Eclogites and eclogites in the Western Gneiss Region, Norwegian Caledonides€¦ · Eclogites and eclogites in the Western Gneiss Region, Norwegian Caledonides S.J. Cuthbert a,),

Ž .Lithos 52 2000 165–195www.elsevier.nlrlocaterlithos

Eclogites and eclogites in the Western Gneiss Region, NorwegianCaledonides

S.J. Cuthbert a,), D.A. Carswell b, E.J. Krogh-Ravna c, A. Wain d,1

a Geology Section, Department of CiÕil Structural and EnÕironmental Engineering, UniÕersity of Paisley, High Street,Paisley PA1 2BE, UK

b Department of Earth Sciences, UniÕersity of Sheffield, Dainton Building, Brookhill, Sheffield S3 7HF, UKc Institute of Geology, UniÕersity of Tromsø, Tromsø 9000, Norway

d Department of Earth Sciences, UniÕersity of Oxford, Oxford OX1 3PR, UK

Abstract

Ž .The Western Gneiss Region WGR marks the outcrop of a composite terrane consisting of variably re-workedProterozoic basement and parautochthonous or autochthonous cover units. The WGR exhibits a gross structural, petrographic

Ž .and thermobarometric zonation from southeast to northwest, reflecting an increasing intensity of Scandian late PalaeozoicŽ .metamorphic and structural imprint. Scandian-aged eclogites have been widely though for kinetic reasons not invariably

stabilised in metabasic rocks but have suffered varying degrees of retrogression during exhumation. In the region betweenthe Jostedal mountains and Nordfjord, eclogites commonly have distinctively prograde-zoned garnets with amphibolite or

Ž w x .epidote–amphibolite facies solid inclusion suites and lack any evidence for stability of coesite high pressure HP eclogites .In the south of this area, in Sunnfjord, eclogites locally contain glaucophane as an inclusion or matrix phase. North ofNordfjord, eclogites mostly lack prograde zoning and evidence for coesite, either as relics or replacive polycrystalline quartz,

Ž w x .is present in both eclogites ultrahigh pressure UHP eclogites and, rarely, gneisses. Coesite or polycrystalline quartz aftercoesite has now been found in eight new localities, including one close to a microdiamond-bearing gneiss. These newdiscoveries suggest that, by a conservative estimate, the UHP terrane in the WGR covers a coastal strip of about 5000 km2

between outer Nordfjord and Moldefjord. A ‘‘mixed HPrUHP zone’’ containing both HP and UHP eclogites is confirmedby our observations, and is extended a further 40 km east along Nordfjord. Thermobarometry on phengite-bearing eclogites

Ž . Ž .has been used to quantify the regional distribution of pressure P and temperature T across the WGR. Overall, a scenarioemerges where P and T progressively increase from 5008C and 16 kbar in Sunnfjord to )8008C and 32 kbar in outerMoldefjord, respectively, in line with the distribution of eclogite petrographic features. Results are usually consistent withthe silica polymorph present or inferred. The P–T conditions define a linear array in the P–T plane with a slope of roughly58Crkm, with averages for petrographic groups lying along the trend according to their geographic distribution from SE toNW, hence defining a clear field gradient. This P–T gradient might be used to support the frequently postulated model fornorthwesterly subduction of the WGC as a coherent body. However, the WGC is clearly a composite edifice built fromseveral tectonic units. Furthermore, the mixed HPrUHP zone seems to mark a step in the regional P gradient, indicating apossible tectonic break and tectonic juxtaposition of the HP and UHP units. Lack of other clear evidence for a tectonic breakin the mixed zone dictates caution in this interpretation, and we cannot discount the possibility that the mixed zone is, at

) Corresponding author. E-mail: [email protected] Present address: British Geological Survey, Murchison House, West Mains Road, Edinburgh EH9 3LA, UK.

0024-4937r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S0024-4937 99 00090-0

( )S.J. Cuthbert et al.rLithos 52 2000 165–195166

least, partly a result of kinetic factors operating near the HP–UHP transition. Overall, if the WGC has been subducted duringthe Scandian orogeny, it has retained its general down-slab pattern of P and T in spite of any disruption during exhumation.Garnetiferous peridotites derived from subcontinental lithospheric mantle may be restricted to the UHP terrane and appear todecorate basement-cover contacts in many cases. P–T conditions calculated from previously published data for both relictŽ . Ž .Proterozoic lithospheric mantle? porphyroclast assemblages and Scandian subduction-related? neoblastic assemblages donot define such a clear field gradient, but probably record a combination of their pre-orogenic P–T record with Scandianre-working during and after subduction entrainment. A crude linear array in the P–T plane defined by peridotite samplesmay be, in part, an artifact of errors in the geobarometric methods. A spatial association of mantle-derived peridotites withthe UHP terrane and with basement-cover contacts is consistent with a hypothesis for entrainment of at least some of themas ‘‘foreign’’ fragments into a crustal UHP terrane during subduction of the Baltic continental margin to depths of )100km, and encourages a more mobilistic view of the assembly of the WGC from its component lithotectonic elements. q 2000Elsevier Science B.V. All rights reserved.

Keywords: Eclogite; Western Gneiss Region; Norwegian Caledonides

1. Introduction

In order to determine the significance of ultrahighŽ .pressure UHP metamorphism in deep crustal geo-

dynamic processes, it is clearly important to establishthe nature and extent of the UHP lithotectonic units,the nature of their boundaries, and their orogenic

Ž .setting. The Western Gneiss Region WGR in theNorwegian Caledonides is well-known for its spec-tacular occurrences of eclogites and garnetiferous

Ž .peridotites. This extensive high pressure HP meta-morphic terrane is accessible, very well-exposed andits orogenic setting has been well-characterised.Hence, the WGR provides us with an excellentnatural laboratory for the examination of the setting,nature and distribution of HP and UHP metamor-phism, and the dynamics, physical properties andbehaviour of subducted continental crust.

Coesite was discovered in eclogites from the WGRŽsoon after its discovery in the Alps Chopin, 1984;

.Smith, 1984 . On the basis of seven scattered loca-Ž .lities showing evidence for coesite, Smith 1988

established a ‘‘coesite–eclogite province’’ in theŽ .Stadlandet area of coastal SW Norway Fig. 1 .

Ž .Subsequently, Wain 1997a; b expanded the databaseof known localities where coesite or polycrystallineaggregates after coesite are known to a total of 24

Ž .localities and Krabbendam and Wain 1997 wereable to map a boundary zone to the coesite–eclogiteprovince, distinguishing a group of UHP eclogitesfrom a group of normal HP eclogites to the south.

Ž .Wain 1997a; b established a pressure difference of)4 kbar between HP and UHP eclogites in the

boundary to the UHP province, and suggested thatthis may mark a tectonic boundary to a distinct UHPlithotectonic unit.

Ž .In spite of the large area of the WGR Fig. 1 ,UHP eclogite discoveries have generally remainedrestricted to the same general area in the earlydiscoveries, i.e., in outer Nordfjord and StadlandetŽ . Ž .Figs. 1–3 of Smith 1984; 1988 . In the meantime,the tantalising discovery of rare microdiamondsŽ .Dobrzhinetskya et al., 1995 from pelitic andmigmatitic gneisses at Fjørtoft, about 70 km NE of

Ž .Stadlandet Figs. 1 and 2 indicated that thecoesite–eclogite province may be much more exten-sive.

This contribution describes some new discoveriesof UHP rocks which extend the area of the UHPterrane in the WGR. We also examine the regionalcontext of UHP metamorphism in relation to theadjacent HP eclogite terranes and mappable lithotec-tonic units, and finally make some comments on thepossible nature of the boundaries to the UHP terranein terms of tectonic and kinetic factors.

2. Geological setting

The WGR occupies an area of approximately5=104 km2 along the coast of southwest Norway

Ž .between Bergen and Trondheim Fig. 1 . The WGRcontains the lowest exposed structural level in thesouthern Scandinavian Caledonides. It lies within alarge tectonic window, and is surrounded by the

( )S.J. Cuthbert et al.rLithos 52 2000 165–195 167

Ž .Fig. 1. Simplified lithotectonic and metamorphic map of the southern Scandinavian Caledonides after Roberts and Gee 1985 , showing thedistribution of eclogite facies metamorphism. M — Maløy; Mo— Molde; R — Romsdal; S — Stryn; F — Førde; D — Dalsfjord.˚

Ž . Ž .Location of diamond gneiss from Dobrzhinetskya et al. 1995 , locations of garnetiferous ultramafic rocks from Krogh and Carswell 1995 .Question marks indicate uncertainty over location of UHP terrane or mixed HPrUHP zone.

outcrop of the Caledonide allochthon which consistsof a thick stack of thrust sheets. To the southeast andeast of the allochthon lies the Baltic foreland, mostlyconsisting of Proterozoic high-grade gneisses,metasediments and metavolcanics and their au-

tochthonous or para-autochthonous cover of late Pro-terozoic and lower Palaeozoic sediments. The al-lochthon is composed of sheets of Baltic basementgneisses and their lower Palaeozoic sedimentarycover, overthrust by ophiolitic, arc-related and some-


Fig. 2. Distribution of petrographic types of eclogites in the coastal parts of the WGR between Sognfjord and Molde, also showing theŽ . Ž . Ž .location of cover units, peridotites and microdiamond. Geological units after Kildal 1970 , Robinson 1995 , Tveten 1995 , Tveten and

Ž . Ž . Ž . Ž . Ž . Ž .Lutro 1995a; b . Eclogite localities from Krogh 1980; 1982 , Cuthbert 1985 , Griffin et al. 1985 , Smith 1988 , Bailey 1989 , ChauvetŽ . Ž .et al. 1992 and Krabbendam and Wain 1997 and additional unpublished data of the authors.

times continental units with a provenance outboardŽof the Baltican margin Roberts and Gee, 1985;

.Stephens, 1988 .

The contact of the basement gneisses and theirautochthonous cover with the base of the allochthonhas had a complex evolution and shows evidence for


Fig. 3. Lithotectonic and metamorphic map of the outer Nordfjord–Stadlandet area showing the distribution of petrographic types of eclogiteand associated HP and UHP gneisses, and peridotites. Dashed lines mark the limits of the HP and UHP zones, and the extent of the mixedHPrUHP zone.

southeast- or eastward-directed thrusting, followedby a reversal of motion in a normal sense to the

Žnorthwest or west Hossack, 1984; Norton, 1986,1987; Chauvet and Brunel, 1988; Chauvet and Ser-anne, 1989; Fossen and Rykkelid, 1992; Wilks and

.Cuthbert, 1994; Wennberg, 1996 . This extensionalŽreworking took place in at least two stages Ander-

sen and Jamtveit, 1990; Cuthbert, 1991; Cuthbertand Carswell, 1990; Hartz et al., 1994; Wilks and

.Cuthbert, 1994; Milnes et al., 1997 with earlier,


amphibolite facies, ductile shear zones at lower tomid-crustal levels being reactivated or cut across bylater, Middle Devonian, brittle normal and strike–slipfaults within the cover which accommodated the fillof the late orogenic Old Red Sandstone molasse

Ž .basins now lying along the west coast Fig. 2 . Asignificant element of transtensional or transpres-sional deformation played a role during this period,

Žespecially in the north of the WGR Seranne, 1992;´Krabbendam and Dewey, 1998; Robinson and Terry,

.1998 . The net result of the extensional tectonicphases was to bring the WGR rocks up against thecover in the footwall of a system of low-anglenormal faults, so that the WGR rocks now lie withina large metamorphic core complex. Evidence for theearliest, eclogite-facies, subduction and exhumationfabrics is sporadic and highly localised, and thus isdifficult to fit into the regional context.

The predominant lithologies exposed in the WGRare broadly granitoid migmatitic and augen or-thogneisses which have a Proterozoic provenanceŽBrueckner, 1972, 1979; Pidgeon and Raheim, 1972;˚Skjerlie and Pringle, 1978; Lappin et al., 1979;

.Harvey, 1983; Tucker et al., 1990 . These gneissesŽ .form the Jostedal Complex of Bryhni 1966 , the

Ž .Fetvatn Gneiss of Brueckner 1977 , the LønsetŽ .gneiss of Krill 1985 and the Baltica Basement of

Ž .Robinson 1995 , and correlate with the Baltic Pro-Žterozoic basement in the foreland Milnes et al.,

.1997 . They are referred to as ‘‘basement’’ here.Other common lithologies are meta-anorthosites andmetamorphosed pyroxene syeniticrmonzonitic rocksŽ .‘‘mangerites’’ , pelitic schists, quartzites, calc-sili-cates and marbles. These diverse lithologies are oftenspatially associated and lie within synclinal ‘‘keels’’downfolded into the dominant orthogneissic base-

Žment Muret, 1960; Bryhni, 1966, 1989; Bryhni andGrimstad, 1970; Brueckner, 1977; Krill, 1985;Tveten, 1995; Robinson, 1995, 1997; Robinson and

.Terry, 1998; Terry and Robinson, 1998 . This lattercharacteristic assemblage of lithologies is typical ofthe Lower and Middle Allochthons, and part of the

Ž .Upper Allochthon Blahø–Surna nappe in the Scan-˚Ž .dinavian Caledonides Roberts and Gee, 1985 . Such

rocks have been designated as ‘‘supracrustal’’ byŽ . Ž . Ž .Bryhni 1966; 1989 and Tveten 1995 . Krill 1985 ,

Ž . Ž .Robinson 1995 and Terry and Robinson 1998have directly correlated such supracrustal rocks in

the northern part of the WGR with the main al-˚lochthon to the northeast. To the south of Alesund,

correlations cannot be made so directly but the com-mon association of anorthosites, mangerites,quartzites, marbles and pelitic schists is typical of theautochthonous cover and overlying nappes in the

ŽJotunheim area southeast of the WGR Strand, 1969;.Brueckner, 1977; Bryhni, 1989 which form the

Lower and Middle allochthons. Hence, by analogywith the known Caledonide autochthon and al-lochthonous nappes, we tentatively designate suchassemblages of lithologies outcropping within the

Ž .WGR as ‘‘cover’’ units lying upon a para- auto-chthonous Baltic basement, although we cannot en-tirely rule out the possibility that some anorthosites,mangerites and metasediments could be an integralpart of the autochthonous Proterozoic Baltic base-

Ž .ment. Furthermore, Krabbendam and Wain 1997Ž .and Wain 1998 have shown that some pelitic schist

Ž .units described by Bryhni 1966 from Nordfjord are,in fact, eclogite-facies equivalents of the commongranodioritic basement gneisses rather than sedimen-tary cover. Thus, at this stage, the extent of Caledo-nian cover within the Western Gneiss ComplexŽ .WGC remains a matter of debate. Nevertheless,mappable units of diverse lithology and origin dooutcrop in the WGR, some of which have certainlybeen emplaced by thrusting or possibly by exten-sional faulting. It follows that the WGR contains anassemblage of lithotectonic units which has not actedas a coherent entity throughout its history. For thisreason, the term ‘‘Western Gneiss Complex’’ will beused here to refer to the assemblage of rock unitsoccupying the WGR and the term ‘‘Western GneissRegion’’ will more properly refer to the geographicalarea in which the WGC outcrops, following Milnes

Ž .et al. 1997 .The WGC shows a systematic change in its struc-

tural and metamorphic characteristics from theŽ .southeast to the northwest and west Fig. 1 . In the

SE, close to the contact with the main allochthon atJotunheimen, eclogites are absent and Proterozoicstructures and parageneses are predominantly well-preserved, being similar to those in the foreland onthe SE side of the allochthon, where eclogites are

Ž .also lacking Milnes et al., 1988, 1997 . Passingwestwards and northwestwards, Caledonide deforma-tion becomes more pervasive and Proterozoic struc-


tures are progressively obliterated. Eclogites are firstrecognised about 40 km NW of the edge of the

Ž .allochthon NW of the Jostedal mountains and arerelatively common throughout the rest of the WGCwithin both basement and cover rock sequences.Early geothermometric studies on eclogites indicateda progressive increase in temperature towards the

Ž .NW Krogh, 1977; Griffin et al., 1985 consistentwith the pattern of increasing intensity of Caledonidetectonometamorphic reworking, but the pressure dis-tribution for peak eclogite facies metamorphism waspoorly constrained due to a lack of reliable geo-barometers. In the extreme NW of the WGR, eclog-ites show evidence for partial melting during ex-

Ž .humation Cuthbert, 1995, 1997 . The overall patternof increasingly intense Caledonide deformation andmetamorphism has been the basis for suggestionsthat the WGC has been subducted towards the NWŽ .in the present frame of reference during the Cale-donian collision between Baltica and LaurentiaŽGriffin et al., 1985; Cuthbert and Carswell, 1990;

.Cuthbert et al., 1983 .Ultramafic rocks are common throughout the

WGC but garnetiferous ultramafites are only knownŽfrom the region to the north of Nordfjord Figs. 1

.and 2 where they often occur as pods and sheetsŽlying along basement-cover contacts Brueckner,

.1977; Bryhni, 1989; Bucher-Nurminen, 1991 andattest to considerable crust–mantle tectonic interac-

Žtion during the Caledonide plate collision Jamtveit.et al., 1991; Brueckner and Medaris, 1998 .

