PETROLOGY OF HIGH.AI-HORNBLENDE- AND MAGMATIC … · classification of Streckeisen 1976), with quartz diorite being the most abundant lithology. The dominant major minerals are plagioclase

Canadion MineralogistVol. 30, pp.377-392 (1992)

PETROLOGY OF HIGH.AI-HORNBLENDE- AND MAGMATIC-EPIDOTE-BEARING PLUTONSIN THE SOUTHEASTERN CAPE BRETON HIGHLANDS, NOVA SCOTIA

CATHARINE E.G. FARROW" ENO SANDRA M. BARRDepartment of Geology, Acadia University, Wolfuille, Nova Scotia BOP IXO

ABSTRACT

Six large late Precambrian plutons in the southeastern Cape Breton Highlands of Nova Scotia are composeddominantly of diorite, quartz diorite, tonalite, and granodiorite. Their petrochemical characteristics indicate that theyare the plutonic equivalents of moderate- to high-K orogenic andesites, formed in a continental margin subductionzone. The occurrence of high-Al hornblende and magmatic epidote in the plutons in the northwest suggests thatcrusal levels of ca. 20 km are exposed in that area. Calculated pressures of hornblende crystallization decreasesystematically to the southeast, indicating that granodiorites exposed in that sector crystallized in the epizone. Themagmas may have formed by variable amounts of partial melting of mafic granulitic source-rocks, followed byfractionation of mafic minerals and some plagioclase, to produce the internal variation within each pluton.

Keywords:. diorite, quartz diorite, tonalite, granodiorite, I-type granite, calc-alkaline, subduction, catazonal,mesozonal, epizonal, Cape Breton Island.

SoMMarnE,

Nous d6crivons six plutons volumineux, d'6ge prdcambrien, situ€s dans la partie sud-ouest du Cap Breton(Nouvelle-Ecosse); ils contiennent surtout diorite, diorite quartzifCre, tonalite et granodiorite. Leurs caract6ristiquesp6trochimiques en font les 6quivalents plutoniques d'une suite anddsitique i teneur en potassium moyenne i 6lev6e,typique d'un milieu de subduction prbs d'une marge continentale. La pr6sence de hornblende riche en Al et d'dpidotemagmatique dans les plutons du secteur nord-ouest fait penser que des niveaux d'environ 20 km de profondeur yaffleurent. Les pressions de cristallisation de la hornblende diminuent progressivement vers le sud-est, et lesgranodiorites de ce secteur semblent dpizonales. Les magmas pourraient r6sulter d'un taux variable de fusion partielled'une suite granulitique mafique; un fractionnement de mineraux mafiques (surtout) et de plagioclase est responsablede la variation interne de chaque pluton.

(Iraduit par la R6daction)

Mots-cl4s: diorite, diorite quartzifdre, tonalite, granodiorite, granite de tlpe I, suite calco-alcaline, subduction,catazone, m6sozone, 6pizone, lle du Cap-Breton.

INTRODUcTIoN

Six large plutons composed of hornblende-bear-ing diorite, quartz diorite, tonalite, andgranodiorite are major components of thesoutheastern Cape Breton Highlands of NovaScotia (Fig. 1). U-Pb dating has indicated that theseplutons crystallized in the time interval betweenabout 565 and 555 Ma (Dunning et al, 1990), andhence are approximately contemporaneous. Threeof the plutons locally contain epidote havingcharacteristics indicative of magmatic origin, ac-

"Present address: Department ofEarth Sciences, CarletonUniversity, Ottawa, Ontario KlS 586.

cording to the criteria of Zen & Hammarstrom(1984a, b, 1988), and all have mineral assemblagesappropriate for the application of the hornblendegeobarometer (Hammarstrom & Zen 1986, Hol-lister el al. 1987, Johnson & Rutherford 1989,Rutter et al. 1989D.

The purpose of this paper is to document themineral chemistry and petrological characteristicsof these plutons, and to use the occurrence ofmagmatic epidote, combined with the hornblendegeobarometer, to interpret their depths of crystal-lization. This study is particularly significant giventhe continuing controversy about the applicabilityof hornblende compositions as geobarometersversus Eeolhermometers (e.g., Blundy & Holland1990, Vyhnal et al. l99l).

377

378 THE CANADIAN MINERALOGIST

Ftc. l. Simplified geological map of the southeastern Cape Breton Highlands showing plutons that are the focus ofthis study and associated units (after Barr et al, 1985, Farrow 1989, Raeside & Barr, in press). Hadrynianmetamorphic suites are Bateman Brook Metamorphic Suite (BB), McMillan Flowage Formation (MF), andBarachois River Metamorphic Suite (BR). Inset map shows location of the study area in eastern Cape BretonIsland.

HORNBLENDE. AND EPIDOTE-BEARING PLUTONS, CAPE BRETON 379

GsoLocIceL SBrrtNc

Plutons included in this study are the KathyRoad Dioritic Suite, Timber Lake Dioritic Suite,Gisborne Flowage Quartz Diorite, Wreck CoveDioritic Suite, Ingonish River Tonalite, and IndianBrook Granodiorite, as mapped and named by Barr

et at. (1985,1987) and Raeside & Barr (in press).Together, they comprise about half the area of theBras d'Or terrane (Barr & Raeside 1989) of thesoutheastern Cape Breton Highlands (Figs. l, 2,inset). Also present in the area are more felsichornblende-free pluton$ that range in age fromHadrynian to Devonian; these are not included in

Frc. 2. Simplified geological map of the southeastern Cape Breton Highlands (legendas in Fig. l), showing locations for samples in which mineral compositions wereanalyzed; stars indicate locations of samples in which magmatic epidote wasobserved: circles indicate locations of other samples. Adjacent numbers areaverage pressures (in MPa) of hornblende crystallization calculated using theequation of Johnson & Rutherford (1989). The data are from Table 2' in whichsamples are listed from north 10 south within each pluton. Inset map showsproposed teranes in Cape Breton trsland after Barr & Raeside (1989).

380 THE CANADIAN MINERALOGIST

the present study. Metamorphic rocks of theBateman Brook Metamorphic Suite, McMillanFlowage Formation, Barachois River MetamorphicSuite, and Price Point Formation occur in associa-tion with the plutonic units (Fig. l) (Barr et al,1985, Raeside & Barr 1990, in press). Althoughcontacts are not well exposed and faulting iswidespread, all of these units appear to have beenintruded by one or more of the 565-555 Maplutons.

Because they are generally separated by otherunits, contact relations among the plutons areknown in only some cases, and relative ages aremainly based on radiometric ages. U-Pb (zircon)data indicate that the Kathy Road Dioritic Suite(560 a 2 Ma) and the Gisborne Flowage QuartzDiorite (564 + 2 Ma) are of similar age (Dunninget ol. 1990). Although undated, the Timber LakeDioritic Suite is similar in petrological features tothe Kathy Road Dioritic Suite, and hence isassumed to be of essentially the same age. TheIngonish River Tonalite is somewhat younger, witha U-Pb (zircon) age of 555 x. 2 Ma (Dunning elol. 1990). The tonalite contains dioritic xenolithsnear contacts with the Wreck Cove Dioritic Suite,indicating that the Wreck Cove unit is older thanthe tonalite, consistent with the maximumaoAr/3eAr hornblende plateau age of about 561 Mafor the Wreck Cove diorite (Reynolds et al. 1989).U-Pb dating of zircon and titanite hasdemonstrated minimum and maximum ages ofcrystallization of 564 t 5 Ma and 575 Ma for theIndian Brook Granodiorite (Dunning et al, 1990).

