MINERALOGICAL JOURNAL, VOL. 5, No. 1, pp. 21-43, AUG., 1966
REVIEW OF PYROXENE RELATIONS IN
TERRESTRIAL ROCKS IN THE LIGHT
OF RECENT EXPERIMENTAL WORKS
HISASHI KUNO
Geological Institute, University of Tokyo, Tokyo, Japan
ABSTRACT
Recent experimental works have shown that clinopyroxene of MgSiO3-FeSiO3 series has a stability field at high pressures and at temperatures lower than several hundred degrees. But review of the data on natural
pyroxenes suggests that pigeonite crystallizes from magmas within its own stability field at temperatures higher than that of orthopyroxene stability. There are also some experimental evidences in favor of the existence of a form of clinopyroxene stable at high temperatures. Crystallization of ortho
pyroxene and pigeonite from basalt magmas is discussed on the basis of hypothetical phase diagrams of MgSiO3-FeSiO3 system and also of (Mg, Fe) SiO3 Ca (Mg, Fe) Si2O6 system. No drastic change is needed for the existing theories on the crystallization of pyroxenes from magmas. It is pointed out that an existence of the stability field of Mg-Fe clinopyroxene at low pressures and low temperatures raises many serious questions which are difficult to answer.
Introduction
Pyroxenes are among the best-studied rock-forming minerals.
Since BOWEN and SCHAIRER'S (1935) experiment on the stability
relation between Mg-Fe clinopyroxene and orthopyroxene at atmos
pheric pressure, it has generally been believed that pigeonite in
natural rocks forms at higher temperatures than does orthopyroxene
of the same Mg: Fe+2 ratio. Almost all of the observed facts on
natural pyroxenes appeared to be in accord with BOWEN and
SCHAIRER'S experiment at least qualitatively.
Quite recently however, experimental works by SCLAR and others
41964), BOYD and ENGLAND (1965), AKIMOTO and others (1965), and
22 Review of pyroxene relations in terrestrial rocks
LINDSLEY (1965) have established that Mg-Fe clinopyroxene has ay
stability field at temperatures lower than those of orthopyroxene
stability at least under high pressures. This result was indeed a
shocking news to observational petrologists because there has been
no indication of natural pigeonite having been formed at low
temperatures.
On the other hand, FOSTER (1951) and ATLAS (1952) demonstrated
experimentally that protoenstatite is a stable form at temperatures.
higher than the stability temperatures of orthoenstatite and that
the former changes very promptly to clinoenstatite upon quenching
It might be supposed that all the natural pigeonites are products
of inversion from Mg-Fe protopyroxenes stable at magmatic tem
peratures, if such ever exist. The complete absence of Mg-Fe-
clinopyroxene in rocks which are supposed to have been formed,
well within its stability field determined experimentally is also a
problem to be answered. It is now necessary to review the observa
tions so far made on natural pyroxenes and reinterprete their-
relations in the light of the recent experimental works.
The term "pigeonite" is used in this paper for all Ca-poor
clinopyroxenes with various Mg: Fe+2 ratios occurring in terrestrial-
rocks.
Acknowledgements
The present discussion is based on many unpublished analyses,
of pyroxenes carried out by H. HARAMURA of the Geological Institute
to whom my thanks are due. The cost of the pyroxene studies,
was defrayed by the Japanese Government Fund for Scientific
Research which is greatly appreciated. I am very much obliged to
J. E. THOMPSON of the Bureau of Mineral Resources, Canberra for
a gift of clinoenstatite-bearing andesite and single crystals of proto
enstatite from Papua and also to D. H. GREEN of the Australian
National University for a manuscript of their paper on this andesite . I. KUSHIRO and A. MIYASHIRO kindly offered constructive criticism
H. KUNO 23
on the result of this study. K. YAGI supplied a sample of 1945 dacite
of Usu. Volcano from which hypersthene phenocrysts were sepa
rated for analysis.
Summary of recent experiments
The stability relations of MgSiO3 polymorphs are shown in Fig.
1. The discrepancy between the position of the ortho- and clino-
pyroxene stability boundary given by BOYD and ENGLAND (1965)
and that by SCLAR and others (1964) bears no significance in the
present discussion. The inversion temperature between these pyroxenes is little affected by pressure, whereas that between proto- and
Fig. 1. Stability relations of MgSiO3 polymorphs after BOYD and ENGLAND (1965). The stability boundary between orth- and clinoenstatite given by Sciar and others (1964) is also shown. Continental and oceanic geotherms calculated by CLARK and RINGWOOD (1964) and also by MACDONALD
(1964) are shown by dotted lines.
24 Review of pyroxene relations in terrestrial rocks
orthoenstatite is strongly affected by pressure.
FOSTER (1951) and ATLAS (1952) found that protoenstatite (proba
bly orthorhombic according to Atlas) is stable at temperatures
higher than the stability temperatures of orthoenstatite and that
the former promptly changes to clinoenstatite, but not to ortho
enstatite, upon quenching. This was later confirmed by many
experiments. This clinoenstatite is most probably the one indicated
as "clinoenstatite " in Fig. 1.
