PHASE-RELATIONS OF AN ARMALCOLITE - PHLOGOPITE LAMPROITE FROM SMOKY BUTTE, MONTANA APPLICATIONS TO LAMPROITE GENESIS

Suprasolidus phase relations at pressures from 8 to 30 kb and temperatures from 950 to 1380-degrees-C have been determined experimentally for a glassy armalcolite-phlogopite lamproite from the chilled margin of a medium-grained lamproite from Smoky Butte, Montana. The armalcolite-phlogopite lamproite has microphenocrysts of olivine in a groundmass of phlogopite, sanidine, armalcolite, clinopyroxene, chromite, priderite, apatite, and abundant glass. The lamproite is SiO2-rich and has high F/H2O relative to lamproites that have been investigated in previous experimental studies. Our data show that with decreasing temperature from the liquidus at pressures above approximately 12 kb, melt coexists successively with: olivine; orthopyroxene + clinopyroxene; orthopyroxene + clinopyroxene + phlogopite; clinopyroxene + phlogopite; and clinopyroxene + orthopyroxene + K-richterite. Below 12 kb, the assemblage succession is: olivine; olivine + clinopyroxene; olivine + clinopyroxene + phlogopite; and olivine + clinopyroxene + phlogopite + armalcolite. The main difference from the natural paragenesis is that the rock does not contain any orthopyroxene-a feature that is rather remarkable inasmuch as it has approximately 16% normative hypersthene-and the rock differs also in that it contains sanidine and priderite. In the experiments, sanidine is observed only as ghostlike domains in some of the glass and appears to have formed during quenching. The solid phases crystallized experimentally are generally compositionally similar to the minerals in the rock. These similarities and the experimental phase relations support the concept of a rapid initial magma ascent with only a small temperature drop and crystallization of olivine, but not of orthopyroxene. At lower pressures, less than approximately 12 kb, it appears that the magma ascended more slowly with a larger temperature drop suggested by the similarity of the experimentally determined sequence of assemblages to the paragenesis of the rock. No quasi-invariant multiphase-saturation point was found such as might be indicative of pressure and temperature conditions for formation of the lamproite magma by eutectic-type partial melting of a mantle source. The occurrence of olivine, orthopyroxene, and clinopyroxene near the liquidus, and the high proportion of normative hypersthene in the melt suggest that lherzolite or harzburgite was probable in the magma source rock. The high SiO2 and MgO contents of the Smoky Butte lamproites may indicate that orthopyroxene was a source mineral even though it did not crystallize under near-surface conditions. The curve defining the appearance of phlogopite appears at progressively lower temperatures from the liquidus as pressure increases, so it would appear that either phlogopite was not the mantle K-reservoir, or it was entirely consumed during the partial melting process. The composition of the near-liquidus glass in the experiments is likely to be the composition of the bulk rock less the very small amounts of olivine + clinopyroxene + orthopyroxene crystallized within a few degrees below the liquidus. From the inferred composition of this glass, anhydrous phlogopite is a potential mineral. The principal variable that determines whether phlogopite crystallizes as a near-liquidus mineral is F/H2O; low values of this ratio promote the presence of phlogopite as a near-liquidus mineral whereas high values deter its crystallization. The common practice of adding H2O but not F in experiments to compensate for degassing may obscure the role of phlogopite in the evolution of lamproite magmas.