In spite of the general trend of increased rework-ing towards the northwest, there is evidence formetastability on a large range of scales during eclog-ite facies metamorphism resulting in the localisedsurvival of pre-Caledonian parageneses throughout

Žthe WGC Gjelsvik, 1952; Griffin and Raheim, 1973;˚Cuthbert, 1985; Mørk, 1985; Wain, 1997a,c, 1998;

.Austrheim, 1998 . Metastability is also manifested inthe survival of eclogite facies relics in the WGC,which is now predominantly an amphibolite-faciesterrane. Clearly, the efficiency of conversion to dense,eclogite-facies parageneses and subsequent retro-grade metamorphism to lower density parageneseswas highly variable and was probably a complex

Žfunction of temperature, strain and fluid flux see.Austrheim, 1998 . In this context, it seems likely that

metamorphic transformations within the eclogite fa-

cies may also have varied in their efficiency and itmay be possible to find evidence preserved for meta-morphic processes operating at the HP–UHP transi-tion. We return to this theme in Section 6.

The timing of Caledonian metamorphism in theWGR is based on a rather limited data set but bothU–Pb zircon and Nd–Sm mineral isochrons foreclogites, presumed to approximate to timing ofeclogite crystallisation, give late Silurian ages of

Ž400–447 Ma with an average of 418 Ma Krogh etal., 1974; Griffin and Brueckner, 1980; Griffin and

.Brueckner, 1985; Mørk and Mearns, 1986 . K–Ar,Ar–Ar and Rb–Sr mineral ages are mostly youngerŽ .410–370 Ma and probably represent cooling ages

Žduring decompression Lux, 1985; Wilks and Cuth-bert, 1994; Boundy et al., 1996; see also the compi-

.lation of Kullerud et al., 1986 . These ages areconsistent with structural and stratigraphic evidencefor the major ‘‘Scandian’’ phase of thrust sheetemplacement and consequent subduction of the WGC

Žoccurring during the late Silurian Roberts and Gee,.1985 . No indisputably UHP eclogites have yet been

dated, so the age of UHP metamorphism is notwell-constrained. No age difference is discernablebetween eclogites from basement and cover unitswithin the WGC. However, eclogites are present insome units of the main outcrop of the allochthon and

Ž .these give pre-Scandian ages )450–505 Ma ,Žmostly older than those in the WGR Dallmeyer and

Gee, 1986; Mørk et al., 1988; Boundy et al., 1996,1997; Essex et al., 1997 but see Bingen et al., 1998for a Scandian U–Pb zircon age in the Bergen Arcs

.anorthosite nappe . If cover units within the WGCcorrelate with those in the main allochthon, it ispossible that some eclogites in the WGC may also bepre-Scandian, or have undergone two phases ofeclogite-facies metamorphism. Overall, it is clearthat an extensive isotopic dating programme is re-quired to ascertain the timing of HP–UHP metamor-phism in the WGR.

3. Field and petrographic characteristics of eclog-ites in the WGC

Eclogites in the WGC are usually found as lentic-ular pods or tabular layers from decimeters to a fewtens of meters thick within most of the other litholo-


gies comprising the WGC. A few much larger eclog-ites occur as large, tabular bodies several kilometersin outcrop extent, some of which were originally

Žlayered, mafic intrusions Mysen and Heier, 1971;.Cuthbert, 1985; Jamtveit, 1987 .

Most eclogites have retrograde selvages of amphi-bolite against their host gneisses. The gneisses them-selves usually have amphibolite facies mineral para-geneses, so that most eclogites are not co-facial withtheir host rocks. However, Krabbendam and WainŽ . Ž .1997 and Wain 1998 have shown that swarms ofeclogite pods commonly enclose, or are envelopedby, schists or gneisses with abundant relics of eclog-ite-facies parageneses including phengite, kyanite,zoisite, garnet and quartz"omphacite, or their sym-plectic breakdown products. Eclogites within suchmasses tend to lack retrogressive amphibolitic sel-vages and have fabric elements colinear with those

Žin the host rocks Krabbendam and Wain, 1997;Terry and Robinson, 1998; Terry et al., 1999; Wain,

.1998 . Some such gneisses contain polycrystallineŽ .quartz after coesite Wain, 1997a,b . Krabbendam

Ž . Ž .and Wain 1997 and Wain 1998 have suggestedthat these eclogite facies schist or gneiss envelopesare metamorphic equivalents of the common gran-

Ž .odioritic basement gneisses. While this may be truein some cases, the phengitic schistose varieties oftenhave pelitic compositions and are associated withquartzite and marble, and the curious zoisite-rich

Ž .gneiss at Otnheim Fig. 3 has a mineralogy consis-tent with an anorthositic composition. These litholo-gies are typical of those designated to cover units.The reason for apparently enhanced survival of

Ž .eclogite facies felsic rocks and metabasic eclogitesin such units is unclear at present, although Krabben-

Ž .dam and Wain 1997 argue they are preserved inlow strain zones that have escaped late-orogenicdeformation and recrystallisation. In addition to theseexamples, phengite relics or replacement inter-growths of biotiteqplagioclase"kyanite are rarely,but widely, found in quartzo-feldspathic gneisses of

Ž .the WGC Griffin, 1987; Wain, 1998 , possibly indi-cating that such HP, non-mafic rocks were oncemuch more common. Considerations of the effi-ciency of hydration vs. dehydration reactions ineclogites and gneisses suggest that eclogites shouldbe less prone to loss of HP parageneses during

Žretrogression than gneisses Heinrich, 1982; Straume,

.1997; Wain, 1998 ; hence, the apparently incompati-ble mineral facies of eclogites and gneisses may be,at least in part, due to differential retrogression.

The ‘‘internal’’ eclogites, which are found aslayers and lenses within the masses of mantle-de-

Ž . Žrived dunite or garnetiferous peridotite Fig. 2 Lap-pin, 1966, 1974; Carswell, 1974; Medaris, 1980;

.Jamtveit, 1984; Griffin and Qvale, 1985 , will beconsidered further in Section 5 below.

Apart from the essential garnet and omphacite,common matrix phases in the ‘‘external’’ eclogitesŽ .those enclosed within schists or gneisses are quartz,rutile, zoisiterclinozoisite, phengite, paragonite,kyanite, orthopyroxene, phlogopitic mica, amphi-boles of a wide range of compositions, carbonatesŽ . Ždolomite, calcite, magnesite , pyrite and apatite fora more exhaustive account of the peculiarities of

.eclogite mineralogy in the WGC, see Smith, 1988 .Orthopyroxene is commonly associated with phlogo-pitic mica while kyanite, phengite and zoisite areabsent from orthopyroxene eclogites. Mineral inclu-sion suites and compositional zoning are common ineclogite matrix phases.

While the mineralogy and fabrics of eclogitesvary considerably, even within single bodies, a num-ber of features show clear geographic distributions.South of the latitude of the southern end of Stadlan-det, and southeast of a line approximately joining the

Ž .inland mouths of the major fjords Figs. 1–3 , eclog-ite garnets are commonly idioblastic and have spes-sartine-rich cores containing amphibolite facies in-clusion suites and eclogite facies inclusion suites

Ž .near their rims Fig. 4 . The change in inclusionassemblage corresponds to a sharp increase in MgrFe

Ž .and is interpreted as representing the epidote- am-phibolite-facies to eclogite-facies transition during

Žprograde growth of garnet Bryhni and Griffin, 1971;Krogh, 1980, 1982; Cuthbert, 1985; Cuthbert and

.Carswell, 1990 . To the north of outer NordfjordŽ .Figs. 1–3 , eclogite garnets are often more xeno-blastic, lack prograde zoning and have marginalretrograde zoning, and have eclogite facies inclu-sions suites similar to matrix parageneses. KroghŽ .1982 noted that the change from prograde-zonedgarnets to unzoned garnets lies approximately alongthe 7008C isotherm as defined by the regional patternof garnet–clinopyroxene Fe–Mg geothermometryŽ .Krogh, 1977 , and attributed the lack of zoning at


Ž .Fig. 4. Photomicrographs showing petrographic features of HP and UHP eclogites. a Prograde-zoned garnet typical of quartz-stable, HPŽ . Ž .eclogites with a dark core containing hornblende Hbl inclusions and ringed by monocrystalline quartz inclusions Qz , and a paler rim

Ž . Ž . Ž .clear of inclusions; from Evja, 3 km SE of Maløy, PPL. b Relict coesite Cs surrounded by replacive quartz pallisade textured PCQ˚Ž . Ž . Ž .within garnet Grt from eclogite at Salta, PPL. c Granular polycrystalline quartz PCQ inclusion considered to have replaced coesite, in

Ž . Ž . Ž . Ž .omphacite Omp , with adjacent garnet Grt in eclogite from Bryggja, Nordfjord, CPL. d Polycrystalline quartz PCQ inclusionsŽ . Ž .considered to have replaced coesite, within garnet Grt also containing inclusions of omphacite Omp ; Fjørtoftvika, Fjørtoft, Nordøyane,

CPL.

higher temperatures to diffusional homogenisation.Ž .However, Wain 1997a; b; 1998 has claimed that

the change in the Nordfjord area from prograde-zonedidioblastic garnets to unzoned xenoblastic garnetscorresponds to the HP–UHP transition rather thansimply to an increase in temperature.

The prograde growth of garnets as defined bycompositional zoning, inclusion suites and matrixassemblages, shows a systematic spatial variationŽBryhni and Griffin, 1971; Krogh, 1982; Krogh and

.Carswell, 1995; Cuthbert, 1985; Bailey, 1989 . InŽ .inner Sunnfjord Fig. 3 , an early epidote amphibo-

lite stage was followed by a glaucophane or bar-roisite eclogite stage. In outer Sunnfjord, an early HPbarroisite-bearing epidote amphibolite stage was fol-

lowed by a barroisite eclogite stage. In outer Nor-Ž .fjord and in Romsdal 40 km SE of Molde; Fig. 1 ,

an early amphibolite or epidote amphibolite stagewas followed by an eclogite stage.

Retrograde mineral assemblages in eclogites andthe predominant parageneses in gneisses appear to bebroadly co-facial, and show a similar pattern toprograde assemblages, with epidote–amphibolite ininner Sunnfjord, epidote amphibolite or amphibolitein outer Sunnfjord, and amphibolite in Nordfjord,Stadlandet and coastal Sunnmøre. In the northernparts of the WGR from Molde to Kristiansund,eclogites tend to have retrograded to garnet gran-

Ž .ulites Krogh, 1981, 1982; Krogh and Carswell, 1995and some eclogite bodies contain trondhjemitic par-


Ž .tial melt seggregations Cuthbert, 1995, 1997 . In thenorthern extension to the WGR northeast of Trond-

Ž .heim Vestranden Gneiss Region or VGR , eclogiteshave not been found, but HP granulites are common

Ž .in basic and felsic lithologies Moller, 1988 . The¨southwest extension of the VGR, rocks along strikewould lie offshore to the NW of Kristiansund andMoldefjord, and could represent the part of the WGCwhere temperatures were highest during retrograde

Ž .HP UHP? metamorphism, and where the efficiencyof overprinting of eclogite facies parageneses hasbeen greatest. Hence, a broad pattern emerges acrossthe WGR which is consistent with lowest tempera-tures in the SE and highest temperatures in the NWduring prograde, peak and retrograde Scandian meta-morphism. Some variations may be superimposed

Ž .upon this pattern; e.g., Wain 1998 suggests that theexhumation path for eclogites in the Nordfjord areais cooler than that for eclogites further north.

Ž .Smith 1984; 1988; 1995 , Krabbendam and WainŽ . Ž .1997 and Wain 1997b; 1998 have identified co-

Ž .esite or polycrystalline quartz PCQ aggregates at-tributable to the breakdown of pre-existing coesite ata total of 24 localities concentrated in the outer

Ž .Nordfjord and Stadlandet area Fig. 3 , but also inŽ . ŽHareid Dimnøy and Hessdalen in Sunnmøre Fig.

.2 . In the present study, we have confirmed all ofthese occurrences and found a further coesite eclog-

Ž .ite locality at Flister Figs. 2 and 3 . Coesite or PCQis found exclusively as inclusions in other phases,including garnet, omphacite, kyanite and zoisite, andhas been found in both eclogites and, rarely, gneissesŽ .Krabbendam and Wain, 1997; Wain, 1997a, 1998 .We have also found eight further localities with PCQ

Ž .at Maurstad and Bryggja Figs. 3–5 , Nordfjord:north Grytting, Djupedalsvatnet and SandviknesetŽ . ŽStadlandet ; Hornet and Brautene south of

. ŽAlmklovdalen ; Stigedalen south of Bjørkedalsvat-. Ž . Ž .net , and Fjørtoft island Nordøyane Figs. 2–5 .

The HornetrBrautene and Stigedal localities extendthe known occurrences of probable coesite eclogites40 km eastwards from previous localities. The Hor-net eclogite is a large, tabular mass 4.5 km long and0.5 km wide in outcrop lying 1.5 km south of the

Ž .Almklovdalen peridotite massif Fig. 3 . Evidencefor coesite is in the form of granular polycrystallinequartz included within garnet. In Stigedalen, eclogiteoccurs as meter scale pods within mica schists and

Ž .quartzites cover unit 1–2 km south of the BjørkedalŽ .anorthositergarnet peridotite massif Fig. 2 . Evi-

dence for coesite is pallisade-textured PCQ includedwithin garnet. Eclogite with prograde-zoned garnetsand no evidence for coesite is also found in Stigedalwithin this metasedimentary cover sequence. AtFjørtoftvika on the north coast of the island of

Ž .Fjørtoft, Nordøyane Figs. 2 and 4 , a kyanite eclog-ite with optically unzoned garnets containing om-phacite and phengite inclusions also has granularpolycrystalline quartz included in garnet. This ex-tends the range of UHP eclogites 55 km northwest ofthe previous northernmost occurrence on Dimnøy,

Ž .Hareidlandet Smith, 1988 . The Fjørtoftvika eclog-Žite occurs in a basement unit Ulla Gneiss; Terry and

.Robinson, 1998 . Polycrystalline quartz has also beenreported recently from a kyanite eclogite in a cover

Ž . Žunit Blahø nappe on Fjørtoft at Nytun Terry et al.,˚.1999 . Interestingly, this cover unit also contains the

microdiamond-yielding gneiss reported by Do-Ž .brzhinetskya et al. 1995 . The contact of this unit

with the underlying basement gneisses is marked bya zone of mylonitic eclogite-facies gneisses contain-

Žing bodies of garnetiferous peridotite Terry and.Robinson, 1998 . Thus, these new discoveries signif-

icantly increase the size of the coesite–eclogite ter-rane in the WGR, confirming the suggestion of

Ž .Smith 1995 that his Coesite Eclogite Province ex-tends along the coast to Fjørtoft. There appears to bea close connection between coesite eclogites andgarnetiferous peridotites, and with the previouslyenigmatic discovery of microdiamond.

