Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Thermobarometry of pelitic rocks using equilibria between quartz-garnet-aluminosilicate-muscovite-biotite,… McMullin, David William Augustine 1990

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1991_A1 M37.pdf [ 30.04MB ]
Metadata
JSON: 831-1.0052348.json
JSON-LD: 831-1.0052348-ld.json
RDF/XML (Pretty): 831-1.0052348-rdf.xml
RDF/JSON: 831-1.0052348-rdf.json
Turtle: 831-1.0052348-turtle.txt
N-Triples: 831-1.0052348-rdf-ntriples.txt
Original Record: 831-1.0052348-source.json
Full Text
831-1.0052348-fulltext.txt
Citation
831-1.0052348.ris

Full Text

T H E R M O B A R O M E T R Y OF PELITIC ROCKS USING EQUILIBRIA B E T W E E N QUARTZ-GARNET-ALUMINOSILICATE-MUSCOVITE-BIOTITE, WITH APPLICATION TO ROCKS OF T H E QUESNEL L A K E A R E A , BRITISH COLUMBIA. By David William Augustine McMullin B.Sc. (Honours) National University of Ireland, 1982 M.Sc. Acadia University, 1984 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY THE FACULTY OF GRADUATE STUDIES GEOLOGICAL SCIENCES We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October 1990 © David William Augustine McMullin, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada Date DE-6 (2/88) Abstract Rocks of the Quesnel Lake area are divided into three units: unit 1, a continental margin sequence; unit 2, The Crooked Amphibolite (an ocean-floor sequence); and unit 3, the Quesnel sedimentary and volcanic sequence. Two conglomerate localities within unit 3 contain clasts identified as being derived from deformed rocks of units 1 and 2. Deformation of the combined package of units 1 and 2 must have accompanied the emplacement of unit 2 onto unit 1 sometime between the deposition of unit 2 (Mississippian - Permian) and the deposition of unit 3 (Triassic - Jurassic). Rocks of unit 1 have been divided by earlier workers into the Barkerville and Cariboo terranes, separated by the Pleasant Valley Thrust. An extensive review shows that the two terranes are stratigraphically similar and share most of their structural history. The Pleasant Valley Thrust, if it exists, is an extremely early structure. These data do not satisfy the criteria for naming these units 'terranes'. The rocks of unit 1 and 2 experienced an extra phase of deformation not seen in rocks of unit 3. A total of five phases of folding are present. Phases 1 through 4 are approximately coaxial with northwest axes and variably oriented axial planes. Phase 5 has northeast trending axes and vertical axial planes. F i is seen in units 1 and 2 only and is visible in outcrop as rootless isoclinal folds and a transposed foliation. In thin section, S i is only preserved within the earliest garnet porphyroblasts. F 2 folding is the major deformational event. Peak metamorphism accompanied and outlasted it. Major F2 folds are present in the field and are accompanied by an axial planar foliation. In thin section, S2 wraps around earlier porphyroblasts but is overgrown by later ones (staurolite, kyanite). F 3 folding is responsible for the major map-scale structures. It ii postdated the peak of metamorphism and isograds axe folded by it. In thin section S 3 is commonly a crenulation cleavage or transposed foliation. Some late mineral growth accompanied the early stages of F 3 . F 4 and F5 are buckle folds and kinks and may be conjugate fold sets from a single deformational event. They are not generally visible in thin section. The assemblage silica - garnet - aluminosilicate - mica (SGAM) is common in amphibolite grade meta-pelitic rocks, and can be used as a thermobarometer if the activities of muscovite and biotite can be calculated accurately. A new method of calculating the ideal activity of mica components is proposed. Standard models do not adequately account for the degree of coupled substitution that takes place. The proposed method stores the site occupancies in a 4-dimensional array and manipulates the entries to satisfy three criteria. 1: That non-permitted ionic configurations (species) have an activity of zero. 2: That the sum of all activities is unity. 3: That the sum of all activities of species containing a particular ion in a particular site is the site occupancy of that ion. The method is computationally simple and yields activity values that satisfy the distribution of species equations of an ideal complex solution model. Standard state properties for annite and Margules solution parameters for biotite are determined using mathematical programming techniques on published experimental and natural assemblage data. Published volume data indicate that Fe-Mg mixing in biotite is ideal. The data permit the calculation of four Margules parameters (MgTi, FeTi, MgAl, FeAl). The differences MgTi - FeTi and MgAl - FeAl are similar to those found by previous workers but the treatment of the data suggests moderately large individual values for the Margules parameters (up to 75 kJ/mol). Using these activity models the SGAM thermobarometer is applied to several sets of published analyses which show that this calibration offers distinct improvements over previous calibrations. Pressures determined using the new calibration are consistent with other iii barometers and the aluminosilicate polymorph present. In addition, several data sets show field gradients, particularly in P, not previously recognized and which agree with field observations. The SGAM barometer applied to the analytical data from the Quesnel Lake area yields pressures and temperatures that are consistent with the mapped isograds. The pressure and temperature gradients indicate that the final setting of the thermobarome-ter was diachronous across the area and during the early stages of F 3 folding. Hot rocks in cores of anticlines 'set' at later times and at shallower depths than cooler rocks in adjacent synclines. Tight spacing of isograds is more consistent with post-metamorphic folding than with high thermal gradients. iv T a b l e o f C o n t e n t s A b s t r a c t i i L i s t o f T a b l e s x L i s t o f F i g u r e s x v i i A c k n o w l e d g e m e n t s x x i P r e f a c e x x i i i 1 F o l i a t e d p e b b l e s i n Q u e s n e l T e r r a n e 1 1.1 Abstract 1 1.2 Introduction 3 1.3 Conglomerate Localities 4 1.3.1 Wingdam 4 1.3.2 Quesnel Lake 9 1.4 Origin of the Clasts 12 1.5 Conclusions 13 2 S t r u c t u r e a n d S t r a t i g r a p h y : A s y n t h e s i s 15 2.1 Introduction 15 2.2 Previous work 17 2.3 Terminology 19 v 2.4 Tectonic subdivisions of the Canadian Cordillera in the Quesnel Lake area — old and new 21 2.5 New tectonic units — stratigraphy 23 2.5.1 Unit 1 — The continental margin sequence 23 2.5.2 Unit 2 — The Crooked Amphibolite (CA) 40 2.5.3 Unit 3 — Quesnel sedimentary and volcanic sequence 45 2.6 Structure 51 2.6.1 Phases of Folding 52 2.7 Summary 61 3 Thermodynamic activity of micas 65 3.1 Introduction 65 3.2 Ideal activity as probability 66 3.3 Charge balanced species in micas and the distribution of species equations 69 3.4 An algorithm for calculating activity in species with coupled substitution 72 3.5 Calculating activity in micas using the new algorithm 77 3.6 Values of activity for common micas using the new technique 83 3.7 Conclusions, recommendations and further work 85 4 Calibration of S G A M thermobarometer 86 4.1 Abstract 86 4.2 Introduction 87 4.3 Approaches to calculating P and T 88 4.4 SGAM - previous calibrations 94 4.5 Properties and parameters to be determined 98 4.6 Data Sources for MAP 100 4.7 Phase equilibrium data 104 vi 4.7.1 Treatment of experimental data 104 4.7.2 Treatment of natural data 106 4.8 MAP analysis 109 4.9 Testing the calibration of SGAM against independent data 112 4.10 Discussion and Conclusions 122 5 Metamorphism of the Quesnel Lake area 124 5.1 Introduction 124 5.2 Mineral growth and 'isograds' 126 5.2.1 Micas 127 5.2.2 Garnet 131 5.2.3 Staurolite 139 5.2.4 Kyanite 146 5.2.5 Sillimanite 149 5.2.6 Andalusite 149 5.3 Microprobed specimens: Mineral inhomogeneity and zonation 154 5.3.1 Garnet 154 5.3.2 Biotite and muscovite 170 5.4 Pressure and temperature determinations 171 5.4.1 Introduction 171 5.4.2 Microprobed specimens: Pressures and temperatures 174 5.5 Pressure - temperature trends and tectonic implication 188 5.6 Summary and conclusions 198 References 200 Appendices 213 vii A Points of confusion arising from the literature 213 B Computer programs to calculate activities of mica components 215 B.l Charge balanced species in micas 215 B.2 Equilibria among charge balanced species in mica 217 B.3 Mass action equations for charge balanced species in mica 220 B.4 MICAC.PAS program 222 B.5 Unit DISTSPEC.PAS called by MICAC.PAS 242 B.6 Input files required by MICAC.PAS 248 B.6.1 Element weights (ELEMENT.DAT) 248 B.6.2 Input analyses (*.IN) 248 B.7 Output files generated by MICAC.PAS 249 C Calculated pressures and temperatures from published analyses 252 D Microprobed samples: Petrography, analyses, P and T 256 D.l Introduction 256 D.2 Summary petrography and selected microprobe analyses 256 D.2.1 #1: JSG-81-234 (Getsinger, 1985) 258 D.2.2 #2: JSG-81-278 (Getsinger, 1985) 258 D.2.3 #3: JSG-80-30 (Getsinger, 1985) 259. D.2.4 #4: DM-85-9 260 D.2.5 #5: PDL-447 (Lewis, 1987) 260 D.2.6 #6: DM-85-42 262 D.2.7 #7: SLG-370A (Garwin, 1987) 263 D.2.8 #8: DM-85-87 263 D.2.9 #9: CJNF-9 (Fletcher, 1972) 264 viii D.2.10 #10: DM-86-70 266 D.2.11 #11: DM-86-173 266 D.2.12 #12: DM-86-197 267 D.2.13 #13: DM-86-205 268 D.2.14 #14: DM-86-227 269 D.2.15 #15: DM-86-270 269 D.2.16 #16: DM-86-281 271 D.2.17 #17: DM-86-351 272 D.2.18 #18: DM-86-369 272 D.2.19 #19: DM-86-384 273 D.2.20 #20: JRM-0626-36 (Montgomery, 1985) 274 D.2.21 #21: JRM-0716-76 (Montgomery, 1985) 275 D.2.22 #22: LCP-40 (Pigage, 1978) 276 D.2.23 #23: LCP-223 (Pigage, 1978) . . 277 D.2.24 #24: DM-87-76 277 D.2.25 #25: DM-87-71 279 D.2.26 #26: DM-86-45 280 D.2.27 #27: JAF-7-9-29 (FiUipone, 1985) 280 D.2.28 #28: DM-87-8 282 D.2.29 #29: DM-87-17 284 D.2.30 #30: DM-87-24 285 D.2.31 #31: DM-87-32 287 D.2.32 #32: DM-87-35 288 D.2.33 #33: DM-87-39 289 D.2.34 #34: DM-87-54 290 D.2.35 #35: JKR-88-4 (Radloff, 1989) 290 ix List of Tables 2.1 Names and correlations given by previous workers to rocks of unit 2, the Crooked Amphibolite 41 2.2 Nomenclature applied by previous workers in the Quesnel Lake area to rocks of unit 3, the Quesnel sedimentary and volcanic sequence 47 2.3 Correlation of the phases of deformation applied by previous workers to rocks of the Quesnel Lake area 53 4.1 Glossary of notation and symbols used 89 4.2 Mineral names, formulae and abbreviations used in text 91 4.3 Phase equilibrium data used in MAP 101 4.4 Enthalpies, entropies and volumes 102 4.5 Solution (Margules) parameters 103 5.1 Summary table of rim inhomogeneity and core zoning in garnet 157 5.2 Summary table of pressures and temperatures for microprobed samples. 175 C. l Pressures and temperatures calculated from data of Holdaway et al. (1988).253 C.2 Pressures and temperatures calculated from data of Pigage (1982) and Hodges and Spear (1982) 254 C.3 Temperatures calculated for 6 pairs of samples from data of Chipera and Perkins (1988). Sample pairs are from the same or adjacent localities. . 255 x D.l Elements analyzed, standard number (U.B.C. microprobe catalog), the standard nature (rough mineral name), the unknown sample (map index #) for which each standard was used, and detection limits for each element.257 D.2 Selected garnet, biotite, and muscovite analyses for sample JSG-81-234 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d. = below detection limit given in table D.l) . . . . 258 D.3 Selected garnet, biotite, and muscovite analyses for sample JSG-81-278 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d. = below detection limit given in table D.l) . . . . 259 D.4 Selected garnet, biotite, and muscovite analyses for sample JSG-80-30 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 260 D.5 Selected garnet, biotite, and muscovite analyses for sample DM-85-9 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d. = below detection limit given in table D.l) . . . . 261 D.6 Selected garnet, biotite, and muscovite analyses for sample PDL-447 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 261 D.7 Selected garnet, biotite, and muscovite analyses for sample DM-85-42 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 262 D.8 Selected garnet, biotite, and muscovite analyses for sample SLG-370A which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 263 xi D.9 Selected garnet, biotite, and muscovite analyses for sample DM-85-87 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 264 D.10 Selected garnet, biotite, and muscovite analyses for sample CJNF-9 which yield the 'typical' P and T (first population) given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in D .H Selected garnet, biotite, and muscovite analyses for sample CJNF-9 which yield the 'typical' P and T (second population) given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit D.12 Selected garnet, biotite, and muscovite analyses for sample DM-86-70 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 266 D.13 Selected garnet, biotite, and muscovite analyses for sample DM-86-173 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 267 D.14 Selected garnet, biotite, and muscovite analyses for sample DM-86-197 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 268 D.15 Selected garnet, biotite, and muscovite analyses for sample DM-86-205 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 269 D.16 Selected garnet, biotite, and muscovite analyses for sample DM-86-227 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 270 table D.l) 265 given in table D.l) 265 xii D.17 Selected garnet, biotite, and muscovite analyses for sample DM-86-270 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 270 D.18 Selected garnet, biotite, and muscovite analyses for sample DM-86-281 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 271 D.19 Selected garnet, biotite, and muscovite analyses for sample DM-86-351 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 272 D.20 Selected garnet, biotite, and muscovite analyses for sample DM-86-369 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 273 D.21 Selected garnet, biotite, and muscovite analyses for sample DM-86-384 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 274 D.22 Selected garnet, biotite, and muscovite analyses for sample JRM-0626-36 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 275 D.23 Selected garnet, biotite, and muscovite analyses for sample JRM-0716-76 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 276 D.24 Selected garnet, biotite, and muscovite analyses for sample LCP-40 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 277 xiii D.25 Selected garnet, biotite, and muscovite analyses for sample LCP-223 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 278 D.26 Selected garnet, biotite, and muscovite analyses for sample DM-87-76 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 278 D.27 Selected garnet, biotite, and muscovite analyses for sample DM-87-71 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 279 D.28 Selected garnet, biotite, and muscovite analyses for sample DM-87-71 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 280 D.29 Selected garnet, biotite, and muscovite analyses for sample DM-86-45 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 281 D.30 Selected garnet, biotite, and muscovite analyses for sample JAF-7-9-29 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 281 D.31 Selected garnet, biotite, and muscovite analyses for sample JAF-7-9-29 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 282 D.32 Selected garnet, biotite, and muscovite analyses for sample DM-87-8 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 283 xiv D.33 Selected gaxnet, biotite, and muscovite analyses for sample DM-87-8 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 283 D.34 Selected garnet, biotite, and muscovite analyses for sample DM-87-17 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 284 D.35 Selected garnet, biotite, and muscovite analyses for sample DM-87-17 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 285 D.36 Selected garnet, biotite, and muscovite analyses for sample DM-87-24 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 286 D.37 Selected garnet, biotite, and muscovite analyses for sample DM-87-24 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 286 D.38 Selected garnet, biotite, and muscovite analyses for sample DM-87-32 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 287 D.39 Selected garnet, biotite, and muscovite analyses for sample DM-87-35 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 288 D.40 Selected garnet, biotite, and muscovite analyses for sample DM-87-39 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . . . . 289 xv D.41 Selected garnet, biotite, and muscovite analyses for sample JKR-88-4 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 291 D.42 Selected garnet, biotite, and muscovite analyses for sample JKR-88-4 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) . 291 xvi List of Figures 1.1 Map of Quesnel Lake area in southwestern British Columbia showing localities examined 2 1.2 A: Outcrop with foliated gneiss boulder. B: Polished foliated gneiss fragment 5 1.2 C: Outcrop of serpentine-talc clasts. D: Polished slab with quartzite pebbles and fuchsite-rich patches 6 1.3 A: Polished slab with graphitic phyllite fragment. B: Photomicrograph of graphitic phyllite fragment 7 1.4 A: Polished banded chert pebble. B: Polished grit pebbles 10 1.4 C: Photomicrograph of grit pebble. D: Photomicrograph of mylonitized grit 11 2.1 Sketch map of the Quesnel Lake area showing the areas of previous theses. 18 2.2 Tectono-stratigraphic units of the Quesnel Lake area 20 2.3 Mutually intersecting thrust faults shown in section E - E ' of Struik (1988a). 39 2.4 Section C-C' -C" reproduced from Rees (1987) 60 4.1 Equilibrium curves calculated from A: an equilibrium assemblage and B: a disequilibrium assemblage 93 4.2 Equilibrium curves calculated from a database containing poor data for one phase 95 4.3 Equilibria between Almandine, Annite, Muscovite, Phlogopite, Pyrope, Quartz and Sillimanite 97 xvii 4.4 Adjustments to experimental brackets used in MAP 105 4.5 Brackets as defined for natural assemblages: A: P and T defined by GASP and garnet-biotite, or A^SiOs and garnet-biotite thermometer and B: P and T close to A^SiOs boundary 107 4.6 Pressures and temperatures for data of Pigage (1982) 113 4.7 Map showing area! distribution of pressures calculated for data of Pigage (1982) 115 4.8 Pressures and temperatures calculated for the data of Holdaway et al. (1988) 116 4.9 Pressures and temperatures calculated for the data of Hodges and Spear (1982) 118 4.10 Map showing SGAM and GASP pressures for data of Hodges and Spear (1982) 119 5.1 Metamorphic isograds for the Quesnel Lake area of Campbell (1978). . . 125 5.2 Photomicrograph of muscovite phyllite showing two foliations, S2 and S 3 . 128 5.3 Photomicrograph of a garnet schist sample PDL-473 of Lewis (1987) showing S2 bent around F 3 folds 129 5.4 Photomicrograph of muscovite growth parallel to S3 in the hinges of F 3 folds 130 5.5 A: Garnet almost entirely replaced by sericite-chlorite mix. B: Kyanite replaced by sericite 132 5.6 Sketch of complex inclusion pattern seen in sample PDL-473 of Lewis (1987) 133 5.7 Photomicrograph of inclusion patterns within garnet in sample DM-86-154 134 xviii 5.8 Photomicrograph of garnet with two stages of growth, each of which has an associated inclusion pattern (sample DM-85-16) 135 5.9 Photomicrograph of small quartz inclusions in garnet in sample DM-87-54 from unit 3 137 5.10 Photomicrograph of randomly oriented mica inclusions in garnet from sample DM-87-35 from unit 3 138 5.11 Sketch of retrograded staurolite with garnet and opaque grain inclusions defining S2. S3 postdated staurolite 140 5.12 Photomicrograph of sample DM-86-173 showing quartz inclusions with-in staurolite asymptotic to external foliation, S 2 / S 3 141 5.13 Photomicrograph of part of staurolite grain from sample DM-87-54 showing two distinct zones 143 5.14 Photomicrograph of sample DM-87-34 showing crenulation cleavage, S 3 , wrapping around staurolite grain 144 5.15 Photomicrograph of part of sample JVR-2 (RadlofF, 1989) showing zoned staurolite grain microboudinaged with the extension direction in the S3 plane 145 5.16 Photomicrograph of sample DM-85-42 showing a large poikiloblastic kyanite grain 147 5.17 Sketch of microboudinaged kyanite grain from sample DM-87-42. Grain is aligned parallel to S3 with remanent S2 between the domains of S 3 . . 148 5.18 Photomicrograph of andalusite grain in sample JSG-81-278 151 5.19 Photomicrograph of andalusite grain in sample DM-87-19 partly enclos-ing a kyanite grain 152 5.20 Photomicrograph of sample DM-87-17 showing andalusite partly enclos-ing a symplectite of biotite and sillimanite (after garnet?) 153 xix 5.21 Gaxnet zoning profiles for samples JSG-81-278, DM-85-42, and SLG-370A 159 5.22 Garnet zoning profiles for samples DM-85-87, DM-86-70, and DM-86-384 160 5.23 Garnet zoning profiles for samples JRM-0716-76, DM-87-8, and D M -87-17 161 5.24 Garnet zoning profiles for samples DM-87-24, DM-87-32, and DM-87-54.162 5.25 A: Plot of Aim and Prp concentrations versus Sps concentration for 'rim' analyses from sample DM-87-24. B: Plot of Mg/(Mg+Fe) ratio versus Sps content for same analyses 165 5.26 Plot of Aim and Prp concentrations versus Sps concentration for rim analyses from sample DM-87-17 166 5.27 Pressure - Temperatures plot showing average P-T values for all samples listed in table 5.2. Numbers are index #'s used on map (in back pocket). 177 5.28 Trends in the values of P and T calculated using garnet with resorption zoning modified by diffusion. A: Sample #27 (JAF-7-9-29). B: Sample #30 (DM-87-24) 183 5.29 Areal distribution of the average pressures listed in table 5.2 190 5.30 Areal distribution of the average temperatures listed in table 5.2 191 5.31 Values of pressure from domain 3, as projected orthogonally onto line A-A' shown in figure 5.29 192 5.32 Values of temperature from domain 3, as projected orthogonally onto line A-A' shown in figure 5.30 193 5.33 Hypothetical P-T-t paths taken by two samples that undergo folding followed by uplift to the exposed surface 195 xx Acknowledgements There are many people who have helped at every stage of this thesis but first and foremost I thank Hugh Greenwood for his guidance, patience, good humour and enthu-siasm. There were some dangerous shoals and I needed the help of an expert navigator. Thanks also to Rob Berman for his patient help with many stages of LIP fitting and for keeping my thinking less fuzzy than it might have been. Thanks to John Knight for much cheerful help in killing bugs on the electron microprobe. To John Ross for help in seeing clearly in the field. Other help in the field from Jeff (all the coffee he could drink) Fillipone, Len (flip the boat over) Gal and Brian (that bear ate my down parka) Pataky. I also want to thank my fellow graduate students for much useful and learned (sometimes) discussion particularly Steve Garwin, Bruce James, Steve Juras, Peter Lewis, Urs Mader, Judy Radloff and Cliff Stanley. There have been almost hundreds of people who have helped not just with the thesis but in keeping my head on an even keel. I can't name you all. Particular thanks go to Steve and Maryann Juras, James Ceaser, Cliff and Sonya Stanley, Louise Wootton, Jeff Fillipone, Mary Anne Bloodgood, Urs and Ursula Mader, Paul and Dawn Holland and Robert and Janet Reid. I also want to thank Pauline Ascoli for just the right mix of bomb-scares, travel, skating, skiing and just enough insanity to keep me laughing in the past two years. Permission to work in the Wells Grey Provincial Park was facilitated by Tom Bell, permit officer, Parks and Outdoor Recreation Division, Ministry of Lands, Parks and Housing. Funding for this study came from NSERC grants to H.J. Greenwood (A-4222), R.G. Berman (OGP0037234), and J.V. Ross (A-2134). I also received additional xxi support in the form of a UBC Summer University Graduate Fellowship and a 2-year University Graduate Fellowship. Last, but not least, I want to thank my parents and brothers and sisters, who, though far away, were always at my side. xxii Preface By 1984 a large number of workers, most of them graduate students at U.B.C., had examined various portions of the Quesnel Lake area of south-central British Columbia. Since 1984 several other studies have been completed. Many of these works concen-trated, of necessity, on the stratigraphic and structural problems of that area. All examined the metamorphism to some extent but only a few examined it in any detail. Therefore, it was felt that a study of the metamorphism on a regional scale was war-ranted. Such a study would use the stratigraphic and structural detail of the earlier works as well as the wealth of collected specimens in the U.B.C. thesis collection. This was the idea behind the origin of this thesis. The study was originally envisioned in three parts; a stratigraphic and structural synthesis; a detailed petrographic study of a large number of collected specimens, thereby producing an extensive and detailed picture of the metamorphic grade and zonation patterns; and a quantitative study utilizing geothermometry and barometry linked to the petrographic study to produce a pressure-temperature-time history. A variety of problems encountered during the study have changed the emphasis of the resulting thesis. Firstly, the preparation of a stratigraphic and structural synthesis was hampered by extensive contradiction and inconsistency within the literature (both published and unpublished). It was not possible to do additional fieldwork on the large area involved (7,500 km2) and thus key localities were sought which would help elucidate problems. The data from two such localities and the initial interpretations based thereon are presented in chapter 1. These data helped in the re-evaluation of the extensive literature xxiii of the Quesnel Lake area. Much of the contradiction and inconsistency seen in the literature is the result of fitting observations to preconceived tectonic models. The data presented in chapter 1 suggest a different tectonic model for the area to that used previously and using this model it is possible to reconcile opposing viewpoints presented in the literature. Chapter 2 presents the re-evaluation of the data for the area and presents a new tectonic model. Many of the problems of contradiction and inconsistency appear to arise from poorly defined terminology. Most of the units within the area are not formally named or described thus it is often unclear if two authors using the same name are in fact referring to the same unit. In addition, the use of the term 'terrane' for several of the tectonic units is deemed incorrect as these units lack the appropriate definitive features. Discussion of several of these previously defined 'terranes' and their bounding 'faults' forms the bulk of chapter 2. From its inception, this study focussed on methods of geothermometry and barom-etry. However, standard methods of thermometry and barometry have a number of drawbacks. In particular, there are only two geobarometers commonly used for pelitic rocks of amphibolite grade and both have limitations. One is of limited precision, the other requires an assemblage that is neither particularly common nor easy to recognise. The recent developement of an extensive data base of thermodynamic data and soft-ware to compute equilibria (Berman et al., 1987; Berman, 1988, 1990) presented some alternative approaches to thermometry and barometry. Various mathematical prob-lems and inconsistencies are eliminated and the software presented the opportunity to calibrate a new barometer. In doing this, it was necessary to develop standard state properties for Fe-biotite (annite), which were not present in the data b 3>S€ ^  3>S well as to calibrate activity models for micas. The calculation of ideal activity for end-members in micas is difficult because of xxiv complex ionic substitution. Chapter 3 presents a new algorithm and program that gives values for the ideal activities of components of mica solutions. The standard state properties of annite and Margules parameters for biotite were determined us-ing mathematical programming techniques on various experimental and natural data. These permit the use of a new geobarometer. The methods used, the resulting ther-modynamic properties, and solution parameters are presented in chapter 4 along with a discussion of the advantages of the new barometer. Chapter 5 focusses on the metamorphic petrology of the Quesnel Lake area. About 1000 thin sections were examined and the data collected is summarized in map form. A suite of 35 samples was chosen for detailed microprobe analysis and the application of the new geobarometer. The calculated pressures and temperatures show a number of interesting features, particularly gradients, which support some tectonic models for the area. Oral presentations on several aspects of the thesis have been given at Saskatoon, St. John's, and Vancouver (McMullin and Greenwood, 1987, 1988b; McMullin and Berman, 1990). From its inception, it was intended that this thesis be written as a series of papers. Each chapter is written with this in mind. Chapter 1 has already been submitted and accepted (May, 1990) for publication in Geology. Chapter 2 will be submitted, with some modification, to the Canadian Journal of Earth Sciences, proba-bly in late 1990. Chapter 3 will probably be submitted to the American Mineralogist. The material in chapter 4 was the subject of an oral presentation at the Greenwood Symposium on Quantitative Methods in Petrology (McMullin and Berman, 1990) and chapter 4 is to be submitted to the Canadian Mineralogist for inclusion in the special volume commemorating this event. Chapter 5, after slight modification, will probably be submitted to the Canadian Journal of Earth Sciences or the Journal of Metamorphic Geology. xxv The five chapters are intended as separate papers. Therefore there is some repetition among them. Chapters 1 and 4, which are closest to publication, have individual abstracts. To increase readability, the references for all five chapters are placed in a single bibliography. The pagination, table of contents, and lists of figures and tables treat the thesis as a unit in order to conform to library requirements. xxvi C h a p t e r 1 P e b b l e s f r o m B a r k e r v i l l e a n d S l i d e M o u n t a i n t e r r a n e s i n a Q u e s n e l t e r r a n e c o n g l o m e r a t e : E v i d e n c e for p r e - J u r a s s i c d e f o r m a t i o n o f the B a r k e r v i l l e a n d S l i d e M o u n t a i n t e r r a n e s 1 1.1 A b s t r a c t Rocks of the Quesnel Lake area belong to three terranes. These are, from east to west, the Barkerville terrane (a continental prism sequence), the Slide Mountain terrane (an ocean-floor sequence), and the Quesnel terrane (an island arc-marginal basin sequence). The major deformation of these rocks occurred during the Jurassic. There has been renewed discussion recently as to whether the Barkerville terrane was deformed prior to the Jurassic. Two conglomerate localities within the Quesnel terrane contain clasts we identify as being derived from deformed rocks of the Barkerville and Slide Moun-tain terranes. These lithologies include gneiss, orthoquartzite, graphitic phyllite, and grit from the Barkerville, and serpentine-talc and chromite fragments from the Slide Mountain terrane. Some clasts inferred to be from the Barkerville terrane show two predeposition foliations implying two phases of deformation prior to the deposition of the conglomerate (Middle Jurassic at the latest). One of these events may be deforma-tion associated with the intrusion of Devonian granitoid bodies. Deformation must also have accompanied the emplacement of the Slide Mountain terrane some time between its deposition (Mississippian - Permian) and its erosion (Triassic - Jurassic). 1 Chapter to be published in 1990 as a paper in Geology, volume 18, co-authored by D . W . A . McMul l in , H . J . Greenwood, and J . V . Ross. 1 Chapter 1. Foliated pebbles in Quesnel Terrane 2 Figure 1.1: Map of Quesnel Lake area in southeastern British Columbia showing locali-ties examined. Terrane boundaries from Struik (1986a, 1988a). Inset map shows study area and tectonic belts of Canadian Cordillera from Wheeler and Gabrielse (1972). Chapter 1. Foliated pebbles in Quesnel Terrane 3 1.2 Introduction Rocks from the Quesnel Lake area of British Columbia (Fig. 1.1) belong to the Bark-erville, Quesnel, and Slide Mountain terranes (Struik, 1986a, 1988a; Monger et al., 1982; Monger, 1984). The Barkerville terrane is a continental margin sequence (Struik, 1986a) of Hadrynian to Paleozoic clastic sediments and carbonates. The Quesnel ter-rane is an island-arc marginal basin sequence (Struik, 1986a) of volcanic and fine-grained sedimentary rocks of Triassic and Jurassic age. The Slide Mountain terrane is an ocean-floor sequence (Struik and Orchard, 1985; Struik, 1988a) consisting of pillow basalt, ultramafic rock, and chert of Mississippian-Permian age. Major deformation of these terranes occurred during the Jurassic (Monger, 1984). In the Quesnel Lake area, thin slivers of mafic and ultramafic rocks at the base of the Quesnel terrane are called Crooked amphibolite (Struik, 1986a) and are correlated with the Slide Moun-tain terrane (Struik, 1988a). Both the Quesnel and Slide Mountain terranes have been interpreted to be thrust onto the Barkerville along the Eureka thrust (Struik, 1988a). One of the major points of discussion about the rocks of the Quesnel Lake area is whether there is a pre-Jurassic phase (or phases) of deformation in the Barkerville terrane (e.g., Bloodgood, 1987a; Fillipone and Ross, 1990; Garwin, 1987; Rees, 1987; Ross et al., 1985; Struik, 1988a). About 400 km southeast of the Quesnel Lake area such an extra phase of deformation in rocks equivalent to the Barkerville (Struik, 1986a) was first shown by Wheeler (1966). See also Klepacki and Wheeler (1985) and Klepacki et al. (1985) and references therein. The arguments in favor of pre-Jurassic deformation in the Barkerville include mainly detailed structural observations (Ross et al., 1985, 1989; Fillipone and Ross, 1990). Arguments against are based on regional stratigraphic correlations and structural interpretations (e.g., Rees, 1987; Struik, 1987, 1988a). The presence or absence of an extra phase of deformation in the Barkerville is of fundamental Chapter 1. Foliated pebbles in Quesnel Terrane 4 importance in the interpretation of metamorphic textures, mineral growth histories, and, ultimately, the tectonic history. At present, the pros and cons of an extra phase of deformation are argued from different standpoints (detailed structural analysis of small areas versus regional stratigraphic and structural interpretations). Thus, the discovery reported here of unambiguous stratigraphic evidence of pre-Jurassic deformation helps clarify the matter. 1.3 Conglomerate Localities 1.3.1 Wingdam The first conglomerate we examined crops out on Highway 26 at Wingdam, 50 km east of the town of Quesnel (Fig. 1.1). It was described by Struik (1988a) as a thin (15 m) subunit near the base of Struik's Triassic Black Phyllite unit. Struik (1988a, p. 79) described the conglomerate as being ". . . light grey consisting of rounded pebbles to small boulders of quartzite and smaller clasts of the same with quartz and feldspar. The matrix is mainly of silica and locally dolomite . . . ." Struik's (1988a) Triassic age for the phyllite unit was based on correlation with a similar phyllite at Spanish Lake (Struik, 1988a, 1988b). However, the conglomerate may be of Jurassic age. About 3.5 km southeast of Wingdam, the conglomerate unit appears to become a finer grained feldspathic, fossiliferous sandstone or siltstone (Struik, 1981b, 1985c, 1990, personal commun.). The fossils are characteristic of the Early Jurassic but range in age from Late Triassic to Middle Jurassic (Struik, 1981b, 1990, personal commun.). Note also that this fossiliferous sandstone rests with an apparent angular unconformity on strongly foliated Slide Mountain terrane serpentinite (Struik, 1981b, 1985c, 1990, personal commun.). Our observations of the lithologic character of the conglomerate at Wingdam diifer from those of Struik (1988a) in that it contains a variety of pebble types not detailed by Chapter 1. Foliated pebbles in Quesnel Terrane 5 Figure 1.2: A: Outcrop at Wingdam showing foliated gneiss boulder (30 cm across) with pegmatite veining. Also visible are quartzite pebbles of various sizes set in sandy matrix. Bedding and cleavage (steeper) dip to left (northwest). B: Polished slab show-ing foliated gneiss fragment discordant to foliation in groundmass (coin diameter 18 mm). D Figure 1.2: C: Outcrop showing light colored serpentine-talc clasts (lens-cap diameter 5.5 cm). D: Polished slab showing small foliated and bedded quartzite pebbles and dark fuchsite-rich patches in normally tan groundmass (coin diameter 18 mm). Chapter 1. Foliated pebbles in Quesnel Terrane 7 B 0.5 mm Figure 1.3: A: Polished slab showing graphitic phyllite fragment. Youngest foliation (F3) is crude pressure-solution cleavage (horizontal) cutting both groundmass and peb-ble. Earlier foliation ( F 2 , vertical) in pebble is crenulation cleavage. Note F 2 fold hinges between cleavage domains at top and lower right of sample (coin diameter 18 mm). B: Photomicrograph of graphitic phyllite fragment showing Fi buckled into F 2 folds (F 2 crenulation cleavage is vertical). Chapter 1. Foliated pebbles in Quesnel Terrane 8 Struik (1988a). Figure 1.2A shows part of the outcrop illustrated by Struik (1988a, Fig. 58, p. 79). The large (30 cm) boulder in Figure 1.2A (not shown in Struik, 1988a, Fig. 58) is a fragment of foliated gneiss with a quartz-feldspar pegmatite vein. In addition, there are rare nonfoliated fine-grained granitoid (aplitic) clasts, fairly abundant frag-ments of dark gray graphite-quartz-muscovite phyllite (Fig. 1.3A), and talc-serpentinite fragments (Fig. 1.2C). The groundmass of the conglomerate is tan to red-brown and consists of granule-size (1-4 mm) grains of quartz and feldspar with finer (< 1 mm) muscovite and ankerite. The micaceous matrix of the conglomerate is bright green around dark green to black (ultramafic?) fragments (Fig. 1.2D). Many of the fragments show one or more internal foliations discordant to the folia-tion in the matrix (Fig. 1.2B, 1.3A, 1.3B). Figure 1.2B shows a foliated gneiss cobble, which consists of quartz, feldspar (plagioclase and K-feldspar), and biotite. The sur-rounding matrix consists of quartz, ankerite, chlorite, biotite, and very fine-grained Fe-oxides and hydroxides with a recrystallized, mylonitic texture near the gneiss cob-ble. Biotite in the matrix surrounding this clast is not in chemical equilibrium with the rest of the matrix and is probably the result of comminution of the gneiss fragment during the latest deformation. In hand sample, graphite-quartz-muscovite phyllite fragments (Fig. 1.3A) show a foliation at an angle to the foliation in the matrix. Thin section examination reveals three foliations in the phyllite fragments. The youngest (F 3), is a crude pressure-solution cleavage axial planar to small (1-3 mm) kinks parallel to the cleavage in the groundmass (horizontal in Fig. 1.3A). The next youngest foliation (F2) is the dominant banding seen (Fig. 1.3, A and B). It is a crenulation cleavage and clearly folds an earlier foliation (Fig. 1.3B). The oldest foliation (Fi), defined by parallel alignment of tiny mica grains, is clearly visible between F 2 cleavage domains where it is bent around F 2 fold hinges (Fig. 1.3B). Chapter 1. Foliated pebbles in Quesnel Terrane 9 The bright green patches seen in hand sample (Fig. 1.2D) are fuchsite-rich reaction haloes around detrital chromite fragments. 1.3.2 Quesnel Lake The second conglomerate we examined crops out near the town of Likely at the west end of Quesnel Lake (Fig. 1). It was examined by Campbell (1978), Rees (1987), and Bailey (1988). Campbell (1978) described it (his lmJs unit) as conglomerate (with local granitic clasts), graywacke, and shale with an Early to (?)Middle Jurassic age (Pliensbachian to (?)Bajocian fossils). Rees (1987) noted clasts up to 20 cm across of chert, limestone, and various volcanic lithologies. He also noted possible plutonic clasts and white and gray "quartz clasts". Bailey (1988, p. 150) described the conglomerate (his Unit 6) as clast-supported and containing chert, limestone, argillite, sandstone, and greenstone. Our observations are similar to those of Bailey, Rees, and Campbell with some dif-ferences. Clasts are well rounded and most of them appear to be fragments of banded chert (Fig. 1.4A) and volcanic rocks. In addition, there is a significant number of pebbles of cream to pale gray, bedded and recrystallized orthoquartzite; white vein quartz streaked with red and green; and buff or gray felsic hypabyssal granitoid. Gran-itoid pebbles are rare and contain grains up to 3 mm across of altered euhedral mica (biotite?) and euhedral or slightly resorbed quartz set in a creamy, fine-grained or amorphous (altered feldspar?) groundmass. There are also cobbles of indurated grit (Fig. 1.4B). Most of these do not appear to have a foliation that predates deposition of the conglomerate (Fig. 1.4B), but we found one pebble of foliated grit. The grit consists of angular to subrounded white, gray or blueish quartz and red-dened feldspar(?) fragments up to 2 cm across set in a brown to purple groundmass (Fig. 1.4B). Most feldspathic fragments are highly altered (sericitized and hematitized) Figure 1.4: A: Cut and polished banded chert pebble from Quesnel Lake locality (coin diameter 18 mm). B: Cut and polished grit pebbles from Quesnel Lake locality (coin diameter 18 mm). D 1 mm • • • • • • • • • • • • • • • Figure 1.4: C: Photomicrograph of grit pebble showing slightly deformed quartz frag-ment (upper right), highly deformed, polycrystalline quartz (lower left), and highly altered feldspar (upper left). D: Photomicrograph of mylonitized grit. En-echelon quartz grains at bottom were originally single grain. Chapter 1. Foliated pebbles in Quesnel Terrane 12 and range from single feldspar grains to feldspar-quartz aggregates of igneous deriva-tion (Fig. 1.4C). Quartz grains vary from single crystals to polycrystalline aggregates to cryptocrystalline fragments (Fig 1.4C). Constituent grains within the grit show ex-treme variation in the degree of deformation (Fig. 1.4C) indicating derivation from a deformed source. In addition, cobbles of foliated and mylonitized grit (Fig. 1.4D) indi-cate deformation of the grit after its deposition but before its erosion and redeposition in the Quesnel terrane conglomerate. 1.4 Origin of the Clasts The clasts in the conglomerate at Wingdam must be derived from the nearby Bark-erville and Slide Mountain terranes. The clasts are neither well rounded nor particu-larly resistant (with the exception of the granitic clasts). Therefore they cannot have traveled far. All of the lithologies found in the pebbles are present in the subjacent Slide Mountain (serpentine-talc, chromite) and Barkerville (quartzite, gneiss, graphitic schist) terranes. May and Butler (1986) indicated that the Quesnel and the Slide Mountain terranes were not affected by significant postdepositional strike-slip (north-south) motion. Thus, the clasts cannot be derived from terranes now distant from the localities examined. The clasts cannot be matched to individual lithologies in the Barkerville and Slide Mountain terranes because of the later episode of metamorphism in the Jurassic (Pigage, 1977; Struik, 1988a; Mortensen et al., 1987). The origin of the clasts at Quesnel Lake is less certain. The clasts are well rounded, chemically stable, and hard, indicating transportation over long distances or in a high-energy environment. Volcanic and hypabyssal granitoid clasts may be from the Quesnel terrane volcanic package just below the conglomerate horizon (Rees, 1987; Bloodgood, 1988; Bailey, 1988). The limestone and argillite of Rees (1987) and Bailey (1988) may Chapter 1. Foliated pebbles in Quesnel Terrane 13 also be from underlying Triassic Black Phyllite (Bloodgood, 1988; Rees, 1987; Struik, 1988b). However, the quartzite, grit, and chert must be from another source. Rees (1987) and Bailey (1988) proposed a westerly derivation from the Permian Cache Creek terrane — a tectonic melange terrane with chert, argillite, limestone, and basalt. Our identification of quartzite and particularly of grit throws doubt on this interpretation because grits are not associated with the (oceanic) Cache Creek terrane. We propose an easterly provenance for the clasts. All of the clast lithologies are present to the east — chert, basalt (and argillite) in the Slide Mountain terrane (Struik and Orchard, 1985; Monger, 1984) and quartzite, grit, limestone, argillite, and granitoid rocks in the Barkerville terrane (Struik, 1986a). As at Wingdam, one cannot match clasts with individual lithologies in the Barkerville because of the later Jurassic metamorphism. 1.5 Conclusions Our main conclusion is that the Barkerville and Slide Mountain terranes were deformed before Middle Jurassic time. This supports the work of Ross et al. (1985, 1989) and Fillipone and Ross (1990) who noted that the earliest foliation visible in some Barkerville terrane rocks predates the deformation and metamorphism ascribed to the major Jurassic event. If the two foliations visible in the graphitic phyllite fragments are attributable to two distinct events, the first of these may be of Devonian age. Rb/Sr and U/Pb dating indicate that the Barkerville terrane was intruded in the Devonian by the Quesnel Lake gneiss (Getsinger, 1985; Mortensen et al., 1987; Rees, 1987; R. L. Armstrong, 1989, personal commun.). The second deformation may be of Permian or Triassic age. The presence of serpentinite fragments at the Wingdam locality indicates that the ultramafic rocks of the (Mississippian-Permian) Slide Mountain terrane were emplaced Chapter 1. Foliated pebbles in Quesnel Terrane 14 and uplifted prior to the Middle Jurassic. These discoveries must lead to a major reevaluation of the tectonic history of the Quesnel Lake area. The work of Struik (1987, 1988a and references therein) presents a picture of punctuated tectonism for this part of the Cordillera with a long period of tectonic quiet followed by the intense tectonic activity of the Jurassic. We propose that the tectonic activity was much more prolonged. A period of Devonian-Mississippian deformation was originally proposed by Sutherland Brown (1963) on the basis of his own and earlier observations but was discounted by Struik (1981a, p. 1774, conclusion #4). In addition, it has generally been accepted that the Quesnel and Slide Mountain terranes (and others) were joined together as a single superterrane (Terrane I of Monger et al., 1982; Intermontane belt shown in Fig. 1) before their accretion to the margin of North America in the Jurassic (Monger, 1984; Monger et al., 1982). The presence of deformed and metamorphosed(?) Slide Mountain and Barkerville terrane fragments at Wingdam indicate that the Slide Mountain terrane was not accreted to the Quesnel terrane but probably to the Barkerville before the deposition of parts of the Quesnel terrane. An important question arises from these conclusions. What is the nature of the contact between the three terranes? If the Slide Mountain terrane was emplaced on the Barkerville before the Jurassic then the basal contact (Eureka thrust?) must only be present where the slivers of Slide Mountain terrane remain. The contact between the Quesnel terrane and the Barkerville and Slide Mountain terranes is poorly exposed throughout the area and is described as tectonic in some places (e.g. Struik, 1988a; Bloodgood, 1988) and as depositional in others (Struik, 1981b, 1985c). Ultimately, the nature of all these terranes and their relationships at all scales must be reevaluated. Chapter 2 Stratigraphy and structure of the Quesnel Lake area: a synthesis and discussion 2.1 Introduction The Quesnel Lake area has an extent of about 7500 km2 and many of the rocks have been metamorphosed to amphibolite grade. In the past few years there have been many studies of the structural and stratigraphic history of the area and localized studies of the metamorphism. There has been little work done on the metamorphic history of the area as a whole. However, before a metamorphic synthesis can be prepared, a stratigraphic and structural framework is required. This chapter presents such a framework on the basis of re-evaluation of the existing documentation coupled with the new data presented in chapter 1. There is abundant published material on the Quesnel Lake area, most of which has appeared during the last twenty years. This material shows the normal progression of ideas and accumulated data, but also displays an unusual degree of internal contradic-tion and differences of opinion. The paragraphs that follow bring out these differences and contradictions in the course of developing an internally consistent synthesis of the structural and stratigraphic history. Most of the written material has appeared in various forums which do not provide rigorous peer review. The few peer-reviewed papers in formal journals tend to be "broad brush" tectonic syntheses relying heavily on the non-reviewed material. The 15 Chapter 2. Structure and Stratigraphy: A synthesis 16 non-reviewed material consists mainly of theses (mostly done at UBC) and the open file reports, preliminary maps, fieldwork reports and reports of activities of the Geological Survey of Canada and the British Columbia Ministry of Energy, Mines and Petroleum Resources. There are two underlying sources of confusion in the Quesnel Lake literature. The first results from the fact that none of the many map-units has been formally named and recognized. As a consequence, the large number of informal names and tentative correlations leads to doubts that different authors refer to the same units. The second results from a tendency to fit the observations to a preconceived tectonic framework. This has resulted in different authors working in different localities deducing structures that conflict with structures deduced by others, even when both assume the same tec-tonic framework. Consequently, the full set of data and inferences therefrom presented in the literature are not only internally inconsistent, they do not satisfy the originally assumed tectonic framework. This paper presents a new tectonic framework within which most of the observations made by recent workers in the Quesnel Lake area are consistent. In evaluating the material, the following set of criteria was followed. • Where there is conflict between an interpretive work and a descriptive one the descriptive work is given preference. • Where two works, which are equally descriptive, conflict in their interpretation, an interpretation which matches both descriptions is chosen or formulated. • Where two works contradict one another in their description, a third source, or the author's own observations are used to resolve the problem. • Work containing inconsistencies or self-contradiction is de-emphasized. Chapter 2. Structure and Stratigraphy: A synthesis 17 Much of the established wisdom on the geology of the Quesnel Lake area is ques-tioned in the following pages mainly due to the level of published contradiction and inconsistency. What is presented in its place is perhaps less complete than some of the 'pictures' presented in the literature but it is thought to be at least internally consistent. 2.2 Previous work The earliest mapping of the strata found in the Quesnel Lake area was done to the northwest in the Cariboo River, Yanks Peak - Roundtop Mountain and Antler Creek areas by Holland (1954) and Sutherland Brown (1957, 1963). They described and established the first stratigraphic sections for these strata and attempted to determine the nature of the deformation and of the mineralization in the area. Most of their work has been completely reinterpreted and is now mainly of historical interest. The earliest regional mapping in the Quesnel Lake area was done by R.B. Campbell of the Geological Survey of Canada in the late 1950's and early 1960's (R.B. Campbell, 1961; 1963). Since that time, more detailed work has been done by Campbell and his co-workers (K.V. Campbell & R.B. Campbell, 1969; R.B. Campbell, 1968, 1970, 1973, 1978; and R.B. Campbell et al., 1973, 1982). Additionally, detailed mapping has been done by a large number of M.Sc. and Ph.D. students mostly at the University of British Columbia (Bloodgood, 1987a, Carye, 1986, Elsby, 1985, Engi, 1984, Fillipone, 1985, Fletcher, 1972, Garwin, 1987, Getsinger, 1985, Lewis, 1987, J.R. Montgomery, 1985, Pigage, 1978 and Radloff, 1989 all at UBC; K.V. Campbell, 1971 at the University of Washington, Seattle; S.L. Montgomery, 1978 at Cornell University, Ithaca, New York; and Rees, 1987 at Carleton University, Ottawa, Ontario). Figure 2.1 shows the distribution of these thesis areas. Most of this work remains in the original unpublished thesis format but some has entered the published literature (Bloodgood, 1987b; Fillipone & Ross, Chapter 2. Structure and Stratigraphy: A synthesis 18 Figure 2.1: Sketch map of the Quesnel Lake area showing the areas of previous theses. 1: Bloodgood (1987a), 2: K.V. Campbell (1971), 3: Carye (1986), 4: Elsby, (1985), 5: Engi (1984), 6: Fillipone (1985), 7: Fletcher (1972), 8: Garwin (1987), 9: Getsinger (1985), 10: Lewis (1987), 11: J.R. Montgomery (1985), 12: S.L. Montgomery (1978), 13: Pigage (1978), 14: Radloff (1989), 15: Rees (1987). Inset map shows location of the Quesnel Lake area. Chapter 2. Structure and Stratigraphy: A synthesis 19 1990; Fletcher & Greenwood, 1979; Getsinger, 1982; Klepacki, 1981; Mortenson et ai, 1987; Pigage, 1982; Pigage & Greenwood, 1982; Rees, 1981; Rees & Ferri, 1983 and Ross et al., 1985, 1989). There has also been some recent work done on the Quesnel Lake area by members of the British Columbia Ministry of Energy, Mines and Petroleum Resources (Bloodgood, 1988; Bailey, 1988). Finally, L.C. Struik of the Geological Survey of Canada has contributed a large body of published work much of it in open file format (Struik, 1979, 1981a, 1981b, 1982, 1983a, 1983b, 1984, 1985a, 1985b, 1985c, 1986a, 1986b, 1987, 1988a, 1988b,1988c). 2.3 Terminology Examination of the literature for the Quesnel Lakes area indicates that there are three units of major areal extent, based on lithology, age, and geologic history. These three major units have been subdivided into, classified as, and correlated with various ter-ranes. Some of the rock units within the area have been labelled 'terranes' yet at least two of these 'terranes' lack one or more of the fundamental features of terranes as defined by Jones et al. (1983): "...fault-bounded geologic entities of regional extent, each characterized by a geologic history that is different from the histories of contiguous terr-anes " Several of the units in the Quesnel Lake area which have been called terranes are not fault-bounded and some of those that are fault-bounded are not noticeably distinct, in terms of geologic history, from their neighbour(s). Thus the three major units within the Quesnel Lake area are simply labeled units 1, 2, and 3 (fig. 2.2). Chapter 2. Structure and Stratigraphy: A synthesis 20 Figure 2.2: Tectono-stratigraphic units of the Quesnel Lake area. Unit 1: the continen-tal margin sequence, unit 2: the Crooked Amphibolite, unit 3: the Quesnel sedimen-tary and volcanic sequence. Inset map shows the tectonic subdivisions of the Canadian Cordillera of Wheeler and Gabrielse (1972). Chapter 2. Structure and Stratigraphy: A synthesis 21 2.4 Tectonic subdivisions of the Canadian Cordillera in the Quesnel Lake area — old and new The Canadian Cordillera has been divided into 5 major tectonic belts and numerous tectonic terranes (Wheeler & Gabrielse, 1972; Monger et al., 1982). The Quesnel Lake area straddles two of the major belts, the Intermontane Belt and the Omineca Belt, but includes mostly Omineca Belt rocks which display the highest grade of metamorphism. Unit 1 (fig. 2.2) is part of the 'Omineca Belt' and units 2 and 3 (fig. 2.2) are part of the 'Intermontane Belt'. The Intermontane Belt has been interpreted to be a collage of terranes which amal-gamated prior to its collision with the margin of North America in the Jurassic (Monger et al., 1982). The Omineca Belt has been described as the autochthonous or para-autochthonous clastic wedge deposited at the western margin of the North American craton (Struik, 1986a, 1987, 1988c) which was deformed and metamorphosed as a result of the collision between the two belts (Wheeler & Gabrielse, 1972; Monger et al., 1982; Struik, 1986a). In the Quesnel Lake area, the 'Omineca Belt' rocks (unit 1) have been further subdivided into two terranes, the Barkerville and Cariboo terranes (Struik, 1986a). The rocks of the 'Intermontane Belt' have been assigned to two terranes. Unit 2 has been called the Slide Mountain terrane and unit 3, the Quesnel terrane (Monger, 1984). However, this division of the rocks into two major belts and the identification of the major tectonic events associated with their collision has led to problems in reconciling observations made on the smaller scale (i.e. the observations of most of the workers listed in section 2.2). The most controversial topic has been the number of phases of deformation experienced by each of these two major belts. As has been shown in chapter 1, the rocks of unit 3 ("Quesnel terrane") probably lie in unconformable Chapter 2. Structure and Stratigraphy: A synthesis 22 contact with rocks of units 1 and 2. Thus the assignment of unit 1 to the Omineca Belt and units 2 and 3 to the Intermontane Belt is misleading. The identification of phases of deformation with the collision of the Intermontane Belt and Omineca Belt, as previously defined, is fraught with possible contradictions, particularly with regard to the correlation of various phases of deformation. The emplacement of unit 2 on unit 1 before deposition of parts of unit 3 requires that units 1 and 2 display more phases of deformation than unit 3. The separation of three different tectonic units permits the delineation of a slightly different geologic history for the Quesnel Lake area. The preferred sequence of geologic events for which the evidence is given in the following pages is as follows (starting with oldest): 1. Deposition of unit 1, the continental margin sequence, from Precambrian to upper Paleozoic time (Struik, 1988a). 