• anatectic rocks;
  • contact aureole;
  • deformation mechanisms;
  • melt distribution;
  • melt segregation


The grain-scale spatial arrangement of melt in layer-parallel leucosomes in two anatectic rocks from two different contact aureoles located in central Maine, USA, is documented and used to constrain the controls on grain-scale melt localization. The spatial distribution of grain-scale melt is inferred from microstructural criteria for recognition of mineral pseudomorphs after melt and mineral grains of the solid matrix that hosted the melt. In both rocks, feldspar mimics the grain-scale distribution of melt, and quartz is the major constituent of the solid matrix. The feldspar pockets consist of individual feldspar grains or aggregates of feldspar grains that show cuspate outlines. They have low average width/length ratios (0.54 and 0.55, respectively), and are interstitial between more rounded and equant (width/length ratios 0.65 for both samples) quartz grains. In two dimensions, the feldspar pockets extend over distances equivalent to multiple quartz grain diameters, possibly forming a connected three-dimensional intergranular network. Both samples show similar mesoscopic structural elements and in both samples the feldspar pockets have a shape-preferred orientation. In one sample, feldspar inferred to replace melt is aligned subparallel to the shape-preferred orientation of quartz, indicating that pre- or syn-anatectic strain controlled the grain-scale distribution of melt. In the other sample, the preferred orientation of feldspar inferred to replace melt is different from the orientations of all other mesoscopic or microscopic structures in the rock, indicating that differential stress controlled grain-scale melt localization. This is probably facilitated by conditions of higher differential stress, which may have promoted microfracturing. Grain-scale melt distribution and inferred melt localization controls give insight into possible grain-scale deformation mechanisms in melt-bearing rocks. Application of these results to the interpretation of deep crustal anatectic rocks suggests that grain-scale melt distribution should be controlled primarily by pre- or syn-anatectic deformation. Feedback relations between melt localization and deformation are to be expected, with important implications for deformation and tectonic evolution of melt-bearing rocks.