EconomicGeology Vol. 86, 1991, pp. 960-982

Zoning and Genesisof the Darwin Pb-Zn-Ag SkarnDeposit, California: A ReinterpretationBased on New Data

RAINER J. NEWSERR¾, Departmentof Geology,University of Alaska,Fairbanks, Alaska 99775

MARCO T. EINAUDI, Department of Applied Earth Sciences,Stanford University,Stanford, California 94035

AND HARVEY S. EASTMAN

Bond Gold Exploration,4600 S. Ulster Street,•400, Denver, Colorado80237

Abstract The > 1-million-metric-tonDarwin Pb-Zn-Ag-Wskarn deposit has been previously described as a group of sulfidereplacement bodies zoned away from the Darwin quartz monzonite plutonand formed from magmatic fluids at •325øC. Detailedsurface mapping and available radiometricdata, however,indicate that the Pb-Zn skarnsulfide bodies are appreciably(>20 Ma) youngerthan the Darwin pluton, and undergroundmapping and core loggingindicate there are severalskarn sulfide pipes with strongconcentric zoning. One of the pipesis zoned arounda deepgranite porphyry plug. The pipesexhibit outward zoning in wt percentPb/Zn and oz/ton Ag/wt percentPb (both ratios<0.5 core, >1.0 margin).The pipesshow minera- logicalzoning, with a core definedby higher sphalerite-galena,higher chalcopyrite,darker sphalerite,more abundantpyrite inclusionsin sphalerite,and evidencefor multiple sulfide depositionalevents. In contrast,both graphitein marble and pyrrhotite in sulfideores are zonedaround the Darwinpluton, which suggests that pyrrhotitestability is influencedby pre- Pb-Zn skarn(Darwin pluton related?) bleaching of marblebeds. Garnet zoning is highlycom- plex, with four generationsidentified by petrographicand compositionalrelations; younger garnettypes are moreabundant in upperand lateral parts of the pipe. Retrogradealteration of garnet is concentratedin the upper and laterally distalparts of the skarn,but garnet in apparentequilibrium with sulfideis present throughout the verticalextent of skarn.Systematic mineral compositionalpatterns include outward increase in hedenbergite + johannsenite componentsin clinopyroxene(<2->20 outward),increase in Sb-I- Bi contentsof galena,initial increasefollowed by decreasein Mn contentsof sphalerite(range from <0.2->1% Mn), and an initial increasefollowed by outwardvariable increase and decreasein FeS contentsof sphalerite(range of <3->20% FeS).Previously published sulfur isotope data are compatible with a decreasein sulfurisotope ratios outward around the pipe core.Published isotopic data combinedwith temperatureestimates from phase homogenization and arsenopyrite-sphalerite geothermometryshow a systematicdecrease in temperaturefrom the skarnsulfide pipe center (>425øC) to the margin(<300øC). Comparisonof stopemaps to isothermcross sections indicates that the bulkof minedsulfides were from areassurrounding the pipe core,in whichtemperatures declined from approximately 375 ø to 300øC (gradient of løC/m). Combined mineral compositionand assemblageand sulfurisotope systematics indicate that asthe ore fluidsflowed outward they underwentpro- gressivedecrease in oxidationstate ('•'1 log unit) and increasein pH (2-3 units);upward- movingfluids underwent initial decreasein oxidationstate and increase in pH followedby a reversalto higheroxidation state and lower pH. The processof ore depositionwas chemically complexand may have involvedremobilization of earlier depositedsulfides. Realistic ore depositionalmodels at Darwin require simultaneouschanges in (at least)temperature, pH, and oxidation state. Introduction which wasdirect-shipping, high-grade material, and 0.1 milliontons of W ore. Averagegrades of sulfide THE DARWINdistrict, in southeastCalifornia (Fig. 1), ore mined after 1945 were approximately6 percent containsPb-Zn-Ag as well as Cu- and W-rich veins Pb, 6 percentZn, 0.2 percentCu, and 6 oz/ton Ag and skarns(Hall and MacKevett, 1962; Newberry, (Hall and MacKevett, 1962). The Darwin mine per 1987). Production from the district includes more se consistsof a seriesof isolatedPb-Zn-Ag _ W ore- than 1 million metric tonsof Ag-Pb-Znore, muchof bodies(Defiance, Thompson, Essex, Independence;

036!-0128/91/1248/960-2353.00 960 DARWINPb-Zn-Ag SKARN DEPOSIT 961

KEY --]Tert-Quaternarysediments .:•Granite porphyry,aplite, pegmatite ?• 'Coso-type'granite •tzß rnonzonite- qtzsyenite !-.'•Diorite •/•, -qtz monzonite -•Penn-Permiancarbonate turbidites

ii!..i•Devono-Mis sippian carbonates •t•fm•t• thrtmtfm•t • fold •xi•

Mineralization

ekam•,.I- vein• veinsc•

.Pb-Zn-Ag . X Davis thrusl system Cu_+Au n ß

w o ß c•,/r,ovearm• • • Darwin I = Defiance ae*le'N+ townsite 2 = Essex 117ø36 ' W

O• 0..5i 1I kilometer8

FIG. 1. Generalizedgeologic map of the Darwin district,showing distribution of variousore types. The Darwin mineconsists of severalorebodies located along the westmargin of the Darwin stock;two of the majororebodies are the Essex(1) andDefiance (2). Modifiedfrom Newberry(1987) andStone et al. (1989).

Fig. 2) whichwere linked by a commonhaulage level enization(Czamanske and Hall, 1975), and fluid in- in 1945. clusionsystematics (Rye et al., 1974). Basedon their Althougha relativelysmall deposit, Darwin is no- study,Rye et al. (1974) concludedthat the Pb-Zn-Ag table for a pioneeringS isotopestudy (Rye et al., sulfidestage of mineralizationoccurred at tempera- 1974) in whichthe authorsintegrated light stable iso- turesof 325 ø _ 55øC andinvolved magmatic hydro- tope analyseswith the knowngeology; and for com- thermal fluidsflowing out of the magmawhich pro- parativeattempts at determiningmineral formation ducedthe adjacentDarwin stock.These authors noted temperaturesby techniquessuch as minor element horizontalzoning of S isotopevalues away from the distribution(Hall et al., 1971), sulfidephase homog- Darwin stockand concluded that ore depositiontook 962 NEWBERRY,EINAUDI, AND EASTMAN

EXPLANATION

•[] GRANODIORITECOSO

r'•GRANITEPEGMATITE, PORPHYRY APLITE

'•GRANITE MATRIX BRECClAPORPHYRY-

•_• • SKARN/ENDOSKARN--MATRIX BRECCIA

• ,,=.•MONZONITEDARWIN QUARTZ &

,'• DARWINMONZODIORITEQUARTZ & QUARTZ MONZON L[• QUARTZDIORITE PORPHYRYSYENITE

Z

•. • UNDIFFERENTIATEDHORNFELS MAR.LE

,,Z," .• CHERTYLIMESTONEARO •1 &GRAPHITIC LIME

i_.

SULFIDE-QUARTZ- ,• CALCITE_.+SK•

• FAULT

• THRUSTFAULT-- SAWTEETH ON UPPER PLATE

._....-- CONTACT

FIG. 2. Detailed geologicmap of the Darwin mine area, basedon unpublishedmapping by R. Newberry andT. Sisson(1980-1982). Note the Pb-Zn-Agskarn veins that cut and displacethe Darwin pluton. placeprimarily due to pH increaseat constantfoe and mineralized and virtually unaltered (Hall and temperature,caused by fluid reactionwith carbonate MacKevett, 1962; Newberry, 1987), which may be and calc-silicate rocks. incompatiblewith its beingthe sourceof hydrother- Althoughwidely quoted in the isotopicliterature, mal fluids;and (3) the associationof ore-gradeW with there are severalfeatures of the Darwinstudy as pre- ore-gradePb-Zn-Ag is uncommonin limestonere- sentedby Rye et al. (1974) whichinvite scrutiny: (1) placementand/or skarn deposits (e.g., Einaudiet al., the Darwin quartz monzodiorite,with probablesol- 1981), raisingthe possibilityof multiple, unrelated idustemperatures > 800øC (Piwinskii,1973; New- hydrothermalevents. There alsoare problemswith berry, 1987), is located<50 m from the sulfidede- the relative timing of plutonismand Pb-Zn-Ag ore posits,but if thisstock was the sourceof hydrothermal deposition;the 174-Ma (U-Pb; Chen, 1977) Darwin fluids, then temperaturesof sulfide ore formation stockis cut by the Davisthrust (Eastman, 1980; New- could be expectedto range up to severalhundred berry, 1987), with movementdated at 154 to 148 degreeshigher than 325 ø +_55øC and skarn-related Ma (Dunne et al., 1978), but Pb-Zn-Agores cut the ores shouldbe present; (2) the Darwin stockis un- thrust and hence postdate thrusting (Hall and DARWINPb-Zn-Ag SKARN DEPOSIT 963

MacKevett,1962; Eastman,1980) andmust be at least (Sylvester et al., 1978). Contact metamorphicre- 20 m.y. youngerthan the Darwin stock. crystallizationof impurecarbonate rocks formed ido- Newberry (1987) resolvedsome of theseapparent crase-wollastonitecalc-silicate hornfels, garnet-rich conflictsby showingthat there are severaldifferent skarnoid, and bleached marble around the Darwin plutonicsuites associated with differentmetasomatic stock (Fig. 2; Hall and MacKevett, 1962; Eastman, skarnsuites in the Darwin district. Newberry (1987) 1980; Newberry, 1987). Tungsten-bearingskarns are further showedthat the contact-typeW skarnsare locally presentalong contactsof the more differen- geneticallyrelated to the Darwin stockand that the tiated unitsof the Darwin stockwith the surrounding Pb-Zn-Ag veins and skarnsare related to younger carbonate-bearingrocks and are distributedsymmet- quartzporphyry bodies. Supporting geologic data in- rically aroundthe Darwin stock(e.g., Fig. 2; New- cludethe presenceofPb-Zn skarnsalong faults which berry, 1987). Thrustingalong the Davis fault system cut acrossand displacethe Darwin stock(Figs. i and (Fig. 1) took place at 154 to 148 Ma (Dunne et al., 2) and graniteporphyry and brecciabodies (Fig. 2) 1978) and resultedin 1- to 3-km eastwarddisplace- which intrude the Darwin stock and contain clasts of ment of the upper plate,juxtaposing the Cosobatho- W-bearingbut Pb-Zn-poorskarn. Newberry (1987) lith andits adjacentCu skarnswith the Darwin pluton suggestedthat the Pb-Zn-Ag-Wore associationmay and its adjacentW skarns(Newberry, 1987). Subse- have resultedfrom superimpositionof Pb-Zn skarns quent to thrusting,a seriesof graniteporphyry dikes on older W skarns. and brecciapipes (Fig. 2) intrudedthe metamorphic • Thispaper investigates the apparentcontradiction and igneousrocks along the northwestmargin of the betweenisotopic studies of Rye et al. (1974), which Darwin stock; aplite geobarometryindicates these concludedthat ore waszoned around and caused by rockscrystallized at a pressureof approximately0.5 the Darwin stock, and petrologic-geologicstudies kbars (depth -- 1.5 km, assumingP = lithostatic; presentedby Newberry (1987), whichindicated that Newberry, 1987). Pb-Zn-Ag skarns(restricted to the Pb-Zn-Agmineralization was unrelated to the Darwin west sideof the Darwin pluton)and Pb-Zn-Agveins stock.As there are nopublished maps for the deposit of the Darwin district formed alongsteeply dipping in which the skarnsare distinguishedor described faults,which are commonlymarginal to the granite andbecause pyroxene-rich skarns--so characteristic breccia bodies and, less commonly,along granite of Pb-Znskarns (Einaudi et al., 1981)--have notbeen porphyry dike contactsand alongthe Davis thrust reportedfrom the Darwinarea, the investigationbe- (Fig. 2). Minor extensionalreactivation of the Davis gan with systematicsurface and undergroundmap- thrust accompanyingCenozoic basin range uplift pingand core logging. Because Rye et al. (1974) had (Dunneet al., 1978) causedslight deformation of ores stressedthe horizontalzoning of oresand sulfuriso- in the vicinity of the thrust. tope ratios around the Darwin stock,a further inves- tigation of zoning-•based on ore and calc-silicate Investigativetechniques minerals,mineral compositions,and metal ratios-- This studyis basedon underground(100-900 lev- wasundertaken. Given the new resultsfrom zoning els) and surfacemapping of the Darwin mines at studies,the thermaland sulfur isotope models of Rye 1:2,400, detailedlogging of approximately1,000 m et al. (1974) werereinterpreted. The zoningdata of diamonddrill core,examination of approximately ported herein are mostconsistent with a seriesof sub- 300 polishedand thin sections,approximately 450 vertical skarnpipes and related bedded skarnsand microprobeanalyses representing 80 thin sectionsand veins, centered 100 to 300 m west of the Darwin grain mounts,compilation of assaymaps and drill as- stock.The sulfurisotope data are consistentwith a says,and sulfurisotope measurements for a galena- modelof verticalfluid flowthrough and away from sphalerite pair and two galena samples.Mineral thesepipes. abundanceswere estimatedusing standard thin sec- tion point-countingtechniques. Electron microprobe Generalgeology analyseswere performed using a 9-spectrometerARL Darwin area regional geologyis summarizedin microprobeat the Universityof California,Berkeley Dunne et al. (1978) and Stoneet al. (1989). Upper (silicates),as described in Newberry(1987), anda 3- Paleozoicsedimentary rocks (Stone, 1984; Stevens, spectrometerChimeca microprobeat Washington 1986) in the immediate Darwin mine area were de- State University (sulfides),as describedin Meinert formed into broad folds in late Triassic time and sub- (1987). Mineral standardswere employedand each sequentlyintruded by mid-Jurassicalkalic plutons analyticalpoint representedthe averageof at least (includingthe 174-MaDarwin stock; Chen, 1977) three analyses.Sulfide mineral separateswere ana- and,on the westside, by calc-alkalicgranite plutons lyzedfor sulfurisotope ratios by KruegerEnterprises, (includingthe 156-MaCoso batholith; Chen, 1977). Cambridge,Massachusetts. Mine assay,production, A 4- to 6-km depthof emplacement(Plithostati½ = 1-2 and drill hole assaydata were compiledonto level kbars)has been suggested for the calc-alkalicplutons maps,averaged into 10-m (minimum)blocks, and a 964 NEWBERRY,EINAUDI, AND EASTMAN

