American Mineralogist, Volume 73, pages 993-1006, 1988

Calcic amphiboles of mafic rocks of the Jeffers Brook plutonic complex,Nova Scotia, Canada

Gnoncr,c, h-PrpnnDepartment of Geology, St. Mary's University, Halifax, Nova Scotia B3H 3C3, Catada

Ansrucr

Amphiboles from mafic facies within a Late Proterozoic granodioritic complex plutonshow a variety of textures similar to those reported for amphiboles from porphyry-coppergranodiorites. Complexly zoned phenocrysts of hornblende have cores of pargasite, cli-nopyroxene, or actinolite and actinolitic hornblende, and some have rims of actinolitichornblende. Hornblende shows patchy compositional domains with gradational bound-aries and also occurs as crystal aggregates. Actinolite occurs as small euhedral crystals inpatches ofactinolitic hornblende. Electron-microprobe analyses show clusters ofdata de-fining four compositional types of amphibole-pargasite, hornblende, actinolitic horn-blende, and actinolite. Of these, actinolite shows coupled substitutions different from thoseofhornblende and pargasite, and variability in actinolitic hornblende falls on either trend.Early minerals crystallized under conditions of low oxygen fugacity, indicated by low Mgcontent of amphiboles and common sulfide inclusions. Later minerals crystallized underconditions of increasingly higher oxygen fugacity, indicated by hrgh Mg content of am-phiboles and abundance ofopaque oxides. There is no evidence that the development ofactinolite is associated with a discrete late hydrothermal event; rather, it is the result ofsubsolidus alteration of clinopyroxene. The development of these amphibole textures ap-pears more dependent on the evolution of volatiles than the chemical composition of themagma.

INrnooucrroN Origin of complex amphiboles in subalkaline infrusions

Geologic setting Amphiboles similar to those identified in the JeffersThe Jeffers Brook pluton is a complex dioritic pluton Brook pluton have been described from a number of plu-

of Late Proterozoic age in the Avalon terrane of the tons associated with porphyry-copper mineralization. InnorthernAppalachians(Fig.l;Pe-PiperandPiper,l98T). a detailed study of the Koloula Igneous Complex in theTheplutoniccomplexconsistsmainlyofgranodioritewith Solomon Islands, Chivas (1981) identified two types ofminor quartz diorite and tonalite (IUGS nomenclature of poorly zoned amphibole: a hornblende and an actinoliticStreckeisen, 1976).Alltheselithologiescontainabundant hornblende associated with magnetite. The amphibolesfine-grained enclaves of quartz diorite, diorite, and gab- did not show any regular core to rim zoning (except thatbro. A few gabbroic and dioritic bodies are found as pods the outermost rim in some crystals was actinolitic horn-either within or marginal to the pluton. blende). Most of the crystals contained patchy domains

The Jeffers Brook Pluton shows several similarities to with compositions ranging from hornblende to actinolite,calc-alkaline intrusions that are associated with porphy- but with a general trend of increasing Mg content as crys-ry-copper deposits and occur above subduction zones, tallization progressed. Similar patchy domains were firstnotably in the geochemistry of the predominant grano- described by Czamanske and Wones (1973) and have alsodiorite facies (Table 1), the evidence for a significant role been reported by Mason (1978) and Hendry et al. (1985).of volatiles, and the detailed mineral assemblages (dis- Chivas (1981) and Hendry et al. (1985) showed that therecussed below). It differs in that the more mafic facies was a general lack of amphiboles with Si between 7.2 andshow some alkaline characteristics such as relatively high 7.3.I-eake (1978) suggested that igneous amphiboles haveTi, Zr, Y, and Nb (Table l). a maximum Si of 7.3, and Chivas (1981) and Hendry et

The small mafic bodies and inclusions within the Jef- al. (1985) proposed that actinolitic hornblende with Sifers Brook granodiorite contain distinctive amphibole as- >7.3 crystallized under subsolidus conditions in the pres-semblages. This study describes the textural relationships ence of a fluid. These Mg-rich amphiboles were devel-and chemistry of co-existing calcic amphiboles present in oped under conditions of high oxygen fugacity and weresix samples from these minor gabbro and diorite facies preferentially depleted in elements that partition intowithin the Jeffers Brook pluton. "late-magmatic" hydrothermal solutions. Chivas (1981)

0003-004x/88/09104993$02.00 993

994

found that actinolite occurred as a clear hydrothermalalteration product in most rocks, and from a comparisonof mineralized and unmineralized rocks, he proposed thatexsolved fluid controlled the degree ofoxidation and thatamphiboles developed by replacement of pyroxene undersubsolidus conditions in the presence of this "late-mag-matic" exsolved fluid. In contrast, in a calc-alkaline quartzdiorite, Yamaguchi (1985) identif ied late-magmatichornblende that showed continuous compositional zon-ing from hornblende (Si : 7.24) to actinolite (Si : 7.6)with an outer rim of ferro-actinolitic hornblende com-position. The data of Chivas (1981) from Koloula andHendry et al. (1985) from Christmas, Arizona, alsoshowed continuous compositional trends from horn-blende to actinolite when various elements were plottedagainst Si; one exception was Mn at Koloula, whichreached a maximum in actinolitic hornblende.

Studies of actinolites and hornblendes principally inmetamorphic rocks have shown a compositional gap un-der conditions below the amphibolite facies (Robinson etal., 1982), interpreted by some authors as defining a sol-vus in the actinolite-hornblende system and by others asa disequilibrium feature. The present study provides nonew insights into the definition of the solvus in the horn-blende-actinolite system.

With the exception of limited data in Chivas (1981),the studies of igneous rocks cited above investigated am-

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

phiboles in granodiorites. In the present study, the rocksinvestigated are gabbros and diorites. Furthermore, theyshow some alkaline chemical characteristics, in contrastto the calc-alkaline rocks described by previous authors(compare with chemical analyses given by Mason andMcDonald, 1978). They thus provide a different set ofinformation on the origin of these amphiboles.

