American Mineralogist, Volume 74, pages 573-585, 1989 Hydrothermal orthoamphibole-bearing assemblages from the GAsborn area, West Bergslagen, central Sweden Annxo H. DavrNr,q'N Institute of Earth Sciences, Free University, de Boelelaan 1085, 1081 HV Amsterdam, The Netherlands Ansrnlcr In the Gisborn area, West Bergslagen, central Sweden,two mineral assemblages are found in conduit zonesfor hydrothermal fluids from which manganiferous iron ores were deposited:(l) an older assemblage consistingof quartz, albite, phengite,chlorite, andZn- and Mn-bearing magnetite and (2) a younger, overprinting assemblage, consistingof bio- tite, cordierite, orthoamphiboles, cummingtonite, Zn- and Mn-free magnetite, hercynite, and corundum. The chemistry of biotite and coexisting orthoamphiboles that have de- veloped along hydrothermal veins associated with a high-level anorogenic (A-type) granite varies with distancefrom the veins. This suggests that the replacement of the older assem- blage by the younger took place under influence ofhydrothermal fluids expelledfrom the verns. The orthoamphibole compositions span the entire field from anthophyllite to sodium gedrite, indicating that they were formed at temperatures above the crest of the solvus of the anthophyllite-gedritemiscibility gap. The maximum temperatureand lithostatic pres- sure during the formation of the hydrothermal vein system are estimated at 560 .C and 1.0 kbar. Orthoamphibole and coexisting biotite in the hydrothermal veins crystallized after quarrz,plagioclase, and cordierite, suggesting that they were formed at a temperature of < 560 oC and implying that the crestof the solvus of the anthophyllite-gedrite miscibility gap at 1.0 kbar must also lie below this temperature. The relative fugacity of HF (/{r" : fr"/.frrJ varies between veins from l0-3 s3 to l0-4 5E and the relative fugacity of HCI U'ro, i.e., fr"r/f.ro) vaies from l0-1 68 to l0-2 60. In and along eachhydrothermalvein,f'r, is relatively constant,and the amount of F incorporated in coexisting orthoamphibole and biotite is controlled by their respective Mg and Fe contents. In and along each vein, /fi., decreases when going from the vein outward, and X* [: Mg/(Me + Fe'?*)] values of coexistingorthoamphibole and biotite show a negative correlation with their respective Cl contents,suggesting that along each vein, variations in /fi., and Xr, mutually influence one another The Mg and Fe contents of orthoamphibole and biotite appear to be controlled by the differences in fL, and fL, between veins. Ba and Ti are preferentially incorporated into biotite and show a positive correlation with Cl. Mn is preferentially incorporated into the orthoamphiboles, but does not show a correlation with either Cl or F. INrnolucrroN In the GAsbornarea,West Bergslagen, central Sweden, the high-level Ostra Hdjden granite(Damman, 1988a)- which is assumed (Damman, in preparation) to belong to the group ofolder granites ofthe Bergslagen district (Aberg et al., 1983a, 1983b; Oen et al.,1984; Baker,1985; Bill- strdm et al., 1988)-is intruded into a pile of silicic meta- volcanics with intercalated marbles, metacherts, meta- morphosed mafic lavas, and a metamorphosed manganiferousiron ore horizon (Fig. l). This volcano- sedimentary succession belongs to the Upper Leptite- Hiilleflinta and SlateGroup of the Bergslagen supracrust- al series (Oen et al., 1982). A hydrothermal vein system associated with the gran- ite is characterized by an estimated maximum tempera- 0003-004x/89/05064573$02.00 ture (T) of 560 "C, lithostatic pressure(P) of approxi- mately 1.0 kbar, and a maximum fluid pressurein the veins (Pr) of 4 kbar (at T: 560 "C). With decreasing I, P, rapidly drops to values below P' (Damman, manu- script). The hydrothermal vein system consistspredom- inantly of quartz + feldspar veins, some containing an- dalusite, sekaninaite (Fe-cordierite), biotite, muscovite, fluorite, and accessory oxides and sulfides.In one vein, sekaninaite and andalusiteconstitute over 90 volo/o ofthe vein, which furthermore contains minor plagioclase, to- paz, sillimanite, qtartz, subsilicic sodium gedrite,biotite, and accessory oxidesand sulfides (Damman, 1988a). Under influence of the hydrothermal activity, marbles and included metavolcaniteand metachertlenses (Ic, Fig. 1) are locally altered into garnet + pyroxene skarnscon- taining minor calcic amphibole, epidote, phlogopite, 573
Transcript
American Mineralogist, Volume 74, pages 573-585, 1989
Hydrothermal orthoamphibole-bearing assemblages from the GAsborn area,West Bergslagen, central Sweden
Annxo H. DavrNr,q'NInstitute of Earth Sciences, Free University, de Boelelaan 1085, 1081 HV Amsterdam, The Netherlands
Ansrnlcr
In the Gisborn area, West Bergslagen, central Sweden, two mineral assemblages arefound in conduit zones for hydrothermal fluids from which manganiferous iron ores weredeposited: (l) an older assemblage consisting of quartz, albite, phengite, chlorite, andZn-and Mn-bearing magnetite and (2) a younger, overprinting assemblage, consisting of bio-tite, cordierite, orthoamphiboles, cummingtonite, Zn- and Mn-free magnetite, hercynite,and corundum. The chemistry of biotite and coexisting orthoamphiboles that have de-veloped along hydrothermal veins associated with a high-level anorogenic (A-type) granitevaries with distance from the veins. This suggests that the replacement of the older assem-blage by the younger took place under influence ofhydrothermal fluids expelled from theverns.
The orthoamphibole compositions span the entire field from anthophyllite to sodiumgedrite, indicating that they were formed at temperatures above the crest of the solvus ofthe anthophyllite-gedrite miscibility gap. The maximum temperature and lithostatic pres-sure during the formation of the hydrothermal vein system are estimated at 560 .C and1.0 kbar. Orthoamphibole and coexisting biotite in the hydrothermal veins crystallizedafter quarrz, plagioclase, and cordierite, suggesting that they were formed at a temperatureof < 560 oC and implying that the crest of the solvus of the anthophyllite-gedrite miscibilitygap at 1.0 kbar must also lie below this temperature.
The relative fugacity of HF (/{r" : fr"/.frrJ varies between veins from l0-3 s3 to l0-4 5E
and the relative fugacity of HCI U'ro, i.e., fr"r/f.ro) vaies from l0-1 68 to l0-2 60. In andalong each hydrothermalvein,f'r, is relatively constant, and the amount of F incorporatedin coexisting orthoamphibole and biotite is controlled by their respective Mg and Fecontents. In and along each vein, /fi., decreases when going from the vein outward, andX* [: Mg/(Me + Fe'?*)] values of coexisting orthoamphibole and biotite show a negativecorrelation with their respective Cl contents, suggesting that along each vein, variations in
/fi., and Xr, mutually influence one anotherThe Mg and Fe contents of orthoamphibole and biotite appear to be controlled by the
differences in fL, and fL, between veins. Ba and Ti are preferentially incorporated intobiotite and show a positive correlation with Cl. Mn is preferentially incorporated into theorthoamphiboles, but does not show a correlation with either Cl or F.
