American Mineralogist, Volume 70, pages 395408, 1985

Occurrenceand alteration of phosphateminerals at the Stewart , Pala District,lSan Diego County, California

J,c,MssE. Stncr,svr exo GonooN E. BnowN, Jn. Department of Geology St anford U niuer sity, St anford, Califurnia 94 3 0 5

Abstract Parageiibticrelationships among Li(Mn, Fe)2+ PO4 and its alteration products are describedfor one of the complex granitic pegmatitesin the Pala pegmatitedistrict, San Diego County, California. At the Stewart pegmatite, lithiophilite occurs in the upper- intermediate,microcline- zone while spodumeneand are found in the quartz core. The sequenceof primary formation within the pegmatite reflectsan increasein the activities of both speciesand volatile components(phosphorus and fluorine).Extensive alteration oflithiophilite involved oxidation, hydration, and cation leach- ing, but suprisingly little metasomatism.Secondary presentinclude sicklerite,hu- reaulite, , stewartite,phosphosiderite, several incompletely identified phases,and various manganeseoxides. Secondaryphosphates formed during the initial stagesof alter- ation pseudomorphouslyreplaced lithiophilite as a result of its limited hydrothermalinterac- tion with late-stagepegmatitic fluids. Later membersof the alteration sequencerepresent supergene weathering products. The Stewart pegmatite crystallized from a highly- diflerentiatedgranitic magmaat shallowcrustal depths(about 3-5 km). The lack of extensive metasomaticreplacement, which is so evident among the phosphatemineral assemblagesof other granitic ,is thought to primarily be a result of the relatively rapid cooling and volatile fluid loss at the shallow depths of formation of the Stewart pegmatite.Under theseconditions, there was little opportunity for metasomaticreaction to take place between lithiophilite and residualpegmatitic fluids.

Introduction generalsusceptibility to alteration, the potential usefulness of the phosphateminerals as indicators of changingcon- For almost a century the granitic pegmatitesof southern ditions during the post-crystallizationhistory of a pegma- California have been a well-known source of gemstones tite has long been recognized(Mason, 1941; Fisher, 1958; and other minerals(Kunz, 1905;Merrill, 1914;Donnelly, eerni, 1970;van Wambeke,l97L; Moore, 1982).More- 1936;Jahns and Wright, 1951; Sinkankas,1957; Foord, over, the widespreadoccurrence and diversity of the peg- 1976').ln addition to gem material, some of thesepegma- matitic phosphatesalso provide a rationale for using these tites contain interesting accessoryminerals, The present mineralsto better understandthe detailsof pegmatitecrys- study of the primary phosphatemineral lithiophilite and its tallization. However, the complex paragenesesof these secondary alteration products from the famous Stewart phaseshave only recentlybegun to be elucidated(Moore, pegmatitenear Pala was undertakento better understand 197Q,1971, 1972a, 1981, 1982; Moore and Molin-Case, the occurrenceand paragenesisof phosphateminerals in 1974i Fontan et al., 1976i Fransolet, 1976; Shigley and granitic pegmatites. Brown, 1980;Miicke, 1981;Segeler et al., 1981;London Phosphateminerals such as apatite, amblygonite,lithi- and Burt, l982al. While the broad outlines of ophilite and others are common minor constituentsof nu- crystal chemistryand the crystal structuresof a number of merousgranitic pegmatites(Moore, 1973,1982\. These peg- phosphateminerals have been establishedas a result of matitic phosphatesare generally found to be altered by theserecent studies, details of the thermodynamicrelation- oxidation, hydration, cation leaching,and metasomaticre- phosphateminerals have yet to be addressed placementreactions. Such alteration is responsiblefor most ships among manner. of the nearly 150 secondaryphosphate species now recog- in a comprehensive (1912) first describedthe phosphate minerals nized from pegmatites.The extent of this phosphatealter- Schaller from Pala, but unfortunatelyhis completestudy of pegma- ation appearsto be somewhatgreater than is the casefor tite mineralogy of this area was never published.Beyond many of the associatedpegmatitic silicates. Because of this some additional information reported by Murdoch (1943) and Jahns and Wright (1951),there has been little subse- I Presentaddress: Research Department, Gemological Institute quent investigation of the Pala .This is in of America,1660 Stewart Street, Santa Monica, California 904O4. nrarked contrast with the extensiverecent studies of phos- 000H04x/85/0304-{395$02.00 395 3% SHIGLEY AND BROWN: PHOSPHATE MINERALS, STEWAR.TPEGMATI.IE

Table 1. Primary and secondaryphosphate minerals from the consistingof hundredsof individual plutons ranging up to Stewartpegmatite. severalkilometers or larger in diameter.These plutons in- trude into prebatholithicmetasedimentary and metavolca- Mineral* Chmlcal fomula** nic rocks. They are composedof older gabbro grading to tonalite, granodiorite, quartz monzonite, younger LlEbaophlIlte Li (Mn, Fe) 2+Po4 and granite; with the first two rock types being the most vol- Slcklerite r,rr_*tunf]-'rel+)roo uminousconstituents. r"y (po3oH) {u", l+turo;oGo4) 2 2 The pegmatitedikes at Pala are generallyconfined to the Purpurlte (m, re) 3+?oo gabbroic plutons of the batholith, and are believedto rep- resent products of the final stagesof its magmatic differ- 3+po4 Heterosite (Fe,Mn) entiation and crystallization (Jahns, 1947; Jahns and Phosphosiderlte r.3+(nzo)zpor, Wright, 1951).The observeduniformity in the attitude and S lewar t ite u"2+1nro)o 2H20 tabular form of these pegmatiteshas {r.]+{on)2e12o) 2(po I ). been taken as evi- Anblygonlre+ 5 (Li,Na)A1(Po4) (F,oH) dencefor their formation by the crystallizationof late-stage residual magmasas vein fillings along sheetlike fractures Trlpllte+q (Mn,re)2+r(Po4) in the older and more competentgabbro. The pegmatites q Trlphylite+ Li(Ie,Mn)- PO4 range from centimetersto metersin thicknessand up to a kilometer in exposedlength. They are composedprimarily

List does not include several incompleiely idef,tlfled of feldspars,qvartz, and micas,and most can be considered phosphate phases nored during lhls study. Taken fron Moore (1982) and I'leischer (l983). "simple" pegmatiteson the basis of their homogeneous t Primry phosphate ninerals. Reported to occur by Jahns and Wrtghr (1951), but not internal structure.However, several ofthe larger dikessuch exelned during this study, as the Stewarthave a more complexinternal structureand diverse mineralogy. Lithium minerals and gem materials, as well as the more unusualaccessory minerals such as the phosphates,are virtually restrictedto these"complex" peg- phate minerals from other granitic pegmatites(see Miicke, matites. 1981;Segeler et al., 1981;London and Burt, 1982a).Thus, the purposesof the presentinvestigation were (1) to care- The Stewart pegmatite fully reexamine phosphate the mineralogy of one of the The Stewartpegmatite occurs as an elongateoutcrop of Pala pegmatites to extendSchaller's earlier work in light of light-coloredrock on the south-facingslope of more recent ideas; (2) to use this information to evaluate Queen mountain two kilometers north of Pala. It extends the geologic pegmatite; history of the and (3) to compare over a distanceof a kilometer with a northerly strike and a these minerals with the phosphatesreported from other moderatewesterly dip, and attains a maximum thicknessof granitic pegmatites.The Stewartpegmatite was chosenfor 25 meters.The StewartLithia mine, near the southernend study becauseit contains abundant lithiophilite crystals, of the diko, is a major gem producer(Jahns et al., 1974\. and its interior portions are partly accessiblethrough the While an accurate modal composition is unavailable, existingunderground workings of the StewartLithia mine. Jahns(1953) published the following bulk compositionfor Table 1 lists the phosphateminerals reported from the peg- matite. Data gatheredon the phosphatemineralogy and alteration sequenceare used to interpret aspectsof the overall geologic history of the Stewart pegmatite in the context of the Jahns-Burnhammodel of granitic pegmatite formation (Jahnsand Burnham, 1969;Jahns, 1982).

