EconomicGeology

Vol. 81, 1986, pp. 640-659

Kiruna-typeDeposits:Their Origin and Relationshipto Intermediate

Subvolcanic
Plutonsin the Great Bear MagmaticZone, NorthwestCanada*
ROBERT S. HILDEBRAND

GeologicalSurveyof Canada,588 BoothStreet,Ottawa, Ontario KIA OE4,Canada

Abstract

Magnetite-apatite-actinolite
rocks,similarto thoseat Kiruna,Sweden,andthe St. Francois
Mountains,Missouri,are associated
with sevenepizonalplutonsthat intrudeandesiticstratovolcanoes
of a broadlyfolded,earlyProterozoiccontinentalmagmaticarc locatedin the
northwestern
Canadian
Shield.The plutons,mainlymedium-grained
quartzmonzonite-mon-

zodiorite-diorite
(IUGS)sheets
andlaccoliths
5 to25kmindiameter,
arespatially,
temporally,
andcompositionally
relatedto hostandesitic
stratovolcanoes.
Theywereemplaced
at 2 to 3
km depth,arecompositionally
heterogeneous,
metasomatically
altered,andhadplagioclase
asthe liquidusphase.In additiontheyhavekilometer-wide
alterationhaloscomprising
an
innerzoneofnearlycomplete
wall-rock
albitization,
anintermediate
zoneofmagnetite-apatiteactinoliteveins,pods,brecciasandwall-rockreplacement,andanouterzoneof disseminated

sulfides.
Onepluton(Balachey)
hasa halothatismoreor lesscompletely
outsidethebody,
but another(RainyLake)hasa halothatis partlysuperimposed
overit. Separation
of iron
phosphate
andsilicatemeltsby liquidimmiscibility
is notsupported
by: (1) the gradualreplacement
of plagioclase
by chessboard
albitetowardthe roofof the RainyLakepluton,(2)
thechemical
trendsintheRainyLakeintrusion,
(3)themineralogical
zoning,lowtemperature
mineralogy,
andreplacement
texturesof the magnetite-apatite-actinolite,
and(4) the relationshipof the magnetite-apatite-actinolite
veinsand bodiesto the plutons.However,all

available
dataarecompatible
with a deutericorigin.It is concluded
thathightemperature,
lowinitialwatercontent,andshallow-level
emplacement
by intermediate
plutonsarenecessary
for the development
of Kiruna-typedeposits.
Introduction

KIRUNA-TYPE
depositsare variable concentrations
of magnetite-fiuorapatite-actinolite
foundin volcanoplutonicterranesfrom the Proterozoicto the Cenozoic. The best knownand probablylargestdeposits
are thosefoundnearKiruna,Sweden(Frietsch,1978;
Geijer andOdman,1974), but other well-knowndeposits occur in Missouri (Snyder, 1969), Chile
(Bookstrom,1977), andat numerouslocationsaround
the PacificOceanbasin(Park, 1972).
There have been many detailed studieson the
mineralogyandoccurrenceof the deposits,but to date
nogeneralmodelfor their genesis
hasbeenaccepted.
In fact, the number of modelsfor their origin approachesthe numberof deposits.For example,the
origin of Kiruna-typedepositshasbeen ascribedto
hydrothermal,or late magmatic,processes(Crane,
1912; Geijer, 1915, 1930; Geijer and Odman, 1974;
Bookstrom,
1977; Frietsch,1978), liquidimmiscibi|ity (Daly, 1915;BadhamandMorton,1976;Lundberg
and Smellie, 1979; Sme|lie, 1980), exhalative sedimentaryprocesses(Hegemannand Albrecht, 1954;
Partk,1973, 1975, 1985), andremobilizationof iron
and phosphorusfrom older sedimentaryrocks(Lan-

* GeologicalSurveyof CanadaContributionNo. 10486.

0361-0128/86/527/640-2052.50

640

dergren, 1948; Park, 1961; Frutos and Oyarzun,

1975).
A folded 1.87-b.y.-old subduction-relatedmagmatic arc exposedin the northwesternCanadian
Shield at Great Bear Lake containsthick sequences
of calc-alkalineandesitcinterpreted as ancientstratovolcanoes(Hildebrand,1984). Emplacedinto, and
in somecasesjust beneath,the andesitesare intermediateconcordantplutons,collectively.termed the
"earlyintermediateintrusivesuite."All of the plutons
have concentrationsof magnetite-apatite-actinolite
and similar spatial,temporal, and compositionalrelationshipsto their respectivehostandesites.Thispa-

per focuseson two of the plutons,the Rainy Lake

IntrusiveComplexandthe Ba|acheypluton,in order
to understandthe genesisof Kiruna-typedeposits.
Regional Geology

The plutonslie alongthe westernmarginof Great

Bearmagmaticzone (Fig. 1), the youngestmagmatic
belt of the 2.1- to 1.8-b.y.-oldWopmayorogen(Hoffman, 1980; Hoffmanand Bowring, 1984). The zone,
at least800 km longby 100 km wide, containsthick
successions
of subgreenschist
faciesvolcanicandsedimentary rocks(Hoffmanand McGlynn, 1977) dated
by U-Pb on zirconat 1.87 to 1.86 b.y. (Bowringand
Van Schmus,in press).The volcano-sedimentary
sequenceslie unconformablyon an older, deformed

GREAT BEARZONE KIRUNA-TYPEDEPOSITS:ORIGIN

641

:( post-orogenic
cover
GreatBearplutonics
"'"'::
GreatBearvolcanics
; Coronation
Prism
HottahTerrane
Great
Bear

0oOooooo
o o
o

50

I00

km

FIG. 1. Regionaltectonic map showingdistributionof volcanicand plutonic rocksin the northern

part of the GreatBearmagmaticzone,basementto the west(Hottahterrane)andto the east(Coronation
prism), youngercover, and the locationof Figure 2.

sialicbasementcomplex(Hildebrand et al., 1983;

Hildebrand,1984) andare cut by biotite-hornblende
and hornblende-bearing
granitoidplutons,someof
which are demonstrablysynvolcanic.The volcanic
rocks are mainly calc-alkalineand form a compositional seriesrangingcontinuouslyfrom basaltto rhyolite. Basedon major and trace element chemistry,
mineralogy, eruptive styles, and regional tectonic
setting, the supracrustalsequencesof the area are
interpretedto representpart of a continentalmagmatic arc related to eastward directed subduction of

axes,obliqueto the zone,andthenintrudedby a suite

of mainlybiotite-bearinggranitoidbodieswhoseage
isabout1.84 b.y. (BowringandVan Schmus,in press).
Later (1.84-1.80 b.y.), the entire Great Bear magmaticzonewasout by a swarmof northeast-trending,
right-lateraltranscurrentfaultsthat belongto a re-

gionalsystemof conjugatetranscurrentfaultsproducingeast-westshorteningand north-southextension.

The localhostrocksfor the early intermediateintrusive suite are a diverse assemblageof volcanic
and sedimentaryrocksthat crop out in two areassep-

oceaniclithosphere(Hildebrand, 1981, 1982a; HildebrandandBowring,1984). Shortlyafter 1.86 b.y, aratedby faultsandyoungergranitoidplutons(Fig.

unitsare not correlative
the entire zonewasfoldedaboutnorthwest-trending 2). Individualstratigraphic

642

ROBERT S. HILDEBRAND

zonedalterationhalocomprising
aninnerzoneof intenselyalbitizedwall rock,an intermediate
zoneof
magnetite-apatite-actinolite
veins,pods,anddissem-

+++++++
+

Port

Radii

inations,andanouterpyritic zone.In somecasesthe

sulfide zone hosted rich, but now mined out, poly-

metallicveinscontainingnativeAg andBi, Ni-Co arsenides,and pitchblende.It was one of these vein

systems,
locatedat PortRadium,whichprovidedthe
uraniumore usedin the ManhattanProject (Anonymous,1977). Bouldersof plutonicrock, similarin
mineralogyandtextureto one of the plutons,occur

in conglomerate
nearthebaseofthe ash-flow
tuffsequence.Thisindicates
thatthe plutonwasemplaced
prior to eruptionof the tuffs,duringor very shortly

granite
++++++++++
+

after andesiticvolcanism.In fact,the compositionand

locationof the plutonssuggestthat they all maybe

+
+

+
+

subvolcanic intrusions related to andesitic volcanism.

+
+

The geologyof the southernareaiscomplexin detail but canbe dividedinto two cycles,eachrelated
to collapseof a major caldera(Hildebrand,1984,
1985a). Initially, voluminousrhyolitic magmawas

+
+

eruptedasashflows.Thisled to collapse
of the first
cauldron,which was then filled with fluvial and la-

Balachey ::
+
+

+
+

+
+

+
+

RainyLake

custrinesedimentary
rocksandseveralkilometersof
andesiticlavasandbreccias,tuff, andepiclasticrocks.
The secondcauldron,causedby eruption of dacitic

magma,formedsoonafterwardandthe topographic
depression
remaining
aftercollapse
wasfilledby fluvio-lacustrinesedimentaryrocksand andesiticlavas
and associated breccias.

The southernarea was intruded by two interme+

+
+

+
+

+
+

+
+
.

km

20

FIG. 2. Simplifiedgeologicmapshowingthe distributionof

supracrustal
rocks(screened),early intermediateintrusivesuite
(checkered),
and youngergranitoidplutons(crosses)
alongthe
easternshoreof GreatBearLake. Balacheyplutonandthe Rainy
Lake Intrusive Complexare labeled.

diate plutonssimilarto thoseof the northernarea:

the Balacheypluton and the RainyLake Intrusive
Complex.The Balacheyplutonintrudesthe postcollapseandesitesof the first cauldroncycle and the
Rainy Lake pluton intrudesthe underlyingintracauldronfaciestuff and fiuvio-lacustrinedepositsdepositedin the samecauldron.

The ageof emplacement

of theBalachey
plutonis
tightly constrained
by both stratigraphy
and geochronology
because
muchofthenortheastern
margin

overlainby deposits
between the areas, but similarities in U-Pb ages, of the intrusionisunconformably
relatedto collapseof the second
rock types, and regionalstratigraphicpositionsug- of coarsemesobreccia

cauldron
whoseoutflowsheetlieson
gestthat volcanism
in both areaswasroughlycon- majorash-flow
postcollapse
andesites
ofthefirstcauldron
cycle(HilThe geologyof the northernareacanbe divided debrand,1984, 1985a).Thisindicatesthat it waseminto two main eruptiveperiods(Hildebrand,1981, placedduringor shortlyaftereruptionof theandesof discordant
zirconfroma pluton
1982b):development
of largeandesitic
stratovolca-ites.Populations
emplaced
intothe center
noes,andsubsequent
emplacement
of largevolume interpretedto be resurgent
ash-flowtuff sheetsand associatedcauldroncollapse. of the secondcauldron(Hildebrand,1984, 1985a)

temporaneous.

2 m.y. (Bowringand Van

The stratovolcanoes,
up to 3 km thick andcomprising yield an age of 1,868 +__

in press).The U-Pbupperinterceptageof
mainlyaugitc-andhornblende-bearing,
plagioclase Schmus,
the
Balachey
plutonis 1,868 +_8 m.y. (Bowringand
porphyriticandesitc,are folded and exposedin
Van Schmus,
in press).Therefore,the plutonis synobliquecrosssections.
anditsage,asconstrained
bybothU-PbgeoAt leastfive sill-likeintermediateplutonsintrude volcanic
and stratigraphic
arguments,
is about
the stratovolcano
complexboth at high stratigraphic chronology
levels and at its base. Each one of the plutonshas a

1,868 m.y.