1422m

1453m 1361m 1393m : 100 meters.

FIG.3. Generalizedoutlines ofskarn body in theEssex pipe area as a functionof elevation.Variable skarnshapes are due to theintersections of the northwest-trending Essex zone with folded carbonate beds;larger skarn bodies are in the hingeregion of the Darwinantiform. Most of the sulfideis in or immediatelyadjacent to the skarn.Based on undergroundmapping by R. Newberryand G. Wilson (unpub.map, 1982). Surface projections of cross-section lines are shown in Figure2. ratiowas calculated. Ratios were then projected onto skarnappears to replacemarble. Unreplaced marble mine crosssections. Petrographic data and micro- is locallypresent between skarn and the Darwin stock. probeanalyses (including those of Eastman,1980), (Fig. 5). Sulfidesoccur disseminated throughout the analysesofmineral separates, Sisotope data, and sul- skarn,in minorveins in the skarn,as through-going fidedeposition temperatures (including those deter- veins,and as bedded replacement bodies beyond the minedby earlier workers) were plotted on mine cross skarn. sections.Because the data set for this paper integrates severaldifferent data sets acquired at differenttimes by differentworkers and published on different mine crosssections, the data cannot be presented in anideal fashion.Different projections for thepetrographic, metalratio, and microprobe data were employed to minimizeprojection distance, to best comparethe resultsof thisstudy with those of Ryeet al. (1974), and to illustrate the three-dimensionalcharacter of the Essexorebody. TheEssex orebody (Fig. 2) of theDarwin deposit wasselected for intensivestudy because all levels were accessiblefor undergroundmapping and be- causewell-located sample collections from this ore- bodywere available at StanfordUniversity. Geologyof theDarwin Pb-Zn-Ag skarns Pb-Zn-Agores of theDarwin deposit occur pre- KEY: dominantlyasbedded replacements and veins (Hall andMacKevett, 1962; Eastman, 1980) in skarn,calc- •] replacement"Bedded" sulfide silicatehornfels, and marble. There are, in addition, [•1 Sulfide-Qtz-Carb+_Garnet veins at leasttwo majorpipelike orebodies, the Defiance andEssex pipes. The Essex pipe is an irregular body ,• Garnet-dominantskarn .•.• Calc-silicatehornfels elongatedtothe northwest-southeast andplunging at • Pyroxene-dominantskarn '• Mica-quartz hornfels about70 ø to the southwest. Itsshape is controlled by intersectionsof the N 65 ø W Essexfissure zone with •.[• Darwinskarnstock-related • Marble receptivecarbonate units near the hingezone of the 50 meters doublyplunging, N 30ø W-trending, Darwin antiform • Darwinmonzodioritequartz ,;;• Mineworkings (Figs.3 and 4). Major beddedskarn zones also are presentadjacent to the mainpipe in marblebeds im- FIG. 4. Geologicmap of the Essexpipe, 600 level,showing mediatelyunderlying sills of the Darwin pluton. This distributionof variousskarn and ore types and relations to struc- is especiallythe casein the upperlevels of the mine. tureand !ithology. Most of the skarnreplaces marble; some py- roxene-richskarn may replacecalc-silicate hornfels. The bulk of In theEssex pipe area narrow skarn and/or sulfide thesulfides are restricted to the skarns. Mapping by R. Newberry veinsare presentin calc-silicatehornfels; the bulk of andG. Wilson(unpub. map, 1982). DARWINBb-Zn-Ag SKARN DEPOSIT 965

ELEV. A'sw C';C B'B • NE .. • . :-.:.'-.::'.'-.:.... •; ,•r.'.'. ':.:: ß ß ß

...'-:--_.--.-'.•?,.,...-..-'..'.-i-i-i-:-'- [".'-:•tx•\: .'.'.'

•..-..-.-';- ':•. -•- e-'-".••::% •:.-.-.-.-..tl ....•ß""'• // ....:.:...... ' z. '"'%"::,....,• .: ...... :':/•Si j / .:.•.--•'• -""'-'.'"- '-.-.:-:::::::::. . ..•. \.'.'.'. •1 , .. :'_'•.•.- •.'::.:.::::. NO ,...... -. •• •::'•••': ,.•o.•.,o.I.'.'.'.'11 ß ...... :•::-../;.•.*.-.;.;.;.-.....*.-.-.-...... •4/0 meters100 '•, *************ilI

EXPLANATION

FAULT • Sulfide-Qtz-Cavein .-,Garnet -• Darwinstock CONTACT 'Beddedreplacement' sulfide • Dioriteporphyry • Marble [• Garnet-dominantskam • Mica-quartzhomfels • Pyro.xene-dominantseam i• Calc-silicateand skarn hornfels D Undifferentiatedhornfels

FIG. 5. Northeast-southwestcross section through the Essexpipe, at approximatelyright anglesto the pipe elongation,showing pipe morphologyand southwestrake to the pipe. Skarnsdominantly replace marble, but narrow replacementsof calc-silicatehornfels also are present. Massivesulfide replacementsand veins are more abundanttoward the surface;sulfide-bearing skarns are present throughoutthe vertical exposure.Based on undergroundmapping by R. Newberry and G. Wilson (unpub. map, 1982).

Lateraland vertical zoning in the Essexpipe is il- and show no evidence of replacingpyroxene-rich lustratedby a vertical (A;A) crosssection (Fig. 5) skarnat the marblecontact. Pyroxene skarn does not constructedat approximatelyright angles to the contactmarble. Garnet skarnscut acrossand replace northwest-southeastelongation of the pipe (Fig. 3). calc-silicate hornfels and skarnoid beds and form lat- The Essexpipe containssubequal amounts of pyrox- erally extensivestrata-bound replacement bodies in ene and garnet-richskarn at depth;pyroxene skarn marbleand hornfelsadjacent to the pipe (Fig. 5). In is both cut and verticallysupplanted by garnet-rich many casesthe light-colored,green granditegarnet skarns.Coarse-grained, vuggy, garnet-rich skarns are skarnscould not easilybe distinguishedfrom the light- in metasomaticcontact with marblein manylocalities colored, green idocrasegarnet hornfelsduring un- 966 NEWBERRY, EINAUDI, AND EASTMAN derground mapping and AX core logging; most of level the pipe horsetailsinto severalquartz-carbonate- thesecontacts probably are gradational.Rocks were sulfideñ garnetveins which lie alongthe beddingof, classifiedwith confidenceas skarn during under- and partly replace,garnet skarn (Hall and MacKevett, groundmapping and core loggingif they contained 1962; Fig. 2). coarse-grained(green) garnet ñ pyroxenewith inter- Reconnaissancelogging and examination of stitialsulfides and/or obviousgarnet veins. •10,000 m of AX drill core and loggingof 1,000 m With increasingelevation, sulfide-rich, garnet- of BX drill coresuggests that undergroundand surface bearing veins become common; however, coarse- exposuresof ore accessiblein 1980-1982 are rep- grainedgarnet skarn with interstitialgalena is present resentativeof the ore mined and sampledbefore throughoutthe vertical exposuresof the pipe (e.g., 1970. If this is the case, then some generalizations as the surfaceexpression of the Essexpipe; Fig. 2). can be made about the relative amounts of the various Sulfidesmostly occur as disseminationsin the skarn ore typespresent. In the Essexarea, the bulk of ore below the 500 level. Sulfide-richveins commonly (•75%) below the 200 level consistedof sulfidesin containa few to 20 percentfine- to medium-grained skarn,with intergrownsulfide-garnet-pyroxene tex- euhedralgarnet and are surroundedby an envelope tures more common than obvious sulfide veins. Above of garnetñ idocraseskarn where they cut hornfelsor the 200 level, sulfide-bearingskarns and sulfide-rich marble (Eastman,1980). Relationsbetween these (ñ garnet)veins are subequalin abundance.The De- sulfide-richveins and normal, sulfide-bearingskarn fiancepipe is dominatedby sulfidesdisseminated in are problematic;the veinscut skarnbut containeu- skarnbelow the 570 level, with highly retrograded hedralskarn minerals and are grosslytransitional (with skarn,sulfide-replaced marble, and sulfide-richveins depth) to garnet-rich skarn (Fig. 5). We interpret more common than simple sulfide-bearingskarn theseveins as representinga stageof hydrothermal above the 400 level. activitywhich postdatesthe bulk of normalskarn for- Skarn Mineralogy mation but under conditionswhere garnet and ido- crasecontinued to form. A few massivequartz-car- The mineralogyof the Darwin Pb-Zn-Agskarns is bonate-sulfideñ garnet bodies are present as re- relativelysimple, although there are complexmineral placementsof marble adjacentto sulfide-richveins compositionalpatterns. The most abundantmineral (Fig. 5); crosscuttingrelations indicate that these is granditc garnet, with compositionsranging from bodies were formed after the skarns,possibly con- 100 to 5 mole percent andradite(Fig. 6). Four gen- temporaneouslywith the sulfide-richveins. Although erations of garnet have been identified, based on sulfides,especially galena, commonly enclose or fill overgrowthtextures, rare vein relationships,and op- vugsand fractures in skarnminerals, unaltered garnet ticaland compositional properties (Figs. 6, 7). Several is presentadjacent to sulfidethroughout the pipe. garnettypes are typicallypresent in the samesample Very late veins (generally <0.3 m wide), with ex- and in many casesthree generationsare present in tremely coarsegrained calcite and pyrite and with the samegrain (Fig. 7A). sporadicnative gold and tellurides, are especially The earliest garnet (generation 1) possesses commonalong lithologic contactsand faults. These blotchy,irregular birefringence and typicallyoccurs veinswere generallynot mined and are not shownin as cores to later garnet types. These garnetshave the mapsand crosssections. Where the late quartz- compositionsof Ad6o_92with 1 percentspessartine and calciteveins cut skarn,they are surroundedby poorly showsystematic enrichment in andraditecomponent defined0.1- to 1-m-wideenvelopes of calcite-quartz- from core to margin.Blotchy birefringent garnets are hematite alteration of the skarn. surroundedby a narrow rim of yellowish,isotropic The Defiancepipe is a steeplydipping body located 700 m southeastof the Essexpipe (Fig. 2). This part of the Darwin depositis currentlyaccessible only at the 400 and 570 levels,and consequentlyit wasless YELLOWBIREFRINC-•NT "DOUBLEBANDED"BIREFRINGENT ½.... r/me 1-2%S10ess L well studied.On the 400 level it containsquartz, car- ß -- IL'", bonate,bustamite, garnet, and retrograded pyroxene ...3-•.%•ß ß --I• • •.•J with massiveto disseminatedsulfides. Limited logging '"• •" <1% ' SPLOTCHY"BIREFI•NGENT apeas of AX drill coreindicates that garnet-pyroxeneskarns • ...... • r/me are present at deeper levels. Drill core also shows (• 1• 2• 3'0 40 50 6•) 7•) 8'0 9• 1{•0 mole percent Andradlte that a smallbody or bodiesof graniteporphyry (with quartz-Kfeldspar veinlets), porphyry matrix breccia, FIC. 6. Variationsin compositionofPb-Zn skarngarnets, Dar- and skarnmatrix breccia (which contains galena and win mine, basedon electron microprobeanalyses. The order of garnet types, as shown in the diagram, is consistentlyseen sphalerite)occurs below the 1000 level.The granite throughoutthe mine. All garnetshave lessthan ] percentcom- porphyryand porphyry matrix breccia are surrounded bined almandine+ pyrope contents;spessartine contents are as by galena-sphalerite-bearingskarn. Above the 400 marked.Data from Eastman(1980) and this study. DARWIN Pb-Zn-AgSKARN DEPOSIT 967

