Jökull - 01.12.1983, Qupperneq 66

from its deep gorge, the erosion surface has quite a
steep southerly dip as it rises up onto Skógafjall.
Thus clearly Skógafjall was already a prominent
topographic ridge at the time of eruption of U nit D.
On the southern slopes of Skógafjall the erosion
surface is marked by a prominent dark weathering
conglomerate, the so-called „Black-Band” con-
glomerate. Pockets of conglomerate and sometimes
also bedded black tuffs are also intermittently pre-
served on the erosion surface in the canyon. At one
place conglomerate was observed filling a narrow
but deep fissure down into the underlying Unit C
rocks although the actual erosion surface at this
locality had been wiped clean ofglacial debris.
Massive basalt is relatively scarce in the lower
and in particular middle parts of this unit, but the
tops of the canyons are cut through a fairly regular
basalt flow, or perhaps series of associated flow
units, the eruption of which probably marked the
end of the episode of volcanic activity which pro-
duced this unit. Thus the ridges between the cany-
ons are largely capped with basalt and show crude
trap featuring.
Thus unit as a whole is some 100 m thick near the
mouths of the canyons but undoubtedly thickens
northwards. Eruption over a glaciated surface and
the thickness and predominance of palagonite tuff-
breccias among the volcanic products clearly indi-
cates eruption beneath a thick ice sheet. Thin beds
of black tuffs observed in places intercalated within
bedded yellow palagonite tuffs may perhaps be int-
erpreted as already cold volcanic ashes washed off
the surface of the ice sheet during the eruption. This
observation and the fact that the sandurs in the
Sólheimajökull area, which are thought to have
been formed as a result of massive glacier floods
tiggered by sub-glacial eruptions, consists of con-
siderable thickness of unpalagonitised black vol-
canic ashes and fragments, together indicate that
the palagonite forming reaction (that is the leaching
and hydration of basaltic glass) cannot have occ-
urred instantaneously or even rapidly when basalt
ashes and fragments were in contact with water or
steam but must have occurred during the slow-cool-
ing of the thick pile of hot, or at least warm, porous
and water saturated fragmented volcanic rocks sub-
sequent to the eruption (ci.Jakobsson 1978).
The basalt magma type erupted in Unit D appe-
ars to have been largely non-porphyritic, although
minute laths of plagioclase are sometimes visible in
hand specimen. Towards the bottom of the unit,
however, the basalt is rather more porphyritic with
labradorite plagioclase, augite and olivine pheno-
crysts.
The common sequence of events for a sub-glacial
eruption, such as exemplified by the Unit D, was
probably as follows. Firstly a fissure eruption ben-
eath a thick ice sheet with initially the lava flowing
sub-glacially, semi-consolidating and brecciating
due to rapid cooling in an aqueous environment as
the overlying ice sheet starts to melt. Thus the lower
part of Unit D shows irregular highly jointed lobes
of basalt doming up into the surrounding tuffs and
breccias in which the black glassy matrix of the
vesicular basalt fragments of varying sizes has sub-
sequently been marginally, or even completely, re-
placed by yellow-brown palagonite. Similarly the
black glassy skins round the large lava lobes or in
places more regular pillows have also subsequently
been either completely or partly replaced by pala-
gonite. A notable feature of this predominantly
palagonite tuff-breccia unit, shown in the steep
walls of the canyons, is how in the middle and upper
parts of the volcanic pile the tuff-breccias are
commonly at least crudely stratified. In places quite
regular stratification with broad channel structures
and even graded bedding may be seen. Thus as the
eruption proceeded it seems likely that the water
activity built up due to continued melting of the ice
sheet leading to the rapid reworking and redeposit-
ion of some of the tuffbreccias. In some places the
stratified tuff-breccias can be seen to be banked up
quite steeply against the earlier consolidated mass-
es of basalt and unstratified breccia. It is in the
upper stratified tuffs-breccias that one finds the
pockets of reworked black tuffs, thich were comm-
ented on earlier. At some stage the eruption must
have broken through the ice sheet, so that in the
final stages it seems likely that the volcanic pile was
built up above the level of the ice sheet and glacial
melt water and one had the sub-aerial eruption ofa
series of lava flow units, forming the so-called
„capping" basalt.
Unit E: Following Unit D glacial conditions seem
to have become established once again so that this
unit, again consisting largely of palagonite tuff-
breccias, was likewise formed as a result of a major
sub-glacial eruption. The basalt magma type, simil-
ar to that in Unit D, was essentially aphyric al-
though scarce olivine phenocrysts may be observed.
The erosion surface at the base ofthis unit is marked
by a thick conglomerate band, the so-called „Dipp-
ing“ conglomerate, which dips fairly steeply west-
wards down the western flank of Skógafjall. Thus
64 JÖKULL 33. ÁR