American Mineralogist, Volume 67, pages 483493, 1982

Lithiumatuminosiliffie,""""r:iffi ilTJsti5il.xlll"sandthelithium

Devro LoNooNr eNo DoNer-o M. Bunr

Department of GeologyArizona State University

Tempe, Arizona 85281

Abstract

The minerals eucryptite (LiAlSiO4), spodumene (LiAlSi2O6), and petalite (LiAlSi4Or6)are important constituents of lithium pegmatites, yet the stability relations of theseminerals are poorly understood. Natural occurrences and available experimental data havebeen utilized to construct a P-T diagram for phases in the binary system LiAlSiO4-SiO2.The calculated lithium aluminosilicate stability relations in the degenerate (n+3)-phasemultisystem eucryptite, spodumene, B-spodumene (p-LiAlSi2O6), petalite, and quartzimply that eucryptite + quartz is stable at low P and T, spodumene is stable at high P,spodumene + petalite is stable at moderate P and ?, petalite is stable at moderate Tand lowP, and B-spodumene is stable only at high T. The proposed lithium aluminosilicate phasediagram accounts for the stable occurrence of eucryptite + qluarlz at low P and T, and forthe observed replacement of petalite by spodumene + quartz (at relatively high P anddecreasing T) or by eucryptite -l quartz (at relatively low P and decreasing Z). The phase B-spodumene is probably stable only well above the lithium pegmatite solidus, thusaccounting for the absence of p-spodumene in nature.

Introduction

The three lithium aluminosilicates, eucryptite(LiAlSiO4), spodumene (LiAlSi2O6), and petalite(LiAlSi4Olq) have long been of interest to mineral-ogists, igneous petrologists, and economic geolo-gists (e.g., Julien, 1879; Brush and Dana, 1880)(Table 1). Much of what is known about the geologyof these minerals stems from field studies of pegma-tites conducted during World War II, when thewartime demand for mica and beryl was high. Theresults of these and subsequent investigations aresummarized by Cameron et aI. (1949,1954),Page etal. (1953), Jahns (1955), eern! (1975), and Heinrich(1975). In addition, studies of pegmatites in theSoviet Union have provided a large body of infor-mation pertaining to the geology of lithium pegma-tites and occurrences of the lithium aluminosili-cates; much of this work is summarized by Fersman(1931), Ginzburg (1955), Beus (1960), Vlasov (1961,1964-66), Gordienko (1970), and Ovchinnikov andKuz'menko (1976.

Of the three lithium aluminosilicates, spodumene

1 Present address: Geophysical Laboratory, 2801 UptonStreet, N.W., Washington, D.C. 20008.

0003-004)v8210506-0483$02. 00

is by far the most widespread and abundant. Peta-lite is second in abundance and distribution, and itconstitutes the principal lithium aluminosilicate inseveral pegmatite districts. Some of the importantlocalities and modes of occurrence of spodumeneand petalite are summarized by Heinrich (1975).Eucryptite is the rarest and least abundant of thesethree minerals; however, it has now been reportedfrom at least twenty localities (Table 2). The rarityof eucryptite may be due in part to the fact that it isdfficult to identify (its name stems from the Greekeu kryptos, meaning well-concealed). It commonlyoccurs in extremely fine-grained intergrowths withalbite or qtrartz, and its optical and physical proper-ties are virtually identical to those of quartz (e.9.,eerny,1972).

Pegmatitic deposits of spodumene and petaliteare mined throughout the world for use in glassceramics and in the manufacture of numerous lithi-um-based compounds (Heinrich, 197 5; Yine, 1976;Singleton, 1979).The only known economic depositof eucryptite is in pegmatites of Bikita, Zimbabwe(Cooper, 1964). In addition to lithium ores, ores ofberyllium, niobium, tantalum, tin, cesium, rubidi-um, bismuth, and other rare elements commonly

483

484 LONDON AND BTJRT: LITHIUM ALUMINOSIUCATE PHASE DIAGRAM

Table l. Names and compositions of phases discussed in thispaper

Pha se Abbreviat ion used Conposlt lon

eucrypt i te

spodumene

p e c a l l t e

quartz

B-eucrypt i te

B-spodunene

Ecr

spd

P e t

q tz

Bec

B s P

o-LiA1S 1O4

a-LlA1S 1205

LiAlsi4olO

sio2

B-LlAlS104

B-LlA1Si2o6

are concentrated in the inner, lithium aluminosili-cate-bearing zones of pegmatites (Vlasov, 1961);the richest of these ores typically occur in albite orlepidolite masses that appear to have formed at theexpense of spodumene-bearing pegmatite (Beus,1960; Vlasov, 1961). Therefore, the processes re-sponsible for the formation and destruction of lithi-um aluminosilicates also may be responsible for thedramatic concentration of rare elements in thesezones. Studies of lithium aluminosilicate stabilitiesmay yield information pertinent to rare-metal min-eralization and thus may aid future geochemicalexploration for these elements.

The purpose of this study is to define the condi-tions of formation and alteration of the lithiumaluminosilicates in terms of pressure, temperature,cooling history, and alteration by late fluids. Themelting and crystallization equilibria of systemscontaining lithium aluminosilicates have been stud-ied extensively at low pressures; phase relationsnear the solidus in the system NaAlSi3Os-LiAl-SiO4-SiO2-H2O at 2 kbar PH,s zr€ discussed byStewart (1978). In contrast, the subsolidus stabil-ities of the lithium aluminosilicates have receivedrelatively little attention, even though lithium alu-minosilicates, especially spodumene, commonlydisplay a variety of replacement textures and as-semblages that result from late-stage pegmatiticprocesses (e.9., Brush and Dana, 1880: Graham,1975; Jahns and Ewing, 1976, Chakoumakos, 1978).This paper will focus on the subsolidus stabilities ofthe three lithium aluminosilicates named above.Virgilite, Li*Al*Si3-*O6 (French, Jazek, and Apple-man, 1978), and bikitaite, LiAlSi2Oe .HzO (Hurl-but, 1958; Phinney and Stewart, 196l; Drysdale,1971) will not be considered here. This paper pre-sents a qualitative P-Z phase diagram for the por-tion of the lithium aluminosilicate system LiAl-SiO4-SiO2 (Table 1) that is relevant to geologic

conditions. In a following paper (London and Burt,1982b) we will consider the effects of added compo-nents in the melt or coexisting fluid (as opposed tosolid solution components) on lithium aluminosili-cate stabilities. The preliminary results of experi-mental phase equilibrium studies in the systemLiAlSiO4-SiO2-H2O are available in London(1981) .

