The excavations (c. 1 dunam; Figs. 2, 3) were conducted on the eastern bank of the Tyropoeon Valley, along the outer face of the Old City wall, c. 25 m east of Damascus Gate and c. 70 m west of the entrance to Zedekiah’s Cave. The excavation uncovered walls that predate the Middle Ages; sections of a fortification wall and an aqueduct dating from the Crusader–Ayyubid periods; foundations of the early Ottoman city wall; and a complex water system and remains of a building from the late Ottoman–British Mandate period. The remains were first uncovered with a backhoe to a maximum depth of 6 m beneath fills of soil and debris that drop steeply from Sultan Suleiman Street in the north to the Ottoman Old City wall in the south.
In the second half of the nineteenth century CE, British explorers uncovered an ancient water system running from east of the Damascus Gate, within the boundaries of the current excavation, to the Temple Mount (Fig. 4). The system consists of three parts: a northern channel, the Struthion Pool, and a southern channel, also known as the Hasmonean aqueduct. The northern channel was first documented in 1871 by Conrad Schick (Wilson 1872:47–51; Warren and Conder 1884:263–264) but had not been examined since then. Its presumed route passes in the vicinity of Zedekiah’s Cave, continuing from there along the northern side of the wall through the part that was re-examined in the current excavation (below, Channel 108), then beneath the Ottoman wall and through the northern part of the Muslim Quarter, as far as the Struthion Pool. The original description reads:
"… the aqueduct is from two and a half to three feet wide [76–91 cm], and so high that a man can walk through it easily, the height rising, occasionally, to twelve feet and more [3.60 m]; it is partly hewn out of the rock, and partly of masonry, the channel being covered by a vault in which numerous openings, now closed by rubbish, were noticed … In building the present city wall, which stands on old foundations, the upper portion of the aqueduct was destroyed" (Wilson 1872:47–48).
The total length of the channel was not given, but estimates place it at c. 400 m. The southern channel leads from the Struthion Pool to the Temple Mount via the northwestern corner of Herod’s wall. It was first discovered in 1864 by Charles Wilson, explored by Charles Warren (Wilson 1872:47–51) and documented anew in the 1980s by Dan Bahat (1994).
Since its discovery, the system has been widely discussed by scholars and estimates differ as to its date, its source and the stratigraphic link between the channels and the pool (for discussion and history of the research, see Weksler-Bdolah 2011a). Bahat dated the southern channel to the Hasmonean period, basing his dating on the fact that Herod’s wall was built on top of it, and its attribution to the Baris (Bahat 1994). Following Bahat, the southern channel is now generally called the Hasmonean Aqueduct. Other scholars, such as Wilson and Barkay (Wilson 1872; Barkay 2007), have suggested that the channels formed a single water-supply system dating from the First Temple period; furthermore, they have suggested to identify it with the conduit of the Upper Pool mentioned in the Bible in association with military campaigns and the siege of Jerusalem. Wilson based his dating on his attribution of the damage identified in the channel’s northern section to the quarrying of Zedekiah’s Cave during the Herodian era. Barkay argues that since no typical Hasmonean-period tool marks are visible in the southern channel, it cannot be dated to this period. In contrast, Weksler-Bdolah suggests that the two channels belong to one long water course (the Damascus Gate Aqueduct) that was used in different periods, the earliest of which is pre-Herodian. In her opinion, even if the channel was first hewn in the First Temple period, it may well have been incorporated in the Hasmonean aqueduct (Weksler-Bdolah 2011a:42).
