Proceedings of the Symposium on the ECONOMIC MINERAL DEPOSITS OF THE SOUTHEAST: METALLIC ORE DEPOSITS Sponsored by the Southeastern Section of the Geological Society of America, Atlanta, Georgia April 6, 1989 Edited by Robert B. Cook Georgia Department of Natural Resources Environmental Protection Division Georgia Geologic Survey BULLETIN 117 I The papers presented herewith have not undergone review procedures of the Georgia Geologic Survey's Quality Assurance Plan. They do not necessarily reflect the technical opinions of the Georgia Geologic Survey. Proceedings of the Symposium on the ECONOMIC MINERAL DEPOSITS OF THE SOUTHEAST: METALLIC ORE DEPOSITS Sponsored by the Southeastern Section of the Geological Society of America, Atlanta, Georgia April 6, 1989 Edited by Robert B. Cook Department of Geology Auburn University Georgia Department of Natural Resources Lonice C. Barrett, Commissioner Environmental Protection Division Harold F. Reheis, Assistant Director Georgia Geologic Survey William H. McLemore, State Geologist Atlanta 1990 BULLETIN 117 PREFACE This Bulletin contains 10 of the 11 papers presented in the symposium "Economic Mineral Deposits of the Southeast: Metallic Ore Deposits" held as part of the Geological Society of America Southeastern Section Annual Meeting in Atlanta, April 6-7, 1989. Also included is an invited paper by Peng Sha and C. Michael Lesher that complements two other volume papers on the gold occurrences of the Goldville District of Tallapoosa County, Alabama. Seven of the volume's papers deal with various aspects of specific gold districts or occurrences including Georgia's Dahlonega district and Royal Vindicator deposit. Depositional types vary widely and include probable synmetamorph_ic, meta -epithermal, shear zone-hosted epigenetic, and placer. Other papers present details of such varied topics as the tin occurrences related to the Rockford Granite of the Alabama Northern Piedmont, the metallogeny of a central Piedmont accretionary plate margin, Zn-Pb resources of the central Kentucky district, and the geochemistry of America's first known PtAs occurrence. The idea for this proceedings volume is not unique. It represents the second in what hopefully will be regularly though intermittently published collections of papers based on similar symposia dealing with the economic geology of the metallic resources of the southeastern United States. The first such work, "Volcanogenic Sulfide and Precious Metal Mineralization in the Southern Appalachians" (edited by Kula c. Misra), is still available as Studies in Geology #16 of the University of Tennessee Department of Geological Sciences. My sincere appreciation is expressed to the authors who have taken the time to formalize and revise the text and illustrations of their symposium presentations. Special thanks go to the numerous reviewers who were invariably thorough, prompt and unselfish in their constructive criticism. Publication of this bulletin would not have been possible without the continued support of the Georgia Geologic survey and in particular the editorial assistance of Ms. Patricia Allgood. The diligent efforts and wordprocessing skills of Ms. Eva Lilly and Ms. Sheila Arington contributed markedly to the successful completion of this volume. Robert B. Cook Department of Geology Auburn University Auburn, AL 36849-5305 ii TABLE OF CONTENTS Thennobarometry of Synmetamorphic Gold Deposits, Goldville District, Northern Piedmont, Alabama (Harold H. Stowell and C. Michael Lesher).......................................................................... 1 Microthennometry of Fluid Inclusions in Auriferous Quartz Veins, Northern Piedmont, Alabama (Peng Sha and Michael Lesher)....................................................................................................... 18 Stable Isotope Geochemistry of Lode Gold Deposits, Northern Piedmont, Alabama (C. Michael Lesher, Harold H. Stowell, and Nathan L. Green)....................................... .. .................................... 32 The Royal Vindicator Gold Deposit, Haralson County, Georgia: A Metamorphosed Epithermal Hot-Springs Type Gold Deposit (Travis A. Paris) ...... .. ......................................................... ................... 50 Gold Deposits of the Dahlonega Belt, Northeast Georgia (George V. Albino) ......................................................... ......................................... 85 Paleoplacer Gold in the Lilesville Sand and Gravel Deposits, North Carolina (James R. Craig and John E. Callahan) .................................................................................................................. 121 The Use of Relict Phases with Placer Gold in Characterizing Lodes (Alan J. Driscoll, Jr., Donald L. Hall, and James R. Craig) .................................................................................... 142 A Geochemical Investigation of the Mason Mountain Sperrylite Occurrence, Macon County, North Carolina (Robert B. Cook and James R. Bumell) ................................................................. 163 The Central Piedmont Metallogenic Province, South Carolina (Steven K. Mittwede) ............................................................................................. 177 The Rockford Tin District, Alabama Piedmont: A Result of Na-Metasomatism Associated with Trondhjemitic Alteration of the Rockford Granite? (David J. Wesolowski and MarkS. Drummond) .................................................................................. 208 Polymetallic Sulfide/Fluorspar Mineralization in Western Kentucky: Lead and Zinc Reserves for the 21st Century (Ed L. Schrader, D. E. Gann and A: D. Bishop) ....................................................................................................................... 229 111 THERMOBAROMETRY OF SYNMETAMORPHIC GOLD DEPOSITS, GOLDVILLE DISTRICT, NORTHERN PIEDMONT, ALABAMA Harold H. Stowell and C. Michael Lesher Department of Geology, University of Alabama Tuscaloosa, AL 35487-0338 ABSTRACT The Goldville district comprises a 16 km long, northeast-trending group of inactive gold mines and prospects in the southeast part of the Ashland-Wedowee belt. Host rocks are metasedimentary rocks of the Wedowee Group that are polydeformed and have been metamorphosed to garnet zone (low- to medium-grade). Gold mines and prospects are associated exclusively with multi-generation quartzplagioclasepyrrhotite veins that are variably s boudinaged, folded, sheared, and transposed into an 2 axial planar fabric. Many veins are bordered by very thin (<1 em) selvages of massive chlorite and/or biotite. Veins, vein selvages, and host rocks are locally sheared by "late" chlorite-muscovite rich shear zones that apparently postdate quartz vein emplacement and peak metamorphism. The chemistry and texture of garnets in vein selvages, host rocks, and shear zones suggest prograde metamorphic growth with compositional control by host lithologies followed by minor "retrograde" modification. Garnet grains within late shear zones are rimmed with chlorite suggesting retrograde replacement. Synmetamorphic vein emplacement is indicated by therrnobarometric data from veins, vein selvages, host rocks in the Goldville district, and a sample from the Wedowee Group. Garnetbiotite, plagioclase-muscovite, and plagioclase-biotite-muscovite-garnet equilibria yield metamorphic conditions of 49050C ca. 5.8 kb for the Goldville district. Results from all textural occurrences in the district and within the Wedowee Group are indistinguishable. Garnet-biotite and plagioclase-biotite-muscovite-garnet equilibria yield higher metamorphic temperatures of about 575C ca. 6.7 kb for Emuckfaw Group schists within the Alexander City fault zone suggesting that this fault zone is a boundary between low-grade (greenschist facies) and medium-grade (amphibolite facies) metamorphic rocks. Textural relationships, mineral chemistry, and thermobarometric data indicate that the present distribution of gold results from mobilization during near-peak metamorphism and progressive D1-D2 deformation. INTRODUCTION The Goldville district comprises a linear group of 17 inactive mines and prospects trending northeast for 16 km across north-central Tallapoosa County (Fig. 1). The regional geology of the Northern Piedmont is described by Tuff (1978), Thomas and others (1980), and Guthrie and Lesher (in press). Gold mineralization in the district is hosted primarily by metasedimentary rocks of the Wedowee Group (Neathery and Reynolds, 1973; Guthrie and Lesher, in press), although the Dutch Bend and Hog Mountain deposits to the west are hosted by tonalite plutons (see Guthrie and Lesher, in press). The rocks in the Goldville district have experienced polyphase deformation (Muangnoicharoen, 1975; Guthrie and Dean, 1989) and greenschist to amphibolite facies 1 metamorphism (Neathery and Reynolds, 1973; Muangnoicharoen, 1975; Guthrie and Dean, 1989; Stowell et al., in press). Gold mines and prospects in the Goldville and Hog Mountain districts are associated exclusively with quartzplagioclasesulfide veins (Adams, 1930; Pardee and Park, 1948). Adams (1930) interpreted the mineralization as epigenetic, derived from granitic magmas at depth . Paris (1986) interpreted some of the host lithologies as volcaniclastic rocks, considered the gold mineralization to be strata-bound, and favored an exhalative origin for the mineralization. Neal (1986) and Neal and Cook (in press) also interpreted the gold to be strata-bound and syngenetic, but interpreted. the host rocks as metapelites. Recent work suggests that vein emplacement/mineralization within metamorphic rocks in the Goldville district occurred during greenschist facies metamorphism (Stowell et al., in press). The mineralogy, and K/Ar and Rb/Sr ages from alteration zones at Hog Mountain and at the Lowe mine suggest that mineralization at both locations likely occurred during the early stages of metamorphism (Green, Lesher, and Stowell, unpublished data). The principal aim of this report is to characterize temperatures and pressures during vein emplacement/gold mineralization and compare these with metamorphic conditions calculated for rocks in the Wedowee and Emuckfaw Groups adjacent to the district. Diamond drill core from the Jones, Lowe, and Tallapoosa mines, and samples from the Wedowee and Emuckfaw Groups provided the basis for detailed petrographic, mineralogical, and petrological studies. This study confirms that vein emplacement/gold mineralization occurred at 49050C and 5.81 kb, and indicates that metamorphism of the Emuckfaw Group, south of the Goldville district, in the Alexander City fault zone occurred at somewhat higher temperatures and pressures of 57550C and 6.71 kb. GEOLOGIC SETTING Host Lithologies Metasediment-hosted gold deposits in the Goldville district occur in greenschist to amphibolite facies lithologies of the Wedowee Group described by Guthrie and Dean (1989; Fig . 1). Lithologic associations and compositions, thin parallel bedding, and rarely-preserved graded bedding suggest that the rocks were deposited as distal turbidites (Thomas and others, 1980; Guthrie and Dean, 1989; Stowell et al., in press). South of the Goldville district the Wedowee Group is truncated by the Alexander City fault zone (Fig. 1). The Emuckfaw Group within this zone contains amphibolite facies metamorphic mineral assemblages (Maungnoicharoen, 1975; Guthrie and Dean, 1989). The proximity of these schists to phyllites which contain greenschist facies mineral assemblages has been attributed to widespread retrograde metamorphism of the Wedowee Group (Tull, 1978; Steltenpohl and Moore, 1988). Deformation Five phases of deformation have been recognized based on local cross-cutting relationships in the Goldville district (Guthrie and Dean , 1989) : isoclinal folds (D1), close to isoclinal quasiflexural folds and ductile to semi-brittle shear zones with northwest-directed transport (D2), upright to inclined quasi-flexural slip folds (D3), upright open folds (D4), and dextral strike-slip faults (D5). The regional deformations correlate in a general manner with structures recognized by Tull (1978; cf. Guthrie and Lesher, in press) . D1 through D3 structures are interpreted to result from progressive deformation during northwest-southeast oriented shortening (Stowell et al., in press). 2 EXPLANATION D. Hog Mountain tonalite zg Zana Granite ';-g~ "~"'I Elkahatchee Quartz Diorite Gneiss Cutnose Gneiss: c<: -, -,-,~-,~j ~~~- ~~~~ gr phyllite .0N.. 3 .! [ill grt-bt-chl-ser-qtz, . bt-chl-ser-qtz, and gr phyllite a. ::l 0 ~ a> Co) ,':\,;."";,-~ ~-/_"{~~/:-: -:;:;::::::::::::::..:: ~:'- / .rJ'-< ~< ..Y: : ;. ~0 . -> :: :7- : :;!.;~y,_:-~ j't/.;.;;1' \.. D ~ -1 0... grt-bt-chl-ser-qtz phyllite it 0 "0 a> ~ < "cC:l wee Cragford Phyllite Ill )( < :;( weu Wedowee Group undifferentiated emu Emuckfaw Group Abandoned mine N X Prospect pit 0 1 Mile 0 1 Kilometer \ I IL I Figure 1. Generalized geologic map of the New Site quadrangle showing mines in the Goldville district and locations of samples used for thermobarometry (modified after Guthrie and Dean, 1989). Mineralization Gold mines and prospects in the district are associated exclusively with multi-generation s, quartzplagioclasesulfide veins that cross-cut compositional layering and the schistosity (Stowell et al., in press). Veins are variably boudinaged , folded, sheared, and transposed into an s S2 axial planar fabric, suggesting emplacement during 2 development. Many veins are bordered by very thin (<1 em) selvages of massive chlorite and/or biotite, and a few are bordered by "late" chlorite- , muscovite-, and quartz-rich shear zones that apparently post-date quartz vein emplacement and peak metamorphism. These sharply defined shear zones locally truncate layering and veins, contain helicitic garnet porphyroblasts which are locally altered to chlorite, and commonly contain significantly smaller grain sizes than host rocks and adjacent vein selvages. Metamorphism Regional metamorphic mineral assemblages in the Wedowee Group have been attributed to an early amphibolite facies event during D1, followed by widespread greenschist facies retrograde metamorphism coeval with D2 (Tull, 1975, 1978; Steltenpohl and Moore, 1988). The earli est metamorphic event generated staurolite to kyanite zone assemblages (Neathery and Reynolds, 1973; Muangnoicharoen, 1975; Steltenpohl and Moore, 1988). Sillimanite zone was attained adjacent to some intrusive bodies (Drummond, 1986). Retrograde chlorite-muscovite zone assemblages occur locally (Steltenpohl and Moore, 1988). Phyllites and schists in the Wedowee Group contain garnet to sillimanite zone assemblages (Drummond, 1986; Steltenpohl and Moore, 1988). Many of the phyllites and schists in the Goldville district are characterized by garnet-biotite-muscovite-quartz and garnet-chloritemuscovite-quartz assemblages (Stowell et al., in press). A single sample (L0-2) containing partially replaced staurolite in a vein selvage suggests that metamorphic conditions locally reached those of staurolite zone. Peak metamorphic assemblages are locally retrograded to chlorite-sericite-quartzbiotite. This biotite zone assemblage indicates low grade conditions of the greenschist facies and is similar to "late" metamorphic assemblages recognized elsewhere in the Wedowee Group (Muangnoicharoen , 1975; Tull, 1978; Steltenpohl and Moore, 1988; Guthrie and Dean, 1989). ANALYTICAL METHODS Microanalyses of mineral grains were obtained by wavelength dispersive spectrometry on a CAMECA SX-50 electron microprobes at the University of Tennessee, Knoxville and the University of South Carolina, Columbia, or a JEOL 733 electron microprobe at Southern Methodist University, Dallas. All of the analyses were performed at 15 kV accelerating voltage, 15-20 nA specimen current, maximum of 30 second/1 00,000 counts on peaks, 10 second counts on backgrounds, and a beam diameter of 1-10 (.1m. PAP (Pouchou and Pichoir, 1985) and BenceAlbee (Bence and Albee, 1968) data reduction methods were used on the CAMECA ahd JEOL instruments, respectively. Comparison of analyses of a single grain obtained from two of the instruments and averaging of multiple analyses from one spot obtained on one instrument indicates a precision better than 2% (relative) for major elements, and a precision better than 5% (relative) for minor elements. PETROGRAPHY AND MINERAL CHEMISTRY The textural relationships and chemistry of minerals provide considerable information about equilibrium between veins and host rocks, diHerent generations of veins, coexisting vein minerals, and regional metamorphic minerals. These data provide a basis for constraining the timing of vein emplacement, metamorphic conditions during vein emplacement, and the conditions of regional metamorphism. 4 Garnet Garnet occurs as anhedral to euhedral grains within all common lithologies, as well as most vein selvages (Stowell et al., in press). Many grains contain helicitic structures, which are discordant with the fabric outside the porphyroblasts, indicating growth over an early foliation defined by ilmenite and quartz. Garnets commonly have sharp contacts with biotite, muscovite, and early chlorite . Some grains within shear zones are replaced by chlorite along margins and fractures; none of these grains were analyzed. All of the garnets can be modeled as four component aluminum garnets which are composed predominantly of almandine with lesser amounts of spessartine, grossular, and pyrope (Table 1). Gamet grains from the Tallapoosa mine are strongly zoned, whereas grains from the Lowe mine show only weak zonation (Stowell et al., in press). All zoned garnets exhibit zoning patterns consistent with prograde metamorphic growth (see Stowell et al., in press). Table 1. Representative garnet analyses from the Goldvllle district, Wedowee Group, and Emuckfaw Group, Tallapoosa County, Alabama Sample L02 Grain G2e6 Lowe mine L0-13 L0-15 L0-18 G1e G3e4 G1e L0-31d G4e 1 Tallapoosa mine TA-11 TA14 G2e G1e Wedowee Gp. Emuckfaw Gp. GG-101 GG-167 NS4a G2e G1e56 G2e Si02 Al 2o3 Ti02 Fe01 MgO MnO CaO total 37.03 20 .5 7 0.10 26.37 1.86 7.29 6.29 99.51 37.13 20.81 0.04 35.71 1.99 1.51 2.97 100.14 36.89 20.88 0.09 27.05 1.68 9.27 4.77 100.63 36 .93 20.67 0.06 35.07 1.93 2.45 2.68 99.77 36.99 21.18 0.01 36.58 1.60 1. 71 3.66 101.74 37.50 21.44 0.05 30.30 1.66 4.52 4.35 99.82 37.21 20.97 0.06 33.25 1.51 3.44 3.79 100.23 37.19 21.20 0.04 34.11 2.67 1.06 3.36 99.63 37.09 21.36 0.03 36.43 2.50 1.14 3.29 101.85 37.46 21.48 0.05 28.07 1.78 0.92 10.91 100.66 Cations based on 12 oxygens Si 2.996 3.003 2.973 AI 1.962 1.983 1.983 Ti Fe+2 0.006 0.002 0.005 1.784 2.415 1.823 Mg 0.224 0.239 0.202 Mn 0.499 0.104 0.633 Ca 0.545 0.257 0.412 total 8.017 8.004 8.030 3.002 1.980 0.003 2.384 0.233 0.169 0.233 8.005 2.963 2.000 0.000 2.451 0.192 0.116 0.315 8.036 3.014 2.031 0.003 2.037 0.199 0.308 0.375 7.967 3.005 1.995 0.004 2.245 0.182 0.235 0.328 7.994 2.996 2.013 0.002 2.299 0.321 0.072 0.290 7.994 2.954 2.005 0.002 2.426 0.297 0.077 0.281 8.042 2.966 2.004 0.003 1.859 0.210 0.062 0.926 8.029 Mg#2 Aim Pyr Sps Grs 0.112 0.584 0.074 0.164 0.179 0.090 0.801 0.079 0.034 0.085 0.100 0.594 0.066 0.206 0.134 0.089 0.790 0.077 0.056 0.077 0.073 0.798 0.062 0.038 0.102 0.089 0.698 0.068 0.105 0.128 0.075 0.751 0.061 0.079 0.110 0. 122 0.771 0.108 0.024 0.097 0.109 0.787 0.096 0.025 0.091 0 .101 0.608 0.069 0.020 0.303 1 total Fe calculated as FeO 2 Mg# = Mg/(Mg+Fe) The almandine content of garnets ranges from 0.58 to 0.80; compositions from both extremes may be found within 75 m at the Lowe mine. Despite large variations in almandine, spessartine, and grossular contents (Table 1), garnets within vein selvages and those within the host rocks 5 cannot be readily distinguished chemically, suggesting that bulk rock chemistry controlled garnet compositions, not externally-derived fluids. Phylloslllcates Phyllosi licates occur both as disseminated grains and segregations (<1 mm max. dimension) within the host rocks, and as coarse (1 -2 mm) aggregates in selvages adjacent to veins. Within the Wedowee Group, biotite-muscovite and chlorite-sericitebiotite define the S1 and S2 foliations, respectively. Biotite and muscovite are the dominant phyllosilicates defining the most prominent foliation in the Emuckfaw Group. Biotite is common within the district and throughout phyllites in the Wedowee Group. Small grains (<1 mm) are found disseminated in most samples and coarse-grained (1 -2 mm) selvages are found adjacent to many quartz veins. Biotite is not found in retrograde shear zones . Biotite compositions in vein selvages, in the matrix of host rocks, and in the Wedowee and Emucktaw Groups outside the Goldville district were compared (Rg. 2, Table 2) . All of the trioctahedral mica analyzed contain between 2.46 and 2.58 tetrahedral AI cations and 0.69 to 0.86 octahedral AI cations per 22 oxygen atoms (Fig. 2). These compositions are very close to AI annite-AI phlogopite solid so lution, but will be called biotite for convenience . The Mg# [Mg# = Mg/(Mg+Fe)] of these biotites varies between 0.37 and 0.51 . Biotite from the Lowe and Tallapoosa mines are compositionally indistinguishable from biotite in the host rocks. siderophyllite 2.0 1.5 octahedral AI 22 oxygens 10 0 .5 .-.-..~~ 0.0 0.0 0.2 annlte 0.4 0.6 Mg I (Mg +Fe) 0.8 1.0 phlogopite A 0 Goldville district Regional Figure 2. Biotite compositions from Table 2 plotted in terms of octahedral AI vs. Mg/(Mg+Fe). 6 Table 2. Representative biotite analyses from the Goldvllle district, Wedowee Group, and Emuckfaw Group, Tallapoosa County, Alabama Sample L0-2 Grain 82 Lowe mine L0-13 L0-15 L018 82 84 B1 L0-31 852 Tallapoosa mine TA-11 TA-14 891 81 WedoweeGp. GG101 821 Emuckfaw Gp. GG-167 NS-4a 821 B3 Si02 AI203 Ti02 FeQ1 MgO MnO CaO Na2o K20 Cl F H202 total 35.67 17.67 1.53 19.23 10.78 0.11 0.02 0.15 8.94 0.01 0 .0 0 3.90 97.99 35.45 18 .48 1.43 21 .59 8.99 0.06 0.02 0 .23 8.08 0.00 0 .25 3.76 98.21 35.50 17.86 1.87 19.76 9.56 0.13 0.02 0.11 8.60 0 .00 0 .10 3 .81 97.25 35.38 18.06 1.75 20.99 9 . 10 0.05 0.02 0.16 8.79 O.o1 0.00 3.88 9 8 . 17 35.46 19.11 1.50 23.58 7.83 0.07 0.00 0.00 8.99 0.03 0.00 3.92 100.49 36.41 18.46 1.84 21.61 8.64 0.11 0.01 0.08 8.37 0.00 0.09 3.90 99.48 34.97 18.22 1.47 22.73 8 .17 0.10 0.04 0.19 8.26 0.00 0.00 3.85 98.00 36.27 18.38 1.48 19.11 11.12 0.07 0.01 0 .24 7.91 0.00 0.25 3.83 98.56 35.73 18.62 1.64 20.36 9.03 0.00 0 .00 0 .18 8.98 0.00 0.00 3 .90 98.43 35.58 18.18 2.28 21 .40 8.42 0.13 0 .00 0 .13 9 .59 O.o1 0 .00 3 .93 99.62 Cations based on 11 oxygens Si 2.744 2.734 AI 1.603 1.680 Ti 0.089 0.083 Fe+2 1.237 1.392 Mg 1.236 1.034 Mn 0.007 0.004 Ca 0.001 0.002 Na 0.022 0.034 K 0.877 0.795 H 1.999 1.939 Cl 0.001 0.000 F 0.000 0.061 total 7.816 7.758 2.753 1.632 0 . 109 1.281 1.105 0.008 0 .002 0 .017 0.850 1.975 0 .000 0 .025 7.756 2.736 1.646 0.102 1.358 1.049 0.003 0 .002 0.024 0.867 1.999 0 .001 0.000 7.785 2.706 1.718 0 .086 1.505 0.890 0.004 0.000 0.000 0.876 1.996 0.004 0 .000 7.786 2.768 1.654 0.105 1.374 0.979 0.007 0.001 0.012 0.812 1.978 0.000 0.022 7.712 2.726 1.673 0.086 1.482 0.949 0.007 0.003 0.029 0.821 2.000 0.000 0.000 7.776 2.748 1.641 0.084 1.211 1.256 0.004 0.001 0.035 0.765 1.940 0.000 0.060 7.747 2.744 1.685 0.095 1.308 1.033 0 .000 0.000 0.026 0.880 2.000 0.000 0.000 7.771 2.727 1.642 0.131 1.372 0 .962 0 .009 0.000 0.020 0 .937 1.999 0.001 0.000 7.799 0.500 1.256 0.346 0.426 1.266 0.414 0.463 1.247 0 .385 0.436 1.264 0 .381 0.372 1.294 0.425 0.416 1.232 0.422 0.391 1.274 0.399 0.509 1.252 0.390 0 .441 1.256 0.429 0.412 1.273 0.369 1 total Fe calculated as FeO 2 H20 calculated assuming 2 OH, Cl, and/or F for every 10 oxygens 3 Mg# = Mg/(Mg+Fe) AI IV, AIVI =tetrahedral and octahedral aluminum, respectively White Mica White mica is abundant throughout the Wedowee Group in fine-grained (<0 .5 mm) foliated masses in strongly deformed samples, and coarse (<1 mm) randomly oriented intergrowths with biotite and chlorite in vein selvages within the district. In addition, coarse white mica is locally intergrown with large feldspar crystals in veins at the Lowe and Tallapoosa mines. White mica is 7 less abundant in the Emuckfaw Group; some amphibole-bearing schists only contain secondary sericite. White micas from both host rocks and veins are close to muscovite-paragonite mixtures and are 79 to 85% muscovite (Table 3). All of the white micas will be subsequently referred to as muscovite. Table 3. Representative muscovite analyses from the Goldvllle district, Wedowee Group, and Emuckfaw Group, Tallapoosa County, Alabama Sample L0-2 Grain M11 Lowe mine Tallapoosa mine Emuckfaw Gp. L0-13 L0-15 TA-14 GG-167 M31 M11 M2 M7 Si02 Al20 3 Ti02 Fe01 MgO MnO CaO SaO Na2o K20 Cl F H202 total 46.13 34.32 0.34 1.37 0.96 0.01 0.02 0.34 0.80 9.25 0.01 0.00 4.44 97.98 47.45 34.41 0.37 1.93 0.94 0.00 0.01 0 .27 1.11 9.31 0.00 0.00 4.53 100.33 45.99 33.44 0.40 1.59 0.99 0.01 0.10 0 .97 1.00 8.28 0.00 0.00 4.38 97.15 45.50 34.14 0.24 4. 13 1.58 0.00 0.04 0.33 1.26 7.37 0.00 0.00 4.46 99.03 48.03 35.68 0.54 1.17 0.75 0.00 0.00 0 .00 0.81 7.50 0.00 0.00 4.57 99.04 Cations based on 11 oxygens Si 3.119 3.138 3.145 AI 2.734 2.682 2.695 Ti 0.017 0.018 0.021 Fe+2 0.077 0.107 0.091 Mg 0.097 0.093 0.101 Mn 0.001 0.000 0.001 Ca 0.001 0.001 0.007 Ba 0.009 0.007 0.026 Na 0.105 0.142 0.133 K 0.798 0.785 0.722 Cl 0.001 0.000 0.000 F 0.000 0.000 0.000 total 6.958 6.973 6.941 3.062 2.707 0.012 0.232 0.158 0.000 0.003 0.009 0.164 0.633 0.000 0.000 6.980 3.152 2.760 0.027 0.064 O.o73 0 .000 0.000 0.000 0.102 0.628 0.000 0.000 6.807 Mg#3 AI IV AI VI 0.555 0.881 1.853 0.465 0.862 1.820 0.526 0.855 1.840 0.405 0.938 1.769 0.532 0.848 1.912 1 total Fe calculated as FeO 2 H20 calculated assuming 2 OH, Cl, and/or F for every 10 oxygens 3 Mg# = Mg/(Mg+Fe) AI'V, AIVI =tetrahedral and octahedral aluminum, respectively 8 Feldspar Feldspar occurs as small grains (<1 mm) disseminated within the host rocks of the district, medium-sized grains (1-2 mm) within vein selvages, and large grains (<10 mm) within veins in the district. Many of these large grains are partly replaced by sericite and/or calcite. Schists from the Emuckfaw Group contain numerous lenses and disseminated coarse grains {1 -2 mm) of feldspar. All of the analyzed grains are albite-rich plagioclase with most compositions ranging from An 16 to An35 (Table 4). Most plagioclase is zoned from core to rim; grains from host rocks are weakly zoned with higher anorthite content in rim compositions. Grains from samples containing quartz veins vary from strongly to weakly zoned with greater or lesser anorthite in rim compositions (Stowell et al., in press} . Table 4. Representative plagioclase analyses from the Goldvllle district, Wedowee Group, and Emuckfaw Group, Tallapoosa County, Alabama Lowe mine Tallapoosa mine Emuckfaw Gp. Sample L02 L0-13 L015 TA14 GG-167 Grain P1e P1 PS P4e P1 Si02 59.23 AI203 24.85 Fe2Q31 0.06 CaO 7.22 Na20 7.52 K20 0.12 total 98.99 63.40 23.20 0.04 4.12 9.41 0.06 100.24 61.13 24.12 0.20 5.27 8.62 0.06 99.42 62.80 22.17 0.14 3.20 9.64 0 .06 98.01 62.95 23.70 0.07 4.47 8.32 0.07 99.58 Cations based on 8 oxygens Si 2.669 2.795 2.730 AI 1.320 1.205 1.269 Fe+3 0.002 0.001 0.006 Ca 0.349 0.193 0.251 Na 0.657 0.803 0.746 K 0.007 0.002 0.002 total 5.002 5.000 5.004 2.827 1.176 0.005 0.154 0.840 0 .002 5 .003 2.785 1.236 0.002 0.212 0.714 0.004 4.953 An 0.345 0.194 0.251 0. 155 0.228 Ab 0.649 0.804 0.747 0.843 0.768 Or 0.006 0.002 0.002 0.002 0.004 1 total Fe calculated as Fe203 METAMORPHISM The textural arguments presented above and in Stowell and others (in press) for vein emplacement during metamorphism allow the direct use of metamorphic minerals to determine conditiQDS during mineralization. If the gold was locally derived from syngenetic deposits in the Wedowee Group as suggested by Paris (1986), Neal (1986), and Neal and Cook {in press), 9 pressure and temperature conditions determined in this manner could represent conditions during syndeformational remobilization of gold. Equilibrium Mineral Assemblages and Petrogenetic Grids Mineral assemblages in the Wedowee and Emuckfaw Groups (Table 5) are generally divariant In pressure-temperature space. Host rock and vein selvage minerals from the Goldville district, and mineral assemblages in the Wedowee Group indicate low to medium grade metamorphism: almandine-rich garnet+chlorite+muscovite zone. Although the assemblage garnet+chlorite+biotite is common, chlorite and biotite compositions vary appreciably (Stowell et al., in press). The variability of mineral chemistry in vein samples suggests a strong lithologic control on vein selvage chemistry. Staurolite occurs as a trace phase within a single vein selvage implying that metamorphism reached amphibolite facies or medium grade at the Lowe mine (Stowell et al., in press). The stability of quartz with muscovite in peak metamorphic assemblages in many of the samples from the Wedowee and Emuckfaw Groups requires that temperatures did not exceed those required for the reaction muscovite + quartz = potassium feldspar+ aluminum silicate+ vapor (Fig. 3). The Emuckfaw Group includes amphibolite facies mineral assemblages containing staurolite and kyanite (Maungnoicharoen , 1975). Sample NS-4a contains abundant hornblende compatible with amphibolite facies or medium grade metamorphism. Thermobarometry Metamorphic temperatures and pressures were estimated using experimentally- or empirically-calibrated elemental exchange equilibria . The calibrations and activity models used for thermobarometric calculations are discussed by Stowell and others (in press). Only coexisting grains exhibiting no textural evidence of disequi librium were selected for thermobarometric study . Mineral assemblages for samples used for thermobarometry are given in Table 5. All calculations are based on electron microprobe analyses of individual mineral grains. Most of the errors quoted below for temperature and pressure calculations are based on analytical and thermodynamic uncertainties. Errors in the relative differences between results based on one thermobarometer may be estimated from analytical uncertainty alone. These relative errors are ~50% less than absolute errors. Garnet-Biotite Thermometry Metamorphic temperatures were calculated for host rocks and vein selvages from the Lowe and Tallapoosa mines, ahd for the parts of the Wedowee and Emuckfaw Groups outside the district (Fig. 1) using analyzed mineral compositions and Fe-Mg exchange between garnet and biotite (Ferry and Spear, 1978; Ganguly and Saxena 1984, 1985). The uncertainty in temperature determinations based on this exchange is about 50C (Ferry and Spear. 1978). Drill core samples from the Lowe and Tallapoosa mines containing adjacent unaltered grains of garnet and biotite were selected for geothermometry. Samples were collected from directly adjacent to veins (L0-2, L0-13, and TA-14), some of which have anomalous gold values, and 5 m or more from veins (L0-18, L0-15, and TA-11). In addition, samples were selected from outside the district (GG-1 01, GG-167, NS-4a) to insure that a valid comparison was made between temperatures calculated for minerals adjacent to the mineralization and regional metamorphic conditions . 10 Table 5. Mineral assemblages of samples used for thermobarometry Equilibrium Assemblages Minor or trace phases LOWE AND TALLPOOSA MINE Veins L0-2: TA-14: 1) pl-qtz 2) car-qtz-sul 1) ms-pl-qtz-ap ser1, op1 ser1 Vein selvages adjacent to quartz and sulfide veins L0-2: 1) grt-bt-ms-pl-qtz ilm, po, py 2) grt-bt-st L0-13b: 1) grt-bt-chl-ms-pl-qtz ilm, py 2) grt-bt-ms-qtz Host rocks L0-15: 1) grt-bt-chl-ms-pl-qtz 2) grt-bt-cal-qtz L0-18 : 1) grt-bt-chl-ms-qtz L0-31c: 1) grt-bt-qtz L0-31 d: 1) grt-bt-qtz TA-11 : 1) grt-bt-ms-qtz TA-14: 1) grt-bt-chl-ms-pl-qtz ilm, ap ilm, gr. tur, ap tur, chl1, ms1, ilm chl1, ms1, ilm ap WEDOWEE GROUP GG-1 01: 1) grt-bt-chl-ms-qtz ilm EMUCKFAW GROUP GG-167: 1) grt-bt-ms-pl-qtz NS-4a: 1) grt-bt-hbl-ep-pl-qtz ms1, chJ1 1 secondary or retrograde phases Mineral symbols: ap =apatite, bt =biotite, car= carbonate, chi = chlorite, ep = epidote, gr =graphite, grt = garnet, hbl = hornblende, ilm =ilmenite, ms =muscovite, op =opaque, pi= plagioclase, po = pyrrhotite, py =pyrite, qtz =quartz, ser = sericite, st =staurolite, tur =tourmaline. Thermometric results range from 480 to 595C for all of the mineral pairs and average 49012C for the district (Table 6). Temperatures estimated from mineral chemistry are less than temperatures estimated from staurolite-biotite stability. This apparent discrepancy may result from PH 2o < Ptotal (see Stowell et al., in press) or from chemical impurities that stabilized staurolite at temperatures below those obtained in the experimental studies. Temperatures calculated using rim compositions of large, subhedral, zoned garnets and adjacent biotite in the host rocks do not differ significantly from temperatures calculated using rim compositions of small, euhedral, unzoned garnets and adjacent biotite in vein selvages. The single temperature estimate for the Wedowee Group is indistinguishable from these results. This suggests that the rims of all garnet and biotite grains in the district and throughout the Wedowee Group equilibrated at upper greenschist to lower amphibolite facies conditions. Thermometric results clearly indicate higher metamorphic temperatures for the Emuckfaw Group (Table 6). 11 10 9 8 7 6 p (kb) 5 4 3 2 1 0 100 Goldville district kyanite st chi/ I + + I I bt ms 1 andalusite sillimanite 200 300 400 500 600 700 800 900 T(C) Figure 3. Pressure vs. temperature plot for Goldville district, Wedowee Group, and Emuckfaw Group samples. The rhombic fields show the average pressure and temperature conditions and estimated errors (analytic and thermodynamic) calculated for vein selvage and host rock mineral chemiistry from the Goldville district, and for regional metamorphic minerals in the Emuckfaw Group. Muscovite + quartz stability after Cha1terjee and Johannes (1974); aluminum silicate phase diagram from Holdaway {1971); chlorite + muscovite = staurolite + biotite + quartz + vapor reaction field from Hoschek (1969) ; all reaction curves involving vapor are for PH2o = Ptotal Plagioclase-Biotite-Garnet-Muscovite Barometry Metamorphic pressures were calculated for vein selvage and host rock samples using analyzed mineral compositions and Fe-Mg-Ca-K exchange between plagioclase, biotite, muscovite, and garnet (Ghent and Stout, 1981; Hodges and Crowley, 1985). The uncertainty in pressure determinations based on this exchange is greater than 1 kb (Hodges and Crowley, 1985). Five samples were selected for geobarometry: two from selvages adjacent to quartzplagioclase-sulfide veins at the Lowe mine, one from quartz-rich phyllitic host rock at the Lowe mine, and one from a vein selvage at the Tallapoosa mine. Mineral compositions from the edges of garnet and plagioclase grains and the centers of phyllosilicate grains were used in the calculations. Barometric results range from 5.1 to 6.7 kb. No discernible difference was found between barometric results from vein selvages and host rock samples in the district. However, the highest 12 pressure was calculated for a sample from the Emuckfaw Group, suggesting that rocks within the Alexander City fault zone experienced higher pressures than the Wedowee Group. Table 6. Pressures (kb) and temperatures (0 C} calculated from elemental exchange equilibria. T-FS1 T-GS2 P-HC 4 LOWE AND TALLPOOSA MINE Veins TA-14 485 540 510 6.4 Vein selvages adjacent to quartz and sulfide veins L0-2 480 525 5.1 L0-13 515 550 6 .4 L0-31 485 539 Host Rocks L0-15 485 535 5.6 L0-18 495 530 TA-11 500 555 All Lowe and Tallapoosa mine samples AverageS 49012 54011 510 5.80.6 WEDOWEE GROUP GG-101 500 500 EMUCKFAW GROUP GG-167 555 565 NS-4a 595 615 6.7 Analytical and thermodynamic uncertainties yield errors ca. 50C (T-FS, T-GS), 80C (T-GU), and 1 kb (P-HC) 1 grt-bt temperature: Ferry and Spear (1978) with Hodges and Crowley (1985) garnet activity model 2 grt-bt temperature: Ganguly and Saxena (1984) 3 pl-ms temperature : Green and Usdansky (1986) 4 pl-bt-grt-ms pressure: Hodges and Crowley (1985) 5 1 standard deviation Plagioclase-Muscovite Thermometry The temperature of vein formation at the Tallapoosa mine was calculated using mineral compositions and Na-K exchange between feldspar and muscovite (Green and Usdansky, 1986). Errors for this calculation are 80C based on analytical uncertainty alone. The mineraJ grains used for this calculation are from within a vein (TA-14), and show no evidence of recrystallization, or disequilibrium. Therefore , this temperature should accurately reflect the conditions of vein formation . The sample contains intergrown coarse muscovite and plagioclase; mineral compositions from directly adjacent to their mutual grain boundaries were utilized to determine the temperature of equilibration. 13 Thermometric results are 51 080C, consistent with low- to medium-grade (upper greenschist to lower amphibolite facies) metamorphism during vein emplacement. DISCUSSION Timing of Deformation, Metamorphism, and Mineralization Mineral textures and structural relationships indicate that D1 through 0 3 structures result from progressive deformation during northwest-southeast oriented shortening (Stowell et al.. in press) . Garnet and biotite chemistry and textures indicate that peak metamorphic conditions postdate D1 and predate 0 3. These data and chlorite-muscovite zone assemblages in late shear zones are compatible with a single metamorphic event which peaked between D1 and 0 2 , and included retrograde mineral growth during 03 . Paris (1986), Neal (1986), and Neal and Cook (in press) considered the gold deposits to be strata-bound and interpreted the gold-su lfide mineralization as syngenetic with later remobilization of gold into quartz veins during metamorphism. Detailed mapping and petrography indicate that the deposits are transgressive to the stratigraphy (Guthrie and Dean. 1989) and that mineralization appears to postdate the initial stages of metamorphism and deformation (Stowell et al., in press). o2-D3 shear Some zones . mineralized veins cross-cut o1 fabrics , others are deformed Mineral assemblages suggest that shearing occurred during and and bordered by subsequent to peak metamorphism. Peak metamorphic shear zones contain stable biotite and garnet, whereas post metamorphic shear zones contain stable chlorite and muscovite. Preliminary data indicate that retrograde shear zones exhibit variable amounts of alteration (Lesher et al., in press) and gold enrichment (Stowell et al., in press). Stable isotopic data from veins, vein selvages, and an "altered" shear zone (L0-31) suggest that isotopically similar fluids were present during metamorphism, vein emplacement. and post-peak metamorphic shearing (Lesher et al., in press) . Physical Conditions During Vein Emplacement and Metamorphism Mineral assemblages, chemistry, and equilibrium relationships for veins, vein selvages, and 02 mineral fabrics indicate that peak metamorphism and vein emplacement occurred at about 500C at 5-6 kb. These results are similar to those determined by stable isotope exchange thermometry for plagioclase-quartz and biotite-quartz pairs in veins and vein se lvages at the Lowe mine (Lesher et al., in press). Textures and mineralogy in the Goldville samples suggest that metamorphism never surpassed upper low-grade or lower medium-grade conditions (greenschist or lower amphibolite facies) and that veln emplacement occurred during metamorphism. Temperatures and pressures estimated for metamorphism and vein emplacement in the Wedowee Group are significantly lower than temperatures and pressures estimated for metamorphism in the Emuckfaw Group. This suggests that the Alexander City fault zone marks an important syn- to post-metamorphic boundary. Thermobarometric results, the nature of garnet zonation (Stowell et al. , in press) . the paucity of minerals that define lower medium-grade metamorphism, and absence of minerals that define upper medium-grade metamorphism in the Goldville district indicate that regiona l metamorphic conditions have been overestimated by previous workers. The peak metamorphic mineral assemblages observed In the Wedowee Group during this study (low- to medium-grade or upper greenschist facies) are similar to previously proposed retrograde assemblages (greenschist facies : Tull, 1975: Steltenpoh l and Moore, 1988} and most peak metamorphic mineral assemblages in the district have not been obviously affected by retrograde metamorphism. Therefore, the importance of retrograde metamorphism appears to have been overemphasized for this part of the Piedmont. 14 Genesis of Gold Mineralization At present there is no unequivocal evidence for the ultimate source of gold in the Goldville district. However, our data, and the compositions and densities of fluids from primary inclusions in vein quartz (Sha and Lesher, in press) are compatible with gold emplacement in veins along shear zones which were active synchronous with the peak of low- to medium-grade metamorphism. The presence of vein mineralization in the Goldville district and it's apparent absence in the Alexander City fault zone, as defined, suggest that rocks within the Goldville district were more permeable. It is likely that the Alexander City fau lt zone comprises a wide zone of hetereogenously-sheared rocks exhibiting variable strain, structural style (folding, faulting, ductile shearing), and permeability. This deformation is considered to have accompanied thrusting of the Emuckfaw Group northwestward onto the Wedowee Group (Guthrie and Dean, 1989). Although no strain data are presently available, it is possible that the Goldville district represents a zone of fold-dominated deformation characterized by lower relative strain, dilation, and higher effective permeability, whereas the Alexander City fault zone represents a zone of shear-dominated deformation characterized by higher relative strain, constriction, and lower effective permeability. CONCLUSIONS Vein emplacement occurred during 0 1-02, synchronous with low- to medium-grade (garnet to staurolite zone) metamorphism. Thermobarometric studies, utilizing a variety of well-constrained elemental exchange equilibria, indicate that vein emplacement initiated during near-peak metamorphic conditions of 49050C and 5.81 kb in the Wedowee Group. Textures and mineral assemblages indicate that gold-bearing veins were emplaced within shear zones which were active during the peak of metamorphism. Late-stage shear zones adjacent to veins suggest that shearing during low-grade metamorphism (chlorite-muscovite zone) occurred subsequent to mineralization. Thermobarometric results from the Emuckfaw Group, adjacent to the Goldville district, suggest that metamorphism within the Alexander City fault zone occurred under higher grade conditions (57550C and 6.71 kb) than mineralization and metamorphism in the Wedowee Group. ACKNOWLEDGEMENTS U.S. Borax donated diamond drill core from the Goldville district to the Geological Survey of Alabama and Russell Lands Inc. provided access to mines and prospects in the Goldville district. Greg Guthrie provided two of the samples used for thermobarometry. Jim Saunders kindly reviewed the manuscript. This research was supported by The University of Alabama Research Grants Committee (Grant No. 1340 to C.M.L. and Grant No. 1370 to H.H .S.) and School of Mines and Energy Development (Grant No. 2-29115 to C.M.L. and H.H.S.). REFERENCES Adams, G.l., 1930, Gold deposits of Alabama, and occurrences of copper, pyrite, arsenic and tin: Alabama Geological Survey Bulletin, 40, 91p. Bence, A.E., and Albee, A.L., 1968, Empirical correction factors for electron microanalysis of silicates and oxides: Journal of Geology, v. 76, p. 37-67. 