2. Possible deformation of unit 1 in the Devonian(?) synchronous with intrusion of the Quesnel Lake Gneiss (chapter 1; Getsinger, 1985; Mortenson et al., 1987). 3. Deposition of unit 2, the Crooked Amphibolite, an ocean or marginal basin se-quence, in late Paleozoic time (Struik, 1988a; Struik &; Orchard, 1985). 4. Emplacement of unit 2 on unit 1 in late Paleozoic or early Mesozoic (Permian -Early Jurassic) time (chapter 1). 5. Erosion of units 1 and 2 and deposition of unit 3 in Early Jurassic time (chapter !)• 6. Deformation and metamorphism of all 3 units in Middle Jurassic (Pigage, 1977; Gerasimoff, 1988). Chapter 2. Structure and Stratigraphy: A synthesis 23 With this tectonic framework it is easier to reconcile some of the contradictions in the literature. The discussion in the following pages is divided into two parts. The first deals with the geology on the large scale, particularly the stratigraphy and the tectonic features that have permitted the division of the rocks into the three tectonic units. The second part deals mainly with the geology on the small scale, particularly the structural features, such as folds, foliations, etc., in order to present a structural framework within which the metamorphic history will be fit (chapter 5). These two parts are not divorced from one another but the division into two parts aids in the clarification of controversy. 2.5 New tectonic units — stratigraphy Detailed descriptions of the units and their boundaries is beyond the scope of this chapter. Where possible, a summary of the lithologic information available for each unit is given and the reader is referred to more comprehensive references on the topic. In the case of unit 1, only a very brief description of lithology is given for the reasons listed below. Units 2 and 3 are described in more detail. Some emphasis is given to the contacts within and between the various units because it is the understanding of the nature of these contacts that leads to a clearer picture of the stratigraphic and tectonic history of the area. Each unit is dealt with individually and thus a certain amount of repetition is unavoidable. 2.5.1 Uni t 1 — The continental margin sequence 2.5.1.1 Stratigraphy Unit 1, as shown in figure (2.2), is the most extensively described of all the units in the area. A large number of workers have generated a voluminous amount of descriptive Chapter 2. Structure and Stratigraphy: A synthesis 24 material on these rocks. Yet it is impossible to generate a single, detailed, stratigraphic picture for this unit because it is inhomogeneous on the small scale but homogeneous at larger ones. Inhomogeneity results from original lithological (facies) variation and from polydeformation and metamorphism. Thus, at all scales, units are commonly reported as showing extreme changes of present thickness (e.g., Getsinger, 1982, 1985; Garwin, 1987) due to variation in the original sedimentary thickness enhanced by de-formation. They also show variation in lithology along strike due to facies variations (e.g., Fletcher, 1972), and variations in mineralogy due to changes in metamorphic grade. Although inhomogeneous, the rocks over large areas are rather nondescript. Therefore, although almost all workers in the area have prepared detailed stratigraphic columns, the stratigraphic units and boundaries defined by each worker tend to be based on subtle lithologic differences. Consequently the stratigraphic columns for indi-vidual subareas cannot be correlated with the columns for adjacent subareas with any confidence. Rocks of unit 1 have been ascribed to three different lithologic groups, the Kaza, Cariboo, and Snowshoe Groups. The Snowshoe Group is of more recent informal definition (see appendix A, item 1) having been upgraded from a formation within the Cariboo Group (Struik, 1983a). Rocks of unit 1 are considered to have been part of the continental margin prism which was deposited from Proterozoic time to late Paleozoic or early Mesozoic time (Struik, 1987). The Kaza Group consists of coarse grained siliciclastic sedimentary rocks which are considered to grade upwards into the more pelitic and carbonate-rich rocks of the Cariboo Group. The Snowshoe Group is a heterogeneous siliciclastic, pelitic, and carbonate-bearing sequence. The Kaza and Cariboo Groups, and the Black Stuart Group, which lies to the north of the Quesnel Lake area, are considered to be the proximal portion of the continental margin prism. The Snowshoe Group is considered to be the heterogeneous distal portion. Most of Chapter 2. Structure and Stratigraphy: A synthesis 25 the area designated as unit 1 on figure 2.2 is underlain by Snowshoe Group although some of the rocks on the eastern side of the Quesnel Lake area have been ascribed to the Kaza and Cariboo groups. Because the Snowshoe Group is considered to be the lateral equivalent of the Kaza and Cariboo Groups it is difficult to say where one stops and the other starts, particularly at the grade of metamorphism shown by the rocks in the Quesnel Lake area. In recent years the rocks of the Snowshoe Group have been assigned to a separate terrane, the Barkerville terrane, and the Kaza, Cariboo, and Black Stuart Groups have been assigned to the Cariboo terrane. As is elaborated below, this distinction is difficult to justify because the two packages are remarkably similar in both stratigraphy and structure. Because of the lack of fossils they are also difficult to separate in the field. Therefore, the single designation, unit 1, the continental margin prism, is used here. This is sufficient to define a structural and tectonic framework on which to base a metamorphic history. Brief lithologic description of the three groups present in the Quesnel Lake area are given below. The reader is referred to other, more detailed works, in particular the M.Sc. and Ph.D. theses done in the area. 2.5.1.1.1 The Kaza Group The Kaza Group outcrops only at the easternmost part of the Quesnel Lake area (Fletcher, 1972; Pigage, 1978; and Engi, 1984). However, at least some of the strata at the western edges of their respective areas labeled as Kaza are now interpreted as Snowshoe Group (Struik, 1986a,b). Some of the earlier studies in the western parts of the Quesnel Lake area designated unit 1 rocks as Kaza Group. These rocks are now referred to as Snowshoe group (e.g., K.V. Campbell, 1971; S.L. Montgomery, 1978; Rees, 1981). The type Kaza Group was described by Sutherland Brown (1963) in the Bowron Lakes area as massive (60-90 m), thick-bedded, feld-spathic and micaceous quartzite with silver-green phyllite, schist and schistose granule conglomerate. In the Bowron Lakes area the Kaza has a present thickness of more than Chapter 2. Structure and Stratigraphy: A synthesis 26 3900 m. The descriptions of Fletcher (1972), Pigage (1978), Engi (1984) and Struik (1988a) follow this. At the higher metamorphic grades the lithologies are gneisses and schists. 2.5.1.1.2 The Cariboo Group Only two of the formations of the Cariboo Group are said to outcrop to any extent in the Quesnel Lake area, the Isaac and Cunning-ham Formations. Struik (1983b) noted small isolated outcrops of the Yankee Belle Formation in the east and north part of the area. Struik (1988a), gave the most recent descriptions of these formations as they occur at the northernmost part of the Quesnel Lake area and northward. He described the Isaac Formation as dark grey to black phyllite, slate, limestone and minor calcareous sandstone; the Cunningham Formation as limestone, dolostone and fine grained marble; and the Yankee Belle Formation as green-gray micaceous quartzite, siltite, grey-green shale, slate and phyllite, limestone and sandy limestone. Isaac Formation rocks have been described by Fletcher (1972), Pigage (1978), Struik (1983a,b) and Engi (1984). The Isaac Formation rocks described by Fletcher, Pigage, Engi and Struik (1983b) are at higher grade than those noted by Struik (1988a) and thus are mostly mica schist, marble, and calcareous schist. Cun-ningham Formation rocks are mapped in the Quesnel Lake area by Pigage (1978) and Struik (1983a) only. Pigage describes the Cunningham Formation as variable, massive to slabby, grey to white marble. Struik (1983b) indicated that the Cunningham Forma-tion consisted of limestone and dolostone. However, a brief examination by the writer in 1985 of some of the rocks designated by Struik (1983b) as Cunningham Formation showed them to be massive, well-crystallized marble not limestone. 2.5.1.1.3 The Snowshoe Group Snowshoe Group rocks make up most of the exposures of unit 1 in the Quesnel Lake area. There are no formal subdivisions of Chapter 2. Structure and Stratigraphy: A synthesis 27 this group. Struik (1988a) has studied the Snowshoe Group over a greater area than any other writer. The various theses present more detailed data on these rocks for different subareas. Struik (1988a) has presented 14 different informal subdivisions of the Snowshoe Group, referred to as 'successions'. Even at the lower metamorphic grade north of the Quesnel Lake area Struik (1988a, p. 47) noted that there were ". . . uncertainties concerning stratigraphic order..." of these informal successions. Struik (1988a) was not certain whether some successions were lateral equivalents of, or lay below or above others. Struik (1983a,b) extended sev-eral of these successions southwards into the Quesnel Lake area. These are the 'Ramos', 'Downey', and 'Bralco' successions. Struik (1983a,b) is the only worker to use these terms for the Snowshoe Group rocks in the area. Other workers have defined their own stratigraphic columns for their own small areas. It is not possible to correlate these stratigraphic columns. Compare, for example, the Snowshoe Group stratigraphy de-fined by Rees (1987), Getsinger (1985), Garwin (1987), Lewis (1987), J.R. Mongomery (1985), Elsby (1985), and Fillipone (1985). The single designation 'Snowshoe Group' is used here with no attempt to subdivide it. Over the Quesnel Lake area the Snowshoe Group shows some gross variation. To the west (the base) it consists of interlayered, micaceous quartzite, quartzose schist with minor impure carbonate, calcsilicate and amphibolitic gneiss (Getsinger, 1985; Lewis, 1987; Fillipone, 1985). This portion of the Snowshoe Group also includes extensive bodies of gneiss (see below). The eastern (up-per) portions of the Snowshoe Group are more carbonate-rich and contain various pure and impure marbles as well as calcsilicate and amphibolite units. Some of the marble units reach present thicknesses of 500 m (Getsinger, 1985; Garwin, 1987). The east-ern (upper) portions of the Snowshoe Group grade eastwards into the carbonate-rich Chapter 2. Structure and Stratigraphy: A synthesis 28 rocks assigned to the Cariboo Group (Struik 1983b). It is impossible to distinguish be-tween the rocks of the two groups, particularly at the amphibolite grade metamorphism experienced by these units within the Quesnel Lake area. 2.5.1.1.4 The Quesnel Lake Gneiss The lower part of the Snowshoe Group is intruded by large bodies of gneiss throughout the Quesnel Lake area. Three major bodies have been identified. The largest stretches from northwest to southeast across Quesnel Lake and has a total length of about 60 km. Parts of it have been studied by Rees and Ferri (1983), Rees (1987), Fletcher (1972) and J.R. Montgomery (1985). Smaller bodies have also been identified at Mount Perseus (Elsby, 1985) and Boss Mountain (Fillipone, 1985). All appear to be sill-like bodies and all appear to have the same age (Mortenson et al., 1987). They are therefore referred to collectively as Quesnel Lake Gneiss. In addition to the large bodies, small bodies (mostly sills) have been noted throughout the Snowshoe Group (Getsinger, 1985; Garwin, 1987; S.L. Mongomery, 1978; Radloff, 1989). Geochronologic studies have been hampered by the apparent large scale assimilation of radiogenic isotopes during intrusion of the gneiss and by redistribution of isotopes during metamorphism. Rb/Sr isotope studies have given very different ages for the Quesnel Lake Gneiss which have been attributed to • variable 8 7 Sr/ 8 6 Sr initial ratios, • assimilation of radiogenic 8 7Sr during emplacement, • remobilization of isotopes during metamorphism (Getsinger, 1985; Rees, 1987; Blenkinsop, 1972; Montgomery, 1985; Mortenson et al, 1987). Although U-bearing minerals (zircon, sphene, monazite) show some evidence of inheritance from source rocks and loss due to metamorphism, U/Pb data indicate a Chapter 2. Structure and Stratigraphy: A synthesis 29 Late Devonian to mid-Mississippi an age of 335 - 375 Ma (Getsinger, 1985; Okulitch, 1985; Mortenson et al., 1987). The Quesnel Lake gneiss shows considerable variation in composition (Fletcher, 1972; Getsinger, 1985; Montgomery, 1985; Montgomery and Ross, 1989). It varies in composition from diorite or quartz diorite to granite and syenite (Getsinger, 1985; Montgomery, 1985; Montgomery and Ross 1989). This expanded compositional range, the presence of hornblende in the mafic compositions and the relatively high Na content in comparison to K content (Fletcher, 1972; Montgomery, 1985; Montgomery and Ross, 1989) all indicate that the Quesnel Lake gneiss is an I-type intrusive unit (Chappell and White, 1974). Montgomery and Ross (1989) classified it as S-type on the basis of isotopic composition (high initial 8 7 Sr/ 8 6 Sr ratio) and the presence of muscovite in felsic units. These features, normally found in S-type granitoid rocks, must be discounted because, as noted above, the Rb/Sr systematics indicate considerable assimilation of radiogenic Sr during intrusion. Secondly, metamorphism would favour the production of subsolidus muscovite in these rocks. 2.5.1.2 Barkerville and Cariboo terranes and the Pleasant Valley Thrust In recent years, Struik (1982, 1983a, 1983b, 1985a, 1985b, 1985c, 1986a, 1986b, 1988a) has presented a tectonic picture of unit 1 as belonging to two terranes. The Snowshoe Group rocks have been ascribed to the Barkerville terrane and the Kaza and Cariboo Group rocks to the Cariboo terrane (Struik, 1983b, 1986a, 1988a). These terranes have been considered to be separated by the Pleasant Valley Thrust originally defined for rocks north of the Quesnel Lake area (Struik, 1982, 1986a, 1988a). As mentioned above, the use of the term terrane implies fault separation and a distinct geologic history for the separated packages. In the Quesnel Lake area the 'terranes' of unit 1, as they have been defined, and the fault separating them do not fit the criteria for such Chapter 2. Structure and Stratigraphy: A synthesis 30 a definition. Because Struik is the only author to have discussed the Barkerville and Cariboo terranes the following discussion concentrates on his work. The Pleasant Valley Thrust (originally the Pleasant Valley fault) was first described by Struik (1982) for rocks north of the Quesnel Lake area as having ". . . contrasting rock suites and structural complexites between its hanging and footwall successions. It... may continue southeast as the Little River Fault (Klepacki, 1980)" (sic). This would imply that the Pleasant Valley Thrust (fault) is a late or postmetamorphic fault, as is the Little River fault (Klepacki, 1981; Getsinger, 1985). The next reference to the structure is by Struik (1983a), a map of the northern part of the Quesnel Lake area with notes. Struik (1983a) stated that ". . . the structure and metamorphism vary across the east-dipping Pleasant Valley Fault..." and ". . . east of the fault the folds are open to tight on the macroscopic scale and isoclinal on the mesoscopic scale... " and ". . . west of the Pleasant Valley Fault the folds are tight to isoclinal " These statements do not permit unambiguous discrimination between rocks to the east and west. The distinction of different 'styles' of folding without reference to relative age is impossible. Furthermore, it is not clear, from the text, whether the Pleasant Valley Fault is an early or late (pre- or post-folding and metamorphism) structure. The map shows the Pleasant Valley Fault truncating the trace of the Little River Fault, which is known to be postmetamorphic as indicated by significant metamorphic difference across it (Klepacki, 1981; Getsinger, 1985). Struik (1983b), on a map with notes, referred to this structure as the Pleasant Valley Thrust and stated that it Chapter 2. Structure and Stratigraphy: A synthesis 31 ". . . emplaces Cariboo Terrane westward over Barkerville Terrane. The thrust intersects bedding at shallow angles and is folded isoclinally..." and ". . . as the final generation of metamorphic minerals formed after the last pervasive cleavage it is reasonable to assume that the fault was there prior to the freezing of those minerals." Thus, the Pleasant Valley Thrust is an early structure and predates much of the de-formation and metamorphism. Yet the map accompanying these descriptions shows the Pleasant Valley Thrust truncating the postmetamorphic Little River Fault. In the same notes, Struik (1983b) described the Cariboo and Barkerville 'terranes' as being ". . . distinguished by their stratigraphy " This is the first work to refer to these stratigraphic packages as terranes. The supposed terrane boundary, the Pleasant Valley Thrust, is described as a very early structure — thus the two 'terranes' it separates have been joined for a considerable part of their structural and metamorphic history. These cannot, then, be properly called terranes (i.e., fault-bounded bodies characterized by geological histories different from those of contiguous terranes). The writer examined in 1985 the outcrops at Mount Watt in the centre of the area covered by Struik (1983b). There is no evidence whatever of a structural break at the horizon indicated by Struik. The peak of Mount Watt is composed of a thick carbonate unit which is metamorphosed to a coarsely crystalline marble not noticeably different from other thick marble units in the area. Struik (1983b) named the thick marble on Mount Watt Cunningham Limestone, making it part of the Cariboo Terrane. The underlying garnet-kyanite schist and marble he called Snowshoe Group, making these rocks part of the Barkerville Terrane. It is not clear from the text Chapter 2. Structure and Stratigraphy: A synthesis 32 of Struik (1983b) whether the terranes are defined by the presence of a fault or the fault by the presence of 'terranes'. More detailed work in the Niagara Peak area, slightly to the south of Mount Watt, by Garwin (1987) also failed to show any evidence of a major thrust structure. In addition, Garwin showed that the marble units ascribed by Struik (1983b) to the Snowshoe Group (Bralco Marble) are compositionally inhomogeneous, discontinuous, and vary considerably in thickness due to both original sedimentary variations (facies changes) and to the degree of deformation. Garwin (1987, p. 25) also considered that the supposed 'truncations' ascribed by Struik to the Pleasant Valley Thrust are of sedimentary origin. The Pleasant Valley Thrust was described later by Struik (1985a) as ".. . a ductile fault separating Cariboo from underlying Barkerville Terrane. It is moderately dipping and classified as a thrust fault because it puts older over younger in the vicinity of Wells." In this description the definition of the fault is based on stratigraphic, not structural evidence. Struik (1985b) described rocks in the east and southeastern part of the Quesnel Lake area and defined ". . . the low-angle fault that carried Cariboo Terrane onto Barkerville Ter-rane is best exposed at the headwaters of Ovis Creek... where Cunningham and Isaac formations.. .are faulted onto grit, phyllite, marble and schist of the Barkerville Terrane. The low-angle fault is the same in style and relative age as the Pleasant Valley Thrust at Mount Watt near the North Arm of Quesnel Lake (Struik, 1983b). It was mapped by Pigage (1978) as a premetamorpbic slide." Chapter 2. Structure and Stratigraphy: A synthesis 33 Firstly, Pigage (1978, p. 6, 11, 23, 75, 77) did not map a premetamorphic slide — but inferred postmetamorphic movement on a premetamorphic (Fj) surface (perhaps a fold axial surface). Pigage (1978, p. 75) suggested up to 100°C difference across this 'slide' which is similar to the temperature difference noted by Klepacki (1981) across the Little River Fault. Secondly, careful examination of the maps of Struik (1985b) and Pigage (1978) shows that the surface trace shown by Struik is not in the same place as that defined by Pigage. Struik (1986a) presented an extended discussion of the terranes for the Quesnel Lake area. The discussion is prefaced with the statement that ". . . each fault-bounded stratigraphically distinct package throughout this paper is called a "terrane," which hereafter refers to a fault-bounded se-quence of stratigraphy whose relationship to sequences across the bounding faults is not known because of differences of stratigraphy and possibly struc-tural history. This definition follows that of Coney et al. (1980) " In fact, this definition is significantly different to that of Coney et al. (1980, p. 329) who stated ". . . the boundaries between terranes are fundamental discontinuities in stratigraphy that cannot be explained by facies changes or unconformity. Most boundaries separate totally distinct temporal or physical rock sequ-ences. . . ." Struik's (1986a) definition is also different from that of Jones et al. (1983) as quoted above (p. 19). Using Struik's modified definition of 'terrane' the Cariboo and Bark-erville terranes can then be defined in purely stratigraphic terms (see Struik, 1986a, p. 1054-1056). These are reported as being separated by the Pleasant Valley Thrust, Chapter 2. Structure and Stratigraphy: A synthesis 34 ". . . a moderate- to low-angle, east-dipping fault that places Cariboo terrane over Barkerville terrane. The fault is folded at Mount Watt. It has rocks of comparable metamorphic grade in both hanging wall and footwall. The inferred position of the thrust can be traced from northwest of Barkerville to east of Wells Gray Provincial Park." (italics mine). In addition Struik (1986a) stated that ". . . generally, the actual fault has not been located even where its position is well constrained by the angular intersection of the beds of the Cariboo onto Barkerville terrane... ." Thus the Pleasant Valley Thrust is apparently defined by the stratigraphy on either side. It is described as being layer-parallel (since it is isoclinally folded, Struik, 1983b) and in a photograph (Struik, 1986a, p. 1052) as a fault with large-scale, structural discordance across it. Struik (1986a) also correlated rocks ascribed to Cariboo and Barkerville terranes in the Wells - Barkerville area northwest of Quesnel Lake area to rocks at the northwest corner of Wells Gray Provincial park. This correlation was elaborated upon in a second paper (Struik, 1986b). Struik (1986b) correlated the Bralco marble of the Snowshoe Group (near Wells), which contains Paleozoic fossil fragments, with a similar marble, which contains no fossils, in Wells Gray Provincial Park. This correlation is over a distance of more than 140 km through strata that have been polydeformed and metamorphosed to sillimanite grade. However, the strata at Wells Gray Provincial Park were mapped by Pell (1984) as Precambrian Horsethief Creek and Kaza Groups. Struik (1986a, p. 1054 and 1986b, p. 591) stated that the correlation ". . . is a contradiction because the Horsethief Creek and Kaza groups are Chapter 2. Structure and Stratigraphy: A synthesis 35 thought to be entirely Precambrian and the parts of the Showshoe Group linked to them are Paleozoic." This is the only statement made by Struik (1986a,b) concerning this contradiction and the problem is neither discussed nor resolved. The implication is that the 'Horsethief Creek' strata of Pell (1984) are to be considered as Paleozoic Snowshoe Group, yet Struik (1986a,b) shows correlated stratigraphic sections with the original group names. Struik (1986a,b) went on to state that "...the Snowshoe Group through Wells Gray Provincial Park. ..is over-thrust by the Cariboo and Kaza groups... the overlying Hadrynian Kaza Group is, therefore, inferred to be thrust onto the "Horsethief Creek Group" in the southeastern Cariboo Mountains along the easterly extension of the Pleasant Valley Thrust." This appears to be assignment of stratigraphic names in order to define a structure. Struik (1986b p. 591 - 593) details 'outcrops' of the thrust strata. However, Murphy (1987 p. 741 - 742) visited the same outcrops and found no evidence in support of thrust (or any fault) in the position given by Struik (1986b). Most recently, Struik (1988a) focussed primarily on the stratigraphy of the various 'terranes' and the only significant reference to the Pleasant Valley Thrust (Struik, 1988a, p. 8) indicates that it is ". . . the contact between the Cariboo and Barkerville terranes... ." In addition Struik noted that the Pleasant Valley Thrust ".. .is marked by cataclastic rocks north of Cariboo Lake and by mylonite south of the lake..." and ".. . to the south, mylonitic rocks of the fault zone are exposed in the northeast bank of the Little River." Chapter 2. Structure and Stratigraphy: A synthesis 36 Here the fault is again defined on the basis of the rocks it separates. More importantly, this is the only work in which the thrust is described as being marked by cataclastic and mylonitic rocks. Yet, it has been described in previous works as being premetamorphic and isoclinally folded. The only mylonitic fabric seen during a short visit to the Little River Valley during this study is that associated with the postmetamorphic Little River Fault (Klepacki, 1981). The above paragraphs have detailed exhaustively the problems associated with the definition and description of the Pleasant Valley Thrust and thus with the definition of the Cariboo and Barkerville terranes. The problems are not restricted to a single reference on the subject but extend across all the documentation. In order to be a true terrane boundary the fault must separate bodies of rock that are distinct in terms of stratigraphy and structure. It does not. The two terranes share most, if not all, of their structural history. The rock units ascribed to the two terranes are more likely laterally equivalent units, where the Snowshoe Group strata represent the distal strata and the Kaza and Cariboo groups represent the proximal strata of a continental margin sequence. It is not necessary for a terrane boundary to be an explicitly observable structure — many are not. But as shown above, the features ascribed to the Pleasant Valley Thrust are contradictory. It is described as premetamorphic and isoclinally folded but also said to truncate postmetamorphic faults, cut strata at high angles, and have cataclastic and mylonitic fabrics. It seems more likely that the cataclastic and mylonitic fabrics (and high angle features) are postmetamorphic and are probably related to the postmetamorphic tectonic denudation of the high grade rocks by the Little River and related faults (Klepacki, 1981; Getsinger, 1985). Once these features are accounted for, the Pleasant Valley Thrust becomes entirely cryptic and is defined only where one can recognize, stratigraphically, the two terranes it supposedly separates. Therefore, Chapter 2. Structure and Stratigraphy: A synthesis 37 the evidence indicates that all the strata assigned to unit 1 in figure 2.2 belong to the same tectonic unit, namely a continental margin sequence, which has been shortened, thickened, intruded, and metamorphosed. There appears to be no irrefutable evidence of the Pleasant Valley Thrust in the Quesnel Lake area, nor of two distinct terranes. 2.5.1.3 Cross-sections and reconstructions There has been some discussion in the past concerning the presence of a pre-Jurassic phase of deformation in unit 1. One of the primary pieces of evidence against such a phase is the palinspastic reconstruction originally given by Struik (1981a) and repro-duced by Struik (1988a). Cross-sections A-A' of Struik (1981a) and E - E ' of Struik (1988a) and the palinspastic reconstructions based on them are for the same line of section in the northern part of the Quesnel Lake area. There are significant weaknesses regarding the construction of these cross-sections and reconstructions. The preparation of cross-sections is an interpretive process and there are several conditions that must be met to produce valid cross-sections. 1. Unit thicknesses must remain constant or vary in a predictable manner. 2. The orientation of various measured surfaces (bedding, foliation, etc.) must vary in a predictable manner. 3. Folds must have cylindrical geometry. 4. Detailed map data and, if possible, other three-dimensional data are required to provide constraints on strata that do not intersect the surface. These criteria are of particular importance if one is removing deformation in a palinspas-tic reconstruction. The fewer of these criteria that are met the less valid the cross-section and reconstruction therefrom. The cross-sections and reconstructions given Chapter 2. Structure and Stratigraphy: A synthesis 38 by Struik (1981a, 1988a) are at best cartoons because none of these criteria are met. Firstly, rocks in the Quesnel Lake area show extreme changes in layer thickness as a result of polydeformation and metamorphism (Getsinger, 1985; Lewis, 1987; Gar-win, 1987). Secondly, because of the isoclinal folding, surfaces (bedding, foliations etc.) have a single orientation and the presence of large scale features such as map-scale folds cannot be demonstrated in the absence of detailed lithological mapping. Thirdly, the folds are not cylindrical (Struik, 1988a, p. 42). Fourthly, map data are not of sufficient detail to determine the structure at depth. Struik stated (p. 42-43) that because the folds have a conical geometry and because of the paucity of outcrop, the cross-sections ". . . are schematic in relation to amplitude, wavelength and symmetry of included beds." Furthermore, section A-A' of Struik (1981a) and section E—E' of Struik (1988a) il-lustrate different structures yet the palinspastic reconstructions based on them are identical. Finally, in both both section A-A' of Struik (1981a) and section E - E ' of Struik (1988a), two thrusts are shown truncating each other. That is, one fault is shown cutting the second in one place and the second cuts the first in another (par-tially reproduced in figure 2.3). This structure is mechanically impossible. Because of these weaknesses little weight can be given to the palinspastic isopach maps (Struik, 1988a, p. 13,16) or 'restored cross-section' (Struik, 1981a, p. 1773; 1988a, p. 46). Interpretations based on them must also be treated with caution. Specifically, the interpretation of Struik (1981a, 1988a) based on these sections that there has been no significant Paleozoic deformation is unfounded. The evidence in favour of pre-Jurassic deformation for the Quesnel Lake area is presented in chapter 1. Chapter 2. Structure and Stratigraphy: A synthesis 39 B C Figure 2.3: A: Section E - E ' reproduced from Struik (1988a) showing mutually trun-cating faults at right of center, labelled X, Y, and Z. Note that the Mural Formation is unpatterned in the section not hatchered as shown in the legend of Struik (1988a). B: Relationship bf Faults X, Y, and Z as abstracted from A above. C: Removal of offset on fault Z shows fault X truncates fault Y at point 'a' and fault Y truncates fault X at 'b'. This relationship is shown by truncation of layers below and above these structures in A. Chapter 2. Structure and Stratigraphy: A synthesis 40 2.5.2 Unit 2 — The Crooked Amphibolite (CA) Unit 2, hereafter referred to as the Crooked Amphibolite (CA), has been studied by a number of different workers (table 2.1) each of whom offered a preferred name and correlation. It has been interpreted, at different times, to be part of both the Quesnel terrane (e.g., Struik, 1985a) and of the Slide Mountain terrane (e.g., Struik, 1988a). The ambiguity arises because the CA occupies the same tectonic postion as the Slide Mountain terrane but is markedly more deformed and metamorphosed. Further con-fusion has arisen in the past because workers have assumed, a priori, that the Quesnel and Slide Mountain terranes were part of the Intermontane Belt and that the emplace-ment of this on the Omineca Belt caused many of the structural and metamorphic features. In contrast to unit 1, unit 2 (CA) is an easily recognisable and mappable unit and the descriptions in the literature are much easier to condense. 2.5.2.1 Lithology The name Crooked Amphibolite was first used by Struik (1985a). The unit is dis-continuous and commonly less than 300 m thick but reaches 800-1200 m in the hinge zones of major folds (Campbell, 1971; Carye, 1986; Bloodgood, 1987a, 1987b; Rees, 1987). Struik (1986a) described the CA as consisting mainly of amphibolite with minor amounts of serpentinite and sheared ultramafic rocks. He assigned a Mississippian -Permian age on the assumption that the Crooked Amphibolite is the metamorphosed equivalent of the Antler Formation of the Slide Mountain Group. Over the Quesnel Lake area the CA is variable in both thickness and constituent lithologies. In the northwest (Spanish Lake area), it occurs only locally but reaches a maximum thickness of 800 m (Rees, 1987). Three lithologic types have been recognised (Rees, 1987): Chapter 2. Structure and Stratigraphy: A synthesis 41 Author Unit 2, the Crooked Amphibolite Campbell, K.V. 1971. Slide Mountain Group. Campbell, R.B. 1978. Partly Pennsylvanian - Permian Slide Mountain Group; partly Paleozoic or Mesozoic ultramafic intrusives; partly Paleozoic or Mesozoic Redfern Complex Montgomery, S.L. 1978. Upper Paleozoic Black Riders Complex. Tipper, H.W. et al. 1979. Partly Mississippian or younger Slide Mountain Group; partly Permian-Triassic Trembleur in-trusives. Rees, C.J. 1981. Mississippian Antler Formation (Slide Moun-tain Group). Rees, C.J. & Ferri, F. 1983. Upper Paleozoic Antler Formation. Struik, L.C. 1983a. Upper Paleozoic Slide Mountain Group. Elsby, D.C. 1985. Mississippian - Permian Slide Mountain Group. Fillipone, J.A. 1985. Antler Formation (Slide Mountain Group). Carye, J.A. 1986. Upper Paleozoic Antler Formation (Slide Mountain Group) Struik, L.C. 1986a. Crooked Amphibolite (possibly Antler Forma-tion equivalent). Bloodgood, M.A. 1987a. Upper Paleozoic Crooked Amphibolite. Bloodgood,M.A. 1987b. Mississippian - Permian Crooked Amphibolite. Rees, C.J. 1987. Upper Paleozoic Antler Formation (not part of Slide Mountain Terrane). Bloodgood, M.A., 1988 Mississippian - Permian Crooked Amphibolite. Struik, L.C. 1988a. Crooked Amphibolite (may be part of Slide Mountain Terrane). Struik, L.C. 1988b. Crooked Amphibolite (part of Slide Mountain Terrane). Fillipone, J.A. & Ross, J.V. 1990. Crooked Amphibolite. Radloff, J.K. 1989. Redfern Complex (equivalent to Crooked Am-phibolite). Table 2.1: Names and correlations given by previous workers to rocks of unit 2, the Crooked Amphibolite Chapter 2. Structure and Stratigraphy: A synthesis 42 1. Greenstone (chlorite - albite ± epidote ± quartz). 2. Metagabbro (chlorite - actinolite - hornblende, with relict cumulate texture). 3. Meta-ultramafic rocks (serpentinite - steatite). Rees (1981, 1987), Rees and Ferri (1983) and Struik (1983a) all referred to these rocks as Antler Formation and assigned an upper Paleozoic age (Mississippian - Permian). Rees (1987) referred to these rocks as Antler Formation but he distinguished them from the Antler Formation of the Slide Mountain Terrane to the north (Wells - Barkerville area, see Struik, 1988a). He did this because he considered that the CA at Spanish Lake is in depositional contact with the overlying Quesnel Terrane strata and therefore part of the Quesnel Terrane (see below). South of Quesnel Lake the CA appears to be more continuous. It has been described from Quesnel Lake to south of the Boss Mountain area (Campbell, 1971; Carye, 1986; Fillipone, 1985; Elsby, 1986; Bloodgood, 1987a). Thickness varies from less than 10 m (Campbell, 1971) to a maximum of 1200 m (Carye, 1986). Here the CA appears to be predominantly an amphibolitic unit and there are two dominant lithologies: 1. A lower, fine-grained chloritic schist (chlorite - amphibole ± feldspar; see Camp-bell, 1971; Fillipone, 1985; Bloodgood, 1987a). 2. An upper, darker, coarser-grained hornblende garbenschiefer (Bloodgood, 1987a) with sporadic garnet (Campbell, 1971). Elsby (1985) noted quartzose interlayers and Fillipone (1985) described quartz - feld-spar - amphibole - chlorite rocks. There are also isolated outcrops of ultramafic rocks (Campbell, 1971; Carye, 1986; Fillipone, 1985; Elsby, 1986; Bloodgood, 1987a,b; Mc-Mullin and Greenwood, 1988a). These are pods of serpentinite or steatite several tens Chapter 2. Structure and Stratigraphy: A synthesis 43 of meters long most of which appear to be tectonic inclusions at (or slightly below) the boundary between the Intermontane Belt and the Omineca Belt. A l l workers prior to 1986 referred to the rocks as Antler Formation and thereby assigned a Mississippian - Permian age. Bloodgood (1987a,b) referred to the unit as Crooked Amphibolite of upper Paleozoic age possibly equivalent to the Antler Formation. In the southernmost portion of the map area an isolated area of rocks interpreted as a klippe of the C A outcrops on the east side of Deception Creek. These rocks have been studied in detail by S.L. Montgomery (1978) and Radloff (1989). The klippe contains rocks interpreted as a tectonized and partly metamorphosed ophiolitic complex. It was named the Redfern Complex by Campbell (1978) but also called the Black Riders Complex by Montgomery (1978). Radloff (1989) preferred the former name. The complex is between 1500 m (Radloff, 1989) and 3000 m (Montgomery, 1978) thick. Radloff (1989) divided the Redfern Complex into 3 subunits: 1. Tectonized dunite (contains relict olivine and chromite and metamorphic serpen-tine, talc, tremolite/actinolite and anthophyllite). 2. Layered cumulates (pyroxenite and dunite interlayered on a scale of 10's of cen-timeters, altered and metamorphosed). 3. Amphibolite (variably textured, containing hornblende -f plagioclase ± quartz ± garnet). The boundaries of the subunits are tectonized with some intercalation of one sub-unit with another. Radloff (1989) interpreted the dunite as the top of the 'ultramafic complex' of an ophiolite, the layered cumulates as the lower portion of the 'gabbroic complex' and the amphibolite as metamorphosed 'mafic volcanic complex' (Coleman, 1977). The Redfern Complex is thinner than the classic ophiolite and Radloff concluded Chapter 2. Structure and Stratigraphy: A synthesis 44 that missing portions (most notably the sheeted dyke complex) have been tectonically removed during emplacement. It is also possible that it has been removed by erosion (chapter 1). On the basis of geochemistry, Montgomery (1978) correlated the Redfern Complex (his Black Riders Complex) with the CA to the north in Campbell's (1971) area and thereby equated it with the Antler Formation. Radloff (1989), on the basis of extensive whole-rock and mineral geochemistry, concluded that the Redfern Complex and the CA are very similar chemically and thereby correlated them but she did not correlate the CA with the Antler Formation. 2.5.2.2 Contacts The contact between the CA and the underlying unit 1 (Barkerville terrane) in the Quesnel Lake area is a thrust and was first described as such by Campbell (1971). Montgomery (1978) noted a fault at the base of the Redfern Complex (his Black Riders Complex) and considered it to be the fault upon which the ophiolite was obducted. In the Spanish Lake area, Rees (1981) identified a wide zone of mylonitization and interslicing of the CA and the underlying rocks. Rees and Ferri (1983) did a kinematic analysis of one outcrop of gneiss in unit 1 and concluded an upper plate (CA) to the east sense of motion. The criteria used by Rees and Ferri (1983) are based on the work of Berthe et al. (1979) and are subject to some doubt (Simpson and Schmid, 1983) but the determined sense of motion is reasonable (appendix A, item 2). Brown and Rees (1981) named the shear zone the Quesnel Lake Shear Zone (also used by Rees, 1987). Elsby (1985), Carye (1986), Fillipone (1985) and Bloodgood (1987a,b, 1988) all considered the boundary to be tectonic based on extensive mylonitization but none appear to have identified an individual fault surface. The contact between the CA and underlying unit 1 rocks was named the Eureka thrust by Struik (1985a) (see appendix Chapter 2. Structure and Stratigraphy: A synthesis 45 A, item 3) The upper contact of the CA is shown as a thrust in virtually all the published literature yet there is considerable evidence that this is primarily a sedimentary con-tact. The data detailed in chapter 1 indicate that the overlying unit 3 sedimentary rocks derived some of their material from the CA (and subjacent unit 1). In addition, Struik (1985c) noted a siltite horizon in unit 3 in apparent unconformable contact with the subjacent CA. Similarly, Radloff (1989) noted that sediments above the Redfern complex appear to be in stratigraphic contact. Bloodgood (1987a,b) and Fillipone and Ross (1990) also noted a stratigraphic contact in the Eureka Peak area, and nearby, both Carye (1986) and Elsby (1985) both noted sedimentary contacts, albeit variably tectonized ones. In the Spanish Lake area, Rees (1987) considered the upper boundary of the CA to be a sedimentary contact thus forcing his classification of the CA as part of the Quesnel terrane. In addition, Rees (1987, p. 73) noted thin 'lenses' of medium grained chloritic greenstone and bright green, fuchsite-bearing, micaceous foliae near the base of unit 3. These are interpreted to be similar to the fuchsitic conglomerate described in chapter 1 and indicate derivation of some of the material of unit 3 from underlying ultramafics in unit 2. In the same area Bloodgood (1988) showed the con-tact between units 2 and 3 to be at least partly tectonized. Struik (1988a) shows the contact as a stratigraphic contact and in the text (Struik, 1988a, p. 75 [^8) indicates the contact is stratigraphic. Therefore it is concluded that the upper contact of the CA is an unconformity along which there have been variable amounts of movement during later deformation. 2.5.3 Unit 3 — Quesnel sedimentary and volcanic sequence Unit 3 has been called the Quesnel terrane because it was considered to be in fault contact with the subjacent rocks. However, as shown in chapter 1, some beds in unit 3 Chapter 2. Structure and Stratigraphy: A synthesis 46 contain clasts of units 1 and 2. Also, the unit has been reported as being in stratigraphic contact with the subjacent unit 2, at least. There appears to be no recorded observation of a contact between unit 3 and unit 1 but the presence of clasts of unit 1 in unit 3 is considered to indicate a stratigraphic contact, probably a tectonized one. Because the lower contact of unit 3 is sedimentary, the unit cannot be termed a terrane. Therefore the unit is informally referred to here as the Quesnel sedimentary and volcanic sequence. 2.5.3.1 Lithology Rocks of unit 3 are divided into two lithostratigraphic packages; 1. a weakly to strongly metamorphosed, dominantly fine-grained sedimentary pack-age at the base, 2. and a weakly metamorphosed, dominantly volcanic package on top. To date, neither of these packages has been formally assigned to officially recognized formations or groups (North American Commission on Stratigraphic Nomenclature, 1983). Various authors have worked on the rocks in different areas and each has assigned his or her preferred informal nomenclature (table 2.2). 2.5.3.1.1 Package 1 — Triassic Black Phyllite (TBP) Unit 2, the Crooked Amphibolite, is everywhere overlain by a package of dominantly phyllitic rocks. The unit has never been formally named but it is most commonly referred to as the Triassic Black Phyllite (TBP). Campbell (1978) assigned it to Tipper's (1978) Quesnel River Group but this does not seem to have been formally accepted and is not in common use (table 2.2; appendix A, item 4). The unit is dated as Middle to Upper Triassic (Ladinian - Norian) from both macro-fossil evidence (Campbell & Tipper, 1971) and from conodont assemblages (Struik, Chapter 2. Structure and Stratigraphy: A synthesis 47 Author Package 1 Package 2 Campbell, K.V. 1971. Upper Triassic rock unit (graphitic pelite). Upper Triassic - Lower Jurassic rock unit (volcanic & sedimentary). Campbell, R.B. 1978. Upper Triassic Quesnel River Group. Triassic - Jurassic Quesnel River Group. Tipper, H.W. et al. 1979. Upper Triassic black phyllite. Triassic - Jurassic Takla Group. Rees, C.J. 1981. Upper Triassic 'Black Phyllite'. Triassic - Jurassic Takla Group. Rees, C.J. k Ferri, F. 1983. Middle - Upper Triassic 'Black Phyllite'. Triassic - Jurassic Takla Group. Struik, L.C. 1983a. Upper Triassic black shale, slate argillite etc. Elsby, D.C. 1985. Triassic graphitic phyllites & argillites Fillipone, J.A. 1985. Triassic Phyllite. — Carye, J.A. 1986. Upper Triassic - Lower Jurassic 'Eureka Quartzite & Crooked Lake Phyllite'. Jurassic Takla Group. Struik, L.C. 1986a. Middle - Upper Triassic black pelite. Middle Triassic - Lower Jurassic volcanics. Bloodgood, M.A. 1987a,b. Middle - Upper Triassic Quesnel River Group. Upper Triassic - Lower Jurassic Takla Group. Rees, C.J. 1987. Upper Triassic Black Phyllite. Takla Group. Bailey, D.G. 1988. Triassic - Jurassic Nicola Group. Triassic - Jurassic Nicola Group. Bloodgood, M.A. 1988. Triassic Black Phyllite (possibly Quesnel River or Nicola Group). Triassic - Jurassic volcanics (possibly Takla or Nicola Group). Struik, L.C. 1988a. Middle - Upper Triassic pelite (Slocan - King Salmon Assemblage). Middle Triassic - Lower Jurassic volcanics (Takla -Nicola Assemblage). Struik, L.C. 1988b. Middle - Upper? Triassic black pelite. Upper Triassic - Lower Jurassic volcanics of Quesnel River Group. Radloff, J.K. 1989. Triassic Black Phyllite — Fillipine, J.A. & Ross, J.V. 1990. Nicola Group — Triassic phyllite. Nicola Group — Jurassic volcanic rocks. Table 2.2: Nomenclature applied by previous workers in the Quesnel Lake area to rocks of unit 3, the Quesnel sedimentary and volcanic sequence. Chapter 2. Structure and Stratigraphy: A synthesis 48 1988a). The package has been studied by a number of authors in the Quesnel Lake area but most of them only dealt with it peripherally. It was most extensively studied by Bloodgood (1987a,b, 1988) who set up a detailed stratigraphic column. In the Crooked Lake - Eureka Peak area Bloodgood (1987a,b) indicated a total thickness varying from 2.5 to 4.0 km and she distinguished 7 lithologic units which are briefly described here (from bottom to top): 1. Pale buff or rusty-weathering quartzite (10 - 150 m). 2. Siliceous, graphitic, dark grey or black phyllite with very shiny phyllitic foliation. 3. Interbedded light and dark grey silty slates. Minor (1 - 3m) interbeds of dark grey siliceous limestone. 4. Well laminated grey phyllite. 5. Blue-black graphitic phyllite interbedded with dark grey siltstone and silty slate. Local green tuff lenses. Very thin, laminated, pale, quartz sandstone throughout. 6. Tripartite black - grey - black phyllites. Bedding defined by very thin (< 2cm) laminated siltstone. Some limestone near top. 7. Dark grey to black phyllite interbedded with green tuff near base. Tuffs increase upwards interbedded with grey - black phyllite, massive sandstone and minor limestone. As noted above the nature of the present contact with the underlying CA appears to be a variably tectonized depositional one. Additionally the TBP is highly tectonized internally and not all of the seven units of Bloodgood (1987a,b) are present throughout Chapter 2. Structure and Stratigraphy: A synthesis 49 the Quesnel Lake area. In the Spanish Lake area units 5 - 7 are present (Bloodgood, 1988). To the south, in the Mica Mountain area, a small infold of phyllite in Omineca Belt rocks was identified (McMullin and Greenwood, 1988a) which contains rocks ten-tatively identified as units 5 and 6. Radloff (1989) noted units 1, 5, and 6 overlying the Redfern Complex. The identification at these last two localities is hindered by the fact that the TBP is metamorphosed to kyanite-staurolite grade. The contact with the overlying package appears everywhere to be tectonic (Spanish Thrust — see below) (Bloodgood, 1987a,b, 1988; Rees, 1987; Struik, 1988a). 2.5.3.1.2 Package 2 — Triassic — Jurassic Volcanics (TJV) This package of rocks outcrops over a limited portion of the Quesnel Lake area. It was not examined in any detail for this study as the grade of metamorphism is very low (zeolite to greenschist facies; see Bloodgood, 1987a,b, 1988; Bailey, 1988; Rees, 1987). The unit has never been formally named in the Quesnel Lake area but various authors have either assigned informal names or correlated these rocks with formally described units outside the area — not always the same formal units. Informally it has been called the Triassic - Jurassic Volcanic unit (used here). It has also been called the Quesnel River Group (appendix A, item 4) by Campbell (1978) who included the underlying TBP. The package has also been called both Takla and Nicola Groups (table 2.2), sometimes including the subjacent phyllites in those correlations. The package contains Middle Triassic (Anisian) conodonts at its base (Struik, 1988a). Outside the Quesnel Lake area (to the west) macrofossils of upper Early Jurassic (Pliensbachian) age have been documented (Bailey, 1988). Within the Quesnel Lake area only Campbell (1971) and Bloodgood (1987a) have described the TJV in any detail. Both note that there is little lateral continuity to the Chapter 2. Structure and Stratigraphy: A synthesis 50 lithologic units probably representing facies variations between different volcanic cen-ters in an island arc. Campbell (1971) reported the basal volcanic unit to be pillowed, augite porphyry andesites (not present everywhere) overlain by brecciated "phyllitic siltstone and greenstone" with a total thickness of about 300 m. Above this he re-ported 1500 m of fairly uniform "pyroxene porphyry breccia, crystal tuff and flows" with "minor interbeds of phyllitic siltstone and breccia". Bloodgood (1987a) reported a considerably thinner sequence of (Takla Group) volcanics in the Eureka Peak area. She distinguished 4 units of variable thickness and extent: 1. A 0 - 50 m thick basal green (buff weathering) tuff — massive, homogeneous and locally banded. 2. Locally overlying the tuff is a 0 - 10 m basaltic pillow lava — green to blue-black on fresh surfaces, weathers pale grey or buff. 3. A thick (300 m) sequence of flow and flow breccias characterized by brecciated bases and more homogeneous, coarsely porphyritic upper portions. 4. Locally occuring altered (epidote) porphyritic flows, breccias and tuffs, charac-teristically rusty weathering and rotten. Outside the Quesnel Lake area (to the west) Bailey (1988) and Panteleyev (1987) documented an equally variable sequence of volcanic and minor associated sedimen-tary rocks. Bloodgood (1987a) and Panteleyev (1987) did extensive geochemistry and petrographic analysis and concluded that although remobilization of elements during alteration and metamorphism was extensive, the volcanics probably represent the ear-liest (alkalic to calc-alkalic) extrusions in a volcanic arc setting. Chapter 2. Structure and Stratigraphy: A synthesis 51 2.5.3.2 Contacts As described above the contact between the TBP of unit 3 and the subjacent rocks is a partly tectonized depositional one as indicated by observed stratigraphic (uncon-formable) contacts and the presence of material derived from units 1 and 2 within beds of unit 3 (chapter 1). The recognition of this contact as stratigraphic contradicts the definition of unit 3 as a terrane (Jones et ai, 1983). The contact between the TBP and TJV is considered to be a fault. In the Quesnel Lake area Bloodgood (1987a,b), Rees (1981, 1987), and Struik (1988b) are the only workers to have examined this in any detail. Bloodgood (1987a) considered the TBP and TJV to be time-equivalent units based on paleontological data of Tipper (1978) and Campbell (1978). She considered the TJV to be emplaced on top of the TBP along a shallow dipping thrust fault (Bloodgood, 1987a,b). Rees (1981) considered this contact to be tectonic in origin but later (Rees, 1987, p. 78, f 1) considered it to be stratigraphic. Struik (1988a) collected conodont assemblages from both packages (TBP and TJV) and concluded that they were time-equivalent and therefore in thrust contact (appendix A, items 5, 6). He named the contact the Spanish Thrust. 2.6 Structure Comparing and synthesizing all the structural analyses of previous workers in the area is difficult because structural description is an interpretive process and few authors see the same area in precisely the same way and there is little consistency in the use of terms. The criteria given in the introduction have been applied in an attempt to generate a single coherent structural picture. Different authors working in separate areas defined different numbers of phases of deformation. Where the resulting discussion becomes confusing as to which phase of deformation is being referred to the phase assigned here Chapter 2. Structure and Stratigraphy: A synthesis 52 is used as a reference point and is always written in square brackets (e.g., [Fi]). 