full text (368 k)

12410003–004X/97/1112–1241$05.00

American Mineralogist, Volume 82, pages 1241–1254, 1997

High-grade contact metamorphism of calcareous rocks from the Oslo Rift,Southern Norway

BJøRN JAMTVEIT,1 SVEN DAHLGREN,2 AND HAAKON AUSTRHEIM3

1Department of Geology, University of Oslo, P.O. Box 1047 Blindern, N-0316 Oslo, Norway2The Norwegian Petroleum Directorate, P.O. Box 600, 4001 Stavanger, Norway

3Mineralogical-Geological Museum, Sars gt.1, N-0562 Oslo, Norway

ABSTRACTShallow-level plutons caused extensive contact metamorphism of Lower Paleozoic shale

and carbonate sequences in the Permian Oslo Rift. A .500 m long and 100 m wide shale-limestone xenolith embedded within monzonites belonging to the Skrim plutonic complexexperienced high-grade contact metamorphism and generation of minerals and mineralassemblages rarely reported from metamorphic rocks. The peak metamorphic (Stage I)assemblages in calcite-saturated rocks include wollastonite, melilites, fassaitic pyroxenes,phlogopite, titanian grossular, kalsilite, nepheline, perovskite, cuspidine, baghdadite, pyr-rhotite, and occasional graphite. Mineral reactions involving detrital apatite produced aseries of silicate apatites, including the new mineral species Ca3.5(Th,U)1.5Si3O12(OH). Thisassemblage equilibrated at T 5 820–870 8C with a C-rich, internally buffered pore-fluid(20–40 mol% CO2 1 CH4). During cooling the shale-limestone xenolith experienced infil-tration of C-poor (, 0.1 mol% CO2) fluids, triggering the formation of retrograde (StageII) mineral assemblages comprising monticellite, tilleyite, vesuvianite, grandite garnets,diopside, and occasional hillebrandite. Rare potassium iron sulfides (rasvumite and djer-fisherite) formed at the expense of primary pyrrhotite. These assemblages probably formednear 700 8C. Formation of diffuse sodalite-bearing veinlets was associated with breakdownof nepheline and the replacement of kalsilite and wollastonite by potassium feldspar. Thesodalite-bearing Stage III assemblage formed by the infiltration of saline brines at a max-imum temperature of 550 8C. Low-temperature (Stage IV) retrogression of the Stage I-IIIassemblage produced scawtite, giuseppettite, hydrogrossulars, phillipsite, thomsonite, andthree hitherto undescribed mineral species.

INTRODUCTIONMineral assemblages and mineral reactions occurring

during contact metamorphism of calcareous rocks havebeen extensively studied and described since the pio-neering work by Goldschmidt (1911) in the Oslo rift. Ob-served mineral assemblages reflect a wide range of tem-perature and fluid-composition space (Tracy and Frost1991). Equilibrium temperatures approaching and evenexceeding 1000 8C are reported for limestone assem-blages in aureoles around mafic intrusions (e.g., Joesten1983). Because most mafic intrusions crystallize withoutreleasing large quantities of volatiles, most high-gradecontact metamorphic carbonate rocks have buffered thelocal pore-fluid composition during heating and devola-tilization. High-grade assemblages reflecting equilibrationwith H2O-dominated fluids are rarely reported from shal-low level (P # 1000 bars) environments (see however,Williams-Jones 1981 and Inderst 1987).In the Oslo area, contact metamorphism is related to

various shallow level intrusions. Frequently devolatiliza-tion reactions in calcareous rocks are driven by infiltra-tion of externally derived H2O-dominated fluids (Jamtveit

et al. 1992a, 1992b; Jamtveit and Andersen 1993; Sven-sen and Jamtveit 1998). Here, we present the first de-scription of high-grade mineral assemblages in contactmetamorphic carbonates from the Oslo rift. A large xe-nolith of Lower Paleozoic sedimentary rock within thePermian Skrim monzonite complex in the Southern Oslorift experienced metamorphic temperatures exceeding thesolidus of the crystallizing monzonite. Mineral reactionsinvolving carbonate and silicate minerals led to the pro-duction of various of Si-poor silicates, including numer-ous phases rarely or never previously reported from meta-morphic rocks. Reactions between silicates and pyrrhotiteproduced rare potassium iron sulfides, and reactions in-volving detritial apatite produced LREE and Th-rich sil-icate apatites including a new mineral species[Ca3.5Th1.5Si3O12 (OH)].

GEOLOGICAL SETTINGThe Permo-Carboniferous Oslo paleorift is situated in

the southwestern part of the Baltic shield. The exposedparts of the rift are composed of two half grabens filledwith Carboniferous sedimentary rocks, alkaline volcanic

1243JAMTVEIT ET AL.: CONTACT METAMORPHISM OF CALCAREOUS ROCKS

TABLE 1. Abbreviations and formulas for observed minerals in limestones and calc-silicate rocks from Flekkeren

Name Formula Petrografic status

abadakalanapau

albiteandraditeakermanitealabanditeanorthiteapatiteaugite (fassaite)

NaAlSi3O8Ca3Fe2Si3O12Ca2MgSi2O7MnSCaAl2Si2O8Ca5(PO4)3(OH)Ca(Mg,Fe,Al,Ti)(Si,Al)2O6

Stage IVStage IVStage IStage I?Stage I*detritalStage I

babgbr

baddeleyitebaghdaditebritholite

ZrO2Ca3ZrSi2O9(LREE,Ca)5(SiO4,PO4)3(OH)

inclusions in zirconStage IStage I

cccrcu

calsitechromitecuspidine

CaCO3FeCr2O4Ca4Si2O7(F,OH)2

all assemblagesdetritalStage I

didj

diopsidedjerfisherite

CaMgSi2O6K6(Fe,Ni)25S26Cl

Stage IIStage II

gagpgrgu

galenagraphitegrossulargiuseppettite

PbSCCa3Al2Si3O12(Na,K,Ca)7–8(Si,Al)12O24(SO4,Cl)1–2

detritalStage I?Stage IIStage IV

hbhi

hillebranditehibschite

Ca2SiO3(OH)2Ca3Al2Si32xO12(OH)4x (x 5 0.2–1.5)

Stage II?Stage IV

kakfskiks

katoiteK-feldsparkimzeyitekalsilite

Ca3Al2Si32xO12(OH)4x (x 5 1.5–3)KAlSi3O8Ca3(Zr,Ti)2(Si,Al)3O12KAlSiO4

Stage IVStage II or IIIStage IStage I

lo lollingite FeAs2 detrital?mamcmgmo

marialitemarcasitemagnetitemonticellite

Na4Al3Si9O24ClFeS2Fe3O4CaMgSiO4

Stage I*Stage IVdetritalStage II

ne nepheline NaAlSiO4 Stage Iop opal SiO2 Stage IVpephpiptpypyh

perovskitephlogopitephillipsitepentlanditepyritepyrrhotite

CaTiO3KMg3AlSi3O10(OH)2(K,Na,Ca)1–2(Si,Al)8O16·6 H2O(Fe,Ni)9S8FeS2Fe12xS

Stage IStage IStage IVStage IStage IVStage I

qz quartz SiO2 Stage I*rs rasvumite KFe2S3 Stage IIsascsdshsosph

saponitescawtitesodalitescheeliteschorlomitesphalerite

(Ca/2,Na)0.3(Mg,Fe)3(Si,Al)4O10(OH)2·4H2OCa7Si6(CO3)O18·2 H2ONa8Al6Si6O24Cl2CaWO4Ca3Ti2(Fe,Si)3O12ZnS

