Relationship Between Rock Type, Metamorphic Grade

Canadian Mineralogist Vol. 25, pp. 485-498(1987)

RELATIONSHIPBETWEEN ROCK TYPE, METAMORPHIC GRADE, AND FLUID- PHASECOMPOSITION IN THE GRENVILLESUPERGROUP, LIMERICKTOWNSHIP, ONTARIO

P. JAMES LgANDERSON Depaftmentof Geologyand GeologicalEngineering, Colorado School of Mines, Goldm, Colorado8M01, U.S.A.

JAMES L. MUNOZ Depafimentof GeologicalSciences, University of Colorado,Boulder, Colorado80309-0250, U.S.A,

ABSTRACT entre535 et 550'C dansles marbres et 402et 543"Cou plus dans les lithologies pauvres en carbonate. La valerir de The srudy area in Limerick Township in the southern X(CO) seraitentre 0,45 et 0.55 dansles marbreset moins Grenville Province consistsof a sequenceof late Proterozoic de 0. l0 dansles milieux pauvresen carbonate.La relatlon metabasalts,noncalcareous and calcareousquartzofeld- entre degr€de m6tamorphismeet lithologie refldte la dimi spatlfc granofels,and marblesintruded by two gabbro com- nution du X(CO) du marbre vers les unit€s moins riches plexes.During the Grenville Orogeny,metamorphism of en calcaire.Cette relation pourrait 6tre attribude l) d un marble producedassemblages that are below the titanite, tamponnageinterne de la phasefluide par lesr6actions iso- actinolite - calcite,and hornblende- K-feldsparisograds, gradiquesi des valeurs 6levdesde X(CO) dans les mar- but above the hornblende - K-feldspar isograd in the bres et des valeurs inf6rieures dans les unit6s pauvres en quartzofeldspathicunits and amphibolitss. Calcite-dolomite carbonate, ou 2) d un tamponnagedes fluides i desvaleurs and two-feldspargeothennometry, and calculationof the faibles de X(CO) par les r€actionsisogradiques suite i displacementof intersectingequilibria due to solid-solution, l'infiltration de fluides aqueux au travers des unit€s pau- bracket equilibrium temperaturesbetween 535 and 550'C vres ou sanscarbonate. in the marbles and between402 and 543'C or higher in rocks with low carbonatecontents. X(CO) is estimated Mots-cles:isogrades, marbre, granofels,intersection d'6qui- between0.45 and 0.55 in the marblesand lessthan 0.10 libres, degr6 de m€tamorphisme,composition de la in the carbonate-poorunits. The relationshipbetween meta- phaseaqueuse interstitielle, Grenville, canton de Lime- morphic grade and rock type is due to a decreasein rick, Ontario. X(CO) from the marble to the lesscalcareous units. This relationship may be due to: 1) internal buffering of the fluid INTRODUCTION by the isogradic reactionsto high X(CO, valuesin the marbles and low X(CO) values in the carbonate-poor Experimental investigations (e.g., Skippen 1971, units, or 2) buffering of the fluids at low X(CO, by rhe 1974, Slaughter et al,1975) and theoretical studies isogradic reactions during infiltration of H2O-rich fluids (ag., Greenwood 1962, 1967,1975) of mixed-volatile through the carbonate-deficient and carbonate-absent equilibria have made possible the interpretation of rocks. metamorphosed, interlayered calcareous, pelitic and quartzofeldspathic sedimentary rocks. The tempe- Keywords: isograds, marble, granofels, intersecting not equilibria, metamorphic grade, pore-fluid composition, rature of mixed-volatile equilibria is dependent Grenville, Limerick Township, Ontario. only on pressure but on fluid composition, especially the ratio of H2O to COr, or X(CO). "High-grade" Souvarnr areas defined by mixed-volatile isograds, therefore, do not necessarily indicate regions of high tempera- Dans le canton de Limerick (Ontario), dansla partie sud ture, unlike dehydration reactions in nongraphitic de la province du Grenville, une s6quenced'dge Protdro- pelitic rocks. Mixed-volatile isograds may, and often zoique tardif de mdtabasaltes, granofels quartzo- do, reflect variations in X(CO). feldspathiquesavec calcaires ou non, et marbres est recou- Carmichael (1970) found that isograds based on pde par deux complexesgabbroi'ques. Pendant I'orogendse mixed-volatile reactions intersect one another in Lake grenvillienne,le m6tamorphismedes marbres a donn6 des Township in the southern Grenville Province, Onta- qui assemblages setrouvent au dessousdes isogrades tita- rio. This was interpreted to be the resuh of migra- nite, actinote- calcite,et hornblende- feldspathpotassi tion of H2O from the Whetstone granite into the que, mais au dessusde ce dernierisograde dans les unitds Kerrick (1976) described a simi- quartzofeldspathiqueset lesamphibolites. La gdothermo- wallrocks. Moore & m6trie, utilisant les dquilibres impliquant calcite et dolo- lar situation in the Alta aureole in Utah. mite et les deux feldspaths, ainsi que le calcul des ddplace- Locally, in southeastern Maine, interbedded ments des intersectionsdes r6actions dus aux solutions pelites, micaceous quartzites and calcareous rocks solidesprdsentes, indiquent des temp6raturesd'6quilibre strike parallel to the direction of the thermal gra- 485

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 486 THE CANADIAN MINERALOGIST

45C)52'

I Mil-es N GRENVILLEPROVTNCE 0 20 ffn PALEOZOTC J_ 0km30 WTF'ELO AREA METAMQRPHIC- BOUNDARIES FIc. l. The field area is located in Limerick Township, southern Ontario, 100 km northeastof Toronto (after Lumbers 1964).The field area of Allen (1976)is located in Tudor Township, immediatelysouth of Limerick Township, and that of Car- michael(1970) in Lake Township, immediatelysouthwest of Limerick Township.

