Canadian MineralogistVol. 19, pp. 25-34 (1981)

ABSTRACT

"Muscovite" (wbite mica) is the most commonmineralogical indicator of strongly peraluminouscomposition in plutonic rocks and, by inference, intheir parental magmas. Its presence has been usedto constrain depth of crystallization; based uponexperimental data, approximately 3 kbar (11 km)is commonly considered the minimum pressure atwhich primary igneous muscovite can crystallize.Recent suggestions that independent criteria requiredepths < 1l km for emplacement of some graniteswith texturally primaryJooking muscovite, so thatsuch mica would in fact be secondary, raise ques-tions about the use of apparently primary muscoviteas an indication either of depth or of magma com-position. New data from 4l samples representing16 plutons in North America and Europe are rel-evant to the paragenesis of muscovite in igneousrocks. Formulas of the analyzed micas are typi-cally about K6.s1Na!.67Fes + o.2oFe2 + 6.n5Mge.roTio.o,Al2.55Sis.10O1o(OH1.ssF0.oz), with very slight trioctahe-dral substitution (2.00 to 2.04 octahedral cations).Primary- and secondary-looking grains are generallysimilar, but primary ones are richer in Ti, Na andAl and poorer in Mg and Si. Plutonic muscovite isso far from ideal KAl"Si"Olo[OHl, in compositionthat it is difficult to evaluate its paragenesis interms of existing experimental data. The manyadditional components may enhance the stabilityfield sufficiently to explain occurrences of primarymuscovite at surprisingly shallow depths.

Keywords; muscovite, paragenesis, mineral chemis-try, granitic rocks, peraluminous granites.

*Current address: Martin-Trost Associates, 1510Washington St., Golden, Colorado 80401, U.S.A.

COMPOSITION OF PLUTONIC MUSGOVITE: GENETIG IMPLICATIONS

CALVIN F. MILLERDepartntettt ol Geology, Vanderbilt University, Nashville, Tennessee 37235, U.S.A.

EDWARD F. STODDARDDeparlntent of Geosciences, North Carolina Stute University, Raleigh,

North Carolina 27650. U.S.A.

LARRY J. BRADFISH*Department ol Geosciences, Uttit,ersit:u of Arizona, Tncsotr, Arizona 85721, U.S.A.

WAYNE A. DOLLASEDepartmerrt of Earth and Space Sciences, University of California, Los Angeles,

Calilornia 90024, U.S.A.

Souvretnn

La 'omuscovite" (dite mica blanc) est, dans uneroche plutonique, I'indice te plus constant d'unecomposition fortement hyperalumineuse de la rocheet, par induction, du magma originel. Sa pr6sencesert ir assigner des limites i la profondeur de cristal-lisation; d'aprds les r6sultats exp6rimentaux, unepression d'au moins 3 kbar (correspondant i 1l km)serait n€cessaire pour que la muscovite primairepuisse cristalliser au sein du magma. Toutefois, descritbres ind6pendants semblent indiquer une mise-en-place, i une profondeur inf6rieure Dr 1 I km, decertains granites ir muscovite d'apparence primaire.Si telle muscovite est r6ellement secondaire, elle nepeut indiquer ni la profondeur. ni la composition dumagma, Des donn6es nouvelles, 6tablies sur 4l6chantillons tir6s de 16 plutons d'Am6rique duNord et d'Europe, contribuent h pr6ciser la para-genbse de la muscovite plutonique. IJne compo-sition typique, (Ko.grNao.oz):o.ge(Fes+o.roFe2*n ooMgn.rnTin.n"Al,.no ) ::.03( Alo.eosii. r 0 ) :aOro( OHr.qrFn.nt ) , mon-tre un l6ger excEs en cations octa6driques (de 2.00i 2.04 cations). Les cristaux d'aspect primaire ousecondaire se ressemblent, sauf que les muscovitesprimaires contiennent plus de Ti, Na et Al et moinsde Mg et Si. Ces muscovites primaires s'6cartentd un point tel de la formule id6ale que les donn6esexp6rimentales sur la stabilit6 de KAlzAlSi,,Oro(OH)o n'ont gudre de rapport avec les conditionsparag6n6tiques. Les nombreuses composantes quiviennent s'aiouter au systeme pourraient 6lareir suf-fisamment le champ de stabilit6 pour nous per-mettre d'expliquer la pr€sence insolite de la mus-covite primaire ir faible profondeur.

(Traduit par la R6daction)

Mots-dAst muscovite, paragendse, composition mi-n6rale, roches granitiques. granite hyperalumineux.

25

26 THE CANADIAN MINERALOGIST

INtnooucrtoN

The paragenesis of plutonic muscovite is ofconsiderable petrological importance becausethe occurrence of apparently primary muscoviteis commonly taken as an indication of bothmagma composition (strongly peraluminous)and depth of crystallization (greater than aboutll km). Although other distinctive aluminousminerals (such as garnet, cordierite, sillimaniteand andalusite) may indicate a strongly per-aluminous composition (defined here as con-taining more AlzOs than can be accommodatedin feldspars f aluminous biotite), muscovite isthe most widespread and could be consideredthe most characteristic. The presumed pressrlreconstraint imposed by the intersection of thegranite solidus and the muscovite breakdowncurve has varied from about 3 to 4 kbar ( I Ito 15 km), depending upon the granite compo-sition and muscovite breakdown reaction used,and upon which experimental results are ac-cepted (e.9.. discussion in Thompson 1974). butshallow, "epizonal" depths appear to be pre-cluded.