PernocnapHv

Dioritic plutons

Rocks of the Kathy Road Dioritic Suite, TimberLake Dioritic Suite, Wreck Cove Dioritic Suite, andGisborne Flowage Quartz Diorite vary from dioritethrough quartz diorite to tonalite (using theclassification of Streckeisen 1976), with quartzdiorite being the most abundant lithology. Thedominant major minerals are plagioclase (andesine)and hornblende, with much less abundant quartzand biotite. K-feldspar is a minor interstitialcomponent in many samples, and clinopyroxene,partially replaced by amphibole, occurs rarely inthe Kathy Road, Timber Lake, and Wreck Coveplutons. Locally, diorites ofthese three plutons lackbiotite and are very hornblende-rich. The GisborneFlowage pluton is characterized by a much greaterabundance of biotite compared to the other threedioritic plutons.

Textures are typically medium-grained hypidio-morphic granular to inequigranular, and rarelyporphyritic. In the latter case, the phenocrysts are

of plagioclase or hornblende. Compositional andtextural variations appear to be gradational withineach suite; Wreck Cove is the most varied pluton,and in some areas, changes in both composition(from diorite to quartz diorite or tonalite) andtexture (from fine- to coarse-grained and locallypegmatitic and from porphyritic to equigranular)occur within a single large outcrop. In someoutcrops of the Kathy Road and Timber Lakeplutons, compositional banding suggestive ofigneous layering is present. Foliation is presentlocally in all four dioritic plutons. Generally, itappears to have been the result of shearing nearthe margins; however, in places flow foliationdefined by alignment of hornblende and plagioclasehas been preserved.

Accessory minerals include epidote, titanite,apatite, allanite, zircon, ilmenite, and magnetite.Secondary (alteration) products include epidote,chlorite, actinolitic amphibole, and "saussurite".Although secondary epidote is abundant, some ofthe epidote in most samples from the GisborneFlowage Quartz Diorite and in many samples fromthe central and northern parts of the Kathy RoadDioritic Suite is inferred to be of magmatic origin(Fig. 2) because it displays the characteristicfeatures described by Zer & Hammarstrom(l98 a). These include epidote with euhedralcontacts with biotite, epidote forming overgrowthson partially resorbed, embayed hornblende, epidotewith zoning and with zones of inclusions, andepidote displaying wormy intergrowths withplagioclase. In both plutons, magmatic epidoteoccurs only in samples that contain modal biotite.

Ingonish River Tonolite

The Ingonish River Tonalite is similar in bothmajor and accessory mineralogy to the dioriticplutons but is more leucocratic, with moreabundant plagioclase and quartz relative to maficminerals. It also contains more biotite relative tohornblende than the dioritic plutons, with these twomafic minerals occurring in approximately equalabundance. The amount of K-feldspar (microper-thitic microcline) increases in abundance toward thesouth, where the pluton is locally granodioritic tomonzogranitic in composition. The texture isgenerally medium- to coarse-grained hypidiomor-phic granular, with mafic minerals tending to occurin clusters. Magmatic epidote is a prominentaccessory mineral in many samples, and displaysfeatures like those described in the dioritic plutons.Other accessory minerals and secondary mineralsin the tonalite also are similar to those in the dioriticplutons.

Indion Brook Granodiorite

The Indian Brook Granodiorite consistsdominantly of zoned plagioclase (oligoclase-an-desine), perthitic microcline, quartz, hornblende,and biotite. The texture is medium grained,hypidiomorphic granular. Abundant large intersti-tial grains of titanite are typically present; otheraccessory minerals include zircon, apatite, andmagnetite. Alteration tends to be more intense thanin the other plutons, and includes moderate tointense saussuritization of plagioclase, chloritiza-tion of mafic minerals, replacement of hornblendeby actinolite, and pervasive hematitization. Al-though secondary epidote is abundant, no epidotewith features indicative of magmatic origin wasobserved.

AMPHIBoLE CHEMISTRY

Chemical compositions

Representative compositions of the amphiboleare presented in Table l. The amphibole is calcic,as defined by Leake (1978), with [Ca+Na]s greaterthan 1.34 and Nas less than 0.67. In most cases,the amphibole has (Na + K)6 less than 0.5, and ismainly tschermakite to tschermakitic hornblende inthe Kathy Road and Timber Lake suites, tscher-makitic hornblende to magnesio-hornblende in theGisborne Flowage, Wreck Cove, and IngonishRiver plutons, and magnesio-hornblende in theIndian Brook Granodiorite. However, some grains,mostly in the Gisborne Flowage Quartz Diorite,have (Na + K)o slightly more than 0.5 and consistof magnesian hastingsite, magnesian hastingsitic

TSG 1. |lmsmAm COMTSmONS OF NntrctE mOM &UTONS OF n6 WDY

381

hornblende, edenitic hornblende, ferrotscher-makite, or ferroan pargasitic hornblende. How-ever, for simplicity in subsequent discussion, all theamphibole is referred to as hornblende.

The suite of hornblende compositions displayvariation in aluminum content, both from onepluton to another and within individual plutons.On average, hornblende from the Kathy RoadDioritic Suite contains the highest total aluminum,and hornblende in the Wreck Cove and IndianBrook plutons has the lowest aluminum content(e.g., Table l). The differences in aluminumcontent appear to be reflected in color differencesin thin section. Hornblende in the Wreck Cove andIndian Brook plutons is pleochroic, with Z = darkgreen, I : medium green, and X = beige to paleyellow, whereas higher-aluminum hornblende inthe Kathy Road, Gisborne Flowage, and IngonishRiver plutons has Z : dark blue or blue-green, Y= medium blue-green, and X = beige to paleyellow.

Hornblende geobarometry and geothermometry

Geobarometers based on the aluminum contentof hornblende equilibrated with quartz have beenproposed by Hammarstrom &Zen (1986), Hollisteret at. (1987), and Johnson & Rutherford (1989) forrocks containing the mineral assemblage quartz +plagioclase + K-feldspar + hornblende + biotite+ titanite + an oxide phase (magnetite orilmenite). The equations proposed by these inves-tigators give somewhat different results, with thelowest calculated pressures obtained from theequation of Johnson & Rutherford (1989). For thepresent study, pressures were calculated using all

HORNBLENDE. AND EPIDOTE-BEARING PLUTONS, CAPE BRETON

N&36a il.#B6 n f u 6 t € m

kms mS GrO tm rcDSsMlg AM€4"53 KO9-S21 K09-S103 N&52 ff€8{9@@ dm 6ru dm 6m dm 6G fu @re tu @n nd

ImNSam0lo ses4{9

SOr 41.79 42.72 42,16 42.20rc, 0.77 0.73 0.76 0.94A.O, 14.18 13,53 12.26 12.13G,o, 0.05 O O.l0 O.1OF€ 10.0e 17.04 18.07 18.50MnO 0.32 0.33 O.4 O.4lMSO 9.86 S.7S 9.77 9.01&o 11.42 11.91 11.81 11.76Mp 1.62 1.45 1.41 1.61K,O 0.S6 0.71 1.4 1,26l@t 97.82 98.21 9S.02 98.31

41 .23 42.49 4.83 €.991.19 1.O7 0.62 0.57

11.16 10.99 12.6 12.380.06 0.10 0,10 0.m

18.49 10.50 19.83 19.08o.42 0.42 0.68 0.539.72 9.73 8.59 8.&

12.07 11.98 12.02 11.U1.54 1.AO 1.49 f . i l1.67 1.50 1.m 1.28

37.4€ 99.37 97.29 98.83

42.55 43.35 41.831.O2 1.03 0.90

12.U 1r.56 10.840.03 0 0.08

18.12 17.13 18.790-62 0.52 0.80

10.02 10,70 9.€9r 1,90 t 1.93 I 1.551.09 1.16 1.91.02 0.90 1.u

98.28 98.17 97.16

41.72 45.82 46.590.96 0.65 0.89

t l ,o3 8.93 9.21o.o8 0.@ o.r0

19.@ 14.88 14.900.67 0.42 0.4s.68 12.99 12.82

11,75 12.12 12.201.2A 1.33 1.241.37 0.95 0.94

97.4 90.08 98.Or

47.03 48.98 4A.73 47.750,99 1. i l 1.23 [email protected] 7.33 7.23 6.710.& 0.04 0.03 0.04

