Sunteți pe pagina 1din 13

Gondwana Research, V 5, No. 2, p p . 261 -273.

02002 International Association for Gondwana Research, Japan


ISSN: 1342-93ix

Granite: A Planetary Point of View


B. Bonin, J. B6bien and P. Masson
C N R S - U M R 861 6 ’Orsayterre’, Dipartement des Sciences de la Terre, Universite‘ de Paris-Sud, F-91405 Orsay Cedex, France

(Manuscript received October 3,2001; accepted October 16,2001)

Abstract
The telluric planets and the asteroid belt display the same internal structure with a metallic inner core and a silicate
outer shell. Experimental data and petrological evidence in silicate systems show that granite can be produced through
all types of magmatic processes.
On Moon, 4 4 3 . 9 Ga granite clasts displaying dry mineral assemblages correspond to discrete intrusive events.
Large K/Ca enrichment and low REE abundances in granite relative to KREEP are consistent with silicate liquid
immiscibility, a process observed in melt inclusions within olivlne of Lunar basalts. On Mars, black-and-white rhythmic
layers observed on the Tharsis rise along the flanks of Valles Marinens and the peripheral scarps of the Tharsis Montes
giant volcanoes suggest the eruption of possibly felsic pyroclastites Samples analysed In the Mars Pathfinder landing
site yield basic to intermediate (45 to 62 wt.% SiO,) compositions. Though no granites were found so far in the Martian
SNC meteorites, a component close to terrestrial continental (granitic) crust is inferred from trace element and isotope
systematics. Like the Martian northern hemisphere, Venus has suffered extensive volcanic resurfacing, whereas folded
and faulted areas resemble terrestrial continents. The hypothesis of a granitic component is again ‘tantalizing’. Finally,
extra-terrestrial granite is also found as enclaves within iron meteorites and ordinary chondrites.
Granite, a major component of the Earth’s crust, can be generated in all geodynamic settings. The low density of
granite favours continental accretion. Thus, the occurrence or absence of granite and of associated silicic volcanism
within the other telluric planets is not a trivial question. Granite is generally thought to be produced through ‘wet’
processes Lunai evidence shows that dry conditions may apply as well. In Venus, a large planet with high rates of
magma production, it is speculated that significant volumes of granite can develop.
Key words: Granites, geochemistry, extra terrestrial, Lunar, Martian.

Introduction order of abundance), that are issued from erosion and


weathering of continental (granitic) formations.
On Earth, granitic rocks were extensivelystudied during Orthogneisses, metapelites and metagreywackes are the
the last centuries for the following reasons: (i) they are chief components of the metamorphic group of rocks.
the most abundant rocks in the Earth’s upper continental Various estimates of the intrusive igneous rocks range
crust and, simply for volumetric reasons, should not be around 86-88 vol.% granites and 14-12 vol.% other rock
ignored, (ii) like other igneous rocks, they represent probes types, such as gabbros, diorites, etc. From the average
into the deep planetary interiors, i.e., the crust and even composition of the three groups of rocks (sedinieptary,
the upper mantle and (iii) they are closely connected with metamorphic, igneous), it can be concluded that roughly
tectonics and geodynamics. For reviews of the conflicting 86 vol.% of the upper continental crust is granitic in
ideas until the present consensus on the magmatic origin composition. Though occasionally present, granite occurs
of granite, see e.g., Eskola (1932), Clarke (1992), Pitcher in relatively insignificantamounts within the oceanic crust
(1993), Bonin et al. (1997). and the upper mantle. The total mass of Earth’s granitic
With an average thickness of 40 km, the Earth’s rocks amounts to at least 10lyt, which corresponds,
continental crust has a total mass estimated at 2.4 * 10lYt. assuming an average density of 2.67, to a volume of
The upper continental crust forms 43% of this mass, and approximately 3.74 * lo9 km3.
consists of sedimentary, metamorphic, and igneous rocks Whether granite is likely to occur or not, outside the
in the rough proportion of 1:4:4 (Wedepohl, 1991). Earth, is therefore, an inviting topic that will be addressed.
Sedimentary rocks comprise 78 vol.% detrital rocks The fundamentals of the paper are based on the following
(shales, greywackes and quartz sandstones in the decreasing two lines of evidence: (i) the telluric planets, including
262 B. BONIN ET AL.

the Earth and the asteroid belt, display the same internal unadulterated by other materials. The vast majority of lunar
structure with a metallic inner core and a silicate outer ‘granites’ are non-pristine, polymict products of impact-
shell, (ii) experimental petrology provides ample evidence induced mixing of lunar plus minor meteoritic materials.
that granite melts at relatively shallow depths and at Granite clasts are relatively common at the Apollo 14
temperatures of the order of 600-900°C, which are the site. Based on the abundance of felsite in Apollo 14 soils
lowest recorded so far in silicate systems (only and regolith breccias, granites are estimated to constitute
natrocarbonatite, which is not a silicate rock, yields lower 0.5-2 vol.% of the crust here (Shervais and Taylor, 1983).
melting temperatures). As granite is eniplaced in fairly The largest pristine lunar granite (labelled 14321,1027)
large quantities within the earth’s crust, there is no reason was found as a 16 x 7 mm outcrop on breccia 14321 (total
why this could not happen within the silicate outer shell mass 9 kg); the original mass of the clast is estimated to
of one or the other telluric planet. The available evidence be 1.8 g. A smaller clast (14303,204) was found as an
for observed and/or inferred occurrences of felsic granite- 11 x 6 mm outcrop on breccia 14303, with original mass
like materials outside the earth will be critically reviewed of only 0.17 g (Warren et al., 1983). In the IUGS
and discussed, with later emphasis on the critical problem classification (Le Maitre, 1989), 14321, 1027 is alkali
of continents within the telluric planets. feldspar granite, with 40% quartz and -60% K-feldspar,
while 14303,204is monzogranite, with 23% quartz, 32%
The Lunar Evidence K-feldspar and 33% plagioclase. Where not brecciated or
shock - melted (30-50 vol.%), the undamaged igneous
The Moon has its origin and composition closely texture is dominantly granophyric, feldspar and quartz
connected to the Earth (O’Neill, 1991a, 1991b). The
occurring as 0.05 mm up to 2 mm intergrown crystals. As
isotopic record provides evidence for extreme chemical
the rocks are devoid of volatile-rich phases, volatiles
differentiation within 200 my of its formation, so that the
cannot be a prerequisite for formation of the graphic
mantle appears to be much more differentiated on the
intergrowths.
Moon than on the Earth (Carlson, 1994). The Moon’s outer
The felsic mineralogy is quite uncommon: alkali
shells originated from the differentiation of the lunar
feldspar lies in the Or80-Or97range, while plagioclase is
magma ocean. The major products of this Moon-wide
calcic. The alkali feldspar granite contains no albite, only
melting event include the anorthositic lunar crust, thought
Ba-poor K-feldspar with a composition close to the pure
to have resulted from the flotation of cumulus plagioclase,
end-member, bracketed between Or,, and Or,l. In the
and the ultramafic lunar mantle, thought to have formed
monzogranite, barian alkali feldspar (Or,,,, 1.54.8 wt.%
by sinking of dense mafic silicate minerals such,as olivine
and pyroxene. The last magmatic dregs of the lunar magma BaO) and normally zoned An,,.,, plagioclase constitute
ocean constituted ‘KREEP’, an acronym that reflects its the main feldspar assemblage. Moreover, scarce ternary
characteristic high concentrations of potassium, rare earth feldspar intergrown with quartz has average compositions
elements, and phosphorus (Carlson, 1994). KREEP is a (Or45Ab25An30, -1.5 wt.% BaO) plotting in the ‘forbidden
globally distributed chemical component that permeates zone’ of the feldspar triangle. The association of almost
a wide variety of lunar rocks, but which is never found in pure K-feldspar with highly calcic plagioclase is unknown
its pristine, unadulterated state. within terrestrial monzogranites.
Anhydrous mafic and accessory minerals include
Lunar granite samples olivine, clinopyroxene (pigeonite, augite, hedenbergite),
As granite is abundant in the Earth’s crust, it was ilmenite, Fe-Ni metal, occasionally zircon, Y-bearing
tempting, before the Surveyor program, to speculate that fluorapatite, whitlockite, and troilite. The alkali feldspar
granitic rocks predominate in the lunar crust. But lunar granite yields extremely Fe-rich olivine and clinopyroxene,
sample recovery has failed to provide large, pristine compared with the monzogranite, i.e., Fo, (fayalite) vs.
granites. Actually,lunar ‘granites’occur either as components F O ~and
~ , hedenbergite vs. zoned En,,,, augite, respectively
in polymict breccias, or as small clasts, or as felsites. As they are poorer in Ni (from less than 0.2 to at most
‘Felsite’ is a descriptive term for pale brown K-Si-rich 2.8 wt.%) than meteoritic metals ( > 4 wt.%), Fe-Ni grains
glasses that are found in many regolith breccias and soil are considered by Warren et al. (1983) to belong to the
samples. Felsites contain a few quench phases and rare Moon’s materials, suggesting crystallization at f 0 , close
relict mafic phases, and almost certainly represent total to the IW(iron-wiistite)buffer. Shervais and McGee (1999)
shock melts of a granite precursor, not endogenous rhyolite demonstrate that variations in mineral assemblages and
magmatism (Shervais and McGee, 1999). In the lunar compositions ( e g , BaO in alkali feldspar, Mg/Fe ratios
vocabulary, a pristine rock is compositionally endogenous in mafic phases) imply the involvement of at least four
(igneous or thermally metamorphosed igneous), distinct parent magmas.