Some of the outstanding questions concerning the am-phiboles that are addressed in this study are as follows:(1) Are distinctive hornblende-actinolite relationships ofthe type described from calc-alkaline granodiorites alsofound in more alkaline mafic rocks? (2) What are theconditions under which more Mg-rich calcic amphiboleforms; in particular, is actinolite a result of later hydro-thermal activity, and are actinolitic hornblende and(magnesio) hornblende the result of subsolidus alteration,as proposed by Chivas ( I 98 lX (3) Is Si : 7 .3 the limit ofigneous hornblende, and is there a compositional gap forSi >7.2? (4) Is there evidence from chemical analyses ofcoexisting amphiboles for the nature of this composition-al gap? (5) Is there continuous chemical variation fromhornblende through actinolitic hornblende to actinolite?

MethodsThe Jefers Brook pluton was mapped in the field in order to

establish the distribution of the major rock facies. Thin sectionswere studied from a large number of outcrops. This mineralog-

tiii*r

x x xx x x

x x x2 ^ x " x " x x x

$;l;l'xftr granodiori te, tonal i te \ . , tr t t t\x.,x(,

A d io r i te , quar tz d io r i te i :

! Je f fe rs Group

//>> dykes, si l ts

\ -

$-/\II

II\

\

'xFrya|e;.\""r'?fie tx - 1 x - , x - * L ^ . ^ r x x x x \ x

11: , : , : , : , : x ] x , x ,x , x 5

t t t " t ^ t ^ t ^ t ^ x ^ x ' x Y x r x x x x

iir'riffi\ -{+-L'z I

. / ' - -

\, / \

/ \

l iFig. l. General geologic map of the Jeffers Brook plutonic complex. Inset shows location in northern Nova Scotia.

SAMPLE1112 t950

995

1965 1995

sio, 43.90Tio, 2.24Alro3 17.50(Fe,OJ" 13.10MnO 0.24MgO 6.81CaO 9.08Na.O 2.28K,O 1.66PrO. 0.12L.O.r 1.85

Total 98.78

48.26 49.741.18 2 .80

19.20 13.449.16 12.780.20 0.296.19 5.288.25 8.792.57 2.132.17 '1.52

0.19 0.442.23 1.70

99.60 98.91

491 45985 40

363 27720 4978 2869 1 61 3

4 0 018 24

243 14038 4623 42

266 35132 841 1.50 n.d.27.00 n.d.15.00 n.d.2.93 n.d.0.97 n.d.0.50 n.d.1.56 n.d.O.27 n.d.

52.40 54.301 . 1 4 1 . 1 1

18.20 17.701 0.10 8.680.19 0.233.55 3.556.65 6.553.82 3.801.69 1.580.30 0.251.85 1.62

99.89 99.37

577 52656 62

441 48634 26

164 701 5 93 47 1 5

16 2082 8341 229 1 0

169 16115 28n.d. 22.80n.d. 40.00n.d. 18.00n.d. 4.82n.d. 1.69n.d . 1 .10n.d. 2.37n.d. o.42

TABLE 1. Representative chemical analyses of maior rock faciesin the Jeffers Brook pluton

Sample: 1964 1950 3588 1 807 1993 2106

2291 182

c

R

c

Fig.2. Cartoon showing schematically the diferent texturalrelationships in amphibole phenocrysts (p : pargasite; h: horn-blende; ah : actinolitic hornblende; a : actinolite; c : clino-pyroxene); the figures in which each of these phenocrysts is il-lustrated; the abundance ofeach type in representative samples(C : common, R : rare, otherwise absent); and opaque mineralphases present in the representative samples.

degree ofalteration to sericite. Some cores appear highlyaltered, but many crystals have very fresh rims. Clino-pyroxene has been found in only a few samples (repre-sented by C1965). Quartz, if present, appears to havebeen the last mineral to crystallize, forming either ratherpatternless intergrowths with the feldspar or filling em-bayments (possibly due to corrosion) in the plagioclaselaths. One sample examined (C1950) contains distinctivelarge grains of amphibole (described below); this samplecontains only trace amounts of opaque minerals.

The quartz diorites (represented by sample Cl82) showa coarse-grained hypidiomorphic texture and consist ofplagioclase (55-700/o), amphibole (20-30o/o), biotite (l-100/o), opaques (l-2o/o), quartz (l-50/o), K-feldspar (-2o/o),and accessory apatite and zircon. Occasional chlorite ispresent.

The hornblende pegrnatite (sample C2294) is coarsegrained with a subophitic texture produced by partial in-clusion of feldspar in hornblende. Plagioclase (500/o), am-phibole (35o/o), qtartz (50/o), K-feldspar (50/o), and acces-sory sphene and opaques are present. Small amounts ofchlorite, epidote, and calcite are present in some samples.

Amphiboles

Description. The following amphiboles (using the no-menclature of Leake, 1978) can be optically distinguishedprimarily on the basis of color and pleochroism: actin-olite, actinolitic hornblende, (magnesio-) hornblende, and

BaRbSr

ZrNbThPbGaZnCuNi

CrLACeNdSmEuTbYbLu

/Vofej Samples are as follows: 1964-hornblende pegmatite vein fromeastern margin of pluton; 1950-diorite from western margin of pluton;3588-Gabbro from western margin of pluton; 18o7-diorite from north-west margin of pluton; 1993-quartz diorite from central part of pluton;2106-typical main granodiorite phase of pluton.

ical study is based on detailed examination ofsix representativesamples. Mineral analyses were made with a rpol-rus electronmicroprobe with four wavelength spectrometers and a TracorNorthern 145-eV energy-dispersive detector. Operating condi-tions were 15 kV at 5-nA beam current. Geological standardswere used. Data were reduced using a Tracor Northern ZAFmatrix correction program.

