Following the results of our previous activities since 2010, two main areas were selected for excavation during the 2013 season, Structure 20 (ST 20 on figs. 2. 3), standing on the NE edge of the plateau and having been identified as a luxurious bathing installations by the previous excavations carried out there in 2011 and 2012, and Structure 19 (ST 19 on fig. 2), being located on a second terrace, some meters inside the plateau.
One of the structures that provided the most intelligible results during the 2010 survey was Structure 20 (ST 20 on figs. 2. 3), standing on the NE edge of the plateau, prominently overlooking the city centre. The results from both the survey season as well as from the excavations carried out in 2011 and 2012 strongly suggest this structure was a bathing installation. The structure consists of several clearly visible rooms. From the S end of the building, a partially rock-cut and partially built water channel brings in water to the structure. The water is collected in a substantial basin, identified by its greyish hydraulic mortar containing charcoal fragments which improved its waterproofing qualities.
In 2013 our activities focussed on the Southern part of the building in order to try to define its southern end and, therefore, the overall plan of the installation. Indeed, it seems that we did find the outer wall of the building. From the huge main hall of the building, excavated in 2011 and 2012, a row of four steps is leading upwards to a substantial doorway, initially measuring 125 cm in width (fig. 4). The lowest step is directly cut into the bedrock on a level of 1129,99 m asl, while the threshold is located on 1130,52 m asl. In a later phase, the stairs as well as the door were narrowed and put on a considerable higher level (fig. 5). In this later and last phase, the door measured only 65 cm in width, while the steps start at 1130,34 m asl and the threshold is at 1130,84 m asl. According to pottery finds from this area, the later phase would correspond to the Late Roman period and this phase of the building was eventually destroyed by the earthquake of 363 CE.
Immediately in front of the door and, therefore, outside the building we discovered a water channel that brought the water into the bathing installation (fig. 6). The channel consists of a series of clay pipes, embedded into a thick layer of the typical greyish mortar and placed into a small construction of two later walls and irregular covering slabs. According to the pottery fragments found in the stone built channel as well as in the mortar embedding the clay pipes, this system was not built before around 100 CE and possibly even later. The last covering slab in front of the threshold is situated at 1130,81 m asl, and can therefore impossibly have been used in the same period as the initial threshold with 1130,52 m asl. It is much more likely that the channel as described functioned with the later steps and threshold. However, an earlier water channel must have existed on a slightly higher level on the bedrock next to the building. Remains of a stone built channel and a rock cut canalisation are still clearly visible although partially destroyed.
Adjacent to the stairway on the E we exposed a small room measuring 4m x 2m in its interior dimensions (fig. 7). The level of this small room was considerably lower than in the adjacent stairway but also in relation to the huge main hall of the building excavated in previous seasons. Although no floor level was discovered, the wall plaster partially discovered in situ and showing a lower limit of 1129,10 m asl probably indicates the initial floor level. Also, a small niche in the W wall of the room and a bench, built directly on the bedrock, in the S wall of the room, could indicate its use as a kind of storage room. Since we know that the building was initially used as a bathing installation, one could tentatively interpret this room as the apodyterium of the bathing installation, i.e. the changing room.
Clearly, the room was reused in a later period. Testifying of this reuse is the considerable amount of pottery, including African Red Slip ware (fig. 8) and a complete Late Roman/ Early Byzantine lamp (fig. 9). In the destruction debris of the room we discovered three fragments of the base and the cloak, and a left knee of a marble statue. It was immediately clear that these fragments belong to the same statue of a young boy carrying a water jar that we found in 2011 and 2012 respectively. Indeed, all parts fit with fragments from previous years (fig. 10) and the upper body of the sculpture, currently in the exhibition „Petra – Wonder in the Desert. In the footsteps of Sheikh Ibrahim“ in Basel (Switzerland), will also fit to these parts. It is planned to have the sculpture restored as far as possible in a coming season, once the upper body has been brought back to Petra.
With the excavations carried out in 2013 the exploration of Structure 20 is completed and its ground plan known, with the exception of the parts that already fall down the cliff on the E part. In order to protect the structures we exposed from 2011 to 2013, all trenches were completely backfilled (fig. 18).
In order to obtain information also for structures that are not built directly on the edge of the plateau, we started exploring Structure 19 (ST 19 on figs. 2. 3), a rectangular building located a few meters more to the inner part of the plateau of Umm al-Biyara. During the 2013 season, the NW corner of the building was excavated, revealing that this part of the building was used for various functional purposes, at least during its last phase.
In the NW corner, a room measuring 2.75m x 4m in its interior dimensions was exposed (fig. 11). Two benches constructed towards the W wall indicate that some activity of preparing or processing must have taken place here, though we were not able to precise the exact nature of these activities. Directly constructed on the bedrock, the floor of the room declines from W to E, with two floor slabs giving access to a door and a corridor-like structure to the E. From this corridor-like structure several more rooms and installations are accessible. Directly N of it is a small rectangular space that in a later phase was reused as an oven (fig. 12). Not only the construction with its heavily damaged stones due to the heating indicate such an use, but also the big amount of bones and some olive pits found in and around the structure within massive layers of fine ashes (fig. 13). Next to the W follows another rectangular room. The clearest evidence for its use comes from the two toilet seats that were found in situ. However, as is clearly indicated by the fact that a third opening was deliberately broken off in order to make one of the seats fit the room (fig. 14), these toilet seats were reused. Since we did not find any toilet seats related to the latrine excavated in 2011 and belonging to Structure 20, it seems as if the toilet seats discovered in 2013 within Structure 19 could initially have belonged to Structure 20.
Clearly, the later phase of Structure 19, including the oven and probably also the construction of the toilet, can be dated to the Late Roman period, as indicated by a lot of pottery, including a complete Late Roman/ Early Byzantine lamp (fig. 15) and many glass fragments. However, the long use of the structure is indicated by earlier pottery, including a complete unguentarium (fig. 16), found in the channel that brought water to the toilet. This channel runs underneath and behind the oven and became visible only when we dismantled the oven in order to find the initial construction of the area (fig. 17).