INrnolucrroN
In the GAsborn area, West Bergslagen, central Sweden,the high-level Ostra Hdjden granite (Damman, 1988a)-which is assumed (Damman, in preparation) to belong tothe group ofolder granites ofthe Bergslagen district (Aberget al., 1983a, 1983b; Oen et al.,1984; Baker, 1985; Bil l-strdm et al., 1988)-is intruded into a pile of silicic meta-volcanics with intercalated marbles, metacherts, meta-morphosed mafic lavas, and a metamorphosedmanganiferous iron ore horizon (Fig. l). This volcano-sedimentary succession belongs to the Upper Leptite-Hiilleflinta and Slate Group of the Bergslagen supracrust-al series (Oen et al., 1982).
A hydrothermal vein system associated with the gran-ite is characterized by an estimated maximum tempera-
0003-004x/89/05064573$02.00
ture (T) of 560 "C, lithostatic pressure (P) of approxi-mately 1.0 kbar, and a maximum fluid pressure in theveins (Pr) of 4 kbar (at T: 560 "C). With decreasing I,P, rapidly drops to values below P' (Damman, manu-script). The hydrothermal vein system consists predom-inantly of quartz + feldspar veins, some containing an-dalusite, sekaninaite (Fe-cordierite), biotite, muscovite,fluorite, and accessory oxides and sulfides. In one vein,sekaninaite and andalusite constitute over 90 volo/o ofthevein, which furthermore contains minor plagioclase, to-paz, sillimanite, qtartz, subsilicic sodium gedrite, biotite,and accessory oxides and sulfides (Damman, 1988a).
Under influence of the hydrothermal activity, marblesand included metavolcanite and metachert lenses (Ic, Fig.1) are locally altered into garnet + pyroxene skarns con-taining minor calcic amphibole, epidote, phlogopite,
573
574
Fig. 1. A model for skarn formation and related processes inthe GAsborn area. I-deposition of the GAsborn manganiferousiron ore horizon, consisting of (Damman, 1988b) (Ia) two feederzones, containing Mn-poor iron ore-bearing magnesian meta-somatically altered metavolcanics; (Ib) Mn-poor iron ore-bear-ing metavolcanics; (Ic) Mn-poor iron ore-bearing marbles; (Id)manganiferous iron ore-bearing metavolcanics and metacherts,followed by deposition of metavolcanics and marbles overlyingthe manganiferous iron ore horizon (II). il-intrusion of theOstra H<ijden granite and subsequent formation of calc-silicatereaction skarns in marbles (IIc; Damman, in preparation) andmanganese silicate reaction skarns in manganiferous rron ore-
bearing metavolcanics and metacherts (IId; Damman, in prep-aration). IIa and IIb are not labeled because nothing happenedduring event II in the conduit zones or metavolcanics, respec-tively. Ill-emplacement of the Ostra Hrijden granite and sub-sequent (IIIa) formation of orthoamphibole, cordierite, biotite,and cummingtonite in the Mg-enriched conduit zones for fluidsfrom which the manganiferous iron ores were deposited (thispaper); (IIIb) biotitization of metavolcanics; (IIIc; Damman,unpublished manuscript) formation of metasomatic infiltrationskarns in marbles and manganiferous iron ore-bearing meta-volcanics and metacherts (IIId; Damman, unpublished manu-script).
fluorite, scheelite, and sulfides and calcic amphibole +biotite * orthoamphibole schist (IIc and IIIc, Fig. 1);manganiferous iron ores (Id, Fig. l) are altered into man-ganese silicate + magnetite * jacobsite skarns (IId andIIId, Fig. l); and felsic metavolcanics (Ib, Fig. l) are lo-cally altered to biotite (IIIb, Fig. l). Hydrothermal veinsassociated with the Ostra Hdjden granite also cut acrosstwo earlier conduit zones for fluids from which manganif-erous iron ores are deposited (Damman, 1988b; Ia, Fig.
TABLE 1, Counting times and errors in microprobe analyses ofminerals in the present studv
t (s) 2o (ok)
l). This paper describes the petrography and mineralchemistry of the rocks in the conduit zones. Evidence willbe presented that cordierite, biotite, orthoamphiboles,cummingtonite, hercynite, and corundum in these rockswere formed out of phengite, chlorite, quartz, and albiteby metasomatic reactions coeval with the intrusion of theOstra Hiijden granite (IIIa, Fig. l).
AN,c.LyrrcA.r, PRoCEDURES
Electron-microprobe analyses were made with a Cam-bridge Instrument Company Microscan 9 operated at anacceleration potential of 20 kV and a sample current of25 nA. Raw data were corrected with the Mark 9 onlineZAF computer program. Standards used were marialite(Cl), orthoclase (K), diopside (Ca, Si), ilmenite (Ti), rho-donite (Mn), fayalite (Fe), jadeite (Na), corundum (Al),and synthetic ZnO (Zn). F was measured at a samplecurrent of 40 nA, using fluorite as a standard. Countingtimes (s) and analytical errors (2o) are presented in Ta-ble l.
Relative (see below) HF fugacities ("firJ and relativeHCI fugacities (,fi,J of fluids with which biotite and or-thoamphibole equilibrated were calculated by using amodification of the method of Munoz (1984), who sug-gested that, from electron-microprobe analyses, the fol-
where Xr,o : (Iutx"r/0.167) x (l - X*"), Xx: | - (X*u* X*o), XM": Mg(Mg * Fe'?+), and X"" : 4 - Xo -
X.,. X, and X., are the values given for F and CI (for-mula units) in Tables 5 and 7. IV(F) and IV(Cl)-theF and Cl intercept values-are single numerical valuesdesigned to express the relative degree ofF and Cl en-richment in a mica corrected for the effects of Fe-F andMg-Cl avoidance (Munoz, 1984).
The hydrothermal solution under whose influence theorthoamphibole and biotite were formed (see below)consists almost entirely of saline water, with only mi-nor other elements. All orthoamphibole and biotite un-der discussion were formed at approximately the samepressure and temperature, implying that it can safelybe assumed thatfnro between all hydrothermal veins inand among which orthoamphibole and biotite wereformed (see below under petrographic descriptions)during orthoamphibole and biotite formation had sim-ilar values (: constant, or c), changing Equations I and3 into log (c/f",) : 2100/T + IV(F) and log (c/f".,) :
5151/T + IV(CI), respectively. Omitting the constantfrom these modified equations leads to the followingrelative fugacities of F (/fi,) and Cl (/i'J:
or
and
los.fLu: -12r00/T + IV(DI,
f,HF: l1-I2too/r+ rv(Dl (1,)
losfh.,: -l5l5r/T + IV(CI)],
f;"r: 1g-ts:srlr + rv(cDt, (3,)
For reasons discussed below, a T of 530 "C (803 K) isused in the above calculations of f'"o and f'".r.
