Warren’s Gate was first identified by Charles Warren and Charles Wilson as an entrance into the compound on the Herodian Temple Mount, and has since been the topic of discussion in many studies (Wilson 1880
:30, note ++). It is generally assumed that the gate dates to the Second Temple period, when the Herodian Temple Mount was enlarged. The current gate is ascribed by some researchers to the Roman renovations from the time of Aelia Capitolina (Wilson and Warren 1871
:17; Weksler-Bdolah 2015
:132–133). Most researchers believe that the current gate was renovated between the Umayyad and the Fatimid periods (Burgoyne 1992
:215; Ritmeyer 2006
:34–35; Mazar 2011
:93; Mazar, Shalev and Reuven 2011
:264–265; Bahat 2013
:254). After the gate ceased use, a water reservoir was installed in its space (No. 30 on the British survey map, Warren and Conder 1884
:116 [plan], 119). The western facade of Warren’s Gate was exposed in the 1970s and 1980s (Bahat 2013
:254). Another archaeological excavation was conducted at the front of the gate in the 1990s, in which a monumental flagstone pavement attributed to the Second Temple period was discovered (Bahat 2013
:245–256). Other excavations located near the current site were carried out in the cavities of the Great Causeway on the top of the Street of the Chain directed by D. Bahat, followed by A. Onn (Onn, Weksler-Bdolah and Bar-Nathan 2011; Bahat 2013:28–137).
An enormous cross-shaped underground hall is located south of the excavation site. The vaulted hall supports a Mamluk madrasa structure above it (Bahat 2013
:222–241). The cross-shaped hall was built inside part of a large ancient pool. The pool’s eastern wall was supported for its entire length on the western wall of the Temple Mount. The wall is 36 m long between Wilson’s Arch and Warren’s Gate. Most researchers believe that the pool postdates the Second Temple period and was constructed in the Roman (Bahat 2013
:236–240) or Byzantine periods. In the authors’ opinion, it is possible that the pool dates to as early as the Second Temple period, after the Herodian Temple Mount was enlarged. Its location north of a long Second Temple-period dam, which crosses the Tyropoeon channel, lends credence to this assumption.
Remains belonging to five construction phases in five strata were identified in the current excavation.
Stratum 1. Middle Ages(?)—post-tenth century CE
The pointed vault built north of the northern part of the cross-shaped hall in the Western Wall tunnels, c. 45 m north of Wilson’s Arch, is ascribed to this period. The excavation (length 5 m, width 2.7 m) was conducted between two north–south walls (W51 on the west and W52 on the east; Fig. 3). Soil fill (L1901, depth c. 0.7 m) was excavated; it contained fragments of pottery vessels, the latest of which is a sherd of a glazed bowl that dates to the twelfth or thirteenth century CE (Crusader or Ayyubid period). The date of the vault is not known for certain; many changes and repairs were carried out on the walls (W51, W52) that support the vault. The bottom courses of the walls were founded on earthen fill containing finds from the ninth–tenth centuries CE. Therefore, it was possible to determine that the current vault postdates the tenth century CE.
Stratum 2. Early Islamic period, ninth–tenth centuries CE (Fig. 4)
A wall (W55, width 0.8 m, height 1.3 m) was discovered beneath the earthen fill of L1901 and below the walls of the vault (W51, W52). A narrow staircase of three steps descending from west to east (L1903, preserved width 0.6 m) was built along its northern face. The latest pottery sherds discovered, when the staircase was dismantled, date to the ninth–tenth centuries CE, thus indicating that the staircase is not earlier than this time. The soil fill accumulation above the staircase (L1902) also contained fragments of pottery vessels, the latest of which are from the Early Islamic period (ninth–tenth centuries CE).
Stratum 3. Byzantine period, sixth century CE (Fig. 4, section 1–1)
Soil fill (L1904) was excavated beneath the staircase in L1903 and W55. A stone slab pavement (L1907–L1908) that extended across a small area was discovered below the staircase. The material sealed beneath it (L1911) included fragments of pottery vessels from the fourth–sixth centuries CE, leading us to ascribe its construction to the late Byzantine period (sixth century), at the earliest.
Stratum 4. Roman period, third–fourth centuries CE (Fig. 5)
A system of water channels is attributed to this phase. Channel 1919a was well-constructed and descended from east to west (exposed length 4.3 m, inner width 0.45–0.50 m, depth 1.1–1.2 m). The walls of Channel 1919a were built of ashlars in secondary use and small rough-hewn stones. The inner face of the channel’s walls and its bottom were coated with a layer of thick plaster and it was covered with stone slabs. Channel 1919a was adjoined at an angle of 90° from the south by a slightly narrower, similarly built channel. Channel 1919b (width 0.35 m), which conveyed water from south to north, was also covered with stone slabs. The channels presumably carried water toward the city drainage channel in the Tyropoeon Valley (known in several places beneath the pavement of the Eastern Cardo). The fill that clogged the channels contained fragments of pottery vessels from the third–fifth centuries CE, as well as a Roman provincial coin minted in 100–270 CE (L1919, B3040, IAA 134942). It seems that the channels were blocked and went out of use in the early Byzantine period (fourth or fifth centuries CE).
