Jökull - 01.12.1977, Qupperneq 45

be explained by simple mixing of silicic and
basaltic andesitic end-members. On the other
liand, mixing between intermediate and basaltic
andesitic liquids seems to offer a possible ex-
planation why initial silica content in the
historical Hekla eruptions is a function of the
length of the preceding repose.
To explain this observed chemical variation
Sigvaldason proposes both compositional layer-
lng in the magma column, for instance formed
by accumulation of increasingly basic liquid by
stepwise or semicontinuous partial melting, pro-
ducing liquid with silica content ranging from
74 to 65 weight %, and mixing between the
dacite and basaltic andesites to form liquid
with 55 to 65 weight % silica.
The silica content in the H4 pumice ranges
from 74% to 57%. Only five bulk analyses exist,
two from the wliite pumice (sub-units (c) and (d))
and one from each of the remaining units, for
tvhich samples were collected from the middle
°f the units at different localities.
TABLE III
Silica content in H± % sío2
L White pumice (sub-unit (c)) 73.9
2. „ „ (sub-unit (d)) 69.0
3. Greyish-yellow pumice 64.4
4. Greyish-brown pumice 61.4
5. Brownish-black pumice 57.5
4 here is no difference in colour between pu- mice with silica 74% and 69% — both are bright
white.
This implies that within each unit (or colour)
there is a gradation in chemical composition.
•Heasurenicnts of the refractive index of the
glass support this.
A detailed investigation of the chemistry and
petrology of one of the acid tephra layers of
stage 2 is now in preparation. No continuous
profile 0f Hi exists so H3, which was deposited
during more stable weather conditions, was
selected for the first part of the analytical work.
Sarnples were collected at intervals of 40 cm
fcom the same 6 meters profile where all four
units are found (Fig. 23). It was found
Fig. 23. H3 in the north bank of the ravine
Ófærugil, 8 km NW of Hekla.
Mynd 23. H3 í norðurbarmi Ófœrugils, 8 km
NV af Heklutindi. — Photo S. Thorarinsson.
that some of the pumice clasts in the three
lowermost units of H3 were striped or veined,
but no sucli clasts were found in the black,
uppermost unit. The veins are white and tlius
become less prominent as tlie pumice becomes
lighter in colour and are absent in the lower-
most 50 cm. The veins are most conspicuous
at the boundary between the brown and greyish-
pink units (Fig. 24). The appearance of the
pumice indicates that sorne kind of mixing does
indeed takc place beneath Hekla and the boun-
daries between the units where the colour
changes occur seem to be rnixing contacts.
The analyses that have been undertaken so
far are microprobe analyses of the glass com-
ponent in the H3 tephra. No attempt is made
to give a detailed account of the chemistry.
The diagrams in Figs. 25 and 26 show tlie silica
JÖKULL 27. ÁR 43