Ž . ŽKrabbendam and Wain 1997 and Wain 1997a;.b; 1998 have shown that the boundary between the

HP and UHP terranes on the north side of NordfjordŽ .Figs. 3 and 5 is not simple, but is marked by a;10 km wide ‘‘mixed zone’’ within which both HPand UHP eclogites are found, sometimes -0.5 kmapart from each other. They mapped the mixed zoneas a SW–NE-trending strip to the east of Nordpollenwhich swings westward across Sørpollen. Based oncurrently known localities where evidence for coesiteis well-established, the UHP terrane underlies all ofStadlandet and extends into Almklovdalen. Our moreextensive survey shows that the southern boundaryof the mixed zone probably trends E–W parallel toinner Nordfjord and includes the Stigedalen eclog-ites. The northern boundary lies between Vagsøy and˚


Fig. 5. Lithological and thermobarometric map of the Kroken–Maurstad area, Nordfjord showing the distribution of HP and UHP eclogites.Ž .Numbers in boxes are calculated P–T conditions using the Krogh-Ravna 1999 garnet–clinopyroxene Fe–Mg thermometer combined with

Ž .the Waters and Martin 1983 garnet–clinopyroxene–phengite geobarometer using the garnet activity model of Newton and HaseltonŽ . Ž . Ž .1981 . PCQspolycrystalline quartz after coesite UHP eclogite . QPZsquartz eclogite with prograde zoned garnets HP eclogite .

Ž . Ž .Outcrop of eclogite, eclogite facies gneiss and dunite from Bryhni et al. 1969 , Krabbendam and Wain 1997 and the authors’ unpublisheddata.

Stadlandet in the west, but is less certain in the eastdue to a paucity of data. The new finding of PCQ at

Ž .Hornet south of Almklovdalen Fig. 3 must liewithin the mixed zone or UHP zone; until better dataemerge, we tentatively suggest that the presence ofprograde-zoned garnets in an internal eclogite in the

ŽAlmklovdalen peridotite massif Griffin and Qvale,.1985 and north of the Hornet eclogite places the

northern boundary of the mixed zone to the north ofAlmklovdalen in a similar position to that of

Ž .Krabbendam and Wain 1997 . This would be con-sistent with the trace of the gneissic foliation. Thereare no good data to fix the northern boundary of themixed zone further east at this stage. Scattered locali-

Ž .ties in outer Sunnmøre Dimnøy, Hessdalen andŽ .Nordøyane Fjørtoftvika and Nytun allow the UHP

terrane to be tentatively extended in a strip up to 40km wide along the coast between Stadlandet andFjørtoft, so that the total area underlain by the UHPterrane may be tentatively estimated at about 5000

km2. The existence and positions of the mixed zoneand the UHP terrane are poorly constrained to theeast and northeast of Stigedalen, and inner Sunnmøre.However, the association of UHP eclogites withgarnetiferous peridotites leads us to speculate thatthe UHP terrane may extend as far eastwards asTafjord, so it could conceivably cover an area of up

2 Ž .to 8000 km . The Romsdal eclogite Krogh, 1982has prograde-zoned garnet, which may constrain theeasternmost limit of the UHP terrane in the northernpart of the WGR. The total area underlain by UHPeclogites is, on current evidence, considerably smaller

Ž .than that suggested by Coleman and Wang 1995whose estimate of 52,500 km2 encompassed theentire WGR, including the recognisable HP eclogiteterrane south of Nordfjord.

The present database does not reveal a clearrelationship between the distribution of petrographictypes of eclogite and lithotectonic units. Cover rocksŽ .quartzites and mica schists in Stigedalen contain


both HP and UHP eclogites, and both types are alsofound within basement lithologies. In Nordøyane,UHP eclogite and gneiss occur both in a cover unitŽ .Blahø nappe; Terry et al., 1999 and in the underly-˚

Ž .ing basement gneisses Ulla Gneiss . However, theSætra and Risberget cover nappes, outcropping alongnearby Moldefjord and lying structurally between theBlahø nappe and basement, contain eclogites with˚prograde-zoned garnets and no evidence for coesiteŽ .Robinson, 1995 .

Ž .Within the mixed zone, Wain 1998 consideredthat individual eclogite bodies were internally uni-form in petrographic type and that HP and UHPtypes were not found in the same body, and sheobserved that eclogites of the same type appeared tolie along discrete horizons in the gneisses. Theseobservations suggest that, even in the mixed zone,HP and UHP eclogite bodies have distinct and se-parate origins. However, typical prograde-zonedgarnets with hornblende inclusions are found in

˚the Arsheimneset polycrystalline quartz-bearing,Žorthopyroxene eclogite Lappin and Smith, 1978;

.Carswell et al., 1985; Smith, 1988 . At Vetrhus,Nordpollen, prograde-zoned garnets and polycrys-talline or multicrystalline quartz inclusions are foundwithin eclogite and the enveloping pelitic schists.These examples suggest that an early prograde his-tory may, rarely, be preserved in rocks which havewitnessed UHP conditions.

( )4. Distribution of pressure P and temperature( )T

Early attempts to determine the distribution of Tacross the WGR used garnet–clinopyroxene Fe–Mgexchange geothermometry, and defined a regionalincrease towards the NW with isotherms trending

ŽNE–SW parallel to the coast north of Maløy Krogh,˚.1977; Griffin et al., 1985 . The pattern was not

well-defined further south between Førdefjord andNordfjord due to the large area covered by the

Ž .allochthon and Devonian basins Fig. 2 . Placementof isotherms was necessarily highly generalised dueto uncertainties in T estimates and scarcity of dataaway from the coast. Estimation of P was hamperedby the lack of geobarometers available for the com-

Ž .mon Opx-free eclogites. Griffin et al. 1985 esti-mated minimum pressures based on the assumed

coexistence of omphacite with albite in eclogites andusing the omphacite activity model of HollandŽ .1980 . The resulting apparent trend of increasing Ptowards the NW was largely an artifact of the d PrdTslope of the employed barometer, although the cur-rently known distribution of coesite clearly indicateshigher pressures in the north of the WGR than in thesouth.

New geobarometers are now available for thedetermination of absolute pressures in eclogites, andprovide an opportunity to obtain a much more reli-able evaluation of trends in pressure across the WGR.

Ž .Wain 1997b; 1998 used the garnet–omphacite–Žphengite barometer Waters and Martin, 1983 up-

dated 1996; see http:rrwww.earth.ox.ac.ukr;.davewarresearchrecbar.html and found a bimodal

distribution of P and T in the Stadlandet — outerNordfjord area — with a significant pressure gap of)4 kbar between the two petrographically definedgroups of eclogites, and calculated pressures whichare broadly consistent with the type of silica poly-morph present. Pressures for UHP eclogites mostlylay close to, and just above, the coesite–quartz equi-

Ž .librium. This was taken by Wain 1997b; 1998 andŽ .Krabbendam and Wain 1997 to indicate a tectonic

break between the HP and UHP terranes, with themixed zone being a zone of interfolding or imbrica-tion of the two terranes.

We have established a more geographically ex-Ž .tended database of P–T determinations Table 1 ,Ž .concentrated on the area studied by Wain op. cit.

but also including samples from Sunnfjord andSunnmøre. Mineral phases from the Kvineset eclog-

Ž .ite Table 1 were analysed on the ARL-EMx elec-tron microprobe at the Central Institute for IndustrialResearch, Oslo with WDS with matrix corrections

Ž .according to Bence and Albee 1969 , or on theARL-EMX microprobe at the Mineralogy–GeologyMuseum, Oslo by EDS with matrix corrections cal-

Ž .culated by the program ZAF-4 of Statham 1976 ,both with running conditions of 15 kV, 10 nA. Theremaining samples were analysed on either theCameca Camebax microprobe at the Mineralogy–Geology Museum, Oslo using WDS at 15 kV and 20nA with matrix corrections using the Cameca PAPprogram, or the JEOL840 SEM at the University ofTromsø using EDS at 20 kV, 3 nA with matrixcorrections by ZAF.


Eclogites with well-equilibrated textures and lack-ing extensive retrogression or abundant pre-eclogiticrelics were selected for analysis wherever possible.Where garnets exhibit prograde zoning, near-rimcompositions were used. T and P were calculatedusing a combination of the updated Grt–Cpx Fe–Mg

Ž .exchange thermometer of Krogh-Ravna 1999 , andthe Grt–Cpx–Phe barometer of Waters and MartinŽ .1983 . Ferric iron in Grt and Cpx were estimated by

Ž .the charge balance method of Droop 1987 . For thephengite barometer, the procedures recommended by

Ž .Carswell et al. 1997 were used and calculationsmade with the garnet activity models of BermanŽ . Ž .1990 and Newton and Haselton 1981 .

In Fig. 6A and B, results are plotted for twochoices of phase compositions. For the favouredcombination of maximum a =a2 in garnet,pyr gros

maximum jadeite in omphacite and maximum Si inŽ .phengite ‘‘max gros’’ dataset in Table 1 , P–T

Ž . Žpoints for UHP coesite stable and HP quartz sta-.ble eclogites are generally consistent with the posi-

tion of the CssQz equilibrium within the uncer-Ž .tainty of the thermobarometers. The Berman 1990

Ž .garnet activity model Fig. 6A gives slightly higherpressures, with all UHP eclogites lying on or abovethe CssQz equilibrium, but some HP eclogites lieon the HP side of the CssQz equilibrium. The

Ž .Newton and Haselton 1981 garnet activity modelŽ .gives pressures ;2 kbar lower Fig. 6B and all the

HP eclogites lie in the quartz field, but two UHPeclogites, from Hornet in Almklovdalen, and Fla-traket Harbour also lie in the quartz field. For the

Ž .Flatraket Harbour eclogite, Wain 1998 calculated ahigher pressure in the coesite field for this body, andour low result may possibly be again due to retro-grade re-equilibration of phengite to lower Si com-positions, as perhaps is also be the case for theHornet eclogite. In general, the Newton and Haseltongarnet activity model seems to give results moreconsistent with the observed type of silica poly-morph for this data set, and indeed seems to fit betterwith experimental data and well-calibrated natural

Žexamples D. Waters, 1999, personal communica-.tion .

Ž .Wain 1997b; 1998 calculated a similar P–Tdistribution to the ‘‘max gros’’ dataset for the Nord-fjord–Stadlandet area, using a different set of HPand UHP samples and using average compositions of

Ž .omphacite and garnet rims for HP garnets , andhighest silica phengites. The similarity of resultsfrom these two different sample sets and differentapproaches to selecting phase compositions strength-ens confidence in our conclusions.

Selecting the alternative combination of ma-ximum pyrope in garnet, minimum MgrFe in

Ž .omphacite and minimum MgrFe and low Si inŽ .phengite ‘‘max py’’ dataset in Table 1 , nearly all

Ž .samples fall in the quartz field Fig. 6A and B , atlower pressures than for the ‘‘max gros’’ set. Theselower apparent pressures may be largely controlledby the retrograde decrease of Si in phengite and it isunlikely that this choice of phase compositions wasin equilibrium at peak P.

Temperatures vary widely within both data sets,probably partly due to uncertainties in ferric ironestimation done by charge balance. The UHP groupforms a fairly distinct cluster, except for the Hol-

Ž .mane eclogite. Wain 1997b also calculated a lowerŽ .T for the Holmane body when the Droop 1987

ferric iron calculation was applied, but when calcu-lated on the basis of total Fe as ferrous, it gave aresult more consistent with the rest of its group.

We have selected the ‘‘max gros’’ dataset usingthe Newton and Haselton garnet activity model toshow the relationships between P, T and the petro-

Ž .graphic and geographic groupings Fig. 6C . A gen-eral trend emerges in which P and T increase fromsoutheast to northwest, with lowest values for theglaucophane eclogites of inner Sunnfjord and highestvalues for the coesite eclogites of outer Nordfjord,Stadlandet and Fjørtoft. Values for the mixedHPrUHP zone overlap to some extent with thoseoutside the mixed zone, but it is perhaps significantthat all but two HP eclogite samples from the mixedzone cluster at slightly higher P than those in the HPzone to the south. A single UHP sample from the

Ž .Blahø nappe on Fjørtoft Terry et al., 1999 , close to˚the diamond gneiss, lacks phengite in the matrix, incommon with many eclogites in the northern WGR,but inclusions of omphacite and phengite are presentin garnet. Phengite inclusions are retrograded tobiotiteqplagioclase symplectite at margins, possiblyresulting in loss of Si; hence, the calculated pressure

Ž .of 7648C at 28 kbar Terry et al., 1999 is likely tobe a minimum value. The isopleth for the Grt–Cpx

Ž .thermometer Fig. 6C intersects the quartzscoesite


Table 1Sample locations, P and T calculations and parageneses of selected eclogites from the WGR

Ž . Ž .GtCp — Grt–Cpx Fe–Mg thermometer of Krogh-Ravna 1999 ; Pphe Berm — Grt–Cpx–Phe barometer of Waters and Martin 1983Ž . Ž .using the garnet activity model of Berman 1990 ; Pphe N and H — Grt–Cpx–Phe barometer of Waters and Martin 1983 using the garnet

Ž .activity model of Newton and Haselton 1981 ; Phe — phengite present if marked x; additional phases — kyskyanite, glsglaucophane,Žbarsbarroisite, hblshornblende, czosclinozoisite, zoszoisite, dolsdolomite, parsparagonite; PCQ — polycrystalline quartz after

.coesite present; Cs — coesite present. Values in italics for the phengite-free matrix of the Fjørtoft sample give T for the Grt–Cpx Fe–MgŽ . Ž .thermometer of Krogh-Ravna 1999 at a nominal pressure of 25 kbar. Question marks in UTM map grid reference column indicates some

uncertainty over location of samples collected by other workers.

