Comment on “Compositional convection in a reactive crystalline mush and melt differentiation” by Stephen Tait and Claude Jaupart

At the end of their analysis of convection in magma chambers, Tait and Jaupart (1992) 'speculated' that the platiniferous dunitic pipes in the Bushveld Complex, South Africa, are 'fossil chimney structure', which developed by a process of 'compositional convection.' A hypothesis that accounts for all of the characteristics of the platiniferous dunitis pipes has not been establihsed, bu the evidence against Tait and Jaupart's suggestions is considerable: (1) a two-stage hypotheis should by entrained; (2) the magneusium dunites are primitive rocks that are unlikely to be related to residual liquids; (3) residual liquids evolved from the Lower Zone (LZ) - Lower Critical Zone (LCZ) would not be sufficiently differentiated to account for the iron-rich assemblages; (4) suitably iron-rich residual melts may be derived from the Upper Critical Zone (UCZ), but they would be extremely dense amd liable to drain downward; and (5) the absence of plagioclase is not compatible with 'typical' interstitial liquids.