Basalt melting, hydrous

The formation of a hemisphere of andesite is not easily explained, especially in the absence of subduction. Fractionation of basaltic magma to produce andesitic residual melt is inefficient, although less fractionation is required if the magma is hydrous. Evidence has been presented that shergottite magmas contained water prior to eruption (McSween et al., 2001), although this hypothesis remains controversial. 

Although shallow-mantle xenoliths, hosted in alkali basalts, commonly contain C02-rich fluid inclusions (see below), there have been no reports, to the author s knowledge, of H20-rich fluid inclusions in these samples. The C02-rich fluid inclusions are commonly attributed to late, possibly magma-derived, metasomatism of the samples. If such metasomatism was produced by silicate- or carbonate-rich melts, ascent of such a melt could produce saturation in a C02-rich vapor, but H2O would partition strongly into either residual melt or hydrous phases such as phlogopite or amphibole. Thus, the absence of H2O in the fluid inclusions in these samples cannot be taken as evidence that the metasomatic agent was anhydrous. 

Hirose K. and Kawamoto T. (1995) Hydrous partial melting of Iherzolite at 1 GPa the effect of H2O on the genesis of basaltic magmas. Earth Planet. Set Lett. 133, 463-473. 

---