he earliest element uncovered in the area is a wide solid wall (W4493; over 8 m wide) built on a southwest–northeast axis. The wall was constructed by casting yellowish mortar mixed with small fieldstones, and its southeastern face was faced with a row of large fieldstones; the northwestern face was not exposed as it lay beyond the excavated area. The northeastern part of the wall was cut by the Western Wall (W4205), which it therefore clearly predates. Similarly constructed and aligned wall segments have been discovered in several places to the west of the excavation area, and they are probably parts of the same massive wall (Onn, Weksler-Bdolah and Bar-Nathan 2011; Onn and Weksler-Bdolah 2016). The function of this wall has been widely discussed: it has been proposed that it was part of the First Wall (Warren and Conder 1884:206–207; Uziel, Lieberman and Solomon 2017), the foundation of the great causeway (Hamilton 1932; 1933), a dam wall (Bahat 2000:38) and a road-bearing dam wall (Weksler-Bdolah 2015). Based on the stratigraphy and data from previous excavations (Onn, Weksler-Bdolah and Bar-Nathan 2011; Bahat 2013), this solid wall probably dates to the Hasmonean period.

The pier of Wilson’s Arch (W4202, W4204), the Western Wall (W4205) and drainage channels (L4513, L4517) are attributed to this stratum (Figs. 2, 3). Three construction phases (7C–7A) have been identified in the western pier supporting Wilson’s Arch (W4202, W4204; c. 15 m total length; Fig. 4). This pier is composed of two adjacent piers, a northern one W4204; 5.0 × 7.5 m) and a southern one (W4202; 5.0 × 7.5 m), exhibiting different methods and dates of construction. In the earliest phase (7C), in the Early Roman period, the northern pier (W4204) was built of eleven courses of stones with drafted margins. Pier 4204 probably originally bore an earlier arch that was as wide as the pier (7.5 m; not preserved). A small cavity built in the center of the pier (L4402; 2.45 m long) had an arched opening leading to it (1.3 m wide. 3.15 m max. height) with a stone lintel surmounted by a relieving arch. Pier 4204 was built on top of the Stratum 8 massive wall (W4493), which served as its base. In the interim phase (7B), within the Early Roman period, the southern pier (L4202) was built of six courses of large stones without drafted margins. Wilson’s Arch itself was also built of similar large stones, and the southern pier and the existing arch were probably constructed at the same phase. Two small rectangular cavities (L4404—1.5 × 2.4 m, height 3.5 m; L4405—1.5 × 1.8 m, height 3.5 m) were built inside the southern pier. The openings to the two cavities are constructed in a similar manner, with a large stone lintel supported by the entrance doorjambs. Cavity 4405 was excavated in the past by Warren while digging the western of the two shafts that he dug at the site. The Western Wall was evidently built in this phase (below). In the latest phase of the stratum (7A), probably in the Late Roman period, two additional cavities (L4401—1.4 × 2.7 m; L4403—2.1 × 2.3 m) were cut into the northern pier (W4204). It seems that W4493 from Stratum 8 was probably leveled and used as the floor for the two cavities. Despite the different dimensions of the cavities in the northern and southern piers, their construction technique is similar, and they were therefore probably opened up at the same phase. Sockets for lintels were cut in the upper courses of the entrances to the two cavities; the lintels, presumably made of wood, were not extant. Since the socket in the southern jamb of cavity L4403 was cut into the southern pier (W4202), it can be established that the two cavities post-date the construction of the W4202 pier, and that they belong to the final construction phase (7A) of the western pier of Wilson’s Arch.

The Western Wall stands c. 13 m to the east of the Wilson’s Arch pier and it was probably built in Phase 7B, together with the southern pier (W4202). The excavation unearthed nine additional well-preserved, stone courses of the Western Wall (Fig. 5), dressed with double drafted margins. Based on the historical sources and archaeological excavations, the Western Wall was constructed in the Early Roman period, namely between the rule of Herod the Great and the period of the Roman procurators (Reich and Baruch 2017).

