0361-0128/00/3075/1049-17 $6.00 1049

IntroductionTHE ROUYN-NORANDA area is well known for its Cu-Zn mas-sive sulfide deposits, in spite of the fact that the firstprospector went into the area in 1922 for gold, one year be-fore the discovery of the Horne deposit. Since then, most ofthe gold production has come from the Horne massive sul-fide deposit (350 t) and from mesothermal vein-type de-posits mainly associated with Archean tonalitic plutons (77 t;Carrier, 1994). The Silidor mine was discovered in 1985 byNoranda Exploration, as a result of statistical studies onquartz vein orientations and interpretation of aerial pho-tographs (Fig. 1). The initial diamond drill hole at Silidor hit6.4 m at 12.7 g/t Au (Picard, 1990). Production began in1990 and ended in 1997 for a total production of 15 t Au atan average grade of 5.1 g/t Au.

Mesothermal vein-type deposits in the Abitibi greenstonebelt have been thoroughly studied over the last two decades.Some of the deposits, such as the Sigma deposit near Val d’Or(Robert and Brown, 1986 a, b), are considered as classic mod-els for this style of mineralization. There are, however, somedifferences between deposits of this class (e.g., Hodgson,1990; Groves et al., 1992). Such variations could be related tonumerous factors: nature of host rock, conditions of deposi-tion related to different timing (Couture et al., 1994; Robert,1994), or variations in pressure-temperature conditions (con-tinuum model, Groves et al., 1992). Within plutonic rocks,several mineralization models have been proposed, resultingin differences in mineral assemblage, alteration, and tectonicregime (Jébrak, 1991).

Accurate age dates for these deposits have been providedby Wong et al. (1991), Carignan and Gariépy (1993), Corfu(1993), and Kerrich and Cassidy (1994); however, recent dis-cussions have shown that even detailed and high-quality mea-surements can be difficult to interpret, because of mineral

The Silidor Deposit, Rouyn-Noranda District, Abitibi Belt: Geology, Structural Evolution, and Paleostress Modeling of an Au Quartz Vein-Type Deposit

in an Archean Trondhjemite

ALAIN CARRIER,†,* MICHEL JÉBRAK,Université du Québec à Montréal, Département des Sciences de la Terre, C.P. 8888, succ. Centre-ville, Montréal, Québec, Canada H3C 3P8

JACQUES ANGELIER,Laboratoire de Tectonique Quantitative, Université Pierre et Marie Curie, 4 place Jussieu, 75252 Paris 05, France

AND PETER HOLYLAND

Terra Sancta Inc., 48 Peoples Avenue, Gooseberry Hill, Western Australia, Australia 6076

AbstractLate orogenic gold-bearing quartz vein deposits within the Rouyn-Noranda mining district occur mainly

within Archean tonalitic plutons of the Blake River Group (southern Abitibi belt). The Silidor deposit is a rep-resentative example of a pluton-hosted lode gold deposit. The Silidor mine contained 2.95 million tons (Mt;mined and estimated reserves), grading 5.1 g/t Au (~15 t Au). The mineralized zone is 900 m in length and hasa vertical extent of 900 m, an average thickness of 3.5 m, and trends northwest-southeast with a dip of 50° to70° NE. Its alteration envelope consists of a red hematite-altered trondhjemite. Measured δ18O values forquartz veins (7.7–10.9‰) and δ34S values for pyrite (–6.1 to –9.4‰) imply oxidizing conditions during gold de-position. The mineralized zone comprises: vein quartz (white, gray, and smoky varieties), beige mineralizedtrondhjemite, and green carbonate-sericite-fuchsite breccia, which resulted from shearing and metasomatismof an early northwest-southeast dioritic dike.

A quantitative microtectonic study (>400 measurements) was carried out on the Powell tonalitic sill in theSilidor mine area to reconstruct the deformation before, during, and after the mineralizing events. The re-duced stress tensors display large variations in σ1 orientation, trending successively northwest-southeast, north-east-southwest, and finally north-south. The Silidor deposit is the result of several vein-filling events, whichoccurred during evolution from strike-slip faulting to reverse faulting regimes, with σ1 remaining northeast-southwest but with exchange of orientation of σ2 and σ3. Such variations may reflect an oblique collision andprocesses of stress permutation.

Paleostress mapping of Archean terranes can be used as a targeting tool for mesothermal lode gold deposits.Reduced stress tensors were used for the computation of paleostress maps, using a distinct element model. TheSilidor Au quartz mineralization appears on these paleostress maps in an area characterized by low mean stressthroughout the deformation history of the area near the Horne Creek fault. This study emphasizes the role ofsecond-order faulting in the location of low- and high-pressure zones in the Archean crust and the possible roleof a tectonic indentor in the location of Au mineralization.

Economic GeologyVol. 95, 2000, pp. 1049–1065

† Corresponding author: email, acarrier@lino.com* Present address: McWatters Mining, Exploration, 1281, 7e rue, Val-d’Or,

Québec, Canada J9P 3S1.

inheritance and thermal resetting (Kerrich and Cassidy,1994). Structural reconstruction allows us to correlate theemplacement of an ore deposit with a regional deformationevent and, therefore, has the potential to tightly constrain therelative age of mineral deposition (Paradis and Faure, 1994;Jébrak and Hernandez, 1995; Faure et al., 1996).

In the Abitibi greenstone belt, the end of the Archean ismarked by the Kenoran orogeny, for which previous studieshave distinguished two main phases of deformation (Hubertet al., 1984; Robert, 1990; Corfu 1993). The first phase oc-curred after deposition of the volcanic pile and before the de-position of Timiskaming-type sediments (ca. 2686 Ma). Theabsence of associated metamorphism suggests that this phasemay be related to thin-skinned tectonic processes. The sec-ond phase occurred after the deposition of Timiskaming-typesediments and was marked by intense folding and faulting, es-pecially along regional corridors of deformation. Good exam-ples of such large deformation zones are the Cadillac-LarderLake and the Destor-Porcupine breaks. Deformation duringthis phase probably evolved rapidly from ductile to brittle,with rotation of the paleostress field (Robert, 1990; Jébrakand Faure, 1992). There is evidence of thermal resetting afterthis tectonic event (Kerrich and Cassidy, 1994), but few dataare yet available on the structural context of this resetting.

The objectives of this study are twofold: (1) to use detailedstructural analysis to correlate the evolution of a mesothermalgold deposit with the regional structural evolution, and (2) toshow how structural analysis may help in modeling of the lo-cation of such deposits. Ductile deformation leads to verycomplex patterns in a polyphase environment; however, brit-tle deformation can be deciphered with less ambiguity, usingmineralogical and structural markers. Detailed tectonicanalysis of brittle deformation allows precise determination ofboth the chronology of events and the paleostress regimes

(Angelier, 1994). It is, therefore, possible to distinguish thedifferent episodes of deformation during gold mineralizationand to reconstruct the stress field during mineral deposition.Furthermore, using the determined different stress tensors, itis also possible to apply a numerical modeling technique forthe prediction of fluid movement and the possible location ofthe zones of enhanced fluid flow. Paleostress studies thushave the potential to provide information on the possible lo-cation of a mineral deposit at the property scale.

This paper is organized as follows: after a description of theregional geology and main ore deposits of the Powell pluton,a comprehensive description of the Silidor deposit is given.Structural measurements are used to reconstruct the evolu-tion of the paleostress and hence to explain the geometry ofthe deposit. Paleostress mapping allows us to reconstruct thepossible paleohydrology of Au-mineralized fluids at the endof the Archean, following the approach of Holyland (1990),Oliver et al. (1990), Ridley (1993), Vearncombe and Holyland(1995), and Holyland and Ojala (1997). We emphasize therole of the regional stress pattern in controlling the movementof hydrothermal fluids from high- to low-lithostatic pressurezones. The location of the mesothermal gold mineralizationappears to be related to indentations along major faults.

Regional GeologyThe Silidor deposit lies within the Powell pluton of the

Blake River Group. The Powell intrusion forms a subhori-zontal, 1-km-thick sill, shallowly dipping to the east (Keating,1992). It is considered to have originally been part of theFlavrian pluton, dated at 2700 ± 2 Ma, from which it was sub-sequently separated by faulting (Mortensen, 1993). A chrono-logical succession of intrusive phases of the Powell pluton hasbeen proposed by Goldie (1976): (1) the Héré unit granodi-orite, (2) tonalitic hybrid rocks, (3) trondhjemite, (4) tonalitic

1050 CARRIER ET AL.

0361-0128/98/000/000-00 $6.00 1050

Tonalite, trondhjemiteGranodioriteRhyoliteBasalt, andesite VHMS deposits

Gold deposits

Sedimentary rocksGabbro, dioriteSyenite

Blake River Group (Abitibi Subprovince)

Horne VHMS

PontiacSubprovince

Cobalt

Cadillac

Kewagama

Kinojevis Group

MalarticGroup

Blake RiverGroup

Doyon-BousquetGold District

Central NorandaCamp Mine Series

0 7.5 15km

DPFZ

CLLFZ

Caste

Timiskaming

48 30'o

Figure 2SILIDOR

Duparquet

79 15'o 79 00'o 78 45'o

48 15'o

48 30'o

78 45'o79 00'o79 15'o

48 15'o

a

b

FIG. 1. a. Location in the Blake River Group, south-central Abitibi greenstone belt; legend: CLLFZ = Cadillac-LarderLake fault zone, DPFZ = Destor-Porcupine fault zone. b. Aerial photograph displaying the Silidor lineament before the dis-covery.

breccia, and (5) diorite dikes (northwest-southeast and east-northeast–west-southwest) and rhyolitic quartz-feldspar por-phyry dikes (east-northeast–west-southwest). During miningat Silidor, host rocks were named tonalite, but Goldie (1976)and Gaulin (1992) considered these rocks as true igneoustrondhjemites based on geochemical features and feldsparcomposition. In this paper, the term “trondhjemite” will beused to described the Silidor host rocks even if this rock mayalso be an albitized granodiorite. The intrusion phases of thePowell pluton are believed to be related to the upper BlakeRiver Group metavolcanic rocks (Paradis et al., 1988), whichhave been interpreted as a volcanic cauldron (Kerr and Gib-son, 1993).

The Powell pluton is bounded by east-northeast–west-southwest faults: the Beauchastel fault to the north and theHorne Creek fault to the south. There are three sets of majorfaults in the Silidor area, oriented east-northeast–west-south-west, northwest-southeast, and northeast-southwest (Fig. 2).The major east-northeast–west-southwest-trending faults aresubvertical or north dipping and are second-order faults sub-parallel to the Cadillac break (Robert, 1989). They probablybranch from the break at depth (Jackson et al., 1995; Ver-paelst et al., 1995). These fault zones were active during theentire structural evolution and display evidence of both duc-tile and brittle behavior.

The structural history of the area includes: an early synvol-canic extensional event (around 2700 Ma) and two main peri-ods of deformation, D1 and D2 (Dimroth et al., 1983; Hubertet al., 1984). Synvolcanic extension is deduced from the

paleogeography of the volcanic series, which displays normalmovement and facies changes along east-northeast–west-southwest faults (de Rosen-Spence, 1976; Kerr and Gibson,1993). The northwest-southeast faults also show evidence of aprotracted tectonic evolution; they control the intrusion offeeder dikes and synvolcanic hydrothermal fluid flow (Setter-field et al., 1995).

