Fibrolite in contact aureoles of Donegal, Ireland DBnnrr.r ... › ammin › AM72 ›...

American Mineralogist, Volume 72, pages 240-254, 1987

Fibrolite in contact aureoles of Donegal, Ireland

DBnnrr.r. M. KrnnrcrDepartment of Geosciences, The Pennsylvania State University,

University Park, Pennsylvania 16802, U.S.A.

AssrRAcr

Fibrolite was examined in the Ardara and Fanad contact aureoles of Donegal, Ireland.Textural evidence suggests that fibrolite formed in the later stages of contact metamor-phism. Garnet-biotite geothermometry and garnet-plagioclase-AlrSiOr-quartz geobarom-etry in the Ardara aureole suggest that fibrolite formed metastably at P-?"conditions withinthe andalusite stability field. Most fibrolite is considered to have formed from the decom-position of biotite through a reaction involving acidic volatiles emanating from the adja-cent intrusion. Nucleation of fibrolite, rather than the stable polymorph andalusite, mayreflect the relatively high activation energy in transforming tetrahedrally coordinated Alof the biotite structure into the fivefold coordination of Al in andalusite. In contrast tofibrolite, the isograd marking the first appearance of coarse-grained sillimanite appears tocorrelate well with the andalusite : sillimanite equilibrium. Because fibrolitized biotite iscommon in metapelites of contact and regional metamorphism, interpretation of fibroliteas an equilibrium phase synonymous with coarse-grained sillimanite must be viewed withcaution.

INrnooucrtoN dislocations is insufficient to perturb the stability rela-The nature and significance of fibrolite, a common con- tions of fibrolite compared to coarse-grained sillimanite.

stituent of medium- to high-grade metapelites, has long A potentially significant excess energetic contributionbeen an enigma. Holdaway (197 1,p. 103) concluded that to fibrolite arises from surface free energy. The positive"fibrolite is probably a protosillimanite which crystallizes surface-free-energy contribution will have the effect of de-rapidly from an irreversible mineral reaction. It may shbilizing fibrotite compared to coarse sillimanite. Nit-well contain excess silica and water and may have Al-Si kiewicz and Kerrick (unpub. ms.) provide a detailed ex-disorder." As shown by Bell and Nord (197 4) for a sam- amination of the surface-free-energy contribution ofple of fibrolite from Brandywine Springs, Delaware, fi- fibrolite.brolite aggregates may indeed contain excess silica (as In both contact and regional metamorphic regimes, thequartz). From their neutron-difraction study, Finger and shtus offibrolite as an equilibrium phase has been ques-Prince (1972) suggested that some Al-Si disorder may tionedbyseveralauthors(PitcherandRead, 1963; Pitch-exist in Brandywine Springs fibrolite. Navrotsky et al. er, 1965; Chinner, 1966;Znn, 1969; PitcherandBerger,(1973) carried out molten oxide calorimetric measure- 1972: Vernon, 1979; Speer, 1982). Because the thermalments on a sample of heat-treated Brandywine Springs regime and history of contact aureoles are, in general,fibrolite and concluded that Al-Si disorder is insignificant simpler than regional parageneses, this paper focuses at-at temperatures below about 800"C. Experiments in the tention on the role of fibrolite in thermal aureoles. How-author's laboratory revealed that the P-7 conditions of ever, some comments regarding fibrolite in regionallythe andalusite : sillimanite equilibrium are identical re- metamorphosed pelites are given at the end of this paper.gardless of whether coarse-grained sillimanite or fibroliteis used as starting material (r"rrict and Heninger, 1984). ANA.L\"rrcA.r- METHODS

These results, coupled with the fact that coarse-grained Analyses of minerals were carried out with an automated ETEcsillimanite appears to be fully ordered (Robie and Hem- electron-microprobe analyzer at Pennsylvania State University.ingway, 1984; Peterson and McMullan, 1986), s rggest Data reduction followed the Bence-Albee scheme using a factorsthat fibrolite may also be ordered. However, Salje (1986) computed according to the procedure of Albee and Ray (1970).

concluded that the heat capacity of fibrolite is signifi- Each m^ineral analysis reported in this paper represents an av-

cantly larger than that of coarse sillimanite. It";";;;l;, erage of two to five individual crystals; the composition of in-

the resultanr entropy contributions yield a ""bi;;l;j

i::t-ty] crystals was obtained bv averaging four to six point

dalusite : fibrolite equilibrium thai is at a considerably "ttfilfrl;, biotite, and plagioclase were analyzed for geother-higher temperature than the equilibrium involving an- mometry and geobarometry (Tables l-3). In oider to utilize gar-dalusite and coarse-grained sillimanite (Salje, 1986). net-biotite geothermometry, the garnet analyses represent aver-

Kerrick (1986) has shown that strain energy arising from ages of rim compositions.

0003{04x/87/03044240$02.00 240

KERRICK: FIBROLITE IN CONTACT AUREOLES

TABLE 1. Ardara aureole samples: Electron-microprobe analyses of garnet

ARD-69-4 ARD-69-6 ARD-69-7 ARD-69-21 ARO-69-23 ARD-69-25 ARD-69-34 ARD-69-39 ARD-69-45 ARD-69-63 ARD-69-658

24r

sio,Alro3Tio,MgoFeOMnOCaONaroKrO

SiAITiMgFeMnCaNaK

Xo,Xpv

x",X"o

38.26 37.9421.33 21 .060.08 0.172.03 2.20

32.83 31.455.66 6.351.48 1 .640.00 0.010.03 0.05

1 01 .70 100.85

3.03s 3.0331.996 1 .9850.005 0.0100.240 0.2632.178 21020.380 0.4300.126 0 .140.q001 0.002S,004 0.005

38.26 38.0921.07 20.790.04 0.182.29 2.27

31.23 32.156.01 5.001 98 1 .810.05 0.010.05 0.04

100.98 100.34

3.047 3.0531.979 1.9650.002 0.0110.272 0.2712.080 2.1550.405 0.3390.169 0.1550.008 0.0010.006 0.004

37.2721.450.021.98

31.896.451.490.000.01

100.56

3.058 3.009 2.9981.960 2.011 2.0340.003 0.003 0.0010.232 0.358 0.2372.366 2.420 2.1450.218 0.067 0.4390.119 0 .113 0 .1280.001 0.003 0.0000.003 0.001 0.0017.961 7.985 7.985

0.806 0.818 0.7270.079 0j21 0.0800.041 0.038 0.044o.o74 0.022 0.149

37.17 37.6624.56 21 250.12 0 .171.22 1.52

26.20 27.255.76 5.155.68 6.930.00 0.000.03 0.05

100.74 99.98

37.86 37.89 37.66 37.8021 .37 21 .74 20.48 21.420.05 0.07 0.05 0.051.78 2.26 1.92 3.02

34.95 28.49 34.84 36.353.78 7.04 3.17 0.991 .62 2.93 1.37 1 .330.00 0.06 0.00 0.020.03 0.02 0.03 0.01

101 .44 100.50 99.52 100.99

Formula basis: 12 oxygens3.020 3.0182.010 2.0400.003 0.0040.211 0.2672.332 1.8990.255 0.4750.138 0.2500.000 0.0090.003 0.0027 .973 7.964

2.9242.2680.0070.1431.7250.3850.4810.0000.003

3.0122.0040.0100.181't.8220.3490.5940.0000.0057.9787.9367.964

o.7450.0820.0430.130

7.968

0.7160.0890.0480.1 46

7.968

0.71 10.0930.0580.1 39

7.956

0.738 0.794 0.6560.093 0072 0.0930.053 0.047 0.0870.116 0 .087 0 .164

0.634 0.6180.053 0.0610.173 0.2010.140 0 .119

ARD-85-1A ARD-85-18 ARD-85-1C ARD-85-2A ARD-85-15 ARD-85-16 ARD-85-17 ARD-8!26AARD-8531AARD-8S324 ARD-8533

sio,AlrosMgoFeOMnOCaO

SiAIMgFeMnCa

X^,X"yXu,X"o

37.81 37.91 37.1821.38 20.69 21.013.59 3.34 3.22

36.37 35.64 36.101.05 0.82 1 .311.05 1 .97 1 .38

1 01 .25 100.37 1 00.20

3.000 3.033 2.9922.001 1.951 1 .9940.425 0.398 0.3862.414 2.385 2.4300.071 0.056 0.0890.089 0.169 0.1 198.000 7.992 8.011

0.805 0.793 0.8040.14:2 0.132 0.1280.030 0.056 0.0390.024 0.018 0 030

38.13 37.1020.63 21.173.24 2.51

36.24 29.261 .13 8 .411.44 2 .12

100.81 1 00.57

37.58 37.9121 .38 21.482.52 2.63

31 .61 31.476.35 5.801.80 2.03

101 .24 101 .32

38.12 37.4520.80 21.272.57 2.32

26.72 30.384.77 6.837.36 2.90

100.34 1 00.34

3.027 3.0081.947 2.0140.304 0.2781.774 2.0410.321 0.4650.627 0.1,808.000 7.985

0.586 0.6890.101 0.0940.207 0.0610.106 0.157

37.69 37.5621.17 20.582.16 2 .16

34.03 30.384.53 8.701.73 1 .20

101 .30 1 00.58

3.011 3.0291.995 1.9570.257 0.2592.274 2.0490.306 0.s940.148 0.1047.991 7.992

0.762 0.6820.086 0.0860.050 0.0340.103 0.198

Formula basis: 12 oxygens3.043 2.983 2.997 3.0101.941 2.007 2.011 2.0110.385 0.300 0.300 0.3112.418 1.968 2.108 2.0890.077 0.573 0.429 0.3900.123 0.183 0.154 0.1737 .987 8.014 7.998 7.985

0.805 0.651 0.705 0.7050.128 0.099 0.100 0.1050.041 0.061 0.051 0.0580.026 0.189 0.143 0.132

The AIrSiO, phases were analyzed (Table 4) in order to eval-uate the role of minor-element solid solution, with particularemphasis on the effect ofsolid solution on the andalusite : fi-brolite equilibrium. Microprobe analyses of fibrolite were ob-tained in selected areas that contained minirnal contaminantphases. Quartz was found to be the major contaminant of fi-brolite. Areas containing quartz were readily identified by thecharacteristic blue cathodoluminescence ofthis phase and/or bySiO, contents in excess of the AlrSiO, stoichiometry. The amountof quartz included within the fibrolites examined in this studywas computed assuming a linear relation between the SiO, con-tent of stoichiometric AlrSiOs (SiOz : 37.09 wt9o) and quartz(SiOr: 100 wto/o). In essence, the SiOr in excess of 37.09 utto/owas assigned to quartz. From the SiO, values offibrolite in Table4, the quartz contamination was a maximum of 1.4 wt% (sampleARD-85-16); fibrolites analyzed in most other samples containless than I wt% quartz.