Nomenclature

The numerous pegmatite classification schemes(e.9., Cameron et al., 1949; Jahns, 1955; Brotzen,1959; Beus, 1960; Vlasov,19611' Uebel, 1977)haveestablished a fair degree of uniformity in the termi-nology applied to the types and internal features ofpegmatite bodies, yet there are some significantdifferences among these models in their proposedcrystallization sequences and zonal assemblages.These discrepancies stem largely from a lack ofuniformity in the criteria used to distinguish theprimary or secondary (subsolidus or replacement)nature of pegmatite minerals. Pegmatite textures

Table 2. Reported occurrences of eucryptite

Boulder county , co lo radoUSA

Branchw1l le , Connect l -c u t , U S A

Center S t ra f fo rd , NsHanpsh i re , USA

Et ta n lne , South Dakora ,U S A

k ld ing n1n€,Nes uex tco ,USA

Hiddent te , Nor thCaro l ina , USA

Klngs Utn ,Nor thCaro t ina , USA

P a t a d i s t , , C a 1 1 f o r n 1 a ,USA

ml le P icacho d ls t ,A i i zona, USA

Naklna, htarlo, CanadaTanco mlne, Uanltoba

Canada

Biklta, Ztubabwe

ktunba, Rwandab l le fon te tn , S . f f l t ca

Sa l lsbury d1s t . , Z tubabwe

Central kzakstan, USSR

klb tnsk reg lon , USSRb1a Penlnsula, USSRSlber la , gSSR

Soviet Central Astu

Nanllng boges, S. Chldahndonderry, We6t

fus t ra l1a

usM #128648*

Brush and Dana, 1880

M r o s e , 1 9 5 3

Schwar tz & L@nard , I926J o n e s , 1 9 6 4u r o 6 e , 1 9 5 3

usw /110014

Leves et aL.,1968

Jahns, 1979

Brtt et aL., 1977

hndarino & narr ls, 1968cern' ,1972, 7975

Hul lbut, 1957, 1958, 1952

von horrtng, 1966Schr 'dder & Eoff@n, 1957

Grubb, I973

tuchtnntkov & (uz'Eenko,r97 6

G l n z b u r g , 1 9 5 9Sosedko & Gordlenb, 1957r u g o v s k t l e t q L , 1 9 7 7R o s s o v s k t t , 1 9 7 1

chao, 1964usw #146420

fine-gratuedaggregate ;assoc tu ted phases

Ecr+Alb after Spd

Ecr+AIb

mss lve Ecr ; Ecr+

Qtz a f te ! Spdten ta t l ve ; Ecr -Spd-Qtzf rac tu re f111 ingEcr+Qtz f rac tu ref lU1ngassoc la t lon no t

Ec !+ [b a f te r Spd

Ecr -SpdEcdh lx* ; mass iveEcr ; Ecr+Qtz a f te rsPdEcr -Q!z -Btk* a f te rPet ; Ecr+Qrz a f te rPet ; nass ive EcrEcr+Alb after SpdaBaoc la t lon ro !

Ecr -Spd-Pet -Qtz

aBsoc la t lon no t

Ecr+flb after SpdEc*A.lb after SpdEcr-Spd-Pet-QtzEcHtz a f te r Spd

Btth L1b, lptt*etc.Ecr+Qtz qfter Pe!

* USMI = U.S. Nati.@Z trh8ew, SfritbmLm l@titution, W@hi,ngtm, D.C.,USA** AnX = @aLe'ine, Bik = bikituite, Lik = Z{-beyite, Lpt = Lepidotite

LONDON AND BURT: LITHIUM ALUMINOSILICATE PHASE DIAGRAM 4E5

are extremely complex and varied, and recognitionof the paragenetic sequence consequently is ambig-uous and subjective.

In this and the following paper (London and Burt,1982b) inferred equilibrium assemblages are indicat-ed by mineral names separated by plus signs (+);mineral associations that probably do not or maynot constitute a single paragenesis are representedby mineral names separated by hyphens (-). Theterm primary is applied to a mineral or an assem-blage that generally appears to have crystallizeddirectly from a melt (with or without the participa-tion of an aqueous fluid phase). The granitic borderzones and most of the extremely coarse-grainedpegmatitic units are interpreted as primary in origin.In general, minerals are regarded as secondary orreplacement assemblages only if recognizable evi-dence of replacem ent (e . g ., pseudomorphism, crys-tallographically continuous relicts of a host phase,embayment, etc.) is present.

Natural (rccurrences

Spodumene and petalite usually form extremelylarge primary crystals in the interior portions ofzoned lithium pegmatites. In such occurrences,both minerals typically are embedded in masses ofqtaftz or of quartz plus other minerals (e.g., albite,microcline, amblygonite-montebrasite). In contrast,eucryptite seldom, if ever, forms as a primaryphase. Relatively coarse (up to 2 cm in dimension)subhedral crystals of eucryptite described by Hurl-but (1962) from Bikita, Zimbabwe, may be primary,although they do not approach the giant dimensionsof the other two lithium aluminosilicates.

Primary spodumene and petalite frequently ex-hibit partial to complete pseudomorphism by avariety of fine-grained mineral assemblages. Al-though alteration may be pervasive, it usually issporadic within individual pegmatites and evenwithin single host crystals. In most cases, thisreplacement along cleavage surfaces and crystalboundaries of the host lithium aluminosilicates canbe clearly attributed to pegmatitic processes andnot to subsequent regional metamorphism or localhydrothermal alteration (e.g., London and Burt,1982a). Alteration of primary spodumene or petalitemay be essentially isochemical or may involvecation exchange. This paper considers only iso-chemical replacement processes; the following pa-per (London and Burt, 1982b) treats lithium alumi-nosilicate alterations involving cation exchange.

Fig. l. Photomicrographs showing (top) the alteration ofpetalite to spodumene + quartz from Bikita, Zi4babwe(specimen courtesy of C. S. Hurlbut, Jr., and Carl A. Francis,Harvard University), and (bottom) spodumene + quartzreplacement of petalite from the Tanco pegmatite, Bernic Lake,Manitoba, Canada. Crossed polars.

Isochemical replacement

The most common isochemical replacement oflithium aluminosilicates involves the breakdown ofpetalite to a symplectic or fibrous intergrowth ofspodumene + quartz (Fig. 1). eernf and Ferguson(1972) and Cernf (1975) have reviewed the eightlocalities from which this isochemical replacementof petalite has been reported. In addition to thetextural similarities of the symplectic or fibrousintergrowths, descriptions from several of theselocalities indicate that the long or c axes of spodu-mene grains or fibers are oriented parallel to the{001} cleavage in the host petalite. In samples fromthe Tanco pegmatite in Manitoba, Cernf and Fergu-son (1972) observed (optically) a crystallographicorientation of the secondary spodumene with re-spect to petalite, such that c,o6 is parallel to bpet.