The early D1 episode of tectonism produced numerousmajor northwest-southeast-trending folds. The Silidor area islocated on the gently dipping upright limb of one of thesefolds. Associated D1 shear zones typically consist of bedding-parallel zones of schistosity. Powell et al. (1995a, b) suggestedthat the development of D1 regional folds predated regionalmetamorphism.

The main cleavage-forming event was D2, which resulted inthe formation of east-west-trending folds and an associatedcleavage (S2). This deformation has been recorded in theRouyn-Noranda and Val d’Or areas (Hubert et al., 1984;Daigneault and Archambault, 1990; Robert, 1990), where it isassociated with the emplacement of syntectonic granitic plu-tons (Feng et al., 1993). The Powell pluton was weakly af-fected by the S2 subvertical foliation along a large east-westanticline (Goldie, 1976). Early movement on the Héré Creekfault, ductile fabric, and hematitic alteration could be associ-ated with the end of the D2 event (Pelletier, 1992). Locally inthe Silidor area, the S2 foliation is folded as a result of dextralstrike-slip movement, compatible with a northeast-south-west-trending schistosity. It is still unclear however whetherthe two schistosities represent a single event (D2) or two

SILIDOR DEPOSIT, ROUYN-NORANDA DISTRICT, ABITIBI BELT 1051

0361-0128/98/000/000-00 $6.00 1051

6 DonRouyn mine

1 Silidor mine

2 New Marlon mine 4 Anglo-Rouyn mine

3 Powell-Rouyn mine 7 Joliet mine5 Senator-Rouyn mine

10 Bagamac mine

9 Chadbourne mine

8 Horne mine

Gold bearing brecciaMassive sulphide depositPorphyry copper deposit

Basalt

RhyoliteGold-quartz veins Surface Mine levelMajors faults

Minor faultsDolerite dikes

Trondhjemite

Tonalitic breccia

Quartz diorite

45o

85o

Trend and dip Schistosity

Horne creek fault

Beauchastel fault

Trémoylake

Noranda lake

Hérélake

1km

Flavrianpluton

Héréunit

Héréunit

Felsic dikes swarm

70o

60o

50 o

65 o

55 o

65 o 80o

82o

76o

85o

16

2

3

4

7

8

9

10

5

83o

79o

78o

81o

60o

Hér

é fa

ult

Jolie

t fau

lt

Smokey creek fault

Rouyn-Noranda

Powell pluton

FIG. 2. Regional geologic context, showing the different lithologies and the major ore deposits (volcanogenic massive sul-fide deposits, lode gold deposits, east-northeast, northwest, and northeast faults, and occurrences of porphyry-type Mo-Cu-Au mineralization).

distinct events (D2 and a possible D3). The S2 foliation is as-sociated with regional metamorphism and occurred after thedeposition of Timiskaming-type sediments (Powell et al.,1995b) and, therefore, after 2673 ± 3 Ma (Davis, 1991).

Late events (post-D2) also affected the area; the Silidor veinand shear cut the S2 foliation. The Héré Creek fault is markedby late brittle movement (gouge) which cuts and displaces theSilidor mineralized structure. Late intrusions within faults arealso known, an example of this would be the strong positivemagnetic anomaly caused by a Proterozoic dolerite dike inthe Smokey Creek fault (Figs. 1 and 3). Late movements onsome of these faults is also demonstrated by offsets of theDonalda gold-bearing quartz veins and Proterozoic doleritedikes. Both the Beauchastel and the Horne Creek faults showa late sinistral movement, which is considered Proterozoic inage (Kerr and Mason, 1990; Riverin et al., 1990).

MineralizationThe Powell pluton contains two types of deposits (Fig. 3): a

gold-copper-molybdenum porphyry deposit in the DonRouyn area (Kotila, 1975; Goldie et al., 1979; Jébrak et al.,1995; Carrier and Jébrak, 1996) and shear zone-relatedmesothermal gold-bearing quartz vein deposits such as Sili-dor (McMurchy, 1948; Wilson, 1962; Carrier, 1994). Thegold-bearing quartz vein deposits are related to brittle-ductilefaults and shear zones which are spatially associated withnorthwest-southeast dioritic dikes in the Powell pluton.These mesothermal gold deposits follow structures with anorthwest-southeast to approximately north-south orienta-tion, dipping to the northeast and east (Table 1; Fig. 3). Min-eral phases identified in these mesothermal deposits includequartz, fuchsite-sericite, and carbonate (dolomite, calcite),

with pyrite, hematite, gold with some silver, tellurobismuthi-nite, galena, chalcopyrite, and AuTe.

At Silidor, the auriferous ore zone included (Fig.4): (1) acontinuous brittle-ductile quartz lode, which is a shear veinwith multiple infilling events (e.g., Silidor vein); (2) less con-tinuous pyritized and albitized wall rocks (e.g., mineralizedtrondhjemite); and 3) a discontinuous sheared and metasom-atized dioritic dike (e.g., carbonate-sericite-fuchsite breccia).

Geometry and deformation

The extent of the Silidor gold-bearing ore zone (e.g., Silidorvein + mineralized trondhjemite + carbonate-sericite-fuch-site breccia) is similar to the extent of the Silidor brittle-duc-tile vein. The Silidor vein extends along a strike length of 900m and over a depth of 900 m and has an average thickness of3.5 m. The Silidor vein is oriented N 340° with an average dipof 60° toward the northeast. It has numerous variations instrike, dip, and thickness. At the northern end of the deposit,the vein splits into a horsetail-like structure, with a shallowdip. The dip of the Silidor auriferous ore zone changes grad-ually from 54° in the northern part of the deposit to 68° in thesouth. The Silidor vein itself shows larger variations in dip, inthe range of 45° E to 70° E. Pinch and swell structures occurboth on vertical and horizontal axes, but the general elonga-tion of the lenses defines a pitch of 60° toward the northeast.The vein shows a sharp change in strike, with usually barreneast-west-oriented segments. A less extensive vein, call theBonus vein, is parallel in strike and has a shallower dip; it joinsthe main vein at the 220-m level. In the southern part of thedeposit, the Silidor vein is sinistrally offset along the HéréCreek fault by 50 m. This zone is marked by increased thick-nesses of ore and alteration zones.

1052 CARRIER ET AL.

0361-0128/98/000/000-00 $6.00 1052

200 m

100 m

Héré

Silidor

200 m

SILIDOR

220 m

273 m

460 m

ANGLO-ROUYN

NEWMARLON

POWELL-ROUYN

SENATOR-ROUYN SILIDOR

200 m

200 m

200 m

N

2775'2625'2325'1875'

1625'

1375'

1250'1000'750'500'250'

460 m

100 m

273 m220 m

Level5

Level5

Level2Level1

No.1Shaft

No.2Shaft

Level8

Surface

Hér

éfau

lt

Hér

éfau

lt

Surface

300' 650'1550' 1850'

2000'2300'

2450'

FIG. 3. a. Plan view projection of Silidor lode gold deposit and kink-band interpretation under northeast-southwest com-pressional effect. b. Plan view projection of drifts of all lode gold deposits of the Powell pluton at the same scale.

a

b

The Silidor mineralized zone has a complex structural his-tory. Although most of the deformation is brittle in style, an

early ductile deformation event is recorded in the northwest-southeast dioritic dike. This northwest-southeast dioritic dikeis the locus of the Silidor structure. The dike is locally alteredto a carbonate-sericite-fuchsite ore assemblage (describedbelow with the alteration). The dioritic dike is commonlytransformed to a zone of mylonite, with S-C structuresmarked by sericite, which indicates a reverse-dextral move-ment. This deformation clearly crosscuts the S2 regional east-west schistosity and is crosscut by quartz and carbonate vein-lets. These veinlets are themselves folded and boudinaged.The strong deformation of the dioritic dike, in contrast to thebrittle deformation exhibited by the trondhjemite, and thesimilarity of the deformation history, as discussed later,strongly suggests that high-strain partitioning took place dur-ing the early stages of vein infilling. In the presence of suchlayer anisotropy (e.g., NW-SE dioritic dike in the Powelltrondhjemite), the orientations and kinematics of shear zonesand veins deviate from simple patterns predicted for homo-geneous rock bodies (Belkabir et al., 1993) and their under-standing requires a more sophisticated tectonic analysis, suchas stress tensor determination for brittle deformation.

Infilling

The Silidor vein is composed of three different facies:highly mineralized smoky quartz, gray quartz, and poorlymineralized white quartz (Figs. 4 and 5). In all facies, gold oc-curs principally in association with pyrite both coating grainsor as inclusions. The gold grade shows a correlation withpyrite concentration and this relationship was used duringmining as an estimate of gold grade, with 1 percent pyriterepresenting about 1 g/t Au (G. Laperle, oral commun.). Theaverage silver content of the gold is between 7 and 8 percentAg and locally up to 16 percent Ag (Carson, 1986). Pyrite,chalcopyrite, rutile, and specular hematite are associated withgold.

SILIDOR DEPOSIT, ROUYN-NORANDA DISTRICT, ABITIBI BELT 1053

0361-0128/98/000/000-00 $6.00 1053

TABLE 1. Comparisons between the Lode Gold Deposits Hosted in the Powell Pluton (after McMurchy, 1948; Wilson, 1962; Cousineau, 1981; and Carrier 1994)

Mine (gold Discovery Tonnage (t) and Trend Length, width, Lode Wall-rock quartz vein) and production grade (g Au/t) and dip and depth (m) mineralogy1 alteration2

Powell-Rouyn 1922 2,715, 000 600 Au-Ag, Py, Carb, (1937-1956) @ 4.4 330°/65° E 3 Cp, Qz, Cb Sil,

1,000 Hem

Anglo-Rouyn 1923 132,150 300 Au-Ag, Py, Cp, Carb, (1948-1951) @ 8.6 345°/60° E 2 Qz, Cb, Tm, Gr Sil,

200 Hem

New Marlon 1934 981,140 200 Au-Ag, Py, Cp, Carb, (1947-1949) @ 6.1 345°/70° E 2 Qz, Cb, Fu Sil,

300 Hem

Senator-Rouyn 1936 1,632,960 200 Au-Ag, Py, Cp, Qz, Carb, (1940-1955) @ 4.8 325°/50° E 23 Cb, Fu, Tm, Gr Sil,

850 Hem

Silidor 1985 2,946,180 900 Au-Ag, Py, Cp, Carb, (1990-1997) @ 5.1 330°/55° E 3.5 Qz, Cb, Fu, Gr Sil,

900 Hem

1 Ag = silver, Au = gold, Cb = carbonate, Cp = chalcopyrite, Fu = fuchsite; Gr = graphite, Py = pyrite, Qz = quartz, Tm = tourmaline2 Cb = carbonte, Hem = hematite, Sil = silica

c

10 m

S-W N-E

Cross-section

b d

Hematite-alteredtrondhjemite

Mineralizedtrondhjemite

Carbonate-sericite-fuchsite breccia

Gray quartz Smoky quartzWhite quartz

Gouge

South-West North-East

1 m Ore zonea

FIG. 4. a. Sketch from photograph illustrating all the different compo-nents of the ore zone from the Silidor lode gold deposit. The ore zone in-cludes metasomatized wall rocks (mineralized trondhjemite and carbonate-fuchsite breccia), vein infilling such as smoky quartz, gray quartz, and latewhite quartz breccia. The thickness and grade of the ore zone are highly vari-able and some of its components can be present or absent and localized ei-ther at the footwall or at the hanging wall. The (b), (c), and (d) sketches il-lustrate an opening mechanism for the quartz lode. b. Early development ofthe gray and smoky quartz veins and metasomatism of the wall rocks along areverse fault. c. and d. Development of the late white quartz cement brecciain local transtension zone, which includes fragments of earlier mineralizedveins and wall rocks.