(generation2) garnetwith compositionsof Ad96-•oo. roxene grains(with similar optical orientations)sur- Andraditegarnet commonly exhibits incipient alter- roundedby murky carbonate+ sulfide;this texture ationto very finegrained opaque minerals. The third may indicate extensiveretrograde alteration of an generationof garnetis bandedbirefringent, essen- originallypyroxene-rich skarn pipe. tially identicalin compositionand appearance to the Pyroxeneassociated with Pb-Zn skarnshows a nar- double-bandedbirefringent garnets of the Capote row rangeof composition,Hdo_3o, with Fe/Mn ratios basin,Cananea district, Mexico, described by Meinert varyingslightly around 2/1 (Fig. 8). Below the 600 (1982). Thisgeneration (Fig. 7A) is characterizedby level in the Essexpipe, pyroxeneis enrichedin iron narrow(.01-.002 mm) bandsof alternatingbirefrino from core to rim; abovethe 600 level pyroxenegen- gentand nearly isotropic garnet, with alternating high erally is depletedin iron from core to rim. Rare, late andlow iron contents.These garnets have composi- pyroxene,occurring with evenlybirefringent yellow tionsof Ad94_6awith 1 to 2 percentspessartine. The garnet,is nearlypure diopside. final(fourth) garnet generation is yellowand evenly Other nonsulfide minerals in the Pb-Zn skarns oc- birefringent.This garnetoccurs as the final rims on cur in severaldifferent modes. Fluorite (especially bandedbirefringent garnets and more rarely asiso- commonin upper parts of the mine) and K feldspar lated grainsin upper levelsof the mine (Fig. 7B). are sporadicallypresent in garnet-linedvugs. Wide- Thesegarnets have compositions of Adas_•with a sys- spreadoccurrence of fluoriteis confirmedby analyses tematic decrease in iron content rimward. Lowest of skarnand horfels grab samples(unpub. Anaconda iron, evenlybirefringent garnets also contain a 5 to Co. data),which indicate 550 to 16,000 ppm F. Wol- 7 percent spessartinecomponent. lastonite is common in the calc-silicate hornfels but Garnetsshow complex relations with sulfides.Ini- is typicallyretrograded to fine-grainedquartz + cal- tial sphaleriteand galena deposition apparently post- cite where near Pb-Zn skarn veins. Idocrase also is datedthird generation banded birefringent garnet, as commonin the hornfels,is present in somePb-Zn thesesulfides commonly occur filling vugs between skarns(remnant metamorphic grains?), occurs in small adjacentbanded birefringent garnets, rather than as amountswith bustamite,quartz, calcite, and grossu- inclusionswithin thesegarnets. The bandedbirefrin- lariticgarnet in the Defiancepipe, andis locally pres- gent garnetsadjacent to interstitial sulfidesshow no ent adjacent to sulfide veins in marble (Eastman, signsof alteration,suggesting sulfides were stable with 1980). Epidote, chlorite, and actinoliticamphibole thesegarnets, but andraditecores to the bandedbi- are commonin the uppermostparts of the mine, refringentgarnets commonly exhibit incipient alter- formedby retrogradealteration of garnetand pyrox- ation to magnetite-calcite-quartz(Fig. 7C). Isolated ene near galenaosphaleriteveins. Ilvaite is reported grainsof andraditegarnet are commonly replaced by (Eastman, 1980) from pyrrhotite-bearingsulfide sulfides+ quartz+ calcite,even in deeperlevels of veins.Chlorite is commonas an alterationproduct of the mine (Fig. 7D). Late, evenlybirefringent, gros- garnetnear calciteopyriteveins. sular-spessartinegarnets, on the other hand, were As discussedin Czamanskeand Hall (1975), the depositedcontemporaneously with sulfidein many ore mineralogyat Darwin is dominatedby sphalerite cases;for example,as tiny (0.5 mm) garnetsinter- andslightly argentiferous galena. Galena in the upper grown with sulfide,magnetite, quartz, calcite, and parts of the mine is enrichedin Ag, Sb, Bi, and Se bustamitein the Defiancepipe (Fig. 7B). Theselate relative to galenain deeper levelsand containssul- garnetsalso are presentas envelopesaround galena fosaltinclusions, interpreted from thermalstudies to veinsin calc-silicatehornfels. Garnet of all typesis be exsolutionproducts. Bismuth-bearing galena and stronglyaltered, typicallyto fine-grained,chlorite- sulfosaltminerals are especiallycommon; Bi contents calcite-epidote-magnetite-pyrite,where adjacentto of skarngrab samplesrange from 74 to 16,000 ppm calcite-pyritefissures. (unpub.Anaconda Co. data).Late carbonate-quartz Pyroxeneis much less commonthan garnet in sulfideveins, with a poorlycharacterized distribution, presentlyknown exposures of the Darwin skarns,and contain tellurium-rich minerals, sulfosalts,and native unlikegarnet, pyroxene deposition probably predated gold.Pyrite is commonin all the ore types;it appears the bulk of sulfidedeposition. Massive pyroxene to be earlier than galena.Pyrrhotite occursin the skarns,only currently known below the 500 levelof westernpart of the Essexorebody, approximately 200 the Essexpipe, contain 0.5- to 2-cm-long,bladed py- m from the main Darwin stockcontact (as measured roxenes,partly replacedby garnetand sulfides (Fig. alongcross sections E•oE and A•-A; Fig. 2). As noted 7E). Pyroxene has not been identified in the Essex by Eastman(1980), pyrite-pyrrhotiterelations are pipe Pb-Zn skarnsabove the 3A level, either because complexin the westernpart of the Essexzone, with it hasbeen totally replacedby garnet+_ sulfide, or, a paragenesisof (1) early pyrite, (2) hexagonalpyr- morelikely, becauseit never formedin the upper rhotite, (3) pyrite + galena+ sphalerite+_ magnetite, portionsof the Essexpipe. In contrast,the Defiance (4) monoclinicpyrrhotite + porouspyrite. Chalco- pipe (400 level) containssmall clots of isolatedpy- pyrite is presentin deeper portionsof the deposit. 9{}8 NEWBERRY,EINAUDI, AND EASTMAN

.4

ß t DARWINPb-Zn-Ag SKARN DEPOSIT 969

JOHANNSENITE Homfels W-skerns Pb-Zn skams and veins

wo!lmstontb , , tdocrue pyroxeue lerne• k-feldsptr ochee!lf.e epldote ß ? hex pyrr'hoUb I mouo p:yTrhoUt.e ersenop•'•e ophalerit.e chalcop•rtte lederie

cMorit. e calctte

FIG. 9. Simplifiedparagenesis diagram for the Darwin ores.

Hall and MacKevett (1962) largely in indicatinga considerableoverlap in time of formationbetween

ß skarn (garnet, idocrase,epidote) and ore (pyrite, DIOP$1DE (E•r=•and Butt, 1982) HEDENBERGITE sphalerite,chalcopyrite, galena). We attribute this FIG. 8. Compositionalvariations in Darwin Pb-Zn skarneli- differencein the diagramsto the availabilityof deep- nopyroxenes(98 electron microprobeanalyses), in termsof the level exposuresand samples for thisstudy and to our Mg (diopside),Fe (hedenbergite),and Mn (johannsenite)end successfuldiscrimination between metamorphic members(mole %), relativeto otherskarns and skarn types. Note (hornfels-related)and metasomatic(skarn-related) the lowJo + Hd contents(unusual for Pb-Znskarns) and the sym- patheticincrease in Jo and Hd (not seenin porphyryCu skarns). talc-silicateminerals. Given that the developmentof The Caporebasin Zn-Cu skarns(Meinert, 1982) have clinopy- hornfelsaccompanied intrusion of the Darwin stock roxeneswith compositionssimilar to Darwin, but even those are (Hall and MacKevett, 1962; Eastman,1980; New- lessMg rich than Darwin pyroxenes.Darwin data from Eastman berry, 1987), structuralconstraints indicate that third (1980) and this study. and fourth generationgarnet and Pb-Zn ore deposi- tion postdatedformation of garnetin the calc-sfiicate Tetrahedrite-tennantiteand arsenopyrite are present hornfelsby more than 20 Ma. in smallamounts (skarn grab samples have As between 30 and 720 ppm), especiallyin upper parts of the Mine-Scale Zonation Patterns deposit.Other rare and trace mineralsare described Constraints and limitations by Czamanskeand Hall (1975). A generalizedparagenetic diagram (Fig. 9) for the Metal zoningstudies in the Darwin minesare com- minerals at Darwin illustrates the series of metamor- plicatedby the factthat assaysand production statis- phic-skarnevents and the complexitiesof the sulfide- tics reflect a combinationof skarn,vein, and super- silicateand sulfide-sulfiderelationships for the main gene-enrichedores. Ratios above the 3A level (Fig. Pb-Zn ore event. This diagramdiffers from that of 7A) are commonlybased on supergene-enrichedores.

