The Eastern Rectangle (51 squares)
In Area E, the buildings excavated by Reich in 1986 (1989) were re-exposed in order to align the new excavations with the architectural remains exposed in the previous excavations. Here, a layer of sea sand and vegetation was excavated on the surface (0.2–0.7 m thick in the east, 0.1–0.2 m in the west). Below it, a washed layer of mud-brick debris from the disintegrated buildings and the fortification wall (ranging from a few centimeters to 0.6 m thick in the south and north) overlay the building’s walls and foundation courses. The mud-brick and plaster pieces that accumulated inside the buildings reflect the gradual collapse of the buildings.
The main architectural remains were uncovered in Areas C, E and K; additional remains were found between Areas C and E, as well as in Area B, an area partially dug by Naveh (Naveh 1962). Areas G and L were not excavated.
 
Area B. West of Room C, an accumulated sand layer was removed, and the wall tops of a row of four units were revealed abutting the northern fortification wall; these units had previously been uncovered in Naveh’s excavations (1962). Naveh also excavated part of the westernmost unit. The plan of this complex of units resembles the complex uncovered in Area E (see below).
 
Area C. This area of the fortress was damaged the most by the mechanical works that preceded the excavation, the main damage being to the fortification wall north and east of the small room uncovered in Naveh’s excavations (Room C; 3 × 3 m; Naveh 1962). After clearing the damaged mud-brick debris, the entire outline of the room was visible, revealing that about one fifth of the room, adjacent to the fortification wall, had not been excavated in the past. Here, a mixed accumulation layer of sand and mud-brick debris overlay the leveled bedrock. Within this layer, smashed pottery vessels, including local storage jars, a jug and a cooking pot, as well as sherds of a variety of imported vessels, were retrieved (Fig. 3). Parts of walls of additional rooms were found east of Room C. To the north of Room C, part of the northern fortification wall (length c. 25 m, width 4.2 m) was exposed below the mud-brick debris layer, extending eastward to the northeastern corner of the fortress. The tops of the stones of the wall’s foundation were exposed, presenting well-dressed kurkar stones set carefully the width of the foundation. The reddish-brown mud-bricks (c. 0.30 × 0.35 m) set on the foundation layer were mostly only preserved for a single course. The bricks in the northernmost row in the wall were set as stretchers running east to west, while the other bricks were lain as headers running south to north. It was evident that the bricklayers aimed to achieve uniform continuous rows along the wall. Some of the bricks had changed color and some were even vitrified, showing signs of a major conflagration; samples from this area were submitted for an analysis of their material and for radiocarbon dating. A section cut in the mud-brick collapse next to the northeastern corner of the fortress exposed the foundations of the outer northern face of the wall that consisted of small kurkar stones set on a rock-cut terrace; two dressed kurkar stone courses were extant on the terrace, overlain by two courses of mud-bricks (Fig. 4). Beneath the mud-brick collapse, the line of a gray plaster layer was observed sloping down to the north from the foundations of the northern fortification wall; a similar plaster line was visible outside the southern and eastern walls. This plaster layer was possibly lain to prevent vegetation from growing between the stones, or to improve the surface runoff drainage from the wall. Beneath the washed mud-brick debris accumulation, a light gray plaster floor was found abutting Room C to the west. This floor was laid over a sand fill layer that leveled out the natural slope of the rock. A fragment of an early Corinthian aryballos was found on the floor.
In the area between Areas C and E, a room (5.0 × 5.5 m) with a central partition wall built of gray mud-bricks, was exposed abutting the eastern fortification wall. In the southeastern corner of the room, sherds of three Greek cooking pots that were mended, an olpe and a rim of a small local bowl, were found beneath a heap of fallen mud-bricks and plaster. A well-preserved portion of the eastern fortification wall (width 3.2 m) was exposed between the northeastern corner of the fortress and Area E.
 
