Canadian MineralogistVol. 16, pp.325-333 (1978)

ABSTRACT

Pollucite forms large lenticular bodies and podsin the upp€r intermediate zone of the Tanco peg-matite, in association with quafiz, microclineperthite, petalite (spodumene * quartz) and ambly-gonite. It was formed as a late constituent ofprimary zonal crystallization and shows no evidenceof metasomatic oriein. The composition of pollu-cite, established chemically or estimated opticallyon 148 samples, is centred at 32 wt,Vo CszO (ap-nroximately PollrgAnalrr). Some pollucite bodiestend to be enriched in Na along their margins. Csand Na are the only major alka.lis; KRb and Liare subordinate. Negative Cs/Na and positive Rb/Tlcorrelations are the only well-defined trends amongthe alkalis. Refractive indices can be roughlycorrelated with tle (Cs, Na) substitution. pollucitebodies are veined by quartz, mostly non-perthiticmicrocline (A 0.92), low albite (/ 1.10.) locallyrimmed by cesian analcime, and purple lithianmuscovite (2M) with lepidolite (lM), A late net-work of braided veinlets carries white lithian mus-covita (ZMr) and some spodumene, followed bysanidine-type adularia. Greenish kaolinite andmontmorillonite (-r calcite) formed last in thealteration sequenc€.

Sorrarrtarns

La pollucite se pr6sente en amas lenticulairesdans la zone moyenne sup6rieure de la pegmatiteTanco, accomDagnde de quartz, microcline perthi-tique, p6talite (spodumdne + quartz) et ambligonite.Elle s'est form6e tardivement lors de la cristallisa-tion zonaire primaire et ne montre nul indice d'ori-gine m6tasomatique. Sa composition, 6tablie paranalyse chimique ou d6duite des propri6t6s optiquesde 148 6chantillons, se situe verc 32Vo (en poids)de CsrO, i peu prds Poll6Analrr. Certains amasde pollucite ont tendance i s'enrichir en sodium,en bordure. K, Rb et Li sont nettement subordon-n6s i (Cs * Na); les corr6lations Cs/Na (n6gative)et Rb/Tl (nositive) sont les seules nettes pour lesm6taux alcalins. ks indices de r6fraction d6pendent

*Publication No. 26, Centre for PrecambrianStudies.

tPresent address: 75 Linden Avenue, Winnipeg,Manitoba R3K 0M7.

THE TANCO PEGMATITE AT BERNIC LAKE, MANITOBAX. POLLUCITE*

P. dNNNf AND F. M. SIMPSONTDepartment ol Earth Sciences, University ol Manitoba, l(innipeg, Manitoba RJT 2N2

grosso modo du dee!6 de substitution (Cs,Na). Lesamas de pollucite sont recoup6s par des filons dequartz, microcline non-perthitique (A 0.92), albiteordonn6e (/ 1.10') Oord6e localement tl'analcimec6sique) et muscovite lithique violac6e QMt) ave*l6pidolite 1M. Des entrelac,s de filonnets tardifscontiennent de la muscovito lithique blanche (2MJet un peu de spodumdne, suivis d'adulaire d6sor-donn6e. Le dernier stade d'alt6ration est repr6sent6par de la kaolinite verdatre et de la montmorillonite(avec ou sans calcite).

Clraduit par la R6daction)

INrnoouctroN

Pollucite was discovered in the Tanco peg-matite in 1960 when Dr. R. W. Hutchinsonnoticed peculiar physical characteristics of whathad been previously logged as a quartz-feldsparintergrowth. Subsequent laboratory examinationby Dr. W. Moorhouse led to its correct identifi-cation @rohberg 1,967). A substantial pollucitebody in the pegmatite was established shortlyafterwards. Recent underground excavations inthe western part of the pegmalile have estab-lished the shape and size of several other bodiesdetected earlier by drilling. The total reserveof pollucite in the deposit exceeds 350,000 tonsaveraging 23.3 wt.Vo CszO; this represents thelargest accumulation of this mineral known inthe world today.

With the exception of Nickel's (1961) de-scription, there are no mineralogical data on theTanco pollucite available in the literature, andits paragenetic position in the pegmatite israther poorly understood ,(Hutchinson 1959,Wright 1963, Crouse & Cernf 1972). Also,mining and exploration at Tanco offer a rrniqueopportunity to study compositional and altera-tion patterns of pollucite bodies in three dimen-sions; all extant descriptions of pollucite fromindividual localities are based on examinationof one or two specimens. These considerationsinitiated the present study; it is based partlyon a thesis by Simpson (1974), partly on theinvestigations of the first author before andsince the completion of this thesis.

325

326 THE CANADIAN IVIINERALOGIST

Senapr.rxo ergo ExpSRIMBNTAL METHoDS

Underground and drill-core sampling yielded148 specimens of pollucite for laboratory study.An additional 30 specimens were collected tosllmins the alteration processes. The suite of148 samples was checked for refractive indices(ru) and 41 specimens were then selected forpartial chemical analyses. Three samples con-sidered representative of pollucite with low,average, and high z were chosen for a detailedmineralogical study.