Stage IVStage IVStage IIIdetritalStage Idetrital

tatbthtitmtoty

thaumasite?tobermoritethoritetitanitethomsonitethorianitetilleyite

Ca6Si2(CO3)2(SO4)2(OH)12·24 H2OCa9Si12O30(OH)6·4 H2OThSiO4CaTiSiO5NaCa2Al5Si5O20·6 H2OThO2Ca5Si2O7(CO3)2

Stage IVStage IVdetritalStage I*Stage IVStage IVStage II

uruvvswozrn

uraniniteuvarovitevesuvianitewollastonitezircon

UO2Ca3Cr2Si3O12Ca10Mg2Al4(SiO4)5(Si2O7)2(OH)4CaSiO3ZrSiO4

detritalStage IStage IIStage I–IIIStage I*

X1X2X3X4Y1Y2

new speciesnew zeolite?new species?new species?unid. speciesunid. species

Ca3.5(Th,U)1.5(SiO4)3(OH)Na2CaAl4Si4O16·X H2ONa2Ca4Si6O15(OH)4Ca¯0.8Al¯0.2SiOn(OH)m·X H2OLREE,Ca-Silicate(stillwellite ?)Ca,LREE,Th,U,Ti,Nb-oxide

Stage IStage IVStage IVStage IVStage I??detrital

* Only in peak assemblage of calcite-free rocks.

The Flekkeren limestone raft is located within the south-western part of the Permian Skrim batholith (Fig. 1; Dahl-gren 1998). The Skrim batholith consists largely of a seriesof alkaline plutons of larvikite (monzonite with ternary feld-spar), syenites and granites, which transects north-trending

Sveconorwegian (1200–1000 Ma) gneisses, quartzites, andamphibolites, easterly dipping Cambro-Silurian sedimentaryrocks, and Upper Paleozoic basaltic volcanic rocks. Thelimestone raft (Fig. 1) is totally embedded in larvikite thatwas emplaced at 281 6 1 Ma (Pedersen et al. 1995). Both

1244 JAMTVEIT ET AL.: CONTACT METAMORPHISM OF CALCAREOUS ROCKS

TABLE 2. Observed mineral assemblages in contactmetamorphic limestones

Sample

Stage I

FL15 79D 79A 79A

Stage II

79B 79B 79D FL17

StageIII

FL32

ccwoMelGrtCpxphcupeksne

xxxxxx

xxx

xxx

xxx

xxxxx

xxx

xxxx

xxxx

x

x

xxx

x

x

xxx

x

x

xxx

x

x

x

xxx

xx

x

x

vsmoty

xxx

xxx

x x

kfssdti

xxx

Notes: Not including phases outside the KNaCaFeMgAlTiSiCOH systemand late (post Stage III) reaction products. Abbreviations used for mineralgroups, rather than end-members, start with an upper case letter as fol-lows: Grt 5 garnet; Mel 5 melilite, Cpx 5 clinopyroxene; Sca 5 scapolite.

TABLE 3. Representative mineral analyses

Sample

Garnet

Stage IMelanite

Stage IUvarovite

Stage IKimzeyite

Stage IIGrossular

Stage IVKatoite

Clinopyroxene

Stage IAugite

Stage IIDiopside

Melilite

Stage I Stage I

Phlogopite

Stage I Stage I

Kalsilite

Stage I

Nepheline

Stage Ino. FL17 FL-1 FL-B FL-C FL-C FL-A FL-E FL-A FL-B FL-A FL-A FL-A FL21

SiO2TiO2Al2O3Fe2O3MnOMgOCaONa2OK2OCr2O3ZrO2F2O

31.689.447.0911.020.121.2334.66

4.150.57

35.401.320.602.390.120.2233.16

27.390.00

19.6810.018.136.480.061.5231.73

1.8718.41

39.290.1117.985.200.050.0037.27

0.00

21.950.0423.600.290.220.1542.18

0.04

46.401.638.144.590.0512.5525.190.11

0.67

52.420.561.703.650.1814.9025.300.28

0.64

42.980.005.502.080.049.2436.772.390.160.00

34.890.0017.910.240.126.2338.871.780.120.00

39.570.4717.193.970.0424.060.050.089.900.24

0.80

38.300.5817.393.890.0023.000.080.0810.160.24

0.88

38.890.0030.960.140.010.022.130.1626.810.00

41.670.0034.480.190.000.000.0616.238.120.00

Total 99.95 100.60 97.89 99.91 88.47 99.34 99.66 99.15 100.17 96.36 94.64 99.12 100.75

N.B.O. Ca 1 Mn 1 Mg 5 3000 cations5 4000

cations5 5000

cations 2Na 2 K 514000

Si 1 Al 52000

SiTiAlFeMnMgCaNaKCrZrOH*F

2.4330.5470.6430.6390.0080.1412.859

0.2530.0211.940

2.9540.0830.0600.1500.0080.0272.973

1.8120.0000.184

1.6250.6230.7920.4030.0050.1882.812

0.1220.743

2.9490.0061.5920.2940.0030.0003.000

0.000

0.204

1.4440.0021.8370.0150.0120.0142.986

0.002

6.224

1.7290.0460.3580.1430.0020.6971.0060.008

0.020

1.9440.0160.0740.1130.0060.8241.0050.020

0.019

1.9670.0000.2970.0800.0020.6301.8040.2100.009

1.5740.0000.9520.0090.0040.4191.8790.1600.007

5.5230.0492.8270.4630.0045.0040.0080.0201.7620.027

0.360

5.4830.0632.9340.4650.0004.9080.0120.0231.8560.027

0.400

1.0320.0000.9680.0030.0000.0010.0610.0080.908

1.0120.0000.9880.0040.0000.0000.0010.7640.252

Note: N.B.O. 5 Normalization based on.* OH 5 4*[3000 2 Si 2 Ti(41)].

the limestone raft and the larvikite are transected by veinsfrom a younger alkali syenite pluton that occurs immediate-ly east of Flekkeren. The limestone raft is separated by alarvikite screen less than 10 m thick from Sveconorwegianmetagabbros in the west. The stratigraphic position of themetasediments in the Flekkeren raft is not evident becausethe protolith was modified by the strong metamorphism.Most likely it consists of the Upper Ordovician (Caradoc-Ashgill) nodular limestones (the Steinvika formation) in thewest, shales of the Venstøp formation in the middle, andnodular limestones (the Herøya formation) in the east. Southof Lake Flekkeren the upper part of the Steinvika formationis underlain by about 350 m of Cambrian and Ordoviciansediments (Dahlgren 1998). These underlying sediments aremissing at Lake Flekkeren, and, because the distance be-tween the basement and the limestone raft is less than 10m, evidently the raft is displaced from its original position.The Flekkeren body is poorly exposed on the northern shoreof the small lake Flekkeren, and as scattered small outcropsin the boggy area north of the lake. It constitutes a lens atleast 500 m long and less than 100 m wide, and it is locallytransected by small veins injected from the host plutons. Theoriginal composition of the initial sedimentary layers (mil-limeter to centimeter thick) varied from nearly pure lime-stone (calcite) layers to marly shales. The described samplematerial was collected from loose blocks from a recent roadblasting and is limited to samples of marly limestones.