Area E. In 1986, Reich uncovered the eastern residential unit (6.5 × 11.0 m), which abutted the eastern fortification wall; it comprised a north–south-oriented rectangular space (3.0 × 6.5 m; courtyard?), two square rooms to its west and an additional, east–west-oriented room to the south (Reich 1989). The excavation in the latter room, which was only partially exposed by Reich, was completed. Two units, similar in size and internal subdivision to the eastern unit, were unearthed to the west. In the southern room of the middle unit, a floor of mud-brick debris set on the bedrock, was exposed below a layer of collapsed mud-bricks and plaster fragments. Sherds of various vessels were found on the floor, including an amphora from Samos and a local storage jar. Several restorable sherds of an Ionian cup were found. Adjacent to the northern wall of the room, remains of a square mud-brick installation abutting the wall, possibly a hearth, were exposed; many pieces of coal were found in the installation, and signs of burning were visible on the bricks (Fig. 5). An alleyway separated Area E from another unit (Area L), of which only the western wall had been uncovered in the past (Naveh 1962).
 
Area F. In the southeastern, highest point of the fortress, there was a large stone heap (c. 6 × 12 m), partly excavated by Naveh (Naveh 1962) and partly by Reich, on its northeastern side. A comparison with the archive photos of Area F from Naveh’s excavation, prior to the current excavation, showed that Naveh’s trench had collapsed and was filled with stratified sand. The removal of the fallen stones from Naveh’s trench, and the dismantling of other collapsed stones in the center of the stone heap, revealed the outline of a rectangular structure. It seems that the structure consisted partly of foundation stones that had been robbed from the area of the southern and eastern fortification walls, and from the foundations of some of the structures that abutted the southern wall (see below Area K; Fig. 6). The nature of the structure is not sufficiently clear. Few ceramic finds, as well as bullets found between the collapsed stones, may indicate that the building was used as a military post in the late Ottoman or the British Mandate period.
 
Area K. An alleyway (c. 2 m wide) runs between Areas E and K. Area K was partially disturbed, possibly in the Ottoman period when the foundations of some of the buildings and of part of the southern wall may have been robbed to construct the building in Area F (see above). The excavation uncovered a robber trench, and parts of the wall foundations in the southwestern part of the area, permitting the reliable reconstruction of the westernmost unit, whose size and interior subdivision resembles the units uncovered in Areas B, C and E (see above). East of this unit, the tops of mud-brick walls were discovered beneath the mud-brick debris, probably walls of three or four small rooms, presumably of two more units located here (Fig. 7). These rooms were better preserved than those in the western unit. One of the rooms had been partially excavated by Naveh. The eastern room, abutting the fortress wall, was found sealed with a pile of fallen bricks.
 
The Western Rectangle (11 squares)
The excavations focused on the area of the fortress gate (Area A), a room abutting the southern wall to the west of Area G, and the northwestern corner of the fortress (Area M). A sand accumulation was removed from the wall stumps in the northeastern corner of the fortress (Area D).
 
Area A. Four squares opened on the northern side of the fortress gate revealed that the northern tower was not preserved, apart from a few in situ foundation stones extant in rock cuttings intended to anchor the foundations. Two of the rooms in the southern part of the gate complex excavated by Naveh (1962), were also unearthed: Room 16 and Room/Building 18 (Fig. 2). Naveh uncovered the external walls outline of these rooms, but he did not excavate the rooms.
In Room 16 (2.6 × 3.4 m), a sand accumulation layer, overlying a heap of mud-brick debris mixed with plaster, was removed. In Room 18, the sand accumulation layer overlay a thick mud-brick debris layer (max. 1.0 m) that had collapsed from the walls. Beneath this layer, a long room (3.4 × 7.0 m), with an entrance between two pilasters protruding out from the end of the long walls was found. A hard limestone slab served as a threshold. About half of the northern area of the room was paved with Glycymeris shells with their convex side face-up. The shells were embedded in a thin layer of grayish plaster set on a sand fill layer (0.1 m thick). A small area (0.50 × 0.95 m) in the middle of the floor was not paved with shells, revealing the nature of floor’s foundation. A few sherds discovered on the shell floor included storage, cooking, serving and drinking vessels, some local, others imported. South of the southern wall of Room 18, a sand accumulation layer containing a particularly large quantity of imported vessels (L15) was excavated. These join similar vessels found in Naveh’s excavation in this area (Naveh 1962).
 
