Jökull - 01.12.1983, Side 71

series oflava flows capping the predominantly pala-
gonite tuff/breccia Unit B, appears to be a pre-
cursor of the transitional alkali basalts which occur
in the succeeding units. Thus basalt samples S5, S7,
S30 and S33 are likewise all of this transitional
alkalic type, whilst the palagonite breccia S4 was
also clearly formed from an eruption of a similar
basalt magma type. Such basalts are characterised
by high Ti and Fe contents, expressed modally in
relatively high titano-magnetite contents, and high
Na^O and KoO contents which result in N'a, O +
K20:Si02 ratios which straddle the various divid-
ing lines which have been proposed between the
alkali olivine basalts and tholeiitic Iiasal Is.Jaknbssnn
(1972, 1979) has indicated that comparable trans-
itional alkali basalts and evolved magma types,
such as basaltic andesites, andesites and comenditic
rhyolites related to this magma series, characterise
the neighbouring Postglacial volcanic systems of
both Eyjafjöll and Katla. The transitional alkali
basalts of the Katla volcanic system have particul-
arly high Ti02 contents (mean value 4.47 wt.%,
range 4.01-4.72 wt.% — Jakobsson 1979) in contrast
to the rather poorly known Eyjafjöll volcanic system
where it would appear that the basalts have some-
what lowerTiQ2 contents and related intermediate
and acid rocks are much more conspicuous. This
suggests that at least the majority ofthe transitional
alkali basalts (mean TiO, value 3.78 wt.% range
2.76-4.36 wt.%) in the studied area belong to the
Eyjafjöll, rather than the Katla, volcanicsystem.
VVith normative plagioclase contents <An50 and
Thornton-Tuttle Differentiation Index values >30,
it is tempting to call certain of the transitional alkali
basalts (in particular S31, S30 and S33) basaltic
andesites or basaltic icelandites, However, their
SiQ contents appear to be rather too low for bas-
altic andesites (icelandites). These particular rocks
have compositions essentally identical to certain
lavas in Easter Island (Baker and Buckley 1974),
which they have classified as hawaiites. However,
hawaiite would seem to be a rather unsatisfactory
name for both these Easter Island and comparable
Icelandic lavas in view of their moderate normative
hypersthene contents and the fact that by definition
(MacDonald 1960) hawaiites should contain both
modal and normative andesine plagioclase, whilst
these lavas contain labradorite plagioclase pheno-
crysts.
In view of the essentially black appearance of the
“Ringing” Ash in the field and the Iack ofrhyolitic
lavas in the Sólheimajökull area, it came as some
surprise to discover that this strikingly thick tephra
deposit is of rhyolite composition. With its some-
what lower SiO, content (68.16wt.%) and higher
total alkali content, (8.75 wt.%) than typical tholei-
itic rhyolites (cf. Charmichael 1964; Sigurðsson 1970)
this rhyolite appears to have mildly alkaline afFin-
ities in keeping with its field associatíons with the
transitíonal alkali basalts.
Although there are no rhyolitíc lavas in the im-
mediate vicinity, basaltic andesites and rarer acid
lavas of, in general, similar mildly alkaline chemical
afíinities are documented (Sigurðsson 1970;Jakobs-
son 1979) in the Eyjafjöll volcanic system. In add-
ition a rhyolitic ignimbrite flow of overall similar
chemistry to the “Ringing” Ashoccursalittlefurth-
er to the north in Þórsmörk. Thus, although we
have no direct evidence of the precise site oferupt-
ion of this thick tephra deposit it seems likely that it
is related to the Eyjafjöll volcanic system. The ass-
ociation in the Sólheimajökull area of a mildly alka-
line rhyolite with transitional alkali basalts is in line
with observations elsewhere in Iceland (Sigurðsson
1970; Jakobsson 1972, 1979) and on balance would
appear to indicate an origin of the acid rock through
low pressure fractionation of the basaltic magma
held in fissure systems or magma “chambers” at
relatively shallow depth.
Volcanic rock sample S8 was collected from the
prominent east-west volcanic ridge with associated
cinder cones along the northern side of the col bet-
ween the Eyjafjallajökull and Mýrdalsjökull ice
caps, and is tho ught to be representative of the most
recent volcanic activity in the Sólheimajökull area
and its immediate vicinity. This rock has the high-
est alkali content of all the lavas analysed from the
area and has undoubted alkali olivine basalt afliniti-
es although its very low normative nepheline cont-
ent indicates that it is only slightly silica under-
saturated. Its normative plagioclase and DifTerenti-
ation Index values indicate that it is best classified
as a hawaiite (basaltic andesite) rather than a bas-
alt. In fact its compositíon is very close to that of the
type hawaiites from Hawaii (MacDonald andKalsura
1964) and not all that dissimilar (although some-
what less silica undersaturated) to the hawaiite
erupted during the 1973 Heimaey eruption in the
nearby Vestmánnaeyjar. Jakobsson (1979) has also
commented on the broad similarity in petrography
and chemistry between the recent mildly alkalic
lava suite erupted on Vestmannaeyjar and the lavas
of the Eyjafjöll volcanic system. However, as the
majority of the Vestmannaeyjar basalts are nephe-
5
JÖKULL 33. ÁR 69