It must be remembered that no one has observed inversion from
orthoenstatite to clinoenstatite by cooling the former at atmospheric
pressure or in hydrothermal range, as mentioned by BOYD and ENGLAND (1965). Thus this inversion appears to be extremely
Fig. 2. Stability relations of FeSiO3 polymorphs given by LINDSLEY (1965). The stability boundary between fayalite-quartz assemblage and FeSiO3 pyroxenes and that between ortho- and clinoferrosilite as given by AKIMOTO and others (1965) are also shown.
H. KUNO 25
sluggish.
BOYD and SCHAIRER (1964) found an existence of another form
of Mg-rich pyroxene stable above 1385•Ž at atmospheric pressure.
According to their description, this pyroxene may be monoclinic
with a cell size different from that of the clinoenstatite formed by
inversion of protoenstatite. Thus the field of "protoenstatite"
shown in Fig. 1 may be divided into two portions, the higher-
temperature portion representing the stability field of this form of
clinopyroxene and the lower-temperature portion that of protoensta
tite.
The stability relations of FeSiO3 polymorphs are shown in Fig.
2. The discrepancy between the positions of the ortho- and clinofer
rosilite stability boundary given by different authors is great, but
we are not concerned with this problem here, because FeSiO3 pyrox
enes are not stable under the pressures within the earth's crust and
are replaced by fayalite-quartz assemblage, and also because all the
mantle rocks contain pyroxenes poor in Fe.
The field of proto?-ferrosilite in Fig. 2 was originally divided
into two portions: the portion of" ferrosilite III " in the lower
pressure side and that of "clinoferrosilite" in the higher-pressure
side (LINDSLEY et al., 1964). LATER, LINDSLEY (1965) interpreted the
"ferrosilite III" and "clinoferrosilite" as formed by inversion from
a different form which he designated as " proto?-ferrosilite". Thus,
Fig. 2 represents LINDSLEY'S latest diagram. His interpretation
appears to be based partly on his confirmation of the existence of
the clinoferrosilite stability field at low temperatures, a finding also
reported independently by AKIMOTO and others (1965), and partly
on the analogy with the phase relations in MgSiO3 pyroxenes.
It is possible that "ferrosilite III" of LINDSLEY and others (1964)
represents FeSiO3 analogue of the clinopyroxene synthesized by
BOYD and SCHAIRER (1964) above 1385•Ž.
LINDSLEY (1965) and AKIMOTO and others (1966) confirmed that
clinopyroxenes of intermediate compositions also have their stability
26 Review of pyroxene relations in terrestrial rocks
field below several hundred degrees at high pressures. Thus it
appears that the existence of a stability field for all members of
Mg-Fe clinopyroxenes at temperatures below those of orthopyroxene
stability has been established for high pressures. However, it has
not yet been confirmed that the same stability field does extend to
low pressures.
Pyroxene relations in rocks
Fig. 3 shows schematically the kinds and composition ranges of
Ca-poor pyroxenes in different groups of rocks.
Pigeonite is confined to rocks derived from high-temperature
basalt magmas (tholeiite, high-alumina basalt, and rarely alkali olivine
basalt) and their differentiation products. Rocks of the hypersthenic
rock series and gabbro-diorite-granite series of orogenic belts, which
are supposed to have been formed from magmas with lower crystal
lization temperatures, do not contain pigeonite. These facts strongly
Fig. 3. Kinds of Ca-poor pyroxenes and their composition ranges
in terms of molecular per cent of FeSiO3 in different groups of
rocks. CPx and OPx are abbreviations for pigeonites and ortho-
pyroxenes respectively.
H. KUNO 27
indicate that pigeonite can form only through high-temperature
crystallization, and orthopyroxene can form even at temperatures
of metamorphism.
Fig. 4 shows composition ranges of analysed Ca-poor pyroxenes
from different groups of rocks. The most Fe-rich pyroxene is eulyte
(Fs 88) from Wang-chang-tzu, Manchuria (Kuno, 1954). The absence
of any pyroxenes more Fe-rich than this is in accord with the
experiments (LINDSLEY et al., 1964; AKIMOTO et al., 1965).
Fig. 4. Chemical compositions of analysed Ca-poor pyroxenes. Point A represents eulyte phenocrysts in dacite from Asio, Japan whose com
position was inferred from r index and unit cell size (KUNG, 1954). The mineral is associated with pigeonite (see Fig. 5), fayalite, and almandine phenocrysts.