TABLE 3—Extended

Sample no.

Cuspidine

Stage IFL-B

Stage IFL-B

Baghdadite

Stage IFL-B

Stage IFL-B

Vesuvianite

Stage IIFL-D

Stage IIFL-D

Monticellite

Stage IIFL17

Stage IIFL-B

SiO2TiO2Al2O3Fe2O3MnOMgOCaONa2OK2OCr2O3ZrO2F2O

32.240.02

0.000.03

60.05

1.3210.46

32.540.02

0.000.00

59.93

1.7010.34

29.092.70

41.42

25.52

29.363.02

42.16

23.72

37.181.0316.671.850.003.3836.810.030.02

36.940.7716.931.840.043.3236.530.030.02

36.840.001.479.510.4616.2435.12

0.01

37.500.000.021.750.7022.9536.30

0.00

Total 99.34* 99.80* 98.73 98.25 96.98 96.48 99.65 99.22

N.B.O. cations 5 6000 cations 5 25 000

SiTiAlFeMnMgCaNaKCrZrOH*F

1.9890.001

0.0000.001

3.969

0.040

2.041

2.0000.001

0.0000.000

3.947

0.051

2.010

1.9840.139

3.028

0.849

1.9920.154

3.066

0.785

10.8290.1775.0540.2710.0001.1397.5150.0100.010

10.8060.1325.1560.2710.0061.1217.4890.0100.010

1.0160.0000.0480.2190.0110.6671.038

0.000

0.9980.0000.0010.0390.0160.9111.035

0.000

* Total 2 0 [ F

PETROGRAPHY AND MINERAL CHEMISTRY

The analyzed samples from Flekkeren contain morethan 60 different mineral species formed at various stagesduring the contact metamorphic history. Petrographic andmicroprobe studies were carried out by the JEOL JSM-840 electron microscope at the Department of Geologyand by the CAMECA CAMEBAX electron microprobeat the Mineralogical-Geological Museum, both at theUniversity of Oslo, using natural and synthetic standards.All analyses were run at an accelerating voltage of 15 kV.Beam currents varied from 10 to 20 nA depending on thephase to be analyzed. Peak and background intensitieswere counted for 10 s for the most abundant elements(e.g., Si) and for 20 s for elements with lower concentra-tions. Mineral abbreviations, end-member compositions,and mode of occurrence are given in Table 1 and ob-served mineral assemblages relevant for the phase pet-rological analysis are listed in Table 2. Representativemineral analyses are given in Tables 3 and 4.In the following, the observed minerals and mineral

assemblages are related to four different stages (I–IV) inthe contact metamorphic evolution. This does not implythat we assume that the metamorphic process has takenplace as a series of events, but is an attempt to separateminerals and assemblages that seem to be related in spaceand time.

Stage IStage I represents the peak metamorphic conditions. At

this stage, calcite in carbonate-rich rocks is typicallycoarse grained (grain size of several millimeters) and dis-plays polygonal foam textures. Carbonate-poor rocks aremore fine grained (typical grain diameter 0.01–1 mm),and the silicate minerals commonly display textures in-dicative of very short transport distances during growth(such as complex graphic intergrowths). Carbonate-bear-ing lithologies contain wollastonite, kalsilite, nepheline,melanite garnet, perovskite, cuspidine, baghdadite, pyr-rhotite (Fe/S 5 0.84–0.93), occasional alabandite, one ormore apatite-group minerals, and occasional graphite asscattered, apparently corroded grains.The bulk composition of the analyzed layered samples

varies significantly on millimeter scale. Some domains con-tain the assemblage Mel1Cpx1Ph1Wo1Cc1Ks (Fig. 2a),which represents an isobaric invariant assemblage in theKAlO2-CaO-MgO-SiO2-H2O-CO2 system.Clinopyroxene commonly occurs as poikiloblasts (occa-

sionally many millimeters across) with inclusions of wol-lastonite, phlogopite, and kalsilite. A continuous range oc-curs in clinopyroxene compositions from the Stage I augites(fassaites) containing up to 10.9 wt% Al2O3 (0.5 Al apfu),2.6 wt% TiO2, and 0.75 wt% Cr2O3 to the nearly pure di-opsides formed during retrogression (Stage II; see Table 3).


TABLE 4. Representative silicate-phosphate analyses

Sampleno. FL-7 FL-7 FL-7 FL-7 FL-7 FL-7 FL-7 FL-7 FL-7 FL-7 FL-7 FL-7

SiO2P2O5CaOCe2O3La2O3Nd2O3ThO2UO2SO3Cl

19.521.6523.273.012.360.4443.583.170.100.22

19.063.4524.744.253.120.8340.572.840.200.19

18.781.9523.873.712.800.9139.412.300.000.16

18.814.2124.834.263.210.8038.582.630.330.18

20.321.1120.0512.088.012.3625.166.150.120.16

20.660.9819.4511.857.852.6826.576.090.150.18

17.466.1326.3611.207.102.2120.964.610.570.21

17.635.6125.8911.027.172.3920.564.540.650.13

18.056.1825.3013.768.752.5916.374.950.550.19

17.396.5425.6111.777.682.2517.436.410.450.16

18.684.8821.5120.6112.954.648.533.380.350.19

0.4540.6356.230.480.320.000.180.070.020.00

Total 97.31 99.24 93.90 97.85 95.52 96.46 96.82 95.58 96.69 95.69 95.72 98.38Normalization based on Ca 1 REE 1 Th 1 U 5 5000

SiPCaLaCeNdThUSCl

2.5790.1863.2940.1100.1500.0201.3100.0900.0100.050

2.4020.3723.3410.1500.2000.0401.1600.0800.0200.040

2.4650.2193.3570.1400.1800.0401.1800.0700.0000.040

2.3940.4583.3860.1500.2000.0401.1200.0700.0300.040

2.7000.1262.8540.3900.5900.1100.7600.1800.0100.040

2.7630.1122.7870.3900.5800.1300.8100.1800.0200.040

2.0610.6193.3350.3100.4800.0900.5600.1200.0500.040

2.1080.5733.3170.3200.4800.1000.5600.1200.0600.030

2.1470.6283.2230.3800.6000.1100.4400.1300.0500.040

2.0910.6723.2990.3400.5200.1000.4800.1700.0400.030

2.2570.5042.7850.5800.9100.2000.2300.0900.0300.040

0.0372.8644.9710.0100.0100.0000.0000.0000.0000.000

Melilite compositions vary continuously from almostpure akermanite to akermanite55gehlenite45. The Mg/(Mg1Fe21 ) ratio is generally high (0.96–0.98), but themost akermanite-rich compositions show more extensiveFeMg21 exchange [Mg/(Mg 1 Fe21 ) 5 0.88–0.90].Phlogopite from the Flekkeren locality is characterized

by Mg/(Mg 1 Fe) . 0.9 and considerable substitutiontoward the eastonite end-member (2.8–3.1 Al apfu whennormalized to total cations 2 K 5 14.00). F typicallysubstitutes for about 10% of the OH.Kalsilite commonly occurs intergrown with wollaston-

ite. Spectacular graphic intergrowths of kalsilite and wol-lastonite with minor nepheline resemble textures formedby spinoidal decomposition (Fig. 2b). As these phaseshardly can form by decomposition of a single pre-existingsolid phase, we speculate that this texture results fromrapid (metastable) reaction between the original potassi-um feldspar and calcite:

