Rit (Vísindafélag Íslendinga) - 01.06.1984, Blaðsíða 294

PÁI.L IMSI.AND
290
relevant rocks. Si in no case shows noticeable misfit. Fe occasionally only
shows a minor misfit. This is in the case of ankaramite and basalt fractiona-
tion. The misfit is not systematic. Likewise Ca may show a minor misfit in
ankaramites and basalts, but it is then systematically too low in the
calculated “daughters”. In the basalts and the low-silica intermediate rocks
Mg occasionally shows a moderate rnisfít. In all these cases the calculated
“daughters” are too high in Mg. Al, likewise, occasionally shows a moderate
misfit. In this case the calculated “daughters” are too poor in Al, if
primitive, but too rich in Al, if evolved. Ti frequently shows a misfit that
exceeds the standard deviation of the analyses. In the case of the ankara-
mites, the less evolved basalts and trachytes the calculated “daughter” is too
low in Ti, while in the evolved basalts, basaltic tristanites and tristanites the
“daughter” is too high in Ti. This happens in spite of the fact that the
fractionate is generally Ti poor in the former case, but Ti rich in the later
case, where two different Ti containing minerals are present. The alkali
elements Na and K show the greatest misfit. It occurs in all rock groups and
frequently it appears in the majority of the calculations made within each
group. This misfit is high and in numerous cases it dominates in the total
misfit figure. The pattern of the misfit is however not as clear cut as in some
of the less serious misfit cases mentioned above. The calculated “daughter”
is too high in Na in the case of the basalts, basaltic tristanites and the
trachytes of Nord-Jan, while it is too low in Na in the case of the
ankaramites, tristanites and the trachytes of Sör-Jan. K, on the other hand,
is too high in the calculated “daughter” of some of the basalts, all of the
tristanites and the Sör-Jan trachytes, but too low in the rest of the basalts,
the ankaramites, the basaltic tristanites and the Nord-Jan trachytes. In the
case of the alkali elements there is no system between the calculated
“daughter” being too rich or poor in these elements and the fractionate
being alkali rich or poor. The reasons for the misfit thus apparently lie
outside the calculation procedure or its restrictions. Probably the most
serious misfit is shown by Mn. This misfit is only a small portion of the total
misfit figure, but it is serious because it occurs in a minor element where it
may amount up to 50 per cent of the concentration in the sample, and
because it is systematically too high in the calculated “daughter liquid”. It
occurs in all rock groups except the ankaramites. Mn is present in numerous
minerals of the rock suite as a minor element and behaves there as in other
rock suites, i.e. showing enrichment in late crystallized phases relative to
early crystallized ones and being high in minerals of evolved rocks as
compared to those of primitive rocks. The fractionated minerals were
analysed for Mn and it is thus clear that this misfit is not introduced by the
calculation procedure.
On the other hand, this misfit pattern, especially that of Mn, might
indicate that in spite of a good total fit, the mineral proportions in the
fractionate are rather far from correct. Such errors are easily introduced by a