Area A
Area A was the easternmost excavation area. Two squares (5 × 5 m each; Figs. 3, 4, 7) were opened in a narrow section of the wadi streambed that flows west from Qastal and joins Nahal Kesalon. The excavation uncovered the remains of a massive terrace wall (W102), probably a retaining wall or dam, which seems to have been eventually cancelled by a road. Four occupation phases were identified (Strata IV–I); the remains are described in stratigraphic order.
Stratum IV (Figs. 3, 4). The remains of the earliest stratum in Area A consist of a single wall (W102; exposed length 6 m, width 1 m; Fig. 5) built in a north–south orientation of two rows of large boulders, with a fill of smaller chinking stones. The wall seems to rest on a step in the bedrock, as evinced by the level of the bedrock west of the wall, which is a meter below that to the east (Fig. 4: Section 4–4). East of the wall, a layer of reddish earth (L106, L108) covered the bedrock up to the top level of the wall. The soil matrix west of the wall (L107, L112) contained many more large stones and cobbles compared to the soil to its east. This may be a natural feature of the soil; more likely, though, it may result from the collapsing of the upper courses of W102, or is an intentional fill for the construction of the later road (Strata III–I).
Strata III–I (Figs. 3, 4, 7). The remains of a road were uncovered directly above the northern end of W102. During the excavation, it became clear that the road was in use for considerable time and had undergone three successive phases of resurfacing (Strata III–I). The remains of the original road were bounded by curb walls: a line of large, flat-topped stones to its north (W114); and a wall constructed of much smaller stones to its south (W115). Between the curb walls, and abutting their upper level, was a stone pavement composed of large stone slabs and small chinking stones (L116; 2.7 m wide; Fig. 6). No remains of the road were identified in the northeast corner of Area A, where a small probe (L110) was dug, going down below the level of the road.
The initial pavement eventually became obsolete for reasons that are not clear and was repaved with a dense layer of small, flat stone slabs over a thin layer of earth (L111, Stratum II; Fig. 8). This second pavement covered and cancelled curb walls W114 and W115. In fact, there were no clear remains of either curb or retaining walls that could be associated with this phase, and thus the exact width of the new road cannot be determined. In yet a later construction phase, uncovered just below the topsoil (Stratum I), a pavement of compacted cobble-stones (L100) was set over a layer of earth fill above the Stratum II pavement. At its northern extent, this pavement abuts the upper level of a curb wall built from a single row of small stones (W113; Fig 8). Wall 113 ran somewhat to the south of the original, Statum III curb wall (W114), and thus represents either a shift southward in the course of the road or a decrease in its width.
Very little pottery finds was found in Area A, and most of them were non-diagnostic fragments, heavily worn and eroded. In the absence of good contexts and identifiable material remains, it was impossible to date any of the archaeological strata or the associated remains.
 
Area B
An excavation square (2.5 × 7.0 m; Fig. 9) was opened along the delineation of Terrace Wall 4 (Fig. 2). Exploration revealed that the terrace wall was formed by two separate, superimposed walls (W200, W203; Fig. 9: Section 1–1). The construction began with W203, which was built of two parallel rows of fieldstones with an earth and cobblestone fill between them (1.5 m wide; Fig. 10). Only a single course of W203 was preserved. Interestingly, it seems that great care was taken during construction to ensure that the inner (southern) face of the wall was straight. Excavation below the base of the wall showed it to have been founded upon a layer of soil with a high concentration of rocks. Removal of this fill exposed the naturally flat bedrock.
Wall 200 was built directly above the southern face of W203, of a single row of medium-sized fieldstones. The two were separated by a layer of light brown earth (Fig. 11). The fill between the walls indicates that perhaps the earlier phase (W203) was abandoned for some considerable time prior to the construction of W200. As W200 was visible prior to the excavation and several of its stones exhibited rounded scars made by a large drill, this part of the wall may in fact be a modern construction.
 