Gondwana Research, -!? 5, No. 2,2002


GRANITE: A PLANETARY POINT OF VIEW 263

The clasts analysed are interpreted as probably free terrestrial mantle-normalizedarachnograms (Fig. lb). Both
of foreign contamination, according to the following LILE and HFSE yield fairly high abundances, with up to
criteria: (i) the glassy, shock-melted areas have 800 times the primitive terrestrial mantle for Ba and Th
compositions appropriate for mixtures of the crystals and 20-80 times for Yb and Lu. LILE are highly fractionated,
from the remainder of the clasts, (ii) the siderophiie while HFSE, including REE, are not. Pronounced Sr and
elements, Ir and Au, are well below the concentrations Ti negative anomalies are probably related to apatite +
recommended (Warren and Wasson, 1977) as a cut-off ilmenite fractionation. These features may be considered
for distinguishing pristine lunar rocks from nonpristine as diagnostic of extensive crystal-liquid fractionation. By
ones. Thus, chemical compositions of lunar granites may comparison, the terrestrial crust shows constantly
be examined with the same eye as terrestrial granites fractionated patterns, with no Eu negative anomaly (Fig.
(Figs. 1 and 2). 2a), and lower LILE and HFSE abundances (Fig. 2b).
Chrondrite-normalized flat REE patterns (Fig. l a ) are Despite their evolved nature, depicted by high silica
unfractionated, characterized by fairly high abundances contents, low Mg/Fe ratios, etc., lunar granites are low in

Lunar granites / chondrites (REE) 1000 3 Lunar granites I primitive mantle (Sun & McDonough)

loo0 7
500

100 o 14321,1027
o 14321,1027 14303,204
14303,204 12013
100
n 12013 rn 73255c
rn 7325%
50 10 1

10
RbBaTh K T a N b L a C e S r Nd P Hf ZrSmEu Ti Tb Y YbLu

Fig. 1. Compositional diagrams for lunar granites (data extracted from table 1, Warren et al., 1983). Samples 14321, 1027, 14303, 1204 and
73255c are clasts recovered from breccias, while sample 12013 corresponds to the average of the light portions, probably not pure felsite, of
the 12013 breccia. (a) Chondrite-normalized ME patterns. (b) Primitive terrestrial mantle (Sun and McDonough, 1989) normalized patterns.
Though not analysed, P contents are inferred to be low on the basis of modes, apatite being extremely rare.

Earth'scrust and lunar granites/chondriies (REE) Eanh'scrust and lunar oraniles I Drirnnive mantle (Sun & McDonoushl
1000
1000

Lunarclasl
~i
Lumrclast
100 A Felsite men in breccia 100 U Lunar clasl
n Lunarclast
Lunar clad
Felsite men In breccia
0 Lower CNSI Lunar clad
10 o Upper crust 10
0 Lower c ~ s t
0 Upper crust

(a)
1 T i i - - r - ~ - i - ~I 7 r
La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Trn yb Lu
i I 1- i
I ' I I I I I I I I I 1 r 7 T-TY i-rl--T
RbBaTh K TaNbLaCeSrNd P HI DSrnEuTi Tb Y Y b L U

Fig. 2. Chemical compositions of lunar granites vs. terrestrial crust (lower and upper crust, from Wedepohl, 1991). (a) Chondrite-normalized REE
patterns. (b) Primitive terrestrial mantle normalized patterns.

(40-200 times the chondrites), (La,/Yb,) ratios near 1.0, any REE, but the heaviest Yb and Lu, with respect to
or even lower, and pronounced Eu negative anomalies KREEP, and yield extremely high WREE ratios. These
(Eu/Eu* ranging from 0.5 to 0.2), with the exception of features are better explained by silicate liquid immiscibility
73255c clast. Similar features are portrayed by primitive (Watson, 1976).

Gondwana Research, I? 5, No. 2,2002


264 B. BONIN ET AL.

Melt inclusions, further evidence f o r silicate liquid Isotopic record of lunar granites
immiscibility
The timing and extent of lunar magmatism was
The lunar samples display crystals that grew from a determined primarily by the internal thermal evolution
silicate magma, and just as in most terrestrial igneous of the Moon.Two major events marked the Moon’s
rocks, irregularities in the growth process resulted in the geological history: (i) the lunar magma ocean, resulting
trapping of small portions of the surrounding fluid silicate in the formation of the first anorthositic lunar crust -4.45
liquid. Unlike terrestrial rocks, almost all the inclusions Ga ago and (ii) the lunar cataclysm -3.9 Ga ago, due to
in the lunar samples are primary and can readily provide a late, heavy bombardment producing most of the
information on the magmatic environment, i.e., the observable lunar basins. Then, more volcanism took place,
temperature and pressure of trapping, and the composition yielding discrete ages of flooding ranging from 3.9-3.6
of the liquid. Ga for the ‘middle-aged’high-Ti basalts to 3.4-3.2 Ga for
A systematic survey of inclusions within crystals of all the ‘young’low and very low-Ti basalts. The isotopic record
lunar samples from Apollo missions was undertaken. The (for compilation of data, see Nyquist and Shih, 1992)
results are reported in the numerous papers stemming -
substantiates at most a 1.2 Gy-long magmatic history.
from the continuing series of Lunar Science Conferences, The lunar granites have been dated by isotopic methods
since the first paper dedicated to Apollo 11 samples including K-Ar or 40Ar-39Ar,Rb-Sr, Sm-Nd, U-Pb and
(Roedder and Weiblen, 1970). In addition, the unmanned Pb-Pb systems.Ages range from 4.373 0.03 Ga to 3.87rt 0.01
Luna 24 spacecraft returned with drillcore materials Ga, indicating a 500 My-long granitic emplacement
recovered from the regolith (Roedder and Weiblen, 1978). history. Four samples have been analysed simultaneously
Melt inclusions were systematically observed within by different methods; three samples yield concordant ages,
crystals of olivine, plagioclase, pyroxene, spinel, ilmenite, but one sample, 14321, 1062, actually a subsample of
cristobalite, and as interstitial inclusions at grain the 14321, 1027 granitic clast, gives a -230 My time lag
boundaries. The major result is that the parent magma between the dates obtained by Sm-Nd and 40Ar-39Ar
evolved during cooling systematicallyto late-stage residual methods.
liquids that split ultimately into immiscible high-iron and The 4oK-40Casystem is considered as more resistant to
high-silica liquids. impact reheating than the Rb-Sr and K-Ar systems, and
The high-silica melt inclusions (average 75 wt.% SO,) three lunar granites have been analysed by this method
are chemically identical to the granite clasts, and (Shih et al., 1993). The sample 14321,1062 yields a
correspond to low-Mg (<0.5 wt.% MgO) and Na (<0.9 K-Ca isochron age of 4.0610.07 Ga, in agreement with
wt.% Na,O), high-Fe (average 2.8 wt.% FeO, up to the 4.09f0.03 Ga (Rb-Sr) and 4.1110.20 Ga (Sm-Nd)
5.8 wt.%) and K (6.8-9.7 wt.% K,O) potassic granite. ages obtained previously. The concordancy of K-Ca,
The coexisting, immiscible high-iron melt inclusions Rb-Sr and Sm-Nd isochron ages strongly suggests that
(average 30 wt.% FeO, range 2311.7%) correspond to the granite crystallized from a magma -4.1 Ga ago and
basic-ultrabasic (<43 wt.% SiO,) ferro-pyroxenite, rich underwent the brecciation event 3.88k0.03 Ga ago
in Ti (up to 8.7 wt.% TiO,), and fairly rich in Ca (up to recorded by the 40Ar-39Arsystem. The sample 14303,206,
12 wt.% CaO, where not trapped by plagioclase). a subsample of the 14303, 204 granitic clast, has an
The close assocation of high-silica and high-iron glasses average K-Ca model age of 4.290.12 Ga, in agreement
within one single inclusion resemble the immiscible high- with the more precise U-Pb model age of 4.314&0.008
silica and high-iron liquids produced exprimentally by low Ga determined from zircon crystals. The third sample
pressure (10.1MPa) fractionation of a basaltic magma 12013, 141, corresponding to the light portion (‘felsite’)
under anhydrous conditions near the iron-wiistite buffer of the 12013 breccia (Warren et al., 1983), has an
(Dixon and Rutherford, 1979; Dixon and Rutherford, imprecise K-Ca model age of 3.91 Ga, in agreement with
1983). The liquid line of descent indicates that 88 to the more precise Rb-Sr isochron ages of 3.99t0.05 Ga
+
95% of an [olivine -t- plagioclase pyroxene f ilmenite] and 4.0lk0.09 Ga determined on other felsite fragments
mineral assemblage should crystallize before the onset of of the same 12013 breccia.
silicate liquid immiscibility. There are still enough fluids, Large K/Ca enrichment during genesis of the granitic
making feasible the physical separation of the two sample 14321, 1062, -16 times K/Ca of the source, is
immiscible liquid phases. In the lunar samples, two required by the initial 40Ca-44Caratio of 47.14E 0.010.
additional liquids, metallic Fe and iron sulphide in The same WCa enrichment is observed in experimental
composition, were apparently present, making four study of silicate-liquid immiscibility using lunar basalts
simultaneously immiscible liquids (Koedder and Weiblen, as starting material. The lunar granites crystallized from
1972). parental magmas of a variety of ages. From a detailed