The gray quartz and the smoky quartz form thin veins (≤40cm), with high gold grade (20 g/t Au). They are usually lo-cated along the footwall of the vein. Strike and dip of theseveins varies from east-west to northwest-southeast and from30° to 40° to the northeast. The smoky quartz (Fig. 5a) hasbeen intensely recrystallized, with large crystals and bands ofpolygonal quartz neograins containing disseminated sulfides(chalcopyrite, pyrite with octahedral faces). Gold and chal-copyrite fill fractures in pyrite. The pyrite in the smoky quartzvein has commonly been broken, with the crystals showing alarge number of mainly octahedral faces. This smoky quartzlocally crosscuts a minor phase of dark quartz which is rich indisseminated tabular hematite.

The white quartz forms thick zones, up to 5 m in thickness.It occurs as lenses along the fault with a horizontal elongationand a high aspect ratio. The strike and dip of the white quartzveins are more consistent than for the smoky quartz veins,varying from north-northwest–south-southeast to west-north-west–east-southeast. They have a brecciated texture, locallywith fragments of deformed host rock, fresh or altered, andfragments of smoky quartz (Fig. 5c and d). The angular na-ture of the fragments and the structure of the breccia indicatemainly hydraulic brecciation, with a jigsaw pattern, evolvingtoward collapse brecciation, with shingle-shaped fragmentslocated at the bottom of rhombic cavities (Jébrak, 1997). Thisquartz is low grade (0.3 g/t Au) but may reach 4 g/t Au be-cause of the abundance of mineralized fragments. The quartz

consists of large recrystallized grains, with both undulatoryextinction and small sericite crystals being located on the bor-ders of neograins. Sulfides consist mainly of chalcopyrite,cubes of pyrite, and minor molybdenite. The white quartz islocally cut by late transverse veinlets, composed of quartz,carbonates, pyrite, chalcopyrite, gold, electrum, hessite,freibergite, and petzite. Gold, electrum, and hessite are en-closed in pyrite. The similar sulfide association of these vein-lets and the smoky quartz veins suggests a protracted miner-alizing event rather than a late remobilization in differentpressure-temperature conditions. The presence of such lateveinlets reflects the continuity and the dynamics of the min-eralized tectonic system.

The Silidor vein is associated with rare centimeter-scalehorizontal tension cracks which have been filled with quartz-hematite, quartz-calcite-chalcopyrite, quartz-chlorite, andlate pink calcite. Although these structures are poorly devel-oped, they indicate that σ3 was at one stage subvertical.

Postore mineral infilling includes: thin barite veinlets(Gaulin, 1992), small gypsum steep north-south veinlets, andclay-filled fault. Gypsum veinlets have been associated eitherwith recent circulation of brines in the Superior province orwith the last stage of hydrothermal activity in gold deposits(Frape and Fritz, 1987; Couture and Pilote, 1993).

A clay-filled fault which follows the whole Silidor structurewas used as a key marker during mining. The same type of fill-ing is known to follow the Héré Creek fault. The mineralogical

1054 CARRIER ET AL.

0361-0128/98/000/000-00 $6.00 1054

FIG. 5. Field and slab photographs from the Silidor lode gold deposit. a. Smoky quartz with abundant pyrite and chal-copyrite. b. Carbonate-sericite-fuchsite breccia enclosed within white quartz and localized in the center of the Silidor orezone. c. Example of hydraulic brecciation in the white quartz cement breccia; note the jigsaw pattern of the fragment in thecenter of the photograph. d. White quartz cement breccia in contact with the hematite-altered trondhjemite of the footwall.

assemblage is quartz, calcite, chlorite, smectite, kaolinite, andlow-crystallinity illite, indicating a relatively low temperatureof the fluid, probably below 200°C (Velde and Vasseur, 1992).

Alteration

Four mappable alteration facies were distinguished system-atically during mining. From the distal to the proximal zone(Fig. 4) they consist of: chloritized trondhjemite, hematizedtrondhjemite, mineralized trondhjemite, and carbonate-sericite-fuchsite breccia (Carson, 1986, Picard, 1990, Gaulin, 1992,Carrier, 1994).

The chloritized trondhjemite is the least altered unit. It has,however, been affected by regional chloritization. It consistsof quartz (35%), oligoclase (36%), muscovite-sericite (10%),chlorite (9%), carbonates (7%), hematite (<2%), rutile andleucoxene (1%), and sulfide (<1%). This unit is locally cross-cut by quartz-sericite-chlorite veinlets.

The hematized trondhjemite forms a unit with an averagethickness of 30 m. This thickness increases to 80 m in the areawhere the Héré fault crosscuts the Silidor vein and attains140 m in the northern horsetail structure. The thickness ofhematization increases at depth and can be followed down to600 m below the surface. This alteration affected not only thetrondhjemite but also quartz porphyry and diorite. Rocks be-come red in color due to the presence of finely dispersedhematite. Albite forms up to 75 percent of the rock. Rem-nants of green chlorite are abundant and small crystals of ap-atite, carbonates, and rutile bordered by ilmenite are also pre-sent. They are relatively rich in gold (400–575 ppb), locallywith some nickel (53 ppm) and molybdenum. This hematiza-tion could be the product of oxidation accompanied by sub-stantial alkali metasomatism.

The mineralized trondhjemite is beige in color. It formsunits of limited extent, 10 m long and up to 2 to 3 m thick oneach side of the quartz vein. There is a weak correlation be-tween the thickness of the quartz vein and the thickness ofthe mineralized trondhjemite, caused by the polyphase na-ture of the quartz infilling (Fig. 4). Hematite is not present inthis zone. Albite and Mg carbonate are abundant, with acicu-lar crystals of magnetite. Pyrite is locally abundant and con-tains ovoid inclusions of chalcopyrite and sphalerite. Gold val-ues are usually economic, with an average grade of 4.2 g/t Au,with some tungsten and copper enrichment. No arsenic orantimony has been encountered. Using the isocon approachof Grant (1986) relative to unaltered rocks, a strong leachingis inferred for LREE, K2O, Ba, Th, Cr, and U, and slight en-richments in Au, W, CO2-H2O, CaO, and Na2O. Other ele-ments do not vary significantly (Gaulin, 1992; Carrier, 1994).

The carbonate-sericite-fuchsite breccia alteration facies isin the Silidor structure. This alteration facies is the result ofthe metasomatism of the northwest-southeast dioritic dike.The carbonate-sericite-fuchsite breccia is an auriferous oreassemblage which has a characteristic apple green color. Theoriginal texture of the dioritic dike has been sheared andoverprinted. The dike has been transformed into an assem-blage of chlorite, Cr2O3-rich sericite, quartz, and fuchsite(Carson, 1986). Sulfides are abundant in this alteration faciesand include chalcopyrite, pyrite, proustite-pyrargirite, gold,electrum, hessite, and freibergite, with disseminatedhematite and rutile.

The formation of the Silidor vein: A summary

Reconstruction of the deposition history of the Silidor veinhas been accomplished through detailed study of the miner-alization. Alteration at Silidor overprints early ductile fabricssuch as S2 foliation. Following the metamorphic event and theassociated deformation, three main episodes have been dis-tinguished, based exclusively on crosscutting relationships:(1) hematization, (2) mineralization, and (3) white quartzdeposition.

Hematization was the first metasomatic phase; it is a distaland pervasive alteration which is known to follow the Silidorand Héré structures. Hematization is also widespread in thePowell pluton and could be related to the emplacement of theDon Rouyn porphyritic system. The Silidor vein is located inthe outer aureole of the Don Rouyn porphyry system, as iden-tified by Goldie et al. (1979). The abundance of hematite sug-gests a high oxygen fugacity during this episode of alteration.

The mineralization of the trondhjemite is temporally asso-ciated with smoky quartz deposition. The association of al-bite-carbonate metasomatic zones with quartz deposition iswell known in mesothermal gold deposits, indicating thatsodium was the dominant cation in solution (Robert andBrown, 1984; Mikucki and Ridley, 1993). Such associationshave been observed frequently in lode gold deposits laiddown at low temperature (225°–400°C) and low pressures(<1 kbar; McCuaig and Kerrich, 1994). Although such alter-ation may occur in porphyry systems (Morasse et al., 1995), inthe Silidor vein-type deposit it appears clearly associated withshearing and deformation.

Few data are available on the composition of the fluid re-sponsible for the Silidor mineralization. Six measurements onthe Silidor quartz gave a δ18O value of 10.1 ± 1.2 per mil, sim-ilar to that measured in adjacent quartz veins at Elder andPierre-Beauchemin (9.2–10.3‰; Kennedy and Kerrich,1982). In addition, δ34S has been measured on four pyritesfrom Silidor with values between –6.1 and –9.4 per mil, sim-ilar to that measured in lode gold deposits such as Francoeur(–9.4 to –11.4‰; Couture and Pilote, 1993), Hemlo, Ontario(0 to –17.5‰), Golden Mile, Western Australia (–3.1 to –9.6‰;McCuaig and Kerrich, 1994). Lange et al. (1995) have shownthat such negative values could be the result of fO2 values closeto the ΣSO4-H2S buffer (Ohmoto and Rye, 1979) and couldbe related to several processes, including extensive interac-tion of hydrothermal fluids with Fe2+-bearing silicates in wallrocks or intrinsically oxidizing magmatic fluids (Lambert etal., 1984; Cameron and Hattori, 1987). The nature of themineralized trondhjemite alteration, preliminary observationson fluid inclusions, and comparisons to similar deposits in thePowell gold district suggest that fluids were of CO2-H2O com-position. The association of gold with sulfide suggests thatgold was transported as a reduced sulfur complex and thatsulfidation of wall rocks could have been a very efficient gold-depositing mechanism.

Late white quartz belongs to the same metallogenic eventas smoky quartz; however, observations indicate that modifi-cation of the hydrothermal system to a lower sulfur fugacityhad occurred, either due to lower capacity of metal trans-portation of the fluid or to less interaction with wall rocks (=iron), because they were already altered.

SILIDOR DEPOSIT, ROUYN-NORANDA DISTRICT, ABITIBI BELT 1055

0361-0128/98/000/000-00 $6.00 1055

Paleostress EvolutionThe analysis of brittle faults and tension gashes reveals that

the Powell pluton underwent several phases of deformation.Most of the faults are brittle, so it is possible to reconstructthe different paleostress regimes using quantitative methodsbased on shear-stress relationships.

Methodology

Orientation data were collected for dilatant veins, dikes,and striated faults. The fault slip data were collected from theunderground workings at Silidor. They were compared todata independently collected from the Don Rouyn quarry, lo-cated 1 km east of Silidor. Data from more than 350 striatedfault planes and 50 veins or dikes were used to reconstructthe paleostress history of the Silidor deposit area. Faults andveins were easily identifiable because of the presence ofslickenside lineations and mineral infill (especially quartz,chlorite, hematite, and carbonates).

Our analysis is based on the concepts of mechanical rela-tionships between brittle features (especially striated faults)and paleostress first developed by Carey and Brunier (1974),using the stress-shear relationship described by Wallace(1951) and Bott (1959). Over the last few decades, such analy-ses have proved to be successful for identification and char-acterization of tectonic events. It is, therefore, extensivelyused in order to interpret fault slip patterns. In this study, theinverse method is used to determine the local stress tensorsfrom directions and senses of slip on numerous faults of vari-ous orientations. This approach is based on the principles andmethods described by Angelier (1975, 1984, 1989, and 1991).