Fnrnocru.prrv

In this study, approximately 40 amphibole-bearingsamples were examined from diorite, quartz diorite, ahornblende pegmatite, and diorite inclusions and en-claves. Representative chemical analyses of the host rocksare presented in Table l. Textural relations between dif-ferent amphiboles have been studied optically. Fromthese, six representative samples have been selected fordetailed mineral-chemistry analysis. In thin section, thediorites (represented by samples Clll2, C1950, C1965,Cl995) show a hypidiomorphic granular (subophitic) tex-ture and consist of feldspars (60-650/o), amphiboles (25-300/o), biotite (2-l0o/o), opaques (principally ilmenite andmagnetite) (l-50/o), and accessory apatite and zircon. Theymay also contain epidote and chlorite (0-2o/o). Some ofthe plagioclase is zoned, and it shows a quite variable

46362

4472260718

1 889821 8

47728.30

20.0015.003.541.360.501.63o.28

63.900.54

16.305.080 .13't.87

4.464.341 .640 .171 . 1 6

99.59

6 1 545

37220

12685I

1 544

6o

693418.0040.0015.004.311 .590.602.400.45

Mn-it mt mt Dv mt mtpy i l bgy i l i l

py sro cpyPY

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

PHENOCRYST

@@-il.l\o_7

@"@(G\\1",m".@",'@"0poque minerols

996

Fig. 3. Photomicrographs showing textural relationships ofamphiboles in mafic rocks of the Jeffers Brook pluton. Numbersindicate analyses in Table 2. Abbreviations as in Fig. 2, e :

epidote. (a) Subhedral hornblende crystal with an inhomogenouscore made up mainly of actinolite and some patches of actinolitichornblende. Sample C182. Plane-polarized light; field of viewapprox. 2.7 mm. (b) Zoned amphibole crystal with green-brownpargasite core and green hornblende margin. Sample C I I I 2. Plane-polarized light; field of view approx. 1 mm. (c) Hornblende crystalwith an inhomogeneous core made up primarily of actinolite withsome patches of actinolitic hornblende. Sample C1965. Plane-

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

polarized light; field of view approx. 2.7 mm. (d) Hornblendecrystal with rim of actinolitic hornblende and interstitial epidote.Specimen Cl112. Plane-polarized light; field of view approx. Imm. (e) Colorless to very pale green actinolite prisms or laths,with occasional Fe-Ti oxide gtains, overgrown by discrete grainsof hornblende. Crystals of actinolitic hornblende occur at themaryin of the actinolite core and in the hornblende rim. SampleC1950. Plane-polarized light; field of view approx. 2.7 mm. (f)Clinopyroxene crystal overgrown by hornblende, with actinoliteprisms developing at the expense ofclinopyroxene. Sample I 965.Plane-polarized light; field of view approx. 1.08 mm.

pargasite. The hornblende is pleochroic dark green to ol-ive-green or yellowish. Actinolite is slightly pleochroicfrom pale green to colorless, and the actinolitic horn-blende is pleochroic from pale blue-green to green. Thepargasite is pleochroic dark brown to yellowish-brown.These optically distinct groups also show chemical dif-ferences, discussed in more detail below. The correlationbetween weight percent alumina and the amphibole typescan be generalized as follows: actinolite 0.5-3.50/o; actin-olitic hornblende 3.7 -4.7 o/o; hornblende 5. 0- I 0.00/o; andpargasite >l0o/0.

In the hornblende pegmatite (C2294), anhedral pargas-ite cores are in sharp contact with green hornblende rims(Figs. 2, 3b). The hc'rnblende rims are homogenous ex-cept that they show a slight zoning to higher Mg contenttoward the margin. The amphibole crystals in this rockappear fragmented, with secondary chlorite along thefractures and the edges ofthe grains. There is no trace ofactinolite along these fractures. Opaque oxides (manga-niferous ilmenite) are rare in this sample.

In the quartz diorites (sample Cl82), amphibole phe-nocrysts generally have a poikilitic texture and compriseeuhedral to subhedral hornblende with irregular actinolit-ic patches (Fig. 2), similar to the patchy domains de-scribed by Czamanske and Wones (1973).In a few crys-tals, cores with irregular patches ofactinolitic hornblendeand euhedral prisms of actinolite are in sharp but irreg-ular contact with homogeneous hornblende rims (Fig. 3a).The sharp, euhedral boundaries of this actinolite diferfrom those in the patchy actinolite domains in horn-blende.

In the diorites, hornblende occurs both as phenocrystsand microphenocrysts. Some of the phenocrysts show asystematic zoning with pargasite cores in sharp contactwith homogeneous or slightly zoned hornblende rims (Fig.3b). The cores of other phenocrysts consist of actinoliteprisms (Fig. 3c), in some cases also with patches of actin-olitic hornblende and hornblende. In sample Clll2,phenocrysts of homogeneous, slightly zoned hornblendeare rimmed by a discrete zone of actinolitic hornblende(Fie. 3d). In some samples (e.e., C1112), actinolite alsoforms aggregates of subhedral microphenocrysts and fillsinterstices between discrete hornblende phenocrysts. Ti-tano-magnetite is common in these diorites, but appearsto be a late-crystallized mineral.

One diorite sample (C1950; Figs. 3e and 4) containslarge composite amphibole grains, in which a core of ac-tinolite is rimmed by a number of discrete grains of horn-blende, actinolitic hornblende, and rare pargasitic horn-blende. The sharp boundaries between the hornblendegrains differ from those of the patchy domains describedby Czamanske and Wones (1973). The actinolite in thecore forms small euhedral to subhedral prisms or lathswith occasional sulfide inclusions. Crystals of actinolitichornblende are also found toward the margin of the ac-tinolite cores, and biotite inclusions occur in these crys-tals. Rare pyrite, chalcopyrite, and siderite are also pres-ent in this sample.

997

Fig. 4. Backscattered-electron scanning image ofpart ofthesame composite grain illustrated in Fig. 3e. Numbered spotsindicate analyses in Table 2. Bar is 0.1 mm.

In a few samples (represented by C1965, Fig. 3f), smallactinolite prisms are found in clinopyroxene phenocryststhat are rimmed by actinolitic hornblende. These rocksalso contain clinopyroxene microphenocrysts and smallnumbers of actinolitic hornblende microlites.

