Implications of experimental petrology to the evolution of ultrapotassic rocks

The contributions of experimental studies pertinent to ultrapotassic rocks of Groups I (lamproites) and II (kamafugites and related rocks) are discussed in terms of synthetic systems, ultrapotassic rock compositions, experiments on characteristic minerals in these rocks and experiments designed to model mantle metasomatism. These studies indicate that the majority of ultrapotassic magmas are derived by partial melting of a metasomatically enriched mantle source at depths of 100 km or greater, and under fluid conditions represented by the C---O---H system with fluorine that may be reduced or oxidized relative to other compositions. Many lamproitic magmas may be derived from a phlogopite-harzburgite with volatiles that are predominantly H₂O and F₁ whereas kamafugitic type ultrapotassic magmas may be products of partial melts of a more wehrlitic mantle source in which the main volatiles are H₂O, CO₂ and possibly F. Experimental and theoretical considerations of mantle metasomatism suggest that it occurs at of f_O2 in the range of the FMQ buffer. Metasomatism involves low density mantle fluids (melts?) in which H₂O and CO₂ are the important volatiles, buffered by amphibole, phlogopite and carbonates. Results of recent experiments suggest that the reactions causing metasomatism may be decoupled and cyclic and occur at different depths.