In recent years many occurrences of primary-looking muscovite have been reported in plutonsthat have been emplaced at depths interpretedto be in the range 5 to l0 km (e.'g., Sylvesteret al. 1978, Nelson & Sylvester 1971, Benoit1971, Bradfish 1979, Swanson 1978, Banks1977). These depth estimates have been basedon structural and stratigraphic reconstmctionsas well as on the metamorphic environment(commonly andalusite-bearing aureoles in un-metamorphosed terranes). If these plutons havein fact been emplaced at such shallow depths,then either the muscovite is secondary in spiteof its appearance, or past evaluations of ex-perimental data are inapplicable to plutonicmnscovite. In either case, the use of muscovitefor depth estimation would be invalidated. andif it is indeed secondary, its use as an indicatorof magma composition becomes questionabie,because it may have forrned metasomatically.

The purpose of this paper is to present dataon the composition of plutonic mnscovite thathave a bearing upon its paragenesis and petrol-ogical interpretation.

PRocEDURE

Samples investigated

A total of 186 muscovite grains from 4lsamples representing 16 plutons were analyzed(Table 1). A majority of the samples came from

the Old Woman*Piute Range in southeasternCalifornia and the Teacup granodiorite ofsouthern Arizona.

Analytical rnethods

The results of electron*microprobe mineralanalyses presented here are from three differentIaboratories. Analyses from the Teacup grano-diorite were done at the University of Arizona;the remainder were done at University of Cali-fornia at Los Angeles (UCLA) and at VirginiaPolytechnic Institute. Accelerating voltage was15 kV in all three laboratories: sample currentwas 2O nA (U. Ariz.) or 15 nA. Althoughsome details of data collection varied among thelaboratories" all used the correction factors ofAlbee & Ray (1970) in calculating weight per-cent oxides. Analytical accuracy and precisionare estimated to be within + 37o of the amountpresent for major elements, ! 5% for mostminor elements (including Na) and -+ 10%o forF and Ba. For the determination of Fee+/Fe2+,otFe Mdssbauer spectra were measured at UCLAon powdered, handpicked mineral separates ofcoarse white mica. Further details of analyticaltechniques are available from the authors.

Primary versus Secondary miccts: textural criteria

One of our goals in this study was to deter-mine whether there were distinct compositionaldifferences between primary and secondarymuscovite in plutonic rocks or, conversely, tosee whether compositional criteria were suffi-cient to distinguish between primary and $econd-ary grains. To do this, we established texturalcriteria that would distinguish a primary (P)from a secondary (S) origin. For a P origin(c1., Saavedra 1978), a mica grain must (l)have relatively coarse grain size, comparable toobviously primary phases; (2) be cleanly termin-ated, ideally with subhedral or euhedral form;(3) not be enclosed by, or raggedly enclose, amineral (or other possible alteration productsof a mineral) from which the muscovite mayhave formed by alteration (e.g., feldspar, alu-minum silicate); (4) be in a rock with clean,unaltered, igneous (in most cases hypidiomor-phic-granular) texture.

We interpret any grain that meets all theabove criteria as primary lbut see, for example,Benoit (1971) for an opposing viewl. Somegrains that do not meet all of these criteriamay also be primary; in particular, there maybe coarse, subhedral primary muscovites re-maining in an altered rock with secondary

COMPOSITION OF PLUTONIC MUSCOVITE

IABLE I. SA'.IPLES ANALYZED

27

L e l l t y

old f@n-Pl ut6 R€nge,lbjow Desert , CA(5 plutong)

T€acup gr€nodlor l to,South-€ntral AZ

P€pasg Flat plqton,Inyo l . l tns., o€stern CA

glrch Creek pluton,Uhlte l . l tns., 6astern CA

La Rata plutd,Penlnsular R€nge',B q J € C € l l f . , i i 6 x l @

Stone l lount€ln gr€nlte,Inn€r Plgdnont, GA

l l h l t e s l d e p l l t o n ,B l @ R l d g e , N C

Spruco Plno Als3kl te,B I u e R l d g e , N C

Wyatt Stat lon pluton,R € l e l g h B e l t , N c

S l n s p l u t o n ,East6rn Slate Belt , NC

eresh€mrs L€ke pl l ton,R a l e l g h B e l t , N C

Argentera l , lassl f ,l . l € r l t l @ A l p 3 ,Fr€nce - l ta ly

sarnpl os./grelns analy"6d

21 /71

7 /58

1/ '

v5

1t7

t/7

1 l l

1/3

r/5

1 /7

3/9

2 / to

| l l ner€lAssenlblages*

m b ( + m t + l l m ) .'bs T1 'tf,tr€r

mgbg

m ( + n t ) .mb T+ mt + ru).nls (1 mtf,ry

mb (mt + l lm)

Repres6ntat I.vel,lugcovlte FotuulsS**

K0. 88Nso. 09FEo.23lr90. 06T I o. 03-A12,6 ,2s13. 12016 (0H: r .96Fg. 16) : l*AUf iSs?f;36if; sfi ;.llll3:il i,s

Ko. 9qN€0. 05Feo.29rs0. t2Tl 0.05-A l 2 .50S13. lZ0 lO (0H2Fn€) :P

ro.95No0.04Fe0.32ltso, r6Tlo.02-A l2 .4oS 13. 19010 (0H2Fn€) !S

tu*;fi&tf iigolSHn i t6118 :?:i,,''A?;i;9.9!:;hi3i3fi ;i3l13 :3ti,'h.85Nr0 .OBFeo. to lso .osT lo .o2 :