Geologic setting Pala is located in the northwest corner of San Diego County (Fig. 1). Several hundred granitic pegmatites, chiefly dikes, are exposedon the hillsidesimmediately sur- rounding the town. The geology of the area has been de- scribed by Larsen (1948, 1951)and Jahns (1954a 1954b, r979\. Much of the Pala pegmatitedistrict is underlain by ig- neous intrusive rocks of the Cretaceous-agePeninsular Ranges (or Southern California) batholith. U-Pb and K-Ar radiometric age dates for batholithic rocks fall within the 130-90m.y. time span (Banks and Silver, 1969; 20 XIrl Krummenacher et al,, 1975; Dalrymple, 1976). This Fig. 1. Map of San Diego County showing the location of the orogenic-typebatholith is a large,composite intrusive body StewartLithia mine near Pala. SHIGLEY AND BROWN: PHOSPHATE MINERALS, S.TEWARTPEGMATI,IE 397 the southern portion of the pegmatite (in wt.% oxides): SiOz 74.9,AlzO3 14.9, CaO 0.1,NarO 3.6,K2O 5.2,LizO 0.7,F 0.4, H2O 0.4, total 100.2.Major constituentsinclude microcline-perthite, , quartz, muscovite, , ,amblygonite, and tourmaline,while additional acc€ssoryminerals are rich in Li, P, Be, B, and Mn (for details,see Jahns and Wright, 1951).Within the pegmatite, mineral assemblagesare arranged in subparallel,layered zones both above and below a discontinuous quartz- spodumenecore (Fig. 2). Zones above the core are gener- ally coarser-grainedand rich in microcline,whereas those below contain both coarse-and fine-grainedmaterial and have more albite. Gem-bearing pockets and massive lepidolite orebodies are located directly underneath the quartz core. This zonal arrangementcorresponds to the generalizedinternal zonal sequenceof Cameron et al. (1949)and Norton (1983). Fig. 3. Photograph of a lithiophilite crystal embedded in peg- Occurrence of phosphate minerals matite host rock in the old underground workings of the Stewart Lithia mine. The outline of the crystal has been highlighted with a Phosphateminerals can be recognizedas dark-stained dashed white line. massesor noduleson the tunnel walls in the StewartLithia mine (Fig. 3). Thesemasses are as much as 40 cm across, are somewhatequidimensional in shape,and occur in the upper quartz-microcline intermediate zone (Fig. 2). Al- during early mining operations.Thus, neither it nor its though quite altered,they representlarge, subhedral to eu- possible alteration products were characterizedin this hedral crystals of lithiophilite whose exterior surfacesare study (see,however, information in Murdoch and Webb, covered by either a few simple crystal faces (Fig. 4) or 1966,p. 78 and 234). irregular growth surfaces.We infer lithiophilite to be a Schaller(1912) and Jahnsand Wright (1951)reported primary pegmatite constituent on the basis of its crystal small amounts of primary and from the morphology, coarsegrain size,textural relationshipswith mine, but samplesof neither phase could be collectedto adjoining silicates,restricted zonal occurrence,and finally a establishtheir parageneses. lack of evidencethat it has replaced any earlier-formed Mineral characterization mineral. Some 130 crystals or crystal fragmentsfrom the from the pegmatiteare extensively mine were examinedduring the course of this study and Lithiophilite crystals While a few contain small areasof remnant lithio- are now housed in the Stanford University mineral col- altered. philite, entirely composedof secondaryphosphate lection. most are manganeseoxides present as impure massive Amblygonite occurs as coarse-grainedcrystalline aggre- mineralsand grains, veinlets, rarely as tiny gateswith lepidolite in portions of the quartz-spodumene areas,irregular small and core (Jahnsand Wright, l95l). It is now virtually impossi- crystals.Optical propertieswere measuredfrom fragments ble to collect in the mine becauseit was largely removed immersedin refractiveindex liquids. X-ray data, obtained with a Philips-Norelco diffractometer,were refined using the poDEx2computer program of A. Sleight(Dupont Cen- STEWARTPEGMATITE GENEMLIZED VERTICALCROSS SECTION tral ResearchLaboratories). Compositional data (summa- rized in Table 2) were primarily obtained using an mr GABBRO PRESENTGROUND SUFFACE EMx-sMelectron microprobe for all constituent elements WALL MTCROCLINECUARTZ-MUSCoVITE-ALBITE (cRAPHIC ZONEr GRANITE) scans. Reduction of OUTER INIERMEDIATE ZONE MICROCLINE{UARTZ.MUSCOVITE,ALSITE SCHORL initially identified by wavelength MIDDLE INTERMEDIATEZONE MICROCLINE-OUARTZ background-and drift-correcteddata was carried out using INNER INTERMEDIATEZONE OUARTZ the MAGICIV computer program of Colby (1968).Selected sampleswere analyzedfor lithium by atomic absorption and water by a micro-coulometrictechnique (Table 3), and MIDDLE INTERMEDIATEZONEr ALBITE{UARTZ-MICROCLINE for their traceelements by emissionspectroscopy (Table 4). wALL zONEr ALBITE{UARTZ MICFOCLINE-SCHORL (LINE ROCK) To date only a small number of secondaryphosphates GABBRO have beenidentified from the StewartLithia mine. In gen- |_OCCU RENCE OF LITHIOPHI LII E phasesare intergrown on a fine scale,and as SYMBO6: A}MICROCLINE XOUARTZ a LEPIDOLITE . AMBLYGjNTE eral these

^ ALBITE . MUSCOVITE O SPODUMENE TOURMALINE such could not be separatedas homogeneousgrains. Thus, Fig. 2. Generalizedvertical crosssection through the southern their completecharacterization was impossiblein somein- portion of the Stewart pegmatiteillustrating the mineral assem- stances.Phases that could not be fully identified are desig- 'K') blagesthat comprisethe internal zoningstructure. nated by a letter (e.g.,phase in the following dis- 398 SHIGLEY AND BRowN: PH)|PHATE MINERALS, STEvART PEGMATITE

Table 2. Representativemicroprobe analyses of selcctedprimary and secondaryphosphates from the Stewartp€gmatite.

Microprobe analyses (velght percent oxides)*

li"eef Lirhiophiltre Sicklerlte Hureaulite I llureaullte II Sanple /l 7621 p54 P25 P54 P25 P54 P25 7621 /l of analyses I 2 122 63 138

P2o5 45.87(6t)+ 46.57(99) 46.68(75) 47.Otl(46) 39.36(61 ) 39.49(37) 31.36(279) 38.03(163) FeO 8.0r(8s) 8.02(r) 5.94(166) 7.89( 18) 9.37(148) 3.77(289) Fe2O3 8. 06(36) 9.O1 (7') unO 36.46(s2) 37.8r(44) 37.12(32) 38.53(22) 39.26(115) 39.62(35) 31.98(482) 38.73(328) Mn2O3

CaO 0.05(2) 0.06(r) 0.12(8) 0.08(3) 0.80(15) 0.32(4) 1.8s(76) r.s2(ros) Mgo n.d.! n.d. n.d. n.d. 0.08(7) 0.07(3) 0.r0(9) 0.r0(3) K2o 0,03(2) n.d. 0.04(3) n.d, n.d, n.d. 0.18(23) 0.04(3) Na2O 0.03(3) 0.04( r ) 0. r3(7) 0.04(1) 0.08(4) n. d. 0.3s(20) 0.27(25'

x 90.45 92.50 92.t5 94.76 85.52 87.39 8r.19 82.46 ** (Li2GlH2o) 9. 55 7. 50 7.85 14,48 t2,61 18.8l L7.54

CelI contents (nmber of atons)f+

Anlon basls 4 4 4 4 20 20 20 20

P l. 008 I .034 l. 00r 1.O23 3. 98t 4. 039 3,546 3.662 Fe 0, I73 o.176 0.153 o.t74 0,595 0.799 0.876 0.355 Mn 0.802 0. 840 o.796 0.838 3,967 4,054 3.040 3.730

ca 0.001 0. 001 o. 003 0.002 0. 100 0. 043 o.222 0.184 Ue 0.014 0.0r4 0.0r3 0.0I4 K 0.00r 0.001 o.o27 0.005 Na 0.001 0. 002 0.006 0.002 0.014 0.067 0.055

(Li+n) 0.999 o.79L 0,943 o.676 lo.7 46 9.822 13.859 t3.O77 t catlons 2,985 2.844 2,903 2.715 19.4r7 18,77r 21.650 2r.O82

'Fr Mineral Ilureaulite III I{ureaulite IV Phase Purpurite I rhase E Phosphoslderite Phase'K' Sauple /l P25 P30 8r22 P30 P30 P25 P30 P30 /l of analyses 8 4 2 8 2 I 2LL

P205 38.s2 (93) 39.s I (49) 40.13(6) 39.7r(s2r) 48.08(66) 36.6l 37.82(95) 40.84(rl6) FeO 0. 48(66) o.27(23\ o.22(r) Fe203 - 8.94(329) 8.22(27) 30.29 39.72(r53) 7.82

CaO 0.2r(16) o.27(33) 0.r0(1) 16.77

Ce1l contents (nuber of atons)

Anion ba6is 20 20 4 6-

P 3.836 3. 961 4,016 1.012 1.001 Fe 0.049 0.028 0.02I 0. r54 0. 934 Mn 4,592 4,636 4.6t4 0.818 0.075

ca 0.028 0,036 0.014 0.004 0.002 Ms o. oo5 0. 007 0. 007 0.002 K- 0.004 0,005 0.001 0.004 Na : 0.012 0,011

(Li+n) 11.389 10,788 to, 592 0, 008 3.942 X cations I 9. 899 19.460 t9 .296 2.OO9 5.896