A drainage channel (L4513, L4517; c. 0.95 m wide; Fig. 6) unearthed in the center of the excavation area was built on north–south alignment, parallel to the Western Wall and c. 7 m to its west. The channel was sealed beneath a building attributed to Stratum 6 (below). The northern part of the channel (L4517) was hewn into W4493 of Stratum 8, whereas its southern section (L4513) was constructed of two thick walls (0.55 m wide) built of ashlar stones. The upper courses of the built channel (L4513) appears to have been dismantled when the Stratum 6 theater-like structure was built. The floor of the northern rock-hewn channel is higher than that of the southern, stone-built channel. The northern channel may have been hewn in an early phase of the stratum (7C), veering slightly to the west; the southern section may have been built in a later phase (7B), with the widening of the Wilson’s Arch pier and renewed planning of the area.

A small theater-like structure (Fig. 6) was unearthed that contained an orchestra, a stage area, entrance passages, staircases and a seating area; the building was not completed. The structure was built over the Stratum 7 drainage channel, and above Stratum 8 Wall 4493, parts of which were removed and leveled to create a base for the structure. The outer perimeter of the structure is rectangular, and it is delineated by Wilson’s Arch in the west, the Western Wall in the east and by walls in the south (W4430) and in the north (W4451). A semi-circular wall (W4420) curving delineated the area of the orchestra (6 m diam.) was preserved in the center of the structure; pilasters, only partially finished, were preserved at each end of the wall. The stage is narrow (c. 3.0 × 10.5 m) and bounded to north and south by two pedestals (0.7 × 1.0 m), that are 1.2 m higher than the stage; the profiles of the pedestals were not fully finished. The stage was founded mainly on top of W4493, but large protrusions discovered in the wall in this area are evidence of preparations for a superstructure that was not completed; the stage was apparently therefore not finished. The passages (aditi maximi; c. 1.5 m wide) led into the structure from the north and south and their walls were built of well-dressed stones. A staircase (0.6 m wide) found at the end of each passage led to the seats, whose construction was not completed. Steps were installed in the seating area as a base for an array of stone seats (cavea); the seats themselves were not found, either because they were never built, or as a result of secondary use of the stones in a later period (see for example, Reich and Billig 2000). The steps found in the area of W4493 were built into the wall itself, whereas the steps to the east and south of W4493 were built using a system of supporting walls with large stones between them. Architectural elements from the walls of the Temple Mount were found within these stone fills (for a review of methods of fixing theater seats, see Sear 2006:71–80). A retaining wall (W4409) uncovered in the center of the seating area and built of roughly dressed small and medium-sized stones, abutted the Western Wall. The wall, which clearly post-dates the Western Wall, provides a stratigraphic anchor with which to date the structure. Two sections were excavated in the foundations of the seating, south of W4409 and W4451, in an attempt to date the structure. The sections retrieved ceramic finds as well as roof tiles, bricks and clay pipes characteristic of the period following Jerusalem’s destruction in 70 CE (first half of second century CE; for discussion see Uziel, Lieberman and Solomon 2017), thus dating the structure’s construction. The theater-like structure unearthed in the excavation is the smallest to be discovered so far in the country (Segal 2000: Theater Table Appendix; Mazor 2007; Weiss 2014:81–100). Although the building was not fully completed, it may have been used by the citizens of the Roman colony of Aelia Capitolina from the time when it was built early in the second century CE until it was deliberately buried under a fill of earth and stones in the latter half of the third century CE (below, Stratum 5).

A uniform fill of gray-brown soil uncovered throughout the excavation area contained a large quantity of small stones and gravel (L4258, L4305, L4339, L4341; at least 3 m thick; see Fig. 3). The fill was deposited from different angles, abutting the Western Wall and the piers (L4202, L4204) of Wilson’s Arch, as well as W4420 of the theater-like structure. The fill sealed the Stratum 6 theater-like structure. A water channel (L4261; 0.3 m inner width, 0.64 m outer width, depth 0.5 m; Fig. 7) unearthed in the upper part of the fill led from north to south, veering westward in the southern part of the excavation area. The channel had small fieldstone walls and a stone slab roof, and was not plastered, and its floor was covered with a hard layer of moist brown sediment. The channel was cut by the foundation trench of a wall (W4206) from a previously excavated Building F, attributed to Stratum 3 (below). Based on pottery finds and coins retrieved from the stratum, the channel probably dates from the latter half of the third century CE (for further details, see Uziel, Lieberman and Solomon 2017).

