The Northern Area. In the western part of the excavation, below the soil fill, a soft limestone surface and a level of small and medium stones (L103; width c. 3 m) were exposed. The western side of the stone level reached the bedrock in the form of a built wall plastered on its outer side. On the eastern side, the stone fill and limestone bedrock served as the western side of a built and hewn water channel (L102; Figs. 4, 5) that was aligned north–south. The channel was coated with white plaster mixed with small ground potsherds and large amounts of soot (exposed length c. 3 m, width c. 0.4 m, depth 0.6 m). Another section (length 1.5 m) was exposed in the northern excavation square (L105) and an excavated trench (L104; width 1 m) was exposed parallel to it in the western part of the excavation.
The Southern Area. Two parallel rock-cut channels, oriented north–south, were exposed. The eastern channel (L106; exposed length 6.5 m, width 0.4 m, depth 0.6 m; Fig. 6), hewn in soft limestone bedrock, was coated with white plaster mixed with small stones, small potsherds and charcoal. The channel went out of use and was canceled as a result of a deliberate blockage of small and medium fieldstones. The western channel (L107; exposed length 4.5 m, width 0.4 m, depth 0.25 m), hewn in soft limestone bedrock, was not plastered and seemed to be an unsuccessful quarrying attempt. Shallow rock-cuttings, whose nature is unclear, were exposed between the two channels. The bedrock descended to the east along the eastern side of the eastern channel. The builders of the channel leveled the area (L108) with a wall built of large fieldstones in dry construction that ran parallel to the channel in a north–south direction, as well as with small fieldstone fill that abutted the stone wall.
A complex comprising a hewn and built water channel and a nearby parallel road was exposed in the excavation. The overall length of the two channel sections was ten meters. The bottom of the channel descended from an elevation of 761.90 m at the southern end to 760.95 m at the northern end. The water channel was built in a north–south direction along the watershed. No artifacts inside the channel or in the fill of the road, which can assist in dating the complex, were discovered. A C14 analysis of the white plaster samples taken from inside the channel (L102) and of the gray plaster from the blockage phase and the road (L101, L106) date the construction of the channel to the first–second centuries CE.
Based on the route of the channel, the manner of its construction and the kind of plaster used, this installation was probably part of the High Level Aqueduct from the Roman-Byzantine period that conveyed water from Solomon’s Pools south of Bethlehem to the Upper City of Jerusalem. Other sections of the aqueduct were excavated at Giv‘at Ha-Matos (HA-ESI 111:105*) and along Derekh Hevron (ESI 15:78–79).