Eucryptite, which is chemically and structurallysimilar to the feldspathoids nepheline and kalsilite,

486

Fig. 2. Photomicrographs showing (top) the intergrowth ofeucryptite + quartz after petalite from Bikita, Zimbabwe(specimen courtesy of C. S. Hurlbut, Jr., and Carl A. Francis,Harvard University), and (bottom) eucryptite + quartzreplacement of spodumene from the Harding pegmatite, TaosCounty, New Mexico. Crossed polars.

surprisingly occurs intergrown with quartz afterpetalite (Fig. 2) from Bikita, Zimbabwe (Hurlbut,1962) and from pegmatites in the Central SovietUnion (Rossovskii, 197 0). Estimates by Rossovskii,and our own observations from thin sections ofBikita samples, indicate that the intergrowths areapproximately 60-80 volVo etcryptite and 40-20volVo qvartz. Cernf (1975) has described a eucryp-tite + quartz intergrowth that appears to havereplaced spodumene at the Tanco pegmatite, andwe have observed eucryptite + qnartz intergrowthsafter spodumene from the Harding pegmatite, TaosCounty, New Mexico (Fie. 3). In addition, well-formed crystals of eucryptite and quartz occur asfracture fillings in pegmatite from Kings Mountain,North Carol ina (e.g., Smithsonian specimenUSNM ll997l) and from Hiddenite, North Carolina(e.9., tentative identification of Smithsonian speci-men USNM 10014). These natural occurrences ledLeavens et al. (1968) and Burt et ql. (1977) to

LONDON AND BURT: LITHIUM ALUMINOSILICATE PHASE DIAGRAM

propose that eucryptite * quartz is a stable assem-blage under some geologic conditions.

Fluid inclusion studies

The available data on the homogenization tem-peratures (i.e., minimum filling temperatures) ofmultiphase fluid inclusions suggest that spodumeneor spodumene-bearing pegmatite crystallizes in therange of 300"-650"C (Sheshulin, 1963; Bazarov andMotorina, 1969; Brookins et al.,1979;Taylor et al.,1979). Sheshulin (1963) identified two generationsof spodumene in pegmatites from the Soviet Union;homogenization temperatures were 500'-650'C forfirst-generation spodumene, and 290'-390"C forsecond-generation spodumene. Estimates of fillingpressure for saline or CO2-rich inclusions indicatepegmatite crystallization pressures in the range of1.5-2.5 kbar (Bazarov and Motorina, 1969; Brook-ins et al., 1979;Taylor et al., 1979). The results ofthese studies imply that spodumene (-rquartz) is

(Bsp,etz)

9 . r 'l/

,fffn

. oO(Brp,Ecr)

T, "C5(n 600

Fig. 3. Summary of the principal experimental work on thelithium aluminosilicate system under geologic conditions(several kilobars of pressure). Dot-dash line and triangle:spodumene synthesis experiments of Shternberg, et al. (1973);dashed line: stability limit of petalite reported by Stewart (1963);dotted line: spodumene-p-spodumene inversion determined byMunoz (1969); dash-dash-dot line: the a-eucryptite-B-eucryptiteinversion reported by Isaacs and Roy (1958); solid line andsquares: reversal of reaction (Bsp,Ecr) reported by Stewart(1963); hexagon: reversal of reaction (Bsp,Ecr) reported byMunoz (1971); circle: spodumene + quartz + petalite,unreversed reaction (D. J. Drysdale, unpublished data, personalcommunication, l9E0). Equilibrium boundaries for the reactionsprobably lie in the directions indicated by arrows.

P, kbar

LONDON AND BURT: LITHIUM ALUMINOSILICATE PHASE DIAGRAM

Table 3. Cell dimensions and calculated molar volumes for eucryptite and p-eucryptite

487

_ 0 oco ' A vo, 3ouLes/bar

Blkita, Zinbabwe (Hurlbut, 1962)Klngs Mountain, North Carollna (Leavens et al.,1968)Tanco mine, Bernlc Lake, Manltoba (Eernf, L972)synthetlc (Tlen and H"*el, 1964)synthetlc (WlnkLer, 1948)synthet ic (Rieder, 1971)synthet lc (Stewart , 1960)

Buerger (1948)Winkl-er (L948)Tlen and Hmel (1964)Pillars and Peacor (1973)

eucryptite1 3 . 4 8 0 ( 5 )1 3 . 4 9 0 ( 7 )L3.472(r)1 3 . 5 11 3 . 5 3l_3. s20 (4)13.476(3)

B-eucrvptlte10 . 555 .27 (bao )5.249 (+ao)10 .497 (3 )

9 . 010 (5 )9 .0L6 (7 )9 . ooo (1)9 . 0 79 . 0 49 . 0 3 s ( 4 )9 . 003 (1 )

TT,22IL.2511 . 18611 . 200 (5 )

4 . 7 4 44 . 7 5 54 . 7 3 34 . 7 9 74 . 7 9 64 . 7 8 64 . 7 3 8

> . 4 t 6

) . q J z

5 . 3595 .364

stable to pressures at least as low as 2 kbar at low tomoderate temperatures.

Calorimetric and experimental data

Construction of the lithium aluminosilicate phasediagram in P-T space is hampered by a lack ofrelevant thermochemical data and experimentalwork in the system LiAlOrSiOz. The stabilityrelations among the lithium aluminosilicates inquartz-saturated and undersaturated systems havereceived considerable attention from ceramists(e.9., Hatch, 1943: Roy et al., 1950, Tien andHummel, 1964) because of the low thermal expan-sion and fluxing properties of lithium alumino-silicate glasses and ceramics. Most of this work wasconducted at atmospheric pressure. The principalinvestigations under geologic conditions (severalkilobars of pressure) are those of Isaacs and Roy(1958), Skinner and Evans (1960), Stewart (1960,1963, 1978), Phinney and Stewart (1961), Edgar(1968), Munoz (1969, 1971), Drysdale (1971, 1975),Grubb (1973), and Shternberg et al. (1973). Thethermochemical properties of spodumene, &spodu-mene, and B-eucryptite have been determined fromsolution and adiabatic calorimetry by Barany andAdami (1966) and Pankratz and Weller (1967); thethermochemical properties of petalite have beendetermined calorimetrically by Bennington er a/.(1980) and by Hemingway (unpublished data, 1980).Although many aspects of lithium aluminosilicatestabilities have been elucidated, these experimentalstudies have not provided a coherent and geologi-cally relevant lithium aluminosilicate phase diagram

in P-T space. There are significant discrepanciesand apparent conflicts in much of the experimentalwork (discussed below). Also, none of the previousstudies has accounted for the stable association ofeucryptite I quartz, even though this assemblagehas been reported from several localities (Table 2).

