Acta naturalia Islandica - 01.02.1946, Blaðsíða 69

ORIGIN OF THE BASIC TUFFS OF ICELAND
63
in this paper, among which are found classical, and some of the
most convincing cases of Pjeturss’ “moraines”.
The real moraines of Iceland are, as far as known to the present
writer, not in any way connected with the sideromelan of the com-
posite plateau series. They might and probably do occur with younger
glassy masses, but such cases have no influence on the interpretation
of our observations.
With the rejection of Pjeturss’ interpretation of the grey con-
glomerates the glacial theory of the fonnation of the pyroclastic
masses loses its main support. But the most direct evidence against
the theory is that observation fails to reveal facts which decisively
point to an ultra-rapid cooling, whereas we find many facts pointing
to the reverse: Thick masses of compact sideromelan, sideromelan in-
fillings of vesicles in lavas and patches of this glass in the middle
of very thick lavas, among other evidences.
The common occurrence of microlites, especially of plagioclase, in
the sideromelan is also a fact which should be noted, as it not only
shows that there has been no “ultra-rapid” cooling but also that the
“separation of opaque ores” is generally not so facile as assumed by
Peacock.
The observations indicate that we are concerned with that type of
volcanism which is characterized by low temperature of the magma,
high viscosity and high water content. Extremely thick fine-grained
and partly glassy lavas are the clearest indication of high viscosity
and as to the other factors they are evidenced by the general porosity
of the rocks and their porphyritic texture, the phenocrysts often be-
ing of very high percentage.
It is pretty clear that a magma may be fluid enough to flow in a
large mass, yet, at the same time be too viscous to crystallize markedly
in the time at disposal, and in that way masses of glass would be
formed. This is, for instance, clearly evidenced by the post-glacial
lavas of obsidian in Iceland. These lavas have spread out with con-
siderable fluidity, yet, without any sign of an exceptionally rapid cool-
ing, they have solidified as glass.
Crystallization in a melt becomes extremely slow when the vis-
cosity is higher than 104—105 c.g.s. But the melt then flows with
ease. Glass is blown at a viscosity of 107 and even at a viscosity of
1010 a thick mass flows perceptibly. Take for instance a layer of the