NORDIC VOLACANOLOGICAL INSTITUTE
C'OJ
GABBRO NODULES FROM A PICRI'l'IC PILLOW BASALT,
MiBFELL, SW ICELAND.
Helmi Risku-Norja
Fil.lie. avhandling i geologi mineralogi
Geol.-mineral. Inst. vid Abo Akademi
1985
t l O IW 1 C VU LC MW LOG l t :\ L
JI i ST l T lJ TL
CAHHHO IWUULES FHOt·: A PICHITIC PILLOW HASALT,
MIUFELL, SW ICELAND.
Helm1 kisku-Norja
Fil.lie. dVhundlinq j qeoloqi och mineraloqt
Geol.-mineral. Inst. vid Abo Akddcrni
1985
ABSTRACT
The gabbroic nodules from Mi6fell, Sw Iceland consist of 53%
plagioclase (An 88 mol%), 11% clinopyroxene (Fs 5, En 50, Wo
45 mol%), 7% olivine (Fo 87 mol%) and 29% interstitial
ground mass.
The nodule bearing host rock is a picritic
pillow basalt with olivine, plagioclase, clinopyroxene and
spinel as phenocryst phases. The glassy pillow rims have a
MgO content of 9.7 w% and FeOt content of 9.2 w%. The com­
position of the glass is the most primitive found in Iceland
so far. The minerals and the confining glass are apparently
close to equilibrium. The liquidus temperatures calculated
on the basis of the equilibrium compositions are about 1250
oC. There is no essential difference in the chemical compo­
sition of the nodule minerals and the host rock phenocrysts.
The
extremely
narrow
compositional variation in the
minerals, lack of zoning and the extensively overlapping
ranges are explained by crystallization in open magma sys­
tems from different batches of primitive magma from a common
source with similar composition.
The nodules could have
originated from olivine gabbro dikes formed at pressures
where crystallization proceeds along olivine and spinel,
spinel and clinopyroxene, clinopyroxene and plagioclase
equilibrium stages. Replenishment with new batches of prim­
itive magma from the depth occasionally reversed the cry­
stallization order in the nodules. The host used same chan­
nels when ascending to a slightly higher niveau where it
reached olivine-plagioclase equilibrium stage.
TABLE OF CONTENTS
1•
INTRODUCTION .•.....
1
2.
GEOLOGICAL SETTING.
5
3.
PETROGRAPHY••..•...
3 • 1 • Nodules ..
3.1.1. Plagioclase..
3.1.2. Clinopyroxene..
Olivine...
3. 1.3
Host rock.
3•2•
3.2.1. Olivine..
3.2.2. Plagioclase..
3.2.3. Clinopyroxene.
3•2• 4•
Spinel. ••..
4.
5.
6.
MAJOR ELEMENT ANALYSES..
4. 1 •
Plagioclase...
01 ivine.......
4•2•
4•3•
Clinoyroxene.
4. 4.
Spinel ........
Glass chemistry
4. 5.
..............
....
......
.....
.....
"
.....
......
.....
.......
TRACE ELEMENT ANALYSES...........•.............
EQUILIBRIUM RELATIONS AND TEMPERATURE ESTIMATES•..
6•1•
Olivine-liquid........
6. 2.
Olivine-clinopyroxene.
Clinopyroxene-liquid..
6•3•
6.4. Plagioclase-liquid.
6• 5•
7.
......
......
Summary ............
7
7
7
10
10
11
12
13
16
16
18
18
20
21
24
28
33
37
38
40
40
43
48
DISCUSSION......... .
51
AKNOWLEDGEMENTS.
60
FOOTNOTES.......
60
SVENSK SAMMANFATTNING.
61
REFERENCES
-1 -
1. INTRODUCTION
The present study is on the mineralogy and chemistry of
mantle
derived
gabbroic
Midfell, SW Iceland.
primitive
picritic
The
and
and
in
found
very
a
they may therefore provide
of
evolution
host rock from
its
are
nodules
basalt
the
constraints on
nodules
the
magmas
within
the
oceanic upper mantle that more evolved nodules do not.
The
hypothesis
tectonics
nature
ultimate
sea-floor
plate
and
spreading
assumes the creation of new oceanic crust at mid-
ocean ridges.
the
of
In recent years attention has focused
of
on
to
the magmas produced at the ridges and their
origin.
The
location
of
Iceland
across
Mid­
Atlantic Ridge axis favors the study of the processes taking
place at
crests
suggested
of
oceanic
ridges.
With
sixties, the
the
theory
question
The
reconsidered.
few
of
to
has
the
arguments
against
been
discussion
seriously
discussed
and
is now very much focused on
ridge
basalts
(MORB)
what extent they reflect chemical properties of the
underlying mantle.
of
(1928)
ocean-floor spreading, in the
the characteristics of the mid-ocean
and
Bowen
that basalts represent primary, unmodified magmas
derived from the mantle, there were
this.
When
mantle
The fundamental question is
derived
primary
magmas
the
nature
and the processes
involved in their formation and subsequent evolution.
Models on basalt petrogenesis at the ocean
be
divided
ridges
can
into two opposing categories: The proponents of
the one view point argue that the composition
of
the
most
-2-
primitive
MORB:s
magmas are
beneath
is
generated
mid-ocean
close
by
to
the
partial
ridges
at
primary magmas. These
melting
the
These magmas have been only slightly
in
the
mantle
pressures from 8-10 kb.
modified
after
their
segregation from the source region (Green and Ringwood 1967;
Kushiro, 1973; Frey et al., 1974; Blanchard
Bryan
1977;
Langmuir
Presnall et
Bougault,
al.,
et
al.,
1976;
et al., 1977; Bougault et al., 1979;
1979;
Rhodes
et
al.,
1979;
1983; Takahashi and Kushiro, 1983).
view maintains that the primary
magmas
Fuji
and
The opposing
are
picritic,
and
undergone extensive fractionation of mainly olivine to
have
yield the
generation
most
of
primitive
MORB:s.
As
to
30
of
kb
(O'Hara,
1968a; Yoder, 1976; Elthon, 1979; Green et al., 1979;
Jacques and Green, 1980; Stolper, 1980; Elthon
1980,
depth
picritic magmas the opinions diverge and the
suggested pressure region ranges from 5 to
1968,
the
1984).
Presnall
and
Scarfe,
et al. (op.cit.) maintain that the
existence of picritic liquids are of no importance in basalt
petrogenesis.
It
has
also
been
argued that even though
picritic liquids might prevail at depth, they will not reach
the
surface
because
of their high density (Sparks et al.,
1980; Stolper and Walker, 1980; Smewing, 1981).
The study of nodules could enhance our knowledge on the
processes operating in depth. If the relation of the nodules
to the host can be established this possibly will, depending
on
their
origin,
elucidate crustal magma processes, early
fractionation history of the host as well
as
the
physical
and chemical properties of the crust and upper mantle.
Many
-3-
outstandin�J studies have
been
made
ultramafic
of
nodules in continental basalts and kimberlites, the
derived
(1977),
compilation of which is found e.g. in Sobolev
and
mantle
(1979)
Meyer
have
studies
liquids,
These
and Hawkesworth and Norry (1983).
helped
in
identifying
possible
Boyd
primitive
in constructing chemical and mineralogical models
for mantle, in estimating pressure ranges for equilibriation
of nodule assembladges, in interpreting deformation textures
and in estimating velocity of eruption and volatile
of
nodule
the
bearing
represent refractory mantle fragments, are
kimberlites
and alkali basalts.
that
Nodules,
host.
most
content
undoubtly
common
in
At least one occurrence of
spinel-lherzolite nodules is also reported in
a
tholeiitic
basalt in Tasmania (Sutherland 1974).
In
oceanic
ultramafic
environment
is
xenoliths
the
reported
restricted
1970;
Imsland,
1984).
of
alkaline
basalts
Jackson and
Wright,
to
(Daly, 1965; White, 1966; Bennel, 1969;
occurrence
On the other hand mafic nodules and
megacrysts of various origin are
geographically
widespread
in oceanic tholeiites (Muir and Tilley, 1964, 1966; Nicholls
et al., 1966; Aumento, 1968; Melson et al., 1968;
et
Miyashiro
al., 1969; Melson and Thompson, 1971; Frey et al., 1974;
1974;
Bougault and Hekinian,
Rhodes
et
al.,
1979).
Donaldson
and
Brown,
The scarcity of ultramafic mantle
fragments in oceanic tholeiitic suites is explained
rare
by
the
occurrence of the primitive mantle derived basalts. In
part it also may reflect the different nature of
mantle
1977;
beneath
the
oceanic
and
continental
the
upper
crusts, the
-4-
former
plagioclase
havinq
and
peridot.itic
the
latter
peridotitic composition.
nodules
Gabbroic
previously
several
by
1973;
J'onsson,
authors
(p'orarinsson,
1978;
1963,
1971;
Pedersen
1953;
et
Jakobsson
1966, 1979;
Jakobsson
and
al • , 1982; Hoj, 1984).
et
reported
have
s�mundsson, 1967; Larsen, 1979;
Ragnarsd'ottir
been
Iceland
Kristmannsd'ottir,
Tryggvasson, 1963;
al. ,
in
1982;
Hald,
Some nodules are
interpreted as true xenoliths while others are considered to
be
cognate
with
the
confining magma. A crustal origin is
generally suggested.
Nodules from picritic rocks in Iceland
studied
mentioned
gabbro
Microprobe
discussed.
nodule minerals
in
the
of
analyses
and in the
picritic
type
the
means
to
as
the
of
above
rock
evaluate
the
mantle derived magma provided that
glass
and
basalt
As the
they
the
work
is
host
rock
are
to
may
fractionation of a
genetic
relationship
the magma and the nodules can be established.
the primary aim of
is
mineral
are reported.
early
a
magma
major elements in the
coexisting
host
Miofell
nodules are contained in a very primitive
between
been
In the following the primitive
nodules.
nature and possibly deep origin
provide
not
previously. Comparison shows unequivocally that the
nodules in Miofell are not of the same
phases
have
examine
whether
Thus
the
or
if
they
represent different magmas from a common source or
if
they
nodules
and
the
comagmatic
are derived from totally different magmas.
- �)
2. GEOLOGICAL S[TTLNC
A zone of active
Neovolcanic
and
rifting
volcanism,
recent
Zone (NVZ), transverses Iceland from north in a
N-S direction to central Iceland and then changes
to
SW
fig.
The
eastern
via
Myrdalsjokull
volcanic
zones.
zone continues from the northern part and runs
zone
to
Vestmannaeyjar
extends
Mid-Atlantic
and
The
Surtsey.
through pingvellir and the Reykjanes
peninsula to the Reykjanes Ridge and
Ridge.
The
NVZ
flanked
is
pattern.
The
age
the
connects
formations and Tertiary plateau basalts
symmetrical
direction
In central Iceland the zone is divided
into two branches, the eastern and western
western
the
by
forming
NVZ
to
Quartenary
a
roughly
of the rocks increases with
increasing distance from the NVZ so that oldest rocks, about
16 m.y., are found in NW and in E Iceland (S�mundsson 1978).
I
I
Thingvallava
I
_
I
I
n
,
I
/
/
I
/ ."
/
/
/
/
//
/ '
I
'
1
/
/
/
/�
II
/
/'
I
I
,
I
11
I
/
!
II
f
1
/ /
I
/
I
I
1
/
I
/
I
I
1
1
I
'
/
I
MIDFELL
-
/QI frill
1
/ Hen
volcanic
cen re
/
I�
(!"
/��
I J / CJ
/.}
/ / ��
-.::-
f
\,' .·'��/1�/
�
km
µ�--1�0__2�0
!_ ______
_
_
Figure 1.
Index map showing the Neovolcanic Zone in Iceland,
the fissure swarms on the Reykjancs peninsula and
the samp]c location Midfell at the Jake �ingvallavatn.
-6-
Midfell is situated on the western branch of
of
Reykjanes
system.
volcanic
easternmost
It
Peninsula.
The
the
of
is
Heng ill
five
a
part
central
of
E-W
the Hengill
volcano
is
the
fissure systems of the Reykjanes
peninsula arranged in NE-SW direction and en
an
NVZ north
echelon
along
directed plate boundary (S�mundsson 1978, Jakobsson
et al. 1978) .
subglacial
Miodfell is a hyaloclastite ridge formed in
fissure
eruptions.
It
the lake pingvallavatn (
is situated on the eastern shore of
fig.1).
hill
The
about
is
300
meters high and extends 3 kilometers in SW-NE direction, the
same
general
trend
as
observed
for
the
fissures
topography in the Hengill area (S�mundsson 1967).
consists of two units of a
slightly
different
and
The ridge
The
age.
northeastern part is a plagioclase porphyritic hyaloclastite
with
only
sparse
inclusions
olivine
phenocrysts
or
without
of chromian spine!, and with microphenocrysts of
plagioclase, olivine and spine!.
photographs
older.
with
and
field
relations
As
inferred
aerial
the southwestern part is
It is a picritic pillow lava with
phenocrysts,
from
abundant
and is always spine! bearing.
olivine
Plagioclase and
clinopyroxene are also present in most samples.
The nodules are found in the
different
basalt
in
They
two
localities about 1 km apart on either side of the
The size of the nodules ranges from few cm up to
ridge.
cm.
picritic
10
are generally round, friable and appear corroded
with interstitial ground mass derived from the host picritic
basalt.
-I-
J.
PETROGRAPHY
The petrography of the nodules and
first
described
host
the
rock
is
and compared in this chapter. The chemical
composition is discussed in the following two chapters.
3.1. Nodules
The gabbroic nodules are pieces
grained
holocrystalline
inequigranular texture.
rocks
of
medium
with
a
to
coarse
hypidiomorphic
The mineral assemblage consists
of
plagioclase, clinopyroxene and olivine. Spine! is found only
as rare inclusions in olivine and in pyroxene and within the
intervening
host
rock
material.
The
minerals
contain
inclusions of glass in some cases. The nodule minerals
show
signs of resorption by a rounded outline. The averaged modal
analysis of the nodules gives
follows:
53
the
mineral
proportions
as
% plagioclase, 11 % clinopyroxene, 7 % olivine
and 29 % interstitial ground mass.