GREAT BEAR ZONE KIR UNA-TYPE DEPOSITS: ORIGIN

643

The Rainy Lake IntrusiveComplexhasnot been

datedanditsageis notconstrained
by fieldrelations,
but its compositional
and texturalsimilaritiesto the
andesitesand the Balacheypluton suggestthat it is
roughly coevalwith both. Thus both the Balachey
pluton and the Rainy Lake Intrusive Complex are
consideredto be temporallyrelated to postcollapse
andesitesof the first cauldroncycle.
Like the plutonsin the northernbelt, there is no
evidenceto suggest
that eitherthe Balacheyor Rainy
Lake bodiesventedto the surfacefrom their present
level of emplacement;however,their closecompositionalandtemporalrelationshipto the andesitesindicatesthat the lavasandplutonsmayrepresentpart
of the samebatchof magma.If so,they mightbe similar in compositionand sizeto bodiesthat at deeper
levelswere the sourceof the eruptedandesites.Possibly both plutonswere able to rise and invadethe
volcanicedificebecausethey did not erupt and thus
dehydrate.
The BalacheyPluton

This northwest-trending
plutoncropsout continuouslyover a lengthof 20 km and a width of I to 6

km (Fig. 3). The majorrocktypesof the bodyare

medium- to coarse-grained,seriate, hornblende
quartzmonzonite(IUGS nomenclature)in the northwest and quartz monzodioritein the southeast.Lo-

cally,thereisa plagioclase
porphyriticborderphase.
Contacts
betweeninternalphases
aregradational,
and
moderatemineralogical
variations(2-10%) occuron

:.,Qyounger
rocks
( Balachey
Pluton
older rocks

alblte
zone
O mag-ap-act
zone
::'O
sulphide
zone
O lessaltered
rocks

km

scalesranging from kilometers to centimeters.A

FIG. 3. Geologicsketchmapof the northwesthalf of the Balcommonfeaturepresentthroughoutthe intrusionis acheyplutonshowingthe distributionof rocksyoungerthanthe

the occurrenceof fracturesalongwhichactinolitic

amphiboleis concentrated.

pluton and alterationzonesin rocksolder than the pluton. The

Balacheyplutonoccupiesthe core of a northwest-plunging
anticline.Locationsof the samplesin Tables1 and2 are indicatedby

All contactsof the Balacheyplutonwith surround- dots.

ingcountryrocksare sharpand,like the bodyitself,
trendnorthwest.Alongthe entiresouthwestern
margin and part of the northeastern
marginthe pluton ferred to be
intrudeslavas,tuffs, and sedimentaryrocksof the (Fig. 4).
postcollapse
andesitepile, alongwith smallrelated

sill-like and folded into a broad anticline

intrusionsof monzonite and diorite, which are all

Petrography
stronglyaltered up to 2 km from the contact.The
contactdipsawayfrom the pluton at aboutthe same
The mainpart of the intrusionis massiveandcon-

inclinationasbeddingin the lavasandsedimentary sistsdominantlyof euhedral,seriticizedplagioclase

rocks(about30).Rocksyoungerthantheplutonare phenocrysts,
I to 4 mm long, in a matrixof quartz
folded;bothsidesof theplutonareflankedby north- andmicroperthite,typicallyforminggranophyric
inwesterlytrendingsynclines.Therefore, Hildebrand tergrowths.Concentricshellsof sericiteoutlineorig(1984) arguedthatthe Balacheyplutonis folded,oc- inal zoningin the plagioclasephenocrysts.Quartz
cupiesthecoreof a northwesterly
plunginganticline, alongwith microperthiteappearto replaceplagioandwasintrudedasa largesill.The lowercontactof clasein manyplaces.Fibrousgreenamphibole,probthe Balachey
plutonis not exposed,
but by analogy ably actinoliteor actinolitichornblende,is subhedral,
with otherplutonsof the earlyintermediate
intrusive replacedby chloritealongcleavageplanes,andforms
suitewhichare exposedin obliquecrosssectionand clots2 to 4 mmacross.Someclotsarepartlyreplaced
are concordant,the lower contactof the Balachey by epidote.AnhedralFe-Ti oxidesare mostlyconplutonis probablyfairlyfiat.Thus,the plutonis in- centratedwithin the clotsof amphibole,but sparse

644

ROBERT S. HILDEBRAND

SW

Norex
syncline

NE

FIC. 4. Schematiccrosssection(not to scale),modifiedafterTirrul (1976), showingmajor foldsof

the southern

area.

hexagonal
platesof hematitearescatteredthroughout. matic arcs.The patternsare nearly identical to the
Hexagonalprismsof apatite lessthan 0.5 mm in di- mostsiliceoussamplesof lava from the postcollapse
ameterandtiny subhedralcrystalsof zircon are com- andesitcpile exceptthat samplesfrom the plutonhave
mon accessoryminerals.
larger Eu anomalies.
Samplescollectedfrom near the roof of the pluton
haveplagioclase
phenocrysts
that arepartlyreplaced Alteration of wall rocks
by an interlockingmosaicof anhedralquartz and alWall rocksof the Balacheyplutonare stronglyalbite. Vestigesof originalphenocrysts
occur,but most tered for a distanceof at least1 km out from the pluphenocrysts
havebeen completelyreplacedaround ton, assuming
that the contactcontinuesto dip at 30
their marginssuchthat they no longerappeareuhe- beneaththe surface.Three alterationzones(Fig. 3)
dral. Instead,they have marginsthat gradeinto and were mapped in the field: an inner zone of intense
interlockwith a groundmass
mosaicof quartz and al- bleachingand albitization, an intermediate zone of
bite. Felted matsandirregularclotsof actinoliteform magnetite-apatite-actinolite,and an outer zone of
pseudomorphs
of ferromagnesianminerals,perhaps mainly pyrite. The criteria usedto define the zones
hornblendeor pyroxene.The bordersof the clotsand were asfollows:(1) the boundarybetweenthe inner
aggregates
are raggedandfuzzy.OpaqueFe-Ti oxides and intermediatezoneswas placed at the first apare muchlesscommonthan in the main body of the pearanceof the assemblage
magnetite-apatite-actinpluton. Minute apatite needlesoccur in every thin olite; (2) the boundarybetweenthe intermediateand
section examined.
outer zoneswasmappedat the disappearanceof the
magnetite-apatite-actinoliteassemblage;and (3) the
Whole-rockchemistry
outer margin of the sulfide zone was placed at the
Eight samplesfrom the Balacheypluton were an- disappearance
of visiblegossan.Mappedin thismanalyzed at the Memorial University of Newfoundland ner, albite is presentin all three zonesand sulfides
for majorandtraceelementsandthe resultsare listed occurin the outer part of the magnetite-apatite-acin Table 1. The major elementswere analyzed by tinolitezone.The alterationzonesaretruncatedalong
atomic absorptionspectrometryexcept P205 which the northwestsideof the intrusionby the unconforwas determined colorimetrically. Trace element mity at the youngercauldronmargin.
abundances
were determinedby X-ray spectrometry
The inner,or albitezone,ischaracterized
by nearly
on a Phillips 1450 instrumentusingpressedpowder completealbitizationof the postcollapse
andesiticlapellets. For the most part, ten rare earth elements vasand sedimentaryrocks.Most originaltexturesare
were determined in each of three samplesby first obliteratedandthe rocksweatherwhite to palepink.
separatingthe rare earth elementsfrom other oxides Nearly all bedded rocks are intensely brecciated;
by cationexchangetechniquesandthen analyzingthe however,vestigesof beddingcanstillbe seenwithin
residueby X-ray fluorescenceon the samePhillips individual fragments.A fine-grained,pre-Balachey
1450 instrument.
monzoniticintrusionin this zone is completelyalbiThe analysesshowthat the Balacheypluton is of tized adjacentto fractures,which gives the rock a
typicalcalc-alkalineintermediatecomposition.Silica stripedto mottledappearance.Andesiticlavaswithin
contentof the plutonrangesfrom about60 to nearly the albitezoneare commonlyreplacedby granoblas66 percent and the least silica-richsamplesoverlap tic-polygonalalbitewith onlya few specksof chlorite.
compositionallywith postcollapseandesiticlavasof Hildebrand(1982a) comparedthe alteredrockswith
the firstcauldroncycle.Rare earth elementanalyses relativelyunalteredsamplesfrom outsidethe haloand
exhibitlight rare earth elementenrichmentpatterns concludedthat virtually all elementsoriginallypresandhavethe high overallabundances
(Fig. 5) typical ent in the lavasandsedimentaryrocks,includingthe
of high K andesitesfrom Cenozoiccontinentalmag- so-called "immobile elements," have been mobilized

GREAT BEAR ZONE KIRUNA-TYPE DEPOSITS:ORIGIN

TABLE1.

Sampleno.
SiO2
TiO2

J-79-93
65.4
0.33

H-79-40
65.5
0.34

645

Major and Trace Element Analyses,BalacheyPluton

J-79-62

J-79-92

Jo79-66

Jo79o61

Ho80-26

60.5
0.59

64.3
0.45

65.7
0.45

62.5
0.44

64.0
0.58

H-80-24
63.0
0.63

AlcOa

15.4

15.9

14.1

1.5.5

15.5

14.4

14.1

14.4

FeOa*

4.86

4.12

6.49

4.54

3.65

5.59

5.80

6.74

MnO

0. ! 0

0.04

0.36

0.09

0. ! 1

0. ! 1

0.10

0.15

MgO

1.94

1.51

3.33

1.53

1.56

2.98

2.24

2.80

CaO

!.02

3.71

4.76

2.1!

2.09

3.32

3.74

3.80

NaO
KO
P205

3.00
4.98
0.08

3.25
4.03
0. ! 0

2.45
4.15
0. ! 4

2.98
5.16
0.09

3.43
5.13
0.08

3.08
4.19
0.09

2.61
4.07
0.07

2.61
4.16
0.11

L.O.I.

