Canadian MineralogistVol. 17, pp.287-298 (1979)

ABSTRACT

The R&na synorogenic Caledonide intrusion innorth Norway contains one known nickel depositof significaoce, Bruvann. The mineralization consistsof pyrrhotite*pentlandite+chalcopyrite*pyrite. oc-curing interstitially to olivine and orthopyroxene inperidotite and grading up to O.8Vo sulfide nickel.Locally, associated with certain deformation zones,interstitial sulfide dissemination passes into massivemobilized sulfide with ap to 5/o nickel. Reservesare 43 million metric tons with 0.337o sulfidenickel, 0.087o copper and approximately O.|lSVocobalt. Tho RAna intrusion consists of a peripheralzone of norite containing bands and lenses ofperidotite and pyroxenite, and a core mainly ofquartz norite. The R&na mass thus has a grosss8atigraphy that conforms to the pattern of manylayered intrusions, but, over much of the intrusion,primary structures have been disturbed by the laterCaledonian fold phases which also involved localoverthrusting; these movements resulted in an in-folding and thrusting of units of semipelitic andcalc-silicato gneiss and black schist into the in-tnrsion. The My has the form of an inverted,Dossibly truncated cone with its axis plunging north-westward at a moderate angle. There are problemsin placing the Rina intrusion within existing classi-fications of mafic intrusions. The peridotites showtro obvious systematic variation of sulfide or si.licatemineralogy across strike. The presence of sulfide-bearing black schists on or close to ttre contactsof the intrusion and emplaced within it alongshear zones and the occurrence of graphite withinsulfide disseminated in peridotite suggest assimila-tion of sulfur from the country rocks. Sulfurisotope studies do not, however, offer confirmationof the hypothesis that atr external source of sulfurhas had more than very local siguificance at R6na.

Sourvrenr

L'intrusion synorog6nique Cal6donienne de R6na(Norvdge septentrionale) ne contient qu'un seulgirement nickelifBre important, celui de Bruvann.Ics sulfures pyrrhotine + pentlandite 4 chalco-pyrite + pyrite, associ6s interstitiellement b l'olivineet lorthopyroxbne de la p€ridotite, contiennent jus-qu'd, 0,8Vo de nickel. Dans certaines zones de

THE NICKEL MINERALIZATION OF THE RANA MAFIC INTRUSION,NORDLAND, NORWAY

ROGNVALD BOYD AND CARL O. MATHIESEN

Norges Geologiske Unders/kelse, Postboks 3006, 7001 Trondheim, Norway

d6formation, les diss6minations interstitielles passenti des sulfures massifs remobilis6s dont la teneuratteint 5Vo Ni. Le gisement possbde des r6servesde 43 x 100 t de sulfures d 0.33/o Ni. 0.087o Ou ete O.0lSVo Co. L'intrusion comprend uno zone denorito avec des lits et des lentilles de p6ridotite etde pyrox6nite, autour d'un noyau de norite quart-zifdre. Stratieraphiquement, elle ressemble gtossomodo d nombre de complexes stratiformeg maisen beaucoup d'endroits, les structures primaires ont6t6 d6rang6es par les plissements et les chovauche-ments Cal6doniens. Des lambeaux de gueiss semi-p6litiques I calc-silicates et des schistes noirs prisdans ces plissements, ont 6t6 enfouis au sein deI'intrusion. Le massif se pr6sente sous forme d'unc6ne renvers6, peut-Otre tronqu6, dont I'axe s'enfon-ce l6gdrement vers le nord-ouest. Liintrusion doRAna n'est pas facile i classifier dans les sch6mascourants. ks p6ridotites De motrtrent aucune varia-tion min6ralogique 6vidente, ni en sulfures ni ensilicates, perpendiculairement au [tage. Irs schistesnoirs i sulfures observ6s au contact de I'intrusionou i I'int6rieur de celle-ci le long de zones decisaillement, ainsi que le graphite d6cel6 dans letdiss6minations de sulfures, semblent indiquer uneassimilation de soufre des roches encaissantes.D'aprds une 6tude des isotopes du soufre, pareilapport do soufre d'une sour@ externe n'a 6t6 gu'unph6nomdne loca.l insignifiant.

Cfraduit par la Rddaction)

INtt'ooucrroN

The RAna mafic intrusion lies at approxi-mately 68o30t{, 20 km southwest of the iron-ore port of Narvik in northern Norway (Fig. 1).This paper will give a brief description of thegeneral geology of the complex, with a moredetailed consideration of the sulfide-bearing areaat Bruvann (Fig. 2), the only known majordeposit in the intrusion, and of the genetic im-plications involved. The information presentedhere results from an investigation of the Rinamass and its nickel mineralization begun in1971, The investigation has been financed by asteel company, Stavanger Staal A/S, and by theNorwegian state, and carried out mainly bythe Geological Survey of Norway.

287

288 THE CANADIAN TVTINERALOGIST

genic belt, in two-mica gneisses classified as theNarvik Group (Gustavson 1972 afr. eadierpapers). These gneisses include thin units ofvariably hornfelsed calc-silicate gneiss up toseveral hundred m thick and commonly occur-ring on or slose to the outer contact of the in-trusion. These units consist mainly of plagro-clase * clinozoisite * hornblende { sphene Sdiopside f microcline { garnet -F quartz:! sar-bonates, and also contain bands of black schistrich in graphite and pyrrhotite.

The country rocks have been affected by atleast four phases of folding inchding nappeemplacement at an early stage. The RAna in-trusion appears to have been emplaced beforethe third phase of folding under garnet-am-phibolite facies conditions that prevailed through-out the solidification ,period of the mass. Tliisis indicated by the plesence of orthopyroxenetamphibole*amphibole/spinel symplectite coro-uas in olivine-plagioclase bearing rocks of theintrusion and by the assemblage garnet*horn-blende*plagioslase in rorcks within shear zonesin the norite. Roddick (L977) dated the in-trusion at 4OO m.y. on the basis of Rb/Sr wholerock and mineral isochrons; tle country rocksgive a metamorphic age of 4@'t-16 m.y.

