Jökull - 01.12.1988, Qupperneq 60

altered glass. The middle flow is coarser grained
than the other two. Thin sections bear this out
showing samples 83017 and 83018 to be coarser
grained than the other two. All thin sections exhibit
occasional plagioclase phenocrysts and some
olivine, some of which is altered. Interstitial glass,
occasionally altered, in the groundmass is present in
all sections. The bottom flow however, being finer
grained, exhibits some compact and uncracked sec-
tions of relatively fresh material. The thin section of
83013 shows little interstitial glass. All in all, 83013
and 83022 are finer grained and look fresher than
83017 and 83018.
TABLEI. K/Ar ages from the Fl-1 borehole on
Flatey.
TAFLAI. K-Ar aldursákvarðanir í borholukjarna
Fl-1 frá Flatey.
Sample number 40 Ar j •™rad mm3/ga) 40 Ar j ■^rrad %b) K20 %wt. Age±error Ma
83013 0.0519 6.40 0.200 0.81±0.08
83017 0.0908 10.03 0.203 1.39±0.10
83018 0.0765 8.11 0.142 1.67±0.18
83022 0.0695 5.66 0.110 1.96±0.33
a) Radiogenic volume is measured in mm3/g ■ ÍCT4.
b) Ratio of radiogenic 40 Ar against total 40 Ar measured.
Error in age represents 2 standard deviations. IUGS con-
stants (cf. Steiger and Jager 1977): kp = 4.962 ■ 10_l°;
ke = 5.81 • KT11; 40K/Ktotal = 1.167 • KT4.
RESULTS
The age determinations are listed in Table I. The
ages are in agreement with the stratigraphic order.
All lava flows have been found to be reversely mag-
netized (Gunnarsson et al., 1984). One sample,
83013 collected at a depth of 2.40 m in the hole
shows an age of 0.81 ± 0.08 Ma. Two samples from
the middle flow were dated; sample 83017 from
37.60 m and sample 83018 from 40.15 m depth. The
ages yielded by these samples are 1.39 ±0.10 Ma
and 1.67 ±0.18 Ma respectively. The reasons for
the age difference are as yet unexplained. Alteration
of some of the interstitial glass seen in thin sections
may be the cause. The analytical difference may,
however, be real. It is generally assumed that argon
is uniformly distributed within each rock unit. This
is one of the fundamental assumptions for the K/Ar
radiometric method. There are, however, cases to
the contrary. Dalrymple and Hirooka (1965) showed
that inhomogeneity in both the argon and potassium
concentrations can give rise to an age spread of up
to 6% within one rock unit and a total spread of both
Ar and K values of about 20%. The data for this
lava flow in Table I indicate indeed a spread of both
K and Ar values, the K20 differing by some 43%
and the argon volume by just under 20%.
The last sample, 83022 represents the third and
lowest lava flow penetrated in the borehole at a
depth of 394 m. This sample shows an age of
1.96 ± 0.33 Ma. The analytical error associated with
this age is largely explained by an error of about 6%
in the potassium analysis. This error renders it
insignificantly different to the sample 83018. Their
stratigraphic difference is, however, undisputable.
CONCLUSIONS
Fig. 2 shows the the preferred correlation between
the geomagnetic polarity time scale as shown by
Mankinen and Dalrymple (1979) and the geological
sections from Flatey and Tjömes peninsula.
These results confirm the conclusions drawn by
Eiríksson et al. (1987) that the whole sequence was
accumulated within the Matuyama chron. The
uppermost lava flow was extruded at 0.81 ± 0.08 Ma
ago, i.e. some time after the Jaramillo subchron.
The middle flow shows ages of 1.39 ±0.10 Ma and
1.67 ±0.18 Ma This lava flow was erupted some
time between the Jaramillo and Olduvai subchrons.
The lowermost flow predates the Olduvai subchron
and shows an age of 1.96 ± 0.33 Ma. As all the lava
flows are reversely magnetized the K/Ar ages are in
agreement with the geomagnetic polarity time scale
as shown by Mankinen and Dalrymple (1979).
Fig. 2 shows a correlation between the Flatey and
Tjömes sections which indicates that the glacial hor-
izons beneath the oldest Flatey lava flow may
58 JÖKULL, No. 38, 1988