Because Fe3* can only be determined accurately bywet-chemical analysis (Guidotti, 1984), no attempt wasmade to calculate Fe3+ in biotite. According to Munoz(1984), the absence of Fe3+ in biotite may introduce amaximal positive effor as large as 0.25 IV unit in cal-culations of IV(F) and thus most probably also in cal-culations of IV(CI), implying that all IV(F) and IV(CI)values presented in this paper should be considered asmaximal values. Analyses of orthoamphiboles dis-cussed below suggest that only a small amount of Fe3+is present in orthoamphibole. Biotite either occurs sta-bly next to orthoamphibole or was formed slightly lat-er, suggesting that the amount of Fe3+ in biotite is alsovery small.
Representative analyses of orthoamphiboles, biotite,magnetite, hercynite, and corundum are presented inTables 2,3, 4,5, 7, and 8; structural formulae are cal-culated on the basis of 23 oxygens for amphiboles,22oxygens for biotite, 4 oxygens for magnetite and her-cynite, and 3 oxygens for corundum. Total tetrahedralSi + Al is constrained to equal 8 in calculations ofbiotite and amphiboles. Several analyses of orthoam-phibole (Tables 2, 4, and 8) show a total of slightlymore than 7 cations in the M1-M4 sites. Forcing theseanalyses to 7 (Robinson et a1., 1982) requires that amaximum of about 0.0065 Fe3+ should be present inthe orthoamphiboles. According to Robinson et al.(1982), a good orthoamphibole analysis has about 0'08Na in the M4 site. A recalculation of the orthoamphi-bole analyses presented in Tables 2, 4, and 8 for 0.08Na in the M4 site suggests that, according to Robinsonet al. (1982), 0.02 (+0.0065) Fe3+ should be present inthe orthoamphiboles.
The suggestion ofRobinson et al. (1982) is opposedby Berg (1985), who suggested that Fe3+ can be neglect-ed, implying that in his calculations the only Na pres-
ent in the M4 site is that amount necessary to force thesum of the cations in the M1-M4 sites to 7 (see below).As most of the analyses presented in Tables 2,4, and8 as well as the P- ?' conditions of orthoamphibole for-mation (Table 6) are relatively close to those of Berg,his suggestions are considered valid for the Gisbornarea, and no attempt was made to recalculate Na inM4 and consequently Fe3*.
In orthoamphiboles with a total of less than 7 cationsin the M1-M4 sites, Na is transferred from the A siteto the M4 site, following the procedure of Berg (1985)'
PnrnocupHY oF THE RocKs rN THECONDUIT ZONES
The conduit zones at the derelict Gustav and JacobPers iron mines (Magnusson, 1930) occur at a 200-mdistance from one another in the Gtsborn area (Dam-
man, 1988b). Magnusson (1930) described the wall rocksof the ore bodies in these mines as a metavolcanic gneiss(in this paper referred to as metavolcanite), locally alteredinto cordierite + gedrite qluartzite. The ore bodies consistof magnetite- and spinel-rich biotite * chlorite + or-thoamphibole "skarn," crosscut by numerous quaftz +feldspar veinlets. Samples for the present study were col-lected from dumps and outcrops surrounding the mines.
The fine- to medium-grain metavolcanites around theore bodies show a granoblastic texture and consist ofal-bite (35-50 volo/o), quartz (30-40 volVd, and microcline(5-10 volTo). Locally l0-15 volo/o biotite, cummingtonitereplaced by calcic amphibole, cordierite, andalusite, andopaque minerals are also present. Biotite, cummington-ite, calcic amphibole, cordierite, and andalusite occur asup to 400-pm subhedral poikiloblasts, enclosing quartzand feldspars. The opaque minerals occur as tiny grainsintergrown with biotite.
The cordierite + gedrite quartzite (Magnusson, 1930)
Fig. 2. "(Na * K) vs. ratAl for all orthamphiboles analyzedin the GAsborn area; open symbols, this paper; filled symbols,Damman, 1988a. Edenite (Ed) substitution: Na (A) + I4rAl : !A* Si; tschermakite (Ts) substitution: I4rAl + I6rAl : Mg + Si.
shows a fine- to medium-grained matrix of quartz, albite,chlorite, phengite, magnetite, and ilmenite enclosed byaggregates of orthoamphibole, cordierite, and biotite.Quartz and albite occur as up to 50-pm anhedral crystals.Chlorite occurs as up to 200-pcm colorless, anhedral crys-tals, and phengite occurs as up to 20-pm anhedral crystalsintergrown with chlorite. Chlorite and phengite replacequartz and albite.
The orthoamphiboles, which vary in composition fromanthophyllite to sodium gedrite (Table 2 and Fig. 2) occuras up to 500-pm subhedral crystals, replacing qtrartz, al-bite, phengite, and chlorite. The most Si-rich and Al-poororthoamphiboles (analyses 1 and 2, Table 2) occur wherebiotite is absent. Orthoamphiboles coexisting with biotitebecome more Na- and Al-rich and Si-poor with increas-ing amounts of biotite and decreasing amounts of quartzand albite present. With quartz and albite present, theorthoamphibole compositions (analyses l-6, Table 2) fallin the fields of anthophyllite and gedrite (Leake, 1978,Fig. 2). Sodium gedrite (analyses 7 and 8, Table 2) inter-grown with biotite is only found where quartz and albiteare absent. Cordierite occurs as up to 500-pm anhedralcrystals, always intergrown with orthoamphibole andbiotite. Biotite occurs as up to 200-pm pale brown andpale green pleochroic crystals, intergrown with orthoam-phibole and cordierite.
Magnetite, occurring as tiny crystals in the quartz- andalbite-bearing rocks, is characterized by small amountsof Zn and Mn (analysis l, Table 3). Magnetite in rockswithout quartz and albite is free of Zn and Mn (analysis2, Table 3) and frequently shows coronitic rims of inter-grown hercynite and corundum (analyses 3 and 4, Table3). Locally, the coronitic rims contain small relict chloritecrystals. Magnetite of both described types often containilmenite lamellae and tiny inclusions of chalcopyrite andpynhotite.
The magnetite- and spinel-rich biotite + chlorite +orthoamphibole skarn of Magnusson (1930) contains fourassemblages bearing cummingtonite + orthoamphibole+ biotite that have minerals similar to those in the cor-dierite + gedrite quartzite, but with higher proportionsof magnetite; the magnetite contents increase from theouter toward the inner parts ofthe ore bodies.
Where hydrothermal veins associated with the OstraH<ijden granite cut across the central parts of the ore bod-ies, a variation in the composition of biotite and the or-thoamphiboles is found with distance (on the order ofcentimeters) from the hydrothermal veins. Where mostperfectly developed, the following chemical variation isshown by orthoamphiboles and biotite in and around thehydrothermal veins: (l) The veins consist of quartz andfeldspars with locally some cordierite and pale blue-brownpleochroic anthophyllite (analyses 3 and 4, Table 4) in-tergrown with dark brown pleochroic biotite, with rela-tively high amounts of Fe, Mn, Cl, Ba, and Ti (analysesI and2, Table 5). Anthophyllite and biotite replace quafiz,feldspars, and, ifpresent, cordierite. (2) At the contact ofveins and wall rock, a thin rim of pale blue pleochroic,up to 700-prm, subhedral anthophyllite (analyses I and 2,Table 4) is developed. (3) In the wall rocks near the hy-drothermal veins, a more Na- and Al-rich gedrite (anal-yses 5 and 6, Table 4) than that in the veins is foundintergrown with less Fe-, Mn-, C1-, Ba-, and Ti-rich bio-tite than that in the veins (analyses 3 and 4, Table 5). (4)With increasing distance from the veins, gedrite becomesprogressively enriched in Na, Al, and F and depleted inFe and Si (analyses 6-8, Table 4); the color ofthe inter-grown biotites changes from pale brown to pale greenpleochroic, and they become progressively depleted inFe, Cl, Ba, and Ti, and enriched in Mg and F (analyses6-8, Table 5).