dient. Calc-silicateisograds are extremelyirregular eousquafizofeldspathic granofels, and marble @ig. and form local, high-gradeareas related to centres 2). The Brinklow Lake and Ormsby Junction of intrusion, and exhibit local cross-cuttingrelations- amphibolite belts consistpredominantly of massive hips. Ferry (1976, 1980)interpreted thesepatterns amphibolites of assumedvolcanic origin. Volumetri- asthe result of a high flux of heat relatedto the plu- cally minor, calcareous,locally schistoseamphibo- tons, introduction of HrO from the intrusive lite may haveformed through the reworking of vol- bodies, dehydration of the pelitic rocks, and inter- canic rocks and introduction of clastic material. The nal control of the fugacities of H2O and CO2by the calcareoushorizons are in greatestabundance in the mineral assemblages. Ormsby Junction amphibolite belt. An ageof 1310 In Grenvillemarble and calcareousmetasedimen- Ma was obtained by Silver & Lumbers (1966) on tary rocks in Tudor Township, southern Ontario zircon from the mafic Tudor metavolcanicrocks (Fig. l), Allen (1976)observed an increasein meta- south of the study area.The similarity of the Tudor morphic grade toward a seriesof metamorphosed volcanicrocks and associatedrock units to thosein tonalite, trondhjemite, granodiorite, and gabbro Limerick Township gives an approximate age of the intrusive bodies. This was attributed to a decrease rocks in the study area. in X(CO) of the intergranular gas-phasetoward the Rusty schistssurround the amphibolite belts and igneous contact, i.e., the relic of a pre-Grenville, occur as discontinuouslenses in them and in the other contact-metamorphicaureole. units. Small, exhalativemassive-sulfide deposits con- taining pyrite, pyrrhotite, sphaleriteand chalcopyrite GENERALGsot-ocv occur in the rusty schistsadjacent to the amphibolites. Their composition, clastic host-rocks, and A sequenceof Grenville amphibolitesand quartzo- associationwith mafic volcanic rocks suggestthat feldspathic and calcareousrocks surrounds the Tha- they are Besshi-typedeposits (Franklin et al. l98l). net anil Jocko Lake gabbro complexes(Lumbers Furthermore, it is likely that the volcanic rocks were 1967,1969, LeAnderson 198a, b) approximately22 the heat sourcefor the deposits.This indicatesthat km south of Bancroft, Ontario, in Limerick Town- the amphibolites are the oldest unit in the field area. ship (Figs. l, 2). The areasstudied by Carmichael The quartzofeldspathic granofels is generally (1970) and Allen (1976) lie to the southwestand arkosic in compositionand is assumedto represent south, respectively(Fig. 1). clastic sediments. Calcereous quaftzofeldspathic The most commonly observedsequence of rock granofels contains similar assemblagesin addition to units is amphibolite, rusty (sulfidic) quartzofeld- up to 10Vocalcite and dolomite. Lensesand layers spathicschist, quartzofeldspathic granofels, calcar- of each rock type are found in the others.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 ROCK TYPE, METAMORPHIC GRADE AND FLUID-PHASE COMPOSITION IN THE GRENVILLE >":i,"ic:)i ="'7?!3-;{ j; .1',,',.j;ct6i1- tt" '',,-.i z"- Loke, ^"- \oor- Compler w "-\ ( "' -. iiii:l-.:"",-

.:T'F{ t,l G.:-.)..it])'T::i " )-i=rt: o.o 05 r.o a-"1; r-;r."";,=.'\i :-.', =, gYi [,,j3.:"-i "i lm

Ftc.2. Generalizedgeological map of the study area (after LeAnderson 1978a).Symbols: rs rusty schists,qf quartzofeldspathicgranofels, cqf calcareousquartzofeldspathic granofels, m marble, cs calcsilicate.

Marble makesup the youngestof the sedimentary enite and hornblendemetagabbro (Lumbers 1967); units and is presentin four areas(Fig. 2). The miner- its ageis unknown but assumedto be approximately alogy, color and texture are a function of meta- that of the Thanetcomplex. A premetamorphicdate morphic grade.Below the actinolite-calciteisograd for the Thanet complex (Berger& York 1981)and the marble is grey and commonly containschlorite, methmorphictextures in both indicatethat the com- biotite, calcite, dolomite, quartz, plagioclase, plexeswere emplacedprior to the Grenville meta- titanite, pyrrhotite, sphalerite,rutile, graphite and, morphic event. locally, epidote,ilmenite and pyrite. Calcic amphi- Numerousgranitic pegmatitesoccur in the Ormsby bole, rangingfrom colorlesstremolite to dark green Junction Amphibolite Belt immediately northeast of actinolite, appearsat the actinolite-calciteisograd. Highway 62. No other intrusions occurred. Above this isogradthe marblesare mostly very fine During the Grenville orogeny, the Grenville Super- grained and grey-green. Hornblende, epidote, group was buried, deformed, and metamorphosed. microcline and diopside appearwith increasinggrade The metamorphicgrade is in the upper greenschist at the expenseof chlorite, dolomite and biotite facies near Madoc (Fie. l) and increasesto upper (mineral compositions are given in LeAnderson amphibolite facies to the west, north and east. 1978a).Layers and lensesof sulfidic schist, quartzo- Paleozoicrocks are exposedto the south. Tempera- feldspathic and calcareous quartzofeldspathic tures calculatedusing the calcite-dolomitegeother- granofels are ubiquitous. A calc-silicatehorizon, mometerfor rocks in Tudor Township,south of the intermediatein compositionbetween the calcareous study area,range from 400 to 550'C (Allen 1970, and quartzofeldspathicgranofels and marble, is located from 475to 630'C to the north, in the Bancroft area in the southwesternpart of the area (Fig. 2). (Sobol& Essene1973). The northeasterncorner of the 1200Ma (Berger An estimation of load pressureat the time of & York l98l) Thanet gabbro complexis exposedin metamorphismis made from the distribution of the the map area(Fig. 2). Basedon gradationalinternal aluminum silicate minerals and the AlrSiO, phase contactsand intrusions of hybrid magma,Lumbers diagram of Holdaway (1971). Aluminum silicate (1967)suggested that the majority of the intrusive minerals were not observedin the study area, but units were emplacedover a short period of time. The andalusite is found approximately l0 km to the Jocko Lake complex(Fig. 2) consistsof metapyrox- north-northeast(Lumbers 1967).Andalusite is also

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 488 THE CANADIAN MINERALOGIST

found 16km to the southwestin centralLake Town- rent of 100 /..A. Raw data were reducedusing the ship, 8 km east of the kyanite-sillimanite isograd Multiple Least Squarescomputer program (written (D.M. Carmichael,pers. corrun., 1973).Pressure in by Northern Scientific for the PDP-ll computer) central Lake Township was probably closeto that with the methodsand constantsof Bence& Albee of the aluminum silicatetriple point. In the absence (1968). of more definitive data, it is assumedthat the pres- Oxide totals of 98-10290were considered accept- sure of the study areawas lessthan, but perhapsclose able for analysesof anhydrous minerals. Acceptable to, that defined by the triple point. totals for the hydrous mineralswere determinedby The pressureof the triple point determinedby subtracting the averagewater content using data Holdaway(1971) is 3.76 t 0.3 kbar at a tempera- listed by Deer et al. (1962a,b, 1963).The reprodu- ture of 501 t 20'C. Calorimetric work (Anderson cibilities of the mole fractions of Al, Mg, Ca, Na et al, lW7) and a thermodynamic analysisof the alu- and K in their respectivesites are lessthan or equal minum silicatesystem (Day & Kumin 1980)support to 0.01 for the minerals studied, except for Ca in Holdaway's data, as doesthe assemblageepidote- epidote (X&: a 0.02; LeAnderson 1978a).The sillimanite found in lhe ReadingProng, N.Y., area results of the chemicalanalyses, activiry models, and by Murray (1976). The pressurein the map area, calculatedactivities are given in LeAnderson (1978a, therefore, is assumednot to have exceeded4 kbar, le8t). the figure used in the calculations. ISoGRADS ANALYTTcALTucnNreurs Five isogradswere mappedin the calcareousrocks Mineral analyseswere obtained at Queen'sUniver- based on model reactions ("isoreactiongrads" of sity using an ARL-EMX electron microprobe Winkler 1974).Inorder of increasinggrade, they are: equipped with an energy-dispersionsystem and a (l) rutile + calcite+ quartz = titanite + COz PDP-I I computer. The systemwas operatedat an (2) 5 dolomite + 8 quartz+ H2O = acceleratingvoltage of 15 kV and an emissioncur- tremolite + 3 calcite+ 7 CO"

