KENCHREAI - 2009

Kenchreai, Koutsongila ridge. E. Korka (Ministry of Culture and Tourism), J. Rife (ASCSA/Vanderbilt University) and K. Kissas (Director, ΚΖ' ΕΠΚΑ), with P. Kasimi (ΚΖ' ΕΠΚΑ), report on the 2009 excavation season on the Koutsongila ridge north of the ancient harbour (Figs 1-2).

In area A, activity concentrated on the wall found in 2007 and the areas to the west and east. A further 4m long stretch was uncovered, continuing south (Fig. 3). The wall rests at the base of a small hill; a dense Early to Middle Roman sherd deposit, representing pottery washed downslope against the wall, was found to continue along its interior (west) face. To the east, a broad depression (ca. 4.5m wide) extends from the wall to a vertical face in the bedrock outcrop upon which sits the rectangular building of probable ritual function explored in 2007−2008. This depression may be an important traffic route, although no road surface or wheel ruts were preserved in its relatively soft calcareous bed. The long wall was thus erected to arrest slope erosion and protect the depression from sediment and debris. A passageway or road along the depression would have approached the harbour from the north across the middle of the ridge, and entered the harbour front from behind the complex at the base of the north mole.

Renewed exploration of the area west of the long wall and on top of the hillock uncovered evidence for burial and cremation (trench A-09-02). Although this seems to have been an open area in antiquity, a channel cut across the trench contained a distinct layer of ash mixed with intensely burned human and animal bone, buried haphazardly under rubble − debris from one or more pyres deposited here after multiple cremations and then covered. The area formed part of a larger cremation zone (a smaller secondary deposit of similarly burned bones was found nearby in 2007). The primary position of the cremation pyres is uncertain.

Area B, near the southeast edge of Koutsongila, was the main focus in 2009 (Figs 4-5). The function of the structures uncovered in 2008 was clarified. In the area of the putative Early to Middle Roman villa, excavation of the well was completed to modern sea level (ca. 7.4m deep), though in antiquity it clearly went deeper. The well was filled with structural and domestic debris from a fifth-century cleaning operation following the destruction and disuse of the villa: this may be associated with the erection of a major structure nearby.

Excavation immediately to the south revealed a large octagonal structure (14.25m maximum interior width) erected in an apparently open area in late antiquity (Fig. 6). This had a massive outer wall in mixed mortar, rubble and brick, founded on platforms set into the bedrock on the north side and in shallow footing trenches on the south. Inside, a series of angular piers divided an outer ambulatory from an interior chamber. One wide entry was found in the north wall: others may have existed, but the evidence is lost. Numerous fragments of blueish-grey schistose marble with acanthus leaves in relief show that the interior walls were richly finished. The floor of the ambulatory was in places paved with simple tiles but elsewhere consisted of hard-packed earth. In contrast, the central chamber had a mosaic pavement of interlocking geometric patterns with diamonds, stars and circles in black, white and red. A building of this scale with such an impressive design and decor, located prominently above the harbour, must have dominated the landscape. Associated artefacts and stratigraphy provide a broad outline of the building’s history. Along the exterior base of the north wall, a narrow trench opened into a small elliptical pit containing numerous intact lamps and coins deliberately placed (Fig. 7). This foundation deposit dates to the early fifth century. When the Octagon fell into disuse, sediment collected over the decaying floors, and transients made fires and built a screen wall inside the building (Fig. 8). Finds from the thin deposits over the floor do not postdate the mid sixth century. Around that time, the building collapsed in a catastrophic event that toppled the walls into one massive layer of structural debris over the whole area. Few artefacts relate to the Octagon’s use, leaving its design and surroundings as the best evidence for its function. Comparable octagonal buildings were erected in late antiquity as palace-halls (e.g. Thessaloniki), baptisteries (e.g. Ravenna), Imperial mausolea (e.g. Romanuliana) and martyria (e.g. Phrygian Hierapolis). No walls were found to indicate that the Octagon formed part of a larger architectural plan, making identification with a palatial or ecclesiastical complex unlikely. However, seven fifth- or sixth-century cist graves (G63−G69) of the finest construction were clustered around it and aligned with its walls (Fig. 7). The intentional placing of these distinctive graves suggests that the Octagon had a funerary function, but whether as the burial site of a member of the local élite or a holy figure is uncertain. If so, the Octagon is a rare case in the known architectural repertoire of late antique Greece.

In area C, in the heart of the Roman cemetery, one chamber tomb, tomb 2, was explored (Fig. 9). The rectangular chamber (4.15m x 4.6m) resembles others on Koutsongila: it is rock-cut, has loculi and niches in the walls for inhumation and cremation, a bench on the front wall and an altar on the back, and a vaulted ceiling. The walls preserve painted plaster with polychrome-banded decoration and a schematic garland in red, while the floor was finished with cement. The construction apparently dates to the mid/late first century, as other tombs on the ridge. The chamber was looted at an early date, and the numerous cups, bowls, cookware lids, lamps and a miniature stone altar were scattered. Thereafter, apart from occasional disturbances by robbers, the chamber gradually filled with a combination of disintegrated mortar, plaster and rock that had collapsed from the walls, and washed-in colluvial sediment. Exploration outside the tomb revealed that a building in mortar and rubble with a monumental façade once enclosed the narrow, rock-cut stairway descending to the stomion. No structures or artefacts were found to identify graveside rituals, which must have occurred elsewhere and/or inside the chamber. Finally, wheel ruts running north-south through this area predate the construction of the tomb, and thus give a sense of the traffic along the coast before the development of the cemetery.