Cqnadian MineralogistVol. 17, pp.857-869 (1979)

PARAGENESIS OF SEBPENTINE ASSEMBLAGES IN HARZBURGITETECTONITE AND DUNITE CUMUTATE FROM THE OUEBEC APPALACHIANS

R. LAURENT EI.{N Y. HEBERTDCpaltement de glologie, Universiti Laval, Citi universitaire, Qulbec, Qtt6, GIK 7P4

Assrnecr

Two types of peridotite occur within the ophiolitesuites of the Quebec Appalachians: (1) a harzbur-gite tectonite of mantle origin, which forms thebasal unit of the ophiolite complexes, and (2)an overlying dunite of cumulate origin. Both typesare serpentinized, but only the first clearly showsubiguitous tectonic fabrics and bears rich asbestosore. The rocks are polymetamorphic, and serpenti-nization developed in two main episodes, earlyoceanic and late continental. The oceanic episodetook place under low activity of oxygen after theperidotites had cooled to below 350oC. This wasa pervasive serpentinization characterized by thepseudomorphic replacement of olivine and ortho-pyroxene by lizardite -r chrysotile -r brucite -rmagnetite -r awaruite, which ran to completion inthe cumulate peridotites but only partly affectedthe harzburgite tectonite. The extent of pervasiveserpentinization controlled the later development ofthe asbestos veins. When the ophiolites were em-placed by obduction, a relatively low degree ofserpentinization allowed the formation of fracturesin the harzburgite tectonite, whereas the high de-gree of serpentinization of the dunite cumulateinhibited its fracturing. The action of oxygen-richwaters started the growth of asbestos veins in dila-tion fractures throughout the harzburgite during thetectonic transport and emplacement of the ophi-olites. The typical paragenesis in asbestos veinsis chrysotile -r magnetite -+ brucite, which wasformed under low-greenschist-facies PJ conditionsas shown by the mineral composition of the Cam-brian metabasaltic country rock. In shear zones.platy and fibrous antigorite -r brucite developedat the expense of mesh lizardite and vein chrysotile.

SoMMATRE

Deux types de p6ridotite se rencontrent dans lecortage ophiolitique des Appalaches du Qu6bec:(1) une harzburgite tectonite provenant du manteauet formant I'unit€ de base des complexes ophioliti-ques; (2) surmontant celle-ci, une dunite cumulat.Ces deux pEridotites sont serpentinis6es, mais seulela premiire possbde une fabrique 6vidente de tec-tonite et contient les plus riches min6ralisationsd'amiante. Ces roches sont polym6tamorphiques.La serpentinisation s'y est d6veloppde au cours dedeux 6pisodes principaux, le premier oc6aoique etle second continental. L'6pisode oc6anique s'est

d6roul6 dans des conditions de faible activit6 d'oxy-gdne et aprds le refroidissement des p6ridotites )rune tempdrature inf6rieure A 350oC. Cette serpen-tinisation, de caractdre p6n6trant, est caract6ris6epar le remplacement pseudomorphique de I'olivineet de I'orthopyroxdne par lizardite -+ chrysotile '+brucite t magnitite + awaruite; compldte dansla dunite cumulat, elle n'est que partielle dans laharzburgite tectonite. Le degr6 de serpentinisationp6n6trante gouverne le d6veloppement ultdrieur desveines d'amiante. Au moment de Ia mise en placedes ophiolites par obduction, le faible degr6 deserpentinisation de la harzburgite tectonite y afavoris6 la formation d'un grand nombre de cassu-res, ce qui n'a pas pu se produire dans la dunitecumulat entierement serpentinis6e. L'action de solu-tions aqueuses riches en oxygEne a provoqu6 laformation des veines d'amiante dans les cassuresqui s'ouvraient dans la harzburgite pendant letransport et la mise en place tectoniques des ophio-lites. Dans les veines d'asbestos, la paragendse typi-qle, chrysotile t magnitite -r brucite, r6sulte desconditions de basses pression et temp6rature dud6but du facies schiste vert, comme I'indique lacomposition min6ralogique des roches encaissantesm6tabasaltiques d'6ge Cambrien, Dans les zones decisaillement, I'antigorite 6cailleuse ou fibreuse et labrucite rcmplacent la lizardite r6ticul6e ou le chry-sotile filonien.

INtnopucrrohr

Commercial mining of chrysotile from theserpentinized peridotites of the Qu6bec Appala-chians started a century ago, in 1878. Sincethen, the region has developed into one of theworld's major producers of asbestos fibre.Studies by Cirkel (1910), Dresser (1913), Gra-ham (1917), Cooke (1937), Faessler & Ba-dollet (1947), Riordon (1955, 1975) andAumento (1970) have led to many sontrover-sial hypotheses on the origin of the mineraliza-tion. These works, however, fail to answer sev-eral fundamental questions about (1) the mesh-anism of fibre growth, (2) the physical condi-tions that control the parageneses and (3) thetemporal and spatial framework of serpentiniza-tion. A better understanding of the origin of as-bestos deposits in the Qu6bec Appalachians re-quircs new knowledge about the petrogenesis and

857

858 THE cANADTAN

evolution of the host rocks. Structures, texturesand mineral assemblages in peridotites front theQu6bec Appalachians are complex because theyare the results of a long evolution. It is thepurpose of our paper to outline the majorstages of this evolution and to discuss texturalfeatures, mineral and rock compositions andphysicochemical conditions that bear on themineral parageneses observed. We stress whatwe consider to be the most significant featuresof the rocks studied; hence, countless detailswill be omitted in an attempt to give the readera clear picture of the main facts. Our account,however, is based on observations collectedduring seven field seasons of mapping in theophiolite belt of the Qu6bec Appalachians (Lau-rent 1973, 1975a, 1977, Laurent et al. L979a)and on detailed studies of the petrography ofthe serpentinized peridotites and associatedrocks. Our field and petrographic work wascarried out in collaboration with students whosetheses give extensive descriptions of the fieldrelationships, textures, mineralogy and chemis-try of the ophiolite rocks (Y. Hibert 1974,Lamothe 1978, Beullac t979, R. H6bert 1979,Rodrigue 1979, J. Beaudin in prep., Y. H6bertin prep.).