:",!2"/2:!,(i;1";f,Q)'-ii .i1'1;""_11-';o //" Loke \\ i Compter

""\ V.o, cql )L/p -1fl_$i:--"" / (Ju

- - - o.= \iffir:;,)'srrr s s =7'- ) 2\ rz 0 -(z- a 1r.-.. /% I l (.t..:;.:',1.,1'3)]:"1::"t,"io,l::2Z-:ir' o.o ._; r.o * i{S,1"=,ir"-.\i :'-,, J

Frc.3. The titanite isograd, basedon the model titanite-producingreaction (l) (seetext). A plus indicatesthe reactant assemblage,a circle, the product assemblage,and a plus and circle, sampleswith both tlte reactant and product assemblages.Hachures are drawn on the high-gradeside of the isograd.For clarity, isogradsare not drawn around [he two occurrencesof the reactant assemblagein the marble.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 ROCK TYPE. METAMORPHIC GRADE AND FLUID.PHASE COMPOSITION IN THE GRENVILLE 489

(3) 5 phlogopite + 6 calcite* 24 quarrz: fraction and mole fraction versas distribution- 3 tremolite + 5 K-feldspar + 6 CO2+ 2 }J2O coefficient diagrams,suggest that at leaston the scale (4) tremolite + 3 calcite* 2 quartz : of a thin section,equilibrium betweenminerals was 5 diopside+3CO2+H2O closely approached(LeAnderson 1978a).Thus we (5) 3 tremolite + 6 calcite+ K-feldspar : assume that these assemblagesrepresent mosaic 12 diopside+ phlogopite + 6 CO2+ 2 }t2O. equilibrium (Osberg l97l); this hypothesisis sup- ported by the presenceof mosaic textures (LeAnder- Although most of the minerals are membersof son 1978a). solid-solution series,the model reactionsmay be used The titanite isograd marks the appearanceof as a basisfor the isogradsprovided that the phases titanite in rutile-bearing,siliceous, calcareous rocks. in the model reactionsare membersof the samesolid- The isograd occurs in the northwest corner of the solution series. The model reactions are labeled map area and around two samplesin the west-central accordingto the product phases,and the isograds part of the area (Fig. 3). There are only five occur- by the product minerals. rencesof the assemblagerutile - calcite - quartz, but An assemblageis used as a constraint on an numerousoccurrences of the four-phaseassemblage. isograd only if the minerals involved in the model This may reflect buffering of the gasphase, partici- reactionwere in equilibrium during metamorphism. pation of plagioclasein the reaction (the titanite con- The assumptionof equilibrium appliesequally to cal- tains between 1.7 and 4,2t/o N2O3; LeAnderson culationsthat usechemical compositions of coexist- .L978a),or both. ing minerals. The range of composition for each The actinolite - calciteisograd, based on the model mineral analyzedin single samplesin this study is tremolite - calcitereaction (2), definesfour low-grade generallygreater than the reproducibilityof the ana- areasin the northwesternand southeasterncorners' lyses.This suggeststhat adjacentgrains of the same and the central and west-centralparts of the map area mineral, e.9., hornblende-hornblende, did not (Fig. ). It coincideswith the titanite isograd at equilibratewith one another. Nevertheless,plots of severallocalities. Calcite inclusionsin actinolite at analytical means of coexisting minerals, e.g., this grade provide supportingevidence that aclino- diopside-hornblende,on mole fraction versusmole lite was producedby this reaction or one similar to

L":f,Ki-o=i

i;p|":li{liI a Jocko- -'' - ,,,! s .if*\\\n\\ d \- "rt , -,-- 7[ i: Loke .i I \o=o- Complgx

'\ I Lh ^-/ et[\ * -\f\rr-L, cor ?!'-)" ;-:'"

rhoneico1c1e1-: l,l .

Frc.4. The actinolite-calciteisograd, basedon the model tremolite- and calcite-producingreaction (2). SeeFigure 3 for explanation.The numbersrefer to samplesfor which temperatureswere calculated using the calcite- dolomite geothermometer(see Table 4).

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 490 THE CANADIAN MINERAI,OGIST

it. The numerous occurrencesof the assemblage The temperatureof isogradicreactions depends in quartz - dolomite - actinolite - calcitemay be the part on mineral composition.The effect of mineral result of buffering of the gasphase by the reaction, compositionon the temperatureof the hornblende the presenceof Fe and Al in actinolite, or both. The - K-feldsparand diopsidereactions was determined Fe and Al may have beencontributed by dolomite, by calculation of the displacementof the correspond- plagioclase,or chlorite (LeAnderson 1978a), ing ph-cc-qtz-tr-kf and tr-cc-qtz-di equilibria An isograd basedon the model reaction2 zoisite (mineral-phaseabbreviations are given in Table l) + CO2: calcite+ 3 anorthite + H2O could not be at arLX(CO) of 0.10 and a pressureof 4 kbar using mapped. It is representedin the field by the assem- the activity models listed in Table l. The tempera- blage epidote - plagioclase- calcite. Below the ture displacementof the equilibriais correlatedwith actinolite - calciteisograd, however, the high-grade the magnesium content of hornblende (Xffi*nt'. [high temperature and high X(CO)] assemblage Table 2) because:hornblende exhibits the greatest plagioclase- calciteis very common,whereas above variation in compositionof the mineralsin the diop- this isograd, epidote - calcite predominates. side reaction (isograd);the magnesiumconcentration The hornblende- K-feldspar isogradis basedon has the greatesteffect on a,c;--ptt' and the mag- the model tremolite - K-feldspar reaction (3). The nesiumcontent is relatedto the color of the mineral, isogradlies between the actinolite- calciteand diop- a feature readily observablein hand specimenand side isograds and defines high-grade areas in the thin section. Brinklow Lake amphibolite belt and around the The displacement of the ph-cc-qtz-tr-kf Thanet complex(Fig. 5). It doesnot occur around equilibrium was calculatedusing the experimentally the OrmsbyJunction Amphibolite Belt or the Jocko determinedequilibrium

- .\ ,{ _\-r a.l cf (" _* =\ Irld^4, crr ,ff -jjo ) :" fuo",,..-*;-"T'<{ir; -= =-/.-- -- rrl-srrt -s(l- )2a2t,, ).=.r.... rhonst coqo,lrlj lrl j<_J.-*-)i.,\ ) 2-l<;: o.o 05 t.o *i"S,\,ir"=..\I :i,"! | 1.; ;_: 1.1=. i km

Ftc.5. The hornblende- K-feldsparisograd, based on the model tremolite- and K-feldspar-producingreaction (3). See Figure 3 for explanation.The numbersrefer to samplesused in the calculationof the displacementof the ph-cc- qtz-tr-kf equilibrium and intersectionof the ph-cc-qtz-tr-kf, tr-cc-qtz-di and cc-an-zo equilibria (Iables 2, 5).