Mafic and ultramafic rocks

The surface geology of the intrusion has acrudely concentric arran'gement, with a peri-pheral zone of norite and a core of quartznorite, both,locally amphibole-bearing and bothlocally gabbroic, particularly the quartz norite.The norite zone contains irregrlar bands andlenses of peridotite (the main host-rock for thenickel .mineralization at Bruvann) and pyroxe-nite. These rocks are not evenly disposed aroundthe peripheral zone but, where common, tend tooccur towards the outer margin of the mass. Atleast locally the norite can be shoqrn to intrudethe ultramafic rocks; whether this is due tosynmagmatic disturbance or to emplacementin separate magma pulses is not yet knorrn.South of the main intrusion are two outliers, oneon Tverrfjell east of RAndal, and the other onthe north side of Kvan6.kertind west of thevalley (Fig. 2). The Tverrfjell outlier is theonly part of the intrusion in which regular band-ing on scales below 1@ m is common; thesestructures, wh.ich include cross- and gradedbedding, are always 'right way up', with theexception of a few dubious cases. Elsevrhere inthe iutrusion banding other than the gross zona-tion already described is rare.

The northern and northwestern contacts ofthe intrusions generally dip outward at moderate

Fro. 1. Map of Norway showing the location ofthe RAna intrusion.

The intrusion has .been described by Foslie(1921), and the general geology of the sur-rounding area by Gustavson (1972),:, see aboreferences therein. The area lies within tle re-cently published 1 :25Q000 geological map-sheetNarvik (Gustavson L974). The mass containsthe only nickel deposits discovered in theNorwegian Caledonides with;n u radius of 100km, though the region does contain many sub-economic deposits of Cu, Fb, 7s, and Fe.

The Bruvann deposit is reetricted to an areaof approximately half a square km near thenorthwestern periphery of tle intrusion, whereit outcrops between 40O and 5@ m above sealevel (Figs. 2,3), A total of 7,00O m had beeudrilled in tlis area up to 1960. The presentinvestigation, in addition to geological and geo-physical matrrping, has entailed 28,500 m ofdrining.

GSNSRAL Gror.ooy oF THE fNtx,ustoNaND rrs BxvrnoNs

Country rock

The intrusion lies within the Caledonian oro-

E i;fl't?ifi[. $9[fi'n''a,'.,ffiffij ironrTt. GABBRoESS]igi ISIRONGLY OEFORIIED LOCATIY}

----- col{lA(l tAPPnox[{ArEt

Frc. 2. Geologic map of tle Rina intrusion.

to steep angles, whereas the southern and easterncontacts dip inward under the mass at anglesthat are shallorp to moderate in the east andsoutheast. The sontact between the norite peri-phery and the quartz-norite core is distinct onlywhere it is tectonic, in which case its dip isusually similar to that of the outer contact.Ifowever, in the eastern part of the intrusionthe nority'quartz-norite contact dips outward.The geometry of the contacts and .gravity data(Sindre & Boyd 1977) suggest that the intrusionhas the approximate form of an inverted, possi-bly truncated cone with an axis plunging north-westward and with exposure nearing the roofin the east. The distribution of rock types withinthis framework and the nature of the magoaticbanding, where not affected by later deforma-tion, suggest a gr6s stratigraphy similar to thatof many wElldesc,ribed layered intrusions em-placed in cratonic environments. Because of

289

the complex and often contrasting nature ofthe Rina strustures and because of publishedreseryations on the classical cumulus theory(Ca;mpbell 1978), judgnent is reserved as tothe exact processes iovolved in producing theinternal variations of the RAna body.

The most @tnmon variety of peridotite foandin the intrusion is harzburgite with 4V6OToolivine (Foee-ez based on nine microprobe anal-yses of samples from, Bruvann), though lher-zofitic and dunitis varieties also ocour; locallnsmall a,rnounts of intentitial plagioclase may bepresent. trn the usage of Wager et oJ. (1960),the bulk of the peridotite is olivine ortho- ormesocumulate, though olivine-hypersthene or-thocumulate and locally olivine adcumulate arepresent. Poikilitic intercumulus ortrhopyroxene,less commonly augite and phlogopite, and rarelyplagioclase are present. The remaining primaryphases, in addition to the sulfides described later,

MCKEL IN TEE NAUE *TETTC INTRUSION

llllll Hffir';P*ii.i*t*,^..,, llffi crrrss : cAcL(srr'|(Art

lii]l]l eeorerrrt lT::-l AilPrrBofrrr FIIII neo scntsr

A FAULT

290 TIIB CANADIAN MINERALOC'ISIT

Ittg t F

Fro. 3. Geological map of the northwestern part

are picotite spinel, hernatite, ilmenite and mag-netite, all in small amounts. Where olivine hasoriginally abutted against plagioclase, coro{tasare developed as mentioned earlier. Olivineshows a greater tendency to idiomorphismagainst sulfides than against silicates; beingofteu rounded and internally corroded in thelatter case. Talc and anthophyllite are commonin deformation zones.

T\e pyroxenite is generally an ortho- or meso-cumulate with up to about 80% hypersthene;adcu'nulate textures occur locally. The inter-sumulus material consists of augite, phlogopite,plagioclase and ocsasionally, opaq.ue minerals.Olivine-bearing varieties occur, but usually as atransition to peridotite. The hypersthene showsslight marginal zoning and may contain augiteexsolution lamellae.

Norite is used here to cover a wide rangeof rock Epes, including limited amounts ofboth primary olivine-bearing and gabbroicvarieties. Alteration, deformation and recrys-tallization impose further complexities. Therock is, where fres\ generally a hypersthene-

of the R&na intrusion.

plagioclase heteradcumulate (adcumulus plagro-clase), commonly rrith strong larnination of theoumulus laths of orthopyroxene and plagioolase.The hypersthene may be zoned, and only rarelycontains exsolved lamellae of clinopyroxene.Hypersthene and plagioclase may contain inclu-sions, one of the other. Clinopyroxene, wherepresent, has a mode of occurrence sirnilar tothat of hypersthene. Primary amphibole is afeature of some of tle norites in certain areas.Secondary amphibole and clinozoisite are com-mon, and, where deformed, the rock is generallyrecrystallized to amphibolesfplagioclase*clino-zoisite*mica. Where olivine-bearing, the noritemay contain ss6s disseminated nickeliferous sul-fide. Olivine-free norite may contain some sul-fide, but essentially nickel-free. Locally non-nickeliferous sulfide @curs together with gra-phite as small blebs or streaks in deformednorite.