* ARL EMX-SM electron nicroprobe, operatl.ng at 15 kV and O,l uA (sample current 0.02 pA), spot dleeter < 10 un. - Mlneral standards spessartlne (Mn), apatite (P), willenite (Mg), olivine (Fe), albite (Na,K,ca), orthoclase (Na,K). t'lelght Percentagea of Fe and l4n oxldes reflned on the basls of asaumed ldeal stolchloDetry. f Nunber in parentheses represents the estimated standard deviatlon (esd) tn tems of least units cited. ! n.d. = not detected; values less than 0.03 wt.Z. ** Volatlle content calculated by dlfference, t+ calculated on the basis of assued volatile content (H2o:Li2o in wt.Z) derived from atomic absorption results - llthioPhtlite (0:total), slcklerlte (2:total-2), hureailtte-r (12:totat-I2), all orher phases (eota1:0). SHIGLEY AND BROWN: PHOSPHATE MINERALS. STEWART PEGMATITE 399

Table 3. Data on water and lithium content for selected minerals of lithiophilite from the pegmatite,nor for any compo- from the Stewart pegmatite. sitional zonation within a given crystal. Compared to specimensfrom other pegmatites,Stewart lithiophilite is Mineral H20(wt.Z)* ** No.** L!2o(wE,7.)r No, most similar to material from Branchville (Brush and Average Range Dana, 1878),Varutriisk (Quensel,1940), Wodgina (Mason, rro l-lthioph111te 6.47 4,36-9.61 5 1941),Viitaniemi (Volborth, 1954),and the White Picacho

SlckleriEe 2.0 1 .3.7r 2.27-5.s2rl area(London and Burt, 1982a).At theseoccurrences, lithi-

Hureaulite 1 12.0 I 0.83 0.32-3.r0 ophilite is somewhatmore abundant than triphylite, thus reflectingtheir overall "-rich"pegmatite chemis- Hureeulite Il t4.9 I 0.38 0.37-O,78 3 try. In contrast, triphylite and its secondary alteration

l{ureaulite III 12.0 l 0.0 0.0 I products dominate the phosphate paragenesesat more

Phosphosiderite 18.0 I f,. d. n. d. 0 "iron-rich" granitic pegmatitessuch as those in the Black

'F' Hills (Moore, 1973, 1982),in the New England area Phase 3.7 I 0. 18 0.00-0. 37 2 (Moore, 1973; Segeleret al., 1981),and at Hagendorf Manganese Oxide(s) 35.0 I 0.0 0.0 8 (Strunz,1952; Miicke, 1981).

* llaEer contents oeasured on representatlve, one-nilligram smp1e6 of the indicated ninerals by a nicrocoulometric SickleriteLi, -,( Mnl! * F el* I POn technlque uslng a Dupont Moistute halyzer. Perfomed by M. Cremer, Analytical Branch, U.S. ceological Suney, Basedon textural relationships,sicklerite appears to be Menlo Park (Job /lKS42). ** Number of sanples analyzed for each nlneral. the initial replacementproduct of Stewart lithiophilite. t Lithlun contenEs measured on representative samples of Ehe Sickleriteis typically referredto as a specieswith the above indicated nlnerals by the atomic absotptlon nethod. Using a Perkln-Eluer Model 303 SpeclrophoEmeter, standard solulions simplified formula where the iron is supposedlyall triva- covering lhe range of 1-20 ppn Ll were tun agalnst unknom solutions containlng the dissolved n1nera1s. Values listed lent, the manganeseall divalent, and with a proportionate repiesent caLculated I-i20 contents for each nlneral. tt n.d. = not Deasured. Sufficlent unallered liEhiophilire could amount of lithium removedby leaching(Mason, 1941; not be obtained for a water content deEerfrination. Fleischer,1983). However, analyses of sickleritefrequently showappreciable Mn3* (e.g.,Schaller, 1912; Mason, 1941; Fontan et al.,1976),and consideringthe potential dillicul- ties of analyzingimpure materialssuch as somesecondary cussion. Table 5 summarizes mineralogical data on the phosphatesfor severalvalence states ofiron and/or manga- varioussecondary phosphates examinedduring this study. nese, this formulation for sicklerite may not be entirely correct. Sickleriteis structurally and chemicallyvery simi- Descriptions of individual minerals lar to lithiophilite, and may not representa distinct species but rather only a less well defined intermediatestage be- LithiophiliteLi(Mn, Fe)2+POn tween lithiophilite and purpurite (seeMoore, 1982).Thus, The large crystals of lithiophilite from the mine are further work on pure material, if such becomesavailable, equant to tabular in shapeand have a simplemorphology appearsnecessary to establisha suitableformula for sick- (similar to that noted by Goldschmidt(1923) and Chapman lerite. (1943). From measurementswith a contact goniometeron 68 crystals, the more important crystal facesin terms of both development and frequency of occurrence are: d Table 4. Data on trace elements in a sample of lithiophilite from {011},I {02U, b {010}-present on all crystals;e {120},t the Stewart pegmatite. {110},e {101}-present on somecrystals; and a {100}, v {203}-rarely present. Most specimensrepresent single Concenttatlon Elenents* crystals,but a few consistof two or more intergrown crys- tals (Fig. 4). No twinning relationshipswere noted in such major P, Mn,Fe, LI instances.Frequently, crystal facesare slightly curved or > o. 17, Al, Ca,Na,( otherwisedeformed due perhapsto uneven initial growth <0.t7, Ge,Hf ,U, La or to secondaryalteration ofthe parent lithiophilite. <0.012 Sb,As,8i,3, Cd,Cs, Cr, Co,Cu, Au, In hand specimen,remnant lithiophilite has a pale ln, Ir, Pb,149,Hg,Mo,Ni,Os,Pd, ?t, pinkish-brown color, while under the microscopeit is seen St, Sn,V,Y, Zn, Zr, Th as colorless,blocky areasreplaced by either reddish-orange <0.00r2 Ba, Be,Ga, Ag, S!, Ce hureaulite (Fig. 5) or pale yellow sicklerite. Microprobe not detected Re,Rb, Rh, Ru, Sc, Ta, Te, TI, Ti, W analyses(Table 2) show Stewart lithiophilite to have an Mn/(Mn + Fe) ratio of about 0.8; it displays optical and * S@iquantltatlve spectrographic analysis of a physicalproperties and unit-cellparameters consistent with representative sample of llthiophllite from the Stevart peg@lite. Sample prepared and analyzed valuesexpected on the basis of this composition(Penfleld by M. Pyzyna (Center for Materlals Research, Stanford and Pratt, 1895;Blanchard, 1981; Fransolet et al., 1982). Unlversity) usitrg a Jarlell-Ash 3.4-meter eoission spectlograph. Analyses of sevelal secondary pho6phates All lithiophilite crystals examined have a similar Mn-Fe frofr the pegnatite gave shilar results. content.There is no evidencefor more than one generation 400 SHIGLEY AND BROWN: PHOSPHATE MINERALS, STEWART PEGMATITE

Table 5. Mineralogicaldata for selectedprimary and secondaryphosphate minerals from the Stewartpegmatite.

Mineral Lithlophllite Sicklerite Hureaullte I Hureaulite II Hureaulite III Hureaulite IV Phosphosiderite Sample /l 762r 7 62r P25 P25 P25 862r 8669

(Mn/Mn+Fe) rarlo 0.82 0. 80-0. 82 0.83-0.92 0.71-0,91 0, 99 o.99 0.07-0.08 D 3.42(19)+ 3.37(25) 3.04(26) 2.76(2s) 3.28(6s) 2.65(r0)

Unit-cel1 parametersS

a (8) 6.074(2) 6.030(1s) r7.s89(3) r7.547(r3) 17.623(3) t7.598 s.331(6) !.(4) r0.4sr (2) 10.082(9) 9.097 (2) 9.O7 7(2r) 9 . r27 (3) 9.066 9.795(7) c (H) 4.727 (L) 4. 7s0(s ) 9.433 (3) 9.437(r4) 9.490(4) 9.398 8.710(8)

B (o) 96.78(4) 96.57 ( r 3) 96.s7 (33) 96.58 90.91(r9) V (R3) 300.1 l (8) 288.76 (s2) 1498.79(54) 1493.2r(23r) r516.62(64) r489.50 454.81 (56) /i of reflectlons T9 I7 33 25 37 16

Optical parameters

o 1.669(1) t .7 IO-t ,720 r.648(r) n. d. 1.652-r.654 r.640(2) >r.69 B 1.673(1) r,730-r.738 r.6s9(1) n. d. I.656-r.657 r.649(2) >1.69 Y 1.680(1) >l. 738 r.662ll) n.d. >1.660 r. 6ss(2) >r,69 2V 7oo ' 600 7o-8ou '600 >700 77" moderate (+) (-) (-) (-) (-) (-) n,d.

dispersion n,d. n.d. relief moderate moderate mode!ate mooeiaEe nodera t e mod erate moderate 1ow node!ate moderate noderate nodera t e moderate moderate

color hand specimen pidk-brom yellow- to yellow- to brom- to plnk- to nlnL-rad vJ^l or r^ red brom red brom gray-yellow red-btom blue-violet thln section colorless brom-ye1lo{ colorleasto ye11ow-green colorless colorless pale blue to yellow-green pale vlolet pleochrois none pale to deep colorless to pale green to none nrla hl'ra t^ yellow red-orange red-orange pale vlolet

t Nubers in parentheses represent the estirated standard devlation (eed) in tems of least units cited for the value to thel-r imediate 1eft, TT n.d. = not detemlned due to a lack of sultable naterlal. 5 Reflned unit-cell paraneters derlved from X-ray powder diffraction patterns for all minerals shom *cept Hureaulite IV, whose pa!4eters were neasured from precesslon photographs. A silicon metal lnternal standard (Hubbard gLg,L,, 1975) was used in the preparatl.on of the diffractometer patterns.

