Until the exposure of the row of large stones, no archaeological remains of any period had been exposed in this specific vicinity, which was a productive pomegranate orchard. The c. 50 sq m excavation area revealed that the row of large stones was in fact part of a paved road. The direction of the road suggests that it may have provided access to and from the spring to Karm er-Ras, located c. 300 m to the northwest.
The excavation reached a total depth of 3 m below surface, without attaining bedrock, although the packed nature of the earth indicated that it was close. Almost 2 m of soil were removed by a mechanical digger, once it was established that this was a thick layer of agricultural soil, entirely void of stones and archaeological finds. This soil may have been transported by the spring waters, although the complete absence of stones or pebbles is surprising. Two wide stone walls (W103, W104; width c. 1.2–1.5 m) were uncovered beneath the soil layer, running parallel at a distance of 2.0–2.5 m (Fig. 1). Both walls were built in a similar fashion of two faces with a core of small stones (Fig. 2). The outer face of each wall consisted of a single course of mostly vertical large stones (average dimensions 0.3 × 0.4 × 0.5 m), placed directly on the soil layer. The inner face of both walls was built of about three courses of slightly smaller stones (average dimensions 0.3 × 0.4 m) that were placed horizontally, one on top of the other. Some of these stones seem to have fallen out of line over time. A section (depth 1 m) was cut between W103 and W104, exposing several superimposed layers of small to miniature stones (c. 0.1 × 0.1 m), packed closely together with trodden earth to form a solid road or path. The two lowest surfaces (L110, L111) were only exposed over a small area, laying respectively 0.5 and 0.25 m below the base of the inner face of the walls. This suggests that these earliest road surfaces must have existed prior to the construction of the walls. The upper two surfaces (L102, L109) were clearly bordered by the walls. In effect, since the whole depth was filled with stones, it was not entirely clear whether there were four or more distinct road surfaces or if the stone layers had simply built up over time. These layers, however, were clearly trodden surfaces and not water-transported layers. The well-preserved upper surface (L102) was exposed for the whole length of the excavation (6.5 m) and obviously extended beyond the limits of the excavation area in both directions.
The road surfaces incorporated fragmentary potsherds, predominantly fragments of jar handles that survived better than body sherds due to their thickness. It was difficult to identify the potsherds on the basis of their form because of their wear. Identifiable potsherds in the lowest layers included a Hellenistic jar (Fig. 3:1); Middle and Late Roman ceramics, including Kefar Hananya ware (Fig. 3:2, 3) and a Shikhin jar (Fig. 3:4). The potsherds in the upper road (L102) included bowl fragments dated to the Byzantine period (Fig. 3:5, 6), a bowl dating to the Crusader period (Fig. 3:7) and another bowl from the Mamluk period (Fig. 3:8); identifying the vases was based more on ware than on form. Occasional animal bone fragments were found in an extremely worn state.
It seems that a paved path was located here since at least the Roman period. At some stage, possibly in the Byzantine period, two low and wide parallel walls were apparently built as retaining walls to support the paved road. The outer vertically placed stone face of these walls must have been backfilled, while the inner face was originally free-standing (height c. 0.6 m) and bordered on the road surface on either side. The road became higher over time, either by intentional depositions of additional small stone layers, or by the treading of accumulated small stones; at the latest use, the stone walls were simply wide curbs, c. 0.2 m above the stone road.