The Excavation
A probe was opened in the nearly square western half of the eastern chamber (c. 3.0 × 3.5 m), following the removal of the modern concrete floor and a modern fill (thickness c. 1 m) in both halves of the eastern chamber (Fig. 2). The excavation reached down to the water table, about 3.5 m below the modern floor.
 
The Crusader period (thirteenth century CE)
Phase 1. The remains of the earlier phase exposed in the probe comprise two–three courses of a well-constructed,east–west wall, built of hewn, bossed ashlars, typical of Frankish architecture (W104; Fig. 3). The wall was set on what seems as a debesh foundation (W165; Fig. 3); only a small section of this foundation was exposed. The wall extends to the east (W104 East) and west (W104 West) of the excavated probe. East of the probe,the wall was preserved to a height of 20 courses and continues at least up to the eastern end of the eastern chamber (8.3 m long, c. 7 m high; Figs. 2, 4 [left], 5); a slit in the wall blocked by modern concrete was cleared during the excavationand identified as an arrow slit (Shotten-Hallel, below; Fig. 2). An opening (width at least 2.8 m; Figs. 3–5) with a clearly visible threshold built of flat stones (W162) was discernable in W104; it probably served with an arched entranceway (gate? Shotten-Hallel, below).
A series of dirt layers (total thickness c. 1.5 m) covered the debesh foundation and abutted W104 almost up to the threshold. The two bottom layers (L164, L163) contained numerous pottery sherds dating from the Crusader period (thirteenth century CE) and a few earlier body sherds, apparently from the Roman and Byzantine periods. The top layer (L161) comprised dry, gray dirt with only a few ceramic finds, all dating from the thirteenth century CE, probably an occupation level.
Phase 2. At a later stage during the Crusader period, parts of W104 or a nearby, similarly built wall, collapsed, as evidenced by collapsed bossed ashlars, some bearing plaster (L159; Fig. 4) found in the northern part of the probe. To its south, in a strip (width 0.5–1.0 m) near the threshold, was a second gray fill of dry dirt that contained only few thirteenth-century CE pottery sherds (L60). It was most probable during this phase that the opening in W104 was blocked by a wall (W148A), of which four courses have survived. These were built of typical Frankish-style bossed ashlars in secondary use with small stones in between them (Figs. 3–5). These stones may have been taken from the southern part of Collapse 159, which explains the strip of dirt closest to the wall. This wall may have closed off a partially ruined part of the building until the Ottoman period.
 
The Ottoman period (eighteenth–nineteenth century CE)
Phase 1. The earliest Ottoman-period activity is represented by a dark grey, packed and moist dirt fill (thickness 0.1–0.4 m), which sealed and leveled the remains of the Crusader period. It contained eighteenth-century CE pottery. Above this fill were the partial remains of a stone-paved floor.
Phase 2. A fill (thickness c. 0.3 m) containing eighteenth-century CE pottery carried the remains of a stone-paved floor (L150; Fig. 6), made up of well-hewn, very regular limestone pavers. Two stone courses above W148A (W148B; Figs. 4, 5) are ascribed to this phase. Floor 150 came up to the base of W148B.
Phase 3. A wall (W147; Figs. 4, 6) exposed along the eastern section of the probe seems to have been built in this phase. Its remains comprise the two first courses, built of well-hewn rectangular stones, set on a foundation (depth over 3 m) of irregular stones and hard, light gray mortar that extends down into the Crusader-period strata. It includes a constructive arch, a common feature in Ottoman-period architecture (cf. Stern 2010: Fig. 2). A floor (L152; Fig. 6) of well-hewn flagstones abutted W147. It was set on a fill (thickness c. 0.8 m) containing nineteenth-century CE pottery.
 
Architectural Survey
Vardit Shotten-Hallel
 
The architectural survey focused on in-situ elements. Three elements are from the Crusader period (twelfth–thirteenth centuries CE): the springers of an arched entrance set in W104, possibly a gate; the fortified wall to its east (W104 East); and an arrow slit in W104 East. A transverse arch extending between two groin vaults along the middle of the southern chamber was ascribed to a post-Crusader phase.
 
The Crusader Period
Arched entrance (Figs. 1:A; 5). The springing line of a large, arched opening (span of 2.94 m) was identified on both sides of the now-blocked opening in W104. This was most probably a pointed arch, built of bossed voussoirs, as evident by a single springer preserved on its western side (Fig. 7).
 
Fortified wall (W104; Figs. 1:B; 5). The well-preserved sections of the wall extending to the east and west of the arched entrance were built of bossed ashlars. The preserved ashlars are rectangular blocks (0.33 × 0.40–0.70 m; Figs. 8, 9), with a projecting boss chamfered diagonally on all edges, creating an enhanced light-and-shadow effect on the wall. This type of boss is typical to the construction in ‘Akko, and can be seen also in other monuments, e.g. on the western façade of Burj al-Sultan in Khan al-Shawarda and in the Hospitaller compound, in the passage between the latrine hall and the northern halls.
 
Arrow slit (Figs. 1:C; 5, 10). The arrow slit in W104 East is located 2.7 m east of the entrance. Unfortunately, its upper part was not preserved (preserved height 1.7 m), and the stone courses above it seem to belong to a later phase. It was probably one of the five arrow slits documented in Kesten’s (1993) plan of Crusader-period Acre.
 
