Crystallization of the Little Three layered pegmatite-aplite dike, Ramona District, California

Subhorizontally layered pegmatite-aplite bodies are characterized by fine-g rained, sodic to granitic aplite that is usually juxtaposed abruptly above by much coarser-grained, commonly graphic potassic pegmatite. Although well studied, there currently is little concensus as to how such dikes form. Th e Little Three dike near Ramona. California, is representative of such zone d bodies in this and other regions, and contains discontinuous miarolitic p ockets near the base of the graphic pegmatite zone. Tourmaline, garnet, bio tite, and muscovite show no changes in major- or minor-element compositions indicative of progressive magmatic fractionation until the immediate vicin ity of the main miarolitic zone, where they record abrupt and extreme enric hments in Li, F, and Mn. There is no correlation of chemical changes in the dike with the appearance of small miarolitic vugs well below the main miar olitic zone, nor is there any indication that the aplite, graphic pegmatite , or miarolitic pockets represent separate magma injections. The chemistrie s of tourmaline? garnet, and micas, however. preclude conventional models o f Rayleigh fractionation or traditional zone refining. Textural features an d modeled cooling histories indicate that the dike cooled quickly and might have solidified partially or totally to glass before crystallization comme nced. Geothermometry based on the compositions of coexisting plagioclase an d homogeneous, nonperthitic K-feldspar indicates inward crystallization of the dike, from similar to 400-435 degrees C at the margins to similar to 35 0-390 degrees C within 20-30 cm of the pocket horizon, then a sharp decreas e to 240-275 degrees C in the pockets where K-feldspar is perthitic. We int erpret the feldspar geothermometry (except perhaps in the miarolitic caviti es) to reflect the temperatures at crystallization fronts that advanced int o the pegmatite, first from the foot wall and eventually joined by a simila r front downward from the hanging wall. Crystallization down from the hangi ng wall may have commenced after similar to 70-80% of the foot wall aplite had crystallized. The very abrupt increases of Li, Mn, and F in tourmaline and garnet near the miarolitic zone appear to be explained best by the proc ess of constitutional zone refining, in which a fluxed crystallization fron t sweeps an incompatible element-enriched boundary layer through a solid or semi-solid. After these two highly fluxed boundary layers merged near the main miarolitic zone, compositional evolution could have proceeded by cryst al-melt fractionation.