15 Chatterjee, N.D., and Johannes, W., 1974, Thermal stability and standard thermodynamic properties of synthetic 2M1-muscovite, KAI2[AISi30 10(0H)2]: Contributions to Mineralogy and Petrology, v. 48, p. 89-114. Drummond, M.S., 1986, Igneous, metamorphic, and structural history of the Alabama Tin Belt, Coosa County, Alabama: Florida State University, unpublished Ph.D. thesis, 411 p. Ferry, J.M., and Spear, F.S., 1978, Experimental calibration of the partitioning of Fe and Mg between biotite and garnet: Contributions to Mineralogy and Petrology, v. 66, p. 113-117. Ganguly, J., and Saxena, S.K., 1984, Mixing properties of aluminosilicate garnets: constraints from natural and experimental data, and applications to geothermo-barometry: American Mineralogist, v. 69, p. 88-97. Ganguly, J., and Saxena, S.K., 1985, Mixing properties of aluminosilicate garnets: constraints from natural and experimental data, and applications to geothermo-barometry: clarifications: American Mineralogist, v. 70, p. 1320. Ghent, E. D., and Stout, M.Z., 1981, Geobarometry and geothermometry of plagioclase-biotitegarnet-muscovite assemblages: Contributions to Mineralogy and Petrology, v. 76, p. 92-97. Green, N.L., and Usdansky, S.l., 1986, Toward a practical plagioclase-muscovite thermometer: American Mineralogist, v. 71, p. 1109-1117. Guthrie, G.M., and Dean, L.S., 1989, Geology of the New Site 7.5-minute quadrangle, Tallapoosa County, Alabama: Geological Survey of Alabama Quadrangle Series Map 9. Guthrie, G.M., and Lesher, C.M., in press, Geologic setting of lode gold deposits in the Northern Piedmont and Brevard Zone, Alabama, in Cook, R.B., Dean, L.S., and Lesher, C.M., eds.: Gold deposits of Alabama. Alabama Geological Survey Bulletin 136, p. 11-32. Hodges, K.V., and Crowley, P.D., 1985, Error estimation and empirical geobarometry for pelitic systems: American Mineralogist, v. 70, p. 70:~-709. Holdaway, M.J., 1971, Stability of andalusite and the aluminum silicate phase diagram: American Journal of Science, v. 271, p. 97-131. Hoschek, G., 1969, The stability of staurolite and chloritoid and their significance in metamorphism of pelitic rocks: Contributions to Mineralogy and Petrology, v. 22, p. 208-232. Lesher, C.M., Stowell, H.H., and Green, N.L., in press, Stable isotope geochemistry of lode gold mineralization in the Northern Piedmont, Alabama, in Cook, R.B. ed., Economic Mineral Deposits of the Southeast, Metallic Ore Deposits: Georgia Geological Survey Bulletin 117, this volume. Muangnoicharoen, N., 1975, The geology and structure of a portions of the northern Piedmont, east-central Alabama: University of Alabama, unpublished M.S. thesis, 74 p. Neal, W.L., 1986, Geochemical relationships between gold, silver, antimony, arsenic, and bismuth in gold mineralization of the Goldville District, Tallapoosa County, Alabama: Auburn University, unpublished M.S. thesis, 151 p. 16 Neal W.L., and Cook, in press, Mineralogy, petrology, and geochemistry of phyllite-hosted gold deposits, Goldville district, Tallapoosa County, Alabama, in Cook, A.B., Dean, L.S., and Lesher, C.M., eds.: Gold deposits of Alabama. Alabama Geological Survey Bulletin 136, p. 101-114. Neathery, T.L., and Reynolds, J.W., 1973, Stratigraphy and metamorphism of the Wedowee group: a reconnaissance: American Journal of Science, v. 273, p. 723-741. Pardee, J.T., and Park, C.F., Jr., 1948, Gold deposits of the southern Piedmont: U.S. Geological Survey Professional Paper 213, 156 p. Paris, T.A., 1986, The Goldville Project: Results of an exploration project for gold in the northern Alabama Piedmont, in Misra, K.C., ed., Volcanogenic sulfide and precious metal mineralization in the southern Appalachians: University of Tennessee, Department of Geological Sciences, Studies in Geology, v. 16, p. 182-205. Pouchou, J.L., and Pichoir, F., 1985, "PAP" <)>(pZ) procedure for improved quantitative microanalysis: in Armstrong, J.T., ed.: Microbeam Analysis-1985, p. 104-106. Sha, P., and Lesher, C.M., in press, Microthermometry of fluid inclusions in auriferous quartz veins, Goldville district, Northern Piedmont, Alabama, in Cook, A.B., ed., Economic Mineral Deposits of the Southeast, Metallic Ore Deposits: Georgia Geological Survey Bulletin 117, this volume. Steltenpohl, M.G., and Moore, W.B., 1988, Metamorphism in the Alabama Piedmont: Geological Survey of Alabama Circular 138, 27 p. Stowell, H.H., Guthrie, G.M., and Lesher, C.M, in press, Metamorphism and gold mineralization in the Wedowee Group, northern Piedmont, Alabama, In Cook, A.B., Dean, L.S., and Lesher, C.M., eds.: Gold deposits of Alabama. Alabama Geological Survey Bulletin 136, p. 133-158. Thomas, W.A., Tull, J.F., Bearce, D.N ., Russell, G., and Odom, A.L., 1980, Geologic synthesis of the southernmost Appalachians, Alabama and Georgia: in Wones, D.R., ed., Proceedings of the International Geological Correlation Program Caledonide Project 27, Department of Geological Sciences, Virginia Po lytechnic Institute and State University, Blacksburg, VA, Memoir no. 2, The Caledonides in the U.S.A., ed., p. 91-98. Tull, J.F., 1975, Structural and metamorphic features of the northern Alabama Piedmont, in Neathery, T.L., and Tull, J.F., eds., Geologic profiles of the northern Alabama Piedmont: Alabama Geological Society, 13th Annual Field Trip Guidebook, p. 63-86. Tull, J.F., 1978, Structural development of the northern Alabama Piedmont northwest of the Brevard zone: American Journal of Science, v. 278, p. 442-460. 17 MICROTHERMOMETRY OF FLUID INCLUSIONS IN AURIFEROUS QUARTZ VEINS, NORTHERN PIEDMONT, ALABAMA Peng Sha and C. Michael Lesher Department of Geology, University of Alabama, Tuscaloosa, Alabama 35487-0338 ABSTRACT Lode gold deposits in the northern Alabama Piedmont are associated exclusively with synkinematic, synmetamorphic quartz veins. Five compositional types of fluids are distinguishable in fluid inclusions in quartz from metapelite-hosted veins in the Goldville district and tonalitehosted veins at the Hog Mountain mine: type 1 very high salinity aqueous, type 2 high salinity C02-bearing aqueous, type 3 moderate salinity C02-bearing aqueous, type 4 moderate-low salinity aqueous, and type 5 C02-rich. Fluid inclusions in the Goldville district formed during four stages: stage I minor, isolated, primary, irregularly-shaped types 1, 2, and 3, stage II abundant, secondary, elliptically-shaped types 3 and 4 in planar arrays, stage Ill minor, secondary, irregularly-shaped type 4 in linear arrays, and stage IV abundant. secondary, irregularly-shaped type 5 In planar arrays. In contrast, fluid inclusions at the Hog Mountain mine formed during two stages, both secondary: stage I abundant, irregularly-, elliptically-, or tubularly-shaped types 3 and 4 in planar arrays, and stage II rare, irregu larly-shaped type 5 in isolation or in linear arrays. P-V-T-X properties of primary, high salinity (type 1 and 2) aqueous fluids in the Goldville district are not consistent with entrapment at pressures and temperatures determined by independent methods and appear to have been modified subsequent to entrapment. P-V-T-X properties of primary and early secondary, moderate salinity, low-C02 (type 3) aqueous fluids are consistent with entrapment during prograde metamorphism and appear to represent the vein -forming metamorphic fluid. Secondary, moderate-low salinity (type 4) aqueous fluids evolved toward higher salinity with decreasing temperature in the Goldville district and lower salinity with decreasing temperature at the Hog Mountain mine, probably reflecting decreasing degrees of interaction with pelitic and tonalitic host rocks, respectively. Late secondary C02-rich (type 5) fluids in both areas were most likely derived by oxidation of graphite in metasedimentary country rocks at low temperatures. The compositional and textural similarities of fluid inclusions in the Goldville district and at the Hog Mountain mine indicate a common metamorphic fluid source and timing of entrapment. Gold precipitation was probably Influenced by fluid-rock interactions. INTRODUCTION Lode gold deposits in the northern Alabama Piedmont are hosted by a variety of greenschistamphibolite facies pelitic metasedimentary, mafic metavolcanic, and felsic intrusive lithologies, but are associated exclusively with synkinematic, synmetamorphic quartz sulfide plagioclase carbonate veins . There is considerable controversy regarding the ultimate source of gold (syngenetic vs. epigenetic) in these deposits (cf. Guthrie and Lesher, in press; Neal and Cook, in press; Paris, 1986; Stowell et al. , in press; Lesher et at., this volume), but all workers agree that gold has been remobilized to some degree by vein-forming fluids. The aim of this paper is to define the composition and nature of the vein-forming fluids and to further constrain the physical and chemical conditions of gold mineralizatlon/remobilization in the southern Appalachians. Materials from three deposits in the northern Piedmont were studied: the Lowe and Tallapoosa mines in the northeastern and southwestern parts of the Goldville district and the Hog Mountain mine 10 km to the northwest (see fig. 1: Stowell and Lesher, this volume). All three deposits have been subjected to recent exploration programs involving diamond core drilling which provided fresh samples of vein quartz for detailed petrographic and microthermometrlc study. Three types of veins occur in these deposits (Lesher et al., this volume) : type I barren, deformed monomineralic quartz veins that occur in barren and mineralized lithologies throughout the region, type II mineralized quartz-sulfide veins (Hog Mountain) and quartz-plagioclase sulfide veins (Go ldvi lle) that are restricted to mine areas, and type Ill calcite sulfide quartz veins that cross-cut other veins. Only type II veins were sampled for this study. 18 Goldvllle District FLUID INCLUSION PETROGRAPHY Two mineral assemblages occur in mineralized (type II) veins in the Goldville district: 1) early, coarse-grained quartz plagioclase biotite garnet chlorite, similar to that in the metapelitic host rocks, and 2) late, fine -grained (<0.01 mm) recrystallized quartz sericite carbonate sulfide, generally distributed around early minerals . Free gold is extremely rare except in the weathered zone; sulfides are interpreted to house gold in solid solution or as submicroscopic inclusions. All quartz contains abundant fluid inclusions, both primary and secondary, which vary in size, shape, phase, and distribution (Fig.1 ). On the basis of cross-cutting relationships, they are grouped into four stages (Table 1). Table 1. Petrographic characteristics of fluid Inclusions In vein quartz, Goldvllle district and Hog Mountain mine, northern Alabama Piedmont Stage{%) Size m Shape Distribution Type{%) Phases Primary I {1 0) 2-35 Secondary II (60) 0.5-10 Ill (5) 0.5-6 irregular spherical, elliptical, tubular irregular GOLDVILLE DISTRICT isolated or grouped in centers of grains 3 (55) 2 (35) 1 (1 0) C02(vtH20(1)-co2(1) C02(v)"H20(1)-C02(1)salts(s) H20(v)-H20(I)"salts(s) planar arrays along margins of coarse grains linear arrays along margins of grains 4 (90) 3 (10) 4 H20(v)"H20(I) C02(v)-H20(I)C02(I) H 20(vtH~(I) IV {25) 15-40 irregular planar or linear 5 arrays in fractures cutting all gains C02(1)"CH4(1) Secondary I (90) 0.5-30 irregular, elliptical, tubular HOG MOUNTAIN MINE planar arrays in fractures 4 (95) 3 (5) H20(v)"H20(I) C02(vtH20(1tC02(1) II {1 0) 15-40 irregular isolated or linear 5 arrays in fractures C02(1)"CH4(I) Abbreviations: v =vapor, I = liquid, s =solid Bubble Volume % 15-60, mostly 20-30 10-15 1-5 0 5-15 0 Primary fluid inclusions, classified as stage I in this study, are generally large (5-30 J.lm) and have irregular shapes or negative crystal forms. At room temperature they contain one H20-rich or C02-rich gas phase + one H20-rich liquid phase one C02-rich liquid phase multiple solid phases. The daughter minerals are too small to identify (<0.5 llm); most are anisotropic, some have rectangular shapes (gypsum?) and some have rhombic shapes (carbonate?) . Vapor bubble volumes range 20-30%, but a few are 50-60%. Most primary fluid inclusions are isolated, but some occur in groups. All are distributed in early quartz, primarily in centers of grains. Along grain boundaries, crushing and recrystallization have overprinted the primary inclusions. Three stages of secondary fluid inclusions have been recognized: stage II ellipsoidal, tubular, or spherical liquid-gas inclusions occurring in planar arrays, stage Ill irregularly-shaped liquid-gas inclusions occurring in linear arrays (healed fractures), and stage IV irregularly-shaped liquid inclusions occurring in planar or linear arrays. Their petrographic relationships to stage I inclusions are illustrated in Figure 1. Stage II and Ill inclusions are much smaller (0.5-8 J.lffi) than stage I, but stage IV Inclusions are usually larger (20-40 J.J.m). At room temperature stage II and Ill 19 inclusions contain one H20-rich or C02-rich gas phase +one H20-rich liquid phase one C02rich liquid phase, whereas stage IV inclusions contain only a single C02-rich liquid phase. Volumes of vapor bubbles in secondary inclusions are less than those in primary inclusions, 1015% in stage II and 1-5% in stage Ill. Stage II and Ill secondary fluid inclusions occur along the margins of early stage quartz grains; stage IV secondary inclusions occur in arrays along fractures that cross-cut several quartz grains. 0 10 20 30~ !J Stage II (i) (jJ ~@ (} Figure 1. Sketch showing petrographic characteristics of different stages of fluid Inclusions In vein quanz from the Goldvllle district. H20-rich gas, H20-rich liquid, C02-rich liquid, and daughter minerals occur in various proportions in fluid inclusions at room temperature. On the basis of their relative abundance, five types of fluids have been identified: type 1 very high salinity aqueous fluids, type 2 high salinity C02-bearing fluids, type 3 moderate salinity C02-bearing aqueous fluids, type 4 moderate-low salinity aqueous fluids, and type 5 C02-rich fluids Their frequency distributions in the different stages of fluid inclusions are listed in Table 1. Hog Mountain Only secondary fluid Inclusions have been identified in mineralized (type II) quartz veins at the Hog Mountain mine. Two stages have been recognized; stage I ellipsoidal, tubular, or irregu lar liquid-gas inclusions occurring in planar arrays, and stage II irregular-shaped liquid inclusions occurring in isolation or in planar arrays. Their petrographic characteristics are summarized in Table 1. Stage I inclusions at the Hog Mountain mine are similar to stage II inclusions in the Goldville district, and stage II inclusions at the Hog Mountain mine are similar to stage IV inclusions in the Goldville district, except that they are less abundant. The apparent absence of primary fluid inclusions in vein quartz at the Hog Mountain mine may indicate that as a population they are too small to characterize with the optical microscope (1 0% of all detectable inclusions are <5J..1.m) or that they have been destroyed by later deformation (all veins exhibit evidence of post-emplacement deformation). 20 Methods FLUID INCLUSION MICROTHERMOMETRY Fluid inclusions were studied using a Fluid Inc.-modified U.S.G.S. gas-flow heating-freezing stage mounted on a Nikon Optiphot-Pol microscope fitted with 15x eyepieces and Leitz EF4x, EF1 Ox, and UT40x objectives. Freezing runs were carried out prior to homogenization runs. Inclusions were quickly frozen to ca. -180C and warmed slowly to observe the temperatures of phase transitions. Measurements included C02 melting temperature (Tmco 2). eutectic temperature (Te), ice melting temperature (TmiceL daughter mineral melting temperature (Tmdm), clathrate melting temperature (Tmclath), C02 homogenization temperature (Thco2), and total homogenization temperature (Th). The size of the inclusions often precluded determination of Te, Tmdm and Tmcrath; the presence of C02 often hampered determination of Tmice As the identities of the daughter mineral(s) in the fluid inclusions are not known and little experimental data are available for complex systems, fluid salinities were calculated as NaCI- or MgCI2-equivalent depending on eutectic temperature (Roedder, 1984). Salinities of low salinity H20-rich (type 3 and 4) fluids were calculated from measured Tmice using the NaCI solubility equation of Potter et al. (1978) or using an equation (wt .% MgCI2 = 0.2089 - 1.583 T- 5.044x1 o-2 T2 - 6.626x1 o-4 T3, r2 = 0.999) derived by regression of MgCI2 solubility data given by Linke (1958). Salinities of high salinity H20-rich (type 1 and 2) fluids were calculated from measured Tffidm using an equation (wt.% NaCI = 0.3825- 1.280x10-3 T + 5.585x1o-s r2 - 4.685x1o-9 r3, r2 = 0.999) derived by regression of NaCI solubility data given by Keevil (1942). Densities of the C02 phase were calculated from measured Thco2 using the P-T-density relation of pure C02 given by Angus et al. (1976). Densities of H20-rich liquids were calculated assuming unit density for pure H20 at room temperature for C02-rich inclusions (Burruss, 1981) or using the equation of Bodnar (1983) for C02-tree inclusions. Bulk densities and compositions of fluid inclusions were calculated using bubble volumes estimated at room temperature (Burruss, 1981 ). As Tmco2 varies narrowly around -58C, CH4 was ignored in most cases (Heyen et al., 1982), except for stage II inclusions at the Hog Mountain mine (Tmco2 =ca. -61 C). Measured Tfficlath data were not used to calculate fluid compositions (Collins, 1979), because the relative error in measurement (15%) is larger than for measurement of Tmco2 (2%) and Thc02 (4%). Microthermometric data for the Goldville district and Hog Mountain mine are summarized in Table 2. Fluid Compositions Calculated fluid compositions and densities are given in Table 3 and summarized in Table 4. Type 1 and 2 fluids in the Goldville district have anomalously high salinities, whereas stage I and 11type 3 fluids have moderate salinities and C02 contents. Stage 11-type 4 fluids have low to moderate salinities with negligible C02 and have been divided into two subtypes : type 4a fluids are characterized by moderate salinities (XNaCI(eq.) = 0.02-0.04) and a relatively narrow range of homogenization temperatures (Th = 225-119C), whereas type 4b fluids are characterized by low salinities (XNaCI(eq.) = 0.00-0.02) and a wider range of homogenization temperatures (Th = 238940C). Stage Ill-type 4 fluids are moderately saline with negligible C02, similar to stage If-type 4a fluids . Stage IV-type 5 fluids are dominated by C02; CH4 is only a minor component. Stage 1-type 3 and 4 fluids at the Hog Mountain mine (Table 3) are compositionally similar to stage 11 -type 3 and 4 fluids in the Goldville district, respectively, except for some with very low = eutectic temperatures (Te -27 to -35C) and low melting temperatures (Tmice = -14C to -34C) . Available experimental data (summarized by Roedder, 1984) suggest that MgCI2 is a necessary component in these fluids. On this basis stage 1-type 4 fluids have been divided into three subtypes: stage 1-type 4a{,i) moderate salinity H20-NaCI inclusions (XNaCI(eq.) = 0.04-0.07); stage 1-type 4a(ii) moderate salinity H20-MgCI2 inclusions (XMgct2(eq.) = 0.03-0.05) ; and stage !-type 4b low salinity NaCI-H20 inclusions (XNaCI(eq.) = 0.00-0.04). Stage If-type 5 fluids at the Hog Mountain mine are also dominated by C02, similar to stage IV-type 5 fluids in the Goldville district, but appear to contain more CH4 (Tmco2 =ca. -61C). 21 Table 2. Summary of mlcrothermometrlc data, Goldvllle district and Hog Mountain mine, northern Alabama Piedmont Stage-Type Tmco2 Primary 1- 1 I -2 -57.0 to -57.4 -57.2 (0.1) 6 1-3 -57.0 to -58.0 56.6 (0.3) 9 1-? -56.5 to -57.9 -57.2 (0.4) 23 Secondary 11-3 -57.3 -57.3 2 ll-4a ll-4b II-? -57.4 to -58.2 -57.7 (0.3) 5 111-4 IV-5 Secondary 1-3 -57.3 to -59.2 -58.3 (1 .3) 2 l-4a(i) 1-4a(ii) l-4b 11-5 -58.0 to -61.8 -60.8 (1.8) 4 ? Te Tmice Tmclath GOLDVILLE DISTRICT -20.8 to -21.3 -21.1 (0.3) 3 -20.4 to -20.6 -20.5 (0.1) 3 -3.9 to -8.5 -5.7(1 .6) 9 -4.3 to -4.9 -4.6 (0.4) 2 -4.8 to -7.9 -6.5 (0.7) 28 -0.1 to -3.9 -1.8 (0.9) 26 7.4-6.0 6.9 (0.8) 3 7.8-6.5 7.1 (0.5) 12 7.7 -7.8 to -10.4 -9.4 (0.8) 18 HOG MOUNTAIN MINE -27.1 to -34.8 -32.8 (2.7) 9 -19.7 to -21.3 -20.9 (0.8) 4 -8.6 to -19.3 -14.0 (7.6) 2 -7.1 to -15.8 -9.9 (1.9) 26 -13.6 to -34.4 -22.5 (4.7) 19 -0.1 to -7.1 -4.2(1.5) 33 Th002 Tllldm 25.4-16.3 21.7 (3.6) 6 28.3-15.0 23.0 (3.9) 9 27.5-3.6 18.2 (6.7) 23 29.7-28.5 29.1 (0.8) 2 300-294 297 (4) 2 198-183 192 (6) 6 28.1-19.6 24.9 (3.6) 5 24.6 to -3.5 16.1 (8.2) 12 10.2-9.4 9.8 (0.6) 2 7.9 to -13.2 -6.7 (10.0) 4 Th 345-326 336 (14) 2 353-255 300 (34) 6 401-242 300 (55) 7 370-189 303 (38) 44 278-269 274 (6) 2 225-119 158 (42) 17 238-94 198 (41) 23 345-126 244 (44) 21 105-76 89 (11) 5 271 333-92 253 (65) 21 356-104 201 (82) 16 251-114 165 (35) 27 376-72 237 (67) 65 Temperatures in oc; data given as range (line 1), mean and standard deviation (line 2), and number of measurements (line 3); ? =undifferentiated stage or type 22 Table 3. Summary of calculated densities and compositions of Inclusion fluids, Goldvllle district and Hog Mountain mine, northern Alabama Piedmont Stage Type Density XH2<) Xco2 XNaCI (eq.) GOLDVILLE DISTRICT Primary I 1 2 3 1.38-1.40 0.82-0.83 0.17-0.18 1.39 0 .83 nil 0.17 1.16-1.21 0.79-0.83 0.05-0.10 0.10-0.11 1.18(0.02) 0.82 (0.02) 0.07 (0.02) 0.11 (0.02) 0.93-1.07 0.81-0.93 0.04-0.17 0.02-0.04 1.00 .(0.07) 0.88 (0.05) 0.09 (0.05) 0.03 (0.34) Secondary II 3 0.93 0 .93 0.85-0.88 0.10-0.13 0.86 0.12 0.02 0.02 4a 0.89-1.09 0.96-0.98 1.01 (0.06) 0.97 (<0.01) nil 0.02-0.04 0.03 (0.1 0) 4b 0.83-1.05 0.88-1.00 0.00-0.02 0.90 (0.06) 0.99 (0.02) nil 0.01 (0.01) IU 4 1.06-1.12 0.95-0.96 0.04-0.05 1.09 (0.02) 0.96 (<0.01) nil 0.05 (0.07) IV 5 0.77-0.95 0.83 (0.05) nil 0.95 nil XMgCI2(eq.) XCH4 n nil nil 2 nil 0.001 6 nil 0.002 9 nil 0.002 2 nil nil 28 nil nil 26 nil nil 18 nil 0.05 12 HOG MOUNTAIN MINE Secondary I 3 1.00-1.07 1.04 0.77-0.78 0.16-0.19 0.03-0.07 0.77 0.18 0.05 nil 0.005 2 48(i) 0.77-1.07 0.93-0.96 0.04-0 .07 0.91 (0.08) 0.95 (0.01) nil 0.05 (0.01) nil nil 26 4a(ii) 0.95-0.97 n.d . 0.96 (0.01) nil 0.03-0.05 n.d . 0.04 (<0.01) nil 19 4b 0.79-1.00 0.96-1.00 0.00-0.04 0.95 (0.04) 0.98 (0.01) nil 0.02 (0.01) nil nil 33 II 5 0.88-0.99 0.96 {0.05l nil 0.78 nil nil 0.22 4 Data given as range (line 1) and mean and standard deviation (line 2); n =number of inclusions, n.d. = not determined 23 Homogenization Temperatures Th data for 127 stage I through Ill inclusions in the Goldville district are summarized in Figure 2; stage IV inclusions are homogenous at room temperature. Four hydrothermal events can be recognized with temperature maxima at: 350-230C, 290-185C, 170-110C, and 11 0-65C. These data record successive entrapment of stage 1-111 inclusions with declining temperature, consistent with the relative timing of entrapment based on cross-cutting relationships. 12 10 Goldvllle Stage I 8 6 4 2 0 10 Stage II 8 6 4 2 0 3 Stage Ill c::::J ll-4a I:!ZJ ll-4b mill II-? n 2 1 0 14 Hog Mountain Stage I 12 10 8 c::J l-4a EEl 1-4b - 1-? 6 4 2 0 50 100 150 200 250 300 350 400 Th (C) Figure 2. Histograms of homogenization temperatures (Th) of fluid Inclusions In vein quartz from the Goldvllle district and the Hog Mountain mine. Inclusion types as defined In text. Th data for 130 stage I inclusions at the Hog Mountain mine are summarized in Figure 2. Only two hydrothermal events can be recognized with temperature maxima at 340-215C and 2151200C. Th of stage Hype 4a(i) and stage 1-type 4a(ii) inclusions at the Hog Mountain mine are indistinguishable {Table 2) and span the entire range of Th for stage 11-type 4 inclusions in the Goldville district, but stage Hype 4b inclusions at the Hog Mountain mine exhibit a narrower range of lower Th, similar to that of stage IHype 4a inclusions in the Goldville district. 24 DISCUSSION Stage I fluids in the Goldville district are compositionally highly variable, ranging from rare to minor, highly saline fluids (types 1 and 2) to more abundant, moderately saline fluids (type 3). As the different fluids do not occur systematically in close proximity, they could not have formed by necking down and phase segregation of a single precursor fluid. They vary principally in salinity and C02 content, and should have been miscible at the inferred conditions of entrapment. Density isochores for type 1 and 2 fluids (Fig. 3) , calculated using the computer program FLINCOR (Brown, 1989) and the modified Redlich-Kwong state equation given by Bowers and Helgeson (1983), do not intersect the prograde metamorphic P-T field (49550C, 5.81 kb) defined by elemental exchange thermobarometry (Stowell and Lesher, this volume) and supported by oxygen isotope thermometry (Lesher et at, this volume), suggesting that type 1 and 2 fluids have been modified subsequent to entrapment. 10 1-3 8 two-phases -6 :0 .::tt. D.. 4 ' Peak metamorphic P-T field for Goldville district solvus for 1-2 2 ' one-phase .. . solvus for 1-3 0 100 200 300 400 500 600 700 800 Figure 3. P-T diagram showing fluid density lsochores for stage I, type 1, 2, and 3 fluids and estimated P-T conditions of vein emplacement for Goldvllle district (explanation In text). H20-C02 solvus calculated using experimental data of Sourlrajan and Kennedy (1965). The P-V-T-X properties of stage 1-type 3 fluids in the Goldville district are compatible with entrapment at the P-T conditions of prograde metamorphism (Fig. 3) and most likely represent the vein-forming fluid. Their compositions are similar to those of stage 11-type 3 fluids in the Goldville district and stage Hype 3 fluids at the Hog Mountain mine (Table 4), indicating links between primary and secondary fluids in the Goldville district and secondary flu ids at the Hog Mountain mine (Table 4) and suggesting derivation from a similar source. Structural, petrographic, and thermobarometric data (Stowell et al., in press) indicate that quartz veins in the Goldville district are synmetamorphic and synkinematic, and stable isotope data (Lesher et al., this volume) suggest that vein-forming fluids were of metamorphic origin. Slightly saline, low-C02, aqueous fluids, such as those of type 3 in these deposits, are typical of metamorphically-generated fluids (Roedder, 1984; Crawford and Hollister, 1986) and may have been derived from adjacent country rocks during metamorphism or from similar rocks deeper in the metamorphic pile. The latter source is more consistent with stable and radiogenic isotopic data (Lesher et al., this volume). 25 Table 4. Summary of velnformlng fluid compositions, Goldvllle district and Hog Mountain mine, northern Alabama Piedmont Type Composition GOLDVILLE Stage lh(OC) HOG MOUNTAIN Stage Th(C) aqueous, very high salinity 345-326 2 aqueous, high salinity, minor C02 353-255 3 aqueous, moderate salinity, minor C02 I 401-242 II 278-269 271 4 aqueous, moderate-low salinity 5 C02-rich, minor-significant CH4 n.d. "" not determined II 238-94 Ill 105-76 IV n.d. 356-92 II n.d. Secondary stage 11-type 4 fluids in the Goldville district and stage 1-type 4 fluids at the Hog Mountain mine have been subdivided into two subtypes, moderate salinity subtype 4a fluids and low salinity subtype 4b fluids. Both exhibit systematic increases in density with decreasing temperature (Figs. 4 and 5). Using the experimental data of Bodnar (1983), these variations can be modeled as cooling trends in the system H20-NaCI, suggesting that two compositionally distinct fluids were present over a wide temperature interval. As the two fluids should have been miscible at the inferred temperatures of entrapment, they could have not had an opportunity to mix, suggesting that the differences were generated locally. Mass balance calculations indicate that fluid-tonalite interactions at the Hog Mountain mine removed Rb, K, Ba, Fe, Co, W, Au, S, and OH from the fluid and added Si, Na, Sr, Ca, Eu, Mg, Mn, and Zn into the fluid (Lesher and Green, 1987), whereas analogous, albeit considerably less extensive, fluid-petite interactions at the Lowe mine removed Rb, K, Ba, Mo, Au, and OH from the fluid and added Si, Na, Sr, Ca, P, S, and C02 into the fluid (C.M.L., unpubl. data). Vein mineralogies indicate that Si, Fe, Au, and S (in quartz-pyrrhotite) precipitated at the Hog Mountain mine, whereas Si, AI, Na, Sr, and Ca (in quartzalbite) precipitated in the Goldville district. The net result is that fluid-wall rock interaction and vein crystallization should have increased fluid salinity at the Hog Mountain mine and decreased fluid salinity in the Goldville district. This is consistent with the somewhat higher salinities of type 3 and 4 fluids at the Hog Mountain mine compared to equivalents in the Goldville district (Table 3). Secondary fluids in the Goldville district appear to exhibit an overall general trend with decreasing temperature toward greater abundances of higher salinity subtypes (ll-4b ~ ll-4a ~ 1114: Fig. 4), whereas secondary fluids at the Hog Mountain mine exhibit an analogous trend toward greater abundances of /ower salinity subtypes (l-4a ~ l-4b: Fig. 5). Based on the alteration trends discussed above, these trends probably reflect decreasing degrees of fluid-rock interaction with declining temperature and evolution of fluid compositions towards that Of the precursor metamorphic fluid. Alteration studies at the Hog Mountain mine indicate that fluid-wall rock interactions probably induced gold precipitation (Lesher and Green, 1987). Wall rock alteration is much less pronounced in the Goldville district, but whole-rock geochemical and stable isotopic data define similar alteration trends for the Goldville district and the Hog Mountain mine (C.M.L., unpubl. data; Lesher et al., this volume) and the compositional variations of secondary fluids in both areas are consistent with modification by reaction with wall rocks. Gold precipitation in the Goldville district was probably also influenced by such interactions. 26 1.5 1.4 - 1.3 C') I E c(,), ->. =Uc) 1.2 1.1 Cl) Q 1.0 0.9 x>'xx \ 0.8 50 100 150 200 250 300 350 400 Th (C) 50 40 *! 30 ~ ~.5 20 10 0 50 100 150 200 250 300 350 400 1-1 1-2 1-3 b. 11-3 [] ll-4a <> ll-4b X 111-4 1-1 1-2 1-3 b. 11-3 [] ll-4a <> ll-4b X 111-4 Figure 4. Homogenization temperature (Th) vs. fluid density (top) and salinity (bottom) of fluid Inclusions In vein quartz from the Goldvllle district. Reference curves In top diagram are density evolution trends of fluids In the H20NaCI system with the Indicated salinities (calculation method In text). Salinity trend In bottom diagram reflects decreasing degrees of fluldwall rock Interaction with decreasing temperature. 27 1.1 1.0 C') I E --~ C) 0.9 >c;; c Cl) 0 0.8 A 8%NaCI [] A 1-3 <> l-4a [] l-4b 0.7 50 100 150 200 250 300 350 400 Th (0 C) 25 0 20 0 -#. i 15 ->- .~ a.5; 10 (fJ 5 0 50 100 150 200 250 300 350 400 6 1-3 <> l-4a(i) 0 l-4a(ii) [] l-4b Figure 5. Homogenization temperature (Th) vs. fluid density (top) and salinity (bottom) of fluid Inclusions In vein quartz from the Hog Mountain mine. Density evolution curves In top diagram as in Fig. 4. Salinity trends In bottom diagram reflect cooling (light stipple) and decreasing degrees of fluid-wall rock Interaction (heavy stipple). 28 The final hydrothermal event in the region was dominated by C02-rich fluids (stage IV in the Goldville district and stage II at the Hog Mountain mine). This cannot be attributed to fluid immiscibility in the H20-C02 system with declining temperature (Takenouchl and Kennedy, 1964). because there are no conjugate H20-rich fluid inclusions in the same stage. The Wedowee country rocks contain abundant graphite, which is more easily oxidized at lower temperatures (Roble et al., 1978) and C02 is considerably more soluble at lower (and higher} temperatures than at moderate (1 00-250C) temperatures (Ellis and Golding , 1963), so generation of C02-rich fluids can probably be attributed to oxidation of graphite at low temperatures. CONCLUSIONS 1) Synkinematic, synmetamorphic auriferous quartz veins at the Hog Mountain, Lowe, and Tallapoosa mines in the northern Alabama Piedmont contain numerous fluid inclusions that can be subdivided into five compositional types depending on salinities and C02 contents. Fluid Inclusions formed during four recognizable stages at the Lowe and Tallapoosa mines, one primary and three secondary, and two stages at the Hog Mountain mine, both secondary. No primary inclusions have been recognized at the Hog Mountain mine. 2) The P-V-T-X properties of rare high salinity "primary" fluids in the Goldville district are not compatible with entrapment at the temperatures and pressures (49550C and 5.81 kb) determined by independent methods and appear to have been modified subsequent to entrapment. 3) Primary and early secondary fluids in the Goldville district and early secondary fluids at the Hog Mountain mine are H20-rich with moderate salinities, minor C02, and only very minor CH4. Their P-V-T-X properties are compatible with entrapment at the temperatures and pressures of vein formation. Compositional similarities suggest that they were derived from similar sources, probably by devolatilization of underlying sedimentary lithologies during regional metamorphism. 4) Secondary fluids in the Goldville district and at the Hog Mountain mine are H20-rich with moderate-low salinities (predominantly NaCI, some MgCI2 at Hog Mountain) and negligible C02. Variations in salinity reflect variable degrees of interaction with wall rocks. Deposition of sulfides and gold in the quartz veins and alteration zones was probably influenced by fluid-wall rock interactions. 5) Secondary fluids exhibit overall general trends with decreasing temperature toward fluids with higher salinity in the Goldville district and toward fluids with lower salinity in the Hog Mountain district, reflecting decreasing degrees of fluid-wall rock interaction with declining temperature and evolution of the fluid toward that of the metamorphic precursor. 6) Late secondary fluids in the Goldville district and at the Hog Mountain mine are C02-rich with minor (Goldville) to significant (Hog Mountain) CH4. They were probably derived by oxidation of graphite in carbonaceous country rocks. ACKNOWLEDGMENTS We are very grateful to U.S. Borax and Newmont Exploration Ltd. for donating diamond drill core to the Department of Geology and th e Geological Survey of Alabama and for providing copies of unpublished company maps and reports. Harold Stowell kindly provided access to his fluid inclusion stage, helpful advice during data collection and reduction, and insightful comments on the manuscript. Reviews by A.J. Driscoll, Jr. and J.R. Craig resulted in several improvements in the presentation. This research was supported by the University of Alabama Research Grants Committee (grant nos. 1340 to C.M.L. and 1370 to H.H.S.) and School of Mines and Energy Development (grant nos. 2-29107 and 2-29115 to C.M.L.). 29 REFERENCES Angus, S., Armstrong, B., deReuck, K.M., Altunin, V.V., Gadetskii, O.G., Chapela, G.A., and Rowlinson, J.S., 1976, International Thermodynamic Tables of the Fluid State. Volume 3, Carbon Dioxide: Pergammon Press, Oxford, 385 p. Bodnar, R.J., 1983, A method of calculating fluid inclusion volumes based on vapor bubble diameters and P-V-T-X properties of inclusion fluids: Economic Geology, v. 78, p. 535-542. Bowers, T.S., and Helgeson, H.C., 1983, Calculation of the thermodynamic and geochemical consequences of nonideal mixing In the system H20-C02-NaCI on phase relations in geologic systems: Equation of state for H20- C0 2-NaCI fluids at high pressures and temperatures: Geochimica et Cosmochimica Acta, v. 47, p. 1247-1275 Brown, P.E., 1989, FLINCOR fluid inclusion recalculation program: WISCWARE, University of Wisconsin, Madison. Burruss, R.C., 1981, Analysis of fluid inclusions: Phase equilibria at constant volume: American Journal of Science, v. 281, p. 1104-1126. Collins, P.L.F., 1979, Gas hydrates in C02-bearing fluid inclusions and the use of freezing data for estimation of salinity: Economic Geology, v. 74, p. 1435-1444. Crawford, M.L., and Hollister, L.S., 1986, Metamorphic fluids : The evidence from fluid inclusions: in Walther, J.V., and Wood, B.J. (editors), Advances jn Physjcal Geochemistry, Volume 5, Fluid-Rock Interactions Qurjng Metamorphism, Springer-Verlag, New York, p. 1-35. oooc Ellis, A.J. and Golding, R.M., 1963, The solubility of carbon dioxide above 1 in water and in sodium chloride solutions: American Journal of Science, v. 261, p. 47-60. Guthrie, G.M., and Lesher, C.M., in press, Geologic setting of lode gold deposits in the northern piedmont and Brevard zone, Alabama: in Lesher, C.M., Cook, R.B., and Dean, L.S. (editors), Gold Deposits of Alabama, Geological Survey of Alabama Bulletin 136, p. 11 -32. Heyen, G., Ramboz, C., and Dubessy, J., 1982, Simulation des equilibres de dans le systeme C02-CH4 en dessous de 50C et de 100 bar: Application aux inclusions fluides: Comptes Rendus des Seances de f'Academie des Sciences (Serie II), v. 294, p. 203-206. Keevil, N.B., 1942, Vapor pressures of aqueous solutions at high temperatures: Journal of the American Chemical Society, v. 64, p. 841-850. Lesher, C.M., and Green, N.L., 1987, Alteration and gold mineralization, Hog Mountain pluton, northern Alabama Piedmont: in Drummond, M.S., Green, N.L., and Neilson, M.J. (editors), The Granites of Alabama, 24th Annual Field Trip Guidebook, Alabama Geological Society, p. 13 -22 . Lesher, C.M., Stowell, H.H., and Green, N.L., this volume, Stable isotope geochemistry of lode gold mineralization in the northern Piedmont, Alabama: in Cook, R.B. (editor), Economic Mineral Deposits of the Southern Appalachians: Metallic Ore Deposits: Georgia Geologic Survey Bulletin 117. Linke, W.F., 1958, Solubilities of Inorganic and Metal-Organic Compounds, v. 1, 4th ed., van Nostrand, Princeton, 1487 p. Neal, W.L., and Cook, R.B., in press, Mineralogy, petrology, and geochemistry of phyllite-hosted gold deposits, Goldville District, Tallapoosa county, Alabama: in Lesher, C.M., Cook, R.B., and 30 Dean, L.S. (editors), Gold Deposits of Alabama, Geological Survey of Alabama Bulletin 136, p. 173-184. Paris, T.A., 1986, The Goldville project: Results of an exploration project for gold in the northern Alabama Piedmont: in Misra, K.C. (editor), Volcanogenic Sulfide and Precious Metal Mineralization in the Southern Appalachians: University of Tennessee Department of Geological Sciences Studies in Geology 16, p. 185-205. Potter, R.W., Clynne, M.A., and Brown, D.L., 1978, Freezing point depression of aqueous sodium chloride solutions: Economic Geology.. v. 73, p. 284-285 Robie, R.A., Hemingway, B.S., and Fisher, J.S., 1978, Thermodynamic properties of Minerals and Related Substances at 298.15 K and 1 Bar (105 Pascals) Pressure and at Higher Temperatures: U.S. Geological Survey Bulletin 1452, 456 p. Roedder, E., 1984, Fluid Inclusions, Volume 12, Reviews in Mineralogy: Mineralogical Society of America, Washington, 644 p. Sourirajan, S. and Kennedy, G.C., 1965, The solubility of carbon dioxide in NaCI solutions at high temperatures and pressures : American Journal of Science, v. 263, p. 445-454. Stowell, H.H., and Lesher, C.M., this volume, Thermobarometry of synmetamorphic gold deposits, Go ldville district, northern Piedmont, Alabama: in Cook, R.B. (editor), Economic Mineral Deposits of the Southern Appalachians : MetaWc Ore Deposits: Georgia Geologic Survey Bulletin 117. Stowell, H.H., Guthrie, G.M., and Lesher, C.M., in press, Metamorphism and Gold mineralization in the Wedowee group, Goldville district, Northern Alabama Piedmont: In Lesher, C.M., Cook, R.B., and Dean, L.S. (editors), Gold Deposits of Alabama: Geological Survey of Alabama Bulletin 136, p. 133-158. Takenouchi, S. and Kennedy, G.C., 1964, The binary system H20-C02 at high temperatures and pressures: American Journal of Science, v. 262, p. 1055-1074. 31 STABLE ISOTOPE GEOCHEMISTRY OF LODE GOLD DEPOSITS, NORTHERN PIEDMONT, ALABAMA C. Michael Lesher, Harold H. Stowell, and Nathan L. Green Department of Geology, University of Alabama, Tuscaloosa, Alabama 35487-0338 ABSTRACT Synmetamorphic, synkinematic lode gold mineralization in the central part of the AshlandWedowee belt is hosted by Wedowee Group metapelites in the Goldville district and by a tonalite pluton intruded into Wedowee Group metapelites at Hog Mountain. Quartz-plagioclase pyrrhotite carbonate veins in the Goldville district were emplaced during progressive Di -D2 deformation and peak metamorphism at 490 sooc and 5.8 1kb (garnet-biotite-plagioclasemuscovite equilibria); most are bordered by thin chlorite/biotite garnet selvages and/or narrow AI-RbK-Ba-Mg-Fe-Cr-enriched chlorite-muscovite alteration envelopes. Quartz-pyrrhotite carbonate veins at Hog Mountain were emplaced at 475-275C (plagioclase-muscovite equilibria) during progressive D2-D3 deformation, following peak metamorphism at 460 50C and 4 .5 1kb (garnet-biotite-plagioclase-muscovite-chlorite equilibria); most exhibit well-defined Rb-K-BaFe-Co-W-S-Au-enriched quartz-muscovite-pyrrhotite alteration envelopes that have been dated at 294 15 Ma (K/Ar whole-rock) and 315 16 Ma (Rb-Sr whole-rock isochron; 87SrfBBsrinitial = 0.7061 0.0012) . H isotopic compositions of tonalites at the Hog Mountain mine vary with degree of alteration (least-altered tonalites = -78 to -81%o oD, moderately-altered tonalftes = -79 to -74%ooD, strongly- a(tered tona tites = -50 to -53%o oD , and sheared, altered tonalites =58 to -62%o oD), reflecting se lective incorporation of D into muscovite during alteration and re-equilibration during postalteration shearing ; 0 Isotopic compositions are variable, but nonsystematic (9 .7 to 12.0%o o1Bo, reflecting the larger reservoir for 0 in the rocks and offsetting fractionation between plagioclase, biotite, and muscovite; S isotopic compositions are also variable and nonsystematic (9.6 to 11.3%.. l)34s). probably reflecting mixing of internally-derived magmatic and externally-derived sedimentary S. Analyzed mineral separates yield 0 isotopic equilibration temperatures and fluid compositions of 475 45C (qtz-ms) and 335 40C (qtz-cal); o1 80fluid = 7 to 11%o; oDttuid = -10 to -31%o; o13Cfluid = ca. -6%o; and o34Sfluid = ca. 10%o. Discordant 0 and H isotopic fractionation between quartz, muscovite, and calcite in veins suggests that 0 in quartz and muscovite records the temperature and fluid composition during initial alteration, whereas H in muscovite and 0 in calcite record conditions at lower temperatures. H and 0 isotopic compositions of metapelites at.the Lowe and Jones mines in the Goldville district also vary with degree of alteration (least-altered metapelites = 12.7 to 12.9'Yoo o18o and -82 = to -95%o oD;sericitized-chloritized metapefites 11.4 to 11 .7%o o1 8 0 and -66 to -69%o oD) reflecting selective incorporation of D and exclusion of 0 by muscovite-chlorite during alteration; limited S isotopic data are constant at ca. 19%o o34s. Analyzed mineral separates yield 0 isotopic equilibration temperatures and fluid compositions of 550-340C (qtz-pl), 520 30C (qtz-bt), and 350 15C (qtz-chl); o180t1uid = 9 to 13%o; 8Dtluid = -46 to -51'Yoo, and o34Sfluid = ca. 19. The stable isotopic data, together with field, petrographic, microthermometric, and geochemical data, suggest that vein-forming fluids in both areas equilibrated to varying degrees with, and were probably derived from, pelitic metasedimentary rocks, but were not derived from adjacent country rocks. Mineralizing fluids were most likely derived from underlying thrust sheets and/or the underlying North American continental shelf. 