COMPARISON OF THE ACCESSORY MINERAL CONTENT FROM THE ANCIENT MARBLE QUARRIES OF MT. PENTELIKON (ATTICA, GREECE) AND PROCONNESUS (MARMARA ISLAND, TURKEY): METAMORPHIC PETROLOGY APPLIED TO PROVENANCE

The current standard for marble provenance studies involves a combination of stable isotopes δ¹³C and δ¹⁸O and the maximum grain size (MGS) of an artifact sample. When these variables are plotted against one another, it is expected that marble of the same region will plot into a distinctive field. When this field overlaps with another, more variables are required for characterization. While it is nearly impossible to definitively assign one quarry location to an unknown artifact sample, many quarry regions can be excluded by quantitative methods. In this study we analyzed the accessory mineral content of 30 samples from the quarries of Proconnesus (Marmara Island, Turkey) and 10 samples from a quarry located on Mt. Pentelikon (Attica, Greece) using an electron microprobe. This was done to further develop a method that will assist Classical archaeology in accurate provenance and to provide mineralogical information relating to the Mediterranean’s metamorphic history.

The presence or absence of certain accessory minerals is dependent upon the composition of the protolith the sample (limestone or dolostone) and its metamorphic history. The accessory minerals in marble should be somewhat unique by area. Previous studies using this method have examined about 2-4 samples per quarry, while this study expanded the dataset for statistical purposes.

Marble from Proconnesus was found to be calcite + dolomite with accessory minerals that included pyrite, siderite, apatite, and fluorite. Pentelic marble was calcitic and did not contain dolomite. Some of the major accessory minerals included pyrite, apatite, quartz, and muscovite.

Results were statistically processed with δ¹³C and δ¹⁸O and MGS using discriminant function analysis. Once the mineralogy of these quarries is better understood, artifact samples may be analyzed and their mineral content compared to known quarry content. Combined statistically with other variables, the method adds more certainty to provenance studies.