2.41

0.64

2.88

2.30

! .90

2.17

1.33

1.55

Total

99.52

99.14

99.69

98.99

99.60

98.87

98.64

99.95

14

12

13

13

12

13

13

Zr
Y
Sr
U

Nb

169
30
134
9

187
36
280
5

185
25
264
4

179
38
199
7

184
32
214
I

156
31
229
6

162
31
222
2

203
31
271
1

Rb
Th
Pb

184
14
10

142
17
20

160
15
62

177
17
24

132
17
19

148
22
22

155
19
18

155
14
16

Ga
Zn
Cu
Ni
Cr
V

10
73
20
14
11
68

13
29
22
13
14
59

94
113

14
45
19
19
16
63

13
70
22
16
11
58

10
90
30
27
57
95

14
59
0
5
18
100

16
84
0
4
20
115

834

1,185

1,131

979

1,048

1,077

125
63

Ba

1,145

973

La
Ce
Pr
Nd
Sm
Eu

15.06
36.58
4.64
17.74
3.08
0.22

23.73
55.91
6.90
24.43
4.91
0.63

16.79
47.04
6.13
23.07
4.82
0.72

Gd

3.36

3.84

3.89

Dy

4.56

3.65

4.11

Er

1.94

2.29

2.32

Yb

2.32

3.13

2.86

Oxidesin weightpercent,traceelementsin ppm;0 = not detected;L.O.I. -- losson ignition;FeOa* = total Fe asFeOa

and partiallyremoved(Table 2). The albite zone is

entirely outsidethe pluton at its northwesternend
but partly within the southeastsideof the intrusion.
The zone of magnetite-apatite-actinolite
contains
thosemineralsaspods,veins,disseminations,
and as

suspectedto be vesicles.Someepidote-linedcavities
are completelyfilled with coarsepink apatite.
The sulfidehalocomprisesabundantgossans
up to
10 m across,which in the early 1930s probablyattractedprospectors
to the areain searchof silverand
rosettes with albite. Most common are abundant small uranium.The pyrite is disseminated
throughoutthe
veins(1-2 cmwide) in whichfibrousgreenamphibole, areain concentrations
ashigh as25 percent.The zone
orientedperpendicularto the vein margins,occurs containedlocalpolymetallicveins,nowminedout, of
with interstitialanhedralpink apatiteand octahedra native Ag, Bi, and Ni-Co arsenides.

of magnetiteor martite.Podsof magnetite-apatiteThe Rainy Lake Intrusive Complex

actinolite,up to 2 m across,typicallycontaincoarse
bladesof amphiboleto 30 cm long,magnetiteoctaThe Rainy Lake Intrusive Complex(Tirrul, 1976)
hedrato 5 or 6 cm, andpatchesof anhedralapatite is a compositionallyand mineralogically zoned,
rangingup to 20 cm. Rosettesof bladedalbite, up to

sheetlikepluton about1.5 km thick and 10 to 11 km

across(Hoffmanet al., 1976). The plutonwasfolded
flowsand haveinterstitialchlorite,amphibole,and after intrusionand is now exposedin oblique cross
magnetite.Epidote is anothercommonmineralwithin sectionon the limb of a majorsyncline(Fig. 4). It has
this alteration zone and in a few lavas it lines cavities
a fiat roof that is roughlyconcordantwith beddingof
15 cm in diameter, are common in altered andesite

646

ROBERT S. HILDEBRAND

100
l
7O

5O

3O

2O

Bal(cheyPluton
J-79,93

\ /

H-79.40

J-79.92

1La cle plr N'dm slm-uG'db Iy HoE'r ;m ib

FIG. 5.

Chondrite-normalized

rare earth elements from the

Balacheypluton.

Tirrul (1976) recognizedthat the plutonwasmineralogicallyandcompositionally

zonedparallelto its
flat roof; he mappedfive major divisions.From top
to bottom they are: a monzoniticborder phase,a
syeniticphasethat he dividedinto a fine-grainedupper part and a coarse-grainedlower part, a central
monzonite,and a lower monzodiorite(Fig. 6). In addition, there is a lower border monzonitethat presumablywasconnectedto the upper border phase
prior to intrusionof youngerplutonsanderosion.For
this studythe fine- and coarse-grained
syenitesare
groupedtogetherbecausethere are onlyminormineralogicaland chemical differencesbetween them.
Major mineral phasesof 18 samplescollectedfrom
near the base to the roof of the intrusion are listed in

Table 3 and are plottedagainstheightin Figure 7.

Border phase

the countryrocksand a floor that is slightlyconvex

downward, such that the thickest parts occur near
the center(Fig. 6). The plutonis intrudedby a younger unrelated syenogranitein the southeast,and
therefore,the lower contactof the RainyLake IntrusiveComplexis not preservedthere.
TABLE2.
Sample no.

The upper border phaseof the pluton is well exposedin manyplacesand is up to 20 m thick. It com-

prisesabout37 percentintenselyseriticizedplagioclasephenocrystsup to i cm long, i to 2 percent

relict clinopyroxenephenocrysts,lessthan 2 percent
serpentinized
orthopyroxene
phenocrysts,
andragged

Major and Trace Element Analysesof Altered AndesiticRocksfrom the BalacheyPluton Halo
H-79-131

H-79-132

H-79-133

H-79-126

H-79-124

SiO2
TiO2

59.1
0.28

71.3
tr.

65.0
0.00

59.4
0.20

61.5
0.45

A1O3
FeO3*

18.2
3.47

15.5
0.65

20.3
0.61

17.1
4.42

16.6
2.73

MnO

0.26

0.07

0.04

0.06

0.12

MgO

3.39

0.39

0.19

1.68

1.42

CaO

4.54

1.41

1.02

3.33

2.14

NaO
KO
P205

7.50
1.12
0.02

8.36
0.37
0.01

9.36
1.46
0.00

7.98
0.88
0.34

8.82
0.88
0.38

1.82
99.88

1.70
98.68

3.48
98.87

3.12
98.16

L.O.I.
Total

Nb

2.53
100.41

22

35

Zr
Y
Sr
U

175
19
370
0

35
0
83
0

171
6
203
0

206
32
78
5

203
45
45
5

Rb
Th
Pb

25
0
6

6
0
15

49
0
0

9
14
3

13
12
1

Ga
Zn
Cu

20
147
8

20

20
20
6

20
55
13

21

24
15

Ni

Cr

31

La

11

17

50

64

Ce

11

17

72

122

Ba

1,034

262

419

204

Samples131, 126, and 124 are andesiticlavas; 132 and 133 are andesiticsandstones
Oxidesin weight percent; trace elementsin ppm; 0 -- not detected; L.O.I. = losson ignition
FeO3* = total Fe as FeOa

15
38
0
0
122

163

GREAT BEAR ZONE KIRUNA-TYPE DEPOSITS: ORIGIN

RLc
O
RLB
0
RL
Al

20

40

647

12

20

40

20

, ,

: Albite
+Perthlte

Alblte

0
0

20

40

60

80

20

40

60

20

40

60

80

FIG.7. Modal
composition
ofsamples
collected
fromtheRainy

LakeIntrusive
Complex
plottedwithheightin theintrusion.
Ver-

km

tical axisin this and subsequentdiagramsis 1.5 km. Zonesare as

in Figure 6. F = ferromagnesianminerals,O = opaqueminerals,

FIG. 6. Geologicsketchmap of the Rainy Lake Instrusive

Plagioelase
= plagi0elase
phenoerysts,'
diamonds
= albititedikes,
ComPlexshowing
the surrounding
geology.Internalzonationof dots= other samples.
the Pluton:RLA= monzoniticborderphase,RL8 = lower monzodiorite,RLc = centralmononite,RLv = upper syenite,A
= intracauldron
faciestuffandotheroldbrrocks,B = postcollapse
intracauldron
lacustrineand fluvialrocks,C = postcollapse
an- stroyed,but the microproberevealedtiny domains
desiticlavas,breccias,
tuff,andepiclastic
rocks,D = quartzdiorite,
E = intrusivep!agioclase
porphyry,F = youngergran{te,XY of unalteredandesinc.Most of the phenocrystsare
rimmed with fresh-appearing
albite that' interlocks
= locations
of samples
analyzedin thisstudy.

with the matrix. Slenderneedlesoapatite, up to 2

mm long, are a'commonconstituentof the ground-

maficclots(3 mm)ofchlorite,amphibole,
carbonate, mass.
andopaqueoxidesin a muchfinergrainedgroundmass Magnetite-apatite-actinoliteveins, up to 30 em
of chlorites,carbonate,chessboardalbite, titanitc,

across,cut the bordermonzoniteandtypicallytrend

epidote,'amphibo.
le, and minorquartz.Concentric normalto the outer contact,locallycuttingabrossit.
layersof sericitein someof the plagioclase
pheno- In the veinscoarse,fibrousamphiboleisorientedwith
crystsoutlineoriginalzoning.Underthe microscope itslongaxisperpendicular
to thevein-margins.
Th
all originalplagi0clase
appearsto havebeen de- central zone of 'each vein, like in those around the
TABLE3.

Modal Analyses,RainyLake IntrusiveComplex

Sample
no. plagioclase
Perthite
QuartzAlbiteAmphibole
Chlorite
'-Epidote
Opaques
ApatiteCarbonate
Others

H-79-16
C-79-27
C-79-26
C-79-25

40.5
32.9

2.1

C-79-24

13.3

C-79-23
C-79-22
C-79-21

4.5
1.9

3.2

4.6
3.8
11.3
8.7

28.6
40.9
78.6
66.4

16.0
13.8
1.3
6.8

0.4
0.2
1.1
1.3

0.2
0.5
1.3

2.4
3.4
3.6
4.4

0.3

18.7

5.5

49.5

10.3

1.0

0.6

0.9

0.1

15.4
16.4
0.3

13.5
8.2
10.5

8.3
3.4
6.1

49.3
58.9
77.7

8.0
10.9
0.2

2.5
0.7
2.3

1.8
i.2
0.2

0.6
0.1
i.6

0.2
0.2

0.4

C-79-20

15.4

25.3

44.6

11.1

1.6

1.2

0.2

0.1

C~79-19

26.8

38.2

0.2

19.0

10.1

0.2

4.0

1.2

0.2

0.2

C-79-t8
C-79-17

39.5
51.5

32.1
19.2

3.2
0.7

7.3
7.6

13.0
17.4

0.3
0.4

2.3
2.2

1.5
0.9

0.1

C-79-16

52.2

17.5

0.3

5.2

18.2

1.1

3.2

1.3

1.0

C-79-15
C-79-!4

59.7
66.4

12.3
14.0

1.7
4.9

2.6

17.6
4.9

1.0
7.3

2.8
0.9

2.1
1.5

0.1

C-79-13

65.6

11.6

5.4

14.7

0.8

0.2

1.7

C-79-12

59.8

16.2

1.8

0.9

12.2

7.7

1.4

C-79-11

52.7

27.3

1.9

9.7

0.6

2.7

2.3

Othersincludecpxandopxin H-79-16 andmainlygranophyre

in othersamples

1.1
1.4

0.3
0.9
3.6
0.7

1.6
0.6

4.8

0.1

1.1

0.3

0.5

0.5

0.1

0.1

0.2

1.6

648

ROBERT S. HILDEBRAND

Balacheypluton,istypicallyfilledwith anhedralapa- felty in appearance.Quartz is sparsenear the base