In practice, one searches for the best fit between all slip datacollected in a rock mass which experienced faulting during agiven tectonic event and a common unknown stress tensor.Consideration of tension data, such as for tension gashes, veinsor dikes, provides useful additional information. The assump-tion that all faults which moved during the same tectonic eventin a rock mass were moving independently but consistentlyunder a single homogeneous stress regime expressed by a sin-gle stress tensor is an obvious approximation, but theoreticaland practical evaluations show that results are consistent if faultslip data sets are large and include a variety of orientations(Dupin et al., 1993; Angelier, 1994, and references therein).

A major difficulty, however, arises in the Silidor area be-cause several brittle tectonic events have obviously affectedthe rock mass. This situation may also occur during a singledeformational event, where changes in the stress regime oc-curred, resulting in rotation or permutation of principal axesduring brittle deformation. Block rotations during a tectonicevent produce similar effects. In all these cases, brittle tec-tonic features observed in a single outcrop correspond to acertain number of distinct stress regimes, which themselvescorrespond to one or more tectonic events. In such situationsof apparent or actual polyphase tectonism, reliable determi-nation of paleostress regimes requires separation of fault slipdata. From the tectonic point of view, it is important to dis-tinguish brittle structures related to distinct tectonic regimes,and it is still more important to separate structures resultingfrom different tectonic events. The analysis of syntectonicmineral infilling provides valuable information because

mineral development corresponds to pressure-temperatureand fluid-chemical conditions which may differ from oneevent to another. Particular attention was thus paid, in ourstudy, to the relationships between mineral development andshear or tension brittle deformation. The mineralogical na-ture of vein infill and of mineral steps on striated faults wasconsequently examined in detail, in close connection with thecollection of structural data.

The relative chronology of structures was also considered.In particular, numerous data on crosscutting relationships be-tween different generations of brittle structures (faults, veins,and dikes), as well as superimposed slickenside lineations,were collected. These data allowed distinction of tectonicevents based on a matrix method described previously by An-gelier (1994), taking into account the mechanical consistencyexpected for each stress regime. In most cases, a given stri-ated fault surface (or a vein) could be assigned to a specifictectonic regime. Detailed observations and the list of the ob-served crosscutting relationships are given in Carrier (1994).An automatic separation of stress tensors and subsets of faultslip data was also performed on the Silidor data, following aprocedure proposed by Angelier and Manoussis (1980) anddiscussed by Angelier (1983). This procedure is based oncluster analysis applied to stress-slip relationships combinedwith stress tensor determinations. Such an automatic separa-tion does not indicate chronology but allows rigorous separa-tion of mechanically consistent fault slip subsets. Finally,compilation of these different sources of information (syntec-tonic mineral development, relative chronology, mechanicalconsistency, and identification of neoformed fracture pat-terns) enabled us to establish the general chronology of brit-tle events at Silidor.

Methodological and technical aspects of stress tensor de-termination were described and discussed in detail by Ange-lier (1983, 1991, and 1994). The actual stress tensor has sixdegrees of freedom, but consideration of shear directions andsenses alone allows determination of four of these variables,namely the three principal stress axes σ1, σ2, and σ3 (maxi-mum, intermediate, and minimum principal stresses, respec-tively, with pressure considered to have a positive sign), and aratio between principal stress differences, which is conve-niently expressed by the φ ratio defined as (Angelier, 1975) φ= (σ2-σ3)/(σ1-σ3).

The two remaining stress tensor variables cannot be deter-mined unless rock mechanic properties (rupture and friction)and/or depth of overburden and fluid pressure condition (atthe time of brittle deformation) can be evaluated (see Ange-lier, 1989, for details). These two variables, which togetherwith the ratio φ control the magnitudes of all principalstresses, could not be determined with the data available inthe present case study.

Finally, an attempt was made to determine the geologic sig-nificance of the paleostress regimes in terms of both the type(reverse, strike slip, or normal), the trend of extension orcompression, and the φ ratio. Note, however, that the valuesof φ should be considered with care because reliable deter-mination is highly dependent on the presence of oblique-slipfaults, and large variations of φ are caused by stress variationsin rock masses. The computed orientations of axes (trendsand plunges) are much more stable, despite local effects.

1056 CARRIER ET AL.

0361-0128/98/000/000-00 $6.00 1056

Results

Eight brittle tectonic episodes have been reconstructed inthe Silidor mine. The numerical results of paleostress deter-minations are summarized in Table 2 and Figure 6.

The early event A is marked by dextral east-west to east-northeast–west-southwest and sinistral northwest-southeastfaults. Faults are filled with chlorite-ankerite. Tension cracksdisplaying the same association are northwest-southeast ori-ented and subvertical. All these elements mark northwest-southeast compression, with a pure compressive ellipsoid(Guiraud et al., 1989). Event B is the most important event interms of measured faults and slickensides. Most of the faultsstrike north-south to northeast-southwest and dip moderatelyto the west. Horizontal tension cracks are associated withthese faults and mark the first movement on the Silidor struc-ture. They also contain chlorite and carbonates, with locallysome hematite. These faults formed as as result of a clearnorthwest-southeast compression, with similar σ1 axes to theA episode. Note that A and B episodes simply differ by a per-mutation of σ2 and σ3, which suggests that they belong to thesame tectonic event.

Scarce dextral northeast-southwest-oriented and sinistraleast-west structures possibly reveal a third, C episode. Asso-ciated tension cracks are subvertical and strike northeast-southwest. Chlorite and ankerite are present in these faultsand veins, notably in the Silidor deposit. These elements in-dicate a possible east-northeast–west-southwest compression.The reconstruction of this episode, however, requires moredocumentation.

In contrast, the D episode is very well marked at the Silidordeposit, with numerous northwest-southeast reverse faultsdipping to the northeast. It corresponds to the faults occur-ring in the Silidor vein. Subhorizontal tension cracks are alsorelatively abundant and are filled with quartz, pyrite, chal-copyrite, and carbonates. Scarce faults showing the same styleof infilling strike north-south and east-west. Inverse modelingshows that they formed as a result of northeast-southwestcompression.

The E event is the first observed extensional event. It ismarked by conjugate sets of normal faults characterized by east-northeast–west-southwest and west-northwest–east-southeastdirections. The Silidor vein shows evidence of such exten-sional deformation, with normal offsets of moderate throw.

Associated tension cracks are subvertical and contain chlorite,sericite, and some sulfides.

Numerous, well-developed northwest-southeast strike-slipdextral faults and some northeast-southwest sinistral faultscharacterize event F. A sinistral displacement along the Héréfault, offsetting the Silidor vein, is related to this episode (Fig.7). Chlorite and ankerite crystallized on the slickensides. Awell-defined north-south compression is calculated with hor-izontal east-west-oriented extension.

The G event is also marked by numerous faults, trendingnortheast-southwest to east-west, with slickensides indicatingreverse movement. It is especially developed on the clay-filled Silidor and Héré faults. A north-south compression iswell defined, with σ1 axes identical to the F event. The F andG events are linked through a transition from strike-slipregime to reverse fault regime (permutation of σ2 and σ3).They may thus correspond to a single major tectonic event.

The latest episode reconstructed in Silidor, event H, ismarked by northeast-southwest normal faults (Héré fault).Downdip slickensides indicate normal movement and over-print slickensides from the A event. This last tectonic regimecorresponds to an extension with σ1 vertical, σ2 northeast-southwest, and σ3 (main extension) northwest-southeast.

Summary of the tectonic evolution

The tectonic evolution of the Silidor area, as for the wholePowell pluton, is clearly polyphase. Preore deformational fea-tures include the regional S2 foliation. Early movements re-lated to D2 are ductile and have not been quantitatively ana-lyzed in this study. The presence of the east-west-trendingschistosity (S2), however, suggests an approximately north-south-trending compression. The precise age of the D2 com-pression is unknown but it certainly occurred after 2673 Ma(see “Regional Geology”).

At Silidor, the east-west schistosity (S2) is locally crosscut byanother schistosity (S3), showing an oblique northeast-south-west orientation. This S3 foliation suggests a dextral move-ment on east-west faults which would be related to a north-west-southeast shortening. Such dextral displacements havebeen documented along the Cadillac-Larder Lake fault(Daigneault and Archambault, 1990; Robert, 1990; Bardouxet al., 1993) and are known to affect the Timiskaming sedi-ments. This S3 transpressive schistosity presents the same

SILIDOR DEPOSIT, ROUYN-NORANDA DISTRICT, ABITIBI BELT 1057

0361-0128/98/000/000-00 $6.00 1057

TABLE 2. Paleostress Results Determined at the Silidor Deposit

Phase Number Mean of faults variation1 σ1 σ2 σ3 φ Chronology

A 20 28° 299°/19° 124°/71° 030°/01° 0.43 PreoreB 135 10° 307°/02° 217°/09° 048°/81° 0.56 PreoreC 15 14° 077°/12° 253°/78° 346°/01° 0.18 SynoreD 49 11° 229°/06° 139°/02° 033°/84° 0.68 SynoreE 15 14° 246°/79° 108°/08° 017°/07° 0.26 SynoreF 38 16° 356°/05° 250°/71° 088°/18° 0.26 PostoreG 58 9° 349°/06° 259°/01° 155°/84° 0.68 PostoreH 19 13° 023°/80° 223°/10° 132°/03° 0.34 Postore

For each stress tensor the following parameters are presented: the number of fault slip data used; a quality estimator corresponding to the 1mean devia-tion between theoretical and measured striation directions; the orientation of the maximum and minimum stress axes (σ1 = maximum stress axes, σ2 = in-termediate stress axes, σ3 minimum stress axes); the calculated value of the ratio between stress differences, φ = (σ2-σ3)/(σ1-σ3); these eight episodes are as-signed pre-, syn-, and postore using mineralogical markers

main compressive axes as the A and B tectonic period. Thetwo early northwest-southeast compressive events (A and B)were clearly developed before the mineralizing event of theSilidor deposit. This deformation may also be synchronouswith the deposition of the copper-gold-molybdenum (early)association in the Don Rouyn deposit and with chlorite-sericite alteration (Goldie et al., 1979). Deformation of thisage seems to control the development of hematitic alteration,both regionally and along the preore Silidor structure.

The synore tectonic period is marked by a northeast-south-west compression. This period includes the C and D com-pression episodes, with northeast-oriented σ1 axes. Duringthis period, the Silidor structure acted as a reverse fault, in re-sponse to a northeast-southwest regional compressive event.Ductile deformation was localized in the northwest-southeastdioritic dike, and large variations in deformation intensity areprobably related to strain partitioning. Strain softening led tolocalization of the slip into and near the core of the fault andwas probably assisted by suprahydrostatic pore-fluid pressure(Little, 1995). Reverse movement continued after the firstmineral deposition, as shown by the distribution of white bar-ren breccia along horizontal opening zones. Variations in fluidpressure explain the existence of both tensional and compres-sive local features. The mineralizing episodes (C and D) havebeen dated at 2562 ± 12 Ma, using Pb-Pb in pyrite, quartz,and chlorite (Carignan and Gariépy, 1993). Such an age is

1058 CARRIER ET AL.

0361-0128/98/000/000-00 $6.00 1058

E e

F f

G g

H h

A a

B b

C c

D d

NW

-SE

CO

MP

RE

SS

ION

NE

-SW

CO

MP

RE

SS

ION

NN

E E

XTE

NS

ION

N-S

CO

MP

RE

SS

ION

NW

-SE

EX

TEN

SIO

N

FIG. 6. Chronological succession and illustration of the stress tensors obtained by the Angelier method. a. Phase A. b.Phase B. c. Phase C. d. Phase D. e. Phase E. f. Phase F. g. Phase G. h. Phase H. By using mineralogical markers, phases Aand B were assigned preore, phases C, D, and E were assigned synore, and phases F, G, and H were assigned postore. Thefive, four, and three branch stars, respectively, correspond to σ1, σ2, and σ3. Arrows indicate extensional or compressionalaxes. Lower hemisphere projections.