Pigeonites have distinctly higher CaO contents than any ortho-
pyroxenes. Orthopyroxenes from igneous rocks have wider limit
of CaO solubility than those from metamorphic rocks. KUNO (1964)
showed that orthopyroxenes from basaltic and andesitic rocks have
generally higher CaO than those from dacitic rocks. These facts
28 Review of pyroxene relations in terrestrial rocks
Fig. 5. Compositions of associated orthopyroxenes (squares), pigeonites (circles), and augites (circles). Solid marks are for compositions of analysed pyroxenes and open marks for compositions inferred optically by using diagrams of HESS (1949) and KUNO (1954). Ca contents of the orthopyroxenes were estimated simply by analogy with analyses of orthopyroxenes from basaltic rocks. K-phenocrysts in basalt from Kuro-hana-yama, Hunakata Volcano (AOKI, 1960). TU-microphenocrysts in
andesite from Tengu-zawa and Usugoya-zawa, Hakone Volcano (KUNO .& NAGASHIMA, 1952). HA-phenocrysts in andesite from Hakone-toge, Hakone Volcano (Kuiao, 1950a; see also Table 1). U-phenocrysts in andesite from Weiselberg, Germany (KuNo, 1947). HI-phenocrysts in andesite from Higashi-yama Volcano, Hatizyo-zima Island (ISSHIKI, 1963). 0-phenocrysts in andesite from Okubo-yama, aMinami-Aizu (KUNO & INOUE, 1949). B-iron-rich dolerite from Beaver Bay, Minnesota (MUIR, 1954). A-phenocrysts in dacite from Asia (KUNG, unpublished data).
imply that orthopyroxene can take up more CaO at higher temper
atures but the limit of CaO solubility is narrower than that in
pigeonite. The last-mentioned relation is more clearly shown by
plotting the compositions of co-existing orthopyroxenes, pigeonites, and augites in Di-En-Fs-Hd quadrilateral (Fig. 5). Except for those
from iron-rich dolerite from Beaver Bay, all the pyroxenes plotted
in Fig. 5 are from basalt, andesite, and dacite. Pyroxenes from two
andesites from Hakone Volcano were chemically analysed and are
plotted in the figure by solid circles and solid squares. The analyses
of the associated hypersthene, pigeonite, and augite from one of
these andesites are given in columns 3a, 3b, 3c, and 3d of Table 1.Fig. 5 shows that the pigeonites are invariably more Ca-rich
H. KUNO 29
Table 1. Analyses of orthopyroxene phenocrysts from Mauna Loa,
and Syowa-Sinzan and of associated pyroxene phenocrysts in pigeonite
andesite from Hakone.
1. Bronzite phenocrysts in olivine-bronzite basalt (HK61100502). A cindercone on the southwestern rift of Mauna Loa. Analyst, H. HARAMURA (KUNO, 1964).
2. Hypersthene phenocrysts (YAGI, 52092003) of hypersthene dacite. A dacite dome protruded in 1945, Usu Volcano, Hokkaido. Analyst, H.
HARAMURA.3a. Hypersthene phenocrysts in augite-pigeonite-hypersthene andesite (HK
33022001) from Hakone-toge, Hakone Volcano. Analysts, K. TADA and M. HUZIMOTO (KUNO, 1954).
3b. Mixture of hypersthene, pigeonite, and possibly augite phenocrysts of the same rock as 3a. The amount of hypersthene is a little larger than
that of clinopyroxenes in the analysed material. Analyst, H. HARAMURA.3c. Pigeonite phenocrysts, possibly with a little admixture of augite pheno
crysts, of the same rock as 3a. The large amount of Na20 is partly due to contamination of groundmass feldspar. Analyst, T. SAMESUIMA
(KUNO, 1952).3d. Augite phenocrysts of the same rock as 3a. Analyst, H. HARAMURA.
30 Review of pyroxene relations in terrestrial rocks
than the associated orthopyroxenes, and except for the pyroxenes
from Beaver Bay and Tengu-zawa and Usugoya-zawa, they are
higher in Fe: Mg ratio than the associated orthopyroxenes.
Crystallization sequence of orthopyroxene and pigeonite in high-
temperature basalt magmas (tholeiite, high-alumina basalt, and alkali
-olivine basalt) has been carefully studied in a number of differentiated
gabbro and dolerite intrusions and of volcanic rock series.
In these intrusions, orthopyroxene separates from the magma
in the earlier stage, and when it reaches the composition of about
Fs30, its place is taken by pigeonite which continues crystallizing until
the later stage (WAGER & DEER, 1939; HESS, 1941; POLDERVAART,
1946; WALKER & POLDERVAART, 1949; BROWN, 1957; KUSHIRO, 1964).
