The Byzantine Mill and Workshop Complex in Terrace House 2

A cooperation between the OeAI and the Institute for the Study of Ancient Culture of the Austrian Academy of Sciences as well as the Römisch-Germanischen Zentralmuseum Mainz has made possible the investigation of one of the most important mill- and workshop-complexes of the early Middle Ages. The project, which should last for three years, is essentially sponsored by the Leibniz Association.

Archaeology

Within the ruins of Terrace House 2, lying in debris and ashes since the third quarter of the 3rd century A.D. – an earthquake had destroyed the once lavishly equipped residential complex located in the centre of the city – a number of mills and workshops were set up in the 6th/7th century A.D. These were connected to each other by a conduit which took advantage of the slope of the Bülbüldağ. A total of eight water wheels provided an equal number of production units with energy. The already completed and published research on the Masonry Saw driven by water in Room WT2 allows the expectation that the other hydromechanical operations also conformed to the latest technological standards of the time. It is clear that the numerous large millstones which are preserved were driven by the available water power above vertical water wheels and a gearing mechanism.

The following questions regarding the history of technology as well as environmental history and economic history form the core of the interdisciplinary research:

How were the individual water wheels, with their accompanying machines, constructed?
What output did they provide?
Can technical alterations be comprehended?
How long were the mills and workshops in operation?
Which relative-chronological statements can be deduced from the structural finds?
Which millstone types were used?
Where were the millstones produced?
How many people could be provided with flour using the early industrial machine?

Survey

At the i3mainz the geometric documentation, as well as work on the presentation of the collected and generated data, was carried out using appropriate analogue and digital media. During the recording, in addition to contact-free survey methods, factors such as height information, colour values and the information content of the data played a crucial role. All of these factors resulted in the implementation of a terrestrial 3-D-laser scanner in combination with a digital single lens reflex camera. Using these means, a coloured 3-D-scatter diagram of the entire mill complex was created as a spatial data foundation for interpretations, reconstructions and visualisations.

Contact

Anja Cramer, i3mainz

Geology

The millstones consist of ignimbrite, hardened pyroclastic flows (›melted tuff‹). They are composed of a basic mass of ashes of differing coloration, with clayey, pumice lapilli weathered by hydrothermal alteration and feldspar as well as a variety of xenolith fragments.

In order to analyse their origin, drill cores from 20 representative millstones were taken. For the petrographic examination, in addition to rock- and crystal inclusions, characteristics such as devitrification, hydrothermal alteration, alterations to the clay minerals and zeolites are important.

In order to identify the origin, the geochemistry of the mafic foreign stone inclusions and of the phenocrysts and xenocrysts will be determined by means of electron beam-microprobe and laser-ICP-MS, and compared to the body of data of mafic lava from the region of western Turkey and the Aegean. In the case of positive findings, walking through the territory will be required in order on the one hand to gather comparative material and to verify the results, and on the other hand in these circumstances to localise the quarries.

Contact

Tatjana M. Gluhak, Institut für Geowissenschaften der Johannes Gutenberg-Universität Mainz

Geology: Sinter-Analyses

The majority of the conduits show deposits of calc-sinter. Four samples were taken from different locations, which by means of a variety of analytical methods should provide information regarding the usage of the water channels, the paleo-clima and earthquakes. The sinter chemistry and the microstructure will be examined with the aid of μ-RFA, REM, EBSD and X-ray tomography. The sinter samples from Ephesos have in general a high porosity and are therefore hard to interpret. At present the focus lies on the chemistry of the samples, in order to be able to assign the Terrace House sinter probably to the Değirmendere aqueduct or also to the Pollio aqueduct. In addition an attempt is also being made to establish the minimum duration of usage of the water channels.