Virtually every conceivable model to explain the internal evolution of granitic pegmatites had been proposed by the 1920s. Two of these hypotheses have prevailed: (1) the fractional crystallization of flux-bearing granitic melt inward from the margins of the pegmatite body to the center, and (2) the buoyant separation of an aqueous fluid from the silicate melt and its effects on the redistribution of components. A recent model combining aspects of both concepts invokes the formation of a flux-enriched boundary layer of silicate liquid in advance of a crystallization front. Though most of the internal chemical and textural features of pegmatites can now be reconciled, the puzzle of pegmatites is far from solved.