FIG. 7. Photomicrographsshowing characteristic features of the Darwin skarns.A. Garnet from skarn on the 400 level, showingmultiple growth events.Nonbanded, andradite core (generation2) garnet with incipient alteration to magnetite+ pyrite, surroundedby (generation3A) bandedrim, in turn roundedand surroundedby (generation3B) new bandedrim, finally surroundedby nonbanded, evenlybirefringent (generation 4) garnet.Transmitted light, 0.35 mm field of view. B. Defiancepipe, 400 level. Detail of turbid calciteassemblage which hasreplaced pyroxene: sphalerite (gray), garnet (generation4; dodecahedrons),and quartz (gray) in calcite (white). Crossedpolars with condenser lens, 0.5 mm field of view. C. Garnet-sphaleriteskarn, 400 level, Essexpipe: garnetswith slightly altered cores(generation 2) and bandedbirefringent (generation3) rims. Vugs formed by coalescing generation3 garnetscontain sphalerite with sulfidation"exsolution" pyrrhotite (blobs) and chalcopyrite inclusions(rectangles). Sphalerite deposition in this samplepostdates, but is in apparentequilibrium width,generation 3 garnet. Transmittedlight with condenserlens, 1.5 mm field of view. D. Garnet- sphaleriteskarn, 700 level, Essexpipe: generation2 garnet (brokengray) partly replacedby sphalerite (dark gray), pyrite (black),and calcite (white). Generation2 garnetis clearly unstableduring sulfide depositionin this sample.Transmitted light, 5 mm field of view. E. Pyroxeneskarn, 600 level, Essex pipe:coarse-grained euhedral pyroxene crystals partly replaced by sphaleriteand galena (black). Crossed polars,5 mm field of view. 970 NEWBERRY, EINAUDI, AND EASTMAN

rB'C' Pb/Zn opt Agl%Pb E SecliOnc.cI ^.^I 1 .

>ø5

E

FIG. 10. North-south(A, B) and east-west(C, D) metal ratio crosssections for the Essexpipe. Pb and Zn are in percent,Ag is in troy ounces/shortton. Note the invertedcup morphology,centered on the skarnpipe (el., Fig. 5), definedby metalratios. Black dots represent location of datafrom unpublished assaymaps, drill logs,and assaybooks, provided by the AnacondaCompany.

If currentexposures are representative,then the bulk samplesfrom the Defiancearea was also taken from of ore below the 3A level was from skarn or skarn- skarn. bearingvein material,but the informationavailable doesnot permit a detailedanalysis of metal ratiosas Metal zoning--Essexpipe a function of host rock. Combinedhorizontal and vertical zoning in the Es- Mineral, isotopic,and mineralcompositional data sexpipe is illustratedby metal ratio crosssections. havebeen derivedfrom severalpublished studies by Figure 10A and B showsmetal zoningalong north- several different authors. The data from these studies southsection C-C' andFigure 10C andD showsmetal havebeen presentedin severaldifferent mine cross zoningalong east-west section B-B' (both located in sections.Detailed sampledescriptions are not gen- Fig. 2). BothPb/Zn and Ag/Pb ratio contours describe erally availablefor isotopicsamples. Underground an inverted cup morphologyfor highestvalues of mappingand core loggingin areasfrom which pre- theseratios. For example,the Ag/Pb contoursare re- viouslypublished studies were performed suggest that markablysimilar in both east-westand north-south the bulk (more than 2/3) of previoussamples from views,both showingdeep skirts of >2 (oz/ton/wt%) the Essex area was derived from skarn ores with the extendingto roughlythe 800 level on all sides.The remainderprimarily from garnet-bearing sulfide veins invertedcup defined by metalratios has a coreof low in skarnand hornfels.Virtually none of the previous values(<0.5 for both metalratios) that corresponds samplesrepresents the late pyrite-ealeiteevent. Ores with the center of the garnet-pyroxeneskarn pipe from the Defiance area are more difficult to catego- (Fig. 5). The horizontalwidth of skarnin Figure 10 rize, giventhe limitedamount of exposuresavailable isapproximately 80 to 120 m (notethat Fig. 5 isdrawn for thisstudy. We suspectthat the bulk of the previous alongthe narrowestportion of the skarnbody; cf. DARWINPb-Zn-Ag SKARN DEPOSIT 971

FIG. 11. Approximatelyeast-west, Pb/Zn (A) and Ag/Pb (B) metal ratio crosssections through the Defiancepipe, below the 400 level. Units as in Figure 10. Note the invertedcup morphologyto the metal ratios, centered on a granite porphyry plug. Locationof granite porphyry determinedfrom loggingdrill core;black dots represent location of data on metalratios from unpublishedassay maps, drill logs,and assaybooks, provided by the AnacondaCompany.

Figs.3 and4), sothat the zonesof highestmetal ratios zone (Fig. 12) is usedto illustrateseveral types of in Figure 10 are commonlyin skarnor the outer mar- mineralzoning patterns around the centralskarn pipe ginsof skarn. as definedby geologicmapping (Figs. 3, 4, and 5) Within the Essexpipe, the high-gradeore stopes andby metalratio studies(Fig. 10). are locatedin areaswith very highAg/Pb and Pb/Zn Galena/sphaleriteratios, basedon underground ratios,and the bulk of ore minedwas within the Ag! mapping,core logging,and hand specimenand pol- Pb andPb/Zn > 1 contours.The localizationof high- ished sectionexamination (Fig. 12A) showthat the gradeore is causedin part by supergeneenrichment skarn pipe has a zinc-rich core and a lead-rich pe- of Ag andPb (Hall andMacKevett, 1962, p. 66) and riphery. Galena/sphaleriteratios are not clearlyde- in part by chemical-thermalzonation within the skarn fined outsideof the orebody,where abundancesof pipe (see below). Although not noted by previous both these mineralsare low relative to pyrite and workers,stope cross sections of the Essexbody outline pyrrhotite,and the relativeproportions of galenato an invertedcup morphology (of. Czamanske and Hall, sphaleritemight be low on the fringesof the orebody. 1975, fig. 1), in agreementwith the metal ratio see- Thesemineralogical ratios are in agreementwith the tions. Pb/Zn ratiosdefined by ore gradesand supportthe hypogeneorigin of the metal zoningpattern within Metal ratios--Defiancepipe the pipes.The more economiczones in the mine are Metal ratiosfor the Defiancepipe crosssection be- definedby highsulfide abundance together with high low the 400 level (D-D', Fig. 11), showa patternre- galena/sphaleriteratios (0.5-3), generallyfound in markablysimilar to that of the Essexbody. Symmet- the upper partsof the deposit. rical zonationis presentaround a deep porphyry Chalcopyritein the Essexpipe skarnoccurs both body,which is surrounded by low sulfideskarn matrix asisolated grains and commonly as very finegrained, brecciathat grades outward to ore-bearingskarn. The subhedral,and oriented grains in sphalerite.Esti- coreof thispipe is definedby low Zn/Pband Pb/Ag mated abundanceof chalcopyritein sphaleritehas ratios,and an invertedcup morphology is definedby beencontoured (Fig. 12B);these data, combined with high (>2) Pb/Zn and Ag/Pb ratios.As in the Essex the distributionof samplescontaining chalcopyrite as area,high values for Pb/Znabove the 400 levelin the separategrains, define a copper-richcore to the ore- Defiancebody partly reflect supergene enrichment. body. Hall (1971) presenteddata for Cu contentof sphalerite;these data also outline a copper-richcore Mineral zoningmEssexpipe to the orebody.The Cu data alsoare consistentwith Mineral zonationin the Essexpipe is clearlyseen limited Cu assays,which indicatevalues up to 0.65 in the distribution of both sulfide and nonsulfide min- percentCu in deeperparts of the mine(in comparison erals.A crosssection (A-A') through the Essexskarn to typicalgrades 9f <0.2% Cu for the ores).The cop- 972 NEWBERRY, EINAUDI, AND EASTMAN

galena/ sphaler ite % chalcopyrite in . aphalerite lchalcopyrite <0.1 grain '-

>2 DARWIN STOCK DARWIN s'r•

<5

>2

.•alerite color (in thin galena-a•ha•r•e section)[3 IMraoenesi8 by oal ß o oran9e

ever•ts

50m•

% Pyrite in sph•erite

DARWIN

.o

(•rnet t•pe' selected

KEY KEY ß DOUBLE o bleachedmarble ß yellow

Ilrl.Jscovile•

eO 'K K undedirled)

/ BIREFR. garnet

I o I •

FIG. 12. Mineralogicalvariations along cross section A'-A (Figs.2 and 5) in the Essexpipe, based on thin section,polished section, and hand specimenstudies. Data pointslocated within 50 m of the cross-sectionline were projectedonto that line, so that the lateral distributionof skarnsamples is greaterthan the lateralextent ofskarn shown in Figure5. A. Galena/sphaleriteratio (note correspondence betweenthe low galena-sphaleritecore and the low Pb/Zn ratiosof Fig. 10C). B. Chalcopyritedistri- bution, showinga high chalcopyritecore similar in locationto the low galena-sphaleritezone. C. Sphaleritecolor (thin sectiontransmitted light). D. Galena-sphaleriteparagenesis. E. Pyrrhotite dis- tribution. F. Distributionof pyrite inclusionsin sphalerite.G. Distributionof garnettypes and garnet zoning.Each shapewithin a compositesymbol represents a differentgarnet type, with the zoningas seenin thin sectionreplicated in the compositesymbol. Note the restrictionof majorandradite (isotropic) garnetto the lower partsof the pipe. H. Distributionof selectednonsulfide minerals, showing boundary DARWINPb-Zn-Ag SKARN DEPOSIT 973 per-richcore (Fig. 11B)corresponds closely with the marginsof the Essex skarn pipe and some of these are zone of lowestPb/Zn ratiosin the ore (Fig. 11A). locatedbetween the skarnand the Darwin stock(Figs. Sphaleritecolor changeswithin the ore zone; 4 and5). The marbleclose to the stockis bleached, sphaleritein the deep, central,part of the mine is whereas marble far from the stock contains dissemi- dark colored,whereas sphalerite in upper and high natedgraphite (Fig. 12H). This distributionpattern fringeparts of the skarnis pale colored(Fig. 11C). suggeststhat bleachingand other contactmetamor- Individual grainsare not color zoned. Comparison phiceffects occurred during intrusion of the Darwin with compositionsdetermined by electron micro- stock(Eastman, 1980); subsequentlyboth bleached probeindicates that thispattern only partly reflects and unbleachedmarble was overprinted by the Pb- variationsin combinediron, manganese, and cadmium Zn skarnhydrothermal system. Eastman (1980) noted contentof sphalerite;another important factor might thatrare graphite in theEssex skarn is restrictedto a be sulfur content(Scott and Barnes,1972). zone located )150 m west of the main Darwin stock Polishedsection study of crosscuttingand inclusion contact.The spatial coincidence of unbleachedmarble texturesinvolving galena and sphalerite indicates that (Fig. 12H), graphitein skarn,and pyrrhotite-bearing in a central zone multiple periodsof ore deposition ores(Fig. 12E) suggests that progressive reduction of led to complexrelative ages,whereas in peripheral the ore fluid by reactionwith organicmatter in the zonesand at depth,sphalerite was replaced by galena distal carbonaterocks controlled the depositionof (Fig. 12D). Thisdepositional history is consistent with pyrrhotiteand graphite in skarn. Hall andMacKevett (1962) andEastman (1980), who Silicate minerals also show systematicpatterns reportgalena generally younger than, but commonly aroundthe core of the orebody.Garnet generations overlappingwith, sphalerite. (cf.Fig. 7A) aresystematically zoned (Fig. 12G), with Pyrrhotiteis present locally in the westpart of the the earliestgarnet being characteristic of the lower- Essexorebody as individualgrains and as minute in- mostskarn in the Essexpipe and the youngestgarnet clusionsin sphalerite.The contactbetween pyrrho- characteristicof the upperand upper peripheral parts tite-bearingand pyrrhotite-absentassemblages is of the skarn. Garnet with final bandedbirefringent roughlyparallel to the mainDarwin stock contact (Fig. rims outlinesthe central skarncore and yellow bi- 12E), asnoted by Rye et al. (1974). Within the pyr- refringentgarnet characterizes the upper parts of the rhotite-bearingzone, however,there is no apparent Essexpipe awayfrom the centralcore. spatialtrend to the abundanceofpyrrhotite inclusions K feldsparis a traceconstituent in skarnabove the in sphalerite.The originof thesepyrrhotite inclusions 600 level, where it commonlyfills vugs between gar- is unclear;they may representsulfidation as repre- net grainsand is in texturalequilibrium with sulfides sentedby an equationlike andfluorite. Fine-grained, texturally late, muscovite- quartzalteration of the K feldsparin skarnis erratic (1 - x) FeS (in sphalerite)+ x/2S2'-• Fe•-xS. (1) but limited to the uppermostpart of the orebody, If this is the case,sphalerite was altered after initial abovethe 400 level (Fig. 12H). A similarpattern is precipitationby a hydrothermalfluid with a relatively seenfor chlorite-calcite-ironoxide alteration of garnet highersulfur fugacity. and for bustamitealteration of pyroxene(Fig. 12H), The distributionof fine-grainedpyrite inclusions whichboth are concentratedat the top andperiphery in sphaleriteis more regular,as there appearsto be of the Essexorebody. These patternsindicate that a core zonein the upperparts of the Essexorebody lower temperature,postskarn alteration, dominated in whichsphalerite contains a highabundance of py- by hydrolysisand sulfidation, was concentrated in the rite inclusions(Fig. 11F). Suchinclusions may rep- upperand peripheral parts of the systemand suggest resentfurther sulfidation,as representedby that conditions(T-fo•-Xco•) remained within garnet (_ K feldspar)stability in the lower centralportions FeS (in sphalerite)+ 1/2S2--• FeS•. (2) of the skarnpipe. The locusof highestabundance of iron sulfideinclu- sionsin sphalerite(Fig. 12E andF) is fundamentally Zoningof mineral compositions--Essexpipe basedon the originallocus of Fe-richsphalerite (see Due to the highly complexzoning of individual below) combinedwith the locusof flow of later sul- garnetgrains (Figs. 6 and7A), the deposit-widezon- fidizinghydrothermal fluids. ing of garnet compositionsis alsohighly complex. Unreplacedmarble beds are commonbeyond the Giventhe rangeof garnetcompositions observed in

