The most important building remains discovered during the excavations are related to Jerusalem’s fortification system south of the Pool of Siloam at the time of the Second Temple period. Most of them were already documented in studies by Guthe and Bliss and Dickie in the late nineteenth century and, once again, but in part, in Areas F and X in Kenyon’s excavation in the 1960s (Guthe 1882:136–141; Bliss and Dickie 1897:97–119, General plan II: H2, N2, O2, P2; Kenyon 1964:11, Pl. 5A–B [Area F]; 1965:14–17, Fig. 2, Pls. 6A, 7A–B , 8A [Area F]). In archaeological excavations carried out in the southern part of the City of David since the 1970s, other remains have been uncovered that are connected to the finds currently discovered (Adan-Bayewitz 1979; De Groot, Cohen and Caspi 1992 [Area A in Yigal Shiloh’s excavation]; De-Groot 1996; Reich and Shukron 2006; Greenhut 2011; Z. Greenhut and G. Mazor, pers. comm.; the principal architectural finds from their 2009 excavation [A-5653] are incorporated into the plan in Fig. 2 with their permission; for a review of the situation prior to the current excavations, see Weksler-Bdolah 2013:171–176).
The excavation (max. depth c. 3 m; Figs. 3, 4) reached down from the present road level to the natural bedrock. The remains were better preserved in Area B, the eastern area, where the bedrock surface was rather deep, whereas in large parts of Area C, the western area, the natural bedrock was discovered just below the modern road; in the latter, the archaeological levels above the bedrock were removed in the past. The principal remains date to the Iron Age II (Stratum 4; eighth century BCE), and the Late Hellenistic (Hasmonean) and the Early Roman periods (Stratum 3; first century BCE–first century CE). Scant remains were also revealed from the Late Roman period (Stratum 2; second century CE) and the Byzantine–Umayyad periods (Stratum 1; sixth–eighth centuries CE).
The important architectural remains include a city wall, which we propose be identified with the ‘First Wall’ of the Second Temple period; a gate incorporated in this wall south of the Pool of Siloam; and a gate-tower that juts out from the wall south of the gateway. The finds are described below from earliest to latest.
Stratum 4: The Iron Age II (eighth century BCE)
Phase 4b. Pottery sherds from the eighth century BCE are ascribed to the oldest habitation phase. These were discovered inside a layer of soil that accumulated in the natural bedrock pockets and at the bottom of a shallow, natural cave. A straightened level of small stones covering a small area, perhaps the floor of an installation, was preserved in one of the notches. This layer was sealed below the core of a wide wall from Phase 4a.
Phase 4a. In this phase, a well-built stone foundation formed the base of a wide wall (L220, W31; Figs. 3–5). The foundation was built of partly hewn fieldstones (average size 0.3 × 0.4 × 0.4 m) packed together without mortar; two courses were preserved inside a natural depression in the bedrock. Most of the pottery sherds discovered between the stones of the foundation date to the eighth century BCE, although several later potsherds that correspond to Stratum 3 were discovered in a few baskets. It can be cautiously suggested that this stone foundation was part of the core of a wide wall. The wall, aligned along a southwest–northeast axis, was situated directly below the ‘First Wall’ (W11).
A rock-hewn drainage channel, which curved from north to southeast, may have also belonged to this phase (L133; Fig. 3). The channel (exposed length 2 m, width c. 0.4 m, excavated to a max. depth of only 0.5 m) was covered with large stone slabs. It was deliberately blocked and canceled when the gateway was installed in the ‘First Wall’ (Stratum 3, below). This channel is very similar to a channel of the First Temple period previously discovered nearby, in Area F of Kenyon’s excavation (Kenyon 1965: Pl. 7B, Fig. 2).
Stratum 3: The Second Temple Period—The Hasmonean and Herodian Periods (second–first centuries BCE–first century CE)
Most of the remains discovered in the excavation belong to six different phases of this stratum.
Phase 3f. Remains of stone quarries in the natural bedrock (L222, L245; Fig. 6) are ascribed to the earliest phase (3f). They included quarrying steps and severance channels of rectangular stone blocks (length greater than 1 m, average width 0.6–0.8 m).
Phase 3e. This phase comprises of a city wall, a gateway and a tower that protrudes eastward from the wall (Fig. 9).