In thin section the nodules seem to have
to
disintegrated
a varying degree or they are in a state of reaction with
in
the basalt.
Some have compact cores
intervening
material has separated the nodule minerals from
while
others
the
each other (fig. 2., 3.).
3.1.1. Plagioclase
Plagioclase is the dominating nodule mineral, up to
% of the nodule constituents.
70
One of the sampled nodules is
made almost entirely of plagioclase with only very minor
-8-
olivine.
varies
size
The
1
from
of
to
8
individual
The
mm.
Both
glomerocrysts.
the
grains
glomerocrysts
and
plagioclase grains
often
occur
in
individual grains
usually show well rounded surfaces with marginal resorption,
in
particular
those
close
to the margins of the nodules.
Oscillatory zoning is often very pronounced in the crystals.'
sometimes
inclusions,
Liquid
zonally
ubiquitous phenomenon in the plagioclases.
resorption varies.
corroded, has
penetrating
fresh
appearance,
an
The
extent
of
In some samples plagioclase is extremely
rugged
into
are
arranged,
the
outlines
grains.
only
signs
and
interstitial
material
In others the crystals have
of
resorption
being
then
rounded margins.
Figure 2.
Photograph on the gabbro nodules in the picritic
pillow basalt, as exposed at the sampling site.
Midfell, SW Iceland.
-9-
Figure 3 a).
Microscope photograph on a gabbro nodule
consisting of plagioclase, clinopyroxene and
small olivine.grains. Pronounced oscillatory
zoning in plagioclase.
Figure 3 b).
Disintegrating nodule consisting of plagio­
clase and clinopyroxene. Nicols not crossed.
-10-
3.1.2. Clinopyroxene
The
clinopyroxene
chromian
weakly
green,
a
j_ s
pleochroic
diopside ranging in size up to 1 cm.
It is always
anhedral, sometimes almost spherical. It contains inclusions
of
olivine, plagioclase, glass and probably also of spinel.
The
Twinning is common.
mineral
nodule
disintegration
and
having
intruded
along
sometimes
to
the
such
pyroxene
shows
it
is
the
the
greatest
interstitial
host
margins
the
an
and
that
amount
most
corroded
degree
of
rock
material
cleavage
planes,
only
relics
of
the
clinopyroxene are recognizable.
In some samples the clinopyroxene shows enough
continuity,
in
plagioclase
spite
of the extensive corrosion, that the
crystals
poikilitically
optical
be
clearly
can
seen
to
be
by the large rounded clinopyroxene
enclosed
grains.
3.1.3. Olivine
The nodule olivines are smaller in size (from
mm,
in
clinopyroxene.
grains
about
average
1 •5
mm)
than
0.5 to 4
the plagioclase and
The crystal habit is poorly developed as the
are squeezed in between the main crystalline phases.
The olivines contain occasional small
inclusions
of
glass
and spinel(?).
From the general appearance of the minerals in the thin
sections
it
clinopyroxene
seems
and
that
the
olivine,
three
phases,
crystallized
more
plagioclase,
or
less
-1 1-
As
simultaneously.
the
pyroxene
inclusions
contains
of
olivine, at least some of the olivine must have crystallized
The smallish subhedral olivine
to the clinopyroxene.
prior
grains in between the larger pyroxenes and plagioclases
on
the
are
other hand products of a later crystallization. The
sparse
inclusions
spinel
and
included
olivine
in
clinopyroxene constitute the first crystalline phases.
3.2. Host Rock
The host rock of the gabbroic
pillow
basalt
formed
nodules
subglacially.
is
a
picritic
The macro crystalline
mineral phases are olivine, plagioclase, chromian spinel and
The
clinopyroxene.
cryptocrystalline.
texture
mass
ground
glassy
is
It has an intersertal
to
through
microgranular
with small plagioclase laths, pyroxene, olivine and
opaque minerals and glass. The amount of macrocrysts show
great
variation
variation is
within
the
ranging
largely
pillows
due
from
to
about
the
10
to
gravitational
in
%.
the
This
settling
after the eruption (fig. 4.).
occurrences have been described earlier from
picrite
50
a
Similar
M�elifell
the Hengill area (S�mundsson 1967) and from the
Stapafell olivine tholeiite from Reykjanes (Matthews et
al.
1964).
host
rock
olivine,
0.5%
The average phenocryst mode of the picritic
estimated
plagioclase,
qround
by
point
1. 3%
counting
spinel,
0.1%
gives
13%
clinopyroxene
and
84.6%
mass for the samples from the glassy pillow rims and
2�% olivine, 5% plagioclase, 5% clinopyroxene, 1% spinel and
-12-
64%
crystallized ground mass for the samples taken from the
inner parts of the pillows.
Figure 4.
Photograph on the nodule bearing host pillow
basalt showing the inhomogeneous distribution
of the olivine phenocrysts within the pillows.
3.2.1. Olivine
Olivine is
the
picrite samples.
either
shows
with
pyroxene.
phenocryst
a
other
average
tendency
olivines
2-3
towards
or
In
mm.
in
the
some
cases
the
a glomerophyric growth
with
plagioclase
and/or
The grain size is usually somewhat smaller in the
samples from the glassy pillow rims.
mostly
phase
The size of the phenocrysts varies between
0.5-6 mm, being in
olivine
dominating
subhedral
The
with rounded outline,
crystal
form
is
but all gradations
-1 3-
from euhedral to
is
by
manifested
microruptured
the
as
grains
again
suggest
features implying
observed.
Euhedral
inclusion
in
microphenocrysts
marginal
also contain occasional glass
the
are
sometimes
is commonly found as
grains
olivine
to
Skeletal
phenomena.
spinel
larger
rounded
Strongly
crystallization
and
pronounced
very
crystals.
chromian
the
a
by
resorption
rapid
a
as
well
This
kinkbanded
of
occurrence
undulating extinction in some
outlines
seen.
are
crystals
anhedral
olivines show signs of mechanical stress.
the
Often
completely
olivines.
The olivines
The
inclusions.
as
and
morphology
of the olivines is revealed in the photo series in fig. 5.
3.2.2. Plagioclase
phenocrysts
Plagioclase
in
found
are
size
two
fractions. The larger megacrysts reach up to 5 mm, are often
found
as
together
glomerocrysts
with
olivine
grains are also
these
several
and/or
macrophenocrysts
outlines,
6. ) .
rich
is
The
arranged and with a frequent
too,
to
marginal
large
appearance
of
that of the nodule
are
inclusions,
sometimes
Separate
general
similar
These,
in
crystals,
pyroxene.
encountered.
plagioclase (fig.
rounded
of
subhedral
with
sometimes zonally
oscillatory
zoning.
The
smaller crystals (less than 0.5 mm) are almost euhedral
and
usually
contain
much
fewer
or
no
inclusions.
Occasionally these are found as inclusions in pyroxene.
-14-
Figure 5 a).
Euhedral olivine phenocryst with inclusions of
chromian spine!.
Figure 5 b).
Kinkbanded anhedral olivine in the picritic host.
-15-
Figure 5 c).
Figure 6.
Anhedral, resorbed olivine in the host rock.
Glomerocrystic plagioclase megacryst in the host rock.
-16-
3.2.3. Clinopyroxene
Some clinopyroxene is
sections
of
present
in
most
of
the
thin
the picritic basalt, but in the glassy rims it
is a very minor
constituent.
In
general
appearance
the
pyroxene
is similar to that in the nodules (fig. 6.) having
rounded
outlines
disintegration.
and
showing
Glomerophyric
clinopyroxene and/or olivine and/or
found.
The
size
of
the
Twinning is often observed.
grains
varying
degree
clusters
with
plagioclase
varies
are
of
other
often
from 2 to 7 mm.
The inclusions contained within
the pyroxenes are olivine, spinel, plagioclase and glass.
Figure 7. Clinopyroxene megacryst in the host rock
containing inclusions of olivine and
chromian spine!. The bigger olivine in­
clusion also has chromian spine! inclusions.
The euhedral olivine to the right contains
liquid inclusions.
-17-
3.2.4. Spi.nel
Spine! is found in the picrite
phenocryst
and
clusters
mainly
inclusions either in olivine or,
(fig.
5. '
7.).
basalt
as
phenocrysts
marginal to olivine or as
more
seldom
in
pyroxene
No spine! inclusions were found within the
plagioclases.
The spinels are in most cases
a
slight
rounding
than 0.01 to 0.6 mm.
difference
in
was observed.
the
is observed.
euhedral, but
sometimes
The size varies from less
In average it is
about
0.2
mm.
No
sizes of the different types of spinels
The color is deep brown.
-18-
4. MAJOR ELEMENT CHEMISTRY
Averages of the microprobe analyses
of
the
different
mineral phases in the nodules and in the host rock are given
in tables 1, 2, 3 and 4 and of glass in table 5.
4.1. Plagioclase
Both core
analyzed.
and
The
rim
of
composition
the
nodule
ranges
(fig.
limits
of
and
An86-90
About one third of the analyzed grains showed a
8• ) •
slight reversed compositional zoning the difference
core
were
from An83.1 to An92.5,
the
most of the grains falling between
plagioclase
between
rim being 0.6-4.1 mol%. In the rest no zoning was
observed.
In
the
plagioclase
host
picritic
ranges
8. ) .
Despite
most
of
considerable
content in nodule plagioclase is
than
in
content
the
nodule
overlap
generally
of
the
the
plagioclase
anorthite
slightly
lower
host rock plagioclase. This slight but significant
difference will be discussed later.
eight
An
from 86.0 to 90.9 mol%, this range thus
being similar to that of
(fig.
basalt
of the analyzed grains.
Zoning was observed
In most cases it is reversed
(6 grains) the difference between core and rim varying
1.5 to 3.7 mol%.
in
from
Normal zoning was insignificant (less than
1 mol%) or not observed.
- 1 ':) -
TAB L 1� 1 .
Ave l' a q c: micro pro l1 e analyses on pl .:1 qi o cl as e
minerals from Mi6fell, SW Iceland.
Si02 A1203 FeOt
An%
CaO
Na20 K20 total
---- ---------------------------------------------------
no
nodule plg:
1
46.9 32.7
46.7 33.4
2
47.3 32.9
3
4
46.4 34.0
host rock plg:
46.2 34.2
5
44.3 33.6
6
47.6 33.7
7
47.5 33.4
8
47.0 33.2
9
12
46.4 34.0
47.5 33. 5
11
0.46
0.49
0.51
0.46
16.6
17.2
1 6.7
17 . 5
1 .6 5
1 • 54
1 .67
1.42
0.02
0.01
0.02
0.02
98.3
99.3
99.1
99.8
85.0
86.1
84.7
87.2
0.49
0.47
0.53
0.38
0.49
0.46
0.40
17.7
17.6
17.6
16.7
17 . 5
17 . 5
17.6
1 • 36
1 • 47
1 • 49
1 • 36
1. 5 2
1.42
1 • 38
0.02
0.02
0.02
0.02
0.03
0.02
0.04
100.0
99.5
101.0
99.4
99.7
99.8
100.4
87.8
86.9
86.7
87.2
86.4
87.2
87.6
-------------------------------------------------------
rl
70!
60'50-
hatched area=nodule plagioclase
unhatched area=host rock plagioclase
I
40_
30-
�
20�
r
1 O>-
80
Figure 8.
85
90
95 An %
Histogram of the plagioclase composition given
as percentage anorthite for all the analyzed
samples from Miofell.
-20-
4.2. Olivine
In nodule olivines the forsterite content
136.4
to
The average for core and rim
88.8 mol% (fig. 9.).
is similar, but
within
from
ranges
.individual
grains
a
slight
very
compositional zoning is commonly observed the cores having a
Fo-content of 86.4-87.3 and the rims of 86.4-88.1 mol%. This
zoning is both normal and reversed.
The
host
rock
olivines
also
very
show
narrow
compositional variation (fig. 9.). The great majority of the
phenocrysts, both from the crystallized parts and the glassy
ground
as
mass,
well
as
the
microphenocrysts
quenched rims fall within the limits
range
thus
is
similar
from the
of 86-91mol% Fo.
for nodule and host rock olivines.
Some host rock grains had a slightly higher Fo content
the
nodule
olivines.
The total
the host rock olivines is from Fo81.8-90.7.
all the analyzed grains had
(81.8
to
85 mol%).
than
This extends the range in Fo content
of the host rock olivines few mol% higher.
of
The
a
markedly
range
Only one of
lower
Fo
content
In the olivines from the glassy pillow
rims no difference between the composition of the
core
and
rim was observed. A slight compositional zoning is, however,
frequently observed in the olivines in crystallized samples.
The
zoning
is
more
often
reversed
than
normal.
The
difference between core and rim within the individual grains
varies
from 0.3 to 3.2 mol% in the reversely zoned and from
0.2 to
1.1
strained
mol%
and
in
partly
indistinguishable from
the
normally
zoned
crystals.
The
resorbed olivines are compositionally
the
rest
( fig.
9• ) •
Fo
content
-? 1-
ranqPs
from
85.5
90.1
to
mol%,
and
minor zonin�J, when
detected, is reversed.
TABLE 2. Average microprobe analyses on olivines from
Mi6fell, SW Iceland.
no
total
Fo%
CaO NiO
Si02 FeOt MnO
MgO
----------------------------------------------------------nodule 01:
1
40.0
3
40.2
host rock ol:
5
39.3
6
39.6
39.7
7
8
40.3
9
40.4
10
39.9
11
40.3
13
40.7
14
39.8
11.5
11 • 8
0.19
0.19
47.8
46.6
0.24
0.24
0.24
0.18
100.0
99.2
88.1
87.6
11.9
11.8
12.0
11.2
11 • 5
11 • 0
10.7
12.0
11 • 8
0.18
0.18
0. 19
0.18
0.18
n.a.
n.a.
0.17
0.18
47.7
46.9
46.6
48.9
47.6
47.3
46.9
4 7.1
46.7
0.25
0.23
0.24
0.24
0.25
n.a.
n.a.
0.26
0.24
0.21
0.25
0.23
0.22
0.26
n.a.
n.a.
