Effects of quench methods on Fe (super 3+) /Fe (super 2+) ratios; discussion and reply

In a recent paper with the above title, Dyar et al. (1987) stated, “The poor precision of wet-chemical Fe³⁺ measurements is disconcerting, but is of particular concern in reduced specimens containing small amounts of Fe'⁺…. In such cases, the Fe³⁺ values or calculated ratios (i.e., Fe³⁺/Fe²⁺ or Fe³⁺/Fe_tot) can have a very high uncertainty (50-100%). This is not a trivial problem in experimental studies requiring reducing conditions. One particular application for which this could present a problem is in the calculation of magmatic f_O2 values from volcanic glass compositions. A 10% change in the ratio Fe³⁺/Fe²⁺changes the f_O2estimate calculated by the method of Kilinc et al. (1983) by one log unit.” (p. 799; my emphasis). Intuitively, this last assertion should seem implausible to the reader, and indeed it is in error, on the high side, by a factor of about 5.