Kfs 1 Cc 5 Ks 1 Wo 1 CO2 (1)

Nepheline commonly occurs as irregular lamellae with-in kalsilite grains, but may also form separate grains thatfrequently are fringed by complex intergrowths of can-crinite and zeolite-group minerals (see below). Kalsilitecontains very little Na whereas nepheline contains about25 mol% kalsilite component.The Flekkeren samples contain an extreme variety of

garnets (see Table 3). Because of the presence of OH andZr in many of these garnets, the structural formula hasbeen calculated by assuming that Ca 1 Mg 1 Mn 5 3.0.For the melanites, structural formulas have also been cal-culated by normalization to a total charge of 24, but theresults are only marginally different. Three garnet typesare classified as Stage I phases. Uvarovite (Xuv # 0.9) thatpredates or dates the peak assemblage nucleated on de-

tritial chromite grains and is responsible for the stronggreen color of some hand specimens. Rare zirconian gar-net (kimzeyite) was found in veinlets through the pre-Stage I uvarovite along with the Stage I Zr phase bagh-dadite (see below) and probably grew in a locally Zr-en-riched system. Kimzeyite contains up to 18.4 wt% ZrO2

(0.7 Zr pfu) and are low in Si because of the extensiveZr61Al41Si4121Al6121 substitution. Ti- and Cr-contents arehigh: 10–11 wt% TiO2 and 1.7–2.4 wt% Cr2O3 (0.6–0.7Ti and 0.1–0.15 Cr pfu). The dominating Stage I garnetgrains are subhedral to anhedral melanite, occasionallywith a core of uvarovite. The melanite compositions showgood positive correlation between Mg, Ti, Cr, and Zr,reaching a maximum of approximately 9.4 wt% TiO2, 5.7wt% Cr2O3, and 2.0 wt% ZrO2. A good negative corre-lation exists between the Mg and Si content of these gar-nets (Fig. 3), presumably reflecting increasing Ti41-con-tent with increasing Ti-content. The uvarovitecompositions plot on the Mg vs. Ti and Mg vs. Si trendsdefined by the melanite and Stage II grandite. Ignoringthe uvarovites, there seems to be a gap in garnet com-position in the range Mg 5 0.06–0.1 apfu, separatingStage I melanites from Stage II grandite. Furthermore thisgap reflects abrupt changes in the composition of zonedgarnet with melanitic cores and granditic rims.Cuspidine [Ca4Si2O7(F,OH)2] and baghdadite (Ca3ZrSi2O9)

are additional Ca phases that occur as small (¯ 100 mm)grains in equilibrium with Cc1Wo1Mel (Fig. 4a). Bagh-dadite was originally described from a melilite skarn byAl-Hermezi et al. (1986). This occurrence of baghdadite isto our knowledge only the second reported. Baghdadite pre-sumably formed by the breakdown of original detrital zirconthrough the reaction

Zrn 1 2Cc 1 Wo 5 Bg 1 2CO2. (2)


FIGURE 2. (a) Peak metamorphic assemblage composed ofcalcite, wollastonite, clinopyroxene (fassaite), phlogopite, meli-lite, and titanian garnet (melanite). Field of view 1 mm long. (b)BSE-image showing graphic intergrowth of wollastonite and kal-silite. Note variations in the orientation of the wollastonite la-mellas along boundaries interpreted as original grain boundaries.

(c) Formation of Stage II phases tilleyite, monticellite, and gran-dite garnet along original calcite-wollastonite grain boundaries.Also note the late growth of scawtite along fractures through thetilleyite and along the tilleyite-wollastonite boundary. (d) Ras-vumite (core) and djerfisherite (rim) replacing an original pyr-rhotite grain.

Baghdadite analyses show nearly perfect stoichiomet-ric compatibility with the formula given by Al-Hermeziet al. (1986) (Table 3). In the Flekkeren samples about15% of the Zr is substituted by Ti, giving the approximateformula Ca3Zr0.85Ti0.15Si2O9. The cuspidine is close to theF end-member and contains variable Zr contents (0–1.7wt% ZrO2). The most Zr-rich samples coexist with bagh-dadite and may in fact reflect limited solid solution to-ward baghdadite, which also is a member of the wohler-ite-cuspidine group.Calcite-free, calc-silicate layers have a Stage I assem-

blage composed of wollastonite, clinopyroxene, melaniticgarnet or uvarovite, and potassium feldspar. An additionalphase is scapolite, which has an ambiguous petrographicstatus in these rocks. It sometimes occurs as millimeter-sized poikiloblasts with inclusions of the peak assemblageminerals and in many cases clearly replaces potassiumfeldspar. Scapolite growth is associated with alkali-ex-change and introduction of Cl to the metasedimentary

rocks. Accessory minerals include zircon, pentlandite,pyrrhotite, lollingite, chromite, scheelite, galena, sphal-erite, uraninite, a LREE-silicate (possibly stillwellite), anunidentified oxide of Ca, LREE, Th, U, Ti, and Nb andSi-rich apatite group minerals. Zircon is recrystallizedand remobilized during the high-grade metamorphismand occurs as euhedral porphyroblasts or as irregularmasses along the silicate grain-boundaries.Various Th and LREE-enriched silicate apatites formed

at the expense of original apatite in layers enriched inheavy accessory minerals during contact metamorphism.These can occasionally be observed as coronas aroundthe original apatite grains (Fig. 4b). Brownish, Th-rich,varieties (X1 in Table 1) are commonly associated withthorite. Thorianite is observed as a reaction product whenthe original silicate apatites break down during hydrationof metamict grains to an amorphous substance with ap-atite-like composition. Silicate phosphate analyses arelisted in Table 4. The composition space of the silicate


FIGURE 3. Cr, Ti, and Si apfu vs. Mg atoms for the varioustypes of garnet observed in the Flekkeren samples. Crosses(zoned melanite) refer to data from a single-zoned garnet grain.

