HARVARD UNIVERSITY

Library of thè

Museum of

Comparative Zoology

■

.

Volume XXX - Fascicolo I

della Società Italiana

di Scienze Naturali

e del Museo Civico

di Storia Naturale di Milano

MINER ALO G Y AND PETROLOGY OF SHALLOW

DEPTH PEGMATITES

Papers from thè First International Workshop

Edited by

FEDERICO PEZZOTTA

MILANO 29 SETTEMBRE 2000

Elenco delle Memorie della Società Italiana di Scienze Naturali

e del Museo Civico di Storia Naturale di Milano

Volume I

I - Cornalia E., 1865 - Descrizione di una nuova specie dei

genere Felis: Felis jacobita (Corn.). 9 pp., 1 tav.

II - Magni-Gkiffi F., 1865 - Di una specie d 'Hippolais nuova

per l’Italia. 6 pp., 1 tav.

Ili - Gastaldi B.. 1865 - Sulla riescavazione dei bacini lacustri

per opera degli antichi ghiacciai .30 pp.,2 figg., 2 tavv.

IV - Sequenza G, 1865 - Paleontologia malacologica dei terre¬

ni terziarii del distretto di Messina. 88 pp., 8 tavv.

V - Gibelli G., 1865 - Sugli organi riproduttori del genere Ver¬

rucaria. 16 pp., 1 tav.

VI - Beggiato F. S., 1865 - Antracoterio di Zovencedo e di

Monteviale nel Vicentino. 10 pp., 1 tav.

VII - Cocchi I..1865 - Di alcuni resti umani e degli oggetti di

umana industria dei tempi preistorici raccolti in Toscana.

32 pp., 4 tavv.

Vili - Targioni-Tozzetti A. 1866 - Come sia fatto l’organo che

fa lume nella lucciola volante dell'Italia centrale (Luciola

italica) e come le fibre muscolari in questo ed altri Insetti

ed Artropodi. 28 pp., 2 tavv.

IX - Maggi L., 1865 - Intorno al genere Aeolosoma. 18 pp., 2

tavv.

X - Cornalia E., 1865 - Sopra i caratteri microscopici offerti

dalle Cantaridi e da altri Coleotteri facili a confondersi con

esse. 40 pp., 4 tavv.

Volume II

I - Issel A., 1866 - Dei Molluschi raccolti nella provincia di Pi¬

sa. 38 pp.

II - Gentilli A., 1866 - Quelques considérations sur l’origine

des bassins lacustres, àpropos des sondages du Lac de Co¬

me. 12 pp., 8 tavv.

Ili - Molon E, 1867 - Sulla flora terziaria delle Prealpi venete.

140 pp.

IV - D'Achiardi A., 1866 - Corallarj fossili del terreno num-

mulitico delle Alpi venete. 54 pp., 5 tavv.

V - Cocchi I., 1866 - Sulla geologia dell’alta Valle di Magra. 18

pp., 1 tav.

VI - Sequenza G., 1866 - Sulle importanti relazioni paleontolo¬

giche di talune rocce cretacee della Calabria con alcuni ter¬

reni di Sicilia e dell’Africa settentrionale. 18 pp., 1 tav.

VII - Cocchi I., 1866 - L’uomo fossile nell’Italia centrale. 82 pp.,

21 figg., 4 tavv.

Vili - Garovaglio S., 1866 - Manzonia cantiana, novum Liche-

num Angiocarporum genus propositum atque descriptum. 8

pp. 1 tav.

IX - Seguenza G., 1867 - Paleontologia malacologica dei terre¬

ni terziarii del distretto di Messina (Pteropodi ed Eteropo-

di). 22 pp., 1 tav.

X - Durer B„ 1867 - Osservazioni meteorologiche fatte alla

Villa Carlotta sul lago di Como. ecc. 48 pp. 11 tavv.

Volume III

I - Emery C., 1873 - Studii anatomici sulla Vipera Redii. 16

pp., 1 tav.

II - Garovaglio S., 1867 - Thelopsis, Belonia, Weitenwebera et

Limboria, quatuor Lichenum Angiocarporum genera reco-

gnita iconibusque illustrata. 12 pp., 2 tavv.

Ili - Targioni-Tozzetti A., 1867 - Studii sulle Cocciniglie. 88

pp., 7 tavv.

IV - Claparède E. R. e Panceri P, 1867 - Nota sopra un Al-

ciopide parassito della Cydippe densa Forsk. 8 pp. 1 tavv.

V - Garovaglio S., 1871 - De Pertusariis Europae medine

commentano. 40 pp., 4 tavv.

Volume IV

I - D’Achiardi A., 1868 - Corallarj fossili del terreno num-

mulitico dell’ Alpi venete. Parte 11. 32 pp. 8 tavv.

II - Garovaglio S., 1868 - Octona Lichenum genera vel adhuc

controversa, vel sedis prorsus incertae in systemate, novis

descriptionibus iconibusque accuratissimi s illustrata. 18 pp.,

2 tavv.

Ili - Marinoni C., 1868 - Le abitazioni lacustri e gli avanzi di

umana industria in Lombardia. 66 pp., 5 figg., 7 tavv.

IV - (Non pubblicato).

V - Marinoni C, 1871 - Nuovi avanzi preistorici in Lombar¬

dia. 28 pp., 3 figg., 2 tavv.

NUOVA SERIE

Volume V

I - Martorelli G., 1895 - Monografia illustrata degli uccelli

di rapina in Italia. 216 pp., 46 figg., 4 tavv.

Volume VI

I - De Alessandri G., 1897 - La pietra da cantoni di Rosi-

gnano e di Vignale. Studi stratigrafici e paleontologici. 104

pp., 2 tavv., 1 carta.

II - Martorelli G. 1898 - Le forme e le simmetrie delle mac¬

chie nel piumaggio. Memoria ornitologica. 112 pp., 63 figg.,

1 tavv.

Ili - Pavesi R, 1901- L’abbate Spallanzani a Pavia. 68 pp., 14

figg., 1 tav.

Volume VII

I - De Alessandri G., 1910 - Studi sui pesci triasici della Lom¬

bardia. 164 pp., 9 tavv.

Volume Vili

I - Repossi E., 1915 - La bassa Valle della Mera. Studi petro¬

grafia e geologici. Parte I .pp. 1-46, 5 figg., 3 tavv.

II - Repossi E., 1916 (1917) - La bassa Valle della Mera. Studi

petrografici e geologici. Parte II. pp. 47-186, 5 figg. 9 tavv.

Ili - Airaghi C., 1917 - Sui molari d’elefante delle alluvioni

lombarde, con osservazioni sulla filogenia e scomparsa di

alcuni Proboscidati. pp. 187-242, 4 figg., 3 tavv.

Volume IX

I - Bezzi M. 1918 - Studi sulla ditterofauna nivale delle Alpi

italiane, pp. 1-164, 7 figg. 2 tavv.

II - Sera G. L., 1920 - Sui rapporti della conformazione della

base del cranio colle forme craniensi e colle strutture della

faccia nelle razze umane. (Saggio di una nuova dottrina

craniologica con particolare riguardo dei principali cranii

fossili), pp. 165-262, 7 figg., 2 tavv.

Ili - De Beaux O. e Festa E., 1927 - La ricomparsa del Cin¬

ghiale nell’Italia settentrionale-occidentale, pp. 263-320, 13

figg., 7 tavv.

Volume X

I - Desio A., 1929 - Studi geologici sulla regione dell’Albenza

(Prealpi Bergamasche), pp. 1-156, 27 figg., 1 tav., 1 carta.

II - Scortecci G., 1937 - Gli organi di senso della pelle degli

Agamidi.pp. 157-208, 39 figg. 2 tavv.

Ili - Scortecci G., 1941- 1 recettori degli Agamidi.pp. 209-326,

80 figg.

Volume XI

I - Guigilia D., 1944 - Gli Sfecidi italiani del Museo di Mila¬

no ( Hymen .). pp. 1-44, 4 figg., 5 tavv.

n-m- Giacomini V. e Pignatti S., 1955 - Flora e Vegetazione del¬

l’Alta Valle del Braulio. Con speciale riferimento ai pasco¬

li di altitudine, pp. 45-238, 31 figg., 1 carta.

Volume XII

I - Vialli V., 1956 - Sul rinoceronte e l’elefante dei livelli su¬

periori della serie lacustre di Leffe (Bergamo), pp. 1-70, 4

figg. 6 tavv.

I - Venzo S., 1957 - Rilevamento geologico dell’anfiteatro

morenico del Garda. Parte I: Tratto occidentale Gardone-

Desenzano pp. 71-140, 14 figg., 6 tavv., 1 carta.

Ili - Vialli V., 1959 - Ammoniti sinemuriane del Monte Alben-

za (Bergamo), pp. 141-188,2 figg., 5 tavv.

MINERALOGY AND PETROLOGY OF SHALLOW

DEPTH PEGMATITES

Papers of thè First International Workshop

edited by

Federico Pezzotta

Mineralogy Section of Museo Civico di Storia Naturale of Milan

Volume XXX - Fascicolo I

29 settembre 2000

Memorie della Società Italiana di Scienze Naturali

e del Museo Civico di Storia Naturale di Milano

© Società Italiana di Scienze Naturali e

Museo Civico di Storia Naturale di Milano

corso Venezia, 55 - 20121 Milano

In copertina: disegno di Federico Pezzotta

Registrato al Tribunale di Milano al n. 6694

Direttore responsabile: Anna Alessandrello

Direttore scientifico: Luigi Cagnolaro

Stampa: Litografia Solari, Peschiera Borromeo - settembre 2000

ISSN 0376-2726

3

FOREWORD

For thè first time in Italy, in September 11 to 13,

1997, an international meeting dedicateci to thè

petrology and mineralogy of pegmatic rocks, with ti-

tle "Petrology, rare minerals and gemstones of shal-

low - depth pegmatites", was organized in thè Museo

Civico di Storia Naturale of Milan. The meeting was

dedicated to Ettore Artini (1866-1928), excellent

mineralogist and petrologist, former curator of Min¬

eralogy (1893-1912) and Director (1912-1928) of thè

Museo Civico di Storia Naturale of Milan.

This volume contains papers presented at thè Con-

ference. I thank heartily thè invited speakers, and thè

other authors for having prepared thè manuscripts.

All papers were submitted to thè normal refereeing

procedures of Memorie. A special thank is to Dr.

Paolo Arduini curator of Vertebrate Paleofitology of

thè Museo Civico di Storia Naturale, for having col-

laborated to thè organization of thè Meeting, and in

particular for having recovered thè necessary finan-

cial support.

This Conference "Petrology, rare minerals and

gemstones of shallow - depth pegmatites" was orga¬

nized by thè Museo di Storia Naturale of Milan, with

thè collaboration of Società Italiana di Mineralogia e

Petrografia. The Conference was integrated with an

exhibition of thè minerals of thè localities of Baveno

and Cuasso al Monte, of thè historic collections of thè

Museum.

The field trip program consisted of two excursions

of two days. TÌie first one, on September 9-10, orga¬

nized by Museo Civico di Storia Naturale of Milan in

collaboration with Centro Studi per la Geodinamica

Alpina e Quaternaria, of thè C.N.R., had title "The

contaminated pegmatites of thè Sissone Valley".This

excursion allowed thè visit of a swarm of pegmatitic

dikes related with thè Masino-Bregaglia Alpine plu-

ton and hosted in carbonatic marbles.The second ex¬

cursion, on September 14-15, had title "The mineral¬

ogy of thè pegmatitic miarolitic cavities of thè

Baveno pink granite, of thè Montorfano granitic

stock and of thè Cuasso al Monte granophyre". Dur-

ing thè excursion it was possible to observe in many

quarries and naturai outcrops thè miaroltic cavities

and a number of aplitic and pegmatitic veins typical

of these post-Hercynian granitoids.

The papers presented in this volume were submit¬

ted in September-October 1997, and were accepted in

October 1998.

Federico Pezzotta

4

.

.

: ■

:

Memorie della Società Italiana di Scienze Naturali e del Museo Civico di Storia Naturale di Milano

Volume XXX - pp.: 5-12, 2000

Constitution, petrology, affìliations and categories

of miarolitic pegmatites

Petr Cerny*

Geologica! Sciences, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2

Abstract - Miarolitic cavities form in late stages of primary solidification of granitic pegmatites, and mainly in those

which are geochemically evolved. Miarolitic vugs are encountered in facial pegmatitic pods and crosscutting pegmatite dikes

interior to their granitic parents, commonly accumulated upwards, and in dikes emplaced into thè metamorphic envelope

of parent plutons. Within individuai pegmatites, thè miarolitic vugs are centrally located, or in thè upper parts of thè peg¬

matite bodies. Early (i) lining of thè cavities consists of euhedral crystals which terminate concentric aggregates of rock-

forming magmatic minerals, coarsening toward thè open space and exhibiting prominent textural and minerai zoning. Sub-

sequent (ii) coating of thè lining is formed by a variety of rare-element minerals (Li, Cs, Be, Y, REE, Sn,Ti, Nb,Ta, Sb), com¬

monly with significant contents of volatile components (F, B, P, H20), deposited from a supercrutical fluid which frequent-

ly corroded thè lining. The last generation of vug-coating or -filling minerals consists of (iii) low-temperature hydrothermal

minerals, in part formed at thè expense of earlier phases and from components leached out from broader vicinity of thè cav¬

ities; micaceous and clay minerals are typical, but low-temperature phases with Li, Be, Y, REE, Ti and P also are found.

Miarolitic cavities form by exsolution of a hydrous supercritical fluid phase from pegmatite melt, whenever it becomes

volatile-oversaturated in (usually) late stages of its consolidation. Build-up of thè exsolved fluid within thè pegmatite is a

prerequisite, but commonly hindered by escape of thè fluid into thè host rocks. Exsolution is triggered by (i) vapor satura-

tion during thè progress of more or less isobaric magma crystallization, or by (ii) reduction of pressure in shallow-intruded

pegmatites, or by (iii) massive stabilization of minerals containing components which enhance solubility of water (B, P, F,

Li). Whereas (i) can generally operate in any pegmatite, independent of any special set of conditions, thè pressure quench

(ii) is operating mainly in pegmatites of thè subaluminous, A-type NYF family, and thè Chemical quench (iii) is typical of (in

part deeper-seated) pegmatites of thè peraluminous, S-type LCT family. Collaboration of pressure and Chemical quenches

is probable in shallow-seated LCT pegmatites. Shattering of thè vug-coating crystals is encountered in pegmatite popula-

tions in which pressure of thè exsolved fluid ruptured thè enclosing pegmatite and explosively escaped; however, subsequent

implosion could also have participated. Extensive leaching and corrosion of thè cavity-lining and -coating minerals indicates

prominent variations in thè chemistry of thè vug fluid.

Recent data on thè conditions of formation and solidification of thè miarolitic pegmatites, in conjunction with better un-

derstanding of other pegmatite categories, indicate a need to reconsider thè classification systematics of granitic pegmatites.

Riassunto - Le cavità miarolitiche si formano negli stadi più tardivi della solidificazione delle pegmatiti granitiche, ed in

maggior misura in quelle geochimicamente evolute. Cavità miarolitiche sono presenti entro ammassi e filoni pegmatitici for¬

matisi entro il magma parentale, frequentemente concentrati nelle parti più superficiali dei plutoni, ed entro filoni messisi

in posto nelle rocce dell’aureola metamorfica attorno alle masse intrusive dalle quali si sono originati. Entro i filoni peg¬

matitici le cavità miarolitiche sono presenti nelle porzioni centrali, oppure nelle porzioni superiori. Le cavità sono rivestite

nelle fasi più precoci della loro formazione (i) da cristalli idiomorfi di minerali costituenti fondamentali della roccia mag¬

matica. Tali cristalli non sono altro che le terminazioni di aggregati concentrici di cristalli caratterizzati da grana più grossa

avvicinandosi alle cavità e da evidenti zonature tessiturali e mineralogiche. Nelle fasi più tardive della formazione (ii) i

cristalli precedentemente formatisi vengono ricoperti da una varietà di minerali contenenti elementi rari (Li, Cs, Be, Y,

REE, Sn,Ti, Nb,Ta, Sb), frequentemente con significativi contenuti di componenti volatili (F, B, P, FLO), depositatisi da un

fluido supercritico in grado frequentemente di corrodere i cristalli di minerali formatisi più precocemente. L'ultima gener¬

azione di minerali (iii), i quali ricoprono o riempiono completamente le cavità, si forma da soluzioni idrotermali di bassa

temperatura. Tali minerali si formano in parte a spese di fasi precedentemente cristallizatesi entro le cavità o nelle imme¬

diate vicinanze delle stesse. Di questo stadio sono tipici i minerali micacei e le argille, ma si possono anche formare miner¬

ali contenenti Li, Be, Y, REE, Ti e P.

Le cavità miaroltiche si formano per essoluzione di una fase fluida acquosa supercritica da un fuso pegmatitico, allorché

diventa supersaturo in volatili durante (solitamente) gli stadi più tardivi della solidificazione. Per la formazione di cavità mi¬

arolitiche l’essoluzione di fluidi è un prerequisito, anche se comunemente i fludi, al posto di concentrarsi a dare cavità, sfug¬

gono nelle rocce incassanti i filoni. L’essoluzione di fluidi è favorita da (i) saturazione durante la cristallizzazione più o meno

isobarica del magma, (ii) dalla riduzione di pressione in pegmatiti intrusesi a bassa pressione, (iii) dalla raggiunta stabilità

di minerali contenenti elementi in grado di aumentare la solubilità dell’acqua nel fuso pegmatitico (B, P, F, Li). Sebbene (i)

la diminuzione di pressione possa operare in qualunque pegmatite, indipendentemente da qualunque condizione, questa (ii)

è responsabile della formazione di cavità soprattutto nelle pegmatiti suballuminose, della famiglia A-type NYF. I fattori

chimici (iii) sono responsabili della formazione di cavità in pegmatiti perallumminose (talvolta anche di maggiore profon¬

dità), della famiglia S-type LCT. Il contributo combinato di fattori barici e fattori chimici nella formazione di cavità è prob¬

abile nelle pegmatiti LCT intrusesi a bassa profondità. Frantumazioni dei cristalli sono osservabili in popolazioni di peg¬

matiti nella quali la pressione dei fluidi essoluti è stata tale da fratturare le pegmatiti e sfuggire in modo esplosivo; tuttavia,

una successiva implosione potrebbe a sua volta provocare ulteriori frantumazioni. Abbondanti dissoluzioni e corrosioni dei

minerali che ricoprono le cavità indicano significative variazioni della composizione chimica dei fluidi nelle geodi.

* E-mail: p_cerny@umanitoba.ca

6

PETR CERNY

Dati recenti sulle condizioni di formazione e solidificazione delle pegmatiti miarolitiche, insieme ad una migliore com¬

prensione delle altre categorie di pegmatiti, indicano la necessità di riconsiderare la sistematica classificativa delle pegmati¬

ti granitiche.

Key words: granitic pegmatite, fluid exsolution, miarolitic cavity, igneous petrology, classification.

Introduction

(Submitted Octobre 1997, accepted as amended

Octobre 1998, published 2000).

Miarolitic pegmatites sensu lato are pegmatites

which characteristically contain open cavities

formed during thè final stages of primary pegmatite

consolidation. The italics emphasize three principal

features of these pegmatites: (i) their diversity, (ii) a

quantitative aspect of thè cavity occurrence, and

(iii) a genetic aspect of thè cavities.

(i) Miarolitic cavities (vugs, pockets) are encoun-

tered in all petrogenetic categories of granitic peg¬

matites, but their abundance is highly variable.

Abyssal-class and muscovite-class pegmatites are al-

most devoid of miarolitic vugs, but they occur with

increasing frequency in pegmatites of thè rare-ele-

ment class and particularly in pegmatites of thè mi¬

arolitic class proper (cf. Cerny, 1991a for pegmatite

classification, but see comments in thè section on

“Geologie Classification” below).

(ii) However, an odd pocket here and there in a

pegmatite body or a swarm of cogenetic pegmatites

does not qualify them for thè miarolitic designation.

The term should be reserved for pegmatites and

pegmatite populations which contain significant

numbers of miarolitic vugs (e.g. thè Ramona, Pala

and Mesa Grande distriets of southern California -

Foord, 1976, 1977, Shigley et al., 1986, Foord et al.

1989 and thè Lake George intrusive center, Harris

Park, Wigwam Creek and other areas in thè Pikes

Peak batholith. Colorado - Foord, 1982), relative to

their cavity-poor to cavity-free counterparts (e.g.,

thè Winnipeg River district of southeastern Manito¬

ba - Cerny et al., 1981, and thè South Piatte district

in thè Pikes Peak batholith, Colorado - Simmons

and Heinrich, 1980, Simmons et al., 1987).

(iii) Miarolitic cavities originate in late stages of

consolidation of thè host pegmatites, showing a dis-

tinct textural and paragenetic evidence of continuity

of primary magmatic crystallization from thè sur-

rounding massive pegmatite into thè walls of thè

cavities (e.g., Némec, 1992). This is in sharp contrast

to secondary cavities formed by dissolution of mas¬

sive minerai assemblages, triggered by low-tempera-

ture fluids which corroded and leached thè primary

phases and redeposited (some of) their components

in thè form of hydrothermal assemblages (e.g., Lan-

des, 1925, Cerny, 1972).

Distribution of miarolitic cavities

Miarolitic cavities are encountered mainly in two

kinds of granitic pegmatites: in facial pegmatitic

pods within their parent granites, and in intrusive

pegmatite bodies which are interior up to (domi-

Fig. 1 - Asymmetric miarolitic pegmatite of thè Korosten pluton,

Ukraine, in pian (top) and vertical section (bottom), with thè cav¬

ity located in its upper part: L’s - graphic granite, z’s - coarse

graphic to apographic pegmatite, crosshatched - K-feldspar, ruled

- quartz (after Lazarenko et al., 1973).

nantly) exterior to their plutonio sources.

Facial pods of miarolitic pegmatites commonly

are concentrated in thè cupolas of thè host granites.

They usually are surrounded by haloes somewhat

enriched in mafie phases (predominantly biotite)

and they are strongly zoned in terms of mineralogy

and texture. Coarse-grained granitic assemblages

are followed by a graphic feldspar(s) + quartz zone,

coarsening in grain size toward thè more-or-less

centrai cavity and transitional into blocky feldspar

and quartz in its immediate vicinity (Fig. 1, 2; e.g.,

Jahns, 1954, Lazarenko et al., 1973).

In dikes and sills of pegmatites which crosscut

their granitic parents or their metamorphic enve-

lope, miarolitic vugs are concentrated in late units of

primary consolidation of their host pegmatites, such

as thè blocky core margin (Fig. 3) or lepidolite units.

The cavities tend to accumulate in centrai or upper

parts of pegmatite bodies (Fig. 4, 5), although thè

vertical range of cavity distribution may be quite

substantial even in steeply-dipping veins (e.g., Lyck-

berg and Rosskov, 1997). In quartzo-feldspathic as-

CONSTITUTION , PETROLOGY. AFFILIATIONS AND CATEGORIES OF MIAROLITIC PEGMATITES

7

fi Ili n g of clinochlore

1 m

_ i

Fig. 2 - Schematic cross-section through a typical concentric miarolitic pegmatite of thè subvolcanic portions of thè Pikes

Peak batholith (from Foord et al., 1995).

Fig. 3 - Schematic structure of thè Mursinka pegmatites, Alabash-

ka region, Ural Mts.: a - granite, b - aplitic zone, c - graphic zone,

d - blocky K-feldspar and quartz, e - centrally located miarolitic

cavities, in part connected by quartz veinlets (after Fersman,

1960).

CONSTITUTION OF MIAROLITIC CAVITIES

The minerai paragenesis of thè miarolitic cavities

proper can be roughly subdivided into three succes¬

sive assemblages: (i) early minerals constituting thè

walls of thè vugs, (ii) a coating of later minerals de-

posited on thè walls, and (iii) a further low-tempera-

ture assemblage covering thè preceding phases or fill-

ing thè remaining space.

(i) The early minerals lining miarolitic vugs typi-

cally are thè rock-forming magmatic phases of thè

host pegmatite zone, typically K-feldspar, albitic pla-

gioclase, quartz +/- tourmaline, muscovite (but also

lepidolite and quartz in late units). The minerals ex-

fend their growth from thè surrounding massive peg¬

matite into thè open space of cavities, with euhedral

terminations. In dose vicinity of thè vugs, thè fabric

of thè minerai aggregates commonly turns from ran-

domly disoriented to concentric, indicative of a

growth pattern oriented toward thè open space of thè

cavity. The grain size strongly tends to increase in thè

same direction. Graphic intergrowths of feldspar +

quartz or schorl + quartz, if present in thè massive

N W

Portions excavated

previous to 1908

^ISchist and gneiss

Pocket-hearing zone

Normal unproductive

pegmatite

Fig. 4 - Schematic section of thè classic deposit of gem tourmaline at Mount Mica, Maine (from Bastin, 1911).

semblages, thè cavities are surrounded by coarse-

grained and leucocratic assemblages similar to those

encountered in thè facial pods described above.

However, thè presence of graphic textures is much

less frequent here, and thè outer mafie haloes are

virtually absent.

matrix, rarely extend into thè cavity-lining crystals. If

they do, they terminate in aggregates of uniformly

oriented individuate (Fig. 6; cf. Wahlstrom, 1939, Fers¬

man, 1960). Rare as they are on global scale, such

multiple terminations seem to be typical of a given

pegmatite population (e.g., Lazarenko et al, 1973).

8

PETR CERNY

v ^ / > ' .

" / / \ - * » — ' 1 ’

HORNBLENDE

_ NQRIT£ .

N” * GRAPHIC

^* /)* •* A "* * massive

Z0NE

* ■ >; v • • • * • ■ ■;••. \/LCn b; -

POCKET

ZONE

LAYERED

AND

NON-LAYERED

APLITIC

ZONE

DIKE

BOUNDARY

HORNBLENDE

NORITE

Fig. 5 - Vertical section through a typical asymmetric, layered peg-

matite of southern California: thè “line rock” at bottoni is marked

by layers of tourmaline or garnet, and locally also cone-shaped K-

feldspar with graphic quartz (M) which is much more abundant in

thè hanging-wall portion (from Foord, 1976).

Fig. 6 - Euhedral termination of graphic K-feldspar + quartz in-

tergrowth in thè lining of a miarolitic cavity (modified from Fers-

man, 1960).

(ii) In geochemically evolved pegmatites, thè ear-

ly lining of miarolitic cavities is commonly coated by

minerals of rare and volatile components, such as

beryl, phenakite, spessartine, elbaite, topaz, lepidolite,

zinnwaldite, fluorite, danburite, hambergite, fluorap-

atite, montebrasite, hematite, manganocolumbite,

manganotantalite, stibiotantalite, microlite and oth-

ers, including quartz, albite and adularia (Francis,

1987, Zagorskyi and Peretyazhko, 1992, Wise et al,

1994, Foord et al, 1995, Orlandi and Pezzotta, 1996).

In some cases, these minerals form a simple coating

on thè undisturbed faces of early feldspars and

quartz, in others thè substrate is distinctly etched or

corroded.

(iii) The last generation of vug-coating (to vug-fill-

ing) minerals usually consists of boromuscovite,

cookeite, hydroxyl-herderite, roscherite, carbonate-

apatite and other secondary phosphates, micaceous

phases, clay minerals, carbonates, goethite and hol-

landite (e.g., King, 1975, Foord et al, 1986, 1991a). In

some pegmatite populations, these phases form at thè

expense of primary magmatic minerals and coating

phases which suffer extensive leaching, particularly in

and beneath thè bottom parts of thè cavities (Fig. 7;

Lazarenko et al, 1973, Foord, 1982). Beryl and topaz

are found in thè leached rocks underlying thè cavities

at Korosten, Ukraine (Wm.B. Simmons, pers. comm.

1997). Rutile and brookite, zircon, fluorapatite, il-

menite, monazite, cheralite and hematite populate

thè leached, quartz-depleted vug substrate in peg¬

matites of thè Wausau pluton (Falster, 1981, Martin

and Falster, 1986). However, in many cases thè sub¬

strate and surrounding assemblages show no indica-

tion of low-temperature breakdown (Jahns, 1954,

Boggs, 1986, 1992).

Fig. 7 - Miarolitic pegmatite of thè Korosten pluton, Ukraine:

crosses - granite, small L’s - graphic granite, large L’s - coarse

graphic to apographic pegmatite, triple dashes - mica-bearing

coarse-grained to blocky pegmatite, ruled - blocky K-feldspar,

crosshatched - quartz; other symbols - diverse types of mica; dot-

ted area below thè pegmatite indicates thè extent of leaching

(from Lazarenko et al, 1973).

In many miarolitic pegmatites, thè distribution of

zones surrounding thè cavity, thè location of thè vug,

and thè minerals lining plus coating thè cavity are un-

evenly distributed in thè vertical direction. Despite

many exceptions, thè pockets tend to be locateci in

thè upper parts of thè facial pegmatites (Fig. 1), thè

CONSTITUTION. PETROLOGY. AFFILIATIONS AND CATEGORIES OF MIAROLITIC PEGMATITES

9

distribution of feldspars and quartz may be highly

asymmetric (Fig. 7; Lazarenko et al. 1973), and thè

rare-element coating also may be stratified (e.g., Fig.

3 in Stern et al., 1986).

In many pegmatite populations, crystals of

feldspars, quartz, tourmaline, beryl or topaz are bro-

ken off, or shattered and covering thè cavity floor.

Etching and corrosion of these fragments is com-

monly followed by partial regeneration; secondary

growth of fibrous tourmaline on thè broken surfaces,

paralel to thè orientation of thè fragments is particu-

larly widespread (e.g., Foord, 1976), but many cases

also are documented for feldspars, quartz and beryl

(e.g., Feklitchev, 1964).

PROCESSES LEADING TO THE FORMATION OF

MIAROLITIC CAVITIES

Miarolitic vugs form in late stages of pegmatite

consolidation, in consequence of oversaturation of

thè residuai pegmatite melt by volatile components

in generai, and by water in particular. Miarolitic cav-

ities do not form if thè supercritical aqueous fluid, ex-

solved in consequence of thè oversaturation, can

more or less continuously escape from thè pegmatite

into thè host rocks. However, if trapped inside thè

pegmatite, a segregated “bubble” of thè vapor-like

fluid becomes thè site of a vug.

Three principal mechanisms can promote separa-

tion of thè aqueous fluid from thè pegmatite melt: (i)

vapor saturation via progress of isobaric magma so-

lidification, (ii) reduction of confining pressure in thè

ascending pegmatite magma - thè classic concept of

pressure quench, or (iii) stabilization of minerals

which extract melt components facilitating high solu-

bility of water in thè magma, namely B, F, P and Li -

thè Chemical quench of London (1987, 1990).

Progressive crystallization of pegmatite-generating

magma

Incompatible behavior of volatiles in generai, and

of water in particular, leads to vapor saturation and

exsolution as a consequence of graduai build-up of

volatile concentration in progressively crystallizing

(and volumetrically shrinking) pegmatite magma.

Reaction to undecooling and/or gradually decreasing

temperature are thè controlling factors, at more or

less Constant pressure.

This process is omnipresent, as high initial con¬

centration of water (± F, B, Li, P) defines pegmatite

melts and enhances their chances of volatile oversat¬

uration. However, this process may also be rather

slow, proportional to thè rate of pegmatite solidifica-

tion. Thus thè gradually evolving supercritical fluid

can, in most cases, gradually escape from pegmatite

to thè host rocks. That’s why miarolitic cavities are

absent from most granitic pegmatites, or developed

in only negligible numbers and sizes, and why exo-

morphic haloes in wallrocks are very common, gen-

erated by outflow of F,B,C02,Li,K-bearing fluids

from thè consolidating pegmatite bodies.

In contrast to thè above, thè pressure and Chemi¬

cal quenches are distinctly more productive, as far as

formation of miarolitic cavities is concerned.

Reduction of pressure - thè pressure quench

This mechanism corresponds in principle to thè

well-known component of forces driving volcanic ex-

trusions and explosions - exsolution of a gas phase by

release of pressure, which reduces thè solubility of

water in silicate melts. In granite and pegmatite pop¬

ulations, this process affects mainly shallow-level in-

trusions.

Shallow-seated miarolitic pegmatites are typical

of thè NYF family of granite + pegmatite systems, lo-

cally intruded at subvolcanic levels and Consolidated

at pressures as low as 1.5 to 1 kbar. Typical examples

include thè amazonite-bearing pegmatites of thè

Pikes Peak batholith in Colorado, namely thè Crystal

Peak area within thè Lake George intrusive center

and similar occurrences (Foord, 1986); thè Sawtooth

batholith and related intrusions in Idaho (Boggs,

1986, 1992), thè pegmatite-bearing units of thè com-

plex Korosten intrusion in Ukraine (Lazarenko et al.,

1973), and thè Baveno granite in Italy.

The granites are mainly postorogenic to anoro-

genic. A- (to I-) type, subaluminous to metaluminous,

derived by second anatexis of previously depleted

crustal lithologies or from juveniL sources with a

substantial mantle component (cf. Cerny, 1 991 b for

detailed discussion). The pegmatites are dominantly

interior facial pods or crosscutting dikes; they rarely

intrude outside their plutonio parents. The NYF sig¬

nature is prominent in thè accessory minerals of thè

plutonio rocks as well as of thè pegmatites them-

selves: minerals of Li, Cs, Ta, B and P are largely ab¬

sent or very minor, but phases containing Nb, Ti, Zr,

REE, Y, U, Th and F are prominent. Typical rare-ele-

ment minerals of these pegmatites include topaz,

gadolinite, allanite, beryl, helvite, bertrandite, baven-

ite, zircon, thorite, xenotime, monazite, cheralite,

columbite, ixiolite, niobian rutile, anatase, brookite,

pyrochlore, titanite and fluorite (Foord and Martin,

1979, Falster, 1981, Martin and Falster, 1986). In con¬

trast, zinnwaldite, garnet, tourmaline, and cassiterite

are not widespread, although characteristic of some

locai pegmatite populations (Foord, 1982, Foord et

al, 1995, Kile and Foord, 1998). Among thè rock-

forming minerals, amazonite is common in thè vugs,

as is smoky quartz; albite is usually subordinate.

Depletion of thè B, F, P, Li content of thè melt - thè

Chemical quench

Massive precipitation of B,F,P,Li-bearing minerals

commonly occurs in late stages of pegmatite solidifi-

cation: tourmaline, lepidolite, amblygonite, triphylite-

lithiophilite, spodumene or petalite extract thè above

elements which enhance thè solubility of water in thè

melt, and depress its solidus. Considerable quantities

of supercritical aqueous fluid are consequently re-

leased, and rapid solidification of thè residuai magma

is promoted disregarding thè depth of pegmatite em-

placement and thè corresponding regime of pressure.

Classic examples of miarolitic pegmatites of this cat-

egory include thè gem-bearing pegmatites of Maine

(Bastin, 1911, King and Foord, 1994), southern Cali¬

fornia (Jahns and Wright, 1951, Foord, 1976, Jahns,

1979, Shigley et al., 1986), Afghanistan (Rossovskyi,

1981), Pakistan (Laurs et al., 1998) and centrai Trans-

baikalia (Zagorskyi and Peretyazhko, 1992).

10

PETR CERNY

Granite magmas which differentiate residuai

melts generating thè above pegmatites are over-

whelmingly late-orogenic to postorogenic, S- (to

rarely I-) type, peraluminous and leucocratic, gener-

ated by partial^anatexis of undepleted supracrustal

lithologies (cf. Cerny, 1991b).The pegmatites are on-

ly exceptionally found within their parent granites.

They typically form exterior bodies, commonly in thè

outer parts of regionally zoned pegmatite aureoles

which surround their plutonic sources. The LCT sig¬

nature is characteristic of these peraluminous granite

+ pegmatite systems: minerals of Li, Rb, Cs, B, Sn,Ta,

B and P are widespread, whereas those of Ti, Nb,Th,

REE, Y are minor to absent. Typical phases encoun-

tered in thè pegmatites of this category, and particu-

larly in their miarolitic pockets include muscovite,

beryl, spodumene, petalite, lepidolite, elbaite, foitite,

rossmanite, spessartine, pollucite, topaz, apatite, tan-

talite, stibiotantalite, microlite, cassiterite and locally

also hambergite, danburite and axinite (King, 1975,

Francis, 1987, Foord et al, 1991a, Zagorskyi and

Peretyazhko, 1992, Francis et al, 1993, King and Fo¬

ord, 1994, Wise et al, 1994).

CONDITIONS OF CRYSTALLIZATION IN MIAROLITIC

CAVITIES

It seems to be beyond reasonable doubt that thè

early lining (i) of thè miarolitic vugs - thè rock-form-

ing silicates rooted in thè surrounding massive peg¬

matite - crystallized from thè last vestiges of peg¬

matite melt but in thè presence of, and in direct con¬

tact with, thè exsolved supercritical aqueous fluid.

Despite thè occasionally steep gradients in trace-ele-

ment concentrations and dramatic textural changes,

thè bulk minerai compositions and textural patterns

plus fluid inclusions indicate a volatile-saturated melt

as thè parent medium. Depending on thè degree of

fractionation of thè pegmatite melt, thè temperature

of crystallization of thè vug lining may vary from

~550°C for geochemically primitive pegmatites to as

low as ~450°C for highly evolved LCT pegmatites

(e.g., London et al, 1989, London, 1997).

In contrast, thè following generation of wall-coat-

ing rare-element minerals (ii) represents a total com-

positional break from thè “wall-rock” substrate

above. It most probably precipitated from a super¬

critical aqueous fluid (to high-temperature hy-

drothermal solution) at -450 to 200°C (e.g., Cerny

and Chapman, 1984), which was enriched in compo-

nents partitioned out of thè residuai magma. Miner¬

als of Li, (Rb), Cs, Be, Y, REE, Ti, U, Th, Zr, Nb, Ta, B

and F are typically represented.

The last minerai assemblage (iii), covering thè cav-

ity-lining and cavity-coating minerals, is undoubtedly

of low-temperature origin, ca. -250 to 150°C, having

crystallized from hydrothermal solution and/or coex-

isting gas phase. In some cases, this assemblage is ev-

idently materially linked to decomposition of early

minerals, or to leaching of thè adjacent massive peg¬

matite (Fig. 7). Boromuscovite (Foord et al 1991b),

cookeite, apatite, tourmaline, zeolites, carbonates are

occasionally deposited in thè “snow-on-the-roof”

manner; smectites are thè dominant clay minerals,

but may be followed by kaolinite (Foord et al, 1986;

Taylor and Foord, 1993).

The breakage and shattering of thè wall-coating

minerals such as elbaite, beryl and topaz is ascribed

to explosive ruptures of thè vugs (Foord, 1976, Foord

et al., 1991a).The overpressure of thè exsolved aque¬

ous fluid may eventually surpass thè confining pres¬

sure and thè mechanical coherence of thè surround¬

ing pegmatite and wall-rock. The violent exit of thè

aqueous fluid would easily destroy thè fragile crystals

on thè cavity walls (Foord, 1976, Jahns, 1979, Stern et

al, 1986). Moreover, subsequent implosion is also

currently advocated for some cavities in southern

Californian pegmatites: it would have similar physi-

cally destructive effects, and it would explain thè

presence of late minerals containing externally de-

rived components such as Mg, Ti from gabbroic host

rocks (M.C. Taylor, pers. comm. 1997).

Etching, corrosion and leaching of cavity-lining

and wall-coating assemblages, complex changes in

chemistry and structural state of feldspar minerals, as

well as thè regeneration and “healing” of broken

fragments of thè crystals all indicate changes in com-

position and reactivity of thè cavity fluids (Foord and

Martin, 1979, Martin and Falster, 1986, Foord et al.,

1995, Kile and Foord 1998). The medium precipitat-

ing a given minerai assemblage is frequently out of

equilibrium with preceding minerals; its parameters

may change in response to pressure fluctuations, in¬

teraction with wallrocks and pocket lining (e.g., Fo¬

ord et al., 1995, Kile and Foord, 1998). Some minerals

or minerai assemblages indicate an acidic environ-

ment (e.g., albite + muscovite on corroded perthitic

K-feldspar in geochemically primitive vugs, and

bertrandite, helvite-danalite and kaolinite at thè low-

temperature stage). Other phases could only crystal-

lize from alkaline fluids (e.g., boromuscovite, ham¬

bergite, danburite, bavenite, phenakite + K-feldspar,

carbonates, smectites), whereas tourmaline and beryl

require more or less neutral conditions. However, thè

information on thè conditions of crystallization is stili

very fragmented, and studies aimed at deciphering

thè full history of fluid + solids evolution in different

types of cavities are sorely needed.

Geologic classification of miarolitic

PEGMATITES

Historically, miarolitic pegmatites were collective-

ly classified into a single category, disregarding their

paragenetic, geochemical and geological diversity. In

thè relatively recent classification schemes by Rodi-

onov (1964), Rudenko et al. (1975), Ginsburg et al.

(1979) and Ginsburg (1984), miarolitic (or shallow-

level) class of granitic pegmatites was established, en-

compassing all miarolitic pegmatites as low-pressure

phenomena with cavities formed by pressure quench.

This concept was initially adopted by Cerny

(1989), but exception was taken to this ali-inclusive

approach at a later date (Cerny, 1992), once Lon¬

don (1986) documented thè relatively high pres-

sures required to crystallize cavity-hosted spo¬

dumene and associateci phases in thè pegmatites of

southern California and Afghanistan: thè pressures

of 2.8 to 2.4 kbar were found well within thè range

of consolidation in thè rare-element class (Fig. 8).

These findings coincided in time with thè develop-

ment of thè concept of Chemical quench (London,

1987, 1990). This interpretation explained thè for-

mation of miarolitic pockets in medium-pressure

CONSTITUTION, PETROLOGY. AFF1LIATIONS AND CATEGORIES OF MIAROLITIC PEGMATITES

11

regimes, in which thè pressure quench would not

necessarily be effective.

The need to redefine thè miarolitic class became

obvious at this stage, as thè pressure-quenched and

chemically quenched miarolitic pegmatites aparently

required separation: thè NYF-related pressure-

quench pegmatites would be thè only category con-

stituting thè shallow-seated, miarolitic-class peg¬

matites, whereas thè medium-depth, LCT-family peg¬

matites with vugs generated by Chemical quench

would become a mere variety of thè rare-element

class (Fig. 8).

Fig. 8 - P-T fields of environments hosting pegmatite populations

of thè abyssal (AB), muscovite (MS), transitional rare-element +

muscovite (RE-MS), rare-element (RE) and miarolitic (MI) class-

es; arrows indicate regional fractionation trends relative to thè

metamorphic grades of thè host rocks (cf. Cerny, 1991 a for further

details). The miarolitic class as illustrated here is typical of thè

shallow-seated, pressure-quenched NYF-family pegmatites,

whereas thè LCT-family miarolitic pegmatites cover thè P-T con-

ditions of both thè miarolitic and rare-element classes.

However, thè situation is turning out even more

complicated these days, as primary magmatic petalite

was identified in thè southern Califomian pegmatites

(M.C. Taylor and D. London, pers. comm. 1997). It is

quite possible that thè cavity-grown spodumene of

these pegmatites results from low temperature com-

bined with fluid overpressure in thè vugs, thè latter hav-

ing marginally transgressed into thè stability field of

spodumene, but thè pressure regime during thè bulk of

magmatic crystallization was confined to thè petalite

field. The estimate of 1.5 kbar (Simmons et al, 1997) for

magmatic crystallization is probably unrealistic (Wm.B.

Simmons, pers. comm. 1997). However, thè mineralogy

of thè southern Californian pegmatites does not pro¬

vide pressure indicators other than thè Li-aluminosili-

cates; thus thè question of thè pressure regime and its

variations remains to a high degree open.

In contrast, thè generai geologie estimate of 1.5

kbar for thè crystallization of thè very young, ~7 Ma

LCT pegmatites of western Elba (Ruggeri and Lat-

tanzi, 1992) may prove to be rather accurate, as is thè

1.5 to 2 kbar estimate for thè Stak Naia pegmatites in

Pakistan (Laurs et al., 1998). If applicable to Califor¬

nia, pressure quench could have been combined with

Chemical quench in all these Li,B,F-rich pegmatites.

The net result, suggested by thè odd bits of infor-

mation currently available, could be that at least

some of thè LCT miarolitic pegmatites should be re-

tained within thè traditional shallow-seated mi¬

arolitic class, but not necessarily all of them.The final

verdict would have to be based on careful examina-

tion of individuai cases, as fast and easy earmarks do

not seem to be available. Moreover, thè distinction

between thè “shallow-seated, miarolitic-class, LCT-

family pegmatites” and thè “medium-depth, rare-ele-

ment-cìass, LCT-family pegmatites with sporadic mi¬

arolitic cavities” has an excellent potential of becom-

ing blurred by a continuous gradation from one “typ¬

ical category” into thè other.

All of thè above stresses thè need to improve thè

current classifications of granitic pegmatites, also in-

dicated by depth overlap encountered between thè

abyssal and muscovite classes (Cerny, 1991a, 1992),

and by thè poorly subdivided NYF family. The pri¬

mary control of pegmatite classes, thè depth of em-

placement, seems to require reconsideration.

Acknowledgements

This study is based on long-term research sup-

ported by thè Naturai Sciences and Engineering Re¬

search Council of Canada, Canada Department of

Energy, Mines and Resources, Manitoba Department

of Energy and Mines, and Tantalum Mining Corpora¬

tion of Canada, Ltd. Extensive discussions with many

colleagues-pegmatologists significantly contributed

to thè development of ideas formulateci in this paper.

Particular thanks go to thè late E. E. Foord, to D. Lon¬

don, F. Pezzotta and an anonymous reviewer for crit¬

icai reading of thè manuscript and constructive sug-

gestions which led to its improvement.

Reference

Bastin E.S., 1911 - Geology of thè pegmatites and associated

rocks of Maine including feldspar, quartz, mica and gem de-

posits. U. S. Geol. Survey Bulletin, 445, 152 pp.

Boggs R.C., 1986 - Miarolitic cavity and pegmatite mineralogy of

Eocene anorogenic granite plutons in thè northwestern USA.

Internai Minerai. Association, 14th General Meeting Stanford,

Abstracts w. Program, 58.

Boggs R.C., 1992 - A manganese-rich miarolitic pegmatite as-

semblage from thè Sawtooth batholith, south centrai Idaho,

U.S.A. Abstracts of Papers, Lepidolite 200 Symposium, Nové

Mesto na Morave, Czechoslovakia: 15-16.

Cerny R, 1972 - The Tanco pegmatite at Bernic Lake, Manitoba.

Vili. Secondary minerals from thè spodumene-rich zone.

Canad. Mineralogist, 11: 714-726.

Cerny P., 1989 - Characteristics of pegmatite deposits of tanta¬

lum. In: Moller P., Cerny P., Saupé F. eds. Lanthanides, Tanta¬

lum and Niobium, Springer-Verlag : 271-299.

Cerny P., 1991a - Rare-element granitic pegmatites. Part I: Anato-

my and internai evolution of pegmatite deposits. Geoscience

Canada, 18: 49-67.

Cerny P., 1991b - Fertile granites in Precambrian rare-element

pegmatite fields: is geochemistry controlled by tectonic setting

or source lithologies? Precambrian Research, 51: 429-468.

Cerny P., 1992 - Geochemical and petrogenetic features of min-

eralization in rare-element granitic pegmatites in thè light of

current research. Applied Geochemistry, 7: 393-416.

Cerny P. & Chapman R., 1984 - Paragenesis, chemistry and struc-

tural state of adularia from granitic pegmatites. Bull.

Minéralogie, 107: 363-384.

Cerny R, Trueman D.L., Ziehlke D.V., Goad B.E. & Paul B.J.,

1981 - The Cat Lake - Winnipeg River and thè Wekusko Lake

pegmatite fields, Manitoba. Manitoba Dept. Energy and

Mines, Minerai Resources Div., Economie Geology Report,

ER80-1, 234 pp.

12

PETR CERNY

Falster A.U., 1981 - Minerals of thè Wausau pluton. Minerai.

Record, 12: 93-97.

Feklitchev V.G., 1964 - Beryl. morphology, composition and

structure of crystals. Nauka, Moscow, 98 pp.

Fersman A.E., 1960 - Graphic structure of granitic pegmatites. In:

Selected Works VI. Academy of Sciences of thè USSR,

Moscow. 645-659 (in Russian).

Foord E.E., 1976 - Mineralogy and petrogenesis of layered peg-

matite-aplite dikes in thè Mesa Grande district. San Diego

County, California. Unpubl. PhD diesis, Stanford University,

326 pp.

Foord, E. E., 1977 - The Himalaya dike System, Mesa Grande Dis¬

trict, San Diego County, California. Minerai. Record, 8: 461-

474.

Foord E. E., 1982 - Amazonite-bearing pegmatites of thè Lake

George intrusive center. In: Cerny R et al. Granitic Pegmatites

of thè Black Hills, South Dakota and Front Range, Colorado.

Geol. Assoc. Canada - Minerai. Assoc. Canada Field Trip

Guidebookj 12:51-56.

Foord E. E., Cerny P., Jackson L.L., Sherman D.M. & Eby

R.K., 1995 - Mineralogical and geochemical evolution of mi-

cas from miarolitic pegmatites of thè anorogenic Pikes Peak

batholith. Colorado. Mineralogy and Petrology, 55: 1-26.

Foord E. E., London D., Kampf A.R., Shigley J.E. & Snee L.W.,

1991a - Gem-bearing pegmatites of San Diego County, Cali¬

fornia. Geol. Soc. America Ann. Meeting San Diego, Field Trip

Guide, 9, 22 pp.

Foord E. E. & Martin R.F., 1979 - Amazonite from thè Pikes

Peak batholith. Mineralogical Record, 10: 373-384.

Foord E. E., Martin R.F., Fitzpatrick J.J., Taggart J.E., Jr. &

Crock J.G., 1991b - Boromuscovite, a new member of thè mi¬

ca group, from thè Little Three Mine pegmatite, Ramona dis¬

trict, San Diego County, California. Amer. Mineralogist, 76:

1998-2002.

Foord E. E., Spaulding L.B.Jr., Mason R.A. & Martin R.F.,

1989 - Mineralogy and paragenesis of thè Little Three peg¬

matites, Ramona District, San Diego County, California. Min¬

erai. Record, 20: 101-127.

Foord E. E., Starkey H.C. & Taggart J.E. Jr., 1986 - Mineralo¬

gy and paragenesis of “pocket” clays and associated minerals

in complex granitic pegmatites, San Diego County, California.

Amer. Mineralogist, 71:428-439.

Francis C.A., 1987 - Minerals of thè Topsham, Maine pegmatite

district. Rocks and Minerals, 62: 407-415.

Francis C.A., Wise M.A., Kampf A.R., Brown C.C. & Whit-

more R.W., 1993 - Granitic pegmatites in northern New Eng-

land. In: Cheney J.T. & Elepburn J.C., ed’s. Field Trip Guide-

book for thè northeastern United States; GSA Meeting 1993

Boston, 1: E1-E24.

Ginsburg A.I., 1984 - The geological conditions of thè location

and thè formation of granitic pegmatites. Proc. 27th Internai

Geol. Congress, 15: 245-260.

Ginsburg A.I.,Timofeyev I.N. & Feldman L.G., 1979 - Principles

of geology of granitic pegmatites. Nedra, Moscow, 296 pp. (in

Russian).

Jahns R.H., 1954 - Pegmatites of southern California. Geology of

southern California, Calif Div. Mines Bull., 170 (7): 37-50.

Jahns R.H., 1979 - Gem-bearing pegmatites in San Diego Coun¬

ty, California: thè Stewart mine, Pala District, and thè Hi¬

malaya mine, Mesa Grande District. In: Abbot P.L. & Todd

V.R., ed’s. Mesozoic crystalline rocks: Peninsular Range

batholith and pegmatites, Point Sai ophiolite. Dept. Geol. Sci¬

ences, San Diego State Univ., 1-38.

Jahns R.H. & Wright L.A., 1951 - Gem- and lithium-bearing

pegmatites of thè Pala district. San Diego county, California.

Calif. Div. Mines Spec. Rept., 7-A, 72 pp.

Kile D.E. & Foord E. E., 1998 - Micas from thè Pikes Peak

Batholith and its cogenetic granitic pegmatites: optical prop-

erties, composition, and correlation with pegmatite evolution.

Canad. Mineralogist, 36: 463-482.

King V.T., 1975 - Newry, Maine: a pegmatite phosphate locality.