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 ROCK TYPE, METAMORPHIC GRADE AND FLUID.PHASE COMPOSITION IN THE GRENVILLE 49I

(LeAnderson1978a, l98l), thereis little or no dis- TABLE2, CALCULATEDDISPLACEI'{ENT OF PH-CC-QTZ-TR-KF ANDTR-CC-QTZ.DI EQU1LIBRIA DUE TO SOLID placementat an ,ft*eh of 0.50, byt th9 curve is soLurI0N displacedto lower teirperaturesat ffnh lessthan 0.50, and to higher temperalures.atX!,f;*eh higher ph-cc-qtz-tr-kf equiI lbrl un than 0.50. Differencesin )fr*mnn of abbut 0.l0 or Sample "Ca-amph. D'lsplacement lessin an areawith evena shallowthermal gradient Nunber "Mg at X(C02)" 0.1 would have little effect on the isograd pattern. 389 0.30 -4ooc Nevertheless,a seriesof hornblende - K-feldspar 91 0.s1 ooc isogradsbased on specific compositions of horn- L 29 0.58 +1ooc blendewould reducethe effectsof compositionon tr-cc-qtz-d'i equllibrlum the isograd patterns and assistin interpretation of SanpI e vCa-mph. spatial variations in temperatureand fluid compo- Nunber ^l4s oisplacenentat X(CoZ)- 0.1 sition. Using samplescoincident with the mappedpos! 429 +Looc tion of the hornblende- K-feldspar isograd, Car- 428 0.58 *2ooc t3ooc michael (1970)calculated the following reaction: 398 0.52 L2L 0.49 +4ooc 139 0.48 +4ooc biotite + calcite+ quartz + epidote* plagioclase: 434 0.45 +4ooc hornblende+ K-feldspar + titanite + magnetite(or hematite). Samplelocatlons and rock tjoes are glven ln Flgures 5 and 7. See text for explanationild Table 1 for the actlvlty mdels ild mlneral abbreviations. Becauseconstant pressureis assumed,an isograd basedon this reactionis sensitiveonly to variations plexes(Figs. 7, l0). The isograd coincideswith the in temperatureand X(CO). hornblende- K-feldspar and diopside- phlogopite Unfortunately, neither magnetitenor hematiteis isogradsat severallocalities (Fig. 10). presentin samplesrelated to this isograd.Neverthe- Common occurrencesof the assemblagehorn- less,an isogradbased on the silicateminerals in the blende- calcite- qtrartz- diopsideare probably the reaction is still lesssensitive to variablesother than result of Na, K, Al and Ti substitutionsin hornblende temperatureand X(COr) than the hornblende- K- and, to a lesserextent, in diopside (LeAnderson feldsparisograd in Figure 5 becausemore minerals 1978a).The sensitivity of the diopside isograd to and, therefore, fewer degrees of freedom are mineral composition is displayedas a function of involved. Although based on fewer samples,the )(ffi*nh (Table 2) at an X(co) of 0.10 using the hornblende- K-feldspar - titanite isograd (Fig. 6) hasthe samegeneral pattern asthe hornblende- K- feldspar isograd. The diopside isograd based on reaction (4) lies betweenthe hornblende- K-feldspar and diopside - phlogopiteisograds around the two gabbro com-

TABLE1. ACTIVITYI.IODELS OF END.MEIitsERCOIT1PONENTS IN MINERALS

ulinrorrr,orotor),'(K) (Ms/3)3(oH/2)2 cuzz)2(Ms/s 2 "3fffilrrorrtont,' {tnfl ur*) t )5(oH/2) a!!r., - (f6u) . ca/(ca+ Mn+ Fe+ Ms)

u[1,r.,,0u. (xK). K/(K+ Na+ ca + Mn+ l,ls) uBlrrsiro,- *ilt ,;!"lurl"*,lrl i"tl ;,try andY " 1'276' orvllle (1972) u8laurrsr.ou . (r-Fe3{+ (si/3)3 toH, 1ca/212 Mn) uSlrnr,ro,- (c6)(r.rg) (sr/2)2

SeeLeAn,lerson (1978a) for derlvatlon of activlty nodels. The mlneral Dhaseabbreviatlons used here. in the text- and renalnlno = - Tables and Flgures are an anorthlte, bl " blotlt6. Ca-arnph" calclumilphlbole, cc - calclte, di = dlopslde,dol . dolonlte, Frc;.6.The hornblende- K-feldspar- titaniteisograd' See = ep. epldote,kf. potasslunfeldspar, ph phloqoplte,pl. Figure 3 for explanationand the text for derivation of plagloclase,qtz " quartz, rut = rutlle. ttn. tltillte, t. - trerbllte andzo. zolslte. the isogradicreaction.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 492 THE CANADIAN MINERALOGIST experimentallydetermined equilibrium-constant of Skippen(1974) (see LeAnderson 1978a, l98l). As expected,even small differencesin ffi*nn result in largeshifts in the position of the equilibrium curves. Such shifts profoundly affect the position of the isograd evenif there are relatively steepthermal gra- dients. An isograd basedon the following reaction:

hornblende+ calcite + quartz = diopside+ microcline + plagioclase+ epidote+ titanite

was mapped(Fig. 8) becauseit is a closerapproxi- Frc.7. The diopsideisograd, based on the model diopside- mation to the naturally occurring reaction. It defines producing reaction (4), SeeFigure 3 for explanation. the samehigh- and low-gradeareas as the diopside The diopsideisograd is also presentaround the Jocko isograd. Lake complexbased on one occurrenceof diopsidenear The diopside- biotite isograd,based on the model the contact with the complex. The position of this diopside - phlogopite reaction (5), is the highest- isocrad is shownin Figure 10. The numbersrefer to sam- grade isograd and lies near, and locally coincides plesused in the calculationof the displacementof tr- with, the diopside and hornblende - K-feldspar cc-qtz-di equilibrium and inlersectionof the ph-cc- isogradsnear the Thanet complex (Figs. 9, l0). qtz-tr-kf, tr-cc-qtz-di, and cc-an-zo equilibria (Tables 2,s). In summary,the isogradsdefine low-gradeareas below the titanite, actinolite - calcite, and hornblende - K-feldspar isograds(Figs. 3, 4, 5, l0) in the marble, and an increasein grade toward and into the quartzofeldspathicunits and the two amphibolite belts, and toward the gabbro complexes(Fig. l0). The generalrange of metamorphic grade for each rock tlpe is as follows: l) marble: generallyfrom below the titanite isograd to below the hornblende - K-feldsparisograd;2) quartzofeldspathic units and amphibolite belts: generallyabove the actinolite - calciteisograd; 3) calc-silicateunits: generallyfrom above the actinolite - calcite or hornblende - K- feldsparisograd to abovethe diopside- phlogopite isograd. The grade around the gabbro complexes, whereisograds could be mapped,ranges from above the hornblende - K-feldspar isograd to above the - - - Frc. 8. The diopside microcline plagioclase epidote diopside - biotite isograd. - titanite isograd.See Figure 3 for explanationand the text for derivation of the isogradicreaction. CelcurerroN oF UNTvARTANTEeurLrBRrA rN IsoBARrcT-X(CO) Sracr FOR THE MODEL SYSTSN4

