[2] The construction and design of the dam was related by the East Roman historian Procopius around 560 AD in his treatise on the architectural achievements of the Justinian I era (De Aedificiis II.3).
[3] Procopius makes two essential points clear: first, the dam had a curved plan in order to withstand the water pressure, and did not follow a more or less curved line just because this was where the most solid bedrock was found, and secondly, the thrust of the water was not contained by the sheer weight of the structure (as in gravity dams), but transferred by the abutments to the wing walls of the gorge through the curvature of the lying arch.
Procopius writes: At a place about forty feet removed from the outer fortifications of the city, between the two cliffs between which the river runs, he constructed a barrier of proper thickness and height.
Field work on the ground by the German scholar Günther Garbrecht in the late 1980s has raised some doubts whether Procopius's account are to be interpreted as referring to an arch dam.
[5] By his own admission, however, his observations in situ fell short of a systematic hydrological and topographical field survey which he urged in view of the continuing deterioration of the ancient site.