Considerable confusion suffounds the thermo-chemical properties and structural state of eucryp-tite. There are two principal forms of LiAlSiO4.One is the natural, low-temperature form known aseucryptite, low-eucryptite, or a-eucryptite. Its di-morph is a synthetic, high-temperature form re-ferred to as high-eucryptite or B-eucryptite. Thecrystal structures of eucryptite and of peucryptiteare analogous to those of a-qtartz and ftquartz,respectively, (i.e., they are stuffed derivatives ofthe silica minerals: Buerger, 1954). Eucryptite pos-sesses space group symmetry R3 (Mrose, 1953;Hurlbut, 1962) whereas that of peucryptite is P6a22(Winkler, 1948). The nature of the polymorphictransition, which has not been studied in detail,appears to be different from that of a- and ftqtartz.It may be a first-order reconstructive transforma-tion, inasmuch as B-eucryptite is metastably pre-served at room temperature (Pillars and Peacor,1973).In addition, the molar volumes of eucryptiteand B-eucryptite are quite different (Table 3). Thedata presented in Table 3 serve to clearly identifythe LiAlSiOa polymorphs studied by other workers.

The thermochemical data reported for "eucryp-tite" by Robie et al. (1978) are derived from thecalorimetric work of Barany and Adami (1966) andof Pankratz and Weller (1967) and thus belong to B-

LONDON AND BURT: LITHIUM ALUMINOSILICATE PHASE DIAGRAM

EcrBecsPdBsp

Pe!

Q t z

Table 4. Molar volume and thermochemical data on phases in thesystem LiAlSiOa-SiO2*

Phases vo, Joules/bar so, joules/bote-K Af l?.298, kjoules/nole

noz (1971) gave an equilibrium temperature of470"C at the same pressure, and D. J. Drysdale(unpublished data; personal communication, 1980)observed that petalite is stable down to tempera-tures of 440'C or lower for this same reaction at 2kbar pressure (Fig. 3). While this discrepancy maynot change the fundamental relationship of qaartz-spodumene-petalite equilibria, it would have adrastic effect on the thermochemical properties ofpetalite as derived from Stewart's work. Recentcalorimetric determinations for Sles.15 of petalite byBennington et al. (1980) and by B. S. Hemingway(unpublished data, personal communication, 1980)are in close agreement (Table 4), and both differsubstantially from the value derived from the workof Stewart (1963). The value for Sie3.15 of petalitereported by Bennington et c/. (1980) has been usedhere in reaction slope calculations involving petal-ite. In addition, the work of Stewart (1963) andMunoz (1971) on reaction (Bsp, Ecr) appears to bein conflict with the results of synthesis experiments(reported as an unreversed reaction boundary) byShternberg et al . (1973) involving react ion(Bsp,Qtz) (Fig. 3). This conflict underscores theneed for further qxperimental work in the lithiumaluminosilicate system.

The lithium aluminosilicate phase diagram

The lithium aluminosilicate phase diagram (Fig.4) involves the (n+3)-phase multisystem eucryptite,spodumene, B-spodumene, petalite, and quartz (Ta-ble 5). Figure 4 is drawn for temperatures sufficient-ly low that B-eucryptite is unstable (c/. Isaacs andRoy, 1958). In addition, B-spodumene is assumed tobe stoichiometric, although high-temperature B-spodumene- ftqnarlz solid solutions have been re-ported by numerous experimentalists (e.g., Hatch,1943; Roy, et a1.,1950; Skinner and Evans, 1960;Munoz, 1969,l97l; Stewart, 1978). The assumptionof stoichiometry in p-spodumene permits the use ofcalorimetric data compiled by Robie et al. (1978).This degenerate multisystem contains five possiblestable or metastable binary invariant points in P-Tspace. These and one topology of all possibleunivariant reactions in this system are presented inFigure 4. The topology shown implies that theassemblage spodumene * petalite is stable (i.e.,invariant points [Spd] and [Pet] are metastable). Analternative configuration (not shown) would inter-change stable and metastable invariant points, andby "the stable-metastable correspondence" (Burt,1978), would imply that the assemblage quartz *

4 . 7 9 7 ( 5 \ * *5 . 3 6 3 ( 5 )5 . 8 3 7 < 2 )7 . 8 2 5 ( 4 )

t 2 , 8 4 0 ( 5 ) * * * *

2 . 2 6 8 8 ( 1 )

90 (2) ***103.80 (80)129.30(80)154.40(1 . 20)

270(L)+232.2OO(92>++

4r.46<20)

-2138 (3) ***-2123. 300 (1 . 980)-3053.500 (2 . 790)- 3025. 300 (2 . 790)

-4832(2)+-4886.5(12 .5)++

-910,700(1 .000)

*Data from Robie et a-1. (1978), except as foll*s:**fron Tieh and Hutuel (7964)***calcuiatd as expfained in tle texX****cafculatd ftom x-tag data (JcpN fiTe card L4-go)lcaTcufated from xhe @tk of Stewatt (f963) as expfajnd jn the texttifron Bemirryxon ex al. (1980)

+++unpubljshd data fton E. s, Eqingvag (Ict*nai comunicaxion, lgBO)

eucryptite. Nevertheless, the thermochemical prop-erties of eucryptite can be estimated by utilizing thework oflsaacs and Roy (1958). They have bracket-ed the reaction o-LiAlSiOa : B-LiAlSiO4 and re-port a transformation reaction slope of 23.5 bars/K.Using the molar volume data of Tien and Hummel(1964) and assuming a constant LV.-p :0.56 joules/bar (Table 4), then the Clausius-Clapeyron equa-tion, dPld? : AS/AV, may be solved for AS, andASo-p : 13.19 joules/mole-K. Subtraction of thisquantity from so2e3.15 for B-eucryptite (Robie er a/.,1978) yields an estimate for ^S"zqs.rs of eucryptite(Table 4). From the experimentally determined re-action line for the eucryptite-B-eucryptite transi-tion, Isaacs and Roy (1958) determined that LH._u: 15.7+3 kj/mole; subtraction of this quantity fromAIlF,2e8.15 for B-eucryptite (Robie et a1.,1978) pro-vides an estimate for Alli,2es.15 for eucryptite (Table4).