Equilibrium constants for the model isogradic reactionsand related equilibria in the systemK2O- AI2O3-CaO-MgO-SiO2-H2O-CO,were calculated using the standard enthalpies,entropies, volumes, and heat capacitieslisted in the internally consistent data-setofHelgeson et al. (1978).The derivedcon- stantslisted in Table 3 result from combining standard thermodynamicrelations with the Maier-Kelley form of the heat-capacityequation (Helgesonel a/. 1978). Isobaric T-X(CO) univariant curves and FIc.9. The diopside- biotite isograd,based on the model their associatedinvariant points (Fie. I l) weregener- diopside-and phlogopite-producingreaction (5). See atedat a total pressureof 4 kbar usinga BASIC pro- Figure 3 for explanation. gram that solvesthe equation:

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 ROCK TYPE, METAMORPHIC GRADE AND FLUID-PHASE COMPOSITION IN THE GRENVILLE

=;:r'rXDt-i V-i, i, z,,it t '--'"a!)3.=ot)i)o 1 ilir;111;;;

Complex

r'\1|(:\5* / I \ t .J-4+.i;)\,1'.:, ,/ d1(\ ,lS:;^\Y ffi.<{ir; )ilii"'-=*S // | ' l' a*N.{l1,li,?aUtz.1.* o.o 05 t.o

Ftc. 10. Compositeisograd map. The isogradsare numberedthe sameas the correspondingreactions given in the rexr and Figures3,4,5,7,9, The hornblende- K-feldspar- titaniteand diopside- microclne- plagioclase- epidote - titanite isogradsare not plotted. Hachuresare on the high-gradeside of the isograds.

ln .l(",1: 0: In fr,r-(P-l)AVs + the equilibrium constantsare modified to account RT for the activitiesof the standard-statecomponents zserln(Pl[S2 Xco) + zgreln[PfffS(l-Xcor)] (l) in the mineral phases.This may be done by solving the equation: where./fp.r and Kl.r are the equilibrium constants for the reaction at P and T and at I bar and T, ln K",1 - ln@,=,.rai): a Q) respectively;A\ is the volume changefor the pure mineral phasesin the reaction 'vs2s as written; zs6, and where Kp,, refers to pure phases(equation l); y ?re the stoichiometric coefficients for CO] and representsthe total number of phaseswith variable HrO, respectively;f[$, and ffr$ are the fugacity activitiesin the reactions;ai is the activity of the rth coefficients for CO2 and H2d in the nonideal standard-statecomponent in the rth phase,and z1is CO2-H2O fluid, calculated according to a hard- the stoichiometriccoefficient of the rth phase. spheremodified Redlich-Kwong equation of stare We have neglectedCHa as a possiblefluid-phase (Kerrick& Jacobs1981, Jacobs & Kerrick l98la). componentalthough, under someconditions, CH, The programuses an iterativemethod to solvefor can be a major constituentin the systemC-O-H-S the equilibrium X(COr) at fixed P and T. (Ohmoto & Kerrick 1977).CIJa concentrations are Equilibrium (l) doesnot appearin Table 3 because highest in graphite-bearingrocks at low X(COr). data for rutile and titanite are not included in the Where present in significant amounts, CH, can compilation of Helgesonet ol. (1978).The approxi- lower the temperatureof invariant points by tens of mate position of the titanite equilibrium at 4 kbar degrees.The invariant point correspondingto the is shown as a dashedline in Figure I I basedon the assemblagegrossular - zoisite - calcite - anorthite position of the titanite isograd.The position of the - quartz (not presentin Fig. I l) is loweredby 40'C equilibrium in Figure I I is in good agreementwith at 2 kbat (Ohmoto & Kerrick 1977).The assump- the position determined experimentally by Jacobs& tion that Ptotur: P(H2O)+ P(CO' is reasonable Kerrick (l98lb). for the calculationsin this study becausethe high- The model eqrrilibriacan be usedto estimateboth grade rocks, which were characterized by low temperaturesand fluid compositionsprovided that X(CO) values, do not contain graphite.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 494 THE CANADIAN MINERALOGIST

TABTE3. COEFFICIENTSFOR EQUILIBRIA PLOTTED IN FIGURE11

ton *r.,r. a(log T - 0.4343)- bI + c1272- d/T + e + hP - constants Reactlon Number a bx1o3 cxlo-5 dxlo- e

16.257 In olo -1.491 2.3299 19.1224 0.6248 5Z.ttl I/.bIJ -19.025 1.8438 -64.5981 0,7903 4 8.937 4.530 -3.509 1 .5147 11.8141 0.5153 q 7.349 -4.618 3.2665 41.2736 1.0786 rr.027 -0.3526 o -0.298 0.049 0.837 0.3041 6.0467 7 9.237 7.641, 0.405 5.0909 77.5539 -1.0370 q 8.042 4.676 -0.998 2.4269 29.9541 -0.5425 9 50.321 17.905 -14,002 3.6683 -?8.3178 -1.3253 10 17.L52 10.793 -4.002 r.4176 0.9823 1.6826 l1 -0.668 3.041 2.9L0 L.0292 24.3930 0.2168 L2 0.227 2.895 0.399 0.1169 6.2530 L,2745

The standard state for al1 equillbrlun constants ls 1 bar and T for fluid (H20+ c02) and P and T for pure nlneral phases. ThermodJmamlcdata and mineral stolchionetry-are ta[en from Helgesong!gl. (19i8). P js ln kilobars andT ls in kelvlns.