Sample Place name UTM co- Map T Pphe T Pphe Phe Other PCQrordinates sheet GtCp Berm GtCp N and H phases Cs

Inner Sunnfjord HP eclogitesMax gros193 Kvineset 227,292 1217 IV 573 18.7 561 16.4 x Gl –194 Kvineset 227,292 1217 IV 528 18.0 517 15.6 x Gl –196 Kvineset 227,292 1217 IV 569 17.9 558 15.6 x Bar –85 Kvineset 227,292 1217 IV 543 19.3 533 17.0 x Bar –56 Kvineset 227,292 1217 IV 483 19.7 472 17.0 x Bar –Max pyr193 Kvineset 227,292 1217 IV 540 16.3 528 14.0 x Gl –194 Kvineset 227,292 1217 IV 522 17.1 511 14.7 x Gl –196 Kvineset 227,292 1217 IV 599 18.2 588 15.9 x Bar –85 Kvineset 227,292 1217 IV 515 17.1 504 14.7 x Bar –56 Kvineset 227,292 1217 IV 540 16.2 528 14.0 x Bar –

Outer Sunnfjord HP eclogitesMax grosIB-1C Sellevoll ? 1117 II 590 23.2 581 21.2 x Ky Bar –

˚IB-2C N Asen ? ? 589 22.2 580 20.0 x Ky Bar –Max pyrIB-1R Sellevoll ? 1117 II 611 21.2 598 18.9 x Ky Bar –

˚IB-2R N Asen ? ? 604 21.8 594 19.6 x Ky Bar –

Nordfjord HP eclogitesMax grosIB-7C Davik 182,684 1218 IV 644 25.3 636 23.4 x Ky Hbl –UHPM-33 Skavøypollen 972,710 1118 I 575 22.9 566 20.8 x Ky Czo –UHPM-36 Halnes 946,719 1118 IV 738 25.2 728 23.3 x Ky Zo Hbl –UHPM-59 Biskjelneset 995,686 1118 IV 607 22.8 598 20.9 x Ky Czo –45r97 Høgefossen 480,696 1218 I 573 25.6 563 23.4 x Ky Hbl –UHPM-24 Levdal 167,715 1218 IV 655 24.1 646 22.2 x Ky Dol –Max pyrIB-7R Davik 182,684 1218 IV 702 23.2 690 21.1 x Ky Hbl –UHPM-33 Skavøypollen 972,710 1118 I 677 22.5 666 20.4 x Ky –UHPM-36 Halnes 946,719 1118 IV 687 21.1 673 18.7 x Ky –UHPM-59 Biskjelneset 995,686 1118 IV 710 20.3 699 18.4 x Ky –UHPM-24 Levdal 167,715 1218 IV 542 22.1 531 19.8 x Ky –45r97 Høgefossen 480,696 1218 I 537 24.8 527 22.6 x Ky –

Nordfjord–Stadlandet mixed zone HP eclogitesMax grosUHPM-51 S Raudeberg 977,754 1118 I 736 27.0 725 25.0 x Ky Zo ?UHPM-57 N. Vagsøy 984,787 1118 I 785 27.9 773 25.7 x Ky –˚UHPM-46 Maløy N 969,740 1118 I 760 28.3 748 26.2 x Ky –˚UHPM-43 N Oppedal 2 940,772 1118 IV 705 27.4 694 25.3 x Ky Czo –UHPM-42 N Oppedal 1 938,768 1118 IV 583 25.4 574 23.4 x Ky? –


Ž .Table 1 continued

Sample Place name UTM co- Map T Pphe T Pphe Phe Other PCQrordinates sheet GtCp Berm GtCp N and H phases Cs

Nordfjord–Stadlandet mixed zone HP eclogitesMax gros

IB-20 N Oppedal 938,768 1118 IV – Ky par –84r97 Flatrakt Bekk 28,765 1118 I 630 25.0 619 23.4 x Czo –89r97 Maløy Fjel 950,727 1118 I 598 26.7 587 24.5 x – –˚UHPM-10 Maløy South 957,723? 1118 IV 606 26.4 597 24.4 x Ky –˚TA-1 Kroken 080,704 1118 I 616 26.5 607 24.4 x Hbl dol –KROKEN Kroken 080,704 1118 I 602 26.0 593 24.1 x Hbl dol –UHPM-27 Almenning 029,700 1118 I 611 27.1 601 25.0 x Hbl –ALM-1 Almenning 029,700 1118 I 629 28.7 616 26.1 x Hbl –Max pyr

UHPM-51 S Raudeberg 977,754 1118 I 706 23.8 695 21.9 x Ky Zo ?UHPM-57 N. Vagsøy 984,787 1118 I 762 23.0 750 20.9 x Ky –˚UHPM-46 Maløy N 969,740 1118 I 676 24.0 665 21.9 x Ky –˚UHPM-43 N Oppedal 2 940,772 1118 IV 589 24.9 579 22.9 x Ky Czo –UHPM-42 N Oppedal 1 938,768 1118 IV 618 22.7 609 20.7 x Ky? –IB-20 N Oppedal 938,768 1118 IV – Ky par –84r97 Flatrak Bekk 28,765 1118 I 680 19.9 668 17.6 x Czo –89r97 Maløy Fjel 950,727 1118 I 586 24.3 575 22.1 x – –˚UHPM-10 Maløy South 957,723? 1118 IV 713 22.1 703 20.2 x Ky –˚TA-1 Kroken 080,704 1118 I 601 24.3 591 22.1 x Hbl dol –KROKEN Kroken 080,704 1118 I 602 20.6 592 18.4 x Hbl dol –UHPM-27 Almenning 029,700 1118 I 601 25.6 591 23.5 x Hbl –ALM-1 Almenning 029,700 1118 I 547 22.7 533 19.9 x Hbl –

Nordfjord–Stadlandet mixed zone UHP eclogitesMax gros

UHPM-6 Bryggja 136,728 1118 I 739 31.5 730 29.8 x Ky? Zo PCQUHPM-13 Maurstad 143,728 1118 I 706 30.4 694 28.2 x Ky Zo PCQUHPM-70 Totland 096,713 1118 I 769 31.0 759 29.2 x Ky Czo PCQUHPM-76 Holmane 103,712 1118 I 594 28.6 585 26.5 x Ky? Czo PCQF-13A Flatraket 026,775 1118 I 718 27.4 700 24.4 x Ky CsUHPM-60 Hornet 247,777 1218 IV 630 26.7 621 24.8 x Ky? PCQMax pyr

UHPM-6 Bryggja 136,728 1118 I 731 30.0 722 28.3 x Ky? Zo PCQUHPM-13 Maurstad 143,728 1118 I 671 25.1 661 23.2 x Ky Zo PCQUHPM-70 Totland 096,713 1118 I 765 27.0 756 25.3 x Ky Czo PCQUHPM-76 Holmane 103,712 1118 I 799 25.8 787 23.8 x Ky? Czo PCQF-13A Flatraket 026,775 1118 I 654 22.4 636 19.3 x Ky CsUHPM-60 Hornet 247,777 1218 IV 665 25.1 654 23.1 x Ky? PCQ

Nordfjord–Stadlandet UHP eclogiteMax gros

3r97 Drage 015,914 1019 II 768 31.8 758 29.9 x Ky Zo xMax py

3r97 Drage 015,914 1019 II 742 26.9 732 25.0 x Ky Zo x

Hareidlandet and Nordøyane UHP eclogitesMax gros

Terry et al., Fjortoft 659,561 1220 III 748 25 x Ky x1999


equilibrium at 7608C, 26.5 kbar and the graphitesdiamond equilibrium at 8008C, 33.5 kbar, the formergiving a minimum pressure for this body and thelatter being consistent with the presence of diamondin nearby gneisses.

In order to identify any real trends in P and Tacross the WGR, it is necessary to evaluate the likelyuncertainties in the P and T estimates. Uncertaintiesdue to thermobarometer calibration and microprobeanalysis are likely to be about "508C and "1.5kbar for garnet q omphacite q phengite thermo-

Ž .barometry Wain, 1998 . Related uncertainty due toferric iron calculation, especially in omphacite, mayresult in even larger errors. Furthermore, an addi-tional source of uncertainty arises due to progradeand retrograde zonation of mineral phases and thedifficulties in selecting phase compositions whichwere in equilibrium at the peak metamorphic condi-tions recorded by individual specimens. We haveattempted to evaluate the overall uncertainty due toall these factors by inspection of our ‘‘max gros’’dataset. Breaking the dataset into its petrographicand geographic subsets, we have found the range of

Žvalues recorded as "1 standard deviation from the.mean for subsets with a reasonably large sample

size. Unfortunately, this introduces real geographicalvariation in P and T as an additional source of errorbut, in the absence of any detailed evaluation ofuncertainties from individual eclogite bodies, we re-gard our selection of sample subsets as the bestcompromise possible with the data currently avail-able. Averages and uncertainties for the sample sub-sets are plotted in Fig. 6D. Uncertainty in T is likelyto be about "758C. Uncertainty in pressure could beup to "2 kbar, although the spread of P is smaller

Žin the HP eclogites than the UHP eclogites largely.based on samples from outer Nordfjord . The larger

spread of data for UHP eclogites could be reduced tothat for HP eclogites if those plotting below the

ŽQzsCs equilibrium are ignored on the basis thatthey have suffered retrogressive re-adjustment of

.phengite composition .These large uncertainties allow only gross gener-

alisations to be made about regional trends. Never-

Fig. 6. P – T plots of thermobarometry results from eclogites inthe WGR. See Table 1 for data, locations and literature sources.Ž . Ž .A Krogh-Ravna 1999 garnet–clinopyroxene Fe–Mg exchangethermometer combined with the garnet–clinopyroxene–phengite

Ž .barometer of Waters and Martin 1983 , using the garnet activityŽ .model of Berman 1990 . Circles — HP eclogites; squares —

UHP eclogites; filled symbols — ‘‘max gros’’ dataset; openŽsymbols — ‘‘max pyr’’ dataset see text for explanation of

. Ž . Ž . Ž .datasets . B As for A , using Newton and Haselton 1981Ž .garnet activity model for garnet. C Relationships of P and T

with petrographic type of eclogite and geographic area using theŽ .Krogh-Ravna 1999 garnet–clinopyroxene Fe–Mg exchange

Ž .thermometer combined with Waters and Martin 1983 garnet–clinopyroxene–phengite barometer, using the ‘‘max gros’’ datasetonly. Diamonds — glaucophane eclogites from Inner Sunnfjord;triangles — Outer Sunnfjord HP eclogites; circles — Nordfjord–Stad HP eclogites; squares — Nordfjord–Stad UHP eclogites.Grey symbols are from the mixed HPrUHP zone. Solid lines are

Ž . Ž . Ž .for the graphite Gph sdiamond Dm and quartz Qz scoesiteŽ . Ž .Cs equilibria Bundy, 1980; Bohlen and Boettcher, 1982 . Dashed

Ž .line is the Krogh-Ravna 1999 grt–cpx Fe–Mg exchangeŽ .geothermometer isopleth for the sample from Fjortøft. D Aver-

age P and T of petrographic and geographical groupings, withŽ .symbols as in plot C . Error bars are "1 sd. Dash–dot line is a

reference ‘‘geotherm’’ of 58Crkm. Error brackets with bar termi-nations are calculated from the dataset. Error brackets withoutterminating bars are for small sample sets where the applieduncertainty is taken from a larger sample set from an adjacentarea.


theless, by inspection of Fig. 6D, it is clear thatrecorded temperatures in the northwestern, coastalUHP terrane of the WGR are significantly higherthan those from eclogites in the HP terrane in Sunn-fjord and Nordfjord. Pressures show a progressiveregional increase from southeast to northwest from15 kbar in inner Sunnfjord to about 30 kbar inStadlandet.

The eclogites from the mixed HPrUHP zoneŽshow a large overlap in temperature. Wain 1997a; b;

.1998 , using a different sample set, found that pres-sures and temperatures from eclogites in the mixedzone clustered in a bimodal distribution with a Pdifference of )4 kbar between the HP and UHPtypes. HP and UHP eclogites lying only a few

Ž .hundred meters apart were found by Wain op. cit.to give pressures differing by up to 8 kbar. Applyingthe rather cautious error brackets discussed aboveŽ ."2 kbar , any calculated pressure difference of 4kbar or less could not be considered to be signifi-cantly different. Inspection of Fig. 5, which uses theP–T data set from this study, confirms the largepressure differences between adjacent eclogite bod-ies; for instance, the Kroken HP eclogite gives apressure 6 kbar lower than the Totland UHP eclogiteless than 2.5 km map distance away or in fact about1 km across strike. The Maurstad UHP eclogite givesa pressure 6 kbar higher than the Levdal eclogite,

Ž .also less than 2.5 km away 2 km across strike .Even within the bounds of uncertainty defined here,these variations in pressure in the mixed zone seemto be real in some cases. As the apparent pressuregradients are much larger than the normal lithostaticpressure gradient across likely structural distancesbetween these eclogites, we must infer either signifi-cant crustal shortening between them, or equilibra-tion of adjacent eclogites at different times on aP–T–t path.

Average P and T for each petrographic grouparranged by geographic area lie along a linear trendŽ .Fig. 6D , with a slope of approximately 58Crkm.Mixed HPrUHP zone eclogites straddle the CssQzequilibrium line, and the group averages lie along thetrend in geographic order from south to north withincreased P and T. Interestingly, the Grt–Cpx ther-mometer isopleth for the Fjørtoft sample intersectsthe diamondsgraphite line on the HP extrapolationof the trend, suggesting conditions consistent with its

extreme northwesterly position in the WGR. Deter-mination of the actual pressure recorded by thissample awaits further developments in geobarome-try, so the significance of the Fjørtoft rocks remainstantalisingly uncertain. The slope and order of thetrend are consistent with the hypothesis of equilibra-tion at steadily increasing depths along a geothermalgradient similar to that of a subduction zone, al-though such an interpretation must be made withcaution due to the likely transient thermal structureof a continental collision zone and the possibilitythat eclogite formation was diachronous, or evenpolycyclic, across the WGC. We also reiterate thepoint that a wide spread of P–T results within somegroups has been averaged, within which some of thevariations may be real.

5. Orthopyroxene external eclogites and garnetif-erous peridotites

Orthopyroxene–clinopyroxene–garnet assem-blages suitable for quantitative thermobarometric

Ž .evaluation see review of Carswell and Harley, 1990are found in ‘‘external’’ Opx eclogites enclosedwithin gneisses, as well as in ‘‘internal’’ eclogitesenclosed within the garnetiferous peridotite bodiesŽsee review of Krogh and Carswell, 1995 and refer-

.ences therein . ‘‘External’’ Opx eclogites are foundŽexclusively from Nordfjord northwards Carswell et

.al., 1995 . Some are petrographically UHP eclogites˚Žwith PCQ e.g., Arsheimneset, Sandvikneset,

.Otnheim, Hessdalen and occur within our desig-nated UHP terrane. Most lack free silica and all lackphengite and kyanite.

Several cation exchange and net transfer geother-mometers and barometers are available for use withOpx eclogites, offering an alternative to phengite-based thermobarometry. However, previous esti-mates of absolute P from Opx eclogites based on Alin Opx in equilibrium with garnet varied widelydepending on interpretation of Al zoning patterns in

Ž .Opx. For example, Lappin and Smith 1978 esti-mated P–T conditions of 30–45 kbar at 700–8508Cfor external Opx eclogites, assuming equilibriumbetween garnet and low Al Opx cores. Lappin andSmith’s samples all came from the UHP terrane asdefined above and certainly give P–T conditionsconsistent with UHP metamorphism. In contrast,


Table 2Calculated P–T conditions for garnetiferous peridotites and orthopyroxene external eclogites in the WGR using analyses from previouslypublished and unpublished sourcesSee text for details of calculations.