25 m

58

HÉRÉ FAULT

57

Hematite-alteredtrondhjemiteMineralizedtrondhjemite

Dioritic dike

QFP felsic dike

Vein

Fault breccia

FIG. 7. Plan view of the Héré fault on the 220-m south level of the Silidormine, showing a late postmineralization strike-slip movement along this fault.

similar to resetting ages of numerous other gold deposits inthe southern Abitibi province (Kerrich and Cassidy, 1994).Late barren infilling in the Silidor vein indicates a late brittlenorth-northeast-oriented extension (episode E).

The postore period was a time of north-south compression(episodes F and G) and of a late northwest-southeast (episodeH) extension. Clay lining of faults suggests low temperaturesduring movement and no obvious connection with the earliermineralizing fluids. Weak alteration of host rocks suggeststhat fluid flow was preferentially within the fault. The ages ofthese postore episodes are unknown but may be correlatedregionally through Ontario and Québec. Cobalt group sedi-ments are crosscut by numerous faults which indicate north-south Proterozoic compression (Faure et al., 1991; Powelland Hodgson, 1992). East of the Rouyn-Noranda area, sinis-tral movements along the Bousquet or Kinojevis northeast-southwest faults offset the Cadillac-Larder Lake break (Bar-doux et al., 1993). Within the Flavrian pluton, the HunterCreek fault is parallel to the Horne fault and shows a late re-verse movement (Camiré and Watkinson, 1990). All these ob-servations suggest a Proterozoic age for these episodes (F andG, and event H), possibly during the Grenvillian orogeny,around 1250 Ma, as shown by the thermal resetting of iso-topic systems (Corfu, 1993).

Paleostress MappingThe Silidor-Don Rouyn area constitutes a homogeneous

district, considering both the tectonic style and the nature ofthe infilling. Lode gold deposits feature similar mineralogy,isotopic signature, and style of deposition (cf. Kerrich, 1983;Robert, 1991). This provinciality probably reflects the pres-ence of large volumes of hydrothermal solutions of similarcomposition (i.e., reservoirs).

The distribution and movement of hydrothermal fluidswithin a district are related to several parameters. The mainfactors are the geometry of the fissured media and the distri-bution of pressure gradients within the crust at that moment.Pressure gradients can be set up by deformation, thermalprocesses, or fluid buoyancy (Oliver et al., 1990). As has beennoted in several other areas of the Abitibi greenstone belt(Hodgson, 1990; Robert, 1991), the distinct association be-tween deformation and ore deposition in the Silidor area isindicative that pressure gradients are mainly induced bydeformation.

Methodology

Deposition of gold in shear zones is controlled by themovement of hydrothermal solutions within the midcrust.The hydrology of such circulation is not accessible to directobservation, but theoretical calculations and analogy with themigration of hydrocarbons in sedimentary basins can be madebased on the distribution of pressure in the crust and fluidbuoyancy (Oliver et al., 1990). At a few kilometers depth inthe crust, fluid pressures are most likely to be greater than hy-drostatic and are controlled in part by rock pressure. Fluidflow is generally upward directed (Holyland et al., 1993), butpetrologic investigations and analogies with petroleum reser-voirs indicate that it could also be lateral, over large distances,following the lithologic units with the highest permeability(Ferry, 1994). Hydrothermal fluids will accumulate either

near zones of low permeability (which act as barriers) or inzones of low pressure (= low mean stress). Such areas may ap-pear due to the stress pattern in relationship to rheologicalvariations, particularly competency contrasts (Oliver et al.,1990; Ridley, 1993; Holyland and Ojala, 1997).

Our tectonic model allows us to develop a techniquewhereby variations of the mean stress can be reconstructed,using the orientation of the main stress axes and geologicmapping. Stress mapping technology (STM™) based on fi-nite-difference techniques which is well suited for discontin-uum modeling (Lemos et al., 1985; Holyland, 1990) was de-veloped by Terra Sancta Inc. It examines the variation instrain and stress through an inhomogeneous terrane upon im-position of a regional stress field. The geology of the area istreated as a mosaic of polygonal blocks composed of homoge-neous rock units (modeled as Coulomb materials) and joints(faults and shear zones). It is, however, necessary to recon-struct the geology as it was during the mineralizing event,therefore, around 2.56 Ga.

Production of the stress map requires estimates of the mag-nitudes and orientations of the far-field horizontal stresses.The latter are given by the previous quantitative analysis ofthe brittle structures, but the former are difficult to obtainwith the present state of knowledge. Empirical values previ-ously used in similar terranes of Western Australia and Africawere adopted (Holyland, 1990; Vearncombe and Holyland,1995; Holyland and Ojala, 1997). For a depth of about 4 to 8km, assuming pressure increases with depth at about 25MPa/km, the chosen values for the stress field for σ1 and σ3are 100 and 70 Mpa, respectively. Rock and fault deformationproperties during the Archean are also required to model theinteraction of an assemblage of blocks. The required variablesinclude tensile and compressive strength, bulk and shearmoduli, and density for each lithology, as well as the frictionangle and stiffness for each contact and fault.

Modeling was carried out using a two-dimensional distinctelement code (UDEC) on a PC-compatible computer. Themethod utilizes an explicit time-stepping dynamic algorithmwhich allows displacements and rotations and general nonlin-ear constitutive behavior for both the matrix and discontinu-ities. When an external stress is applied to this system, theblocks are juggled and internally deformed until equilibriumis attained (Lemos et al., 1985). The product of these compu-tations is a stress contour map. Mean or minimum stress mapscontour the value of the mean or minimum compressivestress and show typically the same anomalies.

Computing was done by one of the authors (P.H.) withoutknowledge of the location of the ore deposit: this approachcould be compared to the blind method used in biological sci-ences. A highly simplified map of the Powell pluton was used,showing only the three main units (trondhjemite, diorite, andvolcanic rocks) and some of the known veins. The relativestrengths of the rock are, in decreasing order, diorite > vol-canic rocks > trondhjemite > shear zone. Volcanic rocks arecomposed of thick horizons of andesite and thin layers of rhy-olite. Rheologic properties are that of andesite. Fault zoneswere divided into first- and second-order faults. Three caseswere studied, using northwest-, north-, and northeast-di-rected compression, corresponding to the result of the fieldobservations and paleostress tensor reconstruction.

SILIDOR DEPOSIT, ROUYN-NORANDA DISTRICT, ABITIBI BELT 1059

0361-0128/98/000/000-00 $6.00 1059

Results

The results of district modeling experiments show an inho-mogeneous paleostress distribution, with alternating zones ofhigh and low pressures along the Horne Creek fault (Fig. 8).The direction of the main compression axes does not drasti-cally modify the distribution of paleopressure zones. Themain features of the north-south compression modeling ex-periment are the following (Fig. 8d): (1) a distinctive zonewith alternating low- and high-pressure areas along theHorne Creek fault. Low-pressure zones are associated withthe intersection of the east-northeast–west-southwest HorneCreek fault and northwest-southeast-striking faults, eitherknown (Silidor, Smokey Creek faults) or inferred; and (2) thecontrasting behavior of the Beauchastel fault, which, despitehaving a similar orientation and mechanical character to theHorne Creek fault, did not develop areas of high and lowpressure along its length. This result is related to the homo-geneity of the block to the north. The only zone of low pres-sure is located at the end of the Silidor-Smokey Creek north-west-southeast-striking faults.

Modeling of a northwest-southeast compression (Fig. 8b)shows a similar distribution. Two more zones of contrastinghigh pressure occur along the Horne Creek fault betweenSilidor and the eastern end of the Powell pluton and at thecontact between diorite and volcanic rocks. These zonescould be related to the embayment of volcanic rocks betweentwo resistant dioritic masses. The high-pressure zone of Sili-dor moves toward the southwest. A new zone of low pressureappears at the eastern border of the map. The northeast-southwest compression paleostress modeling features thesame characteristics with high pressure oriented perpendicu-lar to σ1 and a particularly well-developed low-pressure arealocated at the Silidor deposit (Fig. 8c).

The location of the Silidor gold-bearing quartz vein is pre-dicted by the paleostress modeling. The best fit is observedwith the northeast-southwest compression, which is the di-rection of compression deduced from the microtectonicanalysis. The presence of zones of relatively high pressurealong some of the preexisting joints helps to explain the non-percolation of the hydrothermal solutions and the uneven dis-tribution of mineralization. New potential gold-bearing zonesare also suggested from these maps.

1060 CARRIER ET AL.

0361-0128/98/000/000-00 $6.00 1060

FIG. 8. Results of stress mapping modeling of the Silidor area (low-mean stress plot). a. Simplified geologic map of theeastern part of the Powell pluton. Results of the paleostress modeling, using b. Northwest-southeast, c. Northeast-southwest,and d. North-south compression. The Silidor lode is illustrated in black on the four maps. Legend: Crosses = Powell pluton,Hachures = diorite, White = Blake River Group volcanic rocks, full line = major faults, dashed line = minor faults, dottedline = inferred faults, H = high-pressure zone, L = low-pressure zone.

Discussion

The evolution of the Silidor deposit constitutes a model forlode gold deposits. It shows some similarities with the classiclode gold deposits, such as Sigma (Robert and Brown, 1986a),but also significant differences (Table 3).

Despite their similar ages, the Sigma and Silidor depositsformed in different tectonic settings. The Sigma deposit wasemplaced under a compressive north-south deformationregime, with the hydrothermal conduit being a high-angle re-verse fault zone during fault-vaulting episodes (Sibson et al.,1988). The crosscutting relationships at Sigma imply that thestress pattern remained constant during the whole mineral-ization process. On the other hand, the Silidor deposit dis-plays an evolving stress field during vein infilling, with σ1 re-maining northeast-southwest, and an exchange between σ2and σ3 stress axes, resulting in an evolution from strike-slipfaulting to reverse faulting. The Silidor deposit shows lesscyclicity during the hydrothermal events. This could be be-cause the hydrothermal fluid flow operated under a near-con-stant hydraulic gradient (as at the Francoeur deposit inAbitibi; Couture and Pilote, 1993) or because of the greatertime variability of the local stress. These local stress variationscould be related to an oblique collision mechanism. A similartranspressional environment has been described in theNorseman-Wiluna belt of Western Australia (Hagemann etal., 1992). Silidor shares numerous features with this Aus-tralian district, especially the strike-slip movement of thefaults, the scarcity of extension veins, and the abundance of

implosion breccias. Such features are typical of a brittle envi-ronment, with few pressure seals. Fluid movement couldhave been related to suction pump mechanisms (Sibson et al.,1988).

