Abstract The ability of Rcrust software to conduct path-dependent phase equilibrium modelling with automated changing bulk compositions allows for a phase equilibrium approach to investigate an array of source controls for their effect on the bulk compositions of melts produced by sequential melting events. The following source controls of the rock system are considered: (1) initial magnesium and iron content; (2) initial sodium and calcium content; (3) pressure–temperature path followed by the system; and (4) threshold by which melt extractions in the system are triggered. These source controls are investigated in a water-restricted system and a water-in-excess system. The permutation of these cases resulted in 128 different modelled pressure–temperature bulk composition paths investigating the melting of an average pelite composition. The resultant melt compositions are compared to that of a natural granite dataset and provide a good fit for the incompatible elements Na 2 O and K 2 O with the allowance that granites most likely form as magmas consisting of melt and ferromagnesian-rich crystals. The fluid state of the system is shown to have the strongest control on melt compositions, with the pressure–temperature path having subordinate control on the volume and composition of melts produced.

Abstract The Buddusò Pluton in NE Sardinia (Italy) is a normally zoned intrusion composed of three units with chemical composition ranging from hornblende-bearing tonalites (SiO 2 ∼ 65 wt%) to leucocratic monzogranites (SiO 2 ∼ 76 wt%). Zircon crystals in the pluton are dated at 292.2 ± 0.7 Ma and have ε Hf values ranging from −4 to −8, with no systematic differences observed between the units. The pluton, which is isotopically homogeneous at the whole-rock scale in terms of Sr and Nd isotopes, shows textural evidence indicating local crystal–melt segregation. In this paper, we have implemented a novel approach based on path-dependent phase-equilibria modelling to test the hypothesis that the internal chemical variability of the pluton was generated by crystallization differentiation of a homogeneous parental magma. Our modelling indicates that this hypothesis is valid if the mechanism by which this occurs is compaction in a rheologically locked crystal-rich magma and if the separation occurs at 0.3 GPa from a tonalitic magma with water content >2 wt%. Finally, a subset of the magmatic enclaves in the pluton are considered to be autoliths, formed by the disruption of the compacted crystal mush and interaction between these cumulates and the felsic melt.