andtop of the zone(Fig. 7).
tite and octahedraof magnetite.
Lower monzodiorite

Monzonite: The transition from monzodiorite

to

monzoniteis gradationaland occursover a distance

Exceptfor the lower border phase,seriatemon- of severalmeters.Mineralogicallythe changeischarzodioriteisthe lowestunit mappedwithinthe pluton. acterizedby an increasein the sizeof perthitegrains
The monzodioriteconsistsof 50 to 66 percent eu- to 7 mm andby the appearanceof chessboard
albite.
of plagioclase
phenocrysts
decreases
hedral, partly sericitized plagioclasephenocrysts The abundance
rangingin sizefrom i or 2 mm up to i cm, with in- regularlyfromabout40 percentnearthe baseof the
terstitialpalegreenamphibole,opaqueFe-Ti oxides, zone to about 15 percentnear the top. In addition,
perthite,granophyre,
andalterationproducts.
Modal minorchloriteandamphiboleare foundin the cores
plagioclasephenocrystcontentvariessystematically of plagioclase
phenocrysts.
Bothunalteredalbiteand
from about 53 percent near the baseto 66 percent perthitereplaceplagioclase
to the pointwhere only
coresremain.Thus,part of the decrease
in the centralportionanddecreases
in abundanceto oval-shaped
about50 percentnear the top of the zone.The pla- in modalplagioclaseis due to replacement,but even
gioclasephenocrysts
are tabletsof oscillatoryzoned when the effects are restored, there is still a considandesinethat are packedcloselytogethersuchthat erable decreasein plagioclasecontent. Overall, the
eachisincontact
withitsimmediate
neighbors;
there monzoniteappearsmuchfresherin thin sectionthan
arenoovergrowths
linkingadjacentcrystals.
The rock the monzodiorite. This is because so much of the
is an aggregateof plagioclase
phenocrysts
andinter- originalplagioclaseis now replacedby unalteredalstitialphases.
On someoutcrops
thephenocrysts
de- bite and perthite.
Syenite:
The uppercontactof the monzonitewith
fine a weakigneousfoliationthat led Tirrul (1976) to
suggestthat the monzodioriteis of cumulateorigin; thesyenite,liketheloweronewiththe monzodiorite,
in fact, it is an orthocumulatein the nomenclatureof is gradationalover severalmeters.The contactbetween the syeniteand the upper border phaseis
Wager et al. (1960).
the border
The pale greenamphiboleoccursmostlyasinter- sharp.Locally,the syenitetransgresses
stitialclots(3 mm)consisting
of fibrousmaterialwith phase,reducingits thicknesssuchthat it is absentor
randomopticalorientations,but in a few caseslight nearly so.
The syeniteweatherspink,probablyowingto the
brown to green pleochroicactinolitichornblende
hematite.Numerous
displaysuniformopticalorientationand is probably presenceof finelydisseminated
albititeup to 30 cmacrossand
magmatic.The crystalsof actinolitichornblendeare dikesof fine-grained
margins
cutthe syeniticphase.Their
2 to 3 mm acrossand partly fill areasbetweenpla- withgradational
measuredduring
gioclasephenocrysts.Anhedral opaque oxidesare orientationwasnot systematically
to the roof
ubiquitousin the clotsof amphibolebut rare in the mapping,but manytrendperpendicular
actinolitic hornblendes. Epidote and chlorite are of the pluton.Pinkfiuorapatite,
similarto thatwhich
veinsofthe
mostlyalterationproductsof amphiboleand are in- occursin themagnetite-apatite-actinolite
cludedwith it in Figure 7.
borderphase,coatsfracturesurfaces
in the top 10 m
Anhedralperthite, about4 mm acrossin maximum of the syenite.The fracturestrend nearlynormalto
diameter,is alsointerstitialand is commonlyinter- the roof of the intrusion.
grown with amphibole.In placesit appearsto have
The syenitediffersfromthemonzonitein thatperreplacedmarginalareasof plagioclasephenocrysts. thite decreasesin both abundanceand size (from 7
of
Granophyric
intergrowths
of quartzandmicroperthite to 4 mm),there is an increasein the abundance
albite, carbonateappearsin amphibole
have a similarmode of occurrencein that they are chessboard
is lessthan
interstitial,intergrownwith amphibole,and appear clots,and the modalpercentplagioclase
to replaceplagioclase.
In otherareasquartzdoesnot 16 percent.The destructionof the plagioclasepheoccurin association
with perthitebut assmallanhedral nocrysts
is sointensethat onlysparseellipticalrelics
grains.Tiny prismsof apatiteare a commonaccessory remain,heavilyreplacedby chlorite.Asin the central
mineral.Zircon wasnot foundin any of the thin sec- monzonite,when the originalprealterationsize and
tions examinedand even 100 kg of rock, collected shapeof the plagioclase
phenocrysts
are taken into
for U-Pb dating,failedto yield any.
account,the syenitecontains
fewerphenocrysts
than
In thin sectionsmadefrom samplescollectedpro- anyotherphaseof theintrusion.
Apatitehasa bimodal
either very commonor virtuallyabsent,
gressively
upwardin the monzodiorite,the plagioclase abundance,
phenocrysts
are moresericitized,havewider rimsof with no gradationbetween the two. Quartz content
albite, contain zoisite or clinozoisite in their central increases
dramaticallyfromabout5 percentnearthe
parts,andhavealbitereplacingthem alongcleavage baseof the zone to over 10 percent near the top.
planes.The clotsof amphibolebecomesmallerand Granophyreisabsent.In general,the amphiboleclots
the amphiboles
themselves
becomemoreraggedand arerarerandsmallerin the syenitethanin lowerparts

GREAT BEAR ZONE KIRUNA-TYPE DEPOSITS:ORIGIN

of the intrusion. Fe-Ti oxides are concentrated

in the

clotsasin lower partsof the pluton,but similarto the

clots,they becomefiner and more disseminatedupward. However, this is deceiving, becauseeven
thoughthe opaquemineralsbecomefiner they
tuallybecomemoreabundantin the uppermostparts
of the syenite(Fig. 7). Chessboardalbite increasesin
sizeupwardwithinthe syenite,andperthitedecreases
in size. In the uppermostpart of the unit perthite is
absentand tiny blebsof quartz are common.
Whole-rock chemistry: Seventeen samplescollectedfrom top to bottomthroughthe intrusion(Fig.
6) were analyzedfor major and trace elements,also
at Memorial University of Newfoundland,usingthe
samemethodsof analysisasfor samplesfromthe Balacheypluton.The resultsof whole-rockanalyses
from
the pluton are presentedin Table 4. Concentrations

of manyelementsare displayedgraphicallyin Figure

8.

SiO2, Na20, Nb, Y, Zr, and the heavy rare earth

elementsincreasefrom baseto top of the body.Other
oxides and elements, such as CaO, AlcOa, K20, Sr,

Ba,Rb,MnO, MgO, Cr, andNi decrease

upward.Note
that the abundancesof mostelementsin the upper
border phasegenerallyfall in betweenthoseof the
upper syenite and those of the lower monzodiorite.

Also of particularinterestis the couplednature of

NaO and KO; that is, as NaO increases, K20
decreases.KO, Rb, and Ba decrease with increasing SiOn.

Alteration:The RainyLake IntrusiveComplexhas

an alterationhalo similarto that of the Balacheypluton, but the zonation is not as well defined and its

649

Some areas of the body are brecciascontaining

abundant,mostlyangularblocksof countryrock,some
of which are partly replacedby magnetite-apatiteactinolitein a matrixof magnetite-apatite-actinolite.
Distinctive

blocks found in the breccia are similar in

grossappearanceto rhythmicallyinterbeddedlimestone-argilliteof the intracauldronfill sequenceexcept that the blocksare composedof alternationsof
magnetite-apatite-actinolite
with argillite. This suggeststhat the original limestonewas replaced by
magnetite-apatite-actinolite.
Overall, the brecciated
zonesappearsimilarto typical zonesof ore breccia
that occur at Kiruna, Sweden (Geijer and Odman,
1974).
Althoughmuchof the bodyappearsat firstglance
to be massive,closeinspectionshowsthat it isbanded
in many places.The bandsare visible becausethere
are slight modal variationsin the amountsof magnetite, apatite, and actinolite. Whether or not they
reflect originalbeddingin their hostrocksis debatable, but elsewherein the body as in rocksaround
the Balacheypluton, it is clear that sucha phenomenon does occur (see photographsin Hildebrand,
1984).
Veinsof magnetite-apatite-aetinolite
commonlycut
the upperborderphaseof the pluton.Suchveinsare
up to 30 em acrosswith marginalzonesof aetinolite
growing normal to the walls and central zones of
coarseapatiteandoetahedraof magnetite.Manyveins
containonly the assemblages
magnetiteq- apatiteor
apatite q- aetinolite.
The amphibolein the veins and larger bodiesis
aetinolite.It hasvery low A1(IV) K20, and NaO contents (Table 5). The magnetiteis characterizedby

spatialrelationshipto the plutonis different.In part,

the halo hasbeen partly destroyedby the intrusion very low TiO, CrO3, and MnO contentsand modof youngerplutons(Fig. 6). Magnetite-apatite-actin-erate amountsof vanadium.All of the apatitestudied
olite veinscut the upper border phaseand apatite is fiuorapatitewith fluorinecontentsaround3 percoatsfracturesin the uppermostportion of the sy- cent. Overall, the compositions
of the three mineral
enite. There are alsobodiesof magnetite-apatite-ac- phasesare similarto thoseof other Kiruna-typedetinolite abovethe roof of the pluton and they are posits.
mostlylarger than thosethat occur aroundthe BalOrigin of the Magnetite-Apatite-Actinolite
acheypluton. Only smallgossans
remain abovethe
magnetite-apatite-actinolite
bodiesandthere is some
The bodiesand veins of magnetite-apatite-actincontroversyamongworkersin the areaasto whether olite adjacentto the Rainy Lake Intrusive Complex
they are related to the pluton or are earlier synsedi- andBalacheyplutonled BadhamandMorton (1976),
mentarydeposits.
basedon experimentalwork by Philpotts(1967), to
The bestexposedbodiesof magnetite-apatite-ac-suggestthat an immiscibleiron phosphatemelt septinolite consideredto be related to the Rainy Lake arated from a silicic melt; however, this conclusion
Intrusive Complexare foundat the northwestend of appearsunlikely for the followingreasons:
the pluton aboveits roof. The largestis about50 m
1. Many of the veins containonly amphiboleor
across,hassharpoutermargins,andisheterogeneous
with respectto relative abundancesof magnetite, amphiboleq- apatitewithout magnetite.Thus, there
apatite, and actinolite.In placesalongthe contact, is commonlymore silicathan iron in the veins.
. The veinsare commonlyzonedfrom apatitein
massivemagnetite-apatite-actinoliteoccurs, but in
at the margin,a texotherplacesbandedmagnetite,apatite,andactinolite the coreto amphibole-magnetite
or only a thin veneer of apatite or actinolitewere ture incompatiblewith their derivationfrom an iron
found.
phosphatemelt.

650

ROBERTS. HILDEBRAND
TABLE

4.

Major andTrace ElementAnalyses,RainyLake IntrusiveComplex

Lower monzodiorite

Sampleno.