The grade of metamorphism and vein morphologies of theSigma and Silidor deposits also differ (Table 3). The Silidorveins show some similarities with the high-level brittle-stylemineralization at Wiluna in Australia (Vearncombe et al.,1989; Hagemann et al., 1992). In the Rouyn-Noranda district,the Francoeur deposit, located 15 km southwest of Silidor,shows the same chronological order for alteration assem-blages as at Silidor, with hematite preceding the albite-pyritealteration. The Francoeur deposit is a ductile-shear dissemi-nated replacement type mesothermal deposit (quartz-carbon-ate veins absent; Couture and Pilote, 1993), hosted in highermetamorphic grade rocks (greenschist-amphibolite facies:actinolite-oligoclase zone; Powell et al., 1995b). In light of thedepth continuum model of Groves et al. (1992), Francoeurcould be a deeper expression and Silidor a high-crustal levelexpression of a regional mesothermal system.

The geologic and rheologic settings at Sigma and Silidor aredifferent. The Silidor lode gold deposit is a rather simple min-eralized structure hosted in a relatively homogeneous geo-logic context (trondhjemite sill). In this case, anisotropiczones such as mafic dikes played a major role in the localiza-tion of shear zones. Vein styles at Silidor are characteristic ofother lode gold deposits which are hosted in similarly homo-geneous and competent environments. The Elder and Pierre-Beauchemin veins in the Flavrian pluton, Rouyn-Noranda

SILIDOR DEPOSIT, ROUYN-NORANDA DISTRICT, ABITIBI BELT 1061

0361-0128/98/000/000-00 $6.00 1061

TABLE 3. Comparison between the Gold-Bearing Quartz Deposits of the Rouyn-Noranda and the Sigma-Val d’Or Districts

Characteristics Sigma deposit1 Silidor deposit2

(Val d’Or district, southeastern Abitibi belt) (Rouyn district, south-central Abitibi belt)

Structural context Reverse faulting (high-angle fault) Strike-slip, then reverse faulting (moderate-angle fault)

Shear veins Abundant (0.2–3 m width) Few (0.05–0.3 m width)

Extension veins Numerous and gold rich (1–100 cm width) Few and noneconomic (1–20 cm width)

Tension crack Abundant Limited

Breccia Hydraulic breccia Hydraulic breccia and collapse breccia with buck quartz infilling

Open-space filling Comb quartz grown in flat extension veins Abundant small drusy veinlets and vugs (both flat and and cavities fault-parallel)

No cavities

Few cavities (of 1–2 m scale) Mineral assemblage Qz, Tm, Cb, Py, Sh, Po, Cp, Bt, Qz, Cb, Py, Cp, Au-Ag, Fu, Ser, Ms, Hm, Gn, Mo,

Au-Ag, Pz, Tb, Mo, Fro, Il, Sp Gr, Tb, AuTe, Pz, El, Hs, Fre, Il, Pr, Ba

Alteration Chlorite-carbonates-white micas, carbonates-white Quartz-carbonates-hematite, carbonate-sericite-fuchsite, micas, carbonates-albite carbonates-albite

fO2Below the ΣSO4-H2S buffer On the ΣSO4-H2S buffer

Regional level of Greenschist-amphibolite facies transition Greenschist facies (actinolite-albite)3

metamorphism (actinolite-albite, near surface, to actinolite-oligoclase and hornblende at depth)3

1 Sources for the Sigma deposit: Robert and Brown (1984), Robert and Brown (1986a and b), Robert and Kelly (1987)2 Sources for the Silidor deposit: Carson (1986), Gaulin (1992), Carrier (1994)3 Source : Powell et al. (1995b) Mineral abbreviations: Ag = silver, Au = gold, AuTe = gold tellurides, Ba = barite, Bt = biotite, Cb = carbonates, Cp = chalcopyrite, El = electrum,

Fre = freibergite, Fro = frohbergite, Fu = fuchsite, Gn =galena, Hm = hematite, Hs = hessite, Il=ilmenite, Mo = molybdenite, Ms = muscovite,Po = pyrrhotite, Pr = proustite-pyrargyrite, Py = pyrite, Pz = petzite, Qz = quartz, Ser = sericite, Sh = scheelite, Sp = sphalerite, Tb = tellurobismuthinite,Tm = tourmaline

area (Gaulin and Trudel, 1990; Richard et al., 1990), showsimilar veins styles to Silidor. The numerous veins hosted inthe Bourlamaque pluton, Val d’Or area (Vu, 1990; Belkabir etal., 1993), show similarities with Silidor but without thestrike-slip movement along the fault.

At Silidor, the combination of homogeneous and compe-tent units and east-northeast–west-southwest-bounding faults(Horne and Beauchastel) in a strike-slip regime can also ex-plain differences with Sigma. The distribution on a set ofnorthwest-trending faults of the lode deposits of the Powellpluton resembles the results of simple shear experiments in asandbox (cf. Naylor et al., 1986). The distribution of the lodedeposits may also reflect the periodicity of northwest-trend-ing diorite dikes. The early northwest-southeast compressiveevent (A and B stress states) produced dextral movement oneast-northeast–west-southwest faults (Horne and Beauchas-tel). This also led to the development of northwest-southeastveins, which could be tension cracks or rotated Riedel shears(preore hematized shear zones) of a helicoidal form (Nayloret al., 1986; Sylvester, 1988). This event is followed by thenortheast-southwest synore compression event (C and Dstress states), which would have caused sinistral movement onthe Horne and Beauchastel faults and the development of re-verse faulting in the lode deposits.

The rheological contrast between units and the existence ofsignificant permeability allows a vertical transfer of solutions.The presence of a low-pressure zone which remains stable inthe southern part of the Silidor deposit, where it crosses theHorne Creek fault, suggests that this area could have acted asa sink for hydrothermal fluid throughout the tectonic history.The horsetail located at the northern part of the deposit sug-gests that fracture propagation occurred from south to north,activated by the concentration of fluid in the south. Variationsin the direction of the main compression slightly modified thedistribution of the paleobarometric zones. The paleostressmap suggests that circulating fluid mainly followed majorductile structures such as the Horne Creek and Cadillacbreaks or the areas of low stress that these large structurescreated, with possible oblique fluid flow. Very few changesappear, with even a drastic rotation of the stress field, whichindicates that different structural contexts, marking differenttimes of ore deposition, could result in repeated fluid flow ata deposit.

On a regional scale, the Silidor area is located near a zoneof relative indentation, south of the Horne fault, correspond-ing to an area of dioritic rocks within volcanic rocks. There isalso abundant evidence of simultaneous high-fluid pressurezones, as demonstrated by hydraulic breccias and fluid flowduring reverse faulting. Ridley (1993) has concluded that low-mean stress zones would be zones in which hydrostatic pres-sures of fractures would be relatively high relative to rockpressure.

The hydrothermal system at Silidor has a lower content ofB and As and a higher content of Mo, compared to Sigma, butboth carried Au and W. Conditions of gold deposition wereclearly much more oxidizing at Silidor than at Sigma, asdemonstrated by the early abundance of hematite and theδ18O values of quartz. Oxidizing conditions of gold depositionhave also been documented in other deposits of the Rouyn-Noranda district, by δ18O and δD values in adjacent quartz

veins at Elder and Pierre-Beauchemin (Kennedy and Ker-rich, 1982), and by δ34S values at Francoeur (Couture and Pi-lote, 1993). Oxidizing conditions of gold deposition seem tobe a characteristic of the lode gold deposits of the Rouyn-No-randa district and can explain the lack of boron and arsenic inthese deposits.

ConclusionsThe Silidor deposit is the result of several events of vein fill-

ing (white, gray, and smoky quartz) and associated metaso-matism of its host rocks (mineralized trondhjemite and car-bonate- fuchsite breccia). The early hematized alterationenvelope and δ34S values for pyrite imply oxidizing conditionsduring gold deposition at Silidor, and this seems to be a gen-eral characteristic for most of the late-orogenic lode gold de-posits of the Rouyn-Noranda area.

The structural evolution at Silidor was polyphase, charac-terized by large variations in orientations of the reducedstress tensors. Two major phases occurred before the miner-alizing events: a ductile north-south and a brittle-ductilenorthwest-southeast compressive episode. The tectonic andmetallogenic significance of the northwest-southeast com-pressive episode is illustrated by the deposition of the copper-gold-molybdenum porphyry-type mineralization at DonRouyn, and it also played a role in the development of thehematite alteration. The Silidor deposit is related to a north-east-southwest compressive episode, during which σ1 re-mained northeast-southwest but σ2 and σ3 inverted at somestage during the event. The stress axes inversion during themineralizing event implies a change from strike-slip faultingto reverse faulting and could be related to an oblique collisionmechanism (Fig. 9).

1062 CARRIER ET AL.

0361-0128/98/000/000-00 $6.00 1062

FIG. 9. Block diagram showing relationship between a first-order fault(Cadillac break), second-order faults (Horne Creek, Beauchastel, SmokeyCreek faults), and the third-order faults (the mineralized faults). Arrows il-lustrates upward-ascending hydrothermal fluids. Legend: A = Astoria, AR =Anglo-Rouyn, AT = Arntfield, AU = Augmitto, BF = Buffam, BH = BoulderHill, BZ = Bazooka, C = Cinderella, E = Peel-Elder, GR = Graham, H = Hal-liwell, L = Lac Pelletier, MC = MacDonald, NM = New Marlon, PB =Pierre-Beauchemin (Eldrich), PR = Powell-Rouyn, Q = Quésabé, SD =Stadacona, SIL = Silidor, SY = Sylvie, SR = Senator-Rouyn, W = Wingate,WA = Wasamac.

On a regional scale, the paleostress model maps show theposition of the Silidor deposit as an area of permanent lowmean stress near the Horne Creek fault during the LateArchean tectonic history. The timing of the Silidor mineral-ization, related to the northeast-southwest compressiveepisode, coincides with the best fit for the Silidor locationunder a northeast-southwest compression. In the Rouyn-No-randa area, gold deposits are associated with east-north-east–west-southwest and northwest-southeast second-orderfaults. However, our paleostress modeling suggests that someof these faults, like the Beauchastel fault, did not act as anarea of high and low pressure because of the homogeneity ofthe blocks on either side. Modeling suggests that inhomo-geneity or indentation along these faults was needed to createdistinctive zones of low and high pressure, which seem to becharacteristic for the mineralized second-order faults.

The combination of paleostress studies and inverse model-ing of the distribution of paleopressure offers a new way toexplain the distribution of mesothermal ore deposits and topredict the location of new gold targets.

AcknowledgmentsWe are grateful to Hemlo Gold Mines and Cambior Inc.,

owners of Les Mines Silidor Inc., for financial support and ac-cess to the mine. We thank the chief geologist C.J. Kloerenand the geologic staff, G. Laperle, S. Picard, V. Fillion, K.Germain, M. Dionne, and M. Gagnon, for valuable assistanceand numerous discussions. Assistance and support to the firstauthor was also provided by Y. Lei, M. Gauthier, M. Bardoux,J. Lafrance, J. Bourne, T.C. Birkett, J. Moorhead, and T. Arm-strong. We also thank two Economic Geology reviewers, W.Powell and J. Ojala, and J. Ridley from the Editorial Board,for their insightful corrections and very helpful suggestionswhich greatly improved the final version of the manuscript.Funding to A.C. was provided by a grant for an M.Sc. studyfrom the Fonds pour la formation de chercheurs et l’aide à larecherche (FCAR). Tom Powell found the first gold depositin this area in 1922 and a team of geologists from NorandaExploration found the Silidor deposit 63 years after Powell’sdiscovery, only 3 miles away.

January 8, 1998 ; April 3, 2000

REFERENCESAngelier, J., 1975, Sur l’analyse de mesure recueillis dans des sites faillés: l’u-

nité d’une confrontation entre les méthodes dynamiques et cinématiques:Comptes rendus de l’Académie des Sciences de Paris, v. 281, ser. D, p.1805–1808.

