Geochemistry of Mineralizing Fluids in the Bralorne-Pioneer Mesothermal.pdf

Economic Geology Vol. 86, 1991, pp. 318-353

Geochemistry of Mineralizing Fluids in the Bralorne-Pioneer Mesothermal Gold Vein Deposit, British Columbia, Canada

C. H. B. LEITCH*, C. I. GODWIN, T. H. BROWN, Department of Geological Sciences, University of British Columbia, Vancouver, British Columbia, Canada V6T 2B4

AND B. E. TAYLOR

Geological Survey of Canada, 601 Booth Street, Ottawa, Ontario, Canada K1A OE8

Abstract

Gold quartz veins at the Bralorne mesothermal vein deposit have extensive quartz-ankeritic carbonate-muscovite hydrothermal alteration envelopes that grade outward to chlorite-cal½ite- albite. Minor pyrite with traces of sphalerite, galena, and tetrahedrite are found in the veins with native gold; more abundant pyrite and arsenopyrite, with lesser pyrrhotite and chalcopyrite occur in altered wall rock adjacent to the veins. Envelopes are characterized by addition of KgO, COg, S, As, and Au, with corresponding depletion of NagO, FeO total, and MgO; SiOg and CaO are locally depleted and reconcentrated. Fluid inclusions suggest that deposition of gold was from dilute aqueous solutions (<5 wt % NaC1 equiv) that had a significant carbonic component (5-15 mole % COg + CH4), at 350øC and 1.75 kbars. Progressive dilution of this fluid by cooler meteoric water led to fluids at 250øC and 0.5 to I kbar with 1 wt percent or less NaC1 equivalent and no detectable carbonic component. Sulfur isotope ratios of sulfides associated with gold mineralization range from -7 to +9 per mil, clustering about a magmatic signature of 0 per mil. The ore fluid had a •so value of 13 _ I and •3C of-11 _ 2 per mil, based on measurements in coexisting vein quartz, carbonate, and muscovite. Temperature of the mineralizing fluids appears to have increased with depth at a normal geothermal gradient of approximately 30 øC/km.

Gold, sulfides, and the associated alteration assemblages are modeled, using observed mineralogy and thermodynamic data for chloride complexes, as having been deposited from a slightly acid solution (pH 4.5) with an Na/K ratio of at least 8:1 and a high content of dissolved COg (log fugacity = 2.5). Conditions were strongly reducing, as suggested by the log fugacity of CH4 (0.5), with fo2 about 10 -30 bars andfs2 about 10 -7 bars. Precipitation of gold in the immediately adjacent, highly quartz-sericite-ankerite-pyrite-altered wall rock, was due to reaction with (carbonation of) the wall rock that caused the pH of the ore fluid to rise, destabilizing aurous chloride complexes. Such a shift is in the direction of slightly increasing or constant stability of gold thiosulfide ions, and at a relatively constant fo2. The most significant result of this modeling is that in detail, the actual precipitation of gold is by reduction of aurous ions by electrons donated through concurrent oxidation of S -g in HgS to S- in pyrite (FeSg). This explains the empirically observed correlation between pyrite and gold in this type of deposit, and in the model between consumption of Fe +g and production of pyrite on the one hand, with precipitation of gold on the other.

Introduction

THE Bralorne-Pioneer mesothermal gold vein produced 130 metric tons of gold over a 70-year span, the only deposit in the Canadian Cordillera to ap- proach the output of major deposits in the Precam- brian shield (Bertoni, 1983; Phillips, 1986). In this study, petrographic and chemical analyses of hydrothermal alteration, fluid inclusion and stable isotope studies, and thermodynamic modeling, have led to estimates of the pressure-temperature-composition (P-T-X) of the mineralizing fluids. Stable isotope studies confirm that significant interaction of wall

* Present address: Geological Survey of Canada, 100 West Pender Street, Vancouver, British Columbia V6B 1R8.

rocks with the ore fluid took place and shed light on the source of the fluids.

Geologic Setting of the Bralorne Deposit

The Bralorne gold vein is in the Bridge River mining camp, 180 km north of Vancouver (Fig. 1). The tectonic setting, just east of the Coast Plutonic Com- plex, is described by Rusmore et al. (1988); detailed geology and previous work on the deposit are in Leitch (1990). In the Bridge River camp, two main lithologic assemblages can be distinguished: one dominantly oceanic and the other dominantly island arc. The former is represented by the Permian(?) to Jurassic Bridge River Complex that comprises basalts and thick accumulations of ribbon chert with minor

0361-0128/91/1185/318-3653.00 318

BRALORNE Au DEPOSIT, B. C. 319

•E'/dox"'•d-ø/X•tx'/ \\ • I • •••••• 51ø00 Geologic Contact

h--hornblende k = K -Ar BR b=biotite r =Rb-Sr • J

in o LAKE 15 z=zircon u =U-Pb ••1 ngress

0 5 10

EOCENE L E G E N D k44 ••• •u •hk2• BR BR •• • •v•';•

•Rexmount Porphyry n,o. •r 63 •9 •' © • • • vvv • Bendor Plutons Lake coas,.,u,o.c co..,ex c.c co hk284zu272* * * + + +•• • -••

•MethowTerrane '(• •''-t,n•--••.•+:•+•fi7'+ + +*+*$ .•/ EARLY PERMIAN-EAR1. Y CRETACEOUS • - •-2 + + + +

• Cadwallader ,errane •• PERMIAN-EARLY JURASSIC .... YUKON BR

BRIDGE RIVER TERRANE ,• -- • .... ] •XX• % • • •'•+l • /'Anderson • :. • • +'+ +• • •+ •hk•+'/ / • Lake

• Bridge River Assembledge ' • • • •+ • •:•••5 ••• +• BR • • Prince • • % +• •+ t'x x• • / +hk63. + • •'" •( • George•

ULTRAMAFICROCKS •[ -•

•Cadwallade. Fault Zone '+ • usa hk73 • • •:t•• -- 50ø34' FIG. 1. Geology, mineral deposits, and isotopic dates of the Bridge River district, southwestern

British Columbia. Geology is after Woodsworth (1977) and Church (1987). Mineral deposits correspond to the following numbered open triangle symbols: 1 = Bralorne, 2 = Ida May, 3 = Pioneer, 4 = P. E. Gold, 5 = BRX (Arizona), 6 = BRX (California), 7 = Veritas, 8 = Waterloo, 9 = Lucky Gem, 10 = Lucky Strike, 11 = Minto, 12 = Golden Sidewalk, 13 = Congress, 14 = Peerless, 15 = Summit, 16 = Kelvin, 17 = Olympic, 18 = Matson, 19 = Greyrock, 20 = Piebiter (Chopper).

limestone and coarse elastic rocks. Alpine-type ultramarie rocks in lensold to very elongate bodies form part of the stratified assemblage (Schiarizza et al., 1989). The ultramarie rocks may mark the sites of major crustal shortening that were later foci for trans- current movements. The island-arc assemblage is represented by the Cadwallader Group of Permian(?) to Triassic age (Cairnes, 1937; Rusmore, 1987). It is a basaltic andesitc pile with minor felsic volcanics (Pi- oneer Formation) overlain by volcaniclastics with mi- nor limestone (Hurley Formation).

The Bridge River and Cadwallader terranes containing these two assemblages form small lozengelike fault-bounded slices sutured between the Insular su-

perterrane on the west and the Intermontane superterrane on the east (Monger, 1984). Potter (1986) and Rusmore (1987) proposed that a collapsed back- arc basin (the Bridge River terrane) was thrust onto the continent margin together with its offshore volcanic arc (the contemporaneous Cadwallader terrane). Others propose that the two terranes were unrelated (Wheeler and McFeely, 1987). If so, they represent

true suspect terranes that could have been transported for large distances from their original places of deposition. In any case, the terranes have been displaced northward for up to 150 km from correlative rocks of the Hozameen terrane by Tertiary movement along the Fraser strike-slip fault (Haugerud, 1985; Ray, 1986). The Hozameen terrane hosts the Carolin gold deposit, which, although of a more dispersed or stockwork nature and hosted in younger (Methow) rocks, has similarities to Bralorne in alteration style and mesothermal character.

The Bralorne block containing the Bralorne and Pioneer deposits, shown on the 8 level underground in Figure 2, is bounded by the northwest-trending Fergusson and Cadwallader strike-slip faults, which are marked along their length by sinuous serpentinized ultramarie bodies. Latest movement may have been right lateral (Stevenson, 1958), the same as the Fraser fault system. Bralorne diorite makes up the bulk of the elongate stock hosting the vein gold deposits; its northeastern contact with Pioneer and Hur- ley Formations has been intruded by a tabular body

320 LEITCH, GODWIN, BROWN, AND TAYLOR

/ /

/ \

*'o

/


of soda granite, almost 3 km long by up to 500 m wide. These intrusions taper southeast near the Pi- oneer mine; farther southeast at the P. E. Gold property they are dikelike (Nordine, 1983). Thus the major ore host is diorite at Bralorne but volcanics at Pi- oneer. The diorite-soda granite complex is spatially associated with gold mineralization but predates it by at least 150 m.y. (Leitch et al., 1991). Mineralization was related instead to a swarm of early Late Creta- ceous (90 Ma) "albitite" dikes that parallel the veins and postdate the metamorphic peak at ca. 100 Ma. These dikes are cut by, and altered along, veins. They are gradational to usually unaltered and therefore largely postmineral green hornblende porphyry dikes (86 Ma).

Workings at the Bralorne mine extend over a vertical range of 2 km from the 0 level at surface at 1,400- m elevation, where the Cosmopolitan vein crops out in the old King mine area (Fig. 2), down to the 44 level at 600 m below sea level. Much of the production of the upper levels of the mine came from the 51 vein; the 51B FW vein still contains some 150,000 metric tons at 15 g/metric ton gold above the 8 level (Bellamy and Arnold, 1985). Farther to the southeast, the largest vein in the mine, the 77, accounted for most of the production at depth. These major gold- bearing veins strike about 110 ø and dip north at 70 ø. Slickensides plunge 45 ø east and steps indicate the last movement was reverse. Major ore shoots in the veins occupy less than 20 percent of the vein and plunge steeply west, roughly perpendicular to the slickensides. The veins contain thin dark ribbons of

fine-grained sulfide in massive milky quartz with mi- nor calcite. The "fault valve" hypothesis of Sibson et al. (1988) offers an explanation for fluid migration

and is consistent with the main features of the ribbon- banded yet coarsely crystalline quartz veins at Bra- lorne. A more detailed structural analysis is in Leitch (1990).

Quartz-carbonate-sericite-chlorite envelopes, up to several meters wide, are extensive around the veins. They contain disseminated pyrite, pyrrhotite, and lesser chalcopyrite; arsenopyrite is confined to vein margins. Minor sphalerite and especially galena correlate with portions of the veins that are richer in gold. Traces of tetrahedrite and stibnite have been observed but tellurides have not. Gold is principally found as thin smeared flakes of native metal in black sulfidic septae of the strongly ribboned shear veins. Gold is rarely found by itself in the quartz, usually as rich pockets in the relatively rare extensional veins (oriented at 060ø), such as the 27 vein in the Pioneer mine.

Analytical Methods

Mineral analysis utilized X-ray diffraction (XRD), scanning electron microscope-energy dispersive system (SEM-EDS), and CAMECA wavelength dispersive electron microprobe (WDEM). Operating conditions and correction procedures for the SEM-EDS and WDEM studies are in Leitch (1989). Routine analysis of secondary standards by WDEM were within _+5 percent of the accepted values. Compositions of both the precursor (feldspar and hornblende) and the main alteration minerals were studied. The variation about a median value of major element oxides for all the minerals studied was less than 10 percent in an individual grain or within a sample. Analyses for whole- rock chemical compositions were by X-ray fluores- cence spectrometry (XRF), mainly on pressed powder

TABLE 1. Summary of Modal Mineralogy for Primary Rocks and Alteration Envelopes around the Bralorne Veins

Unaltered (model vol %)• Diorite Soda granite Fresh Fresh

Altered (modal vol %)1 Outer Central Inner

cl-ep cb-asqns qz-ms-cb

Di SG DI SG DI SG

Qz 10 37 5 43 12 40 25 50 Ab 55 52 30 42 30 35 5 5 Hb 33 102 5 a Ms 10 3 13 7 30 25

Ca 15 3 23 13 15 10

Ak 2 5 2 20 7 C1 20 7 12 Bi 2 a

Ep 10 Ox 2 <1 2 1 1 1 tr tr

Sx tr i i i 2 2 5 3

Abbreviations: ab = albite, ak = ankerite, bi = biotite, ca = calcite, cb = carbonate, el = chlorite, DI = diorite, ep = epidote, hb = hornblende, ms = serieite (muscovite), ox = oxides (ilmenite, rutile, leucoxene, sphene), qz = quartz, SG = soda granite, sx = sulfides (pyrite, arsenopyrite, pyrrhotite, ehalcopyrite), tr = trace

i Averaged from visual thin section estimates in Leitch and Godwin (1988, table 2-4-,5) 2 Identification not certain a Not common


TABLE 2. Electron Microprobe Analyses

Carbonates

Sample no. C094A 84-49/ 84-49/ C116-1 812' 795'

Type Calcite Calcite Calcite Calcite Location

Spot (n) 2-10(2) 9 2-1(4) MgCOa 0.51 2.19 1.4 1.82 CaCOa 97.10 95.90 97.1 95.22 FeCOa 2.38 3.01 1.5 1.57 Total 99.99 101.1 100.0 98.61

Fe/Mg 0.77 0.50 0.43 0.39

Sample no.

SB84-49/795'

Ankerite Cores

7(3) 3(6) 30.57 31.96

62.92 62.32

7.79 6.59

101.3 100.9 0.16 0.15

Siderite Rims

2(4) 47.39 15.26

38.27 100.9

0.38

19-51 FW I

Type Calcites Ankerites Location Cores Rim Margins Spot 4 5B 6A 1 6B 3 6 5 MgCOa 3.3 1.6 2.1 34.11 36.94 34.81 37.99 31.12 CaCO3 95.2 96.6 96.8 51.79 51.48 54.75 51.23 49.29 FeCO3 1.5 1.8 1.1 16.23 11.35 11.57 11.82 15.64 Total 99.9 99.8 100.0 102.1 99.77 101.1 101.0 96.06

Fe/Mg 0.25 0.46 0.27 0.26 0.18 0.19 0.18 0.27

8(3) 44.44

18.18

34.60

97.22

0.35

Sample no. C093A Type Site (n) 5(1) 6(7) SiO2 67.79 67.21 A12Oa 19.51 19.87 CaO 0.69 0.72

Na•O 11.33 11.11 K•O 0.12 0.12 Total 99.44 99.03 Ab 96 96 An 3 3 Or 1 1

Rims

5A

33.5

5O.5 16.0

100.0

0.26

Feldspars

C094A C182-2

Albite

1-2(6) (6) 67.77 68.58 19.82 19.85

0.52 0.26

11.67 11.15 0.03 0.05

99.81 99.89 97 99

3 1 0 0

19-51 FW1

68.76

19.72

0.25

11.84

0.09

100.7

98.5 1.0

0.5

Sample no. Type Site (n) SiO• AlcOa TiO2 MgO FeO

K•O H•O Total

Fe/Mg

C182-2

(3) 48.26

37.19 0.30 1.10

0.90

7.81 4.67

100.2 0.31

Micas

84-49/795' C094A Muscovite

6(2) 5(2) 2-8 48.48 49.46 54.30 34.18 34.47 35.30

0.25 0.42 0.05 1.60 2.63 0.19 1.06 2.28 0.73 8.44 5.47 4.65 4.54 4.72 4.77

98.55 99.45 99.99 0.26 0.33 0.68

19-51FW1 C116-1 84-49/795' Sericite Biotite

7, 8(2) 4A, 2(2) 4 4 49.27 47.17 47.60 39.72 33.00 37.01 32.33 18.55

0.23 0.08 0.51 1.00 2.16 0.82 2.62 16.90 0.88 0.77 1.89 10.41 9.54 8.40 8.29 7.39 4.56 4.56 4.47 4.15

99.64 98.81 97.72 98.13 0.19 0.34 0.28 0.26

Cations

Si 6.69 6.40 6.48 6.82 6.48 6.21 6.39 5.74 Ti 0.02 0.02 0.04 0.00 0.02 0.01 0.05 0.10 A1 5.14 5.32 5.35 5.24 5.10 5.75 5.11 3.16 Mg 0.18 0.31 0.53 0.04 0.42 0.16 0.53 3.64 Fe 0.08 0.11 0.25 0.08 0.09 0.08 0.21 1.26 K 1.24 1.42 0.91 0.75 1.60 1.41 1.42 1.36 Total 13.3 13.6 13.6 12.9 13.7 13.6 13.7 15.3

BRALORNE Au DEPOSIT, B.C. 323

TABLE 2. (Cont.)

Chlorites Hornblende and others

Sample C182-2 C116-1 84-49/795 C093A C094A C093A Type RIP P'TH RIP C'PH RIP APR HBL PREH EP (,0 (7) (2) (5) (5) (5) (6) SiO2 25.70 19.74 26.13 17.47 27.44 24.83 52.74 44.32 41.23 TiO2 0.03 0.09 0.27 ND 0.10 AlcOa 22.94 18.49 23.94 15.17 19.63 19.89 4.07 24.66 30.21 MgO 18.38 13.5 ! 19.03 28.57 20.40 8.78 17.01 0.06 0.08 FeO 20.05 33.82 18.96 27.85 18.53 32.74 12.27 0.27 3.88 CaO 0.07 0.06 11.48 27.17 23.12

Na•O 0.09 0.14 0.71 ND 0.79 K•O 0.15 0.07 H•O 11.88 10.33 11.91 10.94 11.56 10.73 Total 98.95 95.89 100.3 100.0 97.85 96.97 98.62 96.49 99.42

Fe/Mg 0.38 0.59 0.36 0.35 0.33 0.67

Cations

Si 5.19 4.58 5.25 3.83 5.68 5.55

A1 5.51 5.06 5.71 3.92 4.78 5.24

Mg 5.87 4.67 5.75 9.34 6.37 2.93 Fe 3.35 6.56 3.20 5.11 3.19 6.12

Total 19.9 20.9 19.9 22.2 20.0 19.8

Operating conditions and standards used are in given Leitch (1989); chlorite classifications from Hey (1954) Sample locations are given in text Rock types: C093A, SB84-49/795, 812, 19-51FW1, Cl16-1 = diorite; C094A, C182-2 = soda granite (n) -- number of separate sites analyzed; where no number is given, only one analysis was obtained; carbonate compositions are

plotted in Figure 8; Ab, An, Or = mole fractions of albite, anorthite, orthoclase, respectively; not reliable for calcites due to low Fe and Mg contents; blanks indicate elements not analyzed for; ND = not detected; Fe/Mg -- total Fe (as FeO) divided by (total Fe as FeO, plus Mg); H20 contents assigned by difference; chlorite compositions are plotted in Figure 10

Abbreviations: APR -- aphrosiderite; C'PH -- corundophillite; EP -- epidote; HBL = hornblende; PREH = prehnite; P'TH = psuedothuringite; RIP = ripidolite

pellets but with checks on fused discs (precisions var- ied from 5% for major elements to 20% for certain minor elements; details of procedure, error analysis, and interlaboratory comparisons are in Leitch, 1989). Complete whole-rock chemical analyses and norms of altered and unaltered host rocks are in Leitch and

Godwin (1986, 1987).

Petrography and Chemical Composition of Altered Wall Rocks

A striking feature of the Bralorne-Pioneer deposit is the intense wall-rock alteration that accompanied vein formation and gold deposition. Vein envelopes vary from less than 0.1 up to 10 m wide, and in places coalesce to 50 m thickness. The dominant alteration

is carbonitization, although silicification and sericit- ization commonly occur adjacent to the veins. Biotite alteration, rare in the Bralorne mine, is more common at the Pioneer and at the P. E. Gold property, 2 km southeast of the Pioneer. Mineralogical and textural features were observed in about 300 thin and polished sections in ten detailed traverses across altered wall

rock at various levels from surface to 2-km depth. Visually estimated modal compositions of the altered rocks are in Table 1.