C-79-11

C-79-12

C-79-13

C-79-14

Central monzonite

C-79-15

C-79-16

C-79-17

C-79-18

C-79-19

Co79o20

SiO
TiO
AltOs
FelOn*

55.0
0.66
19.7
6.02

53.6
0.51
20.6
5.98

53.6
0.63
20.5
6.04

52.8
0.72
18.9
7.75

54.2
0.94
19.2
7.68

55.5
0.92
18.7
6.75

54.8
1.03
18.6
6.54

55.9
0.73
17.8
7.04

58.3
0.86
17.0
5.38

60.8
0.83
16.9
5.25

MnO

0.29

0.10

0.25

0.22

0.37

0.18

0.28

0.15

0.14

0.06

MgO

2.26

2.24

2.25

2.54

2.58

2.48

2.62

2.18

2.18

1.77

CaO

4.29

5.49

5.42

5.14

5.17

5.49

5.01

3.65

3.68

3.16

Na20
KO
P205

3.89
4.73
0.26

4.18
3.51
0.26

3.51
3.89
0.27

3.21
3.72
0.55

3.68
3.72
0.51

4.64
3.60
0.51

4.48
4.25
0.57

4.14
4.69
0.42

6.26
2.92
0.48

7.38
2.56
0.35

1.72
100.55

1.82
100.00

2.17
98.87

1.50
98.71

L.O.I.
Total

2.66
99.76

2.81
99.28

2.57
98.75

2.81
98.36

2.23
100.28

0.90
99.96

Nb
Zr
Y
Sr

8
129
25
530

3
64
15
642

4
87
18
560

8
89
24
579

4
83
24
516

9
107
34
522

6
126
35
495

14
167
31
286

13
160
41
227

13
180
36
202

U
Rb
Th
Pb
Ga
Zn
Cu
Ni
Cr
V

2
150
10
13
20
275
8
13
18
131

0
110
9
5
19
77
4
22
22
104

2
126
6
10
19
83
38
21
14
145

0
113
1
21
21
170
38
9
10
212

1
120
8
24
16
304
4
17
0
193

4
92
2
1
22
115
0
8
4
155

1
111
10
22
15
306
0
16
0
168

0
109
1
10
22
108
0
1
0
148

2
50
11
7
20
102
0
3
0
130

1
43
11
21
20
43
6
6
0
109

904

1,158

Ba

1,028

1,160

1,423

La

20.86

21.94

24.10

22.09

25.87

40.56

23.38

26.55

18.97

Ce
Pr
Nd
Sm
Eu

44.99
5.49
22.97
3.80
0.84
3.54

46.95
5.11
21.30
3.85
0.99
2.51

52.11
6.43
23.55
3.99
1.29
2.92

44.14
28.52
5.46
1.69
4.77

53.75
7.41
31.82
6.54
1.28
5.13

80.57
10.27
44.43
9.39
1.33
6.46

50.72
7.41
30.70
6.92
0.99
5.07

50.72
7.07
30.88
6.15
0.96
4.92

44.80
6.12
27.44
6.38
1.33
4.45

Gd

Dy

935

1,614

984

819

849

2.14

3.46

3.72

4.02

5.80

4.81

5.38

5.63

Er

1.82

1.47

1.33

2.21

2.63

2.09

2.63

2.60

2.69

Yb

1.49

1.20

1.69

2.19

3.28

2.34

3.27

2.10

2.65

Oxidesin weight percent;trace elementsin ppm; 0 = not detected;L.O.I. = losson ignition

FelOn* = total Fe as FeO

3. Granularmagnetite-apatite-actinolite
replaces
sedimentaryrocksand andesiticlava flows.
4. Apatite coatsfracturesin the upper syenitic
phaseof the RainyLake IntrusiveComplexwhichindicatesthat it wassufficientlysolidto fracturewhen
the apatitecrystallized.
5. The composition(low Na20, K20, andTiO) of
the amphibolein the veinsandbodiessuggests
crystallization temperatures(Helz, 1979, 1982; Hammarstromand Zen, in press)too low to maintainan
iron phosphatemelt. In fact, experiments(I. G. Reichenbach;pers.commun.)showthat samplesof the
magnetite-apatite-actinolite
fail to melt completely
after 48 hoursat 1,150Cunderdry conditionsat 1
atmosphere.

6. The magnetitecontainsvery smallamountsof

TiO which is inconsistentwith a parentalmelt that
wasever in equilibriumwith an intermediatesilicate
melt.

7. Globules of iron phosphatemelt would sink

in a silicicmelt owingto the large contrastin density between them (Daly, 1915). Therefore, if the
magnetite-apatite-actinolite
bodieswere derivedby
a processin which an immiscibleiron phosphate
melt separatedfrom a silicicmelt, they shouldoccur
near the base of the melt zone not above the intrusive
bodies.

If the magnetite-apatite-actinolite
bodiesare not
generated
by the separation
of anironphosphate
melt

GREAT BEAR ZONE KIRUNA-TYPE DEPOSITS: ORIGIN

651

T,BLE4--(Continued)
Border monzonite

Upper syenite

Sample no.
SiO2
TiO2

C-79-21

C-79-22

C-79-23

C-79-24

C-79-25

C-79-26

C-79-27

61.4
0.73
16.7
4.80

61.1
0.76

61.3
0.86

64.0
0.63

66.0
0.59

AlcOa
FeOa*

66.1
0.92
17.6
1.53

16.1
5.33

16.5
6.35

14.8
4.95

15.3
2.06

57.9
0.65
17.5
6.03

MnO

0.05

0.06

0.10

0.08

0.12

0.04

0.18

MgO

0.51

1.88

1.66

1.82

1.31

0.49

2.79

CaO

0.88

2.48

2.22

2.13

2.22

2.49

3.99

NazO
KzO
P205

9.41
0.72
0.03

7.82
1.28
0.44

5.80
3.14
0.44

5.95
3.24
0.51

7.72
0.64
0.46

8.88
0.38
0.02

6.01
2.27
0.32

L.O.I.
Total

1.37
99.27

1.19
98.78

2.12
98.77

1.64
100.38

1.78
98.63

2.56
98.81

2.68
100.32

Nb

H-79-16

19

17

14

16

20

19

11

247
25
54
1

208
38
138
1

213
38
165
4

199
43
150
3

271
58
72
2

260
23
68
9

150
30
288
3

143
27

Rb
Th
Pb

15
0
5

42
13
6

86
12
15

82
8
23

4
11
4

58
12
15

104
10
10

Ga

23

19

21

20

22

21

18

Zn

49

44

53

62

58

126

Cu
Ni
Cr
V

2
0
0
24

0
I
0
58

17
4
0
90

16
5
0
85

0
4
0
117

25
3
0
44

0
26
58
151

Ba

112

479

1,002

869

74

87

687

La
Ce
Pr

15.61
2.07

47.62
6.49

Nd

7.76

28.61

Sm
Eu

1.67
0.27

5.82
1.08

Gd
Dy

1.36
1.92

Er

Yb

Zr
Y
Sr
U

20.65
47.45

125

388
4
347
12
17
12

139
4

55
131
183

1,153

47.52
7.17

107.81
234.18
24.89

19.72
37.09
4.13

28.63
63.59
7.50

28.00

28.57

94.60

18.81

30.33

6.47
1.42

5.59
1.02

13.78
1.30

4.27
1.06

5.54
1.25

5.02
5.55

5.40
5.09

5.42
4.33

7.47
7.14

3.45
3.03

4.03
1.55

4.77

1.58

2.62

2.44

2.54

3.05

1.55

1.68

2.12

1.98

3.34

2.49

2.67

3.21

2.23

2.06

31.75
65.50
7.57
31.63
6.32
1.32

4.90

Oxidesin weightpercent;traceelementsin ppm;0 = not detected;L.O.I. = losson ignition

Fe2Oa* = total Fe as FeOa

froma silicate
melt,towhatprocess
dotheyowetheir Thesearereasonable
assumptions
to makein thiscase
origin?I takea somewhat
traditional
viewandsuggest becausethere is no evidencefor late-stagemixingof
that all availableevidenceindicatesthat they result the RainyLake melt with a more maficmelt nor are
fromtheexsolution
ofa volatilephaselatein thecrys- thereanyof the textualfeaturescommonlyassociated
tallizationhistoryof the plutons.The followingdis- with periodicallyrefilled magmachambers.The apcussionreliesheavilyon data from the Rainy Lake proachusedhereisto estimatesomeof the parameters
IntrusiveComplexbecause
completesections
through of the RainyLake magma,suchascomposition,HO
it are availablefor studyandbecauserelationships content, and temperature at the time of intrusionin
withinthe bodyare well exposed.
orderto constrainthe originof the magnetite-apatiteAs an initial condition it is assumed that for much

actinolitesegregations.
The best estimateof the original compositionof
Complexbehavedasa closedsystemafter intrusion. the RainyLake magmacomesmainlyfrom the simiThat is, therewere no influxesof new magmainto larity of the borderphaseto the andesites
just above
the chamber after initial intrusion nor was there a its roof.The followinglinesof evidenceindicatethat
reservoirof hotter, more maficmagmabeneathit. the originalmagmathatcooledto formthe RainyLake

of itscrystallization
history,theRainyLakeIntrusive

652

ROBERT S. HILDEBRAND

..'

$i02

:5

56

50

62

68

:5

,.,

,,

St'

, i.......i I , I

500

Nb

50

150

.-

4:0

e
:

I: ' !

,t

.,

AI20:5
, , , i i iI i i
14

18

22

1200

M,
gO
0
i
i

:"5
2
i

l!

.I

.2

.5

.4 o

IO

500

.5

.7

.9

I.I

Fe203,
4

..

Ti02

0'"

MgO
i
,

0 ....

'0

200

2_

5O

Cr

20

4b

ioo

'

200 0

..

40

Ce

i''

FIG. 8. Anhydrouswhole-rockmajorandtraceelementanalyses
of the RainyLake IntrusiveComplex
plottedwith height in the intrusion.The verticalaxisis 1.5 km in height.Symbolsand zonesas in
Figures 6 and 7.

IntrusiveComplexwasandesitic:(1) the upperborder is similar to the andsites;and (3) there is a close

phaseof the intrusionis very similarto postcollapse spatial,and probably temporal, relationshipof the
andesiticlavasof the firstcauldroncyclein thatboth plutonwith the andesires.
Therefore,the RainyLake
the lavasand the border phasecontainsubstantial magmais assumedto be of silicic,high K andesitic
quantitiesof large oscillatoryzoned andesine,and composition.
lesseramountsof pyroxene,in a muchfiner groundNext it is necessaryto constrainthe depth of emmass;(2) the composition
of the border phase,ne- placementof the pluton.This canbe estimatedfrom
gleetingthe elementsmostmobileduringalteration, the thicknessof the overlyingsedimentaryand vol-

GRF_T BF_R ZONE KIRUNA-TYPE DEPOSITS: ORIGIN

653

becanicrocksasthe top of the andesitcpile is marked phasein an andesiticmeltandthatit crystallized

by a thicksectionof conglomerate
andbrecciainter- tween about 0.5 and 10 kb, then the melt contained
pretedto be relatedto the destruction
of the andesitc 2 percentor lessH20 (Hamiltonet al., 1964; Marsh,
stratovolcano
just prior to, or duringcollapseof, a 1976). Alternatively,if anandesiticmagmaemplaced
major ash-flowcaldera(Hildebrand,1984). Because at 1 kb is saturatedwith respectto H20, thenorthowouldboth crystallize
the Balacheyplutonand,by inference,theRainyLake pyroxeneand clinopyroxene
IntrusiveComplexpredatethe caldera,the thickness prior to plagioclase(Sekineet al., 1979). As stated
of sedimentaryandvolcanicrocksabovethe roof of earlier, the border phaseof the plutoncontains37%
the complexcouldnot havebeen muchgreaterthan plagioclase
phenocrysts,
minor relict clinopyroxene
This
at present,i.e., about 3 km. Geobarometryon mag- and minor altered orthopyroxenephenocrysts.
maticamphiboleswithin the intrusionusingthe em- impliesthat when the magmafirstintruded,it wasa
mixtureof plagioclase,orpirical hornblendegeobarometerof Hammarstrom relativelyhomogeneous
andZen (1985, in press)supportsthisconclusion
be- thopyroxene,
andclinopyroxene
crystals
plusmeltand
causemosthave Altotal
approximatelyequalto 1.
that those parts adjacent to the country rocks
In order to estimatethe originalH20 contentof quenchedto formtheborderphase.Because
the prothe RainyLake magmaonecanusepublishedexper- portionof plagioclase
to clinopyroxene
and orthoimentaldataon andesiticmagmas,but to do this the pyroxeneis very large and becausethe plagioclase
crystallization
sequencemustbe ascertained.
For ex- phenocrysts
do not containinclusions
of either minample,if one knowsthat plagioclaseis the liquidus eral, plagioclasewaslikely the first mineralto crysT,BLE5.
Sample