In hand specimensicklerite is present as dark brown that these two cations are relatively immobile during the masses.Under the microscopeit is usually seenas being lithiophilite alteration sequencewhen little metasomatism intergrown with other secondaryphosphates (Fig. 6). The is involved (Mason, 1941; Moore, 1973,1982). Stewart brownish-yellowsicklerite has an optical orientation and sicklerite resemblesmaterial from the pegmatitesat Viita- faint relict cleavageinherited from the parent lithiophilite. Sickleritecompositional data (Table 2) agreewith earlier publishedinformation of Schaller(1915). The consistent Mn/(Mn+Fe) ratio (0.80-0.82)in this material confirms

L-J.

Fig. 4. Photograph and superimposed drawing of a lithiophi- lite crystal group from the Stewart Lithia mine. The group con- Fig. 5. Photomicrograph showing the alteration of lithiophilite sists of several intergrown crystals oriented in a parallel fashion. (lith) to hureaulite I (hurl). The lithiophilite appears as colorless, Crystal faces are outlined and are identified by conventional letter high-relief grains surrounded by colorless hureaulite and other designations (seetext). minor secondary phosphates. Plain light. SHIGLEY AND BROWN: PHOSPHATE MINERALS. STEWART PEGMATITE 401 niemi (Mason, l94l; Volborth, 1954),Wodgina (Mason, 1941), the White Picacho district (London and Burt, 1982a),and a locality in Siberia(Kosals, 1968),and repre- sentsone of the more Mn-rich membersof the sicklerite- ferrisickleriteseries yet described(see Fontan et a1.,1976\. pO) pO3OH Hureaulite ( Mn,Fe)!+ ( H 20 ) 4( 2( ) 2 Hureaulite has previously been reported from Pala (Schaller,1912, l9l5;' Mason, 1941; Jahns and Wright, 1951).However, from data gatheredduring this study,four types of hureaulite (labeledI-IV) can be distinguishedon the basis of their chemistry and appearance.We suggest that these are only compositional varieties of the same mineral (or an impure mixture of it and some unknown phase(s)in the caseof hureauliteII) that may or may not be representedin phosphatematerial from other pegma- Fig. 6. Photomicrographof massive,yellow sicklerite(sckl) tites. We do not propose that thesetype designationsbe beingreplaced by palegreen hureaulite II (hur2).Both minerals formally adopted unless it could be shown that their arecut by narrowveinlets of colorlesshureaulite III (hur3).Plain Stewart occurrencein not unique. With the exception of light. hureauliteII, powder diffraction data for the other varieties of hureaulite are consistentwith the calculated and ob- identifiedin hureauliteII powder patterns.From a reexam- servedX-ray patternsfor this mineral (Fisher,1964; Moore ination of Stewartmaterial studiedearlier by Fisher (1958), and Ito, 1978)with no indication of other admixedphases. Moore and Ito (1978)showed "salmonsite" to be a mixture Hureaulite I is a pinkish-, reddish-,or yellowish-brown of hureaulite and jahnsite. Our hureaulite II may be the massivephase associatedwith lithiophilite and sicklerite. sameas the material they described,but jahnsite could not Although normally colorlessin thin section,it sometimes be recognizedin thin sectionor X-ray diffraction patterns. displays vivid orange-redpleochroism. The causeof this A lack of suitablesamples of hureauliteII preventedits full color variation is probably due to differencesin the Mn/Fe characterization.We suggestit is the samephase as Schal- content or to the incipient oxidation of thesecations. Com- ler's "salmonsite"and, at least in part, the sameas the positional data (Table 2) indicate this material is an Fe- material examinedby Moore and Ito (1978). bearinghureaulite, Mn/(Mn+Fe) : 0.83-0.92.This range The third type of Stewart hureaulite, which occurs as of Fe content is consistentwith both the Mn-Fe variabilitv pinkish veinlets,is equivalentto the phase"palaite" of noted in lithiophiliteand sicklerite,and what has beenre- Schaller(1912; also Mason, 1941 ; Moore and Araki, 1973). ported for hureaulite from other localities (Moore and Under the microscope,these veinlets are colorlessand have Araki, 1973;Fransolet, 1976). a fibrous appearance(Figs. 6, 7, and 8). In contrast to the Hureaulite II is more difficult to clearly identify because two previous types, hureaulite III has almost no Fe, of its intergrown associationwith other secondaryphos- Mn/(Mn+Fe) : 0.99,or other minor constituents(Table phates. In hand specimen it appears brownish- or yellowish-gray,while under the microscopeit is pale green with occasionalorange-red (Figs. 6 and 7). With crossed nicols, a random, mosaic arrangement of irregular-shapedgrains or "domains" exhibiting anoma- lous birefringenceand extinction is visible.Like hureaulite I, hureauliteII containssome Fe, Mn/(Mn+Fe) : 0.71- 0.91,but it exhibits Ereatervariability in other constituent oxides.Both types may representthe same generationof this mineral, but the lack of uniform extinction, greater variation in composition,and apparent inhomogeneityof the type II material suggestthat they are not the same. Hureaulite II may possiblybe a mixture of hureaulite and some other unknown phase,although the identity of this other phases(s)could not be clearly establishedfrom X-ray data. Hureaulite II seemsto be related to the enigmaticmin- eral "salmonsite" reported by Schaller (1912).While his Fig. 7. Photomicrograph of massive, pale green hureaulite II description matchesthe observedappearance of our hu- (hur2) containing small areas of violet-red purpurite II (pur2), 'K' reaulite II, X-ray diffraction peaksattributed to "salmon- greenish-yellow veinlets of phase (K), and transected by later site" by Fisher and Atlas (rcros File 13-337)could not be veinlets of colorless hureaulite III (hur3). Plain light. 402 SHIGLEY AND BROWN: PHOSPHATE MINERALS, STEWART PEGMATITE

2). Textural evidencesuggests that material on either side of hureaulite III veinletswas either spreadapart or some- what replacedduring veinletformation. Hureaulite IV occursas small,rose-red euhedral crystals as much as I mm across.It is found in exterior cavitiesin the altered lithiophilite along with phosphosideriteand stewartite(see Murdoch, 1943).It is the least abundant of the four varieties,and like type III, it contains little Fe (Mn/(Mn+Fe) : 0.99;see Table 2). These four types of Stewart hureaulite correspond to material from the pegmatitesat Hagendorf(Strunz, 1954), Mangualde (M6rio de Jesus,1933), La Vilate (Des Cloi- zeaux, 1858),Viitaniemi (Volborth, 1954),and Branchville (Brush and Dana, 1890).From a study of lithiophilite alter- ation at several localities, Fransolet (1976) distinguished yellow (sckl) two genetic types of hureaulite: (1) an early iron-bearing Fig. 8. Photomicrograph of massive, sicklerite and minor pale green hureaulite II (hur2) cut by veinlets of colorlessvariety similar to our types I and II, and (2) a yellowish-greenphase 'K' (K) and then by veinlets of colorless later, iron-poor reddish variety that correspondsto our hureauliteIII (hur3).Plain light. typesIII and IV. ( + Purpurite Mn,F e)3 POn Insufficient material could be found for complete Purpurite, first described from Pala by Graton and characterization, thus making it one of the least well docu- Schaller(1905), is found in two forms in the alteredlithio- mented of the Stewart phosphates. philite. Purpurite I occurs as bright red rims on some yellow sicklerite-these rims are optically continuouswith Minor phases the sicklerite and display a vivid red-greenpleochroism. Several additional secondary phosphates were noted in Purpurite II is presentas red- to brownish-violetareas with the altered lithiophilite but could not be fully identified a fibrous, radiating habit (Fig. 7). Neither type was found because of their limited abundance and impure condition. 'K' in sufficient, homogeneous amounts for complete Phase (Fig. 7, 8) occurs as thin, greenish-yellow vein- characterization,but compositionaldata (Table 2) indicate lets, and may be related to a second type of stewartite 'F'(Fig. both have similar chemistry,Mn/(Mn+Fe) : 0.80.We according to Schaller (1912). Phase 9), found as believe that both types representthe same generationof reddish-brown areas in sicklerite, appears to be a calcic- purpurite. sicklerite similar to the material described by Jahns (1952). Several violet grains, referred to as phase'E', have a rela- PhosphosideriteFe3 + ( H O PO 2 ) 2 4 tively high potassium content, and may be leucophosphite Phosphosideriteis occasionallyfound as conspicuous, (Moore, 1972b\, but this could not be further substantiated. bright, blue-violet microcrystallineaggregates along with The abundant manganese oxides that coat the altered lithi- stewartite in exterior cavities in the altered lithiophilite. ophilite crystals were examined in a preliminary manner by Schaller (1912) referred to this material as "." infrared spectroscopy, which indicated this material to be a Phosphosideritecan sometimesbe confusedwith purpurite II becauseof their similar appearance,but it is more bluish-violet in color whereasthe latter is reddish-violet. Phosphosideriteis the most Fe-rich secondaryphosphate from the pegmatite(Table 2; also Schaller,1915), which servesto further differentiateit from purpurite. In compari- son to material from the pegmatitesat Boqueirio (Mur- doch, 1958) and Pleystein(Wilk, 1960),Stewart phos- phosideriteis remarkablyMn-rich (about3 wt.% MnO). In spite of this, X-ray data are consistentwith the Pleystein material(McConnell, 1939; Wilk, 1960).