T1ne quartz-norite core of the intrusion ap-pears to be less well-defined than suggested byFoslie (1921), and to be more variable inter-nally. Plagioclase is more abundant than in

NICKEL IN TIIB NAT*E TVTETTC INTRUSION 291

the norite, and is commonly idiomorphic againstquartz, present in only small amounts. Clino-pyroxene increases relative to orthopyroxene, andprimary a,mphibole and biotite are present inaddition to rare orthoclase. Superimpoced onthese variations are the effects of deformationand recrystallization, especially marked in thearea of Eiterdal.

Srnucruner, Gsolocy

Though the RA,na intrusion may postdate tlestongest phases of regional deformation, theeffects of the later phases have locally beenvery pronounced. This applies particularly to thearea north of a line from Bruvann to Saltvik-vann and to a belt from Kvan&kertind to Eiterdal(Fig. 2). Deformation in these areas, corrcsponding to the regional F movements, probablybegan with the development of segregation band-ing in norite/.metanorite, with subsequent foldingof the bands. Folding of the contact southeast

of Bruvann and south of Sepmolfjell and thedevelopment of the Tverrfjell synform seem todate from this period. The climax of this phaseof defor'mation involved disruption of the massalong two major zones: (1) The overthrustingof the part of the intrusion known as theArneshesten block northwest of a line from Bru-vann to R0,na (Fig. 3) torrards the southeaslThe thrust is defined by a strongly deformedarcuate zone that dips steeply northwest andbecomes less steep at deeper levels. The Arnes-hesten block has a complex internal structuredue to folding and faulting, some of which isrelated to the overthrusting. In addition todisrupting the primary structures of the block,these movements resutrted in the emplacementwithin the block of lenses and slabs of countryrock including black schist (Figs. 3, 4). Relativeto most other parts of .the norite periphery, thisarea is distinguished by its higb proportion ofultramafic rock to norite and by the presenceof large volumes of peridotite with nickel-bear-

L_9.c9!g

@ ourin" beoring mck3, mointt p.ridotit€

ffil Py.or"nttr

ffiffi lorita

mrya .c|s|r,llll srophifo/sulride beorinq

! ni"tg-r of slfids nirket 7o t4.0mm '1%l

--- lnhrprrtqtin gaotogicqt boundory

L Driuiol.

g l l t t :S

coordinatrs in natnrs

Fra. 4. Geological interpretation of profile 2600 E in the Bruvann area; tle coordinates refer to thosystem also used in Figures 2 & 3.

292 THE CANADIAN MINERALq}IST

ing sulfide dissemination. This is not surprisingin an upthrust block, if the model of the intru-sion suggested earlier is correct. (2) The south-ern part of the intrusion was affected by theformation of synclinal structures both east andwest of RAndal, those at Tverrfjell and north ofKvan0,kertind becoming separated from the restof the intrusion by deJormation zones contain-ing bands of country rock.

The Fa trfiase of folding seen in the sur-rounding gneisses seems to correspond to north-south faults common in well-exposed areas onthe south side of the mass, and also on Sep-molfjell and in Arneshesten block. Subhori-zontal pegmatites emplaced late in the historyof the mass occur in many parts of the intrusion.

Ons Gsor.ocy

The Bruvann area

The deposit contains a calculated 43.6 millioumetric tons of sulfide-bearing peridotite aver-aging O.337o sulfide nickel, O.O8Vo copper andca. O.O1,SVo cobalt (cutoff was taken as O.LSVosulfide nickel). Restricted parts of ,this tonnageaverage as much as O.6Vo, and individual anal-yses reach O.8Vo. \\ese figures are based onanalysis of the sulfide content (bromine soluble)of 2 m core sections. The deposit has approxi-mate dimensions of 9@ m east-west and 500 mnorth-south. In the middle of the area it is cutby a northeast-southwest hinge fault (down-throw to the west), the tlrow of which in-creases northeastward from about.200 m closeto Bruvann.

East of this fault the peridotite is underlainby pyroxenite which in turn lies on norite. Thebase of the peridotite is flat lying in north-south section and rises to the east. In the norththe peridotite stops abruptly along a steep east-west contact against a ridge of norite, and it iscyt off in the east by faulting with upthrow tothe east. Thus, the norite underlying the Bru-vann peridotite is thought to be equivalent to thelargely noritic rocks on Arneshesten. The mainmineralized zone, which lies in peridotite, strikeseast-west parallel to the rock contasts. Wheredeformation is leas! as in profile 26@E (Fig.4), it has a dip of about 45o near the surface,flattening out in depth towards the sourth. Abovethis main zone occurs a series of thinner. lesscontinuous zones of mineralization. pvroxeniteinterbanded with mineralized peridotite com-monly contains interstitial sulfide of lower gradethan the average for the minera:lized peridbtite.Spane mineralization also ocsurs locallv in oli-

vine norite. A simplified stratigraphy for theeastern area is, from top to bottom: 5O-1@ msf nn,minslalized peridotite with thin, erraticmineralized zones and pyroxenite bands; 2G-50m of mineralized peridotite (the main zone)qriffo thin pyroxenite bands; O-2@ m of unmine-ralized peridoti.te, G-70 m of pyroxenite and150 m of norite (to explored depth).

Much of the area is complicated by deforma-tion zones and by bodies of other rock typesincluding gneiss, whose geometric relationshipsare often not amenable to interpretation (Fig.4). The soutlern part of the area is more com-plex; mineralized horizons weaken and disappearand the peridoti,te units finger out and pass intogenerally noritic rocks of rapidly and irregutrar-Iy varying character.

