Rit (Vísindafélag Íslendinga) - 01.06.1984, Side 234

230
PÁU. IMSI.AND
magmas at depths are poorly known. The relative amounts of water and
halogens, both in fumarolic gases and in gas samples of erupting magmas,
clearly show that water is greatly in excess of the halogens and this is most
probably the case at depth as well.
A relatively early crystallizing hydrous mineral would thus be expected to
contain OH>F>Cl. After some crystallization and/or degassing the relative
amounts of OH and F of the magma will have changed (OH leaving the
magma greatly in excess of F). At this later point, a crystallizing hydrous
mineral could be expected to contain F>OH>Cl.
A hydrous mineral phase crystallizing from a basic magma would thus
tend to become enriched in F at the expense of OH as crystallization
proceeds; i.e. starting as a F poor crystal, íinishing as a F rich crystal and
being C1 poor throughout the whole sequence. A hydrous mineral crystal-
lizing from a more evolved magma would be expected to be F rich in the
beginning, but might, in some cases, become relatively enriched in OH as
the crystallization continued, because of F depletion of the magma caused
by the earlier crystallization of the mineral. These minerals should also be
C1 poor.
In keeping with the above generalized model the Jan Mayen apatites are
poor in C1 relative to both OH and F, but on the average the apatites of the
basic rocks are slightly richer in C1 than those of the intermediate rocks.
Furthermore the basic rock apatites are, on the average, richer in OH and
poorer in F than those of the intermediate rocks, as has been described
above. A plot of magmatic evolution (D.I. of the rocks or another such
parameter) versus amount of these elements in the apatites does not,
however, give a smooth trend of volatile changes. l'he pattern only appears
in the average figures, because of the grouping of the apatites into the groups
of high OH-low F and low OH-high F, as previously described.
This grouping may indicate that the apatites are of different origin as
regards time of crystallization, i.e. relatively early crystallized apatites (high
OH-low F) and relatively late ones (low OH-high F). On the basis of other
criteria, the apatites ofjan 329, being of the high OH-low F group, have
been taken as the first minerals of that rock to crystallize. Jan 70, as well,
contains apatites of the high OH-low F group. This sample is in fact aphyric
except for the small amounts of small plagioclase, iron-titanium oxide and
apatite microphenocrysts, which are then the first minerals to crystallize in
the sample. Another way to look at the cause of this grouping is, instead of
stressing the time factor, to look upon the OH rich-F poor apatites as being
crystallized from relatively primitive liquids and the other group as having
originated in relatively evolved liquids. Most of the lava samples containing
apatite, other than the two samples mentioned above, contain other pheno-
crysts as well. Among these, internally resorbed plagioclases are common,
indicating, as previously discussed, that magmas of different temperatures,
compositions and state of crystallization have mixed before final crystalliza-