On either side of Channel 1919a was constructive fill consisting of stones, the tops of which had been leveled (L1921 on the north, L1918 on the south). A narrow section of a plaster pavement (L1920) was discovered south of Level 1918. Fragments of pottery vessels discovered during the dismantling of these stone layers and Floor 1920 facilitated their dating to the Roman period (third–fourth centuries CE). Beneath them were soil fills that contained sherds from the first–third centuries CE.
Stratum 5. Second Temple period, first century CE
Phase 5A. The principal remains are of a monumental stone pavement from the Second Temple period (L29052, Figs. 6, 7), whose elevation was c. 2.5 m lower than that of the Roman channels. Coins from the reign of Agrippa I (41–44 CE; IAA 134943) and from Year Two of the Great Revolt (IAA 134944) were discovered in the fill that had accumulated above the paving stones (L29022). Stone Pavement 29052 (c. 4 × 4 m) consisted of three rows of rectangular stones of dressed hard indigenous limestone (mizzi hilu) with smoothed surfaces. The bottom corner of a single stone was discovered that survived from the northern row of stones. Three stones (length 1.5 m, width 1.1 m, thickness 0.3 m) were preserved in the middle row and two stones (length 1.65 m, width 1.15 m, thickness 0.3 m) were in the southern row. The paving stones were placed lengthwise, side by side. The orientation of the pavers more or less corresponded to the cardinal directions. The upper surface of the stones had been carefully smoothed and their edges were chiseled and straightened so they fit together without mortar. The paving stones were embedded in a poured foundation that was hard as stone (L29051, thickness 0.15–0.20 m). A coin from Year Two of the Great Revolt (67–68 CE) (B290057; IAA 135066), was discovered inside the dark soil fill (L29056) that had accumulated where a stone was missing in the northern row of Pavement 29052. The large paving stones were lifted with a crane and the excavation continued underneath them.
Phase 5B. A plastered water channel that was covered with large stone slabs (L29055, Figs. 8, 9) was revealed under the middle row of paving stones. The channel crossed the excavation area in a straight line and descended from west to east (width c. 0.5 m, average depth c. 0.9 m). Its walls were built of courses of small and medium-sized fieldstones bonded with pale yellow-gray material. The sides of the channel—its walls and bottom—were treated with a thick layer of plaster. Pottery sherds from the first century CE were found in the fill that blocked the channel. To its south and below Foundation 29051 was an earlier level (L29053; 2 × 4 m) of rectangular pavers made of relatively soft limestone (1.0 × 1.5 m, average thickness 0.3 m) with small stones inserted between them. The paving stones were placed on a relatively hard surface made of small pieces of limestone and gray mortar (L29059). A coin from the time of Tiberius (30–31 CE; IAA 135067) was discovered inside the foundation, in a place that was not sealed by a stone paver. It was not possible to determine if Pavement 29053 served as a base/foundation for the upper pavement (L29051) or was used earlier as a pavement level.
Below the pavement stones, which were dismantled, was a pile of collapsed stones. Among those stones were two coins from the time of Tiberius (IAA 135068, 135069).
All the remains discovered from the Second Temple period (Stratum 5) were dated to the first century CE, between the years 30/31 and the destruction in 70 CE.
The remains that were revealed are very likely related to the entrance gate to the Temple Mount, known as Warren’s Gate, located c. 8–13 m east of the excavation area.
The monumental stone pavements from the Second Temple period and the drainage channel flowing from west to east probably belonged to a street or public space that was formed where the streets converged in front of Warren’s Gate. The elevation of the upper pavement (c. 724 m asl) and that of the lower pavement (c. 723.5 m asl) were about c. 3 m lower than the level of the Herodian pavement in front of Warren’s Gate (726.7 m asl). The short distance and relatively large elevation differential between the paved section discovered now and the opening of the gate led us to suggest that the pavement exposed was part of a public plaza that extended west of the gate, from which people ascended stairs to the other side of the gate. The staircase that led up to the gate is not known.
In the Roman period, the surface was raised and a new drainage system was installed that conformed to the orthogonal urban layout. The area now drained westward, probably toward the city drainage channel installed beneath the eastern Roman Cardo.
During the Byzantine period, the surface was raised once again and was paved. It was not possible to determine if the section of pavement that was discovered belonged to a street or an open area.
In the Early Islamic period, the surface in the excavation site was higher than the level of Warren’s Gate. The steps from this period descended to the level of the gate.
The pointed arched structure of Stratum 1 was probably part of the underground support structures that were meant to raise the adjacent surface adjoining the western wall of the Temple Mount.