Gondwana Research, V. 5, No. 2,2002


GRANITE: A PLANETARY POINT OF VIEW 265

analysis of the more retentive ages (Shih et al., 1993), considered to have been derived from Mars (McSween,
eight age peaks are recognizable from the age distribution 1994). Though no granitic samples have been discovered
curve and probably represent eight discrete magmatic so far, the data collected from natural samples and remote-
events. Variable ages also imply that most of the lunar sensing imagery suggest that intermediate and silicic
granites did not crystallize directly from the primordial magmatism may have taken place on Mars.
magma ocean, but were produced as plutonic rocks in
Mars Pathfinder in-situ samples
the lunar crust after the solidificationof the magma ocean.
Few natural samples have been analysed in situ on
lmplications f o r granite generation and volumes
Mars. The first source of information is the incomplete
The lunar granites have been sampled within the XRF analyses of basaltic materials by the two Viking
regolith on the Moon's surface. The regolith is composed Landers (Clark and Baird, 1979). The second source comes
of small to tiny particles, that either form loose soils or from the more accurate APXS analyses performed on board
are set within polymict breccias, and consists mostly of the rover of the Mars Pathfinder mission in 1997 (Rieder
the lunar crust pulverized during a protracted meteoritic et al., 1997; McSween et al., 1999). The site is covered
impact history. As such, it is representative of the lunar with pebbles, cobbles and boulders that comprise 16%of
crust at the vicinity of the impacts. It is probably no the surface area, the remainder consisting of fines, as
coincidence that the largest pristine lunar granite clasts windblown aeolian bedforms and as deposits strewn with
are from Apollo 14. Remote sensing data show that on small pebbles, characteristics consistent with deposition
the Moon's surface, K, Th and U are concentrated in the by catastrophic floods, possibly overprinted by ejecta from
longitude range O-5O0W, which corresponds to the a nearby impact crater. However, the angularity of many
Western Highland Province. After a first crude estimation blocks examined suggests that rocks at the Mars Pathfinder
of 4% granitic glass in Apollo 14 soils (Warren et al., site are locally derived. Textures on rock surfaces may
1983), later modeling results reveal that they cannot indicate volcanic, sedimentary, or impact-generated rocks,
contain more than -1% granite (Shervais and McGee, but aeolian abrasion and dust coatings prevent
1999). It is likely that the rest of the Moon's surface, lower unambiguous interpretation.
in K, Th and U, contains far lower quantities of granitic Multispectral imaging has resolved four classes of
materials. objects variously coated by fine ferric aeolian dust and
The lunar evidence tells that granite can be generated other minerals, probably maghmemite or ferrihydrite.
through differentiation of basaltic parental magmas via Nearly linear chemical trends in APXS compositions are
crystal fractionation and silicate-liquidimmiscibilityunder interpreted as mixing lines between rock and adhering
anydrous conditions at low pressures (less than 0.1 MPa)
and low oxygen fugacities (near the iron-wustite buffer).
Compared with the Earth's protracted -4.5 Gy-long
magmatic history, the volumes of granite produced during
the Moon's shorter, i.e., -1.2 Gy-long magmatic history
are insignificant.

The Martian Evidence


Like the Moon, Mars underwent early differentiation
through crystallization of an early magma ocean (Carlson,
1994). The spacecrafts that flew over Mars (Smith et al.,
1999; Bandfield et al., 2000), or landed on it (Clark and
Baird, 1979; Rieder e t al., 1997) provided ample
information on the Martian landscapes governed by
meteoritic cratering in a soft medium (Thomas and
Masson, 1986), valleys suggesting the presence of a fluid
capable of flowage (Aguirre-Puente et al., 1994), and
abundant, gigantic volcanic activity (Greeley and Spudis, Fig. 3. AFM diagram comparing tholeiitic a n d calc-alkaline
1981; Wilson and Head, 1994). The current knowledge differentiation trends. The S-free rock, Pathfinder analysed rocks,
and SNC liquids plot within the tholeiitic field. The anorogenic
of Martian magmatic rocks is largely based on the SNC icelandite suites are from Iceland and the Galapagos Spreading
(shergottite-nakhlite-chassignite) meteorites, widely Centre (from McSween et al., 1999).

