Phase 1. Apparently, most of the preserved vaulted ceiling of the hall dates from the Ottoman period.
Phase 2. The remains of the original southern perimeter of the Mamluk vaulted hall (Fig. 3: Section 1–1) comprise a foundation wall (W4) and a superstructure: a wall (W2), a pillar, partially built with ashlars (W1) and a narrow, brick-built section (W12), connecting W2 and W1 (Figs. 4, 5). An arched entrance in this wall, which led south to the Daraj el-‘Ein complex, was found blocked (L9). To the north of this wall were a partially excavated pillar (W14) and a large ashlar block (W5), probably the remains of a wall; a pavement made of very glossy flagstones (L6; c. 0.15 m thick; Fig. 6) abutted the pillar on the south and Ashlar 5 on the east. A similar pavement was discovered further south, at the Ohel Yizhak site; the excavators identified it as the original floor of the Mamluk bathhouse changing room. Pavement 6 was significantly robbed, but its bedding—a layer of plaster (L7)—extended southward and abutted W4. Ashlar 5 and Bedding 7 covered the foundation wall (W11) for W4 and the floor (Fig. 7).
Phase 3 (Figs. 8, 9)comprised large-scale remains: a stone pavement (L13), which was largely robbed, and a substantial north–south drainage channel (L16) that ran under the pavement. The channel was constructed of medium-sized stone slabs and was covered by the large stone slabs of Pavement 13. One of the larger these slabs was removed during the excavation to probe within the channel, which was found full of muddy soil (Fig. 10). As no finds were revealed within the channel, its precise date could not be established. No diagnostic finds were found below Pavement 13 either. Based on known parallels of drainage channels and the stratigraphic position of the remains, they can be dated with caution to the Byzantine or Early Islamic period.
Phase 4. Remains of this phase were exposed in a limited area, where the slabs of Pavement 13 were removed, close to Drain Channel 16 (see Fig. 9). At a level 0.15–0.20 m below Pavement 13 was another, earlier, stone pavement (L18; Fig. 11), which was cut by Drainage Channel 16. One of the stone slabs of this pavement bore the typical engravings of a Roman gameboard (Fig. 12). It is not clear, however, whether this stone is in situ or in secondary use.
Phase 5(?). A small section of a street pavement (L17; Fig. 13) was discovered to the west of the excavation square and at an elevation nearly 2 m higher than that of Pavement 18 (Phase 4). The street pavement was discovered following the removal of a twentieth-century stone staircase, which led from the modern street to the vaulted hall. It was found under fill layers, on which the stairs were founded; the fills contained modern pottery, glass and plastic fragments. Street Pavement 17 comprised diagonally laid large flagstones and had a clear eastern border, showing that it abutted a stylobate wall that did not survive. These features and the precise location of this pavement suggest that it was part of the Roman-period Eastern Cardo (Weksler-Bdolah and Onn 2017). An identical pavement was exposed just two meters to the south, in the Ohel Yizhak excavation, identified by the excavators as the Roman-Byzantine Cardo (Fig. 14; Barbé and De’adle 2007).
In the absence of numismatic and ceramic finds, the relative dating of Pavements 17 and 18 and their stratigraphic relation remain the most pressing issues in need of explanation. As it seems that Pavement 17 is most probably part of the Roman-Byzantine Cardo, two options are proposed to explain the substantial height difference, depending on whether the stone bearing the incised Roman gameboard in Pavement 18 (see Fig. 12) was discovered in situ or in secondary use:
1. The stone is in situ, meaning that Pavement 18 is contemporary with Pavement 17 of the Roman-period Cardo. In such a case, the substantial height differential between them can be explained by the existence of construction, such as a paved market area, a piazza or a pool, to the east of the Cardo and its eastern stylobate. The existence of a paved street or paved piazza in this part of the city was proposed in the past, based on ambiguous archaeological and ceramic evidence unearthed during excavations on el-Wad street (for a summary and discussion, see Kloner and Bar-Nathan 2017); unfortunately, as in the case of the current excavation, it was not possible to provide an absolute date for the finds. 
2. The stone is in secondary use and was placed in its current location during the construction of Drainage Channel 16 (Phase 3), to serve as part of the channel’s wall, when the channel cut through Pavement 18 (Phase 4). In this case, Pavement 18 is later than Pavement 17, which would explain the height difference.
The excavation at Hammam el-‘Ein adds additional archaeological data to the complex picture of the evolution of one of the major urban arteries of the city, from its relatively early stages up to the late medieval period. The results also contribute to an understanding of the development of the environs of the Cotton Market in the Mamluk quarter, with its numerous public structures.