Dense deposits of ash, moist red clayey soil and chalk were discovered (L4304, L4310; max. thickness 2 m; Fig. 8) on a southwesterly descending slope. The deposits yielded a large quantity of organic matter. Judging by the gradient, the accumulations were probably not habitation levels but were the result of prolonged dumping of refuse. Based on the pottery found in the stratum, it probably dates from the fourth century CE (for further details, see Uziel, Lieberman and Solomon 2017).

The northern face of the northern wall of Building F that was uncovered in previous excavations was unearthed (W4206; Onn, Weksler-Bdolah and Bar-Nathan 2011; Onn and Weksler-Bdolah 2016). Wall 4206 (12.6 m long) was built of dressed stones and preserved to a height of ten courses (5.2 m preserved height), including the upper part of the wall foundation, which was wider than the wall. The foundation trench of W4206 cut into the Stratum 4 accumulations and the Stratum 5 Channel 4261 (see Fig. 3), dating Building F accordingly to not before the end of the fourth century CE. Wall 4206 abuts the western end of the northern wall of previously excavated Building E, that was dated to the second–third centuries CE (Onn, Weksler-Bdolah and Bar-Nathan 2011).

Also attributed to this stratum was a channel with a trapezoidal cross-section hewn into the Western Wall stones, which was discovered at the bottom of Course D of the Western Wall (according to Warren’s numbering; Warren 1884: Plan XXXIII). This channel was unearthed in Warren’s eastern shaft next to the Western Wall (Wilson and Warren 1871:81), and it continues to the northern end of Robinson’s Arch. Four rectangular hewn niches unearthed at the bottom of Course E may possibly also be attributed to this stratum; they lie c. 3.5 m apart and were apparently used to support wooden beams. The niches were evidently hewn in the Western Wall stones after the destruction of the Second Temple, and they may be related to the construction of a large water reservoir to the north of the excavation area.

Pits and walls (W4244, W4246) unearthed cutting into the Stratum 4 accumulations, yielded Early Islamic pottery. These pits and walls are sealed beneath the foundations of the Stratum 1 water reservoir (‘Birkat al-Burāq’; below) and therefore predate its construction.

A plastered water reservoir was unearthed that extended across the entire excavation area (Birkat al-Burāq’; at least 15 m long, 13 m wide; Fig. 8); it was dismantled at the beginning of the excavation and hence does not appear on the plan. It was encountered directly beneath the synagogue’s stone floor. The eastern part of the reservoir may continue southward, as documented by Wilson (Wilson 1865:28–29; Bahat 2013:222–241). The reservoir was founded in the west on the western pier of Wilson’s Arch and in the east on the Western Wall. The base of the reservoir walls is curved at the junction with the floor. Two construction stages were identified in the reservoir, the earlier from the Mamluk period and the later from the Ottoman period. In the earlier phase, the reservoir was built on a thick foundation of stones and hard mortar, coated with two consecutive layers of pinkish plaster. In the later phase, the floor of the reservoir was covered with another layer of stones and bonding material, and the reservoir was coated with a new layer of pinkish plaster. Based on the traces of plaster preserved on the Western Wall, the reservoir wall after its later plastering evidently stood to a height of 3.2 m.

• Fig. 1 - Location map
  Figure 1
  
  Map showing the area of the Temple Mount, the Great Causeway and Wilson's Arch.
  
  Uziel et al. (2019)
  from Uziel et al. (2019)
• Fig. 2 - Plan and Photo
  Figure 2
  
  Plan of the excavations, showing remains from Strata 8, 7 and 6a, b. Figure 2a:
  
  

  • pink = Stratum 8
  • green= Stratum 7C
  • orange = Stratum 7B
  • black = Stratum 6

  
  
  Uziel et al. (2019)
  of excavations from Uziel et al. (2019)
• Fig. 3 - Photo of the newly
  Figure 3
  
  View of the newly exposed section of the Western Wall beneath Wilson's Arch, looking east.
  
  Uziel et al. (2019)
  exposed section of the Western Wall beneath Wilson's Arch from Uziel et al. (2019)
• Fig. 4 - Drawing of the pier
  Figure 4
  
  View of the pier of Wilson's Arch, looking west. Numbers refer to strata as discussed in article.
  
  Uziel et al. (2019)
  of Wilson's Arch from Uziel et al. (2019)
• Fig. 5 - Photo of Stratum 7
  Figure 5
  
  View of Stratum 7 drainage channel.
  