Gondwana Research, V. 5, No. 2, 2002


266 B. BONIN ET AL.

dust, a conclusion supported by correlation between Anhydrous partial melting experiments on a dry
sulphur abundances. Extrapolations of regression lines Martian mantle model show that the dominant solidus
down to zero sulphur give the composition of a assemblage in the source regions for most of the Martian
presumably igneous rock plotting near the andesite-dacite mantle is a spinel lherzolite (Bertka and Holloway, 1994a).
boundary (SiO, = 62.Qk2.7wt.Yo). The inferred andesite Primary liquid compositions resemble terrestrial basaltic
composition is akin to the terrestrial icelandites formed komatiite, except they are markedly Fe-rich and Al-poor,
by fractionation of anorogenic tholeiitic basaltic magmas, resulting in low plagioclase abundances and its late
except for its lower TiO, abundance (Fig. 3). Terrestrial appearance along the crystallisation path. Parent magmas
tholeiitic suites derive from fractionation of picritic to of SNC meteorites were likely derived from a different
basaltic parent magmas and include andesitic to rhyolitic source region that had been previously depleted in an
residual liquids. Andesite occurrence was unexpected; it aluminous phase (Bertka and Holloway, 1994b).
opens a window on the probability of more silicic The nature of the crust is a critical, yet unresolved
formations exposed on Mars. problem. The older crust is supposedly compositionally
similar to basaltic shergottite, except for higher K and P,
SNC meteorites, magmatic water and a possible granitic and lower Ca abundances (McSween, 1994). The most
component recent estimation of the composition of the global Martian
The SNC meteorites and the related ALH84001 crust (Dreibus, 2000) involves combination of basaltic
meteorite are thought to be igneous Martian rocks, based shergottite and Mars Pathfinder data, as a mixture of S-
on their crystallisation ages and a close match between free andesite (-54%), basaltic shergottite (-29%), plus
the composition of gases implanted in them during shock additional MgSO, (-9%), haematite (-7%), and ilmenite
and the atmosphere of Mars (McSween, 1994). The (-1%).
petrological and geochemical study of such Martian rocks The abundance and nature of water constitute a key
give additional clues to the origin and evolution of Mars. issue in the Martian geology. Using analysed H,O contents
Lithologies include: basalt, lherzolite, harzburgite of kaersutite, occurring as inclusions within all SNC types,
(shergottites), clinopyroxenite, wehrlite (nakhlites), water abundances in the parent magmas are calculated
dunite (chassignites) and orthopyroxenite (ALH84001). in the 100-1000 ppm range, implying a rather dry (1-35
No andesite resembling the Mars Pathfinder rocks has been ppm H,O) mantle (Mysen et al., 1998). However, trace
recognized so far within the Martian meteorites. element distributions within shergottite pyroxene as well
Isotopic data for shergottites define model ages of about as crystallisation paths, compared with hydrous and
4.5 Ga, interpreted as the time of planetary differentiation. anhydrous experimental results, imply that pre-eruptive
They suggest that Martian global differentiation, including magmas at 1120°C and 55 MPa, i.e., a depth of 5 km,
core formation, was essentially contemporaneous with the could have contained up to 1.8wt.% water and that most
completion of accretion. Interestingly, associated thermal water was ultimately lost during magma ascent to the
models necessitate a radioactive element-enriched surface (McSween et al., 2001). The presence of such a
component. Crystallisatiqn ages of shergottites are a quantity of water is inconsistent with the dry mantle of
controversial issue, the currently available dates ranging the conventional Mars model. Several hundreds ppm of
from 1.3-0.9 G a (generally assumed to represent water, comparable to the terrestrial mantle, seem a more
crystallisation ages) down to 150 Ma (interpreted either likely value. The water could have been added in the
as time of shock metamorphism, or as crystallisation ages). magma during assimilation of Martian crustal materials
Based on trace element and radiogenic isotope or by interaction with groundwater systems.
systematics, Shergotty and Zagami shergottites are
Martian surface compositions
unrelated with Antarctic shergottites, implying separate
reservoirs of differing compositions early in the history Mars is distinctive in that it displays a well marked
of Mars and the participation of a highly evolved residual planetary dichotomy separating the ancient, heavily
liquid resulting from early differentiation of the parent cratered and elevated plateaus of the southern hemisphere
body (Shih et al., 1982). This contaminant may resemble from the younger plains of the northern hemisphere. The
the terrestrial continental (granitic) crust (McSween, planetary dichotomy and the concentration of younger
1994). Nakhlite and chassignite also yield isotopic ages volcanism into a few large provinces raise the problem of
of 1.3 Ga. The age of ALH84001 is poorly defined, but it plate tectonics (Sleep, 1994) vs. mantle plume (s) (Harder
seems plausible that it may represent an older (4.5 Ga ?) and Christensen, 1996) processes in the global evolution
crust. Shock metamorphism ages are definitively younger, of the planet, a relevant issue to the probabilityof occurrence
from 150 Ma down to <15 Ma. on Mars of silicic magmatism, including granite.

Gondwana Research, V. 5, No. 2,2002


GRANITE: A PLANETARY POINT OF VIEW 267
~

Instruments are installed on board the Mars Global differing from what is currently observed on Earth: lava
Surveyor spacecraft to determine the topography, gravity, flows with a typical length of 200-350 km, giant volcanoes
and surface compositions of Mars. Global models of of the order of 500-700 km in diameter, probably more
topography from the Mars Orbiter Laser Altimeter (MOM) common plinian fall deposits and pyroclastic flows (Wilson
evidence two distinct crustal zones that do not correlate and Head, 1994). Based on Viking orbiters imagery
with the planetary dichotomy (Zuber et al., 2000). A first compared with unaided satellite studies on Earth, Francis
zone, that thins progressively from south (60 km)to north and Wood (1982) conclude provisionally that large
(40 km), encompasses much of the southern highlands explosive (and by implication, silicic) structures are rare
and the Tharsis province, whereas a second zone of to nonexistent. This conclusion was later challenged by
approximatelyuniform crustal thickness (40 km) includes evidence of air fall deposits (Mouginis-Market al., 1982;
the northern lowlands and Arabia Terra. The elastic Crown and Greeley, 1993) and pyroclastic flows
lithosphere yields thicknesses (only 20 km in the southern (Mouginis-Mark et al., 1988). Even ignimbrites are
hemisphere) that increase with time of loading in the postulated (Scott and Tanaka, 1982).
northern plains (100 km) and Tharsis ( > l o 0 km), The conditions prevailing on Mars have two interesting
consistent with volcanic resurfacing at more recent times consequences: (i) the production of large-scale plinian
than previously assumed (Hartmann et al., 1999). deposits may not signal the presence of silicic
Thermal Emission Spectrometer (TES) data from the compositions, but rather be linked to enhanced
Mars Global Surveyor (MGS) are used to determine fragmentation of magmas of any compositions, from
compositions and distributions of the Martian surface basaltic to silicic, (ii) the lack of steep-sided domes formed
formations (Bandfield et al., 2000). Low-albedo areas, by viscous, more silicic magma cannot be cited as
making up -50 % of the surface, are selected to minimize conclusive evidence for the absence of this composition
the contributions due to heterogeneous, fine-grained (Wilson and Head, 1994). In addition, the widespread
bright dust and two spectral signatures are identified. By occurrence of late, presumably basaltic to andesitic lava
comparison with spectra of terrestrial rocks, deconvolution flows (Greely and Spudis, 1981; Cipa, 1994) capping the
results indicate that the two compositions are volcanic in giant volcanoes cannot rule out the eruption of earlier,
origin: (i) basaltic, dominated by plagioclase and more silicic formations, as observed within large terrestrial
clinopyroxene and (ii) andesitic, dominated by plagioclase volcanoes, such as Cantal (Nehlig et al., 2001).
and volcanic glass. Sheet silicates, such as clays and micas, On Mars, high-albedosurfaces, coated by heterogeneous,
are identified constantly at levels of around 15%, which fine-grained bright dust, cannot be used as evidence for
represent an upper limit for the abundance of weathering silicic compositions. More interesting is the occurrence of
products, as weathering occurs primarily on the exterior ‘black-and-white’layered formations on slopes of canyons
surfaces. The dark volcanic materials vary significantly and gullies. Such banded formations are exposed on the
across the planet, the distribution of the two compositions flanks of Olympus Mons, the largest volcano of the solar
splitting roughly along the planetary dichotomy. The system (Tanaka, 1985; Cipa, 1994). On the basal scarp,
basaltic composition is confined to the older surfaces, materials underlying the late lava flows had been
whereas the andesitic composition is concentrated in the uncovered by large landslides, and display a spur-and-
younger northern plains. gullied pattern. Subhorizontal bright-dark bands are
It is no suprise that Mars Pathfinder rocks, exposed at clearly identified. Simple pile up of lava flows cannot
the mouth of Are Vallis in the northern hemisphere, are plausibly explain their regular banding as well as their
andesite, not basalt. By contrast, plagioclase abundance low mechanical resistance evidenced by intense gullying
(up to 65%) calculated in basalt of the southern and gravitational instabilities. Calculated shear stresses
hemisphere is inconsistent with the scarcity of plagioclase in the 5-10 MPa range are consistent with low cohesive
observed in basaltic shergottite, where it crystallised late strength and low internal friction materials, with a high
and is now converted into shock-induced diaplectic initial porosity filled with ground ice or other easily
maskelynite glass (McSween, 1994). This discrepancy decayed minerals, such as hydrous calcium sulphates.
supports the hypothesis of contrasting source regions in Such materials may correspond to thick, more or less
the mantle, suggested by experimental data (Bertka and consolidated pyroclastic deposits of silicic composition
Holloway, 1994b). (hence, their bright look) with intercalated dark lava flows
(Cipa et al., 1995).
Possible bimodal volcanism
The Mars Orbiter Camera (MOC) imagery reveals, in
The combination of lower gravity and lower portions of the Valles Marineris canyon system, ubiquitous
atmospheric pressure would result in volcanic features horizontal layering to depths of at least 8 km in the

