Off the headland at Santa Liberata on the Argentario peninsular, north of Rome (map), four large Roman concrete piers (Latin: pilae)have been built in the sea on the western side of the fishpond associated with the Roman villa on the promontory. Two of these blocks are set onto the shoreline while the other two are isolated. During the 2003 ROMACONS season a core sample was taken from the pier closest to the fishpond on its north west corner (view). We realized however, that the further pier, 70 m to the west of the fishpond and in front of the remains of a first-century B.C. villa, was potentially more interesting (view). This structure is exceptionally well preserved, although tilted slightly towards the south east and fractured down the middle (drawing). It is not clear what function the pila served, but given its enormous scale, the need must have been critical—perhaps the base for a wooden pier serving the villa, or a wave breaker to protect the fishpond. The pier measures 8.9, 9.0, 7.7 and 7.6 m along its sides and is still 5.9 m tall. The mass is approximately 420 cubic m. How was this very large free standing block of concrete cast? Its upper surface is now between 0.5 m and 1.5 m below sea level and allowing for settlement and sea level rise, the concrete must have been laid within an inundated form at a depth of over 4.5 m. In order to determine how the concrete varied in quality with the depth of the water it was decided to take a core down through the whole height of the block (view of coring, coring with waves, underwater view, underwater view). The Guardia di Finanza kindly provided a boat as a base for our equipment, compressor, and generator (view). On the 28th-29th September, 2005 a core 6.14 m long and 0.09 m diameter was taken vertically down through the middle of the pier (removal of core from tube, view of core). Unfortunately we did not have sufficient core barrels to reach the bottom of the block, since even after all six of the 1.0-meter long barrels had been fitted and drilled into the concrete we still had not reached the ancient seabed. Nevertheless, the extremely long sample recovered will provide unique data on deep water ancient concreting. Initial examination of the core has revealed, for example, that the proportion of mortar to aggregate seems to be less in the lower and upper third of the block (view of core) than in the central third (detail of core). Further analysis of the components of the mortar, and of its structural characteristics will be carried out in the Italcemementi laboratory in Bergamo.

This exacting task could not have been completed with the continuing support of Italcementi and the generous assistance of Dr. Pamela Gambogi, Paolo Volpe and Archangelo Alessandiri of the Nucleo Operativo SubAcqueo (personnel). We also owe a great debt to members of the Guardia di Finanza, who provided a boat and other important logistical support (boat, crew): Paolo Gennaro, Sergio DiMauro, Romolo Pero, Gianfranco Atzori, Umberto Martini and Enzo Timordido..