15.01 15.09 14.03 13.980.89 0.88 0.58 0.61

12,49 12.06 13.20 13.6312,02 12.01 11.99 12.000.97 0.90 1.19 0.970.60 0.64 0.71 0.58

97.19 97.411 98.92 97.29

NUMBM OF CAIONS ON NE NS OF 23 ONGS Arc6

8.31 A.72 8.?O 8.96 6,93 e.92 7.001.70 r.28 1.S 1.05 1,07 r.08 l [email protected] 0.27 0.29 0.24 0.20 0.18 0,17o,ol o.o1 0.01 o.o1 0.ol 0.m 000.88 0,4 0.4 0,30 0.9 o.2a 034o,t I o,o7 0.o8 0,1 I 0.12 0.14 0.1 12ia 2.a2 2.78 2.76 2.78 2.91 2.981-71 1.42 1,43 1.58 1.62 1.49 1.38o.o7 0.o5 o.oo 0.09 0.0s 0.07 06l.9o 1.91 1,92 l .S 1,90 1.S 1.890,38 0,38 0.35 0.2A 0.28 0.34 0.28o.2o 0.18 0.19 0, l l o. l2 0.13 0.11

st 8.21 8.31 8.31 63r e.27 6,39 8.23 A,27 8.31 A.40 8.33!A r.?s 1,89 1.89 l .63 1.73 r.81 1.77 1.73 1.89 1.60 1.87sA 0.69 0.67 o,47 o.4 0.27 0,34 0,€ o.rc 0,41 o.4t o.28& 0,01 0 0.01 0.01 0.01 0.01 0.01 0.01 0.@ 0 0.o1F6!' 0,45 0.4 0.m 0.50 0.4 0.3 0.9 0.€ o.87 0.60 0.88T 0.09 0.08 0.o8 0.09 0.14 0.12 0,o7 0.07 0.11 0.11 0.loMs 2.14 2.16 2lA 2i4 2.2O 2.19 1.96 1.S5 2.21 2.35 2,23F€f. 1.84 1.87 1.78 1.81 1.99 1.99 1.e7 1.99 1.58 lA2 1,70Mn 0.04 0.04 0.08 0.08 O.05 O.O5 0.07 0,O7 0.07 0.07 0'04& 1,88 1.89 1.88 l .S8 1.97 1.93 1.97 r.94 1.89 1.e7 1.87Ns 0,4 O,42 o.41 o.4 0.4€ O.47 o.{ 0.€ 0.31 0.33 o.39K 0.18 0.13 0.2A O,24 0.30 0.2S O.27 0.26 0.19 0.17 0.29

AbbBtsdoB KRDS hfrytusd &ddcS,mstubLste Dondc&iF, Groctrm RMS(]ljeM, tf lqffiTonalt6,WSWd@oMc&iE, tsGlndstl &dG|dbtu'MysaMUrfr6dtybyJs!-733do@nm@pr@m!@wmstn

382 THE CANADIAN MINERALOGIST

TABLE 2. CATCULATED PRESSURES AND TEMPERATURES BASED ON HORNBLENDE

Ptsute(MPal

Temp.fc)

J&F B&HH&Z H stal

SB-8/r-49SB€/t-42sB-84-41sB-84-134sB-84-124

AM-84-53

K10-521cF€8-14RR{4-8

K09-S103KOg-S89K10-S8sB-86-31 84AM-84-52

K10-S14cF-88409

AM-8+7RR-87-5503AM-84-38AAM-84-35AM-44-74AM€4.75

AM-84-86AM-84-r6IBG.1

1J

o

2.292.19

2.O12,31

1 . 8 11 . 8 1

'r.80

6 1 .46 6 .7 r4 1.25 7.O47 1.30 6.917 1.30 6.90, t . o u o . o u

4 1.40 6.82

7 0 .87 7 .186 0 .89 7 .104 1 .09 6 .91

2,40 4.262.94 6.232.37 6.272.13 6.321.41 6.90

0.610.61o.5 l0 .51

0.640.640.64

0.500.600.500.500.60

0.52o.62

o.65

0.660.65

0.660.65u.oo

8781013e a l

7263 1 9

624

669 855771 843657 863T A t A E A

260 769

e l E

800679317

589 479 837Tl"DS

GFOD

IRT

324

4

8.276.20

6.406.33

6.49

6.67

694 742 668 816760 816 A8 823710 759 680 425

619 958 604 862770 827 631 847589 424 479 818518 646 420 830618 546 420 844

513 539 416 420403 477 322 809

342 347 272 763237 229 183 703262 267 204 729262 267 204 732413 426 331 761312 305 246 743

48 16 22 69756 26 30 710

156 r39 1 16 742

IBGWestl

IBGGast)

rPluton abbr*iations as in Table 1. Prsssuras were ulculated 6ing th€ eq@tio6 of Hammarstrom & Zen1986 (H&Zl, Hollistsr 6t al. 1987 (H et al.), and Johmon and Ruthsrtord 1989 (J&F). Calculated prssures othomblsnd€ crystalllzation ars multipli€d by 100 to givg vslues In MPa. Tomparaturos wg.g €lculated singtho sqmtions ot Bludy & Hollend (19901 with th€ calcutatod prossuros from thB eqution of Johmon &Rutherford (19891. n = number of homblsnds rim amlys€s usBd to €lculate averag€ total alminum (Af) andavetaoe sillcon (Sil per iomula unit. Ab ls tho ccsxisting plagislase composition.

three equations for comparison (Table 2); thepatterns displayed by the data are the same in allthree cases, but the pressures calculated using theequation of Johnson & Rutherford (1989) seemmost compatible with adjacent metamorphic unitsand are cited in the following discussion and shownon Figure 2.

The application of the hornblende geobarometerhas been questioned, on the basis that the Alcontent of hornblende is more sensitive to tempera-ture than to pressure, and a geothermometer hasbeen proposed for hornblende coexisting withplagioclase in silica-saturated rocks (Blundy &Holland 1990). Temperatures have been calculatedfor the units of this study (Table 2) using theequation of Blundy & Holland (1990) and urilizingthe pressures calculated by the equation of Johnson& Rutherford (1989).

The calculated pressures are highest for samplesfrom the northern part of the Kathy Road DioriticSuite, and they generally decrease to the east andsouth (Table 2,Fig.2), The pressures suggest thatthe nofihern part of the Kathy Road Dioritic Suite,as well as the adjacent Gisborne Flowage Quartz

Diorite, crystallized at mesozonal to catazonaldepths, whereas the southernmost part of the KathyRoad Dioritic Suite formed at a much shallowerdepth. This variation is broadly consistent with achange in metamorphic grade documented in theadjacent McMillan Flowage Formation from am-phibolite facies in the north to greenschist facies inthe south, although the pressures indicated bymineral assemblages in the northern part of theMcMillan Flowage Formation are not as high asthose indicated by the hornblende geobarometer foradjacent parts of the Kathy Road Dioritic Suite andGisborne Flowage Quartz Diorite (Raeside & Barr1986, 1990).

The Timber Lake, Ingonish River, and WreckCove units yield intermediate pressures of crystal-lization, and the lowest pressures are from thesoutheastern part of the Indian BrookGranodiorite; these results indicate crystallizationat a pressure less than 100 MPa (epizonal). Thelatter is consistent with the preservation in that arcaof subgreenschist-facies volcanic rocks of the PricePoint Formation (Fig. 1) (Raeside & Barr 1990).

