Jökull - 01.12.1971, Qupperneq 59

The peneplain, cutting these rocks, almost cer-
tainly separates them completely in time from
the later Pleistocene main glaciations, as they
have been worked out in Europe and America.
Granting this, it is a mistake to try to find
a correspondence between the cold climate
signs in the upper parts of our plateau basalts
and the Pleistocene scale of major glaciations.
Returning to the lack of unconformity be-
tween the cold climate vestiges and the lower
parts of the plateau basalts, the inference is
most important. If, from our Fljótsdalsheidi
area, we proceed down along the whole pile
of basalts, the sediments are very insignificant,
mostly blown sand as is common in lava fields,
and there are no unconformities, there are no
indications of a span of erosion, according to
the extensive mapping by Walker (1959). For
this reason, the whole building up of the basalt
pile must be taken to span a geologically short
time. Estimates of jA million years have ap-
peared in the literature, and 1—2 million years
may be an upper limit. Now, Moorbaíli et al.
(1968) get for acid and basaltic lavas at the
very base of the visible pile a K/Ar-date of
12.5 ± 0.2 My, the measurements being very
consistent, and for andesitic rocks at a strati-
graphic height of 2400 m they find 11.9 ±0.3
My, a very good confirmation of the geological
inference of a short time. By multiplying this
interval by a factor 3 we would get 1.6 My for
the whole time of the formation of the plateau.
It is more correct to say that the figures in-
dicate a time span between 0.3 and 3.3 My.
If we use the figure 2.5 My, the age of the
topmost, colcl climate part of the plateau basalts
is 10 My. This gives us the badly needed tirne
for peneplanation and the valley development.
In short, complete agreement of all the funda-
mental data is obtained in this way. These
K/Ar-datings seem, therefore, most trustworthy.
But our conclusions are in obvious disagree-
nient with such K/Ar-datings which give an
age of only 1.6—3 My for the pre-peneplain
upper part of the plateau basalts. Taking the
fundamental geological facts into consideration,
the disagreement' seerns to be an internal one
between various K/Ar-datings.
Recent datings of rocks from the Skagafjord
— Vídidalur area in northern Iceland, are a
further important indication of the higher age
of the pre-peneplain plateau groups (Everts et
al. 1972). A survey of these rocks and their re-
lation to the history of valley development was
given by me (Einarsson, 1958, 1959, 1962, where
the former studies, especially by Líndal, are
quoted). Later, these rocks were mapped in cle-
tail by Everts and Koerfer (not yet published),
and in connection with that work the samples
for dating were collected.
References to my classification of the erosio-
nal phases of the plateau basalts were already
given. The strandflat of Skagi, on which rest,
first, reversely polarized lavas, has been assumed
by me to be at least 0.7 My old. Four of the
new datings give 1.05 to 1.5 My, or an average
of 1.11 My for the reverse lavas. The Höskulds-
stadanúpur lava lies on river gravel on a remn-
ant of a valley floor at 300 m above sea level.
This is very likely the 2nd 300 m valley stage.
For the age of this lava the two values 1.7 ±
0.2 My and 2.1 ± 0.6 My are given.
The Bakkabrúnir sediments have in my view
been formed after the broad valley Vídidalur
had been graded in the strandflat (or 100 m)
phase. The age of the first lavas whicii cover
these sediments is given as 2.2 ± 0.4 My, 1.4
± 0.4 My and 1.9 ± 0.3 My, 2.1 ± 0.4 My.
According to these datings, even this latest main
phase of valley erosion goes back to 1.5 — 2 My.
This seems very acceptable from the morpho-
logical point of view. It has seemed to me that
for the older valley phases and the preceding
peneplanation at least 5 times longer time is
required, which would make the topmost pre-
peneplain plateau groups something like 10 My
old, in agreement with the above considerations.
We come now, finally, to the dating results
of McDougall (McDougall and Wensink, 1966).
For normal lavas in the NG he gets the values
1.66 ± 0.07, 1.58 ± 0.05, and 1.57 ± 0.05 My.
(“Gilsá event”).
McDougall finds the values 2.51, 3.10, ancl
3.10. My for Wensink’s N2, a magnetic group at
some depth in the dipping and peneplained
plateau basalts. There seems to be no agreement
with our above conclusions. In this connection,
it may be mentioned that even very fresh-look-
ing samples from the cold climate basalt groups,
which have been sent to two different labora-
tories in the U. S. A. for dating, have been
found to be unfit, too altered for dating.
JÖKULL 21. ÁR 57