Area C
Area C is the westernmost section of the excavation, between Terrace Walls 1 and 2 (Fig. 2). Two squares (4 × 4 m each; Fig. 12) were excavated, and sections of additional terrace walls were uncovered. Unlike the other terrace walls, these had not been visible on the surface; they were located during trenching prior to the excavation, and therefore seem to have been out of use for a long time.
Square C1. Removal of the topsoil cover (L301) uncovered a terrace wall (W300; Fig. 13) built in an east–west orientation, directly upon a flat bedrock surface. To the south it abuts a rise in the bedrock. The western section of the wall was constructed of two rows of medium-sized fieldstone, whereas the eastern section was built of a single row.
Square C3. Two parallel terrace walls (W303, W304; Fig. 14) built in an east–west orientation were uncovered. Wall 304, on the south, was built of two rows of medium-sized fieldstones and set directly below and to the north of a bedrock ridge, a position similar to that of W300 (Sq C1). It is possible that W300 and W304, having a similar orientation, are remains of the same wall, but this cannot be confirmed because the area between them was not fully exposed. Wall 303, which runs north of W304, was built of a single row of fieldstones. The excavation to its north (L305) uncovered the flat bedrock surface upon which it was set. A small probe (L306) in the western area of the square, between the two walls, exposed a fill, perhaps intentional, of reddish soil with a high concentration of cobbles.
Only a few ceramic sherds were found in Area C. The range of forms and vessel-types dates mainly to the Second Temple and early Byzantine periods. These finds, however, do not date the construction, use, or maintenance of the terrace walls.
 
Area D
Area D (c. 5 × 8 m; Figs. 15, 16) is located north of Terrace Wall 2, near the highway (Fig. 2). Remains of a road similar to the one uncovered in Area A were found here. The road was oriented approximately east–west, and its northern part was damaged by the modern highway. Along its southern edge, part of a curb wall (W400) was preserved. It was built of a single row of large, unworked boulders. To the north of W400 and abutting its upper level, a high concentration of loosely packed stones formed a pavement (L405). Fragments of a glass bottle that were uncovered during the removal of the cobbles indicate that the road was probably modern. Excavation south of W400 to below the level of its base uncovered a layer of earth fill above the bedrock (L401). The fill contained a small quantity of sherds dateable to either the Second Temple or the Early Byzantine periods.
 
Area E
Two squares (4 × 4 m each; Fig. 17) were excavated at the easternmost extent of Terrace Wall 5, to the west of Area A (Fig. 2). Part of the W5 was exposed in the western square, and a second wall (W503)—in the eastern one, directly under the topsoil (L502). Wall 5 is oriented east–west and was visible on the surface prior to the excavation. It seems to end abruptly, and may have been dismantled. Below the topsoil (L501), it was constructed of a single row of stones. Excavation under the base of W5 showed that it was built upon a layer of soil above the bedrock (Fig. 18).
Wall 503 was built of two rows of fieldstones, set in a north–south orientation directly upon a flat bedrock surface. This orientation, perpendicular to most of the other terrace walls in the area, may indicate that it served as an agricultural plot marker or, less likely, as a dam, similar to W102 in Area A.
 
The excavation uncovered a small portion of a regional agricultural system. The system is perhaps best defined by the terrace walls that dominate the mountain slopes in the region. As is usually the case, the terraces were probably reused and routinely maintained over a long period of time. Dating terrace walls by traditional methods, such as ceramic typologies, is extremely problematic. Hence, soil samples were taken in strategic locations around the terrace walls in the hope that the OSL (Optically Stimulated Luminescence) dating method would help determine the date of construction of the walls. We hope to publish the OSL results in the future as part of a larger regional study.
The two road sections uncovered in the excavation share characteristics indicative of ancient roads, such as curb walls and pavements. However, they do not seem to have been major roadways although this area has always served as a crucial link between Jerusalem and the coastal plain. It is likely that these remains belonged to side-roads that allowed access to agricultural plots, connected villages, and possibly served at times as a link to the main Jaffa–Jerusalem road. The segment of road in Area D seems to have been modern, but the road uncovered in Area A cannot be dated with certainty. Like the terrace walls, this road was also sampled for OSL dating.