Mineralogical zoning

Unaltered wall rocks display sub- or lower greenschist facies metamorphic mineral assemblages, which set broad P-T limits of 2 to 3 kbars and 325 ø to 375øC

(Leitch, 1989). Altered wall-rock envelopes adjacent to the veins including occasional biotite indicate similar or slightly higher temperatures. The three prin- cipal rock types hosting veins are Bralorne diorite, soda granite, and Pioneer greenstone. These three rock types vary in their response to alteration, but in general there is a consistent zoning of alteration minerals, with an associated progressive destruction of texture, over about 5 m from unaltered wall rock to the vein. The sequence is usually from (1) a green chlorite-epidote outer zone that faithfully preserves the original rock textures, to (2) a buff carbonate-albite + sericite intermediate zone, divided into outer texture-preserved and inner texture-destroyed sub- zones, to (3) a foliated, cream-colored quartz-sericite (_+ fuchsite)-carbonate innermost zone that locally becomes a "paper schist." The composition of the carbonate changes from calcite to ankerite as the vein is approached. The zonal arrangement is similar to that described by Robert and Brown (1986b) at the


Sigma mine in Quebec, by Kishida and Kerrich (1987) at the Kerr-Addison mine in Ontario, and by Landfeldt (1987) and Albino (1987) for the Mother Lode district in California.

The chlorite-epidote outer zone is gradational outward with surrounding rocks of greenschist facies metamorphism. It can also be confused with an early, widespread, barren quartz-epidote-carbonate- prehnite stockwork alteration in the diorite and soda granite. Chlorite occurs as fine (0.05 mm or less), dark green, scaly aggregates usually replacing primary hornblende in the diorite and soda granite and commonly intimately intergrown with lesser amounts of calcite and sericite. Chlorite in the diorite lacks pleo- chroism or anomalous interference colors; in contrast, chlorite in the soda granite is coarser (to 0.5 mm), bright green and pleochroic and displays strong pur- ple to blue anomalous interference colors. These optical characteristics suggest that the chlorite in the diorite is a prochlorite with higher Mg and A1, and less Si than the variety found in the soda granite that may be more Fe-rich pennine (Kerr, 1959). This is supported by the mineral chemistry (Table 9.) and re- fleets the more marie (Mg-rich) character of the diorite compared to the soda granite. Epidote usually appears to be early and is replaced by chlorite, carbonate, sericite, and quartz. It most commonly occurs as fine (0.01-0.1 ram), anhedral, semitranslucent grains. The common brown color in plane-polarized light appears to be due to an alteration mineral of unknown identity; the epidote is darkest where it is most altered. Massive epidote (clinozoisite) locally forms a creamy, dense (specific gravity about 3.1), hard rock that looks like a separate intrusive phase but is actually an alteration product. These areas appear to be the product of un- usually intense, early barren quartz-epidote-carbonate-prehnite stockworking that escaped later alteration.

The carbonate-albite-sericite intermediate zone is

divided into an outer texture-preserved and an inner texture-destroyed subzone. Carbonates (up to 65%) are usually the most abundant alteration minerals in the intermediate zone in altered diorite but are less

common in altered soda granite (Table 1). Albite may make up to 45 percent; chlorite remnants from the outer zone may not be completely replaced in the intermediate zone, forming up to 15 percent of the rock. The outer subzone, marked by the onset of carbonate alteration as a vein is approached, is distinguished in the diorite as a buff calcite replacement of hornblende crystals (Fig. 3a) that preserves the original rock texture but makes the rock look like a dif-

ferent phase. This transition from the outer chlorite- epidote zone to the onset of carbonate alteration roughly corresponds to the "cryptic-visible" boundary of Robert and Brown (1986b). The buff or creamy inner subzone is marked by progressive destruction of original igneous texture toward the vein. This sub-

FIG. 3. (a). Thiu sectiou view of outer subzone of central zone aheration (ab = albite, ca = calcite, eli = ohiorite, ms = muscovite), 3 m in the hanging wall, 79 vein, 41 level. Width of field of view is I cm; crossed polars. (b). Thin section view of inner subzone of central zone alteration, 0.2 m into footwall, 79 vein, 32 level. Large porphyroblastie grains of calcite with brigliter ankerite rims are set iu a matrix of sericite and calcite, plus minor chlorite that is dark. Width of field of view is 1.2 cm; crossed polars.

zone commonly weathers a rusty orange-brown due to increasing amounts of ankeritic carbonate (which does not react to cold dilute HC1) as the vein is approached. The range of ankerite-calcite ratios is illustrated in Figure 4a to d (iron and magnesium contents are confirmed by microprobe analyses in Table 2). Ankeritic carbonate forms anhedral interlocking grains ranging from 0.1 to 0.5 mm in diameter. Com- posite grains with clear calcite cores and ankerite rims are common (Fig. 3b). The lack of a strong X-ray peak for ankerite (Fig. 4a and b) in many samples may be due to the volumetrically minor ankeritic rim to the composite grains. Calcite is the only carbonate in late reopenings at the centers of major veins, in crosscutting veinlets, and in the early barren quartz-carbonate-epidote-prehnite stockwork. It may be of explo- 'ration significance that only the main-stage (ore-related) carbonate is ankeritic. Coarse secondary albite


5[3 84-49 / 794.5' •

, I • , , I , , , I 31 30 29

5B 8i-49/795' 31 30 29

b

19-51 FW1

31 3 2 C

C080

,,,1.

•3• 30 28 d

FIG. 4. X-ray diffraction scans of carbonate minerals from the intensely altered inner zone. (a). Sample SB-84-49 at 794.5 ft. (b). The same hole at 795 ft. Both show a small peak for ankerite compared to calcite. (c) Sample 19-51 FW1. It shows a larger ankerite peak. (d) Sample C080. It shows that ankerite is m. ore abundant than calcite. Full-scale deflection is 4 X 102 counts in (a) m•d (b) but l03 counts in (e) and (d).

long. It completely replaces original plagioclase and forms up to 30 percent of the rock. Bright green fuchsite is locally present in altered diorite and forms up to 5 percent of the rock, usually rimming cores of chromite. Green mica, variably identified as Cr-bearing, is also known in Archean mesothermal gold vein deposits such as the Casa Berardi (Pattison et al., 1986) and Lac Shortt (Morasse et al., 1986), both in Quebec. Carbonate includes siderite, ankerite, and minor calcite. The modal estimates of the carbonate minerals in Table 1 are tentative because of the dif-

ficulty in distinguishing them in thin section; even SEM studies failed to identify them confidently, and microprobe analyses were required.

Quartz-albite silica-flooded zones form in the central portions of the soda granite and in albitite dikes. This alteration is fracture controlled and grades in

is next in abundance after carbonate and forms grains up to 0.5 mm across, in veins, vein envelopes, or with replacement textures that include "chessboard," "patchwork," and "irregular" (untwinned) albite (Battey, 1951; Leitch, 1981). Albite alteration is more prevalent in the soda granite, reflecting the differences in original composition between the more marie diorite and the more felsic soda granite.

The quartz-sericite-carbonate innermost zone consists of intense replacement by secondary quartz, sericite (_ fuchsite), ankeritic carbonate, and pyrite adjacent to the vein. This zone, up to 50 cm wide, is schistose (Fig. 5a and b) or occasionally brecciated. Fine (0.1 mm) quartz forms up to 60 percent of the rock, although some of this, especially in the soda granite, is merely isochemically recrystallized. Chemical analyses corrected for volume changes confirm this observation, although the inclusion of minor quartz veins in the samples cause fluctuations in SiO2 content (cf. Sketchley and Sinclair, 1987a). Sericite (muscovite: analyses in Table 2) occurs as pale green to yellowish masses of flakes less than 0.05 mm

• .

'L ' '"P' ":*', ' -..•½• • :'•' .. :*% ..•

•C. 5. (a). Section o[core around the 5]• • vein on the • level, s•owin• the progressive destruction o[ texture in both t•e soda •ra)ite (top) an• •iorite (bottom) [roma coarse-•raine•, m•sive rote to a [oliateS, ki•!y •u•z-sedcite-earbonate-altered rock in the inner zone. The position of the vein is at the gap at the center of the box. (b). Thin section view of foliated, intensely qu•z-se•eite-carbonate-altered inner zone rock, from immediate footwall of the 51 vein, 8 level, near the Empire sh•. Field of view is 1 em wide; crossed polars.


intensity from a crackling or sparse stockwork of quartz-sericite-carbonate-chlorite-epidote-albite mi- croveinlets, to a breccia composed of a network of the same minerals enclosing fragments of soda granite, to a rock composed almost entirely of quartz and albite. Accompanying changes in feldspars proceed from chessboard albite through veinlet albite to stockworked soda granite (Fig. 6) and finally to an intensely silicified white rock composed of an unusual symplectic intergrowth of quartz and albite that locally destroys the original texture (Fig. 7). These in- tergrowths are up to 3 mm across, with a radiating, crudely hexagonal pattern, probably controlled by quartz crystallography because the quartz and albite crystallized simultaneously. Since the alteration is later than the albitite dikes, it appears to be hydrothermal. However, there are no well-defined quartz veins or gold values associated with these silica- flooded zones.

Biotite envelopes occur at Bralorne only near the surface in green hornblende porphyry dikes near the 51 vein on the 4 level and in altered diorite below

1,700 m depth around the 77 vein. It is more common to the southeast on the Pioneer (Joubin, 1948) and P. E. Gold (Nordine, 1983) properties where the Pi- oneer greenstones, prevalent on these two properties, alter more readily to biotite. The occurrence of biotite may be similar to that at the Sigma mine in Quebec (Robert and Brown, 1986) where biotite is prominent as an alteration mineral in the lower levels of the mine.

Consequently, the Pioneer and P. E. Gold exposures might represent deeper levels of the vein system than at Bralorne.

Black calcite replacement is a distinctive alteration present only locally in the soda granite, crosscutting the main-stage alteration. It begins along hairline fractures and may replace the whole rock. The black

FIG. 6. "Crackled" soda granite, cut by a stockwork of fine- grained quartz, sericite, calcite, epidote, pyrite, and minor ohiorite, from drill hole 8E-1174 at 880 ft. Crossed polars, field of view 1 era.

FIG. 7. Primary texture of soda granite replaced by symplectic growths of quartz and albite in a radiating, psuedohexagonal pattern (DDH SB-14-80/93 fl.). Field of view is 1 era; crossed polars.

and nearly opaque character is caused by myriads of extremely fine (1-2 jam) opaque inclusions that may be amorphous carbon. Tests by X-ray diffraction, SEM-EDS, and electron microprobe failed to confidently identify this material.

Garnet-quartz-calcite-pyrite and quartz-tourmaline (schorl) alteration was found in specimens collected from Pioneer volcanics in the Pioneer mine by F. R. Joubin, which are now in the University of British Columbia Economic Geology collection. Neither alteration facies is common, although tourmaline is abundant in late fractures cutting the extensional 27 vein (Joubin, 1948). The presence of borosilicate may be significant; although apparently rare at Bralorne, it is common in mesothermal systems in the Canadian Shield (e.g., Sigma: Robert and Brown, 1986a).

Mineral chemistry Chemical zonation in carbonates is from calcite in

the outer and intermediate zones to ankerite in the

inner zone (Table 2 and Fig. 8). In the least altered rocks (sample C094A), the carbonate is almost pure calcite (0.97 mole fraction CaCOa). Similar calcite is found in the intermediate zone (sample SB-84-49, 812 ft depth), 3 m away from the major 5lB vein. SEM- EDS studies revealed that inner zone carbonates are

composite grains with calcite cores rimmed by ankerite, or less commonly, ankerite cores with siderite rims (Fig. 9a and b). The fine grain size (equal to the beam diameter of the probe) causes contamination of some ankerite analyses by calcite and of siderite analyses by ankerite (Fig. 8), resulting in the apparently metastable plotted points (cf. Anovitz and Essene, 1987). Adjacent to the vein (sample SB-84-49, 795 ft depth), much of the carbonate is ankerite, with the cores of 0.03-mm-diameter grains (spots 3 and 7, Fig. 9a and b; analyses in Table 2) being about Cco.•Dolo.aSido.• (Cc, Dol, and Sid indicate mole frae-


Calc• le

SB 84-49/795• 812' 19-51FW1, Cll 6-1

Ca( O 3 CALCITE Ca[c• te

DOLOMITE

cores

• SB 84-49 / 795' Ankente

r., m .s•_._.../ ANKERITE

SI DERI TE •_g•] S,der,te r,ms •SB 84-49/795 . ß -

MC

..

FeCO 3

FIG. 8. Triangular diagram showing fields of carbonate compositions from altered wall rocks (atomic proportions). Analyses are given in Table 2. Stability fields at 400øC (stippled) for calcite and siderite (calcite group) and for ankerite (dolomite group) are from Anovitz and Essene (1987).

tion CaCO3, MgCO3, and FeCO3, respectively). The rims of these grains are only 5 •m thick but are distinguished in thin section by higher relief than the cores. These rims (spots 8 and 2, Table 2) are Mg and Fe rich (Cc0.0s_0.3.sDo10.sSid0. Ls_0.45), and plot as sid- erites (Fig. 8) with higher Fe/(Fe + Mg) of around 0.4 than the 0.1 to 0.2 ratios typical of the ankeritic core. However, even adjacent to this major vein a few grains of calcite are present (Table 2, spot 9; see also sample Cl16-1 adjacent to the 79 vein on the 41 level). Analyses from the inner zone around the 51 vein on the 19 level (sample 19-51 FW1, Table 2) are similar to those just discussed. Calcite probably formed initially but was progressively replaced with iron-magnesium carbonate by advancing alteration; calcite formed simultaneously farther from the vein. This explains the bulk zoning and zoning in the carbonate grains from Fe-Mg-rich rims to Ca-rich cores. The Ca, Fe, and Mg in the carbonates was probably derived from primary hornblende. However, a portion of the Ca, Fe, and Mg became available for deposition elsewhere, since total Ca, Fe, and in particular Mg, show depletion near the veins (see below). Zoning of calcite farther from, and ankerite closer to, mesothermal gold veins has been described at other deposits. It is noted at the Yellowknife district (Boyle, 1961); Lac Shortt, Quebec (Morasse et al., 1986); Casa Ber- ardi, Quebec (Pattison et al., 1986); the McDermott deposit at Kirkland Lake, Ontario (Workman, 1986);

Cameron Lake, Ontario (Melling et al., 1986); and the Red Lake district, Ontario (MacGeehan et al., 1989.).

Feldspars are ubiquitous, almost pure albite, as confirmed by X-ray, SEM, and microprobe (Table 2); no K feldspar was found. The average albite in least altered rock (samples C093A and C094A) is Ab•6_, with a negligible orthoclase molecule; this may be greenschist facies metamorphic albite. There is a slight increase in the albite molecule, to Ab,_•00, in the texturally distinct alteration albite in and around veins (samples 19-51 FW1 and C182-2, respectively, in Table 2). The very low anorthite component of both the metamorphic and the hydrothermal plagioclase implies a formation temperature of less than 400øC (Winkler, 1971). This agrees with the estimates of ore fluid temperatures of 350øC from fluid inclusion and sulfur isotope studies (below). The narrow range of compositional variability observed suggests thorough homogenization of the feldspars by metamorphic and hydrothermal fluids.

Micas in all the vein envelope zones (inner to outer) are primarily sericite (muscovite). Sericites analyzed were from various rock types and various parts of the deposit (sample SB-84-49, 795-ft depth in green hornblende porphyry, C 182-2 in crackled soda granite, 19-51 FW1 and Cl16-1 in diorite: Table 2). They all have similar compositions, with low Fe, Mg, and Ti contents. Only one analysis of biotite is reported


FIC. 9. (a) and (b). SEM-EDS backscattered pictures of composite carbonate grains in sample SB-84-49 at 795 ft, found in the inner zone alteration of diorite in the hanging wall of the 5lB FW vein near the 5 level. The grains show ankerite cores (gray; spots 3 and 7) and siderite rims (brighter due to higher atomic number of Fe compared to calcium and magnesium; spots 2 and 8). Dark spots are holes; analyses are given in Table 2.

(Table 2), from an unusual biotitic envelope in a green hornblende porphyry dike on the 4 level near the 5lB vein.

Chlorite, distinguished optically, is pennine or prochlorite; the two types are also chemically distinct. Pennine, present only in the least altered soda granite (C094A), is a high Fe chlorite with Fe/(Fe + Mg) about 0.7 and Si (cations) about 5.5 that approximates aphrosiderite, a variety ofripidolite (Table 2 and Fig. 10). A few analyses plot as pseudothuringite, with Si cations about 4.7 and Fe/(Fe + Mg) about 0.6. Pro- chlorite, more abundant than pennine, occurs both in unaltered diorite (sample C093A: Si cations about 5.6-6.0) and strongly altered diorite (C 116-1) and in soda granite (C182-2: Si cations about 5.0 to 5.4). It is richer in Mg with Fe/(Fe + Mg) about 0.35, i.e.,

between ripidolite and pycnochlorite (Fig. 10); one analysis plots as corundophillite. The general ripidolite-pseudothuringite compositions are similar to those described by Robert and Brown (1986b) for chlorite at the Sigma gold quartz vein deposit in Quebec.

The original, possibly metamorphic, prochlorite in the diorite is thus more magnesian than that in the soda granite. This magnesian chlorite is also stable in the presence of hydrothermal fluids, as it is present in altered diorite and soda granite. It implies a high Mg/Fe ratio in the hydrothermal fluids, which is supported by the results of computer modeling (below). Analyses of hornblende (Table 2), the precursor mineral to the chlorite in the diorite, also show Mg/Fe ratios greater than unity. The similarity of Mg/Fe ratios in the precursor hornblende and alteration chlorite suggests that the chlorite replaced some hornblende grains on a one-for-one basis, although in other grains, some Fe, Mg, and Ca cations were trapped as carbonate or released to the fluid.

Chemical changes

Losses and gains of elements during alteration, corrected for volume changes (Leitch and Day, 1990), were determined by Gresens' (1967) technique from whole-rock analyses (Leitch and Godwin, 1988) along ten detailed traverses across envelopes in the footwall and hanging wall of major veins. Traverses were made in both soda granite and diorite host rocks, from surface to the 44 level, an interval of 2,000 m. This cor-

0'8

0'67

6-O

FIG. 10. Classification of the observed chlorite compositions, according to the scheme of Hey (1954). A -- aphrosiderite, filled circles -- least altered host rocks, crosses = altered rocks.


responds to studies elsewhere (Kerrich and Watson, 1984; Robert and Brown, 1986a; Sketchley and Sin- clair, 1987).

Two examples of the chemical changes adjacent to the major veins are in Figure 11a and b. There are no significant differences in gains and losses between the footwall and the hanging wall. Other traverses (Leitch, 1989) show the same general patterns. The main changes within the alteration envelopes on ap- proaching the vein, with mineralogical explanations, are (1) loss of Na20 and gain of K20, due to destruction of albite in the inner zone and sericite replacement; (2) loss of MgO and Fe•O3, due to destruction of amphibole and replacement by sericite and carbonate; (3) net loss of SiO• in the inner and inter-

VOLUME

FACTOR

VOLUME

FAC TO R

Sampl e.

160t AS ..... /No. ß • • / \",, coo3/2 1.20._•_ .......... • ................. _'•__ '1 O'801 .""- • /3

AI203 TiO 2 Zr' Avg.

O•/ .......... •

AI203 Ti 02 Zr Avg,

Fe203 2 2

-2

-

K2 o

0:5 •b D•stance toVe,n (m)

K20

i i i i i 2 4 6 8 10 1• Distance to vein (m)

FIG. 11. Oresens' plots of weight percent loss or gain of oxides, plotted versus distance from the vein, from program ORESPLOT (Leitch and Day, 1990). Volume factor is based on alumina and titania contents in sample C093, the least altered diorite host, from drill hole UB-81o17 at 350 to 400 ft. (a). Hanging wall of the 51 vein, 8 level, near the Empire shaft. (b). Footwall of the 51 vein, 15 level, near the Crown shaft.