0.27
0.26
99.5
99.0
99.0
101.0
100.0
98.4
97.8
100.5
99.0
87.7
87.6
87.4
88.6
88.0
88.5
88.7
87.5
87.6
--------------------------------------------------------1
oo;I
90f""°
L
i
80t-
70
�
"f
50
N
Ir
L
! unhatched area=host rock ol;v1ne
hatched area.nodule olivine
crosshatched area=strained olivines
m•microphenocryst s
�
4
0
r
r
30�
20�
10
f._
,_
I
80
Fiqure 9.
85
90
95
Fo %
Histogram of the olivine composition given as
rercentage forsterite For all the analyzed
samples from Midfell.
-� ' ' ' (
4.3. Clinopyroxenc
The analyses of
narrow
chemical
nodule
clinopyroxene
(fig.
variation
show
10.
a
rather
tab 1 e
3• ) •
Compositional zoning is negligible except for sample Mid-11,
where
two cases the rims proved to be 1-2 mol% lower in
in
Fs and higher in En
than
the
cores.
consistent variation between the rims
Chromium
shows
and the cores.
Only two of the analyzed host rock grains showed
chromium
content
no
lower
(0.25-0.39 w% Cr203) and at the same time
the highest values for Fs (9-10 mol%). Also in the host rock
the
compositional
zoning
is
generally
nonexistent,
in significant or at least inconsistent, where observed.
J
En3�
r
*
*
En
1
Fs
1
15 ... Fs
10
5
1:.= clinopyroxene composition in nodules
*=clinopyroxene composition in host rock
•=-average clinopyroxene composition corrected
for Fe0!Fe,p3
*=calculated clinopyroxene composition
(Nielsen and Drake 1979)
Figure 10.
Summary of the clinopyroxene composition in the
Midfell samples plotted in the En-Fs-Hd-Di
quadrilateral. As a reference the area occupied
hy the chromian diopsides from Midfell within the
whole quadrilateral is given.
-23-
The ranges in Fs%, En% and Wo% contents
rock
clinopyroxene
and
calculated,
the
composition are seen in fig. 10.
In
general
for
the
host
equilibrium
the
chromian
diopsides show remarkably little variation as to the Fs, En,
Wo
and Cr203 content.
in
the
nodules
Essentially they also are
identical
and in the host rock megacrysts.
Chromium
content ranges from 0.49 to 1.23 w% for nodule and from 0.25
to 1.32 w% for host rock clinopyroxene.
grains have Cr203 content about
the
clinopyroxenens
1 w%.
The majority of the
Compositionally
all
fall within the limited diopside field
in the pyroxene quadrilateral (fig. 10.).
TABLE 3.
Microprobe analyses on clinopyroxenes from
Mi6fell, SW Iceland.
no Si02 Ti02 Al203 Cr203 FeOt MnO MgO CaO NaO
total Fs% En% Wo%
-----------------------------------------------------------------------
nodule cpx:
51.2 0.35
51.3 0.27
3 51.0 0.30
4 51.1 0.24
host rock cpx:
51.1 0.29
6 50.6 0.26
51.0 0.23
8 50.8 0.33
50.8 0.28
10 50.6 0.27
11 51. 7 0.27
16 51. 1
0.32
17 50.3 0.26
1
4.39
4.63
3.92
3.97
1.02
0.94
0.58
0.80
4.88
4.25
4.69
4.23
0.12
0.09
0.13
0.09
16.9
16.6
16.4
16.6
20.9
21.1
21 • 1
21.2
0.23
0.20
0.19
0.18
100.0
99.4
98.3
98.4
7.9
7.0
7.7
6.9
48.8
48.6
48.0
48.5
43.3
44.4
44.3
44.6
4.96
4.76
4.72
4.93
4.84
5.32
4. 70
4.62
5.10
1 • 17
1.06
0.90
0.98
1. 04
1 .17
1.20
0.,90
0.94
4.23
4.30
4.17
4.74
4.46
4.07
4.07
4.47
4.17
0.10
0.10
0.11
0.11
0.10
0.10
0.10
0.11
0.11
16.3
16.7
16.8
16.8
16.5
16.4
16.7
16.4
16.8
21 • 4
20.9
21.1
20.9
20.9
21.1
21.2
20.3
20.6
0.20
0.20
0.19
0.20
0.20
0.23
0.21
0.21
0.21
99.8
98.9
99.2
99.8
99.1
99.3
100.2
98.4
98.5
7.0
7.1
6.8
7.7
7.4
6.7
6.7
7.5
6.9
47.9
48.9
49.0
48.7
48.5
48.5
48.8
49.0
49.5
45.1
44.0
44.2
43.6
44.1
44.8
44.5
43.5
43.6
-----------------------------------------------------------------------
-24-
4.4. Spine!
Spine! phenocrysts, phenocrysts marginal to olivine and
spinel inclusions in olivines and in pyroxenes were analyzed
(table 4.). Because of the small
exceptionally
possible
core and the rim.
to
grain
make
size
it
was
only
a distinction between the
Where this was done, the core
proved
in
general, but not without exception, to be more chromian than
the rim, whereas FeOt content was generally slightly
higher
in the rim.
TABLE 4.
Micro probe analyses on spine ls from Midfell,
SW Iceland.
Al203 total
Ti02 Cr203 MnO
FeOt MgO
------------------------------------------------------------------
no
inclusions in olivine
separate phenocrysts
0.26
0.29
30.5
26.7
0.19
0.19
16.9
18.4
16.2
17.2
34.9
36.5
99.0
99.3
inclusions in olivine
inclusions in pyroxene
marginal phenocrysts
0.26
0.25
0.24
35.3
37.5
35.9
0.17
0.17
0.20
16.9
17.0
16.7
15.5
15.7
15.7
30.4
28.6
31.0
98.5
99.2
99.7
inclusions in olivine
inclusions in pyroxene
separate phenocrysts
0.27
0.40
0.27
33.7
33.6
30.0
0.20
0.22
0.20
17.4
19 • 1
17.5
15.9
15.2
15.3
32.0
30.4
34.7
99.5
98.9
98.0
inclusions in olivine
0.25
30.5
0.20
17.3
16.2
34.1
98.1
inclusions in olivine
separate phenocrysts
0.26
0.25
34.6
35.4
0.21
0.20
16.6
17.0
16.4
16.0
31.8
31.0
99.8
99.9
inclusions in olivine
marginal phenocrysts
10
separate phenocrysts
11
inclusions in olivine
inclusions in pyroxene
separate phenocrysts
0.26
0.25
34.0
34.9
0.19
0.21
17.3
17 • 1
15.7
15.6
31. 7
30.2
99.2
98.3
0.25
29.0
0.18
16.6
16.4
35.6
98.0
0.26
0.42
0.22
31. 3
31.8
27.8
0.19
0.20
0.18
16.8
19.0
16.5
16.6
15.6
16.4
33.8
33.1
38.2
99.0
100.1
99.4
separate phenocrysts
0.28
33.2· 0.20
16.7
16.3
32.3
99.0
5
6
7
8
9
17
14
0.24 31. 5 0.17 17.4 16.3 32.9
98.5
separate phenocrysts
-----------------------------------------------------------------
.. 2 5-
a)
b)
.------�- -�·----,
•
t
a
..>:
•
II
•
�·•
:r-·�
di\
>
C
'1 7
1·
>
o
J15
\ or' DOC
o<J:• ••• <l
•
••
>
:J
D
c
C
15
D
a
o o
•
14
--�c------��s------��-�
18
2_ 0
16
'f .
14
I
��
II
0
O'l
�
FeOt
�..:::.:=-=-L
I
<
a
g
+a
•
<J
�
I
.. · .. .. �.
1
I
>
1
j:::
c:
F �" , '.,\:, " u"=
: • • •••
�
�,·, , �
b -:,·._• �.&':!
s·�-.:Y\;:,
J
,.,
I '. ' • ,
ff, :: ·
?+
_
a•
<l
g+Fe )
0. 66
J
_l________L
0.70
0. 6 8
0.72
00
. 4
a
•
>
0.74
ll•spinel inclusions in olivine
i:::::J=,spinel inclusion!:> in clinopyroxene
�=separate spi_nel phenoc.rysts
t,=sp1nels marginal to olivine
Figure 11.
Microprobe analyses the chromian spinels from
Midfell.
In fig. 11 c) the spinel analyses are
calculated to 32 oxygens an 24 cations, total
iron is redistributed in proportions required for
the spinel stochiometry (Finger, 1972).
The chemical
v d r i .1 h l c
e c, p c c i
.t
1
n
composition
contrast
l
()
of
other
the
m1
spjnels
s
ncral
l I y t r u c f o t' t h c !\ l ? 0 J I C I' ? Cl ] r " t 1 o ,, ti 1 c il ·� h o 11 :;
hi qh 1y
This
.-1
1 .,_)·
lar�Je varinti.on (fiq.
constant
11
ratio
Th c t·I q O IF e O t
a.).
is
and the variation is mainly due to the FeO content
(fig. 11 b.). The chromium content varies from 21.8 to
(micro)phenocrysts,
weight% in the separate
from
27.S
to
the mar ginal phenocrysts and from 30.0 to
in
weight%
39.S
from 22.1 to 36.S
weight% in the inclusions within olivines,
35.9
more
43.2 weight% in the inclusions within the pyroxenes.
a)
:.....
•
0
I
�
r' 1
C') !
r
•
D
• •
0.09
0.08
C>
•
(I)'
u..
!�-
•
;��
c.,I >
u..
•
•
0.66
•• • 0.05
0.68
•
•
<l
0.72
0.70
0.04
II
0.74
D
b)
•
I
•
•
•
0 ::J
• •• •
0 • •••
s.·
D
>
D
<
c
D
;;,O O
C>
�I
<l
Cl
D
•
0.5
>J'
•
I'! "4•
•
i • •
D
D
C>
••••
• •
•
C> Cit, C>
•
•
<l :- .':A
= �i.
,
..
.
•
. ..
•
•<tJ<]
,,,,. •
•
0.70
0.68
0.66
:..
•
• C> • a
�.
....
0.72
.,
;,
0.4
C>
0.3
C>
D
c)
•
C')
u..
t
•
D
...... ,
I
•
;:
•
2
.• ...•
•
< 0
:::
<
0.50
<
D
<
0 ofo
0.68
•
0.40
0.30
<
0.66
0.45
0.35
....
Fiqur1: 12.
0.06
<J
• ,1
!,>
•
0.07
<l
�
D
0.70
•
0.72
<:
-,0. 2 f.i
0.74
symbols as in fig. 11.
Microprobe analy ses on the chrc ··.:_Jn sp:i.nels from
Minfell.
The analyses are calculated to 32 oxy­
qens nnd 24, total iron redistributed between FeZ+
and Fe3+ in stochiometric proportions (Finqer, 1972)
-27-
6.0
5.8
i
5.4
I
5.2
Mg;_ 2+
i Fe
2.0
2.2
2.8
2.6
2.4
10
9
Al
Cr
4
Figure 13.
5
7
6
symbols as in fig.11. and 12.
Chromian spinels from Miofell. The microprobe
analyses recalculated as in fig. 11 c) and 12.
The ideal sums for normal spinel stochiometry
Mg+Fe2+=8 and Al+Cr=16 are shown by the lines
in fig. 13 a) and b).
Spinel inclusions in the pyroxenes are slightly
in
Fe
and
Ti
than the other spinels (fig. 11 b., 11 c.).
The range in chromium content also extends higher.
trend
richer
is
obvious
The same
between
the
trivalent cations is plotted against the ratios between
the
general
divalent
cations
(fig.
when
12• ) •
the
ratios
The difference is small but
consistent throughout when all the analyses are
There
considered.
is, however, extensive overlap when all the grains of
different spinel types are considered.
clinopyroxenes
In spite of this the
that contain spincl inclusions are identical
in chemical composition with those without spinels.
-28-
In fig. 13. the substitution of Mg and Fe2+ and
sums,
8
respectively
Even
structure.
Al
The Miofell spinels are very close to the
and Cr are shown.
ideal
of
though
for
16,
they
the
normal
spinel
deviate slightly from these
sums, the slope is exactly the same and continuous,
a
fact
which indicates a strict stochiometric control.
4.5.
Glass chemistry
The glassy pillow rims
black
in
brown,
hand specimen.
translucent,
tachylytic
glass
the rims and
contain
porous
and
friable,
and
In thin section the glass is pale
isotropic
is
within
are
sideromelane.
Opaque,
sometimes found in the inner parts of
the
pillows.
The
analyzed
glasses
phenocrysts and microphenocrysts of olivine, spinel
and frequently of some plagioclase whereas clinopyroxene
is
only rarely found. The amount of phenocrysts in the analyzed
samples is in average about 16%.
TABLE 5.
Average microprobe analyses on the glass
phase from Miofell, SW Iceland.
P205 tot
no
MgO
Na20 K20
CaO
Si02 Ti02 Al203 FeOt MnO
---------------------------------------------------------------------13
14
14
15
48.4
48.5
48.7
48.5
0.76
0.76
0.76
0.80
15.7
15. 5
15.5
15.4
9. 21
9.17
9.21
9.12
0.1 6
0.18
0.17
0.17
9.74
9.74
9.58
9.62
15.4
15 • 5
15.3
1 5. 2
1.54
1 • 54
1.55
1.57
0.00
0.01
0.01
0.00
0.04
0.04
0.03
0.04
101 . 0
100.9
100.8
100.4
-----------------------------------------------------------------------
-29-
The glass analyses are listed in table 5.
and
In figures 14
15 they are compared with the glasses from other areas.
The glass is shown to represent the most basic and the
primitive
in
found
Iceland.
The
low
FeOt/MgO ratio of
0.94-0.95 and FeOt/(FeOt+MgO) of 0.48-0.49 and the very
1<20
content
low
of less than 0.01 weight% and the low contents
of Ti02 (0.76-0.80 weight%)
indicate
most
and
P205
(0.03-0.04
primitive nature of glass.
the
weight%)
The exceptionally
high content of CaO and Al203, both exceeding 15 weight% and
the
being close to unity, is worth pointing out as a
ratio
distinguishing property of the Mi6fell magma.
e
Icelandic basaltic glasses (Makipiia, 1978)
• glass analyses on Krafla lavas
(Granvold and Makip!ia, 1978)
,4alkali basaltic glasses (Thy, 1982 )
*glass analyses from Midfell ,this study
6
..