Close to the base of the mineralized perido-tite in this area occur acoumulations of mas$iveepigenetic sulfide with up to SVo nickel. Thistype of mineralization, first found at Bruvannr\ 1974, is now known from several localitiesalong an east-west line (125O N) close to thesouthern contact of the deposit. These inter-sections usually have in colnmon a proxirnity toperidotite, or sther olivine-bearing rocks withdisseminated sulfide (Fig. 4), and location inzones of deforma.tion.

West of the fault that dissects the Bruv,annarea (Fig. 3) the situation is more com,trrlex.This part of the deposit has a gTeater tonnagethan the eastern area. It is nowhere exposed andwas not discovered !\ttl 1972. The western areacontains two major mineralized peridotite unitsthat lie rather fla,t in the south and genemlly in-crease in dip to the northwest. The present strati-graphy, partly the result of complex deforma-tion, is, in simplified form, from top to bottom:50-250 m of complex, irregularly interbandednorite, peridotite and pyroxenite; 5O-10O rn ofupper mineralized peridotite with thin pyroxe-nite bands; 40-100 m of interbanded peridoti,teand pyroxenite with sub'sidiary norite; 20-100m of lower mineralized peridotite; 6O-120 mof unmineralized peridotite; 0*50 m of pyroxe-nite, and norite (to depth drilled, approximately400 m below sea level).

The western area has several features in com-mon witl the eastern area. fn both cases theperidotites terminate abrupdy to the northagainst a ridge of norite (under coun'try rockin the western area). In both cases the dissemi-nation tends to be richer and more homogeneousin tle north and weaker or absent in the south.The major minsl6li2sd peridotite in the easternpart of the Bruvann area and the lower mine-ralized peridotite in the west were probably

MCKBL rN THE niNe rvrerrc rNTRUsroN 293

originaily continuous; the upper peridotite inthe west may be a tectonic repetition of thelower one.

Typical sulfide dissemination in the peridg-titesit Bruvann contains pyrrhotite (SV7OVo),pentlandite (10-35Vo), chalcopyrite (5-l5Vo\and. minute amounts of pyrite, 1O5Vo accordingto a detailed point-count analysis by Malvik(t977). The sulfide minslalegy of olivine noriteand pyroxenite in the area is similar. The mine-ralization occurs discontinuously and intersti-tially to olivine and hypersthene but, in richerportions, has tle 'net' texture described byNaldrett (1973). Pentlandite ocsurs mainly asdiscrste, irregular grains close to or at themargins of pyrrhotite and in cracks tlroughpyrrhotite grains. 'Flanes' and lamellae ofpentlandite are present in the pyrrhotite nearfractures and associated with grain boundaries,but only in minor amounts, of the order of lToof the total pentlandite (Malvik t977). Chalco-pyrite occurs mainly as free grains marginal topyrrhotite, but also as crack fi[ings cuttingacross other sulfides and locally penetra'tingadjacent silicates. Pyrite forms isolated idio-morphic grains in pynhotite. Small amounts ofmagnetite ocsur, coulmonly near the marginsof olivine grains; intergtown hematite and il-menite occur interstitially. In cornmon withRamdohr (1969), the writers have found minoramounts of graphite within sulfide disseminatedin peridotite.

As noted eatlier, accumulations of massiveore occur locally along an east-west-trendingzone 5O-100 m north of the southern borderof the Bruvann area, and their precise locationis apparently to some extent testonicatlV c9n-trolled. In most cases, this mineralization liesbeneath interstitial sulfide in peridotite andabove a zone of rock with variable amounts ofsulfide and fragments of altered, foliated maficrock. In certain drillhols the massive sulfidemay be several rn thick; in others it has beeninjeoted as veins from a ferr cm to several dmacross. The massive ore seems to have ac-cumulated subsequent to the solidification ofthe peridotite and along a 7'ane of disruptionbetwlen the more homogeneous thicknesses ofperidotite to the north and the more variablerock types to the south. The presence of thesulfide-rich sheared rock beneath the massivemineralization may reflect the accumulation ofthe massive snrlfide in dilations within or ir-regularities on a deformation zone. Note thatthis nni"eraliza,tion has been intersected at threelocalities only and that these differ to an extentthat makes generalization difficult.

Other qeas of mineralization

Other areas of the intrusion contain rnine-ralization of lesser significance. In Rinbogen(Figs. 2, 3), sulfides are present in smalleramounts as blebs and as interstitial dissemina-tion in peridotite, as interstitial disseminationin norite and as massive sulfide wi,th graphitein dilations along shear zones. Common to allthese mineralizations are low Ni:S ratios.

The Eiterdal deposit (Fig. 2) contains mine'ralization in olivine norite along the underlyingsheared contact of the intrusion, here in con-tact with calc-silicate rock containing blackschist. The nsrils ssafa'ins lenses of countryrock and several types of mineralization includ-ing 'blebt sulfide and interstitial mineralizationtransitional into more massive sulfide.

ANer,Yrrcer, oere'

Composition of sulfid.e minerals

Preliminary microprobe determinations sug-gest that the pentlandites have compositionsfa[ing within the range characteristic for theassemlhge pyrrhotite*pentlandite+chalcopy-ritelpyrite (Harris & Nickel 1972, Graterol &Naldrett L97I). Pentlandite'flarnes' in the pyr'rhotite contain several percent less nickel thandiscrete grains and also considerably less cobalt.

The microprobe was also used to search forvariations in olivine or sulfide compositionsrelative to olivine-sulfide grain boundaries; nosuch variations were found. This and the tex'tural evidence (a greater tendency for idio-morphism in olivine against sulfide than Sgatlstothei silicates) suggest that direct sulfurizationof olivine (Kullerud & Yoder 1965, Naldrett1966, 1969) has not been an important processin the RA.na peridotites.

