Four excavation areas were opened (Fig 1): Area A comprised a row of twelve squares along the northwestern section of Roslan Street; Area B encircled the water fountain, wrongly known as Sabil Abu Nabbut, near the Mahmoudiyye Mosque; Area C comprised seven squares, running along Mifraz Shelomo Street, between the Jaffa Museum courtyard and Ha-Zorfim Street; and Area D comprised 11 squares along the southeastern section of Roslan Street. The Excavation reached a depth of c. 3 m. Four layers and several building phases were identified: horizons of Iron Age and Hellenistic pottery; Early Islamic and medieval architectural remains, mostly from the Crusader period; and extensive remains of the Late Ottoman city wall and gate complex, as well as a fountain, a street and dwellings, some of which destroyed as late as 1936.
 
The Iron Age through the Hellenistic period
In Area B, a concentration of exclusively Iron Age pottery was found within a dense layer of clay and sand (thickness 0.1 m), in which ash lenses were discerned. The layer was deposited over the bedrock. The assemblage consisted of Iron Age II–III storage jars and cooking pots. Fragmentary remains of a wall were found at the upper part of this layer. In Areas C and D, pottery dating from the late Iron Age through the fifth or forth centuries BCE was associated with a dense clay deposit, similar to that in Area B, but lacking the ash lenses. In Area D, Hellenistic pottery dominated the deeper layers, although no affiliated architecture was found. While the frequency of Iron Age potsherds increased as the excavation deepened, no occupation layer from this period was discerned before work was discontinued due to safety considerations. Although scanty, these ceramic finds are important, as there are relatively few known Iron Age contexts in Jaffa, and in the lower city were found only in several isolated locations (the Flea Market – HA-ESI 120; Ganor compound – HA-ESI 121).
 
The Early Islamic and Crusader periods
Area A. Part of an installation (depth 0.8 m) with a semispherical sump (diam. 0.6 m, depth 0.2 m; Fig. 2) that resembles a winepress collection vat was discovered. The wall, floor and pit were coated with plaster solidified with potsherds. No associated floors were found. According to its stratigraphic position under the floors and foundations of an Ottoman building, and the potsherds found in unsealed related soil accumulations, it can be ascribed only a general medieval date. Wall foundations unearthed approximately three meters to its south were in a similar stratigraphic position, with Crusader-period pottery in related loci.
 
Area B. Fragmentary walls and floors of either the Early Islamic or Crusader periods were found in the northern part of the area. To their south were the remains of several walls, two forming a straight-angled corner, and packed-earth floors dating from the twelfth–thirteenth centuries CE. Most of the walls were built of fieldstones, and laid on a southeast–northwest trajectory, similar to architectural complexes of the same period in various other sites in Jaffa. Finds include various imported glazed bowls, a glazed oil lamp, brown glazed cooking vessels and plain jars and jugs. Glass sherds and coins of probable medieval date were also collected.
 
Area C. The narrow dimensions of the excavation trench and safety considerations hampered the investigation of the relatively deep pre-Ottoman remains in this area. Nevertheless, fragmentary architecture was exposed under Ottoman-period remains. The walls were built of kurkar or nari fieldstones with some compact, lime-based mortar. Their dating is uncertain due to lack of clearly associated sealed contexts. Most of the pottery found in an affiliated fill, comprising a few sherds and a stamped handle, dated from the ninth through the eleventh centuries CE; some glazed sherds of Crusader origins were found as well. With no clear stratigraphic indications, it seems likely that these buildings were erected in the later part of the Early Islamic period and continued to be in use, probably after some adaptations were made, during the Crusader occupation of Jaffa.
 
