probably in the fourth century CE, is assumed on the basis of absent pottery.


Stratum I. Subsequent to the period of abandonment, occupation at the site was renewed to a limited extent. The remains from this stratum comprised a segment of a single wall (W322) and a packed-earth floor (L321), lying about 0.2 m below surface. The finds consisted of animal bones and potsherds from the fifth century CE.