The Ni:S ratio for disseminations in perido-tite at Bruvann varies from 1:2 downward butis olustered between l:2 and 1:5-6, with amedian at approximately 1:4. This is typical ofthe values for nickel sulfide deposits in perido-tites given by Wilson & Anderson (1959). Themassii" sulfide has a Ni:S ratio of less than1:5. The Cu/(Cu+Ni) ratio at Bruvann, O.2,lies within the range found in deposits in ultra-ma,fic rocls (Naldrett & Cabri 1976, Rajamani &Naldrett 1978). The majority of depoaits forwhich s,uch metal ratios are cited as charactet-istic occur in host complexes that as a wholeare much more mafic than the Rtna mass(e.g., deposits of the Abitibi and Manitoba belts) ''

bitt"itt metal:sulfur ratios seelns to be

294 THB CANADIAN MINERALOGIST

sharacteristic of different parts of the intrusion.At Bruvann, where the dominant mineralizationis in_peridotite, the Ni:S ratio is 1:4; in Eiterdal,in olivine norite it is 1:10 and in RA.nbogen,c&ere the mineralization is predominantly innorite, the ratio is l:14 or less. A similar corre-1,"9o:.*E tock type has been shown by Hiikli(197L). This trend sugg$ts that the'sulfide9agma occurring in various parts of the intru-sion attained- a degree of local homogeneitnthe composition being dependent at liast inpart on the nature of the silicate magma present.

Sulfur isotope study

- The presence of small amounts of graphitein various rocks in the intrusion, the juitaposi-tion of sulfide-rich black schist with mineraiizedperidotite and massive sulfide, and variable Ni:Sratios led the writers to suspect an externalsource for some of the sulfur in the deposits.Graphite has been described from otheruufia"-bearing mafic intrusive 6odies: the Bushveldco-nplel _(Uebenberg L970), Kotalahti (Haa-pala 1969), Hitura (papunen 1970), theHarriman and Warren Aeposits in Maine'(Ra,in-ville & ,Park 1976) and from the Watei Henlntrusion in the Duluth complex (Mainwaring &Naldrett 1977). fn a number of these cie"country-rock xenoliths are prominent. Sulfuri19t9ne studies based on the-difference in sS/sS between sulfur in blask schists and othermetasedimentary rocks aud sulfur from normaluncontaminated mantle-derived magma havebeen used to indicate a partly exteria{ ,orrr""

of sulfur for the sulfides in certain mafic in-!Tyy" complexes (Godlevskii & Grinenko 1963,Jlafdrett 1966, Liebenberg 1969, Mainwaring &Naldrett 1977).

Table I shorps the results of sulfur isotopeanalyses, perforrned on R&na samples. On theassumption tlat investigations of the origin ofsulfur in ore deposits should be based on thesulfur- isotope ratio for the total sulfide content,not that of the individual minerals (Rye &Ohmoto 1974), the sulfide minerals weri notseparated individually. The samples consistedo! -mole than 95% pure bulk suftiae fractious,with the- rnineralogy for individual rock typesas described earlier. The writers have not foundpublished sulfur isotope data for separatedpentlandite, but data on separated mineral frac-tions of coexist:ng pyrite, pyrrhotite and chal_llPYnte .(MltceH 1974, Rye & Ohmoto 1974,Mainwaring & Naldrett L977, and others) sug-gest that the degree of isotope fractionationexpected ig this paragenesis and at high temper-atures of deposi,tion is very limited. In addition,the changes in sulfur isotope ratio so far de-monstrated as due to wal.lrock contamination_(Liebenberg 1.969, Godlevskii & Grinenko 1963,Mainwaring & Naldrett 1977) are almost anorder of magnitude greater than those whichwould be expeoted from isotope fractionation..-The results for samples from R6na peridotites(Table 1) resem-ble those cited bv Stanton(L972) for a number of deposits in mafic andultramafic rocks and interpreted as indicatinguncontaminated magma from a deep-seateds_ource.. This applies equally to the samples ofdissemination in norite; the samples from-pyrox-enite and massive sulfide contain ,trgnUyheavier sulfur but the difference is small anddoes not seem related to variations in the pro-portions of sulfide miuerals.

The sulfur in the bands of black schist, bothwithin and outside the intrusion, is on the aver-lge markedly lighter than meteoritic sulfur, audis clearly lighter than the values obtained fo,disseminated sulfide in peridotite. The valuesfall within the range cited by Stanron (Lg7Z)for ancient sedimentary rockJ with no volcaniccomponent. These resul,ts suggest that externalsulfur was not significant in the formation ofmineralization in the peridotite at Bruvann,or in peridotites elsewhere in the intrusion. Onlvvery locally (and only in samples in which gra-phite is clearly visible megasCopically) has-ex-ternal sulfur had an obvious influence. A corol-lary of this conclusion is that some of the gra-phite (e.9., the flakes of graphite found ininterstitial sulfide in peridotite) must clearly

IABU .I. SIJI.II.IARY oF SULFUR IsoToPE DATA

f inera l i za t ion type A v e .

o34s Range o f 6345l , lo . o fsamples

P e r l d o t i t e d l s s e m i n s t l o n

Pyroxen l te d lsseminat ion

N o r ' l t e d l s s e n l n a t l o n

f l o r l t e d . l s s e m l n a t l o n r l t hg r a p h i t e

U a s s i v e s u l f i d e

U a s s l v € s u l f i d € r l t hgraph l te

S u l f l d e f r o m b l a c l s c h l s tr l t h l n t h e l n t f u s i o n

S u l f i d a f r o m b l a c k s c h l s to u t s l d e t h e . l n t r u s l o n

+ 0 . 3 - 2 . 5 t o + ' 1 . 6

r 2 . 7

r 0 . 7 - 0 . 1 t o + 1 . 6

- 5 . 9 - 3 . 7 t o - 7 . 6

! 2 . 1 r 0 . 4 t o + 4 . 2

- 1 2 . 9 - 3 . 2 t o - 1 7 . 6

- 7 , 4 - 3 . 6 t o - 1 3 . 6

- 9 . 0 + 1 . 0 t o - 1 3 . 5

l l

I

6

l 0

7

The ana lyses r ,e re per fo rmed by Geochron Labora tor les

l n c . , B o s t o n . 6 3 4 s i . r u a . r r " d r e l a t l v e t o 3 4 5 7 3 2 ,

f o r C a f t o n D l a b l o t r o t t . l t e .