Minerai Record, 6: 189-204.

King V.T. & Foord E. E., 1994 - Mineralogy of Maine, Volume 1:

Descriptive mineralogy. Maine Geol. Survey, Dept. Conserva-

don, 418 pp.

Landes K.K., 1925 - The paragenesis of thè granite pegmatites of

centrai Maine. Amer. Mineralogist, 10: 374-411.

Laurs B.M., Dilles J.H., Wairrach Y., Kausar A.B. & Snee

L.W., 1998 - Geological setting and petrogenesis of symmetri-

eally zoned, miarolitic granitic pegmatites at Stak Naia, Nan-

ga Parbat - Haramosh Massif, northern Pakistan. Canad. Min¬

eralogist, 36: 1-47.

Lazarenko E.K., Pavlishin V.I., Latysh V.T. & Sorokin Yu.G.,

1973 - Mineralogy and genesis of thè chamber pegmatites of

Volynia. Lvov State Univ. Pubi House Lvov, 359 pp. (in Russ¬

ian).

London D., 1986 - Formation of tourmaline-rich gem pockets in

miarolitic pegmatites. Amer. Mineralogist, 71: 396-405.

London D., 1987 - Internai differentiation of rare-element peg¬

matites: effects of boron, phosphorus and fluorine. Geochim.

Cosmochim. Acta, 51: 403-420.

London D., 1990 - Internai differentiation of rare-element peg¬

matites; a synthesis of recent research. Geol. Soc. America

Spec. Paper, 246: 35-50.

London D., 1997 - Volatile abundance and thè formation of mi¬

arolitic cavities in evolved silicic magmas. Abstracts, lst Inter-

nat. Workshop “Petrology, Rare Minerals and Gemstones of

Shallow-Depth Pegmatites”. Museo Civico di Storia Naturale

di Milano , 23.

London D., Morgan G.B. & Hervig R.L., 1989 - Vapor-under-

saturated experiments Macusani glass + H20 at 200 MPa, and

thè internai differentiation of granitic pegmatites. Contrib.

Minerai Petrol., 102: 1-17.

Lyckberg P. & Rosskov A., 1997 - Prediction of gem pockets in

shallow depth pegmatites in thè Alabashka area, Mursinka in-

trusion, Ural Mountains, Russia. Abstracts, lst Internat. Work¬

shop “Petrology, Rare Minerals and Gemstones of Shallow-

Depth Pegmatites”. Museo Civico di Storia Naturale di Mi¬

lano, 20.

Martin R.F. & Falster A.U., 1986 - Proterozoic sanidine in peg¬

matite, Wausau complex, Wisconsin. Canad. Mineralogist, 26:

709-716.

Némec D., 1992 - Pegmatites with druse cavities in thè west-Mora-

vian crystalline terrain. Acta Sci. Nat. Musei Moraviae Occi-

dent., Trebic, 18: 13-23.

Orlandi P & Pezzotta E, 1996 - Minerali dell’isola d’Elba. Edi¬

zioni Novecento Grafico, Bergamo, 245 pp.

Rodionov G., 1964 - Pegmatite types and some specific features

of pegmatite formation. Internat. Geol. Congress, 22nd sess. In¬

dia, Rept. pt. VI: 140-156.

Rossovskyi L.N., 1981 - Rare-element pegmatites with precious

stones and conditions of their formation. Internai Geol. Re-

view, 23: 1312-1320.

Rudenko S.A., Romanov V.A., Morakhovskyi V.N., Tarasov

E.B., Galkin G.A. & Dorokhin V.K., 1975 - Conditions of

formation and Controls of distribution of muscovite objects of

thè north-Baikal muscovite province, and some generai prob-

lems of pegmatite consolidation. In: Gordiyenko V.V. ed..

Muscovite pegmatites of thè USSR. Nauka, Leningrad: 174-

182 (in Russian).

Ruggeri G. & Lattanzi L., 1992 - Fluid inclusion studies on Mte.

Capanne pegmatites. Isola d'Elba, Italy. Eur. Jour. Mineralogy,

4: 1085-1096.

Shigley J.E., Kampf A.R., Foord E. E. & London D., 1986 - Gem

pegmatites of southern California. 14th Gen. Meeting, Internai

Minerai Assoc. Stanford, Field Trip Guidebook, 39 pp.

Simmons Wm.B. & Heinrich E.Wm., 1980 - Rare-earthpegmatites

of thè South Piatte district. Colorado. Colorado Geol. Survey,

Resource Ser., 11, 131 pp.

Simmons Wm.B., Lee M.T. & Brewster R.H., 1987 - Geochem-

istry and evolution of thè South Piatte granite-pegmatite Sys¬

tem, Jefferson County, Colorado. Geochim. Cosmochim. Acta,

51:455-471.

Simmons Wm.B., Webber K.L. & Falster A.U., 1997 - Cooling

rates and crystallization dynamics of pegmatites from San

Diego County, California, USA. Abstracts, lst Internat. Work¬

shop “Petrology, Rare Minerals and Gemstones of Shallow-

Seated Pegmatites” Museo Civico di Storia Naturale di Mi¬

lano, 30.

Stern L. A., Brown G.E. Jr., Bird D.K., Jahns R.H., Foord E.E.,

Shigley J.E. & Spaulding L.B. Jr, 1986 - Mineralogy and geo¬

chemical evolution of thè Little Three pegmatite-aplite lay¬

ered intrusive, Ramona, California. Amer. Mineralogist 71:

406-427.

Wahlstrom E. E., 1939 - Graphic granite. Amer. Mineralogist 24:

681-698.

Wise M.A., Rose T.R. & Holden R.E. Jr., 1994 - Mineralogy of

thè Bennet pegmatite, Oxford County, Maine. Minerai

Record, 25: 175-184.

Zagorskyi V.E. & Peretyazhko I.S., 1992 - Gem pegmatites of

centrai Transbaikalia. Nauka, Novosibirsk, 222 pp.

Memorie della Società Italiana di Scienze Naturali e del Museo Civico di Storia Naturale di Milano

Volume XXX - pp.: 13-28, 2000

Pegmatites and Pegmatite Minerals of thè Wausau Complex,

Marathon County, Wisconsin

Alexander U. Falster*1, Wm. B. Simmons1,

Karen L. Webber1 & Tom Buchholz2

' Department of Geology and Geophysics University of New Orleans - New Orleans, Louisiana 70148

2 1140 12th Street North, Wisconsin Rapids, Wisconsin 54494

Abstract - The Wausau Complex is a Proterozoic (1.52-1.48 Ga) shallow-level, anorogenic intrusive complex, consisting

of four distinct intrusive centers: thè Stettin, thè Wausau, thè Rib Mountain, and thè Nine Mile plutons. Of thè four, thè

syenitic Stettin pluton is thè oldest and most alkalic. The pluton is roughly concentrically zoned with two nepheline syenite

rings, one as thè outermost unit, thè other as a core margin around a pyroxene syenite core. The three other plutons, thè

Wausau, Rib Mountain, and Nine Mile, are progressively more silica-rich. The Nine Mile pluton consists of alkali granite

and monzonite. Miarolitic pegmatites are found in all four intrusive centers. Pegmatites in thè Stettin pluton range from

nepheline-bearing to pyroxene- and amphibole-bearing types. Minor minerals include magnetite, zircon, apatite, ilmenite,

xenotime, cheralite, and titanite. Rare accessories include eudialyte, catapleiite, fluorite, pyrochlore, allanite, thorite, and

thorogummite. Pegmatites in thè Nine Mile pluton are characterized by feldspars, quartz and minor biotite (siderophyllite).

Minor minerals include magnetite, hematite, titanium oxide minerals, and siderite. Rare minerals include phenakite, bertran-

dite, cheralite, xenotime, zircon, and pyrite. Significantly, columbite-tantalite group minerals are virtually absent. Pegmatites

of thè Wausau and Rib Mountain plutons have characteristics between those of thè Stettin and Nine Mile plutons.

Within thè entire Wausau Complex thè bulk B content is negligible. However, thè bulk Li content is relatively higher

than B due to thè Li content in amphiboles and micas (Li20 contents in excess of one percent).Typically, such anorogenic

pegmatites have been classified as NYF. However, in thè Wausau Complex, thè pegmatites are generally poor in Nb and F,

but enriched in Fe, Ti, LREE, HREE, and to a lesser extent in Be and Li.

Riassunto - Il Wausau Complex è un complesso intrusivo anorogenico di età proterozoica (1.52-1.48 Ga) di bassa pro¬

fondità, costituito da quattro distinti centri intrusivi: i plutoni denominati Stettin, Wausau, Rib Mountain e Nine Mile. Dei

quattro, il plutone sienitico Stettin è il più antico ed alcalino. Tale plutone è grossomodo concentrico con due anelli sieniti-

ci nefelinitici, dei quali uno costituisce l’unità più periferica e l’altro costituisce un anello attorno ad un nucleo di sienite

pirossenitica. Gli altri tre plutoni, il Wausau, il Rib Mountain ed il Nine Mile, sono progressivamente più ricchi in silice. Il

plutone Nine Mile è costituito da un alkali-granito e da monzonite. Pegmatiti miarolitiche sono state trovate in tutti e quat¬

tro le masse intrusive. Le pegmatiti nel plutone Stettin vanno da tipi a nefelina a tipi a pirosseno ed anfibolo. Tra i minerali

accessori vi sono magnetite, zircone, apatite, ilmenite, xenotime, cheralite e titanite. Tra gli accessori rari vi sono eudialyte,

catapleiite, fluorite, pirocloro, allanite, thorite e thorogummite. Le pegmatiti nel plutone Nine Mile sono caratterizzate da

feldspati, quarzo ed in maniera minore biotite (siderofillite).Tra i minerali accessori vi sono magnetite, ematite, minerali del

gruppo degli ossidi di titanio e siderite. Accessori rari sono fenacite, bertrandite, cheralite, xenotime, zircone e pirite. E’ da

notare che minerali del gruppo della columbite-tantalite sono pressoché assenti. Le pegmatiti dei plutoni Wausau e Rib

Mountain hanno caratteristiche intermedie tra quelle dei plutoni Stettin e Nine Mile.

Nell’intero Wausau Complex il contenuto totale di B è trascurabile. Tuttavia, il contenuto totale di Li è relativamente

alto in confronto a quello del B, a causa del contenuto di Li di anfiboli e miche (contenuti in Li20 oltre l’uno per cento).

Come tipico, queste pegmatiti anorogeniche sono state classificate come NYF, Tuttavia, nel Wausau Complex, le pegmatiti

sono generalmente povere in Nb e F, ma arricchite in Fe, Ti, LREE, HREE, ed in modo minore in Be e Li.

Keywords: Wausau Complex, Stettin pluton, Wausau pluton, Rib Mountain pluton, Nine Mile pluton, Wisconsin, peg¬

matite, NYF.

This work is dedicated to thè memories of Ursula Falster, Walter Prey, and Marie Prey for their support and guidance

in thè past and for being more than mother and grandparents, for being my friends.

* E-mail: afalster@uno.edu

14

ALEXANDER U. FALSTER - WM. B. SIMMONS - KAREN L. WEBBER & TOM BUCHHOLZ

Introduction

The Wausau Complex is a composite alkalic intru¬

sive exposed in Marathon County, Wisconsin (Fig. 1).

Zircons from thè complex yield a U-Pb age of about

1.48-1.52 Ga (Van Schmus, 1975 a and b). The com¬

plex consists of four intrusive centers which are from

oldest to youngest: thè concentric Stettin pluton; thè

Wausau pluton; thè Rib Mountain pluton; and thè

Nine Mile pluton (Myers et al. , 1984). Pegmatites and

pegmatitic segregations occur throughout thè

Wausau Complex (Falster, 1981, 1986, 1987). The

complex is rift-related and bears a distinct alkalic sig¬

nature. The Stettin pluton is thè most alkalic of thè

four intrusive centers (Myers et al., 1984). The

Wausau pluton consists of pyroxene and amphibole

syenites (Sood, 1980). Only thè northern segment of

thè Wausau pluton is now exposed, thè southern part

is obscured by thè younger quartz-syenitic Rib

Mountain pluton. The Nine Mile pluton intrudes and

partially obscures thè Rib Mountain pluton and con¬

sists primarily of alkali granite and quartz monzonite

(Myers et al., 1984). Xenoliths are common in thè

Wausau Complex and consist dominantly of

quartzites, schists, and amphibolites.

Pegmatites of thè miarolitic type are abundant in

thè Wausau Complex, indicating shallow levels of em-

placement (Falster, 1981, 1986, 1987). Most peg¬

matites are of modest size (less than 100 m in maxi¬

mum dimension) and occur within thè complex itself.

Only in a few cases do pegmatites intrude thè sur-

rounding country rock. Pegmatite mineralogies are

distinct within each of thè four intrusive centers.

Methods

Instrumental Neutron Activation Analysis

INAA was performed by thè Phoenix Memorial

Laboratories, University of Michigan, Ann Arbor.

Whole rock samples were pulverized in a Spex ball

mill using a tungsten Carbide container for 30 minutes

after a contamination run of thè same material for 10

minutes. Minerai separates were crushed to a particle

size of less than 1 mm in diameter. All samples were

fused into quartz tubes.

AMRAY 1820 digitai scanning electron microscope

Samples were cleaned for 10 minutes in an ultra-

sonic cleaner, rinsed with distilled water and dried

overnight in a desiccator. Samples were coated with

250 Angstroms of carbon under a vacuum of 1 x IO5

torr.

An acceleration voltage of 15 to 25 kV was used

for energy dispersive spectral analysis. The working

distance was generally kept at 18 mm and a 300 mi¬

cron final aperture was used. Sample tilt was kept at

0 degrees for backscattered electron imaging and at

45 degrees for EDS analyses.

Photomicrography was performed using T-Max

100 ASA 4” x 5” sheet film. Exposure times were kept

at 64 seconds. Films and prints were developed ac-

cording to standard procedures.

X-ray diffraction

A SCINTAG XDS 2000 X-ray diffractometer was

used and operated at 35 kV potential and 15 mA cur-

rent. Samples were prepared according to thè powder

method and run at a scan range of 2 to 70 ° 2 theta

and a scan rate of 0.5 0 per minute.

Stettin pluton pegmatites

In thè Stettin pluton thè following pegmatite

types are found:

Nepheline-bearing pegmatites (NP) are restricted

to thè nepheline syenites. They are found in thè out-

er or inner ring nepheline syenite around thè pyrox¬

ene syenite core. Pegmatitic nepheline syenite con¬

sists of grains up to 30 centimeters in maximum di¬

mension.

Amphibole- and pyroxene-bearing pegmatites

(AP) are abundant in thè amphibole syenites. They

are well zoned and have a complex mineralogy. Struc-

turally, they consist of a wall zone, an intermediate

zone, and a core zone. Aplite units are abundant and

may constitute up to 35% of thè pegmatite. The wall

zone typically consists of microcline, large crystals of

arfvedsonite, which are mantled by aegirine-augite,

and lesser amounts of sodic plagioclase. The interme¬

diate zone consists of microcline and abundant radi-

ating sprays of aegirine-augite. This zone is typically

vuggy with individuai vugs up to several centimeters

in maximum dimension. In some areas, abundant tiny

quartz crystals line thè vugs. Accessory minerals

found in thè intermediate zone include bario-cerio-

pyrochlore, fluorite, zircon, cheralite, thorite,

thorogummite, fergusonite, allanite-(Ce), apatite,

columbite-tantalite, goethite (fine crystals) and sever¬

al unidentified species. In some cases, larger mi¬

arolitic cavities have been found in thè intermediate

zone. These larger cavities typically contain large

pseudomorphs of goethite after siderite, microcline

with epitaxial sodic plagioclase, minor ilmenite, zir¬

con, bario-cerio-pyrochlore and late quartz crystals

as drusy overgrowths. Pink aplite is prominent in thè

intermediate zone. Aplite appears to have invaded

thè intermediate zone. Cataclastic texture of crystal

fragments of arfvedsonite, microcline, and aegirine-

augite attest to moderately violent intrusion of thè

aplite. The aplite consists dominantly of microcline.

An interesting accessory minerai in thè aplite is near

Fig. 1 - Generalized map of thè Wausau Complex (after Myers et al., 1984), showing thè location of thè Stettin pluton, thè Wausau pluton,

thè Rib Mountain pluton, and thè Nine Mile pluton. E-W distance: 24 km. Symbols used are as follows: a = alluvium, go = glacial outwash,

gt = glacial tilt, db = diabase dike, qp = quartz monzonite porphyry, ga = granite aplite, ng = Nine Mile granite and quartz monzonite, qs =

quartz syenite, as = amphibole syenite, ps = pyroxene syenite, sv = syenitized volcanics, sa = syenite aplite, ls = hybrid lensoidal syenite, ns

= nepheline syenite. ts = tabular syenite, lg = leucocratic granite, qm = quartz monzonite, qd = quartz diorite and diorite, vs = volcanogenic

sedimentary rocks, fv = felsic volcanic rocks, iv = intermediate volcanic rocks, mv = mafie volcanic rocks, q = quartzite, bs = biotite schist,

am = amphibolite.

PEGMATITES AND PEGMATITE MINERALS OF THE WAUSAU COMPLEX

15

16

ALEXANDER U. FALSTER - WM. B. SIMMONS - KAREN L. WEBBER & TOM BUCHHOLZ

end-member bario-pyrochlore. The core zone con-

sists in all cases of a quartz lens. Some pegmatites

have a quartz-zircon lens, containing up to 20 % zir-

con. Minor aegirine-augite may surround thè core

margin.

Some pegmatites contain pods of quartz and zircon

in thè intermediate zone near thè core margins. These

pods are generally finer grained (millimeter range)

than thè core zone but may contain small gemmy zir-

cons of yellow, red, or brown color.

Feldspathic pegmatites (FP). This type of peg-

matite occurs commonly in thè pyroxene syenite and

is typically simple, consisting largely of feldspars with

minor amounts of fluorite and Fe-oxides. The peg¬

matites are frequently vuggy with thè inner zones re-

sembling a sponge. Pyroxene and amphibole are sub¬

ordinate or absent and quartz does not occur in thè

pegmatites in thè pyroxene syenites. The accessory

minerals associateti with thè pegmatites of thè am¬

phibole syenites are generally absent. Some peg¬

matites in thè pyroxene syenite consist dominantly of

large grains of anorthoclase up to 40 centimeters in

diameter along with minor pyroxene and occasional-

ly amphibole. The anorthoclase exhibits a delicate

blue schiller-effect. Material of this type has been

mined and sold as “Wisconsin moonstone”.

Wausau pluton pegmatites

Initially, thè three southern plutons, thè Wausau

pluton, thè Rib Mountain pluton, and thè Nine Mile

pluton, were lumped together as thè Wausau pluton

(Falster, 1981). Subsequent field mapping conducted

by Gene LaBerge and Paul Myers revealed separate

intrusive centers in what was then known as thè

Wausau pluton. As a result of this work, thè nomen¬

clature was revised and thè three newly defined in¬

trusive centers were named thè Wausau pluton, thè

Rib Mountain pluton, and thè Nine Mile pluton (My¬

ers et ai, 1984) (Fig. 1).

The quartz-cored feldspathic pegmatites (QWP)

in thè Wausau pluton resemble those in thè Stettin

pluton but are generally more silica-rich. They are

zoned with a feldspathic wall zone and a quartz core

in thè simple pegmatites, but may show an intermedi¬

ate zone with coarse quartz and feldspar. Miarolitic

cavities generally occur just below thè quartz core.

The size of these cavities may reach several decime-

ters in maximum dimension. Sodic amphibole is com¬

monly seen lining fracture planes across these peg¬

matites. One unusual pegmatite in thè Wausau pluton

partially invaded thè surrounding metavolcanic

country rock. This pegmatite reaches over 10 m in

maximum dimension and has a footwall portion com-

prised of major potassium feldspar, plagioclase,

quartz, and micas along with wollastonite, miserite,

agrellite, and titanite as accessory minerals (Medaris

and Guggenheim, 1983).

Rib mountain pluton pegmatites

Quartz-cored, biotite-bearing pegmatites (QRP)

in thè Rib Mountain pluton are generally zoned with

a quartz core and a wall zone of quartz and feldspars

with elongate, lath-shaped biotite. In some, an inter¬

mediate zone of graphic quartz and potassium

feldspar is seen. Books of biotite are commonly

found in thè core margin. No pyroxenes or amphi-

boles are found in thè pegmatites of thè Rib Moun¬

tain pluton.

Nine mile pluton pegmatites

Pegmatites in thè Nine Mile pluton are thè best

studied in thè complex owing to thè extensive mining

of thè decomposed syenites and alkali granites of thè

Wausau Complex for road gravel. Over 950 peg¬

matites have been studied in thè Nine Mile pluton

and they have been broken down into four major

groups.

Group 1: Simple pegmatitic pods (PP) comprise

this group (Fig. 2, #1). The size of these bodies reach¬

es at most 3 meters in length. Contacts with thè sur¬

rounding alkali granite are typically gradational. Peg¬

matitic pods are generally simple in structure with a

wall zone consisting of feldspars, quartz, and biotite,

and a core of monomineralic quartz. Miarolitic cavi¬

ties typically occur directly below thè core. The cavi¬

ties extend for a distance of half thè pegmatite width.

The mineralogy of these pegmatitic pods consists pri-

marily of quartz and feldspars with minor goethite re-

placing siderite rhombs. Accessory minerals include

hematite, phenakite, bertrandite, fluorite, apatite,

cheralite, and Ti-oxides (generally anatase).

Group 2: Simple, vuggy pegmatites (VP). The peg¬

matites in this group are commonly large but zona-

tion is very simple and zone boundaries are grada¬

tional (Fig. 2, #2). A typical wall zone and an inner

zone comprised of vuggy potassium feldspar and

quartz are thè dominant zones. Rarely, scattered

coarse patches of white quartz resembling a quartz

core occur. Small cavities are filled with well-crystal-

lized feldspars and quartz (at times white). The most

characteristic accessory minerai in these cavities is

hematite, which is generally very abundant. Less

common are siderite and pyrite, both of which can be

replaced by goethite or hematite. The average size of

thè vugs ranges from about 0.5 to 1.5 centimeters.

Group 3: The zoned pegmatites (ZP) of group 3

are some of thè largest pegmatites in thè Nine Mile

pluton (Fig. 2, #3), exceeding 100 meters or more.

Contacts between thè pegmatites and thè host rock

are generally sharp, with a visible thin chili margin in

many instances. The wall zone is generally extensive

and consists of quartz, feldspars, and elongate biotite.

In some group 3 pegmatites, thè footwall wall zone is

distinctly thicker than thè hanging wall wall zone. The

intermediate zones typically consist of graphic quartz

and feldspar or almost monomineralic potassium

feldspar. The core margin contains large book bi-

otites. The core is monomineralic quartz. Aplites oc¬

cur as cross cutting units or as footwall bodies, which

may make up thè entire footwall portion of thè peg¬

matite. The footwall portion of thè pegmatites is typ¬

ically more sodic and generally finer grained than thè

more potassic hanging wall. Miarolitic cavities, which

may exceed 4 m in maximum dimension, exist either

below thè quartz core, within thè graphic portions of

thè intermediate zone or below monomineralic

feldspar masses. The common rock-forming minerals

quartz and feldspar (potassium feldspars are domi-

PEGMATITES AND PEG MATITE MINERALS OF THE WAUSAU COMPLEX

17

*_x x

/ * X. X X

X * < X- x X x x - •*■

- x. y v > * * * •* * *■•**■ * ■* _

- x * x x * x * x * x X v y *

*7L~£- — * ~r~? r / / / i ✓ . , / ^ \ -

\

. . &-/C/

/CXS^T

\

.1 I

— I — ' ^

/ \y — / - I ^ " \ " " /

^ N / s _W \ —

V ’ " ‘ /A ^ )-, -

\ / \ \ X

=f

T

\ rr^7 ,. . . -

X V / \ / ^ ^ / \ 1 / \ / \ s ,

i <11 > ' *

' / /

\

- s — - x x x * x. * x j<~x~z~ir*

2T7 • X X X A x X X *

. -. J . I ■ ^••X* • X xx xxxx,

x _ ^ v *■ x x

X * X K

y x x x x

^ X X X^X X X X

’ *”•*.' '‘/y'V XX

x x

X X X V X , \ / '

At 1

/ K /\ /C/ ',<} i

/

!

Cxxxp/ , \ \ / _ _

"CA~ /WN

_'/\/

■

X X X

xxx

Explanation of Symbols

Api ite contact zone

Wall Zone

Quartz Core

Coarse quartz- fel dspar

Coarse feldspar

Pocket (or aplite unit)

So 1 ution-etched region

Fig. 2 - Simplified cross sections of pegmatites in thè Nine Mile pluton, including group 1 - simple pegmatite pods (PP),

group 2 - simple vuggy pegmatites (VP), group 3 - zoned pegmatites (ZP), and group 4 - solution etched pegmatites (SEP).

nant) make up thè bulk of thè pegmatite. They occur

with thè following diverse assemblage of accessory

cavity minerals: siderite (typically replaced by

goethite, lepidocrocite, or hematite), phenakite,

bertrandite, hematite, ilmenite, anatase, rutile,

brookite, zircon, pyrite (typically replaced by

goethite or hematite, at times associated with

jarosite), cheralite, apatite, fluorite, and beryl. Near

thè final stages of pocket formation, violent pocket

rupture was evidently a common phenomenon in thè

pegmatites, as fractured pocket constituents and fine-

grained precipitates of structurally disordered potas-

sium feldspar can be found in many miarolitic cavi-

ties. It appears that a build-up of internai pressure

over confining pressure with associated rupture and

rapid decompression was a commonplace event in

thè shallowly emplaced pegmatites of thè Wausau

Complex. This lead to thè rapid crystallization of pri-

mary high sanidine and quartz. In some cases, thè at-

tendant cooling prevented any further structural

change so that thè structural state is highly disor¬

dered. The disordered sanidine did not crystallize at

high temperature as it loosely covers pre-existing mi¬

crocline, but instead crystallized rapidly at low tem¬

perature. Thus, thè persistence of high sanidine in

Proterozoic pegmatite pockets is an exceedingly rare

event (Martin and Falster, 1986). All pegmatites in

thè Wausau Complex evidently exsolved a free aque-

ous fluid phase during thè latter stages of pegmatite

formation. However, thè volume of this aqueous flu¬

id was generally small as evidenced by thè small vol¬

ume of miarolitic cavities in thè pegmatites as a

whole. Cavities are estimated to contribute less than

0.1 % to thè pegmatite volume.

Group 4: Solution-etched pegmatites (SEP). Min-

eralogically and structurally, these pegmatites are

similar to group 3 but show extensive solution etch-

ing of thè quartz/feldspar footwall immediately be-

low miarolitic cavities (Fig. 2, #4). The large miarolitic

cavities typically show little evidence of pocket rup¬

ture and rapid degassing. Instead, in a zone directly

below thè pocket floor, thè quartz-feldspar footwall is

severely etched and quartz is either completely or

partially removed. The areas closest to thè pockets

are generally free of quartz. Quartz becomes more

abundant as distance from thè pocket floor increases.

In thè voids left behind after quartz had been re¬

moved, an assemblage of minerals not unlike an

18

ALEXANDER U. FALSTER - WM. B. SIMMONS - KAREN L. WEBBER & TOM BUCHHOLZ

Alpine cleft assemblale was left behind. The com¬

mon association consists of potassium feldspar of

simple, adularia habit, albite, siderite replaced by

goethite, muscovite, hematite or ilmenite, anatase,

brookite, rutile, zircon, xenotime, monazite, cheralite,

almandine, pyrite, and rarely bertrandite.

Location of wausau complex pegmatites

Pegmatites in thè Wausau Complex tend to occur

in swarms within non-pegmatitic host rocks. The peg¬

matites dip most steeply near thè country rock con-

tacts and are almost horizontal in thè interior of thè

complex. The attitude of thè various pegmatites sug-

gests that only minor portions of thè cupola of thè

complex has been eroded. The various types of peg¬

matites in thè Nine Mile pluton, thè best studied of

thè intrusions, tend to form clusters of one dominant

type.

MlAROLITIC CAVITIES IN THE NINE MILE PLUTON

Primary (those miarolitic cavities which do not ap-

pear to have formed after another minerai phase has

been dissolved from massive pegmatite) miarolitic

cavities in thè group 3 pegmatites (Fig. 2, #3) are typi-

cally flat-bottomed with a dome-shaped pocket roof.

The footwall is generally more plagioclase-rich and

finer grained than thè hanging wall, which is more

potassic. Pocket rubble, consisting of fragments and

cleavage pieces of thè common pocket minerals, is

sometimes cemented by a secondary overgrowth of

quartz. Some miarolitic cavities contain several layers

of platelets of feldspars and quartz. This very unusual

collapsed house-of-cards texture may be thè result of

thè formation of crystallized platelets forming at thè

boundary of two immiscible fluids, maybe one aque-

ous, thè other CO.-rich.The platelets eventually settled

to thè pocket floor, not unlike rafts of halite forming in

saline lakes. Secondary miarolitic cavities are restrict-

ed to a halo around thè footwall portions of larger pri¬

mary miarolitic cavities (Fig. 2, #4). Characteristically,

they are found in thè spaces formed by thè dissolution

of quartz from quartz-feldspar intergrowths just below

thè pocket floor. With increasing distance from thè

pocket floor, quartz remnants become more common.

Typical associations in secondary miarolitic cavities in¬

clude anatase (thè most common Ti oxide), brookite,

rutile (thè rarest Ti oxide in these cavities), secondary

quartz, feldspars of simple morphology, ilmenite,

hematite, goethite, zircon, cheralite, monazite, xeno¬

time, and other minerai species.

Pegmatite mineralogy

A brief description of all minerai species identi-

fied to date in thè Wausau Complex is discussed be¬

low and summarized in Table 1.

Quartz

Quartz occurs abundantly as smoky quartz in peg¬

matites of thè more siliceous plutons, especially thè

Nine Mile pluton, and can be more than 80 centime-

ters in length.The color is typically a deep brown, thè

surfaces are commonly slightly frosted. However, in a

few miarolitic cavities, very lustrous and light colored

crystals have been found. The habit of thè quartz in

thè pegmatites of thè Wausau Complex has a charac-

teristic barrei shape. An interesting feature of rup-

tured pockets is that they are partially filled with

fractured and re-cemented quartz and feldspar frag¬

ments (“pocket hash”). This is likely due to thè vio-

lent nature of pocket rupture and rapid degassing. In

thè same pockets, oddly shaped quartz crystals are

very common. These represent shards of fractured

quartz, which continued to grow in pockets that had

resealed.

Plagioclase feldspar

Plagioclase is typically more abundant in thè foot¬

wall portions of both pegmatites as well as miarolitic

cavities, rather than in thè more potassic hanging

wall. Compositionally, thè plagioclase in thè peg¬

matites is albite to oligoclase. Maximum size of pla¬

gioclase crystals reaches about 4 centimeters. The col-

or ranges from almost white to pink to reddish with

thè most intense coloration in thè outer zones of thè

crystals. Some cores are clear and colorless. The char-

acteristic color of plagioclase from thè Wausau Com¬

plex is an intense pink to red. This color is quite un¬

usual for plagioclase and they in fact are darker than

thè associated potassium feldspars.

Potassium feldspar

Potassium feldspar, particularly maximum micro¬

cline, is exceedingly abundant throughout thè com¬

plex. In miarolitic cavities, large maximum microcline

crystals up to 40 centimeters can be found. They are

most commonly pink to reddish in color. A few pock¬

ets yielded white potassium feldspar and three peg¬

matites to date have produced green amazonite,

which ranges in color from sky blue to deep green.

Twinning is rare in Wausau Complex feldspars.

Baveno and Manebach twins are more common than

Carlsbad twins. In ruptured pockets metastable, com-

pletely disordered potassium feldspar has been found

as a pressure quench-coating over earlier pocket min¬

erals (Martin and Falster, 1986). A high percentage of

these precipitates are high sanidine, with some ortho-

clase or intermediate microcline.

Anorthoclase

Anorthoclase is a common constituent in thè py-

roxene syenites of thè Stettin pluton. In pegmatitic

portions, crystal grains may reach 40 centimeters in

maximum dimension. The color is gray to dark gray

but lightens upon weathering. A beautiful blue

schiller, a result of exsolution, is characteristic.

Amphibole

Amphibole is abundant in thè Stettin pluton and

in thè other plutons of thè Wausau Complex. In thè

PEGMATITES AND PEGMATITE MINERALS OF THE WAUSAU COMPLEX

19

Table 1: Distribution of minerals in pegmatites of thè Wausau Complex. Column headings as follows:

SP-NP = nepheline-bearing pegmatite, Stettin pluton; SP-AP = amphibole- and pyroxene-bearing pegmatite, Stet-

tin pluton; SP-FP = feldspathic pegmatite, Stettin pluton; WP-QWP = quartz-bearing pegmatite, Wausau pluton;

RMP-QRP = quartz-bearing pegmatite, Rib Mountain pluton; NMP-PP = pegmatitic pod, Nine Mile pluton; NMP-

ZP = zoned pegmatite, Nine Mite pluton; NMP-SEP = solution-etched pegmatite, Nine Mile pluton; and NMP-VP

= vuggy pegmatite, Nine Mile pluton. Minerai abundance in thè pegmatites are described by thè following:

M = major rock-forming minerai; S = subordinate minerai; A = common accessory minerai; and R = rare acces-

sory minerai.

20

ALEXANDER U. FALSTER - WM. B. SIMMONS - KAREN L. WEBBER & TOM BUCHHOLZ

pegmatites of thè Stettin pluton, notably in those oc-

curring in thè amphibole syenites, large arfvedsonites

up to 25 centimeters in maximum dimension occur.

They are characteristically rimmed with aegirine-

augite. Riebeckite may also occur as oriented over-

growths or as solitary crystals in small miarolitic cav-

ities. Fracture planes across pegmatites and syenitic

rocks in thè Wausau pluton commonly are coated

with sodic amphiboles (arfvedsonite or riebeckite).

Pyroxene

Both sodic and calcic clinopyroxene occur in thè

Stettin pluton (Myers et al., 1984). Aegirine and ae-

girine-augite are abundant in most of thè pegmatites

of thè Stettin pluton and occur in some pegmatites of

thè Wausau pluton. Calcic pyroxene (diopside-

hedenbergite series) is iron-rich with up to 10 % ae¬

girine content (Koellner, 1974).

Biotite

Biotite is very abundant in all pegmatites. Peg-

matite wall zones are particularly rich in biotite,

which displays a characteristic spear-shaped habit. At

thè core margin, biotite occurs in large, thin books up

to 40 centimeters across. In other pegmatite units, bi¬

otite occurs in thicker, but smaller books. Near pock¬

et zones, biotite is typically heavily altered.

Minor and accessory minerals

Sulfides and sulfosalts

Arsenopyrite

Arsenopyrite is very rare, having been found in

only a few miarolitic cavities in pegmatites of thè

Nine Mile pluton. Most crystals, unless they were pro-

tected within quartz crystals, have been replaced by

Fe oxides or hydroxides. The maximum size of these

crystals reaches 2 millimeters.

Jamesonite

Jamesonite and possibly boulangerite have been

found as inclusions in smoky quartz from miarolitic

cavities only. These sulfosalts form acicular inclusions

of gray crystals, which are commonly bent and twist-

ed. They reach a maximum length of 3 centimeters

but a maximum thickness of only 0.2 millimeters.

Galena

Galena was identified as inclusions in smoky

quartz in several pegmatites in thè Nine Mile pluton.

The maximum size of thè tiny cubes was approxi-

mately 80 microns. Galena was associated with pyrite,

sphalerite, and jamesonite in thè smoky quartz. The

presence of galena and other sulfides in trace

amounts indicates that thè sulfur activity was rela-

tively high in these pegmatites, thus only rarely was

any Pb left over to partition into silicates. Pb in

feldspar structures, coupled with H.O, is responsible

for amazonite’s color (Plyushnin, 1969; Smith, 1974).

Pyrite

Pyrite is only preserved in thè pegmatites where it

was enclosed in smoky quartz. These crystals measure

up to 1 millimeter in maximum dimension. They are

commonly cubes or modified cubes, but in some peg¬

matites beautiful acicular pyrite was found. It typical¬

ly forms stender extensions from thè cube faces and

results in a star-like growth. Pyrite also occurs as an

early component in massive pegmatite where it forms

crystals up to 1 centimeter. Illese crystals are always

replaced by goethite or hematite. One pegmatite con-

tained up to 20 % pyrite.

Sphalerite

Sphalerite has been identified as inclusions in

smoky quartz in several pegmatites in thè Nine Mile

pluton. The maximum size of thè tiny tetrahedra was

about 80 microns. Sphalerite was associated with

pyrite, galena, and jamesonite in thè smoky quartz.

Halides

Fluorite

Fluorite is very rare or absent in most pegmatites

of thè Wausau Complex. However, it is abundant at

one site in thè Stettin pluton and another area in thè

southern most Nine Mile pluton. The color ranges

from purple to blue, green, and almost colorless.

Oxides

Anatase

Anatase is thè dominant Ti-oxide in thè Wausau

Complex. It is found in primary and, more common¬

ly, in secondary miarolitic cavities. It forms black to

blue-black, rarely brown or yellow crystals up to 9

millimeters in maximum dimension (Fig. 3). The av-

erage size is about 1 millimeter. The forms of thè

dipyramid are characteristic. Pinacoids are present in

many examples. Second order dipyramids and prism

faces are less abundant.

Brookite

Brookite is less common than anatase but stili

more abundant than rutile in thè Wausau Complex. It

is most commonly found as thick tabular black or

dark brown crystals in secondary miarolitic cavities

(Fig. 4). It may reach 6 millimeters in maximum di¬

mension and is typically associated with either

anatase or rutile. At times, all three polymorphs occur

together, although at least one of thè polymorphs will

have undergone paramorphic replacement. X-ray dif-

fraction indicates that most brookite has not convert-

ed to other TiO.phases. Fe is thè dominant contami-

PEGMATITES AND PEG MATITE MINERALS OF THE WAUSAU COMPLEX

Fig. 3 - Secondary electron image of anatase from thè secondary

miarolitic cavities in a group 3 pegmatite in thè Nine Mile pluton.

Photograph by Dr. E. E. Foord.

Fig. 4 - Secondary electron photomicrograph of brookite from thè

secondary miarolitic cavities in a group 3 pegmatite in thè Nine

Mile pluton.

nating ion. Microprobe analysis indicates that up to

0.9 weight percent FeO may be present in brookite.

Lighter colored brookites are nearly pure TiO,.

Columbite-tantalite

Members of thè columbite-tantalite group are

very rare in thè Stettin pluton. They are occasionally

found as millimeter-sized masses in thè zircon-rich

quartz cores of pegmatites in thè amphibole syenites.

In some of thè pegmatites of thè Nine Mile pluton,

ferrocolumbite is very sparingly found as centimeter-

sized crystals in thè massive hanging wall portion di-

rectly above pockets. Inside thè pockets, columbite-

tantalite is exceedingly rare.

Goethite

Goethite is most commonly found replacing

siderite (Fig. 5). It frequently forms fine, platy crystals

of brown to golden color. In some cases, these crystals

Fig. 5 - Backscattered electron photomicrograph of goethite from

thè interior of a replaced siderite rhomb in a cavity of a group 3

pegmatite in thè Nine Mile pluton.

appear to be composed of polysynthetic twins. Maxi¬

mum size of goethite crystals approaches 6 millime-

ters. Overgrowths of goethite on quartz and other

pocket minerals is rare but has been found in a few

pegmatites.

Hematite

Hematite occurs in 4 rnodes in thè Wausau Com-

plex: 1) Discrete platy crystals up to 1 centimeter in

diameter in primary and secondary miarolitic cavi¬

ties. At times, “rose-like” clusters are found, not un-

like thè Alpine “Iron-roses”; 2) As a replacement of

siderite and biotite. Biotite in thè massive pegmatite

is frequently replaced by massive hematite; 3) As al-

most monomineralic “pocket-clay”. Bright red, clay-

like hematite occurs in many miarolitic cavities. No-

tably those which show effects of intense corrosion of

pocket constituents commonly contain large amounts

of this hematite. Hisingerite may accompany it; and

4) As late stains and coatings over crystals and frac-

ture planes.

Ilmenite

Ilmenite is a common accessory minerai in many

primary and secondary miarolitic cavities. It forms

platy crystals up to 1 centimeter across. In a few pock¬

ets, rosettes of millimeter-sized ilmenites were found.

Some massive ilmenite occurs in thè nepheline syen-

ite pegmatites of thè Stettin pluton. X-ray diffraction

and microprobe analysis confirms thè stoichiometry

and crystal structure.

Lepidocrocite

Lepidocrocite occurs only in thè cavities of re¬

placed siderite, commonly associated with goethite or

hematite. The crystals are very thin, tabular, and have

a bright red color. Maximum size approaches 1 mil-

limeter.

22

ALEXANDER U. FALSTER - WM. B. SIMMONS - KAREN L. WEBBER & TOM BUCHHOLZ

Magnetite

Magnetite is common in thè nepheline syenites and

thè nepheline syenite pegmatites of thè Stettin pluton

but is rare elsewhere in thè Wausau Complex. Mag¬

netite occurs most abundantly in grains and crystals up

to 1.5 centimeters in diameter in thè nepheline syenite

ring around thè pyroxene syenite core of thè Stettin plu¬

ton. It also occurs abundantly as detritai grains in

creeks. In some non-pegmatitic rocks of thè Stettin plu¬

ton and thè Wausau pluton, segregation masses of mag¬

netite up to several centimeters across have been found.

Mn-oxides

Mn-oxides such as romanechite are common as a

late stain on pocket minerals and fracture planes.

Opal

A colorless to white common opal is locally seen

coating fracture surfaces. It is of meteoric origin. This

opal exhibits a bright green fluorescence under ultra-

violet light.

Pyrochlore

Members of thè pyrochlore group are restricted to

thè pegmatites of thè Stettin pluton. They form small

crystals that rarely exceed 1 millimeter in maximum

dimension and are dark brown, black, or waxy yellow

in color. Octahedral faces are dominant with subordi¬

nate cube faces modifying them in some pegmatites.

Bario-pyrochlore is largely restricted to thè aplites in

pegmatites of thè amphibole syenites, whereas bario-

cerio-pyrochlore typically occurs in thè vuggy inter¬

mediate zone of thè same pegmatites. A pyrochlore

with several percent Y,0, is rarely found as tiny (< 0.1

millimeters) crystals in aplitic portions in some peg¬

matites of thè pyroxene and amphibole syenites. Sev¬

eral pyrochlore analyses of pyrochlores are reported

in Weidman (1907).

Rutile

Rutile is thè least common TiO.polymorph in thè

Wausau Complex. It typically forms millimeter-sized

networks of intergrown crystals (“sagenite”).The col-

or ranges from black to straw yellow. It is typically as-

sociated with other Ti-oxides.

X-ray diffraction confirmed rutile but in one case,

a paramorph of anatase after rutile was encountered.

According to microprobe analysis, FeO (up to 0.7

weight percent) is thè only common impurity.

Carbonates

Calcite

Calcite has two distinct modes of occurrence: 1) It

has been found as inclusions in smoky quartz from

pegmatites in thè more siliceous rocks of thè com¬

plex. These crystals reach 5 millimeters in maximum

dimension and are simple rhombohedra. Rhombohe-

dral impressions on thè surface of many quartz crys¬

tals indicate that calcite has been present in thè pock-

ets but subsequently dissolved; and 2) In thè pyrox¬

ene syenite, calcite is commonly seen in some peg-

matitic segregations as well as in thè pyroxene syen¬

ite itself. Carbonate segregations may be an indica-

tion of an incipient carbonatite body in thè pyroxene

syenite. The agpaitic trend in pyroxene syenite fur-

ther supports thè notion that these carbonate segre¬

gations could be carbonatitic in origin.

Parisite-synchisite

Parisite-synchisite has been noted in several peg¬

matites as crystals less than 1 millimeter in maximum

dimension. It appears to be more common in thè

Stettin pluton.

Siderite

Siderite is a very common miarolitic cavity minerai

in all plutons of thè Wausau Complex. However, unless

siderite was completely enclosed in quartz, it was al-

tered to goethite, lepidocrocite, or hematite. In most

cases, these pseudomorphs preserved thè delicate sur¬

face detail of thè former siderite crystals. The interiors

of these pseudomorphs are commonly lined with well-

formed goethite, hematite, or lepidocrocite crystals.

Maximum size of these brown or ochre colored

pseudomorphs may exceed 15 centimeters on an edge.

Secondary REE-bearing minerals

Crusts of secondary REE-bearing minerals are

commonly encountered along fractures and as coat-

ings on pocket minerals. Both LREE and HREE-

bearing components have been identified, frequently

in dose proximity to each other. These species are

probably carbonates or fluoro-carbonates.

Sulfates

Anhydrite

Anhydrite has been found with barite as tiny, less

than 1 millimeter inclusions in pocket quartz crystals.

Barite

Barite has been observed in a few pocket quartz

crystals as inclusions less than 1 millimeter in length.

They may have had their origin in thè alteration of

some feldspars with minor Ba-content.

Phosphates

Apatite

Apatite occurs throughout thè complex as an ac-

cessory minerai in all rock types and in thè peg-

PEGMATITES AND PEGMATITE MINERALS OF THE WAUSAU COMPLEX

23

matites. Well-formed euhedral crystals are found in

some miarolitic cavities, notably in thè pegmatites of

thè Stettin pluton. Maximum size of crystals may

reach 1 centimeter but most crystals remain below 1

millimeter in size. The color is white to colorless. Ul-

traviolet fluorescence is generally tan to yellowish in

color. Crystals characteristically are short to long

prismatic with large pinacoids and small dipyramids.

Cheralite

Silicates

Agrellite

Agrellite was noted by Medaris and Guggenheim

(1983), in a Wausau pluton pegmatite where it ap-

pears in thè keelward portion associated with wollas-

tonite, miserite, titanite, and feldspars. This is a very

unusual occurrence and only thè second known loca¬

tion for this minerai. Agrellite forms white prisms up

to 6 millimeters in length.

Cheralite is thè dominant LREE minerai in thè

Wausau Complex. It occurs in all 4 plutons but is

most abundant in thè Stettin pluton and in thè Nine

Mile pluton (Fig. 6). It is found in primary and sec-

ondary miarolitic cavities and in massive pegmatite

dose to miarolitic cavities. Cheralite forms flattened,

disk-shaped crystals, which aggregate to form flower-

shaped rosettes or parallel stacks in thè Stettin plu¬

ton. Here cheralite has a higher Th-content and is

commonly overgrown with small fergusonite crystals.

In thè other Wausau Complex plutons, cheralite

forms pseudohexagonal dipyramids. The maximum

size reaches 1 centimeter. The color is typically yel¬

lowish on thè interior of crystals and brick red on thè

exterior. Basai cleavage is evident.

Fig. 6 - Secondary electron photomicrograph of cheralite with fer¬

gusonite from a small miarolitic cavity in a pegmatite in thè am-

phibole syenites of thè Nine Mile pluton.

Allanite

Allanite occurs in many of thè pegmatites in thè

Stettin and Wausau plutons as an accessory minerai.

It typically forms black, platy crystals frozen in matrix

up to 3 centimeters in maximum dimension. A few

isolated pegmatites in thè Nine Mile pluton contain

rare allanite crystals frozen in matrix in thè coarse

feldspar-quartz intergrowth of thè hanging wall di-

rectly above miarolitic cavities.

Almandine

Almandine occurs rarely as millimeter size brown

crystals in some primary and secondary miarolitic

cavities in thè Nine Mile pluton.

Andradite

Andradite has been found in secondary miarolitic

cavities of several pegmatites in thè Nine Mile pluton.

The grains are clear, sulfur yellow in color, and reach 4

millimeters in diameter. They were identified by X-ray

diffraction and energy dispersive spectral analysis.

Bavenite

Bavenite occurs in some pegmatites of thè Nine

Mile pluton as white, fibrous inclusions in smoky

quartz. Maximum size reaches 3 millimeters. Baven¬

ite may also have been a pocket minerai but probably

did not survive exposure to weathering and cleaning

of thè samples. Identification was performed by X-

ray diffraction.

Monazite

Monazite is a very abundant accessory minerai in

thè pegmatites of thè Wausau Complex and is most

commonly found in secondary miarolitic cavities or

in small primary miarolitic cavities. Crystals are com¬

monly platy and brown in color.

Xenotime

Xenotime is found widespread as millimeter-size

yellow to pink or reddish crystals in secondary mi¬

arolitic cavities. It resembles zircon in morphology

and color. Forms exhibited are prisms and dipyra¬

mids. In some crystals, pinacoids and second order

prisms have been detected.

Bertrandite

Bertrandite is, next to phenakite, thè most abun¬

dant Be minerai in thè Wausau Complex. It is most

common in thè pegmatites of thè Nine Mile pluton

where it occurs as complex and well-formed crystals

up to 7 millimeters in maximum dimension, with an

average dimension of about 0.5 millimeters. Most of

thè crystals are spectacular twins and trillings (Fig. 7).

Bertrandite occurs near altered phenakite and is

strictly a pocket minerai.

Beryl

Beryl is very rare in thè Wausau Complex. It has

been found in only two miarolitic cavities in peg-

24

ALEXANDER U. FALSTER - WM. B. SIMMONS - KAREN L. WEBBER & TOM BUCHHOLZ

à

XI 080 ìaeitV- U . n!*cT «8882

Fig. 7 - Secondary electron photomicrograph of bertrandite from

thè primary miarolitic cavities in a group 3 pegmatite in thè Nine

Mile pluton.

matites in thè Nine Mile pluton but never in massive

pegmatite. Beryl hearing pegmatites are closely asso¬

ciateci with partially assimilateci metasediments. It is

possible that these metasediments supplied thè addi-

tional Al necessary to reach thè stability field of

beryl. Without this additional source of Al thè gener-

ally alkalic nature of thè complex would destabilize

beryl in favor of phenakite or bertrandite. Interest-

ingly, thè two beryl hearing pegmatites also contained

phenakite and bertrandite, further indicating that

conditions were barely within thè stability field of

beryl. Beryl crystals were clear and colorless or sky-

blue aquamarines, with a maximum size of two cen-

timeters. Crystal forms noted include prisms, pina-

coids, and dipyramids.

12 centimeters across in thè interior of some peg¬

matites of thè Nine Mile pluton.

Fergusonite

Fergusonite occurs only in thè pegmatites of thè

Stettin pluton. In particular, pegmatites in thè amphi-

bole syenites yield abundant tiny fergusonites. The

maximum size of thè reddish brown crystals does not

exceed 100 microns.

Hisingerite

Hisingerite has been identified as a phase in thè

pocket fills of hematite-rich pockets. The minerai

forms crude micaceous plates and clayey deposits.

Identification was done by X-ray diffraction.

Miserite

Miserite is found only in thè agrellite-bearing peg¬

matite in thè Wausau pluton where it forms off-white

to tan prisms up to 1 centimeter in length.

Muscovite

Muscovite is a rare component in thè pegmatites.

It typically occurs as sericite in miarolitic cavities and

as a rock-forming component in some pegmatites of

thè amphibole syenites in thè Stettin pluton. The col-

or is grayish to pale yellow, notably in thè zinnwaldite

(uirvingite”)-bearing pegmatite in thè Stettin pluton.

Cancrinite

Cancrinite occurs in thè pegmatitic and non-peg-

matitic portions of nepheline syenite as orange patch-

es up to 2 centimeters in diameter.

Natrolite

Natrolite is one of thè alteration products of

nepheline in thè pegmatitic and non-pegmatitic por¬

tions of thè Stettin pluton. It has not been noted else-

where in thè complex.

Catapleiite

Catapleiite occurs very sparingly in pegmatites of

thè Stettin pluton. Thin colorless crystals of pseudo-

hexagonal catapleiite up to 0.5 millimeters are found

very rarely in small miarolitic cavities.

Eudialyte

Eudialyte has been detected in very small

amounts in massive nepheline syenite pegmatite of

thè outer portions of thè Stettin pluton as deep red to

brown color rounded grains up to 4 millimeters

across.

Fayalite

Fayalite occurs in thè alkali granite and monzonite

ot thè Nine Mile pluton as an accessory minerai. It al¬

so occurs sporadically as slightly altered masses up to

Nepheline

Nepheline occurs only in thè outer rim of thè Stet¬

tin pluton and in thè nepheline syenites which rim thè

pyroxene syenite core of thè Stettin pluton. In thè

pegmatitic portions of these rock types, nepheline

crystals up to 30 centimeters in length have been

found. Crystals are characteristically crude and some-

times corroded or altered, and generally gray to off-

white in color.