32 INTRODUCTION Lode gold mineralization in the northern Piedmont of Alabama is associated with synmetamorphic, synkinematic quartz sulfide plagioclase carbonate veins in a variety of host rocks, predominantly metapelites with lesser mafic metavolcanic rocks and granitoids (Pardee and Park, 1948; Guthrie and Lesher, in press). Some workers have interpreted the deposits as syngenetic and suggested that gold was mobilized into veins during metamorphism (McConnell et al., 1986; Paris, 1986; Neal and Cook, in press), but detailed stratigraphic, structural, metamorphic, and geochemical studies (e.g. Lesher and Green, 1987; Green et al., 1988; Stowell et al., in press) indicate that most of the mineralization is epigenetic. The aim of this paper is to report the results of a stable isotopic study of the Lowe and Jones mines in the northeastern and southwestern parts of the Goldville district and the Hog Mountain mine 10 km to the northwest (see fig. 1: Stowell and Lesher, this volume). All three deposits have been subjected to recent exploration programs involving diamond core drilling which provided fresh samples of vein minerals, alteration selvages, and alteration envelopes for geochemical analysis. The isotopic data constrain the conditions of vein emplacement and the sources of veinforming fluids in this part of the southern Appalachians. Our research indicates that the that the veins were emplaced during prograde metamorphism, that vein minerals and alteration assemblages crystallized from the same fluids, and that veinforming fluids were derived externally from the zones of mineralization. These data confirm that the gold mineralization in this region is epigenetic. GEOLOGIC SETTING The Hog Mountain deposit and Goldville district occur in the central part of the AshlandWedowee belt of the northern Alabama Piedmont (Guthrie and Lesher, in press; see fig. 1: Stowell and Lesher, this volume). Deposits in the Goldville district are hosted by rnetapelitic phyllite and schist of the Cutnose Gneiss member of the Wedowee Group (Neathery and Reynolds, 1973), whereas the Hog Mountain deposit, 10 km to the northwest, is hosted by a tonalite pluton intruded into the same metasediments. lhe rocks in this region have experienced polyphase deformation (Guthrie and Dean, 1989) and upper greenschist to lower amphibolite facies metamorphism (Stowell and Lesher, this volume). Rve phases of deformation have been recognized in the area based on local cross-cutting relationships (Guthrie and Dean, 1989): D1 is characterized by poorly-preserved microscopic to mesoscopic, intrafolial isoclinal, similar folds (F1) and a pervasive axial planar schistosity (S1); D2 is manifested as NE-trending, NW-verging, microscopic to macroscopic, close to isoclinal, quasiflexural folds (F2}. coaxial with F1. a discontinuous domainal cleavage to dominant foliation (S2l that transposes S1. and NE-striking, ductile to semi-brittle shear zones with northwest-directed transport; D3 is characterized by NNE-trending , NW-verging, upright to inclined flexural slip folds (F3), a solution crenulation cleavage (S3), and semi-brittle fault zones; D4 is manifested as SEplunging, upright open folds (F4) and a localized crenulation or fracture cleavage (S4); and Ds is characterized by isoclinal to broad, inclined, plunging extensional orenulations (Ss) associated with dextral strike-slip faults. These deformational styles correlate in a general manner with regional structures defined by Tull (1978; cf. Guthrie and Lesher, in press). D1 through D3 structures are interpreted to result from progressive deformation during northwest-directed shortening. Structural and petrographic studies at Hog Mountain (Lesher and Green, 1987; Guthrie and Lesher, 1988) and Goldville (Stowell et al., in press) indicate that the mineralized veins were emplaced progressively during D1 -D3, synchronous with metamorphism. Petrographic and thermobarometric data indicate that metamorphism in the area peaked between 01 and 02 and was followed by retrograde metamorphism during 03 (Stowell and Lesher, this volume). Metamorphic grade appears to increase southeastward across the area: 460 50C and 4.5 1kb at the Hog Mountain mine (Green and Lesher, 1987), 490 50C and 5.8 1kb at the Lowe mine (Stowell et al., in press), and 575 50C and 6.7 1 kb within the Alexander City fault zone (Stowell and Lesher, this volume). 33 Hog Mountain Deposit The geology and gold mineralization at Hog Mountain have been described by Park (1935), Pardee and Park (1948), Green and Lesher (1987, 1988), and Lesher and Green (1987). Hog Mountain is underlain by a small, elongate (1500 m x 400 m), tonalite pluton intruded into interlayered garnet-biotite-chlorite-quartz phyllite, chlorite-sericite phyllite, and graphite phyllite about 20 km northeast of Alexander City, Alabama (see fig. 1 of Stowell and Lesher, this volume). The pluton transgresses So and S1 in the country rocks, contains inclusions of foliated (S1 ?) country rocks with prograde mineral assemblages, and is cross-cut by quartz-sulfide veins with foliated alteration envelopes that are deformed, boudinaged, and transposed into the S2 fabric along the margins of the pluton. As such, the pluton is interpreted to have been emplaced during prograde metamorphism; after D1, but prior to or during D2. Owing to a narrow range of Rb/Sr ratios, the emplacement age of the Hog Mountain tonalite has been precisely determined, however, vein-related alteration is dated at 294 15 Ma (K/Ar) and 315 16 Ma (Rb/Sr whole-rock isochron, B7sr;86Sqnitial = 0.7061 0.0012: N.L.G., C.M.L., and A.K. Sinha, unpubl. data) . Three types of veins have been identified at Hog Mountain: I) deformed, milky white quartz veins with small (1-5 mm) lenses of transparent, clear quartz, II) massive, vitreous, grey quartzsulfide veins, and Ill) calcite-sulfide quartz veins. Silica mobility during wall rock alteration has also produced abundant diffuse quartz veins and silicified stockworks in the tonalite. Type I veins are restricted to the penetratively-deformed metasedimentary country rocks, whereas types II and Ill are restricted to the less-deformed tonalite. Type Ill veins are paragenetically later and less abundant than type II veins, but some veins are composite and multi-generational. Type 11-111 veins are bordered by muscovite-quartz-pyrrhotite alteration envelopes that are systematically enriched in Rb-K-Ba, Fe-Co, W, S, and Au relative to least-altered tonalite (Lesher and Green, 1987). Alteration envelopes range from massive through foliated to sheared, indicating that variable degrees of strain existed during alteration. Some veins are transgressive to the foliation in alteration envelopes, indicating emplacement into pre-existing fractures; others exhibit deformed fabrics, reflecting deformation subsequent to emplacement. Progressive emplacement/deformation is supported by plagioclase-muscovite Na-K exchange thermometry (Lesher and Green, 1987 and unpubl. data} which yields a broad range of temperatures between 425 50 (most type II veins) and 325 50C (most type Ill veins). Primary fluid inclusions have not been identified in vein quartz at Hog Mountain, but sparse early secondary inclusions with fluid compositions similar to those in some primary and early secondary inclusions in the Goldville district are H20-rich and moderately saline with minor C02 and very minor CH4 contents (Sha and Lesher, this volume). Most secondary fluid inclusions are H20-rich and slightly-moderately saline with minor C02 contents; anomalously low eutectic temperatures suggest that multiple salts (NaCI MgCI CaCI KCI) are present in some fluids. Late secondary fluid inclusions are C02-rich with significant CH4 contents. Goldvllle District The Goldville district comprises a 16 km long, northeast-trending group of inactive gold mines and prospects in the central part of the Ashland-Wedowee belt that have been described by Pardee and Park (1948), Neal (1986), Paris (1986), Guthrie and Dean (1989), Neal and Cook (in press), Neal et al. (in press), and Stowell et al. (in press) . Host rocks are garnet-biotite-chloritesericite-quartz phyllite in the northeast part of the district and graphite phyllite in the southwest part. The phyllites are polydeformed, have been metamorphosed to garnet zone, and sedimentary structures such as thin bedding, cross bedding, and graded bedding are rarely preserved. Gold mines and prospects in the district are associated exclusively with multi- generation quartz plagioclase pyrrhotite veins that cross-cut regional stratigraphy, compositional layering, and the earliest recognizable foliation (S1). Veins are variably boudinaged, folded, sheared, and transposed into an S2 axial planar fabric. Three types of veins have been identified in the Goldville district: I) monomineralic quartz veins, II) quartz-plagioclase carbonate sulfide veins, and Ill) calcite quartz sulfide veins. 34 Types I and II may be genetically related, as they exhibit similar structural relationships and a continuum of feldspar/quartz ratios. Type I veins are common at the Lowe and Tallapoosa mines and in the host rocks throughout the district. Type II veins appear to be restricted to the mine areas, although their absence in surface exposures could be due to weathering of plagioclase, hindering recognition. Type Ill veins are relatively rare and cross-cut type 1-11 veins and S1-S2 fabrics. Synmetamorphic vein emplacement is confirmed by thermobarometric studies {Stowell et at., in press; Stowell and Lesher, this volume). Fe-Mg exchange between coexisting biotite and garnet from host rocks and biotite-rich selvages adjacent to veins yield indistinguishable equilibration temperatures of 490 50C; Na-K exchange between coexisting muscovite and plagioclase In veins indicate similar, albeit less precise, equilibration temperatures of 510 80C; and plagioclase-biotite-muscovite-garnet equilibria indicate pressures of 5.8 1 kb. Some phyllites adjacent to veins exhibit little or no signs of alteration, however, many are bordered by narrow (<1 em) selvages of massive chlorite and/or biotite, and some are bordered by narrow (<1 0 em) zones of chlorite-muscovite garnet that locally truncate layering and veins, contain helicitic garnet porphyroblasts which are locally altered to chlorite, and commonly contain significantly smaller grain sizes than host rocks and adjacent vein selvages. These zones are enriched in Rb-K-Ba, Mg-Fe, and AI-Cr relative to least-altered metapelites (C.M.L. and H.H.S., unpubl. data) and appear to represent hydrothermally-altered shear zones. Primary and early secondary fluid inclusions at the Lowe and Tallapoosa mines in the Goldville district are H20-rich and moderately saline with minor C02 and very minor CH4 contents (Sha and Lesher, this volume). Most secondary fluid inclusions, equivalent to most of those at Hog Mountain, are H20-rich and slightly-moderately saline with minor C02 contents. Late secondary fluid inclusions are C02-rich with minor CH4 contents. STABLE ISOTOPE GEOCHEMISTRY Sampling and Analytical Methods All samples for stable isotopic study were taken from diamond drill core beneath the zone of weathering. The suite from Hog Mountain includes whole-rock samples of least-altered tonalite; two profiles across alteration envelopes adjacent to type II veins, one moderately-altered and ohe strongly altered; and several vein mineral separates from type II {qtz-po) and type Ill veins (qtz-poms-cal). The suite from the Goldville district {Lowe and Jones mines) includes two whole-rock samples of least-altered metapelite and adjacent alteration zones; and several vein mineral separates from type II veins (qtz-pl, qtz-bt, and qtz-chl). Whole-rock samples were crushed in an alumina ceramic jaw crusher and pulverized in an alumina ceramic shatterbox. Mineral separates were hand-picked under a binocular microscope {>99% pure), cleaned ultrasonically in reagentgrade acetone, and pulverized with an agate mortar and pestle. Stable isotope ratios were determined by mass spectrometry in the Department of Geology at Indiana University under the supervision of Dr. E.M. Ripley, using a 9 em Finnigan-MAT Delta-E spectrometer for H, C, and 0, and a 6" 60-sector Nuclide spectrometer for S. H was llberated as H20 and converted to H2 with Zn (Kendall and Coplen, 1982), C was liberated as C02 with phosphoric acid (Rosenbaum and Sheppard, 1986), 0 was converted to C02 using BrFs (Clayton and Mayeda, 1963), and S was combusted with excess CuO to produce S02 (Fritz et at., 1974). H and 0 data are reported as per mil deviations relative to standard mean ocean water (SMOW), C data relative to Pedee Formation belemnite (PDB{' and S data relative to Canon Diablo troilite (COT). Routine analytical precision is 0.05%o for~ 3c and ~1Bo, 0.1%o for ~34s, and 1%o for ~D (E.M. Ripley, pers.l comm., 1988). Results Stable isotopic data for Hog Mountain and Goldville samples are given in Tables 1 and 2, are illustrated in Figures 1-7, and are summarized below. Corresponding whole-rock major and trace element geochemical data for Hog Mountain are given in Green and Lesher (1988). 35 Table 1. Stable Isotope data for least-altered tonalltes, altered tonalltes, and type 11111 vein mineral separates from the Hog Mountain mine, Tallapoosa County, Alabama Sample ~0 ~180 ~34s ~13c (o/oo) (o/oo) (o/oo) ("/oo) Least-Altered Tonalltes HM-2/85.0 I HM-3/126.0 I HM-7/281.6 I -81 11.2 10.9 -78 10.9 11.3 -78 9.7 10.2 Altered Tonalltes Adjacent to Type II Vein HM-1/234.81 HM-1/235.1 m HM-11235.3 s HM-1/235.7 s HM-11235.9 s HM-1/23S.1 SS HM-1/236.4 s-s HM-1/236.8 s-s HM-11237.0 qv -75 11.2 10.3 -69 10.1 10.3 -50 10.4 10.5 -52 10.2 9.6 -53 9.8 10.3 -62 10.7 10.5 -61 1 0.8 9.6 -58 11.0 10.3 -88 11.4 9.8 Altered Tonalltes Adjacent to Type II Vein HM-2/187.8 I HM-2/188.1 I HM-2/188.5 s HM-2/188.6 s -77 11.4 11.2 -82 11.2 10.7 -74 11.2 10.7 -75 11.9 10.3 HM-21188. 7 m -79 12.0 9.9 Type II Vein Mineral Separates HM-11235 Muscovite -49* HM-1/274 Quartz Pyrrhotite 13.4 9.9 4255oPM -238 9.91.1 ca. 10 Composite Type 11-111 Vein Mineral Separates HM-61235 Quartz Muscovite Calcite Pyrrhotite 13.4 11.1 -63 11.8 47545QM 10. 70. 7 3355oPM -229 -8.21 3354oOC 9.0 7.61.2 HM-2/122 Arsenopyrite -4.1 ca. 9 -5.91.3 Notes: analytical methods, conventions, and calcula1ion methods given in text; errors based on analytical uncertainties; alteration intensity (as defined by Lesher and Green, 1987): I = incipient, I= slight, m = moderate. s - strong. s-s .. strong with shearing, qv =quartz vein; *oD value for muscovite in HM1/235 calculated by extrapolation of who le-rock oD -K20 data (Fig. 1); geothermometers: PM = pl-ms Na-K exchange, QM = qtz-ms 0 Isotope exchange, QC = qtz-cal 0 isotope exchange;. 36 Table 2. Stable Isotope data for metapelltes, alteration zones, and type II vein mineral separates from the Lowe mine and Jones vein, Tallapoosa County, Alabama Sample Type ~D ~180 ~34s (o/oo) ("/oo) ("/oo) Temp. (OC) ~180tluid ~Dtluid ~34Sfluid (o/oo) ("/oo) ("/oo) Metapelltes L0-23C L0-23A L0-31A L0-318 L0-31C country rock alteration zone country rock alteration zone alteration zone -82 12.7 -69 11.7 -95 12.9 18.6 -79 13.2 18.8 -66 11.4 ca. 19 ca. 19 Type II Vein Mineral Separates J0-5 quartz 15.2 plagioclase 13.4 L0-26 quartz biotite 14.5 -85 9.9 L0-30 quartz 15.2 plagioclase 13.9 L0-31 quartz 14.5 plagioclase 13.0 L0-32 quartz chlorite 14.7 -78 8.7 38545 52030 47545 43540 35015 10.90.7 12.40.2 12.50.7 -482 11.20.7 9.40.4 -501 Notes: analytical methods, conventions, and calculation methods given in text; errors based on analytical uncertainties; LO .. Lowe mine, JO =Jones mine. Three least-altered tonalites at Hog Mountain with 1.7-1.9% K20 and 0.05-0.93% S range between 9.7 and 11.2'Yoo a18Q, between -78 and -81'Yoo oD, and between 10.2 and 11.3'Yoo a34s. These ranges exceed analytical uncertainties and reflect a combination of original compositional inhomogeneity in the pluton (66-72% Si02, 0.2-0.6% Ti02, 16-18% AI203) and pervasive incipient alteration (Green and Lesher, 1987; Lesher and Green, 1987). H isotopic compositions of altered rocks vary systematically with degree of alteration and deformation, whereas 0 and S are more uniform. Four mode;ately-sericitized tonalites with 2.0-3.9% K20 and 0.24-0.69% S range between 11.2 and 12.0'Yoo a18Q, between -79 and -74'Yoo oD, and between 9.9 and 11.2'Yoo a34S; three decussate-textured, strongly sericitized-sulfidated tonalites with 6.3-6.8% K20 and 1.8- 3.0% S range between 9.8 and 10.4'Yoo a18o, between -50 and -53%o oD, and between 9.6 and 10.5'Yoo a34s; and two strongly foliated, sericitized-sulfidated tonalites with 5.9-8.8% K20 and 2.2- 3.0% S range between 10.7 and 11.0'Yoo a18o, between -58 and -62'Yoo aD, and between 10.3 and 10.5'Yoo a34s. An impure quartz vein with 87% Si02 contains 11.4'Yoo a18Q, -88%o oD, and 9.8'Yoo o34s. Linear regression of samples from a single vein alteration envelope (Fig. 1), confirms that muscovite is the principal host for D and K, and yields a oD value of -49'Yoo at 7.1% K20 (the average composition of muscovite determined by electron flrobe microanalysis: Green and Lesher, 1988). Mineral separates from a type II vein yield o Boftlz = 13.4 and 34spo = 9.9. Mineral separates from a type Ill vein and alteration selvage yield 1 Oqtz = 13.4, a180ms = 11 .1, o180cal = 11.77, oDms = -63, o13Ccal = -8.21, and o34Spo =9.0. 37 -40 -50 -- -60 0 ~ Cl c-o -70 <50 =- 49 @ K20 =7.1 ._.....o -80 -90 0 1 2 3 <50 = 5.23 K20- 86, r2 =0.956 6 7 8 Figure 1. Plot of SD vs. K20 for altered tonalltes In HM-1/234-237m at Hog Mountain; regression excludes sample 236.1 (Table 1). H Isotopic composition of average muscovite with 7.1% K20 (determined by electron probe microanalysis) Is -49%o SD. Two least-altered metapelites at Goldville exhibit 12.7 and 12.9%o s1BQ, -82 and -95%o SD and 18.6'Yoo S34s . H and 0 isotopic compositions of altered rocks vary systematically with degree of alteration, whereas S appears to be more uniform. Three altered metapelites range between 11.4 and 13.4'Yoo s1Bo and between -79 and -66%o SD, and exhibit 18.8%o s34s . Mineral separates yield S180qtz = 14.5-15.2%o, o1 80pl =13.0-13.9%o, o180bt =9.9'Yoo, and o180chl = 8.7'Yoo. Stable Isotope Geothermometry Samples utilized for stable isotope thermometry comprise coexisting mineral pairs which are inferred to be in equilibrium, however, different pairs have been sampled from different veins and are not necessarily cogenetic. Plagioclase and quartz are intimately intergrown within the veins at Goldville and are most likely to be in equilibrium. Coarse-grained muscovite, biotite, and chlorite are uncommon within quartz veins at Goldville or Hog Mountain and were separated (with difficulty) from vein selvages; elemental exchange thermometry suggests that muscovite/biotite selvages and veins equilibrated at the same temperatures (Stowell et al., in press; Stowell and Lesher, this volume). Local replacement of biotite by chlorite in other samples suggests that some chlorite postdates other minerals. Temperatures of equilibration may be calculated using coexisting minerals and experimentally-determined or theoretically/empirically-derived 0 isotopic exchange equilibria. There are substantial discrepancies between experimentally- and theoretically- or empiricallydetermined fractionation factors (see reviews by O'Neil, 1986, and Kyser, 1987). Experimentallydetermined values were used wherever possible; they are considered to be more applicable and appear to give results more consistent with our elemental exchange thermometry. 38 13 12 11 10 -0 9 oP.~ ........ 8 ...... r..o 7 6 5 4 3 2 I~-ms ~ ~~~ .? . , - HM6/235 qtz ...........-~-- HM6/235 ms / . , .,- 6~ _,_ HM6/235 cal ~--.... ~ ~~"" - - - - - qtz-ms - - - - - - - - - - grt-bt - - - - - - - - - - - - pl-ms pl-ms (II) - - - - - pl-ms (Ill) qtz-cal 188$81li!818181S8881l8Eil118811uae;;lll)l;::;mH 200 300 400 500 600 700 Figure 2. Calculated 51so values of aqueous fluids In equilibrium with quartz (short dashed line), muscovite (solid line), and calcite (long dashed line) mineral separates from the Hog Mountain mine (Table 1). Also shown are temperature ranges calculated from quartz-muscovite and quartz-calcite 0 Isotope fractionation, garnet-biotite Fe-Mg exchange thermometry, and plagioclase-muscovite Na-K exchange thermometry for stage II (qtz-pocal) and stage Ill (cal-poqtz) veins. 0 isotopic fractionation factors are taken from Matsuhisa et al. (1979: quartz-water and plagioclase-water), O'Neil and Taylor (1969: muscovite-water), Bottinga and Javoy (1975: biotitewater), Wenner and Taylor (1971: chlorite-water), and O'Neil et al. (1969: calcite-water); H isotopic fractionation factors are taken from Suzuoki and Epstein (1976: muscovite-water and biotitewater), Graham et al. (1987: chlorite-water); and C isotopic fractionation factors are taken from Bottinga (1969: calcite-C02 and graphite-C02). Simultaneous solutions of mineral-water equations for coexisting phases yield temperatures of equilibration and isotopic compositions of aqueous fluids (Tables 1 and 2; Figs. 2 and 5). Based on the reduced sulfide assemblages (po cpy sph mo), calculated low oxidation states of fluid inclusions (P. Sha and C.M.L., unpubl. data), and relatively high temperatures of equilibration calculated below, the oxidation state of the hydrothermal fluid is assumed to have been below the S02/H2S boundary and isotopic fractionation between H2S-s2 or Hs--s-2 is inferred to have been negligible (see Ohmoto, 1986). Temperatures of equilibration of veins and selvages at Hog Mountain (Table 1; Fig. 2) are 475 45C calculated from quartz-muscovite 0 isotope fractionation and 335 40C calculated from quartz-calcite 0 isotope fractionation. These temperatures correspond closely to the 425 50C and 325 50C values calculated by Green and Lesher (1988) for early (type II qtz-pocal) and late (type Ill cal-poqtz) stage veins using plagioclase-muscovite Na-K exchange thermometry. 39 0 -10 -20 -0 -30 .~ ......... 0r..o -40 -50 -60 -70 200 I .... ...... 1111 qtz-ms ' 1111 ' - . ... " . ---- pql-tmz-sca(Illl) HM1/235 ms (calc) ._. HM6/235 ms - - - - - pl-ms (II) ----------------------- pl-ms - - - - - - - - - - - grt-bt qtz-ms 300 600 700 Figure 3. Calculated oD values of aqueous fluids In equilibrium with early muscovite (solid line: composition calculated In Fig. 1) and late muscovite (dashed line: mineral separate) from the Hog Mountain mine. Temperature ranges as In Figure 2. Temperatures of equilibration of veins and selvages at Goldville (Table 2; Fig. 5) are 5503400C calculated from quartz-plagioclase 0 isotope fractionation (3 pairs; low precision attributable to similar slopes of quartz and plagioclase fractionation curves: Fig. 5), 520 30C calculated from quartz-biotite 0 isotope fractionation, and 35015C calculated from quartz-chlorite 0 isotope fractionation . Isotopic re-equilibration by self-diffusion is not considered to have affected the results, as diffusion calculations (Cole and Ohrnoto, 1986) suggest that re-equilibration would be insignificant for plagioclase, mica, and quartz. Unfortunately, no diffusion data are available for chlorite. Fluid Compositions The isotopic composition of the early (stage II) vein-forming fluid at the Hog Mountain deposit o18o (Table 1; Figs. 2-4) is 9.91.1%o calculated from quartz-water 0 isotope fractionation at the o18o plagioclase-muscovite Na-K exchange temperature, 10.70.7"A.o calculated from quartz- muscovite 0 isotope fractionation, -238,-oo oD calculated from muscovite-water H isotope fractionation at the plagioclase-muscovite Na-K exchange temperature, and ca. 1o,-oo o34s assuming negligible pyrrhotite-waterS isotope fractionation. The isotopic composition of the late o18o (stage Ill) vein-forming fluid at Hog Mountain (Table 1; Figs. 2-4) is 7.61.2%o calculated from quartz-calcite 0 isotope fractionation, -229,-oo oD calculated from muscovite-water H isotope o13c fractionation at the plagioclase-muscovite Na-K exchange temperature, -5.91.3%o calculated from calcite-water C isotope fractionation at the quartz-calcite 0 temperature, and ca. 9%o o34s assuming negligible pyrrhotite-water S isotope fractionation. 40 -4 -6 0 -8 --~ ~ ..1.10 -10 -12 I > ....... cal-water '._ ~_,._...-..-.------- HM6/235 cal .. F ~. --....... ...... ........ ..._ ..._ ...... ~ .. ... ' MD-293wr -14 - - - - pql-tmz-sca(Illl) -.... -... - - -- - - - - pl-ms (II) - - pl-ms .._-. JG~136wr grt-bt -16 100 200 300 500 600 700 Figure 4. Calculated o13c value of aqueous fluid In equilibrium with a calcite (solid line) mineral separate from the Hog Mountain mine (Table 1). Also shown Is the graphite-C02 fractionation curve for graphite-rich (MD-293: 9.1o/oC, -19.9%o o13C) and normal graphitic phyllite (JG-136: 0.8% C, -23.8%o o13C) from the Wedowee Group (Wesolowski et al., 1987). Temperature ranges as In Figure 2. The isotopic composition of the vein-forming fluid in the Goldville district (Table 2; Figs. 5-6) is 10.2-13.2"oo o18o calculated from quartz-plagioclase (3 pairs) and quartz-biotite 0 isotope fractionation, -48 2%o oD calculated from biotite-water H isotope fractionation at the quartz-biotite 0 temperature, -50 Woo oD calculated from chlorite-water H isotope fractionation at the quartz- chlorite 0 temperature, and ca. 19%o o34s assuming negligible pyrrhotite-water S isotope fractionation. The isotopic composition of the late-sta~e fluid (Table 2; Figs. 5-6), calculated from o quartz-chlorite 0 isotope fractionation is 9.4 0.4"oo so. Hog Mountain Tonalite DISCUSSION All tonalites exhibit incipient alteration: sericitization of plagioclase and introduction of hydrothermal quartz. As quartz is markedly enriched In 18o relative to most other rock-forming silicates (see summaries by O'Neil, 1986 and Kyser, 1987}, silicification would systematically increase o18o values of altered tonalites . Thus, unaltered tonalites were probably characterized by slightly lower values, of the order of 9-10"oo o18o, similar to those of other pre- and synkinematic granitoids in the northern Alabama Piedlmont (Wesolowski et al., 1987). The strong ly positiveS isotopic values of least-altered tonalites (10-11%o o34S) are clearly non-mantle (01%o: Ohmoto, 1986) and may reflect incorporation of sed imentary S during generation and/or emplacement, but pervasive incipient alteration of the tonalite and the low abundance of Sin the least-altered rocks (<0.3%: Green ahd Lesher, 1988) preclude an unique interpretation. 41 o18o Figure 5. Calculated values of aqueous fluids In equilibrium with quartz {thin solid lines), plagioclase {dashed lines), biotite (stippled line), and chlorite (thick solid line) mineral separates from the Lowe mine and Jones vein (Table 1). Also shown are temperature ranges calculated from quartz-plagioclase, quartz-biotite, and quartz-chlorite 0 Isotope fractionation, garnet-biotite Fe-Mg exchange thermometry, and plagioclase-muscovite Na-K exchange thermometry. Vein Emplacement The restrictive association of gold mineralization with structurally-controlled veins, their transgressive relationship to stratigraphy, and thermobarometric data all indicate that the present distribution of gold is a result of metamorphism during progressive 01-02 deformation. Temperatures of 0 isotopic equilibration in veins and alteration selvages at Hog Mountain (475 45C for qtz-ms and 335 40C for qtz-cal) correspond very closely to temperatures calculated using plagioclase-muscovite Na-K exchange thermometry (Lesher and Green, 1987 and unpubl. data) and to maximum inferred trapping temperatures of fluid inclusions in vein quartz (Sha and Lesher, this volume) . These temperatures are similar to slightly lower than estimates of peak metamorphism in the pluton (460 50C and 4.5 1kb: Lesher and Green, 1987 and unpubl. data), indicating that vein emplacement occurred during or slightly after the metamorphic peak, consistent with observed field relationships. The discordant D and 0 isotopic fractionation between quartz, muscovite, and calcite at Hog Mountain (Table 1) suggests that 0 in quartz and muscovite records the temperature (and fluid composition) during initial alteration, whereas H in muscovite and 0 in calcite record conditions at lower temperatures. 42 -35 -40 -45 -"()' ~ 0 -50 c..o Ibt-water I chi-water ------------ - - - - - L0-32 chi -55 L0-26 bt - qtz-chl -60 - - - qtz-bt gnt-bt - - - - - - - - - pl-ms -65 200 250 300 350 400 450 500 550 600 650 700 T (C) Figure 6. Calculated oD values of aqueous fluids In equilibrium with biotite (solid line) and chlorite (dashed line) mineral separates from the Lowe mine. Temperature ranges as In Figure 5. Temperatures of 0 isotopic equilibration at Goldvllle (550-340C for qtz-pl, 520 30C tor qtz- bt, and 350 ~soc for qtz-chl) overlap the 490 50C and 510 80C temperatures calculated by Stowell et al. (in press) and Stowell and Lesher (this volume) using garnet-biotite Fe-Mg and plagioclase-muscovite Na-K exchange tl1ermometry, respectively, and are consistent with the PV-T-X properties of fluid inclusions in vein quartz (Sha and Lesher, this volume). The garnetbiotite, quartz-biotite, and quartz-plagioclase temperatures calculated from vein minerals and vein selvages are identical, within errors, to prograde metamorphic temperatures determined in the country rocks. These data and prograde zoning of garnet grains indicates that the veins were emplaced during peak metamorphism. The quartz-chlorite temperatures probably reflect retrograde metamorphic conditions; the differences (plagioclase = biotite > chlorite) are consistent with observed paragenetic relationships. Hydrothermal Alteration The systematic enrichment of D in altered tonalites at Hog Mountain and altered pelites at Goldville reflects selective incorporation of D into phyllosilicates during alteration. At the calculated temperatures of vein formation, muscovite is enriched in D relative to biotite by ca. 20%o (relatively temperature independent) and chlorite is enriched in D relative to biotite by ca. 15-30%., (strongly temperature dependent) (Suzuoki and Epstein, 1976; Graham et al., 1987). These enrichments are consistent both in sense and magnitude with the observed positive H isotopic shifts in the altered rocks. At Hog Mountain, oD values of strongly-altered decussate-textured tonalites (-50 and -53'Yoo) and the corresponding calculated muscovite (-49'Yoo) are higher than those of stronglyaltered foliated rocks (-58 and -62'Yoo) and a corresponding muscovite separate (-63'/'oo), suggesting that the latter re-equilibrated with the fluid at lower temperatures during postemplacement deformation. 43 40 20 0 -'() ~ 0c.o -20 -40 -60 -80 -100 -20 ''' ~::..S MWL ''''' ''' M ...... .....~~ -.LO inc. alteration -10 20 30 Figure 7. Plot of BD vs. B1Bo showing whole-rock Isotopic compositions of least-altered and altered Hog Mountain tonalltes (solid diamonds) and Wedowee metapelltes (open diamonds) and calculated compositional fields of veins-forming fluids at the Hog Mountain mine (HM) and the Lowe mine (LO), compared to calculated compositional fields of deep sedimentary brines (DSB), metamorphic fluids (M/M), magmatic fluids (M), meteoric waters (MWL), and standard mean ocean water (S) (Sheppard, 1986). As noted above, incipient sericitization and silicification are probably responsible for the minor dispersion in B1Bo values of least-altered tonalite's (Fig. 7). At the calculated temperatures of vein formation, muscovite is depleted in 18Q relative to quartz by 2-3%o, depleted relative to plagioclase by 1-2%o, and enriched relative to biotite by 3-4%o (O'Neil and Taylor, 1969; Matsuhisa et al., 1979; Bottinga and Javoy, 1975). The trend of slightly decreasing o18Q with increasing intensity of alteration is consistent with the original proportions of quartz, plagioclase, and biotite in the tonalite (pl>qtz>bt), the small magnitudes of fractionation involved, and the offsetting directions of 18o enrichment (qtz>pl>ms>bt) relative to alteration trends (pl~ms+qtz, bt~ms) . Although there are insufficient data to quantitatively model fluid-rock isotopic fractionation during alteration, the trends of D enrichment and 1Bo depletion with increasing degree of alteration of tonalites at the Hog Mountain mine and metapelites at the Lowe mine are similar and are consistent with alteration by fluids of compositions calculated independently from the isotopic compositions of mineral separates (Fig. 7). For example, least-altered tonalites (solid diamonds: ca. -80%o SO and 10%oo1BQ) would be in Hand 0 isotopic equilibrium with magmatic fluids (ca. 50%oBD and 8%oo18Q) at granite solidus temperatures, whereas strongly-altered tonalites (solid diamonds: ca. -50%o oD and 1Oroo o1BQ) would be in isotopic equilibrium with calculated veinforming fluids (HM) at metamorphic temperatures. At the same temperatures, least-altered 44 metapelites (open diamonds: ca. -90%o SO and 13%o s1SO) would be in H and 0 isotopic equilibrium with metamorphic fluids (M/M : ca. -70%o SO and 11%o s18Q). whereas strongly-altered metapelites (open diamonds: ca. -70,..oo SO and 11%o o18Q) would be in isotopic equilibrium with calculated vein-forming fluids (LO) . At lower (diagenetic) temperatures, the metapelites would be in isotopic equilibrium with deep sedimentary brines (DSB). The correspondence between rock alteration trends and fluid compositions calculated from vein mineral separates indicates that the same fluids from which vein minerals crystallized were also responsible for alteration of the country rocks and introduction of gold. Fluid Source The presence of vein-associated alteration at Hog Mountain and Goldville indicates that the vein-forming fluids were not in chemical or isotopic equilibrium with the country rocks and could not have been derived locally by "lateral secretion.'' This is supported by Sr isotopic data at Hog Mountain which preclude derivation from Wedowee metapelites. Strontium isotopes do not fractionate, so the 87Srt86sr ratio of the fluid will be equal to that of the 0.7061 0.0012 initial 87srt86sr ratio of the hydrothermal alteration zone (Green et al., 1988). This value is significantly less radiogenic than average Wedowee Group metapelites which exh ibit initial 87Sr/B6sr = ca. 0.715 (Russell, 1978), but more radiogenic than the least-altered tonalites (87srt86Sr;nitial = 0.7050: N.L.G., C.M.L., and A.K. Sinha, unpubl. data). The limited C isotopic data are equivocal. Although the calcite at Hog Mountain could have been in C isotopic equilibrium with the most e -rich graphitic phyllonite analyzed by Wesolowski et al. (1987), most Wedowee metapelites exhibit more negative s13c values (down to -26.5%o) and could not have been in equilibrium with the fluids that equilibrated with the calcite . The vein-forming fluid at Hog Mountain was characterized by 7 to 1Woo s1Bo, -15 to -31%o SO,5 to -~oo s13c, and ca. 9,..oo o34S, whereas the vein-forming fluid at Goldville was characterized by 9 to 13%o o1So, -44 to -48,..oo SO, ca. 19%o o34s . The H and 0 isotopic composition of the Hog Mountain fluid is similar to that of deep sedimentary brines or metamorphic fluids, but is markedly different from seawater, meteoric waters, or magmatic fluids (Fig. 7) . The Goldville fluid is also within the field of metamorphic fluids, but is significantly depleted in D and enriched in 34S re lative to the Hog Mountain vein-forming fluid. Although these fluids fall within the range that might form by mixing of metamorphic fluids and meteoric waters, the alteration trends do not define mixing lines between such fluids. The positive S isotopic signatures of the fluids are also consistent with an oxidized, nonmagmatic, probable metasedimentary source for the S (see Ohmoto, 1986). The vein-forming fluid at Hog Mountain was enriched in 0, Rb, K, Ba, Fe, W, S, and Au, and the vein-forming fluid at Goldville was enriched in D, Rb, K, Ba, Mg , Fe, AI, and Au. All these elements are enriched in pelitic sedimentary rocks and could be mobilized during metamorphism. As such, vein-forming fluids were most likely derived from deeper parts of the metamorphic pile, metasediments of the Ashland-Wedowee belt underlying the Goldville-Hog Mountain area, metasediments of underlying thrust sheets, and/or metasediments of the underlying North American continental shelf (Green et al., 1988), and asce nded along permeable zones that developed during deformation. CONCLUSIONS Auriferous quartz veins in the central part of the Ashland-Wedowee belt were emplaced after the initiation of deformation and during to slightly after peak metamorphism. Temperatures of 0 isotopic equilibration at Hog Mountain are 475 45C for quartz-muscovite and 335 40C for quartz-calcite, consistent with 425 50 and 325 50C temperatures obtained by plagioclase- muscovite Na-K exchange thermometry for type II and type Ill veins, respectively. Temperatures of 0 isotopic equilibration at Goldville are 550-340C for quartz-plagioclase, 520 30C for quartz- biotite, and 350 15C for quartz-chlorite, consistent with 49050C temperatures obtained by garnet-biotite Mg-Fe exchange thermometry and 51 080C temperature obtained by plagioclasemuscovite Na-K exchango thermometry. Discordant D and 0 isotopic fractionation between 45 quartz, muscovite, and calcite at Hog Mountain suggests that 0 in quartz and muscovite records the temperature (and fluid composition) during initial alteration, whereas H in muscovite and 0 in calcite record conditions at lower temperatures. The presence of alteration envelopes adjacent to mineralized veins at the Hog Mountain deposit and at deposits in the Goldvllle district indicates that the country rocks in both areas were not in equilibrium with vein-forming fluids and could not have been derived locally, negating models involving local mobilization ("lateral secretion") of syngenetic gold into quartz veins during metamorphism. Altered tonalites at Hog Mountain and altered metapelites at Goldville are markedly enriched in D relative to least-altered equivalents, reflecting selective incorporation of D in phyllosilicates. Altered metapelites at Goldville are also depleted in 1So, reflecting exclusion by phyllosilicates. The vein-forming fluid at Hog Mountain was characterized by 7 to 11%o s1ao, -15 to -31%o SD, - 5 to -7%o a13c, and ca. 9%.. s34s, whereas the vein-forming fluid at Goldville was characterized by 9 to 13%., s1Bo, -46 to -51%o oD, and ca. 19%o s34s. The Hand 0 isotopic compositions of such fluids are similar to those of basinal brines or metamorphic fluids, but are markedly different from seawater, meteoric waters, or magmatic fluids. The stable isotopic data, together with field, petrographic, microthermometric, and geochemical data, suggest that vein-forming fluids in both areas equilibrated to varying degrees with, and were probably derived from, pelitic metasedimentary rocks, but were not derived from adjacent country rocks. Mineralizing fluids were most likely derived from underlying thrust sheets and/or the underlying North American continental shelf. ACKNOWLEDGEMENTS We are grateful to Newmont Exploration Ltd. and U.S. Borax for donating diamond drill core to the University of Alabama and the Geological Survey of Alabama and for providing copies of unpublished drill logs, maps, and reports; to Russell Lands Inc. for access to the deposits; to Joe Thacker and Peng Sha for performing mineral separations; to Ed Ripley for performing the stable isotope analyses; and to Mark Drummond for his insightful comments on the manuscript. We have benefitted greatly from discussions with many colleagues, especially Bob Cook and Greg Guthrie. Our research on gold deposits in Alabama has been supported by the University of Alabama Research Grants Committee (grant nos. ~340 to C.M.L., ~345 to N.L.G., and 1370 to H.H.S.) and School of Mines and Energy Development (grant nos. 2-29107 to N.L.G. and C.M.L., and 2-291 ~ 5 to C.M.L., N.L.G., and H .H.S.). REFERENCES Bottinga, Y., 1969, Calculated fractionation factors for carbon and hydrogen isotope exchange in the system calcite-carbon dioxide-graphite-methane-hydrogen-water vapor: Geochimica et Cosmochimica Acta, v. 33, p. 49-64. Bottinga, Y., and Javoy, M., 1975, Oxygen isotope partitioning among the minerals in igneous and metamorphic rocks : Reviews in Geophysics and Space Physics, v. 13, p. 401-418. Clayton, R.N., and Mayeda, T.K., 1963, The use of bromine pentafluoride in the extraction of oxygen from oxides and silicates for isotopic analysis: Geochimica et Cosmochimica Acta, v. 27, p. 43-52. Cole , D.R ., and Ohmoto, H., 1986, Kinetics of isotopic exchange at elevated temperatures and pressures, in Valley, J.W., Taylor, H.P., and O'Neil, J.R. (editors), Stable Isotopes in High Temperature Geological Processes : Mineralogical Society of America, Reviews in Mineralogy, v. 16, p. 41 -87. Fritz, P., Drimmie, R.J., and Nowickie, V.K., 1974, Preparation of sulfur dioxide for mass spectrometer analysis by combustion of sulfides with copper oxide: Analytical Chemistry, v. 76, p. 164. 