Mineral Analysesof Magnetite-ApatiteActivities

Kiruna

RLBP-1

3.86

2.80

CrO3

0.02

0.00

FeO
MnO

0.04
0.04

0.88
0.02

RLA- 1

FEMTN

Balach

RLMRP

3.05

3.61

3.13

3.03

0.00

0.02

0.00

0.00

0.78
0.02

0.07
0.00

0.21
0.01

0.09
0.03

Apatite analyses

MgO

0.00

0.00

0.00

0.00

0.00

0.00

CaO
SrO

54.45
0.03

52.30
0.05

52.95
0.04

52.76
0.02

51.86
0.07

54.03
0.03

CIO
PO5

0.04
40.62

0.20
38.95

0.08
40.20

0.33
39.81

0.34
39.61

0.83
40.43

Total

99.10

95.20

97.12

96.62

95.23

98.47

TiO

0.12

Oxide analyses
0.04

0.34

0.96

FeO

31.01

30.28

31.21

31.69

FeO3

68.52

65.91

68.12

67.04

MnO

0.03

0.09

0.07

0.20

CrO
VO

0.01
0.27

0.02
0.05

0.05
0.46

0.00
0.29

Total

99.96

96.39

100.25

100.19

Amphibole analyses
SiO
TiO
AIO
CrO

54.54
0.04
0.65
0.12

51.74
0.09
4.03
0.08

53.86
0.09
0.56
0.08

FeO

16.88

16.29

18.49

MnO

0.13

0.40

0.20

MgO

14.18

14.01

13.12

CaO

12.11

11.92

11.96

NaO
KO
Total

0.00
0.09
98.74

0.00
0.13
98.68

0.00
0.07
98.43

RLBP-1 = magnetite-apatite-actinolite
vein cuttingRainyLake border phase
RLA-1 -- amphibole-apatitevein in upper syenite,Rainy Lake pluton
FEMTN -- magnetite-apatitevein cutting andesiticlava, quarry near Iron Mountain, St. FrancoisMountains,Missouri
Balach-- magnetite-apatitebody adjacentto Balacheypluton
RLMRP -- magnetite-apatite-actinolite
body aboveroof of Rainy Lake pluton

654

ROBERT S. HILDEBRAND

tallize. Anotherline of evidencewhichsuggests

that
plagioclase
wasthe firstliquidusphaseoccursin the
lower monzodioritewhere all other mineralphases
are interstitial to the plagioclasephenocrysts,thus
indicatingthatthe interstitialmineralsgrewafterplagioclasehad accumulated
in the lower part of the intrusion.

It couldbe arguedthat the plagioclase

crystalsdid
not crystallize from the host melt but are artifacts

from the sourceregion,or originatedelsewhereand

were incorporatedinto the melt duringascent.This
wouldinvalidateany estimateof H20 content.However, evenwithout the phenocrysts,the melt fraction
is of andesiticcompositionand thereforemusthave

contained
little H20 (lessthan4%) in orderfor it to

their hostmagmaand settled(Hildebrand,1985b).

The exactrelationshipbetweencrystalsettling,convection,andsolidification
in the RainyLakeIntrusive
Complexwill be dealtwith in a forthcomingreport.
The interplay between crystal settling rates and
solidification
ratescontrolsthe distribution
of plagioclasein the body,but it alonecannotbe responsible
for the overallmineralogicaland chemicallayering;
namely,it is impossibleto createthe overallchemical
trends(Fig. 9), suchasdecreasing
Rb, Ba, and K20
with increasingsilica,by separationof any observed
or even reasonablyhypothesizedmineral phaseor
phases.Moreover, sincethe pluton solidifiedboth
from the baseupward and the roof downward,the
most evolved compositionsshouldoccur some dis-

risewithin 3 km of the surface(Burnham,1979a and tance in from the roof and floor, as observedin some
b). Furthermore,the plagioclasephenocrysts
show maficintrusions,
yet in theRainyLakeplutonthe most
no evidenceof resorptionand are compositionally silicarich samplesoccurin the uppermostpart. In
typical of thoseknownto have crystallizedfrom an- order to understand the chemical variations in the
desiticmelts.There is no evidenceto suggestthat pluton,it is importantto recallthat the composition
they did not crystallizefrom their hostmagma;thus, of the lower monzodioritemainly reflectsthe conthe followingdiscussion
assumes
thatplagioclase
was centrationof plagioclase
phenocrysts
there. The contheliquidusphaseandthatwhenit beganto crystallize sistentbreak for manyelementsat the monzodioritethe melt contained2 percentHsO or less.
monzoniteboundaryis a functionof the decreasein
The distributionof plagioclase
phenocrysts
in the the volumeof plagioclase
phenocrysts
andanincrease
Rainy Lake Intrusive Complex indicatesthat the in perthiteand albite.Bothmineralsclearlyreplace
crystalswere mechanicallyconcentratedin the lower plagioclase
to somedegree.Compositions
of the uppart of the magmachamber.Becausethe size and per syeniteare remarkablefor their high NasO conshapeof the phenocrysts
throughoutthe intrusionare tents, but as stated above such an increase cannot be
not much different from thosein the border phase the resultof somesort of liquid-crystalfractionation
andbecausethere are no overgrowthsotherthanlate involvingany combinationof observedphases.This
albite on the plagioclasephenocrystsin the lower can be visualizedin the systemalbite-anorthite-ormonzodiorite,there waslittle growthof plagioclase thoclase(Fig. 10) where removalof plagioclase,
orbetween the time of intrusion and their concentration
thopyroxene,andclinopyroxenewill forcethe comin the lower portionof the magmachamber.This sug- positionsof residualliquidsto move toward the orgeststhat accumulationof the phenocrystsin the thoclaseapex;yet, samplesof the monzoniteandthe
lowerpartofthechambermusthavebeenfairlyrapid. syeniteplot progressivelycloserto the albite apexas
In fact,the observeddistributionof plagioclase
phe- a nearlydirectfunctionof their heightin the intrusion.
nocrystswithin the body closelyresemblessinusoidal Note that samplesof altered andesiticrocks from
distributionpatternsof olivine found in lava lakes within the albitezoneof the Balacheyhaloplot in the
(Moore and Evans, 1967; Evansand Moore, 1968) sameregion as samplesfrom the syenite.Likewise,
and some mafic sills (Murata and Richter, 1961; in the systemalbite-orthoclase-quartz,
fractionating
Simkin,1967). Suchdistributioncurvesoriginatefrom meltsmove progressivelycloserto the granite minithe interplaybetweencrystalsettlingand solidifica- mumwith time,but suchisnotthe casewith the Rainy
tion both from the roof downwardand the floor up- Lake intrusion(Fig. 10).
ward (B. D. Marsh,pers. commun.;Gray and Crain,
An additionalline of evidencethat clearlyindicates
1969). That is, settlingcrystalscanbe trappedby the that the upperpart of the intrusionis not part of any
downwardsolidification
of the roofif the growthrate fractionation scheme is found in the rare earth eleof the roof zone exceedsthe settlingvelocityof the ment distributionpatterns.The overalltrendsof the
crystals.Sucha mechanismleadsto a decreasein ob- rare earth elementsare clearlynot the resultof any
servedphenocrystdistributionfrom the top down- sortof liquid-crystalfractionationbecausethere islitward. In the caseof the floor of the chamberpro- tle or noincreasein lightrareearthelementsaswould
gressivecrystallizationleadsto an increasein modal be expectedwith any liquid-crystalfractionationinabundanceof settledphenocrysts
upward.The end volvingknownphenocrystic
phases.
In fact,quitethe
resultis a somewhatsinusoidaldistributioncurve.My oppositeoccursasthere is an increasein heavy rare
calculationsindicate that plagioclasephenocrysts earth elementsupward throughthe pluton. Therfoundin the intrusionwere 0.34 g/cm3 denserthan mogravitationaldiffusionas advocatedby Hildreth

GREAT BEAR ZONE KIRUNA-TYPE DEPOSITS: ORIGIN

2OOO

655

I000'

K20

Na20

I00

Rb

.%.

,5,

,5,

/',

500-

,5,

,5,

Sr

CaO

ee

ee

20-'

ee e

o o
MgO

/,,

,5,

40-

Fe20

20'

,5,

300-

20-

Zf'

200-

15-

.,xo
zx

54

56

SiO2 in wt. %

7O

SiO2 in wt. %

FIG. 9. Harker variationdiagramsfor selectedoxides(wt % anhydrous)andtraceelements(ppm)

in the RainyLake (dots)andBalachey(triangles)plutons.Note that in the RainyLake body,assilica
increases,K20, Rb, and Ba decrease.

(1979, 1981) can apparentlyproducean increasein

heavyrare earthelementstowardthe top of a magma
chamberin magmasof certaincomposition,
but at the
sametime it createsa lower zone enriched in light

(<2%) and was crystallizinginterstitial amphiboleso

the exsolutionof volatilesprobablydid not occuruntil

rare earth elements, a feature not seen in the data

floor, the volatiles would be concentrated in the more

the very lateststagesof crystallization.

Becausecrystallizationproceededinward from both the roof and

fromthe RainyLake IntrusiveComplex.Furthermore, central, latest crystallizing,portion of the chamber.

the largevariationin plagioclase
contentin the pluton As a volatilephaseexsolvesfrom a melt it rises,due
shouldproducevariationsin the concentrationsof Eu to the densitycontrast,andflowsupwardthroughthe
(Arth, 1976), a featurefoundonly in the lower part overlyingsolidifiedor mostlysolidifiedmelt. This exof the Rainy Lake complex.
plainswhy only the upper half of the intrusionis inAll of the above,coupledwith the evidenceof re- tenselyaltered.
placement textures within the syenite, suggestthat
Exsolvingvolatilesincreasethe internal pressure
the presentcompositionof the syeniteis notthe result of the chamberand eventually,when the roof rocks
of fractionation but rather resulted from the alteration
can no longercontainsuchpressure,they fail along
of alreadysolid,or nearlysolid,rockby fluids.There fractures oriented normal to the horizontal roof of
canbe little doubtthat a volatile phasewould even- the body(KoideandBhattacharji,1975; Burnhamand
tuallyexsolvefromthe RainyLakemeltbecausesuch Ohmoto, 1980; Knapp and Norton, 1981) and volaa phenomenonis a natural and necessaryresult of the tiles stream out of the intrusion along them. Such
crystallization of calc-alkaline melts emplaced at fractures,filled with magnetite-apatite-actinolite,
are
shallowcrustallevels (Burnham,1979a and b; Burn- commonin the borderphaseof the RainyLake intruham and Ohmoto, 1980). As already shown from sion.The rapid dehydrationwould not only lead to
phaserelations,the melt had a low initial H20 content rapidquenchingof anyremainingmelt but couldeas-

656

ROBERT S. HILDEBRAND

Qz

An

Rainy Lake
border phase
syenite
,

monzonite

1 monzodiorite

Or
FIG. 10. Cationnormafivecompositions
of samplesof the Balacheypluton,the RainyLake Instrusive
Complex,andsomealbitized andesiticrocksfrom within the albite zone of the Balacheypluton (Table
2) plottedin the systemsalbite-anorthite-orthoclase
(Ab-An-Or)and albite-orthoclase-quartz
(Ab-OrQtz). Averageandesiteis an averagecompositionof postcollapse
andesites.

