Geochemical confirmation of the lunar magmasphere hypothesis

The lunar magmasphere (or magma ocean) hypothesis was originally conceived to account for the enrichment of cumulus plagioclase (Al and Ca) in the main (highlands) portion of the crust. The great age of the highlands, and the complementary pattern of Eu anomalies between the highlands and the younger mare basalts, helped convince most specialists that the magmasphere hypothesis is correct. Doubts persist, however, particularly among physicists concerned about heat sources. It was shown in 1976 that a plot of Na/Ca vs. Mg/Fe for pristine highlands cumulates manifests a profound bimodality: One group, the Mg-rich rocks, plots along a normal igneous trend of inverse correlation between Na/Ca and Mg/Fe; the other group, the ferroan anorthosites (FAN), features low Na/Ca and low Mg/Fe. Only the FAN group can be plausibly linked to plag. flotation over the magmasphere.