Tertiary Volcanic Plateau. 387 
so that the stages in the process of decomposition can be studied by 
observation to the periphery. The glass becomes brown and red and 
finally a deep blood-red, this progressive change being accompanied 
by a progressive increase of refractive index. So far these changes 
do not affect the homogeneity of the glass, but this condition reaches 
a limiting point and the glass breaks down. The brown oxide 
separates into spots, some of which, under a high magnification, are 
seen to possess asteroid forms. The glass cover was taken off the 
section, which was placed in a saturated water solution of nigrocene. 
In half an hour the brown oxide was stained blue. Then the section 
was i1mmersed in strong hydrochloric acid for seven hours. The 
yellow and red colours of the glass were removed. The section was 
then reimmersed in the nigrocene for two hours, after which the 
original red and yellow elass was stained blue, the stain intensity 
being a maximum for the first-named. The class retained its shape, 
but on probing with a steel point was seen to be plastic. Thus the 
brown oxide is colloidal, and the glass owes its colour to iron oxide, 
which can be removed by strong hydrochloric acid. The glass 
fragments contain occasional crystals of felspar and pyroxene, but 
these are quite unaltered. This is even the case where the mother 
glass has been thoroughly broken down and the crystals are left in 
a matrix of brown oxide. 
The most important fact arising out of the petrological inquiry is 
the resistance offered by the minerals to the weathering agencies and 
the collapse of the glass. Helland noted how the crystals of felspar 
and pyroxene remained unaltered in these rocks (4, p. 81). The 
palagonitization process may work so completely in the glass that it 
finally breaks down, whilst the felspar and pyroxene crystals remain 
practically unaffected. Whatever the particular rationale of the 
palagonitic decomposition process may be, glass is far more readily 
attacked by it than minerals, i.e. volcanic ash than lava. Of course, 
_this follows not only from the lesser resistance offered by the 
non-crystalline material to the decomposing agencies, but also from 
its fine fragmental nature, whereby the rock is attacked in all parts 
at once. Seeing that palagonite rock will take far less time to form 
from ash than lava, the importance of determining the nature of the 
original material of the red partings is obvious. (Here we are 
speaking of the main mass of the lava. The thin frothy upper 
surface layer of some streams would succumb almost as readily as 
the ash.) 
In the rock just described the original ash character of the deposit 
is easily seen, but in most of the red partings alteration has been so 
thorough that it is very difficult and often impossible to settle the 
question by microscopical examination. If the red rock represented 
weathered lava, investigation in the field would show a passage to 
crumbling and decomposing lava surfaces, but I have hardly ever 
seen clear evidence of this in the case of a basalt covered by a red 
_ parting. As there are two or three hundred such partings in the 
Icelandic Tertiary series, detailed individual examination is out of 
the question, but further light on them can be obtained from general 
considerations. 