Interpretation. Individual composite phenocrysts showa range of amphibole compositions, from pargasite to ac-tinolite. In many cases, phenocrysts have cores with sharpcontacts with rims of different composition. Cores consistof clinopyroxene, pargasite, actinolite, and small crystalsof actinolite within actinolitic hornblende. Patchy com-positional domains in phenocrysts are similar to thosereported from granodiorites at Finnmarka (Czamanskeand Wones, 1973), Koloula (Chivas, l98l), and Christ-mas (Hendry et al., 1985). The complex amphiboles withdistinct cores of actinolite are similar to those describedfrom Koloula gabbros and diorites by Chivas (1981),where the presence of actinolite was believed to be as-sociated with late hydrothermal veining.

Textural data is typically difficult to interpret unequiv-ocally. Euhedral to subhedral homogeneous or zonedcrystals are taken to indicate magmatic crystallization.Rims of discrete crystals on phenocrysts generally indi-cate a reaction relationship with the melt. Subsolidusexsolution generally results in discrete exsolution lamel-lae. Patchy domains in amphiboles appear to result fromsubsolidus reactions in the presence of exsolved fluids

PE-PIPER:CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

998 PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

TaeLe 2. Chemical analyses and structural formulae of representative microprobe analyses of amphiboles from the Jeffers Brookpluton

182-10 182-11 182-12 1112-17 1112-18 1112-11 1112-8 1112-6 1965-14 1965-15 1965-2

sio,Tio,Al203FeO,",MnOMgoCaONaroKrO

Total

5l

TiratAlr6tAl

Fe3*Fe.MnMgCaNaKMg/(Mg + Fe,*)

47.95 48.551 .01 1.006.44 5.95

14.91 14.470.48 0.41

12.63 12.7412.22 12.710.97 0.740.47 0.50

97.08 97.06

7.099 7.1880.112 0 .1110.901 0.8120.223 0.2270.157 0.0391 .689 1.7520.060 0.0512.787 2.8111 .939 2.0160.278 0.2't20.089 0.0940.62 0.62

54.79 47.510.13 1 .061.22 7.57

10.98 15.260.38 0.39

16.33 1 2.6913.07 11.960.20 1.030.03 0.32

97.13 97.79

42.99 46.98 48.492.55 1.04 0.71

11.35 8.17 6.6516.58 15.88 14.470.38 0.40 0.36

10.74 12.55 13.851 1 .55 11.97 12.17

1 .68 1.10 0.920.44 0.38 0.08

98.25 98.38 97.70

52.92 48.650.25 1.012.92 6.35

13.05 13.950.50 0.38

16.61 13.2211.42 12.030.45 1.020.06 0.32

98.16 96.94

7.497 7.1650.027 0.1120.488 0.835

0.2680.625 0.1040.921 1 .6150.060 0.0473.507 2.9021.734 1 .8990.124 0.2910.011 0.0600.79 0.64

51.05 54.930.46 0.024.58 0.67

12.72 12.680.43 0.32

14.39 14.831 1 .99 13.980.69 0.040.06

96.36 97.46

7.469 7.9980.051 0.0020.531 0.0020.259 0.1130.0961.460 1 .5440.053 0.0393.137 3.2181.880 2.1810.196 0.01 10 .0110.68 0.63

Structural formulae based on O : 237.894 6.952 6.355 6.840 7.0380.014 0j17 0.283 0.114 0.0780.1 06 1 .048 1 .645 1 .160 0.9620.101 0.258 0.333 0.242 0.176

0.349 0.376 0.46s 0.4851 .323 1 .519 1.674 1 .468 1 .2710.046 0.048 0.048 0.049 0.0443.507 2.767 2.366 2.723 2.9962.018 1.875 1.829 1.867 1.8930.056 0.292 0.482 0.311 0.2590.006 0.060 0.083 0.071 0.015o.71 0.62 0.59 0.65 0.70

1950-3 1 950-4 1 950-5 | 950-6 1 950-7 1950-8 1950-9 1950-10 1950-11 2294-9 2294-7

sio,Tio,Al2o3FeO",MnOMgoCaONaroKrO

Total

siTir+rAlr6tAlFeP.Fe.MnMgCaNaK

46.95 43.371.05 1 .517.70 10.33

15.75 17 .190.32 0.34

12.48 10.611 1 .59 11.641.23 1.440.45 0.73

97 .52 97.16

6.911 6.4990.1 16 0 .1701.089 1 .5010.247 0.3240.369 0.4341.570 1.7200.040 0.0432.738 2.3701.828 1.8690.351 0.4180.085 0.1400.64 0.58

48.35 49.040.82 0.896.66 5.96't5.77 14.180.46 0.33

12.88 13.731 1 . 1 6 1 1 . 9 51.25 0.870.37 0.39

97.72 97.34

54.83 54.840.19 0.211 .38 1.64

1 1 . 10 11 . 210.37 0.32

17.15 17.'181 1.94 12.130.30 0.270.06 0.11

97.32 97.91

54.69 55.290.14 0 .160.96 1 .17

1 1 .58 11 .350.41 0.33

17.68 17.671',t.34 1 1 .540.22 0.320.11 0.07

97.13 97.90

7.778 7.8080.015 0.0170.1 61 0.192

0.0030.471 0.3310.906 1.0090.049 0.0393.747 3.7191.728 1.7460.061 0.0880.020 0.0130.81 0.79

49.25 41.980.88 3.155.79 12.80

14.85 11 .780.58 0.21

13.65 12.7811.84 11 .990.88 2.240.21 0.36

97.92 97.88

7.156 6.1890.096 0.3490.844 1 .8110.146 0.4140.399 0.1071.406 1 .3450.071 0.0262.956 2.8081 .843 1.8940.248 0.6400.039 0.0680.68 0.68

48.550.946.53

14.800.34

13.3411.720.950.51

97.68

Structural formulae based on O : 237.094 7.079 7.169 7.817 7.7770.103 0.090 0.098 0.020 0.0220.906 0.921 0.831 0.183 0.2230.219 0.229 0.196 0.049 0.0510.303 0.366 0.265 0.191 0.2041 .506 1 .565 1 .469 1 .133 1.'1260.042 0.057 0.041 0.045 0.0382.905 2.810 2.991 3.644 3.6311 .835 1 .751 1 .871 '1.824 1 .8430.269 0.355 0.247 0.083 0.0740.095 0.069 0.073 0.011 0.020

Mg/(Mg + FeF-) 0.64 0.67 0.76 0.76

A/ote: The positions 9! these analyses-except 1112-6, 1112-8, 1112-11, 2294-7,and 2294-9-are shown in Figures 3 and 4. The analyses 1112-6,1112-8,and1112- l l cometromhornblendeovergrowths(1112-11)onacoreof act inol i tepr ims(1112-6)wi thminoract inol i t ichornblendeandhornblende(1 1 1 2-8) similar to that illustrated in Fig. 3c. Analysis 2294-7 comes from the brown pargasite core and analysis 2294-9 comes from the green hornblendemargin of a zoned phenocryst, similar to that illustrated in Fig. 3b.

