Greinar (Vísindafélag Íslendinga) - 01.01.1977, Page 101

99
We have a very clear evidence in Stapafell in Iceland (77),
where it is seen that olivine phenocrysts of a few mm dia-
meter and even smaller ones, have sunk into the lower half of
each pillow in a pile of pillows formed under seawater. The
time of sinking, measureable by the estimated time of cooling
by conduction towards the cold surface of each pillow, has
been a fraction of an hour. So, for the 30—40 years time span,
there would be no difficulty in understanding the sinking of
such heavy crystals, if the basaltic melt has no much greater
viscosity due to pressure than in the pillows.
The inferred sinking crystals must have grown in equili-
brium with the basaltic melt, and then olivine or rather pyrox-
ene seem to be our only choice.
The quantity of Th to be extracted from the melt, naturally
corresponds only to that of a trace element, and a small amount
of pyroxene must suffice for this. The chemical composition at
the mantle level where the Etna magma production takes place,
would thus be that of Etna lavas with a small addition of
pyroxene.
What we expect of increased shear in the present connection,
is increase in the formation of magma. Further, this should
lead to increased expansive force as a factor in the opening
of an eruptive fracture from below.1)
If the upper mantle, at magma production depth, were just
formed of old extruded basalt, that had gradually been de-
pressed to that depth, we seem to lack Th for the formation
1) In this conection a further factor should be mentioned. A lecture on
strain measurements, i.a. on Etna, was given on Oct. 9, 1975 at the Uni-
versity of Iceland by Mr. Roger Bilham of Lamont-Doherty Observatory.
Assuming a magma chamber connected with a vertical fracture, filled
with magma, Bilham indicated a theoretical strain magnification in the
fracture to be a factor of 2h/d, where h is the height of the fracture and
d its width. In agreement, a tiltmeter was found to indicate periodic
changes as a result of body-tides, until an eruption started.
Thus, body-tides seem to add greatly to the opening of a magma re-
servoir, the rise of the magma, and the initiation of the eruption. When
the magma has in that way reached a relatively shallow depth, the shear
stress may finally and suddenly open the way by the formation of en
echelon fractures.