DoNnc.cl AUREoLES

General geologic setting

The geology ofthe Donegal region ofnorthern Irelandhas been thoroughly reviewed by Pitcher and Berger(1972), and the chronology of the sedimentary, tectonic,and igneous history has been summarized by Powell andPhillips (1985). In brief, the country rocks of Donegalconsist of Dalradian metasedimentary and meta-igneousrocks that range in age ftom late Precambrian to MiddleCambrian. This group of rocks subsequently underwentmajor deformation (the Caledonian orogeny) during LateCambrian to Late Ordovician time. Intrusion ofthe Don-egal granites is considered to have occurred at about 390-405 Ma (Halliday et al., 1980).

242 KERRICK: FIBROLITE IN CONTACT AUREOLES

TneLe 2. Ardara aureole samples: Electron-microprobe analyses of biotite

ARD-69-4 ARD-69-6 ARD-69-7 ARD-69-21 ARD-69-23 ARD-69-25 ARD-69-34 ARD-69-39 ARD-69-45 ARD-69-63 ARD-69-658

sio,Al,03Tio,MgoFeOMnO

NaroKrO

35.9120.322.008.07

21.320. ' t40.000.187.19

95.1 3

5.4292.5711.051o.2281 .8192.6960.0185.812

0.0000.0541.3861.441

0.3140.465

0.301 0.3640.489 0.420

0.356 0.3970.395 0.396

sirvAlvrAlTiMgFeMn2

NaK)X,nXonn

0.0000.0671.5331.600

0.3350.463

sio,Alro3Tio,MgoFeOMnOCaONaroKrO

5 lIVAI

'vAl

TiMgFeMn2

CaNaK

X"nXo""

37.31 36.0021 .O7 19.801.39 1.849 . 1 6 1 0 . 1 1

1 8 1 1 1 7 9 80.07 0.160.00 0.010.14 0 207.36 106

94.61 94.16

5 559 54482 441 2.5521.260 0.9800.156 0 .2102.034 2.2812.257 2.2760.009 0.0215.715 5.767

0.000 0.0020.040 0.0591 399 1 5561 .439 1.617

35.36 34.79 36.21 35.651 8.55 20.18 19.79 18.982.37 1.98 1.91 2.477.65 7.69 9.25 7.85

21.74 22.27 18.76 21.630.16 0 .04 0 .14 0 .120.00 0.00 0.02 0.000.09 0.14 0.22 0.217.83 7.36 8.31 7.35

93.76 94.45 94.61 94.26

Formula basis: 22 oxygens5.482 5.345 5.476 5.4702.518 2.655 2.524 2.5300.874 0.999 1.008 0.9050.276 0.229 0.218 0.2851.768 1.761 2.073 1.7972.819 2.861 2.386 2.7760.022 0.006 0.018 0.0155.758 5.857 5.703 5.777

0.000 0.000 0.003 0.0000027 0.041 0.063 0.0631 .549 1.443 1 .607 1 .4391 576 't.484 1 .673 1 .502

35.17 31.07 34.8419.44 21 .50 19.962.37 0.89 1.427.78 12.06 10.72

20.74 20.53 18.320.15 0.1 1 0.070.01 0.03 0.000.11 0.04 0.238.34 3.18 6.35

94.11 89.41 91.91

5.419 4.907 5.3682.581 3.093 2.6320.950 0.937 0.9950.275 0.104 0.1651.787 2.862 2.4642.673 2.730 2.3630.019 0.015 0.0095.704 6.648 5.995

0.0020.0321.6381.673

0.3140.470

0.006 0.0000.013 0.0700.628 1.2480.647 1.317

0.428 0.4110.411 0.395

35.2320.132.248.21

20.320.080.000.357.97

94.53

5.3772.6230.9990.2571.8682.5940.0105.728

0.0000.104't.552

1.655

0.3080.491

0.3120.482

0.3270.454

ARD-85-1A ARD-85-18 ARD-85-1C ARD-85-2A ARD-85-15 ARD-85-16 ARD-85-17 ARD-85-26AARD-85-31AARD-85-32AARD-85-33

35.54 35.591 9.95 19.372.20 2.509.10 10.27

20.07 18.350.03 0 040.00 0.000.34 0.267.88 7.86

95.11 94.24

5.377 5.3942.623 2.6060.936 0.8560.2s0 0.2852.0s3 2.3202.539 2.3260.004 0.0065.782 5.793

0.000 0.0000.100 0.0771 .521 1 .5191.622 1 .596

0.355 0.4010.439 0.402

35.51 34.9719.89 20.052 j2 2 .588.60 8 .10

21.18 20.420 00 0.070.00 0.000.23 0.217 93 8.99

95.46 95.39

5.381 5.3262.619 2.6740.934 0.9280.242 0.2961.942 1.8392.684 2.6010.000 0.009s.801 5.672

0.0000.0631.7471 .810

0.3250.459

36.05 35.21 36.5119.51 20.11 20.031.67 1 .92 1 .57

11.10 10 .39 10 .8518.01 18.02 '17 370.1s 0.08 0.070.02 0.00 0.000.18 0.30 0.327.s6 7.95 7.95

94,25 93.98 94.67

Formula basis: 22 oxygens5.437 5.347 5.4662.563 2.653 2.5340.907 0.948 1.0020.190 0.220 0.1772.495 2.352 2.4212.272 2.289 2.1750.019 0.01 0 0,0095.883 5.819 5.784

35.32 34.7919.47 19.741.90 2.498.51 7.93

20.97 20.980.01 0.130.00 0.000.20 0.207.70 6.94

94.04 93.38

5.427 5.424 5.3722.573 2.576 2.6280.970 0.949 0.9650.1 80 0.219 0.2892.411 1.947 1.8252.274 2 693 2.7090.016 0.002 0.0415.850 5.811 5.829

0.000 0.000 0.0000.062 0.060 0.0601.482 1 .508 1.3671 .543 1 .568 1.427

0.0030.0521.4541.509

0.000 0.0000.088 0.0941.540 1 .5181.628 1 .612

35.5619.491.40

14.1414.500.150.000.206.96

92.40

5.3702.6300.8410.1593.1841.8320.0206.036

0.0000.0591.3401.398

0.5290.304

36.0819.981.59

10.7518.070.130.000.217.72

94.53

0A25 0.405 0.4190.387 0.394 0.377

0.413 0.3350.390 0.464

0.3150.468

Ardara aureole

The Ardara pluton (Fig. l) represents a forcibly em-placed, diapiric intrusion (Holder, 1979; Sanderson andMeneilly, 1981). This pluton consists of an outer cara-pace of quartz monzodiorite (230-900 m thick) enclosinga $anodiorite core (Akaad, 1956a). Pitcher and Berger(1972, p. 182) considered that these two intrusive phaseswere closely spaced in time. The highly foliated structureof the outer qvartz moruodiorite, coupled with strongdeformation ofthe country rocks, provides clear evidencefor the diapiric character of the Ardara pluton.

The northern Ardara aureole, which is the focus of thepresent investigation, consists ofa pelitic horizon (locallyreferred to as the Clooney Pelitic Group) and an adjacentlimestone unit (the Portnoo Limestone). Pitcher and Ber-ger (1972) correlated these units with the respective Up-per Falcarragh Pelites and the Falcarragh Limestone,which are members of the Creeslough Succession thatforms the country rock for most of the Donegal intrusions(see Pitcher and Berger, 1972, Fig. l-l). In detail, thepelitic horizon ofthe northern Ardara aureole consists ofinterlayered aluminous pelites and semipelites. knses and

KERRICK: FIBROLITE IN CONTACT AUREOLES 243

sio? 57.94 61.44At,o3 26.15 23.90FeOCaO 8.80 6.08Na,O 6.13 7.44K.O 0.02 0.03

99.04 98.89

57.92 67.7226.42 24.120.059.90 6.745.75 6.97N.D. N.D.

100.1 1 100.55

56.77 62.1227.39 23.47N.D. N.D.9.05 4.916.27 8.78

99.49 99.29

TneLe 3. Ardara aureole samples: Electron-microprobe analy-ses of zoned plagioclase

ARD-85-18 ARD-85-1C ARD-85-10

rick and Speer (1987) have concluded that the anda-lusite - sillimanite isograd reaction represents a closeanalogue to the univariant andalusite : sillimanite equi-librium involving pure phases. Pitcher and Read (1963)concluded that the metamorphic zone characterized byfibrolite is relatively narow; however, fibrolite is actuallyabundant in andalusite-bearing rocks, extending to a dis-tance ofabout 400 m from the intrusive contact (Fig. lB).