Phenakite

Phenakite is thè dominant Be-mineral in mi¬

arolitic cavities in pegmatites of thè Nine Mile pluton

(Falster, 1977). It typically forms interpenetration

twins up to 3 centimeters in length and is colorless, or

yellow to white in color. Single crystals are common-

ly simple rhombs up to 1 centimeter in maximum di-

mension. Partial or complete replacement of

phenakite by bertrandite is common (Fig. 8).

PEGMATITES AND PEGMATITE MINERALS OF THE WAUSAU COMPLEX

25

Fig. 8 - Secondary electron photomicrograph of phenakite from

thè primary miarolitic cavities in a group 3 pegmatite in thè Nine

Mile pluton.

Thorite

Thorite occurs in pegmatites of thè amphibole

syenites in thè Stettin pluton. It is found in thè inter¬

mediate and core margins of these pegmatites, com-

monly associated with allanite and often altered to

thorogummite. As a result of thè high levels of ra-

dioactivity detected in some of thè pegmatites in thè

Stettin pluton in thè late 1950s, mining of these Th-

minerals was attempted but was abandoned when it

was found that thè desired U was present in small

amounts. Thorite reaches several centimeters across

and may occur in large clusters. It is originally black

in color but turns brown when it alters to thorogum¬

mite.

Thorogummite

Thorogummite is a common alteration product of

thorite. It may form masses up to 10 centimeters in di-

ameter, which may represent clusters of originai thor¬

ite crystals. Thorogummite is typically chocolate

brown and resinous in appearance.

Titanite

Titanite occurs abundantly in thè Stettin pluton,

notably in nepheline syenite pegmatites where it may

reach a size of up to 2 centimeters across. Crystals are

typically shiny and black and may comprise 10 per-

cent of some pegmatites. Titanite also occurs in some

pegmatites of thè Wausau pluton, although not as

abundantly as in thè Stettin pluton.

Tourmaline

Two grams of tourmaline in thè form of thin, long-

prismatic crystals of dark green color have been

found in one large miarolitic cavity in thè Nine Mile

pluton. The minerai was identified as elbaite by X-ray

diffraction.

Wollastonite

Wollastonite was noted only in thè agrellite-bear-

ing pegmatite where it forms white prisms up to 8

millimeters in length.

Zinnwaldite, variety “Irvingite”

A variety of zinnwaldite called “irvingite” (Weid-

man, 1907), occurs only in a restricted area of thè

Stettin pluton where it forms bright yellow mica crys¬

tals up to 2 centimeters in diameter in a poorly ex-

posed pegmatite. This minerai represents thè only Li

minerai in thè Wausau Complex.

Zircon

Zircon is most abundant in thè Stettin pluton

where it formed early in thè crystallization history of

thè tabular syenite and is associated with purple flu¬

orite. Crystals are brown dipyramids, rarely with thin

prism faces, and reach 2 cm in maximum dimension.

Quartz-zircon cores occur in some larger pegmatites

in thè Stettin pluton. These crystals can be yellow or

reddish to lilac in color and are commonly clear with

a maximum size that reaches 1 cm. In thè pegmatites

of thè other plutons, zircon is a typical pocket miner¬

ai. It occurs as gemmy, millimeter-size crystals, very

similar to thè xenotime described above and is par-

ticularly abundant in secondary miarolitic cavities.

Some deep brown zircon is found in massive peg¬

matite and may reach up to several millimeters in

maximum dimension. These crystals are elongated

and typically consist of prisms and dipyramids.

Preliminary geochemical data

Whole rock geochemical data are available from

Sood et al. , (1980) and from thè authors’ neutron ac-

tivation results for pegmatite wall zones from several

pegmatites in thè Stettin pluton and pegmatite

groups 2, 3, and 4 in thè Nine Mile pluton (Tables 2-

3, Figs. 9-11). Chondrite normalized plots of REE

(rare earth elements) for pegmatite wall zones (and a

few other zones) in thè Stettin pluton and thè Nine

Mile pluton reveal interesting trends, particularly

when compared to thè average REE-content in thè

Fig. 9 - Chondrite normalized plot of average REE distribution in

thè foot wall wall zone, miarolitic cavity zone, and hanging wall

wall zone average of a group 3 pegmatite in thè Nine Mile pluton.

26

ALEXANDER U. FALSTER - WM. B. SIMMONS - KAREN L. WEBBER & TOM BUCHHOLZ

Fig. 10 - Chondrite normalized plot of average REE distribution

inìhe average wall zones of thè Wausau complex and South Piat¬

te district pegmatites.

Fig. 1 1 - Chondrite normalized plot of average REE distribution

in thè wall zones of Stettin and Nine Mile pluton pegmatites com-

pared to thè wall zone composition of South Piatte pegmatites.

wall zones of pegmatites from South Piatte, Colorado

(Simmons et al., 1987). The wall zones of thè Stettin

pluton pegmatites, particularly in thè amphibole- and

pyroxene-bearing pegmatites, have thè highest REE

contents. It is obvious from Figures 9 through 11, that

thè REE in thè Wausau complex are only slightly

LREE-enriched.

The wall zones of pegmatites in thè Nine Mile plu¬

ton, have a noticeably lower concentration of REE

than those in thè Stettin pluton (Fig. 10).

The distribution of modest sized REE-bearing

minerals in thè Wausau Complex pegmatites is gen-

erally diffuse, compared to similar REE-bearing

minerals which occur in large masses in replacement

units in thè pegmatites of thè South Piatte district

(Simmons et al, 1987). Possibly, thè higher F activi-

ty in thè South Piatte pegmatites enhanced thè mi-

gration and enrichment of REE-bearing phases in

late stage pegmatite units. The Wausau Complex is

generally low in F and thus REE appear to remain

more evenly distributed in thè pegmatites. A frac-

tionation is noticeable into thè foot wall and thè in¬

termediate zones of thè pegmatites in thè Wausau

Complex (Fig. 9).

Table 2: Neutron activation data for Wausau Complex pegmatites (all data in ppm). Chondrite normalizing val-

ues from Taylor and McLennan (1985). n.d. = not detected. Sample numbers refer to thè following pegmatites:

WC-1: Wall zone of an unzoned zircon-bearing feldspar pegmatite from thè tabular syenite in thè outer rim of

thè Stettin pluton.

WC-2: Wall zone of an arfvedsonite-aegerine-augite-bearing pegmatite from thè amphibole syenite of thè Stet¬

tin pluton.

WC-3: Wall zone of an unzoned nepheline syenite pegmatite from thè outer rim of thè Stettin pluton.

WC-4: Wall zone (hanging wall) of pegmatite 917A (group 3 pegmatite) in thè Nine Mile pluton.

WC-5: Zone of secondary miarolitic cavities in pegmatite 917A (group 3 pegmatite) in thè Nine Mile pluton.

WC-6: Wall zone (footwall) of pegmatite 917A (group 3 pegmatite) in thè Nine Mile pluton.

WC-7: Wall zone (hanging wall) of pegmatite 808A (group 4 pegmatite) in thè Nine Mile pluton.

PEGMATITES AND PEG MATITE MINERALS OF THE WAUSAU COMPLEX

27

Table 3: Neutron activation data for Wausau Complex pegmatites (all data in ppm)

Chondrite normalizing values from Taylor and McLennan (1985). n.d. = not detected Sample numbers refer to

thè following pegmatites:

WC-8: Hanging wall aplite of pegmatite 808A (group 4 pegmatite) in thè Nine Mile pluton.

WC-9: Wall zone (footwall) of pegmatite 808A (group 4 pegmatite) in thè Nine Mile pluton.

WC-10: Zone of secondary miarolitic cavities of pegmatite 22C (group 3 pegmatite) in thè Nine Mile pluton.

WC-11: Wall zone (hanging wall) of pegmatite 190H (group 3 pegmatite) in'the Nine Mile pluton.

WC-12: Intermediate zone of pegmatite 190H (group 3 pegmatite) in thè Nine Mile pluton.

WC-13: Intermediate zone of pegmatite 900A (group 2 pegmatite) in thè Nine Mile pluton.

WC-14: Miarolitic cavity floor of pegmatite 900A (group 2 pegmatite) in thè Nine Mile pluton.

CONCLUSIONS

The Wausau Complex is a rift-related alkalic com¬

plex and as such bears a distinct geochemical signa¬

ture. Cerny (1991) divides pegmatites into two types,

a LCT type (Li, Cs, and Ta enriched) of orogenic peg¬

matites and an NYF type (Nb, Y, and F enriched) of

anorogenic affinity. The LCT type pegmatites are fur-

ther subdivided into classes based on chemistry and

mineralogy. However, this has not yet been done for

thè NYF type anorogenic pegmatites. A classical ex-

ample of NYF pegmatites, thè South Piatte pegmatite

district in Colorado, is presented in Simmons et al.

(1987).Tectonically, thè Wausau Complex pegmatites

are clearly anorogenic but they do show some devia-

tions from thè generai NYF type of pegmatites. Some

F-, Nb-, Zr, and Th-enrichments are evident in parts

of thè Stettin pluton but these enrichment trends are

lacking in thè Nine Mile pluton. Therefore, thè

Wausau Complex might be more properly included

in a separate class of pegmatite types, as it does not

neatly fit into either thè LCT type or NYF type peg¬

matites as described by Cerny (1991). Since thè en¬

riched elements common to all four intrusive centers

are Fe, Ti, and REE, an anorogenic pegmatite type

“FeTiREE”, indicating Fe, Ti, and REE-enrichment

could be introduced here. With thè exception of some

minor Li and B mineralization, thè Wausau Complex

is depleted in B, which is similar to thè South Piatte

district (Simmons et Al., 1987). Be is evident in thè

Nine Mile pluton but appears to be lacking in thè

Stettin pluton. Be is not common in thè South Piatte

district but is present in a few pegmatites as beryl or

gadolinite (Simmons et al., 1987). The data presented

here suggest that thè highest REE contents occur in

thè wall zones of pegmatites in thè Stettin pluton.

A comprehensive study of thè mineralogy and

geochemistry of thè Wausau Complex and its peg¬

matites is in progress and will address some of thè

problems discussed above.

Acknowledgments

We are indebted to thè following property owners

in thè Wausau area who graciously permitted sam¬

pling on their land: Robert Dehnel, Rainerd Zunker,

Cari Wimmer, Robert Thurber, Thomas O’Brian,

Paul Knopp, and Harold Beilke

Valuable input was received from thè following

minerai collectors of thè Wausau area: Dr. Earl John¬

son, David Bierbrauer, Dean Rein, Sr., Walter Tam-

minen, David Feck.

Field assistance was provided by Walter F. Prey

and Ursula H. Falster.

Dr. Eugene E. Foord (USGS, Denver, Colorado),

Thomas Campbell (Rapid City, South Dakota), and

Kenneth Keester (Santa Barbara, California) assisted

in thè early stages of minerai identification and mi-

crophotography. Instrumental neutron activation

analyses were carried out by thè Phoenix Memorial

Laboratory of thè University of Michigan, Ann Ar-

bor, through a grant to Wm. B. Simmons.

28

ALEXANDER U. FALSTER - WM. B. SIMMONS - KAREN L. WEBBER & TOM BUCHHOLZ

References

Cerny P., 1991 - Fertile Granites of Precambrian Rare-element

Fields: Is Geochemistry Controlled by Tectonic Setting or

Source Lithologies? Precambrian Research, 51:429-468.

Falster A. U., 1977 - Phenakite from Marathon County, Wiscon¬

sin. The Mineralogical Record , 8: 389-390.

Falster A. U., 1981 - Minerals of thè Wausau pluton. The Miner¬

alogical Record. 12: 93-97.

Falster A. U., 1986 - Minerals of thè Stettin pluton, Marathon

County, Wisconsin. Rocks and Minerals , 61: 74-78.

Falster A. U., 1987 - Minerals of thè Pegmatitic Bodies in thè

Wausau pluton, Marathon County, Wisconsin. Rocks and Min¬

erals, 62: 188-193.

Koellner S. E., 1974 - The Stettin Syenite Complex, Marathon

County, Wisconsin: Petrography and Minerai Chemistry of

Olivine, Pyroxene, Amphibole, Biotite, and Nepheline. Un-

published M. S. Thesis, University of Wisconsin, Madison.

Martin R. F. & Falster A. U., 1986 - Proterozoic Sanidine and

Microcline in Pegmatite, Wausau Complex, Wisconsin. The

Canadian Mineralogist, 24: 709-716.

Medaris L. G. & Guggenheim S., 1983 - An Occurrence of Agrel-

lite and Miserite in thè Wausau Complex, Marathon County,

Wisconsin. In: 1983 Symposium on Alkaline Complexes,

Wausau, Wisconsin. Abstract Volume.

Myers P. E, Sood M. K., Berlin L. A. & Falster A. U., 1984 -

The Wausau Syenite Complex, Central Wisconsin. 30th Annu-

al Institute on Lake Superior Geology. Fieldtrip Guidebook 3,

69 p.

Plyushnin G. S., 1969 - On thè coloration of amazonite. Zap. Vs-

es. Minerai. Obshchest., 98: 3-17 (in Russian).

Simmons Wm. B., Lee M.T. & Brewster R. H. 1987 - Geochem¬

istry and Evolution of thè South Piatte Granite-pegmatite

System, Jefferson County, Colorado. Geochimica et Cos¬

mochimica, 51: 455-471.

Smith J. V., 1974 - Feldspar minerals. Volume 1: Crystal structure

and physical properties. Ch. 12. Miscellaneous physical prop-

erties : 556-600.

Sood M. K., Myers P. E. & Berlin L. A., 1980 - The Petrology,

Geochemistry, and Contact Relations of thè Stettin and

Wausau Syenite Plutons, Central Wisconsin. 26th Annual In-

stitute on Lake Superior Geology. Fieldtrip guidebook, 3, 59 p.

Taylor S. R. & McLennan S. M. 1985 - The Continental Crust: Its

Composition and Evolution. Blackwell Scientific Publications,

Oxford, 312 p.

Van Schmus W. R., Medaris L. G. & Banks P. O., 1975a - Chronol-

ogy of Precambrian Rocks in Wisconsin, I: The Wolf River

Batholith, a Rapakivi Massif Approximately 1500 m. y. old.

Geological Society of America Bulletin, 86: 907-914.

Van Schmus W. R., Thurman E. M. & Peterman Z. E., 1975b -

Geology and Chronology of Precambrian Rocks in Wisconsin,

II: Rb-Sr Data for thè Older Rocks in Eastern and Central

Wisconsin. Geological Society of America Bulletin, 86: 1255-

1265.

Weidman S., 1907 - The Geology of North Central Wisconsin. Wis¬

consin Geological and Naturai History Survey, XVI, 697 p.

Memorie della Società Italiana di Scienze Naturali e del Museo Civico di Storia Naturale di Milano

Volume XXX - pp.: 29-43, 2000

Internai structures, parageneses and classification of thè

miarolitic Li-bearing complex pegmatites of Elba Island (Italy)

Federico Pezzotta*

Museo Civico di Storia Naturale of Milan, Corso Venezia 55 - 20121 Milan - Italy

Abstract - This paper reports a study on thè internai structure, thè mineralogy and thè paragenetic sequence in pockets

of thè miarolitic complex pegmatites of Elba Island (Italy). Depending on structural and mineralogical characteristics these

dikes are classified into four categories. Listed in order of increasing degree of evolution they are: 1) dikes without Li min¬

erai; 2) Li-bearing dikes with complex asymmetric zoning; 3) Li-bearing dikes with simple asymmetric zoning; 4) irregu-

larly zoned Li-bearing dikes. The variable degree of evolution is reflected in a roughly zoned occurrence across thè margin

of thè Mt. Capanne pluton, thè more evolved dikes being located farthest away from thè contacts in thè thermometamor-

phic aureole. The dikes without Li minerals are barren miarolitic dikes with dispersed pegmatitic pods in thè aplitic rock,

and with a primitive pocket mineralogy. The Li-bearing dikes with complex asymmetric zoning have a centrai portion with

well developed asymmetric zoning, characterised by a mainly aplitic sodic lower unit, and a potassium rich pegmatitic up¬

per unit, with large miarolitic cavities and an irregular, aplitic-pegmatitic peripheral portion, near thè termination of thè

dike. The pocket mineralogy of thè centrai portion is uniform and displays a moderate degree of evolution. In thè periph¬

eral portion thè mineralogy of cavities ranges from primitive to highly evolved. The Li-bearing dikes with simple asymmet¬

ric zoning show structural and mineralogical characteristics similar to those of thè centrai portion of thè second category

and have a uniform highly evolved pocket mineralogy. The irregularly zoned Li-bearing dikes are cavity-poor and display

an irregular zoning which may be locally absent. These dikes, hosted in thè hornfels, suffered massive loss of fluids during

thè crystallisation, as shown by alteration of thè host rocks. This prevented internai structural differentiation and thè for-

mation of miarolitic cavities.

The pocket evolution is complex, especially in thè large cavities of dikes of category 2, and includes pocket ruptures and

one, and locally two, hydrothermal stages. The structural, mineralogical, and paragenetic characteristics of thè dikes of thè

third category and thè centrai portion of thè ones of thè second category, have many similarities with those reported in thè

literature for miarolitic, gem-bearing dikes from California and Afghanistan.

Riassunto - Questa pubblicazione riporta uno studio sulle strutture interne, sulla mineralogia e sulla paragenesi delle

cavità delle pegmatiti miarolitiche complesse dellTsola d’Elba (Italia). In base alle caratteristiche strutturali e miner¬

alogiche questi filoni sono stati classificati in quattro categorie. Elencate in base al progressivo grado di evoluzione esse

sono: 1) dicchi privi di minerali di Li; 2) dicchi contenenti Li con zonatura asimmetrica complessa; 3) dicchi contenenti Li

con zonatura asimmetrica semplice; 4) dicchi contenenti Li irregolarmente zonati. Il variabile grado di evoluzione si riflette

in una zonatura di affioramento attraverso il margine del M.te Capanne, dove i dicchi a più alto grado di evoluzione si

trovano nell’anello di rocce termometamorfiche a maggior distanza dal contatto con il plutone. I dicchi privi di minerali di

Li sono dicchi sterili miarolitici con concentrazioni pegmatitiche sporadiche entro la roccia aplitica, e presentano una min¬

eralogia delle cavità poco evoluta. I dicchi contenenti Li con zonatura asimmetrica complessa presentano una porzione cen¬

trale con una ben sviluppata zonatura asimmetrica, caratterizzata da una unità inferiore aplitica prevalentemente sodica, e

da un’unità superiore pegmatitica ricca in potassio, con ampie cavità miarolitiche. Questi filoni presentano, in prossimità

delle loro estremità, porzioni periferiche irregolari aplitico-pegmatitiche. La mineralogia delle geodi della porzione centrale

è uniforme e mostra un moderato grado di evoluzione. Nelle porzioni periferiche la mineralogia delle cavità va da primiti¬

va a fortemente evoluta. I dicchi contenenti Li con zonatura asimmetrica semplice mostrano caratteristiche strutturali e

mineralogiche simili a quelle della porzione centrale dei filoni della seconda categoria e presentano una mineralogia delle

cavità uniforme e fortemente evoluta. I dicchi contenenti Li irregolarmente zonati sono poveri di cavità e mostrano una

zonatura irregolare, talvolta assente. Questi dicchi, messisi in posto negli hornfels, hanno subito una massiccia perdita di flu¬

idi durante la cristallizzazione, come indicato dalla alterazione della roccia incassante. Questo fenomeno ha impedito la dif¬

ferenziazione con formazione di strutture interne regolari e la formazione di cavità miarolitiche.

L’evoluzione delle geodi è complessa, soprattutto nel caso delle ampie cavità dei dicchi della seconda categoria, e com¬

prende eventi di fratturazione e localmente uno, se non due, stadi idrotermali. Le caratteristiche strutturali, mineralogiche

e paragenetiche dei dicchi della terza categoria e delle porzioni centrali di quelli della seconda categoria, hanno molte simil¬

itudini con quelle riportate in letteratura per i dicchi miarolitici, ricchi di gemme, della California e dell’Afghanistan.

Key words - granitic pegmatite, miarolitic cavity, Elba Island.

* E-mail: fpezzotta@yahoo.com

30

FEDERICO PEZZOTTA

Introduction

The aplite-pegmatite intrusions of thè Mt. Ca¬

panne Pluton, locateci in thè western part of thè Elba

island (Tyrrhenian Sea, Italy), are known for their ex-

cellent pocket collectibles and rare accessory miner¬

ai, produced mainly in thè past. Moreover, these in¬

trusions are thè only representatives of Li-bearing

pegmatites in Italy.

Although thè mineralogy of these dikes has been

investigated since thè 18th century, little is known

about their geochemistry, petrogenesis and internai

structure. Recently some studies were performed on

these topics, and thè first results are reported in Rug-

geri and Lattanzi (1992), Aurisicchio et al (1992),

Pezzotta (1993a-b), Aurisicchio et al. (1994) and Pez-

zotta (1994), Orlandi and Pezzotta (1996).

One approach to thè study of zoned pegmatites is

thè use of minerals with variable compositions and

textures (e.g. tourmaline, beryl, micas, and feldspars)

as recorders of petrogenetic information (e.g. Foord,

1976; Jahns, 1982; Shearer et al., 1985; Jollif et al.,

1986). The colour of tourmalines is roughly indicative

of their chemistry, and may be used to reveal thè frac-

tionation trends and relative degree of evolution of

thè rocks that host these minerals (Jollif et al, 1986,

and reference therein). Distinct types of textures of

many minerals (e.g. feldspars, tourmaline, micas, etc.),

confined to particular zones of a structurally complex

pegmatite, also evidence thè evolution of thè physico-

chemical conditions of crystallisation (e.g., London,

1992).

Here I report thè structural, mineralogical and

paragenetic features of thè Mt. Capanne pegmatite

field. The description and classification of thè Elba

pegmatites are based on abundance, distribution, tex-

ture, and parageneses of rock-forming minerals and

accessory phases. This paper also discusses thè petro¬

genesis and thè internai evolution of these intrusions,

as well as thè regional zonal distribution of thè dif-

ferent pegmatite types. Systematic analyses of miner¬

ai compositions and quantitative petrologie studies of

selected dikes are in progress.

Much of thè volume of thè famous Elba peg¬

matites has been mined out before World War I (e.g.

Sinkankas, 1981). During this study, detailed field

work since 1985 resulted in finding about 45 unex-

plored dikes along thè eastern margin of thè Mt. Ca¬

panne pluton. They are rich in accessory phases and

miarolitic cavities. The three-dimensional sampling

and description of a number of selected dikes have

been performed by mining, with collaboration of

minerai collectors. Data on thè famous but exhausted

dikes were obtained from thè literature, as well as

from observations on veins outeropping in old quar-

ries, pegmatite fragments in dumps, and specimens in

public and private mineralogical collections.

The Mt. Capanne pluton

The Mt. Capanne pluton is thè largest exposed in-

trusion of thè Tuscan magmatic province, and makes

up most of thè western part of thè Elba island (Fig.

l).The radiometrie (Rb/Sr) age of thè intrusion is 6.7-

6.9 Ma (Ferrara and Tonarini, 1985).

The classification of thè intrusive rocks of thè Mt.

Capanne pluton is problematic. Considering mode,

chemistry, zircon morphology (Pupin, 1976, 1980) and

abundance of mafie microgranular enclaves, they

show calc-alkaline characteristics. In contrast, thè

abundant aluminous xenoliths, thè presence of a

cordierite-bearing facies and certain dementai ratios

(Th/Ta, Th/U, La/Ta) are typical of peraluminous

granitoids (Bussy, 1990). Poli et al (1989), on thè ba-

sis of geochemical and isotopie data, suggest that thè

Mt. Capanne granite originated from anatectic mag-

mas formed by thè partial melting of metapelitic

rocks, metasomatically enriched in volatile elements.

A subcrustal potassic magma might have been thè

heat source for thè anatexis of thè metapelitic rocks.

The pluton intruded, deformed and injected rocks

of thè tectonic unit IV described by Keller and Pialli

(1990, and references therein), including ophiolites,

radiolarites, limestones and flysch. The intrusion also

produced a well-developed thermal aureole in thè

high horneblende hornfels facies.

The monzogranite dome

The “main facies” of thè pluton consists of a medi-

um-grained monzogranite. Near thè contact with thè

country rocks, thè equigranular granite becomes por-

phyritic, with large K-feldspar megacrysts. It contains

quartz, plagioclase (zoned from An40 to Anl2), K-

feldspar and biotite, accessory apatite, zircon, titanite,

tourmaline, ilmenite, secondary chlorite, sericite, cal¬

cite, and a small amount of alteration minerals. Some

dispersed grains of hereynite and locai cordierite are

distributed near thè contact (Poli et al, 1989).

Silicic dike rocks

Acidic dikes of thè M.te Capanne aureole are

shown in thè diagram of figure 2. These intrusions,

mainly composed of leucogranites and aplites (see

thè legend of Fig. 1) strike N-S and dip 30°- 75° E.

They are hosted by thè granodiorite phase of thè plu¬

ton, along thè contact of thè granodiorite and ther-

momethamorphic rocks, and in thè hornfels of thè

thermometamorphic aureole. Width of thè dikes

ranges from few centimetres up to 20 meters. The

leucogranites have a medium grain size and a locally

oriented magmatic texture. These rocks contain

Fig. 1 - Schematic geological map of thè eastern margin of thè Mt. Capanne pluton, Isola d’Elba (Italy). Simplified after

thè unpublished geological field map 1:5000 of thè author.

Q, quaternary sediments, C, limestones and sandstones of complex V; Se, metasedimentary hornfels of complex IV,

composed by biotitic schists with locai lenses of marbles; S, metabasic rocks of complex IV, composed by tremolite,

antofillite, and tale hearing serpentinites, with locai intense hydrothermal-thermometamorphic alteration associated with

magnesite deposits; 1— — J Mt. Capanne monzogranite;

J Leucogranite and aplite dikes; 7i, granitic porphyries.

Miarolitic aplite-pegmatite dikes: (1) dikes without Li minerals; (2) Li-bearing dikes with complex asymmetric zonation;

(3) Li-beanng dikes with simple asymmetric zonation; (4) irregularly zoned to unzoned Li-beraing dikes. Old pegmatite

quarries also are marked.

INTERNAL STRUCTURES. PARAGENESES AND CLASSIFICATION OF THE MIAROLITICI L-BEARING COMPLEX PEGNI AT1TES

31

Colle Reciso

'i' Pietra Acuta :<D

Elba Island

Fosso dei Fordoni

Ca' Mas taglino-^ (J)

Catri

Fosso

S. ILARIO IN CAMPO

Fosso Marcianelia

Fosso S. Francesco

Graziano

CD Grotta D 'Oi

Fosso Bovalico

S. PIERO IN CAMPO

qcciatoia

' Fonte del Prete

'asso Foresi

MARINA DI CAMPO

1 Km

32

FEDERICO PEZZOTTA

N

Fig. 2 - Pole diagram of thè acid dikes outcropping along thè E

margin of thè Mt. Capanne pluton. Triangles: leucogranite and

aplite dikes; black points: miarolitic aplite-pegmatite dikes and as-

sociated aplites. Stars indicate fractures coated by late black tour-

maline, feldspars, and quartz.

quartz, K-feldspar, plagioclase (5-30% An), subordi¬

nate biotite and white micas, and locally abundant

cordierite. Tourmaline, zircon, apatite, titanite, an-

dalusite, sulphides and other opaque minerals, are ac-

cessory phases. A late circulation of magmatic fluid is

responsible for thè locai alteration of feldspars and

Fe-Mg minerals. The aplites show a variatile grain

size, with rare patches of quartz-rich massive peg-

matite. Flow textures, as well as multiple intrusion

contacts and magmatic breccias, are widespread.

Quartz, K-feldspar, plagioclase (5-20% An), scarce

biotite and/or white micas are thè rock-forming min¬

erals. Zircon, apatite and andalusite are rare accesso-

ry phases. A tourmaline-rich facies, with abundant

subhedral to anhedral tourmaline crystals, is wide¬

spread. This tourmaline-rich facies locally displays a

strong deuteric alteration, with almost complete dis-

coloration and/or chloritization of biotite, formation

of iron hydroxides, and kaolinization and sericitiza-

tion of feldspars.

Miarolitic Li-bearing aplite-pegmatite dikes

The aplite-pegmatite dikes containing thè world-

famous miarolitic cavities lined by euhedral crystals

of several minerals species (Orlandi and Pezzotta,

1996) are widespread along thè eastern contact of thè

pluton in its peripheral parts and in thè surrounding

thermometamorphic rocks (Aloisi, 1920; Marinelli,

1959; Carobbi and Rodolico, 1976). The best known

occurrences are located near thè villages of S.Piero in

Campo and S. Ilario in Campo (Fig. 1). The size of

these dikes ranges from few centimetres up to 2 me-

ters in width, and up to 20 meters along strike.

In thè pockets, two main stages of thè paragenetic

sequence have been already recognised (e.g. Carobbi

and Rodolico, 1976):

- an “aplitic-pegmatitic” stage, consisting of K-

feldspar, albite, quartz, tourmaline, beryl, lepidolite,

petalite, pollucite and garnet.

- a “hydrothermal stage” responsible for thè crys-

tallisation of several zeolites, possibly pink elbaite

and calcite, and thè alteration of K-feldspar into

kaolinite and/or zeolites.

Ruggeri and Lattanzi (1992) described these peg-

matites as “miarolitic (shallow depth) pegmatites” in

thè classification scheme of Cerny (1982a).The same

authors, studying fluid inclusions in vug crystals of

quartz, tourmaline and beryl, calculated thè upper

pressure limit for pegmatitic crystallisation at ~ 2

kbar. These data are consistent with independent ge-

ological evidence.

The attitude of thè miarolitic, Li-bearing peg¬

matites, and thè associated aplite dikes swarms, oc-

curring along thè western margin of thè M.te Ca¬

panne pluton between Le Caviere (S of S.Piero) and

Fosso dei Forcioni (N of S. Ilario), are shown in thè

pole diagram of figure 2. These dikes (categorised as

(1), (2), (3), and (4) in Fig. 1) strike N-S and dip 35°-

80° W. They are hosted in thè granodiorite, in thè

larger leucogranitic and aplitic apophyses and dikes

outcropping along thè margin of thè pluton, and in

thè thermometamorphic rocks.

The structural evidence indicates that thè mi¬

arolitic, Li-bearing, pegmatities are thè latest intru-

sions emplaced into thè aureole of thè Mt. Capanne

pluton.

Fracture fillings and hydrothermal alteration

The attitude of a fracture System in thè granodior¬

ite of thè eastern margin of thè Mt. Capanne pluton,

associated with quartz, feldspars, and black tourma¬

line crystallisation and locai hydrothermal alter-

ations, is also shown in thè diagram of figure 2. These

brittle structures, with negligible vertical movement

slip, locally crosscut thè miarolitic pegmatites (Auris-

icchio et al, 1999).

The miarolitic aplite-pegmatite dikes

The study and thè classification of thè aplite-peg¬

matite dikes of thè E margin of thè Mt. Capanne plu¬

ton were performed taking into account thè internai

structure, textural patterns, rock-forming mineralogy

and thè abundance and distribution of accessory

phases, colour of tourmaline, and pocket parageneses.

Microstructural observations were performed on

thin sections cut along traverses across representative

dikes. Pocket minerai parageneses have been deci-

phered by microscopie observations of thè growth re-

lationships among thè different minerals. Besides a

large number of small cavities, about sixteen large

pockets have been observed in place and sampled by

thè author during this study.

Identification of minerals has been carried out by

qualitative SEM analysis performed with an EDAX

equipment attached on a Philips PW515 Scanning

Electron Microscope; quantitative WDS analysis was

carried out on an ARL SEMQ electron microprobe

in wavelength-dispersive mode. An automated

Rigaku difractometer and CuK(alpha) radiation

were used for X-ray powder diffraction. Selected

compositions of some minerals of thè studied dikes

INTERNAI. STRUCTURES. PARAGENESES AND CLASSIFICATION OF THE MIAROLITICI L-BEARING COMPLEX PEGMATITES

33

are reported in: Aurisicchio et al. (1988 - beryl); Or¬

landi et al. (1990 - Nb-Ta oxides); Aurisicchio et al.

(1993 - uranopolycrase); Orlandi and Pezzotta (1993

- sekaninaite); Pezzotta (1994 - helvite); Pezzotta et

al. (1996 - tourmaline); Aurisicchio et al. 1999 (tour-

maline); Aurisicchio et al. 1998 (Nb-Ta oxides); Teer-

stra et al. 1998 (rubidium feldspars and pollucite).

The aplite-pegmatite dikes were classified into 4

categories: 1) dikes without Li minerals;2) Li-bearing

dikes with complex asymmetric zoning; 3) Li-bearing

dikes with simple asymmetric zoning; 4) irregularly

zoned to unzoned Li-bearing dikes.

1) Dikes without Li minerals

The miarolitic dikes of this category are charac-

terised by lack of Li-Cs minerals and by thè presence

of tourmaline crystals of exclusively black colour.

These intrusions are abundant and are hosted by thè

monzogranite (in both thè equigranular and por-

phyritic facies). Locally they also are hosted by thè

aplite and leucogranite masses emplaced along thè

contact between granodiorite and hornfels (Fig. 1).

The dikes are a few centimetres to 1 meter thick and

2 to 20 meters long. Networks of small (1-20 cm

thick) crosscutting veins can also be observed in thè

monzogranite. These intrusions do not display signif-

icant textural or mineralogical zoning. They are main-

ly aplitic, with thè pegmatitic portions dispersed as ir-

regular pods and veins. Pegmatitic dikes lacking

aplite are exceptional. Miarolitic cavities, rarely ex-

ceeding 10 cm in diameter, occur only in pegmatitic

portions.

The main components of thè aplitic rock are

quartz, K-feldspar, plagioclase and biotite. Black

tourmaline is irregularly distributed. Blue beryl and

zircon are rare accessory phases. Tourmaline occurs

as primary, zoned, prismatic, euhedral crystals in

aplite and pegmatite, and as radiating crystal groups

and graphic intergrowths with quartz, albitic plagio¬

clase, and perthitic-graphic K-feldspar in pegmatite.

A second generation of intergranular tourmaline is

common.

The mineralogy of miarolitic cavities reflects thè

primitive nature of these pegmatites. Only tourma¬

line and, exceptionally, zoned garnet (25-40% Alm,

60-75% Sps, with Mn-enriched rim) are locally abun¬

dant. Tourmaline needles (syntaxial on primitive

crystals) and piumose aggregates are associated with

thè second generation of tourmaline. Other accesso-

Late stage

massive pegm. Pocket minerai

Minerals growth growth

Quartz

Feldspars

Tourmaline

Beryl

Cassiterite

Nb-Ta Oxides

Zircon

Apatite

Garnet

Early Late

Fig. 3 - Paragenetic sequence in pockets developed in dikes with¬

out Li minerals. (1) Black prismatic crystals; (2) black needles; (3)

pale blue beryl; (4) colourìess to pale green beryl.

ry phases are rare; thè pocket paragenesis (Fig. 3) is

very simple. No significant hydrothermal process of

minerai alteration is developed.

2) Li-bearing dikes with complex asymmetric zoning

The miarolitic dikes of this category are charac-

terised by Li-bearing accessory phases, and a com¬

plex three-dimensional asymmetric zonation. These

dikes display a marked variation in grain size, in

abundance and in textures of rock forming minerals,

and in distribution as well as composition of accesso¬

ry phases. Pocket parageneses range from primitive

to highly evolved types. It is possible to assume that

most of thè famous dikes of Grotta d’Oggi, La Sper¬

anza, Fonte del Prete and Fosso Gorgolinato (Millo-

sevich, 1914),belong to this category (see Fig. l).Two

of these dikes (named NI and N2 in Pezzotta, 1993a)

located about 50m SW from thè famous locality “La

Speranza”, are of particular interest, having large di-

mensions and being only little eroded by naturai

agents. These two pegmatites, exploited for minerai

specimens from 1990 to 1994, are here considered as

representatives of thè structures and mineralogy of

thè Li-bearing dikes of this category.

The dikes belonging to this category are hosted by

thè porphyritic monzogranite and, exceptionally,

along thè contact of thè leucogranite with basic ther-

mometamorphic rocks (Fig. 1). Most of them are 6-10

m long and 1-2 m thick. The dip can range from 25° to

75°. The contacts along strike are generally sharp, and

thè termination consist of a fingering-out network of

small veins. In some cases, thin anastomising aplitic

apophyses can run continuously for several tens of

meters.

The NI and N2 dikes are described here as typical

examples of Li-bearing dikes with complex asymmet¬

ric zoning. They are mutually similar but thè larger

one (N2) displays a much more complex zoning,

which is a multiple of that displayed by in thè NI

dike.

The S. Silvestro Dike (NI dike)

Figure 4 represents a horizontal section across thè

NI dike. A large “centrai portion” displays a well de¬

veloped asymmetric zoning, with an aplitic footwall

rich in Na and relatively depleted in K, and a coarse

grained overlying quartz-perthite-albite pegmatite,

enriched in K and poorer in Na. From thè floor up to

thè hanging wall, 4 units have been distinguished: 1)

aplitic footwall unit; 2) aplite-pegmatite unit; 3) peg¬

matitic core with pocket zone; 4) aplite-pegmatite

hanging-wall unit. Near thè termination of thè dike,

thè centrai portion rapidly disappears into “peripher-

al portions”. The peripheral portions are charac-

terised by aplitic grain size with dispersed pegmatitic

pods and veins and abundant monzogranite frag-

ments, with locally extensive tourmalinization.

The footwall part of thè NI dike consists of 5-25

cm thick aplitic unit. This massive aplite has a sugary

appearance and locally displays a banded texture

which consists of 0.1-1 cm thick layers. These layers

reflect variation in grain size and in concentration of

plagioclase, K-feldspar and biotite. The rock consists

of abundant plagioclase and quartz, scarce K-

Distribution of accessory phases

34

FEDERICO PEZZOTTA

Porphyric Aplitic

monzo granite unjt

Aplite-pegmatite

unit

Pegmatitic

core

Aplite-pegmatite Porphyric

unit monzo granite

PERIPHERAL PART

INTERNAL STRUCTURES. PARAGENESES AND CLASSIFICATION OF THE MIAROLITICI L-BEARING COMPLEX PEGNI ATITES

35

feldspar, and sparsely distributed biotite. Plagioclase

forms euhedral to anhedral progressively zoned crys-

tals (14-5% An). Little euhedral and progressively

zoned Ca-richer (20-8%) plagioclase crystals also are

abundant in thè mass of aplite and locally as inclu-

sions in thè K-feldspar. K-feldspar is present as an¬

hedral grains and in larger (up to 1.5 cm) perthitic

crystals with graphic quartz, elongated roughly per-

pendicular to thè wallrock contact. Biotite is present

as fine-grained flakes at thè contact with thè host wall

rock, and as medium-grained elongated crystals in

thè others portion of thè aplite. Micrographic biotite-

quartz intergrowth, as well as replacement and/or

overgrowth by white micas, are common along thè

margins of thè biotite crystals. Tourmaline is normal-

ly present as a rare accessory phase; it is concentrat-

ed in anhedral fracture-filling grains, preferentially

associated with K-feldspar.

The aplite unit grades upward into a roughly strat-

ified 30-50 cm thick aplite-pegmatite unit (Fig. 5). The

layering is concordant with thè attitude of thè dike

and consists of 2 to 4 cm thick discontinuous lenses of

coarse-grained miarolitic pegmatite, alternating with

fine-grained aplite. The pegmatitic lenses are charac-

terised by: enrichment in K-feldspar, small miarolitic

cavities (few cubie centimetres in volume), absence

of biotite near thè pockets and graphic tourmaline-

quartz intergrowths. Small pockets (maximum 1 cm3

in volume) also are present in thè aplitic lenses. Pla¬

gioclase grains (8-4% An) are mainly euhedral.

Perthitic K-feldspar with graphic quartz are concen-

trated in pegmatitic lenses. Biotite, in baguette-like

crystals, is rimmed by white micas and partially dis-

coloured and/or chloritised. Graphic tourmaline-

quartz intergrowths are common in pegmatite and lo¬

cally also form fracture fillings in K-feldspar. Blue to

colourless beryl is a rare accessory phase. In thè up¬

per part of thè aplite-pegmatite unit a weakly albi-

tised pethitic K-feldspar contains anhedral to euhe¬

dral albite-rich plagioclase crystals in fractures.

The aplite-pegmatite unit grades upward into a

20-45 cm thick giant-textured pegmatite. Almost all

of thè volume of thè miarolitic cavities contained in

thè dike is confined to this pegmatitic core. A large,

fiat, irregularly oblate pocket (2.30 x 0.35 x 0.80 me-

ters) was discovered in December 1990 in thè peg¬

matitic core of NI dike. The volume of this “Main

Pocket” (approximately 500 dm3) largely exceeds thè

sum of thè volumes of all other pockets found in thè

dike. This miarolitic cavity locally extends down into

thè aplite-pegmatite unit.

A “pocket horizon” divides thè pegmatitic core in¬

to a lower and an upper part, with partially different

textural and modal characteristics.

The lower part has a 3-6 cm grain size, common

small miarolitic cavities a few cubie centimetres in

volume, relatively abundant albite (4-0% An) mainly

in coarse graphic intergrowth with quartz; perthitic

K-feldspar with graphic quartz (4-6 cm across); rare

quartz-tourmaline graphic intergrowth and biotite in

baguette-like crystals up to 5 cm, confined to thè con¬

tact with thè lower unit, largely transformed into

white micas.

The part of thè pegmatitic core lying above thè

pocket horizon is characterised by a blocky texture

(up to 25 cm in grain size); absence of miarolitic cav¬

ities; very scarce albite which becomes abundant on-

ly dose to large cavities; abundant perthitic K-

feldspar (15-25 cm across) with graphic quartz; local¬

ly abundant graphic quartz-tourmaline intergrowths

rapidly turning into radiating groups of euhedral

crystals, or into single crystals projecting downward

perpendicular to thè roof of thè pocket; and anhedral

quartz masses up to 10 cm in diameter. Biotite is con¬

fined to thè upper contact of thè unit. All around thè

large miarolitic cavities, but mainly around thè “main

pocket”, radiai microfractures are present, filled with

albite and rarely also quartz and/or green tourmaline.

These brittle deformations, particularly evident on

microscopie observation, control a widespread and

abundant albitisation of thè K-feldspar.

The Aplite-Pegmatite Hanging-Wall Unit (3 to 12

cm thick), has a medium grain size (1-2 cm) and con¬

sists of abundant perthitic K-feldspar with graphic

quartz, plagioclase (8-2% An), and biotite. Smaller

grain size is typical at thè contact with thè host rock.

This thin aplitic portion contains scarce plagioclase in

anhedral spots and as rare small euhedral crystals,

moderately Ca-rich, similar to those of thè aplitic

footwall unit. Albitisation of K-feldspar is wide¬

spread. Tourmaline is present both as cavity-grown

and fracture-filling crystals.

The peripheral portions are fine to medium-

grained with irregularly dispersed pods, spots and

veins of pegmatitic grain size. Pockets rarely reach 1

dm3 in volume. Monzogranite xenoliths, and locally

also thè monzogranite portions beneath thè peg¬

matite veins, display moderate to complete tourma-

linisation. In this process, feldspars and biotite are

completely replaced by tourmaline crystal aggre-

gates, generating masses of tourmaline and quartz.

The contact of aplitic-pegmatitic dike with mon¬

zogranite is sharp. The oscillatory zoned plagioclase

and K-feldspar crystals of thè monzogranite are cut

by thè dike, and display well-developed syntaxic

overgrowth in thè endocontact. A progressive com-

positional zoning (25-10% An), which contrasts with

thè oscillatory zoning (40-12% An) of thè crystals of

thè monzogranite, characterises thè plagioclase over¬

growth. In a 1-2 cm thick band of thè exocontact, bi¬

otite displays a slight alteration, and is locally rimmed

by a thin overgrowth of white mica (probably zin-

nwaldite; C. Aurisicchio, pers. comm.). In thè same

area, myrmekite is abundant, in particular along thè

Fig. 4 - Internai structure, tourmaline colours, and accessory phase distribution in thè NI Dike. Dike section perpendicular

to dip and parallel to strike. _ _

_ I aplite; E9 pegmatite;

■j£L-

porphyric monzogranite;

bnr

tourmalinized monzogranite;

miarolitic cavities; I^JlI quartz-tourmaline graphic aggregates; tourmaline sprays locally degenerating in comò

texture; ÉZI biotite; I ^ 1 lepidolite; SD petalite; USI pockets hydrothermal clays.

lexiure, I— : 1 uiuinc, « — 1 - - r ^ _ _ ,, r\ —

Tourmaline colors: B = black; BN = black needles; C = colorless; G = green; GS = green sprays; Y - yellow; O - orange,

P = pink; PB = pale blue; PP = pale pink; PU = purple; R = red.

36

FEDERICO PEZZOTTA

Fig. 5 - (a) NI Dike in a vertical section, perpendicular to strike; (b) N2 Dike in a vertical section, perpendicular to strike.

The insert has thè same legend as figure 4.

dike contact. These phenomena, associated with al-

bitisation of K-feldspar, are mudi more abundant in

thè monzogranite of thè upper exocontact of thè cen¬

trai portion than in other parts of thè contacts.

A vertical section of NI dike, perpendicular to thè

strike (Fig. 5a), reveals a progressive thinning of thè

pegmatitic core. The aplitic and aplitic-pegmatitic

units suddenly lose their characters when granitic en-

claves occur in thè dike, indicating thè limit of thè

centrai portion and thè transition to thè peripheral

portion.

The Rosina Dike (N2 dike)

The N2 dike is similar to thè NI dike but is larger

(10 m long and 2 m thick), and displays a complex

structure of thè centrai portion, which is a multiple

repetition of thè more simpler NI structure.

A vertical section of thè N2 dike, perpendicular to

thè strike (Fig. 5b), displays a sharp irregular contact

with thè monzogranite at thè hanging-wall contact; a

network of aplitic veins characterises thè floor. The

dike, from thè floor to thè hanging wall, can be divid-

ed into a lower unit (up to 80 cm thick), mainly

aplitic, and an upper unit (up to 1.20 cm thick) con-

sisting of 2 or even 3 repetitions of “aplitic footwall

unit-pegmatitic hangingwall unit” pairs.

The massive aplitic lower unit is locally roughly

stratified and contains xenolits of monzogranite, of

biotitic metasedimentary rock and of stratified aplite.

Pods and fracture-filling pegmatitic veins also are

present. In September 1993, a single subspherical

large pocket (about 75 dm3 in volume) was found in

a pegmatitic pod completely hosted in this aplitic

unit.

In thè upper unit of thè N2 dike, abundant small

to medium size pockets (up to 20 dm3 in volume) are

developed in a single “aplitic footwall unit-pegmatitic

hangingwall unit” pair. A series of medium to large

pockets (up to 80 dm3 in volume) resides in a core re-

gion derived from thè coalescence of many “aplitic

footwall unit-pegmatitic hangingwall unit” pair. A pe-

culiarity of thè core region is thè presence of sekani-

naite in micrographic intergrowth with quartz (Or¬

landi and Pezzotta, 1993), which is abundant along

thè contact of aplite and pegmatite, below thè pocket

horizon. Individuai “aplitic footwall unit-pegmatitic

hangingwall unit” pairs are frequently bordered by

fracture-like planes filled by elongated, multicenti-

metric, biotite crystals. The grain size of thè “core re¬

gion” can be exceptionally coarse, and at larger pock¬

ets perthitic K-feldspar with graphic quartz as well as

quartz masses can attain 30 cm in diameter. Tourma-

line crystals are up to 40 x 4.5 cm in size.

A vertical NW-SE fracture crosscuts thè dike (and

also some pockets) inducing partial alteration and re-

crystallisation of feldspars and quartz.

Pocket parageneses

The pocket parageneses of thè dikes of this group

range from primitive to highly evolved. The greatest

variability of pocket mineralogy is characteristic of

thè peripheral portions. The large pockets of thè cen¬

trai zone show relatively uniform parageneses of in¬

termediate diversification.

The colours of tourmaline crystals and thè distrib-

ution of thè associated accessory phases in pockets of

thè NI dike are shown in figure 4. Cavities with black

tourmaline in peripheral portions are only indicated

where this minerai is abundant. Moreover, rare, prim¬

itive small to minute pockets also are widespread in

thè rest of thè peripheral portions as well as in thè

aplite-pegmatite unit of thè centrai portion. In cen¬

trai and peripheral portions, in an area adjacent to

thè “main pocket”, an overgrowth of green radiating

tourmaline needles on black crystals can be ob-

served.The primitive assemblage of accessory phases

(indicated with -a- in Fig. 4) displays a paragenetic se-

quence similar to thè simple one described in figure

3. The succession of colours in thè polychrome tour¬

maline (from black, to green, yellow, and pink) indi-

cates a high degree of evolution in some pockets of

thè peripheral zones. All tourmaline crystals in thè

INTERNAL STRUCTURES. PARAGENESES AND CLASSIFICATION OF THE MIAROLITICI L-BEARING COMPLEX PEGMATITES

37

pockets of thè pegmatitic core have similar colours,

indicating an intermediate degree of evolution. In thè

pockets of thè centrai portions (e.g., NI Dike Main

Pocket), differentiation strongly concentrates minor

elements including Li, B, and Mn as well as trace ele-

ments such as Ta, Cs, Ti, W, U, and Be.

Late stage Zeolites

massive pegm. Pocket minerai hydrothermal

Minerals growth growth stage

Early Late

Fig. 6 - Paragenetic sequence observed in thè large, aplite-hosted

pocket found in thè N2 dike. (1) Black tourmaline; (2) Yellow-

brown tourmaline; (3) dark brown tourmaline; (4) dark purple

tourmaline; (5) pale blue beryl; (6) colourless beryl. Bold line: so-

lution-alteration.

Figures 6, 7, 8, and 9 illustrate progressive evolu¬

tion of thè paragenetic sequences in representative

pockets of thè NI and N2 dikes.

The moderately primitive paragenetic sequence

reported in figure 6 characterises thè large aplite-

hosted pocket found in thè N2 Dike. The crystals of

K-feldspar, albite, quartz, and tourmaline were

deeply rooted to thè pocket walls, and K-feldspar was

mainly concentrated at thè roof of thè pocket, where-

as albite dominates on thè floor. Garnet crystals (0.2

to 1.0 cm across, progressively zoned; 93-97% Sps, 7-

3% Alm) are present in a large number in cavities.

Moreover, preliminary analytical data indicate that

Late stage Zeolites Second Late

massive pegm. Pocket minerai Pocket Post-rupture hydrothermal pocket hydroth.

Minerals growth . growth rupture growth stage rupture stage

Early Late

Fig. 7 - Paragenetic sequence observed in thè pocket System in thè

pegmatite core of thè N2 Dike. (1) Probable beginning of thè re¬

verse alkali differentiation; (2) crystallisation mainly of albite, par-

tially coating tourmaline and other previously crystallised miner¬

als; (3) black to dark brown tourmaline; (4) narrow overgrowth

and fracture filling of green tourmaline; (5) Pale blue-colourless

beryl; (6) Pale blue-colourless beryl of a new generation; (7) milky

quartz overgrowth, abundant nucleation of minute crystals, and

replacement of zeolites; (8) abundant albite and glassy-colourless

K-feldspar, nucleation and overgrowth on older crystals; (9) tour¬

maline overgrowth and nucleation of black needles. Bold line: so-

lution-alteration.

thè Mn content of tourmaline crystals from this cavi-

ty progressively increased during crystallisation to

7.5% MnO (a common feature of thè Elba tourma-

lines; Aurisicchio et al. , in progress) and dropped

abruptly to less than 0.2% when thè nucleation of

garnet began.

A moderately more evolved paragenetic sequence

is observed in thè series of partially interconnected

large cavities, in thè core of thè N2 Dike (Fig. 7). The

evolution of these pockets includes:

(a) a moderate minerai segregation associated

with a reverse alkali distribution similar to that de-

scribed by Stern et al. (1986, page 420) in a pegmatite

from California (albite, tourmaline and beryl prefer-

entially along thè roof of thè pocket; K-feldspar is

more abundant on thè floor of thè cavity, in many

crystals of small size, whereas in thè roof it composes

rare but large crystals);

(b) thè presence of pollucite and of a late over¬

growth of green tourmaline;

(c) thè presence of moderate amount of helvite in

tetrahedral crystals (up to 4 cm across, now largely al-

tered and dissolved; Pezzotta, 1994);

(d) crystallisation of garnet at primitive stages of

evolution of thè pockets;

(e) pocket rupture during thè minerai growth

which caused a partial collapse of thè roof of thè cav¬

ity;

(f) a new generation of beryl after thè pocket rup¬

ture.

