Rit (Vísindafélag Íslendinga) - 01.06.1984, Blaðsíða 302

298
PÁLLIMSLAND
In the model calculations the number of minerals used in the fractiona-
tion increases as the rocks become more evolved. At the same time the total
misfit has a tendency to increase. In addition, the misfit, repeatedly shown
by some elements, tends to become more serious, as for instance in the case
of Mn and Ti. This clearly shows that, as the rocks become more evolved,
the best fit model fractionation increasingly deviates from the possible
actual one, and indicates that crystal fractionation has decreasing influence
on the rock suite evolution.
There is a recognisable pattern to the misfit of the model calculations
where certain elements dominate in sample after sample of the various rock
groups. This may indicate: 1) that the relative amounts of minerals used in
the fractionation calculations are wrong compared to the actual fractionate,
2) that the actual crystal fractionation is systematically accompanied by
another process, leaving this misfit pattern as an imprint on the fractional
crystallization result, or 3) a combination of these two.
In the basalts, the dominating rock group, there is a discrepancy as
regards model fractionation results, between basalts of relatively high and
low K content. The high-K basalts easily fall into the model fractionation
pattern, i.e. they show the same kind of fit and misfit as do the other rock
groups, while tha low-K basalts do not. The low-K basalts show a notice-
ably higher total misfit than the other rock groups and the misfit of
individual elements strikes elements other than those typical of the pattern
mentioned. Here Fe, Ca, Al, and Mg show the most serious misfit but are
accompanied by Na and Ti. These four first mentioned elements do not
usually show noticeable misfit at all in other cases. This discrepancy is
rather diíficult to understand within the framework of uncomplexed crystal
fractionation evolution of the rock suite. A fractionation of totally diíferent
characteristics from that of the rest of the basalts is a possible explanation.
This is a fractionation not accounted for by the model, which treats all
basalts as if they were a part of one continuous fractionation trend. There
are, however, no indications in this direction in the general petrology or
mineralogy of the rocks. A partial modification of the fractionation evolution
by some elemental loss, including K, could be postulated, but such a
postulation is of questionable relevance. A general alkali enrichment of the
rock suite, except for the low-K basalts, could also be postulated and this is
probably less quesdonable than the first proposal, but it will not be
discussed further at this point.
This discrepancy shown by the basalts directs the attention towards the
problem of the origin of the alkaline, and especially the potassic, nature of
the rock suite as such, a problem that, in spite of its fundamental import-
ance, will not be considered at this point.
The final conclusions that may be drawn, as regards the role of crystal
fractionation in the evolution of the Jan Mayen rock suite, are as follows. In
the case of the ankaramites, the compositional range is probably mostly the