Ž . Ž .Sample T 8C P kbar Locality Data source

MgCr peridotites — coresrporphyroclastsES-4 898 42.3 Flemsøy, Nordoyane Mørk, unpublishedF12 719 21.7 Fjørtoft Nordoyane Jamtveit et al., 1991U95 661 27.7 Ugelvik, Otrøy Carswell, 1986U92 724 25.6 Ugelvik, Otrøy Carswell, 1986U8 739 22.8 Raudhaugane, Otrøy Carswell, 1986UCS-7A 708 27.9 Raudhaugane Otrøy Medaris, 1984U539 697 26.9 Raudhaugane Otrøy Jamtveit et al., 1991119-111 777 30.3 Sandvika Gurskøy Medaris, 1984Sandvika 2 637 21.1 Sandvika Gurskøy Krogh-Ravna, upublished dataJAMT 947 43.8 Gurskebotn, Gurskøy Jamtveit, 198496-3-26A 712 29.8 Aldalen Medaris, 1984NT-3 870 40.0 Kalskaret, Tafjord Medaris, 1984T96 882 38.8 Kalskaret Tafjord Jamtveit et al., 1991T153 832 35.5 Kalskaret Tafjord Jamtveit et al., 1991SU-1K 781 29.8 Almklovdalen Medaris, 19847174 760 26.9 Bjørkedalen Brastad, 19857998 786 24.4 Bjørkedalen Brastad, 1985

MgCr peridotites — rims or neoblastsES-4 828 29.7 Flemsøy, Nordoyane Mørk, unpublishedF12 719 21.7 Fjørtoft Nordoyane Jamtveit et al., 1991U92 662 16.1 Ugelvik, Otrøy Carswell, 1986U8 736 26.1 Raudhaugane, Otrøy Carswell, 1986U8 740 30.2 Raudhaugane, Otrøy Carswell, 1986U8 721 29.6 Raudhaugane, Otrøy Carswell, 1986UCS-7A 639 27.6 Raudhaugane, Otrøy Medaris, 1984U539 685 18.7 Raudhaugane, Otrøy Jamtveit et al., 1991JAMT 659 20.8 Gurskebotn, Gurskøy Jamtveit, 198496-3-26A 759 31.1 Aldalen Medaris, 1984NT-3 790 30.2 Kalskaret Tafjord Medaris, 1984T96 648 15.1 Kalskaret Tafjord Jamtveit et al., 1991T153 685 22.0 Kalskaret Tafjord Jamtveit et al., 1991SU-1K 669 25.4 Almklovdalen Medaris, 19847998 785 24.3 Bjørkedalen Brastad, 1985

FeTi peridotites — coresH185 765 26.0 Kolmannskog, Moldefjord Carswell et al., 1983U125 675 21.0 Raknestangen, Midøy Carswell et al., 1983E4 675 32.0 Eiksundal, Hareidlandet Carswell et al., 1983

External opx eclogites — cores or unzoned grainsU 206 733 22.5 Solholm, Otrøy Carswell et al., 1985U 243 695 22.3 Solholm, Otrøy Carswell et al., 1985U 19 727 19.6 Skarsh, Otrøy Carswell et al., 1985KR 4 734 23.0 Kvalvag Carswell et al., 1985˚KR-3 760 25.8 Hjørungavag Carswell et al., 1985˚D171 773 29.3 Liset Stadlandet Lappin and Smith, 1978D171A 713 35.2 Liset Stadlandet Lappin and Smith, 1978KR 5 676 25.1 Grytting Stadlandet Carswell et al., 1985A44 688 30.7 Grytting Stadlandet Carswell et al., 1985IB-21A 750 33.4 Grytting Stadlandet Krogh-Ravna, upublished data


Ž .Table 2 continued

Ž . Ž .Sample T 8C P kbar Locality Data source

External opx eclogites — cores or unzoned grainsIB-21B 752 30.8 Grytting Stadlandet Krogh-Ravna, upublished dataIB-21C 750 34.2 Grytting Stadlandet Krogh-Ravna, upublished data8G 776 34.7 Grytting Stadlandet Lappin and Smith, 197878r19 734 35.5 Grytting Stadlandet Lappin and Smith, 1978KR 6 677 29.7 Hellesylt Carswell et al., 1985V80r19 660 22.9 Høydalsneset Carswell et al., 1985C414 774 39.1 Nybø, Sørpollen Lappin and Smith, 1978C411A 745 37.9 Nybø, Sørpollen Lappin and Smith, 1978C411H 820 47.6 Nybø, Sørpollen Lappin and Smith, 1978C411C 740 37.7 Nybø, Sørpollen Lappin and Smith, 1978Ly 707 33.5 Lyngenes, Sørpollen Lappin and Smith, 1978KR 1 612 15.7 Hornindal Carswell et al., 1985KR-2A 509 19.5 Kvalneset, Totland Carswell et al., 1985KR-2B 501 18.0 Kvalneset, Totland Carswell et al., 1985127M 683 30.3 ? Lappin and Smith, 197868.87.1 787 31.0 ? Lappin and Smith, 1978

External opx eclogites — rims or neoblastsUHPM 7r97 645 21.7 Otnheimneset, Stadlandet Krogh-Ravna, unpublishedKR 5 694 22.3 Grytting Stadlandet Carswell et al., 1985A44 639 20.5 Grytting Stadlandet Carswell et al., 1985IB-21A 624 18.0 Grytting Stadlandet Krogh-Ravna, unpublishedIB-21B 576 16.0 Grytting Stadlandet Krogh-Ravna, unpublishedIB-21C 680 26.5 Grytting Stadlandet Krogh-Ravna, unpublishedKR 6 616 18.5 Hellesylt Carswell et al., 1985KR-2A 607 19.1 Kvalneset, Totland Carswell et al., 1985KR-2B 589 15.2 Kvalneset, Totland Carswell et al., 1985

Ž .Carswell et al. 1985 calculated much lower P–Tconditions of around 700–7508C and 17–18 kbarusing the higher Al rims of Opx grains. It seemsclear now that the high Al rims on Opx are a resultof retrograde growth of amphibole and that higherpressures are more likely in some cases. Neverthe-less, problems with combination of appropriate graincompositions in what are often highly retrogradedeclogites remain a serious problem.

Peridotites and serpentinites occur throughout theŽ . Ž .WGR Fig. 2 . Bryhni 1966 noted that peridotite

bodies occur in linear belts or swarms across theNordfjord area. One of these swarms, stretchingfrom Vagsøy through Sørpollen to Nordpollen, was˚

Ž .noted by Wain 1998 and Krabbendam and WainŽ .1997 to coincide with their mixed HPrUHPboundary zone, although such swarms of ultramafitesare not unique to this zone and are also found in theHP zone to the south of Nordfjord. Further north,

two more belts of peridotites stretch from Tafjord toGurskøy and parallel to Moldefjord through Otrøy to

Ž .Nordøyane. Bucher-Nurminen 1991 noted thatperidotites ‘‘decorated’’ basement-cover contacts,and there is a particularly strong association of peri-dotites with anorthosite horizons, e.g., at TafjordŽ .Brueckner, 1977 and to the south of NordfjordŽ .Bryhni, 1966 .

Preserved garnetiferous peridotites are rare andrestricted to the region north of Nordfjord. Carswell

Ž .et al. 1983 distinguished two groups of peridotitesbased on mineralogy, bulk composition and fieldrelationships. An Fe–Ti-enriched group is associatedwith metamorphosed layered mafic complexes ormeta-anorthosites and probably has a Caledonidemetamorphic origin. A Mg–Cr-enriched group isprobably derived from tectonically introduced litho-spheric or sub-lithospheric mantle. Mg–Cr peri-dotites in different ultramafic belts seem to have


distinctive characteristics. For example, in the south-ern belts, ferrobasaltic eclogites enclosed within peri-

Ž .dotite bodies ‘‘internal’’ eclogites are found withprograde-zoned garnets more typical of HP ‘‘exter-nal’’ eclogites, possibly indicative of a prograde

ŽCaledonian evolution Jamtveit, 1984; Griffin and. Ž .Qvale, 1985 . Carswell et al. 1983 defined several

stages of development for the Mg–Cr group, withmegacrystsrporphyroclasts having a mantle origin

Ž .during the mid-Proterozoic Jamtveit et al., 1991and grain rims and neoblasts having equilibrated atlower pressures and temperatures, considered to becompatible with external eclogites. The remarkablerecent discovery from garnetiferous peridotites onOtrøy and Nordøyane of evidence for majoritic com-positions in some garnet megaclasts is consistentwith a Proterozoic deep mantle plume evolution forsome Mg–Cr peridotites prior to emplacement into

Žthe crust van Roermund, 1998; van Roermund and.Drury, 1999; Terry et al., 1999 .

We have made a preliminary re-evaluation ofpublished data for some Opx external eclogites and

Ž .garnetiferous peridotites Table 2 and Fig. 7 . Theaverage of results from the Grt–Cpx and Grt–OpxFe–Mg exchange barometers of Carswell and HarleyŽ . Ž .1990 and Krogh-Ravna 1999 was combined with

Ž .the Carswell and Harley 1990 Grt–Opx Al barome-ter. For the external Opx eclogites and Fe–Ti peri-

Ž .dotites Fig. 7A , rim compositions give generallylower P and T than grain cores, probably due toretrogressive diffusive re-adjustment or growth ofamphibole as discussed above. Grain cores fromexternal Opx eclogites and Fe–Ti peridotites have abroad P–T scatter which is steeper than the trend forphengite and kyanite eclogites. There is no cleargeographic pattern. The most northerly localities from

ŽMoldefjord Solholmen and Skarsholmen on Otrøy,.and Kolmannskog northwest of Molde give P–T

conditions within the quartz field, in contrast to thecoesite eclogites along strike on Fjørtoft. In contrast,grain cores from the Eiksundal Fe–Ti peridotite,

Ž .Hareid Fig. 2 , give a very high pressure, just intothe diamond field. Grain cores from the Grytting

Ž .orthopyroxene eclogite on Statlandet Figs. 2 and 3give a broad group of values whose lower range is

Ž .similar to local phengite eclogites e.g., Drage butcalculated pressures vary by )5 kbar for this singlelocality. The Nybø eclogite Sørpollen lies within the

Fig. 7. Garnet–orthopyroxene–clinopyroxene thermobarometryŽ .calculated from published data references in Table 2 using the

Ž .average of the Carswell and Harley 1990 garnet–orthopyroxeneŽ .thermometer and Krogh-Ravna 1999 garnet–clinopyroxene ther-

Ž .mometers, combined with the Carswell and Harley 1990garnet–orthopyroxene Al-barometer. Cs–Qz and Dm–Gph equi-

Ž .libria, and 58C reference geotherm as in Fig. 6. A Externalorthopyroxene eclogites and Fe–Ti garnet peridotites. Squares —external Opx eclogites; circles — FeTi garnet peridotites, open

Ž .symbols — grain rims; filled symbols — grain cores. B Mg–CrŽ .mantle-derived garnet peridotites. Filled symbols are for cores ofzoned grains and porphyroclasts, or unzoned grains, and opensymbols are for rims of zoned grains and porphyroclasts, andneoblasts.

Ž .mixed HPrUHP zone Fig. 3 and gives extremepressures well within the diamond field, consistentwith those previously determined by Lappin and

Ž . Ž .Smith 1978 and Smith 1988 . Unfortunately, theNybø eclogite lacks free silica, phengite or kyanite,making independent thermobarometric corroborationof these results difficult. The most southerly Opx

Ž .eclogites, at Kvalneset near Totland, Nordfjord andŽ .Hornindal Fig. 2 , give the lowest P and T as might

be expected although T for the Kvalneset eclogite ismuch lower than calculated for nearby phengiteeclogites. Overall, the difficulties in identifyingwell-equilibrated phases in the Opx eclogites con-tinue to make interpretation of thermobarometric es-timates from these rocks problematic and controver-sial.

The Mg–Cr peridotites appear to give a moreŽ .rational pattern Fig. 7B , and probably have the

better textural and chronological control for theirŽ .analyses see Carswell, 1986; Jamtveit et al., 1991 .


Nevertheless, interpretation of geothermobarometricresults remains difficult. Cores of porphyroclasts orunzoned large grains give a range of T and P from

Ž6398C to 8288C and 21.1 to 43.8 kbar which espe-.cially at higher pressures scatters along a linear

array lying close to the ;58Crkm field gradientidentified for phengite- and kyanite-bearing eclogitesŽ .Fig. 6D . However, once again, there is no cleargeographic pattern; indeed, samples from the samebody or area may be widely separated in P and T , asshown by the two samples from Sandvika, Guskøy

Ž .and the nearby Gurskebotn sample Table 2 . InŽcontrast, samples from two other areas Ugelvik and

.Raudhaugene, Otrøy and Kalskaret, Tafjord formdistinct clusters at significantly different positionsalong the array. Porphyroclast rims and neoblasts lie

Žin a steep elongate cluster in the P–T plane Fig..7B , departing from the trend for phengite external

eclogites and overlapping the low P end of theporphyroclastrgrain core array. Pressures are usu-ally, but not always, lower than for grain cores fromthe same body. Only the upper part of the array forrims and neoblasts is compatible with conditions fornearby external eclogites.

While the P–T array for Mg–Cr peridotite graincores and porphyroclasts might feasibly relate to thethermal regime associated with Scandian continentallithosphere subduction, such an interpretation is in-consistent with the Proterozoic age for the garnetperidotite porphyroclast assemblages indicated by

Ž .the Sm–Nd data of Jamtveit et al. 1991 . On theother hand, with such a low dTrd P gradient, theP–T array is thought unlikely to correspond to anambient mid-Proterozoic mantle geotherm and is evenless likely to represent the thermal regime associatedwith emplacement into the sub-continental litho-spheric mantle of a sub-lithospheric mantle plume asdeduced from the relict association of early majoriticgarnets and high-Al orthopyroxenes in garnet peri-

Ždotite bodies on Otrøy van Roermund, 1998; van.Roermund and Drury, 1999; Terry et al., 1999 .

To account for the P–T variations observed forgarnet peridotite samples from within local areas,two factors need to be considered — re-equilibrationof older parageneses and disequilibrium. It might besuggested that, as the bodies were entrained by thecrust and moved within the subduction channel dur-ing Scandian continental collision, they suffered

variably effective recrystallisation and re-equilibra-tion to the conditions along the channel. It is notablethat the P–T results for grain rims and neoblastsŽ .Fig. 7B overlap both the lower end of the porphy-roclast array and the field of phengite UHP eclogitesŽ .Fig. 6D , consistent with the hypothesis that theycrystallised or recrystallised during or after emplace-

Žment in the subducted WGC crust Carswell, 1986;.Jamtveit et al., 1991 . Hence, the overall apparent

P–T array in Fig. 7B may well correspond to acomposite record of Proterozoic and Scandian P–Tconditions. Even so, the spread of data for individualbodies is large and obscures any possible relation-ship between P–T conditions for individual peri-dotite bodies and their adjacent external eclogites.

Consideration of possible disequilibrium dictatesthat caution is needed in the interpretation of theperidotite array. The dTrd P gradient of isoplethsfor the garnet–Al in Opx equilibrium used for pres-

Ž .sure evaluation see Carswell and Harley, 1990happens to equate with a geothermal gradient ofaround 58Crkm, so it would appear that the linearP–T array for garnet peridotite samples may belargely an artifact of the P–T evaluation methodol-ogy, under circumstances where the net transfer reac-tion equilibrium for Al between garnet and orthopy-

Ž .roxene monitoring P fails to keep pace with theŽ .Fe–Mg cation exchange equilibria monitoring T .

Hence, at this stage, we conclude that there is somelimited evidence for individual Mg–Cr peridotiteshaving been sourced from distinct positions in thelithospheric mantle and subsequently partially equili-brating under similar conditions to the externaleclogites and their enclosing crustal rocks. However,uncertainties in the thermobarometry mask any clearpattern and create artifacts which require that anyapparent ‘‘geothermal gradient’’ is viewed with cau-tion.

6. Discussion

Three key issues in the study of UHP rocks areŽthe scale of the UHP units indicative of the volume

of buoyant continental crust capable of being sub-.ducted , the extent to which they equilibrated to

Žambient P–T–fluid conditions controlling the petro-


physical properties of the subducted crust such as.density and strength , and the motion history of UHP

units relative to their surrounding crust and mantleŽ .reflecting their mechanical response to subduction .Our observations of evidence for the distribution ofcoesite indicate that the UHP terrane in the WGR issignificantly more extensive than was previously in-

Ž .dicated by Smith 1988; 1995 , although our moreconservative estimates are smaller than the excessive

Ž .estimate by Coleman and Wang 1995 of 350=150km2 and their rather optimistic estimates for the

Ž .extent of the UHP terranes in Dabieshan China andŽ .Kokchetav Russia of 450=100 and 300=150

km2, respectively. The tectonic boundaries to mostUHP terranes are poorly defined, due to overprintingby late orogenic processes. While some UHP ter-ranes appear to be well-defined, allochthonous litho-tectonic units of limited thickness are bounded by

Žshear zones against low pressure rocks e.g.,.Kokchetav Massif; Kaneko et al., 1998 ; others such

as the Dora Maira Massif have tectonic breaks de-fined principally on petrographic criteria, with noconclusive evidence relating these to present-day

Ž .structures e.g., Michard et al., 1995 . The situationin the WGR is also not well-defined but bears somesimilarities to the large South-Central Dabie Terranein Dabieshan where the eclogites show a generalnorthward increase in P and T , but are divided by apossible tectonic break into an HP unit with zonedgarnets and an UHP unit with coesite and microdia-

Ž .mond Wang et al., 1995; Carswell et al., 1997 .However, unlike the structurally low WGC, theSouth-Central Dabie Terrane appears to occupy ahigh structural level in the orogen, at least at itssouthern margin where it overthrusts other lower Punits near the foreland.