46 Graham, C.M., Viglino, J., and Harmon, R.S., 1987, Experimental study of hydrogen isotope exchange between aluminous chlorite and water and of hydrogen diffusion in chlorite: American Mineralogist, v. 72, p. 566-579. Green, N.L., and Lesher, C.M., 1987, Mineralogy and geochemistry of the Hog Mountain pluton, Northern Alabama Piedmont, in M.S. Drummond, N.L. Green, and M.J. Neilson (editors), Iha Granites of Alabama: Alabama Geological Society, 24th Annual Field Trip Guidebook, p. 1-12. Green, N.L., and Lesher, C.M., 1988, Geology of the Hog Mountain tonalite and associated lode gold deposits, northern Alabama Piedmont: University of Alabama School of Mines and Energy Development Faculty Research Report 88, 42 p. Green, N.L., Sinha, A.K., and Lesher, C.M., 1988, Gold associated with retrograded shear zones: An Alleghanian episode of mineralization in the southern Appalachians: Geological Society of America, Annual Meeting, Abstracts with Programs, v. 21, no. 7, p. A277. Guthrie, G.M., and Dean, L.S., 1989, Geology of the New Site 7.5-Minute Quadrangle, Tallapoosa and Clay Counties, Alabama: Alabama Geological Survey Quadrangle Map 9, 41 p. Guthrie, G.M., and Lesher, C.M., 1988, Structural and metamorphic controls on lode gold mineralization, Tallapoosa County, Alabama: Geological Society of America, Southeastern Section Meeting, Abstracts with Programs, v. 20, no. 4, p. 268. Guthrie, G.M., and Lesher, C.M., in press, Stratigraphic, metamorphic, and structural settings of lode gold deposits in the northern Alabama Piedmont, in Lesher, C.M., Cook, A.B., and Dean, L.S. (editors), Gold Oeposjts of Alabama: Geological Survey of Alabama Bulletin 136, p. 11-32. Kendall, C., and Coplen, T.B., 1982, Multisample conversion of water to hydrogen by Zn for stable isotope determination: Analytical Chemistry, v. 57, p. 1437-1440. Kyser, T.K., 1987, Equilibrium fractionation factors for stable isotopes, in Kyser, T.K. (editor), Stable Isotope Geochemistry of Low Temperature Fluids: Mineralogical Association of Canada, Short Course, v. 13, p. 1-84. Lesher, C.M., and Green, N.L., 1988, Syn-metamorphic hydrothermal alteration, Hog Mountain gold deposit, Northern Alabama Piedmont. Geological Society of America, Southeastern Section Meeting, Abstracts with Programs, v. 20, no. 4, p. 277. Lesher, C.M., and Green, N.L., 1987, Alteration and gold mineralization, Hog Mountain pluton, Northern Alabama Piedmont, in M.S. Drummond, N.L. Green, and M.J. Neilson (editors), Iha Granites of Alabama, 24th Annual Field Trip Guidebook, Alabama Geological Society, p. 1322. Matsuhisa, Y., Goldsmith, J.R., and Clayton, R.N., 1979, Oxygen isotope fractionation in the system quartz-albite-anorthite-water: Geochimica et Cosmochimica Acta, v. 43, p. 1131-1140. McConnell, K.l., Abrams, C.E., and German, J.M., 1986, Gold deposits in the Northern Piedmont of Georgia and Alabama and their relationship to lithostratigraphic units, in Misra, K.C. (editor), volcanogenic Sulfide and Precious Metal Mineralization jn the Southern Appalachians : University of Tennessee, Department of Geological Sciences, Studies in Geology 16, p. 164181. Neal, W.L., 1986, Geochemical relationships between gold, silver, antimony, arsenic, and bismuth in gold mineralization of the Goldville District, Tallapoosa County, Alabama: Auburn University, unpublished M.S. thesis, 151 p. 47 Neal, W.L., and Cook, A.B., in press, Mineralogy, petrology, and geochemistry of phyllite-hosted gold deposits, Goldville district, Tallapoosa, Alabama, in Lesher, C.M., Cook, A.B., and Dean, L.S. (editors), Gold Deposits of Alabama: Geological Survey of Alabama Bulletin 136, p. 101114. Neal, W.L., Cook, A.B., and Paris, T.A., in press, Gold and related trace metal geochemistry at the Tallapoosa and Lowe mines, Goldville district, Tallapoosa County, Alabama, in Lesher, C.M., Cook, A.B., and Dean, L.S. (editors). Gold Deposits of Alabama : Geological Survey of Alabama Bulletin 136, p. 115-132. Neathery, T.L., and Reynolds, J.W., 1973, Stratigraphy and metamorphism of the Wedowee group: a reconnaissance: American Journal of Science, v. 273, p. 723-741. Ohmoto, H., 1986, Stable isotope geochemistry of ore deposits, in Valley, J.W., Taylor, H.P., Jr., and O'Neil, J.R. (editors), Stable Isotopes jn High Temperature Geological Processes: Mineralogical Society of America, Reviews in Mineralogy, v. 16, p. 491-559. O'Neil, J.R., 1986, Theoretical and experimental aspects of isotope fractionation ,in Valley, J.W., Taylor, H.P., Jr., and O'Neil, J.R. (editors) , Stable Isotopes in High Temperature Geological Processes: Mineralogical Society of America, Reviews in Mineralogy, v. 16, p.1-40. O'Neil, J.R., and Taylor, H.P., Jr., 1969, Oxygen isotope equilibrium between muscovite and water: Journal of Geophysical Research, v. 74, p. 6012-6022. O'Neil, J.R., Clayton, R.N., and Mayeda, T.K., 1969, Oxygen isotope fractionation in divalent metal carbonates: Journal of Chemical Physics, v. 51, p. 5547-5558. Pardee, J.T., and Park, C. F., Jr., 1948, Gold deposits of the southern Piedmont: U.S. Geological Survey Professional Paper 213, 156 p. Paris, T.A., 1986, The Goldville project: ResuHs of an exploration project for gold in the northern Alabama Piedmont, in Misra, K.C. (editor), Volcanogenic Su lfide and Precious Metal Mineralization in the Southern Appalachians: University of Tennessee, Department of Geological Sciences, Studies in Geology 16, p. 185-205. Park, C.F., Jr., 1935, Hog mountain gold district, Alabama: American Institute of Mining and Metallurgical Engineers Transactions, Mining Geology, v. 115, p. 209-228. Rosenbaum, J. and Sheppard, S.M.S., 1986, An isotopic study of siderites, dolomites, and ankerites at high temperatures. Geochimica et Cosmochimica Acta, v. 50, p. 1147-1150. Russell, G.S., 1978, U-Pb, Rb-Sr, and K-Ar isotope studies bearing on the tectonic development of the southernmost Appalachian orogen, Alabama: Florida State University, unpublished Ph.D. dissertation, 197 p. Sha, P., and Lesher, C.M., this volume, Microthermometry of fluid inclusions in auriferous quartz veins, Hog Mountain-Goldville district, northern Alabama Piedmont, in Cook, A.B. (editor), Economic Mjneral Deposits of the Southern Appalachians: Metallic Ore Deposits: Georgia Geologic Survey Bulletin 117. Sheppard, S.M.F., 1986, Characterization and isotopic variation in natural waters, in Valley, J.W., Taylor, H.P., Jr., and O'Neil, J.R. (editors), Stable lsotooes jn High Temperature Geological Processes: Mineralogical Society of America, Reviews in Mineralogy, v. 16, p. 165-183. Stowell, H.H., and Lesher, C.M., this volume, Thermobarometry of synmetamorphic gold deposits, Goldville district, Northern Piedmont, Alabama, in Cook, A.B. (editor), Economic 48 Mineral Qeposjts of the Southern Appalachians: Metallic Ore Qeposjts: Georgia Geologic Survey Bulletin 117. Stowell, H.H., Guthrie, G.M., and Lesher, C.M., 1989, Metamorphism and gold mineralization in the Wedowee Group, Goldville district, Northern Alabama Piedmont, in Lesher, C.M., Cook, R.B., and Dean, L.S. (editors), Gold Deposits of Alabama: Geological Survey of Alabama Bulletin 136, p. 133-158. Suzuoki, T., and Epstein, S., 1976, Hydrogen isotope fractionation between OH-bearing minerals and water: Geochimica et Cosmochimica Acta, v. 40, p. 1229-1240. Tull, J.F., 1978, Structural development of the Alabama Piedmont northwest of the Brevard zone: American Journal of Science, v. 278, p. 442-460. Wenner, D.B., and Taylor, H.P., Jr., 1971, Temperatures of serpentinization of ultramafic rocks based on 180/ 16o fractionation between co-existing serpentine and magnetite: Contributions to Mineralogy and Petrology, v. 32, p. 165-185. Wesolowski, D., Drummond, M.S., and Gomolka, J., 1987, Oxygen and carbon isotope distributions in granitoids and metasediments of the Rockford district and adjacent portions of the northern Alabama Piedmont and Inner Piedmont, in Drummond, M.S., and Green, N.L. (editors) , Granites of Alabama: Geological Survey of Alabama, p. 221-238. 49 THE ROYAL VINDICATOR GOLD DEPOSIT, HARALSON COUNTY, GEORGIA: A METAMORPHOSED EPITHERMAL HOT-SPRINGS TYPE GOLD DEPOSIT Travis A. Paris Consulting Geologist 5604 Malmsbury Road, Knoxville, TN 37921-3827 ABSTRACT The Royal Vindicator gold deposit occurs in the northern Piedmont of western Georgia near the boundary between the Talladega Group and the Wedowee Group. The deposit is hosted in the lower part of an alternating sequence of felsic and mafic metavolcanics interpreted to be equivalent to the Hil labee Greenstone. Regional strike of foliation/bedding at the deposit is approximately east-west, dipping 15-20 to the south. The overlying Wedowee Group is in gradational contdct with the underlying Hillabee metavolcanics, not in fault contact a s is common in Alabama and as suggested by Heuler and Tull (1989). The gold ores consist of strata-bound pyritic, siliceous bodies hosted in altered felsic metavolcanics. Gold particles observed in reflected light microscopy range from 0.0005 mm to 2.0 mm in longest dimension~ and occur in four different habits: (1) discrete ovoid to ragged particles of native gold ; (2) ovoid to ragged gold grains within and on pyrite grains ; (3) composite grains of gold associated with h ighly reflective, silvery white and duller gray, unidentified minerals; and (4 ) subhedral grains intergrown with magnetite or hematite. Geochemical analyses of the ore bodies reveal a strong pos~tive correlation of gold with sil ver and molybdenum and a sub tle correlation with mercury, arsenic, copper, and lead. The gold mineralization is interpreted to have been deposited synchronously with felsic tuffs along the flanks of shallow, elongate basins by a volcanic hot springs system. The hot springs solutions formed siliceous sinters, altering and replacing tuffs as they were being deposited. During the Devonian-aged Acadian(?) orogeny the deposit was metamorphosed to the biotite zone and possibly the garnet zone of the greenschist facies and deformed. Altered tuffs are now siliceous, auriferous, pyritic magnetite-sericite- quartz schists. The metasinters are now fine-grained, alternating layers of auriferous quartz and pyrite, oriented parallel to bedding/foliation. These auriferous horizons contrast with barren milky quartz veins which both parallel and cut across bedding/foliation. 50 INTRODUCTION This paper describes the geology of the Royal Vindicator gold deposit and presents data on the geologic setting, geochemistry of ore and country rocks, modes of gold occurrence and distribution, and interpretation of the origin of the deposit. LOCATION The Royal Vindicator gold mine is a former gold producer currently being reevaluated by industry concerns. It is located in southwestern Haralson County, Georgia, approximately 5 km east of the Georgia-Alabama state line, and 4.8 km south of the town of Tallapoosa (fig. l). The countryside surrounding the site is comprised of small farms and timberlands. PRODUCTION HISTORY The exact point in time at which gold was discovered at the Royal Vindicator deposit is obscure. Yeates and others (1896) reported that gold was originally discovered at the site by a farmer named Owens around 1840. More recently, however, Reeves (1975) reported t hat gold may h ave been discovered at the site as ear l y as 1 829 . SheaLer (1917) noted that the deposit was be i ng operated as th e Ho llins (Holland) mine in 1834. The depos i t was operated as a successful placer and saprolite mine with minor production from shallow shafts from the time of its discovery until the Civil War (Yeates and others, 1896; Shearer, 1917). Production after the war was sporadic with various operators depleting the surface deposits and switching to underground methods. These later mining attempts lasted for only a few years at a time because of poor economics resulting from the low recovery of gold from increas ingl y sul fi dic ores. During one such period, beginning in 188 7, t he d eposit was known as the Camille mine (Yeates and ot hers , 1896 ; She arer , 19 17 ). In 1896 the deposit was purchased by t h e Roy al Gol d Mining Company which erected a concentrating mi l l , roa sting f u rnace , and chlorination (modified Thies) pl ant for treat ing t he sulfidic ores (Yeates and others, 1896; Lane, 1900; Jones, 1909). This milling operation was also unsuccessful in obtaining an economic recovery of gold from the ores. Underground mining prior to 1900 was through inclined shafts, but by 1910 a vertical shaft had been sunk to 155 feet (47.2 m). The mine was last worked through this shaft from 1914 to 1917 by the Royal Vindicator Mines Company before finally closing permanently (Shearer, 1917; Ruhl, 1917; Reeves, 1975). The mine was dewatered, sampled, and evaluated with two core holes in the mid-1930's. This effort resulted in no renewed production (Maillot, 1940). 51 Figure 1. Location of the Royal Vindicator gold deposit, Tallapoosa County, Georgia. Map base taken from the Tallapoosa, Georgia, 7.5-minute quadrangle. Contour interval is 20 feet. 52 The most recent exploration of the deposit began in late 1979. U.S. Borax conducted regional reconnaissance and detailed geologic mapping of the mine region followed by soil and stream sediment geochemistry, ground magnetic and VLF-EM surveys, and core and reverse circulation drilling. In 1987 the prospect was acquired by a joint venture between Atlantic Goldfields, Inc., and Jascan Resources, Inc., who conducted further core and shallow auger drilling. The main shaft was also dewatered and reconditioned to allow mapping and channel sampling of the 105 foot (32 m) level workings. Exploration to the end of 1988 has identified approximately 75,000 tons of old tailings averaging 0.045 oz/tn gold and outlined south-southeast trending mineralized bodies containing approximately 350,000 tons grading 0.145 oz/tn gold. Yeates and others (1896) and Jones (1909) reported that the Royal mine yielded approximately 5,000 ounces of gold from surficial and shallow underground workings. More recent studies, however, indicate as much as 40,000 ounces of gold with minor silver credits were produced from an estimated 250,000 tons of ore. This conclusion is based on the extent of surficial and underground workings, mine assays, and the volume and grade of tailings (West Georgia College, unpublished reports; Ruhl, 1917; Maillot, 1940). PREVIOUS INVESTIGATIONS The Royal Vindicator gold mine occurs in one of a group of greenstone and felsic gneiss units containing gold and copper mineralization which has historically been called the Dahlonega gold belt (Yeates and others, 1896; Jones, 1909). The structural and stratigraphic setting of the Royal Vindicator mine area has only recently been examined in greater detail and is still not well defined. Hurst and Crawford (1969) located the Royal mine in an arcuate-shaped unit of chlorite schist and amphibolite flanked to the north by quartzite and phyllite of the Talladega Group and to the south by garnetiferous, graphitic mica schists of the Wedowee Group. They interpreted the gold bearing quartz and sulfide mineralization as the product of hydrothermal alteration along a shear zone. Cressler (1970) indicated the greenstone outcrops along Georgia Highway 100 immediately east of the mine and in the Walker Creek valley are Hillabee Chlorite Schist (Hillabee Greenstone). The quartzite ridge l km north of the mine and forming Tally Mountain to the northeast was correlated with the Cheaha Sandstone Member of the Talladega Slate. Neathery and Hollister (1984) show the Royal deposit is hosted in chlorite schist associated with other mineralized mafic units of the western Georgia Piedmont associated with 53 the Wedowee Group. Geologists of the Georgia Geologic Survey have recently placed the Royal Vindicator deposit in various geologic units. While the majority of gold production in the Dahlonega gold belt occurred in the New Georgia Group (Abrams and McConnell, 1984~ McConnell and Abrams, 1982, 1984~ McConnell and others, 1986~ and German, 1985), the Royal Vindicator and a number of other west Georgia gold deposits fall outside of this stratigraphic unit. Abrams and McConnell (1984), McConnell and Abrams (1984), and McConnell and others (1986) placed the Royal Vindicator in the stratigraphically younger Dog River Formation of the western Sandy Springs Group, which structurally overlies rocks of the Talladega Group to the northwest, and is separated from them by the Allatoona Fault. More recently, German (1988) assigned the Royal- Vindicator deposit to an unnamed greenstone and quartzofeldspathic gneiss unit flanked to the north by the Talladega Group and to the south by the Andy Mountain Formation of the Sandy Springs Group (western belt). The presence of a fault separating the various units is not indicated, and German (1988; in press) correlates the greenstone unit with the Hillabee Greenstone, which is in gradational contact with underlying Talladega Group metasediments. Heuler and Tull (1989) have extended the Hillabee Greenstone into Georgia. They conclude the arcuate chlorite schist unit in the deposit vicinity, as well as a linear chlorite schist unit between Tallapoosa and Buchanan, Georgia, is Hillabee Greenstone. The Jemison Chert and Butting Ram Sandstones of the Talladega Group metasediments are extended by Heuler and Tull (1989) into the mine area as the quartzite and phyllite units which underlie the prominent ridge north of the mine site and Tally Mountain to the northeast. Wedowee Group metasediments structurally overlie the Hillabee Greenstone and are separated from it by the Hollins Line Fault. However, Heuler and Tull (1989) place the Royal Vindicator gold deposit within a felsic schist unit contained in the Wedowee Group, not in the Hillabee Greenstone. Paris (1986, l989a) has recognized that the Royal Vindicator gold deposit is hosted by siliceous schist in a felsic metavolcanic unit interbedded with mafic and felsic metavolcanics which correlate stratigraphically and geochemically with the Hillabee Greenstone. This metavolcanic sequence is stratigraphically underlain by metasediments of the Talladega Group and stratigraphically overlain by metasediments of the Wedowee Group (Paris, 1986). 54 STRATIGRAPHY Introduction The Royal Vindicator mine area occupies predominantly forested uplands and some swamp areas which are deeply weathered and contain only scattered, small rock outcrops. Three major lithostratigraphic units crop out in the mine area - the Talladega Group, the Hillabee Greenstone, and the Wedowee Group. Gold mineralization of potential economic grade occurs within the upper part of a silicified, pyritized felsic unit in the lower part of the Hillabee Greenstone sequence. Talladega Group The oldest lithostratigraphic unit in the mine area is the Talladega Group. Muscovite-quartz schist and phyllite intercalated with thin quartzite beds underlie the prominent ridge approximately 350m north of the mine (fig. 2). These rocks are suggested to be correlative with the Jemison Chert and Butting Ram Sandstone sequence of the Talladega Group, in agreement with Heuler and Tull (1989). Exploration drilling approached to within 5- to 10-m of the Hillabee Greenstone-Talladega Group contact but did not cut it. More recently, however, this stratigraphic contact has been penetrated by core drilling at the mine site by the Georgia Geologic Survey (German, in press), and confirms that mafic metavolcanics are in gradational contact with siliceous and calcareous metasediments of the Talladega Group. Similar contact relationships have been recognized in Alabama by Tull and others (1978), Tull and Stow (1980, 1982), and Paris (l989b). Wedowee Group Bluffs and uplands south of the mine area and Walker Creek are underlain by rocks correlated with the Wedowee Group which structurally and stratigraphically (?) overlies the Hillabee Greenstone (fig. 2). In Georgia, rocks of the Wedowee Group are known as the Andy Mountain Formation of the Sandy Springs Group, western belt (Abrams and McConnell, 19 81) . The Wedowee Group, as seen in both core and outcrop at the Royal Vindicator deposit, is comprised mainly of graphitic phyllite, sulfide-bearing chlorite-garnet-biotitemuscovite+graphite phyllite, and massive to locally micaceous quartzite~ Also present are lesser amounts of biotitechlorite+amphibole schist, garnetiferous quartzite, and milky quartz veins. These lithologies are similar to those reported for the Cragford Phyllite of the Wedowee Group in Alabama by Neathery and Reynolds (1975). 55 Tallapoosa ---------~ I I I I I I EXPLANATION Wedowee Group undlfferentleted Hlllabee Greenstone Banded-Iron formation Talladega Group undifferentiated N J Figur7 2 . General deposlt (RV) area, Tgaelollaopgoyosa ocfounttyh,e Royal V.indicator Georgia. gold 56 Twelve core holes penetrate the contact of the Wedowee Group metasediments with the underlying Hillabee Greenstone metavolcanic sequence. No fault contact or shear zone appears to be present. The contact appears to be a gradational zone with Wedowee lithologies interbedded with those of the Hillabee. Additionally, a graphitic phyllite and quartzite unit up to 45 m thick which correlates well with the Wedowee Group lies roughly 70 m below the contact of and within the Hillabee Greenstone. Sparse mineralization was also encountered in core which penetrated the Wedowee Group. Base-metal mineralization, comprised of 1- to 2-mm grains of galena and red-brown sphalerite,, occurs in 1- to 3-cm thick massive to banded quartzite horizons hosted in garnet-chlorite-biotite- muscovite phyllite and graphitic phyllite. Also present, disseminated within the phyllites and spatially associated with the base-metal mineralization, are 2- to 3-mm long arsenopyrite crystals. This mineral assemblage is similar to that seen in the Tallapoosa mine, Tallapoosa County, Alabama, a deposit which also occurs in the Wedowee Group (Paris, 1985). Hillabee Greenstone A unit of white, laminated to banded, quartz-sericite schist, striking approximately east-west and dipping at 15 to 50 southeast, was mapped in saprolite exposed in the walls of the old mine cuts. Foliation/bedding is locally modified by chevron folding. Thin glassy quartz lenses and bands locally occur parallel to foliation/bedding; major quartz veining is absent. Weathering of pyrite in the schist leaves cubic- to ragged-shaped, iron-oxide filled voids and stains the saprolite to a tan to red-orange color. No readily identifiable shear zone or other potential control of the gold mineralization is apparent in outcrop. The quartz-sericite schist unit is underlain by weathered chlorite schist in two mine cuts. Weathered epidote-quartz-feldspar+chlorite gneiss and schist overlies the quartz-sericite schist in the southern part of the mine cut and in drainage trenches. Rock types found in outcrops away from the mine workings are predominantly chlorite schist and minor chlorite-epidotefeldspar gneiss, felsic schist, and banded iron formation. All of these rocks, based on lithologies, metamorphic grade, stratigraphic position, and lithogeochemical analyses of drill core, are correlated with the Hillabee Greenstone (Paris, 1986, l989a). Core drilling to delineate the extent and tenor of gold mineralization also allowed the development of a mine area stratigraphy (Table 1). To date, 51 NX/NQ (51-mm diameter) core holes (totaling approximately 7153.2 m) and five 133-mm 57 Table 1. Stratigraphic/Lithologic Column Based on Core Drilling, Royal Vindicator Gold Deposit, Haralson County, Georgia Name Lithology Thickness (meters) Cragford Phyllite (after Neathery and Reynolds, 1975) graphitic phyllite; chloritegarnet-graphite-phyllite; chlorite schist; garnet quartzite >110 Upper Volcanic Sequence chlorite schist; sericite-quartz schist; biotite-quartz-feldspar gneiss; chlorite-epidote-quartzfeldspargarnet gneiss 50-71 Upper Sedimentary Sequence graphite phyllite; chlorite- graphite-mica schist; mica schist; quartzite 0-45 Mixed Chlorite and Sericite Schist Zone chlorite schist; sericitequartzchlorite schist 7-10 Upper Greenstone II chlorite schist 7-15 Upper Gneiss II Ccll 0 .j.) cIll Upper Greenstone Cll Cll ~ (.? Cll Cll .Q Upper Gneiss Ill .-j .-j ~ :r: Hanging Wall Greenstone biotite-quartz-feldspar garnet gneiss chlorite schist with cpy and po bearing epidote-quartz granofels (lapilli or amygdules); minor banded iron-formation at top chlorite-epidote-quartzfeldspartgarnetbiotite gneiss chlorite schist; epidote-chlorite schist 4-23 9-19 10-26 0-9 Siliceous Zone pyrite-magnetite-quartz-sericite schist; quartzite; quartz; sericite-feldspar-quartz epidotegarnet schist; pyrite bands 2-34 Footwall Greenstone chlorite schist; epidosite 2-5 Lower Gneiss biotite-muscovite-quartz feldspargarnet gneiss 11-17 Lower Greenstone chlorite schist; banded ironformation near base 16-17 58 diameter, reverse circulation holes (totaling approximately 359.4 m) have been drilled at the Royal Vindicator in order to outline the gold-bearing zones. Rocks of the Hillabee Greenstone in the mine area appear to have been deposited as an alternating felsic-mafic metavolcanic sequence approximately 250-m thick, overlain by graphitic metasediments. At least one hiatus in volcanism is indicated by a wedge of graphitic phyllite and quartzite in the middle of the volcanic section. The Hillabee stratigraphy in the upper parts of the drill holes is variable from hole to hole, probably representing rapid facies changes in the volcanic environment, possibly tied to waning volcanism. The stratigraphy in the lower portions of the holes is much more consiste nt and appears to represent more stable, cyclic volcanism. The major mafic and felsic cycles are informally named in the order of their occurrence above or below the main gold-bearing unit (Table 1). Mafic Units The mafic metavolcanic units are comprised predominantly of biotite-chlorite schist, chlorite+amphibole schist, and epidote-biotite-chlorite+garnet schist. These units vary from individual beds a -few centimeters to several meters thick. Units comprised of different or similar mafic lithologies vary from 9- to 20-m thick. Possible relict volcanic features include 2- to 5-mrn blocky, saussuritized feldspar megacrysts interpreted as relict phenocrysts, elliptical epidote-quartz clasts (0.5- to 4-cm long) thought to represent compacted lapilli, and rounded calcite-quartz and epidote-quartz structures 2 to 10 mm in diameter which may be filled vesicles or amygdules. Individual mafic metavolcanic units are not traceable in outcrop due to poor exposure and similar gross lithologies. Upper greenstone At least one mafic metavolcanic unit is distinctive and is identifiable and correlatable in core, based on trace mineral content. This unit, the Upper greenstone, contains scattered elliptical epidote-quartz clasts (0.5- to 4-cm long) which contain traces of pyrrhotite, chalcopyrite, and local pyrite in minute fractures in the clast boundaries and interiors. The clasts are interpreted to be deformed lapilli, possibly compressed pumice fragments. Banded Iron Formation Associated with some mafic units are thin (7- to 15-cm thick), banded magnetite-quartz+pyrite granofels units interpreted to be Algoma-type banded iron formation (BIF: Gross, 1980). These banded iron formations crop out in the area and have been intersected by core drilling in two different mafic units at the mine. In core, a highly 59 discontinuous BIF occurs at the top of the Upper greenstone associated with an equally discontinuous felsic lapilli tuff with a chloritic matrix. It has not been identified in outcrop. A second BIF located just above the base of the lower greenstone can be traced discontinuously in outcrop for several kilometers to the east and southeast of the mine area, roughly marking the Hillabee Greenstone-Talladega Group contact. The BIF which occurs at the top of the upper greenstone has not been identified in outcrop. At least two other BIF's are known in the Royal Vindicator mine area (fig. 2). Highly deformed BIF crops out on a low knoll approximately 1.6 km east-northeast of the mine site at what is thought to be the same stratigraphic level as the goldmineralized zone, but contains no gold. Another BIF can be traced discontinuously for approximately 100 m in the vicinity of old prospect workings 0.9 km southeast of the major Royal Vindicator mine workings. Felsic Units Felsic lithologies cored at the Royal Vindicator mine include, sericite-quartz+pyrite schist, sericite-quartz- feldsparipyrite schist, muscovite-biotite-quartz-feldspar +garnet gneiss, and chlorite-epidote-quartz-feldspar+pyrite +garnet+biotite gneiss. These units also cannot be -traced for any-distance beyond or correlated outside of the mine area due to weathering and poor exposure. In the upper portion of the core holes the felsic schistose units are intercalated with chloritic schist. The gneissic units occur in the lower parts of the holes as thin-bedded to massive- bedded horizons 4- to 30-m thick. The dominant minerals of the felsic lithologies are sericite, chlorite, quartz and feldspar. Biotite, epidote, garnet, pyrite, and calcite are accessory minerals. Quartz occurs as fine-grained material of the groundmass, in veinlets of gray, glassy quartz and milky white quartz, and locally as rounded to elliptical (2- to 6-mm across) , bluish quartz "eyes'' which are interpreted as relict phenocrysts. Feldspar (plagioclase) is commonly fine-grained or forms blocky shaped grains (2- to 4-mm across) which are variably saussuritized. Individual blocky grains in the fine-grained groundmass are interpreted as relict phenocrysts. Epidote occurs as fine-grained material in the well foliated to massive groundmass with fine-grained sericite and chlorite, as round radiating crystal masses interpreted as filled amygdules, and as elliptical quartz-epidote masses to 4 em in length interpreted as altered lapilli. Biotite forms 1- to 4-mm brown laths oriented parallel to foliation. Garnet occurs as single, scattered, red to red-purple, subhedral grains 0.5- to 3-mm in diameter, and in bands of grains one or two grains thick. Pyrite varies widely in abundance throughout the metavolcanic section, occurring as anhedral grains (<0.5- to 3-mm), rounded grains or equant to flattened cubes up to 10 mm across, and as bands of subhedral grains 60 parallel to foliation/bedding. Calcite is found as 0.5- to 3-mm disseminated grains in the groundmass and as 0.5- to 1mm thick bands parallel to foliation. The most distinctive felsic units are the Upper gneiss and the Siliceous zone. Upper Gneiss The Upper gneiss is a distinctive marker unit 10- to 26-m thick. It is comprised predominantly of chloriteepidote-quartz-feldspar+pyrite+garnet+biotite gneiss which occurs immediately above the -Siliceous zone or is separated from it by a chlorite schist unit as much as 10-m thick (the Hanging wall greenstone) . The greatest abundance of epidote and chlorite occurs structurally above and laterally displaced to the east of the greatest underlying gold concentrations. Epidote abundance decreases rapidly away from the gold mineralization. The lack of facing criteria allows two interpretations: (1) the epidote and chlorite concentration in the Upper gneiss unit represents an alteration product formed originally beneath the gold horizon, or (2) it represents a volcanic unit above the gold horizon that was altered by mineralizing solutions during and after gold deposition. Siliceous Zone The Siliceous zone is a variably altered, laminated, siliceous, felsic metavolcanic unit. The most intense alteration and gold mineralization are contained in the uppermost 5- to 15-m of the unit. Core drilling indicates that the overall thickness of the Siliceous zone varies from 2.1 to 32.8 meters. The variations in thickness of the Siliceous zone are believed to define two depositional basins with an hourglass shape in plan. The axes of these basins now trend down-dip approximately SBO E. In underground workings and drill core, the unit is represented by intercalated, banded white, gray, and tan feldspathic schist, quartz-sericite schist, and fine-grained silica bands with disseminated and banded pyrite. Gold occurs with strongly altered, siliceous rock containing a distinctive pyrite and iron-oxide assemblage. Silicification occurs as fine-grained quartz permeating the original tuff (alteration), as micro-crystalline bands interpreted to be metachert, and as milky quartz vein segregations oriented parallel to bedding/foliation. The protolith of the Siliceous zone has been logged in drill core away from the altered, gold-bearing zone. There, the rock is an unaltered muscovite-biotite-quartz-feldspar +garnet+chlorite+epidote schistose gneiss (rhyodacitic metavolcanic) . Blocky 1- to 2-mm feldspar grains and 2- to 3-mm rounded blue quartz grains are scattered through the rock, 61 giving it a similar appearance to that of the felsic metavolcanics in other parts of the section. Footwall greenstone A thin (0.6- to 2.4-m thick) chlorite schist occurs at the base of the Siliceous zone and is known as the Footwall greenstone. This unit forms a distinctive marker bed within which drilling was usually terminated. STRUCTURE Abrams and McConnell (1984) and McConnell and Abrams (1984) define "Greater Atla nsta1 as the Area'', isoclinal, recumbent fol dominant and th ding. i foliation in s fabric is They indicat rocks of related e that the mtoajFo1r outcrop to open patterns, folds. however, are the expression of F2 isoclinal This are German fabric responsi bi(s1le988rfoe) lraretmecodosgt notiozuetcs rFos2p2isaposactttlheinrenadsl otmioninotahpneet S surface. n folds which west Georgia area. The predominant foliation in the prospect area is a schistosity defined by preferably oriented platy minerals (micas and chlorite). The f abric parallels compositional layering developed (bedding) fabrics and include is t ermed s minor axial 1p. l Weaker anar fol , i at i poorly on in some isoclinal fold noses and minor chevron folds whose axial fsoulrifaatcieosn/ibnetdedrisnegctvsa1r ies at a widely low angle. The strike of from west-northwest to east- northeast. Dips are almost exclusively to the south at 10 to 50. Rare dip reversals to 15 northwest were noted in underground workings. The rocks in the mine area have been multiply deformed, and although a rigorous examination of structures has not been made, certain observations can be described. The major outcrop trend in the mine area defines a zigzag fold (fig. 2), represented by plunging southeast wahnichF1 ove has rtu in r ned synform turn been -antiform deformed pair by F folds which locally folds. The synform the deflection of ptt rhoaerntsi opeonasseot -fn oSt f.rliteh eFi n1atsofto- ltdtrhe nei sd i nFgr2e prri iendstger nae tf oe djlui absyft north of the mine to a south-southwes t direct ion at Tally Mountain. The trend of this fold can also be traced by discontinuous outcrops of a banded iron formation adjacent to the Talladega Group-Hillabee Greenstone contact. The Royal Vindicator mine occurs in the upright limb of the synform. Broad north-northeast-trending trending workings, Fa4nwd aarpres are easily probably the Fs3e warp en in s and the underg northwestround mine cause of some of the northwest to northerly strikes of f oliation seen in surface outcrops. 62 F3 and F4 are similar to those recorded by German (1988). Faulting is comprised mainly of post-metamorphic, brittle, high-angle reverse faults trending north-northwest. These are best exposed in underground mine workings. No evidence is seen for the presence of the Hollins Line thrust fault as mapped in the vicinity by Heuler and Tull (1989). Core drilling through the Wedowee Group-Hillabee Greenstone contact and limited exposures of the contact in outcrop do not appear to support the presence of this fault in the Royal Vindicator mine area. METAMORPHISM Paris (1986) recognized that the Royal Vindicator gold deposit lithologies lie in the biotite zone of the greenschist facies, while German (1988) indicated that the general metamorphic grade may be as high as the staurolite grade of the amphibolite facies. However, German (1988) indicates that the metamorphic grade at the Royal Vindicator mine is no higher than garnet grade. Metamorphic grade in the immediate mine area appears to be the garnet grade of the greenschist facies. Epidote, biotite, and chlorite characterize mafic and felsic rocks of the Hillabee Greenstone sequence, and red garnets 0.1- to 2mm in diameter are common in the felsic units but are rare in the chlorite schists. Amphiboles are rarely observed, and these appear to be actinolite. Although a local mineral collector has reported staurolite in the vicinity of the Royal Vindicator mine, none was observed during this study. The overlying Wedowee Group lithologies also contain abundant biotite, chlorite, and garnet. Although no kyanite or staurolite is seen in the immediate Royal Vindicator mine area, German (1988) reports their presence in the overlying Andy Mountain Formation (Wedowee Group) rocks in other areas. Mafic units in the Hillabee Greenstone and Wedowee Group in the vicinity of the Royal Vindicator mine are predominantly chloritic schists that are locally actinolitic; rocks dominated by amphiboles are very rare. The absence of abundant amphiboles may be attributed to intense magnesian alteration of the rocks, which allowed certain chlorite species to be stable into the lower amphibolite facies. GEOCHEMISTRY Geochemistry of the Metavolcanics The metavolcanic sequence hosting the Royal Vindicator mine correlates well with the Hillabee Greenstone based on its stratigraphic position in relation to the Talladega group 63 lithologies. In addition, major element whole rock geochemical analyses of this and other studies (McConnell and Abrams, 1984; German, 1988) are very similar to those of the Hillabee Greenstone reported by Tull and others (1978) in Alabama. Lithologies seen in mine cuts and core indicate a strong bimodal distribution of mafic to felsic rock types, and a standard AFM plot (fig. 3) of 30 core analyses (Appendix 1) confirms this distribution. The chlorite schists (mafic metavolcanics) are all basaltic in composition with tholeiitic affinities, and felsic metavolcanics (quartz- sericite schist and gneiss) are dacitic to rhyolitic in composition with both tholeiitic and calc-alkaline affinities. Although alteration is definitely present in portions of the stratigraphic section (Siliceous zone and Upper gneiss) and is probably pervasive throughout the section, the AFM plot of Royal Vindicator samples shows a similar distribution for gross lithologic types as that found by Tull and others (1978) for unaltered lithologies of the Hillabee Greenstone in Alabama. This would indicate that major remobilization of most elements has either not taken place or is pervasive in all samples. Exploration Geochemistry Surficial exploration for gold in the area of the old mine and the surrounding area was conducted using soil geochemistry. Samples were collected from the B-horizon at 15.3 m intervals on north-south oriented grid lines spaced 61 m part. Samples were analyzed for Au by fire assay-atomic absorption methods and Ag, Cu, Pb, Zn, As, Mo, and Ba were analyzed by atomic absorption techniques. This sampling revealed only a localized gold anomaly in soils in the immediate pit area and two weak coincident Au-Mo+Cu anomalies to the east of the mine. In comparison, analyses of 30 selected core samples for Au, Ag, Cu, Pb, Zn, As, Mo, and Ba (Appendix 1) revealed a distinctive correlation of Au with Ag and Mo (Table 2). A subtle, but recognizable correlation of Au with Hg, As, Cu, and Pb is also present. ALTERATION AND MINERALIZATION PATTERNS Yeates and others (1896) and Jones (1909) report the gold mineralization occurs in intercalated quartz veins and schist dipping steeply to the south. Descriptions obtained from old reports and maps of companies that had worked and sampled the mine, along with unpublished notes from the Georgia Geologic Survey (Shearer, 1917) indicates that the gold mineralization is associated with a zone of shallow dipping siliceous schist with quartz veining, possibly a shear zone. Current work confirms the association of gold 64 TABLE 2 Correlation coefficients for trace metals in 30 selected core samples from the Royal Vindicator gold deposit, Haralson County, Georgia. Au Ag As Ba Cu Cr Hg Li Mo Pb Zn Au Ag 0.669 As 0.193 0.285 CD Ba -0.061 0.137 -0.054 01 Cu 0.164 0.213 0.208 -0.141 Cr -0.136 -0.118 0.471 -0.085 0.029 Hg 0.337 0.243 -0.047 0.133 -0.094 -0.231 Li -0.083 0.538 0.069 0.550 0.194 0.087 -0.032 Mo 0.885 0.521 0.056 -0.002 -0.170 -0.001 0.210 -0.129 Pb 0.349 0.766 0.397 0.329 0.356 0.040 0.265 0.660 0.256 Zn -0.213 0.186 -0.038 -0.225 0.666 -0.045 -0.143 0.304 -0.223 0.362 Felsic metavolcanic& + Sericite-quartz achiat (siliceous zone) 0 Epidote gneiaa (upper gneiaa) 0 Mafic metavolcanic& FeO MgO Figure 3. AFM plot of selected Royal Vindicator gold deposit lithologies. 66 mineralization with a siliceous felsic schist dipping to the south at a shallow angle but casts doubt on the role of shearing and quartz veining. Assays from core and reverse circulation drilling show that gold mineralization occurs predominantly in stratabound, altered and silicified felsic metavolcanics of the Siliceous zone ; however, trace amounts of gold do occur at other points in the stratigraphic section. The upper limit of the altered, gold-mineralized sequence begins at the basal contact of the Upper gneiss or Hanging wall greenstone with the Siliceous zone. Below this contact, mineralized, altered portions of the Siliceous zone felsic volcanics are generally 3- to 15-m thick. The base of the gold mineralized section is marked by a thin (2- to 10-mm thick) zone containing pyrrhotite and black-green chlorite intergrown with ironcarbonate (siderite?) lenses. Below this horizon, gold and magnetite are not present, pyrite is rare, and the degree of silicification rapidly diminishes. Gold values appear to be directly proportional to the abundance of silica, pyrite, and iron oxides. These mineral phases form the alteration pattern which distinguishes the Siliceous zone from other felsic metavolcanic units in the section. Gold values of milky quartz veins, the classic or typical ore at many gold deposits and which cross cut the Siliceous zone at various places, are only trace to nondetectable. Secondary silicification and quartz banding parallel to bedding/foliation is the major alteration effect which blurs primary volcanic textures. Accompanying the widespread alteration but less obvious is an iron enrichment consisting of iron oxides and iron sulfides. Disseminated to weakly banded zones of magnetite and hematite grains (0.5- to 2.0- mm in diameter) constitute up to 5 percent of the rock in areas of strongest silicification and gold mineralization. Pyrite is also present as equant to distorted cubes 0.5- to 20.0-mm across; as 0.5- to 20-mm thick bands of subhedral to euhedral crystals (fig. 4); as 30-cm long pods associated with silica bands (fig. 5); and as disseminated, anhedral grains 0.5- to 2.0-mrn across. Pyrite content in altered areas varies from l to 20 percent and averages 5 to 7 percent. Based on euhedral shape and pressure shadows, much of the pyrite associated with the gold was probably recrystallized during metamorphism. Examination of polished sections of "ore" and sulfide concentrates has identified chalcopyrite, galena, sphalerite, arsenopyrite, stibnite, covellite, and chalcocite as trace minerals within the Siliceous zone. Molybdenum is present at elevated concentrations associated with higher gold assays, but no molybdenite has been identified. The highly reflective phase associated with gold in composite grains may be a bismuth or telluride mineral. 67 GGHr-602 Figure 4. Photograph of prominent pyrite bands (Py) in core from the Siliceous zone~ sample is from drill hole GGHr-602. Figure 5. Photograph of pyrite pods (Py) in banded ore of the Siliceous zone on the 105-foot (32-meter) mine level. Quarter (24 mm) for scale. 68 Gold and silver are the only precious metals identified at the Royal Vindicator mine. No attempt has been made to assay for platinum group metals. Native gold is the only precious metal mineral identified to date. Silver assays appear directly related to those of gold. The silver is probably alloyed with the gold, but it may also occur in the lattice of other minerals such as galena or as a separate, unidentified mineral phase. Shearer (1917) and Maillot (1940) stated that in unoxidized ores, free milling gold comprised approximately 20 to 30 percent of the gold recovered, with the rest being contained in heavy mineral concentrates (predominantly pyrite). Gold occurs in a number of different habits at the Royal Vindicator mine: (1) native gold in quartz (fig. 6); (2) gold grains within or attached to pyrite grains (fig. 7); (3) subhedral crystals associated with magnetite and hematite (fig. 8); and (4) composite grains composed of gold, a highly reflective, silvery colored metallic mineral, and a pinkish colored metallic mineral in quartz and pyrite. Gold particles have variable shapes, occurring as ragged flakes, spindles, rounded or ovoid forms, and as blocky, subhedral crystals. Grain sizes (longest dimension) vary from 0.0005- to 2.0-mm with an average length to width ratio of 2:1. Visible gold particles in core and non-oxidized samples of the Siliceous Zone are rare. Measurement of 12 gold grains viewed in polished ore samples yield an average grain length of 0.028 mm. Measurements of a population of gold grains obtained from a heavy mineral separate and a sulfide concentrate by Cuesta Research Ltd. emphasized that gold grains 0.004- to 0.0005-mm in length, and averaging 0.001 mm in length, dominated the samples. The variety of forms of gold mineralization suggests the metal has been preserved in part, in its original habits. While there may have been some remobilization and recrystallization due to metamorphism and deformation, the distribution of gold based on core assays is restricted to a maximum of four enriched zones separated by zones of low grade. These zones maintain their equivalent stratigraphic positions regardless of the overall tenor of the deposit (fig. 9). This distribution suggests either a strata-bound, syngenetic or stratiform epigenetic origin for the goldbearing portions of the Siliceous zone. In addition to the strata-bound appearance of the gold mineralization, a further control on its localization is indicated by the distribution of gold mineralization grade to thickness of the Siliceous zone. Utilizing an isopach map of the Siliceous zone (fig. 10), ore grade mineralization 69 Figure 6. Photomicrograph of a 0.02-mm long anhedral gold grain (Au) in quartz; plane-polarized light. Sample is from a depth of 86 meters (283 feet) in drill hole GGHr-605 . . . t' - ~- '-;. -:. .,. L 4 . 