NTcKEL IN THB nANe rvrerrc rNTRUsroN

be of an early or primary magmatic nature,whereas locally it has probably been derivedfrom the country rock with some sulfur.

DrscussroN

A slassification of ultramafic and associatedmafic rocks was proposed by Naldrett & Gas-parrini (L97I) and later modified (Naldrett1973, Naldrett & Cabri L976). All three papenemphasize the implications of the classificationfor metal deposits (especially of nickel) asso.ciated wi,th these rocks. With more generalgeological considerations in view, ctrassificationsof mafic and ultramafic rocks have been pub-lished by Jackson & Thayer (1972), Thayer &Jackson (L972), Moores (1973) and Wyllie(1969). The RA,na intrusion and a number ofothers in the Caledonides, including those ofthe Seiland province (Robins & Gardner 1974),Ireland (Leake 1970, Kanaris-Sotiriou & AngusL976) and northeast Scotland (Ashcroft &Munro 1978, Boyd & Munro L978), do notfall easily into any of the published classifica-tions. This problem and many of ,the difficultiesin interpreting the Rlna mass and its mineraliza-tion arise from the fact that, though it hasmany of the features of the stratiform or funnel-shaped intrusions that occur in stable areas, itlies in an orogenic belt and also, at least locally,has many of the features of the intrusions classi-fied as alpine (Jackson & Thayer L972), e,9.,tectonite fabrics, metamorphirc and igneouslayers, discordant and irregular banding, poorc.ontinuity of banding and dykes of mafic rock(at RAna, norite intruding earlier formed rocktypes). In this context the use of the termsstratiform and concentric, descriptive of theform of the intrusion, as opposed to alpine,descriptive of present geological environment(Jaclson & Thayer 1,972), seems unfortunateas is the tendency to equate alpine ultramalicrocks with ophiolite (Naldrett & Cabrj 1976,Garson & Plant 1973). Some related reservationson existing classifications of mafic and trltra-mafic rocks have been expressed by Chaltis(1969), Challis & Lauder (1966), Nesbitt er a/.(1970), Moore (L973), Robins & Gardner(1974) and others.

The form of the contacts of the R&na in-trusion suggests ,the shape of an inverted cone,possibly truncated, with its axis plunging north-westward and with the highest section of thecone, possibly near the roof of the intnrsion,in the east and the deepest section in the north-west. In a very general sense the distribution of

the peridotite bodies lends support to thismodel: they are not common on the easternrim but are mostly found'along the northernand northwestern contact. Further confirmationof this model is given by the nature of theArneshesten block which, having been upthrustrepresents the deepest level exposed in the massand contains not only peridotites of substantialthickness, but also the only ones in the intrusionknown to be significantly mineralized over largevolumes. The individual peridotites, however,may be subparallel to,the outer contact (easternR&nbogen, TVerrfjell) (Fig. 2) or discordant toit (western RAnbogen, Bruvann) (Figs. 3, 4)'some bodies apparently changrng along theirtengfh fro,rn concordant to discordant. Thewriters have failed to construct a model ac-counting for all these complexities; however,these must be related to the extensive syn- andpostintrusive deformation of the area.

The nickeliferous sulfides at Bruvann (andelsewhere at R&na) are invariably within orclosely associated with olivine-bearing rocks.However, the mineralization does not occursystematically at or near the u,pper or lowermargins of the peridotites. In addition, thernassive nickel-bearing mineralizatron is gener-ally close to 'nef texture sulfide, bu,t it consti-tutis only a small fraction of the total minerali-zation. Thus it seems likely that the magma

' from which the peridotites began crystallizingwas not saturated in sulfur, that sulfur satura-tron was attained while olivine and locally asmaller a.mount of hypersthene were accumulat-ing, and that the sul'fide subsequently accumu-lated as part of the material interstitial to thecumulus phases. The further obtervation thatolivine sho'ws a tendency to idiomorphism whereit abuts against interstitial sulfide while beingmore rounded or even internally resorbed againstinterstitial silicate suggests that olivine accumu-lated rapidly (possibly having nucleated close tothe floor of the magma chamber) and wasresorbed to form orthopyroxene, mainly afteraccumulation, through reaction with intercumu'lus liquid (the Bowen-Anderson reaction).

the direct cause of the attainment of sulfursaturation is not clear. The isotope data tendto discount any 'major role 'for country-rocksulfur. Irvine (1975, 1977) has suggested theexistence of a mechanism that may explainseveral features of the Bruvann mineralization:an increase in silisa or alkalis in 'magmas ofcertain compositions (either by assimilation orby magpamixing processes) reduces -sulfurs6tuUitity (resulting in formation of an immis-cible suifide magma) and causes a polymeriza-

295

296 THE CANADIAN MINEMLOCIST

tion of the silicate magma (which would prod-uce orthopyroxene instead of olivine). Ions suchas Ni2+ with high octahedral-site-preferenceenergies are then preferentially expelled fromthe silicate liquid and partition into the sulfidemagma. Some evidence for multiple infiluxesof magma at RAna may lie in the reclurenceof thick peridotite layers and in the iatrusiverelationship, at least localln of norite to theultramafic roeks. Late- and postmagmatic de-formation complicates the reconstruction ofevents. That the bulk of the mineralization atBruvann lies in peridotite and not in pyroxenitecould be taken as evidense against the theorybut, on the other hand, once sulfur saturationis attaiued, it can perhaps be assumed that thesulfide ,magma would accumulate more rapidlythan orthopyroxene; the mineralized periditi.re,particularly in its upper reaohes, is interbandedwith pyroxenite which is also mineralized tosome extent.