The chrysotile-asbestos-rich peridotites of theThetford Mines and Asbestos areas are part ofa narrow ophiolite belt that extends discontinu-ously through the Canadian Appalachians fromBaie Verte, Newfoundland, to southern Qu6bec(Church 1972, Laurent'].915a, Williams & Tal-kington 1977, Williams & St*Julien 1978). Theophiolite belt, in part much dismembered andrnetamorphosed, consists of peridotites, gabbrocumulates. diabase sills" tholeiitic pillow lavasand red argillites (Lamarche 1972, St-Julien1972, Laurent 7973, 1975a, 1977, Church1977); these units have been emplaced tectoni-callv within the Inner Zone of the Qu6becAppalachians in Early Ordovician time. Theophiolites are regarded as fragments of theoceanic crust and mantle underlying the Iapetus(Proto-Atlantic) Ocean, or smaller marginalinter-arc basins (Bird & Dewey 1970, Dewey1974). If the peridotites represent parts ofslices of upper mantle and oceanic crust, thefollowing sequence of events must be consi-dered: (1) the formation of the peridotites inan oceanic environment through processes ofoceanic lithosphere accretion and sea-floorspreading (2) the fragmentation and tectonicemplacement of the peridotites on the continent,and (3) deformation of the peridotites with theCambro-Ordovician country rocks. Serpentiniza-tion took place in two main episodes during this

MINERALOGIST

evolution. An early serpentinization event thatproduced massive lizardite is assumed to have?aken place in the oceanic environment, at lowtemperitures and at some remote distanse fromthe spreading axis. Late serpentinization occurredin a dynamic regime, at higher temperafi[esthan the first episode and contemporaneous -withthe tectonic transport and emplacement of therocks into their present setting. The second ser'pentinization produced foliated textures andieveral generations of chrysotile * brucite andantigorite * brucite veins. The peridotites arecut by rootless granitic dykes that by field rela'tionships can be shown to have been em-placed-between the two episodes of serpentiniza-iion. The dykes have yielded early OrdovicianK-Ar ages (Poole et al. 1963).

Pr,nmortrBs BeFone SrnprNttNzerroll

Two types of peridotite are distinguished:(1) a halzburgite tectonite of upper-mantleorigin that forms the base of the obductedophiolit.t and (2) an overlying dunite of cu-

mulate origin. The presence of strong tectonitefabrics in the hariburgite and their general

absence in the dunite cumulate show that thebasal harzburgite had been complexly deformedprior to forming the floor on which the cu-mulates were

- deposited. Ave'Lallement &

Carter (lg7}), Nicolas et al.. (1972) and Green& Radciiffe (1972) have shown that a tectonitefabric in peridotite is likely to result,from-anepisode of slow, extensive plastic deformationand recrystallization under deep crustal -9r. upper-mantle pt"ttu.e and temperature conditions-' It

is assuni6d that the ubiquitous tectonite fabricof the basal harzburgite in the Qu6bec Appala-chians was acquired under a spreading axis byplastic flow during solid-state emplacement intoihe developing oceanic lithosphere (Juteau eta | . 1 9 7 7 ) .

The harzburgite tectonite grades locally intodunite. The harzburgite is homogeneous, but inplaces it contains alternating bands of olivinepyroxenite and dunite from 1 to 30 cm thick'fhe banding is complexly folded and over-printed by the main tectonic foliation of theiock. The tectonic foliation is defined by thepreferential orientation of orthopyroxenes,ienses and stringers of spinels and mosaics ofrecrystallized olivine. The textures vary orrershort distances from granoblastic, with littlecataclasis and only minor recrystallization ofolivine, to mylonitic, with complete recrystalliza-tion of the olivine (Fie. 1). The most commontexture is blastomylonitic, with orthopyroxene

' t . n a m

phenoclasts I to 30 mm long set in a foliatedmicroblastic matrix of olivine grains of twodistinct sizes. The orthopyroxene phenoclastshave been bent, and contain deformation la-mellae parallel to [100] as well as kink bands.The large olivine grains, up to 5 mm in size,are strained and granulated. Mosaics of muchfiner grained polygonal and unstrained neoblastsof olivine, many with triple-point grain boun-daries, surround the large olivine grains, theorthopyroxene phenoclasts and the spinels (Lau-rent 1975b). In some occurrences the harz-burgite is so strongly anisotropic that its fabricis nematoblastic, with olivine grains and ortho-pyroxene phenoclasts both containing sharp de-formation lamellae that have length-to-widthratios as great as 8 and 12, respectively. Anexcellent example of this type of tectonite is thefresh harzburgitic dunite of the Lac du Diablearea at Mont Albert (MacGregor 1962). Inother occurrences in locally developed mylonite,all the olivine has been recrystallized. Similartextnral variations have been described fromperidotite xenoliths in alkali basalts (Mercier &Nicolas 1975) and in kimberlites (Boullier &

859

l"-.+;-*--;{' o ' 5 * * '

Nicolas 1975). We also observed thin veins oforthopyroxenite that cut across the banding andfoliation of the harzburgite, apparently cement-ing joint planes. This early jointing may havedeveloped during decompression following thosolid-state emplacement of this peridotite withinthe oceanic lithosphere. The joint systems werereactivated several times; this permitted thesuccessive emplacement of magmas of gabbroic,dioritic and granitic composition. Bodies ofgabbro and diorite, subsequently rodingitized,are consanguineous with the ophiolite assem-blage. They were emplaced, we believe, at ornear the spreading axis, whereas the graniticdykes were emplaced later, since they postdatethe early episode of oceanic serpentinization.The harzburgite tectonite is composed of 65 to95Vo olivine Foae-sz, 5 to 3570 orthopyroxeneEnro-r, with exsolution lamellae of clinopyrox-ene, and less than lVo chronian spinel of variablecomposition, particularly in CrrO' (35-55Vo)ene, and less than l%o chromian spinel of variableGregor & Smith 1963, Irvine &. Findlay 1972'Kacira 1971, Laurent L975b, Y. [I6bert, inprep.).

sERpENTTNE AsSEMBLAGES FRoM THE euEnsc AppALAcHrANs

Frc. l. (a) Blastomylonitic texture of fresh harzburgite tectonite with orthopyroxene phenoclast set in afoliated microblastic matrix of olivine grains. Crossed nicols. Sample 2-2776, Mont Albert. (b) Foliatedmicroblastic matrix of olivine grains of two distinct sizes. Large olivine grain in the centre is at extinction.Crossed nicols. Same sample as (a).

-tpq

t----:=*-tJ fYta

Frc. 2. (a) Lizardite pseudomorphic texture in chromite-rich dunite cumulate. Chromite contains inclu-sions of fresh olivine. Crossed nicols. Sample 75-26772. Provengal Hill, Black Lake. (b) Lizarditepseudomorphic texture in dunite cumulate. Natural light. Sample 70-26772. Provengal Hill, Black Lake.(c, d) Lizardite hourglass textures in dunite cumulate. Crossed nicols. Sample 2-28674, Caribou Lakqnear Black Lake.