FIG. 12. Volume factor plot (ratio of final volume to initial volume for immobile components) for sample series C003 (a) and C032 (b). Avg = average of volume factors for AI2Oa, TiO2, Zr: losses and gains are plotted in Figure 11.

mediate zones of the vein envelope--this is variable, due to quartz veinlets being included in some samples as the major vein is approached; and (4) increase in loss on ignition (L.O.I.) and CaO due to development of sericite and carbonate in all zones. Increases are

not always linear, depending on composition of the original host rock and the amount of included quartz vein. CaO commonly increases in the outer envelope as calcite becomes stable and decreases in the inter-

mediate and inner envelopes as ankerite becomes stable. The parallel increase of L.O.I. and the calculated volume factor implies that the changes in volume estimated from immobile elements are real, and therefore, the calculated losses and gains in other elements reflect real processes.

The similarity of detailed traverses down to the 44 level indicates that there is no major change in envelope alteration chemistry from the surface to a 2,000-m depth; thin section examination supports this conclusion. Differences between the way diorite and soda granite react to alteration are slight. In the altered soda granite there is a smaller loss of Fe•O3 and MgO, since there are only minor marlcs to destroy, and a greater loss of Na•O, since there is more albite to destroy. Changes in CaO are less noticeable in the soda granite, again since most of the CaO is in mafic minerals in both soda granite and diorite, and the former is less mafic.

Diagrams of losses and gains (Fig. 11) demonstrate the relative immobility of A1203 and TiO•, the components used to estimate volume changes (Gresens, 1967). An example of the volume changes or "volume factor" involved in alteration, for the altered series


of Figure 11a, is given in Figure 12a. In this case there is an initial volume decrease in the intermediate

zone sample (C003/3), possibly due to destruction of albite and hornblende. This is followed by a volume increase in the inner zone samples immediately adjacent to the vein (C003/1,2), caused by strong carbonate alteration as shown by correlated increases in L.O.I. and CaO (Leitch and Day, 1990). Figure 12b also shows the typical agreement in using A1203 or TiO2 in estimation of volume change and the consistent overestimation of volume factor using Zr. In such cases, Zr is rejected from the computation of the average volume factor used in the profiles.

Minor elements also show patterns. There is a marked increase in S, As, and Sb as the veins are approached (Leitch and Godwin, 1988). This is associated with the addition of pyrite, arsenopyrite, and tetrahedrite (like COz, HzO, and KzO) in immediate vein envelopes, by hydrothermal fluids rather than mere remobilization (like CaO, MgO, and FezO3: cf. Ludden et al., 1984). Chalcophile elements, such as Cu and Zn, show a slight decrease toward the vein or are constant. Lithophile elements follow the patterns of certain major elements due to substitution: Ba and Rb follow K20, whereas Sr mimics CaO. Yttrium, like Zr, shows mainly immobile behavior, falling as the volume factor increases and vice versa. Cr and Ni, dependent on the amount of mafic inclusions in the diorite, show correlated behavior. Co and V do not show discernible patterns. Elements present at or near the detection limit (Pb, Nb, Mo, Ag, and W) are erratic.

These patterns are broadly similar to those detailed by Boyle (1979) in his study of the chemical changes attending alteration of the diorite at Bralorne (corrected only for density differences, not for volume changes). The changes are also similar to those for the Yellowknife deposits (Boyle, 1961), which show a marked increase in KzO/NazO and a consistent decrease in SiO2/COz and SiOz/L.O.I. near the veins. Similar chemical profiles are reported for mesothermal gold vein deposits at Macassa in Quebec (Kerrich and Watson, 1984), Kerr-Addison in Ontario (Kishida and Kerrich, 1987), Wawa in Ontario (Studemeister and Kilias, 1987), Cameron Lake in Ontario (Melling et al., 1986), and the Mother Lode in California (Al- bino, 1987). Most vein-type Archcan gold deposits in the Canadian Shield have similar patterns of alteration (Kerrich, 1983).

Fluid Inclusion Studies

Samples for fluid inclusion studies were collected from quartz veins at surface and over a 2-km depth. Sites were the same as those for oxygen isotope sam- pies. Vein quartz forms euhedral crystals up to i cm long (Fig. 13) that are outlined by concentric growth

•.•./laye ring

sulf• des i'•• calci t e quartz k../•clear quartz

J (.•.7-- • lcrn •

cloudy quart z--• (

s,lfides--• •'•/ I 0.5 C rn I

(b)

FIG. 13. Sketch of textures in quartz from the main, ribboned shear veins in the Bralorne deposit; vein walls are oriented east- west in both diagrams. (a). Growth zoning in coarse euhedral quartz crystals that grew perpendicular to vein walls and ribbons (Woodchuck vein, 3 level adit portal). (b). Growth zoning in large euhedral quartz crystals (79 vein, 41 level, Queen mine area, hole Q-144 at 201 ft. Note also the small clear euhedral crystals enclosed in the larger crystals.

zones of minute (1-2 gm diameter) primary fluid inclusions (Fig. 14). None of these inclusions were large enough for microthermometry. Instead, larger isolated (assumed primary) and fracture-controlled (pseudosecondary) inclusions were studied (Fig. 15). Fluid inclusions were measured from the two main

paragenetically distinct types of quartz found in the veins: gray main-stage quartz adjacent to sulfides, and clear quartz, both earlier and later. The main mass of milky quartz provided useable fluid inclusions only in local areas where it was clearer with fewer but

larger inclusions or along fracture planes where larger inclusions were present. Fluid inclusions were also measured in late-stage calcite, mainly as ribbons in the quartz veins, less commonly in apparently contemporaneous calcite grains included within the quartz veins. No particles of gold were seen directly associated with any fluid inclusions.

Preliminary fluid inclusion data for the Bralorne deposit were obtained with Chaixmeca equipment (Leitch and Godwin, 1987, 1988). Further studies, carried out with a FLUID Inc.-adapted U.S.G.S. type gas-flow stage calibrated to better than _0.4øC from -56.6 ø to +660.4øC, characterize the compositions of the fluids in the inclusions in greater detail. Mole fractions of carbonic fluid are derived from visually


FIG. 14. Growth zones outlined by primary fluid inclusions in quartz from the 79 vein on the 32 level. Note hexagonally arranged 1-tam primary inclusions and swarms of 3- to 5-tam see- ondary fluid inclusions on microfractures, that give the quartz its milky color. Width of field of view (a) is 2.5 ram, 1.25 mm in enlargement of lower right-hand portion (b).

estimated vol percent at 25øC (Fig. 16). Although there are uncertainties inherent in this method, pre- cise calculation by the method of Parry (1986) does not change the results significantly for the more common inclusions wi•h modest carbonic contents. Details of calibration and error analysis are in Leitch (1989).

Classification and description of fluid inclusions

Inclusions were classed as primary, pseudosecondary, and secondary by the criteria in Roedder (1984; of. Hollister, 1981). Results for the major types of inclusions suitable for measurements are summarized

in Table 3 and Figure 16. The visual division into primary and pseudosecondary is supported by the general bimodal distribution of Th and T• data in Table 4 (summarized from histograms in Leitch and Godwin, 1988).

Primary (type 1) inclusions display spheroid or negative crystal shapes in quartz (Fig. 15), or are rhomboid in calcite. They range from 3 to 60 /•m across and occur as isolated inclusions or in discrete

clusters not in linear arrays. Many contain three visible phases (Fig. 17): an outer aqueous liquid, an inner carbonic liquid, and an innermost carbonic vapor bubble. Some appear to be simple two-phase inclusions since small amounts of carbonic fluid rimming the vapor bubble are hard to see (but are confirmed by CO• or clathrate melting). The presence of another component in the type 1 primary inclusions, probably mostly methane (CH4) but possibly including trace amounts of hydrogen sulfide (HoS), is indicated by the depression of the melting point of CO9•.

Type la inclusions in quartz, marked "P" in Figure 16, form the majority of the three-phase inclusions

FIG. 15. Isolated, and therefore assumed primary, type la COu-bearing inclusions of typical 5- to 10-tam size, in quartz from breccia portion of the 51 vein on the 8 level. Trails of pseudosecondary and secondary inclusions (types 2 and 3) are 3- to 5- tam size (width of field of view is 200 tam). Higher vapor/liquid ratios distinguish the type I inclusions, which may be either negative crystal shape as in (a) or rounded as in (b). Irregular inclusions in (b) are type 2 (lower vapor/liquid ratio).


20

15

-70 -65 -60 -55

Tmco 2

I.Id . o • ,o ,• •o •, •o•

Thco 2

3O

20

10

0 -1 -2 -3 -& -• -• Tm ic ß

30

20

•0

, [- , o' 5 10 15

Trnc lat hrat ß

-4O -35 -30 -25 -20

Teut ect•c

0 10 20 30 40 50 60 70 60 90 10C V/V+ L (vol %)

FIG. 16. Summary of data for carbonic fluid inclusions for the Bralorne deposit. P, PS, S correspond to primary, pseudosecondary, and secondary inclusions, respectively; black areas to methane-rich (type lb) inclusions. All inclusions are in quartz except for hachured areas which are in calcite. Other abbreviations are given in Table 3.

and usually homogenize to liquid. They contain modest amounts of carbonic fluid (0.1-0.4 mole fraction of the total contents of the inclusions, with a mode of 0.15) and consist of an aqueous phase, carbon dioxide, and minor methane (Xco2 = 0.10, Xcu4 = 0.05). The volume percent of carbonic fluid represents the H20- CO2 ratio at the time of trapping. Variations in carbonic fluid could be due to any one of, or a combination of, necking, trapping of immiscible fluids, or fluid mixing.

Type lb inclusions in quartz, shown in black in Figure 16, are rare, with significantly higher carbonic fluid contents than type 1 a; they usually homogenize to vapor. They are vapor rich and contain a minor aqueous phase plus abundant methane and carbon dioxide (0.3-0.9 mole fraction carbonic component of the total contents, with a mode of 0.5, consisting of Xcu4 = 0.25, Xco2 = 0.25). They occur both in clear quartz that may be fragments included in the main-stage milky quartz, or in main-stage quartz a few millimeters from, but not mixed with, the more common type la inclusions. The high CH4 contents of type lb inclusions are evidenced by strong depression of COe melting points to below -60øC (Fig. 16). The COe homogenization temperatures of type lb inclusions are lower than those of type la, but type lb densities are lower (Table 3) because of high CH4 contents.

Primary inclusions in calcite (hachured in Fig. 16) have homogenization temperatures and fluid compositions similar to pseudosecondary (type 2) inclu-

sions in quartz. They generally have a minor COe content and occur as isolated inclusions or as clusters

not related to fractures. Because most calcite is paragenetically late, it seems reasonable that the primary inclusions in calcite and the pseudosecondary inclusions in quartz both represent the same phase of later, lower temperature, more dilute fluids.

Pseudosecondary (type 2) inclusions are the most common in the Bralorne vein quartz, arranged along myriads of tiny fractures that crisscross the quartz grains with a brushlike or wispy texture. This texture, together with the presence of the primary three-phase inclusions, indicates a mesothermal environment (T. J. Reynolds, pers. commun., 1987). Many of these fractures do not cross grain boundaries; thus, the fluid inclusions in them represent fluids trapped in fractures during growth of the quartz host. Type 2 inclusions in quartz, marked "PS" in Figure 16, homogenize to the liquid phase and contain no methane and only minor carbon dioxide (mode of less than 0.1 mole fraction of the total contents) that may not be visible as a separate phase but is inferred from clathrate melting temperatures. They range from 3 to 20 across, averaging about 5/zm, and have rounded to irregular shapes (Fig. 18) or flattened lensold shapes which may display the highly irregular outlines typical of necking. Even if these inclusions have necked, they still have homogenization temperatures similar to others in the same fracture.

Secondary (type 3) inclusions in both quartz and calcite are localized along through-going fractures,


TABLE 3. Summary of Fluid Inclusion Characteristics for the Bralorne Deposit, Southwest British Columbia

(all temperatures in øC)

Type la primary: three-phase, CO2-bearing, CH4-poor M.F.: H20 = 0.8, CO• = 0.1, CH4 = 0.05, NaC1 = 0.03 Th: range = 235 to 425, average of modes 1 = 280 Thcol: range = 15 to 27, mode • = 20.5 Tmo: range -- 8.2 to 11.0, mode = 9.6

Tm, (.9): range = -2 to -7, average= -3.5 Tmc%: range -- -57.3 to -59.8, mode = -58.8 V/(V + L): range = 10-45 vol %, mode = 20 vol % r carbonic: 0.7 g/cm a r bulk: 0.95 g/cm a Vbar: 23 moles/cm a

Type lb primary: Three-phase, CO•-CH4-rich M.F.I: H•O = 0.45, CO2 = 0.25, CH4 = 0.25,

NaCI = 0.03

Td: range = 230 to 330, mode = 270 Thco,: range = 3.3 to 9.0, mode = 7.5 Tmc•: range = 11.1 to 13.0, mode -- 12.4 Tml (?): not observable

Tmco,: range = -61.3 to -66.5, mode = -63(?) V/(V + L): range = 30-85 vol %, mode = 50 vol % r carbonic: 0.6 g/cm a r bulk: 0.8 g/era a Vbar: 31 moles/cm ø

Type 2 psuedosecondary: Two-phase, HaO-liquid-rich M.F.l: H20 -- 0.95, COa = 0.05, NaC1 < 0.01 Th: range = 160 to 260, average of modes 1 = 200 Thco,: range = 20 to 31, mode = 27.5 Tmo: range = 7.0 to 10.5, mode = 9.5 Trot (.9): range = -0.5 to -2.5, average = -1.8 Tracon: range = -56.6.to -57.5, mode = -56.8 V/(V + L): range = 5-15 vol %, mode = 10 vol % r carbonic: 0.7 g/era a r bulk: 0.97 g/cm a Vbar: 20 moles/cm a

Type 3 Secondary: Two-phase, H•O-liquid-rich M.F.I: H•O = 0.99, NaCI < 0.01 Th: range = 120 to 180, average = 150 Tm,: range = -0.1 to - 1.0, average = -0.5 V/(V + L): 2-7 vol %, mode = 5 vol % r bulk: 1.00 g/cm a Vbar: 18 moles/era a

1 Average Of Th histogram means in Table 4; mean value is 280, standard deviation +__ 36, 276 measurements

2 Mode is from histograms in Figure 16 Abbreviations: M.F. = approximate mole fraction, Th = final

homogenization temperature, Td = decrepitation temperature, Thco, = homogenization temperature of CO•, Tmcl = final melting poiut of clathrate, Tmt• = filial melting point of ice, Tincol = final melting point ofCO•, V/(V + L) -- ratio of vapor bubble to liquid, r = density, Vbar = molar volume

some oriented with their long axes perpendicular to their controlling fracture (cf. Roedder, 1979). These inclusions trapped fluid that was circulating after the quartz had been deposited and fractured. They are simple two-phase aqueous inclusions with small (5 vol

%) vapor bubbles, no detectable carbonic content, and they homogenize to the liquid phase. They are generally small (less than 10 t•m), with lensoidal, an- gular, or negative crystal shapes.

Homogenization temperatures

Homogenization temperatures (Th) for fluid inclusions in quartz and calcite range from 120 ø to 440øC. Most inclusions homogenize to the liquid phase; only a few to the vapor phase. This is unlikely to be due to necking, because the Th for liquid and vapor is similar. Inclusions homogenizing to vapor (mainly type lb vapor- and CH4-rieh inclusions) commonly de- erepirated before homogenization. For type la and type 2 inclusions the few deerepitation temperatures (Ta) measured were usually close to the Th values of other inclusions in the same duster. Only a minimum estimate of 230 ø to 330ø½ (mode 270øC) can be made for the trapping temperature of inclusions that de- erepirated.

Two main populations of Th are indicated in Table 4. The higher temperature group, with a total range of 240 ø to 420øC (avg 280øC), corresponds to the type 1 primary inclusions. The lower temperature peak, at 180 ø to 220øC (avg 200øC), corresponds to the type 2 pseudosecondary inclusions. Occasionally, a weak third peak at lower temperatures (130 ø- 160øC, avg 150øC) corresponds to type 3 secondary inclusions. A similar pattern of primary homogeni- zations at 290 ø to 400øC, with early secondary ho- mogenizations at 250øC and later secondary homog- enizations at 190øC, was observed at the Kolar deposit in India, which is similar to Bralorne in strike (8 km) and depth (3.2 kin) extent (Santosh, 1986). A similar bimodal distribution in Th was found at the Big Hurrah deposit in Alaska (Read and Meinert, 1986) and for homogenization and ice-melting temperatures in the mesothermal Au deposits of the Jungwon area of Ko- rea (Shelton et al., 1988).

With one notable exception (sample 8-51B), in which clear quartz (late vug crystals or early fragments) has anomalously high homogenization temperatures, there is a tendency for T h in primary (type 1) inclusions to increase with depth, from about 235 ø to around 330øC. There is, however, overlap in the temperatures and the zonation would not be con- vincing without the corroboration of deerepitation studies (Sugiyama, 1986; Table 4). The deerepitation method suffers from difficulties relating data to specific inclusion types and to significant or variable overshoot of Th depending on size of the inclusions and the grains studied. However, the advantages are that data may be rapidly gained from many inclusions, too small to measure optically. Optically studied see- tions from surface to the 2,000-m depth show general agreement between Ta and Th (Leiteh, 1989). The gradient suggested by the deerepitation data is about


TABLE 4. Summary of Fluid Inclusion Data, Bralorne Gold Quartz Vein System, Southwestern British Columbia

Level (vein) Depth (m)

Tillice (øC) Th (øC)

Pseudo- Psudo-

Primary secondary Secondary Primary secondary Sample no. Type 1 Type 2 Type 3 Type 1 Type 2

Secondary Type 3

0 (Lorne) C1013 (Cosmopolitan) C1002

0

3 (Woodchuck) C1024 (5lB) C1027

130

8 (5lB FW) (51B Bxa) (51B)

270

15 (51) 580

16 (51) 16-5 i•vV(D 630

26 (85) 1,100

32 (79) 1,400

-2.o _ 1.5

-1.,5 + 1.0 (27) -2.2 + 0.7 (11)

8-51B(V"vV) -2.0 + 1.1 (8) 8-51B -7 + 1..5 (11) -2.5 + 1.7 (39) cos1-1/2

290+ 15(10) 230+30(5) 280 +_ 20 (7) 325-335

235 + 25 (21) 195 q- 25 (9) 250 + 30 (16) 345-380 175

310 i 60 (46) 225+ 15 (5) 270 q- 30 (26)

?200 (4) ½ 350-430

140 q- 20 (16)

15-51(C) -2.5 q- 1.3 (9) 260 + 20 (21) 360-400

-2.3 q- 1.5 (20) -0.5 q- 0.2 (6) ?,280 (2) -1.5 + 0.6 (6) e 220 + 25

Cl18-11 -1.6 q- 0.6 (11) 300 q- 40 365-385

(9) •

(20)

220

190 q- 20 (4) 190

190 q- 20 (18) 160 q- 20 (5) lS0

225 q- 25 (6) 230

Cl17-7 -4 + 1.0 (14) -1.5 q- 0.5 (8) 320 q- 60 (15) 190 _ 30 (3,2) 7140 (1) -2.0 q- 1.o (9) e 205 _+ 35 (16) c

340-385 180

Cl17-5 -3.5 q- 1.5 (9) -1.5 q- 0.7 (18) 270 q- 40 (18) 180 q- 30 (26)

41 (79) Cl16-14 -3.5 q- 1.7 (15) -0.5 q- 0.2 (8) ?265 q- 15 (9) 215 q- 20 (5) 7150 (1) 1,8oo 57o-as5 21o

44 (77) C128-20 -5.0 (3) -1.5 q- 0.8 (11) 280 q- 50 (,25) 190 q- 40 (10) 2,000 C128-19 -5.0 q- 1.1 (9) -1.5 q- 1.3 (29) 330 q- 60 (21) 200 q- 40 (45)

410-425 200

T .... (melting point of ice) figures can only be regarded as crude estimates (see text for details); Tl• = homogenization temperature Data are reported as mean q- one standard deviation, followed by number of determinations in brackets; ? indicates uncertain data

(too few to calculate standard deviation); italic figures beneath Th are decrepitation data (Sugiyama, 1986), with Ta (start of decrepitation) followed by a dash and Tp (peak of decrepitation)

• Measured in calcite; all others from quartz

30øC/km, which corresponds to a reasonable geothermal gradient (cf. Goldfarb et al., 1988). The lower temperature type 2 and 3 inclusions show a significantly lower gradient of 0 ø to 10øC/km.