FIGURE 4. BSE-images showing (a) Coexistence of bagh-dadite and the related mineral cuspidine in a matrix of meliliteand wollastonite. This system is calcite saturated. Scale bar is100 mm. (b) Th-rich silicate apatite (X1) forming as a coronaaround an original apatite grain. Scale bar 10 mm.

apatites can be largely described by the end-members ap-atite [Ca5P3O12(OH)], britholite [LREE3Ca2Si3O12(OH)],and the new species X1 [Ca3.5(Th,U)1.5Si3O12(OH)]. Minorsubstitutions occur along the vectors Cl(OH)21 (0.035 60.005 Cl pfu) and SiSP22 (0.035 6 0.015 S pfu). Silicatephosphate analyses have been normalized assuming thatCa1Th1U1La1Ce12Nd 5 5.000. Doubling of the Ndcontent was done to account for the presence of REEheavier than Nd and is based on the relative concentra-tions of the three analyzed LREE assuming a log-linearREE-cation size vs. concentration pattern. Errors intro-

duced by doing this are small in these extremelyLREE-enriched apatites. Normalization including the tet-rahedral coordinated cations is not possible as these sili-cate apatites may contain significant amounts of CO3 andB, which are not measured by the EMP. The analyzedsamples range in composition from ap17X121br62 toap6X185br9, but cluster around ap20X140-45br40-35 (Fig. 5).The U content is constant (4.5 6 1.5 wt% UO2) and doesnot vary systematically with Th content but is low (2.3–3.2 wt% UO2) in the most Th-rich samples (.1.0 Th pfu).The limited variations in the apatite content of the ana-lyzed samples indicate that the main substitution mech-anism is ThCaLREE22. However, a positive correlationexists between the Th 1 U content and the Si deficiency(defined as Th 1 2LREE 2 Si) in the tetrahedral position,which may be explained by moderate coupled Th 1 Bsubstitution for Ca and P or LREE and P [e.g., along thevector ThB(CaP)21] (Burt 1989). The total sum of theEMP analyses is low for Th-poor samples (Table 3). This


FIGURE 5. Compositional variations among apatites and sili-cate apatites showing that most analyses plot within the field ofthe new end-member species X1. Open symbols show the com-position of grains that give low totals in the electron microprobeanalyses. This is probably partly due to hydration of metamictgrains and partly due to a significant concentration of B and or C.

is probably due to a significant H2O content, possiblyresulting from hydration of metamict grains.

Stage IIThe transition from Stage I to Stage II assemblages in

calcite-saturated layers is marked by the occurrence ofmonticellite, tilleyite, vesuvianite, grandite garnet (poorin Ti and Cr, see Table 3), Al-poor diopside, and occa-sional hillebrandite. Retrogression defining the onset ofStage II occurs pervasively through the Stage I assem-blages. Tilleyite forms as a corona between calcite andwollastonite (Fig. 2c) and is in textural equilibrium withmonticellite probably forming at the expense of melilite.Tilleyite and monticellite may form through the reactions:

2Wo 1 3Cc 5 Ty 1 CO (3)2

3Cc 1 2Ak 5 Ty 1 2Mo 1 CO (4)2

Monticellite also coexists with garnet and vesuvianiteand is observed in reaction rims around melilite with wol-lastonite and calcite. In this case it may form by the fluid-absent reaction:

Ak 5 Wo 1 Mo (5)

Monticellite falls into two compositional groups, onewith Mg/(Mg 1 Fe) 5 0.67 2 0.77 and one coexistingwith tilleyite and with Mg/(Mg 1 Fe) . 0.95.Vesuvianite occasionally forms euhedral porphyro-

blasts with inclusions of wollastonite and melilite or oc-curs as millimeter-thick layers alternating with granditicgarnet layers. The Mg/(Mg 1 Fe) ratios for four samplesare within the range 0.7–0.8 (all Fe assumed divalent).

Grossular-rich Stage II garnet (Xgr . 0.6), typically isobserved as rims on Stage I melanites or as alterationrims between other Stage I minerals but sometimes aseuhedral poikiloblasts with numerous calcite and wollas-tonite inclusions. On the basis of the textural observationsthe most likely candidates for reactions responsible forvesuvianite and garnet formation are:

3Wo 1 2Ak 1 2Ge 1 CO 1 2H O 5 vs 1 Cc (6)2 2

2Wo 1 Ge 1 CO 5 Gr 1 Cc (7)2

Both these reactions include breakdown of the gehlen-ite (ge 5 Ca2Al2SiO7) component of the Stage I melilites.Al-poor diopside formed as a rim or neoblast aroundpeak-metamorphic augitic pyroxene (fassaite) and is com-monly associated with grandite garnet. The formation ofthese phases may be described by the breakdown of thecalcium-tschermaks (cats) component of the pyroxenethrough the continuous equilibrium:

Cats 1 Wo 5 Gr (8)

The Stage I pyrrhotite is frequently surrounded by a co-rona of rasvumite, which again may be replaced by djer-fisherite (Fig. 2d). The formation of these sulfides is spa-tially associated with formation of other Stage IIminerals. Djerfisherite may also replace pyrrhotite with-out any intervening stage of rasvumite growth. Rasvumiteand djerfisherite are rare potassium iron sulfides previ-ously reported from alkaline intrusions (e.g., Czamanskeet al. 1979) and from enstatite chondrites (Heide andWlotzka 1995). These phases are interpreted to haveformed by reactions such as (stoichiometry dependent onthe pyrrhotite composition):

Wo 1 2Cc 1 2Ks 1 6Pyh 5 2Rs 1 Gr 1 2CO (9)2

3Wo 1 6Cc 1 6Ks 1 25Pyh 5 Dj 1 3Gr 1 6CO (10)2

or by the addition of K from externally derived fluids. Ras-vumite is nearly stoichiometric KFe2S3. Djerfisherite hasbeen described as K6Cu(Fe,Ni)12S14, as K6Na(Fe,Ni)24S26Cl,and as K6(Fe,Ni)25S26Cl (cf. Czamanske et al. 1979). Theanalyzed samples fall into two populations, one Cl-bearingand one essentially Cl-free. Ni was only analyzed in theCl-poor samples, which contain 4–5 Ni apfu. Furthermorethe X-ray spectrum indicated significant amounts of Co andCu.

Stage IIIExamining the samples from Flekkeren using UV light

has revealed diffuse veinlets that contain highly fluores-cent sodalite in a symplectite (grain size 10–50 mm) con-taining kalsilite, potassium felspar, and calcite (Fig. 6a).Wollastonite, clinopyroxene, and titanite occur in thissymplectitic matrix. These assemblages are referred to asStage III assemblages below. Clinopyroxene grains inthese veinlets are extensively zoned and show sharp com-positional changes from a green acmite rich core (0.17–0.18 Na pfu), to an intermediate zone (0.071–0.076 Na


FIGURE 6. BSE-images showing (a) Symplectite composedof the Stage III phases kalsilite, K-feldspar, sodalite, nepheline,and calcite between larger wollastonite and calcite grains. Scalebar is 100 mm. (b) Complex intergrowths of guiseppettite, phil-lipsite, thomsonite, calcite, and an unknown zeolite-like mineral(X2) replacing an original nepheline grain. Nepheline can stillbe found as inclusions in the fassaitic clinopyroxene porphyro-blast in the upper right corner of the BSE-image. Field of viewabout 0.7 mm long.

pfu) to a colorless rim (0.048–0.056 Na pfu), which is inoptical discontinuity with the core.