4
0
<ON 2
z
0
>::
.
. ,. .,.;.,,,,.
. _., ..
•
..
46
•
***
46
50
52
SiO�
Figure 14.
Total alkalies versus silica. The Hawaiian
dividing line from MacDonald and Katsura (1964).
--i
-30-
-··-··,-·
�) 4r------·---y· ----
.' • •• •
., ..•.
...
..
••
J "' ,�
·�: �·
.., ..
.. � ..
f1.,
°
l
l
.. .-'-�"�-�-----11·
.. ..
..
...
..
.. ..
..
13
..
t4
'2I ..
r
"I
FeOI
.
.. , ....
...
�·
.. ..
... ... ...
...
...
'......
.
•
.
:. •
....
..
,c
.
..
•
MnO
14
. . .. .
12
Cao
10
�.
.....
..
.. :{.. .t,,
••
• A. ..
"'
., ":.. ·. �··
.
. .... ..
• Do ... .
I
.. ..
••
...
"'
.
........ . ..
a •
.. .... ... .....
..
,.. ...
0. 15
,11
...
...
..
0 8
.
..
....
.....
..
.. ..
• \ ••
•
• 6 ..
... ··�
0.2
••
,.;,,,,,,.,.
8'11
,.
,o
0·0� ----�4=-----� �---- �--�.....
e
e
2
.
.
.
MgO
i
�
---- _L____,
8
ll'
symbols as in fig.
Figure 1 5 •
.
•
1.0
...
...
..
0. 20
......:....
16
,
.. .. .. ... : .....
..
0.25�
.. .. ...
... • • • . .
...
...
0.30
....."
a •
4
..
...
0
.. 1,
..
,• i I
II
1.0
e------'---.,----'-----_.._----�
4
. •. . , .
•.. ....... 41
g
1. 5
.. ....
.. .....
I
.. .... ..
.. .... ..
.. ...........
• "'lo\.
l--c---L------'----�- ----,
..
... ·r· ,,
:�
..
14.
The homogeneous and
MgO variation diagrams.
primitive composition of the MicHel 1 qlass is
compJred with available data on other Icelandic
basaltic qlasscs.
- 31 -
On MgO variation diagrams the
the
general
trends,
Mi6fell
glasses
for
characteristic
the
basalts in Iceland (fig. 15.), i.e. Al203 and
with
increasing
MgO,
decrease with MgO.
alkalies
For
are
tholeiitic
CaO
increase
FeOt P205, K20, Na20 and Ti02
This is also clearly seen when the total
plotted against the Si02 content (fig.
comparison
compositional
while
follow
and
to
trends
get
idea
Icelandic
in
following analyses
have
basaltic
Makipaa
glasses(
an
been
on
in
glass
general
the
glasses
basaltic
included
1978a),
the
14.).
the
diagrams:
analyses on the
Krafla lavas ( Granvold and Makipaa 1978)
and
the
The major element
alkali
basaltic glasses (Thy 1982).
analyses
analyses of the Mi6fell glass show no significant
that
suggesting
the
on
variation
magma was homogeneous on eruption. No
correlation between the amount of phenocrysts and the
glass
composition was observed.
The
results
plagioclase
remarkably
show
the
in
both
that
or
it
is
clinopyroxene
very
minor
zoning
There is no significant difference
primitive
chemical
minerals.
On the basis of similar appearance
can be
seen
and,
in
composition
recalling
a
state
of
is
almost
and inconsistent where
observed.
composition
and
nodules and in the host rock are
Compositional
homogeneous.
nonexistent
olivine,
nodule
in
and
the
very
host rock
and
chemical
that the nodules in some cases
of
disintegration,
the
host's
glomerocrysts as well as clinopyroxene, plagioclase and some
of olivine megacrysts are most likely xenocrysts originating
from the more fully disintegrated nodules. This is confirmed
-32-
when the Middfell samples are compared with
in
rocks:
picrites
and
olivine
spinel
other
are
picritic
obligatory
phenocryst phases, plagioclase is rare and clinopyroxene
not
is
generally found (Blanchard et al., 1976; Sigurdsson and
Shilling, 1976; Jakobsson et al., 1978; Jakobsson, 1983).
The
with
exception from this
highly
variable
uniformity
composition.
are
This
the
spinels
feature remains
unexplained, but not unique as it is reported also in
sample
(Frey et al., 1974; Sigurdsson and Shilling,
suites
1976; Pallister and Hopson, 1981).
observed
other
slight
The
difference
between spinel inclusions in clinopyroxene and the
other types of
spinels,
is
already
pointed
out,
and
a
possible explanation will be given later.
The glass is primitive.
found
in
Iceland, and it has a low content of alkalies and
incompatible
elements
fractionation.
The
indicating
primitive
and of the glass suggests a
the two.
It is the most magnesian glass
low
degree
of
crystal
composition of the minerals
possible
relationship
between
-33-
5. TRACE ELEMENT ANALYSES
analyzed
Trace elements, six transition metals and Sr, were
in
to (a) further evaluate the state of evolution in
order
magma and nodules, (b) to see if there is
trace
difference
a
element composition of nodule and host rock minerals,
(c) to further test host-nodule relations and
and
in
to
( d)
their
origin
ultimately use trace element composition as an
indication of primary liquid composition.
Some of the trace
liquid
elements
have
The trace
preferentially concentrated in specific minerals.
abundances
element
in
the
Relative
changes
fractionation
are
in
trace
much
element
larger
fractional crystallization models.
Ni
these
th an
phases.
particular
concentrations during
in
case
of
major
Trace elements thus provide a sensitive test for
elements.
that
are therefore strongly
liquid
affected by the fractionation of
mineral-
(D values), and they are
coefficients
distribution
high
very
It also has been claimed
content in mantle derived olivine is uniform (Sato
1977) and that the oxidation state of Ni is insensitive over
a wide range of oxygen fugacities of basaltic magmas (Seward
1971). This makes Ni especially suitable when
fractionation
history is to be deciphered.
The elements Ni, Cr,and Co, preferentially included
the mafic minerals, and the
r1nd
Zn
were
clinopyroxene,
analyzed
spinels
in
less compatible elements Mn, Ti
with
microprobe
in
olivine,
and glass. In addition Sr in nodule
plagioclase and ground mass and in host's glass was analyzed
with
XRF.
Trace
elements
were
considered
to
be
-34-
homogeneously distributed within the liquid. This may be
oversimplification
as
analyses
of
the glass phase showed
apparent inhomogeneities within and between samples.
6
an
Table
gives the average trace element content for glass and the
ranges for the
minerals
as
well
as
the
ranges
in
the
calculated distribution coefficients.
With the exception of Co the average content
elements
in
glass
glasses (MgO
(Bougault
>
9
is
comparable
w%)
analyzed
trace
with the most primitive
from
oceanic
tholeiites
and Hekinian, 1974; Frey et al., 1974; table 6.).
The composition of the mineral phases
range
of
in trace element contents.
show
a
rather
wide
This range is not bimodal
excluding the possibility of two chemically distinct sources
for
the
nodules
and
the
host. Neither do the megacrysts
represent refractory mantle xenocrysts, because in that case
a
uniform
composition
ranges in the
cumulate
trace
origin.
It
would
element
be expected.
content
The continuous
instead
indicate
a
is also suggested on the basis of the
trace element variation that some of the olivines are not in
equilibrium with the confining liquid.
No major difference in the
nodule
trace
element
content
of
and host rock minerals can be seen, and in all cases
the ranges overlap.
The analytical data show, however, some
differences, the significance of which is discussed later:
The calculated distribution coefficients are
-for nodule olivine
Der< 1, for host rock olivine 1<Dcr>1
-for nodule olivine
Dni is usually lower than for host olivine
-35-
The incompatible clements
difference
Ti ,
and
Zn
Mn
showed
no
in behaviour when nodule and phenocryst minerals
were compared.
All the elements showed some degree of zoning
cases
when
compared.
cores
and
rims
of
the
mineral
There was no regularity in the
significance is doubtful.
in
some
grains were
zoning,
and
its
- j 6 --
TABLE 6.
glass
Average trace element content in the glass phase,
ranges in trace element contents in the minerals
and distributions coefficients.
nod.al.
------
host O l .
nod cpx.
---------
-------�-
host cpx
--------
IH 150
1* 200
2* 165
1110-1891
104!+-3097
Cr 381
1 * 470
2* 505
108-257
131-644
2236-8826 3598-8403
Co 164
1 * 45
2* 42
167-485
209-466
90-512
Ti 4045
8-258
9-272
1217-2891 1302-2779
Mn 1260
1100-1721
806-1996
549-1063
Zn 91
2* 52
12-85
11-82
23-75
81-455
96-389
168-449
1* 4600
2* 4860
1* 1317
2* 1324
595-1003
spinel
------
812-1202
542-1203
2200-2597ip
1506-1570io
1400-1764
111-562
-------------------------------------------------------------------
Dni
7.4-12.4
7.0-20.6
0.5-3.0
0.6-2.4
Der
0.28-0.67
0.34-1.69
5.87-23.2
9.44-22.1
Dco
1.01-2.96
1.27-2.84
0.55-3.12
1.02-2.74
6. 1 2
0.30-0.71
0.32-0.69
0.3-0.5
0.44-0.84
0.47-0.80
1.11-1.40
0.25-0.82
1.22-6.18
Dti
Dmn
0.91-1.43
0.64-1.58
Dzn
0.13-0.93
0.12-0.90
Sr
1*
glass(host)
ground mass(nod)
plg
Dsr
105-110
100
105-110
250
2.35
6.0-8.0
io=inclusion in oli, ip=inclusion in cpx
1*=basaltic glass from Atlantic Ocean (Frey et al., 1974)
2*=picritic glass from Atlantic Ocean (Bougault and Hekinian, 1974)
- 3 7 --
6 . E Q U I L. I B IU U M R EL A T I O 1-l S A I JD T EM P E HA T U R [ E S T I MA T E S
The compositional similarity of
rock
minerals
same
or
sugge�ts
very
that
similar
crystallization
the
pairs
host
they could be derived from a
type
of
magma.
Whether
the
of the observed minerals has taken place in
equilibrium with the confining liquid can
comparing
and
nodule
be
evaluated
by
the coexisting mineral-mineral and mineral-liquid
with
experimental
compositions
to
correspond
If
results.
the
the
analyzed
equilibrium compositions,
these pairs can then be used as geothermometers to infer the
crystallization
over
place
temperatures.
As the crystallization takes
temperature
a
crystallization
interval,
temperatures
approach
may
highest
the
the
liquidus
temperature.
The equilibrium relations have been evaluated by
coexisting
clinopyroxene-liquid,
olivine-liquid,
clinopyroxene-olivine
and
using
plagioclase-liquid
pairs
by
comparing them with experimentally determined relations. The
equilibrium composition for clinopyroxene was calculated
the
basis of the liquid composition according to the method
of Nielsen and Drake (1979).
As
differentiate
and
glass,
ground
analyzed.
and
on
in
analyses
between
mass
Fe2+
and
the
microprobe
does
not
these components in
Fe3+
clinopyroxene
were
chemically
FeO in glass was 9.03 w%, in ground mass 4.33 w%
clinopyroxene
are
the
2.89
bas.is
w%
for
oxygen fugacity calculations.
The
corrected
rnicroprobe
equilibrium, temperature and
-38-
6.1. Olivine-liquid
The MgO
according
and
FeO
contents
in
olivine
and
in
melt
to the method deviced by Roeder and Emslie (1970)
gives equilibrium constants (KO values) of 0.3052-0.3057 for
host
rock
olivine and of 0.3052-0.3058 for nodule olivine.
These values are in
good
Roeder
(op.cit.)
and
Emslie
conditions.
(MgO)ol
agreement
and
the
results
of
so indicate equilibrium
KD is calculated for the exchange reaction:
(FeO)l ------- (MgO)l
+
with
(FeO)ol
+
parentheses denote molar fractions
As the composition of olivine is thought to be dependent
the
liquid
composition
only
temperature is also dependent
Emslie,
cit.)
op.
liquid
the
phenocryst
olivine
temperatures
of
on
since
the
composition
liquidus
(Roeder
and
the temperature of equilibration can be
calculated using either
between
and
on
the
and
distribution
the
olivine.
compositions
1214
oC
for
give
of
The
MgO
or
FeO
nodule
and
estimated
average
MgO and and 1243 oC for FeO
(tab 1 e 7.).
Oxygen fugacity was calculated according to the formula
of
Roeder
and
Emslie
assumed to be ferrous.
scatter,
(op.cit.).
The values
In olivine all iron was
so
obtained
show
some
the negative logarithms of f02 for the host ranges
from 8.2507-12.0152 and for the nodules from 8.7118-12.5901.
The
average
value
of about 10 would suggest that both the
nodule and host have crystallized at
QFM buffer.
fugacities
below
the
-39-
16 a)
/
•
-0 5-
-l
/
�
f-
-1.5� /
:.:-
0,
::.
'
"
/
/
• Yo
••
•
•
/
/
0
o/
/
/0
/ /
*
lJ..
Cl)
0
- 1 .5
't
-1
log(Fe/Mg)ol
[
16 b)
0
-0.
0
,• �:�JI,,'/
'
·� ,/f
;�� 1
�s ,,,/, I
t
Liii
V
i
":-
i-21
-1
solid line -t•qui!ibrium according to Mttkip&l\(1978)
d11!1hed line equilibrium in Ouke'a experimenti,{1976)
N-new line possibly indicating eaulllbrium for Icelandic tholeitte,
Midte!I ui.mples
O 6amples from MMkiptUi's study
*
:-�
�
,
-0.5
O
0.5
log(Fe/Mg)liq
Figure 16.
Logarithmic values of Fe/Mg in clinopyroxene plotted
against those in olivine (16 a) .. The Miofell samples
coincide with continuation of the 'equilibrium line
of the Icelandic basalts. Fe/Mg in clinopyroxene
versus Fe/Mg in liquid (16 b) shows poor correlation
with earlier experiments, but the new line probably
indicates equilibrium for the whole range of Ice­
landic tholeiites.