In some cavities intersected by a fracture cross-

cutting thè dike (belonging to thè same fracture Sys¬

tem as shown in Fig. 2) a second pocket rupture oc-

curs. In these cavities a second hydrothermal event is

associated with a new generation of milky quartz, al¬

bite, K-feldspar, and black tourmaline (with syntaxic

overgrowth on older crystals and abundant new nu¬

cleation), and a partial dissolution of beryl, and an al-

most complete dissolution of petalite, pollucite,

helvite, zeolites, and hyalite. Quartz and albite micro-

crystals, mainly concentrated on thè floor of thè

pockets, replace locally thè zeolites.

Figure 8 shows thè paragenetic sequence observed

in thè main pocket of thè NI Dike. Compared to thè

pockets illustrated in figure 7, thè more evolved par-

Late stage

massive pegm

Minerals growth

Pocket minerà! Pocket

growth rupture

Zeolites

Post-rupture hydrothermal

growth stage

Quartz

Feldspars

Tourmaline

Beryl

Lepidolite

Garnet

Cassiterite

Nb-Ta Oxides

Zircon

Petalite

Pollucite

Hydroth. Clays

Zeolites

Amorph.Silica

}

Early

Late

Fig. 8 - Paragenetic sequence observed in thè main pocket of thè

NI Dike. (1) Probable starting of thè reverse alkali differentia¬

tion; (2) crystallisation mainly of albite; (3) black tourmaline; (4)

yellow-green tourmaline; (5) pale blue beryl; (6) colourless to

pink beryl; (7) pink beryl of new generation; (8) second genera¬

tion of lepidolite. Bold line: solution-alteration.

38

FEDERICO PEZZOTTA

agenesis of this cavity is indicateci by thè presence of

lepidolite and green to yellow tourmaline. Other re-

markable differences are demonstrated by a more

evolved minerai segregation and reverse alleali distri-

bution developed since thè first stages of pocket min¬

erai growth (with lepidolite preferentially spread on

thè floor of thè pocket), by crystallisation of mainly

albite among feldspars, during thè post-rupture min¬

erai growth, covering thè previous crystallised miner¬

ai with a 1 - 2 cm thick layer, and by thè crystallisa¬

tion, during thè post-rupture growth, of a new gener¬

ation of lepidolite in small masses of minute pink

crystals, distributed on thè floor of thè pocket (local-

ly replacing petalite) and encrusting larger, pre-exist-

ing minerai aggregates.

Minerals

Late stage Zeolites

massive pegm. Pocket minerai Pocket Post-rupture hydrothermal

growth growth rupture growth stage

Earlv

Late

Fig. 9 - Paragenetic sequence observed in highly evolved pockets

in thè N2 Dike. (1) Black tourmaline; (2) green to pink tourma¬

line; (3) growth of floating, doubly terminated tourmaline “pen-

cils”; (4) abundant new nucleation of tourmaline with overgrowth

on previous crystals; (5) colourless beryl; (6) pink beryl; (7)

colourless to pale pink beryl of a new generation; (8) lepidolite

partially pseudomorphous after pink tourmaline, K-feldspar, and

petalite. Bold line: solution-alteration.

Eariy

crystallizjed

minerals

Late

minerals

Fig. 10 - Associations of oxide and wolframate minerals in thè

studied pegmatites, depending on thè degree of evolution of thè

mineralogy in thè pockets. (3), (6), (7), (8), and (9), indicate thè as-

sociation of oxides and wolframates observed in thè pockets re-

ported respectively in figure 3, 6, 7, 8 and 9. (*) indicates an un-

certain position of thè minerai in thè diagram, because of limited

analytical and paragenetic data reported in thè literature and ob-

tained by thè author.

(2) (6) (7) (8) (9)

Primitive Highly evolved

pockets pockets

Figure 10 details thè association of oxide and wol¬

framate minerals in these pegmatites, depending on

thè evolution of thè mineralogy of thè pockets. These

minerals can be found intergrown mutually, locally in

epitaxial aggregates, and associated with zircon, mon¬

azite and xenotime.

The study of miarolitic cavities from other Li-

bearing dikes with complex asymmetric zoning, in thè

S.Piero and S.Ilario areas, indicates minerai assem-

blages and pocket parageneses dose to those de-

scribed for thè NI and N2 dikes. Even if garnet

and/or petalite and pollucite are locally rare (up to

absent), no significant variation of thè paragenesis of

thè other minerals can be observed. It is noteworthy

that pocket albite, in thè NI and N2 dikes as well as

in all thè other observed dikes, occurs in white to pale

azure stocky crystals and curved aggregates (over-

grown on graphic intergrowths of albite plus quartz)

but never develops clevelandite morphology.

3) Li-bearing dikes with simple asymmetric zoning

The paragenetic sequence, observed in thè highly

evolved miarolitic cavities (of moderately small size)

in thè peripheral parts of NI and N2 dikes, is report¬

ed in figure 9. Remarkable differences from thè par¬

agenetic sequence of thè Main Pocket of thè NI Dike

are thè absence of minerai segregation, a complex

history of tourmaline growth, thè more precocious

crystallisation of garnet, and thè abundant presence

of a second generation of lepidolite, in large aggre¬

gates of millimetric crystals, partially replacing pink

tourmaline, K-feldspar and petalite. K-feldspar in

these pockets becomes grey to pink-red in colour.

The change of thè tourmaline colour from black to

green, and locally also from green to pink, occurred

before thè pocket growth of crystals. Before thè

pocket rupture, rare, floating, doubly terminated

tourmaline “pencils” crystallised, usually pink in

colour. After thè pocket rupture, a locally abundant

new tourmaline generation (green to pink) produced

thousands minute doubly terminated needles. These

submillimetric tourmalines, locally together with

minute crystals of albite and mica, coated all thè oth¬

er minerals as a “snow on thè roof”, reflecting a near-

ly stagnant parent medium.

The miarolitic dikes of this group are characterised

by Li-bearing accessory phases, and by a simple asym¬

metric zoning. The grain size, thè abundance and thè

textures of rock-forming minerals, change in thè direc¬

tion perpendicular to thè walls. The pockets paragene¬

ses, thè tourmaline colours, and thè minerai assem-

blages, usually highly evolved, are similar (or were

identical) in all thè cavities of thè same dike. Consid-

ering old samples and thè outerops in abandoned

quarries, it is possible to assume that at least a part of

thè dikes of thè famous localities of Masso Foresi, Fac-

ciatoia, Catri, Fosso Gorgolinato, and Fosso dei For-

cioni (Millosevich, 1914) belong to this type.

These dikes are hosted in thè porphyritic monzo-

granite dose to thè contact with thè country rocks, in

thè E-dipping aplite (locally deuterised and with ac¬

cessory tourmaline) and leucogranite dikes, and in

thè thermometamorphic (extensively hydrothermally

altered) mafie rocks of thè contact aureole. The peg¬

matites have a fiat shape and are often intruded “en

echelon”. Low-angle intersections of two or three

dikes can be locally observed. The dikes are small:

they are 6-8 meters long and rarely exceed thè thick-

ness of 25 centimetres. Miarolitic cavities are locally

INTERNAL STRUCTURES. PARAGENESES AND CLASSIFICATION OF THE MIAROLITICI L-BEARING COMPLEX PEGMATITES

39

abundant, and thè largest ones (exceptionally up to

120 x 80 x 25 centimetres) have been observed in dike

intersections. Along thè exocontacts thè same miner¬

ai alterations occur as described for thè second cate-

gory of dikes.

Two kinds of Li-bearing dikes with asymmetric zona-

tion have been distinguished depending on minerai tex-

tures:(l) a “Facciatoia Type”, and (2) a “Catri Type”.

On a small scale, thè intrusions of thè Facciatoia

Type have structures and minerai textures similar to

those described for thè centrai portion of thè dikes of

thè second category. Indeed, in a dike 15-20 cm

across, an aplitic footwall unit transitional to a peg-

matitic-miarolitic zone, and an aplitic-pegmatitic

hangingwall unit can be observed. Tourmaline (as

graphic intergrowth with quartz) and lepidolite (as

isolated centimetric primary crystals, and as micro-

crystalline small masses of late replacive origin on

petalite, pollucite, and pink tourmaline, and probably

also feldspars and quartz) are abundant along thè

pocket horizon.

Altered

leucogranite

Layered

aplite

X

* * *

: •f-vj.y-nìry

•••• . 'V i

v..

* • * • * • ' «4

fp

X *

X X

X X

X X X x

X X X »

X X

X

'••un

^iiiiinninn,.

K X

X X

K

ini i nit li i inn

a x x x

A A * *

X * X x

X K X X X X

a x x x x x

X X X X ^ X

20 cm

I - 1

Fig. 1 1 - Internai structure of a Catri Type dike, with multiple

aplite-pegmatite layers. Dike section perpendicular to dip and

parallel to strike. Note thè pocket with tourmaline crystals prefer-

entially lining thè roof, and with large altered petalite crystals, in

thè lower part of thè figure.

The dikes of thè Catri Type are stratified, and

characterised by texturally and paragenetically segre-

gated layered aplitic and pegmatitic units (Fig. 11). A

Constant feature is thè presence of tourmaline crys¬

tals, grown perpendicularly to thè walls downwards

(comb texture). Sodium-enriched aplite has sugary

appearance and is composed of albitic plagioclase (7-

2% An), quartz, rare K-feldspar, and accessory white

to pink micas in small fiat crystals, rare biotite, and

sekaninaite. In contrast, pegmatitic (1 to 4 cm in grain

size), potassium-enriched units are composed of K-

feldspar, quartz, scarce albitic plagioclase, tourma¬

line, and lepidolite. Albitic plagioclase is mainly con-

centrated below thè pocket horizon. K-feldspar is

perthitic and commonly deeply albitised; crystals

grow perpendicular to thè roof, and have a graphic

intergrowth with quartz at thè root. At thè base of

comb tourmaline crystals, a tourmaline-quartz graph¬

ic symplectite rapidly turns into a single tourmaline

crystal with a quartz core, and finally into a quartz-

free tourmaline prism. In subvertical dikes, tourma¬

line comb texture can be developed concentrically,

from both walls inwards.

The rare dikes hosted in extensively hydrother-

mally altered mafie rocks have a reaction aureole in

thè exocontact, locally rich in radiating crystals of

black tourmaline. These intrusions, mainly of peg¬

matitic nature, are intermediate between Facciatoia

and Catri Type, and have tourmaline in graphic inter¬

growth with quartz, as well as in radiating crystals

and/or in comb texture.

Minerai assemblages and pocket paragenesis of

thè III category of dikes are similar to thè highly

evolved ones illustrated in figure 9. Remarkable dif-

ferences are in thè presence of a strong minerai seg-

regation, particularly developed in thè Catri Type

(more than 90% of tourmaline crystals are lined on

thè pockets roof), locally associated with a moderate

“inverse” alkali differentiation; thè absence of evi-

dent pocket ruptures during thè pocket crystallisa-

tion; and a single-phase tourmaline crystal growth. In

some pockets a late fluid circulation occurred, similar

to thè one described in figure 7 (after thè second

pocket rupture). This process, which can be only

weakly developed, is associated with late microfrac-

tures (thè same System as shown in Fig. 2), and is re¬

sponsive for thè crystallisation of a “black hat” on

thè top of tourmaline crystals.

Kit. Capanne

monzogranite

Pluton

margin

Thermometamorphic

aureole with leucogranitic

and aplitic intrusions

Thermometamorphic

rocks

Dikes of group 1

Dikes of group 2

Dikes of group 3

Dikes of group 4

_ Incrasing dislance from thè centre of thè pluton - ►

Fig. 12 - Qualitative diagram of thè zoned distribution of Elba

pegmatites.

40

FEDERICO PEZZOTTA

4) Irregularly zoned to unzoned Li-bearing dikcs

The Li-bearing dikes of this category are irregu¬

larly zoned to unzoned, and have a random distribu-

tion of accessory phases. The host rocks are thè hy-

drothermally altered mafie rocks and thè metasedi-

mentary hornfels of thè thermometamorphic aureole

of thè pluton. These dikes can have irregular shapes

and changes in attitude. The thickness ranges from a

few centimetres up to 2 meters and thè length can at-

tain 6-7 meters. Black tourmaline and white micas

can occur along thè exocontacts together with hy-

drothermal alteration produets.

The internai structure of these dikes ranges from

irregularly distributed masses of aplitic and peg-

matitic rocks, with locai concentrations of tourmaline

and white to pink micas, to a medium-grained rock

with homogeneous distribution of minerals. In thè

pegmatitic masses, tourmaline occurs locally in radi-

ating groups or comb textures, as well as randomly

oriented prismatic crystals. Tourmaline is normally

black, but it is occasionally present as polychrome

crystals with black to green and pink colour. White to

pink micas, locally abundant, can be concentrated in

granular masses many decimetres across, in contact

with fine-grained spongy albite aggregates, and en-

closing green to pink, partially corroded tourmaline

crystals. Plagioclase is always of albitic composition.

Miarolitic cavities are rare, and absent in some dikes.

The pockets are concentrated dose to areas with

tourmaline in comb texture or in spongy albite. The

minerai assemblage is simple; only small crystals of

polychrome tourmaline, white to pink micas, and oc¬

casionally pink apatite, were found as accessory phas¬

es. Some multicentimetric dissolution cavities have a

partially polyhedral shape and are associated with

pink tourmaline and micas; they might have originat-

ed after petalite or pollucite crystals. In dikes hosted

in metabasic rocks, thè quartz of thè graphic K-

feldspar along thè walls is occasionally leached, prob-

ably by supergene Mg-rich Solutions (Cerny and

Novàk, 1992; P. Cerny, pers. comm. 1994).

Dikes with transitional characteristics

Dikes with characteristics transitional between

those of thè four categories described above are not

uncommon. The observed variations are:

(a) Dikes transitional between category 1 and cat¬

egory 2: intrusions with black tourmaline but lacking

Li-bearing minerals, and locally with a rough asym-

metrical structure; mainly aplitic dikes with irregular¬

ly distributed pegmatitic pods and veins, with a few

small miarolitic cavities with green to yellow tourma¬

line.

(b) Dikes transitional between category 2 and cat¬

egory 3. Masses of medium to small size with a large

centrai portion, and small peripheral portions.

(c) Dikes intermediate between group 3 and

group 4. Irregularly zoned intrusions, up to 2 meters

across, hosted in metabasic rocks, having portions

with tourmaline comb texture and with layered

aplite, together with rare but highly evolved pockets.

(d) Dikes intermediate between group 1 and 4. In-

trusions hosted in thè hornfels, dose to thè granodi-

orite contact, similar to those of group 4 but without

Li-bearing minerals.

Discussion

The zoned pattern of occurrence of dikes

Abundant literature on rare-element-bearing peg-

matites reports that differentiation during thè evolu-

tion of pegmatite magma leads to a progressive en-

richment of minor and trace elements such as F, Li, B,

Nb, Ta, Rb, Cs, W, Sb, Bi, U, Ce, La, Sn, Be, in residuai

melt (e.g. Walker et al. , 1986; London, 1990). The phe-

nomena related to thè progressive differentiation of

a pegmatite magma include: changes in composition

of minerals such as garnet (from almandine-rich to

spessartine-rich; Cerny & Hawthorne, 1982), tourma¬

line (from schorl to elbaite; e.g. Jolliff et al. , 1986), and

micas from biotite to muscovite to lepidolite (e.g.

Hawthorne & Cerny, 1982, Cerny et al, 1995); in-

crease in Li, Cs and decrease in Fe content in beryl

(e.g. Burt, 1982); occurrence of minerals such as pol¬

lucite and microlite in thè most evolved rocks. Com-

pared to these observations, thè four groups of mi¬

arolitic Elba pegmatites can be considered produets

of crystallisation of residuai melts with variable de-

grees of evolution. Residuai melts ranged from prim¬

itive (pegmatites of group 1) through moderately

evolved (pegmatites of group 2) to highly evolved

(pegmatites of groups 3 and 4).

Figure 12 qualitatively highlights thè zoned pat¬

tern of occurrence of thè studied pegmatites. Al-

though an overlap is evident, dikes without Li-miner-

als are mainly hosted in thè monzogranite, whereas

thè others (groups 2, 3, and 4) are distributed at in-

creasing distance, from thè centre of thè pluton, host¬

ed in thè monzogranite, in thè aplitic and

leucogranitic lenses, and in thè hornfels of thè ther¬

mometamorphic aureole. Similar zoned patterns of

pegmatites are well known in thè literature (e.g.True-

man and Cerny, 1982), even if descriptions of zoned

fields of miarolitic dikes are rare. The regional zoning

of pegmatites is explained (Cerny, 1982b, 1991b,

1992) by differences in thè contents of volatile and

other liquidus- and viscosity-depressing constituents

(Li, F, H20, B, P; Luth, 1976; Chorlton and Martin,

1978) among thè individuai batches of separated peg¬

matite melts. These differences result in intrusions

with variable thermal stability, reflected in diversified

distances of pegmatite migration. At Elba, thè narrow

contact aureole, thè presence of abundant aplite and

leucogranite apophyses, and thè partially convective

mechanism of heat transfer (manifested by thè pres¬

ence of irregularly distributed magnesite deposits of

thermometamorphic origin in metabasic rocks) gen-

erated a steep and irregular heat gradient around thè

Mt. Capanne pluton that could be responsible for thè

partial overlapping distribution of thè four groups of

Elba pegmatites.

Evolutionary characteristics

Based on thè minerai parageneses, rock textures

and structures of thè dikes, thè examined Elba peg¬

matites can be considered members of thè rare-ele-

ment class and of thè LCT petrogenetic family in thè

current classification of Cerny (1991). In addition, thè

shallow level of intrusion (indicateti by thè data re-

ported by Ruggeri and Lattanzi (1992) and by thè

INTERNAL STRUCTURES. PARAGENESES AND CLASSIFICATION OF THE MIAROLITICI L-BEARING COMPLEX PEGMATITES

41

abundant presence of miarolitic cavities) and some

structural features are similar to those of thè mi¬

arolitic class (cf. Cerny, 2000, about problems plagu-

ing thè definition of this class).

The compositional and textural asymmetric zon-

ing of thè dikes of category 2 and 3 are typical of thè

miarolitic gem-bearing pegmatites of many localities

in thè world (e.g. Cerny, 1982b). In particular, thè

dikes of thè Catri type (category 3), and thè centrai

portion of thè dikes of category 2, display on small

scale rock structures and minerai textures similar to

those described by London (1986) and Stern et al.

(1986) respectively for thè Himalaya dike and thè

Little Three dikes in thè San Diego County, Califor¬

nia.

The bodies of group 2 are of particular interest.

Indeed, whereas variations of accessory phase assem-

blages are common and well known in thè literature

on massive pegmatites, thè presence of primitive to

highly evolved parageneses in thè pockets belonging

to different zones of one and thè same dike is unusu-

al in miarolitic pegmatites. The complex three-dimen-

sional structure of these intrusions documents a com¬

plex and heterogeneous internai evolution.

The presence of miarolitic cavities, which consti-

tute evidence of vapour saturation prior to thè com¬

plete crystallisation of pegmatite-forming melt, ap-

pears to be controlled by thè timing and extent of

volatile exolution from silicate melt (London, 1986).

The pocket parageneses reported in this paper docu-

ment a crystallisation history for thè miarolitic cavi¬

ties of Elba pegmatites similar to that reported for

analogous pegmatites from other localities (e.g. Fo-

ord et al. , 1986; London, 1986; Stern et al., 1986). In

particular, thè main phenomena observed in thè Elba

dikes can be summarised as follows: 1) More evolved

cavities have a much more complicated pocket para-

genesis and history of crystallisation. 2) The first

pocket rupture commonly observed in thè largest

cavities of dikes of group 2 generated an abrupt

change in thè chemical-physical conditions of crys¬

tallisation, testified by new nucleation and new mor-

phology of tourmaline, beryl and mica crystals. 3) The

garnet precipitation seems to control at least thè Mn

content in tourmaline. More evolved pockets show

garnet crystallisation at earlier stages. 4) The crys¬

tallisation in pockets occurred essentially in a closed

System, up to thè transition to thè hydrothermal zeo-

lite stage. Deviations from this behaviour can occur

during thè pocket rupture and possibly also during

thè latest stages of pocket crystallisation, with thè es-

cape of significant amounts of fluids. These phenom¬

ena are documented by changes in thè crystallisation

conditions (new nucleation of tourmaline, beryl and

micas), by abrupt changes of colours in tourmaline

crystals, by thè presence of overgrowth of coloured

tourmaline in fractures adjacent to thè largest pock¬

ets, and by thè presence of extensive albitisation of K-

feldspar along thè fractures radiating around thè cav¬

ities. 5) Albite never occurs as cleavelandite, in con-

trast to many other similar dikes described in thè lit¬

erature from other localities. 6) Minerai segregation

with tourmaline and beryl mainly confined to thè

roof of thè pocket, and lepidolite to thè floor, was ob¬

served in thè largest cavities of thè second-category

dikes and in almost all thè cavities of thè dikes of thè

third category. However, an inverse alkali differenti-

ation, similar to that described by Stern et al. (1986)

in a California pegmatite, has been observed in some

of thè largest pockets. 7) The formation of zeolites, to-

gether with partial alteration of feldspars, corrosion

of petalite, pollucite and other minerals, and deposi-

tion of hydrothermal clays, is typical of highly

evolved pockets, but is absent in primitive cavities. 8)

At least one late hydrothermal stage is associated

with a System of late fractures, which locally crosscut

thè dikes all over thè studied area. This widespread

process opened thè pocket System and generated par¬

tial collapse of thè crystals, deep corrosion of petalite,

pollucite, beryl, and helvite, locai corrosion of

coloured tourmaline, replacement of zeolites by al¬

bite and quartz microcrystals, and a new generation

of quartz, albite, K-feldspar, and black tourmaline de-

posited over all thè other minerals. This process is re¬

sponsive of thè black “hat” on tourmaline crystals

occurring in some dikes.

Pegmatite - host rock interactions

The exsolution of fluids during thè crystallisation

of a pegmatitic melt can result in thè formation of

dispersion halos around thè pegmatite intrusions. The

alterations of thè host rocks are of metasomatic na¬

ture and include generation of secondary minerai as-

semblages and modification of thè originai composi-

tion of primary minerals (e.g. Shearer et al., 1986;

Morgan and London, 1987). The influx of fluids from

thè wall rock to thè pegmatite is difficult to ascertain

and is not well documented. Some mineralogical and

fluid inclusion data indicate that wall-rock-derived

fluids may infiltrate pegmatite before thè cessation of

internai subsolidus recrystallisation (London, 1990).

In thè Elba pegmatites, escape of fluids is evident

in thè dikes of thè second and third category (hosted

in monzogranite, aplites and leucogranites) by thè

presence of locai tourmalinisation, albitisation of K-

feldspar, biotite alteration, and possibly myrmekite

growth along contacts.The interactions between peg¬

matite and wallrock were probably much more im-

portant in dikes intruded in thè metabasic and

metasedimentary rocks. The data reported in this

work allow thè following considerations: 1) The ser-

pentinites which host pegmatitic dikes are always

deeply altered, with abundant magnesite veins, and

occasionally narrow feldspatic and tourmaline-rich

veins. 2) Pegmatites hosted in sepentinites show tour¬

maline veins in thè exocontact, and an endocontact

consisting of micrographic quartz-K-feldspar aggre-

gates. These features are absent in dikes hosted in in¬

trusive rocks. 3) In serpentinites thè dikes of catego¬

ry 2 and 3 are rare if not exceptional, whereas thè

dikes of category 4 are common. In metasedimentary

rocks only dikes of thè fourth category are present. 4)

In pegmatites hosted in metasedimentary rocks, mi¬

cas and tourmaline can be abundant along thè dike

margins. 5) The dikes of category 4 are poor in, or to-

tally lacking of miarolitic-cavities. 6) Dikes with

primitive mineralogical composition hosted in ther-

mometamorphic rocks (see dikes intermediate be¬

tween category 1 and 4) display rock structures and

minerai textures identical to those of thè dikes of cat¬

egory 4.

42

FEDERICO PEZZOTTA

CONCLUSIONS

Restricted as they are in area distribution and in

size of individuai bodies, thè Elba pegmatites display

a great variety of structures and complex minerai

parageneses. They have been classified into 4 cate-

gories with progressive degree of evolution, distin-

guished on thè basis of internai structures, minerai

textures, abundance and distribution of rock-forming

minerals and accessory phases, tourmaline colour,

and pocket parageneses. Across thè narrow contact

aureole, thè dikes have a roughly zoned distribution

with more evolved bodies emplaced at farther dis-

tance from thè margin of thè pluton. The internai

zoning of thè pegmatites richer in miarolitic cavities

(groups 2 and 3) is similar to those of many Califor¬

nia miarolitic gem-bearing pegmatites, well known in

thè literature. The pocket parageneses range from

simple, in primitive cavities, to complex in highly

evolved cavities, with a crystallisation history includ-

ing minerai segregation, pocket ruptures, probable

openings of thè chemical-physical System, and late-

stage hydrothermal fluids circulation. Dikes em¬

placed in thermometamorphic rocks probably lost a

large amount of thè fluids exsolved during thè crys-

tallisation thè host rocks, and developed an irregular

zoning with only occasionai miarolitic cavities.

Acknowledgements

I wish to thank thè Consiglio Nazionale delle

Ricerche (Italy) which permitted thè use of thè ARL

SEMQ electron microprobe, in thè laboratories of

thè Dipartimento di Scienze della Terra of thè Uni¬

versity of Milano; thè Dipartimento di Chimica Strut¬

turale e Stereochimica Inorganica of thè University

of Milano which permitted SEM analysis. I also ac-

knowledge valuable contributions by thè following

colleagues: R Orlandi for some SEM minerai analysis

and for encouragement; G. Liborio for some powder

X-ray diffraction data, C. Aurisicchio, C.M. Gramac-

cioli and A. Boriani for profitable discussions. I am al¬

so grateful to thè minerai collector K. Wirt for much

valuable field information. A special thanks goes to P.

Cerny for encouragement during this study, for in-

spiring suggestions and for reviewing thè manuscript.

References

Aloisi P., 1920 - Il Monte Capanne. Tipogr. Succ. F. Nistri, Pisa, 303 pp.

Aurisicchio G, Conte A.M., Fioravanti G. and Freda G, 1992

- Il ruolo delle tormaline nell’evoluzione delle pegmatiti del¬

l’Isola d’Elba. L’appennino settentrionale. Società Italiana di

Mineralogia e Petrologia, Convegno “Minerogenesi Appen¬

ninica”, Firenze, 21-23 settembre 1992, Riassunti, 265.

Aurisicchio C., Conte A.M. and Cerny P., 1994 - Petrochemical

features of igneous rocks in thè rare-element pegmatite field

of Elba, Italy. International Mineralogical Association , 16th

General Meeting, Pisa, Italy, Abstracts, 21.

Aurisicchio C., Fioravanti G., Grubessi O. and Zanazzi P.F.,

1988 - Reappraisal of thè crystal chemistry of beryl. American

Mineralogist, 73: 826-837.

Aurisicchio C., Orlandi R, Pasero M. and Perchiazzi N., 1993

- Uranopolycrase, thè uranium-dominant analogue of poly-

crase-(y), a new minerai from Elba Island, Italy, and its crystal

structure. European Journal of Mineralogy, 5: 1161-1165.

Aurisicchio C., Ottolini L. and Pezzotta’F., 1999 - Electron-

and ion-microprobe analyses, and genetic inferences of tour-

malines of thè foitite-schorl solid solution, Elba Island (Italy).

European Journal of Mineralogy , 11: 217-225.

Burt D.M., 1982 - Minerals of beryllium. In: Cerny R (ed.). Short

course in granitic pegmatites in Science and industry. Miner¬

alogical association of Canada, 8: 135-146.

Bussy F., 1990 - Pétrogénèse des enclaves microgrenues associées

aux granitoi'des calco-alcalins: exemple des massifs varisques

du Mont-Blanc (Alpes occidentales) et miocène du Monte

Capanne (Ile d’Elbe, Italie). Mém. Géol. Lausanne, 7, 309 pp.

Carobbi G. and Rodolico E, 1976 - Minerali connessi alle rocce

dell’arcipelago. In: I minerali della Toscana. Olshki, Firenze:

47-76.

Cerny P, 1982a - Anatomy and classification of granitic peg¬

matites. In: Cerny P. (ed.). Short course in granitic pegmatites

in science and industry. Mineralogical association of Canada.

8: 1-40.

Cerny P, 1982b - Petrogenesis of granitic pegmatites. In: Cerny P.

(ed.). Short course in granitic pegmatites in science and in-

dustry. Mineralogical association of Canada. 8: 405-462.

Cerny R, 1991 a - Rare-element Granitic Pegmatites. Part I:

Anatomy and internai evolution of pegmatite deposits. Geo-

science Canada, 18: 49-67.

Cerny P, 1991a - Rare-element Granitic Pegmatites. Part II: Re-

gional to Global Environments and Petrogenesis. Geoscience

Canada, 18: 68-81.

Cerny P, 1992 - Regional zoning of pegmatite populations and its

interpretation. Vortrag vor der Osterreichischen Mineralogis-

chen Gesellschaft, 137: 99-107.

Cerny R, 2000 - Constitution, Petrology, Affiliations and Cate-

gories of Miarolitic Pegmatites. This volume.

Cerny P. and Hawthorne F.C., 1982 - Selected peraluminous

minerals. In: Cerny P. (ed.). Short course in granitic pegmatites

in science and industry. Mineralogical association of Canada,

8: 163-179.

Cerny P. and Novak M., 1992 - Locality No.2: Vezna near

Nedvedice, a pegmatite dike of thè beryl-columbite subtype

penetrating serpentinite. In: Lepidolite 200. International

Symposium on thè Mineralogy, Petrology and Geochemistry

of granitic pegmatites, Field Trip Guidebook. Novi Mesto na

Morave , Czechoslovakia: 27-32.

Chorlton L.B. and Martin R.F., 1978 - The effect of boron on

thè granite solidus. Canadian Mineralogist, 16: 239-244.

Ferrara G. and Tonarini S., 1985 - Radiometrie Geochronology

inTuscany: results and problems. Rendiconti Società Italiana di

Mineralogia e petrologia , 40: 111-124.

Foord E. E., 1976 - Mineralogy and petrogenesis of layered peg-

matite-aplite dikes in thè in thè Mesa Grande District, San

Diego County, California. Ph.D. thesis, Stanford University,

Stanford, California.

Foord E.E., Starkey FI.C. and Taggart Jr. J.E., 1986 - Mineral¬

ogy and paragenesis of “pocket” clays and associated minerals

in complex granitic pegmatites. San Diego County, California.

American Mineralogist,! 1: 428-439.

Hawthorne F.C. and Cerny P, 1992: The mica group. In: Cerny

P. (ed.). Short course in granitic pegmatites in science and in¬

dustry. Mineralogical association of Canada, 8: 63-91.

Jahns R.H., 1982 - Internai evolution of granitic pegmatites. In

Cerny P. (ed.). Short course in granitic pegmatites in science

and industry. Mineralogical association of Canada, 8: 293-328.

Jollif B.L., Papike J.J. and Shearer C.K., 1986 - Tourmaline as a

recorder of pegmatite evolution: Bob Ingersoll pegmatite,

Black Hills, South Dakota. American Mineralogist, 71: 472-500.

Keller V. A. and Pialli G., 1990 - Tectonics of thè Island of Elba:

a rappraisal. Bollettino della Società geologica Italiana , 109:

413-425.

London D., 1986 - Formation of tourmaline-rich gem pockets in

miarolitic pegmatites. American Mineralogist , 71: 396-405.

London D., 1990 - Internai differentiation of rare-element peg¬

matites. A synhesis of recent research. Geological Society of

America, Special Paper 246: 35-49.

London D., 1992 - The application of experimental petrology to

thè genesis and crystallization of granitic pegmatites. Canadi¬

an Mineralogist, 30: 499-540.

Luth W.C., 1976 - Granitic rocks. In: D.K. Bailey and R.A. Mac-

Donald Eds. The evolution of thè crystalline rocks. Acad.

Press N. Y., 484 pp.

Marinelli G., 1959 - Le intrusioni Terziarie dellTsola d’Elba. Atti

della Società Toscana di Scienze naturali Memorie, 66: 50-253.

Millosevich F., 1914 - I 5000 Elbani del Museo di Firenze. Con¬

tributo alla conoscenza della mineralogia dellTsola d’Elba.

Studio Editoriale Insubria , Reprint Milano 1978, 96 pp.

Morgan VI G.B. and London D., 1987 - Alteration of amphi-

bolitic wallrocks around thè Tanco rare-element pegmatite,

Bernic Lake, Manitoba. American Mineralogist ,72: 1097-1121.

Orlandi P, Pasero M., Perchiazzi N„ 1990: Nb-Ta oxides from

Elba island pegmatites. Atti Soc. Tose. Sci. Nat. Meni., 97: 161-

170.

INTERNAL STRUCTURES. PARAGENESES AND CLASSIFICATION OF THE MIAROLITICI L-BEARING COMPLEX PEGNI ATITES

43

Orlandi P. and Pezzotta F.. 1993 - La sekaninaite dei filoni peg-

matitici elbani. Atti Soc. Tose. Sci. Nat. Meni.. 100: 85-91.

Orlandi P. and Pezzotta F., 1996 - Minerali dell'Isola d’Elba, i

minerali dei giacimenti metalliferi dell'Elba orientale e delle

pegmatiti del M.te Capanne. Ed. Novecento Grafico, Berg¬

amo. 248 pp.

Pezzotta E, 1993a - Preliminary data on structure. petrography

and mineralogy of two pegmatitic dikes in Isola d'Elba (Italy).

Terra abstraets. Abstract supplement No. 1 to Terra Nova, 5,

528 pp.

Pezzotta E, 1993b - Osservazioni strutturali, petrografiche e

classificative sui filoni aplitico-pegmatitici litiniferi del settore

occidentale del M.te Capanne (Isola d’Elba). Plinius , 10: 208-

209.

Pezzotta E, 1994 - Helvite of a M.te Capanne pluton pegmatite

(Elba island, Italy): Chemical, X-ray diffraction data and de-

scription of thè occurrence. Rendiconti Fis. Accademia dei Lin¬

cei s. 9 (5): 355-362.

Pezzotta E, Howthorne E, Cooper M. A. and Teerstra D.,

1996 - Fibrous foitite from S. Piero in Campo, Elba, Italy.

Canadian Minaralogist, 34, 741-744.

Poli G., Manetti P. and Tommasioni S., 1989 - A petrological re-

view on Miocene-Pliocene intrusive rocks from southern Tus-

cany and Tyrrhenian sea (Italy). Periodico di Mineralogia, 58:

109-126.

Pupin J.P., 1976 - Signification des caractères morphologiques du

zircon commun des roches en pétrologie. Base de la méthode

typologique. Applications. Thèse Doct. Etat, Univ. Nice, 394 pp.

Pupin J.P, 1980 - Zircon and Granite Petrology. Contrib. Minerai

Petrol., 73: 207-220.

Ruggeri G. and Lattanzi L., 1992 - Fluid inclusion studies on

Mt. Capanne pegmatites. Isola d'Elba, Tuscany, Italy. Euro¬

pea) ‘i Journal of Mineralogy, 4: 1085-1096.

Shearer C.K., Papike J.J. and Laul J.C., 1985 - Chemistry of

potassium feldspars from three zoned pegmatites, Black Hills,

South Dakota: Implications concerning pegmatite evolution.

Geochimica et Cosmochimica Acta, 49: 663-673.

Shearer C.K., Papike J.J., Simon S.B. and Laul J.C., 1986 - Peg-

matite-wallrock interactions, Black Hills, South Dakota: Inter-

actions between pegmatite-derived fluids and quartz-mica

schist wallrock. American Mineralogist, 71: 518-539.

Sinkankas J., 1981 - Emerald and Other Beryls. Chilton Book Co.

Nelson Canada, Radnor Pa-Scarborough Ontario, 225pp.

Stern L.A., Brown G.E., Bird D.K., Jahns R.H., Foord E. E.,

Shigley J.E. and Spaulding Jr. L.B., 1986 - Mineralogy and

geochemical evolution of thè Little Three pegmatite-aplite

layered intrusive, Ramona, California. American Mineralogist,

71:406-427.

Teerstra D.K., Cerny P. and Hawthorne F.C., 1998 - Rubidium

fedlspars in granitic pegmatites. Canadian Mineralogist, 36:

483-496.

Trueman D.L. and Cernì P, 1982 - Exploration for rare-element

granitic pegmatites. In: Cerny P (ed.). Short course in granitic

pegmatites in Science and industry. Mineralogical association

of Canada, 8: 463-491.

Walker R.J., Hanson G.N., Papike J.J., O’Neil J.R. and Laul J.C.,

1986 - Internai evolution of thè Tin Mountain pegmatite, Black

Hills, South Dakota. American Mineralogist, 71: 440-459.

'

..

:

Memorie della Società Italiana di Scienze Naturali e del Museo Civico di Storia naturale di Milano

Volume XXX - pp.: 45-56, 2000

Compositional pathways of tourmaline evolution during

primary (magmatic) crystallization in complex (Li)

pegmatites of thè Moldanubicum, Czech Republic

Milan Novak*

Department of Mineralogy and Petrography, Moravian Museum, Zelny trh 6, 659 37 Brno, Czech Republic

Abstract -A comparative study of compositional trends of tourmaline in 16 individuai pegmatite dikes of thè complex

type in thè Moldanubicum, Czech Republic, yielded distinct trends characteristic for thè individuai pegmatite subtypes (lep-

idolite, elbaite and “masutomilite”). Their formation is discussed in thè respect to thè following factors: a crystal-chemical

factor, geochemical factor, element-partitioning factor and contamination factor. The compositional trends of thè Y-site

cations, X-site cations and fluorine are illustrated at several diagrams where decreasing (Fe + Mg + Ti) apfu was selected as

a suitable index (FMT index) of progressive crystallization (fractionation) in tourmaline.

The trends of thè Y-site cations (Ti, Mg, Fe, Zn, Mn, Al) are mostly dose to thè schematic diagram of cations covaria-

tion derived from thè cation field strengths. Deviations were found in behaviour of Mn in lepidolite pegmatites and Al in

masutomilite pegmatite at Kracovice. Concentration of thè Y-site cations seems to be controlled particularly by thè crystal-

chemical factor and geochemical factor; thè element-partitioning factor and contamination factor are very likely less im-

portant.

The X-site cations (Na, Ca) and F exhibit contrasting evolution trends in (i) Ca-poor and in (ii) Ca-rich tourmaline (Ca

> 0.2 apfu). (i) Two distinct compositional trends were observed in thè lepidolite pegmatites and in elbaite pegmatites. Dis-

regarding thè pegmatite subtypes, these tourmalines also exhibit very good positive correlation of Na versus F and negative

correlation of X-site vacancy versus F. (ii) The compositional trends exhibit a decrease in Ca from thè outermost zones to

centre; elevated Ca very likely reflects thè influence of Ca-rich host rocks. Increase in Ca during thè tail-end of primary crys¬

tallization indicates high Ca-activity of parent medium in late stages. The concentrations of Ca indicate elevated control of

thè geochemical factor and contamination factor; thè element-partitioning factor and crystal-chemical factor seem to have

played an important role particularly in Na and F concentrations of Ca-poor tourmaline.

Riassunto - Uno studio comparato dei trend composizionali delle tormaline in 16 diversi dicchi pegmatitici di tipo com¬

plesso della Moldavia, Repubblica Ceca, ha evidenziato trend distinti caratteristici per i diversi subtipi pegmatitici ( a lepi¬

dolite, elbaite e “masutomilite”). La loro formazione c stata discussa in base ai seguenti fattori: fattore cristallo-chimico, fat¬

tore geochimico, fattore di ripartizione degli elementi e fattore di contaminazione. I trend composizionali dei cationi del sito

Y, dei cationi del sito X e del fluoro sono illustrati in numerosi diagrammi verso la somma (Fe + Mg + Ti) apfu, considera¬

ta un indice adatto (indice FMT) ad indicate la progressiva cristallizzazione (frazionamento) nella tormalina. I trend dei

cationi del sito Y (Ti, Mg, Fe, Zn, Mn, Al) sono prevalentemente simili ai diagrammi di variazione dei cationi derivati dai

valori delle forze di campo. Deviazioni sono state osservate nel comportamento del Mn nelle pegmatiti a lepidolite e del-

l’Al nelle pegmatiti a masutomilite a Kracovic. Le concentrazioni dei cationi del sito Y sembrano essere particolarmente

controllate da fattori cristallochimici e fattori geochimici; i fattori di ripartizione degli elementi e di contaminazione sono

molto verosimilmente meno importanti.

I cationi del sito X (Na, Ca) ed il F mostrano andamenti evolutivi contrastanti nelle (i) tormaline povere in Ca e nelle

(ii) tormaline ricche in Ca (Ca > 0.2 apfu). (i) Due trend composizionali distinti sono stati osservati nelle pegmatiti a lepi¬

dolite e nelle pegmatiti ad elbaite. Indipendentemente dal subtipo della pagmatite, queste tormaline mostrano anche una

correlazione positiva molto buona del Na verso il F ed una correlazione negativa della vacanza del sito X verso il F. (ii) I

trend composizionali mostrano un decremento del Ca dalle zone periferiche al centro dei filoni; elevati contenuti in Ca,

molto probabilmente, riflettono l’influenza di rocce incassanti ricche in questo elemento. Un aumento del Ca durante gli

stadi finali della cristallizzazione primaria indica una elevata attivitr del Ca nel mezzo dal quale la tormalina si c formata. I

valori di concentrazione del Ca indicano un elevato controllo da parte del fattore geochimico e di contaminazione; il fat¬

tore di ripartizione degli elementi e cristallochimico sembrano aver giocato un ruolo importante soprattutto nelle concen¬

trazioni di Na e F nelle tormaline povere in Ca.

Key words: tourmaline, electron microprobe, geochemistry, complex pegmatites, Moldanubicum, Czech Republic

* E-mail: mnovak@sci.muni.cz

46

MILAN NOVAK

Introduction

Compositional variation in tourmaline from com-

plex pegmatites have been studied for several

decades. Staatz et al. (1955) established systematic

compositional trends in tourmaline from thè wall

zone inward in thè Brown Derby pegmatite, Gunni-

son County, Colorado; thè major element variation

from outer zone to thè core brings an increase in Li

concomitant with a decrease in Fe. Foord (1976) cor-

roborated thè generai compositional trend character-

ized by increasing Al and Li, and decreasing Fe in lay-

ered miarolitic dikes of thè Himalaya pegmatite-

aplite System, Mesa Grande, California. However,

rather complicated, locally even oscillatory zoning

was recorded in tourmaline crystals from open vugs.

The most detailed study to date, devoted to composi¬

tional evolution of tourmaline in complex pegmatites

- thè lepidolite pegmatite Bob Ingersol I, Black Hills,

South Dakota (Jolliff et al. 1986) - showed: Mg and Ti

decreasing abruptly from thè country rock through

border zone to thè wall zone; Fe decresing and

(Al+Li) increasing from thè wall zone to thè core;

and thè minor elements Mn, Zn and Ca generally in¬

crease toward thè core. Novak and Povondra (1995)

recognized apparent differences in compositional

trends of tourmaline within thè pegmatites of thè lep¬

idolite and elbaite subtype, respectively. The compo¬

sitional trends of major elements are in most aspects

similar to those described by Foord (1976) and Jolliff

et al. (1986); however, elbaite and rare liddicoatite

from pegmatites of thè elbaite subtype appears to be

Mn,F-rich and exhibits relatively low vacancies in thè

X-site relative to Mn-poor elbaite and rare rossman-

ite from lepidolite pegmatites (Novak and Povondra

1995, Novak et al. 1997, Selway et al. 1998, 1999).

A comparative study of compositional variations

in tourmaline from 16 individuai pegmatite dikes of

thè complex type in thè Moldanubicum, Czech Re-

public found out several distinct trends typical for thè

individuai pegmatite subtypes. Their formation is dis-

cussed in thè respect to various factors controlling

tourmaline composition. Compositional trends also

were compared with thè schematic covariation of Y-

site cations in tourmaline, given by Jolliff et al. (1986).

Late hydrothermal tourmaline from fissures and fis-

sure-like pockets (Novak & Taylor 1996, Novak 1996,

Novak & Selway 1997) are not included in discussion

and diagrams.

Complex pegmatites of thè moldanubicum

Granitic pegmatites, typically tourmaline-bearing,

are widespread particularly within thè metamorphic

terraine of thè Moldanubicum, Czech Republic (Fig.

1) (Novak et al. 1992). Following subtypes of complex

pegmatites (LCT family), mostly consistent with thè

currently used subdivision of pegmatites of thè rare-

element class (Cerny 1991, Novak and Povondra

1995), can be distinguished:

(i) Pegmatites of thè lepidolite subtype are thè

most widespread and most variable. Disregarding

lepidolite (abundant trilithionite to minor

polylithionite) as a dominant Li-bearing minerai, el¬

baite and rare rossmanite are commonly present in

COMPOSITIONAL PATHWAYS OF TOURMALINE EVOLUTION DURING PRIMARY

47

subordinate to accessory amount; petalite and ambly-

gonite-montebrasite are rare (Novak and Povondra

1995, Selway et al., 1999). Other typical minor to ac¬

cessory minerals include schorl, foitite, garnet, beryl,

cassiterite, manganocolumbite, apatite, zircon, and

topaz. A transitional petalite-lepidolite subtype

seems^ to be represented by thè locality Nova Ves

near Cesky Krumlov (Stanék 1992).

(ii) Pegmatites of thè elbaite subtype are relative-

ly widespread (Novak and Povondra 1995). Elbaite

(liddicoatite),the most abundant and locally even thè

only Li-bearing minerai, predominates over rare lep¬

idolite (typically polylithionite), muscovite is mostly

absent or very rare. Besides minor to accessory

schorl, garnet, beryl, cassiterite, manganocolumbite

and zircon, some B-rich minerals such as hambergite,

tusionite and boromuscovite also occur (e.g., Cooper

et al. 1994, Novak et al. 1998, 1999).

(iii) Pegmatite from Kracovice near Trebfè is

characterized by masutomilite to Mn-rich lepidolite

as typical Li-bearing minerals beside relatively rare

Mn-rich elbaite. It does not fit well to thè divided

pegmatite subtypes.The locality exhibits some excep-

tional features relative to thè other Moldanubian

complex pegmatites, absence of muscovite, very

abundant topaz, accessory minerals such as euxenite,

wolframoixiolite, pyrochlore, scheelite, and F-rich

hambergite (Novak et al. 1998). Minerai assemblages,

chemistry of micas (zinnwaldite, masutomilite) and

spatial relations to rare-earth beryl-bearing peg¬

matites with relatively abundant REE-oxide miner¬

als (aeschynite) in thè Trebfé vicinity indicate that

thè Kracovice pegmatite may be transitional to or be-

long to thè NYF family.

Two distinct patterns of internai structure of thè

complex pegmatites were recognized in thè

Moldanubicum disregarding pegmatite subtypes out-

lined above. Symmetrically or almost symmetrically

zoned pegmatite dikes (all lepidolite pegmatites, ma¬

sutomilite pegmatite from Kracovice and some el¬

baite pegmatites) are evidently more widespread,

than homogeneous to subhomogeneous elbaite peg¬

matites typically hosted in Ca-rich rocks (marbles,

Fe-skarn, pyroxene gneiss).

Localities, samples, experimental conditions

AND NORMALIZATION PROCEDURE

Tourmaline was studied at 16 pegmatite dikes

(Fig. 1). Principal data concerning pegmatite subtype,

minerai paragenesis, internai structure, size, and host

rock are given in Table 1. At each locality, tourmaline

grains (14 to 5) from all or almost all textural-para-

genetic units (1 to 5 grains per unit) were analyzed on

an electron microprobe. No apparent hiatus in tour¬

maline crystallization known from complex peg¬

matites (London et al. 1996) was observed in most lo¬

calities. The following examples of pegmatites of thè

individuai pegmatite subtypes, with various internai

structure and host rock, were selected to illustrate

typical compositional pathways. The symmetrically

zoned pegmatite dikes from 1 to 3 m thick hosted in

biotite gneiss: (i) thè lepidolite pegmatite from Las-

tovicky, (ii) elbaite pegmatite from Pikàrec, (iii) ma¬

sutomilite pegmatite from Kracovice; and (iv) subho¬

mogeneous dike of thè elbaite pegmatite from

Vlastéjovice hosted in Fe-skarn. A detailed descrip-

tion of these pegmatite dikes and their minerai para¬

genesis is given by Stanèk (1973), Houzar (1987),

Nèmec (1990), Novak and Hyrsl (1992) and Novak

and Povondra (1995).

Electron microprobe analyses were carried out in

thè WDS (wavelength-dispersive) mode on a Cameca

SX-50 instrument, Department of Geological Sci¬

ences, University of Manitoba, Winnipeg, with a beam

diameter 4-5 mm, accelerating potential 15 kV. A sam-

ple current 20 mA was used for Si, Al. Ti, Fe, Mn, Mg,

Ca, Na, and K, thè current of 40 mA for Zn, F and P;

a counting time for all elements 20 seconds. The fol¬

lowing standards were used for Ka X-ray lines: Si, Ca

- diopside. Al - kyanite, Fe - fayalite. Ti - rutile, Mg -

pyrope, Mn - spessartine, Na - albite, K - orthoclase, P,

F - apatite and Zn - gahnite. Data were reduced using

thè PAP routine (Pouchou and Pichoir 1985).

The Chemical formulae from microprobe data

used in thè diagrams and tables were normalized at 6

Si apfu, as was used by Jolliff et al. (1986). However,

thè trivalent cations Al and perhaps B may enter thè

T-site of thè tourmaline structure (Burns et al. 1994,

Hawthorne 1996) and their presence in thè T-site

causes overestimation of Al in such calculation. The

accurate formula calculation for tourmaline minerals

recommended by Grice and Ercit (1993) or other for¬

mula calculations from microprobe data (e.g., Man-

ning 1982, Morgan and London 1987, Henry and

Dutrow 1990, 1996, Selway and Novak 1997) were

not used due to: (1) undetermined H20, Li20 and

B203 contents, (2) highly variable lithium content in

thè investigated tourmalines, (3) lack of structural da¬

ta. The used style of recalculation also is supported by

relatively negligible deviations from 6 Si apfu (5.84 -

6.21) found by Povondra (1981), Povondra et al.

(1985), Stanék and Povondra (1987) and Novak and

Povondra (1995) in thè wet Chemical analyses of tour¬

maline from thè Moldanubian rare element peg¬

matites.

Decreasing sum of Fe + Mg + Ti atoms per formu¬

la unit (apfu) was selected as a suitable index (FMT

index) of progressive crystallization in tourmaline.

The contents of these elements should decrease dur-

ing crystallization, as suggested by Foit and Rosen¬

berg (1979) and Jolliff et al. (1986), and thus reflect

degree of fractionation.

Chemical composition of tourmaline and compo¬

sitional pathways are controlled by several factors

due to continuously changing PT conditions, extreme

geochemical evolution of parent medium (evolved

pegmatite melt ± vapour) in complex pegmatites

(London 1996) relative to most magmatic and meta-

morphic rocks, and variable Chemical composition of

host rocks. In order to simplify thè below discussion,

following factors were defined: (i) a crystal-chemical

factor involves crystal-chemical constraints

(Hawthorne 1996) mostly controlled by decreasing

temperature; (ii) geochemical factor corresponds to

increasing degree of fractionation and alkalinity vari-

ation of pegmatitic melt (London et al. 1989); (iii) el-

ement-partitioning factor characterizes influence of

co-crystallizing minerals, particularly micas and gar-

nets; (iv) contamination factor reflects Chemical com¬

position of a pegmatite host rock. However, it must

be emphasized that thè effects of thè individuai fac¬

tors are mostly combined and, moreover, thè defini-

tions of thè individuai factors given above are mostly

rather wide.

48

MILAN NOVAK

Table 1 - List of examined pegmatites. * - localities illustrateci in thè figures 2 - 5; subtype - L-lepidolite, PL-tran-

sitional lepidolite-petalite, E-elbaite, Ma-masutomilite; micas - M-muscovite, L-lepidolite, Ma-masutomilite; zon-

ality - Z-zonal, H-subhomogeneous to homogeneous.