Some previous geodynamic models for the south-ern Scandinavian Caledonides have tended to treatthe WGC as a coherent body during subduction orhave indicated only in a very general way that it is

Žlikely to have been tectonically imbricated e.g.,Carswell and Cuthbert, 1986; Cuthbert and Carswell,1990; Cuthbert et al., 1983; Andersen et al., 1991;

.Dewey et al., 1993 . A simplistic view of the generaltrends in eclogites in the WGR described in theforegoing sections might be that the WGC was,indeed, a continuous, unbroken structural unit during

Žeclogite facies metamorphism although later modi-

.fied during exhumation , and that the UHP part ofthe WGC was simply a higher-grade metamorphiczone which was more deeply subducted than con-tiguous rocks to the south. In this view, the firstappearance of coesite is an isograd.

Other geodynamic models for the WGC havetaken a more mobilistic view; for instance, SmithŽ .1980 proposed a model of a tectonic mega-melange,

Ž .and Austrheim 1991 proposed a melange-like,gravity-differentiated model based on observations inthe Bergen Arcs Anorthosite Nappe. The large extentof mappable lithological units across the WGR dic-tates against such extreme, melange-like models ex-cept perhaps on a local scale, but nevertheless, theWGC is clearly a composite terrane assembled froma number of lithotectonic units. The Mg–Cr peri-dotites are generally agreed to have been introducedinto the WGC from the mantle, and they are UHProcks. Allochthonous or parautochthonous cover units

Žare an important component of the WGC Muret,1960; Brueckner, 1977; Krill, 1985; Bryhni, 1989;

.Robinson, 1995, 1997; Tveten, 1995 . In the Molde-fjord–Nordøyane area, cover and basement appear tohave distinctive assemblages of eclogites. UHP

Ž .eclogites are present in the basement Ulla Gneiss .In the Sætra and Risberget cover nappes, eclogiteshave prograde-zoned garnets typical of HP eclogitesŽ .Robinson, 1995 . The Blahø cover nappe has spo-˚radic occurrences of eclogite, but importantly con-tains the UHP eclogite and diamond-gneiss onFjørtoft, separated from basement by eclogitic my-lonites and pods of garnet peridotite. Evidence fordeep-level emplacement of the Blahø nappe against˚

Žbasement is well-preserved here Terry and Robin-son, 1998; Robinson and Terry, 1999; Terry et al.,

.1999 .In the Norfjord–Stadlandet area, where evidence

for UHP metamorphism has been most commonlyfound, UHP eclogites appear to be found in bothbasement and cover, and late amphibolite faciesre-working has largely obscured any earlier HP fab-rics which might have recorded the kinematics ofUHP nappe emplacement. Assessment of the coher-ence of the WGC in this area during collisionrsub-duction depends critically on interpretation of themixed UHP–HP zone. We shall consider below twopossible end-member scenarios for the origins of themixed zone — tectonic juxtaposition of originally


separate HP and UHP terranes, and a single terraneŽ .exhibiting a metamorphic transition isograd .

The spatial distribution of P across the WGR,while admittedly based upon unevenly distributeddata, does not appear to be smooth and regular inspite of the linear array in P–T space. Based uponthe dataset used in this study, augmented with that of

Ž . ŽKrabbendam and Wain 1997 and Wain 1997a; b;.1998 , pressures vary from about 14–17 kbar in

inner Sunnfjord to 20–23 kbar immediately to thesouth of the mixed HPrUHP zone at Nordfjord,giving a P change of up to 6"3 kbar over a mapdistance of 55 km across the strike of the gneissicfoliation. The southernmost eclogites in the UHP

Žzone proper give pressures of about 28 kbar Wain,.1998 , indicating an apparent pressure change of

6.5"1.5 kbar over a map distance of about 10 kmacross the HPrUHP mixed zone. The highestrecorded pressures are about 33 kbar, which arefound about 40 km across strike from the north sideof the mixed zone, indicating a pressure change of 5kbar over this distance. It therefore appears that themixed zoned represents a strong telescoping ofeclogite-facies isobars. It is difficult to relate theseapparent pressure gradients to true structural thick-ness due to the effects of late folds, but in a morelocal study in the Nordfjord–Stad area where struc-tures are reasonably well-constrained, Krabbendam

Ž .and Wain 1997 demonstrated a strong telescopingof isobars across the mixed zone related either todifferential attenuation of a continuous crustal sec-tion or a distinct tectonic break. Within the mixedzone, pressures derived from adjacent UHP and HPeclogites indicate strong, non-lithostatic apparent

Žpressure gradients Fig. 5; see also Krabbendam and.Wain, 1997; Wain, 1997b . The maximum estimates

of D P between adjacent HP and UHP eclogites areabout 8 kbar over 500 m. Foliation dips are steepalong this part of Nordfjord, so assuming that thisdistance approximates to maximum structural thick-ness and that pressures were purely lithostatic, avertical shortening of about 26 km can be estimatedbetween these eclogites. However, given the cautiousview of uncertainties in pressure estimates adoptedhere, D P between the two groups of HP and UHPeclogites could be as low as 4 kbar, equivalent toabout 13 km vertical shortening over a minimum of;500 m. Following the same reasoning and using

Ž .the estimate of Krabbendam and Wain 1997 of upto 4–6 km present-day structural thickness, a bulkvertical shortening from 13 to 26 km across theentire mixed HPrUHP zone can be calculated. How-ever, this bulk figure may be meaningless, owing tothe close proximity of HP and UHP eclogites withinthis zone.

While the distribution of eclogite types and calcu-lated pressures does seem to support some shorten-ing of crustal section, and hence, tectonic transportbetween the HP and UHP zones and possible tec-tonic mixing within the boundary zone between them,the lithological assemblages in the country-rocks inthe HP and UHP zones are substantially similar andneither zone can be attributed wholly to either base-ment or cover units, as both are present in each zoneand in the mixed zone. Furthermore, the mixed zoneis not marked by any increase in strain intensity, norare the intervening country-rocks between adjacentHP and UHP eclogites within the mixed zone. Hence,it can be concluded that the HP and UHP zones oneither side of Nordfjord have similar crustal prove-nances. The fabrics in the country-rock gneisses aredominated by amphibolite-facies parageneses defin-ing L-tectonite fabrics indicating strong vertical flat-

Ž .tening and E–W extension Andersen et al., 1994 ,Žsometimes constrictional Krabbendam and Dewey,

.1998 . These strains are likely to have made a con-siderable contribution to the juxtaposition of the HPand UHP zones, but as these strains are not unique tothe mixed zone, some extra transport, specificallywithin the mixed zone, and pre-dating the presentfabrics in the country rock gneisses seem likely. Thesimilarity of lithologies in the HP and UHP zoneswould suggest that the UHP terrane is not ‘‘exotic’’to the rest of the WGC, but simply forms a moredeeply subducted part of the same crustal assem-blage which has been tectonically dismembered. Thepresence of numerous peridotite bodies within the

Žmixed zone Krabbendam and Wain, 1997; Wain,.1998 is consistent with such a scenario if they were

incorporated from the mantle bounding the subduc-tion zone; this important observation needs furtherinvestigation. This may not be the only tectonicbreak in the central-southern WGR. In the Dalsford–

Ž .Førde area, inner Sunnfjord Fig. 2 glaucophaneeclogites in the Jostedal Complex basement givetemperatures up to 1008C lower than adjacent bar-


roisite eclogites in the Fjordane complex coverŽ .Cuthbert, 1985; Krogh-Ravna, unpublished data .

Inference of tectonic juxtaposition and loss ofcrustal section from P–T conditions recorded in theHP and UHP eclogites assume that these eclogitesequilibrated at approximately the same time, and thatequilibration was maintained as P, T and fluidcomposition changed. There are currently no isotopicdates available for UHP eclogites, and isotopic dise-quilibrium makes dating of HP eclogites difficultŽ . Ž .Griffin and Brueckner, 1985 . Wain 1997c; 1998 ,

Ž . Ž .Austrheim 1998 , Austrheim and Engvik 1997 andŽ .Austrheim et al. 1997 have shown that disequilib-

rium, as signified by survival of pre-eclogitic plagio-clase-bearing parageneses, is common in dry igneousor granulite protoliths in eclogite-facies terranes, andthe catalytic effect of fluids and deformation are ofgreat importance in transforming crust to eclogitefacies parageneses. The efficiency of transformationswithin the eclogite facies, such as that between HPand UHP rocks, is less well-known, but it seemspossible that HP eclogites could survive metastablyunder UHP conditions. If a HP terrane underwentcontinued subduction and the overstep in P androrT required for the transformation was large, transi-tions between reacted and unreacted rock masseswould record a significant jump in pressure, andpossibly temperature. UHP eclogites could alsorecord younger isotopic ages than HP eclogites, givena precise-enough dating technique.

Disequilibrium within a crustal slab that remainedcoherent during eclogite facies metamorphism couldprovide an alternative explanation for the nature ofthe HP–UHP transition in the WGC. Pressures

Ž .recorded in the HP–UHP transition mixed zone arequite close to the coesite–quartz inversion and tem-peratures are close to those which might be expectedto homogenise garnet compositional zoning for rea-sonable timescales and grain sizes. Unequivocal evi-dence for transformation of HP to UHP eclogite is

Ž .currently lacking; Wain 1998 found no evidence offluid or deformation controlling the development ofUHP assemblages in eclogites from the Nordford–Stadlandet area, except in local retrograde effects.Features such as occurrence of UHP eclogite inhigher strain zones or fluid infiltration channelswithin HP eclogites, or volumes of HP eclogite inlow strain zones in UHP bodies have not yet been

found, and eclogite bodies appear to be uniform inŽ .petrographic type Wain, 1997a, 1998 . Also, within

the mixed zone, eclogite bodies of the same typeŽ .follow distinct horizons in the gneiss Wain, 1998 ,

suggesting tectonic interleaving of HP and UHP rockunits. The survival of a large volume of plagioclase-

Ž .bearing rock in the Flatraket mangerite body Fig. 3would have been more likely if it had only experi-enced HP conditions, as the overstep on the plagio-clase breakdown reaction would have been smallerthan at UHP. The Flatraket body contains HP eclog-ites but is sandwiched between UHP eclogites, sothis line of reasoning requires tectonic juxtaposition

Ž .of HP and UHP rock masses Wain, 1997a,c, 1998 .However, the occurrences of prograde-zoned garnetin UHP eclogites and their surrounding gneisses atAsheimneset, Stadlandet and Vetrhuset, Nordpollenindicates that prograde features can, rarely, surviveUHP conditions. In the light of these tantalisingexceptions, the influence of kinetic factors such asfluids, temperature and deformation in controllingthe transformation from HP to UHP eclogite de-serves closer scrutiny.

Preservation of zoning and inclusion suites in HPeclogites precludes their formation by retrogrademetamorphism of UHP eclogites at eclogite facies,but retrograde effects are seen in the form of changesin phengite chemistry and thin retrograde-zoned rimson garnets. These have the effect of smearing out thedistribution of calculated P and T and creating largeuncertainties in determination of any pressure gapbetween HP and UHP eclogites.

Thus, at this stage, the relative roles of tectonicand kinetic factors in controlling the distribution ofHP and UHP eclogites are uncertain, but both haveundoubtedly contributed to the present distribution ofeclogite types in the WGC. The existence of UHPeclogites in both basement and cover argues for UHPmetamorphism prior to or during tectonic assemblyof lithotectonic units. This assemblage may thenhave suffered some dismembering, certainly duringthe later stages of exhumation at amphibolite facies,but probably also at deeper levels. The present distri-bution of HP and UHP eclogites and their recordedapparent P–T conditions was controlled by thisdeformation, but may also have been affected bykinetically controlled survival of HP eclogite andlater retrogressive adjustment of phase compositions


Ž .especially phengite . Better understanding is likelyto come from a more extensive database of thermo-barometric data, identification of critical areas whereearly structures and parageneses are better preserved,and detailed fabric studies to fully evaluate the im-portance of fluids, strain and recrystallisation in thetransformation from HP to UHP eclogite.

Given the broad pattern of systematic P–T changeacross the WGR, we speculate that the Baltica crustŽ .possibly with a nappe pile already in place above itwas subducted and eclogites formed, equilibrating to

Ž .the transient? geothermal gradient in the subduc-tion zone. The subducted crust became more or lessimbricated in the subduction zone, and was verticallyflattened during the later stages of exhumation, butnot disrupted to the extent that the pattern of P andT recorded along the subducted slab was severely

Žmodified. Imbrication prior to, during or after UHP.metamorphism was apparently effective in entrain-

ing masses of peridotite from the mantle boundingthe subduction channel. The peridotites may havebeen derived from a range of depths in the subduc-tion channel as suggested by their distinct character-istics and clustering of P–T conditions in some

Ž .cases Fig. 7A and B . They probably had diverseorigins and subduction histories with possibilities forentrainment under deep-level UHP conditions andrecrystallisation during transport up the subductionchannel, or transport down the subduction channeland subjection to prograde HPrUHP metamorphism,as seems likely for the Almklovdalen body with itsenclosed ferroan eclogites bearing prograde-zoned

Ž .garnets Griffin and Qvale, 1985 . The commonlocation of peridotites along basement-cover contactsindicates that such surfaces were important detach-ment horizons during subduction or return and, alongwith the mixed HP–UHP transition zone, emphasisethe importance of taking a, more or less, mobilisticview of the development of the WGC.

7. Conclusions

Ž .1 Coesite, or polycrystalline quartz after coesite,has now been found from 32 localities in the WGRand new discoveries significantly extend the

Ž .coesite-bearing UHP terrane to the north and east.

We tentatively suggest that a conservative estimatefor the size of the UHP terrain is ;5000 km2.

Ž .2 Polycrystalline quartz has been found in theBlahø cover nappe and in the adjacent basement˚gneiss, close to the locality where microdiamond hasbeen separated by thermochemical methods from apelitic gneiss, confirming the UHP nature of thisunit.

Ž .3 A review of petrographic features in eclogitessuch as silica polymorph type, garnet zoning andinclusion patterns, matrix parageneses and retrogres-sive evolution shows a spatial variation indicating anincrease of T and P from southeast to northwestacross the WGR, although with some local varia-tions. The presence of a mixed UHPrHP zone as

Ž .described by Wain 1997b; 1998 and KrabbendamŽ .and Wain 1997 is confirmed and extended to the

east.Ž .4 Geothermobarometry using garnet–ompha-

Žcite–phengite assemblages Waters and Martin, 1983;.Krogh-Ravna, 1999 gives P–T conditions generally

consistent with the distribution of silica polymorphs.ŽGiven a cautious evaluation of uncertainties P up to

."2 kbar, T"758C , the variation in T across theWGR suggested by eclogite petrography and early

Ž .geothermometry studies Griffin et al., 1985 is con-firmed and shown to correlate with an increase in Pfrom 16 kbar in the south to )32 kbar in thenorthwest.

Ž .5 P–T conditions for phengite eclogites lie ingeographical order from south to north along a lineararray in the P–T plane, corresponding to a fairlylow ‘‘geothermal gradient’’ of ca. 58Crkm. Thisgradient could, of course, be diachronous, but exist-ing geochronological data are not sufficient to re-solve this.

Ž .6 The mixed HPrUHP zone marks the transi-tion between an HP terrane to the south and an UHPterrane to the north. The mixed zone appears torepresent an area of telescoped isobars within theregional trend of increasing P across the WGR.Within this zone, apparent pressure gradients be-tween HP and UHP eclogites exceed lithostatic gra-dients, even assuming rather cautious uncertainties in

Ž .P "2 kbar and in estimating true structural thick-ness. While telescoping of isobars probably results,at least in part, from tectonic shortening, the litholog-ical similarities of the HP and UHP terranes suggest


that the UHP terrane is not ‘‘exotic’’ with respect tothe HP zone and both have a similar crustal prove-nance, their juxtaposition possibly resulting from a-26 km vertical loss of crustal section, over a zone;4 km in present-day thickness. Vertical flatteningat amphibolite facies may have contributed to thisshortening, with possible earlier crustal imbricationat greater depths to explain the very local juxtaposi-tion of HP and UHP eclogites in this zone. Consider-ation of kinetic factors in the HP–UHP transitionsuggests that survival of HP eclogites under UHPconditions could contribute to the formation of themixed zone, but direct evidence for this is currentlytenuous.