Salinities (ice-melting temperatures)

Salinities of type 1 and 2 inclusions at Bralorne are difficult to estimate because of the presence of variable amounts of both carbon dioxide and methane (cf. Collins, 1979). Thus the T,,• e data for type 1 and 2 inclusions (5 and 3 wt % NaC1 equiv, respectively; Table 4) overestimates the salinity of the aqueous solutions using the equation of Potter et al. (1978). If only carbon dioxide is present, the salinity can be correctly estimated from clathrate-melting temperatures (Bozzo et al., 1973). A clathrate-melting temperature of 9.5øC (mode from P, PS peak in Fig. 16),

suggests 1 wt percent NaC1 equiv for the type 1 and 2 inclusions in quartz and calcite. However, in type lb inclusions, ice melting is not seen, and clathrate melting is above 10øC. This reflects methane and indicates that the method of Bozzo et al. (1973) is not applicable. Correlation between mole fraction CH4 and Tinice in type la and lb inclusions suggests roughly equivalent salinities (cf. Linnen, 1985). Therefore the salinity of the type 1 inclusions is between 1 and 5 (i.e., 3 _ 2) wt percent NaC1 equiv. Type 3 inclusions have ice-melting temperatures just below that of pure water (-0.5øC), implying 0.8 wt percent NaC1 equiv salinities. Vaguely detectable eutectic melting in inclusions in quartz (Fig. 16) shows modes at -20.5øC in type 2, suggesting only NaC1, and -23øC in type 1, suggesting the presence of minor KC1 (Roedder, 1984). This conclusion is supported by the computer

BRALORNE Au DEPOSIT, B. (2. 335

;-. .

,

.

FIG. 17. Large (30-50 #m) three-phase type la inclusions in quartz from the 51B FW vein on the 8 level. The inclusions contain an easily visible innermost carbonic (COg + CH4) vapor bubble, surrounded by a carbonic liquid phase and an outer aqueous phase (at 25øC). In most of the inclusions, vapor/liquid ratios are about 20 to 50 vol percent; molar carbonic fractions are about 0.15 to 0.25. Width of field of view: 375 #m in (a), 200 #m in (b).

type' 2. Bulk densities, equivalent mole fraction CO2, and molar volumes (V•,•) for the inclusions were estimated from Brown and Lamb (1986), assuming a density of 1.02 g/cm a for a 3 wt percent NaC1 equiv solution (Potter and Brown, 1977). Equivalent CO2 contents for inclusions containing both methane and carbon dioxide were computed by Swanenberg's (1979) method.

Pressure estimates

Entrapment pressures for primary and pseudosecondary fluids, estimated by several methods, are compared in Table 5. Independent estimates of trapping temperatures (Tt) from sulfur isotope fractionations (350øC for primary inclusions and 250øC for pseudosecondary inclusions, see below) give pressure estimates of 1 and 0.5 kbars, respectively, based on the difference between Th and Tt of 70 ø and 50øC for primary and pseudosecondary inclusions, and the

modeling of the ore fluid, below, which suggests an Na/K ratio of about 8:1.

Densities

Data in Table 3 indicate that the carbonic (non- aqueous) component of type lb methane-rich inclusions has a lower density (mode 0.60, range 0.57- 0.64) than in type la carbon dioxide-rich inclusions (mode 0.70, range 0.67-0.75) or type 2 inclusions (mode 0.70, range 0.63-0.75). These densities are based on the temperatures of homogenization of the carbonic portion of the inclusion (Thco2) and data from Swanenberg (1979) and Hollister (1981). The Thco2 is difficult to measure reliably for type lb inclusions because of the overlap with the clathrate-melting temperatures, but it is well defined for type l a and

FI•;. 18. Typical pseudosecondary (type 2) inclusions, without visible carbonic fluid (i.e., containing only two phases, aqueous liquid and vapor, with vapor/liquid ratios about 5 to 10 vol percent) in quartz from the 51 vein on the 16 level. The characteristic planar distribution along healed microfractures is shown in (a), in which the field of view is 400 #m. An enlargement of the lower left portion (field of view = 160 #m) is shown in (b).


TABLE 5. Pressure Estimates for the Bralorne Deposit Gathered from the Different Approaches Used

in the References Listed

Primary luciusions Pseudosecondary inclusions

Type la Type lb Type 2 kbars (MPa) kbars (MPa) kbars (MPa) Reference

1.o (100) o.s (so) 1 >1.0 (>100) >1.0(>100) 2

1.5 (•50) 0.7,5 (7,5) 3 1.75(175) 1.5(150) 1.O (lO0) 4

Figures given are rounded off; uncertainty is ñ0.25 kbars References (see text for details of the method used in each

case): 1 = Leitch and Godwin (1988), 2 = Leroy (1979), 3 = Hol- lister and Burruss (1975), 4 = Bowers and Helgeson (1983b)

assumption of pure HsO (Roedder, 1984). However, such estimates do not consider the presence of carbonic fluid. Estimates of the bulk density of these inclusion fluids define isochores on a pressure-temperature (P-T) plot, which can be projected to the estimated entrapment temperatures to provide trapping pressures of 1.5 and 0.75 kbars, respectively (Hollis- ter and Burruss, 1976).

The P-T conditions of entrapment of the primary and pseudosecondary fluids can be evaluated more rigorously using solubility relations for HsO and COs in salt solution (Bowers and Helgeson, 1983a). Fluids in type l a inclusions, with mole fractions of Xcoa = 0.1 and XCH4 = 0.05, salinities of about 3 wt percent NaC1 equiv, and Th less than or equal to 300øC, would have been supercritical at pressures above 1.5 kbars. Projection of the 0.95 g/cm 3 isochore to the solvus for these inclusion compositions yields a pressure of entrapment of 1.75 kbars at the estimated trapping temperature of 350øC (Bowers and Helgeson, 1983b). Similarly, type 2 fluids would have been supercritical above 0.75 kbars; projection of the 0.97 g/cm 3 isochore to the solvus for these inclusions at the estimated trapping temperature of 250øC con- strains pressure of entrapment to 1.0 kbar, lower than we suggest for most vein filling. For type lb fluids (mean carbonic mole fraction of 0.5, bulk density of 0.8 g/cm • and salinity of 3 _ 2 wt % NaC1 equiv) decrepitation temperatures of 230 ø to 300øC give minimum estimates of trapping pressures of 1.5 kbars.

The presence of occasional vapor- and CH4-rich type lb inclusions with the far more abundant vapor- and CH4-poor type 1 a inclusions raises the possibility of trapping at subcritical temperatures. However, type lb inclusions are so rare in comparison to type la that most trapping was probably at or above the solvus. Also, continued fault movement at lower temperatures could have produced heterogeneous populations of inclusions when earlier inclusions were broken.

Most type lb (vapor-rich), and a few type la, inclusions decrepitate before homogenizing at temperatures ranging from 230 ø to 330øC. Deerepitation over such a small range of temperature is consistent with rapid pressure increases in carbonic fluids at temperatures above 250øC (Malinin, 1974). Since in- ternal pressures of up to 1.2 kbars are required to decrepitate 12- to 13-#m-size inclusions in quartz, and up to 2.7 kbars for smaller inclusions (Leroy, 1979), minimum trapping pressures implied are in agreement with the pressure estimates above.

Evolution of mineralizing fluids Paragenetie relations are obvious between type 2

and 3 inclusions, but less certain between type 1 and type 2. However, based on the temperature drop from type 1 to type 2, a progression in fluid composition at Bralorne from type 1 to type 3 is postulated as follows. Type 1 primary (?early) fluids were high temperature (350øC) and carbon dioxide and methane rich (Xco• + XCH4 = 0.15), with low salinities (approximately 3 wt % NaCl equiv). Type 2 pseudosecondary fluids were lower temperature (250øC), with less carbon dioxide (Xco• = 0.05), no detectable methane, and had lower salinities (less than 1 wt % NaG1 equiv). Later type 3 secondary fluids were cool- est (180øC), with no detectable carbon dioxide (<0.03 mole fraction), but the same low salinity as the pseudosecondary fluids. The fluid evolution can be explained by mixing of hotter, more saline, carbon dioxide- and methane-rich fluid with cold, more dilute meteoric water as the hydrothermal system waned.

Estimated pressures of entrapment are lower for type 2 pseudosecondary inclusions than for type 1 primary inclusions, and the geothermal gradient was probably lower (< 10øC/km) at the time of entrapment of the pseudosecondary fluids than it was for the primary fluids (30øC/km). These two features suggest that the deposit had been partly unroofed and the rock mass had cooled by the time fluids in the pseudosecondary and secondary inclusions were trapped. Pressure estimates from primary and pseudosecondary inclusions (Table 5) indicate that the early (main) mineralization took place at depths of 4 to 7 km from fluids of type 1. Later deposition took place at 2- to 4-km-depth from fluids of type 2. These pressures and depths are comparable to those estimated by Coveney (1981) for the Alleghany mesothermal gold lodes in California, or the 1.3 to 2.9 kbars (5-12 km) estimated by Smith et al. (1984, but el. Brown and Lamb, 1986) for a similar mesothermal deposit at Timmins, Ontario.

Stable Isotope Studies

Stable isotope analyses of sulfides, carbonates, and silicates constrain the origin of ore fluids at Bralorne and permit estimates of the temperature of mineral-

BRALORNE Au DEPOSIT, B. C. 3 37

ization. Lateral and vertical zonation within the de-

posit and extensive fluid-wall rock interactions are indicated. Minerals observed to be in contact or close

proximity to each other prior to crushing were sep- arated by handpicking. For most samples, isotopic equilibrium is indicated by a normal enrichment order; samples not displaying this order are "reversed." Inconsistent equilibrium relations among some samples may reflect deposition over a range of temperatures and/or chemical conditions.

Sulfur isotopes

Sulfur isotopes were analyzed in galena, sphalerite, pyrite, pyrrhotite, chalcopyrite, and tetrahedrite from samples at different levels of the mine and from nearby deposits in the Bralorne area (Pioneer, P. E. Gold, and B.R.X.). The results in Table 6 and Figure 19 show that •i34S values range from -7 to q-11 per mil. They cluster around -5 and q-5 per rail and have an overall average of q-2. There is no horizontal or ver-

tical zonation for any given sulfide. The narrow range of b34S values for Bralorne and other deposits in the Bridge River camp, averaging '2 per mil, is characteristic of some vein deposits (Rye and Ohmoto, 1974) and implies an igneous source. The absence of sulfate minerals at Bralorne is characteristic of Mesozoic me-

sothermal vein deposits (Taylor, 1987) and suggests that H2S was the dominant sulfur species in the ore fluid (Ohmoto and Rye, 1979). Computer modeling, below, indicates that H2S was the dominant sulfur species in the ore fluid.

The b34S value of pyrite, from -5 to 4-11 per mil, is similar to that in deposits of the Mother Lode district, California (Weir and Kerrick, 1987; Taylor, 1987), interpreted by Taylor to be consistent with derivation of sulfur from shales, greenstones, or felsic plutons, or a combination of all three. In the Bridge River camp, Bralorne and other deposits are spatially and temporally associated with Late Cretaceous felsic intrusions, and Pb isotope evidence suggests that

TABLE 6. Sulfur Isotope Data for Several Deposits of the Bridge River Camp, Southwestern British Columbia Data are plotted in Figure 19.

Measured ba4S for minerals

Deposit and sample no. gn sl tt py cp po As•.g, T (øC)

Bralorne host rocks C082A

C082B

C092B C093A

C094A

C095A

Bralorne-Pioneer E73.004.048

Bralorne Mines Suite 60

Bralorne (41 level, 79 vein) Cl16-14

Bralorne (surface, Ida May) E73.004.047

Pioneer (14 level) E3519

Pioneer (5 level) Joubin

P.E. Gold P-85-03.450.5m

2.35 4.13

-6.48 -3.35

[-6.321

5.95 4.68

2.36

-5.30 -2.63

1.94 1.00

2.15

4.53

10.7

8.7

3.5

8.5

4.0

3.4

1.78 370

3.03 260

8.99 6.42 4.17 (R)

2.67 295

I1.141 (R)

-0.08 1.78 2.85 2.02 1.86 360

[1.78]

-4.22 -4.92 -5.76 3.28 240

BRX Arizona

Unlocated -7.50 Blackdome

Uniocated 0.64

Sulfur isotope ratios are reported as rSa4S, in per mil, relative to the Canyon Diablo troilite (CDT) standard, determined on SOs gas extracted by high-temperature combustion of the sulfide with cupric oxide; analyses were carried out at the Stable Isotope Laboratory, University of Calgary, under the supervision of Roy Krouse, and at the University of Ottawa, under the supervision of Bruce Taylor; no systematic differences are apparent in the data sets from the two laboratories; precision of analysis was evaluated by duplicate analyses, which are given as bracketed figures; difference in duplicate rSa4S values is less than ñ0.1 per rail; equilibration temperatures were calculated from formulas of Ohmoto and Rye (1979)

Abbreviations: cp = chalcopyrite; gn = galena, po = pyrrhotite, py = pyrite, sl = sphalerite, tt = tetrahedrite, R = reversed


po

,• py

x $1

gn

the main ore event. Sphalerite and galena are probably in equilibrium; experimental curves for tetrahedrite are not available. Sphalerite and galena pairs give temperatures of formation ranging from 240 ø to 370øC. These temperatures cannot be compared directly with fluid inclusion homogenization data since the only sample for which both sulfur isotope and fluid inclusion data are available (C 116-14) showed reversed b34S values.

-10 -5 0 *5 -1 0

/34S,O/0o FIG. 19. Sulfur isotope data for various sulfides from the Bridge

River camp. Mineral abbreviations: cp = chalcopyrite, gn = galena, po = pyrrhotite, py = pyrite, sl = sphalerite, tt = tetrahedrite. Solid circles are for pyrite from least altered host rocks; open circles are for vein pyrite (see Table 6 for individual data).

large-scale hydrothermal systems were operative (Leitch et al., 1989). It therefore seems likely that sulfur could have been derived from a felsic igneous source as well as from sedimentary and greenstone sources in the intruded rocks.

Calculated temperatures of equilibration between various sulfide pairs are given in Table 6. Tempera- tures calculated using data from minerals other than galena and sphalerite are erratic and unrealistically high. Pyrite is stable over a wide temperature range and typically does not equilibrate as well as galena and sphalerite (cf. Godwin et al., 1986). Pyrrhotite, chalcopyrite, and pyrite are more widespread in the deposit and probably precipitated at several times over the history of ore formation. Tetrahedrite, sphalerite, and galena, which are closely associated with the gold, only formed contemporaneously during

Carbon and oxygen isotopes of carbonates Carbon and oxygen isotope measurements in vein

calcites and coexisting quartz are given in Table 7. Values vary over only small ranges, t•3C = -10.7 _ 1.5 and (•180 ---- +15.2 _+ 1.5 per mil. The one exception is the t•lsO of 7.2 for sample C1035, a clear, vug-filling calcite that is paragenetically late and of restricted occurrence (near the 77 vein on the 20 level). Despite the relative insensitivity of the quartz- calcite geothermometer (20øC per 0.2%0 variation in delta quartz-calcite), the 2 to 3 per mil quartz-calcite fractionation indicates calcite formation temperatures of 145 ø to 290øC (Table 7), similar to the 150 ø to 250 øC range for secondary and pseudosecondary fluid inclusions. The single high temperature of 345øC (sample 8-51Bsp) is anomalous, but fluid inclusions in clear quartz in this sample also yield anomalously high Th data. The calcites are enriched (Fig. 20) in 180 compared to the mean values (13 ___ 2%0) of carbonates from mesothermal gold vein deposits in the Precambrian shield in Ontario (Timrains, Hollinger) and in Western Australia (Kalgoorlie) but are depleted in lSO relative to carbonates in the Mother Lode veins of California (17 ___ 5%0: Taylor, 1987).

Calcite analyzed from Bralorne has a range oft•l•C values from -12.1 to -9.5 per mil. This is more negative in t•13C (Fig. 20, from Taylor, 1987) than vein carbonates from other similar deposits such as the

TABLE 7. Carbon and Oxygen Isotope Ratios in Vein Calcite, and Oxygen Isotope Ratios in Coexisting Vein Quartz, in the Bralorne Deposit, Southwestern British Columbia

Data are plotted in Figure 20.

Sample no. Level Vein (• 13Cca]cite I (• 18Ocalcite2 (• 18Oquartz3 Aquart .... lcite 4 T(øC)

8-51B(FW) 8 5lB -11.3 14.6 18.0 3.4 145 8-51B(SP) 8 51B -9.9 15.0 16.6 1.6 345 C081-2 8 5lB -12.1 15.0 18.4 3.4 145 15-51(c) 15 51 -9.6 17.0 18.9 1.9 290 16-51(E) 16 51 -11.0 15.6 18.5 2.9 180 C1035 20 775 -9.5 7.2 -- -- low Cl17-7 32 79 -11.3 13.8 -- -- --

• In per mil relative to PDB; CO2 liberated from calcite by reaction with 100 percent phosphoric acid at 25øC (McCrea, 1950) and collected until reaction was essentially complete; reproducibility of b•3C values is ___0.2 per mil

2 In per mil relative to SMOW a Dash indicates there was no quartz present in the vein 4 Per mil quartz-calcite fractionation s Taken from a small carbonate veinlet near the major 77 vein

BRALORNE Au DEPOSIT, B. C. 3 3 9

o

•.4

-8

MARBL•//• ""•__jo•VEIN & ALTERATION

H • •- • CARBONATE

•ø0 •T o* o

- 1• I I I I I I I I I I I ß '8 -12 -16 *20 +24 +28 •32

FIG. 20. Plot of $]3C vs. (•]SO (Taylor, 1987) for carbonates in the Mother Lode area of California, showing averages and fields for Bralorne (B), Timmins (T), Hollinger (H), and Kalgoorlie (K).

Mother Lode in California (-5 _+ 4%o) or the Timrains and Hollinger deposits in Ontario and the Kalgoorlie deposit in Western Australia (-3 _+ 1%0). The Bralorne carbonate •3C values are most similar to those measured at Panasquiera, Portugal, and Pasto Bueno, Peru (high-temperature Sn-W deposits), which range from -12 to -7 per mil (Rye and Ohmoto, 1974). At Bra- lorne (see below), the dominant carbon species in the ore fluids was H2603(aq), and the temperature of carbonate deposition was higher than 200øC, so • Coaltit e

13 = • CH•co3 -- •SC•uld (Rye and Ohmoto, 1974). Thus the increase in •sC to -9.5 per mil in sample C1035, which is a late-stage calcite, suggests that the •sCf increased in the later stages of mineralization.

The •sC of the hydrothermal fluids at Bralorne is too negative for carbon to have been derived from purely magmatic sources (-3 to -7.5: Taylor and Gerlach, 1986). If magmatic and/or sedimentary carbonate (•3C = -2 to +2) carbon were the dominant source(s), then these more negative values are possibly due to oxidized organic carbon present in sediments through which the fluids have passed (Ohmoto and Kerrick, 1977), or to fractionation between CH4 and CO2 (Taylor, 1987). Similar •sC values (-10 _+ 0.05%0) and •]so values (+16.4 _+ 0.15%0) have been observed at the Agnico-Eagle deposit in Quebec but only for siderite in premineral chemical carbonate sediments; ankerite associated with gold has the more typical values noted above of •3C = -3 _+ 0.12 and •So = +11.4 _+ 0.29 per rail (Kerrich, 1987). As Ohmoto and Rye (1979) point out, the source of carbon for carbonates in hydrothermal vein deposits with •sC values between -5 and -10 per mil is difficult to define.

Oxygen isotopes of silicates

One hundred oxygen isotope analyses (Table 8) of mineral separates and whole-rock samples from 15 different vein systems in the deposit were obtained, including ten duplicate analyses of quartz and one of sericite. Vein quartz was sampled at the surface over 6 km, from the Cosmopolitan vein at the northwest end to the P. E. Gold vein at the southeast end of the

vein system (Fig. 21a). Variation with depth was in- vestigated in samples collected to 2 km deep in the Bralorne mine (Fig. 2lb).