'I ...., ... ' . J. ,- f I .- ~ \ \ -. Figure 7. Photomicrograph of 0. 05-mm long spindle-shaped gold grain (Au) in pyrite (Py) ; plane-polarized reflected light. Sample is from a depth of 86 meters (283 feet) in drill hole GGHr-605. 70 Figure 8. Photomicrograph of subhedral gold grains (Au) in contact with magnetite (Mag) and hematite (Hem) in quartz matrix. The larger gold grain is 0.2-mm long; planepolarized reflected light. Sample is from a depth of 85 meters (281 feet) in drill hole GGHr-605. 71 A A 0.080 ...... 1\) l 0.149 -----0.257 /( A - 2.897 0.260 0.055 0 unit D to153m l:J165m T.D. B 0.095 c D 158mT.D. L 3 meters 0 0 0.1 ozjtn Au Figure 9. Assay sections of core holes showing gold horizon distribution in the Siliceous zone (unit C) and correlation of the gold-bearing zones between drill holes. The base of the Hanging wall greenstone (unit B) is used as the reference datum plane. Other lithologies are Upper gneiss (A), Footwall greenstone (D), and Lower gneiss (E). See figure 10 for drill hole locations. 0 N 1 0............_2_0._0__..--..4.J0 Feet Meters 0 100 Isopach Interval= 5 feet 0 Figure 10. Isopach map of the Siliceous zone. Contour interval is 5 feet. Areas bounded by dashed lines represent zones of rock at least 8-feet (2.4-meters) thick containing at least o. 08 ozjton gold. Solid circles are locations of drill holes intersecting the above grades and thicknesses. Half-solid circles are locations of drill holes intersecting at least 8 feet (2.4 meters) of mineralized rock containing between 0.04 oz;ton gold and 0.08 ozjton gold. Open circles are the locations of drill holes that failed to intersect significant mineralization. 73 (>2.4 m @ 0.08 oz/tn Au) can be shown to occur along the eastern flank of a south-southeast plunging basin. This concentration of gold mineralization also corresponds with the greatest thickness of the overlying Upper gneiss unit which is considered to be an alteration zone spatially related to the gold mineralization. CONCLUSIONS The metavolcanic sequence hosting the Royal Vindicator gold deposit is correlated both stratigraphically and geochemically with the Hillabee Greenstone. In Alabama, the Hillabee Greenstone and its associated mineralization, and the underlying Talladega Group metasediments form a distinctive stratigraphic succession and topographic expression. Prominent base-metal sulfide deposits, and to a lesser extent gold deposits, occur low in the Hillabee Greenstone section. The Pyriton sulfide deposits in Clay County, Alabama occur in Hillabee Greenstone approximately 40 m above the contact with the Jemison Chert (Stow and Tull, 1979). The Crutchfield and Anna-Howe gold deposits in Cleburne County and the Dempsey gold-pyrite prospect of Clay County are both hosted in Hillabee Greenstone and are approximately 40- to 50-m above the Hillabee Greenstone- Talladega Group contact (Paris l989b, l989c). Gold prospects such as the Stewart, Jemison, and B. T. Childers and sulfide prospects such as the Garrett (fig. ll) are noted to occur low in the Hillabee section, near the contact with the Talladega Group (Paris and Cook, 1989). Many of these deposits contain significant felsic or siliceous units and are spatially associated with banded iron formation. The Talladega Group near the contact usually underlies a subdued to pronounced ridge. The Royal Vindicator deposit exhibits a similar overall stratigraphic and topographic section. The major element chemistry of the greenstones and felsic metavolcanics of the Hillabee, as analyzed by Tull and others (1978) is also similar to that of the mafic and felsic metavolcanics examined in this study at the Royal Vindicator. Tull and Stow (1982) suggest that the sulfide deposits of the Hillabee Greenstone formed in a continental margin-arc environment. An alternative interpretation suggested by this study is that the volcanic sequence formed in a failed rift near the continental margin of a back-arc basin. A similar model is proposed by German (1988). The abundance of apparent pyroclastic textures indicates the rocks were deposited in shallow marine to slightly emergent subaerial conditions. An alternative interpretation of the pyroclastic textures is that they were related to highly gas-charged submarine domes and flows; however, no discrete domes have been identified to date. It is proposed that gold mineralization at the Royal Vindicator deposit is the result of deposition of gold by 74 ..----- - - -{l CHILTON- ~ Stewart i _,_] - - - -~HAR-A-LSOJN'l [!CLEBURNE I ,..J ( ' ' i I I I I Anna -Howe, ~ Crutchfield -~ Arb~tc'oo~ heal..ollf"""",.._......,. ,s, / - \ Royai- VIndloar: J ~- ' _ j _ T-~~~ Hfc~-~roG DO!Tlp& o y l ' an JI atfon al Pyrite Copper Company I --- \ ~'\~ \ a I0 ~3~rt~- SCALE 0 10 20 30 40 Kilometers Figure 11. Location of base- and precious-metal occurrences hosted by Hillabee Greenstone or correlative units, Alabama and western Georgia. 75 hot-springs fluids along an irregular, west-facing slope on the eastern edge of what is now a narrow, south-southeast- trending basin. Deposition took place during a period of felsic volcanism in an alternating mafic-felsic volcanic sequence. Gold mineralization appears to have been deposited in three or more pulses in hot springs terraces, or in hot springs terraces and more permeable zones in immediately underlying or contemporaneously deposited felsic pyroclastics. Precious-and base-metal mineralization and associated volcanism probably took place over a relatively short time span. Following deposition of the gold-bearing horizons the deposits were buried by further volcanism which was in turn finally overwhelmed by sedimentation which formed the Wedowee Group. The deposits and enclosing rocks were subsequently multiply deformed and metamorphosed to the upper greenschist facies during the Acadian orogeny (Tull, 1982). ACKNOWLEDGEMENTS The author served as project geologist at the Royal Vindicator project for U.S. Borax from 1979 to 1987 and for the Atlantic Goldfields and Jascan Resources joint venture from 1987 to 1989. I would like to thank u.s. Borax, Atlantic Goldfields, Inc., and Jascan Resources, Inc. for permission to publish information about the deposit derived from their exploration. Gratitude is also expressed to Jerry German, Keith McConnell, Charlotte Abrams, Jim Tull, Robert B. Cook, and George Heuler for their insights into the geology of the western Georgia and eastern Alabama Piedmont and lively discussions of the geologic setting of the mine area and deposit genesis. Special appreciation is expressed to Ken Gillon for his critical review and editing which markedly improved the paper. Final interpretations are the author's responsibility. REFERENCES CITED Abrams, C.E. and McConnell, K.I., 1981, Stratigraphy of the area around the Austell-Frolona antiform, west-central Georgia, in Wigley, P.B., ed., Latest thinking on the stratigraphy of selected areas in Georgia: Georgia Geologic Survey Information Circular 54-A, p. 55-67. , 1984, Geologic setting of volcanogenic base and precious ----------~ metals deposits of the west Georgia Piedmont: A multiply deformed metavolcanic terrain: Economic Geology, Vol. 70, no. 7, pp. 1521-1539. Cressler, C.W., 1970, Geology and ground-water resources of Floyd and Polk Counties, Georgia: The Geological Survey of Georgia, Information Circular 39, 95 p. Cuesta Research Limited, 1988, Mineralogy, gold ore concentrates: Atlantic Goldfields, Inc. and Jascan Resources, Inc. Royal Vindicator Gold Mine: 76 unpublished report for A.C.A. Howe International, Ltd., 17 p. German, J.M., 1985, The geology of the northeastern portion of the Dahlonega gold belt: Georgia Geologic Survey Bulletin 100, 41 p. -------w--e-s-t--c, e1n9t8r8a,l The geology of Georgia Piedmont: gold The occurrences in the Carroll County gold belt and the southwestern portion of the Dahlonega gold belt: Georgia Geologic Survey Bulletin 107, 47 p., maps in pocket. , Stratigraphic implications from core drilling in the vicinity ------~---- of the Royal-Vindicator gold mine, Haralson County, Georgia: Georgia Geologic Survey Information Circular 84, in press. Gross, G.A., 1980, A classification of iron formations based on depositional environments: Canadian Mineralogist, Vol. 18, p. 215-222. Heuler, G.F., and Tull, J.F., 1989, Geologic mapping of the Talladega belt of Haralson County, Georgia [abs.]: Geological Society of America Abstracts with Programs, Vol. 21, no. 3, p. 22. Hurst, V.J. and Crawford, T.J., 1969, Sulfide Deposits, Coosa Valley Area, Georgia: Economic Development Administration, U.S. Department of Commerce, 190 p. Jones, S.P., 1909, Second report on the gold deposits of Georgia: Geological Survey of Georgia Bulletin 19, 283 p. Lane, C., 1900, Royal Vindicator Mines: unpublished report for the Royal Vindicator Mines Company, 3 p. Maillot, E.E., 1940, Royal Vindicator Mine, Georgia: unpublished prospectus, assay map, 14 p. McConnell, K.I., and Abrams, C.E., 1982, Geology of the New Georgia Group and associated massive sulfide and gold deposits: West-central Georgia: Guidebook in Exploration for metallic resources in the southeast symposium, University of Georgia Center for Continuing Education, Athens Georgia, p. 143-170. --- -- --G-e-o--r-g,i a 1984, Geol The ogic geology of the Survey Bulletin Greater 96, 127 Atlanta p. Region: McConnell, K.I., Abrams, C.E., and German, J.M., 1986, Gold deposits in the northern Piedmont of Georgia and Alabama and their relationship to lithostratigraphic units, in Misra, K.C. (ed.), Volcanogenic Sulfide and Precious Metal Mineralization in the Southern Appalachians: University of Tennessee Department of Geological Sciences Studies in Geology 16, p. 164-181. Neathery, T.L., and Hollister, V.F., 1984, Volcanogenic sulfide deposits in the southernmost Appalachians: Economic Geology, Vol. 79, p. 1540-1560. Neathery, T.L., and Reynolds, J.W., 1975, Geology of the Lineville East, Ofelia, Wadley North, and Mellow Valley Quadrangles, Alabama: Geological Survey of Alabama Bulletin 109, 120 p., maps. Paris, T.A., 1985, The Goldville project: Results of an exploration project for gold in the northern Alabama 77 Piedmont, in Misra, K.C. (ed.), Volcanogenic Sulfide and Precious Metal Mineralization in the Southern Appalachians: University of Tennessee Department of Geological Sciences Studies in Geology 16, p. 182-205. ------~--- ' 1986, Geology of the Royal-Vindicator gold deposit, Haralson County, Georgia [abs.]: Geological Society of America Abstracts with Programs, Vol. 18, no. 1, p. 60. ___________ , 1989a, The Royal Vindicator gold deposit: A metamorphosed epithermal hot springs type gold deposit [abs.]: Geological Society of America Abstracts with Programs, Vol. 21, no. 3, p. 53. --------~- ' 1989b, Geology of the Crutchfield and Anna-Howe gold mines, Cleburne County, Alabama, in Gold Deposits of Alabama: Geological Survey of Alabama Bulletin 136, in press. ___________ , 1989c, Geology of the Dempsey gold-pyrite prospect, Clay County, Alabama, in Gold Deposits of Alabama: Geological Survey of Alabama Bulletin 136, in press. Paris, T.A., and Cook, R.B., 1989, Exploration geology of selected gold prospects traditionally related to the Hillabee Greenstone, Alabama and Georgia, in Geological Survey of Alabama Bulletin 136, in press. -- Reeves, M.; 1975, Tallapoosa Gold Mine, in The History of Mining in Haralson County, Georgia: unpublished manuscript, West Georgia College Library Special Collections, 4 p., map. Ruhl, 0., 1917, Royal Vindicator Mine: unpublished consultants report for the Royal Vindicator Mines Company, 6 p., assay map. Shearer, H.K., 1917, Old Royal Gold Mine: unpublished field notes, Georgia Geologic Survey, 8 p. Stow, S.H. and Tull, J.F., 1979, Sulfide mineralization in Tull, J.F. and Stow, S.H., eds., The Hillabee meta- volcanic complex and associated rock sequences: Alabama Geological Society, 17th Annual Field Trip Guidebook, p. 33-36. Tull, J.F., 1982, Stratigraphic framework of the Talladega slate belt, Alabama Appalachians, in Bearce, D. N. , Black, W.W., Kish, S.A., and Tull, J.F., eds. , Tectonic Studies in the Talladega and Carolina Slate Belts, Southern Appalachian Orogen: Geological Society of America, Special Paper 191, p. 3-18. Tull, J.F., and Stow, S.H., 1980, Structural and stratigraphic setting of arc volcanism in the Talladega slate belt of Alabama, in Frey, R.~v., ed., Excursions in Southeastern Geology, -vol. 2: American Geological Institute, p. 545-581. , 1982, Geologic setting of the Hillabee meta- ------v-o~l~c-a-nics and associated strata-bound sulfide in the Appalachian Piedmont of Alabama: Economic Geology, Vol. 77, no. 2, p. 312-321. Tull, J.F., Stow, S.H., Long, L., and Hayes-Davis, B., 1978, The Hillabee Greenstone: Geochemistry, structure, 78 mineralization and theories of origin: University of Alabama Mineral Resources Institute, Research Report Number l, 100 p. West Georgia College Library, Royal Gold Mine File: a collection of letters and an unsigned report from the period 1909-1911 relating to the Royal Gold Mine, obtained by West Georgia from the University of Missouri Western Historical Manuscript Collection--Rolla, 10 p. Yeates, W.S., McCallie, S.W., and King, F.P., 1896, A preliminary report on a part of the gold deposits of Georgia: Geological Survey of Georgia, Bulletin 4-A, 542 p. 79 Appendix 1 Analyses of metavolcanic (Hillabee Greenstone) lithologies at the Royal Vindicator gold deposit. Major element data in percent; trace element data in ppm. GGHr-6D. All Total iro analyses n by asUF.Se 2. o3B orSaxampRlee numbers should be search Corporation, prefixed Anaheim, California. ------------- --- -- ----- --- ----------- -- -----------~- - ---------- - -------- Sample No. 7-160 16-100 18-150 19-90 22-255 18-265 ==================================== == = =======~=~= ====================== 6 Si0 TiO 2 AFF1ee62o33 68.60 0.37 13.40 5.10 75.20 0.23 11.70 3.70 71.00 0.33 12.70 5.10 76.10 0.25 12.10 3.40 71.70 0.33 12.70 5.10 72.10 0.25 13.20 3.10 MnO 0.14 0.06 0.13 0.07 0.14 0.07 MgO 2.50 0.81 2.10 1.20 1.20 0.83 CaO KN2a6o 1.37 5.20 0.43 1.44 5.40 0.18 1. 74 5.50 0.17 0.96 5.80 0.36 1.50 5.40 0.24 2.30 4.60 2.00 P205 L.O.I. 1.89 -- -- -- --0--.8--2 -- 1--. -7-9 1.34 1.18 ------ ------ 1.36 ------ TOTAL 99.00 99.54 100.56 101.58 99.49 99.81 Au <0.02 <0.02 <0.02 <0.02 <0.02 <0.02 Ag 0.7 0.5 0.7 0.5 0.5 0.5 As <2.0 <2.0 <2.0 <2.0 <2.0 <2.0 Ba 22.0 4.0 22.0 19.0 26.0 223.0 Cu 7.0 7.0 25.0 4.0 39.0 4.0 Cr <136.0 <136.0 <136.0 <136.0 <136.0 <136.0 Hg 0.06 <0.05 0.10 <0.05 <0.05 <0.05 Li <5.0 <5.0 <5.0 <5.0 <5.0 <5.0 Mo <5.0 <5.0 5.0 <5.0 <5.0 <5.0 Rb <9.0 <9.0 <9.0 9.0 <9.0 82.0 Pb 10.0 <5.0 10.0 <5.0 7.0 10.0 Sr 14.0 29.0 29.0 <10.0 29.0 34.0 Zn 118.0 33.0 76.0 41.0 102.0 31.0 Rock Type EFG EFG EFG EFG EFG LG EFG = chlorite-epidote-quartz-feldspar gneiss (Upper gneiss unit); LG = biotite-muscovite-quartz-feldspar gneiss (Lower gneiss unit). 80 Appendix 1 (contd.) ----------------------------------~--- -- ----------- ---- --- -- ---- ----- -- - Sample No. 22-20 22-140 7-115 18-95 22-60 22-170 ==========---===================-=-=============-=====-==================== Si0 TiO 2 6 Al Fe 2o33 3 Fe 92.50 0.03 3.30 1.65 71.90 0.25 13.40 2.60 50.20 1.00 14.50 10.90 48.20 1.00 13.60 12.30 53.40 0.87 13.80 10.30 48.40 1.60 12.70 13.90 MnO 0.02 0.04 0.28 0.28 0.16 0.24 MgO 0.73 0.68 8.30 8.30 5.80 7.10 CaO NK2a3o 0.04 1.18 0.15 2.40 4.20 2.50 8.00 3.90 0.20 11.00 1.97 0.36 8.00 2.50 0.31 10.60 2.80 0.34 P205 L.O.I. 0.69 ------- 1.22 ------- ---2-.3--5 ---2-.1--8 4.69 ------ 1.40 ------ TOTAL 100.29 99.19 99.63 99.19 99.83 99.08 Au <0.02 <0.02 <0.02 <0.02 <0.02 <0.02 Ag 0.2 0.2 0.7 1.0 1.0 0.7 As <2.0 <2.0 <2.0 <2.0 7.0 <2.0 Ba 10.0 156.0 7.0 38.0 67.0 24.0 Cu 21.0 26.0 7.0 12.0 92.0 139.0 Cr 205.0 <136.0 136.0 136.0 273.0 136.0 Hg <0.05 <0.05 <0.05 0.09 0.05 0.09 Li <5.0 <5.0 <5.0 <5.0 10.0 <5.0 Mo 8.0 <5.0 <5.0 <5.0 <5.0 <5.0 Rb <9.0 82.0 <9.0 9.0 18.0 <9.0 Pb <5.0 10.0 10.0 12.0 17.0 12.0 Sr <10.0 62.0 19.0 58.0 67.0 62.0 Zn 12.0 19.0 71.0 60.0 68.0 55.0 Rock Type UGII UGII UGST UGST UGSTII UGST UG = chlorite schist with chalcopyrite and pyrrhotite bearing epidosite clasts (Upper greenstone unit); UGSTII = chlorite schist (Upper greenstone II unit); FWG = chlorite schist (Footwall greenstone unit); HWG = chlorite schist (Hanging wall greenstone unit). 81 Appendix 1 (contd.) ----- ----- ---------- --- -- -- -- ---- --- ---- ---- ------- -- ----- ------- -- ---- Sample No. 7-280 16-120 16-175 18-255 19-180 22-335 =====================================================================c= TSii0O2 6 Al Fe 6 Fe 2o33 48 . 20 1.60 13.80 13.20 45. 10 1.00 14.90 12.40 47.40 1.30 14.70 12.90 50.50 0.85 13.80 10.00 47.30 1. 70 13.60 13.70 49.10 1.20 14.20 11.60 MnO 0.27 0.14 0.21 0.17 0.28 0.24 MgO 7.60 7.60 8.10 7.30 8.00 9.00 CaO 9.10 6.30 7.40 8.40 9.40 8.70 KN2a6o 3.40 0.26 3.00 1.64 3.50 0.58 3.90 0.11 2.80 0.37 3.20 0.16 P205 L.O.I. 2.03 6.53 3.18 3.57 2.31 ------- ------- ------- ------- ------ __ 2__.9.,.4_ TOTAL 99.46 98.61 99.27 98.60 99.46 100.34 Au <0.02 <0.02 <0.02 <0.02 <0.02 <0.02 Ag 1.0 1.9 1.4 1.0 1.0 0.7 As <2.0 <2.0 <2.0 <2.0 <2.0 <2.0 Ba 14.0 355.0 110.0 12.0 19.0 12.0 Cu 133.0 23 . 0 132.0 17.0 162.0 6.0 Cr <136.0 136.0 <136.0 136.0 <136.0 273.0 Hg <0,05 0.07 <0.05 <0.05 0.06 <0.05 Li 10.0 48.0 14.0 10.0 14.0 14.0 Mo <5.0 <5.0 5.0 <5.0 <5.0 <5.0 Rb <9.0 73.0 18.0 <9.0 9.0 <9.0 Pb 14.0 24.0 12.0 14.0 17.0 10.0 Sr 38.0 96.0 29.0 58.0 43.0 53.0 Zn 240.0 92.0 58.0 63.0 359.0 76.0 Rock Type FWG HWG FWG FWG FWG FWG FWG = chlorite schist (Footwall greenstone unit); HWG = chlorite schist (Hanging wall greenstone unit). 82 Appendix 1 (contd.) ----------------------------------------------------------------------- Sample No. 7-1A 7-16 16-1 16-12 18-2 18-12 ======================================================================= ATSiilO02623 Fet3 Fe 77.20 0.15 10.20 2.50 73.40 0.18 12.30 2.50 70.50 0.20 11.70 5.70 74.00 0.17 12.10 2.60 82.30 0.17 7.70 2.20 82.10 0.08 8.30 2.10 MnO 0.37 0.03 0.02 0.03 0.03 0.02 MgO 0.38 0 . 43 0.76 0.39 0.34 0.10 CaO 1.45 3.10 0.82 1.69 1.04 0.99 KN2a6o 5.50 0.37 4.40 1.18 6.10 0.75 4.40 1. 75 4.40 0.28 4.70 0.05 P205 L.O.I. ----1-.2--2 2.61 ------- 2.08 ------- 2.09 ------- 0.87 ------ --1--.1--4 TOTAL 99.34 100.13 98.63 99.22 99.33 99.58 Au 1.95 <0.02 2.31 <0.02 1.19 7.86 Ag 1.0 0.2 1.4 0.5 0.7 2.4 As <2.0 <2.0 5.0 <2.0 <2.0 2.0 Ba 50.0 65.0 53.0 127.0 36.0 53.0 Cu 8.0 6.0 48.0 54.0 12.0 7.0 Cr <136.0 <136.0 <136.0 <136. 0 <136.0 <136.0 Hg 0.09 0.06 0.06 0.05 0.09 0.09 Li <5.0 <5.0 10.0 <5.0 <5.0 <5.0 Mo 13.0 <5.0 5.0 <5.0 <5.0 59.0 Rb <9.0 27.0 27.0 64.0 9. 0 9.0 Pb 11.0 12.0 17.0 7.0 10.0 19.0 Sr 29.0 43.0 29.0 48.0 24.0 24.0 Zn 18.0 19.0 36.0 24.0 18.0 6.0 Rock sz sz sz sz sz sz Type sz = sericite-quartz+pyrite+feldspar schist (Siliceous zone unit). 83 Appendix 1 (contd.) ----------------------------------------------------------------------- Sample No. 18-26 19-2 19-13 19-24 22-5 22-20 ===========================================================~-========== 6 Si0 TiO 2 A1 2 3 Fet3 Fe 73.50 0.17 11.90 2.40 73.70 0.30 11.30 4.90 77.50 0.12 11.50 2.50 74.50 0.17 12.30 2.50 79.10 0.08 9.60 2.10 94.20 0.02 2.80 1.65 MnO 0.03 0.06 0.04 0.03 0.04 0.02 MgO 0.44 0.69 0.26 0.44 0.03 0.74 CaO 2.50 1.26 1.36 2.40 1.52 0.04 ~a3o LP~2Oo~I 4.40 1.25 2.49 ------- 5.80 0.82 ---- -- 4.60 1.64 0.79 ------ 4.70 1.46 2.41 ------- 5.20 0.60 1.48 ------- 1.17 0.16 0.49 ------ TOTAL 99.08 98.83 100.31 100.91 99.75 101.29 Au <0.02 1. 71 0.16 <0.02 1.17 <0.02 Ag 0.4 1.0 0.5 0.5 0.7 0.2 As . <2.0 <2.0 <2.0 <2.0 <2.0 <2.0 Ba 85.0 110.0 204.0 77.0 86.0 173.0 Cu 6.0 17.0 9.0 6.0 6.0 16.0 Cr <136.0 <136.0 <136.0 <136.0 <136.0 205.0 Hg 0,09 0.09 0.11 <0.05 <0.05 0.06 Li <5.0 10.0 <5.0 <5.0 <5.0 <5.0 Mo <5.0 <5.0 <5.0 <5.0 17.0 20.0 Rb 36.0 27.0 36.0 36.0 18.0 <9.0 Pb 8.0 10.0 10.0 <5.0 7.0 10.0 Sr 34.0 24.0 34.0 29.0 29.0 38.0 Zn 19.0 48.0 8.0 13.0 14.0 8.0 Rock sz sz sz sz sz sz Type sz = sericite-quartz+pyrite+feldspar schist (Siliceous zone unit). 84 GOLD DEPOSITS OF THE DAHLONEGA BELT, NORTHEAST GEORGIA: George V. Albino* Department of Geology, Bryn Mawr College Bryn Mawr, PA, 19010 Gold deposits of the Dahlonega Belt occur in a one to two km wide zone marked by retrograde metamorphism and well-developed near-vertical foliation and sub-horizontal stretching lineation. A number of features indicate that the gold belt is a zone of high shear strain. These include ribbon mylonites, development of s-c and C-C' fabrics, parallelism between fold axes and stretching lineations, and rare sheath folds. Kinematic indicators including s-c and C-C' relationships and assymetric quartz c- axis patterns suggest movement along the Dahlonega Belt shear zone was dextral. Southwest of Dahlonega there is evidence for a later phase of cratonward reverse faulting. Retrogressive alteration along the Dahlonega Belt resulted in formation of kyanite and staurolite at the expense of early sillimanite-bearing assemblages. Elongate, pre-kinematic and synkinematic plutons of granitic to tonalitic composition occur within the belt and along its southeast margin. Gold deposits of the Dahlonega Belt include well-defined quartz veins, zones of sheeted quartz veins, and broad zones containing lenticular quartz-carbonate-sulfide pods. Many of the auriferous zones are spatially associated with pre and syn-kinematic felsic intrusive rocks. Auriferous veins consist dominantly of quartz, with lesser calcite or dolomite, tourmaline, muscovite, biotite, hornblende, garnet and chlorite. Pyrrhotite and pyrite are both common and are accompanied by minor magnetite, chalcopyrite, galena, sphalerite, and gahnite. Gold occurs both as the native metal and in the tellurides petzite, calaverite, and sylvanite. In addition to gold and silver the ores contain low but locally anomalous concentrations of arsenic, antimony, tungsten, molybdenum, and mercury. Alteration assemblages vary with host rock 1ithology, with carbonatization and biotitization dominant in metabasites, and base leaching in quartzofeldspathic rocks. Mineral assemblages and compositions indicate alteration at temperatures >500 C and at pressures of >4.5 to 5 kb accompanying infiltration of a fluid with Xc02 of l0.2. INTRODUCTION The Dahlonega Gold Belt in the Eastern Blue Ridge of northeast Georgia (Figure 1) has been one of the most important gold-producing areas of the southeastern U.S .. It probably accounted for the bulk of Georgia's production of more than 870,000 ounces, from first discovery of go1d 1n 1829 until production essent i a11 y ceased in the 1ate 1930's *-Present address, Corona Gold Inc., 940 Matley Lane, 15, Reno, NV, 89502 85 AI ' I'' ,> .I',I ', '- -- --- I '"----- - - -~J -- ~ .1..0.0.k..m. Figure 1. Location map. showing subtUvlslons of the Southern Appalachians (Blue Ridge in dash pattern). Approzlmate location of Dahlonega Belt indicated by arrow. 86 (Koschmann and Bergendahl, 1968). Most gold produced from the Dahlonega Belt has come from placers and hydraulic mining of saprolite (Pardee and Park, 1948), with little gold recovered from primary unoxidized ores. Although its total production is not large by world standards, the Dahlonega Belt is of considerable scientific interest as a nearly unique example of a significant gold belt hosted by high grade metamorphic rocks. The purpose of the present paper is to describe the geological setting of the gold occurrences of the Dahlonega Belt and the nature of the auriferous zones and their associated alteration assemblages, and to discuss the conditions under which mineralization occurred and the nature of the ore-fermi ng process. The genetic re 1at ionshi p between the Dahlonega Belt occurrences and the more important group of deposits hosted by greenschist facies metamorphic rocks (e.g. Kerrich, 1983, Phillips, 1985, Weir and Kerrick, 1987) will also be discussed. REGIONAL GEOLOGY The Dahlonega Belt is in the Eastern Blue Ridge of the Appalachian Orogen. The Eastern Blue Ridge is separated from the Piedmont Province to the east by the Brevard Zone, and from the Western Blue Ridge, which consists mainly of the Late Precambrian to cambrian Ocoee and Great Smoky Groups, by the Hayesville Thrust (Figure 2). The Eastern Blue Ridge is dominantly a region of high grade metamorphic rocks, probably mostly of Late Precambrian to Early Paleozoic age, which were metamorphosed and thrust eastward over the North American Craton during the Taconic Orogeny (Hatcher, 1978, 1987). In northeast Georgia the Eastern Blue Ridge is divided into three belts, from southeast to northwest, the Tallulah Falls, Dahlonega, and Richard Russell Belts (Figure 2). The Tallulah Falls and Richard Russell Belts are very similar in terms of rock types present, their relative proportions, metamorphic grade, and structural style. They probably at one time were parts of a single continuous thrust sheet (the Hayesville Thrust Sheet) consisting mainly of coarse-grained, weakly foliated, banded psammitic gneiss with lesser amounts of pelitic schist and gneiss, am- phibolite, and minor ultramafic rocks and aluminous schist. Psammitic and pelitic rocks are migmatitic (Figure 3A), and contain granitic (K-feldspar rich) and trondhjemitoid (plagioclase-rich) leucosome. Dioritic to granitic intrusive rocks are common in the Tallulah Falls Belt along its contact with the Dahlonega Belt, but are largely absent in the Richard Russell Belt. Mafic rocks are minor in the Richard Russell and Tallulah Falls Belts, except for the mafic to ultramafic lake Burton Complex, consisting mainly of amphibolite and metagabbro (Hopson, 1988), which partly rings the Tallullah Falls Dome in the Tallulah Falls Belt. Its position flanking the core of the dome indicates it is structurally lower than most of the exposed metasediments. Metamorphic grade in the Tallullah Falls and Richard Russell Belts is upper amphibolite (Bryant and Reed, 1979; Carpenter, 1970). Migmatitic metasediments contain assemb1ages diagnostic of the upper kyan ite or sillimanite zone with either aluminosilicate+melt or aluminosilicate+K- feldspar in both belts. Along strike to the northeast, in rocks equivalent to the Richard Russell Belt, 'granulite facies assemblages occur suggesting peak metamorphic temperatures of at least 750, and pressures 87 --- -{f- -- - .. ......,i' .f NC .,.,.... -,.,.... : -G-A 1/) ------- , , -:",, :'_,~~::t"' ~ ~ --"" ,..... : ,.. I RRB ~ : ~H .:' - -- Je>'~............ : - - .,.......-. Q.V0. _..-- TFB -, D : ....... -- - - - - -- ~ _: --......... ----: .-- ---- ......~_...,..... --- - _... -- - ........ ..... ... ... .. : '~sc GA,_. ' '- -~~---------- -10 km Figure 2. General geologic map of the Eastem Blue Ridge ln northeast Georgia. DB = Dahlonega Belt; RRB = Richard Russell Belt; TFB = Tallulah Falls = = Belt; GSTS = Great Smoky Thrust Sheet; HF Hayesville Fault; BZ = Brevard Zone, and; MS Murphy Syncline. Also shown for reference are the towns of Dahlonega (D) and Helen (H). Modified after Nelson (1985). 88 Figure 3. Rocks of the Dahlonega Belt and Oanklng belts. A, Coarse- grained mlgmatitic quartz-biotite-plagioclase gneiss from the Tallulah Falls Belt. B, Psammitic schist from DB near Helen showing "pinstripe" tenure defined by biotite laminae. C, Quartz-ribbon textures In psammitic schist from DB east of Helen. Rock consists of polycrystalllne quartz ribbons in a much finer-grained quartz-plagloclase-biotlte matrix. D, Medlwn-gralned, well-foliated banded amphibolite from CreightonFranklin Mine. Hornblende occurs as oriented needle-like grains In dark, amphibole-rich layers. E, Amphibolite from southeast margin of Dahlonega Belt near Lake Allatoona. Reduction of grain size and production of a penetrative schistosity is apparent. 89 of 6.5 to 7.0 kb (Absher and McSween, 1985). The ent1re sequence of rocks of the Eastern B1ue Ridge has been thrust over autochthonous Precambrian strata of the North American miogeocline, as evidenced by the presence of windows through the high grade crystalline rocks of the Hayesville thrust sheet, in which Great Smoky Group strata are exposed (Nelson, 1985). Movement along the Hayesville Thrust is generally considered to have been pre-metamorphic, but there must have been some later displacement either along the Hayesville Fault or some unrecognized parallel structure (see below). This movement probably occurred during Alleghanian thrusting (Albino, 1987; Vauchez and Dallmeyer, 1989). In northeast Georgia the Hayesville Thrust Sheet appears to be a composite nappe consisting of a stack of thinner thrust sheets. Evidence in support of this interpretation includes inverted thermal gradients in several areas in the Richard Russell Belt and its extension into North Carolina (Albino, 1987; Tso, 1987; Kittelson and McSween, 1987) and the occurrence of thin fault - bouned slivers of basement gneiss exposed around the margin of the Tallulah Falls Dome (Stieve et al., 1988). The pattern of metamorphic isograds around the Tallullah Falls Dome, with lower kyanite zone succeeded outward by upper kyanite and upper sillimanite zones (Hatcher, 1973) , is indicative of an inverted metamorphic gradient, providing another piece of evidence supporting correlation of the Richard Russell and Tallulah Falls Belts. THE DAHLONEGA BELT The Dah1onega Be1t is def 1ned by: 1) its straight trend, and the parallelism between its boundaries and orientation of foliation and lithologic contacts within the belt; 2) better developed foliation and lineation, as compared to the flanking rocks; 3) low metamorphic grade compared to the flanking belts; 4) occurrence of felsic plutons and dikes, as well as felsic gneisses within the belt and along its southeast margin; 5) occurrence of abundant amphibolite, and; 6) occurrence of numerous gold deposits and a small number of massive sulfide bodies. The Dahlonega Belt was considered a zone of shearing by early workers in the area (Jones, 1909; Crickmay, 1952; Furcron and Teague , 1945), but has more recently been described as a largely intact stratigraphic sequence (Hatcher, 1979; McConnell, 1980; McConnell and Abrams, 1983, 1984, 1988; German, 1985, 1988), or as a pile of thrust sheets, including melange units (Higgins et al., 1984, 1988). Rock Types Rock types in the Dahlonega Belt are similar to those in flanking areas, dominantly psammitic gneiss, micaceous schist and gneiss, amphibolite, and lesser amounts of ultramafic rock. The greater proportion of amphibolite in the Dahlonega Belt, as well as the occurrence of felsic igneous rocks are points of dissimilarity. Small amounts of Faand Mn-rich metasediments are also locally abundant in the Dahlonega Belt (McConnell and Abrams, 1983, 1984; German, 1985, 1988). Psammitic gneiss or semi-schist is the most abundant rock type. It is generally fine- to medium-grained and consists mainly of quartz, 90 plagioclase (An20 ), and biotite, with variable amounts of muscovite, garnet, clinozoisite, ilmenite, and locally magnetite or pyrrhotite. It has a well-developed foliation defined by the orientation of biotite flakes, thin mica seams, and millimeter- to centimeter-scale compositional banding (Figure 3B). Folding of this fabric is common on scales ranging from a few centimeters to several tens of centimeters. The folds are typically isoclinal, and some 'eye' patterns typical of sheath folds have been observed. Muscovite is quite variable in abundance, with more muscovite in gneisses from the southeastern part of the Dahlonega Belt (note that muscovite also increases along strike to the southeast as will be discussed below). In many samples muscovite is obviously late-formed, as it is partly pseudomorphous after biotite or occurs as unoriented poikiloblastic flakes in strongly foliated gneiss. Quartz occurs both as small xenoblastic grains and as large ribbon-like aggregates in a finer- grained quartz-plagioclase-biotite matrix (Figure 3C). The polycrystall ine ribbons correspond to Type 2A of Bouiller and Boucher (1978). Garnets are typically well-formed, and are euhedral against biotite; unusual discoidal garnets, which are much thinner perpendicular to foliation than they are wide, are locally common. Garnets of the Dahlonega Belt are richer in cao and MnO than those in the flanking belts (Albino, unpub. data). Mica schist is coarser-grained than psammitic rocks, and is essentially a quartz-biotite rock with minor plagioclase and muscovite. Locally muscovite is dominant, and garnet, staurolite and/or kyanite may be present. Muscovite occurs mostly as thin blades later than, and partly replacing biotite laths. Aluminosilcate minerals occur as porphyroblasts or poi k11 ob1asts, and 1oca 11 y reach sizes of up to 3-4 em. Rocks with abundant aluminosilicates in some instances contain small amounts of dark blue-green hornblende as ragged, diffuse poikiloblasts. The mica schists are well-foliated with fabric defined by orientation of the mica sheets and locally by fine-scale compositional banding. Folding and crumpling of the rna in fabric is common, as are shear bands. Garnet, and 1ess commonly staurolite, poikiloblasts commonly have internal inclusions in trails discordant to the foliation external to the porphyroblasts suggesting syn-metamorphic deformation as desribed elsewhere by Bell et al. (1986). Amphibolites are abundant in the Dahlonega Belt. They occur as persistent units hundreds of metres thick, as pods with strike lengths of metres to a few hundred metres, or as bands, as thin as a few em, in micaceous or psammitic metasediments (McConnell, 1980; Gillon, 1982). Amphibol ites are locally banded, green to blue-green hornblende, and and consist variabl o e f plagio amounts cla of sqeu(aArntz20, _ 3i5l)m, ednairtke and clinozoisite. They are typically well foliated, and possess a strong lineation defined by long axes of hornblende prisms (Figure 3D). Polycrystalline quartz ribbons with strong crystallographic preferred orientations are common in quartz-rich varieties. Coarse-grained granoblastic amphibolite occurs in small pods throughout the belt, and locally the more typical, medium-grained foliated amphibolite may be seen to have formed through shearing of these rocks (Figure 3E). Small bodies of metamorphosed ultrmafic rock occur adjacent to some amphibolites (e.g. Gillon, 1982), and locally with aluminous schist zones in the metasedimentary sequence. These are either olivine- or anthophyllite- and chlorite-rich. The different mineral assemblages probably represent primary chemical differences, with anthophyllite-rich 91. rocks derived from pyroxenites. Felsic to intermediate igneous rocks are an important component of the Dahlonega Belt, and plutonic rocks identical to those in the Dahlonega Belt occur in the Tallulah Falls Belt along its boundary with the Dahlonega Belt. There are two general types of felsic rocks in the Dahlonega Belt, felsic gneisses and massive to foliated cross-cutting i ntrus iva bodies. Fe1sic gneisses are pervas iva 1y foliated, strongly recrystallized and contain metamorphic minerals such as garnet and hornblende similar to those found in adjacent rocks; such minerals are absent in the cross-cutting intrusive rocks, which, although commonly foliated, are mineralogically similar to typical plutonic rocks. Felsic gneiss occurs mostly as discontinuous lenses along linear belts as much as 20 km in length, with individual lenses from a few metres to > 100m thick, and with strike lengths of tens of metres to a few kilometres. The Barlow Gneiss (German, 1988) occurs principally near Dahlonega, and contains ovoid, 3-6 mm long blue quartz 'eyes' and plagioclase porphyroclasts in a matrix of quartz, plagioclase, biotite, and minor clinozoisite and sphene or ilmenite (Figure 4A). Similar rocks with blue quartz eyes occur at the Charles Mine in Forsyth County. The Galt's Ferry Gneiss (McConnell, 1980; McConnell and Abrams, 1984), occurs in the southwest part of the Dahlonega Belt. Quartz eyes are not common, and the rocks consist of granob1ast i c p1agi oc 1ase and quartz, with euhedral garnet crystals and radiating sheaf-like aggregates of hornblende along foliation planes (Figure 4B). Clinozoisite is locally abundant, and occurs as monominerallic patches as large as 10 em, as well as disseminated throughout the rock. The Galt's Ferry Gneiss also differs from the generally homogeneous Barlow Gneiss in the common presence of compositional banding, defined by variation in hornblende content. Both the Barlow and Galt's Ferry Gneisses are well-foliated to locally mylonitic in appearance. Where probable primary igneous textures are preserved, the felsic gneisses are typically medium to coarse-grained hypidiomorphic-granular rocks, similar to typical granitic intrusive rocks. The strong foliation in the felsic gneisses, combined with the abundance of typical metamorphic minerals in the Galt's Ferry Gneiss indicates that these rocks were part of the Dahlonega Belt during deformation and metamorphism. This has led McConnell and Abrams (1983, 1984) and German (1988a, 1988b) to suggest that they are of felsic flows or sub-volcanic intrusions. On the other hand, the petrographic and geochemical similarity between the Galt's Ferry Gneiss and the Cavender Creek Pluton (Cook et al., 1984) near Dahlonega (Albino, unpub. data), which, although recrystallized and containing metamorphic hornblende and garnet, clearly truncates foliation at its contact with the Dahlonega Belt metasediments, argues for a syn-tectonic intrusive origin for the Galt's Ferry Gneiss. As well, the Barlow Gneiss at the Barlow Mine, its type locality, truncates foliation in amphibolite (Pardee and Park, 1948), demonstrating the intrusive origin of this unit. In addition to the felsic gneisses there are several varieties of intrusive rocks in the Dahlonega Belt. The least felsic variety is diorite to granodiorite which occurs along the southeast margin of the Dahlonega Belt in close association with two-mica granite containing sparse magmatic garnet. It is medium-grained, generally weakly foliated, and consists of 2 to 3 mm subhedral to euhedral plagioclase in a matrix of biotite, quartz, plagioclase orthoclase and minor sphene, apatite and zircon. Associated granites are white, very leucocratic and range from 92 b d Figure 4. Igneous rocks of the Dahlonega Belt. A. Barlow Gnelu from small branch off Cane Creek southwest of Dahlonega. Blue quartz eyes are In plagioclase-quartz-biotite groundmass. B. Galt's Ferry Gneiss. Lake Allatoona. Rock consists of dark grey rlbbon-llke quartz grains In pale grey to white plagioclase-quartz groundmass. Abundant coarse-grained black hornblende occurs In sprays and rosettes oriented parallel to follatlon. c. Coarse-grained. weakly follated trondhjemite from Yahoola Creek near Dahlonega Consolldated Mine. D. Albite and biotite-porphyritic trondhjemite from dike at Dahlonega Consolldated Mine. Sample Is sllghtly deformed. 93 weakly foliated, southeast of the Dahlonega Belt, to quartz ribbon mylonite within the belt. Granite consists of quartz, 1 to 2 em microcline phenocrysts, plagioclase and muscovite. Biotite occurs locally, and pink to red garnet is present in trace amounts. Southeast of the Dahlonega Belt the granite is in very elongate bodies as much as 20 to 30 km in length, oriented parallel to the Dahlonega Belt-Tallulah Falls belt contact. These bodies are variably referred to as Hightower Granite (Crickmay, 1952), Rabun Gneiss (Hatcher, 1974), or Yonah Gneiss (Gillon, 1982). Within the Dahlonega Belt granite is as thin, discontinuous bands of strongly deformed to mylonitic granite near the SE margin of the belt, and as irregular pods of weakly-deformed, coarse- grained to pegmatitic granite throughout the Dahlonega Belt. As a result of its coarse-grained nature and the abundant microcl ine phenocrysts, kinematic indicators such as S/C relationships and assymmetric tails like those described elsewhere by Simpson and Schmid (1983 and Passchier and Simpson (1986) are well-developed in the more deformed bodies. Chemically, these high K20 and K/Na, rocks and are low charact to very erized by low conte only moderate nts of high fi Si0 eld 2 Ci70%), strength elements such as zirconium, niobium and yttrium (Albino, unpub. data). The third variety of intrusive rock in the Dahlonega Belt was observed only in the valley of Yahoola Creek near the Dahlonega- Consolidated property. It is a coarse-grained, equigranular biotite- muscovite trondhjemite to granodiorite. It consists of abundant sodic plagioclase, quartz, biotite and muscovite. Microcline is present locally (Figure 4C). The muscovite is at least partly secondary, as it partially pseudomorphs biotite. Trondhjemite crops out in two irregular oval bodies which are not elongate parallel to the Dahlonega Belt trend. This, combined with the generally weakly foliated nature of the rocks, suggests that it was intruded fairly late in the deformational history of the Dahlonega Belt. Locally, however, it is intensely foliated, and grades into fine-grained quartz-muscovite-pyrite schist. Chemically the trondhjemite is similar to the granite described above in terms of Si02 content and abundance of HFS elements, but has much higher Na/K and contains less barium and rubidium, and more strontium than the granite (Albino, unpub. data). The fourth variety of intrusive rock in the Dahlonega Belt is fine- grained plagioclase porphyritic trondhjemite, chemically similar to coarse-grained trondhjemite (Figure 4D). Porphyritic trondhjemite occurs as dikes ranging from tens of em's tCl 2-3 metres in width, mostly at the Dahlonega Consolidated property along Yahoola Creek, but also in a single undeformed dike a few metres wide in Rabun County at the northeast end of the gold belt (Hatcher, 1974). Where unaltered this rock consists of albite to oligoclase phenocrysts in a fine-grained felsitic groundmass of quartz, albite, biotite, and minor sphene. The dikes at the Dahlonega- Consolidated Mine have been folded, are foliated, and locally grade into fine-grained muscovite-albite-quartzpyrite phyllonite. Compositionally the porphyritic trondhjemite is similar to coarse-grained trondhjemite, differing mainly in higher differentiated equivalent of Si0 the 2c, o suggesting arse-grained that rock. they may be a more Titanium, zirconium and yttrium are low in both varieties of trondhjemite, while rubidium, barium, and strontium are also present in slightly greater amounts. The combination of spatial relationship, similar timing with respect to gold mineralization, and their similar compositions suggest that the fine- 94 grained porphyritic and coarse-grained equigranular trondhjemites are part of the same igneous suite. Structural Geology of the Dahlonega Belt The Dahlonega Belt has a well-developed foliation and lineation, and the linear trend of this fabric (Figure 5) is very different from the situation in flanking belts, in which foliation trends are very variable, and complex interference patterns occur on both outcrop and regional scales (Figure 5). The planar fabric within the Dahlonega Belt is defined by compositional banding on a fine (mm) to megascopic (map unit) scale and by the preferred orientation of micas and amphiboles. The sub-horizontal lineation typical of the belt is defined by the long axes of prismatic hornblende in amphibolite, shapes of elongate mica aggregates in metasediments, and orientations of fold axes. The foliation trends northeast along the belt, and northeast of Dahlonega dips very steeply to the NW or is sub-vertical. In the area around and southwest of Dahlonega the dip of the fabric changes to moderate (to steep) to the southeast. In the same area marked by the change in attitude of foliation there is a distinct widening of the belt, there is considerable folding of flexural-slip type of the Dahlonega Belt fabric, and there are more numerous zones of very fine-grained mica-rich phyllonite. This change in attitude of foliation corresponds to a change in strike of the Hayesville Fault and where the trace of the Hayesville Fault merges with that of the Dahlonega Belt (Figure 2). The distinctive fabric of the Dahlonega Belt formed through rotation of pre-existing fabrics and pressure solution of quartz and feldspar in the northwest part of the belt, and by recrystallization during ductile shearing, paticularly in the southeast part of the belt. This last mechanism was responsible for mylonitic textures, including spectacular coarse ribbons in quartz-rich amphibolites (Figure 6A). The occurrence of parallel stretching lineations and fold axes, as well as rare sheath folds (Figure 6B) are interpreted as evidence of high shear strain (Bell, 1978; Escher and Watterson, 1974; Cobbold and Quinquis, 1980), with the sub-horizontal orientation of lineations suggesting movement was essentially strike-slip in nature. There are several types of features which give an indication of the sense of shearing across the Dahlonega Belt, including S/C fabrics in granitic rocks, asymmetric quartz c-axis patterns in ribbon mylonites, 'snowball '-type garnets (Figure 7A), and local shear bands in micaceous schists (Figure 7B). All indicators suggest movement along the Dahlonega Belt was dextral. The large size and lack of subgrains or other evidence of incomplete recovery in quartz ribbons indicate that mylonitization took place under relatively high grade metamorphic conditions (Bouchez, 1987). This conclusion is supported by the occurrence of smal grains of other minerals in the ribbons, indicating grain growth and the occurrence of maxima in the quartz c-axis patterns attributable to prism slip (Bouiller and Bouchz, 1978) . Well-preserved quartz c-axis patterns suggest shearing was syn- or post-peak metamorphic, as the patterns would not likely have survived prograde metamorphism. Evidence of flattening across the belt includes widespread foliation boudinage (Platt and Vissers, 1980), common 'discoidal' garnets with short dimensions perpendicular to foliation (Figure 7C), and abundant mica 95 20 km Figure IS. Trend-surface map of the Dahlonega Belt and Oanktng belts. The Dahlonega Belt Ia outlined by the striped line; curving lines represent strike of dominant planar fabric. SoUd squares Indicate location of gold deposits and occurrences; locations of Canton (C), Dahlonega (D), and Helen (H) areshown for reference. Structural data compUed from numerous sources, locations of gold deposits from German (1985). 