Within the andalusite-bearing rocks, garnet occurs assmall, rounded porphyroblasts. In contrast, within thesillimanite zone, garnet occurs as larger crystals that are,in several specimens, irregular in shape. The garnet tex-ture, as well as the overall texture of the rock, suggeststhat the transformation from the andalusite to sillimanitezones was accompanied by a wholesale textural reconsti-tution.

In several samples, andalusite porphlroblasts have pinkcores surrounded by colorless rims. This zoning is quitecommon in porphyroblastic andalusite (Okrusch and Ev-ans, 1970). Electron-microprobe analysis (Kerrick andSpeer, 1987) shows that the cores contain higher con-centrations of Fe compared to the rims; however, thereappear to be no significant differences in the concentra-tions of other minor elements between cores and rims.

Replacement reaction has been significant in manysamples from the northern Ardara aureole. Andalusitehas undergone varying degrees of replacement by mus-covite. In many samples, replacement is complete, yield-ing coarse-grained ovoids of interlocking muscovites sur-rounded by coarse-grained biotite. In several specimens,well-developed muscovite porphyroblasts have replacedbiotite (Fie. 2A). The temporal relationship of these por-phyroblasts to the muscovite that has replaced andalusiteporphyroblasts is unknown.

SiAIFe

NaKX^"Xoo

Xo,

Formula basis: 8 oxygens2.611 2.748 2.589 2.755 2.555 2,7681 .389 1 .260 1 .396 1.248 1 .452 1 .231

0.001 0.0010.425 0.291 0.474 0.316 0.436 0 2340.536 0.645 0.497 0.593 0.546 0.7580 001 0.002 N.D. N.D.0.442 0.310 0.488 0.348 0.444 0.2360.557 0.688 0.512 0.652 0.556 0.7640 001 0.002 N.D. N.D.

pods of metadolerite are common in a zone extendingabout 200 m from the plutonic contact (Akaad, 1956b).

Figure I shows the metamorphic zones in the northernArdara aureole. Accordingly, the aureole is subdividedinto an outer zone characterized by andalusite + kyanite,an intermediate zone with andalusite, and an inner zonecontaining coarse, prismatic sillimanite (hereafter re-ferred to as sillimanite). The extent of the andalusite +kyanite zone was not confirmed in this study, as kyanitewas identified in only one sample (ARD-85-30). If fi-brolite is excluded, the sillimanite isograd is abrupt andwell defined. From electron-microprobe analyses of mi-nor-element contents of andalusite and sillimanite, Ker-

Trele 4. Ardara aureole samples: Electron-microprobe analyses of aluminum silicates

ARD- ARD-85-1A 85-2A

F F

ARD- ARD-85-3E 85-5A

F T

ARD. ARD-85-10 85-15

F A

ARD-85-16 ARD-85-17 ARD- ARD- ARD.85-32A 69-34 69-658

sio, 36.95 37 52At,o3 61.66 62.59TiO, 0.01 n.d.Fe,O3 0.27 0.38CrrO3 0.15 0.02MnO 0.05 0 01MgO 0.00 0.07> 99.09 100.59

1 007 1.0071 981 1.9810.000 n.d.0.006 0.0080.003 0.0000.001 0.0000.000 0.0032.998 2.999

0.010 0.01 10.990 0.989

37.65 37.5461.63 61 .1 1

n.d. 0.070.1s 0 .180.01 0.020 02 0.030.02 0.01

99 48 98.96

1.020 1.0231 .969 1.963n.d. 0.0010.003 0.0040.000 0.0000.000 0.0010.001 0.0012.994 2.992

0.004 0.0070.996 0.993

36.94 37.8062.52 61.380.00 0.000.17 0.620.00 0.000.00 0.000.07 0.27

99.70 100.07

1 .000 1.0201.995 1.9530.000 0.0000.003 0.0130.000 0.0000.000 0.0000.003 0.0113.001 2.997

0.006 0.0230.994 0.977

,(: 0.983

36.88 37.05 37.0961.81 61.83 62.240.03 0.06 0.o20.29 0 .19 0 .190.03 0 .12 0 .130.01 0.01 0.020.05 0.o2 0.04

99.10 99.28 99.73

1.005 1.007 1.0041.985 1.982 1.9860.001 0.001 0.0010.006 0.004 0.0040.001 0.003 0.0030.000 0.000 0.0010.002 0.001 0.0022.999 2.998 2.999

J I

AITiFe

MnMg

X,"Xu'at

37.51 36.27 36.66 37.9662.05 62.73 62.53 60.720.01 0.05 0.05 0.000.18 0.26 0.26 0.430.13 0.04 0.03 0.010.02 0.01 0.01 0.030.00 n.d. 0.02 O.22

99.90 99.36 99.56 99.37

Formula basis: 5 oxygens1 .013 0.986 0.994 1 .0311.976 2.011 2.000 1 .9440.000 0.001 0.001 0.0000.004 0.005 0.005 0.0090.003 0.001 0.00{ 0.0000.000 0.000 0.000 0.0010.000 n.d. 0.001 0.0092.996 3.005 3.002 2.993

0.007 0.007 0.008 0.0190.993 0.993 0.992 0.981

K: 0.989

0.0100.990

0.009 0.0090.991 0.991

Notei A : andalusite, S : sillimanite, F: fibrolite. K: Xgg,Ase/XffiNp..


Fig. l. (A) Geology of the northern portion of the Ardara pluton and surrounding country rock (from Pitcher and Berger, I 972,Fig. 14-6). The rectangular area encompassing Clooney is the area covered by Fig. lB. The locations of two samples (ARD-69-63and ARD-69-65B) are shown near the outer limit of the aureole. (B) SampleJocality map of the northern Ardara aureole, showingaluminum silicate assemblages. The ARD prefix (used to designate these samples in the text and tables) is omitted for clarity.

Biotite has undergone two major replacement reac-tions. First, chloritization of biotite occurs in some sam-ples. Second, and of paramount importance to this in-vestigation, biotite in many samples throughout theaureole has undergone varying degrees of replacement byfibrolite. As illustrated by Figures 28 and 2C, the fibro-litization of biotite is clearly a replacement reaction (cf.Chinner, l96l). Fibrolite also occurs in stringers that areparallel to the foliation. In many samples it is obvious

that these stringers replace biotite, but in others there isno clear evidence of a biotite precursor. Within the silli-manite zone, fibrolitization of biotite is well advanced inmany specimens examined. However, in this zone, fi-brolite also occurs as a product ofgarnet decomposition(Fig. 2E) and as concentrations at grain boundaries (Fig.2n.

In a few samples, fibrolite occurs within muscovitecrystals. The relative timing of growth of these phases is

WETffi

Andalus i te &Kyani te

Andalus i te

Si l l imani te

^ r i : I c!o.o1e.1l ' a r " / ' \ \( r z r - r ) / . ' - : , !- ' / - , l - t \ ) /

- . .

ffif.i:ril*;*yslirilifi t?atu,,-lb

,r\1r) '-r 'r l l

C l o o n e y L o u g h

Y( ,l -, '

ir -.' r 1 /r ino\e ,F.r-u1Q r,'/ \ \- ' -'-

\'r )' t',) i-,"{ i j' \'-= :,:' )r'ii, 7 rL'r',' , { i f i { 6 s - 1 i \ ' - , ' , 1 7 _ - r \ ' \ / \ - ' . : . / - ) \ ' . \ /T-t ; : " f " t - ' " / - \ . . - o"- K- \ ' n- .u\" ] : \ ) \ / - / r ' / - -B-

| - / . , / , , / 1 ' /

/ t , , t - -

\ \

I(ERRICK: FIBROLITE IN CONTACT AUREOLES 245

uncertain, although the presence of well-developed fi-brolite needles suggests that the muscovite did not re-place fibrolite. Pitcher and Read (1963) concluded thatinclusions of fibrolite within muscovite suggest contem-poraneous growth ofthese phases.

In most samples, fibrolite is spatially unassociated withandalusite or sillimanite (Fig. 2D). In essence, fibrolitedid not nucleate on or within a pre-existing AlrSiO, min-eral. Furthermore, within the andalusite and andalusite-kyanite zones, fibrolitized biotite occurs in several sam-ples that lack andalusite and that show no texturalevidence (i.e., muscovite-rich patches) suggestive of an-dalusite porphyroblasts that have been completely re-placed.

Fanad aureole

The Fanad granodiorite is the northernmost of theDonegal granites. This pluton consists of an outer zoneof migmatitic quartzmonzonite enclosing a xenolith-richgranodiorite core. As with the Ardara pluton, the folia-tion becomes more prominent as the external contact isapproached.

The aureole ofthe Fanad granodiorite was studied ina pelitic horizon on the eastern shore of Mulroy Bay,north of the hamlet of Glinsk (Fig. 3). This area was thesubject of several previous investigations (Pitcher andRead, 1963; Naggar and Atherton,1970; Edmunds andAtherton, l97l). The pelitic horizon in this region is partof the Sessiagh-Clonmass Formation, which is part of theDalradian Creeslough Succession. Pitcher and Berger(1972) consider that the Sessiagh-Clonmass Formation isolder than the upper Falcarragh pelites of the northernArdara aureole. Outside of the aureole, the pelite is agarnetiferous schist. Accordingly, the Fanad aureoleformed from a protolith that had been metamorphosedto the greenschist facies by an earlier regional metamor-phism.