Pollucite from Tanco is generally homo'geneous on a small scale. Four randomly se-lected hand specimens were checked fot n aI 3to 5 spots on their surfaces. The ranges- of n -tnindividual samples did not exceed 0.001' theequivalent of about 0.6 v'rt.Vo C*O. The onlyexceptions are the specimens veined by latealteration products; these were excluded fromthe study.

- Fine veining by mica and spodu-

mene proved to be much more of an obstaclefor obiaining reliable compositional data thanany inhomogeneity in the pollucite itself. Sevenof the 41 partial analyses had to be discardedbecause of complex contamination of the sam-ple, and 17 other analyses were adjusted for ad-iri*t.rres of spodumene or muscovite, quanti-tatively determined by X-ray powder diffraction.

The methods of optical study, density deter'mindies, chemical analyses, and- X-ray powderdiffraction are as described bv Cernf & Simeson (1977), with the only exception that thestandards used 'for determination of sommonconstituents were heavily spiked with cesium,to balance the Cs content of pollucite and itseffects on determination of other elements.

OccunnnNcp IN THE PBoMATTTE

Pollucite is restricted to the upper parts ofthe eastern and western 1l6nks of the pegmatitewhere it occurs within, or adjacent to, the upperintermediate zone (5 in Fig. 1). This zoneconsists mainly of quartz and giant-size crystalsof microcline perthite, petalite altered to spodu-mene * qaartz, and amblygoilte (cl. Cerni &Simpson L977, Table 1 for detailed mineralogyof all zones). Pollucite forms numerous podsand nodules I Io 2 m in diameter and severallarge lenticular bodies that constitute pollu-citezone (8). The largest of these, 180 X 75 x !2yin size, is located in the eastern flank of theoeernatite. and three smaller bodies are foundioil" *"tt"tn part; their positions are projectedon the section plane in Figure 1.

Masses of pollucite show simple smoothboundaries against quartz, and are anhedral

LAKE LEVEL

Frffiffi

I[F,A

Flo. 1. Longitudinal sectiotr through tle Tanco pecmatiq'.Most of tirceasxern pollucite body is exposed in this section; some of its boundariesshown 6ere and all ihe western pollucite bodres are projected on tlissection to show ttreir full extent.

flflTtTflil|mffiT'Ttv

| d @ a ( 4 ,

@ | @ & 6 t

@ @ a ( a )

POLLUCITE AT BERNIC LAKB

Fro. 2, Typical yeining pattern in the eastern pollucite body, shown hereseparated from blocky microcline perthite (upper left) by a dark selvageof quartz. Dark veinlets of quartz, feldspars and lepidolite dissectpollucite into polygonal blocks. Faintly visible system of subparallelwhitish veinlets, gently sloping to the right, carries muscovite andspodumene. The hammerhead is 20 cm long.

327

with respect to all other pdmary constituents ofzone (5). Pollucite may locally cement fracturedor brecciated silicate assemblages but it is neverobserved as a replacement.

Veining and replacement by quartz, feldspars,micas and spodumene described later in-thisarticle are widespread, though minor in volume(Fig. 2). These constitute the only major con-taminants of the bulk of the pollucite bodies;these were essentially meneminslslic beforethe onset of secondary processes.

CotrposltroNer, VanresrlrryThe variation of refractive index of pollucite

with the Cs/(Na*H:O) substitution has beenwell established (e.g., Richmond & Gonyer 1938,Nel 1944, denf L974. Thus an optical surveyof 148 pollucite specimens was undertaken asa first step in determining their compositionalrange.

The mean value of n is 1.5206, with a stand-ard deviation -+- 0.0013 that encompasses most

TABLE l. ATKALI CoNTENTS AND n 0F THE TANC0 P0LIUCI'IT

Zone Sample # Li20 Na20 K20 Rb20 Cs20 Tl* n

!o

Eo

7

R-HR-2-3R-3-4R-4-2R-2-6r - t - cP-2-4P-4-6P-3-2P-4-4834- r 47. 5834-1 57.5834- l 87.5834-206.6M5-205.5M4- '165.5M4- 1 63.5

**u39-95.8U J Y - O Z . !

u39-75.5fB12-207.2nB12-226.2

. 09

.06, l lno

. I tt o

. l l

. 2 5

.08

t q. 2 0. 1 4no

. 2 3

. 2 8? l

.29

.43

' r . 88 . t 2

t . t u . a a' r .65 .09t a E n q

t . o d . J /I . 6 4 . l t' | .87 . t5' I . 6 3 . t 9

1 . 7 2 . 0 7

r . 84 . 08i . 7 6 . t 01 . 6 2 . 2 5' r .68 . 10

I . 46 . 191 . 4 3 . 2 0' I . 3 4

. l 9' I . 98 .17' 1 . 9 6

. 1 81 . 7 3 . 2 3' I . 31 . l 0

1 . 7 4 . 1 6

1 . 4 5 . 2 4 . 7 7l E a ? q e ?' I . 45 . 15 . 67' I . 51 . 22 1 , 22

t . 6 t . a I . J 3r . 48 . 20 . 851 . 4 8 . 1 6 . 8 91 . 3 r . 25 . 941 . 4 8 . l 9 I . 1 61 .60 . 22 . 201 . 7 0 . ? 6 . 1 91 . 7 0 . 2 0 . 5 1l . ! ! . t f , . 1 6