StewartiteMnz+ 1H, O) n(Fe3r+ ( OH ) 2(H zO) z (PO+)r) . 2H2O Stewartiteis perhapsthe most noticeablebut at the same time one of the rarest of the Pala phosphatesdescribed by Schaller(1912). It was found as small,bright yellow crystals 'F' on only two alteredlithiophilite crystals,and was identified Fig. 9. Photomicrograph of reddish-brown phase (F) in primarily by its describedassociation with phosphosiderite. massive, yellow sicklerite (sck1). Plain lieht. SHIGLEY AND BROWN: PHOSPHATE MINERALS, STEWART PEGMATITE 403

mixture of severalphases (G. Rossman,oral comm., 1980). phosphatesnoted at other pegmatitesseems to be absentat Characterizationof these minor phasesawaits their dis- the Stewartmine. coveryin quantitiessuitable for completestudy. Secondaryreplacement of lithiophilite was not a con- Several silicate minerals were discovered in small stant volume process,since most of the altered crystals amounts within the altered lithiophilite crystals. Thin exhibit evidencefor changesin size and shape (bulging quartz veinletswith minor pale green muscoviteand blue surfaces,curved fractures, filled veinlets, etc.). However, tourmaline fill fracturesin severalaltered crystals. In addi- there are few if any open cavities or boxwork structures tion, small euhedralcrystals of blue tourmaline within the from which material wasleached out on a large scale. massivelithiophilite is evidencefor the contemporaneous crystallizationof both minerals. Alteration sequence Basedon their own observationsand those of Schaller Secondary alterarion of lithiophilite (1912),Jahns and Wright (1951)suggested the following alteration sequencefor Stewartlithiophilite: lithiophilite - Generalobseruations hureaulite - sicklerite + "salmonsite" + purpurite + The overall extent of lithiophilite alteration seemsto "palaite" + stewartite * phosphosiderite+ manganese have beenindependent of the size of the original crystals. oxides.Our modifled sequenceis shown in Figure 10.The There is no apparentdifference in the nature or extent of replacementof lithiophilite by sicklerite, purpurite, and this alteration with respect to either the kinds of sur- manganeseoxides, first proposedby Quensel(1937,1940\ rounding silicatesor the relative position of a particular and then Mason (19a1),is commonly recognizedat numer- crystal within the intermediatezone of the pegmatite.Due ous localities(Palache et al., L95l; Moore, 1973).The ini- to limited accessibilityof the older mine workings, the tial Mn/(Mn + Fe) ratio of the Stewartlithiophilite is main- degreeof alteration relative to the proximity of either the tained in the secondaryphosphates in the early stagesof present-daytopography or the ground-water level could the sequence,but departsfrom this value in later phases(as not be fully evaluated,but no suchdifferences were noted. noted by Mason, l94l; and sinceby others).Thus, Mn-rich A similar suite of secondaryphosphates was found in lithiophilite is eventually replaced by both Mn-rich and each of the altered crystals,suggesting that they were all Fe-rich secondaryphosphates. originally lithiophilite and not some other primary phos- Alteration reactionsaffecting the Stewartlithiophilite in- phate. The temporal and spatial successionof secondary volved oxidation of Fe and Mn, concomitantleaching of Li phosphateformation was establishedon the basis of tex- (and ultimately P), and hydration. With the exceptionof tural relationships.While there are different degreesof certain minor phases,none of the secondaryphosphates alterationevident among various crystals,the samerelative seemsto have resultedfrom the metasomaticintroduction parageneticsequence among the secondaryphosphates was of externally-derivedcomponents. Thus, almost all of the alwaysobserved. We concludethat the Stewartlithiophilite constituentsrequired for the formation of the observedsec- crystalswere all subjectedto similar alterationconditions. ondary phosphatesoriginated within the lithiophilite crys- Replacementreactions proceeded from the outsideof the tals themselves. original crystalsinward at rates that varied with direction. This variation is apparentin the non-uniform width of the Discussion concentricbands of secondaryphosphates that rim the cen- While the alteration sequencein Figure 10 shows an tral, remnant areasof lithiophilite. No preferentialcrystal- overall temporal relationshipamong the various secondary lographic or compositional control of alteration was evi- phosphates,their intergrown nature indicates that they dent. Initial lithiophilite alteration involved gradual but probably formed in part contemporaneously.Thus, there pervasivereplacement by sicklerite,followed at some later appearsto be no strictly sequentialdevelopment of alter- time by the formation of more hydrated and more oxidized ation reactions,but rather a tendencyfor them to partly phasessuch as hureaulite and purpurite. One important overlapone anotherin time. factor governing alteration was the location of fractures It has been suggestedthat lithium-rich granitic pegma- and cleavageplanes in the original lithiophilite, sincemany tites containing gem pockets,such as the Stewart,crystal- of the secondaryphosphates replaced one another along lized at shallow crustal depths of only severalkilometers suchfeatures. Small-scale transport of componentsby solu- (Ginzburg,1960; Jahns and Burnham,1969;(ern!, 1982; tion also played a role as indicatedby the extensiveveinlet Jahns,1982). If correct,this would indicatethat lithiophi- formation in the altered crystals.However, there is little lite alteration in the Stewart pegmatite occurred at the evidencefor the introduction of externally-derivedconstit- shallow depthsof pegmatiteformation, and then in a near- uents into the lithiophilite alteration products (exceptfor surfaceweathering environment when the pegmatite was 'F'). severalof the uncommon phasessuch as Neither is exposedby uplift and erosion. All secondaryphosphates there any indication for the transport and redistribution of were formed within the original lithiophilite crystals.We lithiophilite-derivedcomponents to other nearby portions believethat lithiophilite replacementbegan during the final of the pegmatite,since secondaryphosphates have so far stagesof pegmatitecrystallization. only beenfound within the alteredcrystals. The stainingof Although availabledata arelimited, the neighboringsili- silicate minerals around phosphatenodules by secondary catesin the pegmatiteappear to have undergonealteration 404 SHIGLEY AND BROWN: PHOSPHATE MINERALS, STEWART PEGMATITE

LI lH I OPHI L I ]E AL]ERATI ON SEOLENCE,S]EWART PEGI4ATI TE

MNGANESEOXIIES

LITH OXIDATION/ LEACHING TIME i 1,7 V> HYDRATIOI'I

Fig. 10. A two-dimensional representation of the proposed alteration sequence for Stewart lithiophilite. Time succession proceeds from the lower left to upp€r right. In general, more hydrated phases (water and/or hydroxyl) are shown progressively to the right, while more oxidized and cationleached phases are shown progressively upward. Symbols next to each element indicate its relative loss (f) or 'F' gain ( f) during alteration reactions. Phase is shown as calcic-sickterite. similar to that of the lithiophilite. The intermediatezones batholith. These magmas were injected into pre-existing of the pegmatite contains some clays of hydrothermal near-surfacefractures in the batholithic host rocks, and origin, but thereare few indicationsof extensivemetasoma- there began to crystallize at temperaturesaround 800"C tic replacementsince most primary silicatesin this portion and pressuresof 1-1.5kbar (3-5 km; seeJahns and Wright, of the pegmatiteare relativelyfresh. 1951; Foord, 1976; Taylor et al., 1979). This lack of extensivemetasomatism among the second- The primary nature of Stewart lithiophilite is clearly ary phosphateswas the mostsurprising result ofthis study, demonstratedby its textural relationships,and we infer a since most secondaryphosphates seem to form by such similar origin for amblygonite.Both minerals crystallized processes(Moore, 1973, 1982).Thus, the extent of meta- within the pegmatitemagma system along with their re- somatismis largely responsiblefor the diversity of second- spectiveassociated primary silicates.Their restricted spa- ary phosphateassemblages at various pegmatites(see Shi- tial distribution within the pegmatite suggeststhat they gley,1982). Since the types of secondaryphosphates reflect formed at specificand possiblylimited periods of time. In the conditions and environmentof pegmatiteparagenesis, both instancestheir occurrenceis related to the internal the phosphate assemblageof the Stewart pegmatite sug- zonal structureof the pegmatite,and not to the locationsof gestsa different set ofalteration conditionsas comparedto later replacementmineralization. many other localities.We believethese different conditions The absenceof primary phosphatesamong the first- are primarily the result of rapid cooling and volatile loss formed mineral assemblagesof the pegmatite wall and from the pegmatiteat the shallowdepth of its formation. outer intermediatezones suggests the need for a period of magmaticdifferentiation to take placewithin the pegmatite magma before the phosphoruscontent reachesa suffrcient Conditions of lithiophilite formation level to permit the crystallization of lithiophilite and The Stewartpegmatite is a typical exampleof a complex amblygonite.The exact level of phosphorussaturation in granitic pegmatite(Cameron et al., 1949; Jahns, 1955; granitic pegmatitic magmasis unknown, but experimental Stewart, 1978; Norton, 1983),and its formation seems resultsof Shigleyand Brown (1982)demonstrate that lithi- compatiblewith the Jahns-Burnhamgenetic model (1969; ophilite can be crystallizedfrom hydrous aluminosilicate Jahns,1982). The graniticpegmatites around Pala crystal- melts containing about 2 wt.oh PrOt. For less siliceous lized from residual,highly-differentiated magmas left over felsic magmas,Watson and Capobianco(1981) suggest a following the consolidation of the Peninsular Ranges saturationlevel of 1.4wt.Yo. SHIGLEY AND BROWN: PHOSPHATE MINERALS. STEWART PEGMATITE 405

