Morphology of quench crystals in submarine basalts

Abstract

Submarine basalts from the mid-Atlantic ridge, Red Sea rift, and Joides site 105 in the western Atlantic have been studied in ultra thin, doubly polished thin sections. Most of the samples are pillow lava fragments containing a variety of skeletal crystal growth forms that can be related to three major textural zones in the pillows. Olivine appears as diffuse, lattice-like growths near quenched pillow margins, as elongated chainlike or hollow lanternlike forms within a partly crystallized variolitic zone, and as larger lantern shaped crystals within the holocrystalline pillow centers. Plagioclase forms acicular hollow prisms elongated parallel to a in the variolitic zone. In cross section, these crystals often show sector zoning developed on the 001 and 010 faces. Each sector is also progressively zoned outward from a more calcic to a more sodic composition. In holocrystalline pillow cores, plagioclase forms more nearly equidimensional solid crystals with no trace of sector zoning. Pyroxene is rarely well crystallized and is typically found as feathery crystals intimately intergrown with plagioclase in the groundmass. Spinel and magnetite are ubiquitous in the samples examined, although the glassy pillow rims usually contain spinel but not magnetite. Within the variolitic zone and holocrystalline pillow interiors, magnetite appears both as skeletal crystals and as overgrowths on spinel. Sulphides (mainly pyrite) are especially abundant in the Red Sea lavas. Phenocrysts of olivine and plagioclase and subhedral and euhedral groundmass olivine and plagioclase show no evidence of growth by filling in of hollow or lattice-like crystals similar to those of the outer pillow zones, It is suggested that development of these skeletal forms is restricted to supercooled, highly viscous magma in which relatively rapid growth is combined with low rates of diffusion. Slower growth or more rapid diffusion rates or both at crystallization temperatures maintained close to the liquidus appear to promote more even growth rates and compositional homogeneity on all crystal faces; this possibility accounts for the more nearly equidimensional form and lack of sector zoning in phenocrysts and microphenocrysts. Petrographic observations support the division of sea floor basalts into ‘Olivine tholeiites’ and ‘plagioclase tholeiites’ which has been suggested by other workers. The only large pyroxene crystals observed in the samples examined appear to be xenocrysts, and pyroxene is not a prominent quench (liquidus) phase in this or other samples studied.