Gondwana Research, V. 5, No.2,2002


268 B. BONIN ET AL

canyons, with 5-50 m-thick bright-dark bands clearly in three iron meteorites and clasts in two ordinary
visible in the spurs. The banding consists of alternate dark chondrites (Shih et al., 1993).
cliffs or steep slopes and brighter, shallower intervening The Colomera iron meteorite is highly interesting in
slopes. Dark strata are likely to be made up of volcanic that large silicate enclaves containing alkali feldspar
flows resembling terrestrial flood basalts, whereas the phenocrysts up to several centimetres in length are
brighter, shallower slopes may be interpreted as resulting reported at its surface. The absence of large enclaves in
only from aeolian dust coatings (McEwen et al., 1999). the core of the meteorite and the liquid droplet form of
An alternative explanation, which does not exclude the the internal silicate enclaves indicate that Colomera
former interpretation, is offered by bimodal magmatism, formed rather as a molten iron segregation of a small
with intermediate (andesite-dacite) to rhyolitic units mass contained in a silicate matrix. The drop-like silicate
intercalated within predominating basaltic flows. A large enclaves are considered as globules of molten silicate
number of terrestrial continental flood basalt provinces, which were trapped into the freezing Fe-Ni liquid. Albite,
such as Karoo, ParanA-Etendeka,and Deccan (Hawkesworth sanidine, anorthite (?), diopside, whitlockite, and clear
et al., 1999), are characterised by such bimodal magmatism. and pink glasses yield a Rb-Sr isochron age of 4.5E0.04 Ga
To sum up, granite (or rhyolite) has not been observed (recalculated with the decay constant of 1.42 * a-l).
so far in the Martian materials. There is, however, The Colomera data are consistent with a simple
circumstantial direct and indirect evidence that materials evolutionary history during the formation of the solar
yielding more evolved, silicic compositions may exist on system. An upper limit of 4 2 7 my can be set for the time
Mars, either at the surface level as bright pyroclastic layers, interval between dispersion of the silicate in the metal
or at depths as a radioactive crustal component. These phase and final Sr isotopic equilibration (Sanz et al.,
hints offer, therefore, a tantalizing glimpse at the ‘Martian 1970). The Weekeroo Station iron meteorite, though
crust’ (McSween, 1994), but more data should be collected lacking large alkali feldspar phenocrysts, is geochemically
to substantiate the hypothesis. similar (Burnett and Wasserburg, 1967a).
The Kodaikanal iron meteorite is characterised by
Elsewhere? abundant upto 2 cm-large silicate enclaves containing
diopside, orthopyroxene, two species of alkali feldspar,
The terrestrial planets, from Mercury to the asteroid and alkali feldspar glass. A well-defined Rb-Sr isochron
belt, issue an attractive challenge to granite inventory yields a recalculated age of 3.720.10 Ga, with an initial
within the solar system. Meteorites provide ample and ratio of 0.713k0.020. Kodaikanal is interpreted as a
useful information on the nature and evolution of ‘secondary’ body produced by ill-defined differentiation
extraterrestrial matter, especially in the asteroid belt. process about 0.78 Gy after the formation of the solar
Comparatively, Mercury and Venus are less well known. system (Burnett and Wasserburg, 1967b).
The Adzhi-Bogdo meteorite is a breccia, belonging to
Granitic enclaves within meteorites the LL group, i.e., very low in iron, of the ordinary
Meteorites are broadly subdivided into undifferentiated chondrite class. The breccia contains various types of
and differentiated families (Bischoff, 2001). The clasts, including impact glasses, some of which are K-rich,
undifferentiated family comprises four distinct classes of and alkali feldspar granite fragments. The alkali feldspar
chondrite defining 12 groups, and may include three granite consists chiefly of K-feldspar plus either quartz,
classes of primitive achondrite. The differentiated family or tridymite, and minor albite, C1-apatite, whitlockite,
is made up of five classes of metal-poor achondrite, two zircon, ilmenite, Ca-poor Fe-rich pyroxene, and a Na,
classes of stony-iron meteorite, and seven main classes of Ti-bearing silicate (Bischoff et al., 1993). Bulk rock
iron meteorite. The Martian SNC and the Lunar meteorites compositions (Bischoff, 1993) are Si, K-rich (72-79 wt.%
form two supplementary classes that are generally SiO,, 5-10 wt.% K,O) and Na-poor (I2 wt.% Na,O). So
included within the achondrites. During the 60s and 70s, far K-rich clasts are known in 14 LL chondritic breccias,
pioneering isotope studies have been performed on but only Vishnupur and Adzhi-Bogdo yield alkali fledspar
meteorites to cast light on the time sequences in the granite and associated pyroxenite clasts (Jackel and
formation of planetary bodies and their evolution. Owing Bischoff, 1997). ltyo types of K-rich clasts are identified:
to the specificity of Rb-Sr and K-Ar isotope systematics, (i) the ‘coarse-grained’(-100 pm grain size) granitic clasts
alkali-bearing material was searched systematically in (up to 700 pm in length) may result from slow cooling of
meteorites, even the iron ones, to obtain reliable ages. As the chondritic parent magma differentiating ultimately
a result, extraterrestrial ‘granites’were known well before into alkali feldspar granite residual liquid and pyroxenite
their discovery on the Moon. They make up small enclaves cumulates, (ii) the fine-grained (-1 pm grain size) clasts

Gondwana Research, V; 5, No. 2,2002


GRANITE: A PLANETARY POINT OF VIEW 269

(several mm in length), composed of euhedral olivine and m to km, as pancake domes. Terrestrial analogues of
skeletal pyroxene set in a K-rich groundmass, crystallised basaltic composition might be, e.g., the subglacial
from liquids formed by impact melting of various mixtures volcanoes of Iceland and the flat-topped volcanic cones
of already crystallised granite and LL chrondritic parent- in the Pacific and Atlantic Oceans (Clague et al., 2000),
body. which were emplaced under a hydrostatic pressure of the
Granite droplets and enclaves in stony meteorites may same order of magnitude as the Venusian atmospheric
constitute a case for either silicate-metal liquid pressure. Alternatively, pancake domes might be made
immiscibility, or silicate liquid differentiation, or both. up from multiple pulses (Fink et al., 1993) of silicic
Granite fragments in ordinary chondrites are more likely ( ~ 6 0 - 6 5wt.% SO,) magma, with viscosities ranging
issued from the differentation of a silicate magma. between 1014 and 1017Pa s (McKenzie et al., 1992). A
Whatever their ultimate origin and despite their scarcity statistical study of shapes and size distributions (Smith,
and small sizes, the granitic material observed in 1996) shows that the pancake domes on Venus differ
contrasting types of meteorites shed a glaring light on significantly from the terrestrial flat-topped submarine
the feasibility of highly evolved liquids of alkali feldspar volcanoes by their height-to-diameter ratios and flatness
granite composition on the asteroids. values, so that one or more of the variables governing
their formation, such as magma composition, crystallinity,
Granite on Mercury and Venus? Facts and speculations volume, effusion rate, cooling rate, pre-existingtopography,
Compared with the the Moon and Mars, relatively may differ. The association of pancake domes, coronae
scarce information on Mercury and Venus is available. and dyke swarms lends support to the notion that granite
Mercury is currently known only by Mariner 10 imagery bodies with batholithic sizes might exist on and beneath
acquired in 1974. Three tectonic stages are identified, like the surface of Venus (Petford, 2001).
on the Moon, though they differ in intensity (Thomas et Some Venusian rocks have been analysed in-situ by XRF
al., 1982). It seems reasonable to assume that Mercury installed on the Venera stations, that have landed
geology is similar to that of the Moon. As lunar granites exist, successively during 1972 to 1985. The crust at the landing
occurrence of granite on Mercury cannot be ruled out. sites appears to be dominated by low-K tholeiitic basalts
Despite similar size (Venus has a diameter of 0.94 times (Surkov et al., 1982). However, two sites (Venera 8 and
the terrestrial one), mass, density and global compostion, 13) reveal compositions with 4.0-4.8 wt.% K,O contents,
Venus and the Earth are geologically different. The and high U, Th abundances, suggesting the possible
Venusian surface is governed by the following factors: (i) occurrence of more silicic (granitic) compositions (Surkov
the atmospheric pressure is 9.8 MPa (only 0.1 MPa on the et al., 1986).
Earth), (ii) the outer shell of the planet is considered as It is speculated (Petford, 2001) that partial (about 20%)
constituting a single lithospheric plate, (iii) low cratering fluid-absent melting of altered rocks, like the tholeiitic
densities imply that a large resurfacing (catastrophic?) basalts analysed by Venera landers, can leave a garnet-
event occurred 500 Ma ago, or perhaps later (Namiki and bearing pyroxenite residue and produce high-Na and Al,
Solomon, 19941, (iv) altimetric data, acquired by radar low-K, silicic (>70 wt.% SiO,) liquids, resembling the
imagery since 1978 (Pioneer Venus) to 1990 (Magellan), Archaean and more recent adakites and trondhjemites
evidence a unimodal population with a mean elevation (Martin, 1994; Rapp, 1997). Repeated partial melting
of O.OkO.5 km. Two highland areas (Ishtar and Aphrodite episodes generating silicic magmas in the garnet stability
Terra), with elevations not exceeding 7 km, are analogous field would result over time in a bulk increase of crustal
in topographic style to the terrestrial continents and yield density, with important long-term consequences for
evidence for both compressional and extensional tectonic topographic and isostatic evolution of lithosphere (Namiki
regimes (Vorder Bruegge and Head, 1989; Ansan, 1993). and Solomon, 1993). The combination of these lines of
By analogy with the terrestrial continents, Ishtar and evidence adds fuel to more speculations about the nature
Aphrodite Terra can contain felsic rocks resembling the and evolution of a granitic crustal component in Venus.
Archaean continental nuclei (Turcotte, 1996).
The volcanic evidence (Head et al., 1992) is more Discussion
ambiguous. Besides large caldera-bearing shield volcanoes
resembling those on the Earth and on Mars, small volcanic The survey of the literature devoted to extraterrestrial
edifices were emplaced in the plains, the origin of which granite results in a paradox: so far, granite is found in
is not fully understood (Guest et al., 1992). The Venusian chondrites, iron meteorites and on the Moon, i.e., in small
literature is referring to the typical steep-sided domes, planetary bodies. On the contrary, large planets, such as
with a mean diameter of 24 km and heights of several Mars and Venus, provide only faint, indirect, or ambiguous