This relation was interpreted by HESS (1941) as due to successive
increase of Fe: Mg ratio in the . pyroxene during crystallization by
which its composition crosses the orthopyroxene-clinopyroxene
inversion interval of BOWEN and SCHAIRER (1935). A similar relation
-is also found for phenocrysts of volcanic rocks (KUNO, 1950a; UCHI-MIZU
, 1966), although the composition at which orthopyroxene changes to pigeonite differs greatly in different rock suites probably because
of variability of crystallization temperature of magmas.In rapidly cooled rocks such as volcanic rocks and some dolerites,
pigeonite once crystallized from magma is preserved as such even .at room temperature, whereas in slowly cooled rocks such as gabbros
and some dolerites, the mineral undergoes inversion to produce
orthopyroxene with exsolution lamellae of augite parallel to (001) of
the original monoclinic pyroxene (HESS, 1941; POLDERVAART and
HESS, 1951). Such pyroxene is called "orthopyroxene of the Stillwater
type" (HESS, 1960). The exsolution of augite is obviously due to
the narrower limit of CaO solubility in orthopyroxene than in
pigeonite (Fig. 4). In rocks of intermediate rate of cooling such as some dolerites, the exsolved augite forms irregular blebs . Such
pyroxene is called here "orthopyroxene of the Kintoki-san type" for the reason described below.
H. KUNO 31
Kintoki-san, one of the parasitic cones of Hakone Volcano, is made up of lavas and scoria of basaltic andesite intruded by radial
dikes of the same rock type (KUNG, 1950b). These rocks contain microphenocrysts of pigeonite having a composition apparently
similar to that of Usugoya-zawa pigeonite (Fig. 5). The radial dikes converge at the center of the crater of Kintoki-san where a thick dike of coarse-grained andesite occurs. This andesite contains
microphenocrysts of orthopyroxene with extremely fine-grained, irregular blebs of monoclinic pyroxene, probably augite. This ortho-
pyroxene has been formed probably through inversion of the pigeonite which is preserved as such in the finer-grained rocks of the same volcano.
These inversion behaviors are shown schematically in Fig. 6.
Orthopyroxene crystallizing from the high-temperature basalt magmas contain 3.0 to 1.5 wt. % CaO (Fig. 4 and column 1, Table 1). Upon rapid cooling, it forms homogeneous crystals such as found as
phenocrysts in volcanic rocks, whereas upon slow cooling, it exsolves augite lamellae parallel to (100) ("orthopyroxene of the Bushveld
Fig. 6. Cooling behavior of Ca-poor pyroxenes crystallizing from
high-temperature basalt magmas.
32 Review of pyroxene relations in terrestrial rocks
Fig. 7. Cooling behavior of orthopyroxenes crystallizing from low-temperature magmas (hypersthenic rock series and gabbro-diorite-granite series of orogenic belts) and of orthopyroxenes in metamorphic and mantle rocks.
type" as named by HESS and PHILLIPS, 1938) (Fig. 6). This is
obviously due to the narrower limit of CaO solubility in orthopyrox
ene at lower temperatures.
Fig. 7 further shows schematically how the exsolution lamellae
of augite develop or do not develop in orthopyroxene depending on
original CaO content in the orthopyroxene and on rate of cooling
(see also HESS, 1952).No examples have been reported for orthopyroxene having
inverted to clinopyroxene during cooling.
Suggested phase relations of Ca-poor pyroxenes
crystallizing from magmas
From the experiments mentioned above, one might suggest
that the inversion of orthopyroxenes to clinopyroxenes upon heating,
as described by BOWEN and SCHAIRER (1935), was really an inversion
from ortho- to protopyroxenes, the latter having been changed upon
quenching to clinopyroxenes stable at low temperatures. If this
H. KUNO 33
suggestion be accepted, then all natural pigeonites may be interpreted
as inversion products of protopyroxenes crystallizing from magma.
However, the existence of protopyroxenes of intermediate composi
tions has not been confirmed yet. Moreover, there are some facts
which strongly indicate that pigeonite is distinct from protopyroxene,
as will be described below.
1) THOMPSON and DALLWITZ recently found an andesite from
Papua in which they identified clinoenstatite and bronzite phenocrysts
(see also TILLEY et al., 1964). They interpreted the clinoenstatite
as having been inverted from protoenstatite (DALLWITZ et al., 1966).
This is indeed the first finding of protoenstatite in natural rocks.
GREEN determined the compositions of the co-existing pyroxenes by
microprobe X-ray analyser and found that the clinoenstatite (original-
ly protoenstatite) is more Mg-rich and Ca-poor than the bronzite
(DALLWITZ et al., 1966). The lower Ca content in the supposed proto
enstatite is in good agreement with the relation between synthetic
protoenstatite and orthoenstatite solid solutions as shown in Fig. 9
after BOYD and SCHAIRER (1964). On the contrary, natural pigeonites
have higher Ca and Fe contents than the associated orthopyroxenes
(Fig. 5). Following a similar reasoning, BOYD and SCHAIRER (1964)
also maintain that protopyroxene and pigeonite are different forms.
2) The clinoenstatite from Papua shows polysynthetic twinning
characteristic of clinoenstatite produced by quenching synthetic
protoenstatite. Pigeonite phenocrysts in andesites do not show such
polysynthetic twinning. Only two or three twinning lamellae on
(100) are usually seen in a single crystal of porphyritic pigeonite up
to 1 mm in length. Untwinned crystals are also common. Thus
there is no indication of inversion from another mineral.