In contrast to pressures, the temperatures

HORNBLENDE. AND EPIDOTE.BEARING PLUTONS, CAPE BRETON 383

calculated using the equation of Blundy & Holland(1990) show a relatively small range within eachpluton that does not exceed the reported uncertain-ty for their calibration (t 75"C). As would beexpected, higher temperatures are generally indi-cated for the dioritic plutons compared to theIndian Brook Granodiorite.

EPIDOTE CHEMISTRY

Chemicol compositions

Epidote grains of inferred magmatic origin fromthe Kathy Road, Gisborne Flowage, and IngonishRiver plutons commonly show a decrease in Al andincrease in Fe contents from core to rim (Table 3).This is attributed to replacement of Al by Fe3+ nearthe margins of magmatic epidote crystals andsubsequent growth of more Fe-rich epidote. Similarcompositional differences are apparent betweenmagmatic and secondary epidote, with composi-tions of secondary epidote similar to rim composi-tions on magmatic epidote grains. Secondaryepidote also tends to be lower in Ti; Cr, Mg, Mn,and Na contents are variable, whereas Ca contentis relatively constant (Table 3).

The average proportion of the pistacite com-ponent ranges from 24.50/o it the Kathy RoadDioritic Suite to 26.8t/o irt the Ingonish RiverTonalite. These values are slightly lower than those

of epidote with textural relationships (e. g. , replace-ment of plagioclase) that indicate a secondaryorigin; such epidote generally contains between 26andnn/o (Table 3). However, both types of epidotehave pistacite components within the range 25 to29 th^.has been reported to be typical of magmaticepidote (Tulloch 1979,1986, Vyhnal et ol. l99l).

The epidote compositions documented in thisstudy differ markedly from those analyzed byVyhnal et at. (1991) from monzogranitic plutons inthe southern Appalachian Orogen. The lattertypically have a higher proportion of the pistacitecomponent, lower Si, Al, and Ca, and higher Fe'

Pressure imPlicotions

Naney (1983) showed that epidote is stable insynthetic H2O-saturated tonalitic and granodioriticmelts at /(Or; values between the nickel-bunsenite(NB) and himatite-magnetite (HM) buffers, ati.mperatutes between the solidus (600'C) and

approximately 700oC, and at pressures at least inthe 600 to 800 MPa range. Zen & Hammarstrom(1984a) proposed that magmatic epidote forms nearthe solidus of the crystallizing magma according tothe following schematic reaction:2 hornblende + 2 alkali feldspar + I magnetite+ nHrO + mO2 : 3 epidote + 3 biotite + 6qvarlz.

Minimum pressures of crystallization lower than

TABI-E 3. REPRESENTATIVE COMPOSMONS OF MAGMAIC AND SECONDARY EPIDOTE

Secondary Epidote

KRDS IRT WCDS

Magmatic Epidote

KRDS GFOD IRTCore Rim Core Rim Core Rim

sio2TlozAlro"Cf2o!FeolMnOMsoCaONa20Total

38.020 .16

26.1 8o.04

1 0.67o.320.07

23.310.07

97.74

st 3.06At 2.39Fo' 0.76Cr 0.003-ri o.01Mg 0.01Mn O.O2Ca 2.O1Na 0.01

Ps% 22.96

37.68 57.74 38.240.15 0.1 1 0.10

23.77 24.89 24.640.08 0 0.03

12.63 11.53 12.230.08 0.23 0.240.03 0 0

23.54 25.55 23.710.09 0.01 0

98.05 98.05 99.12

3.06 3.05 3.072.2A 237 2.320.86 0.78 0.820.01 0 0.0020.01 0.01 0.010.004 0 00.01 0.o2 0.022.05 2.O4 2.O40.01 0.002 0

27.37 24.73 26.12

37.80 37.610.37 0.12

24.11 22.930.07 0.06

1 1 .36 13.090.30 0.230.05 0

23.44 23.460.09 0.05

97.65 97.55

3.07 3.092.31 2.22o.77 0.900.004 0.0040.02 0.010.01 00.o2 0.o22.O4 2.060.01 0.01

25.05 2a.A2

37.84 37.69 37.360.o5 0.12 0.o7

22.92 22.68 23.190.06 0.o2 0.08

13.23 12.97 12.110.32 0.16 0.480.05 0 0.19

23.25 23.88 23.250.03 0.09 0.08

97.74 97.49 96.80

3.10 3.10 3.082.21 2.19 2.250.91 0.89 0.830.003 0.001 0.010.003 0.01 0.004o.o1 0 0.020.02 0.01 0.032.04 2.10 2.O50.005 0.01 0.01

29.05 28.97 27.03

NUMBEF OF CA'NONS ON THE BASIS OF 25 OXYGEN ATOMS

Epidote analyses by slectron microprobe as in Table 1Ps% = pistacits componefi.Pluton abbrgviations as in Table l.

384 THE CANADIAN MINERALOGIST

600 MPa were suggested for more silica-rich rockssuch as calc-alkaline granite, whereas pressureshigher than 800 MPa were proposed for lesssilica-rich rocks, including quartz diorite anddiorite (Zen & Hammarstrom 1986). Magmaticepidote in two-mica granite commonly is morepistacite-rich and formed at considerably lowerpressure than is indicated by the hornblende-epidote association (Zen & Hammarstrom 1988).

More recently, the high-pressure origin ofmagmatic epidote in hornblende-bearing rocks hasbeen questioned, because epidote with features thatsuggest magmatic origin occur in rocks for whichthe calculated pressures of crystallization forcoexisting hornblende are as low as 280 t 50 MPa(Vyhnal et al. 1990). However, the bulk composi-tions of the rocks are monzogranitic rather thantonalitic and granodioritic, and a high-pressureorigin for epidote in rocks of the latter composi-tions appears still to be valid.

The occurrence of magmatic epidote in the KathyRoad Dioritic Suite, Gisborne Flowage QuartzDiorite, and Ingonish River Tonalite is consistentwith the high calculated pressures of hornblendecrystallization (Fig. 2). The Timber Lake andWreck Cove dioritic suites and the Indian BrookGranodiorite lack magmatic epidote, and in theseplutons, hornblende geobarometry has yieldedpressures of crystallization too low to be com-patible with crystallization of magmatic epidote in

rocks of quartz dioritic and granodioritic composi-tion.

Wrrorr-Rocr CHEMISTRY

General char acteristics

Means and standard deviations for major andtrace element concentrations are compiled in Table4. Results of the analyses and sample locations forthe dioritic and tonalitic units are presented byFarrow (1989). Data for the Indian BrookGranodiorite are given in Table 5.

The Timber Lake, Wreck Cove, and Kathy Roaddioritic plutons have average silica contents typicalof mafic rocks (less thar 52t/o), with the lowestaverage in the Timber Lake Dioritic Suite. Thelowest silica values are from samples with highconcentrations of mafic minerals (mainlyhornblende), and each unit has a range in silicacontents and corresponding variations in otherelements, as illustrated in Figure 3, consistent withthe variation in modal composition of the samplesthat range from diorite through to quartz dioriteand tonalite. The Gisborne Flowage Dioritic Suitehas generally higher silica content than the otherthree dioritic suites, averaging about 5590, al-though some samples have SiO2 values as low as5090 (Fie. 3).