F I

SERPENTINE ASSEMBLAGES FROM THE QUEBEC APPALACHIANS 861

The lower part of the cumulates overlyingthe harzburgite consists of alternating layersof dunite and chromite-rich dunite. Beds ofdunite several metres thick are fine grained andisomodal, whereas chromite-rich layers are thinand exhibit graded bedding. Textures of theserocksn where not obliterated by younger cata-clastic deformation, are everywhere allotriomor-phic (adcumulus). They are not foliated inspite of the fact that locally they are tightlyfolded around axes parallel to their bedding.Since cumulate textures were not obliterated"folding of these rocks must have oscurredplastically before the end of their crystalliza-tion, i.e., at relatively high temperatures and inthe presence of residual liquid. This suggeststhat during the sedimentation of the cumulatesthe magma chamber was tectonically active.Plastic folding was followed by formation offractures (now filled by veins of pyroxenite orgabbro), showing that tectonic activity continuedafter the cumulates had cooled enough to un-dergo brittle failure. Olivines in the cumulatedunite are completely replaced by lizardite withmesh (Figs. 2a & b) and hourglass (Figs. 2c &d) textures, although fresh olivine (Foel-*) ispreserved as inclusions in poikilitic chromite(Kacira 197I, Y. H6bert, in prep.). Comparedwith the chromian spinels from the harzburgitetectonite, cumulate chromite shows a narrowercompositional range but iso on the average,slightly richer in CrzOr (48-60%) and poorer inAlrOa (7-16Vo). Olivirc and chromite of thecumulate dunites are interpreted as the earliestfractionation produc8 of the differentiation of aparental magma of picritic composition (Laurentet aI. L979b).

OcnlNrc SnnprNrntzertoN

The first episode of serpentinization was per-vasive; it was characterized by the replacementof olivine and orthopyroxene by lizardite. Per-vasive serpentinization ran to completion in thedunite cumulate but affected the harzburgitetectonite only in part: serpentinized rocks areirregularly distributed and rocks that range fromfresh to wholly serpentinized can be observed. Afirst episode of rodingitization is related to thisepisode of serpentinization. The early gabbroicintrusive rocks, and to a lesser extent, the dio-rites, have been converted to rodingitic assem-blages of clinozoisite, diopside and hydrogros-sular (Olsen 196L, De 1972). Minerals of thisrodingitic assemblage are overprinted by cata-clastic defoimation that predates the widespreaddevelopment of reddish-brown biotite in the in-trusive rocks ('Olsen 196I).

Pervasive serpentinization of olivine in thedunite cumulate produced lizardite pseudo-morphs in mesh and hourglass textures witheither isotropic or hourglass lizardite in the meshcentres (Fig. 2). Concentrates of magnetiteformed during the process occur at serpentinegrain-boundaries. These textures are illustratedby Wicks et al, (1977, Figs, lc, ld, 3b) and cor-respond to their cases L and 3. Here, serpenti-nization advances at right angles to grain boun-daries, either wholly replacing olivine at auniform rate and constant volume or leavinga central core of unaltered olivine which is re-placed by randomly oriented serpentine in alater episode.

On the other hand, pervasive serpentinizationof the harzburgite tectonite produced a suite ofrocks with a great variety of textures owing tovariations in the pre-existing foliated fabrics andto a system of early fractures. Preferential pseu-domorphic growth of lizirdite proceeded alongthe foliation and on fractures so that the resultingserpentinization is anisotropic (Fig. 3). Weobserved a lepidoblastic matrix of apparentlyfibrous lizardite enclosing subunits of olivine and

Frc. 3. Lepidoblastic texture of serpentinized harz-burgite tectonite. Natural light. Sample G-16873,Quany Hill, Black Lake.

862 THE CANADIAN MINERALOGIST

TABIT 1. EI,ECTION-MICROPROBE AIIAIiSES OF LIZAil)ITB TROt.l TBETIORD MIJ{ES

v t . Z ' A 3 c D E f c u r J

sto2 4L.o2 4L.02 41.70 40.83 38.92 39.06 4r.66 4O.2r 40.68 41.53 40.67 4L.64

A12O3 0 .00 0 .08 0 .04 0 .07 0 .06 0 .04 0 .02 0 .00 0 .00 0 .06 0 .00 1 .48

Fdro t 2 .73 2 .56 1 .84 1 .92 3 .03 2 .80 2 .39 2 .58 3 .15 2 .50 2 .62 3 .01

lteo 4L.97 42.L5 42.39 42.99 43.88 43.99 4L38 42,99 42.67 47.67 43.19 39.37

C"2O3 0 .78 0 .57 0 .49 0 .69 0 .s1 0 .62 O,29 0 .17 0 .00 0 ,07 0 .02 O.94

N10 0 .00 0 .02 0 .04 0 .00 0 .10 0 .04 0 ,76 0 .55 0 .00 0 .67 0 .00 /cdo 0 .06

AtoElc ploportlotra

s l .6979 .6840 .72L4 .70A2 .6078 .6163 .67L9 .6287 .6526 .6843 .6776 .67

otl

r " l, t [ -

, t lNll

.0 .0030 .0014 .0030.0022 .0014

.0383 .0356 .0267 .0278.0396 .0369

x 1 .0489 1 .0501 1 .0929 1 .1114 1 ,0213 1 .0345

.0208 .0150 .0134 .0188.0126 .0154

.0 .0003 .0005 .0 .0014 .0005

.0008 .0 .0 ,oo22 .0 .05

.0322 .0337 ,0423 .0344 .0365 .04

.9946 1.0007 7,0204 r.O234 L.0723 .95

.0074 .0042 .0 .0020 .0004 .o2

.0099 .0070 .0 .0089 .O / c^ .00

1 . 1 0 8 0 1 . 1 0 2 5 1 . 1 3 4 9 1 . 1 6 1 1 . 0 7 7 1 1 . 0 8 8 7 1 , . M 4 9 r . 0 4 5 6 1 . 0 6 2 7 1 . 0 7 0 9 1 . 1 0 9 2 1 . 0 8

r . 5 9 1 . 6 1 r . 5 7 1 " . 6 4 1 " . 7 7 r . 7 7 1 . 5 6 1 . 6 6 1 . 6 3 1 . 5 6 r . 6 4 1 . 5 9

! g _ _ l n ^ 9 6 . 5 9 6 . 9 9 7 . 6 9 7 , 6 9 6 , 3 9 6 . 6 9 6 , 9 9 6 . 7 9 6 . 0 9 6 . 7 9 6 . 7 j 5 . 3UqrFe_- '

