Abstract:
Imperfections in the perfect fractional melting (PFM) process can ameliorate the drastic “stripping” of incompatible trace elements from the residue which characterises ideal PFM, and the consequent catastrophic collapse of the relative concentrations of these elements in all further aliquots of melt which may be extracted. Once the numerical value of the mass fraction of partial melt extracted in PFM exceeds the numerical value of the crystal-liquid distribution coefficient for an element, then reversed discrimination (reversed, that is, relative to the initial concentrations in the system undergoing PFM) appears in all further aliquots of melt extracted, i.e. the relative concentration of an incompatible element in the next drop of liquid becomes lower than that of a slightly more compatible element, contrary to the effect of equilibrium (batch) melting. However, in PFM there is only an extremely narrow range of (very low) values of the melt fraction extracted within which reversed discrimination occurs in the next drop of liquid to be extracted while the relative concentrations of both highly incompatible elements are still high relative to the source region. Thereafter, as melting proceeds, these relative concentrations collapse to extremely low values. Subdued discrimination, in which the relative concentration of the more incompatible element is lower than expected (from the equilibrium melting equation) relative to that of a more compatible element, given the ratios of these two elements in the source material, is only achieved while the relative concentrations of the two elements are still high, but the range of mass fraction of previous and current melt extraction permitted are again extremely small — necessarily lower even than those required to produce reversed discrimination with high relative concentrations.