Area D. Crusader layers were reached in almost all squares. A straight-angled corner, comprising two substantial walls running east–west and north–south and built of large ashlars was discovered in a deep probe (Fig. 3). They belonged, in all likelihood, to a public or defense structure; standard domestic units of this period have been found in various medieval strata in Jaffa and comprise markedly humbler architecture. No floors were discerned, but finds from the abutting sediment belonged to the Crusader period, and included typical imported glazed wares, as well as arrowheads and a well-preserved dagger blade, metallurgically proven to be pre-modern.
While parts of Jaffa's thirteenth-century fortification walls have been discovered in the Flea Market excavations (HA-ESI 118; HA-ESI 120), the twelfth-century city walls, attested to in numerous historical sources, remain evasive. The only potentially associated remains were found in the intersection of ‘Ami‘ad and Ha-Zorfim Streets (HA-ESI 122). These resemble in style, appearance, materials and orientation the impressive walls in Area D, although no decisive linkage between the two areas could be established.
 
The Ottoman period
Area A. A segment of the late-eighteenth century fortification wall (length 5 m; Fig. 4), including part of a gate, was discovered at the northern end of the Area. Only the western jamb and 1.6 m of the gate's breadth could be exposed. The gate's entrance was paved with flagstones. The wall, surviving 12 courses high (height 2.9 m, width 1.4 m), comprised a core of fieldstones mixed with packed soil, and façades built of kurkar ashlars. As is standard in large Ottoman structures in Jaffa, the wall was supported by subterranean foundation arches (Fig. 5). A longer section of the wall, ending at a hemispheric tower, was discovered within the Qishle compound (HA-ESI 121: Stratum II). The wall, tower and gate can be tentatively identified in the 1799 Jacotin map of Jaffa.
A well-preserved segment of a stone-paved street (length 28 m, width 4 m; Fig. 6) was discovered along the central part of Area A. The street paving consisted mostly of dressed limestone slabs; kurkar blocks were used to flank the paving. The street was laid on a bedding of sand (thickness 0.10–0.15 m), which covered a packed layer of clay and plaster. The northernmost eight meters of the street veered slightly westward, and connected with the eighteenth-century gate; it ends at slightly a higher level than that of the gate floor. Ottoman Gaza sherds and Marseilles tiles uncovered from the street's foundation bed date it to the second half of the nineteenth century CE, when stone-paved streets were first introduced to Ottoman Jaffa. Similar streets were recently discovered at the Ha-Zorfim Street complex (HA-ESI 122) and along Yehuda Ha-Yammit Street (HA-ESI 123; HA-ESI 124).
A segment of a plastered drain channel (length 5.5 m), covered with stone slabs, was discovered north of the wall and gate. It was probably part of Jaffa's Late Ottoman drain system, which transported waste water into the sea. Like the paved streets, Jaffa's public drain systems date from the later decades of the nineteenth century CE.
The foundations of the western perimeter wall of the late-nineteenth century Qishle were exposed nearby. The corner of a room with two superimposed tile floors, was unearthed near the southern face of the eighteenth-century wall. An additional wall, of probable early-nineteenth century date, was found outside the fortification. Several other fragmentary structures, belonging to at least two Late Ottoman phases, were uncovered in this area. Two Late Ottoman building phases were identified at the southern end of Area A. The earlier phase consisted of the northwestern corner of a structure with sturdy, thick walls (width 0.9 m) and a related plaster floor. The later phase included a storage installation associated with the paved street.
 
Area B. The remains of an unusual storage or industrial installation, a channel, and several walls dating from the nineteenth century CE were discovered west of Sabil Abu Nabbut. The installation consisted of an uneven oblong platform built of fieldstones (1.0×4.4 m; Fig. 7), into which five large ceramic vessels were inserted. The vessels were set on flat, circular frames, and were secured into place with stones and hard plaster. Inverted marble pillar bases were used for two of the frames, and the remaining three were made of beach-rock slabs. The function of this installation could not be determined.
 
