Rit (Vísindafélag Íslendinga) - 01.06.1984, Page 295

PETROGENETIC RELATIONSHIPS
291
fractionate, which contains up to five different minerals, including a com-
plex clinopyroxene, which in fact is quite similar in chemical composition to
the basic rocks themselves.
VII. The trace elements in the crystal fractionation
Incompatible trace elements are those rejected by the rock forming
minerals during magmatic crystallization. During early crystallization these
are thus continuously enriched in the remaining magma, as the crystalliza-
tion advances. In later stages of crystallization, crystals may become stable
that tolerate or even select these elements, which then no longer are
incompatible. At the same time other elements may become incompatible,
that were not previously. Incompatible trace elements have a crystal-liquid
distribution coeíficient of zero. In all crystal-liquid reactions the ratio of two
such incompatible elements will thus not change. A plot of one incompatible
elements versus another for a rock suite of an uncomplexed evolution will
thus show a linear trend going through the origin of the diagram.
Zr, Y an Rb are usually incompatible trace elements in the early evolution
of rock suites, i.e. if phlogopite or amphibole are not participating minerals.
These elements have been analysed in all the Jan Mayen samples. Their
concentrations are shown in the tables and figures of chapter 8. Zirconium
crystals have been found in one trachyte sample only, but these are small
and few. Biotite and alkali feldspar are common minerals in the tristanites
and trachytes. Rb thus ceases to be an incompatible element in the most
evolved part of the rock suite, while Y and Zr probably do not.
In Fig. 127, these elements are plotted against each other. The pattern is
linear for the basic rocks, but the scatter increases as these evolve . In the
more evolved rocks (higher concentrations of incompatible elements), the
Rb-Zr diagram still shows an increase of this scatter, while in the Y-Zr
diagram the trend swings towards higher Zr values at nearly constant Y
values. The reason for the scatter is unknown, but within it is contained the
constant ratio concept of truly incompatible behaviour.
Treuil & Varet (1973) have pointed out the usefulness of plotting an
incompatible element ratio against the concentration of the numerator
element of the ratio. This plot allows the recognition of the pure crystal
fractionation, as long as the incompatible behaviour is undisturbed, as the
ratio is kept constant by the process, while the elements increase in amount.
Fig. 128 shows this for Zr and Y. The plot indicates that the highly evolved
rocks are not all evolved through the same fractionation, if fractionation is
the process behind their evolution. The Nord-Jan tristanites and trachytes
might be fractionated from the basic rocks of the rock suite, while the Sör-
Jan tristanites and trachytes apparently are not fractionated from the basic
rocks ofjan Mayen at all. IfY is here substituded by Rb, which is not purely
incompatible throughout the rock suite, the possibility arises of identifying