* RecaLculated to 86.52 aasmidg L3.52 E2O conteot of serpeatltre. (A to J) psod@rph of ol1vl.ne

fu dblte c@ulate; 0() pseud@olph of oLivla€ 1n harztnrrglte tectonttei (L) bastXte pseudo-

Botph of orlhoproaeoe 1tr harzburglte tectonlte.

orthopyroxene that are about 1 cm in diameter.The texture of this matrix resulted from replace-ment concomitant with expansion. Evidence ofexpansion is afforded by lepidoblastic lizarditewhich fills dilation crystal fractures in spinels(Laurent I975b, Fig. 4b). Orthopyroxene andolivine in the subunits are replaced in part bypseudomorphic lizardite. The orthopyroxene isaltered to bastite along grain boundaries, frac-tures and cleavages (Wicks & Whittaker 1977,Figs. 2b, 2c), whereas olivine has lizarditemesh-rims that surround fresh relics in meshcentres (Wicks et al. L977, Fig. 1a).

Microprobe analyses of lizardite pseudo-morphs after olivine (A to J) in dunite and of liz-ardite after olivine (K) and orthopyroxene (L) inharzburgite are given in Table 1. The analyseswere done with an ARL electron microprobeat 20 kV and 1O mA using natural olivine anddiopside standards provided by the GeophysicalLaboratory, Carnegie Institution of Washington.The results given are average values of a mini-mum of three selected spots in each lizarditeanalyzed; identification of the samples was madeindependently by X-ray diffraction using Guinierand Gandolfi cameras. Because mesh pseudo-morphs are multiphase assemblages, the results

obtained show chemical variations which, inpart, may reflect the presence of ( I ) micro-scopic grains of magnetite or awaruite and (2)chrysotile and brucite intergrown with lizardite.Lizardite pseudomorphs have a higher Mg/Mg* Fe ratio than their parent olivine and ortho-pyroxene. The iron liberated by serpentinizationhas formed magnetite and minor amounts ofawaruite. Some iron has also entered brucitethat is intimately intergrown with lizardite; notethat we have found no evidence of widespreadformation of brucite during the first episode ofserpentinization. Ideally, lizardite contains aboutl3.5Ve H,O by weight and is enriched in SiOr,or impoverished in MgO, with respect to olivine.With few exceptions, serpentinization texturesprovide strong evidence that the reactions run atconstant volume (Thayer L966, L967) or thatany increase in volume is minimal. Hence, it ismore likely that magnesium was removed insolution than that silicon was added. Our dataalso suggest that the serpentinization of ortho-pyroxene in harzburgite released large amountsof silicon and smaller quantities of iron, alumi-nurn and calcium. The following equation prob-ably is a vali.d approximation of the major ex-changes that took place during the first episode

SERPENTINE ASSEMBLAGES FROM THE QUEBEC APPALACHIANS 863

of serpentinization (modified from ColemanI97l): olivine f orthopyroxene * HzO -+ ser-pentine -r brucite -+ magnetite t awaruite :texcess of Si, Al, Fe, Ca, Mg (removed).

The role of minor elements is complex. Liz-ardite after olivine is richer in chromium in thedunite cumulate than in the harzburgite tecto-nite. Also, chromium (as well as aluminum andcalcium, Table 1) is higher in lizardite (bastite)after orthopyroxene than in lizardite after olivine(Wicks & Plant 1979). Nickel contents of thelizardite in tectonite peridotites define a rangefrom ,0.02 to 0.2AVo NiO (Nickel 1971, Table1), which is narrower than the range of 0.O toO.76Vo found in the lizardite pseudomorphs afterolivine from the dunite cumulate (Table 1).

The trends of both major and minor elementsmobilized during the first episode of serpenti-nization can be ascertained. Aside from the ironand nickel retained in secondary magnetite andawaruite, calcium, magnesium and aluminumwere, at least in part, taken up by the rodingiti-zation of gabbro and diorite spatially associatedwith the serpentinized rocks (Cooke 1937,Faessler & Badollet 1947, De t9'12). In an ex-cellent review of the literature, Moody (1976b)pointed out that serpentinization can be eitherisochemical or allochemical, and that allochem-ical serpentinization (such as the first episodeof serpentinization in the Qu6bec Appalachians)implies the presence of an open system with afree exchange between circulating waters and therocks. Studies of oceanic rocks provide ampleevidence that seawater and oceanic crust in-teract and exchange chemical components (Bis-choff & Dickson 1975, Hajash 1975, Wolery &Sleep 1976, Bloch & Hofmann 1978). Conse-quently, serpentinization in the oceanic environ-ment takes place under the conditions of anopen system.

Oceanic serpentinization is mostly of thelizardite-chrysotile type. Experimental and anal-ytical evidence suggests that this paragenesisforms at temperatures below 350'C (Evans1977, p. 434). Alteration of peridotites in theQu6bec Appalachians during the early phase ofserpentinization has produced lizardite t mag-netite, an assemblage that is common in oceanicserpentinites.

Conditions of low oxygen-fugacitibs accom-panying the oxidation reactions during the firstepisode of serpentinization are indicated by thepresence of awaruite (Ni'Fe) and nickel sulfidessuch as heazlewoodite and millerite (NickelL959, L97t, Chamberlain 1966), and by tracesof N-alkanes with molecular weights betweenCre and Gu (Gibbs 197I). On the basis of

a detailed study of opaque minerals in serpen-tinites, Eckstrand (1975) has suggested thatserpentinization may generate reducing condi-tions. Calculations by Moody (L976a) based onthermodynamic data confirm that the as$em-blages iron-magnetite and magnetite-awaruitedefine a very oxygen-deficient but hydrogen-rich environment.

ffhe presence of awaruite and other oxygen-poor compounds has led several researchers todescribe this type of serpentinization as takingplace in a reducing environment. In fact, it is acomplex oxidation and reduction process inwhich 55 to SOVo of the iron in the rock isoxidized at the expense of water, which is re-duced to gaseous hydrogen. The hydrogen inturn is responsible for producing awaruite andrelated minerals by reduction. However, awaruiterarely occurs in amounts greater than lVo andis often absent. Magnetite occurs in amounts ofup to at least 1,O/o and is clearly the dominantiron-bearing phase. Thus, classifying the processas reducing puts too much emphasis on a minor,but important, phase. This has been discussed indetail in Wicks & Whittaker (1977):, Ed. note.f

The cumulate dunites are wholly serpenti-nized, whereas the underlying harzburgite tec-tonite is not. This spatial distribution of theintensity of serpentinlzation implies that the-source of water was located above rather thanbelow the rocks and, therefore, that seawaterwas probably responsible for the first episode ofserpentinization.