——1983, Analyses qualitative et quantitative des populations de jeux defailles: Bulletin de la Société géologique de France, v. 25, p. 661–672.

——1984, Tectonic analysis of fault slip data sets: Journal of Geophysical Re-search, v. 89, p. 5835–5848.

——1989, From orientation to magnitudes in paleostress determinationsusing fault slip data: Journal of Structural Geology, v. 11, p. 37–50.

——1991, Inversion directe et recherche 4-D: comparaison physique etmathématique de deux modes de détermination des tenseurs de paléocon-traintes en tectonique de failles: Compte Rendus de l’Académie des Sci-ences de Paris, v. 312, ser. II, p. 1213–1218.

——1994, Fault slip analysis and paleostress reconstruction, in Hancock,P.L., ed., Continental deformation: Tarrytown, NY, Pergamon Press, p.53–100.

Angelier, J., and Manoussis, S., 1980, Classification automatique et distinc-tion des phases superposées en tectonique de failles: Compte Rendus del’Académie des Sciences de Paris, v. 290, ser. D, p. 651–654.

Bardoux, M., Goulet, N., Jébrak, M., and Morin, D., 1993, Tecto-nostratigraphie et minéralisations en or sur la faille Cadillac (McWatters,Abitibi, Québec): Ministère de l’Énergie et des Ressources naturelles duQuébec DV 92-07, p. 9–32.

Belkabir, A., Robert, F., Vu, L., and Hubert, C., 1993, The influence of dikeson auriferous shear-zone development within granitoid intrusions: theBourlamaque pluton, Val d’Or district, Abitibi greenstone belt: CanadianJournal of Earth Sciences, v. 30, p. 1924–1933.

Bott, M.H.P., 1959, The mechanics of oblique slip faulting: Geological Mag-azine, v. 96, p. 109–117.

Cameron, E.M., and Hattori, K., 1987, Archean gold mineralization and oxi-dized hydrothermal fluids: ECONOMIC GEOLOGY, v. 82, p. 1177–1191.

Camiré, G., and Watkinson, D.H., 1990, Volcanic stratigraphy and structurein the Hunter Creek fault area, Rouyn-Noranda, Québec: Canadian Jour-nal of Earth Sciences, v. 27, p. 1348–1358.

Carey, E., and Brunier, B., 1974, Analyse théorique et numérique d’un mod-èle mécanique élémentaire appliqué à l’étude d’une population de failles:Comptes rendus des séances de l’Académie des sciences, v. 279, ser. D, p.891–894.

Carignan, J., and Gariépy, C., 1993, Pb isotope geochemistry of the Silidorand Launay gold deposits: Implications for the source of Archean Au in theAbitibi subprovince: ECONOMIC GEOLOGY, v. 82, p. 1931–1957.

Carrier, A., 1994, Évolution structurale et métallogénique du gisement au-rifère Silidor, Abitibi, Québec: Montréal, Québec, Unpublished M.Sc. the-sis, Université du Québec à Montréal, 272 p.

Carrier, A., and Jébrak, M., 1996, The Silidor-Don Rouyn area of the CentralNoranda Camp: An illustration of three successive metallogenic events inthe southern Blake River Group, Abitibi, Québec, in Péloquin, A.S., ed.,The Blake River Group: A new look at an old volcano: A do-it yourself fieldguide, p. 1–11.

Carson, D.J., 1986, Mineralogy and occurrence of gold at the Silidor prop-erty: Rouyn-Noranda, Quebec, Mines Silidor Inc., Internal report, 35 p.

Corfu, F., 1993, The evolution of the southern Abitibi greenstone belt in lightof precise U-Pb geochronology: ECONOMIC GEOLOGY, v. 88, p. 1323–1340.

Cousineau, P.A., 1981, Compilation de la mine Anglo-Rouyn, Projet Héré(98-446): Sainte-Foy, Québec, Société québécoise d’exploration minérale,Unpublished report, 29 p.

Couture, J.F., and Pilote, P., 1993, The geology and alteration patterns of adisseminated, shear zone-hosted mesothermal gold deposit: The Francoeur3 deposit, Rouyn-Noranda, Quebec: ECONOMIC GEOLOGY, v. 88, p. 1664–1684.

Couture, J.F., Pilote, P., Machado, N., and Desrochers, J.P., 1994, Timing ofgold mineralization in the Val d’Or district, southern Abitibi belt: Evidence fortwo distinct mineralizing events: ECONOMIC GEOLOGY, v. 89, p. 1542–1551.

Daigneault, R., and Archambault, G., 1990, Les grands couloirs de déforma-tion de la sous-province de l’Abitibi: Canadian Institute of Mining and Met-allurgy, Special Volume 43, p. 43–64.

Davis, D.W., 1991, Age constraints on deposition and provenance of Archeansediments in the southern Abitibi and Pontiac subprovinces from U-Pbanalysis of detrital zircons. Abitibi-Greenville transect: Lithoprobe Work-shop, Montréal, Report 25, p. 147–150.

de Rosen Spence, A.F. , 1976, Stratigraphy, development and petrogenesis ofthe central Noranda volcanic pile, Noranda, Québec: Unpublished Ph.D.thesis, Toronto, Ontario, University of Toronto, 166 p.

Dimroth, E., Imreh, L., Goulet, N., and Rocheleau, M., 1983, Evolution ofthe south-central part of the Archean Abitibi belt, Quebec. Part II: Tec-tonic evolution and geomechanical model: Canadian Journal of Earth Sci-ences, v. 20, p. 1355–1373.

Dupin, J.-M., Sassi, W., and Angelier, J., 1993, Homogeneous stress hypoth-esis and actual fault slip: A distinct element analysis: Journal of StructuralGeology, v. 15, p. 1033–1043.

Faure, S., Jébrak, M., and Angelier, J., 1991, E-W compression in the Abitibigreenstone belt: A numerical approach [abs.]: Geological Association ofCanada-Mineralogical Association of Canada, Program with Abstracts, v.16, p. A 36.

——1996, Structural evolution of Les Mines Selbaie, northern Abitibi belt,Québec, Canada: Exploration and Mining Geology, v. 5, p. 215–230.

Feng, R., Fan, J., and Kerrich, R., 1993, Noble metal abundances and char-acteristics of six granitic magma series, Archean Abitibi belt, Pontiac sub-province: Relationships to metallogeny and overprinting of mesothermalgold deposits: ECONOMIC GEOLOGY, v. 88, p. 1376–1401.

Ferry, J.M., 1994, Overview of the petrologic record of fluid flow during re-gional metamorphism in northern New England: American Journal of Sci-ence, v. 294, p. 905–988.

SILIDOR DEPOSIT, ROUYN-NORANDA DISTRICT, ABITIBI BELT 1063

0361-0128/98/000/000-00 $6.00 1063

Frape, S.K., and Fritz, P., 1987, Geochemical trends from groundwatersfrom the Canadian Shield: Geological Association of Canada Special Paper33, p. 19–38.

Gaulin, R., 1992, Métallogénie de la mine Silidor, Abitibi, Québec: Ministèrede l’Énergie et des Ressources du Québec MB 92-17, 58 p.

Gaulin, R., and Trudel, P., 1990, Caractéristiques pétrographiques etgéochimiques de la minéralisation aurifère à la mine Elder, Abitibi,Québec: Canadian Journal of Earth Sciences, v. 27, p. 1637–1650.

Goldie, R.J., 1976, The Flavrian and Powell plutons, Noranda area, Québec:Unpublished Ph.D. thesis, Kingston, Ontario, Queen’s University, 356 p.

Goldie, R.J., Kotila, B., and Seward, D., 1979, The Don Rouyn mine: AnArchean porphyry copper deposit near Noranda, Quebec: ECONOMIC GE-OLOGY, v. 74, p. 1680–1684.

Grant, J.A., 1986, The isocon diagram: A simple solution to Gresens’equationfor metasomatic alteration: ECONOMIC GEOLOGY, v. 81, p. 1976–1982.

Groves, D.I., Barley, M.E., Cassidy, K.C., Fare, R.J., Hagemann, S.G., Ho,S.E., Hronsky, J.M.A., Mikucki, E.J., Mueller, A.G., McNaughton, N.J.,Perring, C.S., Ridley, J.R., and Vearncombe, J.R., 1992, Subgreenschist togranulite-hosted Archean lode-gold deposits of the Yilgarn craton, A depo-sitional continuum from deep-sourced hydrothermal fluids in crustal-scaleplumbing systems: International Archean Symposium, 3rd, University ofWestern Australia, Geology Department University Extension, Proceed-ings, v. 22, p. 325–337.

Guiraud, M., Laborde, O., and Philip, H., 1989, Characterization of varioustypes of deformation and their corresponding deviatoric stress tensorsusing microfault analysis: Tectonophysics, v. 170, p. 289–316.

Hagemann, S.G., Groves, D.I., Ridley, J.R., and Vearncombe, J.R., 1992, TheArchean lode gold deposits at Wiluna, Western Australia: High-level brit-tle-style mineralization in a strike-slip regime: ECONOMIC GEOLOGY, v. 87,p. 1022–1053.

Hodgson, C.J., 1990, An overview of the geological characteristics of gold de-posits in the Abitibi subprovince: University of Western Australia, GeologyDepartment (Key Centre) and University Extension Publication 24, p.63–100.

Holyland, P.W., 1990, Targeting of epithermal ore deposits using stress map-ping techniques: Australia Institution of Mining and Metallury Pacific RimCongress, Melbourne, Proceedings, v. 3, p. 337–341.

Holyland, P.W., and Ojala, V.J., 1997, Computer-aided structural targeting inmineral exploration: Two- and three-dimensional stress mapping: Aus-tralian Journal of Earth Sciences, v. 44, p. 421–432.

Holyland, P., Ridley, J.R., and Vearncombe, J.R., 1993, Stress mappingtechnology (STMTM) [abs.]: Geological Society of London Geofluid’s 93,Contributions to an International Conference on Fluid Evolution,Migration and Interaction in Rocks, Torquay, England, Abstracts, p.272–275.

Hubert, C., Trudel, P., and Gélinas, L., 1984, Archean wrench fault tectonicsand structural evolution of the Blake River Group, Abitibi belt, Québec:Canadian Journal of Earth Sciences, v. 21, p. 1024–1032.

Jackson, S.L., Cruden, A.R., White, D., and Milkereit, B., 1995, A seismic-reflection-based regional cross section of the southern Abitibi greenstonebelt: Canadian Journal of Earth Sciences, v. 32, p. 135–148.

Jébrak, M., 1991, Les gisements d’or dans les tonalites archéennes (Abitibi,Québec): Mineralium Deposita, v. 27, p. 1–9.

——1997, Hydrothermal breccias in vein-type ore deposits: A review ofmechanisms, morphology and size distribution: Ore Geology Reviews, v.12, p. 111–134.

Jébrak, M., and Faure, S., 1992, Paleo-stress evolution of the southernAbitibi greenstone belt. [abs.]: Geological Association of Canada-Miner-alogical Association of Canada, Program with Abstracts, v. 17, p. A52.

Jébrak, M., and Hernandez, A. 1995, Tectonic deposition of mercury in theAlmadén district (Las Cuevas deposit, Ciudad Real, Spain): MineraliumDeposita, v. 30, p. 413–423.

Jébrak, M. , Harnois, L., Carrier, A., and Lafrance, J., 1995, The Don RouynCu-Au porphyry system: Metallogeny and Tectonic Evolution of theRouyn-Noranda Region, Precambrian ‘95, Canadian Institute of Miningand Metallurgy Field Trip Boooklet, p. 83–87.