The Ingonish River Tonalite has a higher average

TABLE 4. MEANS AND STANDARD DEVIATIONS OF CHEMICAL COMPOSMONS OF PLUTONS'OF THE STUDY AREA

PLUTON TLDS(n) ln=51

wcDs(n =91

KRDS(n= 161

GFOD(n= 101

IRT(n=201

IBG( n = 1 1 1

sio2 47.31 * 4.45ro" 0.76 o.24Ar2q 18.34 2.97Fe2O.t 10.42 1.62MnO O.19 0.02MsO 6.64 2.04CaO 10.1 1 1 .77Na2O 2.1O 0.79KrO 't.43 O.21PrOr 0.35 0.38LOr 2.22 0.65Total 99.86 0.42

5 1 . 4 9 r 6 . 1 90.81 0.25

1 6 . 1 6 1 . 4 310.09 2.6',1o .17 0 .035.89 1 .819.O2 2.402.24 0.671.45 0.460 . 1 8 0 . 1 02.32 0.95

99.82 0.59

51.91 r 4 .410.82 0.27

17.47 1.439.70 1 .550.18 0 .035.34 0.779.26 1.452.62 0.390.88 0.430.16 0 .051.24 0.60

99.58 0.42

55.15 r 2 .580.81 0 .15

17.76 0.628.03 1 .12o.17 0 .o23.63 0.527,48 1.403.27 0.312.16 0 .77o.25 0.09o.97 0.44

69,2'l x 2.720.53 0.09

17,08 0.676.94 1 .1 1o .14 0 .o23.39 0,57

2.83 0.2',11 .71 0 .420.16 0 .021.38 0.43

100.04 0.62

64.10 * 4.370.48 0.16

16.27 1.604.47 1.640.10 0.042,10 0 .984,O7 1.673.33 0.423.24 't.O2

0.16 0 .061.8 r 0 .86

99.63 0.83

Ba 224 57 216 97 166 76 491 132 326 S6 439 106Rb 55 I 43 20 2S 17 62 26 49 ,18 95 37sr 439 97 385 70 368 61 588 1,16 426 5? 343 132Y 2 1 7 2 1 7 2 2 6 2 5 3 1 8 2 2 2 4Zt 62 16 93 66 82 30 i73 56 93 18 164 27N b 5 1 7 2 6 1 9 2 7 2 1 0 2Cu 7a 23 132 71 91 38 44 30 61 24 39 3gP b 1 6 1 6 9 2 1 0 < 1 1 0 1 8 3 1 0 3zn 1O4 20 94 24 91 ,t7 90 ,t2 79 I 64 17Ni 31 25 30 21 26 12 .10 2 6 3 I 3cr 72 77 45 45 55 61 14 9 24 9 39 14v 297 124 302 125 301 63 209 40 198 39 123 66G a 1 8 3 1 8 3 1 8 2 1 9 2 1 8 2 1 6 2T h 1 0 0 8 3 9 3 9 3 4 3 1 5 1 3

tPluton abbreviatioG as in Table 1. Amlytical data for TLDS, wcDs, KRDS, GFoD and IRT from Farow (1 9991,and tor IBG from Table 5. Amlyses by X-ray fluorescenc€ as describ€d by Fanow (1 9g91.

HORNBLENDE. AND EPIDOTE-BEARING PLUTONS. CAPE BRETON

TAAIE 5. CHEMICAL COMPOSMON OF INDIAN BROOK GFANODIORIrE

385

1 2 9 4 5 0 t 9

2e7 670 459 487 647 329 308 624 553 514 399'123 r38 rO8 143 135 46 74 109 60 1 10 4g60 242 335 216 288 462 350 360 40S 408 5m2 6 " 2 6 2 1 2 6 1 8 1 5 2 A 1 7 2 6 2 2

217 . 167 113 177 1e7 142 161 188 187 1551 0 " 1 2 1 1 l 0 8 1 3 l O I 1 0 55 a 78 5 5 38 rO5 58 1r 51 20

1 0 9 1 0 1 0 1 0 5 1 0 1 4 1 2 "62 68 49 36 71 7A 72 69 64 57 907 " 1 4 0 I I 8 r r 6 6 1 0

6 7 - 3 7 4 1 5 1 5 8 3 0 3 5 3 7 2 3 1 448 - 94 43 137 222 146 140 62 110 2241 4 " 1 4 1 4 1 6 r 8 1 8 1 7 1 0 1 8 1 74 3 - 1 1 2 7 1 0 2 1 0 1 2 6 1 2

so.To.&qFqq'MroMgoCaONaPKpPP!LOITo@l

Ba&Sr

ZrM

tu2nMCr

Gan

7r.08 87.6E 65,160.37 0.42 0.43

11.88 t3.54 16.983.50 3.41 4.470.00 o.o7 o.o81 . 1 4 l . t 6 1 . 0 92.O7 2.90 3.722.49 3.4it 3.464.42 4.0S 3.340.r0 0.13 0.r43.O8 2.91 1.70

100.64 99.87 99.59

66.66 63.4'l0.r0 0.54

$.42 16.472,42 4.980.05 0.081.83 1.622.82 3.28s.21 3.053.S0 4.14o.o8 0.r31.66 3.26

97.1S 99.84

57.65 81.320.6t 0.53

16,63 1€.137.04 6.060,15 0,133,81 2.797.12 5.653.24 3,34'f .83 2.73o . 1 7 0 . 1 71.50 0.99

99.55 99.84

e4.19 67.010.54 0.40

1 5 , 1 8 1 0 . r O

o . t 2 0 . 1 02.47 1.423.A2 3.503.24 3.883.90 2.450.r€ 0.161.21 0.97

9S.86 99,49

€3.S6 58.730,64 0.73

15.49 18.975.0a 7.sl0,1' t 0,18L& 3.584,14 6.073.23 4.103.67 1.32o,l7 0.t21.10 1.60

99.?6 99.51

Amlts by l-q R@t!@ @ d@ibed by Frr@ 11 9991. Smple lcathB qtoon In mil|& ot ladnde 46'N 8d mlnd6 ol lomMo@Ht 1,20.76 &32.4t 2,2O.7 &32.6i 3.20,76 & 33.1;4, lg.3 & 33.7; 5,20.55 & 33.8; 6, 30.8 & 2S.85; 7, 30,85 & 28.s; S,26.95& 29.95; 9, 25.95 & 32.7:10,20.8 932.2t 11,28,5 & 33.4.

SiO2 content of about 5990, but overlaps in SiOtcontent with more silicic samples from the dioriticunits, and with the more mafic samples from theIndian Brook Granodiorite (Fie. 3). The IndianBrook Granodiorite shows a wide range in silicacontent from about 57 to 72t/0, with an averagevalue of 64V0. This range in chemical compositioncorresponds with variations in the proportions offeldspar and mafic minerals in the unit; thesevariations seem to be gradational.

Taken as a group, the plutons range more or lesscontinuously from mafic to felsic (Fig. 3). Mostmajor element oxides display linear trends on silicavariation diagrams (e.9., Fig. 3); TiO2, Al2Or,Fe2O3t, MnO, MgO, and CaO display definitenegative correlations (correlation coefficients morethan -0.7) with silica, whereas K2O and Napdisplay definite positive correlations (correlationcoefficients more than +0.7) with silica. P2O5shows only a moderate correlation (correlationcoefficient -0.55), mainly because of scatter inmafic samples that appears to correspond tovarying apatite content.

Among the trace elements, Zn, Y, and Ga showstrong negative correlation with silica (e.g., Fig. 3).Ni and Cu show only moderate correlation, withmost of the variation in the mafic plutons. Sr showsmoderate negative correlation (correlation coeffi-cient about -0.5) with SiO2, and both Ba and Rbshow positive correlation (+0.5 and +0.7, respec-tively) with SiO2. Nb contents are low, betweenabout 5 and 10 ppm. Zr values are high in theGisborne Flowage Quartz Diorite compared to theother dioritic plutons (Fie. 3).