2.6.1 Phases of Folding The rocks in the Quesnel Lake area are structurally complex. Previous workers in the area have identified several phases of deformation as shown by superimposed genera-tions of folds and associated foliations. The minimum number of phases recognized is 2, the maximum is 5. Table 2.3 shows the correlations made here between the vari-ous phases identified by previous workers. The correlations are made on the basis of fold (and foliation) orientation and the relationship to the metamorphism. Confusion and ambiguity arises in part because phases Fi through F 4 are approximately coaxial. Thus, only works which have examined small areas in detail are of use in distinguising the various phases. Although Struik (1983a, 1983b, 1985b, 1986a, 1986b, 1987, 1988a, 1988b) has done extensive work on the Quesnel Lake area, these works are not useful in defining phases of folding because they are at large scales and contain little detail. In addition, Struik (1983b, 1988a) avoided defining 'phases of folding' in the sense of generations of superimposed folds. Instead Struik (1988a) identified 3 categories of deformation, which were labeled from oldest to youngest as 'flow', 'ductile shortening', and 'brittle shortening and extension'. Struik (1988a, p. 37) stated that ". . . the strain environment is emphasized over the superposition of struc-tural elements. As a consequence little importance is attatched to desig-nating structures as Si or S2 etc.; though for comparison purposes this is done " This scheme is not useful for deducing a history of events. The terms 'flow', 'ductile', and 'brittle' are interpretive not descriptive terms and strictly speaking refer to pro-cesses. It is unclear how an observation could be assigned to one of these categories. Chapter 2. Structure and Stratigraphy: A synthesis 53 Author Unit Fx F 2 F 3 F 4 F 5 Fletcher (1972) 1 F0 Fx F 2 F 3 Montgomery (1978) 1 Fi F 2 F 3 2 Fi F 2 F 3 Pigage (1978) 1 Fo Fx F 2 F 3 F 4 Engi (1984) 1 Fi F 2 F 3 F 4 1 Fa F 2 F 3 F 4 F 5 Elsby (1985) 2 ? F 2 F 3 ?F4 F 5 3 F 2 F 3 ?F4 F 5 1 Fx F 2 F 3 F 4 F 5 Fillipone (1985) 2 ? Fx F 2 F 3 F 4 3 Fx F 2 F 3 F 4 Getsinger (1985) 1 Fla.b F 2 F 3 F 4 Montgomery (1985) 1 Fx F 2 F 3 F 4 F 5 1 Fx F 2 F 3 F 4 Carye (1986) 2 Fx F 2 F 3 F 4 3 F 2 F 3 F 4 Bloodgood (1987a) 2 Fx F 2 3 Fx F 2 Garwin (1987) 1 Fx F 2 F 3 F 4 Lewis (1987) 1 Fx F 2 F 3 F 4 1 Fx F 2 F 3 Rees (1987) 2 Fx F 2 F 3 3 ? F 2 F 3 1 Fx F 2 F 3 F 4 Radloff (1989) 2 Fo Fx F 2 ?F3 3 Fx F 2 ?F3 Table 2.3: Correlation of the phases of deformation applied by previous workers to rocks of the Quesnel Lake area. Italic F indicates a phase interpreted here for observations made by previous worker (see text). Question mark indicates uncertainty regarding presence of a phase or a tentative correlation because of insufficient or ambiguous data in the source. Chapter 2. Structure and Stratigraphy: A synthesis 54 Struik (1988a) refers to isoclinal folds in his 'flow' category and in later paragraphs refers to "cleavage, folds and faults" of the 'ductile shortening' category as features which "display characteristics of ductile flow." It is not clear how 'ductile flow' fea-tures differ from 'flow' features. These terms apply to processes and thus to conditions of deformation such as pressure, temperature, and strain rate. Thus each category may encompass several phases of superimposed folding. Alternatively, one phase of folding may have taken place under different conditions (ductile vs. brittle) in different local-ities. A brief description of the essential features of each of the fold phases shown in table 2.3 is given below followed by more detailed discussion of the correlations made to the phases defined by previous workers. The first phase of deformation [Fx] is seen only in units 1 and 2. It is seen only as rare, rootless, isoclinal folds which have an axial planar cleavage defined by the alignment of platy minerals [Si]. Si is parallel to lithological layering except in Fi fold hinges where mica flakes cut across folded layering. Fold axes trend roughly northwest but Si orientations are variable depending on the position with respect to later fold phases. There are no identified regional-scale Fi folds. Phase two folds are visible in all three units. However, in unit 3 it is the first phase of deformation visible. In units 1 and 2, phase two folds are tight to isoclinal in form and also have an axial planar cleavage [S2]. Fold axes trend northwest with variable orientations of S2 depending on the position on the limbs of later, particularly F 3 folds. In limb regions of F 2 folds S2 is parallel to layering and in some cases is visible as a crenulation cleavage. In hinge zones of F 2 folds S2 is clearly seen as a crenulation cleavage. In unit 3, F 2 folds are also tight to isoclinal, particularly near its base, but the foliation [S2] is never visible as a crenulation cleavage. In all three units, peak metamorphic mineral growth is synchronous with or slightly postdates the development of S2. This is indicated by crenulated inclusion trails in high grade Chapter 2. Structure and Stratigraphy: A synthesis 55 minerals in units 1 and 2 but not in unit 3. Because of the very thick sequence of rocks involved, metamorphism is not instantaneous nor strictly contemporaneous between units 1, 2, and 3. Metamorphic recrystallization in structually higher rocks (unit 3) may have started slightly after and finished slightly before those deeper in the crust (unit 1). Regional scale F 2 folds have been recognized for the area, particularly in unit 1, but they have been extensively refolded by phase 3 folds (Elsby, 1985; Fillipone, 1985). Most previous workers have given 'vergence' for the F 2 folds but there is often conflict between workers in adjacent areas. Because of the superimposed F 3 folding and because there is a marked absence of sedimentary 'way-up' criteria, the sense of 'vergence' indicated by each author must be treated with caution (appendix A, item 7)-Phase three folds are visible in all three units. These folds are responsible for the map scale features shown by many authors for the Quesnel Lake area. F 3 folds are open to tight and have an associated cleavage. Fold axes trend northwest and the axial plane dips steeply, usually to the northeast. At high structural levels S3 is a well developed crenulation or fracture cleavage. At deeper levels it is a crenulation cleavage that can be indistinguishable from S2. At higher structural levels, S3 (the second cleavage in unit 3) is seen to wrap around high grade metamorphic minerals. However, at lower levels, peak metamorphic recrystallization continued for slightly longer and thus clouds the distinction between S2 and S 3 . Phase four and phase five structures do not seem to be recorded simultaneously in the same outcrop and thus they may be conjugate systems of a single deformation phase. F 4 is represented mainly by open buckles and warps with upright axial planes which trend north or northwest and poorly developed axial planar fracture cleavage. F 5 is mainly represented by small buckles or crenulations with associated crenulation cleavage. It is distinctive because it is the only fold phase for which the trend of the Chapter 2. Structure and Stratigraphy: A synthesis 56 axial plane is northeastward. Both F 4 and F 5 are post peak metamorphism. Fletcher (1972) recognized only 3 phases of deformation the first of which was syn-chronous with metamorphism. However, examination of the thin section drawings (e.g., Fletcher, 1972, fig. 5, #'s 3, 4, 5, 6) and of the thin-sections indicates a premetamor-phic foliation, indicated as F0 on table 2.3. Fletcher's Fi [F2] is synmetamorphic and F 2 [F3] which are small scale folds with crenulation cleavage and upright axial planes, are syn- to postmetamorphic. Fletcher's F 3 folds are postmetamorphic, associated with retrogression(?), and have northwest trending axes and are therefore phase four [F4]. Montgomery (1978) defined only three phases of deformation. Secondly he did not identify any unit 3 lithologies in the area. However, later work (Radloff, 1989; McMullin and Greenwood, 1988a) has identified infolded unit 3 rocks in the area. From the description and knowledge of the area from these later works it seems likely that most of Montgomery's 'high grade Kaza' rocks are unit 3 phyllites. Montgomery's Fi features, which affected both units 1 and 2, seem to be a combination of both phases one and two described here. It is difficult to separate the features because of the misidentification of the unit 3 rocks. Montgomery's Fi folds are described as rootless isoclines but also as being synmetamorphic. Other large scale isoclinal folds were identified by Montgomery as F 2 which he stated were also synmetamorphic and were responsible for the map scale features [ F 3 ] . However, the map scale folds are known to have folded isograds (Campbell, K.V. , 1971: Campbell, R.B., 1978; Bloodgood, 1987a,b). Montgomery's (1978, p. 150-157) description of fabrics does not follow standard modern technique so it is difficult to determine which 'foliation' is associated by him with each phase of deformation. There is no access to Montgomery's samples but those of Radloff (1989) who examined the same area are discussed below. Chapter 2. Structure and Stratigraphy: A synthesis 57 Pigage (1978) defined 4 phases of deformation. However, his first phase foliation was described as synmetamorphic. Examination of Pigage's drawings and the thin-sections reveals that the peak metamorphic minerals commonly enclose a crenulation cleavage. Thus Pigage's Fi folds fold an earlier foliation [Fo]. Pigage's descriptions of his Fi to F 4 follow the descriptions of F 2 to F 5 defined here. Engi's (1984) phases of deformation follow those given in table 2.3 except that her F 4 is equivalent to F 5 defined here. Elsby (1985) defined five phases of folding plus a sixth phase of brittle faulting. Elsby's Fi to F 5 follow closely the definitions given here for Fi to F 5 . Elsby looked at rocks from units 1, 2, and 3. However, he stated that his Fi is restricted to unit 1. He considered that units 2 and 3 both acted together (as the Intermontane Belt) and thus the absence of Fi structures from unit 3 is used as evidence as absence from unit 2. Secondly, unit 2 is thin in Elsby's area and where exposed is homogeneous, thus rootless isoclines are not present. As these isoclines, with the associated cleavage cutting layering in hinges, are the only diagnostic feature of F i , their absence cannot be used as evidence for the absence of Fx. Fillipone (1985), like Elsby (1985), assumed that units 2 and 3 acted as a single tectonic unit (Fillipone's 'cover sequence'). Fillipone's phases Fi through F 5 for his 'basement' (unit 1) match the definitions of Fi to F 5 above. Fillipone (1985) defined the phases of deformation in his 'cover' separately. That is, his Fi for his 'cover sequence' was equivalent to F2 for his 'basement', hence the apparent mismatch on table 2.3. Like Elsby (1985), Fillipone did not identify any Fx structures in unit 2 rocks, probably because the unit is very poorly exposed in Fillipone's area. Getsinger (1985), who examined only rocks of unit 1, identified five phases of defor-mation but labeled the first two as F J a and Fib. F i a and Fib [Fi] were both described by Getsinger as isoclinal folds which were indistinguishable from one another on the basis Chapter 2. Structure and Stratigraphy: A synthesis 58 of fold orientation and geometry except in some cases. At these localities Getsinger noted isoclinal folds which were refolded isoclinally or isoclinal folds cut by intruded gneiss which is in turn isoclinally folded. These Fib folds do not appear to have a sep-arate foliation associated with them. Getsinger reported that F 2 was synmetamorphic and examination of her drawings and thin-sections indicates that metamorphism was synchronous with the second foliation [S2]. Montgomery (1985), who examined unit 1 rocks only, defined five phases of defor-mation which match the definitions given above. Carye (1986) examined all three units in the Quesnel Lake area. Carye's area contains significant outcrop of unit 2 and thus he is the only worker to have examined it in detail. Carye (1986) differed from other workers in describing the same number of phases of deformation in unit 2 as in unit 1. Thus, Carye (1986) proposed that unit 2 was emplaced on unit 1 before deposition of unit 3. This is in agreement with the interpretations made in chapter 1 on independent criteria. Carye (1986) identified four phases which match the definitions of Fi to F 4 above. Bloodgood (1987a) examined only units 2 and 3. She did not identify any phase one [Fi] structures in unit 2. She defined only two phases of deformation which match phases two and three [F2, F 3 ] above. Garwin (1987) and Lewis (1987) mapped adjacent areas containing only unit 1 0 rocks. They described only four phases of deformation. These are correlated as shown on table 2.3. The work of Rees (1987) is particularly difficult of correlate with other works done in the Quesnel Lake area. Like previous workers, Rees assumed that units 2 and 3 were part of a single tectonic unit, the Intermontane Belt. However, because of significantly better outcrop of unit 2 when compared to other workers (except Carye, 1986) he noted, as Carye did, that unit 2 rocks contain the same number of phases of deformation as Chapter 2. Structure and Stratigraphy: A synthesis 59 unit 1. He therefore proposed that all three units underwent the same number of phases of deformation. He identified three phases of deformation and determined that peak metamorphism was associated with the second of these. The correlation by Rees of the first phase of folding in unit 3 with the first phase in units 1 and 2 is probably mistaken. Thus the complex structural history and complicated fold style and vergence patterns interpreted by Rees from the outcrop data are probably in error. The complicated pattern of folding of Rees (1987) resulted in some peculiar anomalies at least one of which Rees was unable to explain. Rees (1987, p. 153 and maps) shows a '?klippe' of Triassic Black Phyllite (unit 3) with a few outcrops of Crooked Amphibolite (unit 2) surrounded by unit 1 rocks at the southern end of Cariboo Lake. This is to the northeast of the supposed folded contact between units 1 and 2 (Rees 1987, section C - C ' - C " reproduced in figure 2.4). The proposed westerly vergence for the major folds shown put the '?klippe' in the hinge zone of a syncline with unit 1 and 2 rocks above it! Rees (1987, p. 154) stated that ". . . there doesn't seem to be any practical way of infolding the Antler For-mation (unit 2) and Black Phyllite (unit 3) to the extent needed to ac-count for their postulated existence on south Cariboo Lake. It remains an enigma." Re-examination of the outcrop data in conjunction with the understanding that units 1 and 2 may have been deformed prior to deposition of unit 3 may help explain this enigma. This re-examination was not possible on the basis of the discussions in Rees (1987) which are extremely interpretive in tone. Radloff (1989), examined virtually the same area seen by Montgomery (1978) and recognised rocks of all three units. Radloff identified four phases of folding in units 1 and 2 and three in unit 3. However, the first deformation in unit 2 was only identified Chapter 2. Structure and Stratigraphy: A synthesis 60 P P A Figure 2.4: Section C-C' -C" with labelling reproduced from Rees (1987). Rocks of HPS and DMWQg belong to unit 1 defined here. Rocks of P P A are part of unit 2 defined here and rocks of uTrBP, Tr J T , and ImJ are part of unit 3 defined here. Note the ?klippe (Rees, 1987) of upper Triassic Black Phyllite (uTrBP) in the hinge of a proposed westerly verging synform below rocks of units 1, 2, and 3. Chapter 2. Structure and Stratigraphy: A synthesis 61 in the basal (peridotitic) rocks and Radloff ascribed this to mantle deformation prior to emplacement. Thus she ascribed an extra phase of deformation of unit 1 only. The apparent lack of Fi structures in rocks of unit 2 may be due to the very thick section and their mechanically homogeneous nature. However, Radloff noted that strongly foliated gneiss intruded the contact between units 1 and 2 with a chilled margin. Radloff assigned the foliation within the gneiss to her F 2 . This seems to be completely analogous to the quartz-diorite gneiss seen by Getsinger (1985) to intrude between development of F i a [Fi] and Fib [F2] isoclinal folds. Thus, the evidence indicates that unit 2 was emplaced on unit one prior to F 2 , presumably synchronous with Fi (indicated as Fo on table 2.3). Note that Radloff (1989) defined separate deformation schemes for unit 1 and units 2 and 3. There is no orientation information given by Radloff for her F 3 in units 2 and 3. It is correlated with F 5 defined above on the basis that it is the same as Radloff's F 4 in unit 1. 2.7 S u m m a r y The careful re-examination of the literature on the geology of the Quesnal Lake area leads to the following conclusions. 1. The rocks are divisible into three tectonic units: (a) Unit 1, the continental margin prism. (b) Unit 2, the Crooked Amphibolite, remnant of deformed oceanic crust, prob-ably the correlative of the late Paleozoic (Pennsylvanian-Permian) Antler Formation of the Slide Mountain terrane. (c) Unit 3, the Quesnel sedimentary and volcanic rocks. Chapter 2. Structure and Stratigraphy: A synthesis 62 2. Unit 1 is not divided into the previously defined Barkerville and Cariboo terranes because they do not satisfy the criteria for the definition of terranes. (a) The two 'terranes' contain rocks which are facies variants. (b) The fault which separates the two terranes (the Pleasant Valley Thrust) as it has been defined is an ambiguous and cryptic feature. Even if present in the Quesnel Lake area it is an extremely early feature and thus does not separate bodies of rock with markedly different structural histories. 3. Unit 2, although poorly exposed, appears to have been emplaced on unit 1 before deposition of unit 3. (a) Unit 3 rocks contain clasts of deformed lithologies from units 1 and 2 (chapter !)• (b) Sparse evidence indicates that units 1 and 2 show the same number of phases of deformation, one more than unit 3. 4. Unit 3 cannot be considered a separate terrane because it has a depositions!, albeit tectonized, contact with subjacent unit 2 (and unit 1?). 5. Five phases of deformation are identified for the Quesnel Lake area. Phases one through four were coaxial with northwest trending axes and variably dipping axial planes. Phase five structures have axes trending northeast with vertical axial planes. (a) Fi is represented by rootless isoclinal folds at outcrop scale only. The ax-ial planar cleavage is never visible as a crenulation cleavage and is seen to crosscut layering in the hinges of the rootless isoclines. Chapter 2. Structure and Stratigraphy: A synthesis 63 (b) F 2 folds are also isoclinal but the associated axial planar cleavage is com-monly a crenulation cleavage, particularly in the hinges of F 2 folds. Map scale F 2 folds have been identified and have been refolded by F 3 . F 2 is synmetamorphic. (c) F 3 folds are generally upright and are the folds responsible for the major map-scale features. F 3 cleavage is always a crenulation cleavage. Some metamorphic mineral growth is associated with F 3 but in general F 3 folds are postmetamorphic and fold mineral isograds. (d) F 4 and F 5 are never viewed in the same outcrop and may be conjugate fold systems. Both are represented by open small-scale buckles and crenulations and are post the peak of metamorphism. Retrogression has been documented with F 4 . 6. The preferred sequence of tectonic events is as follows. (a) Deposition of unit 1 from Proterozoic to late Paleozoic (or early Mesozoic) time. (b) Intrusion of Devonian-Mississippian Quesnel Lake gneiss. (c) Deposition of Late Paleozoic (Pennsylvanian-Permian) Crooked Amphibo-lite (unit 2), assuming it is correlative with the Antler Formation of the Slide Mountain terrane. (d) Emplacement of unit 2 on unit 1. Contact between units 1 and 2 is Eureka thrust. First phase of deformation recognized in units 1 and 2 [Fi]. (e) Uplift and erosion of combined units 1 and 2. (f) Deposition of unit 3 on eroded units 1 and 2, including fragments of units 1 and 2 in some layers of unit 3. Chapter 2. Structure and Stratigraphy: A synthesis 64 (g) Deformation of all three units producing F 2 structures and peak of meta-morphism. Possible renewed movement on Eureka thrust and tectonization of lower contact of unit 3. (h) Continued deformation of all 3 units. F 3 folds of major contact between units 1, 2, and 3. (i) F 4 and F 5 folding and retrogression. Chapter 3 A n algorithm and program for calculating ideal activity of components of minerals showing coupled substitution, with particular reference to micas 3.1 Introduction The calculation of pressure and temperature of equilibration of metamorphic assem-blages is an important task of the petrologist. This is usually done by calculating the P and T of intersection in P-T space of two or more displaced equilibria. In order to apply thermodynamic parameters for any particular end-member to natural minerals it is necessary to calculate the activity of the end-member in solid solution. Activ-ity models are divided into two parts, the 'ideal' and 'non-ideal' activity. The 'ideal' activity, which is related to the configurational entropy, is the probability of a partic-ular configuration of ions within the structure of a mineral. The 'non-ideal' activity describes departures from whichever 'ideal' model is chosen as reference. Margules parameters, which describe non-ideality, may be determined from experiments or from natural assemblages (chapter 4). The method of calculating ideal activity proposed here utilizes two ideas: • Ideal activity is a probability; therefore the sum of the ideal activity of all possible ionic configurations within a mineral must be unity. • Matrices can be used to store, keep track of, and manipulate values of ideal activ-ity. Such a matrix has one dimension for each site which undergoes substitution. Each dimension in the matrix has the same number of entries (rows, or columns 65 Chapter 3. Thermodynamic activity of micas 66 etc.) as there are substituting entities in that site. Values within such a matrix can be manipulated so that: — The activity of each non-permitted ionic configuration is zero. — The sum of all activities (all entries in the matrix) is unity. — The sum of the activities of all species containing some specific element in a particular site equals the site occupancy of the element for that site. A matrix is initially filled by the activity values determined using an ideal site-mixing model (Greenwood, 1977). These values are modified to produce values which satisfy the mass-balance and mass-action equations that can be written using a distribution of species approach in an ideal complex solution (Brown, 1977). 3.2 Ideal activity as probability Standard texts on thermodynamics (Wood and Fraser, 1977; Nordstrom and Munoz, 1986) show that ideal activity is best considered as a probability. In any mineral the ideal activity of an end-member is the probability that the appropriate ionic configura-tion exists within the crystal structure. In a Utopian world we could send some demon into a crystal with instructions to count the number of times a particular configuration of ions occurs within the crystal. This number, normalized by the total number of ob-servations made, is the activity of that configuration of ions (i.e., that species). In the real world, we analyse the crystal (by microprobe perhaps) and calculate a structural formula. From that formula we calculate an ideal activity usually making the assump-tion that the probability of the ionic configuration of interest is the continued product of the probabilities of all the individual ions within individual sites. For example, the Chapter 3. Thermodynamic activity of micas 67 activity of almandine (Fe 3Al 2Si 30i2) in aluminous garnet ((Mg,Fe,Ca,Mn)3Al2Si30i2) is where P£ is the probability of finding element X in site A. The assumption is made that the probability of finding a particular ion in a particular site is the same as the mole fraction of that ion in that site. Because the probability of finding Al in the octahedral sites and Si in the tetrahedral sites is unity, equation 3.1 becomes This is the ideal site-mixing model and typifies the case where there is no coupling between subtitutions on different sites. It also applies to minerals which show 'unique' coupled substitution. That is, minerals for which one substitution is accompanied by a unique co-substitution. For example, the substitution of Na for Ca in plagio-clase ((Na,Ca)(Al,Si)AlSi20g) is always accompanied by a substitution of Si for Al. Therefore, calculation of the activities of anorthite or albite requires either Na-Ca site occupancy or the tetrahedral site occupancy. However, minerals which show 'non-unique' coupled substitution pose a problem. 'Non-unique' coupled substitution occurs when substitution on one site must be ac-companied by a co-substitutions, but there is no unique way of co-substitution. For ex-ample, in biotite ((K,Na,Ba,Ca)1(Fe,Mg,Mn,Ti,Al)3Al(Si,Al)3O10(OH,F,Cl)2) the sub-stitution of Al for Fe in an octahedral site may be accompanied by replacement of K by a vacancy in the 'A' site, or by the replacement of Si by Al in a tetrahedral site, or by the replacement of Fe in another octahedral site by a vacancy and the replace-ment of Al by Si in a tetrahedral site. In the case of simple substitution or unique coupled substitution the probabilities are either completely independent or completely coupled. That is, the value of one probability is independent of the value of another or G r t r > C u b r j C u b - p C u b - p O c t - p O c t r>Tet -pTet T j T e t a A l m — r F e " r F e ' r F e ' r A l ' r A l ' r S i " r S i ' r S i (3-1) a A l m — A F e ' A F e ' A F e • G r t _ Y C u * > v C u b v C u b (3.2) Chapter 3. Thermodynamic activity of micas 68 it is unity. In the case of non-unique coupled substitution the values of the individual site probabilities are neither completely independent nor completely coupled. That is, many of the end-members that one might wish to define for an ideal simple model are not permissible because of the requirement for local charge balance. For example, al-though Ti substitutes for Fe in the octahedral sites in biotite there is no Ti end-member with the same structural formula as annite (i.e., KTi3AlSi30io(OH)2 is not possible). Neither is there a unique substitution that accompanies Ti substitution, such as the SiAl_i substitution that accompanies NaCa_i in plagioclase. Because of this, activity of micas should not be represented by ideal simple site-mixing models, such as is com-monly done. For micas the number of species is considerably larger than the number of components and activity should be calculated using an ideal complex model. Ideal complex models have rarely been used in the literature for solid solutions (Engi, 1983 is an exception), though papers dealing with activity of fluid species routinely use such models. Ideal simple site-mixing models usually suffice. The formulation of complex solution models is recommended by Brown and Greenwood (in prep.) for all solutions. Ideal complex models have the drawback that they can require considerable com-puting facilities. It is necessary to determine the number of species involved and then to write an equal number of mass-balance and mass-action equations by considering the distribution of species. This system of n equations in n unknowns must then be solved simultaneously. These equations are non-linear and must be solved numerically. In the micas it is possible to list 510 charge-balanced species, requiring the solution of 510 non-linear equations in 510 unknowns. Unlike systems of linear equations, systems of non-linear equations usually have several roots, each of which is a valid mathematical solution to the problem. Numerical methods require an initial estimate of the values of the root and this must be as close to the 'real' solution as possible. If this guess is too far away from the 'real' set of values the numerical method may iterate to a root that Chapter 3. Thermodynamic activity of micas 69 is mathematically valid but which is not the 'real' solution to the physical problem. Some (but not all) of these roots may be discarded easily. For example, a root may have negative values (negative activities) which are impossible in the real world. It is possible to combine site-mixing models with the distribution of species model and thereby to reduce the number of unknowns and equations. This makes the problem somewhat more tractable but the problem of an accurate initial guess and the possibility of multiple roots still remain. The distribution of species model for micas with 75 species is shown in the following section. The distribution of species approach is followed by a matrix manipulation algorithm devised from an intuitive approach to activity. This matrix method is computationally simple, is relatively fast even on a personal computer, and generates a set of values for the activity for all the charge-balanced species which satisfy all the linear and non-linear equations formulated from the distribution of species approach. 3.3 Charge balanced species in micas and the distribution of species equa-tions Micas (both di- and trioctahedral) have the general formula (K, Na, Ca, Ba, • ^(Fe, Mg, Mn, Al, Ti , • )3(A1, Si) 2Si 2Oi 0(OH, F, Cl) 2. Using a site-mixing model there would be 5 x 63 x 22 x 32 possible ionic permutations (i.e., 38,880 species considering KFeMgFeAlSi30io(OH)2 to be a different species from KMgFeFeAlSi30io(OH)2, etc.). However, because the concern is primarily with charge balanced species one can write the formula for micas as (A, B, • )x (P, Q, R, • )3(A1, Si) 2Si 2O 1 0(M) 2 Chapter 3. Thermodynamic activity of micas 70 where A and B are all univalent and divalent cations in the A site respectively, P are tetravalent, Q are trivalent and R are divalent cations in the octahedral site, and M are the univalent anions in the XH-fold site. The symbol, • , represents a vacancy. There are now only 3 x 43 x 22 (768) permutations. However, if we consider that Al and Si substitution in the two tetrahedral sites is only in response to charge balance constraints these can be ignored. Therefore we need only consider the A site and three octahedral sites (i.e., 3 x 43 = 192 permutations). Of these 192 species only 75 are charge balanced. These 75 species are given in appendix B. Having determined the composition of these species it is necessary to determine the equilibria between them. This was done by using the methods outlined by Greenwood (1967, 1968). • The composition of all the species are written in matrix form, where the matrix has 75 rows, for 75 species, and n columns, for n components. The components can be oxides or elements. • The rank of the matrix (the minimum set of components required to write all compositions) and set of independent rows are determined. In this case the rank was determined to be 6, and species 1, 2, 6, 17, 20, and 45 as given in appendix B are taken as the basis. • The sub-matrix containing only the independent rows (species 1, 2, 6, 17, 20, and 45) written in the minimum 6 components is inverted. • A basis change is performed on the full matrix of 75 species in 6 components by multiplying it by the inverted matrix. The resulting matrix contains the compositions of all species written in terms of the independent species. The entries within the matrix are the coefficients of the reactions that link each species with the independent species. For example the 19th row in the new matrix is: Chapter 3. Thermodynamic activity of micas 71 Si S2 S6 Sl7 S20 S45 Sl9 -0.75 0.5 1.0 0.5 0.25 -0.5 thus the reaction concerning S19 is: S19 + 0.75Si + O.5S45 = 0.5S2 + S6 + 0.5SIT + 0.25S20. The 69 mass action equations given in appendix B were determined in this way. Note that initially all 69 species were written in terms of the 6 independent species, but the system can be simplified by noting that many of the species have the same compositions but with different ionic configurations. In addition to the 69 mass action equations an extra 6 mass balance equations are required so as to have 75 equations for 75 unknowns. These equations state that the total amount of a component is the sum of the activities of the species which contain that component times the number of sites in that species containing the component. For example, the mole fraction of R (divalent ions in the octahedral site) in the mica is 75 i=l where v\ is the number of sites containing R in species i and X; is the mole-fraction (activity) of species i. The six mass balance equations include the five linking the total abundance of A, B, P, Q, and R to mole fractions of species. The sixth mass balance equation is, the sum of the mole fractions (activities) of all species equals unity. All 75 equations for micas are given in appendix B. Available algorithms for solving systems of non-linear equations either handle fewer than 20 equations or require forming a matrix of derivatives of these equations. If Chapter 3. Thermodynamic activity of micas 72 this matrix is singular the method fails, though this does not indicate that the set of equations is singular. This is the case for the equations given in appendix B. Conse-quently an alternative approach was developed which results in values that satisfy all 75 equations, although these equations were not used in the solution process. 3.4 A n algorithm for calculating activity in species with coupled substitu-tion The standard site-mixing model for activities as described above assumes • that site occupancies are probabilities and • that the activity, also a probability, is the coproduct of the individual site occu-pancies. For any site in a mineral, the site occupancy is a list of numbers, or a one-dimensional array. Activity is the product of site occupancies and thus the activities of all species can be displayed in an array with as many dimensions as there are sites in the mineral. Let us take for example a mineral with two sites, site A and site B, which undergo substitution. A real mineral of this form could be the pyroxenes ((Mg, Fe, Ca, Na, Li)a(Mg, Fe, Al, Ti) 1Si 20 6). Let us say that site A contains a number, m, of different elements A;, site B contains n elements Bj. In the pyroxene example Ai is Mg, A 2 is Fe, etc., Bi is Mg, and B 2 is Fe, etc. The total amount of element Ai in site A is ai, the amount of A 2 is a 2 and the amount of A; is a,. Likewise the amount of element Bj is /3j. Using an ideal site-mixing model for activity, the activity of the species containing Aj in the A site and Bj in the B site is (a;j) which is calculated as a\ x /3j. In the pyroxene example above, the activity Chapter 3. Thermodynamic activity of micas 73 of diopside (CaMgSi^Oe) is the site occupancy of Ca in the M l site multiplied by the site occupancy of Mg in the M2 site. The values of all the activities for a mineral with sites A and B can be displayed as a two dimensional array as follows: Site B Pi 02 0n ai,i ai,2 ai,n Site A a 2 a2,i a2,2 a 2,n ; ; ; am,2 ^m,n The rows are related to the values for the A site, the columns to the values for the B site. That is, row 1 contains the activities of species containing element Ai in the A site, row 2 the activities of species with element A 2 in the A site and so on. Likewise column 1 contains the activities of species with element Bi in the B site and so on. The values of activity stored in this fashion satisfy three rules. • Rule 1: The sum of all activities is unity. That is m n i = i j = i • Rule 2: The sum of all activities in any row or column is a known site occupancy. That is, n ^2 a ; j = U i = a ; (for any i) j=i m X a i j = v i = & (for any J)-i = l • Rule 3: The activities of all the non-permitted species (i.e., species that are not charge balanced) must be zero. Chapter 3. Thermodynamic activity of micas 74 Recognizing these three rules, an algorithm has been formulated which produces a ma-trix of activities with these properties. To aid in illustration in the following discussion, two dimensional matrices are used as examples but the properties and procedures apply equally to matrices with greater numbers of dimensions. The algorithm has three steps. The first step is to fill a matrix with the values of activity that would be produced using an ideal site-mixing model. Using the mineral above, which has sites A and B (such as the pyroxene example), let us say that the A site has four elements with site occupancy of 0.2, 0.2, 0.1, 0.5, and the B site has four elements with a site occupancy of 0.1, 0.4, 0.2, and 0.3. If the mineral was pyroxene the structural formula might be (Aio. 2 A 2 0 2 A 3 0 . i A 4 0 S )(Bi 0 1 B 2 o 4 B 3 o 2 B 4 o 3 )Si 20 6 The matrix of ideal site-mixing activities looks like this: Site B 0.1 0.4 0.2 0.3 0.2 0.02000 0.08000 0.04000 0.06000 Site A 0.2 0.02000 0.08000 0.04000 0.06000 0.1 0.01000 0.04000 0.02000 0.03000 0.5 0.05000 0.20000 0.10000 0.15000 However, let us say that some of the species are not permitted for reasons of charge balance (e.g.,, say, species at positions 1,4; 2,1; and 4,3). Therefore, although the above matrix satisfies rules 1 and 2, it does not satisfy rule 3. The second step is therefore to make the entries for non-permitted species zero: Chapter 3. Thermodynamic activity of micas 75 Site B 0.1 0.4 0.2 0.3 0.2 0.02000 0.08000 0.04000 — Site A 0.2 — 0.08000 0.04000 0.06000 0.1 0.01000 0.04000 0.02000 0.03000 0.5 0.05000 0.20000 — 0.15000 Now the matrix satisfies rule 3 but not rules 1 and 2. The sum of all entries is less than unity by the sum of the deleted values (rule 1 violated) and only column 2 and row three satisfy rule 2. The third step is to modify the matrix so as to increase all the non-zero values so as to satisfy rules 1 and 2. This is done iteratively in a series of repeated row-wise and column-wise normalizations. After zeroing 'non-permitted' activities the row sums and column sums are either less than or equal to the correct value. That is, in general n a 'J = u ' 7^  a i ( f° r a n y i) j=i and m Zl aiJ = VJ ^ A (for any j). i=l In the first step values are normalized row-wise. For each row the values of activity are changed by multiplying each entry in that row by the theoretical row sum divided by the actual row sum (i.e., by ai/u;). In the above matrix the a's are 0.2, 0.2, 0.1, 0.5 and the u's are 0.14, 0.18, 0.1, 0.4. Therefore all values in row 1 are multiplied by Chapter 3. Thermodynamic activity of micas 76 0.2/0.14, values in row 2 by 0.2/0.18, values in row 3 remain the same, values in row 4 are multiplied by 0.5/0.4. The resulting matrix is: Site B 0.1 0.4 0.2 0.3 0.2 0.02857 0.11429 0.05714 — Site A 0.2 — 0.08889 0.04444 0.06667 0.1 0.01000 0.04000 0.02000 0.03000 0.5 0.06250 0.25000 — 0.18750 Row sums are now correct, but column sums are not. The same normalization process is done column-wise, i.e., each ajj is multiplied by fij/vy The resulting matrix is: 0.1 Sit< 0.4 3 B 0.2 0.3 0.2 0.02827 0.09269 0.09399 — Site A 0.2 — 0.07210 0.07311 0.07038 0.1 0.00989 0.03244 0.03290 0.03167 0.5 0.06184 0.20277 — 0.19795 This pair of row-wise/column-wise normalizations constitutes a single iteration. For a two-dimensional matrix, an initial activity value a; j becomes ( A , V i a i - i ) After any iteration, the new column sums are correct (see above matrix) but the row sums are not. Repetition of row-wise/column-wise normalization eventually leads to Chapter 3. Thermodynamic activity of micas 77 convergence, i.e., both row and column sums are correct, and therefore repeated 'nor-malzation' cannot further change the values within the matrix. The above matrix converges to Site B 0.1 0.4 0.2 0.3 0.2 0.02459 0.08238 0.09303 — Site A 0.2 — 0.06506 0.07348 0.06145 0.1 0.00885 0.02965 0.03349 0.02801 0.5 0.06655 0.22291 — 0.21054 (values are rounded in final decimal place). Final values do not depend on the order in which normalization is done (i.e., row-wise/column-wise or vice versa). The above matrix satisfies all three rules stated and constitutes a matrix of activities for a mineral with coupled substitution. An analytical proof of convergence has not been found but in the several thousand examples done to date convergence has occurred. The algorithm is easily expanded to a larger number of dimensions to calculate activities for real minerals such as micas. After convergence the values of activity are tested against the appropriate mass-action and mass-balance equations and in all cases these equations have been satisfied. 3.5 Calculat ing activity in micas using the new algorithm The computer program MICAC (for MICa Activity Calculator) given in appendix B utilizes this algorithm. The algorithm is easily applied to natural minerals. All that is required is a matrix with as many dimensions as there are sites undergoing coupled substitution. In the case of micas there are six sites undergoing substitution Chapter 3. Thermodynamic activity of micas 78 but a four dimensional matrix is sufficient. The substitutions in the two tetrahedral sites can be accounted for by noting that there are only three possible configurations of the tetrahedral sites and anions. These are • • • Ai2Si20io(M)2, • • • AlSi30io(M)2, and • • • Si40io(M)2, which have charges of -8, -7, and -6 respectively. It is sufficient therefore to note that, for charge balance, the A site and three octahedral sites must have 6, 7, or 8 positive charges. The rules outlined above for a 2-dimensional case can be restated for the 4-dimensional case as follows. Rule 1: The sum of all activities is unity; m n o p i= l j= lk= l1=1 Rule 2: The sum of all activities of species containing a specific ion in a specific site is the site occupancy of that ion in that site; n o p miY,aio.kj = u i = ( f o r a n y i ) j = l k=l1=1 m o p U S Yl aio.k,i = v i = Pi ( f o r any j) i= i k=i1=1 m n p XmiZ ^ J W = w k = *k (for any k) i = i j = i 1=1 m n o X) 2 ^jw = x i = e i ( f o r a n y !)• i= l j = l k= l Rule 3: The activities of non-permitted (not charge balanced) species are zero. Note that in the case of the site occupancy sums (Rule 2) summing is done over 3 dimensions, or one less than the total number of dimensions. These 3-dimensional subunits are referred to as hyperplanes in the following discussion and in the comments for MICAC. The sums are referred to as hyperplane (or plane) sums. The vectors of site occupancies, which in the 2-dimensional case above were written along the margins of the matrix, are referred to as marginal vectors. There are three values in the A site Chapter 3. Thermodynamic activity of micas 79 marginal vector. These are the site occupancies for divalent ions, univalent ions, and vacancies. There are four values in the octahedral site marginal vectors. These are site occupancies for tetravalent, trivalent, and divalent ions, and vacancies. The three marginal vectors for the three octahedral sites are identical. Thus the matrix contains 3 x 43 entries. MICAC performs 9 steps which include the same three basic steps of the algorithm described above. 1. Input. The program asks a number of questions with regard to input format and the amount of data to be processed. An additional question is asked regarding possible run-time problems (see below). 2. From the input mica analysis the program calculates a structural formula. Steps 2 to 9 are repeated for each analysis in the input file. 3. Using the calculated structural formula, the program prepares the marginal vec-tors. 4. A matrix is then filled with values for activity using the ideal site-mixing ap-proach, i.e., aij,k,i = o=i • /3j • <$k • This is step 1 in the algorithm as described above. 5. The program calculates the total charge associated with each position within the matrix. If this lies outside the permitted range (6 to 8) the value is made zero (step 2 of the algorithm above). 6. The matrix is normalized in each dimension successively. As stated above, the order in which this is done does not affect the final values determined. The values Chapter 3. Thermodynamic activity of micas 80 in each hyperplane are summed and compared with the theoretical value (one of the values in the marginal vector). The values are then modified by the ratio of marginal vector to the hyperplane sum. For the first dimension, in general n o p aio-k.i = u i 7^ «i (for any i) j=l k=l 1=1 Any value ajj^i is multiplied by a;/u;. This is repeated for each dimension. After each set of 4 normalizations the ratios of the hyperplane sums to marginal vectors are calculated. If these are within a specified limit (at present 100 ± 0.5%) the program states that convergence has occurred and the values within the matrix now satisfy the three rules stated above. The convergence criteria can be made smaller (e.g., 100 ± 0.1%) but this increases the length of time necessary for calculation and the change to individual values of activity within the matrix is too small to warrant such a stringent criterion. 7. The program calculates the activities of real species from the activities of the charge-balanced forms. Nine such species are considered at present. These are, muscovite = KAl 2 nAlSi 3 O 1 0 (OH) 2 paragonite = NaAl2DAlSi3Oio(OH)2 celadonite = KAlMgDSi4Oio(OH)2 annite = KFe3AlSi3Oio(OH)2 phlogopite = KMg3AlSi3Oio(OH)2 margarite = CaAl 2nAl 2Si 2Oio(OH) 2 oellacherite = BaAl 2aAl 2Si 2Oio(OH) 2 pyrophyllite = • Al 2aSi 4Oio(OH) 2 and talc = • Mg 3Si 4Oi 0(OH) 2. Chapter 3. Thermodynamic activity of micas 81 However, the activity of any charge balanced species can be calculated. The program will also present (in one form of the output) values of activity that are produced using various 'old' formulations of activity, all of which are modified ideal site-mixing models. 8. Program prints out various types of output (see appendix B). 9. Finally, the values determined using this method are substituted into the equa-tions formulated using the distribution of species approach. The status of this check is printed to the screen and an output file. If the values satisfy the distribu-tion of species equations a short message to that effect is printed. If not, the list of equations and the evaluated RHS and LHS for the equations are printed. Be-cause the matrix normalization approach stops when the hyperplane sums come within a specified range of the marginal vectors, the individual values within the matrix will not satisfy the distribution of species equations exactly and a similar criterion of closeness must be used (see appendix B). All of the possible run-time errors that commonly occur due to problems with the input are simple ones, with one exception. Some muscovite analyses, though they conform to the structural formula, do not contain sufficent Al to do the normalization phase. The program tries to normalize but cannot bring the hyperplane sums close enough to the values in the marginal vectors. This is probably due to error in the input analysis, particularly in the amount of Al. When the program encounters a data point where repeated normalization will not cause convergence (150 iterations) it can artificially augment the amount of Al in the analysis by 0.5 wt% and redo the procedure. The user has the choice of having the program do this automatically or of interactively stating a preference. This procedure does not seem to make a significant difference to the final calculated value for the activities of muscovite and paragonite, Chapter 3. Thermodynamic activity of micas 82 which axe primarily a function of the K + and Na + contents respectively. The celadonite and pyrophyllite activities are probably more affected. In any event, data points for which this procedure has been followed are flagged with a symbol. Other run-time problems axe more easily understood. Data points where the ana-lytical total is less than 93.5 and more than 98.5 weight percent (excluding the 0H~, F~ and C l - content) are rejected. The program also requires that the analysis fit the usual structural formula. That is, the A site cannot have more than one ion per unit formula, and the octahedral sites can contain no more than three and no less than two ions per unit formula. In some muscovite analyses, an excess in the A site occupancy is linked to a deficiency in the octahedral site occupancy probably due to analytical error and a low value for Al. The program can artificially augment (just once) the Al content by 0.5% in order to increase the octahedral site occupancy and decrease the A site occupancy. The user has the choice at the start of the program of having this done automatically or on an interactive, point-by-point basis. Finally, the program performs a check on the amount of tetrahedral Si required to 'make' the total activity of all species in the matrix. The total mole fraction of species within the matrix with a charge of 6+ requires four tetrahedral Si ions, the fraction with 7+ requires three tetrahedral Si, and the fraction with 8-f requires 2 tetrahedral Si. The total Si required is checked against the actual amount given in the initial analysis. In the case of a discrepancy the output is flagged with the symbol '*'. This never seems to occur. Chapter 3. Thermodynamic activity of micas 83 3.6 Values of activity for common micas using the new technique The method described above produces values for activities of all the charge balanced species in micas which are consistent with the distribution of species approach of an ideal complex model. In the case of the 2-A1 dioctahedral micas (muscovite, paragonite, margarite, oellacherite) the new method does not produce values which are significantly different from those given using the most commonly accepted site-mixing model, a M s = X j f • (X™8)2 • ( X ^ ) 2 where v M s _ K A K — K + Na + Ca + Ba + • ' X M s = — and' A 1 Al + Fe + Mg + Mn + Ti OH v M s O H ~ OH + F + CI' However, for the trioctahedral micas, the values produced using MICAC are signif-icantly different from those using the simple site-mixing model. This is to be expected, as different site-mixing models produce significantly different values. Generally the val-ues for the activity of annite (and phlogopite) are intermediate between those produced using a A n n = X g M X ^ M X o n ) 2 where v B t _ K  A K — K + Na + Ca + Ba + • ' X B t = Fe d F e Al + Fe -f Mg + Mn + Ti a n  Y B t OH O H OH + F + Cl and those using the same expression except with v B t _ Fe X F e - y Chapter 3. Thermodynamic activity of micas 84 The values produced using the former are commonly 50% larger than those produced using the latter. The values produced using MICAC are generally closer to the smaller value. See the example given in appendix B. It should be noted that the ratio of the activities of annite and phlogopite produced with MICAC is exactly the same as the ratio of values derived using any site-mixing model. This is because annite and phlogopite have the same charge balanced form in MICAC and their individual activites are produced from a single value within the matrix manipulated by MICAC. Thermometers (exchange reactions) involving biotite use only the ratio of activities of annite and phlogopite. Therefore the values of activity produced using MICAC do not yield different temperatures (all other parameters in the thermometric calculation being equal). However, the position in P-T space of any net transfer reaction involving either one of the biotite end-members alone or both end-members in different proportions is sensitive to the absolute values of the activities of the end-members. Therefore net transfer equilibria will plot in a different positions in P-T space when different activity models are used. In the case of the other end-member activities, only celadonite is used on a regular basis. There is no universally accepted approximation for the activity of celadonite though _ vMs vMs yMs aCel — A K • JV M G • A A, is most commonly used (the various mole fractions can be defined in a variety of ways). The value for the activity of celadonite in muscovite determined using MICAC is usually about twice the value produced using the above expression. The activities of pyrophyllite and talc in mica are rarely, if ever, calculated. The values calculated using MICAC are valid ideal activities but in the absence of non-ideal solution parameters for these end-members in mica it is not possible to use these activities to find the positions Chapter 3. Thermodynamic activity of micas 85 of equilibria involving these phases. 3.7 Conclusions, recommendations and further work The program MICAC, based on the algorithm described, gives values of activity for all charge balanced species in the micas (both di and tri-octahedral) which are deemed more valid than standard ideal simple (site-mixing) models because they satisfy the three rules given above. These values of activity are roots to the set of distribution of species equations formulated using a complex ideal model for activity. The new algorithm does not require the sophisticated computational facilities necessary for the solution of the large set of distribution of species equations. In the case of the di-octahedral (2-A1) micas, the values produced by MICAC are not significantly different from those produced using standard site-mixing models. However, the values for tri-octahedral micas and for celadonite, pyrophyllite and talc are different from those produced using standard methods. The calculation of the positions of all equilibria involving muscovite, paragonite, margarite, and oellacherite, as well as exchange equilibria involving biotite, can be done using site-mixing models for activity. However, the accurate placement of net transfer equilibria involving the biotite end-members and celadonite require the accurate values of activity produced using MICAC. In addition, any attempt to formulate non-ideal solution parameters for any of these end-members using natural phases require that the ideal portion of the activity terms be accurate and the values produced using MICAC are recommended. Further programs involving larger numbers of of sites are under consideration, though initial work indicates that the very large matrices and deeply nested 'do loops' result in unreasonable execution times. Chapter 4 Cal ibrat ion of the S G A M thermobarometer for pelitic rocks using data f rom phase equil ibrium experiments and natural assemblages 1 4.1 Abstract The assemblage Silica - Garnet - Aluminosilicate - Mica (SGAM) is common in am-phibolite grade metapelitic rocks, yet the equilibrium Aim + Ms = Ann + 2 Al 2 Si0 5 + Qtz is rarely used as a barometer because of poorly constrained standard state properties for annite and solution parameters for biotite. These parameters have been deter-mined using mathematical programming analysis on sets of phase equilibrium exper-imental data and natural assemblage data. Pressure and temperature of the natural assemblages were independently constrained by aluminosilicate polymorph occurrences, grossular - aluminosilicate - silica - plagioclase (GASP) barometry and garnet-biotite thermometry. Recent garnet solution models indicate only small deviations from ideality on the almandine-pyrope binary. Using Ferry and Spear's (1978) garnet-biotite experiments, Fe-Mg mixing in biotite appears to be ideal, permitting calibration of the standard state properties of annite. A significant problem in applying these data to natural assemblages is the unknown magnitude of non-ideal mixing between Mg/Fe and Ti /Al . Submitted to Canadian Mineralogist in September 1990 under this title co-authored by D.W.A. McMullin, R.G. Berman, and H.J. Greenwood. 86 Chapter 4. Calibration of SGAM thermobarometer 87 The differences, W B 3 l - W B t (MgTi - FeTi) and W B t - W B t (MgAl - FeAl) are about one-third the magnitude of those found by Indares and Martignole (1985) but the treatment of the data suggests moderately large individual values for the Margules parameters (up to 75 kJ/mol). Application to several sets of published analyses show that this calibration offers distinct improvements over previous calibrations. Pressures determined using the new calibration are consistent with GASP barometry and the aluminosilicate polymorph present. The effect of Ti and Al in biotite is satisfactorily accounted for and variations in these amounts do not produce systematic variation in the calculated results. In addi-tion, several data sets show field gradients, particularly in P, not previously recognized and which agree with field observations. 4.2 Introduct ion There are two major barometers for pelitic rocks, GASP (Garnet - Aluminosilicate - Silica - Plagioclase) and GRAIL (Garnet - Rutile - Aluminosilicate - Ilmenite -Quartz) initially calibrated by Ghent (1976) and Bohlen et al. (1983) respectively. However, both have disadvantages. Metamorphic plagioclase is usually heterogeneous, leading to uncertainty over which composition to take as representative of equilibrium. Thus, GASP pressures are usually not very sensitive. GRAIL is not commonly used for two reasons. Firstly, metamorphic assemblages with both ilmenite and rutile are not very common. Secondly, small grain size and small modal amounts of the Ti-bearing phases make potential GRAIL rocks difficult to recognise in the field. A third barom-eter, using the assemblage Garnet - Biotite - Plagioclase - Muscovite was calibrated by Ghent and Stout (1981), and is less commonly used. It has the same disadvantage Chapter 4. Calibration of SGAM thermobarometer 88 as the GASP barometer because of the use of plagioclase. In addition, all four miner-als used are in solid solution, and uncertainties in the solution models propagate into greater uncertainty in the resultant pressures. For these reasons the system silica - garnet - aluminosilicate - mica (hereafter re-ferred to as SGAM) has been investigated as a possible thermobarometer. I used the program PTAX (part of GeO-Calc software of Berman et al., 1987) and the thermo-dynamic database of Berman (1988, 1990) to calculate pressure and temperature. A significant problem in calculating the position of SGAM equilibria is that non-ideal mixing effects in biotite are largely unknown. Indares and Martignole (1985) estimated biotite solution parameters using the Fe-Mg exchange equilibrium between garnet and biotite and composition data from natural assemblages. However, because it is an exchange reaction, they could only determine the differences between the Fe and Mg Margules parameters (i.e., MgTi — FeTi and MgAl — FeAl). They assumed that MgFe mixing is ideal (MgFe parameter zero). This study determines a set of individual Mar-gules parameters for biotite using the technique of Mathematical Programming (MAP) to analyse well-constrained natural assemblages in conjunction with the existing exper-imental data. 4.3 A p p r o a c h e s t o c a l c u l a t i n g P a n d T Two approaches can be taken in calculating temperatures and pressures of equilibration of mineral assemblages. These approaches are based on the same fundamental ther-modynamic principles but differ in assumptions made in their use. The most common way to calculate P or T has been to calibrate the chosen equilibrium, from phase equi-librium experiments or thermodynamic data, and to formulate some expression in T, P and the composition of the participating phases. From thermodynamic first principles Chapter 4. Calibration of SGAM thermobarometer 89 P, Pr Pressure (bars, kilobars), Reference pressure (1 bar). T , T r Temperature (K, °C) Reference temperature (298.15 K). A G f ' T Gibbs free energy of formation of phase i from the elements at P and T . A r G p , T Gibbs free energy of reaction at P and T . A H P , T Apparent enthalpy of formation of phase i from the elements at P and T (kJ/mol). AfH° Enthalpy of formation of phase t from the elements (kJ/mol). A r H p ' T Enthalpy of reaction at P and T . gP,T Entropy of phase i at P and T (J/mol/K). Third law entropy of phase i (J/mol/K). A r S p - T Entropy of reaction at P and T . Molar volume of i (J/bar). Cp, Heat capacity of phase i. A r V p - T Volume of reaction at P and T. W H l 2 , W S l 2 , W V l 2 Margule's parameters for excess enthalpy (kJ/mol), entropy (J/mol/K) and volume (J/bar). ai d Ideal activity of component i. H Activity coefficient of component i. a; Activity of phase i = ajd • 7;. K Equilibrium constant. KD Distribution coefficient. R Gas constant. Table 4.1: Glossary of notation and symbols used, a barometer is usually derived as P = a + b - T + c - R - T - l n K which can be rewritten to give the usual form of a thermometer, rp _ x + y - P z - R - l n K (4.1) (4.2) Chapter 4. Calibration of SGAM thermobarometer 90 where a, b, c, x, y, and z are constants derived by calibration (see table 4.1 for other notation). For example, Ferry and Spear's (1978) calibration of the equilibrium 2 Phi + Aim = Ann + Prp (4.3) to give their garnet-biotite geothermometer 2089 + 0.0095P 0.7820 +In K { ' ' where T is in degrees K, P is in bars, and _ (Mg/Fe)Grt (Mg/Fe)Bt ' See table 4.2 for mineral abbreviations and formulae. For convenience this approach to thermobarometry is referred to as the single equation approach. A second approach is to calculate the position in P-T space where the Gibbs free energy equals zero for all equilibria implied by a given assemblage using a computer program (e.g., PTAX from GeO-Calc, Berman et al., 1987) and a database of thermo-dynamic properties and solution parameters (e.g., Berman, 1988, 1990). This second approach can be used to formulate the equations used in the single equation approach, such as 4.4, but not vice versa. This way of doing thermobarometry is referred to as the multi-equilibrium approach. The single equation approach has the advantage of being computationally simple to use, but several problems arise. Firstly, thermometers (and barometers) are usu-ally calibrated by experiments using pure phases, or pure phases diluted by a single solution phase, whereas the natural minerals are complex solid solutions. Therefore, solution models are required to calculate end-member activities in the analyzed miner-als. For many thermometers and barometers there is a number of different calibrations 2 For convenience, all equilibria are written with the high T assemblage on the right hand side. Chapter 4. Calibration of SGAM thermobarometer 91 Almandine Aim Fe 3 Al 2 Si30 1 2 Andalusite And Al 2 Si0 5 Annite Ann KFe 3AlSi 30 1 0(OH) 2 Biotite Bt (K,Na,Ba)1(Fe,Mg,Mn,Ti,Al)3Al(Si,Al)301o(OH,F,Cl)2 Enstatite En MgSi0 3 Fayalite Fa Fe 2Si0 4 Ferrosilite Fs FeSi03 Forsterite Fo Mg 2Si0 4 Grossular Grs Ca 3 Al 2 Si 3 0i 2 Garnet Grt (Mg,Fe,Ca,Mn) 3Al 2Si 30 1 2 Kyanite Ky Al 2 Si0 5 Muscovite Ms KAl 2 AlSi 3 O 1 0 (OH) 2 Phlogopite Phi KM g 3 AlSi 3 O 1 0 (OH) 2 Pyrope Prp Mg 3 Al 2 Si 3 0i 2 Quartz Qtz Si0 2 Sillimanite Sil Al 2 Si0 5 Table 4.2: Mineral names, formulae and abbreviations used in text. Abbreviations are those suggested by Kretz (1983). Chapter 4. Calibration of SGAM thermobarometer 92 using different 'preferred' solution models. For example, Newton and Hazelton, (1981), Hodges and Spear (1982), Ganguly and Saxena (1984, 1985), Indares and Martignole (1985), and Hoinkes (1986) have all proposed modified versions of Ferry and Spear's (1978) thermometer (4.4). As each of these corrections gives a different T for any given set of garnet and biotite compositions the problem arises, which is the 'correct' one to use? It is not unusual to find publications where the author(s) has tabulated T calcu-lated using several of these corrections and then decided arbitrarily (more or less) that one correction yields the 'best' temperatures. This problem of 'correction' becomes worse when the thermometer or barometer was calibrated from natural assemblages. Usually simple solution models are invoked on the understanding that the calibrated thermometer (barometer) be used only on minerals of 'similar' composition to those used in the calibration. It is often difficult to specify what constitutes minerals of 'similar' composition. A second and related problem is that individually calibrated equilibria commonly invoke different solution models either implicitly or explicitly for the same phase and thus they are inconsistent with one another. Thirdly, calibrations which utilize the same solution models but which determine different values for the a, b, c, or x, y, z terms in equations 4.1 and 4.2 imply different standard state properties for the minerals involved. Any P - T point defined by two inconsistent thermometers and barometers is in error. These problems can be avoided using the multi-equilibrium approach and a pro-gram such as P T A X (Berman et al., 1987) and an internally consistent database of standard state properties and solution parameters. The accuracy of any defined P - T point still depends on the accuracy of the database and of the mineral analyses but erro-neous results caused by inconsistencies among underlying data and solution models are eh'minated. Furthermore, the multi-equilibrium calculation approach can help identify Figure 4.1: A: Cluster of equilibrium curves defined by equilibrium assemblage (assum-ing thermodynamic data are correct). B: Curves defined by disequilibrium assemblage. Chapter 4. Calibration of SGAM thermobarometer 94 samples which display disequilibrium as well as identifying minerals whose standard state or solution properties need refinement. For example, in figure 4.1A the calculated equilibria for a sample form an obvious cluster at a single P-T suggesting equilibration at that P and T. The excellent internal agreement strongly implies equilibrium among all the minerals as well as the accuracy of the underlying thermodynamic data. In figure 4.IB the equilibria do not cluster suggesting that the reactions did not 'freeze' at the same P-T conditions. In figure 4.2 all the equilibria involving the phase i (dashed curves) are displaced away from the P-T point defined by the other phases, indicating that either phase i is not in equilibrium with the remaining phases or that the stan-dard state or solution properties of i are in error. If this second hypothesis is likely, this sample may be used to constrain further the properties of the offending phase. In this respect, multi-equilibrium calculations are the inverse of MAP, in that, the former process utilizes compositions and thermodynamic data to determine pressure and temperature whereas MAP utilizes pressure, temperature, and composition data to determine thermodynamic data. Thus the techniques can be combined iteratively to identify suspect data and to improve poorly known thermodynamic data. The fol-lowing analysis uses these techniques to calculate a consistent set of biotite solution properties and calibrate the SGAM barometer. 