between bleached and unbleachedmarble at a location near the pyrite-pyrrhotite boundary of 12E, presenceof K feldsparthroughout the skarnabove the 600 level, restrictionof major alterationof garnet to the upper and marginalparts of the Essexpipe. Data (someof the pyrite, pyrrhotite, and chalcopyriteabundance data) from Eastman (1980) andthis study(remaining data). 974 NEWBERRY,EINAUDI, AND EASTMAN

A maxmole % Jo+ Hdin skarnpyroxene iis iwt' %Section$b+ CC'Bi in 'alena point50•t•sSam101e•

/ D recalculatedmole % FeS in sphalerite

FIG. 13. Distributionof mineralcompositions along cross section E'-E, Essex pipe. The dataare all consistentwith zoningaround a centralskarn pipe, with some perturbations related to highergraphite contentsin marbleson the westside of the pipe.A. Maximummole percent johannsenite (Jo) + hed- enbergite(Hd) componentsin skarnclinopyroxene (electron microprobe analyses). B. Averagewt percentBi q-Sb contents in galena(microprobe analyses and atomic absorption analyses of high-quality mineralseparates). C. Wt percentMn in sphalerite.D. Estimatedmole percent FeS in sphaleriteprior to sulfidation(see text), based on (a) electronmicroprobe analyses adjusted for inclusionpyrrhotite and pyrite and (b) atomicabsorption analyses of mineralseparates. Data from Hall and MacKevett (1962),Hall (1971),Czamanske and Hall (1975),Eastman (1980), and this study [most of the sphalerite and pyroxeneelectron microprobe analyses]. a singlegrain, average compositions show no recog- roxene and chalcopyritedata, this indicatesthat the nizablepatterns. The restrictionof the secondgen- hydrothermalfluid becamerelatively enriched in Sb, erationyellow isotropicgarnet to lower portionsof Bi, Fe, andMn andrelatively depleted in Mg and Cu the Essexskarn pipe, however,(Fig. 12G) indicates with distance from its source. that andraditegarnet is restrictedin distributionand Average manganesecontent of sphalerite is also is onlyabundant in lowerparts of the Essexpipe. zoned(Fig. 13C), with an initial upwardand outward Pyroxenealso shows complex zoning patterns (e.g., increasefrom the deepcenter of the pipe anda return Newberry,1987) but a narrowerrange in compositionto low Mn contentsat the fringesof the pipe. This thanthat of garnetwithin a singlesample. Maximum patternbears some similarity to the patternof Mn in hedenbergite(Hd) + johannsenite(Jo) content in py- pyroxene(Fig. 13A) in terms of the initial outward roxeneshows a simplepattern (Fig. 13A),increasing increasein Mn; lower temperatureson the fringe of progressivelyaway from the deep core zone where the pipe (seebelow) may have restrictedMnS solu- pyroxeneis closeto diopsidein composition. bility in sphalerite,causing the low Mn characteristic Compositionalzoning of galena(Fig. 13B), based of the skarnfringe. on the data of Hall (1971) and Czamanskeand Hall The interpretationof analyticaldata on the iron (1975),indicates that galena became increasingly en- contentof sphaleriteis complicatedby the presence richedin Sb + Bi with increasingdistance from the of very fine grainedpyrite andpyrrhotite inclusions deepcore of the pipe.In conjunctionwith the py- (cf. Fig. 12E and F). Iron contentsdetermined by DARWIHPb-Zn-Ag $KARN DEPOSIT 975

mole% Johannsenlte + bulk chemical techniquesfor sphaleriteseparates In skarn (Hall, 1971) include the iron sulfideblebs, but elec- pyroxene) tron microprobeanalyses do not. If it is acceptedthat J+H 21 * the Fe sulfideinclusions in sphaleriteresulted from sulfidationalone (as argued above)mwith no addition of Fe--than the bulkFe/Zn ratiosof thesecomposites represent the original (presulfidation)sphalerite J+H 23 compositions.Therefore, the originalbulk composi- .. tionsof inclusion-bearingsphalerite grains were es- timated from electron microprobeanalyses in con- junction with estimatedmodal abundance of Fe sul- • ,'•'-'- stope fide inclusions,and these were used to indicate the 4%,':3 f contours originalhydrothermal zoning pattern. This recalcu- /mole%Jo •fHd lation results in small increases in the FeS content *,"456 4 / pyroxeneInskarn .• andis quantitatively important only in thoseareas with Sb+Bi ,' >5 vol percentiron sulfideinclusions in sphalerite la!,' / (cf. Fig. 12E andF). Compositionsfor no other min- /3.5• erals were treated in this way becausethere is no ; ._7•wt%Sb +Bi ; 375 in galena petrographicevidence that the late sulfidationevent ; f causedchanges in mineralcomposition for other min- mole ß 8-isotope erals considered here. FeS porphyrytemp. (•C) 50 METERS Zoning of recalculatedsphalerite compositions is partially asymmetricwith respectto the pipe core FIG. 14. Crosssection through the Defiancepipe (sectionD'- (Fig. 13D). There is a generalincrease in recalculated D, Fig. I I), showingprogressive outward zoning of (a) maximum Jo + Hd componentin clinopyroxeneand (b) wt percent Bi + Sb molepercent FeS upwardand toward the periphery in galena,similar to shapeof Pb/Zn >I metal ratio zone (Fig. of the pipe (Fig. 13D), althoughthe increasein FeS IIA). Also shownare averageFeS contentsof sphaleriteand S from the core toward the pluton is much lessthan isotope-basedtemperature estimates, which show little variation. thatfrom the coreaway from the pluton.Additionally, Major stope locations(taken from unpub. AnacondaCompany a zoneof loweriron sphaleriteappears in the upper- maps)are marginalto the porphyry plug and show an inverted cup morphology.Analytical data from Hall andMacKevett (1962), mostlevels of the Essexpipe. Hall (1971), Rye et al. (1974), Czamanskeand Hall (1975), East- Compositionalvariations in sphalerite,recalculated man (1980) lmostsphalerite and pyroxenemicroprobe analyses], to reflectinitial deposittonal environments (Fig. 13D), and this study. Abbreviations:Hd = hedenbergite,Jo = Johon- canbe interpretedin termsof the sulfidationstate of nsenite. hydrothermal fluids, because the iron content of sphalerite in equilibrium with an iron sulfide is a functionof temperatureand fugacity of sulfur(Scott Zoningof mineral compositionsBDefiancepipe and Barnes,1971). Thus, iron contentof sphalerite Limited data for mineral compositionsfrom the increasedas sulfidationstate decreased upward and Defiancepipe (Fig. 14) is compatiblewith the data toward the periphery of the pipe. The decreasein presentedfor the Essexpipe. Iron + manganesecon- sulfidationstate from the core toward graphitic marble tents of pyroxeneincrease upward and outward, as (awayfrom the pluton)was muchgreater than that do the Bi + Sb contentsof galena.Pyroxene compo- fromthe coretoward the pluton,reflecting the lower sitionalisopleths are concordantwith metal zoning sulfidation-oxidation environment that also stabilized and are symmetricallydistributed above the granite pyrrhotiterather than pyrite. The upwarddecrease porphyry plug. Sphaleriteshows virtually no com- in sulfidationis reversedin the uppermostlevels positional variation, however, indicating minimal (above3A level;Fig. 7A), wherelow iron sphalerite changesin the sulfidationstate despite major changes reflectsrelatively high sulfidation states. If thissphal- in the pyroxenecompositions. erite wasdeposited during main-stage Pb-Zn miner- alizationand if the recalculationof sphaleritecom- Zoningof sulfur isotoperatios positionsto reflect main-stageconditions was valid, Three galenasamples and one sphaleritefrom the thanthe high-levelsphalerite signals a spatialreversal deep central core of the Essexpipe were analyzed towardhigher sulfidation states relative to deeperand for sulfurisotope ratios for this study (Table 1) in lateraltrends. Because this area is spattallycoincident order to complement the coverage of Rye et al. with the locusof extensiveretrograde alteration of (1974), particularlyin the deep central core of the garnet and pyroxene(Fig. 13H), it alsois possible Essexpipe. Sampleswere selected,crushed, and then that the low iron contentof sphaleritereflects late handpickedunder a binocularmicroscope. The com- sulfidizingfluids that were appreciably cooler and/or bineddata set is illustrated in Figure15A, where/ia4S more oxidizedthan deeperand earlier fluids. valuesof galenaare interpretedas zoned around the 976 HEWBERRY,EINAUDI, AND EASTMAN

