Jökull - 01.12.1975, Qupperneq 21

an eruption fissure. When the eruption comes
to an end, the final filling of the fissure, at
least up to a few liundred metres from the sur-
face, will contain excess argon. This magma
forms a dyke, and the possibility exists that
this dyke wifl give a far too high K/Ar age. In
fact, a dyke of K/Ar age 260 My cuts the British
Tertiary lava flows, the age of which is found
by the same method to be 50—60 My (Mussett
et al.j 1972). As these authors point out, this
case is a warning against using dykes for dating
work. The chances are that dykes may contain
more Ar than lavas of the same age.
On the background of our above discussion
we cannot, however, go so far as to predict
that dykes should generally have been per-
manently retentive for an initial excess argon;
we have seen that argon may be lost very tho-
roughly by groundwater action.
We should also consider large intrusive bodies
in this connection. The answer for dykes and
other intrusives might possibly be similar to
the one we gave for surface lavas at the be-
ginning of this section, i. e. the excess argon
might by the crystallization largely go into the
interstitial spaces, frorn which it could be trans-
ported by percolating groundwater very soon.
But we cannot exclude the possibility that much
of the initial argon was incorporated as foieign
atoms in minerals. This means that excess argon
might give too high ages in single crystal dat-
ing as well as in whole-rock dating of large
intrusives and dykes. It appears that only in
fine-grained surface-, and shallow water glassy
igneous rocks one cair take very early loss of
the excess argon for granted. But in other re-
spects the latter material has obvious short-
comings for dating.
3. THE ARGON LEVEL IN GROUND-
WATER COMPARED WITH THE RATE
OF ACCUMULATION OF RADIOGENIC
ARGON IN POTASSIUM-BEARING
MINERALS.
In spite of the just mentioned possibilities
of excess argon, it is generally assumed that a
potassium-bearing mineral starts with zero
radiogenic argon content. In the present sec-
tion we shall do so, and point out a consequent
perplexing situation due to argon in ground-
water.
Rocks are generally soaked with interstitial
groundwater which contains atmospheric argon,
in the ratio 40Ar/36Ar = 295.5. To clarify the
perplexity of the situation, we consider a sani-
dine crystal, a potassium-rich and important
mineral, from the point of view of radiometric
dating.
By the normal composition of sanidine,
K(AlSÍ30g), potassium is 14% by weight. By a
density of 2.55 g/crn3 for sanidine, we have
0.358 g K/cm3. Of this, 40K makes out 1.19 • 10-4,
and again 11% of this produces argon. Hence,
the mass of argon-producing 40K is 4.68 • 10-°
g/cm3, and the number of such 40K atoms is
7.00 • 101(i per cm3. As the decay constant is
5.85 • 10-14/year, the annual number of 40Ar
atoms produced is 4.09 • 10° per cm3.
The absorbed atmospheric argon in ground-
water is about 0.5 cm3 gas per litre (gas re-
duced to 1 atm., 0 °C). This corresponds to
1.345 • 1016 atoms of argon per cm3 in the
groundwater. Hence, it takes 3.28 • 10° years,
or about 73% of the age of the earth, for the
sanidine to produce the same concentration of
argon/cm3, as there was all the time in the
surrounding groundwater per cm3. The radio-
genic argon in the sanidine is thus, in the early
stages, produced within a deep argon-low.
Later, the situation becomes most problematic.
Will argon tend to diffuse into the mineral
from the groundwater or the other way round?
The clue to this question lies in the equi-
librium of concentration between water and
mineral. For comparison, we may notice that
in equilibrium, there are 10 cm3 argon in the
air ancl 0.5 cm3 argon in water for a 1000 cm3
volume; the equilibrium concentration of argon
is thus 20 to 1 between air and water.
In argon equilibrium between water and
sanidine, there would most probably be far less
concentration in the sanidine. But as far as we
know, there are no measurements concerning
that equilibrium. Diffusivity of argon in sani-
dine varies by many orders of magnitude frorn
case to case in the temperature range where
such measurements have been made, i. e. 400—
1150 °C. This variability has been suggested
to be due to lattice dislocations, which are
highly irregular and unpredictable. Diffusivity
in the really significant range of 0—100 °C,
seems not to have been measured, and there is
JÖKULL 25. ÁR 1 9