Roman Baths: July 13-17, 2008

View of the gap in the "restored" foundations of the sixth century: the mortared rubble fill 5 (sixth century has "imploded" above, where it was 2.10 m wide and is filled in with debris, most probably originating from the foundation trenches' soxth-century fill. The smaller rubble wall is below (width only 1.45 m).

During the second week of the campaign, the team working in the Roman Baths was split into two groups, with RB 1 supervised by Johan Claeys (K.U.Leuven) and Serkan Demir (Trakya University, Edirne) focused on researching the southern facade. As there were strong indications that this wall had been repaired in Late Antiquity, the group dug two sounding along the southern wall of frigidarium 1. These test trenches will give us information about the original building phase of the terraces supporting the huge bath complex on the steep slope east of the Lower Agora.

The first sounding is immediately south of an apparent fissure in the southern facade of the frigidarium and is aimed at establishing how the Sagalassians coped with the substantial structural damage here caused by the earthquake ca. A.D. 500. The fissure starts in the rows of ashlars above the euthyntheria (the projecting level corresponding more or less with the past walking level). The second sounding adjoins the southwestern ashlar pillar of same room. It should help us gain knowledge about the structural layout of these enormous pillars at the projecting corner of the bath complex.

The first sounding has already yielded some remarkable results. It soon became obvious that the crack in the wall went all the way down into the foundation support of the outside wall and that most damaged had occurred in the foundation layer of at least ca. 5 m deep and consisting of rougher limestone ashlars than those of the visible parts of the wall. At 1.10 m below the euthyntheria, at least five rows of ashlars with a total height of 3.55 m had disappeared over almost the total length (2.10 m) of the sounding trench (the gap is wider above than below). Probably only the fact that the corner foundations of frigidarium 1 stayed in place over the total height of the building kept the upper wall layers above the gap in place, but made them crack. For safety reasons, the test sounding was stopped at a depth of 4.50 m below the original surface. Apparently, during the sixth century a huge effort had been put in reconstructing this part of the foundations, whereby mortared rubble with fragmented spolia (reused architectural fragments), slightly protruding two to three times from the rest of the original ashlar wall, replaced the missing ashlars. This mortared rubble fill is 2.10 m wide above and 1.45 m below, but continues deeper. Yet, as its upper part is located 1.45 m below the level of the uppermost course--this difference in height being filled in with debris--it is clear that the repairs did not hold the external pressure everywhere, suggesting the presence of another not yet excavated vaulted room below frigidarium 1. In fact, as the mosaic floor along the southern inner all of frigidarium 1 is still intact, the collapse inside the gap in the sixth century mortared rubble did not come from above, but from outside, as otherwise the mosaic floor would have sunk as well, and therefore cannot have been the result of the later sixth-early seventh-century killer quake. This new damage must have occurred at some point between the two earthquakes.

The ceramic material that we could associate with the deepest layer (= a refill after the restoration following the first earthquake took place) has been assigned to the sixth century A.D., indicating the major role that the baths still fulfilled then. Little imagination is necessary to foresee that this part of the wall would remain the weakest link in the whole construction, eventually giving away under the pressure from the foundation trench's fill. It also explains why, when the "big one' struck the city around A.D. 540-620, most probably ca. A.D. 590, the south wall of the baths collapsed much more than the other outer walls.

As already suggested, the depth of this terrace/facade wall raises new questions about the possible presence of underground vaults beneath this part of the building. The ground floor of the Roman Baths is not yet completely documented, because of (partially) collapsed corridors that have prevented further research. The fact that a section of the repaired foundations seems to have given away under pressure from outside, seems to corroborate this hypothesis. As most of the vaulted rooms at the ground level of the basis had as only function to support an artificial terrace, the fact that rubble filling the inside of the supposed vault below the south end of frigidarium 1 should therefore not come as a surprise. After the first earthquake, the northwest corner room of the baths, previously accessible from outside and inside, was blocked off completely and its debris--reaching almost to the ceiling--was never removed. On the other hand, the fact that the floor of the large frigidarium did not give away, when the huge weight of the collapsed ceilings fell down on it--a phenomenon we encountered in other rooms that had substructures (hypocaust and/or underground vaults)--suggests that the supposed vault carrying frigidarium 1's southern part, must have remained standing.

Based on our experiences in the first sounding, the necessary preparations were taken to enable the team to reach an ever greater depth in the second sounding just outside the southwester corner pillar of frigidarium 1. The initially planned working area will be extended in all directions, creating a larger, stable terrace from where one can continue the work.