Fonnn,l,rroN oF THE oRTHoAMpHTBoLE-BEARTNGASSEMBLAGES
The rocks described in this paper show a relict olderassemblage consisting of quartz, albite, magnetite, phen-gite, and chlorite, which is replaced by a younger assem-blage consisting of biotite, orthoamphiboles, cordierite,cummingtonite, a second magnetite, hercynite, and co-rundum. Damman (1988b) suggested that the relict olderassemblage was formed by magnesian metasomatic alter-ation of metavolcanics in two conduit zones for hydro-thermal fluids from which overlying manganiferous ironores were deposited. The formation of orthoamphibole,cordierite, and biotite at the expense ofchlorite, phengite,quafiz, and albite may occur according to metamorphicreactions such as (l) chlorite I quartz: orthoamphibole+ cordierite (Akella and Winkler, 1966) and (2) phengite* chlorite : biotite + cordierite (Helmers, 1984). How-ever, biotite and orthoamphibole in and along the hydro-thermal veins associated with the Ostra Hdjden graniteshow rapid changes in Ti, Ba, Mn, Fe, Mg, F, and Clcontents with distance from the veins (Tables 4 and 5),suggesting that the orthoamphibole-, biotite-, and cor-
TABLE 2. Chemical analyses of anthophyllite, gedrite, and sodium gedrite from the outer parts of the conduit zones
577
Anthophyllite Gedrite Sodium gedrite
8H25.3
1 2H25.8 H25.7
e
H25.144
H25.135
H25.17o
H25.10 H25.1
sio,Alro3Tio,FeOMnOMgoCaONaroF
Total
5 lratAl
L6lAl
TiFeMnMgCaNaF
52.022.730.07
24.920.25
16.330.220.350 . 1 7
97,06
7.720.280.200.013.090.033.610.040 . 1 00.08
1 5 . 1 6
51.474.370.07
25.240.27
15.840.220.610.19
98.28
7.560.440.320.013 1 00.033.470.030 . 1 70.09
15.22
0 . 1 30.530.300.42
48.326.150.13
24.950.21
14.96
0 8 00.18
95.70
7.320.680.420 0 13 . 1 60.033 3 8
0.240.08
15.32
46.609.62o.17
24.820.27
14.290.321 . 1 90.29
97.57
6.941.050.64o.o2a n o
0.033 1 8U . U J
0.340.14
15.48
0.340.510.330.48
43.8712.290.22
25.080.31
12.540.341.80o.47
96.92
o.04
1.360.830.023.170.042.830.06U.CJ
0.2315.71
0.480.470.350.54
41.9416.06o.17
24.200.39
12.470.612.360.20
96.40
6.241.761.060.023.010.052.760.100.680 . 1 0
15.78
0.680.480.38U.OJ
39.2719.310.33
21.100.40
13.040.373.050.20
97.07
5.872.131.270.042.640.052.900.060.880 . 1 0
15.94
38.9419.610.35
21.500.37
13.020.362.940.21
97.30
5.812 . 1 91.260.042.680.052.900.060.850.10
15.94TotalANA
XvsaNa/r4tAl
Ed/Ts
0.080.540.290.40
0.840.520.400.65
0.240.520.33u.5c
0.840.520.380.66
dierite-forming reactions took place under influence ofhydrothermal fluids expelled from the veins, or, in otherwords, they were formed by high-temperature metaso-matic reactions.
Hercynite and corundum are always associated withmagnetite, biotite, and orthoamphiboles where the con-duit zones do not contain any q\artz or feldspars. Thissuggests that the formation of biotite and orthoamphiboleat the expense ofquartz, albite, phengite, and chlorite isrelated to the formation of hercynite and corundum.Magnetite occurring where the conduit zones containquartz and albite frequently contains small amounts ofZnand Mn (analysis l, Table 3), while magnetite rimmedby intergrown hercynite and corundum is free of theseelements (analysis 2,Table 3). The rims of hercynite andcorundum sometimes contain small relict chlorite frag-ments. On the basis of the above observations, the fol-lowing reaction is proposed for the formation of hercyniteand corundum: Zn- and Mn-bearing magnetite + someMg and Al derived from the breakdown of chlorite : Zn-and Mn-free magnetite * hercynite * corundum.
EsrrrvrlrroN oF P-T coNDrrroNs DURTNGFORMATION OF THE
ORTHOAMPHIBOLE.BEARING ASSEMBLAGES
Damman (1988a) estimated the maximum tempera-tures and pressures of formation of the hydrothermal veinsystem associated with the Ostra Hdjden granite at 550-600'C and less than 3 kbar. A fluid-inclusion study ofthe granite and the hydrothermal veins (Damman, inpreparation) suggests maximum temperatures of em-placement around 560 "C, confirming the above I esti-
mates, but the latter study also shows that the greatestIithostatic pressure was about 1.0 kbar and that maxi-mum fluid pressures in the hydrothermal veins werearound 4 kbar.
The orthoamphiboles in the veins crystallized relative-ly late (after qtartz, feldspars, and cordierite), suggestingthat they were formed at T < 560 'C. Several papers(Robinson et al., 1969, 1970, l97l; Ross et al., 1969)have reported a miscibility gap between anthophyllite andgedrite. Figure 2 shows that the orthoamphiboles pre-
TABLE 3. Chemical analyses of magnetite, hercynite, and co-rundum
Magnetite Hercynite Corundum'I
G2.12 3 4
G11.Mt G11.HCT G11.Cor
sio,Alro3FerO.FeOMnOMgoZnO
Total
SiAIFe3*Fe.*MnMgZn
Total
U.JC
0.3268.8329.570.22
0.3568.4330.77
99.55
58.492.76
31.480.535.741.86
100.85
1.940.06o.740.010.240.043.03
98.091.95
1 00.1 4
1 .980.02
2.00
0.3499.73
0.010.021.990.960.01
0.013.00
0.021.981.00
3.00
Note: Analysis 1 is of magnetite that occurs without hercynite or corun-dum. Analysis 2 is of magnetite that occurs with hercynite (analysis 3) andcorundum (analysis 4).
sented in this paper show a complete range of A-site andtarAl values of 0.02-0.89 and 0.10-2.32, respectively, im-plying that they were formed under supersolvus condi-tions (for discussion of data, see below under orthoam-phibole chemistry). Crowley and Spear (1981) andRobinson et al. (1982) suggesred that, with respect to A-siteoccupancy, the miscibility gap widens and the tempera-ture of the crest of the solyus increases, going from Mg-rich to Fe-rich bulk compositions.