Gondwana Research, V. 5, No. 2, 2002


270 B. BONIN ET AL

Table 1. The genetic classification of granitoids based on the number of partial melting events from an original peridotitic mantle source
(Clarke, 1992).

Number of partial Stages Genetic type


meltinrr events
1 Ia. Partial melting of mantle peridotite produces a basaltic Primary granitoids
magma that undergoes extensive fractional crystallization, oceanic plagiogranites
or liquid immiscibility, to produce a granitoid rock non-orogenic (per)alkaline granitoids
most Proterozoic TTG suites
2 Ib. Partial melting of mantle peridotite produces a basaltic Secondary granitoids
magma that crystallizes directly or indirectly as basalt, gabbro, some plagiogranites most Archaean TTG suites
amphibolite or cclogite, and forms ancient terrestrial crust or continental arc metaluminous
modern oceanic crust tonalite-granodiorite suites
11. Partial melting of these mafic rocks produces K-poor felsic
magma volumctrically important in forming continental crust

3 Ib. (as above) Tertiary granitoids


11. (as above) Peraluminous granitoids
111. Direct melting of pre-existing granitoid rocks, or weathering
and erosion of intermediate to felsic igneous rocks of continental
crust to produce quartzo-feldspathic and pelitic sedimentary rocks
that subsequently partially melt to produce even more silicic magmas

evidence for the occurrence of granite. Sampling bias produce a granitic liquid by partial melting. In a provocative
obviously constitutes a formidable problem, that cannot paper, Campbell and Taylor (1983) argue that the absence
be solved easily. Even on the Earth, if the area covered by of water on Venus precluded the formation of extensive
the ocean floor is considered, basalt and related rocks are continental crust. This idea is questionable because liquid
better represented than granite. Any discussion on water is absent on Venus, but the H,O compound can
probability of extraterrestrial granite in the solar system occur. Water at the liquid state should not be mistaken
should take this fact into account. for water as the H,O compound. In the terrestrial
Granite compositions correspond to thermal minima magmatic processes, H,O is never in its liquid state. It is
in experimental silicate systems, which means that they either bound in hydrous minerals, such as amphiboles
can be produced through different ways, summarized in and micas, or dissolved in silicate liquids, or constituting
table 1 (from Clarke, 1992). The main processes capable a discrete fluid phase above the pressure (21.8 MPa) and
of generating the different granite types are partial melting temperature (374°C) critical values. Thus, there is no
and magma differentiation, including solid-liquid reason to suppose that liquid water is necessary for the
fractionation and liquid-liquid immiscibility. The latter production of granitic liquids.
process, that was considered as insignificant on the Earth, Experimental evidence shows that granitic liquids are
is clearly identified in the iron meteorites and on the easier to produce in the presence of H,O than in dry
Moon. Completely dry granites, having crystallized under systems. Water acts as a catalyst in magmatic processes,
low oxygen fugacities, evidence that water is not a and promotes generation and segregation of granitic
prerequisite for the generation of granitic liquids. On the liquids either by partial melting (anatexis) of a solid
Earth, the oceanic crust yields a main basaltic composition, matrix, or by complete differentiation of a basic parent
but can contain up to 3% granites, which correspond to magma into a highly evolved granitic residual liquid and
dry to water-deficient felsic residual liquids issued from mafic-ultramafic cumulates. Accordingly, water plays a
MORB by crystal fractionation (Dixon and Rutherford, role, not in the qualitative process of the generation of
1983). Accordingly, the basic crusts of Mars and Venus granite, but in the quantitative process of continental
are likely to contain granitic materials. growth, since the continental crust is typified by large
The Earth is the only planet in the solar system to volumes of granitic rocks.
retain liquid water. In the literature on planetary
differentiation, it is frequently argued that water in the Conclusion
Earth is the major factor controlling the generation of
granite and the development of continental crust. The Continents and plate tectonics are mutually necessary,
‘wet’solidus is reached at much lower temperatures than and there is a large room for doubt concerning continental
the dry one, implying that less energy is required to growth on the terrestrial planets other than the Earth.

Gondwuna Research, V. 5, No. 2,2002


GRANITE: A PLANETARY POINT OF VIEW 271

Turcotte (1996) states that only the Earth has continents Carlson, R.W. (1994) Mechanisms of Earth differentiation:
and plate tectonics, though this hypothesis is sometimes consequences for the chemical structure of the mantle. Rev.
challenged (see Sleep, 1994). The Moon and the other Geophys., v. 32, pp. 337-361.
Cipa, A. (1994) L r volcan Olympus Mons et ses environs, Mars.
small planetary bodies are unlikely to have produced
Ph.D. Thesis, UniversitC de Paris-Sud, Orsay (France), 243p.
enough granite to make continents, e.g., the K-rich portion Cipa, A., BCbien, J. and Bonin, B. (1995) Do flank terraces and
of the Moon can contain no more than -1% granites basal scarp indicate large-scale gravitational gliding on the
(Shervais and McGee, 1999). On Mars, no evidence exists Martian volcano Olympus Mons? C. R. Acad. Sci. Paris,
for continental differentiation, though silicic formations V. 322, pp. 369-376.