4.4 S G A M — previous calibrations SGAM, which includes the minerals quartz, garnet, aluminosilicate, muscovite, and biotite, shows promise as a thermobarometer. The three equilibria that can be written for the SGAM assemblage are the garnet-biotite exchange reaction (equilibrium 4.3) and the Fe and Mg end-member net transfer reactions Aim + Ms = Ann + 2 A l 2 S i 0 5 + Qtz (4.5) Chapter 4. Calibration of SGAM thermobarometer 95 Figure 4.2: Equilibrium curves calculated from a database containing poor data for one phase. Curves involving phase i (dashed) displaced from P-T intersection (A) defined by equilibria involving remaining minerals of the assemblege. Chapter 4. Calibration of SGAM thermobarometer 96 and Phi + 2 A l 2 S i 0 5 + Qtz = Prp + Ms. (4.6) Only two of these are linearly independent and the position of the invariant point defined (Fig. 4.3) is dependant on mineral compositions. S G A M , though a widespread assemblage, has not been calibrated experimentally. Despite this, reaction 4.5 has been examined and used in thermobarometric studies in recent years. A . B . Thompson (1976a) deduced the relative stabilities of various assem-blages but did not suggest a quantitative barometer. Tracy et al. (1976) and Tracy (1978) used equilibrium 4.5 qualitatively to make rough estimates of pressure (e.g., <7 kb). Fletcher and Greenwood (1977) calculated the position for equilibrium 4.5 us-ing a multi-equilibrium calculation approach and a thermodynamic database from the published literature. Although the pressures determined by Fletcher and Greenwood using S G A M were consistent with those determined using other barometers, Fletcher and Greenwood were reluctant to place weight on the results because the thermody-namic data came from a variety of different sources. "Quantitative" barometers based on equilibrium 4.5 were derived by Robinson (1983), Hodges and Crowley (1985), and Holdaway et al. (1988). A l l followed the 'single equation' approach and all derived their equations from natural data. Robinson's (1983) calibration appears to be in er-ror (reported in an abstract). The calibration utilizes the Fe content (Xp e) of garnet and biotite (not the F e / M g ratio) and the K and A l contents of coexisting biotite and muscovite are accounted for by a constant. A l l pressures utilizing this barometer fall approximately along a fine in P - T space, and at temperatures below about 475°C the barometer yields negative pressures. The calibrations of Hodges and Crowley (1985) and Holdaway et al. (1988) are compared later with the new calibration. Chapter 4. Calibration of SGAM thermobarometer 97 Figure 4.3: Equilibria possible between the phases Almandine, Annite, Muscovite, Phlogopite, Pyrope, Quartz, Sillimanite. Intersection defines P and T of equilibration of assemblage. Chapter 4. Calibration of SGAM thermobarometer 98 4.5 Properties and parameters to be determined The positions of equilibria in P-T space, such as those in figure 4.7, are calculated by evaluating the equation A r G p , T = phases o = £ « AfH? -T-Sf [ +V°(P - Pr) + - V°)dP J + R - T - l n K (4.7) where Vi is the reaction coefficient for phase i. Evaluation of expression 4.7 requires values for enthalpies of formation, third law entropies, heat capacity terms, and expan-sivity and compressibility terms for all phases. If the phases involved are not stoichio-metric end-members one requires, in addition, activity models in order to evaluate the equilibrium constant, K. For equilibrium 4.5 K = a A l m ' a M s (4.8) where a A n n a A l m a M s = a id Ann T A r — a A I m ' 7 A l m — a M s ' 7Ms assuming that the activities of SiG"2 in quartz and A^SiOs in aluininosilicate are unity. The ideal (configurational) terms for both micas are evaluated using the algorithm and program described in chapter 3. For almandine a A i m = ( X F e r t ) 3 - T h e activity coeffi-cients are described by Margules parameters using the activity formulation proposed by Berman and Brown (1984). Activity models for garnet and muscovite were taken from Berman (1990) and Chatterjee and Froese (1975) respectively. Chapter 4. Calibration of SGAM thermobarometer 99 In the evaluation of non-ideality in biotite there are several site substitutions to be considered. K, Na, Ca, and Ba occur on the XH-fold or 'A' site; Fe, Mg, Ti, Al, and Mn occur on the octahedral sites; and OH, F, and Cl on the 'OH' site. In biotite, the 'A' site is predominantly filled by K (>0.9 ions per unit formula). Although muscovite shows considerable non-ideality in Na-K mixing on the 'A' site, in biotite it is considered to be negligible as it rarely contains more than 0.05 Na ions per unit formula (Guidotti, 1984). In metamorphic biotite there is some substitution of F and Cl for OH but except in F-rich parageneses, F and Cl usually constitute less than 0.10 ions per unit formula (2 sites). Therefore non-ideal mixing on this site has been ignored. The major substitution in biotite occurs in the octahedral sites. Mn is usually present in negligible amounts (< 0.01 atoms per 3 sites (Guidotti, 1984)). Thus the important interactions are Mg-Fe, Mg-Ti, Mg-Al, Fe-Ti, Fe-Al, and Ti-Al . As a first approximation only symmetric interactions on these binaries are considered as the data do not warrant a more complicated model. Evaluation of 7Ann requires the Margules parameters W B t (MgFe), W B t (MgTi), W B t (MgAl), WgJ (FeTi), and W B t (FeAl). In this study, the four parameters Wj^, Wfg, WB 4', and WB 4' have been determined in order to calculate the positions of the Mg- and Fe-SGAM equilibria accurately. Because the data are not sufficiently constraining to define the Ti-Al interaction parameter it is assumed to be zero. The Fe-Mg binary is assumed to be ideal (i.e., W B 2 = 0) on the basis of the observation that there is no excess volume of mixing between annite and phlogopite (Hewitt and Wones, 1984). In fitting for the parameters indicated above I used the method of mathematical programming (MAP), which is an extension of the linear programming method first introduced to the petrologic literature by Greenwood (1967, 1968) and expanded on by Gordon (1973, 1977). Linear and non-linear programming (collectively MAP) have been used extensively since then by a number of workers (e.g., Day and Halbach, Chapter 4. Calibration of SGAM thermobarometer 100 1979; Day and Kumin, 1980; Halbach and Chatterjee, 1982; Berman and Brown, 1984; Berman et al., 1986; and Berman, 1988, 1990). The techniques, their advantages and disadvantages are discussed extensively by the authors above, particularly Berman et al. (1986). 4.6 Data Sources for M A P Two kinds of data are considered here. Firstly, there are P - T - X data from phase equi-librium experiments and from natural assemblages. These data are used as constraints on the parameters being fit for during MAP analysis. The sources for these data and the equilibria used are given in table 4.3. Secondly, there are thermodynamic data including standard state properties, heat capacities, expansivity and compressibility values and solution parameters. These values are used as bounds during MAP analy-sis. Enthalpies of formation determined in this study and the enthalpies, entropies, and molar volumes for the phases involved in SGAM are given in table 4.4. The heat ca-pacity terms and thermal expansivity and compressibility terms for all phases in MAP analysis were taken from Berman (1988, 1990). Solution parameters for muscovite are from Chatterjee and Froese (1975), those for garnet from Berman (1990), and both sets are given, along with the solution parameters determined during this study, in table 4.5. S° of annite was calculated as 421.01 J/mol/K using the exchange reaction Phi + 3 FeO = Ann + 3 MgO, (4.9) entropies of phlogopite and periclase from Berman (1990) and (1988) respectively, and the entropy of FeO from Helgeson et al. (1978). The value for the electronic config-uration and crystal field stabilization contribution (Helgeson et al., 1978, p. 51) was Chapter 4. Calibration of SGAM thermobarometer 101 Equilibrium Ref.* Parameters constrained.§ Uncertainties for MAP. Phi + Aim = 1 A f H A n n 6°C, 0.01 X A n n (see text) Ann + Prp 2t wB3 l - w2B3s wf4l - 1.0 kbar, 30°C (see text) 3t 11 1.0 kbar, 30°C (see text) 4 t 11 (see text) 5 t 11 1.0 kbar, 50°C (see text) 6 t 11 0.1 kbar, 30°C (see text) Aim + Ms = 2t wB t W B t W B t v v 1 3 ' v v 1 4 ' v v 2 3 > 0.5 kbar, 30°C (see text) Ann + 2 Sil + Qtz W B t v v 2 4 3t 11 0.5 kbar, 30°C (see text) Aim + Ms — 2t ii 0.5 kbar, 30°C (see text) Ann + 2 Ky + Qtz Aim + Ms = 4t ii 0.5 kbar, 30°C (see text) Ann + 2 And + Qtz * Refs: (1) Ferry & Spear (1978), (2) Pigage (1982), (3) Holdaway et al. (1988), (4) Ferry (1980), (5) Indares & Martignole (1985), (6) LeBreton & Thompson (1988). § Component order for Grt: 1 = Ca, 2 = Mg, 3 = Fe, 4 = Mn. Bt: 1 = Mg, 2 = Fe, 3 = Ti, 4 = Al. Margules notation of Berman & Brown (1984). f Data from natural assemblages. % Experiments with natural minerals. Table 4.3: Phase equilibrium data (experiments and natural assemblages) used to derive enthalpy of annite and biotite solution properties. Chapter 4. Calibration of SGAM thermobarometer 102 AH f°(kJ/mol) S°(J/mol/K) V° (J/bar) Aim -5267.216* 340.007* 11.511 And -2589.972 91.434 5.147 Ann -5142.000* 421.010f 15.483§ Ky -2594.220 82.430 4.412 Ms -5976.740 293.157 14.087 Phi -6210.391$ 334.346* 14.977* Prp -6286.548 266.359 11.316 aQtz -910.700 41.460 2.269 Sil -2586.091 95.930 4.983 * Value fit by MAP in this study. f Calculated from oxide data (see text). § From Hewitt and Wones (1984). * From Berman (1990). All other values from Berman (1988). Table 4.4: Enthalpies, entropies and volumes. Chapter 4. Calibration of SGAM thermobarometer 103 Biotite:* ij W H i j (kJ/mol) W S i j (J/mol/K) W V i j (J/bar) 13 58.865 14 75.000 23 30.921 24 63.721 Component order: 1 = Mg, 2 = Fe, 3 = Ti, 4 = Al. Garnet :f ijk W H i j k (kJ/mol) W S i j k (J/mol/K) W V i j k (J/bar) 112 21.560 18.79 0.100 122 69.200 18.79 0.100 113 20.320 5.08 0.170 133 2.620 5.08 0.090 223 0.230 0.00 0.010 233 3.720 0.00 0.060 123 58.825 23.87 0.265 124 45.424 18.79 0.100 134 11.470 5.08 0.130 234 1.975 0.00 0.035 Component order: 1 = Ca, 2 = Mg, 3 = Fe, 4 = Mn. Muscovite: § ijk W H i j k (kJ/mol) W S i j k (J/mol/K) W V i j k (J/bar) 112 12.230 -0.7104 0.6653 122 19.456 -1.6543 -0.4561 Component order: 1 = K, 2 = Na. * Values fit during this study. f Solution model of Berman (1990). § Solution model of Chatterjee and Froese (1975). Table 4.5: Solution (Margules) parameters. Component order indicated for each min-eral follows the notation of Berman and Brown (1984). Chapter 4. Calibration of SGAM thermobarometer 104 determined as 4.7 J/cation by averaging the values for the two exchange equilibria (entropies of Fo, Fa, En, and Fs from Berman, 1988). The molar volume of annite was set at 15.483 J/bar (from Hewitt and Wones, 1984). 4.7 Phase equilibrium data Phase equilibrium data on 4 equilibria from 6 sources were used to determine the enthalpy of annite, and the solution properties of biotite (table 4.3). One study of the garnet-biotite equilibrium (4.3), notable by its absence from table 4.3, is that of Perchuck and Lavrent'eva (1983). This data set was considered in the early phase of this study because Perchuk and Lavrent'eva used natural garnet and biotite compositions. However, there are several problems with using these data. Firstly, I was unable to determine what the starting compositions were in their experiments. The tabulated compositions of starting minerals for their experiments are given as Mg numbers only and do not match the few quoted mineral analyses. Secondly, the run products were extremely inhomogeneous (particularly garnet) and one cannot tell which compositions were in equilibrium. Thirdly, the primary interest in their data was due to their use of Ti and Al rich biotite, yet the run product compositions are listed only as Mg number. The data therefore cannot be used as constraints on Ti and Al solution parameters. 4.7.1 Treatment of experimental data For reversed experimental data the P and T quoted for any experiment is known to lie on one side or the other of the equilibrium involved. Knowing this, it is possible to use in Fo + 2 FeO = Fa + 2 MgO (4.10) En + FeO = Fs + MgO (4.11) Chapter 4. Calibration of SGAM thermobarometer 105 Temperature Figure 4.4: Adjusted experimental brackets for MAP analysis of data for a reaction with positive slope and phase i as reactant. Chapter 4. Calibration of SGAM thermobarometer 106 the MAP analysis values of P and T for an experiment that have been adjusted to take account of the experimental uncertainties. The uncertainties for MAP analysis, listed in column 4 of table 4.3, were utilized as follows. The nominal P, T, and composition for the experiment were adjusted by the uncertainties given in table 4.3 so as to move the P - T - X constraint away from the equilibrium position. For example, for an equilibrium with positive slope and phase i as a reactant, the nominal pressure for a 'reactants stable' experiment was increased, the nominal temperature was decreased, and X, was decreased (if i is a product then X, was increased). For a 'product stable' experiment, P is decreased, T increased and X, increased (see fig. 4.4). In all cases these adjustments render the constraint provided by the experiment less stringent. The temperature listed for the data of Ferry and Spear (1978) are those quoted by Ferry and Spear. Two of Ferry and Spear's biotite compositions were adjusted to in-clude 0.06 mole fraction Al. This is the amount proposed by Ferry and Spear to account for slight excess volume in the synthetic pure annite used as starting material for these two experiments. LeBreton and Thompson's (1988) experiments were melting experi-ments on assemblages of natural minerals, not phase equilibrium reversals. Therefore, these experiments were treated as equilibrated natural assemblages (see below). 4.7.2 Treatment of natural data Data from equilibrated natural assemblages were treated differently. The first step was to estimate a pressure and temperature of equilibration using garnet-biotite thermom-etry and either GASP barometry or an aluminosilicate boundary (for rocks containing 2 aluminosilicate polymorphs). This estimate is based on standard state properties and garnet solution parameters from Berman (1988, 1990). Provisional biotite solution properties were assigned from the parameters of Indares and Martignole (1975) and attributed entirely to the Mg interactions (see Berman, 1990). I used the plagioclase Chapter 4. Calibration of SGAM thermobarometer 107 P - T where r e a c t a n t s a s s u m e d stable ^ C D t_ 3 OT V) CD 'nominal ' P - T for natural a s s e m b l a g e from G A S P / G r t - B t or A S K b o u n d a r y / G r t - B t P - T where p roduc ts a s s u m e d s tab le B Temperature Figure 4.5: A: P-T 'brackets' used for natural assemblages where nominal P-T de-fined by GASP/Grt-Bt (one ASK polymorph), or where nominal P-T defined by ASK boundary (two or more Al 2 Si0 5 polymorphs) and Grt-Bt thermometer. B: 'Brack-ets' defined for assemblages with one ASK polymorph (e.g., Sil) but which give a GASP/Grt-Bt P -T close to an Al 2 Si0 5 polymorphic transformation (e.g., Sil/Ky). For example, an assemblage containing Sil with 'nominal' P-T close to the Ky/Sil boundary, T for the low T - high P 'bracket' was determined as nominal T - A T and pressure was determined as the P of the Ky/Sil boundary at this T. Chapter 4. Calibration of SGAM thermobarometer 108 solution model of Fuhrman and Lindsley (1988) for the activity of anorthite in the GASP equilibrium. Embedded in this first estimate is a provisional assumption re-garding Mg-interactions in biotite and reliance on 2 independent equilibria. The next step in the process is to generate a 'bracket' on the equilibrium P - T - X conditions ap-propriate to the level of confidence in the estimates of P, T, and analytical error. This results in a 'bracket' that is identical in algebraic structure to the experimental bracket (fig. 4.5A) and allows experimental and natural data to be treated simultaneously. Conservative estimates of the uncertainties were taken so as to prevent estimated natu-ral P-T coordinates from exerting excessive control over the results. Thus once a P and T of equilibration had been determined it was assumed that the reactants and products were stable at P-T points an arbitrary AP and A T removed from the preliminary P-T estimates (fig. 4.5A). For samples containing one ASK polymorph in which the initial P-T estimate from garnet-biotite and GASP lay close to one of the ASK boundaries, that boundary was taken as one limit of the adjustment offset, while the other limit was set by a judgement of the uncertainty in the P-T estimate (fig. 4.5B). As with experimental data, the AP and A T offsets were applied so as to make the data point less constraining on the final parameters than the unadjusted nominal P and T. The MAP uncertainties for natural assemblages given in table 4.3 are the values used in the final fits. Larger uncertainties were used in the initial stages of fitting. Some of the data sets were gradually tightened as described below. The data of LeBreton and Thompson (1988) are from melting experiments on natural minerals, not reversals. Thus, the data were treated in the same way as the natural assemblages, in that each experiment provided two brackets. However, the P uncertainty used was much smaller than that used for natural assemblages (100 bars). LeBreton and Thompson's run products were somewhat inhomogeneous probably due to the large grain size garnets used (LeBreton and Thompson, 1988, p. 229). However, there is no estimate of the Chapter 4. Calibration of SGAM thermobarometer 109 amount of inhomogeneity. In the absence of specific homogeneity data, larger uncer-tainties in temperature were used. Indares and Martignole (1985) quote composition ranges rather than single analyses and these ranges are incorporated as uncertainties in composition for these data. 4.8 M A P analysis Three considerations influenced the MAP analysis. 1. Experimental data were to be fit within their uncertainties (table 4.3). 2. The natural data were to be fit within the narrowest brackets possible while maintaining consistency within the total set of all data considered. 3. Absolute values of the derived Margules parameters were to be kept as small as possible. This third point arose because it was found that the data set used was not as constrain-ing as hoped, and during 'tightening' of the P-T brackets on natural data the Margules parameters tended to become large while retaining modest differences. This process does not guarantee the accuracy of the solution parameters for biotite. However, the resulting solution parameters are consistent with the P-T constraints used and with the present thermodynamic database. To determine the correctness of the determined parameters will require further experiments and application to natural assemblages. Until more constraints can be applied, one can only test the model by applying it to other field data which have not been used in constraining the parameters (see below). In the initial stages of MAP analysis the uncertainties for natural assemblages were set at ± 1000 bars and ± 30°C. This produced a feasible solution but the parameters were not narrowly constrained. I then tightened the uncertainties on the natural data. Chapter 4. Calibration of SGAM thermobarometer 1 110 Addition of natural assemblage data to the MAP analysis constraint sets and tightening of the P-T brackets on the natural assemblage data was done step-wise in the following order to give the parameters listed in the tables. 1. The experimental data (Ferry and Spear, 1978) were augmented by the data of Indares and Martignole (1985). In the initial stages of fitting the Indares and Martignole data were tightened progressively to explore the effect on the resulting solution parameters. However, later fitting runs showed that tight (± 10°) constraints on the Indares and Martignole data produce unrealistically large biotite solution parameters (> 120 kJ/mol). The final adopted error brackets of ± 50° permit more realistic solution parameters and are reasonable considering the overall uncertainty in the natural assemblage data. 2. The remaining natural assemblage data sets were added one at a time with wide brackets. The resulting fit parameters were examined to see which data sets were most constraining and which parameters were affected most by each addition. This produced an initial feasible solution utilizing all the relevant experimental and natural data. 3. The initial feasible solution was then progressively more tightly constrained by narrowing the brackets assigned for the data of Pigage (1982) and Holdaway et al. (1988) on the Fe-SGAM reaction (4.5). Brackets were closed to ± 500 bars and ± 30°C and then some relaxed slightly to maintain feasibility. Two of the three samples of Pigage (1982) which contain sillimanite only could not be constrained to fall within the sillimanite field, perhaps due to some retrograde resetting of the garnet-biotite compositions. Chapter 4. Calibration of SGAM thermobarometer 111 4. In addition, one of the andalusite-bearing samples of Ferry (1980) plots in the sillimanite field by 30°. It was not possible to 'force' this data point into the andalusite stability field. 5. At this point the natural assemblage data were considered to be as closely con-straining as possible. The objective function was then varied to maximize and minimize various fit parameters thus delimiting the possible ranges and allowing comparison with such calorimetric data as exists. The value for the enthalpy of annite determined is largely independent of objective functions applied to the solution parameters, and is constrained by the chosen thermodynamic properties and by data points in the Ferry and Spear (1978) data set (which are Ti and Al free). 6. Finally, the biotite solution parameters were minimized, in effect producing a minimum sum. However, the data are not sufficiently constraining on the absolute value of any of the parameters. Minimization of all four solution parameters in the absence of any other restrictions produces small values for W B3 and Wfg (approximately 20 and 0 kJ/mol respectively) but unacceptably large values for W B t and W B t (> lOOkJ/mol). The final set of parameters was determined after iteratively decreasing the value of the largest of the four parameters in steps of 5 kJ/mol, yielding upper limit of 75 kJ/mol. The values quoted in table 4.5 result from minimizing all four parameters within that restriction. The final set of fit parameters (standard state values and solution parameters) are given in tables 4.4 and 4.5. It is emphasised that these values are not unique because the data used are not sufficiently constraining. However, they are self consistent and useful for thermobarometry. The annite enthalpy of formation (-5142.000 kJ/mol) is similar Chapter 4. Calibration of SGAM thermobarometer 112 to that (-5142.800 kJ/mol) of Berman (1990). Note that Berman (1990) determined a slightly different entropy and volume for annite as well. 4.9 Testing the calibration of S G A M against independent data The utility of the new calibration can be tested first against the data used in the calibration — a closed circle designed to detect errors. More interestingly, it can be tested against independent published calibrations to determine whether it offers any improvement in precision, consistency, and ability to perform well when applied to minerals having compositions far from those used to calibrate the model. An 'improved calibration' should not only satisfy data which older models could not, but should also deal well with data that were previously handled satisfactorily. In the following paragraphs it is shown that these tests are well met by the new calibration, and in particular, that it is more sensitive than GASP and in several cases can be used to demonstrate a pressure gradient where none was seen before. Pressures and temperatures have been calculated for the sample compositions given by Pigage (1982) and are shown on figure 4.6 and tabulated in appendix C. The large box in the center of the figure gives the P-T range defined for the area using the GASP barometer and garnet-biotite thermometer (plagioclase solution model of Fuhrman and Lindsley, 1988; garnet solution model of Berman 1990; and biotite model from this study). The new calibration of the SGAM barometer yields pressures and temperatures consistent with this as required by inclusion of this data set in the fitting. The results are also consistent with the field evidence, as the Ky/Sil isograd crosses Pigage's area. However, the calibrations of Holdaway et al. (1988) and Hodges and Crowley (1985) yield pressures which are too low. The difference between the new calibration and that of Holdaway et al. (1988) results from the fact that the garnets analyzed by Pigage Chapter 4. Calibration of SGAM thermobarometer 113 6500 6000 5500 u 5 0 0 0 L. ra CO 4500 J L o of/ o 0 > 0 0 o SGfln v H&C85 A HDH88 520 560 600 640 Temperature (C) 680 Figure 4.6: Pressures and temperatures determined for data of Pigage (1982) using three different calibrations. Values are tabulated in appendix C. The box in the centre of the diagram shows the P-T range denned by GASP for the same samples. Note that the new calibration (diamonds) is consistent with GASP. Also shown are the P's and T's determined using the calibrations of Holdaway et al. (1988) (upright triangles) and Hodges and Crowley (1985) (inverted triangles). These last two calibrations both give pressures which are judged to be too low, because the Pigage samples come from an area which straddles the Ky/Sil isograd. Chapter 4. Calibration of SGAM thermobarometer 114 (1982) have a quarter of the Mn content of those used by Holdaway et al. (1988) in their calibration. Hodges and Crowley (1985) were very careful to note that their calibration is enormously 'imprecise' (on the order of 10 kbar) due to the nature of the data used in the calibration. One of the most important features of the new calibration of SGAM is seen when the data are plotted on a map. Figure 4.7 is a map showing the areal distribution of pressures calculated from the data of Pigage (1982). SGAM pressure seems to show a rough trend, increasing from northeast to southwest and from approximately 4.9 kbar to approximately 5.7 kbar. In addition, this increase in pressure matches the general increase in grade in the area and the amount of pressure increase approximately matches what one might expect from lithostatic pressure given the general stratigraphic dip for the area. This pressure gradient is close to the limit of resolution expected from the calibration and from the error in microprobe analysis alone. The GASP pressures for the area show no recognizable pressure gradient (Fig. 4.7) although the mean pressure is the same. Although the data of Pigage were used in the MAP analysis, the gradient shown was not part of the constraints because such narrow constraints on individual samples could not be justified and because such a gradient was previously unknown. Thus the gradient shown is a feature purely of the data and not an artifact of the MAP analysis. Changes in the fit parameters change the absolute values of the pressures but the sense, with P increasing to the southwest, remains the same. Thus, it is felt that this gradient is a geologically significant feature required by the data. Figure 4.8 shows the pressures and temperatures calculated for the data of Hold-away et al. (1988). The data are tabulated in appendix C. The central box shows the P-T range defined by GASP barometry and garnet-biotite thermometry. The new calibration is required to be consistent with this range by use of MAP on this data set. Generally all three calibrations give the same pressures and temperatures. However, Chapter 4. Calibration of SGAM thermobarometer 115 Figure 4.7: Map from Pigage (1982) showing the distribution of SGAM pressures cal-culated with the calibration of SGAM presented here. Values in parentheses are GASP pressures for the same samples. Chapter 4. Calibration of SGAM thermobarometer 116 Figure 4.8: Pressures and temperatures for the data of Holdaway et al. (1988). The box represents the P-T range for the samples using GASP and garnet-biotite thermometry. The new calibration of the SGAM barometer (diamonds) is consistent with this P-T range. The calibrations of Holdaway et al. (1988) (upright triangles) and Hodges and Crowley (1985) (inverted triangles) are in the same range, with slightly more scatter. Chapter 4. Calibration of SGAM thermobarometer 117 the new calibration gives a slightly narrower range pressure for the samples than do the two older calibrations. The Holdaway et al. (1988) calibration gives slightly higher temperatures because they incorporate Mn non-ideality in garnet (Ganguly and Sax-ena, 1984, 1985) which is quite Mn-rich in this area. The wide spread of the P-T data represents a field gradient in P (and T?) for the area (Holdaway et ai, 1988). The difference between the Holdaway et al. calibration and the GASP and garnet-biotite box arises from the fact that Holdaway et al. utilized staurolite breakdown criteria and field evidence to estimate an average P for the area of 3.1 kbar on which they calibrated the barometer. The areal distribution of the pressures determined for the Holdaway et al. (1988) data has been examined. The picture is complicated but is similar to that given by Holdaway et al. In general the pressures increase from northeast to southwest. There are some irregularities which may be the result of later plutonism or postmetamorphic folding as shown by folded isograds (Holdaway et al. 1988). Figure 4.9 shows the pressures and temperatures calculated using the data of Hodges and Spear (1982). The data are tabulated in appendix C. This data set was not part of the MAP analysis. The central box is again the P-T range defined using GASP barom-etry and garnet-biotite thermometry. It is clearly seen that only the new calibration is consistent with these P's and T's. The new calibration is also consistent with the field observations that note the occurrence of all three aluminosilicate polymorphs in the area. The other two calibrations yield pressures which are much too low. Figure 4.10 shows the areal distribution of the pressures calculated using the data of Hodges and Spear (1982). These data show an increase in pressure from east to west. The GASP pressures do not show this gradient. The aluminosilicate bathograd drawn by Hodges and Spear (Fig. 4.10) is based on the absence of andalusite and is thus only a crude estimate of an isobar. Nevertheless isobars drawn using these data Chapter 4. Calibration o f S G A M thermobarometer 118 5000 -4500 4000 -1 m 3500 O J h 3000 tn i n O J CL 2500 o SGRM v H8C85 A HDH88 0 o o 0 o . o 2000 -1500 v V V 420 440 460 480 500 520 T e m p e r a t u r e ( C ) Figure 4.9: Pressures and temperatures for the data of Hodges and Spear (1982). Box represents the GASP and garnet-biotite pressures and temperatures for the area. Other symbols as for previous diagrams. Chapter 4. Calibration of SGAM thermobarometer 119 Figure 4.10: Map showing the distribution of SGAM pressures from the data of Hodges and Spear (1982). GASP pressures are given in parentheses for comparison. Isobar trace of Hodges and Spear (1982). Chapter 4. Calibration of SGAM thermobarometer 120 would be subparallel to the aluminosilicate bathograd as drawn — within the precision of the calibration. As noted above these data were not used in the MAP analysis and thus the results here indicate that the pressures calculated with the new calibration seem to be both accurate and precise. Temperatures for the above data sets do not show gradients. This is due to the poorer sensitivity of the garnet-biotite thermometer. For garnet compositions common to pelitic rocks temperature precision is rarely better than ±30°C. Thus the garnet-biotite thermometer demonstrates quantitative temperature gradients only in areas of high thermal gradient (aureoles) or over large areas. The data of Raeside et al. (1988), from a thermal aureole, and the data of Chipera and Perkins (1988), a large area showing pronounced change in grade, are suitable for examinations. The new parameters give almost identical temperatures to the calibration preferred by Raeside et al. (1988) on the basis of the field assemblages. The only discrepancy is at lower temperatures. Raeside et al. (1988) determined a range in T of 317-675°C. The low temperature end is based on a single sample considerably lower in T than the next highest at 484°C. The new calibration gives a range of 431-654°C. Chipera and Perkins (1988) examined a number of calibrations of the garnet-biotite thermometer using data from a upper amphibolite to granultite assemblage. Their preferred calibration (that of Perchuk and Lavrent'eva, 1983) yielded what Chipera and Perkins considered the most accurate and precise temperatures for the area (600-750°C). The precision of each calibration was evaluated by trend surface analysis of the calculated T's. Chipera and Perkins noted two features of the Perchuk and Lavrent'eva calibration: 1. The scatter from the trend surface for any degree of polynomial was smaller for the Perchuk and Lavrent'eva calibration than for the 7 other calibrations considered. Chapter 4. Calibration of SGAM thermobarometer 121 2. The Perchuk and Lavrent'eva calibration was less prone to yield different tem-peratures for samples from the same locality (or ones close together) which had different compositions. This second point is open to debate. Chipera and Perkins (1988, p. 46) cite only 2 examples of pairs of samples with different compositions which give anomalously differing temperatures with other calibrations (sample pairs BL1083C-BL1083E and HS1483C-HS1583C). Yet in the case of the BL sample pair, all of the other calibrations used by Chipera and Perkins, with the exception of the Ganguly and Saxena (1984) calibration, give similar temperatures (appendix C). In the case of the HS sample pair, the K Q ' S are very different and only the Perchuk and Lavrent'eva (1988) calibration and one of the Goldman and Albee (1977) calibrations give similar temperatures. The temperatures determined using the new calibration show slight differences to Chipera and Perkins' preferred values. The temperatures are slightly higher and show a slightly wider range, from 550 to 820°C, still consistent with the mineralogical estimates of the grade of the rocks examined. Secondly, similar temperatures (within the ±30°C range) were determined for 5 out of 6 sample pairs which have different compositions but which come from the same locality, or from close by. Temperatures for these sample pairs and the temperatures calculated by Chipera and Perkins using 8 different calibrations are tabulated in appendix C. The new calibration has been applied to a number of data sets from the literature (to date, 17) and to new data collected for a regional metamorphic study (chapter 5). A limiting factor in applying this calibration to older data is the relative paucity of studies which give detailed microprobe analyses of all phases, particularly of muscovite. In most of the cases studied, the SGAM pressures are similar (within 2 kb at most) to the GASP pressures. The cases in which a strong discrepancy occurs always give GASP Chapter 4. Calibration of SGAM thermobarometer 122 pressures considerably higher than SGAM. This probably indicates that the mechanism for the re-equilibration of the GASP assemblage is more sluggish than that of SGAM. Nevertheless, the new calibration seems to work well for many rocks. Cross-checks with GASP and GRAIL may help identify disequilibrium problems which in turn may yield information on the metamorphic history of the rocks involved. 4.10 Discussion and Conclusions There are 3 major conclusions to be made: 1. The new SGAM calibration produces temperatures and pressures which are con-sistent with GASP pressures and with the field assemblages. 2. Temperatures determined with the garnet-biotite thermometer using the solu-tion parameters presented here are as precise as those determined using previous calibrations. 3. Pressures (and temperatures) determined using the new calibration commonly show gradients. In cases where this has been seen the gradients match both the sense and amount that might be expected from field evidence. Given these observations, it must be emphasised that the parameters determined need not necessarily be the 'right' ones. Mathematical programming only constrains solutions to be consistent with the input data. Of the six binary solutions considered, Fe-Mg was assumed to be ideal on the basis of volume considerations, but the as-sumption of ideality on the Ti-Al binary is probably not correct. However, the present data do not constrain the Ti-Al binary sufficiently. In addition, all the binary solutions were considered to be symmetrical and no ternary interactions were considered. Further data, particularly experimental data on T i - and Al-rich biotites, will help constrain Chapter 4. Calibration of SGAM thermobarometer 123 these parameters better. Further application of the calibration to natural compositions will also help. Finally, it was assumed here that the SGAM barometer 'closes' at the same condi-tions as GASP, which need not be true. The effects of re-equilibration during retro-gression (garnet-biotite exchange) on the calculated SGAM pressures also need to be evaluated. Chapter 5 Metamorphism of the Quesnel Lake area: Petrography, mineral isograds, pressures, and temperatures 5.1 Introduction This chapter describes the salient features of the metamorphic history of the Quesnel Lake area. The area is large (approximately 7,500 km2) and parts of it have been described by previous workers in various degrees of detail. Most of the pre-existing data on the metamorphism for the area is in the 15 theses done on parts of the area in the last 20 years. These are the same works that form the basis of the stratigraphic and structural history presented in chapter 2. In addition to these theses, Campbell (1978) determined isograds for the area as a whole and these are shown in figure 5.1. The field areas covered by previous theses are shown in inset to map (back pocket). The size of the area has precluded detailed collection, examination, and description of specimens from the entire area and from all lithological compositions. This study concentrates on the pelitic assemblages. The suite of rocks examined included the portions of thesis collections still available at UBC as well as an additional 600 samples collected in the 1985, 1986 and 1987 seasons from localities between the study areas of the previous workers. The bulk of the additional samples come from the following areas and are indicated by the concentrations of symbols on the map (in back pocket): • scattered localities between the north and east arms of Quesnel Lake north of the field areas of Lewis (1987) and Garwin (1987), 124 Chapter 5. Metamorphism of the Quesnel Lake area 125 Figure 5.1: Sketch geologic map of the Quesnel Lake area showing the metamorphic isograds of Campbell (1978). Heavy dashed line is boundary between units 1 and 2. Inset map shows the tectonic subdivision of the Canadian Cordillera of Wheeler and Gabrielse (1972). Chapter 5. Metamorphism of the Quesnel Lake area 126 • along the shores of the east and north arms of Quesnel Lake, • along the shores of Clearwater and Azure Lakes, • and between the areas of Fillipone (1985) and Radloff (1989). Including old collections, a total of over 1000 thin sections was examined. One of the primary objectives of this study was to make quantitative estimates of pressures and temperatures for the rocks of the area. The chosen thermobarometer, SGAM, described in chapter 4, requires the assemblage quartz, garnet, aluminosilicate, muscovite, and biotite; therefore this study focussed on assemblages of garnet grade or higher. From the suite of 1000 samples examined in thin section, a suite of 35 key samples was chosen for detailed microprobe analysis. All 35 samples contained the required SGAM assemblage for thermobarometric determinations. These samples were chosen from as widely spaced localities as possible in order to give some idea of the distribution of pressure and temperature for the area as a whole. 5.2 Mineral growth and 'isograds' As noted above, rocks of at least garnet grade were examined, but the emphasis was on rocks containing one or more of the aluminosilicate polymorphs (andalusite, silli-manite, or kyanite). Consequently the collection and examination process discussed here focusses on these rocks to the exclusion of other, equally important, lithologies. A number of factors limit the detail possible in the study. 1. 'Good' pelitic assemblages are not common in rocks of unit 1 (chapter 2) which underlies the bulk of the Quesnel Lake area (map, in back pocket). As described in chapter 2, unit 1, the continental margin sequence, is a heterogeneous siliciclastic and carbonate package intruded by gneiss. The carbonate, psammitic, and gneiss Chapter 5. Metamorphism of the Quesnel Lake area 127 lithologies make up a large portion of unit 1 so that the samples used to define isograds and mineral zones are widely spaced. Pelitic rocks are absent from unit 2. Rocks of unit 3 are highly aluminous pelites but, except for the southernmost portion of the area, are not metamorphosed above biotite grade. The paucity of appropriate mineral assemblages not only affects the quantitative determinations of pressure and temperature but also the examination of the textural relationships of each of the major metamorphic minerals to each other and to the phases of deformation. 2. Only a single thin section was cut for each specimen. Ideally several thin sections, cut at different orientations, help display all the textural relationships present within any single specimen. 3. Many of the rocks in the Quesnel Lake area have been subjected to retrogression. Virtually all rocks show some chloritization and/or sericitization of the high grade minerals — particularly garnet, staurolite, and kyanite. In many rocks these are completely pseudomorphed. Retrogression not only renders a rock unusable for P-T determinations, it also obscures the textural relationships of the prograde assemblage. 5.2.1 Micas Micas (muscovite and biotite) define the prominent foliations seen in all the rocks of the Quesnel Lake area. In the polydeformed rocks there are commonly two foliations present. The lowest grade rocks examined, which come from unit 3, have a pronounced crenulation cleavage, exemplified by several samples collected by Bloodgood (1987a). A typical example of a muscovite phyllite is shown in figure 5.2. The two foliations shown in figure 5.2 are considered to be S2 and S3. These basic features are repeated in Chapter 5. Metamorphism of the Quesnel Lake area 128 Figure 5.2: Photomicrograph of muscovite phyllite showing two foliations (sample 06-04 of Bloodgood, 1987a). The earlier foliation is considered S2, the later foliation, a crenulation cleavage, is S 3 . Plane-polarized light. Chapter 5. Metamorphism of the Quesnel Lake area 129 Figure 5.3: Photomicrograph of a garnet schist sample PDL-473 of Lewis (1987) show-ing S2 bent around F 3 folds. Mica flakes do not show strain. Crossed polars. samples from all metamorphic grades with various further complications.1989fxe rocks of unit 1, which have been metamorphosed to garnet grade or higher, there may be up to 3 foliations. The earliest foliation is only preserved in poikiloblasts (see garnet section below). Figure 5.3 shows S2 and S3 in a garnet schist collected by Lewis (1987). Recrystallization of muscovite up to the time of F 3 is indicated by the presence of polygonal arcs of mica, though some samples show strained micas in the hinges of F 3 folds. Secondly, some samples show sparse grains of muscovite growing parallel to S3 in the hinges of F 3 folds (fig. 5.4). Biotite displays similar features to those shown by muscovite, except that biotite grain growth parallel to S3 in F 3 fold-hinges was not noted. Muscovite, in the form of sericite, is an important mineral in many rocks. Sericite, as a product of retrogression, is very common in the area and affects a wide belt of Chapter 5. Metamorphism of the Quesnel Lake area 130 i ^ H M H H 1 mm Figure 5.4: Photomicrograph of muscovite growth parallel to S3 in the hinges of F 3 folds at lower left. Sample SLG-262 of Garwin (1987). Crossed polars. Capter 5. Metamorphism of the Quesnel Lake area 131 rocks from Clearwater Lake, in the south, to Cariboo Lake, in the north (see map, in back pocket). Severe retrogression was noted by Getsinger (1985), Lewis (1987), Garwin (1987), and during this study (McMullin and Greenwood, 1986, 1988a). Ret-rograded samples seen during this study are indicated on the map. Sericite, with or without chlorite, replaces garnet, staurolite, and kyanite (fig. 5.5). The replacement of these minerals is a very late event as relict grains at the center of pseudomorphs show strain attributed to the final phases of deformation whereas the pseudomorphic mica is undeformed. A biotite isograd has not been drawn on the map because all the rocks studied in detail are metamorphosed to garnet grade or above. 5.2.2 Garnet Garnet is the most important porphyroblastic mineral formed. In addition to being part of the thermobarometric assemblage, garnet porphyroblasts show the widest variety of included textures and help unravel the relationship of other minerals to the phases of deformation. This is probably due to the fact that garnet seems to have grown through a protracted period of the metamorphic history of these rocks. Garnets present in rocks of unit 1 are particularly inclusion rich. However, the rocks of unit 3 are more aluminous and garnets therein are virtually inclusion free. The multi-deformational textures preserved in and around garnet from unit 1 are exemplified by the sample illustrated in figure 5.6. This example (PDL-473 of Lewis, 1987) clearly shows a folded foliation wrapping around the garnet. However, the poik-iloblastic pattern within the garnet illustrates that the folded foliation outside the garnet was itself a crenulation cleavage. The three foliations are interpreted as S i , the crenulated foliation within the garnet, S2, the crenulation cleavage inside the garnet and the earlier of the two external foliations, and S 3 , the final crenulation cleavage external to the garnet. Textures this clearly developed are not common but they are Chapter 5. Metamorphism of the Quesnel Lake area 132 Figure 5.5: A: Garnet almost entirely replaced by sericite-chlorite mix (sample DM-86-153a). B: Kyanite (or staurolite?) replaced by sericite (sample DM-86-130). Both photographs crossed polars. Chapter 5. Metamorphism of the Quesnel Lake area 133 Figure 5.6: Sketch of complex inclusion pattern seen in sample PDL-473 of Lewis (1987). Two foliations are seen external to the garnet and two are seen in the inclusion pattern. The second internal foliation is equated with the first external foliation (both S2). The earliest foliation (internal) is considered to be S i , the latest (external) foliation is considered to be S 3 . reasonably widespread in rocks of unit 1. Several examples were noted in the samples collected by Lewis (1987), Getsinger (1985), and Fletcher (1972) and additional cases were noted in samples collected for this study (fig. 5.7). There are many examples where garnet porphyroblasts contain highly contorted inclusion patterns whose rela-tionship to the external foliation is unknown (fig. 5.8). In some cases foliations within the garnet have a distinct relationship. For example in figures 5.6 and 5.7 the second internal foliation is a crenulation of the first. However, in the case of the sample shown in figure 5.8 the two foliations occur in separate parts of the garnet grain and one Chapter 5. Metamorphism of the Quesnel Lake area 134 Figure 5.7: Photomicrograph of inclusion patterns within garnet in sample DM-86-154. Inclusions indicate a crenulation cleavage (S2) which is subparallel to the external foliation. However, the external foliation is defined by strongly deformed mica 'fish' and may represent a transposed foliation S 2 / S 3 . Chapter 5. Metamorphism of the Quesnel Lake area 135 Figure 5.8: Photomicrograph of garnet with two stages of growth, each of which has an associated inclusion pattern (sample DM-85-16). Core has almost straight inclusion pattern (Si?), rims show sigmoid inclusion trails (S2?) which are discordant to the external foliation (S3?). Chapter 5. Metamorphism of the Quesnel Lake area 136 cannot tell the relationship between them except to state that the foliation in the core is presumably older than that in the rim which is in turn older than that shown in the matrix. There are many cases where swirled and contorted inclusion patterns cannot be easily assigned to individual phases of deformation. The simplest conclusion to be made is that garnet growth in unit 1 started after Fi and continued through F 2 . Garnet growth may have continued through the initial stages of F 3 , especially in the highest grade rocks. However, the garnets in higher grade rocks were formed, with abundant fibrolitic sillimanite, at the expense of ear-lier porphyroblastic minerals (garnet and staurolite) and are small, idioblastic, and inclusion-free (see also Fletcher, 1972; Getsinger, 1985; Garwin, 1987). Thus the rela-tionship of these garnets to earlier formed ones is difficult to determine. In addition, many samples contain garnet which has been retrograded (to chlorite ± sericite) which also clouds the relationship of garnet growth to foliation formation. Because the rocks of unit 3 are more aluminous than those of unit 1, garnet in these rocks is less poikiloblastic. Quartz inclusions are sparse and very small but where present show straight inclusion trains not the complex patterns seen in unit 1 garnets (fig. 5.9). Mica inclusions are more common in unit 3 garnets and commonly are randomly oriented (fig. 5.10). These observations indicate that garnet growth occurred at the same time as the first phase of deformation. However, the external foliation, which postdates the garnet growth, is S 3 . The isograds are folded by F 3 folds such as the syncline east of Crooked Lake (see map). This indicates that the first foliation in these rocks is S2, confirming the observations made in chapters 1 and 2 that the rocks of unit 3 have experienced one less phase of deformation than those of unit 1 (and 2). The garnet isograd is shown on the map (back pocket) and has been folded by F 3 folds. The isograd can only be roughly positioned in the unit 1 rocks at Quesnel Chapter 5. Metamorphism of the Quesnel Lake area 137 Figure 5.9: Photomicrograph of small quartz inclusion in garnet in sample DM-87-54 from unit 3. Garnet has been fractured and pulled apart parallel to F 3 extension direction. Adjacent kyanite grain show kink bands associated with F 3 . Crossed polars. Chapter 5. Metamorphism of the Quesnel Lake area 138 Figure 5.10: Photomicrograph of randomly oriented mica inclusions in garnet from sample DM-87-35 from unit 3. Garnet is included in staurolite grain which has has been wrapped by S3 and which has an internal foliation S2. Crossed polars. Chapter 5. Metamorphism of the Quesnel Lake area 139 Lake. This is because of the paucity of garnetiferous assemblages. Gneiss, quartzite, psammite and carbonate lithologies constitute the bulk of unit 1 close to the boundary with unit 2 in the Quesnel Lake area. 5.2.3 Staurolite The relationship between garnet and staurolite growth is not clear and likely varies throughout the area. In general it appears that staurolite growth postdated some of the garnet growth as indicated by the presence of garnet porphyroblasts within staurolite grains (fig. 5.10). However, in most cases garnet and staurolite appear to have grown together, particularly in the later stages of staurolite growth. At the highest grades staurolite breaks down to sillimanite probably with the growth of biotite or the small, idioblastic garnets mentioned above. Possible reactions are: 6 St + 25 Qtz = 46 Sil + 8 Grt + 12 H 2 0 and 6 St+ 8 Ms+ 17 Qtz = 62 Sil + 8 Bt + 12 H 2 0 . Figure 5.11 shows a rare example of the relationship between staurolite and garnet growth and the development of the foliations. The figure shows a staurolite grain, now significantly altered to sericite, which enclosed several garnet grains. A foliation defined by opaque grains is seen internal to the staurolite but which wraps around the garnet. This foliation is interpreted as S2, on the basis of the arguments presented in the garnet section above. Thus the staurolite growth postdated F 2 deformation. The foliation external to the staurolite is considered to be S 3 . This example indicates staurolite growth over S2 without any rotation. There are also examples which show rotation of the staurolite during the development of S2- Figure 5.12 shows inclusion trains within staurolite asymptotic to the external foliation (which is interpreted as S2 transposed parallel to S 3 ) . This may indicate some staurolite growth up to the early Chapter 5. Metamorphism of the Quesnel Lake area 140 • • • • 4 mm Figure 5.11: Sketch of a retrograded staurolite grain from sample DM-86-354 show-ing garnet porphyroblasts with a foliation wrapping around them included within the staurolite. Foliation external to the garnet and internal to the staurolite is considered S2- Foliation external to staurolite is considered to be S 3 . Chapter 5. Metamorphism of the Quesnel Lake area 141 Figure 5.12: Photomicrograph of sample DM-86-173 showing quartz inclusions within staurolite asymptotic to external foliation S 2 / S 3 . Crossed polars. stages of S 3 . Both the above examples come from rocks of unit 1. Rocks of unit 3 from the southern part of the Quesnel Lake area also show spectac-ular inclusion patterns. There, a small infolded syncline of unit 2 and unit 3 rocks have been metamorphosed to staurolite/kyanite grade. Radloff (1989) studied some of these rocks and concluded that there were two stages of staurolite growth, one 'prograde', one 'retrograde'. It is not clear from the text of Radloff (1989) what criteria were used to define 'prograde' versus 'retrograde'. Examination of the samples of Radloff and those collected for this study just to the north of Radloff's area indicate that stauro-lite shows spectacular zonation patterns such as that shown in figure 5.13. These do not necessarily indicate several phases of growth in the sense of different metamorphic conditions. Secondly there do not seem to be two generations of staurolite crystals but rather a continuum of grain sizes representing an extended period of growth. Thirdly, Chapter 5. Metamorphism of the Quesnel Lake area 142 whereas staurolite grains do have spectacularly developed inclusion patterns these are not geometrically complex. The orientation of the inclusion trails appears to coincide with the zoning patterns. The cores are inclusion rich (particularly with opaque mate-rial) and the defined foliation is generally straight and at a high angle to the external foliation. The rims have few inclusions (in comparison to the cores) and have inclusion trails that are sigmoidal and asymptotic to the external foliation. The external foli-ation is S3 and is clearly a crenulation cleavage (fig. 5.14). The sigmoid, asymptotic inclusion trails in rim zones indicate some growth of staurolite during the development of S 3 . The straight inclusion trails in the cores indicate that the earliest staurolite growth was over a pre-existing foliation, S2 (fig- 5.13). Microboudinage of staurolite during the formation of S3 is well developed in some samples (fig. 5.15). The evidence indicates that growth of staurolite started after that of garnet. In rocks of unit 1, staurolite contains garnet porphyroblasts and traces of an enveloping foliation that is considered to be the second foliation present in these rocks. The foliation external to the staurolite is considered to be S3 (or later). In rocks of unit 3, staurolite contains garnet porphyroblasts but the internal foliations are the first foliation seen in these rocks and are also considered to be S2. The external foliation, S 3 , is a crenulation cleavage or transposed foliation and staurolite growth may have continued during its earliest stages. However, staurolite grains are commonly microboudinaged with the extension direction in the plane of S 3 . A staurolite (only) isograd was drawn by Getsinger (1985) for the area northwest of Quesnel Lake (shown on map in back pocket). Lewis (1987) also drew a staurolite isograd. However, Lewis also noted that the appearance of staurolite before kyanite seemed to be a function of bulk rock composition. In general, for the Quesnel Lake area, separate staurolite and kyanite isograds cannot be distinguished and a combined staurolite/kyanite isograd is shown on the map. Getsinger (1985) also distinguished Chapter 5. Metamorphism of the Quesnel Lake area 143 • • H H 1 mm Figure 5.13: Photomicrograph of part of staurolite grain from sample DM-87-54 show-ing two distinct zones. Inner zone has small quartz inclusions defining a planar foliation at a high angle to the external foliation. The rim zone is less poikiloblastic and has inclusion trails asymptotic to the external foliation. Crossed polars. Chapter 5. Metamorphism of the Quesnel Lake area 144 Figure 5.14: Photomicrograph of sample DM-87-34 showing crenulation cleavage, S 3 , wrapping around staurolite grain. Staurolite grain shows well developed zoning. Crossed polars. Chapter 5. Metamorphism of the Quesnel Lake area 145 Figure 5.15: Photomicrograph of part of sample JVR-2 (Radloff, 1989) showing zoned staurolite grain microboudinaged with the extension direction in the S3 plane. Chapter 5. Metamorphism of the Quesnel Lake area 146 a staurolite out isograd. In that area Getsinger showed that staurolite 'disappeared' in both sillimanite and kyanite fields. In the high grade rocks to the southeast of the map area, staurolite is seen to be reacting out and many samples are staurolite-free. However, it is present in several of the microprobed samples taken from the sillimanite zone and thus it is not possible to draw a staurolite out isograd for these rocks. 