-40-
6.2. Olivine-clinopyroxene
The
mole
ratios
FeOt/MgO
of
in
coexisting
clinopyroxenes and olivines when plotted against each other,
give KD values close to unity
samples
coincide
indicating
(Makipaa
with
equilibrium
1978).
( f ig •
the
in
a.).
continuation
Icelandic
Mi6fell
The
of
the
tholeiitic
line
basalts
The temperatures were calculated using Dni
values of coexisting olivines and
to
16
clinopyroxenes
Hakli and Wright (1967), and a reasonable
according
agreement was
attained (table 7.).
6.3. Clinopyroxene-liquid
the
When
of
FeOt/MgO
ratios
plotted
against
that
in the liquid the
Mi6fell samples show no correlation
with
the
clinopyroxene
results
of
is
extrapolated
fig.16
b.).
from
If
the
all
the
trend results (fig. 16 b.).
obtained might
represent
the line of
results
data
(op.cit.) and Mi6fell are compiled together,
reasonable
experimental
Duke (1976) and they also fall off
equilibration
(op.cit.;
coexisting
the
equilibrium
of
Makipaa
from
Makipaa
an
apparently
The new relation so
conditions
for
the
can
be
whole range of the Icelandic tholeiitic basalts.
The
estimated
clinopyroxene
using
the
glass
the
composition
iterative method of Nielsen and Drake
(1979) starting with a
Mi�fell
equilibrium
known
glass
calculated
composition is about En 53.5%,
composition.
equilibrium
Fs
6.8%
and
For
the
clinopyroxene
Wo
40%,
not
- 41 -
markedly different from the analyzed composition (fig. 10.).
The clinopyroxene therefore seems to be close to equilibrium
with
the
confining glass.
The temperature calculated with
this method is 1210 oC.
Temperature can also be estimated on basis of the known
equilibrium
Drake
composition
1979).
correlated
of
Reciprocal
with
the
of
melt and pyroxene (Nielsen and
absolute
equilibrium
constant
(K)
with
generalized linear regression equation; lnK=A/T+B, A
are
constants
and
is
temperature
T is absolute temperature.
the
and
B
K has been
calculated for the reaction:
(CaO)l+(MgO)l+(Fe0)1+(3Si02),---�(CaSi03)px+(MgSi03)px+(FeSi03)px.
Parentheses denote activities, which are for the melt components:
aSi02=XSi02/(XNaOo.S+XKOO.S+XSi02)
aMg0=XMg0/(XMgO+XFeO+XCaO+(XA101.5-XNa00.5-XK00.5)
-"-
aFeo=XFeO/
aCaO=XCaO/
- fl -
and for pyroxene:
aMgSi03=XMg/XMg+XFe+XCa+2XNa+XCr+XTi+1/2(XAl+XNa))
-"-
aCaSi03=XCa/
aFeSi03=XFe/
(X=atomic fractions)
- II -
Nielsen and Drake (op.cit.) give different
the
for
constants A and B depending on the assumed mixing model
for the melt components.
only
values
single
In
clinopyroxene
the
mixing
macroscopic 'M' site is considered here.
simple oxides in the melt are assumed
to
mix
ideally
at
When
the
- t+ 2 -
equation of linear regression to correlate K and T is:
lnK=13596/T-4.47
This mixing model
Emslie
was
first
deviced
by
Roeder
and
Nielsen and Drake (op.cit.) showed that it
(1970).
is adequate only for low alkali basalts.
When the same macroscopic 'M'
site
mixing
model
for
clinopyroxene components as assumed above and a mixing model
(1972)
modified from Bottinga and Weill
for
the
melt
is
consist
of
used, the equation is written:
lnK=14600/T-9.92
Bottinga and Weill considered the melt
network
to
forming (NF) and network modifying (NM) components.
When K, Na, Ba, Sr, and Ca, in this order, combine with
tetrahedrally
coordinated
KA102, NaA102,
BaA1204,
Al
in
SrA1204
the melt, the components
and
CaA1204
(Bottinga and Weill, op.cit.; Drake, 1976).
are
formed
These, together
with SiOZ, make up the network forming components. MgO,
and
the
CaO
Weill
assumed
to
FeO
excess after having combined with Al, are the
most important network modifiers.
and
the
In the modified
Bottinga
model (Nielsen and Drake, op.cit.) all the Ca is
act
distributed
as
between
network
CaA1204
modifier
(NF)
instead
and
CaO
of
(NM).
being
The
clinopyroxene temperatures are seen in table 5. The modified
Bottinga-Weill
mixing model gives good correlation with the
FeO temperatures for olivine (table 7.).
6 •4
P 1 ,, q i o c l a s 1' - l i q u i d
relations
Equilibrium
distribution
of
can
the
Ca, Na, Al and Si in plagioclase and melt.
some
The temperature must be estimated using
independent
using
estimated
be
other
method
plagioclase composition (Drake 1976).
of
the temperature estimated
with
olivine
and
clinopyroxene
geothermometers were used.
sr-
O
06
I
r
04- _____
-0.2
a
0-
Na
0
61
o.
-0.4
1.0
0
.,,-..--
�
C
0
4r
-0.2-
0]
o
!} 0
•
In D o I
.2r
orI
11 0
v
f
0
;:::-
�
0
0.2
--- --I
---- ---
---- ----
I
1150
1200
1250
r--: Ca
/
�
�
/
Al
,,.,.f;f"
----
/
/
_-9-
0.8
0.6
0.4
0.2
Si
o:['-
a
Q
9 �---,,,._.,�......,.��.3
. =--�5�_�8�6��
. �
65 ---=6�.6,,-�6�7
4
10 tT( ° K)
h1okr,n l1n(' C!Q11d1bt1.Jm accouJtng to Drake ( 1076)
$(,lid i.,,r. best 1,1 line for !r.clandic blU81'!!1 (Mtildpliti 1970)
v,·rt1L,11 l1nl'
,anqf! 1n In D \IOluf's tor M1dloll 9,amole·a at 12so"c
,1v1",h!P hiO IP1 M1cJ!1dl ri!1:1g,oc !;tse
, , '''!"'It'll•·•!', cui l'\1itll,1 i..tv;,,.. ((,dl11vrJlrl ,,nd
Mtik1pllA. 1978)
;
r·!hPr ..,vmh01<'.
�amp!1•s l,n.,.1 KrAfla la..;ss (Gronvold end Mblt.1pal.1 976)
(r,·drrtw:1 <1ll1tr Grtln..,n!d and Miikipri0.1978)
"'i":r
Figure 17.
Here
Distribution coefficients D for plagioclase­
qlass in the Mic.Hell samples compared with
those determined by Drake (1976), Makipaa
(197A) �nd Gr�nvoJd and Makipaa (1978).
-44-
The distribution of the major elements is compared with
the
experimental data of Drake (1975, 1976), Makipaa (1978)
and Granvold and Maki.paa (1978; fig 17.).
agreement
the values determined for Icelandic basalts
with
indicating
that
equilibrium
the
with
the
plagioclase
confining
composition
melt.
Because
negligible slope of the regression lines the
distribution
possible
close
There is a
can
major
is
in
of
the
element
not be used as a geothermometer, with the
exception
of
Dna.
disagreement
The
of
the
experimental data sets plotted clearly demonstrates that the
bulk composition has a marked effect on distribution of
the
major elements.
The
temperature
composition
base
on
element
the method of Kudo + Weill (1970) and
it's different modifications
Granvold, 1984).
major
using
estimates
(Mathez,
1973;
Drake,
1976;
The generalized form for the equation is:
l n A!E + AO'' IT = BT - C
A,B,C = constants
T = absolute temperature
B'= xCa + xAl -xNa - xSi
A = xNa xSi/xCa xAl
C
= XAb/XAn
x - atomic fraction in liquid
X - mole fraction in plagioclase
All the temperatures except for those calculated for
partial
H20
a
pressure of 5 kb are considerably higher than
the temperatures
arrived
at
by
using
clinopyroxene
and
olivine geothermometers. Only the temperatures calculated on
-45-
basis of
dry
agreement
experiments
are
compiled
in
table
7.
An
was achieved only by using equation from Grdnvold
(1984). That equation is based on experiments
and
tholeiites
demonstrates
an
obvious
on
Icelandic
compositional
dependence of temperature estimates.
Sr distribution between plagioclase and melt
temperature
estimates
when used unmodified.
can
give
The equations
of Drake (1972) and Drake and Weill (1975) give temperatures
of
about
1220
oC,
temperature of only
calculated
also
but
that
of
Sun et al. (1974) gives
about 1140 oC .
using
the
molar
coefficients according to Sun et al.
give
also
rather
Temperatures
fraction
can
be
distribution
(op.cit.),
but
these
low values, about 1110 to 1125 oC. This
discrepancy in Sr temperatures was recognized also by Sun et
al.
(1974),
and
they
regard
it
to be mainly due to the
effect of the bulk composition on the Sr distribution.
Equilibrium constant Ksr was calculated for reaction:
SrA1204(1)+2Si02(1),---�Al2Si208(plg)
Using this
complex
melt
K
value
components
and
assuming
simple
mixing
temperatures based on equation of
Drake and Weill (1975) range from 1270 to 1310 oC.
so
in
reasonable
comparable
with
geothermometers.
agreement,
the
The
of
They are
even though on high side,and
temperatures
calculated
with
other
compound equilibrium constant KSr/Ca
for the exchange reaction:
(CaA12Si208)plg+(SrA1204)1�---�(SrA12Si208)plg+(CaA1204)1
-46-
gives far too low temperatures, from about 1000 to 1150
the
Also
compound
distribution
coefficient
Dsr/ca gives
similarly low temperatures. The temperatures calculated
Dea
Kea
and
defined
values
the
in
same
are
all
1400 oC and so obviously unrealistic.
above
and
to
for
manner as the
corresponding values for Sr (Drake and Weill 1975)
close
oC.
The results from the temperature estimates are summarized in
table 7.
The use of
liquid
to
Sr
estimate
distribution
between
plagioclase
and
temperature results therefore in great
discrepancies between the (compound) equilibrium
constants,
the molar distribution coefficients and the straight forward
distribution coefficient.
This shows that the compositional
effects can not be disregarded. Involving many compositional
parameters does not seem to
discrepancy
for Sr.
can
be
is
improve
the
results
and
the
more pronounced for the major elements than
The simple weight ratio Sr distribution coefficient
used
as a geothermometer when the caution is taken
not to extrapolate the
temperature
dependence
of
the
partitioning to the rocks of very different composition.
Sr
-47-
TABLE 7: Temperature estimates with different geothermometers
Geothermometer
Temperature
Reference
1
olivine (Mg)
1210-1220
Roeder and Emslie, 1970
2
olivine (Fe)
1243-1264
3*
clinopyroxene
1210
4*
clinopyroxene
1248-1261
5*
clinopyroxene
1291-1301
6*
clinopyroxene
1270-1280
7
plagioclase
1380-1345
Kudo and Weil, 1970
8
pl.agioclase
1329-1369
Mathez, 1973
9
plagioclase
1342-1379
Drake, 1976
10
plagioclase
1243-1291
Granvold, 1984
11
plagioclase(Dsr)
1220
Drake, 1972
12
plagiocl.ase(l<sr)
1271-130
Drake and Weill, 1975
13
plagioclase(Dsr)
1138
Sun et al., 1974
14*
plagioclase(Ksr)
1106-1120
.1 5
plagioclase(Ksr/ca)
989-1153
16
pl.agioclase(Dca)
1377-1505
17
plagioclase(l<ca)
1382-1439
18
olivine/clinopyroxene
1230-1289
3
4
5
6
Calculated from the liquid composition
Modified Bottinga-Weill mixing model of melt components
Simple oxide mixing model of melt components
Same as 5, but activities for clinopyroxene components are
approximated with their molar fractions
K denotes molar fraction distribution coefficient
See text for further explanations.
14
-"-
Nielsen and Drake, 1979
"-"
II - II
ti - If
11 - II
Drake and Weill,1975
II_ It
ff - "
Hakli and Wright, 1967
-48-
6.5.
Summary
The
inronsistency
temperatures
of
reflects
the
f02, composition) which
crystallization
the
of
role of other factors (P, PH20,
probably
have
plagioclase.
an
effect
These
have
adequately counted for in the equations of
(1970),
Mathez
plagioclase
calculated
not
the
been
Weill
and
Kudo
(1973) and Drake (1976).
on
It seems that the
major element composition of plagioclase alone should not be
used
as
a
geothermometer,
independent
temperature
but
only in conjunction with
an
as
estimates
affirmative
parameter.
The plagioclase geothermometer tends to
higher
temperatures
than
noted also by others.
coincides
to
give
rest, a fact which has been
the
On the other hand the Dsr temperature
with the lowest olivine temperatures and is close
those
clinopyroxene
composition
was
used
temperatures
in calculations.
where
higher.
generally
variation
Even
speaking
( 0f
though
is
1-2
the
very
w%)
clinopyroxene
homogeneous,
in
the
the
liquid
When clinopyroxene
analyses are corrected for FeO, the calculated
are
somewhat
FeOt
temperatures
composition
there
is
content.
In
some
the
composition sensitive temperature calculations this is bound
to cause some scatter.
The
temperatures
obtained
different
with
geothermometers seem to show an internal agreement accepting
the difference of few tens
temperatures.
This
shows
of
degrees
j_ n
the
calculated
that the crystallization of the
different mineral phases has taken place
within
a
limited
-49-
range.
temperature
The liquidus temperatures obtained are
close to 1250 oC.
Higher crystallization temperatures than 1250 oC may be
inferred by comparing Dmn values for olivines from this work
with those
Within
experimentally
temperature
determined
interval
(1980).
Makipaa
1125-1250 oC Makipaa reported
Dmn values consistently higher than
with
by
As
1 •
decreases
Dmn
increasing temperature the negative Dmn values in some
of the Miofell samples
might
indicate
crystallization
at
temperatures above this upper limit.
The fundamental question is whether the phenocrysts are
in
equilibrium with the melt.
data
unequivocally
minerals
that
shows
corresponds
Comparison with experimental
to
the
composition
equilibrium
of
the
composition.
The
homogeneous composition of the minerals and
compositional
zoning
corresponding
necessarily
to
imply
the
negligible
indicate this, too. But a composition
equil.ibrium
that
the
composition
does
not
phenocrysts crystallized from
this particular magma, as they could have crystallized
also
from another, but very similar type of magma.