ily create the intensely brecciated zones above its

roof. Similar breccias, also attributed to volatile

streaming, occur above other epizonal intrusions

(Fiske et al., 1963).
The mostintenselyaltered areasof the upper part
of the intrusion,which appearas the albitite dikes
discussed
earlier, are interpretedto representmajor
zonesof fluid migration.Note that the two samples
of the dikelikebodiesare markedlylower in totaliron,
phosphorus,and magnesiumthan any other part of
the intrusion.This suggeststhat the zonesmay have
been a major sourceof the constituentsfor the magnetite-apatite-actinoliteveins and bodies. Massbalancecalculationsshowthat the amountsof iron, magnesium,and phosphorusin the veinsand bodiesare
smallcomparedto thoselost from the intrusion.
Becausethe alteration in the upper part of the
Rainy Lake Intrusive Complex is so similar,both in
compositionand mineralogy,to the albite alteration
halo aroundthe Balacheypluton, it is reasonableto
relatethe two zonesto a commonoriginandto suggest
that for some reason, perhaps related to different
depthsof emplacementor coolinghistories,the albite
zone related to the Rainy Lake is telescopedover the

upperpart of the intrusionitself.Without fluid inclusiondata it is not possibleto determine the compositionof the volatilephaseresponsiblefor the alteration or the controlson mineral zoning, but someinsight can be gained from experimentalwork and
elementalmobility within the systemitself. Experimental work (Killinc and Burnham, 1972) indicates
that in meltswith the compositions
of the RainyLake
andBalacheyplutonsan exsolvingfluid phasewould
be enriched in chlorine; suchfluidsat high temperaturesare able to transportall the necessaryconstituents for the alteration associatedwith the plutons
(Burnham and Ohmoto, 1980). The rare earth element data from the Rainy Lake Intrusive Complex
are compatiblewith this conclusionin that the upper
part of the body is depletedin light rare earth elements,which can be complexedand mobilizedrelative to heavy rare earth elements,by chlorine-dominated fluids(Taylor and Fryer, 1980, 1982, 1983).
Seyfriedand Mottl (1982), althoughworkingon seafloorbasalts,haveshownexperimentallythat the type
of alteration found in the plutonsdescribedhere is
the result of low water to rock ratios,low pH, chlorine-rich, high-temperaturesystems.This contrasts

GREAT BEAR ZONE KIRUNA-TYPE DEPOSITS: ORIGIN

with the typicallow-temperatureandhigh water to

rockratioscommonlyassociated
with subsolidus
hydrothermal convection.
Discussion

The magnetite-apatite-actinolite
bodiesdescribed
here are so much like other worldwide

occurrences

describedin the literaturethat they probablyoriginated by the same mechanism.In the Mesosoicto
Paleogenearc of Chile there are large numbersof
suchdeposits.Bookstrom(1977) relatedthoseat E1
Romeral to alteration

associated with intermediate

epizonalplutons.Fiskeet al. (1963) describedmagnetite-apatite-actinolitebrecciasabovethe roof of the

Tatooshpluton,whichunderliesMount Rainierin the

657

nesiumwere selectivelyremovedfromthe intrusion.

The famous,and somewhatenigmatic,magnetiteapatite flows associatedwith an andesitic volcano
(Frutos J. and Oyarzfn M., 1975; Henriquez and
Martin, 1978) at E1Laco, Chile, deservea brief discussion
becausethey are oftencitedasdirectevidence
that ironphosphatemeltsexistin nature.Mostvisitors
to E1Lacowith whomI havespokenreport that they
look in every way like lava flows.I haveno reasonto
doubt their observations,
yet there are severallines
of evidence that suggestthat they originatedin a
mannersimilarto the depositsdescribedhere:(1) they
areriddiedwith largegascavitiesin whichmagnetite
and hematite crystalssublimated(Henriquez and
Martin, 1978; see alsophotographsin Park, 1961);
(2) thereare meter-highsublimatepipesof magnetite
_ apatitenearthe vents(R.F. Martin, pers.commun.);
(3) they apparentlyextendonly a hundredmetersor
sofromthe vent areas;and(4) the magnetitecontains
only smallamountsof TiO2 (FrutosJ., and Oyarzfin

CascadeRange,and suggested
that they resultedas
volatilesstreamedout of the pluton. Crane (1912)
and, more recently,Pannoand Hood (1983) relate
the well-knownPilot Knob depositsof Missourito
hydrothermalreplacementof ash-flowtuffsand sed- M., 1975). These observationsindicatethat the flows
imentaryrocks,but the heat sourcefor the alteration may have originatedwhen a pluton beneaththe ashasnotbeenfounddueto the flat-lyingnatureof the sociatedandesiticvolcanodehydratedat a rapid rate
volcano-sedimentary
sequences
(Sideset al., 1981). andthe volatilephase,enrichedin iron andphosphoAlthoughthey have been metamorphosed
and de- rus, streamedto the surfaceas a supercriticalfluid.
formedto varyingdegreesafterdeposition,
the classic Upon arrivalat the surfacethe fluid phasemightsubKiruna ores (seereferencesin Introduction)are re- limate magnetiteandapatiteat sucha rate and tem-

markablysimilarto thosedescribedhere in termsof perature that the material could be fluid enoughto
mineralogy,textures,and setting.The commonoc- flowdownslope.
The fluidphasepossibly
reachedthe
currenceof intenselyaltered and albitized rocksin surfacebecausethe plutonmayhavebeenemplaced
the footwallat Kiruna(Geijer, 1915), interpretedin at a shallowerlevel than thosediscussed
in thispaper.
part to represent subvolcanicintrusions,suggests In summary,Kiruna-typedepositsare composed
generationby deutericprocesses
as originallyhy- of magnetite-apatite-actinolite
typically associated
pothesizedby Geijer (1930).
with intermediate compositionplutons and Na20
For a numberof yearsPartk(1973, 1975, 1985) metasomatism.
They tend to occurabovethe roof of
hasarguedthatKiruna-typedeposits
arenotmagmatic the magmabodies,thereby ruling out an originby
in origin but rather are metamorphosed
chemical iron phosphate-silicate
melt immiscibilitybecause
sediments.
His argumentsare basedmainlyon the iron phosphatemeltsare much denserthan silicate
seeminglystrata-bound,or bedded,nature of the de- melts;instead,they areinterpretedto be the products
positsand the presenceof sedimentarystructures; of high-temperature,chlorine-dominated
fluidsgenhowever, similarfeaturesoccurin both the Great Bear erated duringvolatileexsolution.
magmaticzone (Hildebrand,1984) andthe St. FranAlthough several facets of the model presented
coisMountains(PannoandHood, 1983). In bothareas here are poorly understood,particularlythe mechait is clear that suchfeaturesresultfrom replacement nismsof transportanddeposition,it appearsdifficult
to refute the observation that the common link beof older sedimentaryandvolcanicrocks.
J. Hoover Mackin (1968) related magnetite de- tween most Kiruna-typedepositsis an intimate aspositsin the Iron Springsregion of Utah to deuteric sociationwith intermediate plutons, in many cases
alteration of intermediate compositionlaccoliths. related to andesitic volcanoes of continental-arc setMagnetite veinlets within the intrusionsoccur in a tings.Thus, a critical factor in the developmentof
zoneintermediateto the borderphaseandthe central Kiruna-type depositsis that an intermediate pluton
portions.He convincinglydemonstrated
that the iron must containlow enoughH20 so that it can rise to
deposits
outsidethe plutonformedby deutericpro- high crustallevels.Basalticmagmastypicallycontain
cessesrelatedto the coolingof the intrusions.In ac- toolittle HzO to exsolvea fluidphase,but whenthey
cordancewith the modelpresentedhere he showed do,ironlossandalbitizationcanoccur(Eldson,1982).
thatnotonlyiron,butphosphorus,
fluorine,andmag- Althoughmanyrhyoliticmagmas
haveenoughwater,

658

ROBERT S. HILDEBRAND

they commonlydo not contain enoughmagnesium Fiske,R. S., Hopson,C. A., andWaters,A. C., 1963, Geologyof
Mount Rainier National Park, Washington:U.S. Geol. Survey
and phosphorus.
Prof. Paper 444, 93 p.
Acknowledgments

Frietsch,R., 1978, On the magmaticorigin of iron ores of the

Kiruna type: ECON.GEOL.,v. 73, p. 478-485.
Frutos,J.,J., andOyarzhn,M., J., 1975, Tectonicandgeochemical
evidenceconcerning
the genesis
of El Lacomagnetitelavaflow

The generalideaspresentedin this paper developedduringPh.D. studiesat MemorialUniversityof

deposits,Chile: ECON. GEOL., v. 70, p. 988-990.
Newfoundland.
Regionalfieldmappingwassupported Geijer,
P., 1915, Someproblemsin iron ore geologyin Sweden
by the Departmentof Indian and Northern Affairs
and in America: ECON.GEOL.,v. 10, p. 299-329.
andthe GeologicalSurveyof Canada.I am indebted -1930, The iron oresof the Kiruna type: Geographicaldisto G. Andrewsand D. Press(both at MUN) for assis- tribution,geologicalcharacters,andorigin:SverigesGeol.Unders/3kning,ser. C, no. 367, 39 p.
tancewith the whole-rockanalyses,
E. Hurdle (GSC)
Geijer, P., and Odman,O. H., 1974, The emplacement
of the
for themodalanalyses,
andthemanagement
andstaff Kiruna iron ores and related deposits:SverigesGeol. Underof Terra MiningandExplorationLtd. for allowingacs/3kning,ser. C, no. 700, 48 p.
cess,oftenwith little or no notice,to their properties Gray, N. L., and Crain, I. K., 1969, Crystal settlingin sills:A
modelfor suspension
settling:CanadianJour.Earth Sci., v. 6,

in the Camsell River area. I have had valuable and

p. 1211-1216.
informativediscussions
overthe yearswith B. J. Fryer, Hamilton,
D. L., Burnham, C. W., and Osborn, E. F., 1964, The
D. F. Strong,R. F. Martin, S. A. Bowring,P. F. Hoffsolubilityof waterandthe effectsof oxygenfugacityandwater
man,R. P. Taylor, R. Tirrul, andS. M. Roscoe.A specontenton crystallizationof maficmagmas:Jour.Petrology,v.
5, p. 21-39.
cial thanksto BruceD. Marshfor explainingthe significanceof the phenocrystdistributioncurve,a fea- Hammarstrom,J. M., and Zen, E., 1985, An empiricalequation
for igneouscalcicamphibolegeobarometrylabs.]:Geol. Soc.
ture I ponderedover for manyyears.Review of the
America Abstractswith Programs,v. 17, p. 602.
text by I. G. Reichenbach,P. F. Hoffman,M. B. Lam- -in press,Aluminumin hornblende:An empiricaligneous
geobarometer:Am. Minerologist,
bert, D. F. Strong,R. F. Martin, and J. G. Arth has
Hegemann,F., andAlbrecht,F., 1954,Zur Geochemieoxydischer
greatlyimprovedits substance
andform.