The salient features of prograde metamorphism in theGlinsk area were summarized by Naggar and Atherton(1970), Edmunds and Atherton (1971), and Pitcher andBerger (1972). The outer limit of the aureole is markedby the first appearance of biotite within the precursorquartz-chlorite-muscovite-garnet schist. Prograde meta-morphism is delineated by a series of well-defined iso-grads (Fig. 3). Edmunds and Atherton (1971) traced theprograde chemical evolution of garnet within the contactaureole. Through microprobe analysis they were able todistinguish garnet cores, correlative with unresorbed re-gionally metamorphosed garnet, and rim compositionsrepresenting garnet grown during contact metamorphism.

Isograd number 8 in Figure 3 was designated by Ed-munds (1969) as the disappearance of muscovite coupledwith the appearance of fibrolite. However, samples col-lected by the author result in a different location for thefibrolite-in isograd (isograd number 6 in Fig. 3). Further-more, in contrast to Edmunds' (1969) study, this studyshows that muscovite persists into the sillimanite zone.Although some muscovite in the higher-grade rocks ap-

pears to be secondary, muscovite within biotite-rich knots(which were probably garnet porphyroblasts in the pro-tolith) could well be primary.

As in the Ardara aureole, fibrolite is dominantly inter-grown with biotite. This biotite-fibrolite association oc-curs from the fibrolite isograd through the sillimanite zone.As shown in Figures 2G and 2H, textures demonstratethat fibrolite formed from the breakdown of biotite. Withrare exception, fibrolite shows no evidence ofnucleatingon or within andalusite or sillimanite.

The sillimanite zone is characterized by the coexistenceof sillimanite, andalusite, and fibrolite. Sillimanite is in-tergrown with, and has replaced, andalusite.

AN.qlvsrs oF FrBRoLrrE FoRMATToN

P-I conditions

To assess the physicochemical conditions of contactmetamorphism, and to address the P-T conditions of fi-brolite formation, thermobarometry was deterrnined us-ing the garnet-biotite Mg-Fe exchange equilibrium (Ferryand Spear, 1978) and the garnet-plagioclase-AlrSiO'-quartz geobarometer (Ghent, 1976; Ghent et al., 1979).

The garnet-biotite thermometry is summarized in Ta-ble 5. Uncertainties given in this table for the tempera-tures obtained by the Ferry-Spear equation were derivedusing a standard error for the equilibrium constant com-puted as in Hodges and Spear (1982,p. 1126).In so doing,it was assumed that each of the mole fractions used incomputing the equilibrium constant had standard errorsof 3o/o relative. Ferry and Spear (1978) recommend thattheir garnet-biotite geothermometer may be used with(Ca + Mn)/(Ca + Mn * Fe * Me) < 0.2 in garnet and(u'Al + Ti)/("IAl + Ti + Fe + Mg) < 0.15 in biotite.This requirement is fulfilled for most of the analyzed gar-nets (Table 1), although most biotites analyzed from theArdara samples (Table 2) have ("'Al + Ti)/(vrAl + Ti +Fe + Mg) = 0.20. Hodges and Spear (1982) reformulatedthe Ferry and Spear (1978) geothermometer by account-ing for nonideal mixing in garnet. The Hodges and Spear(1982) activity coefrcients for the pyrope and almandinecomponents were derived from a symmetric, four-com-ponent regular solution model using the Ca-Mg binaryinteraction parameter (W.**u) of Newton et al. (1977)and assuming the other binary interaction parameters tobe zero. As shown in Table 5, the temperatures calculatedfrom the Ferry and Spear (1978) versus Hodges and Spear(1982) formulations differ between 2 and 8 deg for mostsamples. However, the temperature differences are 20-25deg for three samples (69-63, 69-658, and 85-264) thatcontain significant Ca contents. In comparison with mostother samples, biotites in these three samples have lowermicroprobe totals, in part attributable to lower KrO con-tents. Although not obvious petrographically, these bio-tites may be altered. Therefore, the temperatures deter-mined by garnet-biotite thermometry for these threesamples may not represent those of the thermal maxima.

Using the garnet-plagioclase-AlrSiOs-9uartz geoba-rometer in sillimanite-zone rocks of the Ardara aureole,


a " - i .

r l

; . J

,_, _ __Jr

B.

tt

,,*'& q'.,

' - t iT

h;*! . i

LJ

, .- t

f.-try.

Fig. 2. Photomicrographs of samples from the Ardara (ARD prefix) and Fanad (GL prefix) aureoles. (A) Late muscovite (M)

cross-cutting biotite (B) (sample ARD-85-31). The foliation of the biotite is preserved as inclusion trails in the muscovite. Widthof photo corresponds to 1.3 mm. (B) Biotite (B) fringedby fibrolite (F) (sample ARD-85-10). Width of photo corresponds to 2.6mm. (C) Biotite with surrounding fibrolite (sample ARD-85-3E). Width of photo corresponds to 0.65 mm. (D) Prismatic sillimanite(S) with adjacent layers of fibrolitized biotite (dark layers in lower portion of photo) (sample ARD-85-10). Width of photo corre-sponds to 2.6 mm. (E) Gamet (G) replaced by fibrolite + p)'rrhotite (sample ARD-85-lC). Width of photo corresponds to 2.6 mm.(F) Fibrolite at grain boundaries (sample ARD-85-3E). Selected grains of biotite (B) and quartz (Q) are noted. Width of photo

corresponds to 1.3 mm.

the rim compositions of zoned plagioclase yield pressures plagioclase consists of a homogeneous, calcic core sur-

(4-5 kbar) well beyond the stability limit of andalusite; rounded by a sodic rim (see Table 3) and that there is a

however, core compositions yield pressures in the range sharp discontinuity between the core and rim. The sodic

of2-3kbar.Electron-microprobetraversesshowthatthe rims have compositions similar to the plagioclase in


- } t

' ' i/t ? t . i .

,4t'Vr

afl'

*,

'}94=' a *

, 'i''

d.n\

I\

" /2 .

.l ' ::1 ,,"

'Y*"tr, .J,lr*eF

'#{Fig.2.-Continued. (G) Biotite fringed and cut by fibrolite (sample GL-85-10). Width of photo corresponds to 1.3 mm. (H)

Remnants of biotite within fibrolite (sample GL-85-8). This represents an advanced stage in the fibrolitization of biotite. Width ofphoto corresponds to 1.3 mm. (I) Triangular arrangement of fibrolite within the basal plane of biotite (sample ARD-85-1C). Widthof photo corresponds to 0.055 mm. (J) Triangular arrangement of fibrolite within the basal plane of biotite. Specimen is a garnet-fibrolite-muscovite-biotite-kyanite-quartz gneiss from Mount Tuckahoe, New York. Width of photo corresponds to 1.3 mm.

quartzofeldspathic dikes and segregations. Thus, growthof the sodic rims may have been influenced by later al-kali-rich fluids emanating from the intrusion. The calciccores thus appear to provide a better representation ofthe plagioclase equilibrated with the earlier-formed meta-morphic assemblage. Accordingly, a pressure of 2.5 + 0.5kbar is assumed for the Ardara aureole.

In the kyanite-andalusite zone, kyanite and stauroliteoccur as inclusions within andalusite porphyroblasts andin the surrounding matrix. Naggar and Atherton (1970),Pitcher and Berger (1972), and Holder (1979) concludedthat kyanite and staurolite are aureole minerals thatformed before andalusite. This paragenetic sequence iscompatible with a prograding aureole passing from thekyanite stability field through to the andalusite stabilityfield. As shown by Kerrick (1968) and Haas and Holda-way (1973), the assemblage kyanite + qtrartz has a veryrestricted stability range at the pressure (2.5 kbar) deter-mined by geobarometry. Assuming that the early-formedkyanite crystallized at equilibrium and that the geoba-rometry is essentially correct, this may suggest that the

fluid phase was somewhat impure, such that in relationto the condition P"ro : P,o,",, the pyrophyllite breakdowncurve is shifted to lower temperatures and the kyanite +quartz stability field is thereby enlarged. The coexistenceofandalusite + kyanite probably reflects the sluggishnessofthe kyanite - andalusite transformation. Thus, kyaniteis considered as a metastable phase during the later an-dalusite-forming event. Naggar and Atherton (1970)demonstrated that kyanite is restricted to Mg-rich rocks;south of the kyanite zone, higher bulk-rock Fe/(Fe + Mg)ratios precluded kyanite as a phase in the assemblage.Thus, the boundary between the kyanite-andalusite zoneand the andalusite zone cannot be interpreted as a pro-grade isograd reaction.