1 .48 . 21 . 811 . 4 2 . 2 0 1 . 0 6

.71 32 .90 1 .522,94 32.65 105 1,5224,79 34 .10 1 .524.89 3t .35 1 , s l 98.86 3l .50 1 .522.97 33.04 1.5224.88 33 .20 1 .5225.74 32.00 1.5205.72 3? .20 I . 5198.85 31 .56 1 .5 i 92.80 31 .52 1 .5185' I . 10 32 .00 1 .5205

.59 3 l . 60 1 . 522

.73 32.40 1.5202

.82 32,26 1.5202

.80 3 l . 67 1 .5206

.82 32 .03 I . 5218

.77 30 .60 r00 1 .517

.44 29 .67 1 .518

.86 31.27 r .5204

.68 32.57 80 1.5232

.74 32.20 '1.5204

plimary pollucites from other pegmatites(Cernf 1974, Fis, L). Extreme values are 1.517anld t.524. Frequency distributions of z valuesin eastern and western bodies are very similarand close to gaussian when combined.

Chemical analyses of 37 specimens yield amean of 31.96 wI.Vo CsrO, with standard devia-tion :L .96 and extreme values of 29.67 alid34.10 (fable 1). This mean is also close to theCssO content most frequently quoted for pri-mary pollucites. However, frequency distribu-tions of CsrO in the eastern and western bodiesand in all the bodies combined are asvmmetric.

Analysts: R.M. Hi'11 and R. Chapman (Dept. Earth Sciences"Univ. of ltlanitoba)

* 'lhallium detenninations by C. thllaire (Laboratoire deGeochimie Analytique, Ecole Polytechnique, Montreal) ;data glven in ppn.

* Full analyses of these specinEns arc given in Table 2.

M20- 1 79.5 .53M?0-181.2 .41u - a u l . 5 . 5 0TL-19 . i lTL-20 .15TL-21 .21TL-84 .1 6TL-85 .12TL-86 .14CsA-5 .12CsA-8 .'17CsA-3 .18csA-7 .12llEx-c . t 3l,lEx-D .18

33 .10 1 .523830 .70 1 .520231 .75 i . 5203 r . 50

' r 69 I . 5185

30.20 47 1.517633.60 107 1.522833 .60 120 1 . s2232 .80 108 1 .521232.20 169 1.520831 .40 1 .519430 .90 I . 51 983 l . 50 1 .521432.30 1.523431 .20 I .51 9632. 30 I .5214

328

33

wt%Csa O

3 t

THE CANADIAN MINERALOGIST

ta a

- a -

ta '

^

nFIc. 3. Plot of wt.Vo CszO against refractive indices

n of the Tanco pollucites; dots, eastern body;triangles, w€stern bodies. Distinct positive corre-lation with considerable lateral dispersion. Thecross indicates standard errors of compositionaland optical data.

skewed towards high values. Correlation of nwith Cs"O shows that the refractive indices canbe used as an approximate measure of the Cscontent of Tanco pollucites, despite consider-able transversal scatter (Fig. 3).

Optical data were used in assessing possibleregularities in spatial distribution of composi-tional variations. No meaningful patterns werefound in horizontal sections. Several verticalprofiles through the eastern body reveal Cs/Naratios decreasing along, or close to, its margins'However, this pattern is not persistent, and someprofiles show a rather erratic variation.

MTNSRAI-ooy on TrnBn SPEcIMENS

Tbree specimens representative of polluciteswith low, intermediate, and high n were selectedfor a detailed study. The data obtained arelisted in Table 2.

The crystal chemistry of these pollucitescorresponds to the general characteristics of thespecies as derived by Beger (1969) and con-firmed bv eernf (1974). The Si/Al ratio istypically higher than 2.00, and the chargebalance Al3+/( ++2R2+) deviates from the idealvalue of 1.00 only within the limits of analyticalerror. The sum of Cs * HrO should equal 1.00but is somewhat lower, possibly due to incom-plete dehydration of pollucite on heating(Fleischer & Ksanda 1940, Barrer & McCallum1951). The unit-cell dimensions show no appre-ciable variation with the changing cationic con-tent as they are significantly influenced onlyby the Si/Al ratio.

Generally, the relationships between differentcrystallochemical and physical properties corre-spond to those established earlier (Cenf 1974,Figs. 3,4,6 and 7). The only noteworthy devia-tion within individual sets of otherwise well-balanced properties is a difference between themeasured and calculated densities. This variesbetween 0.03 and 0.04. with the measuredvalues invariably higher than the calculated data;a reverse relationship is much more common.This puzzling discrepancy will be investigatedin a crystallochemical study of pollucite inprogress.

