formation, with chloritoid breakdown supplying sufficient material to form the large amounts (c. 25 vol%) of staurolite found in the rock. This reaction produces an excess of SiO 2 , which leaves the crystal domain as SiO 2aq and thus caused the formation of the inclusion-free areas in the staurolite and precipitation of quartz in the matrix. In the sillimanite zone, staurolite is consumed forming new garnet. The newly formed garnet has less quartz inclusions than its core due to a proportionally greater consumption of quartz by the second garnet-forming reaction than by the initial, garnet-grade reactions. Textural and thermodynamic data both suggest that inclusions in these porphyroblasts represent leftovers of a preferentially overgrown matrix than coproducts of the porphyroblast-forming reaction.

No abstract

The Cycladic blueschist belt, Greece, is mostly submerged below sea level and regional correlations are difficult to establish. Marbles are widespread within the belt and locally used as marker horizons to subdivide monotonous schist sequences. However, owing to the lack of distinctive petrographic characteristics, the marbles have not been used for island-to-island correlations. This study aims to investigate the potential of Sr-, C- and O-isotope compositions of marbles as a tool for unravelling the litho- and/or tectonostratigraphic relationships across the Cycladic islands, and as a proxy for the time of sediment formation. For this purpose, we have studied metamorphic carbonate rocks from the islands of Tinos, Andros, Syros, Sifnos and Naxos. Identical87Sr/86Sr values for certain marble horizons occurring on Tinos, Andros and Sifnos are interpreted to document coeval regional carbonate precipitation. The87Sr/86Sr values of the apparently least altered samples intersect the seawater curve multiple times within the most likely time interval of original carbonate precipitation (< 240 Ma; as indicated by previously published ion probe U–Pb zircon data) and thus an unequivocal age assignment is not possible. Very broad temporal correlations are possible, but more subtle distinctions are not feasible. On Andros, the overlapping Sr-isotope values of marbles representing the lowest and highest parts of the metamorphic succession are in accordance with a model suggesting isoclinal folding or thrusting of a single horizon, or very fast sedimentation. In contrast, distinct87Sr/86Sr values for samples from Tinos, representing different levels of the metamorphic succession, suggest that these rocks represent a temporal succession and not the tectonic repetition of a single horizon. Based on Sr-, O- and C-isotope characteristics alone the time equivalence of marbles occurring on different islands could not be documented unambiguously. However, by using various combinations of these parameters, some occurrences can be discriminated from the overall sample population. The new data further accentuate the general potential of coupled Sr-, C- and O-isotope characteristics for identification of archaeological provenance and complement existing datasets for Aegean marbles.

No abstract