The topology of the isobaric T-X(CO) grid 500oC range based on the reaction clinozoisite+ presentedin this paper differs from that calculated elJatlz= grossular+ anorthite + H2O (Boettcher by LeAnderson(1978a, 1981). The earlierversion 1970,Bird & Helgeson1981). used mostly experimentaldata, and did not incor- The temperatureof metamorphism for the marble porate either heat-capacityequations or models to is calculatedusing the MgCO, content for coexist- reflect a nonideal fluid. For example,in Figure 1I, ing calcite and dolomite. The MgCO3 content of reactions(3) and (4) do not intersect.However, if calcite in equilibrium with dolomite is strongly the data of Skippen(1974) are usedfor the diopside dependenton temperatureand is not significantly equilibrium and those of Hoschek (1973) for the affected by pressurein the range 1-10,000 bars tremolite - K-feldspar equilibrium, an intersection (Goldsmith & Newton 1969).The effectsof FeCO3 is obtained(LeAnderson 1978a, l98l). and MnCO, on the solvus are not known, but are Using the data of Helgesonet ol. (ln8), the cc- not consideredsignificant for mole fractions lessthan an-zo equilibrium (reaction 6) plots at a higher l9o, the content of FeCO3and MnCO, in the cal- X(CO2) than when the experimental data of cite in this study (LeAnderson 1978a). Johannes& Orville (1972)arc used;it plots closeto The rim of calcite grains in six samplesof low- the position of the equilibrium determined grade marble (Fig. 4) was analyzed. Grains with experimentallyby Allen & Fawcett (1982). In the suturedmargins or included dolomite blebs, or grains model system,the cc-an-zo equilibrium intersects that are in contact with elongatedolomite grains that the phl-cc-qtz-tr-kf, tr-cc-qtz-di, and tr-cc-kf- may representexsolved dolomite, were not analyzed. di-ph equilibria, which generatescc-absent equilibria Temperaturescalculated using the relationship: at each intersection(Fie. 1l). Similarly, cc-absent equilibria are generatedby the intersectionof the cc- 1.6 = (log Mgco. + 0.223)103/1.727 an-zo equilibrium with the dol-kf-ph-cc and dol- qtz-tr-cc equilibria (Fig. ll). (Sheppard& Schwarcz 1970) range from 535 to 595"C (Table4). A rangebetween 5 and 170'C for CoNorrroNs or MsrAMonpHrsN4 individual samples may indicate partial re- equilibration of calciteduring retrogrademetamor- The isogradsrecord an increasein gradefrom the phism (Hutcheon & Moore 1973). However, the marble into the quartzofeldspathicunits and the two highest temperaturescalculated are for sampleswith amphibolitebelts, as well astoward the two gabbro the greatestrange in temperatures,which suggests complexes.Prehnite is not presentin the study area. that Mg-rich domains were inadvertentlyanalyzed This absence, and the presence of coexisting in somegrains. If so, the temperatureof metamor- amphibole - epidote - diopside - zoisite - qtartz, phism of the marbleswas probably betweenapprox- indicatea lower-temperaturelimit during metamor- imately 535 and 550"C. phism of approximately 350oC. Thesephases are Temperatures are calculated for five samples of related by the reaction prehnite + tremolite : high-gradecalc-silicate and calcareousquartzofeld- diopside+ zoisite+ quartz + H2O, which was cal- spathicgranofels near the Thanet complex(Table 5, culatedat2and 5 kbar by Rice (1983)using the data Figs. 5, 7) using the two-feldspar geothermometer of Helgesonet al. (1978).The absenceof grossular of Whitney & Stormer (1977)and by calculationof and presenceof coexistingepidote, plagioclase and the displacementof intersectingequilibria (LeAnder- quartz suggesta maximum temperaturein the upper son 1978a,b, l98l). The activity models are given

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 \

ROCK TYPE, METAMORPHIC GRADE AND FLUID.PHASE COMPOSITION IN THE GRENVILLE 495 = t*ZO* COz 4OOOOot:r,

550

540 12dlr ph. tg onI g

550 122532o2O.O57 520

5lf o5hff lSonr g 500 11r181o, ,t 490 Y:t:o 480

470

460

450 o.o o.o5 o.to xco2 o.t5

Ftc. 11. The T-X(CO, grid'at4 kbar. Equilibrium constantsare from Helgesonet al. (1978)(Table 3). Equilibria l-5 are usedas model reaclions for the isograds(Figs. 3-10). The abbreviationsfor the mineral phasesare listed in Table l.

'IEMPERATURES in Table 1. The equilibriainvolved are cc-an-zo and TABLE4. CALCULAIEDFOR THE LOII-GMDE MARBLE USING ph-cc-qtz-tr-kf (sample389), cc-an-zo and tr-cc- CALCITE-DOLOI.IITEGEOTHERIIOI.IETRY qtz-di (A28 and A34), and tr-cc-qtz-di and ph-cc- qtz-tr-kf (M9 and 398).The temperaturescalculated SampI e Numberof calculated Tffiperature Nunber ^Moco^ using the experimentallydetermined equilibrium- Analyses Range (oC) l'|rximum (oC) constantsof Skippen(1974) (tr-cc-qtz-di), Hoschek (1973) (ph-cc-qtz-tr-kf and Johannes& Orville 98 3 0.0408-0.0505 50 535 4-8-2630.0505-0.05105540 (1972) (cc-an-zo) and the two-feldspar georher- 455 3 0.0404-0.0532 70 550 mometer of Whitney & Stormer (1977)(LeAnder- 40 4 0.0404-0.0619 110 590 51 8 0.0323-0.0638 L70 595 son 1978a,l98l) are givenin Table 5, togetherwith 54 1 0.0638 sgs the temperaturescalculated using the data of Helge- (1978) sotJet al. (Table 3). xfifiro,. r,lsl(l'ts+ Ca + Fe+ Mn). Temperaturescalculated for samplesA28 and 389, T"c E (lo9 Mgc03+ using the experimentaldata of Johannes& Orville 0.223)/0.001727.(sheppard & schwarcz1.970). (1972),are unrealisticallylow (LeAnderson 1978a, Samplelocatlons are glven in Figure 4. 1981)and arenot consideredfurther. However,temperaturescalculated for thesesame samples using the Temperaturesfor A29 and 398 could not be two-feldspar geothermometerand the data of Helge- obtained using the data of Helgesonet al. (1978) son et al. (1978) at 4 kbar are reasonableand in because the tr-cc-qtz-di and ph-cc-qtz.tr-kf agreementfor eachsample, although the tempera- equilibria do not intersect. Intersection can' be tures calculatedfor A28 are higher than for 389. obtained, however, with a slight increasein the

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 7

496 THE CANADIAN MINERALOGIST

TABLE5. CALCULATEDTEMPEMTURE (OC) AND X(CO,) VALUES FOR FIVE HIGH-GMDE SAMPLES

Feldsoar Intersectlnqequllibria Sanple Geotnbr- at 4 kbar (LeAnderson Intersectingequltibria Intersectlng-equillbrla Numier noreter 1978a, 19i8b, i981) at 4 kbar (this study) at 3 kbar (this study)