  Uziel et al. (2019)
  drainage channel from Uziel et al. (2019)
• Fig. 9 - Section of the excavations
  Figure 9
  
  Section of the excavations beneath Wilson's Arch, looking east. Numbers refer to strata as discussed in the article. Stratum 5B is not shown in the figure as it covered Strata 6
  
  Uziel et al. (2019)
  beneath Wilson's Arch from Uziel et al. (2019)

The earliest feature uncovered in the excavation is a thick wall (Wall 4493) that runs southwest to northeast (Fig. 2). The southern face of the wall is built of large, uncut boulders, and a core of smaller fieldstones mixed with hard, orange-yellow mortar. The 8 m wide wall was exposed for a length of 13 m, although it clearly extended farther to the west and to the north. Stratigraphically, the wall is sealed by Stratum 7 remains that use it as a foundation, as well as Stratum 6 Building 4420 (see below). Wall 4493 is cut by the Western Wall of the Temple Mount (Stratum 7), indicating that it precedes the Herodian platform (for further discussion on the dating of the Herodian Temple Mount, see, e.g., Shukron 2012; Patrich and Edelcopp 2015; Weksler-Bdolah 2015; Reich and Baruch 2016). The early date of this feature is also supported by its alignment, different from all Herodian and later architecture, which is oriented north−south, east−west. The seeming western extension of this feature―or at least a similar feature using this yellow mortar, has been noted in many previous excavations in the Western Wall Tunnels compound (Wilson 1880; Warren and Conder 1884; Onn, Weksler-Bdolah and Bar-Nathan 2011; Onn and Weksler-Bdolah 2016). Warren and Conder (1884) suggested that this may be part of the First Wall, mentioned in the writings of Josephus Flavius, and dating to the Hasmonean period. Hamilton (1932, 1933) proposed that it served as the foundation of a bridge, while Bahat (2000) interpreted it as a dam. Weksler-Bdolah (2015) expanded on this idea, arguing that the proportions of the dam were very wide (Hamilton 1933 documents the width at a total of 12 m) since it also served as a major pathway leading to the Temple. Although our excavations exposed only a small portion of this wall, this is the first time that a clearly constructed stone façade was exposed, possibly supporting the theory that this may have served as a fortification wall―even possibly the First Wall, as suggested by Warren and Conder. From a chronological perspective, the initial dating of this feature to the Hasmonean−early Herodian period may be fitting. Furthermore, the orientation and location of the wall follows the line of fortification―probably the First Wall―discovered by Avigad in the Jewish Quarter (Geva and Avigad 2000: 224−225). This proposal must be treated with caution as the dating of the feature is far from certain, and since the other sections of the First Wall were not built of this orange-yellow mortar. Still, it is possible that these differences are due to changes in the topography.³

Stratum 6, dated to the first half of the 2nd century CE, features an impressive theatrelike structure (Building 4420; Figs. 2, 6) that was built above Stratum 8’s Wall 4493 and Stratum 7’s drainage channel. The interior of the building was semi-circular and the exterior rectangular. It was enclosed on the east by the Western Wall, on the west by the pier of Wilson’s arch, on the north by Wall 4451 and on the south by Wall 4430. A narrow stage area was located on the western side of the building, with pedestals extending from either end. The stage area was approximately 2 m wide and 10 m long, including the pedestals. The building was entered from two aditi maximi, located north and south of the orchestra. These 1.5 m wide passages brought audiences to two small staircases that led up to the cavea. The orchestra was a mere 6 m in diameter, attesting to the diminutive size of the building. This is seemingly the smallest known theatre-like structure in the eastern Roman Empire (Segal 2000: Table; Mazor 2007b; Wiess 2014), even smaller than the private theatre recently discovered at Herodium (Porat, Chachy and Kalman 2015).

In all, four staircases were discovered leading to the cavea―two, mentioned above at the end of the aditi maximi, and an additional two staircases at the centre of the podium wall (4420), separating the cavea and the orchestra. Clear signs on the staircases leading to the cavea showed that the carving was incomplete (Fig. 7), meaning that the building was never finished. Supporting this assumption is the incomplete molding of the cornices of the pedestals. Most of the building’s cornices and the podium wall were carved in stepped fashion, with no additional decoration (see, e.g., Peleg-Barkat 2017: 78−79). Such profiles are known for example in the odeon in Beth Shean (Mazor and Amos 2007: Figs. 9.6, 9.7). Certain cornices of the building, however, were left uncarved, remaining flat, with a triangular section. It appears that the stones were transported to the site in advance of final carving, which would have been undertaken after they were set in place. At the southern end of the podium wall, an incision was noted, indicating where a stone should have been carved as a pilaster; this carving never occurred, leaving the stone unfinished.