Area C. An octagonal fountain, situated at the heart of an open, paved courtyard (Fig. 8), was discovered in the center of the area. The fountain was built of stones and compact plaster, over a foundation that was constructed of cement-plaster and stones. A layer of plaster coated the fountain's basin and walls, and remains of paint could be discerned on the outer walls. Water sprouted from a spout set in a small polygon-shaped marble post, which was vertically set at the center of the basin, most probably in secondary use. Water drained through a perforation in the basin’s floor. The courtyard was paved with stone slabs, with crossing central strips of marble tiles. Benches were constructed along the walls on its eastern and western perimeters. The western bench was dismantled during the excavation, revealing a drain made of flat ceramic tiles and leading northward, toward the sea.
Flanking the fountain courtyard on its west and east were the remains of three Late Ottoman buildings, which were demolished in 1936 (see below). Of the western building, only the southern corner and limited parts of the southern and eastern walls with related plaster floors could be exposed. The walls were sturdily built of kurkar, cemented with compact plaster and set on substantial foundations.
A structure comprising two adjacent rooms, built on different levels with no separating wall, was discovered east of the fountain (Fig. 9). The southern, lower level, possibly the front room (divan) of the dwelling, had doorways in its southern and western walls; the western entrance, apparently a secondary adjustment, opened onto the fountain courtyard. The room’s walls were plastered and the floor was tiled with limestone slabs. The northern space was an elevated platform paved with tiles (0.2×0.2 m). At its center, the tiles formed a colorful geometric and floral pattern (Fig. 10), in a style typical of the late nineteenth and especially early twentieth centuries CE. The room was accessed from the divan by a square brick step that abutted the central part of the elevated platform. Elevated platforms were sometimes built in traditional houses and served for hosting.
Remains of an early phase of this building were identified. Two drain conduits leading north, toward the sea, were discovered under the eastern wall of the southern room. One was constructed of ceramic pipes, and the other was constructed of mortar and covered with stone slabs. The floor of the northern room lay on a foundation comprising plain and colored tiles, stone slabs and plaster – probably the original floor, which was destroyed when a British Mandate pipe was laid. This pipe and the known date of the destruction of the house date the latest phase of the structure between 1920 and 1936.
A large water cistern or vaulted room comprised the remains of a second structure to the east of the courtyard. Only its southwestern corner was excavated. The installation’s walls were plastered on the inside and painted red, as was its plaster floor. The wall's trajectory is congruent with those of the other two buildings uncovered around the fountain courtyard. Associated finds included ceramic and glass sherds as well as metal scrap, all of Late Ottoman or British mandate origins.
Fragmentary remains of several Ottoman walls were discovered at the eastern extremity of Area C. The walls were built of stones solidified with packed clay or hard lime, methods that were frequently used in Jaffa’s Ottoman architecture. Traces of white plaster coating could be discerned on one of the walls. A small marble pillar base of classical origin was incorporated in the core of the westernmost wall. The walls could not be securely associated with one another or with other exposed architecture.
 