Figure 4 presents the results ofgarnet-biotite geother-mometry in samples from the Ardara aureole assumingP : 2.5 kbar. Some samples were omitted because of theevidence for significant retrogression; i.e., samples ARD-85lC and ARD-85-3 contain garnet retrograded to bio-tite + fibrolite (Fig. 2E), and sample ARD-85-5 containsbiotite that has been produced by retrogression ofanda-

248 KTRRICK: FIBROLITE IN CONTACT AUREOLES

TABLE 5. Ardara aureole samples: Temperatures ("C) from gar-net-biotite geothermometry at P: 2.5 kbar

Sample no. G" L"t Sample no. G". I""t

ARD-69-4ARD-69-6ARD-69-7ARD-69-21ARD-69-23ARD-69-25ARD-69-34ARD-69-45ARD-69-63ARD-69.658ARD-85-14

541 + 19492 + 17477 + 16612 + 235 1 0 + 1 8542 + 19621 + 23544 + 19367 + 12401 + 13644 + 24

547498484620R17

cc+

626550387425648

ARD-85-18 548 + 19ARD-85-1C 647 + 24ARD-85-2A 656 + 25ARD-85-44 496 + 17ARD-8S-54 527 + 19ARD-85-15 493 + 17ARD-85-16 492 + 17ARD-85-17 484 + 17ARD-85-26A 410 + 13ARD-85-314 473 ! 16ARD-85-32A 528 + 19ARD-85-33 587 + 21

ccb653661501534501499491435481535592

Fig. 3. Sample localities and isograds within a pelitic horizonmetamorphosed by the Fanad pluton. The samples are desig-nated with a GL prefix in the text and tables. The isograds are(1) outer limit of new biotite growth, (2) south of this line newbiotite is restricted to knots ofchlorite, (3) new garnet appears,(4) andalusite appears and chlorite disappears, (5) cordierite ap-pears, (6) fibrolite appears, (7) K-feldspar appears, (8) muscovitedisappears, (9) andalusite becomes pleochroic, (10) sillimaniteappears. With the exception of isograd (6), which was deter-mined in this study, the isograds are from Pitcher and Berger(1972,Fie. ru3).

lusite porphyroblasts. Three other samples were rejectedon the basis of the presence of quartzofeldspathic dikesor pegmatitic segegations within the thin sections. In thesesamples, late-stage retrogression may have been fluxedby fluids evolved from the nearby magmatic source.

Figure 4 suggests that the temperature at the sillimaniteisograd was between 550 and 600"C. This is in goodagreement with Holdaway's (1971) determination of theandalusite : sillimanite equilibrium at 2.5 kbar, thusproviding independent support for the pressure estimateobtained from the garnet-plagioclase-AlrSiO5 -quartzequilibrium using core compositions of plagioclase.

The close conformity of the sillimanite isograd to theform of the external intrusive contact supports the inter-pretation of a uniform thermal profile during metamor-phism (as in Fig. 4). Excluding fibrolite, and consideringthe minor-element contents of andalusite and sillimanitein this aureole; Kerrick and Speer (1987) have conclud-ed that the sillimanite isograd represents a close approx-imation to the univariant andalusite : sillimanite equi-librium involving pure endmember phases.

Figure 4 suggests that the fibrolite-bearing samples be-yond a distance of about 200 m from the intrusive con-

' I,s = temperatures using the equation of Ferry and spear (1978). Theuncertainties were computed using the standard-error-propaga-tion equation of Hodges and Spear (1 982, p. 1 1 26).

f I"": temperatures using the Ferry and Spear (1978) equation andgarnet activity coefficients derived from Equation 9 of Hodgesand Spear (1982).

tact formed well below the sillimanite stability field. It ispossible that fibrolite in this zone formed at a higher tem-perature than that recorded by garnet-biotite geother-mometry. Accordingly, fibrolite could have formed at theclimax of the main thermal event, or by a second thermalpulse, perhaps correlated with the intrusion of the coregranodiorite. However, one would expect that the garnet-biotite compositions would have recorded such a higher-temperature history. If contact metamorphism in the Ar-dara aureole represents a single thermal pulse, as arguedby Pitcher and Read (1963), it seems improbable thatfibrolitization would have occurred upon progradationwithout readjustment of the garnet (rimfbiotite compo-sitions to increasing temperature. That garnet-biotite pairsnear the sillimanite isograd represent the thermal maxi-mum is supported by the fact that this geothermometeryields reasonable agreement with Holdaway's (1971) an-dalusite : sillimanite equilibrium at the prevailing pres-sure. It is therefore suggested that fibrolite formed meta-stably at temperatures within the andalusite stability field.

These conclusions are based on temperatures derivedfrom the Ferry and Spear (1978) geothermometer. If weinstead adopt temperatures from the Hodges and Spear(1982) formulation (Table 5), the thermal profile in Fig-ure 4 would be shallower at distances of 0.5 to 1.0 kmfrom the intrusive contact. However, because the datapoints for samples near the fibrolite isograd would essen-tially remain as in Figure 4, there would be little changein the temperatures implied for this isograd. As statedpreviously, the thermometry for samples 69-63,69-658,and 85-264 may be suspect. However, even if these sam-ples were excluded from Figure 4, the remaining datawould nevertheless suggest that fibrolite formed at tem-peratures significantly below the sillimanite stability field.

The thermometry data in Figure 4 are dependent onthe assumption that the garnet rim analyses representequilibrium compositions. For each sample, the garnetanalyses given in Table I represent the average of several


4 0 0

o . 4 o 6

Distance (Km)

Fig. 4. Plot of temperature as a function of distance from the contact of the Ardara pluton. Temperatures for the data pointswere derived using the Ferry and Spear (1978) garnet-biotite geothermometer (Table 5). The solid line was obtained by least-squaresregression using the equation T : Cr 'l Crln D (where C, and, C, are constants and D : distance). The shaded area encompassesthe 950/o confidence level about the mean (solid line). The dashed line was calculated assuming heat transfer by conduction onlyand an intrusion shaped as a vertical cvlinder.

oo 5 0 0F

o 80 . 2

point analyses randomly taken within the outer 100 pmof crystals. The random-point procedure was adoptedlargely as a result of the nature of garnets within anda-lusite-bearing samples. The presence of abundant, smallinclusions within such garnets considerably limited thenumber of suitable areas for analysis. In these samples,zoning profiles in garnet are unattainable owing to thepresence of inclusion-rich cores, or to atoll or skeletalstructures. Complete garnet zoning profiles were obtainedin three samples (Fig. 5). Garnet-biotite thermometry cal-culated from the analytical data for sample ARD-85-26A(Fig. 6) yields no consistent temperature variation acrossthis garnet. However, zoning is apparent in the outer por-tions of garnets in sillimanite-bearing samples ARD-85-lA and ARD-85-IC (Fig. 5). This zoning is reflected inthe temperature profiles shown in Figure 6. [Without fur-ther analysis, such as that described by Spear and Sel-

verstone (1983), the temperatures derived from garnetcore compositions coupled with matrix biotite composi-tions lack geologic significance.l Thus, for samples ARD-85-1A and ARD-85-lC, garnet compositions determinedfrom averaging random-point analyses in the outer 100pm provide a source of error in the geothermometry. Thisfactor undoubtedly contributes to the scatter ofdata pointsplotted in Figure 4.

It is possible that the coexistence of fibrolite with an-dalusite and/or sillimanite is permitted because of non-stoichiometry. Of particular importance is the coexis-tence of fibrolite with andalusite, as nonstoichiometrycould stabilize this mineral pair to lower metamorphicgrades than the sillimanite zone (cf. Fig. 4). To addressthis question, microprobe analyses were carried out onfibrolite and coexisting andalusite in two specimens fromthe Ardara aureole (Table 4). [The presence of contami-

. 6 9 - 2 1

a 9 - 4- \ 6 9 - 4 5I 1 r : } - r s 9 - 2 5

oe_e\/ . .{qlEr!r

a 5 - 1 6 \ | | 6 5 - r

8 5 - 3 1 4 - -

8 5 - 2 6 A .

Si l l imani te- lsograd

D I S T A N C E ( M I C R O N S )

Fig. 5. Mole percentages of components (Al : almandine, Py : pyrope, Gr : grossular, Sp : spessartine) calculated fromelectron-microprobe traverses across selected garnets from the Ardara aureole.


@

T (C.)

ffi

425

4m

375

o aa

a a

" a '

a

a

a

l-Bs- LA Iroo 200 3@ @

fss-lc I . oo t o o a r a a '

o a ' a

r@ zoo o 4@t s 600 zm

f 85264-l.a. 't a t - a a a a

- - a a a a a

r @ m m @ m @ m

DISTANCE (MICRONS}

Fig. 6. Temperatures calculated for electron-microprobe tra-verses across garnets (see Fig. 5) using the Ferry and Spear (1978)garnet-biotite geothermometer.

nant phases (especially quartz) in fibrolite precluded anal-ysis of fibrolite * andalusite pairs in other samples.l Asthe andalusite in these samples is colorless, these micro-probe analyses represent rim compositions (Kerrick andSpeer, 1987). At equilibrium,

A G . : 0 : A G 9 + R T l n K , ( l )

where

K : af;rbr"ror/ aAird.,", : XTif;1ta15;6rlX$ff1ua,51i,r,

This formulation of the equilibrium constant assumesideal mixing in andalusite and fibrolite. For selected Kvalues, the equilibria in Figure 7 were calculated fromEquation I using the venrex program of Connolly andKerrick (1984, 1987). The standard-state free-energychange for the reaction (AGf) was computed using theheat-capacity and enthalpy data of Robie and Heming-way (1984) and the molar volume, thermal expansion,and compressibility data of Robinson et al. (1982).

As shown in Table 4, samples ARD-85-16 and ARD-85-17 have respective K values of 0.989 and 0.983. At2.5 kbar, these equilibrium constants would conespondto respective temperatures of 585 to 570"C (Fig. 7), i.e.,20-35 deg below the curve for the pure andalusite : sil-limanite equilibrium. However, garnet-biotite thermom-etry for these samples yields temperatures around 490"C.Providing that the garnet-biotite thermometry yields peakmetamorphic temperatures, it is concluded that nonstoi-chiometry is insufficient to stabilize fibrolite in these sam-ples.

