Acta naturalia Islandica - 01.07.1964, Side 15

THE ANKARAMITES OF HVAMMSMÚLI, EYJAFJÖLL, SOUTHERN ICELAND 13
phenocrysts. Indeed, in some sections the presence of such cavities is as much
the rule as is their absence. Such cavities may in certain rocks be explained
by a rapid or incomplete growth in situ (Drever and Johnston 1957), and
in the case at hand the earlier phase of crystallization may have occured
in this way, fallowed later by the crystallization of the rim and the cavity
fillings. We are therefore compelled to postulate either that these cavities
represent a transverse section of a ,,tunnel“ which in fact is open to the ground-
mass, or the crystallization of the inclusion from liquid trapped by the crystal-
lizing olivine earlier on.
Most of the olivine crystals are very clear and free of inclusions, but
exsolutions of magnetite are present in some. They form small, circular dend-
ritic growths orientated parallel to a cleavage or they may form apparently
irregular pattern of trails of inclusions winding about the crystal. The distri-
bution pattem is partly dependent on the orientation of the crystal in the thin
section. An extreme case of this type is seen in the top zone of the outcrop
just SW of the quarry. Here the olivines are filled with „inclusions" almost
to such an extent as to make the host opaque in thin section (Pl. IV a and b).
The „black olivines" are completely fresh and unaltered save for the ore ex-
solutions. Usually they are euhedral, whereas the accompanying pyroxene is
rounded in shape. The margin is always bordered by a dense growth of ore
which may extend into the groundmass as well. Then follows a narrow zone
clear of inclusions. The main central area is crowded with ore. The magnetite
occurs in irregular clots, or branching streaks ramifying the surface of the
crystal section. The occasional parallelism of the streaks suggests that the
distribution of ore is in fact not random, but is governed by the crystal struc-
ture to some extent. Some crystals are completely blackened, or light may pene-
trate through the crystal as through a thick cloth. The clear marginal zone
is, however, present in most cases. Where the degree of exsolution is less in-
tense, the crystal may be comparatively clear in the centre, and grow blacker
towards the margin. The probable course of the process of exsolution may be
deduced from the apparently various stages or degrees seen in the section:
First the margin becomes crowded with magnetite (Pl. IV a), which
appears to be rather on the crystal than inside it. Then, (and probably simul-
taneously), exsolutions of rods or clots of magnetite start to form within
the crystal, most intensely nearest to the margin (but leaving the clear zone
at the margin) and decreasing in intensity inwards. The clots join up to form
streaks, which in turn connect to form a dense net of ore. Finally the speces
of the „net“ are filled up.