Ž .7 A review of existing data for orthopyroxeneeclogites, all within the UHP terrane, gives resultsbroadly consistent with an UHP origin, but no sys-tematic geographical pattern emerges and the scatterof data from individual localities is large, re-empha-sising the problems of recognising equilibrium com-binations of phase compositions in these petrographi-cally complex rocks.

Ž .8 Mantle-derived, Mg–Cr-enriched garnetifer-ous peridotites are possibly restricted to the UHP andmixed zones of the WGR. Peridotites often decoratebasement-cover contacts and may be useful indica-tors of imbrication, tectonic transport and crust–mantle interaction in the subduction zone. Garnet–opx–cpx geothermobarometry on early Proterozoicporphyroclast assemblages from Mg–Cr garnet peri-dotites gives P–T conditions lying along a lineararray unlikely to correspond to ambient conditionsalong a Proterozoic subcontinental geotherm.Neoblast assemblages and rim compositions of prob-able Scandian age show a trend of P and T depart-ing from the trend for Scandian external phengiteeclogites but overlapping with the UHP eclogites.The overall P–T array for porphyroclastic andneoblastic assemblages in the garnet peridotitesprobably corresponds to a composite record of Pro-terozoic and Scandian P–T conditions. The similar-

Ž .ity of the HP Proterozoic segment of the array toŽ .that defined by the external Caledonian phengite

eclogites is considered to be fortuitous and may bean artifact of the P–T evaluation methodology.

Ž .9 The overall pattern of petrographic character-istics, silica polymorphs, mantle peridotite distribu-tion and P–T conditions across the WGR is consis-

Žtent with previous views e.g., Cuthbert et al., 1983;.Griffin et al., 1985 that the WGC has been sub-

ducted towards the northwest during the Scandiancollision episode, and the general order of P and Talong the subducted slab has been preserved in spiteof severe Scandian reworking during exhumation.

ŽHowever, evidence from the northern WGR Terry.and Robinson, 1998 suggests that significant move-

ment between lithotectonic basement and cover unitshas occurred along with entrainment of mantle peri-dotite. The extent of tectonic modification and differ-ent metamorphic regimes in basement and cover inthe Nordfjord–Stadlandet area is less clear, but themixed HPrUHP zone may represent a significantdisjunction and is also marked by a concentration ofperidotite bodies. Unfortunately, as in all UHP ter-ranes, regional scale eclogite facies structures, whichcan give direct evidence for deep-level tectonic evo-lution, have been largely obliterated by late, lowerpressure deformation and metamorphism and it isnecessary to infer tectonic breaks from discordancesin recorded P and T. Such discordances could alsoresult from kinetic factors operating in a coherent

Ž .crustal unit. Although Wain 1998 found no evi-dence for kinetic controls on the distribution of HPand UHP mineralogies within eclogites from themixed HPrUHP transition zone in the WGR, thiscannot be positively ruled out and some tentativeevidence supports survival of prograde textures un-der UHP conditions. Hence, the interpretation ofP–T jumps in tectonic breaks must be treated withcaution.

Ž .10 A complete understanding of the dynamics ofcontinental collision and subduction requires aknowledge of the interactions and feedback betweenexternal tectonic controls, thermal evolution, meta-morphic reactions and reaction rates, fluid flux andchanges in petrophysical properties of the crust andmantle. The WGR provides a well-exposed and ac-cessible example of a UHP terrane with the potentialto provide information on all these aspects. Furtherwork should concentrate on more completely defin-ing the full extent of the UHP terrane, establishing abetter geochronologic framework for the regionalfield gradient, detailed structural examination of thenature of basement-cover contacts, examination ofperidotites to determine the nature of crust–mantleinteractions, and detailed textural studies to ascertain


the influence of kinetic factors in the creation andpreservation of HP and UHP rocks and relatedchanges in their petrophysical properties.

Acknowledgements

We acknowledge support for recent studies inNorway from the Norwegian Research Council, the

Ž .Norwegian Geological Survey NGU , the BritishCouncil, the Carnegie Trust for the Scottish Univer-sities, the Natural Environment Research Counciland the Universities of Paisley, Sheffield, Tromsøand Oxford. Hakon Austrheim and Dave Waters are˚thanked for their constructive reviews which resultedin significant improvements of the manuscript, andwe are grateful to Hannes Brueckner for suggestionsfor improvements and valuable discussions on thesignificance of garnet peridotites.

References

Andersen, T.B., Jamtveit, B., 1990. Uplift of deep crust duringorogenic extensional collapse: a model based on field studiesin the Sogn–Sunnfjord region of Western Norway. Tectonics9, 1097–1111.

Andersen, T.B., Jamtveit, B., Dewey, J.F., Swensson, E., 1991.Subduction and eduction of continental crust: major mecha-nisms during continent–continent collision and orogenic col-lapse, a model based on the south Norwegian Caledonides.Terra Nova 3, 303–310.

Andersen, T.B., Osmundsen, P.T., Jolivet, L., 1994. Deep crustalfabrics and a model for the extensional collapse of the south-west Norwegian Caledonides. Journal of Structural Geology16, 1191–1203.

Austrheim, H., 1991. Eclogite formation and dynamics of crustalroots under continental collision zones. Terra Nova 3, 492–499.

Austrheim, H., 1998. The influence of fluid and deformation onmetamorphism of the deep crust and consequences for thegeodynamics of collision zones. In: Hacker, B., Liou, J.G.Ž .Eds. , When Continents Collide: Geodynamics and Geochem-istry of Ultrahigh Pressure Rocks. Chapman & Hall, pp.297–323.

Austrheim, H., Engvik, A.K., 1997. Fluid transport, deformationand metamorphism at depth in a collision zone. In: Jamtveit,

Ž .B., Yardley, B. Eds. , Fluid Flow and Transport in Rocks:Mechanisms and Effects. Chapman & Hall, pp. 123–136.

Austrheim, H., Erambert, H., Engvik, A.K., 1997. Processing ofcrust in the root zone of the Caledonian continental collisionzone: the role of eclogitization. In: Touret, J.L.R., Austrheim,

Ž .H. Eds. , Collision Orogens: Zones of Active Transfer Be-

tween Crust and Mantle. Tectonophysics, Vol. 273, pp. 129–156.

Bailey, D.E., 1989. Metamorphic evolution of the crust of south-ern Norway: an example from Sognefjord. PhD thesis, Univer-sity of Oxford.

Bence, A.E., Albee, A.L., 1969. Empirical correction factors forelectron probe microanalyses of silicates and oxides. Journalof Geology 76, 382–403.

Berman, R.G., 1990. Mixing properties of Ca–Mg–Fe–Mn gar-nets. American Mineralogist 75, 328–344.

Bohlen, S.R., Boettcher, A.L., 1982. The quartz–coesite transfor-mation: a pressure determination and the effects of othercomponents. Journal of Geophysical Research 87, 7073–7078.

Boundy, T.M., Essene, E.J., Hall, C.M., Austrheim, H., Halliday,A.N., 1996. Rapid exhumation of lower crust during conti-nent–continent collision and late extension: evidence from40Arr39Ar incremental heating of hornblendes and muscovites,Caledonian orogen, western Norway. Geological Society ofAmerica Bulletin 108, 1425–1437.

Boundy, T.M., Mezger, K., Essene, E.J., 1997. Temporal andtectonic evolution of the granulite–eclogite association fromthe Bergen Arcs, western Norway. Lithos 39, 159–178.

Brueckner, H.K., 1977. A structural, stratigraphic and petrologicstudy of anorthosites, eclogites and ultramafic rocks and theircountry rocks, Tafjord area, western South Norway. NorgesGeologiske Undersøkelse Bulletin 241, 1–53.

Bryhni, I., 1966. Reconnaissance studies of gneisses, ultrabasites,eclogites and anorthosites in outer Nordfjord, Western Nor-way. Norges Geologiske Undersøkelse Bulletin 241, 1–68.

Bryhni, I., 1989. Status of the supracrustal rocks in the WesternŽ .Gneiss Region, S. Norway. In: Gayer, R.A. Ed. , The Cale-

donide Geology of Scandinavia. Graham and Trotman, pp.221–228.

Bryhni, I., Bollingberg, H.J., Graff, P.-R., 1969. Eclogites inquartzo-feldspathic gneisses of Nordfjord, West Norway. NorskGeologisk Tidsskrift 49, 193–225.

Bryhni, I., Grimstad, E., 1970. Supracrustal and infracrustal rocksin the Gneiss Region of the Caledonides west of Breimsvatn.Norges Geologiske Undersøkelse 266, 105–150.

Bryhni, I., Griffin, W.L., 1971. Zoning in eclogite garnets fromNordfjord, west Norway. Contributions to Mineralogy andPetrology 32, 112–125.

Bucher-Nurminen, K., 1991. Mantle fragments in the Scandina-vian Caledonides. Tectonophysics 190, 173–192.

Bundy, F.R., 1980. The P, T phase and reaction diagrams forelemental carbon. Journal of Geophysical Research 85, 6930–6936.

Carswell, D.A., 1974. Comparative equilibration temperatures andpressures of garnet lherzolites in Norwegian gneisses andkimberlite. Lithos 7, 113–121.

Carswell, D.A., 1986. The metamorphic evolution of Mg–Cr-typeNorwegian garnet peridotites. Lithos 19, 279–297.

Carswell, D.A., Cuthbert, S.J., 1986. Eclogite facies metamor-phism in the lower continental crust. In: Dawson, J.B., Car-

Ž .swell, D.A., Hall, J., Wedepohl, K.H. Eds. , The Nature ofthe Lower Continental Crust. Geological Society Special Pub-lication, Vol. 25, pp. 193–209.


Carswell, D.A., Harley, S.L., 1990. Mineral barometry and ther-Ž .mometry. In: Carswell, D.A. Ed. , Eclogite Facies Rocks.

Blackie, Glasgow, pp. 83–110.Carswell, D.A., Harvey, M.A., Al-Samman, A., 1983. The petro-

genesis of contrasting Fe–Ti and Mg–Cr garnet peridotitetypes in the high grade gneiss complex of Western Norway.Bulletin Mineralogique 106, 727–750.´

Carswell, D.A., Krogh, E.J., Griffin, W.L., 1985. Norwegianorthopyroxene eclogites: calculated equilibration conditions

Ž .and petrogenetic implications. In: Gee, D.G., Sturt, B.A. Eds. ,The Caledonide Orogen — Scandinavia and Related Areas.Wiley, Chichester, pp. 823–841.

Carswell, D.A., O’Brien, P.J., Wilson, R.N., Zhai, M., 1997.Thermobarometry of phengite-bearing eclogites in the DabieMountains of central China. Journal of Metamorphic Geology15, 239–252.

Chauvet, A., Brunel, M., 1988. La grande faille normale ductiledu Sunnfjord. Une extension NW–SE dans la chaıneˆcaledonienne de l’Ouest Norvege. C. R. Acad. Sci. Paris, Ser.´ `II 307, 415–422.

Chauvet, A., Seranne, M., 1989. Microtectonic evidence of Devo-nian extensional westward shearing in Southwest Norway. In:

Ž .Gayer, R.A. Ed. , The Caledonide Geology of Scandinavia.Graham and Trotman, pp. 245–254.

Chopin, C., 1984. Coesite and pure pyrope in high-gradeblueschists of the Western Alps: a first record and someconsequences. Contributions to Mineralogy and Petrology 86,107–118.

Coleman, R.G., Wang, X., 1995. Overview of the geology andŽ .tectonics of UHPM. In: Coleman, R.G., Wang, X. Eds. ,

Ultrahigh Pressure Metamorphism. Cambridge Univ. Press,pp. 1–32.

Cuthbert, S.J., 1985. Petrology and tectonic setting of relativelylow temperature eclogites and related rocks in the Dalsfjordarea, Sunnfjord, West Norway. PhD thesis, University ofSheffield.

Cuthbert, S.J., 1991. Evolution of the Devonian Hornelen Basin,west Norway: new constraints from petrological studies ofmetamorphic clasts. In: Morton, A.C., Todd, S.P., Haughton,

Ž .P.W.D. Eds. , Developments in Sedimentary ProvenanceStudies. Geological Society Special Publication No. 57, pp.343–360.

Cuthbert, S.J., 1995. Trondhjemite veins in eclogite from theWestern Gneiss Region, Norwegian Caledonides; evidence forpartial melting. Chinese Science Bulletin 40, 103–104, Abstr.Suppl. 3rd Int. Eclogite Field Symp.

Cuthbert, S.J., 1997. Exhumation-related partial melting of eclog-ites and gneisses in the Western Gneiss Region, Norwegian

ŽCaledonides. Terra Nova 9, Abstract No. 121 Abstr. Suppl. 1,.5th Int. Eclogite Conf. .

Cuthbert, S.J., Carswell, D.A., 1990. Formation and exhumationof medium-temperature eclogites in the Scandinavian Cale-

Ž .donides. In: Carswell, D.A. Ed. , Eclogite Facies Rocks.Blackie, Glasgow, pp. 180–203.

Cuthbert, S.J., Harvey, M.A., Carswell, D.A., 1983. A tectonicmodel for the metamorphic evolution of the Basal Gneiss

Complex, western south Norway. Journal of MetamorphicGeology 1, 63–90.

Dallmeyer, R.D., Gee, D.G., 1986. 40Arr39Ar mineral dates fromretrogressed eclogites within the Baltoscandian miogeocline:implications for a polyphase Caledonian orogenic evolution.Geological Society of America Bulletin 97, 26–34.

Dewey, J.F., Ryan, P.D., Andersen, T.B., 1993. Orogenic upliftand collapse, crustal thickness, fabrics and metamorphic phasechanges: the role of eclogites. In: Pritchard, H.M., Alabaster,

Ž .T., Harris, N.B.W., Creary, C.R. Eds. , Magmatic Processesand Plate Tectonics. Geological Society Special PublicationNo. 76, pp. 325–343.

Dobrzhinetskya, L.F., Eide, E.A., Larsen, R.B., Sturt, B.A.,Trønnes, R.G., Taylor, W.R., Poshukhova, T.V., 1995. Micro-diamonds in high-grade metamorphic rocks from the WesternGneiss Region, Norway. Geology 23, 597–600.

Droop, G.T.R., 1987. A general equation for estimating Fe3q

concentrations in ferromagnesian silicates and oxides frommicroprobe analyses, using stoichiometric criterion. Miner-alogical Magazine 51, 431–435.

Essex, R.M., Gromet, L.P., Andreasson, P.-G., Albrecht, L., 1997.Early Ordovician U–Pb metamorphic ages of the eclogite-bearing Seve Nappes, Northern Scandinavian Caledonides.Journal of Metamorphic Geology 15, 665–676.

Fossen, H., Rykkelid, E., 1992. Postcollisional extension of theCaledonide orogen in Scandinavia: structural expressions andtectonic significance. Geology 20, 737–740.

Griffin, W.L., 1987. ‘‘On the eclogites of Norway’’ — 65 yearslater. Mineralogical Magazine 51, 333–343.

Griffin, W.L., Austrheim, H., Brastad, K., Bryhni, I., Krill, A.G.,Krogh, E.J., Mørk, M.B.E., Qvale, H., Tørudbakken, B., 1985.High-pressure metamorphism in the Scandinavian Cale-

Ž .donides. In: Gee, D.G., Sturt, B.A. Eds. , The CaledonideOrogen — Scandinavia and Related Areas. Wiley, Chichester,pp. 783–801.

Griffin, W.L., Brueckner, H.K., 1985. REE, Rb–Sr and Sm–NdŽstudies of Norwegian eclogites. Chemical Geology Isotope

.Geoscience Section 52, 249–271.Griffin, W.L., Qvale, H., 1985. Superferrian eclogites and the

crustal origin of garnet peridotites, Almklovdalen, Norway. In:Ž .Gee, D.G., Sturt, B.A. Eds. , The Caledonide Orogen —

Scandinavia and Related Areas. Wiley, Chichester, pp. 803–812.