Because of the difficulty in quantitatively analyzingsmall amounts of light elements with the electron micro-probe, B was not analyzed in the aluminum silicates in-vestigated in this study. Grew and Hinthorne (1983) con-

P (k bod

T EC)

Fig.7. P-T diagram showing isopleths ofselected values ofthe equilibrium constant for andalusite: fibrolite equilibrium(K: lSPi'rlhrsro5/Xtililtilfo5). The heavy line (K: 1.0) corre-sponds to the equilibrium with stoichiometrically pure phases.

cluded that the relatively low B contents in sillimanitewere insufficient to significantly alter the stability rela-tions of this phase. Providing that the crystal chemistryof fibrolite is identical to sillimanite, it is concluded thatB substitution is insufficient to significantly alter the sta-bility of fibrolite.

Salje's (1986) heat-capacity data suggest that the P-Tstability field of fibrolite is significantly smaller than thatof coarse-grained sillimanite. He has concluded that thedestabilization of fibrolite results from excess energy aris-ing from lattice defects (stacking faults). Furthermore,surface free energy would have the effect ofdestabilizingfibrolite in relation to coarse-grained sillimanite. Accord-ingly, within a given contact aureole, fibrolite would beconfined to higher metamorphic grades than sillimanite.However, in the Ardara aureole, fibrolite first appears atlower metamorphic grades than coarse-grained silliman-ite. It is therefore concluded that the excess energy offibrolite relative to sillimanite is not a significant factorin the appearance and distribution offibrolite in this au-reole.

Within the Fanad aureole, only three samples collectedwithin the fibrolite zone contain garnet. Thus, this au-reole yielded considerably less geothermometric data thansamples from the Ardara aureole. In general, tempera-tures within the andalusite-K-feldspar zone of the Fanadaureole are in the vicinity of 600"C.

Moorr- FoR FTBRoLITE FoRMATTON

The mineral textures in the aureoles investigated in thisstudy clearly reveal that fibrolite formed from the break-down of biotite. In several specimens from the Ardara


aureole, this reaction occurred in rocks lacking otherAlrSiO5 phases; thus, it is concluded that fibrolitizationcannot be considered as a polymorphic transformation.Furthermore, Iocal closed-system ionic equilibria linkingtwo aluminum silicate phases, such as suggested by Car-michael (1969) for the kyanite r sillimanite transfor-mation, are precluded for rocks with fibrolite as the solealuminum silicate.

In the aureoles investigated in this study, evidence hasbeen presented suggesting that the fibrolite formed late inthe thermal history of the aureoles. The bulk of the fi-brolite appears to have formed from biotite decomposi-tion. The reaction

2K(Mg"Fe,-,)3AlSi3O,o(OH), + l4HCl: Alrsio5 + 5sio, + 2KCl

+ 6(Mg"Fe, ,)Cl, + 9HrO (l)

would account for the fibrolitization of biotite. Removalof K, Mg, and Fe as chloride complexes is supported bythe efficacy of chlorides as complexing agents. However,petrographic evidence suggests that fibrolite is the dom-inant phase in the fibrolite-quartz aggregates replacingbiotite. Thus, the actual reaction may involve the releaseof Si as an aqueous complex (e.g., HoSiOo). With the ex-ception of one sample (ARD-85-3E), opaque minerals areconspicuously absent in the fibrolite aggregates that havereplaced biotite. Accordingly, along with K and Mg, Feis considered to have been released into the fluid phase.The ultimate fate of the components released into thefluid is unknown. Silica transported as HoSiOo may bedeposited nearby as quartz. If late-stage muscovite formedcontemporaneouslywith fibrolite, the muscovite may haveprovided a local sink for K released from biotite decom-position.

As described by Chinner (1961) for metapelites in theScottish Highlands, samples from the Donegal aureoleswere found to contain fibrolite arranged in a regular tri-gonal pattern on basal sections of biotite (Fig. 2I). Theauthor has observed this texture in fibrolitized biotite fromother localities (Fig. 2J). Flat-stage measurements of theacute angles between the oriented fibrolites typically yieldvalues in the range 58-62. As suggested by Chinner(1961), this geometric arrangement matches that of thetetrahedral layer of the mica structure (see Liebau, 1980,Fig. 3d). Thus, the nucleation of fibrolite, which containstetrahedrally coordinated chains parallel to the c axis, wasfacilitated by the chainlike structure in the tetrahedrallycoordinated layers of the host biotite. The formation ofthe sillimanite (fibrolite) structure from a mica can beenvisioned by considering that the tetrahedral chains ofthe fibrolite were inherited from biotite and that the oc-tahedral sheet of biotite was rearranged upon expulsionof the octahedrally coordinated Mg2+ and Fe2+. Loss ofthe interlayer K+ resulted in juxtaposition of adjacenttetrahedral chains (as in the sillimanite structure). A keyproblem is why fibrolite formed rather than andalusite,the polymorph postulated to be stable at the P-I condi-tions of andalusite-zone rocks. Aside from octahedrally

TneLe 6. Electron-microprobe analyses of fibrolitized vs. non-fibrolitized biotite

GL-85-8ARD-85-10-

FibrolitizedNonfibroli-

tized

sio,Alr03Tio,M9oFeOMnONaroK.ocl

34.57 34.5720.08 20.123.41 2.544.86 5.44

23.67 23.900.20 0.150.09 0.238.85 8.75

95.73 95.70

5.3232.6780 9680.3941 . 1 1 53.0480.0255.550

5.3332.6670.9900.2921.2493.0820.0175.630

0.0681.7201.7880.2230.549

35 1319.293.326.70

22.700.180.208.770.03

96.32

34.8420.132.447.12

22.230.140.147.850.02

94.91

5.3402.660n 476

o.2791.6242.8610.0185.757

Formula basis: 5 oxygensSirvAlvrAlTiMgFeMntCaNaK

X"n,X^""

5.3512.6490.8110.3801.5202.8920.0225.b25

0.0581.7031.761o.2710.516

0.0401.5341.5740.2830.499

0.0231.7301.753o.2020.552

. Biotite rimmed by fibrolite (see discussion in text).

coordinated Al, the remainder of the Al in the andalusitestructure is in fivefold coordination, whereas that in sil-limanite is in tetrahedral coordination. The model of fi-brolite formation assumes that tetrahedral Al was inher-ited from the biotite structure. Accordingly, it is postulatedthat andalusite was precluded because of the additionalenergy required to rearrange the tetrahedrally coordinat-ed Al of biotite into the AlO5 hexahedra of andalusite.

Fibrolite is considered to have formed in the waningstages of the thermal event. According to Reaction l,fibrolitization of biotite would be enhanced by high a".'.In a recent review, Eugster (1985) has concluded that HCIis evolved from crystallizing granitic magmas. Supportfor Cl as a major component in volatiles released fromthe intrusion comes from the fact that Cl is strongly par-titioned into the vapor phase (Pichavant, 1983). Hydro-lysis reactions such as 2NaCl-.'. + HrO : 2HCl""o". +NarO-",, (Eugster, 1985) may be responsible for produc-tion of HCI in the magmatic vapor phase. Significantamounts of HCI can also be produced by the crystalli-zation of metal sulfides within the intrusion (Burnham,r979).

Interpretation of the garnet-biotite geothermometrycould be questioned on the basis ofthe widespread late-stage fibrolitization of biotite. That is, the thermometrywould not record maximum temperatures if there werechanges in the composition of biotite resulting from thefibrolitization reaction. To test this, electron-microprobeanalyses were carried out on samples containing both fi-


brolite-free and fibrolitized biotite. In sample GL-85-8(from the Fanad aureole), there are no significant difler-ences in the composition of fibrolitized versus fibrolite-free biotite (Table 6). Within a dike cutting sample ARD-85-10 there are relatively large crystals of biotite withperipheral fibrolitization (Fig. 2B). The biotite adjacentto the fibrolite has a slightly different interference colorcompared to the central portion of the biotite. Electron-microprobe analyses suggest that the biotite adjacent tofibrolite is depleted in Fe and Ti compared to the centralportion (Table 6). Nevertheless, the diference in molefraction of annite and phlogopite between core and rimbiotite is approximately I mol0/0. Thus, for thermometricanalysis, it is concluded that the fibrolitization reactionhas not significantly affected biotite compositions.

Within the sillimanite zone of the Ardara and Fanadaureoles, fibrolite occurs in anastomosing stringers thatprobably represent fracture fillings. These features werealso described by Pitcher and Read (1963) and Pitcherand Berger (1972). The author has observed these tex-tures in metapelites from other contact aureoles. Fibrolitein some samples from the sillimanite zones of the Ardaraand Fanad aureoles is confined to grain boundaries ofquartz and feldspar (Fig. 2F). However, the bulk of thefibrolite within the andalusite and andalusite-kyanitezones appears to be a replacement product of biotite. Ac-cordingly, the Al for such fibrolite was obtained from thebiotite precursor.

Late-stage muscovitization is particularly prominent inthe Ardara aureole. Pitcher and Read (1963) and Pitcherand Berger (1972) concluded that fibrolite and muscoviteare contemporaneous (or nearly so). On the basis of tex-tures of fibrolite-muscovite intergrowths examined in thisstudy, it is concluded that there is no cogent argumentagainst the contemporaneity of fibrolite and muscovite.