The Bowen-Anderson reaction probablycaused partial resorbtion of cu,mulus olivine toproduce intercumulus orthopyroxene. That this13ctio1 may take place between cumulus crys-tals and intercumulus liquid has been advocaiedin t}re case of the Stillwater complex by Jackson(1961) and in the case of rocki of the Seilandprovince by Gardner & Robins (1974). Thepartition coefficienr Ol"'/Hf, (HekU 1971) issuch that tlis process, though affecting only aportion of the original amount of olivine crvs-t4ljzed, would recycle substantial quantities ofnickel to the silicate magma because it woildaffect a large volume of olivine. Where sulf,iderya'gma rilas present interstitially, it would absorbthis nickel preferentially with respect to crystal-Iizing orthopyroxene. The effect bf this pro,cesswould be enhanced if: (1) there was expulsionof intercumulus liquid from olivine cuirulatesformed before sulfur saturation, and (2) inter-cumulus sulfide magma had, because of itsgxeater density, intimate contact witl a greatervolume of olivine than that against wnicn itultimately solidified, i.e., if sulfide masma Der-colated downward through a pile of

-oumirlus

olivine crystals displacing intercumulus silicatemagma. These processes may have contributedto the relatively high nickel content of theR6na sulfides.

A number of factors suggest tha,t part of thesulfur in the R0,na mineral deposit was derivedfrom black schist lenses wirhin and around tleintrusion. The sul.f,ur isotope data seem to indi-cate tlat this process may have been active, butonly_locally and apparently not at a stage atwhich the major mineralization at Bruvann-could

be affected. This tends to suggest tbat the tec-tonic juxtaposition of black schist and mineral-ized rock seen in the Bruvann area occurrd toola.te to have had any influence, even on themassive sulfide. Conversely, to have had morethan local inflsejss, sush contamination pro-cesses would seem to require the incorpomtionof country rock into the magma. very early inits crystallization history.

ACKNowl,BDGEMENTs

The writers thank Stavanger Staal A/S forpermission to publish the Rana material,BArd Tltdal for assistance with microprobcana;lyses, Astrid Hemrming, Lars Holipkk andTore Skauge for help in preparing the figuresand Magne Gustavson, Gunnar Juve, IngvarLindahl, David Roberts, John Thom,pson andFrank Vokes for their constructive critioism ofthe text. In the cou$e of the exploration workat Rflna many people at the Geological Surveyof Norway have been involved; their help isgatefully acknowledged.

RsreRENcEs

Asncnorr, W.A. & Murno, M. (197E): Thestructure of the eastern part of the Insch maficintrusion, Aberdeenshire. Scot. l. Geol, t4, 55-79,

Bovo, R. & MuNno, M. (1978): Deformation ofthe Belhelvie mass, Aberdeenshire. Jcot. l. Geol.L4, 2944.

CeMrnEr.L, IJI. (1978): Some problems with thecumulus theory. Lithos L1., 3Il-323.

Cnenrs, G-4,. (1969): Discussion on the paper"The origin of ultramafic and ultrabasic rocls"by P.J. Wyllie. Tectonophys. 7, 495-505.

& Leuonn, W.R. (1966): The genetic posi-tion of "Alpine" tygn ultramafic rocks, Bull.Volc.29, 283-306.

Fosr.rg S. (f921): Field observations in NorthernNorway bearing on magmatic differentiation. J.Geol. 29, 7Ol-719.

GeroNrn, P.M. & Rornrs, B, (t974): The olivine-plagroclase reaction: geological evidence fromthe Seiland petrographic province, nortlernNorway. Contr. Mineral, Petrology 44, 149-156.

Gensor.r, M.S. & PL$tr. I. (1973): Alpine typeultramafio rocks and episodic mountain buildingin the Scottish Hiehlands. Nature Phys. Sct.242,3+38.

Gorr.Bvsrn, M.N. & Gnnnxro, L.N. (1963):Some data on the isotopic compositioa of sulfurin the sulfides of the Norilrsk deposit. Geochem.,354r.

NICKEL IN TIIE N.&NE TVTETTC INTRUSION 297

Gnernnor-, M. & Nnr-oxetr, A.J. (1971): Mine-ralogy of the Marbridge No. 3 and No. 4 nicke!-iron sulfide deposits. Econ. Geol- 66, 866'900.

GusrevsoN, M, (L972): The Caledonian mountainchain of the southern Troms and Ofoten areas.III. Structures and structural history. Norg. GeoI.Unders. 288.

(L974): Geologisk kart over Norge, berg-grunnsgeoloeisk kart - Narvik' 1:250 000. Norg.Geol. Unders.

llalpar.l, P.S. (1969): Fennoscandian nickel de-posits. Iz Magmatic Ore Deposits (II.D.B. Wit'son, ed.). Econ. Geol. Mon. 4,262-275,

Hiiru, T.A. (1971): Silicate niskel and its appli'cation to tle exploration of nickel ores. Geol.Soc. Finland Bull. 48' 247'263.

Ilmrug D.C. & NtcxEL, EJI. (1972): Pentlanditecompositions and associations in some mineraldeposits. Can, Mineral. 11, 861'878.

Invnw, T.N. (1975): Crystallization sequences iothe Muskox intrusion and other layered intrusions.II. Origin of chromitite layers and similar de'posits of other magmatic orq. Geochim. Cos-mochim. Acta 39, 991'1020.

(t977): Origin of chromitite layers in theMuskox intrusion and other stratiform intrusions:a new interpretatiot Geol. 6, 273'277.

fAcKsoN, E.D. (1961): Primary textures and mine'ral associations in the tjltramafic Zone of tleStillwater complex, Montana. US. Geol. Sum.Prof. Pap.858.

& TtrevsR, T.P. (1972): Some criteria fordisting;uishing between stratiform, concentric andabinJ peridotite-gabbro complexes' Int' Geol'Congr. 24th, 2, 289-296.

Klxanrs.sorrnrou, R. & ANcus, N.S. (1976):The Currywongaun-Doughruagh syntectonic in-trusion, Connemara, Ireland, t. Geol. Soc, Inn'don L82,485-508.

