Jökull - 01.12.1985, Qupperneq 50

of lateral features associated with subglacial water flow.
One of the major problems in interpreting landscape
history in Iceland is that while the evidence of large-
scale, sub-continental glaciation is clear all over Ice-
land, the traces of its disappearance are few and so
small-scale that they seem to bear no relationship even
to stadial glaciation.
A similar situation was apparent during an examin-
ation of the Borgarfjörður-Mýrar area of W Iceland,
(Ashwell, 1975), and it was suggested there that the
main late-Glacial process had beenengiacial and subgla-
cial water flow. This and the lack of end moraines are
readily explicable when viewed in terms of Iceland’s
volcanic and glacial history.
Rifting and volcanism have been taking place in
Iceland since the Tertiary. The presence of the Pleis-
tocene ice cover modified the volcanic processes and the
resulting structures, formed subglacially, (Kjartansson
1943), were distinctive landscape features such as table-
mountains, (van Bemmelen and Rutten 1955), known
collectively as the Móberg formation.
Volcanic activity continues today under Vatnajökull,
Mýrdalsjökull, and, sporadically, under Öraefajökull.
The result of this activity has been a complete modifica-
tion of the coastline of S Iceland for a length of some
130 km in the approximately 1100 years of historical
time, (Thorarinsson, 1974) by mainly volcanic jökul-
hlaups. In the case of the Grímsvötn caldera under
Vatnajökull it has been estimated, (Björnsson, 1983),
that some 10% of the total inflow of magma to the area
reaches the glacier base, that is, about 5x 106m3yr_1, and
about 3%, that is 1.5xl06m3yr"1, is erupted into the
glacier as tephra, on average in recent time. The volume
of the resulting floods, carrying Iarge amounts of solid
material of volcanic origin, is of the order of several
cubic kilometres. In the case of Katla, under Mýrdals-
jökull, Jónsson (1982) suggests that the process has
been a volcanoglacial debris flow on occasions, rather
than a jökulhlaup, because some 80% of the total
volume of up to ^xlO^m^s""1 at peak flow may be of
solid material.
It must be assumed that similar processes have occur-
red relatively frequently on a geological timescale
throughout the later Pleistocene and Holocene, and
that melting must have occurred whether the overlying
ice was Polar or Temperate since present theory sug-
gests that the pillow-lavas forming the core of the
Móberg structures were extruded into water, (van
Bemmelen and Rutten, 1955; Saemundsson, 1979).
Since it is difficult to estimate the rate at which the
Móberg structures were built up there would be little
point in trying to calculate the amount of water involved
or the distance which it could have travelled to escape
under the ice. It can be suggested, however, that in the
case of the largest structures water could have reached
the margins of the ice by flow along existing channels,
mainly pre-Glacial river valleys. Where these valleys
did not follow the most direct escape routes, normally
following the direction of ice surface slope and there-
fore of the flow of ice, (Shreve, 1972), to deeper waters
beyond the ice edge on the continental shelf, water
would flow sub-glacially across ridges or englacially
across valleys to maintain the direct flow. There is no
theoretical reason why such flow is not possible, (Röth-
lisberger, 1972), especially when the source is high up
on a plateau providing sufficient head for considerable
upward or downward movement. Under these condi-
tions esker formation can be expected in the lower parts
of rising flows, (Shreve, 1972). The process would have
accelerated as ice thickness decreased and ice margins
retreated to reduce flow distances. Once these margins
were established on land the existence of end-moraines
in valleys would be short because of the periodical
jökulhlaups and they would be more likely to have
survived on higher ground.
The Snaefell massif to the SW has clearly been the
source of both water and deposited material for Fljóts-
dalur and lower Jökuldalur. The volume of volcanic
material in the system, resting on a basement of Terti-
ary rocks at about 650 m elevation, rising to 1833 m on
Snaefell itself and covering an area of 20x12 km, is
enormous and the corresponding amount of rock and
water released during the sub-glacial building process
must have been very large, with sufficient head to be
capable of substantial transport and erosion until either
the volcanic centre became extinct or the ice cover
disappeared.
In support of this theory it should be noted that, apart
from a thin covering of unsorted material covering the
widespread eskers and perhaps representing the final
downwasting of ice, neither of the deposits of unsorted
material in Fljótsdalur and the lower part of Jökuldalur
is found in the main valley. The deposits N of Skriðu-
klaustur are marginal to the valley and also show some
signs of layering: Only the material on Fellaheiði, now
being cut by Rangá, is true moraine or drift, filling a
depression between Sandvatn stóra and Rangárhnjúkur
and well away from any water flow in the main valleys.
It appears that this upland area must have been covered
by an ice tongue under pressure, with water from the S
following its NW edge through the Tröllagjót notch into
Jökuldalur, and round its E edge past Rangárhnjúkur.
48 JÖKULL 35. ÁR