The western section (L120; central map ref. NIG 252557/769202, OIG 202557/269202) consisted of a shallow curving channel (depth 0.3–0.4 m; Fig. 2) of uniform elevation (241.22–241.23 m above sea level), part of which had been exposed in the past (‘Atiqot 28:1*–7*, Fig. 2:III [Hebrew]).
The eastern section (map ref. NIG 252615–55/769200–5, OIG 202615–55/269200–5) contained a well-preserved, narrow (bottom width 0.1–0.3 m, upper width 0.5–0.6 m) and deep (average depth 1 m; Figs. 1: Sections 1–3; 3) channel, whose quarrying was both calculated and arduous. In the western part of the section (Loci 240, 250, 260) the channel curved gently on its way east (Fig. 4). Point IV (‘Atiqot 28: Figs. 2, 6) close to the western end of the segment (L250) and Point V (‘Atiqot 28: Fig. 2) at the eastern end of the segment (L310) were re-exposed. The excavation of the aqueduct’s course between these two points was completed; although no signs of the channel were visible on surface prior to the excavation, it was ascertained that this was a continuous and well-preserved channel. The channel becomes narrower and very deep (depth 1.7 m) for a length of 15 m in the middle of the segment (from map ref. NIG 252625/769201; OIG 202625/269201 eastward; Loci 270, 280, 290), where it appears as a large deep fissure in bedrock, resembling a canyon. It is feasible that the quarrymen exploited a natural crevice in bedrock, widened it and integrated it in the course of the channel (Fig. 5). Three cupmarks, cut in bedrock surface above the southern side of the channel, were discovered; two were perpendicular to the channel (L291; diam. 0.3 m, depth 0.15 and 0.2 m; Figs. 1: Section 1; 6). The bottom elevation of the channel in the eastern segment (240.84–241.03 m asl) was 0.2–0.4 m lower than the western segment (241.23 m asl), indicating a 0.2% gradient. For the 110 m long section the gradient was gentler than the average overall gradient measured in the past (0.5% for a 1.1 km distance from west to east; ‘Atiqot 28: 2*). In the last 80 m of the channel, before it disappears to the east, the bottom elevations of the channel were measured at the points where it was visible along the surface (241.02, 240.87, 240.70, 240.35 m asl, from west to east), showing a steeper gradient close to 1% in this section.
Along this aqueduct section (Loci 260, 270, 280, 290) sixty-three potsherds were collected from the soil fill that had accumulated in the channel, mostly in the higher elevations of the soil fill, c. 0.1 m or more above the channel’s bottom. Most of the potsherds dated to Middle Bronze II and included a bowl (Fig. 7:1), kraters (Fig. 7:2, 3), a jar (Fig. 7:4) and a handle (Fig. 7:5), as well as four potsherds from the Roman-Byzantine periods and the Middle Ages. Fourteen body fragments, ascribed to the Middle Bronze Age, were discovered on the bottom of the channel (L270), with no other later finds. Between the eastern and western segments of the channel only the bedrock surface was discovered. The eastern end of the western segment (Fig. 8) and the western end of the eastern segment (Fig. 9) were not cut abruptly but rather disappeared gradually. Bedrock surface between the segments had no obvious upright rock protrusions and in most cases, it was no higher than the vicinity of the sections where the channel was preserved (Fig. 10). It therefore seems that in the past the channel had existed uninterrupted but over time it eroded away and was not preserved in this area. Notwithstanding, the possibility that the channel was not hewn at all in this region cannot be negated. Based on this option, which has yet to be proven, the aqueduct project was never completed, probably because of engineering problems. At the western end of the eastern segment (L240) and close to the eastern end of the eastern segment (L300), signs of a calcium deposit (travertine?; Fig. 11) were detected on bedrock surface. This, however, could have been deposited after the channel was filled with soil and does not necessarily attest to a continuous flow of water over a prolonged period.