Complete chemical analyses, including ferric and ferrous iron, H2O
contents and delta D values for 16 phlogopite and biotite and 2
hornblende separates are presented. Samples were obtained from volcanic
rocks from four localities: (1) phlogopite phenocrysts from minette
lavas from the western Mexico continental are, (2) biotite and
hornblende phenocrysts from andesite lavas from Mono Basin, California,
(3) phlogopite and biotite from clinopyroxenite nodules entrained in
potassic lavas from the East African Rift, Uganda, and (4) phlogopite
phenocrysts from a wyomingite lava in the Leucite Hills, Wyoming. The
Fe2O3 contents in the micas range from 0.8 to 10.5 wt%, corresponding
to 0.09 to 1.15 Fe3+ per formula unit (pfu). Water contents vary from
1.6 to 3.0 wt%, corresponding to 1.58 to 3.04 OH pfu, significantly
less than would be expected for a site fully occupied by hydroxyl.
Cation- and anion-based normalization procedures provide accurate
mineral formulae with respect to most cations and anions, but are unable
to generate accurate estimates of Fe3+/Fe-T, and overestimate OH at the
expense of O on the hydroxyl site. These inaccuracies are present
despite acceptable adjusted totals and stoichiometric calculated site
occupancies. The phlogopite and biotite phenocrysts in arc-related lavas
from western Mexico and eastern California have the highest Fe3+/Fe-T
ratios (56-87%), reflecting high magmatic oxygen fugacities (Delta NNO
= +2 to +5), in contrast to those from Uganda (25-40%) and the Leucite
Hills (23%). There is no correlation between the OH content and the
Fe3+/Fe-T ratio in the micas. Values of K-D(Mg/Fe2+) (+/- 2 sigma
errors) were calculated for three phlogopite-olivine pairs (0.12 +/-
0.12, 0.26 +/- 0.14, 0.09 +/- 0.12), two biotite-hornblende pairs (0.73
+/- 0.08 and 1.22 +/- 0.10) and a single phlogopite-augite pair (1.15
+/- 0.12). Values of K-D(F/OH) for two biotite and hornblende pairs
could not be determined without significant error because of the
extremely low F contents (< 0.2 wt%) of the four phases. The delta D
values obtained in this study encompass a large range (-137 to -43 parts
per thousand). The phlogopite and biotite separates from Uganda have
delta D values of -70 to -49 parts per thousand, which overlap those
believed to represent `'primary'' mantle. There is a larger range in
delta D values (-137 to -43 parts per thousand) for phlogopite
phenocrysts from western Mexico minette lavas, although their range in
delta(18)O values (5.2-6.2 parts per thousand) is consistent with
`'normal'' mantle. It is unlikely, therefore, that the variable delta D
values reflect heterogeneity in the mantle source region of the minette
magmas. Nor can the extremely low delta D values reflect degassing of
H-2 or H2O since almost 100% loss of dissolved water in the magma is
required, an unrealistic scenario given the stability of the hydrous
phenocrysts. The very low delta D values of the Mascota minette
phlogopites require that the hydrogen be introduced from an external
source (e.g., meteoric water). Whatever the process responsible for the
observed hydrogen isotope composition, it had no effect on the
delta(18)O value, f(O2), a(H2O) or bulk composition of the host magmas.