are likely to occur. Finally, Venus has apparently accreted Clague, D.A., Moore, J.G. and Reynolds, J.R. (2000) Formation
small continental terrains during its early history. Hence, of submarine flat-topped volcanic cones in Hawaii. Volcanol.
Bull., V. 62, pp. 214-233.
the ultimate question is not whether granite occurs in a Clark, B.C. and Baird, A.K. (1979) Chemical analyses of Martian
given planetary body, but rather whether sufficient surface materials: status report. Lunar Planet. Sci., v. X,
volumes of granitic materials were produced to constitute pp. 215-217.
stable continental nuclei. Clarke, D.B. (1992) Granitoid rocks. Topics in the earth sciences
7, Chapman and Hall, London, 283p.
Crown, D.A. and Greeley, R. (1993) Volcanic geology of Hadriaca
References Patera and the eastern Hellas region of Mars. J. Geophys.
Res., v. 98, pp. 3431-3451.
Aguirre-Puente, J., Costard, F. and Posado-Cano, R. (1994)
Dixon, S. and Rutherford, M.J. (1979) Plagiogranites as late-
Contribution to the study of thermal erosion on Mars. J.
stage immiscible liquids in ophiolite and mid-ocean ridge
Geophys. Res., v. 99, pp. 5657-5667.
suites: an experimental study. Earth Planet. Sci. Lett., v. 45,
Ansan, V. (1993) InteprCtations gkologiques et structurales de
Venus i partir des images radar Venera 15-16 et Magellan. pp. 45-60.
Dixon, S. and Rutherford, M.J. (1983) The origin of rhyolite
Ph.D. Thesis, UniversitC de Paris-Sud, Orsay (France), 270p.
Bandfield, J.L., Hamilton, V.E. and Christensen, P.R. (2000) A and plagiogranite in oceanic crust: an experimental study.
global view of Martian surface compositions from MGS-TES. J. Petrol., v. 24, pp. 1-25.
Science, v. 287, pp. 1626-1630. Dreibus, G. (2000) The chemical composition of the Martian
Bertka, C. and Holloway, J.R. (1994a) Anhydrous partial melting crust (abst.). 3lStIGC, Rio de Janeiro (Brazil).
of an iron-rich mantle I: subsolidus assemblages and partial Eskola, P. (1932) Origin of granites. Mineral. Petrol. Mitteil., v.
melting phase relations at 10 to 30 kb. Contrib. Mineral. 42, pp. 455-481.
Petrol., v. 115, pp. 313-322. Fink, J.H., Bridges, N.T. and Grimm, R.E. (1993) Shapes of
Bertka, C. and Holloway, J.R. (1994b) Anhydrous partial melting Venusian ‘pancake’ domes imply episodic emplacement and
of an iron-rich mantle 11: primary melt compositions at 15 silicic composition. Geophys. Res. Lett., v. 20, pp. 261-264.
kb. Contrib. Mineral. Petrol., v. 115, pp. 323-338. Francis, PW. and Wood, C.A. (1982) Absence of silicic volcanism
Bischoff, A. (1993) Alkali-granitoids as fragments within the on Mars: implications for crustal compositions and volatile
ordinary chondrite Adzhi-Bogdo: evidence for highly abundance. J. Geophys. Res., v. 87, pp. 9881-9889.
fractionated, alkali-granitic liquids on asteroids (abst.). Lunar Greeley, R. and Spudis, P.D. (1981) Volcanism on Mars. Rev.
Planet. Sci., v. XXW, pp. 113-114. Geophys. Space Phys., v. 19, pp. 13-41.
Bischoff, A. (2001) Meteorite classification and the definition Guest, J.E., Bulmer, M.H., Aubele, J., Beratan, K., Greeley, R.,
of new chondrite classes as a result of successful meteorite Head, J.W., Michaels, G., Weitz, C. and Wiles, C. (1992)
search in hot and cold deserts. Planet. Space Sci., v. 49, pp. Small volcanic edifices and volcanism in the plains of Venus.
769-776. J. Geophys. Res., v. 97, pp. 15949-15966.
Bischoff, A., Geiger, T., Palme, H., Spettel, B., Schultz, L., Scherer, Harder, H. and Christensen, U.R. (1996) A one-plume model of
P., Schluter, J. and Lkhamsuren, J. (1993) Mineralogy, Martian mantle convection. Nature, v. 380, pp. 507-509.
chemistry, and noble gas contents of Adzhi-Bogdo - an LL3- Hartmann, W.K., Malin, M., McEwen, A., Can; M., Soderblom,
6 chondritic breccia with L-chondritic and granitoidal clasts. L., Thomas, P., Danielson, E., James, E! andveverka, J. (1999)
Meteoritics, v. 28, pp. 570-578. Evidence for young volcanism on Mars from crater counts.
Bonin, B., Dubois, R. and Gohau, G. (1997) Le metamorphisme Nature, v. 397, pp. 586-589.
et la formation des granites. Evolution des idCes et concepts Hawkesworth, C., Kelley, S., Turner, S., Le Roex, A. and Storey,
actuels. Fac Sciences, Nathan-UniversitC, Paris, 317p. B. (1999) Mantle processes during Gondwana break-up and
Burnett, D.S. and Wasserburg, G.J. (1967a) 87Rb-87Srages of dispersal. J. Afr. Earth Sci., v. 28, pp. 239-261.
silicate inclusions in iron meteorites. Earth Planet. Sci. Lett., Head, J.W., Crumpler, L.S., Aubele, J.C., Guest, J.E. and
V. 2, pp. 397-408. Saunders, R.S. (1992) Venus volcanism: classification of
Burnett, D.S. and Wasserburg, G.J. (1967b) Evidence for the volcanic features and structures, associations, and global
formation of an iron meteorite at 3.8 x lo9 years. Earth distribution from Magellan data. J. Geophys. Res., v. 97,
Planet. Sci. Lett., v. 2, pp. 137-147. pp. 13153-13197.
Campbell, I.H. and Taylor, S.R. (1983) No water, no granites - Jackel, A. and Bischoff, A. (1997) Potassium-rich fragments in
no oceans, no continents. Geophys. Res. Lett., v. 10, LL-chondritic breccias (abst.). Meteoritics and Planet. Sci.,
pp. 1061-1064. v. 32, A66.