In the Ardara aureole, tourmaline is a prominent ac-cessory in rocks with abundant fibrolite and muscovite.Thus, this study supports Pitcher and Berger's (1972)metasomatic fibrolite-muscovite-tourmaline association.In the Ardara aureole, support for a magmatic source fortourmaline comes from the fact that large tourmalinecrystals occur in dikes cutting pelitic schists near the in-trusive contact. In fact, Akaad (1956b) noted that azoneof tourmaline + quara (varying from 3 to 25 cm in thick-ness) occurs along the northern and northeastern contactsof the Ardara pluton and that a narrow zone (ca. 2-3 mm)of the adjacent pelite has been replaced by tourmaline.The author examined specimens collected from progres-sively metamorphosed pelites in the Lettermackawardarea, which is considered to be within the aureole of theThorr pluton. As in the study of Pitcher and Read (1963),there is an increase in the abundance of late-stage fibro-lite + muscovite as the intrusive contact is approached.In fact, Pitcher and Berger (1972, p.307) concluded that"this late-stage encroachment of a metasomatic zoneseems to be a general feature of the Donegal aureoles."Thus, the conclusion is arrived at that, in addition toHCl, the fluids evolved from the cooling intrusion re-

leased B. Such metasomatic fluids may also have beenactive in some regionally metamorphosed rocks. For ex-ample, the presence of B-enriched fluids in the late-stagefibrolitization of granulite-facies rocks was documentedin the study of Ahmad and Wilson (1981).

Experimental data show that, for systems containingCl and B, the magmatic vapor phase should be enrichedin K and Na. Thus, it is probable that the fluid carriedNa in addition to K. In the Ardara aureole, it is possiblethat the Na-enriched rims on plagioclase reflect the sinkfor the Na of the metasomatic fluid.

There is no cogent evidence bearing on the relativetiming of the fibrolitization of biotite versus the devel-opment of the metasomatic fibrolite-muscovite-tourma-line assemblage. If we assume that these processes werecontemporaneous, fibrolitization by Reaction I would beenhanced by low a"., in the metasomatic fluid. Providingthat the development of the sodic rims on plagioclase wascontemporaneous with a single fibrolite-forming event, itis possible that the metasomatic fluid had a relativelyhigh a*u.,/a"., ratio. This is in concert with the experi-mental data of Burnham(1967), which showed that fluidsin equilibrium with granitic material at 650'C and 6 kbarhave molar NaCl/KCl ratios in the range 1.8-1.9. De-velopment of muscovite would have been favored by KCIreleased from the fibrolitization ofbiotite (Reaction l).

Aside from the areas investigated in this study, thereare other contact aureoles where fibrolite first appears atlower grade than coarse sillimanite (Woodland, 1963;Loomis, 1972; Bliimel and Schreyer,1976; Woodsworth,1979; Speer, 1982). In the aureoles studied by Loomis(1972), Woodsworth (1979), and Speer (1982), the bio-tite-fibrolite association is common. Thus, the mecha-nism proposed in this study may be a common phenom-enon in contact-metamorphosed metapelites.

Late-stage fibrolitization has also been postulated inseveral studies of regionally metamorphosed pelites(Chinner, 1966; Chakraborty and Sen, 1967; Ashworth,1975). In a regionally metamorphosed sequence in theMount Mossilauke area of New Hampshire, Rumble(1973) showed that fibrolite is widely distributed in rockscontaining sillimanite and/or andalusite. The fibrolite istypically associated with biotite or muscovite. In thePlymouth quadrangle, Rumble (1973, Fig. l) showed nu-merous localities where fibrolite occurs without coexist-ing andalusite or sillimanite. Accordingly, this area mayhave experienced late-stage fibrolitization. The numerousgranitic intrusions in this area could have provided asource for acidic volatiles needed for the fibrolitization.Within the Ticino thermal high of the Lepontine Alps ofSwitzerland, representing a Barrovian regional metamor-phic regime, the author has found an extensive zone (12km in width) of fibrolite within the upper kyanite zone.Much of the fibrolite in this area formed as a replacementof biotite. Biihl (1981) concluded that fibrolite formedduring retrograde metamorphism in sillimanite-zonegneisses of this area. Within the Lepontine Alps thereoccur spectacular AlrSiOr-quartz-feldspar segregations


containing abundant tourmaline; these segretations areconsidered to have formed in the waning stages of thethermal event (Kerrick et al., 1985). Thus, as in the Don-egal aureoles, B is considered as a constituent ofthe late-stage fluids.

Ferry's (1983) study suggests that acid metasomatismmay be important in regional metamorphism. Thus, themechanism of fibrolitization of biotite by influx of HCI-bearing fluids could well account for fibrolitized biotitein many regionally metamorphosed pelites. It must beemphasized, however, that fibrolite is not a late-stagemineral in all regionally metamorphosed pelites. In theTruchas Peak region of New Mexico (Grambling, l98l)and in the Mica Creek region of British Columbia (E. D.Ghent, pers. comm.), fibrolite and coarse sillimanite ap-pear together at the first sillimanite isograd in these re-gionally metamorphosed terranes.

Note added inprool Subsequent to final checking ofthepage proofs of this paper, I became aware of J.A.T. Smel-lie's (1974) paper on the genesis of atoll garnets in metape-lites ofthe Ardara aureole. He concluded that most garnetsdisplay varying degrees of replacement by biotite + pla-gioclase + quartz + opaques. In some samples, this re-placement reaction progressed inward from the exteriorportions ofgarnet crystals, whereas in others, the replace-ment reaction began in the interiors of garnet crystals.Smellie concluded that garnet atoll structures formed inthe advanced stages ofreplacement. From analysis ofgar-net zoning profiles, it was concluded that the replacementreaction yielded reversed zoning characterized by deple-tion in Ca and Mg in the peripheral regions of resorbedgarnets. The garnet profiles shown in Figure 5 ofthe pres-ent contribution were obtained on relatively large, round-ed garnet crystals. According to Smellie's interpretation,such garnets would be least affected by the replacementreaction. The garnet zoning profiles of samples 85-lA and85-264 Frg. 5) were obtained from crystals lacking tex-tural evidence for replacement. However, the garnet insample 85-1C contains fibrolite in the core, thereby im-plying that replacement occurred in the interior. Follow-ing Smellie's analysis, the lack of detectable Ca depletionin the rims of the garnets plotted in Figure 5, coupled withsimilar compositions of the rims on each garnet crystal(Fig. 5), suggests that the peripheral regions ofgarnets inFigure 5 were not affected by the replacement reaction.Nevertheless, Smellie suggested that reverse zoning iscommon in the outer 25 pm of garnet crystals of meta-pelites in the Ardara aureole. Accordingly, his data sup-port the conclusion (p. 2a9) that compositional zoningprovides a source oferror for the garnet analyses ofTable1, which were obtained by averaging random points with-in the outer 100 pm of crystals.

Garnet atoll structure is present in samples 69-7,85-lC, 85-17, and 85-32A. Garnets in samples 69-34,69-45,69-63,85-l 5, and 85-33 have "dusty" centers thatmay reflect incipient replacement of the central portions.Garnets analyzed in other samples (Table l) lack textural

evidence of replacement. Inspection of the geothermo-metric data as a function of distance from the intrusivecontact (Fig. a) reveals no consistent diferences betweenthe above three textural groups of garnet. Accordingly,the garnet replacement reaction does not appear to com-plicate the interpretation of the thermal profile of the au-reole as derived from garnet-biotite thermometry Gig. a).

AcxNowr,nlcMENTS

This research was supported by a grant from the Earth Sciences Sectionofthe National Science Foundation (Grant no EAR 83-07682). I am verygrateful to J. A. Grambling and J M. Rice for critically reviewing an earlyversion of this paper. J. D. Kerrick kindly provided advice regardingstatistical error analysis. J. Bowers provided considerable assistance inthis project Mary Frank cheerfully and efficiently typed the manuscript.

RnrnnnNcnsAhmad, R., and Wilson, C.J.L. (1981) Uranium and boron distributions

related to metamorphic microstructure-Evidence for metamorphic fluidactivity. Contributions to Mineralogy and Petrology, 76,24-32.

Akaad, M.K. (1956a) The Ardara granitic diapir of Co. Donegal, Ireland.Geological Society ofLondon Quarterly Journal, I 12, 263-288.

- ( 1956b) The northern aureole of the Ardara pluton of Co. Donegal.Geological Magazine, 93, 37 7 -392.

Albee, A.L., and Ray, L. (1970) Corection factors for electron probe

microanalysis of silicates, oxides, carbonates, phosphates and sulfates.Anall.tical Chemistry, 42, 1408-1414.

Ashworth, J.R. (1975) The sillimanite zones of the Huntly-Portsoy areain the north-east Dalradian, Scotland Geological Magazine, I 12, I 13-

Bell, P.M., and Nord, G. (1974) Microscopic and electron diftaction studyof fibrolitic sillimanite. Carnegie Institution of Washington Year Book73,444-448.

Bliimel, P., and Schreyer, W (1976) Progtessive regional low-pressuremetamorphism in Moldanubian metapelites of northern Bavarian For-est, Germany. Krystalinikum, 12, 7-30.

Btihl, H. (1981) Zur sillimanitbildung in den gneisen der zone von Bellin-zona. Schweizerische Mineralogische und Petrographische Mitteilun-gen, 6l , 27 5-295.

Bumham, C.W (1967) Hydrothermal fluids at the magmatic stage. InH L. Barnes, Ed. Geochemistry of hydrothermal ore deposits, p. 3'l-76. Holt, Rinehart, & Winston, New York.

- (1979) Magmas and hydrothermal fluids. In H.L. Bames, Ed. Geo-chemistry of hydrothermal ore deposits, p. 7l-136. Wiley, New York.

Carmichael, D.M. (1969) On the mechanism of prograde metamorphicreactions in quartz-bearing pelitic rocks. Contributions to Mineralogyand Petrology. 20, 244-267.

Chakrabony, K.R., and Sen, S.K. (1967) Regional metamorphism of pe-

litic rocks around Kandra, Singhbhum, Bihar. Contributions to Min-eralogy and Petrology, 16, Zl0-232.