One of the purposes of the trace element
see,
was
equilibrium
determined.
mineral
If
phase
constants
equilibrium
concerned
(K)
to
between
coefficients
those
the
a"_,
a
reliable
(D)
experimentally
liquid
and
the
or between two coexisting mineral
phases can be ascertained, trace element composition can
used
to
if the trace element composition can be used as a test
for equilibrium by comparing distribution
and
work
be
geothermometer, since large amount of
-50-
experimental data
literature
on
is
A
available.
because
contents.
review
of
of
the
large
impossible or at least
scatter
in
trace element
It is obvious that some of the olivines
in equilibrium with confining liquid.
the
In the
the subject is given by Irving (1978).
case of Midfell this proved, however,
dubious
qood
are
not
With such wide ranges
there are always some minerals with D values that agree with
the
experimental
values. To attain reliable and comparable
equilibrium D:s experimental work on
rocks
would
be necessary.
with these
particular
Trace elements can, however, be
used to infer fractionation history from the olivines.
-51-
7. DISCUSSION
When compared
from
nodules
with
Iceland
the
the
available
Mio fell
data
clinopyroxene,
spinel.
major
with
plagioclase,
The nodule minerals show
element
qabbroic
nodules show some very
distinctive features. The mineral assemblage is
only
on
limited
and
olivine
to
probably
very limited variation in
mineral composition. They are more primitive
much higher content of both An and Fo in
plagioclase
and in olivine respectively (fig. 8., 9.). They also contain
Cr-diopside with with a restricted
the
compositionally
These
10.).
primitive
nodules
nature
augite
therefore
appear
found (fig.
usually
to
of
instead
be
more
a
of
than the other nodules previously studied
from tholeiitic and
mineral
variable
composition
olivine
assemblage
tholeiitic
the
in
Miofell
lava
suites.
nodules
The
not
does
correspond to a low pressure crystallization (Bender et al.,
Presnall
1978;
et
al., 1979; Fisk et al., 1980; Olafsson,
pers. comm.). Furthermore the
chromian
diopside,
which
Mio fell
clinopyrexene
is
a
has been reported only as a high
pressure mineral phase (Arculus, 1975; Donaldsson and Brown,
Eggler et al., 1979; Wass, 1979; Imsland, 1984).
1977;
The main difference between the mineralogy of the
basalt
host
and the nodules is the common occurrence of chromian
spinel in the host rock and it's almost total absence in the
nodules.
plagioclase
suggests
The
general
appearance
of
clinopyroxene
and
and also of some olivine megacrysts in the lava
that
they
were
not
eruption hut disintegrating or
crystallizing
reacting
with
during
the
the
Liquid.
-52-
close
The
similarity
of
these
to
minerals
nodule
the
they
suggests genetic relationship or
at least
derived
On basis of the appearance
from
a
similar magma.
and identical chemical
large
clinopyroxene
composition
and
the
that
are
and
glomerocrysts
plagioclase megacrysts as well as
some olivine megacrysts are taken
to
be
xenocrysts
which
represent pieces of more fully disintegrated nodules.
The composition of spine! is highly variable. There
a
slight,
but consistent difference in spinels included in
clinopyroxene and other type
13. ).
is
This
difference
of
spinels
(fig.
11 .,
12.,
can be due to a slightly different
evolutionary stage between the two groups of spinels.
The temperature estimates based
equilibrium
compositions
on
an
give
calculations
approximate
temperature for the formation of the nodules of
oC.
with
minimum
about
1250
The geothermal gradient in Reykjanes area is about 100
oC/km for the crust and 1 oC/km for the
Bj5rnsson, 1980).
mantle
(Beblo
and
Estimated thickness of the crust is about
8-9 km (Palmasson and S�mundsson, 1974). This indicates that
the
nodules
have
been
formed
at the approximate minimum
depth of crust mantle boundary or in the upper mantle.
The glass analyses (table 5, fig.
that
the
of
and
oceanic
glasses
analyzed
It belongs to
that
show
those
have MgO content
greater than 9 w%. Low degree of fractionation is
by
15.)
liquid that carried the nodules to the surface is
most magnesian glass found in Iceland.
rare
14.
indicated
the extremely low content of incompatible elements.
trace element content of the glass is similar to that
The
found
-53-
in
the ]east fractionated ocean-floor basaltic glasses.
content in the glass is about half of the estimated
Ni
average
for primary oceanic magmas (Sato, 1977). These features can,
of course, partly reflect the
original
source
composition
and degree of partial melting.
The high Al203 and CaO content in
are
naturally
explained
by
the
glass
Miodfell
an early enrichment when only
olivine along with spinel was fractionated. Comparison
the
mineral
phases
other
in
picrites confirms this. The
ascending host magma is thus a picritic
has
reached
with
liquid
which
only
the stage of fractionating olivine and spinel,
and continues to do so when it comes in contact with olivine
gabbro
yielding
the nodules.
It represents primitive melt
from the mantle as has been suggested to
be
the
case
for
Icelandic picrites and rift zone rocks in general (Jakobsson
et al., 1978; Maaloe and Jakobsson, 1980; Oskarsson et
1982;
Haroard'ottir,
1984).
The
primitive liquid giving
rise to the Miofell picrite is accordingly
derivation
and
is
"primary",
at
al. ,
also
of
mantle
least on a local scale.
Comparison with the Icelandic tholeiites suggests even, that
the
Miofell
samples
represent
from which the whole range
been
15.,
derived
and
which
of
the mantle derived parent,
Icelandic
tholeiites
have
has not been sampled before (fig.
18.).
The
nodules
plagioclase-(spinel)
represent
equilibrium
cumulates from the host.
element
contents,
olivine-clinopyroxene-
nodule
But,
and
as
They
stage.
judged
from
phenocrystic
are
the
olivine
not
trace
have
crystal L
very
from
L7-ed
similar
type
of
ITJiHJllla.
comagmatic, xenocrystic origin seems thus most plausible for
the nodules, and
within
the
they
stability
represent
crystallization
field
plagioclase
of
products
in the upper
mantle.
/
/
/
0.70�
/
.,,
I
,, /
/,
�\
"'
�\
"'
i
0.65
/I0
I
\I l
f
os
l
\
i
I�
ti
0.454
6
Figure 18.
\
I
\
I
0
is
I6
"\
6
6
6
t,
II
...
v
/11
6
J
::r:"'I I
6
66 6
·�
6
6
6
I
66 6
I
\
6
6
I
I 2 L'.)
6 6 I
6 "66
I
6
"
I
I
tJ
/
;REYKJA�ES 6
RIDGE
/
IC ELAND
ol \ 6
6
.......
.,,
*
I
6 /
'r- .,,
/
I
MIElFELL
I
47
I
49
48
5
51
FeOt/(FeOt+MgO) versus Si02 for Icelandic and
Reykjanes Ridge basalts and for the Miofell
samples. Fractionation trends of Skaergaard
intrusion (Wager, 1960),Thingmuli (Carmichael,
1964) and Kilauea volcano (Tilley, 1960) are
shown for comparison.
The upper pressure
nodules
J
6
I 1
I 6
\
o o�
'\.
J6
I
I
"
6
I
f
/A I
oO
0
11
1
I I
o 1 °1
I o1
/1
r•
0.5
I
/
/
\
1
o
,.,_·<:-"',;
(o
0. 60
//
"(:'-'}:
7
Cl)\
,,,_...
set
by
limit
for
the
formation
presence of plagioclase.
exceed
7-8
kb.
The
partial
the
According to
Green and Hibberson (1970) and to Fjsk (1978) this
greatly
of
can
not
H20 pressure, the
-55-
effect of other volatiles and of the minor element
within
the minerals, which are unknown in
Biggar,
1983).
et
Presnall
demonstrated that olivine,
this case, might
(Presnall
have unpredictable consequences, though
1978;
contents
al.
plagioclase,
et
al.,
(1978)
have
clinopyroxene
and
spine! can coexist only within rather narrow pressure limits
between 5-10 kb. The
fine
spinels
is
(fig.
13.)
chemically strictly
pressure
exhibited
the
by
also likely to indicate a physico-
controlled
this
Within
environment.
range the crystallization proceeds along different
lines from those
spine!,
stochiometry
at
spine!
plagioclase
atmospheric
and
pressures:
clinopyroxene
instead of olivine
and
and
olivine
and
clinopyroxene
and
olivine
and
spine!,
plagioclase and plagioclase and clinopyroxene (fig. 19.).
The highest Ni contents in Midfell olivines
to
the
Midfell
samples
basal cumulus
cumulus
is
dunite,
and
gabbro
similar
wehrl.itic
in
some
of
The whole
range
in
to that of olivines from
rocks
the
cumulus gabbro itself described from
complex.
close
highest values reported in the olivines from ocean­
floor basalts (Muir and Tilley, 1964).
the
are
interlayered
with
olivines within the
the
Samail
ophiolite
These olivines with high Ni contents represent the
earliest crystallization products from primary magmas
after
separation from mantle peridotite (Sato, 1977; Pallister and
Hopson,
Because
1981).
olivine/liquid
Ni
forming olivines.
fractionation
of
is
of
the
rapidly
Dni
high
concentrated
value
.in
the first
According to Sato (op.cit.) only
olivine
depletes
Ni
to
for
this
6-12
%
point.
/
- (J t,
-
Fr<1ctionation
not
much cxct;Cdi.n�J
Ti02 is plotted ag�inst Ng/(Mg+FeZ+)
Figure 19.
{rc(lr;,. ..n olu,, l'ro,$n.J:lel111.19711)
Phase diagrams showing crystallization paths
at 1 kb and 10 kb pressures. See text for
further explanations.
Cr content in some
greater
than
reported
of
the
in
olivines
MicHell
olivines
e•g•
partly
due
to
contamination
chromian spinel inclusions. Mainly
that
clinopyroxene
was
olivines
crystallized.
olivines
marks
the
it
drop
beginning
in
This
minute
the
far
may
undetected
indicates,
fractionating
not
The
with
is
from the Samail
ophiolite sequence (Pallister and Hopson, 1981).
be
when
(fig. 20).
10 kbar
1 atm
indicated
S'\i is also
however,
when the first
Cr
content
in
crystallization of chromian
spine! and later that of chromian diopside.
- 'j 7 -
2
1L
I
I
*:::
Midfell samples
, 5%
5%
\
0
I
0.55
Figure 20.
0.70
0.65
0.60
Mg/(Mg+Fe2 ..)
(rec:rawn after Rhodes and Dungan. 1978)
Weight% Ti02 versus Mg/(Mg+Fe2+) in the glass
indicating degree of olivine fractionation.
Der
The ranges in Dni and
olivines
1
10%
as
well
as
the
values
the
of
ranges
of
the
host
anorthite
rock
and
forsterite contents in some of the host's phenocrysts extend
slightly
higher than those in the nodule minerals (fig. 8.,
9., table 6.). This small difference might be of
as
it
indicates
that
some
of
crystallized from a magma with
composition.
pressure
a
the
host
slightly
importance
minerals
more
have
primitive
This crystallization has also taken place in a
environment,
where
olivine-spinel-plagioclase
cotecticum was reached (fig. 19.).
The nodule crystallization has proceeded
along
olivine,
clinopyroxene-
plagioclase
cotectic.
plagioclase and spinel are found as
c]inopyroxene,
an c1
olivine
is
in
the
main
However,
inclusions
in
a minor constituent of the
-58-
nodule mineral assemblage. Accordingly
taken
have
mu�; t
rlace
very
the
crystallization
close to, if not exactly, at
olivine-spinel-clinopyroxene-plagioclase invariant point.
The close similarity in the chemical composition of the
and
nodule
host
the
rock
minerals
overlapping ranges can be explained only by
open
in
magma
systems
Extremely
interval.
within
limited
compositional
negligible
rather
chemical
zoning,
extensively
with
crystallization
pressure
limited
variation
often
and
reversed
the
when
detected, of the minerals are a further evidence of this, as
is the pronounced oscillatory zoning in plagioclases.
A plausible mechanism for the formation of the
is
nodules
as follows: The primary liquid after separation from the
mantle source, fractionates
This
olivine
and
chromian
spinel.
fractionation brings the melt composition via olivine-
spinel-clinopyroxene
plagioclase
equilibrium
cotectic.
spinels are trapped
explaining
the
Some
of
inclusions
as
the
to
the
in
clinopyroxene-
last
crystallizing
clinopyroxene
thus
observed compositional trend when different
types of spinels are compared.
Repeated
replenishment
and
mixing with primitive magma from the depth prevents the melt
from
evolving
Instead,
influxes
composition
occasionally
more
towards
back
of
and
new
forth
fractionated
primitive
across
magma
phase
drive
melt
boundaries,
even causing a reversed crystallization order.
Crystal aggregates formed from the first
nodules,
compositions.
batches
i.e.
the
consist of plagioclase, clinopyroxene and olivine.
They were plastered against the walls of the magma conduits,
-59-
where
they event11ally form olivine gabbro dikes cutting the
mantle peridotite.
of
host's
The signs of mechanical stress
megacrystic
olivines
imply
in
some
that the rock from
which these olivines were derived once was in a solid state.
Most
likely
belong
they
to the nodule assemblage and are
derived from the same olivine gabbro dikes.
primitive
magma
ascend
along
the
somewhat higher levels and reach
equilibrium
stage.
During
resorbed
appearance
of
the
of
channels towards
olivine-spinel-plagioclase
ascent
gabbro is incorporated and partly
same
New batches
earlier formed olivine
assimilated
nodules.
causing
the
During the Middfell
eruption a more sudden surge of primitive magma was forceful
and rapid enough to carry pieces of the dike rock as nodules
to the surface.
-60-
A c l<IHl\\ LI D c I- r t l l r �)
lhe project
w d :_)
O sk d rs son
for
st1qqcsted
rny
two
Volcano]oqical Institute.
project
1. ()
me by I< ar 1
years
Crbnvold
fellowship
This work
is
the
at
a
and
part
Niels
Nordic
of
their
dealinq with mineral-glass equilibrium relations in
the most primitive I�elandic basalts
constraints
to
order
provide
on the melting relations of these magmas within
the oceanic upper mantle.
supervised
in
the
study and
of the manuscript.
been completed.