Crowley and Spear (1981) also suggested that the slopeof the critical curve for the solvus has a positive dP/dT.The orthoamphiboles presented in this paper were formedat T < 560 "C, implying that the temperature of the crestof the solvus of the anthophyllite-gedrite miscibility gapmust also be below 560'C at 1.0 kbar and X."values ofthe orthoamphibole between 0.35 and 0.62 (Fig. 3; fordiscussion of data, see below under orthoamphibolechemistry). This result confirms the suggestion of Crow-ley and Spear (1981).
Onrno,qlrpHrBor,E cHEMrsrRY
In Figure 2, all orthoamphibole analyses made for thepresent study and some analyses of subsilicic sodium ged-rite presented by Damman (1988a, see below) are plottedin a diagram of maximum occupancy of Na in the A siteversus rarAl (Berg, 1985); the orthoamphiboles show thelargest variation in the ratios of Na in the A site to t4lAlever reported for one single locality. Also shown in Figure2 is the subdivision of the orthoamphiboles according tothe nomenclature of Leake (1978).
The subsilicic sodium gedrite (Damman, 1988a; anal-yses 5 and 6, Table 8) presented in Figure 2 is found ina hydrothermal vein that belongs to the same hydrother-mal vein system as the veins crosscutting the conduitzones described in this paper. Instead ofcrosscutting therocks of the conduit zones that were already Mg enrichedprior to hydrothermal vein formation (Damman, 1988b),however, this vein crosscuts Mg-poor metavolcanics sev-eral hundred meters away from the conduit zones (Dam-man, 1988a). The vein shows the following crystallizationsequence (Damman, 1988a): andalusite, andalusite +cordierite, cordierite * subsilicic sodium gedrite, biotite,and quartz * albite * sillimanite * oxides + sulfides.
Robinson etal. (1971,1982) and Spear (1980) showeda moderate inverse correlation between the edenite/tschermakite ratio [calculated as A(Na + K)/[r4rAl - r(Na
+ K)1, see Berg (1985) and Fig. 21. Berg (1985) showed asimilar moderate inverse correlation between the A(K +Na)/tatf ratio and X*. These correlations suggest thatthe edenite/tschermakite ratio of orthoamphiboles is crys-tallographically and crystallochemically controlled by theincorporation of Mg and Fe in the Ml-M3 sites of theorthoamphibole structure. In Figure 3, the orthoamphi-boles analyzed for the present study and subsilicic sodiumgedrite (Damman, 1988a) are plotted in an A-site occu-pancy per tarAl vs. X*" diagram similar to that used byBerg (1985). This figure shows a wide variation in X""and ratios of A-site occupancy to ralAl, but only shows amoderate positive correlation between them.
In Table 6 a comparison is made between P-Z condi-
tions of formation and edenite/tschermakite ratios of or-thoamphiboles analyzed for the present study, subsilicicsodium gedrite (Damman, 1988a), and orthoamphibolesfrom other localities derived from the literature. Al-though there is a large variation of P-T conditions andedenite/tschermakite ratios, it is interesting to notice thatthe orthoamphiboles with the highest edenite/tschermak-ite ratios (Berg, 1985; this paper) were formed at the low-est pressures, suggesting that a decrease in lithostaticpressure may favor the edenite substitution at expense ofthe tschermakite substitution, given that enough Na, Al,and Si to fulfill this condition are available (see below).
The subsilicic sodium gedrite (Damman, 1988a) wasformed under the same P-7 conditions as the orthoam-phiboles in the conduit zones described in this paper, butshows much lower edenite/tschermakite ratios. The Siand Al content of the subsilicic sodium gedrite are con-trolled by those of coexisting cordierite, while its Na isderived from a hydrothermal fluid (Damman, 1988a).Biotite (analyses 5 and 6, Table 7) and subsilicic sodiumgedrite (analyses 5 and 6, Table 8) from the subsilicicsodium gedrite-bearing hydrothermal vein are very richin Fe and only contain minor Mg. The country rockaround this vein does not contain these elements, sug-gesting that they are fluid derived and implying that thevein-forming fluid was relatively rich in Fe and did notcontain much Mg. The data presented in Tables 4 and 5show that orthoamphibole and biotite in hydrothermal
e
D
z
o . 5 3
o . + e
o . 4 3
o . 3 e
o . 3 3
o . 2 g
o . 2 3o o . 2 0 . 4 0 . 6
x v go . e
Fig. 3. A-site occupancy per tatAl vs. Xr" for all orthoam-phiboles analyzed in the GAsbom area. For symbols, see Fig. 2.
veins crosscutting the conduit zones described in this pa-per are also relatively Fe rich (analyses l-4, Table 4, andanalyses I and 2, Table 5). Going from the veins into thecountry rock, these minerals change in composition fromFe rich to more Mg rich, suggesting that the Mg-enrichedrocks of the conduit zones (Damman, 1988b) acted as asource of the Mg incorporated in orthoamphibole andbiotite.
TABLE 5. Variaton in biotite chemistry with distance from the hydrothermal veins in the central part of the conduit zones
This paperDamman (1 988a)Berg (1 985)Otten (1 984)Spear (1 980)Abraham and Schreyer
(1 973)Zotov and Siderenko
(1 s67)James et al. (1978)Lal and Moorehouse
(1 969)Kroonenberg (1 976)
0.33-0.880.31-0.500.54-0.86
0.680.05-0.50
approx. 0.33-0.63
0.21approx. 0.36-0.52
0.26-0.270.33
TABLE 6. A comparison of edenite/tschermakite ratios (EdlTs)and P and f of formation for orthoamphiboles fromdifferent localities
Ed/Ts P(kba0 lfc)
which biotite equilibrates may be related to their Fe con-tents in a similar way as the incorporation of F is relatedto Mg. According to Munoz (1984) and Volfinger et al.(1985) this so-called Mg-Cl avoidance is confirmed bymany analyses of natural biotites and amphiboles.
The few available data on F in orthoamphiboles ledBerg (1985) to suggest that F incorporation in orthoam-phibole is related to Mg contents, in a way similar to thecase in biotite.
No published data on Cl in orthoamphibole could befound in the literature, and none of the orthoamphibolesdescribed here contains over 0. l0 wto/o Cl.
To evaluate the above suggestions, biotite and or-thoamphibole occurring in and along (maximal distancefrom the vein-wall rock contact : 20 cm) three hydro-thermal veins associated with the Ostra Hdjden granite,which crosscut the conduit zones described in this paper(Fig. l), are plotted in diagrams of Xr, vs. F (Fig. 4a),Xr, vs. Cl (Fig. 4b), and I6rAl vs. F (Fig. 4c) for biotiteand in a diagram of Xr, vs. F (Fig. 4d) for orthoamphi-bole. In the veins, biotite and coexisting orthoamphibolereplace quartz, plagioclase, and, ifpresent, cordierite. Inthe wall rock of the veins, biotite and orthoamphiboleoccur in intergrown masses, replacing quartz, albite,phengite, and chlorite, which were formed by magnesianmetasomatic alteration of volcanics during deposition ofthe Glsborn manganiferous iron ores (Fig. 1; Damman,1988b). For comparison, biotite and subsilicic sodiumgedrite from the subsilicic sodium gedrite-bearing veindescribed by Damman (1988a) are plotted as well. Fromthe discussion presented in the section on analytical pro-cedures, it follows lhatf'", andffi., are used instead/"rolfno and fnro/fr"r.
