Greinar (Vísindafélag Íslendinga) - 01.01.1977, Page 87

85
from the differentiation diagram of Hekla lavas, that the
source of the Hekla magma, with which the content of the
reservoir would all the time be in hydrostatic connection, cor-
responds to a great basaltic magma flow, an about 6000 years
old lava (Gunnarsholt-lava). But magma of that primary com-
position was erupted 1878 and 1913 on a fissure only 3 km
east of the Hekla fracture, as we have stated in Chapter 3.
Thus the primary magma does not differentiate for thousands
of years close to Hekla. Only in a fracture-shaped Hekla re-
servoir it does so, the Si02 rising by at least from 54—64%
during a century of dormancy. There is undeniably a differ-
ence between a “latent” or a “potential” magma layer, and
magma that has already risen up into, or by some other way
is the filling of a reservoir. Growing evidence shows that we
must think of such reservoirs as being essentially of a vertical
dimension, because only a vertical order of composition in
the reservoir can reasonably explain why this is changed into
a chemistry-time order during an eruption. This outcome was
all clear in the Hekla eruption. But was Hekla the only vertical
reservoir, in which differentiation took place? If potential
magma was changed into real magma in a deep fissure, or in
a kind of vertical funnel at the crossing of fractures (cf. our
conclusion on shieldvolcanes) might not some differentiation
take place in the reservoir, either by gravitative settling or
by assimilation? If it took some time for the magma mass of
a prospective eruption to form or collect in such a vertical re-
servoir, a vertical chemical order would then be turned into
time order during the eruption. Were there any such indica-
tions ?
In lectures at four universities and in the Royal Academy
in the Netherlands in 1952, the present author stated his belief
that the magma of shield volcanoes in Iceland came from a
greater depth than that of the fissure eruptions. Some time
later, he suspected that, the chemistry might also be slightly
different. He felt that only a geochemist should evaluate the
significance of the known slight variations of composition, and
put the question before one. But the latter definitely denied