Rit (Vísindafélag Íslendinga) - 01.06.1984, Blaðsíða 137

MINERAL CHEMISTRY AND REEATIONSHIPS
133
heterogeneous source on a small volume scale or that the magma was
collected í'rom a relatively great volume of source rock but was however able
to transport with it the residual minerals, which means that at least 50 per
cent of the material must be liquid according to O’Hara (1967). The
mineral variation thus seems to exclude the possibilities that these minerals
represent: a) pieces of the mantle, unmodiíied by magma extraction, or b)
crystal residues after some magma production. The restriction of the
wehrlite occurrence to ankaramites and the characteristic occurrence of the
wehrlite minerals in the ankaramites further seems to exclude the possibility
that these are accidentally caught fragments. l’he possibility then remains
that these wehrlite minerals are early crystallization products of the primary
magmas, which accumulated at depth as phenocrysts in the liquid to
produce the ankaramites and more coherent pieces in the case of the
wehrlite (cognate xenoliths). The compositional variations of these minerals
then reflect changing physical or chemical conditions during their cryst-
allization sequence. The stressed nature of the olivine is difficult to explain
in this case. Stressed minerals in ultramafic nodules seem however to be
quite common, whether the nodules are explained as having cumulated
from liquids or as parts of primary mantle rocks. There seems to be no
reason not to believe that the more fayalitic olivines and the titan-salites are
true phenocrysts.
According to Roeder & Emslie (1970) the composition ofolivines crystal-
lized from a liquid is independent of temperature and only depends on the
MgO/FeO ratio of the liquid. Thus the composition of the olivines permits
the determination of the composition of the liquid from which the olivines
crystallized according to the equation: (MgO/FeO)oiX0.3=(MgO/FeO)liq
(in mol. per cent).
Duke (1976) argued that the distribution of Fe and Mg between clino-
pyroxene and liquid is independent of temperature but varies as a function
of composition. He presented the equilibrium conditions for clinopyroxene
crystallizing from a liquid, as being: log(Fe/Mg)cpx=—0.564 + 0.755 log(Fe/
Mg)llq. Mákipáa (1978) in a study of equilibrium relations of clinopyroxenes
and olivines of Icelandic tholeiites obtained an equilibrium line ofsomewhat
different slope. The compositions of clinopyroxenes are more complex and
much more variable than those of olivine. The clinopyroxene-liquid equili-
brium relationships might thus be more complex than the olivine-liquid
relationships and the ones referred to here may not be as reliable as those of
the olivine-liquid.
The actual compositions of the liquids crystallizing both the olivines and
the clinopyroxenes are unknown, but equilibrium relationships may show
the compositions of these liquids, as regards iron and magnesium content,
compared to the whole rock composition. Thus a qualitative study of the
equilibrium diagrams may help in gaining some general ideas regarding
these liquids and the origin of the minerals. In Figures 84 and 85 the iron-