The walls of the aditi maximi are constructed of finely cut ashlars, overlying the Stratum 8 mortar. It appears that in the entire northern half of the building, the mortar of Wall 4493 was carved and straightened to serve as the base of the orchestra and the seats of the cavea. Here, too, it is clear that the process of construction was never completed, notably in the stage area, where the mortar of Stratum 8’s Wall 4493 was only partially carved; a significant portion remained uncarved and was higher than the intended height of the stage.

The use of the earlier Wall 4493 as a foundation is clearest in the area of the cavea, where it was carved in a stepped fashion, in order to fit the stone seats that would have been placed over it. The stone seats were not found in the excavation, save for one possible fragment that was not found in situ.⁴ At the centre of the cavea foundations stood a small wall (4409), which abuts the Western Wall, clearly indicating that the theatre-like structure was built after the construction of the Western Wall. South of Wall 4409, heaps of large stones created an artificial slope upon which the cavea could have been built. The heap included architectural elements that could be attributed chronologically to the Early Roman period, similar to those exposed near Robinson’s Arch (Peleg-Barkat 2017).

The structure found beneath Wilson’s Arch is small in comparison to other such structures, both locally (Segal 2000; Wiess 2014) and throughout the Roman world (Sear 2006). Therefore, it is evident that the size of the audience would have been limited. That said, there is still quite a range in the calculation of spectators (see Sear 2006: 26 for further discussion). The width of the cavea at the centre―extending between the Western Wall and the Podium―is 4 m. The width of the seats in a small, theatre-like structure ranges from 0.6 to 0.81 m (Sear 2006: Table 3.4). It is logical, considering other features in the building, to assume that the seats were narrow. The negative left in the carving of the mortar of Wall 4493 measures 0.50 m, suggesting that a minimal width should be considered for the rows― somewhere between 0.5−0.6 m―meaning the building had between 6−8 rows. If this is the case, following Sear’s (ibid.: 27−28) equation, the capacity was between 150−200 people.

In order to date the construction of the theatre-like structure, two probes were excavated into the foundations of the cavea.⁵ The vessels presented in Fig. 8 derived from these probes. The majority of the ceramic forms are typical of the 2nd century CE. Large basins and mortars, such as the rilled rim basin (Fig. 8: 1) or the thumb indented mortarium (Fig. 8: 2) are typical of the legionary pottery production in the Binyanei HaUma kilns (Magness 2005, Rosenthal-Heginbottom 2005, 2011). The cooking cauldron (Fig. 8: 3) and jug (Fig. 8: 4) also date to the 2nd century CE. Production of ceramic building supplies such as roof tiles, bricks and pipes began after the destruction of the Second Temple. The foundations of the theatre-like structure yielded hundreds of examples of ceramic building supplies such as the tegula roof tile presented in Fig. 8: 5. A few of the roof tiles uncovered in the foundations carried legionary stamps on them. Based on the above data as well as the negative evidence, the absence of ceramic forms typical of the late 3rd century CE (see below, Stratum 5), this layer and consequently the construction of the theatre-like building should be dated to the 2nd century CE.

Further supporting the dating of the structure to the period after the Roman destruction of Jerusalem in 70 CE is the presence of architectural elements that apparently originated from the Second Temple period Temple Mount (for the secondary use of earlier architectural elements as building materials, see, e.g., Stiebel 2008: 222−224). Lichtenberger (2006: 293) argued that the theatre seats found near Robinson’s Arch (Reich and Billig 2000) should be dated to the period subsequent to the destruction in 70 CE. While the relationship between these seats and Building 4420 cannot be securely established, the moldings of both are parallel.