Figure 4 shows relationships between pairsof alkali elements. based on data from Table 1.

A A

. t , l^ . o

a

a ' ab .t a t

a

TABU 2. CHgitCAr @mosntd Am ft|tStCA! PRoPERTIIS 0F nruE m[ltrllt sA]fLES

812-n7.2 [39-95,8 S12-46.2 812-207.2

slAI

sr02Ala%FuZo:}!0C!0Lt20&zo

t0saoCs20H2sHao-

47. @'15.70

.01

.(tr3

.04

. 81 . 9 8. 1 ?.77

30.60

.(E

46.09

.01

.@3

.0t

.431 . 7 4

. 1 6

.7432.m

.04

47.42I 6.02

. 0 1

.01

.06

. aI .30

. 1 0

32.57

.04

2.161 2.112

.836 .884

X etr,sr/ArCaMLIsXCailg

2.997 2.W 3.007

2.58 2.t9 2.51

.589 .6A .530

. 1 7 3 . 1 5 6 . l 1 4

.05t .079 .053

,02. ,022 .0m

.ol0 .009 .006

.@e - .@3.003

99.54 9.53

D @6. (e/@:)r 2.s 2.8

o @lc. (s/@r) 2.81 2.85

f r . 5 r 7 r . 5 2 0 4

A^(l) 13.571(r ) 13.67211l

vi l3t z5so.76(43) zs5s.57(43)

X (n"rnzt) .g47 .89g .a2999.87

2 . W2.831.5232

13.672(l )2555.57(43)

H2o

Ca+H20Pol l+cnk

'

.33 .27 .21

.92 .90 .84

69.6 70.2 75.7

73.2 73.7 n.7

A@l$t3. ( nelll, R.lt. Hlll, and i. Ch!@n (DoPt. [!rth scls@s, Unlv. of ilanltoba).

A&lc @!on!9 ba6od on r 3ubcsll ,tti 7 orygoa ln lnhydrcB fo@18.. AEAgs of 3 to 4 9nln9 of !6lyz€d dtoial.F Aveago of.nllyzod @terl!I, ,lth f1!.t!a!to6 e ! 0.001.

a pol l . (cs r '100)/(ca { Ll i t | ! r Rb + K).* cRr ! [ (ca + nb+ () r loo]/(ca + Rb + ( ] Ll ] tu).

POLLUCITE AT BERNIC LAKE 329

r3o 140 1 .50 160 t .70 1 .80 r90 200N o a 0 , w t %

20 .40 .60 .80 1.00 t.20Rb2O, wt. %

The plot of CsrO vr. Na2O displays the expectednegative correlation but also a wide scatter ofpoints, attributable to either variable contentsof other alkalis, to slight variation in the Si/Alratio which regulates the alkali total, or torelatively high standard deviation in the cesiumdetermination. RbzO shows a poor positivecorrelation with CszO, in accordance with othergeochemical and crystallochemical data on thesealkalis (e.9., Stavrov 1963). Surprisingly, theI(zO v^r. RbzO diagram shows a random scatterof data, unless we accept the predominant hori-zontal belt of points as indicative of variableRb at a constant K content. In contrast, theRb/Tl ratio shows a well-defined correlation,with individual values ranging from 82 to 63.These are the lowest found in the silicates ofthe Tanco pegmatite (dernf, unpubl. data) butstill lie within the range of 300-65 given byde Albuquerque & Shaw (1972) for terrestrialigneous rocks.

ALTERATIoN PRocEssEsAs in most other localities, the pollucite

bodies in the Tanco deposit underwent severalstages of low-temperature hydrothermal altera-tion. Three consecutive alteration stages may bedistinguished: (i) coarse polygonal veining bymicrocline, albite, quartz and lepidolite; (i1)fine braided veining which carries spodumene,lithian muscovite and adularia, and (iii) replace-ment by kaolinite and montmorillonite (:Lquattz, calcite).

Frc. 4. Major and minor alkalis in the Tanco pollucite. The CseO/Na2Oand CszO/Rb2O correlations are poorly defined, and the KrO/RbrOplot seems to be inconclusive, whereas the Rb,/Tl diagram shows goodpositive alignment of data. Symbols as in Figure 3.

40 60 80 t@ t20 r40 l@T l , p p m

Coarse veins

Coarse veins, 5 to 40 mm across, fill inpolygonal fracture patterns as shown in Figures1 and 5a. Microcline is usually confined to themarginal parts of pollucite bodies, whereas al-bite and quartz penetrate farther. Telescopingof these three minerals along a single fractureis common. Columnar spodumene is a rare con-$titutent of some of the quartz veins. Albite is

TABLE 3. PROPERTIES OF MICROCLINE

NaZo KZO Rbr0 Csro Cao Pbo

CsaO

wf.70

33

351- (b)I

34fI

"fcgo rz l-- l

w f . % - . ^

' 1 . . .

