Rit (Vísindafélag Íslendinga) - 01.06.1984, Blaðsíða 208

204
PÁLL IMSLAND
MgO (Fig. 112) in the feldspars ranges up to 0.25 wt. per cent. It was
detected in 91 per cent of the plagioclase analyses but only in 64 per cent of
the alkali feldspar analyses. Mg is thus generally present in plagioclases and
occurs in higher concentrations than in the alkali feldspars, where it is
mostly below 0.1 per cent. In the basic coarse-grained xenoliths, Mg in
plagioclases tends to be lower than in the plagioclases of comparable An
content occurring in the basic lavas, while in the intermediate coarse-
grained xenoliths the feldspars tend to be slightly richer in Mg than those of
the lavas. As for Fe this distribution pattern of Mg in the Jan Mayen basic
rocks seems to confirm the findings of Longhi et al. (op.cit.) that the
distribution coefficient for Mg between plagioclase and liquid is higher in
the case of extrusive basic rocks than plutonic ones. In the intermediate
rocks, on the other hand, the relations are reversed, but in these rocks
plagioclase is not the only feldspar present, which may complicate matters.
In the case of the plagioclases Mg, like Fe, tends to be concentrated in the
late formed crystals relative to the early formed ones. Thus groundmass
grains or phenocryst margins are usually slightly richer in Mg than the
phenocryst cores (Table 32). The same holds for alkali feldspars. Where
plagioclase and alkali feldspar occur together, the plagioclases are slightly
richer in Mg than the alkali feldspars. Longhi et al. (op.cit.) found Mg to be
more readily incorporated into plagioclase than ferrous iron (Fe2+). Bryan
(1974), in a study of the Fe-Mg relationships in sector zoned plagioclases of
submarine basalts, found the plagioclase to be generally richer in Fe than
Mg. This is the case with the Jan Mayen feldspars as well as most other
published feldspar analyses which give both Fe and Mg (see e.g. Rahman &
MacKenzie, 1969; Frey et al., 1974; Bence et al., 1975; Myers et al., 1975;
Ayuso et al., 1976 and Whitney & Stormer, 1977). Bryan (op.cit.) found
50—83 per cent of the Fe in the plagioclases to be Fe2+. In his case the Fe in
excess of Mg can thus hardly be explained by most of the Fe being Fe3+. If
the plagioclase Fe is, on the other hand, primarily in the Fe3+ valency state,
then the finding of Longhi et al. (op.cit.) may be a general case.
SrO was detected in 78 per cent of the plagioclase analyses and in 42 per
cent of the alkali feldspar analyses. Plotted against the An content of the
feldspars (Fig. 112) it shows a broad maximum in feldspars of intermediate
An content. The most An rich plagioclases contain up to 0.25 wt. per cent
SrO. The plagioclases of the intermediate An content contain up to 0.5 per
cent SrO. In the Na- and K rich feldspars Sr is lower. The K rich alkali
feldspars contain mostly less than 0.1 per cent SrO. The Sr maximum found
here in plagioclases of intermediate An contents is similar to that found by
Heier (1962) in plagioclases of a collection of rocks and by Wager &
Mitchell (1951) in the plagioclases of the Skaergaard intrusion. Compared
to the plagioclases of the lavas, the plagioclases (of comparable An content)
of the coarse-grained xenoliths are low in Sr except for the hydrous mineral
xenolith. The first crystallized plagioclases of each sample usually contain