Rit (Vísindafélag Íslendinga) - 01.06.1984, Blaðsíða 318

314
PÁLL IMSLAND
ing minerals of the primary maíic magmas of the Jan Mayen magma system.
These primary magmas are similar in composition to the more Mg rich
ankaramites and are thus quite primitive. Equilibrium considerations show
the clinopyroxene to crystallize before olivine in numerous rock samples and
the various features show it to be of far greater influence on the early rock
suite and magma evolution than the olivine. The accessory chrome-spinel
has a minor influence on the magma, except for its Cr content, but it is
apparently the first mineral of these three early ones to crystallize. This
early chrome-spinel crystallization is interrupted by the crystallization of
the chromian diopside. The early crystallization of these mafic minerals
starts at — 1350 to 1400°C and at about 80 km depth. This crystallization
produces the wehrlite, which in a disintegrated state or still uncumulated to
a coherent rock, approaches the surface in primitive magmas along with
other minerals produced en route. The ankaramite crystallization, i.e. the
early titan-salite and the olivine, slightly less forsteritic than the most
primitive one, represent this early en route crystallization. According to
least-squares calculations of the crystal fractionation model, it is possible to
produce the ankaramitic basalts by subtracting these early minerals from
the most primitive ankaramitic compositions, in probable portions, with
small modifications of the K content.
2) After the wehrlite and ankaramite crystallization, the crystallizing
mineral assemblage changes. Iron-titanium oxides, especially titanomagne-
tite, and plagioclase, which soon attains a major role, start to crystallize.
The olivine and titan-salite further evolve. This gabbro type crystallization
takes place above 40 km depth and primarily between 1200 and 1100°C.
The gabbro xenoliths are produced by this crystallization and the pheno-
cryst assemblage of the basalts. Derivation of the basalts from the ankara-
mitic basalts, by fractionating these minerals is, according to the least-
squares model, difficult without modifications of the alkalies, Mn and Ti.
3) In going from the basalts to the more evolved rocks, the mineralogy
changes drastically. The plagioclase evolves and becomes a major phase at
the expense of the clinopyroxene, and the olivine influence becomes neglible.
Least-squares calculations, from Mg poor basalts to these evolved rocks,
give a relatively good total fit, but the misfit is characterized by the same
elements as in the case of the basalts. Trace elements clearly argue against
crystal fractionation in this step.
4) In the most evolved rocks, the tristanites and trachytes, the alkalinity
of the feldspar becomes of great importance as does the appearance ofbiotite
as a phenocryst phase. Olivine may still be present but is extremely scarce.
Likewise the influence of the clinopyroxene becomes relatively small. The
crystallization of the phenocrysts of these rocks primarily takes place within
the crust and at temperatures up to ~1100°C. To derive the trachytes from
the tristanites by the least-squares calculation model, a separate “mother
liquid” for Nord- and Sör-Jan, and a different fractionate mineralogy, offive