Petrogenesis of Silali volcano, Gregory Rift, Kenya.

Basalts of the Silali volcano are of high-Fe transitional affinity and range from plagioclase-phyric varieties, probably formed by selective accumulation, to aphyric. They have compositional similarity to oceanic island basalts, but are isotopically variable and have strong, negative Zr anomalies on chondrite-normalized plots. Major and trace element and isotopic evidence indicates that the basalts represent several liquid lines of descent, each equilibrated at crustal pressures. All Silali basalts are relatively evolved; primitive basalts were probably held close to the Moho, where they evolved by gabbro fractionation to produce basalts with less than 8% MgO. These magmas were subsequently held at several levels in the crust, including the rift-axial lava-sediment sequence. Fractionation was along a high-Fe trend through mugearite to metaluminous, two-feldspar trachyte and thence to peralkaline, one-feldspar, silica-oversaturated and -undersaturated trachytes. Magmas intermediate in composition between basalt and trachyte were very rarely erupted at Silali, but their evolution can be deduced from an extensive suite of dolerite blocks, which contain residual glasses varying from mugearitic to peralkaline phonolitic compositions. Magma mixing between basalt and trachyte has been common at Silali, implying complexity of the plumbing system. Evolution of the trachytes was by fractional crystallization combined with assimilation of crust during or after fractionation