Lateral and vertical isotopic zoning in quartz veins at Bralorne is not well defined. Quartz veins are ho- mogeneous across their width (see the 51 vein on the 16 level, samples 1-7, in Table 8). Analyses of vein quartz from five surface localities (Fig. 21a) suggest a possible district-scale variation in •so from 14.9 in the southeast to 19.4 per mil in the northwest (analyses from the Pioneer mine and in the 77 vein at the Bralorne mine are almost within analytical error of each other). Mineralogy (more biotite to the southeast; Joubin, 1948; Nordine, 1983) and structural data (a deeper level of exposure to the southeast; Joubin, 1948) suggest that the geographic increase in •so of quartz reflects a variation in vein temperature rather than •so of the hydrothermal fluid. Similarly, a decrease in •so of quartz with depth (Fig. 21b) suggests increasing temperatures with depth as indicated by fluid inclusions.

Geothermometry: Quartz-sericite,-chlorite,-albite, and -hornblende mineral pairs were analyzed from veins and wall rocks to estimate the temperature of mineralization based on oxygen isotope fractionations (Table 8). For quartz-sericite, calculated temperatures for milky quartz from the main veins (Clayton et al., 1972; O'Neil and Taylor, 1969; Table 8) vary from 200 ø to 240øC; paragenetically distinct, clear euhedral quartz (with anomalously high Th) and associated mica yielded a higher temperature of 305øC. The temperature of 530øC from the 79 vein on the 41 level is too high due to contamination of the sericite by fine quartz. Temperature estimates for quartz-sericite in altered wall rocks range from 370 ø to 400øC. Using the equation of Wenner and Taylor (1971), most of the quartz-chlorite pairs analyzed do not provide valid temperatures. Only one sample, from altered wall rock around the 51 vein on the 8 level

(C002FW4), gave a geologically reasonable temperature of 325øC. The indicated isotopic disequilibrium between quartz and chlorite in the veins may stem from crystallization at different times and/or reequilibration of the chlorite with lower temperature fluids. Using the quartz-albite pair (Matthews et al., 1983) a weakly altered diorite sample, 10 m from the 51 vein on the 15 level (C033-9), gives a temperature of 290øC, apparently reset toward lower tempera-


TABLE 8. Oxygen Isotope Compositions of Minerals and Rocks in the Braiorne-Pioneer Mesothermai Gold Vein System

Sample Level Vein Host rock bX8OX Tt •18Ofiuid

Vein quartz (all milky quartz except clear quartz, marked*)

C1002 0 Cosmopolitan Diorite 19.4 C 1027 3 51 (surface) Diorite 19.1 320 12.9 C048 4 77 (surface) Diorite 17.3 C082 4 Pioneer HW Soda gr 18.0 C049 710 PE gold Pioneer 14.9 360 9.9 8-51B(FW) 8 5lB, in FW Diorite 18.0 385 13.6 8-51Bsp 8 5lB split Diorite 16.6 8-51Bsp* 8 5lB split Diorite 17.4 380 12.8 C081-2 8 5 lB main Diorite 18.4 350 13.1

Cl11-28 8 5lB FW Soda gr 18.4 375 13.7 15-51(c) 15 51, center Diorite 18.9 340 13.3 16-51(E) 16 51, in FW Diorite 18.1 ?350 11.7 16-51 C 16 Composite Diorite 18.4 16-51(1) 16 1 cm from HW Diorite 19.4

(2) 16 3 cm from HW Diorite 19.0 (3) 16 6 cm from HW Diorite 18.3 (4) 16 10 cm from HW Diorite 18.9 ?350 12.5 (5) 16 13 cm from HW Diorite 18.6 (6) 16 17 cm from HW Diorite 18.9 (7) 16 20 cm from HW Diorite 18.6

C118-11 26 85 vein Soda gr 14.8 C118-11' 26 85 vein Soda gr 15.5 365 10.6 C116-3/16 41 79 vein Diorite 17.3 400 13.2 Cl16-14 41 79 vein Diorite 17.1 400 13.3 C128-20 44 77 vein Diorite 17.1 410 13.0

Quartz-mineral fractionation

Sample Level Vein Mineral •jXSOmi .... I (%0) Tf Wt

Vein sericite and chlorite

Pioneer 0 HW Main Sericite 13.6 CO 49 ? 10 PE Gold Chlorite 12.0

8-51Bsp 8 51B Sericite 13.9 C111-28 8 5lB FW Sericite 14.1

C117-7 32 79 Sericite 10.7

C 116-3/16 41 77 Sericite 16.0 C128-20 44 79 Sericite 12.1

Rock •J 18Omineral Sample type qz ab hb mica

4.4 400 (230) 2.8 1,260 3.5 465 (305) 4.3 405 (240)

(no qz) 1.3 845 (530) S.O 360 (200)

Quartz- mineral fractionation

ab hb mica

360 380

375

400

410

T

(øc)

Wall-rock minerals: Least altered rocks

C092A Albitite 10.6 14.0 C071 Albitite 9.5 C4141 Albitite 12.8 C093A Diorite 14.3 13.8

C094B Soda gr 11.5 12.3

C033-9 Diorite 16.1

C033-5/6 Diorite 16.2 C033-1/2 Diorite 17.5 C002FW4 Diorite 15.3

C002FW1 Diorite 16.4

C080 Albitite

11.7 c

6.2

9.9 c

Wall-rock minerals: Altered rocks

14.5 8.3

14.7 TM 7.8 •

13.7 TM 13.8 TM

(R)

(R) 0.5

1.6

8.1

8.3

(R)

1.6 e

2.8 TM 2.7 •

2.7 TM

730

340

1,040

290

350

530

325 560

BRALORNE A•t DEPOSIT, B. C. 3 41

TABLE 8. (Cont.)

Sample Host rock 61sO D (m) a Sample Host rock 61SO 1 D (m)

Whole rocks: Least altered rocks

C083A Hb porphyry 10.0 > 10 C094B Soda granite 12.7 > 10 C092A Albitite 13.5 > 10 C095A Pioneer 6.5 > 10 C093A Diorite 10.6 >10 C4141 Albitite 13.4 >10

Whole rocks: Altered rocks

C1027HW3 Diorite 11.1 2.3 C116-23 Diorite 11.7 10.0 HW2 Diorite 12.9 0.3 -22 Diorite 13.0 5.0

HW1 Diorite 13.7 0.1 -21 Diorite 13.2 3.0

C002FW4 Diorite 13.9 3.5 -20 Diorite 13.0 2.0 FW3 Diorite 14.2 1.5 -19 Diorite 13.4 1.0

FW2 Diorite 15.2 0.4 - 18 Diorite 12.8 0.3

FW1 Diorite 16.1 0.1 C128 Diorite 11.0 6.0

C033-9 Diorite 11.8 10.0 -5 Diorite 11.9 4.0 -10 Diorite 13.0 10.0 -3 Diorite 11.6 2.0 -8 Diorite 14.1 8.0 - 1 Diorite 12.1 0.1

-7 Diorite 13.4 5.0 C111-29 Soda gr 13.7 0.5 -5/6 Diorite 14.9 3.5 -30 Soda gr 13.8 1.0 -3/4 Diorite 16.1 1.5 -31 Soda gr 13.9 3.5 -1/2 Diorite 16.4 0.3

Data are reported in the usual 6 notation relative to SMOW (standard mean ocean water); oxygen was extracted with CF3 at 575øC (Borthwick and Harmon, 1982), isotopic ratios were determined on CO2 produced by reaction with a hot carbon rod at 750øC (Clayton and Mayeda, 1963); mass spectrometric analyses are normalized to both SMOW and SLAP (standard light Antarctic precipitation); mean variation of 6 from duplicate analyses is 0.09 _+ 0.04 per mil; our 6•so value for NBS-28 is 9.6 +__ 0.1

Tt = trapping temperature estimated from fluid inclusions (italic figure is from sulfur isotopes); for quartz-mineral temperatures (Tf), figures for quartz-muscovite are from Bottinga and Javoy (1975); parenthetical figures are from Clayton et al. (1972) and O'Neill and Taylor (1969); all temperatures are in øC

Abbreviations: FW = footwall; HW = hanging wall; (R) = reversal of b•So values; ab = albite, ch = chlorite, hb = hornblende, qz = quartz; c = chlorite, m = muscovite; Soda gr = soda granite, Pioneer = Pioneer greenstones, Albitite = albitite dike, Hb porphyry = hornblende porphyry dike

• Average of duplicates is used where available 2 Distance from vein in meters

tures by the hydrothermal alteration. Least altered wall rock (C093A) far from veining gives a quartz- albite temperature of 730øC (although the albite here is not likely of magmatic origin). For quartz-hornblende (Bottinga and Javoy, 1975), least altered Bra- lorne diorite (C093A) far from the main veins gives a similar temperature (340øC) to that from quartz- hornblende in sample C033-9, 10 m from the 51 vein on the 15 level, which gives 350øC. Both these figures are extrapolated slightly outside the range of the equation used. However, as for the quartz-albite pairs above, they imply a strong interaction of the wall rock with the altering fluids, (i.e., a high water/rock ratio). The interaction is apparently more marked for the hydrous minerals, chlorite and hornblende, than for albite.

Implications for water/rock ratios: Samples of whole rocks and separate minerals from five traverses across altered wall rocks (the 0, 8, 15, 41, and 44 levels, at 0, 350, 650, 1,800, and 2,000 m depth) were analyzed for their oxygen isotope ratios to test for zoning. Alteration varies from quartz-sericite-ankerite up to 1 m from the vein, through albite-calcite from 1 to 5 m, to chlorite-epidote 5 m or more from

the vein. Both whole-rock and mineral oxygen isotope ratios increase as the veins are approached (Table 8). For instance, in the 10-m series in altered diorite near the 51 vein on the 15 level (C033-1 to 10), there is a progression from ½5•sO = 10.6 per mil in least altered rock (estimated by C093A), to 16.4 per mil immediately adjacent to the vein; the minor reversal between 7 and 8 is due to proximity to a subsidiary fracture off the main vein. The ½5•sO varies in a similar way among samples from the other series in diorite, although gradients are not as pronounced at deeper mine levels (the Cl16 and C128 series from the 41 and 44 levels) which are in much more mafic wall rock than the typical diorite, containing less than 5 percent quartz. In soda granite, whole-rock ½5•sO values are about 1 per mil higher within 3.5 m of the vein. There is thus a clear pattern of reequilibration of the whole-rock ½5•sO values due to hydrothermal alteration for up to 10 m from veins.

Quartz in the veins and altered envelopes has a high ½5•sO value (up to 19.4%0) compared to quartz in the least altered rock (14.3%0), and its ½5•SO increases systematically toward the veins. The other major rock-forming minerals in the diorite also show


(14.9), , co49 •

, ¾½ •,,,•P. E. ß (19.4) ,%_ ,_• •/GOL D

• • T '-- -•' 'd10•7 z + ½•C082 /

• / • - CADWALLADER VAULT •(Z •• PIo•er HW a LEVEL DEPTH (m) COSMOPOLITAN' '51'VEIN '77'VEIN'51BFW' 'HW' PE,GOLD

O 13• 19.4 3 19.1 • •9' •

15 65o 5 •

26 •70 32 •450

41

44 2oo0 1 ? • b

FIG. 2I. (a), Geographic zoning of oxygen isotope ratios of vein quartz from the Bralorne-Pioneer vein system. The "+" symbols indicate extent of Bralorne intrusions; sample numbers are C1002, etc.; •80 values are given in parentheses; vein designation is contained in single quotes, with vein strike and steep northerly dip indicated. (b). Vertical zoning in oxygen isotope ratios in the Bralorne-Pioneer vein system is shown by variation in {518Oquartz values with reference to the mine levels and schematic position in the plane of the vein system. The host rock is diorite except for those samples marked * (51BFW and Pioneer HW veins) in soda granite, and the P, E. Gold vein marked *ø in Pioneer greenstone.

small increases in h•80 values toward the veins until

they become unstable (plagioclase from 13.8-14.5%0, and hornblende from 6.2-8.3%0). These effects explain the shifts in whole-rock •i]sO patterns toward higher values as the vein is approached in hydro- thermally altered wall-rock envelopes. In the more mafic host rocks at deeper levels, the lack of alteration quartz may explain the smaller shifts observed. The marked shifts in h]80 values for both mineral separates and whole-rock samples imply high water/rock ratios during the alteration, a feature common to other earbonate alteration zones around mesothermal gold quartz veins (e.g., Kerrich, 1983; Taylor, 1987).

Oxygen isotope composition of the ore fluids (Table 8) were calculated from isotopic analyses of vein quartz, trapping temperatures from fluid inclusion studies, and the quartz-water fraedonation equation of Clayton et al. (1972). Values appear to have been reasonably constant at about 12.6 + 1.3 per rail from the bottom to top of the vein system except for two samples, from the P. E. Gold vein and the 51 vein on the 26 level, which have lower values.

The small-scale variation in h]sO values of vein quartz (decreasing h values with depth, i.e., increasing temperature) is similar to the regional lateral variation in observed h]sO values of quartz described by Ma- heux et al. (1987) and Nesbitt et al. (1987). Deposits in the camp are zoned from the high-temperature Bralorne-type Au quartz veins (•18Oquartz = 17.5 + 1.0%0), to intermediate temperature Sb-Ag-Au veins (21.0 + 1.0%0), to low-temperature Hg deposits

(29.0 __+ 2.0%0). Deposition was from the same or similar deeply circulating, highly evolved (]SO-shifted) fluids of around 11.5 __+ 2 per mil. This value, corrected to the estimated trapping temperature of 350øC, is within analytical uncertainty of the mean value obtained for Bralorne in this study. The strong enrichment in •i]sO values of quartz, combined with D/H studies of fluid inclusions, led Nesbitt et al. (1986, 1987) to propose that the ore-forming fluids were composed of deeply circulated meteoric water (cf. Magaritz and Taylor, 1976, 1986).

The calculated h]80 fluid of 12.6 per mil at Bra- lorne is close to values for the Coquihalla deposits near Hope, B. C., and the Mother Lode deposits in California (Taylor, 1987) and is in the middle of the range for metamorphic waters. However, the concept of "metamorphic water" must be clarified, because its original isotopic definition (e.g., Taylor, 1974) is not meant to imply water of dehydration (see Taylor, 1987). Inasmuch as most water/rock reactions involve either seawater or meteoric water, the evolved meteoric water hypothesis seems reasonable from the evidence to date but might be further constrained by data from hydrogen isotope studies (in progress). In similar deposits in the Juneau area of Alaska, •iD values of-20 to -30 per mil indicate a deep-seated metamorphic fluid source (Goldfarb et al., 1989). In the Mother Lode of California, sericite and mariposite associated with alteration and gold mineralization formed from waters with somewhat higher •iD than waters in some fluid inclusions in adjacent quartz


veins, indicating mixing of waters with different evo- lutionary histories (Taylor, 1987). Lead isotope evidence (Leitch et al., 1989) suggests that a magmatic source was involved in the formation of galena.

Although meteoric water may end up with a metamorphic buffered CO2 value (Nesbitt et al., 1987), the high CO2 contents typically found in the ore fluids are usually interpreted as derived by metamorphic dehydration (e.g., Kerrich, 1983; Andrews et al., 1986). Such an origin is espoused by many workers on gold vein deposits in the Precambrian shield of Ontario and Quebec (e.g., Colvine et al., 1984). Hodgson and MacGeehan (1982) favor a variation of metamorphic origin for the fluids by expulsion of pore water (after Lydon, 1977). At the Mother Lode deposit in California, Landfeldt (1987) proposed that the ore fluids may represent a CO•-rich fluid phase separation (Trommsdorffand Skippen, 1986) derived from a reservoir generated at 20- to 30-km depth during rebound of an isotherm depressed by subduc- tion. The high CO• contents are also consistent with fluids produced by degassing of mantle magmas (e.g., Cameron, 1989a, b).

A more complex, mixed origin for H-O-C-S in the mineralizing fluids must be considered for fluids circulating along structural and tectonic boundaries such as the Bralorne fault zone. For example, emplacement of mafic magmas along major tectonic zones provides a source of CO• isotopically similar to that found in the Mother Lode (Taylor, 1979, 1987). Thus the ore fluids at Bralorne could also have been derived from

mafic magmas. Sulfur isotope values in the sulfides, with a range from -7 to +9 per mil, clustered around zero, would support such an interpretation. Carbon and oxygen isotope ratios in carbonate minerals are most similar to those in high-temperature (Sn-W) vein deposits, but the observed range in •3C values could be produced from many sources. Contamination of magmatic carbon by oxidized organic carbon or by dissolved carbonate carbon in sediments could produce the observed range of •3C (-12 to -7%0). It seems likely that deeply circulating meteoric waters were mixed with a more uniform crust-equilibrated reservoir of waters being driven off by metamorphism (cf. Kerrich, 1987; Taylor, 1987). Early Late Creta- ceous intrusions, closely associated with mineralization, provided heat to drive the flow of the ore fluids; they may also have provided a magmatic component.

Discussion: Deposition, Transport, and Source of Gold

Characteristics of the ore-forming fluids

The results presented above suggest that the mineralizing fluids at Bralorne evolved from initially high- temperature, CO•-rich to low-temperature, H20- dominant ones (cf. the deposits in the Red Lake camp

of Ontario; Andrews et al., 1986). It is not certain that the gold was deposited with the earlier carbonic fluids (as found at many mesothermal gold vein deposits, e.g., Robert and Kelly, 1987, or Walsh et al., 1988) or with the later H•O-rich fluid (cf. Andrews et al., 1986). However, since the later fluids are like those found in inclusions in calcite, and no gold is associated with calcite, it seems most likely that gold deposition accompanied the earlier fluids. Thus the mineralizing fluids at Bralorne are probably similar to other mesothermal gold vein deposits in the Ca- nadian Shield, such as the Hollinger-McIntyre, where the composition of the primary fluid was CO• rich and low salinity: 93 mole percent H•O, 6 equiv mole percent CO• and 1 equiv mole percent NaC1, and deposition was at 280 ø +__ 50øC (Smith et al., 1984; Wood et al., 1986). Similar fluids have also been described at Doyon, Quebec (Guha et al., 1982) and Pamour, Ontario (Walsh et al., 1988), as well as for the Kalgoorlie-Kambalda area, Western Australia (Groves et al., 1984). The salinity of the Bralorne ore fluid, at about 3 wt percent NaC1 (approximately 0.5 m) is the same as predicted by the experiments of Wood (1987) to be best capable of mobilizing gold (with a high Au/Ag ratio, as observed at Bralorne) from source rocks while leaving base metals behind, a process suggested by Kerrich (1983) as important in forming mesothermal gold deposits as opposed to base metal deposits.

Further details of the ore fluid were modeled using the PATH computer program (Helgeson et al., 1970; Perkins, 1980). Thermodynamic data for aqueous species are from Helgeson (1969); data for solid phases are from Helgeson et al. (1978). Data for gold chloride complexes were included (Wood, 1987, has shown experimentally that gold solubilities in the 5 to 500 ppb range are attained in solutions of 0.5 m NaC1). Other workers have postulated that gold, present as reduced sulfur complexes, was deposited in response to fluid oxidation and the lowering of total sulfur in solution, both as a result of pyritization (e.g., Phillips and Groves, 1983, 1984; Romberger, 1986). Carbonyl complexes may also be responsible for solution and transport of Au in a reduced species, in accordance with the low redox state (Fe+2/Fetotal = 0.9) and the ubiquitous gold-carbonate association (Fyfe and Kerrich, 1984). Carbonyl, cyanide, and/or thiocyanide complexes were also proposed as gold- transporting agents by Hutchinson and Burlington (1984). At the time of the study, data for carbonyl, cyanide, and thiosulfide complexes were not available. Including thiosulfide, however, would only increase the gold in solution and total gold deposited (Seward, 1973, 1984); the mechanism of precipitation would not be altered.