96 Figure 6. Features lnctlcatlve of shearing In the Dahlonega Belt. A, Ribbon mylonite, southwest of Lake Burton. Very large, continuous planar quartz ribbons are In matrix of ftne-gralned cllnozolslte and quartz. B, Sheath fold In quartz veined ampblboUte near Canton. A:r.ls of fold ls sub-horizontal, suggesUng strlke-sllp movement along DB. 97 Figure 7. Deformational features of the Dahlonega Belt. A, Rotated garnets In fine-grained psammltlc gneiss, Turkey Hill Mine. Nature of Inclusion tra11a In garnets and asymmetry of quartz pressure shadows Indicate dextral shear. B, Psammltlc schist, Wildcat Creek near Lake Burton. Shear bands (SB) are well-developed (Note sUde Is reversed &om orientation In field, so actual sense of shearing ls dextral). C, Discoidal garnet typical of psammltlc gneisses of Dahlonega Belt, Turkey Hill Mine. Such textures are the result of shortening across the belt. 98 seams, formed by pressure solution of quartz and feldspar, giving rise to the common pinstripe nature of the psammitic metasediments. This combination of shearing parallel to, and shortening across the Dahlonega Belt indicates that deformation probably occurred in a transpressive environment as defined by Sanderson and Marchini (1984). Metamorphism of the Dahlonega Belt The Dah 1onega Be1t has traditiona11 y been regarded as a be1t of relatively low grade metamorphism (kyanite or staurolite zone) surrounded by rocks of the higher grade (sillimanite zone) Tallullah Falls and Richard Russell Belts. Sillimanite does, however, occur in incompletely retrogressed pelitic schist from several localities in the Dahlonega Belt (Gillon, 1982, this study) with muscovite and quartz and in the assemblage sillimanite-almandine-biotite, which suggests a temperature of 525 to 550 (Figure 8). This early, highest temperature assemblage was succeeded by kyanite-almandine-biotite and staurolite- garnet-biotite. This sequence of assemblages suggests that the retrograde P- T path was either isobaric or followed a cooling path more or less parallel to a normal orogenic geothermal gradient (Figure 8), indicating retrogression was not related to uplift and decompression . In the latest stages of retrograde metamorphism, staurolite was partly replaced by muscovite, and locally abundant pale green chlorite crystallized in the pressure shadows around garnets. Pressure and temperature during early stages of retrograde alteration were more than 5 kb and 500 to >550C based on the garnet- biotite thermometer of Thompson (1976) and Ghent's (1976) garnet- plagioclase geobarometer (Albino, unpub. data). This pressure represents conditions during the early stages of retrogression when kyanite+garnet were stable. Retrograde alteration in rocks flanking the Dahlonega Belt is manifested by the appearance of muscovite, which was not part of the stable assemblage during peak metamorphism. Muscovite growth was in part at the expense of garnet, and garnets in the Richard Russell Belt are commonly partly surrounded by symplectic muscovite-biotite-quartz inter- growths. Retrograde temperatures of 525-575C have been obtained for these samples (Albino, unpub. data) using the garnet-biotite geothermometer of Thompson (1976). Retrograde alteration of amphiholites is pervasive, and resulted in the formation of c1inozoisite- bear1lng assemblages, in contrast to the upper amphibolite grade metabasites of the flanking belts which contain very calcic stages of plagioclase (up retrogression to are Anc7o0m+)maonndly no epidote-group minerals. Later associated with formation of carbonate-rich assemblages, and is largely the result of infiltration of C02-rich fluids. The rare ultramafic rocks of the Dahlonega Belt also provide some important data on the temperature history of the area. The occurrence of anthophyllite in these rocks indicates that they were at one time subjected to temperatures of 600-650 or higher (Evans and Trommsdorrf 1970). The anthophyllite has been largely replaced by mixtures of talc, serpentine after metamorphic olivine, and calcic amphibole. The existence of the early anthophyllite, and the high temperatures it implies, are important evidence of an early stage of high grade metamorphism in the Dahlonega Belt. In the part of the Dahlonega Belt southwest of 99 8 6 p And 2 500 1 Ms-Ab.Q=ASL 2 .'Glii1:.s:MiTJ. 3 BiMsQ=GtKfsV 600 700 800 T Figure 8. Pressure-temperature grid showing locations of reactions relevant to prograde and retrograde metamorphism In the Dahlonega Belt and Oanldng belts. Boxes indicate peak and retrograde metamorphic conditions In the Richard Russell and Tallulah Falls Belts calculated from mineral compositions and Inferred from assemblages; arrow represents Inferred path of retrogression In the Dahlonega Belt based on mineral compositions and assemblages. 100 Dahlonega the degree of retrogade metamorphism increases, and there are zones of very fine-grained, intensely foliated phyllonite, which have mineral assemblages typical of greenschist to lowest amphibolite grade metamorphism, such as muscovite-chlorite-biotite and hornblende-epidotealbite. The phyllonitic rocks with these low grade assemblages sometimes contain shear sense 1ndi cators demonstrating northwest-directed thrusting, contemporaneous with retrogression. This low grade metamorphic event may be Alleghanian, and based on field relationships is later than most or all of the gold deposits of the Dahlonega Belt. GOLD DEPOSITS OF THE DAHLONEGA BELT Gold deposits of the Dahlonega Belt include: 1) broad zones of very low-grade ore; 2) medium to narrow, generally tabular pyritic quartz veins or sets of sheeted veins; 3) sets of thin sheeted quartz veins in trondhjemite dikes, and; 4) narrow high grade quartz veins containing abundant visible gold. Detailed descriptions of many individual properties are given by Pardee and Park (1948), Jones (1909), and Yeates et a1. ( 1896). Deposit of the first type could generally be mined only where deep weathering had resulted in formation of thick saprolite horizons amenable to hydraulic mining. These deposits have, however, been a significant source of go 1d in the Dah 1onega Be1t, as they inc 1ude the deposits of Findley Ridge and the Barlow Cut near Dahlonega. Gold occurs in generally tabular zones containing abundant pod-like quartz veins (Figure 9A), containing a few percent pyrite and traces of other sulfides, in a wide zone of mica-, pyrite/pyrrhotite-, and locally carbonate-rich (altered) rocks. Gold content of unweathered rocks are low, typically less than 100 ppb. This low primary grade argues that supergene migration of gold, as well as freeing of refractory gold during sulfide breakdown, contributed to the formation of these deposits (Lesure, 1971). The second type of deposit, in which gold occurs in tabular quartz veins (Figure 9B) or zones of tabular sheeted veins containing considerable pyrite or arsenopyrite, has also been a major source of gold in the Dahlonega Belt, as it includes the Creighton-Franklin, Cherokee and several other important producers in the area. Ore shoots within these veins tend to be steeply plunging rods with plunge length much greater than strike length, in which the veins are thicker and richer than average (Pardee and Park, 1948). Gold grades in thse veins are moderate, ranging from a few ppm to as much as 30-40 ppm, greater abundances occurring in sulfide-rich parts of veins. In some deposits, notably the CreightonFranklin Mine, sulfide-bearing wallrock adjacent to the veins contains gold in possibly economic amounts, 2-5 ppm (Table 1). The third type of deposit, thin sheeted quartz veins in dikes, is at the Dahlonega-Consolidated Mine, where the Knight 'Vein', an important source of ore from both surface and underground workings, is actually an intensely foliated, 2-3 m wide albite porphyritic dike, which contains abundant 3 mm to 2 em wide glassy quartz veins parallel to the dike contact. The veins contain a few percent 1-2 mm pyrite cubes (Figure 9C) which are closely associated with gold and precious metal tellurides. Auriferous quartz veins occur in coarse-grained and fine-grained porphyritic trondhjemite at the Dahlonega Consolidated Mine (Lindgren, 1906), flanked by zones in which biotite has been replaced by muscovite. 101 Table 1. Abundances of selected ore metals in altered and mineralized samples from the Dahlonega Belt. SAMPLE BB-4 BB-5 BB-9 BY-7 CF-2 CF-3 CF-6 CF-12 CF-16 CF-20 CF-23 CF-24 CH-1 CH-2-109 CH-2-123 CH-2-130 CH-2-153 CH-2-169 CH-2-225 CH-3-83 CH-3-120 CH-3-155 CH-3-174 CH-3-182 CH-3-281 DC-1A DC-2 DC-3 DC-4 DC-10 DC-11 DC-105 DC-108 JMV-1 JSV-5 JSV-7 JSV-8 JSV-9 RBT-2 TH-2 TH-4 WC-3 WC-9 WC-14 WC-16 Au 0.035 <0.06 0.021 17.000 7.500 0.270 0.022 1. 100 0.089 0.010 0.720 0.012 0.060 0.037 <0.005 <0.005 0.021 <0.006 0.029 0.008 0.041 0.012 0.007 0.022 <0.008 11.000 0.420 1.900 4.100 0.200 0.810 0.016 0.025 0.410 0.008 <0.005 <0.005 <0.005 <0.005 <0.005 6.400 <0.005 <0.005 0.005 <0.005 Ag As Sb Hg <5 <2 0.2 0.018 <5 <2 0.3 <5 <2 <0.2 10 <2 <0.2 0.007 <5 <2 <.2 0,007 <5 <2 <0.2 <5 <2 0.2 <5 7 <0.2 0.005 <5 <2 <0.2 <5 <2 <0.2 <5 <2 <0.2 <5 <2 <0.2 14 9200 3.8 0.028 <5 410 0.3 <5 33 <0.2 <5 7 0.2 <5 8 <0.2 <5 310 0.3 26 9300 2.9 0.007 <5 65 0.5 63 34 1.1 0.021 <5 12 <0.2 <5 21 <0.2 <5 1400 0.5 <0.005 <5 2600 1 0.005 7 2 0.2 <0.005 <5 <2 0.2 <5 <2 0.2 0.006 <5 <2 0.2 <5 <2 <0.2 <5 <2 0.3 <5 <2 0.2 0.005 <5 <2 0.3 0.005 <5 <2 <0.2 <0.005 <5 <2 0.3 <5 2 0.2 <5 <2 0.2 <5 <2 0.2 <5 <2 0.2 <5 5 <0.2 <5 170 <0.2 0.050 <5 <2 0.2 <5 <2 0.3 <5 <2 0.3 <0.005 <5 <2 <0.2 w Mo <4 12 <4 <5 <4 <5 <4 <5 16 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 5 5 <4 <5 <4 <5 <4 11 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 6 <4 <5 <4 5 <4 <5 <4 7 5 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 <5 <4 6 <4 <5 <4 <5 <4 <5 <4 <5 All abundances reported in parts per million. Analyses by X-ray Assay Laboratories, Don Mills, Ontario. Au, Ag, As, Sb, W, and Mo analysed by INAA; Te determined by flameless AAS, and Hg by graphite furnace AAS (-not analyzed). BB = Battle Branch, CF = Creighton-Franklin, CH = Charles, DC = Dahlonega-Consolidated, JMV and JSV = Jones Vein/Lot 10, RBT = outcrops outside Robertstown, TH =Turkey Hill, WC =Wildcat Creek 102 Thin tongues of intrusive rock cut auriferous quartz veins at the White County (Pardee and Park, 1948) and the Dahlonega-Consolidated Mines, suggesting mi nera1i zat ion occurred assent i a11 y synchronous with plutonism. The fourth deposit type consists of thin, high grade quartz veins containing considerable visible gold, sometimes as the only ore mineral, but also commonly with abundant galena or arsenopyrite; they commonly cross-cut schistosity of enclosing rocks. They are mostly in the vicinity of Dahlonega and include some veins of the Calhoun, Turkey Hi 11, and Battle Branch Mines and the Loud Mine in White County. These veins commonly are close to deposits of the first three types. Auriferous veins of the Dahlonega belt are mostly quartz, commonly with lesse muscovite, r amounts and biot of ite .p l aHgoi orncbllaesnede(,Anz2o0 _i3s0i)t,e cal /cli cite or nozois dolomite, apatite, ite, and tourmaline are common in veins hosted by amphibolite, whereas garnet is a common constituent of metasediment-hosted deposits; kyanite has been observed in gold-bearing veins at the Battle Branch (Pardee and Park, 1948) and Charles Mines (this study). The quartz ranges from white to dark grey, and is glassy to sugary. It typically shows considerable evidence of recrystallization, and in places it occurs as elongate ribbons similar to those in mylonites. Accessory vein minerals occur either along vein margins (especially apatite which occurs as pale green stubby prisms to >1 em), or along partings in the central parts of veins, convnonly associated with pyrite and pyrrhotite (Figure 9B). Banding is fairly common in the larger veins, where thick pyrite-rich or garnet bands occur. Opaque mineral assemblages are typically simple, with pyrite, pyrrhotite or arsenopyrite dominant. Where pyrite or arsenopyrite coexist with pyrrhotite, textural relationships indicate that pyrrhotite deposition was late (Figure 9D, 9E), although composite pyrrhotite- chalcopyrite blebs do occur as inclusions in gold-bearing pyrite. The common occurrence of pyrhhotite in pressure shadows (Figure 9E) is one manifestation of the paragentic relationships among the Fe sulfides, as is the occurrence of pyrrhotite in small, discontinuous veinlets which sharply cross-cut foliation. Pyrrhotite is most common in amphibolite- hosted deposits, where it occurs in wall rock adjacent to veins, without accompanying pyrite, as well as in the veins themselves. In felsic and metasedimentary rocks pyrite is always dominant, both in the veins, where it occurs with pyrrhotite, and in the wall rocks. Chalcopyrite is common in small amounts in the Dahlonega Belt, although rarely visible in hand specimen. It occurs entirely in close association with pyrrhotite, commonly as composite grains associated with pyrite, some of which have well-developed crystal forms (Figure 9F). This probably indicates these composite grains formed initially as intermediate solid solution (iss - (CuFe) 1+xS), and that pyrite and iss were deposited contempraneously. This assemblage is stable only at temperatures of >450C (Barton and Skinner, 1979). Minor galena occurs at a number of properties (Pardee and Park), and is commonly associated with high grade ore shoots. Galena does not universally indicate high gold contents, however, as galena-bearing samples fom the Battle Branch and Charles Mines, are not all enriched in gold (samples BB-4 and CH-3-120, Table 1). Oxide minerals occur in small amounts in many Dahlonega Belt quartz veins. Ilmenite, as thin, sometimes curved plates is most common. Magnetite, as isolated euhedra and locally as masses a few em in size, is 103 Figure 9. Veins, ore textures and ore mineralogy of DB gold deposits. A, Tabular zone of pod-like quartz veins in sheared and carbonatlzed amphibolite, near Canton. B, Banded quartz vein from CreightonFranklin Mine. Vein is mainly white quartz, with dark bands of hornblende (hb), and coarse pyrite euhedra with pyrrhotite pressure shadow (arrow). C, Mylonltlzed porphyritic trondhjemite, Dahlonega-consolidated Mine. Gold Is in thin veinlets parallel to dike contacts or in euhedral pyrite grains occurring along foliation (arrow). D,Pyrlte (py) surrounded and partly replaced by later pyrrhotite (po) In quartz vein, Creighton-Franklin Mine. (Continued on next page.) .. 104 Figure 9 (Cont.). E, Coarse subhedral arsenopyrite (asp) with pyrrhotite (po) In pressure shadow. F, Composite chalcopyrite (cp) -pyrrhotite (po) grain In quartz vein, showing apparent crystal outllne, Creighton-Franklln Mine. G, Composite pyrite-chalcopyrite grain In pyrite occurring with calverite (cv), petzite (ptz), and native gold (Au) In ore from Dahlonega-consolldated Mine. H, Sylvanite (syl) ln fractures ln pyrite, Dahlonega- Consolldated Mine. Back-scattered electron Image. 105 fairly widespread. Gahnite, the Zn spinel was noted by Lindgren (1906) at the Standard Mine near Dahlonega. Gold- and gold-silver tellurides occur in most unweathered samples containing more than 1 ppm gold. These minerals are all extremely finegrained, ranging from 1 to a maximum of 10 to 20 microns in size. In many cases they were only noted optically after they had been detected by EDS analysis. The tellurides found include petzite, calverite, and rare sylvanite. Sylvanite was observed in a single sample in fractures in pyrite (Figure 9H), in contrast to petzite and calaverite which occur in a large number of samples. These two phases (petzite > calaverite) are typically found in composite grains with small amounts of native gold (Figure 9G). These composite grains are closely associated with pyrite, and occur either as inclusions in pyrite or along pyrite-gangue grain boundaries. Where included in pyrite the gold-bearing phases are commonly accompanied by pyrrhotite-chalcopyrite composite grains. Assay data on a limited number of samples from the Dahlonega Belt substantiate the correlation between gold and tellurium (Figure 10F). The consistent association of native gold, petzite and calaverite are suggestive that the three minerals represent the breakdown products of some mineral which has decomposed, probably on cooling. The data of Cabri (1965) suggest that the original gold-silver-tellurium phase was gold-rich petzite-hessite88 , which is likely stable only at temperatures above approximately 400C at pressures of 4.5 to 5 kb, further confirming the high temperature origin of the Dahlonega Belt gold deposits. The occurrence of the gold-telluride mineral intergrowths as inclusions in pyrite also provide strong evidence that gold deposition was fairly early during vein formation. In addition to recoverable quantities of gold and silver, the ores of the Dahlonega Belt contain variable but generally small amounts of arsenic, tungsten, molybdenum, and tellurium, and traces of antimony and mercury (Figure 10; Table 1). Arsenic is abundant (up to percent level) in a few localities where arsenopyrite is the main iron sulfide phase, but in most occurrences is present in the low ppm range (Figure 10B). Antimony is present in much smaller amounts (Figure 10C), and in general is strongly correlated with arsenic (Figure 10H), suggesting that it also occurs in arsenopyrite. A similar relationship is observed for mercury, which is present in even lesser quantities (Figure 100). There are at least two, and possibly three, distinct populations with respect to gold content (Figure 10A, E). The larger group contains gold in the range of <5 to approximately 60-100 ppb, and may itself be divisible into two sub-groups, one with gold <10 ppb, and the second containing 10 to 100 ppb gold. These groups appear to correspond to the two populations noted by Cook and Burnell (1985), with the higher grade population ecorresponding to their 'chemical sedimentary' unit. The second major group consists of 'ore' samples with gold content close to or greater than 1 ppm (Table 1, Figure 10E). The two populations which contain anomalous amounts of gold may reflect different mechanisms of gold deposition, with gold-silver telluride deposition playing an important role in the higher grade group, and gold concentration controlled by nonequilibrium processes, possibly co-precipitation with pyrite or arsenopyrite, in the low-grade group (see, for instance, Romberger, 1986). There is no correlation between gold and arsenic abundances in the sample population as a whole (Figure 10G), but if only data from gold-poor rocks is included there is a general association of gold with arsenic. 106 (A) U) 12 Ill ..1 10 A. Ic 8 U) 6 IL 4 0 2 0 AU AU (PPM - LOG SCALE) (C) 20 U) Ill 18 ..1 16 A. 14 cI 12 U) 10 8 1L 6 0 4 SEI 2 0 0.1 1 10 100 1000 10000 Sll (P..M - LOG 8CALI!) ( E) CUMULATIVE FREQUENCY 100 #. 80 (G) ~ ::3 tI 12! 10 ~ .... 8- :;) 6} c 4 .. -f :t~-r ~ 2 4 6 8 10 AS(PPM) (THOU8ANDS) (B) 26 U) Ill 24 ..1 A. 20 cI 18 U) 12 IL 0 8 4 0 A& 10 100 1000 AS (PPM - LOG SCALE) (D) 12 HG U) 11 Ill 10 ..1 9 i 8 c 7 8 en 5 4 ~ 3 2 I l l 1 O~~~~~~~TTTTTT~TT~ 0.01 0.1 1 10 100 1000 10000 HG (PPM - LOG SCALE) ( F ) 18 TE va. ~~u 16 ~ 14 I 12 A. ~ :8L..A.... 10 . 1 0 ,_,...I,-r-111,......,...1..,.......,I.,.....,-,.......,I..,-.--.11-r-1"T""TI'I...,..-i1 2 4 6 8 10 12 14 16 AU(PPM) ( H ) S B v a . A S_ __ _ , 4-r---=..;;~.......;;-=.;......_.......;.. 3.5 2 4 6 8 10 AS (PPM) (THOUSANDS) Figure 10. Metal abundances and correlations In altered and mlnerallzed samples from the Dahlonega Belt. Data from Table 1. 107 HYDOTHERMAL ALTERATION Mineral Assemblages and Textures Mineralogical changes accompanying gold mineralization are greatest in amphibolites, and are much less obvious in granitic rocks and metasediments, where alteration is commonly expressed only by modal changes without appearance or disappearance of phases. Alteration was synchronous with to slightly later than retrograde metamorphism and shearing, based on the sequence of alteration assemblages in pelitic rocks, from kyanite- garnet to kyanite-staurolite to staurolite-muscovite, and on the occurrence of alteration minerals both with preferred orientations and as decussate aggregates. In selvages flanking late- formed brittle quartz-tourmaline-pyritepyrrhotite veins, well-developed quartz ribbons are preserved, but the matrix has been altered to biotite, muscovite, calcite and pyrrhotite and locally tourmaline, indicating that alteration outlasted mylonitization. Alteration in amphibol ites is of two types involving either C02 addition or Fe-Mg depletion. The first type (C02 addition) is restricted to the immediate vicinity of quartz veins, and is more closely related to significant gold deposits such as the Creighton-Franklin Mine. The main mineralogical changes in this type of alteration include: 1) decrease or total destruction of plagioclase; 2) increase in modal quartz; 3) increase in clinozoisite-zoisite abundance, and; 4) the appearance of biotite, calcite and/or dolomite, and pyrrhotite (Figure 11A). More locally chlorite, muscovite and tourmaline are present in altered amphibolite (Figure 11B). Textures of the altered rocks are variable, they range form well-foliated and lineated to very coarse-grained decussate, with very large sprays (>1 em) of prismatic hornblende and clinozoisite. In a few zones carbonatization is accompanied by iron metasomatism, which has led to development of abundant Fa-silicates such as almandine and grunerite, accompanied 1oca 11 y by significant amounts of magnetite and/or i ron sulfides. Iron and magnesium-depleted zones are either thin zones flanking quartz veins, or broad (metres wide) zones associated with intense shearing along 1ithologic contacts; locally sulfide-bearing irregular quartz-cl inozoisite-hornblende-magnet ite veins occur in some of these zones. Although no major gold deposits appear to be related to these zones, there are small hydraulic pits in saprolite where sulfidic quartz veins occur with the altered amphibolites. The altered zones are mainly fine-gra amounts in o e f d c granular linozoisi t q e u artz (ofte and pla n quite giaobculnadsaent)(Ana3n0d_ 40), with variable small amounts of magnetite, garnet, and hornblende. Compositions of plagioclase and hornblende are very similar to those in unaltered rocks. The main mineralogical changes are an increase in modal quartz and plagioclase, and a large decrease in modal hornblende. Altered intrusive rocks occur at the Dahlonega-Consolidated, Barclay, and Charles Mines. At the Dahlonega- Consolidated mine auriferous quartz veins occur in both coarse-grained and fine-grained porphyritic trondhjemite. Biotite in coarse-grained trondhjemite flanking veins is partly replaced by both coarse and shreddy muscovite. Albite and muscovite are more abundant, and biotite much less abundant, in porphyritic trondhjemite near veins (Figure 11C). Locally thin biotite- hornblende-apatite veinlets cut otherwise fresh trondhjemite, but these 108 Figure 11. Hydrothermal alteration lD gold deposits of the Dahlonega Belt. A. Biotlte-homblende-cllDozoisite-calcite zone Oanklng pyritic and auriferous quartz velD with central homblende parting, CreightonFranklin Mine. B, Altered selvage around late, cross-cutting quartzpyrite-pyrrhotlte-tourmallne (T) velD lD amphibolite, CreightonFranklin Mine. Bleaching around velD is result of muscovite-dolomite replacement of amphibolite assemblage of hornblende and plagioclase. C, Porphyritic trondhjemite, Dahlonega Consolidated Mine, showing bleaching from an Increase lD albite and muscovite, and loss of biotite. Slmtlar features also occur on larger scales. D, Coarse-grained gamet (GT) lD muscovite schist, Battle Branch Mine. Gamet shows evidence of rotation, but kyanlte and staurolite lD fine-grained muscovite schist are randomly oriented. 109 do not appear to be associated with gold- bearing quartz. Thin quartzcalcite-pyrite veinlets occur near the auriferous zones. Disseminated pyrite appears to be rare, with the exception of zones of strongly foliated quartz-muscovite- pyrite schist traceable into coarse-grained trondhjemite at the Consolidated property. At the Barclay Mine the main auriferous quartz vein cuts the Persimmon Creek Gneiss, a recrystallized and foliated tonalite of uncertain age (Hatcher, 1974). Alteration around the vein has resulted in muscovite replacement of biotite. In the 'Quarry' Zone at the same mine gold occurs in deformed pegmatitic stringers in a gneissic tonalite matrix with no obvious mineralogical effects. Two types of alteration occur in metasediments along the Dahlonega Belt, discrete alteration zones adjacent to veins, and broad zones containing auriferous quartz veins and abundant garnet, staurolite and kyanite, which occur mostly as syn- to post-tectonic porphyroblasts as large as 3 to 4 em (Figure 11D). Aluminosilicate-rich zones are commonly poorly foliated, suggesting that these assemblages formed late in the deformation of the Dahlonega Belt. These rocks consist of quartz, biotite, ilmenite and garnet with or without staurolite or kyanite; plagioclase is generally absent. Quartz veins, commonly containing large garnet and locally staurolite and kyanite, occur in some aluminosilicaterich zones. Gold is locally closely associated with the garnet and kyanite (Pardee and Park, 1948). Coarse-grained apatite occurs in veins and aluminosilicate- rich vein selvages at several localities, particularly the Battle Branch and Jones-Lot 10 Mines. Coarse-grained, highly poikiloblastic hornblende, rare in metasedimentary rocks of the Dahlonega Belt, is present locally. Late-formed chlorite and muscovite occurs in pressure shadows around porphyrob1asts and as part i a1 rep1acement of aluminosilcate minerals; replacement of foliation-parallel biotite by unoriented muscovite blades is also common. Quartz, calcite and pyrite are mainly restricted to veinlets, with the exception of a broad altered and slightly auriferous zone at the Charles Mine, where calcite-quartzbiotite-garnet-hornblende-plagioclaseFe sulfide is the common assemblage. Altered selvages flanking quartz veins are characterized abundant muscovite, with secondary muscovite as fine-grained, generally unoriented scaley aggregates, commonly with associated tourmaline. Biotite is minor in these zones, reflecting the growth of muscovite at its expense; garnet, too, commonly is corroded in appearance, apparently as a result of muscovite replacement. Minor calcite and pyrite are common and locally abundant. These zones of muscovite- and pyrite-rich schist have themselves been important sources of gold in some of the saprolite 'belts', for instance those around Helen (Jones, 1909; Yeates et al., 1896). Assays from unweathered pyritic muscovite-rich schist, however, indicate that the gold content of unweathered rock is low (Table 1). Conditions of Mineralization and Composition of the Ore-Forming Fluid Garnet-biotite geothermometry (Thompson, 1976) and garnet-plagioclase (+quartz and Al 2Si05 ) geobarometry (Ghent, 1976) indicate that retrograde metamorphism and alteration took place at temperatures of >500C and pressures of 5kb. The common occurrence of pyrite or arsenopyrite with pyrrhotite, iss, ilmenite and magnetite indicate that the mineralizing fluids had moderate values of f 02 and f52 (Barton and Skinner, 1979). The 110 presumed occurrence of iss+pyrite, and gold-rich petzite-hessite88 also support a high temperature origin for the sulfides and gold mineralization (see above). The mineral assemblages in altered amphibolites suggest that the ore- forming fluid was and a pressure of 5C0k2b- rich. indic a Ca te lcula that tions the fl based uids on has a temperature of 550C an Xc02 as high as 0.25 (Hoschek, 1980) in order to form the dominant biotite-hornblende-calcite- quartz assemblage (Figure 12). The later and less abundant muscovite- chlorite-dolomite-quartz zones reflect alteration either by more C02-r1ch fluids or at a lower temperature (Figure 12). The low abundance of carbonate in altered zones in other rock types reflects the role of bulk composition in governing alteration mineralogy; only calcium-rich rocks like the amphibolites will tend to be carbonatized at the temperatures prevalent in this area. Alteration of quartzofeldspathic rocks such as trondhjemite or tonalite and psammitic metasediments consisted of H+ metasomatism, expressed by the formation of muscovite, with attendant loss of potassium and possibly some sodium. The potassium gained by biotite formation in some amphibolites may have been derived from the metasedimentary rocks in this way. The prevalence of H+ metasomatism as an important form of alteration in the Dahlonega Belt gold deposits is a resu 1t of the temperature dependence of reactions between white micas (muscovite or paragonite) and feldspars (Meyer and Hamley, 1967), because, at a constant K+;w (or Na;H+), formation of micas will be favored over feldspars with decreasing temperature. That is, if a fluid in equilibrium with muscovite and albite at a given temperature is allowed to react with a rock at a lower temperature (but with K+/H+ held constant), the result will be muscovite replacement of feldspar (or biotite), such as is seen in the Dahlonega Belt and other gold deposits in metamorphic belts (Weir and Kerrick, 1987). The occurrence of muscovite+feldspar as the common assemblage in these rocks indicates that the mineralizing fluid was likely slightly acid to neutral (Meyer and Hamley, 1967). The lesser abundance of carbonate in altered rocks of the Dahlonega Belt, as compared to greenschist-hosted Mother Lode-type deposits, is explained by the relationship between temperature and equilibrium Xc02 for mixed volatile reactions on isobaric T-Xc02 sections (Figure 12). At higher temperatures carbonatization equilibria are displaced to higher avsasleumesbloafgeXsc0w2 ,ilalndbet herefor formed e c in alcite mafic -bearin rocks, g, re rather sulting than dolo in lower mite-bearing C02 contents in altered rocks. As well, in Ca-poor lithologies such as quartzofeldspathic metasediments and felsic igneous rocks, even calcite will not typically form from fluids with Xc02 in the range calculated for the Dahlonega Belt fluids. Similar arguments can be applied to sulfidation reactions which would tend to produce pyrrhotite or pyrite in altered wall rocks. Because of the positive slope of sulfidation equlibria on T-f92 diagrams, formation of sulfide minerals from Fe-oxides or silicates is less likely to occur at high than low temperatures, which explains why rocks with more than a few percent sulfide are absent in the Dahlonega Belt, in contrast to the more abundant Fa-sulfides in alteration zone of deposits formed at lower temperatures (e.g. Phillips, 1985). The very close association between gold and tellurides in most deposits examined is also an unusual feature compared to greenschist- hosted Mother Lode ores, where a gold-tellurium association is usually part of a late stage of ore formation (Weir and Kerrick; Phillips, 1985). This relationship in the Dahlonega Belt ores may indicate that Au 111 500 0 0.1 P=5.5 kb 0.3 0.4 Figure 12. T-Xc02 section showing stablUty fields for assemblages In altered amphlbolltes of the Dahlonega Belt. Note that early assemblages such as blotlte-cllnozolslte formed at higher Xc02 or lower temperature than later chlorite- or muscovite-bearing assemblages. 112 saturation is not normally reached at the high temperatures at which the Dahlonega Belt ores were deposited, and that the presence of sufficient aqueous tellurium to cause saturation and precipitation of gold-rich petzite-hessite was a prerequisite for gold deposition. Precipitation of gold-silver 9t8elluride may have been caused by cooling, as the change in Au solubility as a function of temperature is much greater for tellurides than for the equilibrium involving Au0 (using data for the Au telluride, calverite, as no data exist for the high temperature phase which probably was deposited in the Dahlonega Belt mineralized zones). Deposition of the gold-silver telluride may have resulted in depletion of tellurium in the fluid, explaining why gold tellurides are not abundant in the main stage mineralization in most Mother Lode-type deposits. The occurrence of very small amounts of gold (<100 ppb) in a number of samples lacking tellurides may be the result of gold incorporation in the lattice of common sulfide minerals, probaly arsenopyrite or pyrite. Incorporation of gold in sulfide lattices has long been a hotly debated topic, but the recent data of Cathelineau et al. (1988) provides firm evidence that gold does occur in the lattice of arsenopyrite up to several thousand ppm. Mining of such low grade deposits was probably only possible where intense weathering over along period of time produced thick saprolite zones~ which were enriched by accumulation of gold from an overlying 'collapsed' zone of weathered rock. Tectonic Setting and Timing of Events in the Dahlonega Belt There are few published radiometric age determinations for the Eastern Blue Ridge, so timing of events along the Dahlonega Belt must be inferred indirectly. Peak metamorphism in the area is temporally related to nappe emp1acement, based on the evidence of inverted metamorph 1c gradients. This deformation is generally considered to be Taconic in age (e.g. Glover et al., 1983). The Dahlonega Belt shearing is associated with retrogression of the peak assemblages, and thus is probably post-Taconic. The association of greenschist facies retrogression with westward-directed thrusting in the southwestern part of the Dahlonega Belt indicates that there was a later, lower-grade retrogression-thrusting event, probably Alleghanian in age by analogy with the Brevard Zone (Edelman et al., 1987). The single hornblende 39Ar-40Ar age determination from the Dahlonega Belt (Dallmeyer, 1988) of 331 Ma probably records cooling through 450-500C, which, based on temperatures of retrogression and mineralization, may have been later than Dahlonega Belt metamorphism and mineralization. This date, does however, provide a lower limit for the age of the belt. The synchrohous nature of at least some plutonism, deformation and gold deposition, however, suggests an Acadian age (360-380 Ma) for the belt, based on: 1) probable Acadian ages for syn-kinematic granitic rocks southeast of the Dahlonega Belt (Fullagar and Goldberg, 1988) and similar plutons 1n Alabama (Drummond et al., 1988); 2) evidence of Acadian transpression in the sedimentary record of the southern and central Appalachian (Ferrill and Thomas, 1988), and; 3) the occurrence of a number of major Acadian dextral strike-slip faults in the northern Appalachians (Williams and Leger, 1988). A younger, early Alleghanian age, as suggested by Lesher and Green (1988) for gold deposits at Hog Mountain, Talapoosa County, Alabama, is also consistent with available data. 113 Whether gold deposits of the Dahlonega Belt are Acadian or Alleghanian in age, the structural evidence indicates that deformation took place 1n a transpressive setting, which may have been due to nonothorgonal plate convergence. Where plate covergence is significantly oblique (l 22) the orogen-para l lel component of plate motion is typically partitioned into strike-slip faults (trench-linked strike-slip faults of Woodcock, 1986). In a transpressional regime these strike-slip faults may be accompanied by 'flower structures', which are zones in which shortening across a fault is accomodated by vertical extrusion of material to higher structural levels . The combination of shear parallel to the belt and flattening across it is consistent with the Dahlonega Belt being the deep level equivalent of a flower structure. The greater abundance of amphibolite in the Dahlonega Belt relative to the flanking areas may also be attributable to upward movement of the structurally lower mafic/ultramafic Lake Burton Complex as a result of shortening across the belt. The Dahlonega Belt has been affected by post-mineralization retrograde metamorphism associated with continentward thrusting of the crystalline rocks of the Blue Ridge. The effects of this superposed deformation are only evident southwest of Dahlonega, where published regional maps (e.g. Nelson, 1985) show the Hayesville Thrust curving sharply southward, and joining the Dahlonega Belt. These features suggest that the Hayesv i 11 e Thrust, or some c1ose1y re1ated structure, have undergone reactivation during Alleghanian thrusting (Vauchez and Dallmeyer, 1989). CONCLUSIONS In brief, the Dahlonega Belt is interpreted as a 1-2 km wide ductile shear zone which underwent shortening and right-lateral displacement in a transpressive regime during the accretion of the Avalon Terrane to North America in the Acadian Orogeny. Shearing was contemporaneous with retrograde metamorphism at temperatures of 500-550C and pressures of ca. 5 kb. Emplacement of felsic plutonic rocks, including both K- and Na-rich varieties, and formation of auriferous quartz veins al so occurred during shearing. Formerly exploited gold deposits are of several types, all of which are characterized by an abundance of vein quartz. The mineralizing fluid was C02-rich, and interaction with amphibol ites resulted in formation of calcite-, clinozoisite, and biotite-bearing assemblages. Alteration in quartzofeldspathic by the same fluids rocks consisted mainly of muscovite formation at the expense of feldspars and biotite. High gold abundances are accompanied by tellurium enrichment in most deposits examined, and the ubiquitous association of gold, petzite, and calaverite is consistent with gold precipitation in the form of gold-rich petzitehessite88, a phase stable only at high temperature. The occurrence of widespread, very low-grade gold mineralization may be the result of incorporation of gold in the lattice of arsenopyrite or pyrite. Other elements slightly enriched in Dahlonega Belt occurrences are arsenic, antimony, mercury , tungsten, and molybdenum. Similarities between the deposits of the Dahlonega Belt and the classical greenschist-hosted Mother Lode-type deposits suggests that the Dahlonega Belt ores are the midcrustal equivalent of Mother Lode-type deposits such as those in the Sierra Nevada of California, the Abitibi Greenstone Belt of the Canadian 114 Shield, and the Eastern Goldfields province of the Yilgarn Block in Western Australia. ACKNOWLEDGEMENTS A large number of people and organizations have contributed to this study. These include BP Minerals, which funded the work reported here, as well as the property holders in the Dahlonega Belt who allowed me access to their properties and core. Discussions with workers in the area, including Bob Cook, Jerry German, Janet Hopson, Travis Paris, and in particular Ken Gillon, whose 1986 thesis provided an extremely important geologic base in the Helen area, were also very helpful. Reviews of various incarnations of this paper by Bob Hodder, Bob Cook, and Jerry German led to substantial improvements. I must emphasize, however, that the conclusions of this paper are my own, and that many of them are not shared by the geologists mentioned above. REFERENCES Absher, B.S. and McSween, H.Y., Jr., 1985, Granulites at Winding Stair Gap, North Carolina: the thermal axis of Paleozoic metamorphism in the southern Appalachians: Geol. Soc. America Bull., v. 96, p. 588599. Albino, G.V., 1987, The Dahlonega Belt, southeast Georgia Blue Ridge. I) Deformation textures and history: Geological Society of America Abstracts with Programs, v. 19, p. 73. Barton, P.B., Jr. and Skinner, B.J., 1979, Sulfide mineral stabilities: in Barnes, H.L. (ed.), Geochemistry of Hydrothermal Ore Deposits (2nd edition), New York, John Wiley & Sons, p. 278-403. Bayley, W.S., 1928, Geology of the Tate quadrangle, Georgia: Georgia Geol. Survey Bull. 43, 170 p. Becker, G.F., 1894, Reconnaissance of the gold fields of the southern Appalachians: U.S. Geol. Surv. 16th Annual Report, Part 3, p. 251319. Bell, T.H., 1978, Progressive deformation and reorientation of fold axes in a ductile mylonite zone: the Woodroofe Thrust: Tectonophysics, v. 44, p. 285-320. ------, Rubenach, M.J., and Fleming, P.O., 1986, Porphyroblast nucleation, growth and dissolution in regional metamorphic rocks as a function of deformation partitioning and foliation development: Jour. Met. Geol., v. 4, p. 37-67. 