Chinner, G.A. (1961) The origin of sillimanite in Glen Clova, Angus.Journal of Petrology, 2, 312-323.

-(1966) The distribution of pressure and temperature during Dal-radian metamorphism. Geological Society of I-ondon Quart€rly J.our-nal.122, I59-186.

Connolly, J.A.D., and Kerrick, D.M. (1984) venrrx: A computer algo-rithm for the calculation of stable phase assemblages and Schreine-makers-type phase diagrams (abs.). EOS, 65, 287.

-(1987) An algorithm and computer program for calculating com-position phase diagrams. CALPHAD, 11, l-61.

Edmunds, W.M. (1969) A chemical and mineralogical study of hornfelsesin County Donegal. Ph.D. thesis, University of Liverpool.

Edmunds, W.M., and Atherton, M.P. (1971) Polymetamorphic evolutionofgarnet in the Fanad aureole, Donegal, Eire. Lithos, 4,147-161.

Eugster, H.P. (1985) Granites and hydrothermal ore deposits: A geochem-ical framework. Mineralogical Magazine, 49, 7 -23.

254 I(ERRICK: FIBROLITE IN CONTACT AUREOLES

Ferry, J.M (1983) Regional metamorphism of the Vassalboro Formation,south-central Maine, USA: A case study of the role of fluid in meta-morphic petrogenesis. Geological Society oflondon Joumal, 140, 55 l-576.

Ferry, J.M., and Spear, F.S. (1978) Experimental calibration of the par-titioning of Fe and Mg between biotite and garnet. Contributions toMineralogy and Petrology, 66, ll3-117.

Finger, L.W., and Prince, E. (1972) Neutron diffraction studies: Andalu-site and sillimanite. Carnegie Institution of Washinglon Year Book 71,496-500.

Ghent, E.D. ( I 976) Plagioclase-garnet-Alrsio s-quartz: A potential geo-barometer-geothermometer. American Mineralogist, 6 l, 7 l0-7 14.

Ghent, E.D., Robbins, D.B., and Stout, M.Z. (1979) Geothermometry,geobarometry, and fluid compositions of metamorphosed calc-silicatesand pelites, Mica Creek, British Columbia. American Mineralogist, 64,87.s-885.

Grambling, J.A. (1981) Kyanite, andalusite, sillimanite, and related as-semblages in the Truchas Peaks region, New Mexico. American Min-eralogist, 66, 7 02-7 22.

Grew, E.S., and Hinthorne, J.R. (1983) Boron in sillimanite. Science,22l,547-549.

Haas, H., and Holdaway, M.J. (1973) Equilibria in the system Al,O,-SiOaHrO involving the stability limits of pyrophyllite. American Jour-nal of Science. 27 3. 449-464.

Halliday, A.N., Aftalion, M., and I-eake, B.E. (1980) A revised age forthe Donegal granites. Nature, 284, 736-737.

Hodges, K.V., and Spear, F.S. (1982) Geothermometry, geobarometryand the AlrSiOs triple point at Mr. Moosilauke, New Hampshire.American Mineralogist, 67, I I l8-l 134.

Holdaway, M.J. (1971) Stability of andalusite and the aluminum silicatephase diagram. American Joumal of Science,27 l, 97-131.

Holder, M.T. (1979) An emplacement mechanism for post-tectonic gran-ites and its implication for their geochemical features. In M.P. Athertonand J. Tarney, Eds Origin ofgranite batholiths: Geochemical evidence,p. I 16-128. Shiva.

Kerrick, D.M. (1968) Experirnents on the upper stability limit of pyro-phyllite at 1.8 kb and 3.9 kb water pressure. American Journal ofSci-ence. 266. 204-214.

- ( I 986) Dislocation strain energy in the AlrSiO, polymorphs. Phys-ics and Chemistry of Minerals, 13, 221-226.

Kerrick, D.M., and Heninger, S.c. (1984) The andalusite-sillimaniteequilibriurh revisited. Geological Society of America Abstracts withPrograms, 16, 558.

Kerrick, D.M., and Speer, J.A. (1987) The role of minor element solidsolution on the andalusite-sillimanite equilibrium in metapelites andperaluminous granitoids. American Joumal of Science, il press.

Kerrick, D.M., Mullis, J., and Klein, H. (1985) AlSiO,-bearing segrega-tions in the Lepontine Alps, Switzerland: Aluminum mobility in meta-pelites. Geological Society of America Abstracts with Programs, 17,627.

Liebau, F. (1980) Classification of silicates. Mineralogical Society ofAmerica Reviews in Mineralogy, 5, l-24.

Loomis, T.P. (1972) Contact metamorphism of pelitic rock by the Rondaultramafic intrusion, southern Spain. Geological Society of AmericaBulletin. 83. 2449-247 4.

Naggar, M.H., and Atherton, M.P. (1970) The composition and meta-morphic history of some aluminum silicate-bearing rocks from the au-reoles of the Donegal granites. Journal of Petrology, I l, 549-589.

Navrotsky, A., Newton, R.C., and Kleppa, O.J. (1973) Sillimanite-dis-ordering enthalpy by calorimetry. Geochimica et Cosmochimica Acta,37, 2497-2508

Nefion, R.C., Charlu, T.V., and Kleppa, O.J. (1977) Thermochemistryof the high pressure garnets and clinopyroxenes in the system CaO-MgO-AlOr. Geochimica et Cosmochimica Acta, 41, 369-377 .

Okrusch, M., and Evans, B.W. (1970) Minor element relationships incoexisting andalusite and sillimanite. Lithos, 3, 261-268.

Peterson, R.C., and McMullan, R.K. (1986) Neutron diffraction studiesof sillimanite. American Mineralogist, 71, 7 42-745.

Pichavant, M. (1983) Melt-fluid interaction deduced from studies of sil-icate-B,Or-H:O systems at I kbar. Bulletin de Min6ralogie, 106,201-2 t t .

Pitcher, W.S. (1965) The aluminum silicate polymorphs. In W.S. Pitcherand G.W. Flinn, Eds. Controls of metamorphism, p. 327-341. Wlley,New York.

Pitcher, W S., and Berger, A.R. (1972) The geology ofDonegal: A studyof granite emplacement and unroofing, 435 p. Wiley, New York.

Pitcher, W.S., and Read, H H (1963) Contact metamorphism in relationto manner ofemplacement ofthe granites of Donegal, Ireland. Journalof Geology, 7 1, 261-296.

Powell, D., and Phillips, W.E.A. (1985) Time and deformation in theCaledonide orogen ofBritain and Ireland. In P.L. Harris, Ed. The na-ture and timing of orogenic activities in the Caledonian rocks of theBritish Isles. Geological Society of London Memoir 9, 17-39.

Robie, R.A., and Hemingway, B.S. (1984) Entropies ofkyanite, andalu-site, and sillimanite: Additional constraints on the pressure and tem-perature of the ALSiO, triple point. American Mineralogist, 69, 298-306

Robinson, G R., Jr., Haas, J.L., Jr., Schafer, C.M., and Haselton, H.T.,Jr. (1982) Thermodynamic and thermophysical properties of selectedphases in the MgO-SiOr-HuO-CO', CaO-AlrO3-SiOr-HrO-CO:, and Fe-FeO-FerOr-SiO, chemical systems, with special emphasis on the prop-erties of basalts and their mineral components. U.S. Geological SurveyOpen-File Report 83-79, 429 p.

Rumble, D., IIL (1973) Andalusite, kyanite and sillimanite from the MountMoosilauke region, New Hampshire. Geological Society of AmericaBulletin. 84. 2423-2430.

Salje, E. (1986) Heat capacities and entropies ofandalusite and silliman-ite: The influence offibrotitization on the phase diagam ofthe AlrSiO'polymorphs. American Mineralogist, 7 l, 1366-137 l.

Sanderson, D.J., and Meneilly, A.W. (1981) Analysis of three-dimen-sional strain modified uniform distributions: Andalusite fabrics from agranite aureole Journal of Structural Geology, 3, 109-1 16.

Smellie, J.A.T. (1 974) Formation of atoll garnets from the aureole of theArdara pluton, Co Donegal, Ireland. Mineralogical Magazine, 39, 878-88E .

Spear, F. S., and Selverstone, J. (l 983) Quantitative P- f paths from zonedminerals: Theory and tectonic applications. Contributions to Miner-alogy and Petrology, 83, 348-357.

Speer, J.A. (1982) Metamorphism of the pelitic rocks of the Snyder Groupin the contact aureole of the Kiglapait layered intrusion, labrador:Effects ofbuffering partial pressures ofwater. Canadian Joumal ofEarthSciences. 19. 1888-1909.

Vernon, R.H. (1979) Formation of late sillimanite by hydrogen metaso-matism (baseJeaching) in some high-grade gneisses. Lithos, 12, 143-152.

Woodland, B.G. (1963) A petrographic study of thermally metamor-phosed pelitic rocks in the Burke Area, northeastern Vermont. Amer-ican Journal of Science, 261, 354-37 5.

Woodsworth, G.J (1979) Metamorphism, deformation, and plutonism inthe Mount Raleigh pendant, Coast Mountains, British Columbia. Ca-nadian Geological Survey Bulletin, 295,58 p.

Znn,E-An. (1969) The stability relations ofthe polymorphs ofaluminumsilicate: A survey and some comments. American Joumal of Science,276,297-309.

MeNuscmsr REcETvED Mev 30, 19E6M,c,NUscRrPr ACCEPTED DEcer"rsen l, 1986

Date post:	05-Feb-2021
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

Fibrolite in contact aureoles of Donegal, Ireland DBnnrr.r ... › ammin › AM72 ›...

Documents