Keating, P., 1992, Interpretation of the gravity anomaly field in the NorandaVal d’Or region, Abitibi Greenstone belt, Canadian Shield: Canadian Jour-nal of Earth Sciences, v. 25, p. 962–971.

Kennedy, L.P., and Kerrich, R., 1982, Transition from marine to meteoriticwater hydrothermal regimes in an emerging Archean intrusive complex:18O evidence from the Morran pluton [abs.]: American Geophysical Union,San Francisco, 1982, Abstracts,

Kerr, D.J., and Gibson, H.L., 1993, A comparison of the Horne volcanogenicmassive sulfide deposit and intracauldron deposits of the mine sequence,Noranda, Quebec: ECONOMIC GEOLOGY, v. 88, p. 1419–1442.

Kerr, D.J., and Mason, R., 1990, A re-appraisal of the geology and ore de-posits of the Horne mine complex at Rouyn-Noranda, Quebec: CanadianInstitute of Mining and Metallurgy, Special Volume 43, p. 153–166.

Kerrich, R., 1983, Geochemistry of gold deposits in the Abitibi greenstonebelt: Canadian Institute of Mining and Metallurgy, Special Volume 27, 75p.

Kerrich, R., and Cassidy, K.F., 1994, Temporal relationships of lode gold min-eralization to accretion, magmatism, metamorphism and deformation—Archean to present: A review: Ore Geology Reviews, v. 9, p. 263–310.

Kotila, B., W., 1975, An Archean porphyry copper deposit: The Don Rouynmine, Noranda, Québec: Kingston, Ontario, Unpublished B.Sc. thesis,Queen’s University, 21 p.

Lambert, I. B., Phillips, G.N., and Groves, D.L., 1984, Sulphur isotope com-positions and the genesis of Archean gold mineralization, Australia andZimbabwe: Geological Society of Zimbabwe, Special Publication 1, p.389–416.

Lange, I.M., Krouse, H.R., and Woodward, L.A., 1995, The Golden Mes-senger mine, York district, Montana: Geologic and isotopic constraints:ECONOMIC GEOLOGY, v. 90, p. 1322–1328.

Lemos, J.V., Hart, R.D., and Cundall, P.A., 1985, A generalised program formodeling jointed rock-mass—a keynote lecture, in Fundamentals of RockJoints: Sweden, Centek Luleå, p. 335–344.

Little, T.A., 1995, Brittle deformation adjacent to the Awatere strike-slip faultin New Zealand: Faulting patterns, scaling relationships, and displacementpartitioning: Geological Society of America Bulletin, v. 107, p. 1255–1271.

McCuaig, T.C., and Kerrich, R., 1994, P-T-t-deformation-fluid characteris-tics of lode gold deposits: Evidence from alteration systematics: GeologicalAssociation of Canada, Short Course Notes, v. 11, p. 339–379.

McMurchy, R.C., 1948, Powell-Rouyn mine, in Structural geology of Cana-dian ore deposits: Canadian Institute of Mining and Metallurgy, p.581–589.

Mikucki, E.J., and Ridley, J.R., 1993, The hydrothermal fluid of Archaeanlode-gold deposits at different metamorphic grades: Compositional con-straints from ore and wall rock alteration assemblages: Mineralium De-posita, v. 28, p. 469–481.

Morasse, S., Wasteneys, H.A., Cormier, M., Helmstaedt, H., and Mason, R.1995, A pre-2686 Ma intrusion-related gold deposit at the Kiena mine, Vald’Or, Québec. southern Abitibi subprovince: ECONOMIC GEOLOGY, v. 90, p.1310–1321.

Mortensen, J.K., 1993, U-Pb geochronology of the eastern Abitibi sub-province. Part 2: Noranda-Kirkland Lake area: Canadian Journal of EarthSciences, v. 30, p. 29–41.

Naylor, M., Mandl, G., and Sijpesteijn, C.H.K., 1986, Fault geometries inbasement-induced wrench faulting under different initial stress states:Journal of Structural Geology, v. 8, p. 737–752.

Ohmoto, H., and Rye, R.O., 1979, Isotopes of sulfur and carbon, in BarnesH.L. ed., Geochemistry of hydrothermal ore deposits, 2nd ed.: New York,NY, Wiley and Sons, p. 509–567.

Oliver, N.H.S., Valenta, R.K., and Wall, V.J., 1990, The effect of heteroge-neous stress and strain on metamorphic fluid flow, Mary Kathleen, Aus-tralia, and a model for large-scale fluid circulation: Journal of MetamorphicGeology, v. 8, p. 311–331.

Paradis, S., and Faure, S., 1994, Relationships between deformation andcupriferous mineralization in the external nappe domain of the Quebec Ap-palachians: ECONOMIC GEOLOGY, v. 89, p. 936–943.

Paradis S., Ludden, J., and Gélinas, L., 1988, Evidence for contrasting com-positional spectra in comagmatic intrusive and extrusive rocks of lateArchean Blake River Group, Abitibi, Québec: Canadian Journal of EarthSciences, v. 25, p. 134–144.

Pelletier, C., 1992, La faille Héré, mine Silidor: Unpublished B.Sc. thesis,Montréal, Québec, Université du Québec à Montréal, 22 p.

Picard, S., 1990, Le gisement Silidor: Canadian Institute of Mining and Met-allurgy, Special Volume 43, p. 175–184.

Powell, W.G., and Hodgson, C.J., 1992, Deformation of the Gowganda For-mation, Matachewan area, Ontario, by post-Early Proterozoic reactivationof the Archean Larder Lake-Cadillac break, with implications for gold ex-ploration: Canadian Journal of Earth Sciences, v. 29, p. 1580–1589.

Powell, W.G., Hodgson, C.J., Hanes, J.A., Carmichael, D.M., McBride, S.,and Farrar, E., 1995a, 40Ar/39Ar geochronological evidence for multiplepostmetamorphic hydrothermal events focused along faults in the southern

1064 CARRIER ET AL.

0361-0128/98/000/000-00 $6.00 1064

Abitibi greenstone belt: Canadian Journal of Earth Sciences, v. 32, p.768–786.

Powell, W.G., Carmichael, D.M. and Hodgson, C.J. 1995b, Conditions andtiming of metamorphism in the southern Abitibi greenstone belt, Quebec:Canadian Journal of Earth Sciences, v. 32, p. 787–805.

Richard, M., Hubert, C., Brown, A.C., and Sirois, R., 1990, The Pierre Beau-chemin gold mine: A structurally controlled deposit within a sub-horizon-tal layered composite granitoid: Canadian Institute of Mining and Metal-lurgy, Special Volume 43, p. 211–220.

Ridley, J., 1993, The relations between mean rock stress and fluid flow in thecrust: With reference to vein- and lode-style gold deposits: Ore GeologyReviews, v. 8, p. 23–37.

Riverin, G., Bernard, D., and Boily, B., 1990, The Donalda gold deposit,Rouyn-Noranda, Québec: Canadian Institute of Mining and Metallurgy,Special Volume 43, p. 199–210.

Robert, F., 1989, Structural setting and control of gold-quartz veins of the Vald’Or area, southeastern Abitibi subprovince: University of Western Aus-tralia, Geology Department (Key Centre) and University Extension, ShortCourse Notes, no. 24, p.167–209.

——1990, Internal structure of the Cadillac tectonic zone southeast of Vald’Or, Abitibi belt, Québec: Canadian Journal of Earth Sciences, v. 26, p.2661–2675.

——1991, Gold metallogeny of greenstone belts: Considerations from theeastern Abitibi subprovince, Canada: in Ladeira, E.A., ed., Brazil Gold ’91,Rotterdam, Balkema, p. 83–86.

——1994, Vein fields in gold districts: The example of Val d’Or, southeasternAbitibi subprovince, Quebec: Geological Survey of Canada, Current Re-search Paper 1994-1C, p. 295–302.

Robert, F., and Brown, A.C., 1984, Progressive alteration associated withgold-quartz-tourmaline veins at the Sigma mine, Abitibi Greenstone belt,Quebec: ECONOMIC GEOLOGY, v. 79, p. 393–399.

——1986a, Archean gold-bearing quartz veins at the Sigma mine, AbitibiGreenstone belt, Quebec. Part I. Geologic relations and formation of thevein system: ECONOMIC GEOLOGY, v. 81, p. 578–592.

——1986b, Archean gold-bearing quartz veins at the Sigma mine, AbitibiGreenstone belt, Quebec. Part II. Vein paragenesis and hydrothermal al-teration: ECONOMIC GEOLOGY, v. 81, p. 593–616.

Robert, F., and Kelly, W.C., 1987, Ore-forming fluids in Archean gold-bear-ing quartz veins at the Sigma mine, Abitibi Greenstone belt, Quebec,Canada: ECONOMIC GEOLOGY, v. 82, p. 1464–1482.

Setterfield, T.N., Hodder, R.W., Gilbson, H.L., and Watkins, J.J., 1995, TheMcDougall-Despina fault set, Noranda, Quebec: Evidence for fault-con-trolled volcanism and hydrothermal fluid flow: Exploration and MiningGeology, v. 4, p. 381–393.

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

Sylvester, A.G., 1988, Strike-slip faults: Geological Society of America Bul-letin, Centennial Volume, v. 100, p. 1666–1703.

Vearncombe, J.R., and Holyland, P.W., 1995. Rheology, stress mapping, andhydrothermal mineralization—the example of structurally-controlled gold-antimony deposits in South Africa: South African Journal of Geology, v. 98,p. 415–429.

Vearncombe, J.R., Barley, M.E., Eisenlohr, B.N., Groves, D.I., Houstoun,S.M., Skwarnecki, M.S., Grigson, M.W., and Partington, G.A., 1989, Struc-tural controls on mesothermal gold mineralization: Examples from theArchean terranes of Southern Africa and Western Australia: ECONOMICGEOLOGY MONOGRAPH 6, p. 124–134.

Velde, B., and Vasseur, G., 1992, Estimation of the diagenetic-smectite to il-lite transformation in time-temperature space: American Mineralogist, v.77, p. 967–976.

Verpaelst, P., Péloquin, S., Adam, E, Barnes, A.E., Ludden, J., Dion, D.J.,Hubert, C., Milhereit, B., and Labrie, M., 1995, Seismic reflection profilesacross the “ Mines Series ” in the Noranda camp of the Abitibi belt, east-ern Canada: Canadian Journal of Earth Sciences, v. 32, p. 167–176.

Vu, L., 1990, Geology of the Ferderber gold deposit and gold potentiel of theBourlamaque batholith, Belmoral Mines Ltd., Val d’Or, Quebec: CanadianInstitute of Mining and Metallurgy, Special Volume 43, p. 882–891.

Wallace, R.E., 1951, Geometry of shearing stress and relation to faulting:Journal of Geology, v. 59, p. 118–130.

Wilson, M.E., 1962, Rouyn-Beauchastel map area, Québec: Geological Sur-vey of Canada Memoir 315, 140 p.

Wong, L., Davis D.W., Krogh, T.E., and Robert, F., 1991, U-Pb zircon andrutile chronology of Archean greenstone formation and gold mineralizationin the Val d’Or region, Quebec: Earth and Planetary Sciences Letters, v.104, p. 325–336.

SILIDOR DEPOSIT, ROUYN-NORANDA DISTRICT, ABITIBI BELT 1065

0361-0128/98/000/000-00 $6.00 1065

0361-0128/98/000/000-00 $6.00 1066