Values for Siq.15 and AIlies.15 of petalite can bederived from the experimental work of Stewart(1963) on the eucryptite-absent (Bsp,Ecr) reaction:Spd + ZQtz : Pet. Stewart (1963) gives two reac-tion reversals that define a line with slope 23.5 bars/K. Using this information and data from Table 4,Siss.rs and Alli,2e6.15 of petalite have been calculat-ed from the above Clausius-Clapeyron equation andfrom the equation, LG : LH - ZA^t + (P-L)LV,assuming that AV and AS are constant over a widerange of P and T. Other workers have presentedexperimental results involving reaction (Bsp,Ecr)that differ significantly from those of Stewart(1963). Stewart reported a reversal of reaction(Bsp,Ecr) at 550"C and 2 kbar Ps,e, whereas Mu-

LONDON AND BURT: LITHIUM ALUMINOSILICATE PHASE DIAGRAM

(Prt,8pd)

II

, l

il I

!||l

[a.o]

r-r-r 8,'

(8pd,Erp) (Qtr!po)

Fig. 4. P-T phase diagram illustrating the stable topology of univariant reactions in the (n+3)-phase multisystem eucryptite,spodumene, pspodumene, petalite, and quartz in the system LiAlSiO4-SiOr. Solid lines: stable reactions; dashed and dotted lines:metastable reactions; solid circles: stable invariant points; open circles: metastable invariant points. The triangular scale marker inthis and the following diagram indicates the relative slopes of reaction lines. Positions of phases on the binary LiAlSiOa-SiOz areindicated in the upper left portion of this diagram. See Table I for abbreviations of phase names.

LlAtSlo4 8lO2

/' 8)

eucryptite + Fspodumene (not known in nature) isstable. The configuration shown in Figure 4 istherefore assumed to be the stable one.

The univariant reactions around point [Bsp] in-volve only the naturally occurring phases, and thusthe topology of reactions about this point is particu-larly relevant to geologic environments (Fig. 4). Theconfiguration of phases on the binary join LiAlSiOa-SiOz Gig. 4) implies four possible univariant reac-tions at invariant point [Bsp]. Reaction slope calcu-lations based on the data of Table 4 indicate that

tl

489

(Prt,Qte,Eo:l

/a8 (Eor,8pC)

- (Brp,Qtr)

these four reactions have positive slopes in P-Tspace. The uncertainties in the thermochemicalproperties of eucryptite and the discrepancies re-ported for the 2 kbar reversal of reaction (Bsp,Ecr)make the precise calculation of the location of point

[Bsp] a meaningless effort. Our feeling, based onthe natural occurrences discussed above, is that thepoint [Bsp] occurs at considerably lower P and Tthan does the aluminosilicate triple point(Holdaway, 1971). Natural occurrences and thelimited experimental work then suggest that the

tK

T

490 LONDON AND BURT: LITHIUM ALUMINOSILICATE PHASE DIAGRAM

Table 5. Univariant reactions in the system LiAlSiO4-SiO2(Fie. a)

P h a s e s a b s e n E R e a c t i o n s ' " 2 9 8 dP/dTbars K

substitutions that make the system non-binary, thelithium aluminosilicate phase diagram here pro-posed (Figures 4 and 5) explains some commonlyobserved field associations and experimental re-sults. The breakdown of petalite to B-spodumenereported from hydrothermal runs with bulk compo-sitions between LiAlSiO4 and LiAlSiaOlq (e.9., Royet al., 1950; Skinner and Evans, 1960; Stewart,1963, 1978; Munoz, 1969) may occur via reactions(Pet, Qtz, Ecr) and (Qtz,Spd) (Table 5 and Figure 4)at moderate Tand low to moderate P. However, thetopology of reactions depicted in Figure 4 suggeststhat petalite is stable to high I for bulk composi-tions from LiAlSi4Orq to SiO2 (Rossovskii, 1967;Rossovskii and Matrosov, 1974; c/. also Stewart,1963, and Bennington et al., 1980, p. 13). Theabsence of B-spodumene phases in nature may beexplained by the fact that the bulk compositionalrange of lithium-rich zones in pegmatites falls be-tween LiAlSi4Oro and SiOz (Stewart, 1978), andreaction (Ecr,Spd) (Table 5 and Figure 5) probablylies at sufficiently high Z that B-spodumene, ifproduced, would react completely with excess sili-ca to form petalite above the lithium pegmatitesolidus (see Jahns and Burnham, 1958).

The proposed stability relations among the natu-rally occurring lithium aluminosilicates are illustrat-ed at invariant point [Bsp] in Figure 5. With de-creasing Z, petalite may break down to spodumene+ qaartz at moderate to high P (via reaction(Bsp,Ecr)) or to eucryptite * quartz at moderate to

(Prt,Qtr,Ecrl

SPODUMEI,IE

(8pd,trp)

Fig. 5. P-Zphase diagram illustrating the stability fields of thelithium aluminosilicates in (natural) quartz-saturated environ-ments.

""29a

( P e t , B s p )

(Spd, Bsp)

( B s p , n c r )

( B s p , Q t z )

( P e t , S p d )

( P e r , Q t z , E c ! )

( Q r z , S p d )

- 2 . 1 6 + 1 . 8

+ 1 7 . 4 2 + t 4 . 4

+ 1 9 . 9 8 + 8 , 1

- 2 4 . 3 0 + 4 . 9

+ 2 2 . 9 4 + 3 0 . 2

+ 2 5 , 1 0 + 1 2 . 6

+ 5 1 - 0 0 + 4 9 . o

- 5 . 1 2 - I 0 . 7

four univariant lines define a spodumene field atrelatively high P and moderate T, a petalite field athigh T and low to moderate P, a eucryptitg field atlow to moderate P and T, and a petalite * spodu-mene field over a narrow range of moderate to highP and ?. Inasmuch as the bulk compositions oflithium-rich zones in pegmatites normally fall be-tween petalite and quartz on the binary LiAlSiO4-SiO2 (Stewart, 1978), the quartz-saturated phasediagram (Fig. 5) is appropriate for showing thestability fields of the lithium aluminosilicates innature.

Reactions among the lithium aluminosilicatesmay be useful indicators of pressure and tempera-ture. Shternberg et al. (1973) suggested that spodu-mene may be useful as a depth indicator, aldFigures 4 and 5 indicate that reaction (Pet,Bsp) mayconstitute a geobarometer in that it establishes aminimum cyrstallization pressure for spodumene-bearing pegmatites. The ferric iron analogue ofspodumene, LiFe3+Si2O6, has been synthesized atmoderate to very low pressure (Barrer and White,1951; Drysdale, 1965), and substitution of iron andother elements in spodumene might therefore ex-tend its stability field to quite low pressures (Apple-man and Stewart, 1968). Natural spodumenes, how-ever, usually contain less than two or three molepercent of impurities (Heinrich, 1975;Deer et al.,1978), and these impurities exist largely in the formof micro-inclusions of other minerals (Graham,1975). Therefore, the activity of LiAlSi2O6 in natu-ral spodumene probably is near unity.

