The Transition from Granite to Pegmatite
There is no known causal relation between aqueous vapor saturation of melt and the formation of giant crystals. Growth of giant crystals in pegmatites requires more rapid transport (mass, if not rate) of Si and Al than an aqueous vapor can provide in the timeframe of cooling. In analogue experiments, a flux-rich melt moves 107 times the mass of Si as aqueous vapor at 107 times the rate of diffusion in hydrous haplogranitic melt at 800oC and 200 MPa. A hydrous flux-rich melt created via constitutional zone refining at crystal-melt interfaces provides the necessary mass transport in pegmatite-forming environments.
There is no known causal relation between aqueous vapor saturation of melt and the mineralogical zonation that is prevalent in many granitic pegmatites. Crystallization is sequential (or oscillatory) in highly undercooled granitic melts, even those of invariant composition, because of (1) differences in the curvature of the liquidus surfaces for quartz and feldspars, and (2) interactions among fluxing components and major melt components that suppress the nucleation of rock-forming minerals.
Empirical evidence indicates that granitic pegmatites become saturated in H2O only toward the end of their consolidation. Fractionation models for pegmatites imply that their source magmas should be H2O-saturated prior to the derivation of pegmatite-forming melts. Resolution of this question, when granitic melts become saturated in H2O and the evidence for this transition (if not pegmatites), remains as one of the most significant problems in igneous petrology today.