Silicate MI can record a variety of magmatic processes that control the evolution of igneous systems. The following subsections briefly list relevant tools and techniques (+ appropriate references) that have been used to study mixing, crystallization, and other phenomena of interest to workers undertaking a study of a magma-hydrothermal system.
Silicate MI may be useful indicators of magma mixing, particularly when a rock has undergone some alteration or extensive crystallization of the groundmass. Under such circumstances, the MI may remain as reliable indicators of melt compositions during magmatic crystallization. Hervig & Dunbar (1992) studied both the Bishop and Bandelier magmatic systems, and used MI as evidence that magma mixing was an important process in producing the zonation within both of these magma bodies. Their data for MI of the Lower Bandelier tuff can be divided into two groups. Matrix glass plots within the less-differentiated group; this is inconsistent with crystal fractionation, wherein the matrix should represent the most differentiated composition. The data were interpreted to indicate mixing between low- and high-Ti melts. In earlier studies, Anderson (1976) and Rhodes et al. (1979) used MI as evidence for mixing events during the formation of andesitic magmas and ocean-floor basalts, respectively. Halsor (1989) studied large clear MI within plagioclase from andesites erupted at Toliman volcano in Guatemala. He concluded that MI were not in equilibrium with either bulk rock or matrix liquid, and that a mixing event caused rapid growth of the plagioclase and entrapment of large MI.
Crystallization and fractionation may be recorded by the compositions of a suite of MI trapped at different times. As opposed to mixing trends, which are linear on trace-element scatter plots, crystallization will result in data trends that are curved for compatible elements. Differentiation may be observed by analysis of MI within a single crystal or a group of phenocrysts. Lu et al. (1992) used trace-element concentrations in MI to calculate the amount of crystal fractionation during crystallization of the Bishop tuff magma. Vaggelli et al. (1993) used changes in MI compositions between cores and rims of clinopyroxene crystals as evidence for crystal fractionation in recent Vesuvius lavas.
Because MI approach constant-volume closed systems, they can be used as experimental pressure vessels for high-temperature petrological experiments. In effect, most homogenization experiments are exercises in experimental petrology. Pressure in the inclusion is a function of temperature (Fig. 4) and can be varied by means of a heating stage or furnace. Ideally, one can determine the solidus and liquidus of magmas by heating naturally crystallized inclusions. Clocchiatti & Massare (1985) used silicate MI within plagioclase as experimental vessels that they could monitor as a function of equilibration temperature. By doing experiments at different temperatures on separate groups of MI, they were able to replicate melt and phenocryst compositions obtained with high-pressure experimental apparata on similar tholeiite compositions. Roedder (1976) presented similar thermometric data on MI from lunar and Hawaiian basalts, though without EPMA analyses of the "run products." Roedder (1972) performed laboratory experiments on MI to estimate magmatic conditions during formation of Ascension Island trachytes.
One of the most recent uses of MI has been the correlation of tonsteins, or bentonized ash-beds within ancient lithologic sequences. Because the volcanic ash has been completely altered to clay, whole rock compositions and glass shard compositions cannot be used for stratigraphic correlations. However, Delano et al. (1994) were able to use MI to differentiate a variety of individual units and to define regional-scale stratigraphic time-planes. Even though they studied Ordovician rocks, the inclusions were still glassy and seemingly unaltered. Economic geologists can use MI to define the stratigraphy of mineralized terrains and to correlate units in highly altered areas. Potentially, MI can be used to study the mass balance of bentonite-style (or any other kind of) alteration or to characterize the volatile concentrations in ancient eruptive sequences (Webster et al. 1994).