Results

Compositional variations in tourmaline from indi¬

viduai pegmatite subtypes outlined above are illus-

trated on figures 2 through 5 and representative com-

positions are given in tables 2 through 5. Variation of

nine elements: Ti, Mg, Fe, Zn, Mn and Al as major to

minor Y-site cations; Na and Ca as major to minor X-

site cations; and F in thè 0(1) site are discussed in de-

tail.

Y-site cations

These elements are discussed in thè order of de-

creasing cation field strength, as shown by Jolliff et al.

(1986).

Titanium and Mg behave in a very similar way.

Their concentration commonly drop abruptly from

tourmaline in thè outermost unit (tourmaline locally

associated with or replaced biotite) to concentrations

below thè detection limit (Fig. 2, 5a).

Iron concentration slightly increases from Mg-en-

riched schorl to foitite in thè outermost unit and then

decreases gradually to Fe-free tourmaline in thè end

of primary (magmatic) pegmatite crystallization at

most localities (Fig. 3, 5b).

Zinc as a typical trace element in tourmaline ex-

hibits very similar behaviour at almost all localities,

disregarding their subtype, internai structure, minerai

assemblage of tourmaline and host rock of pegmatite

(Fig. 2, 5a). Initial low concentrations increase in thè

1. 4-0.7 range of thè FMT index up to about 0.1 Zn ap-

fu, and then fall off.

Variations in Mn versus thè FMT index display

distinct patterns depending on thè individuai peg¬

matite subtypes (Fig. 3, 5b). In thè lepidolite peg¬

matites, Mn is low throughout thè pegmatite crystal¬

lization (Fig. 3a), only some localities display a rather

exceptional increase up to 0.5 apfu at thè FMT index

of about 0.7 to 0 (Puklice I, Radkovice, Roznà-

Hradisko, Roznà-Borovina). Elbaite pegmatites are

characterized by significant increase in Mn during

Fe + Mg + Ti

Fig. 2. Variation in concentrations of Ti, Mg and Zn per formula

unit versus thè FMT index. A - Lastovicky, B - Pikàrec, C - Kra-

covice.

COMPOSITIONAL PATHWAYS OF TOURMALINE EVOLUTION DURING PRIMARY

49

3.0 2.5 2.0 1.5 1.0 0.5 0.0

Fe + Mg + Ti

Fig. 3. Variation in concentrations of Fe, Mn, and Al per formula

unit versus thè FMT index. The same localities as in Fig. 2. Trian-

gles (Fe) below imaginary line connecting sum (Fe + Mg + Ti) = 0

and sum Y-cations = 3 indicate Mg and Ti amount, increased par-

ticularly in early stages.

pegmatite crystallization up to 1.1 apfu (Pikàrec); a

decrease in Mn is locally recorded at thè end of crys¬

tallization (Fig. 3b, 5b). Manganese behaviour in thè

masutomilite pegmatite from Kracovice is very simi-

lar to those observed in elbaite pegmatites (Fig. 3c).

Among thè other Y-site cations. Al exhibits a

rather complicated evolution trend. Decrease in thè

FMT index is concomitant with increase in Al in both

lepidolite and elbaite pegmatites; however, details of

thè trends are quite different, particularly in thè reai

values of Al apfu in thè lepidolite pegmatites. The Al-

contents in thè Y-site (Z-site is proposed to be fully

occupied by Al) varies from 0.5 to 2.2 apfu, and there

is a significant variation from 1.5 to 2.2 apfu for Fe-

free tourmaline with thè FMT index values of about

0 (Fig. 3a). The pathway of thè Al variation is rela-

tively regular. Behaviour of Al in tourmaline from el¬

baite pegmatites is quite similar to those in lepidolite

pegmatites except Vlastéjovice where Al varies over

a much broader range from -1.3 to 2.2 Al apfu (nega¬

tive value corresponds to low Al content, when 1.3

apfu in thè Z-site is occupied by other transitional el-

ements - Fe3+, Fe2+ and/or Mg) (Fig. 5b). Significant

variation from 1.3 to 2.2 Al for thè FMT index dose

to 0 is also found.The range is very likely smaller due

to overestimation of Al during formula normalization

on 6 Si. The behaviour of Al in tourmaline from thè

masutomilite pegmatite in Kracovice is in a conspic-

uous contrast to those from all other studied peg¬

matites, either lepidolite or elbaite subtype. It varies

from 0.5 to 1.1 Al apfu and is rather Constant through

thè variation in thè FMT index (Fig. 3c).

X-site cations

The X-site is occupied by Na and Ca; however, sig¬

nificant vacancies also are recorded (Povondra 1981,

MacDonald et al. 1993, Novak and Povondra 1995,

Novak and Selway 1997, Selway et al. 1998, 1999); K

is mostly below thè detection limit. Distinctly differ-

Fig. 4. Variation in concentrations of Na, Ca and F per formula

unit versus thè FMT index. The same localities as in Fig. 2.

50

MILAN NOVAK

Table 2 - Representative compositions of tourmaline from thè zoned lepidolite pegmatite Lastovicky.

ent patterns of Na variation were observed in thè in¬

dividuai pegmatite subtypes. All lepidolite peg-

matites exhibit an increase of Na values, ranging from

0.4 to 0.6 apfu in schorl to foitite of thè outermost

textural-paragenetic unit up to about 0.8 apfu in Fe-

rich elbaite with thè FMT index dose to 1, followed

by apparent Na decrease at 0.6 to 0.4 apfu in elbaite

to rossmanite at thè end of crystallization with thè

FMT index dose to 0 (Fig. 4a). Elbaite pegmatites, al-

though fairly consistent in compositional evolution of

thè Y-site cations, display two distinct trends of thè X-

site cations. Calcium-poor tourmalines from zoned

pegmatites are characterized by graduai Na increase

from 0.4 to 0.85 apfu during crystallization (Fig 4b).

Tourmalines with high Ca-content (Ca > 0.2 apfu, lo-

cally Ca > Na in thè Bliznà and Recice localities)

from mostly homogeneous pegmatites show Constant

Na at about 0.6 apfu throughout thè crystallization,

but slight to abrupt decrease is recorded at thè end of

crystallization at Vlastéjovice (Fig. 5c) and Bliznà

(Novak et al. 1997). The masutomilite pegmatite at

Kracovice exhibits relatively high Na in tourmaline,

ranging from 0.6 up to 0.93 apfu at thè end of crystal¬

lization (Fig. 4c).

In most pegmatites characterized by Ca-poor

tourmalines, Ca occurs mostly in very low concentra-

tions throughout thè pegmatite crystallization, and

only slight increase is recorded in late stages. In el¬

baite pegmatites with Ca-rich tourmalines, thè rela¬

tively high Ca concentrations abruptly decrease in

early tourmaline, and then Ca abruptly increases in

thè tail-end of primary crystallization (Fig. 5c)

(Novak et al. 1997).

Fluorine

Fluorine varies in a wide range; very low contents

of about 0.0 to 0.2 apfu F were found in early crystal-

lized schorl to foitite from outermost situated peg¬

matite units of lepidolite pegmatites and in relatively

late Fe-poor elbaite to rossmanite in lepidolite-rich

units. Highest contents of F are typical for Fe-rich el¬

baite in lepidolite pegmatites and Mn-rich elbaite in

elbaite pegmatites. Very high F-contents up to 0.85

apfu were found particularly in elbaite from thè ma¬

sutomilite pegmatite at Kracovice (Fig. 4c), relative

to thè tourmaline (schorl to elbaite) from other com-

plex pegmatites (Chaudhry and Howie 1976, Sahama

et al. 1979, Povondra et al. 1985, Jolliff et al. 1986,

Novak and Selway 1997, Selway et al., 1999).

Fluorine exhibits evolution trend remarkably sim-

COMPOSITIONAL PATHWAYS OFTOURMALINE EVOLUTION DURING PRIMARY

51

Table 3 - Representative compositions of tourmaline from thè zoned elbaite pegmatite Pikàrec.

ilar to that of Na in thè X-site of Ca-poor tourmalines

(Fig. 4); very good positive correlations Na versus F

and negative correlation X-site vacancy versus F

were recorded. In Ca-rich tourmalines, F commonly

increases during crystallization (Fig. 5c).

Discussion

In order to discuss thè compositional trends in

tourmaline throughout thè pegmatite crystallization

in distinct pegmatite subtypes, thè trends were com-

pared with thè schematic covariation of Y-site cations

in tourmaline, published by Jolliff et al. (1986) (Fig.

6). Results of thè comparison as well as a discussion

of thè controlling factors are given in thè sequence: Y-

site cations in thè order of decreasing field strength,

X-site cations and fluorine.

Y-site cations

1. Titanium and Mg drop abruptly from tourma¬

line in thè outermost unit inwards, as observed in oth-

er complex pegmatites (e.g., Foord 1976, Jolliff et al.

1986, Aurisicchio and Freda 1992), and as corre-

sponds to thè schematic evolution of Ti and Mg (Fig.

6). However, in most elbaite pegmatites with homo-

geneous or subhomogeneous internai structure and

Ca-rich tourmalines (e.g., Bliznà; Novak et al. 1997),

and in thè masutomilite pegmatite in Kracovice, Mg

and locally also Ti maintain relatively high concen-

trations throughout thè crystallization, despite a

graduai decrease (Fig. 2c, 5a). Most of these peg¬

matites are hosted by Ca, Mg-rich rocks (dolomite

marble, pyroxene gneiss); thus a weak influx of Mg

from thè host rock into pegmatite during its crystal¬

lization seems to be an explanation. However, in-

creased Ti concentration in tourmaline from thè

Bliznà pegmatite (Novak et al. 1997) penetrating

dolomite-calcite marble with graphite and elevateci

Mg in tourmaline from Kracovice enclosed in gneiss

are not explained sufficiently.

2. Iron slightly increases from tourmaline in thè

outermost unit inwards, then decreases to thè centre.

Thus, covariations of Fe and other Y-site cations are

very dose to thè schematic (see Fig. 6) and to those

described from other localities (e.g., Jolliff et al.

1986). Because Fe is thè dominant element in thè cal-

culated FMT index, thè discussion of Fe behaviour in

detail is fairly limited.

3. Low concentrations of Zn were found in early-

52

MILAN NOVAK

Fe + Mg + Ti

Fig. 5. Variation in concentrations of thè Y-site and X-site cations

and fluorine versus thè FMT index in elbaite pegmatite from

Vlastéiovice. Missing points in thè diagram A vary from 0.57 to

0.22 apfu Mg.

crystallized tourmaline from almost all pegmatites,

disregarding their subtype, internai structure and

host rock. Zinc increases slightly in Fe-rich elbaite,

then fall off (Fig. 2). This trend is very dose to thè

schematic one (Fig. 6).

4. Behaviour of Mn exhibits two evidently distinct

trends. Mostly very low Mn concentrations are found

in lepidolite pegmatites; however, minor increase in

thè end of crystallization is locally recorded. Man¬

ganese generally increases during crystallization in

elbaite pegmatites and in thè masutomilite pegmatite

from Kracovice; some localities, however, display

abrupt decrease of Mn in thè tail-end of crystalliza¬

tion (Fig. 3b, 5b). Such trend is very similar to thè

schematic pattern (Fig. 6), and it was found in other

elbaite pegmatites (e.g., Zagorskyi and Peretyazhko

1992, Aurisicchio and Freda 1992, Novak and Povon-

dra 1995, Novak et al. 1997). In lepidolite pegmatites,

a cation partitioning with abundant co-crystallizing

muscovite and lepidolite may result in dominant Mn-

poor tourmalines. However, muscovite is also an in-

dicator of elevated acidity (Burt 1981, London 1982,

Gordeenko and Ponomarova 1988). Hence, thè acid-

Fig. 6. Schematic illustration of ideal covariation of Y-site cations

of tourmaline in response to decreasing temperature and increas-

ing fractionation of melt (Jolliff et al. 1986).

ity-alkalinity of a parent medium, which appears to

be different in lepidolite and elbaite pegmatites

(Némec and Povondra 1993, Novak and Povondra

1995), may have played an important role. Garnet

(spessartine-almandine), which locally occurs in peg¬

matites along with tourmaline (Némec 1983, 1990),

exhibits increased Mn contents in elbaite pegmatites

relative to lepidolite ones (Novak and Povondra

1995) . It is not clear whether this fact indicates in¬

creased overall activity of Mn in this particular peg¬

matite subtype or crystallization of abundant micas

(muscovite, lepidolite) in lepidolite pegmatites de-

pressed activity of Mn in melt. However, thè control-

ling role of garnet seems to be limited, because simi¬

lar compositional trends and concentrations of Mn

are found in tourmaline from both garnet-rich peg¬

matites (Pikàrec, elbaite with up to 7.93 wt.% MnO)

and garnet-free elbaite pegmatites (Vlastéjovice,' el¬

baite with up to 6.60 wt.% MnO).These observations

are in a contrast with those found in elbaite peg¬

matites from thè Elba Island by Aurisicchio and Pez-

zotta (1997); where thè Mn contents in elbaite from

pockets associated with spessartine-rich garnet are

evidently depressed relative to those from garnet-

free pockets.

5. Variations in Al characterized by a generai in¬

crease throughout crystallization are consistent with

all other published data (Foord 1976, Sahama et al.

1979, Jolliff et al. 1986, 1987, Némec 1989, Aurisicchio

and Freda 1992, Novak and Povondra 1995), and thè

schematic covariation model (Fig. 6), too. A wide

range of Al concentrations in Fe-free tourmalines

was found in lepidolite pegmatites (Novak and

Povondra 1995, Novak and Selway 1997, Selway et al. ,

1999), and also in some elbaite pegmatites (Fig. 3b).

It indicates, that in thè Fe-depleted end of thè prima-

ry crystallization, thè trend emerges toward Fe-free,

Al-rich tourmaline components. In thè elbaite peg¬

matites with Ca-rich tourmaline, thè Al increase may

result from an increase of thè olenite component. In

lepidolite pegmatites, Fe-free elbaite is characterized

by high rossmanite component, which may locally

predominate over thè elbaite component (Roznà-

Hradisko, Lastovicky; Novak and Selway 1997, Sel¬

way et al., 1998, 1999) (Fig. 3a). It is not clear, whether

thè increase of thè rossmanite component is con-

trolled by crystal-chemical constraints (Hawthorne

1996) , cation partitioning (Li, Al) with associated lep¬

idolite or some other factors.

The Al concentrations in tourmaline from thè ma¬

sutomilite pegmatite at Kracovice are relatively con-

COMPOSITI ON AL PATHWAYS OF TOURMALINE EVOLUTION DURING PRIMARY

53

Table 4 - Representative compositions of tourmaline from thè zoned masutomilite pegmatite Kracovice.

stant throughout pegmatite crystallization. It is in a

conspicuous contrast with all data published to date

(e.g., Staatz et al. 1955, Foord 1976, Sahama et al.

1979, Povondra et al. 1985, Jolliff et al. 1986,

Zagorskyi and Peretyazhko 1992, Aurisicchio and

Freda 1992, Novak and Povondra 1995, Novak and

Selway 1997) with thè exception of elbaite pegmatite

Belo Horizonte No.l, Peninsular Ranges Batholith,

southern California (Taylor et al. 1997). Perhaps

some specific compositional characteristics of thè

parent medium, such as very high activity of F (abun-

dant topaz, relatively common F-rich hambergite),

high activity of B (F-rich hambergite), and/or in-

creased alkalinity (absence of muscovite, presence of

zinnwaldite-masutomilite micas, abundance of F-rich

hambergite) may have generated nearly Constant Al

contents - a unique compositional trend in tourma¬

line from complex pegmatites.

A comparison of schematic covariations on thè Y-

site cations based on thè field strength calculations

(Jolliff et al. 1986), and actual variations in thè indi¬

viduai pegmatite subtypes described in this work,

yielded remarkable similarities in thè evolution of Ti,

Mg, Fe, Zn and mostly also of Mn and Al. This obser-

vation may indicate that thè crystal-chemical factor

in combination with thè geochemical factor is thè

most important to control tourmaline composition

during primary crystallization. Some difference in

evolution of Mn from thè theoretical schematic mod¬

el found in lepidolite pegmatites seems to be con-

trolled by (i) a element-partitioning among simulta-

neously crystallized phases, when mica (muscovite

and lepidolite) buffers Mn-activity to a low level, or

(ii) increased acidity of a parent medium indicated by

abundant muscovite and trilithionite (Gordeenko

and Ponomareva 1988). Behaviour of Al is generally

consistent with thè schematic model (Fig. 6). The al-

most Constant concentrations of Al in tourmaline

throughout crystallization in thè masutomilite peg¬

matite from Kracovice seems to be thè most signifi-

cant deviation from thè schematic model found in

this work. It may be controlled by a specific composi¬

tion of thè parent medium. Rather elevated contents

of Mg in tourmaline from homogeneous pegmatites

cutting Mg-rich rocks, and Fe in subhomogeneous

pegmatite at Vlastéjovice cutting Fe-skarn (Fig. 5a,

b), respectively, show evident participation of thè

contamination factor in thè elbaite pegmatites with a

homogeneous and subhomogeneous internai struc-

ture. Zoned pegmatites display only fairly negligible

influence on thè tourmaline composition disregard-

54

MILAN NOVAK

Table 5 - Representative compositions of tourmaline from thè homogeneous elbaite pegmatite Vlastejovice.

ing thè composition of host rocks. Pegmatites en-

closed in serpentinite (Nova Ves), amphibolite (Do-

brà Voda) or marble (Krasonice), respectively, do not

virtually differ from those enclosed in various

metapelites (see Table 1) or thè difference is quite a

negligible.

X-site cations and fluorine

Behaviour of cations in thè X-site was studied by

Foord (1976) and Jolliff et al. (1986), and also dis-

cussed for distinct subtypes of complex pegmatites

(Novak and Povondra 1995, Novak and Selway 1997,

Novak et al. 1997 and Selway et al., submitted). As

considerable covariations of Na, Ca and F have been

observed in this study, these elements will be dis-

cussed together.

Disregarding pegmatite subtypes described here-

in, two principal groups were recognized in thè trends

of thè X-site cations and F in tourmaline; Ca-poor

tourmalines in various pegmatite subtypes, and Ca-

rich tourmalines (Ca > 0.2 apfu) from elbaite peg¬

matites typically with homogeneous to subhomoge-

neous internai structure.

The first group is characterized by a very good

positive correlation of Na versus F and negative cor-

relation X-site vacancy versus F throughout thè crys-

tallization. In thè lepidolite pegmatites, apparent de-

crease of Na and F is recorded in thè end of tourma¬

line crystallization. Mostly Fe-poor elbaite (or ross-

manite) is commonly associated with lepidolite (Li,

F-rich), but locally also with albite (Na-rich); some

tourmaline occurs in a quartz core, but lepidolite and

locally albite are mostly present in a subordinate

amount.The Na partitioning between tourmaline and

albite to control Na activity seems to have only a very

minor significance if any, because thè same behaviour

of Na was found in tourmaline completely enclosed

in lepidolite. However, thè partitioning of F into lep¬

idolite over tourmaline is very likely. Such conse-

quence is also supported by thè fact that change into

thè reverse decreasing trend of Na and F in tourma¬

line is apparently related to thè substitution of mus-

covite by lepidolite in thè tourmaline assemblages at

thè localities examined. Lepidolite crystallization

seems to have significantly depleted thè melt of F.

Graduai increase in Na and F of Ca-poor tourmalines

from elbaite pegmatites (muscovite and lepidolite are

rare or absent) and masutomilite pegmatite (zin-

COMPOSITIONAL PATHWAYS OF TOURMALINE EVOLUTION DURING PRIMARY

55

nwaldite, Mn-rich lepidolite and masutomilite are

abundant) throughout thè overall pegmatite crystal-

lization indicates elevated activity of F; however, in-

creasing alkalinity of parent medium may have

played a role. Remarkably positive correlation of Na

and F in all Ca-poor tourmalines is very likely result-

ing from crystaì-chemical constraints, possibly a sig-

nificant interaction Na+- F in thè crystal structure of

tourmaline (Robert et al. 1997). Hence, activity of F

in melt may control entering of Na into thè tourma¬

line structure.

In Ca-rich tourmalines (Ca > 0.2 apfu), entirely

different compositional trends were found. There is

no positive correlation of Na and F. Increased Ca

contents from thè outermost zones in some peg-

matites and subsequent decrease seem to reflect thè

influence of Ca-rich host rocks (marble, Fe-skarn, py-

roxene gneiss). However, Ca increase in thè tail-end

of primary crystallization, found in Ca-rich tourma¬

line from elbaite pegmatites (Fig. 5c) and also in

some lepidolite pegmatites, indicates rather elevated

activity of Ca in late stages of magmatic crystalliza¬

tion (London et al. 1989, London 1992) than an influx

of Ca from host rock.

CONCLUSIONS

A comparative study of compositional variations

in tourmaline from 16 individuai pegmatite dikes of

thè complex type in thè Moldanubian region, Czech

Republic found out several distinct trends of thè Y-

site cations, X-site cations and F typical for thè indi¬

viduai pegmatite subtypes.

Compositional trends of thè Y-site cations (Ti, Mg,

Fe, Zn, Mn and Al) are dose to thè schematic dia-

gram of cations covariation derived from thè cation

field strengths. Concentration of thè Y-site cations

seems to be controlled particularly by thè crystal-

chemical factor in a combination with thè geochemi-

cal factor, whereas thè element-partitioning factor is

less important. The contamination factor operates in

Ca-rich tourmaline from elbaite pegmatites with ho-

mogeneous internai structure cutting Mg- and/or Fe-

rich rocks, particularly in early stage of magmatic

crystallization. Zoned complex pegmatites exhibit

quite a negligible control of thè contamination factor.

The X-site cations (Na and Ca) and F exhibit con-

trasting evolution trends in Ca-poor and Ca-rich tour¬

maline, respectively, disregarding their pegmatite sub-

type. Their concentrations and thè X-site vacancy indi¬

cate elevated control of thè geochemical factor (in¬

creased amounts of Ca in late stages of crystallization),

contamination factor (elevated amounts of Ca in early

stage of crystallization) and particularly element-parti¬

tioning factor combined with crystal-chemical factor

(Na - F relations in Ca-poor tourmaline).

Distinct compositional trends found in tourmaline

from lepidolite and elbaite pegmatites in thè

Moldanubicum (Novak and Povondra 1995) were

generally corroborated in thè present study. Behav-

iour of Al in tourmaline from masutomilite pegmatite

in Kracovice is unique and worth a detailed study as

' well as element-partitioning between tourmaline and

mica and tourmaline and garnet throughout peg¬

matite evolution.

Acknowledgements

The study was supported by thè Dean of Science,

University of Manitoba Postdoctoral Fellowship to

thè author in 1991 to 1993 and by thè NSERC Re¬

search plus Major Installation grants to Petr Cerny.

Part of thè work was also supported by Grant Agency

of Czech Republic, grant No. 205/96/0855. I am in-

debted to Petr Cerny, Federico Pezzotta and an

anonymous reviewer for heplful comments, and to

Ron Chapman for assistance with analytical work

and computer.

References

Aurisicchio C. & Freda C., 1992 - The role of thè tourmaline

minerals in thè evolution of thè Elban pegmatites: Tuscany,

Italy. In: Abstracts of Papers, International Symposium on thè

Mineralogy, Petrology and Geochemistry of Granitic Peg¬

matites, Lepidolite 200, Nové Mesto na Moravé, August-Sep-

tember 1992: 12.

Aurisicchio C. & Pezzotta F., 1997 - Tourmaline-group miner¬

als of thè LCT miarolitic pegmatites of thè Elba Island,

(Italy): Chemical composition and genetic and paragenetic in-

ferences. In: Abstracts, International Symposium on Tourma¬

line, Nové Mèsto na Moravé, July 1997: 1-2.

Burns P.C., Macdonald D.J. & Hawthorne F.C., 1994 - The crystal

chemistry of manganese-bearing elbaite. Canad. Mineralogist ,

32:31-41.

Burt D.M., 1981 - Acidity-salinity diagrams-application to greisen

and porphyry depositis. Econ. Geology, 76: 832-843.

Cerny P, 1991 - Rare-element Granitic Pegmatites. Part I: Anato-

my and Internai Evolution of Pegmatite Deposits. Geoscience

Canada , 18: 49-67.

Chaudhury M.N. & FIowie R.A., 1976 - Lithium tourmalines

from thè Meldon aplite, Devonshire, England. Miner. Mag ., 40:

747-751.

Cooper M„ Hawthorne F.C., Novak M. & Taylor M.C., 1994 -

The crystal structure of tusionite Mn2+Sn4+(BO,)2, a dolomite

structure borate. Canad. Mineralogist , 32: 903-907.

Foit F.F. jr. & Rosenberg P.E., 1979 - The structure of vanadium-

bearing tourmaline and its implications regarding tourmaline

solid-solutions. Amen Mineralogist, 64: 788-798.

Foord E. E., 1976 - Mineralogy and petrogenesis of layered peg-

matite-aplite dikes in thè Mesa Grande district. San Diego

County, California. PhD. Thesis, Stanford University , Califor¬

nia, 326 p.

Gorgeenko V.V. & Ponomareva I.N., 1988 - Physico-chemical

stability conditions of lithium micas of thè lepidolite series.

Zap. Vses. Minerai. Obch., 117: 633-638.

Grice J.D. & ErcitT.S., 1993 - Ordering of Fe and Mg in thè tour¬

maline crystal structure: The correct formula. Neues. Jb. Min¬

er. Abh., 165: 245-266.

Hawthorne F.C., 1996 - Structural mechanism for light-element

variations in tourmaline. Canad. Mineralogist, 34: 123-132.

Henry D.J. & Dutrow L.B., 1990 - Ca substitution in Li-poor alu-

minous tourmaline. Canad. Mineralogist, 28: 111-124.

Henry D.J. & Dutrow L.B., 1996 - Metamorphic tourmaline and

its petrologie applications. In: Grew E.S. & Anovitz L.M.

(eds): Reviews in Mineralogy. Boron Mineralogy, Petrology

and Geochemistry, 33: 503-557.

Houzar S., 1987 - New lithium-bearing pegmatite from Kracov¬

ice nearTrebfc. Acta Sci. nat. Musei Moraviae occident. Tre bit'.

1-4. (in Czech).

Jolliff B.L., Papike J.J. & Shearer C.K., 1986 - Tourmaline as a

recorder of pegmatite evolution: Bob Ingersoll pegmatite,

Black Hills, South Dakota. Amen Mineralogist, 71:472-500.

Jolliff B.L., Papike J.J. & Shearer C.K., 1987 - Fractionation

trends in mica and tourmaline as indicators of pegmatite in¬

ternai evolution: Bob Ingersol pegmatite, Black Hills, South

Dakota. Geochim. Cosmochim. Acta, 51: 519-534.

London D., 1982 - Stability of spodumene in acidic and saline flu-

orine-rich environments. Carnegie Institution of Washington

Year Book, 81: 331-334.

London D., 1992 - The application of experimental petrology to

thè genesis and crystallization of granitic pegmatites. Canad.

Mineralogist, 30: 499-540.

London D., 1996 - Granitic pegmatites. Trans. Royal Soc. Edin¬

burgh: Earth Sciences, 87: 305-319.

56

MILAN NOVAK

London D., Morgan G.B. Vi & Hervig R.L., 1989 - Vapor-un-

dersaturated experiments with Macusani glass + H20 at 200

MPa, and internai differentiation of granitic pegmatites. Con-

trib. Minerai. Petrol. , 102: 1-17.

London D„ Morgan G.B. VI. & Wolf M.B., 1996 - Boron in

granitic rocks and their contact aureoles. In: Grew E.S. &

Anovitz L.M. (eds): Reviews in Mineralogy. Boron Mineralo¬

gy, Petrology and Geochemistry, 33: 299-330.

Macdonald D.J., Hawthorne F.C. & Grice J.D., 1993 - Foitite,

Q[Fe2+2(Al, Fe3+)] Al^O^ (B03)3(0H)4, a new alkali-defi-

cient tourmaline: Description and crystal structure. Amen

Mineralogist , 78: 1299-1303.

Manning D.A.C., 1982 - Chemical and morphological variation in

tourmalines from thè Flub Kapong batholith in peninsular

Thailand. Miner. Mag.,45: 139-147.

Morgan G.B. VI & London D„ 1987 - Alteration of amphibolitic

wallrocks around thè Tanco rare-element pegmatite, Bernic

Lake, Manitoba. Amer. Mineralogist, 72: 1097-1121.

Némec D., 1983 - Zinnwaldit in Moldanubischen Lithium-Peg-

matiten. Chem. Erde, 42: 197-204.

Némec D., 1989 - Chemical composition of tourmaline in peg¬

matites of thè Jihlava Massif (Western Moravia, CSSR).

Chem. Erde, 49: 317-340.

Némec D., 1990 - Neues zur Mineralogie eines Hambergite-

fiihrenden Pegmatitgangs von Kracovice (bei Trebic, West-

Morava, CSFR). Z geol. Wiss., 18: 1105-1115.

Némec D. & Povondra P, 1993 - Chemical composition of lepi-

dolite and thè acidity-alkalinity of its parent pegmatite medi¬

um. Scripta Fac. Sci. Nat. Univ. Masaryk Brun., Geology , 23:

45-53.

Novàk M., 1996 - Tourmaline environments in complex peg¬

matites; a basis for mineralogical and geochemical study. Acta

Miner. Petrogr., Szeged, 37: 85.

Novàk M., Burns P.C. & Morgan G.B. VI, 1998 - Fluorine vari¬

ation in hambergite from granitic pegmatites. Canad. Mineral¬

ogist, 36:^441-446.

Novàk M., Cerny P, Cooper M., Hawthorne F.C., Otollini L.,

Zhi Xu & Liang, J.-J., 1999 - Boron-bearing 2M, polylithion-

ite and 2M,+ 1M boromuscovite from an elbaite pegmatite at

Reècice, western Moravia, Czech Republic. Eur. J. Minerai.,

11,669-678.

Novàk M., Cerny R, Cech F. & Stanèk I, 1992 - Granitic peg¬

matites in thè territory of thè Moravian and Bohemian

Moldanubicum. In: Novàk M. & Cerny P. (eds). Field Trip

Guidebook, International Symposium on Mineralogy, Petrol¬

ogy and Geochemistry of Granitic Pegmatites, Lepidolite 200,

Nové Mèsto na Moravé, August-September 1992: 11-20.

Novàk M. & Hyrsl J., 1992 - Locality No. 3: Vlastéjovice near

Zruc nad Sàzavou, pegmatites with fluorite penetrating skarn.

In: Novàk M. & Cerny P. (eds). Field Trip Guidebook, Inter¬

national Symposium on Mineralogy, Petrology and Geochem¬

istry of Granitic Pegmatites, Lepidolite 200, Nové Mèsto na

Moravé, August-September 1992: 33-37.

Novàk M. & Povondra P, 1995 - Elbaite pegmatites in thè

Moldanubicum; a new subtype of thè rare-element class. Min¬

erai. Petrol., 55; 159-176.

Novàk M. & Selway J.B., 1997 - Locality No. 1: Roznà near

Bystrice nad Pernstejnem, Hradisko hill, a large lepidolite

subtype pegmatite dike. In: Novàk M. & Selway J.B. (eds.).

Field Trip Guidebook, International Symposium on Tourma¬

line, Nové Mèsto na Moravé, June 1997: 23-38.

Novàk M., Selway J.B., Korbel P. & Sarbach M., 1997 - Locali¬

ty No. 4: Bliznà near Cernà v PosumavI, elbaite subtype peg¬

matite penetrating marble with exomorphic tourmaline. In:

Novàk M. & Selway J.B. (eds.). Field Trip Guidebook, Inter¬

national Symposium on Tourmaline, Nové Mèsto na Moravé,

June 1997: 55-70.

Novàk M., & Taylor M.C., 1996 - New occurrences of foitite and

its position in pegmatite evolution. In: Abstracts GAC/MAC

Annual Meeting, May 27-29, Winnipeg 1996: A 70.

Pouchou J.L. & Pichoir F., 1985 - “PAP” procedure for improved

quantitative microanalysis. Microbeam Anal., 20: 104-105.

Povondra P., 1981 - The crystal chemistry of tourmalines of thè

schorl-dravite series. Acta Univ. Carol., Geol.: 223-264.

Povondra R, Cech F. & Stanèk J., 1985 - Crystal chemistry of el-

baites from some pegmatites of thè Czech Massif. Acta Univ.

Carol., Geol.: 1-24.

Robert J.-L., Gourdant J.-P, Linnen R.L., Rouer O. & Benoist

P, 1997 - Crystal-chemical relationships between OH, F and

Na in tourmalines. In: Abstracts, International Symposium on

Tourmaline, Nové Mèsto na Moravé, July 1997: 84-85.

Sahama Th.G., v.Knorring O. & Tornroos R., 1979 - On tour¬

maline. Lithos, 12: 109-114.

Selway J.B. & Novàk M., 1997 - Experimental conditions, nor-

malization procedures and used nomenclature fro tourmaline.

In: Novàk M. & Selway J.B. (eds.). Field Trip Guidebook, In¬

ternational Symposium on Tourmaline, Nové Mèsto na

Moravé, June 1997: 19-21.

Selway J.B., Novàk M., Cerny P. & Hawthorne F.C., 1997 -

Tourmaline from lepidolite-subtype pegmatites. In: Abstracts,

International Symposium on Tourmaline, Nové Mèsto na

Moravé, July 1997,91-92.

Selway J.B., Novàk M., Cerny P. & Hawthorne F.C. 1999 -

Compositional evolution of tourmaline in lepidolite-subtype

pegmatites. Eur. J. Miner., 11, 569-584.

Selway J.B., Novàk M., Hawthorne F.C., Cerny R, Ottolini L.

& Kyser T.K., 1998 - Rossmanite, Q[LiAl2] A^S^O^ (B03)3

(OH)4, a new alkali-deficient tourmaline: Description and

crystal structure. Amer. Mineralogist, 83: 896-900.

Staatz M.H., Murata K.J. & Glass J.J., 1955 - Variation of com¬

position and physical properties of tourmaline with its posi¬

tion in thè pegmatite. Amer. Mineralogist, 40: 789-804.

Stanèk 1, 1973 - Minerai paragenesis of thè new lithium-bearing

pegmatite at Lastovicky, western Moravia, Czechoslovakia.

Scripta Nat. Fac. UJEP, 3: 1-14.

Stanèk J., 1992 - Novà Ves near Cesky Krumlov, a pegmatite dike

of thè petalite subtype penetrating serpentinite. In: Novàk M.

& Cerny P. (eds). Field Trip Guidebook, International Sympo¬

sium on Mineralogy, Petrology and Geochemistry of Granitic

Pegmatites, Lepidolite 200, Nové Mèsto na Moravé, August-

September 1992: 57-62.

Stanèk J. & Povondra P, 1987 - Elbaites from Recice, western

Moravia. Acta Mas. Moraviae, Sci. nat., 72: 35-42.

Taylor M.C., Selway J.B. & Novàk M., 1997 - Tourmaline as a

recorder of pegmatite evolution: Belo Horizonte No. 1 peg¬

matite, Peninsular Ranges Batholith, southern California. In:

Abstracts, International Symposium on Tourmaline, Nové

Mèsto na Moravé, July 1997: 121-122.

Zagorskyi V.E. & Peretyazhko I.C., 1992 - Pegmatites with

gemstones of Central Transbaikalia. Nauka, Novosibirsk. (in

Russian).

Memorie della Società Italiana di Scienze Naturali e del Museo Civico di Storia Naturale di Milano

Volume XXX - pp.: 57-72, 2000

New Data for Feldspars and Micas from Granitic Pegmatites in thè

Southwestern USA

Boris M. Shmakin*1, Victor Ye. Zagorsky1, Eugene E. Foord\

Michael E. Brownfield2 & Paul H. Briggs2

1 Vinogradov Institute of Geochemistry Siberian Division, Russian Academy of Sciences P.O. Box 4019

Irkutsk, Russia 664033

2 M.S. 905 U.S. Geologica l Survey Box 25046 Denver Federai Center Denver ; CO 80225

Abstract - A total of 73 K-feldspars were partially chemically analyzed (Na, K and selected trace elements) and X-ray

structural States were determined on a total of 43 (of thè 73 total samples). A total of 46 new analyses (major, minor and

trace elements) for muscovite (30), biotite (1), zinnwaldite (protolithionite and Li-phengite-muscovite) (2), lepidolite (12)

and cookeite (1) were also determined. Samples come from granitic pegmatites in thè southwestern USA, and include: Mey-

ers Quarry, Eight Mile Park district, near Canyon City, CO; Harding mine, Dixon, NM; Globe mine, Petaca district, NM;

White Picacho district, AZ; and several pegmatite districts in San Diego Co., CA.

Pegmatites examined range from simple to highly complexly evolved. The Meyers Quarry pegmatite is of thè rare-ele-

ment class, beryl-type; thè Harding pegmatite is also of thè rare-element class, complex-type, spodumene subtype, and thè

pegmatites of thè Petaca district are of thè rare-metal-muscovite class. The Harding pegmatite contains holmquistite, bityite,

and zinnwaldite in thè immediately adjacent schist, particularly on thè upper contact, indicative of loss of Li from thè peg¬

matite. The White Picacho pegmatites are of thè rare-element class, complex type, spodumene subtype. All of thè pegmatites

examined in San Diego Co., CA are of thè rare-element class, complex type, and mostly of thè lepidolite subtype (Cerny,

1991) with some in thè Pala district being of thè spodumene subtype. They contain miarolitic cavities as well.

As a whole, thè dementai contents in thè K-feldspars are dose to those from other rare-metal and miarolitic pegmatites of

thè world. Some regional geochemical pecularities for thè White Picacho rare-element pegmatites and for thè Little Three mi¬

arolitic pegmatites exist. In all of thè pegmatites studied, thè later generations of K-feldspar contain more Rb, Cs and TI than

earlier ones in thè same pegmatite body, reflecting fractionation and concentration of rare alkalies. The Rb/Cs ratio approach-

es 1 in some crystals from miarolitic cavities. In late-stage overgrowths (rims) of K-feldspar crystals (2 examples), there is an in-

creased content of Ba, reflecting resetting of thè pegmatite ‘Chemical clock’ due to opening of formerly closed systems.

All structurally studied K-feldspar samples are microcline or orthoclase. No sample has <0.84. However, there are

10 samples with At values <0.5 and 3 samples with Àt between 0.60 and 0.64. Five orthoclase samples from miarolitic cavi¬

ties, including two overgrowth samples have At values between 0.0 and 0.3, and At” values between 0.16 and 0.34, which are

lower than in thè surrounding microcline matrix. Because of efficient interchange between monoclinic K-feldspar (i.e. or¬

thoclase) and thè contained fluid in thè miarolitic cavities, continued ordering produced well-ordered orthoclase, e.g. Little

Three mine. The amount of cooling and thè influence of temperature and speed of crystallization are other important Con¬

trols on thè structural state of thè feldspar.

The muscovite and to a lesser extent, thè lepidolite compositions group into distinct populations. Lepidolites range from

about 30 to 57 atomic % Li, clustering around trilithionite. Trace and minor elements detected in significant quantities

(>0.03%) include: B, Ba, Nb, Sn and Zn. Sn and Zn are highest in muscovites. Ga is low in lepidolites but may be as much

as 400 ppm in muscovites. Boron is low in all lepidolites except for one sample from thè Little Three mine, and may be as

much as 1500 ppm in muscovite.

Riassunto - Sono stati parzialmente analizzati 73 campioni di K-feldspato ( Na, K ed alcuni elementi in tracce); tra

questi, 43 sono stati studiati ai raggi X per determinarne lo stato strutturale. Sono state inoltre eseguite 46 nuove analisi

(maggiori, minori ed elementi in tracce) su campioni di muscovite (30), biotite (1), zinnwaldite (protolitionite e Li-fengite-

muscovite) (2), lepidolite (12) e cookeite (1). I campioni provengono da pegmatiti granitiche nel sud-ovest degli USA, e

comprendono: la miniera di Meyers Quarry, nel distretto di Eight Mile Park, vicino a Canyon City, CO; la miniera Harding,

Dixon, NM; la miniera Globe, nel distretto di Petaca, NM; il White Picacho district, AZ; e numerosi distretti pegmatitici nel¬

la contea di San Diego, CA.

Le pegmatiti esaminate vanno da semplici a fortemente complesse ed evolute. La pegmatite di Meyers Quarry è della

classe ad elementi rari, del tipo a berillo; La pegmatite di Harding è ancora della classe ad elementi rari, di tipo complesso,

del subtipo a spodumene, e le pegmatiti del distretto di Petaca sono della classe ad elementi rari a muscovite. La pegmatite

di Harding contiene holmquistite, bityite, e zinnwaldite negli scisti immediatamente adiacenti, soprattutto al contatto di tet¬

to, ad indicare una perdita di Li durante la cristallizzazione. Le pegmatiti di White Picacho sono della classe ad elementi

rari, di tipo complesso, del subtipo a spodumene. Tutte le pegmatiti prese in considerazione nella contea di San Diego, CA,

sono della classe ad elementi rari, di tipo complesso, e per lo più del subtipo a lepidolite (Cerny, 1991) con alcune, nel dis¬

tretto di Pala, del subtipo a spodumene. Anch’esse contengono cavità miarolitiche.

In generale, i contenuti in elementi nei K-feldspati sono prossimi a quelli di altre pegmatiti miarolitiche ad elementi rari

nel mondo. Vi sono tuttavia alcune particolarità geochimiche regionali nelle pegmatiti ad elementi rari di W hite Picacho e

nelle pegmatiti miarolitiche della miniera Little Three. In tutte le pegmatiti studiate, le generazioni più tardive di K-leldspa-

to contengono più Rb, Cs e TI rispetto alle generazioni precedenti, indicando un trazionamento ed una concentrazione di

* E-mail: pgmigor@igc.irk.ru

58

BORIS M. SHMAKIN - VICTOR YE. ZAGORSKY - EUGENE E. FOORD - MICHAEL F. BROWNFIELD & PAUL H. BRIGGS

alcali. Il rapporto Rb/Cs si avvicina a 1 in alcuni cristalli provenienti dalle cavità miarolitiche. Nelle sovracrescite tardive

(bordi) dei cristalli di K-feldspato (2 esempi), vi è un incremento del contenuto in Ba, che riflette un disturbo chimico a

causa dell’apertura del sistema.

Tutti i K-feldspati studiati da un punto di vista strutturale sono microclino ed ortoclasio. Nessun campione ha £t, < 0.84.

Tuttavia, vi sono 10 campioni con valori di At < 0.5 e 3 campioni con At compreso tra 0.60 e 0.64. Cinque campioni di orto¬

clasio provenienti da cavità miarolitiche, che comprendono due campioni con sovracrescite, hanno valori di Dt tra 0.0 e 0.3,

e valori di At” tra 0.16 e 0.34, che sono minori rispetto a quelli del microclino della matrice circostante. A causa dell’effi¬

ciente interscambio tra il K-feldspato monoclino (ad esempio ortoclasio) ed i fluidi contenuti nelle cavità miarolitiche, un

continuo processo di ordinamento ha prodotto un ortoclasio ben ordinato, come ad esempio alla miniera Little Three. Le

modalità del raffreddamento e l’influenza della temperatura e della velocità di cristallizzazione sono ulteriori importanti

fattori di controllo dello stato strutturale dei feldspati.

Le composizioni della muscovite ed in minor misura, della lepidolite, formano due distinte popolazioni. Le lepidoliti

presentano un contenuto in Li tra il 30 ed il 57% di occupanza del sito, addensandosi attorno alla trilitionite. Elementi in

traccia e minori rinvenuti in quantità significative (>0.03%) comprendono: B, Ba, Nb, Sn e Zn. Sn e Zn sono più abbondanti

nella muscovite. Il contenuto in Ga è basso nelle lepidoliti ma può raggiungere il valore di 400 ppm nelle muscoviti. Il boro

è basso in tutte le lepidoliti ad eccezione di un campione della miniera Little Three. Tale elemento può raggiungere valori

di oltre 1500 ppm nella muscovite.

Key-words: K-feldspar, muscovite, lepidolite, granitic pegmatites.

Introduction

In January, 1991, four first authors visited a num-

ber of pegmatite districts in thè southwestern USA

for thè purpose of collecting feldspar and mica sam-

ples to further study thè chemistry, geochemistry,

structural States of feldspars, and genesis of granitic

pegmatites of different modes of origin, and from

widely varied host rocks of different ages. All of thè

pegmatites examined are of Precambrian («1.2-1. 4

BY) age except for those of thè Southern California

Batholith, which are mid-Cretaceous (about 100

Ma). The Precambrian pegmatites are emplaced in-

to stable cratonic rocks while those of thè SCB are

related to orogenic processes (subduction). A suite

of 73 feldspar and 46 mica samples were chemically

analyzed at thè Vinogradov Institute of Geochem¬

istry, Russian Academy of Sciences, Irkutsk, Russia.

The feldspars were partially-, and thè micas were

completely-analyzed. Major, minor and trace ele-

ments (including boron) for thè micas were deter-

mined by ICP-AES at thè U.S. Geological Survey,

Denver, CO.

In addition, X-ray structural States were deter-

mined for 43 potash feldspars from granitic peg¬

matites of late Precambrian (Harding pegmatite, NM

(«1300 Ma) and White Picacho district pegmatites,

AZ) and Cretaceous (pegmatites in San Diego Coun-

ty, CA («95-100 Ma) age.

The geology and mineralogy of thè well-known

Harding pegmatite is best summarized in Jahns and

Ewing (1976), Brookins et al. (1979), and Chak-

oumakos and Lumpkin (1990). While extensive geo¬

logie and mineralogie studies of thè Harding peg¬

matite exist, detailed studies of thè potash feldspars

are stili lacking. Knowledge of thè geology and min¬

eralogy of thè pegmatites in thè White Picacho dis¬

trict is based on studies by Jahns (1952), London

(1979), and London and Burt (1978, 1982). Numerous

geologie and mineralogie studies exist of thè gem-

and specimen-bearing granitic pegmatites of San

Diego County, California, and only a few of thè more

important and review-type papers are cited here: Fo-

ord (1976, 1977), Foord et al. (1986, 1989, 1991), Jahns

(1979), Jahns and Wright (1951), Shigley and Brown

(1985), Stern et al. (1986), and Taylor et al. (1979).

Sample locations

Samples of micas and feldspars were collected

from thè following pegmatites: Meyers Quarry, Eight

Mile Park district, Canyon City, CO; Harding mine,

Dixon, NM; Globe mine, Petaca district, NM; White

Picacho pegmatites, NE of Wickenburg, AZ; and var-

ious pegmatites in San Diego Co., CA. The localities

sampled in San Diego Co. include: Pala district- Eliz¬

abeth R-Ocean View mine, White Queen mine, Pala

Chief mine, Stewart mine, and thè Naylor Rock (a

float block); Ramona district- Little Three mine;

Mesa Grande district- Himalaya mine and San Diego

Tourmaline mine. Fig. 1 is a map of thè southwestern

USA showing thè major areas visited.

Figure 2 shows thè locations of major pegmatites

in thè White Picacho district, AZ and Fig. 3 shows thè

locations of thè major pegmatite districts in San

Diego Co., CA.

Fig. 1 - Map of thè southwestern USA showing locations of peg¬

matite districts sampled.

NEW DATA FORD FELDSPARS AND MICAS FROM GRANITIC PEGMATITES IN THE SOUTHWESTERN USA

59

Sample descriptions

Location information, no. of samples analyzed from

each location, brief sample descriptions and Chemical

data for potash feldspar samples examined are given in

Table 1 below.Table 2 contains location and sample de-

scription information for mica samples examined in

this study. Similar information is also included for 16

additional samples of micas from San Diego County

that were previously analyzed and are in thè literature.

Data from a total of 17 mines (including those

from previous literature) are given here.

Analytical methods

Feldspars

Samples of potash feldspars were collected from

various zones of thè pegmatites. The samples were

purified by hand-picking under a binocular micro¬

scope and in heavy liquids. Chemical analyses for thè

alkalies (Na, K, Rb, Cs) were done by L.S. Tauson in

thè Institute of Geochemistry, Irkutsk. Ba, Sr, Pb and

TI were determined by quantitative atomic emission

spectroscopy by S. K. Yaroshenko and N.L. Chu-

makova, also in thè Institute of Geochemistry.

X-ray powder diffraction data were obtained us-

ing a DRÒN-3 diffractometer, with thè following op-

erating conditions: 35 Kv, 20 Ma, Ni-filtered Cu Ka

radiation,scan speed 0.5o/min.,chart speed 1 cm/min.

The sum of Al in Tj positions^h) was determined

with a rapid method using thè formula: = 7.3280 -

0.7266(À 20(204-060). The difference in Al occupan-

cies in TjO and T,m (At) was calculated as At =

0.0144+1. 244(A 20(131-131). These two equations

were derived by Afonina (1995) and were calculated

on thè basis of study of K-feldspar samples for which

structures and celi parameters are well known (Dal

Negro et al ., 1978, 1980; Griffen and Johnson, 1984;

Kroll and Ribbe, 1983, 1987; Afonina, 1995).

Additional parameters, At’ and At” were also cal¬

culated for cases involving one-step and two-step

types of ordering respectively (Afonina, 1995): At’ =

24.629 - 2.710 (A 20(204-060) and At” = 9.521-1.018 (A

20(204-060) when thè At value is relatively small, or

At” = 37.836 - 4.225 (A 20(204-060), when thè At val¬

ue is relatively large.

In samples with At <0.2, thè value of At” is dose to

At (first step of thè two-step ordering type) whereas

in samples with At >0.4, thè value of At normally is

dose to At’ (one-step type of ordering). For samples

with At between 0.2 and 0.4, as a rule, we have

At”<At<At’ because of intermediate type of ordering.

60

BORIS M. SHMAKIN - VICTOR YE. ZAGORSKY - EUGENE E. FOORD - MICHAEL F. BROWNFIELD & PAUL H. BRIGGS

r

n

For K-feldspar samples with At=0, thè value of At”

is greater than 0. As a rule, they have >0.72. As

was shown by TEM studies (Bambauer et al, 1989)

there is a monoclinic celi in K-feldspar with ^b <0.72.

For thè other K-feldspars with monoclinic X-ray sym-

metry, twinning or combination of microcline with

twinning structure in potash feldspar is present. This

is why determination of At for sudi feldspars will be

more correct using A20(204-060) or linear celi para-

meters, than using A20(131-131).

Micas

Mica samples were prepared by a variety of meth-

ods, depending on both sample and grain size and puri-

ty. Large lbooks’ or pieces of pure mica were cut up in

a rotary mica cutter and then ground to less than 200

mesh for Chemical analysis. Impure, and or fine-grained

micas were purified by hand-picking and heavy-liquid

separations if necessary. Then they were ground to less

than 200 mesh as for thè coarser samples.

Chemical analyses for major element oxides were

done by V.A. Grigorieva (Inst. of Geochemistry,

Irkutsk). Atomic-emission spectrographic analyses

for Ba, Sr, Pb, Zn, Sn and TI were done for a total of

53 micas by S.K. Yaroshenko and A.I. Kuznetsova, al-

so of thè Institute of Geochemistry (Irkutsk). A total

of 46 mica samples were analyzed by 40 element ICP-

AES methods by P.F. Briggs (USGS) using an Na202-

sinter method in carbon crucibles, so that B could be

quantitatively determined.

If only total Fe was determined, then a split of

FeO and Fe203 to 0.1 wt. % Fe203, with thè remain¬

der as FeO was made arbitrarilly for thè purpose of

formula calculations.

Data

Feldspars

Table 1 presents Chemical data for 73 K-feldspars

from thè Harding pegmatite, NM, pegmatites in thè

White Picacho district, AZ, and San Diego County, CA.

Table 3 presents structural state data for 43 of thè

K-feldspars.

NEW DATA FORD FELDSPARS AND MICAS FROM GRANITIC PEGMATITES IN THE SOUTHWESTERN USA

61

Table 1 - Contents of K, Na (Wt.%) and selected minor and trace elements (ppm) in potash feldpars from pec-

matites in thè southwestern USA. Number of samples used in composites shown in parenthesies (e.g. WP-C3 (2)).

62

BORIS M. SHMAKIN - VICTOR YE. ZAGORSKY - EUGENE E. FOORD - MICHAEL F. BROWNFIELD & PAUL H. BRIGGS

Table 2 - Locations and descriptions for micas from granitic pegmatites in thè southwestern USA.