5.2.4 Kyanite Kyanite appears to have developed at the same time as staurolite. However, kyanite is generally non-poikiloblastic and thus not easily placed in a time sequence relative to the surrounding foliations. One rare example of a poikiloblastic kyanite grain is shown in figure 5.16. This shows a straight internal foliation continuous with the external foliation which is crenulated into broad arcuate folds. The earlier foliation is considered to be S2 and the later folds are S 3 . In addition, like staurolite, kyanite is occasionally seen boudinaged with the extension direction in the plane of S 3 , particularly in rocks of unit 3 in the southern part of the area (fig 5.17). Kyanite growth probably ceased before the development of F 3 features as F 3 deformation features such as the kink bands seen in figure 5.9 are unannealed. In addition there is evidence that the F 3 deformation seen in the southern part of the area occurred in the andalusite field of stability. The microboudinaged kyanite grain shown in figure 5.17 has a small attached grain of andalusite and the presence of small sphene inclusions in both kyanite and the adjacent andalusite indicates inversion of kyanite to andalusite during F 3 deformation. See andalusite section below for further details. A kyanite isograd was drawn by Getsinger (1985) for the area north-west of Quesnel Lake. Lewis (1987) also showed a kyanite isograd but considerably closer to the stau-rolite isograd. In addition, Lewis stated that the appearance of staurolite or kyanite in rocks was mostly a function of bulk composition. In general, one cannot distinguish Chapter 5. Metamorphism of the Quesnel Lake area 147 Figure 5.16: Photomicrograph of sample DM-85-42 showing a large poikiloblastic kyan-ite grain. The inclusions (quartz and mica) define a foliation that is planar. This is continuous with the foliation outside the grain which is bent into arcuate folds. The folds are considered F 3 and the earlier foliation is S2. Crossed polars. Chapter 5. Metamorphism of the Quesnel Lake area 148 Figure 5.17: Sketch of microbouclinaged kyanite grain from sample DM-87-42. Grain is aligned parallel to S3 with remanent S2 between the domains of S 3 . A small area of andalusite at the margin of kyanite fragment containing the same included minerals as kyanite indicates inversion of kyanite to andalusite, probably during F 3 . Chapter 5. Metamorphism of the Quesnel Lake area 149 separate staurolite and kyanite isograds for the area, because of the paucity of pelitic assemblages and because of the degree of retrogression of many samples. 5.2.5 Sillimanite Sillimanite growth textures provide virtually no information on the timing of growth. This is mainly due to the tendency for mimetic growth of sillimanite on micas, par-ticularly biotite. However, it is probably synchronous with or slightly postdates F 2 deformation. Sillimanite is mostly fibrolitic in form with coarser grains appearing only in the highest grade rocks. Sillimanite appears at the expense of staurolite, garnet, kyanite, and ultimately muscovite. No unambiguous reaction textures are seen but probable reactions include the two reactions given in the staurolite section above as well as: Grt + Ms = Bt + 2 Sil + Qtz, Ky = Sil and Ms + Qtz = Kfs + Sil + H 2 0. Only a single sillimanite isograd is drawn on the map. Previous workers have drawn several sillimanite 'isograds' that mark the appearance of sillimanite at the expense of different, lower grade minerals (e.g., Fletcher, 1972). A single isograd is preferred as several different minerals may be reacting to form sillimanite at any pressure and temperature and the persistence of a particular lower grade mineral is probably more related to the mineral composition and to its abundance in the rock. 5.2.6 Andalusite Campbell (1971), was the first and only worker to date to document the presence of andalusite in rocks of this area. The area studied by Campbell was visited during Chapter 5. Metamorphism of the Quesnel Lake area 150 this study to collect samples because the original specimens were not readily available. Andalusite was found in a total of 4 samples from unit 1 in the southernmost part of the area (samples DM-87-17, -19, -24, and -42). In addition, a sample originally collected by Getsinger (1985) from northwest of Quesnel Lake (JSG-81-278) also contained a single grain of andalusite. In most cases very little andalusite is present (usually only a single grain) and was identified using both optical properties and X-ray diffraction (Gandolfi Camera). Figure 5.18 shows the andalusite in sample JSG-81-278 (unit 1). The andalusite is poikiloblastic and overgrows the major foliation (S3?). This feature is seen in several of the samples from the southern part of the area (rocks of unit 3). Figure 5.19 shows an andalusite grain in sample DM-87-19 that is partly overgrowing and enclosing a kyanite porphyroblast and in sample DM-87-17 (fig 5.20) andalusite partly surrounds a symplectite of sillimanite and biotite (after garnet?). All these features indicate that andalusite is one of the last minerals formed in these rocks and probably represents a decompression of these rocks while they were still hot (see thermobarometry below). The andalusite in most samples is deformed and shows extensive sub-grain development. This matches the kink banding seen in kyanite throughout the area and may be late F 3 deformation or later (F4). An andalusite isograd is not drawn on the map because • very few andalusite-bearing samples were discovered, • andalusite appears to be a late, retrograde(?) or decompression phase and is not part of the normal prograde assemblage. Chapter 5. Metamorphism of the Quesnel Lake area 151 Chapter 5. Metamorphism of the Quesnel Lake area 152 Chapter 5. Metamorphism of the Quesnel Lake area 153 H B B ^ ^ H 1 mm Figure 5.20: Photomicrograph of sample DM-87-17 showing andalusite (high relief) partly enclosing a symplectite of biotite and sillimanite (after garnet?) at bottom right. Plane polarized light. Short dimension is 3.2 mm. Chapter 5. Metamorphism of the Quesnel Lake area 154 5.3 Microprobed specimens: Mineral inhomogeneity and zonation Mineral inhomogeneity and zonation is an important feature of the rocks analyzed for this study. It is necessary to try to identify the extent, the causes, and the effects on thermobarometric determinations. In the following discussion the term inhomogeneity refers to random (or apparently random) variation in composition particularly of the rims of garnet grains. The term zonation refers to compositional variation that is perceived to have a spatial sense (i.e., rim to core variation). In this way garnets can be referred to by any combination of the terms, homogeneous or inhomogeneous and zoned or unzoned. An inhomogeneous, unzoned garnet shows variation in rim compositions but not systematic variation across the grain. The following discussion deals at length with the variation seen in garnet. 5.3.1 Garnet Garnet is the most commonly zoned mineral in metamorphic rocks and has been exten-sively studied since the advent of the electron microprobe in the early 1960's (see Tracy, 1982 for a recent review). Tracy (1982) noted two major types of zoning, growth zoning and diffusion zoning. Both occur because intracrystalline diffusion in garnet appears to be slow except perhaps at very high grade (700+ °C). Growth zoning occurs when the continued growth of garnet sequesters elements that are present in the rock in low concentrations thereby lowering them further (e.g., Mn). Because diffusion is slow, Mn does not move from the relatively Mn-rich core to the Mn-poor rim. Growth zoning need not be the result of changes in metamorphic conditions. Diffusion zoning occurs when a change in metamorphic conditions causes re-equilibration between the garnet rim and the adjacent matrix minerals. Because diffusion is slow only a small volume of the garnet changes to the new equilibrium composition and a diffusion profile forms Chapter 5. Metamorphism of the Quesnel Lake area 155 between this 'new' composition and the old. This process need not require a change in the modal amount of garnet present. However, to these two types of zoning should probably be added a third, 'reaction' zoning, and a fourth 'resorption' zoning. 'Reac-tion' zoning is produced by a change in the reaction that is producing garnet. This can produce discontinuities in the zoning profile whereas growth zoning and diffusion zoning as defined above do not. 'Resorption' zoning occurs when some of the garnet constituents become reactants in net transfer reactions thus reducing the modal abun-dance of some of the garnet constituents. Constituents not involved in the reaction (e.g., spessartine) become highly concentrated at the sites of reaction. Natural garnets generally display some combination of these types of zoning. The garnets examined during this study show a wide variety of zonation patterns and in some cases two garnets from the same sample show different patterns. However, one caveat should be made. Detailed studies of garnet zoning such as that of A.B. Thomp-son et al. (1977) or Tracy et al. (1976) require a large number of analyses per grain, at least 50 and preferably 100 or more. Thompson et al. studied a single grain that showed extremely complex zoning and Tracy (1982) specifically warned that random profiles across such a grain would be very difficult to interpret and he suggested that, in some studies, small numbers of analyses have lead to incorrect interpretations of the origins of the zoning. In this study, the primary goal was the determination of pressures and temperatures for the 35 chosen samples. Consequently, the focus is on determining rim compositions of as many garnet grains within the sample as possible. In virtually all cases some analyses of core compositions were done, but these usually consisted of single traverse of 10 to 20 points across a single grain within the sample. No more than 65 analyses of garnet were done in any individual sample and in samples with very small, skeletal, or corroded garnet, as few as 10 were possible. A total of 1100 garnet analyses were done on the 35 samples. Chapter 5. Metamorphism of the Quesnel Lake area 156 Table 5.1 gives a summary of the zoning information for all garnet analyses. Zoning profiles for 12 of the samples are given in figures 5.21 to 5.24 in the following pages. Some details of the analytical conditions, the standards used, element detection limits, as well as selected garnet, biotite, and muscovite analyses for each sample are given in appendix D. Samples are referred to both by the map index number (prefaced by a '#' symbol) and by the original sample number. The most notable feature of garnet compositions is their variability. Except for some of the samples from the highest grade, all samples have inhomogeneous rims (ranges greater than 2 mole % in one or more of the constituents). Also, most samples contain zoned garnet. Several samples show different zoning in different garnet grains within that sample (sample #'s 2, 14, and 19, table 5.1). The four types of zoning described above refer to four processes which can produce zoning. Any individual garnet is likely the result of these and probably other, unrecognized processes. In the following discussion, the zoning profiles are divided into morphological 'classes' and labelled using the following convention. The label, a group of four letters, each of which may be one of the letters u (up), d (down), and s (stable) and describe the direction of change of the amount of grossular, pyrope, almandine, and spessartine (in that order) f r o m r i m t o core. That is, the label 'udds' describes a profile in which, in going from rim to core, grossular increases, pyrope decreases, almandine decreases, and spessartine remains stable. There are 81 combinations of the letters u, d, and s in groups of four but only 50 possible classes. Because the sum of all constituents is constant (100%) profiles such as 'uuuu' are not possible. The 26 examples of zoning given in table 5.1 belong to 16 of the 50 classes, but 15 of the samples belong to 3 classes or related classes. There are also general trends. In 20 of the 26 samples grossular increases (from rim to core) or remains stable. In 16 of the samples pyrope decreases or remains stable, in 21, almandine decreases or remains Chapter 5. Metamorphism of the Quesnel Lake area 157 Sample Spot / Rim / Homog. # Number Profile Core n Zoning Grs Prp Aim Sps 1 JSG-81-234 s r c 42 ir? uz 18-20 6-10 71-73 2 2 JSG-81-278 s r c 65 ir z T 5-9 f 9-14 1 78-74 1 7-2 1 11-10 I 10-7 T l 74-78-73 IT 5-1-9 3 JSG-80-30 s p r c 35 ir z? 1 7-5 T 9-10 i 82-80 T 3-5 4 DM-85-9 s r 28 ir 9-26 7-14 60-77 0-6 5 PDL-447 s r 22 ir 9-13 10-14 77 0 6 DM-85-42 s p r c 23 ir z T 8-25 I 13-5 i 76-63 T 1-7 7 SLG-370A s p r c 37 ir z T l 5-11-8 I 13-10 1 77-74 1 5-2 8 DM-85-87 s p r c 51 ir z I 8-5 10-11 1 80-75 T 0-8 9 CJNF-9 s r 28 ir? 5-7 12-13 78-79 3-4 10 DM-86-70 s p r c 33 ir z T 13-18 i 13-8 T 70-74 0-1 11 DM-86-173 s r c 37 ir? z T 7-10 I 15-10 77-79 1 4-0 12 DM-86-197 8 r 10 ir 5-6 10-11 73-77 5-9 13 DM-86-205 s r 25 ir 6-9 10 74-77 5-8 14 DM-86-227 8 r c 27 hr z T 5-17 I 10-6 1 78-66 T 6-11 T 5-13 I 10-8 77-78 I 6-0 15 DM-86-270 S r c 31 hr z T 5-7 t 10-14 71-73 i 11-7 16 DM-86-281 S r c 46 hr z 7 T 10-16 1 76-74 1 6-2 17 DM-86-351 S r 14 ir 7-18 9-13 77-82 0-2 18 DM-86-369 S r 14 ir? 5-6 10-12 78-80 3-4 19 DM-86-384 s P r c 58 hr z T 8-17 [ 14-10 1 70-64 1 7-4 T 11-17 i 14-10 1 69-60 T 6-11 20 JRM-0626-36 s r 23 ir 5-17 10-16 68-74 1-6 21 JRM-0716-76 s p r c 29 hr z T 4-9 I 14-11 79-80 1 2-0 22 LCP-40 s r c 28 hrz? 1 5-4 I 12-10 79-80 T 2-5 23 LCP-223 s p r c 29 hrz? 1 9-7 T 10-11 77-78 T 3-4 24 DM-87-76 s p r c 39 ir? z 5 13 1 77-70 T 4-12 25 DM-87-71 s p r c 35 ir? z 1 5-4 T 9-11 79-80 1 6-4 26 DM-86-45 s r 10 ir? 5-6 12-14 73-74 7-8 27 JAF-7-9-29 s r c 34 ir? z 6-7 11-14 T 74-77 1 7-3 28 DM-87-8 s p r c 14 ir z T 3-20 I 13-5 1 83-73 0-6 29 DM-87-17 s p r c 43 ir z T 3-15 T 11-16 I 78-70 1 7-1 30 DM-87-24 s p r c 36 ir z 4-5 T 10-16 T 69-75 1 17-5 31 DM-87-32 s p r c 30 ir z T 5-16 I 15-10 1 75-70 T 1-10 32 DM-87-35 s p r c 34 ir z? 5-11 10-13 75-78 2-4 33 DM-87-39 s r c 37 ir? z 5 T 11-14 T 75-77 1 9-4 34 DM-87-54 s p r c 37 ir z T 6-20 1 13-5 1 73-67 T 2-10 35 JKR-88-4 s r 25 ir 6-10 10-13 77-80 0-4 Table 5.1: Summary table of the inhomogeneity of rim compositions and zoning in cores of microprobed garnets. Column 1 gives index number (from map), column 2 gives samples number (see appendix D for petrographic data and original source), column 3 indicates whether analyses are individual spots (s) or automated profiles (p), column 4 indicates whether rims (r) or cores (c) were analyzed, column 5 gives number of analyses done, and column 6 indicates whether rims are homogeneous (hr) or inhomogeneous (ir) and if cores are zoned (z) or unzoned (uz). Columns 7 to 14 indicate the ranges in composition for garnet in that sample. For zoned garnets, the arrows indicate sense of zoning from rim to core (| constituent increasing, j constituent decreasing). Ranges without arrows indicate change in composition is small (unzoned) or there is no particular sense to the change (inhomogeneous). Chapter 5. Metamorphism of the Quesnel Lake area 158 stable, and in 16, spessartine decreases or remains stable. 5.3.1.1 Growth and modified growth zoning The most common class is 'uddu' which is the classic growth zoning profile and is seen in sample #'s 6 (DM-85-42 profile, fig. 5.21), 14, 19, 31 and 34 (DM-87-32 and D M -87-54 profiles fig. 5.24). The inner portion of the profile of sample #2 (JSG-81-278) is class 'sddu' and may also be growth zoning (fig. 5.21). The next most common class is 'uddd' seen in sample #'s 7 (SLG-370A profile, fig. 5.21), 19 (DM-86-384 profile, fig. 5.22), and 28 (DM-87-8 profile, fig. 5.1). Class 'udds' profiles seen in sample #'s 11, 14, and 21 are considered the same as 'uddd'. These classes are probably also growth zoning slightly modified by 'reaction zoning' or 'resorption zoning' as described above. The outer portion of the profile of sample #2 (JSG-81-278 profile, fig. 5.21) may also be resorption zoning. The distinguishing feature, higher Mn contents at the rim, may result from the outer rim forming from a different mineral assemblage than the core, or, slight resorption of the garnet by reaction of pyrope, grossular, and almandine. A reaction mechanism for this is discussed below. 5.3.1.2 Resorption zoning The third most common class of profile is 'suud' which is seen in sample #'s 27, 30 (DM-87-24 profile, fig. 5.24), and 33. These are exemplified by variable low to very high Mn contents at the rim. The data from sample #30 (DM-87-24) yield some clues as to the processes involved. The following observations of textures and chemical variation are important. 1. Garnet is not abundant, and the grains present appear to be corroded fragments of earlier grains. Chapter 5. Metamorphism of the Quesnel Lake area 159 Figure 5.21: Garnet zoning profiles for samples JSG-81-278, DM-85-42, and SLG-370A (f^'s 2, 6, and 7 on map). Mole fraction of constituents are plotted along Y-axis. The length of the profile is indicated on the X-axis as are the positional nature of the end points (rim, core). Analyses (larger dots) are from automated runs and are equally spaced across the profile. Constituents are almandine (Aim, solid line), pyrope (Prp, long dash), grossular (Grs, short dash), and spessartine (Sps, dotted line). Gaps in the profile indicate bad analysis (cracks or inclusions). Chapter 5. Metamorphism of the Quesnel Lake area 160 Figure 5.22: Garnet zoning profiles for samples DM-85-87, DM-86-70, and DM-86-384 (#'s 8, 10, and 19 on map). Mole fraction of constituents are plotted along Y-axis. The length of the profile is indicated on the X-axis as are the positional nature of the end points (rim, core). Analyses (larger dots) are from automated runs and are equally spaced across the profile. Constituents are, almandine (Aim, solid fine), pyrope (Prp, long dash), gTossular (Grs, short dash), and spessartine (Sps, dotted line). Gaps in the profile indicate bad analysis (cracks or inclusions). Chapter 5. Metamorphism of the Quesnel Lake area 161 0.9 0.8 -I eg 0.7 I 0.2 0.1 0.0 JRM-0716-76 . Aim . -«.. Grs Sp8 rim 1.4 mm rim 0.8 0.7 0.6 0.2 0.1 • 0.0 DM-87-8 , Aim ——•—•—. Grs Prp Sps 0.8 mm core 0.8 i 0.7 0.6 0.2 • 0.1 0.0 DM-87-17 Aim rim T».«-' X,— Prp — ' Grs Sps '•••i 1.1 mm core Figure 5.23: Garnet zoning profiles for samples JRM-0716-76, DM-87-8, and DM-87-17 (#'s 21, 28, and 29 on map). Mole fraction of constituents are plotted along Y-axis. The length of the profile is indicated on the X-axis as are the positional nature of the end points (rim, core). Analyses (larger dots) are from automated runs and are equally spaced across the profile. Constituents are, almandine (Aim, solid line), pyrope (Prp, long dash), grossular (Grs, short dash) and spessartine (Sps, dotted line). Gaps in the profile indicate bad analysis (cracks or inclusions). Chapter 5. Metamorphism of the Quesnel Lake area 162 0.8- DM-87-24 0.7 o 0.6 to Z i ? 0.2 1 Aim - » . Sps 0.1 V. Prp —• Qrs 0.0 rim 0J2nvn rn 0.8 n 0.7 g 0.6 1 1 0 .2 0.1 DM-87-32 Aim ........ 0.0 rim Prp Grs Sps 1.1 mm rim 03 • 0.7 • 6 0.6 • | 0.2 • 0.1 • 0.0 DM-87-54 Aim Qrs Sps Prp 1.2 mm Figure 5.24: Garnet zoning profiles for samples DM-87-24, DM-87-32, and DM-87-54 (#'s 30, 31, and 34 on map). Mole fraction of constituents are plotted along Y-axis. The length of the profile is indicated on the X-axis as are the positional nature of the end points (rim, core). Analyses (larger dots) are from automated runs and are equally spaced across the profile. Constituents are, almandine (Aim, solid line), pyrope (Prp, long dash), grossular (Grs, short dash) and spessartine (Sps, dotted line). Gaps in the profile indicate bad analysis (cracks or inclusions). Chapter 5. Metamorphism of the Quesnel Lake area 163 2. The remaining garnet grains are surrounded by biotite and plagioclase. 3. The garnets are characterized by very large variations in Mn content at the 'rim'. 4. Matrix biotite composition (homogeneous) is different from that of biotite inclu-sions. Of particular interest is the lower Al content of matrix biotite. The reaction Ca 3 Al 2 Si 3 0 1 2 - f (Fe,Mg)3Al2Si30i2 + KAl 2Si 3A10 1 0(OH) 2 = K(Fe,Mg) 3Si 3A10 1 0(OH) 2 + 3 CaAl 2 Si 2 0 8 (5.1) may be responsible for both textures and the zoning. Figure 5.25A is a plot of the almandine and pyrope concentrations versus spessartine concentration for 12 rim analyses of a single grain from sample #30. Analyses are from spots within 20 u-m of the grain edge. The extremely linear trends are not a function of closure alone (components sum to a constant) and are not seen in other inhomogeneous rim analyses from other samples (e.g., sample #29 shown in figure 5.26). The only requirement of closure in a four component system is that any one component varies inversely to the s u m of the others. There is no requirement that any two show a linear relationship. Therefore, the linear trends shown in figure 5.25 indicate that although all analyses were done on the 'rim' of a garnet grain they define some profile. If the profile is the result of exchange (e.g., Fe-Mg with biotite) linear variation is only visible in the exchanging components (i.e., Sps and Grs will be constant). Therefore the profile in sample #30 is not the result of exchange. If the profile is the result of net transfer of a component, the other components will vary inversely with that component but the ratio of the other components will remain constant. As shown in figure 5.25B the ratio of Mg/(Mg+Fe) varies for sample #30. Thus net transfer alone will not produce the Chapter 5. Metamorphism of the Quesnel Lake area 164 profile. However, a combination of net transfer and diffusion will produce the profile seen in sample #30. The progress of reaction 5.1 from left to right as written results in resorption of gar-net, the appearance of biotite with plagioclase and, because spessartine is not involved, relict garnet will become significantly Mn-enriched. This will produce linear trends in the amounts of all the components. If one starts with the least Mn-rich composition shown in figure 5.25, the dotted lines on 5.25A show the expected variation in Aim and Prp contents with changing Sps enrichment by reaction 5.1 alone. No exchange occurs and the Mg/(Mg+Fe) ratio remains constant for changing Mn enrichment (fig. 5.25B). However, the lines defined by real compositions have rotated away from dot-ted line positions indicating that exchange with the matrix has occurred. This can be explained by the possibility that the progress of reaction 5.1 generates biotite with the same Mg/Mg+Fe ratio as the original garnet (i.e., Fe-rich, Mg-poor). It may also be Al poor if the replacement of octahedral Al in muscovite by the Mg and Fe from garnet is 'efficient'. This 'new' biotite will homogenize with the existing 'old' biotite much faster than garnet will homogenize. If a significant amount of garnet was resorbed during the reaction the final biotite composition will be significantly different to the biotite present before initiation of reaction 5.1. The garnet Mg/Mg+Fe ratio, which has not changed by the progress of reaction 5.1, is not in equilibrium with the new biotite (even if the reaction occurs under isothermal conditions). Thus diffusion of Mg and Fe between the two phases must take place to restore chemical equilibrium. The closer in space the garnet is to the new biotite the greater the amount of diffusion that will have occurred. The most Mn-rich garnet must be at the rim of the garnet and therefore must show the greatest change in Mg/(Mg+Fe) ratio in agreement with figure 5.25. There are two conclusions from the foregoing arguments: Chapter 5. Metamorphism of the Quesnel Lake area 165 .76-.74 .70-.68 0>: O Prp • Aim —i 1 1 1 1 1 1 1 1 1 1 1 ' .06 .08 .10 .12 .14 .16 .18 Sps .16 .14 -.10 -.08 .18-e .16-u. + co 2 CD 2 .14 .12 .06 .08 .10 .12 Sps .14 • i .16 - l — .18 Figure 5.25: A: Plot of Aim and Prp content versus Sps content for 'rim' analyses from sample DM-87-24, showing linear trends. B: Plot of Mg/(Mg+Fe) for the same analyses. Compositions along dashed lines in A define values of Mg/(Mg+Fe) ratio defined by dashed line in B. Dotted lines in A define coexisting Aim and Prp contents that yield a constant Mg/(Mg+Fe) ratio as shown by dotted line in B. Chapter 5. Metamorphism of the Quesnel Lake area 166 O Prp • Aim .79" -.16 .77- -.14 E < 8 * . TJ .75- -.12 O Ot .73- -.10 —i 1 1 1 1 1 1 1 ' r .01 .03 .05 .07 .09 Sps Figure 5.26: Plot of Aim and Prp content versus Sps content for rim analyses from sample DM-87-17. The linear trends shown on previous figure are absent. Chapter 5. Metamorphism of the Quesnel Lake area 167 1. Although these analyses were taken at the rim (all within 20 /xm of the edge) they come from different positions on a well developed, if poorly known, profile. 2. The garnet that has the highest Mn content, and shows the greatest change in Mg/(Mg+Fe) ratio must be the closest to the composition that is in equilibrium with the matrix biotite. Rough mass-balance calculations for equilibrium 5.1 using core garnet, biotite inclu-sions, and muscovite inclusions as an initial assemblage can produce the compositions of the rim garnet and matrix biotite and muscovite compositions. 1. The initial garnet composition is: 0.053 Grs; 0.168 Prp; 0.725 Aim; and 0.054 Sps. 2. Removal of 70% of the effective amount of garnet produces garnet of composition: 0.047 Grs; 0.149 Prp; 0.644 Aim; and 0.160 Sps. The effective amount of garnet is related to amount of garnet enriched in Mn after reaction. If the Mn-enriched garnet is volumetrically small (i.e., a rim) then one does not require large absolute change in the modal abundance of garnet. The grossular content of the final garnet is a cross-check on the amount of reaction, because it is not affected by the Mg-Fe exchange. The grossular content is low in the garnet analyzed; thus the amount of change is small. However, the calculated change from 0.053 to 0.047 matches the amount seen in the analyses. 3. The garnet that has been destroyed had the Mg/Fe ratio 0.188/0.812, thus the new biotite must have that ratio. The final biotite composition has lower Al content than the initial one; thus the new biotite must contain less Al than the old. The new biotite is generated by replacing octahedral Al in the muscovite by the Fe and Mg. As there is significantly more biotite than garnet in the final Chapter 5. Metamorphism of the Quesnel Lake area 168 assemblage, it is assumed that the progress of reaction 5.1 alone is what generates the final biotite composition. The best composition of 'new' biotite is: 0.196 Mg; 0.742 Fe; 0.012 Ti; and 0.050 Al. Adding this 'new' biotite to the old biotite (0.449 Mg; 0.343 Fe; 0.025 Ti; 0.183 Al) in the proportions 68:100 produces a biotite with the composition 0.346 Mg; 0.504 Fe; 0.20 Ti; and 0.129 Al. This is very similar to the present matrix biotite composition (a typical analysis is 0.343 Mg; 0.482 Fe; 0.046 Ti; 0.129 Al). The Mg/Mg+Fe ratios are similar. The only discrepancy is in the amount of Ti which depends entirely on the initial biotite composition chosen. Other biotite inclusions have higher Ti contents. 4. A small volume of the remanent garnet re-equilibrates with the large volume of biotite, thereby generating the zonation in Mg/Mg+Fe ratio shown in figure 5.25B. These calculations are first order approximations only. The enrichment of garnet in Mn, the generation of new biotite, homogenization of new and old biotite, and the final re-equilibration of garnet with matrix biotite are treated as discrete and isolated steps. In reality, they are continuous and concurrent processes. However, these calculations produce the final compositions from the initial ones remarkably closely. Resorption zoning appears to be an important feature of some garnet from samples in the area. The data from sample DM-87-24 are relatively simple and support the argument for extensive resorption zoning. Other examples from the area are not so clear. Small amounts of resorption or excessively steeply zoned rims will be difficult to identify. If the matrix biotite is homogeneous, plotting one garnet component ver-sus another will help identify any resorption profiles which would be characterized by variability in the Mn content but constant Mg/(Mg+Fe) ratios. However, using garnet compositions from a zoning profile affects the calculated pressures and temperatures Chapter 5. Metamorphism of the Quesnel Lake area 169 in a particular manner and this may also help identify where resorption has occurred. Resorption, its tectonic causes and the effects on the determination of P and T are discussed in the section on P-T determinations below. 5.3.1.3 Other zoning The remaining samples show a wide variety of classes of profile. It is difficult to interpret many of these profiles because of insufficient data. Three examples are given in figures 5.22 and 5.23. Sample #8 (DM-85-87 profile, fig. 5.22) is the lone example of 'dsdu' zoning. The feature of this profile that is difficult to interpret is the clearly opposite behaviour of Ca and Mn. This may be the result of reaction zoning. The increase in Ca content towards the rims may be the result of the consumption of originally zoned plagioclase or perhaps of epidote or of calcite. A drop in the activity of C 0 2 during prograde metamorphism could result in the breakdown of calcite and the release of Ca for incorporation into other phases. Mn shows the classic growth zoning pattern. The higher Fe content in the rim may be the result of retrograde diffusion with Fe-Mg minerals (although Mg remains fairly constant). If the increase in Fe is related to that of Ca, a reaction consuming epidote may be responsible. The profile shown in sample #10 (DM-86-70 profile, fig. 5.22) is labelled as class 'udus'. The extremely low Mn content and almost flat profile make interpretation difficult. However, Ca and Fe appear to increase inwards where Mg decreases. This may be very subdued growth zoning. Elements present in minute quantities or in abundance should not show detectable growth zoning. In the former case there is no element (Mn) to sequester within cores, and in the latter case (Ca) the garnet continues to grow in an infinite reservoir. The only zoning present would reflect changes in the metamorphic conditions (Mg increasing, Fe decreasing toward the rim). The final profile, sample #29 (DM-87-17 profile, fig. 5.23), shows erratic changes Chapter 5. Metamorphism of the Quesnel Lake area 170 in composition from rim to core. This probably is the result of several generations of growth (reaction zoning). Large garnets may have included smaller, earlier garnets of different compositions. This results in the highly complex zoning documented by A.B. Thompson et al. (1977) and Tracy et al. (1976). Random profiles across such garnet would show the complex and erratic behaviour seen in sample #29. 5.3.2 Biotite and muscovite Within any sample there is some inhomogeneity seen in the micas analyzed. However, individual grains, unless they are very large, do not appear to be zoned. Different grains do seem to vary in composition, particularly biotite. Variation is usually positional, in that micas closer to garnet have slightly different compositions to those further into the matrix. One would not expect to see significant zoning in the micas because grains are generally small and micas have higher diffusion rates and therefore homogenize faster than commonly zoned minerals, such as garnet (Tracy, 1982, p. 388). In addition there are analytical problems that make the identification of inhomogeneity difficult. 1. Micas are subject to electron beam damage and it is necessary to use large beam diameters thus analyzing large portions of mica at any one 'spot'. The larger the 'spot' the more homogeneous the mineral appears. In this study, a 2 /zm spot was used for garnet and a 10 fim spot for micas. 2. Mica analyses are sensitive to the orientation of the grains. Grains mounted on their edge give slightly different results to those lying flat. Therefore, even in a completely homogeneous sample, analyses will appear slightly inhomogeneous. For these reasons the slight variation seen in the mica analyses from most of the samples are not considered significant. Nevertheless care was taken to ensure that the mica Chapter 5. Metamorphism of the Quesnel Lake area 171 compositions used for P-T determination are those that appear to be in equilibrium with adjacent garnet. 5.4 Pressure and temperature determinations 5.4.1 Introduction The determination of pressures and temperatures is a necessary part of any discussion and description of the metamorphism of an area and most recent metamorphic stud-ies present P-T calculations. The method and some of the procedures used here in calculating pressure and temperature are somewhat different from standard practice. In many studies, pressure and temperature are calculated using two different equilib-ria (on different assemblages). For pelites, it has been common practice to calculate T using the garnet-biotite thermometer (first calibrated by Ferry and Spear, 1978) and to calculate pressure using either the GASP (Ghent, 1976) or GRAIL (Bohlen et al., 1983) barometers. The problems that arise from thermodynamic inconsistency between two independently calibrated equilibria have been discussed in chapter 4. However there are also some kinetic problems. Independent assemblages have different kinetic properties and may record or 'freeze' under different conditions. Using the two together may yield P and T that was not experienced by the rock. This study emphasizes P-T calculations using individual analyses, rather than av-erage analyses. Most studies in the past have used average analyses to calculate a single P and T for the sample. Estimates of the 'error' in P and T are propagated from the standard deviations from the 'average' compositions. In this study, individual mineral analyses are used in the calculations to generate a range of pressures and temperatures for each sample. These can be examined for dispersion and for trends and averages can be calculated if required. This approach has some advantages. Chapter 5. Metamorphism of the Quesnel Lake area 172 1. 'Average' compositions are mathematical entities and do not necessarily represent compositions that would be in equilibrium at the P and T at which the rock equilibrated. Calculating P and T for adjacent mineral grains has the potential to show situations of disequilibrium (i.e., outlier P-T points) that may not be evident from the original composition data. 2. There are two common assumptions made in using average analyses. (a) That all the mineral grains have equilibrated at the same P and T. (b) That the variation seen in mineral composition and as expressed in the standard deviation (or variance) can be propagated into an estimate of the error in calculating this P and T (see for example Hodges and McKenna, 1987). There are several objections to these assumptions. (a) However unlikely, it is possible that groups of mineral grains at different places within the sample have recorded different conditions. Using average analyses destroys useful P-T information. (b) Inhomogeneity may be the result of very small domains of bulk composition, not of disequilibrium. That is, mineral pairs in different parts of a sample may have different compositions yet all may have equilibrated at the same P and T. In such a case, the P and T calculated from individual analyses will be the same and have standard deviations of zero (subject to analytical error only). Whereas using average analyses will yield the same P and T, propagating standard deviations in composition into P and T yields a non-existent P-T 'error'. Unless the variation in composition in one phase is truly independent of the variation in another, propagating standard deviations Chapter 5. Metamorphism of the Quesnel Lake area 173 of composition into P and T will always produce larger 'errors' than the standard deviations of P and T calculated for individual analyses. (c) Inhomogeneity may also be the result of disequilibrium. In extreme cases, this is recognized in the original compositions and such analyses are usually discarded when calculating average analyses. It is suggested, however, that P's and T's be calculated using these analyses, not because such P's and T's represent any real metamorphic conditions but because different processes, which produce inhomogeneity, may produce different patterns in apparent P and T. In this way one may be able to identify the process which produced strong inhomogeneity. Secondly, having identified a particular process as the cause of variation in calculated P and T one may be able to make a decision on which P-T value represents 'equilibrium'. Thirdly, in cases where these processes have operated but where inhomogeneity was not recognized the distinctive patterns in P and T (if they exist) would act as indicators that a particular process has operated. Using average analyses would destroy such information. In fact, the zoning described above for sample #30 (DM-87-24) was first identified because the P's and T's calculated for this sample define a conspicuous trend in P-T space. This trend was subsequently identified in several other samples in which inhomogeneity was not so pronounced, indicating that these samples have undergone the same process that affected sample #30. This is discussed further below. In all the scenarios above, using average analyses and standard deviations will destroy useful information and will lead the petrologist to think that the estimates of pressure and temperature are subject to larger errors than is in fact the case. Chapter 5. Metamorphism of the Quesnel Lake area 174 Another practice that is common in thermobarometric studies using the garnet bi-otite thermometer is to assume that core garnet and matrix biotite compositions are those that were at equilibrium at the 'peak' of metamorphism (e.g., Chamberlain, 1986, p. 70, f 2). Such assumptions are often made when rim compositions give metamor-phic conditions that do not match what is known about the metamorphic history of the rocks involved. This may be the case for garnets that have been completely homog-enized at the highest grade and show evidence of retrograde re-equilibration (diffusion zoning only). However, it is very difficult to satisfy the criterion that biotite has not significantly changed composition during retrogression. In this study, P's and T's were determined using core garnet and matrix biotite and muscovite but it was not as-sumed, a priori, that these represent 'peak' conditions. Such calculations were done as an adjunct to rim composition determinations. 5.4.2 Microprobed specimens: Pressures and temperatures Slightly more than 600 pressure-temperature calculations were done on 34 of the 35 mi-croprobed samples. In the remaining sample (#34, DM-87-54), none of the muscovite analyses was usable. It is not feasible to present all these data here — the information presented in table 5.2 summarizes the P-T data garnered from the samples. These data, along with brief discussions of the individual samples in the following pages, are intended to present the salient features of the P-T data without presenting a mass of individual P -T points. In table 5.2 samples are indicated by both the index number, which is used on the map (back pocket), and by the sample number used by the original collector (columns 1 and 2). Columns 3 and 4 give the average P and T calculated from the individual P's and T's determined using individual triplets of garnet, biotite, and muscovite analyses. Columns 5 and 6 give the standard deviation of these P's and T's. Column 7 gives Chapter 5. Metamorphism of the Quesnel Lake area 175 Index # Average Std. dev. Typical (see map) Sample # P (bars) T (°C) P (bars) T (°C) n N P (bars) T (°C) 1 JSG-81-234 6301 413 437 6 19 25 6249 413 2 JSG-81-278 3964 494 289 13 6 7 3967 494 3 JSG-80-30 4245 551 135 12 14 14 4321 554 4 DM-85-9 . 6694 593 740 6 10 14 6605 598 5 PDL-447 6375 621 100 5 3 15 6455 620 6 DM-85-42 6203 603 65 0 4 16 6206 603 7 SLG-370A 5294 597 358 18 7 8 5072 575 8 DM-85-87 5413 539 255 15 10 18 5388 543 9 CJNF-9 5825 617 177 9 3 15 5869 619 5655 556 104 7 12 5704 553 10 DM-86-70 4548 536 351 18 20 28 4632 533 11 DM-86-173 6322 575 380 14 15 22 6207 577 12 DM-86-197 4880 613 245 17 12 12 4684 618 13 DM-86-205 4554 556 160 12 20 20 4482 556 14 DM-86-227 4587 550 215 18 14 15 4639 551 15 DM-86-270 5339 525 384 13 18 20 5133 522 16 DM-86-281 5612 577 414 13 7 23 5910 582 17 DM-86-351 6718 586 671 26 9 10 6740 579 18 DM-86-369 6262 538 342 10 9 9 6391 542 19 DM-86-384 6332 573 160 12 5 19 6306 572 20 JRM-0626-36 8283 565 301 22 10 10 8156 557 21 JRM-0716-76 5996 596 232 19 6 8 5973 604 22 LCP-40 6125 569 241 14 22 25 6272 575 23 LCP-223 5422 552 181 15 20 20 5412 558 24 DM-87-76 5881 614 224 17 21 26 5851 611 25 DM-87-71 - - - - - 27 5541 739 - - - - - 4113 617 26 DM-86-45 5770 641 365 15 12 12 5789 638 27 JAF-7-9-29 - - - - - 23 6658 710 - - - - - 4998 609 28 DM-87-8 - - - - - 30 8671 805 - - - - - 4601 599 29 DM-87-17 - - - - - 21 6998 726 - - - - - 4782 598 30 DM-87-24 - - - - - 19 6157 756 - - - - - 3654 614 31 DM-87-32 6538 619 403 9 14 17 6417 614 32 DM-87-35 7087 582 436 13 17 20 7043 584 33 DM-87-39 4476 631 948 32 22 24 4499 641 34 DM-87-54 no usable muscovite analyses 35 JKR-88-4 - - - - - 15 6811 624 - - - - - 4490 536 Table 5.2: Summary table of pressures and temperatures for microprobed samples. Index # is used on map (back pocket). Sample #is number used by original collector (see appendix D for petrography and references). Average P and T calculated from 'n' P's and T's calculated using 'n' triplets of garnet-biotite-muscovite analyses. N is the total number of triplets used to calculate P and T. See text for discussion of the selection process of the 'n' P-T points from 'N' such points. 'Typical' P and T are actual values close to the average and were determined using the triplets of garnet-biotite-muscovite analyses given in appendix D for each sample. Chapter 5. Metamorphism of the Quesnel Lake area 176 the number, n, of P's and T's that were included to produce the average and column 8 gives the number, N, which is the total number of P-T determinations done for that sample. In most cases n is smaller than N. There are several reasons for this. In many samples, P's and T's were calculated for mica inclusions, or for various combinations of core garnets and matrix micas and these were generally not included when calculating the averages. However, in some cases, P's and T's were highly scattered. The sections were examined to see if valid criteria could be found to justify omitting points from the calculation of the average. Where this was possible, n is considerably lower than N but the standard deviations of P and T are low. Where such criteria could not be identified n is close to N but the standard deviations of P and T are much higher in some cases. This selection process is discussed more fully below for the samples for which it is important. The final two columns of table 5.2 give a typical P and T similar to the average. This is a 'real' P -T value determined using a particular triplet of garnet, biotite, and muscovite analyses from this sample. These analyses are given for each sample in appendix D. To graphically display the relationship between samples the average P-T's are plotted for all samples in figure 5.27. The samples can be split into several groups based on the apparent quality of the P-T determinations. The first group includes the samples with the highest quality, that is, samples with the lowest standard deviations in P and T, particularly in P. Typically for these samples n is also close to N, that is, the averages were calculated with minimum filtering of anomalous points. These group 1 samples are #'s 2, 3, 5, 6, 8, 9, 11, 12, 13, 14, 15, 18, 20, 21, 22, 23, 24, and 26. In addition, the pressures and temperatures determined lie in or close to the fields of stability of the aluminosilicate polymorphs present in these samples. The second group of samples are those that have an extremely wide range in pressures and temperatures but which plot along linear trends in in P-T space. Averages have not been calculated for these samples. These Figure 5.27: Pressure - Temperatures plot showing average P-T values for all samples listed in table 5.2. Numbers are index #'s used on map (in back pocket). Chapter 5. Metamorphism of the Quesnel Lake area 178 are sample #'s 25, 27, 28, 29, 30, and 35. The resorption process described for sample 30 above is the likely cause of these P-T patterns (see below). The third and final group of samples have poor groupings of P-T points. In some cases these P-T points define rough linear trends and may be similar to group 2. Samples in this group are #'s 1, 4, 7, 10, 16, 17, 19, 31, 32, and 33. 5.4.2.1 Group 1 The average pressure and temperature determined for sample #2 (JSG-81-278) from 6 of 7 individual P-T points falls within the kyanite field very close to the aluminsilicate triple point (fig. 5.27). The seventh P-T point was clearly an outlier and was discarded. This value of P-T supports the petrographic observation of all three aluminosilicate polymorphs in this sample (appendix D). Andalusite is the last polymorph formed and it is probable that this rock has followed the classic clockwise metamorphic path around the aluminosilicate triple point. Kyanite would have formed early, followed by sillimanite at the peak of metamorphism. This sample lies just on the sillimanite isograd (map). The andalusite may have formed during the uplift that occurred on the northern side of the fault inferred to pass along the north arm of Quesnel Lake and that has juxtaposed high grade rocks north of the north arm against low grade to the south (map). Thus the P-T point recorded in this rock is not the peak of metamorphism but rather, some point on the uplift and cooling path at which re-equilibration between minerals ceased (see section on P-T trends). The P-T determined for sample #3 (JSG-80-30) lies within the sillimanite field (fig. 5.27) but also at low pressures consistent with the above uplift hypothesis. Note that no 'filtering' was done on the P-T calculations and although the garnet rims are mildly inhomogeneous in this sample (table 5.1) the standard deviation in the P-T's determined is very small. This seems to be an example of equilibrated compositional Chapter 5. Metamorphism of the Quesnel Lake area 179 subdomains within this specimen. The average P-T determined for sample #5 (PDL-447) is based on only 3 of the 15 individual P's and T's determined. These 3 points are for rim garnet compositions and adjacent biotite and matrix muscovite. This is one of the samples in which there is a clear change in the biotite composition away from the rim of the garnet. The other P-T determinations used either matrix biotite with rim garnet or core garnet compositions (away from rim) with rim biotites. In the first ,case, biotites further away from the rim give lower T's and higher P's (for the same garnet and muscovite compositions) and in the second case garnets further into the core yielded lower temperatures and lower pressures (for the same biotite and muscovite compositions). These trends are not interpreted as 'real' P -T trends but as artifacts of using compositions along zoning profiles. The determined P-T lies within the kyanite field (fig. 5.27) in agreement with the observed assemblage. The P-T for sample #6 (DM-85-42) plots within the kyanite field (fig. 5.27) consistent with the observed mineralogy but is based on only 4 individual P-T calcu-lations. The remaining calculations were discarded because core garnet compositions were used with matrix muscovite and biotite. They define a P-T point close to 400 °C and 2000 bars. This is probably not a real estimate of early metamorphic conditions, because it would require that the SGAM assemblage (chapter 4) was present under these conditions. This is unlikely. This example clearly contradicts the assumption of some workers, mentioned above, that core garnet and matrix biotite represent the peak metamorphic assemblage. The P-T for sample #8 (DM-85-87) is based on 10 of 18 individual P-T calcu-lations. The remaining points were done using biotite inclusions within garnet and the adjacent garnet composition and matrix muscovite. The rim P-T conditions are within the kyanite field (fig. 5.27) which is inconsistent with the observed presence of Chapter 5. Metamorphism of the Quesnel Lake area 180 sillimanite. The biotite inclusions yield a slightly lower average T that is not statisti-cally different to the rim T. It is concluded that this sample has re-equilibrated during cooling. Sample #9 (CJNF-9) is the only sample for which two average P-T's are presented. These are for two distinct, tight clusters of individual P's and T's. The high P, high T points are given by garnet with adjacent biotite. The lower P, lower T points are defined by garnet with matrix biotite. Muscovite is homogeneous. The volume of rim biotite is small so that the re-equilibration of garnet with it may not have changed the garnet composition much, nor the muscovite composition. Thus these two P-T points may represent two points on the P-T path followed by this rock. The fact that the rim compositions yield the higher temperatures indicates that this is probably the later set of conditions. That is, these points lie on the prograde part of the P-T loop. These two points fall on opposite sides of the kyanite-sillimanite boundary and the rock contains both sillimanite and kyanite. The high P, high T point is plotted in figure 5.27. Fifteen of 22 individual P-T determinations are used to define the average P-T for sample ffll (DM-86-173). The remaining 7 were done using core garnet compositions, various biotites and matrix muscovite. The determined average P-T falls within the kyanite field (fig. 5.27) in agreement with the mineral assemblage. In this case, the P-T conditions determined using the core garnet compositions are higher than those for the rim and could be interpreted as indicating 'peak' metamorphic conditions, but these core P-T's define a linear trend similar to those in seen in group 2 samples below. These linear trends are the result of using homogeneous biotite with various garnet compositions from a diffusion profile. Sample #'s 12 (DM-86-197), 13 (DM-86-205), and 14 (DM-86-227) all have very similar average P-T's. This is to be expected as these samples come from adjacent localities. No data were discarded for samples 12 and 13 and the single point discarded Chapter 5. Metamorphism of the Quesnel Lake area 181 from sample 14 is a P-T determination using a biotite inclusion. The P-T's of samples 13 and 14 are identical, whereas sample 12 gives a slightly higher P (not statistically significant) and a higher T and all plot within the stability field of sillimanite (fig. 5.27). This agrees with the field and petrographic data. Interestingly, the biotite inclusion discarded from sample 14 gives P-T values of 5213 bars, 634 °C which is slightly further into the sillimanite stability field along the line joining samples 13 and 14, and 12. All these data indicate that although these rocks plot in the appropriate aluminosilicate field they probably have all been retrograded slightly (see section on P-T gradients, below). Sample #'s 15 (DM-86-270) and 18 (DM-86-369) have been included in group 1, but the individual P-T's in both cases define a fairly loose cluster, similar to samples of group 3. The average P-T for sample #15 is defined from 18 of 20 individual P's and T's (two outliers discarded). No data were discarded for sample #18. Both P-T's fall within the kyanite stability field (fig. 5.27), which agrees with the observed mineral assemblages. Sample #'s 20 (JRM-0626-36), 21 (JRM-0716-76), 22 (LCP-40), 23 (LCP-223), 24 (DM-87-76), and 26 (DM-86-45) stretch in a band across the southeastern part of the area and all produce good P-T estimates. In calculating the average P-T's, two individual P-T were discarded from sample #21 (obvious outliers), 3 were discarded from #22 (biotite inclusions used), and 5 were discarded from #24 (biotite inclusions used). All the sample P-T's plot within the appropriate aluminosilicate field except #'s 22 and 23 which plot inside the kyanite field (fig. 5.27). Sample #22 contains only sillimanite and #23 contains both sillimanite and kyanite. The lower temperatures seen in these samples are probably the result of re-equilibration during cooling. The same feature was seen in sample #8 (above), which lies along strike (NW) from these. The variation in the pressures from west to east as seen in these samples is remarkably Chapter 5. Metamorphism of the Quesnel Lake area 182 consistent. Sample #20 gives the highest P observed within the area, but the P-T conditions are consistent with the observed mineral assemblage. Sample #'s 21, 26, 24, and 22 (moving from west to east) yield virtually identical pressures, and sample #23 (easternmost) gives the lowest P in this traverse. This feature and those seen in other group 1 samples are discussed in the final section below. 