0.66

(Chivas, l98l). In general, retrograde metamorphism orsubsolidus alteration results in chemical disequilibriumand patchy development of minerals. Late-stage circulat-ing hydrothermal fluids are associated with alterationalong veins and cracks.

The textural data suggest that pargasite and clinopy-roxene were early phases to crystallize, forming the coresto phenocrysts with hornblende or, less commonly, actin-olitic hornblende rims. All of these phases appear fromtheir textures to have crystallized under magmatic con-ditions. Actinolite, or actinolite and actinolitic horn-blende, form crystalline aggregates at the expense of cli-nopyroxene in many phenocrysts. Crystal boundaries are

sharp; patchy compositional domains are not developed,and the textures are typical of subsolidus crystallization.Actinolite also occurs as patchy compositional domainswithin hornblende phenocrysts and exceptionally is foundin the interstices between hornblende crystals. There isno evidence for actinolite resulting from late hydrother-mal activity; it is not associated with fractures or veins.

In contrast, the textural evidence suggests that someactinolitic hornblende is ofigneous origin, because it oc-curs as homogeneous rims to hornblende phenocrysts andas microphenocrysts. Actinolitic hornblende also occursin an apparent subsolidus relationship with actinolite inthe cores of phenocrysts.

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

Porgosit

O 1

€Porgositicromblende

tr

a

Edenitichornbtende

o

D i o r i t e : ̂ C 1 1 1 2

" c1950r c1965' C1995

Quortz dior i te: o Cl82

Hornblende pegmoti te: . C2294

oct inol i t ic hornblende r ims

Ischermokihornblende

/ oI

II[^

lic \\ \ ^. ̂̂ -l *t

v v 6

-t jr{----^^

:- 'Hornbtende

-{--'-}

t

i t ichornbtende

\o,

/7'

I\.

/ : \\.r"' ]j\il/,

Actinolite

Fig. 5. Plot of (Na + K),^, versus Si. Nomenclature according to Leake (1978). Symbols indicate ditrerent rock samples. Dashedlines show concentration ofsamples in the hornblende, actinolitic homblende, and actinolite fields.

999

AT

OL6 ISi

The homcgeneous hornblende crystals, which in placeshave cores ofpargasite or actinolite and actinolitic horn-blende and have rims of actinolitic hornblende, appearto be ofigneous origin. Aggregates ofhornblende grainsprobably resulted from reaction ofclinopyroxene pheno-crysts with the magma. Hornblende grains with patchydomains of actinolite are similar to those described byChivas (1981) as resulting from subsolidus reaction withexsolved fluids.

Opaque minerals

Opaque minerals in three samples have been analyzedin detail by electron microprobe (Fig. 2). The clinopyrox-ene-bearing diorite (sample C1965) contains abundantopaque minerals, both as small independent crystals andas aggregates of similar crystals. Four types of indepen-dent crystals are distinguished: magnetite, ilmenite, chal-copyrite, and pyrite. Some of the oxide crystals containsmall inclusions of sulfide. The aggregates consist ofopaque oxides and sulfide crystals. The textural relation-ships in this sample suggest that amphibole co-precipi-tated with the opaque minerals.

The diorite with large amphibole grains (C1950) con-tains

r000

Structural formulae have been calculated from micro-probe analyses by using crystal-chemical constraints ac-cording to the method described by Stout (1972) andRobinson et al. (1982). First, the structural formulae werecalculated assuming total cations to be 13, exclusive ofK, Na, and Ca. The resulting calculated value for Fe3* isthe maximum consistent with stoichiometry. This cal-culation gave a range ofvalues for Fe3*/Fe.., from 0.12to 0.39 for actinolite, 0.01 to 0.37 for actinolitic horn-blende, and 0.01 to 0.36 for hornblende. Many of thesevalues are unreasonably high. Second, the structural for-mulae were calculated assuming total cations to be 15,exclusive of K and Na. The resulting calculated value forFe3* is the minimum consistent with stoichiometry. Theaverage Fe3+/Fe,o, ratios range from 0.01 for actinolite to0.09 for hornblende. These averages are probably a littlelow. We therefore use the mean values of the two meth-ods ofcalculation (13 and 15 cations) in discussing thecomposition of these amphiboles (Table 2).

The chemical classification of these amphiboles isshown in Figure 5. The classification ofLeake (1978) hasbeen used throughout this paper, except that the group ofsimilar amphiboles falling in the pargasite, pargasitichornblende, and tschermakitic hornblende fields are re-ferred to simply as pargasite. Two samples that just fallwithin the edenitic hornblende field are regarded as horn-blende. The plot in Figure 5 also shows that most ana-

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

oo

,

^^ : i o " "

. t _J ^ ^ - a o t

I . r- -t 'oo\ v ' d

r o o t o' _ \ r

- - - \ \ o -V'-- \ ' . \- - /i - ) . I Hornb lende' \ - l I .z

- - - ' "Ac t ino l i t i c ho rnb lendet o o \

O J- - r?r- o Act ino l i teo

o AT o.7

Fig. 6. Plot of rorAl vs. A-site occupancy, showing compositional gaps between actinolite, actinolitic hornblende, and hornblende.

Symbols as in Fig. 5.