NEW DATA FORD FELDSPARS AND MICAS FROM GRANITIC PEGMATITES IN THE SOUTHWESTERN USA

63

Micas

Table 2 contains location data, and sample de-

scriptions for thè 44 mica samples studied; Table 4

contains major element chemistry for micas, includ-

ing some from thè literature, and Table 5 contains

trace and minor element data for thè micas. For com-

parison analyces of lepidalites published earlier

(Stevens, 1938) are included.

Discussion or data

Feldspars

Principal observations that may be made from

these Chemical and structural state data are: (1) thè

levels of K and Na, as well as trace and minor ele-

ments are very different, but are within thè limits of

studied K-feldspars of other rare-metal and mi-

arolitic cavity type granitic pegmatites (Foord, 1976,

1977; Foord et al., 1979; Horsky and Martin, 1977;

Shmakin, 1979; Zagorsky and Kuznetsova, 1990;

Zagorsky and Peretyazhko, 1992); (2) lithium con-

tents are very low or essentially absent (<30 ppm) in

K-feldspars from thè Harding pegmatite, pegmatites

of thè White Picacho district, as well as thè Little

Three and Himalaya mines, but feldspars from peg¬

matites in thè Pala district contain from 60 to 400

ppm (this report; Foord et al., 1979). As much as 700

ppm Li is present in K-feldspars from thè Ocean

View-Elizabeth R pegmatite in thè Pala district (E. E.

Foord, unpub. data); (3) thè contents of Rb and Cs in

pegmatites of thè White Picacho district, AZ are rel-

atively high, but thè contents of Pb and TI are lower

than for feldspars from thè Harding mine. This ap-

pears to be a locai regional phenomenon; (4) thè

maximum overall concentration levels of Rb, Cs, Pb

Table 3 - X-ray crystallographic results for potash feldspars from pegmatites in New Mexico, Arizona and Cali¬

fornia.

64

BORIS M. SHMAKIN - VICTOR YE. ZAGORSKY - EUGENE E. FOORD - MICHAEL F. BROWNFIELD & PAUL H. BRIGGS

Table 4 - Major element chemistry for pegmatite micas from thè southwestern USA.

NEW DATA FORD FELDSPARS AND MICAS FROM GRANITIC PEGMATITES IN THE SOUTHWESTERN USA

65

Table 5 - Selected trace-element composition of pegmatite micas from thè southwestern USA. ND. Not deter-

mined.

and TI are in K-feldspars from thè Harding mine

rare-metal pegmatite, NM.; (5) thè lowest contents of

all examined elements for miarolitic-cavity-bearing

pegmatites are in K-feldspars from thè Little Three

mine, Ramona, CA and thè Pala Chief mine, Pala,

CA; (6) moderate levels and minimal variations of

thè studied elements and relatively high Rb contents

are found in K-feldspars from thè Himalaya dike Sys¬

tem; (7) medium levels of Rb, Cs and Tl, but very low

Pb levels are present in K-feldspars from thè Stewart

mine; (8) wide variations in Rb and Cs contents are

present in K-feldspars from thè Elizabeth R mine

and for Pb and Tl from thè White Queen mine; (9) in

thè majority of cases, thè later generations of K-

feldspar contain more Rb, Cs and Tl than earlier

ones; productive dikes have higher levels of these el¬

ements than non-productive ones; (10) thè geochem-

ical data for Sr are not sufficient (many values are

less than 20 ppm) to permit satisfactory interpreta-

tion; (11) as a rule, K-feldspars contain more Ba than

Sr; only thè K-feldspars from thè Harding mine con¬

tain more Sr than Ba (3 of four samples); (12) higher

levels of Ba are present in late-stage overgrowths

(rims) on two substrate-overgrowth pairs; (13) thè

maximum contents of Ba and Sr are present in early-

stage K-feldspar from fine-grained, graphic peg¬

matite from thè Stewart mine, and thè Cs content is

minimal (sample no. 165). Li and Rb show a bimodal

66

BORIS M. SHMAKIN - VICTOR YE. ZAGORSKY - EUGENE E. FOORD - MICHAEL F. BROWNFIELD & PAUL H. BRIGGS

Pegmatite Feldspars, SW USA

Li (ppm)

Fig. 4 - Li vs. Rb distribution for feldspars.

distribution (Fig. 4): 1) a positive correlation between

thè two elements, and 2) for one group an enrichment

in Rb but not in Li. Samples from thè Pala district are

high in Li and are from pegmatites that contain spo-

dumene. Samples from thè White Picacho, Harding

(has spodumene), Little Three and Mesa Grande are

low in Li. Rb vs. K+Na (Fig. 5) and Rb vs. K also show

a binary distribution: 1) a low positive slope and 2)

high positive slope. Cs is concentrated in higher total

alkali containing feldspar. In generai, Cs increases

along with Rb. The Cs/Rb ratio approaches 1 in

feldspars from miarolitic cavities. Single samples

from thè Harding, White Queen and Little Three

mines contain >0.2 wt. % Cs.

A steep trend for Rb vs. Na+K is shown by sam¬

ples from thè Harding, White Picacho, and Mesa

Grande pegmatites, and a shallower, nearly fiat, trend

is shown for samples from thè Pala Chief, Elizabeth

R, White Queen and Little Three mines. Samples

from thè Stewart mine lie on both trend lines.

All 43 studied K-feldspar samples are microclines

or orthoclase with Az not less than 0.68. Microcline

from thè Harding pegmatite. Pala Chief mine, and

Himalaya mine have Az greater than 0.88 and Jb

greater than 0.94. No sample studied has a value of

Xb less than 0.84, which is usually indicative of no

monoclinic feldspars being present. However, based

on studies by Foord et al. (1979 and 1989), some or¬

thoclase is present in miarolitic cavities from thè Hi¬

malaya mine, thè Little Three mine, thè White Queen

mine and thè Elizabeth R mine. There are ten sam¬

ples with At values less than 0.5 and three samples

have At values between 0.60 and 0.64. These samples

have lowered values of Az, from 0.68 to 0.81 in thè

first group (ten samples) and 0.86 to 0.89 in thè sec-

ond one (three samples). All five samples of K-

feldspar from miarolitic cavities have At values be¬

tween 0.0 and 0.3, and At” values between 0.16 and

0.34. In thè translucent overgrowth (rim) on sample

D-94, thè values for all measured indexes are lower

than in thè centrai portion of thè crystal. Other sam¬

ples with lowered values of Az and At are from coarse-

grained and blocky feldspar crystals near quartz

cores (D-26, D-74, D-136, D-156 and D-173) or from

graphic structures (D-61, D-100 and D-165). Two of

them, contain extremely high contents of Rb (D-173)

or Ba (D-165), which can influence thè process of

structural ordering (Afonina, et al., 1978). Based on

additional structural state determinations of K-

feldspars from San Diego County (Foord et al., 1979,

1989), thè crystallization history of K-feldspar may be

summarized as such: two factors contribute to effi-

cient ordering of thè K-feldspar: 1) availability of an

aqueous fluid and 2) slow rate of cooling. Down to

about 350°C and perhaps lower, thè miarolitic cavity

area is dominateci by a fluid phase, and thè rate of

heat loss probably was slower there than near thè

contacts of thè dikes. As long as there is efficient in-

terchange between thè monoclinic K-feldspar and

thè fluid, ordering will continue. The result is well-or-

NEW DATA FORD FELDSPARS AND MICAS FROM GRANIT1C PEGMATITES IN THE SOUTHWESTERN USA

67

Pegmatite Feldspars, SW USA

10000

8000

£ 6000

Q_

Q.

_q 4000

a :

2000

0

12.5 13 13.5 14

K+Na (Wt. %)

Fig. 5 - Rb vs. Na+K distribution for feldspars.

dered orthoclase. As thè temperature falls below

about 400°C in a pegmatite body, thè monoclinic K-

feldspar should transform to triclinic K-feldspar (i.e.

microcline). The partly ordered orthoclase near thè

contacts inverts readily, and appears today as well-or-

dered microcline. Closer to thè ‘pocket zone’, where

thè orthoclase had become better ordered, thè inver-

sion was less efficient in thè time available, probably

because of impurity elements and because thè differ-

ence in free energy between thè (metastable) better-

ordered orthoclase and thè microcline was smaller

than between thè more poorly ordered orthoclase

and microcline near thè contacts. As a result, inter¬

mediate microcline is found. Finally in thè pockets

themselves, thè difference in free energy between

very well-ordered orthoclase and microcline was so

small that thè inversion did not occur in some cases.

Some pockets contain substrates and overgrowths of

well-ordered microcline, presumably as a result of

relatively quick cooling through thè stability field of

orthoclase, and others in which both are “stuck”

metastably in thè well-ordered orthoclase state. Al-

ternatively, this could be a result of crystallization at

low temperatures and/or possible irradiation effects

(Shmakin and Afonina, 1967).There are also Ba- and

Rb-containing feldspars which are “stuck”

metastably in thè orthoclase state. Sudden evacuation

of thè fluid medium from thè pocket environment af¬

ter formation of thè overgrowth could also prevent

nucleation of thè low-temperature polymorph.

Micas

Figure 6 is a summary (RrR2-R3 plot) of thè mica

compositions encountered. It is interesting to observe

thè presence of brown zinnwaldite (protolithionite

according to thè classification of Lapides et al. , 1977)

and a light gray, intermediate zinnwaldite-muscovite

(Li-phengite-muscovite according to Lapides et al.,

1977) in thè Vadito Formation (quartzite, schist and

amphibolite), immediately adjacent to thè Harding

pegmatite. Such micas are typical for exocontacts of

pegmatites, where two isomorphous series were es-

tablished: biotite — protolithionite — zinnwaldite (al-

kaline stage) and zinnwaldite or protolithionite — Li-

phengite-miscovite (with increasing degree of acidity,

see Zagorsky and Makrygin, 1976; Kuznetsova and

Zagorsky, 1984). The same micas are typical of rare

metal granites as well (Koval et al., 1972; Lapides et

al, 1977). The presence of an Li-enriched zone im¬

mediately adjacent to thè pegmatite has been previ-

ously recognized (Jahns and Ewing, 1976, 1977). The

presence of holmquistite (Li-bearing amphibole)

(London, 1986) and bityite, (an Li-bearing sheet sili-

cate) are other minerals indicative of Li-migration

out of thè Harding pegmatite. No lepidolite was ana-

lyzed in this study from thè Harding mine, even

though thè “rose muscovite” (Heinrich and Levinson,

1953), for which thè Harding mine is famous, iooks’

like lepidolite. Jahns and Ewing (1976) have summa-

rized thè variations in appearance and compositions

of thè principal micas occurring in thè Main Quarry

68

BORIS M. SHMAKIN - VICTOR YE. ZAGORSKY - EUGENE E. FOORD - MICHAEL F. BROWNFIELD & PAUL H. BRIGGS

of thè Harding mine. Lepidolite is most abundant in

thè “spotted rock” zone and other interior zones.

Much rose muscovite replaces spodumene but is also

present in a cleavelandite-rose muscovite zone. Much

of thè muscovite may be purple or lilac in color and

is difficult to distinguish from true lepidolite except

by analytical or optical methods. Based on similar

chemistry to thè above analyzed zinnwaldites (pro-

tolithionite, Li-phengite-muscovite), two zin¬

nwaldites were partially-analyzed by Post and Austin

(1993), though one was called ‘biotite’. The Globe

and other pegmatites in thè Petaca district are mini-

mally evolved, and contain Fe-bearing muscovite.

Similar results for total iron were obtained by Post

and Austin (1993). The determined partial chemistry

of thè micas from 26 mica samples from 16 mines in

thè Petaca District was remarkably uniform (Post

and Austin, 1993). All 36 muscovites studied by Post

and Austin (1993) from various pegmatite districts in

northern New Mexico were of thè 2Mj polytype. The

pegmatites in thè Petaca district are minimally

evolved and are of thè rare metal-muscovite class.

Two lepidolites from thè White Picacho district are

low Li-lepidolites. The lepidolites from pegmatites in

San Diego Co., range from several low-Li lepidolites

(147, 154, EEF-2, STV-2, STV-3, STV-4, STV-5), with

thè remainder spread out from ideal trilithionite to

higher and lower Li-contents (Fig. 6).

The trace element data indicate that muscovite

may be enriched in Sn more than four times that of

lepidolite and more than 10 times that of biotite and

zinnwaldite (Fig. 7). Sn content is highest (to 1100

ppm) in Himalaya mine muscovite and in some cases

may be due to minute inclusions of cassiterite. Sn is

distinctly lower in lepidolites from this mine, reflect-

ing prior extraction of Sn by cassiterite and mus¬

covite. Samples of muscovite from thè Harding mine

have thè lowest Sn content of all of thè studied micas,

indicative of thè high degree of fractionation of that

pegmatite. Zn contents are high in biotite and some

muscovites and lepidolites. This is a function of thè

geochemical individualities of thè host pegmatite(s).

More than 1000 ppm Zn is present in biotite from thè

Meyers Quarry as well as in thè muscovite from thè

Globe mine. Zn contents in micas from thè White Pi¬

cacho pegmatites are also elevated. TI is elevated in

lepidolite relative to most, but by no means all mus¬

covites. TI and Rb show a linear positive trend re-

flecting thè association of Rb and TI (Fig. 8). For

higher concentrations (e.g. Harding and Stewart

L i

Fig. 6 - Ternary diagram (Li-Al+Fe3+-Fe2++Mg) showing all mica compositions determined.

TI (ppm) f Sn (ppm)

NEW DATA FORD FELDSPARS AND MICAS FROM GRANITIC PEGMATITES IN THE SOUTHWESTERN USA

69

Pegmatite Micas, SW USA

1400

1200

1000

800

600

400

200

0

35 40 45 50 55

Si02 (wt. %)

- Sn vs. Si02 distribution in micas.

Pegmatite Micas, SW USA

Fig. 8 - TI vs. Rb:0 distribution in micas.

Li20 (wt. %) % Li20 (wt. %)

70

BORIS M. SHMAKIN - VICTOR YE. ZAGORSKY - EUGENE E. FOORD - MICHAEL F. BROWNFIELD & PAUL H. BRIGGS

Pegmatite Micas, SW USA

35 40 45 50 55

Si02 (wt. %)

f Meyers ▲ Harding + Petaca O White Picacho • Stewart

7 Mesa Grande V Little Three A Elizabeth R ■ Pala Chief X White Queen

- Li:0 vs. Si02 distribution in micas.

Pegmatite Micas, SW USA

0 2 4 6 8

F (wt. %)

Fig. 10- Li:0 vs. F distribution in micas.

NEW DATA FORD FELDSPARS AND MICAS FROM GRANITIC PEGMATITES IN THE SOUTHWESTERN USA

71

mines), this correlation is not so obvious. Ga is low

(about 100 ppm) in lepidolite relative to muscovite

(100 to 400 ppm). Nb contents vary from low to high

for both muscovites and lepidolites. B is low (about

300 ppm) in lepidolite (except for one sample, EEF-

2, from thè Little Three mine) but may be 600 ppm or

more in some muscovites. The Little Three mine is

anomalous in its behavior of B, as evidenced by B-

containing lepidolite and thè presence of thè new

minerai boromuscovite (Foord et al , 1991). Two sam-

ples of muscovite contain significant B contents: Hi-

malaya mine 220B (1100 ppm B) and White Picacho

38 (1500 ppm B). One sample (EEF-2) of pale purple,

fine-grained, low-Li lepidolite from a frozen pocket

in thè main dike on thè Little Three mine contained

0.35 wt. % B203.

A dose clustering of muscovites and a somewhat

less dense clustering of lepidolites is shown on an

Li20 vs. Si02 plot (Fig. 9). Micas from thè Harding

mine show a negative rather than positive trend for

Li20 vs. Si02. F vs. Si02 is very similar to Li20 vs.

SiÒ2. A linear positive trend is shown by Li20 vs. F

(Fig. 10), reflecting thè mica evolution with time. H20

and F show a strong negative correlation reflecting

thè coupled substitution of thè two (Fig. 11). All of

these plots show thè distinction between thè species

of micas analyzed.

CONCLUSIONS

This study of additional K-feldspars and micas

from previously studied granitic pegmatites in thè

southwestern USA, has confirmed observations and

conclusions published previously. Expected demen¬

tai fractionation trends are present in both K-

feldspars and micas. The greater thè degree of peg-

matite evolution, thè higher thè content of Li, Rb, Cs,

and TI in both feldspars and micas, along with F and

Li in micas. The confirmation of two types (colors) of

zinnwaldite has been established from thè Harding

mine. A significanti greater number of complete mi¬

ca analyses now exists for all of these pegmatites. The

amount of B present in thè pegmatitic micas was of

particular interest, but only three samples were found

to have >1100 ppm B.

Structural state studies of K-feldspars show thè

existence of orthoclase rims and crystals in cavities

(Pala, CA) with Àt values of 0.0-0. 3 and Àt” values of

0.16-0.34. Rim feldspar is enriched in Ba and Sr along

with Pb but depleted in Rb, Cs, and TI in comparison

with core feldspar.

Acknowledgments

The authors thank V. A. Grigoryeva of thè Vino-

gradov Institute of Geochemistry, Russian Academy

of Sciences, Irkutsk, Russia for her careful analyses of

43 mica samples. We also thank Gaiina G. Afonina

(now deceased) for X-ray determination of thè struc¬

tural States of thè feldspars examined. We thank thè

following mine owners in San Diego Co., CA for ac-

cess to their mines for sampling purposes: Blue Shep-

pard (Stewart mine); L.B. Spaulding, Jr. (Little Three

mine); Roland R. Reed (Ocean View-Elizabeth R

mines), Robert Dawson (White Queen mine), and

William F. Larson and John McLean (Himalaya

mine). Reviews of this manuscript were provided by

W. B. Simmons, Jr. and Peter J. Modreski.

Pegmatite Micas, SW USA

V

4 6 8 10 12

H20 (Wt. %)

Fig. 11- H:0 vs. F distribution in micas.

72

BORIS M. SHMAKIN - VICTOR YE. ZAGORSKY - EUGENE E. FOORD - MICHAEL F. BROWNFIELD & PAUL H. BRIGGS

References

Afonina G.G., 1995 - Determination of Ai-Si ordering trend and

quantities of Al in tetrahedral positions of potash feldspars us-

ing a powder X-ray method. Zapiski Vseross. Minerai. Ob-

shchestva , 124 (3): 65-79. (In Russian)

Afonina G.G., Makagon V.M. and Shmakin B.M., 1978 - Bari-

ura- and rubidium-containing potassium feldspars. Nauka

Publishing House, Novosibirsk, lllp. (in Russian)

Bambauer H.U., Krause C. and Kroll H„ 1989 - TEM-investi-

gation of thè sanidine/microcline transition across metamor-

phic zones: thè K-feldspar varieties. European Journal of Min-

eralogy, 6: 47-58.

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

Landis G.P. and Register M.R., 1979 - The Harding Peg-

matite. Sunimary of recent research. 30th Field Conference,

New Mexico Geol. Soc. Guidebook : 127-133.

Cerny R, 1991 - Rare-element granitic pegmatites. Part I: anato-

my and internai evolution of pegmatite deposits. Geoscience

Canada, 18 (2): 49-67.

Chakoumakos B.C. and Lumpkin G.R., 1990 - Pressure-tempera-

ture constraints on thè crystallization of thè Harding peg¬

matite, Taos County, New Mexico. Canadian Mineralogist, 28:

287-298.

Dal Negro A., De Pieri R., Quareni S. and Taylor W.H., 1978

- The crystal structures of nine K-feldspars from thè Adamel-

lo massif (northern Italy). Ada Crystallographica, B34: 2699-

2707.

Dal Negro A., De Pieri R. and Quareni S., 1980 - The crystal

structures of nine K-feldspars from thè Adamello massif

(northern Italy). Ada Crystallographica , B36: 3211.

Foord E. E., 1976 - Mineralogy and petrogenesis of layered peg-

matite-aplite dikes in thè Mesa Grande district. San Diego

County, California. Ph.D. thesis Stanford University, Stanford,

CA. 326 pp.

Foord E. E., 1977 - Famous Minerai Localities. The Himalaya

dike System, Mesa Grande district, San Diego, California.

Mineralogical Record , 8: 461-474.

Foord E. E., Martin R.F. and Long P.E., 1979 - Potassio feldspars

from pocket pegmatites, San Diego County, California. Geo-

logical Society of America Annual Meeting, Abstracts with

programs, 11 (7) 427p.

Foord E. E., Starkey H.C. and Taggart J.E. Jr., 1986 - Mineral¬

ogy and paragenesis of ‘pocket clays’ and associated minerals

in complex granitic pegmatites. San Diego County, California.

American Mineralogist, 71: 428-439.

Foord E. E., Spaulding L.B., Mason R.A. and Martin R.F., 1989

- Mineralogy andparagenesis of thè Little Three mine peg¬

matites, Ramona district. San Diego County, California. Min¬

eralogical Record, 20: 101-127.

Foord E.E., London David, Kampf A.R., Shigley J.E. and Snee

L.W., 1991 - Gem-bearing pegmatites of San Diego County,

California. In: Walawender M.J. and Hanan B.B. eds., Geolog-

ical Excursions in Southern California and Mexico. Guide¬

book for thè 1991 Annual Meeting, Geological Society of

America , San Diego, California, October 21-24: 128-146.

Foord E.E., Martin R.F., Fitzpatrick J.J., Taggart J.E. Jr. and

Crock J.G., 1991 - Boromuscovite, a new member of thè mica

group, from thè Little Three mine pegmatite, Ramona district,

San Diego County, California. American Mineralogist, 76:

1998-2002.

Griffen D.T. and Johnson B.T., 1984 - Strain in triclinic alkali

feldspars. A crystal structure study. American Mineralogist, 69:

1072-1077.

Heinrich E.Wm. and Levinson A. A., 1953 - Studies in thè mica

group; mineralogy of thè rose muscovites. American Mineral¬

ogist, 38: 25-49.

Horsky S.J. and Martin R.F., 1977 - The anomalous ion-ex-

change behavior of “ordered” orthoclase. American Mineralo¬

gist, 62: 1191-1199.

Jahns R.H., 1952 - Pegmatite deposits of thè White Picacho Dis¬

trict, Maricopa and Yavapai Counties, Arizona. Arizona Bu¬

reau of Mines Bulletin, 162, 105 pp.

Jahns R.H., 1979 - Gem-bearing pegmatites in San Diego Coun¬

ty, California. The Stewart Mine, Pala district and thè Hi¬

malaya mine, Mesa Grande district. In: Abbott P.L. and Todd

V.R. eds., Mesozoic Crystalline Rocks: San Diego, California.

San Diego State University, Dept. of Geological Sciences : 3-38.

Jahns R.H. and Ewing R.C., 1976 - The Harding mine, Taos

County, New Mexico. New Mexico Geological Society Guide¬

book, 27th Field Conference, Vermejo Park: 263-276.

Jahns R.H. and Ewing R.C., 1977 - The Harding Mine, Taos

County, New Mexico. Mineralogical Record, 2: 115-126.

Jahns R.H. and Wright L.A., 1951 - Gem- and lithium-bearing

pegmatites of thè Pala district, San Diego County, California.

California Division of Mines and Geology, Special Report 7-

A.

Koval P.V., Kovalenko V.I., Kuzmin M.I., Pisarskaya V.A. and

Yurchenko S.A., 1972 - Minerai parageneses, composition,

and nomenclature of micas from rare-metal albite-bearing

granitoids. Doklady Akad. Nauk SSSR, 202 (5): 1174-1177.

Kroll H. and Ribbe P.H., 1983 - Lattice parameters, composition

and Ai-Si ordering in alkali feldspars. In: P.H. Ribbe ed.,

Feldspar mineralogy. Mineralogical Society of America, Re-

views in Mineralogy , 2: 57-99.

Kroll H. and Ribbe P.H., 1987 - Determining (Ai-Si) distribution

and strain in alkali feldspars using lattice parameters and dif-

fraction peak positions. A review. American Mineralogist, 72:

491-506.

Kuznetsova L.G. and Zagorsky V.Ye., 1984 - Micas from meta-

somatites of rare-metal field of spodumene pegmatites. Dok¬

lady Akad. Nauk SSSR, 275 (1): 151-155 (in Russian).

Lag ache M., 1984 - The exchange equilibrium distribution of al¬

kali and alkaline earth elements between feldspars and hy-

drothermal Solutions. In: Brown W. ed. Feldspars and

Feldspathoids. Reidei, Boston: 247-279.

Lapides I.L., Kovalenko V.I. and Koval P.V., 1977 - Micas of

rare-metal granitoids. Nauka Publishing House, Novosibirsk,

104 pp. (In Russian).

London D.A., 1979 - Occurrence and alteration of lithium miner¬

als, White Picacho pegmatites, Arizona. M.S. thesis, Arizona

State Univ., Tempe, AZ.

London D.A., 1986 - Holmquistite as a guide to pegmatitic rare

metal deposits. Economie Geology, 81: 704-712.

London D.A. and Burt D.M., 1978 - Lithium pegmatites of thè

White Picacho District, Maricopa and Yavapai Counties, Ari-

zona. In: Burt D.M. and Pewe T.L. editors. Guidebook to thè

geology of centrai Arizona. Arizona Bureau of Geology and

Minerai Technology, Special Paper 2: 61-72.

London D.A. and Burt D.M., 1982 - Alteration of spodumene,

montebrasite, and lithiophilite in pegmatites of thè White Pi¬

cacho District, Arizona. American Mineralogist, 67: 97-113.

Post J.L. and Austin G.S., 1993 - Geochemistry of micas from

Precambrian rocks of northern New Mexico. Circular 202,

New Mexico Bureau of Mines and Minerai Resources, 20 pp.

Sebastian A. and Lagache M., 1990 - Experimental study of thè

equilibration between pollucite, albite, and hydrothermal flu¬

id in pegmatitic systems. Mineralogical Magazine, 54: 447-454.

Shigley J.E. and Brown G.E., 1985 - Occurrence and alteration

of phosphate minerals at thè Stewart pegmatite, Pala district.

San Diego County, California. American Mineralogist, 70: 395-

408.

Shmakin B.M., 1979 - Composition and structural state of K-

feldspars from some U.S. pegmatites. American Mineralogist,

64: 49-56.

Shmakin B.M. and Afonina G.G., 1967 - X-ray determination of

triclinicity in potash feldspars for thè solution of rock genesis

problems. Doklady Akad. Nauk SSSR, 173 (2): 417-420. (In

Russian)

Stern L. A., Brown G.E. Jr., Bird D.K., Jahns R.H., Foord E.E.,

Shigley J.E. and Spaulding L.B. Jr., 1986 - Mineralogy and

evolution of thè Little Three pegmatite-aplite layered intru¬

sive, Ramona, California. American Mineralogist, 71: 406-427.

Stevens R.L., 1938. New analyses of lepidalites and their inter-

pretation. American Mineralogist, 23: 607-628.

Taylor B.E., Foord E.E. and Friedrichsen H., 1979 - Stable iso¬

tope and fluid-inclusion studies of gem-bearing granitic peg¬

matite-aplite dikes, San Diego County, California. Contribu-

tions to Mineralogy and Petrology, 68: 187-205.

Zagorsky V.Ye. and Makrygin A.I., 1976 - Evolution of mica

composition at thè exocontacts of tantalum-bearing peg¬

matites. Geokhimiya, 9: 1362-1369. (In Russian)

Zagorsky V.Ye. and Kuznetsova L.G., 1990 - Geochemistry of

spodumene pegmatites and alkali-rare metal metasomatites.

Nauka Publishing House, Novosibirsk, 141 p. (In Russian)

Zagorsky V.Ye. and Peretyazhko I.S., 1992 - Gem-bearing peg¬

matites of Central Transbaikalia. Nauka Publishing House,

Novosibirsk, 224 p. (in Russian)

Memorie della Società Italiana di Scienze Naturali e del Museo Civico di Storia Naturale, Milano

Voi. XXX, pp. 73-109, 2000

Crystal structures of alkali feldspars from granitic pegmatites:

A review

Achille Blasi* and Carla De Poi Blasi*

Dipartimento di Scienze della Terra, Università degli Studi, Via Botticelli 23, 20133-Milano, Italy

Abstract - Granitic pegmatites represent a unique naturai laboratory that is extremely efficient in thè synthesis of alkali feld¬

spars of different structural States. The crystal structure determinations available to date for alkali feldspars from granitic pegma¬

tites were obtained from 29 different specimens, of which 22 are K-rich feldspars, 6 are Na-rich feldspars, and 1 is a crypto-perthitic

intergrowth of K- and Na-rich feldspar. The structure determinations of 26 specimens were carried out by single-crystal X-ray dif-

fraction methods; 1 specimen of “ordered orthoclase” and 2 specimens of low albite were investigated by single-crystal neutron dif-

fraction. In view of their importance, thè crystal structures of some specimens were refined using different sets of intensity data

from thè same specimen or re-refined subsequently. Some structure determinations have become obsolete but remain of historical

value. The vast majority of crystal structures were determined using naturai specimens at room temperature: naturai K-rich

feldspars include specimens of low microcline ( Pellotsalo , 2 specimens; Prilep , 2 specimens; Pikes Peak , 3 specimens; Khibiny), in¬

termediate microcline ( Spencer U\ Adamello P17C), “ordered orthoclase” ( Himalaya , investigated by neutron diffraction), ortho¬

clase ( Khibiny ; Spencer C; Itrongay, 2 specimens), and sanidine ( Rabb Canyon ; Back Claim, 4 specimens); naturai Na-rich feldspars

include specimens of low albite ( Ramona ; Amelia , 2 specimens, one of them investigated by neutron diffraction) and high albite

( Rabb Canyon, thè same cryptoperthitic specimen as that used for determining thè crystal structure of thè coexisting sanidine).

Some specimens were thermally treated and quenched to room temperature: quenched specimens were then used in structure re-

finements at room temperature (high sanidine produced by pre-heating of either orthoclase, Itrongay, a third specimen, Spencer C,

or low microcline, Bedford ; high albite produced from pre-heating of low albite, Amelia, a third specimen) or at high temperatures

(monalbite produced from pre-heated low albite, Amelia, a fourth specimen). Some naturai specimens were used in structure de¬

terminations at low temperatures (low microcline at 163 K, Prilep , thè same specimen as one of thè two specimens investigated at

room temperature; orthoclase at 121 K, Itrongay, thè same specimen as that investigated at room temperature; low albite at 93 K,

Ramona, thè same specimen as that investigated at room temperature; low albite at 13 K, Amelia, a fifth specimen, investigated by

neutron diffraction). Appropriate Tables are presented in thè Appendix; these contain information on geological occurrence of thè

specimen as well as compositional and crystallographic data. A review is presented for each specimen as regards metric and struc¬

tural properties as well as composition and degree of lattice strain due to coherency between exsolved phases. Aspects of particu-

lar interest are emphasized and indicated as subjects for future work. Reference values for Al-O and Si-O distances have been pro-

posed for low microcline and low albite on thè basis of thè data available to date, including those from non-pegmatitic alkali

feldspars.

Riassunto - Le pegmatiti granitiche rappresentano un laboratorio naturale unico, che è estremamente efficiente nel sintetizzare

feldspati alcalini aventi stati strutturali differenti. Le determinazioni di struttura cristallina attualmente disponibili per feldspati al¬

calini provenienti da pegmatiti granitiche furono ottenute da 29 differenti campioni, dei quali 22 sono feldspati ricchi di K, 6 sono

feldspati ricchi di Na, e 1 è una intercrescita criptopertitica di feldspato ricco di K e di feldspato ricco di Na. Le determinazioni di

struttura di 26 campioni furono effettuate mediante metodi diffrattometrici a raggi X su cristallo singolo; 1 campione di “ortocla¬

sio ordinato” e 2 campioni di albite bassa furono studiati mediante diffrazione neutronica su cristallo singolo. Data la loro impor¬

tanza, le strutture cristalline di alcuni campioni furono raffinate usando serie diverse di dati di intensità raccolti per lo stesso cam¬

pione, oppure furono ri-raffinate in tempi successivi. Alcune determinazioni di struttura sono diventate obsolete ma conservano un

valore storico. In larga maggioranza le strutture cristalline furono determinate a temperatura ambiente usando campioni naturali:

i feldspati naturali ricchi di K includono campioni di microclino basso ( Pellotsalo , 2 campioni; Prilep, 2 campioni; Pikes Peak, 3 cam¬

pioni; Khibiny), di microclino intermedio ( Spencer U\ Adamello P17C ), di "ortoclasio ordinato" ( Himalaya , studiato mediante dif¬

frazione neutronica), di ortoclasio ( Khibiny ; Spencer C; Itrongay, 2 campioni ) e di sanidino ( Rabb Canyon', Back Claim, 4 campi¬

oni); i feldspati naturali ricchi di Na includono campioni di albite bassa ( Ramona ; Amelia , 2 campioni, uno dei quali studiato medi¬

ante diffrazione neutronica) e di albite alta ( Rabb Canyon, lo stesso campione criptopertitico usato per determinare la struttura

cristallina del sanidino coesistente). Alcuni campioni furono trattati termicamente e “quenched a temperatura ambiente: i cam-pi-

oni “quenched” furono poi usati per raffinamenti di struttura a temperatura ambiente (sanidino alto prodotto mediante pre-riscal-

damento sia di ortoclasio, Itrongay, un terzo campione, Spencer C, sia di microclino basso, Bedford ; albite alta prodotta me-diante

pre-riscaldamento di albite bassa, Amelia, un terzo campione) o ad alte temperature (monalbite prodotta mediante pre-riscalda-

mento di albite bassa, Amelia, un quarto campione). Alcuni campioni naturali furono usati in determinazioni di struttura a basse

temperature (microclino basso a 163 K, Prilep, lo stesso campione di uno dei due campioni studiati a temperatura ambiente, orto¬

clasio a 121 K, Itrongay, lo stesso campione studiato a temperatura ambiente; albite bassa a 93 K, Ramona, lo stesso campione stu¬

diato a temperatura ambiente; albite bassa a 13 K, Amelia, un quinto campione, studiato mediante diffrazione neutronica). Nella

Appendice vengono presentate appropriate Tabelle le quali contengono informazioni sulla provenienza geologica dei campioni in¬

sieme con i loro dati composizionali e cristallografici. Ogni campione è oggetto di “review per quanto riguarda proprietà metriche

e strutturali, come pure composizione e grado di “strain reticolare dovuto a coerenza tra fasi essolte. Aspetti di particolare inter¬

esse vengono evidenziati e indicati come argomenti per lavori futuri. Valori di riferimento per le distanze Al-O e Si-O vengono pro¬

posti per microclino basso e albite bassa sulla base dei dati disponibili finora, i quali includono anche i dati di feldspati alcalini non

pegmatitici.

Key words: alkali feldspar, crystal structure, Al-O distance, Si-O distance, granitic pegmatite.

* E-mail address: umimin@icil64.cilea.it

74

ACHILLE BLASI & CARLA DE POL BLASI

Introduction

Granitic pegmatites represent geological environ-

ments of particular significance for alleali feldspar min-

eralogy. The large amount of data available in thè lit-

erature on metric properties determined by routine X-ray

powder diffraction methods shows that alkali feldspars

from pegmatites exhibit a variety of structural States. This

means that granitic pegmatites can be regarded as a

unique naturai laboratory that is extremely efficient in

synthesizing different alkali feldspars.

The efficiency of thè pegmatite lab in producing a

variety of feldspar structures is largely due to thè action

of aqueous fluids. Their abundance influences thè sub-

solidus history of alkali feldspars in promoting exsolu¬

tion and Si,Al ordering, thè latter proceeding often to

completion. A sudden decrease in P(fluid) can be fol-

lowed by rapid nucleation and growth of highly disor-

dered feldspar material. In other circumstances, an ex-

plosive release of thè fluid medium from pocket systems

may occur, with consequent interruption of exsolution

and Si, Al ordering processes. As a result, granitic peg¬

matites may contain essentially fully ordered alkali feld¬

spars inverted from originally monoclinic precursors, as

well as structurally intermediate and essentially fully

disordered K-rich feldspars.

Circulation of aqueous fluids at low temperatures in

open systems can produce crystals of highly ordered K-

rich feldspar, that have developed directly within thè

field of stability of microcline. Otherwise, low-tempera-

ture hydrothermal processes can produce late genera-

tions of K-rich feldspars showing a range of structural

States. Rare examples of quenched pegmatites contain

sanidine and high albite in cryptoperthitic intergrowth.

Attention will be focused on all thè environmental

situations that were recorded by pegmatitic alkali feld¬

spars used in structure determinations. Metric and

structural properties, along with composition and degree

of lattice strain due to coherency between exsolved

phases, enable thè fine details of thè nature of alkali feld¬

spar to be properly explored.

To accomplish this thoroughly, a number of pegma¬

titic alkali feldspars were thermally treated and then

quenched in laboratory experiments: thè crystal structure

of these specimens was determined either at room tem¬

perature or at high temperatures. Otherwise, thè crystal

structures of unheated specimens were investigated at

low temperatures.

We will present a review of all these specimens, in-

cluding pioneering investigations that were of particular

significance to settle problems and to contribute to thè

development of modern aspeets of alkali feldspar min-

eralogy.

Methods

I he estimate of thè structural state from metric prop¬

erties was obtained by means of thè A method, which is

based on thè structural indicators A (bc) and A (b*c*),

normatized between 0.5 and 1.0, and A(ay) and A(a*y*),

normali/ed between 0.0 and 1.0 (see Blasi & De Poi

Blasi, i 794). The values of thè structural indicators were

calculated with thè procedure of Blasi (1977) and thè ref-

erence values suggested by Blasi & De Poi Blasi (1994, p.

68) for thè lattice constants of thè alkali feldspar end-

members. These reference values include thè data pro-

posed by Stewart & Wright (1974) for high sanidine and

high albite, those proposed by Kroll & Ribbe (1987) for

high albite (analbite), and those obtained from thè aver-

age of thè direct lattice constants proposed by Smith

(1974a), Stewart & Wright (1974), and Kroll & Ribbe

(1987) for low microcline. The new data for low microcline

are: a = 8.5931, b = 12.9643, c = 7.2221 À, a = 90.636, [3 =

115.948, y = 87.668°, V = 722.84 À3, a* = 0.129522, b* =

0.077201, c* = 0.153991 A1, a* = 90.428, (3* = 64.056, y* =

92.284°, V * = 0.0013834 À-3; thè lattice constants for thè

other end-members are reported in Table 5 of Blasi & De

Poi Blasi (1994). The resulting values of thè structural indi¬

cators A (bc), A (b*c*), A(ay), and A(cx*y*) may be slightly

different from thè corresponding data calculated by earli-

er authors using a different set of lattice constants for thè

alkali feldspar end-members: e.g., thè difference A (bc) -

A (b*c*), which was - 0.049 in Blasi et al. (1984b) for thè

Pellotsalo microcline (see also thè relevant section in this

work), becomes -0.047, as can be seen from thè data in

Table Al-a.

Compositions in terms of N0r contents were calculated

from V and a with thè equations of Kroll et al. (1986) and

Hovis (1986) listed in Table 1 of Blasi & De Poi Blasi

(1994). The equation number referred to a given N0r value

in thè present work, as is thè case of “N0r from V: 0.964

(eq. 8)” or “N0r(F: eq. 8) = 0.964”, corresponds to thè

equation number in Table 1 of Blasi & De Poi Blasi (1994).

The evaluation of thè degree of lattice strain due to co¬

herency between exsolved K- and Na-rich phases was ob¬

tained from thè difference N0r(a) - N0r(F) (Blasi & De Poi

Blasi, 1980, 1994; Blasi et al. , 1984b, 1987a) and thè strain

index A a - n(obs) - a(est) (Stewart & Wright, 1974), where

<z(est) is obtained from a contoured bc plot. Other deter¬

minative methods which are summarized in Blasi & De

Poi Blasi (1994), can also be used.

For thè sake of uniformity, thè values of mean T-O dis-

tances, <T- 0>, were recalculated with 4 decimals from in¬

dividuai T-O distances. The o values in (T-O) distances

were calculated from thè Law of Propagation of Errors

(see Blasi, 1977) using standard errors of individuai T-O

distances and assuming that (a) individuai T-O distances

can be considered as uncorrelated, and (b) co-variance

terms can be neglected.

The values of thè Al contents of thè T sites were cal¬

culated with thè procedure of Blasi & De Poi Blasi (1994,

eqs. 18 and 19) using individuai T-O distances as starting

data and thè new reference values for Al-O and Si-O sug¬

gested in this work (see thè relevant section). Thus, equa¬

tions (18) and (19) of Blasi & De Poi Blasi (1994) become:

ti = 0.25«rrO) - 1.614/«(T-0)> - 1.614) (1)

t, = 0.25 + 0.75«7>O) - <(T-0»/(1.738 - «T-O») (2)

for K-rich feldspars, and:

t, = 0.25«7>O) - 1.613/(«T-0)) - 1.613) (3)

ti = 0.25 + 0.75«r,-O> - «T-0)))/(1.742 - ((T-O))) (4)

CRYSTAL STRUCTURES OF ALKALI FELDSPARS FROM GRANITIC PEGMATITES: A REVIEW

75

for Na-rich feldspars. In these equations, ((T- O)) denotes

thè grand mean T-O distance, which is obtained from thè

average of thè 16 individuai T-O distances in thè 4 tetra-

hedra T\o, Tini , T2o, and Tini. Equations (1) and (2), or

(3) and (4), are to be used with (Tr O) distances shorter

and longer than ((T-O)), respectively. The relevant values

of a were calculated from thè Law of Propagation of Er-

rors (see Blasi, 1977) using standard errors of individuai

T-O distances, adopting thè same assumptions (a) and (b)

as those employed for calculating o values in (T-O) dis¬

tances, and assuming that Al-O and Si-O distances are

constants.

Tables Al and A2 of thè Appendix present information

on specimen description and geological occurrence, crystal

dimensions, lattice constants, structural indicators A (bc),

A (b*c*), A(ay), and A(a*y*), composition in terms of N0r

contents calculated from V and a, discrepancy factors and

number of diffractions used in structure refinement,

<T-0> distances and Al contents of T sites for K- and Na-

rich feldspars, respectively, from granitic pegmatites. In

each feldspar group thè sequence of specimens is present-

ed in decreasing Si,Al order, and, under thè same degree

of Si,Al order, in chronological order of publication of thè

papers concerned (thè oldest first). The information as-

sembled includes data regarding: (/) room-temperature

crystal structures of naturai specimens, (ii) room-tempera¬

ture crystal structures of pre-heated specimens, (Hi) high-

temperature crystal structures of pre-heated specimens,

(iv) low-temperature crystal structures of naturai speci¬

mens. Items (i), (ii), and (iv) are covered in both Tables Al

and A2; item (iii) is covered in Table A2. The organization

and contents of thè text follow thè presentation of data

given in Tables Al and A2.

Most of thè sets of reciprocai lattice constants were re-

calculated during this study: in such cases thè relevant s

values were not given. The values of N0r contents obtained

from V and a are meaningless for crystal structures deter-

mined at low or high temperatures. However, these values

were given in Tables Al and A2 for purposes of compari-

son with thè corresponding data obtained at room tem¬

perature.

In spite of thè considerations on thè swinging-arm ef-

fect in thè section on thè reference values for Al-O and Si-

O distances in alkali feldspars, thè Al contents of thè T

sites in K- and Na-rich feldspars used in structure de-

terminations at non-ambient temperatures were present-

ed in Tables Al and A2 for thè sake of comparison with

thè corresponding data obtained at room temperature.

Room-temperature crystal structures of

NATURAL K-RICH FELDSPARS

Pellotsalo low microcline. The Pellotsalo microcline is

a feldspar known in thè mineralogical literature since thè

end of thè last century (see Blasi et al., 1987a, for histori-

cal notes). B.E. Brown & Bailey (1964) attributed its geo¬

logical provenance to a granitic body on thè island of Pel¬

lotsalo, Lake Ladoga, Russia. By contrast, Blasi et al.

(1987a) documented that thè Pellotsalo microcline comes

from a granitic pegmatite. Among thè references men-

tioned for this purpose by Blasi et al. (1987a), see espe-

cially Eskola (1951, pp. 39-40). The originai sampling was

made before thè region changed from Finnish to Russian

control (see B.E. Brown & Bailey, 1964; also Eskola, 1951):

that is thè reason why thè sample locality is frequently de-

scribed in mineralogical literature as being from Finland.

The crystal structure determined by B.E. Brown & Bai¬

ley (1964) represented for more than twenty years thè best

reference point for thè structural state of low microcline.

Blasi et al. (1984b) analyzed in detail thè structure deter¬

mined by B.E. Brown & Bailey (1964) in thè context of a

study concerning thè crystal structures of two crystal frag-

ments from thè Pikes Peak low microcline. To accomplish

their analysis, Blasi et al. (1984b) prepared a total of 81 di-

agrams by plotting data points of individuai interatomic

distances and angles versus thè (Tìo-O) distance for all thè

specimens of K-rich feldspar available at that time. The

structure of thè Pellotsalo microcline, in spite of its vener¬

ale age and its R value of 0.104, showed data points in

good alignment with those for triclinic K-rich feldspars

(except thè Pontiskalk microcline of Finney & Bailey,

1964: an authigenic specimen whose structure was poorly

refined because of thè low number of \Fo\ data used in thè

refinement). Apparently, thè Pellotsalo specimen on thè

basis of its (T-O) distances and thè Al contents calculated

therefrom (see Table Al-a and Fig. 1) could be considered

a hyper-ordered microcline. By contrast, its metric proper-

ties (see Table Al-a) as well as thè optic axial angle and

thè extinction angle on (001) (2VX= 82.5° and X'A a =17°:

B.E. Brown & Bailey, 1964) are consistent with thè struc¬

tural state of a normal low microcline. In addition, Blasi et

al. (1984b) noted in thè Pellotsalo specimen a discrepancy

of A (bc) - A(b*c*) = -0.049. This value (-0.047 from thè da¬

ta in Table Al-a: see thè section on methods) is particular-

ly high when compared with that of other specimens of

low microcline, in thè light of thè findings of Blasi (1980)

on thè behavior of A (bc) and A (b*c*) in alkali feldspar.

All these problems could be solved when a re-exami-

nation of thè Pellotsalo microcline became available by

Blasi et al. (1987a), who investigated a specimen kindly

supplied by S.W. Bailey. The specimen comes from thè

same crystal as that employed by B.E. Brown & Bailey

(1964) to isolate thè fragment that they used in structure

refinement. Single-crystal photographs showed that thè

fragment chosen by Blasi et al. (1987a) for their structure

refinement consists of a dominant K-rich phase and a very

minor Na-rich phase. The K-rich phase shows sharp spots,

some of which are accompanied by short streaks lying on

curves of Constant 0.

The Pellotsalo specimen is coarsely perthitic: this made

it possible to perform microprobe analyses of both thè K-

and thè Na-rich phase, which gave thè compositions

ON3.62i^^6.ioA^().o5Cfi().o2l^Tfo.2i and Orj 05 Abgg07Ano88

mol%, respectively. The microprobe composition of thè K-

rich phase is slightly less potassic than that obtained from

V [N0r(K: eq. 3) 0.956, N0r(E: eq. 8) 0.958: Table Al-a],

probably because thè volume excited by thè electron

beam contains small amounts of an exsolved Na-rich

phase (see Stewart & Wright, 1974; Blasi et al., 1984a,

1984b, 1987a; Blasi & De Poi Blasi, 1994). The small dif-

ference N Qr(a: eq. 9) - Nor(T: eq. 8) of -0.007 (see Table Al-

76

ACHILLE BLASI & CARLA DE POL BLASI

a) and thè value of thè A a index of -0.023 À indicate that

thè Pellotsalo microcline is free of lattice strain due to co-

herency stresses between exsolved K- and Na-rich phases.

The crystal structure was refined to low values of R

and wR (0.027 and 0.030, respectively). The (T- O) dis-

tances are consistent with thè metric properties and thè

structural indicators obtained by means of thè A method

(see Table Al -a) in indicating that thè degree of Si,Al or-

der corresponds to that of a normal low microcline rather

than to that of a hyper-ordered microcline. The contrasting

structural States resulting from old and new refinement of

thè two Pellotsalo specimens are displayed in Fig. 1 in

terms of Al contents derived from (T- O) distances. The

new structure refinement is of great value for estimating

Al-O and Si-O distances in K-rich feldspar, which are nec-

essary to convert T-O distances to Al contents of thè T

sites (see thè relevant section; also Table 1 and Table 2).

Blasi et al. (1987a) also re-investigated thè metric

properties determined by B.E. Brown & Bailey (1964): thè

above-mentioned discrepancy between A (bc) and A (b*c*)

led to discovery of a probable error in thè (3* angle. This

was corrected and thè new value made it possible to cal¬

ciate a new set of lattice constants (see Table Al-a). The

latter, along with thè atomic coordinates of B.E. Brown &

Bailey (1964), were used by Blasi & De Poi Blasi (unpubl.)

to calciate a new set of interatomic distances and angles.

These, however, are not significantly different from those

published by B.E. Brown & Bailey (1964).

Blasi et al. (1987a) reviewed thè contrasting genetic in-

terpretations formulated concerning thè Pellotsalo

feldspar during this century. These are due to thè am-

biguous occurrence of irregular patches of cross-hatching

in thè K-rich phase. Actually, thè diffraction spots in sin-

gle-crystal photographs show no evidence of monoclinic

ancestry in thè K-rich phase. However, thè shape and po-

sition of thè weak diffraction spots from thè coexisting Na-

rich phase may be interpreted in favor of an incipient M-

type association, which would indicate that exsolution

took place in a monoclinic environment. Considering that

several lines of evidence indicate that thè Pellotsalo mi-

croperthite formed by exsolution, thè high Na-content of

thè bulk composition (Or72Ab27An1 mol%: Goldsmith &

Laves, 1961) is of key importance in confirming primary

growth in thè monoclinic field. In addition, according to

Blasi et al. (1987a), thè moderate enrichment in Rb and

thè low concentration of Ba in thè K-rich phase suggest

that thè Pellotsalo microperthite formed at a temperature

just above that of thè monoclinic-triclinic inversion.

Prilep low microcline. The specimen was investigated

by Strob (1983) for his Diplomarbeit. Partial data were

published by Kroll & Ribbe (1983: Table 3; also Ribbe,

1984: Table 3), and Smith & Brown (1988: Fig. 3.6).

The specimen is a coarsely twinned amazonite from a

granitic pegmatite, sampled in thè Caniste II quarry, near

Prilep, Macedonia. The optic axial angle, 2VX = 83°, and

thè extinction angle on (001), X'A a = 17°, are consistent

with thè extreme structural state of low microcline. These

results are confirmed by thè metric properties and thè

structural indicators presented in Table Al-a. The N0r con¬

tents obtained from V indicate that thè composition of thè

Prilep microcline is highly potassic [N0r(F: eq. 3) 0.964,

N0r(V: eq. 8) 0.965: Table Al-a]. The difference N0r(a: eq.

9) - N0r( V: eq. 8) (-0.003: see Table Al-a) is consistent with

thè value of thè A a index (-0.009 A) in indicating absence

of lattice strain due to coherency between exsolved phas¬

es. The crystal structure of thè Prilep microcline was re¬

fined to low values of R and wR (0.0240 and 0.0265, re¬

spectively). The values of (T-O) distances and Al contents

derived therefrom (see Table Al-a and Fig. 1) are consis¬

tent with thè optical and metric data in indicating that thè

specimen is extremely ordered.

The data for thè Prilep microcline are very dose to

those for thè Pellotsalo (Blasi et al., 1987a) and thè Pikes

Peak (Blasi et al., 1984b; Blasi & De Poi Blasi, 1994) spec¬

imens of low microcline. However, unlike thè Pellotsalo

and thè Pikes Peak specimens, thè Prilep microcline is an

amazonite, which usually contains appreciable amounts of

Pb. In spite of this, thè Prilep microcline should be consid-

ered an excellent specimen for evaluating thè Al-O and Si-

O distances in K-rich feldspar (see thè relevant section; al¬

so Table 1 and Table 2).

Pikes Peak low microcline. Blasi et al. (1984b) refined

thè structure of two crystal fragments, 7813A and 7813B,

selected from a 14-mm3 cleavage block (specimen 7813)

from an overgrowth of white to colorless microcline coat-

ing thè late magmatic, bluish green, amazonitic microcline

perthite (specimen 7816) in thè so-called “Ray Ziegler”

pocket. This pocket, which measures roughly 90 x 180 x 30

cm, occurs in a pegmatite at Crystal Peak, Pikes Peak

batholith, Colorado, U.S.A.