Chemical data on some orthoamphibole and biotitefrom one vein (marked with * signs in Figs. 4a-4d) arepresented in Tables 4 and 5, and data on some orthoam-phibole and biotite from the other veins and the subsilicicsodium gedrite-bearing vein are presented in Tables 7and 8 and by Damman (1988a).
Tables 5 and 7 show that a considerable difference in/fi. exists among the four hydrothermal veins under dis-cussion, and they show that for each of the veins, whengoing from the vein outward (compare analyses I and 8,Table 5) fL, remains relatively constant. Tables 5 and 7and Figure 4a show that biotites from different veins withsimilar Xr, values have very different F contents, sug-gesting that these differences in F content are caused bythe between-vein differences in/fi. discussed above (Val-ley et al., 1982). Whenever biotite from a single vein andthe surrounding country rock show a variation in Xr",Xru increases when going from the vein outward, and theF content of the biotite increases with increasitg Xue(compare analyses l-8 in Table 5 and Fig. 4a). These datasuggest that in and along each vein, at more or less con-stant f'rr, the F content of biotite is controlled by theeffect of F-Fe avoidance (Munoz and Ludington, 1974,1977; Petersen et al., 1982; Valley et al., 19821- Munoz,l 984).
< 1 .5 <560<1 .5 <560
2 + 1 6 1 5>5 500-550>4 530
0.5 550-580
< 1 1 ?3-6 625
<10 400-7005-6 700-800
The above data suggest that besides X",, (Robinson etal.,1971, 1982; Spear, 1980; Berg, 1985), pressure andthe paragenesis (or, in other words, the availability of Na,Al, Si, Mg, and Fe) have a large influence on edenite/tschermakite ratios and the Xr, of orthoamphiboles. Thesedata contradict the suggestions ofRobinson et al. (1971,1982), Spear (1980), and Berg (1985), who srrggested thatedenite/tschermakite ratios of orthoamphiboles are onlycrystallochemically controlled by their Xrr.
F a,Nn Cl rN onrrroAMpHrBot,E AND BroTrrE
Field, experimental, and theoretical data on F and CIin biotite (reviews by Munoz and Ludington, 1974, 1977;Petersen et aI., 1982; Valley et al., 1982; and Munoz,1984) show that the extent ofF incorporation in biotitedepends on Xr, [Mg/(Mg + Fe)], the amount of I6rAl pres-ent in the biotite, and the HF fugacity (f,r.) during theformation of the biotite. The efect of X*, is discussed byMunoz and Ludington(1974,1977), Petersen et al. (1982),Valley et al. (1982), and Munoz (1984) who showed thatfor any given ratio of fnro/f^, in fluids with which biotiteequilibrates, the F/OH ratio in biotite increases with Xrr.
Valley et al. (1982) discussed the effect of/r. on theincorporation of F in biotite, and they showed that com-pared to Mg-rich biotites, Fe-rich biotites require a higher/,o to stabilize a given amount of F. Valley et al. (1982)also suggested that at any ratio of f^ro/f* in fluids withwhich biotite equilibrates, an increase of the t6tAl contentenhances incorporation ofF. The latter effect is opposedto the results of experiments of Munoz and Ludington(1974), who did not find any correlation between t6rAl andF in annite and siderophyllite.
The small amount of Cl in biotite, commonly muchlower than that of F, makes it difficult to evaluate thecrystallochemical controls of its incorporation in biotite(Munoz, 1984). Leelanandem (1969) reported that Clcontent in biotite containing 0. 19 to 0.62 wIo/o Cl is notcorrelated with Xrr. Jacobs (1976) reported biotites with0. l-0.3 wto/o Cl that show an increase in Cl content withincreasing Fe contents; this suggests that incorporation ofCl in biotite at any given ratio of f^,o/fr., in fluids with
Trele 7. Analyses of biotite along two hydrothermal veins described in this paper (G32 and G23) and biotite (E7) described byDamman (1988a)
G32.01 G32.02 G23.7 G23.6 E7.2 E7.3
sio,Alro3Tio,BaOFeOMnOMgoNaroKrOclT
Total
SirarAlrotAlTiBahe
MnMgNaKclF
Total
Xug
&rxhrv(F)log fiFrv(cDlog flcl
34.7517.361.09
19.310.04
12.610.208.590 . 1 50.79
94.89c . J /
2.630.530 . 1 3
2.490.012.900.061.690.04n ? o
16.24
0.540.240.221 .92
-4.54-4 .10-2.31
35.731 7 . 1 31.05
19.26
12.500 . 1 68.760 . 1 80.81
95.55
5.472.530.560. ' t2
2.46
2.850.051.710.04n 2 0
1 6 . 1 8
0.530.260.211 .90
-4.51-4.08-2.33
37.8515 .910.06
14.570.09
15.720.858.620 .103.87
97.64
5.712.290.540.01
1 .840.013.530.251.640.021 8 6
17 .69
0.660 .180 .161 . 1 6
-3.80-4.34-2.07
38.4815.510.o2
14.940.05
16.170.908.830.103.68
98.68
5.742.260.47
1.860.013.590.261.680.021.74
17.64
0.660 . 1 60 . 1 81 . 2 1
-3.84-4.32-2.09
35.8117.300.30
22.760.218.780.568.850.103.29
97.96
5.582.420.760.03
2.970.032.04o.171.760.021.62
17.40
0.400.460.140.91
-3.53-3.81-2.60
36.1717.360.65
20.900 . 1 5
10,420.7'l8.780.103.28
98.52
5.552.450.690.08
2.680.022.380.211.72o.021.59
17.40
0.47o.370 . 1 61.03
-3.64-3.94-2.47
Tables 5 and 7 show that /'"o of the fluid with whichbiotite equilibrated also varies considerably between veins.Table 5 shows that/fi., decreases when going from a hy-drothermal vein outward (compare analyses I and 8). TheXr, vs. Cl plot (Fie. ab) shows that biotites with similarXr, values from different hydrothermal veins have dif-ferent Cl contents, and it shows that whenever a variationin X", of biotite along a single vein exists (Table 5), €lshows a negative correlation w'ith XMg. The above datasuggest that (l) between veins, incorporation ofCl in bio-tite is controlled by variation of/'"., and (2) along eachvein, the variation in/i., and the efect of Mg-Cl avoid-ance (Munoz, 1984; Volfinger et al., 1985) mutually in-fluence one another.
The plot of t6rAl vs. F (Fig. 4c) shows a wide range ofF content and a much smaller range in t61Al content, butthese do not show any correlation, indicating that F inbiotite is not related to its I6rAl content (Munoz and Lu-dington,1974).