A thick, uniform fill covered the entire excavation area of Stratum 5 (Fig. 9). Its matrix was characterised by gray-brown earth with many small stones that abutted the Western Wall and the pier of Wilson’s Arch. The fill of Phase 5b abutted circular Podium Wall 4420 of the theatre-like structure; Phase 5a sealed the structure, including the foundations of the cavea. Small drainage Channel 4261, at the top of the fill, led from north to south, turning westward at the southern end. The channel and the fill were cut by the foundation trench of Building F of Stratum 3 to the south, which had been excavated earlier (Onn, Weksler-Bdolah and Bar-Nathan 2011). The channel’s walls are built of small fieldstones covered with flat capstones. There was no plaster in the channel; neither was there any clearly constructed floor—only a hard, damp layer of brown sediment serving as a base. The channel measures 0.30 m wide internally, 0.64 m wide externally and is 0.50 m deep.

The ceramic assemblages of the two phases―5b and 5a―were treated independently. The vessel types constituting the assemblage from Phase 5b can be dated to the 2nd−3rd centuries CE. Many of the forms are similar to the vessels produced and used in the Binyanei HaUma workshop operated by legionary craftsmen during the Late Roman period, until the end of the 2nd century CE (Magness 2005; Rosenthal-Heginbottom 2005). Since several forms that appear in this assemblage were absent in the Binyanei HaUma finds, the date of this phase should be post late 2nd century CE. The numismatic finds and the similarities between the Stratum 5 ceramic assemblage and the one uncovered in L.6032 of the Western Wall Plaza excavations (Rosenthal Heginbottom 2011) indicate that Stratum 5 can be dated to the late 3rd century CE.

Numerous vessel forms occur in both Phases 5b and 5a. Several types of local bowls (Fig. 10a: 1−3), including many rouletted bowls with varied body and rim shapes and surface treatments (Fig. 10a: 4−5) and a number of large basins and mortaria were also found (Fig. 10a: 6−9) The cooking vessels comprised closed and open cooking pots, cooking cauldrons and cooking bowls (Fig. 10a: 10−12, 10b: 1). Jugs and juglets of several forms (Fig. 10b: 2−4), kraters with thumb indention decoration (Fig. 10b: 5) and dolia (Fig. 10b: 6) were unearthed. The typical storage jars of Stratum V exhibit an outfolded, smoothed rim (Fig. 10b: 7−10). Very few examples from this assemblage have an inner protrusion. The Stratum 5 lamps are ovoid with large filling holes or with a decorated disc (Fig. 10b: 11−14). Several slipped body sherds carrying combed and printed decoration were also found (Fig. 10b: 15).

As opposed to the ceramic assemblage of Phase 5a, the Phase 5b fill yielded no vessel types distinctive of the 4th century CE. This may indicate that the fill layers were deposited at different times. Still, the sloping nature of the Phase 4 fill (see below) may have permeated the Phase 5a loci.

Many rouletted bowls with variations of rim and body shape were found in this transition between Phase 4 and Phase 5b (Fig. 11a: 1−4). An imported African Red Slip bowl from 5a dates to 230–300 CE (Fig. 11a: 5, Hayes 1972: 73). Large bowls, basins and mortaria of several types were found: rilled rim, shelf rim, arched rim and cut rim basins (Figs. 11a: 6−10, 11b: 1). Combing decoration or incised wavy lines are visible on the rim or exterior of some examples. The cooking vessels are closed with a grooved rim or a convex neck (Fig. 11b: 2−3); other cooking bowls have horizontal handles (Fig. 11b: 4). Storage jars with a thickened rim were found (Fig. 11b: 5) along with jars with an outfolded rim (Fig. 11b: 6); some of the rims protruded inward (Fig. 11b: 7). Other jars were of the Gaza (Fig. 11b: 8), dolia (Fig. 11b: 9) or holemouth types (Fig. 11b: 10); the latter bears incised lines (letters?) on the shelf rim. The lamps from this layer were ovoid. They had large filling holes as well as items with a decorated disc (Fig. 11b: 13−16).

Stratum 4 is characterised by a dense striation of alternating ash and chalky and red-clayey tiers (Fig. 9). These do not represent surfaces, as they slope sharply to the southwest, and should be interpreted as the dumps of alternate layers within the area beneath the arch. The reasoning behind dumping of this fill is unknown.