A l2 . 86s l j .O40tO (0Ht .98FO.02) :P

-%?i:;93?3:36l3ts; :;ll3:B;;,,h^tli;gntii;?0fii38; llllS : Bli,.Kt .OON€O.O6,FeO. 3 lxgo.O9TlO.O0-

A 1 2 . 5 3 5 1 3 . l 2 0 t O ( 0 H r . 9 7 F O . 0 3 ) : P

brtl'igstii;io?3i3si3ll3 : 3ti,,r€.94N€o . 02Feo.3oMso.29T I o. 03-

A l 2 . 2 3 5 1 3 . 2 6 0 1 0 ( 0 H 1 . 9 | F O . 0 9 ) r S

Ko:glNao,Q5Feo.27l, ' lgo.ozrlo.oz; _A r 2 . 6 0 r r 3 . l 2 u l o t u n l . 9 6 t o . 0 4 , r r

KO. 83NaO. t OFeO. t8l, l9o. 08Tl O. 02-A l2 .7oSl 3 .o90 lo (oH2Fna) :P

References

l , l l l l o r e t a l . 1 9 8 0 ;i l l l e r 6 s t o d d € r d

r978

Bradflsh 1979

Sy lves ter e t a l .1978

Nelson a Sylvestor1971

Gast I I 1975

r rh l tney e t a l .1976

l t l l l e r and K lshr 980

Brobst 1962

Psrker 1979

Wede@ye r 6sprul I I 1980

Parker 1979

De Po l e t a l .r968

rb (mt)

ob

nb

irb

mbg (nt)

m (mt?)

n b g ( m t ) ,nS (: mt)

m9rrug

* v € r l e t e l m l n e r € l s : l ' l - n u s c o v l t e r b - b l o t l t e , g - g o r n e t , s - s l l l l m a n l t e i ( a c c e s s o r y m l n e r a l s ) : ( m t ) - m g n e t l t € , ( l l m ) - 1 1 * n 1 a u ,( r u ) - r u t l l e . * { P E P r l m r y - l @ k l n g , S - s e c o n d a r y - l @ k l n g ( s e e t e x t ) ; a l l F e c a l c u l a t e d a s f e ( Z + ) ; m l n o r f l n , C r , C a , B a , a n d C l@lttedi oH+F assured - 2; n€ - notan€lyzed.

muscovite; in fact, this is the most interestingcase for analysis. Therefore, we assign thosegrains that meet the first three criteria, eventhose found in mildly altered rocks, to the "P-mica" category, whereas small, ragged grains,particularly those that are confined within orat the margins of certain minerals, are con-sidered "S-micas", or secondary. Figure I showstypical examples.

Since by no means were all grains easily

Flc. l. Typical P- and S-micas; bars indicate 1 mm. (a) P: coarse, subhedral, cleanly terminated grain,Old Woman-Piute Range (crossed polars); (b) P: coarse, subhedral, randomly oriented grains parti-ally enclosed by (later?) biotite; Old Woman-Piute Range (plane polars); (c) S: small grains enclosedwithin plagioclase (as alteration?), Teacup granodiorite (crossed polars).

placed in one or the other category, some ofthe designations as P- or S-mica are somewhatuncertain.

Resulrs

The few published analyses of plutonic mus-covite are commonly high in total Fe + Mg(0.2 to 0.5 cations per formula based on 11oxygen atoms), contain noteworthy Ti, and

28 THE CANADIAN MINERALOGIST

55%Si\>) o p' . . i i ' o-

' i " ' { o o o o o " : . . . \

' 1 r - - " ' ; i

i: '.' t

Tr:rt'".{ij. i': . :. -.'r.ia1 ^. 6{\--1.. - ..:

Mg

Mg NoFrc, 2. Compositions of analyzed muscovite grains. Closed circles - P-micas; open circles - $micas;

dotted linis enclose fields of Old Woman-Piute Range micas; dashed lines enclose Teacup grano'

diorite micas. (a) Compositions in terms of Si, Al and other octahedral cations (Fe, Mg, Ti, Mn,Cr) (atomic proportions), The inset shows the location of mica end'members: M muscovite, fluormus-covite and paragbnite, F ferrimuscovite, C celadonite, T end-member formed by total substitution TiMg",-+ 2 Al't, A annite-phlogopite, and S siderophylite-eastonite (see text for formulas). Ruled areashown in figure. (b) Compositions in terms of Ti, Mg and Na (atomic proportions)' Ruled area ininset is the area shown in the main fi€ure.

AI

3 r - t . t t a " 1i *:.,'_::+;' . '3 r- l ' ] lE,' i.z;l:i ":" ":-.."-

, " ;-t2i;

COMPOSITION OF PLUTONIC MUSCOVITE 29

have excess Si and deficient Al compared withideal muscovite (Joyce 1973, Best et al. 1974,Neiva 1975, Mohon 1975, Becker 1978, An-derson et al. 198O, Anderson & Rowley 1981),Muscovites investigated by Castle & Theodore(1972), Hamis (1974) and Guidotti (1978a)approach ideality more closely but still haveabout 0.15 (Fe * Mg)/formula unit.

Our data confirm that plutonic "muscovites"differ significantly from the ideal stoichiometryKAI'rSirOrofOHlr. The average formula (Table1) is approximately Ko.stNao.orFes+n.2sFe'*0.0,Mgo.toTio.orAL.rsSig.toOro(OH,.rgFo.or). The ele-ments Ca, Ba, Mn, Cr, Cl and P are presentin negligible quantities (< 0.01/formula).