J55-ruD 487-517 447-492 :99-:i: 0.007-0.02 0.001-0.02 0.001-0.02 340-400 402-432 365-400 !\?':2',_ 0.007-0.02 0.015-0.04 0.01-0.02 y_?-122 440-505 537-607 497-557 0.02-0.09 0.035-0.10 0.02-0.07 440-450 i:1-::: 0.03-0.04 430-450 t:-i:i 0.02-0.03

oC. The flrst row for each sanple numberglves the calculated tsnperature range 1n The second row for each sanple numberglves the calculated range in X(C0r) (lntersectlng equillbrla only). Intersectlng equlllbrla are cc-an-zo and tr-cc-qtz-di (A28 and A34)i cc-an-qtz and ph-cc-qtz-tr-kf (389); ph-cc-qtz-tr-kf and tr-cc-qtz-dj (A29 and 398). No meaningful 'intersectlons were obtalned for sarnplesA29 and 398 at either 3 or 4 kbar uslng the data tn Helgesong!g1. (1978), owingto the nearly parallel relatlon of the curves tn T-X(CoA)space. SanplesA29, 398, A28 and A34 are calcslllcates and sanple 389 a calcareousquartzofeldspathlc granofels. The samplelocations are given on Flgures 5 and 7 and mlneral abbreviatlons in Flgure L.

entropy of reaction within the experimentalcon- of calciteand plagioclasein the marble, arld of cal- straints for either equilibria or, possibly, by using cite and epidote in rocks with low amoun{sof car- different activity-models.In view of the steepslope bonate, related through the cc-arl-zo equilihrium (6) of both reactions,there is no good justification for (Fig. 11), is in agreementwith this. making such arbitrary changes,because the result- The spatialrelationship between the isogradsand ing uncertaintyin the temperatureof intersectionis rock type indicatesa geneticrelationship. Further- very great. However, the temperaturescalculated for more, a decreasein X(CO) of approximately0.3 to 429 and 398 using both the two-feldspar geother- 0.5 is responsiblefor the increasein metamorphic mometer and the experimentally determined grade from the marble toward the carbonate- equilibrium-conslantscited above are within the deficient rocks nearthe Thanet complex,even though rangecalculated for samplesA28 and 389, although the temperaturedecreased several tens of degrees. the temperaturecalculated f.or A29 using the two- Based on these relationships, we infer that the feldspargeothermometer is higher than that calcu- increasein grade toward the Jocko Lake complex lated at 4 kbar using the experimentally determined and amphibolite belts is also due to a decreasein equilibrium-constants.The temperaturescalculated X(CO). The inverserelationship betweengrade and for A34 usingthe two-feldspargeothermometer and carbonatecontent may be due to: 1) internal buffer- the experimentallydetermined equilibrium-constants ing of the X(CO) of the fluid by the isogradic reac- are in agreement,but the temperatureobtained for tions. All five reactionsproduce CO, or produce this sampleusing the data of Helgesonet al. (1978) more CO2than H2O. Owing to the positiveslope of is the highest of all the temperaturescalculated. the isogradicreactions in isobaricX(CO) spaceat In summary,it appearsthat the temperaturenear low valuesof X(CO) (Fig. ll), rocks with a high the Thanet complexranged from approximately402 carbonatecontent, for examplemarble, would likely (389) to 543'C (A28) and possibly higher (up to produce greater amounts of CO2 than the 607oCi A34). Thus, the temperature range was carbonate-poorrocks. This would explain the high greater,and the averagelower, than for the marble, X(CO) and low grade in the marble, and low which revealsa range of 535 to 550oC. X(CO) and high grade in the carbonate-poor The X(CO) valuesfor the five samplesnear the rocks; or 2) buffering during the influx of external Thanet complex ranged from less than 0.01 to H2O-rich fluids. The quartzofeldspathicrocks and approximately 0.10. The X(CO) for the marble amphibolite belts may have been more permeable was approximately0.45 to 0.55 basedon the tem- than the marble, which would have allowed for perature range of 535 to 550'C and the calculated extensive infiltration of HrO-rich fluids in the position of the model qtz-dol-tr-cc equilibrium carbonate-poor rocks and maintenance of low using the data of Helgesonet al, (1978).The presence X(CO) values, a situation documentedelsewhere

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 \

ROCK TYPE, METAMORPHIC GRADE AND FLUID-PHASE COMPOSITION IN THE GRENVILLE 497

by Ferry (19S0"1983). Influx of heat accompanying & - (1962b): Rock-Forming the influx of fluld would also explain the wider range Minerals.i. SheetSilicates. Longmans, London' of temperaturein the calc-silicatesand calcareous (1963): quartzofeldspathicgranofels in the marble. & - Rock-Formine Minerals. 4. Framework Silicates. Longmans, London.

FeRRv,J.M. (1976): Metamorphism of calcareoussedi- The first author is very to P. Roederand ments in the Waterville-Vassalboro area, south- D.M. Carmichaelfor directing the research.D.M. central Maine, USA: mineral reactions and graphi- Carmichaelread an earlierdraft of the manuscript. cal analysis.Amer. J, Sci.c276,84l-882. Discussionswith A.E. Williams-Jonesand J. Bourne during the progressof the researchand comments (1980): A casestudy of the amount and distri on the manuscriptby G. Skippen,J.M. Ferry, W.E. bution of heat and fluid during metamorphism. Petrologlt 7l' 373'385, Trzcienski, Jr., and R.F. Martin were of great Contr. Mineral, are greatly appreciated. assistanceand - (1983): Regional metamorphism of the Vassal- boro Formation, south-central Maine, USA: a case REFERENCES study of the role of fluid in metamorphic petrogenesis.J. Geol. Soc.London 140,551-576. Ar-r-eN,J.M. (1976)t,Silicate-Carbonate Equilibria in Calcareous Metasediments of the Tudor Township FneNrr-lr, J.M., SeNcsrnn, D.M. & Lvpot't, J.W. Area, Ontario: A Test of the P-T-X(CO)-X(H2O) (1981):Volcanic-associated massive sulfide deposits. Model of Metamorphism. Ph.D. thesis, Queen's Econ. Geol., 75lh Anniv. Vol., 485-627. Univ., Kingston, Ont. GoroslurH. J.R. & NrwroN, R.C. (1969):P-T-Xrela- & Fewcrrr, J.J. (1982):Zoisite - anorthite- tions in the systemCaCO3-MgCO3 at high temper- calcite stability relations in H2O-CO2 fluids at 5000 aturesand pressures.Amer. J. Sci.267A, 160-190. bars: an experimental and SEM study. "I. Petrology 23.215-239. GneBNwooo, H.J. (1962): Metamorphic reactions involving two volatile components. Carnegie Inst. ANoensoN,P.A.M., Nrwrou, R.C. & KrEppe, O.J. Wash. Yesr Book 61,82-88. (1977): The enthalpy change of the andalusite- sillimanite reaction and the Al2SiOi diagram. (1967): Mineral equilibria in the system MgO- Amer. J. Sci.277,585-593. SiO2-H2O-CO2. .In Researchesin Geochemistry '(2nd edition, P.H. Abelson, ed.)' John Wiley & ' BeNcr, A.E. & Alsre, A.L. (1968): Empirical correc- Sons, New York. tion factors for the electron microanalysis of sili- (1975): of pore fluids by metamorphic cates and oxides. 'I. Geol. 76' 382-4f3. - Buffering reactions. Amer. J. Sci. 275, 573-593. (1981): aoAr,u:[1 of Bsncrn, G.W. & Yom, D. daiing Hr,lcpsox, H.C., Drr-a1ry,J.M., NrssITr, H.W. & the Thanet gabbro, Ontario: looking through the Brno, D.K. (1978): Summary and critique of the Grenvillian metamorphic veil and implications for thermodynamic .properties of rock-forming paleomagnetism. Can. ,1. Earth Sci. 18, 266-273. minerals. Amer. J. Sci.27EA, i

