The Merzbacher & Eggler (1984) geohygrometer: a cautionary note on its suitability for high-K suites

The use of the popular Merzbacher & Eggler (1984, Geology 12, 587-590) expe rimental geohygrometer for calcalkaline and andesites and dacites is critic ally evaluated and two pitfalls are found. First, calculation of the correc t projection parameters is problematic because two endmember calculation sc hemes are found in the literature; Baker & Eggler (1983, Journal of Volcano logy and Geothermal Research 18, 387-404) and Walker et al. (1979, Contribu tions to Mineralogy and Petrology 70, 111-125). Although related, these two schemes have crucial differences that can result in very different project ion parameters for the plagioclase component. This is most crucial for high -K compositions; H2O contents estimated using the Walker et al. (1979) sche me can by as much as 100% higher than those estimated using the Baker & Egg ler (1983) projection. Incorrect projection parameter calculation has led t o overestimation of water contents in high-K andesites and dacites from the Central Andes. Second, for medium-K ang high-K andesitic-dacitic compositi ons water contents derived using the Merzbacher & Eggler (1984) geohygromet er deviate considerably from water contents estimated using other methods. Experimental data from the literature, and our studies of water contents in ferred from melt inclusions and plagioclase-melt equilibrium for dacites fr om the Altiplano-Puna Volcanic Complex of the Central Andes indicate that t he Merzbacher & Eggler (1984) geohygrometer should not be applied to compos itions with K2O > 1.9 wt%, as originally calibrated.