3) The protoenstatite now pseudomorphed by clinoenstatite
from Papua is distinctly orthorhombic. Its crystal habit is different
from that of ordinary orthopyroxene. Natural pigeonite appears to
be monoclinic, although (001) face has never been identified in phe
nocrysts because of their irregular outline. The exsolution lamellae
34 Review of pyroxene relations in terrestrial rocks
of augite arranged in the herring bone pattern in the inverted
pigeonite or "orthopyroxene of the Stillwater type" (see Fig. 3 of
HESS, 1941) is a good indication of its originally monoclinic symmetry.
Thus it is almost certain that natural pigeonite is a distinct
form having a stability field at temperatures of basaltic and andesitic
magmas. This form is probably equivalent to the monoclinic form
of MgSiO3-rich pyroxenes synthesized by BOYD and SCHAIRER (1964)
above 1385•Ž, and also to the " proto? -ferrosilite" of LINDSLEY
(1965). This form of Mg-Fe pyroxenes is called here "high clino-
pyroxene" and is distinguished from "low clinopyroxene" found
experimentally at low temperatures and high pressures.
The clinopyroxenes obtained by BOWEN and SCHAIRER (1935) by
heating natural orthopyroxenes are probably high clinopyroxenes.
If the inversion curve (actually a loop) between this form and
orthopyroxene given by BowEN and SCHAIRER is extended, it reaches
the FeSiO3 end of the series a little above 900•Ž. The boundary
line between the" proto? -ferrosilite" and orthoferrosilite fields of
Fig. 2 intersects the zero pressure line also a little above 900•Ž.
A suggested phase relation for the MgSiO3-FeSiO3 series at
atmospheric pressure is shown in Fig. 8.
BOYD and SCHAIRER (1964) suggested two alternative phase
diagrams for Mg-Fe pyroxenes close to MgSiO3 (see Fig. 11 of their
paper). But they gave some reasons for which they preferred the
one shown in the MgSiO3-rich portion of Fig. 8. This type of
diagram also explains the higher Mg: Fe ratio in the protoenstatite
than in the associated bronzite from Papua, as is discussed by
DALLWITZ and others (1966).
The position of the Mg-rich portion of the inversion loop A A'
B' B is fixed by the compositions of orthopyroxene phenocrysts in
recent lavas whose temperatures at the time of extrusion are appro-
ximately known. The data are given in Table 2.
The temperature of the lava containing the Mauna Loa pyroxene
is not directly determined. The temperature of 1190•Ž is assumed
H. KUNO 35
Table 2. Compositions of orthopyroxene phenocrysts and tempera
tures of the lavas in which they occur.
simply by analogy with that of 1959 lava of Kilauea Iki (RICHTER
-and EATON, 1960) and also with the melting temperature of Mauna
Loa olivine basalt determined by TILLEY and others (1964). As the
temperature of lava measured at the surface would be more or less
lower than that at which the phenocrysts crystallized, the curve
A' B' is drawn a little above the points for the Mauna Loa and 0-sima
-orthopyroxenes.
In Fig. 8, a trend line representing composition-temperature
relations of pyroxenes crystallizing at successive stages of Skaergaard
magma based on the description by WAGER and DEER (1939), BROWN
(1957), and WAGER (1960), is given together with that for the magma
of the hypersthenic rock series of Japan. As the rocks of Paricutin
and Usu belong to the latter series, the trend line for this series is
drawn close to them. The pyroxene trend for the pigeonitic rock
series would lie only a little below the Skaergaard trend (KUNG, 1966).
A detailed discussion of mafic mineral evolution along these trend
lines is given elsewhere (KUNO, 1966).
Fig. 9 shows a series of phase diagrams for sections across the
quadrilateral Di-En-Fs-Hd at various MgSiO3: FeSiO3 (wt.) ratios.
These diagrams are drawn so as to explain the Papua and Skaergaar
d pyroxene relations.
36 Review of pyroxene relations in terrestrial rocks
Fig. 8. A hypothetical diagram showing phase relations in the MgSiO3-FeSiO3 series at low pressures. The high-temperature, Mg-rich portion of the diagram is based on BOYD and SCHAIRER'S (1964) hypothetical diagram, and the high-temperature, Fe-rich portion is, based on BOWEN and SCHAIRER'S (1935) experiment. The temperatures for points E and F were inferred by extrapolation of the orthopyroxene and (low) clinopyroxene stability boundary lines given by BOY D and ENGLAND (1965) and by LINDSLEY (1965) respectively (see Figs. 1 and 2). The dotted lines represent composition-temperature relations of pyroxenes crystallizing at successive stages of the magma of Skaergaard and of the hypersthenic rock series. Circles represent compositions and crystallization temperatures of orthopyroxene phenocrysts in basalt from Mauna Loa (M) , in basalt from O-sima Volcano (0, 1950 lava), in basaltic andesite from Paricutin Volcano (P, 1944 lava), and in dacite from Syowa-
Sinzan (U, 1945 lava). See Table 2.