Hartz, E., Andresen, A., Andersen, T.B., 1994. Structural observa-tions adjacent to a large-scale extensional detachment zone inthe hinterland of the Norwegian Caledonides. Tectonophysics231, 123–137.

Harvey, M.A., 1983. A geochemical and Rb–Sr study of theProterozoic augen orthogneisses on the Molde Peninsula, westNorway. Lithos 16, 325–338.

Heinrich, C.A., 1982. Kyanite–eclogite to amphibolite faciesmetamorphism of hydrous mafic rocks in the central alpineAdula nappe. Journal of Petrology 27, 123–154.

Holland, T.J.B., 1980. Activities of components in omphaciticsolid solutions: an application of Landau theory to mixtures.Contributions to Mineralogy and Petrology 105, 446–453.


Hossack, J.R., 1984. The geometry of listric growth faults in theDevonian basins of Sunnfjord, West Norway. Journal of theGeological Society of London 141, 629–637.

Jamtveit, B., 1984. High-P metamorphism and deformation of theGurskebotn garnet peridotite, Sunnmøre, Western Norway.Norsk Geologisk Tidsskrift 64, 97–110.

Jamtveit, B., 1987. Metamorphic evolution of the Eiksunddaleclogite Complex, Western Norway, and some tectonic impli-cations. Contributions to Mineralogy and Petrology 95, 82–99.

Jamtveit, B., Carswell, D.A., Mearns, E.W., 1991. Chronology ofthe high pressure metamorphism of Norwegian garnet peri-dotitesrpyroxenites. Journal of Metamorphic Geology 9, 125–139.

Kaneko, Y., Maruyama, S., Terabayashi, M., Yamamoto, H.,Ishkawa, M., Anma, R., Parkinson, C.D., Ota, T., Nakajima,Y., Yamamoto, J., Masago, H., Liou, J.G., Ogasawara, Y.,1998. Geology of the eastern Kokchetav metamorphic belt,northern Kazakhstan. International Workshop on UHP Meta-morphism and Exhumation. Stanford University, pp. A-86–A-

Ž .90 Abstr. .Kildal, E.S., 1970. Geologisk karte over Norge, berggrunnskart

Maloy 1:250 000, Norske Utgave. Norges GeologiskUndersøkelse.

Krabbendam, M., Dewey, J.F., 1998. Exhumation of UHP rocksby transtension in the Western Gneiss Region, ScandinavianCaledonides. In: Holdsworth, R.E., Strachan, R.A., Dewey,

Ž .J.F. Eds. , Continental Transpressional and TranstensionalTectonics. Geological Society of London Special PublicationNo. 135, pp. 159–181.

Krabbendam, M., Wain, A.L., 1997. Late-Caledonian structures,Ž .differential retrogression and structural position of ultra high

pressure rocks in the Norfjord–Stadlandet area, Western GneissRegion. Norges Geologisk Undersøkelse Bulletin 432, 127–139.

Krill, A., 1985. Relationships between the Western Gneiss Regionand the Trondheim Region: Stockwerk tectonics reconsidered.

Ž .In: Gee, D.G., Sturt, B.A. Eds. , The Caledonide Orogen —Scandinavia and Related Areas. Wiley, Chichester, pp. 475–483.

Krogh, E.J., 1977. Evidence for a Precambrian continent–conti-nent collision in western Norway. Nature 267, 17–19.

Krogh, E.J., 1980. Geochemistry and petrology of glaucophane-bearing eclogites and associated rocks from Sunnfjord, West-ern Norway. Lithos 13, 355–380.

Krogh, E.J., 1981. Compatible P – T conditions for eclogites andsurrounding gneisses in the Kristiansund area, western Nor-way. Contributions to Mineralogy and Petrology 75, 387–393.

Krogh, E.J., 1982. Metamorphic evolution of Norwegian country-rock eclogites, as deduced from mineral inclusions and com-positional zoning in garnets. Lithos 15, 305–321.

Krogh, E.J., Carswell, D.A., 1995. HP and UHP eclogites andgarnet peridotites in the Scandinavian Caledonides. In: Cole-

Ž .man, R.G., Wang, X. Eds. , Ultrahigh Pressure Metamor-phism. Cambridge Univ. Press, pp. 244–298.

Krogh, T.E., Mysen, B.O., Davis, G.L., 1974. A Paleozoic age forthe primary minerals of a Norwegian eclogite. Annual Report

of the Geophysical Laboratory, Vol. 73. Carnegie Institute,WA, pp. 575–576.

Krogh-Ravna, E.J., 1999. The garnet–clinopyroxene geother-mometer — an updated calibration. Journal of MetamorphicGeology, submitted.

Kullerud, L., Tørrudbakken, B.O., Illebekk, S., 1986. A compila-tion of radiometric age determinations from the Western GneissRegion, South Norway. Norges Geologisk Undersøkelse Bul-letin 406, 17–42.

Lappin, M.A., 1966. The field relationships of basic and ultrabasicmasses in the basal gneiss complex of Stadlandet andAlmklovdalen, Nordfjord, southwestern Norway. Norsk Geol-ogisk Tidsskrift 46, 439–495.

Lappin, M.A., 1974. Eclogites of the Sunndal Grubse ultramaficmass, Almklovdalen, Norway and the T – P history of theAlmklovdalen mass. Journal of Petrology 15, 567–601.

Lappin, M.A., Pigeon, R.T., van Breemen, O., 1979. Geochronol-ogy of basal gneisses and mangerite syenites of Stadlandet,West Norway. Norsk Geologisk Tidsskrift 59, 161–181.

Lappin, M.A., Smith, D.C., 1978. Mantle equilibrated eclogitepods from the Basal Gneisses in the Selje district, WesternNorway. Journal of Petrology 19, 530–584.

Lux, D.R., 1985. KrAr ages from the Basal Gneiss Region,Stadlandet area, western Norway. Norsk Geologisk Tidsskrift65, 277–286.

Medaris, L.G. Jr., 1980. Petrogenesis of the Lien peridotite andassociated eclogites, Almklovdalen, western Norway. Lithos13, 339–353.

Medaris, L.G. Jr., 1984. A geothermobarometric investigation ofgarnet peridotites in the Western Gneiss Region of Norway.Contributions to Mineralogy and Petrology 87, 72–86.

Michard, A., Henry, C., Chopin, C., 1995. Structures in UHPMrocks: a case study from the Alps. In: Coleman, R.G., Wang,

Ž .X. Eds. , Ultrahigh Pressure Metamorphism. Cambridge Univ.Press, pp. 132–158.

Milnes, A.G., Dietler, T.N., Koestler, A.G., 1988. The Sognefjordnorthshore log — a 25-km depth section through Caledonised

Ž .basement in Western Norway. In: Kristoffersen, Y. Ed. ,Progress in Studies of the Lithosphere in Norway. NorgesGeologisk Undersøkelse Special Publication No. 3, pp. 114–121.

Milnes, A.G., Wennberg, O.P., Skar, Ø., Koestler, A.G., 1997.˚Contraction, extension and timing in the South NorwegianCaledonides: the Sognefjord transect. In: Burg, J.-P., Ford, M.Ž .Eds. , Orogeny Through Time. Geological Society SpecialPublication No. 121, pp. 123–148.

Moller, C., 1988. Geology and metamorphic evolution of the¨Roan area, Vestranden, Western Gneiss Region, Central Nor-wegian Caledonides. Norges Geologiske Undersøkelse Bul-letin 413, 1–31.

Mørk, M.B.E., 1985. A gabbro to eclogite transition on Flemsøy,Sunnmøre, west Norway. Chemical Geology 50, 283–310.

Mørk, M.B.E., Kullerud, L., Stabel, A., 1988. Sm–Nd dating ofSeve eclogites, Norbotten, Sweden — evidence for early

Ž .Caledonian 505 Ma subduction. Contributions to Mineralogyand Petrology 99, 344–351.


Mørk, M.B.E., Mearns, E.W., 1986. Sm–Nd isotopic systematicsof a gabbro–eclogite transition. Lithos 19, 255–267.

Muret, G., 1960. Partie S.E. de la culmination du Romsdal,Chaine Caledonienne, Norvege. International Geological` `Congress, Norden 19, 28–32.

Mysen, B.O., Heier, K.S., 1971. Petrogenesis of eclogites inhigh-grade metamorphic gneisses exemplified by the Hareid-land eclogite, west Norway. Contributions to Mineralogy andPetrology 36, 73–94.

Newton, R.C., Haselton, H.T., 1981. Thermodynamics of thegarnet–plagioclase–Al SiO –quartz geobarometer. In: New-2 5

Ž .ton, R.C., Navrotsky, A., Wood, B.J. Eds. , Thermodynamicsof Minerals and Melts. Springer-Verlag, New York, pp. 131–147.

Norton, M.G., 1986. Late Caledonide extension in Western Nor-way — a response to extreme crustal thickening. Tectonics 5,195–204.

Norton, M.G., 1987. The Nordfjord–Sogn detachment, W. Nor-way. Norges Geologiske Tidsskrift 67, 93–106.

Pidgeon, R.J., Raheim, A., 1972. Geochronological investigation˚of the gneisses and minor intrusive rocks from Kristiansund,West Norway. Norsk Geologisk Tidsskrift 52, 241–246.

Roberts, D., Gee, D.G., 1985. An introduction to the structure ofthe Scandinavian Caledonides. In: Gee, D.G., Sturt, B.A.Ž .Eds. , The Caledonide Orogen — Scandinavia and RelatedAreas. Wiley, Chichester, pp. 55–68.

Robinson, P., 1995. Extension of Trollheimen tectono-strati-graphic sequence in deep synclines near Molde and Brattvag,˚Western Gneiss Region, southern Norway. Norsk GeologiskTidsskrift 75, 181–198.

Robinson, P., 1997. Quartzite of Reksdalshesten, Moldefjorden,Western Gneiss Region: another parallel with Trollheimen.Norges Geologiske Undersøkelse Bulletin 433, 14–15.

Robinson, P., Terry, M.P., 1998. Evolution of amphibolite-faciesstructural features during transition from convergence to ex-tensional collapse, Western Gneiss Region, Norway. Interna-tional Workshop on UHP Metamorphism and Exhumation.

Ž .Stanford University, pp. A-79–A-83 Abstr. .Robinson, P., Terry, M.P., 1999. Evolution of amphibolite-facies

structural features during transition from convergence to ex-tensional collapse. Western Gneiss Region, Norway. Lithos,submitted.

Seranne, M., 1992. Late Palaeozoic kinematics of the Møre–´Trøndelag Fault Zone and adjacent areas, central Norway.Norsk Geologisk Tidsskrift 72, 141–158.

Smith, D.C., 1980. A tectonic melange of foreign eclogites and`ultramafics in west Norway. Nature 287, 366–367.

Smith, D.C., 1984. Coesite in clinopyroxene in the Caledonidesand its implication for geodynamics. Nature 310, 641–644.

Smith, D.C., 1988. A review of the peculiar mineralogy of the‘‘Norwegian Coesite Eclogite Province’’, with crystal-chem-ical, petrological, geochemical and geodynamical notes and an

Ž .extensive bibliography. In: Smith, D.C. Ed. , Eclogites andEclogite-Facies Rocks. Elsevier, Amsterdam, pp. 1–206.

Smith, D.C., 1995. Microcoesite and microdiamonds in Norway:Ž .an overview. In: Coleman, R.G., Wang, X. Eds. , Ultrahigh

Pressure Metamorphism. Cambridge Univ. Press, pp. 299–355.

Statham, P.J., 1976. A comparative study of techniques for quanti-Ž .tative analysis of the X-ray spectra obtained with a Si Li

detector. X-ray Spectrometry 5, 16–28.Stephens, M.B., 1988. The Scandinavian Caledonides: a complex-

Ž .ity of collisions. Geology Today January–February 198820–26.

˚Straume, A.K., 1997. Retrograde metamorfose av eklogiter fraHaddal–Ulsteinvik, Sunnmøre. Masters Thesis, Universitet iOslo.

Terry, M.P., Robinson, P., 1998. Eclogite facies structural featuresand their bearing on mechanisms for production of highpressure rocks, Western Gneiss Region, Norway, including themicrodiamond-bearing gneiss at Fjørtoft. International Work-shop on UHP Metamorphism and Exhumation. Stanford Uni-

Ž .versity, pp. A-74–A-78 Abstr. .Terry, M.P., Robinson, P., Carswell, D.A., Gasparik, T., 1999.

Evidence for a Proterozoic mantle plume and a thermotectonicmodel for exhumation of garnet peridotites, Western GneissRegion, Norway. American Geophysical Union Spring Meet-ing, EOS Transactions 80, 359–360.

Tucker, R.D., Krogh, T.E., Raheim, A., 1990. Proterozoic evolu-˚tion and age–province boundaries in the central WesternGneiss Region, Norway: results of U–Pb dating of accessoryminerals from Trondheimsfjord to Geiranger. In: Gower, C.F.,

Ž .Rivers, T., Ryan, B. Eds. , Mid-Proterozoic Laurentia–Bal-tica. Geological Association of Canada, Special Paper 38, pp.149–173.

Tveten, E., 1995. Regional foliation patterns in gneisses as an aidto the correlation of metasupracrustal rocks in the WesternGneiss Region, southern Norway. Norges GeologiskeUndersøkelse Bulletin 427, 29–32.

Tveten, E., Lutro, O., 1995a. Geologisk kart over Norge,˚berggrunnskart ALESUND — 1:250.000. Foreløpige Utgave.

Norges Geologiske Undersøkelse.Tveten, E., Lutro, O., 1995b. Geologisk kart over Norge,

berggrunnskart ULSTEINVIK — 1:250.000. Foreløpige Ut-gave. Norges Geologiske Undersøkelse.

Ž .van Roermund, H.L.M., 1998. An ultra-deep d)200 km oro-genic peridotite body in western Norway. Eos TransactionsAmerican Geophysical Union 79, Abstr. V21E-03, F971.

van Roermund, H.L.M., Drury, M.R., 1999. Ultra-high pressureŽ .P )6 GPa garnet peridotites in Western Norway: exhuma-tion of mantle rocks from )185 km depth. Terra Nova 10,295–301.

Wain, A.L., 1997a. Ultrahigh pressure metamorphism in westernNorway: a tectonic or kinetic problem? Terra Nova 9, 94,

Ž .Abstr. Suppl. 1 EUG .Wain, A.L., 1997b. New evidence for coesite in eclogite and

gneisses: defining an ultrahigh pressure province in the West-ern Gneiss Region of Norway. Geology 25, 927–930.

Wain, A.L., 1997c. Plagioclase metastability and phase transfor-mations in the Coesite–Eclogite Province of Western Norway:tectonic or kinetic implications for UHPM. Terra Nova 9, 39,Abstr. Suppl. 1, 5th Int. Eclogite Conf., Ascona.

Wain, A.L., 1998. Ultrahigh pressure metamorphism in the West-ern Gneiss Region of Norway. PhD thesis, University ofOxford.


Wang, X., Zhang, R., Liou, J.G., 1995. UHPM terrane in eastŽ .central China. In: Coleman, R.G., Wang, X. Eds. , Ultrahigh

Pressure Metamorphism. Cambridge Univ. Press, pp. 356–390.Waters, D.J., Martin, H.N., 1983. Geobarometry of phengite-

bearing eclogites. Terra Abstracts 5, 410–411.Wennberg, O.P., 1996. Superimposed fabrics due to reversal of

shear sense: an example from the Bergen Arc Shear Zone,western Norway. Journal of Structural Geology 18, 871–889.

Wilks, W.J., Cuthbert, S.J., 1994. The evolution of the HornelenBasin detachment system, western Norway: implications forthe style of late orogenic extension in the southern Scandina-vian Caledonides. Tectonophysics 238, 1–30.

Documents

Eclogites and eclogites in the Western Gneiss Region, Norwegian Caledonides€¦ · Eclogites and eclogites in the Western Gneiss Region, Norwegian Caledonides S.J. Cuthbert a,),