Area D. The excavation in this area focused on removing a sand accumulation layer above the wall foundations discovered by Naveh (1962).
 
Area G. A single square was excavated abutting the inner face of the southern fortification wall and Area G. The wall tops of a small room (2.0 × 2.8 m), built perpendicular to the fortification wall, presumably of another unit, were discovered. The collapsed mud-brick debris partially excavated around the borders of this room, produced an iron arrowhead similar to those found by Naveh in Area A (Naveh 1962).
 
Area M. About six squares were opened, revealing two adjacent rooms built abutting the western fortification wall. In the northern room (3.4 × 7.0 m) three mud-brick pilasters on an east–west axis, 1.0–1.2 m apart, were found (Fig. 8). A piece of hard limestone, whose eastern side was worn, was found on the eastern pilaster. The absence of a floor in this room might indicate that the pilasters were built on a constructive fill, possibly intended to support a staircase that ascended to the top of the fortification wall.
 
The Finds
The pottery finds from the current excavation are mostly similar to the assemblages of imported and local vessels previously published, dates predominantly to the end of the seventh century BCE (Naveh 1962; Reich 1989; Fantalkin 2001). Noteworthy is the large quantity of Ionian cups in the assemblage, as well as the Greek cooking pots, and a variety of imported amphora types, for example from Miletos and Samos. Some of the vessels bore incisions and potter’s marks. Many of the sherds retrieved from the excavation were friable, and their decorations were almost completely worn away. In Areas E and F, a few fragments of vessels from the Ottoman and British Mandate periods, albeit in negligible quantities, were retrieved. Metal artifacts, mainly made of bronze, were also found, including a spatula, a needle, a fishing weight made of lead, and nails.
Two coins of Antiochus III (242–187 BCE) were retrieved on the surface, joining an identical coin found at or near the site on a visit by A. Kloner in the 1980s (Reich 1989). The stone tool assemblage includes a few fragments of a variety of grinding stones and a small stone weight; some of the stone vessels are not characteristic types in Israel. Pumice stones, light yellowish in color, were found in various areas of the fortress, as well as a very small quantity of flint items, found mainly on the surface. Noteworthy is the complete absence of animal bones at the site, despite the complete sieving of most of the sediment unearthed.
 
Conclusion
The excavation in the upper part of the fortress largely filled in the plan in this area. The excavation in Area K, adjacent to the southern and eastern fortification wall, exposed the extent of the stone robbery that removed entire portions of the foundations of the walls and of some of the structures. It was important to determine which areas of the fortress had been disturbed. The two additional units exposed in Area E, help define the standard characteristic plan of the residential units in the fortress, which can be defined as a three-room house; this plan was also observed in Areas B and K. The remains in Area C, differ in plan and construction from those in Areas B and E, indicating that they may have served a different purpose. The division of space here may have post-dated the original construction; however, this theory requires further scrutiny.
The shell floor discovered in Room 18 in Area A, like the remains of a similar floor in Courtyard 17 in that area (Naveh 1962:94), is a unique feature among the architectural remains from the late seventh century BCE in Israel. A shell flooring was identified in Phase IV of the Tel Kudadi fortress, which according to Fantalkin and Tal (2009:196–201) should be dated between the late eighth century and the first half of the seventh century BCE. In general, shell decorations are known in the southern Levant in the Middle Bronze Age at Tell el-Ajjul, where a shell-covered bench was found (Petrie 1931:6, Pl. XII). As far as is known, shell floors were found at two sites dated to the thirteenth century BCE: at Tel Megiddo, near the Stratum VIII palace (Loud 1948: Figs. 50, 52, Area AA, Stratum VIII–VIIB) and at Tell Kazel on the southern Syrian coast, where a structure with shell floors was found in the southeastern part of the site, as well as shells covering the walls (Badre 2006:80, Area II). Shell floors resembling the floor in Mezad Hashavyahu have been discovered at sites identified as Phoenician colonies dating to the eighth–sixth centuries BCE, in the south Iberian Peninsula and in Portugal (Escancena and Vásquez 2009Ein-Mor, Mermelstein and Korbin, forthcoming).