They also provided
the
samples,
critically read several versions
Without them the work could
never
have
This is gratefully acknowledged.
{foot note) THE ANALYTICAL PROCEDURE:
The analyses were
performed
at
the
Mordic Volcanological
Institute in
Reykjavik with an automatic AHL-SEMU microprobe.
The sample
currents used
were 25 nA for glass and plagioclase, 75 nA
for pyroxene and 90
for olivine and spinel.
Natural
minerals
were
used
as standards.
The results were
recalculaterl according to the correction method by Bence and
Albee
(196b).
Countinq time was 10-20 s for the peak and 4
s for the background.
For trace elements counting time from
40 to 100 s for the peak and 40 s for background were used.
The sample currents were similar.
Higher
acceleratinq
potential of 25 kilo volts did not improve precision and
trace element analyses were also performed mainly with an
acceleratinq potential of 15 kilo volts.
( foot note) TE Ht 1 I trn LO LY :
To avoid too
hasty
genetic
inferenc0s I have used tt1rouqhout the "ork the neutral term
"nodule" to desiqnate recoqnizable mineral aggregates, which
in appearance differ from the main mineral assemblaqe of the
host rock.
-61-
SVENSK SAMMANFATTNING: GABBRONODLJLER FRAN EN PIKRITCSK
PILLOW BASALT, MIDFELL, SV ISLAND.
En subglacial pikritisk "pillow basalt"
fran
Midfell,
SV
Island, innehJller gabbronoduler, som ar av en annan typ
'a n
de
fran
ti d ig ar e
Mineralsallskapet
Island
undersokta
gabbronodulerna.
i Miofell-nodulerna bestar av plagioklas,
klinopyroxen (kromdiopsid) och
olivin.
forekommer
Spinell
som en sallsynt inklusion i klinopyroxen eller i olivin.
Den
nodulbarande
plagioklas
och
makrokristallina
..
pillowkanterna
magnesiumrika
innehaller
h ar
olivin,
kromdiopsidisk
mineralfaser.
ar
mycket
laga
Glaset
patraffats
som
fran
Det
p!
koncentrationer
element,
vilket
spinell,
klinopyroxen
primitivt.
mycket
glaset
inkompatibla
basalten
de
ar
avkylda
mest
det
Island
av
som
och
det
alkalier
och
en
indikerar
1 a9
fraktioneringsgrad.
Den kemiska sammans�ttningen av olivin, plagioklas
klinopyroxen
ar
mycket
begr�nsad och primitiv.
och
Det finns
inte n8gon betydande skillnad i sammans�ttning mellan noduloch "host rock" -mineralerna.
V�rdbasaltens glomerokryster,
klinopyroxen- och plagioklas megakryster samt nagra av
olivinmegakryster
de
mera
grundas
tycks
fullstandigt
pa
det
vara xenokryster h�rstammande fr�n
disintegrerade
till stand
Detta
nodulerna.
likartade utseendet, den identiska kemiska
sammans�ttningen och att nodulerna i vissa fall
ett
dess
av
disintegration.
Den
sammans�ttningen av glaset och mineralen h�nvisar
de �r �enetiskt relaterade.
kan
ses
primitiva
till
att
,.
') - tl {_
I motsats till de ovriga mineralfaserna �r spinellernas
sammans�ttning
mycket
varierande.
overlappningen finns det en
mellan
svag
spinellinklusionerna
Trots
men
konsekvent
skillnad
i klinopyroxen � ena sidan och
spinellinklusionerna i olivin och skilda
a
den vidstr�ckta
spinellfenokryster
den andra. Detta forklaras med olika evolutionara stadier
av spinellerna.
I fr�ga om huvudelementen tycks mineralerna befinna sig
i
j�mvikt
metoder,
med v�tskan.
utgaende
samexisterande
Temperaturer kalkylerade med olika
fran
jamviktsammansattningen
de
mineral-mineralparen och mineral-vatskeparen
ger en liquidustemperatur pa over 1250 oC.
geotermiska
hos
gradienten
och
skorpans
betyder detta att nodulerna har
Betraktande
tjocklek
bildats
i omradet,
gransen
i
den
mellan
skorpan och manteln eller i den ovre manteln.
Sparelementsammansattningen
Variationen
ar
v·asentl igen den
-mineralen.
kontinuerlig
samma
Den
for
uppvisar
stor
och ej bimodal.
b�de
kontinuerliga
nodul-
och
variationen
variation.
Den 'ar ocksa
"host
rock"
indikerar att
mineralerna har bildats genom en kumuluskristallisation
att
i
och
fraga om sparelementen en del kumuluskristaller inte
ar i j�mvikt med v�tskan.
Mineralsallskapet i nodulerna
lagtrycksallskap.
Den
representerar
inte
ovre tryckgransen for uppkomsten av
nodulerna best�ms av plagioklasens narvaro till ca 5-10
motsvarande
ett
djup
av
15-30
km.
foljer kristallisationen en annan vag
tryck:
ett
ol ivin-sp.inell,
kb,
I detta tryckomrade
an
vid
atmosfariskt
spinell-klinopyroxen
och
-63-
klinopyroxen-plagioklas
st',jl let
i
olivin-spinel.l
for
spinell-plagioklas och plagiok.l as-klinopyroxen.
Den p3fallande likheten i den kemiska
hos
nodul-
"host
och
rock"
-mineralen,
begr�nsade kemiska variationen och
zoneringen
forklaras
genom
sammansattningen
den
den
ytterst
obetydliga
kemiska
en kristallisation i ett oppet
magmasystem inom ett begransat tryckomrade.
En sannolik mekanism for uppkomsten
Den
uppstigande
prim.ara
v"atskan
nodu.lerna
av
fraktionerar,
ar:
efter
mantelkallan, olivin och spinell tills
"
vatskesammansatttningen
nar
plagioklas-klinopyroxen
separering
fran
-kotektikum.
Fornyelse
och
blandning
med
nya
primitiva
magmapulser
fran
v'a tskan.
I
stallet
n·arheten
av invarianta punkten for klinopyroxen-plagioklas-
djupet
hindrar
haller
sig
olivin-spinell. Tillfalligtvis
magma
en
vidare utveckling av
v�tskesammans�ttningen
kan
tillflode
av
primitiv
aven fororsaka en omvand kristallisationsordning.
De
forst bildade mineralaggregaten bildar olivingabbrogangar
magmakanalerna
i
ovre
niva,
de
-kotektikum.
Harvid
olivingabbron
delvis,
utseendet
hos
nar
nodulerna.
en
nagot
olivin-plagioklas-spinell
,o
assimileras
vilket
i
Senare pulser av primitiv
manteln.
magma anv�nder samma kana.ler vid uppstigning till
hogre
i
den
tidigare
fororsakar
Mio fell
bildade
det resorberade
-eruptionen
var
tillrackligt kraftig och snabb for att kunna bara stycken av
g�ngbergarten som noduler till ytan.
REFEREMCES
Arculus, R.J., 1975:
Melting behaviour of two basanites in
the range 10-35 kbar and the effect of Ti02 on the
olivine-diopside reactions at high pressures. Carnagie
Inst. Wash. Year Book 74, pp 512-514.
Aumento, F., 1968: The Mid-Atlantic Ridge near 45 o N, II.
Basalts from the area of Confederation Peak. Can. J.
Earth Sci. 5, 1.
Beblo, M. and Bjornsson, A. 1980: A model of electrical resis­
tivity beneath NE-Iceland, correlation with temperature.
J.Geophys. 47:184-190.
Bence, A.E. and Albee, A.L., 1968: Empirical correction fac­
tors for the electron microanalysis of silicates and ox­
ides. J. Geol. 76:382-403.
Bender, J.F., Hodges, F.N. and Bence, A.E., 1978: Petrogenesis
of basalts from the project FAMOUS area: anexperimental
study from O to 15 kbars. Earth Planet.Sci.Lett. 41:277302.
Bennell, M.J., 1969: Ultrabasic and basic nodules
Azores. Proc. Geol.Soc.Lond. 1658:252-253.
from
the
Biggar, G. M., 1983: Crystallization of plagioclase, augite
and olivine in synthetic systems and in tholeiites. Min.
Magaz. 47:161-176.
Blanchard, D.P., Rhodes, J.M., Duncan, M.A., Rodgers, K.V.,
Donaldson, C.H., Brannon, J.C., Jacobs, J.W., and Gibson,
E.K. 1976: The chemistry and petrology of basalts from
Leg 37 of the Deep Sea Drilling Project. J. Geophys.
Res. 81 :4231-4246.
Bottinga, Y. and Weill D.F., 1972: The viscosity of magmatic
silicate liquids: a model for calculation. Amer.J.Sci.
272:438-475.
Bougault, H. and Hekinian, R., 1974: Rift valley in the Atlan­
tic Ocean near 36o50'N:
petrology and geochemistry of
basaltic rocks. Earth Planet.Sci.Lett.24:249-261.
Bougault, H., Cambon, P., Corre, O., Joron, J,L. and Treuil,
M., 1979: Evidence for variability of magmatic processes
and upper mantle heterogeneity in the axial region of the
Mid-Atlantic Ridge near 22oN and 36oN. Tectonophys.
55:11-34.
Boyd, F.R. and Meyer, H.O.A.,(ed.) 1979: The mantle sample:
inclusions in kimberlites and other volcanics. Litho­
Crafts Inc., Chelsea, Michigan, pp 423.
(19?.8, ·1. ed.)1956: The evoluti.on of the
i.gneous
tl.L.
rocks. Dover Puhl. ications, Inc., New York, pp 332.
Bowen,
Bryan, W.B., 1979:
Regional. variation and petrogenesi s of
ba�alt glasses from the FAMOUS area, Mid-Atlantic Ridge.
J. Petrol. 20:7-93-325.
Carmichael, I.S.E., 1964: The petrology of Thingmuli volcano,
a Tertiary volcano in eastern Iceland. J.Petrol. 5:435460.
Daly,
R.A., 1965:
The
geology
of
Proc.Am.Acad.Arts & Sci. 60:1-80.
Ascension
Island.
Donaldson, C.H. and Brown, R.W., 1977: Refractory megacrysts
and magnesium-rich melt inclusions within spinel in oce­
anic tholeiites: indicators of magma mixing and parental
magma composition. Earth Planet. Sci. Lett. 37:81-89.
Drake, M.J., 1972: The distribution of major and trace ele­
ments between plagioclase feldspar and magmatic silicate
liquid: an experimental study.
Ph.D.
dissertation,
University of Oregon, 190 pp.
Drake, M.J., 1975: Plagioclase thermometry and barometry in
igneous systems.
In: Extended abstracts: International
Conference on Geothermometry and Geobarometry, Pennsyl­
vania State University, Pennsylvania.
Drake, M.J., 1976:
Plagioclase-melt
Cosmochim. Acta 40:457-465.
equilibria.
Geochim.
Drake, M.J. and Weill, D.F., 1975: The partition of Sr, Ba,
Ca,
Y,
Eu2+, Eu3+ and other REE between plagioclase
feldspar and magmatic silicate liquid: an experimental
study. Geochim.Cosmochim.Acta 39:689-712.
Duke, J.M., 1976: Distribution of the period four transition
elements among olivine, calcic clinopyroxene and mafic
silicate liquid: experimental results.
J.Petrol. 17:
499-521.
Dungan, M.A. and Rhodes, J.M., 1978: Residual glasses and melt
inclusions in basalts from DSDP Legs 45 and 46: Evidence
for magma mixing. Contrib.Mineral. Petrol. 67:417-431.
Eggler, D.H., McCallum, M.E., and Smith, C.B., 1979: Megacryst
assemblages in kimberlite from northern Colorado and
southern Wyoming:
petrology,
geothermometry-barometry,
and areal distribution. In " The Mantle Sample: Inclu­
sions in Kimberlites and other Volcanics (Boyd, F.R. and
Meyer, H.O.A., ed.), LithoCrafts Inc., Chelsea, Michigan,
pp213-226.
Elthon, D., 1979: High magnesia liquids as the parental
for ocean floor basalts. Nature 278,5:514-518.
magma
Elthon, D. and Scarfe, C.M., 1980: High-pressure face cquili­
bria of a high-magnesia basalt: implications for the ori­
Carnagie Inst. Wash.
gin of mid-ocean ridge basalts.
Yearbook 79:277-281.
Elthon, D. and Scarfe, C.M., 1984: High-pressure phase equili­
bria of a high-magnesia basalt and the genesis of primary
oceanic basalts. Am. Mineral.69:1-15.
Finger, L. W., 1972: The uncertainty in the calculated ferric
iron content of a microprobe analysis. Carnagie Inst.
Wash. Year Book 71:600-603.
Fisk, M.R., 1978: Melting relations and mineral chemistry of
Iceland and Reykjanes Ridge basalts. University of Rhode
Island, Ph.D. thesis.
Fisk, M.R., Shilling, J.-G. and Sigurdsson, H. 1980: An exper­
imental investigation of Iceland and Reykjanes Ridge
Contrib.Min.Petrol.
tholeiites: I. Phase
relations.
74:361-374.
Frey, F.A., Bryan, W.B. and Thompson, G. 1974: Atlantic ocean
floor,
geochemistry and petrology of basalts from Leg 2
Project.
and
3
of
the
Deep
Sea
Drilling
J.Geophys.Res.:79:5507-5527.
Fuji, T. and Bougault, H., 1983: Melting relations of a magne­
sian abyssal tholeiite and the origin of MORBs. Earth
Planet.Sci.Lett.62:283-295.
Green, D.H. and Hibberson, W., 1970: The instability of pla­
gioclase in peridotite at high pressure. Lithos 3:20922 1 •
Green, D.H., Hibberson, W.O. and Jaques, A.L.,
1979:Petro­
genesis of mid-ocean ridge basalts. In: "The Earth: Its
Origin, Structure and Evolution", McElhinney, M.W.(ed.),
Academic Press, London, pp 265-299.
Green, D.H. and Ringwood, A.E., 1967: The genesis of
magmas. Contrib.
Mineral.Petrol. 15:103-190.
basaltic
Grcinvold, K., 1984: Myvatn fires 1724-1729. Chemical composi­
tion of the lava. Nordic Volcanological Institute 8401.
