Criteria for the recognition of partial melting

Abstract

Partial melting changes rocks from single phase (solid) to two phase (solid+melt) systems. The bulk viscosity decreases as the melt fraction increases and this effect raises the rate of deformation and heat transfer, as well as causing crustal differentiation. Therefore, it is important to be able to recognise which rocks have partially melted.Macroscopic textures provide the simplest criteria for recognising partial melting. Melting and deformation are generally synchronous, and when the melt fraction retained is low (<20%) metatexite migmatites are formed. Typically, these are morphologically complex because the melt fraction is squeezed out of the deforming matrix and collects in whatever dilatant sites are present. The presence of melanosome layers and patches provides the best evidence of where the melt formed, and the leucosomes where it collected. Diatexite migmatites can be easily recognised by the presence of a flow foliation, schlieren, enclaves and vein like leucosomes, and are evidence of a high melt fraction and pervasive partial melting. For the unusual case of melting without synchronous deformation, rounded neosome patches containing both the melt and solid fractions of the melt-producing reaction develop and, as the degree of melting increases these enlarge, to form diatexite migmatites. In both cases the characteristic feature is an increased grainsize and loss of pre-migmatization structures. Migmatite textures are robust, they survive later deformation well.Microscopic textures such as: (1) thin films of quartz, plagioclase and K-feldspar along brain boundaries that represent crystallized melt and, (2) melt-solid reaction textures, also provide good criteria for identifying partially melted rocks. However, these textures are fragile and easily destroyed by deformation. The identification of mineral assemblages from which melt-forming reactions can be inferred is another reliable critera for recognising partial melting, but post-migmatization rehydration in granulite terranes can destroy this evidence.Whole rock geochemistry can be used to model the partial melting process, but problems in identifying the palaeosome and an unmodified melt compositions can restrict its application. However, whole rock geochemistry coupled with good field based control, can be used to deduce what processes have occurred in a terrane where the rocks have partially melted.Variations in field appearance, texture and composition are, in large part a consequence of whether, or not, and when, the melt-fraction separated from the solid fraction.