Implications for the crystallization of lithiumpegmatites

Allowing for considerable uncertainty in the de-rived thermochemical data (Table 4) and for ionic

Ecr + Qtz = Spd -1.23

E c r + 3 Q t z = ? e r + 1 . 2 3

S p d + 2 Q t z = P e t + 2 . 4 6

z E c r + P e t = 3 s p d - 4 , 9 2

E c r + Q t z = B s p + 0 . 7 6

S p d = B s p + 1 . 9 9

2Ecr + Pet = 3Bsp +1.05

8sp + 2Qt2 = ?et +0.48

LONDON AND BURT: LITHIUM ALLIMINOSILICATE PHASE DIAGRAM 491

low P (via reaction (Spd,Bsp)). The lithium alumi-nosilicate stability relations presented in Figures 4and 5 explain the isochemical replacement of peta-lite reported from numerous localities and accountfor the observed compatibility of eucryptite +qvartz under geologic conditions of relatively low Pand Z. Based on existing thermochemical data, theproposed reaction (Pet,Bsp) should be highly pres-sure-dependent and would represent a minimumpressure boundary for crystallization of spodu-mene-bearing pegmatites and granites. Ionic substi-tutions in natural spodumene may extend its stabil-ity to somewhat lower pressures at the expense ofeucryptite + quafiz. These factors probably arelargely responsible for the general absence of eu-cryptite + quartz pseudomorphs after spodumeneand for the rarity of eucryptite + quartz assem-blages. The following paper (London and Burt,1982b) considers other chemical processes that ap-pear to operate in lithium pegmatite evolution andthat would also inhibit or preclude the formation ofeucryptite + quartz.

AcknowledgmentsWe thank Bruce S. Hemingway, U. S. Geological Survey, for

the calorimetric determination of Sies of petalite. David B.Stewart, U. S. Geological Survey, and Jack D. Drysdale, Uni-versity of Queensland, provided unpublished experimental datapertinent to this investigation. Eric J. Essene, John M. Ferry,John R. Holloway, John W. Larimer, Alexandra Navrotsky, andDavid R. Veblen provided constructive reviews. Carl A. Fran-cis, Harvard University, and Daniel E. Appleman and John S.White, Jr., Smithsonian Institution, provided samples of lithiumaluminosilicates from numerous localities. We thank KathrynGundersen for typing the final manuscript and Sue Selkirk fordrafting the diagrams.

This work was supported in part by a Sigma Xi Grant-in-Aid ofResearch, by a Graduate Research Grant from the GeologicalSociety of America, and by National Science Foundation Grant#EAR-7814785 to Arizona State University.

ReferencesAppleman, D. E. and Stewart, D. B. (1968) Crystal chemistry of

spodumene-type pyroxenes. (abstr.) Geological Society ofAmerican Special Paper, 101,5-6.

Barany, R. and Adami, L. H. (1966) Heats of formation oflithium sulfate and five potassium and lithium aluminosili-cates. U. S. Bureau of Mines, Report of Investigations, 6873.

Barrer, R. M. and White, E. A. D. (1951) The hydrothermalchemistry of silicates. Part I. Synthetic lithium aluminosili-cates. Journal of the Chemical Society, London, April, 1951,Part II, (283), 1267-1278.

Bazarov, L. S. and Motorina, I. V. (1969) Physicochemicalconditions during formation of rare-metal pegmatite of thesodium-lithium type. Doklady Academii Nauk SSR, 188, 194-197 (transl. Doklady of the Academy of Sciences, USSR,Earth Science Sections, 188, 124-126, 1969).

Bennington, K. O., Stuve, J. M., and Ferrante, M' J. (1980)

Thermodynamic properties of petalite (LirAl2Si8O2o). U. S.

Bureau of Mines, Report of Investigations, 8451.Beus, A. A. (1960) Geochemistry of Beryllium and Genetic

Types of Beryllium Deposits. Akademii Nauk SSSR, Moscow(transl. Freeman, San Francisco, 1967).

Brookins. D. G., Chakoumakos, B. C., Cook, C. W., Ewing' R.

C., Landis, G. P., and Register, M. E. (1979) The Hardingpegmatite: summaxy of recent research. New Mexico Geologi-cal Society Guidebook, 30th Field Conference, Sante Fe

Country, 127-133.Br<itzen, O. (1959) Outline of mineralization in zoned granitic

pegmatites. A qualitative and comparative study' GeologiskaFdreningen i Stockholm Forhandlingar, 81, l-98.

Brush, G. J. and Dana, E. S. (1880) On the mineral locality at

Branchville, Connecticut. Fourth paper. Spodumene and the

results of its alteration. American Joumal of Science' 118,

257-285.Buerger, M. J. (1948) Crystals based on the silica structures

(abstr.). American Mineralogist, 33, 7 5l-:7 52.Buerger, M. J. (1954) The stutred derivatives of the silica

structures. American Mineralogist, 39, 600-614.Burt, D. M. (1978) Multisystems analysis of beryllium mineral

stabilities: the system BeO-A12O3-SiO-H2O. American Min-

eralogist, 63, 664-67 6.Burt. D. M., London, D., and Smith, M. R. (1977) Eucryptite

from Arizona, and the lithium aluminosilicate phase diagram.(abstr.) Geological Society of America Abstracts with Pro-grams,9 (7 ) ,917.

Cameron, E. N., Jahns, R. H., McNair, A. H.' and Page, L. R.(1949) Internal structure of granitic pegmatites' EconomicGeology, Monograph 2.

Cameron, E. N. and others (1954) Pegmatite investigations,lg42-45, New England. U. S. Geological Survey ProfessionalPaper 225.

Cernf, P. (1972) Eucryptite. Canadian Mineralogist, l1' 708-

713.Cern9, P. (1975) Granitic pegmatites and their minerals: selected

examples of recent progress. Fortschritte der Mineralogie, 52,

225-250.Cern9, P. and Ferguson, R. B. (1972) Petalite and spodumene

relations. Canadian Mineralogist, I I (3), 660-678.Chakoumakos, B. C. (1978) Replacement features in the Harding

pegmatite, Taos County, New Mexico. (abstr') New MexicoAcademy of Science Bulletin, 18, 22.

Chao, C.-L. (1964) Liberite (LizBeSiOr), a new lithium-berylli-um silicate mineral from the Nanling Ranges, South China. TiChih Hsuch Pao, 44, 334-342 (not seen; extracted fromChemical Abstracts, 61, #1585ld, 1964).