City Wall. A broad wall (W11; length c. 20 m, width 3.0–3.3 m; Figs. 3, 7, 8) was erected in this phase above the quarries in congruence with the quarrying steps. The wall was previously documented along tens of meters by Bliss and Dickie. We suggest that this wall, extending from southwest to northeast, be identified with the ‘First Wall’ of the Second Temple period. The outer face of the wall just south of the gateway was built of coarsely hewn square stones (average face size 0.2–0.5 m) with white mortar between them. Further south, the wall merged with a hewn bedrock cliff and all that remained of it is the foundation of its core built of dry construction utilizing medium-sized  fieldstones (0.1–0.3 m each dimension) arranged side by side in neat courses. The core of the wall sealed beneath it the core of the broad wall from Phase 4a (above, L220, W31).
Gateway (L127; Fig. 3). The southern part of a previously unknown opening (presumed width c. 3 m) in the northern part of the wall (W11) was exposed. The width and the careful construction of the opening indicate that it was a gate in the city wall. Its full width was reconstructed on the basis of a stepped installation (L124) that was built later in the Second Temple period along the inside face of the city wall, north of the gateway (Phase 3d, below). The gate’s southern doorjamb was constructed of dressed, soft limestone (nari), and north of it were two carefully smoothed threshold stones that were preserved in situ (elevation 634 m asl). The southern wall (W21) delimiting the gateway was built of stones dressed in the Hasmonean style, with drafted margins and a prominent boss (Figs. 3: Section 1–1; 10). The natural bedrock inside the gateway was weathered as a result of karstic activity, and it seems that the gateway was originally paved with stones, which were later dismantled. Part of a stone slab pavement was preserved c. 3 m northwest of the gate threshold, inside the city, at an elevation c. 1 m higher than the threshold itself (L115; 635 m asl; Fig. 3). This pavement was apparently part of an open, stepped area that extended up into the city (Fig. 9). A soil fill (L206) rich with Hasmonean-period potsherds (late second century–mid-first century BCE) was found inside an ancient quarry, the top of which was at the same level as the bottom of Pavement 115; this fill reflects the extensive public works conducted in order to level the area prior to laying the pavement.
On the basis of the finds and construction style it seems that the wall (W11) and the gate that was incorporated in it (L127) were built in the Hasmonean period. We can therefore identify it with the ‘First Wall’ of the Second Temple period.
Tower (L119; presumed outer dimensions c. 5 × 7 m; Fig. 3). Part of a tower that jutted eastward out from the wall was exposed south of the gateway. A water reservoir was installed inside it. Bliss and Dickie referred to the tower as a small, open cistern, and noted its inner dimensions (4.20 × 6.53 m, wall width 0.9 m; Bliss and Dickie 1897:100). In the present excavation, the tower’s southern wall was discovered (W10; length 4.75 m, width 1 m, height 3 m; Figs. 4: Section 2–2; 11); it was built in congruence with the remains of the quarries. The wall had a single foundation course built of large, coarsely hewn stones surmounted by three courses of ashlars dressed in the Hasmonean style with drafted margins and a prominent boss. Some of the stones were damaged, and it was apparent that repairs were made to the wall. The reservoir exposed inside the tower, which was partly excavated, had walls coated with gray hydraulic plaster (Fig. 12). The reservoir’s floor had a uniform foundation built of small fieldstones stacked together and bonded with gray mortar.
A wall (W12; width 0.9–1.0 m) with seven narrow steps (tread 0.5–0.6 m, rise 0.3 m) incorporated in it was built along the reservoir’s western wall, close to the ‘First Wall’. The foundation of a square pillar that apparently supported arches which held up the ceiling was preserved in the center of the reservoir. Although it is not absolutely certain, it seems that the reservoir was installed in the tower’s interior after the tower was already standing (Phase 3d).
