Many interpretations of metamorphic parageneses tacitly assume that a free fluid phase was always present during metamorphic equilibration (fluid-present metamorphism), as opposed to the circumstance that, while volatile species are certainly released from minerals during metamorphism, these species may not saturate the grain boundaries to produce free fluid (fluid-absent metamorphism). H2O-activity has meaning for both fluid-present and fluid-absent metamorphism and may be evaluated if an equilibrium may be written involving H2O and the temperature and rock-pressure of equilibration are known. PH2O has only a physical meaning for fluid-present metamorphism.
Although the effects of an aqueous fluid on displacing reactions in P–T space are well known, as are the enhanced kinetic and deformational effects of excess fluid, there is actually little compelling evidence to suggest that the presence of aqueous fluids is common during metamorphism of the middle and lower crust, and the mantle, for the duration of most dynamic processes, including partial melting. Possible exceptions are discontinuous dehydration reactions which affect large compositional ranges of metamorphic rock over narrow pressure–temperature intervals and which could locally produce a free liquid if production rate exceeds escape rate.
Evidence from stable isotopes and fluid inclusions obviously points to the accumulation of free aqueous fluid in definite locations for limited duration, but cannot, unfortunately, be used as evidence for generalized fluid transport during regional or contact metamorphism. Both mineralogical and textural criteria presently used to infer ubiquitous fluid, must be examined in terms of crustal volume and duration of dynamic processes to deduce the presence or absence of free fluid. If volatile components ever saturate grain boundaries, then the fluid-phase may only exist locally and perhaps only for short periods of time. Flow of metamorphic fluid may be focused into narrow regimes that could be quite widely spaced. Fluid-present metamorphism should not be considered as typical of whole crust evolution.