July 15, November25, 1985

REFERENCES

Anonymous,1977, Gilbert LaBinc 'man who found the atomic

age' (obituary):Toronto Star, June 9, p. A16.
Arth, J. G., 1976, Behaviorof trace elementsduring magmatic
processes--Asummaryof theoreticalmodelsand their applications:U.S. Geol. SurveyJour. Research,v. 4, p. 41-47.
Badham,J.P. N., and Morton, R. D., 1976, Magnetite-apatite
intrusionsandcalc-alkalinemagmatism,CamsellRiver, N.W.T.:
CanadianJour. Earth Sci., v. 13, p. 348-354.
Bookstrom,A. A., 1977, The magnetitedepositsof El Romeral,
Chile: ECON.GEOL., v. 72, p. 1101-1130.
Bowring,S. A., and Van Schmus,R. A., in press,Wopmay orogen:
An exampleof rapid plate tectonics:Tectonics.
Burnham,C. W., 1979a, Magmasand hydrothermalfluids, in
Barnes,H. L., ed., Geochemistryof hydrothermalore deposits:
New York, JohnWiley and Sons,p. 71-136.
-1979b, The importanceof volatile constituents,in Yoder,
H. S., ed., The evolutionof the igneousrocks:Princeton,New
Jersey,Princeton Univ. Press,p. 439-482.
Burnham, C. W., and Davis, N. F., 1974, The role of H20 in
silicatemelts: II. Thermodynamicand phaserelationsin the
systemNaAISiaOs-H20to 10 kilobars,700 to 1100C: Am.
Jour. Sci., v. 274, p. 902-940.
Burnham,C. W., and Ohmoto, H., 1980, Late-stageprocessesof

felsicmagmatism:Soc.Mining GeologistsJapanSpec.Issue8,
p. 1-11.
Crane, G. W., 1912, The iron ores of Missouri: Missouri Bur.

GeologyMines, v. 10, 2nd ser., 434 p.

Daly, R. A., 1915, Origin of the iron ores at Kiruna: Vetensk.O

Prkt.Undersok.
I Lappland,
v. 5, 31 p.
Elsdon,R., 1982, Autometasomaticalterationof gabbro,Kap EdvardHolm intrusivecomplex,eastGreenland:Mineralog.Mag.,
v. 45, p. 219-225.

Evans,B. W., and Moore, J. G., 1968, Mineralogy as a function

of depth in the prehistoricMakaopuhitholeiitic lava lake, Hawaii: Contr. Mineralogy Petrology, v. 17, p. 85-115.

Eisenerze:Chemic der Erde, v. 17, p. 81-103.

Helz, R. T., 1979, Alkaliexchange
betweenhornblendeandmelt:
A temperaturesensitivereaction:Am. Mineralogist,v. 64, p.
953-965.

--

1982, Phaserelationsandcompositions
producedin studies
of the meltingbehaviorof rocks:Mineralog.Soc.America:Reviews in Mineralogy, v. 9B, p. 279-346.
Henriquez, F., and Martin, R. F., 1978, Crystal-growthtextures
in magnetiteflowsandfeederdykes,EL Laco,Chile: Canadian
Mineralogist,v. 16, p. 581-589.

Hildebrand,R. S., 1981, Early ProterozoicLaBincGroup of

Wopmayorogen:Remnantof a continentalvolcanicarcdevelopedduringobliqueconvergence:
CanadaGeol. SurveyPaper
--

--

81-10, p. 133-156.
1982a,An earlyProterozoiccontinentalvolcanicarcat Great
Bear Lake, Northwest Territories: Unpub. Ph.D. thesis, St.

John's,MemorialUniv. of Newfoundland,237 p.
1982b,Geology,EchoBay-MacAlpine
Channelarea:Canada
Geol. Survey Map 1546A.

--

1984, Geologyof the RainyLake and White Eagle Falls

mapareas,Districtof Mackenzie:EarlyProterozoiccauldrons,
stratovolcanoes
and subvolcanicplutons:CanadaGeol. Survey
Paper 83-20, 42 p.

--

1985a, Geology,Rainy Lake-White Eagle Falls map areas,

District of Mackenzie: CanadaGeol. Survey Map 1589A.

--

1985b, Origin of mineralogicalzoningin an intermediate

subvolcanic
plutonby crystalsettlingand conductivecooling
[abs]:Geol. Soc. America Abstractswith Programs,v. 17, p.
610.

Hildebrand,R. S., and Bowring,S. A., 1984, A non-extensional

modelfor the originof continentalintra-arcdepressions,
with
a Proterozoicexamplefrom Wopmayorogen,NorthwestTerritories, Canada:Geology, v. 12, p. 73-77.
Hildebrand,R. S., Bowring,S. A., Steer,M. E., andVan Schmus,
W. R., 1983, GeologyandU-Pbgeochronology
of partsof the
Leith Peninsulaand Riviere Grandin map areas,District of
Mackenzie: CanadaGeol. SurveyPaper 83-1A, p. 329-342.
Hildreth,W., 1979, The BishopTuff:Evidencefor the originof
compositional
zonationin silicicmagmachambers:Geol. Soc.
America Spec. Paper 180, p. 43-75.

GREAT BEAR ZONE KIRUNA-TYPE DEPOSITS: ORIGIN

659

--

1981, Gradientsin silicicmagmachambers:Implications Partk,T., 1973, Rare earthsin the apatiteiron oresof Lappland
together with some data about the Sr, Th and U content of
10153-10192.
theseores:ECON.GEOL.,v. 68, p. 210-221.
Hirano, S., and Somiya,S., 1976, Hydrothermalcrystalgrowth -1975, Kiruna iron oresare not "intrusive-magmaticoresof
of magnetite
in thepresence
of hydrogen:
Jour.CrystalGrowth,
the Kiruna type": ECON.GEOL.,v. 70, p. 1242-1258.
v. 35, p. 273-278.
-1985, Phosphorusin different types of ore, sulfidesin the
Hoffman,P. F., 1980, Wopmayorogen:A Wilsoncycleof early
iron deposits,and the type and origin of oresat Kiruna:ECON.
Proterozoicagein the northwestof the CanadianShield:Geol.
GEOL., v. 80, p. 646-665.
Assoc.CanadaSpec.Paper 20, p. 523-549.
Park, C. F., 1961, A magnetite"flow" in northern Chile: ECON.
Hoffman,P. F., and Bowring,S. A., 1984, Short-lived1.9 Ga
GEOL., v. 56, p. 431-441.
continentalmarginanditsdestruction,
Wopmayorogen,north- -1972, The iron ore depositsof the Pacific basin:ECOa.
west Canada:Geology,v. 12, p. 68-72.
GEOL.,v. 67, p. 339-349.
Hoffman,P. F., andMcGlynn,J. C., 1977, GreatBearbatholith: Philpotts,A. R., 1967, Origin of certain iron-titaniumoxide and
A volcano-plutonic
depression:
Geol.Assoc.
CanadaSpec.Paper
apatiterocks:ECON.GEOL.,v. 62, p. 303-315.

forlithospheric
magmatism:
Jour.Geophys.
Research,
v. 86, p.

16, p. 170-192.

Hoffman,P. F., Bell,I. R., andTirrul, R., 1976,SloanRivermap

area (86K), Great Bear Lake, District of Mackenzie: Canada

Sekine, T., Katsura, T., and Aramaki, S., 1979, Water saturated
phaserelationsof someandesiteswith applicationto the estimationof the initial temperatureandwater pressureat the time
of eruption: Geochim. et Cosmochim.Acta, v. 43, p. 1367-

Geol. SurveyPaper76-1A, p. 353-358.

1396.
Kilinc,I. A., andBurnham,C. W., 1972, Partitioningof chloride
betweena silicatemeltandcoexisting
aqueous
phasefrom2 to Seyfried,W. E., Jr., andMottl, M. J., 1982, Hydrothermalalter8 kilobars:ECON.GEOL.,v. 67, p. 231-235.
ationof basaltby seawaterdominatedconditions:Geochim.et
Cosmochim.Acta, v. 46, p. 985-1002.
Knapp,R. B., andNorton,D., 1981,Preliminarynumericalanalysis
of processes
relatedto magmacrystallization
andstress
evolution Sides,J. R., Bickford,M. E., Shuster,R. D., andNusbaum,R. L.,
1981, Calderas in the Precambrian terrane of the St. Francois
in coolingplutonicenvironments:
Am.Jour.Sci.,v. 281, p. 35-

68.
Mountains,southeasternMissouri:Jour.Geophys.Research,v.
Koide, H., and Bhattacharji,S., 1975, Formationof fractures
86, p. 10349-10364.
aroundmagmaticintrusionsand their role in ore localization: Simkin,T., 1967, Flow differentiationin the picriticsillsof North
ECON. GEOL., v. 70, p. 781-799.
Skye,in Wyllie, P. J., ed., Ultramaficand related rocks;New
Landergren,S., 1948, On the geochemistry
of Swedishiron ores
York, John-WileyandSons,p. 64-69.
andassociated
rocks:Sveriges
Geol.Unders6kning,
ser.C, no. Smellie,J. A. T., 1980, Kiruna-typeiron oresfrom Mertainen and
496, 182 p.
Painirovalabs.I: Geol. F6ren. StockholmF6rh., v. 102, p. 280.
Lundberg,B., andSmellie,J.A. T., 1979,PainirovaandMertainen Snyder,F. G., 1969, Precambrian
irondepositsin Missouri:ECOI.
ironores:Two depositsof the Kirunaironore type in northern
GEOL.MON. 4, p. 231-238.
Sweden:ECON.GEOL.,v. 74, p. 1131-1152.
Taylor, R. P., andFryer, B. J., 1980, Multiple-stagehydrothermal
Mackin,J.H., 1968,Ironoredeposits
oftheIronSprings
district,
alterationin porphyrycoppersystemsin northernTurkey: The
southwestern
Utah, in Ridge,J. D., ed., Ore depositsof the
temporal interplay of potassic,propylitic, and phyllic fluids:
UnitedStates,1933-1967 (Graton-Sales
vol.):New York,Am.
CanadianJour. Earth Sci., v. 17, p. 901-926.
Inst. Mining, Metall. PetroleumEngineers,v. 2, p. 992-1019.
-1982, Rare earth element geochemistryas an aid to interMarsh, B. D., 1976, Some Aleutian andesites:Their nature and
preting hydrothermal ore deposits,in Evans,A.M., ed., Mesource:Jour. Geology,v. 84, p. 27-45.
tallizationassociated
with acidmagmatism:
London,JohnWiley,

Moore, J. G., and Evans, B. W., 1967, The role of olivine in the

385 p.

crystallization
of the prehistoricMakaopuhitholeiiticlavalake, -1983, Rare earth elementlithogeochemistry
of granitoid
mineraldeposits:CanadianMining Metall. Bull., v. 76, no. 12,
Hawaii: Contr. MineralogyPetrology,v. 15, p. 202-223.
p. 74-84.
Murata,K. J., andRichter,D. H., 1961, Magmaticdifferentiation
in the Uwekahuna
laccolith,Kilaueacaldera,Hawaii:Jour.Pe- Tirrul, R., 1976, The geologyof the RainyLake IgneousComplex,
trology, v. 2, p. 424-437.
District of Mackenzie, Northwest Territories: Unpub. B.Sc.
thesis,Kingston,Queen'sUniv., 115 p.
Panno,S. V., andHood, W. C., 1983, Volcanicstratigraphy
of
the Pilot Knob iron deposits,Iron County, Missouri:ECON. Wager, L. R., Brown, G. M., and Wadsworth,W. J., 1960, Types
GEOL.,v. 78, p. 972-982.
of igneouscumulates:Jour.Petrology,v. 1, p. 73-85.