The PATH program models reactions and changes in the ore fluid by progressive, stepwise titration that


TABLE 9. Reaction between Modeled Fluid and Wall Rock at the Bralorne Deposit, Predicted by the PATH Program Only species at concentrations greater than 10 to 6 m were considered.

0.6NaAISiaOs (albite) + 0.06Ca2Mg4SisO22(OH)2 (tremolite) + 0.34FeSiOa (ferrosilite) + 0.19793K + + 0.22535Ca +2 + 0.54649H2S + 0.70719H2COa + 0.007058Mg +2 + 0.003401SO• 2 + 0.0003107H + + 0.00006433Al(OH) +• + 0.002067KSO• + 0.001771CASO4 + 0.039548MGSO4 + 0.000108550AuC1• + 0.031207HSO• q- 0.0021105HC1 + 0.000000015Au + + 0.0005135Mg(OH) + ::> 2.01805SiOg (quartz) + 0.199993KAISiaO•0(OH)g (muscovite) + 0.347118CaMg(COa)• (dolomite) + 0.312016FeS• (pyrite) + 0.000108565Au (native) + 0.60002Na + + 0.0023276C1- + 0.012928HCO• + 0.016685Fe(OH) + + 0.011300Fe +2 + 0.001975H4SIO4 + 0.0000111OH- + 0.0000839Al(OH)• + 0.0004263HS- q- 1.1116H20

All species are aqueous except where noted and are expressed in moles/liter (m)

represents water-rock reactions. The program writes detailed reactions (Table 9), considering as many as 25 reactants and as many products. It first calculates the fluid that would be in equilibrium with the observed alteration assemblage. Such a model for the Bralorne deposit must account for the abundant carbonate in a typical alteration assemblage that also includes quartz, muscovite, albite, chlorite, pyrite, and native gold (thermodynamic data for arsenic com- pounds are not available, so the implications of the arsenopyrite commonly seen in the observed assemblage cannot be assessed). The ore fluid in equilibrium with this typical alteration assemblage is characterized in Table 10. Except for the fugacities of CO2 (10 +2"•) and CH4 (]0+ø'5), the characteristics of these fluids are similar to those modeled by Helgeson and Garrels (1968) for deposition of gold, pyrite, and quartz, although their study considered precipitation in response to a temperature drop rather than reaction with the wall rock. However, the gold content of the ore fluid predicted by the model (1 to 2 x 10 -•ø m, or 0.1 to 0.2 ppb) is much lower than previously estimated (e.g., 20 ppb in both Helgeson and Garrels, 1968, and Kerrich, 1983; 1.5 to 10.5 ppb in Brown, 1986, and 1 to 10 ppb calculated by Shenberger, 1985). The value estimated in this study is much closer to the value of 0.04 ppb measured at Broad- lands, New Zealand (Seward, 1984), or the average from mineralized areas of 0.1 ppb that was measured by McHugh (1988). Such low predicted gold contents emphasize the importance of gold-depositing mechanisms (rather than transporting mechanisms) in order

TABLE 10. Characteristics of Ore Fluid at the Bralorne

Deposit, Predicted by the PATH Program

Temperature = 350øC, pressure = 1.75 kbars, • pH = 4.5 (slightly acid at 350øC and 1.75 kbars)

Na/K = 8:1, [Na] = 0.4 m, [K] = 0.0.5 m, [CI l = 0.5 m, (total salinity = 0.5 m, approximately 3 wt %)2

/coz = 102'S,/ca, = 10øS,/oa = 10-a,J•a = 10 -7 lee+21 -- X X 10 -7 m, [ng+•l -- 0.003 m, [Ca+21 -- 0.01 m, lS-•l : 10 -•ø m, lAu +] = 10 -•ø m (0.1 ppb, as AuCI•)

Estimates from fluid inclusion studies

Agrees with estimate from fluid inclusion studies

to explain adequately the formation of large gold deposits.

Next, the model allows a fluid of the derived composition to react with a fresh rock of the observed composition. The least altered Bralorne diorite (Table 1) was modeled as 60 percent albite, 6 percent tremolite, and 34 percent ferrosilite. The latter two minerals approximate the Fe/Mg ratios in probe analyses of the 40 percent hornblende typically found in the fresh diorite (thermodynamic data are not available for hornblende, a compositionally complex mineral). The pressure and temperature constraints applied were constant at 300øC and 1 kbar, selected to be as close as possible to the fluid inclusion information without extending beyond the experimental data in PATH.

Deposition of gold

Theoretical predictions: When the modeled fluid is allowed to react with the "fresh" (least altered) rock, the alteration mineralogy predicted by the PATH program closely matches the sequence observed outward from the veins at Bralorne (modal mineral data in Table 1). Close to the vein, the model predicts dissolution of the original albite and tremolite q- ferrosilite and precipitation of major amounts of quartz, dolomite, and muscovite, minor amounts of pyrite, and minute amounts of native gold. Gold precipitation is most favored in the early steps of the process. About two-thirds of the gold is deposited immediately adjacent to the vein as the COa-bearing fluids react with the wall rock. Gold precipitation is predicted to decrease sharply as soon as carbon (in trace amounts) becomes stable. Farther out from the vein (i.e., in later steps of the modeled process), chlorite becomes stable (approximated thermodynami- cally by talc, for which data are available). Farther out still, albite becomes stable where muscovite is no longer stable. This would be expected because close to the vein, hydrolysis of albite to muscovite consumes K from the fluid and releases Na into solution; thus the Na/K in the fluid would evolve to higher values as it passed outward (cf. Kerrich, 1983). Concur- rently, as the fugacity of sulfur in the fluid drops, the stable sulfide becomes pyrrhotite rather than pyrite


and the precipitation of Au is no longer favored (cf. Phillips et al., 1983). Where magnetite becomes stable (farthest from the vein), Au is dissolved. The model thus predicts a strong correlation of gold with pyrite and an absence of significant gold from pyrrhotite- or magnetite~bearing assemblages. This is corroborated by evidence from other mesothermal gold vein deposits where magnetite is found as a precursor iron mineral supplying Fe to form pyrite and enhance gold deposition (Phillips et al., 1983; Hall and Rigg, 1986).

The calculated solubilities for gold, as both chloride complexes such as AuCI• and thiosulfide complexes such as Au(HS)•, are illustrated in Figure 22. Under the predicted reducing, slightly acid conditions of log fo• = -30 and pH = 4.5 (neutral at 300øC is 5.2), in the field of stability of pyrite, carbonate, and sericite (as observed in the wall-rock assemblage) and in the H2S stability field (as confirmed by PATH), the dominant Au species appears to be the chloride complex. The observed alteration assemblage, where K feldspar is not stable, requires this pH (contrast this with the

-28

-30

-32

-36

-38

-4(

FIG. 22. Log fo•-pH diagram to illustrate conditions of mineralization at Bralorne, at 300øC and 1 kbar. Stability fields for carbonate (ca), graphite (gr), methane (CH4), muscovite (muse), K-feldspar (kspar), kaolinitc (kaolin), pyrite (py), pyrrhotite (po), hematite (hm), and magnetite (mt) are shown. The solution composition used is total sulfur = 0.5 m, ionic strength = 0.5 m, [Ca +•] -- 0.01 ra, [K +] = 0.05 ra, from PATH modeling. The ore fluid at Bralorne, modeled at log fo• = -30, pH = 4.5, is shown by the cross in the field ofpyrite, muscovite, carbonate, and H•S. As CO• is removed from the solution by carbonation of the wall rock, the solution becomes less acid, moving in the direction of the arrow and causing precipitation of gold (down the chloride complex solubility gradient). The abundances of Au as the chloride complex and the thiosulfide complexes are from Seward (1984).

very high pH conditions, where thiosulfide complexes are more stable, predicted by Walsh et al., 1988). Reaction of this solution with the wall rock would

remove CO2 due to the formation of carbonate, causing the pH to increase and a corresponding shift toward lower AuCI• concentrations, i.e., precipitation of gold (arrow in Fig. 22). Since this is in the direction of slightly increasing or constant stability of gold thiosulfide ions, it is not supportive of Au deposition by destabilization of thiosulfide complexes. Note that it is also at relatively constant fo• (cf. Goldfarb et al., 1989), so that sulfide complexing of Au would not be expected to vary during this process.

A correlation between pyrite and gold has been empirically observed at mesothermal gold mines throughout the world; this model offers a theoretical explanation. The actual precipitation of the gold in the model involves reduction of the aurous gold (Au +) in the AuCI• complex, to native gold (Auø), by donation of an electron. Electrons appear to be donated by a concurrent oxidation reaction, of S -2 in H2S, to sulfur in pyrite, FeS•, which may be thought of as S-. This is suggested by the strong correlation in the model between Fe +• consumption and pyrite production on the one hand and gold produced on the other. Consideration of charge balance requirements for the equation written by the program (Table 9) shows that in this reaction, the small surplus of S -2 (in H•S) oxidized to S- (in pyrite) corresponds to the small amount of gold being precipitated. As soon as carbon becomes stable (C +4 + 4e- = CO), the com- petition for available electrons appears to reduce sharply the possibility of precipitating gold. A similar conclusion has been reached for the mesothermal gold lode deposits of Western Australia (cf. the reactions for reduction of gold in thiosulfide complexes: Phillips et al., 1983). According to the experiments of Bancroft and Jean (1982), the gold ions may be adsorbed on the surface of sulfide grains such as pyrite and then reduced, similar to reduction of gold on activated carbon.

Comparison of predicted and observed assemblages: The predicted outward transition from quartz- sericite-ankerite-pyrite to chlorite-albite-calcite is as observed. Amorphous carbon is seen only rarely in a late-stage carbonate alteration that is black from included finely divided carbon. Assay data show that gold concentrations drop off sharply immediately outside the veins. The bulk of the gold is in the thin black ribbons in the quartz veins, which are highly altered wall-rock slivers composed of finely divided pyrite (carbon has not been identified here), arsenopyrite, sericite, and minor ankerite. Higher gold contents correlate with galena, and to a lesser extent, sphalerite (Dolmage, 1934; Joubin, 1948; Poole, 1955). Gold grains up to several hundred microns across are intergrown with galena (this study) and


sphalerite (Poole, 1955) in some Bralorne-Pioneer specimens, whereas gold inclusions in pyrite and arsenopyrite tend to be less than 15 #m across. Sphal- erite is also an excellent indicator of gold at Pamour, Ontario (Walsh et al., 1988), where it contains inclusions of gold. At Bralorne, pyrrhotite is present in some altered wall rocks, but there is insufficient data to confirm the predicted inverse relation between gold and pyrrhotite. Magnetite is not found in any altered rocks at Bralorne.

The best gold ore shoots were in diorite, peripheral to soda granite bodies (D. H. James and J.P. Weeks, 1961, unpub data; Campbell, 1975). This has been explained by suggesting either a genetic relationship of ore to the soda granite (disproved by isotopic dating: Leitch et al., 1991) or by brittleness of the soda granite causing an inability to sustain large openings. However, the PATH model suggests an alternative explanation: the more abundant iron in the diorite and in the greenstone, the favored host rocks, could increase the precipitation of pyrite, and hence, of gold. A similar correlation between iron-rich host rocks, the best ore, and the most abundant sulfides occurs in the mesothermal gold veins of the North Western mining camp in Zimbabwe (Fuchter and Hodgson, 1986). At the Hunt mine, Kambalda, West- ern Australia (Phillips and Groves, 1984), fluid-wall rock reaction is also interpreted to have caused the deposition of gold and associated ankerite, calcite, biotite, chlorite, and pyrite. Similarly, Au has been found to be concentrated where the shear crosses a competent, chemically favorable Fe-rich host rock at the Lac Shortt deposit in Quebec (Morasse et al., 1986). There is no net increase in iron during such an alteration process (Phillips et al., 1983; Hall and Rigg, 1986; Melling et al., 1986).

Transport of gold

A model for ore deposition at Bralorne must account for two outstanding features of the quartz veins. These are the presence of the ribbons that contain the bulk of the gold, and the syntaxial, coarsely crystalline milky quartz that grew perpendicular to the ribbons or walls of the veins. The latter observation is supported by the strong induced piezoelectric response of the quartz, which indicates alignment of the c-axes of many grains (M. M. Gomshei, pers. commun., 1988). The dark ribbons in the quartz imply a cyclic process, with repeated fracturing and emplacement of quartz. The coarse quartz crystals imply growth in open space (under high, but fluid, pressures: cf. Wood et al., 1986, and Nesbitt, 1988). This con- flicts with the general lack of open space in the mesothermal environment, under pressures of 1 to 2 kbars (4-8 km depth, see above). Hydrostatic pressures are found as deep as 6 to 7 km in Witwatersrand drill holes and to 12 km in the Kola Peninsula hole

(B. E. Nesbitt, pers. commun., 1990), but these are both in noncompressional cratonic settings, unlike Bralorne.

The characteristics of repeated fracturing and deposition of quartz, in a high fluid pressure regime, may be explained by a "fault-valve" model (Sibson et al., 1988). Fluids transporting ore are derived from a geopressured reservoir, possibly formed by metamorphic devolatilization at amphibolite grade, ponded below the ductile-brittle transition zone. The build-up of pressure occurs because of the inappro- priate orientation for slip, in a horizontally compressive stress field, along the steeply dipping faults hosting the Bralorne veins. In order to permit slip, and therefore rupture and fluid flow, the pore pressure must be very high. Such extremely high fluid pressures thus provide the environment for coarse crystal growth to occur at depths of over 4 km. When rupture occurs, fluid will flow into the open space of the fault where the sudden drop in pressure would promote deposition of quartz (Walther and Helgeson, 1977), sulfides (Helgeson and Lichtner, 1987), and gold. Fluid flow would tend to die away slowly in such a system over a period of a few months (Sibson, 1981), possibly allowing time for coarse crystal growth.

As mineral deposition occurred, and the fracture sealed itself, pressure would rise again, leading to a repeat of the same process. This cyclic process explains the ribboning in the Bralorne veins. Each ribbon of quartz would have associated with it a black layer of intensely altered host rock consisting of pyrite, arsenopyrite, sericite, and occasional ankerite, with gold, that formed the vein margin before rupture took place again. The PATH model predicts that the bulk of the gold is precipitated in the immediate envelope to the vein, so that incorporating even a sliver of wall rock into the vein would include most of the gold in the vein.

Source of fluids and gold Lead isotope ratios suggest large-scale mixing of

mantle and upper crustal components, implying widespread circulation of fluids that were then fo- cused along major fault zones to form the deposits. It seems most likely that these fluids were of metamorphic derivation (cf. Kerrich, 1983; Groves et al., 1987), with possible contributions of highly evolved (•80-shifted) meteoric (Nesbitt et al., 1986; Nesbitt, 1988) and magmatic waters. It is not possible to choose between evolved meteoric and metamorphic waters due to a lack of hydrogen isotope data for Bra- lorne. A similar conclusion, that the available data are not sufficient to define a source for the mineralizing fluids of Archean lode gold deposits with conviction, even with hydrogen data, was reached by Roberts (1987). At the Mother Lode in California, stable isotope data merely confirm that the fluids have inter-

BRALORNE Au DEPOSIT, B. C. 3 4 7

acted extensively with rocks (Bohlke and Kistler, 1986; Taylor, 1987), and the ultimate source or sources of the Mother Lode fluids is not clear (Weir and Kerrick, 1987). Objections to the evolved meteoric water model relate mainly to the lack of reli- ability of hydrogen isotope analyses of fluids, which are often extracted from fluid inclusions (Pickthorn et al., 1987). For a few deposits, such as the Jungwon in Korea, the isotopic data are not compatible with ore formation from a metamorphic fluid, and therefore, favor an evolved meteoric fluid (Shelton et al., 1988).

Wall rocks adjacent to veins at Bralorne are strongly enriched in lsO, toward isotopic equilibrium with the veins. This signifies that a fluid-dominated vein system was operative at Bralorne (el. Kerrieh, 1983). A corollary is that lateral diffusion of silica, gold, etc. (el. Boyle, 1979) from wall rock into veins (i.e., a rock-dominated system) is not compatible with the isotopic relations. These results support the conclusion that the fluids involved in transporting the gold at Bralorne were derived from deep metamorphic waters or deeply circulating, evolved meteoric waters, in a fashion similar to that proposed by Fyfe and Henley (1973) or Nesbitt (1988).

A further constraint on the origin of the ore fluids at Bralorne is required to explain the extensive earbonate alteration. This abundant CO2, and the K required for muscovite replacement of albite adjacent to veins, could be supplied by fluids released during greensehist-amphibolite metamorphism, where the relative proportion of CO2/H20 is on the order of 0.2 to 0.5, and K/Na is about 1 (Kerrieh and Fyfe, 1981). A reduced, Na- and Au-bearing solution was found to be produced during the dehydration of amphibo- lites in the metamorphism of spilitized marie volcanic rocks at the Opapimiskan Lake gold deposit in Ontario (Hall and Rigg, 1986); such studies at other deposits (Colvine et al., 1988) support the derivation of the ore fluids from metamorphic dehydration. In the deep crust "magmatie" and "metamorphic" processes are so intertwined that distinction between the two is

probably irrelevant (Cameron, 1989e). At Bralorne, the timing of ore deposition, synchronous with ret- rograde conditions after the metamorphic peak as suggested by isotopic dating and pressures of mineralization lower than metamorphic pressures, supports metamorphic dehydration (el. Phillips and Groves, 1984; Thompson, 1986; Nesbitt and Mueh- lenbachs, 1988).

Ultramarie rocks have been cited as the source of

gold in vein deposits (e.g., Pyke, 1975, 1976, in the Porcupine camp, Ontario) because of their presumed elevated Au abundances. However, analysis of Au by Tilling et al. (1973) suggested that ultramarie rocks have abundances of 0.5 to 2 ppb Au, in the same range as other igneous rocks. More recent work suggests

that carbonation provides a mechanism for release of Au from magnetite and secondary sulfides where it may have been concentrated earlier during serpentinization of ultramafic rocks. Thus the gold may be reconcentrated in carbonate-altered ultramafic equivalents, or listwanites (3-5 ppb: Buisson and Leblanc, 1987). Serpentinized ultramafic rocks are prominent along the faults of the Bralorne zone, adjacent to the gold deposits, although their carbonate- altered equivalents are not. Thus it is not likely that the nearby or subjacent ultramafic rocks of the Bra- lorne area are the source for the gold. Instead, sweeping large volumes of fluid (as suggested by the isotopic studies) through large bodies of rock with slightly elevated gold contents could provide a mechanism for extraction of sufficient gold (cf. Hutchinson and Burlington, 1984). High Mg basalts of the Pioneer Formation in the Bralorne area (Leitch, 1989) could be such a source (cf. Keays, 1984). The low-salinity (0.5 m) solutions found at Bralorne would be capable of nearly complete extraction of gold from the source rock, yet are close enough to saturation to allow for efficient precipitation at the site of deposition (Wood, 1987).

The total production of gold from the Bralorne deposit was 130 metric tons, or 1.3 X 108 g. If this mass of gold was derived by 50 percent efficient leaching (cf. Buisson and Leblanc, 1987) from rocks containing 1 ppb Au (Tilling et al., 1973), then the source volume would be on the order of 100 km 3 of rock (1 km 3 of rock contains 2.8 X 106 g of gold at 1 ppb). If this gold were leached and transported during metamorphic outgassing within the greenschist facies with 2 to 3 percent water liberated from the volcanics and sediments, then the solvent volume would be on the order of 200 km a, and the concentration of gold in solution would be 1.3 X 108 g Au divided by 200 X 10 l'• g of rock, or 0.65 ppb (cf. Kerrich, 1983). If the leaching were less than 50 percent efficient, this value would be less (at 10% efficiency, it would be 0.1 ppb). This is similar to the 0.1 to 0.2 ppb predicted by the PATH model, or the 0.04 to 0.1 ppb range measured in natural waters associated with mineralization presently forming (Seward, 1984; Krupp and Seward, 1987; McHugh, 1988). Calculations by Shenberger (1985) indicate that a concentration of 1 to 10 ppb Au in solution is sufficient to produce economic Au grades, consistent with observations by Brown (1986) and Krupp and Seward (1987) in New Zealand, where economic grades of mineralization are being formed from fluids with 1.5 to 10.5 ppb Au in solution. The length of time necessary to form a deposit of economic size (50-100 metric tons, or 2-3 Moz) may be only several thousand years (Krupp and Seward, 1987).