Tables 4 and 8 and the X,, vs. F plot (Fig. 4d) showthat F content of orthoamphibole is always lower thanthat of coexisting biotite (Fig. 4a), and the tables showthat a large diference in F content exists between or-thoamphiboles with similar Xr, values from different hy-drothermal veins. Table 4 and Figure 4d show that when-ever a variation exists in the Xr, of orthoamphiboles alonga single vein, Xr, increases when going from the vein
TABLE 8. Analyses of orthoamphibole along two hydrothermalveins described in this paper (G32 and G23) and hy-drothermal subsilicic sodium gedrite (E7; Damman,1 988a)
outward, and the F content of the orthoamphiboles in-creases with increasing Xrr The above data suggests that,similar to biotite, the incorporation of F in orthoamphi-bole is controlled between veins by diferences inf'r, and,secondarily, along each vein at constant /lrr, by the Xr,of the orthoamphiboles (Berg, 1985).
A plot of XMg vs. Cl for the orthoamphiboles is notpresented, because Cl in the orthoamphiboles rarely ex-ceeds 0.05. wt0/o Cl (Tables 4 and 8).
DrsrnnurroN oF ELEMENTS BETwEEN coExrsrrNcORTHOAMPHIBOLE AND BIOTITE
The discussion ofFigures 4a-4d shows that local vari-ations of F and CI fugacities are the most important con-
+
+ ao
a A s
1 , 2 5
1 0 0
o.75
o50
trol of the incorporation of these elements in orthoam-phibole and biotite. A comparison of data from Tables 5and 7 and Figure 5a shows that the difference between Fcontents of biotite and coexisting orthoamphibole (F", -
Fo-) increases with increasingfh, of the fluid with whichthey equilibrated, suggesting that an increase inliro favorsthe incorporation of F in biotite over coexisting orthoam-phibole. The ratio X"^;/XW increases with increasingamounts of F in orthoamphiboles and biotite, expressedby F", (Fig. 5b), indicating that the distribution of Mgand Fe between coexisting orthoamphibole and biotite iscontrolled by the F contents of these minerals, which inturn are governed by local variations in the F fugacity ofrhe fluid with which they equilibrated (Tables 5 and 7).
1 0
o8
a4
-;
(I)
(J
d:
:d
tL
+
+ + + ^+ o J
o.2 04 o 2 o4o,Jl ^0,7
Fig. 4. (a) X', vs. F, (b) XMs vs. Cl, (c) I6rAl vs. F of biotite, and (d) Xr" vs. F of orthoamphibole in and around hydrothermalveins associated rvith the Ostra Hiijden granite. Each symbol represents one hydrothermal vein; + : da Gll, see Tables 4 and 5;o :daG32,seeana lyses land2,Tab lesTand8; x :dac23,seeana lyses 3and4,Tab lesTand8;A:daET,seeana lyses5and6, Tables 7 and 8 and Damman, 1988a. For errors in analyzed values, see Table l.
^ 2.:
(I)L L 1
10o 202 04 06 0B 10>t p,,s
O Lo
;EL-
1.oo8061 0o8Ub
-t
Fig. 5. Distribution of elements between coexisting orthoamphibole and biotite: @)logf'", vs. (Fu, - Fo-); (b) Fu, vs. (Xfl;r/Xf,?); (c) logf!., vs. (Clu, - Cl"-); (d) Cl", vs. (X",i/Xi;f); (e) Ba", vs. Cl",; (f) Ti", vs. Ti"-; (g) Ti", vs. Cl",; (h) Mn", vs. Mno-. Forerrors in analyzed values, see Table l.
The subsilicic sodium gedrite described by Damman(1988a) is omitted, becuase the crystallization sequenceof this vein shows that its subsilicic sodium gedrite andbiotite were not formed in equilibrium with one another.A comparison of data from Tables 5 and 7 with Figure5c shows that the difference between the Cl contents ofcoexisting biotite and orthoamphibole (C1", - Cl^-) in-creases with increasi\g.fh., in the fluid with which theyequilibrated, suggesting that an increase in/fi., favors theincorporation of Cl in biotite over coexisting orthoam-phibole. Only few biotites contain over 0.3 wto/o Cl (Fig.4b, Table 5); the coexisting orthoamphiboles never con-tain over 0.1 wto/o Cl (Table 5). These Cl-rich biotiteshave relatively low F contents (Table 5). These biotitesand coexisting orthoamphiboles show a decrease of )pr;"/Xfr3 with increasing Cl contents in both minerals, ex-pressed by Clu, Gig. 5d). The above data indicate thatthe distribution of Mg and Fe between coexisting or-thoamphibole and biotite with low F/Cl ratios is con-trolled by the amount of Cl in biotite, whereas with higherF/Cl ratios, the distribution of F between biotite and or-thoamphibole is the most important controlling factor.
The biotites presented in Table 5 (group shown by *signs, Figs. 4a4c) show a positive correlation betweenBa and Cl (Fig. 5e); biotites along other veins are less Cl-rich (Fig. 4b) and do not contain more than 0.05 wto/o Ba.Coexisting orthoamphiboles do not contain any Ba (Ta-ble 4), indicating that (l) Ba is preferentially incorporatedin biotite over coexisting orthoamphibole and (2) an in-crease in/". of fluids with which biotite equilibrated (Ta-bles 5 and 7) enhances the incorporation ofBa in biotite.The Ti content of biotite is always higher than that ofcoexisting orthoamphibole (Fig. 5f). In biotite, Ti showsa strong positive correlation with Cl (Fig. 5g), implyingthat (l) Ti is preferentially incorporated in biotite overcoexisting orthoamphibole and (2) incorporation of Ti inbiotite, similar to incorporation of Ba, is enhanced by anincrease in /fi., in the fluid with which the biotite equil-ibrated.
The Mn content of orthoamphibole is always higherthan that of coexisting biotite (Fig. 5h), but there is nocorrelation with the Cl or F contents of the orthoamphi-boles (Tables 4 and 8), indicating that (l) Mn is prefer-entially incorporated in orthoamphibole over coexistingbiotite and (2) the incorporation of Mn in orthoamphi-bole is independent of variations in f'", and, f'"o of thefluid with which the orthoamphibole equilibrated.
AcrNowr,nocMENTS
I would like to thank Prof. Dr. L S. Oen and Drs. P. Maaskant, F.Spear, J. Munoz, K. Linthout, W. Lustenhouwer, and A. Boudreau fortheir help and encouragement during the preparation of this paper. Elec-tron-microprobe analyses were performed at the electron-microprobe lab-oratory of the Instituut voor Aardwetenschappen, Vrye Universiteit, Am-sterdam, with financial and personnel support of Z.W.O.-W.A.C.O.M.(research group for analytical chemistry of minerals and rocks subsidizedby the Netherlands Organization for the Advancement ofPure Research).
RnrBnnNcps crrEDAberg, G., Bollmark, B., Bjdrk, L., and Wiklander, U. (1983a) Radio-
metric dating ofthe Horrsjd granite, south-central Sweden. GeologiskaFdrenings i Stockholm Fdrhandlingar, 105, 78-81.
Aberg, G., kvi, 8., and Frederiksson, G. (1983b) Zircon ages of meta-volcanic and synorogenic granites from the Svardsjd and Yxsjtibergareas, south-central Sweden. Geologiska Fiirenings i Stockholm Filr-handlingar, 105, 199-203.