Formulas calculated for muscovites assum-ing all Fe as Fe2+ have 2.M to 2.10 octahedralcations per formula, suggesting that much ofthe Mg and Fe might be present as a trioctahe-dral (biotite) component. However, in most ofthe samples analyzed by Miissbauer spectro-scopy, over 7O% of the Fe is present as Fe"*(total range was 27 to 88%). Recalculatingthe formulas of these samples using the Fe3+/Fetour ratios determined indicate a far smalleroctahedral excess (Table 2); with a reasonable

TABLE 2. EXAI.IPLE OF CORRECTION OFtluscovlTE col.tpostTt0N FoR FERRTC tRoN{

Fe (3+) /Fe (total )-0.80uncorrected corrected

s t 0 2A1203T l 0 2MgoFe2O3Fe0l.ln0Cr2O3Ca0Na20K20bauF0 - F

Total

F o r n u l a , 0 - l l r

s lA l ( l v )

A l ( v l )T If,tgFe (3+)Fe (2+)MnC r

NaKBa

i

estimate of Fet+/Fe.*r (0.75), octahedral siteoccr"rpancy in all samples is 2.02 -r 0.02 (1o)per formula. If all Fe and Mg were triocta-hedral, the octahedral site would have [2 +(Fe * Mg)/31 cations, neglecting minor effectsof Ti. Because Fe * Mg typically total morethan 0.3 cations/formula. it is clear that tri-octahedral substitution is very minor.

Our data do not allow formulas to be recastuniquely as combinations of ideal end-mem-bers because ( I ) the nature of Ti substitutionis not fully understood and (2) for most sam-ples Fes*/Fetour is not known. End membersthat must be present include ideal muscoviteKAlrSLOro[OH]r, ferrimuscovite KFes+:AlSi,rOrnloHl'. celadonite K[Mg,Fe]AlSioorololfl,,paragonite NaAlsSi"O'olOHl', fluor-muscoviteKAl,rSi,rO,oF,, biotite K [Fe.Mg] gAlsiro'o f OHl r-K(Fe,Mg):.rAldis.'OrofOHl' and a Ti-bearingspecies. The Ti substitution may involve (1)J114q.Fe) + Til'' = 2Af', giving K(Mg,Fe)TiAlSLOro(OH)r as the end member; (2) Tf'+ Al'" = Al"' * Si'", giving KTiTALSiO'o(OH)eior (3) Ti"' + tr = 2(Mg.Fe), leading toKTi0.sAlSLOrn(OH), (Dymek & Albee 1977).Guidotti (1978b) considered substitution (1) tobe most important in muscovite.

In Figure 2a our analyses are plotted in termsof Si,Al and the sum of octahedral substituentsFe, Mg, Ti, Mn and Cr. Although the presenceof minor amolrnts of trioctahedral (biotite) andTi-bearing end-members complicates the inter-pretation of the diagram, it is clear that ferri-muscovite and celadonite are important com-ponents, and that our analyses average approxi-mately 75/6 muscovite 1 paragonite + fluor-muscovite. Paragonite and fluor-muscovitetypically total nearly l0%, leaving approxi-mately 65% of the pure muscovite end-member.

Although our data reveal general consistencyamong plutonic muscovite compositions, Figure2 and Table I illustrate considerable variationfrom pluton to pluton and from sample tosample. The variability is not clearly linked tothe coexisting mineral assemblage (Table 3).Table 4 reveals that there are also large differ-ences among grains within a single sample.Much of the within-sample variation seems at-tributable to paragenesis; we believe that dif-ferences in composition may arise dependingupon the stage at which the individual micagrain crystallized or last equilibrated (see be-low).

Coexisting phases

In all samples examined, muscovite coexists

45.83

0.47U . b J

q . o t0 . 0 30 . 0 1o.o20.62

10.250 . 0 50 . q 8

-0 .20

9 \ , 7 3

45.83

o . \ 7

J . ) b

0 . 8 00 . 0 30 . 0 10 . 0 20.52

10.250 . 0 50 . 4 8

-0 .20

95.08

3 . t 2 9 3 . 1 0 30 . 8 7 1 0 . 8 9 7

| .743 r .6950.024 0 .0240.064 2 .062 0 .063 2 .0 i l

0 . 1 8 10.229 0.0460.002 0 .002

0 0

0.002 0 .0020.082 0 .976 0 .081 0 .9680.891 0 .8840 . 0 0 t 0 . 0 0 1

0 . 1 0 2

2.592E A I 2 . 5 1 \

* S a m p l e S U - 8 , m u s c o v l t e - b l o t l q e g r a n l t e , 0 l d w o m a n | l o u n t a l n s ;F e ( 3 + ) / F e ( t o t a l ) f r o m c o a r s e n u s c o v l t e s e p a r a t e , r e s t o fenalysls B average of P-mlca analyses.