(c ).T*ft

"'"

K2o

wr.%

Ptl I - l 7

PMI-I 9

PMI-20

' t .77 13 .001 . 0 8 1 3 . 6 3. 2 s I 3 . 0 0

2 . 1 9 . 2 42.42 .283 . 6 I . 2 1

( .86 - .e5)( 7 - 0 )

.07 .002,17 .00r.004 .001

9 0 0 3 3 . 7 ( r . S )

I 1so56.6( ' r .6 )87053. ' l (1 .7 )

721 .3 (3)83

ob l iqu i ty * ( l l ) : .92(AbrAn)r.** ( l l ) : n

U . c . d ' s o f P M I - 1 9 :dmeas. 8.594(4)R

(upr"d.*t*) s.4sff

b l 2 . e5r (3 )A

c 7 .211 (2 ) f i

o(

prV

n1nr.mb"" of samples): arithrnetic mean (range).

**by x-r"y diffraction (0rv'llle ]967); othenrlse as

under*.*n*f"*

ste$art t Wright (1974, Fig. 1).

330 THE CANADIAN MINERALOGIST

Frc. 5. Late veining in the Tanco pollucite: (a) dark nonperthitic microcline in whitish pollucile, locallywith a black halo of adularia aird clay minerals (polished hand specimen); (b) dark veinlets of mus-covite and spodumene penetrating clear pollucite 6hin section, polarized light); (c) white microclinecorroded by- grey aduliria along contacis with isotropic pollucite (thin section, crossed polars); (d)grey fine-giaiol"d-adol-iu on whlte coarsely granular microcline and along white veinlets of muscovitein isotropic pollucite (1hin section, crossed polars).

1.13o, indicating a well-ordered structure (Smith1974, Fig. 7-43).

Pollucite in the immediate vicinity of thealbite veinlets occasionally shows changes inX-ray powder diffraction intensities, and re-fractive indices as lolY as 1.500. Thase proper-ties indicate cation exchange or recrystallizationinto cesian analcime eernf 1974, Figs. l,- 8),due to the action of albite-producing solution$from the adjacent veins.

Lepidolite consists of a mixture of 2Mt andl,M polytypes; it is deep reddish purple, fineflaked to

-riassive, with r-efractive indices in the

cleavage plane between 1:570 and 1.580. Theseproperties match those of massive lithian mus-covite + lepidolite of zone (9) (Rinaldi er at'1.972, samples R7, R9, REN5).

F in e-braid.e d. v eining

A network of subparallel, locally braidedveinlets of white muscovite with spodumene

occasionally accompanied by manganotantaliteor wodginite. Lepidolite penetrates those partsof pollucite bodies located close to lepidolitemasses of unit (9). Where associated with thefeldspars, lepidolite is always of later origin,filling central parts or margins of the feldsparveins.

Compositional and physical characteristics ofthe microcline are given in Table 3; represent-ative veinlets are shown in Figures 5a, c. Theeiratic variation in the Na:O content found bychemical analyses seems to be caused by vari-able contamination of the analyzed material bygranular albite. The nonperthitic nature of thismicrocline, and the rather low and uniform(Ab + An) content determined by X-ray dif-fraction support this explanation.

Albite shows an almost constant d - I.529on (001) cleavage fragments; this correspondsto Ano-r (Morse 1968). Its average obliquityvalue is 1.10'. with variations from 1.05o to

POLLUCITE AT BERNIC LAKB 331

penetrates most bodies of pollucite (Fig. 5b).Its orientation shows no relation to tle bound-aries of polygonal blocks framed by the pre-ceding vein type, and it often holds for severalmetres s/ithout significant change. The mica-ceous veinlets occasionally penetrate and cross-cut the feldspar veins.

Lath-shaped crystals of spodumene are irre-gularly dispersed or clustered in the mica vein-lets, and are partly replaced by them. Spodu-mene may repre$ent early inclusions in pollu-cite, connected later (as centres of mechanicalweakness) by the braided fracture system in-vaded by mica-generating solutions. This spodu-mene, however, has nevet been found outsidethe micaceous veinlets, and most of it isoriented subparallel to their course. Thus itmost probably represents an early replacementproduct that preceded muscovitization.

Muscovite yields X-ray powder diffractionpatterns of. thie 2Mt polytype, and the averagez on its cleavage flakes varies between about1.600 and 1.590. This indicates that the second-ary mica yaries between typical and somewhatlithian muscovite in composition.

The muscovite veining was locally followedby a second generation of K-feldspar. Fine-grained adularia replaces microcline of thecoarse vein system and pollucite along theircontacts, and it also grows from the marginsof nearby muscovite veinlets into the pollucite(Figs. 5c, d). Gandolfi X-ray diffraction photo-grapbs show the adularia to be of a high sani-dine structural state.