TABI•E1. Sulfur Isotope Data from the EssexPipe, Thermal zoning Darwin Deposit In order to verify the data of previousworkers, Sample •34S •34S additionaltemperature estimates were madefor this no. Level gl spl Aspl-gl TøC study employing sphalerite-arsenopyritecomposi- tional geothermometry(3 samples;Table 2) and Darl 600 +1.3 +2.8 1.5 420 Dar2 900 +2.1 sphalerite-galenasulfur isotope thermometry (1 sam- Dar3 600 -2.0 ple; Table 1). Arsenopyrite-sphaleritetemperatures were estimatedfrom electronmicroprobe analyses of Analysesperformed at GeochronLabs, Cambridge, Massachu- arsenopyriteand sphalerite grains from coexisting ar- setts;fractionation factors from Ohmotoand Rye (1979); samples senopyrite-sphalerite-pyrite(or pyrrhotite) assem- plotted in Figure 15 blages. One of these samplescontained pyrrhotite Abbreviations:gl = galena,sph = sphalerite blebsand these were includedin the estimateof orig- inal sphaleritecomposition. The calculationsused the intersectionsof arsenopyritecompositional isopleths skarnpipe definedby mineral,metal, and mineral (Kretschmarand Scott, 1976) with sphaleritecom- compositionalzoning (Figs. 5, 10, 12, and 13). The positionalisopleths (Scott and Barnes, 1971) on a log •34Svalues of galenaare highest in thecentral, deep j• vs.temperature diagram. part of the pipe anddecrease outward and upward, Temperaturesderived from S isotopefractionation althoughthe coreto marginzoning is greateston the (Rye et al., 1974; and this study),phase homogeni- westside of the pipe. The •34Svalues of sphalerite zation(Czamanske and Hall, 1975), andarsenopyrite fromRye and Ohomoto(1974) alsoare interpreted geothermometry(this study) all are compatiblewith aszoned around a centralcore; •34S values of galena thermalzoning around the Essexpipe. Despitethe areemployed here because Rye et al. (1974)present varietyof techniquesemployed, all temperaturees- a widerspatial distribution of galenathan of sphalerite timatesare >350øC in the coreof the skarnpipe and data. <300øC onthe marginsof the skarn(Fig. 15B).There In contrastto the patternfor the Essexpipe, sulfur appearsto be a smallvertical thermal gradient and a isotoperatios from the Defiance pipe show only minor largehorizontal thermal gradient. The factthat tem- variations(Rye et al., 1974). Valuesdecrease upward peraturesestimated from sphalerite-arsenopyrite-iron in a relativelysystematic manner, e.g., the •34Svalue sulfideequilibria, which reflect presulfidationcon- for pyritechanges gradationally from 4.4 per mil at ditions,are compatiblewith temperaturesestimated the 1200 level to 2.0 per mil near the surface.The from phasehomogenization and sulfurisotope frac- available data are insufficient to define a horizontal tionation(Fig. 15B), suggestthat the latter estimates zonation. alsoreflect presulfidationconditions.

A E'I SectionCC'd©l a• Sgalena

AThis study __ ß Rye et al. (1974)• •

FIG. 15. A. Distributionof sulfurisotope ratios in galena.B. Distributionof sulfidedepositional temperaturesfor cross section E'-E in theEssex pipe. Sulfur isotope ratios from Rye et al. (1974)(with samplelocations from Rye and Ohmoto, 1974; and Czamanske and Hall, 1975)and this study. Tem- peratureestimates based on (1) S isotoperatio fractionationbetween sphalerite and galena, unless noted(from Rye et al., 1974, and this study),(2) minimumtemperatures for rehomogenizationof exsolvedsulfosalts (Czamanske and Hall, 1975),and (3) compositionsof coexisting sphalerite and ar- senopyrite(this study). DARWINPb-Zn-Ag SKARN DEPOSIT 977

, T•',BLE2. Data for Calculationof Arsenopyrite-SphaleriteTemperatures •

Sampleno. 87 87 40 40 82 82 Arsenopyrite Sphalerite Arsenopyrite Sphalerite Arsenopyrite Sphalerite

N 2 4 4 6 3 3 3 Fe 35.25 (.04? 5.58 (.32) 35.31 (.36) 7.55 (.22) 35.47 (.15) 6.62 (.44) Zn 0.05 (.02) 60.54 (.54) 0.02 (.02) 58.13 (.35) 0.02 (.01) 59.45 (.47) Mn 0.02 (.01) 0.28 (.04) 0.01 (.01) 0.52 (.04) 0.02 (.01) 0.28 (.03) As 43.50 (.13) 0.0 (0) 42.48 (.29) 0.0 (0) 43.93 (.21) 0.0 (0) S 21.04 (.10) 33.12 (.07) 21.55 (.28) 33.41 (.41) 20.33 (.11) 33.30 (.18) Total 4 99.91 99.52 99.37 99.61 99.77 99.65 Mole%FeS 9.8 (.5) 196 (2) 11.6 (.6) Mole% As 31.03 (.03) 30.27 (.19) 31.60 (.07) Temp6 375 (5)0C 310 (10)*C 385 (5)0C

Electronmicroprobe analyses (in wt %) performedusing Cameca microprobe at WashingtonState University N = numberof spotanalyses for that sample One standarddeviation in parentheses Electronmicroprobe analyses in Eastman(1980) indicate approximately 0.5 wt percentCd istypically present in Darwinsphalerite Basedon presenceof 7 vol percentpyrrhotite inclusions Uncertaintyfrom microprobe analyses only

Comparisonof isothermsdetermined by the above Hence, the higher and lower temperaturesamples techniquesto the locationsof majorstopes in the Es- reflecttemperatures ofskarn formation and skarn de- sexzone (Fig. 16) suggestthat a dose correlationex- struction,respectively, rather than a time-indepen- isted betweentemperature and ore deposition;ore dent thermal zonation. As all the data are from the wasdeposited in the temperatureinterval from 375 ø centralpart of the pipeor the beddedskarn extension to 300øC. The pattern of isothermsalso is similarto (ef. Hall and MaeKevett,1962), the datacannot re- patternsof mineral composition,metal ratios, and solve a horizontal thermal zonation. The locus of ma- mineral distribution,suggesting a commonthermal jor ore deposition,as indicated by locationsof major control. stopes,shows a bell-shapedpattern centeredon the Isotopictemperatures from the Defianceskarn pipe porphyrybody (Fig. 14) similarto that seenfor the (Fig. 14) alsoare high (>375øC). A sampleyielding Essexpipe. Althoughdefinitive temperature data are an anomalouslylow temperature(295øC) represents not available,by analogywith the Essexpipe, the dis- a zone of major calcite-quartzsulfide alteration (as tributionof ore shootsin the Defiancepipe mayhave indicated by undergroundmapping), whereas the been controlledby thermalpatterns. othersamples are fromsulfide-bearing skarn (sample descriptionsin Hall and MaeKevett, 1962, p. 69). Discussion Metal zoning Metal zoningin the Essexpipe (Fig. 10) is domi- natedby an upwardand outwardincrease in Pb/Zn, similarto thatobserved in otherlead-zinc skarns (e.g., Einaudiet al., 1981; Shimizuand Iiyama, 1982; Yun and Einaudi, 1982; Meinert, 1987; Megaw et al., 1988), followedby a reversalon the outermargins, where relativelylow Pb/Zn ratiosare characteristic of pyrite- and/orpyrrhotite-rieh assemblages low in both Pb and Zn. Similaroutward changes in Pb/Zn ratioshave been observedby Loueksand Petersen (1988). The pattern of Ag/Pb zoningis similarto that of Pb/Zn,in that anupward and outward increase in Ag/ Pb is observedin the highlymineralized portions of the pipe,followed by a reversalto lowerratios on the fringes.Similar zonal patterns have been described FIG. 16. Distributionof stopes(from unpub. AnacondaCo. by Megaw et al. (1988) for severalskarns and re- maps)and sulfide mineral isotherms (fT, om Fig. 15B)for the E•-E crosssection of the Essexpipe, illustrating(1) symmetricalzoning placementsin northernMexico. At Darwin, the lo- of stopesaround the pipecore, and (2) localizationof moststopes cationof.highest Ag/Pb ratiosreflects the presence in the area between the 375* and 300'C isotherms. of paragenetieallyyounger, Ag-rieh galena (Hall, 978 NEWBERRY, EINAUDI, AND EASTMAN

1971; Czamanskeand Hall, 1975) which tendsto oc- and chosea mediantemperature of 325 ø ___55øC. cur outsidethe core of the pipe (Fig. 12D). Rye et al. (1974) did not attemptto placetheir tem- peratureestimates into a thermalzoning context (and Mineral zonation they downplayedtemperature decline as a causeof Perhapsthe key observationconcerning minerals ore deposition),but they recognizedthat "near the and their distribution at Darwin is that the Darwin intrusionon the eastfringe of the mineralizedzone, depositis truly a skarn,rather than a seriesof"massive the isotopictemperatures for someunknown reason replacementbodies in silicatedlimestone" (Rye et are low" (p. 474). The presentstudy explainsthis al., 1974, p. 468). Sulfidemineral as well as metal apparentanomaly; if all the temperatureestimates ratiopatterns mimic (and are partly caused by) silicate basedon sulfurisotope fractionation, phase homog- mineral distributionpatterns, e.g., distributionof enization, and sphalerite-arsenopyritecompositions garnet types and distributionof pyroxenecomposi- are compiledin a singlecross section, and if the tem- tions.Furthermore, the parageneticrelations between peratureestimates are takenat facevalue, a thermal sulfidesand silicatesindicate that duringsulfide de- zoningpattern for the main stageof ore deposition position,the hydrothermalfluid evolvedthrough a emergesthat is consistentwith the zoningof mineral range of physiochemicalconditions from those ap- assemblages,mineral compositions, and metal ratios. propriatefor garnet+ pyroxeneand then garnetsta- This overall zonal pattern is centeredon a vertical bility, to those in which both garnet and pyroxene skarnpipe locatedwest of the Darwin stock. were hydrolyzed. That S isotope-basedtemperatures are compatible with the arsenopyrite-sphaleriteand the phasehoo Mineralizingpluton mogenizationtemperatures (Fig. 15B) suggestthat Metal, mineral, and mineral compositionalzoning, the S isotopefractionations also reflect presulfidation aswell asisotopic zoning, suggest that a verticallocus equilibriumtemperatures. Lack of isotopicequilibra- of upwellinghydrothermal fluids was locatedabout tion of sphaleriteat the lower temperaturesof sulfi- 150 m from the main contact with the Darwin stock dation(•250øC?) is suggestedby the factthat much and that these fluids flowed both toward and away of the pyrrhotite present (Eastman,1980) is the from the stock.On this basis,a genetictie between monoclinicpolymorph (stable below 254øC; Kissen, the Pb-Zn-Agskarns and the Darwin pluton can be 1974), but sulfurisotope temperatures from the pyr- ruled out. Furthermore, structuralevidence (New- rhotitezone (Rye et al., 1974) are consistentlyin the berry, 1987) that indicatesthe Darwin Pb-Zn skarns vicinity of 300øC. Althoughthe sulfidationreactions postdatethe Darwin pluton by at least 20 Ma also describedrequire addition of new sulfurto the sphal- rules out a geneticlink to a deeper off-shootof the erite to precipitatethe tiny inclusionsof pyrrhotite Darwin stock.Major and trace elementcompositions or pyrite, the amountof addedsulfur, and hence po- of the Darwin stockgreatly differ from thoseof the tential change in bulk sulfur isotope ratio of the deepporphyry body at the Defiancepipe (Newberry, sphalerite (assumingthe presulfidationsulfur in 1987), whichsuggests the deepporphyry body is un- sphaleritedoes not isotopically reequilibrate), is small. related to the Darwin stock. For example,if sulfidationof a sphaleriteresults in The problemof determiningthe mineralizingplu- precipitationof 5 modalpercent pyrite (a typicalup- ton is not unique to the Darwin skarn.There is, un- per limit seenat Darwin), andthe addedsulfur is in fortunately, nothing about the spatial proximity of equilibriumwith the new inclusionpyrite at 300øC pluton to skarnwhich guaranteesa genetic relation- (at whichtemperature Apyrite_sphalerite is about 1%0 ), ship, especiallyin districtswhere multiple plutonic addition of the new sulfur results in an increase in the eventsare knownto haveoccurred. For example,the •34Svalue for the altered(composite) sphalerite grain Cirque Lake stockis coincidentallynext to (northof) of about0.05 per mil--well within analyticaluncer- the MacTungW skarn,N.W.T., but zoningof quartz tainties. vein densities,alteration, and mineralizationpoint to The temperaturedistribution pattern illustrated in a hiddenplutonic source at depthsouth of the deposit Figure 15B mayrepresent a time-integratedpattern; (Atkinsonand Baker, 1986). i.e., temperatureestimates from sulfidesin the core of the depositrepresent earlier depositedskarn-as- Temperaturesof ore deposition sociated ores, whereas the temperature estimates Rye et al. (1974) determinedisotopic temperatures from the marginof the depositare from sulfidesde- for the Essexzone that ranged from 365 ø to 255øC positedat a time whenthe entire deposithad cooled and statedthat the uncertaintiesin samplingand iso- to lower temperatures.This is a viable hypothesis; topic techniquescould not explainthe observedvari- but the gradationalchange in temperaturesmeasured ations in isotopictemperatures. These authorscon- from pipe core to marginmight argueagainst it. Al- cludedthat spatialvariations in depositionaltemper- thoughwe have not seeneither the bulk of the sam- atures for sphaleriteogalenaores probably did exist ples which were submittedfor S isotopestudies or DARWINPb-Zn-Ag SKARN DEPOSIT 979 the precise samplinglocations, our underground -28. mappingin the vicinity of previoussample locations suggeststhat the bulk of the samplestaken were from skarn-associatedores. Consequently, we suspectthat the thermal distributionof Figure 15B representsa -29 snapshotof temperaturespresent in the depositdur- ing the mainstage of sulfidedeposition. Regardlessof the precisetemperature distribution, muchof the sulfidedeposition took place at temper- aturesin excessof 375øC, probablyin the presence of a stronglateral thermal gradient. At anXco2 < 0.02 (computedby Rye et al., 1974), pressure<0.5 kbar (indicatedby aplite geobarometry;Newberry, 1987; and by 20% FeS sphaleritewith pyrite-pyrrhotite; Hall and MacKevett, 1962), and oxidationstate (see below)between hematite-magnetite and nickel-nickel -32' oxide, granditc garnet is stable to at least 350øC