30 THE CANADIAN MINERALOCIS|

TABLE 3. }IEAN I.IUSCOVITE COI.IPOSITIONS tN DIFFERENT '{INEML ASSEI'IBLAGES**

Assemblage

# samples# g ra l nss 1024r203Tt 02ltg0Fe0*iln0Ca0Na20K20

0 - F

Anhyd rousTot€ |

m u + b l

o

l 94s.72 + 0.46 (ro)3 2 . \ 3 + 1 . 1 30.67 F 0 .18o . 7 3 T 0 . 1 63.92 7 0.7\0.04 ; o.o30.04 F 0.040.58 T 0 . r8

10.44 t 0 .23o j 3 6 F o . t 0

- 0 . r 5 -

94.78

mu + b l+ 9 t

l 01 6

45.85 + 0.6632.45 T 13\0.45 ; 'o .z r0 .75 F 0 . r94.0\ 7 0.77o.05 T o .020.09 T 0 .04o . 4 r ; ' 0 . r 2

1 0 . 4 3 ; 0 . 3 6N A -

4t 2

45 .30 + 0 .8532.42 T o.820 . 6 4 ; 0 . r 9o . n 7 0 . 0 73 .70 t 0 . 44o .o5 F o .o20 .08 t 0 . 040 .70 t o . r 0

r 0 . 3 8 ; o . 4 o0 .26 F o .o5

-0 .1 t -

94 .15

3 . t 0 4 + 0 . 0 4 20.996

-

1 .6230 .033 + 0 .0 r00.074 F 0.007 2.0450.212 T 0.0250.003 ; o.ool

0.005 + 0.0030 .093 T 9 .91d 1 .0060 .907 ; 0 . 036

0 .060 + 0 .012

2 .619 + 0 .055

m u + b l+ g t +

mu +gt sl l

I4

45.16 + o .2434.09 i 0 .2\0 ,08 t o .o70.61 r 0 .072 . r 5 i 0 . 2 2o . o l t 0 . o l0 .09 t 0 .030.59 I o .o8

10.23 + O,07NA

Forrule, 0 s l1***

s t 3 . 115 + 0 .032Al ( rv) 0.885

-

A l ( v l ) 1 . 7 1 8T l 0 .031 + 0 .010r,rg 0.074 F 0.016 2.052Fe 0.224 T 0.044Hn 0.002 T 0.002

Ca 0.003 + 0.003Na 0 .076 ; 0 .023 0 .988K 0.907 T 0.020

F 0.080 + 0.020

tAl 2.603 + 0.070

94.52

3.124 + 0.022o.876

-

1 .7300 .023 + 0 .0 r t0.076 T 0.020 2.0630 .231 ; 0 . 046o .oo3 F o .0o l

0.007 + 0.003o.o5lr F 0.017 0.9680.907 I 0.034

2.606 + O.O74

93.01

3.093 + 0 .0 r00.907

1 . 8 4 60,004 + 0 .0040.062 ; 0 .008 2 .0350 . 1 2 3 T o . o r 3

o -

0 .007 + 0 .00 !0 .078 i o .o to 0 .9790.894 I o .oo7

2.753 + O.01O

*Fe as Feo. * *0 ld l y 'oman-P lu te Range P-mlcas . * * *A l l

with quartz, K-feldspar and sodic plagioclase.Most also contain Al-rich ( > 1.5 Al/ I I oxygenatoms) biotite, and many have almandine-spes-sartine garnet. The most common accessoryminerals are apatite, zircon and magnetite; il-menite, rutile and primary-looking epidote arepresent in a few samples. Sillimanite was iden-tified in one sample.

The occurrence of Al-rich biotite and garnetis consistent with a strongly peraluminous mag-ma composition.

Previous interpretations of compositionalvariation

In recent years, celadonite-rich micas orphengites commonly have been considered tobe restricted to relatively high-P and low-Tenvironments. There are considerable analyticaldata that support the occurrence of celadoniticmicas in such environments (Miyashiro 1973),and some experimental data suggest that thesemicas are unstable in igneous or high-grade meta-morphic environments [e.9., Velde (1965), butsee Crowley & Roy (1964) for a different inter-pretationl. Brimhall (1972) reported celadonite'

Fe Ca lcu la ted as Fe(2+) - NA not ana lyzed '

rich compositions similar to ours for clearlysecondary, sericitic mica, but these grains arenot low-T alterations: he estimated that theywere formed in the range 55G-690"C. Whitemica in high-grade metapelites, though higherin the muscovite component than our samples,is also not ideal in composition; for example,Guidotti (1978b) reported upper-sillimanite-zone muscovite that has more than 10% celado-nite or celadonite * Ti-bearing component.Anderson & Rowley (1981) propose, on thebasis of thermodynamic considerations, thatceladonitic muscovite should be stable tohigher T than pure muscovite.

Composition as a function ol coexistingassemblage

Table 3 illustrates the variation in muscovitecomposition as a function of coexisting mineralassemblage for samples from the Old Woman-Piute Range. Muscovites coexisting with bio-tite, garneto and garnet * biotite are almostidentical. This is somewhat surprising, becausewhole-rock and biotite chemistry do vary sys-tematically. Muscovites from the single silli-

COMPOSITION OF PLUTONIC MUSCOVITE 3 l

manite-bearing sample are dftinctly differentfrom the rest, with higher Al and lower Fe,$S, fi and Si. Analyses from this sample differfrom all others delermined during this project,with the exception of those from the Li Ravapluton, Baja California. The La Raya rumil"does not contain sillimanite, nor has it beenfound in the pluton as a whole (R.G. Gastil,pers. comm. 1980), but similar analvses havebeen reported for other sillimaniie-bearingplutonic rocks (Guidotti 1978a).

Composition as an indicator of primory orsecondary origin

Several compositional characteristics distin-guish most of our texturally primary-looking (p)from secondaryJooking (S) micas. Most tex-tural P-micas are considerably richer in Ti, Atand Na and poorer in Mg and Si than S-micas(Fig. 2b) [4., Anderson & Rowley (l93l)for rather different resultl. P-micas are typi-cally somewhat closer to ideal muscovite thantheir S-mica counterparts, primarily because oflower Mg (Fig. 2a).