Cooper, D. G. (1964) The geology of the Bikita pegmatite. In S.H. Haughton, Ed., The Geology of Some Ore Deposits inSouthern Africa, Geological Society of South Africa' 2,441-461.

Deer, W. A., Howie, R. A., and Zussman' J. (1978) Rock-forming Minerals. Vol. 2A. Single-chain Silicates, 2nd Ed.

Wiley, New York.Drysdale, D. J. (1971) A synthesis of bikitaite. American Miner-

alogist, 56, 17 18-1723.

Drysdale, D. J. (1975) Hydrothermal synthesis ofvarious spodu-menes. American Mineralogist, 60, 105-110.

Edgar, A. D. (1968) The a-BLiAlSi2O6 (spodumene) transitionfrom 5000 to 45000 lb/in2 Pr"o. In International Mineralogical

492

Association, Papers and Proceedings, 5th General Meeting,Cambridge, England, 1966, Mineralogical Society, London,222-231.

Fersman, A. E. (1931) Pegmatites. Vols. l-3 [in Russian].Akademii Nauk SSSR, Leningrad (transl. in French, Louvain,1951) .

French, B. M., Jazek, P. A., and Appleman, D. E. (197g)Virgilite, a new lithium aluminum silicate mineral from theMacusani glass, Peru. American Mineralogist, 63,461465.

Ginzburg, A. I. (1955) Mineralogical-geochemical characteris-tics of lithium pegmatites [in Russian]. Trudy Mineralogiches-kogo Muzeya Akademii Nauk SSSR, 7,lZ-55.

Ginzburg, A. I. (1959) Spodumene and its alteration processes[in Russian]. Trudy Mineralogicheskogo Muzeya AkademiiNauk SSSR, 9, 19-52 (not seen; extracted from Chemica.lAbstracts, 53, #214t+6e, 1959).

Gordienko, V. A. (1970) Mineralogy, Geochemistry, and Gene-sis of Spodumene Pegmatites [in Russian]. publishing HouseNedra, Leningrad, USSR.

Graham, J. (1975) Some notes on a-spodumene, LiAlSizOo.American Mineralogist, 60, 919-923.

Grubb, P. L. C. (1973) Paragenesis of spodumene and otherlithium minerals in some Rhodesian pegmatites. In Sympo-sium on Granites, Gneisses and Related Rocks, GeologicalSociety of South Africa, 3, 20l-216.

Hatch, R. A. (1943) Phase equilibrium in the system Li2O-Al2O3-SiO2. American Mineralogist, 28, 47 l-496.

Heinrich, E. W. (1975) Economic geology and mineralogy ofpetalite and spodumene pegmatites. Indian Journal of EarthSciences. 2. 18-29.

Holdaway, M. J. (1971) Stability of andalusire and the alumino-silicate phase diagram. American Journal of Science. Z7l. 97_l 3 l .

Hurlbut, C. S., Jr. (1957) Bikitaite, LiAlSi2O6.H2O, a newmineral from Southern Rhodesia. American Mineralogist, 42,792-:797.

Hurbut, C. S., Jr. (1958) Additional data on bikitaite. AmericanMineralogist, 43, 768J70.

Hurbut, C. S., Jr. (1962) Data on eucryptite from Bikira,Southern Rhodesia. American Mineralogist, 4j, 557.

Isaacs, T., and Roy, R. (1958) The a-p inversions of eucryptiteand spodumene. Geochimica et Cosmochimica Acta, 15. Zl3.-217.

Jahns, R. H. (1955) The study of pegmatites. Economic Geology,50th Anniversary Volume, Part 2, 1025-1130.

Jahns, R. H. (1979) Gem-bearing pegmatires in San DiegoCounty. In P. L. Abbott, and V. A. Todd, Eds., MesozoicCrystalline Rocks: Peninsular Ranges Batholith and pegma-tites, Point Sal Ophiolite. Manuscripts and road logs preparedfor Geological Society of America Annual Meeting, San Die-go, CA, 1979; published by Department of Geological Sci-ences, San Diego State University, San Diego, CA, l_3g.

Jahns, R. H., and Burnham, C. W. (1958) Experimental studiesof pegmatite genesis: melting and crystallization of granite andpegmatite. (abstr.). Geological Society of America Bulletin,69, 1592-1593.

Jahns, R. H. and Ewing, R. C. (1976) The Harding pegmatite,Taos County, New Mexico. New Mexico Geological SocietyGuidebook, 27th Field Conference, Vermejo park,263-276.

Jones, R. W., Jr. (1964) Collecting fluorescent minerals. Rocksand Minerals. 39. 339.

Julien, A. A. (1879) On spodumene and its alterations from the

LONDON AND BURT: LITHIUM ALUMINOSILICATE PHASE DIAGRAM

granite-veins of Hampshire County, Massachusetts. Annals ofthe New York Academy of Science, 1, 318-359.

Leavens, P. B., Hurlbut, C. S., Jr., and Nelen, J. A. (1968)Eucryptite and bikitaite from Kings Mountain, North Caroli-na. American Mineralogist, 53, 1202-1207.

London, D. (lgEl) Preliminary experimental results in the sys-tem LiAlSiOa-SiO2-H2O. Yearbook of the Carnegie Institu-tion of Washingron, 80, 341-345.

London, D. and Burt, D. M. (1980) Local lowering of silicaactivity during albitization of spodumene-its relations to theformation of micas and eucryptite. (abstr.) Transactions of theAmerican Geophysical Union, 61, 404.

London, D. and Burt, D. M. (1982a) Alteration of spodumene,montebrasite, and lithiophilite in pegmatites of the WhitePicacho district, Arizona. American Mineralogist, 67 ,97-113.

London, D. and Burt, D. M. (1982b) Chemical models for lithiumaluminosilicate stabilities in pegmatites and granites. Ameri-can Mineralogist, 67, 494-509.

Lugovskii, G. P., Rub, A. K., Ryabtsev, V. V. (1977) Eucryptitefrom Siberian pegmatites. Doklady Akademii Nauk SSSR,237, 195-198 (transl. Doklady of the Academy of Sciences,USSR, Earth Science Sections, 237, 206-208, 1977).

Mandarino, J. A., and Harris, D. C. (1965) New Canadianmineral occurrences: I. Eucryptite, pollucite, rozenite, epso-mite, dawsonite, fairchildite, and butschliite. Canadian Miner-alogist, 8, 377-3E0.

Mrose, M. E. (1953) The a-eucryptite problem. (abstr.) Ameri-can Mineralogist, 38, 353.

Munoz, J. L. (1%9) Stability relations of LiAlSi2Ou at highpressure. Mineralogical Society of America Special Paper 2,203-209.