H. KUNO 37
Fig. 9. Phase diagrams for sections across the pyroxene quadri-
lateral Di-En-Fs-Hd at various MgSiO3: FeSiO3 ratios. HCSS-high clino-
pyroxene solid solution. Fo-forsterite. Disc-diopside solid solution includ-
ing Fe-bearing varieties. Prss-protoenstatite solid solution. OEnss-
aorthoenstatite solid solution. LCSS-low clinopyroxene solid solution.
OPSS-orthopyroxene solid solution. 01SS-olivine solid solution. Tr
tridymite.
38 Review of pyroxene relations in terrestrial rocks
The diagram for MgSiO3 100% is merely a reproduction of BOYD,
and SCHAIRER'S diagram. The diagram for MgSiO3 90% is constructed
on the assumption that the crystallization temperature of the liquid
is so much lowered, possibly owing to the presence of much salic
components and volatiles, that protoenstatite and orthopyroxene co-
exist with liquid A, DALLWITZ and others (1966) inferred that the
crystallization temperature of the Papua pyroxenes was between
1250 and 1150•Ž. The protoenstatite undergoes inversion at about
700•Ž to form low clinopyroxene after the liquid extrudes to the
surface, but orthopyroxene fails to do so.
In the succeeding diagrams, the low clinopyroxene field is omitted
. The diagram for MgSiO3 80% represents the earliest stage or
the hidden zone of Skaergaard. This stage corresponds to that
portion of the Skaergaard trend line lying within the orthopyroxene
field of Fig. 8. Liquid A2 separates orthopyroxene and diopside.
Olivine, although not shown in the diagram, is being resorbed by
liquid A2.
The diagram for MgSiO3 70% represents the lower zone of
Skaergaard or the stage corresponding to that portion of the trend
line lying inside the inversion loop A A' B' B of Fig. 8. Orthopyrox
ene, high clinopyroxene (=pigeonite), diopside, and olivine (not shown
in the diagram) are in equilibrium with liquid A3. Orthopyroxene
is being replaced by high clinopyroxene across the inversion interval,
while olivine is being resorbed by liquid A3. The high clinopyroxene
later undergoes inversion to form orthopyroxene of the Stillwater
type after solidification of the liquid. In this and succeeding dia
grams, high clinopyroxene and diopside are assumed to be separate
phases. Clinopyroxenes of intermediate compositions (subcalcic
augite) can only form as a metastable phase upon rapid cooling
(YODER et al., 1963).
The diagram for MgSiO3 50% represents the middle zone of
H. KUNO 39
Skaergaard or the stage corresponding to that portion of the trendd
line lying within the high clinopyroxene field of Fig. 8. Olivine is
no more present and liquid A, separates only high clinopyroxene
(=pigeonite) and diopside. The high clinopyroxene later changes
to orthopyroxene of the Stillwater type.
The diagram for MgSiO3 10% represents the stage of the portion
of the trend line lying within the high clinopyroxene+olivine+
tridymite field of Fig. 8. In the actual Skaergaard sequence, however,
this stage corresponds to the beginning of the upper zone where
the pyroxene composition is more magnesian than that shown in
Fig. 8. Olivine reappears in the assemblage. Other phases crystal-
lizing from liquid A, are high clinopyroxene (=pigeonite later
inverted to orthopyroxene), diopside, and tridymite.
The whole crystallization sequence of the pigeonitic rock series
can also be explained by the diagrams for MgSiO3 80%, 70%, 50%,
and 10%, although actual pyroxene compositions are slightly different
from those shown in these diagrams.
The whole crystallization sequence of the hypersthenic rock
series is dominated by the phase relation of the diagram for MgSiO3
80%, although the crystallization temperatures are much lower
. Absence of low clinopyroxenes in the
crustal and upper mantle rocks
In Fig. 1, oceanic and continental geotherms are drawn by dotted
lines. The uppermost 25km of the oceanic mantle and uppermost
5 to 20km of the continental mantle lie within the low clinoenstatite
stability field. However, orthoenstatite and bronzite are common
constituents of peridotite and eclogite nodules which are supposed
to have been brought up from the upper mantle, but Mg-Fe clino
pyroxene has never been observed. As these orthopyroxenes contain exsolution lamellae of clinopyroxene, probably diopside, they do not
40 Review of pyroxene relations in terrestrial rocks
appear to have been inverted from low clinopyroxenes when they
were captured by basalt magmas.
A possible explanation is that the upper mantle, even at the
contact with the oceanic Moho, was once heated to the temperatures
of the orthopyroxene stability field and then cooled slowly. Ortho-
pyroxene exsolved diopside lamellae but failed to change to low
clinopyroxene because of the extreme sluggishness of the inversion
(see p. 24). The presence of small amounts of A12O3 and CaO in the mantle orthopyroxenes might widen their stability field toward the
low temperature side.
The non-occurrence of Mg-Fe clinopyroxenes in hornfels, granu
lite, and amphibolite is also a problem. As orthopyroxenes are
common in these rocks, it is difficult to understand why low clino
pyroxenes do not form in the lower-temperature parts of these
metamorphic rocks.
