Jökull - 01.12.1973, Qupperneq 51

huge, mature central volcano and some o£ these
arguments will be briefly mentioned here: Be-
sides the depression indicating the caldera, the
shape o£ Mýrdalsjökull suggests a subsurface of
an irregular mountain massif, whicli is typical
for eroded mature central volcanoes, such as
the Torfajökull area and Kerlingafjöll. Regular
ridges are otherwise characteristic for the vol-
canic zones. Acidic rock abounds both in nuna-
taks (Robson, 1956) and in the debris on the
surrounding glaciofluvial plains (Krossá and
Jökulsá á Sólheimasandi). Plutonic rocks have
also been found there. As indicated by the
name Fúlilækur (The bad smelling river) there
is subglacial geothermal activity hidden under
the ice, at least in one or more places.
KATLA AND THE JÖKULHLAUPS
Returning to the initial question: What is
Katla? Eythorsson (1945), Thorarinsson (1955,
1959), Rist (1955, 1957) and Björnsson (1970)
all discuss the position of Katla on the ice
field on the supposition of a single eruption
centre. Yet, Eythorsson and Thorarinsson hint
at the possibility of other eruption centres.
Eythorsson (1945) concludes, from incomplete
historical data on jökulhlaups in Jökulsá á Sól-
heimasandi, that they have been caused by
cruptions farther to the west in the glacier,
which Thorarinsson (1959) questions. These
authors, i.a. have thoroughly accounted for the
eruptions and jökulhlaups of Katla as far as
records allow. They accept that the eruption
°f Katla 1918 shows the position of Katla on
the glacier.
There exist very scant records on the out-
break of former eruptions. The reverend J.
'iustmann visited the glacier in the company
°f three others after the eruption of 1823 (Lopts-
s°n, 1880). Judging from his description it
seems likely that this eruption took place on
SW—NE striking fissure, not far away from the
centre of activity in 1918, but probably some-
what farther to the southwest. Other descrip-
tions of Katla eruptions are inaccurate to such
a degree that the volcanic centres remain un-
known, except that contemporary written docu-
ments state that the eruption of 1755 broke
out in two places on the snow field, though
still farther west than Katla (/. Sigurdsson’s
account, Loptsson, 1880). Ólafsson and Pálsson
(1843) in their Book of Travels recount that
there were at that time three vents. The jökul-
hlaup of 1955 produced two cauldrons in the
ice field not far from the centre of activity of
1918. In my opinion a short eruption with two
active vents took place at that time without
breaking through the glacier. All evidences
seem to indicate, that the Katla eruptions begin
on rather short fissures, where the volcanic
activity soon is constricted to one or two vents
as in Vestmannaeyjar in 1973.
Far more records are available on the Katla
jökulhlaups than on the eruption centres. All
the jökulhlaups accounted for in reliable sources
have been issued from Höfdabrekkujökull on
to Mýrdalssandur, but there are far less reli-
able records of three outbursts from Sólheima-
jökull, one in the lOth and two in the 13th
century. Although these sources can be ques-
tioned there is some evidence supporting this.
The structure of Sólheimasandur and Skógar-
sandur shows that they are built up of flood
sediments and not as the common glaciofluvial
plains. The records do not make it clear
whether all the jökulhlaups on Mýrdalssandur
did come from Höfdabrekkujökull alone or
whether a part of it came from other glaciers
in the north eastern part of Mýrdalsjökull.
There are no accounts of major jökulhlaups
in Markarfljót, but its channel north of Thórs-
mörk bears unquestionable marks of great
jökulhlaups, but that area has not been studied
for the purpose of tracing their age or origin.
Anyway, it must be considered probable that
they owe their existence to volcanic activity in
the northern part of Mýrdalsjökull.
WHAT CAN BE EXPECTED
IN KATLA ERUPTIONS?
According to the opinions set forth in this
paper, Katla is by no means a single eruption
centre, but rather there only occurs one erup-
tion at each vent. By comparing Askja and
Katla (Figs. 2 and 4), it can be concluded that
the gap south of Kötlukollar (Kötlugjá) corre-
sponds to the Öskjuop, i.e. by far the deepest
pass from the caldera. The jökulhlaup will
appear there as long as the eruptions break out
within the caldera rim.
JÖKULL 23. ÁR 49