5.4.2.2 G r o u p 2 The pressures and temperatures calculated for all the samples in this group show a distinctive trend when plotted on a P-T diagram. Two examples, sample #'s 27 (JAF-7-9-29) and 30 (DM-87-24), are shown in figure 5.28. These trends are not considered to represent 'real' processes that occurred in the rock but are the result of using biotite and muscovite compositions for grains adjacent to the garnet and garnet compositions from a zoning profile. As shown for sample #30 (DM-87-24), the variation in garnet composition at the rim is not random but defines what is interpreted as a resorption zoning profile modified by later diffusion. Rim garnet is still considered to be in equilibrium with the muscovite and biotite, but the resorption and diffusion have produced a profile so steep that apparently random points at the rim intersect different portions of this profile. The important conclusion is that the most Mn-rich compositions are the ones closest to the garnet composition in equilibrium with the biotite. Thus, using the most Mn-rich garnet compositions will yield the closest approximations to the P and T at which the rock equilibrated. The most Mn-rich garnets produce the lowest points on the lines shown in figure 5.28. It is these values that are used as estimates of equilibrium pressure and temperature for the samples classified as group 2, and that are plotted in figure 5.27. The pressures and temperatures defined for the 6 group 2 samples (#'s 25, 27, 28, 29, 30, and 35) using the lowest individual values from each specimen are very similar. Chapter 5. Metamorphism of the Quesnel Lake area 183 JAF-7-9-29 DM-87-24 •• • B 1 1 1 600 650 700 750 Temperature (°C) Figure 5.28: Trends in the values of P and T calculated using garnet with resorp-tion zoning modified by diffusion. A: Sample #27 (JAF-7-9-29). B: Sample #30 (DM-87-24). Chapter 5. Metamorphism of the Quesnel Lake area 184 This should be expected as 5 of these (27, 28, 29, 30, 35) come from adjacent localities in the southernmost part of the area. More importantly, the pressures are all remark-ably low and indicate that significant decompression has occurred in this location. The major structure in this area is a doubly plunging F 3 anticline. Because the isograds in this area have been folded the peak of metamorphism pre-dated the development of F 3 . Thus the development of an F 3 anticline in hot (sillimanite grade) rocks might be ex-pected to produce decompression features. Such decompression features would include the resorption of garnet by reaction 5.1 which goes from left to right with decreasing pressure. Decompression would be seen in lower calculated pressures. Decompression could also lead to the development of low pressure phases. Andalusite was seen in 4 samples from this area, 2 of the microprobed specimens (#'s 29 and 30) and 2 others (see symbols on map). The textural evidence clearly indicates that andalusite post-dates the other aluminosilicate assemblages. The only problem is that although the pressures seen in the microprobed specimens are low, the temperatures are still high and the P-T points are within the sillimanite field (fig. 5.27). There are two possible answers to this. Firstly, microprobe analysis requires the use of a finite 'spot' size thus producing an average analysis over the diameter of the 'spot' (actually a volume up to 10 times the diameter of the chosen surface 'spot'). It is therefore likely that the true garnet compositions in equilibrium with the biotite and muscovite in the group 2 samples are even more Mn-rich than the analyses used. This would push the measured pressures and temperatures further down the P-T trends shown in figure 5.28 possibly as far as the andalusite field. For example, the most Mn-rich analysis for garnet in sample #30 contained 16.8 mole % spessartine (fig. 5.25). If the true rim composition had 19.5 mole % spessartine and the appropriate pyrope and almandine contents as defined by the dashed lines in figure 5.25 the determined Chapter 5. Metamorphism of the Quesnel Lake area 185 temperature and pressure would be 572 °C and 2710 bars, right on the sillimanite-andalusite boundary. A second hypothesis is that the thermobarometer equilibrated at the low pressure-high temperatures points indicated above and that further cooling (isobaric?) moved the rocks into the andalusite stability field where crystallization occurred. Such cooling would have to be sufficiently rapid to 'freeze' the thermobarometer but permit the growth of andalusite. The second hypothesis, though possible, is not considered likely because significant growth of andalusite in the absence of further re-equilibration of the garnet, biotite and muscovite seems unreasonable. The first hypothesis is the more feasible. 5.4.2.3 Group 3 The third group of samples shows neither the well clustered P's and T's of group 1 nor the linear trends in P and T of group 2. Instead pressures and temperatures for these samples are scattered and the samples are characterized on table 5.2 by large standard deviations in P and T. Some individual samples in this group seem transitional to group 1 or group 2. Sample #1 (JSG-81-234) is transitional to group 2. The P-T's plot in two clusters, one minor and one major, with a few points plotting on a line drawn between them. This linear trend is not as pronounced as in the true group 2 rocks. In defining the average P -T given on table 5.2, the small cluster and one intermediate point were discarded. The defined P-T plots within the kyanite field, in agreement with the mineral assemblage (fig. 5.27). The relatively high P in comparison to the low P sillimanite grade rocks just to the south (#'s 2 and 3) indicates that amounts of uplift decrease away from the fault running along the north arm of Quesnel Lake (map). The garnets in sample #4 (DM-85-9) are present as coronas around plagioclase, in Chapter 5. Metamorphism of the Quesnel Lake area 186 turn surrounded by biotite. This sample is from a vein and the texture suggests the progress of reaction 5.1 from right to left during an increase in pressure. The pressures and temperatures are dispersed in a somewhat linear trend but with the major cluster at the upper end. Because the reaction 5.1 operated in an opposite sense compared to group 2 samples it is the P-T points at the upper end that probably represent equilibrium. The 4 discarded points fall at the lower end of the trend or away from the line completely. The resulting P-T plots in the kyanite field (fig. 5.27) and is similar to that defined for sample #5, which comes from an adjacent locality. P-T's for sample #7 (SLG-370A) plot in a diffuse cluster. A single point was discarded and the resulting average P-T lies within the sillimanite field, consistent with the observed assemblage (fig. 5.27). The pressure defined is similar to those of nearby sample #'s 8 and 9. P-T points for sample #10 plot in two clusters, one at higher P and lower T than the other. This relationship can be used to define a line of opposite slope to that shown in group 2 samples. The difference between the groups appears to be in the biotite composition. This could result in retrogression of garnet to biotite, by any reaction, without subsequent re-equilibration of garnet with newly generated biotite. Thus the lower T, higher P cluster is chosen. The resulting average P-T is within the kyanite field (fig. 5.27), in agreement with the mineral assemblage. However, the pressure seems somewhat low in comparison to neighbouring samples indicating that even the analyses chosen were not in equilibrium. Twelve of the 23 P-T determinations done for sample #16 (DM-86-281) are from biotite inclusions. The remaining 11 plot in a very diffuse cluster in P-T, with 7 points defining a smaller cluster and 4 satellite points. These 7 points were used in calculating the average P-T presented in table 5.2. This P-T falls within the kyanite field, in agreement with the mineral assemblage (fig. 5.27). The pressure is similar to Chapter 5. Metamorphism of the Quesnel Lake area 187 those calculated for nearby samples. Sample #17 (DM-86-351) is similar to #10 above. The small number of P-T's determined plot in a diffuse line with negative slope. A single point not on this general trend was discarded and the remaining 9 points used to determine the average P and T. This plots within the kyanite field close to the points for sample #'s 4 and 5 which come from nearby localities (fig. 5.27). The 19 P-T's determined for sample #19 (DM-86-384) were determined on several different combinations of rim and core garnet, matrix and adjacent (to garnet) biotite and muscovite. In only 5 cases were the grains reasonably close together, and it is these that were used to determine the average P and T. Though this small cluster has small standard deviations it is classified with group 3 because most of the P-T's plot in a diffuse envelope around this point. The average P-T plots in the kyanite field close to the positions of sample #4, 5, and 18 which come from nearby localities (fig. 5.27). The P-T's from sample #'s 31 (DM-87-32), 32 (DM-87-35), and 33 (DM-87-39) all define diffuse clusters, though the points for sample #33 also show some alignment in a trend similar to those shown by group 2 samples. The P-T's discarded from the average P-T determinations for sample #'s 31 and 32 are clear outliers. The two points discarded from sample #33 are P-T's calculated from inclusions. Even though the points define clusters with high standard deviations the average P-T's fit into the pattern already seen. The sample from the sillimanite grade rocks in the core of the F 3 anticline, #33, gives low pressures similar to those of the group 2 samples from this area. Sample #'s 31 and 32 which come from the kyanite grade rocks in the adjacent syncline give considerably higher pressures. The average P-T's plot in the field of stability of the appropriate aluminosilicate polymorph (fig. 5.27). Chapter 5. Metamorphism of the Quesnel Lake area 188 5.5 Pressure — temperature trends and tectonic implication The areal distribution of the average pressures and temperatures given in table 5.2 are shown in figures 5.29 and 5.30 respectively. There are three different domains of P-T behaviour. These domains are: • Domain 1: North of the fault running along the north arm of Quesnel Lake (figs. 5.29 and 5.30). • Domain 2: The doubly plunging anticline in the extreme south of the area, called the Boss Mountain anticline by Fillipone (1985), (figs. 5.29 and 5.30). • Domain 3: The remaining central part of the area (figs. 5.29 and 5.30). To show trends in domain 3 more clearly, pressures and temperatures are projected onto line A-A' shown in figures 5.29 and 5.30 and are shown in figures 5.31 and 5.32 respectively. This method is preferred over contouring the data because the standard deviations (as tabulated in table 5.2) can then be shown. Pressures show the following trends: 1. In domain 1, pressures increase northwards from the proposed fault running along the north arm of Quesnel Lake (fig. 5.29). 2. In domain 2, pressures decrease outwards from the core of the Boss Mountain anticline (fig. 5.29). 3. In domain 3, pressures show a general decrease from west to east (fig. 5.29). More detail is revealed when domain 3 points are projected onto line A-A'. Pres-sures initially decrease from west to east then increase slightly and appear to be decreasing again at the easternmost edge. Chapter 5. Metamorphism of the Quesnel Lake area 189 Trends in temperature appear to be more subtle. 1. In domain 1, temperatures decrease northwards (fig. 5.30), in agreement with the observed change in grade (map, in back pocket). 2. In domain 2, temperatures show a slight decrease outwards from the core of the Boss Mountain anticline (fig. 5.30), also in agreement with the observed change in grade (map). 3. In domain 3, the thermal structure is more subdued. T's show a slight domal structure (fig. 5.32, the crest of which coincides with the center of the sillimanite zone. Thus, temperature structure coincides with the grade, as defined by the mineral assemblage. Thus the rocks that have recorded the highest pressures yield the lowest temperatures. This type of field gradient is not the simple gradient which one might expect, showing increasing pressure with increasing temperature coincident with increasing mineralogic grade. The P-T field gradient defined by the thermobarometric determinations cannot be explained by assuming that the conditions recorded in each sample occurred at the same time (i.e., 'peak' of metamorphism). It is necessary to examine what sort of P-T gradients are possible and what processes lead to their formation and preservation. There are five possible types of P-T gradient for any particular area. These are: • Type 1: No variation in either pressure or temperature (a point on a P-T dia-gram). • Type 2: No variation in pressure but variation in temperature (a horizontal line in P-T space). • Type 3: Rocks with higher recorded pressures have higher temperatures (a line of positive slope in P-T space). Chapter 5. Metamorphism of the Quesnel Lake area 190 Figure 5.29: Areal distribution of the average pressures listed in table 5.2. Dashed line is the contact between unit 1 and unit 2, for reference. Domains 1, 2, and 3 show differing trends. Values of P from domain 3, projected orthogonally onto line A-A', are shown in figure 5.31. Chapter 5. Metamorphism of the Quesnel Lake area 191 Figure 5.30: Areal distribution of the average temperatures listed in table 5.2. Dashed line is the contact between units 1 and 2. Domains 1, 2, and 3, show differing trends in T . Values of T in domain 3 are projected orthogonally onto line A - A ' and shown in figure 5.32. Chapter 5. Metamorphism of the Quesnel Lake area 192 8.0-Figure 5.31: Values of pressure from domain 3, as projected orthogonally onto line A-A' shown in figure 5.29. Dots represent localities north of line A-A', squares represent localities to the south of it. Vertical bars show ± 1 standard deviation as given in table 5.2. Chapter 5. Metamorphism of the Quesnel Lake area 193 650-<D 600-CO cu a E 550-500-i t { A ' Figure 5.32: Values of temperature from domain 3, as projected orthogonally onto line A-A' shown in figure 5.30. Dots represent localities north of line A-A', squares represent localities to the south of it. Vertical bars show ± 1 standard deviation as given in table 5.2. Chapter 5. Metamorphism of the Quesnel Lake area 194 • Type 4: No variation in temperature but variation in pressure (a vertical line in P-T space). • Type 5: Rocks with higher recorded pressures have lower temperatures (a line of negative slope in P-T space). These five types of field gradients, of which the fifth is the one seen in the Quesnel Lake area, result from difference in rates of deformation and uplift with respect to the rate of thermal re-equilibration. The following is a simple, but effective, model to explain the five types of P-T gradients listed above. Figure 5.33 shows the P-T paths taken by two arbitrarily chosen rocks, A and B. At time to A and B He at different pressures (depths) and at different temperatures on some arbitrary stable P-T gradient. At time ti, folding of the rocks places rock A in the core of a syncline and rock B in the core of an anticline at the same level within the deformed package. Because the pressure imposed on a rock changes virtually instantaneously whereas temperature changes gradually the P-T paths taken are shown as vertical in figure 5.33. Uniform uplift starts at t2 some time after tj and, at any time tt, A and B are at the same level in the crust and have the temperatures shown on figure 5.33. If the time between folding (ti) and the initiation of uplift (t2) is long, relaxation of the folded isotherms will occur and rocks A and B will move to C and then down a single uplift path. In general, A and B will follow the separate paths shown. The five types of field gradient are produced in the following ways. 1. If the time between ti and t2 is long enough to permit relaxation of the folded isotherms rocks A and B migrate back to the original isotherm at position C and thence down along the single cooling path, producing no field P-T gradient (type !)• Chapter 5. Metamorphism of the Quesnel Lake area 195 B *0 C i t o r f t 4 i / A / A A A A Temperature Figure 5.33: Hypothetical P-T-t paths taken by two samples that undergo folding followed by uplift to the exposed surface. At time t0 A starts at a lower P and T than B. Folding occurs at ti that places A in the core of a syncline and B in the core of an anticline. Relaxation of the folded isotherms followed by uplift would bring both toward C and down along a single cooling path (arrow). Uplift without relaxation drives A and B along separate paths as shown. See text for discussion. Chapter 5. Metamorphism of the Quesnel Lake area 196 2. If the length of time between ti and t2 is short and the rate of uplift rapid the conditions at time ti will be recorded. Rocks in the cores of anticlines will yield higher temperatures than those in the cores of synclines but there will be no variation in the recorded pressure (type 2). 3. If the rate of folding and uplift is fast, the conditions at t 0 will be preserved. Rocks in the cores of anticlines will show higher temperatures and pressures than the rocks in the cores of adjacent synclines (type 3). In these three cases, note that the P-T conditions recorded in rocks A and B were experienced at the same time. In the remaining cases that is not true. Diachronous resetting must occur. The resetting of thermobarometers, as with all reactions, is a function of both temperature and cooling rate. The hotter the rock the more likely it is to be reset during cooling. However, the faster a rock is cooled the less likely it is to be reset. If two rocks at the same crustal level have different temperatures and are uplifted to the surface at the same rate, the hotter rock will have to cool faster than the cooler. Thus temperature and uplift rates will work against one another. 4. If the time between ti (folding) and t2 (initiation of uplift) is short rocks A and B follow the separate cooling paths shown in figure 5.33. Along these paths, the rate of cooling of B is always faster than that of A. If cooling is very slow, then resetting will be a function of temperature. Thus rock A may 'set' at time t4 whereas rock B will 'set' at time t6, producing a field P-T gradient in pressure but not in temperature. Rocks in the cores of anticlines will yield lower pressures than those in the cores of adjacent synclines, but without any temperature difference (type 4 gradient). Chapter 5. Metamorphism of the Quesnel Lake area 197 5. In general, both temperature and rate of cooling work to control the setting temperature. If the path taken is rate-of-cooling dominated then rock B will tend to 'set' before rock A. For example rock B may set at t3, rock A at t,j. The field gradient produced has a positive gradient in both pressure and temperature. Rocks in the cores of anticlines have higher temperatures and pressures than those in the cores of adjacent synclines (type 3). This is similar to the gradient produced by extremely rapid deformation and uplift (type 3). However, in this case the isotherms and isobars have 'relaxed' somewhat and show more subdued fold amplitudes than the layering in the rocks (providing that layering, isotherms, and isobars were originally parallel). Finally, if the cooling path is somewhat less rate-of-cooling dominated, rock A will 'set' before rock B. For example, rock A may set at t4 and rock B at t5. Thus the field gradient will be one of increasing pressure with decreasing temperature and rocks in the cores of anticlines will show lower pressures and higher temperatures than the rocks in an adjacent syncline (type 5). As seen in figures 5.29, 5.30, 5.31 and 5.32 the rocks from the Quesnel Lake area show a type 5 field gradient. It must therefore be interpreted from the P-T data that the highest grade rocks (those that were initially at the greatest depth; rock B in fig. 5.33) should be found in the cores of antiforms and these antiforms must have developed after the peak of metamorphism. This matches exactly what is seen in the outcrop pattern as shown on the map (back pocket). It also is in agreement with the observed textures in thin section that indicate the peak of metamorphism predated F 3 but that some mineral growth continued afterward, particularly in the rocks that attained the highest grades. Thus, although the peak of metamorphism is associated with F2, the distribution of the isograds as mapped from observed assemblages and Chapter 5. Metamorphism of the Quesnel Lake area 198 the distribution of fossil isotherms and isobars are a function of the postmetamorphic folding (F 3 ) . Postmetamorphic folding also explains the distribution of the isograds in the area. Previous workers have noted that the isograds are folded and have concluded simply that the folding postdated the peak of metamorphism. At the same time workers have had considerable trouble explaining the close spacing of isograds in parts of the area. In fact, because the isograds have been folded there must be locations where the spacing between the isotherms becomes narrower, in the limbs of the postmetamorphic folds. Though isograds are imaginary surfaces, they must behave in the same way as real surfaces such as bedding or lithologic layering do. Thus, the isograds should be narrowly spaced in the limbs of F 3 folds and should diverge in the hinges, just as is seen in the rocks of the Quesnel Lake area. 5.6 Summary and conclusions The field, petrographic, and thermobarometric data indicate the following metamorphic history for the area. 1. Metamorphic mineral growth during the first phase of deformation (Fj) was prob-ably confined to the micas. Garnet may have grown late in F i . 2. The major mineral growth occurred during F 2 deformation. The F 2 foliation wraps around the earlier porphyroblasts (garnet), but later porphyroblasts (kyan-ite and staurolite) commonly overgrow it. 3. Metamorphism waned during F 3 and, except in the high grade cores of anticlines, metamorphic recrystallization ceased. Chapter 5. Metamorphism of the Quesnel Lake area 199 Finally, there are several important points to be made regarding the utility of the the SGAM thermobarometer and the techniques used. 1. As stated in chapter 4, the SGAM thermobarometer produces pressures and tem-peratures that are 'real' and that show gradients that are consistent with field and petrographic data. 2. The precision of the technique is in part a product of calculating P's and T's from individual groups of minerals instead of from average analyses. 3. Zoned and/or highly inhomogeneous minerals can yield as much information on the P-T history of a rock as do homogeneous minerals — perhaps more. 4. 'Reset' P-T values can be as informative as 'peak' metamorphic conditions and rocks suspected of having such conditions recorded in them need not be discarded. 5. The agreement between field, petrographic, and thermobarometric data indicate that, in this instance, SGAM has recorded only the latest part of the P-T history. Earlier P-T information may not be accessible to this, or any, thermobarometer. 200 References Bailey, D .G. (1988): Geology of the central Quesnel belt, Hydraulic, south-central British Columbia, in Geologic Fieldwork 1987, British Columbia Ministry of Energy, Mines and Petroleum Resources, Paper 1988-1, p. 147-153. Berman, R . G . (1988): Internally consistent thermodynamic data for minerals in the system Na20-K20-CaO-MgO-FeO-Fe203-Al203-Si02-Ti02-H20-C02. Journal of Petrology, v. 29, p. 445-522. (1990): Mixing properties of Ca-Mg-Fe-Mn garnets. American Mineralogist, v. 75, p. 328-344. , and Brown, T . H . (1984): A thermodynamic model for multicomponent melts, with application to the system CaO-Al203-Si02- Geochimica et Cosmochimica Acta, v. 48, p. 661-678. , , and Perkins, E . H . (1987): GeO-Calc: Software for calculation and display of pressure-temperature-composition phase diagrams. American Mineralogist, v. 72, p. 861-862. , Engi, M. , Greenwood, H .J . , and Brown, T . H . (1986): Derivation of internally consistent thermodynamic data by the technique of mathematical programming: a review with application to the system MgO — SiC*2 — H 20. Journal of Petrology, v. 27, p. 1331-1364. Berthe, D., Choukroune, P., and Jegouzo, P. (1979): Orthogneiss, mylonite and non-coaxial deformation of granites: the example of the South Armorican Shear Zone. Journal of Structural Geology, v. 1, p. 31-42. Blenkinsop, J . (1972): Computer assisted mass spectrometry and its application to rubid-ium-strontium geochronology. Ph.D. thesis, University of British Columbia, Vancouver, 109p. Bloodgood, M .A . (1987a): Deformational history, stratigraphic correlations and geochem-istry of eastern Quesnel Terrane rocks in the Crooked Lake area, east central British Columbia, Canada. M.Sc. thesis, University of British Columbia, Vancouver, 165p. (1987b): Geology of the Triassic Black Phyllite in the Eureka Peak area, Central British Columbia, In Geological Fieldwork 1986, British Columbia Ministry of Energy, References 201 Mines and Petroleum Resources, Paper 1987-1, p. 135-142. (1988): Geology of the Quesnel terrane in the Spanish Lake area, central British Columbia, In Geological Fieldwork 1987, British Columbia Ministry of Energy, Mines and Petroleum Resources, Paper 1988-1, p. 139-145. Bohlen, S.R., Wall, V . J . , and Boettcher, A . L . (1983): Experimental investigations and geological application of equilibria in the system FeO-Ti02-Al203-Si02-H20. American Mineralogist, v. 68, p. 1049-1058. Brown, R.L. , and Rees, C . J . (1981): Tectonic history of the Shuswap terrane of the southern Canadian Cordillera (abstr.). Geological Association of Canada - Mineralogical Association of Canada - Canadian Geophysical Union. Program with abstracts, v. 6, p. A7. Brown, T . H . (1977): Non-ideal and complex solutions. In Greenwood, H.J. (ed.), Short course in Application of thermodynamics to petrology and ore deposits, Mineralogical Association of Canada, p. 126-135. , and Greenwood, H.J . (in prep): A re-examination of ideal solid solutions. Canadian Mineralogist, v. 29 (Greenwood Volume). Campbell, K . V . (1971): Metamorphic petrology and structural geology of the Crooked Lake area, Cariboo Mountains, British Columbia. Ph.D. thesis, University of Washington, Seattle, 192p. and Campbell R.B. (1969): Quesnel Lake map-area, British Columbia, In Report of Activities, Part A, April to October 1969, Geological Survey of Canada Paper 70-1, p. 32-35. Campbell R .B . (1961): Quesnel Lake west half, British Columbia. Geological Survey of Canada, Map 3-1961. (1963): Quesnel Lake east half, British Columbia. Geological Survey of Canada, Map 1-1963. (1968): Canoe River, British Columbia. Geological Survey of Canada, Map 15-1967. (1970): Structural and metamorphic transitions from infrastructure to superstruc-ture, Cariboo Mountians, British Columbia, In Wheeler, J.O. (ed.), Structure of the Cana-dian Cordillera, Geological Association of Canada, Special Paper 6, p. 67-72. References 202 (1973): Structural cross-section and tectonic model of the southeastern Canadian Cordillera. Canadian Journal of Earth Sciences, v. 10, p. 1607-1620. (1978): Quesnel Lake (93A) map-area, British Columbia. Geological Survey of Canada, Open File 574 (map). , Mountjoy, E.W., and Struik, L .C . (1982): Structural cross-section through south-central Rocky and Cariboo Mountains to the Coast Range. Geological Survey of Canada, Open File 844. , Mountjoy, E.W., and Young, F . G . (1973): Geology of McBride map-area, British Columbia. Geological Survey of Canada, Paper 72-35. and Tipper, H.W. (1971): Geology of Bonaparte Lake map-area, British Columbia. Geological Survey of Canada, Memoir 363. Carye, J . A . (1986): Structural geology of part of the Crooked Lake area, Quesnel Highlands, British Columbia. M.Sc. thesis, University of British Columbia, Vancouver, 185p. Chamberlain, C P . (1986): Evidence for repeated folding of isotherms during regional meta-morphism. Journal of Petrology, v. 27, p. 63-89. Chappell, B.W. , and White, A . J . R . (1974): Two contrasting granite types. Pacific Geology, v. 8, p. 173-174. Chatterjee, N.D. , and Froese, E. (1975): A thermodynamic study of the pseudobinary join muscovite-paragonite in the system K A l S i 3 0 8 - N a A l S i 3 0 8 - A l 2 0 3 - S i 0 2 - H 2 0 . American Mineralogist, v. 60, p. 985-993. Chipera, S .J . , and Perkins, D. (1988): Evaluation of biotite-garnet geothermometers: Application to the English River subprovince, Ontario. Contributions to Mineralogy and Petrology, v. 98, p. 40-48. Coleman, R . G . (1977): Ophiolites: Ancient Oceanic Lithosphere? Springer-Verlag, Berlin, 229p. Coney, P .J . , Jones, D.L., and Monger, J . W . H . (1980): Cordilleran suspect terranes. Nature, v. 288, p. 329-333. Day, H.W., and Halbach, H. (1979): The stability field of anthophyllite: The effect of ex-perimental uncertainty on permissible phase diagram topologies. American Mineralogist, References 203 v. 64, p. 809-823. Day, H.W., and Kumin, H.J. (1980): Thermodynamic analysis of the aluminosilicate triple point. American Journal of Science, v. 280, p. 265-287. Elsby, D .C . (1985): Structure and deformation across the Quesnellia - Omineca terrane boundary, Mount Perseus area, east-central British Columbia. M.Sc. thesis, University of British Columbia, Vancouver, 178p. Engi, J . E . (1984): Structure and metamorphism north of Quesnel Lake and east of Nia-gara Creek, Cariboo Mountains, British Columbia. M.Sc. thesis, University of British Columbia, Vancouver, 137p. Engi, M . (1983): Equilibria involving Al-Cr spinel; I. Mg-Fe exchange with olivine; experi-ments, thermodynamic analysis and consequences for geothermometry. In Orville Volume; Studies in metamorphism and metasomatism. Greenwood H.J. (ed.). American Journal of Science, v. 283A, p. 29-71. Ferry, J . M . (1980): A comparative study of geo'thermometers and geobarometers in pelitic schists, south-central Maine. American Mineralogist, v. 65, p. 720-732. , and Spear, F.S. (1978): Experimental calibration of the partitioning of Fe and Mg between biotite and garnet. Contributions to Mineralogy and Petrology, v. 66, p. 113-117. Fillipone, J . A . (1985): Structure and metamorphism at the Intermontane - Omineca bound-ary, near Boss Mountain, east central British Columbia. M.Sc. thesis, University of British Columbia, Vancouver, 149p. , and Ross, J . V . (1990): Deformation of the western margin of the Omineca Belt near Quesnel Lake, east-central British Columbia. Canadian Journal of Earth Sciences, v. 27, p. 414-425. Fletcher, C.J .N . (1972): Metamorphism and structure of Penfold Creek area, near Quesnel Lake, British Columbia. Ph.D. thesis, University of British Columbia, Vancouver, 123p. , and Greenwood, H.J. (1979): Metamorphism and structure in the Penfold Creek area near Quesnel Lake, British Columbia. Journal of Petrology, v. 20, p. 743-794. Fuhrman, M.L . , and Lindsley, D.H. (1988): Ternary-feldspar modelling and thermometry. American Mineralogist, v. 73, p. 201-216. .References 204 Ganguly, J., and Saxena, S.K. (1984): Mixing properties of aluminosilicate garnets: Con-straints from natural and experimental data and applications to geothermo-barometry. American Mineralogist, v. 69, p. 88-97. , and (1985): Mixing properties of aluminosilicate garnets: Constraints from natural and experimental data and applications to geothermo-barometry: Clarifica-tions. American Mineralogist, v. 70, p. 1320. Garwin, S.L. (1987): Structure and metamorphism in the Niagara Peak area, western Cari-boo Mountains, British Columbia. M.Sc. thesis, University of British Columbia, Vancou-ver, 177p. Gerasimoff, M . (1988): The Hobson Lake Pluton, Cariboo Mountains, British Columbia, and its significance to Mesozoic and Early Cenozoic Cordilleran tectonics. M.Sc. thesis, Queen's University, Kingston, Ontario, 196p. Getsinger, J .S . (1982): Metamorphism and structure of Three Ladies Mountain area, Cari-boo Mountains, British Columbia, In Current Research, Part A , Geological Survey of Canada, Paper 82-1A, p. 317-320. (1985): Geology of the Three Ladies Mountain / Mount Stevenson area, Quesnel Highland, British Columbia. Ph.D. thesis, University of British Columbia, Vancouver, 239p. Ghent, E .D. (1976): Plagioclase - garnet - A^SiOs - quartz: A potential geobarometer-geothermometer. American Mineralogist, v. 61, p. 710-714. , and Stout, M.Z. (1981): Geobarometry and geothermometry of plagioclase -biotite - garnet - muscovite assemblages. Contributions to Mineralogy and Petrology, v.76, p. 92-97. Goldman, D.S., and Albee, A . L . (1977): Correlation of Mg/Fe partitioning between garnet and biotite with 0 1 8 / 0 1 6 partitioning between quartz and magnetite. American Journal of Science, v. 277, p. 750-767. Gordon, T . M . (1973): Determination of internally consistent thermodynamic data from phase equilibrium experiments. Journal of Geology, v. 81, p. 199-208. (1977): Derivation of internally consistent thermochemical data from phase equilib-rium experiments using linear programming. In Greenwood, H . J . (ed.), Short course in Application of thermodynamics to petrology and ore deposits, Mineralogical Association References 205 of Canada, p. 185-198. Greenwood, H.J. (1967): The N-dimensional tie-line problem. Geochimica et Cosmochim-ica Acta, v. 31, p. 465-490. (1968): Matrix methods and the phase rule in petrology. XXII International Geo-logical Congress, v. 6, p. 267-279. (1977): Entropy, randomness, and activity. In Greenwood, H.J. (ed.), Short course in Application of thermodynamics to petrology and ore deposits, Mineralogical Association of Canada, p. 38-46. Guidotti, C . V . (1984): Micas in metamorphic rocks. In Bailey, S.W. (ed.), Micas, Washing-ton, D.C, Mineralogical Society of America, Reviews in Mineralogy, v. 13, p. 357-469. Halbach, H. , and Chatterjee, N.D. (1982): The use of linear programming for deter-mining internally consistent thermodynamic data for minerals. In Schreyer, W. (ed.), High-Pressure Researches in Geoscience, E. Schweitzerbart'sche Verlagsbuchhandlung, Stuttgart, 1982, p. 475-491. Helgeson, H .C . , Delaney, J . M . , Nesbitt, H.W., and Bird, D.K. (1978): Summary and critique of the thermodynamic properties of rock-forming minerals. American Journal of Science, v. 278A, 229p. Hewitt, D.A. , and Wones, D.R. (1984): Experimental phase relations of the micas. In Bailey, S.W. (ed.), Micas, Washington, D.C, Mineralogical Society of America, Reviews in Mineralogy, v. 13, p. 201-256. Hodges, K .V . , and Crowley, P.D. (1985): Error estimation and empirical geothermo-barometry for pelitic systems. American Mineralogist, v. 70, p.702-709. Hodges, K .V . , and McKenna L .W. (1987): Realistic propagation of uncertainties in geologic thermobarometry. American Mineralogist, v. 72, p. 671-680. Hodges, K .V . , and Spear, F.S. (1982): Geothermometry, geobarometry and the Al2SiOs triple point at Mt. Moosilauke, New Hampshire. American Mineralogist, v. 67, p. 1118— 1134. 1 Hoinkes, G. (1986): Effect of grossular content in garnet on the partitioning of Fe and Mg between garnet and biotite. Contributions to Mineralogy and Petrology, v. 92, p. 393-399. References 206 Holdaway, M . J . , Dutrow, B.L. , and Hinton, R.W. (1988): Devonian and Carboniferous metamorphism in west-central Maine: The muscovite-almandine geobarometer and the staurolite problem revisited. American Mineralogist, v. 73, p. 20-47. Holland, S.S. (1954): Yanks Peak - Roundtop Mountain area, British Columbia. British Columbia Department of Mines, Bulletin 34. Indares, A . , and Martignole, J . (1985): Biotite-garnet geothermometry in the granulite facies: The influence of Ti and Al in biotite. American Mineralogist, v. 70, p. 272-278. Jones, D.L. , Howell, D .G. , Coney, P.J. , and Monger, J . W . H . (1983): Recognition, character, and analysis of tectonostratigraphic terranes in western North America. In Accretion Tectonics in the Circum-Pacific Regions; Proceedings of the Oji International Seminar on Accretion Tectonics, Hashimoto, M. and Uyeda, S. (eds.), Terra Scientific Publishing Company, Tokyo, p. 21-35. Klepacki, D.W. (1981): The Little River Fault — new low-angle boundary fault of the northern Shuswap Complex, Quesnel Lake, British Columbia (abstract). Geological Asso-ciation of Canada - Mineralogical Association of Canada - Canadian Geophysical Union Joint Annual Meeting, Program with Abstracts, v. 6, p. A-32. , Read, P.B., and Wheeler, J . O . (1985): Geology of the headwaters of Wilson Creek, Lardeau map area, southeastern British Columbia, In Current Research, Part A, Geological Survey of Canada, Paper 85-1A, p. 273-276. and Wheeler, J . O . (1985): Stratigraphic and structural relations of the Milford, Kaslo and Slocan groups, Goat Range, Lardeau and Nelson map areas, British Columbia, In Current Research, Part A, Geological Survey of Canada, Paper 85-1A, p. 277-286. Kretz, R. (1983): Symbols for rock-forming minerals. American Mineralogist, v. 68, p. 277-279. LeBreton, N. , and Thompson, A . B . (1988): Fluid-absent (dehydration) melting of biotite in metapelites in the early stages of crustal anatexis. Contributions to Mineralogy and Petrology, v. 99, p. 226-237. Lewis, P.D. (1987): Polyphase deformation and metamorphism in the western Cariboo Mountains near Ogden Peak, British Columbia. M.Sc. thesis, University of British Columbia, Vancouver, 146p. May, S.R., and Butler, R.F. (1986): North American Jurassic apparent polar wander: References 207 Implications for plate motion paleogeography and Cordilleran tectonics. Journal of Geo-physical Research, v. 91, p. 11519-11544. McMull in, D .W.A. , and Berman, R .G . (1990): Calibration of SGAM barometry for pelitic schists using data from phase equilibrium experiments and natural assemblages. Geological Association of Canada - Mineralogical Association of Canada, Program with Abstracts, v. 15, p. A88. McMull in, D .W.A. , and Greenwood, H.J . (1986): Metamorphic pressures and temper-atures in the Barkerville and Cariboo terranes, Quesnel Lake, British Columbia: Prelimi-nary results. In Current Research, Part B, Geological Survey of Canada, Paper 86-1B, p. 727-732. , and (1987): PT-GRADS: Simultaneous P and T calculations with GeO-Calc's PTSYSTEM. Geological Association of Canada - Mineralogical Association of Canada, Program with Abstracts, v. 12, p. 73. , and (1988a): Metamorphism in and near the norhtern end of the Shuswap Metamorphic Complex, south-central British Columbia, In Current Research, Part E, Geological Survey of Canada, Paper 88-1E, p. 57-64. , and (1988b): Coupled substitution, charge balance and the effects on the activities of the species in an ideal solution. Geological Association of Canada -Mineralogical Association of Canada - Canadian Society of Petroleum Geologists, Program with Abstracts, v. 13, p. A83. Monger, J . W . H . (1984): Cordilleran tectonics: a Canadian perspective. Bulletin de la Societe geologique de France, v. 26(2), p. 255-278. , Price, R .A. , and Tempelman-Kluit, D.J . (1982): Tectonic accretion and the origin of the two major metamorphic and plutonic welts in the Canadian Cordillera. Geology, v. 10, p. 70-75. Montgomery, J .R . (1985): Structural relations of the southern Quesnel Lake gneiss, Isosceles Mountain area, southwest Cariboo Mountains, British Columbia. M.Sc. thesis, University of British Columbia, Vancouver, 96p. , and Ross, J . V . (1989): A note on the Quesnel Lake Gneiss, Caribou (sic) Mountains, British Columbia. Canadian Journal of Earth Sciences, v. 26, p. 1503-1508. Montgomery, S.L. (1978): Structural and metamorphic history of the Dunford Lake map References 208 area, Cariboo Mountains, British Columbia. M.S. thesis, Cornell University, Ithaca, New York, 170p. Mortenson, J . K . , Montgomery, J .R. , and Fillipone, J . A . (1987): U-Pb zircon, mon-azite and sphene ages for granite orthogneiss of the Barkerville terrane, east-central British Columbia. Canadian Journal of Earth Sciences, v. 24, p. 1261-1266. Murphy, D .C. (1987): Kaza Group, eastern Wells Gray Park, British Columbia, In Current Research, Part A, Geological Survey of Canada, Paper 87-1A, p. 735-742. Newton, R .C . , and Hazelton, H.T. (1981): Thermodynamics of the garnet-plagioclase-A^SiOs-quartz geobarometer. In Newton, R.C, Navrotsky, A., and Wood, B.J. (eds.), Thermodynamics of Minerals and Melts. New York, Springer-Verlag, p. 131-147. Nordstrom, D.K. , and Munoz, J . L . (1986): Geochemical Thermodynamics. Blackwell Scientific Publications, Palo Alto, California, 477p. North American Commission on Stratigraphic Nomenclature (1983): North Amer-ican Stratigraphic Code. American Association of Petroleum Geologists Bulletin, v. 67, p. 841-875. Okulitch, A . V . (1985): Paleozoic plutonism in southeastern British Columbia. Canadian Journal of Earth Sciences, v. 22, p. 1409-1424. Panteleyev, A . (1987): Quesnel gold belt - alkalic volcanic terrane between Horsefly and Quesnel Lakes. In Geological Fieldwork, British Columbia Ministry of Energy, Mines, and Petroleum Resources, Paper 1987-1, p. 125-133. Pell, J . (1984): Stratigraphy, structure and metamorphism of Hadrynian strata in south-east Cariboo Mountains, British Columbia. Ph.D. thesis, University of Calgary, Calgary, Alberta, 185p. Perchuk, L .L . , and Lavrent'eva, I .V . (1983): Experimental investigation of exchange equilibria in the system cordierite-garnet-biotite. In Saxena, S.K. (ed.), Kinetics and Equilibrium in Mineral Reactions. New York, Springer-Verlag, Advances in Physical Geochemistry, v. 3, p. 199-239. Pigage, L . C . (1977): Rb-Sr dates for granodiorite intrusions on the northeastern margin of the Shuswap Metamorphic Complex, Cariboo Mountains, British Columbia. Canadian Journal of Earth Sciences, v. 14, p. 1690-1695. References 2 0 9 (1978): Metamorphism and deformation on the northeastern margin of the Shuswap Metamorphic Complex. Ph.D. thesis, University of British Columbia, Vancouver, 289p. (1982): Linear regression analysis of sillimanite forming reactions at Azure Lake, British Columbia. Canadian Mineralogist, v. 20, p. 349-378. , and Greenwood, H .J . (1982): Internally consistent estimates of pressure and temperature: the staurolite problem. American Journal of Science, v. 282, p. 943-968. Radloff, J . K . (1989): Origin and obduction of the ophiolitic Redfern Complex on the Omineca - Intermontane Belt boundary, western Cariboo Mountains, British Columbia. M.Sc. thesis, University of British Columbia, Vancouver, 179p. Raeside, R.P., Hil l , J .D . , and Eddy, B . G . (1988): Metamorphism of Meguma group metasedimentary rocks, Whitehead Harbour area, Guysborough County, Nova Scotia. Maritime Sediments and Atlantic Geology, v. 20, p. 1-9. Rees, C . J . (1981): Western margin of the Omineca Belt at Quesnel Lake, British Columbia, In Current Research, Part A, Geological Survey of Canada, Paper 81-1A, p. 223-226. (1987): The Intermontane - Omineca Belt Boundary in the Quesnel Lake area, east-central British Columbia: tectonic implications based on geology, structure and pa-leomagnetism. Ph.D. thesis, Carleton University, Ottawa, 421p. , and Ferri, F. (1983): A kinematic study of mylonitic rocks in the Omineca - In-termontane Belt tectonic boundary in east-central British Columbia, In Current Research, Part B, Geological Survey of Canada, Paper 83-1B, p. 121-125. Robinson, G.R. , Jr . (1983): Calibration of the muscovite-biotite-quartz-garnet-alumino-silicate geothermobarometer. EOS, v. 64 p. 351. Ross, J . V . , Fillipone, J . A . , Montgomery, J .R. , Elsby, D . C , and Bloodgood, M . A . (1985): Geometry of a convergent zone, central British Columbia, Canada. Tectono-physics, v. 119, p. 285-297. Ross, J . V . , Garwin, S.L., and Lewis, P.D. (1989): Geology of the Quesnel Lake Region, central British Columbia: Geometry and implications, In Proceedings, 7th International Conference on Basement Tectonics: Dorecht, The Netherlands, D. Reidel Publishing Co., p. 1-23. Simpson, C , and Schmid, S .M. (1983): An evaluation of criteria to deduce the sense of References 210 movement in sheared rocks. Geological Society of America Bulletin, v. 94, p. 1281-1288. Struik, L . C . (1979): Stratigraphy and structure of the Barkerville - Cariboo river area, central British Columbia, In Current Research, Part B, Geological Survey of Canada, Paper 79-1B, p. 33-38. — (1981a): A re-examination of the type area of Devono-Mississippian Cariboo Oro-geny, Central British Columbia. Canadian Journal of Earth Sciences, v. 18, p. 1767-1775. (1981b): Snowshoe Formation, central British Columbia, In Current Research, Part A, Geological Survey of Canada, Paper 81-1A, p. 213-216. (1982): Snowshoe Formation (1982), central British Columbia, In Current Research, Part B, Geological Survey of Canada, Paper 82-1B, p. 117-124. (1983a): Spanish Lake (NTS 93A/11) and adjoining map areas, British Columbia. Geological Survey of Canada, Open File 920 (map). (1983b): Bedrock geology of Quesnel Lake (93A/10) and part of Mitchell Lake (93A/15) map areas, central British Columbia. Geological Survey of Canada, Open File 962 (map). (1984): Stratigraphy of Quesnel Terrane near Dragon Lake, Quesnel map area, central British Columbia, In Current Research, Part A, Geological Survey of Canada, Paper 84-1A, p. 113-116. (1985a). Pre-Cretaceous terranes and their thrust and strike-slip contacts, Prince George (east half) and McBride (west half) map areas, British Columbia, In Current Research, Part A, Geological Survey of Canada, Paper 85-1A, p. 267-272. (1985b): Dextral strike-slip through Wells Gray Provincial Park, British Columbia, In Current Research, Part A, Geological Survey of Canada, Paper 85-1A, p. 305-309. (1985c): Thrust and strike-slip faults bounding tectono-stratigraphic terranes, cen-tral British Columbia, In Field guides to geology and mineral deposits in the south-ern Canadian Cordillera. D.J. Tempelman-Kluit (ed.), Geological Society of America, Cordilleran Section, Boulder Colorado, p. 141-148. (1986a): Imbricated terranes of the Cariboo gold belt with correlations and implica-tions for tectonics in southeastern British Columbia. Canadian Journal of Earth Sciences, v. 23, p. 1047-1061. References 211 (1986b): A regional east-dipping thrust places Hadrynian onto probable Paleozoic rocks in Cariboo Mountains, British Columbia, In Current Research, Part A , Geological Survey of Canada, Paper 86-1A, p. 589-594. (1987): The ancient western North American margin: an alpine rift model for the east-central Canadian Cordillera. Geological Survey of Canada, Paper 87-15. (1988a): Structural geology of the Cariboo gold mining district, east-central British Columbia. Geological Survey of Canada, Memoir 421, lOOp. (1988b): Regional imbrication within Quesnel Terrane, central British Columbia as suggested by conodont ages. Canadian Journal of Earth Sciences, v. 25, p. 1608-1617. (1988c): Crustal evolution of the eastern Canadian Cordillera. Tectonics, v. 7, p. 727-747. , and Orchard, M .J . (1985): Late Paleozoic conodonts from ribbon chert delineate thrusts within the Antler formation of the Slide Mountain terrane, British Columbia. Geology, v. 13, p. 794-798. Sutherland Brown, A . (1957): Geology of the Antler Creek area, Cariboo District, British Columbia. British Columbia Department of Mines, Bulletin 38. (1963): Geology of the Cariboo River area, British Columbia. British Columbia Department of Mines and Petroleum Resources, Bulletin 47. 60 p. Thompson, A . B . (1976a): Mineral reactions in pelitic rocks: I. Predictions of P - T - X (Fe-Mg) phase relations. American Journal of Science, v. 276, p. 401-424. (1976b): Mineral reactions in pelitic rocks: II. Calculations of some P - T - X (Fe-Mg) phase relations. American Journal of Science, v. 276, p. 425-454. , Tracey, R .J . , Lyttle, P., and Thompson, J . B . (1977): Prograde reaction histories deduced from compositional zonation and mineral inclusions in garnet from the Gassetts schist, Vermont. American Journal of Science, v. 277, p. 1152-1167. Tipper, H.W. (1978): Northeastern part of Quesnel (93B) map-area British Columbia, In Current Research, Part A , Geological Survey of Canada, Paper 78-1A, p. 67-68. , Campbell, R . B . , Taylor, G . C . , and Stott, D.F. (1979): Parsnip River, British Columbia. Geological Survey of Canada, Map 1424A. References 212 , Woodsworth, G.J., and Gabrielse, H. (1981): Tectonic assemblage map of the Canadian Cordillera and adjacent parts of the United States. Geological Survey of Canada, Map 1505A. Tracy, R.J. (1978): High grade metamorphic reactions and partial melting in pelitic schist, west-central Massachusetts. American Journal of Science, v. 278, p. 150-178. (1982): Compositional zoning and inclusions in metamorphic minerals. In Ferry J.M. (ed.), Characterization of Metamorphism through Mineral Equilibria, Washington, D.C, Mineralogical Society of America, Reviews in Mineralogy, v. 10, p. 355-397. , Robinson, P., and Thompson, A . B . (1976): Garnet compositions and zoning in the determination of temperature and pressure of metamorphism, central Massachusetts. American Mineralogist, v. 61, p. 762-775. Wheeler, J . O . (1966): Lardeau (west half) map-area (82K Wl/2), In Report of activities, Geological Survey of Canada Paper 66-1, p. 102-103. , and Gabrielse, H. (1972): The Cordilleran structural province. In Price, R.A., and Douglas, R.J.W. (eds.), Variations in tectonic styles in Canada. Geological Associa-tion of Canada, Special Paper 11, p. 1-81. Wood, B.J., and Fraser, D .G . (1977): Elementary Thermodynamics for Geologists. Oxford University Press, Oxford, Great Britain, 301p. A p p e n d i x A P o i n t s o f c o n f u s i o n a r i s i n g f r o m t h e l i t e r a t u r e 1. Struik (1988a, p. 47, f8) stated that the Snowshoe Group is a formal designation. Struik (1988a) does not give a reference for this and no formal description of what constitutes the Snowshoe Group was found during this study. The earliest reference to rocks designated as Snowshoe Group is Struik (1983a). 2. In determining the sense of shear on the Eureka Thrust (Quesnel Lake Shear Zone) Rees and Ferri (1983) made a number of important and possibly incorrect assumptions. Firstly, they assumed that the kinematic indicators used (feldspar megacrysts) were randomly oriented prior to movement on the thrust. The outcrop used by Rees and Ferri (1983) is from the margin of a sill-like (Fletcher, 1972; Montgomery, 1985; Rees, 1987) orthogneiss body. It is highly likely that it had an original igneous (flow) foliation, particularly near its margins. Secondly, Rees & Ferri (1983) assume that the phases of deformation which postdate the shearing have not affected the kinematic indicators. Simpson & Schmid (1983) specifically warn that later phases of deformation must be considered, evaluated, and their effects removed. 3. Struik (1985a) is given here as the first reference to the Eureka Thrust. Struik (1985a) notes Struik (1983b) as the primary reference but the Eureka Thrust does not traverse the area covered by the 1983b reference. Struik (1983a) did map the contact but did not show it as a fault. 4. The name 'Quesnel River Group' was originally used by Tipper (1978) for strata in the Dragon Lake area (Quesnel map sheet, 93B) but it does not appear to be a formal name even when applied to those strata. 5. Struik (1988a) dated rocks from the upper portion of the TBP as Ladinian and rocks from the lower portion of the TJV as Anisian. His interpretation that these entire packages are time equivalent and that their present overlap (a possible 150 km) is tectonic is unwarranted without corroborating evidence. 6. Struik (1988b, p. 1611, %10) stated that the Spanish Thrust is present along a strike length of 150 km outside the area, on the basis of Tipper et al. (1981). He stated that the contact can be followed and is ".. .a sharp well-defined contact uninterrupted by inter-digitation and can be traced for over 150 km (Tipper et ai, 1981)". The referenced work 213 Appendix A. Points of confusion arising from the literature 214 is the Tectonic Assemblage Map of the Canadian Cordillera. It only shows the distribu-tion of the interpreted assemblages not the individual lithologies that Struik mapped in the Spanish Lake area. The existence of a thrust along the entire length of that contact outside the Spanish Lake area remains to be demonstrated. 7. The term 'vergence' is commonly used by previous workers in the area. However, in the absence of sedimentary way-up criteria 'vergence' only describes the sense of asymmetry of folds. Apparent conflict between the sense of 'vergence' reported by workers for the same fold phase in adjacent areas can result from the lack of sedimentary way-up criteria. Superimposed phases of folding further complicate the matter. Therefore, the 'vergences' reported for the various phases of folding in the Quesnel Lake area are treated with extreme caution. Appendix B Derivation of distribution of species equations and computer programs to calculate activities of mica components B . l Charge balanced species in micas Ionic substitutions are distinguished only on charge. In the A site, A represents all univalent cations, B represents divalent ions. In the octahedral sites P represents all tetravalent cations, Q represents trivalent cations and R represents divalent cations. • represents a vacancy in any site. Only univalent anions (M) substitute into the XH-fold site. The 75 charge balanced species in mica are: Si - P , • ,Al2Si20io(M)2 s2 = A ,P , Q, • ,Al2Si2Oio(M)2 s3 - • , P , Q a,AlSi 3O 1 0(M) 2 s4 = • , P , R, R,Al2Si2Oio(M)2 S 5 A , P , R, • ,AlSi3O10(M)2 S 6 = B , P , R, • ,Al2Si2O1 0(M)2 s7 = • , P , R, • ,Si4O1 0(M)2 s8 = ° , P , • , P,Al2Si2Oio(M)2 s9 A , P , o, Q,Al2Si2Oio(M)2 Sio = ° , P , • , Q,AlSi3Oio(M)2 Sn = A , P , • , R,AlSi3Oio(M)2 S12 = B , P , • , R,Al2Si2010(M)2 S13 = a , P , • R,Si4Oio(M)2 S14 = B , P , • , • ,Si4O1 0(M)2 Sl5 A , Q P • ,Al2Si2Oio(M)2 Sl6 = • , Q P • ,AlSi3O10(M)2 Sl7 = • , Q Q ,R,Al 2Si 2Oio(M) 2 Sl8 = A , Q Q ,o,AlSi3Oio(M)2 Sl9 = B , Q Q • ,Al2Si2Oio(M)2 215 Appendix B. Computer programs to calculate activities of mica components 216 S20 • , Q, Q, S21 = a, Q, R, S22 = A, Q, R, S23 = • , Q, R, S 2 4 -- A, Q, R, S25 = B, Q, R, S26 = A, Q, • , S27 = • , Q, • , S28 = A, Q, • , S29 - B, Q, • , S30 = • , Q, • , S31 = A, Q, • S32 = B, Q, • , S33 - • , R, P, S34 - A, R, P, S35 = B, R, P, S36 • , R, P, S37 • R, Q S38 - A R, Q S39 - • R, Q S40 - A R, Q S4I = B, R, Q S42 • R R S43 = A R R S44 — • R R S45 = A R R S46 = B R, R S47 — • ,R R S48 = B R R S49 = A ,R • S50 = B R • S51 = • ,R S52 -- A ,R S53 = B ,R • a,Si4Oio(M)2 Q,Al2Si201 0(M)2 R,Al2Si2Oio(M)2 R,AlSi3Oio(M)2 • ,Si4Oio(M)2 • ,AlSi3Oio(M)2 P,Al 2Si 2O 1 0(M) 2 P,AlSi 3O 1 0(M) 2 Q,AlSi3O1 0(M)2 Q,Al 2Si 2O 1 0(M) 2 Q,Si4Oio(M)2 R,Si4Oio(M)2 R,AlSi 30 1 0(M) 2 R,Al 2Si 2O 1 0(M) 2 a,AlSi3Oio(M)2 • ,Al2Si2010(M)2 • ,Si4Oio(M)2 ,Q,Al2Si20io(M)2 ,R,Al 2Si20 1 0(M) 2 ,R,AlSi 30 1 0(M) 2 ,o,Si4Oio(M)2 • ,AlSi 3O 1 0(M) 2 ,P,Al2Si20io(M)2 lQ,Al2Si2Oio(M)2 ,Q,AlSi 3O 1 0(M) 2 ,R,AlSi3Oio(M)2 R,Al 2Si 2O 1 0(M) 2 ,R,Si 40 1 0(M) 2 .•,Si4Oio(M)2 (P,AlSi301 0(M)2 ,P,Al2Si2O10(M)2 ,P,Si 40 1 0(M) 2 ,Q,Si 40 1 0(M) 2 ,Q,AlSi3O10(M)2 Appendix B. Computer programs to calculate activities of mica components 217 S54 = B , R , 0 , S55 - p, S56 = A , o P, S57 = • , • p, S58 A , o p, S59 = B , D , p, S60 - • , • p, S61 = B , D , p, S62 = A , D Q S63 = • , • Q S64 = Q S65 - B , D , Q S66 - • , • Q S67 = Q S68 = B , o , Q S69 = A,o R S70 - B , D , R S71 — • , • R S72 = A ,D R S73 B , D R S74 -- B , D R S 7 5 = B , D D R, Si4Oio(M)2 P,Al 2Si 2O 1 0(M)2 Q,Al 2Si 2O 1 0(M) 2 Q,AlSi 3O 1 0(M) 2 R ,AlSi 3O 1 0(M) 2 R,Al2Si2O10(M)2 R, Si4Oio(M)2 • ,Si4Oio(M)2 ,P,Al 2Si 20 1 0(M)2 ^,^813010(1^)2 ,Q,AlSi3Oio(M)2 Q,Al2Si2Oio(M)2 ,Q,Si4Oio(M)2 ,R,Si 4Oio(M) 2 R,AlSi 3Oio(M) 2 P,AlSi 3Oio(M) 2 P,Al 2Si 2Oio(M) 2 ,P,Si 4Oio(M) 2  1Q,Si 4Oio(M) 2 Q,AlSi3Oio(M)2 R, Si4Oio(M)2 P,Si 4Oio(M) 2 B.2 Equilibria among charge balanced species in mica Using the method described in chapter 3 the following equilibria can be written between the species in previous section. S 3 = 0.5Si + 0.5S 2 0 S 4 + 0.5S2 + 0.25S 2 0 = 0.75Si + O.5S17 + 0.5S 4 5 S5 + O.5S17 = 0.25Si + 0.5S2 + 0.25S 2 0 + 0.5S 4 5 S 7 + 0.5S2 + 0.5Si7 = 0.75Si + 0.75S 2 0 + 0.5S 4 5 S 8 = Sa Appendix B. Computer programs to calculate activities of mica components 218 Sn = S5 512 = S6 513 = S7 S14 + S17 = + S20 515 = S2 516 — S3 Sis + 0.5Si = S2 + 0.5S 2 O S19 + 0.75Si + 0.5S 4 5 = 0.5S2 + S6 + O.5S17 + 0.25S2o S2I — S17 S 2 2 + 0.25SX + 0.25S2O = 0.5S2 + O.5S17 + 0.5S45 S 2 3 + 0.5S2 = 0.25S! + 0.5S17 + 0.25S2o + 0.5S45 S 2 4 + 0.25Si + O.5S17 = 0.5S2 + 0.75S20 + 0.5S45 S 2 5 + 0.5S! = S6 + 0.5S20 S26 = s2 S27 = S 3 S28 = Sl8 S29 = Sl9 S30 = S20 S31 = S24 S32 = S25 S33 = s4 S34 s5 S 3 5 = S6 S36 — S7 S37 S17 S38 S22 S39 = S23 S4O S24 S41 = S25 S42 = s4 S43 = S22 S44 S23 S 4 6 + 0.25Si + 0.5S2 + 0.25S2O = S6 + 0.5S17 + 0.5S45 S47 + S2 = 0.5Si + 0.5S20 + S 4 5 Appendix B. Computer programs to calculate activities of mica components 219 5S 1 7 = s6 + S49 = s5 S50 = s6 S51 = s7 S52 = S24 s 5 3 • S25 S54 = S48 S55 = Si S56 = s2 s57 = S3 S58 = s5 S59 = S6 S60 = S7 S61 = S14 S62 = s2 S63 = S3 S64 = S18 S65 = Sl9 S66 = S20 S67 = S24 S68 = S25 S69 = S5 S70 = s6 S71 = s7 S72 = S24 S73 = S25 S74 = S48 s75 Sl4 As noted in chapter 3, all species were originally written in terms of the independent species (Si, S2, S17, S2o, and S45). However, it is simpler to note that many species are composition-ally identical and therefore only the first such species in the list of 75 is given in terms of the independent species. Subsequent species are written in terms of the first. Appendix B. Computer programs to calculate activities of mica components 220 B.3 Mass action equations for charge balanced species in mica In the absence of thermodynamic data one must assume that the equilibrium constants for the equilibria in the previous section are all unity. It is therefore a simple task to write the equilibrium constant for each. For the first equilibrium above this is V 0 . 