It may be suggested that the stability field of low clinopyroxene
found at high pressures does not extend to the low pressure range
but instead its place is taken by that of an yet unrecognized form
of orthopyroxene. It is important to find out some X-ray or optical
means to distinguish this orthopyroxene, if ever exist, from high-
temperature orthopyroxene, as well as some means to distinguish
high and low clinopyroxenes.
Conclusion
In spite of the finding of the stability field of Mg-Fe clinopyrox
enes at low temperatures and high pressures, it appears that no
drastic change is needed for the existing theories on the crystalliza
tion of pyroxenes from rock magmas if we assume another form of
clinopyroxene stable at magmatic temperatures (called here high
clinopyroxene). It is important to investigate experimentally the
real existence of low pressure extension of the low clinopyroxene
H. KUNO 41
stability field, and also to study X-ray or optical difference between
high and low clinopyroxenes and also between orthopyroxenes
crystallizing from magmas and those in metamorphic rocks . Does the difference in 2 V variation in volcanic orthopyroxenes and meta
morphic and plutonic orthopyroxenes as found by HESS (1952) and
KUNO (1954) correspond to that between two forms of orthopyroxene?
More detailed X-ray works should be done to solve this problem .
REFERENCES
AKIMOTO, S., KATSURA, T., SYONO, Y., FUJISAWA, H. & KOMADA, E. (1965).
Polymorphic transition of pyroxenes, FeSiO3 and CoSiO3, at high pres
sures and temperatures: J. Geoph. Res., 70, 5269-5278.
AKIMOTO, S., KOMADA, E. & KUSHIRO, I. (1966). Preliminary experiments on
the stability field of natural pigeonite and enstatite : Proc . Jap. Acad.,
42, No. 5, 482-487.
AOKI, K. (1960). Early stage basalts from the Nasu volcanic zone: J. Jap .
Ass. Petro, Min. Econ. Geol., 45, No. 2, 54-65 (in Japanese).
ATLAS, L. (1952). The polymorphism of MgSiO3 and solid-state equilibria
in the system MgSiO3-CaMgSi2O6: J. Geol., 60, No. 2, 125-147.
BOWEN, N. L. & SCHAIRER, J. F. (1935). The system MgO-FeO-SiO2: Amer.
J. Sci., 29, 151-217.
BOYD, F. R. & ENGLAND, J. L. (1965). The rhombic enstatite-clinoenstatite
inversion: Carnegie Inst. Wash. Year Book, 64, 117-120.
BOYD, F. R. & SCHAIRER, J. F. (1964). The system MgSiO3-CaMgSi2O6: J.
Petro., 5, No. 2, 275-309.
BROWN, G. M. (1957). Pyroxenes from the early and middle stages of
fractionation of the Skaergaard intrusion, East Greenland : Min. Mag.,
31, 511-543.
CLARK, S. P. & RINGWOOD, A. E. (1964). Density distribution and constitution
of the mantle: Rev. Geophys., 2, 35-88.
DALLWITZ, W. B., GREEN, D. H. & THOMPSON, J. E. (1966). Clino-enstatite
in a volcanic rock from the Cape Vogel area, Papua: J. Petro., 7, Octo
ber issue (in press).
FOSTER, W. R. (1951). High-temperature X-ray diffraction study of the
polymorphism of MgSiO3: J. Amer. Ceram. Soc., 34, 255-259.
HESS, H. H. (1941). Pyroxenes of common mafic magmas. Part 1 and Part 2:
Amer. Min., 26, 515-535; 573-594.
•\•\ (1949). Chemical composition and optical properties of common clino-
42 Review of pyroxene relations in terrestrial rocks
pyroxenes, Part 1: Amer. Min., 34, Nos. 9, 10, 621-666.
HESS, H. H. (1952). Orthopyroxenes of the Bushveld type, ion substitutions
and changes in unit cell dimensions: Amer. J. Sci., Bowen Volume,
173-187.
•\•\ (1960). Stillwater igneous complex, Montana. A quantitative miner
alogical study: Mem. Geol. Soc. Amer., 80, 1-230.
•\•\ & PHILLIPS, A. H. (1938). Orthopyroxenes of the Bushveld type: Amer.
Min., 23, No. 7, 450-456.
ISSHIKI, N. (1963). Petrology of Hachijo-jima Volcano Group, Seven Izu.
Islands, Japan: J. Fac. Sci., Univ. Tokyo, Sec. 2, 15, Pt. 1, 91-134.
KRAUSKOPF, K. B. (1948). Lava movement at Paricutin Volcano, Mexico
Bull. Geol. Soc. Amer., 59, 1267-1283.
KUNG, H. (1947). Occurrence of porphyritic pigeonite in "weiselbergite"
from Weiselberg, Germany: Proc. Jap. Acad., 23, No. 9, 111-113.
•\•\ (1950a). Petrology of Hakone Volcano and the adjacent areas, Japan
Bull. Geol. Soc. Amer., 61, 957-1020.