Mutual textural relationships imply that lithiophilite tablish the crystallization sequenceamong the primary crystallized along with the surrounding microcline and phosphates.Moore (1973, 1982) proposed the sequence quartz. These features also indicate that the lithiophilite apatite + triphylite + amblygonite.In contrast, London crystals did not grow outward into the magma from some and Burt (1982a)suggested the lithiophilite crystallized point of attachment, as is thought to be the case for min- after montebrasitein pegmatitesof the White Picachodis- erals in the later-formed gem pockets. Rather, field evi- trict. This discrepancymay be a result of differing crys- dence suggests that both lithiophilite and the surrounding tallization conditions at different pegmatites.Our observa- silicates formed contemporaneously in the presence of both tions at the Stewartpegmatite suggest that lithiophilite for- pegmatite magma and exsolved volatile fluid during the mation proceededand may have beenin part concomitant intermediate stage of the Jahns-Burnham model (1969). with the formation of amblygonite,which reflectsthe in- We suggest that phosphorus and other components neces- creasingactivity of phosphorusand fluorine in the pegma- sary for lithiophilite formation were selectively partitioned tite magma system (London and Burt, 1982a, 1982b, into and transported through this interconnected volatile 1982c). fluid to the growing crystals. Similar mechanisms presum- ably contributed to the growth of the neighboring primary silicates. This magmatic behavior on the part of phos- Conditions of lithiophilite alteration phorus is supported by the observation that lithiophilite Secondaryphosphates generally result from the solution, and amblygonite are absent from the lower, albite-rich in- recrystallization,or oxidation of parent phosphate min- termediate zones below the quartz core of the Stewart peg- erals, or by their metasomaticreplacement (eernf, 1970). matite that are believed to have formed at the same time However,relating the phosphateparagenesis to the overall but not in the presenceofthe volatile fluid (Jahns,1982). geologic history of a particular pegmatiteis complicated Moore (1973) concluded that formation of lithiophilite becauseone cannot always directly link the occurrenceof and amblygonite takes place over a temperature range of specific secondaryphosphates to unique alteration con- 500-700'C. Experimental results of Shigley and Brown ditions. A further hinderanceis the lack of thermodynamic (1982) suggest a lower interval of 400-500"C for the crys- data for most phosphateminerals. Thus, while at the pres- tallization of lithiophilite. Deganello (1976) found triphylite ent time the phosphatemineralogy of a pegmatiteat best to be unstable at temperature near 200"C in an oxidizing givesa generalindication of the alteration history, it does atmosphere. The presence of reduced phases such as lithi- provide evidenceregarding this history that is often less ophilite may imply that relatively low oxygen fugacity con- clearlyreflected in other pegmatitemineral assemblages. ditions prevailed during the formation of the Stewart peg- From our observations,all Stewart lithiophilite crystals matite, but this requires conformation. Finally, the crys- were initially alteredin a late-stagehydrothermal environ- tallization of only anhydrous phosphates such as lithiophi- ment, and later under near-surfaceweathering conditions. lite from water-rich pegmatite magmas suggests that hy- Secondaryphosphate formation took placeentirely within drated phosphates are not stable at temperatures ofseveral the altered primary crystals,with releasedcomponents, if hundred degrees Centigrade that are thought to exist any, being carried off by hydrothermal or gtoundwater during the formation of the intermediate zones of a pegma- solutions.Oxidation and cation leachingoccurred in con- tite (Moore, 1973, 1982). junction with hydration and limited introduction of some The scattered distribution of lithiophilite crystals within componentsfrom outsidesources. the intermediate zone of the pegmatite may be indicative of Referring to Figure 10, an indication of the particular a rather slow rate of crystal nucleation with a more rapid conditions of lithiophilite alteration can be gained from growth once crystallization is initiated. Conversely, the severalconsiderations. Both sicklerite and purpurite form components necessary for lithiophilite formation may have under hydrothermal conditions during the final stagesof been limited in the pegmatite magma. The amount of P2Os pegmatiteconsolidation (Moore, 1973;Leavens and Simp- was apparently suffrcient during lithiophilite formation son, 1975; Fontan et al., 1976).Temperatures during this since it was also plentiful during the later formation of stageof alteration are assumedto havebeen approximately amblygonite. The crystallization of lithiophilite could 300-500'C on the basis of secondaryfluid inclusions in rather have been governed by the limited availability of some unalteredlithiophilite. Pressure-correctedhomogen- iron or manganese. Small amounts of both elements could nation temperatures(using the salinity equationsof Potter easily be taken up in garnet or tourmaline, whereas they et al., 1978)fall in the range of 275-350"C for these in- would have been needed as major constituents in lithio- clusions. Such temperaturesfor the hydrothermal alter- philite where MnO + FeO : 45 wt.%,. This may account ation of lithiophilite agree with those of late-stagegem for the absence of lithiophilite as a mineral in the gem pocket formation in the nearby Himalaya pegmatite at pockets, which formed during a later stage when the avail- Mesa Grande (Taylor et al., 1979).Subsequently-formed ability of iron and manganese was somewhat reduced. secondaryphosphates apparently crystallized at evenlower Amblygonite, however, is known as a pocket phase in other temperatures.Hureaulite and phosphosideritecrystallized pegmatites(Palache et al., 1951). in the range of about 150-200'C,which seemsto roughly Because of limited access to the Stewart pegmatite and be the upper stability limit for hydrated secondaryphos- the lack of observable amblygonite, it is diflicult to es- phates (Moore, 1973).The final alteration products like 406 SHIGLEY AND BROWN: PHOSPHATE MINERALS, STEWART PEGMATITE stewartiteand the manganeseoxides formed under super- of the geologyof the Stewartpegmatite have yet to receive geneweathering conditions. careful study. There is little publishedinformation regard- Internally-generatedhydrothermal fluids seem to have ing the lithium aluminosilicateminerals of the pegmatite beenresponsible for some of the lithiophilite replacement. which Stewart(1978) and London and Burt (1982a,1982b, The changingchemistry of theseresidual fluids is reflected 1982c)have shown to provide key information on the geo- in the compositionsof the severalhureaulites-types I and logic evolution of complexgranitic pegmatites.In addition, II contain someiron while typesIII and IV are iron-free.A further laboratory experimentalstudies of pegmatitecrys- similar manganeseenrichment of late-stage pegmatitic tallization, suchas thoseof Shigleyand Brown (1982)to be fluids was noted by Moore and Araki (1973),Fransolet describedin detail in a forthcoming article, are neededto (1976),and Foord (1976).On the other hand, the presence better understandgranitic pegrnatitegenesis. of stewartite and phosphosideriteindicates at least some local enrichmentof iron in theselate pegmatiticfluids. Acknowledgments The lack of extensivemetasomatism among the second- The researchreported above representsa portion of the senior ary phosphatessuggests that the conditions conduciveto author's Ph.D. dissertationat StanfordUniversity under the direc- G. Liou of the suchphosphate alteration apparently did not existfor ex- tion of Profs. G. E. Brown Jr.. R. H. Jahns,and J. Geology Department and Mr. R. C. Erd of the U.S. Geological tended periods during the crystallization of the Stewart Survey,Menlo Park. Accessto the Stewart Lithia mine was pro- pegmatite. vided by Mr. E. Swoboda.We thank M. F. Hochella,D. Fletcher, R. H. Brigham,T. Frost, R. N. Guillemette,D. Pohl, K. Keefer,P. Summary M. Fenn, R. Osiecki,M. L. Zientek,W. E. Dibble, B. H. W. S. de The Stewart pegmatiteis an outstanding exarnpleof a Jong, A. R. Kampl and P. B. Moore for their helpful suggestions, complex, lithium-rich granitic pegmatitewhich exhibits a assistance,and encouragement.The researchproject was support- vertical asymmetryin terms of many of its internal features. ed by the National ScienceFoundation (NSF grant 80-16911)and The sequenceof primary mineral formation is diffrcult to the Geology Department of Stanford University. S. Kingsbury ascertain,but assumingthat crystallizationgenerally pro- drafted the illustrationsin Figures I and 2. M. Havstad prepared ceededfrom the border zone towards the core, then this the photographs in Figure 4. Critical reviews by E. Foord, R. sequencereflects an increasein Li-species(spodumene fol- Cobban,and M. Rossgreatly improved the original versionof this lows microcline) and in volatile content (lepidolite and paper. amblygonite occur along with spodumene).Lithiophilite References and presumablyamblygonite represent primary pegmatite Banks, P. O. and Silver, L. T. (1969)U-Pb isotope analysesof minerals. zircons from Cretaceousplutons of the Peninsularand Trans- Secondaryalteration of lithiophilite involved oxidation, verseRanges, Southern California. (abstr.) Geological Society of hyration, and cation leachingbut little metasomatism.As a AmedcaSpecial Paper, l2l,17-18. result of this study, most of the phosphateminerals first Blanchard,F. N. (1981)X-ray powder diffraction data for lithio- 4, 4V53. describedfrom this areaby Schaller(1912)have been more philite. Florida Scientist, Brush,G. J. and Dana, E. S. (1878)On a new and remarkable fully characterized.The phosphateassemblage at the Stew- mineral locality in Fairfield County, Connecticut; with a de- art pegmatiterepresents a classiclithiophilite alteration se- scription of severalnew speciesoccurring there. American Jour- quencesimilar to that reportedfrom other pegmatitelocal- nal of Science,116,3346 and lL4-123. ities. No new phosphatemineral assemblagesor paragene- Brush, G. J. and Dana, E. S. (1890)On the mineral locality at tic relationshipswere noted. Further study of the possible Branchville, Connecticut. Fifth Paper. Analyses of several alteration of amblygonite,as well as the phosphateminer- manganesianphosphates. American Journal of Science,139, alogy of the other important Pala pegmatites,is needed. 20r-216. The phosphatemineralogy of the Stewartpegmatite pro- Cameron, E. N., Jahns, R. H., McNair, A. H., and Page, L. R. Ge- videsclues regarding mineral alteration during its late- and (1949)Internal Structureof Granitic Pegmatites.Economic post-crystallizationhistory in both hydrothermaland near- ology, Monograph 2. eernj', P. (1970)Review of somesecondary hypogene parageneses surfaceweathering environments. The limited nature of the after early pegrnatite minerals. Freiberger Forschungshefte, phosphateassemblage appears to be primarily due to the ReiheC, (BergakademieFreiberg), 2'10, 4'l-67. lack of metasomatismwhich is so apparent in the phos- dern!, P., Ed. (1982)Granitic Pegrnatitesin Scienceand Industry. phate mineralogyof other pegmatitelocalities. R. H. Jahns Short Course Handbook, vol. 8, Mineralogical Associationof (oral comm., 1978)has suggestedthat the extent of meta- Canada,Winnipeg. somatismcan be related to the depth of pegmatiteforma- Chapman,C. A. (1943)Large magnesia-richtriphylite crystalsin tion. In shallow pegmatites,such as the Stewart,there ap- pegmatite.American Mineralogist, 28, 90-98. parently was little opportunity for residual pegmatitic Colby, J. W. (1968)Quantitative analysisof thin insulating films' fluids to interact with the primary phosphatesunder elev- Advancesin X-ray Analysis,11,287-303. (1976)K-Ar age of San Marcos gabbro and ated temperatureand pressureconditions for an extended Dalrymple, G. B. related gabbroic rocks of the Southern California Batholith. period. Suchconditions that favor extensivemetasomatism Isochron/West,no. 15,35-36. are more likely to exist in pegmatitesformed at greater Deganello,S. (1976)Study of the oxidation of triphylite in the depths. range 2O to 900"C. Zeitschrift fiir Kristallographie, 14, 393- Despiteits notoriety as a gem producer,certain aspects 400. SHIGLEY AND BROWN: PHOSPHATE MINERALS, STEWART PEGMATITE 407