Stage IVStage IV includes minerals occurring as late products

of retrogressive reactions affecting Stage I–III mineralsand in most cases their occurrence is confined to localsubdomains in the samples. Many of these minerals arevery fine grained and care was taken to avoid inter-growths and poorly defined grains during electron micro-probe analysis. All phases described below show uniformextinction in cross-polarized light and are homogenous atthe 1 mm scale. Microprobe analyses of different grainshave given consistent results.Alteration products after matrix nepheline grains are com-

plexly intergrown (Fig. 6b) and include several phases in-cluding the sulfate-bearing cancrinite-group mineral giusep-

pettite [representative formula: K2.5Na4.5Ca0.9Al6Si6O24(SO4)1.3-Cl0.3] and the zeolites thomsonite (nearly ideal end-membercomposition) and phillipsite (representative formula: KNa0.1-Ca1.1Al3.3Si4.7O16·6H2O). An additional, and to our knowledgenot yet described, zeolite-like mineral with the approximateformula Na2.3Ca0.8Al4Si4O16 · X H2O (X2 in Table 1) is alsopresent. Low-grade tectosilicates formed along with calciteduring wollastonite consumption. Scawtite occurs as fringesalong grain boundaries and fractures through tilleyite (Fig.2c).Stage IV hydrogrossular is invariably Al-rich (.1.8 Al

pfu) and occurs along grain boundaries or fracturesthrough Stage II grandite garnet. It contains 1.45–1.7 Sipfu, (5.2–6.2 OH-groups pfu) and are classified as hib-schites and katoites (cf. Passaglia and Rinaldi 1984).Fractures through large wollastonite grains (100 mm to

1 mm long) in the Stage III veinlets invariably contain acolorless unidentified phase (X3 in Table 1) that may rep-resent a new mineral species. The formula inferred fromelectron microprobe analyses is [Na2Ca4Si6O15(OH)4]. An-other unknown Ca,Si-rich mineral species (X4 in Table1) was found as fracture fillings through clinopyroxenegrains along with mineral X3. Tobermorite and smectite(saponite) represent the latest and very fine-grained alter-ation products from calc-silicates observed in these sam-ples, and opal was identified as a coating on late fractures.

PETROLOGYIn the following, we attempt to determine the varying

temperature and fluid composition conditions during themetamorphic history of these rocks through an analysisof heterogeneous phase equilibria. As the activity-com-position relations for Ti-rich garnets and vesuvianite sol-id-solutions are essentially unknown, the Flekkeren as-semblages were analyzed within the composition spaceKAlO2-CaO-MgO-SiO2-H2O-CO2. Pressure estimates onthe basis of stratigraphic reconstructions (Vogt 1907;Dahlgren, unpublished results) and fluid inclusion work(Olsen and Griffin 1984; Andersen 1990) are in the range700–1000 bar. Figure 7 and 8 show isobaric T-XCO2 andT-X(O) diagrams for this end-member system calculatedat 1000 bar, applying the VERTEX program of Connolly(1990) and a revised version of the thermodynamic da-tabase of Holland and Powell (1990). For binary H2O-CO2 fluids, the CORK equation of state (Holland andPowell 1991) was used. For graphite saturated assem-blages, we applied the GCOH and GCOHS equations ofstate of Connolly and Cesare (1993).As the petrographic status of the observed graphite in

these rocks is uncertain, phase diagrams were calculatedby assuming binary H2O-CO2 fluids or assuming graphitesaturation. The observed Stage I mineral assemblageCc-Wo-Ph-Cpx-Ak-Ks-fluid (Table 2) corresponds toequilibration near the isobaric Invariant Point 8 in Figure7a, located at T ¯ 870 8C, X ¯ 0.38. The diopside-CO2

absent reaction is stable on the low temperature side ofInvariant Point 8, and thus there is a net production ofpyroxene in this point (cf. Carmichael 1991). This is con-


FIGURE 7. Simplified T-X diagrams calculated at P 5 1000CO2

bar for the pure KAlO2-CaO-MgO-SiO2-H2O-CO2 system ex-cluding equilibria that are irrelevant with the present variabilityin bulk composition (i.e., equilibria involving forsterite, Ca-poorpyroxene, and amphibole). Reactions are always written with thehigh-temperature assemblage to the right of the equality sign. (a)T-X diagram for the region relevant to the Stage I assemblage.CO2

(b) T-X diagram showing in more detail equilibria relevant forCO2

the retrograde Stage II and III assemblage. The regions of pos-sible stability of these assemblages are limited by the reactionsshown by bold lines. Abbreviations not included in Table 1: Sa5 sanidine; Lc 5 leucite; Q 5 quartz, Bq 5 beta quartz; Spu 5spurrite.

FIGURE 8. T-X(O) projections calculated at P 5 1000 bar forthe pure KAlO2-CaO-MgO-SiO2-C-O-H system with graphitesaturation.

sistent with the observed poikiloblastic texture of the py-roxene in these rocks. Uncertainties in the position ofInvariant Point 8 mainly stems from possible errors in theactivity corrections of the phase components Di, Ak, andPh. The univariant reactions intersect at acute angles andthus minor errors in the relevant activities give large ef-fects on the calculated X -value. Test experiments withCO2

several activity models and with varying compositions ofthe phases involved suggest a maximum range for T of820–870 8C and X of 0.2–0.4. Lowering the pressureCO2

to 700 bar gives a reduction in T by about 25 8C and anincrease in X by 0.03 units for the end-member system.CO2

The inferred peak temperature is slightly higher than theinferred solidus (approximately 800 8C) for a composi-tionally similar monzonite intrusion in the Sande Caul-dron further north in the Oslo Rift (Andersen 1984).The stability of the retrograde assemblage Ty-Mo-Wo-

Cc, can be inferred from Figure 7b. Wollastonite 1 mon-ticellite puts an upper limit on the metamorphic temper-ature of about 710 8C, and the presence of tilleyite at suchtemperatures requires that X # 0.016. Coexistence ofCO2

monticellite and kalsilite restricts the stability of the StageII assemblage to a region near Invariant Point 10 at T ¯710 8C, X ¯ 0.005. Hence, these diagrams suggest aCO2

change from 20–40% CO2 in the metamorphic fluid dur-ing peak metamorphic conditions to almost C-free fluidsduring formation of the Stage II assemblage.The Stage III assemblage Ks-Kfs-Wo-Cc constrain the

maximum temperature to about 550 8C and the compo-sition of a binary H2O-CO2fluid to X , 0.016 (Fig. 7b).CO2

At graphite saturation, the assumption of a binary H2O-CO2 fluid is incorrect. Figure 8 shows a T-X(O) diagramconstructed for the end-member system at 1000 bars. A


FIGURE 9. Schematic temperature-time evolution for theFlekkeren megaxenolith superimposed on a T-t curve believedto have a shape typical for regions near the top of a coolingintrusive. Heating during the prograde stage lead to devolatili-zation reactions causing a relatively high C content in the porefluids at Stage I. This was followed by an infiltration controlledincrease in the H2O content through influx of magmatic (upwarddirected arrows) and later of meteoric fluids (downward directedarrows) during Stage IV.