IVAI

lyzed amphiboles cluster in either the hornblende or ac-tinolite fields. There are fewer analyses of pargasite oractinolitic hornblende composition. Actinolitic horn-blende analyses show a distinct cluster on plots ofAl andSi composition (Figs. 5-l l), whereas there is a substantialspread of pargasite analyses.

Amphibole types that have been distinguished optical-ly also cluster into groups on the basis of mineral chem-istry. Few chemical analyses fall in the compositionalranges between actinolite, actinolitic hornblende, andhornblende (Figs. 5 and 6), but there is no evidence foran actual compositional gap. Chivas (1981) also found asimilar distribution for amphiboles inferred as subsoli-dus. Yamaguchi (1985) found rapid compositionalchanges in zoned amphibole in the compositional rangeofactinolitic hornblende. There is also a paucity ofanal-yses falling between hornblende and pargasite, which isclearly seen in only in variation plots of Mg (Fig. l0) andA1,", (Fig. I l).

Plots (not reproduced here) similar to those in Figures5 to I l, in which analyses are distinguished on the basisof crystal form (fine and fibrous, coarser and euhedral),show that there are not systematic changes in mineralchemistry between different forms of the same mineral.In particular, there are no systematic differences betweenfibrous and prismatic actinolite.

All but one ofthe screened analyses ofactinolitic horn-

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

.i

{To { l v lo fs3 'o lJ i L

+ Fe3* + 2Tiat. Trend line shows predicted values for perfect substitution. Symbols

l00 l

2

IVAI

Fig. 7. Plot of tatAl against A-site ions + t6tAlas in Fig. 5.

blende are from crystals interpreted as subsolidus in or-igin, i.e., the crystals are either from patchy domains to-gether with hornblende or from cores containing smalleuhedral actinolite crystals. Other analyses of actinolitichornblende rims have been scrsened out because ofan-alytical totals as low as 95.1o/o. When these analyses areplotted (fields shown in Figs. 5-l l), they are indistin-guishable from other actinolitic hornblendes. These ac-tinolitic hornblendes show a compositional range similarto the outer parts of zoned hornblende-actinolite crystalsdescribed by Yamaguchi (1985). Ferro-actinolitic horn-blende of the type described by Yamaguchi was not rec-ognized.

Coupled substitutions

The data presented by Chivas (1981) and Yamaguchi(1985) indicate that there is continuous chemical varia-tion from hornblende through actinolitic hornblende toactinolite. This variation has been studied in the JeffersBrook pluton samples by examining some of the principalcoupled substitutions operative in the calcic amphibolesthrough the use of binary-element-variation plots (Figs.7-1 l) and through comparison of the data of Chivas

0lz0

(1981) and Yamaguchi (1985), recalculated in the samemanner as the analyses in this paper.

Substitution of t4lAl for Si in tetrahedral sites in all fourtypes of amphibole is dominantly compensated by sub-stitution of t6rAl, Fe3*, and Ti in octahedral sites and par-tial occupancy of the A sites by Na and I! as illustratedby the linear relationship in Figure 7. The slight deficien-cy in tatAl indicates that the excess A-site and octahedralcharge is not completely balanced by the t4tAl substitu-tion. This excess octahedral and A-site charge might bebalanced by substitution of Na for Ca in the M4 sites.

Edenite-type substitutions of I4rAl plus (Na + K),", forSi and A-site vacancy is illustrated in Figure 8 [in which,following Czamanske et al. (1981), richterite-type substi-tution eflects are removed by subtraction of Na,'.,]. Am-phiboles analyzed by Czamanske et al. (1981) showed al:l relation between t4lAl and A-site - Naro, with anintercept at about 0.6 l4rAl for 0.0 A-site - Na*.. Such atrend line describes well the variability in samples C2294and C I I 12 and also in samples anaTyzed by Chivas ( I 98 I )and Yamaguchi (1985). Other samples are best describedby a trend line with a similar slope, but an intercept ofabout 0.5 I4lAl. This trend line reflects less coupling withAl3+. Fe3+. and Ti in octahedral sites. Data for actinolite

i .

1002 PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

IVAI

4T -poM4 0.7Fig. 8. Plot of tarAl against A-site ions - Naly+'. Solid trend line from Czamanske et al. (1981). Dashed lines are trend lines for

Jefers Pluton data discussed in text. Symbols as in Fig. 5.

0

and actinolitic hornblende show a trend line with a 4:lrelation between t4rAl and A-site - NaM4, with an inter-cept at 0.0 lorAl.

A plot of t4rAl against (octahedral) Ti (Fig. 9) likewiseshows different trends for actinolite and actinolitic horn-blende compared with hornblende and pargasite. For thehornblende, there is a good linear correlation with a ratioof l:4 and a projected intercept at 0.5 r4tAl for zero Ti(Fig. 9), similar to the trend found by Czamanske et al.(1981) for Japanese granitoids. The analyses of Chivas(1981) also lie on this trend. Actinolite and actinolitichornblende, with low t4rAl and Ti show an indistinct lin-ear trend with a l:12 rutio passing through the origin.This trend may be in part an artifact of the method ofrecalculating analyses. Nevertheless, there is only limitedcoupling of rotAl with Ti in actinolite and actinolitic horn-blende, which is consistent with the trend shown in Fig-ure 8, in which there is little or no residual of rolAl sub-stitution unrelated to A-site occupancy. Samples analyzedby Yamaguchi (1985) contain very low Ti over a widerange of t4tAl values.

There is only a narrow range of values of Mg/(Mg +Fe'z*) (total Fe expressed as Fe2+) so that with uncertain-ties in partitioning Fe, caution must be used in inter-preting variations in Mg. There is, however, a strong neg-

ative correlation between Mg and I4rAl for hornblende,actinolitic hornblende, and actinoliteo whereas pargasiteshows a positive correlation (Fig. 10). Furthermore, in-dividual zoned crystals of hornblende show an increasein Mg/(Mg * Fe'?+) from core to rim; such zoning hasbeen interpreted to reflect increasing oxygen fugacity asthe magma became more siliceous (Czamanske andWones, 1973; Yamaguchi, 1985). This indicates that oxy-gen fugacity may have decreased during the crystalliza-tion ofthe pargasite cores, but increased steadily duringthe crystallization of hornblende.

