Greinar (Vísindafélag Íslendinga) - 01.01.1977, Blaðsíða 95

93
Concerning our case, primary basalt magma, if it formed a
fluid layer at an early state of the differentiated earth, could
possibly — because of P-T and time conditions — have develop-
ed into a state, which cannot be reproduced in laboratory after
that state had once disappeared, when it changed into active
magma, and flowed to the surface, to turn into crystallized
rock — with all the ionic exchanges that crystallization would
have caused. This is just a general reminder, not an attempt
to escape from solving the problem.
In this connection, the importance of shear in crystallization
should again be stressed (cf. Chapter 5 and 6). If shear was
absent in such a hypothetical fluid basaltic layer, into which
state would the fluid change by very slow cooling under high
pressure ?
In the author’s studies of the Hekla eruption of 1947—48,
already quoted, he found reasons for the question: could not
the process: magma —> crystallization be irreversible, such
that remelting did not reproduce the magma ? There had been
attempts to guess the ionic state of a magma, but in reality
we do not know that, nor that of the remolten rock. In 1952,
x-ray analyses of basaltic glasses were performed in Utrecht
for the author, but without any obvious significant result. And
for reasons, which have nothing to do with science, my Hekla
papers did not reach the distribution which was planned by
the Soc. Sci. Isl. and thus did not, to our knowledge, inspire
others to take up this question.
By now, there are various new data which seem to aid us.
We shall first consider suggestive data on silica. We quote
from (72): “Diffraction patterns of both silica and germania
glasses are consistent with a structure in which nearly all
the atoms belong to tridymite-like regions of up to about 20
angstroms or more, that are bonded efficiently together in a
manner analogous to that found in twinned crystals” (italics
ours). This is a remarkable state in glass. It is largely formed
of angstrom-size crystals which, because of their random dis-
tribution, make the matter isotropic and translucent in the
optical range, and thus glassy by usual definition.