Single-crystal photographs indicate that thè Pikes Peak

microcline is virtually free from exsolved Na-rich phase

and twins or related structures, but produces sharp dif¬

fraction spots suggesting structural homogeneity. X-ray

powder diffractometer patterns show extraordinarily

sharp peaks indicating that thè material investigated is

particularly well crystallized and structurally homogene-

ous, at least within thè limits of thè 14-mm3 cleavage block.

In fact, thè values of thè lattice constants from powder da¬

ta are very dose to those determined by single-crystal dif¬

fractometer data for crystal fragments 7813A and 7813B

(see Table Al-a).

A series of electron microprobe analyses of several

fragments from thè same 14-mm3 feldspar block gave thè

composition Or943Ab52An00Cn01Rbf04 mol%. As with thè

Pellotsalo microcline (see thè relevant section), thè micro¬

probe composition is slightly less potassic than that ob¬

tained from celi volume V [N0r(V: eq. 3) 0.973, N0r(F: eq.

8) 0.974: powder 7813; 0.971, 0.973; fragment 7813A; 0.962,

0.963: fragment 7813B; Table Al-a], These compositions

are slightly more potassic than those determined for thè

Pellotsalo microcline [N0r(F: eq. 3) 0.956, N0r(F: eq. 8)

0.958: Table Al-a]. The differences N0r(a: eq. 9) - N0r(F:

eq. 8) (-0.011: powder 7813; -0.017: fragment 7813A; -0.010:

fragment 7813B; Table Al-a), and thè A a index (-0.027 À:

powder 7813; -0.037: fragment 7813A; -0.028: fragment

7813B) indicate absence of lattice strain due to coherency

between exsolved phases.

The crystal structure of both specimens was refined to

R and wR values of 0.033 and 0.033, respectively, for spec¬

imen 781 3 A, and 0.031 and 0.034, respectively, for speci¬

men 7813B. These values are comparable with those ob-

CRYSTAL STRUCTURES OF ALKALI FELDSPARS FROM GRANITIC PEGMATITES: A REVIEW

77

tained for thè Pellotsalo microcline re-examined by Blasi

et al. (1987a) (see Table Al-a).The (T- O) distances and thè

Al contents calculated therefrom (see Table Al-a and Fig.

1) are consistent with thè lattice constants and thè struc-

tural indicators derived therefrom (see Table Al-a) in in-

dicating that thè degree of Si, Al order of both Pikes Peak

crystal fragments corresponds to that of an extremely or-

dered microcline. The 7813A and 7813B specimens are ex-

cellent candidates for use to estimate thè Al-O and Si-O

distances in K-rich feldspar (see thè relevant section; also

Table 1 and Table 2).

The findings by Blasi et al. (1984b) also contributed to

a better understanding of thè end stages of evolution of a

typical pocket assemblage in thè anorogenic Pikes Peak

complex. Microcline overgrowth 7813 probably formed

below thè temperature of thè monoclinic-triclinic inver-

sion. This is consistent with thè low concentration of Ba

(BaO: 0.04 wt%, either by electron microprobe or spec-

trographic analyses), thè Virtual absence of Sr (SrO: 0.00

and <0.001 wt%, by electron microprobe and spec-

trographic analyses, respectively) and thè high concen¬

tration of Rb (Rb20: 0.14 wt%, either by electron mi¬

croprobe or spectrographic analyses). Lead, which is en-

riched in thè amazonitic microcline substrate 7816 (PbO:

1.62 wt%, N.M. Conklin, priv. comm., in Blasi et al. 1984b),

is virtually absent in 7813 (PbO: <0.02 wt%, by spectro¬

graphic analysis), which may imply an important hiatus

between host and overgrowth and a modification of thè

fluid’s composition. The high degree of Si, Al order in mi¬

crocline overgrowth 7813 is considered a reflection of thè

mild alkalinity of thè residuai supercritical aqueous fluid

phase that operated at thè end stage of a protracted histo-

ry of differentiation of an anorogenic basic magma.

Blasi & De Poi Blasi (1994) subjected a third crystal

from thè Pikes Peak overgrowth to structure refinement.

The specimen, denoted as 7813C, is a cleavage fragment

separated from thè same 14-mm3 feldspar block employed

by Blasi et al. (1984b) to obtain thè two crystals 7813A and

7813B.

The lattice constants determined for specimen 7813C

are very dose to those obtained for thè other feldspar ma¬

terial from overgrowth 7813 (see Table Al-a), and thè

same applies to quantities derived from lattice constants,

e.g. thè composition in terms of N0r [N0r(V: eq. 3) 0.962,

N0r(F: eq. 8) 0.964: Table Al-a], thè difference N0r(a: eq. 9)

- N0r(F: eq. 8) (-0.008: Table Al-a), and thè thè A a index (-

0.018 À).

Blasi & De Poi Blasi (1994) made a thorough collec-

tion of intensity data measuring a total of 14966 diffrac-

tions in thè 20(MoKa) range 3-100°. Two standard struc¬

ture refinements were done in thè 20(Mo/Ca) ranges 3-60

[refinement (a)] and 3-100° [refinement (b)], respectively

(see Table Al-a). The R and wR values are 0.016 and 0.018,

respectively, in thè former, and 0.019 and 0.020, respec¬

tively, in thè latter. These values are definitely lower com-

pared with thè corresponding ones in thè other low micro¬

cline specimens(see Table Al-a). The values of o in (T-O)

distances, which amount to 0.001 À in thè Pellotsalo mi¬

crocline and in Pikes Peak specimens 7813A and 7813B,

reduce to 0.0004 and 0.0003 À in cleavage fragment

7813C, with reference to refinement (a) and refinement

(b), respectively (see Table Al-a). The latter values of a

are similar to those obtained by Strob (1983) for his Prilep

microcline. The (T-O) distances from refinements (a) and

(b) represent excellent data for estimating thè Al-O and

Si-O distances in K-rich feldspar (see thè relevant section;

also Table 1 and Table 2). Blasi & De Poi Blasi (1994) ob-

served that thè values of thè T-O distances are very simi¬

lar to thè corresponding values in specimens 7813A and

7813B. However, they noted that thè (Tlm-O) distance in

microcline 7813C, i.e. 1.6159(4) [refinement (a)] or

1.6158(3) A [refinement (b)], seems to be somewhat

longer than in specimens 7813A and 7813B, in which it is

equal to 1.614(1) (see Table Al-a). This is, however, an

idiosyncratic feature of specimen 7813C, because thè same

value also resulted from other data collections obtained

for thè same specimen with thè same or a different dif-

fractometer.

In view of thè high quality of thè intensity data col-

lected for specimen 7813C, Blasi & De Poi Blasi (1994) ex-

plored several models for determining thè composition in

K-rich feldspar from K,Na site refinements. The results of

thè investigation will be considered in thè section dealing

with thè Rabb Canyon sanidine. In addition, Blasi & De

Poi Blasi (1994) used thè intensity data collected for spec¬

imen 7813C in an attempt to explore thè possibility of de¬

termining T-site occupancies from Si,Al site refinements.

The attempt was successful: thè separation of Si and Al

atoms occurred in a number of cycles of least-squares site

refinement depending on thè type of thè model employed.

Several models of T-site refinement were investigated and

thè results were critically discussed.

Khibiny low microcline. Borutskiy et al. (1985) re-

fined thè crystal structures of an adularia and a low mi¬

crocline occurring in pegmatites from urtites and ris-

chorrites, respectively, in thè centrai are of thè Khibiny

massif, Kola Peninsula, Russia. As regards thè adularia

specimen, see thè section on thè Khibiny adularia.

Turning to thè low microcline, this is specimen No. 1667

from a pegmatite in thè pyroxene rischorrites of thè “Eu-

dialyte Ridge” on Mt. Kukisvumchorr in thè Khibiny mas¬

sif. Borutskiy et al. (1985) presented three complete sets of

Chemical analyses for thè Khibiny microcline showing con-

trasting results, especially for Ca. In ternary terms, thè

composition adopted was Or96Ab35Rbf03 mol% (thè Eng-

lish translation contains a misprint of thè Rbf content, i.e.

0.8 instead of 0.3 mol%).The specimen, however, also con¬

tains an appreciable amount of ferric iron (0.45-0.46 wt%

Fe203), which corresponds to ~1.5 mol% KFeSi3Os

Borutskiy et al. (1985) reported eight sets of lattice

constants for their specimen, which were determined by

several investigators. The set of lattice constants presented

in Table Al-a is that adopted in their work. The N0r con¬

tents calculated from V yielded exceptionally pure com-

positions [N0r(F: eq. 3) 0.994, N0r(F: eq. 8) 0.996: Table

Al-a]. The difference N0r(n: eq. 9) - N0r(F: eq. 8) (0.005:

see Table Al-a) and thè A a index (0.018 À) indicate that

thè specimen is free from lattice strain due to coherency

between exsolved phases. The values of thè structural in¬

dicators calculated from metric properties (see Table Al-

a) indicate that thè specimen is a low microcline showing

residuai Si, Al disorder.

78

ACHILLE BLASI & CARLA DE POL BLASI

The crystal structure was refined to an R value of

0.036. The (T-O) distances and thè Al contents derived

therefrom confimi that thè specimen is not completely or-

dered (seeTable Al-a and Fig. l).The Khibiny low micro¬

cline, therefore, cannot be used for determining thè refer-

ence values for Al-O and Si-O distances.

According to Borutskiy et al. (1985) thè properties of

thè alkali feldspars may be useful in exploration for apa¬

tite in thè Khibiny massif. In particular, they can be used

to resolve thè problem of assigning rocks of uncertain ori-

gin to thè nepheline syenite complex or to thè melteigite-

urtite complex, associated with thè famous Khibiny ap¬

atite deposits. Borutskiy et al. (1985) observed that thè co-

existence of adularia and low microcline in thè melteigite-

urtite and thè rischorrites of Khibiny reflects either a mul-

tistage formation for these rocks, or unusual conditions

during which thè simultaneous formation of different

structural modifications of K-rich feldspar occurred.

Spencer U intermediate microcline. The specimen is

classified as a microcline-microperthite in thè suite of al¬

kali feldspars described by Spencer (1937), and comes

from thè mica-bearing pegmatite of Kodarma, Bihar, In¬

dia. It is almost glass-clear and shows a medium coarse mi-

croperthitic texture of thè string or film types (Spencer,

1937). The K-rich phase shows patches of cross-hatching

(Bailey & Taylor, 1955).

The first structure analysis of thè Spencer U microcline

was undertaken by Bailey & Taylor (1955), who cut small

untwinned cubes with a razor biade from regions in be-

tween thè cross-hatched areas of a cleavage flake. The lat¬

tice constants, obtained from weighted average of data

from 6 feldspar cubes, showed interaxial angles intermedi¬

ate between those for monoclinic symmetry and those for

thè highest degree of triclinic geometry. The collection of

Weissenberg intensity data started with cube No. 1 and, af¬

ter its loss, continued with cube No. 4 (see Table Al-a).

Combined data from thè two feldspar cubes were used in

2D ( R = 0.095 for 0 kl diffractions and R = 0.098 for hkO

diffractions) and 3D (R = 0.154 for all diffractions and R =

0.138 for a new set of \Fc\ values calculated on thè basis of

thè last atomic parameters) Fourier refinements. A total of

2652 diffractions was employed.The (T-O) distances from

3D results gave clear proof of partial ordering in micro¬

cline: (7ìo-0> = 1.700, (Tlm-O) = 1.645, (TZo-O) = 1.614,

and (Tlm-O) - 1.611 À (sub-sites o and m are inter-

changed in respect to those of thè originai study, in accor-

dance with Laves’s (1951) convention for thè orientation

of axes in microcline).

Later, Bailey (1969) made a least-squares refinement

of Spencer U microcline using 2124 diffractions selected

from thè originai data employed by Bailey & Taylor

(1955). Adopting Laves’s (1951) convention, Bailey (1969)

interchanged thè sub-sites o and m, transformed thè lattice

angles a and y, and corrected thè (3 value from that of thè

originai study (see Table Al-a). The R values dropped to

0.080, and thè (T-O) distances, listed in Table Al-a, gave

delinite proof that thè Spencer U specimen is an interme¬

diate microcline. This may be seen from thè diagram in

Fig. 1, derived from (T-O) distances, and is also confirmed

by thè structural indicators calculated from lattice con¬

stants in Table Al-a.

The extent of thè perthitic exsolution (bulk composi-

tion Or^Ab^An^o wt%: Spencer, 1937; N0r(F) = 0.94:

Table Al-a) and thè irregular patches of cross-hatching in

thè K-rich phase indicate that thè Spencer U specimen is

a complex feldspar material.

The difference N 0r(a: eq. 9) - N0r(L: eq. 8) = 0.033

(Table Al-a: set of lattice constants from Bailey, 1969) sug-

gests that thè K-rich phase could be affected by lattice

strain due to coherency with thè exsolved Na-rich phase.

o

This effect is confirmed by a A a value of 0.068 A, and is

consistent with thè bulk composition observed. The set of

lattice constants determined later by Wright & Stewart

(1968) from X-ray powder diffractometer data confirms

lattice strain (seeTable Al-a; also Stewart & Wright, 1974,

Fig. 9). In addition, thè extinction angle on (010) is 7.5°

(Spencer, 1937): a value slightly higher than normal, which

suggests cryptoperthitic exsolution, and hence lattice

strain (see Stewart, 1974; Stewart & Wright, 1974).

The Spencer U microcline shows a variable IV x: 76.5°,

measured by Spencer (1937), and a range of values from

69 to 72°, measured by Finney & Bailey (1964) in thè crys¬

tal chip used by Bailey & Taylor (1955) for structure analy¬

sis. Finney & Bailey (1964) adopted a value of 70.5°, thè

middle of their observed range of 2VX values.

In their crystal fragments, Bailey & Taylor (1955) ob¬

served weak extra diffractions from thè exsolved Na-rich

phase, which did not seriously disturb thè refinement. In

different crystal fragments of Spencer U specimen, Smith

& MacKenzie (1959) and Smith (1974a, Fig. 6-13; also

1974b, p. 455) observed diffuse diffractions elongated in

thè direction of thè b axis and centered on position hkl

with ( h + k) odd. These diffractions violate C- face center-

ing and are indicative of a P2x/a structural arrangement

(see Blasi & De Poi Blasi, 1994, pp. 90-92, for review of

models). The same specimens also showed thè main spots

with ( h + k) even to be associated with diffuse streaks in

thè direction of spots for thè M-type association of micro¬

cline.

In conclusion, thè Spencer U microcline, a historical

example of intermediate structure between low microcline

and C2/m K-rich feldspars, comes from a complex microp-

erthite: thè K-rich phase shows irregular optical cross-

hatching, slight lattice strain, rather variable 2VX , and X-

ray evidence of fine scale domain textures, including thè

P2{la structural arrangement.

Adamello P17C high microcline. The specimen is

from a suite of 9 K-rich feldspars, which were subjected to

structure refinement by Dal Negro et al. (1978; 1980; see

also De Pieri, 1979). The 9 specimens come from tonalite,

quartz-diorite, aplite and pegmatite of thè Adamello mas¬

sif, Northern Italy.

Single-crystal X-ray photographs show that all speci¬

mens are untwinned homogeneous crystals, free from

diffuse streaks, and do not display thè extra diffractions

hkl with ( h + k) odd, commonly observed in specimens

of adularia. One of thè Adamello feldspars is clearly

monoclinic, two are metrically monoclinic and struc-

turally triclinic, and thè others are definitely triclinic.

The structural state of thè triclinic specimens ranges

from high to low microcline, and thè feldspar suite de-

CRYSTAL STRUCTURES OF ALKALI FELDSPARS FROM GRANITIC PEGMATITES: A REVIEW

79

0.5

ti o + tl m or 2t1

0.6 0.9

1 .0

0.5

0.6 0.7

Fig. 1 - Relation between sum and difference of Al contents in Tlo and Tlm sites,i.e. tlo + tlm and tlo - tlm, in specimens of K-rich feldspar

from granitic pegmatites used in structure refinements. The values of tlo + tlm (in thè case of triclinic simmetry) or 2tl (in thè case of mon¬

oclinic simmetry) and tlo - tlm were obtained from (T- O) distances following thè procedure developed by Blasi & De Poi Blasi (1994, eqs.

18 and 19) and using thè reference values for Al-O and Si-O distances given in thè relevant section of this work. Room-temperature crystal

structures of naturai specimens'. 1 Pellotsalo low microcline (B.E. Brown & Bailey, 1964); 2 Pellotsalo low microcline (Blasi et ai, 1987a); 3

Prilep low microcline (Strob, 1983); 4 and 5 Pikes Peak low microcline 7813A and 7813B, respectively (Blasi et al.. 1984b); 6 and 7 Pikes

Peak low microcline 7813C refinements (a) and (b), respectively (Blasi & De Poi Blasi, 1994); 8 Khibiny low microcline (Borutskiy et al.,

1985); 9 Spencer U intermediate microcline (Bailey, 1969); 10 Adamello P17C high microcline (Dal Negro et al., 1978); 11 Himalaya "or-

dered orthoclase" (Prince et al., 1973); 12 Khibiny adularia (Borutskiy et al., 1985); 13 Spencer C orthoclase (Colville & Ribbe, 1968); 14

Itrongay orthoclase (Rimata et al., 1996); 15 Itrongay orthoclase (Nyfeler et al., 1998); 16 Rabb Canon sanidine (Keefer & Brown, 1978); 17,

18, 19 and 20 Buck Claim adularia Buck-18-IV, Buck-18-III, Buck-18-II and Buck-18-I, respectively (Ferguson et al., 1991). Room-tempera¬

ture crystal structures ofpre-heated specimens'. 21 Itrongay high sanidine (Nyfeler et al., 1998) ; 22 Spencer C high sanidine (Ribbe, 1963); 23

Bedford high sanidine (Blasi et al., 1987a). Low-temperature crystal structure of naturai specimen'. 24 Itrongay orthoclase at 121 K (Rimata

et al., 1996).

scribes, on thè whole, a typical intermediate two-step or-

dering path.

According to Dal Negro et al. (1978), thè triclinic spec¬

imens crystallized (or recrystallized) directly with triclinic

symmetry. On thè whole, thè Adamello feldspars would

represent structural States corresponding to equilibrium

conditions at thè moment of crystallization or possible

subsequent recrystallization.

Specimen P17C is from a tourmaline-muscovite-

bearing pegmatite dike. Its bulk composition, obtained

from Chemical analysis, is Or87 ^Ab^oAnojCnooRbfog

mol%. Unlike thè other 8 non-pegmatitic feldspars from

thè Adamello massif, specimen P17C does contain Rb

and is free from Ba. The composition from V, [N0r 0.92:

Table Al-a], is consistent with weak exsolution, as is

expected in C2lm K-rich feldspars rather than in CI K-

rich feldspars, in which exsolution proceeds to larger

extents with a simultaneous increase in Si, Al order. In-

deed, metric properties and structural indicators in Table

Al-a indicate that thè P17C specimen is a highly disor-

dered triclinic K-rich feldspar. The difference N0r(a: eq.

9) - N0r(V: eq. 8) = 0.024 (see Table Al-a) and thè

value of thè A a index, 0.028 À, are consistent with absence

of lattice strain due to coherency between exsolved

phases.

The crystal structure was refined to an R value of

0.032. The (T- O) distances and thè Al contents derived

therefrom (Table Al-a and Fig. 1) confirm that thè speci¬

men is a highly disordered microcline. Together with other

specimens of a similar structural state, P17C is of great val¬

ue for clarifying thè position of thè switchover between

C2/m and Ci symmetry in K-rich feldspars.

Himalaya “ordered orthoclase The specimen, in-

vestigated by Prince et al. (1973), is from thè Himalaya

tourmaline mine (Jahns, 1954: map 8, locality 20), in thè

Mesa Grande pegmatite district. San Diego County, Cali¬

fornia, U.S.A. The pegmatite at Himalaya mine is an ex-

ample of asymmetrically-zoned pegmatite dike with

aplitic footwall portion, as displayed by thè schematic

cross section in Fig. 5B of Jahns & Burnham (1969; also

Jahns, 1982). A detailed survey on thè Himalaya peg-

matite-aplite dike System was given by Foord (1977), and

further data can be found in Taylor et al. (1979),

Cerny (1982), Martin (1982), G.E. Brown & Ewing (1986),

Foord et al. (1986), and London (1986).

Prince et al. (1973) described in detail thè occurrence

of thè Himalaya specimen: long pendants of microcline

perthite, growing from thè hanging wall of thè dike, grade

into orthoclase perthite in thè centrai portion of thè peg-

80

ACHILLE BLASI & CARLA DE POL BLASI

matite, near thè geni pockets; thè orthoclase perthites are

coated by a gem-quality, single-phase orthoclase over-

growth, up to 1 cm thick, where they project into thè gem

pockets.

Prince et al. (1973) performed a neutron diffraction

structure refinement on a cleavage fragment taken from

thè gem-quality orthoclase overgrowth. The compositions

determined by wet Chemical analysis and by electron mi¬

croprobe analysis are Or874Ab104An03Cn05 Rbf15 mol%

and Or88 6Ab99An00Cn00Rbf15 mol%, respectively (recal-

culated in this work from originai data).

The wet Chemical analysis shows 0.44 wt% H20+, liber-

ated between 117 and 1000 °C, which was initially attrib-

uted by Prince et al. (1973) to hydroxyl-for-oxygen substi-

tution in thè crystal structure. Both analyses show a rela-

tively high Rb content, whereas thè amount of Ba is small

and thè relevant data are contrasting. The Ca content is

very small and thè relevant data are also contradictory.

The Na contents are very dose to each other in both

analyses, and are in thè range observed in thè monoclinic

K-rich phase of many microperthites. The Himalaya spec¬

imen, however, is largely free from any exsolved Na-rich

phase, as indicated by thè dose interrelationship of thè fol-

lowing data: Or = 87.4 mol% (from wet Chemical analysis),

Or = 88.6 mol% (from microprobe analysis), NOr(I0 =

0.891 (from eq. 2,Table Al-a), N0r(T) = 0.883 (from eq. 6,

Table Al-a), and N0r(fl) = 0.894 (from eq. 7,Table Al-a).

Further data come from another crystal fragment

of thè Himalaya specimen heated by Horsky & Martin

(1977) to 1000 °C for 1 h. After heating, thè fragment

appeared to be milky, and this unexpected behavior was

attributed by Horsky & Martin (1977) to decrepitation of

fluid inclusions and propagation of thè fluid along cracks

and cleavages in thè chip. The compositions from V and

a in this specimen [N0r(T: eq. 2)= 0.906, N0r(T: eq. 6)=

0.896, and N0r(a: eq. 7) = 0.899: calculated as indicated

in thè section on methods, from originai data in Table 2

of Horsky & Martin, 1977], compared with thè data for

thè unheated specimen, would confirm that thè Himalaya

K-rich feldspar is essentially a single-phase crystal. This

is consistent with thè absence of lattice strain due to co-

herency between exsolved phases, as indicated in thè

unheated specimen by thè difference N0r(a: eq. 7) -

N0r(T: eq. 6) = 0.011 (see Table Al-a) and by thè value

o

of A a = 0.048 A. The experiment mentioned, performed

by Horsky & Martin (1977), was also useful for con-

firming that H20+ from thè wet Chemical analysis in thè

Himalaya specimen is not related to hydroxyl-for-oxy¬

gen substitution in thè structure, as initially envisaged by

Prince et al. (1973).

The values of b and c, or b* and c*, of thè Himalaya

specimen investigated by Prince et al. (1973) are consistent

with strong Si, Al order in T sites: A (bc) - 0.918 and

A (b*c*) = 0.869 (see Table Al-a). The difference A (bc) -

A (b*c*) is 0.049: a value that is somewhat high, and possi-

bly related to thè value of thè (3 angle, 116.073(9)°, which

in turn is somewhat high for a K-rich feldspar (see Blasi,

1980).

Prince et al. (1973) made precession and Weissen-

berg X-ray photographs of single crystals from thè same

specimen as that employed for neutron diffraction. Al-

though thcse photographs were exposed to Cu/Ca radia-

tion for periods of time of up to 100 h, Prince et al.

(1973) did not observe thè diffuse extra diffractions hkl

with ( h + k) odd, typical of adularia specimens, nor could

they detect any evidence of twinning.

Using neutron-diffraction data, Prince et al. (1973)

refined thè crystal structure of thè Himalaya specimen to

R = 0.031 (wR = 0.035). The (T- O) distances are consis¬

tent with strong Si, Al order in thè T sites: thè value of

2tl is 0.900(7) (see Fig. 1 and Table Al-a), which is

very dose to thè above-mentioned values of A (bc) and

A (b*c*). By contrast, direct refinement of T-site occu-

pancies indicates that thè Himalaya specimen reached thè

highest possible Si, Al order compatible with monoclinic

symmetry, i.e. tl = 0.5 and t2 = 0, since thè experimental re-

sults gave tl = 0.516(29) and t2 = -0.016(29). On thè basis

of thè results from T-site refinement, Prince et al. (1973)

adopted thè name “ordered orthoclase” for thè Himalaya

specimen. Later, Horsky & Martin (1977) concluded that

thè Himalaya orthoclase is distinct, thè first documented

example of thè “theoretical” maximum low sanidine or

“theoretical” orthoclase structure type.

Stewart & Wright (1974) cast doubt upon thè reli-

ability of neutron site refinement because thè metric

properties, thè (T- O) distances and thè optic axial angle

(2VX - 63°) of thè Himalaya specimen indicate incomplete

Si,Al order for monoclinic symmetry. Furthermore, Stew¬

art & Wright (1974) reported that Wones et al. (1967) and

Crosby (1971) described thè occurrence of monoclinic K-

rich feldspars with metric properties implying more Si, Al

order than that of thè Himalaya specimen. The existence

of more ordered specimens, according to Stewart &

Wright (1974), also indicates that thè Himalaya specimen

is not completely ordered.

Smith (1974a, p. 68; also Smith & Brown, 1988, p. 44)

reported that, in thin section, thè extinction in thè Hi¬

malaya specimen is uneven; lattice refinement in thè neu¬

tron diffractometer gave indications of triclinic geometry;

single-crystal X-ray patterns show diffuseness of diffrac¬

tions which would be split in twinned low microcline. From

these data Smith (1974a, p. 68; also Smith & Brown, 1988,

p. 44) concluded that thè Himalaya specimen is not thè

same as theoretical low sanidine, and that it is a domain-

twinned microcline with such strong coherence that thè

physical properties are pseudo-monoclinic.

To sum up: optic axial angle, lattice constants and (T-

O) distances are in dose agreement in indicating that thè

Himalaya specimen is extremely ordered. The Al contents

from T-site refinement of neutron data are probably bi-

ased. On thè other hand, it is known that thè determi-

nation of 7-site occupancies from Si,Al site refinements of

neutron scattering lengths may present problems (e.g.,

Kroll & Ribbe, 1983, p. 58 and p. 67; Ribbe, 1984, p. 38; Fitz

Gerald et al., 1986, p. 1404; Ribbe, 1994, p. 9). In spite of

this, thè Himalaya specimen might be considered as a

unique example of an average feldspar structure made up

of a domain-twinned microcline with domains so perfectly

balanced and strongly coherent as to mimic thè structure

of a low sanidine, dose to thè theoretical end-member low

sanidine.

The genetic conditions of thè Himalaya gem-quality,

single-phase “ordered orthoclase” are rather difficult to

understand: indeed, thè specimen coexists with K-rich

CRYSTAL STRUCTURES OF ALKALI FELDSPARS FROM GRANITIC PEGMATITES: A REVIEW

81

feldspars of different structural States within thè same peg-

matite dike, and formed in a pegmatite gem pocket, which

is a water-rich environment, likely to promote Si.Al order-

ing in K-rich feldspar.

According to Prince et al. (1973), thè crystal growth of

thè Himalaya “ordered orthoclase” from hydrothermal

fluids in a gem-bearing pocket was presumably very slow,

in agreement with thè regional cooling, that occurred in a

mesozonal plutonic environment (thè country rock is gab-

broic). Prince et al. (1973) proposed that thè Himalaya or¬

thoclase persisted metastably to lower temperatures.prob-

ably because of its stable, balanced Al distribution in thè

71 sites, and suggested that “ordered orthoclase” does

have a field of thermodynamic stability dose to thè peg¬

matite solidus.

Thereafter Martin (1974) proposed a stable-phase-

equilibrium diagram for alkali feldspars with two first-or-

der transitions for KAlSi3Os: one between fully disordered

sanidine and “ordered orthoclase”, at approximately 800

°C, and thè other between “ordered orthoclase” and max¬

imum microcline, at approximately 590 °C.

Later, Horsky & Martin (1977) emphasized that thè or¬

dered structure of thè Himalaya orthoclase survived

metastably, perhaphs because of sudden and explosive loss

from thè pegmatitic System of a suitable aqueous medium

for Si,Al ordering to proceed.

On thè basis of data obtained by Taylor et al. (1979),

Martin (1982) stated that thè Si, Al ordering in thè Hima¬

laya “ordered orthoclase” occurred in gem pockets at

565-525 °C. Indeed, thè data of Taylor et al. (1979), ob¬

tained from stable isotope and fluid inclusion studies,

showed that minerals in gem-bearing pockets from peg-

matite-aplite dikes in San Diego County grew from aque¬

ous fluids from 565 to 520 °C, and perhaps lower, under

pressures of at least 2000 bars. These data were criticized

and corrected by London (1986) to approximately 475 and

425 °C between 2800 and 2400 bars. Foord et al. (1986) re-

ported P-T data for pocket formation as 540-520 °C at 2

kbar (referred to Taylor et al., 1979, and Stern et al., 1986)

and 450-500 °C at 2-3 kbar (referred to London, 1986). The

comparison and evaluation of these data are baffling. The

P-T data suggested by London (1986) disclose new

prospects for interpretation of genetic conditions of thè

Himalaya specimen. In fact, they would mean that thè Hi¬

malaya specimen formed at, or below, thè inversion tem¬

perature, which is estimated to be 450-500 °C, at low pres¬

sure, for K-rich feldspar (e.g., Smith & Brown, 1988, Fig.

1.2). In this case, thè conditions of formation of thè Hi¬

malaya overgrowth would be somewhat similar to those

for alpine adularia specimens, which are unstable at all

temperatures. In this respect, »ern_ (1994, p. 528) consid-

ered thè Himalaya overgrowth as a late generation of K-

rich feldspar, which formed as a low-temperature

hydrothermal phase.

In conclusion, thè genetic conditions of thè Himalaya

K-rich feldspars need further investigation. Determination

of thè structure of all K-rich feldspar types in thè Hi¬

malaya dike would be helpful, along with thè bulk compo-

• sitions of both orthoclase and microcline perthites from

thè pendants. These data, along with thè P-T data assessed

for pocket formation, could be used to enter thè phase di¬

agram of alkali feldspars properly.

Khibiny adularia. The specimen. No. 1481/E, is from a

pegmatite in massive urtites from thè Material'naya shaft

in thè Yakspor apatite deposit in thè centrai are of thè

Khibiny massif, Kola Peninsula, Russia. Borutskiy et al.

(1985) refined thè crystal structure of thè adularia and that

of a microcline occurring in pegmatites from rischorrites

of thè same massif (see thè section on thè Khibiny low mi¬

crocline).

Borutskiy et al. (1985) reported two complete sets of

Chemical analyses for thè Khibiny adularia showing some

discrepancies, especially for Ca, as in thè case of thè

Khibiny microcline (see thè relevant section). In ternary

terms, thè composition adopted was Or9]Ab7g Cn12 mol%.

As in thè case of thè Khibiny microcline, thè Khibiny adu¬

laria also contains an appreciable amount of ferric iron

(1.09-1.16 wt% Fe203), which corresponds to ~3.8 mol%

KFeSi308.

Borutskiy et al. (1985) reported four sets of lattice

constants for thè Khibiny adularia, which were deter-

mined by several investigators. The set of lattice con¬

stants presented in Table Al-a is that adopted in their

work. The composition from V is N0r = 0.92 (see Table

Al-a). The difference N0r(a: eq. 7) - N0r(F: eq. 6) = -0.019

o

(see Table Al-a) and thè value of thè A a index, -0.048 A,

are consistent in indicating that thè specimen is free of lat¬

tice strain due to coherency between exsolved phases. The

values of thè structural indicators calculated from metric

properties (see Table Al-a) indicate that thè degree of

Si, Al order of thè specimen is rather high for monoclinic

symmetry.

Borutskiy et al. (1985) refined thè crystal structure of

thè Khibiny adularia to an R value of 0.033 in thè space

group CI. However, metric properties are monoclinic,

and thè discrepancies between values of (T-O) distances

for o and m subsites are too small to be indicative of a tri-

clinic structure (see Table Al-a). Note thè abnormality of

thè (Tlo-O) distance, which is shorter than thè (Tlm-O)

distance.The Khibiny adularia is slightly less ordered than

thè Spencer B adularia studied by Colville & Ribbe (1968)

[((71o-0) + (71ra-0))/2 = 1.662 versus (71-0) = 1.664 À],

The Khibiny adularia lies at thè switchover between mon¬

oclinic and triclinic symmetry in K-rich feldspars (see Fig. 1).

Borutskiy et al. (1985) reported that scans made on a Weis-

senberg X-ray goniometer showed neither signs of devia-

tion from C2lm symmetry nor evidence of exsolution.

Spencer C orthoclase. The specimen is classified as an

orthoclase-microperthite in thè suite of alkali feldspars de-

scribed by Spencer (1937). It is a colorless water-worn

pebble, about 2.4 cm across, collected by Spencer (1930)

from thè gem-washing residues at thè Mogok Ruby Mines,

Northern Burma. According to Spencer (1930, p. 330)

these residues appear to be derived from thè weathering

of pegmatites intruded into a country-rock of crystalline

limestone.The feldspar is optically homogeneous with 2VX

= 43.6° and Chemical composition Or905Ab71An24 mol%

(Spencer, 1930, p. 331; also Spencer, 1937, Table I; compo¬

sition recalculated from originai data). The specimen was

denoted C by Spencer (1937). Later, specimen C became

Spencer C.

The first structure analysis on thè Spencer C speci¬

men was made by Chao et al. (1940) using two 2D prò-

82

ACHILLE BLASI & CARLA DE POL BLASI

jections on [100] and [010]. The difference obtained be-

tween (T-O) distances (Tl-OÒ = 1.68 and (T2-0) = 1.58

À) was interpreted as a result of partial Si, Al ordering.

However, Cole et al. (1949) documented that thè accuracy

of thè data obtained by Chao et al. (1940) was too poor to

conclude that thè values of thè of (T-O) distances were i

dicative of partial disorder in thè Spencer C specimen.

Later, Jones & Taylor (1961a, 1961b) refined thè crys-

tal structure of thè Spencer C orthoclase to a final R value

of 0.072 using four 2D projections on [100], [010], [001],

and [T], obtained from Weissenberg data collected from

three different crystal fragments (see Table Al-a). The

new (T-O) distances established that thè specimen is actu-

ally partially ordered ((Tì-O) = 1.652 and (72-0) = 1.633

À). Jones & Taylor (1961a, 1961b) adopted in their work

thè lattice constants determined previously by Cole et al.

(1949) on an other specimen of Spencer C orthoclase (see

Table Al-a).

Colville & Ribbe (1968) reinvestigated thè Spencer C

specimen. They reported a microprobe analysis made by

Smith & Ribbe (1966): thè composition Or909Ab81

Antì 5Cn0 2Srf03 mol%, recalculated in this work from orig¬

inai data, is very dose to that obtained by Spencer (1930;

1937: see above) from wet Chemical analysis. The crystal

structure was redetermined by Colville & Ribbe (1968) by

a 3D least-squares refinement using \Fo\ data listed by

Jones & Taylor ( 1961 b). The R value dropped to 0.054, and

thè (T-O) distances became: (Tl- O) = 1.6555(28) and (72-

O) = 1.6283(26) À (see Table Al-a). Colville & Ribbe

(1968) adopted in their work a set of lattice constants de¬

termined by D.B. Stewart by X-ray powder diffractometry.

These lattice constants are very dose to those obtained by

Cole et al. (1949). Later, Wright & Stewart (1968) pub-

lished another set of lattice constants for thè Spencer C or¬

thoclase (see Table Al-a), which is slightly different from

thè set obtained by D.B. Stewart.

All three sets of lattice constants give N0r contents

from V and a that are in dose agreement with thè “Chem¬

ical” and microprobe compositions mentioned above [thè

average of N0r(V/) values is ~ 0.89: see Table Al-a]. In ad-

dition, thè small values of thè differences N0r(«: eq. 7) -

N0r(R: eq. 6), in thè range -0.002-0.006 (see Table Al-a),

and thè A a values, in thè range -0.003-0.019 À, indicate

that thè Spencer C specimen is free of lattice strain due to

coherency between exsolved phases. The agreement be-

tween “Chemical”, microprobe and N0r data, as well as thè

absence of lattice strain validate thè assumption of

Spencer (1930, pp. 330-331), according to which most of

thè Na content in thè Spencer C orthoclase is retained in

solid solution.

The values of A (bc), in thè range 0.716-0.726, and those

of A in thè range 0.702-0.710, are in dose agree¬

ment with thè 2tl value of 0.713 calculated from thè (T-O)

distances determined by Colville & Ribbe (1968) (see

Table Al-a). In Fig. 1, thè representative point of thè

Spencer C orthoclase plots very dose to thè two represen¬

tative points of thè Itrongay orthoclase investigated at

room temperature by Rimata et al. (1996) and Nyfeler et

al. (1998) (see thè relevant sections).

Single-crystal X-ray photographs show that thè

strongest ditfraction spots of thè Spencer C orthoclase are

accompanied by diffuse streaks in thè relation for thè M-

type association of low microcline (see Smith & MacKen-

zie, 1959; Jones & Taylor, 1961a, 1961 b; McConnell, 1965;

Colville & Ribbe, 1968). This poses a major problem con-

cerning thè symmetry of thè Spencer C orthoclase. The

problem was stirred up by Laves in Jones & Taylor (1961a;

see also Laves, 1950, 1952). The specimen would not actu-

ally be monoclinic: rather, it could be a domain-twinned

microcline.

Jones & Taylor (1961a, 1961 b) ignored thè weak

streaks in their structure analysis, and concluded that an

attempt to discuss thè structure of thè Spencer C ortho¬

clase in terms of submicroscopically twinned microcline is

unlikely to be physically realistic. Colville & Ribbe (1968),

more realistically, regarded thè structure of thè Spencer C

orthoclase as an average structure.

Smith (1974a, p. 73, p. 247; 1974b, p. 455) reported that

single-crystal oscillation photographs taken by Smith &

MacKenzie (1959) showed that thè weaker and longer dif¬

fuse streaks of Spencer C orthoclase have an intensity

variation different from that of thè stronger and shorter

diffuse streaks occurring in intermediate microcline as a

result of a domain-twinned texture. According to Smith

(1974a, pp. 67-68; also Smith & Brown, 1988, p. 44), thè for-

mer type of streaks might be interpreted as a result of

slight development of rudimentary domains in a structure

that is essentially monoclinic.

However, thè diffuse streaks about thè main diffrac-

tion spots in many specimens of monoclinic K-rich feld-

spar result from their tweed texture, which is commonly

observed in TEM images. The tweed texture corresponds

to a modulated pattern representing an incipient stage of

thè coarser tartan texture, which results from thè M-type

association occurring in low microcline. The degree of

Si, Al order at which thè tweed texture develops in K-rich

feldspars is of great interest in alkali feldspar mineralogy,

because it represents thè switchover from monoclinic to

triclinic symmetry.

McLaren & Fitz Gerald (1987) found a value of 2tl =

0.67 in a tweed orthoclase investigated by CBED (conver¬

gei beam electron diffraction) and ALCHEMI (atom lo¬

cation by channelling enhanced microanalysis). TEM in-

vestigations performed by Bambauer et al. (1989) on thè

sanidine-microcline transition across metamorphic zones

showed that tweed orthoclase develops at 2tl = 0.69. Lat¬

er, thè latter value was changed to 0.68 by Kroll et al.

(1991; also Kroll & Knitter, 1991). On thè basis of TEM

studies, Nord (1992, p. 489) stated that thè presence of

tweed textures in orthoclase occurs at 2tl = 0.7 to 0.8. The

latter 2tl range was also adopted by Brown & Parsons

(1989, p. 31). Using thè Landau theory, Carpenter & Salje

(1994) found a value of Qt = 0.40, i.e. 2tl = 0.70, at 753 K

for thè transition between non-modulated and modulated

K-rich feldspars.

The above 2tl values, at which thè beginning of thè

tweed texture would occur in K-rich feldspars, are some-

what smaller than those proposed by other investigators

[e.g., Smith, 1974a, Fig. 9-11, A (b*c*) = 0.55, normalized

between 0 and 1, corresponding to A (b*c*) = 2tl = 0.78,

when normalized between 0.5 and 1, as in this work;

Eggleton & Buseck, 1980, 2tl = ~ 0.80; Ribbe, 1983, p. 10,

and 1984, p. 7, 2tl = 0.74; Su et al., 1984, Fig. 4, 2tl = 0.74;

Smith & Brown, 1988, p. 45, (Tlo-O) = (71-0) = ~ 1.660 À,

corresponding to 2tl = ~ 0.74].

CRYSTAL STRUCTURES OF ALKALI FELDSPARS FROM GRANITIC PEGMATITES: A REVIEW

83

Of all thè data mentioned. a 2tl value of 0.68 may rep-

resent an appropriate estimate of thè inception of thè

tweed texture in K-rich feldspars. Thus, thè Spencer C

specimen, with A (bc) - 0.716-0.726, A(b*c*) = 0.702-0.710,

and 2tl from (T-O) distances equal to 0.713 (see above),is

expected to show a tweed texture when investigated by

TEM techniques. In this light, thè Spencer C specimen

should be considered an orthoclase rather than a low sani-

dine. The situation is similar to that discussed earlier for

thè Himalaya “ordered orthoclase”.

From a historical point of view, it should be mentioned

that thè changeover from low sanidine to orthoclase was

assumed by Wright & Stewart (1968) to be marked by

their P50-56F orthoclase, a specimen with A (bc) = 0.732

and A (b*c*) - 0.721, i.e. with a structural state similar to

that of thè Spencer C orthoclase.

Itrongay orthoclase. This is a yellow, ferriferous K-rich

feldspar of gem quality from a miarolitic pegmatite at

Itrongay, near Fianarantsoa, Madagascar. Outcrops of

pegmatite are small, and minerals occur on thè surface as

crystals and fragments in part enclosed in a calcareous tu-

fa of recent formation (see Coombs, 1954). According to

Martin (1982) thè pegmatite at Itrongay is unusual, and

poorly understood from a petrological viewpoint.

With rare exceptions, thè Itrongay feldspar is homo-

geneous and its structural state, on thè basis of 2VX, ranges

from high- to low sanidine (Coombs, 1954). In spite of this,

thè feldspar is commonly denoted as thè Itrongay ortho¬

clase.

The Chemical analyses published by Coombs (1954) in¬

dicate that thè Itrongay feldspar may contain up to 3.25

wt% Fe203, which corresponds to 12.5 wt% Fe-feld. High

iron contents in specimens of gem quality and without any

turbidity or visible fine inclusions would involve that most

of thè iron replaces Al in thè T sites. Note that wet Chem¬

ical analyses reported by Coombs (1954) indicate that di-

valent iron may also occur.

According to Hofmeister & Rossman (1983, Fig. 1), thè

yellow color of thè Itrongay feldspar is due to absorption

bands in thè blue region at 418 and 445 nm, due to tetra-

hedral Fe3+. Further data obtained by Hofmeister & Ross¬

man (1984) by means of optical absorption and electron

paramagnetic resonance confirmed that all Fe3+ is tetrahe-

drally coordinated. Behrens et al. (1995) investigated thè

iron content in specimens of Itrongay orthoclase with 0.2-

2.4 wt% Fe203 by means of optical spectroscopy, 57Fe

Mòssbauer resonance absorption, at 7 K, and electron

paramagnetic resonance. Their findings imply that more

than 50% of thè iron is due to a submicroscopic exsolution

of a separate iron-rich phase that may be an iron oxide

(Fe203, Fe304) or other stili unknown phases, which con-

tribute to thè magnetic effect observed.

Rimata et al. (1996) performed a crystal structure in-

vestigation on an Itrongay K-rich feldspar at 296 and 121

K. Their paper was devoted to thè low temperature be-

havior of an orthoclase, rather than to structural study of

thè Itrongay specimen in itself. Electron microprobe

* analyses made by Rimata et al. (1996) yielded thè com-

position Si02 65.88, A1203 16.52, Fe203 1.26, and K20 16.55

J wt%, which shows excess Si02, deficient A1203 + Fe203,

and deficient K20 for a K-rich feldspar. From these data,

Rimata et al. (1996) recalculated thè Chemical formula

(K0.946Do.o54)[( Alo.902Feo.044Si3.054) Os], in which thè deviations

from thè stoichiometry of a K-rich feldspar are related to

thè anomalous Chemical contents mentioned. Taken at

face value, thè Fe content determined corresponds to 4.6

mol% Fe-feld. It should be noted that, in this specimen,

there is a complete absence of Na20.

This section will account for thè results concerning

Itrongay feldspar at 296 K, whereas its behavior at 121 K

will be considered in thè relevant section.

Turning to thè specimen investigated at 296 K, thè ab¬

sence of Na20 would be consistent with either thè high

N0r contents determined from V and a [thè average of

N0r(F) values is ~ 0.99: see Table Al-a] or thè absence of

lattice strain [N0r(a: eq. 7) - N0r(F: eq. 6) = -0.001: see

Table Al-a; A a - -0.006 À]. Note, however, that substitu-

tion of Fe3+ for Al in K-rich feldspar produces an expan-

sion of celi edge a (see Smith, 1974a, p. 276).

Refinement of thè iron scattering power for thè T sites

performed by Rimata et al. (1996) indicated complete Fe

disordering in thè framework. Disregarding thè influence

of iron content, thè (T-O) distances (see Table Al-a) indi¬

cate that apparently thè structure of thè Itrongay speci¬

men is very dose to that of thè Spencer C orthoclase (see

thè relevant section). Rimata et al. (1996) calculated thè

values of tl and t2 from (T-O) distances by means of eq.

(19) of Blasi & De Poi Blasi (1994). In thè present case,

however, eqs. (19) and (18) of Blasi & De Poi Blasi (1994)

should be used to calculate tl and t2, respectively. The use

of eq. (19) to calculate t2 gives rise to an overestimate of

thè Al content (0.148 versus 0.130: see Table Al-a). Fur-

thermore, it should be noted that Blasi & De Poi Blasi’s

(1994) equations were developed to calculate Al contents

from (T-O) distances in K-rich feldspars with AlSi3 com-

position. Ignoring thè iron content, thè Al contents of thè

Itrongay feldspar given in Table Al-a were plotted in Fig.

1 for comparison with other K-rich feldspars from granitic

pegmatites.

More recently, Nyfeler et al. (1998) investigated Si,

Al, and Fe order-disorder in another specimen of Itron-

gay orthoclase to understand thè mechanism of thè Ar

diffusion through thè feldspar structure in view of its im-

portance to K-Ar geochronology. The specimen was sev-

eral cm in size: thè crystal chip selected for microprobe

analyses had a surface of 2-5 mm2, and appeared to be

transparent and homogeneous under thè optical micro¬

scope. The average composition of thè specimen is

( K0 95N a0 05 ) [ ( Al0 95Fe0 05Si3) Og] , with Na in thè range 0.04-

0.08 (98% of thè analyses showed 0.04 < Na < 0.06) and Fe

in thè range 0.01-0.05 (98% of thè analyses showed 0.04 <

Fe < 0.05). Chemically, thè Itrongay feldspars investigated

by Nyfeler et al. (1998) and Rimata et al. (1996) differ in

their Na contents, and are very similar as regards their Fe

contents.

Nyfeler et al. (1998) investigated by X-ray diffraction

several feldspar fragments from thè crystal chip analyzed

by thè electron microprobe. The fragments were studied in

their naturai state and after annealing at temperatures up

to 950 °C for various periods of time.The lattice constants

were determined using a conventional single-crystal dif-

fractometer, and thè intensity data were collected by

means of a single-crystal diffractometer equipped with a

84

ACHILLE BLASI & CARLA DE POL BLASI

CCD detector. Complete sets of lattice constants, atomic

coordinates and displacement parameters were given for

two different crystal fragments, one of them in its naturai

state (see Table Al-a) and thè other after annealing at 900

°C for 672 h (see Table Al-b). For both of them, thè indi¬

viduai 7-0 distances were calculated in this work using

thè ORFFE4 program of Busing et al. (1985). Note that

thè values for thè atomic coordinates for 71 are inverted

with thè corresponding data for 72 in Table 5 of Nyfeler et

al. (1998). This section will account for thè results con-

cerning thè unheated Itrongay feldspar, whereas thè re¬

sults for thè annealed specimen will be considered in thè

section on thè Itrongay high sanidine.

Turning to thè unheated specimen of Nyfeler et al.

(1998), thè values of V and a are clearly different from thè

corresponding values obtained by Kimata et al. (1996) for

their specimen. As a result, thè compositions obtained

from V and a in thè former specimen are more sodic than

those observed in thè latter [thè average of N0r(K) values

is ~ 0.95 versus ~ 0.99: see Table Al-a]: in both specimens,

thè compositions from V and a are very dose to those de-

termined from microprobe analyses. As with thè feldspar

of Kimata et al. (1996), thè specimen of Nyfeler et al.

(1998) is also free from lattice strain [N0r(a: eq. 7) - N0r(K:

eq. 6) = -0.002: see Table Al-a; Da = -0.011 À],

In both specimens, thè corresponding values of b and

c celi edges are very dose to each other, and thè same

applies to thè values of thè structural indicators A (bc)

and A (b*c*) (see Table Al-a). The corresponding values of

thè (7-0) distances are virtually indistinguishable in both

specimens (see Table Al-a). As a result, thè data points for

thè two Itrongay specimens are very dose to each other in

Fig. 1.

In contrast to thè results obtained by Kimata et al.

(1996), Nyfeler et al. (1998) found that Fe is enriched in

thè 71 site, as shown by thè structural-chemical formula

(Ko.95Na005)[( Alo^FeoosSi! 28g)71( Al028gSii In this

formula, thè 7-site occupancies for Al + Fe3+ and for Si

were obtained by Nyfeler et al. (1998) from thè average

value of tl calculated from (a) b and c celi edges and (b)

(7-0) distances, using equations (IOa) and (5) of Kroll &

Ribbe (1983), respectively. The atomic proportion of Fe =

0.05, obtained from microprobe analyses (see above), was

distributed between thè 71 and 72 sites by an empirical

method based on thè values of thè electron-density differ-

ence observed and calculated for thè 71 site and for thè 72

site (see also thè section on Itrongay high sanidine).

As observed in thè section on thè Spencer C ortho-

clase, a value of 2tl = 0.68 may correspond to thè inception

of thè tweed texture in K-rich feldspars. Therefore, thè

Itrongay feldspars investigated by Kimata et al. (1996) and

Nyfeler et al. (1998), with 2tl = 0.71 from -7-OÒ distances

(see Fig. 1 and Table Al-a), could actually be orthoclase

specimens.

Further studies are needed on Itrongay K-rich feldspar

to ascertain its maximum iron content, and thè possibility

that iron orders, or not, in thè 7 sites, and also occurs in thè

form of submicroscopic exsolution. Eventually, a thorough

knowledge of thè relationship between iron behavior and

structural state would be helpful to clarify thè ordering

process in ferriferous K-rich feldspar and genetic aspects

of thè pegmatite at Itrongay.

Rabb Canyon sanidine. Keefer & G.E. Brown (1978)

refined thè crystal structures of a sanidine and a high al-

bite coexisting in partially coherent cryptoperthitic inter-

growth in a rare example of quenched pegmatite at Rabb

Canyon, Grant County, New Mexico, U.S.A. In fact, in thè

Rabb Park complex, a hypabyssal intrusive rhyolite

porhyry carries a cognate suite of quenched fragments of

incompletely crystallized granite, aplite, and pegmatite (cf.

O'Brient, 1986).