A wall (W30; length 3.4 m, width 1 m, height within the current excavation 0.9 m; Figs. 4: Section 1–1; 13) was discovered west of the gateway and inside the city wall. It ran in a southeast–northwest direction, in line with the gateway’s southern doorjamb, and presumably delimited a street or an open square inside the gate. The wall was founded on an upright bedrock cliff (height c. 1.1 m), the remains of an ancient quarry (L222). A single course of stones (length 0.5–0.7 m, average width 0.5 m, height 0.55 m) dressed in the Hasmonean style, with drafted margins and a prominent boss, was preserved in the wall above a foundation course. Two of the stones bore on their edge delta-shaped engraved mason’s marks (Fig. 14). The wall was documented in the past by Bliss and Dickie; its continuation, outside the boundaries of the current excavation, was unearthed by A. De-Groot (1996). Presumably, there was an opening (width 2.75 m) that led to the tower between the city wall (W11) and W30. The opening was blocked in a later phase with proper construction of stone courses devoid of drafted margins or a boss. This blockage is clearly visible in Bliss and Dickie’s drawings, which depicted the entire wall to a greater height.
Two building phases were apparent in the part of the city wall south of the gateway, which was preserved to a height of two courses. The upper course of stones was set in a fashion that did not correspond with the lower course, indicating that the two courses were constructed at different times. It therefore seems that the upper course reflects a repair or a rebuilding of the city wall in this part of the city, and it may indicate that the gateway was negated in this phase. It should however be noted that on the basis of the small finds discovered between the stones of the wall, the later phase also occurred during the Second Temple period.

Phases 3d–3c. Phase 3d reflects changes that occurred in the layout of the buildings. The First Wall was repaired and the gateway was probably blocked. The water reservoir was installed inside the tower that protruded from the line of the wall and the opening that led from the gateway toward the tower (in W30) was blocked with proper construction. In this phase, a stepped plastered pool that was partially exposed was added north of the gateway; evidence of changes and repairs carried out in the pool (L124; length 3 m, width 2.5 m, depth c. 2.5 m; Fig. 3) were also discovered. Six narrow steps (tread depth 0.4 m, rise c. 0.4 m) were installed in the pool. The southern wall of the stepped pool was built above Pavement 115 and therefore postdates it. The northern part of this pool was exposed in Greenhut and Mazor’s 2009 excavation (Fig. 2). At some point, after Phase 3e (3d or 3c), meager walls were built south of the gate-tower, parallel and perpendicular to the tower’s southern wall (W18, W20, W25; Fig. 3); several of the walls were leaning on the tower. Pottery sherds, mainly of vessels dating to the first century CE were found in the soil fill next to these walls. It seems the walls were meant to stabilize the soil and for constructing terraces on the steep lower slopes outside the city wall.
Destruction in 70 CE (Phases 3b, 3a). South of the tower and above the terrace wall tops was soil fill sealed by a level of black ash (L223; Phase 3b; Fig. 4: Section 3–3), the top of which abutted W10. This level sloped from north to south, following the topography. Above it were collapsed ashlars (L122; Phase 3a), some of which had drafted margins and a prominent boss. The ashlars probably originated in the city wall and gate-tower, as well as from the dwellings inside the city wall, when they were destroyed. Also found among the debris were parts of columns, stucco fragments from wall paintings and a stone roof-roller, all of which likely came from the dwellings inside the city wall (Figs. 15, 16).
Stratum 2: The Roman Period (post 70 CE – the second century CE)
Pottery vessels that first appear in Jerusalem in the Roman period after 70 CE and are common in the city, mainly in second-century CE contexts, were found in the soil fill that sealed the collapsed ashlars and among the collapsed, undressed stones discovered on the high bedrock east of the wall. Some of the vessels are imports and most are local table and storage ware, presumably produced in the Tenth Legion kilnworks at Binyanei Ha-Umma (petrographic analyses have not yet been performed). A special find from this stratum was a column fragment that has a coarse square base, at the bottom of which is a square dressed protuberance that was most likely meant to anchor and stabilize the column (Figs. 17, 18). This might be the base of a milestone or border stone. The absence of an inscription makes it impossible to date the stone, but its ascription to Stratum 2 and to the finds presumably attributed to the soldiers of the Tenth Legion seem reasonable. The provenance of the finds could not be determined with certainty, but the proximity to the Pool of Siloam—the source of the city’s water also after 70 CE—allows us to suggest that Legion soldiers probably guarded the pool.
Stratum 1
This stratum consists of a meager wall consisting of one row of stones (W13) and a small section of a plaster floor that were discovered above Stratum 2. Their date cannot be determined accurately, except that they are from the Byzantine or Early Islamic period (sixth–eighth centuries CE), the same time as the pottery sherds in the fills discovered alongside the wall, or later.