Des Cloizeaux,A. (1858)D6termination des formes cristallineset Jahns,R. H. and Burnham, C. W. (1958)Experimental studies of des propri6tesoptiques de I'hureaulite.Annales de Chimie et de pegmatite genesis:melting and crystallization of granite and Physique,Serie III, 53,293-302. pegrnatite.(abstr.) Geological Society of America Bulletin, 69, Donnelly, M. (1936) Notes on the lithium pegmatitesof pala, 1592-1593. California.Pacific Mineralogist, 3, 8-12. Jahns,R. H. and Burnham, C. W. (1969)Experimental studies of Fisher, D. J. (1958) Pegrnatitephosphates and their problems. pegmatite genesis.I. A model for the derivation and crys- AmericanMineralogist, 43, 181107 and 609-610. tallization of granitic p€gmatites.Economic Geology, 64, 847 Fleischer,M. (1983)Glossary of Mineral Species.Mineralogical 864. Record,Inc., P.O. Box 35565,Tucson, Arizona. Jahns, R. H., Swoboda, E. R., and Larson, W. F. (1974) Fontan, R., Huvelin, P., Orliac, M., and Permingeat,F. (1976)La Tourmaline-bearingpockets in pegmatite of the Pala district, ferrisickl€rite des pegmatitesde Sidi Bou Othmane (Jebilet, San Diego County, California. (abstr.) Geological Society of Maroc) et le groupe min6raux d structurede triphylite. Bulletin America,Abstracts with Programs,6,197-198. de la Soci6t6frangaise de Min6ralogieet de Cristallographie,99, Jahns,R. H. and Wright, L. A. (1951)Gcm- and lithium-bearing 27+286. pegmatitesof the Pala district, San Diego County, California. Foord, E. E. (1976) Mineralogy and petrogenesisof layered California Division of Mines SpecialReport 7A. pegmatite-aplitedikes in the Mesa Grande district, San Diego Kampf, A. R. (1981)The phosphatemineralogy of the Palermo County, California. Ph.D. thesis,Stanford University, Stanford, pegmatite.(abstr.) American Mineralogist, 67, 187-188. California. Kosals,Y. A. (1968)Discoveries of lithiophilite, manganotantalite- Fransolet,A-M. (1976)L'hur6aulite: ses propri6t6s min6ralogiques ,and in zonesof albitization of gneisses.Inter- et son r6le dansl'6volution g6n6tiquedes phases Li(Fe,Mn)PO.. national GeologyReview, lO, 23-25. Bulletin de la Soci6t6frangaise de Min6ralogie et de Cristallo- Krumrnenacher,D., Gastil, R. G., Bushee,J., and Doupont, J. graphie,99,261-273. (1975)K-Ar apparentages, Peninsular Range Batholith, South- Fransolet, A-M., Antenucci, D., Speetjens,J-M., and Tarte, P. ern California and Baja California. Geological Society of (1982)An X-ray determinativemethod for the divalent cation AmericaBulletin, 86,760-768. ratio in the triphyliteJithiophilite series.(abstr.) Geological As- Kunz, G. F. (1905) Gems, jeweler's materials, and ornamental sociationof Canada,Abstracts with Programs,7, 50. stones of California. California State Mining Bureau Bulletin Ginzburg, A. I. (1960)Specific geochemical features of the pegna- Jt. titic process.21st International GeologicalCongress, Copenha- Larsen, E. S. (1948)Batholith and associatedrocks of Corona, gen,77,lll-121. Elsinore, and San Luis Rey quadrangles,Southern California. Goldschmidt, V. (1923) Atlas der Kristallformen. Carl Winters GeologicalSociety of AmericaMemoir 29. UniversitdtsbuchhandlungHeidelberg. Larsen,E. S. (1951)Crystalline rocks of Corona,Elsinore, and San Graton, L. C. and Schaller,W. T. (1905)Purpurite-a new min- Luis Rey quadrangles,Southern California. California Division eral.American Journal of Science,l7O, 146-151. of Mines Bulletin, 159,7-5O. Hubbard, C. R., Swanson,H. E., and Mauer, F. A. (1975)A silicon Leavens,P. B. and Simpson,T. A. (1975)Iron-manganese phos- powder diffraction standard referencematerial. Journal of Ap- phatesfrom the Williams pegmatite,Coosa County, Alabama. plied Crystallography,8, 45-48. MineralogicalRecord, 6, 67-73. Jahns,R. H. (1947)Internal structure of the Pala pegrnatites,San London, D. and Burt, D. M. (1982a)Alteration of spodumene, Diego County, California. (abstr.) Geological Society of montebrasite,and lithiophilite in pegmatitesof the White Pi- AmericaBulletin, 58, 1254-1255. cachodistrict, Arizona.American Mineralogist, 67, 97 -l 13. Jahns,R. H. (1952)Pegrnatite deposits of the White Picacho dis- London, D. and Burt, D. M. (1982b)Lithium aluminosilicateoc- trict, Maricopa and Yavapai Counties, Arizona. Arizona currencesin pegmatitesand the lithium aluminosilicatephase Bureauof Mines Bulletin 162. diagram.American Mineralogist, 6'1, 483493. Jahns, R. H. (1953)The genesisof pegmatites.II. Quantitative London, D. and Burt, D. M. (1982c)Chemical models for lithium analysis of lithium-bearing pegmatite, Mora County, New aluminosilicatestabilities in pegmatitesand granites.American Mexico.American Mineralogist, 38, 1078-l I 12. Mineralogist,67, 494-509. Jahns,R. H. (1954a)Geology of the PeninsularRange Province, Mirio de Jesus, A. (1933) Pegmatites mangano-litiniferasda Southern California and Baja California. In R. H. Jahns,Ed., regiaode Mangualde.Communicag6es dos ServigosGeol6gicos Geology of Southern California, California Division of Mines de Portugal, 19, 65-210. Bulletin 170,Chapter 2, Part 3, 29-52. Mason, B. (1941) Minerals of the Varutrdsk pegmatite.XXIII. Jahns,R. H. (1954b)Geologic guide for the northern part of the Someiron-manganese phosphates and their alterationproducts, Peninsular Range Province, Southern California. In R. H. with specialreference to material from Varutrdsk. Geologiska Jahns,Ed., Geology of SouthernCalifornia, California Division Foereningeni StockholmFoerhandlingar, 61, ll7 -L7 5. of Mines Bulletin 170,Guidebook 5. McConnell, D. (1939) Symmetry of phosphosiderite.American Jahns,R. H. (1955)The study of pegrnatites.Economic Geology, Mineralogist,24, 636-&2. Fiftieth AnniversaryVolume, 1025-1130. (1914) Jahns,R. H. (1979)Gem-bearing pegmatites in SanDiego County. Merrill, F. J. H. Mines and mineral resourcesof San Diego In P. L. Abbott and V. R. Todd, Eds., Mesozoic Crystalline County, California. California State Mining Bureau Report, 14, 691-7ffi and 706-708. Rocks: PeninsularRanges Batholith and Pegrnatites,Point Sal Ophiolite. Geological Society of America, Annual Meeting Moore, P. B. (1970) Crystal chemistry of the basic iron phos- Guidebook,l-38. phates.American Mineralogist, 55, 135-169. Jahns,R. H. (1982)Internal evolution of pegmatitebodies. In P. Moore, P. B. (1971)Crystal chemistryof the alluaudite structure eernf, Ed., Granitic Pegmatitesin Scienceand Industry, Miner- type: contribution to the paragenesisof pegmatitephosphate alogical Association of Canada, Short Course Handbook, 8, giant crystals.American Mineralogist, 56, 1955-L97 5. 293-327. Moore, P. B. (1972a)Phosphate paragenesis in granitic pegrna- 408 SHIGLEY AND BROWN: PHOSPHATE MINERALS, STEWART PEGMATITE

