Introduction
The microstructural analysis of (former) partially melted rocks has a long history (reviewed by Vernon, 2004 and Sawyer, 2008) and has been particularly focused on the crystallization of igneous rocks (e.g., Vernon, 2010, 2011) and on the relationships between crystallization and deformation of cooling magmas or partially solidified igneous rocks (e.g., Vernon 1999; Holness, 2010). Melting and crystallization have also been investigated by experimental approaches, using both natural samples (e.g., Mehnert et al., 1973; Büsch et al., 1974; Fenn, 1977) and synthetic analogue materials (e.g., Means, 1989, Means and Park, 1994). However, the microstructural aspects of partial melting that can be observed in the common source rocks of granites (i.e., migmatites and granulites), and especially in the residual rock domains that are left after melt extraction (e.g., melanosomes), have only become a subject of research during the last three decades. This is due in part to the intrinsic difficulty of studying systems where the melt phase is not present anymore, and where the post-melting history may have overprinted part of, or all the indications of its former presence. Along with reaction textures that are indicative of melting, the microstructural evidence of melt in migmatites has been mainly studied, and looked for, in the intergranular melt films, layers and pools that characterize an anatectic system (see the recent reviews by Sawyer, 2008, Holness et al., 2011 and Vernon 2011). More recently, a new object of study – melt inclusions - is gaining attention as a way of complementing the more traditional and well-established approaches.
In this contribution we review the main outcomes of almost 15 years of research on melt inclusions in anatectic rocks, and discuss the main outlooks, describing first the glass inclusions in minerals from enclaves in peraluminous felsic lavas, and then the melt inclusions and nanogranites in garnet and other peritectic minerals from regional migmatites and granulites. Their very small size (<25 µm) has probably caused inclusions to go almost unnoticed in the past, but nowadays the spatial resolution of imaging and analytical tools makes their characterization possible in terms of both microstructures and composition.
Although the bulk chemical composition of the melt inclusions is of paramount importance because they probably represent the best examples of anatectic melts occurring in Nature, the paper is focused mainly on the microstructural aspects, the main topic of this Special Volume. In this sense it represents an expansion of the concepts presented in Cesare (2008).