Dozens of excavations conducted beside and beneath the Ottoman Old City wall have uncovered ancient fortification systems—including walls, towers and a moat—dating from the First and Second Temple periods, the Roman–Byzantine periods (fourth–fifth centuries CE), the end of the Early Islamic period (tenth–eleventh centuries CE) and the Crusader–Ayyubid periods (twelfth–thirteenth centuries CE; for history of the research, see Wightman 1993; Weksler-Bdolah 2011b; Geva 2015). From the 1930s to the present day, many excavations conducted in the Damascus Gate area have uncovered architectural remains and fortifications dating from the Second Temple until the Ottoman periods. The principal remains include a gate that most archaeologists date to the time of Aelia Capitolina (Hamilton 1944; Magen 2000); a church, a gate and a wall corresponding to the Ottoman wall that are dated to the Crusader period (twelfth century CE; Hennessy 1970; Wightman 1989) or the Ayyubid period (thirteenth century CE; Geva and Bahat 1998); a complex of three vaulted halls from the Ayyubid–Mamluk periods (Magen 2000); and two Ottoman buildings (Zelinger and Mashiach 2018:15–16).
Pre-Medieval Periods
Two fortification walls were uncovered in the southeast of the excavation area (W100A, W101A; Figs. 5, 6). They were built of large, dressed stones, some with drafted margins, and bonded with crushed white chalk, charcoal grits and a few crushed potsherds. The walls protrude outward from the line of the Ottoman wall that was built on top of them (W100B, W101B; below) and are architecturally different from it. The two walls are dated to the pre-medieval period, based on a water channel from the Crusader–Ayyubid periods (L108, below) that was added beside them.
Crusader–Ayyubid Periods (twelfth–thirteenth centuries CE)
Two segments of a fortification wall (W105—excavated length 6.8 m, width 3 m; W111—excavated width 1.5 m; Figs. 5, 6) were uncovered in the southeastern part of the area. Wall 111, which was only partially preserved, was perpendicular to and incorporated in W105. Wall 105 is massive and broad, and it continues southwestward, beyond the limits of the excavation area. Both walls were built in a similar manner, with heaps of dressed stones of various sizes bonded together with a hard, beige-colored mortar containing earth, lime and charcoal. The walls’ western faces were not preserved and may have been robbed, whereas their eastern faces (W106) border a water channel that follows an L-shaped course (L108). The wall is dated to the Crusader–Ayyubid periods (mid-twelfth–mid-thirteenth centuries CE) based on fragments of a glazed bowl produced in Beirut that was found while dismantling parts of W105 (Fig. 7). Also found while dismantling the wall were pottery shards from the Early Roman and Late Byzantine periods (not drawn), a Byzantine-Arab coin (640–670 CE; IAA 165698), a white marble fragment and a shell.
Channel 108 (length 8.85 m, width 0.8–1.1 m, depth c. 1 m) sloped from northeast to southwest before turning at a right angle, from northwest to southeast and into the Old City. The channel’s western wall (W106) was built of small and medium-sized dressed stones, some with typical Crusader diagonal drafting marks; it was preserved to a maximum height of four courses. The eastern wall of the channel (W107) ran alongside the ancient fortification wall (W101A) and was built of medium-sized ashlars, some also bearing diagonal drafting marks. The lowest course of W107 incorporated an architectural feature in secondary use—a dressed stone truncated on one of its sides (width 8 cm), typical of the Crusader period. In its southward continuation, the channel is delimited on the east by W101A (length 4.65 m), which was lined in this phase with small, flat dressed stones bonded with white mortar composed mostly of lime and charcoal with a few crushed potsherds. The channel’s floor was made of small and medium-sized dressed stones bonded with gray mortar containing lime and charcoal, and it was coated with light gray plaster. The channel floor was set on brown soil that yielded pottery dating from Iron Age II, the Hasmonean period, and the Early Roman period (second century BCE–first century CE; not drawn). The channel is dated to the Crusader–Ayyubid periods based on the diagonal drafting marks on the stones in W107 and on the dating of the fortification wall (W105) on which it rests.
Early Ottoman Period (fifteenth century CE)
The Ottoman wall (W100B, W101B, W103) was founded on top of ancient architectural remains. Walls 100B and 101B were built over the earlier walls (W100A, W101A), whereas W103 was built over Channel 108 from the Crusader–Ayyubid periods. A narrow opening (L108A; width 0.2 m, height 0.4 m; Figs. 3: Section 3–3; 8) set in W103 enabled the channel to enter the Old City. The current excavation uncovered the lower part of W103. This is the wall’s foundation, composed of four courses of large building stones with drafted margins; these courses differ from the stone courses above them.