It is possible that the high heat flow in evidence at Bralorne today, and presumably present at the time


of mineralization, was both indicative of the flow of large volumes of hydrothermal solutions, and favorable for efficient gold mobilization from source rocks at depth. The importance of a high geothermal gradient to generation of auriferous metamorphic fluids has been noted by Groves et al. (1987). The well- developed fracturing within the major Bralorne fault zone provided excellent channelways, or plumbing, to focus solutions from large volumes of rock and is probably the main reason that these deposits are found in this zone (cf. Groves et al., 1987; Nesbitt, 1988).

Conclusions

The Bralorne deposit is located within a regional brittle fault zone. Such major crustal structures are also associated with many of the large mining camps of the Superior province in Canada (Hodgson et al., 1982) and the Yilgarn block in Australia (Phillips, 1986). Rocks in the mine area show sub- to lower greenschist facies metamorphism, at 325 ø to 375øC and 2 to 3 kbars H20 pressures. As at other mesothermal gold quartz vein deposits, alteration envelopes grade outward from intensely foliated quartz-ankeritic carbonate-sericite (_ fuchsite) to less sheared calcite- chlorite-albite to unsheared epidote-calcite. Chemical studies of the altered rocks show that there has been

addition of K20, CO=, S, As, and Au; Na=O, FeO, and MgO have been depleted; SiO2 and CaO are locally depleted and reconcentrated. These results reflect replacement of albite by muscovite, and hornblende by chlorite and ankerite. Gold is found mainly as thin smeared flakes in the black sulfidic septae of the strongly ribboned shear veins. Concentrations of pyrite, pyrrhotite, and much lesser chalcopyrite drop off sharply within a few meters of the veins; arsenopyrite is found immediately adjacent to the veins. Small amounts of sphalerite, and especially galena, correlate with gold-rich portions of the veins.

A fault-valve model best explains the ribboned, yet euhedral, coarsely crystalline milky quartz veins. Cyclic build-up of pressure in a reservoir of geopressured fluids below the brittle-ductile transition caused

overpressuring, invoking failure by reactivation of the previously formed steeply dipping faults, unfavorably oriented for failure at a high angle to the maximum compressive stress in a transpressive regime. Failure provoked strong discharge of the ponded fluids and the pressure release promoted deposition of quartz and sulfides, with zoned quartz crystals deposited in space held open by the high pore pressures. Sealing of the fault by this mineral deposition allowed pressure to build and the cycle to repeat. Each of the ribbons of sulfide, with minor gold, may represent a sliver of highly replaced wall rock that was included in the vein when the next episode of fracturing, di- lation, and mineral deposition occurred.

A vertical zonation of gradually increasing temperature with depth at about 30øC/km is apparent from studies of fluid inclusions by both optical and decrepitation methods and is reflected by the systematic variation in the o;xygen isotope ratios measured in vein quartz. However, there is no appreciable change in ore grade, sulfide or gangue mineral assemblages, or rock alteration, from surface to an almost 2-km depth. Such a lack of zoning implies ore deposition from a large, widely circulating hydrothermal system that essentially cooled only slowly upward at a geothermal gradient. The Bralorne fault zone was instrumental in providing abundant ground prepa- ration, or plumbing, for the veins, and in focusing the discharge of large volumes of fluid containing minute quantities of gold.

Oxygen isotope ratios of vein quartz and associated micas show that there has been significant wall-rock reaction with the mineralizing fluids, suggesting high water-rock ratios near the veins. Calculated oxygen isotope compositions of mineralizing fluids are approximately constant at 13 _ 1.5 per rail over the 2- km vertical extent of the deposit. Measures of vein carbonates suggest that the carbon isotope ratio ranged from -12 per mil in the early ore fluids to -9.5 per mil in the latest stages. These values are too negative to be derived purely from magmatic sources but are consistent with the circulation of fluids

through sediments containing organic matter. Sulfur isotope ratios of sulfides associated with the gold mineralization range from -7 to +9 per mil, clustering about 0 per mil. It is possible to derive these values from a felsic igneous source as well as from sedimentary and greenstone sources in the intruded rocks. The isotopic evidence, coupled with the need for CO• to cause the widespread carbonate alteration, suggests that the source for the ore fluid probably was metamorphic devolatilization at greenschist or amphibolite facies. Contributions from deeply circulated, highly evolved meteoric and magmatic waters are possible.

Although the source(s) of the ore fluid remains un- resolved, its characteristics can be reasonably well defined. Based on fluid inclusion studies, primary ore deposition appears to have been from fluids of low salinity (3 wt % NaC1 equiv) with a significant CO• content (mole fraction of 0.10, rarely 0.25), minor CH4 content (mole fraction of 0.05, rarely 0.25), and densities of 0.95 g/cm 3, at temperatures of 350øC and pressures of up to 1.75 kbars (7 km depth). Later pseudosecondary fluids were even more dilute (1 wt % NaC1 equiv), with much lower CO2 (0.05 mole fraction), no CH4, and densities of 0.97 g/cm 3, at temperatures below 250øC and pressures of 1.0 kbar. Secondary fluids contained less than 1 wt percent NaC1 equiv, with nondetectable CO• and densities of 1.0 g/cm 3, at temperatures of 180øC. The progression of fluids may be explained by dilution of a hot, car-


bonic fluid with increasing amounts of cool, less saline meteoric water.

Computer modeling of the fluid responsible for the observed alteration assemblages and gold deposition, using chloride complexes, is supported by the observed CO2 and CH4 in the fluid inclusions. The ore fluid was slightly acid (pH = 4.5) at the 350øC and 1.75 kbars suggested by fluid inclusion studies and was a dilute solution of about 3 wt percent NaC1 equiv (Na/K = 8:1), containing significant CO2 (log fugacity -- 2.5) and minor CH4. It was reducing (fo• about 10 -aø bars, andJ• about 10 -7 bars) and contained a calculated 0.2 X 10-" m AuC12 (0.2 ppb Au). This is significantly less than most previously published estimates of 20 ppb but is about the same as average present-day mineralizing fluids. The main deposition of gold was accompanied by deposition of quartz, muscovite, ankerite, pyrite, and minor graphite. Chlorite, pyrrhotite, and albite-stable areas are not compatible with gold deposition. Gold precipitation appears to have been by reduction ofAu + by donation of an electron from a concurrent oxidation of sulfur in H•S to form pyrite. Deposition occurred mainly in response to reaction with, and pyritization of, wall rock. The results of this study suggest that further thermodynamic modeling, using thiosulfide complexes and considering temperature variations, is warranted.

Acknowledgments

Thanks are extended to the Corona Corporation for access to underground workings and field support. Critical reviews of the manuscript by two Economic Geology reviewers are appreciated. Research at the University of British Columbia has been supported by an I. W. Killam predoctoral fellowship to the senior author and grants to Godwin from the Natural Sci- ences and Engineering Research Council of Canada, the British Columbia Ministry of Energy, Mines and Petroleum Resources (BCMEMPR), and the Canada- British Columbia Mineral Development Agreement. Fluid inclusion studies were completed using equipment at the BCMEMPR in Victoria. CO2 from oxygen isotope extractions, performed at the Geological Sur- vey of Canada (Ottawa) with the assistance of W. C. Cornell, was analyzed in a mass spectrometer at the Ottawa Center for Geoscience Studies/Geological Survey of Canada Stable Isotope Laboratory by G. St. Jean.

May 15, December 6, 1990 REFERENCES

Albino, G. V., 1987, Contrasting styles of gold mineralization in the Sierra Nevada metamorphic belt, California: Unpub. manuscript, Univ. Western Ontario, Dept. Geology, 26 p.

Andrews, A. J., Jugon, H., Durocher, M., Corfu, F., and Lavigne, M. J., 1986, The anatomy of a gold-bearing greenstone belt:

Red Lake, northwestern Ontario, Canada, in Macdonald, A. J., ed., Gold '86: Willowdale, Ontario, Konsult Internat., p. 3-26.

Anovitz, L. M., and Essene, E. J., 1987, Phase equilibria in the system CaCO•-MgCO3-FeCO•: Jour. Petrology, v. 28, p. 389- 414.

Bancroft, G. M., and Jean, G., 1982, Gold deposition at low temperature on sulphide minerals: Nature, v. 298, p. 730-731.

Battey, M. H., 1951, Alkali metasomatism and the petrology of some keratophyres: Geol. Mag., v. 92, p. 104-126.

Bellamy, J. R., and Arnold, R., 1985, Bralorne mine project: Report on the 1984 exploration work: Vancouver, British Columbia, E&B Explorations Inc., March 1985, unpub. rept., 25 p.

Bertoni, C. H., 1983, Gold production in the Superior province of the Canadian Shield: Canadian Inst. Mining Metallurgy Bull., v. 76, no. 857, p. 62-69.

Bohlke, J. K., and Kistler, R. W., 1986, Rb-Sr, K-At, and stable isotope evidence for the ages and sources of fluid components of gold-bearing quartz veins in the northern Sierra Nevada foothills metamorphic belt, California: ECON. GEOL., v. 81, p. 296-322.

Borthwick, J., and Harmon, R. S., 1982, A note regarding CIF3 as an alternative to BrF5 for oxygen isotope analysis: Geochim. et Cosmochim. Acta, v. 46, p. 1665-1668.

Bottinga, Y., and Javoy, M., 1975, Oxygen isotope partitioning among the minerals in igneous and metamorphic rocks: Rev. Geophysics Space Physics, v. 13, p. 401-418.

Bowers, T. S., and Helgeson, H. C., 1983a, Calculation of the thermodynamic and geochemical consequences of nonideal mixing in the system H•O-CO•-NaCI on phase relations in geologic systems: Equation of state for H•O-CO•-NaCI fluids at high pressures and temperatures: Geochim. et Cosmochim. Acta, v. 47, p. 1244-1275.

1983b, Calculation of thermodynamic and geochemical cousequenees of nonideal mixing in the system H•O-CO•-NaCI on phase relations in geologic systems: Metamorphic equilibria at high pressures and temperatures: Am. Mineralogist, v. 68, p. 1059-1075.

Boyle, B. W., 1961, The geology, geochemistry and origin of the gold deposits of the Yellowknife district: Canada Geol. Survey Mem. 310, 135 p.

• 1979, The geochemistry of gold and its deposits: Canada Geol. Survey Bull. 280, p. 228-232.

Bozzo, A. T., Chen, J. B., and Barduhn, A. J., 1973, The properties of the hydrates of chlorine and carbon dioxide, in Delyanni, A., and Delyanni, E., eds., Internat. Symposium on Fresh Water from the Sea, 4th, v. 3, p. 437-451.

Brown, K. L., 1986, Gold deposition from geothermal discharges in New Zealand: ECON. GEOL., v. 81, p. 979-983.

Brown, P. E., and Lamb, W. M., 1986, Mixing of H•O-CO• in fluid inclusions; geobarometry and Arebean gold deposits: Geochim. et Cosmochim. Acta, v. 50, p. 847-852.

Buisson, G., and Leblanc, M., 1987, Gold in mantle peridotites from upper Proterozoic ophiolites in Arabia, Mall, and Morocco: ECON. GEOL., v. 82, p. 2091-2097.

Cairnes, C. E. (1937): Geology and mineral deposits of the Bridge River mining camp, B.C.: Canada Geol. Survey Mem. 213, 140 P.

Campbell, D. D., 1975, Bridge River gold camp: Vancouver, B.C., Dolmage Campbell and Assoc., Ltd., unpub. rept., 7 p.

Cameron, E. M., 1989a, Derivation of gold by oxidative metamorphism of a deep ductile shear zone: Part 1. Conceptual model: Jour. Geochem. Explor., v. 31, p. 135-144.

-- 1989b, Derivation of gold by oxidative metamorphism of a deep ductile shear zone: Part 2. Evidence from the Ramble belt, south Norway: Jour. Geochem. Explor., v. 31, p. 145- 149.

-- 1989c, Review of"Archean Lode Gold Deposits in Ontario," by A. C. Colvine, J. A. Fyon, K. B. Heather, S. Marmount, P.M. Smith and D. G. Troop, Ontario Geological Survey Mis-


cellaneous Paper 139, 1988: Jour. Geochem. Explor., v. 31, p. 329-334.

Church, B. N., 1987, Geology and mineralization of the Bridge River mining camp (92J/15, 920/2, 92J/10): British Columbia Ministry Energy, Mines and Petroleum Resources, Geol. Field- work, 1986, Paper 1987-1, p. 23-29.

Clayton, R. N., and Mayeda, T. K., 1963, The use of bromine pentafiuoride in the extraction of oxygen from oxides and silicates for isotopic analyses: Geochim. et Cosmochim. Acta, v. 27, p. 43-52.

Clayton, R. N., O'Neil, J. R., and Mayeda, T. K., 1972, Oxygen isotope exchange between quartz and water: Jour. Geophys. Research, v. 77, p. 3057-3067.

Collins, P. L. F., 1979, Gas hydrates in COg-bearing fluid inclusions and the use of freezing data for estimation of salinity: ECON. GEOL., v. 74, p. 1435-1444.

Colvine, A. C., Andrews, A. J., Cherry, M. E., Durocher, M. E., Fyon, A. J., Lavigne, M. J. Jr., Macdonald, A. J., Marmount, S., Poulsen, K. H., Springer, J. S., and Troop, D. G., 1984, An integrated model for the origin of Archean lode gold deposits: Ontario Geol. Survey Open-File Rept. 5524, 103 p.

Colvine, A. C., Fyon, J. A., Heather, K. B., Marmount, S., Smith, P.M., and Troop, D. G., 1988, Archean lode gold deposits in Ontario: Ontario Geol. Survey Misc. Paper 139, 136 p.

Coveney, R. M., Jr., 1981, Gold quartz veins and auriferous granite at the Oriental mine, Alleghany district, California: ECON. GEOL., v, 76, p. 2176-2199.

Dolmage, V., 1934, The Cariboo and Bridge River goldfields, British Columbia: Canadian Inst. Mining Metall. Trans., v. 37, p. 405-430.

Fuchter, W. A. H., and Hodgson, C. J., 1986, Gold deposits of the North Western mining camp, Gwanda greenstone belt, Zimbabwe, in Macdonald, A. J., ed., Gold '86: Willowdale, On- tario, Konsult Internat., p. 255-269.

Fyfe, W. S., and Henley, R. W., 1973, Some thoughts on chemical transport processes with particular reference to gold: Minerals Sci. Eng., v. 5, p. 295-303.

Fyfe, W. S., and Kerrich, R., 1984, Gold: Natural concentration processes, in Foster, R. P., ed., Gold '82: Rotterdam, A. A. Balkema Pub., p. 99-127.

Godwin, C. I., Watson, P. H., and Shen, K., 1986, Genesis of the Lass vein system, Beaverdell silver camp, south-central British Columbia: Canadian Jour. Earth Sci., v. 23, p. 1615-1626.

Goldfarb, R. J., Leach, D. L., Pickthorn, W. J., and Paterson, C. J., 1988, Origin of lode-gold deposits of the Juneau gold belt, southeastern Alaska: Geology, v. 16, p. 440-443.

Goldfarb, R. J., Leach, D. L., Rose, S.C., and Landis, G. P., 1989, Fluid inclusion geochemistry of gold-bearing quartz veins of the Juneau gold belt, southeastern Alaska: Implications for ore genesis: ECON. GEOL. MON. 6, p. 363-375.

Gresens, R, L., 1967, Composition-volume relationships of metasomatism: Chem. Geology, v. 2, p. 47-55.

Groves, D. I., Phillips, G. N., Ho, S. E., Henderson, C. A., Clark, M. E., and Woad, G. M., 1984, Controls on distribution of Ar- chean hydrothermal gold deposits in Western Australia, in Fos- ter, R. P., ed., Gold '82: Rotterdam, A. A. Balkema Pub., p. 689-711.

Groves, D, I., Phillips, G. N., Ho, S. E., Houston, S. M., and Stand- ing, C. A., 1987, Craton-scale distribution of Archeau greenstone gold deposits: Predictive capacity of the metamorphic model: ECON. GEOL., v. 82, p. 2045-2058.

Guha, J., Gauthier, A., Vallee, M., Descarreaux, J., and Lange- brard, F., 1982, Gold mineralization patterns at the Doyon mine (Silverstack, Bousquet, Quebec): Canadian Inst. Mining Met- allurgy Spec. Vol. 24, p. 50-57.

Hall, R. S., and Rigg, D. M., 1986, Geology of the West Anticline zone, Musselwhite prospect, Opapimiskan Lake, Ontario, Can- ada, in Macdonald, A. J., ed., Gold '86: Willowdale, Ontario, Konsult Internat., p. 124-136.

Haugerud, R., 1985, Geology of the Hozameen Group and Ross Lake shear zone, Maselpanic area, northern C•tscades, southwestern British Columbia: Unpub. Ph.D. thesis, Seattle, Univ. Washington, 263 p.

Helgeson, H. C., 1969, Thermodynamics of hydrothermal systems at elevated temperatures and pressures: Am. Jour. Sci., v. 167, p. 729-804.

Helgeson, H. C., and Garrels, R. M., 1968, Hydrothermal transport and deposition of gold: ECON. GEOL., v. 63, p. 622-635.

Helgeson, H. C., and Lichtner, P. C., 1987, Fluid flow and mineral reactions at high temperatures and pressures: Geol. Soc. London Jour., v. 144, p. 313-326.

Helgeson, H. C., Brown, T. H., Nigrini, A., and Jones, T. A., 1970, Calculation of mass transfer in geochemical processes involving aqueous solutions: Geochim. et Cosmochim. Acta, v. 34, p. 569- 592.

Helgeson, H. C., Delany, J. M., Nesbitt, H. W., and Bird, D. K., 1978, Summary and critique of the thermodynamic properties of rock forming minerals: Am. Jour. Sci., v. 278-A, 229 p.

Hodgson, C. J., and MacGeehan, P. J., 1982, A review of the geologic characteristics of "gold only" deposits in the Superior province of the Canadian Shield: Canadian Inst. Mining Met- allurgy Spec. Vol. 24, p. 211-229.

Hodgson, C. J., Chapman, R. S. G., and MacGeehan, P. J., 1982, Application of exploration criteria for gold deposits in the Su- perior province of the Canadian Shield to gold exploration in the Cordillera, in Levinson, A. A., ed., Precious metals in the northern cordillera: Calgary, Alberta, Assoc. Explor. Geo- chemists, p. 174-206.

Hollister, L. S., 1981, Information intrinsically available from fluid inclusions: Mineralog. Assoc. Canada Short Course Handbook, v. 6, p. 1-12.

Hollister, L. S., and Burruss, R. C., 1976, Phase equilibria in fluid inclusions from the Khtada Lake metamorphic complex: Geo- chim. et Cosmochim. Acta, v. 40, p. 163-175.

Hutchinson, R. W., and Burlington, J. L., 1984, Some broad characteristics of greenstone belt gold lodes, in Macdonald, A. J., ed., Gold '86: Willowdale, Ontario, Konsult Internat., p. 339- 371.

Joubin, F. R., 1948, Structural geology of the Bralorne and Pioneer mines, Bridge River district, British Columbia: Western Miner, July 1948, p. 39-50.

Keays, R. R., 1984, Archeau gold deposits and their source rocks: The upper mantle com•ection, in Macdonald, A. J., ed., Gold '86: Willowdale, Outario, Konsult Internat., p. 17-51.

Kerr, P. F., 1959, Optical mineralogy, 3rd ed.: New York, McGraw-Hill, 442 p.

Kerrich, R., 1983, Geochemistry of gold deposits in the Abitibi greenstone belt: Canadian Inst. Mining Metallurgy Spec. Vol. 27, 75 p.

-- 1987, The stable isotope geochemistry of Au-Ag vein deposits in metamorphic rocks: Mineralog. Assoc. Canada Short Course Handbook, v. 13, p. 287-336.

Kerrich, R., and Fyfe, W. S., 1981, The gold-carbonate •tssociation: Source of COg, and COg fixation reactions in Archeau lode deposits: Chem. Geology, v. 33, p. 265-294.