Argillic replacement

Clay minerals occur in two major forms.Most abundant are white spheres up to 3 mm insize, randomly disseminated through the mus-covite-veined pollucite. These consists of dusty"clouds" producing faint 14A and 7A reflectionsin Gandolfi photographs. Restricted to the mar-gins of pollucite bodies are apple-green patchesand veinlets of illite, montmorillonite and kao-linite, determined by optical, X-ray powder dif-fraction and differential thermal methods(dernf 1978). Larger pods of the green clayminerals usually contain some quartz and cal-cite. Partial chemical analysis of the green clay,in part contaminated by residual pollucite, gavecsso 5.00, cao 4.95, Kro 4.94, NazO 1.92,MgO 1.76, Rb,O 0.18 and LizO 0.09 wt,Vo.

GBNnrrc CoNspsRATroNs

The position ol pollucite in the crystallizationsequence

The timing of pollucite crystallization in peg-

matites has been discussed in literature sincethe recognition of pollucite as a mineral species.Amongst modern authors Dymkov (1953),Quensel (1956), Ginzburg (e.e., 1960) and Solo-dov (1960) advosate a late metasomatic originwhereas Beus (1960), Melentyev (1961), Sy-mons (1961), Cooper (1964) and Heinrich (1976)consider pollucite as a primary constituent ofcentral parts of pegmatite bodies.

In the Tanco pegmatite, Hutchinson (1959)and Wright (1963) interpreted pollucite bodiesas late metasomatic features, in a close spatialrelationship with the obviously metasomaticlepidolite units (9), developed at the expenseof microcline-rich zone (6). Wright also observedpollucite replacing spodumene on a microscopicscale. As previously indicated '(Crouse &eern.i 1,972), the evidence is questicnable; morerecent work supports the primary nature ofpollucite.

The distribution of lepidolite units generallyfollows, though not closely, that of the pollucitebodies. Large segments of pollucite segregationsare not underlain by lepidolite, and some lepi-dolite units lack ttre overlying pollucite. Lepi-dolite and pollucite bodies rarely come intodirect contact; they are usually separated by aquartz selvage, and lepidolite veins commonlycriss-cross pollucite. Lepidolite bodies are foundmainly withiu the K-feldspar-rich central inter-mediate zone (6) and along the lower marginsof the upper intermediate zone (5), whereaspollucite bodies are confined to the latter. Thusthe differences in distribution and relative agerule against cornmon origin.

The pollucite-like mineral observed byWright (1963) as a replacement of spodumeneis most probably cesian analcime, a late hydro-thermal alteration product (eernf 1972), It wasrecently found to vein and replace eucryptiteand amblygonite. Primary pollucite has neverbeen observed to replace petalite or spodumene* quartz aggregates formed by its retrogradethermal breakdown.

The evidence pointing to the primary natureof the Tanco pollucite may be summarized asfollows: (i) Pollucite occurs in lenticular, loaf-shaped or ellipsoidal bodies, within or in directcontact with the upper intermediate zone (5).This zone consists of giant-size petalite, ambly-gonite and microcline perthite imbedded inabundant quartz. Smdl rounded blebs of pollu-cite n'float" in quartz among crystals of theabove-mentioned minerals without any texturalindication of a late origin. (ii) Simple outlinesand rather straight boundaries of even the larg-est pollucite bodies show no indication of cross-cutting or metasomatis relationships with other

332 TIIE CANADIAN IVTINERALOCIST

minerals or assemblages. Pollucite is anhedraiagainst other minerals but does not replacethem. Petalite adjacent to ttre pollucite bodieslocally contains round grains of pollucite, regu-larly distributed, and somewhat graphic in out-line. This may suggest a simultaneous crystal-Iization of both minerals. (iii) In close vicinityto the albitized zone (6), pollucite is veined byalbite containing manganotantalite and wd-ginite. Fine-flaked lithian mussovite and lepi-dolite filling fractures in pollucite show thesame physical properties as the micas in lepidolite units (9). Thus, both well-establishedmetasomatis events in the pegmatitq the(Ta,Be,Sn,Zr,HD enricbment associated withalbitization and the lepidolite units, postdatethe crystallization of pollucite.

C ompositional v ariations

Most of ttre Tanco pollucite is rather uniformia ssmposition but a considerable percentage ofsamples covers a range of about PollegAnalrr -PollzoAnalgo Na-rich compositions occasionallytend to be located close to the margins of pol-lucite bodies; otherwise tleir distribution israther erratic. Available evidence suggests thatat least some of the variation in Cs,/Na comesfrom alkali exchange along feldspar-bearingveins; the presence and extent of primary varia-tions cannot be evaluated at present.