Munoz, L L. (1971) Hydrothermal stability relations of syntheticlepidolite. American Mineralogist, 56, 2069-2087.

Ovchinnikov, L. N. and Kuz'menko, M. V., Eds. (1976) Fieldsof Rare Metal Granitic Pegmatites [in Russian]. Nauka Press,Moscow.

Page, L. R. and others (1953) Pegmatite investigations,194245,Black Hills, South Dakota. U. S. Geological Survey Profes-sional Paper, 247.

Pankratz, L. B. and Weller, W. W. (1967) Thermodynamicproperties of three lithium aluminosilicates. U. S. Bureau ofMines, Report of Investigations, 7001.

Phinney, W. A., and Stewart, D. B. (1961) Some physicalproperties of bikitaite and its dehydration and decompositionproducts. U.S. Geological Survey Professional Paper 424,D-353-D357.

Pillars, W. W. and Peacor, D. R. (1973) The crystal structure ofbeta eucryptite as a function of temperature. American Miner-alogist, 58, 6E1-690.

Rieder, M. (1971) Stability and physical properties of syntheticlithium-iron micas. American Mineralogist, 56, 256-280.

Robie, R. A., Hemingway, B. S., and Fisher, J. R. (1978)Thermodynamic properties of minerals and related substancesat 298.15 K and I bar (105 Pascals) pressure and at highertemperatures. U. S. Geological Survey Bulletin 1452.

Rossovskii, L. N. (1967) Formation pecularities of petalite andspodumene pegmatites [in Russian]. Geologiya Rudnykh Mes-torozhdeniyakh,9, 19-29 (not seen; extracted from ChemicalAbstracts, 68, #E0336f, 1968).

Rossovskii, L. N. (1970) Eucryptite in petalite-microcline peg-matites of Soviet Central Asia. Doklady Akademii NaukSSSR, 197, 1398-1401 (transl. Doklady of the Academy of

LONDON AND BURT: LITHIUM ALUMINOSILICATE PHASE DIAGRAM 493

Sciences. US SR, Earth Science Section s, 197, 143 -1 46, l97 l).Rossovskii, L. N. and Matrosov, I. I. (1974) Pseudomorphs of

quartz and spodumene after petalite and their importance tothe pegmatite-forming process. Doklady Akademii NaukSSSR, 216 (5), l135-1137 (transl. Doklady ofthe Academy ofSciences, USSR, Earth Science Section s, 216, 164-1ffi , 197 4).

Roy, R., Roy, D. M., and Osborn, E. F. (1950) Compositionaland stability relationships among the lithium aluminosilicates:eucryptite, spodumene, and petalite. Journal ofthe AmericanCeramic Society, 33, 152-159.

Schrcider, A. and Hoffman, W. (1957) The optical constants oflow eucryptite from Mollenfontein (South Africa). NeuesJahrbuch ftir Mineralogie, Monatshefte , 73:75.

Schwartz, G. M., and Leonard, R. J. (1926) Altered spodumenefrom the Etta Mine. South Dakota. American Journal ofScience. 2ll. 257-264.

Sheshulin, G. L (1963) Composition of gas-liquid inclusions inthe minerals of spodumene pegmatites. In A. I. Ginzburg, Ed.,New Data on Rare Element Mineralogy. Nauka Press, Mos-cow (transl. Consultants Bureau, New York, 1963).

Shternberg, A. A., Ivanova, T. N., and Kuznetsov, V. A. (1972)Spodumene-a mineral depth indicator. Doklady AkademiiNauk SSSR, 202, 175-178 (transl. Doklady of the Academy ofSciences, USSR, Earth Science Sections, 2O2, lll-114, 1973).

Singleton, R. H. (1979) Lithium. U. S. Bureau of Mines, MineralCommodities Profile.

Skinner, B. J. and Evans, H. T. (1960) B-spodumene solidsolutions and the join Li2O-Al2O3-SiO2. American Journal ofScience, 258A, 312-i24.

Sosedko, A. F. and Gordienko (1957) Eucryptite from thepegmatite of the northern part of the Kola Peninsula. DokladyAkademii Nauk SSSR, 116, 135-136 (transl, Doklady of theAcademy of Sciences, Earth Science Sections, l16, E73-E75,r9s7).

Stewart, D. B. (1960) The system LiAlSiO4-NaAlSi:OrHuO at2fi)0 bars. International Geological Congress, XXI Session,Paxt 17, 15-30.

Stewart, D. B. (1963) Petrogenesis and mineral assemblages of

lithium-rich pegmatites. (abstr.) Geological Society of Ameri-

ca Special Paper,76, 159.Stewart, D. B. (1978) Petrogenesis of lithium-rich pegmatites'

American Mineralogist, 63, 970-980.Taylor, B. E., Foord, E. 8., and Friedrichsen, H. (1979) Stable

isotope and fluid inclusion studies of gem-bearing granitic

pegmatite-aplite dikes, San Diego County, California. Contri-

butions to Mineralogy and Petrology,68' 187-205.

Tien, T. Y. and Hummel, F. A' (1964) Studies in lithium oxide

systems: XIII. Li2O'Al2O3'2SiO2-Li2O'AlzO:'2GeOz. Journal

of the American Ceramic Society, 47, 582-584'Uebel, P.-J. (1977) Internal structure ofpegmatites, its origin and

nonmenclature. Neues Jahrbuch fiir Mineralogie Abhandlun-g e n , 1 3 1 , 8 3 - 1 1 3 .

Vine, J. D., Ed. (1976) Lithium resources and requirements by

the year 2000. U. S. Geological Survey Professional Paper

1005.Vlasov, K. A. (1961) Principles of classifying granite pegmatites

and their textural-paragenetic types. Izvestiya Akademii Nauk

SSSR, Seriya Geologicheskaya, 1,5-20 (transl. Izvestiya of

the Academy of Sciences, USSR, Geologic Series' l, 5-20,

l96l).Vlasov, K. A. (1964-66) Geochemistry and Mineralogy of Rare

Elements and Genetic Types of Their Deposits (3 vols')'

Nauka Press, Moscow (transl. Israel Program for Scientific

Translations, Jerusalem, l%6-l%E).von Knorring, O. (1966) Pegmatite mineral studies. Research

Institute of African Geology and Department of Earth Sci-

ences, l0th Annual Report, University of Leeds, 32-33 (not

seen; extracted from Mineralogical Abstracts ' LE' 126).

Winkler, H. G. F. (1948) Synthese und Kristallstruktur des

Eukryptits, LiAlSiO4. Acta Crystallographia, Section B' I'

27-34.

Manuscript received, APril 6, 1981;acceptedfor publication, Ianuary |E' 19E2.