5 . Y 0 . 5 K = l = 1 * 2 0 X 3 or X _ v0.5 v-0.5 3 — -^1 - -^-20 • Note that for computational purposes, terms are squared, cubed or multiplied to the fourth power to remove exponents less than one so that the previous equation becomes X 3 = X i • X 2 o . The equations from the above equilibria are (in the same order): x| = X j • X20 •n- X20 - x?- Y 2 A 1 7 . Y2 A 4 5 x 5 - Y2 A 1 7 = X i - x l - X20 • Y 2 A 4 5 X 7 • x\. Y2 = x?- Y 3 A20 . Y 2 A 4 5 x 8 - X ! x 9 = x 2 X10 = x 3 X n — x 5 X12 - x 6 X13 = x 7 X14 • X17 = x 6 - X20 X15 = x 2 Xl6 - x 3 Y 2 A 1 8 • X i - x*. X20 Y 4  A 1 9 •x?. Y 2 A 4 5 = x\. Y4 . A 6 Y 2 . A 1 7 X20 X21 = X 1 7 Y 4 A22 • X i - X20 — x\. Y2 A 1 7 . Y 2 A 4 5 X 2 3 • xl = X i - Y 2  A 1 7 • X20 . Y 2 A 4 5 Y 4  A 2 4 • X i - x\7 x\. Y3  A20 • Y2 A 4 5 Y 2 A 2 5 •Xx = x i - X20 Appendix B. Computer programs to calculate activities of mica components 221 X 2 6 x 2 X 2 7 = x 3 X 2 8 = X i 8 X29 = X 1 9 X 3 0 = X 2 0 X31 - X24 X32 = X25 X 3 3 - x 4 X34 x 5 X 3 5 - x 6 X 3 6 = x 7 X 3 7 X17 X 3 8 X22 X39 X 2 3 X 4 0 - X24 X41 X 2 5 X42 - x 4 X 4 3 = X 2 2 X44 = X23 X4 A 4 6 •X! X 2 0 = X(5 • Xj 7 . X2 A 4 5 A47 • xl = Xi • X2rj . X2  A 4 5 Y 4 A 4 8 •X! X2 A 1 7 = X6 ' XJO . X2 A 4 5 X 4 9 = x 5 X 5 0 — x 6 X51 x 7 X 5 2 - X24 X 5 3 X25 ^54 - X 4 8 X55 — Xi X 5 6 x 2 X 5 7 x 3 X 5 8 --- x 5 X 5 9 = x 6 X60 - x 7 Appendix B. Computer programs to calculate activities of mica components 222 X61 = X14 ^ 6 2 = x 2 X-63 - x 3 ^ 6 4 = X 1 8 X65 = X 1 9 X 6 6 = X 2 0 X67 = X24 X 6 8 = X 2 5 X69 - x 5 X 7 0 = x 6 X71 = x 7 X72 - X 2 4 X 7 3 - X 2 5 X74 - X 4 8 X 7 5 X14 These equations along with an additional 6 mass balance equations are evaluated for micas using unit DistSpec below. The equations in DistSpec are in a different order to those above though the species numbers are the same. The first 6 equations in DistSpec are the mass balance equations. The non-linear equations in the above set are grouped and evaluated last by DistSpec. B.4 MICAC.PAS program {**^ ********************************^  PROGRAM MICa_Activi ty_Calculator; -[***************** ***>*^********^^ •C VERSION 1.0 Program to calculate the a c t i v i t i e s of the dioctahedral micas (muscovite - paragonite) , ce ladonite , t r i o c t h e d r a l micas ( b i o t i t e ) , margarite oe l lacher i t e , pyrophy l l i t e and t a l c i n a s o l i d so lut ion based on the assumption that a l l species i n a s o l i d solut ion must be charge balanced. Input can be in the form of oxide weight percent, element weight percent or as molecular formula. Output choice of short (table of 'new' a c t i v i t i e s only) or long ( l i s t of 'new' and ' o l d ' a c t i v i t i e s ) format. F i n a l values calculated are tested against the equations derived from d i s t r i b u t i o n of species (unit DistSpec) . } -[*******************************^ Appendix B. Computer programs to calculate activities of mica components 223 USES Dos , C r t , P r i n t e r , D i s t S p e c ; {**************************^ LABEL BOTTOM,B0TT0M2,REPOCT; CONST NumElt = 13; M i n P e r c e n t = 0 . 9 9 5 ; MaxPercen t = 1 .005; M i n P e r c e n t 2 = 0 . 9 9 2 5 ; MaxPercen t2 = 1.0075; MuscMin = 6; MuscMax = 8; MaxDim = 4 ; •[**************#**##*****^  TYPE RVec = ARRAY[Ind2] o f REAL; OxVec = A R R A Y [ 1 . . N u m E l t ] o f REAL; N o r M a t r i x = A R R A Y [ 1 . . 2 , 1 . . N u m E l t ] o f REAL; I n t A r r = ARRAY[Ind2 , Ind2] o f BYTE; R e a l A r r = ARRAY[Ind2 , Ind2] o f REAL; T i t l e = STRING[14] ; E l S t r = STRING[3] ; E I N A r r = A R R A Y [ 1 . . N u m E l t ] o f E l S t r ; VAR c h e c k , c h e c k 2 , c o u n t , i , i 1 , i 2 , i 3 , i 4 , i n d e x , i t e r , j 1 , j 2 , j 3 , j 4 , k l , k 2 , n , n s t o p , n u m : B Y T E ; A l P l u s , A 2 P l u s , A c t B r i t , A c t C e l , A c t D i , A c t P y p h , A c t T r i , A c t T a l c . A n n A c t , A n n A c t 0 1 d , A n n A c t 0 1 d 2 , A S i t e 0 c c , C e l A c t , C e l A c t O l d , M a r g A c t , M a r g A c t O l d , M u s c A c t , M u s c A c t O l d , m u l t , 0 c t 2 P l u s , 0 c t 3 P l u s , O c t a l , O c t S i t e O c c , O e l l A c t , O e l l A c t O l d , O H O c c , 0 S i t 0 c c 2 . P a r a A c t , P a r a A c t 0 1 d , P h l A c t . P h l A c t O l d , P h l A c t 0 1 d 2 , P r e d S i , P y p h A c t , T a l c A c t , t o t a l , X A 1 1 , X B a , X C a , X E i g h t , X F e , X F e 1 , X K , X K 1 , X M g , X M g l , X N a , X N a l , X O H m o d , X 0 H m o d 2 , x r e a l . X S e v e n . X S i x . X T i l : R E A L ; xcheck : R A r r 5 ; FN : R A r r 2 7 5 ; imat : I M a t r i x ; amat ,dmat , tmat : R M a t r i x ; norm : N o r M a t r i x ; Appendix B. Computer programs to calculate activities of mica components 224 A t W t , E l t o O x , v a l e n c e PISum.marg charge ndim a n s 1 , a n s 2 , a n s 3 , a n s 4 , a u t o , o u t i n d Elname f i l e a , f i l e l , f i l e 2 , f i l e 3 , f i l e 4 , l a b e l a , n a m e c o n d i t i o n , i n c o d e E l W t , F o r m u l a , I n p u t , O u t f r a c , O u t p u t , O u t p u t 1 , 0 u t p u t 2 :OxVec; : R e a l A r r ; : I n t A r r ; : I V e c ; :CHAR; : E l N A r r ; : T i t l e ; :BOOLEAN; :TEXT; {****************************^  BEGIN C l r S c r ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e C Inpu t d a t a f i l e ( . I N e x t e n s i o n u n d e r s t o o d ) . : ' ) ; r e a d l n ( f i l e a ) ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e C S top a f t e r how many p o i n t s (0 f o r f u l l f i l e ) . : ' ) ; r e a d l n ( n s t o p ) ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e C Do you w i s h t o c a l c u l a t e f o r a p a r t i c u l a r p o i n t i n t h e f i l e ? : ' ) ; r e a d l n ( a n s l ) ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e C Do you want l o n g <L> o r s h o r t <S> ou tpu t fo rma t? : ' ) ; r e a d l n ( a n s 3 ) ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e C Do you want an automated (A) o r i n t e r a c t i v e ( I ) run? : ' ) ; r e a d l n ( a u t o ) ; i f ( U p C a s e ( a n s l ) = ' Y ' ) Appendix B. Computer programs to calculate activities of mica components 225 t h e n b e g i n w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e ( ' What i s t h a t p o i n t l a b e l ? (Remember upper and l o w e r case ) : ' ) ; r e a d l n ( l a b e l a ) ; end; f o r k2 := L E N G T H ( l a b e l a ) t o 14 do l a b e l a := C O N C A T Q a b e l a , ' ' ) ; i f ( f i l e a = " ) t h e n f i l e a := ' M i c a ' ; f i l e l := C 0 N C A T ( f i l e a , ' . I N ' ) ; f i l e 2 := C 0 N C A T ( f i l e a , ' . S H T ' ) ; f i l e 3 := C 0 N C A T ( f i l e a , ' . M L F ' ) ; f i l e 4 := C 0 N C A T ( f i l e a , ' . L O N ' ) ; a s s i g n ( I n p u t , f i l e l ) ; r e s e t ( I n p u t ) ; a s s i g n ( E l W t , ' e l e m e n t . d a t ' ) ; r e s e t ( E l W t ) ; a s s i g n ( O u t p u t , f i l e 2 ) ; r e w r i t e ( O u t p u t ) ; a s s i g n ( O u t p u t 1 , ' n o n l i n . d a t ' ) ; r e w r i t e ( O u t p u t l ) ; a s s i g n ( 0 u t p u t 2 , f i l e 4 ) ; r e w r i t e ( 0 u t p u t 2 ) ; a s s i g n ( O u t F r a c , f i l e 3 ) ; r e w r i t e ( O u t F r a c ) ; { a s s i g n ( F o r m u l a , ' F o r m u l a . d a t ' ) ; r e w r i t e ( F o r m u l a ) ; } i n c o d e := F A L S E ; w h i l e no t i n c o d e do b e g i n C l r S c r ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e l n ; w r i t e l n C A r e i n p u t d a t a as element we igh t p e r c e n t ( " E " ) : ' ) ; { Ass ignment o f i n p u t } { and ou tpu t f i l e s . } Appendix B. Computer programs to calculate activities of mica components 226 writelnC or as oxide weight percent C ' O ' ' ) : ' ); writeC or as molecular formula format C ' M ' ' ) : ' ) ; readln(ans2); i f (UpCase(ans2) = 'E') or (UpCase(ans2) = '0') or (UpCase(ans2) = 'M') then incode := TRUE; end; ClrScr; ndim[l] := 3; ndim[2] := 4; { # of elements per > ndim[3] := 4; •C dimension. > ndim[4] := 4; charge[1,1] = 1 charge[1,2] = 2 { Assignment of the } charge[1,3] 0 { charges on each } charge[2,1] = 4 { element in each > charge[2,2] = 3 { site. } charge[2,3] = 2 charge[2,4] = 0 for k2 := 3 to 4 do for kl := 1 to 4 do charge[k2,kl] := charge[2,kl]; for k2 := 1 to NumElt do begin readln(ElWt,AtWt[k2],Elto0x[k2], { Read at. weights, } valence[k2],Elname[k2]); { 2 factors & symbol. } for kl := LENGTH(Elname[k2]) to 3 do Elname[k2] := CONCAT(Elname[k2],' '); end; {**************************************^  { Major loop, repeated for each data point ( i .e . each analysis). } n := 0; while not EOF(Input) do begin count := 0; outind := ' ' ; n := n + 1; i f (n > nstop) and (nstop <> 0) then goto B0TT0M2; readln(Input.name); { Read in data pt. ID } Appendix B. Computer programs to calculate activities of mica components f o r k2 := LENGTH(name) t o 14 do name := CONCAT(name,' ' ) ; i f ( U p C a s e ( a n s l ) = ' Y ' ) and (name <> l a b e l a ) t h e n b e g i n r e a d l n ( I n p u t ) ; go to BOTTOM; end; i f UpCase(ans3) = ' S ' t h e n b e g i n i f ( ( n - 1 ) mod 40 = 0) o r ( U p C a s e ( a n s l ) = ' Y ' ) t h e n b e g i n w r i t e l n ( O u t p u t ) ; w r i t e l n ( O u t F r a c ) ; w r i t e l n ; w r i t e l n ( 0 u t p u t , ' Sample Muse. P a r a . C e l a . A n n i . P h l o . M a r g . ' , ' O e l l . P y p h . T a l c . ' ) ; w r i t e l n ( ' S a m p l e Muse. P a r a . C e l a . A n n i . P h l o . M a r g . ' , ' O e l l . P y p h . T a l c . ' ) ; w r i t e l n ( O u t F r a c , ' Sample [ - X M g - ] [ - X F e - ] [ - X T i - ] [ - X A 1 - ] [ - X K — ] [ X O H b i ] ' , ' [XNamu][XOHmu] ' ) ; end; w r i t e ( O u t p u t , ' ' . n a m e ) ; w r i t e l n ( O u t p u t 1 , ' ' . n a m e ) ; w r i t e ( O u t F r a c , ' ' . n a m e ) ; w r i t e ( n a m e ) ; end e l s e b e g i n w r i t e l n ( 0 u t p u t 2 ) ; w r i t e l n ; w r i t e l n ( 0 u t p u t 2 , ' " - - — ') w r i t e l n C ' ) w r i t e l n ( 0 u t p u t 2 ) ; w r i t e l n ; w r i t e l n ( O u t p u t l , ' ' . n a m e ) ; w r i t e l n C ' . n a m e ) ; w r i t e l n ( 0 u t p u t 2 , ' ' . n a m e ) ; end; •[********^*************************^ { Loop t o r e a d i n c o m p o s i t i o n s o f unknowns i n terms o f e l e m e n t s , o x i d e s , Appendix B. Computer programs to calculate activities of mica components 228 s t r u c t u r a l f o r m u l a . I f d a t a i s e i t h e r e lement o r o x i d e c o n c e n t r a t i o n s program c a l c u l a t e s a s t r u c t u r a l f o r m u l a based on 22 charges (anhydrous f o r m u l a ) . E l e m e n t s / O x i d e s a r e : 1, S i 0 2 : 2 , A1203: 3 , T i 0 2 : 4 , Fe203 : 5 , FeO: 6 , MgO: 7 , MnO: 8, CaO: 9 , BaO: 10 , Na20: 1 1 , K20: 12 , F : 13 , C l : } {*************************^ ans4 := ' N ' ; REPOCT: i f (UpCase(ans2) = ' M ' ) t h e n b e g i n f o r k2 := 1 t o NumElt do r e a d ( I n p u t , n o r m [ 2 , k 2 ] ) ; r e a d l n ( I n p u t ) ; end e l s e b e g i n i f (UpCase(ans2) = ' 0 ' ) and (UpCase(ans4) <> ' V ) t h e n b e g i n f o r k2 := 1 t o NumElt do b e g i n r e a d ( I n p u t , n o r m [ l , k 2 ] ) ; n o r m [ l , k 2 ] := n o r m [ l , k 2 ] / E l t o O x [ k 2 ] ; end ; r e a d l n ( I n p u t ) ; end ; i f (UpCase(ans2) = ' E ' ) and (UpCase(ans4) <> ' Y ' ) t h e n b e g i n f o r k2 := 1 t o NumElt do r e a d ( I n p u t , n o r m [ l , k 2 ] ) ; r e a d l n ( I n p u t ) ; end ; i f (UpCase(ans2) = ' 0 ' ) o r (UpCase(ans2) = ' E ' ) t h e n b e g i n t o t a l := 0 ; f o r k2 := 1 t o NumElt do b e g i n n o r m [ 2 , k 2 ] := n o r m [ l , k 2 ] / AtWt[k2] * v a l e n c e [ k 2 ] ; t o t a l := t o t a l + n o r m [ l , k 2 ] * E l t o 0 x [ k 2 ] ; end ; t o t a l := t o t a l - ( n o r m [ l , 1 2 ] * 0 .42107) - ( n o r m [ l , 1 3 ] * 0 . 2 2 5 6 4 ) ; end ; i f ( t o t a l < 9 3 . 5 ) o r ( t o t a l >98.5) Appendix B. Computer programs to calculate activities of mica components 229 t h e n b e g i n w r i t e l n ( O u t p u t , ' T o t a l wt '/, o x i d e ( ' . t o t a l : 5 : 2 , ' ) o u t s i d e p e r m i t t e d r a n g e . ' ) w r i t e l n ( 0 u t p u t 2 , ' T o t a l wt '/. o x i d e ( ' . t o t a l : 5 : 2 , ' ) o u t s i d e p e r m i t t e d r a n g e . ' ) w r i t e l n ( O u t F r a c , ' 0 ' , ' T o t a l wt '/, o x i d e ( ' . t o t a l : 5 : 2 , ' ) o u t s i d e p e r m i t t e d r a n g e . ' ) w r i t e l n ( ' T o t a l wt '/, o x i d e ( ' . t o t a l : 5 : 2 , ' ) o u t s i d e p e r m i t t e d r a n g e . ' ) go to BOTTOM end; t o t a l := 0 ; f o r k2 := 1 t o NumEl t -2 do t o t a l := t o t a l + n o r m [ 2 , k 2 ] ; mu l t := 2 2 . 0 / t o t a l ; f o r k2 := 1 t o NumElt do b e g i n n o r m [ 2 , k 2 ] := norm[2 ,k2] * m u l t ; no rm[2 ,k2 ] := norm[2 ,k2] / v a l e n c e [ k 2 ] ; end; end; { w r i t e l n ( F o r m u l a ) ; w r i t e l n ( F o r m u l a . n a m e ) ; f o r k2 := 1 t o NumElt do w r i t e l n ( F o r m u l a , E l n a m e [ k 2 ] , , n o r m [ 2 , k 2 ] : 4 : 3 ) ; > A S i t e O c c := no rm[2 ,8 ] + norm[2 ,9] + norm[2 ,10] + n o r m [ 2 , l l ] ; O c t A l := no rm[2 ,2 ] - (4 - n o r m [ 2 , l ] ) ; O c t S i t e O c c := O c t A l + norm[2,3] + norm[2 ,4] + norm[2 ,5 ] + norm[2,6] + norm[ 2 .7 ] i f O c t A l < 0 t h e n b e g i n w r i t e l n ( 0 u t p u t 2 , ' T - S i t e o c c . ( ' , 4 + 0 c t A l : 5 : 3 , ' ) < 4 . ' ) ; w r i t e ( O u t p u t , ' T - S i t e o c c . ( ' , 4 + 0 c t A l : 5 : 3 , ' ) < 4 . ' ) ; w r i t e ( 0 u t F r a c , ' 0 ' , ' T - S i t e o c c . ( ' , 4 + 0 c t A l : 5 : 3 , ' ) < 4 . ' ) ; w r i t e C T - S i t e o c c . ( ' , 4 + 0 c t A l : 5 : 3 , ' ) < 4 . ' ) ; end; { C a l c u l a t i o n o f A } { s i t e and o c t . s i t e } { occupancy . } { W r i t e message i f } { t e t r a h e d r a l s i t e } { occupancy i s l e s s } •C t h a n 4 . > Appendix B. Computer programs to calculate activities of mica components 230 i f A S i t e O c c > 1 t h e n b e g i n w r i t e l n ( 0 u t p u t 2 , ' A - S i t e o c c . ( ' , A S i t e 0 c c : 5 : 3 , ' ) > 1. ' ) w r i t e ( O u t p u t , * A - S i t e o c c . ( ' , A S i t e 0 c c : 5 : 3 , ' ) > 1. ' ) ; w r i t e ( O u t F r a c , ' 0 • A - S i t e o c c . ( ' , A S i t e 0 c c : 5 : 3 , ' ) > 1. ' ) ; w r i t e ( ' A - S i t e o c c . ( ' , A S i t e 0 c c : 5 : 3 , ' ) > 1. ' ) ; end; { W r i t e message i f A { s i t e occupancy i s { g r e a t e r t h a n 1. i f O c t S i t e O c c < 2 t h e n b e g i n w r i t e l n ( 0 u t p u t 2 , ' 0 - S i t e o c c . ( ' , 0 c t S i t e 0 c c : 6 : 4 , ' ) < 2 . ' ) ; w r i t e ( O u t p u t , ' 0 - S i t e o c c . ( ' , 0 c t S i t e 0 c c : 6 : 4 , ' ) < 2 . ' ) ; w r i t e ( O u t F r a c , ' 0 ' , ' 0 - S i t e o c c . ( ' , 0 c t S i t e 0 c c : 6 : 4 , ' ) < 2 . ' ) ; w r i t e ( ' 0 - S i t e o c c . ( ' , 0 c t S i t e 0 c c : 6 : 4 , ' ) < 2 . ' ) ; end; { W r i t e message i f { o c t a h e d r a l s i t e { occupancy i s l e s s < t h a n 2 } } } } } } > i f O c t S i t e O c c > 3 t h e n b e g i n w r i t e l n ( 0 u t p u t 2 , ' 0 - S i t e o c c . ( ' , 0 c t S i t e 0 c c : 6 : 4 , ' ) > 3 . ' ) ; w r i t e ( O u t p u t , ' 0 - S i t e o c c . ( ' , 0 c t S i t e 0 c c : 6 : 4 , ' ) > 3 . ' ) ; w r i t e ( O u t F r a c , ' 0 ' 0 - S i t e o c c . ( ' , 0 c t S i t e 0 c c : 6 : 4 , ' ) > 3 . ' ) ; w r i t e C 0 - S i t e o c c . ( ' . O c t S i t e O c c : 6 : 4 , ' ) > 3 . ' ) ; end; { o r g r e a t e r } •C t h a n 3 . } i f ( A s i t e O c c > 1) o r ( O c t S i t e O c c < 2) o r ( O c t S i t e O c c > 3) o r ( O c t A l < 0) t h e n b e g i n w r i t e l n ( O u t p u t ) ; w r i t e l n ( O u t F r a c ) ; w r i t e l n ; i f ( A S i t e O c c > i ) o r ( O c t S i t e O c c < 2) t h e n b e g i n ans4 := ' N * ; i f ( coun t < 1) and (UpCase(au to) = ' I ' ) { S k i p t o nex t > { a n a l y s i s i f s i t e } { occupancy i s no t } { g o o d . > Appendix B. Computer programs to calculate activities of mica components 231 then begin writeC Do you wish to boost the A l content t o solve this? readln(ans4); end; if (UpCase(auto) = ' A ' ) then ans4 := ' Y ' ; if (UpCase(ans4) = ' Y ' ) and (count < 1) then begin count := count + 1; outind := '#'; norm[l,2] := norm[l,2] + 0 . 5 ; writeC '); write (Output , ' ' ) ; write(0utfrac,' '); goto REPOCT; end; end; goto BOTTOM end; ); m a r g [ l , l ] := norm[2 ,10] + n o r m [ 2 , l l ] ; m a r g [ l , 2 ] := norm[2 ,8] + n o r m [ 2 , 9 ] ; m a r g [ l , 3 ] := 1 - A S i t e O c c ; m a r g [ 2 , l ] := norm[2 ,3] / 3 ; ma rg [2 ,2 ] := ( O c t A l + n o r m [ 2 , 4 ] ) / 3 ; m a r g [ 2 , 3 ] := (norm[2 ,5] + norm[2 ,6 ] + n o r m [ 2 , 7 ] ) / 3 ; ma rg [2 ,4 ] := 1.0 - O c t S i t e O c c / 3 ; f o r k2 := 3 t o 4 do f o r k l := 1 t o 4 do m a r g [ k 2 , k l ] := m a r g [ 2 , k l ] ; f o r i 4 := 1 t o ndim[4] do f o r i 3 := 1 t o ndim[3] do f o r i 2 := 1 t o ndim[2] do f o r i l := 1 t o n d i m [ l ] do b e g i n a m a t [ i 4 , i 3 , i 2 , i l ] := m a r g [ 4 , i 4 ] * m a r g [ 3 , i 3 ] * m a r g [ 2 , i 2 ] * m a r g [ l , i l ] ; check := c h a r g e [ 4 , i 4 ] + c h a r g e [ 3 , i 3 ] • c h a r g e [ 2 , i 2 ] + c h a x g e [ l , i l ] ; i f ( check >= MuscMin) and (check <= MuscMax) t h e n i m a t [ i 4 , i 3 , i 2 , i l ] := TRUE e l s e i m a t [ i 4 , i 3 , i 2 , i l ] := F A L S E ; check := 0 { Loop t o f i l l t h e } { m a r g i n a l v e c t o r s . } •( F i l l i n g i n i t i a l } { m a t r i x . } { Check o f t o t a l } { cha rge f o r t h e f o u r } { s i t e s i n m i c a s . } { A s s i g n TRUE i f > { t o t a l charge i s } { between 6 and 8 } { e l s e FALSE } Appendix B. Computer programs to calculate activities of mica components 232 end; { R e d u c t i o n o f a c t i v i t i e s o f n o n - p e r m i t t e d s p e c i e s t o z e r o . } f o r i 4 := 1 t o ndim[4] do f o r i 3 := 1 t o ndim[3] do f o r i 2 := 1 t o ndim[2] do f o r i l := 1 t o n d i m [ l ] do i f no t i m a t [ i 4 , i 3 , i 2 , i l ] t h e n a m a t [ i 4 , i 3 , i 2 , i l ] := 0 ; {********^*****************^ { S t a r t o f h y p e r p l a n e r e b a l a n c i n g s e c t i o n . Makes su re t h a t the h y p e r p l a n e sums e q u a l t h e m a r g i n a l v e c t o r s w i t h i n s p e c i f i e d m a r g i n o f e r r o r ( i . e . between M i n P e r c e n t and M a x P e r c e n t ) . } -[********************^ c o n d i t i o n := F A L S E ; i t e r := 0 ; r e p e a t i t e r := i t e r + 1; f o r k2 := 1 t o MaxDim do f o r k l := 1 t o ndim[k2] do P l S u m [ k 2 , k l ] := 0 ; { I n i t i a l i z a t i o n o f } •C p l a n e sums t o 0 . } f o r i 4 := 1 t o ndim[4] do f o r i 3 := 1 t o ndim[3] do f o r i 2 := 1 t o ndim[2] do f o r i l := 1 t o n d i m [ l ] do P l S u m [ 4 , i 4 ] := P l S u m [ 4 , i 4 ] + a m a t [ i 4 , i 3 , i 2 , i 1 ] ; { C a l c u l a t i o n o f D-4 } { h y p e r p l a n e sums } f o r i 4 := 1 t o ndim[4] do f o r i 3 := 1 t o ndim[3] do f o r i 2 := 1 t o ndim[2] do f o r i l := 1 t o n d i m [ l ] do i f a m a t [ i 4 , i 3 , i 2 , i l ] <> 0 t h e n a m a t [ i 4 , i 3 , i 2 , i l ] := a m a t [ i 4 , i 3 , i 2 , i l ] * m a r g [ 4 , i 4 ] / P I S u m [ 4 , i 4 ] ; f o r i 4 := 1 t o ndim[4] do f o r i 3 := 1 t o ndim[3] do f o r i 2 := 1 t o ndim[2] do f o r i l := 1 t o n d i m [ l ] do { N o r m a l i z a t i o n o f > •( D-4 h y p e r p l a n e sums } { C a l c u l a t i o n o f D-3 } { h y p e r p l a n e sums > Appendix B. Computer programs to calculate activities of mica components 233 P l S u m [ 3 , i 3 ] := P l S u m [ 3 , i 3 ] + a m a t [ i 4 , i 3 , i 2 , i l ] ; f o r i 4 := 1 t o nd im[4] do f o r i 3 := 1 t o nd im[3] do f o r i 2 := 1 t o nd im[2] do f o r i l := 1 t o n d i m [ l ] do i f a m a t [ i 4 , i 3 , i 2 , i i ] <> 0 t h e n a m a t [ i 4 , i 3 , i 2 , i l ] := a m a t [ i 4 , i 3 , i 2 , i l ] * m a r g [ 3 , i 3 ] / P l S u m [ 3 , i 3 ] ; f o r i 4 := 1 t o nd im[4] do f o r i 3 : - 1 t o nd im[3] do f o r i 2 := 1 t o ndim[2] do f o r i l := 1 t o n d i m [ l ] do P l S u m [ 2 , i 2 ] := P l S u m [ 2 , i 2 ] + a m a t [ i 4 , i 3 , i 2 , i l ] ; f o r i 4 := 1 t o nd im[4] do f o r i 3 := 1 t o ndim[3] do f o r i 2 := 1 t o nd im[2] do f o r i l := 1 t o n d i m [ l ] do i f a m a t [ i 4 , i 3 , i 2 , i l ] <> 0 t h e n a m a t [ i 4 , i 3 , i 2 , i l ] := a m a t [ i 4 , i 3 , i 2 , i l ] * m a r g [ 2 , i 2 ] / P l S u m [ 2 , i 2 ] ; { N o r m a l i z a t i o n o f } { D-3 h y p e r p l a n e sums } { C a l c u l a t i o n o f D-2 } { h y p e r p l a n e sums } { N o r m a l i z a t i o n o f } { D-2 h y p e r p l a n e sums } f o r i 4 := 1 t o nd im[4] do f o r i 3 := 1 t o ndim[3] do f o r i 2 : - 1 t o ndim[2] do f o r i l := 1 t o n d i m [ l ] do P l S u m C l . i l ] := P l S u m C l . i l ] + a m a t [ i 4 , i 3 , i 2 , i 1 ] ; { C a l c u l a t i o n o f D - l } { h y p e r p l a n e sums } f o r i 4 := 1 t o nd im[4] do f o r i 3 := 1 t o nd im[3] do f o r i 2 := 1 t o ndim[2] do f o r i l := 1 t o n d i m [ l ] do i f a m a t [ i 4 , i 3 , i 2 , i l ] <> 0 t h e n a m a t [ i 4 , i 3 , i 2 , i l ] := a m a t [ i 4 , i 3 , i 2 , i l ] * m a r g C l . i l ] / P l S u m C l . i l ] ; f o r k2 := 1 t o MaxDim do f o r k l := 1 t o nd im[k2] do P l S u m C k 2 . k l ] := 0 ; { N o r m a l i z a t i o n o f > { D - l h y p e r p l a n e sums } { I n i t i a l i z a t i o n o f } { p l a n e sums t o 0 . } f o r i 4 := 1 t o ndimC4] do f o r i 3 := 1 t o nd im[3] do Appendix B. Computer programs to calculate activities of mica components 234 f o r i 2 := 1 t o ndim[2] do f o r i l := 1 t o n d i m [ l ] do b e g i n P l S u m [ 4 , i 4 ] P l S u m [ 3 , i 3 ] P l S u m [ 2 , i 2 ] P l S u m C l . i l ] end; check2 := 0 ; num := 0 ; = P l S u m [ 4 , i 4 ] + a m a t [ i 4 , i 3 , i 2 , i l ] = P l S u m [ 3 , i 3 ] + a m a t [ i 4 , i 3 , i 2 , i l ] = P l S u m [ 2 , i 2 ] + a m a t [ i 4 , i 3 , i 2 , i l ] = P l S u m C l . i l ] + a m a t [ i 4 , i 3 , i 2 , i l ] = 1 t o MaxDim do := 1 t o ndimCk2] do f o r k2 : f o r k l b e g i n num := num + 1; i f ( P l S u m C k 2 , k l ] = 0) and ( m a r g [ k 2 , k l ] = 0) t h e n x r e a l : = 1 e l s e x r e a l : = m a r g [ k 2 , k l ] / P l S u m [ k 2 , k l ] ; i f ( x r e a l > M i n P e r c e n t ) and ( x r e a K MaxPercen t ) t h e n check2 := check2 + 1; P l S u m [ k 2 , k l ] := 0 end; { G e n e r a t i o n o f t he { p l a n e sums. > } Check o f p l a n e sums as a f r a c t i o n o f t h e m a r g i n a l v e c t o r s . P e r m i t t e d m a r g i n o f e r r o r i s 1 0 0 * ( l - M i n P e r c e n t ) ' / , > i f check2 = num t h e n c o n d i t i o n := TRUE; i f i t e r = 150 t h e n b e g i n w r i t e l n ( O u t p u t , ' A t t empt t o f o r c e d i m e n s i o n ' , ' sums f a i l e d . (150 i t e r a t i o n s ) ' ) ; w r i t e l n ( 0 u t F r a c , ' 0 ' , ' A t t empt t o f o r c e d i m e n s i o n ' , ' sums f a i l e d . (150 i t e r a t i o n s ) ' ) ; w r i t e l n ( ' At tempt t o f o r c e d i m e n s i o n ' , ' sums f a i l e d . (150 i t e r a t i o n s ) ' ) ; w r i t e l n ( o u t p u t 2 , { E r r o r message i f } { p l a n e sums cannot } { be b rough t w i t h i n } { t h e m a r g i n o f e r r o r } i Appendix B. Computer programs to calculate activities of mica components 235 ' At tempt t o f o r c e d i m e n s i o n ' , ' sums f a i l e d . (150 i t e r a t i o n s ) ' ) ; ans4 := ' N ' ; i f ( count < 1) and (UpCase(au to ) = 'I') t h e n b e g i n writeC Do you wish t o b o o s t t h e A l c o n t e n t t o solve this? : '); readln(ans4); end; i f (UpCase(au to) = ' A ' ) t h e n ans4 := ' Y ' ; i f (UpCase(ans4) = ' Y ' ) and (count < 1) t h e n b e g i n count := count + 1; o u t i n d := ' # ' ; n o r m [ l , 2 ] := n o r m [ l , 2 ] + 0 . 5 ; writeC '); write (Output , ' ' ) ; write (Ou t f r ac , ' ' ) ; go to REPOCT; end; go to BOTTOM end; u n t i l c o n d i t i o n ; {***************************************************************************} { End o f h y p e r p l a n e n o r m a l i z a t i o n s e c t i o n . } {*************************************************************************** -{I***************************************************************************]. { Loop t o c a l c u l a t e t h e a c t i v i t i e s o f t h e micas o f i n t e r e s t . > {***************************************************************************} A c t T r i := 0 ; A c t D i := 0 ; A c t C e l := 0 ; A c t B r i t := 0 ; A c t P y p h := 0 ; A c t T a l c := 0 ; X E i g h t := 0 ; XSeven := 0 ; X S i x := 0 ; f o r i 4 := 1 t o nd im[4] do f o r i 3 := 1 t o nd im[3] do f o r i 2 := 1 t o nd im[2] do f o r i l := 1 t o n d i m [ l ] do b e g i n i f ( c h a r g e [ 4 , i 4 ] + c h a r g e [ 3 , i 3 ] + c h a r g e [ 2 , i 2 ] + c h a r g e [ l , i l ] = 8) t h e n X E i g h t := X E i g h t + a m a t [ i 4 , i 3 , i 2 , i l ] ; { C a l c u l a t e s t he } { t o t a l a c t i v i t y o f } { s p e c i e s w i t h charge } { o f 8. } Appendix B. Computer programs to calculate activities of mica components 236 i f (charge[4,i4]+charge[3,i3] +charge[2,i2]+charge[l,il] = 7) then XSeven := XSeven + amat [ i 4 , i 3 , i 2 , i l ] ; i f (charge[4,i4]+charge[3,i3] +charge[2,i2]+charge[l,il] = 6) then X S i x := X S i x + amat [ i 4 , i 3 , i 2 , i l ] ; i f i m a t [ i 4 , i 3 , i 2 , i l ] and (charge[4,i4]+charge[3,i3] +charge[2,i2] = 6) and (chargeCl.il] = 1) then begin i f (chargeC4,i4] = 2) and (chargeC3,i3] = 2 ) then A c t T r i := A c t T r i + amat [ i 4 , i 3 , i 2 , i l ] ; i f (chargeC4,i4] = 3) o r ( c h a r g e C 3 , i 3 ] = 3 ) o r (charge[2,i2] = 3 ) then A c t D i := A c t D i + amat [ i 4 , i 3 , i 2 , i l ] ; end; i f i m a t C i 4 , i 3 , i 2 , i l ] and (charge C4,i4]+charge[3,i3] +charge[2,i2] = 5) and (chargeCl.il] = 1) then A c t C e l := A c t C e l + amatCi4 , i3,i2 , i l ] ; { C a l c u l a t e s t h e } { t o t a l a c t i v i t y o f } { s p e c i e s w i t h charge } { o f 7 . } { C a l c u l a t e s t he } { t o t a l a c t i v i t y o f } { s p e c i e s w i t h charge } { o f 6 . } { Loop checks a l l } { s p e c i e s w i t h 6+ on } { o c t a h e d r a l s i t e s } { & 1+ on A s i t e . } { Subloop f o r t r i o c t . { micas { Subloop f o r d i o c t . •C micas { Loop checks a l l { s p e c i e s w i t h 5+ on { o c t a h e d r a l s i t e s { & 1+ i n A s i t e . { ( C e l a d o n i t e ) } } } } i f imatCi4,i3,i2,il] and (charge C4,i4]+charge C3,i3] +chargeC2,i2] = 6) and (chargeCl.il] = 2) then i f (chargeC4,i4] = 3) or (charge[3,i3] = 3) or (chargeC2,i2] =3) then ActBrit := ActBrit + amatCi4,i3,i2,il] i f imatCi4,i3,i2,il] and (charge C4,i4]+charge C3,i3] +chargeC2,i2] = 6) and (chargeCl.il] = 0) { Loop checks a l l } { s p e c i e s w i t h 6+ on } { two o c t a h e d r a l } { s i t e s & 2+ on A } { s i t e . ( B r i t t l e } { micas ) } { Loop checks a l l } { species with 6+ on } { octahedral s i t e s } { and 0 i n A s i t e . } Appendix B. Computer programs to calculate activities of mica components 237 t h e n b e g i n i f ( c h a r g e [ 4 , i 4 ] = 3) o r ( c h a r g e [ 3 , i 3 ] = 3) o r ( c h a r g e [ 2 , i 2 ] = 3) t h e n A c t P y p h := A c t P y p h + a m a t [ i 4 , i 3 , i 2 , i l ] ; i f ( c h a r g e [ 4 , i 4 ] = 2 ) and ( c h a r g e [ 3 , i 3 ] = 2) and ( c h a r g e [ 2 , i 2 ] = 2) t h e n A c t T a l c := A c t T a l c + a m a t [ i 4 , i 3 , i 2 , i l ] ; end ; end ; { Sub loop f o r d i o c t . } { m i c a ( p y r o p h y l l i t e ) } { Sub loop f o r t r i o c t . } { m i c a ( t a l c ) . } 0 c t 2 P l u s := no rm[2 ,5 ] + norm[2,6] + n o r m [ 2 , 7 ] ; A 2 P l u s ;= no rm[2 ,8 ] + n o r m [ 2 , 9 ] ; XFe := no rm[2 ,5 ] / 0 c t 2 p l u s ; XMg := no rm[2 ,6 ] / 0 c t 2 p l u s ; OHOcc := 1 - n o r m [ 2 , 1 2 ] / 2 - n o r m [ 2 , 1 3 ] / 2 ; A c t T r i * OHOcc *0H0cc; A c t D i * OHOcc *0H0cc; A c t C e l * OHOcc *0H0cc; A c t B r i t * OHOcc *0H0cc; * OHOcc *0H0cc; * OHOcc *0H0cc; A c t P y p h A c t T a l c 0 A c t T r i A c t D i A c t C e l A c t B r i t A c t P y p h A c t T a l c i f A 2 P l u s = t h e n b e g i n XCa := 0 ; XBa := 0 ; end e l s e b e g i n XCa := no rm[2 ,8 ] XBa := no rm[2 ,9 ] end; / A 2 P l u s ; / A 2 P l u s ; A l p l u s := norm[2 ,10] + n o r m [ 2 , 1 1 ] ; 0 c t 3 p l u s := O c t A l + n o r m [ 2 , 4 ] ; XK := n o r m [ 2 , l l ] / A l p l u s ; XNa := no rm[2 ,10] / A l p l u s ; MuscAct := XK * A c t D i ; := XNa * A c t D i ; * A c t B r i t ; * A c t B r i t ; XFe * XFe * XMg * XMg * P a r a A c t MargAct O e l l A c t AnnAct P h l A c t C e l A c t := XCa := XBa := XK * := XK * := XK * •C C a l c u l a t i o n o f new > { a c t i v i t i e s . } A c t T r i ; A c t T r i ; PyphAct XFe — 0 » XMg „ , XMg * ( O c t A l / O c t 3 p l u s ) * A c t C e l ; := ( O c t A l / 0 c t 3 p l u s ) * ( O c t A l / 0 c t 3 p l u s ) Appendix B. Computer programs to calculate activities of mica components 238 PhlActOld := norm[2,11] * (norm[2,6] / 3) * ActPyph; TalcAct := XHg * XMg * XHg * Ac tTa lc ; AnnActOld := norm[2,11] * (norm[2,5] / 3) * (norm[2,5] / 3) * (norm[2,5] / 3) * OHOcc * OHOcc; AnnAct01d2 := (norm[2,11] / ASiteOcc) * (norm[2,5] / OctSiteOcc) * (norm[2,5] / OctSiteOcc) * (norm[2,5] / OctSiteOcc) * OHOcc * OHOcc; * (norm[2,6] / 3) * (norm[2,6] / 3) * OHOcc * OHOcc; PhlAct01d2 := (norm[2,11] / ASiteOcc) * (norm[2,6] / OctSiteOcc) * (norm[2,6] / OctSiteOcc) * (norm[2,6] / OctSiteOcc) * OHOcc * OHOcc; MuscActOld := n o r m [ 2 , l l ] * (OctAl / OctSiteOcc) * (OctAl / OctSiteOcc) * OHOcc * OHOcc; ParaActOld := norm[2,10] * (OctAl / OctSiteOcc) * (OctAl / OctSiteOcc) * OHOcc * OHOcc; CelActOld := norm[2,11] * (norm[2,6] / 2) * (OctAl / OctSiteOcc) * OHOcc * OHOcc; MargActOld := norm[2,8] * (OctAl / OctSiteOcc) * (OctAl / OctSiteOcc) * OHOcc * OHOcc; Oe l lAc tOld := norm[2,9] * (OctAl / OctSiteOcc) * (OctAl / OctSiteOcc) * OHOcc * OHOcc; PredSi := 2 + 2 * XSix + XSeven; { C a l c u l a t i o n of } { a c t i v i t i e s using } { some ' o l d ' formulae } 0Si t0cc2 := norm[2,3] + norm[2,5] + norm[2,6] + O c t A l ; XMgl := norm[2 ,6] /0Si t0cc2 ; XFel := norm[2 ,5] /0Si t0cc2 ; X T i l := norm[2,3]/OSitOcc2; XA11 := 0 c t A l / 0 S i t 0 c c 2 ; XK1 := XK; XNal := XNa; XOHmod := SqRT(AnnAct / (XFel*XFel*XFel*XKl) ) ; X0Hmod2 := SQRT(MuscAc t / (XAl l*XAl l*XKl ) ) ; {***************************************************************************} -C Output. } •[***************************************************************************} i f (UpCase(ans3) = ' L ' ) then begin wr i te ln (Outpu t2 , ' C a l c u l a t e d a c t i v i t i e s : ' ) ; w r i t e l n ( ' C a l c u l a t e d a c t i v i t i e s : ' ) ; Appendix B. Computer programs to calculate activities of mica components w r i t e l n ( 0 u t p u t 2 , M u s c o v i t e M u s c o v i t e 0 u t p u t 2 , P a r a g o n i t e P a r a g o n i t e 0 u t p u t 2 , A n n i t e A n n i t e O u t p u t 2 , P h l o g o p i t e P h l o g o p i t e 0 u t p u t 2 , C e l a d o n i t e C e l a d o n i t e 0 u t p u t 2 , M a r g a r i t e M a r g a r i t e 0 u t p u t 2 , O e l l a c h e r i t e O e l l a c h e r i t e 0 u t p u t 2 , P y r o p h y l l i t e P y r o p h y l l i t e 0 u t p u t 2 , T a l c T a l c 0 u t p u t 2 ) ; ' , M u s c A c t : 7 : 5 ) ; ' , M u s c A c t : 7 : 5 ) ; ' , P a r a A c t : 7 : 5 ) ; ' , P a r a A c t : 7 : 5 ) ; ' , A n n A c t : 7 : 5 ) ; ' , A n n A c t : 7 : 5 ) ; \ P h l A c t : 7 : 5 ) ; * , P h l A c t : 7 : 5 ) ; ' , C e l A c t : 7 : 5 ) ; * , C e l A c t : 7 : 5 ) ; ' , M a r g A c t : 7 : 5 ) ; ' , M a r g A c t : 7 : 5 ) ; ' , 0 e l l A c t : 7 : 5 ) ; ' , 0 e l l A c t : 7 : 5 ) ; ' , P y p h A c t : 7 : 5 ) ; ' , P y p h A c t : 7 : 5 ) ; ' , T a l c A c t : 7 : 5 ) ; ' , T a l c A c t : 7 : 5 ) ; 0 u t p u t 2 , ' O l d ' ' f o r m u l a t i o n s o f a c t i v i t y : ' ) ; ' O l d ' ' f o r m u l a t i o n s o f a c t i v i t y : ' ) ; O u t p u t 2 , XA1**2} XA1**2> M u s c o v i t e {XK . M u s c o v i t e {XK . 0 u t p u t 2 , P a r a g o n i t e {XNa P a r a g o n i t e {XNa 0 u t p u t 2 , A n n i t e {XK A n n i t e {XK 0 u t p u t 2 , A n n i t e {XK A n n i t e {XK 0 u t p u t 2 , P h l o g o p i t e {XK P h l o g o p i t e {XK 0 u t p u t 2 , P h l o g o p i t e { X K . (Mg/SUM)**3> P h l o g o p i t e { X K . (Mg/SUM)**3> O u t p u t 2 , , M u s c A c t 0 1 d : 7 : 5 ) , M u s c A c t 0 1 d : 7 : 5 ) XA1**2> XA1**2} (Fe /3)**3> (Fe/3)**3> (Fe/SUM)**3> (Fe/SUM)**3> (Mg/3)**3} (Mg/3)**3> ' , P a r a A c t 0 1 d : 7 : 5 ) ; ' , P a r a A c t 0 1 d : 7 : 5 ) ; ' , A n n A c t 0 1 d : 7 : 5 ) ; ' , A n n A c t 0 1 d : 7 : 5 ) ; ' , A n n A c t 0 1 d 2 : 7 : 5 ) ; ' , A n n A c t 0 1 d 2 : 7 : 5 ) : ' , P h l A c t 0 1 d : 7 : 5 ) ; ' , P h l A c t 0 1 d : 7 : 5 ) ; ' , P h l A c t 0 1 d 2 : 7 : 5 ) : ' , P h l A c t 0 1 d 2 : 7 : 5 ) : Appendix B. Computer programs to calculate activities of mica components 240 ' C e l a d o n i t e {XK . XA1 . XMg} ' , C e l A c t 0 1 d : 7 : 5 ) ; w r i t e l n C C e l a d o n i t e {XK . XA1 . XMg} ' , C e l A c t 0 1 d : 7 : 5 ) ; w r i t e l n ( 0 u t p u t 2 , ' M a r g a r i t e {XCa . XA1 . XA1} ' , M a r g A c t 0 1 d : 7 : 5 ) ; w r i t e l n C M a r g a r i t e {XCa . XA1 . XA1} ' , M a r g A c t 0 1 d : 7 : 5 ) ; w r i t e l n ( 0 u t p u t 2 , ' O e l l a c h e r i t e {XBa . XA1**2} ' , 0 e l l A c t 0 1 d : 7 : 5 ) ; w r i t e l n C O e l l a c h e r i t e {XBa . XA1**2} ' , 0 e l l A c t 0 1 d : 7 : 5 ) ; w r i t e l n ( 0 u t p u t 2 ) ; w r i t e l n ; w r i t e l n ( 0 u t p u t 2 , ' P r e d i c t e d S i c o n t e n t = ' , P r e d S i : 5 : 3 ) ; w r i t e l n C P r e d i c t e d S i c o n t e n t = ' , P r e d S i : 5 : 3 ) ; w r i t e l n ( 0 u t p u t 2 , ' A c t u a l S i c o n t e n t = n o r m [ 2 , 1 ] : 5 : 3 ) ; w r i t e l n C A c t u a l S i c o n t e n t = norm[2 ,1 ] : 5 : 3 ) ; i f ( P r e d S i / n o r m [ 2 , l ] > MaxPercen t ) o r ( P r e d S i / n o r m [ 2 , l ] < M i n P e r c e n t ) t h e n b e g i n w r i t e l n ( 0 u t p u t 2 , ' *** Note S i d i s c r e p a n c y * * * ' ) ; w r i t e l n C *** Note S i d i s c r e p a n c y * * * ' ) ; end; i f ( o u t i n d = ' # ' ) t h e n b e g i n w r i t e l n ( 0 u t p u t 2 , ' *** A l has been augmented * * * » ) ; w r i t e l n C ' *** A l has been augmented * * * ' ) ; end; end e l s e b e g i n i f ( P r e d s i / n o r m [ 2 , 1 ] > M a x P e r c e n t ) o r ( P r e d S i / n o r m [ 2 , 1 ] < M i n P e r c e n t ) t h e n o u t i n d := ' * ' ; w r i t e l n ( O u t p u t , M u s c A c t : 7 : 4 , P a r a A c t : 7 : 4 , C e l A c t : 7 : 4 , A n n A c t : 7 : 4 , P h l A c t : 7 : 4 , M a r g A c t : 7 : 4 , O e l l A c t : 7 : 4 , P y p h A c t : 7 : 4 , T a l c A c t : 7 : 4 , o u t i n d ) ; w r i t e l n ( O u t F r a c , X M g l : 7 : 3 , X F e l : 7 : 3 , X T i l : 7 : 3 , X A l l : 7 : 3 , X K l : 7 : 3 , X 0 H m o d : 7 : 3 , X N a l : 7 : 3 , X 0 H m o d 2 : 7 : 3 , ' \ o u t i n d ) ; w r i t e l n ( M u s c A c t : 7 : 4 , P a r a A c t : 7 : 4 , C e l A c t : 7 : 4 , A n n A c t : 7 : 4 , P h l A c t : 7 : 4 , M a r g A c t : 7 : 4 , 0 e l l A c t : 7 : 4 , P y p h A c t : 7 : 4 , T a l c A c t : 7 : 4 , ' ' . o u t i n d ) ; end; {% + + + X + %t + + + + + + + X + + + + + + + + + + + + + + + + # + + + + )LjL + + **********************************} Appendix B. Computer programs to calculate activities of mica components 241 { C h e c k i n g o f c a l c u l a t e d a c t i v i t y v a l u e s a g a i n s t d i s t r i b u t i o n o f s p e c i e s e q u a t i o n s . } x c h e c k [ l ] x c h e c k [ 2 ] x c h e c k [ 3 ] xcheck[4] x c h e c k [ 5 ] = A l P l u s ; = A 2 P l u s ; = 0 c t 2 P l u s ; = no rm[2 ,3 ] ; = 0 c t 3 P l u s ; NonL i n C h e c k ( a m a t , i m a t , F N , n d i m , X C h e c k ) ; check := 0 ; f o r i := 1 t o 75 do i f ( F N C l . i ] = 0) and ( F N [ 2 , i ] = 0) t h e n check := check + 1 e l s e i f ( ( F N [ l , i ] <> 0) and ( F N [ 2 , i ] <> 0 ) ) and ( ( F N [ l , i ] / F N [ 2 , i ] > M i n P e r c e n t 2 ) and ( F N [ l , i ] / F N [ 2 , i ] < MaxPercen t2 ) ) t h e n check := check + 1 e l s e b e g i n w r i t e l n ( ' F ( l , ' , i : 2 , ' ) = ' ,FN[1 , i ] : 9 : 7 , ' F ( 2 , ' , i : 2 , > ) = ' , F N [ 2 , i ] : 9 : 7 , R a t i o = ' , F N [ l , i ] / F N [ 2 , i ] : 9 : 7 ) ; w r i t e l n ( O u t p u t l , ' F ( l , ' , i : 2 , ' ) = ' , F N [ 1 , i ] : 9 : 7 , F ( 2 , ' , i : 2 , ' ) = ' , F N [ 2 , i ] : 9 : 7 , ' R a t i o = ' , F N [ l , i ] / F N [ 2 , i ] : 9 : 7 ) end ; { C a l l t o e x t e r n a l } { u n i t wh ich r e t u r n s } { v a l u e s f o r LHS and } •[ RHS of d i s t r i b . o f } { s p e c i e s eqns . } { Check t h a t t he > { LHS ( F N [ l , i ] ) and } { RHS ( F N [ 2 , i ] ) o f > { d i s t r i b u t i o n o f } { s p e c i e s e q u a t i o n s } { a re e q u a l w i t h i n } { l i m i t s ( M i n P e r c e n t 2 } { and M a x P e r c e n t 2 ) . } i f check = 75 t h e n b e g i n w r i t e l n C D i s t r i b u t i o n o f s p e c i e s w r i t e l n ( O u t p u t 1 , ' D i s t r i b u t i o n o f end e l s e b e g i n w r i t e l n ( 7 5 - c h e c k : 2 , ' d i s t r i b u t i o n w r i t e l n ( O u t p u t 1 , 7 5 - c h e c k : 2 , ' d i s t r i b u t i o n end; e q u a t i o n s a re s a t i s f i e d . ' ) ; s p e c i e s e q u a t i o n s are s a t i s f i e d . ' ) ; o f s p e c i e s e q u a t i o n s v i o l a t e d . ' ) ; o f s p e c i e s e q u a t i o n s v i o l a t e d . ' ) ; Appendix B. Computer programs to calculate activities of mica components 242 BOTTOM: end; { End o f sample l o o p . } {**************************^ B0TT0M2: i f (UpCase(ans3) = ' L ' ) t h e n b e g i n w r i t e l n ( 0 u t p u t 2 ) ; w r i t e l n ; w r i t e l n ( 0 u t p u t 2 , . >); w r i t e l n C ' ) ; w r i t e l n ( 0 u t p u t 2 ) ; w r i t e l n ; end; i f (UpCase(ans3) = 'S>) t h e n b e g i n w r i t e l n ( O u t p u t ) ; w r i t e l n ( O u t f r a c ) ; w r i t e l n ( O u t p u t , ' " # " i n d i c a t e s .5 wt'/, A l added . ' ) ; w r i t e l n ( O u t p u t , ' »»*»> i n d i c a t e s p r e d i c t e d S i no t e q u a l t o a c t u a l S i . ' ) ; end; c l o s e ( O u t p u t ) ; c l o s e ( O u t p u t l ) ; c l o s e ( o u t p u t 2 ) ; c l o s e ( O u t F r a c ) ; { c l o s e ( F o r m u l a ) ; } END. { M I C a _ A c t i v i t y _ C a l c u l a t o r > B.5 Unit D I S T S P E G . P A S called by M I C A C . P A S UNIT D i s t S p e c ; {*************************^ { U n i t t o c a l c u l a t e v a l u e s f o r LHS and RHS o f 75 d i s t r i b u t i o n o f s p e c i e s e q u a t i o n s . E q u a t i o n s 1 - 6 are mass b a l a n c e e q u a t i o n s , 7 - 7 5 a re mass •0 =: [T]X JO =: [ ! ' Z ] i { -o. iaz o% senxBA } : Q = : [ T ' T ] i { 1 0 uo iq .BZTXBiq . tu i } u t l e q ° P SI oq. T =: T r o ? NI03H : 3 I A 8 : j p s u p ' xepux ' T ' l f ' S r ' S f ' f r f ' T T ' c ' T ' e T ' f r T HVA {***************************************************************************} N0IIVlN3W31dWI {********************************************************** : (gjJVy:X3S^DXW :seAl:mTaNW J S Z S ^ v y : i HVA ! XTJIITSWI : a«mrw HVA f x t J ^ W H : ^nn/W HVA)Xoe^ouTTUON 3HflQ3D0Hd {*******************************************^  : 3 I A H j o |>- -T]AVHHV = o e A l •TV3H JO [9*"I]AVHHV = S " V H •TV3H JO [SZ* 'T'Z' 'T]AVHHV = S Z S ^ V H !1V3H j o [92.' ' T] AVHHV = 3 Z " V H JNV3100H JO [ T P ^ I ' s p u i ' Z V U I ' 2 P U I ] A V H H V = XTJa*WI : 1V3H Jo [ T P U I ' 2 P U I ' S P U l ' 2 p u i ] A V H H V = xTJa«WH •v''I - SP«I • E * ' I = TP«I 3dAI {***************************************** **********************************} sssn -[***************************************************************************} 30ViH3XNI {***************************************************************************} { •suoxq.^nbe u o t a o t 2fZ s^uduodvaoo vowa jo SQifiAjfov d^vmoreo of snrejSojd Jd^ndxuoQ -Q xipuoddy Appendix B. Computer programs to calculate activities of mica components 244 end; index := 0; for i4 := 1 to mndim[4] do for i3 := 1 to mndim[3] do { Loading of 1-D } { with activities of } { permitted species } { (permits easier } { coding). } for i2 := 1 to mndim[2] do for i l := 1 to mndim[l] do i f Mimat[i4,i3,i2,i l] then begin index := index + 1; X[index] := Mamat[i4,i3,i2,il]; end; {****************************^  { Evaluation of LHS (F[ l , i ] ) and RHS (F[2,i]) for 75 equations. > {*****************************^  F [1,1] : = X [1] +X [2] +X [3] +X [4] +X [5] +X [6] +X [7] +X [8] +X [9] + X[10] +X[11] +X[12]+X[13]+X[14]+X[15]+X [16]+X[17]+X[18]+X [19] + X [20] +X [21] +X [22] +X [23] +X [24] +X [25] +X [26] +X [27] +X [28] +X [29] + X [30] +X [31] +X [32] +X [33] +X [34] +X [35] +X [36] +X [37] +X [38] +X [39] + X [40] +X [41] +X [42] +X [43] +X [44] +X [45] +X [46] +X [47] +X [48] +X [49] + X [50] +X[51] +X[52]+X[53]+X[54]+X [55]+X [56]+X [57]+X[58]+X [59] + X [60] +X [61] +X [62] +X [63] +X [64] +X [65] +X [66] +X [67] +X [68] +X [69] + X [70] +X [71] +X [72] +X [73] +X [74] +X [75] ; F[2,l] := 1; F[l,2] := X[2]+X[5]+X[9]+X[ll]+X[15]+X[18]+X[22]+X[24]+X[26] + X [28] +X[31] +X[34]+X[38]+X[40]+X[43]+X[45]+X[49]+X[52]+X [56] + X[58] +X[62]+X[64]+X[67]+X[69]+X[72]; F[2,2] := MXcheck[l]; F[l,3] := X[6]+X[12]+X[14]+X[19]+X[25]+X[29]+X[32]+X[35]+X[41] + X [46] +X[48]+X[50]+X[53]+X[54]+X[59]+X [61]+X[65]+X[68]+X [70] + X[73]+X[74]+X[75] ; F[2,3] := MXCheck[2] ; F[l,4] := 2*X[4]+X[5]+X[6]+X[7]+X[11]+X[12]+X[13]+X[17]+X[21] + 2*X[22]+2*X[23]+X[24]+X[25]+X[31]+X[32]+2*X[33]+X[34] + X[35]+X [36]+X [37]+2*X [38]+2*X [39]+X[40]+X[41]+2*X[42] + 2*X[43]+2*X[44]+3*X[45]+3*X[46]+3*X[47]+2*X[48]+X[49]+ X[50]+X[51]+X[52]+X[53]+2*X[54]+X [58]+X [59]+X[60]+X[67] + X[68]+X[69]+X [70]+X [71]+X [72]+X [73]+2*X[74]; F[2,4] := MXCheck[3] ; F [1,5] : = 2*X [1] +X [2] +X [3] +X [4] +X [5] +X [6] +X [7] +2*X [8] +X [9] +X [10] + X [11] +X [12] +X [13] +X [14] +X [15] +X [16] +X [26] +X [27] +X [33] +X [34] + X[35] +X[36]+X[42]+X [49]+X [50]+X [51]+2*X [55]+X[56]+X[57]+X[58] + X [59] +X [60] +X [61] +X [62] +X [63] +X [69] +X [70] +X [71] +X [75] ; F[2,5] := MXCheck[4]; F [1,6] : = X [2] +X [3] +X [9] +X [10] +X [15] +X [16] +2*X [17] +2*X [18] + 2*X[19] +2*X[20]+2*X[21]+X[22]+X[23]+X[24]+X[25]+X[26]+X[27] + 2*X[28]+2*X[29]+2*X[30]+X [31]+X[32]+2*X[37]+X[38]+X[39] + Appendix B. Computer programs to calculate activities of mica components 245 X [40] +X [41] +X [43] +X [44] +X [52] +X [53] +X [56] +X [57] +X [62] +X [63] + 2*X[64]+2*X[65]+2*X[66]+X[67] +X[68]+X[72]+X[73]; F !2, 6] = MXCheck F : i , 7] • X[ l ] ; F 12 7] = X[8]; F ; i 8] = X[2]; F [2 8] • X[9]; F ; i 9] = X[3]; F 12 9] = X[10]; F ; i 10] = X[5]; F \2 10] = X [ l l ] ; F : i 11] = X[6]; F [2 11] = X[12] ; F : i 12] = X[7]; F 12 12] = X[13]; F : i 13] = X[2] ; F 12 13] = X[15] ; F :i 14] X[3]; F .2 14] = X[16]; F :i 15] = X[17] ; F .2 15] SS X[21] ; F [i 16] = X[2]; F .2 16] = X[26] ; F :i 17] = X[3] ; F \2 17] = X[27] ; F :i 18] = X[18] ; F .2 ,18] SS X[28] ; F :i ,19] SS X[19]; F .2 ,19] s s X[29] ; F :i 20] = X[20] ; F .2 ,20] = X[30] ; F :i ,21] s s X[24]; F .2 ,21] - X[31] ; F :i 22] s s X[25]; F .2 ,22] = X[32] ; F ;i ,23] s s X[4]; F .2 ,23] = X[33] ; F ;i ,24] X[5]; F .2 ,24] = X[34] ; F ;i ,25] S3 X[6]; F .2 ,25] = X[35]; F ;i ,26] S3 X[7]; F \2 ,26] S3 X[36] ; F ;i ,27] = X[17]; F .2 ,27] S3 X[37]; F ;i ,28] S3 X[22] ; F \2 ,28] S3 X[38] ; F ; i ,29] S3 X[23] ; F \2 ,29] • S 3 X[39] ; F ; i ,30] = X[24]; Appendix B. Computer programs to calculate activities of mica components 246 F \2 30] = X[40] ; F ; i 31] = X[25] ; F \2 31] = X[41]; F : i 32] = X[4] ; F 12 32] = X[42]; F : i 33] = X[22]; F 12 ,33] = X[43]; F : i ,34] = X[23] ; F [2 ,34] = XC44] ; F : i 35] = X[5]; F 12 ,35] = X[49] ; F :i ,36] = X[6]; F .2 ,36] = X[50]; F r.i ,37] = X[7]; F 12 ,37] = X[51] ; F .1 ,38] = X[24] ; F 12 ,38] = X[52] ; F : i ,39] = X[25] ; F .2 ,39] X[53] ; F :i ,40] = X[48] ; F .2 ,40] = X[54] ; F :i ,41] X[l]; F .2 .41] = X[55]; F :i ,42] X[2]; F .2 ,42] = X[56] ; F .i ,43] = X[3]; F .2 .43] X[57]; F ;i ,44] = X[5]; F .2 ,44] = X[58] ; F :i ,45] = X[6]; F .2 ,45] = X[59] ; F :i ,46] = XC7]; F .2 ,46] = X[60] ; F :i ,47] X[14]; F .2 ,47] = X[61] ; F :i ,48] = X[2] ; F .2 ,48] = X[62]; F ;i ,49] = X[3] ; F .2 ,49] = X[63] ; F ;i ,50] = X[18]; F .2 ,50] = X[64] ; F ;i ,51] = X[19] ; F .2 ,51] = X[65] ; F :i ,52] = X[20] ; F .2 ,52] = X[66] ; F :i ,53] X[24]; F .2 ,53] = X[67] ; F :i ,54] = X[25] ; F .2 ,54] = X[68] ; F ;i ,55] = X[5] ; Appendix B. Computer programs to calculate activities of mica components 247 F 12 55] = X[69] ; F n 56] = X[6]; F 56] = X[70]; F ; i 57] = X[7]; F .2 57] = X [ 7 l ] ; F II 58] = X[24]; F [2 58] = X[72]; F : i 59] = X[25] ; F [2 59] = X[73]; F [1 60] = X[48]; F [2 ,60] = X[74]; F II ,61] s XC14]; F 12 61] = X[75]; F II 62] = X[1]*X[20] ; F 12 ,62] = X[3]*X[3] ; F : i ,63] = X [1] *X [1] *X [1] *X [17] *X [17] *X [45] *X [45] ; F .2 ,63] = X [4] *X [4] *X [4] *X [4] *X [2] *X [2] *X [20] ; F : i ,64] = X [1] *X [2] *X [2] *X [20] *X [45] *X [45] ; F 12 ,64] = X [5] *X [5] *X [5] *X [5] *X [17] *X [17] ; F : i ,65] = X [1] *X [1] *X [1] *X [20] *X [20] *X [20] *X [45] *X [45] ; F 12 ,65] = X [7] *X [7] *X [7] *X [7] *X [2] *X [2] *X[17] *X [17] ; F : i ,66] = X[6]*X[20] ; F .2 ,66] = X[14] *X[17]; F : i ,67] = X[2]*X[2]*X[20] ; F 12 ,67] = X[18] *X[18]*X[1] ; F : i ,68] = X [2] *X [2] *X [6] *X [6] *X [6] *X [6] *X [17] *X [17] *X [20] ; F 12 ,68] = X [19] *X [19] *X [19] *X [19] *X [1] *X [1] *X [1] *X [45] *X [45] ; F : i ,69] = X [2] *X [2] *X [17] *X [17] *X [45] *X [45] ; F 12 ,69] = X [22] *X [22] *X [22] *X [22] *X [1] *X [20] ; F : i ,70] = X [1] *X [17] *X [17] *X [20] *X [45] *X [45] ; F 12 ,70] = X [23] *X [23] *X [23] *X [23] *X [2] *X [2] ; F : i ,71] = X [2] *X [2] *X [20] *X [20] *X [20] *X [45] *X [45] ; F 12 ,71] = X [24] *X [24] *X [24] *X [24] *X [1] *X [17] *X [17] ; F ; i ,72] = X[6]*X[6]*X[20] ; F 12 ,72] = X[25]*X[25]*X[1] ; F II ,73] = X [6] *X [6] *X [6] *X [6] *X [17] *X [17] *X [45] *X [45] ; F 12 ,73] = X [46] *X [46] *X [46] *X [46] *X [1] *X [2] *X [2] *X [20] ; F II ,74] = X [1] *X [20] *X [45] *X [45] ; F 12 ,74] = X [47] *X [47] *X [2] *X [2] ; F II ,75] = X [6] *X [6] *X [6] *X [6] *X [20] *X [20] *X [20] *X [45] *X [45] ; F 12 ,75] = X [48] *X [48] *X [48] *X [48] *X [1] *X [2] *X [2] *X [17] *X [17] ; END; END. { D i s t S p e c } -[************************** *************************************************} Appendix B. Computer programs to calculate activities of mica components 248 B.6 Input files required by M I C A C . P A S B.6.1 Element weights ( E L E M E N T . D A T ) MICAC requires the file ELEMENT.DAT which contains the following element weights, con-version factors to convert oxide weights to element weights (or vice versa), valences, and the element symbol. 28.0855 2.13931 4 Si 26.98154 1.88947 3 Al 47.90000 1.66803 4 Ti 55.84700 1.42973 3 Fe 55.84700 1.28649 2 Fe 24.30500 1.65809 2 Mg 54.93800 1.29123 2 Mn 40.08000 1.39919 2 Ca 137.33000 1.11649 2 Ba 22.98977 1.34797 1 Na 39.09830 1.20459 1 K 18.99840 1.00000 1 F 35.45300 1.00000 1 Cl B.6.2 Input analyses (*.IN) The file of analytical data must have the extension .IN. The program asks for the root (e.g., EXAMPLE) and will look for the file (EXAMPLE.IN). Output files are generated with the same root but various extensions (see below). Analytical data can be input in 3 formats, element weight percents, oxide weight percent, and molecular formula (the molecular formula must be calculated on the basis of 22 positive charges). Each input data point occupies two lines in the input file, the first line contains the analysis label the second contains the weight percents or cation/anion numbers. Elements/oxides/cations-anions must be in the same order as indicated in ELEMENT.DAT (i.e., Si, Al, Ti, Fe3+, Fe2+, Mg, Mn, Ca, Ba, Na, K, F, Cl). The data point label can have a maximum of 14 characters. Anything after these is ignored by the program. The weight percents (or cation/anion numbers) can be given to any number of decimal places and can occur on more than a single line in the input file. The progam searches for 13 consecutive numbers, therefore elements in the above list which were not analysed or which fall below detection limit must be given as zero. Note that Pascal code requires real numbers be given with digits before and after the decimal point (.5 and 5. are both unacceptable and will result in a runtime error). The following examples show the same analysis in all three formats. a). Input as element weight percent. Appendix B. Computer programs to calculate activities of mica components 249 EXAMPLE-1 17.97 10.22 0.964 0.132 14.76 5.752 0.215 0.059 0.225 0.210 7.132 0.301 0.024 b). Input as oxide weight percent. EXAMPLE-1 38.44 19.31 1.608 0.189 18.99 9.537 0.278 0.083 0.251 0.283 8.591 0.301 0.024 c). Input as molecular formula (22 positive charges). EXAMPLE-1 2.820 1.670 0.089 0.010 1.165 1.043 0.017 0.006 0.007 0.040 0.804 0.070 0.003 B.7 Output files generated by MICAC.PAS MICAC generates 4 different output files. Activity for various mica species are written to one of two files. The program asks whether a short or long output is required. Short output, which gives the datum label and values for activity for 9 selected micas, is written to the file with the extension .SHT (e.g., EXAMPLE.SHT). Output in EXAMPLE.SHT has the following form: Sample Muse. Para. Cela. Anni. Phlo. Marg. Oell . Pyph. Talc. EXAMPLE-1 0.0110 0.0006 0.0396 0.0486 0.0349 0.0001 0.0001 0.0016 0.0054 '#' indicates .5 wt'/, Al added. '* ' indicates predicted Si not equal to actual S i . Long output includes 'new' activities and a variety of values for activity calculated using 'old' activity models and are written to the file with extension .LON (e.g., EXAMPLE.LON). Output in EXAMPLE.LON has the following form: EXAMPLE-1 Calculated activit ies: Muscovite 0.01100 Paragonite 0.00055 Annite 0.04862 Phlogopite 0.03491 Celadonite 0.03963 Margarite 0.00010 Oellacherite 0.00011 Appendix B. Computer programs to calculate activities of mica components 250 P y r o p h y l l i t e 0 .00163 T a l c 0 .00540 ' O l d ' f o r m u l a t i o n s o f a c t i v i t y : M u s c o v i t e {XK . XA1**2} 0.02264 P a r a g o n i t e {XNa . XA1**2} 0.00113 A n n i t e {XK . (Fe /3)**3> 0.04373 A n n i t e {XK . (Fe/SUM)**3> 0.06169 P h l o g o p i t e {XK . (Mg/3)**3> 0.03140 P h l o g o p i t e { X K . (Mg/SUM)**3> 0 .04429 C e l a d o n i t e {XK . XA1 . XMg} 0 .06782 M a r g a r i t e {XCa . XA1 . XA1} 0.00018 O e l l a c h e r i t e {XBa . XA1**2} 0 .00020 P r e d i c t e d S i c o n t e n t = 2 .822 A c t u a l S i c o n t e n t = 2 .820 MICAC also writes out modified values of mole fractions to be used with program PTAX of Berman et a/.(1987). These values are written to the file with extension .MLF (e.g., EX-AMPLE.MLF). These modified values are entered in '*.CMP' files used as input to PTAX and result in PTAX using the ideal activities defined by MICAC. Output in EXAMPLE.MLF has the following form: Sample [-XMg-] [ -XFe- ] [ - X T i - ] [ -XA1- ] [ - X K - ] [XOHbi] [XNamu] [XOHmu] EXAMPLE-1 0 .374 0 .418 0 .032 0 .176 0 .952 0 .836 0 .048 0 .611 NONLIN.DAT contains the data labels and the status of the check of the calculated ac-tivity values with the distribution of species equations. If the equations were satisfied a single statement to that effect follows the label. If some of the equations are not satisfied, a list of the values for the LHS and RHS for each violated equation is presented along with the ra-tio LHS/RHS. Ideally LHS/RHS should be unity. In MICAC, a value for LHS/RHS between MinPercent2 (0.9925) and MaxPercent2 (1.0075) is acceptable. This is because MICAC only forces hyperplane sums (see text) to between MinPercent (0.995) and MaxPercent (1.005) of the theoretical value. The closer MinPercent and MaxPercent are to unity the closer MinPercent2 and MaxPercent2 can be to unity. In addition 14 of the distribution of species equations are nonlinear with some terms to the fourth power. If some of the activities are small (such as the margarite and oellacherite activities in the example) then the LHS and RHS of the equations will become extremely small and the ratio, LHS/RHS, becomes more subject to error. Output in NONLIN.DAT has either the following form: Appendix B. Computer programs to calculate activities of mica components 251 EXAMPLE-1 Distribution of species equations are satisfied, or the form: EXAMPLE-1 F(l,63) = 0.0000000 F(2,63) = 0.0000000 Ratio = 1.0050312 F(l,65) = 0.0000000 F(2,65) = 0.0000000 Ratio = 1.0050312 F(l,68) = 0.0000000 F(2,68) = 0.0000000 Ratio = 0.9949939 3 distribution of species equations violated. In the second case the it is clear that equations 63, 65, and 68 involve powers of very small numbers but their ratios are still very close to the required constraints. (The values of MaxPer-cent2 and MinPercent2 were tightened to MaxPercent and MinPercent to produce the second output above.) Appendix C Calculated pressures and temperatures using calibrations of S G A M and GASP 252 Appendix C. Calculated pressures and temperatures from published analyses 253 SGAM GASP This study.* H&C (1985).f HDH (1988)4 This study.* Sample # P T P T P T P T Holdaway et al. (1988) 4a 4024 572 3532 555 2631 568 5211 578 47 4151 566 3406 549 3071 578 4099 566 19 3675 556 3194 552 2944 589 4012 558 63 2977 568 2763 558 2232 610 3885 573 30 2832 560 2735 559 2699 616 3409 563 8 4631 621 4998 611 4056 632 4634 621 140 3908 591 3973 590 3643 632 4719 595 77-3 4514 596 4793 597 4129 619 4496 596 77-2 4522 607 4759 607 4058 630 4693 608 76 3948 597 4540 607 3976 641 3647 596 56 3837 595 4073 593 3575 641 3799 595 91 3550 611 4669 628 3900 669 4799 617 73 3431 587 4225 602 3655 640 3663 588 87 4998 643 5755 629 4274 657 4206 639 86 5676 647 7013 657 6026 675 5299 645 * Calculated using models as given in text {Calculated using the calibration of Hodges and Crowley (1985). ^Calculated using the calibration of Holdaway et al. (1988). Table C. l : Pressures (bars) and temperatures (°C) calculated for data of Holdaway et al. (1988) using SGAM and GASP/garnet-biotite. SGAM calibrations are from this study, Hodges and Crowley (1985) and Holdaway et al. (1988). GASP/garnet-biotite conditions calculated using Fuhrman and Lindsley's (1988) model for plagioclase, gar-net model of Berman (1990), and biotite model from this study. Appendix C. Calculated pressures and temperatures from published analyses 254 SGAM GASP This s tudy.* H&C (1985).t HDH (1988)4 This study.* Sample # P T P T P T P T Pigage (1982) 373 4864 577 - - - - 5999 583 121 5094 566 - - - - 4794 565 367 5149 577 3631 525 2395 560 6078 581 82 5196 554 3710 520 3103 550 5053 553 398 4591 555 3687 535 3092 552 4725 555 492 5455 561 3968 534 3494 555 4805 557 223 5350 564 3834 524 3191 555 4730 560 2-376 5615 588 4416 549 3551 582 6260 592 2-13 4532 555 3062 518 2212 552 5062 558 74 5257 608 4390 571 2225 552 6213 612 59 5555 582 4517 550 3717 576 5666 582 40 5693 563 4153 534 3631 555 4624 558 Hodges anc Spear (1982) 78b 3716 477 1845 466 1863 480 4453 481 80d 3769 525 2098 495 1551 520 3874 525 90a 3770 498 2123 486 1996 496 4426 501 92d 3985 501 2275 491 2123 502 4558 504 145e 3107 505 1836 501 1434 511 4301 511 146d 3565 480 1736 466 1593 469 2524 475 * Calculated using models as given in text. f Calculated using the calibration of Hodges and Crowley (1985). ^Calculated using the calibration of Holdaway et al. (1988). Table C.2: Pressures (bars) and temperatures (°C) calculated for data of Pigage (1982) and Hodges and Spear (1982) using SGAM and GASP/garnet-biotite. SGAM cali-brations are from this study, Hodges and Crowley (1985) and Holdaway et al. (1988). GASP/garnet-biotite conditions calculated using Fuhrman and Lindsley's (1988) model for plagioclase, garnet model of Berman (1990), and biotite model from this study. Appendix C. Calculated pressures and temperatures from published analyses 255 Temperatures Sample # This study TH GA1 GA2 FS PL GS IM1 IM2 BL1083C 744 717 631 656 790 697 548 683 673 BL1083E 763 697 617 604 761 684 685 675 698 DS2183B 622 608 553 553 632 621 592 562 581 DS2483 576 562 520 520 569 587 565 513 558 LS7884 677 654 586 562 697 654 588 595 587 SE0985A 702 675 601 615 728 669 533 626 619 RF2283 686 662 591 582 708 659 583 612 598 RF2583A 663 644 578 578 683 647 558 599 595 VM2683A 713 688 610 622 746 678 602 652 640 VR1383A 701 676 601 595 729 670 640 631 628 HS1483C 691 663 592 569 710 661 566 602 588 HS1583C 784 716 630 635 788 697 700 688 733 Table C.3: Temperatures calculated for 6 pairs of samples from data of Chipera and Perkins (1988). Sample pairs are from the same or adjacent localities. Temperatures using other calibrations are from Chipera and Perkins (1988, table 3). Calibrations: TH = Thompson (1976b); GA1 = Goldman and Albee (1977), second parameter so-lution; GA2 = Goldman and Albee (1977), fifth rank solution; FS = Ferry and Spear (1978); PL = Perchuk and Lavrent'eva (1983); GS = Ganguly and Saxena (1984); IM1 = Indares and Martignole (1985), thermodynamic data only; IM2 = Indares and Mar-tignole (1985), using both thermodynamic and experimental data. See Chipera and Perkins (1988) for further details. Appendix D Microprobed samples: Petrography, mineral analyses, and calculated pressures and temperatures D . l Introduction This appendix presents summary petrographic information for each of the microprobed speci-mens. They are keyed to the map using the index numbers. Low index numbers are assigned to samples from the north of the area and increase southwards. The sample number is given with the map index number and the original source is given. The sample numbers are pre-ceded by the initials of the original collector for easy reference (e.g. JSG samples collected by J.S. Getsinger, 1985). Samples collected by the author have sample numbers starting with the initials 'DM'. Also given for each sample is a triplet of garnet, biotite, and muscovite analyses. These triplets are those that yield the 'typical' pressures and temperatures noted in table 5.2. For samples which yielded clusters of points in P-T space, an average P and T is presented in table 5.1 and the 'typical' P and T is a determination close to that average. One sample (#9: CJNF-9) example yielded two populations of P's and T's and a 'typical' P and T for each population is given in table 5.1. The triplet of analyses that yield each of these 'typical' P's and T's are given below. As discussed in chapter 5, the P's and T's determined for some samples define lines in P-T space and an average P and T was not determined in these cases. The two 'typical' P's and T's given in table 5.1 are individual P-T determinations from the extremes of these lines of P's and T's (see chapter 5). D.2 Summary petrography and selected microprobe analyses Electron microprobe analyses of garnet, biotite and muscovite were done on the Cameca SX-50 microprobe at U.B.C. using standard techniques and data reduction routines. An accelerating potential of 15 kV and a beam current of 20 nA were used. The spot size for garnet was 2 /xm, and for biotite and muscovite was 10 fim. The elements analyzed, the standards used, the nature of each standard, the samples for which each standard was used, and detection limits for each element axe given in table D.l. Details of the compositions of the standards are given in the U.B.C. catalog of microprobe standards.. 