(Taylor and Liou, 1978). This is consistentwith the ! petrographicdata (e.g., Fig. 7B) that suggeststhat muchof the sulfidewas deposited with or in equilib- -3: fromRyeetal. (1974) t rium with garnetskarn. A stronglateral thermal gra- oH dient (Fig. 15B) suggeststhat sulfidedeposition was not isothermal,as is alsosuggested by the observed FIG. 17. Logfo:-pH diagram(350øC) modifiedfrom Rye et al. (1974) showingsimplified vertical and lateral chemicalevo- variationsin sulfide-silicaterelations (e.g., Fig. 12H). lution of the Essexpipe ore-formingfluids, based on sphalerite Indeed, the thermal patternssuggest that the main compositions(Fig. 13D) andsulfur isotope ratios (Fig. 15A). This stageof sulfidedeposition took place over a temper- isothermaldiagram is a grosssimplification of the fluid evolution, ature rangefrom •400 ø to <300øC. as the data in Figure 15B clearly indicatea >125øC variationin fluid temperaturebetween the pipe core and margins.Sulfur iso- Conditionsof sulfidedeposition and deposit tope ratioisopleths (modified from Rye et al., 1974) are for galena formation in equilibrium with a fluid containing0.01 M sulfur and with a bulk (•34Svalue of 5 per mil. Locationsof sulfideoxide field The similarities in characteristics of both the Essex boundariesare somewhatdifferent (although the diagramtopology andDefiance pipes suggest that an integrationof data is maintained)using most recently published log K's;this diagram is retainedto permit direct comparisonwith resultspresented by from both pipesis a valid approachin establishinga Rye et al. (1974). Somecurves were calculatedusing the computer generalizedmodel for skarnsulfide pipes in the Dar- programof Ripley and Ohmoto (1979). Letters A to E represent win district.Sphalerite compositional patterns (Figs. estimatedfluid conditionsfor variousparts of the deposit, dis- 13D and 14) from this studyand S isotopevariations cussedin text. Abbreviations:Ksp = K feldspar,mt -- magnetite, from Rye et al. (1974) andthis studycan be usedto mu = muscovite,po = pyrrhotite, py = pyrite, qtz = quartz, spl = sphalerite. estimatechanges in pH, fo2, andfs2 during ore de- position(Fig. 17). Sulfidedeposition took place over a range of temperatures,from •400 ø to <300øC. However, becausevariations in mineral •34S values toids at oxidation states between nickel-nickel oxide are far moresensitive to changesin pH andoxidation andhematite-magnetite (Ohmoto and Rye, 1979). As statethan to temperaturein the 300 ø to 400øC range pyrrhotiteis not presentin the deepestworkings (but (Ohmoto,1972), an isothermalpH-log fo2 diagram low iron sphaleriteq- pyrite is present),the oxidation canbe employedas a firstapproximation to trackfluid stateof the initial fluid wasabove that of pyrite-pyr- evolution (Fig. 17). Most parameterschosen (tem- rhotite-magnetite.Interaction of this fluid with mi- perature,total sulfur,etc.) are the sameas those em- caceoushornfels (Fig. 5) would have set the initial ployedby Rye et al. (1974) to facilitatecomparison. pH at lessthan 5 (minimumrequired for muscovite- The initial fluidwas either of magmaticderivation K feldspar-quartzstability at 350øC for the fluid or wasin equilibriumwith graniticrocks at hightem- compositionas derivedfrom fluid inclusionsby Rye peratures,because of the high salinitiesof ore fluids et al., 1974). On coolingat constantcomposition from (•24 wt % NaC1equiv; Rye et al., 1974),the presence high temperatureto 410 ø to 430øC (the highestsul- of a quartzporphyry plug at depthunder the Defiance fide depositionaltemperatures determined in this pipe, and the presenceof high-temperatureskarn study)such a fluid would precipitategalena with a mineralsat depthin bothpipes. A bulkfluid •34S value &saSvalue of approximately2 per mil, sphalerite with of 5 per mil is within the magmaticrange for fluids a &saSvalue of approximately3.5 per mil (valueses- in equilibriumwith I-type or magnetiteseries grani- timated using data in Ohmoto, 1972), and with an 980 NEWBERRY, EINAUDI, AND EASTMAN iron contentof about3 molepercent (solid circle la- reaction with carbonatealone did not produce the beled "core" in Fig. 16). high pH characteristicof fluidsin the westernpyr- The first appearancesof K feldsparin the Essex rhotite-pyritezone. Here again,a coupledfo•-pHre- skarnat the 600 level, decreasein galenaand sphal- action,such as in eq (3), probablycaused pH to in- erite•a4S values to about1 and3 permil, respectively, creaseas fo• decreased.The importanceof a pyrite- and slightincrease in iron contentof sphaleriteto precipitatingreaction is suggestedby the lack of about 4 mole percent are consistentwith fluid evo- furtherincrease in pH (i.e., no magnetite-onlyzone) lution to a slightlylower oxidationstate and an ap- oncepyrrhotite becomes the predominantiron sul- preciablyhigher pH (pointB, Fig. 17). Neutralization fide.Variations in sulfurisotope ratios over short dis- and slightreduction of the sulfide-depositingfluids tancesin the pyrrhotite-magnetite-pyritezone (Fig. arelikely consequences of reaction with the pyroxene- 15A) probablywere causedby minorfluctuations in bearingskarn and slightlygraphitic marble present the fluid oxidationstate (as suggested by fluctuations belowthe 600 level (Essexpipe). A majorportion of betweenpyrite, magnetite, and pyrrhotite stability in the sulfidein thispart of the skarnoccurs as replace- this zone; Eastman,1980). ment of iron-poorpyroxene (Fig. 7E). The dataavailable for the Defiancepipe permit an Abovethe 600 level (Essexpipe) garnetskarn with interpretationbroadly similar to thatof the Essexpipe. minor pyroxeneis common,and pyroxene skarn is Most notably,although there are major variationsin rare.Fluids traveling up andto the eastfrom the skarn pyroxeneand galenacompositions in the Defiance pipe coretoward the stock(at the 600'level)followed pipe (Fig. 14), variationsin sphaleritecomposition a path of decreasingoxidation state and increasing (Fig. 14) and sulfurisotope ratios (Rye et al., 1974) pH, asindicated by sphaleritecompositions (Fig. 13D) are quite small.Compositions of sphaleritein located and decreasingS isotoperatios (Fig. 15A) to neutral samplesvary from 3 to 3.5 percentFeS at depthto 5 pH for that temperature(point C, Fig. 17). Increase percentFeS nearthe surface,compatible with an ox- in pH is a logicalconsequence of interactionwith car- idation state decrease similar to that seen in the mid- bonates,but the appreciabledecrease in oxidation deeper parts of the Essexpipe. Sulfur isotoperatios stateis not (giventhe low graphitecontent this mar- from galenaand sphaleriteshow no systematicvari- ble), unlesspH is chemicallylinked to fo•. Onepos- ations,but the values(Rye et al., 1974) of +0.9 to siblelink is precipitationof pyrite from an H2S-pre- -0.3 (galena)and 3 to 1 (sphalerite)are similar to dominant solution: thoseseen in the lower and centralparts of the Essex pipe. The absenceof major variationsin sphalerite Fe+2 + 2H2S+ 1/202 = 2H+ + H20 + FeS2 (3) compositionsand isotopic ratios in the Defiancepipe partly reflectslimited samplingoutside of the pipe for which an increasein pH yieldsa decreasein core region (Fig. 14) and partly reflectsthe lack of logfo•. graphitein hostcarbonate rocks around the Defiance Lower iron sphaleritein the uppermostpart of the pipe (asindicated by surfaceand undergroundmap- Essexpipe (eastside) together with little changein S ping). isotoperatios (Fig. 15A) requiresboth oxidationand The modelpresented here hypothesizesthat fluids pH decrease,that is, a reversalin the overall trend were derivedfrom a sourcebelow the centralpart of (path C--•D, Fig. 17). Decreasein pH is consistent the Essexorebody and not from the adjacentDarwin with sericite-quartzalteration of K feldsparand chlo- stock.Hence, the absenceof mineralizationand ap- rite alterationof garnetin the upperpart of the skarn preciable alterationin the Darwin stockis a conse- (Fig. 12H). Rye et al. (1974) indicatethat hydro- quenceof the lackof geneticconnection between the thermalfluids responsible for (late)carbonate depo- Darwin stock and the ores. The absence of Pb-Zn sitionwere not purely of magmaticderivation; per- skarnseast of the Darwin stock(Fig. 1) is related to hapsmixing of suchnonmagmatic fluids with the up- the absenceof graniteporphyry bodies in that area. welling skarn-formingfluids resultedin oxidation. Further, the distalnature of Pb-Znskarns (Einaudi et Oxidationcould cause a decreasein pH (especiallyat al., 1981) is reaffirmed,as the orebodiesextend out- temperaturesbelow 300øC) by partialconversion of ward from a deepersource, unexposed at presentin the weakacid H2S to the strongacid H2SO4. the Essexpipe, but seenat the Defiancepipe. Finally, Hydrothermalfluids in the Essexpipe flowing from the relativelylow temperatures of orefluids (•400 øC) the core to the west (path B--•E, Fig. 17) follow a canbe reconciledwith a magmaticsource, given that path similarto the east-flowingfluids, but the high (1) the mostlikely sourcelies at depthbelow the ore- graphite content of the marblesto the west of the bodies,(2) the sourcewas a high F granitewith a pipe causeda greater decreasein oxidationstate compositionnear the ternary eutectic(hence, a sol- (equilibriumwith graphitewas not attained,how- idustemperature of about550øC; Newberry, 1987), ever). The presenceof unreplacedcarbonate on both and (3) the ore fluidscooled as they roseand moved sidesof the Essexpipe (e.g., Fig. 5) indicatesthat outward. DARWIN Pb-Zn-AgSKARN DEPOSIT 981