Table 4 illustrates compositional differenceswithin a single sample, and Figtrre 3 showsqypical P- and S-micas from that sample.

Imperfect corlelation between composifionand our textural types is probably a functionof (l) our inability to evaluate.paragenesis ofall grains on a textural basis and (2) the com-plex physical-chemical environment and historyof these grains. Secondary muscovite probablyforms under a variety of conditions, and pri-mary muscovite may be susceptible to subsolidusre-equilibration. Still, the ionsistency of com-positional distinctions is striking and may proveuseful in the interpretation of other muscovitegranites.

The high Ti of apparently primary muscovitehas been noted by Anderson (e.g., Anderson &Rowley l98l); he points out the analogy withhigh-grade metamorphic muscovite, which isalso enriched in Ti compared with low-grademuscovite (Guidotti 1978b). The significanceof high Na is difficult to evaluate because max-imum Na in muscovite increases with T at lowto moderate metamorphic gradesn but near mag-matic T, where the assemblage muscovite +K-feldspar 1 Al-silicate is stable, the trendreverses (c1., Thompson 1974),

The lower celadonite component of P-typemicas would qualitatively agree with Velde's(1965, 1967) conclusion that celadonite abund-ance should diminish with increasing T. How-ever, according to Velde's experimental data,

all of our micas, P or S, have higher celadonitethan should be tolerated at temperatures ap-proaching those of magmas (see also plot inMiyashiro 1973, p. 2O2). The discrepancy be-tween our apparently igneous mica compositionsand those predicted by Velde may be a functionof an oversimplified experimental environment.It is conceivable, though in our view less likely,that the P-micas are either partially re-equili-brated primary micas or coarsely crystalliiedsecondary micas (Bradfish 1979). If so, thedifferences between P and S compositions mustreflect either imperfect equilibration or crystal-lization under different subsolidus conditions.

Pressure constraints implied by primarytnuscovite

The minimum pressure at which muscovirecan coexist with silicate liquid is dependentupon so many factors that are currently poorlyunderstood that we regard precise estimation asimpossible. Both the granite solidus and themuscovite-breakdown (muscovite + albite *in either lead to large uncertainties in their

TABLE 4. TYPICAL P- AND S.MICAS FRO}1 A SINGLE SAilPLE**

A(s )

46 .43? o l A

0 . 3 1 ,I . 54

0 . r 80 .020 . 0 10 . t 9

| 0 . 930 .04u . ) l

-o.21

94.79

3.2050.795

1.5790 . 0 r 60.169 2.O970.3210 . 0 1 10 .001

0 .0010.025 0.9990.9620 .001

0 . 1 1 1

2.374

G r a i n , t y p e :

s t02Al 203T i02tt90Fe0*Mn0Cr2O3Ca0Na20t

32

placement. An error of l0"C in either willchange their intersection by about 0.4 kbar (1.5km). Most workers use the intersection of thecurve muscovite * quartz -> K-feldspar fsillimanite + HrO with the water-saturatedsolidus in the system Ab + Or * Qz to definequartz -+ K-feldspar f Al-silicate + HrO)curve are steepo and therefore small uncertaintiesthe minimum P. The exact location of theformer curve is open to some question (cf.,Day 1973, Althaus et al. 197O) and should inany case be at lower T because of the invol-vement of the Ab component in the reaction(Thompson 1974),leading to an intersection athigher pressures. The granite solidus will beraised by the presence of the An componentand by a(H:O) ( I (therefore higher P in-tersection) and lowered by the presence of maficcomponents, excess AlrOa (Abbott & Clarke1979) and by the presence of boron (Chorlton& Martin 1978), thus lowering the P of inter-section. The effect of boron may be particularlyimportant.

The nonideal composition of plutonic mus-covite also mirst influence the position of thegranite solidus/muscovite breakdown intersec-tion. Extrapolation of Velde's (1965) data sug-gests that muscovites with the compositions wefind could only be stable at unreasonably highP (probably at least 7 or 8 kbar, or 30 km).Thermodynamic considerations lead Anderson& Rowley (1981) to an opposite conclusion,that their muscovite" which is within our com-positional range, has a higher T stability limitthan ideal muscovite and could have crystallizedfrom granodioritic melt at 2 kbar (7.6 km).

'We tentatively support the reasoning of An-derson & Rowley, which leads to the conclusionthat the effect of celadonitic impurities alonecan explain the occurrence of primary mus-covite in plutons emplaced at apparently anom-alously low pressure. Most importantly, how-ever, we wish to emphasize the uncertainty ofthe pressure range in which primary igneousmuscovite can crystallize and to caution againstpetrological interpretations based upon the 3-4kbar (11- '15 km) min imum.

CoNcl-ustoNs

Onr conclusions regarding plutonic mus-covites mav be summarized as folllows:

Frc. 3. Examples of analyzed P- and $micas fromsample Pl5, Old Woman-Piute Range; bar in-dicates I mm. See Table 4 for analyses. (a)grain D (P-mica). (b) grain A (S-mica).