projection of COH-fluid compositions through graphiteleads to a binary representation of the fluid in the H-Osystem where X(O) [5 O/(O 1 H)] varies from 0 to 1 asthe fluid composition changes from the C-H join to theC-O join and is ƒ for pure H2O (Connolly 1995). X(O)is proportional to the oxygen fugacity in GCOH fluids.The estimates of the peak metamorphic temperature (nearInvariant Point 8) is not sensitive to the absence or pres-ence of graphite. In the T-X(O) diagram Invariant Point8 is located at T ¯ 840 8C, X(O) ¯ 0.4. However, thefluid composition corresponding to this X(O) value de-viates significantly from a binary H2O-CO2fluid and con-tains about 14 mol% CH4. Adding S as a fluid componentand buffering the fluid composition by pyrrhotite with aFe/S ratio of 0.93 (the most Fe-rich pyrrhotite measuredin the Flekkeren samples), the resulting fluid compositionis 40% H2O, 30% CO2, 10% H2S, 10% CH4, 5% CO, and5% H2.Phase relations, calculated including Al2O3 as an ad-

ditional thermodynamic component, are consistent withthe formation of garnet and vesuvianite by Reactions 6and 7 at temperatures below peak temperature. However,the exact positions of these equilibria are associated withconsiderable uncertainty because of uncertainties in ther-modynamic data and solid solution models for vesuvi-anite and melanitic garnets.Several solid-solid continuous equilibria can be written

among the possible phase components to constrain meta-morphic temperatures from fluid-independent reactions,but none have proved useful because of small entropychanges and correspondingly great sensitivity to activity-concentration models and uncertainties in available ther-modynamic data. However, several useful observationscan be made from variations in the mineral composition.The extensive Al substitution in the pyroxene reflects thehigh temperature of metamorphism and is buffered byReaction 8, which is characterized by a positive changein both entropy and volume. This equilibrium is probablyone of the most useful continuous reactions in recordingrelative temperature variations within and among contactaureoles in calcareous rocks where the fluid compositionmay be variable. Likewise, the Mg content of the garnetis likely to correlate positively with the metamorphic tem-perature because almost every Fe-Mg exchange equilib-rium between garnet and other ferromagnesian mineralsproduces relative Fe enrichment in the garnet. As the Alcontent in pyroxene and Mg content in garnet correlatepositively with the temperature of equilibration, the Ticontents of both pyroxene and garnet also correlate pos-itively with temperature. Cr shows the same behavior asTi, but the Cr content of the garnet is strongly affectedby local Cr sources in its environment and thus showsmuch more scatter than does the Ti-content.

DISCUSSIONThe mineral assemblages from the Flekkeren locality

are unusual for metamorphic rocks, mainly because rocksof such bulk composition rarely experience such high

temperatures in combination with low pressures. Many ofthe observed phases are indeed more common in someSi-poor alkaline intrusions such as ijolite-series rocks andultramafic lamprophyres (e.g., Sørensen 1979; Dahlgren1987).The large number of minerals present in these rocks is

clearly related to metamorphic reactions taking placethroughout the metamorphic history, and these reactionsare partly driven by the interactions with externally de-rived magmatic and meteoric fluids. The metamorphichistory inferred from the petrology is schematically illus-trated in Figure 9 and is superimposed on a T-t curve thatis believed to be typical for regions near the top of acooling intrusion (cf. Hanson 1995; Fig. 8). We believethat the petrologic analysis gives a reasonably reliableindication of the conditions of formation for the observedStage I mineral assemblage, indicating a relatively CO2-rich (or at least C-rich) pore fluid during peak metamor-phism. However, in the present geological setting, wherea megaxenolith of impure limestone fell into a partly mol-ten intrusion, we cannot expect the rapidly heating meta-sedimentary rock to follow a progressive reaction pathpredicted from an equilibrium phase diagram. The ob-served graphic intergrowths between kalsilite and wollas-tonite (Fig. 2b) reflect short transport distances during thereaction and are more readily explained by rapid non-progressive metamorphic breakdown of original potassi-um feldspar and calcite than by progressive reactionsalong a buffered T-X path that would have lead to theCO2

production of leucite as an intermediate but non-observedspecies.During rapid heating of a calcareous rock embedded in

a partly crystallized intrusion, the fluid pressure withinthe xenolith rapidly rose through devolatilization reac-


tions to reach or even exceed the fluid pressure in thesurrounding intrusive. This hindered major fluid infiltra-tion into the xenolith in agreement with available O iso-tope results obtained from the carbonates. At this stage(Stage I) the fluid composition in the xenolith was eitherbuffered by decarbonation reactions or, in the non-equi-librium case, controlled by the stochiometry of the on-going reactions. This caused a rise in the CO2 content(and CH4 content at reducing conditions) of the meta-morphic fluids toward peak temperature conditions. Whencooling started, the total solid volume of the xenolith wasconsiderably reduced because of volatile loss (occasion-ally exceeding 10% of the initial volume, depending onthe initial carbonate-silicate ratio). Unless this volumeloss was accommodated by compaction, which may havebeen a relatively slow process at the time-scales of theheating of these contact metamorphic rocks, the reductionin total solid volume made the xenolith more permeableand susceptible to pervasive influx of external fluids.Thus, the progressive stage of metamorphism leading tothe Stage I assemblage is strikingly similar to the indus-trial production of highly porous cement clinker by rapidheating at low P of a clay-limestone mixture, with a bulkcomposition similar to some of the Flekkeren samples.This model explains why the post-peak (Stage II–IV)

assemblage reflects C-poor, H2O-dominated magmatic ormeteoric fluids. The pervasive formation of the Stage IIassemblage is in many respects in contrast to the morecommon situation during retrogressive metamorphismwhere volatilization reactions usually depend on fluid ac-cess around fractures, faults, or other highly permeablezones generated after the formation of the peakassemblage.However, Stage III retrogression is associated with

veins or veinlets. This is consistent with retrogression atlower temperature where thermal contraction may haveled to fracturing and more localized access of magmaticand possibly meteoric fluids to generate veins in the meta-sedimentary rocks. The growth of sodalite during StageIII implies an addition of Cl to the system. Influx ofbrines through brittle fractures is a common feature incontact aureoles in the Oslo rift during cooling from peakmetamorphic conditions and is ascribed to the late releaseof saline magmatic liquids that were relatively immobileduring peak metamorphism because of the low perme-ability of the country rocks at this stage (Jamtveit andAndersen 1993).Stage IV marks the onset of the growth of very fine

grained low-temperature phases such as zeolites. At thisstage the magmatic system was completely crystallizedand the most likely source for externally derived fluid ismeteoric water.

ACKNOWLEDGMENTSDiscussions and assistance from Odd Nilsen and Henrik Svensen at the

Department of Geology, University of Oslo, and the supporting engage-ment of Alf Olav Larsen at Norsk Hydro, Research Centre, Porsgrunn, aregreatly acknowledged. Very thorough reviews by an anonymous referee

and by Tom Hoisch significantly improved this paper. This work was sup-ported by grant no. 110578/410 from the Norwegian Research Council.