Gondwana Research, I/; 5, No. 2,2002


272 B. BONIN ET AL.

Le Maitre, R.W. (Ed.), (1989) A Classification of igneous rocks Academic and Professional, Chapman and Hall, London-
and glossary of terms. Recommendations of the IUGS Glasgow, 321p.
subcommission on the systematics of igneous rocks. Rapp, R.P. (1997) Heterogeneous source regions for Archaean
Blackwell Sci. Publ., Oxford, 193p. granitoids: experimental and geochemical evidence. In: de
Martin, H. (1994) The Archean grey gneisses and the genesis of Wit, M.J. and Ashwal, L.D. (Eds.), Greenstone Belts.
continental crust. In: Condie, K.C. (Ed.), Archean crustal Clarendon Press, Oxford, pp. 267-279.
evolution, developments in Precambrian Geology 11, Rieder, R., Economou, T., Wanke, H., Turkevich, A., Crisp, J.,
Elsevier, Amsterdam, pp. 205-259. Bruckner, J., Dreibus, G. and McSween, H.Y. Jr. (1997) The
McEwen, A.S., Malin, M.C., Carr, M.H. and Hartmann, W.K. chemical composition of Martian soil and rocks returned from
(1999) Voluminous volcanism on early Mars revealed in the mobile alpha proton X-ray spectrometer: preliminary
Valles Marineris. Nature, v. 397, pp. 584-586. results from the X-ray mode. Science, v. 278, pp. 1771-1774.
McKenzie, D., Ford, P.G., Lui, F. and Pettengill, G.H. (1992) Roedder, E. and Weiblen, P.W. (1970) Lunar petrology of silicate
Pancake-like domes on Venus. J. Geophys. Res., v. 97, melt inclusions, Apollo 11 rocks. Proc. Apollo 11 Lunar Sci.
pp. 15967-15976. Conf., 1, pp. 801-837.
McSween, H.Y. Jr. (1994) What we have learned about Mars Roedder, E. and Weiblen, P.W. (1972) Petrographic features and
from SNC meteorites. Meteoritics, v. 29, pp. 757-779. petrologic significance of melt inclusions in Apollo 1 4 and
McSween, H.Y. Jr., Murchie, S.L., Crisp, J.A., Bridges, T., 15 rocks. Proc. 3rd Lunar Sci. Conf., 1, pp. 251-279.
Anderson, R.C., Bell, J.E III., Britt, D.T., Briickner, J., Dreibus, Roedder, E. and Weiblen, P.W. (1978) Melt inclusions in Luna
G., Economou, T., Ghosh, A., Golombek, M.P., Greenwood, 24 soil fragments. In: Merrill, R.B. and Papike, J.J. (Eds.),
J.P., Johnson, J.R., Moore, H.J., Morris, R.V., Parker, T.J., Mare Crisium: the view from Luna 24. Pergamon Press, New
Rieder, R., Singer, R. and Wanke, H. (1999) Chemical, York, pp. 495-522.
multispectral, and textural constraints on the composition Sanz, H.G., Burnett, D.S. and Wasserburg, G.J. (1970) Aprecise
and origin of rocks at the Mars Pathfinder landing site. J. 87Rb/87Srage and initial 87Sr/86Srfor the Colomera iron
Geophys. Res., v. 104, pp. 8679-8715. meteorite. Geochim. Cosmochim: Acta, v. 34, pp. 1227-1239.
McSween, H.Y. Jr., Grove, T.L., Lentz, R.C.F., Dann, J.C., Scott, D.H. and Tanaka, K.L. (1982) Ignimbrites of Amazonis
Holzheid, A.H., Riciputi, L.R. and Ryan, J.G. (2001) Planitia region of Mars. J. Geophys. Res., v. 87, pp. 9839-
Geochemical evidence for magmatic water within Mars 985 1.
from pyroxenes in the Shergotty meteorite. Nature, v. 409, Shervais, J.W. and McGee, J.J. (1999) Petrology of the Western
pp. 487-490. Highland Province: ancient crust formation at the Apollo 14
Mouginis-Mark, P.J., Wilson, L. and Head, J.W. I11 (1982) site. J. Geophys. Res., v. 104, pp. 5891-5920.
Explosive volcanism on Hecates Tholus, Mars: investigation Shervais, J.W. and Taylor, L.A. (1983) Micrographic granite:
of eruption conditions. J. Geophys. Res., v. 87, pp. 9890- more from Apollo 1 4 (abst.) Lunar Planet. Sci., v. XIY
9904. pp. 369-370.
Mouginis-Mark, P.J., Wilson, L. and Zimbelman, J.R. (1988) Shih, C.-Y., Nyquist, L.E., Bogard, D.D., McKay, G.A., Wooden,
Polygenic eruptions on Alba Patera, Mars. Bull. Volcanol., J.L., Bansal, B.M. and Wiesmann, H. (1982) Chronology and
V. 50, pp. 361-379. petrogenesis of young achondrites, Shergotty, Zagami, and
Mysen, B.O., Virgo, D., Popp, R.K. and Bertka, C.M. (1998) The ALHA77005: late magmatism on a geologically active planet.
role of H,O in Martian magmatic systems. Amer. Mineral., Geochim. Cosmoschim. Acta, v. 46, pp. 2323-2344.
V. 83, pp. 942-946. Shih, C.-Y., Nyquist, L.E. and Wiesmann, H.(1993) K-Ca
Namiki, N. and Solomon, S.C. (1993) The gabbro-eclogite phase chronology of lunar granites. Geochim. Cosmochim. Acta,
transition and the elevation of mountain belts on Venus. J. V. 57, pp. 4827-4841.
Geophys. Res., v. 98, pp. 15025-15031. Sleep, N.H. (1994) Martian plate tectonics. J. Geophys. Res., v.
Namiki, N. and Solomon, S.C. (1994) impact crater densities 99, pp. 5639-5655.
on volcanoes and coronae on Venus: implications for volcanic Smith, D.E., Zuber, M.T., Solomon, S.C., Phillips, R.J., Head,
resurfacing. Science, v. 265, pp. 929-933. J.W., Garvin, J.B., Banerdt, W.B., Muhleman, D.O., Pettengill,
Nehlig, P., Leyrit, H., Dardon, A., Freour, G., de Goer de Herve, G.H., Neumann, G.A., Lemoine, F.G., Abshire, J.B.,
A., Huguet, D. and ThiCblemont, D. (2001) Repeated Aharonson, O., Brown, C.D., Hauck, S.A., Ivanov, A.B.,
coristruction and destruction of the Cantal stratovolcano. McGovern, P.J., Zwally, H.J. and Duxbury, T.C. (1999) The
SOC.g6ol. Fr. Bull., v. 172, pp. 295-308. global topography of Mars and implications for surface
Nyquist, L.E. and Shih, C.-Y. (1992) The isotopic record of lunar evolution. Science, v. 284, pp. 1495-1503.
volcanism. Geochim. Cosmochim. Acta, v. 56, pp. 2213-2234. Smith, D.K. (1996) Comparison of the shapes and sizes of
ONeill, H.St.C. (1991a) The origin of Moon and the early history seafloor volcanoes on Earth and ‘pancake’domes on Venus.
of the Earth - a chemical model. Part 1:The Moon. Geochim. J. Volcanol. Geoth. Res., v. 73, pp. 47-64.
Cosmochim. Acta, v. 55, pp. 1135-1157. Sun, S.S. and McDonough, W.E (1989) Chemical and isotopic
ONeill, H.St.C. (1991b) The origin ofMoon and the early history systematics of oceanic basalts: implications for mantle
of the Earth - a chemical model. Part 2: The Earth. Geochim. composition and processes. In: Saunders, A.D. and Norry,
Cosmochim. Acta, v. 55, pp. 1159-1172. M.J. (Eds.), Magmatism in the ocean basins, Geol. SOC.
Petford, N. (2001) Dyke widths and ascent rates of silicic magmas London, Spl. Publ. 42, pp. 313-345.
on Venus. Trans. R. SOC.Edinb.: Earth Sci., v. 91, pp. 87-95. Surkov, YA., Moskalyeva, L.P., Shleglov, O.P., Kharyukova, VP.,
Pitcher, W.S. (1993) The nature and origin of granite. Blackie Manvelyan, O.S. and Smirnov, G.G. (1982) Preliminary

Gondwana Research, V. 5, No. 2,2002


GRANITE: A PLANETARY POINT OF VIEW 273
~

results on the determination of the elemental compound of Warren, PH., Taylor, G.F., Keil, K., Shirley, D.N. and Wasson, J.T.
rocks of Venus, from the automatic interplanetary stations (1983) Petrology and chemistry of two ‘large’granite clasts
Venera 13and Venera 14. Lett. Astronom. J., v. 8, pp. 437-443. from the Moon. Earth Planet. Sci. Lett., v. 64, pp. 175-185.
Surkov, Y.A., Kirnozov, F.F., Soborpov, O.B., Glazov, V.N., Warren, P.H. and Wasson, J.T. (1977) Pristine nonmare rocks
Dunkchenko, A.G. and Tatsy, L.P. (1986) Content in uranium, and the nature of lunar crust. Proc. 8th Lunar Sci. Conf.,
thorium and potassium in the rocks of Venus in the region pp. 2215-2235.
of landing of the landers Vega-1 and Vega-2. Lett. Astronom. Watson, E.B. (1976) Two-liquid partition coefficients:
J., V. 12, pp. 111-119. experimental data and geochemical implications. Contrib.
Tanaka, K.L. (1985) Ice-lubricated gravity spreading of Olympus Mineral. Petrol., v. 56, pp. 119-134.
Mons aureole deposits. Icarus, v. 62, pp. 191-206. Wedepohl, K.H. (1991) Chemical composition and fractionation
Thomas, P. and Masson, P. (1986) Martian fluidized crater of the continental crust. Geol. Rundsch., v. 80, pp. 207-223.
distribution: tectonic implications. Earth, Moon and Planets, Wilson, L. and Head, J.W. 111. (1994) Mars: review and analysis
V. 34, pp.169-176. of volcanic eruption theory and relationships to observed
Thomas, P., Masson, P. and Fleitout, L. (1982) Global volcanism landforms. Rev. Geophys., v. 32, pp. 221-263.
and tectonism on Mercury: comparison with the Moon. Earth Zuber, M.T., Solomon, S.C., Phillips, R.J., Smith, D.E., Tyler, G.L.,
Planet. Sci. Lett., v. 58, pp. 95-103. Aharonson, O., Balmino, G., Banerdt, W.B., Head, J.W.,
Turcotte, D.L. (1996) Magellan and comparative planetology. Johnson, C.L., Lemoine, F.G., McGovern, P.J., Neumann,
J. Geophys. Res., v. 101, pp. 4765-4773. G.A., Rowlands, D.D. and Zhong, S. (2000) Internal structure
Vorder Bruegge, R.W. and Head, J.W. (1989) Fortuna Tessera, and early thermal evolution of Mars from Mars Global
Venus: evidence of horizontal convergence and crustal Surveyor topography and gravity. Science, v. 287, pp. 1788-
thickening. Geophys. Res. Lett., v. 16, pp. 699-702. 1793.

Gondwana Research, f 5 , No. 2,2002