115 Bouiller, A-M. and Bouchez, J-L., 1978, Le quartz en rubans dans les mylonites: Bull. Soc. Geol. Fr., val. 22, p. 253-262. Bryant, B. and Reed, J.C., Jr., 1970, Structural and metamorphic history of the southern Blue Ridge: in Fisher, G.W. et al. (eds.), Studies of Appalachian geology-cental and southern: New York, Interscience Publishers, p. 213-225. Cabri, L.J., 1965, Phase relations in the Ag-Au-Te system and their mineralogical significance: Econ. Geol., v. 60, p. 1569-1606. Carpenter, R.H., 1970, Metamorphic history of the Blue Ridge province of Tennessee and North Carol ina: Geol. Soc. America Bull. v. 81, p. 749-762. Cathelineau, M, Boiron, M.C., Holliger, Ph., and Marion, Ph., 1988, Goldrich arsenopyrites: crystal-chemistry, gold location and state, physical and chemical conditions of crystallization: in Goode, A.D.T. and Bosma, L.I. (compilers), Bicentennial Gold 88, Extended Abstracts Oral Program, Geol. Soc. Australia Inc., Abstracts No. 22, p. 235-240. Cobbold, P.R. and Quinquis, H., 1980, Development of sheath folds in shear regimes: Jour. Struct. Geol., v. 2, p. 119-126. Cook, R.B. and Burnell, J.R., Jr., 1985, Trace metal signature of major lithologic units, Dahlonega District, Lumpkin County, Georgia: Geol. Soc. America Abstracts with Programs, v. 17, p. 85. ------ and ------, 1983, Geo1ogy of the Dah 1onega district, Georgi a: Geol. Soc. America Abstracts with Programs, v. 15, p. 109. ------ , ------ , and Sibley, D.E., 1984, Preliminary geologic map of the Dahlonega district, Georgia: Georgia Geol. surv., Open-File Report 85-3. Crickmay, G.W., 1952, Geology of the crystalline rocks of Georgia: Georgia Geol. Surv. Bulletin 58, 54 p. Dallmeyer, R.D., 1988, Late Paleozoic evolution of the western Piedmont and eastern Blue Ridge, Georgia: Controls on the chronology of terrane accretion and transport in the southern Appalachian orogen: Geol. Soc. America Bull., v. 100, p. 702-713. Drummond, M.S., Wesolowski, D., and Allison, D.T., 1988, Generation, diversification, and emplacement of the Rockford Granite, Alabama Appalachians: Mineralogic, petrologic, isotopic (C & 0), and P-T constraints: Jour. Petrol., v. 29, p. 869-898. Edelman, S.H., Liu, A., and Hatcher, R.D., Jr., 1987, The Brevard Zone in South Carolina and adjacent areas: An Alleghanian orogen-scale dextral shear zone reactivated as a thrust fault: Jour. Geology, v. 95, p. 793-806. 116 Escher, A. and Watterson, J., 1974, Stretching fabrics, folds and crustal shortening: Tectonophysics, vol. 22, p. 223-231. Evans, B.W. and Trommsdorrf, V., 1970, Regional metamorphism of ultramafic rocks in the central Alps: Parageneses in the system CaO-MgO-Si02H20: Schweiz. Mineral. Petrog. Mitt., v. 50, p. 481-492. Ferrill, B.A. and Thomas, W.A., 1988, Acadian dextral transpression and synorogenic sedimentary successions in the Appalachians: Geology, v. 16, p. 604-608. Fullagar, P.O. and Goldberg, S.A., 1988, Rb-Sr and U-Pb geochronological investigations in the ADCOH site area, southern Appalachians: Geol. Soc. America Abstracts with Programs, v. 20, p. 264. Furcron, A.S. and Teague, K.H., 1945, Sillimanite and massive kyanite in Georgia: Georgia Geol. Surv. Bulletin 51, 76 p. German, J.M., 1988a, The geology of gold occurrences in the west-central Georgia Piedmont: Georgia Geol. Surv. Bull. 107, 48 p. ------, 1988b, Geology of the Dahlonega and Carroll County gold belts, Georgia: Geol. Soc. America Abstracts with Programs, v. 20, p. 266. ------, 1985, The geology of the northeastern portion of the Dahlonega gold belt: Georgia Geol. Surv. Bull. 100, 41 p. Ghent, E.D., 1976, Plagioclasegeothermometer-geobarometer: g arn Am. e tM-Ainl e2Sria2l0.,5-vq.ua6r1t,z : p. a potential 710-714. Gillon, K.A., 1982, Streatigraphic, sructural, and metamorphic geology of portions of the Cowrock and Helen, Georgia, 71/2 quadrangles: unpublished M.S. thesis, University of Georgia, Athens, Georgia, 236 p. Glover, L., Speer, J.A., Russell, G.S., and Farrar, S.S., 1983, Ages of regional metamorphism and ductile deformation in thr central and southern Appalachians: Lithos, v. 16, p. 223-245. Guthrie, G.M. and Lesher, C.M., 1988, Structural and metamorphic controls on lode gold mineralization, Talapoosa County, Alabama: Gaol. Soc. America Abstracts with Programs, v. 20, p. 268. Hatcher, R.D., Jr., 1987, Tectonics of the southern and central Appalachian intern1des: Ann. Rev. Earth Plan. Sci., v. 15, p. 337362. ------, 1978, Tectonics of the western Piedmont and blue Ridge, southern Appalachians: review and speculations: Am. Jour. Sci., v. 278, p. 276-304. ------, 1979, The Coweeta Group and coweeta Syncline: major features of the North Carolina-Georgia Blue Ridge: Southeastern Gaol., v. 21, no. 1, p. 17-29. 117 ------, 1974, An introduction to the Blue ridge tectonic history of northeast Georgia: Georgia Geol. Soc. Guidebook, 60 p. ------, 1973, Basement versus cover rocks in the Blue Ridge of northeast Georgia, northwestern South Carolina, and adjacent and adjacent North Carolina: Am. Jour. Sci., v. 273, p. 671-685. ------, 1971, The geology of Rabun and Habersham Counties: Georgia Geol. Surv. Bull. 83, 48 p. Higgins, M.W., Atkins, R.L, Crawford , T.L., Crawford, R.F., III, and Cook, R.B., 1984, A brief excursion through two thrust stacks that comprise most of the crystalline terrane of Georgia and Alabama: Georgia Geol . Soc . Guidebook, 19th Annual Field Trip, 67 p. ------,------,Nelson, A.E., and Crawford, T.J., 1988, The Talking Rock Complex: an ancient accretionary complex, northern georgia and southern North Carol ina: Geol. Soc. America Abstracts with Programs, v. 20, p. 270. Hopson, J.L., 1988, Petrogenesis of the Lake Burton mafic-ultramafic complex: Geol. Soc. America Abstracts with Programs, v. 20, p. 271. Hoschek, G., 1980, Phase relations of a simplified marly rock system with applications to the western Hohe Tauern (Austria): Contrib. Mineral. Petrol., v. 73, p. 53-68. Jones, S.P., 1909, Second report on the gold deposits of Georgia: Geol. surv. Georgia Bull. 19, 283 p. Kerrich, R., 1983, Geochemistry of gold deposits in the Abitibi greenstone belt: Can. Inst. Mining Metal. Special Paper 27, 75 pp. Kittelson, R.C. and McSween, H.Y., Jr., 1987, The amphibolite to granulite facies transition in the Blue Ridge, Macon County, NC: Geol. Soc. America Abstracts with Programs, v. 19, p. 93. Koschmann, A.H. and Bergendahl, M.H., 1968, Principal gold-producing districts of the United States: u.s. Geol. Surv. Prof. Paper 610, 283 pp. Lesher, C.M. and Green N.L., 1988, Syn-metamorphic hydrothermal alteration, Hog Mountain gold deposit, northern Alabama Piedmont: Geol. Soc. America Abstracts with Programs, v. 20, p. 277. Lesure, F.G., 1971, Residual enrichment and supergene transport of gold, Calhoun Mine, Lumpkin County, Georgia: Econ. Geol., v. 66, p. 178186. Lindgren, W., 1906, The gold deposits of Dahlonega, Georgia: u.s. Geol. Surv. Bull. 293, p. 119-128. 118 McConnell, K.I., 1980, Origin and correlation of the Pumpkinvine Creek Formation: a new unit in the Piedmont of northern Georgia: Georgia Gaol. Surv. Info. Circ. 52, 19 p. ------ and Abrams, C. E., 1988, Metamorphosed felsic to intermediate volcanic rocks and their association with volcanogenic gold deposits in the northern Piedmont of Georgia: Gaol. Soc. America Abstracts with Programs, v. 20, p. 279. ------ and ------ , 1984, Geology of the Greater Atlanta Region: Georgia Geol. Surv. Bull. 96, 127 p, ------ and ------, 1983, Geology of the New Georgia group and associated sulfide and gold deposits: west-central Georgia: Society of Economic Geologists, Field Trip Guidebook, Atlanta, Georgia, 27 p. Meyer, C. and Hamley, J.J., 1967, Wallrock alteration: in Barnes, H. L. (ed.) Geochemistry of hydothermal ore deposits, New York, Holt Rinehart and Winston, p. 166-245. Murray, J.B., 1973, Geologic map of Forsyth, and north Fulton Counties, Georgia: Georgia Geol. Surv. Bull. 88 (map only). Nelson, A.D., 1985, Major tectonic features and structural elements in the northwest part of the Greenville Quadrangle, Georgia: U.S. Geol. Surv. Bull. 1643., 22 p. ------and Gillon, K.A., 1985, Stratigraphic nomenclature in the Richard Russell and Helen thrust sheets, Georgia and North Carolina: U.S. Geol. Surv. Bull. 1605-A, p. A59-A62. Pardee, J.T. and Park, C.F., Jr., 1948, Gold deposits of the southern Piedmont: U.S. Geol. Surv. Prof. Pap. 213, 156 p. Passchier, C.W. and Simpson, C., Porphyroclast systems as kinematic indicators: Jour. Struct. Geol., v. 8, p. 831-843. Phillips, G.N., 1985, Archean gold deposits of Australia: University of the Witswatersrand Economic Geology Research Unit Information Circular 175, 40 pp. Platt, J.P. and Vissers, R.L.M., 1980, Extensional structures in anisotropic rocks: Jour. Struct. Geol., v. 2, p. 397-410. Romberger, S.B., 1986, The solution chemistry of gold applied to the origin of hydrothermal deposits: in: Clark, L.D. (ed.), Gold Deposits of the Western Shield, Canadian Institute of Mining and Metallurgy Special Volume 38, p. 168-186. Sanderson, D.J. and Marchini, W.R.D., 1984, Transpression: Jour. Struct. Geol., v. 6, p. 449-458. Simpson, C. and Schmid, S.M., 1983, An evaluation of criteria to deduce the sense of movement in sheared rocks: Geol. Soc. America Bull., 119 v. 94, p. 1281-1288. Stieve, A.L., Sinha, A.K., and Hatcher, R.D., Jr., 1988, Rb/Sr whole rock Grenville age for the basement gneisses of Tallulah Falls Dome, northeastern Georgia: Geol. Soc. America Abstracts with Programs, v. 20, p. 318. Thompson, A.B., 1976, Mineral reactions in pelitic rocks: I. Prediction of P-T-XcFe-Ms> phase relations: Am. Jour. Sci., v. 276, p. 401-424. Tso, J.L., 1987, Evidence of inverted metamorphic isograds between Fries and Galax, SW Virginia Blue Ridge: Geol. Soc. America Abstracts with Programs, v. 19, p. 133. Vauchez, A.R., and Dallmeyer, R.D., 1989, Polyphase tectonic history of the Hayesville Fault, Georgia-North Carol ina: Geol. Soc. America Abstracts with Programs, v. 21, no. 3, p. 63. 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. Woodcock, N.H., 1986, The role of strike-slip faults at plate boundaries: Phil. Trans. Royal Soc. London, vol. 317, p. 13-29. Yeates, W.S., McCallie, s.w., and King, F.P., 1896, A preliminary report on a part of the gold deposits of Georgia: Georgia Geol. Surv. Bull. 4-A, 542 p. 120 PALEOPLACER GOLD IN THE LILESVILLE SAND AND GRAVEL DEPOSITS, NORTH CAROLINA James R. Craig Department of Geological Sciences Virginia Polytechnic Institute and State University Blacksburg, VA 24060 and John E. Callahan Department of Geology Appalachian State University Boone, North Carolina 28608 ABSTRACf Paleoplacer gold has been found in Eocene(?) lag gravel deposits of the Citronelle Formation in the Lilesville Area of Anson County, North Carolina. The processing of the sand-, gravel- and cobble-rich beds for commercial construction grade sand and gravel at rates of up to 10,000 tons per day has resulted in concentration of heavy minerals, including gold. The heavy minerals are dominantly ilmenite with minor kyanite-sillimanite, zircon, rutile, magnetite, hornblende, staurolite, epidote and monazite, with traces of cassiterite and gold. The gold grains were apparently derived from gold deposits exposed to the north and west. The gold, present in grains dominantly less than 1 mm but ranging upward to 5 mm in maximum dimension, occurs generally as irregular to flattened nuggets typical of fluviatile transport. The presence of several small euhedral gold crystals, however, may indicate some secondary mobilization and recrystallization. The silver contents of the gold grains range from 0% to 24% (fineness 1000 to 760), and approximately one-third of the grains analyzed by electron microprobe exhibit the development of partial to complete goldrich rims with finenesses of 990-1000. INTRODUCTION Placer deposits of gold are widespread, geographically occurring throughout geologic time along rivers and streams draining areas where lode gold deposits have cropped out. Recent placer occurrences are well known in many parts of the southern and central Appalachians and are especially well documented .in the Charlotte 1o x 2 quadrangle (e.g., D'Agostino and Rowe, 1986; D'Agostino et al., 1987, 1989). Although they served as the initial discovery sites and major early producers in many of the states, few have been worked or considered economic in recent years. Paleoplacer deposits, including the world's largest reserves (the Witwatersrand deposits in South Africa; Pretorius, 1981), are much less common because they are so often either deeply buried or re-eroded and dispersed. Nevertheless, D'Agostino and Rowe (1986) and D'Agostino and Whitlow (1989) report the occurrence of placer gold within Cretaceous gravels and in fossil stream channels cut into the Triassic rocks in the Charlotte 1 X 2 quadrangle of North Carolina. Cook and Nelen (1989) have described paleoplacers in the Cretaceous sediments of the Alabama coastal plain, and Sharp (1988) reports gold in the Cretaceous of the South Carolina coastal plain. The present paper 121 presents new information on paleoplacer gold found within Eocene(?) gravel deposits of the Citronelle Formation that is presently being exploited for sand and gravel. These heretofore undescribed gold occurrences that contain small quantities of gold (Callahan and Craig, 1989) are important because: (1) they, or others like them, may have economic potential; (2) they may provide information on the past erosional history of the mid-Atlantic region; and (3) they may provide infom1ation on the nature of placer gold in general. LOCATION, GENERAL GEOLOGY OF THE DEPOSITS, AND SAl\tlPLING Paleoplacer gold has been found in the gravels of the Citronelle Formation approximately 5 km east of the town of Lilesville in the Piedmont region of Anson County, North Carolina (79.55'W Long.; 34.55'N Lat.; Fig. 1). The Citronelle Formation named by Matson (1916) has been mapped by Coole:y (1970) in this area as a series of isolated patches with a total aerial extent of about 64 km2 adjacent to the present valley of the Pee Dee River. Cooley (1970) has interpreted these beds as a channel-bar complex of a slowly degrading, laterally migrating stream with a current direction roughly parallel to the present valley trend (i.e., north to south). Similar terrace deposits occur, as shown in Figure 1, along many parts of the eastern Piedmont of North Carolina. The age of the deposits is not known with certainty but has been interpreted as late Pliocene (Berry, 1916) and as middle Eocene and late Eocene (Cooley, 1970). The Citronelle beds are up to 10m in thickness and unconformably overlie sands and clays of the Cretaceous age Middendorf Formation which in turn lies unconformably on the Carboniferous age Lilesville granite (Butler and Fullagar, 1975; Cooley, 1970). The layers are distinct, continuous at least for tens to hundreds of meters, and the contact with underlying beds is sharp. The Citronelle Formation is composed of cross-stratified sands with included stringers of gravel. The basal gravel, often the most visible unit because of its resistance to erosion, consists of moderately to well sorted sand and coarse gravels. Abundant white cobbles, commonly in the range of 5-10 em, are composed of 75-91% metaquartzite with the remainder vein quartz (Cooley, 1970). The nearly spherical nature of many of the cobbles and pebbles suggests a significant distance of transport in high energy fluviatile environment. Gravel, cobble and sand beds are poorly consolidated and quite friable and hence have proven to be easily exploitable sources of sand and gravel for construction purposes. The source area(s) of the metaquartzite pebbles and the heavy minerals in the sand fraction associated with them has not been documented. The vein quartz pebbles could be from any of the numerous quartz veins in the Piedmont One possible source for the metaquartzite is the Pilot Mountain, North Carolina. region where similar lithologies are found. The heavy minerals could be derived from lithologies in the Charlotte, Kings Mountain, or Inner Piedmont belts. The samples described in this study were obtained primarily from the Hedrick Industries sand and gravel operation that is about 5 km east of Lilesville along U.S. Route 74. This operation currently processes approximately 6000 tons per day of sediment that is washed and sized into a variety of sand and gravel products. The initial step in processing is the disaggragation of the poorly indurated material by means of a rotating washing tromrnel equipped with high pressure hoses. The slurry is discharged over a riffle of heavy steel bars and then carried by conveyor belt to a series of screens for sizing. The approximately 10 em deep riffle tends to collect tramp iron scrap and heavy minerals as well as some cobbles and sand, all of which becomes tightly packed. Samples of this sediment were extracted using picks and small trowels and were crudely panned on the spot to eliminate the tramp iron and other assorted scrap metal, coins, and lead hunting shot.. 122 -~-~J~.!_A__-------- ----"""B.----------- NORTH CAROLINA Mondoroon , , . Gruneboro Santorcte tf :'qrlt.<:~1~\t .,RSA,'.-.iEl hIG.:Hlr~~~~ r 4.t aoldtboro ttChorlolle t~~Fa)'lllt111 0 0 IOOmiles IOOkm Figure 1. Sketch map of North Carolina showing the location of Lilesville and the distribution of terrace deposits (in black) on the eastern portion of the Piedmont. The insert at the lower right provides greater detail; the Lilesville deposits are shown in the eastern edge of Anson County. From Callahan and Craig (1989) after Brown et al. (1985). 123 The remaining sediment was washed into buckets and returned to the laboratory for careful panning, sampling, and analysis. Repeated hand panning in the laboratory yielded more than 9000 grains of gold ranging from less than 0.1 mm to a maximum of 5 mm (Fig. 2). Representative grains of gold and the associated minerals were selected and mounted in cold setting epoxy for electron microprobe analysis. Mter grinding and polishing by standard techniques (Craig and Vaughan, 1981), grains were analyzed using an ARL-SEMQ employing pure gold, silver, copper, National Bureau of Standards gold-silver wires and livingstonite (HgSb4Sg) as standards. Analysis was carried out using a beam current of 20 nanoamps at 20 kV, and the data were reduced using the Bence-Albee program. 1HE PALEOPLACER GOLD AND ASSOCIATED HEAVY MINERALS The Citronelle Formation where exposed and mined for sand and gravel east of Lilesville, North Carolina, is dominantly composed of quartz as cobbles, gravel, and sand. The heavy mineral content averages 0.3 % of the total, two-thirds of which is composed of ilmenite and its alteration products (Callahan and Craig, 1989). Also present are two populations of zircon, kyanite, magnetite, hornblende, staurolite, epidote, monazite, rutile, and traces of cassiterite and gold (Table 1). Most of these minerals are present as roughly equant 0.1-1.0 mm grains, although larger grains occur. llmenite, the most abundant of the heavy minerals, typically occurs as somewhat rounded grains that contain numemus exsolution lamellae of hematite. Near the margins of the grains, the hematite has weathered out, leaving empty voids as shown in Figure 3A. One of the zircon populations consisted of pinkish euhedral crystals (Fig. 3B) with length-to-width ratios in the range 2:1 to 3:1; these had hafnium contents of 1.0-1.3 wt % Hf02. The other zircons are colorless, have length-to-width ratios in the range 3:1 to 4:1, and have hafnium contents of 0.75-0.90 wt % Hf02. The two populations of zircons as well as the monazite grains are generally euhedral, commonly display growth zoning, and commonly contain small liquid and solid inclusions. The kyanite occur as thin bluish rectangular plate-like fragments up to 5 mm in length that commonly contain inclusions of rutile, ilmenite, and possibly magnetite (Fig. 3C). Magnetite is present only in trace quantities but has been observed in the form of a 1 mm long, doubly terminated octahedron. Much of the magnetic fraction consists of small, hollow spheres of unknown origin (Fig. 3D). As these consist of iron and iron oxide in a somewhat eutectic-like intergrowth, we suspect that they may be a by-product of an iron refining or foundry operation. Furthermore, these appear very similar to the steel shot particles that are attributed to metal refining and processing plants in The Particle Atlas (McCrone et al., 1969). A few 1-2 mm grains of cassiterite were observed and confirmed by means of the electron microprobe. These are very similar in appearance to the much more common yellow-brown, roughly equant grains of staurolite. Small numbers of approximately 0.1 mm gold grains were recovered during the processing of approximately one ton of the average, in-place, Citronelle beds at the Bonsai and Hedrick sand and gravel operations. On the basis of these samples, the average gold content appears to be approximately 0.00247 gm/ton (or 0.000079 troy oz per ton)= 0.003 mg per kilogram (Callahan and Craig, 1989). In contrast, the sampling of the riffles at the washer of the Hedrick operation has yielded over 9000 gold grains. The quantity of raw sand and gravel from which these were concentrated is unknown and could easily exceed one million short tons. The 20 kg of concentrate (after removal of scrap iron and pebbles greater than one-fourth inch (6.5 mm) yielded approximately six grams of gold; this corresponds to a grade of approximately 9.6 troy oz/ton in this concentrate). Although not intended to be statistically rigorous because many of the very smallest gold grains were not recovered, an approximate distribution of gold grains is given in Table 2. The largest single gold nugget recovered was very irregular and measured 3 x 5 mm and weighed 0.408 gm. This nugget contained about 25% by volume quartz; hence, the gold portion 124 Table 1. Approximate weight percentages of heavy minerals from the minus 40-mesh fraction, in the Citronelle Formation in the area of study. From Callahan and Craig (1989). Mineral Weight Percentage llmenite 50 Altered ilmenite 17 Kyanite 11 Zircon 5 Rutile 5 Magnetic (mt +iron spheres) 4 Hornblende 3 Staurolite 2 Epidote 1 Monazite 0.5 Other (including cassiterite and gold) 1.5 125 Figure 2. Examples of paleoplacer gold from the Lilesville, NC, deposits. (A) Typical example of fme-grained gold containing an assortment of irregular and flattened grains. (B) Highly irregularly-shaped, large gold grain. The grid scale visible is in millimeters. 126 Figure 3. Heavy mineral components from the Lilesville, NC, paleoplacer deposits. (A) Photomicrograph of ilmenite from the Lilesville deposits. Note that the exsolved hematite lamellae (white) have been weathered out near the margins of the ilmenite grains. Width of field of view is 0.6 mm. (B) Typical examples of the "pink" zircons, the more common of the two zircon populations. (C) The kyanite laths typically contain small inclusions of ilmenite and rutile. (D) Magnetic metallic spheres believed to represent contaminants from iron refining operations or from steel bird shot. The grid visible in each photograph is in millimeters. 127 Table 2. Approximate size distribution of gold grains recovered from the Citronelle Formation gravels in this study; based on representative sample of 2726 grains. Size Range No. Grains Percentage <0.1 mm 0.1-0.5 mm 0.5-1.0 mm 1.0-2.00 mm 2.0-5.0 mm >5.0mm 1240 885 508 75 15 3 45.5 32.5 18.6 2.8 0.7 0.1 128 was estimated to be approximately 0.39 gm. The second largest gold nugget was 0.148 gm. Several hundred grains weighed 4-5 mg, but the average grain weight was probably only about 0.6 mg. The gold grains display a wide variety of shapes and a moderate range of colors independent of their sizes. Most grains are crudely equant to very slightly flattened as is characteristic of placer gold grains that have been transported considerable distances in streams. There are also numerous gold grains in the form of thin disks that might be evidence of intense pounding during transport. In contrast, a moderate number of grains (generally larger than 1 mm) display angular to sharp and delicate points and irregular projections that display little or no sign of sedimentary abrasion or pounding. One to 2% of the grains that are greater than 0.5 mm display crystal faces or face-like surfaces; it is not clear if these represent gold crystal faces or are the imprints of previously adjacent minerals. The most surprising observation was the discovery of approximately 20 completely to nearly completely terminated gold crystals, all in the size range of <0.5 mm (Fig. 4). Most of these crystals are dodecahedrons that are very well developed and are nearly equant. Very few display any evidences of abrasion or distortion. GOLD CO:MPOSIDONAL AND TEXTURAL DATA In order to carry out ore microscopic examination of the grains and to conduct electron microprobe analysis, randomly selected gold grains were mounted using cold setting epoxy and polished using conventional techniques (Craig and Vaughan, 1981). The individual gold grains were mounted in small custom-cut cavities carved into a blank epoxy section. After grains were placed in their respective holes, a thin coating of fresh epoxy was placed over the grains to cement them into the section for polishing. This technique allowed the mounting of large and small grains so that all would be exposed simultaneously during grinding and polishing. The compositions of the grains were determined using point or small area raster (5 llffi2 to 10 J.l012) analysis. One hundred individual analyses were carried out on 62 grains from two operating sand and gravel pits; Table 3 presents 65 analyses of 47 representative grains from the Hedrick mining operation after discarding obviously faulty analyses and duplicates; Table 4 contains 17 analyses of 15 randomly picked grains from the Bonsai mining operation. The analyses have been tabulated in order of descending fineness of the grain or, ifrimmed, the core of the grain. Copper was analyzed in all grains and mercury in most grains. Both copper and mercury contents are generally very low and show no discernible correlation with gold fineness. Copper was found to be anomalously high in grains no. 20 (0.66%), no. 15 (1.19%), and extraordinarily high in grain no. 11 (13.09%) in Table 3 and in grains no. 2 (3.81%) and no. 7 (1.44%) in Table 4. Many mercury analyses were 0.10% or less, but five values exceed 0.20%; the highest from the Hedrick deposit was 0.52% (no. 46), and the highest from the Bonsai deposit was 0.92% (no. 10). Optical examination using reflected light revealed that all grains were not internally homogeneous, but that many possessed "rims" or regions that displayed different colors than the interior. These rims varied from nearly continuous to very discontinuous and from only a few microns to nearly 100 microns in thickness. Although the colors of the rims were sometimes only subtly different from the cores, it was evident that the boundaries between the rims and cores were extremely sharp, although very irregular, as is shown in Figures 5 and 6. The rim compositions (Tables 3 and 4) ranged compositionally from 836 fine (No. 46 in Table 3) to 999 fine (Nos. 31, 33, 33 in Table 3) with many rims exceeding 990 fine. The high fineness rims exhibit a relatively high degree of porosity, whereas the 129 Figure 4. Gold crystals, many completely euhedral, are found disseminated among the paleoplacer gold grains. The absence of erosion or impact textures on these crystals indicates that either they must be very close to their source area or that they formed or were recrystallized after deposition in the paleoplacer. The largest c1ystal face visible in this figure is 0.5 mm in its largest dimension. 130 Table 3. Electron microprobe analyses o~oldrfoains from the Hedrick Mine in the Citronelle Formation near Liles le, orth Carolina. Analyses arranged in descending order of fineness of grain cores; all data in weight percent. Grain No. Au Ag Cu Hg Total Fineness }a core 2b core 3a core 4C core 5 core 6C core 7C core 8c core 9d core 10 core 11 core 12 core 13 core 14 core 15 core 16 core 17 core 18 core 19 core 20 core 21 core 22 core 23 core 24 core 25 core 26 core 27 core 28 core 29d core 30 core rim 31 core rim 32 core rim 33 core rim 34 core rim 100.12 97.99 99.51 99.73 101.24 102.18 101.28 100.23 98.38 101.13 88.33 99.03 98.60 101.29 95.33 98.59 96.46 98.18 93.58 96.45 96.36 98.42 99.94 97.53 94.78 96.21 94.25 97.00 91.85 91.19 91.64 89.97 102.52 89.73 102.49 89.06 100.37 89.67 102.77 a Surface of flattened gold disk. b Surface of gold sphere. c Interior of flattened gold disk. d Surface of gold crystal. - Not analyzed for 0.00 0.00 0.05 0.07 0.28 0.36 0.45 0.53 0.55 0.59 0.87 1.75 2.22 2.40 2.46 2.83 3.00 3.24 3.31 3.86 3.96 4.41 4.44 5.35 5.26 5.42 5.53 5.90 7.54 9.02 6.97 10.62 0.07 10.63 0.08 11.93 0.08 12.06 0.32 0.04 0.00 0.01 0.00 0.06 0.25 0.06 0.08 0.00 0.22 13.09 0.23 0.08 0.17 1.19 0.10 0.01 0.14 0.10 0.66 0.05 0.05 0.07 0.09 0.08 0.26 0.11 0.08 0.07 0.03 0.13 0.09 0.00 0.02 0.10 0.40 0.00 0.03 0.03 100.16 97.99 99.57 99.80 101.58 102.79 101.79 100.84 98.93 0.11 102.05 102.28 101.01 100.89 0.01 103.87 98.98 0.38 101.90 99.47 101.57 0.10 97.09 100.98 0.27 100.64 0.00 102.88 0.11 99.55 0.01 102.98 0.17 100.29 0.16 102.05 0.00 99.89 0.08 103.06 99.46 100.24 98.74 100.68 102.59 100.38 102.67 101.39 100.45 0.03 101.79 0.14 103.26 1000 1000 999 999 997 997 996 995 994 994 990 983 978 977 975 972 970 968 966 962 961 957 955 948 947 947 944 943 924 910 929 894 999 894 999 882 999 881 997 131 Grain No. 35 core rim 36 core rim 37 core rim 38 core rim 39 core rim 40 core rim 41 core rim 42 core rim 43 core rim 44 core rim 45 core rim 46 core rim 47 core rim Table 3 (continued). Au Ag Cu Hg 89.33 12.10 0.08 0.00 97.62 1.53 0.06 84.81 12.23 0.08 0.10 99.51 0.79 0.03 0.12 88.04 13.29 0.05 0.01 102.07 0.97 0.02 0.17 84.34 13.07 0.02 0.01 99.35 0.49 0.04 0.14 86.29 13.46 0.10 87.53 13.24 0.02 86.62 13.49 0.13 97.80 3.75 0.09 85.20 13.60 0.03 0.01 98.39 0.78 0.02 0.06 87.58 14.19 0.04 100.76 1.67 0.00 86.24 14.13 0.11 92.77 8.09 0.23 86.13 15.60 0.07 102.69 0.90 0.12 84.22 16.48 0.10 98.50 1.94 0.11 80.35 19.50 0.03 0.52 84.45 16.61 0.05 0.17 77.76 22.42 0.00 98.89 0.15 0.14 Total Fineness 101.51 881 99.20 985 97.22 874 100.45 992 101.39 869 103.23 991 97.44 865 100.02 995 99.85 865 100.79 869 100.23 865 100.65 963 98.84 862 99.25 992 101.81 861 102.43 983 100.48 859 101.08 920 101.80 847 103.71 991 100.80 836 100.55 981 100.40 805 101.28 836 100.04 776 99.18 999 132 Table 4. Electron microprobe analyses of gold grains from the W.R. Bonsai Mine in the Citronelle Formation near Lilesville, North Carolina. Analyses arranged in descending order of fineness of grain cores; all data in weight percent. Grain No. Au Ag Cu Hg Total Fineness 1 core 2 core 3 core 4 core 5 core 6 core 7 core 8 core 9 core 10 core 11 core 12 core 13 core rim 14 core 15 core rim 97.61 0.64 0.13 0.31 98.68 994 94.15 0.81 3.81 0.01 99.15 992 97.51 2.66 0.11 0.00 100.27 973 96.33 3.10 0.10 0.00 99.53 969 94.36 3.44 0.13 0.00 97.92 965 92.15 4.76 0.09 0.00 96.99 951 91.73 5.67 1.44 0.17 99.01 942 92.53 5.68 0.10 0.00 98.31 942 89.26 6.17 0.12 0.04 97.91 936 87.63 8.27 0.07 0.92 96.90 914 89.68 9.78 0.02 0.00 99.48 902 81.76 15.58 0.03 0.00 97.38 840 80.06 16.06 0.03 0.04 96.19 833 97.99 0.64 0.00 0.20 98.83 994 75.57 22.65 0.03 0.08 98.33 769 75.69 23.86 0.02 0.00 99.57 760 98.04 0.17 0.05 0.00 98.27 998 133 850 0 5 1 0 1 5 20 25 Distance in Micrometers Figure 5. Many of the gold grains exhibit development of partial to complete rims of higher purity gold at the margins of grains. (A) A reflected light photomicrograph showing the presence of a higher-purity rim (991 fine) on an homogeneous grain with a fineness of 869. The width of the field of view is 0.6 mm. (B) A plot of the fineness from a grain interior through a higher-purity rim (grain no. 37 in Table 3) as determined by electron microprobe analyses along a traverse. The analytical spot size was approximately 1 J.UI1 across, and the steps were 1.0 J.UI1 apart. 134 Figure 6. The nature of the gold enrichment in the rims of grains is evident in images of grain no. 37 of Table 3 on which the rim averages approximately 40 J.U11 in thickness. (A) Backscattered electron image; the slightly lighter area is the gold-enriched margin. (B) The X-ray map showing the distribution of gold. (C) The X-ray image showing the distribution of silver. The width of the field of view is 0.2 mm. (D) The marginal areas of this grain exhibit a high degree of porosity where the gold-enriched rims are developed. The width of the field of view is 1.1 mm. 135 marginal areas where rims are not devleoped exhibit little or no porosity. This porosity is evident by examining the rim areas in the grain (shown in Fig. 6D). The presence of higher purity gold rims on cores of lower fineness was first reported by Porcher (1882) who observed them on placer grains from the Brush Creek locality in Montgomery and Floyd Counties, Virginia. Porcher (p. 3) states, "The metal is in small rounded grains, which have almost exactly the color of fine gold and full metallic luster; but specific gravity is low (15.46) , and on cutting the grains, or scraping them with a knife, it becomes apparent that the color first observed is superficial only, the metal in the intelior being almost white." [The analysis gave gold -65.31, silver 34.01, copper 0.14, iron -0.20 and quartz 0.34.] This phenomenon has subsequently been noted by several authors, including Johnson and Uglow (1924), M.S. Fisher (1935), N.H. Fisher (1945), Stumpfl and Clark (1965), Ramdohr (1965, 1969), Desborough et al. (l970a,b), and Knight and McTaggert (1986), and Cook and Nelen (1989) and has been found to be present in many localities in the southeastern United States (Groen, 1987) Examination of Figure 7, a plot of the data in Table 3, demonstrates that rims were recognized on 17 (or 38%) of the 47 representative and randomly selected grains chosen for analysis from those recovered at the B.V. Hedrick operation. Although there is no correlation of rim fineness with the composition of the underlying core of the grain on which it has developed, it appears that a composition of about 920 fine may represent a general threshold for rim development Every grain with a core fineness of lower than 920 exhibited some degree of rim development, whereas no grain with a core fineness above this level displayed any rim. Most rim compositions (see Table 3 and note the filled black squares on Fig. 7) have been verified by duplicate analyses. However, some of the true compositions of the thinnest rims may be higher than the analyses indicated, because the analytical volume of the electron spot may be larger than the thickness of the rim. Such analyses would report as a mixture of the compositions of the high purity rim and the lower fineness core. The gold grains from the Bonsai deposit (Table 4 and Fig. 8) reveal a somewhat different pattern in terms of rim development Although the grains display a compositional range similar to that of the Hedrick deposit, only two grains were found to exhibit the development of rims. lt is not known if this reflects differences in the site of recovery (the steel grate of the trommel at the Hedrick versus the accumulated heavies at the bottom of the sand screws at the Bonsai) or differences in the transportation or sites of deposition. Traverses of closely spaced (1-3 micrometers) analyses by electron microprobe, as well as randomly spaced spot analyses, have revealed that the interiors of grains are quite homogeneous. Most grains varied by less than 10 fineness units (1 wt% Au); the maximum variation observed was less than 20 fineness units (2 wt % Au). DISCUSSION The origin of the gold recovered from the gravel and sand deposits near Lilesville, NC, is not presently known. Most of the grains display rounded and flattened morphologies indicative of long distances of fluvial transport Other grains, however, retain many sharp and delicate edges that do not appear to have suffered any abrasion or pounding as would be expected from significant transport. Furthermore, the several subhed.ral and euhed.ral gold crystals show no evidence of abrasion. Such crystals certainly could not have experienced any appreciable transport. Accordingly, it appears that the crystals and perhaps some of the irregular grains with sharp margins have either undergone little tran.sport or have been recrystallized or precipitated in situ. 136 1000 0 0 [] [] 0 0 [] [] [][] [] ccc [][] 950 [] Cc c cc [] [] [] [] [] []I (f) 900 (zzwwf) LL 850 800 [] CORE COMPOSITION RIM COMPOSITION [] I 5 10 15 20 25 30 35 40 45 SAMPLE NUMBER Figure 7. Plot of the fineness of 47 randomly selected and representative gold grains from the Hedrick deposit. The compositions of the cores of the grains (numbered as in Table 3) are shown by the open boxes in order of descending fineness. The compositions of the corresponding rims on the grains are plotted in filled boxes. 137 1000 oa a co 950 0 a a a 0 (f) 900 a (zzwwf) u.. 850 0 800 a CORE COMPOSITION RIM COMPOSITION tl 750 5 10 15 20 25 30 35 40 45 SAMPLE NUMBER Figure 8. Plot of the fineness of 15 randomly selected and representative gold grains from the Bonsai deposit. The compositions of the cores of the grains (numbered as in Table 4) are shown by the open boxes in order of descending fineness. The compositions of the corresponding rims on the grains are plotted in filled boxes. 138 Many workers have noted the tendency for the grains size of placer gold grains to decrease downstream from the source area and for grains to become more flattened with distance of transport. Other workers (e.g., Smith, 1913; Ferguson, 1924; Fisher, 1945) have also reported that the fmeness of placer gold grains increases with transport. The distance of transport experienced by the gold grains in the Lilesville deposits is unknown, but the highest fineness measured in grain interiors are, in general, found in the grains whose morphologies (i.e., thin, flattened disks) are most consistent with significant fluviatile transport. The mechanism by which high purity rims develop on placer gold grains is unknown. Desborough (1970) and Knight and McTaggart (1986) speculated that the rims formed through a leaching mechanism whereby the silver is selectively dissolved, leaving a residual gold-enriched zone. Such a mechanism presumably requires diffusion of silver through the gold-silver alloy to the metal-solution interface where the selective solution could occur. Groen (1987), however, has calculated the profile that would be expected to develop as a result of such diffusion and found that the boundary should be a relatively smooth transition from the internal core composition to the pure rim. The actual boundary between the core and rim of each grain on which one is developed is exceptionally sharp, as is visible in Figure 5. This degree of sharpness is very similar to that observed by one of the authors (JRC) and his co-workers on gold grains from numerous localities within the United States. Boyle (1979, p. 384) favors gold-rim formation by precipitation from gold-bearing solutions, and Giusti (1983, 1986) describes the occurrence of porous, "new gold" that he believed to be forming on the exteriors of grains. Mann (1984) and Webster and Mann (1984) describe the formation of dendrites of gold and gold-rich rims developing on nuggets during lateritic weathering; these are attributed to the reprecipitation of secondary gold by reduction from gold chloride-bearing solutions. Warren (1982) has described gold crystals forming in soil zones above placer deposits; he attributed their formation to precipitation from gold lode soluble by cyanide produced by cyanogenic plants. The irregular distribution of the gold-enriched rims, the sharpness of the boundaries between the rims and the cores of grains, and the occurrence of euhedral and unabraded gold crystals are all consistent with a mechanism by which gold is precipitated on preexisting gold grains either during transport or after deposition in the placer. The higher fineness rims on the Lilesville paleoplacer nuggets exhibit a considerable degree of porosity not observed within the cores of the grains nor along the margins of grains without rims. Such porosity is consistent with a silver leaching model of formation but could also result from the precipitation of "new gold" such as suggested by Giusti (1983). ACKNOWLEDGMENTS We are grateful to Hedrick Industries and W.R. Bonsai for permission to sample, work and report upon the gold recovered from their operators. We are indebted to Todd Solberg for the electron microprobe analyses, to John Groen for the SEM photograph, to Marjorie McKinney for sample preparation, to Margie Sentelle for typing the manuscript, and to Llyn Sharp for her help in figure preparation. This research has been supported by the Department of the Interior's Mineral Institutes program administered by the U.S. Bureau of Mines under allotment grant no. G1174151 given to Virginia Polytechnic Institute and State University. 139 REFERENCES Berry, W.E. (1916) The flora of the Citronelle Formation. U.S. Geol. Surv. Prof Paper 98-L, p. 193-108. Boyle, R.W. (1979) The geochemistry of gold and its deposits. Geol. Survey Canada Bull. 280, 584 p. Brown, P. et al. (1985) Geologic Map of North Carolina. North Carolina Dept. Natural Resources and Community Development. Butler, J.R. and Fullagar, P.D. (1975) Lilesville and Pageland plutons and the associated meta-rhyolites, eastern North Carolina slate belt. Geol. Soc. Am., Abstracts with Program, v. 7, p. 475. Callahan, J.E. and Craig, J.R. (1989) The potential for heavy minerals in the Lilesville, North Carolina, sand and gravel deposits. In A.J.W. Zupan, ed., 24th Forum on the Geology of Industrial Minerals, Proc. Vol., in press. Cook, R.B. and Nelen, J.A. (in press) The occurrence and exploration significance of gold in Cretaceous sediments of the Alabama coastal plain. In C.M. Lesher, R.B. Cook, and L.S. Dean, eds., Gold Deposits of Alabama, Geol. Surv. Alabama Bull. Cooley, T.W. (1970) Post-Cretaceous stratigraphy of the central sandhills region, North and South Carolina. Unpub. Ph.D. dissertation, Univ. North Carolina, 137 p. Craig, J.R. and Vaughan, D.J. (1981) Ore Microscopy and Ore Petrography. Wiley Interscience, New York, 406 p. D'Agostino, J.P. and Rowe, W.D. (1986) Mineral Occurrences of the Charlotte 1 x 2 Quadrangle, North Carolina and South Carolina. U.S. Geol. Surv., Misc. Field Studies Map MF-1793. D'Agostino, J.P. and Whitlow, S.I. (1989) Gold in fossil placers in Triassic sedimentary rock of the Wadesboro basin. In J.E. Gair, eds., Mineral Resources of the Charlotte 1 x 2 Quadrangle, North Carolina and South Carolina. Geol. Surv. Prof. Paper 1462, p. 163-166. D'Agostino, J.P., Whitlow, J.W., and Perkins, T. (1987) Locations of native gold seen in panned concentrates from Charlotte 1 x 2 quadrangle, North Carolina and South Carolina. U.S. Geol. Surv. Open-file Report 87-367,27 p. D'Agostino, J.P., Whitlow, S.I., and Gair, J.E. (1989) Colluvial deposits of gold and other minerals. In J.E. Gair, eds., Mineral Resources of the Charlotte 1 x 2 Quadrangle, North Carolina and South Carolina. Geol. Surv. Prof. Paper 1462, p. 160-162. Desborough, G. (1970) Silver depletion indicated by microanalysis of gold from placer occurrences, western United States. Econ. Geol., v. 65, p. 304-311. Desborough, G., Raymond, W.H., and Iagmin, P.J. (1970) Distribution of silver and copper in placer gold derived from the northeastern part of the Colorado Mineral Belt. Econ. Geol., v. 65, p. 937-944. Ferguson, H.G. (1924) Geology and ore deposits of the Manhattan District, Nevada. U.S. Geol. Surv. Bull. 723, 163 p. Fisher, M.S. (1935) The origin and composition of alluvial gold, with special reference to the Morobe Goldfield, New Guinea. Bull. Inst. Mining and Metall. (London), No. 365, 46 p. Fisher, N.H. (1945) The fineness of gold, with special reference to the Morobe Goldfield, New Guinea. Econ. Geol., v. 40, p. 449-495, 537-563. Giusti, L. (1983) The distribution, grades, and mineralogical composition of gold-bearing placers in Alberta. Unpublished M.S. thesis, Univ. Alberta, Edmonton, Alberta, 396 p. Giusti, L. (1986) The morphology, mineralogy, and behavior of "fine-grained" gold from placer deposits of Alberta: sampling and implications for mineral exploration. Can. J. Earth Sci., v. 23, p. 449-495. 