Eight different processes of serpentinizationhave been distinguished by Wicks & Whittaker(1977) on the basis of relative differences intemperature, regime of shearing and antigoritenucleation. The first episode of serpentinizationcorresponds to their process 3, characterized bythe development of lizardite pseudpmorphic tex-tures in the absence of shearing and underfalling or constant temperature conditions. Thespreading axis is a dynamic, high-temperatureenvironment that does not satisfy the abovecriteria. Furthermore, thermal models mitigateagainst widespread serpentinization of layer 3at or near oceanic spreading axes (Fowler1976). We are led to assume that the earlyepisode of serpentinization occurred in theoceanic environment but at a distance removedfrom the spreading axis. Only after havingcooled to below 35'0"C were the peridotitesserpentinized under static conditions.

INrnusrw RocKs

Three groups of rootless intrusive rocks, em-

864 THE CANADIAN MINERALOGIST

Frc. 4. (a) Quartz monzonite dyke in serpentinized harzburgite tectonite, Jeffrey mine, Asbestos. (c)Lens-like body of strongly deformed and altered hornblende diorite in serpentinired harzburgite teto-nite, Jeffrey mine, Asbestos. (b) and (d) Asbestos veins composed of chrysotile cross-fibres, in serpenti-nized harzburgite t€ctonite from Thetford Mines.

placed chiefly in the harzburgite tectonite, canbe distinguished on t}le basis of differences incomposition, alteration and deformation (Lau-rent 1975a, b). The oldest group consists ofisolated lenses of massive rodingite a few metresin length. The rodingite preserves relics ofgabbroic texture, but all primary phases havebeen replaced by calc-silicate minerals (Olsen1961). The second group is made up of ir-regular bodies, up to 100 m thick, of strongly

deformed, foliated, altered and partly rodingi-tized hornblende diorite (Fig. 4c). The primarymineral assemblage of the diorite is andesine,poikilitic hornblende, sphene, apatite and zircon(De 1,972). These first two groups are con-sanguineous with the ophiolite assemblage andwere emplaced in the peridotites prior to thefirst episode of serpentinization. They wentthrough several phases of alteration includingtwo, distinct episodes of rodingitization (Table

IAEIA 2. ST'UXAn| OP SVENTS AUD PASAGENESES

865

rial penetrates the serpentinite to distances thatvary from a few millimetres to several centi-metres, and xenoliths of serpentinite displayinga reaction rim of serpentine-tab-anthophylliteare found within the intrusive rock. These ob-servations provide evidence that the youngestdykes were emplaced in peridotites that al-ready had been serpentinized. The presence ofantigorite within the serpentinite at the contactwith these dykes was reported by Olsen (1961).De (1972) described thermal aureoles 1O to 50cm wide in which lizardite has been dehydratedto antigorite-chlorite. No recrystallized olivinehas been observed. because either this mineralnever formed or it was replaced subsequentlyduring the late episode of serpentinization.Although we cannot yet confirm the identifica-tion of antigorite, the work of Olsen and Desuggests that the most widespread and charac-teristic reaction in the aureoles was the re-placement of lizardite by antigorite. Studies ofprogressively metamorphosed serpentinites com-bined with a theoretical analysis of the serpen-tinite multisystem by Evans et al. (L976) indr-cate that antigorite is the high-temperatureserpentine phase. It replaces lizardite as a resultof proglade metamorphism at temperatures cor-responding to the onset of the greenschist facies(300-350"C), and is stable to the onset of theamphibolite facies (Evans 1977). Serpentinitewithin contact aureoles of the dykes is cut bychrysotile fibres formed in the late episode ofserpentinization (see below); therefore, the mon-zonitic dykes were emplaced between the twoepisodes of serpentinization. The only age deter-

SERpENTINE AssEMBLAcEs FRoM THE eu6src AppALAcHIANs

oDbloutefo@t16

Dtles ofgabbro ed dlorlte

Paeud@rphlc replace@t -of prl@ry Elu€raLgl.lwAttu t arvvsottl,e aZl.@sol.ai.te +t bnbLte ! nvgretita dtopolde + lqd"6@4 @@ita gzoaauLwite

Elp1ac@nl of quartz ll@odte dyke€ edalevelog@t of blotlte

'(K @bg@tld)

forutLd of the chrysotlLFasbestos-rXch orea) f,oLlated tdtures velnB of aoisLte"

Ped.dotitoe

flrBt @turplrl,aBtoc€el.c epiedoof aBtl.c ae4,e!-t!.o&atl.oB

DEf,o@tL@

gec@al @t@rphlaolc@tloetrtal eplaodoof dynaDicEerp6tLnl.zatl@

Iata aLtalatioa(poat-euplac@!t)

,lttt @brgorite + gtoaauLette ,bwtta + @grettte ptel@ttu, Atapai.da,

veeurl@lte, eta,b) sbeatoa vels rlthahtlsottle + bwlte+ nag@tita

tala + ahlot ttee + de"bwtea + q@ta

2). The first produced clinozoisite, diopside andhydrogrossular as pseudomorphs after the prim-ary minerals, whereas the second episode ofrodingitization produced veins of such calc-sili-sate minerals as zoisite, grossular, prehnite, diop-side, vesuvianite, wollastonite and margarite(Olsen L96l). The two episodes were separatedby a period of deformation that was accom-panied by the emplacement of the youngest in-trusive rocks and a widespread development ofbiotite through potassium metasomatism.

The third group of intrusive rocks consistsof relatively undeformed and unfoliated dykesof quartz monzonite I to l0 m thick (Fig. 4a)with potassium feldspar, oligoclase, quartzo bio-tite and mussovite as the main constituents (De1972). The older, foliated intrusive rocks havesharp borndaries, whereas the youngest dykesshow progressiveo uneven sontacts with the en-closing serpentinites. Quartz monzonitic mate-

---E* -

Fto. 5. (a) Stockwork of fractures filled with chrysotile ooss-fibres in harzburgite tectonite, Vimy Ridge,Thetford Mines. (b) Deformed and recrystallized(?) antigorite cross-fibres filling a fracture in serpenti-nized harzburgite tectonite, Jeffrey mine, Asbestos.

866 TI{E CANADIAN MINERALOGIST

minations on the monzonite are two early Or-dovician muscovite K-Ar ages, 477 and 481m.y. (Poole et al. 1963); these, we believe, cor-respond approximately to the time of tectonicemplacement of the ophiolites. We interpret thedyke intrusion as having occurred during frag-mentation of the oceanic lithosphere immedi-ately prior to, or concurrent with, obduction ofthe ophiolites onto the continent.

