Jökull - 01.12.1977, Qupperneq 21

tion, but no counterpart in a known eruption
or jökulhlaup, could be tlie result o£ an erup-
tion in Grímsvötn that failed to trigger a jökul-
hlaup. In 1854 the caldera was but slightly
recovered after the hlaup of 1851. The next
jökulhlaup to follow, possibly accompanied by
a minor eruption, occurred in 1861, resulting
in a repose time of 7 years. During the 17th,
18th and 19th centuries the period of repose
for Grímsvötn eruptions ranged, when known,
between 5 and 12 years, 9 to 10 years being the
most common (6 out of 11 instances in the
range 5 or 12 years; Thorarinsson 1974, p. 224).
The fotir missing Grimsvötn tephras are G 1873,
G 1867, G 1861, and G 1684-85. Tliree of the
four belong to “certain” eruptions — what
causes their non-appearance is uncertain, but it
seems remarkable that three of the four missing
layers belong to consecutive eruptions — i.e. all
tephras from known eruptions in the period
from 1838 to 1875 are missing, and the only
layer that appears is the unknown one of 1854.
In general, a protracted north wind during a
Grímsvötn eruption would prevent the tephra
from reaching Bárdarbunga. Likewise, the re-
distribution of the tephra into dunes might
cause the drill to miss certain layers. When dry,
volcanic ash is blown about easily by the wind
just like snow. In this connection it is of interest
to note the non-appearance (according to the
present interpretation) of the 1903 tephra from
hórdarhyrna, a sample of which was collected
aboard a ship off Langanes NE of Iceland
(Mouritzen & Noe-Nygaard 1950), but whose
main observed bulk fell farther south, on the
highlands between Jökuldalur and Vopnafjörd-
ur (Thorarinsson 1974, p. 121). It seems very
I ikely that some of this tephra fell on Bárdar-
bunga, but subsecjuently was blown into drifts.
Lhe I> 1903 eruption, therefore, affords an in-
stance where an ash layer jjrobably jrresent is
missed by the drill.
G 1873: More tephra fell as a result of this
cruption than known with certainty for any
other Grímsvötn eruption. This applies especial-
ly for the three-day period from the 9th to the
Gth of January (1873). On the morning of the
Oth the volcanic cloucl must have far exceeded
10 km in height. Most of the tephra was carried
to the SE, but on the lOth fine ash fell both
in Eyjafjördur and Thingeyjarsýsla, in direction
across Bárdarbunga (Thorarinsson 1974, p. 91).
This tephra should, therefore, have been de-
jrosited there.
G 1867: In the 1867 eruption very little tephra
fall was observed, and only in the regions south
of Vatnajökull (Thorarinsson op. cit., pp. 80—
81).
G 1861: A major jökulhlaup occurred in Skeid-
ará in 1861, and Thorarinsson (op. cit., p. 78)
states that a “minor eruption may have accom-
panied the jökulhlaup”. In view of the non-
appearance of tephras such as the G 1873 one,
the present evidence is inconclusive as to the
question whether an eruption took place in 1861
or not, and in fact Thorarinsson (1978, pers.
comm.) considers it highly improbable.
G 1684—85: According to contemporary sources
an erujrtion occurred in Vatnajökull in 1685,
and jökulhlaups took place in the winter of
1684—85 both in Skeidará and Jökulsá. Prob-
ably this was the result of eruptions in Kverk-
fjöll or Dyngjufjöll, and in Grímsvötn (Thor-
arinsson ojr. cit., pp. 52—53). Indeed, the 25-
meter ashless stretch towards the bottom of the
drill liole is curious, representing, as it does,
almost 40 years in time.
Kverkfjöll
Two tephras in the Bárdarbunga core corre-
spond in time with jökulhlaups in Jökulsá á
Fjöllum, Kv 1711 and Kv 1717, the latter one
as a result o£ a described major eruption in
Vatnajökull. Chemically both tejrhras are oli-
vine tholeiite, comparable to layer “a” (Kverk-
fjöll ? 1477, Table II no. 12). As mentioned
earlier Thorarinsson (1950) regarded most o£
the six hlaups occurring in Jökulsá between
1655 and 1729 volcanogenic. At least two of
these (i.e. 1684 ancl 1716) were associated witli
eruptions in Grímsvötn — curiously enough
hlaups in Jökulsá often coincide with eruptions
in Grímsvötn (e.g. in 1934, Thorarinsson 1950).
Tlie remaining hlaups, in 1726 and 1729, are
not represented in the core.
JÖKULL 27. ÁR 19