140 Groen, J.C. (1987) Gold-enriched rims on placer gold grains: An evaluation of formational processes. M.S. thesis, Virginia Polytechnic Institute and State University, Blacksburg, 72 p. Johnson, W.A. and Uglow, W.L. (1924) Placer and vein gold deposits of Barkerville and Cariboo District, British Columbia. Geol. Sur\r. Canada Mem. 149, 246 p. Knight, J. and McTaggart, K.C. (1986) The composition of placer and lode gold from the Fraser River drainage area, southwestern British Columbia. Can. Geol. J. CIM, v. 1, p. 21-30. Mann, A.W. (1984) Mobility of gold and silver in lateritic weathering profiles; some observations from Western Australia. Econ. Geol., v. 79, p. 38-49. Matson, C.G. (1916) The Pliocene Citronelle Formation. U.S. Geol. Surv. Prof. Paper 98-L, p. 167-192. McCrone, W.C., Draftz, R.G., and Deily, J.G. (1969) The Particle Atlas. Ann Arbor Science Pubs Inc., 406 p. Porcher, S. (1882) On an interesting specimen of native gold, from Montgomery Co., Virginia. The Virginias, v. 3, p. 3. Pretorius, D.A. (1981) Gold and uranium in quartz-pebble conglomerates. Econ. Geol., 75th Anniv. Vol., p. 117-138. Ramdohr, P. (1965) Rheingold als Seifenmineral. Baden-Wtirtemberg Geol. Landesamtes Jahreshafte, v. 7, p. 87-95. Ramdohr, P. (1969) The Ore Minerals and Their Intergrowths. Pergamon Press, Oxford, England, 1174 p. Sharp, W.E. (1988) Auriferous terrace gravels on Little Fork Creek, Chesterfield County, South Carolina. In D.T. Secor, ed., Southeastern Geological Excursions, Guidebook for Geological Excursions held in connection with meeting of the Southeastern Section of Geological Society of America, p. 19-23. Smith, P.S. (1913) The fineness of gold in the Fairbanks District, Alaska. Econ. Geol., v. 8, p. 449-454. Stumpfl, E.F. and Clark, A.M. (1965) Electron-probe microanalysis of gold-platinoid concentrates from southeast Borneo. Trans. Inst. Mining and Metall., v. 74, p. 933946. Warren, H.V. (1982) The significance of a discovery of gold crystals in overburden. In A.A. Levonson, ed., Precious Metals in the Northern Cordillera. Assoc. Explor. Geochem., p. 45-51. Webster, J.G. and Mann, A.W. (1984) The influence of climate, geomorphology and primary geology on the supergene migration of gold and silver. J. Geochem. Explor., v. 22, p. 21-42. 141 THE USE OF RELICT PHASES WITH PLACER GOLD IN CHARACTERIZING LODES Alan J. Driscoll, Jr.,* Donald L. Hall,** and James R. Craig** *5247 Dunleigh Drive, Burke, Virginia 22015 **Department of Geological Sciences, Virginia Polytechnic Institute and State University Blacksburg, Virginia 24061 ABSTRACf Placer gold deposits represent the earliest exploration guides to lode deposits. It has long beenrecognized thatplacerdeposits ofgold, platinum group metals orotherore minerals indicate the presence of lode deposits upstream, and that the individual placer minerals may provide information on the distance of transport, residence time, and lode composition. Single-phase placer grains or associated detrital minerals generally dominate the fine fraction of sediments, but careful examination commonly reveals the presence of significant numbers of multiphase grains. Quartz is the most frequently observed phase attached to placer gold, but a variety of other phases have also been encountered. The presence of these "relict" phases provides an opportunity to extract valuable information on the nature of the lode deposit. Exploration programs commonly involve soil and stream sediment sampling, followed by drilling. While drilling yields essential information on the geology, grade and tonnage of a deposit, it also requires significant fmancial commitment. An intermediate exploration phase involving careful analysis ofplacer samples can provide useful data about the deposit without major commitments such as land acquisition and drilling. Examination of placer gold from the Brush Creek deposit in southwestern Virginia serves as an example. Fluid inclusion and petrographic studies of relict phases attached to placer gold grains have constrained not only the timing, but also the P-T conditions ofthe lode formation. Relict quartz textures indicate that gold mineralization post-dated mylonitization associated with the Fries fault. Furthermore, fluid inclusion compositions suggest affinities for both detachment fault and epithermal type gold deposits. INTRODUCfiON The discovery of many major gold mines and districts has resulted from the initial disovery of placer gold in streams or rivers draining the area where primary gold-bearing deposits occur. For example, the first gold mine in the United States (the Reed Mine in North Carolina), and mines in the Mother Lode District, California, resulted from the prior discovery ofplacer gold. Undoubtedly, the search for placer gold served as one ofthe earliest and most effective means of exploration. The presence of gold in the alluvium has long been recognized as indicating the existence of a lode deposit upstream. Other minerals in the alluvium, spatially related to the ore minerals may provide information about the lode deposit, but the problem, ofcourse, is that their relationship to the ore minerals is often poorly understood. This situation has been recognized in many provenance studies, including those on the world's largest gold deposits in the Witwatersrand basin of South Africa (Hallbauer and Kable, 1982). 142 In addition to indicating where a lode may occur, placer grains of gold, platinumgroup metals (PGM), and other placer ore minerals can help to characterize the lode. In order for this to have merit, the placer grains must retain chemical or physical characteristics, or signatures, of the lode. Cabri and Harris (1975) have applied the concept of chemical signatures to alluvial PGM alloys with considerable success. They have shown convincingly that there is a strong relationship between the composition of placer PGM and the inferred source rock. Those alluvial grains derived from concentrically zoned or "Alaska Type" intrusions have Pt* 100/ (Pt+Ir+Os) ratios greater than 86, while those derived from ophiolites, or "Alpine Type" deposits have Pt* 100/ (Pt+Ir+Os) ratios of 36 or less. Oearly, though they may have been partially altered, these alluvial samples have retained at least some chemical characteristics of the source lodes. The composition of alluvial gold is not as useful as that of alluvial PGM in evaluating associated lodes, primarily because the ranges and causes of variability of natural gold finenesses are not well understood, and because gold occurs in a wide variety of geological environments. Nevertheless, studies of the variations of natural gold compositions and differences between districts have begun to reveal some potential utility for exploration (Antweiler and Campbell, 1977; Craig and Rimstidt, 1985; Craig and Solberg, 1986). Commonly however, alluvial gold is found with other attached phases that retain both chemical and physical characteristics of the lode from which they were derived. These "relict" phases may provide important information about that lode. This paper will discuss the variety of relict phases that are found with placer gold, the information these phases can offer, and as an example, will discuss what has been learned by this type of work done on alluvial gold grains from the Brush Creek deposit in southwestern Virginia. The term ''relict" will be used throughout the paper to describe phases that have remained attached to, or intergrown with alluvial gold since they were weathered from the lode deposit THE UTILITY OF RELICf PHASES Mineralo~y and Texwres Although the vast majority of placer gold grains consist solely of electrum, a significant percentage of these grains in most placers have small fragments of other minerals included or attached. Quartz is the most commonly reported relict phase found with alluvial gold, but feldspars (Foster et al., 1978; this study), micas (this study), ilmenite (Cox and Singer, 1986; this study), magnetite (Cox and Singer, 1986) pyrite (this study), garnet (Barker, 1986), and rock fragments (Barker, 1986) have also been found. The association of relict phases with alluvial gold may be used to suggest and eliminate possible types of lode deposits from which the gold could have been derived Relict phases may have certain characteristics that can be identified in the placer grain , and then sought in either stream sediments or outcrops. These characteristics may include mineralogy, deformational textures, fluid inclusion character, color, trace element content, and cathodoluminescence. The presence of relict quartz has limited utility in constraining the conditions of ore formation, because quartz is common in such a wide range of geological environments. Nevertheless, quartz can still provide useful information in terms of its texture. Although the quartz may suffer significant abrasion during weathering, this process does not alter its internal structure. Hence, any initial metamorphic or deformational textures are preserved, 143 even in very small quartz relicts. Furthermore, quartz commonly contains fluid inclusions that may preserve information o the conditions and mechanisms of formation. Relict feldspar can not only help constrain the source of the alluvial gold, but can also serve as a geothermometer (Smith, 1974). Furthermore, the structural state of feldspars can be a useful indicator of the environment of ore deposition (Cerny and Chapman, 1986). For instance, crystallization at high temperatures followed by rapid cooling will result in a highly disordered feldspar, while crystallization at lower temperatures or very slow cooling will probably yield more highly ordered feldspar. Relict oxide minerals can permit the estimation of limits on oxygen and/or sulfur activity, depending on the particular phases present. Sulfide minerals, though commonly coexisting with gold in lodes, do not readily survive the rigors of fluvial transportation. Their presence as relicts generally indicates close proximity to the lode source; furthermore, they also permit estimations of limits on sulfur and/or oxygen fugacity. The presence .of a relict phase with gold does not necessarily mean that the two phases are cogenetic, it only means they coexisted in the lode. As with any interpretation then, care and discretion must be exercised. Fluid Inclusions and IsotOJ?eS The nature of fluid inclusions associated with many different types of ore deposits and their utility as petrologic tools has been documented in hundreds of papers and is well summarized by Roedder (1984). Accordingly, the study of fluid inclusions in relict quartz grains may help to not only characterize the trapped fluid, but also to constrain the possible types of deposits represented by the lode. Of particular interest are the compositions of the fluid inclusions. Figure la, prepared from many sources in Roedder (1987), illustrates C02-NaCl relationships in major ore deposit types. The presence of C02 or other noncondensible gases may be quickly determined by crushing experiments, and in some instances may be quantified rnicrothermometrically, or through micro-Raman techniques. Salinities of appropriate inclusions can be measured on a standard heating/freezing stage. Homogenization temperatures, while not as useful as compositional data, can still be important in distinguishing between deposit types. Wilkins et al. (1986) have grouped some major ore deposit types as a function of the salinity and homogenization temperatures of their fluid inclusions (Fig. lb). The relict minerals attached to gold grains may be sufficient to provide isotopic data on the host lode (for instance, oxygen isotopic measurements can be made with as little as 5 mg of quartz). These measurements, coupled with temperature estimates and information on isotopic fractionation can provide an estimate of the isotopic composition of the fluid responsible for mineralization. This in turn may elucidate the source of the fluids and hence the environment of ore deposition. Alternatively, if two coexisting minerals (e.g., two sulfides) can be separated and the fractionation between this pair is known, an estimate of temperature, as well as the isotopic composition of the fluid, can be made. This temperature estimate can then be combined with fluid inclusion compositions and densities to provide an estimate of pressure. In summary, techniques commonly employed in advanced petrologic studies, but considered to be inapplicable to placer deposits, take on a new importance with the application to exploration, as these techniques can yield crucial information for minimal costs. 144 50 C" C1) 40 -0zas 30 0 ~ >-~ 20 !z::: :::i 10 DETACHMENT -20 FAULT-, RELATED ' ...1 z1- METAMORPHIC :::i ) 4 u ~ 2 Q) ~ 0 ~ Q) 8 ~ ~ 6 4 2 n Kirby el.al., 1988 Te -50 to -60 ~ This study Undeformed Qtz Te -21 to -23 Ea Deformed Qtz ,--,-,-'f"'-'-'~ I -35 -30 -25 -20 -15 -10 -5 0 Temperature C Figure 11. Ice melting temperatures of fluid inclusions in barren, mylonitic quartz (top), and auriferous quartz (bottom) from Brush Creek, Virginia. 157 3 ~ 1 e:enJn ~ 0 a. 0 TYPE 1 100 200 300 400 TEMPERATURE (C) Figure 12. Isochores for fluid inclusions in barren (Type 1) quartz, and auriferous (Type 2 and 3) quartz. Schematic uplift path for Brush Creek rocks is initially isobaric, and then strongly isothermal. 158 that the two-phase inclusions partially decrepitated during'uplift in response to internal over-pressures. The brittle nature of decrepitation suggests thatit occurred at relatively low !empe:atur~s and prob.ably fairly late in th~ uplift history: ~~e relationship between mclus10n s1ze and the mternal pressure reqmred for decrepitation (Bodnar et al., 1989) indicates that about 2 kb of internal over-pressure is necessary to initiate decrepitation in these inclusions. This observation has significant implications for P-T conditions of trapping and uplift paths, as described below. Type 2 inclusions display final melting temperatures of -0.4 to -6.6C (Fig. 11), corresponding to salinities of 1 to 10 wt% NaCl equivalent (Hallet al., 1988). Reliable eutectic temperatures could not be measured on these inclusions due to their small size. As most inclusions were one-phase liquids, and vapor-absent phase transitions are generally metastable (Roedder, 1984), melting temperatures were measured on inclusions that were already two-phase, or that were made two-phase by partial decrepitation during heating. The compositions of initially two-phase inclusions and those that were rendered two-phase in the laboratory, are similar, further supporting that the two-phase Type 2 inclusions have partially decrepitated. Crushing studies of Type 2 inclusions revealed that very minor quantities of gases are dissolved in the aqueous phase. When one-phase, Type 2 inclusions are opened during crushing, this gas exsolves, is expelled into the surrounding immersion oil, and dissolves into the oil in a few seconds. The vapor phase expelled from each inclusion has a diameter 1-2 times that of the original inclusion. If the aqueous portion of the inclusion is taken to be pure H20, the density of C02 at 1 atmosphere and 25C indicates that less than 0.5 mole % gas is dissolved in these inclusions. Raman spectroscopy on both one- and twophase Type 2 inclusions failed to detect any C02, Cf4, N2 or sulfur species. This is consistent with our crushing tests, because the detection limit for most of the above species with Raman spectroscopy is on the order of 1-2 mole%. Type 3 inclusions are 2-5 J.lm in size, occur in both deformed and undeformed (i.e. late-stage) quartz, and are associated with gold in undeformed quartz. They are similar to Type 2 inclusions in that they occur along healed intergranular and intragranular microfractures and are mostly one-phase liquids. Two-phase liquid+ vapor inclusions show evidence for partial decrepitation, much like Type 2 inclusions. Apparent eutectic tempera- tures of -21 to -23C were measured for Type 2 inclusions, suggesting NaCl KCl as the dominant salts. Final melting temperatures of ice range from -10 to -15C (Fig. 11, bottom). Modelling these fluids in the system NaCl-KCl-H20 suggests total salinities of 14 to 22 wt% (Hallet al., 1988). The melting temperature of hydrohalite or sylvite along the hydrohalite-ice, or sylvite-ice cotectics, respectively, could not be measured due to the small size of these inclusions. Accordingly, Na/K ratios could not be determined. However, Tm ice of -14.5C implies NaCl/(NaCl+KCl) weight ratios greater than 0.4 (Na/K ratios> 0.85). Crushing tests revealed amounts of dissolved gases comparable to those found in Type 2 inclusions. Isochores were constructed for the three fluid inclusion populations based on their estimated compositions and densities (Fig. 12). Because the isochores of CaC12-NaClH20 solutions are not well-known, these inclusions were treated as NaCl solutions of approximately equivalent salinity and density. The densities of one-phase, Type 2 and 3 inclusions were approximated by assuming the solution was vapor-saturated at room temperature (i.e., that the Th of these two inclusion types is 25C). Internal pressures greater than 1 bar, as verified by crushing studies, suggest that the true isochores may be displaced slightly to higher pressures at a given temperature. This will result in under estimation of trapping pressures if a given trapping temperature is assumed. 159 InterPretation The contrast in deformation states between the auriferous and barren quartz clearly indicates that the gold mineralization post-dated the mylonitization. Furthermore, the two varieties of auriferous quartz (undeformed and moderately deformed) suggest either multiple episodes of mineralization, or that mineralization extended over an interval that was only partly synchronous with deformation. This is born out in part by our fluid inclusion studies. The euhedral relict limonite pseudomorph in one of the alluvial gold samples, the interstitial sulfide mineralization of the breccia (Fig. 7), and the presence of euhedral quartz fragments found in stream sediments with high gold values, all suggest that some of the gold mineralization occurred as open-space ftlling. Integration of fluid inclusion data with mineralogical and textural criteria places reasonable constraints on the P-T conditions of the gold mineralization, and on the P-T path experienced by rocks at Brush Creek (Fig. 12). The common assemblage in the mylonite zone, of albite + chlorite epidote, the rarity of biotite, and the lack of garnet suggests that the prevailing metamorphic grade recorded is lower greenschist facies. Furthermore, the fact that chlorite usually defines the mylonitic foliation in these rocks suggests that this grade was attained during mylonitization. The highest temperature recorded by the mylonitization then, was that of lower greenschist facies or, conservatively, 350C. With this temperature, and the isochore estimated for Type 1 fluid inclusions, the pressure during late mylonitization must have been on the order of 2-2.5 kb. The estimated isochores for Type 2 and 3 inclusions, the types associated with gold mineralization, coupled with the observation that a differential pressure of about 2 kb was attained between Type 2 and 3 inclusions and the matrix during uplift, requires that the uplift path be initially isobaric and then strongly isothermal (Fig. 12). These relationships also imply minimum and maximum temperatures of gold mineralization of 150C and 200C, respectively, and suggest that accompanying pressures were near 2 kb. The data presented in the foregoing discussion indicate that gold mineralization at Brush Creek was probably epithermal, and associated with brittle fracture of rocks within the Fries fault zone, but clearly post-dated the latest mylonitization associated with the fault This precludes the possibility of a "shear zone-type" deposit, and accordingly, has significant implications regarding any proposed exploration strategy. For example, a shear zonetype ore body would likely be tabular, and conformable to the shear zone, whereas a veintype epithermal deposit could have any number of orientations and/or morphologies. CONCLUSIONS Although the technique of studying placer samples to characterize lode deposits may be applicable in only a limited number of cases, it can serve as an excellent exploration method where it is applicable. Recognition of the presence of certain relict phases may lead to more effective soil or stream sediment sampling programs. Recognition of certain textures in relict phases may suggest tectonic or structural relationships that could have influenced mineralization. Finally, knowledge of the P-T conditions during mineralization, as determined from fluid inclusion studies, isotopic studies, other geothennometers, geobarometers, or even geochronometers may help reveal the temporal and spatial relationship of the mineralization to the regional geology. Any of this information could have a significant influence on an exploration strategy. 160 For relatively little cost, and no commitment, a great deal of information can be obtained, in appropriate situations, by careful analysis of placer samples. The techniques are well established in various geological disciplines, but the application of these techniques to placer samples for exploration has been unrealized. ACKNOWLEDGMENTS We are grateful to Todd Solberg for his assistance in the microprobe analyses, to Llyn Sharp for her help with the photography and graphics, to Jean Cline and Ron Sheets for early reviews of the manuscript, and to Cheryl Knight for the Raman analyses. Steve Mittwede's critical review, and suggestions from Don Rimstidt, Larry Turner and Robert Cook are greatly appreciated. Kim Joseph deserves special thanks for her patience. This research has been supported by the Department of the Interior's Minerals Institute Program administered by the U.S. Bureau of Mines under allotment grant number G1174151 given to Virginia Polytechnic Institute and State University. Funding was also provided by the Virginia Division of Mineral Resources. REFERENCES Antweiler, J.C. and Campbell, W .L., 1977, Application of gold compositional analyses to mineral exploration in the United States. J. Geochem. Expl., 8, 17-29. Barker, J.C., 1986, Placer gold deposits of the Eagle Trough, Upper Yukon River Region, Alaska. U.S. Bur. of Mines, Inf. Circ. 9123, 23 p. Bodnar, R.J., Binns, P.R., and Hall, D.L., 1989, Synthetic fluid inclusions. VI. Quantitative evaluation of the decrepitation behavior of fluid inclusions in quartz at one atmosphere confining pressure. J. Met. Geol., 7, pp. 229-242. Cabri, L.J. and Harris, D.C., 1975, Zoning in Os-Ir alloys and the relation of the geological and tectonic environment of the source rocks to the bulk Pt: Pt+lr+Os ratio for placers. Can. Mineral., 13, 266-274. Cerny, P., and Chapman, R., 1986, Adularia from hydrothermal vein deposits: Extremes in structural state. Can. Mineral., 24, 717-728. Craig, J.R., and Rimstidt, J.D., 1985, Gold: Compositional variations in naturally occurring alloys. Geol. Soc. Am., Abstracts with Programs, 17, 55. Craig, J.R. and Solberg, T.N., 1986, Compositional signatures in some Appalachian gold occurrences. Proc. Int.. Min. Assoc., 13th meeting, Varna, Bulgaria, Pub. House of Bulgarian Academy of Sci., p. 213-222. Cox, D.P., and Singer, D.A., eds., 1986, Mineral deposit models. U.S. Geol. Surv. Bull. 1693, 379 p. Driscoll, Alan J., Jr., 1989, Lode gold deposit characterization using evidence from stream sediments: An example from Brush Creek, Montgomery County, Virginia. Unpublished M.S. thesis, VPI & SU, Blacksburg, VA, 71 p. Fontaine, W.M., 1882, Notes on Virginia geology: The Brush Creek, Va., gold district,. In The Virginias, 3, 108. Foster, R.L., Foord, E.E., and Long, P.E., 1978, Mineralogy and composition of Jamison Creek particulate gold, Johnsville Mining District, Plumas Co., Calif. Econ. Geol., 73, 1175-1183. Hall, D.L., Sterner, S.M., and Bodnar, R.J., 1988, Freezing point depression of NaClKCl-H20 solutions. Econ. Geol., 83, 197-202. Hallbauer, D.K., and Kable, E.J.D., 1982, Fluid inclusions and trace element content of quartz and pyrite pebbles from Witwatersrand Conglomerates: Their significance with respect to the genesis of primary deposits. In Amstutz, G.C., et. al., eds., Ore Genesis: The State of the Art, Springer-Verlag, Berlin, p. 742-752. 161 Kaygi, P.B., 1979, The Fries Fault near Riner, Virginia: An example of a polydefonned, ductile deformation zone. Unpublished M.S. thesis, VPI & SU, Blacksburg, VA, 165 p. Kirby, C.S., Driscoll, A.J. Jr., Bodnar, R.J., Law, R.D., 1988, Importance of fluid composition in the mylonitization process: A shear zone as a conduit for CaCJ2NaCl brines. Geol. Soc. Am., Abstracts with Programs, 20, A332. Luttrell, G.W., 1966, Base and precious-metal and related ore deposits of Virginia. Va. Div. of Min. Resources, Min. Resources Report 7, 167 pp. Oakes, C.S., and Bodnar, R.J., 1988, Phase equilibria in the system NaCl-CaClrH20: The ice liquidus. Geol. Soc. Am., Abstracts with Pro$famS, 20, A390. Roedder, E., 1984, Fluid inclusions. In Ribbe, P.H., ed., Mineral. Soc. Am., Reviews in Mineralogy, 644 p. Roedder, E., ed., 1987, Fluid Inclusion Research, Vol. 20. The University of Michigan Press, Ann Arbor, 559 p. Smith, J.V., 1974, Feldspar Minerals, Vol. 1. Springer-Verlag, New York. Sterner, S.M., and Bodnar, R.J., 1984, Synthetic fluid inclusions in natrnal quartz I. Compositional types synthesized and applications to experimental geochemistry. Geochim. Cosmochim. Acta, 48, 2659-2668. Sweet, P.C., 1980, Gold in Virginia, Va. Div. Min. Resources, Pub. 19, 77 p. Sweet, P.C., 1983, Virginia gold-Resource data. Va. Div. Min. Resources, Pub. 45, 196 p. Werre, R.W. Jr., Bodnar, R.J., Bethke, P.M., and Barton, P.B. Jr., 1979, A novel gasflow fluid inclusion heating/freezing stage. Geol. Soc. Am., Abstracts with Programs, 11, 539. Wilkins, J., Jr., Beane, R.E., and Heidrick, T.L., 1986, Mineralization related to detachment faults: A model. In Beatty, B. and Wilkinson, P.A.K., eds., Arizona Geol. Soc. Digest, 16, 108-117. 162 A GEOCHEMICAL INVESTIGATION OF THE MASON MOUNTAIN SPERRYLITE OCCURRENCE, MACON COUNTY, NORTH CAROLINA Robert B. Cook* and James R. Burnell** *Department of Geology, Auburn University, AL 36849 **E.C. Jordan, Inc., P. o. Box 7050, Portland, ME 04112 ABSTRACT Sperrylite United States as (Pt a Arasr2e) was initially discovered in the accessory mineral recovered during the placer mining of gem corundum and rhodolite garnet in streams draining Mason Mountain, Macon County, North Carol ina. In-place sperry! i te was subsequently found associated with sulfide-rich garnet-biotite gneiss near the summit of the mountain. The potential economic significance of the occurrence was ignored until 1982 when it was evaluated by stream-sediment and rock geochemistry. Of 38 representing panne a 50 d km 2 c oncentrate, stream-sediment sa area, ten contained Pt in amounts mples up to 200 ppb. Arsenic contents varied between 2 and 8 ppm. Pd did not exceed 5 ppb in any sample. The highest Pt values were for samples collected from streams draining the western one-half of Mason Mountain. Coincidentally collected silt stream-sediment samples did not contain Pb, Zn, or cu in anomalous amounts. Rock analytical da.ta were uniformly negative with respect to Pt-group elements, with the exception of weakly foliated biotite-garnet-pyrrhotite rock exposed in bald, iron-stained outcrops along the southwest slope of the mountain, a sample of which contained 15 ppb Pt. Sulfide- rich rock from a prospect midway up the south slope of the mountain's west end contained up to 0.81% cu, 3.62% Pb, 0.55% Zn, 36 ppm Ag, and no detectable Pt-group metals or Au. The Co, Ni, and Cr contents of sulfide-rich rock samples are relatively low, reaching maximum values of 80, 95, and 300 ppm, respectively. Siliceous units associated with sulfide occurrences are anomalously aluminous and magnesian with values ranging up to 27.1% sillimanite-biotite schist aAndl 2og3aleannda-b1e3a.4r%ingMggOa for garnetrnet-biotite gneiss, respectively. The sperrylite occurrence does not appear to represent ore-related mineralization and is of mineralogical interest only. INTRODUCTION Platinum group metals were reported as rare constituents of heavy mineral concentrates produced during the early years 163 of gold mining in both Georgia (Stephenson, 1871) and North carolina (Genth, 1891) . The occurrences included rather diverse terranes of both the Blue Ridge and Piedmont Provinces. Although no significant attempt was made to identify the source of the platinum, its occurrence in Georgia was verified almost a century later in streams draining poorly known crystalline rocks generally north of the Brevard Zone in White County by Hurst and Otwell (1964) and in Habersham County by Hurst and Crawford (1964). was In an event unrelated to gold mining, discovered in 1894 at the mouth of sperrylite Ned Wilson (PBtrAans c2h) along the Caler Fork of Cowee Creek, about 5 miles north of Franklin, Macon County, North Carolina (Hidden, 1898; Hidden and Pratt, 1898; Kemp, 1902). The discovery was accidental, resulting from a heavy mineral study of material found during the alluvial mining of corundum and the gem garnet rhodolite. Fifty-five sperrylite crystals, the largest being 0.2 mm in diameter, were initially submitted to S. L. Penfield for identification (Hidden, 1898). The sperrylite occurred in both nugget-like masses and cube-octahedral crystals. According to Hidden, sperrylite was also identified in heavy mineral suites from four other gem placer operations both east and west along the Caler Fork of Cowee Creek. A subsequent paper by Hidden and Pratt (1898) described the separation of 300 sperrylite crystals from heavy mineral concentrate collected from a new location on Mason Branch, about two miles southwest of the initial discovery site. The total weight of these crystals was one milligram with the largest crystal being 0.4 mm in diameter. The heavy mineral suite containing the placer sperrylite was unusual, containing minor gold, relatively abundant corundum, bronzite, cordierite, aluminosilicates, and various spinel group minerals. Ultimately, the apparent source of the placer sperrylite was discovered by Hidden and Pratt 0.8 km north of the Mason Branch gem placer near the summit of Mason Mountain. Here, crystals of sperrylite were panned from weathered residuum derived from an outcrop of garnet-biotitesulfide gneiss. Subsequent published data relating specifically to the Mason Mountain sperrylite occurrence are lacking. Henderson (1931) described the mineralogy of material collected at the "rhodolite quarry" on the west end of the mountain, an occurrence apparently unrelated to that containing the sperryli te. Heinrich ( 1950) made corrections in the mineralogical work of Henderson (1931) and proposed a paragenetic scheme for the quarry rhodolite occurrence. Neither paper describes significant sulfide minerals and Henderson points out that the locality of his study is areally and perhaps geologically distinct from that described by Hidden and Pratt (1898). 164 It appears that the very presence and possible economic significance of the Mason Mountain platinum occurrence has been overlooked by modern exploration geologists, primarily due to the age of the original reports. Consequently, the present study was undertaken to identify the initial in-place sperry! i te discovery site and to generate preliminary geochemical data upon which future work could be based. General Geology Mason Mountain is within a poorly-characterized portion of the western North Carolina Blue Ridge Province lying between the Hayesville and Shope Fork faults (fig. 1). The general tectonic setting of the area has been described by Hatcher and others (1980). Immediately adjacent areas have been described in somewhat greater detail by Mcsween and Hatcher (1985) and Kittelson and McSween (1987). Projection of units shown on maps included in these works indicates that the Mason Mountain area is underlain by undifferentiated rocks of the Coweeta Group and Tallulah Falls Formation. The study area is shown as biotite schist and gneiss, and layered gneiss and migmatite on the geologic map of the Knoxville 1:250,000 quadrangle (Hadley and Nelson, 1971). On the most recent North Carolina state geologic map (North Carolina Department of Natural Resources and Community Development, 1985) the area is included within a general zone of Precambrian hypersthene granulite dominated by locally migmatitic biotite gneiss interlayered with biotite gneiss and amphibolite. METHODS An approximately 50 km2 area centering around Mason Mountain was evaluated by panned stream-sediment concentrate and silt geochemistry, reconnaissance geologic mapping, and rock geochemistry. The examined area includes the southeast part of the Alarka 7-1/2" Quadrangle, the southwest part of the Greens Creek 7-1/2" Quadrangle, the northwest part of the Corbin Knob 7-1/2" Quadrangle and the northeast part of the Franklin 7-1/2" Quadrangle (fig. 1). Heavy mineral concentrate stream-sediment samples were collected at 38 sites by carefully panning approximately equal initial volumes of gravel dug from active channel deposits. Concentrates were analyzed commercially for Pt, Pd, Au, and As. Pt and Pd were determined by neutron activation analysis with 5 ppb detection limits. Au was analyzed by standard fire assay preparation and atomic absorption spectrophotometric (AAS) finish with a detection 1 imit of 5 ppb. Arsenic was determined by standard AAS techniques with an analytical detection limit of 1 ppm. 165 Figure 1 Figure 1. General location of Mason Mountain (MM), Macon County, North Carolina. The 7. 5-minute quadrangles shown are the Alarka (A), Greens Creek (GC), Corbin Knob (CB) and Franklin (F) . 166 Coincidentally collected silt samples were analyzed for cu, Pb, and zn by standard AAS techniques with a detection limit of 1 ppm for each element. Rock samples were analyzed commercially for a variety of geochemical parameters. In addition to Pt, Pd, Au, and As determinations by the above described techniques, a number of rock samples were analyzed for Cu, Pb, Zn, Ag, Ni, and Co by standard AAS methods with 1 ppm detection limits. Obviously mineralized samples were analyzed by neutron activation for Ru, Rh, Pt, Re, Os, Ir, and Pd with detection limits of 5, 1, 5, 1, 3, 0.1, and 5 ppb, respectively. Neutron activation whole rock analyses were obtained commercially for rocks associated with sulfide mineralization. RESULTS The presence of trace Pt in streams draining Mason Mountain is confirmed by stream sediment geochemistry. Drainages containing anomalous Pt are not characterized by coincident Pd, As, cu, Pb, or Zn in elevated amounts. Reconnaissance mapping on Mason Mountain has identified previously unreported prospects on its south slope and west end. Rock samples from the south slope prospects contain very anomalous Cu, Pb, Zn, Ag, and Au. A pyrrhotite-rich sample collected from a prospect area associated with bald sulfidic outcrops contains trace (15 ppb) Pt. Stream-Sediment Geochemistry Heavy mineral concentrate stream-sediment geochemistry (Table 1) has confirmed the presence of sporadic trace Pt in streams draining both the north and south flanks of Mason Mountain (fig. 2) Ten of 38 panned concentrate samples contain Pt in amounts up to 200 ppb. Nine of these ten anomalous samples represent streams originating on the flanks of Mason Mountain. No sample contained Pd in amounts greater than the analytical detection limit. The As content of all samples is low, ranging between 2 and 8 ppm. Trace amounts of Au are widely distributed throughout the area, all samples containing detectable Au in amounts up to 10,300 ppb. Gold was visible in all samples containing greater than 1000 ppb Au. Silt samples collected coincidentally with panned concentrate samples contained cu, Pb, and Zn in low amounts ranging up to 94, 31, and 129 ppm, respectively (Table 2). There is no apparent statistical correlation between Pt, Au, As, cu, Zn and Pb in stream sediment samples, nor is there a 167 Table 1. Panned stream-sediment concentrate analytical data, Mason Mountain area, Macon County, North Carolina Sample # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Ml M2 M3 M4 Pt ppb 150 N.D. N.D. 200 N.D. N.D. N.D. N.D. 100 N.D. 100 100 N.D. N.D. 100 100 100 N.D. N.D. N.D. 100 N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. 100 N.D. N.D. N.D. N.D. Pd ppb N.D. 1 N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D. Au ppb 10300 15 6200 15 140 15 30 5 25 10 30 10 10 5 5 10 10 5 60 15 5 5 345 150 35 35 100 40 15 10 15 5 15 5 65 40 15 50 1Below analytical detection limit. As ppm 4 6 6 6 4 6 3 2 3 2 3 3 3 2 3 4 3 3 3 6 3 3 2 2 3 8 6 3 2 4 3 2 4 3 3 3 3 2 168 ~ J 0 1 Kilomaters ( 2 . ) . / . I "fa2 ( .1_.......--........___ .... 0) co . 30 -}-.... \. .,....B""' 129 / '(8 . EXPLANATION 0 Stream-sediment sample site 0 Pt below detection limit 100200ppb Pt X Prospect 83"22'30H I / _../ ' - <""-- '\..23 \ -35"15" / -~8 J 19I'20' I ( ) I / . / (\ . \ ( ) . i / Figure 2 Figure 2. stream-sediment and rock sample locations on parts of the Alarka, Greens Creek, Corbin Knob, and Franklin 7. 5-minute quadrangles, Macon County, North Carolina. Mason Mountain is indicated by the 725-meter contour line. Table 2. Stream-sediment analytical data, Mason Mountain area, Macon County, North Carolina Sample # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 M1 M2 M3 M4 Cu ppm 39 31 42 82 27 86 25 24 29 20 94 51 31 46 33 22 37 27 10 14 37 37 72 24 31 43 39 28 13 37 21 25 19 20 92 78 88 85 Pb ppm 10 13 8 6 13 7 5 13 7 6 11 6 5 5 9 9 19 9 6 6 18 5 10 6 25 31 26 12 15 31 11 12 17 14 11 9 12 10 Zn ppm 85 67 105 79 53 87 47 52 59 53 78 62 57 121 69 56 80 60 41 44 78 54 129 45 88 92 80 61 46 87 63 77 48 52 64 56 61 72 170 pronounced geographic association between samples containing relatively large amounts of two or more of these elements. Prospect Examinations Traverses walked across the central and western portions of Mason Mountain resulted in the identification of several previously unreported prospects (fig. 2) . Prospects associated with bald outcrops along the south slope of the mountain are characterized by relatively abundant pyrrhotite in float, outcrop and dump material. Prospects contiguous to and on strike with the rhodolite mine at the west end of the mountain contain only sparse sulfides and apparently were dug in search of gem garnet. These prospects are not considered further in this paper. The most extensive prospect found on Mason Mountain lies approximately 0.5 km north of stream sediment sample site M4 (fig. 2). The prospect is marked by an open cut that ends in what appears to be a caved adit, and dumps suggestive of several hundred feet of underground workings. Dominant rock types observed in outcrop and as dump material are biotitequartz gneiss and schist that contain locally very abundant, coarse-grained garnet porphyroblasts of the rhodolite variety. Sillimanite is locally very abundant. Mineralized dump material contains fine- to mediumgrained galena, pyrite, sphalerite and chalcopyrite in a generally fine-grained, gray siliceous biotite-garnet-quartz gneiss. Analyses of grab and chip samples (NCS-21, NCS-22, and NCS-23; Table 3) indicate the presence of Pb, cu and Zn in amounts up to 23,200 ppm, 8120 ppm and 5500 ppm, respectively. Silver occurs in amounts up to 36 ppm while Au, Pt and Pd are not present in amounts above the analytical detection limit. Polished section examination did not result in the identification of platinum group minerals. Approximately 400 m west of the above described prospect is an area of bald, iron-stained, sulfide-bearing siliceous gneiss that is marked by several shallow prospect openings (fig. 2). Composite chip samples collected at this site (Samples NCS-24 and NCS-25; Table 3) contain low though anomalous cu, Pb, Zn and Ag. Both samples contain anomalous gold but do not contain Pt or Pd in detectable amounts. Polished section examination indicates that pyrrhotite is the dominant iron sulfide mineral with paragenetically later, subordinate chalcopyrite, galena and sphalerite. Sperrylite or other platinum group minerals were not identified in polished sections of samples collected at this site. Examination of float boulders downslope from the general bald outcrop area resulted in the discovery of several 171 Table 3. Rock sample trace metal analyses, Mason Mountain prospect areas, Macon County, North Carolina Sample Pt Pd Au Ag Cu Pb Zn Co Ni As 11 ppb ppb ppb ppm ppm ppm ppm ppm ppm ppm NCS-3 15 N.D. N.D. 1 1460 14 460 41 45 3 NCS-21 N.D. 1 N.D. N.D. 25 5375 10,900 370 15 10 2 NCS-22 N.D. N.D. N.D. 27 5700 23,200 5500 6 8 3 NCS-23 N.D. N.D. N.D. 36 8120 18,900 480 7 9 2 NCS-24 N.D. N.D. 100 56 1050 510 410 20 23 3 NCS-25 N.D. N.D. 245 32 230 76 110 19 26 3 1Below analytical detection limit. NCS-3 NCS-21 NCS-22 NCS-23 NCS-24 NCS-25 massive garnet-biotite-pyrrhotite rock. garnet-biotite-quartz gneiss with fine-grained pyrite, chalcopyrite and galena. very siliceous garnet-biotite-sulfide gneiss. siliceous biotite-sulfide gneiss with very coarse-grained garnet porphyroblasts. massive pyrrhotite-garnet rock. massive pyrrhotite-garnet rock. 172 samples of massive pyrrhotite-garnet-biotite-quartz rock. Analysis of this rock (NCS-3; Table 3) indicates the presence of 15 ppb Pt. The rock is unusual with respect to other analyzed samples in that it contains relatively small amounts of Pb and Ag and relatively large amounts of Co and Ni. Analyses for the other Pt group metals identified trace Ir in amounts up to 0.8 ppb. Polished section examination did not result in the identification of Pt group minerals in these samples. Rocks cropping out in the general area of the south slope prospects are predominantly gneisses that contain variable amounts of garnet, quartz, sillimanite and biotite with locally significant pyrrhotite and orthopyroxenes. Whole-rock analyses of major prospect area rock types are given in Table 4. Sample locations are shown on figure 2. Sillimanite-biotite-garnet gneiss and massive garnet- biotite-pyrrhotite rock (NCS-2 and NCS-3) are characterized MpbyygOr,lrohwI