CoNtrNnNtlr- SsnpENrtNrzATroN

The late episode of serpentinization producedthe commercial chrysotile fibres that fill a stock-work of expansion fractures from O.5 to 2,5cm thick, mainly within the harzburgite tectonite(Fig. 5a). The late serpentinization was con-trolled by several systems of fractures, and inmany cases only narrow zones of wall rockalong the fractures were deeply serpentinizedfor the second time. Selvages along margins ofthe chrysotile-asbestos veins (Fig. 4b) consistof cryptocrystalline banded fibrous lizarditethat resulted from colloidal deposition; this inturn grades towards the wall rock into coarsergrained lizardite associated with chrysotile andbrucite. Chrysotile forms cross-fibres in thecentres of the veinso with magnetite concentratedalong the contact between fibres and vein sel-vages and, in smaller amounts, with chrysotileand brucite in the veins (Cooke 1937, Riordon1955, 1975, Riordon & Lalibert6 1972, Laurent1975b, Fig. 9). To explain the mechanism offibre growth, Laurent (L975b) applied theprinciples of incremental strain analysis devel-oped by Durney & Ramsay (L973). Tensile andshear fractures commonly are filled synchron-ously (as they progressively open) with crystalsof fibrous habit derived by solution or diffusionof material from the wall rock. The orientationof the growing fibres in the veins is controllednot by the position of the wall rock but by thedirection of minimum shear stress; expansioncauses fibre growth, and changes in the orienta-tion of the minimum shear stress are expected toproduce changes in the direction of fibte growth.Such changes are recorded by kinked fibres.Different generations of veins exhibit cross-cutting relationships. Chrysotile fibres in earlyveins are bent, kinked, broken and often re-placed by fibrous antigorite or brucite or both(Fig. 5b), whereas late veins are not de-formed (Fig. ad). From shear zones in theJeffrey mine at Asbestos, we have collectedserpentine vein samples of strongly deformedfibrous antigorite up to 20 cm long, cut by unde-formed and unaltered veins of chrvsotile fibres

1 cm long. These observations suggest that"chrysotilization" is syntectonic and that it devel-oped in a dynamic regime. The veins cementfractures caused by tectonic movements thatoccurred at a high level in the crust. Regionalmetamorphism at this level did not exceed low-greenschist facies, as shown by the mineralcomposition of pillow lavas from the CambrianCaldwell Formation adacent to the peridotites(S6guin & Laurent 1978).

Although chrysotile pseudomorphs after oli-vine are commonly observed in the harzburgite(Wicks & Whittaker 1977, Fig. 8c), the ob-served relations of wall rock to asbestos veinssuggest that vein chrysotile was derived fromlizardite rather than from fresh olivine. Duringthe processes of solution, transport and chryso'tile growth, most of the iron accommodated inlizardite was expelled to form additional mag-netite. The observed paragenesis is the follow'ing: lizardite * HrO * Or'+ clinochrysotile *magnetite -+ brucite -f excess Si (removed).

The chemical data of Hahn-Weinheimer &

Frc. 6. Foliated texture of serpentinized dunitecumulate, showing replacement of lizardite olthe early episode of serpentinization by anti-gorite of the late episode of serpentinization.Crossed nicols. Sample 1a-28578' Red Hills,Thetford Mines.

SERPENTINE ASSEMBLAGES FROM THE QUEBEC APPALACIIIANS 867

Hirner (1975) show that the chrysotile is charac-terized by a higher Mg/Mg.* Fe ratio than thecoexisting lizardite and that it has a compositionnear that of ideal serpentine.

During the same episode of serpentiniz.ation,antigorite -+ brusite developed extensively inshear zones within the harzburgite tectonite anddunite cumulate. Nonpseudomorphic platy anti-gorite replaced lizardite, producing foliated tex-tures (Fig. 6) and locally forming antigoriteschists.

The formation of chrysotile veins correspondsto serpentinization process 5 of Wicks & Whit-taker (1,977), defined by the recrystallization ofpseudomorphic lizardite to form nonpseudomor-phic chrysotile i brucite, and process 7, to formnonpseudomorphic antigorite '+ brucite under aregime of rising temperature. The occurrence ofvein antigorite has been reported from MontAdstock near Thetford Mines (Wicks & Whit-taker 1977, Table 3) and from the BritishCanadian mine at Black Lake (Hahn-Wein-heimer & Hirner 1975, Table 1), suggesting thattemperatures had risen above the stability fieldof lizardite and chrysotile in major shear zones.However, temperature alone does not controlthe development of chrysotile at the expeirse oflizardite, since lizardite and chrysotile common-ly occur together in serpentinites. It is knownthat lizardite and chrysotile, because of theirdifferent crystal structures, have distinct pro-perties of nucleation. Lizardite usually grows onan olivine or pyroxene substrate but does notnucleate easily, whereas in synthetic systemschrysotile is favored by crystallization from gelsand oxides (Dungan 1,977). Lizardite was dis-solved along tensile and shear fractures and wassynchronously reprecipitated as chrysotile fibres.This chemical process and the variations of thephysical parameters defining the stress appliedto the peridotite bodies, as well as favorabletemperatures ranging between 150 and 350oC,seem to have been the key factors that promotedthe delelopment of chrysotile veins. Antigoriteformed in the same regime but at higher temper-atures, in shear zones where we can assume thatfrictional energy was converted to heat.

Cumulate dunites do not display extensivearrays of asbestos veins. More often, chrysotileformed (with some magnetite and brucite) dur-ing the late serpentinization in sparse patchesdisseminated through the rock. Fracturing mayhave been inhibited in this peridotite, as itwas wholly serpentinized prior to tectonic em-placement; consequently, the rock did not un-dergo brittle deformation during the late episodeof serpentinization.,

After emplacement in their present setting,the serpentinized peridotites were further al-tered at ambient temperatures by COz- and SiOrbearing groundwater. Serpentine minerals werealtered locally to talc and magnesite; extensivealteration occurred along faults between theperidotites and the country rocks, where broadzones of talc-carbonate-chlorite rock were de-veloped.

ACKNOWLEDGMENTS

We thank T. Feininger, J. Moody and F.J.

Wicks for reviewing this manuscript and forproviding many helpful and informative sug-gestions. This study was supported by grant

A8293 from the Natural Sciences and Engineer-ing Research Council and by a joint grant from

the Direction G6n6rale de l'Enseignement Sup6-

rieur, Qu6bec.