Bno, D.K. & Het-cssor.r,H.C. (1981): Chemical inter- Holoewav, M.J. (1971): Stability of andalusiteand the action of aqueoussolutions with epidote - feldspar aluminum silicate diagram. Amer. J. Sci. 271, mineral assemblages in geologic systems. II. 97-131. Equilibrium constraints in metamorphic,/geothermal processes.Amer. J. Sci. 281, 576-614. Hoscsx, G. (1973): Die Reaktion Phlogopit + Calcit + Quarz = Tremolit + Kalifeldspat + H2O + BoerrcHen, A.L. (1970): The system CaO- " COz. Contr. MineraL Petrology 39,231-237. AI2O3-SiO2-H2O at high pressures and tempera- HurcuroN, I. & Moons, J.M. (1973): The tremolite tures. "I. Petrology 11,337-379. isograd near Marble Lake, Ontario. Can. J. Earth CennarcnrL, D.M. (1970): Intersecting isograds in the ,Sci. 10, 936-947, Whetstone Lake area, Ontario. J. Petrology ll, 147-181. Jecoss, G.K. & Knnnrcr, D.M. (l98la): APL and FORTRAN programs for a new equation of state Dev, H.W. & KuvrN, H.J. (1980): Thermodynamic of H2O, CO2, and their mixtures at supercritical analysis of the aluminum silicate triple point. Amer. conditions, Computers Geosci. 7, l3l-143, J. $ci.279.1135-1159. - (l98lb): Devolatilization equilibria in H2O- Drrn, W.A., Howrc, R.A. & ZussuaN, J. (1962a\: CO2 and H2O-CO2-NaCI fluids: an experimental Rock-Forming Minerals. 2. Chain Silicates. Long- and thermodynamic evaluation at elevated pressures mans, London. and temperatures.Amer. Minerol. 66, I 135-1153.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021 7

498 THE CANADIAN MINERALOGIST

Jonexxns,W. & Onvrlrr, P.M. (1972):Zur Stabilitat Onvrr-rE, P.M. (1912): Plagioclase cation exchange der Mineralparagenesen Muskovit + Calcit ; equilibria with aqueouschloride solution: results at Quarz, Anorthit + Kalifeldspat und Anorthit -r 700oC and 2000 bars in the presence of quartz. Cilcil Fortschr. Mineral. 50, 46-47. Amer. J. Sci. tl2. n4-n2.

Kenmcr, D.M. & Jacoss, G.K. (1981): A modified Osssnc, P.H. (1971): An equilibrium model for Redlich-Kwong equation for H2O, CO2, and Buchan-type metamorphic rocks, south-central H2O-CO2 mixtures at elevated pressuresand tem- Maine, USA. Amer. Mineral. 56, 570-586. peratures. Amer. J. Sci. 2E1, 735-767. Rrcp, J.M. (1983): Metamorphism of rodingites. I. LsANosnsoN, P.J. (1978a): The Metamorphism of Phase relations in a portion of the system CaO- "fmpLtre" Marbles, Calcareous Schists, and MgO-AI2O3-SiO2-CO2-H2O. Amer. J. Sci. 2834, Amphibolites in a Portion of Limerick Township, l2r-150. Ontario. Ph.D, thesis, Queen's Univ., Kingston, Ont. SHsppeno,S.M.F. & Scrwancz, H.P. (1970):Fraction- ation of carbon and oxygen isotopes and magnesium - (1978b): Metamorphism of Grenville marbles between coexisting metamorphic calcite and dolo- near Bancroft, Ontario. Geol. Soc. Amer. Abstr. mite. Contr. Mineral. Petrology 26, 161-198. Programs 10, 442. Srrvpn,L.T. & Lur{sens,S.B. (1966):Geochronologi- (1981):Calculation of temperatureand X(CO) cal studiesin the Bancroft-Madoc area of the Gren- values for tremolite - K-feldspar - diopside - epidote ville Province, Ontario, Canada. Geol. Soc.Amer., assemblages.Can. Mineral. 19, 619-630. Spec.Publ.87, 156 (abstr.).

Lutvtsens,S.B. (1964): Preliminary report on the rela- SnppEN,G. (1971): Experimental data for reactions in tionship of mineral deposits to intrusive rocks and siliceous marbles. I. Geol. 79. 457-481. metamorphism in part of the Grenville Province of southeastern Ontario. Ont. Dep. Mines, Prelim. (1974):Experimental model for low pressure Rep.1964-4, l-33. metamorphism of siliceous dolomitic marble. Amer. J. Sci.274,487-509. (1967): Stratigraphy, Plutonism, and Metamor- phism of the Ottawa River Remnant in the SraucHrBn,J., Krnnrcr, D.M. & Warr, V.J. (1975): Bancroft-Madoc Area of the Grenville Province of Experimental and thermodynamic study of equilibria Southeastern Ontario, Canoda. Ph.D. thesisr Prince- in the system CaO-MgO-SiO2-H2O-CO2. Amer. ton, Univ,, Princeton, N.J. J, Sci.275,143-162.

- (1969): Geology of Limerick and Tudor Town- Sosor, J.W. & EsseNe,E. (1973):Petrology of Gren- ships. Onl. Dept. Mines, Geol. Rep.67. ville marbles from southern Ontario. Geol. Soc. Amer,, Abstr, Program 51 815, Moonn, J.N. & KBnn:cr, D.M. (1976): Equilibria in siliceousdolomites of the Alta aureole.Utah. Amer, WHrrNev,J.A. & Sronvsn, J.C., Jn. (1977):The dis- J. Sci.216,502-524. tribution of NaAlSi3O6 between coexisting microcline and plagioclase and its effect on geother- Munnay, D.P. (1976): Chemical Equilibria in Epidote- mometric calculations. A mer. Mineral. 62. 687-691.. bearing Calcsilicqtes and Basic Gneisses, Reading Prong, New York. Ph.D. thesis, Brown Univ., WrNrr-sn,H.F.G. (1974):Petrogenesis o"{ Metamorphic Providence, Rhode Island. RocAs (3rd ed.). Springer-Verlag, Berlin.

Osrrroro, H. & Ksnntcr, D. (1977): Devolatilization equilibria in graphitic systems.Amer. J. Sci. 271, Receiyed November 23, 1985, revised rnanuscipt 1013-1044. accepted August 29, 1986.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/25/3/485/3420309/485.pdf by guest on 27 September 2021