256 Appendix D. Microprobed samples: Petrography, analyses, P and T 257 Elt. Std # Compound Samples Det. lim. type analysed (El. wt. %) Gaxnet Si S235 Garnet #1-8 0.02 S007 Garnet #9 - 35 Ti S013 Oxide #1-35 0.03 Al S278 Garnet #1-8 0.02 S007 Garnet #9 - 35 Mg S235 Garnet #1-35 0.015 Fe S384 Garnet #1-8 0.05 S237 Olivine #9-35 Mn S015 Garnet #1-35 0.05 Ca S007 Garnet #1-35 0.03 Micas Si S164 Muscovite #1-35 0.02 Al S164 Muscovite #1-35 0.02 Ti S013 Oxide #1-35 0.03 Fe S082 Biotite #1-35 0.04 Mg S024 F-Phlogopite #1-35 0.015 Mn S015 Garnet #1-35 0.05 Cr S222 Oxide #1-35 0.05 K S024 F-Phlogopite #1-35 0.03 Na S229 Amphibole #1-35 0.016 Ca S229 Amphibole #1-35 0.025 Ba S016 Sulphate #1-35 0.12 F S024 F-Phlogopite #1-35 0.15 Cl S286 Halide #1-35 0.03 Table D.l: Elements analyzed, standard number (U.B.C. microprobe catalog), the standard nature (rough mineral name), the unknown sample (map index #) for which each standard was used, and detection limits for each element. Appendix D. Microprobed samples: Petrography, analyses, P and T 258 El. Wt. % Garnet Biotite Muscovite Si 17.161 17.278 21.232 Al 11.190 9.448 18.595 Ti b.d. 0.887 0.228 Fe 25.872 12.508 0.761 Mg 0.977 7.349 0.365 Mn 4.936 b.d. b.d. Cr n.a. b.d. 0.095 Ca 0.679 0.117 b.d. Ba n.a. b.d. 0.804 K n.a. 6.915 7.331 Na n.a. 0.145 1.111 F n.a. 0.801 b.d. Cl n.a. 0.042 b.d. 0 39.744 38.307 43.329 Total 100.559 93.797 93.851 Table D.2: Selected garnet, biotite, and muscovite analyses for sample JSG-81-234 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d. = below detection limit given in table D.l) D.2.1 #1: JSG-81-234 (Getsinger, 1985) Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Graphite(?). Porphyroblastic phases: Garnet, Staurolite, Kyanite, Plagioclase. Textures: All porphyroblastic phases, except garnet extremely poikiloblastic. Internal folia-tions folded (Si and S2) and continuous with the external foliation (S2). S2 foliation in matrix wraps around garnet but not other phases indicating garnet porphyroblasts earlier. Some mi-nor crenulation of external fabric. Extremely poikiloblastic plagioclase appears to have sector zoning and is very similar to cordierite but cordierite not found during qualitative microprobe analyses. Retrogression: None. D.2.2 #2: JSG-81-278 (Getsinger, 1985) Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Sillimanite, Andalusite, Stauro-lite, Plagioclase, Ilmenite, Rutile, Tourmaline. Porphyroblastic phases: Garnet, Staurolite, Kyanite, Andalusite. Textures: Only garnet and andalusite are poikiloblastic but inclusions in garnet (mainly Appendix D. Microprobed samples: Petrography, analyses, P and T 259 El. Wt. % Garnet Biotite Muscovite Si 16.834 16.499 21.838 Al 11.070 9.982 18.624 Ti b.d. 0.925 0.445 Fe 26.736 14.620 0.956 Mg 1.345 6.056 0.444 Mn 1.492 b.d. b.d. Cr n.a. b.d. b.d. Ca 2.326 b.d. b.d. Ba n.a. b.d. 0.188 K n.a. 7.359 7.855 Na n.a. 0.204 0.984 F n.a. 0.404 b.d. Cl n.a. b.d. b.d. 0 38.849 37.896 44.286 Total 98.652 93.945 95.620 Table D.3: Selected garnet, biotite, and muscovite analyses for sample JSG-81-278 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d. = below detection limit given in table D.l) quartz) not oriented. Inclusions in andalusite (mainly muscovite) are parallel'to external foli-ation (S2) (see photograph, fig. 5.18). Foliation in matrix wraps around garnet but not other porphyroblasts indicating garnet growth earlier. Rutile only as inclusions in staurolite and kyanite. Bmenite in matrix and as inclusions in andalusite. Also andalusite seen partly includ-ing kyanite grain. Implies andalusite latest aluminosilicate polymorph to grow. Retrogression: None. D.2.3 #3: J S G - 8 0 - 3 0 (Getsinger, 1985) Assemblage: Quartz, Muscovite, Biotite, Garnet, Sillimanite, Staurolite, Ilmenite, Chlorite. Porphyroblastic phases: None. Textures: The rock is coarse grained with the micas reaching 3-5 mm in length. Thus garnet and staurolite with similar grain sizes are not porphyroblastic. The rock is highly aluminous and the garnet and staurolite lack any internal foliations. The coarse grained mica defines a transposed foliation in which the micas have two orientations denning a well developed herring-bone pattern. The foliation is considered to be S3 because it postdates most of the mineral growth. The earlier foliation is synchronous with or slightly predates the garnet and staurolite formation (thus it is S2). Retrogression: Very slight, chlorite after biotite and perhaps garnet. Appendix D. Microprobed samples: Petrography, analyses, P and T 260 El. Wt. % Garnet Biotite Muscovite Si 16.898 16.539 21.653 Al 11.391 10.076 18.976 Ti b.d. 0.815 0.221 Fe 28.090 16.354 0.807 Mg 1.354 5.313 0.324 Mn 1.535 b.d. b.d. Cr n.a. b.d. b.d. Ca 1.043 b.d. b.d. Ba n.a. b.d. 0.270 K n.a. 7.674 7.720 Na n.a. 0.133 1.007 F n.a. 0.293 b.d. Cl n.a. 0.029 b.d. 0 39.253 38.045 44.090 Total 99.564 95.271 95.068 Table D.4: Selected garnet, biotite, and muscovite analyses for sample JSG-80-30 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) D.2.4 #4: DM-85-9 A s s e m b l a g e : Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Chlorite, Sericite. P o r p h y r o b l a s t i c p h a s e s : Kyanite, Staurolite, Plagioclase. T e x t u r e s : The rock is extremely coarse grained and comes from the margin of a quartz-kyanite vein. Kyanite grains reach 6 cm (2 cm is common). There are no internal textures in any of the porphyroblastic phases. The foliation is crude (due to grain size) and a single mica cleavage shows some buckling and kyanite grains are kinked (F4). Garnet occurs as peculiar coronas around plagioclase 'porphyroblasts' and as inclusions in kyanite blades. Microprobe analysis indicates that garnet is very inhomogeneous (see chapter 5). R e t r o g r e s s i o n : Slight to moderate and variable. Plagioclase is variably sericitised. Kyanite and staurolite commonly have sericite mantle and along fractures. Biotite and garnet variably chloritized (some almost pristine). D.2.5 #5: PDL-447 (Lewis, 1987) A s s e m b l a g e : Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Sericite, Chlorite. P o r p h y r o b l a s t i c phase s : Garnet, Staurolite, Plagioclase, Biotite. Appendix D. Microprobed samples: Petrography, analyses, P and T 261 El. Wt. % Garnet Biotite Muscovite Si 17.010 16.669 21.860 Al 11.591 9.808 18.554 Ti 0.031 0.786 0.222 Fe 25.720 13.733 0.826 Mg 1.937 6.779 0.406 Mn 3.157 b.d. b.d. Cr n.a. b.d. b.d. Ca 0.374 b.d. b.d. Ba n.a. b.d. 0.140 K n.a. 7.601 7.741 Na n.a. 0.137 1.158 F n.a. 0.444 b.d. Cl n.a. 0.045 b.d. 0 39.721 38.066 44.076 Total 99.541 94.068 94.983 Table D.5: Selected garnet, biotite, and muscovite analyses for sample DM-85-9 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d. = below detection limit given in table D.l) El. Wt. % Garnet Biotite Muscovite Si 17.184 16.715 21.396 Al 11.171 9.777 18.767 Ti b.d. 0.835 0.232 Fe 27.044 14.944 0.742 Mg 2.016 6.554 0.351 Mn 2.441 0.065 b.d. Cr n.a. b.d. 0.052 Ca 0.056 b.d. b.d. Ba n.a. b.d. 0.216 K n.a. 7.812 7.571 Na n.a. 0.100 1.273 F n.a. 0.402 b.d. Cl n.a. b.d. b.d. 0 39.587 38.386 43.693 Total 99.499 95.590 94.293 Table D.6: Selected garnet, biotite, and muscovite analyses for sample PDL-447 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Appendix D. Microprobed samples: Petrography, analyses, P and T 262 El. Wt. % Garnet Biotite Muscovite Si 17.059 17.157 21.695 Al 11.432 9.836 18.754 Ti b.d. 1.127 0.253 Fe 25.736 13.997 0.656 Mg 1.883 6.453 0.328 Mn 3.223 0.094 b.d. Cr n.a. b.d. b.d. Ca 0.700 b.d. b.d. Ba n.a. b.d. 0.244 K n.a. 7.208 7.235 Na n.a. 0.197 1.390 F n.a. 0.335 b.d. Cl n.a. b.d. b.d. 0 39.712 38.755 43.964 Total 99.745 95.159 94.519 Table D.7: Selected garnet, biotite, and muscovite analyses for sample DM-85-42 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Textures: The rock has a spectacularly displayed crenulation cleavage. Relict hinges exist be-tween transposed earlier foliation. The presence of sparse opaque inclusions in porphyroblasts parallel to the older foliation in the matrix indicates that the crenulation cleavage is S3. This is supported by the fact that many of the mica flakes in the relict hinge zones and the biotite porphyroblasts are kinked and thus the crenulation cleavage postdates the peak of metamor-phism. Retrogression: Slight to moderate and variable (less than #4). Staurolite has sericite as thin mantles and along fractures. Plagioclase slightly sericitized. Slight chloritization of garnet and biotite. D.2.6 #6: DM-85-42 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Plagioclase, Ilmenite, Rutile, Tour-maline, Chlorite. Porphyroblastic phases: Garnet, Kyanite. Textures: The foliation external to the porphyroblasts is isoclinally folded. The earlier fo-liation is continuous with the foliation internal to kyanite porphyroblasts. The later foliation wraps around the kyanite and has chlorite laths growing in the axial planes. The later foliation is post metamorphic (peak) and considered to be S3. The earlier foliation is S2 (see photograph, fig. 5.16). Rutile only as inclusions in kyanite. Appendix D. Microprobed samples: Petrography, analyses, P and T 263 El. Wt. % Garnet Biotite Muscovite Si 17.206 16.720 21.574 Al 11.306 9.696 18.450 Ti b.d. 1.118 0.260 Fe 26.776 14.777 0.821 Mg 1.708 5.964 0.365 Mn 1.856 b.d. b.d. Cr n.a. 0.078 b.d. Ca 1.679 b.d. b.d. Ba n.a. 0.159 0.279 K n.a. 7.264 7.653 Na n.a. 0.227 1.022 F n.a. 0.279 0.147 Cl n.a. b.d. b.d. 0 39.692 38.080 43.559 Total 100.223 94.362 94.130 Table D.8: Selected garnet, biotite, and muscovite analyses for sample SLG-370A which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Retrogression: Slight. Some chlorite after garnet and/or biotite. Some chlorite not mimetic on earlier minerals and of uncertain derivation. D.2.7 #7: SLG-370A (Garwin, 1987) Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Sillimanite, Staurolite, Plagioclase, Ilmenite, Rutile, Tourmaline, Chlorite. Porphyroblastic phases: Garnet, Staurolite, Plagioclase. Textures: Rock is coarse grained with a crude foliation only. Plagioclase porphyroblasts overgrow muscovite and kyanite aligned parallel to external foliation implying foliation is S2. Most inclusions in garnet are random but one shows vague folding of foliation(?) not continuous with external foliation (Si?). Rutile mantled by ilmenite only as inclusions in plagioclase. Retrogression: Slight. Chlorite overgrowing biotite only. D.2.8 #8: DM-85-87 Assemblage: Quartz, Muscovite, Biotite, Garnet, Sillimanite, Staurolite, Plagioclase, Il-menite, Chlorite. Porphyroblastic phases: None. Appendix D. Microprobed samples: Petrography, analyses, P and T 264 El. Wt. % Garnet Biotite Muscovite Si 16.837 16.824 21.595 Al 11.506 9.800 18.457 Ti b.d. 1.229 0.318 Fe 27.874 14.858 0.894 Mg 1.628 6.350 0.361 Mn 2.140 b.d. b.d. Cr n.a. b.d. b.d. Ca 0.204 b.d. b.d. Ba n.a. b.d. 0.193 K n.a. 7.162 7.652 Na n.a. 0.229 1.127 F n.a. 0.339 b.d. Cl n.a. b.d. b.d. 0 39.391 38.569 43.726 Total 99.580 95.360 94.323 Table D.9: Selected garnet, biotite, and muscovite analyses for sample DM-85-87 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Textures: Rock is coarse grained with a crude foliation only. Garnet and staurolite not signif-icantly different to matrix in grain size. Mica defines a crude crenulation cleavage, with relict hinge zones. Peak mineral growth (garnet, staurolite, sillimanite(?)) postdates crenulation cleavage. Mica in hinge zones is unstrained. Implies that crenulation cleavage is S2 and earlier foliation is S\. Garnet and staurolite grains too small to show internal foliations. Retrogression: Slight. Chlorite overgrowing biotite in matrix and along some fractures in garnet. D.2.9 #9: C J N F - 9 (Fletcher, 1972) Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Sillimanite, Staurolite, Plagio-clase, Ilmenite, Chlorite. Porphyroblastic phases: Garnet, Kyanite, Staurolite. Textures: Single foliation in the matrix wraps around inclusion rich garnet. Inclusion trails do not define good internal foliation. Staurolite and kyanite apparently overgrows foliation implying foliation is S2. Symplectites of sillimanite, biotite, and small inclusion-free garnets, around which foliation wraps, probably represent prograde destruction of earlier garnet. Some kyanite is early (large, ragged grains) some appears to be late (small well formed grains) and grows most commonly over sillimanite or muscovite(?). This appears to be retrograde growth (during cooling). Appendix D. Microprobed samples: Petrography, analyses, P and T 265 El. Wt. % Garnet Biotite Muscovite Si 16.968 16.643 21.585 Al 11.454 9.825 18.313 Ti b.d. 1.925 0.404 Fe 26.875 14.491 0.819 Mg 1.959 5.391 0.391 Mn 1.151 b.d. b.d. Cr n.a. 0.114 b.d. Ca 1.636 b.d. b.d. Ba n.a. b.d. 0.297 K n.a. 7.281 7.510 Na n.a. 0.554 1.029 F n.a. b.d. b.d. Cl n.a. b.d. b.d. 0 39.454 38.391 43.592 Total 99.497 94.615 93.940 Table D.10: Selected garnet, biotite, and muscovite analyses for sample CJNF-9 which yield the 'typical' P and T (first population) given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) El. Wt. % Garnet Biotite Muscovite Si 16.917 16.766 21.585 Al 11.390 9.934 18.313 Ti b.d. 1.058 0.404 Fe 27.587 13.766 0.819 Mg 1.890 6.380 0.391 Mn 1.649 0.062 b.d. Cr n.a. b.d. b.d. Ca 1.222 b.d. b.d. Ba n.a. b.d. 0.297 K n.a. 7.173 7.510 Na n.a. 0.336 1.029 F n.a. 0.298 b.d. Cl n.a. b.d. b.d. 0 39.564 38.305 43.592 Total 100.219 94.078 93.940 Table D . l l : Selected garnet, biotite, and muscovite analyses for sample CJNF-9 which yield the 'typical' P and T (second population) given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Appendix D. Microprobed samples: Petrography, analyses, P and T 266 El. Wt. % Garnet Biotite Muscovite Si 16.900 16.956 21.616 Al 11.339 9.172 17.943 Ti b.d. 0.906 0.251 Fe 25.680 13.821 1.595 Mg 1.500 7.007 0.386 Mn 4.450 b.d. b.d. Cr n.a. b.d. b.d. Ca 0.133 b.d. b.d. Ba n.a. b.d. b.d. K n.a. 7.928 7.284 Na n.a. 0.168 1.292 F n.a. 0.668 b.d. Cl n.a. b.d. b.d. 0 39.499 38.073 43.383 Total 99.501 94.699 93.750 Table D.12: Selected garnet, biotite, and muscovite analyses for sample DM-86-70 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Retrogression: Very slight. Chlorite mimetic after biotite in matrix. Garnet unaffected. D.2.10 #10: DM-86-70 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Rutile, Chlorite, Sericite. Porphyroblastic phases: Garnet, Kyanite, Staurolite. Textures: Coarse grained mica swirled in a complicated pattern showing possibly three phases of deformation. Garnet, kyanite and staurolite growth predates latest folding but, although some mica in hinge zones is strained, there is some new mica growth parallel to the axial plane of the latest folds (F3). Porphyroblast growth appears synchronous with second phase of de-formation (F2). Rutile present only as inclusions in kyanite. Retrogression: Slight. Patches of chlorite after biotite. Thin sericite mantles around stauro-lite and kyanite and along fractures in staurolite. Plagioclase lightly sericitized. D.2.11 #11: DM-86-173 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Rutile, Tourmaline, Chlorite. Porphyroblastic phases: Garnet, Kyanite, Staurolite. Appendix D. Microprobed samples: Petrography, analyses, P and T 267 El. Wt. % Garnet Biotite Muscovite Si 17.227 16.683 21.630 Al 11.230 9.893 18.637 Ti b.d. 1.025 0.153 Fe 27.336 14.519 0.773 Mg 1.805 6.181 0.339 Mn 1.929 b.d. b.d. Cr n.a. b.d. b.d. Ca 0.779 b.d. b.d. Ba n.a. b.d. b.d. K n.a. 7.359 7.446 Na n.a. 0.244 1.156 F n.a. 0.267 b.d. Cl n.a. b.d. b.d. 0 39.642 38.218 43.679 Total 99.948 94.389 93.813 Table D.13: Selected garnet, biotite, and muscovite analyses for sample DM-86-173 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Textures: Both garnet and staurolite have folded inclusion patterns. Garnet inclusion patterns more complex, staurolite inclusion patterns are straight except at the margins. Staurolite inclusions include rutile. Garnet inclusions include ilmenite. Ilmenite in the matrix. The external foliation wraps around both garnet and staurolite, one staurolite grain bent around in hinge(?) of fold but no strain visible in this grain. Some staurolite growth early in the development of the external foliation (S2/S3). Foliation internal to staurolite is considered S2. Figure 5.12 is a photograph of part of this specimen. Retrogression: Very slight. Small, sparse patches of chlorite after biotite(?). D.2.12 #12: DM-86-197 Assemblage: Quartz, Muscovite, Biotite, Garnet, Sillimanite, Plagioclase, Ilmenite, Apatite, Tourmaline, Chlorite. Porphyroblastic phases: Garnet. Textures: Sparsely micaceous psammite with relict, ragged, poilikoblastic garnet. Most gar-nets replaced by dense swirling mats of sillimanite with associated biotite and small, clear idioblastic garnet. External foliation is probably a crenulation cleavege (sparse mica foliae have two orientations of mica grains). All quartz and mica grains show extensive strain with kink bands in mica ( F 4 ? ) perpendicular to the foliation. Major foliation wraps around garnet (and sillimanite-biotite symplectite) and is probably S2. Sillimanite (fibrolite) is unoriented. Appendix D. Microprobed samples: Petrography, analyses, P and T 268 El. Wt. % Garnet Biotite Muscovite Si 16.868 16.549 21.744 Al 11.300 9.038 18.370 Ti b.d. 1.538 0.418 Fe 26.240 15.929 0.862 Mg 1.784 5.543 0.391 Mn 1.079 b.d. b.d. Cr n.a. b.d. b.d. Ca 3.200 b.d. 0.030 Ba n.a. b.d. 0.178 K n.a. 7.337 7.850 Na n.a. 0.178 0.897 F n.a. 0.279 0.163 Cl n.a. 0.040 b.d. 0 39.322 37.581 43.777 Total 99.793 94.012 94.680 Table D.14: Selected garnet, biotite, and muscovite analyses for sample DM-86-197 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Retrogression: Virtually none. Tiny patches of chlorite associated with garnet. D.2.13 #13: DM-86-205 Assemblage: Quartz, Muscovite, Biotite, Garnet, Sillimanite, Plagioclase, Ilmenite, Apatite, Chlorite, Sericite. Porphyroblastic phases: None. Textures: Sparsely micaceous psammite with micaceous and quartzose foliae. Mica is quart-zose foliae commonly at high angle to mica in micaceous foliae indicating that major foliation is a transposed earlier foliation. Although garnet grain size is similar to that of quartz and mica, it is commonly present in clusters of grains probably formed from earlier generations of garnet. Foliation wraps around these clusters. Sillimanite (fibrolite) is oriented parallel to foliation but is probably mimetic on oriented biotite. Mica grains are commonly kinked at high angles to foliation. Quartz is highly strained. Assignment of foliations to individual phases of deformation is not possible. Retrogression: Very slight. Small patches of chlorite replacing biotite apparently where bi-otite is most highly strained. Chlorite associated with amorphous yellow-brown mix of Fe oxide-hydroxide. Plagioclase shows sparse sericitization. Appendix D. Microprobed samples: Petrography, analyses, P and T 269 El. Wt. % Garnet Biotite Muscovite Si 17.079 16.898 21.630 Al 11.202 10.205 18.090 Ti b.d. 1.333 0.430 Fe 26.683 14.556 1.028 Mg 1.547 5.640 0.429 Mn 2.102 b.d. b.d. Cr n.a. b.d. b.d. Ca 1.917 b.d. b.d. Ba n.a. b.d. 0.285 K n.a. 7.562 8.037 Na n.a. 0.251 0.732 F n.a. 0.215 b.d. Cl n.a. 0.033 b.d. 0 39.482 38.648 43.501 Total 100.012 95.341 94.162 Table D.15: Selected garnet, biotite, and muscovite analyses for sample DM-86-205 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) D.2.14 #14: D M - 8 6 - 2 2 7 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Sillimanite, Staurolite, Plagio-clase, Bmenite, Rutile, Tourmaline, Chlorite, Sericite. Porphyroblastic phases: Garnet, Staurolite, Plagioclase. Textures: Highly micaceous schist with chaotically swirled foliation. Individual, consistent foliation orientations not identifiable. Porphyroblasts are inclusion-poor. Quartz/ilmenite in-clusion trails in garnet are planar and at random angles to external foliation. Garnet and staurolite porphyroblasts are densely fractured perpendicular to foliation which wraps around them. Sillimanite associated with corroded garnet, mostly fibrolitic but with some more coarsely crystalline (0.1 mm wide) sillimanite. Rutile only present as sparse inclusions in plagioclase and staurolite. Kyanite grains are rare and are corroded (relict). Mica grains are kinked (no consistent orientation) and quartz shows moderate undulose extinction. Retrogression: Slight to moderate, patchy. Chlorite replaces biotite (with amorphous yellow-brown Fe oxide-hydroxides). Patchy sericitizatibn of plagioclase. D.2.15 #15: D M - 8 6 - 2 7 0 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Sillimanite, Staurolite, Plagioclase, Bmenite, Rutile, Tourmaline, Apatite, "Sericite. Appendix D. Microprobed samples: Petrography, analyses, P and T 270 El. Wt. % Garnet Biotite Muscovite Si 16.980 16.719 21.478 Al 11.219 9.865 18.362 Ti b.d. 1.491 0.398 Fe 27.784 14.917 1.289 Mg 1.600 5.507 0.363 Mn 1.323 b.d. b.d. Cr n.a. b.d. b.d. Ca 2.117 b.d. b.d. Ba n.a. b.d. b.d. K n.a. 7.356 7.795 Na n.a. 0.262 0.894 F n.a. 0.301 b.d. Cl n.a. b.d. b.d. 0 39.481 38.225 43.562 Total 100.504 94.643 94.141 Table D.16: Selected garnet, biotite, and muscovite analyses for sample DM-86-227 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) El. Wt. % Garnet Biotite Muscovite Si 17.095 17.062 21.785 Al 11.085 9.807 18.228 Ti b.d. 1.262 0.441 Fe 26.301 13.767 0.748 Mg 1.723 6.480 0.472 Mn 1.256 b.d. b.d. Cr n.a. b.d. b.d. Ca 3.748 b.d. b.d. Ba n.a. b.d. 0.315 K n.a. 7.360 7.701 Na n.a. 0.291 0.936 F n.a. 0.313 b.d. Cl n.a. b.d. b.d. 0 39.609 38.717 43.789 Total 100.817 95.059 94.415 Table D.l7: Selected garnet, biotite, and muscovite analyses for sample DM-86-270 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Appendix D. Microprobed samples: Petrography, analyses, P and T 271 El. Wt. % Garnet Biotite Muscovite Si 17.030 16.964 21.775 Al 11.091 10.024 18.452 Ti b.d. 0.992 0.378 Fe 26.582 14.157 0.693 Mg 1.909 6.283 0.359 Mn 1.690 0.067 b.d. Cr n.a. b.d. b.d. Ca 1.727 0.036 b.d. Ba n.a. b.d. b.d. K n.a. 7.532 7.482 Na n.a. 0.133 1.118 F n.a. 0.245 b.d. Cl n.a. 0.052 b.d. 0 39.318 38.617 43.831 Total 99.347 95.102 94.088 Table D.18: Selected garnet, biotite, and muscovite analyses for sample DM-86-281 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Porphyroblastic phases: Garnet, Kyanite, Plagioclase. Textures: Psammitic schist with a folded foliation (single fold in thin section). Porphyroblasts growth predates fold development ( F 3 ) . Porphyroblasts are aligned parallel to S3. Sparse evidence (two mica orientations) that the folded foliation (S2) is in turn a crenulation cleavage or a transposed foliation (of Si). Quartz and mica show undulose extinction and some kyanite blades have strong kink bands ( S 4 ? ) . Rutile as inclusions in kyanite only. Staurolite grains small and rare. Sillimanite sparsely developed near garnet. Retrogression: Almost none. Patchy sericitization of plagioclase. D.2.16 #16: D M - 8 6 - 2 8 1 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Chlorite, Sericite. Porphyroblastic phases: None. Textures: Highly quartzose schist with sparse micaceous foliae. Mica grains in micaceous foliae parallel to foliae. Small grains of aluminous phases (garnet, kyanite, staurolite) in micaceous foliae grew synchronously with foliation developmnet (S2?)-Retrogression: Slight to moderate. Slight chloritization of biotite. Well developed, patchy sericitization of plagioclase. Appendix D. Microprobed samples: Petrography, analyses, P and T 272 El. Wt. % Garnet Biotite Muscovite Si 17.067 16.866 21.546 Al 11.131 9.656 18.514 Ti b.d. 0.958 0.192 Fe 27.892 14.607 0.898 Mg 1.912 6.473 0.354 Mn 1.841 0.059 b.d. Cr n.a. b.d. b.d. Ca 0.702 b.d. b.d. Ba n.a. b.d. 0.132 K n.a. 7.135 7.292 Na n.a. 0.150 1.359 F n.a. 0.397 0.201 Cl n.a. 0.037 b.d. 0 39.532 38.249 43.538 Total 100.077 94.587 94.026 Table D.19: Selected garnet, biotite, and muscovite analyses for sample DM-86-351 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) D.2.17 #17: DM-86-351 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Rutile, Tourmaline, Pyrite, Sericite. Porphyroblastic phases: Garnet, Staurolite, Kyanite. Textures: Micaceous schist with a well developed transposed foliation. The transposed fo-liation wraps around garnet porphyroblasts but some staurolite and kyanite grains truncate it. Porphyroblasts are inclusion-poor. Garnet shows inclusion-zoning with almost clear cores, an intermediate zone of large quartz and mica inclusions, and rims which are densely packed with very fine grained opaque material (graphite?). Ilmenite and pyrite occurs in the matrix and as inclusions in garnet. Rutile is present only as inclusions in kyanite and staurolite. The external foliation (transposed) is considered to be S2 on Si due to the relationship with the porphyroblasts. Retrogression: Very slight sericitization of plagioclase. D.2.18 #18: DM-86-369 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Plagioclase, Ilmenite, Rutile. Porphyroblastic phases: None. Textures: Micaceous quartzite with very sparse foliation. Mica grains have two orientations Appendix D. Microprobed samples: Petrography, analyses, P and T 273 El. Wt. % Garnet Biotite Muscovite Si 16.954 16.681 21.585 Al 11.026 9.676 18.069 Ti b.d. 1.202 0.384 Fe 27.421 14.005 0.805 Mg 1.753 6.341 0.412 Mn 1.470 b.d. b.d. Cr n.a. b.d. b.d. Ca 1.444 b.d. b.d. Ba n.a. b.d. 0.195 K n.a. 7.213 8.215 Na n.a. 0.138 0.716 F n.a. 0.285 0.161 Cl n.a. 0.038 b.d. 0 39.139 38.040 43.318 Total 99.207 93.619 93.860 Table D.20: Selected garnet, biotite, and muscovite analyses for sample DM-86-369 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) indicating that the foliation (defined by quartzose and micaceous foliae) is a crenulation cleav-age. Mica grains in micaceous foliae parallel the margins of the foliae, in quartzose layers mica grains make an angle of 30 to 45° to the foliae. The garnet grains are very small and scattered, and only a single kyanite grain was seen in thin section. Garnet grains are too small to de-termine their relationship to the fabric. In contrast to other samples above, both ilmenite and rutile are present in the matrix. Ilmenite is seen mantling (partly or completely) rutile grains. Some rutile grains enclosed within single quartz grains are pristine. Retrogression: None. D.2.19 #19: DM-86-384 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Plagioclase, Ilmenite, Rutile, Zir-con, Chlorite, Sericite. Porphyroblastic phases: Kyanite. Textures: Micaceous schist with a well developed but chaotic foliation. The micas are aligned in the major foliation which is broadly buckled and micas are kinked perpendicular to the folia-tion. The garnet are small but appear to overgrow mica denning the foliation. Garnet grains are included within kyanite porphyroblasts which are aligned in the major foliation. One kyanite grain, athwart the foliation, is very strongly kinked. This implies that metamorphic mineral growth predated the development of this foliation, but the fact that garnet apparently overgrows Appendix D. Microprobed samples: Petrography, analyses, P and T 274 E l . Wt . % Garnet Biotite Muscovite Si 17.231 17.154 21.627 A l 11.304 9.951 18.014 T i b.d. 0.887 0.391 Fe 24.842 13.109 0.795 M g 2.089 7.315 0.503 M n 2.221 0.086 0.061 Cr n.a. b.d. b.d. C a 2.473 b.d. b.d. B a n.a. b.d. 0.315 K n.a. 7.543 7.650 Na n.a. 0.125 1.031 F n.a. 0.330 0.160 C l n.a. b.d. b.d. 0 39.784 39.063 43.394 Total 99.944 95.563 93.941 Table D.21: Selected garnet, biotite, and muscovite analyses for sample D M - 8 6 - 3 8 4 which yield the ' typ ica l ' P and T given i n table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection l imi t given i n table D . l ) the micas which define the foliation implies that the foliation is transposed (S2/S3). Rutile and ilmenite are present within the matrix and ilmenite appears to be a retrograde product of rutile. Inclusions in garnet and kyanite grains and larger mica and plagioclase grains are of rutile. Retrogression: Very slight chloritization of biotite. Nearby rutile commonly mantled by ilmenite. Light but extensive sericitization of plagioclase. D.2.20 #20: JRM-0626-36 (Montgomery, 1985) Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Plagioclase, Ilmenite, Rutile, Tour-maline. Porphyroblastic phases: Garnet, Kyanite. Textures: Micaceous schist with a spectacularly folded foliation. Mica and kyanite grains bent around tight to isoclinal microfolds. Folding postdates the porphyroblast development and is considered S3. The earlier foliation is defined by very well aligned mica and abundant but skeletal kyanite grains. Kyanite growth synchronous with foliation development (F 2 ) . Foliation wraps around garnet which has a discordant internal foliation (Si). Dmenite present in matrix and as inclusions in the cores of garnets. Rutile present mainly as inclusions in kyanite and rims of garnet. Some rutile remains in the matrix mantled by ilmenite. A clear sequence of events is therefore defined by these observations. 1: Development of a foliation (Si) during which i l -menite was the stable titanium oxide. 2: Growth of garnet over this foliation. 3: Development Appendix D. Microprobed samples: Petrography, analyses, P and T 275 El. Wt. % Garnet Biotite Muscovite Si 17.162 16.930 21.670 Al 11.486 9.920 18.227 Ti b.d. 0.989 0.358 Fe 26.231 12.421 0.704 Mg 2.066 6.976 0.450 Mn 2.171 0.065 b.d. Cr n.a. b.d. b.d. Ca 1.769 0.082 b.d. Ba n.a. 0.263 0.345 K n.a. 7.100 7.380 Na n.a. 0.457 1.146 F n.a. 0.363 b.d. Cl n.a. b.d. b.d. 0 40.031 38.480 43.608 Total 100.916 94.046 93.888 Table D.22: Selected garnet, biotite, and muscovite analyses for sample JRM-0626-36 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) of a second foliation (S2) synchronous with the continued growth of garnet and the growth of kyanite. Rutile was the stable titanium oxide. 4: Folding of the S 2 foliation by F 3 (rutile still stable? but micas and kyanite not recrystallized). 5: Final cooling and the development of ilmenite over accessible rutile grains. Retrogression: None, except very minor ilmenite after rutile. D.2.21 #21: JRM-0716-76 (Montgomery, 1985) Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Rutile, Chlorite, Sericite. Porphyroblastic phases: Garnet, Kyanite, Staurolite, Plagioclase. Textures: Micaceous schist with a single foliation which wraps around all porphyroblastic phases. External foliation may be a crenulation cleavage or transposed foliation (two orien-tations of mica). Garnets contain sparse ilmenite inclusions, and kyanite is inclusion-free. Staurolite has notable internal bands of very fine grained opaque material (graphite?) but close examination reveals that these bands commonly terminate abruptly within the staurolite along straight lines and the overall appearance is of a crystal ghost within the staurolite. The bands of opaque material strongly resemble those seen in twinned chloritoid grains and it is proposed that staurolite replaced chloritoid during progressive metamorphism. Most of the remaining opaque material is ilmenite (both in the matrix and as inclusions) but some rutile is present as Appendix D. Microprobed samples: Petrography, analyses, P and T 276 El. Wt. % Garnet Biotite Muscovite Si 17.156 16.616 21.755 Al 11.348 10.096 18.501 Ti b.d. 0.885 0.430 Fe 28.257 15.199 0.831 Mg 1.941 5.904 0.421 Mn 1.277 b.d. b.d. Cr n.a. b.d. 0.079 Ca 0.993 b.d. b.d. Ba n.a. b.d. 0.478 K n.a. 7.418 7.231 Na n.a. 0.198 1.241 F n.a. 0.470 b.d. Cl n.a. b.d. b.d. 0 39.810 38.157 44.037 Total 100.782 94.943 95.004 Table D.23: Selected garnet, biotite, and muscovite analyses for sample JRM-0716-76 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) inclusions within staurolite. Retrogression: Moderate. Staurolite and kyanite have thin mantles of sericite and sericite along fractures. Chlorite replaces biotite and some garnet. The retrogression of garnet in particular centers on points where strain may have been localized. D.2.22 #22: L C P - 4 0 (Pigage, 1978) Assemblage: Quartz, Muscovite, Biotite, Garnet, Sillimanite, Plagioclase, Ilmenite, Tourma-line, Chlorite, Sericite. Porphyroblastic phases: None. Textures: Moderately micaceous schist with a single, poorly denned foliation. Mica grains are coarse (5 mm) and commonly randomly orientled. Some large mica grains surround small, clear, idioblastic garnet grains. Sillimanite is present as small swirling knots often associated with garnet. The age of the crude foliation with respect to peak metamorphism is unknown. Ilmenite is only Ti phase present. Retrogression: Slight to moderate chloritization of biotite (garnet pristine). Plagioclase lightly sericitized. Appendix D. Microprobed samples: Petrography, analyses, P and T 277 El. Wt. % Garnet Biotite Muscovite Si 17.176 16.974 21.410 Al 11.299 10.047 18.786 Ti b.d. 1.139 0.253 Fe 28.556 14.746 0.787 Mg 1.838 5.941 0.266 Mn 1.413 b.d. b.d. Cr n.a. b.d. 0.073 Ca 0.556 b.d. b.d. Ba n.a. 0.172 b.d. K n.a. 6.918 7.201 Na n.a. 0.280 1.321 F n.a. 0.468 b.d. Cl n.a. b.d. b.d. 0 39.745 38.511 43.615 Total 100.583 95.196 93.712 Table D.24: Selected garnet, biotite, and muscovite analyses for sample LCP-40 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) D.2.23 #23: LCP-223 (Pigage, 1978) Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Sillimanite, Plagio-clase, Ilmenite, Tourmaline. Porphyroblastic phases: Garnet, Kyanite, Staurolite. Textures: Moderately micaceous schist with a crenulation cleavage. The foliation wraps around garnet prophyroblasts. Inclusions in garnet porphyroblasts are sufficiently large and numerous to render the garnet skeletal in appearance. There is no apparent foliation internal to the garnet. A single relict kyanite porphyroblast is inclusion free. Fibrolitic sillimanite is most commonly associated with garnet and staurolite. Some staurolite grains are very small, clear and idioblastic and may represent new growth during prograde destruction of earlier stau-rolite. Retrogression: Negligible. D.2.24 #24: DM-87-76 Assemblage: Quartz, Muscovite, Biotite, Garnet, Staurolite, Sillimanite, Plagioclase, Il-menite, Apatite. Porphyroblastic phases: Garnet. Textures: Micaceous schist with chaotically arranged mica grains. The rock (as seen in thin Appendix D. Microprobed samples: Petrography, analyses, P and T 278 El. Wt. % Garnet Biotite Muscovite Si 16.899 17.066 22.432 Al 11.630 10.068 17.983 Ti b.d. 1.139 0.306 Fe 27.411 14.265 0.914 Mg 1.822 6.337 0.614 Mn 1.789 b.d. b.d. Cr n.a. b.d. b.d. Ca 1.274 b.d. b.d. Ba n.a. b.d. b.d. K n.a. 7.210 7.980 Na n.a. 0.258 0.840 F n.a. 0.276 b.d. Cl n.a. b.d. b.d. 0 39.734 38.882 44.327 Total 100.559 95.501 95.396 Table D.25: Selected garnet, biotite, and muscovite analyses for sample LCP-223 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) El. Wt. % Garnet Biotite Muscovite Si 17.003 16.725 21.531 Al 11.434 10.009 18.536 Ti b.d. 1.198 0.363 Fe 27.083 14.696 0.868 Mg 1.940 5.744 0.397 Mn 1.446 b.d. b.d. Cr n.a. b.d. b.d. Ca 1.538 b.d. b.d. Ba n.a. b.d. 0.111 K n.a. 7.880 7.922 Na n.a. 0.162 0.960 F n.a. 0.445 b.d. Cl n.a. b.d. b.d. 0 39.612 38.246 43.752 Total 100.056 95.105 94.440 Table D.26: Selected garnet, biotite, and muscovite analyses for sample DM-87-76 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Appendix D. Microprobed samples: Petrography, analyses, P and T 279 E l . Wt . % Garnet Biotite Muscovite Si 16.960 16.271 21.586 A l 11.458 9.802 17.573 T i b.d. 2.095 1.160 Fe 27.719 17.785 0.950 M g 1.658 3.919 0.454 M n 1.155 0.061 b.d. Cr n.a. b.d. 0.051 Ca 1.604 b.d. b.d. B a n.a. b.d. b.d. K n.a. 7.828 8.703 Na n.a. 0.093 0.409 F n.a. 0.259 b.d. C l n.a. 0.064 b.d. 0 39.476 37.873 43.531 Total 100.030 96.050 94.417 Table D.27: Selected garnet, biotite, and muscovite analyses for sample D M - 8 7 - 7 1 which yield the first ' typ ica l ' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b .d = below detection l imi t given in table D . l ) section) does not have a distinct foliation implying mineral growth in the absence of a directed stress. However, all anisotropic grains show very pronounced undulose extinction implying sig-nificant post-crystallization directed stress. Porphyroblastic garnet is inclusion free except at the outermost rim where slivers of ilmenite define a foliation parallel to the mica grains ouside the garnet. The garnet is densely fractured perpendicular to the crude foliation just outside it. A single staurolite grain is highly fractured but subidioblastic and the absence of staurolite may be a function of bulk composition as of metamorphic grade. Sillimanite is fibrolitic and is associated with biotite. Retrogression: Patchy replacement of biotite (and muscovite?) by chlorite (and sericite?) particularly where these,are highly strained (kinked). D.2.25 #25: D M - 8 7 - 7 1 Assemblage: Quartz, Muscovite, Biotite, Garnet, Sillimanite, Plagioclase, Ilmenite. Porphyroblastic phases: None. Textures: Micaceous quartzite with sparse mica grains defining a single foliation. Garnet grains are small (< 1 mm) and inclusion-free. Foliation wraps around the garnets and they are fractured perpendicular to i t . Sillimanite grains (up to 0.3 mm across) are aligned in the foliation and seem to predate the development of the foliation. Retrogression: None. Appendix D. Microprobed samples: Petrography, analyses, P and T 280 El. Wt. % Garnet Biotite Muscovite Si 16.930 15.953 21.574 Al 11.494 10.590 17.720 Ti b.d. 1.671 0.877 Fe 27.094 17.282 0.873 Mg 1.306 4.264 0.438 Mn 1.186 0.110 b.d. Cr n.a. b.d. b.d. Ca 2.140 0.064 b.d. Ba n.a. b.d. b.d. K n.a. 7.795 8.810 Na n.a. 0.124 0.339 F n.a. 0.267 0.178 Cl n.a. b.d. b.d. 0 39.229 38.065 43.336 Total 99.379 96.185 94.145 Table D.28: Selected garnet, biotite, and muscovite analyses for sample DM-87-71 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b .d = below detection limit given in table D.l) D.2.26 #26: DM-86-45 A s s e m b l a g e : Quartz, Muscovite, Biotite, Garnet, Kyanite, Sillimanite, Plagioclase, Ilmenite, Chlorite. P o r p h y r o b l a s t i c p h a s e s : Garnet. T e x t u r e s : Micaceous quartzite with sparse mica denning a very crude foliation which has been buckled. Garnet grains are small (1 mm) and sparse and inclusion-free. A few small ragged kyanite grains are probably relict. Sillimanite growth is confined to micaceous area and is aligned parallel to the foliation. All anisotropic minerals show pronounced undulose extinction. R e t r o g r e s s i o n : Slight to moderate chloritization of biotite. D.2.27 #27: JAF-7-9-29 (Fillipone, 1985) A s s e m b l a g e : Quartz, Muscovite, Biotite, Garnet, Kyanite, Sillimanite, Staurolite, Plagioclase, Ilmenite, Rutile, Tourmaline. P o r p h y r o b l a s t i c p h a s e s : Kyanite, Staurolite, Garnet. T e x t u r e s : Micaceous quartzite with sparse mica outlining a well defined foliation. In more micaceous areas the foliation appears to be a transposed foliation. Kyanite and staurolite are the largest and most prominent porphyroblasts, whereas garnets are small (1.5 mm). None of these phases are rich in inclusions. Foliation wraps around garnet but the staurolite and kyanite Appendix D. Microprobed samples: Petrography, analyses, P and T 281 El. Wt. % Garnet Biotite Muscovite Si 17.218 17.030 21.604 Al 11.485 10.098 17.982 Ti b.d. 2.048 0.555 Fe 26.120 13.967 0.978 Mg 2.132 5.516 0.488 Mn 1.569 0.102 b.d. Cr n.a. b.d. b.d. Ca 2.662 b.d. b.d. Ba n.a. b.d. 0.329 K n.a. 8.129 8.810 Na n.a. 0.123 0.336 F n.a. 0.163 b.d. Cl n.a. b.d. b.d. 0 40.122 39.075 43.493 Total 101.308 96.251 94.575 Table D.29: Selected garnet, biotite, and muscovite analyses for sample DM-86-45 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) El. Wt. % Garnet Biotite Muscovite Si 16.999 16.342 21.678 Al 11.351 10.007 18.457 Ti b.d. 0.890 0.313 Fe 27.001 16.323 0.716 Mg 2.139 5.594 0.366 Mn 1.833 0.123 b.d. Cr n.a. b.d. b.d. Ca 0.970 b.d. b.d. Ba n.a. b.d. 0.189 K n.a. 7.244 7.795 Na n.a. 0.129 0.890 F n.a. 0.221 b.d. Cl n.a. b.d. b.d. 0 39.623 37.944 43.679 Total 99.916 94.817 94.083 Table D.30: Selected garnet, biotite, and muscovite analyses for sample JAF-7-9-29 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Appendix D. Microprobed samples: Petrography, analyses, P and T 282 El. Wt. % Garnet Biotite Muscovite Si 17.200 16.397 21.761 Al 11.366 10.112 18.173 Ti b.d. 0.911 0.323 Fe 27.310 15.128 0.798 Mg 1.880 5.495 0.448 Mn 1.524 b.d. b.d. Cr n.a. b.d. 0.060 Ca 1.835 b.d. b.d. Ba n.a. b.d. 0.192 K n.a. 7.378 7.950 Na n.a. 0.181 0.863 F n.a. 0.213 b.d. Cl n.a. 0.048 0.026 0 39.913 37.727 43.656 Total 101.028 93.590 94.250 Table D.31: Selected garnet, biotite, and muscovite analyses for sample JAF-7-9-29 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) appears to overgrow the micas denning the foliation. All this indicates that the foliation is S2. Rutile occurs as rare inclusions in staurolite, kyanite, and garnet. Ilmenite occurs in the matrix. There is no evidence for deformation after mineral growth (kinks, undulose extinction, etc.). Retrogression: None. D.2.28 #28: D M - 8 7 - 8 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Sillimanite, Staurolite, Plagioclase, Ilmenite. Porphyroblastic phases: Garnet, Kyanite, Staurolite. Textures: Micaceous quartzite with sparse mica outlining a well denned transposed foliation. This foliation wraps around garnet and in one case causes an adjacent kyanite grain to wrap around a garnet. This foliation is considered to be S3. The same garnet grain has mica wrapping around the garnet in the pressure shadow of S3 at one side and appears to be a remnant of an earlier foliation (S2). Within the garnet, quartz inclusions define a small fold hinge the axial plane of which is parallel to the proposed S2. This indicates that the garnet growth was synchronous with the development of S2. The orientation of the kyanite grain which wraps the garnet grain in S3 indicates that it grew athwart the earlier foliation (S2). The fold preserved within the garnet appears to be a folded foliation (i.e. segregated layering) and therefore is deemed an F 2 fold of Si. The kyanite and staurolite grains are without inclusions and predate Appendix D. Microprobed samples: Petrography, analyses, P and T 283 El. Wt. % Garnet Biotite Muscovite Si 17.174 16.141 21.583 Al 11.338 10.553 18.782 Ti b.d. 0.619 0.172 Fe 28.656 17.484 0.942 Mg 2.307 4.618 0.261 Mn 0.543 0.072 b.d. Cr n.a. b.d. b.d. Ca 0.204 b.d. b.d. Ba n.a. b.d. 0.189 K n.a. 7.474 8.435 Na n.a. 0.135 0.586 F n.a. 0.328 b.d. Cl n.a. 0.033 b.d. 0 39.667 37.696 43.807 Total 99.889 95.153 94.757 Table D.32: Selected garnet, biotite, and muscovite analyses for sample DM-87-8 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) El. Wt. % Garnet Biotite Muscovite Si 16.741 16.405 21.976 Al 11.189 10.792 17.938 Ti b.d. 1.467 0.450 Fe 27.575 15.641 1.063 Mg 1.447 4.515 0.538 Mn 1.513 b.d. b.d. Cr n.a. b.d. b.d. Ca 1.794 b.d. b.d. Ba n.a. b.d. b.d. K n.a. 7.515 8.344 Na n.a. 0.218 0.499 F n.a. 0.308 b.d. Cl n.a. b.d. b.d. 0 39.003 38.232 43.857 Total 99.262 95.093 94.665 Table D.33: Selected garnet, biotite, and muscovite analyses for sample DM-87-8 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Appendix D. Microprobed samples: Petrography, analyses, P and T 284 El. Wt. % Garnet Biotite Muscovite Si 17.081 16.746 21.733 Al 11.503 10.421 18.339 Ti b.d. 1.330 0.299 Fe 26.460 15.429 0.830 Mg 2.142 5.210 0.440 Mn 2.404 0.081 b.d. Cr n.a. b.d. b.d. Ca 0.568 b.d. b.d. Ba n.a. b.d. 0.225 K n.a. 7.334 8.089 Na n.a. 0.212 0.788 F n.a. 0.287 0.142 Cl n.a. b.d. b.d. 0 39.805 38.574 43.702 Total 99.963 95.624 94.587 Table D.34: Selected garnet, biotite, and muscovite analyses for sample DM-87-17 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) S3. Sillimanite is present only as tiny wisps near stauolite grains. The fact that the kyanite bent around garnet in S3 is noticably kinked implies that peak metamorphic conditions existed before F 3 folding. F 3 folding produced the major anticline affecting the rocks from which this sample is taken. Retrogression: None. D.2.29 #29: DM-87-17 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Sillimanite, Andalusite, Plagio-clase, Ilmenite, Rutile, Tourmaline. Porphyroblast ic phases: Garnet, Kyanite, Andalusite. Textures: Quartzose schist with a poorly denned foliation. Narrow zones of aligned mica de-fine this foliation but domains between these zones have little preferred mica orientation. Early garnet porphyroblasts are replaced by swirled sillimanite and biotite symplectite with smaller garnet and the foliation wraps around these symplectites. Kyanite porphyroblasts overgrow mica grains but are deformed by the foliation-producing event. The opaque inclusions in the remanent early garnet are small but appear to be ilmenite. Later garnet contains rutile and ru-tile inclusions are present within the kyanite. Ilmenite is present in the matrix. The andalusite porphyroblasts overgrow the mica but are extensively deformed by the foliation. In addition, andalusite appears to partly enclose a sillimanite-biotite symplectite after garnet (photograph, Appendix D. Microprobed samples: Petrography, analyses, P and T 285 El. Wt. % Garnet Biotite Muscovite Si 16.942 16.720 21.533 Al 11.255 10.414 18.831 Ti b.d. 1.151 0.305 Fe 26.276 15.486 0.699 Mg 1.639 5.312 0.270 Mn 1.254 b.d. b.d. Cr n.a. b.d. b.d. Ca 3.270 b.d. b.d. Ba n.a. b.d. 0.360 K n.a. 7.380 7.740 Na n.a. 0.243 0.859 F n.a. 0.207 b.d. Cl n.a. b.d. b.d. 0 39.378 38.547 43.765 Total 100.014 95.460 94.362 Table D.35: Selected garnet, biotite, and muscovite analyses for sample DM-87-17 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) fig. 5.20). Also, the andalusite contains rutile inclusions variably mantled by ilmenite. The textures are interpreted as follows. The foliation seen is considered to be S3 and postdates porphyroblast growth. The earliest garnet grew synchronously with an early foliation(?) prob-ably S2 and while ilmenite was the stable Ti-oxide. Increasing grade postdating S2 produced continued garnet growth and kyanite, with rutile as the Ti-oxide. Peak temperatures after F 2 produced the sillimanite. Decompression as the result of F 3 caused the appearance of an-dalusite which was deformed along with the earlier phases during the development of S3. The decompression also resulted in the production of ilmenite from rutile. Retrogression: None. D.2.30 #30: DM-87-24 Assemblage: Quartz, Muscovite, Biotite, Garnet, Sillimanite, Staurolite, Andalusite, Plagio-clase, Ilmenite. Porphyroblastic phases: Staurolite, Andalusite. Textures: Schist very similar to the previous specimen (#29) with a poorly defined foliation. Relict hinge zones indicate that the foliation is transposed. Garnet is rare and appears to be relict (see the discussion on the thermobarometry of this specimen in chapter 5). The staurolite grains are inclusion-free and all appear to be mantled by andalusite. The andalusite appears to replace staurolite and to overgrow the mica foliae, where it is usually skeletal with abundant Appendix D. Microprobed samples: Petrography, analyses, P and T 286 El. Wt. % Garnet Biotite Muscovite Si 17.074 16.402 21.589 Al 11.588 9.757 17.865 Ti b.d. 1.275 0.278 Fe 25.946 16.790 2.073 Mg 2.176 5.192 0.347 Mn 1.420 0.270 b.d. Cr n.a. b.d. b.d. Ca 2.946 b.d. b.d. Ba n.a. 0.256 0.296 K n.a. 7.276 7.901 Na n.a. 0.263 0.915 F n.a. 0.254 b.d. Cl n.a. b.d. 0.025 0 40.049 38.026 43.432 Total 101.199 95.761 94.721 Table D.36: Selected garnet, biotite, and muscovite analyses for sample DM-87-24 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) El. Wt. % Garnet Biotite Muscovite Si 16.968 16.528 21.589 Al 11.233 9.818 17.865 Ti b.d. 1.321 0.278 Fe 24.720 16.337 2.073 Mg 1.540 5.143 0.347 Mn 1.148 0.214 b.d. Cr n.a. b.d. b.d. Ca 5.954 b.d. b.d. Ba n.a. 0.239 0.296 K n.a. 7.187 7.901 Na n.a. 0.252 0.915 F n.a. 0.417 b.d. Cl n.a. b.d. 0.025 0 39.609 37.990 43.432 Total 101.172 95.446 94.721 Table D.37: Selected garnet, biotite, and muscovite analyses for sample DM-87-24 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) Appendix D. Microprobed samples: Petrography, analyses, P and T 287 El. Wt. % Garnet Biotite Muscovite Si 17.123 16.940 21.957 Al 11.420 9.918 18.362 Ti b.d. 0.436 0.312 Fe 26.750 14.488 0.934 Mg 2.042 6.460 0.451 Mn 1.976 0.070 b.d. Cr n.a. b.d. b.d. Ca 0.928 b.d. b.d. Ba n.a. b.d. 0.132 K n.a. 7.415 7.651 Na n.a. 0.174 1.109 F n.a. 0.392 b.d. Cl n.a. b.d. b.d. 0 39.742 38.245 44.055 Total 99.981 94.538 94.963 Table D.38: Selected garnet, biotite, and muscovite analyses for sample DM-87-32 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) mica inclusions parallel to the external foliation. None of the grains in this sample show pro-nounced undulose extinction implying that mineral recrystalization continued after the foliation development ( F 3 ) . This rock does not contain rutile. The metamorphic history is considered to be similar to that proposed for the previous sample. The andalusite is clearly the latest mineral to crystallize and decompression while temperatures were maintained is considered likely. This is in agreement with the thermobarometric data as given in chapter 5. Retrogression: None. D . 2 . 3 1 # 3 1 : D M - 8 7 - 3 2 Assemblage: Quartz, Muscovite, Biotite, Garnet, Staurolite, Kyanite, Plagioclase, Ilmenite, Rutile, Tourmaline, Chlorite, Sericite. Porphyroblastic phases: Garnet, Kyanite, Staurolite. Textures: Micaceous schist with a pronounced transposed foliation which postdates the peak of metamorphism. This is implied by the wrapping of this foliation around porphyroblasts and the presence of bowed kyanite grains in hinge zones. Garnet growth preceded that of staurolite and kyanite both of which contain garnet enclusions. Garnets show zoning in the inclusion patterns with moderately clear cores (ilmenite inclusions), an intermediate zone of coarse quartz and mica inclusions and a rim rich in very fine grained opaque material (graphite?). These are similar to the garnets noted in sample #17. Kyanite is inclusion-free (some tourmaline) and Appendix D. Microprobed samples: Petrography, analyses, P and T 288 El. Wt. % Garnet Biotite Muscovite Si 16.967 17.210 21.480 Al 11.376 9.816 19.059 Ti b.d. 0.852 0.227 Fe 27.203 14.236 0.611 Mg 1.928 6.331 0.213 Mn 1.650 b.d. b.d. Cr n.a. b.d. b.d. Ca 1.112 b.d. b.d. Ba n.a. b.d. 0.330 K n.a. 7.306 7.325 Na n.a. 0.232 1.280 F n.a. 0.406 0.158 Cl n.a. b.d. b.d. 0 39.499 38.574 43.821 Total 99.735 94.963 94.504 Table D.39: Selected garnet, biotite, and muscovite analyses for sample DM-87-35 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) inclusions in staurolite are of rutile. The presence of ubiquitous undulose extinction implies no post-deformational recrystallization (F3). Retrogression: Slight chloritization of biotite. Chlorite is usually rutilated. Plagioclase is slightly sericitized. D.2.32 #32: DM-87-35 Assemblage: Quartz, Muscovite, Biotite, Garnet, Staurolite, Kyanite, Plagioclase, Ilmenite, Rutile, Chlorite. Porphyroblastic phases: Garnet, Kyanite, Staurolite. Textures: Micaceous schist with a well developed transposed foliation. Staurolite and kyanite porphyroblasts are aligned parallel to this foliation ( S 3 ) . Staurolite grains are very large (up to 2 cm), contain abundant inclusions of quartz, and appeared to have continued growth during the development of S3. Inclusion trails are asymptotic to the external foliation and the grains are strain-free. Kyanite grains (inclusion-free) are strongly strained with abundant kink bands, particularly where they have wrapped around other porphyroblasts (e.g. garnet), and thus kyanite growth probably ceased prior to the development of S3. The growth history of staurolite appears to be very complex. Staurolite shows pronounced zonation in the density of quartz inclusions. Cores have abundant small inclusions whereas rims have fewer but larger inclusions. The boundaries between zones are planar and parallel the crystal faces. This implies early Appendix D. Microprobed samples: Petrography, analyses, P and T 289 El. Wt. % Garnet Biotite Muscovite Si 17.002 16.643 21.452 Al 11.223 10.246 18.517 Ti b.d. 0.854 0.207 Fe 26.679 15.416 0.823 Mg 1.855 5.263 0.332 Mn 1.431 0.140 b.d. Cr n.a. b.d. b.d. Ca 2.522 b.d. b.d. Ba n.a. b.d. 0.254 K n.a. 7.431 8.217 Na n.a. 0.165 0.622 F n.a. 0.261 b.d. Cl n.a. b.d. b.d. 0 39.522 38.031 43.381 Total 100.234 94.450 93.805 Table D.40: Selected garnet, biotite, and muscovite analyses for sample DM-87-39 which yield the 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) growth over a finer grained matrix followed by a hiatus and then renewed growth over a coarser matrix. In most cases the boundary between the density of inclusions also marks an abrupt change in inclusion trail orientation. The core inclusions are random (or faintly planar) and at a high angle to the external foliation, whereas rim inclusions are strongly oriented and usually become asymptotic to the external foliation. In the largest grains some crenulated inclusion patterns are seen. Garnet grains have a few mica inclusions and these are randomly oriented implying growth over a non-directed texture (see fig. 5.10). This, along with other data given in chapter 5, is one of the pieces of evidence that the rocks of unit 3 (chapter 2 and map) have experienced a phase of deformation less than those of unit 1. These data imply the following metamorphic history: 1: Early growth of mica in the absence of directed stress. 2: Continued temperature rise accompanied by growth of garnet. 3: Continued garnet growth and the development of a foliation (S2). 4: Growth of staurolite and kyanite. Continuous recrystallization of micas and quartz. 5: Final deformation (S3) accompanied by staurolite growth but not kyanite. Retrogression: Slight chloritization of biotite. D.2.33 #33: DM-87-39 Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Plagioclase, Hematite, Sericite. Porphyroblastic phases: Kyanite. Appendix D. Microprobed samples: Petrography, analyses, P and T 290 Textures: Psammitic schist with sparse mica denning a single foliation. Single kyanite grain (4 mm) is inclusion-free, aligned parallel to foliation ( S 3 ? ) , and shows undulose extinction. Garnet grains are very small (< 1 mm), and show no internal foliations. Retrogression: Plagioclase is lightly sericitized and hematite may replace magnetite (or il-menite). D.2.34 #34: DM-87-54 No pressure temperature determinations made because the muscovite analyses were unusable. Garnet zoning profile given in chapter 5. Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Rutile, Graphite. Porphyroblastic phases: Garnet, Kyanite, Staurolite. Textures: Micaceous schist identical to sample #'s 31 and 32. The only difference is the abundance of very fine grained opaque material (graphite?) in the matrix. Photographs of parts of this sample given in figures 5.9 and 5.13. Retrogression: Negligible. D.2.35 #35: J K R - 8 8 - 4 (Radloff, 1989) Assemblage: Quartz, Muscovite, Biotite, Garnet, Kyanite, Staurolite, Plagioclase, Ilmenite, Chlorite. Porphyroblastic phases: Garnet, Kyanite, Staurolite. Textures: Micaceous schist very similar to sample #'s 31, 32, and 34. The foliation ( F 3 ) is defined by narrow zones of strongly deformed mica. Between these zones are domains of less deformed mica usually at a high angle to S3 defining an earlier foliation (S2). The rock is finer grained than those listed above. Other relationships are the same. Retrogression: Moderate chloritization of the garnet. Cores seem most affected and most garnets are considerably embayed. Appendix D. Microprobed samples: Petrography, analyses, P and T 291 El. Wt. % Garnet Biotite Muscovite Si 17.136 17.189 21.613 Al 11.284 9.800 19.190 Ti b.d. 0.942 0.171 Fe 28.156 14.915 0.506 Mg 2.008 5.895 0.226 Mn 1.744 b.d. b.d. Cr n.a. b.d. b.d. Ca 0.189 b.d. b.d. Ba n.a. b.d. b.d. K n.a. 7.572 6.359 Na n.a. 0.145 1.733 F n.a. 0.422 b.d. Cl n.a. b.d. b.d. 0 39.698 38.526 44.026 Total 100.215 95.406 93.824 Table D.41: Selected garnet, biotite, and muscovite analyses for sample JKR-88-4 which yield the first 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) El. Wt. % Garnet Biotite Muscovite Si 17.072 16.846 21.552 Al 11.304 9.714 18.834 Ti b.d. 0.975 0.200 Fe 27.784 14.830 0.593 Mg 1.580 6.004 0.284 Mn 1.570 b.d. b.d. Cr n.a. b.d. b.d. Ca 1.557 b.d. b.d. Ba n.a. b.d. 0.347 K n.a. 7.604 6.743 Na n.a. 0.156 1.618 F n.a. 0.178 0.234 Cl n.a. b.d. b.d. 0 39.583 38.234 43.690 Total 100.450 94.541 94.095 Table D.42: Selected garnet, biotite, and muscovite analyses for sample JKR-88-4 which yield the second 'typical' P and T given in table 5.2 in chapter 5. (n.a. = not analyzed, b.d = below detection limit given in table D.l) 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share