University of Iceland.
Grcinvold, K. and Makipaa, H., 1978: Chemical composition of
Krafla lavas 1975-1977. Nordic Volcanological Institute
78, 16. University of Iceland.
Haldi, T.A. and Wright, T.L., 1967:The fractionation of nickel
between
olivine
and
augite
as a geothermometer.
Geochim.Cosmocirn. Acta 31:877-884.
Haroard'ottir, V., 1984: The petrology of the Hengill volcan­
ic system, southern Iceland. McGill University, M.sc.
thesis (in prep.)
C.J. dfld ,lorry, ll.J.(cd.) 1'.Nd: Cor1t1ncntc1I
,ind m<1ntle xenoliths. Shivil Puhlishinq Ltd., 272 p.
l>dS<l] ts
fi.iwh:t'swurtl1,
H ci .i
'
:J • ' 1 9 8 4 : G d h b r () - cl n () r t h O s i t c n {) d \l l c s f r O fll
K r i �; u v i k
f 1 s !:, t J r c
s y s l cm
,1 n d a c t i v c z o n e o f r i f t i n q ,
I{ (' y k j d n c s p e n i n s u J c1 , S \\' •
l c c l a n cl • ti o r d 1 c V o l c ,l n o l o q i c a l J n s t i. t u t e H 4 U • Li n i v e r s i t y o 1·
l cc l cl n d.
lmsland, P., 19b4: The pctroloqy, mineraloqy and evolution of tl1e
Jan Mayen mc19ma system. V'sind<lf'elag lslcndica 43, 322 pp.
Prentsmidjan Oddi, Reykjavik.
Jrvinq,
A.J.,
1978: A review of experimental
element
trace
crystal/liquid
Ceochim.Cosmochim.Acta 42:743-770,
studies
of
partitioninq.
Jackson, E.D. and \Yriqht, T.L., 1970: Xenoliths in the
volcanic series, Hawaii. J.Petrol. 11:405-430.
Honolulu
Jacques, A.L. and Green, O.H., 1980: Anhydrous melting of perido­
tite at 0-15 kb pressure and the qenesis of tholeiitic
basalts. Contrib.Mineral.Petrol. 73:287-310.
Jakobsson, S., 1966: The Grimsnes lavas,
turalia lslandica 2:6.
Jakobsson, S.P., 1979:
Jokull 29:57-73.
Outline
of
the
SW-Iceland.
petrology
Acta
of
Na-
Iceland.
Jakobsson, S.P., 1979: Petrology of recent basalts in the eastern
volcanic zone, Iceland. Acta Naturalia Islandica 26.
Jakobsson, S.P.,
('ocean'it).
berqtequndir
1983: Islenskar
N'att'urufr�dinqurinn 52:�0-85.
I'
Pikr'it
Jakobsson, S.P., Pedersen, A.K., Rcinsbo, J.C. & Larsen,
L.H.,
1973:
Petrology of mugearite-hawaiite: Early extrusives in
the 1973 Heimaey eruption, Iceland. Lithos 6:203-14.
Jakohsson, S.P., J'onsson, J. and Shido, F., 1978: Petrolo9y of
Petrol.
the
J.
Western Reykjanes Peninsula, Iceland.
19:669-705.
1963:
J.,
J'onsson,
Hnydlingar
N'att'urufr�oingurinn 33,1:9-20.
i
islenzku
J'onsson, J., 1978: Jardfr�dikort Reykjanesskaqa. Report
tofnum OS 7HD 7831.
bergi.
Orkus-
Kristmannsd'ottir, H., 1971: Anorthosite inclusions in tertiary
dolerite from the island groups Hrappsey and Purkey, West
Iceland. J. Geol. 79,741-748.
Kudo, A.M. and Weill, D.F., 1970: An igneous plaqioclase
mometer. Contrih. Mineral.Petrol. 25:52-65.
Kushiro, I., 1973: Origin of some magmas in oceanic
oceanic regions. Tectonophysics 17:211-222.
and
thercircum-
and
Hanson ,
G•N• ,
I. ct n fl m u i r , C • H • , B e n d e r , J . 1- • , B e n c e , A • E •
19 7 7 :
f' et r o genesis of basalts from the FAMOUS are a : Mid Atlantic Ridge. Earth planet.Sci.Lett. 36:133-156.
Larsen, �J.C., 1979:
Glass-bearing gabbro inclusions in hyaloclastites from Tindfjallajbkull, Iceland.
Lithos 12,
289-302.
Maalbe, S. and Jakobsson, S.P., 1980: The PT phase relations
of a primary oceanite from the Reykjanes peninsula, Ice­
land. Lithos 13:237-346.
MacDonald, G.A. and Katsura T., 1964:
Chemical composition of
Hawaiian lavas. J.Petrol. 5:82-133.
Makipaa, H.T., 1978: Mineral equilibria, geothermometers and
geobarometers
in
some
Icelandic
hyaloclastites.
Bull.Geo!. Soc. Finland 50:113-134.
Makipaa, H., 1978a: Petrological relations in some Icelandic
Bull. Geol. Soc. Finland
hyaloclastites.
basaltic
50:81-112.
Makipaa, H.T., 1980: Partitioning of chromium, manganese,
cobalt and nickel between olivine and basaltic liquid: an
experimental study. Bull.Geol.Soc. Finland 52,2:175-191.
Mathews, W.H., porarinsson, S. and Church, N.B., 1964: Gravi­
tative settling of olivine in pillows of an Icelandic
basalt. Am. J. Sci. 262:1036-1040.
Mathez, E.A., 1973: Refinement of the Kudo-Weill plagioclase
thermometer and its application to basaltic rocks. Con­
trib. Mineral. Petrol. 41:61-72.
Melson, W.G., Thompson, G. and Van Andel, T.H., 1968: Volcan­
ism and metamorphism in the Mid-Atlantic Ridge, 22oN la­
titude. J.Geophys.Res. 73:5925-5941.
Melson, W.G. and Thompson, G., 1971: Petrology of a
fault
zone
and
adjacent
ridge
Philos.Trans.Roy.Soc.Lond, A268:423-441.
transform
segments.
Miyashiro, A., Shido, F. and Ewing, M., 1969: Diversity and
origin of abyssal tholeiite from the Mid-Atlantic Ridge
near 240 and 300 north latitude. Contrib.Mineral.Petrol.
23:38Muir, I.D. and Tilley, C.E., 1964: Basalts from the northern
part
of the rift zone of the Mid-Atlantic Ridge.
J.Petrol. 5:409-434.
Muir, I.D. and Tilley, C.E., 1966: Basalts from the northern
part of the Mid-Atlantic Ridge, II. The Atlantis collec­
tion near 30oN. J.Petrol. 7:193-201.
Nicholls, G.D., Mawalk, A.J. and Hays, E.E., 1964: The nature
and composition of rock samples dredged from the Mid­
Atlantic ridge between 22oN and 52oN. Mar.Geol. 1:333-
Niclscn,R.L. and Urake, M.J., 1979: Pyroxene-melt equilibria.
Geochim. Cosmochim. Acta 43:1259-1272.
O'Hara, M.J., 1968: The bearing of phase equilibria studies in
synthetic and natural systems on the origin and evolution
of basic and ultrabasic rocks. Earth Sci.Rev. 4:69-133.
O'Hara, M.J., 1968a: Are ocean floor
nature 220:683-686.
basalts
primc1ry
magma?
O'Hara M.J., Saunders, M.J. and Mercy, E.L.P., 1975: Garnet­
peridotite, primary magma and eclogite: interpretation of
upper mantle processes in kimberlite. Physics and Chem­
istry of the Earth, 9:571-604.
Oskarsson, N., Sigvaldason, G.E. and Steinl=>'orsson, S., 1982:
A dynamic model of rift zone petrogenesis and the region­
al petrology of Iceland. J. Petrol. 22:28-74.
Pallister, J.S. and Hopson, C.A., 1981: Samail ophiolite
suite: Field relations, phase variation, cryptic varia­
tion and layering, and a model of a spreading ridge magma
chamber. J.Geophys.Res. 86:2593-2644.
Palmasson, G. and S.emundsson, K. 1974: Iceland in relation to
the Mid-Atlantic Ridge.
In:"Annual Reviews of Earth
planet.Sci. 2:25-50". (ed. F.A. Donath)
Pedersen, A.K. and Hald, N., 1982: A cummingtonite-porphyritic
dacite with amphibole-rich xenoliths from the Tertiary
central volcano at Krofjo'rdur, NW Iceland.
Lithos
15:137-159.
Presnall,D.C., Dixon, S.A., Dixon, J.R., O'Donnell, T.H.,
Brenner, N.L., Schrock, R.L. and Dycus, D.W., 1978:
Liquidus phase
relations
on
the
join
diopside­
forsterite-anorthite from 1 atm to 20 kbar: their bearing
on the generation and crystallization of basaltic magma.
Contrib. Mineral. Petrol. 66:203-220.
Presnall, D.C., Dixon, J.R., O'Donnell, T.H. and Dixon, S.A.,
1979: Generation of mid-ocean ridge tholeiites. J.Petrol.
20:3-35.
Ragnarsd'ottir, l<.V., Bishop, F.C. and Jakobsson, S.P., 1982:
Crystallization processes in a shallow magma chamber
beneath Seyoish'olat·, Grimsnes, Iceland.
(Abs.). In
"Abstracts
Generation of major basalt types". IAVCEI­
IAGC Scientific Assembly. Reykjavik.
Rhodes, J.M., Dungan, M.A., Blanchard, D.P. and Long, P.E.,
1979: Magma mixing at mid-ocean ridges: evidence from
basalts drilled near 22oN on the Mid-Atlantic Ridge.
Tectonophys. 55:35-62.
Roeder, P.L. and Emslie, R.F., 1970:
Olivine-liquid
brium. Contrib. Mineral. Petrol. 29:275-289.
equili-
Sato, H., 1977: Nickel content of basaltic magmas: identifica­
tion of priinc1ry magmas <:1nd a measure of the degree of
o 1 i v i n e f r a c t i_ o n d t .i o n • L i t h o s 1 0 : 1 1 3 - 1 2 0 •
Seward, F.M., 1971: The distribution of transition elements in
the system CaMgSi206-Na2Si205-H20 at 1000 bars pressure.
Chem.Gcol. 7:73.95.
J-G., 1976: Spinels i_n Mid­
Sigurdsson, H. and Shilling,
Atlantic Ridge basalts: Chemistry and occurrence. Earth
Planet.Sci.Lett. 29:7-20.
Smewing, J.D., 1981: Mixing characteristics and compositional
differences in mantle-derived melts beneath spreading
axes: Evidence from cyclically layered rocks in the
ophiolite of North Oman. J.Geophys.Res. 86:2645-2659.
Sobolev, N.V., 1977: Deep-seated inclusions in kimberlites and
the problem of the composition of the upper mantle. Ed­
wards Brothers Inc., Ann Arbor, Michigan, p 279.
Sparks, R.S.J.,
Meyer, P. and Sigurdsson, H., 1979: Density
variation amongst mid-ocean ridge basalts: Implications
for magma mixing and the scaricity of primitive lavas.
Earth Planet.Sci.Lett. 46:419-430.
Stolper,
E., 1980: A phase diagram for mid-ocean ridge
basalts: preliminary results and implications for petro­
genesis. Contrib.Mineral.Petrol. 74:13-27.
Stolper, E. and Walker, D., 1980: Melt density and the average
composition of basalt. Contrib.Mineral.Petrol. 74:7-12.
Sun, C-0., Williams, R.J. and Sun,
S-S. 1974: Distribution
coefficients of Eu and Sr for plagioclase-liquid and
clinopyroxene-liquid equilibria in oceanic ridge basalt:
an experimental study. Geochim.Cosmochim.Acta 38:14151433.
Sutherland, F.L., 1974: High-pressure inclusions in tholeiitic
basalt and the range of lherzolite-bearing magmas in the
Tasmanian volcanic province.
Earth
Planet.Sci.Lett.
24:317-324.
S�mundsson, K., 1967: Vulkanismus und Tektonik des HengillGebietes in Sudwest-Island.
Acta Naturalia Islandica
2, 7.
S�mundsson, K., 1978: Fissure swarms and Central volcanoes of
the neovolcanic zones of Iceland. In "Crustal evolution
in northwestern Britain and adjacent regions. (Bowes,
D.R. and Leake, B.E., ed.). pp 415-432.
S�mundsson, K., 1979:
Jokull 29:7-28.
Outline
of
the
geology
of
Iceland.
Takahashi, E. and Kushiro, I., 1983: Melting of a dry perido­
tite at high pressures and basalt genesis. Am.Mineral.
68:859-879.
Thompson, H.N. and Til.ley, C.E., 1969:
Melting and crystalli­
zation rPlations of Kilauean basalts in Hawaii. The la­
vas of the 1959-60 Kjlauec1 eruption. Earth Planet.
Sci.
I_ e tt. 5 : 4 6 9 - 4 7 7 •
Phase realations in transitional and alkali
Thy, P., I 982:
bc1saltic: glasses from Iceland. Nordic Volcanological Insti.tute 8202.
Tilley, C.E., 1960: Differentiation of Hawaiian basalts: some
variants in lava suites of dated Kilauean eruptions.
J.Petrol. 1 :47-55.
Tryggvason, T., 1957:
The gabbro bombs at
Bull.Geel.Inst. Uppsala, 30:273-320.
lake
Gr<Pnavatn.
Yoder, H.S., 1976: Generation of basaltic magma.
Academy of Sciences, Washington D.C., 265 p.
porarinsson, S., 1953: Gr<Pnavatn and
grafisk Tidskrift 52:292-302.
National
Geo-
Gestsstaoavatn.
Wager, L.R., 1960: the major element variation of the layered
series of the Skaergaard intrusion and re-estimation of
the average composition of the hidden layered series and
successive residual mgmas. J.Petrol. 70:111-125.
Wass, S.Y., 1979:
Multiple origins of clinopyroxenes
kali basaltic rocks. Lithos 12:115-132.
in
White, R.W., 1966: Ultramafic inclusions in basaltic
from Hawaii. Contr. Min.Petrol. 12:245-314.
al-
rocks