Jökull - 01.12.1967, Qupperneq 32

also where the ice i'low has ceased against very
steep slopes, as in some places in Hagafell and
Jarlhettur, but very distinct changes in colour
on the rock surface show the position of the
ice margin there. The areas overridden by the
glacier are greyish to light-greyish in colour,
but the grey-brownish to rusty-brownish weatli-
ering colour on the bedrocks is preserved in
other places. No single moraine ridge represents
the former ice margin along the SW shore of
the present Hagavatn, but a 400—500 m broad
zone, which consists of moraine material, with
numerous kettle holes, is found there. This
moraine zone has mostly been built up sub-
aqueously in the front of a calving glacier, be-
cause the present outlet of Hagavatn was block-
ed by the ice and the lake level was somewhat
higher than at present. The unusually great
masses of moraine material there have probably
been accumulated because the glacier advanced
through the lake depression pushing the un-
consolidated lake sediments together to form
that moraine zone.
The frontal moraine ridges are in most cases
just a single ridge within 5—10 m in height,
and they do not show any signs of oscillation
of the ice front. No evidence of a glacial ad-
vance is found inside tlie frontal moraine
ridges. It therefore seems most likely that Haga-
fellsjökull Eystri reached its maximum in the
last century. According to Bergthorsson (1967)
the climatic depression did not finish until in
the last decade of the 19th century and no
major increase in the temperature occurred un-
til about 1920. Therefore the retreat of Haga-
fellsjökull Eystri hardly began until after 1890
and was very slow until after 1920 (Thorarins-
son 1943).
lce-clammed marginal lakes
along Hagafellsjökull Eystri
Very great changes in the size and position
of marginal lakes and in all drainage from
them took place as Hagafellsjökull Eystri ad-
vanced and retreated again. During its maxim-
um extent four marginal lakes were formed in
the Jarlhettur Valley between the glacier margin
and the Jarlhettur Range (Fig. 3). The estimat-
ed area of each of the two largest lakes is 0.1
km2. The marginal lakes have left great numb-
ers of abandoned shore lines. Only one of the
marginal lakes reached such a height that it
was drained by a pass in the Jarlhettur Range
(Fig. 3), but it has not been drained through
it for a long time as there is only a minor
erosion channel in the pass even though the
bedrock there is very weak. While the glaciers
blocked the Jarlhettur Valley to the NW of
Stóra Jarlhetta a glacier river issued from its
NE corner and flowed SW along the Jarlhettur
Range (Fig. 3).
The size and position of Hagavatn has been
very variable during the last centuries depend-
ing on the regime of Hagafellsjökull Eystri at
each time. Fig. 3 shows the position of the lake
during the maximum extent of the glacier while
Fig. 4 shows the position of Hagavatn as in-
dependent of the glacier, but of course several
intermediate stages have occurred (Fig. 5).
Wright (1935), Kjartansson (1938), Thorarins-
son (1939, 1966) and Green (1952) have all
written about the history of Hagavatn or some
particular events of it. In the present paper
only the main points of its history, together
with some new observations I have made there,
will be summarizecl.
When Hagafellsjökull Eystri advanced in the
17th century, it reached first as far as the Nýi-
foss Col (Fig. 1) and blocked the present outlet
(Green 1952). In this way the lake level was
gradually raised until the water could escape
through the Leynifoss Col (Fig. 1) which is 9.5
m higher than the Nýifoss Col. The advancing
glacier also blocked the Leynifoss Col, leaving
the Sanddalur Col and the Mosaskard Col as
the lowest open cols from the Hagavatn basin.
The Sanddalur Col (Fig. 1) is 20.5 m higher
than the Leynifoss outlet and it shows no signs
of having been an outlet from the lake at
any time. Tlie water divide in this col is cover-
ed with uneroded and unwashed moraine de-
bris, and Sanddalur itself shows no signs of any
major stream erosion. The same can be said
about the Mosaskard Col which is situated
further to the SW. It is covered by the Lamba-
hraun lava, but no traces of surface drainage
are to be found there.
The skilful cartographer Björn Gunnlaugsson
shows on his map, dating from the years 1834—
1835, a river course from Hagavatn through the
Sanddalur Col, but the map, which is at a scale
of 1:480.000, seems not quite reliable in this
266 JÖKULL 17. ÁR