Area D. Elements belonging to Jaffa’s main gate, known as the Jerusalem Gate or Abu Nabbut’s Gate, were uncovered in the eastern part of the area. The gate is mentioned in nineteenth-century textual sources and appears on maps as well as in artistic depictions and photographs of this period. The sources describe a bridge that crossed a moat and led into an enclosed courtyard. The courtyard was circumscribed by the well-documented water fountain, originally named Sabil Mahmudi, and known today as Sabil Suleiman, on the north; by a high wall on the west; and by the gate structure – a tall building with attached turrets – on the south. Of these elements, only the gate entrance and the sabil were visible prior to the excavations.
The excavations revealed a segment of the wall that enclosed the gate courtyard to its west (thickness 1.4 m; Fig. 11). The course of the wall ran from the northwestern corner of the gate structure to the western end of Sabil Suleiman. It stood on foundation arches, and comprised a dense core of fieldstones, packed together with soil and hard plaster, with ashlars on both façades. Its identification is based on its dimensions, which preclude domestic use, and its location, which is congruent with the delineation of the western wall of the gate courtyard on the 1842 Skyring map, Jaffa’s most detailed cartographic document of the first half of the nineteenth century CE. This wall prevented direct access by potential attackers, especially cavalry charges, into the city by creating a 90° angle toward the south, where the gate entrance was located. It also exposed the yard to the full view of the defenders who were stationed above it.
The excavations revealed the channel that drained the sabil's basin. The channel was built of stones, coated with plaster and covered with slabs. It ran along 2.3 m southward from the center of the sabil, and following a 90° turn it ran c. 20 m eastward (Fig. 12), toward the bridge, and ended outside the city. At some stage, perhaps in the 1860s or 1870s, when the gate still stood but the previously vacant grounds east of it were already being settled, the waste water was diverted into the moat below the bridge.
Outside the gate, a large segment of the nineteenth-century bridge that crossed the moat was unearthed. This was a surprising find, as prior to the excavations it was generally assumed that the bridge was dismantled when the moat was filled in. The bridge consisted of two consecutive arches, stretching c. 9 m over the city moat, and was paved with dressed stones (Fig. 13). The structure survived fairly intact, although most of its paving stones were removed for secondary use, probably in the city's network of paved streets that was constructed during the Late Ottoman period.
Excavations under the bridge penetrated the fill which sealed the moat once the fortifications became obsolete. It was rich with Late Ottoman finds, such as pottery, glass, metal waste and coins, as well as porcelain and imitation porcelain sherds, some of which carry datable producer marks. Future analysis of the porcelain and numismatic finds should help provide the most reliable time frame to be ascertained so far for the dismantling of the gate complex, and the rest of Jaffa’s fortifications.
Ottoman remains found in the western part of Area D included a solid pilaster, from which sprang four vaults, and two pipelines made of Ottoman gray Gaza Ware fragments. The orientation of the pipes indicates that they served as the original drain system for the still-standing late-nineteenth century buildings along the southern side of Roslan Street.
 
The British Mandate
Area C. Excavations in this area unearthed ruins from 'Operation Anchor', carried out by the British Mandatory government in June 1936 in an effort to suppress the Arab Revolt rioting in the city. During the operation, dwellings in the dense urban core of Jaffa were demolished and their remains spread and packed down, to allow the opening of passage routes (see HA-ESI 123: Figs. 5, 6) for heavy military vehicles. These routes were later turned into broad, paved avenues, the easternmost wing becoming Mifraz Shelomo Street. The remains consisted of a massive layer (1.0–1.5 m; Fig. 14) of building stones, many of which were coated with colored plaster (mostly red, yellow and blue), covering the floors and foundations of the destroyed buildings. This debris completely sealed the Late Ottoman buildings, the water fountain and the surrounding courtyard (see Fig. 8). The impact of this destruction is reflected in the change that took place in the area's declivity: from a northeastern one, on which the Late Ottoman buildings were constructed, to the present-day eastern decline. The debris, uncovered immediately under the foundation bed of the modern asphalt road, included large quantities of ceramic sherds, broken glass, and metal refuse from the late nineteenth and early twentieth centuries CE.
 
Area D revealed no such scene of destruction and the Late Ottoman layer was found directly under the modern asphalt. Despite the dismantling of Jaffa’s fortifications and the fast urban development in the areas to the east and north of the historical city, the former gate compound continued to serve as a passageway between the old and new parts of the city, and no dense construction replaced the gate courtyard.
 
The excavations on Roslan and Mifraz Shelomo Streets have added important information about the development of Jaffa in the Late Ottoman period and during the British Mandate. They have enhanced our understanding of (1) the urban layout of the nineteenth-century city, including a hitherto unknown paved street and various architectural adaptations; (2) the eighteenth-century city wall, to which few historical sources refer, and its northern gate; and (3) the main city gate complex. They have also enabled the systematic excavation of the 'Operation Anchor' ruins.
Details were also added to our knowledge of Jaffa’s earlier periods. Although scattered and insufficient for constructing an overall plan, these finds, alongside those from other excavations, help delineate Jaffa’s urban sprawl from the Iron Age onward.