Jökull - 01.12.1983, Qupperneq 72

line normative they have been referred to the alkalic
magma series whilst those of the Eyjafjöll complex
have been taken to relate to the transitional alkalic
magma series. However, this latter assessment was
based on very limited data and there is clearly a
need for further analysis of rocks from the Eyjafjöll
volcanic system.
GENERAL DISCUSSION
The preponderance of palagonite tuff-breccias in
Units A-E suggests that probably all the volcanic
rocks in the Sólheimajökullareaand immediate vic-
inity were erupted in Pleistocene to Recent times.
The presenceofglaciated erosion surfaces and tillit-
es within the volcanic successsion supports the
commonly, although not universally, held view that
the palagonite tuffs-breccias and associated basalt
pillows, lobes or more massive flow-units are largely
the products of sub-glacial eruptions. However, at
least the lower part of Unit C appears Iikely to have
been erupted during an inter-glacial period.
Possible feeder dykes for the eruptions are rarely
seen in the Sólheimajökull area; hence there is no
direct evidence of the precise position and orient-
ations of the fissures from which these rocks were
presumably erupted. On balance it seems likely
that in this area one is in general observing rocks,
which represent the distal portions of volcanic
eruptions. However, itis noteworthy that twodykes
crosscutting Units B-E on Skógafjall and Hvítmaga
have roughly E-W trends, similar to the trends of
recent fissure eruptions in the col area between the
Eyjafjallajökull and Mýrdalsjökull ice caps.
On lithological grounds it is tempting to correlate
all the rocks in the area mapped with the Moberg
Formation (sensu stricto). No undoubted pre-
Pleistocene “plateau-type” lava flows exist in this
area. However, the distinctive, coarsely porphyritic
ankaramite of Unit A apparendy correlates litho-
logically and perhaps also in a time-stratígraphic
sense with rocks elsewhere in south-central Iceland
(notably at Hvammsmúli and Arnarklettur) which
have been previously classified as Old Grey Basalts.
However, more recently Jakobsson (1979) has inter-
preted all the ankaramites associated with the Eyja-
fjöll volcanic system to be of Upper Pleistocene age.
We certainly see no justification for classifying
the volcanic rocks in the canyon sequence to the
west of the Jökulsá river as Old Grey Basalts - as
indicated on the Geological MapofSouth-East Ice-
land (Kjartansson 1962). Our mapping has shown
that these rocks are younger rather than older than
most of the rocks immediately to the east of the
Skógafjallsgil and Jökulsá river. Neither could we
see any real justification for the recognition of the
rocks west of the Skógafjallsgil river and from just
the north of the area mapped in detail up to the col
between Eyjafjallajökull and Mýrdalsjökull as
Pleistocene Young Grey Basalts (Kjartansson 1962).
We consider that these rocks laigely correlate with
Units D and E to the east of Skógafjallsgil. These
rocks to our mind suffered considerable erosion
prior to the deposition of the “Top” conglomerate,
the latest glacial tillite. The crude trap featuring
sometimes apparent on the ridge tops in this area
has resulted from the relative resistance to glacial
erosion of the more massive basalt flow-units within
the “Moberg” succession, such as the capping bas-
alt flow-units of Unit D. However, the deeply incis-
ed river courses show that palagonite tuff-breccias
dominate the thickening volcanic succession right
up to the col, and even the most recent eruption
products there.
The indication of a gradual change in the chem-
istry of the basalt magma type erupted in the Sól-
heimajökull area during the Pleistocene-Recent
time period is of considerable petrogenetic interest,
especially in the light of the evidence both from
Iceland (Jakobsson 1972) and the Mid-Atlantic
Ridge at 45°N (Aumento 1967) that the nature of the
basalt magma type erupted varies with distance
from the actíve spreading ridge crest.Furthermore
the rate of discharge of volcanic rock appears to
decrease with distance from the active ridge crest
(Jakobsson 1972). Thus large volumes of tholeiitic
basalt characterise the active spreading ridge crest
and progressively smaller volumes of first transi-
tional alkali basalts and finally alkali olivine basalts
as one moves across the ridge flanks. From this it
seems reasonable to expect that the basalt magma
type erupted in a particular area may become in-
creasingly alkaline in character with time as the site
of eruption moves progressively away from the act-
ive spreading ridge crest. The foregoing observat-
ions and data on the volcano-stratigraphy in the
Sólheimajökull area are consistent with such a
model, as indeed are data on the Vestmannaeyjar
volcanic system (Jakobsson 1979) where earlier
transitional alkali basalt suite magmas have been
followed by genuine alkali olivine basalt suite
magmas in more recent times.
70 JÖKULL 33. ÁR