Phase 6. A small portion of a white mosaic floor (L125; Fig. 5) that abutted a plastered wall (W60), which delimited it from the east, was exposed. The use of this floor was negated by the construction of the aqueduct (L116) and the plaster floor that adjoined it (L121) in the sixth century CE (below). The aqueduct’s western wall (W57w) was built on top of W60.
Phase 5. Two segments of the aqueduct (L114, L116), generally oriented north–south, were exposed and a rectangular underground pool (7 × 14 m, depth c. 9 m; not excavated), covered with a barrel vault, was revealed between them. The lower parts of the two aqueduct segments were hewn in the bedrock and their upper parts were built of medium and large fieldstones (W57e, W57w); the aqueduct was covered with large stone slabs. The channel, which had a U-shaped cross-section, was coated with a thick layer of plaster. The sides (W59e, W59w) of the southern aqueduct segment (L114) were low and this segment was connected to the pool via a round gutter (diam. 0.17 m). The northern aqueduct segment (L116) turned to the pool at a right angle. A circular entrance shaft (L105) was built above the bend in the aqueduct, probably meaning to clean the silt that accumulated in the water carrier. A white plaster floor (L121) that was discerned in the northeastern balk of the excavation abutted the opening of the shaft. This segment of the aqueduct and Floor 121 negated the mosaic floor (L125) of Phase 6. The fill sealed beneath Floor 121 contained potsherds dating to the sixth century CE, which is the earliest possible date for the construction of the aqueduct.
Phases 4, 3. Two phases were discerned in a section of the aqueduct, which branched off of Aqueduct 114 (L115a, L115b; exposed length 34 m). The aqueduct in this segment bypassed the rectangular underground pool from the east and continued north on a high bedrock terrace that was exposed along the western side of the Ben-Hinnom Valley. The aqueduct (L115b) was built of large, partially worked fieldstones, with almost no mortar in the early phase (Phase 4); today only the outer eastern face of the aqueduct is visible (Fig. 6). The level of the channel in this aqueduct is evident in the plaster and travertine remains that were left on the upper part of the aqueduct’s sides. Another aqueduct (L115a) was built in the late phase (3); its southern part was founded on the aqueduct from the early Phase 4 and its northern part was set on the bedrock and on the eastern side of Aqueduct 116 of Phase 2. Aqueduct 115b was not discovered in the northern part of the excavation; it was probably dismantled at the time Aqueduct 115a was built. The sides of Aqueduct 115a in Phase 3 (W50e, W50w) were built of medium-sized fieldstones and small stones and soil were poured between them. The channel of Aqueduct 115a, which had a rectangular cross-section, was coated with a thick layer of plaster (L108). A cast of small stones and light brown soil (L103) was deposited in the gap between the western side of the aqueduct and the rectangular pool. Several potsherds dating to the Ottoman period were discovered in the foundation of the aqueduct’s sides (W50e, W50w) and in the soil and stone fill (L103).
Phase 2. A black terracotta pipe (L129), set in poured cement, was installed below the bottom of Aqueduct 115a. Round openings were drilled in the bottom of the aqueduct to open blockages and release air pressure in the pipe. A small rectangular tower (L51; 0.9 × 2.2 m, height 2.2 m), founded on a layer of soil (thickness c. 0.5 m) that had accumulated on the bedrock, was built next to the eastern side of the aqueduct. A black terracotta pipe, which was used to divert water from Aqueduct 115a to the Sultan’s Pool and to a sabil that Suleiman I constructed in 1536 near the bridge on Hevron Road, was installed inside the tower.
Phase 1. A boy’s school was built over the rectangular underground pool at the time of the British Mandate. Photographs of the western side of Ben-Hinnom Valley show a rectangular structure with a tiled roof that sloped in two directions. The school was almost completely demolished. The school’s southern wall (W54), its concrete floor and the Marseilles roof tiles that were discovered on the surface are the last remnants of the structure, which were exposed above the southern segment of the ancient aqueduct (L114) and above the southern segment of Aqueduct 115a.
In 1847, William Henry Bartlett drew the region of the excavation from the roof of the rectangular pool, in the direction of the Sultan’s Pool. The aqueduct from the Ottoman period and the rectangular pool were mapped by Charles Wilson (Wilson C.W. 1866. Ordnance Survey of Jerusalem, 1865. London), by F.J. Bliss and A.C. Dickie (Bliss F.J. and Dickie A.C. 1898. Excavations at Jerusalem 1894-1897, General Plan N. II. London) and by C. Schick (Schick C. 1898. Birket es Sultan, Jerusalem, PEFQS 1898:224–229). Schick also surveyed the rectangular tower and the terracotta pipe that connected it to the sabil and the Sultan’s Pool.
The ancient aqueduct (Phase 5) was built at the earliest in the Byzantine period, whereas the next three phases, 4–2, were probably constructed in the Ottoman period. The use of the aqueduct ceased at the time of the British Mandate (Phase 1) and a boy’s school was built above it. The aqueducts and the pool are slated to be incorporated in the museum that will house the Moses Montifiore collection.