The excavation was conducted in an underground tunnel (the Exhibition Tunnel) near the entrance to the City of David Visitor’s Center and below Ma‘alot ‘Ir David street. The purpose of the excavation was to connect the finds from the excavation of the ‘Palace’, which took place about a decade ago (Mazar 2009), with excavations conducted in recent years at the entrance to the visitor’s center (Uziel 2013; Dan-Goor 2018). The tunnel was excavated from both ends simultaneously. The tunnel’s route curves and is rectangular in section (width c. 5.0 m, height c. 3.5 m). Four strata were identified above the natural bedrock (VII, VI, Post-VI, V; the strata were named according to parallel strata in the excavations at the entrance to the visitor’s center), dating from the Hellenistic to the Early Byzantine periods. Most of the remains unearthed in the excavation are continuations of those discovered while excavating the entrance to the visitor’s center (Uziel 2013; Dan-Goor 2018).
 
Stratum VII (Hellenistic period)
Stepped rock-hewn terraces found throughout the current excavation area attest to the presence of a quarry at the site (Figs. 2, 3). The earliest pottery, which was found on the natural bedrock inside the quarry, dates from the Hellenistic and Early Roman periods. The western part of the quarry was excavated previously and yielded similar finds (Dan-Goor and Tchekhanovets 2020). The quarrying ceased before plastered installations were constructed in it during the Early Roman period (Stratum VI).
 
Stratum VI (Early Roman period)
Plastered installations, which are evidently water-related, were found above the remains of the quarry (Fig. 4). A plaster floor (L323) in the southern part of the excavation was the continuation of a plastered installation discovered in the excavations at the entrance to the visitor’s center (L229; min. length 12 m). A stone foundation built directly on top of the bedrock was revealed beneath the installation’s plaster floor. The northern wall of the installation’s was excavated in three sections (W236, W317, W333) due to the casting of concrete pillars that divided it into three. The installation’s northwestern corner was not found, and its southern part extended beyond the limits of the excavation. Two walls that formed a corner (W332, W327) and were contemporaneous with Installation 323 were also found in the southern part of the area.
The northern part of the excavation area contained a square, plastered installation (L337; 2.0 × 2.0 m, depth c. 4.5 m; Figs. 4, 5). Its lower part was hewn into the bedrock, and its upper part was built. The installation continued northward to a kind of entrance area that was blocked with a Late Roman-period wall. The installation yielded copious amounts of pottery dating from the Early Roman period (first century BCE–70 CE). The finds from the installation are similar to those from a reservoir (L173) excavated at the entrance to the visitor’s center.
Another corner of a rock-hewn installation (L131) was unearthed in the eastern part of the excavation area, but because it was located on the outer fringes of the excavation, it could not be excavated in full, or to any depth.
 
Post-Stratum VI (Roman period)
Walls and installations were built on top of the layer containing the Early Roman installations; the date of their construction is uncertain at present, but the previous installations were probably abandoned after 70 CE. A wall built above and into Floor 323, and thus canceling it, was divided into three sections (W232, W316, W334; Fig. 6) by the concrete pillars. To its north was a tabun (L217).
An installation (L326) built beside the upper stones of an earlier wall (W327) was uncovered in the eastern part of the excavation area. The installation was coated with several plaster layers. Its western part was cut by a Late Roman wall.
This stratum represents a transition period between the time when the installations were rendered obsolete, probably because far fewer water installations were required after the destruction of Jerusalem, and the period of renewed building at the top of the hill in the Late Roman period (Stratum V).
 
Stratum V (Late Roman and Early Byzantine periods)
The continuation of a magnificent residential building (the ‘mansion’; Dan-Goor and Tchekhanovets 2020) was excavated, parts of which were uncovered in the three previous excavations mentioned above. As the building’s walls were abutted by two habitation levels, the stratum was divided into two sub-strata: Vb and Va.
 
Stratum Vb (Late Roman period). The mansion’s construction was adapted to the topography at the top of the hill. In places where the bedrock was at a higher elevation, the walls were founded on top of it (e.g., W310, W313, W324; Fig. 7), and in places where the bedrock was relatively low they were built on top of earlier walls (e.g., W149 built on W185 from the Early Roman period) or on foundation walls (e.g., W213, built on foundation W227). Two plaster floors (L322, L335) uncovered in two rooms were attributed to the mansion’s earlier phase, dated according to the pottery to the Late Roman period.
 
Stratum Va (Late Roman–Early Byzantine periods). Previous excavations uncovered a few impressive mosaics (Fig. 8) in several of the mansion’s rooms, above the Stratum Vb floors. Some were white (L166, L171; Dan-Goor 2018), and others were both white and decorative (Uziel 2013: Fig. 3). The current excavation revealed a section of a mosaic floor (1.2 × 2.0 m), which was part white and part patterned in three colors (Figs. 9, 10). The decorative mosaic extends eastward beyond the limits of the excavation. Based on the finds, this stratum can be dated to the Late Roman–Early Byzantine periods (fourth century CE). Other parts of the mansion were excavated in the past (the ‘House of Eusebius’; Macalister and Duncan 1926:89–91; Mazar 2009:87).
 
The excavation inside the tunnel expanded on the archaeological picture that emerged from the excavations at the entrance to the visitor’s center. The impressive mosaic floor in Stratum Va attests to the population’s luxurious lifestyle during the Late Roman–Early Byzantine periods. The Early Roman remains also complete the stratum containing the many water-related installations found in adjacent excavations. The numerous water installations at the top of the hill, in the City of David, attest to the great demand for water for Jerusalem’s many pilgrims, mostly during the major festivals (Dan-Goor 2018). The remains from this period were built directly on top of the natural bedrock, or even hewn into the rock. Widespread construction during the Early Roman period coupled with the Hellenistic quarry obliterated any traces of architectural remains from the Iron Age.