tites. (abstr.) Geological Society of America, Abstracts with Pro- Varutriisk. Geologiska Foereningen i Stockholm Foerhand- grams,4, 601. lingar,62,291-296. Moore, P. B. (1972b) Octahedral tetramer in the Schaller, W. T. (1912) New manganese phosphates from the gem of leucophosphite K2(Fei+(OH)r(HrO)r(PO4)4) . 2HrO. Ameri- tourmaline field of Southern California. Washington Academy can Mineralogist, 57, 397-410. of Science Journal, 2,143-145. Moore, P. B. (1973) Pegmatite phosphates: descriptive mineralogy Schaller, W. T. (1915) Analyses of rocks and minerals. United and crystal chemistry. Mineralogical Record,4, 103-130. States Geological Survey Bulletin, 59I, 348-352. Moore, P. B. (1981) Pegrnatite minerals. In K. Frye, Ed., The Segeler, C. G., Kampf, A. R., Ulrich, W., and Whitmore, R. W. Encyclopedia of Mineralogy, p. 362-167. Hutchinson Ross, (1981) Phosphate minerals of the Palermo No. I pegmatite. Stroudsburg, Pennsylvania. Rocks and Minerals. 56.196-214. Moore, P. B. (1982) Pegmatite minerals of P(V) and B(III). In P. Shigley, J. E. (1982) Phosphate minerals in granitic pegmatites: a eernf, Ed., Granitic Pegmatites in Science and Industry, Miner- study of primary and secondary phosphates from the Stewart alogical Association of Canada, Short Course Handbook, 8, pegmatite, Pala, California. Ph.D. thesis, Stanford University, 267-29r. Stanford. California. Moore, P. B. and Araki, T. (1973) Hureaulite Mn3* Shigley, J. E. and Brown, G. E. Jr. (1980) Phosphate mineralogy of (HrO)4(PO3(OH))r(POn)r: its atomic arrangement. American the Stewart pegmatite, Pala district, San Diego County, Califor- Mineralogist, 58, 302-307. nia. (abstr.) Geological Society of America, Abstracts with Pro- Moore, P. B. and Ito, J. (1978) I. Whiteite, a new species, and a grarns,12,521. proposed nomenclature for the jahnsite-whiteite complex series. Shigley, J. E. and Brown, G. E. Jr. (1982) Reconnaissance experi- II. New data on xanthoxenite. III. Salmonsite discredited. Min- mental study of lithiophilite crystallization from hydrous alumi- eralogical Magazine, 42, 309-323. nosilicate melts. (abstr.) Geological Society of America, Ab- Moore, P. B. and Molin-Case, J. (1974) Contribution to pegmatite stracts with Programs, 14, 616. phosphate giant crystal paragenesis. II. The crystal chemistry of Sinkankas, J. (1957) Recent gem mining at Pala, San Diego wyllieite Na2Fe;+A(PO+):, a primary phase. American Miner- County, California. Gerns and Gemology, 9, 8G-87 and 95. alogist, 59, 280_290. Stewart, D. B. (1978) Petrogenesis of lithium-rich pegmatites. Miicke, A. (1981) The parageneses of the phosphate minerals of American Mineralogist, 63, 970-980. the Hagendorf pegmatite-a general view. Chemie der Erde, 40, Strunz, H. (1952) Die phosphat-quarz-feldspatpegmatite von 217-234. Hagendorf-Pleystein in Bayern. Neues Jahrbuch fiir Mineral- Murdoch, J. (1943) Crystallography of hureaulite. American Min- ogie, Abhandlungen, 84, 77 -92. eralogist,28,19 24. Strunz, H. (1954) Identitiit von wenzelit und baldaufit mit Murdoch, J. (1958) Phosphate minerals of the Borborema pegma- hur6aulith. Neues Jahrbuch fiir Mineralogie, Monatshefte, 166- tites. IL Boqueirflo. American Mineralogist, 43, I 148-1 156. 177. Murdoch, J. and Webb, R. W. (1966) Minerals of California (Cen- Taylor, B. E., Foord, E. E., and Friedrichsen, H. (f979) Stable tennial Volume, 186G1966). California Division of Mines Bulle- isotope and fluid inclusion studies of gem-bearing granitic tin 189. pegrnatite-aplite dikes in San Diego County, California. Contri- Norton, J. J. (1983) Sequence of mineral assemblages in differ- butions to Mineralogy and Petrology, 68,18'1-205. entiated granitic p€gmatites. Economic Geology, 78, 854-87 4. van Wambeke, L. (1971) The problem of cation deficiencies in Palache, C., Berman, H., and Frondel, C. (1951) Dana's System of some phosphates due to alteration processes.American Miner- Mineralogy, 7th edition. volume 2. Wiley, New York, New alogist, 56, 1366-1384. York. Volborth, A. (1954) Phosphatminerale aus dem lithiumpegrnatit Penfield, S. L. and Pratt, J. H. (1895) Effect of the mutual replace- von Viitaniemi, Eriijdrvi, Zentral-Finnland. Annales Academiae ment of manganese and iron on the optical properties of lithio- Scientarium Fennicae,Series A III,39, 1-90. philite and triphylite- American Journal of Science, 150, 387- Watson, E. B. and Capobianco, C. J. (1981) Phosphorus and the 390. rare earth elements in felsic magmas: an assessmentof the role Potter, R W., Clynne, M. A., and Brown, D. L. (1978) Freezing of apatite. Geochimica et Cosmochimi ca Acta, 45, 2349-2358. point depression of aqueous sodium chloride solutions. Eco- Wilk, H. (1960) Phosphosiderit und strengit von Pleystein in nomic Geology, 73, 284-285. Ostbayern. Acta Albertina Ratisbonensia (Regensburger Natur- Quensel, P. (1937) Minerals of the Varutriisk pegmatite. I. The wissenschaften),23, 107-170 (not seen, extracted from Mineral- lithium-manganese phosphates. Geologiska Foereningen i ogical Abstracts, 16, 372). Stockholm Foerhandlingar, 59, 77-96. Quensel, P. (1940) Minerals of the Varutrdsk pegrnatite. XVI. Manuscript receit:ed,January 25, 1984; Lithiophilite, a third primary alkali-manganese phosphate from acceptedfor publication, October 26, 1984.