The amphiboles described by Chivas (1981) and Ya-maguchi (1985) show similar trends of increasing Mg fromhornblende to actinolite. Compared to the Jefers Brookanalyses, the igneous amphiboles analyzed by Yamaguchi(1985) have lower Mg content, and the subsolidus am-phiboles analyzed by Chivas (1981) have higher Mg con-tent.

In summary, there are clear differences between cou-pled substitutions occurring in the analyzed samples ofactinolite and those occurring in igneous hornblende.There are also differences in some of the coupled substi-tutions in pargasite. According to the limited data avail-able, there are no differences between subsolvus actino-litic hornblende and that of apparent igneous origin on

aa

/ o

o . /

/ o

tr

- t l Y t

z 1 o

/t o , ' i

. to t a t ' - \' , 2 " t b

- . 2 ' - ' t - . ,! ^ j r\. "/' -t'

\ ! _ c -

+ o

" o E

/ o o

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS 1003

a

the rims of phenocrysts. Actinolitic hornblende can beregarded as showing trends that continue those seen ineither the hornblendes or the actinolites.

Cnvsr.qll.rzATloN HrsroRy oF THE AMpHTBoLES

Textural evidence for sequence of crystallization

The overall mineralogy of the pegmatites, diorites, andquartz diorites indicates that magma evolution duringcrystallization was initially constrained by crystallizationof plagioclase and clinopyroxene. The first amphibole toform was pargasite; later, hornblende crystallized. Ascrystallization proceeded, clinopyroxene became unsta-ble and was replaced by actinolite. Actinolitic hornblendeappears to have crystallized on the rims of hornblendegrains. The rare actinolitic hornblende microphenocrystsmay also be of primary igneous origin. In some samples,the outer parts of zoned phenocrysts are made up of in-dividual hornblende grains; these are presumably reac-tion products of clinopyroxene with the magma. Patchydomains of actinolite are probably the result of subsoli-dus reactions in the presence ofexsolved fluid.

This inferred crystallization sequence is similar to thatidentified by Chivas (1981) in the Koulala complex, butwith some important differences. In the Jeffers Brook plu-

o ,'z

"./

ton, actinolite (not hornblende and actinolitic horn-blende) is the principal product of alteration of clinopy-roxene. The textural evidence suggests that most of thehornblende and some of the actinolitic hornblende arethe result of magmatic crystallization. Alteration byexsolved late-magmatic fluids, to produce patchy do-mains, is of limited effect.

It might be questioned whether the actinolitic horn-blende rims are of igneous origin. Although their texturesupports this interpretation, chemically they are indistin-guishable from subsolvus amphiboles and have compo-sitions that many authors regard as beyond the range ofigneous amphiboles. Although it might be argued thatsome other distinct phase crystallized on the rims andwas subsequently altered to actinolitic hornblende, suchan interpretation cannot be applied to zoned actinolitichornblende of similar composition recognized by Ya-maguchi (1985), which has an outer rim of ferro-actin-olitic hornblende.

Pressure conditions of crystallization

HammarstromandZen (1986) have suggested that to-tal Al might be used as a geobarometer in igneous am-phiboles. They have demonstrated that there is a linearrelationship between I4lAl and Al,o, for the range 0.5 <

n / ,

t./o

o

)

0.3Fig. 9. Plot of tatAl against Ti, showing different l:4 trend for pargasite and hornblendes and 1: 12 trend for actinolite and actinolitic

hornblende. Symbols as in Fig. 5.

0 Ti

1004

Al o, < 2.6. Their calculations were made on the basis ofall Fe as FeO and 23 oxygens. A similar plot using chem-ical mineralogy as determined in this paper is presentedin Figure ll, in which their corrected regression line isshown. (The difference between the two estimates of Alcomposition is less than 4ol0.) This plot also shows theestimated crystallization pressure based on the Al o, con-tent and the regression equation of Hammarstrom andZen. This equation is clearly not applicable to the actin-olites and actinolitic hornblendes, which show coupledsubstitutions different from those ofthe hornblendes andwhich lie beyond the range of Al compositions investi-gated by Hammarstrom and Zen. The maximum pres-sure of about 7 kbar is found for the pargasite in thepegmatite. Hornblende pargasite cores in the diorite sug-gest pressures of 5-6 kbar, whereas pressures of less than4 kbar are estimated for the later hornblendes in all rocktypes.

Oxygen fugacity

The overall paucity of magnetite in diorites such assample C1950, the absence of magnetite as inclusions inearly-formed phenocrysts, and the occurrence of sulfideinclusions in some diorites are evidence of low oxygen

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

IVAI

005 0.6

MglFeT. MgFig. 10. Plot of tatAl against Mg/(Fe., * MB) to show diference in Fe/Mg ratio between actinolite-actinolitic hornblende and

hornblende-pargasite. Lines join analyses of inferred igneous amphiboles from the same zoned crystal. Symbols as in Fig. 5.

0.7

fugacity during the early stages of crystallization. The de-crease in Mg content in pargasites also indicates decreas-ing oxygen fugacity during early crystallization states.However, during late stages of crystallization, there is evi-dence of higher oxygen fugacity. Ilmenite in the horn-blende pegmatite is rich in Mn, and late-stage magnetite(with a high hematite component) appears in most sam-ples. There is a progressive increase in Mg content ofhornblendes.

CoNcr,usroNs

l. Mineralogical features of amphiboles seen in calc-alkaline granodiorites, such as patchy domains and theabundance of actinolite (Chivas, 1981), are also observedin more mafic alkaline rocks. Distinctive textures include(a) in the hornblende, patchy compositional domains thatlack sharp boundaries (similar to the domains describedby Czamanske and Wones, 1973), (b) euhedral actinolitecrystals in actinolitic hornblende and clinopyroxene coresto amphibole phenocrysts, and (c) outer rims to zonedphenocrysts consisting of discrete crystals of hornblende.Zoned hornblende crystals with sharp rims of actinolitichornblende are also seen.