RersnnNcss

AurrarNro, F. (1970): Serpentine mineralogy ofultrabasic intrusions in Canada and on the Mid-Atlantic Ridge. Geol. Surv. Can. Pap. 69'53.

Avr'Letr.nrvrnNr, H.G. & Cenren, N.L. (1970):Syntectonic recrystallization of olivine and modesof flow in the upper mantle. GeoI. Soc. Amer.Bull. 8L, 2203-2220.

BEULLAC, R. (1979): Etude Pdtologiqae des Ro'ches Ophiolitiques de la Rdgion du Lac Nicolet,Qu6bec. M.Sc. thesis, Univ. Laval, Qu6bec, Qu6.

Bnp, J.M. & Drwrv, J.F. (1970): Lithosphereplate - continental margin tectonics and theevolution of the Appalachian orogen. Geol. Soc.Amer. Bull. 81, 1031-1060.

Brscuorr, J.L. & DrcrsoN, F.W. (1975): Seawater-basalt interactions at 200oC and 500 bars: im-plications for origin of sea-floor heavy-metal de-posits and regulation of seawater chemistry.Earth Planet. Sci, Lett. 26,385-397.

Brocr, S. & HorueNN, A.W. (1978): Magnesiummetasomatism during hydrothermal alteration ofnew oceanic crust. Geology 6,275-277,

BourLLrER, A.M. & Nrcor-es, A. (1975): Classifl-cation of textures and fabrics of peridotite xeno-liths from South African kimberlites. Phys. Chem,Earth 9, 467-475.

CneMsrnr,erN, I.A. ( 1966) : Heazlewoodite andawaruite in serpentinites of the Eastern Town-ships, Quebec. Can. Mineral, 8, 519-522,

CHURCH, W.R. (1972): Ophiolite: its definition,origin as oceanic crust, and mode of emplace-ment in orogenic belts, q/ith special reference tothe Appalachians. In The Ancient OceanicLithosphere, Earth Phys. Branch Publ,, Dep.Energy, Mines and Resoarces. Ottawa, 42(3),7t-85.

868 THE CANADIAN MINERALOGIST

(1977)z The ophiolites of southern Quebec:oceanic crust of Betts Cove typ,. Can. l. EarthSci. L4, 1668-1673.

Ctnxsr,, F. ( 1910) : Chrysotile asbestos: its oc-surrence, exploitation, milling aod uses. Catt.Mines Br. Publ. 69.

CoLEMAN, R.G. (1971): Petrologic and geophysicalnature of serpentinites. Geol. Soc. Amer. Bull,82, 897-918.

CooKr, H.C. (1937): Thetford, Disraeli, and east-ern half of Warwick map areas, Quebec. Geol.Surv. Can. Mem. 2L1.

DE, A, (1972): Petrology of dikes emplaced in theultramafic rocks of southeastern Quebec andorigin of the rodingite. Geol. Soc, Amer, Mem.132, 489-501.

Dswey, J.F. (1974): Continental margins andophiolite obduction: Appalachian Caledoniansystem. In The Geology of Continental Margins(C.A. Burk & C.L. Drake, eds.), Springer-Verlag,New York.

Dnrssrn, JA. (1913): Preliminary report on theserpentine and associated rocks of southern Que-bec. Geol. Surv. Can. Mem. 22,

DUNGAN, M.A. (1977): Metastability in serpen-tine--olivine equilibria. Amer. Mineral. 62, 1018-t029.

DunNnv, D.W. & ReMsey, J.G. (1973): Incre-mental strains measured by syntectonic crystalgrowths. In Graity and Tectonics (K.A. DeJong& R. Scholten, eds.), J. Wiley & Sons, New York.

Ecrsrnenu, O.R. (1975): The Dumont serllentinite:a model for control of nickeliferous opaque min-eral assemblages by alteration reactions in ultra-mafic rocks. Econ. Geol. 70, 183-201.

EvaNs, B.W. (1977): Metamorphism of alpineperidotite and serpentinite. Ann. Rev. EarthPlanet. Sci. 5, 397-447.

----, JoEANNES, W., OrsRDoou, H. & TRourvrs-DoRFF, V. (1976): Stability of chrysotile andantigorite in the serpentinite multisystem.Schweiz. Mineral. Petrog. Mitt. 56, 79-93.

FAESSLER, C. & Beoonrr, M.S. (1947): Theepigenesis of the minerals and rocks of theserpentine belt, Eastern Townships, Quebec. Can.Mining J. 68, 157-167.

FowLEn, C.M.R. (1976): Crustal structure of theMid-Atlantic ridge crest at 37oN. Geophys. l. R.Astr. Soc, 47, 459-49L.

Grnns, G.W. (1971): The organic geochemistry ofchrysotile asbestos from the Eastern Townships,Quebec. Geochim. Cosmochim. Acta 85, 485-502.

Gneuau, R.P.D. (1917): Origin of massive ser-pentine and chrysotile asbestos, Black Lake-Thetford area, Quebec. Econ. Geol. L2, 154-202.

GnsrN, H.W. U & Reocunrq S.V. (1972): Dislo-cation mechanisms in olivine and flow in theupper mantle. Earth Planet. Sci, Lett. t5, 239-247.

HenN-WerNHErMER, P. & Hrnxnn, A. (1975):Major and trace elements in Canadian asb€stosore bodies. Analytica'l results and statistical inter'pretation, Third Int. Conf. Phys. Chem. AsbestosMinerals, Univ. Laval, Qudbec, Paper 6,25.

He;a$r, A. (1975): Hydrothermal processes alongmid-ocean ridges: an experimental investigation.Contr. Mineral. Petrology 58, 205-226,

HisnRT, R. (1979) ; Etude Pdtologique des RochesOphiolitiques dAsbestos et du Mottt Ham (Ham'Sad), Qudbec. M.ft. thesis, Univ. Laval, Qu6bec'Qu6.

H6renr, Y. (1974)z Etude Pdtrologique et Chi-mique de Ia Coupe du I'ac de lEst dans le Com'plexe Ophiolitique de Thetford Mines, Qu6bec.M.Sc. thesis, Univ. Laval, Qu6bec, Qu6.

InvrNE, T.N. & FrNor.lv, T,C. (1972): Alpine-typeperidotite with particular reference to the Bay ofIslands igneous complex, In The Ancient OceanicLithosphere. Eorth Phys. Branch Publ,' Dep,Energy, Mines and Resources, Otnwa, 42(3)'97-t28.