Petrogenesis

Overall, the chemical variations within eachpluton and among the plutons as a group suggesta major role for crystal fractionation in magmaevolution. Plots of Rb versats Sr and Ba versas Sr(Fig. ) show large increases.in Rb and Ba withlittle change in Sr in going from more mafic tomore felsic samples, consistent with fractionationof dominantly hornblende combined with someplagioclase; the latter appears to become moresignificant in the Indian Brook Granodiorite. Thevery low Rb and Ba values in some mafic dioriticsamples may reflect hornblende accumulation.

Although both Fe and Mg show negativecorrelation with silica content (Fig. 3, Table 4), theFeOt/MgO ratio stays approximately constant (Fig.5), indicative of calc-alkaline affinity (Miyashiro1974). However, the wide variation in Cr withapproximately constant V in the dioritic andtonalitic plutons (Fig. 6) is not typical of calc-alkaline suites. Electron-microprobe analyses ofhornblende in these plutons indicate that minoramounts of Cr (up to 0.4v/o Cr2O3) are present(Farrow 1989), and hence Cr variation may haveresulted from hornblende fractionation in theseplutons. In addition, relict pyroxene cores €uepresent in some hornblende grains, suggesting thatearlier fractionation of Cr-bearing pyroxene alsomay have contributed to the range in Cr values.The lack of change in V suggests that magnetitefrilctionation was not significant in the dioritic-tonalitic plutons; in contrast, decrease in V withless change in Cr indicates that magnetite may haveplayed a significant role in producing variation inthe Indian Brook Granodiorite (Fig. 6).

386 THE CANADIAN MINERALOGIST

23

AlrO. tz

20

Fe.of to

oI t

CoO

o

Kzo

o0 v u

ot 50

Zn

o250

Zr

o

KRDSTLDSOFOD

IRTwcDs

IBG

h - l R

n:10n.20n . 9n:11

xD+AIO

'E "1'

t o o f t - .q:Sle

; o o\ t o

Rb ppm 10

pt s\ ,/61" *-\

\\

KRDS X n=15 ),TLDS D n.56FoD * n.10

IRT A n:4WCOS I n=9

IBG O n.11

F'T 1"*,m:*- "

45 50 55 60 65 70 75

s i02

Ftc. 3. Silica variation diagrams for selected majorelement oxides and trace elements in plutons of thisstudy. Data from Farrow (1989) and Table 5.

Distribution in large separate intrusions that insome cases appear to have crystallized over a rangeof depths does not suggest direct genetic links

Bo ppm 1oo

10 100 1000Sr ppm

Frc. 4. Plots of (a) Rb-Sr and (b) Ba-Sr for plutons ofthis study. Vectors show trends for Rayleigh fractiona-tion of 3090 plagioclase (Pl), K-feldspar (Kfs), biotite(Bt), and hornblende (Hbl), calculated using thefollowing KD values (from McCarthy & Hasty 1976,Tindle & Pearce 1981) for plagioclase, K-feldspar,biotite, and hornblende, respectively: Bal. 0.4, 6, 6.36,0.35; Rb: 0.04, 0.8, 3.26, 0.011; Sr: 3.35, 3.6, 0.12,0.058.

among the plutons; they are unlikely to be relatedto one another by differentiation processes. In-stead, each pluton may represent an individualevolving batch of magma, perhaps generated fromsimilar source-rocks by varying degrees of partialmelting and then subject to fractional crystal-lization. The similarity in chemical trends displayedby the plutons (Fig. 3) probably representssimilarity in both source rocks and subsequentevolution of the batches of magma.

\1xXx

tx {

o

"ol WrR&ftt

x r t rxcl ts,.'T4msfu s

HORNBLENDE. AND EPIDOTE.BEARING PLUTONS, CAPE BRETON 387

KRDSTLDSOFOD

IRTwc05

IBG

n=15n E 5n 310

n=20n = 9n =11

6

Xtr+AI

THOLEI IT IC

3

FeoTMso

+lxtbx

, +a /

tr x lI

CALC-ALKALINE

o L40

sio2

Frc. 5. Plot of FeOr/MgO against SiO2, with dividing line between tholeiitic andcalc-alkaline rocks after Miyashiro (1974).

trt ms

807060wlTo

50

T H * C A

KRDS X n.15TLDS tr n.5OFOD * n=10

IFI A n'20WCDS I n'9

IBG O n=10

1 10 't@ t000

Cr ppm

Frc. 6. Plot of V against Cr, with ttroleiitic (TH) andcalc-alkaline (CA) fields after Miyashiro & Shido(r97s).

Rare-earth elements

Rare-eafih element (REE) dala are available fora total of nine samples representing the Kathy RoadDioritic Suite, Gisborne Flowage Quartz Diorite,Ingonish River Tonalite, and Indian Brook

Granodiorite (Table 6). Tolal REE abundance islow in the Kathy Road and Ingonish River samples,and much higher (especially in the case of the lightREE) in samples from the Gisborne Flowage

Quartz Diorite and Indian Brook Granodiorite(Fig. 7). This difference is consistent with theabundance of accessory phases (epidote, allanite,zircon, titanite) in the latter two units.

The Kathy Road and Ingonish River sampleshave essentially identical chondrite-normalizedREE patterns, with slightly lower heavy R.EE in theIngonish River Tonalite (Fig. 8). The three samplesfrom Kathy Road Dioritic Suite represent a rangeof silica contents from low to moderate to high.The low-silica sample has the lowest concentrationsof LREE, but HREE patterns are similar in allthree samples, and very flat.

The Gisborne Flowage and Indian Brooksamples also have similar R.EEpatterns, with a highabundance of heavy REE compared to the KathyRoad and Ingonish River samples (Fie. 8).

Overall, theflat HREE patterns in all these unitsare most consistent with an origin by partial meltingof source rocks in which neither garnet norhornblende was present in the residue. The presenceofgarnet in the residue generally leads to depletionirt HREE, whereas the presence of hornblendetends to result in concave-upward HREE patterns(Hanson 1980), neither of which is indicated bythese data (Fig. 8). A mafic granulitic source, with

388 THE CANADIAN MINERALOGIST

TABT"E 6. FAFE.EAATH ELEMENT DATAI

' t 2 3 4 5 6 7 A 9

KRDS KRDS KRDS IRT IBT GFOD GFOD IBG IBG

ta 9.94Co 23.22Pt 9 .12Nd 13.26sm 3.23Eu 0.81Gd 2.96Tb 0.60Dy 3.36Ho 0.67E! 2.OATm 0.28Yb 1.99Lu 0.31

10,74 7.9424.78 77.2A

3.00 2.3811.70 10 .152.72 2.730.63 0.792.37 2.67o.42 0.482.69 2.940.57 0.651.60 1 .84o,24 0.261.61 1 ,420.26 0.28

26.59 24,7954.68 56.13

6.75 6.8925.86 27.35

5.19 6 .831.21 1 .383.55 4.34o.a7 0.693.46 3.930.64 0.801.82 2 .220.26 0.3r1 .82 2 .17o.28 0.32

31.26 27 .6769.18 62.54

30.87 24.856 . 1 6 4 . 1 31.44 1 .54

o,71 0.47

10.7026.03

3.3814.363.48d 7 q

2.770.45

0.541.640.23

o.24

9.0817.63

1 1 .302,79

2 . 2 5

0.392.37o,47t . J 6

o.211.47 2.49 1.74

u.50 u .z6

I Amlysd 1-7 by ICP-MS at Memo.ial University, St John's, Noffoundtand. Amlysos I & 9 by lnstrumentalNautron Activation at St. Mary's Univorsity, Halifax, Nova Scotia. Sample locations giv€n in minutos offa t f tuds46oNandminutesot long i tuda6Ol i l :1 , 16 .5&45.5 ;2 .23 .7&42.4 :3 ,25 .3&43.6 ;4 ,33 .1 &30.8 ;5 ,23 .3 & 39 .1 ;6 ,32 .1 & 37 .4 ;7 ,31 .6 & 38 .0 ; S , 20 .8 & 32 .2 ;9 , 30 .86 & 28 .9 . Samptes 8 and Iaro numbers 10 and 7, rospectively, in Table 5.