The wide and massive wall (W104) built of bossed ashlars and featuring an arrow slit belonged to a fortified structure that is distinctly Frankish in architecture. Wall 104 can be identified as the northern wall in Structure 52, which appears on two reconstructed maps of the Genoese quarter during the Crusader period (1104–1291; Benvenisti 1970:99; Jacoby 1979). This structure and a similar structure to its west (No. 44) were interpreted as sections of a covered street (Kedar and Stern 1995:109, Plan 2). The direction of the arrow slit implies that the area to the north of W104 was an open space, and therefore the entrance in the wall led southward, into the fortified structure. Excavations conducted in an adjacent square, to the northeast of this structure, yielded a few archaeological remains from the Crusader period. The most significant of these were the remains of a Crusader-period tunnel (Tatcher 1998:14), indicating that there was an open space above it.
An important feature of the urban layout of Crusader-period Acre was the physical division within the city between the various maritime communes and the military orders. Thus, in addition to the separated compounds of the Hospitallers and the Templars, the maritime communes from Genoa, Pisa and Venice each had its own quarter. The fortified structure to which W104 belonged could certainly have been the boundary of the Genoese quarter in the thirteenth century CE. In view of the architectural language of the remains in the excavated area, it seems plausible that the fortified wall served as an outer defensive wall of a well-protected building or area, which also marked the boundary between the Hospitaller area and the Genoese quarter in the thirteenth century CE.
This understanding of the excavated wall nevertheless leaves us with several unanswered questions: What was the street level when the gate was in use? Was the gate elevated from the street and accessed by a ramp? What was the structure’s layout to the south of the excavated square?
 
Post-Crusader Period
Transverse arch (Figs. 1:D; 11; 12). A fully preserved transverse arch separates the two groin vaults in the southern chamber. The arch is built of hewn blocks that project 3–4 cm from the face of the vault. The stones were placed in a zipper-like manner, with ‘tails’ (c. 8 cm each; Fig. 12) placed into alternating directions. The arch springs from two pilasters that are integrated in walls that date from the Crusader period but carry vaults that are later in date.
Although these vaults have been dated to the Ottoman period (Stern et al. 2011), suggesting a similar date for the transverse arch, there is evidence suggesting an earlier date. No similar arches have been found in any of the Ottoman-period buildings in ‘Akko. On the other hand, the arches on the ground floor of Burj al-Sultan (the Sultan’s Tower) in the southwest corner of Khan el-Shawardah in ‘Akko exhibit several similar details. An inscription above the entrance to the Khan, dedicated to al-Malik al-Ashraf Sayf ad-Din Barsbay, the ninth Burji Mamluk sultan of Egypt, dates it to H 840/1436–7 CE (Sharon 1997:31–33). True, in the absence of a detailed survey of the tower, no construction from this period can be pointed out, although Frankish masonry has been identified both in the outer western face of the tower and in several places on the interior western face of the tower. However, major changes seem to have taken place inside the tower following the destruction of its original vaulting, probably after the fall of Acre in 1291, although later modifications should not be precluded. It thus seems plausible to date both the transverse arches of Burj al-Sultan and the one surveyed in the southern chamber to the fifteenth century CE, and not to the late Ottoman period, as previously suggested (Stern et al. 2011).
 
Conclusions
By integrating the stratigraphic data from the excavation with the results of the architectural survey, it seems that W104 was part of a previously unidentified Crusader-period structure. The nature of the structure is still uncertain, although the high-quality masonry and the scale of construction, the arrow slit and the wide entrance, all indicate that this was the outer, northern wall of a fortified structure, which possibly marked a boundary within the city. Thus, the area to the north of the wall was most probably a street or an open space, such as a plaza. Since no clear surface was identified at the level of the threshold (W162), it is possible that the entrance leading southward was approached via a wooden bridge or ramp. This is also suggested by the well-built lower courses of the wall, featuring bossed ashlar stones. Another possibility is that the original floor, at the level of the threshold, was paved with flagstones, which were removed when the entrance was blocked and possibly reused elsewhere in the city. The collapsed stones (L159) and the blocked entrance may be associated with documented historical events, such as the St. Sabas war that occurred in the mid-thirteenth century CE (1256–1258). Since ‘Akko continued to serve as an urban center until its destruction in 1291, the collapse and the blocking may have occurred later in the thirteenth century CE. Although this was the second excavation in the building, only further excavations will provide a more accurate picture of its nature and history in the Crusader period.
The date of the transverse arch in the hall south of the excavated area cannot be agreed upon. Whereas the results of the two excavations within the building suggest that it was built as part of the Ottoman-period building, the results of the survey raise the possibility that it is of an earlier date. Thus, it is impossible at present to dated it with precision.
During the Ottoman period, major changes occurred in the building, resulting in three construction phases. Although a wall (W148) separates the two excavation areas in the building, these three construction phases were identified on both sides of the wall (Stern et al. 2011). The construction phases are represented by a series of fills and well-built floors that correspond with the building of walls and their repair; it is assumed that these construction phases were also accompanied by construction and repair work in the vaults.