COMPOSITION OF PLUTONIC MUSCOVITE 33

(1) The analyzed muscovites are far from idealend-member muscovite. They contain appreci-able Fe, Mg, Ti, Na and F. They are deficientin Al and have a modest excess of Si.(2) Because the analyzed micas are so non-ideal, experimental data are not directly ap-plicable to their stability. Hence, the inter-section of the reaction curve muscovite -1- albite+ quartz - K-feldspar f Al-silicate * HgOwith the granite minimum melting curve is notthe true minimum pressure of crystallization ofprimary igneous muscovite. If, as suggested byAnderson & Rowley (1981), this reaction occursat higher T for celadonitic muscovite, the mini-mum P is lower (they suggest 2 kbar, or 7,6km, for their samples). Alternatively, if oneaccepts the conclusion of Velde (1965, 1967)that celadonitic mica is confined to lower T thanmuscoviten either much higher P is required forcrystallization of these micas as magmatic min-erals, or the mica compositions represent sub-solidus re-equilibration, or the micas are allrelatively low-T secondary minerals. The finalalternative elirninates plutonic muscovite as anindicator of emplacement pressure and castsdoubt on its value as an indicator of primarymagma composition. We tentatively agree withthe interpretation that the impurities increasethe stability field of muscovite (Anderson &Rowley 1981) and that these micas may there-fore have crystallized from magmas at relat-ively shallow depths (< l0 km). We furtheremphasize the many other uncertainties in place-ment of both granite-solidus and muscovite-breakdown curves that make estimates of theirintersection (minimum P of primary muscovite)very imprecise.(3) The fact that primary-looking muscovitealmost invariably coexists with aluminous bio-tite -f garnet suggests that such muscovite isindeed restricted to rocks crystallized fromprimary strongly peraluminous magmas, even ifthe muscovite itself may be secondary.(4) Compositional differences may distinguishprimary from secondary muscovite. Grains thatappear t€xturally to be primary typically havehigher Ti, Na and Al and lower Si and Mg thanthose that appear to be secondary.

AcKNowLEDGMENTs

Samples of muscovite-bearing granite werekindly supplied by R. Gordon Gastil of SanDiego State University (La Raya plutop, BajaCalifornia), Achille Blasi of Universita diMilano (Argentera Massif, France-Italy) andJohn Gillespie, Jr., of the University of Cali-

fornia, Santa Barbara (Papoose Flat and BirchCreek plutons, California). Microprobe workwas assisted by Robert Jones (UCLA), ToddSolberg (VPI) and Tom Teska (University ofArizona). Discussions with J. Lawford Ander-son of the University of Southern Californiacontributed in many ways to the developmentof ideas expressed in the paper. T.P. Loomis,R. Beane and A. Meijer of the University ofArizona advised Bradfish's Master's thesis,which dealt in large part with muscovite para-genesis in the Teacup granodiorite. Research wassupporred bv NSF granr EAR 7823694 and bythe Vanderbilt Universitv Research Council.

REFERENCEs

Annorr, R.N., JR. & Creme, D.B. (1979): Hypo-thetical liquidus relationships in the subsystemAllO,r - FeO - MgO projected from quartz,alkali feldspar and plagioclase for a(H:O) <l. Can. Mineral. L7, 549-560.

At.sru, A.L. & Rev, L. (1970): Correction factorsfor electron probe microanalysis of silicates,oxides, carbonates, phosphates, and sulfates. lrral.Chenr.42, 1408-1414.

ALTHAUS, E., Kenorrr, E., Nrtscu, K,H. & WtNr-rrn, H.G.F. (1970): An experimental re-exami-nation of the upper stability limit of muscoviteplus quartz. Neues Jahrb. Mineral. Monatsh.,325-336.

ANDERSoN, J.L., Curnns, R.L. & vAN ScHMus,W.R. ( 1980) : Anorogenic metaluminous andperaluminous granite plutonism in the M-id-Pro-terozoic of Wisconsin, USA. Contr. Mineral.Petrology 74, 371-328.

& Rowrev, M.C. (1981): Synkinematic in-trusion of peraluminous and associated metal-uminous granitic magmas, Whipple Mountains,California. Can. Mineral. 19, 83-101.

BeNrs, N.G. (1977): A Tertiary igneous-meta-morphic complex in southeastern Atizona. Geol.Soc. Amer. Abstr. Prograns 9, 385.

Brcrrn, S.W. (1978): Petrology of the CuffytownCreek pluton, South Carolina, Geol. Soc. Amer.Ahstr. Programs LO, 162.

BENorr, W.R. (1971):. Yertical Zonitr.g and Dil-ferentiation in Granitic Rocks - Central FlintCreek Range, Montana, M.Sc. thesis, Univ. Mon-tana. Missoula, Montana.

Brsr, M.G., AnusrnoNc, R.L. Gneusrsru, W.C.,EMBREE, G.F. & AHr-nonN, R.C. (1974): Micagranites of the Kern Mountains pluton, easternWhite Pine Country, Nevada: remobilized base-ment of the Cordilleran miogeosyncline? Geol.Soc. Amer, Bull. 86, 1277-1286.

BneorIsH, LJ. (1979): Petrogenesis ol the Tea CupGranodiorite, Pfual County, Arizona. M.Sc. the-sis, Univ. Arizona, Tucson, Arizona.

34 THE CANADIAN MINERALOGIST

Bnrvnerr., G.H., Jn. (1912): Early Hydrothermal,Wallrock Alteration at Butte, Montana. Ph.D.thesis, Univ. California, Berkeley, California.

Bnonst, D.A. (1962): Geology of the Spruce Pinedistrict, Avery, Mitchell, and Yancey Counties,North Carolina. U. S. Geol. Surv. Bull. \L22-A.