?o-

g

lrJ

H 100

tE,p

to 50 60 70Si0a

Ftc. 7. Total REE against SiO2 for samples from KathyRoad Dioritic Suite, Ingonish River Tonalite, Gis-borne Flowage Quartz Diorite, and Indian BrookGranodiorite. Total includes La, Ce, Nd, Sm, Eu, Tb,Yb. and Lu from Table 6.

plagioclase and pyroxene in the residue, couldgenerate melts with REE patterns like thosedisplayed by these units. The lack of strong Euanomalies in most samples suggests either thatfeldspar fractionation was not a major process or,more likely, given the evidence from other majorand trace element data for fractionation ofplagioclase and mafic minerals within each unit,that feldspar fractionation was balanced byhornblende fractionation so as to minimize Euanomalies (Hanson 1980). Fractionation of bothfeldspar and mafic minerals. probably led toenhancement of LREE and depletion in HREE.

Tectonic setting

Overall, the plutons of the southeastern CapeBreton Highlands are typical of compositionallyexpanded calc-alkaline I-type suites formed inassociation with continental-margin subductionzones (e.9., Pitcher 1982,1987, Brown et al. 1984).They are compositionally equivalent to subalkalinebasaltic to dacitic volcanic suites formed in suchsettings (Fig. 9). On diagrams suited for thediscrimination of tectonic settings of mafic volcanicrocks, the mafic dioritic samples (SiO, less than52Vo) generally plot in the volcanic arc fields,although with considerable overlap with ocean-floor fields (Figs. l0a, b).

Further support for origin of these units in asubduction-related setting is their hieh Al contents;the Gisborne Flowage Quartz Diorite is very similarin average major-element composition to theaverage high-K, low-SiO2 orogenic andesite of Gill(1981), and the Ingonish River Tonalite is like theaverage moderate-K, high-SiO2 orogenic andesite(Table 7). These similarities to moderate- andhigh-K arc rocks indicate the presence of thickcontinental crust. We interpret these units to be theplutonic equivalents of volcanic-arc basalts andandesites and their differentiation products,formed in a late Precambrian subduction zone. Thecalc-alkaline volcanic rocks of the Price PointFormation (Macdonald & Barr 1985) may be relicsof the cogenetic volcanic arc suprastructure, mostof which has been removed by erosion farther tothe northwest, where deeper crustal levels areexposed.

The apparent difference in age may be relatedto more rapid cooling and crystallization of thehigh-level magmas. However, the Ingonish River

HORNBLENDE- AND EPIDOTE.BEARING PLUTONS, CAPE BRETON

KRDSacnn

IRTIBG

Ftc. 8. Chondrite-normalized REE patterns for samples from Kathy Road DioriticSuite, Ingonish River Tonalite, Gisborne Flowage Quartz Diodte, and IndianBrook Granodiorite. Normalization to the chondrite values of Evensen el a/.(1978).

389

o:

o

! roo,oE(n X

+A

KRDSTLDSGFOD

IRTwcDs

IBG

xE+AIo

n=15f t = 5

n ='10 Rhyotiten =20n = 9

. n = 1 0

RhyodociteDocite

CornenditePontellerite

OTrochyondesite

Zr/TiO"Andesite

Andesite/Bosolt Alkoline Bosolt

Subolkoline Bosolt

SUBALKALINE SERIES ALKALINESERIES

'-0.0't 0.10 1.0 10

Nb/YFIG. 9. Plot of Zr/TiO2 against M/Y. Names of volcanic fields after Winchester

& Floyd (1977).

Lo Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

390

TABII 7. COMPARISON wlTH HIGH.K AND MEDIUM.K OROGENTCANDESITES'

High-K

THE CANADIAN MINERALOGIST

CONTINENTAL

t l

1 ll - l

irqnsitionqi Atkoline

OCEANFLOOR

Tonalite appears to be significantly younger, andmay represent a later pulse of magmatism.

CoNcr-usroNs

Dioritic, tonalitic, and granodioritic plutons ofthe southeastern Cape Breton Highlands areinterpreted to be the roots of a volcanic arc, formedby late Precambrian subduction. The occurrence ofhigh-Al hornblende and magmatic epidote in thenorthwestern part of the area, the systematicdecrease in calculated pressures of hornblende

ISLAND500 /^Rc

Ti/Y300

200

100o.2 0.3 0.5 't.0

Nb/Y

2xNb KRDSTLDSOFODwcDs

Zr/4 YFrc. 10. Mafic samples (SiO2 less than 52Vo) plotted on the (a) TilY against Nb/Y

diagram of Pearce (1982) and (b) the Nb-Zr-Y triangular diagram of Meschede(1986).

Xtr+I

si02norAl203FeO!MnOMso

NaPK2oProg

55.870.82

r7.998.1 3o .17J . O 6

7.64t ? l

2 . 1 90.25

55.70.93

18.17.6o.184.07.43.4

0.31

80.010.54

17.317.030.143,448.782.871.730.1 5

60.5o,71

17.36.4o .12

6.7

0 .19

'Hloh-K baslc ard6tto and medlm"K acid andshe arc trom Gill (198i 1,@lculated to total I 00% volatilefreo. Avs€ge Gtsbore Flowage OwrtzDlorlte (GFODI and Ingonish Fivq ToElite llRTl 8re t,om Table 4,r@lcqlsGd to 100%, vol€dlelreo.

frn7 M.RB

HORNBLENDE- AND EPIDOTE.BEARING PLUTONS, CAPE BRETON 391

crystallization toward the southeast, and thepreservation of volcanic rocks in the southeasternpart of the area are consistent with the interpreta-tion that progressively deeper crustal levels arepreserved from southeast to northwest across thearea. The hornblende geobarometer may not besufficiently precise to determine the level ofexpo$ure in the northwest, but the co-occurrenceof high-Al hornblende and magmatic epidote,combined with the nature of the mineral as-semblages in adjacent metamorphic units, suggestthat the rocks formed at depths equivalent to ca.600 MPa (ca, 20 km), More detailed studies of theassociated metamorphic units are in progress tofurther constrain the P-T conditions and tectonicevolution of this area.

AcrNowI-SncEMENTS

Field and petrological studies for this projectwere funded by an NSERC Operating Grant toS.M. Barr. C.E.G. Farrow was supported by anNSERC 1967 Science and Engineering Scholarshipduring her graduate studies at Acadia University.We thank R.P. Raeside and A.S. Macdonald fortheir major role in the geological mapping of thesoutheastern Cape Breton Highlands, which madethe present study possible. We thank C.G. Barnesand an anonymous reviewer for their constructivecomments that led to significant changes andimprovements in the concept of this manuscript.

REFERENcES

Benn, S.M. & Resstor, R.P. (1989): Tectono-statigraphic terranes in Cape Breton Island, NovaScotia: implications for the configuration of thenorthern Appalachian orogen. Geology 17, 822-825.

& JaluEsoN, R.A. (1987): Geologicalmap of the igneous and metamorphic rocks ofnorthern Cape Breton Island. Geol. Surv. Can,,Open-File Rep.1594 (six sheets, scale l:50,000).

& MecooNaro, A.S. (1985): Geologyof the southeastern Cape Breton Highlands, NovaScotia. Geol. Surv. Con., Pap. 85-18, 103-109.

