The evolution of the major element compositions of mare basaltic liquids and coexisting crystals in low‐pressure fractionation processes has been modeled in order to examine the postulated link between picritic volcanic glasses and mare basalts. Existing models of equilibrium and fractional crystallization [Longhi, 1977, 1982] have been modified to include expressions for the Cr‐spinel liquidus surface and the boundary between the armalcolite and ilmenite surfaces; average molar partition coefficients for armalcolite/liquid pairs are also included in the modifications. Results of the calculations show that most mare basalts are not derived by simple low‐pressure fractionation from parents with the compositions of picritic glasses as tabulated by Delano [1986a]. Also, there is no indirect evidence as yet for posteruption fractionation of picritic magmas in the form of sufficiently magnesian, mare‐type olivines at the Apollo 12 and 15 landing sites. However, the calculations do suggest that Luna 24 ferrobasalt and Apollo 11 high‐K basalts could have been derived by simple, near‐surface fractionation of known picritic compositions. Also, xenocrystic olivines with mare‐like concentrations of TiO2 and CaO in high‐Ti basalt 74275 provide evidence for the existence of a picritic magma at the Apollo 17 site. More importantly, the calculations indicate that the picritic magmas would have fractionated to produce basalts with bulk and mineral compositions generally similar to those of mare basalts. These latter results favor the hypothesis that mare basalts have fractionated, “nonprimary” compositions and that the small number of observed linkages between basalts and picritic parents is a consequence of limited sampling.
Read full abstract