Kerrich, R., and Watson, G. P., 1984, The Mac•tssa mine Archeau lode gold deposit, Kirkland Lake, Ontario: Geology, patterns of alteration, and hydrothermal regimes: ECON. GEOL., v. 79, p. 1104-1130.

Kishida, A., and Kerrich, R., 1987, Hydrothermal alteration zoning and gold concentration at the Kerr-Addison Archeau lode gold deposit, Kirkland Lake, Ontario: ECON. GEOL., v. 82, p. 649- 690.

Krupp, R. E., and Seward, T. M., 1987, The Rotokawa geothermal system, New Zealand: An active epithermal gold-depositing environment: ECON. GEOL., v. 82, p. 1109-1129.

Landfeldt, L. A., 1987, The geology of the Mother Lode gold belt, Sierra Nevada Foothills metamorphic belt, California, in

BRALORNE An DEPOSIT, B.C. 351

Kisvarsanyi, G., and Grant, S. R., eds., North American conference on tectonic control of ore deposits and the vertical and horizontal extent of ore systems: Proceedings volume: Rolla, Univ. Missouri-Rolla Press, p. 47-56.

Leitch, C. H. B., 1981, Secondary alkali feldspars in porphyry systems: Canadian Inst. Mining Metallurgy Bull., v. 74, no. 831, p. 83-88.

-- 1989, Geology, wallrock alteration, and characteristics of the ore fluid at the Bralorne mesothermal gold quartz vein deposit, southwestern B.C.: Unpub. Ph.D. thesis, Vancouver, Univ. British Columbia, 483 p.

-- 1990, Bralorne: A mesothermal, shield-type vein gold deposit of Cretaceous age in southwestern British Columbia: Ca- nadian Inst. Mining Metallurgy Bull., v. 83, no. 941, p. 53-80.

Leitch, C. H. B., and Day, S. J., 1990, NEWGRES: A TurboPascal program to solve a modified version of Gresens' hydrothermal alteration equation: Computers Geosci., v. 16, p. 925-932.

Leitch, C. H. B., and Godwin, C. I,, 1986, Geology of the Bralorne- Pioneer gold camp (92J/15): British Columbia Ministry Energy, Mines Petroleum Resources, Geol. Fieldwork 1985, Paper 1986-1, p. 311-317.

-- 1987, The Bralorne gold vein deposit: An update (92J/15): British Columbia Ministry Energy, Miues Petroleum Resources, Geol. Fieldwork 1986, Paper 1987-1, p. 35-38.

-- 1988, Isotopic ages, wallrock chemistry and fluid inclusion data from the Bralorne gold vein deposit (92J/15w): British Columbia Ministry Energy, Mines Petroleum Resources, Geol. Fieldwork 1987, Paper 1988-1, p. 301-324.

Leitch, C. H. B., Dawson, K. M., and Godwin, C. I., 1989, Early Late Cretaceous-early Tertiary gold mineralization: A galena lead isotope study of the Bridge River mining camp, southwestern British Columbia, Canada: ECON. GEOL., v. 84, p. 2226--2236.

Leitch, C. H. B., van der Heyden, P., Godwin, C. I., Armstrong, R. L., and Harakal, J., 1991, Geochronometry of the Bridge River camp, southwestern British Columbia: Canadian Jour. Earth Sci., v. 28, no. 2, p. 261-274.

Leroy, J., 1979, Contribution •t l'etalinnage de la pression interne des inclusions fluides lor de leur decrepitation: Bull. Mineral- ogle, v. 102, p. 584-593.

Linnen, R., 1985, Contact metamorphism, wallrock alteration, and mineralization at the Trout Lake stockwork molybdenum deposit, southeastern British Columbia: Unpub. M.Sc. thesis, McGill Univ., 220 p.

Ludden, J. N., Daigneault, R., Robert, F., and Taylor, R. E., 1984, Trace element mobility in alteration zones associated with Ar- chcan Au lode deposits: ECON. GEOL., v. 79, p. 1131-1141.

Lydon, J. W., 1977, The significance of metal ratios of hydrothermal ore deposits: Unpub. Ph.D. thesis, Kingston, Queen's Univ, 353 p.

MaeGeehan, P. J., Sanders, T., and Hodgsou, C. J., 1982, Meter- wide veins and a kilometer-wide anomaly: Wall-rock alteration at the Campbell Red Lake aud Dickenson gold mines, Red Lake district, Ontario: Canadian Inst. Mining Metallurgy Bull., v. 75, no. 841, p. 90-102.

Magaritz, M., and Taylor, H. P., Jr., 1976, •O/•60 and D/H studies along a 500 km traverse across the Coast Range batholith and its country rocks, central British Columbia: Canadian Jour. Earth Sci., v. 13, p. 1514-1536.

1986, •SO/•60 and D/H studies of plutonie granitic and metamorphic rocks across the Cordilleran batholiths of southern British Columbia: Jour. Geophys. Research, v. 91, p. 2193- 2217.

Maheux, P. J., Muehlenbachs, K., aud Nesbitt, B. E., 1987, Evi- dence of highly evolved ore fluids responsible for sulfide associated gold mineralization iu the Bridge River district, B.C. labs.l: Geol. Assoc. Canada Program with Abstracts, v. 12, p. 70.

Malinin, S. D., 1974, Thermodynamics of the H20-CO2 system: Geochem. Internat., v. 11, p. 1060-1085.

Matthews, A., Goldsmith, J. R., and Clayton, R. M., 1983, Oxygen isotope fractionations involving pyroxenes: The calibration of mineral-pair geothermometers: Geochim. et Cosmochim. Acta, v. 47, p. 631-644.

McCrea, J. M., 1950, The isotopic chemistry of carbonates and a palcotemperature scale: Jour. Chem. Physics, v. 18, p. 849- 853.

McHugh, J. B., 1988, Concentration of gold in natural waters: Jour. Geochem. Explor., v. 30, p. 85-94.

Melling, D. R., Watkinson, D. H., Poulsen, K. H., Chorlton, L. B., and Hunter, A.D., 1986, The Cameron Lake gold deposit, northwestern Ontario, Canada: Geological setting, structure, and alteration, in Macdonald, A. J., ed., Gold '86: Willowdale, Ontario, Konsult Internat., p. 149-169.

Monger, J. W. H., 1984, Cordilleran tectonics: A Canadian per- spective: Soc. Geol. France Bull., ser. 7, v. 26, no. 2, p. 255- 278.

Morasse, S., Hodgson, C. J., Guha, J., and Coulombe, A., 1986, Preliminary report on the geology of the Lac Shortt gold deposit, Demaraisville area, Quebec, Canada, in Macdonald, A. J., ed., Gold '86: Willowdale, Ontario, Konsult Internat., p. 191-196.

Nesbitt, B. E., 1988, Gold deposit continuum: A genetic model for lode Au mineralization in the continental crust: Geology, v. 16, p. 1044-1048.

Nesbitt, B. E., and Muehlenbachs, K., 1988, Genetic implications of the association of mesothermal gold deposits with major strike-slip fault systems, in Kisvarsanyi, G., and Grant, S. R,, eds., North American conference on tectonic control of ore deposits and the vertical and horizontal extent of ore systems: Proceedings volume: Rolla, Univ. Missouri-Rolla Press, p. 57- 66.

Nesbitt, B. E., Murowchick, J, B. and Muehlenbachs, K., 1986, Dual origins of lode gold deposits in the Canadian Cordillera: Geology, v. 14, p. 506-509,

-- 1987, Genesis ofAu, Sb, and Hg deposits in accreted terranes of the Canadian Cordillera labs.l: Geol. Assoc. Canada Program with Abstracts, v. 12, p. 70.

Nordine, G., 1983, Geological report on the Pacific Eastern property, Lillooet Mining Division: Vancouver, British Columbia, Amir Mines, Ltd., priv. rept., 48 p.

Ohmoto, H., and Rye, R. O., 1979, Isotopes of sulfur and carbon, in Barnes, H. L., ed., Geochemistry of hydrothermal ore deposits: New York, Wiley Intersci., p. 509-567.

Ohmoto, H., and Kerrick, D. M., 1977, Devolatilization equilibria in graphitic systems: Am. Jour. Sci., v. 277, p. 1013-1044.

O'Neil, J. R., and Taylor, H. P., Jr., 1969, Oxygen isotope equilibrium between muscovite and water: Jour. Geophys. Research, v. 74, p, 6012-6022.

Parry, W. T., 1986, Estimation of Xco•, P, and fluid inclusion volume from fluid inclusion temperature measurements in the system NaCI-CO2-H20: ECON. GEOL., v. 81, p. 1009-1013.

Pattison, E. F., Sauerbrei, J. A. Hannila, J. J., and Church, J. F., 1986, Gold mineralization in the Casa-Berardi area, Quebec, Canada, in Macdonald, A. J., ed., Gold '86: Willowdale, Ontario, Konsult Internat., p. 170-183.

Perkins, E. H., 1980, A re-investigation of the theoretical basis for the calculation of isothermal-isobaric mass transfer in geochemical systems involving an aqueous phase: Unpub. M.Sc. thesis, Vancouver, Univ. British Columbia, 176 p.

Phillips, G. N., 1986, Geology and alteration in the Golden Mile, Kalgoorlie: ECON. GEOL., v. 81, p. 779-808.

Phillips, G. N., and Groves, D. I., 1983, The nature of Archcan gold-bearing fluids as deduced from gold deposits of Western Australia: Geol. Soc. Australia Jour., v. 30, p. 25-39.

-- 1984, Fluid access and fiuid-wallrock interaction in the genesis of the Arcbean gold-quartz vein deposit at Hunt mine,


Kambalda, Western Australia, in Foster, R. P., ed., Gold '82: Rotterdam, A. A. Balkema Pub., p. 389-416.

Phillips, G. N., Groves, D. I., and Martyn, J. E., 1983, An epi- genetic origin for Archean banded iron-formation-hosted gold deposits: E(•ON. GEOL., v. 79, p. 162-171.

Pickthorn, W. L., Goldfarb, R. J., and Leach, D. L., 1987, Com- ment and reply on "Dual origins of lode gold deposits in the Canadian Cordillera": Geology, v. 15, p. 471-472.

Poole, A. W., 1955, The geology and analysis of vein and fault structure of the Bralorne mine: Canadian Inst. Mining Metal- lurgy Bull., November 1955, p. 733-737 (Trans., v. 58, p. 433- 437).

Potter, C. J., 1986, Origin, accretion, and postaccretionary evolution of the Bridge River terrane, southwest British Columbia: Tectonics, v. 5, p. 1027-1041.

Potter, R. W., and Brown, D. L., 1977, The volumetric properties of aqueous sodium chloride solutions from 0 ø to 500øC at pressures up to 2000 bars based on a regression of available data in the literature: U.S. Geol. Survey Bull. 1421-C, 36 p.

Potter, R. W., Clynne, M. A., and Brown, D. L., 1978, Freezing point depression of aqueous sodium chloride solutions: ECON. GEOL., v. 73, p. 284-285.

Ray, G. E., 1986, The Hozameen fault system and related Co- quihalla serpentine belt of southwestern British Columbia: Ca- nadian Jour. Earth Sci., v. 23, p. 1022-1041.

Read, J. J., and Meinert, L. D., 1986, Gold-bearing quartz vein mineralization at the Big Hurrah mine, Seward Peninsula, Alaska: ECON. GEOL., v. 81, p. 1760-1774.

Robert, F., and Brown, A. C., 1986a, Archean gold-bearing quartz veins at the Sigma mine, Abitibi greenstone belt, Quebec: Part I. Geologic relations and formation of the vein system: ECON. GEOL., v. 81, p. 578-592.

-- 1986b, Archean gold-bearing quartz veins at the Sigma mine, Abitibi greenstone belt, Quebec: Part II. Vein paragenesis and hydrothermal alteration: ECON. GEOL., v. 81, p. 593-616.

Robert, F., and Kelly, W. C., 1987, Ore-forming fluids in Archean gold-bearing quartz veins at the Sigma mine, Abitibi greenstone belt, Quebec, Canada: ECON. GEOL., v. 82, p. 1464-1482.

Roberts, R. G., 1987, Ore deposit models #11: Archean lode gold deposits: Geosci. Canada, v. 14, no. 1, p. 37-52.

Roedder, E., 1979, Fluid inclusions as samples of ore fluids, in Barnes, H. L., ed., Geochemistry of hydrothermal ore deposits: New York, Wiley Intersci., p. 684-737.

-- 1984, Fluid inclusions: Rev. Mineralogy, v. 12, 644 p. Romberger, S. B., 1986, The solution chemistry of gold applied

to the origin of hydrothermal deposits: Canadian Inst. Mining Metallurgy Spec. Vol. 38, p. 168-186.

Rusmore, M. E., 1987, Geology of the Cadwallader Group and the Intermontane-Insular superterrane boundary, southwestern British Columbia: Canadian Jour. Earth Sci., v. 24, p. 2279- 2291.

Rusmore, M. E., Potter, C. J., and Umhoefer, P. J., 1988, Middle Jurassic terrane accretion along the western edge of the Inter- montane superterrane, southwestern British Columbia: Geol- ogy, v. 16, p. 891-894.

Rye, R. O., and Ohmoto, H., 1974, Sulfur and carbon isotopes and ore genesis: A review: ECON. GEOL., v. 69, p. 826-842.

Santosh, M., 1986, Ore fluids in the auriferous Champion reef of Kolar, South India: ECON. GEOL., v. 81, p. 1546-1552.

Schiarizza, P., Gaba, R. G., Glover, J. K., and Garver, J. I., 1989, Geology and mineral occurrences of the Tyaughton Creek area (920/2, 92J/15, 16): British Columbia Ministry Energy, Mines Petroleum Resources, Geol. Fieldwork 1988, Paper 1989-1, p. 115-130.

Seward, T. M., 1973, Thio complexes of gold and the transport of gold in hydrothermal ore solutions: Geochim. et Cosmochim. Acta, v. 37, p. 379-399.

-- 1984, The transport and deposition of gold in hydrothermal

systems, in Foster, R. P., ed., Gold '82: Rotterdam, A. A. Bal- kema Pub., p. 165-181.

Shelton, K. L., So, C.-S., and Chang, J.-S., 1988, Gold-rich mesothermal vein deposits of the Republic of Korea: Geochemical studies of the Jungwon gold area: ECON. GEOL., v. 83, p. 1221- 1237.

Shenberger, D. M., 1985, Gold solubility in aqueous sulfide solutions: Unpub. M. Sc. thesis, Pennsylvania State Univ., 102 p.

Sibson, R. H., 1981, Fluid flow accompanying faulting: Field evidence and models: Am. Geophys. Union, Maurice Ewing Ser., v. 4, p. 593-603.

Sibson, R. H., Robert, F., and Poulsen, K. H., 1988, High-angle reverse faults, fluid-pressure cycling, and mesothermal gold- quartz deposits: Geology, v. 16, p. 551-555.

Sketchley, D. A., and Sinclair, A. J., 1987, Multi-element litho- geochemistry of alteration associated with gold-quartz miner- alisation of the Erickson mine, Cassiar district: British Columbia Ministry Energy, Mines Petroleum Resources, Geol. Fieldwork 1986, Paper 1987-1, p. 57-63.

Smith, T. J., Cloke, P. L., and Kesler, S. E., 1984, Geochemistry of fluid inclusions from the McIntyre-Hollinger gold deposit, Timmins, Ontario, Canada: ECON. GEOL., v. 79, p. 1265-1285.

Stevenson, J. S., 1958, Bridge River area, British Columbia: British Columbia Ministry Energy, Mines Petroleum Resources, unpub. rept., 320 p.

Studemeister, P. A., and Kilias, S., 1987, Alteration pattern and fluid inclusions of gold-bearing quartz veins in Archean trondhjemite near Wawa, Ontario, Canada: E(•ON. GEOL., v. 82, p. 429-439.

Sugiyama, T., 1986, Short report on preliminary test of new decrepitation system on samples from University of British Co- lumbia, Canada: Tokyo, Japan, Mitsubishi Metal Corp., Central Research Inst., unpub. rept., 16 p.

Swanenberg, H. E. C., 1979, Phase equilibria in carbonic systems, and their application to freezing studies of fluid inclusions: Cont. Mineralogy Petrology, v. 68, p. 303-306.

Taylor, B. E., 1987, Stable isotope geochemistry of ore-forming fluids: Mineralog. Assoc. Canada Short Course Handbook, v. 13, p. 337-418.

Taylor, B. E., and Gerlach, T. M., 1986, Mantle degassing at Long Valley, Steamboat Springs, and the Coso Range [abs.]: EOS, v. 65, p. 1153.

Taylor, H. P., Jr., 1974, The application of oxygen and hydrogen isotope studies to problems of hydrothermal alteration and ore deposition: ECON. GEOL., v. 69, p. 843-883.

-- 1979, Oxygen and hydrogen isotope relationships in hydrothermal ore deposits, in Barnes, H. L., ed., Geochemistry of hydrothermal ore deposits: New York, Wiley Intersci., p. 236-277.

Thompson, M. L., 1986, Petrology of the Crixas gold deposit, Brazil: Evidence for gold associated with hydrothermal alteration, subsequent to metamorphism, in Macdonald, A. J., ed., Gold '86: Willowdale, Ontario, Konsult Internat., p. 284-296.

Tilling, R. I., Gottfried, D., and Rowe, J. J., 1973, Gold abundance in igneous rocks: Bearing on gold mineralization: ECON. GEOL., v. 68, p. 168-186.

Trommsdorff, V., and Skippen, G., 1986, Vapour loss ("boiling") as a mechanism for fluid evolution in metamorphic rocks: Contr. Mineralogy Petrology, v. 94, p. 317-322.

Walsh, J. F., Kesler, S. E., Duff, D., and Cloke, P. L., 1988, Fluid inclusion geochemistry of high-grade, vein-hosted gold ore at the Pamour mine, Porcupine camp, Ontario: ECON. GEOL., v. 83, p. 1347-1368.

Walther, J. V., and Helgeson, H. C., 1977, Calculation of the thermodynamic properties of aqueous silica and the solubility


of quartz and its polymorphs at high pressures and temperatures: Am. Jour. Sci., v. 277, p. 1315-1351.

Weir, R. H., Jr., and Kerrick, D. M., 1987, Mineralogic, fluid inclusion, and stable isotope studies of several gold mines in the Mother Lode, Tuolumne and Mariposa Counties, California: ECON. GEOL., v. 82, p. 328-344.

Wenner, D. B., and Taylor, H. P., Jr., 1971, Temperatures of serpentinization of ultramarie rocks based on lSO/160 fractionation between coexisting serpentine and magnetite: Contr. Mineralogy Petrology, v. 32, p. 165-185.

Wheeler, J. O., and McFeely, P., 1987, Tectonic assemblage map and correlation chart of the Canadian Cordillera and adjacent parts of the United States of America: Canada Geol. Survey Open-File Rept. 1.56,5 (1:2,000,000 scale map).

Winkler, H. G. F., 197 l, Petrogenesis of metamorphic rocks, 3rd ed.: New York, Springer-Verlag, 237 p.

Wood, P. C., Burrows, D. R., Thomas, A. V., and Spooner, E. T. C., 1986, The Hollinger-McIntyre Au-quartz vein system, Timrains, Ontario, Canada: Geological characteristics, fluid properties and light stable isotopes, in Macdonald, A. J., ed., Gold '86: Willowdale, Ontario, Konsult Internat., p..56-80.

Wood, S. A., 1987, Application of a multiphase ore mineral solubility experiment to the separation of base metal and gold mineralization in Arebean greenstone terrains: ECON. GEOL., v. 82, p. 1044-1048.

Woodsworth, G. J., 1977, Pemberton (92J) map-area, British Co- lumbia: Canada Geol. Survey Open-File Rept. 482 (1:2.50 000 scale map).

Workman, A. W., 1986, Geology of the McDermott gold deposit, Kirkland Lake area, northeastern Ontario, Canada, in Macdon- ald, A. J,, ed., Gold '86: Willowdale, Ontario, Konsult Internat., p. 184-190.

Documents

Geochemistry of Mineralizing Fluids in the Bralorne-Pioneer Mesothermal.pdf