Archaeology is like reading a good detective story; you think you've unravelled the mystery, then a new clue appears at the turn of a page. Based on a striking similarity in the division of the northern and the southern wall into niches in the two sections being excavated, we assumed that the new apodyterium extended over the whole north side of the baths. Further exposure of their floor levels this week proved, however, that we were in reality dealing with two distinct rooms.

Johan Claeys (KULeuven) and Mustafa Kiremitçi (Edirne University) spent the first three days of the week excavating the northwest trench and the last two on winterizing the site and conservation. They cleared the part of the apodyterium's western extension where the large marble floor slabs had been removed. Coins of emperors Arcadius and Honorius (both A.D. 395-408) found in the benches date them to the fourth century, as do three small fragments of a wine jug decorated with a grape harvest scene encountered behind the northern bench. These discoveries suggest that the current arrangement of the apodyterium benches, constructed on top of an already existing opus sectile floor, took place at a later date than the refurbishment of the building's wall veneer. The new arrangement may have served the frigidarium that was discovered last week in the eastern excavation sector. The materials from the benches are distinct from the finds made above floor level, ceramics dating to the fifth to seventh centuries. A coin of the emperor Heraclius (A.D. 610-640) was found close to the floor level, suggesting the room was still in use until at least the early seventh century.

Most of the eastern floor section in the larger apodyterium was still covered with several panels of opus sectile, though it also showed later repairs. In front of the entrance to the private part of caldarium I (see August 17-23), two large inscription fragments were re-used in the floor. Though they clearly originate from the same inscription, the text fragments do not represent a continuous text. The inscription appears to belong to a previously found dedication to the emperors Marcus Aurelius and Lucius Verus in the spring of A.D. 165.

The eastern part of the northern extension of caldarium II was also excavated to its floor level and proved to be a corridor. The floor, completely covered with large white marble slabs, was well preserved in this part of the corridor (contrary to its western half exposed in 2002). A branch of the water channel (see August 17-23) opens here on top of the floor and is covered with two small marble slabs. The function of this channel was apparently to supply water for cleaning the floors of the apodyterium and the corridor leading to caldarium II. This water channel along the apodyterium's south wall came from the west wall of the room separating it from the large central shaft of the baths (see introduction). As the channel is 3.5 m above the shaft's floor level, the water may have come from a pipe collecting rain water from the roof. Within a well-preserved and covered part of this channel, some glass finds--the head and neck of a bottle, fragments of a lamp, and a glass--dated to the seventh century.

Markku Corremans (KULeuven) and Erdal Ünal (Eskisehir University) also reached the floor level in the eastern section of the baths. Here the floor was composed of white marble slabs (approximately 1.1 by 0.4 m). These slabs were present in both the rectangular niche of the south wall and the curved niche in the room's east wall. In the central part of the room, we found a basin surrounded by a brick wall. We believe this basin, built of bricks covered with marble veneer, was a pool. This room was not part of the apodyterium as we previously assumed (see August 17-23). This means there is probably a dividing wall in the unexcavated area between both trenches. We believe this was a cold water pool--we've found no evidence of a hypocaustum system--making the room a second frigidarium. The apodyterium must have been connected to this room.

Large amounts of marble revetment slabs, mainly composed of green-white cipollino from Euboia, white and purple veined pavonazetto from Dokimeion, have been found in the new frigidarium. Also, for the first time at Sagalassos, small profiles in rosso antico (red marble from Cape Matapan in the Peloponnesos) have been found. We also discovered pieces of marble statuary including a foot, a lower leg fragment, a hand holding a staff, and a piece of hair.

Excavation has thus shown that the bath complex, along its northern edge from west to east, was composed of a tepidarium connected with caldarium I, a large service shaft connecting the various floor levels of an apodyterium and a smaller frigidarium, and the northern room with the six ashlar piers. At the end of the week, the floors of the newly excavated rooms were covered with geotextile and filled with pumice and sand.