Keefer & G.E. Brown (1978) reported that thè Rabb

Canyon cryptoperthite forms clear, colorless crystals

about 1 cm across, which occur in a matrix of coarsely crys-

talline quartz. TEM observations showed a fine-scale

lamellar texture, with thè orientation of thè exsolution in¬

terface near (601). The lamellar thickness of thè dominant

K-rich phase is ~ 1000 A, and that of thè Na-rich phase ~

500 À. The K-rich phase is an untwinned sanidine, where¬

as thè Na-rich phase is a Pericline-twinned high albite with

twin lamellae ~ 50 À wide. The bulk composition deter-

mined by electron microprobe is Or51Ab48An01 mol%.

As regards thè structure of thè Rabb Canyon high al¬

bite, see thè relevant section.

Turning to thè Rabb Canyon sanidine, thè values of

(7-0) distances and Al contents (see Table Al-a and

Fig. 1) indicate appreciable Si,Al ordering, which is con-

sistent with that inferred from thè structural indicators

A (bc) and A (b*c*) (see Table Al-a). The specimen is

slightly strained, as indicated by thè difference N0r(a: eq.

5) - N0r(V: eq. 4) = 0.047 (see Table Al-a) and by a value

of A a - 0.078 A. Direct K,Na site refinement gave a value

of N0r = 0.65, which contrasts with thè N0r contents from

V in thè range 0.84-0.82 (see Table Al-a). Blasi & De Poi

Blasi (1994) contended that N0r = 0.65 is an underestimat-

ed composition because N0r values from V are much high-

er, and thè specimen is not so strained as to justify such a

discrepancy of values. Blasi & De Poi Blasi (1994) also ob¬

served that: (1) thè crystal structure of thè Rabb Canyon

sanidine is virtually identical to that of thè Eifel sanidine

#7002, an unstrained feldspar of composition N0r = 0.85

studied by Phillips & Ribbe (1973); (2) thè ( M-O ) distance

and, in particular, thè individuai M-OA2 distance of thè

Rabb Canyon sanidine indicate highly potassic composi¬

tion; (3) thè N0r values estimated by Keefer & G.E. Brown

(1978) using thè method of Robin (1974) and that of Tullis

(1975) are 0.92 and 0.79, respectively.

To clarify thè situation, Blasi & De Poi Blasi (1994) ex-

tensively investigated K,Na site refinement in a K-rich

feldspar of well-known composition: thè Pikes Peak mi¬

crocline 7813C (see thè relevant section). They explored

several models of M-site refinement, and found that thè

resulting compositions tend to be underestimated unless

thè particular procedures described are adopted.

Buck Claim adularla. Ferguson et al. (1991) refined

thè crystal structures of four specimens of adularia from

an Archean granitic pegmatite at thè Buck Claim, Bernic

Lake, Manitoba, Canada.

The pegmatite is of thè complex (Li, Rb, Cs)-enriched

type, petalite subtype (see Cerny, 1989). The adularia oc¬

curs in open fractures on drusy surfaces in thè massive

quartz core of thè pegmatite, closely associated with more

abundant carbonate-rich apatite crystals. The feldspar

CRYSTAL STRUCTURES OF ALKAL1 FELDSPARS FROM GRANITIC PEGMATITES: A REVIEW

85

crystals, up to 0.25 mm across, show thè Felsòbanya-

Maderaner habit of adularia described by Cerny & Chap-

man (1986) and are iron-oxide-stained but otherwise col-

orless. The four crystals used in structure determination

were selected from specimen Buck-18 (see Cerny & Chap-

man, 1984). They were denoted by Ferguson et al. (1991) as

Buck-18-I, -II, -III, and -IV.

Microprobe analyses performed by Ferguson et al.

(1991) on specimens Buck-18-I and -IV, as well as on four

other crystals from specimen Buck-18, yielded ex-

ceptionally pure compositions, consistent with Or100

mol%. Similar compositions are also expected for speci¬

mens Buck-18-II and -III, because thè N0r contents from V

and a are very high in all four Buck Claim specimens (see

Table Al-a).

In thè diagram of b and c celi edges, thè representative

point for specimen Buck-18-IV plots along thè low micro¬

cline - high sanidine join of thè quadrilateral of alkali

feldspars, towards thè corner for high sanidine. In thè

same diagram, thè representative points for specimens

Buck-18-I, -II, and -III lie outside thè above quadrilateral,

near thè corner for high sanidine (Ferguson et al., 1991,

Fig. 1; also Blasi & De Poi Blasi, 1994, Fig. 2). The three

specimens Buck-18-I, -II, and -III deserve careful atten-

tion. Specimens Buck-18-I and -III lie on iso-A(bc) or iso-

A(b*c*) straight lines (see Blasi, 1977, for their definition;

also Blasi & De Poi Blasi, 1994, Fig. 2) which are just out¬

side thè quadrilateral of alkali feldspars. Specimens Buck-

18-11 and -III lie on iso-Or(6c) or iso-Or(6*c*) straight

lines (see Blasi, 1977, for their definition; also Blasi & De

Poi Blasi, 1994, Fig. 2) which are rather far from thè

quadrilateral of alkali feldspars.

The above observations allow proper comparison of all

four specimens Buck-18-I, -II, -III, and -IV to be made in

terms of thè structural indicators A (bc) and A (b*c*) (see

Table Al-a) and thè strain index A a. In terms of A (bc) and

A (ò*c*), specimen Buck-18-I [A (bc) = 0.492 and A(6*c*) =

0.486] and specimen Buck-18-III [A (bc) = 0.472 and

A(b*c*) = 0.469] show strong Si,Al disorder with anom-

alous Al contents. By contrast, specimen Buck-18-II [A(6c)

= A (b*c*) = 0.518] shows strong Si,Al disorder with nor-

mal Al contents, and specimen Buck-18-IV [A (bc) = 0.615

and A(b*c*) = 0.617] appears to be much more ordered. In

terms of A a, specimen Buck-18-II (A a = -0.080 À) and

specimen Buck-18-III (A a = -0.069 À) show thè reverse

strain described by Blasi et al. (1984a). By contrast, speci¬

men Buck-18-I (A a = -0.028 À) and specimen Buck-18-IV

(Aa = -0.003 À) are in a normal range.

The three specimens Buck-18-I, -II, and -III are not

unique in their extreme Si.Al disorder. The occurrence of

other specimens of adularia with compositions dose to thè

pure end-member KAlSi308 was documented in granitic

pegmatites and hydrothermal non-alpine vein deposits by

Cerny & Chapman (1984; 1986). Their investigations, by

X-ray powder diffractometry and infrared absorption

spectrometry, showed unusually high Si,Al disorder in

these specimens. Another adularia, thè Eagle Harbor

specimen of Stewart & Wright (1974), also shows b and c

‘ values in thè range of thè corresponding data obtained by

»ern_ & Chapman (1984; 1986) from their specimens. The

behavior of b and c , as well as b* and c*, in all thè speci¬

mens of adularia investigated by Stewart & Wright (1974),

Cerny Chapman (1984; 1986), and Ferguson et al. (1991)

was displayed by Blasi & De Poi Blasi (1994) in their Fig. 2.

Among all these specimens of adularia, those of Fer¬

guson et al. (1991) were thè sole feldspars used in structure

determination. The R values in thè four crystals are in thè

range 0.042-0.058 (Table Al-a). Ferguson et al. (1991) at-

tributed these high R values to thè poor quality of their

specimens, which optically show a strong wavy extinction.

The (T- O) distances indicate that thè adularia Buck-18-IV

is thè most ordered specimen among thè four crystals

studied (Table Al-a): its representative point plots, in Fig.

I, very dose to thè Rabb Canyon sanidine of Keefer &

G.E. Brown (1978). The representative points for speci¬

mens Buck-18-II and -III plot, in Fig. 1, approximately

half-way between thè representative points for specimen

Buck-18-IV and specimen Buck-18-I, which is thè most

disordered specimen.

Ribbe (1994: caption to Fig. 6a) observed that all four

refinements by Ferguson et al. (1991) show considerably

larger than expected (72-0) and (( T-O )) values. This state¬

ment derives from thè fact that in Fig. 6a of Ribbe (1994)

thè (71-0) distance was plotted for each Buck Claim spec¬

imen versus thè ((T’-O)) distance, instead of thè (72-0) dis¬

tance. The (72-0) and ((T-O)) distances are normal for

specimens Buck-18-II, -III, and -IV: thè (72-0) distance in

specimen Buck-18-I is slightly larger (within 2s) than thè

(Tl-O) distance (see Table Al-a).

For thè three most disordered specimens Buck-18-I, -

II, and -III, Blasi & De Poi Blasi (1994, p. 68) observed: “It

is surprising that thè most ordered* specimen in terms of

T-O distances appears to be thè least ordered* in terms of

lattice constants” (*misprinted as disordered in thè origi¬

nai paper). This statement is justified by a comparison of

2tl values, obtained from (T-O) distances, and thè values

of thè structural indicators A (bc) and A(ó*c*) for thè three

most disordered specimens [Buck-18-I: 2tl = 0.479, A (bc)

= 0.492 and A (b*c*) = 0.486; Buck-18-II: 2tl = 0.533, A (bc)

= A (b*c*) = 0.518; Buck-18-III: 2tl = 0.555, A (bc) = 0.472

and A (b*c*) = 0.469],

For thè sake of completeness, in specimen Buck-18-IV

it is 2tl = 0.613, A (bc) = 0.615 and A(b*c*) = 0.617. To sum

up: in terms of (T-O) distances, Si,Al order decreases from

specimen Buck-18-IV to -III, -II, -I, whereas in terms of

celi parameters Si, Al order decreases from specimen

Buck-18-IV to -II, -I, -III.

Blasi & De Poi Blasi (1994, p. 65) observed that thè

major uncertainties in thè choice of reference values of

lattice constants for thè alkali feldspar end-members oc-

cur in thè case of high sanidine. This is due to thè occur¬

rence of specimens like those investigated by Stewart &

Wright (1974), Cerny & Chapman (1984; 1986), and

Ferguson et al. (1991), which plot outside thè quadrilat¬

eral of alkali feldspars (see Blasi & De Poi Blasi, 1994:

Fig. 2). Such specimens pose severe problems, and their

crystal structure should be determined carefully in order

to verify whether thè Al contents obtained from (T-O)

distances are consistent with those determined from met-

ric properties. A clear example of inconsistency is that

displayed by specimen Buck-18-III, which shows

“anomalous” metric properties when compared with (T-

O) distances that are within a normal range for high

sanidine.

86

ACHILLE BLASI & CARLA DE POL BLASI

Teertstra et al. (1998) confirmed that thè composition

of thè Buck-18-I adularla is dose to that of thè end-mem-

ber KAlSi308, and observed that thè metric and structural

data of this specimen may be taken as thè most reasonable

values for Or100 end-member high sanidine.

Indeed, thè Buck-18-I specimen of adularla is of great

value, because it demonstrates that essentially complete

Si.Al disorder can be found in naturai K-rich feldspars.

However, it is worth investigating further to clarify subtle

details of its crystal structure, e.g., (72-0) distance slightly

greater than thè -71-00 distance. It would be desirable to

know whether thè Buck Cairn feldspars show thè diffuse

extra diffractions hkl with ( h + k ) odd, which commonly

occur in specimens of adularia.

As regards thè conditions of formation of thè Buck

Cairn specimens. Cerny & Chapman (1984) and Ferguson

et al. (1991) accepted as thè likely origin that proposed by

Foord & Martin (1979) and Martin (1982) for late high-

sanidine overgrowths occurring in pockets from granitic

pegmatites in thè Pikes Peak batholith, Colorado, U.S.A.

Foord & Martin (1979) and Martin (1982) suggested that

these pockets were subjected to “pressure quenching”, i.e.,

sudden decrease in T’(fluid), with consequent rapid nucle-

ation and crystal growth of a highly disordered K-rich

feldspar phase. According to Ferguson et al. (1991; also

Cerny & Chapman, 1984; Cerny, 1994) thè extreme Si, Al

disorder in thè Buck Cairn specimens was preserved

metastably because thè low-temperature hydrothermal

precipitation of high sanidine adularia was not followed by

thè action of any subsequent order-promoting agents.

Room-temperature crystal structures of

PRE-HEATED K-RICH FELDSPARS

Itrongay high sanidine. As mentioned earlier (see thè

section on thè Itrongay orthoclase used in structure deter-

mination at room temperature), Nyfeler et al. (1998) in-

vestigated by X-ray diffraction several crystal fragments

from thè Itrongay orthoclase after annealing at tempera-

tures up to 950 °C for various periods of time. The lattice

constants, atomic coordinates and displacement parame-

ters, however, were presented only for one of these

feldspar fragments, which was annealed at 900 °C for 672

h. The fragment is different from that investigated by

Nyfeler et al. (1998) in its naturai state, but both specimens

were isolated from thè same crystal chip.

The compositions from V and a of thè annealed K-rich

feldspar are definitely more sodic than those of thè un-

heated specimen [thè average of N0r(V) values is ~ 0.90

versus ~ 0.95: see Table Al-b and Table Al-a], This sug-

gests that an exsolved Na-rich phase was resorbed during

heating.The absence of lattice strain in thè unheated spec¬

imen [N0r(n: eq. 7) - N0r(F: eq. 6) = -0.002 and A a = -0.011

A: see thè section on thè Itrongay orthoclase investigated

at room temperature], indicates that thè Na-rich phase

might be microperthitic rather than cryptoperthitic. By

contrast, Nyfeler et al. (1998) reported that their specimen

showed optical homogeneity. In thè annealed Itrongay or¬

thoclase, thè difference N0r(a: eq. 5) - N0r(F: eq. 4) be-

comes 0.006 and thè A a value changes to -0.040 À. Note

that negative Aa values seem to be an idiosyncratic feature

in heated K-rich feldspars (see thè section on thè Bedford

high sanidine).

The R values in thè annealed specimen are slightly

higher than those in thè unheated specimen (see Table Al-

b and Table Al-a). The (T- O) distances indicate that thè

heat-treatment produced a high sanidine, with strong but

not complete Si,Al disorder [2tl = 0.545(6): Table Al-b].

Note that thè values of A (bc) and A(6*c*), 0.511 and 0.519,

respectively (Table Al-b), are consistent with rather

stronger Si, Al disorder. In Fig. 1, thè representative point

of thè Itrongay high sanidine plots between that for thè

Spencer C high sanidine and that for thè Buck-18-III adu¬

laria.

Nyfeler et al. (1998) found that prolonged heating at

900 °C renders Fe equally distributed between thè 71

and 72 sites according to thè structural-chemical formula

( K0.95N a0 05 ) [ ( Al0 523Fe0 025Si j 452) T] ( Al0 427Fe0 025S i 1.548) nOs].

In this formula, thè 7-site occupancies were obtained as

described for thè unheated specimen investigated at RT

by Nyfeler et al. (1998) (see thè section on Itrongay ortho¬

clase). Note that this formula does not take into account

thè fact that K,Na homogenization also occurred, as docu-

mented by thè N0r contents obtained from celi volume V.

Nyfeler et al. (1998) also observed that prolonged heat¬

ing at 900 °C and quenching reduce thè value of an empir-

ical extinction parameter, which was proposed by

Sheldrick (1993) to cover both primary and secondary ex¬

tinction. The X-ray extinction parameter is an indicator of

crystal perfection (see Larson, 1970; also Becker & Cop-

pens, 1974). Thus, Nyfeler et al. (1998) deduced that thè

relatively high values of thè extinction parameter in un¬

heated specimens [range: 0.076(3)-0. 147(5)] indicate that

Itrongay orthoclase crystals are rather ideal, with no pro-

nounced mosaic pattern, whereas thè relatively low values

of thè extinction parameter found after heating and

quenching [range: 0.012(2)-0.020(2)] suggest that forma¬

tion of mosaic blocks had to occur.

Nyfeler et al. (1998), who investigated Itrongay or¬

thoclase to understand thè release of Ar from K-rich

feldspar in view of its importance to K-Ar geochronology,

concluded that Ar diffusion upon heating is related to T-

O bond breaking as a result of Si,Al rearrangement.

Spencer C high sanidine. A crystal fragment from thè

Spencer C orthoclase (see thè relevant section) was con-

verted to high sanidine by Cole et al. (1949) by heating to

1075 °C for 300 hours.

Cole et al. (1949) performed a 3D Fourier refinement

of thè sanidinized Spencer C orthoclase using intensity

data collected by Weissenberg photographs. The R value

was 0.14, and thè (71-0) and (72-0) distances were

both equal to 1.642 À, indicating complete Si, Al disorder.

Later, thè intensity data collected by Cole et al. (1949)

were subjected to 3D least-squares refinement by Ribbe

(1963). The R value dropped to 0.099, and thè values of

(Tl-O) and (72-0) distances became equal to 1.645 and

1 .640 A, respectively. These data, as well as thè Al contents

calculated therefrom, indicate that thè structure is not ful-

ly disordered (see Table Al-b). Fig. 1 shows that thè sani¬

dinized Spencer C orthoclase is slightly more disordered

than thè Itrongay orthoclase heated by Nyfeler et al.

(1998).

CRYSTAL STRUCTURES OF ALKALI FELDSPARS FROM GRANITIC PEGMATITES: A REVIEW

87

The lattice constants adopted by Ribbe (1963) were

those determined by Cole et al. (1949) using thè 0-method

of Weisz et al. (1948), which gives very high accuracy. The

compositions determined from V and a [thè average of

N0r(F) values is ~ 0.89: see Table Al-b] are very dose to

those obtained from thè unheated specimen (see Table

Al-a).These data confirm that thè Spencer C orthoclase is

virtually free from exsolved Na-rich phase (see thè section

on thè Spencer C orthoclase). As expected, lattice strain is

absent [N0r(n: eq. 5) - N0r(V: eq. 4) = 0.004: see Table Al-

a; A a - -0.046 A], as in thè unheated Spencer C specimen.

Note thè negative Aa value, which occurs in many heated

K-rich feldspars (see thè section on thè Bedford high sani-

dine). The values of thè structural indicators A (bc) and

A (b*c*), 0.514 and 0.520, respectively, indicate a degree of

Si,Al disorder tangibly higher than that estimated from

thè Al contents calculated from thè (7- O) distances deter¬

mined by Ribbe (1963) (2tl = 0.537: see Table Al-b).

Bedford high sanidine. Blasi et al. (1987b) subjected a

cleavage fragment of low microcline to dry annealing at

1050 °C for 200 days, followed by quenching to room tem¬

perature. The fragment was isolated from a hand specimen

of perthitic coarsely cross-hatched low microcline from a

granitic pegmatite occurring in thè Bedford County peg-

matite district, Virginia, U.S. A.

The originai K-rich feldspar material of bulk com-

position Or783Ab21 ^An^Oio jRbf^Srfoo mol% was puri-

fied by Blasi et al. (1984a) to Or907Ab88An0 jCn^

Rbf0 2Srf0 j mol% by magnetic and gravity separation: both

bulk compositions were determined by flame photometry

(Na and K) and X-ray fluorescence spectrometry (all oth-

er elements). Blasi et al. (1984a) used thè purified feldspar

material in dry-heating experiments at 1050 °C for periods

of time up to 150 days to produce a K-rich feldspar disor-

dering series. The conversion of low microcline to high

sanidine proceeded via a one-step disordering process. Mi¬

croprobe analyses of thè K-rich phase of thè unheated ma¬

terial yielded Or929Ab65An01 Cn03Rbf02Srf01 mol%, and

those of thè extreme end-member of thè disordering series

gave Or90 5 Ab9 2An0 3 Cn02Rbf01Srf00 mol%. As a result of

K,Na homogenization, thè latter composition is very dose

to that determined by flame photometry and X-ray fluo¬

rescence spectrometry. The unheated feldspar material

gave N0r( V: eq. 3) = 0.947, N0r(V: eq. 8) = 0.949, N0r(a: eq.

9) = 0.960, A a = 0.012 À, A (bc) = 0.997, A (b*c*) = 0.993,

A(ay) = 1.003, and A(a*y*) = 1.003; thè most disordered

material gave N0r(V: eq. 1) = 0.923, N0r(T: eq. 4) = 0.897,

N Qr(a: eq. 5) = 0.903, A a = -0.039 À, A (bc) = 0.504, A (b*c*)

= 0.511 (for uniformity, N0r values and structural indica¬

tors were recalculated from originai data, as indicated in

thè section on methods). Thesa data indicate that thè un¬

heated material was slightly perthitic and free from lattice

strain. The most disordered material was an essentially ful-

ly disordered and chemically homogeneous high sanidine:

its N0r values are very dose to thè bulk composition de¬

termined by flame photometry and X-ray fluorescence

spectrometry.

The crystal fragment used in structure refinement by

Blasi et al. (1987b) was isolated from thè same unheated

purified feldspar material as that studied by Blasi et al.

(1984a). Single-crystal X-ray photographs showed that, af¬

ter annealing, thè crystal fragment was transformed into

an untwinned homogeneous C2/m K-rich feldspar show-

ing sharp diffraction maxima. The metric properties are

dose to those determined from X-ray powder dif-

fractometer data for thè extreme end-member of thè dis¬

ordering series prepared by Blasi et al. (1984a).

The compositions determined from V and a [thè av¬

erage of N0r(V) values is ~ 0.90: see Table Al-b] are very

dose to thè bulk composition of thè purified feldspar ma¬

terial determined by flame photometry and X-ray fluores¬

cence spectrometry, as a result of K.Na homogenization

produced by heat-treatment. As expected. thè difference

N0r(n: eq. 5) - N0r(V: eq. 4) = -0.002 (see Table Al-b) and

thè value of A a = -0.060 A are consistent with absence of

lattice strain. Note that many heated K-rich feldspars, in-

cluding thè Itrongay and Spencer C specimens (see thè

relevant sections), tend systematically to show negative A a

values. The effect, designated reverse strain by Fenn &

G.E. Brown (1977), is not yet completely understood (see

discussion in Blasi et al., 1984a). The values of thè struc¬

tural indicators A (bc) and A (b*c*), 0.504 and 0.512, re¬

spectively, indicate almost complete Si,Al disorder (see

Table Al-b).

The structure refinement of thè sanidinized Bedford

low microcline performed by Blasi et al. (1987b) led to

very low R values ( R 0.022 and wR 0.024, see Table Al-b),

with (TI-O) = 1.643 and (77-0) = 1.640 À. These data in¬

dicate a degree of Si,Al disorder tangibly higher than that

achieved by thè sanidinized Spencer C orthoclase of

Ribbe (1963). Thus, thè Bedford sanidine represents thè

most disordered specimen among heated K-rich feldspars.

This can also be seen by thè position of its representative

point in Fig. 1. Note that thè value of 2tl = 0.523, calculat¬

ed from thè (71-0) distance, is very dose to thè above-

mentioned values of A (bc) and A (b*c*) (see Table Al-b).

The results obtained by Blasi et al. (1984a, 1987b)

demonstrate that thè achievement of complete disorder in

dry-annealing experiments is a more sluggish process than

expected, with important implications for alkali feldspar

mineralogy. The occurrence in nature of specimens of high

sanidine such as thè adularia Buck-18-I of Ferguson et al.

(1991) represents an exceptional example of extreme

Si, Al disorder in highly metastable conditions (see also

thè section on thè Buck Claim adularia).

Low-TEMPERATURE CRYSTAL STRUCTURES OF

natural K-rich feldspars

Prilep low microcline at 163 K. A crystal fragment

(0.09 x 011 x 0.18 mm) extracted from thè same specimen

of Prilep microcline as that studied by Strob (1983) (see

thè relevant section) was subjected to structure re¬

finement by Phillips et al. (1988) at 294 (R 2.0%) and 163

K (R 1.9%) in order to investigate thè room-temperature

phase transition described by Openshaw et al. (1979a,

1979b), Wyncke et al. (1981), and Brown et al. (1984; also

Smith & Brown, 1988, p. 75 and p. 162) for low microcline.

The results obtained by Phillips et al. (1988), which were

not published in detail, would not be consistent with thè

inferred room-temperature phase transition in low micro¬

cline (see also Ribbe 1994, p. 10).

88

ACHILLE BLASI & CARLA DE POL BLASI

Itrongay orthoclase at 121 K. This section will account

for thè structural behavior of thè Itrongay orthoclase in¬

vestigateci at 121 K by Rimata et al. (1996). For results on

thè same crystal fragment studied at 296 K by thè same au-

thors, see thè relevant section.

Rimata et al. (1996) observed a strong anisotropie con-

traction of thè structure at 121 R: celi edge a decreases

from 8.600 À at 296 R to 8.574 À at 121 R, thè fi angle in-

creases from 116.03° at 296 R to 116.07° at 121 R, and celi

edges b and c show little variation. As a result, celi volume

V decreases from 722.9 À3 at 296 R to 720.2 À3 at 121 R.

In these circumstances, thè N0r contents in Table Al-c are

meaningless because they also reflect thè contraction of V

and a. In this specimen, an additional effect on V and a is

due to thè presence of iron content (see thè section on thè

Itrongay specimen investigated at room temperature). The

values of A(bc) and A (b*c*), 0.684 and 0.649, respectively,

with a somewhat high difference of 0.035, bracket thè val¬

ues of A (bc) and A (b*c*), 0.670 and 0.674, respectively,

observed in thè Itrongay orthoclase investigated by Rima¬

ta et al. (1996) at room temperature (see Tables Al-c and

Al-a).

The R values are higher at 121 R: R 0.028 at 296 R ver¬

sus 0.044 at 121 R, and wR 0.028 at 296 R versus 0.042 at

121 R. The (71-0) distance increases from 1.655 À at 296

R to 1.660 À at 121 R, whereas thè (72-0) distance shows

little variation (see Table Al-a and Table Al-c). The (R-O)

distance decreases from 2.986 À at 296 R to 2.969 À at 121

R.The apparent expansion of thè (71-0) distance at 121 R

may be due to reduction in thè swinging-arm effect in re-

sponse to decreased thermal motion; see thè section on

reference values for Al-O and Si-O distances in alkali

feldspars for information on thè swinging-arm effect. Ob-

viously, thè reduction in this effect also influences thè Al

contents derived from (7-0) distances. As a result, thè po-

sition of thè representative point of thè Itrongay specimen

at 121 R in Fig. 1 is apparently consistent with a degree of

Si, Al order higher than that displayed by thè representa¬

tive point of thè same specimen at 296 R.

A linear extrapolation of thè displacement parameter

B for thè R atom at 0 R showed that there is a finite in-

tercept of B = 0.3 À2. According to Rimata et al. (1996),

thè intercept is evidence of zero-point motion that would

not depend on thermal energy but might reflect static

Si, Al disorder or dynamic energy resulting from collision

of R and O atoms during data collection under thè Earth’s

gravity.

Room-temperature crystal structures

OF NATURAL Na-RICH FELDSPARS

Ramona low albite. The specimen, denoted as No. 29,

is a low albite from a suite of plagioclase feldspars given

by R.C. Emmons to W.H. Taylor, and Comes from a cavity

in an albitized pegmatite at Little Three mine, Ramona,

San Diego County, California, U.S.A. (Ferguson et al.,

1958; also Emmons et al., 1953). Detailed investigations on

thè mineralogy and genetic aspects of thè Little Three peg-

matit re provided by Stern et al. (1986) and Foord et

al. (1989). Many of thè old references can be found in

these two papers.

Emmons et al. (1953) reported a complete wet Chemi¬

cal analysis for thè Ramona albite indicating that thè spec¬

imen is a nearly pure NaAlSi3Os. From this Chemical

analysis, Ferguson et al. (1958) calculated a composition of

OrU)Ab985An05 mol%. The lattice constants determined

by Ferguson et al. (1958) are given in Table A2-a.The val¬

ues of N0r calculated from V and a are in dose agreement

with thè Chemical composition. As expected, given thè

non-perthitic character of most low albite crystals, these

data are consistent with thè absence of lattice strain due to

coherency between exsolved phases. The values of thè

structural indicators obtained from thè A method indicate

a highly ordered Si,Al distribution. This, however, appears

to be somewhat anomalous, because thè values of thè

structural indicators obtained from celi angles are defi-

nitely higher than those obtained from celi edges: perti-

nent interrelationships among Al contents estimated from

thè structural indicators A (bc), A (b*c*), A(ay), and

A(a*y*) can be seen in Fig. 4 of Blasi & Blasi De Poi

(1977). Ferguson et al. (1958) determined thè crystal struc¬

ture of thè Ramona albite by means of 2D Fourier projec-

tions parallel to all three axes: thè R values were in thè

0.080-0.093 range. The study showed that thè Ramona al¬

bite is highly ordered ((Tlo-O) = 1.742, (71/n-O) = 1.590,

(Tlo-O) = 1.636, ( Tlrn-O ) = 1.616 À), and thè electron-

density distribution of thè Na atom is largely anisotropie.

Later, Ribbe et al. (1969) subjected thè Weissenberg in-

tensity data collected by Ferguson et al. (1958) to a 3D

least-squares refinement. The R value dropped to 0.068,

and thè new (7-0) distances (see Table A2-a) approximate

to those determined by X-ray counter techniques in spec-

imens of low albite (see, for instance, thè data for Amelia

low albite in Table A2-a). The Al contents determined

from (7-0) distances evidence a behavior similar to that

observed above for thè structural indicators derived from

lattice constants. Indeed, thè value of tlo - tlm is tangibly

higher than that of tlo + tlm (see Fig. 2): this seems to be

an idiosyncratic feature of low albite that should be fur-

ther investigated.

Ribbe et al. (1969) also confirmed thè anisotropy of thè

Na atom. They explored split-atom models to see whether

thè anisotropy can be interpreted in terms of space-aver-

age for a half-atom model or a quarter-atom model, rather

than in terms of time-average of a single Na atom giving

rise to anisotropie displacements.

The quarter-atom model was discarded virtually from

thè outset, because even thè half-atom model was at thè

limit of resolution. In thè half-atom model, thè single

anisotropie Na atom was replaced by two isotropie half-

o

atoms separated by 0.36 A along thè direction of thè max¬

imum electron-density distribution in thè single-atom

model.

On thè basis of thè experimental data available, thè

generai conclusion reached by Ribbe et al. (1969) on thè

Na atom anisotropy was rather uncertain. In thè end, their

position was in favor of thè time-average interpretation of

thè Na-atom anisotropy.

The interpretation proposed by Ribbe et al. (1969) was

confirmed by Quareni & Taylor (1971), and, more recent-

ly, by thè findings of Winter et al. (1979) and Smith et al.

(1986) (see thè section on thè Ramona low albite at 93 R

and that on Amelia low albite at 13 R).

CRYSTAL STRUCTURES OF ALKALI FELDSPARS FROM GRANITIC PEGMATITES: A REVIEW

89

Amelia low albite. Harlow & G.E. Brown (1980) car-

ried out a crystal structure investigation on an Amelia low

albite by X-ray and neutron diffraction methods at room

temperature.

The low albite investigated occurs as a cleavelandite in

a pegmatite from thè Rutherford mines, near Amelia,

Amelia County, Virginia, U.S.A. It is a water-clear speci¬

men, known for its high Chemical and phase purity (Wald-

baum & Robie, 1971; also Sinkankas, 1968; Deer et al.,

1975,Table 13).

The specimen, supplied by D.R. Waldbaum, was also

used by Harlow & G.E. Brown (1980) to obtain lattice

constants from X-ray powder diffractometer data (Table

A2-a). Later, Smith et al. (1986) carried out a neutron dif¬

fraction study at 13 K on thè same feldspar material (see

thè relevant section).

Microprobe analysis reported on by Harlow & G.E.

Brown (1980) yielded a composition of Or06Ab993 An01

wt% (see also thè section on Amelia monalbite), which is

in dose agreement with thè values of N0r obtained from V

and a (Table A2-a). As with most specimens of low albite,

these data indicate that thè Amelia albite is non-perthitic

and, therefore, free from lattice strain due to coherency

between exsolved phases. The structural indicators ob¬

tained from lattice constants indicate that thè Amelia Al¬

bite is essentially fully ordered. Unlike thè Ramona albite,

thè values of thè structural indicators obtained from celi

angles in thè Amelia Albite are only very slightly higher

than those obtained from celi edges (see Table A2-a).

Comparison of thè crystal structures determined from

X-ray and neutron data shows that thè positional pa-

rameters are essentially identical, whereas slight differ-

ences occur in thè displacement parameters (e.g., thè val¬

ues of equivalent isotropie temperature factors, B, for T

and Na sites from neutron data are higher than those from

X-ray data).

The values of (T-O) distances from X-ray and neutron

data are very dose to each other, and indicate essentially

complete Si,Al order with more Al in Tlo or in T2m site

than in Tlm site (see Table A2-a). In other words, thè val¬

ues of tlo - tlm are higher than those of tlo + tlm (see Fig.

2), a feature already noticed also in thè crystal structure of

thè Ramona albite at room temperature (see thè preced-

ing section). Refinement of T-site occupancies from X-ray

data was not successful. Al contents from refinement of

thè neutron scattering lengths indicate almost complete

Si, Al ordering with no Al in Tlo and T2m sites: unlike Al-

content data from (T-O) distances, it is, therefore, (tlo -

tlm) < (tlo + tlm), i.e. 0.935 versus 1.005 (see Table A2-a).

Considering that all thè data on thè Amelia albite are

consistent with essentially complete Si, Al order or so, thè

(T-O) distances, obtained from X-ray and neutron data,

can be used to estimate thè Al-O and Si-O distances in Na-

rich feldspars (see thè relevant section).

Harlow & G.E. Brown (1980) found that thè relation-

ship between (T-O) distances and Al contents from neu¬

tron diffraction data (Amelia low albite; Himalaya or-

thoclase: Prince et al., 1973; Eifel low sanidine: G.E.

Brown et al., 1974) is best fulfilled by a 3rd-order poly-

nomial.

Harlow & G.E. Brown (1980) made an analysis of thè

displacement of thè Na atom in low albite: time-av-

erage versus space-average behavior, which was thè

subject of discussion for years. They compared thè sin-

gle-atom (time-average) model with thè half-atom

(space-average) model, and found that thè single-atom

model yields a smaller weighted R factor for both thè

neutron and thè X-ray data. This is in agreement with thè

findings of Winter et al. (1977), who demonstrated that

thè values of thè displacement parameter B for thè Na

atom in specimens of low albite studied at 93 to 1243 K

yield a linear extrapolation to zero at 0 K. indicating that

reai thermal motion is thè cause of thè large Na site ani-

sotropy.

Rabb Canyon high albite. This is thè Pericline-

twinned Na-rich phase coexisting in partially coherent

cryptoperthitic intergrowth with a sanidine in a quenched

pegmatite from Rabb Canyon, Grant County, New Mexi¬

co, U.S.A. (see thè section on Rabb Canyon sanidine).

The cryptoperthite was investigated by Keefer & G.E.

Brown (1978), who determined thè crystal structures of

thè individuai phases. The Rabb Canyon high albite is a

unique example of naturai highly disordered albite used in

structure determination. However, as will be shown below,

there were severe problems with structure refinement.

Keefer & G.E. Brown (1978) made structure refine-

ments of thè high albite using half-atom and quarter-atom

models (see thè section on Amelia high albite). These

models were not satisfactory, and thè single-atom model

was adopted with an alkali site occupancy fixed at

Na08K02. Unlike thè R factors in thè sanidine (R = 0.029

and wR - 0.039), thè R factors in thè high albite were

rather high ( R = 0.083 and wR = 0.096). Keefer & G.E.

Brown (1978) attributed these high values to thè difficul-

ties encountered in measuring thè high-albite intensities.

The amount of high albite was rather small in thè crystal

fragment investigated, and any interference on intensity

data from thè coexisting sanidine was amplified. Lattice

strain may be another source of error, because it may vary

within each albite lamella producing thè high displace¬

ment parameters observed. Note that each lamella is also

Pericline-twinned with twin lamellae ~ 50 À wide, which

may further complicate thè shape of diffraction peaks.

The values of thè (T-O) distances and thè Al contents

in Table A2-a indicate appreciable Si.Al ordering. The val¬

ues of (Tlo-O) and (Tlm-O) distances are virtually identi¬

cal, and those of (T2o-0) and (Tlm-O) distances are very

dose to each other. In space group Ci, this configuration

corresponds to a structural state dose to that of analbite

(tlo = tlm s t2o = t2m) and approximates to a two-step or¬

dering trend (see Fig. 2).

By contrast, thè metric properties in Table A2-a are

consistent with almost complete Si, Al disorder. Note that

thè difference between A (bc) and A(b*c*) values is

0.141, an abnormally high value, that, according to Blasi

(1980), may indicate that unit-cell parameters conceal

errors.

°

Lattice strain in thè Na-rich phase (A a - -0.234 A) of

thè Rabb Canyon cryptoperthite is much higher that in

thè K-rich phase (A a = 0.078 À). These data are con-

firmed by thè contrasting values of N0r from V and a

(see Table A2-a and Table Al-a). This is in agreement

with thè high proportion of thè sanidine phase in thè crys-

90

ACHILLE BLASI & CARLA DE POL BLASI

tal (mole fractions: 0.68 from X-ray scale-factor re-

finement and 0.67 from TEM image).

K,Na site refinement in thè high albite failed to pro¬

duce a chemically reasonable composition. This was

determined by mass balance using thè sanidine compo¬

sition, thè mole fraction of thè sanidine in thè crystal,

and thè bulk composition of thè crystal (see thè section

on Rabb Canyon sanidine) as Or22Ab78. Note that celi

volume V gives N0r values in thè 0.12-0.14 range (Table

A2-a).

Room-temperature crystal structures of

PRE-HEATED NA-RICH FELDSPARS

Amelia high albite. The specimen, denoted as No. 31

in thè suite of feldspars described by Emmons et al. (1953),

is a low albite from a pegmatite from thè Rutherford

mines, near Amelia, Amelia County, Virginia, U.S.A. (see

also thè section on Amelia low albite). The data from a

complete wet Chemical analysis reported on by Emmons et

al. (1953) were used by Ferguson et al. (1958) to calculate

a composition of Or16Ab97 7Anu7 mol%.

Ferguson et al. (1958) heated a few fragments of thè

Amelia low albite at 1065 °C for 16 days. After cooling, op-

tical measurements and X-ray powder diffraction data

showed that thè material has been transformed into high

albite. Most of thè fragments appeared to be twinned poly-

synthetically. A fragment free from twinning was subject-

ed to single-crystal X-ray investigation. The N0r contents

calculated from V are in dose agreement with thè Chemi¬

cal composition, whereas thè N0r value obtained from a is

anomalous. The metric properties are consistent with a

highly disordered Si, Al distribution. However, thè differ-

ence A(bc ) - A (b*c*) - -0.040 (see Table A2-b) is anom¬

alous for a high albite, and, according to thè findings of

Blasi (1980), this may imply that thè lattice constants con-

ceal errors: an unexpected event in this specimen, consid-

ering that thè lattice constants were measured by thè “0-

method” of Weisz et al. (1948), which, in principle, gives

very high accuracy. On thè other hand, thè A (bc) - A (b*c*)

difference evidences different behavior in other speci-

mens of high albite investigated at room temperature [e.g.,

Prewitt et al., 1976: synthetic high albite: A (bc) = 0.498 and

A (b*c*) - 0.494; Winter et al., 1979: Tiburon high albite:

A (bc) = 0.502 and A (b*c*) = 0.492],

The structure analysis performed at RT by means of

2D Fourier projections parallel to all three axes gave R

values in thè 0.14-0.19 range. The results obtained by

Ferguson et al. (1958) showed that thè crystal structure

of thè Amelia high albite is highly disordered ((71<?-0) =

1.652, (71m-0) = 1.639, (TZo-O) = 1.642, (TZm-O) =

1 .647 À) and that thè electron-density distribution of thè

Na atom is much more strongly anisotropie than in low

albite.

Later, Ribbe et al. (1969) used thè Weissenberg in-

tensity data of Ferguson et al. (1958) to perform a 3D

least-squares refinement. The R value reduced to 0.082,

and (he new (T- O) distances in Table A2-b are in dose

agreement with those determined by X-ray counter

techniques in specimens of high albite at RT (Prewitt et

al., 1976: synthetic high albite: (Tìo-O) = 1.646, (Tlm-O) =

1.641, (T2o-0) = 1.641, (T2m-0) = 1.642 À; Winter et al.,

1979: Tiburon high albite: ( T\o-0 ) - 1.649, (71m-0) =

1.642, (T2o-0) = 1.640, (Tlm-O) = 1.642 À). The (T- O)

distances and thè derived Al contents in thè Amelia high

albite indicate that thè effect of heating produced strong,

but not complete, Si, Al disorder (see Table A2-b and

Fig. 2).

Ribbe et al. (1969) confirmed that thè degree of ani-

sotropy of thè Na atom is much stronger in high than in

low albite. The simplest interpretation of thè electron-den¬

sity distribution in terms of time-average of a reai

anisotropie displacement of a single Na atom was con-

sidered quite unrealistic in high albite. The single-atom

model was therefore discarded, and an interpretation in

terms of space-average for split-atom models, i.e. thè half-

atom and quarter-atom models, was explored.

In thè half-atom model, thè single anisotropie Na atom

is replaced by two isotropie half-atoms, separated by 0.61

A along thè direction of thè elongation of thè displace¬

ment ellipsoid in thè single-atom model.

In thè quarter-atom model, thè Na atom is replaced

by four isotropie quarter-atoms, which are nearly collin¬

ear along thè direction of thè maximum anisotropy in thè

single-atom model and symmetrical about thè mean po-

sition; thè two outermost and thè two innermost quarter-

o

atoms being separated by ~ 0.9 and ~ 0.2 A, respec-

tively.

The quarter-atom model was thè split-atom model

that best fitted thè experimental data. The evidence in

high albite of an average structure poses thè problem as

to thè Na atom anisotropy represents a random space-

average or a faulted domain-average. In thè latter case,

bipartite, quadripartite and multipartite unit cells with

faulted domains could be taken into account. In thè case

of high albite, Ribbe et al. (1969) concluded that thè

anisotropy of thè Na atom probably corresponds to a

space-average of a quadripartite-multipartite structure

with faulted domains.

More recent studies by Prewitt et al. (1976) on syn¬

thetic high albite confirmed thè validity of thè quarter-

atom model, but found no evidence of domain texture. On

thè other hand, Winter et al. (1979) found thè quarter-

atom model to be inadequate (see thè section on Amelia

monalbite).

High-temperature crystal structures of

PRE-HEATED NA-RICH FELDSPARS

Amelia monalbite. Duba & Piwinskii (1974) annealed

a fragment of an Amelia low albite at 1080 °C for 21 days,

at 1100 °C for 50 days, and at 1111 °C for 62 days. After

quenching, thè specimen displayed fine lamellar twinning

on thè Albite law. A precession study by Okamura &

Ghose (1975a) showed that thè pairs of dif- fraction spots

for Albite twinning merged to a single spot at 930 °C, in-

dicating that thè specimen inverted to monalbite.

Winter et al. (1979) re-investigated thè same Amelia

specimen as that studied by Okamura & Ghose (1975a).

The composition of thè Amelia albite reported on by

Winter et al. (1979) is Or06Ab993An01. Actually, this

composition was determined with an electron micro-

CRYSTAL STRUCTURES OF ALKALI FELDSPARS FROM GRANITIC PEGMATITES: A REVIEW

91

• 9

•••

tlo + tlm or 2t1

0.5 0.6 0.9 1.0

I - 1 _ I _ I _ I _ i _ 1 - 1 _ I _ L__J

Fig. 2 - Relation between sum and difference of Al contents in 71o

and Tlm sites, i.e. tlo + tlm and tlo - tlm, in specimens of Na-rich

feldspar from granitic pegmatites used in structure refinements.

The values of tlo + tlm (in thè case of triclinic simmetry) or 2tl

(in thè case of monoclinic simmetry) and tlo - tlm were obtained

from (T-O) distances following thè procedure developed by Blasi

& De Poi Blasi (1994, eqs. 18 and 19) and using thè reference val¬

ues for Al-O and Si-O given in thè relevant section of this work.

Room-temperature crystal structures of naturai specimens: 1 Ra¬

mona low albite (Ribbe et al., 1969); 2 and 3 Amelia low albite X-

ray and neutron data, respectively (Harlow & G.E. Brown, 1980);

4 Rabb Canyon high albite (Keefer & G.E. Brown, 1978). Room-

temperature crystal structures of pre-heated specimens : 5 Amelia

high albite (Ribbe et al., 1969). High temperature crystal structures

of pre-heated specimens: 6 and 7 Amelia monalbite at 980 and

1060 °C, respectively (Winter et al., 1979). Low-temperature crys¬

tal structures of naturai specimens: 8 Ramona low albite at 93 K

(Williams & Megaw, 1964); 9 Amelia low albite neutron data at 13

K (Smith et al., 1986).

probe by Prewitt et al. (1976) using another Amelia al¬

bite specimen, and is, surprisingly, thè same as that re-

ported on by Harlow & G.E. Brown (1980) for their

Amelia albite specimen (see thè section on Amelia low al¬

bite). For thè sake of comparison, reference should be

made to 9 Chemical analyses of Amelia albite reported by

Deer et al. (1975).

Winter et al. (1979) confirmed thè inversion tempera¬

ture at 930 °C by means of a single-crystal diffractometer

equipped with a Polaroid land film cassette. Next, they be-

gan with a collection of intensity data at 1060 °C, which

was discontinued after collecting 430 diffractions because

of failure of thè furnace. The results of thè structure re-

finement were published along with thè data from a struc¬

ture refinement already performed by Okamura & Ghose

(1975b) on thè same specimen at 980 °C. We contend that

thè structure refinement at 1060 °C should be disregarded

for thè following reasons: (1) thè number of diffractions

1 used in structure refinement is very small, (2) thè R factors

are comparable to those obtained for thè refinement at

980 °C, but thè individuai a values appear to be clearly bi-

ased, (3) thè value of thè grand mean T-O distance is ab-

normally low [«T-O» = 1.6369(18) À], (4) thè value of thè

displacement pprameter B for thè Na atom is exceedingly

high (B = 15.8 A2) when compared with thè corresponding

value in thè structure at 980 °C ( B = 12.0 À2), (5) thè val-

ue of (71-0) distance and that of tl are anomalous com¬

pared with thè corresponding values for refinement at 980

°C (Table A2-c), (6) thè value of tl + t2 is abnormally low

(Table A2-c).

Winter et al. (1979) reported that thè Amelia specimen

annealed by Duba & Piwinskii (1974) was unsuitable for

structure analysis at room temperature because of twin-

ning. In thè absence of data measured at RT, thè values of

lattice constants of thè Amelia specimen at 980 °C (Table

A2-c) might be compared with thè corresponding values

for thè ideal high albite (analbite) (see data in Table 5 of

Blasi & De Poi Blasi, 1994). At 980 °C, there is strong ex-

pansion of thè structure with an increase in celi edges a

(from ~ 8.159 A at RT to 8.274 À at 980 °C), b (from ~

12.872 À at RT to 12.991 À at 980 °C), and c (from ~ 7.109

À at RT to 7.144 À at 980 °C), and a decrease in thè celi

angles a (from ~ 93.52° at RT to 90.06° at 980 °C), (1 (from

~ 116.42° at RT to 116.13° at 980 °C), and y (from ~ 90.26°

at RT to 90.05° at 980 °C). The celi volume V increases

from ~ 666.9 À3 at RT to 689.4 À3 at 980 °C.

With reference to thè room-temperature position of

thè ideal high albite (analbite) in a diagram of b vs. c, thè

representative point for thè Amelia monalbite moves

roughly towards thè high-sanidine corner, in a direction

that is tangibly outside thè quadrilateral joining thè ref¬

erence points for thè alkali feldspar end-members and cor-

responds approximately to Trend I of Brown et al. (1984).

The structural indicators derived from thè A method yield

“impossible” Al contents (see Table A2-c). Note that in thè

Amelia low albite studied at 13 K by Smith et al. (1986),

thè A method gives a Si, Al distribution that is consistent

with that estimated from (7-0) distances (see thè relevant

section). In both specimens, thè variations in b and c celi

edges are coupled, with some deviation from thè ideal

coupling, which needs no change in A (bc) values in a bc

plot, in response to changes in temperature. The deviation

from thè ideal coupling is higher in thè Amelia monalbite,

and this explains its anomalous values of A (bc) and

A (b*c*) (see Table A2-c). Further studies are needed to

ascertain whether thè deviation of b and c celi edges from

thè ideal coupling in monalbite at high temperatures is

confirmed.

The N0r contents for thè Amelia monalbite in Table

A2-c are meaningless; they merely reflect thè strong ex-

pansion of celi edge a and celi volume V.

The deviations observed between thè values of thè (7-

O) distances (see Table A2-c) and thè value of thè grand

mean 7-0 distance [((7-0)) = 1.6419(5) À] indicate ap-

parent appreciable Si, Al ordering, which is estimated by

thè values of thè derived Al contents in Table A2-c. This

behavior corresponds to a two-step ordering trend in thè

Amelia monalbite (see Fig. 2). Note, however, that thè

swinging-arm effect at high temperatures influences ap-

preciably thè (7-0) distances and thè de-rived Al contents

92

ACHILLE BLASI & CARLA DE POL BLASI

(see thè section on reference values for Al-O and Si-O dis-

tances in alkali feldspars for information on thè swinging-

arm effect).

The Na atom is 9-fold coordinated and lies, along with

thè OA2 oxygen atom, in thè mirror piane. The remaining

oxygen atoms forni a distorted cube obeying thè mirror

piane. The displacement parameter B for thè Na atom in thè

Amelia monalbite follows thè trend observed in thè speci-

mens of high albite used in structure refinement at high

temperatures by Prewitt et al. (1976: synthetic high albite)

and Winter et al. (1979: Tiburon high albite). The latter

showed that Z?(Na) for all these specimens did not extrapo-

late to zero at 0 K, indicating that some static spatial disor-

der must occur. This was attributed by Winter et al. (1979)

to differences in thè locai Si,Al configurations produced by

Si,Al disorder. According to Winter et al. (1979), previous

interpretations restricting thè number of such configura¬

tions to four and assuming isotropie Na-quarter atoms

(Ribbe et al., 1969; Prewitt et al., 1976) are not adequate.

The specimens of high albite investigated by Prewitt et

al. (1976) and Winter et al. (1979) did not invert to monal¬

bite at high temperatures. The high albite studied by Pre¬

witt et al. (1976) is a synthetic specimen Or05Ab983An12

annealed at 1060 °C for 40 days.The high albite studied by

Winter et al. (1979) was obtained by annealing a meta-

morphic low albite Or025Ab9975 An0 from thè Tiburon

Peninsula, Marin County, California, U.S.A., at 1080 °C for

60 days.

Using relevant structural data at various temperatures,

Winter et al. (1979) extrapolated an inversion temperature

for thè Tiburon specimen in thè 1050-1075 °C range and

for thè synthetic high albite in thè 1105-1110 °C range. Ac¬

cording to Winter et al. (1979), thè different inversion tem¬

peratures for these specimens may depend on differences

in thè degree of Si,Al disorder induced by different an¬

nealing treatments.

Undoubtedly, thè Amelia monalbite was subjected to

stronger annealing treatment before X-ray examination at

high temperatures. Thus, thè specimen shows glassy rinds

indicating that partial melting occurred during annealing.

Unfortunately, thè structure of thè Amelia monalbite be¬

fore inversion is not available. It could be an analbite, which

would have rendered easier thè inversion to monalbite.

Low-temperature crystal structures of

NATURAL NA-RICH FELDSPARS

Ramona low albite at93 K. Williams & Megaw (1964)

made an X-ray 2D study of a Ramona low albite and a

synthetic high albite at 93 K. The low albite from thè Ra¬

mona pegmatite. San Diego County, California, U.S.A.,

was thè same as that used by Ferguson et al. (1958) and

Ribbe et al. (1969) in structure determination at room

temperature (see thè relevant section). The investigation

was undertaken to see whether thè anisotropy of thè Na

atom persisted down to a low temperature. The results

were inconclusive and thè data in Table A2-d should be

consif iered of historical value. Note, however, that thè val¬

ues /-( )> distances and Al contents derived therefrom

are no: u from thè expected ones (see Table A2-d and

Fig. 2).

Quareni & Taylor (1971) re-refined thè data collected

by Williams & Megaw (1964) for thè Ramona low albite at

93 K, and used them along with their 2D data at 293, 573,

and 873 K on a low albite from Schmirntal,Tyrol, Austria,

to perform extrapolations of thè displacement parameter

B versus temperature. They found that thè extrapolations

for Na, Al, and Si atoms gave values of B very dose to ze¬

ro at 0 K, whereas thè extrapolation for thè oxygen atoms

give B > 0 at 0 K. They concluded that thè marked

anisotropy of thè Na atom at room temperature is due to

anisotropie thermal vibration, not to positional disorder.

Later, thè data re-refined for thè Ramona albite were

incorporated by Winter et al. (1977) in their extrapolation