Jumnu, T., Nrcoles, A., DusEssY, J., FnucueRu,J.C. & Boucnrz, l.L. (L977)t Structural relation-ships in the Antalya ophiolitic complex, Turkey:possible model for an oceanic ridge. Geol. Soc.Amer. Bull, 88, 174GL748,

Kecrne, N. (1971): Geology of Chromitite Occur-rences and Ultramalic Rocks ol the ThetfordMines - Disraeli Area, Qudbec. Ph.D. thesis,Univ. Western Ontario, London, Ont.

LAMARcHE, R.Y. (1972): Ophiolites of southernQuebec. In The Ancierlt Oceanic Il.ithosphere.Earth Phys. Branch Pabl., Dep. Energy, Minesand Resources, Ottawo, 42(3), 65-69,

Lerrtornn, D. (1978) : Analyse Structurale du M6-lange Ophiolitiqae du Lac Montioie, M.Sc. thesis,Univ. Laval, Qu6bec, Qu6.

LeunrNl R. (1973): The Thetford-Mines ophio-lite, Paleozoic "flake" of oceanic lithosphere inthe Northero Appalachians of Quebec. Geol, Soc.Amer. Program Abstr.5, 188.

(L975a): Occurrences and origin of tbeophiolites of southern Quebec, Northern Appala-chians. Can, l. Earth Sci. t2, 443-455.

(1975b): Petrology of the alpine-tytlc ser-pentinites of Asbestos and Thetford Mines, Que-bcr,. Schweiz. Mineral. Petrog. Mitt, 55, 43L'455'

(1977): Ophiolites of tle Quebec Appala-chians. In North American Ophiolites (R.G.Coleman & W.P. Irwin, eds.), Oregon Dep. Geol.Mineral Ind, Bull. 95,25-40,

-, BEur-l-Ac, R., H6rnnr, R., Hfsnnr, Y. &RooNcuE, G. (1979a): Geology of the QuebecAppalachian ophiolites, and the Brompton litrea-ment. Geol. Assoc. Can,/Mineral. Assoc. Can.Program Abstr. 4, 63,

DELALoYE, M., VuecNlt, M. & WncNnR,J.J. (1979b): Composition of parental basalticmagma in ophiolites. In Proc. Int. OphioliteSymp., Nicosia, Cyprus (in, press).

MecGnrcor, I.D. (1962): Geology, Petrokigy andGeochemistry of the Mont Albert and Asso-ciated Ultrurnofic Bodies of Central Gasp6, Qu€-Dec. M.Sc. thesis, Queen's Univ., Kingston, Ont.

& Surn, CJI. (1963): The use of chromespinels in petrographic studies of ultramafic in-trusions. Can. Mineral. 7, 443-412,

Menorn, J.-C. C. & Nrcor-es, A. (1975): Texturesand fabrics of ulper-mantle peridotites as il-lustrated by xenoliths from basalts. l. PetrologyL6,45M87.

Mooov, J.B. (1976a): An experimental study oftle serpentinization of iron-bearing olivines. Caa.Mineral. L4, 462-478,

(1976b): Serpentinization: a review. Zfihos9, 125-138.

NIcKEL, EJI. (1959): The occurrence of nativenickel-iron in the serpentine rock of the EasternTownships of Quebec Province. Can. Mineral. 6,307-319.

(1971): The distribution of nickel amongminerals of asbestos-bearing rocks in the EasternTownships of Quebec. Mineral Sci. Div,, InternalRep. MS ?l-BL, Dep. Energy, Mines and Re-sources, Qttava.

Nrcor-es, A., Boucnrz, J.L. & Bouorrn, F. (1972)tInteq,r€tation cin6matique des d6formations plasti-ques dans le massif de lherzolite de Lanzo (AlpesPi6montaises), comparaison avec d'autres mas-sifs. Tectonophysic,r 14, L43-171.

Or.seN, E.J. (1961): High temlErature acid rocksassociated with seqpntinite in eastern Quebec.Amer. J. Sci, 259, 329-347.

Poor,r, WJI., B€r.euu, J,R. & WeNlpss, R.K.(1963): Minimum age of Middle Ordovicianrocks in southern Quebec. Geol. Soc. Amer.Bull, 74, 1063-1065.

RronooN, P.H. (1955): The genesis of asbestos inultrabasic rocks. Econ. Geol. 50, 67-81.

(1975)t Geology of asbestos deposit$ ofsoutheastorn Quebec. Ministdre Rich, Nat. Qu6-bec, ES. L8,

& Lsrnnnr6, R. (1972): Asbestos depo-

869

sits of southern Quebec. Int. Geol. Congr. 24th'G uideboo k, E xcur sion B-08.

RoDRrcuE, G. (1979): Etude Pdtrologique des Ro'ches Ophiolitiques du Mont Orford, QuCbec. M.Sc.thesis, Laval Univ., Qu6bec, Qu6.

S:r-Jwrcx, P. (1972): Appalaohian structure andstratigraphy Quebec: Int. Geol. Congr. 24th,Guidebook, Excursion A56-C56.

S€crml, M.K. & Llurexr, R. (1978): Magneticsignatures of ophiolitic atrd contnental marginpillow lavas from the Quebec Appalachians.Arch. Sci. Gendve 8L, 179-204.

THAyER, T.P. (1966): Serpentinization con$ideredas a constant volume metasomatic pro**s. Amer.Mineral.51, 685-710.

(1967): Serpentinization considered as aconstant volume metasomatic process: a reply.Amer. Mineral, 52, 549-553.

WIcKs, FJ. & PLANT, A.G. (1979: Elecfion-micro-probe and X-ray-microbeam studies of serpentinetextures. Can. Mineral. 17' 785-830.

& \ilrurrnrrn, E.J.W. (L977): Serpentinetextures and serpentinization. Can. Mineral. t5'459488.

-, & Zussr"reN, J. (1977): An ideal-ized model for serpentine textures after olivine.Can. Mineral. L5, 44G458,

Wn-r.reus, H. & Sr-JwrnN, P. (1978): The BaieVerte-Brompton line in Neudoundland and re-gional correlations in the Canadian Appalachians.Geol. Surv. Can. Pap. 78-L4,25540,

& Tar-m.roro\ R.W. (1977): Distributionand tectonic setting of ophiolites and ophioliticm6langes in the Appalachian Orogen. /a NorthAmerican Ophiolites (R,G. Coleman & W.P.Irwin, eds.), Oregon Dep. Geol. Mineral Ind.Bull. 95, L-ll,

WoLERy, TJ. &Sr-sEp, NJI. (1976): Hydrothermalcirculation and geochemical flux at mid-oceanridges. .I. Geol. 84, 249-275.

Received February 1979, revised manwcript ac-cepted October 1979.

SERPBNTINE ASSEMBLAGES FROM THE QUEBEC APPALACHIANS