Late alterations

High density sf yeining in the marginal partsof pollucite bodies, and a roughly zoned se-quence of alteratio,n products from the marginsinward suggest an external source of alterationfluids. The principal veining and replacemenlassemblages found in the Tanco pollucite areremarkably similar to those described fromother ogcurrences (Ginzburg 1946, Quensel1956, Neuvonen & Vesasalo 1960, Melentyev1961). Albite, lepidolite and late muscovite(oogilbertite", "oncosine") veining is most wide-spread. K-feldspar veins seem to be typical ofTanco, whereas veinlets of fine spodumene andspodumene-bearing symplectites with otherphases seem to be more characteristic of otherlocalities. The somewhat varied but recurrentpattern of pollucite alteration at many localitiesstresses the fixed position occupied by thismineral in the crystallization sequence of com-plex granitic pegmatites and the qualitativelyuniform character of low-temperature hydro-thermal processes.

SuccnsrtoNs non Futuns RESEARcH

A review of the compositional variations and

physical properties of the natural members ofthe analcime-pollucite series is in progress, un-der the guidance of the first author. New findsof pollucite-like minerals should be checked forphysical properties and thoroughly examine{when sodium-rich compositions are indicated.

Laboratory synthesis of pollucite has beensuccessful over a wide range in temperature(Barrer & McCallum 1953, Barrer et al. L953,Plyushev 1959, Kume & Koizumi 1965, Kopp& Clark 1966, Richerson & Hummel 1972, Suitoet al. 1974, Martin & Lagache 1975). Thestability field, however, has never been estab-lished for either pure Cs-pollucite or Na-bear-ing phases comparable to natural minerals.Similarly, pollucite crystallization in simplifiedlithium pegmatite systems has never been simu-lated experimentally. Research projects intothese aspects are highly desirable.

AcrNowr,epcEMENTS

This study was supported by National Re-search Council of Canada Operating Grants tothe first author and to Dr. R. B. Ferguson, bythe Department of Earth Sciences, Univenity ofManitoba, and by the D. E. M. R. ResearchAgreements awarded to the first author h 1973and 1974. The authors wish to thank the man-agement of the Tanco mine for their coopera-tion and hospitality, in particular Messrs. C. T.Williams and J. C. Hayward. Mr. R. A. Crouseprovided the section of the pegmatite. Thanksare also due to K. Ramlal, R. M. Hill and R.Chaprnan @ept. of Earth Sciences, Universityof Manitoba) and C. Dallaire (Laboratoire deG6ochimie Analytique, Ecole Polytechnique,Montr6al) for carrying out the chemicalanalyses and to Dr. J. Ito (James Franck Insti-tute, University of Chicago) fol helpful advice.

REF'ERENcES

BARRER, R. M., BevNneu, J. W. & McCu-r-uu, N.(1953): Hydrothermal chemistry of silicates V.Compounds structurally related to analcite. J.Chem. Soc. Lond., 40354047.

& McCer,r,urvr, N. (1951): Intraoystallinewater in pollucite. Nature L67, L077.

& - (1953): Hydrothermal chemistryof silicates. IV. Rubidium and caesium alumino-silicates. l. Chem. Soc. Lond, 4029-4035.

Brcnn, R. M. (1969): The crystal structure andchemical composition of pollucite. Z. Krist..Lz9,280-302.

Bsus, A. A. (1960): Geochemistry ol Berylliumand Genetic Types of Beryllium Deposits. Tzdalel.Akad. Nauk S.S.S.R., Moscow (in Russ.).

dERNi,, P. (1972): The Tanco pegmatite at BernicLake, Manitoba. VItr. Secondary minerals fromthe spodumene-rich zones. Can. Mineral. tl,7t4-726.

(1974): The present status of the analcime-pollucite series. Can. Mineral. L2, 334-341.

(1978): Alteration of pollucite in somepegmatites of southeastern Manitoba. Can.Mineral. 16, 89-95.

& SrrursoN, F. M. (1977): The Tanco peg-matite at Bernic Lake, Manitoba, D(. Bervl.Can, Mineral. t5, 489-499,

Cooren, D. G. (1964): The geology of the BikitaFgmatite. In Tbe Geology of Some Ore Depositsi! S,outhern Africa, 2 (S. H. Haughton,

- ed.),

Geol. Soc, S. Africa, Johannesburg.

CRousE, R. & eERNi, p. (1972): The Tanco peg-matite at Bernic Lake, Manitoba. I. Geology andparagenesis. Can. MineraL ll, 591-608.

on Ar,ruqunnqus, C. A. R. & Sruw, D. M. (1972):Thallium. /n Handbook of Geochemistry IIl8,Springer-Verlag, Berlin.

DyMKov. Yu. M. (1953): Morphology of pollucitedeposits and their genesis. Trudy Mineral. Mu-seya Akad. Naa& ^!.J.J.R. 5, 132-L45 (in Russ.).

FLETScHER, M. & Ksaxte, C. J. (1940): Dehydra-tion of pollucite. Amer. Mineral. 25, 666-6i2.

Fnounrno, M. H. (1967): Presidential address. Can.Mineral 9, 278-281,

Grxzoung A. I. (1946): Pollucite in pegmatites ofthe Kalbin Range (eastern Kazakhstan). Dokt.Akad. Nauk S.S.S.R. 52, 335-377 (in Russ.).