BruNnv, J.D. & HoLLaNo, T.J.B. (1990): Calcicamphibole equilibria and a new amphibole-plagioclase geothermometer. Contrib. Mineral,Petrol. 104,208-224.

BnowN, G.C., Tsonpe, R.S. & Wrnr, P,C. (1984):The geochemical characteristics of granitoids incontrasting arcs and comments on magma sources.J. Geol. Soc. London l4l.413-426.

DuNNrNc, G.R., Benn, S.M., Rarstoe, R.P. &JavresoN, R.A. (1990): U-Pb zircon, titanite, andmonazite ages in the Bras d'Or and Aspy terranesof Cape Breton Island, Nova Scotia: implicationsfor igneous and metamorphic history. Geol, Soc,Am. Bull. 702,322-330.

Evrmell, N.M., HavrlroN, P.J. & O'NtoNs, R.K.(1978): Rare-earth abundances in chondriticmeteorites. Geochim. Cosmochim. Acta 42, ll99'12 t2 .

FARnow, C.E.G. (1989): Petrography, MineralChemistry, and Geochemistry of Dioritic andTonalitic Plutons of the Southeastern Cape BretonHighlands, Nova Scotia, M.Sc. thesis, AcadiaUniv., Wolfville, Nova Scotia.

Grrr, J.B. (1981): Orogenic Andesites and PlateTectonics. Springer-Verlag, Berlin.

Helrvensrnolt, J.M. & ZsN, E-An (1986): Aluminumin hornblende: an empirical igneous geobarometer.Am. Mineral. 77, 1297-1313.

HeNsoN, G.N. (1980): Rare earth elements inpetrogenetic studies of igneous systems. Ann, Rev.Earth Planet. Sci. E, 371-406.

HoLLrsrEn, L.S., Gntssotu, G.C., PsrEns, 8.K.,SroweLL, H.H. & SIssoN, V.B. (1987): Confirma-tion of the empirical correlation of Al inhornblende with pressure of solidification ofcalc-alkaline plutons. Am. Mineral. 72, 231-239.

JonxsoN, M.C. & Rururnrono, M.J. (1989): Ex-perimental calibration of the aluminum-in-hornblende geobarometer with application to LongValley caldera (California) volcanic rocks. Geology17,837-841.

Lrars, B.E. (1978): Nomenclature of amphiboles.Can. Mineral. 16. 501-520.

MecooNan, A.S. & Bann, S.M. (1985): Geology andage of polymetallic mineral occurrences in volcanicand granitoid rocks, St. Anns area, Cape BretonIsland, Nova Scoria. Geol. Sum. Can,, Pap. ES'IB,1 t7 -124.

McCanruv, T.S. & Hasrv, R.A. (1976): Trace elementdistribution patterns and their relationship to thecrystallization of granitic melts. Geochim. Cos-mochim. Acta 40, l35l-1358.

MescHeoe, M. (1986): A method of discriminatingbetween different tlpes of mid-ocean ridge basaltsand continental tholeiites with the M-Zr-Ydiagram. Chem. Geol. 56, 207-218.

MrvasHIno, A. (1974): Volcanic rock series in islandarcs and active continental margins. Am. J. Sci,274,321-355.

392 THE CANADIAN MINERALOGIST

& SHroo, F. (1975): Tholeiitic and calc-alkalicseries in relation to the behaviors of titanium,vanadium, chromium, and nickel. Am. J. !ci.275,265-277.

Narnv, M.T. (1983): Phase equilibria of rock-formingferromagnesian silicates in granitic systems. z4z.

"/. Scl. 2E3,993-1033.

PEARCE, J.A. (1982): Trace element characteristics oflavas from destructive plate boundaries. /n An-desites: Orogenic Andesites and Related Rocks(R.S. Thorpe, ed.). Wiley-Interscience, New York(525-548).

Prrcurn, W.S. (1982): Granite type and tectonicenvironment. .In Mountain Building Processes(K.J. Hsii, ed.). Academic Press, New York(19-40).

(1987): Granites and yet more granites fortyyears on. Geol. Rundschau 76, 51-79.

ReEsrDE, R.P. & Benn, S.M. (1986): Stratigraphy andstructure of the southeastern Cape Breton High-lands, Nova Scotia. Mar. Sed. Atl. Geol. 22,264-277.

& - (1990): Geolocy and tectonicdevelopment ofthe Bras d'Or suspect terrane, CapeBreton Island, Nova Scotia. Can. J. Eorth Sci.27,l37 l -1381.

- & - (in press): Geology of the northernand eastern Cape Breton Highlands, Cape BretonIsland, Nova Scotia. Geol. Surv. Con., Pap.89-13.

REyNoLDs, P.H., JavmsoN, R.A., Benn, S.M. &RaEsroE, R.P. (1989): An NArl3eAr study of theCape Breton Highlands, Nova Scotia: thermalhistories and tectonic implications. Can. J. EarthSci. Xi,2081-2091.

RrcHerD, L.R. & CLARKE, D.B. (1990): AMPHIBOL- a program for calculating structural formulae andfor classifying and plotting chemical analyses ofamphiboles. Am. Mineral. 75, 421-423.

RurrrR, M.J., VaN Drt Laar, S,R. & Wylr-rE, P.J.(1989): Experimental data for a proposed empiricaligneous geobarometer: aluminum in hornblende atl0 kbar pressure. Geologlt 17, 897-900.

SrnEcrcrsEN, A. (1976): To each plutonic rock itsproper name. Earth-Sci. Rev.72, l-33.

TrnoLr, A.G. & PEencE, J.A. (1981): Petrogeneticmodelling of in situ fractional crystallization in thezoned Loch Doon pluton, Scotland. Contrib,Mineral. Petrol. 78, 196-207.

Tulrocs, A.J. (1979): Secondary Ca-Al silicates aslow-grade alteration products of granitoid biotite.Contrib. Mineral. Petrol, 69. 105-l 17.

(1986): Comment on "Implications of mag-matic epidote-bearing plutons on crustal evolutionin the accreted terranes of nofihwestern NorthAmerica" and "Magmatic epidote and itspetrologic significance". Geo logy 14, I 86-1 87.

VvHNeL, C.R., McSwrEN, H.Y., Jn. & Srrrn, J.A.(1991): Hornblende chemistry in southern Ap-palachian granitoids: implications for aluminumhornblende thermobarometry and magmaticepidote stability. Am. Mineral. 76, 176-188.

WrNcHrsrrn, J.A. & Frovo, P.A. (1977): Geochemicaldiscrimination of different magma series and theirdifferentiation products using immobile elements.Chem. Geol. 20. 325-343.

ZrN, E-An & Hauuansrnou, J.M. (1984a): Magmaticepidote and its petrologic significance. Geology 12,515-518.

& (t984b): Mineralogy and apetrogenetic model for the tonalite pluton at BushyPoint, Revillagigedo Island, Ketchikan l" x 2oquadrangle, southeastern Alaska. 1r The UnitedStates Geological Survey in Alaska: Accomplish-ments During 1982 (K.M. Reed & S. Bartsch-Winkler, eds.). U.S. Geol. Surv., Circ. 939,tt8-123.

& - (1986): Reply to comments on"Implications of magmatic epidote-bearing plutonson crustal evolution in the accreted terranes ofnorthwestern North America" and "Magmaticepidote and its petrologic significance". Geology14, 188-189.

& - (1988): Plumbing the depth ofplutons by magmatic epidote-hornblende associa-tion: a cautionary review and an example fromRound Valley pluton, western ldaho. Geol. Soc.Am., Abstr. Programs 20,475-476.

Received October 5, 1990, revised manuscript acceptedAugust 27, 1991.