Rit (Vísindafélag Íslendinga) - 01.06.1984, Qupperneq 286

282
PÁLL IMSLAND
total fit. The most serious misfit is that of Na, which is too low in the
calculated “daughter” in all cases. In two cases (of three) A1 is too high. Ti
is slightly too high in two cases, and K and Mg in one case. P is slightly high
in all cases. Mn has been too high in most cases hitherto, but here this
tendency is less serious. This subtraction of minerals from a basaltic
tristanite leaves 35 to 54 per cent of a tristanite “liquid”.
If subtraction of five (diíferent but relevant) minerals in the appropriate
proportions, but with a fairly good total fit, can be taken as supporting
evidence for the operation of the crystal fractionation process, then the
process is a likely candidate for being responsible for the evolution of the
low-silica intermediate rocks. The systematically high Na content of the
calculated “daughters” does, however, cast some doubt on its complete
effectiveness.
V. Crystal fractionation within the highly evolved rocks
The trachytes are moderately porphyritic rocks. A hydrous mineral,
biotite, becomes an important phenocryst phase in this rock suite in the
tristanites, where it occurs in smaller amounts than in the trachytes. There
is thus a close resemblance between the tristanites and the trachytes in this
respect, as well as many others.
Here a tristanite is chosen to enter the trachyte group, but not without
complications. These most highly evolved rocks of the rock suite have
chemical characteristics which indicate that the rocks from Nord-Jan and
Sör-Jan might be of different origin or differently evolved. This is primarily
indicated by the incompatible trace element pattern and is confirmed by the
calculations undertaken here. To derive both Nord- and Sör-Jan trachytes
from a single tristanite by the subtraction of a single fractionate is impossi-
ble, no matter whether a Nord- or Sör-Jan tristanite is used as a “mother”.
By single fractionate subtraction this problem is more easily solved if two
“mother liquids” are permitted (one for Nord-Jan, one for Sör-Jan). The
best result is obtained by using the two “mothers” and two different
fractionates as well, which is done here, Nord-Jan and Sör-Jan trachytes
being fractionated separately and from separate “mothers”. This is in
accordance with the spatial distribution pattern of the rock types and their
mineralogy, as described in chapter 5.
Jan 37, a tristanite from Nord-Jan, is here used as a “mother liquid” in
the calculations of crystal fractionation for the Nord-Jan trachytes (Table
49). Five minerals are needed in the calculations to obtain a good total fit.
This fracdonate is highly dominated by a feldspar, which is an orthoclase
(AnKjAb^Or^) in composition. Among the other minerals is a clino-
pyroxene of an evolved composition. The high Ti and A1 characterizing the
salites of the basalts are here considerably lower, while Fe is still high. The
titanomagnetite is identical to the previous one. The olivine is Fa enriched.