CAsrLE, R.O. & TnsooonE, T.G. (1972): Somegenetic implications of the phase composition ofa simple New England pegmatite. U. S. Geol.Surv. Prof. Pap. 800B, 105-117.

CHoRLroN, L.B. & MARrrN, R.F. (1978): The effectof boron on the granite solidus. Can. Mineral. 16,239-244.

Cnownv, M.S. & Rov, R. (1964): Crystallinesolubility in the muscovite and phlogopite groups.Anrer. Mineral. 45, 348-362.

Dev, H.W. (1973): The high temperature stabilityof muscovite

'plus quartz. Amer. Mineral. 58,

255-262.

DE PoL, C., ScHrAvrNATo, G. & MrNurn, L.(1968): Distribuzione di Na, Rb, Ca, Sr, Ba nelfeldspato potassice dei graniti dell'Argentera(zona centrale del massiccio). Soc. Geol. Itol,Boll. 87, 425-445.

DvMEK, R.F. & Arnss, AL. (1977): Titanium andaluminum in biotite from high-grade Archaeangneisses. Lang$, West Greenland. Amer, Geo-phys. Union Trans 58,525 (abstr.).

GAsrrL, R.G. (1975): Plutonic zones in the Penin-sular Ranges of southern California and northernBaja California. Geology 3, 361-363.

GuIoorTr, C.V. (1978a): Muscovite and K-feldsparfrom two-mica adamellite in northwestern Maine:composition and petrogenetic implications. Amer,Mineral. 69, 750-753.

covite in medium- to high-grade metapelites ofnorthwestern Maine. Amer. Mineral, 63, 878-884.

HARRts, N.B.W. (1974): The petrology and petro-genesis of some muscovite granite sills from theBarousse massif, central Pyrenfies. Contr. Min-erul. Petrologt 45, 215-230.

Jovcr, A.S. (1973): Chemistry of the minerals ofthe granitic Mumrmbidgee batholith, AustralianCapital Territory. Chem. Geol. ll, 271-296.

Lurn, W.C., JeHNs, R.H. & TurrrE, O.F. (1964):The granite system at pr€ssures of 4 to l0 kilo-bars. .I. Geophys. Res. 69, 759-173.

MTLLER, C.F. & lCsH, S.A. (1980): Peraluminoustrondhjemite, Whiteside pluton, North Carolina.Geol. Soc. Amer, Abstr. Programs 12, 2Ol.

---, Srameg 4.K., SToDDARD, E.F., Horscn,T.D. & Hunsl R.W. (1980): Old Woman-PiuteRange plutonic complex, southeastern California.Geol. Soc, Anter. Abstr. Programs 12, l4I.

& Sroooeno, E.F. (1978): Origin of garnetin granitic rocks: an example of the role of

Mn from tfte Old Woman-:Piute Range, Cali-fornia. Geol. Soc. Amer. Abstr. Prograns 10,456.

Mrvesnrno, A. (1973): Metunrcrphis,n und Meta-ntorplric Belts. George Allen & Unwin Ltd.,London.

MonoN, J.D. (1975): Comparative Geothermome-try for the Monte Cristo Pegmatite, YavapaiCounty, Ar'zona. M.Sc. thesis, Univ. Arizona,Tucson. Arizona.

Nrtvn, A.M.R. (1975): Geochemistry of coexistingaplites and p€gmatites and of their minerals fromcentral northern Portugal. Chem, Geol. 16, 153-177.

Nusox, C.A. & Svrvrsrsn, A.G. (1971): Wallrock decarbonation and forcible emplacement ofBirch Creek pluton, southern White Mountains,Cafifornia. Geol. Soc. Anrer. Bull, 82, 2891-2904.

PARKER, J.M. (1979): The geology and mineral re-sources of Wake County, North Carolina. N. C.Dep. Nat. Res. Comm. Devel., Geol. Surv. Sect.Bull. 86.

Slevrou, J. (1978): Geochemical and petrologicalcharacteristics of mineralized granites of the westcentre of Spain. In Metallization Associated withAcid Magmatism 3 (M. Stemprok, L. Burnol& G. Tischendorf. eds.). Geol. Surv. Czech.,279-291.

SweNsoN, S.E. (1978): Petrology of the Rocklinpluton and associated rocks, western SierraNevada, California. Geol. Soc. Amer. Ball. 89.679-686.

Svrwsrsn, A.G.. Ornrel, G., NELsoN, C.A. &CHntsrrr, J.M. (1978): Papoose Flat pluton: agranitic blister in the Inyo Mountains, California.Geol. Soc. Amer. Bull. 89, 1205-1219.

TsorrapsoN, A.B. (1974): Calculation of muscovite-paragonite-alkali feldspar phase relations. Contr.Mineral. Petrology 44, 173-194.

VELDE, B. (1965): Phengite micas: synthesis, sta-bility and natural occurrence. Amer. J. Sci. 263,886-9 13.

(1967): Si+a content of natural phengites.Contr. Mineral. Petrology 14, 250-258,

Wrorvtvrn, R.C. & Srnurr,r, R.K. (1980): Geo-chemistry and geochronology of Sims granite,eastern Carolina State Belt, North Carolina.Ceol. Soc. Amer. Abstr. Programs 12, 2lI.

WHrrNEy, J.A., JoNps, L.M. & Wexnn, R,L.(1976): Agi: and origin of the Stone Mountaingranite, Lithonia district, Georgia. Geol. Soc.Amer. Bull. 87, 1067-1077.

Received May 1980, revised matruscript acceptedOctober 1980.