Kulunuo, G. &YopsR, H.S., Jn. (1965): Sulfide-silicate reactions and their bearing on ore forma-don under magmatic, postmagnatic and meta-morphic conditions. Iz Symposium on Problemsof Postmagrnatic Ore Deltosition n', Geol, Surv.Czechoslov akia, P rague.

LsAKE, B.E. (1970): The ftaguentation of tleConnemara basic and ultrabasic intrusions. InMechanism of lpeous Intrusion (G. Newall &N. Rast. eds.), Geol. l. Spec, Issue 2, I0l'L22.

LrrssNBERg L. (1969): The Sulphides in theInyered Sequence of the Bushveld lgneoas Com'plex. Ph. D. thesis, Univ. Pretoria, Pretoria'South Africa.

(1970): The sulphides in the layered se-quence of the Bushveld igneous complex. la Sym'posium on the Bushveld Igneous Complex andother Layered Intrusions (D.JI. Visser & G. vonGruenewaldt, e&.), Geol. Soc. 5. Atr. Spec.Publ. l, LO8-207.

MAr.rwARrNG, P.R. & Nerunrrr, AJ. (t977)'.Country-rock assimilation and the genesis of Cu-Ni sulfides in the Water Hen intrusion, Duluthcomplex, Minnesota. Econ. Geol.72, 126%L2E4-

wtlirsilt M. (1974)t A study of zulfur isotopes mthe Outokumpu ore deposits, Finland. Geol. Surv.Finland Bull.267.

MALvtK, T. (1977): Mineralogisk underslkelse avbulkkonsentrat av RA,na Ni-malm. Unpubl. Rep.,Norges Tekniske Hdgskole, Trondheim.

Moone, A.c. (1973): Studies of igneous and tec'tonic textures and layering in the rocks of theGosse Pile intrusio& Central Australia. J. Pe-trology 14 49-E0.

MooREs, E.M. (1973): Geotectonic significanceof ultramafic roc,ks. Earth'Sci. Rev. 9, U1'255.

NALDRETT, AJ. (1966): The role of sulturizationin the genesis of iron-nickel sulfide depositsof the Porcupine Distric! Ontario. Can. Inst.Mining Met. Trans.69, L47-155.

(1969): Discussion of papers concernedwith sulfide delpsits. In MagmaJ;ic Ore Deposits(H.D.B.'lVilson, ed.), Econ. Geol. Mon. 4,359'365.

(1973)z Nickel sulfide deposits - -theirclassifiiation and genesis, with special emphasison deposits of volcanic association. Can. Iwt'Mining Met. Traw. 76, I83'2nI.

& CesRr, L.J. (1976): Ijltramafic and re-lated rocks: their classification and genesis withspecial reference to the concentration of nickelsuffides and platinum-group elements. Econ. GeoI'71 ,1131 -1158 .

& GlsrannrNr, E.L. (1971): Archean nickelsulfide deposits in Canada: their classification,gpological setting and genesis witl some iuqces'tions as to exploration' Iz Symposium on ArdleanRocks (J.E. Glover, ed,), Geol. Soc. Aust. Spec'Publ. 3, 201-226.

Nesrrrr, R.W., Gooos, A.D.T., Moonn, A.C. &Horwoon, T.P. (1970): The Giles Complex,central Australia: a stratified sequence of maficand ultr,amafic intrusions. tz Symposium on theBushveld Igneous Complex and other I-ayeredIntrusiong (DJ.L. Visser & G. von Gruenewaldt'eds.), Geol. Soc. S' Alr, Spec. Publ. t, 547-564'

PlpuxsN, H. (1970): Sulfide mineralogy of theKotalahti and Hitura nickel'copper ores, Finland'Ann. Acad. Sci. Fennicae Ser. A. III' 109.

I{ANVrLLB, G.D. & Pem, W.C. (1976): Nickelife-rous pyrrhotite delnsits, Knox County' south'easteni-Maine. /z Studies in New England Geol'ogy (P.C. Lyons & A.H. Brownlow, eds-), Geol'Soc, Amer. Mem. 146' 3t9'347.

RAJAIaANT, V. & Ner.onrtr, A.J. (1978): Parti'tioning of Fe Co, Ni and Cu between sulfideliquid-and basaltic melts and the composition ofNi-Cu zuUide deposits. Econ. Geol. 73, 8?'93'

298 THE CANADIAN MINERALOGIST

RAUDoHR, P. (1969): The Ore Minerals and theirIntergrowths. Pergamonn Oxford.

RoBrhrs, B. & GenoNrn, P.M. (1974): Synorogeniclayered basic intrusions in the Seiland p€trogra-phic province, Finnmark. Norg, Geol. unders.312, 91-130.

Rooorcr, LC.. (L977): Age of the RAna massif,north Norway. Fifth Eur. Colloq, Geochron.Cosmochron, & Isotope Geol. Pisa (Abstr.).

RYF" R.O. & Omamo, II. (t974): Sulfur and carbonisotopes and ore genesis: a review. Econ. Geol.69,82G842.

Snrons, d & Boxo, R. (1927): TyngdemAlinger,RAna, Ballangen, Nordland. Norg. Geol. (Jnders.Rap. 1588 (Unpubl.).

SraNroN, R.L. (1972): Ore Petrology. McGraw-HilI, New York.

TrtrAyER, T.P. & JAcKsoN, B.D. (L972): A slassifi-cation of igneous rocks by their history of crystal-lization and emplacemenl US. Geol. $u.rv. Prof.Pop. 800.8,79-83.

WAGER, L.R., Bnowx, G.M., & Weoswonrn, WJ.(1960): Types of igneous cumulates. J. Petrology1, 73-85.

Vy'u,soN, H.D.B. & ANDERSo\ D.T. (1959): Thecomposition of Canadian sulfide ore deposits.Can. Inst, Mining Met. Trans.62,327-339.

Wyr-r,re, P.J. (1969): The origin of ultra^mafic andulta,basic tooks. Tectottophys. 7, 437455.

Received July 1978, revised manuscrip acceptedDecember 1978.