Late Ottoman Period and British Mandate Era (nineteenth–twentieth centuries CE)
A complex water system uncovered in the northwestern part of the excavation area included a large reservoir (L117; Fig. 9), a settling pit that drained into it (L127), a cistern (L121), an installation (L124), a water channel (L116) and remains of a building (L120, L132). Based on the building style, the mortar and the plaster, the entire system was built at the same time. Both the mortar and the plaster of all the water system’s elements contained tar.
Reservoir 117 (4.1 × 8.1 m, excavated depth 4 m; Figs. 3: Section 4–4; 10) was covered with a barrel-vaulted roof that was only partially preserved; only the upper part of the reservoir was excavated. The reservoir’s walls and vault were built of small and medium-sized dressed stones bonded with mortar containing charcoal, lime and tar. The walls were coated with pink plaster containing lime and pottery grits. A channel (L146; width 0.26 m, depth 0.32 m) set in the upper part of the southern wall led southward, beyond the outer edge of the reservoir. The reservoir was delimited on the west by a wall (W118; length 8.7 m, width c. 0.8 m); the northern part of W118 joined the reservoir (W134). The wall was built with an outer, western face of large and medium-sized dressed building stones and an inner face of piles of stones; the stones were bonded together with mortar similar to that found in the reservoir. Wall 118 was preserved to a height of four courses. It dates from the first half of the nineteenth century CE, based on the pottery, porcelain (below) and glass (not drawn) discovered while dismantling it.
The floor of Settling Pit 127 (1.15 × 1.45 m, depth 1.25 m; Figs. 11, 12) sloped southward toward the reservoir. The settling pit was constructed like the reservoir, and its walls were coated with white plaster. A water channel with a quadrangle cross section (L138; width 0.22 m, depth 0.34 m) led to the northern wall of the pit. A water channel with a round cross section (L140; diam. 0.15 m; Fig. 12) set in the pit’s southern wall led to the reservoir.
Water cistern 121 (2.30 × 2.65 m, excavated depth 2 m; Fig. 13) had thick walls (up to 1 m wide) built of two rows of dressed medium-sized building stones with a core of small stones bonded with white mortar containing lime and charcoal. The walls were coated with pink plaster, some of which was smeared with tar, laid on a bedding of stone chips.
Installation 124 (1.1 × 3.2 m, preserved depth 0.34–0.38 m; Fig. 14) was built of small and medium-sized dressed stones bonded with mortar identical to that inside the cistern. The installation’s western side was not preserved. The installation was coated with pink plaster applied on a bedding of Marseille roof tiles that are dated to the first half of the nineteenth century CE. The floor of the installation sloped westward and was overlain by a layer of tar. The floor bedding yielded finds including glass, pottery and imported European porcelain (not drawn) dating from the first half of the nineteenth century CE at the earliest.
Water Channel 116 (excavated length 14 m, width 0.36 m, depth 0.30–0.34 m; Fig. 15) sloped from north to south and was adjacent to W118 and Cistern 121. Its sides were built of an inner face of medium-sized dressed stones and an outer face of small stones, bonded with mortar and coated with white plaster. The channel was covered with large stone slabs, some of which were preserved in situ.
The building remains were found to the east and west of the reservoir. Two walls on the eastern side (W114—length 3.9 m, width 0.3 m; W115 —length 2.5 m, width 0.33 m; Fig. 16) formed the corner of a building built of large, dressed stones and preserved to the height of a single course; the walls were abutted by a floor of square stone tiles (L120). The floor’s western edge was laid on top of a wall (W125) enclosing the reservoir on the east. The construction of W125 resembled that of W118. A stone-tiled floor (L132) similar to Floor 120 was found on the western side. The building remains and the flooring date from the second half of the nineteenth century CE at the earliest, based on finds from the bedding of Floor 120, which include pottery, Marseille roof tiles, European porcelain imports, glass vessels and metal (not drawn).