2. The first minerals to crystallize were clinopyroxene

o o

D

l oI

v

t D t r - -f ; "

o

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS 1005

lvAl

Porgosite oo

Actinoliteo llo

E U U -ofr"

oto

Hornblende

AI I

lUo'Fig. ll. Plot of tarAl vs. Al,o,. Reference line from Hammarstrom artdZnn (1986), conected for different method of calculating

atomic formulae. Approximate crystallization pressures based on regression equation of Hammarstrom and Zen are also shown forhornblende and pargasite only. Symbols as in Fig. 5.

20

and pargasite. Hornblende crystallized from the magmaon the outer parts of complex phenocrysts. Textures alsoindicate that hornblende formed by reaction of clinopy-roxene with the magma. Actinolite and actinolitic horn-blende formed under subsolidus conditions from clino-pyroxene. In a few samples, patchy domains of actinolitein hornblende result from subsolidus alteration in thepresence of exsolved fluids. Rims of actinolitic horn-blende on some phenocrysts appear to be of igneous or-igin on the basis oftheir texture.

3. The early-formed minerals crystallized under con-ditions oflow oxygen fugacity (indicated by the presenceof sulfides rather than oxides), but later minerals crystal-lized under conditions ofincreasingly higher oxygen fu-gacity (indicated by opaque oxide minerals and the Mgcontents ofhornblendes). The similarity of the amphiboleassemblages in the alkaline diorites studied to those incalc-alkaline granodiorites reported in the literature isprobably the result of similar evolution of volatiles ineach magma type.

4. There is no textural evidence that actinolite formedfrom late hydrothermal activity, as proposed by Chivas

(1981). Actinolite has formed by subsolvus alteration ofclinopyroxene and possibly other phases. In general, it isabsent from cracks and interstices.

5. Chemically, actinolitic hornblende rims of probableigneous origin are indistinguishable from probable sub-solvus actinolitic hornblende. Subsolvus actinolites showdifferent coupled substitutions from igneous hornblendes:the actinolitic hornblendes could represent one end mem-ber of either of these trends. The actinolitic hornblenderims are of similar chemical composition to igneous ac-tinolitic hornblende described by Yamaguchi (1985).Typical hornblendes have Si

1006

7. The calcic amphiboles of the Jeffers Brook plutoncrystallized at a shallow crustal level at low pressure(mostly less than 4 kbar), but with high partial pressuresof water and high oxygen fugacity. One or more of thesespecial conditions may have been responsible for keepingthe calcic amphibole solvus at high temperature and thusfacilitating the crystallization of actinolitic hornblendeunder magmatic conditions.

AcxNowr,nncMENTS

This work was supported by an NSERC operating grant. D. E. Tumerand D.J W Piper have assisted with this work Microprobe analyses werecarried out at the regional microprobe centre at Dalhousie University.Reviews by F. C. Hawthorne, J. A. Speer, and the joumal referees gleatly

improved the original version ofthis manuscript.

RnrunnNcns crrso

Chivas, A.R. (1981) Geochemical evidence for magmatic fluids in por-phyry copper mineralization. Part I. Mafic silicates from the KoloulaIgneous Complex. Contributions to Mineralogy and Petrology, 78, 389-403.

Czamanske, G.K., and Wones, D.R. (1973) Oxidation during magmaticdifferentiation, Finnmarka Complex, Oslo area, Norway: Part 2, Themafic silicates. Joumal ofPetrology, 14, 349-380.

Czamanske, G K , Ishihara, S., and Atkin, S.A. (198 1) Chemistry ofrock-forming minerals of the Cretaceous-Paleogene batholith in southwest-ern Japan and implications for magma genesis. Journal of GeophysicalResearch. 86. 10431-10469.

PE-PIPER: CALCIC AMPHIBOLES IN PLUTONIC MAFIC ROCKS

Hammarstrom, J.M., and Znn, E. (1986) Aluminum in hornblende: An

empirical igneous geobarometer. American Mineralogist' 7 l' 1297 -1313

Hendry, DA.F., Chivas, A.R., Long, J.VP., and Reed, S.J.B. (1985)

Chemical differences between minerals from mineralizing and barren

intrusions from some North American prophyry copper deposits. Con-

tributions to Mineralogy and Petrology, 89,317-329'Leake. B.E. (1978) Nomenclature of amphiboles. Mineralogical Maga-

ztne. 42.533-563.Mason, D.R. (1978) Compositional variations in ferromagnesian minerals

from porphyry copper-generating and barren intrusions in the Western

Highlands, Papua New Guinea. Economic Geology, 73, 878-890

Mason, D.R., and McDonald, J.A (1978) Intrusive rocks and porphyry

copper occurrences ofthe Papua New Guinea-Solomon Islands region:

A reconnaissance study Economic Geology, 73,857-877 -

Pe-Piper, G., and Piper, D.J.W (1987) The pre-Carboniferous rocks of

the western Cobequid Hills, Avalon zone, Nova Scotia. Maritime Sed-

iments and Atlantic Geology, 23, 4148.Robinson, P., Spear, F.S, Schumacher, J C., Laird, J., Klein, C., Evans,

B.W., and Doolan, B.L. (1982) Phase relations of metamorphic am-

phiboles: Natural occurrences and theory. Mineralogical Society of

America Reviews in Mineralogy, 98, l-227

Stout, J.H. (1972) Phase petrology and mineral chemistry ofcoexisting

amphiboles from Telemark, Norway. Journal of Petrology, 13, 99- I 45.

Streckeisen, A (1976) To each plutonic rock its proper name' Eadh-

Science Reviews, 12, l-33.Yamaguchi, Y. (l 985) Hornblende-cummingtonite and hornblende-aclin-

olite intergowths from the Koyama calc-alkaline intrusion, Susa,

southwest Japan. American Mineralogist, 70, 980-986.

MANUscRrpr RECEIVED Aucusr 17, 1987MnNuscrpr AccEprED ApnIr- 26, 1988