REFERENCES CITEDAl-Hermezi, H.M., McKie, D., and Hall, A.J. (1986) Baghdadite, a newcalcium zirconium silicate mineral from Iraq. Mineralogical Magazine,50, 119–123.

Andersen, T. (1984) Crystallization history of a Permian composite mon-zonite-alkali syenite pluton in the Sande Cauldron, Oslo rift, S. Norway.Lithos 17, 153–170.

———(1990) Melt-mineral-fluid interaction in peralkaline silicic intrusions inthe Oslo Rift, Southeast Norway IV: Fluid inclusions in the Sande nord-markite. Norges Geologiske Undersøkelser Bulletin, 317, 41–54.

Bjørlykke, K. (1974) Depositional history and geochemical compositionof Lower Palaeozoic epicontinental sediments from the Oslo Region.Norges Geologiske Undersøkelse Bulletin, 305, 1–81.

Burt, D.M. (1989) Compositional and phase relations among rare earthelement minerals. In Mineralogical Society of America Reviews in Min-eralogy, 21, 259–307.

Carmichael, D.M. (1991) Univariant mixed-volatile reactions: Pressure-temperature phase diagrams and reaction isograds. Canadian Mineral-ogist, 29, 741–754.

Connolly, J.A.D. (1990) Multivariable phase diagrams: an algorithm basedon generalized thermodynamics. American Journal of Science, 290,666–718.

———(1995) Phase diagram methods for graphitic rocks and applicationto the system C-O-H-FeO-TiO2-SiO2. Contributions to Mineralogy andPetrology, 119, 94–116.

Connolly, J.A.D. and Cesare, B. (1993) C-O-H-S fluid composition andoxygen fugacity in graphitic metapelites. Journal of metamorphic ge-ology, 11, 379–388.

Czamanske, G.K., Erd, R.C., Sokolova, M.N., Dobrovol’skaya, M.G., andDmitrieva, M.T. (1979) New data on rasvumite and djerfisherite. Amer-ican Mineralogist, 64, 776–778.

Dahlgren, S. (1987) The satellitic intrusions in the Fen carbonatite-alkalinerock province, Telemark, Southeastern Norway. Ph.D. thesis, 350 p.University of Oslo.

(1998) Map sheet 1713 IV Nordagutu, bedrock geology. Scale 1:50000. Geological Survey of Norway, in press.

Goldschmidt, V.M. (1911) Die Kontaktmetamorphose im Kristianiagebiet. Vi-tenskaps-selskapets Skrifter. I., Matematisk Naturvitenskapelig Klasse,No. 1, 483 p.

Hanson, R.B. (1995) The hydrodynamics of contact metamorphism. Geo-logical Society of America Bulletin, 107, 595–611.

Heide, F. and Wlotzka, F. (1995) Meteorites, 231 p. Springer-Verlag,Heidelberg.

Holland, T.J.B. and Powell, R. (1990) An enlarged and updated internallyconsistent dataset with uncertainties and correlations: the system K2O-Na2O-Cao-MgO-FeO-Fe2O3-Al2O3-TiO2-SiO2-C-H2-O2. Journal ofMetamorphic Geology, 8, 89–124.

———(1991) A compensated-Redlich-Kwong (CORK) equation of statefor volumes and fugacities of CO2 and H2O in the range 1 bar to 50kbar and 100–1600 C. Contributions to Mineralogy and Petrology, 109,265–273.

Inderst, B. (1987) Mischkristalle und fluide phase (H2O-CO2) bei hoch-temperature dekarbonatisierungsreaktionen in zwei thermalaureolen.PhD thesis, 215 p. Ludwig-Maximillans-University, Munchen.

Jamtveit, B. and Andersen, T. (1993) Contact metamorphism of layeredshale-carbonate sequences in the Oslo rift: III. The nature of skarn-forming fluids. Economic Geology, 88, 830–849.

Jamtveit, B., Bucher-Nurminen, K., and Stijfhoorn, D.E. (1992a) Contactmetamorphism of layered shale-carbonate sequences in the Oslo rift: I.Buffering,infiltration and the mechanisms of mass transport. Journal ofPetrology, 33, 377–422.

Jamtveit, B., Grorud, H.F., and Bucher-Nurminen, K. (1992b) Contactmetamorphism of layered shale-carbonate sequences in the Oslo rift: II.Migration of isotopic and reaction fronts around cooling plutons. Earthand Planetary Science Letters, 114, 131–148.

Jamtveit, B., Grorud, H.F., and Ragnarsdottir, K.V. (1997) Flow and trans-port during contact metamorphism and hydrothermal activity: Examples


from the Oslo rift. In B. Jamtveit and B.W.D. Yardley, Eds., Fluid flowand transport in rocks: Mechanisms and effects, p. 57–82. Chapmanand Hall, London.

Joesten, R. (1983) Grain growth and grain-boundary diffusion in quartzfrom the Christmas Mountains (Texas) contact aureole. American Jour-nal of Science, 238-A (Orville Volume), 233–254.

Olsen, K.I. and Griffin, W.L. (1984) Fluid inclusion studies of the Dram-men granite, Oslo Paleorift, Norway: I. Microthermometry. Contribu-tions to Mineralogy and Petrology, 87, 1–14.

Passaglia, E. and Rinaldi, R. (1984) Katoite, a new member of theCa3Al2(SiO4) 3-Ca3Al2(OH)12 series and a new nomenclature for the hy-drogrossular group of minerals. Bulletin Mineralogy, 107, 605–618.

Pedersen, L.E., Heaman, L.M., and Holm, P.M. (1995) Further constraintson the temporal evolution of the Oslo Rift from precise U-Pb zircondating in the Siljan-Skrim area. Lithos, 34, 301–315.

Sørensen, H., Ed. (1979) The alkaline rocks, 622 p. Wiley, New York.Sundvoll, B., Neumann, E.R., Larsen, B.T., and Tuen, E. (1990) Age re

lations among Oslo Rift magmatic rocks: implications for tectonic andmagmatic modelling. Tectonophysics, 178, 67–87.

Svensen, H. (1996) A phase petrological analysis of contact metamorphiclower palaeozoic sedimentary rocks from the Oslo graben, Norway.Ph.D. thesis, 136 p. University of Oslo.

Svensen, H. and Jamtveit, B. (1998) Contact metamorphism of shales andcarbonates from the Grua area, the Oslo rift, Norway: A phase petro-logical study. Norsk Geologisk Tidsskrift, in press.

Tracy, R.J. and Frost, B.R. (1991) Phase equilibria and thermobarometryof calcareous, ultramafic and mafic rocks, and iron formations. In Min-eralogical Society of America Reviews in Mineralogy, 26, 207–289.

Vogt, J.H.L. (1907) Uber die Erzgange zu Traag in Bamble, Norwegen.Zeitschrift zur Praktische Geologie, 15, 210–216.

Williams-Jones, A.E. (1981) Thermal metamorphism of siliceous lime-stone in the aureole of Mount Royal, Quebec. American Journal ofScience, 281, 673–696.

MANUSCRIPT RECEIVED NOVEMBER 20, 1996MANUSCRIPT ACCEPTED JULY 23, 1997

full text (368 k)

Documents