The Finds. The soil accumulations that covered the remains yielded abundant finds, mostly from the southeastern part of the excavation area (L104, L119, L147; the last two are not shown on the plan). The finds are dated to the late Ottoman period and the early British Mandate era (second half of the nineteenth–first half of the twentieth centuries CE) and include assorted fragments of pottery, glass, metal, etc.
The pottery consists of local Gaza wares, including a casserole (Fig. 17:1), a cheese container (burenye in Turkish; Fig. 17:2), an ibrik decorated with orange paint (Fig. 17:3) and a water jar known as a giarra (Fig. 17:4); imported glazed ware, such as a glazed French Vallauris cooking pot (Fig. 17:5) and Turkish green-glazed bowls (Fig. 17:6, 7); stoneware, including a German water bottle (Fig. 17:8); a frit-ware coffee cup from Kütahya in Turkey (Fig. 17:9); a German porcelain coffee cup (Fig. 17:10); hard-paste imported ware, including a French café au lait bowl (Fig. 17:11); a plate imported from Turkey with an inscription in Turkish (Fig. 17:12); English plates decorated with a ‘willow-pattern’ design in blue and black (Fig. 17:13, 14); and hard-paste ware stamped on the base with the mark of the manufacturer or the exporter in England (Fig. 17:15), Belgium (Fig. 17:16), Germany (Fig. 17:17), and France (Fig. 17:18, 19). The pottery assemblage also includes oil lamps (Fig. 17:20), locally made tobacco pipes (Fig. 17:21), tobacco pipes imported from Turkey (Fig. 17:22, 23), black locally made nargilehs (Fig. 17:24), red nargilehs imported from Tophane in Turkey (Fig. 17:25) and a Marseille roof tile (Fig. 17:26).
The glass finds include small bowls (Fig. 18:1), beakers (Fig. 18:2), stoppers (Fig. 18:3) and bracelets (Fig. 18:4–6). The metal finds include pack-animal shoes (Fig. 18:7) and a nargileh nozzle (Fig. 18:8). Filed mother-of-pearl inlays for jewelry and ornaments were also found (Fig. 18:9), as was a bone cigarette holder (Fig. 18:10).
The soil accumulations that covered the remains also yielded two coins from the rule of the sultan ‘Abed al-Majid (Constantine mint [Algeria]; 1857 CE, 1839–1861 CE), a coin from the rule of ‘Abed al-‘Aziz (1861–1875 CE) and a German coin (Berlin mint; 1897 CE).
The excavation uncovered remains of fortifications from two periods beneath and beside the Ottoman city wall, including fortification walls that predate the Middle Ages and parts of a fortification wall from the twelfth–thirteen centuries CE. The twelfth–thirteen-century CE fortifications join other segments that have been excavated beside the Ottoman walls and attributed to Jerusalem’s medieval fortifications.
The renewed exposure of a segment of Channel 108 enables us to reconsider the dating of the water system leading from east of the Damascus Gate to the Temple Mount (the Damascus Gate channel; the Upper Pool conduit; the Hasmonean aqueduct). Some scholars have suggested dating it to the First Temple period and some to the Hasmonean period. The finds show that the channel segment that was excavated is located in the water system’s northern part, as it appears on Schick’s plan (Fig. 4). It appears that the Channel 108 segment was built later than the original hewn channel, although it follows the course of the ancient channel. Although it is impossible to determine when the earlier channel was hewn, the later building phase found in the current excavation is dated to the Crusader–Ayyubid periods by finds retrieved from its bedding, the diagonal drafting marks on the stones of its walls and the date of the fortification wall against which it rests. Future excavations near Damascus Gate and Zedekiah’s Cave may shed light on the course of the channel to the north of the Ottoman wall.
The complex water system uncovered near Damascus Gate was part of the urban infrastructure that was in use at the end of the Ottoman period and during the British Mandate era, and it attests to Jerusalem’s urban growth and the importance of the gate, which served as a main entrance to the city in this period.