A large pottery assemblage dating to the Late Roman period was uncovered within these thin, stratified fill layers. Similar pottery assemblages have been published from Mazar’s excavations near the southwestern corner of the Temple Mount (Mazar and Gordon 2007; De Vincenz 2011a, 2011b). Additional parallels derive from the Ophel excavations (Fleitman and Mazar 2015), the Givati Parking Lot (Balouka 2013) and several contexts in the Jewish Quarter excavations (Magness 2006, 2012).

The assemblage includes late 4th century items such as: bowls with a rouletted decoration, presenting several rim variations (Fig. 12a: 1−2); two imported African Red Slip bowls defined by Hayes as Type 50b dated to 350–400 CE (Fig. 12a: 3−4, Hayes 1972: 73) and Type 53b dated to the 4th–early 5th centuries (Hayes 1972: 78 – 82); rilled rim basins, shelf rim basins and arched rim basins. Several examples carry decoration on the body exterior (Fig. 12a: 5−8). The most frequent vessel uncovered in this assemblage is the bowl-lid. Typically, the bowl had rounded or slightly carinated walls, a thickened rim and a narrow ring base that was used as a handle (Fig. 12a: 9−11). The relatively few cooking pots uncovered were all of a closed, globular form with a convex neck (Fig. 12b: 1−2). The typical storage jars from this phase had an outfolded rim, some with inner protrusions (Fig. 12b: 3), or a thickening on the inner neck (Fig. 12b: 4−5). Other jar forms uncovered include Gaza jars (Fig. 12b: 6), holemouth jars (Fig. 12b: 7) and wide neck jars (Fig. 12b: 8). The representative lamps include ovoid items with large filling holes (Fig. 12b: 9−10) and round ones with decorated discs (Fig. 12b: 11).

Stratum 3 consists of a single building (Building F) found at the southern end of the excavation. The external, northern face of the building was exposed in an earlier excavation farther to the west and dated to the 3rd−4th centuries CE (Onn, Weksler Bdolah and Bar-Nathan 2011). It was interpreted as belonging to an expansion of an earlier structure (Building E). In the current excavation, the final 5 m to the east of the northern wall of the structure were exposed, along with its foundation trench. Stratigraphically it cuts the Stratum 4 fills, suggesting that it cannot date earlier than the 4th century CE. Ten courses of the building’s Wall 4206, built of fine cut ashlars and preserved to a height of 5.2 m, were exposed. The entire length of the wall is 12.65 m, abutting the corner of Building E (ibid.). The lower portion of the wall protrudes; the foundation is apparently wider than the superstructure.

To the north, a large plastered pool, excavated earlier by Bahat (2013: 226−227), belongs to this phase. The southern wall is the northern border of the current excavation. The pool had plaster walls that did not survive but which left a mark on the Western Wall; it occupies an area of 26 × 26 m north of Wilson’s Arch. The exact date is still unclear, as minimal ceramic material was retrieved when it was dismantled

This stratum displays many disturbances in the Stratum 4 fills, and several unclear poorlybuilt walls. All these features are sealed by the Stratum 1 pool (see below). Pottery dating to the late Byzantine and Early Islamic period was found in the intrusions of this phase. The walls positioned at the southern (4246) and northern (4244) ends of the excavation, abut the Western Wall and Building F. Furthermore, the plaster of the Stratum 1 pool abuts the upper part of both walls, signifying that they served as outer walls of the later pool (see below).

The remains of a large plastered pool were discovered directly beneath the modern floor of the current prayer area housed beneath Wilson’s Arch (Fig. 9). This pool was partially excavated and discussed when small excavations were conducted in the southern end of the area (Bahat 1993: 38). In the current excavation, two phases of this pool were discerned. The plaster of the pool coats the pier of the arch and the Western Wall, where the negative of the plaster can be distinguished. Following this negative, the depth of the pool can be estimated as approximately 3.2 m. With the pool covering the entire area under the arch (15 × 13 m), the approximate volume of water that it could have held is 624,000 litres, although it is likely that it would not have been filled to capacity. The foundation of the earlier pool was built of large boulders and hard, concrete-like mortar, coated with two layers of thick, pinkish plaster. In a second stage, for an unknown reason, a layer of hard cement-like mortar covered this plaster and was subsequently coated with another layer of pinkish plaster. No pottery was found within the foundation and plaster of the pool, however organic material was obtained and was set aside for future radiocarbon dating. Early Islamic remains were sealed beneath the pool, meaning that it probably dates to sometime in the Late Islamic period.