Comparisonwith other Pb-Zn skarns 1. The Darwin Pb-Zn-Agdeposit fits in the con- Pb-Znskarns are typicallycharacterized by pyrox- tinuumbetween skarn deposits and carbonate-hosted ene with highto very highFe + Mn contents(Einaudi Pb-Zn-Ag deposits.Metal depositionin the Darwin et al., 1981), very unlike thoseseen at Darwin (Fig. depositclearly began after some,but not all, calc- 8). High Mg contentsin Pb-Zn skarnpyroxenes are silicatedeposition, as indicatedby skarnwith inter- consideredindicative of proximityto the fluid source stitial sulfides,sulfide-rich veins which contain and/ (e.g., Meinert, 1987), which is compatiblewith a or are envelopedby garnet,sulfide-quartz-carbonate granite porphyry body seen deep in the Defiance veinsassociated with destructionof garnet,and sulfide pipe. Lack of evolutionto high Fe-Mn contentsin veinsand replacements in marble.Thermal gradients, Darwinskarn pyroxene may be relatedto the apparent pressureof formation,and longevity of individualfluid instabilityofpyroxene in the aluminousenvironment flow conduits,among other variables,probably de- at Darwin or to the relatively deep level of erosion. termine the degreeof spatialproximity of skarnand Symmetric(and similar) distributions of isotherms, nonskarnores in a given district. metal ratios, and stopes(Figs. 10, 15B, and 16), and 2. Coincidenceof zoning patternsfor skarnsili- asymmetric-inconsistentpatterns of pH-fo• changes catesand for Pb-Zn-Agores at Darwin indicatesthat (Fig. 17) aroundthe skarncore at the Essexpipe sug- althoughore and ganguewere not depositedsimul- gest metal depositiontook place at least partly due taneously,they belong to the sameevolving hydro- to temperaturedecrease. The Darwin depositappears thermalsystem, and they shouldbe treatedas a single to have a much greater thermal gradient than ob- entity. servedin other Pb-Zn skarns;•50øC/km for Provi- 3. Becausesulfur isotope fractionat/on factors are dencia, Mexico (Sawkins,1964) and 23øC/kin (re- relatively insensitiveto temperaturein the 300 ø to gionalgradient) for Groundhog,New Mexico (Mei- 400øC range, metal depositionat Darwin has been nert, 1987) relative to the 100øC/100-m lateral treated asa simple,essentially isothermal process. In gradient for Darwin. Basedon data from the above detail the processwas complex,probably took place studies,Megaw et al. (1988) concludedthat fluid over a temperaturedecrease of >125øC, and prob- cooling is not a major control on mineralizationin ably involved complex (Fig. 15) but interrelated Mexican Pb-Zn skarn manto deposits,as Rye et al. changesin solutiontemperature, pH, oxidationstate, (1974) concludedabout Darwin. However, because and majorelement chemistry. it is nowhere demonstratedthat the pattern of sam- 4. This studyillustrates the difficultyin assigning pling has succeededin isolatingthe factor of time, a skarnto a mineralizingpluton. The spatialproximity i.e., that the measuredtemperatures all reflect the of plutonand skarn,although taken by manyas prima sameinstant of time,it isunlikely that any of the stud- facie evidence for cause and effect, is, in itself, not ies (includingthis one)have measured a true thermal very compellingevidence for a geneticlink. gradient.If one acceptsthe data at face value, there Acknowledgments is still a problem in comparingthe valuesinasmuch as the two thermal gradient studiescited above are The AnacondaCompany allowed accessto under- groundand surfaceworkings of the Darwin mine;ac- gradientsalong the fluid conduit,not perpendicular to the fluid conduit. Meinert (1987) showeda dike cessto drill core, maps,and assaydata; and provided living quartersat the mine site.The AnacondaCom- to marblethermal gradientof 35 ø to 50øC over 5 to pany alsoprovided somefunds for electron micro- 20 m; a higher gradient than that seen at Darwin, probe studies.M. L. Rivers, M. L. Zientek, S. Cor- althougha lower absolutedecrease in temperature. nelius,and L. E. Burnsassisted with the microprobe Two points can be made (1) it is not clear that any authorhas presented a true thermalgradient for a Pb- analyses.T. Sisson,G. H. Pessel, G. Wilson, and Zn skarn,hence the potentialeffects of solutioncool- L. D. Meinert assistedwith the field work and pro- vided adviceand inspiration.Graduate students at the ing on metal depositionin the Pb-Zn skarnenviron- Universityof Alaskaassisted with the compilationof ment are not yet established;and (2) the geometry of isothermsin Pb-Zn skarnsprobably varies consid- mineraland metal zoningdata. This studywas sup- erably. The latter effect is not only importantin de- ported by NSF grant EAR 78-19588 and by a grant from the Universityof Alaska.The manuscriptben- velopingchimney vs. mantomorphology (Megaw et efited from criticismby two EconomicGeology re- al., 1988) but developingPb-Zn skarnmorphology viewers. and zonationpatterns. September21, 1990; February 26, 1991 Conclusions REFERENCES The Darwin depositillustrates several character- Atkinson, D., and Baker, D. J., 1986, Recent developmentson isticsof skarndeposits and several problems inherent the geologicpicture of Mactung: CanadianInst. Mining Metø in skarn studies. allurgy Spec. Vol. 37, p. 234-244. 982 NEWBERRY, EINAUDI, AND EASTMAN

Chen, J. H.-Y., 1977, Uranium-leadisotopic ages from the south- Ohmoto, H., 1972, Systematicsof sulfur and carbonisotopes in ern SierraNevada batholith and adjacentareas, California: hydrothermalore deposits:ECON. GEOL., v. 67, p. 551-578. pub. Ph.D. thesis,Univ. California, SantaBarbara, 368 p. Ohmoto,H., andRye, R. O., 1979, Isotopesof sulfurand carbon, Czamanske,G. K., and Hall, W. E., 1975, The Ag-Bi-Pb-Sb-S-Se- in Barnes, H. L., ed., Geochemistryof hydrothermal ore de- Te mineralogyof the Darwin lead-silver-zincdeposit, southern posits,2nd ed.: New York, Wiley, p. 509-567. California:ECON. GEOL., v. 70, p. 1092-1110. Piwinskii,A. J., 1973, Experimentalstudies of igneousrock series, Dunne, G. C., Gulliver, R. M., and Sylvester,A. G., 1978, Me- centralSierra Nevadabatholith, California. Part II: NeuesJahrb. sozoicevolution of rocksof the White, Inyo, and SlateRanges, MineralogieMonatsh., v. 51, p. 193-215. eastern California, in Howell, D. G., and McDouga]l, K. A., Ripley,E. M., andOhmoto, H., 1979, A FORTRANprogram for eds., Mesozoicpaleogeography of the westernUnited States: plottingmineral stabilities in the Fe-Cu-S-Osystem in termsof Los Angeles,Soc. Econ. PaleontologistsMineralogists, p. 189- log (totso•/tot•2o)or logfo2 or T: ComputersGeosci., v. 5, p. 207. 289-300. Eastman,H. S., 1980, Skarn genesisand sphalerite-pyrrhotite- Rye, R. D., and Ohmoto, H., 1974, Sulfur and carbonisotopes pyriterelationships at the Darwinmine, Inyo County,California: and ore genesis:A review: ECON.GEOL., v. 69, p. 826-842. Unpub. Ph.D. thesis,Stanford Univ., 279 p. Rye, R. O., Hall, W. E., andOhmoto, H., 1974, Carbon,hydrogen, Einaudi,M. T., andBurt, D. M., 1982, Introduction--terminology, oxygen,and sulfur isotopic study of the Darwin lead-silver-zinc classification,and compositionof skarndeposits: ECON. GEOL., deposit,southern California: ECON. GEOL., v. 69, p. 468-481. v. 77, p. 745-554. Sawkins,F. J., 1964, Lead-zincore depositionin the light of fluid Einaudi, M. T., Meinert, L. D., and Newberry, R. J., 1981, Skarn inclusions,Providencia mine, Zacatecas, Mexico: ECON. GEOL., deposits:ECON. GEOL. 75th ANNIV.VOL., p. 317-391. v. 64, p. 629-643. Hall, W. E., 1971, Minor element contentsof the sulfideminerals, Scott,S. D., andBarnes, H. L., 1971, Sphaleritegeothermometry Darwin lead-silver-zincmine, Inyo County, California: Soc. and geobarometry:]•CON. GEOL., v. 66, p. 653-669. Mining GeologistsJapan Spec. Issue 2, p. 119-126. 1972, Sphalerite-wurtzite equilibria and stoichiometry: Hall, W. E., andMacKevett, E. M., 1962, Geologyand ore deposits Geochim. et Cosmochim.Acta, v. 36, p. 1275-1295. of the Darwin quadrangle,Inyo County, California:U.S. Geol. Shimizu,M., and Iiyama, J. T., 1982, Zinc-leadskarn deposits of SurveyProf. Paper 368, 87 p. the Nakatatsumine, central Japan: ECON. GEOL., v. 77, p. 1000- 1012. Kretschmar,U. L., andScott, S. D., 1976, Phaserelations involving arsenopyritein the systemFe-As-S and their application:Ca- Stevens,C. H., 1986, Evolutionof the Ordovicianthrough middle nadianMineralogist, v. 14, p. 364-386. Pennsylvaniancarbonate shelf in east-centralCalifornia: Geol. Soc.America Bull., v. 97, p. 11-25. Loucks,R. R., and Petersen,U., 1988, Polymetallicepithermal Stone, P., 1984, Stratigraphy,depositional history, and paleo- fissurevein mineralization,Topia, Durango, Mexico:Part II. geographicsignificance of Pennsylvanianand Permian rocks in Silvermineral chemistry and high resolutionpatterns of chem- the OwensValley-Death Valley region, California: Unpub. Ph.D. ical zoning in veins:ECON. GEOL., v. 83, p. 1529-1559. thesis,Stanford Univ., 399 p. Megaw, P. K. M., Ruiz, J., and Titley, S. R., 1988, High-temper- Stone, P., Dunne, G. C., Stevens, C. H., and Gulliver, R. M., 1989, ature, carbonate-hostedAg-Pb-Zn(Cu) depositsof northern Geologicmap of Paleozoicand Mesozoicrocks in partsof the Mexico: ECON. GEOL., v. 83, p. 1856-1885. Darwin and adjacent quadrangles,Inyo County, California: Meinert, L. D., 1982, Skarn,manto, and brecciapipe formation U.S. Geol. SurveyMisc. Inv. Ser. Map 1-932, 2 sheets. in sedimentaryrocks of the Cananeamining district, Sonora, Sylvester,A. G., Oertel, G., Nelson, C. A., and Christie, J. M., Mexico: ECON.GEOL., v. 77, p. 919-949. 1978, Papoosefiat pluton:A graniticblister in the Inyo Moun- -- 1987, Skarnzonation and fluid evolutionin the Groundhog tains,California: Geol. Soc. America Bull., v. 89, p. 1205-1219. mine, Central mining district, New Mexico: ECON. GEOL., v. Taylor, B. E., andLiou, J. G., 1978, The low-temperaturestability 82, p. 523-545. of andraditein C-O-H fluids:Am. Mineralogist,v. 63, p. 378- Newberry,R. J., 1987, Use of intrusiveand calc-silicate compo- 393. sitionaldata to distinguishcontrasting skarn types in the Darwin Yun, S. and Einaudi, M. T., 1982, Zinc-lead skarnsof the Yeonhwa- polymetallicskarn district, California,USA: Mineralium De- Ulchin district, SouthKorea: ECON.GEOL., v. 77, p. 1013- posita, v. 22, p. 207-215. 1032.