Abraham, K., and Schreyer, W. (1973) Petrology of a femrginous hornfelsfrom Riegensgluck, Harz Mountains, Germany. Contributions to Min-eralogy and Petrology, 40, 27 5-292.
Akella, J., and Winkler, H.G.F. (1966) Orthorhombic amphibole in somemetamorphic reactions Contributions to Mineralogy and Petrology, 2,t-12.
Baker, J.H. (1 985) The petrology and geochemistry of L9-1.8 Ga graniticmagmatism and related sub-seafloor hydrothermal alteration and ore-forming processes. Ph.D. thesis, 204 p. University of Amsterdam. GUAPapers ofGeology, Series l, No. 21.
Berg, J.H. (1985) Compositional variation in sodium gedrite from Lab-rador. American Mineralogist, 7 0, 1205-1210.
Billstrdm, K., Aberg, G., and Ohlander, B. (1988) Isotopic and geochem-ical data of the Pingstaberg Mo-bearing granite in Bergslagen, south-central Sweden. In J.H. Baker and R.H. Hellingwerf, Eds., TheBergslagen Province, central Sweden: Structure, stratigraphy and ore-forming processes-I.G.C.P. Project 247. Geologie en Mijnbouw, 67,255-263.
Crowley, P.D., and Spear, F.S. (1981) The orthoamphibole volus: P, ?n,X(Fe-Mg). Geological Society of America Abstracts with Programs, 13,435.
Damman, A.H. (1988a) Hydrothermal subsilicic sodium gedrite frorn theGAsborn area, West Bergslagen, central S\ileden. Mineralogical Maga-zine.52. 193-200.
- (1988b) Exhalative-sedimentary manganiferous iron ores from theGisborn area, West Bergslagen, central Sweden. In J.H. Baker and R.H.Hellingwerf, Eds., The Bergslagen Province, central Sweden: Structure,stratigraphy and ore-forming processes- I.G.C.P. Project 247 . Geolo-gie en Mijnbouw, 67, 433442.
Guidotti, C.V. (1984) Micas in metamorphic rocks. Mineralogical Societyof America Reviews in Mineralogy, 13,357468.
Helrners, H. (1984) Stages of ganite intrusion and regional metamor-phism in the Proterozoic rocks ofwestern Bergslagen, central Sweden,as exemplified in the Grengen area. Neues Jahrbuch fiir MineralogischeAbhandlungen, 150, 307 -324.
Jacobs, D.C. (1976) Geochemistry of biotite in the Santa Rita and Han-nover-Fierro stocks, Central mining district, Grant County, New Mex-ico. Ph.D. dissertation, University of Utah, Salt Lake City, Utah.
James, R.A., Grieve, R.A.F., and Paul, L. (1978) The petrology of cor-dierite-anthophyllite gneisses and associated mafic and pelitic gneissesat Manitouwadge, Ontario. American Journal of Science, 278, 4l-63.
Kroonenberg, S.B. (1 976) Amphibolite-facies and granulite-facies meta-morphism in the Coeroeni-Lucie area, southwestern Surinam. Ph.D.thesis, 259 p University of Amsterdam, Amsterdam, Netherlands.
Lal, R.K., and Moorehouse, W.W. (1969) Cordierite-gedrite rocks andassociated gneisses at Fishtail l,ake, Harcourt Township, Ontario. Ca-nadian Joumal ofEarth Sciences, 6, 145-165.
Leake, B.E. (1978) Nomenclature of amphiboles. American Mineralogist,63 ,1023 -1053 .
Leelanandem, C. (1969) H,O., F, and Cl in the chamockitic biotites fromKondapalli, India. Neues Jahrbuch fiir Mineralogie Monatshefte, 461-468.
Magnusson, N.H. (1930) Lengbans malmtrakt. Sveriges Geologiska Un-dersdkning, ser. Ca, 23, I 1 I p.
Munoz, J L. (1984) F-OH and CI-OH exchange in micas with applicationto hydrothermal ore deposits. Mineralogical Society of America Re-views in Mineralogy, 13,469491.
Munoz, J.L., and Ludington, S.D. (1974) Fluoride-hydroxyl exchange inbiotite. American Journal of Science, 27 4, 39G413.
- (1977) Fluorine-hydroxyl exchange in synthetic muscovite and its
applications to muscovite-biotite assemblages. American Mineralogist,62,304-308.
Oen, I.S., Helmers, H., Verschure, R.H., and Wiklander, U. (1982) Ore-deposition in a Proterozoic incipient rift environmenl A tentative modelfor the Filipstad-Gr).th).ttan-Hjulsjii region, Bergslagen, Sweden. Geo-logische Rundschav,, 7 l, 182-94.
Oen, I.S., Venchure, R., and Wiklander, U. (1984) Isotopic age deter-minations in Bergslagen: V. The Horrsjd granite. Geologie en Mijnbouw,63, 8s-88.
Otten, M.T. (1984) Na-Al-rich homblende coexisting with hornblende ina corona between plagioclase and olivine. American Mineralogist, 69,458464.
Petersen, E.U., Essene, E.J., and Peacor, D R. (1982) Fluorine end-mem-ber micas and amphiboles. American Mineralogist, 67,538-544.
Robinson, Peter, Klein, C., and Ross, M. (1969) Equilibrium coexrstenceof three amphiboles. Contributions to Mineralogy and petrology, 22,248-258.
Robinson, Peter, Ross, M., and Jaffe, H.W. (1970) The composition fieldofanthophyllite and the anthophyllite miscibillity gap (abs ). AmericanMineralogist, 55, 307-309.
-(1971) Composition of the anthophyllite-gedrite series, compari-sons of gedrite-hornblende, and the anthophyllite-gedrite solvusAmerican Mineralogist, 56, 1005-1041.
Robinson, Peter, Spear, F.S., Schumacher, J.C., Laird, J., Klein, C., Ev-ans, E.W., and Doolan, B.L. (1982) Phase relations of metarnorphicamphiboles: Natural occurrence and theory. Mineralogical Society ofAmerica Reviews in Mineralogy, 9b, l-228.
Ross, M., Papike, JJ., and Shaw, K.W. (1969) Exsolution textures inamphiboles as indicators ofsubsolidus thermal history MineralogicalSociety of America Special Paper 2, 27 5-299.
Spear, F.S. (1980) The gedrite-anthophyllite solvus and the compositionlimits of orthoamphibole from the Post Pond Volcanics, Vermont.American Mineralogist, 65, 1103-l I18.
Valley, J.W., Petersen, E.U, Essene, E.J, and Bowman, J.R. (1982)Fluorphlogopite and fluortremolite in Adirondack marbles and calcu-lated C-O-H-F compositions. American Mineralogist, 67 , 545-557.
Volfinger, M., Robert, J.-L., Vielzeuf, D., and Neiva, A.M.R. (1985)Structural control on the chlorine content of OH-bearing silicates (mi-cas and amphiboles). Geochimica et Cosmochimica Acta,49,3748.
Zotov, I.A., and Siderenko, G.A. (1967) Magnesian gedrite from thesouthwestern Pamirs. Akademiya Nauk SSSR Doklady, I 80, I 38-14 l.