(1960): Specific geochemical features oftlre pegmatitic process. Rep. 21st Int. Geol. Cong,t7, ItL-tzl.

HntNnrcn, E. W. (1976). A comparison of threemajor lithium pegmatites: Varutriisk, Bikita, andBernic Lake. U.S. Geol. Surv. prof, pap. t008,50-54.

HurcruNsorrr, R. W. (1959): Geology of the Mont_gary pegmatite. Econ. Geol. 84, ISZS-L542.

Korr, O. C. & Cum, G. W. (1966): Syntheticpollucites in the system Cs2O.AlrOs.4SiO, _CsrO.Fe2O..4SiOe - H2O - their phase relation-ships and physical properties: disCussion. ,4.rzer.Mineral. 51, 1243-1245.

Kuvre, S. & Korzurur, M. (1965): Synthetic pollu-cites in the system CsrO.AlrOs.4SiO,Cs2O.psrQr.{SiO, - HrO - their phase relation-ship and physical properties. Amer. Mineral. 80.587-592.

MARrnr, R. F. & Lacacua, M. (1975): Cell edgesand infrared spectra of synthetic leucites andpollucites in the system KAlSirO. - RbAlSirO6 _CsAlSizOu. Can. Mineral, 18, 275-28L.

Mnr.nNrvnv, G. B. (1961): The first finding ofpollucite in the granitic pegmatites of Sayan.Dokl, Acad. ^lci. UJ.,S.R. Earth Sci. Sect. i4L.1287-t289.

POLLUCITE AT BERNIC LAKE 333

NeL, H. J. (1944): Pollucite from Karibib, SouthWest Africa. Anter. Mineral. 29, 443-452.

NEwoNE\ K. J. & VEsAsALo, A. (1960): Pollucitefrom Luolamiki, Somero, Finland. Comm. G6ol.Finlande Bull. 188, 133-148.

NIcKEL, E. H. (1961): The mineralogy of theBernic Lake pegmatite, southeastern Manitoba.Mines Branch Tech. Bull. 20.

Onvu.rr, P. M. (1967): Unit-ceU parameters of themicrocline-low albite and the sanidine-high al-bite solid solution serbs. Arner. Mineral. 52,55-86.

Plvusurv, V. E. (1959): High-temperature syn-thesis of pollucite. Zap, Vses. Mineral. Obshchest.88, 152-156 (in Russ.).

QUENSEL, P. (1956): The paragenesis of the Varu-triisk pegmatite, including a review of its mineralassemblage. Ark. Mineral. Geol. 2, 9-125.

Rrcrcnsox, D. W. & HUMMEL, F. A. (1972): Syn-thesis and thermal expansion of polycrystallinecesium minerals, J. Amer, Ceram. Soc. 55,269-273.

RIcHrv{oND, W. E. & GoNven, F. A. (1938): Onpollucite. Amer. Mineral, 23, 783-789.

RrNar,or, R., dpnNt', P. & FsncusoN, R. B. (19721:The Tanco pegmatite at Bernic Lake, Manitoba.VI. Lithium-rubidium-cesium micas. Can. Mineral11,690-707.

SrnarsoN, F. M. (1974): The Mineralogy ol Pollu-cite and Beryl lrom the Tanco Pegmatite atBernic Lake, Manitoba. M.Sc. thesis, Univ. Mani-toba, Winnipeg, Man.

Surnr, J. Y. (1974): Feldspar Minerals. I. CrystalStructure and Physical Properties. Springer-Ver-Iag, Berlin.

Sorooov, N. A. (1960): Dstribution of alkali metalsand beryllium in minerals of a zoned pegmatiteof the Mongolian Altai. Geochem. 874-885.

Sravnov, O. D. (1963): Principal features of geo-chemistry of Li, Rb, and Cs in the process ofconsolidation of granites and associated p"g-matites. Geol, Mestorozhd. Red. Elementov 21.1-141 (in Russ.).

SrEwARr, D. B. & Wnrcrrr, T. L. (1974): AllSiorder and symmetry of natural feldspars, andthe relationship of strained cell parameters tobulk composition. ,loc. frang. Mindral. Crist. BuIl.97,356-377.

Surro, K., Lacevr, A. & Irvaue, J. T. (1974): Sta-bilit6 des solutions solides de la s6rie pollucite-leucite sous une pression d'eau de 30 kb, C. R.Acad. Sci. Paris 278(D), 2397-2400.

SvuoNs, R. (1961): Operation of Bikita Minerals(Private), Ltd., Southern Rhodesia. Bull. Inst.Mining MeL 661, 1.29-172.

WRIcHT, C. M. (1963): Geology and origin of thepollucite-bearing Montgary pegmatite, Manitoba.Geol. Soc. Amer. BuIl. 74. 919-946.

M9ns1, S. A. (1968): Revised dispersion method Received January 1978, revised manuscript acceptedfor low plagioclase. Arner. Mineral. 5A, 105-115. May, 197g.