•\•\ (1950b). Geology of Hakone Volcano and adjacent areas. Part I: J
. Fac. Sci., Univ. Tokyo, Sec. 2, 7, Pt. 4, 257-279.
•\•\ (1952). Explanatory Text of the Geological Map of Japan "Atami"
Geol. Surv. Japan, 1-141 (in Japanese).
•\•\ (1954). Study of orthopyroxenes from volcanic rocks: Amer. Min., 39,
30-46.
•\•\ (1964). Aluminian augite and bronzite in alkali olivine basalt from.
Taka-sima, North Kyusyu, Japan: "Advancing Frontiers in Geology and
Geophysics " dedicated to Dr. KRISHNAN (India), 205-220.
•\•\ (1966). Differentiation of basalt magmas: "Basaltic rocks" edited by
H. H. HESS (in press).
KUNG, H. & INOUE, T. (1949). On porphyritic pigeonite in andesite from
Okubo-yama, Minami-Aizu, Hukusima Prefecture: Proc. Jap . Acad., 25,
No. 4, 128-132.
KUNG, H. & NAGASHIMA, K. (1952). Chemical compositions of hypersthene
and pigeonite in equilibrium in magma: Amer. Min., 37, 1000-1006.
KUSHIRO, I. (1964). Petrology of the Atumi dolerite, Japan: J. Fac. Sci.,
Univ. Tokyo, Sec. 2, 14, Pt. 2, 135-202.
LINDSLEY, D. H. (1965). Ferrosilite: Carnegie Inst. Wash . Year Book, 64, 1
48-150.
•\•\ , MACGREGOR, I. D. & DAV IS, B. T. C. (1964). Synthesis and stability of
ferrosilite: Carnegie Inst. Wash. Year Book, 63, 174-176.
MACDONALD, G. J. F. (1964). Dependence of the surface heat flow on the
radioactivity of the earth : J. Geoph. Res., 69, No. 14, 2933-2946.
MINAKAMI, T. (1962a). Catalogue of Active Volcanoes of the World Inclu- di
ng Solfatara Fields. Part XI, Japan, Taiwan, and Marianas, compiled
H. KONO 43
by Kuno, 214.
MINAKAMI, T. (1962b). Catalogue of Active Volcanoes of the World Including
Solfatara Fields. Part XI, Japan, Taiwan, and Marianas, compiled by
Kuno, 297.
MUIR, I. D. (1954). Crystallization of pyroxenes in an iron-rich diabase from .
Minnesota: Min. Mag., 30, 376-388.
OBA, Y. (1960). On the phenocrystic pigeonite in a comma lava of Usu.
volcano, Hokkaido: J. Jap. Ass. Petro. Min. Econ. Geol., 46, No. 5, 172-
177 (in Japanese).
POLDERVAART, A. (1946). The petrology of the Mount Arthur dolerite
complex, East Griqualand: Trans. Roy. Soc. South Africa, 31, Pt. 2, 83-
110.•\•\
& HESS, H. H. (1951). Pyroxenes in the crystallization of basaltic
magma: J. Geol., 59, No. 5, 472-489.
RICHTER, D. H. & EATON, J. P. (1960). The 1959-60 eruption of Kilauea.
volcano : New Scientist, 7, 994-997.
SCLAR, C. B., GARRISON, L. C. & SCHWARTZ, C. M. (1964). High-pressure sta
bility field of clinoenstatite and the orthoenstatite-clinoenstatite tran
sition (abstract): Tr., Amer. Geoph. Union 45, 121.
TILLEY, C. E., YODER, H.S., Jr. & SCHAIRER, J. F. (1964). New relations on
melting of basalts: Carnegie Inst. Wash. Year Book, 63, 92-101.
UCHIMIZU, M. (1966). Geology and petrology of alkali rocks from Dogo, Oki
Islands: J. Fac. Sci., Univ. Tokyo, Sec. 2, 16, Pt. 1, 85-159.
WAGER, L. R. (1960). The major element variation of the layered series of
the Skaergaard intrusion and a re-estimation of the average composition
of the hidden layered series and of the successive residual magmas: J.
Petro., 1, 364-398.
WAGER, L. R. & DEER, W. A. (1939). Geological investigations in East
Greenland : Part III, The petrology of the Skaergaard intrusion,.
Kangerdlugssuaq, East Greenland: Med. om Gronland, 105, 1-352.
WALKER, F. & POLDERVAART, A. (1949). Karroo dolerites of the Union of
South Africa: Bull. Geol. Soc. Amer., 60, 591-706.
WILCOX, R. E. (1954). Petrology of Paricutin Volcano, Mexico: Bull. Geol.
Surv., 965-C, 281-353.
YODER, H. S., Jr., TILLEY, C. E. & SCHAIRER, J. F. (1963). Pyroxene quadri
lateral: Carn. Inst. Washington Year Book, 62, 84-95.
Manuscript received 27 June, 1966.