Potter's Quarter: August 6-13, 2007

Study of the Water Infrastructure

Since 1998, Dr. Femke Martens has undertaken a program of test soundings throughout the urban area of Sagalassos to investigate the ancient city's infrastructure, its streets and water supply. The hydraulic devices and stretches of water drainage and supply channels exposed in these soundings so far form part of an elaborate distribution network that was supplied by at least 5 aqueducts entering the town from the east and west. Since 2002, the research on the urban infrastructure has greatly benefited from the geophysical survey at the site, carried out using magnetometry and georadar. Small-scale excavations, however, remain essential to study the evolution of the waterworks in their chronological framework.

To reconstruct the organization of the water-distribution system at Sagalassos, the main and secondary distribution points of the network need to be located and investigated. Surveys of the late 1980s referred to the presence of a main distribution point--where the water was also cleaned by settling tanks or castellum aquae--somewhere northeast of the Potters' Quarter, most likely at the site's northeastern edge. Topographical surveys in this area helped locate a rectangular structure, which may have fulfilled this function. Just west of this building running water from an uncertain source is still tapped for agricultural use.

We carried out a new geophysical search in the area around this building, detecting some anomalies that possibly correspond with water channels. These headed from the direction of the rectangular structure toward the south, next to another, west-east oriented system. The aims of this season's test soundings were to clarify the function the rectangular building by studying its relation to the surrounding elements of the water network. In addition we wanted to shed light on the chronology of the layout of the Upper Eastern Aqueduct, which could not be dated so far, and to determine the exact position where this main source of water entered town.

Based on the results of the geophysical survey, we have opened two trenches so far. The first determined the position where the eastern aqueduct entered the town. The channel, which we found 0.85 m below the surface, was built of mortared rubble. It had an interior width of ca. 0.50 m and a height of ca. 0.70 m. The channel's interior was finished with a mortared floor and the walls were coated with pink hydraulic mortar, as could be reconstructed on the basis of fragments of this wall plaster found within the channel. The level abutting the channel contained pottery of the Julio-Claudian era, but further excavation below the channel will be required to firmly establish a date of this system. The channel followed a contour line, which allowed us to connect it to a stretch of the same channel, which we excavated farther west during the early 1990s ("site F") and to another rock-cut stretch to the east, which was identified by recent cartographical research as the Upper Rock-cut Aqueduct. As far as dating the latter, geomorphological research has established some ten years ago that this aqueduct was a repair and a replacement of an older section that went down during a landslide in historical times. If we can date more exactly the construction of the exposed water channel in the Potters' Quarter, this will also allow us to date exactly or post quem the catastrophic landslide, which took place during the city's heyday.

This latter rock-cut stretch of the Upper Eastern Aqueduct was cleaned, and it proved to have been carved out in the rock to a depth of 1.80 m and to a width of ca. 0.51 m. Its mortared floor was still intact and its walls appear to have been plastered to a level of 1.05 m. The excavation allowed to connect these three stretches running, however, 40 m below the level of the presumed castellum aquae, making it unlikely that both were connected.

We opened a second trench to investigate the relationship of the north-south oriented anomalies with the presumed castellum aquae. In this trench a stretch of a north-south oriented water channel composed of prefabricated terra-cotta tiles (interior width: 7.3 cm, interior height: 9 cm) was found at about 1.10 m below the surface. The channel was laid out upon and ancient scree slope. Further excavation will allow us to date this system. We will open a third trench to trace this and the other north-south oriented anomalies in the direction of the castellum aquae. We want to determine whether or not they were connected to the latter, or to establish and date their real origin.